american fisheries society · status, distribution, and conservation of native freshwater fishes of...

This PDF is for your personal use only, and may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to others is expressly prohibited.

Status, Distribution, andConservation of Native Freshwater Fishes of Western North America:

A Symposium Proceedings

frtmatterWNAF1 5/25/07 1:43 PM Page i

Status, Distribution, andConservation of Native Freshwater Fishes of Western North America:

A Symposium Proceedings

Edited by

Mark J. Brouder

U.S. Fish and Wildlife ServiceAshland Fishery Resources Office

Ashland, Wisconsin 54806, USA

and

Julie A. Scheurer

National Marine Fisheries ServiceAlaska Region

Juneau, Alaska 99802, USA

American Fisheries Society Symposium 53

Proceedings of the Symposium

“Status, Distribution, and Conservation of Native Freshwater

Fishes of Western North America”

Held in Salt Lake City, Utah, USA

1–3 March 2004

American Fisheries Society

Bethesda, Maryland, USA

2007

frtmatterWNAF1 5/25/07 1:43 PM Page iii

Suggested citation formats are

Entire book

Brouder, M. J., and J. A. Scheurer, editors. 2007. Status, distribution, and conservation ofnative freshwater fishes of western North America: a symposium proceedings. Amer-ican Fisheries Society, Symposium 53, Bethesda, Maryland.

Chapter in book

Gill, C. J., K. R. Gelwicks, and R. M. Keith. 2007. Current distribution of blueheadsucker, flannemouth sucker, and roundtail chub in seven subdrainages of theGreen River, Wyoming. Pages 115–122 in M. J. Brouder and J. A. Scheurer, editors.Status, distribution, and conservation of native freshwater fishes of westernNorth America: a symposium proceedings. American Fisheries Society, Sympo-sium 53, Bethesda, Maryland.

© Copyright 2007 by the American Fisheries Society

All rights reserved. Photocopying for internal or personal use, or for the internal or per-sonal use of specific clients, is permitted by the American Fisheries Society (AFS) provid-ed that the appropriate fee is paid directly to Copyright Clearance Center (CCC), 222Rosewood Drive, Danvers, Massachusetts 01923, USA; phone: 978-750-8400. Requestauthorization to make multiple copies for classroom use from CCC. These permissionsdo not extend to electronic distribution or long-term storage of multiple articles or tocopying for resale, promotion, advertising, general distribution, or creation of new col-lective works. For such uses, permission or license must be obtained from AFS.

Printed in the United States of America on acid-free paper.

Library of Congress Control Number 2007923773

ISBN 978-1-888569-89-6ISSN 0892-2284

American Fisheries Society Web site: www.fisheries.org

American Fisheries Society5410 Grosvenor Lane, Suite 110Bethesda, Maryland 20814-2199

USA

frtmatterWNAF1 5/25/07 1:43 PM Page iv

v

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Symbols and Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi

PART 1—STATUS

Great Plains Fishes Declining or Threatened with Extirpation in Montana, Wyoming, or Colorado . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Wayne A. Hubert and Kelly M. Gordon

Population Abundance Estimates for Humpback Chub and Roundtail Chub in Westwater Canyon, Colorado River, Utah, 1998–2000 . . . . . . . . . . . . . . . 15J. Michael Hudson and Julie A. Jackson

Splittail “Delisting”: A Review of Recent Population Trends and Restoration Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25Ted R. Sommer, Randall D. Baxter, and Fredrick Feyrer

Status of June Sucker in Utah Lake and Refuges . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39Matthew E. Andersen, Christopher J. Keleher, Joshua E. Rasmussen, Eriek S. Hansen,

Paul D. Thompson, David W. Speas, M. Douglas Routledge, and Trina N. Hedrick

Status of Native Hawaiian Stream Fishes, a Unique Amphidromous Biota . . . . . . . 59Robert T. Nishimoto and J. Michael Fitzsimons

Biological Status of Leatherside Chub: A Framework for Conservation of Western Freshwater Fishes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67Mark C. Belk and Jerald B. Johnson

Population Status and Trends of Four Bear Lake Endemic Fishes . . . . . . . . . . . . . 77Scott A. Tolentino

The Status of Desert Redband Trout in Southwestern Idaho . . . . . . . . . . . . . . . . . 85Donald W. Johnson and Katie Fite

Improved Status of the Endangered Oregon Chub in the Willamette River, Oregon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91Paul D. Scheerer

Roundtail Chub Population Assessment in the Lower Salt and Verde Rivers, Arizona . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103Scott D. Bryan and Matthew W. Hyatt

Contents

frtmatterWNAF1 5/25/07 1:43 PM Page v

vi

PART 2—DISTRIBUTION

Columbia River Fishes of the Lewis and Clark Expedition . . . . . . . . . . . . . . . . . . 113Dennis Dauble

Current Distribution of Bluehead Sucker, Flannelmouth Sucker, and Roundtail Chub in Seven Subdrainages of the Green River, Wyoming . . . . . . . . 121Curtis J. Gill, Kevin R. Gelwicks, and Robert M. Keith

A Review of the Distribution and Management of Bonytail in the Lower Colorado River Basin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129Charles O. Minckley and Mitch S. Thorson

Distribution and Habitat Use of Cottids in the Lake Washington Basin . . . . . . 135Roger A. Tabor, Kurt L. Fresh, Dwayne K. Paige, Eric J. Warner, and Roger J. Peters

Pacific Lamprey Ammocoete Habitat Utilzation in Red River, Idaho . . . . . . . . . 151Christopher W. Claire, Timothy G. Cochnauer, and George W. LaBar

PART 3—CONSERVATION/MANAGEMENT

What the Status of Utah Chub Tells Us about Conserving Common,Widespread Species . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165Jerald B. Johnson and Mark C. Belk

Native Southwestern Trouts: Conservation with Reference to Physiography, Hydrology, Distribution, and Trends . . . . . . . . . . . . . . . . . . . . . . . 175John N. Rinne and Bob Calamusso

Simulation of Human Effects on Bull Trout Population Dynamics in Rimrock Reservoir, Washington . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191Keith Underwood and Steve P. Cramer

CONTENTS

frtmatterWNAF1 5/25/07 1:43 PM Page vi

viivii

The organizing committee for this sympo-sium was led by Mark Brouder and includ-ed Lynn Starnes, Julie Scheurer, Lori Mar-

tin, Rob Nielsen, and Jim Tilmant. Manyadditional people helped and we thank each fortheir assistance, particularly David Zafft, PaulMarsh, Don MacDonald, Leanne Roulson, SteveWolff, and all members of the Wyoming–Col-orado and Bonneville Chapters’ Meeting Commit-tee who helped with various logistics and meetingarrangements. We also express our sincere thanks

to our anonymous peer reviewers. We thank theAssociation of Fish and Wildlife Agencies’ Multi-State Conservation Grant Program, the NationalFish and Wildlife Foundation’s Bring Back theNatives Grant Program, the David and LucilePackard Foundation, the American Fisheries Soci-ety’s Fisheries Conservation Foundation, and theNational Park Service for providing financial sup-port. We also thank the American Fisheries Soci-ety for their encouragement and cooperation inpreparing this publication.

Acknowledgments

frtmatterWNAF1 5/25/07 1:43 PM Page vii

ix

James T. TilmantChair, Western Native Fishes CommitteeWestern Division of the American Fisheries Society———————————————————

As land development, population growth,

and watershed alterations have continued

throughout the western United States,

Canada, and northern Mexico during the past cen-

tury, the status of many of our western native

freshwater fish species has become questionable.

This is particularly true with regards to non-game

species and those without recognized economic

value. Native freshwater fishes have been exten-

sively impacted from land and watershed develop-

ment, habitat alteration and loss, direct human

harvest, and increased competition from intro-

duced species. As population growth within the

western region continues, a solid understanding of

where remaining populations of our native fish

fauna occur, habitats that are important to their

life histories, and the threats to the continued exis-

tence of these populations is critical for conserva-

tion and restoration. The need to address these

issues led to the formation of the Western Native

Fishes Committee of the Western Division, Amer-

ican Fisheries Society.

The Western Native Fishes Committee con-

vened at the 2003 annual meeting in San Diego,

California. At that meeting, we discussed the Fish-

eries publications by Warren et al. (2000) on the

diversity, distribution, and conservation status of

native freshwater fishes of the southern United

States and by Musick et al. (2000) on North Amer-

ican marine stocks at risk. The committee recog-

nized the need for a similar synthesis for freshwater

fishes of the area encompassed by the Western

Division. Hence, we organized a symposium for

the 2004 Western Division annual meeting in Salt

Lake City, Utah to begin to gather information. The

primary objective of the symposium was to bring

forth previously unpublished information on the

status and distribution of western native freshwater

fish species. The symposium brought together aca-

demic, federal, state, and local scientists and

resource managers. There were more than 30 pre-

sentations at the symposium and manuscripts were

prepared for most of them to be included in these

published proceedings.

These proceedings contribute to the larger

effort by the Western Division Native Fishes Com-

mittee to review and document the conservation

status of all native freshwater fishes within western

North America. The committee is in the process of

creating an accessible database with historical and

current distribution maps for each species as well

as up-to-date information about status, habitat

requirements, and known causes of decline.

The extent to which this symposium provided

new and updated information on many of our

lesser known western native freshwater fish species

is gratifying. Papers included address many species

of special concern within Wyoming, Colorado,

Utah, New Mexico, California, Arizona, Oregon,

Idaho, Hawaii, British Columbia, and Sonora,

Mexico and include multiple species overviews

within several major western watersheds, includ-

ing the Colorado River basin, Gila River, Green

River drainage, Wind River drainage, lower Sacra-

mento River, Virgin River, Yampa River, Verde

River, Willamette River, Columbia River basin, and

the Rio Conchos and Rio Bravo basins of Mexico.

Introduction

frtmatterWNAF1 5/25/07 1:43 PM Page ix

Much of this information is extremely helpful to

the Western Division’s ongoing effort to determine

and document the current status of our western

native freshwater fishes. These papers should also

be immensely helpful to state and federal fisheries

managers with responsibilities to insure the protec-

tion and continued existence of non-game native

fish species within the areas of their jurisdiction.

The Western Division thanks the contributors to

these Proceedings and greatly appreciates the effort

they made to compile and report the new informa-

tion provided.

References

Warren, M. L., Burr, B. M., Walsh, S. J., Bart, H. L., Cashner,R. C., Etnier, D. A., Freeman, B. J., Kuhajda, B. R., May-den, R. L., Robison, H. W., Ross, S. T., and W. C. Starnes.2000. Diversity, distribution, and conservation status ofthe native freshwater fishes of the southern UnitedStates. Fisheries 25(10):7–31.

Musick, J. A., Harbin, M. M., Berkeley, S. A., Burgess, G. H.,Eklund, A. M., Findley, L., Gilmore, R. G., Golden, J. T.,Ha, D. S., Huntsman, G. R., McGovern, J. C., Parker, S. J.,Poss, S. G., Sala, E., Schmidt, T. W., Sedberry, G. R., Weeks,H., and S. G. Wright. 2000. Marine, estuarine, and diadro-mous fish stocks at risk of extinction in North America(exclusive of Pacific salmonids). Fisheries 25(11):6–30.

x

INTRODUCTION

frtmatterWNAF1 5/25/07 1:43 PM Page x

xi

A ampere

AC alternating current

Bq becquerel

C coulomb

°C degrees Celsius

cal calorie

cd candela

cm centimeter

Co. Company

Corp. Corporation

cov covariance

DC direct current; District of Columbia

D dextro (as a prefix)

d day

d dextrorotatory

df degrees of freedom

dL deciliter

E east

E expected value

e base of natural logarithm (2.71828…)

e.g. (exempli gratia) for example

eq equivalent

et al. (et alii) and others

etc. et cetera

eV electron volt

F filial generation; Farad

°F degrees Fahrenheit

fc footcandle (0.0929 lx)

ft foot (30.5 cm)

ft 3/s cubic feet per second (0.0283 m3/s)

g gram

G giga (109, as a prefix)

gal gallon (3.79 L)

Gy gray

h hour

ha hectare (2.47 acres)

hp horsepower (746 W)

Hz hertz

in inch (2.54 cm)

Inc. Incorporated

i.e. (id est) that is

IU international unit

J joule

K Kelvin (degrees above absolute zero)

k kilo (103, as a prefix)

kg kilogram

km kilometer

l levorotatory

L levo (as a prefix)

L liter (0.264 gal, 1.06 qt)

lb pound (0.454 kg, 454g)

lm lumen

log logarithm

Ltd. Limited

M mega (106, as a prefix);

molar (as a suffix or by itself)

m meter (as a suffix or by itself);

milli (10-3, as a prefix)

mi mile (1.61 km)

min minute

mol mole

N normal (for chemistry);

north (for geography); newton

N sample size

NS not significant

n ploidy; nanno (10-9, as a prefix)

o ortho (as a chemical prefix)

oz ounce (28.4 g)

P probability

p para (as a chemical prefix)

p pico (10-12, as a prefix)

Pa pascal

pH negative log of hydrogen ion activity

ppm parts per million

Symbols and Abbreviations

The following symbols and abbreviations may be found in this book without definition. Also undefined arestandard mathematical and statistical symbols given in most dictionaries.

frtmatterWNAF1 5/25/07 1:43 PM Page xi

xii

qt quart (0.946 L)

R multiple correlation or regression coeffi-

cient

r simple correlation or regression coefficient

rad radian

S siemens (for electrical conductance);

south (for geography)

SD standard deviation

SE standard error

s second

T tesla

tris tris(hydroxymethyl)-aminomethane (a

buffer)

UK United Kingdom

U.S. United States (adjective)

USA United States of America (noun)

V volt

V, Var variance (population)

var variance (sample)

W watt (for power); west (for geography)

Wb weber

yd yard (0.914 m, 91.4 cm)

α probability of type I error (false rejection of

null hypothesis)

β probability of type II error (false acceptance

of null hypothesis)

Ω ohm

μ micro (10-6, as a prefix)

' minute (angular)

" second (angular)

° degree (temperature as a prefix, angular as a

suffix)

% per cent (per hundred)

per mille (per thousand)0/00

SYMBOLS AND ABBREVIATIONS

frtmatterWNAF1 5/25/07 1:43 PM Page xii

3

The Great Plains cover about 20% of the landarea of the 48 contiguous United States. TheGreat Plains slope gently eastward from the

foothills of the Rocky Mountains to roughly the457-m (1,500-ft) elevation contour line, the 100thmeridian of longitude, and the 51-cm (20-in) peryear isohyet (i.e., a line on a map connecting pointshaving equal precipitation). The climate is semi-arid with extreme seasonal variability in precipita-tion and air temperature (Kraenzel 1955).

Native fish assemblages in the Great Plains atthe time of settlement by Europeans were a resultof past climatic and geologic events. The area wasa shallow sea during the Cretaceous Period, result-ing in nearly planar sediment deposits (Cross et al.1986; Cross and Moss 1987). Uplift and erosion ofthe Rocky Mountains filled much of the regionwith alluvium leading to stream channels domi-nated by silt and sand (Cross and Moss 1987). Dif-fering fish assemblages evolved in three water-sheds across the Great Plains region, but the

connections of these watersheds into the Missis-sippi–Missouri drainage after the PleistoceneEpoch obscured preglacial faunal distinctions(Cross et al. 1986; Robison 1986). Much of thenatural diversity of the Great Plains is because ofthe continuity established with the MississippiRiver drainage (Burr and Mayden 1992), whichserved as a refuge during glaciation.

Drainage patterns from the Pleistocene Epochto present affect distributions of Great Plains fish-es. The principal drainages are the Missouri, Platte,and Arkansas rivers, which flow west to east (Fig-ure 1). North-south migration is prevented in thewestern portion of these watersheds by a lack ofnorth-south connectivity (Matthews and Zimmer-man 1990; Rahel et al. 1996), so faunal similaritiesamong these drainages decrease east to west (Crosset al. 1986). Also, increasingly harsh environmentalconditions occur with westward progressiontoward the rain shadow of the Rocky Mountains(Cross 1970; Cross et al. 1986; Cross and Moss

Great Plains Fishes Declining or Threatened withExtirpation in Montana, Wyoming, or Colorado

ABSTRACT Of 55 fish species that were likely native to the Great Plains region of Montana, Wyoming, and Col-

orado, 33 (60%) species have been given some kind of conservation designation by at least one of the state fish-

eries management agencies because they are rare or in decline. The species with conservation designations were

generally fishes that inhabit large rivers of the upper Missouri River drainage (14 species); live in small, cool- or

clear-water streams (9 species); or reside in a wide array of habitats but occur at the edge of their ranges in indi-

vidual states (10 species). Changes in riverine habitats due to construction of reservoirs on large rivers and intro-

duction of exotic piscivorous fishes to reservoirs are major causes of decline of riverine species in the Great Plains

region of Montana, Wyoming, and Colorado. Fishes that occur as disjunct, relict populations in small cool- or clear-

water streams or at the periphery of their range in individual states are susceptible to local extirpations caused by

habitat alterations and introductions of exotic piscivorous fishes but may have significant conservation value due to

their genetic diversity. Given the large proportion of native Great Plains fishes that are declining or threatened with

extirpation in individual states, a region-wide effort to maintain native fish assemblages is warranted.

Wayne A. Hubert and Kelly M. Gordon

American Fisheries Society Symposium 53:3–13

© 2007 by the American Fisheries Society

WAYNE A. HUBERT and KELLY M. GORDON U.S. Geological Survey, Wyoming Cooperative Fish and Wildlife Research Unit, University of Wyoming,Department 3166, 1000 East University Avenue, Laramie, Wyoming 82071-3166, USA

NatAmerFishWAPart1-9 5/25/07 1:44 PM Page 3

1987). Consequently, western Great Plains nativefish faunas are subsets of eastern faunas of GreatPlains watersheds.

As the Pleistocene Epoch climate warmed andglaciers retreated, fishes in refugia of the OzarkHighlands dispersed onto the Great Plains. Later, asthe climate warmed further, much of the GreatPlains became inhospitable to many of these fishspecies. However, some relict populations survivedin cool, clear, spring-fed or foothills streams nearthe western edge of the Great Plains (Stevenson etal. 1974; Hawkes et al. 1986).

Wide variations in discharge, turbidity, andwater temperature appear to structure fish assem-blages in Great Plains streams (Ross et al. 1985).Many streams undergo large seasonal fluctuationsin discharge because of flooding due to snowmeltin spring and desiccation due to heating and evap-oration in summer (Matthews 1988). Fish assem-blages in these streams tend to be depauperatecompared to streams with more stable flows (Crossand Moss 1987; Matthews 1987; Bramblett andFausch 1991; Fausch and Bramblett 1991). Howev-er, the species that occur are adapted to the wideranges of physicochemical conditions that occurannually (Matthews 1988; Resh et al. 1988; Dieter-man and Galat 2004).

Human activities have impacted habitats ofnative Great Plains fishes. In general, large riversthat were historically turbid are now clearer, andsmaller streams that were typically clear are nowmore turbid. Dams and reservoirs have altered theflows and affected the fish assemblages of almostevery sizable stream on the Great Plains (Dieter-man and Galat 2004). Reservoirs generally trapsediment, reduce fluctuations in flow, and lowersummer water temperatures. These activitiesreduce turbidity and enhance the ability of sight-feeding piscivores, formerly rare or introduced, toprey on native fishes (Sanders et al. 1993; Rabeni1996; Pigg et al. 1999). Additionally, timing andsuccess of reproduction, population dynamics, andabundance of prey can be altered (Eberle et al.1993; Hesse et al. 1993; Echelle et al. 1995; Wolf et al.1995). Downstream rivers often change from wide,shallow, and braided to narrow, deep, and singlechannels with few side channels or backwaters. Thereduced habitat complexity affects many native fish

species (Hesse et al. 1982; Patton and Hubert1993). Dams on tributaries and small streams canprevent access to spawning and nursery areas usedby Great Plains fishes (Smith 1988; Braaten 1993)and fragment populations to below the minimumnumber needed for viability (Sheldon 1987, 1988).

Agriculture has had multiple effects on GreatPlains streams, including increased turbidity, silta-tion, and nutrients through soil erosion and irriga-tion return flows (Menzel et al. 1984). Erosion andsediment deposition alter substrate characteristics,leading to a decrease in species that require cleangravel for spawning (Berkman and Rabeni 1987).Fishes that are adapted to clear, stable flow havedeclined as water tables have decreased and springshave dried up (Owen et al. 1981; Cross et al. 1985;Schmulbach and Braaten 1993).

Isolation of fishes in upstream reaches ofstreams has resulted from reservoir construction.Fishes that cope with drying or intermittency bymoving downstream may be forced to enter areservoir where they are exposed to a higher risk ofpredation by piscivorous fishes (Winston et al.1991). Alternatively, introduced reservoir speciesmay use upstream segments, increasing predationon native fishes (Cross et al. 1986; Sheldon 1987;Winston et al. 1991; Lynch and Roh 1996).

Introductions of exotic fishes into reservoirs onthe Great Plains, many perceived as desirable gamefishes, have contributed to declines of native fishes(Cross et al. 1986). Numerous exotic fish specieshave been introduced in the western Mississip-pi–Missouri basin (Cross et al. 1986; Rabeni 1996).Reservoirs have increased the proportion of lentichabitat and reduced turbidity and variability ofdischarge, factors conducive to naturalization andrange expansions of many introduced species(Cross et al. 1986; Sanders et al. 1993; Rabeni 1996;Gido and Brown 1999).

Our purpose was to assess the status of nativeGreat Plains fishes in Montana, Wyoming, andColorado. We summarized information on theconservation status of native Great Plains fishes inthis three-state region and assessed the commonfeatures of fishes that have received such designa-tions. Specifically, we were interested in identifyingthe fishes with conservation status that occurred asresidents of larger rivers, as disjunct, relict popula-

4

HUBERT AND GORDON


tions in small, cool- or clear-water streams near thewestern edge of the Great Plains, or as wide rang-ing species across the Great Plains but on theperiphery of their natural range in Montana,Wyoming, or Colorado.

Methods

We used the conservation status of native fishspecies designated by state fish management agen-cies to identify fishes that are likely to be experienc-ing declines in the Great Plains regions of Mon-tana, Wyoming, and Colorado. We consulted Websites and documents produced by agencies in thethree states to identify the conservation status offishes that occur naturally in the Mississippi Riverwatershed of each state (Figure 1).

We classified each of the species with a desig-nated conservation status as (1) large fishes thatoccur in rivers, (2) small fishes that occur in rivers,(3) fishes found in small, cool- or clear-waterstreams, or (4) widespread fishes found in an arrayof habitats. These judgments were made based on

the habitat associations described for each speciesin general references such as Scott and Crossman(1973), Baxter and Stone (1995), and Pflieger(1997). Additionally, we classified the distribu-tions of each of these species as they probablyoccurred just prior to settlement by Europeans:(1) on the periphery of a contiguous range acrossthe Great Plains, (2) disjunct, relict populationsseparated by the Great Plains from core popula-tions to the east or north, or (3) widespread distri-butions in the state(s) where they have conserva-tion status. These judgments were made usingdistribution maps presented in Scott and Cross-man (1973) and Lee et al. (1980).

We identified exotic fishes not found naturallyin the Mississippi River watershed of Montana,Wyoming, or Colorado that have been introducedto the watershed within the three-state area andfishes native to a portion of the Great Plains regionof the three states whose ranges have been pur-posefully extended within the Great Plains regionof the three states. Reasons for introductions orrange extensions were identified along with thefishes that were highly piscivorous. Informationwas obtained from Web sites produced by the stateagencies, Brown (1971), Lee et al. (1980), and Bax-ter and Stone (1995).

Results and Discussion

Fifty-five fish species were identified as native tothe Great Plains region of Montana, Wyoming, andColorado (Table 1). Special conservation status iscurrently conferred on 60% (33 species) of theGreat Plains species by at least one of the stateagencies (Table 1). Additionally, six more species inMontana are identified as possibly in peril, buttheir status is unknown and additional informa-tion is needed before they are designated withsome level of conservation status (Table 1).

Of the native Great Plains fish species, 43% (16of 33 species) in Montana, 24% (9 of 37) inWyoming, and 41% (14 of 34 species) in Coloradohave a conservation status of some kind. Amongthe 33 species having a conservation status in atleast one state, we identified (1) nine species oflarge, riverine fishes, (2) five species of small, river-ine fishes, (3) nine species of small, cool- or clear-

5

GREAT PLAINS FISHES

Figure 1. The focus of this paper was the Great Plains fishesin Montana, Wyoming, and Colorado. Boundaries of the studyarea are highlighted and major river systems are shown.


6

HUBERT AND GORDON

Occurrence and conservation status of native Great Plains fishes in Montana, Wyoming, and Colorado. An X identifies speciesthat occur in a state but do not have a designated conservation status by the state fish management agency. State designationsof conservation status are (1) Montana: S1 = critically imperiled because of extreme rarity or because of some factor(s) of itsbiology making it especially vulnerable to extinction; S2 = imperiled because of rarity or because of other factor(s) demonstrablymaking it very vulnerable to extinction throughout its range; S3 = either very rare and local throughout its range or found locally(even abundantly at some of its locations) in a restricted range or vulnerable to extinction throughout its range because of otherfactor(s); S4 = apparently secure, though it may be quite rare in parts of its range, especially at the periphery; S5 = demonstra-bly secure, though it may be quite rare in parts of its range, especially at the periphery; and SU = possibly in peril, but statusuncertain, more information needed; (2) Wyoming: NSS1 = populations are physically isolated and/or exist at extremely low den-sities throughout the range, habitats are declining or vulnerable and extirpation appears possible; NSS2 = populations are physi-cally isolated and/or exist at extremely low densities throughout range, but their habitat conditions appear stable; and (3) Col-orado: SE = state endangered, ST = state threatened, and SC = state special concern.

State

Common name Scientific name Montana Wyoming Colorado

Pallid sturgeon Scaphirhynchus albus S1

Shovelnose sturgeon S. platorynchus S4 X

Paddlefish Polyodon spathula S1S2

Shortnose gar Lepisosteus platostomus S1

Gizzard shad Dorosoma cepedianum X X

Goldeye Hiodon alosoides X NSS2

Northern pike Esox lucius X

Central stoneroller Campostoma anomalum X X

Lake chub Couesius plumbeus X X SE

Western silvery minnow Hybognathus argyritis S4S5 NSS1

Brassy minnow H. hankinsoni SU X ST

Plains minnow H. placitus SU X SE

Sicklefin chub Hybopsis meeki S1

Sturgeon chub Macrhybopsis gelida S2 NSS1

Flathead chub Platygobio gracilis X X SC

Hornyhead chub Nocomis biguttatus NSS1 X

Red shiner Cyprinella lutrensis X X

Common shiner Luxilus cornutus X ST

Emerald shiner Notropis atherinoides X

Bigmouth shiner N. dorsalis X X

Sand shiner N. stramineus S4 X X

Suckermouth minnow Phenacobius mirabilis NSS1 SE

Northern redbelly dace Phoxinus eos SU SE

Southern redbelly dace P. erythrogaster SE

Finescale dace P. neogaeus NSS1

Fathead minnow Pimephales promelas S4S5 X X

Longnose dace Rhinichthys cataractae X X X

Creek chub Semotilus atromaculatus SU X X

Pearl dace Margariscus margarita S2 NSS1

River carpsucker Carpiodes carpio X X X

Quillback C. cyprinus X

Longnose sucker Catostomus catostomus X X X

White sucker C. commersonii X X X

Mountain sucker C. platyrhynchus X X SC

Blue sucker Cycleptus elongatus S2S3

Smallmouth buffalo Ictiobus bubalus X

Bigmouth buffalo I. cyprinellus S4S5

Shorthead redhorse Moxostoma macrolepidotum X X

Table 1.

(continued)


water stream fishes, and (4) 10 widespread speciesthat occur in an array of habitats (Table 2).

The nine species of large, riverine fishes withconservation status were all found in the upperMissouri River watershed. Eight species have con-servation status in Montana, and one species(goldeye) has conservation status in Wyoming.Among the nine species, seven were widespread inMontana prior to settlement by Europeans, butthe shortnose gar in Montana and the goldeye inWyoming were on the periphery of their naturalrange. Declines in distributions and the conserva-tion status of shortnose gar, pallid sturgeon, shov-elnose sturgeon, paddlefish, blue sucker, big-mouth buffalo, sauger, and freshwater drum inMontana are probably due to the construction ofreservoirs on large rivers in the upper MissouriRiver watershed and the widespread introductionof exotic, piscivorous fishes (Quist et al. 2005).Goldeye receive conservation status in Wyomingbecause of its occurrence in a single river (i.e.,Powder River) that is part of the upper MissouriRiver watershed. Among the nine large, riverinespecies, the pallid sturgeon is the only species list-ed under the Endangered Species Act, where it isclassified as endangered.

The five species of small, riverine fishes withconservation status were all found in the upperMissouri River or North Platte River watershedsprior to settlement by Europeans. The western sil-

very minnow, plains minnow, sicklefin chub, stur-geon chub, and flathead chub are often classified asturbid-river cyprinids and have declined widely inassociation with the construction of reservoirs(Pflieger and Grace 1987; Williams et al. 1989; Pat-ton et al. 1998; Dieterman and Galat 2004; Quist etal. 2004). None of these small, riverine fishes arelisted as threatened or endangered under theEndangered Species Act, but the sturgeon chub hasbeen petitioned for listing.

Seven of the nine species of small, cool- orclear-water stream fishes were identified as occur-ring as disjunct, relict populations in the state(s)where they have received conservation status. Thehornyhead chub, common shiner, northern red-belly dace, southern redbelly dace, finescale dace,pearl dace, and Arkansas darter probably occurredas one or a very few populations in Montana,Wyoming, or Colorado prior to settlement byEuropeans. All seven of these species are commonthroughout the core of their natural ranges to theeast in the United States or to the north in Cana-da. As a result, none of these species are listedunder the Endangered Species Act. However, thesmall distributions of disjunct, relict poplationsmake them vulnerable to extirpation from therefuges where they occurred at the time of settle-ment by Europeans. The primary causes of extir-pations are probably habitat alterations and intro-ductions of exotic piscivorous fishes (Schlosser

7

GREAT PLAINS FISHES

Black bullhead Ameiurus melas X X

Channel catfish Ictalurus punctatus X X X

Stonecat Noturus flavus X X SC

Flathead catfish Pylodictis olivaris X

Trout-perch Percopsis omiscomaycus S2

Burbot Lota lota SU X

Plains topminnow Fundulus sciadicus NSS2 SC

Plains killifish F. zebrinus S4 X X

Brook stickleback Culaea inconstans X

Green sunfish Lepomis cyanellus X

Arkansas darter Etheostoma cragini ST

Iowa darter E. exile SU X SC

Johnny darter E. nigrum X X

Orangethroat darter E. spectabile NSS2 SC

Logperch Percina caprodes X

Sauger Sander canadense S2 X

Freshwater drum Aplodinotus grunniens S4

Table 1. (continued)


1988; Winston et al. 1991; Lohr and Fausch 1996;Quist et al. 2003).

Most of the 10 widespread species that occurin a broad range of habitats are listed in only oneof the three states. These species occurred withinthe state’s boundaries at the time of settlement byEuropeans, but at the periphery of the species’range. Lake chub, brassy minnow, sand shiner,suckermouth minnow, stonecat, trout-perch,plains killifish, Iowa darter, and orangethroatdarter are widespread to the east and north in theGreat Plains but have conservation status in Mon-tana, Wyoming, or Colorado, largely due to small,restricted distributions in a state. These specieshave a high probability of extirpation from a statein association with habitat alterations or exoticspecies introductions in the watersheds wherethey are found.

It is likely that designated conservation statusby state agencies does not identify the full array ofspecies that are declining in the western Great

Plains. Patton et al. (1998) identified 16 of the 37native species of Great Plains fishes in Wyoming ashaving decreased distributions from the 1960s tothe 1990s, but only five of these species have con-servation status in Wyoming (i.e., western silveryminnow, sturgeon chub, hornyhead chub,finescale dace, and plains topminnow). Thedeclining fishes included three species of small,riverine fishes (i.e., western silvery minnow, plainsminnow, and flathead chub), four species of cool-or clear-water stream fishes (i.e., plains topmin-now, hornyhead chub, common shiner, finescaledace, and mountain sucker), and two widespreadspecies (i.e., lake chub and stonecat) that havebeen given conservation status in Montana,Wyoming, or Colorado. However, six species thatare generally considered to be widespread habitatgeneralists were also found to be decreasing (i.e.,river carpsucker, channel catfish, black bullhead,creek chub, longnose sucker, and shorthead red-horse [also known as northern redhorse]). Among

8

HUBERT AND GORDON

Table 2.

Habitat classification and probable distribution patterns prior to settlement by Europeans of 33 native fish species currently list-ed as having special conservation status by state fish management agencies in Montana, Wyoming, or Colorado.

Habitat classification

Distribution class Large, riverine Small, riverine Cool, clear stream Wide array

Disjunct, relict Hornyhead chubCommon shinerNorthern redbelly daceSouthern redbelly daceFinescale dacePearl daceArkansas darter

Periphery of range Shortnose gar Western silvery minnow Plains topminnow Lake chub Goldeye Plains minnow Brassy minnow

Sicklefin chub Sand shinerSuckermouth minnowStonecatTrout perchPlains killifishIowa darterOrangethroat darter

Widespread, contiguous Pallid sturgeon Sturgeon chub Mountain sucker Fathead minnowShovelnose sturgeon Flathead chubPaddlefishBlue suckerBigmouth buffaloSaugerFreshwater drum


the declining fishes in Wyoming, Patton et al.(1998) classified five species as being adapted toturbid rivers with silt and sand substrates (i.e.,flathead chub, plains minnow, western silveryminnow, river carpsucker, and channel catfish)and suggested that they have been impacted byreservoirs and diversion dams that have stabilizedflows and reduced silt loads. They classified fiveother species as tending to occupy small- to medi-um-sized streams having cool, clear water withsilt-free rocky substrates (i.e., common shiner,finescale dace, hornyhead chub, lake chub, andmountain sucker) and suggested that land man-agement and irrigation practices that increase tur-bidity and siltation are impacting these species.Prior to the 1960s, the distributions of severallarge, riverine fishes in Wyoming were alreadyreduced in association with reservoir constructionon large rivers (i.e., shovelnose sturgeon, goldeye,and sauger), so they were not identified as declin-ing by Patton et al. (1998).

When we reviewed records of fish species intro-ductions and distributions for Montana, Wyoming,and Colorado, we identified 23 species and threehybrids produced in hatcheries that have been wide-ly introduced to the Great Plains regions of the threestates (Table 3). Additionally, 11 species with limit-ed natural distributions within the Great Plainsregions of the three states have been widely distrib-uted since settlement by Europeans (Table 3). Theprimary reason for these introductions and rangeextensions has been for the creation and manage-ment of sport fisheries, primarily in reservoirs.More than half of these fishes are piscivorous andmay prey on native species.

It is interesting to note the features of 22 nativespecies found within the Great Plains region of thethree states that have not been given conservation sta-tus in any of the three states (Table 4). Most of themtend to be widespread species that occur across a sub-stantial portion of the Great Plains. Additionally, sixspecies are managed as sport fishes, two species aremanaged as prey species in reservoirs, and four speciestend to be highly invasive (Table 4). Nevertheless, Pat-ton et al. (1998) identified two of the sport fishesnative to Wyoming (i.e., black bullhead and channelcatfish) as showing declines in distributions inWyoming between the 1960s and 1990s.

Most of the species that have received special con-servation status in Montana, Wyoming, or Coloradooccurred as disjunct, relict populations or were onthe periphery of their range at the time of settlementby Europeans. These populations of fishes in stateswhere they have conservation status are likely to havesubstantial conservation value despite the fact thatthe species are globally secure (Lesica and Allendorf1995). Many species at the edges of their ranges occurin unusual or atypical habitats. Separation from corepopulations and occurrence in unusual habitats canfacilitate genetic drift and natural selection resultingin genetically distinct populations and potential forfuture speciation. Genetically divergent populationshave greater conservation value than populations atthe core of a species range, and this may justify theuse of limited conservation resources to protect them(Lesica and Allendorf 1995).

The observations that 60% of the native GreatPlains fish species in Montana, Wyoming, and Col-orado have received a special conservation statusby state agencies and that several additional speciesappear to be declining (Patton et al. 1998) suggestthat native fish assemblages have changed substan-tially since settlement by Europeans and that man-agement is needed to preserve native fishes in theGreat Plains areas of these states. A region-wide,large-scale effort is needed to define better theGreat Plains species in decline, their habitatrequirements, and their susceptibility to predationand competition by introduced fishes. Additional-ly, the locations of remaining populations of rareor declining species need to be catalogued so thatmanagement can focus on preservation of habitatand fish assemblages conducive to their survivalinto the future. Species that have severely reduceddistributions in the region are likely to requirecooperative efforts by state and federal agencies,private land owners, and water users to reducehuman impacts and assure the survival of thesefishes in the western Great Plains.

ReferencesBaxter, G. T., and M. D. Stone. 1995. Fishes of Wyoming.

Wyoming Game and Fish Department, Cheyenne.Berkman, H. E., and C. F. Rabeni. 1987. Effect of siltation on

stream fish communities. Environmental Biology of Fish-es 18:285–294.

9

GREAT PLAINS FISHES


10

HUBERT AND GORDON

Table 3.

Exotic (E) fishes not found naturally in the Mississippi River watershed of Montana, Wyoming, or Colorado that have been intro-duced to the watershed within the three-state region and fishes native (N) to a portion of the Great Plains region of the three stateswhose ranges have been purposefully extended within the Great Plains region of the three states. Reasons for introductions orrange extensions included development of sport fishes in reservoirs (1), small impoundments (2), tailwaters downstream fromreservoirs (3), small streams (4), or rivers (5); enhancement of prey fishes to reservoirs (6); enhancement of subsistence or com-mercial fisheries (7); biological control of nuisance organisms (8); and other factors (9). Piscivorous fishes are identified (X)

Exotic (E) orCommon name Scientific name Native (N) Primary reasons Piscivorous

Gizzard shad Dorosoma cepedianum N 6

Lake whitefish Coregonus clupeaformis N 6, 7

Coho salmon Oncorhynchus kisutch E 1 X

Kokanee O. nerka E 1, 6

Cutthroat trout O. clarkii N 1, 2, 3 X

Rainbow trout O. mykiss E 1, 2, 3 X

Brown trout Salmo trutta E 1, 3 X

Brook trout Salvelinus fontinalis E 4 X

Lake trout S. namaycush E 1 X

Splake S. namaycush x S. fontinalis E 1, 2 X

Arctic grayling Thymallus arcticus E 2

Northern pike Esox lucius N 1 X

Tiger musky E. lucius x muskellunge E. masquinongy E 1,2 X

Goldfish Carassius auratus E 9

Grass carp Ctenopharyngodon idella E 8

Common carp Cyprinus carpio E 3, 4, 7

Golden shiner Notemigonus crysoleucas E 9

Emerald shiner Notropis atherinoides E 6

Spottail shiner N. hudsonius N 6

Channel catfish Ictalurus punctatus N 1, 2 X

Flathead catfish Pylodictis olivaris N 1, 5 X

Western mosquitofish Gambusia affinis E 8

Brook stickleback Culaea inconstans E 9

White bass Morone chrysops E 1 X

Striped bass M. saxatilis E 1 X

Wiper M. chrysops x M. saxatilis E 1 X

Rock bass Ambloplites rupestris E 2, 4 X

Green sunfish Lepomis cyanellus N 2, 9 X

Pumpkinseed L. gibbosus E 2, 9

Bluegill L. macrochirus E 2

Smallmouth bass Micropterus dolomieu E 1, 2, 4 X

Largemouth bass M. salmoides E 1, 2 X

White crappie Pomoxis annularis E 1, 2 X

Black crappie P. nigromaculatus E 1,2 X

Yellow perch Perca flavescens N 1, 2, 9 X

Walleye Sander vitreus N 1, 2 X

Freshwater drum Aplodinotus grunniens E 9


Braaten, P. J. 1993. The influence of habitat structure andenvironmental variability on habitat use by fish in theVermillion River, South Dakota. Master’s thesis. Univer-sity of South Dakota, Vermillion.

Bramblett, R. G., and K. D. Fausch. 1991. Fishes, macroinver-tebrates, and aquatic habitats of the Purgatoire River inPinon Canyon, Colorado. Southwestern Naturalist36:281–294.

Brown, C. J. D. 1971. Fishes of Montana. Big Sky Books,Montana State University, Bozeman.

Burr, B. M., and R. L. Mayden. 1992. Phylogenetics andNorth American freshwater fishes. Pages 287–324 in R. L.Mayden, editor. Systematics, historical ecology, andNorth American freshwater fishes. Stanford UniversityPress, Palo Alto, California.

Cross, F. B. 1970. Fishes as indicators of Pleistocene andrecent environments in the Central Plains. Pages 241–257in W. Dort, Jr. and J. K. June, Jr., editors. Pleistocene andrecent environments of the central Great Plains. Univer-sity of Kansas Press, Lawrence.

Cross, F. B., R. L. Mayden, and J. D. Stewart. 1986. Fishes inthe western Mississippi basin (Missouri, Arkansas, andRed rivers). Pages 363–412 in C. H. Hocutt and E. O.Wiley, editors. The zoogeography of North Americanfreshwater fishes. Wiley, New York.

Cross, F. B., and R. E. Moss. 1987. Historic changes in fishcommunities and aquatic habitat in plains streams ofKansas. Pages 155–165 in W. J. Matthews and D. C.Heins, editors. Community and evolutionary ecology ofNorth American stream fishes. University of OklahomaPress, Norman.

Cross, F. B., R. E. Moss, and J. T. Collins. 1985. Assessment ofdewatering impacts on stream fisheries in the Arkansasand Cimarron rivers. University of Kansas, Museum ofNatural History, Lawrence.

Dieterman, D. J., and D. L. Galat. 2004. Large-scale factorsassociated with sicklefin chub distribution in the Mis-souri and Yellowstone rivers. Transactions of the Ameri-can Fisheries Society 133:577–587.

Eberle, M. E., G. W. Ernsting, J. R. Tomerelli, and S. L. Wells.1993. Assessment of restored streamflow on fish com-munities in the Arkansas River of southwestern Kansas.Transactions of the Kansas Academy of Science96:114–130.

Echelle, A. A., G. R. Luttrell, and R. D. Larson. 1995. Declineof native prairie fishes. Pages 303–305 in E. T. LaRoe, edi-tor. Our living resources: a report to the nation on thedistribution, abundance, and health of U.S. plants, ani-mals, and ecosystems. U.S. Department of the Interior,National Biological Service, Washington, D.C.

Fausch, K. D., and R. G. Bramblett. 1991. Disturbance andfish communities in intermittent tributaries of a westernGreat Plains river. Copeia 1991:659–674.

Gido, K. B., and J. H. Brown. 1999. Invasion of North Amer-ican drainages by alien fish species. Freshwater Biology42:387–399.

Hawkes, C. L., D. L. Miller, and W. G. Layher. 1986. Fishecoregions of Kansas: stream fish assemblage patternsand associated environmental correlates. EnvironmentalBiology of Fishes 17:267–279.

Hesse, L. W., G. E. Mestl, J. W. Robinson. 1993. Status ofselected fishes in the Missouri River in Nebraska withrecommendations for their recovery. Pages 327–340 in L.W. Hesse, C. B. Stalnaker, N. G. Benson, and J. R. Zuboy,editors. Proceedings of the symposium on restorationplanning for the rivers of the Mississippi River ecosys-tem. U.S. Department of the Interior, National Biologi-cal Survey, Washington, D.C.

Hesse, L. W., A. B. Schlesinger, G. L. Hergenrader, S. D. Reetz,and H. S. Lewis. 1982. The Missouri River study- ecolog-ical perspectives. Pages 185–223 in L. W. Hesse, editor.The middle Missouri River. Missouri River Study Group,Norfolk, Nebraska.

Kraenzel, C. F. 1955. The Great Plains in transition. Univer-sity of Oklahoma Press, Norman.

Lee, D. S., C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E.McAllister, J. R. Stauffer. 1980. Atlas of North Americanfreshwater fishes. North Carolina State Museum of Nat-ural History, Raleigh.

11

GREAT PLAINS FISHES

Table 4.

The 22 native fish species not currently listed as having spe-cial conservation status by state fish management agencies inMontana, Wyoming, or Colorado with designation (X) of themas managed sport fishes, managed prey species in reservoirs,invasive species, or showing evidence of decline in Wyomingfrom the 1960s to the 1990s, based on Patton et al. (1998).

Prey in Declining in Species Sport fish reservoirs Invasive Wyoming

Gizzard shad X

Northern pike X

Central stoneroller

Emerald shiner X

Bigmouth shiner

Red shiner X

Longnose dace

Creek chub

River carpsucker X

Quillback

Longnose sucker X

White sucker X

Smallmouth buffalo

Shorthead redhorse X

Black bullhead X X

Channel catfish X X

Flathead catfish X

Burbot X

Brook stickleback X

Green sunfish X X X

Johnny darter

Logperch


Lesica, P., and F. W. Allendorf. 1995. When are peripheralpopulations valuable for conservation? ConservationBiology 9:753–760.

Lohr, S. C., and K. D. Fausch. 1996. Effects of green sunfish(Lepomis cyanellus) predation on survival and habitatuse of plains killifish (Fundulus zebrinus). SouthwesternNaturalist 41:155–160.

Lynch, J. D., and B. R. Roh. 1996. An ichthyological survey ofthe forks of the Platte River in western Nebraska. Trans-actions of the Nebraska Academy of Sciences 23:65–84.

Matthews, W. J. 1987. Physicochemical tolerance and selec-tivity of stream fishes as related to their geographicranges and local distributions. Pages 111–120 in W. J.Matthews and D. C. Heins, editors. Community and evo-lutionary ecology of North American stream fishes. Uni-versity of Oklahoma Press, Norman.

Matthews, W. J. 1988. North American prairie streams as sys-tems of ecological study. Journal of the North AmericanBenthological Society 7:387–409.

Matthews, W. J., and E. G. Zimmerman. 1990. Potentialeffects of global warming on native fishes of the south-ern Great Plains and the Southwest. Fisheries15(6):26–32.

Menzel, B. W., J. B. Barnum, and L. M. Antosch. 1984. Eco-logical alterations of Iowa prairie agricultural streams.Iowa State Journal of Science 59:5–30.

Owen, J. B., D. S. Elsen, and G. W. Russell. 1981. Distributionof fishes in North and South Dakota basins affected bythe Garrison Diversion Unit. University of North Dako-ta, Fisheries Research Unit, Grand Forks.

Patton, T. M., and W. A. Hubert. 1993. Reservoirs on a GreatPlains stream affect downstream habitat and fish assem-blages. Journal of Freshwater Ecology 8:279–286.

Patton, T. M., F. J. Rahel, and W. A. Hubert. 1998. Using his-torical data to assess changes in Wyoming’s fish fauna.Conservation Biology 12:1120–1128.

Pflieger, W. L. 1997. The fishes of Missouri. Missouri Depart-ment of Conservation, Jefferson City.

Pflieger, W. L., and T. B. Grace. 1987. Changes in the fish faunaof the lower Missouri River, 1940–1983. Pages 166–177 inW. J. Matthews and D. C. Heins, editors. Community andevolutionary ecology of North American stream fishes.University of Oklahoma Press, Norman.

Pigg, J., R. Gibbs, and K. K. Cunningham. 1999. Decreasingabundance of the Arkansas River shiner in the SouthCanadian River, Oklahoma. Proceedings of the Okla-homa Academy of Science 79:7–12.

Quist, M. C., W. A. Hubert, and F. J. Rahel. 2003. Exotic pis-civorous fishes and reduced intermittence affect sucker-mouth minnows in a southeastern Wyoming stream.Intermountain Journal of Science 9:62–65.

Quist, M. C., W. A. Hubert, and F. J. Rahel. 2004. Relationsamong habitat characteristics, exotic species, and turbid-river cyprinids in the Missouri River drainage ofWyoming. Transactions of the American Fisheries Soci-ety 133:727–742.

Quist, M. C., W. A. Hubert, and F. J. Rahel. 2005. Fish assem-blage structure following impoundment of a Great Plainsriver. Western North American Naturalist 65:53–63.

Rabeni, C. F. 1996. Prairie legacies–fish and aquaticresources. Pages 111–124 in F. B. Samson and F. Knopf,editors. Prairie conservation: preserving America’s mostendangered ecosystem. Island Press, Washington, D.C.

Rahel, F. J., C. J. Keleher, and J. L. Anderson. 1996. Potentialhabitat loss and population fragmentation for cold waterfish in the North Platte River drainage of the RockyMountains: response to climate warming. Limnologyand Oceanography 41:1116–1123.

Resh, V. H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W. Li,G. W. Minshall, S. R. Reice, A. L. Sheldon, J. B. Wallace,and R. C. Wissmar. 1988. The role of disturbance instream ecology. Journal of the North American Bentho-logical Society 7:433–455.

Robison, H. W. 1986. Zoogeographic implications of theMississippi Basin. Pages 267–285 in C. H. Hocutt and E.O. Wiley, editors. The zoogeography of North Americanfreshwater fishes. Wiley, New York.

Ross, S. T., W. J. Matthews, and A. A. Echelle. 1985. Persis-tence of stream fish assemblages: effects of environmen-tal change. American Naturalist 126:24–40.

Sanders, R. M., D. G. Huggins, and F. B. Cross. 1993. TheKansas River system and its biota. Pages 295–326 in L. W.Hesse, C. B. Stalnaker, N. G. Benson, and J. R. Zuboy, edi-tors. Proceedings of the symposium on restoration plan-ning for the rivers of the Mississippi River ecosystem. U.S.Department of the Interior, National Biological Survey,Biological Report 19, Washington, D.C.

Schlosser, I. J. 1988. Predation rates and the behavioralresponse of adult brassy minnow (Hybognathus hankin-soni) to creek chub and smallmouth bass predation.Copeia 1988:691–697.

Schmulbach, J. C., and P. J. Braaten. 1993. The VermillionRiver: neither red nor dead. Pages 57–69 in L. W. Hesse,C. B. Stalnaker, N. G. Benson, and J. R. Zuboy, editors.Proceedings of the symposium on restoration planningfor the rivers of the Mississippi River ecosystem. U.S.Department of the Interior, National Biological Survey,Biological Report 19, Washington, D.C.

Scott, W. B., and E. J. Crossman. 1973. Freshwater fishes ofCanada. Fisheries Research Board of Canada Bulletin 184.

Sheldon, A. L. 1987. Rarity: patterns and consequences forstream fishes. Pages 203–209 in W. J. Matthews and D. C.Heins, editors. Community and evolutionary ecology ofNorth American stream fishes. University of OklahomaPress, Norman.

Sheldon, A. L. 1988. Conservation of stream fishes: patternsof diversity, rarity, and risk. Conservation Biology2:149–156.

Smith, J. B. 1988. Movement and spawning of fishes in CrazyWoman Creek, a tributary to the Powder River, Wyoming.Master’s thesis. University of Wyoming, Laramie.

Stevenson, M. M., G. D. Schnell, and R. Black. 1974. Factoranalysis of fish distribution patterns in western andcentral Oklahoma. Systematic Zoology 23:202–218.

Williams, J. E., J. E. Johnson, D. A. Hendrickson, S. Contreras-Balderas, J. D. Williams, M. Navarro-Mendoza, D. E. McAl-lister, and J. E. Deacon. 1989. Fish of North America: endan-gered, threatened, or of special concern. Fisheries 14(6):2–20.

12

HUBERT AND GORDON


Winston, M. R., C. M. Taylor, and J. Pigg. 1991. Upstreamextirpation of four minnow species due to damming of aprairie stream. Transactions of the American FisheriesSociety 120:98–105.

Wolf, A. E., D. W. Willis, and G. J. Power. 1995. Larval fishcommunities in the Missouri River below GarrisonDam, North Dakota. Journal of Freshwater Ecology11:11–19.

13

GREAT PLAINS FISHES


15



Pools, riffles, and eddies in high gradientcanyon reaches of the Colorado Riverdrainage provide habitat for humpback

chub Gila cypha and roundtail chub G. robusta.These two species occur sympatrically in theYampa River (Yampa Canyon, Colorado), GreenRiver (Desolation/Gray canyons, Utah), upperColorado River (Black Rocks, Colorado; Westwa-ter Canyon, Utah; and Cataract Canyon, Utah),and Little Colorado River (Grand Canyon, Ari-zona). While humpback chub distribution is lim-ited to this small number of canyons, current dis-tribution of roundtail chub is much broader thanhumpback chub (Bezzerides and Bestgen 2002).Because these two species are closely related andoverlap in these habitats, an understanding of thestatus of these two species over time in sympatricpopulations is desirable.

Due to declines in distribution and abundancethroughout its range, the humpback chub is cur-rently protected under the Endangered Species Actof 1973, as amended (ESA, 16 U.S.C. 1531 et. seq.).General reasons for decline can be attributed toalteration of historic habitat caused by dam con-struction and operation, water diversion and chan-nelization, competition with and predation bynonnative aquatic species, and hybridization withother Gila species. The most recent recovery planwas finalized in 1990 (USFWS 1990), with anamendment and supplement to that plan approvedin 2002 (USFWS 2002). The amendment and sup-plement to the 1990 recovery plan identifies objec-tive and measurable recovery criteria to downlistand delist humpback chub in both the upper basinand the lower basin recovery units. Within theupper basin recovery unit, one of the criteria to

Population Abundance Estimates for HumpbackChub and Roundtail Chub in Westwater Canyon,

Colorado River, Utah, 1998–2000

ABSTRACT Population abundance estimates conducted from 1998 to 2000 were completed for adult (>200

mm) humpback chub Gila cypha and roundtail chub G. robusta in Westwater Canyon on the Colorado River, Utah.

Sampling was conducted annually using a three-pass mark–recapture approach. The primary method of capture

was trammel netting with supplemental electrofishing on one pass per year. Separate abundance estimates were

generated for each year of the study using the null estimator (Mo ) within Program CAPTURE. Results showed a

decline in the adult humpback chub population between 1998 and 1999 and no change in abundance between

1999 and 2000. The adult roundtail chub population abundance in Westwater Canyon during this time period was

relatively stable. Catch per unit effort (CPUE) data from this study and historic interagency standardized monitor-

ing indicated a continued declining trend in mean CPUE for humpback chub that was significant. Mean CPUE for

roundtail chub also showed a continued declining trend, but it was not statistically significant. The results of this

study provide information to assess the current status of these two species and a point of reference for future

population estimates of chub in the upper Colorado River basin.

J. Michael Hudson and Julie A. Jackson

J. MICHAEL HUDSON U.S. Fish and Wildlife Service – Columbia River Fisheries Program Office, 1211 SE Cardinal Court, Suite 100, Vancouver,Washington 98683, USA

JULIE A. JACKSON Utah Division of Wildlife Resources – Moab Field Station, 1165 South Highway 191, Suite 4, Moab, Utah 84532, USA

NatAmerHudson1 5/25/07 1:49 PM Page 15

downlist humpback chub is the maintenance ofone of the five upper basin populations (BlackRocks, Westwater Canyon, Cataract Canyon,Yampa Canyon, Desolation/Gray Canyon) as a corepopulation with a minimum viable population of2,100 adults (≥200 mm) for five consecutive years.To delist humpback chub, the upper basin recoveryunit will have to maintain a minimum viable pop-ulation for 8 years in two of the five upper basinhumpback chub populations. Per the 2002 amend-ment and supplement, the adult humpback chubpopulation will be determined via point estimatesin each of 2–3 consecutive years with 1–2 yearsbetween these groups of abundance estimates tomeasure progress toward achieving and maintain-ing the minimum viable population. Within eachcore population, there must not be a significantdeclining trend in adult point estimates in order todownlist and delist humpback chub.

The roundtail chub is not cur-rently listed as threatened or endan-gered under the ESA. However, thisspecies is listed on the Utah SensitiveSpecies List, is considered a sensitivespecies by the U.S. Forest Service,Rocky Mountain Region (Region 2),and is listed as a Species of SpecialConcern by the National Park Ser-vice, Southeast Utah Group. Many ofthe same impacts affecting hump-back chub also affect roundtail chubdue to their similar life histories andhabitats. While recovery actions arenot identified under a recovery planas with humpback chub, a multistate,multiagency conservation agreementhas recently been developed andimplemented to take steps towardconservation of this sensitive species(UDWR 2004).

This study was conducted from1998 to 2000 with the objective ofobtaining a population abundanceestimate of adult humpback chub inWestwater Canyon. It is one of thefirst upper basin population abun-dance estimates to be conducted inaccordance with the humpback chub

recovery goals. A secondary objective was to obtaina population abundance estimate for roundtailchub in Westwater Canyon.

The results of this study provides three pointestimates that will aid in reclassification of hump-back chub under the ESA (i.e., downlist, delist),provides information for comparison with futurepopulation abundance estimates of chub in theupper Colorado River basin, and expands the exist-ing knowledge of roundtail chub.

Study Area

Westwater Canyon is located on the ColoradoRiver downstream of the Colorado–Utah border(Figure 1). The canyon is approximately 19.3 riverkilometers (RK) long (RK 200.4–181.1). It is char-acterized by the black Proterozoic gneiss and gran-

16

HUDSON AND JACKSON

Figure 1. The five humpback chub populations in the upper Colorado Riverbasin (Yampa Canyon, Desolation Canyon, Cataract Canyon, WestwaterCanyon and Black Rocks). The inset is Westwater Canyon with the threesections of the canyon and the three sampling reaches identified.


ite complex that comprise the inner gorge. Thehabitat in the upper section of the canyon consistsof runs, eddies, and pools interspersed between rif-fles and rapids. Each of these deep canyon habitatsis bounded by a riffle area. The steepest part ofWestwater Canyon is the middle section (RK192.4–187.6) and is characterized by turbulentflows and Class III rapids. The lower section ofWestwater Canyon is a confined canyon reach withreduced gradient and primarily a homogeneousrun habitat where chubs are scarce (Chart andLentsch 1999).

MethodsSampling

Sampling occurred at three sites in the upper por-tion of Westwater Canyon previously establishedfor the Interagency Standardized Monitoring Pro-gram (ISMP): Miners Cabin (RK 199.8–199.3),Cougar Bar (RK 195.6–194.8), and Hades Bar (RK193.2–192.9). The middle section was not includedbecause it could not be effectively sampled to meetthe objectives of this study. However, U.S. Fish andWildlife Service sampled the middle section ofWestwater Canyon during 1979–1981 and hump-back chub were present (Valdez et al. 1982). Thelower section was not included because of the lackof chubs in that portion of the canyon.

Sampling occurred during September andOctober in 1998–2000, with three sampling passesconducted each year. Eight days lapsed between theend of one pass and the beginning of the subse-quent pass. During each pass, Miners Cabin andCougar Bar were each sampled for two nights andHades Bar was sampled for one night. Gear includ-ed trammel nets (23 x 2 m; 2.5 cm and 1.25 cmmesh) and a pulsed DC Coffelt electrofishing unitmounted on an inflatable 16-ft Achilles sport boat.

Trammel nets were set in mid-afternoon andchecked every 2 h until midnight, at which timethey were pulled. Nets were reset before dawn andallowed to fish until late morning while beingchecked every 2 h. Trammel nets were primarily setin deep eddies off boulders or rock faces to targetjuvenile and adult chubs. However, nets were occa-sionally set in relatively shallow riffle/run areas offin-channel boulders. All Gila spp. caught were

removed from the net and placed in a holding penuntil data were collected at the end of each 18-hsampling period.

Electrofishing was conducted at each site dur-ing one of the three sampling passes each year forconsistency with preestablished protocols from theISMP (USFWS 1987). Shoreline habitats were elec-trofished at each site. Electrofishing was conductedwith consistent effort prior to setting nets in lateafternoon and subsequent to nets being pulled atnight during each 18-h sampling period. Elec-trofishing targeted the late juvenile/adult compo-nent of the population and data were used in con-junction with trammel net data for initial capturesto derive a population abundance estimate.

Gila spp. were identified to species using a suiteof diagnostic characters (i.e., degree of frontaldepression, presence of scales on nuchal hump,angle of base of anal fin relative to angle of top ofcaudal fin) in conjunction with the “art of seeingwell” (Douglas et al. 1998). Information collectedfrom all Gila spp. captures included total length(mm) and weight (g). All captured Gila spp. werescanned for a passive integrated transponder (PIT)tag. PIT tag numbers were recorded from recap-tured chubs. Initial chub captures of fish greaterthan 150 mm received a PIT tag and the numberwas recorded.

Data analysis

Population estimates were determined for adulthumpback chub and roundtail chub (>200 mm)in Westwater Canyon using closed populationmodels within Program CAPTURE (Otis et al.1978; White et al. 1982; Rexstad and Burnham1991). Program CAPTURE was used to help deter-mine the most appropriate estimator (Mo [nullestimator], Jackknife Mh , Darroch Mt , Chao Mth ,Chao Mt , and Chao Mh), but assumptions andvariables associated with the choice of the mostappropriate estimator were also considered. Inde-pendent adult population abundance estimateswere calculated for each species in each year. Pro-gram CAPTURE was used to determine confi-dence intervals around the estimate, the coeffi-cient of variation, and the probability of capture.Linear regression analyses were conducted on theresulting population estimates for the respective

17

POPULATION ABUNDANCE ESTIMATES FOR HUMPBACK CHUB AND ROUNDTAIL CHUB


species to examine short-term trends in the popu-lations throughout the period of this study.

Confidence intervals were determined for allestimators; however, profile likelihood intervalswere determined in lieu of 95% confidence inter-vals for Mo (null estimator) and the Darroch Mt.The profile likelihood interval helps to account formodel selection uncertainty by providing widerconfidence intervals (D. R. Anderson and G. C.White, Colorado State University, Ft. Collins, per-sonal communication). In addition, these intervalstend to give more correct confidence intervals forsmall samples (Ross Moore, Mathematics Depart-ment, Macquarie University, Sydney, Austrailia,personal communication).

Catch per unit effort (CPUE) was determinedfor trammel net effort (fish/net hour) for the cap-ture of humpback and roundtail chub through theperiod of this study. CPUE was compared amongpasses within and among years using the Kruskal-Wallis nonparametric analysis of variance(ANOVA) with Dunn’s multiple comparisons testto examine the equality of samples and the two-sample Kolmogorov-Smirnov to compare the dis-tribution of catch rates. In addition, total annualCPUE comparisons were tested among years usingthe same analyses. CPUE for electrofishing effort(fish/electrofishing hour) is not reported due to alimited amount of information to analyze.

ISMP protocols were followed on one pass dur-ing each year of this study to allow comparisonwith historic humpback chub and roundtail chubCPUE information. CPUE was compared amongyears (1988–2000) using the Kruskal-Wallis non-parametric ANOVA with Dunn’s multiple compar-isons test to examine the equality ofsamples and the two-sample Kol-mogorov-Smirnov to compare the dis-tribution of catch rates.

ResultsHumpback chub

Population estimates.—After consider-ing the results of the model selectionfunction, the lack of any real justifica-tion to consider another estimator(i.e., changes in the probability of cap-

turing an individual due to behavior, flow, etc.),and consultation with appropriate expertise (Dr.Ron Ryel, Utah State University, Logan, Utah) weselected the null estimator (Mo) to estimate pop-ulation abundance of adult humpback chub inWestwater Canyon, 1998–2000. The adult hump-back chub population estimate was 4,744 indi-viduals in 1998 (Table 1). The profile likelihoodinterval around this estimate was 3,760–14,665(CV = 0.23; = 0.035). In 1999, the adult hump-back chub population estimate decreased to2,215 individuals (Table 1). The profile likelihoodinterval around this estimate was 1,608–7,508(CV = 0.28) with a slightly higher probability ofcapture ( = 0.041). The adult humpback chubpopulation estimate in 2000 remained approxi-mately the same at 2,201 individuals (Table 1). Theprobability of capture also remained the same ( =0.041). However, the profile likelihood interval wastighter (1,335–4,124; CV = 0.28).

The relationship among these three estimatesindicates a short-term decline in the WestwaterCanyon adult humpback chub population (Figure2). However, the slope of this short-term trend doesnot significantly depart from zero and each pointestimate exceeded the minimum viable populationidentified in the recovery goals (USFWS 2002).

CPUE.—Total trammel net captures for hump-back chub were 501 in 1998, 278 in 1999, and 277in 2000 (Table 2). Total captures in 1998 included486 adults, 12 recaptures, and 3 subadults. In 1999,we captured 267 adults, 8 recaptures, and 3subadults. The total captures for 2000 included 261adults, 11 recaptures, and 5 subadults.

Mean catch per unit effort (CPUE) decreased

p

p

p

18

HUDSON AND JACKSON

Table 1.Population estimate (N) for humpback chub adults (>200 mm) in WestwaterCanyon 1998–2000. Population estimate generated using the null estimator(Mo) within program CAPTURE. The profile likelihood interval, coefficient ofvariation (CV), and probability of capture ( ) are included with the respec-tive population estimates.

Profile likelihood Year N interval CV

1998 4,744 3,760–14,665 0.23 0.035

1999 2,215 1,608–7,508 0.28 0.041

2000 2,201 1,335–4,124 0.28 0.041

p

p


from 1998 to 2000 (Figure 3). Theslope of this decreasing trend sig-nificantly departs from zero (p <0.05). Comparisons of mean CPUEand the distribution of catch ratesbetween years (all passes com-bined) indicated significant differ-ences (p < 0.05) between all years.The Kruskal–Wallis test indicatedthat there were no significant dif-ferences between passes withinyears, but there were significantdifferences (p < 0.05) betweenpasses among years. The Kol-mogorov-Smirnov test indicatedthat there were no significant dif-ferences in the distribution ofcatch rates between passes within1998 and 1999. However, therewere significant differences (p <0.05) between passes within 2000and among years.

Comparison with ISMP.—Thehistoric (1988–2000) ISMP CPUEfor humpback chub in WestwaterCanyon indicates a significantdecreasing trend (Figure 4; p <0.05). Mean catch per unit effortbetween 1988 and 2000 as perISMP sampling protocol was sig-nificantly different among years (p< 0.05; Kruskal-Wallis). Further-more, the distribution of catchrates around the mean was signifi-cantly different among years (p <0.05; Kolmogorov-Smirnov).

Roundtail chub

Population estimates.—Consider-ing the results of the model selec-tion function of Program CAP-TURE and to allow an accuratecomparison with the populationestimate of humpback chub inWestwater Canyon, the null esti-mator (Mo) was used to determinethe population estimates of adultroundtail chub in Westwater

19


Table 2.Total humpback chub captures in Westwater Canyon 1998–2000. Includes number of adult captures, adult recaptures, and subadults caught.

Total Adult Adult SubadultYear captures captures recaptures captures

1998 501 486 12 3

1999 278 267 8 3

2000 277 261 11 5

Figure 2. Short-term trend in point estimates of Westwater Canyon adulthumpback chub population (N) for 1998–2000 and profile likelihood confi-dence intervals for each estimate.

Figure 3. Westwater Canyon humpback chub trammel net catch per unit effortfor 1998–2000. CPUE for each year includes respective standard error. Linerepresents trend in all years. The p-value indicates statistical significance ofthe trend line.


Canyon, 1998–2000. The adult roundtail chubpopulation estimate was 5,005 individuals in 1998(Table 3). The profile likelihood interval aroundthis estimate was 3,586–19,781 (CV = 0.30; =0.026). In 1999, the adult roundtail chub popula-tion estimate decreased to 4,234 individuals (Table3). The profile likelihood interval around this esti-mate was 3,349–12,917 (CV = 0.23) with a slightlyhigher probability of capture ( = 0.037). Theadult roundtail chub population estimate in 2000increased to approximately the 1998 abundance at4,971 individuals (Table 3). While the probability ofcapture decreased at the same time ( = 0.031), theprofile likelihood interval was slightlytighter (3,824–16,641; CV = 0.25). Therelationship among these three estimatesindicates short-term stability in theWestwater Canyon adult roundtail chubpopulation (Figure 5).

CPUE.—Total trammel net capturesfor roundtail chub were 397 in 1998,481 in 1999, and 521 in 2000 (Table 4).Total captures in 1998 included 389adults, 7 recaptures, and 1 subadult. In1999, we captured 457 adults, 12 recap-tures, and 12 subadults. The total cap-

tures for 2000 included 458 adults,10 recaptures, and 53 subadults.

Mean CPUE decreased slightlyacross all passes from 1998 to 2000(Figure 6), although the slope of thisdecreasing trend does not significant-ly depart from zero. Comparisons ofmean CPUE and the distribution ofcatch rates among years (all passescombined) indicated significant dif-ferences (p < 0.05) between all yearsexcept for mean CPUE between 1998and 1999. Contrary to this decreasingtrend, CPUE increased among passeswithin years. The Kruskal–Wallis testsupported this observation indicat-ing significant differences (p < 0.05)between passes within and amongyears, except for 2000. The Kol-mogorov-Smirnov test indicatedthere were significant differences (p <0.05) in the distribution of catch rates

between passes within and among all years.Comparison with ISMP.—The historic

(1988–2000) ISMP catch per unit effort forroundtail chub in Westwater Canyon indicates aslight decreasing trend (Figure 7), but the slopeof this trend does not significantly depart fromzero. Mean catch per unit effort between 1988and 2000 as per ISMP sampling protocol was sig-nificantly different between 1990 and 2000 andbetween 1993 and 2000 (p < 0.05; Kruskal-Wal-lis). The distribution of catch rates around themean was significantly different among years (p< 0.05; Kolmogorov-Smirnov).p

p

p

20

HUDSON AND JACKSON

Table 3.Population estimate (N) for roundtail chub adults (>200 mm) in WestwaterCanyon 1998–2000. Population estimate generated using the null estimator (Mo) within program CAPTURE. The profile likelihood interval,coefficient of variation (CV), and probability of capture ( ) are included with the respective population estimates.

Profile likelihood Year N interval CV

1998 5,005 3,586–19,781 0.30 0.026

1999 4,234 3,349–12,917 0.23 0.037

2000 4,971 3,824–16,641 0.25 0.031

p

Figure 4. Westwater Canyon humpback chub catch per unit effort for sam-ples collected using ISMP protocol from 1988 to 2000. CPUE for each yearincludes respective standard error. Line represents trend among all passesin all years. The p-value indicates statistical significance of the trend line.

p


DiscussionPopulation estimates

Population estimates of humpback chub in West-water Canyon demonstrated a downward trendfrom 1998 to 2000. This trend is consistent with thepoint estimates (1998–2000) of humpback chub inBlack Rocks, Colorado (McAda 2002). Previouspopulation estimates (Chart and Lentsch 1999;Nesler 2000), while not as robust, indicate that thehumpback chub population in Westwater Canyonis highly variable. Chart and Lentsch (1999) deter-mined the population to be 5,621 individuals in1994, 10,148 individuals in 1995, and 5,186 indi-viduals in 1996. Nesler determined the WestwaterCanyon humpback chub population to be any-where from 5,719 in 1993 (90% survival) to 1,164

in 1997 (59% survival). Point esti-mates in Westwater Canyon for 1999and 2000 may be indicative of a lev-eling off prior to a rebound in thehumpback chub population, or thedeclining trend could continue. TheWestwater Canyon roundtail chubpopulation appeared to be relativelystable from 1998 to 2000. Chart andLentsch (1999) indicated that theroundtail chub population wasdeclining in the period from 1993 to1996 (6,809 in 1993, 5,733 in 1994,and 2,551 in 1996). These combineddata sets further support theobserved variability of chub popula-tions in Westwater Canyon. Identicaleffort was applied toward the cap-ture of humpback chub and round-tail chub throughout the 3 years ofthe study, increasing our confidence

that the humpback chub decline is realand not an artifact of sampling bias.

Humpback chub and roundtail chubpopulation abundance patterns observedin this study and by Chart and Lentsch(1999) may indicate that these twospecies coexist in Westwater Canyonthrough equilibrium of populationdynamics. While these two studies arenot strictly comparable due to different

approaches to sampling and analysis of the data,some general observations can be made. From1993–1996, the Westwater Canyon roundtail chubpopulation appeared to be declining. At the sametime, the humpback chub population was variable,but relatively larger. Conversely, from 1998 to 2000,adult humpback chub appeared to decline and sta-bilize while the roundtail chub population was sta-ble and relatively larger. Population estimates ofWestwater Canyon humpback chub scheduled for2003–2005 will contribute to the existing data andfurther clarify the short-term population trends andpopulation dynamics between humpback chub androundtail chub in Westwater Canyon.

This study was not designed to sample forsubadult humpback chub in Westwater Canyonfor the purpose of generating subadult abundance

21


Table 4.Total roundtail chub captures in Westwater Canyon 1998–2000. Includesnumber of adult captures, adult recaptures, and subadults caught.

Total Adult Adult SubadultYear captures captures recaptures captures

1998 397 389 7 1

1999 481 457 12 12

2000 521 458 10 53

Figure 5. Westwater Canyon adult roundtail chub population estimates (N)for 1998–2000. Each point estimate includes respective profile likelihoodconfidence intervals. Line represents short-term trend among the three pointestimates. The p-value indicates statistical significance of the trend line.


estimates. Future efforts should include a compo-nent of sampling designed to capture subadulthumpback chub. This would provide informationtoward the recovery goal component of determin-ing mean estimated recruitment of humpbackchub and also provide additional insight towardpopulation dynamics of the Westwater Canyon

population as revealed by the adultpopulation point estimates.

Confidence intervals aroundhumpback chub point estimatesimproved from 1998 to 2000, butthere was no considerable improve-ment in the coefficient of variationor the probability of capture.Tighter confidence intervals in 1999than in 1998 were an artifact of asmaller population estimate. How-ever, increased trammel net effort in2000 relative to previous years(1,329 h in 1998, 1,306 h in 1999,and 1,951 h in 2000) resulted intighter confidence intervals whilemaintaining a similar point estimateto 1999. Increased effort using alter-native sampling methods (e.g., hoopnets and electrofishing) may improvethe coefficient of variation and prob-ability of capture for humpback chuband roundtail chub. McAda (2002)demonstrated an improvement inthese measures by incorporating afourth pass to the sampling design.

Precision of the abundance esti-mate may be further improved byadding additional sampling siteswithin Westwater Canyon and com-bining Black Rocks and WestwaterCanyon abundance estimates. Addi-tional sampling sites are availablewithin the canyon. Adding these sitesmay improve the probability of cap-ture over all passes. In addition, lim-ited information from this study andMcAda (2002) indicate limitedmovement between Black Rocks andWestwater Canyon. Combining thedata from the two canyons may pro-

vide more precise abundance estimates.Assessing population dynamic trends using

regression analysis can also be considered a weakapproach from two perspectives. Currently, thereare only three point estimates, which provide lit-tle power to the analysis. In addition, as morepoint estimates become available, increases and

22

HUDSON AND JACKSON

Figure 6. Short-term trend in point estimates of Westwater Canyon adultroundtail chub population (N) for 1998–2000 and profile likelihood confi-dence intervals for each estimate.

Figure 7. Westwater Canyon roundtail chub catch per unit effort for samplescollected using ISMP protocol from 1988 to 2000. CPUE for each yearincludes respective standard error. Line represents trend among all passesin all years. The p-value indicates statistical significance of the trend line.


23

decreases in population abundance over time maybe lost in the overall trend analysis. However,improvements in the sampling approach and dataanalysis could result in a more accurate represen-tation of the population dynamics of humpbackchub and roundtail chub populations in theupper Colorado River basin.

Comparison with ISMP

The decreasing trend of historic CPUE was statisti-cally significant across all years for WestwaterCanyon humpback chub. Likewise, the adult pop-ulation point estimates indicated a decreasingtrend for 1998–2000. However, the point estimatesfor 1999 and 2000 were similar. McAda (2002) stat-ed that the marked decline in mean CPUE suggeststhat the Black Rocks humpback chub populationhas declined since 1986. This statement was furthersupported by a decrease in CPUE in 2000 and areflected decrease in the adult population pointestimate. Thus, based upon an extremely limiteddata set, it appears that CPUE may be indicative oftrends in the adult humpback chub population.

Paukert (2004) argues that trammel net CPUEfor rare fishes may not be an appropriate index ofabundance because the sample sizes needed prob-ably require more effort than would be possible. Inpartial support of this argument, increased effortin 2000 resulted in tighter confidence intervals forthe abundance estimate. However, the amount ofeffort required for reliable abundance estimatesand indices of abundance may be obtainable. Addi-tional information from future sampling efforts inWestwater and Black Rocks canyons will aid in fur-ther refining the relationship between CPUE andadult population point estimates.

If a relationship between an index of abundance(CPUE) and abundance estimates can be con-firmed, it may be possible to determine an annualindex of abundance for humpback chub in West-water Canyon, as well as other humpback chubpopulations, that can be related to the recoverygoals and the downlisting/delisting process. Shouldthe downlisting/delisting process for humpbackchub take longer than 8 years, this approach couldprovide more frequent monitoring of populationabundance without the added impacts of intensivemultipass, multiyear abundance estimates.

Acknowledgments

We would like to acknowledge Tom Chart, MelissaTrammell, Steve Meismer, and Pete Cavalli, wholent their considerable biological expertise to vari-ous portions of this study. We would also like tothank Ron Ryel, Utah State University, and KevinBestgen, Colorado State University, for statisticalguidance. Many thanks to Kevin Bestgen, ChuckMcAda, and Rich Valdez for their in depth reviewand critique of the initial report submitted to theRecovery Implementation Program for Endan-gered Fish Species in the Upper Colorado RiverBasin. Finally, we would like to thank the numer-ous technicians and volunteers without whom thisstudy would not have been possible.

This study was funded by the Recovery Imple-mentation Program for Endangered Fish Speciesin the Upper Colorado River Basin. The recoveryprogram is a joint effort of the U.S. Fish andWildlife Service, U.S. Bureau of Reclamation,Western Area Power Administration, states of Col-orado, Utah, and Wyoming, upper basin waterusers, environmental organizations, the ColoradoRiver Energy Distributors Association, and theNational Park Service.

References

Archer, D. L., L. R. Kaeding, B. D. Burdick, and C. W. McAda.1985. A study of the endangered fishes of the Upper Col-orado River. U.S. Fish and Wildlife Service, Grand Junc-tion, Colorado.

Bezzerides, N., and K. Bestgen. 2002. Status review of round-tail chub Gila robusta, flannelmouth sucker Catostomuslatipinnis, and bluehead sucker Catostomus discobolus inthe Colorado River basin. Colorado State University, Lar-val Fish Laboratory Contribution 118, Ft. Collins.

Chart, T. E., and L. Lentsch. 1999. Flow effects on humpbackchub Gila cypha in Westwater Canyon. Utah Division ofWildlife Resources, Publication Number 99–36, SaltLake City.

Chart, T. E., and L. Lentsch. 2000. Reproduction and recruit-ment of Gila spp. and Colorado pikeminnow Pty-chocheilus lucius in the Middle Green River 1992–1996.Utah Division of Wildlife Resources, Publication Num-ber 00–18, Salt Lake City.

Douglas, M. E., R. R. Miller, and W. L. Minckley. 1998. Mul-tivariate discrimination of Colorado plateau Gila spp.:the “art of seeing well” revisited. Transactions of theAmerican Fisheries Society 127:163–173.

McAda, C. W. 2002. Population size and structure of hump-



back chub in Black Rocks, 1998–2000. U.S. Fish andWildlife Service, Colorado River Fisheries ProjectOffice, Grand Junction, Colorado.

Mills, L. S., and F. W. Allendorf. 1996. The one-migrant-per-generation rule in conservation and management.Conservation Biology 10:1509–1518.

Nesler, T. P. 2000. Recovery of the Colorado River endan-gered fishes: biological recovery goals and criteria forColorado pikeminnow, humpback chub, razorbacksucker, and bonytail. Colorado Division of Wildlife,Fort Collins.

Otis, D. L., K. P. Burnham, G. C. White, and D. R. Anderson.1978. Statistical inference from capture data on closedanimal populations. Wildlife Monographs 62:1–135.

Paukert, C. P. 2004. Comparison of electrofishing andtrammel netting variability for sampling native fishes.Journal of Fish Biology 65:1643–1652.

Rexstad, E. and K. Burnham. 1991. User’s guide for interac-tive program CAPTURE. Colorado State University,Colorado Cooperative Fish and Wildlife Research Unit,unpublished report, Fort Collins.

UDWR (Utah Department of Wildlife Resources). 2004.Range-wide conservation agreement for roundtail chubGila robusta, bluehead sucker Catostomus discobolus,and flannelmouth sucker Catostomus latipinnis. Pre-pared for the Colorado Fish and Wildlife Council. UtahDepartment of Natural Resources, Utah Division ofWildlife Resources, Salt Lake City.

USFWS (U.S. Fish and Wildlife Service). 1987. InteragencyStandardized Monitoring Program handbook. U.S. Fish

and Wildlife Service, Colorado River Fisheries ProjectOffice, Grand Junction, Colorado.

USFWS (U.S. Fish and Wildlife Service). 1990. Humpbackchub recovery plan, 2nd revision. Report of ColoradoRiver Fishes Recovery Team to U.S. Fish and WildlifeService, Region 6, Denver.

USFWS (U.S. Fish and Wildlife Service). 2002. Humpbackchub Gila cypha recovery goals: an amendment andsupplement to the humpback chub recovery plan. U.S.Fish and Wildlife Service, Mountain-Prairie Region 6,Denver.

Valdez, R. A., and G. C. Clemmer. 1982. Life history andprospects for recovery of humpback chub and bonytailchub. Pages 109–110 in W. H. Miller, H. M. Tyus, and C.A. Carlson, editors. Fishes of the upper Colorado Riversystem: present and future. American Fisheries Society,Western Division, Bethesda, Maryland.

Valdez, R. A., and R. J. Ryel. 1995. Life history and ecologyof the humpback chub Gila cypha in the ColoradoRiver, Grand Canyon, Arizona. Final Report to theBureau of Reclamation, Contract No. 0-CS-40–09110,Salt Lake City, Utah.

Valdez, R. A., P. Mangan, R. Smith, and B. Nilson. 1982. Part2 Colorado River Fishery Project Final Report FieldInvestigations. U.S. Fish and Wildlife Service and U.S.Bureau of Reclamation, Salt Lake City, Utah.

White, G. C., D. R. Anderson, K. P. Burnham, and D. L. Otis.1982. Capture-recapture and removal methods forsampling closed populations. Los Alamos NationalLaboratory, LA-8787-NERP, Los Alamos, New Mexico.

24

HUDSON AND JACKSON


25

In September 2003, the U.S. Fish and Wild-life Service (USFWS) removed splittail Pogo-nichthys macrolepidotus (Figure 1) from the list

of threatened species under the EndangeredSpecies Act (ESA; USFWS 2003a). This unprece-dented step represented the first extant fish ever tobe removed from the federal list of threatened andendangered species. Splittail had been the subjectof intense scrutiny and research prior to and fol-lowing its initial listing in February 1999 (USFWS1999). The attention to splittail was not limited tothe federal review, as the California Department ofFish and Game (CDFG) listed the fish as a speciesof special concern in 1989, a status that has contin-ued to the present.

Following the extinction of Clear Lake splittailP. ciscoides by the 1970s, splittail presently repre-sents the only surviving member of its genus(Moyle 2002; Moyle et al. 2004). The distinguishingfeature of splittail compared to other minnows is

the enlarged dorsal lobe of the caudal fin. Likemost California cyprinids, the splittail is a relative-ly large member of its family, sometimes exceeding40 cm standard length. Adults are characterized byan elongated body, a distinct nuchal hump (on theback of the head), and a small, blunt head. Splittailadults usually also have barbels at the corners ofthe slightly subterminal mouth, an unusual featureamong North American cyprinids. Splittail are typically dull, silvery-gold on the sides and olive-gray dorsally, but during spawning season, the pec-toral, pelvic, and caudal fins have an orange-redtinge. Males also develop small white nuptialtubercles on the head, with additional tubercles onthe base of the fins.

This paper reviews some of the recent data andactivities that contributed to the 2003 USFWSdecision: (1) trends in abundance and distribution,(2) factors that affect the population, and (3)restoration efforts that could benefit splittail. Over-

Splittail “Delisting”: A Review of Recent Population Trends and Restoration Activities

ABSTRACT Splittail Pogonichthys macrolepidotus, a minnow native to the San Francisco Estuary, was original-

ly listed by the U.S. Fish and Wildlife Service as threatened in 1999. The listing was remanded in 2003 based on

recent evidence about its status and efforts to restore the species. Although young-of-year production declined dur-

ing a 6-year drought prior to the listing, the return of wet conditions in the late 1990s resulted in record indices of

abundance. Much of the minnow's historical off-channel habitat was lost by the early 1900s, but surveys suggest

that the current range of splittail has stabilized. Year-class strength is directly related to the duration of inundation

of remaining floodplain. Adults migrate upstream in winter or early spring to spawn on seasonally inundated vege-

tation. Their offspring rear in the food-rich floodplain habitat before emigrating with receding floodwaters. Based on

the recognition that the species is perhaps one of the most floodplain-dependent fishes in the estuary, floodplain

restoration became a central component of a major agency/stakeholder effort to fix long-standing problems in the

region. Floodplain restoration is likely to substantially improve the long-term status of splittail, although extreme

alterations in the food web from alien species may prevent the minnow from returning to historical levels.

Ted R. Sommer, Randall D. Baxter, and Frederick Feyrer



TED R. SOMMER California Department of Water Resources, 901 P Street, Sacramento, California 95814, USA

RANDALL D. BAXTER California Department of Fish and Game, 4001 North Wilson Way, Stockton, California 95205, USA

FREDERICK FEYRER California Department of Water Resources, 901 P Street, Sacramento, California 95814, USA

NatAmerFishWA-Sommer2.qxd 5/25/07 1:47 PM Page 25

all, our understanding ofsplittail biology hasincreased dramatically overthe past decade, leading tothe development of promis-ing restoration efforts tohelp enhance production.

Study Area

Splittail occur only in theSan Francisco Estuary andits tributaries (Figure 2).The estuary includes exten-sive downstream bays (Sui-sun, San Pablo, and SanFrancisco) and a delta, a network of tidalchannels that receive inflow from the Sacra-mento and San Joaquin rivers. San PabloBay, the lower limit of the present range ofsplittail, has small tributaries, includingPetaluma and Napa rivers, each of whichhas substantial marsh habitat. The mostextensive marshes of the region occur in the34,000-ha Suisun Marsh, which includes anetwork of vegetation-lined tidal sloughs(Matern et al. 2002). Most of the majordelta channels are deep (frequently > 5 m),tidally influenced waterways bordered bysteep, rock-covered banks with a narrowriparian corridor and minimal emergentvegetation (Sommer et al. 2001a, 2001b).Two key features of the delta are the StateWater Project (SWP) and the Central ValleyProject (CVP), large water diversions thatpresently divert from 35% to 65% of deltainflow, depending on the time of year(Brown et al. 1996; Sommer et al. 1997).The primary floodplain of the delta is theYolo Bypass, a 61-km long, partially leveedbasin that floods in winter and spring inabout 60% of years (Sommer et al. 2001a).Land use in the Yolo Bypass was predomi-nantly agricultural during the past severaldecades, but recent restoration and landacquisition have reallocated the majority ofthe area to wildlife habitat. The only othersubstantial area of floodplain within the

26

SOMMER ET AL.

Figure 1. Adult splittail Pogonichthys macrolepidotus.

Figure 2. The range of splittail, including the San Francisco Estuaryand its tributaries. The San Francisco Estuary includes the regionfrom San Francisco Bay upstream to Sacramento and a location 56km upstream of Stockton. The delta represents the portion of theestuary upstream of the confluence of the Sacramento and SanJoaquin rivers. The confluence is located at Chipps Island, whichrepresents river kilometer 0 for both rivers.


delta is the lower Cosumnes River, which has asuite of habitats, including freshwater wetlands,riparian forests, grasslands, and tidal sloughs(Crain et al. 2004). Other tributaries to the deltainclude the Sacramento, Feather, American,Mokelumne, and San Joaquin rivers. These riversare extensively channelized, although several havesetback levees on their lower reaches that providesome floodplain habitat (Mount 1995). For theseupstream areas, the largest contiguous area offloodplain is the Sutter Bypass, a 7,300-ha seasonalflood basin immediately upstream of Yolo Bypass(Sommer et al. 2001a).

The entire range of splittail has been substan-tially altered by a variety of anthropogenic factors,including levees, dams, land reclamation activities,water diversions, and contaminants (Atwater et al.1979). In addition to changes in physical habitat,the biota of the estuary has been altered by a largenumber of species introductions (Cohen and Carl-ton 1998). Native fishes have shown populationdecreases due to multiple factors (Bennett andMoyle 1996), resulting in the listing of Central Val-ley steelhead Oncorhynchus mykiss, two races ofChinook salmon O. tshawytscha, delta smeltHypomesus transpacificus, and, formerly, splittail.

Life History of Splittail

Splittail may live for 8–10 years but do not typical-ly live longer than 5 years (Moyle 2002; Moyle et al.2004). The largest and oldest fish are females,which are highly fecund, laying up to 200,000 eggsper fish (Feyrer and Baxter 1998). Fish usuallyreach sexual maturity by the end of their secondyear (Daniels and Moyle 1983). Spawning successis highly variable among years but is correlatedwith freshwater outflow and the availability of shal-low-water habitat with submerged vegetation(Daniels and Moyle 1983; Sommer et al. 1997). Intypical years, adults begin a gradual upstreammigration towards spawning areas sometimebetween late November and late January, but sub-stantial migration can also occur in spring.Upstream movement appears to coincide with flowpulses that inundate floodplains and riparian areasin which splittail forage and spawn (Harrell andSommer 2003; Moyle et al. 2004). Peak spawning

occurs from the months of March through April,although records of spawning exist for late Januaryto early July (Wang 1986; Moyle 2002). Splittail arethought to lay their adhesive eggs on submergedvegetation in flooded areas in the lower reaches ofrivers and sloughs (Caywood 1974; Moyle 2002).Laboratory studies indicate that developingembryos hatch in 3–5 d at 18.5°C (Moyle et al.2004). Splittail are 7–8 mm total length when theyolk is absorbed and feeding begins, typically onsmall rotifers, at 5–7 d posthatch. Pond studiesindicate that the early life stages of splittail use avariety of different habitats but are strongly associ-ated with shallow water areas (Sommer et al. 2002).

Distribution

Splittail are endemic to the sloughs, lakes and riversof the Central Valley that connect to the estuary(Caywood 1974; Moyle 2002; Moyle et al. 2004). Inthe Sacramento Valley, they were found in earlysurveys as far up the Sacramento River as Redding,up the Feather River as high as Oroville, and in theAmerican River to Folsom (Rutter 1908). Archaeo-logical evidence from the San Joaquin basin indi-cates that splittail were abundant in two large lakes,where they were harvested by native people (Moyle2002; Moyle et al. 2004).

Part of the rationale for the proposed listing ofsplittail in 1994 was that the USFWS believed thatthere had been a recent major restriction in therange of the species (USFWS 1994). Specifically,they concluded that the distribution of splittail hadbecome confined to the Sacramento–San JoaquinDelta and Suisun Bay. The upstream extent of thedistribution of splittail for most of the major riverswithin its range is summarized in Table 1. For com-parative purposes, we list data for (1) the historicalrange of splittail as described by Rutter (1908), (2)the known range in the 1970s (Caywood 1974), (3)results for the mid-1990s (Sommer et al. 1997), (4)our most recent observations (Feyrer et al. 2005),and other previously unpublished observations.These data suggest that splittail has retained muchof its historical range in the major rivers, althoughthe data do not reflect extensive losses of channeledge complexity and riparian habitat or offchannel floodplain habitat due to levee construc-

27

SPLITTAIL “DELISTING”


tion (Mount 1995). Further, Feyrer et al. (2005)demonstrated that upstream distribution of age-0fish in the Sacramento River has remained persist-ent at 232–296 km upstream from the estuary, asmeasured by a monitoring program spanning 28years (1976–2003). In any case, the geographic dis-tribution of splittail is broader than previouslybelieved and has not changed detectably over thepast several decades.

One factor complicating comparisons withhistorical data is that the distribution of splittailchanges seasonally and annually (Sommer et al.1997; Feyrer et al. 2005). At present, the dominantlife history pattern is for the majority of the split-tail population to remain in the estuary duringsummer and early fall, with adults migratingupstream to spawn in the delta and its tributariesin late fall–early spring. Walford (1931) observedseasonal trends in catch of splittail in the delta,strongly suggesting that seasonal shifts in distri-bution also occurred historically. However, thehistorical data are insufficient to determinewhether large numbers of splittail typicallyremained in upstream tributaries on a year-roundbasis. Annual variation in the present distributionis also evident from the juvenile splittail data.Sommer et al. (1997) used beach seine data toexamine the annual distribution of age-0 splittail.The data continue to show that the distribution of

young splittail varies substantially among years(Figure 3). There is some indication that the dis-tribution of splittail tends to be furthest upstreamin dry years, likely because the lack of inundatedfloodplain or vegetated channel boarders withinlow elevation areas forces splittail to migrate fur-ther upstream to find suitable spawning habitat(Feyrer et al. 2005). Again, it is unknown whetherthis trend occurred historically.

The fact that distribution of splittail remainsrelatively broad is likely a major factor that hashelped it avoid extinction. A large potential rangereduces the risk of a localized extinction and helpsexpand the potential area of habitat. By contrast, itsextinct cogener, P. ciscoides, was only known tooccur in Clear Lake (Figure 2) and its tributaries(Moyle 2002). The relatively isolated 17,670-halake has been substantially altered by habitat degra-dation and species introductions, both of whichwere likely primary causes of its extinction.

Splittail Abundance Trends

The historic abundance of splittail is not known,but they were abundant enough to be harvested bynative peoples and commercial fisheries in the 19thand early 20th centuries (Walford 1931; Moyle2002; Gobalet et al. 2004). There has never been aneffort to estimate the population size of splittail.

28

SOMMER ET AL.

Upstream-most locations of historical and recent splittail collections (1998–2002). River kilometer (rkm) is the distance fromthe mouth of the river.

Location (rkm) of splittail collection

Mid-1990s Recent Distance River Historical 1970s (Sommer et al. (Feyrer et al. 05) to first system (Rutter 1908) (Caywood 1974) 1997) unless noted otherwise dama

Sacramento 483 387 331 391b 387Feather 109 Present 94 94c 109American 49 37 19 No new data 37San Joaquin Widespread Present 201 218.5d 295Mokelumne NA 25 63 96e 63Napa NA 21 10 32 NAPetaluma NA 25 8 28 NA

Table 1.

a Lowest dams in each river are Red Bluff (Sacramento), Oroville (Feather), Nimbus (American), Sack (San Joaquin), and Woodbridge (Mokelumne).Woodbridge is a seasonal dam. Napa River is not dammed within the range of splittail; first dam was removed from the Petaluma River in 1994.

b D. Killam, California Department of Fish and Game, personal communication.c B. Oppenheim, NOAA Fisheries, personal communication.d R. Baxter, California Department of Fish and Game, unpublished data.e J. Merz, East Bay Municipal Utility District, personal communication, November 2000.


29


However, there are seven sampling efforts by theInteragency Ecological Program, a consortium ofstate and federal agencies that capture splittail fre-quently enough to allow the development of abun-dance indices (Sommer et al. 1997; Baxter 1999).Abundance data for age-0, age-1, and age-2+ split-tail are summarized for several of the surveys inFigures 3–5. The CDFG fall midwater trawl is amonthly survey that samples 100+ sites from SanPablo Bay to Rio Vista during September throughDecember (Sommer et al. 1997). The survey repre-sents one of the best long-term data sets and coversa large portion of the range of splittail. Key limita-

tions of the survey for splittailinclude a low catch of adults,insufficient coverage of theupstream range, and insuffi-cient data to separate age class-es before 1975. The Universityof California at Davis SuisunMarsh otter trawl is a monthlysurvey that samples sevensloughs in Suisun Marsh. Thesurvey is geographically local-ized and therefore may not berepresentative of the range ofsplittail, but it is particularlyvaluable because it has relative-ly good catch of multiple age-classes. The SWP salvage is anabundance index based on col-lection of migrating splittail atthe water diversion’s fishscreens. Like the Suisun Marshsurvey, SWP salvage is geo-graphically localized; however,it is also considered an impor-tant abundance index becauseit has the largest splittail catchof any of the surveys for allage-classes.

Based on these data, it isclear that splittail abundancevaries widely between years,particularly for age 0 (Figure4). A consistent trend is thatabundance of age-0 splittail isrelatively low during dry years

(Meng and Moyle 1995; Sommer et al. 1997; Bax-ter 1999; Moyle et al. 2004). This effect was mostpronounced during drought years in 1987–1992and 1994. From 1995 to 2001, age-0 splittail abun-dance indices improved substantially, with 1998showing the highest splittail abundance index everrecorded for most surveys. Age-1 year-classes tendto reflect age-0 abundance in the preceding year,with the highest abundance in years following astrong year-class (Figure 5). Overall, age-2+ trendsare much less variable, likely because the multiyearage structure buffers the populations (Figure 6).There is also a discernable increase in adult abun-

Figure 3. Age-0 splittail (>24 mm FL) abundance and distribution based on U.S.Fish and Wildlife Service beach seine survey, 1978–1982, 1992–2002. Data aremean catch per haul by region for May and June. Regions follow Sommer et al.(1997), except for those upstream of the delta: (1) lower Sacramento River(“LowSac.R”—Feather River [river kilometer 129] to American River [river kilome-ter 97]); (2) middle Sacramento River (“MidSac.R.”—Butte Creek [river kilometer222] to Knights Landing [river kilometer 145]); and (3) Upper Sacramento River(“UppSac.R.”—Ord Bend [river kilometer 296] to Colusa State Park [river kilometer239]). Sampling in the latter three regions began in 1981.


dance 2–3 years after high age-0 abundance years,indicating that good year-classes have at least amoderate influence on the adult population size.

Largely because of the variability in age-0abundance, the population trends of splittail havebeen one of the major sources of debate about thespecies. Following the listing of splittail in 1999,the USFWS reopened the comment period for the

final rule on four occasions to seek peer reviewand public comment on various issues, includingabundance trends. The most recent analysis ofpopulation trends in a peer-reviewed journal wasby Kimmerer (2002a), who used a regressionapproach to determine whether freshwater flowor trophic linkages influenced the abundance ofage-0 splittail. He found that the abundance ofage-0 splittail varied positively with freshwateroutflow, but that there was no discernable changein abundance after 1987, the point at which amajor drought began, and the food web in theestuary was substantially modified by the prolif-eration of the introduced clam brackish-watercorbula Corbula amurensis (also known as Peta-mocorbula amurensis). These results are consistentwith an earlier analysis by Sommer et al. (1997),who concluded that there had been no long-termchange in the abundance indices of either adult orjuvenile splittail. Using a conservative 20% levelof statistical significance, the USFWS (2003a)found that almost half of the indices of abun-

30

SOMMER ET AL.

Figure 4. Medium-term trends (late 1970s through 2002)in age-0 splittail abundance, as indexed by three surveys.Asterisks indicate the start of the survey or of lengthrecords for age determination and the most recent indexcalculated (2002). No sampling (NS) and thus no index isindicated by downward arrows. Survey and index calcula-tion methods were as described in Sommer et al. (1997)except for Suisun Marsh, which used a slightly differentapproach: (1) age-0 indices were calculated based onMay–December catch, not just June–August catch (Som-mer et al. 1997); and (2) indices were calculated based onmean catch/trawl for all stations combined, rather thanthe sum of the mean catch for each of seven sloughs(Sommer et al. 1997).

Figure 5. Medium-term trends (late 1970s through 2002)in age-1 splittail abundance, as indexed by two surveys.Asterisks indicate the start of the survey or of lengthrecords for age determination and the most recent indexcalculated (2002). Survey and index calculation methodswere as described for age-0 with different months andsize categories appropriate to age-1.


31


dance showed some evidence of decline; however,they ultimately concluded that this trend isunlikely to lead to extinction.

Factors that May Influence the Abundance and Distribution of Splittail

In its determination of threatened status for split-tail, the USFWS (1999) identified altered hydraulicsand reduced outflow caused by water exports as theprincipal causes of the population decline. Addi-tional threats were listed as direct loss (mortality) atpumping plants and diversions, loss of spawningand nursery habitat as a consequence of drainingand diking for agriculture, reduction in the avail-ability of highly productive brackish-water habitat,urban and agricultural pollution, introducedspecies, and exacerbation of these factors as a resultof 6 years of drought.

Our present understanding of the populationdynamics of splittail is that year-class strength ofthe species is largely determined by the frequency

and duration of floodplain inundation. Of theother factors identified by the USFWS (1999),urban and agricultural pollution and introducedspecies remain potentially major but poorly under-stood threats. There is, however, no evidence tosupport the conclusion that direct loss at pumpingplants has had a significant effect on abundance, atleast since the 1970s. Although diking and drainingof floodplain areas for agriculture have resulted inloss of spawning and nursery habitat, most of thisactivity occurred well before recent observations ofpoor recruitment. These and other factors arereviewed briefly below.

Floodplain inundation

Although much of the historical floodplain in thevalley has been lost to levee construction and riverchannelization, some substantial areas of flood-plain remain in the region. The largest of these arethe Yolo and Sutter bypasses, the primary flood-plains of the Sacramento River. Studies by Sommeret al. (1997) have demonstrated that floodplaininundation represents the primary factor thatdetermines spawning success. They found that theduration of flooding of the Yolo Bypass was strong-ly correlated with splittail year-class strength. Som-mer et al. (1997) also showed that adults moveonto the floodplain in winter and early spring toforage and spawn on flooded vegetation. Afterspawning, adults typically return to the delta,Suisun Bay, and Suisun Marsh to forage during thesummer and fall. Juvenile splittail rear on the YoloBypass floodplain from April through June (Som-mer et al. 2004a). Small-scale floodplain wetlandstudies suggest that young splittail are associatedwith shallow areas (<1 m depth) but that juveniledistribution varies on a diel basis (Sommer et al.2002). Young splittail may become entirely benthicat night, another reason why the inundation oflarge areas of shallow water habitat (thereby creat-ing more benthic resting areas) may help to sup-port high splittail production. Sommer et al.(2004b) found higher abundances of phytoplank-ton, diptera, and terrestrial invertebrates in YoloBypass than the adjacent Sacramento River chan-nel. A major portion of the diet of juvenile splittailis apparently larval chironomids (Kurth andNobriga 2001), which are present in Yolo Bypass at

Figure 6. Medium-term trends (late 1970s through 2002)in age-2+ (adult) splittail abundance, as indexed by twosurveys. Asterisks indicate the start of the survey or oflength records for age determination and the most recentindex calculated (2002). Survey and index calculationmethods were as described for age-0 with differentmonths and size categories appropriate to age-2+.


much higher densities than in that main riverchannels (Sommer et al. 2001b, 2004b). Juvenilesplittail subsequently emigrate to the river chan-nels and estuary as floodwaters recede (Sommer etal. 1997). Observations from the Yolo Bypass areconsistent with studies in the Cosumnes River, anearby undammed watershed that was recentlyidentified as a major spawning and rearing area for splittail (Moyle et al. 2003; Crain et al. 2004).Crain et al. (2004) observed larval splittail rearingin the floodplain during March–May, with peakoccurrence during April and May. As in YoloBypass, young splittail in the Cosumnes River were most commonly associated with submergedterrestrial plants.

Based on these observations, it appears thatsplittail is perhaps the most floodplain-dependentspecies in the San Francisco Estuary (Sommer et al.2001a). It is likely that the decline of splittail dur-ing the 1987–1992 drought was due to the lack ofaccess to floodplain spawning habitat (Sommer etal. 1997). The relatively long life span of the fish(sometimes >5 years) is an important attributethat helps it survive periods without access to thishabitat. Recent studies by Sommer et al. (2002)indicate that the species can be successfullyinduced to spawn in dry years if it is provided withaccess to inundated vegetation. Successful spawn-ing also takes place on vegetated river margins dur-ing years with no floodplain inundation (Moyle etal. 2004; Feyrer et al. 2005).

SWP and CVP water export operations

The two major water projects, the SWP and CVP,seasonally entrain adult and juvenile splittail. Indi-viduals greater than about 20 mm TL are salvagedby fish screening facilities and transported andreleased back to the estuary, although overall sur-vival rates are unknown (Brown et al. 1996).Juveniles are entrained primarily during May–July, and adults are mostly collected duringDecember– March (Meng and Moyle 1995; Som-mer et al. 1997).

Most evidence suggests that entrainment lossesdo not have a major effect on year-class strength.The possible effects of SWP exports were evaluatedby Sommer et al. (1997) using three differentapproaches, none of which provided evidence of

population-level effects. Overall, splittail entrain-ment at the water diversion is correlated withabundance levels. In other words, losses are highestin wet years, when the population is most robust.These trends contrast with delta smelt, a threat-ened species under the ESA, and longfin smeltSpirinchus thaleichthys, a California species of spe-cial concern. Both smelt species were observed atthe water diversion in higher numbers in dry years,when their populations were relatively low. As aconsequence, Sommer et al. (1997) predicted thatthe effects of entrainment on the smelt populationswould be more substantial than for splittail.

Although it is possible that there could be pop-ulation level entrainment effects under some con-ditions (Sommer et al. 1997), there is no evidencethat this has occurred since the 1970s. Nonetheless,entrainment of splittail remains an ongoing issuefor project operations; for example, pumping wascurtailed substantially in 1995 due to the salvage ofmillions of juvenile splittail during a relativelyshort period.

Changes to the food web

The number of alien species introductions to theSan Francisco Estuary has made it perhaps themost invaded estuary in the world (Cohen andCarlton 1998). The food web was changed mostdramatically after the introduction of the clam C.amurensis by the mid-1980s, causing declines inphytoplankton and invertebrates (Jassby et al.2002; Kimmerer 2002a). For splittail, of greatestconcern is a decline in the abundance of Neomysisshrimp, a major food source (Daniels and Moyle1983; Feyrer et al. 2003). However, Kimmerer(2002a) found no evidence that a significantdecrease in Neomysis in 1987 resulted in a signifi-cant decline in age-0 splittail abundance. Nonethe-less, Feyrer et al. (2003) found that food web alter-ations have had a substantial effect on the diets ofolder (age-1 and older) splittail in Suisun Marsh,where splittail populations decreased during thepostclam period. Mysids were formerly the domi-nant prey of splittail but are now are nearly absentfrom gut contents. During the postclam period,splittail diet increasingly focused on bivalves andamphipods. Feyrer and Baxter (1998) found evi-dence that the fecundity of splittail was lower in the

32

SOMMER ET AL.


33


1990s than in the 1980s, lending credence to thehypothesis that C. amurensis had population-leveleffects on splittail by reducing its food supply.

Contaminants

Major pollutants in the San Francisco estuaryinclude a wide variety of chemicals, including heavymetals, pesticides, herbicides, and polycyclic aro-matic hydrocarbons (Nichols et al. 1986). The factthat these compounds vary substantially in time andspace makes it very difficult to predict biologicaleffects; however, many co-occur with vulnerableearly life stages of fishes (Kuivila and Foe 1995;Kuivila and Moon 2004). Toxicity has been docu-mented for a few fishes in the estuary, includingstriped bass Morone saxatilis and Chinook salmon(Saiki et al. 1992; Bailey et al. 1994; Bennett et al.1995), but there is little data for splittail. Under lab-oratory conditions, Teh et al. (2004a) found thatexposure to the widely used pesticide diazinoncaused spinal deformities and decreased growth inyoung splittail. However, little is known about pop-ulation-level effects in the field. Contaminants inthe sediments are potentially the greatest threat tosplittail because these fish are benthic foragers andconsume a large amount of detritus (Daniels andMoyle 1983; Feyrer et al. 2003; Moyle et al. 2004).Perhaps of greatest concern are possible effects ofselenium, which can occur in high concentrations inone of their primary food sources, the clam C.amurensis (Stewart et al. 2004). Because the fish arelong-lived, splittail may accumulate selenium to lev-els that might affect development and survival ofeggs and larvae. Feyrer et al. (2003) found that split-tail diet was largely composed of detritus andbivalves (including C. amurensis) following thedecline in mysid abundance. Feyrer and Baxter(1998) documented lowered fecundity of splittailduring the late 1990s compared to the early 1980s,suggesting that this hypothesis merits further inves-tigation. In another study, results from Teh et al.(2004b) indicate that splittail fed high concentra-tions of selenium grow significantly slower and havehigher liver and muscle selenium concentrations.

Agricultural water diversions

In addition to the large SWP and CVP diversions,the San Francisco Estuary has 2,209 agricultural

diversions in the delta and 366 diversions in SuisunMarsh used for enhancement of waterfowl (Herrenand Kawasaki 2001). These small diversions contin-ue to be a concern for fisheries management as themajority of the structures do not have fish screens.This issue has not been studied in detail; however,the limited evidence suggests that splittail may notbe especially vulnerable to delta agricultural diver-sions. The most intensive study is by Nobriga et al.(2004), who examined entrainment at a diversionnear the confluence of the Sacramento and SanJoaquin rivers during July 2000 and 2001. Theyfound that splittail entrainment was exceptionallylow, just one fish over two intensive sampling peri-ods (M. Nobriga, California Department of WaterResources, personal communication).

Stock recruitment effects

An additional concern for splittail is that a reduc-tion in the number of spawning adults may leadto poor recruitment of young fish to the popula-tion. Sommer et al. (1997) used simple linearregression to test for stock–recruitment effects inseveral different indices of abundance. Datathrough 1995 indicated that there was no signifi-cant stock– recruitment effect for five abundanceindices, but there was a weak statistical relation-ship for the University of California Davis SuisunMarsh survey. As part of the present review, weupdated the analysis using abundance data (log +1 transformed) through 2002 (Figures 4 and 6).For the recent data, there were no statistically sig-nificant relationships between age-2 and olderabundance and age-0 abundance for any of theindices. The lack of strong stock–recruitmentrelationships suggests that environmental factors(i.e., floodplain inundation), not the number ofadults, controls splittail recruitment.

Recreational harvest

Historically a commercial fishery (Walford 1931),splittail continue to be a popular target for a mod-est sport fishery (Moyle 2002; Moyle et al. 2004).The activity primarily occurs from Novemberthrough May, when adult splittail migrate to andfrom spawning habitat. Creel surveys for stripedbass and salmon suggest that up to several hundredadult splittail may be caught on a daily basis.


34

SOMMER ET AL.

Although the USFWS (2003a) concluded that thefishery had little effect on the population, it is pos-sible that sportfishing could affect egg supplybecause nearly all fish are retained, especiallyfemales and the largest individuals.

Restoration

During the mid-1990s, two major restoration effortswere initiated to help address declines in fisheriesresources in the San Francisco Estuary and its water-shed: (1) the Central Valley Project ImprovementAct (CVPIA), and (2) the CALFED program. Cur-rent and planned restoration activities by these twoprograms were identified by USFWS (2003a) as keyreasons for delisting splittail.

Central Valley Project Improvement Act

Enacted by Congress in 1992, the Central ValleyProject Improvement Act (Public Law 102–575)represented the most far-reaching change in CVPoperation since its construction. The most signifi-cant component of this legislation was that itplaced fish and wildlife protection, restoration, andmitigation as project purposes with equal priorityto water diversions for irrigation and domesticpurposes. One major change was that CVPIA real-located 800,000 acre-feet of project yield to be usedfor environmental purposes. A key part of this leg-islation was the Anadromous Fisheries RestorationProgram (AFRP), which sought to double popula-tions of Chinook salmon, steelhead, Americanshad, striped bass, and sturgeon as compared to1967–1991 levels (USFWS 2003b). During1995–2000, the AFRP spent approximately $26 mil-lion on projects to improve fisheries resources andmonitoring (USFWS 2003b). Although this pro-gram does not specifically target splittail, we esti-mate that at least one-third of the funds spent dur-ing 1995–2000 on projects, including habitatprotection, restoration, and fish passage, could havebenefits to this species.

CALFED

While water conflicts in California represent anongoing part of the state’s history, the listing ofwinter-run Chinook salmon and delta smelt in theearly 1990s created dramatic new pressures on

water allocation (Koehler 1995). Specifically, lossesof salmon and smelt became a major operationalconsideration at the SWP and CVP diversions,directly affecting water supply reliability. Severalother species including splittail, longfin smelt,spring-run Chinook salmon, and steelhead werealso being considered for federal listing, raisingconcerns that water diversions would be furthercurtailed. These issues helped bring together manyof the key stakeholders to try to address long-standing resource conflicts in the San FranciscoEstuary and its watershed. The result of thesenegotiations was the formation of CALFED, acooperative effort of more than 20 state and feder-al agencies, including California Department ofWater Resources, California Department of Fishand Game, State Water Resources Control Board,U.S. Bureau of Reclamation, U.S. Fish and WildlifeService, U.S. Geological Survey, and Army Corps ofEngineers. In 2003, CALFED was renamed the Cal-ifornia Bay-Delta Authority (CBDA) after it wasofficially designated as a new state agency.

The CBDA program most relevant to the statusof splittail is the Ecosystem Restoration Program.In the 7 years since this was initiated in 1995, theprogram has invested $335 million in more than300 restoration projects (CBDA 2003). Recent datacollected on splittail had a major effect on habitatrestoration priorities, particularly the importanceof floodplain, shallow water tidal, and riparianhabitat. Projects that targeted these habitat types,or constructed fish screens or passage facilitieswithin the range of splittail, are likely to have the greatest benefits to the species. Since 1995, thetotal expenditure on these categories of projectswas $195 million. The total amount of habitat protected or restored was 45,700 ha, 6,500 ha ofwhich was for floodplain, the habitat type mostlikely to limit splittail abundance. An additional$32 million was spent on improving water andsediment quality and invasive species control thatcould also benefit splittail.

Conclusions

Much of the initial rationale for the listing of split-tail focused on an apparent decrease in range andabundance (USFWS 1994; Meng and Moyle 1995).


35


Over the past decade, it has become clear that therange of the species is much wider than previouslyrealized, encompassing most of the available habi-tat below the major dams. Although there is gener-al agreement that there were substantial losses ofoff-channel habitat and river reaches upstream ofdams during the past century (Meng and Moyle1995; Sommer et al. 1997; USFWS 2003a), data col-lected over the past three decades show no evidenceof a recent range restriction. Moreover, the strongyear-classes produced in the late 1990s demonstratethat the splittail population is much more resilientthan previously understood. The inherent variabil-ity in splittail abundance and the lack of exact pop-ulation estimates have sparked a lively debate aboutwhether there has been at least a modest decline inpopulation levels; however, we concur with USFWS(2003a) that the population has enough resilienceto avoid extinction in the foreseeable future.

While abundance and distribution are oftenprimary considerations in listing decisions, theESA also requires the USFWS to evaluate whetherthere are present or future threats to the habitat orrange of a species. In their recent review, theUSFWS (2003a) identified many of the samepotential threats reviewed in the present paper,including the negative effects of water diversions,contaminants, and invasive species. Despite somepotential threats, we agree with the USFWS findingthat the adverse effects are adequately offset byrecent beneficial actions such as CALFED andCVPIA restoration and cooperative resource man-agement. The scope of these efforts is extensive,representing one of the most ambitious restorationefforts in the United States (Koehler 1995).

From a fisheries perspective, the listing debatefueled new research on splittail and other nativefishes and helped change habitat restoration prior-ities. Until the 1990s, there was little research ormonitoring of native fishes in California other thansalmonids. It is therefore not surprising that theinitial proposal to list splittail assumed that thespecies was declining for reasons similar to othernative fishes, including delta and longfin smelt(USFWS 1994). Since then, splittail has became amajor focus of fisheries research in the San Fran-cisco Estuary, helping to reveal that different nativefishes are responding to different cues (Bennett

and Moyle 1996; Kimmerer et al. 2002b). In partic-ular, it appears that splittail are perhaps the mostfloodplain dependent species in the estuary (Som-mer et al. 2001a), whereas delta and longfin smeltdo not make extensive use of this type of habitat.Splittail-related research has also revealed that thefloodplain is a major nursery habitat for Chinooksalmon and stimulates lower trophic levels in theestuary. This recognition led to a new emphasis onfloodplain restoration in programs such asCALFED that should benefit a suite of aquaticorganisms. Although splittail populations shouldimprove substantially as a consequence of flood-plain restoration, we do not believe that the specieswill ever return to historic levels because of theextensive habitat alteration and large numbers ofalien species in the estuary (Atwater et al. 1979;Mount et 1995; Cohen and Carlton 1998).

The fact that the splittail population has beenable to survive a multitude of stressors over the pastcentury provides a good example of resilience inwestern native fishes. A major reason that it hasbeen able to cope with changes in the physical habi-tat is that splittail are physiologically hardy and ableto tolerate a relatively wide range of temperature,salinity, and dissolved oxygen levels (Young andCech 1996). This tolerance contrasts with severalother native fishes of the San Francisco Estuary,including longfin and delta smelt (Sommer et al.1997). However, physiological hardiness does notadequately explain the resilience of splittail in theface of the large number of species introductions.Unlike many other regions in the west, flood man-agers in California decided to retain substantialparts of the historical floodplain in the lower Sacra-mento Valley (Yolo and Sutter bypasses) to providepassive flood protection for valley communities(Sommer et al. 2001a). Our working hypothesis isthat this large remaining area of floodplain habitatmay be the major reason that the species has notgone extinct. Specifically, the ability to use flood-plain habitat early in the year may provide compet-itive advantages to splittail. Splittail spawn and rearduring winter and spring and are therefore able touse the highly productive seasonal floodplain inun-dated during that period (Sommer et al. 2001a,2004). By contrast, nonnative fishes typically spawnin late spring or summer when the floodplain is


dewatered and the fish are confined to less-produc-tive perennial channels.

Note, however, that these life history characteris-tics will not necessarily protect splittail in the future.For example, it is unclear whether programs such asCALFED and CVPIA are adequate to protect split-tail if there is a repeat of multidecade droughts thatoccurred in the past millennium (Ingram et al.1996). We are also particularly concerned about thepotential spread of northern pike Esox lucius, whichhas recently become abundant in a reservoir on theFeather River (Rischbieter 2000). Unlike most otheralien fishes in the San Francisco Estuary, this highlypredaceous species is able to spawn and maintainrelatively high consumption rates in relatively coldconditions, particularly in shallow, vegetated areaslike Yolo Bypass (Craig 1996). As a result, we believethat it is prudent to retain splittail as a species ofspecial concern under the California EndangeredSpecies Act, and for the USFWS to periodically eval-uate the status of the population (USFWS 2003).

Acknowledgments

Data presented in this study were collected by thededicated staff of the California Department ofWater Resources, CDFG, USFWS, and U.C. Davis.We owe particular thanks to Bruce Herbold foreditorial suggestions. Funding was provided by theInteragency Ecological Program and the CaliforniaBay-Delta Authority.

ReferencesAtwater, B. F., S. G. Conard, J. N. Dowden, C. W. Hedel,

R. L. MacDonald, and W. Savage. 1979. History, land-forms and vegetation of the estuary’s tidal marshes.Pages 347–385 in T. J. Conomos, editor. San Francis-co Bay: the ecosystem. Pacific Division of the Ameri-can Association for the Advancement of Science,San Francisco.

Bailey, H. C., C. Alexander, C. DiGiorgio, M. Miller, S. I.Doroshov, and D. E. Hinton. 1994. The effect of agri-cultural discharge on striped bass (Morone saxatilis).Ecotoxicology 3:123–142.

Baxter, R. D. 1999. Status of splittail in California. Cali-fornia Fish and Game 85:28–30.

Bennett, W. A., D. J. Ostrach, and D. E. Hinton. 1995.Condition of larval striped bass in a drought-strick-en estuary: evaluating pelagic food web limitation.Ecological Applications 5:680–692.

Bennett, W. A., and P. B. Moyle. 1996. Where have all thefishes gone? Interactive factors producing fish declinesin the Sacramento-San Joaquin Estuary. Pages519–542 in J. T. Hollibaugh, editor. San Francisco Bay:the ecosystem. Pacific Division of the American Asso-ciation for the Advancement of Science, San Francisco.

Brown, R., S. Greene, P. Coulston, and S. Barrow. 1996.An evaluation of the effectiveness of fish salvageoperations at the intake to the California aqueduct,1979–1993. Pages 497–518 in J. T. Hollibaugh, editor.San Francisco Bay: the ecosystem. Pacific Division ofthe American Association for the Advancement ofScience, San Francisco.

CBDA (California Bay-Delta Authority). 2003. Ecosys-tem Restoration Program Phase 2 Report. Available:www.ca lwater. ca .gov/Prog r ams/Ecosys temRestoration/ERPProjects_Phase2/Phase2_Report.pdf.

Caywood, M. L. 1974. Contributions to the life history ofthe splittail Pogonichthys macrolepidotus (Ayres). M.S.thesis. California State University, Sacramento.

Cohen, A. N., and J. T. Carlton. 1998. Accelerating invasionrate in a highly invaded estuary. Science 279:555–558.

Craig. J. S. 1996. Pike: biology and exploitation. Chap-man and Hall, London.

Crain, P. K., K. Whitener, and P. B. Moyle. 2004. Use of arestored central California floodplain by larvae ofnative and alien fishes. Pages 125–140 in F. Feyrer, L.R Brown, R. L. Brown, and J. J Orsi, editors. Early lifehistory of fishes in the San Francisco Estuary andwatershed. American Fisheries Society, Symposium39, Bethesda, Maryland.

Daniels, R. A., and P. B. Moyle. 1983. Life history of theSacramento splittail (Cyprinidae: Pogonichthysmacrolepidotus) in the Sacramento–San JoaquinEstuary. Fishery Bulletin 81:647–654.

Feyrer, F., and R. Baxter. 1998. Splittail fecundity and eggsize. California Fish and Game 84:119–126.

Feyrer, F., B. Herbold, S. A. Matern, and P. B. Moyle. 2003.Dietary shifts in a stressed fish assemblage: conse-quences of a bivalve invasion in the San FranciscoEstuary. Environmental Biology of Fishes 67: 277–288.

Feyrer, F., T. R. Sommer, and R. Baxter. 2005. Spatial-tem-poral distribution and habitat associations of age-0splittail in the lower San Francisco Estuary water-shed. Copeia 2005:159–168.

Gobalet, K. W., P. D. Schulz, T. A. Wake, and N. Siefkin.2004. Archeological perspectives on native Americanfisheries of California, with emphasis on steelheadand salmon. Transactions of the American FisheriesSociety 133:801–833.

Harrell, W. C., and T. R. Sommer. 2003. Patterns of adultfish use on California’s Yolo Bypass floodplain. Pages88–93 in I. P. M. Faber, editor. California riparian sys-tems: processes and floodplain management, ecologyand restoration. Riparian Habitat Joint Venture,Sacramento, California.

SOMMER ET AL.

36


37


Herren, J. R., and S. S. Kawasaki. 2001. Inventory of waterdiversions in four geographic areas in California’sCentral Valley. California Department of Fish andGame Fish Bulletin 179:343–355.

Ingram, L. B., J. C. Ingle, and M. E. Conrad. 1996. Iso-topic records of pre-historic river inflow in San Fran-cisco Bay Estuary. Pages 35–61 in J. T. Hollibaugh,editor. San Francisco Bay: the ecosystem. Pacific Divi-sion of the American Association for the Advance-ment of Science, San Francisco.

Jassby, A. D., J. E. Cloern, and B. E. Cole. 2002. Annualprimary production: patterns and mechanisms ofchange in a nutrient-rich tidal ecosystem. Limnologyand Oceanography 47:698–712.

Kimmerer, W. J. 2002a. Effects of freshwater flow onabundance of estuarine organisms: physical effects ortrophic linkages. Marine Ecology Progress Series243:39–55.

Kimmerer, W. J. 2002b. Physical, biological and manage-ment responses to variable freshwater flow into theSan Francisco Estuary. Estuaries 25:1275–1290.

Koehler, C. L. 1995. California’s Bay-Delta agreement: amodel for cooperation. Rivers 5:46–51.

Kuivila, K. M., and C. G. Foe. 1995. Concentrations,transport and biological effects of dormant spraypesticides in the San Francisco Estuary, Calif-ornia. Environmental Toxicology and Chemistry14:1141–1150.

Kuivila, K. M., and G. E. Moon. 2004. Potential exposureof larval and juvenile delta smelt to dissolved pesti-cides in the Sacramento–San Joaquin Delta, Califor-nia. Pages 229–242 in F. Feyrer, L. R. Brown, R. L.Brown, and J. Orsi, editors. Early life history of fishesin the San Francisco Estuary and watershed. Ameri-can Fisheries Society, Symposium 39, Bethesda, Mary-land.

Kurth, R., and M. Nobriga. 2001. Food habits of larvalsplittail. Interagency Ecological Program Newsletter14(2):40–42.

Matern, S. A., P. B. Moyle, and L. C. Pierce. 2002. Nativeand alien fishes in a California estuarine marsh:twenty-one years of changing assemblages. Transac-tions of the American Fisheries Society 131:797–816.

Meng, L., and P. B. Moyle. 1995. Status of splittail in theSacramento–San Joaquin Estuary. Transactions of theAmerican Fisheries Society 124:538–549.

Mount. J. F. 1995. California’s rivers and streams: theconflict between fluvial process and land use. Univer-sity of California Press, Berkeley.

Moyle, P. B. 2002. Inland fishes of California. Universityof California Press, Berkeley.

Moyle, P. B., P. K. Crain, K. Whitener, and J. F. Mount.2003. Alien fishes in natural streams: fish distribu-tion, assemblage structure, and conservation in theCosumnes River, California, U.S.A. EnvironmentalBiology of Fishes 68:143–162.

Moyle, P. B., R. D. Baxter, T. Sommer, T. C. Foin, and S. A.Matern. 2004. Biology and population dynamics ofSacramento splittail (Pogonichthys macrolepidotus) inthe San Francisco Estuary: a review. San FranciscoEstuary and Watershed Science [online serial]2(2):article 4.

Nichols, F. H., J. E. Cloern, S. N. Luoma, and D. H. Peter-son. 1986. The modification of an estuary. Science231:567–573.

Nobriga, M., Z. Matica, and Z. Hymanson. 2004. Evaluat-ing entrainment vulnerability to agricultural irrigationdiversions: a comparison among open-water fishes.Pages 281–295 in F. Feyrer, L. R. Brown, R. L. Brown,and J. J. Orsi, editors. Early life history of fishes in theSan Francisco Estuary and watershed. American Fish-eries Society, Symposium 39, Bethesda, Maryland.

Rischbieter, D. B. C. 2000. Structures to prevent thespread of nuisance fish from Lake Davis, California.North American Journal of Fisheries Management20(3):784–790.

Rutter C. 1908. The fishes of the Sacramento–SanJoaquin basin, with a study of their distribution andvariation. Bulletin of U.S. Bureau of Fisheries27:103–152.

Saiki, M. K., M. R. Jennings, and R. H. Wiedmeyer. 1992.Toxicity of agricultural subsurface drainwater fromthe San Joaquin Valley, California, to juvenile Chi-nook salmon and striped bass. Transactions of theAmerican Fisheries Society 121:78–93.

Sommer, T., R. Baxter, and B. Herbold. 1997. Theresilience of splittail in the Sacramento–San JoaquinEstuary. Transactions of the American Fisheries Soci-ety 126:961–976.

Sommer, T. R., W. C. Harrell, M. Nobriga, R. Brown, P. B.Moyle, W. J. Kimmerer, and L. Schemel. 2001a. Cali-fornia’s Yolo Bypass: evidence that flood control canbe compatible with fish, wetlands, wildlife and agri-culture. Fisheries 26:6–16.

Sommer, T. R., M. L. Nobriga, W. C. Harrell, W. Batham,and W. J. Kimmerer. 2001b. Floodplain rearing ofjuvenile Chinook salmon: evidence of enhancedgrowth and survival. Canadian Journal of Fisheriesand Aquatic Sciences 58:325–333.

Sommer, T., L. Conrad, G. O’Leary, F. Feyrer, and W. Har-rell. 2002. Spawning and rearing of splittail in amodel floodplain wetland. Transactions of the Amer-ican Fisheries Society 131:966–974.

Sommer, T., W. Harrell, G. O’Leary, S. Zeug, R. Kurth,and F. Feyrer. 2004a. Ecological patterns of larval fish-es in a river-floodplain of the San Francisco Estuary.Pages 111–124 in F. Feyrer, L. R. Brown, R. L. Brown,and J. J. Orsi, editors. Early life history of fishes in the San Francisco Estuary and watershed. Amer-ican Fisheries Society, Symposium 39, Bethes-da, Maryland.

Sommer, T., W. C. Harrell, A. Mueller-Solger, B. Tom,


38

SOMMER ET AL.

and W. Kimmerer. 2004b. Effects of flow variation on channel and floodplain biota and habitats ofthe Sacramento River, California, USA. Aquatic Conservation: Marine and Freshwater Ecosystems14:247–261.

Stewart, A. R., S. N. Luoma, C. E. Schlekat, M. A. Doblin,and K. A. Hieb. 2004. Food web pathway determineshow selenium affects aquatic ecosystems: a San Fran-cisco Bay case study. Environmental Science andTechnology 38(17):4519–4526.

Teh, S. J., G. Zang, T. Kimball, and F. C. Teh. 2004a. Lethaland sublethal effects of esfenvalerate and diazinon onsplittail larvae. Pages 243–253 in F. Feyrer, L. R.Brown, R. L. Brown, and J. J. Orsi, editors. Early lifehistory of fishes in the San Francisco Estuary andwatershed. American Fisheries Society, Symposium39, Bethesda, Maryland.

Teh, S. J., X. Deng, D. F. Deng, F. C. Teh, S. S. O. Hung, T.W. M. Fan, J. Liu, and R. M. Higashi. 2004b. Chroniceffects of dietary selenium on juvenile Sacramentosplittail (Pogonichthys macrolepidotus). Environmen-tal Science Technology 38(22):6085–6093.

USFWS (U.S. Fish and Wildlife Service). 1994. Endan-gered and threatened wildlife and plants; proposeddetermination of threatened status for the Sacramen-to splittail. Federal Register 59:862–869.

USFWS (U.S. Fish and Wildlife Service). 1999. Final rule.Endangered and threatened wildlife and plants:determination of threatened status for the Sacramen-to splittail. Federal Register 64(25):5963–5981.

USFWS (U.S. Fish and Wildlife Service). 2003a. Endan-gered and threatened wildlife and plants: Notice ofremanded determination of threatened status for theSacramento splittail (Pogonichthys macrolepidotus).Federal Register 68(183):5140–5166.

USFWS (U.S. Fish and Wildlife Service). 2003b. Anadro-mous Fisheries Restoration Program. Available:http://www.delta.dfg.ca.gov/afrp/projects.asp.

Walford, L. A. 1931. Handbook of common commercialand game fishes of California. Bureau of CommercialFisheries Fish Bulletin 28.

Wang, J. C. S. 1986. Fishes of the Sacramento–SanJoaquin Estuary and adjacent waters, California: aguide to the early life histories. Interagency Ecologi-cal Study Program for the Sacramento–San Joaquin Estuary, Technical Report 9. Department of WaterRe-sources, Sacramento. Available: http://elib.cs.berkeley .edu/kopec/tr9/html/home.html (February2005).

Young, P. S., and J. J. Cech. 1996. Environmental toler-ances and requirements of splittail. Transactions ofthe American Fisheries Society 125:664–678.


The June sucker Chasmistes liorus is one offour extant species of lakesuckers in NorthAmerica. The three other species are the

cui-ui C. cujus of Pyramid Lake, Nevada and short-nose sucker C. brevirostris and Lost River suckerDeltistes luxatus of the Klamath lakes and LostRiver system of Oregon and California (Scoppet-

tone and Vinyard 1991). The Snake River sucker C.muriei, a presumed fifth lakesucker species fromthe upper Snake River system of Wyoming, wasdescribed by Miller and Smith (1981) after itsextinction in the 20th century. The June sucker wasfederally listed as endangered with critical habitatin 1986 (USOFR 1986). Designated critical habitat

39



Status of June Sucker in Utah Lake and Refuges

ABSTRACT The June sucker Chasmistes liorus is endemic to Utah Lake, Utah. Abundant when first described

in the 19th century, the species declined precipitously in the 20th century, leading to it being listed as endangered in

1986. The wild population size at time of listing was estimated to be less than 1,000 and may be even smaller at

present. A multi-partner cooperative program was formally established in 2002 with the dual goals of recovering the

June sucker and allowing continued operation of water facilities for human use. One recovery action of the program

has been collection and artificial propagation of June sucker, yielding more than 46,000 June sucker of varying ages

currently being held outside of Utah Lake. Mature fish held in captivity are beginning to contribute to recovery as they

and their offspring are released into the lake. Dwindling numbers of wild fish combined with the increasing propor-

tions of stocked fish returning to spawn in the Provo River indicates barriers to recruitment that are being addressed

by other program recovery actions. While actions being taken to address environmental threats to June sucker, espe-

cially controlling nonnative fishes and habitat alteration, must continue if artificially and naturally produced June

sucker are to survive in Utah Lake, the ability of this species to thrive and reproduce in habitats outside of Utah Lake

will likely be important to its persistence. Habitat recovery and conservation efforts will be critical for maintaining a

diverse environment where both June sucker and Utah sucker Catostomus ardens can survive. Environmental influ-

ences in Utah Lake appear to have been important for the evolution of sucker feeding habits and the observed mor-

phologies of the two species. June sucker have been kept from going extinct, but should remain listed as endangered.

The goal of this paper is to present information regarding the current status of June sucker and the status of actions

to recover this endangered species, currently dominated by the captive propagation efforts.

Matthew E. Andersen, Christopher J. Keleher, Joshua E. Rasmussen, Eriek S. Hansen,Paul D. Thompson, David W. Speas, M. Douglas Routledge, and Trina N. Hedrick

MATTHEW E. ANDERSEN U.S. Geological Survey, Grand Canyon Moni-toring and Research Center, 2255 North Gemini Drive, Flagstaff, Arizona86001, USA, [email protected].

CHRISTOPHER J. KELEHER Utah Department of Natural Resources,1594 West North Temple, Suite 3710, Salt Lake City, Utah 84114-6301,USA, [email protected].

JOSHUA E. RASMUSSEN Utah Division of Wildlife Resources, CentralRegion, 115 North Main Street, Springville, Utah 84663, USA,[email protected].

ERIEK S. HANSEN 1320 North 200 East, #232, Logan, Utah 84321,USA, [email protected].

PAUL D. THOMPSON Utah Division of Wildlife Resources, NorthernRegion, 515 East 5300 South, Ogden, Utah 84405, USA,[email protected].

DAVID W. SPEAS U.S. Bureau of Reclamation, 125 South State Street,UC-732, Salt Lake City, Utah 84138-1147, USA, [email protected].

M. DOUGLAS ROUTLEDGE Utah Division of Wildlife Resources, FisheriesExperiment Station, 1465 West 200 North, Logan, Utah 84321, USA,[email protected].

TRINA N. HEDRICK Utah Division of Wildlife Resources, 1594 WestNorth Temple, Suite 2110, Salt Lake City 84114-6301, USA,[email protected].


ANDERSEN ET AL.

is the lower 7.8 km of the Provo River (Figure 1). AJune sucker recovery plan was approved in 1999(USFWS 1999).

June sucker were first collected from Utah Lakeby scientists of the Wheeler expedition of the 1870s(Cope and Yarrow 1875) and were formallydescribed by Jordan (1878). The species is endem-ic to Utah Lake and its tributaries. Some firstaccounts of June sucker described a very largespawning population (Jordan 1878; Jordan andEvermann 1903; Carter 1969; Heckmann et al.1981) that supported large harvests of the fish byaboriginal humans as well as settlers of Europeandescent from the east (Carter 1969). The annualspawning run of this species is known today onlyfrom the Provo River.

June sucker was one of 12 native fish speciesfound in Utah Lake at the end of the 19th century.The historic ichthyofauna of Utah Lake alsoincluded Utah chub Gila atraria, leatherside chubSnyderichthys copei, speckled dace Rhinichthys

osculus, longnose dace R. cataractae, Bonnevilleredside shiner Richardsonius balteatus hydroflox,least chub Iotichthys phlegethontis, Utah suckerCatostomus ardens, Bonneville cutthroat troutOncorhyncus clarkii utah, mountain whitefishProsopium williamsoni, mottled sculpin Cottusbeldingii, and Utah Lake sculpin Cottus echinatus.Miller and Smith (1981) suggested that contempo-rary suckers of Utah Lake resulted from hybridiza-tion events between Utah sucker and June suckerwithin the past 80 years, especially during droughtconditions of the 1930s. Hybridization betweenJune and Utah suckers has been noted by modernworkers, though the timing of the initiation ofhybridization is uncertain (Li 1999). Recentmolecular analysis suggests that June and Utahsucker share a common evolutionary history, hav-ing diverged into the modern forms from a com-mon progenitor in recent geologic history (Mocket al. 2006). The two sucker species recognized asoccurring in Utah Lake today are June sucker and

40

Figure 1. Selected localities throughout Utah.


Utah sucker (USFWS 1999; Belk et al. 2001;SWCA 2002).

June sucker are so named because their annualspawning run often occurred in June (Miller andSmith 1981; Shirley 1983; Radant et al. 1987).Recent information suggests that spawning runsnow occur from mid-April to early May (UtahDivision of Wildlife Resources, unpublished data)perhaps in response to altered hydrologic condi-tions resulting from water development (Figure 2).Wild June sucker mature in about their fifth year oflife, based on opercles taken from spawning Junesucker accidentally killed (Belk 1998). Evidencefrom a small number of individuals suggests apotential life span of more than 40 years (Scoppet-tone and Vinyard 1991; Belk 1998).

Water Development History

In March 1849, settlers established the first colonyalong the Provo River, called Fort Utah, and con-structed the first water diversion structure on theriver. The “Bean Ditch” irrigation canal providedwater for more than 80 ha of crops (USBR 1989).By 1850, several larger diversions were constructednear the mouth of Provo Canyon approximately 17km upstream from Utah Lake. Water-propelledindustry, such as sawmills, became common. In1853, the first irrigation company was formed andwas allowed to remove up to half the water of theProvo River.

The average annual inflow (1951–1990) toUtah Lake from all sources is about 898 millioncubic meters (898 x 106 m3). Of this, 428 x 106 m3

is discharged to the Jordan River (only naturaloutlet of Utah Lake; Figure 1) and about 470 x 106

m3 is lost to evaporation. In 1872, a low dam wasplaced across the lake outflow to the Jordan River,changing the function of Utah Lake to a storagereservoir. A pumping plant was built in 1902 sothat the lake could be lowered below the outlet ele-vation. The pumping plant has been modified andenlarged several times. Its present capacity is about31.5 m3/s, and it can lower the lake level 2.4–3 mbelow the “compromise elevation” of 1,368.26 mmean sea level (msl; Utah Division of WaterResources 1997). The compromise elevation is atarget managed lake elevation that the interested

water authorities have agreed to try to maintain.The surface elevation of Utah Lake fluctuates, onaverage, about 1.2–1.5 m in a given year. Over sev-eral years, the surface elevation can fluctuateapproximately 3.4 m. At compromise elevation,the lake has an average depth of 2.9 m and a max-imum depth of 4.2 m (Fuhriman et al. 1981).Approximately 2,590 m of the lower Provo River isinfluenced by the lake level when it is at compro-mise elevation (Keleher et al. 1999).

Deer Creek Reservoir, the principle feature ofthe Provo River Project, was constructed by theBureau of Reclamation in 1941 about 25 kmupstream of Utah Lake (Figure 1). It has an activestorage capacity of 188 x 106 m3 and is operated bythe Provo River Water Users Association. Approxi-mately 149 x 106 m3 of Provo River water is storedin Deer Creek Reservoir. The reservoir also storeswater imported from adjacent Weber and Duch-esne River drainages (Utah Division of WaterResources 1997).

Jordanelle Reservoir, approximately 17 kmupstream from Deer Creek Reservoir, was first filledto capacity in 1996 (Figure 1). It has a storage capac-ity of 460 x 106 m3 and is operated by the CentralUtah Water Conservancy District as a component ofthe Bonneville Unit of the Central Utah Project(Utah Division of Water Resources 1997).

STATUS OF JUNE SUCKER

41

Figure 2. Provo River stream flows (in cubic feet per second) for average water years. Number in parenthesesin legend refers to the number of average water yearsincluded in that time period. Deer Creek Reservoir wasconstructed in 1941, and the June sucker flow workgroup began meeting in 1994.

Provo River Streamflows(Average Water Years Only)


ANDERSEN ET AL.

Historic use of Provo River water in Utah Valleyfor agriculture and municipal purposes by increas-ing numbers of settlers resulted in the loss of anextensive wetland at the confluence of the river andthe lake (Wakefield 1933), channelization of theriver, and armoring of the banks. Extended droughtin the 1930s decreased the area and volume of UtahLake (Miller and Smith 1981). Altered flow regimesin the lower Provo River, along with decreased avail-able spawning habitat as a result of river channeliza-tion and installation of diversion dams that blockmigration, increase the potential for hybridizationwith Utah sucker. Lowest flows of the lower ProvoRiver in recent times were observed in the early1980s. The natural spring runoff peak has beenaltered due to upstream reservoir storage (Radant etal. 1987). A cooperative effort is underway amongwater authorities and providers to restore the timingand duration of a more natural hydrograph byacquiring and releasing water to mimic historicflows. Agencies continue to refine flow recommen-dations and acquire necessary water to maximizehabitat for June sucker spawning.

Threats to June SuckerNonnative fish species

Nonnative fish were introduced into Utah Lake,starting with black bullhead Ameiurus melas in 1871and common carp Cyprinus carpio in 1881, as foodsources in part because overharvest had so dramati-cally reduced the supply of native fish species (Popov1949). Sport species were also introduced through-out much of the 20th century, and although not allspecies have persisted, a large suite of nonnative fishspecies is now present in Utah Lake. Most arethought to directly or indirectly impact June sucker,including goldfish Carassius auratus, common carp,red shiner Cyprinella lutrensis, fathead minnowPimephales promelas, black bullhead, channel catfishIctalurus punctatus, western mosquitofish Gambusiaaffinis, white bass Morone chrysops, green sunfishLepomis cyanellus, bluegill L. macrochirus, large-mouth bass Micropterus salmoides, black crappiePomoxis nigromaculatus, yellow perch Percaflavescens, and walleye Sander vitreus (SWCA 2002).The most recent thorough inventory of the UtahLake fish community was conducted in 1978 and

1979 (Radant and Sakaguchi 1981; Table 1). Of 17species sampled, common carp were the most abun-dant fish collected, representing 66.2% of the catchby number and 90.9% by weight. Common carpwere followed by white bass (26.7% by number,4.2% by weight), walleye (2.0% by number, 2.6% byweight), and black bullhead (1.4% by number, 0.7%by weight). All other fish species collected represent-ed less than 1% both in number and weight of totalcatch. The only native species collected, excluding asingle Utah chub, were June sucker and Utah suckerthat combined represented 0.4% of the total numbercaptured and 0.5% of weight (Table 1). White bassand common carp have the greatest negative impactsthrough direct and indirect effects on June sucker(SWCA 2002).

Population declines

Habitat losses from natural (Miller and Smith1981) and anthropogenic (Wakefield 1933; Radantet al. 1987) causes, introduction of nonnative fish(Table 1), and overharvest (Carter 1969) appear tohave initiated large-scale population declines of theJune sucker (Modde and Muirhead 1994; USFWS1999; Belk et al. 2001). Modde and Muirhead(1994) concluded that recruitment failure of Junesucker in Utah Lake was not attributable to poorreproductive success. June sucker raised in protect-ed locations and released into Utah Lake return tospawn in the Provo River thereby providing furtherevidence that the major impediment to recruit-ment and eventual recovery of the species is lowsurvival of early life stages.

June sucker numbers, counted during the annu-al spawning run, were estimated at below 1,000adults in 1986 when the species was listed asendangered (USFWS 1999). As the need for feder-al listing became apparent, the threat of extinctionof June sucker seemed imminent. Scoppettone andVinyard (1991) stated in their review of the life his-tory and management of the four endangered lake-suckers “…the June sucker population may havedeclined so low that extraordinary efforts will berequired to avert extinction.” Initial efforts to pre-vent extinction included streamside in vitrospawning and taking of fertilized eggs into a hatch-ery for hatching and rearing (UDWR 2004).Opportunistic actions to rear June sucker were

42


taken as additional protected locations becameavailable. Research was initiated in an effort to bet-ter understand the ecology and life history of thespecies and threats to its continued existence.

Recovery ActivitiesRecovery implementation program

A multi-partner June Sucker Recovery Implementa-tion Program (Program) was organized and final-ized in 2002 to address recovery actions described inthe 1999 Recovery Plan. The Program partners haveagreed to pursue recovery of June sucker whileallowing for continued water use to benefit humans.Actions being implemented by the Program addressthe following recovery elements:

a. Nonnative and sportfish management,b. Habitat development and maintenance,c. Water management and protection,d. Genetic integrity and augmentation,e. Research, monitoring, and data

management,f. Information and education.

The Program has conducted feasibility studiesto improve habitat and address nonnative speciescontrol. Concepts for addressing these concernshave been developed and habitat enhancement,along with control of problematic nonnative fish,are expected to reduce or eliminate the recruitmentbottleneck. Improving conditions that promotesurvival of larvae and young-of-year (YOY) fish isan effort to complete the life cycle within the UtahLake system so that natural production can aug-ment the wild population. The transition fromconceptual plans to implementation requiresaddressing multiple challenges, including securingsufficient funds, identifying willing landowners tofacilitate habitat modifications and/or purchases,the logistical and economic challenges associatedwith controlling nonnative fish, and resolving con-flicts with sportfishing interests.

Hatchery production, stocking, and research

To date, captive propagation has been the most vis-ible and active recovery action and is the primarysource of June sucker released into the wild. June


43

Table 1. Relative abundance by number and weight of adult fish collected in 1978 and 1979 inventory of Utah Lake (Randant and Sakaguchi 1981).

Species Total number Percent by number Percent by weight

Common carp 22,717 66.2 90.9

White bass 9,163 26.7 4.2

Walleye 687 2.0 2.6

Black bullhead 484 1.4 0.7

Channel catfish 274 0.8 0.7

Largemouth bass 48 0.1 0.2

Bluegill 109 0.3 0.1

Yellow perch 102 0.3 0.1

Fathead minnow 553 1.6 T

June sucker and Utah sucker 126 0.4 0.5

Black crappie 5 T T

Golden shiner Notemigonus crysoleucas 14 T T

Brown trout Salmo trutta 3 T T

Rainbow trout Oncorhynchus mykiss 2 T T

Green sunfish 3 T T

Redside shiner 1 T T

Utah chub 1 T T

T = less than 0.1%.


ANDERSEN ET AL.

sucker produced by in vitro spawning of runningripe adults in the Provo River have been brought tothe Utah Division of Wildlife Resources’ FisheriesExperiment Station (FES) for rearing. Human-mediated propagation has also been conducted atFES. The original FES facility for June sucker wassmall, 72 m2, had only cool water of 15.6°C, andhad become increasingly crowded. Standard troutfeeds readily available at the station were initiallyfed to these fish. Though the exact cause or causeswere not known, the condition of these captive fishdeteriorated, as indicated by the presence of visibledeformities and as quantified by calculation of thehealth condition profile (HCP) of Goede and Bar-ton (1990). A new facility for June sucker at FES wascompleted in 2001, providing additional space (252m2) and warmer water (18.5°C). Survival of Junesucker hatched in the new facility has improved ascompared to the original facility. Artificial feeddevelopment and artificial propagation experi-ments have been initiated in the new facility. Anadditional June sucker facility at FES initiatedoperation in 2006.

While other recovery actions are beingresearched and implemented, the Program contin-ues to spawn and/or rear June sucker in refuges forrelease into Utah Lake to augment the wild spawn-ing population, support survival in native habitat,and provide fish for research and monitoringefforts. A management plan for June sucker in cap-tivity was completed in 2004 (UDWR 2004). Thisplan, intended to support genetic managementdecisions, including maintaining available geneticdiversity, is being revised with newly availablemolecular data in 2006. The Program has fundedresearch on the genetic make-up of the wild Junesucker and Utah sucker populations in the wild andthe June sucker held in refugia (Mock et al. 2006).Additional research includes investigating ecologicalneeds of young June sucker, responses of June suck-er to habitat variables, ecological and morphologicdifferences of suckers from Utah Lake, and heri-tability of morphologic traits. Propagation and aug-mentation, the recovery actions with the longestrecord of implementation at this time, are begin-ning to exhibit measurable results. The goal of thispaper is to describe the current status of June suck-er and the status of actions to recover this endan-

gered species, currently dominated by the captivepropagation efforts.

Study SitesNatural habitats

Utah Lake.—Utah Lake covers approximately38,400 ha and is located in Utah County, Utah,about 65 km south of the Great Salt Lake (Figure 1).The lake has an average depth of only 2.8 m and amaximum depth of 4.2 m at the compromise eleva-tion of 1,368 msl. It is approximately 38 km longand 21 km wide. Utah Lake has a large area to depthratio and frequent winds prevent thermal stratifica-tion. Scouts from the expedition of Father Escalantein 1776 described a shoreline that included broadpastures, marsh communities with reeds and marshgrasses, and abundant swamps. Settlement of theterritory by humans of European descent started in1847, and over time, nearshore areas were diked,drained, and filled to provide land for agriculturalproduction and grazing, recreational facilities, and amunicipal airport. Historically, the lake had relative-ly stable water levels and was likely less turbid with adeeper littoral zone. Stable lake levels along withclear water and a lake perimeter of productive wet-land habitats and macrophyte beds provided nurs-ery habitat commonly used by native fishes (USFWS1999). Increased nutrient loading from urbaniza-tion, increased lake level fluctuations from down-stream water demands and upstream storage, andintroduction and establishment of common carp,which uproot aquatic vegetation and resuspend bot-tom sediments through foraging, have contributedto the homogeneous, turbid lake devoid of aquaticvegetation that exists today.

Provo River.—Riverine habitat used by spawn-ing adult June sucker and developing larvae wasprobably more extensive historically than today(Radant and Sakaguchi 1981). Prior to settlementof Utah Valley, several large tributaries, includingthe Provo and Spanish Fork rivers and HobbleCreek (Figure 1), provided diverse habitats inbraided, slow, meandering channels. The naturalriver systems provided warm, slow water pools andmarsh habitats suitable for enhanced larval devel-opment and refuges from predation by larger fish-es (historically Bonneville cutthroat trout). River

44


channelization, dredging, irrigation depletions,and upstream water storage have severely impactedthese tributaries, resulting in significantly reduced,less complex habitats. Currently, the only habitatused by spawning June sucker is the lower approx-imately 4.8 km of the Provo River up to an irriga-tion diversion that is a barrier to upstream fishmovement in all but wet years. In very high wateryears, an additional 3 km above this diversionbecomes accessible (USFWS 1999; Figure 1).

Captive Rearing SitesSpringville State Fish Hatchery

The Utah Division of Wildlife Resources (UDWR)operates a state fish hatchery at Springville, UtahCounty, Utah, approximately 10 km from the conflu-ence of the Provo River and Utah Lake (Figure 1). Thefacility is primarily for rainbow trout production andincludes indoor tanks and outdoor concrete raceways.Fertilized June sucker eggs were first brought to thisfacility in 1982 (Shirley 1983), and larvae were raisedin a wetland pond on the hatchery complex, then weretransferred to Camp Creek Reservoir in 1987.

Utah Correctional Institution

Temporary tanks were used for rearing June suck-er at the Utah Correctional Institution (UCI) inDraper, Salt Lake County, Utah from 1987 to 1993.The UCI offered a natural groundwater supply anda protected rearing area in which to raise fish,although care and survival of fish were variable atthis facility. Remaining June sucker were removedfrom the UCI in 1993 and transferred to the Fish-eries Experiment Station (FES).

Brigham Young University

Laboratories in the Widstoe Building at BrighamYoung University in Provo, Utah County, Utah(Figure 1) received and hatched fertilized Junesucker eggs from 1986 to 1991. All surviving fishwere transferred to FES.

Utah State University; Natural Resources Building and Millville

Utah State University (USU) in Logan, Cache Coun-ty, Utah has a laboratory in the Natural ResourcesBuilding (NR 111) where June sucker have been

hatched, reared, and held. No June sucker are held inthis laboratory today. June sucker have also beenheld in an outdoor research facility with man-madeponds maintained by USU in the nearby town ofMillville (Figure 1). A small number of June suckerare held at the Millville ponds today.

Wahweap State Fish Hatchery

Wahweap State Fish Hatchery (Wahweap) is aUDWR facility located on Wahweap Creek, anephemeral tributary to the Colorado River, nowdammed immediately below the confluence by theGlen Canyon Dam to form Lake Powell. This facilityis 3.3 km northwest of Big Water, Kane County, Utahand approximately 50 km northwest of Page, Ari-zona (Figure 1). The hatchery sits on 107 ha ofUDWR land with 35 ponds, including 27 lined and 8earthen ponds. Surface area of the ponds rangesfrom 0.1 to 0.16 ha. The facility has a patented rightto 0.102 m3/s of water from two wells located on thehatchery property.

Fisheries Experiment Station

June sucker have been held at the UDWR FisheriesExperiment Station (FES) in Logan, Utah sinceAugust 1991 (Figure 1). The FES is used for sportand native fish production, research, and fish healthmanagement in addition to June sucker researchand production. The site is on 35 ha and is suppliedby 16 artesian wells with a total flow of 127.4 L/sand water temperature range of 12–18.5°C. Junesucker have been held in fiberglass circular tanksand rectangular troughs in a 6 x 12 m Quonset hutsince 1991. The tanks are fed by a well with 15.6°Cwater that is passed through degassing columns thatutilize Koch-Glitsch cascade mini-rings to removenitrogen gas and increase dissolved oxygen. InDecember 2001, an additional facility was con-structed to meet increased production require-ments. The building is a 12 x 21 m metal structurehousing 37 fiberglass circular tanks of various sizesfor holding family lots, eight fiberglass troughs forhatching eggs and initial rearing, and 15 fiberglasstroughs for conducting research. The facility utilizesa 2 x 2 m quarantine building housing 10 aquaria.During construction, the water source was switchedto a well with 18.5°C water that incorporated a low-head oxygenation system (Water Management


45


ANDERSEN ET AL.

Technology, Inc., Baton Rouge, Louisiana) with aliquid oxygen supply. This system replaced thedegassing columns in the Quonset hut. The needfor additional rearing space for June sucker hasbeen identified, especially to provide fish for releaseinto Utah Lake. National Environmental Policy Actcompliance efforts identified FES as the preferredalternative for this additional facility, which initiat-ed operation in 2006. The new facility utilizesrecirculation technology to raise water tempera-tures in an effort to improve growth rates, reducesusceptibility to disease, increase longevity, andproduce progeny lots for release into Utah Lake orother locations as needed.

Refuge SitesArrowhead and Teal ponds

Arrowhead and Teal ponds at the Ogden NatureCenter are located in Ogden, Utah, approximately63 km north of Salt Lake City (Figure 1). Surfacearea of both ponds is approximately 0.2 ha, andmaximum depth is less than 2 m. The level of theponds is maintained by groundwater that preventscomplete winter freezing. Approximately 20 Junesucker are held at this site as an emergency back-upstock. The site’s limited size restricts the number offish that can be held there.

Red Butte Reservoir

Red Butte Reservoir is located on Red Butte Creek,approximately 2.5 km northeast of Salt Lake City(Figure 1). Red Butte Reservoir has a surface area ofapproximately 5.2 ha and a maximum depth of 10m. The reservoir was constructed in 1930 to supplywater for Fort Douglas, a U.S. Army base that wasdecommissioned in 1991. The federal governmentretained ownership of the dam and reservoir fol-lowing decommmissioning, and lands around RedButte Reservoir are managed by the U.S. ForestService as a research natural area. Because the sur-rounding land management provided increasedsecurity for the reservoir with no competing fish-ery issues, captive June sucker were introduced forholding in 1992. Bonneville cutthroat trout liveand reproduce upstream of the reservoir in RedButte Creek and can occasionally be found in thereservoir. The introduced population of June

sucker unexpectedly reproduced in the reservoirand recruited several year-classes starting inapproximately 1995. This site has potential to serveas a long-term refuge for June sucker, so efforts tosecure the area for this purpose and to repair thedam to meet State of Utah safety standards wereinitiated (U.S. Army Corps of Engineers and U.S.Fish and Wildlife Service 2003). The Central UtahWater Conservation District assumed ownershipand operation of Red Butte Reservoir in 2004. In2005, as the reservoir was being drained to allowfor dam repairs, approximately 9,000 June suckergreater than 150 mm were transferred to UtahLake. Salvaged June sucker smaller than 150 mmwere taken to holding ponds constructed in 2004on the Ensign Ranch in Box Elder County, Utahuntil they can attain lengths greater than 150 mmthought to help them avoid predation.

Ensign ponds

Three 0.4-ha ponds were constructed on a privateranch, the Ensign Ranch, in 2004 to benefit Junesucker and other native fishes. These ponds are fedby mountain springs piped approximately 1 km toa valley depression. They were first used to receiveJune sucker transferred from Red Butte Reservoirin 2005.

Camp Creek Reservoir

Camp Creek Reservoir is located in western BoxElder County, Utah, near the town of Etna,approximately 358 km northwest of Salt LakeCity. The reservoir is located on Camp Creek,which originates in Nevada and flows east intoUtah. The surface area of the reservoir is approx-imately 2 ha, and maximum water depth is 5 m. Itis privately owned and used for irrigation(Thompson 2001). A variety of age-classes of Junesucker have been introduced here, beginning in1987, and they have reproduced. In order tomaintain healthy fish in Camp Creek Reservoirand to support recovery efforts the June suckerpopulation is thinned every 1 to 2 years, and theyare transferred to and released in Utah Lake.Approximately 1,000 June sucker were capturedin Camp Creek Reservoir, passive integratedtransponder (PIT)-tagged, transported, andreleased in Utah Lake in 2004 and 2005.

46


MethodsEmergency actions

The primary goal of biologists who first engaged inconservation actions to protect June sucker was toprevent immediate extinction. Resource agencies,with limited funds and facilities, initiated salvageoperations in 1982 by in vitro spawning of ripeadults from the spawning run and taking of fertil-ized eggs into protected locations. The SpringvilleHatchery was the first facility to receive fertilizedeggs for hatching and rearing. Between 1982 and1993, June sucker were either taken as fertilized eggsfrom the Provo River or transferred fromSpringville as juveniles and held at Utah State Uni-versity, Brigham Young University, Utah Correc-tional Institution, Camp Creek Reservoir, andArrowhead and Teal ponds. Disease considerations,the availability of only cool water, and low percent-age hatching and larval survival led to the search foralternatives to Springville Hatchery. This searchgradually focused increasing attention on thepotential for rearing June sucker at FES where addi-tional facilities and full-time staff were available.Fish were first transferred to FES in 1991 (Table 2).Streamside matings were conducted for a few yearsduring the 1980s, then reinitiated in 1994 withtransfer of fertilized eggs to FES for hatching andrearing. Consistent annual monitoring of the Junesucker spawning run was also initiated in 1994.

Refuges

Multiple populations currently provide refuges forthe species in case of catastrophic loss in the wild.Today, captive June sucker are held at FES, CampCreek Reservoir, Ensign Ponds, Arrowhead Pond,and USU Millville (Table 3). Dam repairs werecompleted, and June sucker were re-introducedinto Red Butte Reservoir in October 2006.

Habitat enhancement and nonnative control

The Program is developing approaches for habitatenhancement by targeting areas believed impor-tant to early life stages of June sucker, associatedprimarily with historic tributary flood plains andtheir deltaic interface with the lake. Most areas cur-rently targeted are privately owned, and the Pro-

gram is committed to working with willing sellers.The Program is also developing approaches fornonnative fish control, including removal of targetspecies, especially common carp and white bass(SWCA 2002).

Broodstock development

The new native fish facility at FES was completedin 2001 and now holds some of the old June suck-er stock and all of the newly received stock. Junesucker currently held at Camp Creek Reservoirmay also prove to be of desirable genetic back-ground and used as broodstock for artificial cross-es. Fish reared in excess of minimum numbersneeded to maintain broodstock are released fromFES either into Utah Lake or into other refuges.

June sucker held at FES are segregated into lots,each the result of a cross of two individual fish.Individuals in the oldest lots have all been PIT-tagged and are held together. They are identified bythe year in which they were produced, except forthose fish that were captured as drifting larvae andare likely the result of more than one fertilizationevent. Collectively, these lots are the brood lots(i.e., parent fish crossed to yield progeny lots;UDWR 2004). Progeny lots are the offspring des-tined for release into Utah Lake and/or protectedrearing locations. Progeny lots will be held untilthey reach a size, currently thought to be 150 mm,which allows them to avoid the majority of thepredatory fish present in Utah Lake. After reaching150 mm, the progeny lots will be released into UtahLake. Current research is investigating the availablegenetic diversity of brood lots and will be used tohelp identify the most appropriate fish for progenylot production.

Propagation research

Crosses of broodstock at FES to produce progenylots were initiated in 1998 (Table 2). A release of 20June sucker progeny fish (offspring spawned in2000 from two fish captured as adults, one in1989, one in 1991) from FES was made in 2003.Since molecular markers are not yet available,biologists attempt to maximize available geneticdiversity by conducting crosses using individualsfrom two different year-classes (UDWR 2004).The genetic diversity of fish held in other refuges


47


48

ANDERSEN ET AL.

Lot number Number on hand Fish/lb Length (in) Gram/fish Length (mm)

89SKJNUSU Brood lot 17 0.72 15.38 629.99 390.65

91SKJNBYU Brood lot 27 0.72 15.38 629.99 390.65

91SKJNUSU Brood lot 19 0.72 15.38 629.99 390.65

92SKJN Brood lot 8 0.72 15.38 629.99 390.65

93SKJNlot2 Brood lot 20 0.72 15.38 629.99 390.65

94SKJNLot-4 Brood lot 87 0.72 15.38 629.99 390.65





Mixed lot Lost tags, 6 0.72 15.38 629.99 390.65lot unknown

Totals 10 296 0.72 15.38 629.99 390.65

990618SKJNPR01 Brood lot larval fish 37 1.19 14.10 381.17 358.14from Provo River 1999

000509SKJNPR01 Brood lot 245 1.41 12.57 321.70 319.28

000525SKJNPR04 Brood lot 84 1.59 11.82 285.28 300.23

000525SKJNPR05 Brood lot 155 2.70 9.99 168.00 253.75

000601SKJNPR07 Half sib hatchery 273 1.06 14.05 427.92 356.87male

000527SKJNFE01 Progeny lot 8 1.19 14.10 381.17 358.1489SKJNUSU female 91SKJNUSU male C16

Totals 5 765 1.59 12.51 316.81 317.65

010424SKJNPR01 Brood lot 141 1.21 14.02 374.87 356.11

010426SKJNPR02 Brood lot 53 1.77 12.35 256.27 313.69

010502SKJNPR03 Brood lot 276 2.23 11.10 203.40 281.94(Feed Study #1)

010516SKJNPR05 Brood lot 305 2.25 11.40 201.60 289.56

010516SKJNPR06 Brood lot 352 1.19 14.10 381.17 358.14

Totals 5 1,127 1.73 12.59 283.46 319.89

020430SKJNPR01 Brood lot 23 4.14 9.58 109.56 243.33

020501SKJNPR02 Brood lot 4 2.15 11.91 210.97 302.51

020506SKJNPR03 Brood lot 90 5.64 8.64 80.42 219.46


020521SKJNPR06 Brood lot 401 1.63 13.07 278.28 331.98

020521SKJNPR07 Brood lot 410 2.14 11.93 211.96 303.02

020528SKJNPR08 Brood lot 305 2.18 11.86 208.07 301.24

020604SKJNPR10 Brood lot larval fish 319 2.72 11.02 166.76 279.91from Provo River 2002

Totals 8 1,847 2.78 11.25 193.04 285.72

Table 2. June sucker lots held at Fisheries Experiment Station as of October 1, 2005.

(continued)


49


Lot number Number on hand Fish/lb Length (in) Gram/fish Length (mm)


030528SKJNPR22 Brood lot 438 9.82 7.18 46.19 182.37

030604SKJNPR25 Brood lot 47 6.96 8.05 65.17 204.47

030616SKJNFE01 Progeny lot 337 6.26 8.34 72.46 211.84



Totals 6 2,361 7.71 7.97 62.50 202.31

040507SKJNPR01 Brood lot 4 24.78 5.27 18.30 133.86

040612SKJNFE01 Progeny lot* 852 61.52 3.90 7.37 99.06

040613SKJNFE03 Progeny lot 256 23.78 4.42 19.07 112.27(Feed Study #4)*

040613SKJNFE04 Progeny lot* 983 26.83 5.14 16.91 130.56(Temperature Study #1)


Totals 7 2,164 54.70 4.34 13.00 110.34

Tag retention Mixed lot 86 2.13 11.95 212.95 303.53

050620SKJNFE26 Progeny lot ~ 3,529 1,624.68 1.31 0.28 33.27


050621SKJNFE29 Progeny lot ~ 120 1,556.64 1.33 0.29 33.78














Totals 16 38,284 1,613.44 1.33 0.29 33.69

Grand totals 42 46,967 168.60 9.14 209.32 232.19

~ Numbers estimated by percentage at last passive integrated transponder tag reading.

* Estimate no total inventories taken to date.



50

ANDERSEN ET AL.

is also not well known at this time. Additionalmolecular and morphologic analyses are beingconducted to quantify diversity available in captiv-ity and to describe genetic structuring of theknown population, including wild and captiveJune sucker.

Studies to evaluate feed regimes for rearingcaptive June sucker are continuing (Hansen2002, 2003a). Experiments are also being con-ducted to increase gamete yield from hatcheryfish (Table 4). Attempts to induce spawning in

FES June sucker broodstock using hormoneshave occurred since 1998, although success inspawning female June sucker has been limited.In 2002, a study to evaluate human chorionicgonadotropin dosage levels resulted in noinjected females ovulating. Possible causes forlack of ovulation were increased water temper-ature 5 months prior and a diet change 4months prior to hormone injection. In 2003, astudy was initiated to evaluate temperaturerequirements and an additional hormone,

Table 3. June sucker stocked out from the Fisheries Experiment Station.

Date Lot number Number stocked Kilograms stocked Fish/kg Grams/fish Stocking/transfer location

1992–1993 89SKJN-USU 1,487 – – – UCI, USU, Red Butte

1992–1993 91SKJN-BYU 13 – – – USU

04/26/94 89SKJN-USU 76 7.85 9.68 103.25 Ogden Nature Park

04/26/94 89SKJN-USU 544 55.70 9.77 102.39 Ogden Nature Park

04/26/94 910523SKJNUL01 707 34.02 20.78 48.12 Ogden Nature Park

10/11/94 89SKJN-USU 146 21.29 6.86 145.84 Utah Lake


10/11/94 91SKJN-BYU 797 47.82 16.67 60.00 Utah Lake


10/24/95 89SKJN-USU 76 16.81 4.52 221.25 Provo River (section 1)

10/24/95 91SKJN-BYU 177 25.32 6.99 143.08 Utah Lake

10/24/95 91SKJN-BYU 199 28.48 6.99 143.10 Provo River (section 1)




06/21/96 91SKJN-BYU 155 31.53 4.92 203.39 Provo River (section 2)


11/14/98 94SKJN Lot-4 352 36.79 9.57 104.51 Millville Ponds (USU)



08/99 94SKJN Lot-4 370 69.40 5.33 187.57 Provo River

08/30/99 990531SKJNFE01 13,532 4.05 3,340.70 0.30 Millville Ponds (USU)


04/09/01 000601SKJNPR07 8,364 7.58 1,102.82 0.91 Wahweap SFH

06/21/01 000601SKJNPR07 40 0.04 928.25 1.08 Mona Reservoir (USU)

07/06/01 000601SKJNPR07 132 0.14 938.73 1.07 Mona Reservoir (USU)

07/12/01 000601SKJNPR07 542 0.59 926.27 1.08 Goshen (USGS), Utah Lake (BYU)

07/19/01 010426SKJNPR01 80 0.01 1,0374.52 0.10 Utah Lake (BYU)


(continued)


51


Ovaprim. Ovaprim is an analog of salmongonadotropin releasing hormone with adopamine blocker (Hansen 2003b). Successwith artificial hatchery spawning has improved,with three crosses yielding tens of thousands offertile eggs each in 2004 and again in 2005.Improving yield and viability from artificialhatchery crosses of June sucker is being studiedfurther by U.S. Fish and Wildlife Service per-sonnel at the Bozeman (Montana) Fish Healthand Technology Center.

Stocking

Approximately 5,400 June sucker in excess ofbroodstock needs at FES were implanted with pas-sive integrated transponders (PIT) tags andreleased into Utah Lake and the Provo River in1994, 1995, 1996, and 1999 (Table 3). Subsequentmonitoring determines if June sucker raised in aprotected location will participate in wild spawningruns. This information has been used to help devel-op stocking protocols that are still being refined.

Adult and juvenile June sucker in excess of

Date Lot number Number stocked Kilograms stocked Fish/kg Grams/fish Stocking/transfer location




07/02/02 020521SKJNPR07 220 0.01 23,095.66 0.04 USU

07/10/02 020521SKJNPR07 4,000 0.80 4,982.11 0.20 USU

09/02 990618SKJNPR01 8 1.21 6.63 150.82 Utah Lake (used in state fair)

07/10/03 030513SKUTPR01 1,302 0.12 11,039.89 0.09 BYU

07/10/03 030513SKUTxNPR02 861 0.09 9,990.25 0.10 BYU

07/10/03 030515SKJNPR03 8,754 0.78 11,220.31 0.09 BYU

07/10/03 030515SKJNxUTPR04 9,297 0.63 14,852.20 0.07 BYU

07/10/03 030520SKJNPR11 1,432 0.06 24,284.36 0.04 BYU

07/10/03 030521SKJNPR13 2,965 0.10 28,419.98 0.04 BYU

07/10/03 030523SKUTPR15 2,097 0.07 30,820.11 0.03 BYU

07/10/03 030528SKJNPR19 3,274 0.09 36,089.07 0.03 BYU

07/11/03 030520SKJNxUTPR12 2,391 0.08 31,006.85 0.03 BYU

07/11/03 030520SKUTPR09 2,057 0.14 15,116.11 0.07 BYU

07/11/03 030520SKUTxJNPR10 2,137 0.15 14,722.50 0.07 BYU

07/11/03 030521SKJNxUTPR14 3,935 0.24 16,064.90 0.06 BYU

07/11/03 030523SKUTxJNPR16 3,027 0.10 30,333.09 0.03 BYU

07/11/03 030528SKJNxUTPR20 2,954 0.09 32,561.73 0.03 BYU

07/17/03 00527SKJNFE01 19 3.27 5.82 171.89 Utah Lake

07/17/03 000523SKJNPR02 2 1.11 1.80 555.66 Utah Lake (lots were combined)

07/17/03 000601SKJNPR06 3

07/17/03 000602SKJNPR08 2

07/17/03 010518SKJNPR08 1

08/19/03 030520SKJNPR11 733 0.06 11,542.58 0.09 USU

08/19/03 030523SKUTPR15 333 0.04 9,176.59 0.11 USU

11/24/03 020521SKJNPR07 400 3.81 105.11 9.51 USU

12/02/03 020521SKJNPR07 67 0.66 101.17 9.88 USU

01/21/04 020521SKJNPR07 30 0.37 80.66 12.37 USU



ANDERSEN ET AL.

those needed to maintain respective populationswere removed from Red Butte and Camp Creekreservoirs in 2001, and 1,600 (approximately 700and 900, respectively) were released into UtahLake (Table 3). Condition of sampled fish indi-cated stunting, possibly due to lower water levelsand/or density, in Camp Creek Reservoir(Thompson 2001). Fish released were intendedto augment the wild population of older fish.Stocked fish were hatched and raised in thehatchery with no attempt to imprint them toUtah Lake tributary water, and it was unknown ifstocked fish would participate in spawningevents in the Provo River.

Approximately 8,300 YOY June sucker fromFES, originally collected as eggs from a streamsidecross on the Provo River, were taken to the Wah-weap State Fish Hatchery in April 2001. The fishwere held at Wahweap until August 2002, at whichtime 2,474 individuals that survived and grew wereimplanted with PIT tags and transported to UtahLake. A subset of the same June sucker lot was heldat FES throughout the study period.

ResultsEmergency actions

The emergency efforts implemented to save Junesucker have been effective in that the species stillexists, albeit in low numbers, in the wild. The adult

population was estimated at less than1,000 in 1999 (USFWS 1999). Less than300 June sucker adults have been cap-tured each year during the annualspawning run up the Provo River in thefirst years of the 21st century (UDWR,unpublished data).

Refuges

The June sucker stocked in Red Butteand Camp Creek reservoirs are spawn-ing, as demonstrated by the presenceof multiple year-classes younger thanthose originally introduced (UDWRfield data, Thompson 2001). In 2003,June sucker eggs were collected intraps and nets along the perimeter ofRed Butte Reservoir and the face of the

dam. Adult fish were observed spawning on theface of the dam and a digital video record of theactivity was collected. Vertebral and operculardeformities have not been seen in June suckerfrom these habitats, although high densities areapparently stunting growth rates. June suckerhave been spawning in Arrowhead Pond in thepresence of green sunfish. The Red Butte Dam isbeing repaired, and June sucker will be reintro-duced into the reservoir as it refills in 2006. CampCreek Reservoir continues to hold June suckerand to produce excess fish that are transferred toUtah Lake.


The June sucker program has acquired lands at themouth of Hobble Creek, one of the primaryinflows to Utah Lake’s Provo Bay (Figure 1). Fol-lowing habitat enhancement this location couldprove to be favorable for June sucker duringspawning, rearing, or other life stages. Attempts bythe June sucker program partners to purchaselands along the Provo River near the mouth havebeen largely unsuccessful to date. The Program ispursuing easements on lands just north of themouth of the Provo River that may provide tempo-rary habitat for June sucker (e.g., to be used asgrow out ponds for YOY fish).

The carp control feasibility study initiated bythe Program has completed two seasons. The cur-

52

Table 4.

Percent survival to swim up compared for three June sucker hatcherycrosses and three crosses received from wild stock in 2003.

Number of fish Percent survival Cross number Number of eggs on feed (swim up) to swim up

Hatchery Cross 1 6,875 3,855 56.07

Hatchery Cross 2 7,497 2,219 29.60

Hatchery Cross 3 8,162 5,281 64.70

Wild Cross 1 23,750 9,371 39.46

Wild Cross 2 32,487 17,391 53.53

Wild Cross 3 31,529 1,832 5.81

* Two lots received from wild stock were not included in this comparisonsince they were discarded prior to projected hatch date due to poor eggquality and condition.


rent estimate of the carp population in Utah Lakeis 7 million (c.i. 6–10 million). The Program is nowassessing how a continuous removal project mightbe made effective and affordable.

Because some of the nonnative fishes in UtahLake and its tributaries are sport species, the desiresof Utah Lake anglers are being considered as non-native fish management actions are developed. Acreel survey is being conducted to assess anglerdesires and habits.


The largest number of June sucker in captivity isheld at FES. About 1% of these fish are mature andthe remainder are juveniles. The oldest June suck-er in captivity, up to 16 years old (Table 2), nowheld at FES, are showing signs of stress, includingosteologic and opercular deformities. Conditionof the FES population as a whole has improved,however, in the new June sucker facility withwarmer water. At this time, FES is holding 10mature brood lots, each made up of less than 50individuals, 23 immature lots ranging in numberfrom 4 to more than 400 individuals, and 23 prog-eny lots of from 8 to more than 5,000 individuals.As of 1 October 2005 the facility was holding morethan 46,000 individual June suckers (Table 2).


Captive feeding studies have concluded that a feedregime of brine shrimp and the Razorback Diet(formulated by the U.S. Fish and Wildlife ServiceBozeman Fish Health and Technology Center andmanufactured by Nelson & Sons, Inc., Bozeman,Montana) is the most suitable diet for June suckercompared to the other diets evaluated. Other dietsfed to June sucker prior to the feed study were twoof the more inferior diets tested. The switch to thepreferred diets has improved overall fish conditionand reduced the occurrence of deformities(Hansen 2002; Hansen 2003a).

Artificial crosses between select individualsfrom brood lots to produce progeny for releasehave resulted in variable production and sur-vival. Induced spawning trials with Ovaprim andvariable water temperatures showed that a sig-nificantly higher number of fish ovulated whenheld in 13°C water for 4 months, compared to

13°C for 1 week and 18°C for more than a year.Human chorionic gonadotropin effectivelyincreased the amount of milt extruded by males.In 2003, there was no significant difference inpercent of eggs hatched from the wild stock ver-sus progeny of induced captive broodstock; how-ever, an order of magnitude fewer eggs were pro-duced from hatchery crosses, compared to thosereceived from streamside crosses (Table 4;Hansen 2003b).

The 8,300 June sucker held at Wahweap from2001 to 2002, fed the diet used at FES in the 1990s,grew an average of 0.2 mm/d. Individuals from thesame lot at FES grew an average of 0.12 mm/d. Sur-vival rates during the study period were 35% atWahweap and 73% at FES. Biologists implanted2,474 June sucker with PIT tags at Wahweap; ofthese, 200 (7%) expired during tagging, and ofthese, 140 (70%) were noticeably deformed. Nomortalities were incurred during transport fromWahweap to Utah Lake.

Stocking monitoringRipe hatchery and/or refuge-reared June suckerbegan to appear in spawning runs in 1995 (Kele-her et al. 1999; Utah Division of WildlifeResources, unpublished data), and the first returnsfrom the augmentation program begun in 1994.Of approximately 6,000 PIT-tagged June suckerstocked into Utah Lake and the Provo River fromFES, Red Butte Reservoir, and Camp Creek Reser-voir, 253 have been captured during the spawn inthe Provo River along with wild fish (Table 5). Thisnumber may not be a true indicator of survival ofstocked fish. Some stocked fish may have not yetreached sexual maturity, and monitoring does notsample all fish in the spawning run. Largest num-bers of returning stocked fish have come from RedButte Reservoir, but individuals from FES andCamp Creek have also been recorded (Table 5).From 1997 through 2003, June sucker returned tospawn in the Provo River between 20 April and 1July. Estimated onset of June sucker spawningactivity was late May in 1997 and 1998, the firstweek of June in 1999, and late April during2000–2004. It was difficult to sample for Junesucker in the Provo River in 2005 due to unseason-ably high spring flows. The proportion of fish in


53


ANDERSEN ET AL.

the spawning run made up of fish reared in captiv-ity has been increasing and consisted of 75% ofthe fish captured in the 2003 spawning run (UtahDivision of Wildlife Resources, unpublished data).

Fish in recent Provo River spawning runsexhibit morphologies consistent with June suck-er, Utah sucker, and hybrids of the two species.Currently available data indicate that there is

measurable diversity within the existing Junesucker genome (Mock et al. 2004, 2006). Theheritability of morphologic traits observed insuckers from Utah Lake is being investigated, asare environmental influences on morphology.

Discussion

Emergency actions

Those interested in the persistence of the Junesucker should be encouraged that the emergencyactions taken on behalf of June sucker beginning in1982 have supported survival of the species to date.Time will tell whether these actions were in time(i.e., whether sufficient genomic diversity has beenmaintained to eventually support a self-sustainingpopulation of June sucker in Utah Lake).

Because the June sucker in the wild do notappear to be successfully reproducing in sufficientnumbers, a network of hatcheries and refuges willbe maintained with June sucker for the foreseeablefuture. Hatcheries and refuges continue to be man-aged to produce large numbers of June sucker forrelease into Utah Lake.

Refuges

The ability of June sucker to spawn in lentic habitats(i.e., Red Butte and Camp Creek reservoirs, andArrowhead Pond) suggests that they may be able tospawn in Utah Lake, and possibly spawned in thelake historically. This reproductive strategy has beenobserved in other lakesuckers. In-lake spawning wasobserved in native lakesuckers of Upper KlamathLake in association with subsurface groundwaterdischarge areas (Buettner and Scoppettone 1990;Perkins et al. 2000). June sucker spawning in UtahLake, if it occurs, will be difficult to observe anddocument until the numbers of June suckers in thelake are increased. Lentic spawning in the reservoirsites has been, and likely will continue to be, animportant source of June sucker that can be stockedinto Utah Lake to help support species survival untilnatural reproduction and environmental remedia-tion are sufficient to support a self-sustaining pop-ulation. Refuges are likely to continue to be impor-tant for conservation of the species. Newly emerginggenetic data will help maximize the available genet-ic diversity in these managed habitats.

54

Table 5. Stocked June sucker recaptured in the Provo River.

Date Stocked Returned Source

10/05/94 760 FES

10/11/94 791 FES

06/01/95 19 FES

08/16/95 235 FES

08/18/95 256 FES

08/23/95 259 FES

08/24/95 180 FES

10/05/95 173 FES

10/05/95 370 FES

10/06/95 194 FES

10/18/95 474 FES

10/18/95 460 FES

10/23/95 129 FES

10/23/95 145 FES

06/01/96 2 FES

06/21/96 295 FES

06/01/97 6 FES

06/01/98 1 FES

06/01/99 6 FES

09/01/99 692 FES

06/01/00 21 FES

05/01/01 479 RB

06/01/01 12 FES

10/04/01 222 RB

10/09/01 587 CC

10/10/01 315 CC

06/01/02 36 FES

06/01/02 65 RB

06/01/02 9 CC

06/01/03 35 FES

06/01/03 97 RB

06/01/03 2 CC

FES = Fisheries Experiment Station.

RB = Red Butte Reservoir.

CC = Camp Creek Reservoir.



Environmental threats to June sucker, especially (1)nonnative fishes that prey on young June suckerand alter habitats, and (2) habitat alteration, mustbe reduced if artificially and naturally producedJune sucker are to survive in Utah Lake and its trib-utaries. Increasing the survival of wild-producedlarvae to recruitment will be necessary to establish-ing a self-sustaining June sucker population. Addi-tional predator avoidance locations and/or mecha-nisms are needed. Predator avoidance locationsmay include more complex habitats where wild-produced larvae can find cover from predators.Selective passage gates and other structures may bethe mechanisms necessary to help young Junesucker survive to recruitment.

The ability of June sucker to recruit in habitatsother than Utah Lake (e.g., Camp Creek and RedButte reservoirs) indicates that current conditionsin the lake are limiting to recovery, especiallydegraded habitat and the presence of nonnative fishspecies. At least some habitat restoration and non-native fish population reduction will be necessaryto reestablish June sucker recruitment. The loss ofhistoric flows on the Provo River has negativelyimpacted spawning adults (Radant et al. 1987).Efforts by Program participants to acquire waterand mimic a natural hydrograph during the spawn-ing period are ongoing and will continue. Channel-ization and land development have limited habitatcomplexity at the mouth of river and reduced avail-able cover for drifting larvae. Habitat investigationssuggest that increased vegetative cover permitsincreased survival of larvae (Petersen 1996; T. A.Crowl and M. C. Belk, June Sucker Recovery Imple-mentation Program, unpublished data). Efforts toincrease the complexity of the river mouth and pro-vide conditions that allow for restoration of aquat-ic macrophytes in key areas will continue but maybe limited by lack of access to privately held lands.Many investigators have documented the negativeeffects of predation by nonnative fish on June suck-er, especially drifting larvae (Modde and Muirhead1994; Petersen 1996; Belk et al. 2001; SWCA 2002).Successful June sucker spawning and recruitment inArrowhead Pond in the presence of green sunfishoffers additional evidence of the importance ofcover for larvae. Habitat modification by common

carp reduces emergent macrophytes that serve ascover for drifting larvae, increasing their exposureto predation (SWCA 2002). Efforts by Programparticipants to develop strategies to reduce andcontrol nonnative fish species, especially white bassand common carp, should continue.

Cooperative efforts to provide minimum flowsin the Provo River to benefit June sucker havelargely been effective, even in drought years. Pro-gram partners are investigating the potential valueof pulsing flows to help distribute larval June suck-er out of the Provo River where a limited amountof cover is available.

June and Utah suckers probably have a sharedevolutionary history. The observed morphologicdifferences between the two species may be theresult of recent ecological selection (Mock et al.,2006). More diverse habitats that have some of thephysical features of the historic Utah Lake habitatwill be necessary for maintaining the environ-ment driving the evolutionary trajectory that hasmaintained the two sucker forms to date. The tra-ditional timing of June sucker spawning, early tomid-June, on the descending limb of the hydro-graph (Jordan 1878, Radant and Sakaguchi 1981;Shirley 1983) before water temperatures hadwarmed (Shirley 1983) may have been a methodfor maintaining a spawning run distinct from theUtah sucker who spawned in April (Radant et al.1987). Personnel from UDWR now capture Junesucker, Utah sucker, and hybrids ascending theProvo River to spawn as early as March. Mainte-nance of two distinct species, June sucker andUtah sucker, may require that a more naturalhydrograph is reestablished so that these speciesmay experience their traditional spawning cues offlow and temperature.


Studies are currently under way to investigate thedegree of relatedness between morphologies andgenomes in suckers ascending the Provo River tospawn. A study is underway investigating heritabil-ity of the observed morphologic traits. Investiga-tors are studying the genetic structure and diversi-ty of captive June sucker and those captured in theannual spawning run. Results of these investiga-tions will help to direct production of additional


55


ANDERSEN ET AL.

fish in refuges, including hatchery crosses. Possibly,the majority of genetic diversity available in wildfish is already well represented in captivity. Alterna-tively, captive fish may represent a fraction of thediversity evident in wild spawning fish, and devel-opment of additional brood lots may be needed tomaximize the diversity present in the brood lotsand resulting progeny lots destined for release intoUtah Lake. The advancing age of the oldest broodlot fish in captivity is also of concern, as some ofthese fish may senesce before they can contributeto brood lots, an outcome that may also require thecollection of additional wild fish for brood lots. Inthe absence of clear genetic markers, broodstockhas been selected based on June sucker morpho-logic characteristics, but continued research ingenetics, morphology, ecology, and heritability isneeded to support or modify this approach.Research into all of these areas has been underwayfor years (e.g., Belk and Benson 2005; Belk et al.2005; Mock et al. 2004, 2006). Data and recom-mendation from the research are providingincreasingly specific information to help guidemanagement actions, especially propagation.


Warmer water and specifically formulated suckerfeeds appear important for June sucker growth andsurvival in the hatchery. The June sucker indoorfacility completed at FES in 2001 has been animportant tool for increasing water and ambient airtemperatures, increasing captive June sucker sur-vival rates, lowering crippling rates, and providingspace for research needs such as feeding and spawn-ing hormone studies. An interim hatchery at FES isplanned for completion in 2006 in order to provideadditional space, take advantage of a limited watersupply, and further increase water temperatures bycirculating water in a closed system.

Producing young June sucker by spawning cap-tive fish in standing water bodies (i.e., extensive cul-ture) appears to produce superior quality fish onaverage, compared to those reared in indoor hatch-ery tanks (i.e., intensive culture), as illustrated by thecomparison of fish from the same lot reared at FESand Wahweap. Lower survival rates and highergrowth rates of fish in extensive culture compared tointensive culture have been documented by Cushing

(1981). Findings suggest that June sucker may bemore fit for the wild if appropriate exterior rearinglocations can be secured (e.g., Belk and Benson 2005;Belk et al. 2005). However, production numbers arecurrently limited by available space. Evaluation ofextensive culture areas must include review of waterquality, dependable water supply, and site security.Current stocking and study demands require main-taining both culture types. A warmwater fish hatch-ery, incorporating intensive and extensive methodsat two locations, has been planned, based on initialestimates of producing more than 3 million Junesucker for release into Utah Lake to support recovery.If producing these large numbers of June sucker areultimately reviewed and approved by Program par-ticipants, additional facilities will be needed. Anadditional protocol of spawning June suckerstreamside, hatching them in protected locations(e.g., FES) and holding them in protected locations(e.g., Red Butte and Camp Creek reservoirs) or pro-tected nearshore areas, will likely be necessary untilthey can successfully avoid predation. Artificialpropagation will be directed by the captive manage-ment plan now being reviewed and revised by Pro-gram participants. The Program partners and FESpersonnel are working closely with the U.S. Fishand Wildlife Service Bozeman Fish Health andTechnology Center to improve captive reproduc-tion of June sucker.

Stocking

The presence of recently stocked June sucker inannual spawning runs in the Provo River indicatesthat initial efforts to artificially increase the spawn-ing population have been successful. Not all fishparticipating in the spawning runs are captured,and more stocked fish probably participate in runs.Additionally, it is likely that some stocked fish havenot reached sexual maturity and will contribute tospawning runs in future years, or some fish mayskip spawning in some years. Propagation andstocking need to continue until other threats limit-ing natural reproduction and recruitment, espe-cially habitat loss and effects due to nonnative fish,have been reduced or eliminated. Propagation andstocking protocols need additional refinement soas to introduce the most robust and geneticallydiverse individuals possible.

56


Both remaining native fish species of Utah Lake,June sucker and Utah sucker, stand to benefit fromthe ecosystem approach of the Program, especiallythe efforts to improve habitat, control nonnativefish, and provide instream flows. Maximizing avail-able genetic diversity should ensure, to the extentpossible, that the population of June sucker inUtah Lake has the greatest possible tolerance forfuture environmental variation. Increasing the wildpopulation size by augmentation with geneticallydiverse June sucker should increase survival poten-tial and recruitment.

Acknowledgments

Since 1982 many agencies have supported theefforts described in this paper. They include UtahDivision of Wildlife Resources, Utah Departmentof Natural Resources, U.S. Fish and Wildlife Ser-vice, Utah Reclamation Mitigation and Conserva-tion Commission, U.S. Department of the Interior,U.S. Bureau of Reclamation, Central Utah WaterConservancy District, Utah State University, andBrigham Young University. June sucker have beenheld at the Red Butte Reservoir with authorityfrom the U.S. Army, U.S. Forest Service, and cur-rent authority from Central Utah Water Conser-vancy District. June sucker are held at Camp CreekReservoir with authority from Boyd Warr. J. E.Wiglama prepared the tables. This paper was great-ly improved by the extensive comments of ananonymous reviewer.

References

Belk, M. C. 1998. Age and growth of June sucker (Chasmistesliorus) from otoliths. Great Basin Naturalist58(4):390–392.

Belk, M. C., M. J. Whitney, and G. B. Schaalje. 2001. Complexeffects of predators: determining vulnerability of theendangered June sucker to an introduced predator. Ani-mal Conservation 2001(4):251–256.

Belk, M. C., and L. J. Benson. 2005. Hatchery-induced mor-phological variation in an endangered fish: a challengefor hatchery-based recovery efforts. Report to the JuneSucker Recovery Implementation Program. BrighamYoung University, Provo, Utah.

Belk, M. C., D. B. Gonzalez, and R. C. Tuckfield. 2005. Den-sity effects on growth, survival, and diet of June suckerChasmistes liorus: a component allee effect in an endan-gered species. Report to the June Sucker Recovery Imple-

mentation Program project V. 02.07. Brigham YoungUniversity, Provo, Utah.

Buettner, M., and G. G. Scoppettone. 1990. Life history andstatus of Catostomids in upper Klamath Lake, Oregon.U.S. Fish and Wildlife Service, National FisheriesResearch Center, Reno Field Station, Completion Report,Reno, Nevada.

Carter, D. R. 1969. A history of commercial fishing in Utah Lake.Master’s thesis. Brigham Young University, Provo, Utah.

Cope, E. D., and H. C. Yarrow. 1875. Report upon the collec-tions of fishes made in portions of Nevada, Utah, Cali-fornia, Colorado, New Mexico, and Arizona, during theyears 1871, 1872, 1873, and 1874. Report of the Geo-graphic and Geologic Exploration and Survey West ofthe 100th Meridian (Wheeler Survey) 5:635–703.

Cushing, D. H. 1981. Fisheries biology: a study in populationdynamics. University of Wisconsin Press, Madison.

Fuhriman, D. K., L. B. Merrit, A. W. Miller, and H. S. Stock.1981. Hydrology and water quality of Utah lake. GreatBasin Naturalist Memoirs 5:43–67.

Goede, R. W., and B. A. Barton. 1990. Organismic indicesand an autopsy-based assessment as indicators of healthand conditions of fish. Pages 93–108 in S. M Adams, edi-tor. Biological indicators of stress in fish. American Fish-eries Society, Symposium 8, Bethesda, Maryland.

Hansen, E. S. 2002. Evaluation of diets for rearing June suck-er (Chasmistes liorus). Ichthyogram. 13(4):9–11. UtahDivision of Wildlife Resources, Fisheries ExperimentStation, Logan.

Hansen, E. S. 2003a. Evaluation of feed regimes for rearingJune sucker (Chasmistes liorus). Ichthyogram 14(1):1, 6,7. Utah Division of Wildlife Resources, Fisheries Experi-ment Station, Logan.

Hansen, E. S. 2003b. Evaluation of induced spawning tech-niques and requirements in captive June sucker (Chas-mistes liorus). Ichthyogram 14(4):6–9. Utah Division ofWildlife Resources, Fisheries Experiment Station, Logan.

Heckmann, R. A., C. W. Thompson, and D. A. White. 1981.Fishes of Utah Lake. Great Basin Naturalist Memoirs 5,Utah Lake Monograph, Brigham Young University,Provo, Utah.

Jordan, D. S. 1878. A synopsis of the family Catostomidae.Contributions to North American Ichthyology III B. U.S.National Museum Bulletin 12:97–220.

Jordan, D. S., and B. W. Evermann. 1903. American food andgame fishes. Doubleday Page, New York.

Keleher C. J., L. D. Lentsch, and C. W. Thompson. 1999. Eval-uation of Flow Requirements for June Sucker (Chas-mistes liorus) in the Provo River: an empirical approach.Utah Division of Wildlife Resources, Publication No.99–06, Salt Lake City.

Li, T. 1999. Genetics of endangered species Chasmistes liorus(June sucker). Doctoral dissertation. Brigham YoungUniversity, Provo, Utah.

Miller, R. R., and G. R. Smith. 1981. Distribution and evolu-tion of Chasmistes (Pisces: Catostomidae) in westernNorth America. Occasional Papers of the Museum ofZoology University of Michigan 696.

Mock, K. E., M. P. Miller, and B. L. Cardall. 2004. Genetic


57


analysis of spawning and refuge June sucker populations.Report submitted in fulfillment of June Sucker RecoveryImplementation Program contracts V. 03.01 (amended)and V. 02.04b. Utah State University, Logan.

Mock, K. E., R. P. Evans, M. Crawford, B. L. Cardall, S. U.Janecke, and M. P. Miller. 2006. Rangewide molecularstructuring in the Utah sucker (Catostomus ardens). Mol-ecular Ecology 15(8):2223–2238.

Modde, T., and N. Muirhead. 1994. Spawning chronologyand larval emergence of June sucker (Chasmistes liorus).Great Basin Naturalist 54(4):366–370.

Perkins, D. L., G. G. Scoppettone, and M. Buettner. 2000.Reproductive biology and demographics of endangeredLost River and shortnose suckers in upper Klamath Lake,Oregon. Report to the U.S. Bureau of Reclamation,Washington, D.C.

Petersen, M. E. 1996. The effects of prey growth, physicalstructure, and piscivore selectivity on the relative preyvulnerability of gizzard shad (Dorosoma cepedianum)and June sucker (Chasmistes liorus). Master’s thesis. UtahState University, Logan.

Popov, B. H. 1949. The introduced fishes, game birds, andgame and fur-bearing mammals of Utah. M.S. thesis.Utah State University, Logan.

Radant, R. D., and D. K. Sakaguchi. 1981. Utah Lake fish-eries inventory. U.S. Bureau of Reclamation Contract8–07-40–50634. Utah Division of Wildlife Resources,Salt Lake City.

Radant, R. D., M. M. Wilson, and D. S. Shirley. 1987. Junesucker Provo River instream flow analysis. Final reportfor the United State Bureau of Reclamation Contract8–07-40–50634, Modification 4. Utah Division ofWildlife Resources, Salt Lake City.

Scoppettone, G. G., and G. Vinyard. 1991. Life history andmanagement of four endangered lacustrine suckers.Pages 359–377 in W. L. Minckley and J. E. Deacon, edi-tors. Battle against extinction; native fish management inthe American West. University of Arizona Press, Tucsonand London.

Shirley, D. L. 1983. Spawning ecology and larval develop-ment of the June sucker. Pages 18–36 in Proceedings ofthe Bonneville Chapter of the American Fisheries Soci-ety 1983. American Fisheries Society, Bonneville Chap-ter, Bethesda, Maryland.

SWCA (Stephen W. Carothers Associates, Inc.). 2002. Non-native Fish Control Feasibility Study to benefit Junesucker in Utah lake. Report to the June Sucker RecoveryImplementation Program, Utah Department of NaturalResources, Salt Lake City.

Thompson, P. 2001. June sucker (Chasmistes liorus) monitor-ing and transfer activities in the northern region, 2001.Utah Division of Wildlife Resources, Publication num-ber 01–24, Salt Lake City.

U.S. Army Corps of Engineers and U.S. Fish and WildlifeService. 2003. Red Butte Reservoir Transfer environmen-tal assessment. U.S. Army Corps of Engineers and U.S.Fish and Wildlife Service, Salt Lake City, Utah.

UDWR (Utah Division of Wildlife Resources). 2004. Man-agement plan for June sucker in captivity. Utah Divisionof Wildlife Resources, Publication number 04–2, SaltLake City.

USBR (U.S. Bureau of Reclamation). 1989. Beyond the Wasatch.Government Printing Office, Slat Lake City, Utah.

USFWS (United States Fish and Wildlife Service). 1999. Junesucker (Chasmistes liorus) recovery plan. United StatesFish and Wildlife Service, Region 6, Denver.

USOFR (U.S. Office of the Federal Register). 1986. Endan-gered and threatened animals and plans; final rule deter-mining the June Sucker (Chasmistes liorus) to be anEndangered Species with Critical Habitat. Code of Fed-eral Regulations, Title 50, Part 17. U.S. GovernmentPrinting Office, Washington, D. C.

Utah Division of Water Resources. 1997. State water plan,Utah Lake basin. Utah Division of Water Resources, SaltLake City.

Wakefield, J. H. 1933. A study of the plant ecology of SaltLake and Utah valleys before the Mormon immigration.Master’s thesis. Brigham Young University, Provo, Utah.

58

ANDERSEN ET AL.


59

Four species of gobies and one eleotrid com-prise the indigenous stream fishes of theHawaiian high islands. Gobies and eleotrids

occur around the world in temperate and tropicalwaters (Nelson 1984), but the evolutionary, behav-ioral, and ecological affinities of Hawaiian streamfishes rests among relatives in the oceanic islands ofthe tropical Pacific to the south and west of theHawaiian Islands rather than to the east amongfishes in coastal streams of North and South Amer-ica (Fitzsimons et al. 2002a). These animals do notconform to traditional definitions as primary orsecondary freshwater fishes because their life cyclesinclude a marine phase. In Hawai’i, adults of threeupstream species are limited to freshwater. Two

other species occur as adults in lower sections ofstreams, estuaries, and river mouths where condi-tions range from completely fresh to brackish withsalinities occasionally approaching that of seawa-ter. These five species, together with two species ofmollusks, an endemic prawn, and a shrimp, exhib-it amphidromy (McDowall 1992), a kind ofdiadromous life cycle in which there are twomigrations between freshwater and the sea. Spawn-ing occurs in freshwater where eggs hatch within48 h, and free-living embryos (sensu Balon 1990)are swept downstream into the ocean. Althoughthe ocean phase of the life cycle for these fishes maybe as long as 6 months (Radtke et al. 1988), the ani-mals usually return to freshwater as glass-clear

Status of Native Hawaiian Stream Fishes,A Unique Amphidromous Biota

ABSTRACT Native Hawaiian stream fishes are represented by only five species belonging to two families,

Gobiidae (‘o’opu nakea Awaous guamensis, ‘o’opu ‘alamo’o Lentipes concolor, ‘o’opu nopili Sicyopterus stimpsoni,and ‘o’opu naniha Stenogobius hawaiiensis) and Eleotridae (‘o’opu ‘akupa Eleotris sandwicensis). All species are

found on each of the main Hawaiian Islands, and none is currently threatened or endangered. These animals are

not true freshwater fishes, but rather share an amphidromous life cycle where adults live and reproduce in streams

and larvae develop at sea. Techniques developed for sampling (electroshocking, seining) and assessment (e.g.,

index of biotic integrity, instream flow incremental methodology) in continental U.S. streams are inappropriate for

Hawaiian streams. Thus, procedures were developed specifically for fishes in streams on oceanic islands of the

tropical Pacific where amphidromy is the predominant life history mode. Geographical information systems-com-

patible data from ongoing statewide native stream fish surveys can soon be viewed on the Web site for the Hawai’i

Division of Aquatic Resources (http://www.hawaii.gov/dlnr/dar). The 2000 Hawai’i Supreme Court decision on the

Waiahole Water Dispute specifically provides for the maintenance of optimum flow for native stream fishes, and the

Division of Aquatic Resources has adopted policies guiding instream water use decisions: (1) no net loss of habitat

for native fishes, (2) use of a watershed or ahupua’a perspective, and (3) maintenance of an open corridor between

the stream and the ocean to facilitate native species migrations. The preservation of indigenous Hawaiian stream

fishes now has been elevated to the highest level of protection in the state.

Robert T. Nishimoto and J. Michael Fitzsimons



ROBERT T. NISHIMOTO Hawai’i Division of Aquatic Resources, Post Office Box 936, Hilo, Hawai’i 96721-0936, USA. Corresponding author:[email protected]

J. MICHAEL FITZSIMONS Museum of Natural Science, Louisiana State University, Baton Rouge, Louisiana 70803, USA and Department of Biolo-gy, University of Hawai’i at Hilo, Hilo, Hawai’i 96720, USA


larvae with forked caudal fins and little or no pig-mentation on the head and body (Tate et al. 1992;Nishimoto and Kuamo’o 1997). The distinctionbetween amphidromy and other types of diadromy(e.g., anadromy and catadromy) is that immatureanimals are involved in both migrations and repro-duction does not immediately follow access tohabitats occupied by adults. There is no evidencefor homing to a natal site, but young fishes migrat-ing inshore toward the mouths of streams almostinvariably orient toward a source of decreasedsalinity in tests offering a choice of full-strengthseawater versus freshwater or even dilute seawater(Smith and Smith 1998).

Hawaiian freshwater fishes mostly occur on thenorth and east slopes of islands where trade windsproduce the orographic rainfall that maintainsperennial streams (Armstrong 1983). Streams witha full complement of native fishes are typicallyclear and cold and have a strong flow all year long.In such streams, there is little accumulation of sed-iment, leaf litter, and other loose debris because ofuninterrupted flow and powerful freshets causedby localized heavy rains in the mountains or by thepassage of weather fronts that can occur any timeof the year but are most frequent during the rainymonths (October to April). Because of frequentflash floods, Hawaiian streams are better describedas dynamic rather than stable; each stream charac-teristically remains in a constant state of recoveryfrom the most recent freshet. Native fishes in thesestreams survive not in spite of episodic floods, butactually because of them (Fitzsimons et al. 1996).Flash floods remove organic debris and sedimentsfrom spawning sites and restrict underwater vege-tation to those species of algae (diatoms and low-growing filamentous species) representing earlystages of succession in an aquatic community.These plant species are essential food items for onespecies of fish that is an obligate herbivore (Fitzsi-mons et al. 2003) and are important also in thediets of at least two other species of fishes. Floodsfacilitate migrations of amphidromous nativestream fishes and macroinvertebrates by openingup the stream where it flows into the sea. There isincreasing evidence that flood waters entering thesea provide a biological signal important in timingonshore migration of larval fishes that live as adults

in freshwater streams. Finally, freshets, especiallyduring the onset of the rainy season, trigger anincrease in reproductive behavior. Ecosystem func-tioning in Hawaiian streams is based on recurringflash floods that maintain the colonizing species offishes, invertebrates, and algae also as the principalmembers of the final stage of biotic succession(Fitzsimons and Nishimoto 1995; Fitzsimons et al.2003). Understanding this relationship is signifi-cant when fishery management and stream-usedecisions seek to preserve natural biodiversity inisland streams.

From this background, our report offers briefdescriptions of Hawaiian stream fishes, assessestheir threatened/endangered status, explains theneed for survey and monitoring procedures devel-oped specifically for island streams, and describeshow ecological and behavioral data have been usedto establish policies for water-use decisions in theHawaiian Islands.

Methods and Materials

Sources for the information provided here are theHawai’i Stream Assessment (HSA), the streamdatabase maintained by the Hawai’i Division ofAquatic Resources (DAR), a manuscript inprogress on the natural history of Hawaiianstreams and stream animals, and references citedherein. The HSA was distributed in December1990 by the Commission on Water Resource Man-agement, Department of Land and NaturalResources, State of Hawai’i; the report includedbaseline information on the location, physical fea-tures, and species present in perennial streamsthroughout the high islands in the southeasternsection of the archipelago. At that time, HSA wasthe principal document for assisting with streamprotection and management decisions. Morerecently, the DAR database has incorporated datafrom HSA along with subsequent surveys into acoding system that provides an eight-digit code forthe identification of individual watershed units;data from animal abundance, distribution, ecology,behavior, recruitment, hydrology, anthropology,and the existence of photos and video recordingsfrom each survey site are linked with the four-digitsystem employing geographic information systems

60

NISHIMOTO AND FITZSIMONS


(GIS) analysis as the umbrella for integratinginformation in response to broad or narrowlydefined queries.

ResultsSpecies description and distribution

In Hawai’i, gobies and eleotrids are referred to col-lectively as ‘o’opu (Pukui et al. 1983). Marinespecies are ‘o’opu kai, the latter word in reference tothe sea. Freshwater species are ‘o’opu wai (meaningriver, stream, or freshwater). Binomials are usedalso for individual fish species. As with binomialscientific names (genus and specific epithet), thefirst Hawaiian name indicates relationship and thesecond is descriptive of the species.

‘O’opu akupa Eleotris sandwicensis (“big-mouthed ‘o’opu”) are endemic eleotrids easily dis-tinguished from stream gobies by a prognathouslower jaw and separated pelvic (ventral) fins.‘O’opu akupa occur in estuaries, lower sections ofstreams, and occasionally in tide pools and anchia-line ponds. These fish usually occur no fartherupstream than the first waterfall. They are sit-and-wait predators on fishes and invertebrates. Obser-vations of eggs, courting pairs, and recruitment oflarvae indicate that akupa likely reproducethroughout the year.

‘O’opu naniha Stenogobius hawaiiensis (“the‘o’opu that avoids”) are endemic gobies readilyidentified by the black “tear drop” that extendsfrom the lower edge of the orbit down and back-ward across the cheek toward the bottom of theoperculum.‘O’opu naniha occur on all high islandsof the Hawaiian Chain and are often the mostcommon gobies in seaside pools, freshwater andbrackish ponds, along the margins of streams, andespecially in lower sections of streams with sandand gravel bottoms. These fish have the fusedpelvic fins that form a sucking disk in true gobies,but, in contrast to upstream species, the structureis elongate, poorly muscled, and not well adaptedfor hanging onto rocks in strong currents. As aresult, ‘o’opu naniha are naturally restricted todownstream areas below the first waterfall and thefish are usually not found in high-gradient moun-tain streams that lack slowly moving water near thestream mouth. They are omnivores on benthic

organisms, and they occasionally swim up to inter-cept items drifting downstream. Most courtshipand spawning have been observed during summermonths.

The indigenous ‘o’opu nakea Awaous guamensis(“light colored ‘o’opu”) are light tan to pale greenwith a dark chain-like stripe extending along themidside of the body and tail and ending in aprominent dark spot or blotch at the base of thecaudal fin. ‘O’opu nakea are the most commonnative freshwater fish in the windward streams ofthe five major islands. Individuals of the ‘o’opunakea may be found anywhere along the length ofa stream from the mouth to several miles inland,but greater numbers of animals are typical of mid-dle sections of streams. Adults often bury them-selves particularly at night in loose sand and grav-el. The ‘o’opu nakea feed on filamentous algae,diatoms, and small stream animals such as fly lar-vae, mollusks, and oligochaetes mostly ingestedwith vegetation. The fish usually bite off largeclumps of algae or take in a mouthful of sand orgravel from which algae and invertebrates areremoved by comb-like gill rakers in the animal’s“pharyngeal mill.” Once cleaned, particles of sandand gravel are spit out or allowed to drop fromunder the lower edge of the gill cover. Althoughtypically bottom feeders, ‘o’opu nakea occasionallyswim up from the bottom and ingest pieces ofmaterial being washed downstream. ‘O’opu nakeaspawn mostly in the lower sections of streams fromAugust through December.

‘O’opu nopili Sicyopterus stimpsoni (“the ‘o’oputhat clings”) is unique among native stream fishesin Hawai’i by having three notches in the upper lip(one median and two lateral versus one or none inother species) and by being an obligate herbivore(diatoms and filamentous algae). The ‘o’opu nopilioccur on all islands and are most common in shal-low, swift parts of a stream well inland from themouth. Their range in a stream usually overlaps theupper extent of ‘o’opu nakea and the lower extentof ‘o’opu ‘alamo’o Lentipes concolor, but there aremany exceptions. Their presence in a stream isoften indicated by conspicuous feeding patches onthe upper surface of rocks and boulders wherealgae are scraped from the substrate with sweepingmovements of the upper jaw armed with elongate

61

STATUS OF NATIVE HAWAIIAN STREAM FISHES


tricuspid teeth. The ‘o’opu nopili spawn all yearlong, but most reproductive activity occurs duringsummer months.

The endemic ‘o’opu alamo’o (“lizard-like‘o’opu”) differ sharply from other stream fishes inHawai’i by having a single median notch in theupper lip and by usually occurring farther inlandand above higher waterfalls than any of the otherstream fishes. Courting males with jet black on thehead and body anterior to the posterior edge of thefirst dorsal-fin base and bright red-orange on theposterior part of the body and tail further distin-guish this species from other stream fishes. ‘O’opu‘alamo’o feed opportunistically on stream inverte-brates. In many streams, the most common fooditems are larvae of extremely abundant chirono-mid flies that lay their eggs on dampened parts ofrocks near the water line. The pattern of recruit-ment of larvae from the ocean and the behavior ofadults indicate that fish of this species spawn yearround.

When migrating young fishes leave the oceanand move upstream, two species (‘o’opu ‘akupaand ‘o’opu naniha) typically penetrate upstreamno further than the first waterfall or cascade(Fitzsimons and Nishimoto 1991). Three specieshave well-developed sucking disks (fused pelvicfins) and are able to climb waterfalls as high asabout 20 m (‘o’opu nakea), 30 m (‘o’opu nopili),and more than 300 m (‘o’opu ‘alamo’o). Streamsthat have a short estuary and gentle grade in thelower reaches are likely to have all five species.Ones with a very long estuary may not include‘o’opu ‘alamo’o. Conversely, streams that end in awaterfall dropping directly on the beach from aheight of 15 m or more are likely to include ‘o’opu‘alamo’o and perhaps ‘o’opu nopili but probablynot the other species.

Species Status

Early stream surveys often missed species, especial-ly ‘o’opu ‘alamo’o (Timbol et al. 1980), because theeffect of stream morphology on species presenceand distribution was unknown. Until recently,‘o’opu ‘alamo’o was thought to be extinct on theisland of O’ahu (Higashi and Yamamoto 1993), butall species are now known (Devick et al. 1992) from

windward streams on each of the major highislands (Kaua’i, O’ahu, Moloka’i, Maui, andHawai’i). Over a decade ago, ‘o’opu ‘alamo’o waslabeled as endangered, and ‘o’opu ‘akupa, ‘o’opunakea, and ‘o’opu nopili were listed as species ofspecial concern in a publication by the AmericanFisheries Society (AFS; Williams et al. 1989). Subse-quent surveys by the Hawai’i Division of AquaticResources leading to the publication of the Hawai’iStream Assessment (Hawaii Cooperative Park Ser-vice Unit 1990) and the establishment of astatewide stream database have shown that the fivespecies are not rare even on the island of O’ahuwhere they are now recorded from seven streams(Devick et al. 1995; Fitzsimons et al. 2002b). Cur-rently, none of the species is listed as endangered,threatened, or of special concern by AFS, the U.S.Fish and Wildlife Service (Threatened and Endan-gered Species System), or the Hawai’i BiologicalSurvey (Bishop Museum). However, because eachisland has a wet and dry side, and dry-side develop-ment requires water, water will always be a covetedand limited resource in Hawai’i and stream animalswill remain vulnerable.

Instream Flow Incremental Methodology,Index of Biotic Integrity, Reference ConditionApproach, and Pacific-Asia Biodiversity Transect Network

Typical Hawaiian streams resemble high qualitytrout streams on the North American continent,but the analogy is difficult to extend beyond super-ficial similarities because of frequent flash floodsand a marine life history stage for fishes and largeraquatic invertebrates (Fitzsimons and Nishimoto1997). These distinctions have made it imperativeto develop stream survey methods that are appro-priate for Hawai’i. Techniques arguably effective inmainland streams either cannot be used in islandstreams or are sharply limited in their applicability.Three examples are mentioned here.

Traditional sampling methods employing elec-troshocking, seining, or installing set nets are diffi-cult to use in boulder-strewn island streams subjectto sudden flooding. Between floods, however,underwater visibility may be 10 m or more; there-fore, visual sampling has become standard.

62



Instream flow incremental methodology(IFIM) was developed by the U.S. Fish and WildlifeService to provide, among other things, standardtechniques for recommending minimum flowrequirements for individual streams on the NorthAmerican mainland (Estes and Osborn 1986).Although certain measurements in IFIM proce-dures can be used in Hawai’i for physical character-izations of streams (depth, velocity, substrate, andcover), the continuous seesaw effect of changingwater levels that occurs naturally in island streamsmakes IFIM mostly inapplicable for Hawai’i. Fre-quent flash floods of a magnitude that would bedisastrous on a continent and the subsequentreturn to greatly reduced flows and clear water inas few as 2 or 3 d make the setting of a precise min-imum flow requirement for a Hawaiian streamboth arbitrary and unachievable.

Procedures for developing an index of bioticintegrity (IBI) or similar measure include usefulitems for describing stream conditions at a discretepoint in time (Karr 1981; Karr et al. 1986). Howev-er, assessing the health (“biotic integrity”) of aHawaiian stream during a single visit by scoringphysical conditions and by using species richnessand numbers of individuals of indigenous fishes(the usual approach) or macroinvertebrates isinappropriate. A stream with a full complement ofanimals can receive, for example, a score of 10 inrespect to water clarity in the morning, a zero dur-ing an afternoon flood, and back to a 10 two orthree days later. Although once used commonly bymainland fishery biologists, evaluating streams bysimply scoring the major aquatic species as presentor absent can be misleading because the occur-rence and local density of Hawaiian species arestrongly influenced by stream topography. Islandstreams differ significantly from continentalstreams in respect to population origins; everynative fish and larger invertebrate (crustaceans,mollusks) in every Hawaiian stream is a migrant.Adult animals in a stream may or may not (the lat-ter is more likely) have begun life in that stream,and the young animals in the same stream proba-bly are not their offspring. Therefore, the numberof age-classes and number of individuals per classcannot be used to estimate a particular stream’sproductivity. Because indigenous stream fishes and

macroinvertebrates are benthic species, counts ofanimals for use in population estimates can also bedeceptive. At low water, the number of individualsper unit volume of water or per area of stream bot-tom will appear higher than the same number ofanimals when water level is higher and the animalsare dispersed. Thus, the ranking of Hawaiianstreams according to their “biological value” froma subjective scoring of physical features, the num-ber of species present, or the number of individu-als per unit area or volume is deceptive. The mis-take is compounded when the composite score fora stream is compared with a reference stream inHawai’i as a basis for recommending maintenance,special protection, or development. A referencestream, regardless of its high score, inevitably willchange. The recently proposed reference conditionapproach (RCA) uses naturally occurring variabil-ity among streams and stream animals minimallyexposed to human stressors as a reference forassessing individual streams (Bailey et al. 2004).This approach is logical for Hawaiian streams.

After more than a decade of collaboration bypersonnel in the Hawai’i Division of AquaticResources and the Louisiana State UniversityMuseum of Natural Science, a procedure forstream surveys in Hawai’i and other islands of thetropical Pacific was designed specifically toaccommodate the amphidromous life cycles ofstream fishes and larger invertebrates and the fre-quent flash floods that characterize high-islandstreams (Fitzsimons et al. 2005). These techniquesare included in chapter 7 of Biodiversity Assess-ment of Tropical Island Ecosystems, Pacific-AsiaBiodiversity Transect Network (PABITRA) Manualfor Interactive Ecology and Management (Fitzsi-mons et al. 2005). Pacific-Asia Biodiversity Tran-sect Network is a “collaborative program for inves-tigating the function of biodiversity and the healthof ecosystems in the tropical Pacific Islands.” Adescription of PABITRA, an outline of the manual,and chapter contents are available at the Web sitehttp://www.botany.hawaii.edu/pabitra/. The pri-mary purpose of freshwater survey and monitoringprocedures described in the PABITRA chapter isnot to rank streams but rather to determinewhether or not a given stream has an expected nat-ural complement of aquatic species. However, the

63



information needed to answer this relatively simplequestion can be applied to inquiries of much broad-er scope. The use of GIS as an organizer for fielddata describing physical and biological features of astream can identify associations over time and spacewithin a single stream or between similar streams.These associations become the basis for recommen-dations regarding the removal of water withoutaffecting habitat for stream animals or the additionof water for stream restoration. The extension of theGIS model to include an entire watershed establish-es the ahupua’a concept (Kamehameha Schools1994) as the framework for management and con-servation decisions among Hawaiian streams. Anahupua’a is a land division used by the early Hawai-ians. It encompasses the area from the back andboth sides of a valley out into the ocean as far as theseaward edge of the coral reef, and it provided all theresources (food, clothing, shelter, etc.) required bypeople living in a well-demarcated ecosystem ofwhich they were an integral part.

Status of Fish Surveys

Surveys of native stream fishes have been complet-ed for 177 streams. These represent 47.1% of the376 perennial streams that occur on the five majorislands. Geographical information systems-compat-ible data compiled from these and ongoing surveyswill soon be available for viewing at the Web site forthe Division of Aquatic Resources (http://www.hawaii.gov/dlnr/dar).

Mandated Protection for Native Stream Fishes

The decline of sugar cane production in Hawai’iduring the past dozen years has prompted demandsfor the return of water to the windward sides ofislands. The most widely publicized example hasbeen the Waiahole contested case, which included 52d of hearings, testimony from 161 witnesses, and 567exhibits introduced into evidence (Dingeman 1997).The outcome was the return of water, diverted sincethe 1920s, back into the basin of Waiahole Stream onO’ahu and a decision in August 2000 by the Hawai’iSupreme Court charging the state of Hawai’i with arequirement to assure “the maintenance of optimumflow for native fishes” throughout the state.

In response to the mandate from the Hawai’iSupreme Court to maintain optimum flow forstream fishes, the Division of Aquatic Resourcesestablished three working principles for decisionmaking that would take into account the inabilityto set precise minimum flow standards because ofnaturally fluctuating water levels, the importance ofrecognizing the entire watershed as a functioningecosystem, and the significance of amphidromouslife cycles among the principal aquatic species.Because Hawaii’s stream fishes and macroinverte-brates are benthic animals with remarkable species-specificity for occupying discrete habitats withincertain sections of streams (Nishimoto and Fitzsi-mons 1986; Fitzsimons and Nishimoto 1991; Fitzsi-mons et al. 1997), it is possible to obtain preciseinformation on the nature, size, and location ofhabitats required by these stream dwellers. Theirstereotypic, predictable behavior allows the investi-gator to focus directly on the animals themselves todetermine the effect of removing or adding water toall or part of a stream in lieu of setting arbitraryminimum-flow requirements that would be biolog-ically irrelevant and unattainable. Policies estab-lished by the Division of Aquatic Resources forwater-use decisions in Hawai’i are (1) no net loss ofhabitat for native fishes, (2) use of a watershed orahupua’a approach, and (3) the maintenance ofopen stream mouths that provide ready access fornative species migrating into and out of the sea.The preservation of indigenous stream fishes nowhas been elevated to the highest possible level ofprotection in the state of Hawai’i.

ReferencesArmstrong, R. W., editor. 1983. Atlas of Hawai’i. 2nd edition.

University of Hawaii Press, Honolulu.Bailey, R. C., R. H. Norris, and T. B. Reynoldson. 2004.

Bioassessment of freshwater ecosystems: using the refer-ence condition approach. Kluwer Academic PublishersNorwell, Massachusetts.

Balon, E. K. 1990. Epigenesis of an epigeneticist: the develop-ment of some alternative concepts on the early ontogenyand evolution of fishes. Guelph Ichthyology Reviews1:1–42.

Devick, W. S., J. M. Fitzsimons, and R. T. Nishimoto. 1992.Conservation of Hawaiian freshwater fishes. State ofHawaii, Department of Land and Natural Resources,Division of Aquatic Resources, Honolulu.

64



Devick, W. S., J. M. Fitzsimons, and R. T. Nishimoto. 1995.Threatened fishes of the world: Lentipes concolor Gill, 1860(Gobiidae). Environmental Biology of Fishes 44:325–326.

Dingeman, R. 1997. Water divided up: state panel drafts planfor Waiahole. Honolulu Advertiser (July 16):A1, A9.

Estes, C. C., and J. F. Osborn. 1986. Review and analysis ofmethods for quantifying instream flow requirements.Water Resources Bulletin 22:3l89–398.

Fitzsimons, J. M., M. G. McRae, H. L. Schoenfuss, and R. T.Nishimoto. 2003. Gardening behavior in the amphidro-mous Hawaiian fish Sicyopterus stimpsoni (Osteichthyes:Gobiidae). Ichthyological Exploration of Freshwaters14:185–191, München, Germany.

Fitzsimons, J. M., and R. T. Nishimoto. 1991. Behavior of gob-ioid fishes from Hawaiian fresh waters. Pages 106–124 inW. S. Devick, editor. New directions in research, manage-ment, and conservation of Hawaiian freshwater streamecosystems: Proceedings of the 1990 Symposium onStream Biology and Fisheries. State of Hawaii, Depart-ment of Land and Natural Resources, Division of AquaticResources, Honolulu.

Fitzsimons, J. M., and R. T. Nishimoto. 1995. Use of fishbehavior in assessing the effects of Hurricane Iniki on theHawaiian island of Kaua’i. Environmental Biology ofFishes 43:39–50.

Fitzsimons, J. M., and R. T. Nishimoto. 1997. Hawaiian streamsand trout streams - an imperfect analogy. Pages 345–353 inW. S. Devick, editor. Proceedings of the 1996 Annual Meet-ing of the Western Association of Fish and Wildlife Agen-cies. Hawaii Division of Aquatic Resources, Honolulu.

Fitzsimons, J. M., R. T. Nishimoto, and W. S. Devick. 1996.Maintaining biodiversity in freshwater ecosystems onoceanic islands of the tropical Pacific. Chinese Biodiversi-ty 4(supplement):23–27.

Fitzsimons, J. M., R. T. Nishimoto, and J. E. Parham. 2005.Stream ecosystems. Pages 105–138 in D. Mueller-Dom-bois and K. Bridges, editors. Biodiversity assessment oftropical island ecosystems: PABITRA manual for interac-tive ecology and management. University of Hawai’iPress, Honolulu.

Fitzsimons, J. M., J. E. Parham, L. K. Benson, and M. G.McRae. 2002b. Biological assessment of Kahana Stream,Island of O’ahu. Report to the State of Hawaii, Depart-ment of Land and Natural Resources, Division of AquaticResources and the Commission on Water Resource Man-agement, Honolulu.

Fitzsimons, J. M., J. E. Parham, and R. T. Nishimoto. 2002a.Similarities in behavioral ecology among amphidro-mous and catadromous fishes on the oceanic islands ofHawai’i and Guam. Environmental Biology of Fishes65:123–129.

Fitzsimons, J. M., H. L. Schoenfuss, and T. C. Schoenfuss.1997. Significance of unimpeded flows in limiting thetransmission of parasites from exotics to Hawaiianstream fishes. Micronesica 30:117–125.

Hawai’i Cooperative Park Service Unit. 1990. Hawai’istream assessment, a preliminary appraisal of Hawai’i

stream resources. Report R84. Prepared for the Com-mission of Water Resources Management. Hawai’iCooperative Park Service Unit, Honolulu.

Higashi, G. R., and M. N. Yamamoto. 1993. Rediscovery of“extinct” Lentipes concolor (Pisces: Gobiidae) on theIsland of O’ahu, Hawai’i. Pacific Science 47:115–117.

Kamehameha Schools. 1994. The Ahupua’a. Life in earlyHawai’i. Kamehameha Schools Bernice Pauahi BishopEstate, Honolulu.

Karr, J. R. 1981. Assessment of biotic integrity using fishcommunities. Fisheries 6(6):21–27.

Karr, J. R., K. D. Fausch, P. L. Angermeier, and P. R. Yant.1986. Assessing biological integrity in running waters: amethod and its rationale. Illinois Natural History Survey,Special Publication 5, Urbana.

McDowall, R. M. 1992. Diadromy: origins and definitions ofterminology. Copeia 1992:248–251.

Nelson, J. S. 1984. Fishes of the world. 2nd edition. Wiley,New York.

Nishimoto, R. T., and J. M. Fitzsimons. 1986. Courtship, ter-ritoriality, and coloration in the endemic Hawaiianfreshwater goby, Lentipes concolor. Pages 811–817 in T.Uyeno, R. Arai, T. Taniuchi, and K. Matsuura, editors.Indo-Pacific fish biology. Ichthyological Society of Japan,Tokyo.

Nishimoto, R. T., and D. G. K. Kuamo’o. 1997. Recruitmentof goby postlarvae into Hakalau Stream, Hawai’i Island.Micronesica 30:41–49.

Pukui, M. K., S. H. Ebert, and E. T. Mookini. 1983. The pock-et dictionary, with concise Hawaiian grammar. Universi-ty Press of Hawaii, Honolulu.

Radtke, R. L., R. A. Kinzie, and S. D. Folsom. 1988. Age atrecruitment of Hawaiian freshwater gobies. Environ-mental Biology of Fishes 23:205–213.

Smith, R. J. F., and M. J. Smith. 1998. Rapid acquisition ofdirectional preferences by migratory juveniles of twoamphidromous Hawaiian gobies, Awaous guamensis andSicyopterus stimpsoni. Environmental Biology of Fishes53:275–282.

Tate, D. C., J. M. Fitzsimons, and R. P. Cody. 1992. Hawaiianfreshwater fishes (Osteichthyes, Gobiodei): a field key tothe species of larvae and postlarvae during recruitmentinto fresh waters. Occasional Papers of the Museum ofNatural Science, Louisiana State University 65:1–10,Baton Rouge.

Timbol, A. S., A. J. Sutter, and J. D. Parrish. 1980. Distribution,relative abundance, and stream environment of Lentipesconcolor (Gill 1860), and associated fauna in Hawaiianstreams. Hawaii Cooperative Fishery Research UnitReport, prepared for Hawaii Division of Fish and Gameand U.S. Fish and Wildlife Service. University of Hawaiiat Manoa, Water Resources Research Center, Honolulu.

Williams, J. E., J. E. Johnson, D. A. Hendrickson, S. Contr-eras-Balderas, J. D. Williams, M. Navarro-Mendoza, D. E.McAllister, and J. E. Deacon. 1989. Fishes of NorthAmerica - endangered, threatened, or of special concern:1989. Fisheries 14(6):2–20.

65



67

Outside of anadromous salmonids and afew endangered species, the biology ofnative freshwater fishes of western North

America is poorly known (Bruton 1995; Ruck-elshaus et al. 2002). Studies on habitat, range of tol-erance to environmental variables, species interac-tions, life history patterns, and other species levelinquiries are conspicuously lacking for nonendan-gered native species. Even rarer are studies thatcharacterize variation in morphology (McElroyand Douglas 1995), behavior, genetics, and life his-tory (Baltz and Moyle 1982; Baker et al. 1998)among populations within a species— variationthat determines evolutionary substructure withinthe species.

This dearth of information is the single greatestbarrier to effective conservation of the uniquewestern fish fauna (Bruton 1995). Although legis-lation designed to conserve species has been in

place for several years (Williams and Deacon1991), lack of information about the status andbiology of species makes it difficult to apply legisla-tive protections in a defensible and effective way.Even for species that do not require legislative pro-tection, lack of biological information makes it dif-ficult to determine appropriate management activ-ities. As such, management activities are usuallybased on anecdotal information or overgeneraliza-tions of broad ecological theories. Similarly,because of the lack of information about variationamong populations within a species, few conserva-tion efforts are implemented at a scale larger thanthe local population.

The antidote to this state of affairs is to focusefforts on understanding native species biologybefore crises of rarity or resource conflicts arise.What do we need to know to create effective man-agement plans for native species? This question is

Biological Status of Leatherside Chub: A Frameworkfor Conservation of Western Freshwater Fishes

ABSTRACT Outside of anadromous salmonids and a few endangered species, the biology of native freshwater

fishes of western North America is poorly known. What do we need to know to effectively manage native species

and avoid decline and extinction? A recent analysis of the role of science in the Pacific salmon controversy outlines

a clear framework for biological evaluation and management of native species. This framework has three compo-

nents: (1) determine the status of populations based on genetic and ecological variation, (2) identify and quantify

threats to populations, and (3) determine actions to alleviate threats and promote conservation of populations. We

use our studies of leatherside chub Gila copei (also known as Snyderichthys copei), a small cyprinid native to the

Bonneville basin and upper Snake River drainage, as a case study to illustrate the application of this research and

management framework. Recent surveys have revealed dramatic reductions in range of leatherside chub over the

last 50 years. Genetic, morphometric, and ecological studies all indicate that leatherside chub comprise two dis-

tinct species. Leatherside chub is threatened by both habitat degradation and introduced brown trout Salmo trut-ta, and the interaction between these two threats exacerbates negative effects.We conclude by showing how stud-

ies of leatherside chub can inform and influence management, conservation, and habitat restoration activities.

Mark C. Belk and Jerald B. Johnson



MARK C. BELK Department of Integrative Biology, Brigham Young University, Provo, Utah 84602, USA, [email protected]

JERALD B. JOHNSON Department of Integrative Biology, Brigham Young University, Provo, Utah 84602, USA; Northwest Fisheries Science Center, 2725 Montlake Boulevard East, Seattle, Washington 98112, USA, [email protected]


faced by all those charged with managing or con-serving a species about which little is known. Onthe one hand, almost any type of information mayprove useful in some way; however, managers mustdecide how best to allocate limited funds and guardagainst paying for studies that provide little usefulinformation for management planning. A recentanalysis of the role of science in the Pacific salmoncontroversy outlines a clear framework for biolog-ical evaluation and management of native species(Ruckelshaus et al. 2002). This framework hasthree components: (1) determine the status of pop-ulations based on genetic and ecological variation,(2) identify and quantify threats to populations,and (3) determine actions to alleviate threats andpromote conservation of populations. Typically,components 1 and 2 are viewed as research projectsto obtain information necessary to implementcomponent 3. However, component 3 can alsoinvolve research, for example, to assess the effec-tiveness of different management actions.

Leatherside chub Gila copei (also known as Sny-derichthys copei [Nelson et al. 2004]) is typical ofmany nongame fish species in western NorthAmerica in that until recently, little was known ofthe status or biology of it. It was originallydescribed from specimens obtained from the BearRiver near Evanston, Wyoming by Jordan andGilbert (1881). Leatherside chub is a small cyprinidnative to streams of the Bonneville basin and upperSnake River drainage. General sources noted thatthey were found in cool flowing waters and thatthey were regarded as “excellent bait minnows”(Sigler and Miller 1963; Baxter and Simon 1970).They were reported as abundant in many locations,and large collections of specimens are available atthe Smithsonian National Museum of Natural His-tory and the University of Michigan’s Museum ofZoology. Beyond these few general observations,nothing was known. Clearly, a better understand-ing of the biology of the species is required to makeinformed decisions about management actions.

In this paper, we use studies of leatherside chubconducted in the last 10 years to assess the currentbiological status of the species and to illustrate theutility of the framework outlined above. Recentsurveys have revealed dramatic range reductionsover the last 50 years, especially in the northern

populations. We show that genetic, ecological, andmorphometric studies all indicate that leathersidechub comprise two distinct species. We show thatleatherside chub is threatened by both habitatdegradation and introduced brown trout Salmotrutta and that the interaction between these twothreats exacerbates negative effects. We highlightan ongoing study to evaluate the threat of rangefragmentation on population persistence andgenetic variability of leatherside chub. We con-clude by showing how studies of leatherside chubhave informed and influenced management, con-servation, and habitat restoration activities, andwe provide recommendations for future manage-ment actions.

Status of Populations, Genetic and Ecological Variation

Determining the status of populations involves twodistinct types of information. The first is to deter-mine the current distributional status of the species.The second is to determine the pattern of variationin genetic, ecological, and environmental relation-ships among populations within the species.

Distributional status

Determining current distribution involves exhaus-tive surveys and comparison to historic surveysand records. Such surveys can be done in combina-tion with surveys for other species. However, unlessthose doing the surveying are carefully instructedas to the importance of documenting all species,small, rare, or inconspicuous species are oftenmissed or not recorded. For example, introduced,piscivorous sport fishes may induce habitat shiftsin vulnerable native species, such that typical sur-vey methods would fail to detect all species (e.g.,Chapman et al. 1996). Understanding the numberand distribution of extant populations, and the rel-ative abundance of the species in each location,provides the basis for assessing threats and formu-lating management plans for the species as a whole(Ruckelshaus et al. 2002).

Surveys of distribution of leatherside chub overthe last decade have revealed some surprising pat-terns of local extirpation and population fragmen-tation. Leatherside chub have disappeared from

68

BELK AND JOHNSON


several entire drainages (e.g., Beaver River, centralUtah, Little Wood River, south-central Idaho), andwithin drainages where they still exist, their distri-bution has declined and become highly fragment-ed (e.g., Sevier River, central Utah, Bear River,Utah, Wyoming, Idaho). For example, comparisonwith historic distributions has resulted in an esti-mate of 42% range reduction in the Sevier Riverdrainage (central Utah) over the last 100 years(Wilson and Belk 2001), and this drainage is con-sidered to have the most extensive and abundantpopulations of leatherside chub anywhere. In con-trast to the trend of declining distribution, severalpopulations, once thought to be eliminated, havebeen rediscovered (e.g., lower Sevier River, JuabCounty, Utah; tributaries of the upper Bear River,Summit County, Utah; Pacific Creek, Teton Coun-ty, Wyoming). Overall, leatherside chub probablyoccupy less than half of their historic range, andmany populations are isolated remnants of a previ-ously highly connected distribution.

Variation among populations

Understanding genetic, ecological, and environ-mental relationships among populations withinthe species involves assessment of genetic variationand ecological similarity within and among popu-lations at various spatial and temporal scales(Crandall et al. 2000; Rader et al. 2005). Tools forobtaining such information include phylogeneticanalysis using mitochondrial and nuclear genesequence data and microsatellites. In addition tothe use of molecular genetic data, it is important touse common-environment or reciprocal transplantexperiments to assess ecological variability and thegenetic basis of phenotypic variation among pop-ulations (Rader et al. 2005). Some species may becomposed of a relatively homogeneous collectionof populations; however, given the naturally frag-mented nature of aquatic systems and the highlyvariable geologic history of western North Ameri-ca, genetic and ecological relationships amongpopulations of native fish species are likely to behighly heterogeneous and complex.

Information about genetic, ecological, andenvironmental relationships among populationswithin the species will allow determination of theextent and boundaries of subdivisions within the

species. Clusters of populations that have experi-enced different evolutionary histories from otherpopulations have been referred to as evolutionarysignificant units (ESUs; Crandall et al. 2000). Eval-uation of ESUs, can help guide the designation ofbiologically relevant subunits of the speciesreferred to as management units (MUs; Vrijenhoek1998; Crandall et al. 2000). In addition, such infor-mation might suggest selectively important envi-ronmental or biological conditions (e.g., tempera-ture regimes, predator–prey relationships, etc.)that might influence future population growth ormanagement actions.

Recent research on genetic and ecological vari-ation among populations of leatherside chub hasled to interesting taxonomic results. Analysis ofmtDNA sequence data (Johnson and Jordan 2000;Dowling et al. 2002), and more recently nuclearDNA sequence data (Johnson et al. 2004), indicatesthat leatherside chub comprise two species ratherthan one as previously assumed. Furthermore,leatherside chub do not belong in the genus Gila,but rather they fit nicely within the genus Lep-idomeda, and the two species appear to be nonsis-ter taxa (Johnson and Jordan 2000; Dowling et al.2002; Johnson et al. 2004).

The two species occupy distinct geographicregions. Leatherside chub found in the Bear River,and a variety of upper Snake River drainages, com-prise one species (hereafter referred to as northernleatherside chub Lepidomeda copei [Johnson et al.2004]), whereas leatherside chub found in the UtahLake drainage (Provo River, Spanish Fork River,and some minor streams) and the Sevier Riverdrainage comprise another species (hereafterreferred to as the southern leatherside chub Lep-idomeda aliciae [Johnson et al. 2004]). The bound-ary between the northern and southern leathersidespecies is formed by the Weber River drainagelocated in north-central Utah. Although it is men-tioned as being native to the Weber River drainage(Sigler and Miller 1963), curiously, there are nomuseum, historical surveys, or current records ofleatherside chub occurrence in the Weber Riverdrainage, suggesting that this gap in the distribu-tion may represent a real and ancient condition.

In addition, morphometric variation in headshape between northern and southern leatherside

69

BIOLOGICAL STATUS OF LEATHERSIDE CHUB


chub corresponds to the phylogenetic separationobserved in molecular genetic data. Northern andsouthern leatherside chub exhibit distinct headshapes. The northern leatherside chub has a short-er, more rounded rostrum and is somewhat dis-tinct compared to the other species of Lepidomeda.The southern leatherside chub has a more pointedrostrum, similar to the other species of Lepidome-da (Johnson et al. 2004).

Ecological variation among populations alsosupports designation of two species of leathersidechub. Common-environment experiments includ-ing representatives of both species show signifi-cant, apparently genetically based differences intemperature-specific growth (Belk et al. 2005).Southern leatherside chub experience highergrowth rates at temperatures above about 19°Ccompared to northern leatherside chub. However,at temperatures below 19°C, northern leathersidechub grow significantly faster than southernleatherside chub. A corresponding analysis of timeof hatching in northern versus southern leather-side chub suggests that the northern species may bespawning at lower temperatures, but at roughly thesame time of year, as the southern species. Takentogether, these data suggest local adaptation to dif-fering environmental temperature regimes innorthern versus southern leatherside chub (John-son et al. 2004; Belk et al. 2005).

Obviously, northern and southern leathersidechub must be managed separately. The next ques-tion is whether there is evolutionarily significantvariation among populations within the twospecies that might affect management or restora-tion activities. The short answer is we do not knowyet. Currently, we are conducting a range-widestudy of genetic variation in the southern leather-side chub aimed at detecting ESUs within thespecies. An added question that will be addressedby this study is what is the effect of recent fragmen-tation on patterns of genetic variation among pop-ulations. In 2 to 3 years, we will have sufficient datato recommend appropriate units for managementthat would correspond to ESUs, and we will also beable to determine the effect of recent barriers ongenetic variation within and among populations.Given the current rarity of the northern species, asimilar study for populations of the northern

leatherside chub is imperative.

Identification of Population Threats

Immediate threats to populations will, in mostcases, be ecological in nature and will consist of (1)habitat loss and degradation, (2) effects of intro-duced species, and possibly (3) overexploitation(see Bruton 1995). Populations that have beenreduced to small size by the above-named ecologi-cal threats may become susceptible to geneticthreats such as inbreeding or hybridization.

Habitat loss

Given the large-scale development of water inwestern North America, there are few, if any, aquat-ic habitats that have not undergone significantchange over the last century (Minckley and Dou-glas 1991). Thus, almost by definition, all aquaticspecies have experienced habitat loss or degrada-tion to some degree. How do we determine the sig-nificance of the threat to species persistence of thischange in habitat? First, we must determine habitatrequirements or preferences for the species. Thiscan be done by measuring habitat use relative toavailability in representative locations. Second, wemust assess the degree of loss or degradation acrossthe species range and the population level conse-quences. Habitat loss leads to an overall loss ordecrease in abundance of populations; however,and maybe more importantly, it can lead to disrup-tion of among population dynamics and otherfragmentation phenomena (e.g., source-sinkdynamics, metapopulations, and gene flow; Meffe1986; Fagan et al. 2002). Knowledge of extent andeffects of fragmentation is essential for effectivespecies-wide conservation efforts.

Leatherside chub are threatened by habitat loss(almost all following observations on threats toleatherside chub refer to the southern species). Fac-tors leading to habitat loss include complete dewa-tering over extended periods, channelization, andconstruction of barriers (e.g., dams, diversions,etc.). Leatherside chub appear to have broad toler-ance of rather extreme environmental conditions(Wilson and Belk 2001; M. C. Belk, personal obser-vation). They can persist in remnant pools in thestreambed for several weeks after the water flow hasbeen completely eliminated. We have found thin,

70

BELK AND JOHNSON


but living leatherside chub in such pools after allother species have died (Belk, personal observa-tion). As such, leatherside chub appear to be adapt-ed to periodic, short-term (a few days to weeks) lowwater conditions such as during seasonal droughts.However, where dewatering has occurred over aprolonged period, leatherside chub have becomelocally extirpated (e.g., Panguitch River below Pan-guitch Reservoir, Garfield County, Utah).

Channelization affects leatherside chub popula-tions by decreasing the complexity of habitat avail-able. In the absence of introduced species, leather-side chub prefer pools and pockets of relativelylow-velocity water in the midst of higher velocityhabitats (Wilson and Belk 2001). Diet analysis sug-gests they consume both terrestrial and aquaticinvertebrate prey from the drift, much like smallsalmonids (Bell and Belk 2004). Channelizationtends to homogenize habitats into run and deepriffle segments of larger than natural extent. Runsand riffles do not provide the habitat necessary forleatherside chub, resulting in local extirpation orreduced population size in channelized areas(Ellsworth 2003).

In areas where introduced salmonids are abun-dant, channelization may be even more disturbingto leatherside chub populations because of therelationship between habitat use and the presenceof predators. In the presence of predatorysalmonids, leatherside chub shift their habitat useaway from main channel pools to off-channel habi-tats such as backwaters and cutoff pools (Walser etal. 1999; Olsen and Belk 2005). Off-channel habi-tats provide refuge for leatherside chub becausesalmonids do not preferentially occupy such habi-tats. Channelization results in the loss of off-chan-nel habitats; thus, leatherside chub have no refugehabitat available, and populations disappear(Ellsworth 2003).

Barriers in flowing water systems affect leather-side chub in two ways. First, leatherside chub donot appear to persist in lakes, ponds, or reservoirs.We are aware of no populations in nonflowingwater. Thus, the creation of reservoirs replacesappropriate habitat with poor habitat for leather-side chub. Second, barriers such as dams and diver-sions impede movements and fragment once con-tinuous populations (e.g., Schaefer et al. 2003). We

are currently studying the effect of recent barrierson genetic structure among populations of leather-side chub. In addition, we are conducting amark–recapture study in several locations to deter-mine the rate of movement and variation indemography among populations.

Introduced species

Another threat to native species is the widespreadintroduction of exotic species. The fish fauna ofwestern North America is depauperate comparedto other regions of the continent, especially if weconsider large-bodied food or sport fishes (onceagain, excepting anadromous salmonids of thePacific Northwest; Minckley and Douglas 1991). Inthe latter part of the 19th and first half of the 20thcenturies, the typical response to such depauperateconditions was to introduce species from otherregions to provide food, recreational opportuni-ties, or other resource management activities (e.g.,control of other fish or mosquitoes). This commonmanagement response led to the introduction(both intentional and unintentional) of hundredsof species of fish throughout western North Amer-ica for everything from food (e.g., carp, catfish) tosport (e.g., bass, walleye) to mosquito control(western mosquitofish Gambusia affinis) (Miller etal. 1991). Many introduced species have becomethe dominant species in the new system to thedetriment of native species.

Introductions of exotic species were donewith little understanding or concern about thepotential effects on native fauna. We still under-stand little about the actual mechanisms bywhich introduced species impact native species.Interactions and mechanisms of impact of intro-duced and native species can be assessed by anexperimental ecology approach. Both competi-tion and predation can be assessed with replicat-ed enclosures or tanks following a factorialapproach (e.g., Mittelbach 1988; Belk 1993; Millset al. 2004). Such experiments will yield informa-tion on the strength of various ecological interac-tions and thus allow determination of specificcomponents of the threat of introduced species.Some species or life stages may not lend them-selves to controlled factorial experiments (e.g.,long-lived or large-bodied species). Inferences

71



about such species must be made from correla-tive and comparative observational studies (e.g.,Levin et al. 2002).

Understanding interactions and mechanisms ofimpact of introduced species can be further com-plicated by the fact that fish populations are size-structured. As such, size-structure of both theintroduced and native species must be consideredin determining effects of introduced species. Dif-ferent size-classes may have different diets, habitat-use patterns, vulnerability to predation, and com-petitive abilities (Werner and Gilliam 1984; Polisand Holt 1992). Thus, the number of potentialinteractions between a native and introducedspecies is actually the product of the types of inter-action (e.g., competition and predation) and thenumber of size-classes or ontogenetic stages. Forexample, two similar-sized species of fish can inter-act as competitors in each of two size-classes (juve-niles and adults). Adults may also compete with orprey on juveniles of the other species. Thus, ratherthan one potential interactive pathway (competi-tion), there may be as many as six interactive path-ways (four competitive pathways and two preda-tion pathways; e.g., Olson et al. 1995). Suchcomplexity of ecological interactions must beexplored and understood to determine effectiveefforts aimed at reducing the negative effect ofintroduced species (Belk et al. 2001).

The most widely distributed and abundantintroduced species in streams and rivers of theBonneville basin and upper Snake River drainage isbrown trout (Sigler and Sigler 1987). Brown troutare resilient and a favorite sport fish. They are com-paratively more piscivorous than trout native towestern North America, and when introduced tonew regions, they can have detrimental impacts onnative species (Garman and Nielsen 1982;Townsend 1996; Penczak 1999; Museth et al. 2003).

Several lines of evidence suggest that browntrout have strong negative effects on leathersidechub. Current distribution of leatherside chub isweakly negatively correlated with abundance ofbrown trout (Wilson and Belk 2001). In the pres-ence of brown trout, leatherside chub occupy refugehabitats almost exclusively (Walser et al. 1999;Olsen and Belk 2005). In short-term survival exper-iments, leatherside chub experienced high mortali-

ty rates from brown trout (Nannini and Belk 2006).The ongoing mark–recapture project mentionedabove is being conducted in populations with andwithout brown trout. Mortality and growth esti-mates from this project will provide information onboth the direct predatory effect of brown trout andthe nonlethal effect resulting from forced use ofrefuge habitats and restricted movement.

It appears that in areas with complex naturalhabitats, leatherside chub may be able to coexistwith brown trout, albeit at relatively low densities(Olsen and Belk 2005). It is not clear whether thiscoexistence is stable or transitory. The ongoingmark–recapture experiment will provide data todetermine whether coexistence of leatherside chuband brown trout is a long-term possibility.

Overexploitation

Finally, in some species, overexploitation byhumans may be an important threat to long-termpersistence of the population (Bruton 1995). Typi-cally, we think overexploitation will manifest as areduction in population abundance or, more often,as a decline in catch. However, in long-livedspecies, such indicators appear to have a long lagtime potentially leading to catastrophic declines(Hutchings and Myers 1994; Shelton and Lilly2000). A better indicator of population abundanceand effects of exploitation is population age struc-ture (Doak and Morris 1999; Holmes and York2003). Age-growth studies provide a good index ofa species life history and provide a baseline toassess effects of exploitation. For exploited popula-tions, information about individual fecundity, age(and size) at maturity, and longevity are critical fordevelopment of predictive models to inform har-vest regulations and other management activities(Holmes and York 2003).

Commercial exploitation is not currently athreat to leatherside chub populations. Some pop-ulations have been harvested for use as bait, butcurrent laws prohibit the take of the species forsuch use. However, most people would not be ableto distinguish between leatherside chub and othersimilar cooccurring species (e.g., redside shinerRichardsonius balteatus, small Utah chub Gilaatraria, and speckled dace Rhinichthys osculus).Thus, inadvertent overexploitation for use as bait

72

BELK AND JOHNSON


may still be a threat especially for small popula-tions. Leatherside chub are relatively long-lived (upto 8 years of age), and they mature at age 2 (John-son et al. 1995). In addition, leatherside chub arehighly vulnerable to seining and electroshockinggears, especially in the fall when they seem toaggregate in deeper pools (Belk, personal observa-tion). For these reasons, leatherside chub may bevulnerable to local overexploitation if baitfish har-vests are not carefully monitored.

Recommendations to Alleviate Threats

Given a reasonable understanding of the status ofpopulations and threats to a species’ persistence,actions can be formulated to promote conserva-tion. One clear advantage to following the aboveframework is that conservation efforts can becoordinated across the species as a whole ratherthan at the single population level. Conservationefforts focused on a single population may harm,or at least not help, the conservation and persist-ence of the species as a whole. For example, usinginappropriate populations as source populationsfor augmentation or repatriation efforts mayharm the source population and result in wastedtime and effort (Stockwell 2003). Some may sug-gest that use of the “nearest neighbor” rule (usethe population in closest geographic proximity)for determining source populations would be suf-ficient; however, this rule assumes that genetic andecological variation follows a simple isolation bydistance pattern. Given the complex geologicalhistory and current disturbed condition of aquat-ic habitats in western North America, departurefrom the simple isolation by distance pattern islikely common. Deciding where to focus conserva-tion efforts and use limited funds to most benefitthe species cannot be done without an under-standing of the relationship among populationsand the nature and extent of threats.

Recent studies on leatherside chub outlinedabove have already influenced management, con-servation, and habitat restoration activities. In1998, leatherside chub was listed as a state sensitivespecies in the state of Utah as a result of the surveywork done in the early 1990s (e.g., Wilson and Belk2001). This designation requires that proposed

activities specifically consider effects on leathersidechub. In addition, a leatherside chub workinggroup within Utah has met several times to coordi-nate research and conservation activities. Twoworkshops (2004 and 2005) have been held withrepresentatives from responsible agencies andorganizations in Idaho, Wyoming, and Utah to dis-cuss, plan, and coordinate conservation activitiesfor leatherside chub. Recent habitat restorationefforts in the Provo River have explicitly includedhabitats designed to encourage persistence andexpansion of populations of leatherside chub.

Additionally, previous and ongoing researchprovides clear recommendations for future conser-vation-related activities for leatherside chub.Genetic and ecological variation among popula-tions suggests that many populations are notexchangeable (Crandall et al. 2000). Additionalclarification of the variance among remnant popu-lations is needed before population augmentationor repatriation efforts proceed, especially amongpopulations of the northern species. Additionalinformation about the genetic structure amongpopulations in the southern species, as part of ourongoing study on effects of fragmentation, will beuseful to determine appropriate source popula-tions for future repatriation efforts.

Our studies of habitat use in leatherside chuband effects of introduced brown trout haverevealed an important interaction between habitatand predation. In streams without brown trout (orother introduced predators), some guaranteedminimal water flow is apparently all that isrequired for leatherside chub to persist. However,in areas with brown trout, habitat restorationactivities must include plans for construction ofoff-channel habitats such as backwaters that pro-vide a refuge for leatherside chub. Simply addingmeanders and pool-riffle sequences will not be suf-ficient to allow leatherside chub to coexist with anintroduced predator. Recent restoration work onthe Provo River in central Utah has provided anopportunity to assess the utility of off-channelhabitats in promoting coexistence of leathersidechub with an abundant population of brown trout.Ongoing monitoring of the development of thefish community in restored sections of the ProvoRiver is needed to assess the success of the project

73



relative to leatherside chub and other native fishes.Because of the aggressive nature and piscivo-

rous habit of brown trout, it may be advisable tostock native cutthroat trout (e.g., Oncorhynchusclarkii utah) for recreational fishing purposes instreams occupied by leatherside chub. Stream-dwelling cutthroat trout are less piscivorous andless likely to pose a significant threat of predationcompared to brown trout. Leatherside chub coe-volved with native cutthroat trout in many loca-tions, so it seems likely that they might be adaptedto persist with native cutthroat trout. We encour-age an experimental evaluation of the relationshipbetween leatherside chub and cutthroat trout.

Finally, fragmentation among populations andpopulation reduction due to brown trout mighthave long-term detrimental effects on populationsof leatherside chub. Historic records indicate thatleatherside chub were quite continuously distrib-uted within major drainages in their range (e.g.,Sevier River, Little Wood River), and they wereoften quite abundant. Species with such distribu-tions may be adapted to high levels of gene flowand social interactions (Meffe 1986). Decreasedlevels of gene flow may lead to isolation and even-tual loss of genetic variability within populations.Populations that are adapted to high populationsizes and attendant social interactions may exhibitAllee effects (e.g., decreased individual fitness atlow population sizes; Courchamp et al. 1999;Stephens 1999) further complicating conservationefforts. We suggest efforts be made to restore habi-tat and decrease brown trout numbers to facilitatereconnection of recently isolated populations ofleatherside chub and to increase population sizes inareas where numbers have been reduced.

Acknowledgments

We acknowledge Utah Division of WildlifeResources, Idaho Fish and Game Department,Wyoming Game and Fish Department, Utah Recla-mation, Mitigation, and Conservation Commissionand the Provo River Project, and the Bonnevillechapter of the American Fisheries Society for pro-viding funding and collecting permits to help withthis research. In addition, we thank the U.S. ForestService in Utah, Idaho, and Wyoming, and Brigham

Young University for funding and other support.We acknowledge the help of D. K. Shiozawa, K.Mock, and T. Dowling. Former postdocs and grad-uate students who have contributed to research onleatherside chub are K. Wilson, M. Smith, C. Walser,D. Houston, D. Olsen, and M. Nannini. We alsoacknowledge numerous undergraduates of BYUwho have worked on various projects.

References

Baker, J. A., S. A. Foster, D. C. Heins, M. A. Bell, and R. W.King. 1998. Variation in female life-history traits amongAlaskan populations of the threespine stickleback, Gas-terosteus aculeatus L. (Pisces: Gasterosteidae). BiologicalJournal of the Linnean Society 63:141–159.

Baltz, D. M., and P. B. Moyle. 1982. Life history characteris-tics of tule perch Hysterocarpus traski populations incontrasting environments. Environmental Biology ofFishes 7:229–242.

Baxter, G. T., and J. R. Simon. 1970. Wyoming fishes.Wyoming Game and Fish Department, Cheyenne.

Belk, M. C. 1993. Growth and mortality of juvenile sunfish-es Lepomis sp. under heavy predation. Ecology of Fresh-water Fish 2:91–98.

Belk, M. C., J. B. Johnson, K. W. Wilson, M. E. Smith, and D.D. Houston. 2005. Variation in intrinsic individualgrowth rate among populations of leatherside chub(Snyderichthys copei Jordan & Gilbert): adaptation totemperature or length of growing season? Ecology ofFreshwater Fish 14:177–184.

Belk, M. C., M. J. Whitney, and G. B. Schaalje. 2001. Complexeffects of predators: determining vulnerability of theendangered June sucker to an introduced predator. Ani-mal Conservation 4:251–256.

Bell, A., and M. C. Belk. 2004. Diet of leatherside chub, Sny-derichthys copei, in the fall. Western North AmericanNaturalist 64:413–416.

Bruton, M. N. 1995. Have fishes had their chips? The dilem-ma of threatened fishes. Environmental Biology of Fish-es 43:1–27.

Chapman, L. J., C. A. Chapman, R. Ogutu-Ohwayo, M.Chandler, L. Kaufman, and A. E. Keiter. 1996. Refugia forendangered fishes from an introduced predator in LakeNabugabo, Uganda. Conservation Biology 10:554–561.

Courchamp, F., T. H. Clutton-Brock, and B. Grenfell. 1999.Inverse density dependence and the Allee effect. Trendsin Ecology and Evolution 14:405–410.

Crandall, K. A., O. R. P. Bininda-Emonds, G. M. Mace, and R.K. Wayne. 2000. Considering evolutionary processes inconservation biology: an alternative to “evolutionary sig-nificant units.” Trends in Ecology and Evolution15:290–295.

Doak, D. E., and W. Morris. 1999. Detecting population-levelconsequences of ongoing environmental change withoutlong-term monitoring. Ecology 80:1537–1551.

74

BELK AND JOHNSON


Dowling, T., C. A. Tibbets, W. L. Minckley, and G. R. Smith.2002. Evolutionary Relationships of the plagopterins(Teleostei: Cyprinidae) from cytochrome b sequences.Copeia 2002:665–678.

Ellsworth, C. M. 2003. Response of the fish community andage class structure to channelization and habitat restora-tion in the Provo River between Deer Creek and Jor-danelle reservoirs. Master’s thesis. Brigham Young Uni-versity, Provo, Utah.

Fagan, W. F., P. J. Unmack, C. Burgess, and W. L. Minckley.2002. Rarity, fragmentation, and extinction risk in desertfishes. Ecology 83:3250–3256.

Garman, G. C., and L. A. Nielsen. 1982. Piscivority bystocked brown trout (Salmo trutta) and its impact on thenongame fish community of Bottom Creek, Virginia.Canadian Journal of Fisheries and Aquatic Sciences39:862–869.

Holmes, E. E., and A. E. York. 2003. Using age structure todetect impacts on threatened populations: a case studywith stellar sea lions. Conservation Biology17:1794–1806.

Hutchings, J. A., and R. A. Myers. 1994. What can be learnedfrom the collapse of a renewable resource? Atlantic cod,Gadus morhua, of Newfoundland and Labrador. Canadi-an Journal of Fisheries and Aquatic Sciences51:2126–2146.

Johnson, J. B., M. C. Belk, and D. K. Shiozawa. 1995. Age,growth, and reproduction of leatherside chub (Gilacopei). The Great Basin Naturalist 55:183–187.

Johnson, J. B., and S. Jordan. 2000. Phylogenetic divergencein leatherside chub (Gila copei) inferred from mitochon-drial cytochrome b sequences. Molecular Ecology9:1029–1035.

Johnson, J. B., T. E. Dowling, and M. C. Belk. 2004. Neglect-ed taxonomy of rare desert fishes: congruent evidencefor two species of leatherside chub. Systematic Biology53:841–855.

Jordan, D. S., and C. H. Gilbert. 1881. Notes on a collectionof fishes from Utah Lake. Proceedings of the UnitedStates National Museum 3:459–465.

Levin, P. S., J. A. Coyer, R. Petrik, and T. P. Good. 2002. Com-munity-wide effects of nonindigenous species on tem-perate rocky reefs. Ecology 83:3182–3193.

McElroy, D. M., and M. E. Douglas. 1995. Patterns of mor-phological variation among endangered populations ofGila robusta and Gila cypha, (Teleostei: Cyprinidae) inthe upper Colorado River basin. Copeia 1995:636–649.

Meffe, G. K. 1986. Conservation genetics and the manage-ment of endangered fishes. Fisheries 11:14–23.

Miller, R. R., C. Hubbs, and F. H. Miller. 1991. Ichthyologicalexploration of the American West: the Hubbs-Miller era,1915–1950. Pages 19–40 in W. L. Minckley and J. E. Dea-con, editors. Battle against extinction: native fish man-agement in the American West. University of ArizonaPress, Tucson.

Mills, M. D., R. B. Rader, and M. C. Belk. 2004. Complexinteractions between native and invasive fish: the simul-taneous effects of multiple negative interactions. Oecolo-gia 141:713–721.

Minckley, W. L., and M. E. Douglas. 1991. Discovery andextinction of western fishes: a blink of the eye in geo-logic time. Pages 7–17 in W. L. Minckley and J. E. Dea-con, editors. Battle against extinction: native fish man-agement in the American West. University of ArizonaPress, Tucson.

Mittelbach, G. G. 1988. Competition among refuging sun-fishes and effects of fish density on littoral zone inverte-brates. Ecology 69:614–623.

Museth, J., R. Borgstrom, T. Hame, and L. A. Holen. 2003.Predation by brown trout: a major mortality factor forsexually mature European minnows. Journal of FishBiology 2003:692–705.

Nannini, M. A., and M. C. Belk. 2006. Antipredator respons-es of two native stream fishes to an introduced predator:does similarity in morphology predict similarity inbehavioral response? Ecology of Freshwater Fish15(4):453–463.

Nelson, J. S., E. J. Crossman, H. Espinosa-Pérez, L. T. Findley,C. R. Gilbert, R. N. Lea, and J. D. Williams. 2004. Com-mon and scientific names of fishes from the UnitedStates, Canada, and Mexico, American Fisheries Society,Special Publication 29, Bethesda, Maryland.

Olsen, D. G., and M. C. Belk. 2005. Relationship of diurnalhabitat use of native stream fishes of the eastern GreatBasin to presence of introduced salmonids. WesternNorth American Naturalist 65:501–506.

Olson, M. H., G. G. Mittelbach, and C. W. Osenberg. 1995.Competition between predator and prey: resource-basedmechanisms and implications for stage-structureddynamics. Ecology 76:1758–1771.

Penczak, T. 1999. Impact of introduced brown trout onnative fish communities in the Pilica River catchment(Poland). Environmental Biology of Fishes 54:237–252.

Polis, G. A., and R. D. Holt. 1992. Intraguild predation: thedynamics of complex trophic interactions. Trends inEcology and Evolution 7:151–154.

Rader, R. B., M. C. Belk, D. K. Shiozawa, and K. A. Crandall.2005. Empirical test for ecological exchangeability. Ani-mal Conservation 8:1–9.

Ruckelshaus, M. H., P. S. Levin, J. B. Johnson, and P. M.Kareiva. 2002. The pacific salmon wars: what sciencebrings to the challenge of recovering species. AnnualReview of Ecology and Systematics 33:665–706.

Schaefer, J. F., E. Marsh-Matthews, D. E. Spooner, K. B. Gido,and W. J. Matthews. 2003. Effects of barriers and thermalrefugia on local movement of the threatened leoparddarter, Percina pantherina. Environmental Biology ofFishes 66:391–400.

Shelton, P. A., and G. R. Lilly. 2000. Interpreting the collapseof the northern cod stock from survey and catch data.Canadian Journal of Fisheries and Aquatic Sciences57:2230–2239.

Sigler, W. F., and R. R. Miller. 1963. Fishes of Utah. Depart-ment of Fish and Game, Salt Lake City, Utah.

Sigler, W. F., and J. W. Sigler. 1987. Fishes of the Great Basin:a natural history. University of Nevada Press, Reno.

Stephens, P. A. 1999. What is the Allee effect? Oikos87:185–190.

75



Stockwell, C. A., A. P. Hendry, and M. T. Kinnison. 2003.Contemporary evolution meets conservation biology.Trends in Ecology and Evolution 18(2):94–101.

Townsend, C. R. 1996. Invasion biology and ecologicalimpacts of brown trout Salmo trutta in New Zealand.Biological Conservation 78:13–22.

Vrijenhoek, R. C. 1998. Conservation genetics of freshwaterfish. Journal of Fish Biology 53:394–412.

Walser, C. A., M. C. Belk, and D. K. Shiozawa. 1999. Habitatuse of leatherside chub in the presence of predatorybrown trout. Great Basin Naturalist 59:272–277.

Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic nicheand species interactions in size-structured populations.Annual Review of Ecology and Systematics 15:393–425.

Williams, C. D., and J. E. Deacon. 1991. Ethics, federal leg-islation, and litigation in the battle against extinction.Pages 109–121 in W. L. Minckley and J. E. Deacon, edi-tors. Battle against extinction: native fish managementin the American West. University of Arizona Press,Tucson.

Wilson, K. W., and M. C. Belk. 2001. Habitat characteristicsof leatherside chub (Gila copei) at two spatial scales.

76

BELK AND JOHNSON


77

Four endemic species of fish are found in BearLake (McConnell et al. 1957; Sigler andMiller 1963; Sigler and Sigler 1987). They

include three species of Prosopium: Bonnevillecisco P. gemmifer, Bonneville whitefish P. spilono-tus, and Bear Lake whitefish P. abyssicola. Thefourth endemic species is the Bear Lake sculpinCottus extensus.

The Bonneville cisco can easily be identifiedfrom the other two Prosopium spp. based on simplemorphological characteristics (Sigler and Work-man 1978; Tolentino and Thompson 2004). Thecisco have a terminal mouth and fusiform body,whereas the whitefish have subterminal mouthsand more a more robust body shape. The cisco is apelagic, schooling fish compared to the whitefish,both of which are benthic, nonschooling fish. Theirpopulation has been monitored using hydroa-coustic technology since 1990 (Luecke et al. 1990;Nielson and Tolentino 1996).

The Bonneville and Bear Lake whitefish are

both benthic species. Although they occur sym-patrically, they spawn allochronously (at differenttimes). This assortative spawning is thought tomaintain the distinct species. Distinguishingamong individuals that are not spawning has beenlimited to those that are over 250 mm total length(TL) using morphologic or meristic characteristics(Snyder 1919; Sigler and Miller 1963; White 1974;Sigler and Sigler 1987; Broughton 2000). Whitefishgreater than 250 mm TL were considered to beBonneville whitefish since Tolentino and Thomp-son (2004) observed few (N = 9 of 333) matureBear Lake whitefish larger than that size. From1973 to 1998, the two species of whitefish werecombined and referred to as the Bear Lake “white-fish complex.” Only recently have more detailed lifehistory, diet, and relative population size investiga-tions been completed (Ward 2001; Thompson2003; Tolentino and Thompson 2004).

Historical literature describes the endemicwhitefishes of Bear Lake in general terms (Snyder

Population Status and Trends of Four Bear Lake Endemic Fishes

ABSTRACT Populations of the four Bear Lake endemic species, Bonneville cisco Prosopium gemmifer, Bon-

neville whitefish P. spilonotus, Bear Lake whitefish P. abyssicola, and Bear Lake sculpin Cottus extensus, were

monitored by Utah Division of Wildlife Resources and/or Utah State University. Bonneville cisco populations have

been monitored annually since 1990 using hydroacoustic sampling. The population has ranged from 1.9 to 9.7

million fish. Because no reliable method was available to determine whitefish species identification prior to 1999,

the whitefish population was monitored from 1973 to 1998 by lumping the catches of both species and referring

to them as the “whitefish complex.” In 1999, the Utah Division of Wildlife Resources began recording the percent-

age of gill-net catch comprised by Bonneville and Bear Lake whitefish. Bonneville whitefish comprised between

84% and 94% of the gill-net catch while Bear Lake whitefish comprised between 6% and 16%. Relative abun-

dance of Bear Lake sculpin was monitored biennially since 1998 using catch rates from bottom trawling. Aver-

age catch per 20-min trawl at both the top and bottom of the thermocline at three different sites ranged from 37

fish to 79 fish.

Scott A. Tolentino



SCOTT A. TOLENTINO Utah Division of Wildlife Resources, Bear Lake Station, Post Office Box 231, Garden City, Utah 84028, USA,[email protected]


1919; Locke 1929; Perry 1943; McConnell et al.1957). Tolentino and Nielson (1999) summarizedthe Bear Lake “whitefish complex” catch and dietinformation, but they were unable to report the dataon individual whitefish species. Additionally, somelimited research has been conducted on the geneticsof the Bear Lake whitefishes by White (1974),Ohlhorst (1985), and Toline et al.(1999). White’s(1974) study was limited to a survey of general pro-teins: lactate dehydrogenase, malate dehydrogenase,and glutamate dehydrogenase associated with tissuefrom blood serum, liver, white muscle, whole eye,and brain from all three Prosopium species fromBear Lake. He found that the isozyme patterns wereremarkably uniform, with no species-specific differ-ences. Ohlhorst (1985) used chromosomes to evalu-ate centromere size and location and used restric-tion enzyme analysis of mitochondrial DNA(mtDNA) to examine the size of nucleotides. Again,no species-specific differences were noted amongthe species. Toline et al. (1999) used bothmtDNA (composite haplotype frequen-cies) and nuclear DNA (nucleotidesequence divergence) to determine thegenetic relatedness of all three Prosopiumspecies in Bear Lake. The results demon-strated a distinct genetic difference betweenBonneville cisco and the two whitefishspecies. In addition, a single, fixed molecu-lar marker differentiated between the BearLake and Bonneville whitefish, which sug-gested that introgression has not occurredbetween these taxa. None of the results ofthe three studies supported a fourth speciesof Prosopium existing in Bear Lake since thethree main taxa have had very little time todiverge evolutionarily.

The Bear Lake sculpin are closely asso-ciated with the benthic community ofBear Lake (McConnell et al. 1957; Siglerand Miller 1963; Sigler and Sigler 1987;Neverman 1989). Smart (1958) describedthe life history of the Bear Lake sculpin,and Utah State University (USU) investi-gated the distribution of the sculpin pop-ulation in the 1960s (Dalton et al. 1965).In the 1980s, Neverman (1989) describedthe feeding ecology of the Bear Lake

sculpin. Following that research, efforts wereundertaken to develop a monitoring method forthis endemic species (Luecke et al. 1990). Benthicareas of the lake were sampled with bottom trawlsin the early 1990s, and minimum population esti-mates were reported (Wurtsbaugh et al. 1993,1994, 1995, 1996, and 1997). The sampling target-ed three depth strata that corresponded to the epil-imnion, metalimnion, and profundal zones.

In this paper, I describe the monitoring methodsand the results of those programs for each species.

Site DescriptionTrend gill netting

Bear Lake is an ultra-oligotrophic lake that over-laps the state border between the states of Utahand Idaho at 1,805 mean sea level (msl; Figure 1).It is approximately 32 km long and 6–13 km wideand covers a surface area of 282 km2. Its maxi-

78

TOLENTINO

Figure 1. Map of Bear Lake, Utah–Idaho.

St. Charles Cr.

Fish Haven Cr.

Bear Lake

Swan Cr.

IDAHOUTAH

North Eden Cr.

Big Spring Cr.


mum and mean depths are 63 and 28 m, respec-tively. There are no major bays or coves. The lakewas formed by tectonic faulting and is considereda tilted fault-block graben (Birdsey 1989). At thenorth shore, a pumping facility connects the lakewith the Bear River, allowing the top 6.5 m to bemanipulated and used as a storage reservoir fordownstream irrigation and hydroelectric powergeneration. The lake is dimictic and the epil-imnion rarely reaches temperatures exceeding20ºC. The lake typically freezes in 4 of 5 years, andice formation can begin in December and lastuntil early May (unpublished Utah Power andLight data).

In addition to the four endemic species of fishmentioned above, indigenous fish found in BearLake include Bonneville cutthroat troutOncorhynchus clarkii, redside shiner Richardso-nius balteatus, Utah sucker Catostomus ardens,Utah chub Gila atraria, and speckled daceRhinichthys osculus. Nonnative fish currentlyfound in Bear lake include rainbow troutOncorhynchus mykiss, lake trout Salvelinusnamaycush, yellow perch Perca flavescens, andcommon carp Cyprinus carpio.

No commercial fishery for any species existson Bear Lake, and it has been and currently ismanaged as a trophy cutthroat trout and laketrout fishery. Cutthroat natural reproduction andrecruitment is limited due to dewatering ofspawning tributaries and naturally reproducedcutthroat contribute to less than 10% of theannual catch (Utah Division of WildlifeResources, unpublished data). In order to main-tain a viable sport fishery, cutthroat trout contin-ue to be stocked into the lake on an annual basis.Lake trout have been stocked in Bear Lake formore than 90 years and have recently beenstocked on a triennial basis since 1995. There hasbeen no measurable natural recruitment of laketrout in Bear Lake, likely due to limited spawninghabitat and the high number of native fish thatfeed on eggs in the lake. In order to ensure no nat-ural recruitment of lake trout in the future, alllake trout stocked since 2001 were sterilized bypressure treatment of the fertilized eggs. Resultsfrom the pressure treatments have yielded 100%sterility results.

Methods

Hydroacoustic sampling

Luecke et al. (1990) developed a hydroacousticsampling regime that provided data used to calcu-late lake-wide Bonneville cisco population esti-mates from 1990 to 1996. A dual-beam BioSonicsmodel 105 echosounder equipped with a 420-khztransducer produced 2 pings per second from aboat traveling at 3–5 m/s to collect these data. Upto 10 transects were sampled annually. Fish targetswere processed with a dual beam BioSonics model281 processor and corresponding software. Lueckeet al. (1990) gives further details of the equipmentand processing technique. From 1996 to 2003, aHydroacoustic Technology Inc. Model 243 split-beam echosounder was used to collect Bonnevillecisco population data during the new moon periodin the month of July (Nielson and Tolentino 2002).The echosounder was equipped with a 200-Khztransducer, which produced five pings per secondfrom a boat traveling at 3–5 m/s. Five transectswere sampled annually, although only four tran-sects were sampled in 1999. In 1996, both the dualbeam and split beam system were used along thesame transects, and the resulting cisco populationestimates were highly correlated (Wurtsbaugh et al.1997; Nielson and Tolentino 2002). Midwatertrawling was conducted annually from 1990 to1995 (Luecke et al. 1990) and every third year from1996 to 1999 (Nielson and Tolentino 2002) to ver-ify the identity of midwater acoustic targets. Bothacoustic systems were calibrated annually, prior tosampling, to a standard target sphere suspendedapproximately 5 m below the transducer.

Gill netting

From 1999 to 2003, the Utah Division of WildlifeResources (UDWR) collected both Bear Lake andBonneville whitefish using experimental sinkinggill nets set three times per year: prior to stratifica-tion, during stratification, and after overturn attwo sites for 48 h duration. The nets were pickedand reset at 24-h intervals. Nielson and Tolentino(2002) give a detailed description of sites andmethodologies used. All whitefish were measured(TL, mm) and up to 10 randomly selected white-fish smaller than 250 mm TL from each depth were

79

POPULATION STATUS AND TRENDS OF BEAR LAKE ENDEMIC FISHES


placed on ice and transported back to the laborato-ry in coolers. Scale counts were either made onfresh fish brought back to the laboratory or the fishwere frozen whole so that scales counts could bemade at a later date. Whitefish larger than 250 mmTL were considered to be Bonneville whitefishsince previous studies indicated that the maximumsize of Bear Lake whitefish was approximately 264mm TL (Ward 2001; Thompson 2003; Tolentinoand Thompson 2004).

The technique developed by Ward (2001) andlater modified by Tolentino and Thompson (2004)was used to separate the two species. This procedureinvolved counting scales both along the lateral lineand above the lateral line to determine the species.The percent composition by species was determinedfor the gill-net catches from 1999 to 2003.

Bottom trawling

On a biennial basis from 1998 to 2002, the UDWRcollected Bear Lake sculpin using a 4.9-m headropesemibottom otter trawl. Wurtsbaugh and Lay (1998)give a detailed description of the trawl dimensionsthat were used in this study, and Nielson andTolentino (2002) describe the sample locations.Trawls in this study were conducted attwo depths that intersected the topand the bottom of the thermocline. Inorder to determine sampling depths,temperature profiles of the water col-umn were taken with a temperaturemeter on the day prior to trawling.Previous studies by Wurtsbaugh et al.(1997) concluded that the highestconcentration of Bear Lake sculpinsoccurred in the benthic zone wherethe thermocline intersected the bot-tom, and for logistical reasons, we tar-geted this area where sculpin catcheswere greater and more consistent.Other areas, both in the littoral andprofundal zones, were not sampledsince previous trawl catches in thesezones were extremely variable and didnot yield results that could be com-pared between years. All trawls weremade for 20 min and were pulled at aspeed of approximately 1 m/s.

Results

Bonneville cisco

Lake-wide population estimates of Bonneville ciscosranged from 1.92 million to 9.7 million fish (Figure2). Analysis of variance (ANOVA) for the cisco pop-ulation estimates collected by USU from the dualbeam hydroacoustic system from 1990 to 1995revealed significant differences between years (F5, 174

= 4.80, p < 0.05) with population numbers increasingslightly from 1993 to 1995. Additionally, populationestimates reported by UDWR were also significantlydifferent between years from 1996 to 2003 (F7,33 =3.71, p < 0.05). There was an obvious increase in theBonneville cisco from 1996 to 2001 and then adecline in 2002 followed by another rebound in 2003.Midwater trawling results indicated that more than98% of the fish caught within the size range of50–220 mm TL were Bonneville cisco, and no adjust-ments were made to the acoustic data to compensatefor other species in the water column.

Bonneville and Bear Lake whitefish

Gill nets were set for a total of 1,728 h in each yearfrom 1999 to 2003. The annual gill-net catch of

80

TOLENTINO

Figure 2. Bonneville cisco population estimates for fish 50–220 mm totallength for Bear Lake 1989–2003. Error bars represent 95% confidenceinterval about the means.*Estimated by USU using a dual beam system.

**Estimated by UDWR using a split beam system.

1989–2003 July Hydroacoustic Estimates(Bonneville cisco — 50mm – 220mm TL)

14

12

10

8

6

4

2

0

89 90 91 92 93 94 95 96* 96** 97 98 99 00 01 02 03

Year

Mill

ions

of c

isco

1989–1996 U.S.U. dual beam

1996–2003 U.D.W.R. split beam


both species of whitefish combined ranged from420 to 1,346 fish. Random subsamples of between119 and 250 whitefish from the total gill-net catchwere collected annually to determine the percentcomposition of the whitefish by species. The sub-sampled whitefish from each year were identifiedto species using scale counts. The percent compo-sition of Bonneville whitefish by year ranged from84% to 94%, and Bear Lake whitefish ranged from6% to 16% (Figure 3).

Bear Lake sculpin

Bottom trawls were completed at depths whereboth the top and the bottom of the thermoclineintersected the bottom at three different sites in1998, 2000, and 2002. Trawl depths ranged from 8to 11 m, which sampled the top of the thermocline,and 12–18 m, which sampled the bottom of thethermocline. Sculpin catches from all sites werecombined and averaged for the top of the thermo-cline and bottom of the thermocline. The averagecatch per 20-min trawl ranged from 37 to 62 (SE13–16) sculpins for the top of the thermocline andfrom 43 to 79 (SE 7–12) sculpins for the bottom ofthe thermocline (Figure 4).

Discussion

The Bonneville cisco population fluctuated betweenapproximately 2 and almost 10 million fish andappears to be stable (1990–1999) or increasing(2000–present). It is not yet known whether this fluc-tuation is natural, in response to drought, or if alter-native causes were responsible.Additional, long-termevaluation of the population is needed. Althoughcisco comprise a large portion of the diet of both cut-throat trout and lake trout in Bear Lake (Nielson andTolentino 2002), both populations of these picivo-rous predators are controlled by stocking (little nat-ural recruitment of either species occurs in BearLake), and therefore, predation on cisco can be con-trolled. Cisco are also caught by anglers, but only forapproximately 7–10 d per year during their spawningrun in mid- to late January. Past creel estimates ofcisco harvest range from approximately 10,000 to100,000 fish annually (Utah Division of WildlifeResources, unpublished data), and there is little riskfrom overexploitation by sportfishing.

81


Figure 3. Percent composition of gill-net catch and average total lengths of Bonneville and Bear Lake whitefish from Bear Lake 1999–2003.

100

80

60

40

20

0

300

250

200

150

100

50

01999 2000 2001 2002 2003

Year

Perc

ent C

atch

by

Spec

ies

Figure 4. Catches rates (number/20-min bottom trawl) ofage-1 and older Bear Lake sculpin from Bear Lake 1998to 2002.

120

100

80

60

40

20

01998 2000 2002

CPUE

(N/ 2

0 m

in.t

raw

l)

Aver

age

Tota

l Len

gth

(mm

)

Bonn %BLWF %

Bonn Avg. TLBLWF Avg. TL

Year


Tolentino and Thompson (2004) were able toachieve correct species identification of all white-fish in their study on Bear Lake when they used acombination of lateral line and above lateral linescale counts. Applying this same technique, allwhitefish collected in the current study on BearLake from 1999 to 2004 were also positively identi-fy to species. This initial effort to document thenumber of each species of whitefish caught in thestandardized gill-net sampling was successful andwill allow continued trend-monitoring of the catchby species into the future.

Recent work by Thompson (2003) and Tolenti-no and Thompson (2004) documented the highestcatches of Bear Lake whitefish at depths greaterthan 40 m. In fact, Thompson (2003) found thatfew Bonneville whitefish are caught a depthsgreater than 40 m and that the species have verydifferent diets. In this study, we only sampled todepths of 35 m. Beginning in 2004, however, anadditional gill net set at the 50 m depth contourwill be added to the UDWR standardized gill-netsampling plan at Bear Lake in order to better sam-ple Bear Lake whitefish.

Although Bear Lake whitefish comprised onlybetween 6% and 16% of the total catch of whitefishfrom gill nets over the 5-year period, 1999–2003, thisspecies has been shown to prefer depths greater than40 m and the highest catches in recent studies wereobserved from depths of 50–60 m (Thompson2003). This habitat selectivity limits them to a smallportion of Bear Lake, which provides those depths.The catch per unit effort of Bear Lake whitefish atdepths of 40 m and deeper was nearly equal to thatobserved for Bonneville whitefish in the 15–25 mdepths where they were most concentrated (Thomp-son 2003). Fewer Bear Lake whitefish were capturedin the overall catch of whitefish when compared toBonneville whitefish. The data indicate a stable Bon-neville whitefish population for the period sampled.

Although we did not report our diet analysesfor the whitefishes in Bear Lake in this paper, weagree with the results of previous studies(McConnell et al. 1957; Tolentino and Thompson2004) that the Bonneville whitefish were far-rang-ing opportunistic feeders. In contrast, the diet ofthe Bear Lake whitefish consists mainly of ostra-cods, which is consistent with the fishes occur-

rence over substrates in deep water where ostra-cods are found.

Other fish, including Bonneville ciscos, BearLake sculpins, Utah suckers, Utah chubs, and com-mon carp prey on the eggs and larval forms of allthree endemic Prosopium spp. at certain times ofthe year. However, no attempt has been made toquantify predation by those species, but it is notlikely to negatively impact the Prosopium popula-tion since the nonnative species are typically lit-toral zone oriented for their entire lives and theindigenous species have coevolved for thousands ofyears (Tolentino and Nielson 1999; Mazur andBeauchamp 2000).

Using the annual gill-net surveys to determinepopulation relative abundance and monitoringangler catch/harvest through regular angler creelsurveys coupled with having regulations in placethat restrict sport harvest and forbid commercialfishing for either species, the possibility of overex-ploiting either species is highly unlikely. Further-more, Bonneville whitefish only support a limited,specialized sport fishery during the winter, andBear Lake whitefish are rarely harvested due totheir diminutive size.

The Bear Lake sculpin has been identified byother researchers as one of the most abundant fishin Bear Lake (McConnell et al. 1957; Dalton et al.1965; Neverman and Wurtsbaugh 1994). Bottomtrawl catches of Bear Lake sculpins in this studyconfirmed that they were the most numerousspecies captured. Although the trawl catches ofBear Lake sculpins were variable during the threeyears sampled (Figure 4), the catch per unit effort(CPUE) data from the current study (1998–2002)compared very closely with CPUE data collected byWurtsbaugh et al. (1993, 1994, 1995, 1996, and1997) from same depth zone. We were unable tomake any statistical comparisons to the data col-lected by Wurtsbaugh et al. since different siteswere used in their study.

The majority of the known spawning substrateof the Bear Lake sculpin has been dewatered dueto drought conditions during the period2000–2003 in which the water level in Bear Lakehas dropped to near historic lows (~1,800 m msl).This is very similar to what was observed in theearly 1990s when lake levels were also near historic

82

TOLENTINO


lows. However, during the same time as the previ-ous drought (1990–1995), Wurtsbaugh’s et al.(1997) data revealed an increasing population ofBear Lake sculpin prior to the lake recovering tonear full-pool in 1998. The increase also occurredat a time when they perceived the predator fishpopulation to be at relatively high levels, so it islikely that the Bear Lake sculpin populations werenot being negatively affected by dewatering ofknown spawning habitat or predators in the sys-tem. It is therefore possible that other sculpinspawning areas exist in Bear Lake, but their loca-tion(s) are unknown at this time.

The monitoring program for endemic speciesin Bear Lake functioned initially to define the sizeof the endemic populations and later to detectchanges in population size or to estimate changesin relative abundances over time. This monitoringprogram has proven very useful. Although onepostulation is that the endemic fish populationsmay be affected by decreased water levels, moni-toring has not yielded any results that would indi-cate this. Fluctuation of the lake level could actu-ally be beneficial to one or more of the endemicspecies since the rise and fall of the water levelscoupled with wind action scours rocky shorelinehabitat and prevents sediment accumulation andalgal growth on the rocky littoral substrate. It isthis same substrate that is used by at least three ofthe four endemic species use for spawning.

Currently other research is being conducted byUSU to further the understanding of life historiesof both of the endemic whitefish species. It is sug-gested that population monitoring for all endemicspecies be continued in order to track changes orfluctuations that occur. When new technologicaladvances and/or further knowledge of the speciesare gained, changes or adjustments to the monitor-ing program should be considered that would serveto better monitor the populations.

Acknowledgments

This project was funded by the UDWR and theFederal Aid in Sport Fish Restoration Act, ProjectF-44-R. I thank Bryce Nielson, retired UDWR BearLake Project Leader, for his assistance with fieldwork and technical suggestions on both the

research and draft manuscript. I also than DonArcher, UDWR special project coordinator, andtwo anonymous reviewers for their editorial com-ments, which improved this paper. Finally, I thankmy former seasonal employees, J. Flinders, J. Hand-ly, M. Mills, K. Labrum, and S. Rust, for their assis-tance with field work during the last five years,sometimes during challenging weather conditions.

References

Birdsey, P. W. 1989. The limnology of Bear Lake, Utah-Idaho,1912–1988: a literature review. Utah Division of WildlifeResources, Publication No. 89–5, Salt Lake City.

Broughton, J. M. 2000. Terminal Pleistocene fish remains fromHomestead Cave, Utah, and implications for fish biogeog-raphy in the Bonneville Basin. Copeia 2000(3):645–656.

Dalton, R. A., T. H. Lee, J. L. Helm, and W. T. Helm. 1965.Distribution of sculpin Cottus extensus, in Bear Lake,Utah-Idaho. Utah Academy of Science 42:70–73.

Locke, S. B. 1929. Whitefish, grayling, trout, and salmon ofthe intermountain region. Appendix V to Report ofCommissioner of Fisheries for the Fiscal Year 1929 -Bureau of Fisheries Document No. 1062. United StatesGovernment Printing Office, Washington, D.C.

Luecke, C., W. A. Wurtsbaugh, and D. Teuscher. 1990. Exami-nation of forage fish monitoring techniques: assessment ofpopulation trends for Bonneville cisco and Bear Lakesculpin. 1994 Utah State University progress report toUtah Division of Wildlife Resources, Salt Lake City.

Mazur, M. M., and D. A. Beauchamp. 2000. Populationdynamics and interactions among fishes of Bear Lake.1999 Utah State University progress report to Utah Divi-sion of Wildlife Resources, Salt Lake City.

McConnell, W. J., W. J. Clark, and W. F. Sigler. 1957. BearLake: its fish and fishing. Utah State Department of Fishand Game, Salt Lake City.

Neverman, D. 1989. The diel vertical migration and feedingof underyearling Bear Lake sculpin (Cottus extensus).Master’s thesis. Utah State University, Logan.

Neverman, D., and W. A. Wurtsbaugh. 1994. The thermoreg-ulatory function of diel vertical migration for a juvenilefish, Cottus extensus. Oecologia 98:247–256.

Nielson, B. R., and S. A. Tolentino. 1996. Bear Lake cutthroattrout enhancement project 5-year report 1990–1994.Federal Aid in Sport Fish Restoration project F-26-R.Utah Division of Wildlife Resources, Publication Num-ber 96–23, Salt Lake City.

Nielson, B. R., and S. A. Tolentino. 2002. Fisheries investiga-tions of the Bear Lake cutthroat trout enhancement proj-ect, 5-year report 1995–1999. Federal Aid in Sport FishRestoration project F-26-R. Utah Division of WildlifeResources, Publication Number 02–09, Salt Lake City.

Ohlhorst, S. L. 1985. The origin of a lacustrine species flock:a test of the intralacustrine reproductive allopatry andallochrony hypothesis. Report to National Science

83



Foundation. Proposal Number R11–85. Utah State Uni-versity, Logan.

Perry, E. L. 1943. Biology and economic significance of thepeaknose cisco of Bear Lake, Idaho and Utah. Doctoraldissertation. Utah State University, Logan.

Sigler, W. F., and R. R. Miller. 1963. Fishes of Utah. Utah StateDepartment of Fish and Game, Salt Lake City.

Sigler, W. F., and J. W. Sigler. 1987. Fishes of the Great Basin.University of Nevada Press, Reno.

Sigler, W. F., and G. W. Workman. 1978. The Bonneville ciscoof Bear Lake, Utah-Idaho. Utah State University, UtahAgricultural Experiment Station, Research Report 33,Logan.

Smart, E. W. 1958. An ecological study of the bottom faunaof Bear Lake Idaho and Utah. Doctoral dissertation. UtahState University, Logan.

Snyder, J. O. 1919. Three new whitefishes from Bear Lake,Idaho and Utah. United States Bureau of Fisheries36:3–9, United States Government Printing Office,Washington, D.C.

Thompson, B. W. 2003. An ecological comparison of twowhitefish species in Bear Lake Utah/Idaho. Master’s the-sis. Utah State University, Logan.

Tolentino, S. A., and B. R. Nielson. 1999. The Bear Lakewhitefish complex: summary report of creel and gill netdata. Federal Aid in Sport Fish Restoration project F-26-R. Utah Division of Wildlife Resources, PublicationNumber 99–4, Salt Lake City.

Tolentino, S. A., and B. W. Thompson. 2004. Meristic differ-ences, habitat selectivity, and diet separation of Prosopi-um spilonotus and P. abyssicola. Annales Zoologici Fenni-ci 41(1):309–317.

Toline, C. A., T. R. Seamons, and C. Davis. 1999. Quantifica-tion of molecular and morphological differentiation ofwhitefish taxa in Bear Lake. Final report of Utah StateUniversity, Department of Fisheries and Wildlife, Logan.

Ward, A. 2001. Morphometric evaluation of the whitefishcomplex in Bear Lake, Utah / Idaho. Master’s thesis. UtahState University, Logan.

White, R. G. 1974. Endemic whitefishes of Bear Lake, Utah-Idaho: a problem in systematics. Doctoral dissertation.Utah State University, Logan.

Wurtsbaugh, W. A., and C. Lay. 1998. Influence of foodresources and temperature on the distribution and ener-gy balance of Bear Lake sculpin, Cottus extensus. 1994Utah State University progress report to Utah Divisionof Wildlife Resources, Salt Lake City.

Wurtsbaugh, W. A., C. Luecke, C. Lay, and R. Orme. 1995.Examination of the abundance and spatial distributionof forage fishes in Bear Lake. 1994 Utah State Universityprogress report to Utah Division of Wildlife Resources,Salt Lake City.

Wurtsbaugh, W. A., C. Luecke, C. Lay, and J. Ruzycki. 1994.Examination of the abundance and spatial distributionof forage fishes in Bear Lake. 1993 Utah State Universityprogress report to Utah Division of Wildlife Resources,Salt Lake City.

Wurtsbaugh, W. A., C. Luecke, C. Lay, and A. Ward. 1996.Examination of the abundance and spatial distributionof forage fishes in Bear Lake. 1995 Utah State Universityprogress report to Utah Division of Wildlife Resources,Salt Lake City.

Wurtsbaugh, W. A., C. Luecke, N. Bouwes, and J. Ruzycki.1993. Examination of the abundance and spatial distri-bution of forage fishes in Bear Lake. 1990–92 Utah StateUniversity progress report to Utah Division of WildlifeResources, Salt Lake City.

Wurtsbaugh, W. A., C. Luecke, and A. Ward. 1997. Exami-nation of the abundance and spatial distribution of for-age fishes in Bear Lake. 1996 Utah State Universityprogress report to Utah Division of Wildlife Resources,Salt Lake City.

84

TOLENTINO


85

Redband trout is the common name for resi-dent (nonanadromous) Oncorhynchus mykissgairdneri populations of the Columbia basin

east of the Cascade Crest. These redband stocks donot have access to the ocean because they are isolat-ed by habitat disrupted by human activity, discon-tinuous drainages, and impassable hydroelectricdams. Populations of redband trout in the SnakeRiver drainages of southern Idaho, northwesternNevada, and southeastern Oregon are historicallyand genetically linked to the Snake River basin steel-head evolutionarily significant unit (ESU). They areseverely threatened by degradation of riparian habi-tat resulting from land use practices (Roberts 1986)as well as decreased streamflows accompanying pro-longed drought.

Redband trout are already extinct throughout72% of their historic range (USDA/USDI 1997).Extant Columbia River basin populations aredepressed in 94% of the watersheds where residentredband are separated from the anadromous steel-head form (Howell 1997). The Endangered SpeciesCommittee of the American Fisheries Society andthe Idaho Department of Fish and Game (IDFG)have considered the inland Columbia basin red-band trout a “species of special concern” since

1989. Oregon redband trout subspecies were desig-nated sensitive in 1991 (Moskowitz and Rahr1994). The redband trout was nominated for list-ing and designated by the U.S. Fish and WildlifeService (USFWS) as a candidate species (C-2;Dodd et al. 1985), but the petition was denied(USFWS 1995).

In 1997, Snake River basin steelhead, theanadromous form of redband trout, were listed asthreatened under the Endangered Species Act(ESA), making a critical distinction between theSnake River basin stock and other ESUs (Schmitten1997). The unique environment resulting from thedrier terrain of the Snake River basin belowShoshone Falls creates warmer stream tempera-tures than found elsewhere in the range of inlandsteelhead and redband trout. Although the desertredband trout can adapt to reduced water flows,their ability to persevere drought and habitatdestruction is currently being exceeded in manyheavily grazed desert drainages (Behnke 1986).Habitat degradation by livestock overgrazing isconsidered the greatest threat to redband troutpopulations (Coffin 1982). More than 70% of theredband habitat in the Boise District of southwest-ern Idaho has been inventoried by U.S. Bureau of

The Status of Desert Redband Trout in Southwestern Idaho

ABSTRACT Ecologically distinct stocks of resident redband trout above the Hells Canyon dam complex in

southwestern Idaho represent physiologically unique components of the evolutionary legacy of the Snake River

steelhead evolutionarily significant unit. These gene pools from the most extreme environments inhabited by this

species, absent introgression produced by pooled hatchery stocks, have the potential to provide the genetic diver-

sity necessary for species survival. Populations of desert redband trout in the Snake River drainages of southern

Idaho are severely threatened by degradation of riparian habitat resulting from land use practices, as well as

decreased streamflows accompanying prolonged drought.

Donald W. Johnson and Katie Fite



DONALD W. JOHNSON AND KATIE FITE Western Watersheds Project, Post Box 2863, Boise, Idaho 83701, USA


Land Management (BLM) as in less than goodcondition (Vinson 1991).

Although the taxonomic status of the desert red-band trout has been unclear for more than 100 years(see Behnke (1992) references to Gilbert and Ever-mann [1892] and Jordan and Evermann [1896]), itsecological adaptations together with its strikingphysical appearance have always set these stocksapart from closely related coastal rainbow trout andother O. mykiss subspecies.

Redband trout are reported to be widely distrib-uted in timbered mountain drainages in the SnakeRiver basin. Some of these stocks have been contam-inated through introgressive hybridization withhatchery-stocked coastal rainbow trout (Williams etal. 1991; Williams and Shiozawa 1993). Desert pop-ulations such as the one in King Hill Creek providethe most promise of pure redband trout stocksbecause of their inaccessibility. Most small isolateddesert trout populations have been ignored andtheir diversity and genetic relationships remainlargely unknown. Conservation of different racescannot be accomplished without a complete under-standing of the diversity of both isolated and non-isolated populations within and among basins(Currens et al. 1990). The status of desert redbandtrout populations is unknown in many drainagessince they have not been adequately studied.

Ecologically distinct stocks of the resident red-band trout above the Hells Canyon dam complexrepresent physiologically unique components ofthe evolutionary legacy of the Snake River basinsteelhead ESU. These gene pools from the mostextreme environments inhabited by this species,absent introgression produced by pooled hatcherystocks, have the potential to provide the geneticdiversity necessary for species survival.

The National Marine Fisheries Service (NMFS)has suggested that available evidence supports list-ing resident redband trout with the listed steelheadESUs where resident fish of native lineage once hadthe ability to interbreed with anadromous fish, butno longer do so because they are currently aboveman-made barriers (Schmitten 1997). Such“reserve” gene pools in the resident freshwaterform of the species may persist through times ofunfavorable conditions for the anadromous form.In 1991, NMFS concluded that the potential loss of

anadromy in distinct population segments may bereason for listing the species as a whole (Schmitten1997). Landlocked adfluvial migrating populationsof redband trout in the Great Basin substantiatethat the migratory-drive is maintained in the genepool many generations after habitat alterationshave restricted migratory routes (Belsky 1997).

Redband trout have been recognized as distinctbased on their tolerance of severe desert environ-ments and from their external appearance withcharacteristics of both rainbow and cutthroat trout(Behnke 1986). Wallace (1981) investigated thephysical variation between nine stocks of southernIdaho trout: one hatchery population and eightnative populations from Owyhee County streams.He analyzed numbers of vertebrae, gill rakers, basi-branchial teeth, pyloric caecae, lateral line scales,and pelvic fin rays. In addition, most native red-band trout had white tips on the fins, elliptical parrmarks, and spots completely or mostly above thelateral line. Berg (1987) suggested recognition ofsubspecies from each geographic area after report-ing that different redband populations were simi-lar to but distinct from coastal rainbow trout andeach other. That distinctiveness is thought to reflectgenetic drift resulting from the semi-isolation ofthese small, uniquely adapted populations.

Biochemical genetic studies (Wishard et al.1980, 1984) supported similarities and differencesbased on physical characteristics. Wallace (1981)concluded that although there may not be anyabsolutely pure populations of native trout left insouthwestern Idaho, they should be treated asessentially pure and managed accordingly; thegenetic studies suggested little, if any, introgressionfrom hatchery rainbow trout. The genetic analysesof Wishard et al. (1980) found geographic relation-ships (Owyhee River versus unassociated SnakeRiver tributary groupings) inconsistent with anysignificant hatchery introgression with a number ofdistinct native trout gene pools defined for OwyheeCounty. Their study suggested that in OwyheeCounty, habitat destruction presented a greaterthreat to extinction than gene pool pollution.

Behnke (1992), in his examination of physicalcharacteristics, found no evidence of decreased dis-tinctiveness of native redband stocks as a result ofhatchery introductions. Wishard et al. (1980) con-

86

JOHNSON AND FITE


cluded that the regional streams were effective pop-ulation units and that habitat degradation mightresult in the loss of native and uniquely adaptedgene pools. A subsequent study (Leary et al. 1983)suggested that these gene pools probably containnonnative hatchery genetic material, although inthe absence of preintroduction data and accuratestocking records, the evidence was inconclusive.More recent investigations of some southern Idahostreams identified various levels of introgressionwith hatchery rainbow trout and only one purenative redband population in King Hill Creek(Williams et al. 1991; Williams and Shiozawa 1993).

Genetic differences among redband trout pop-ulations indicate that their isolation from rainbowtrout has resulted in biochemical differences thathave conferred survival value in a low oxygen envi-ronment (e.g., desert streams in summer) (Klar etal. 1979). Some strains are considered differentenough to be separate subspecies (Meehan 1994).Desert redband populations are adapted to severedesert conditions and, regardless of their origin orcontemporary taxonomic status, are a geneticresource that should be preserved (Berg 1987).

The loss of streamside vegetation from livestockgrazing reduces shading and may increase watertemperatures above trout tolerance limits. The eco-logical costs of grazing to riparian habitat, troutproduction, and the effectiveness of livestock exclo-sures on ecosystem recovery has been reviewed else-where (Armour et al. 1994; Fleischner 1994; Belskyet al. 1999). Efforts to establish instream water rightshave had little impact since most are “junior” rightsin overappropriated watersheds; the Oregon WaterResources Board has issued 40 such fish-protectingrights since 1987, while irrigators and other usershave obtained 2,300 water rights (Moskowitz andRahr 1994). The USFWS in a September 27, 1995,news release stated that degradation of streamsidehabitat and decreased streamflows from irrigationwithdrawals, complicated by prolonged drought,were thought to threaten the species.

Methods and Results

Available regional redband surveys were reviewed;results of that review are presented here. The 1986Elko Resource Management Plan reported redband

trout in two streams in the area with poor habitatcondition in 66% of the 73 streams surveyed.Riparian and stream conditions on the Nevada sideof the Snake River drainage are deteriorated andworsening. The drought, continuing livestock mis-management and mining impacts on public lands,especially since 1986, have made the continuedexistence of redband trout in Nevada uncertain (R.Strickland, Toiyabe Chapter of the Sierra Club, per-sonal communication). Since 1992, the HumboldtNational Forest has recognized the need to protectredband populations; 20 redband streams werereclassified to the most sensitive management cate-gory and attempts have been made to assure thatminimum streamflows are maintained by filing forwater rights (Strickland, personal communication).

A 1992 IDFG Jarbidge River drainage evaluationof bull trout Salvelinus confluentus status reportedthat, while drought conditions and associated highwater temperatures (26°C) were excessive for bulltrout, self-sustaining populations of redband trouthad survived the prolonged drought (six of sevensites; 1.7–16.2 fish/100 m2). The 1992 JarbidgeRiver survey reported that more than 95% of thetrout sampled on the east and west forks of the Jar-bidge River were native redband trout (Warren andPartridge 1993). Water diversions and habitatdegradation were cited as primary factors in the lossof bull trout populations from those waters.

Snake River redband trout occur in Oregon’sImnaha, Burnt, Powder, Malheur, Grand Ronde,and Owyhee River basins; the latter includes Idahoand Nevada watersheds. Status of many popula-tions is unknown, but drought, water withdrawals,and livestock-induced riparian habitat degradationare threats to their survival. The Oregon Depart-ment of Fish and Wildlife has reported decliningpopulations in the Owyhee River basin. During asummer survey in eastern Oregon, 90% of thestream miles sampled had temperatures exceedingstate maximum standards. Grazing, logging, irriga-tion withdrawals, and dams have degraded habitatand dried streams used by cutthroat trout, bulltrout, and redband trout (Moskowitz and Rahr1994). Streams in Owyhee County, Idaho are gen-erally degraded; their banks are unstable, and theycarry heavy sediment loads as a result of grazingpractices (Grunder 1992).

87

DESERT REDBAND TROUT IN SOUTHWESTERN IDAHO


The most recent southwestern Idaho surveysincluded sampling of five locations in the NorthFork Owyhee River basin where viable numbers ofredband trout (8.8 fish/100 m2) were found onlywithin a BLM livestock exclosure (Grunder 1992).A 1993 study of redband trout in Owyhee Countythat inventoried stocks and associated habitatfound that 10 of 16 stream segments sampled con-tained redband trout with densities from 0.3 to 102fish/100 m2. Ten of the segments had been previ-ously sampled (1976, 1977, or 1991); fish densitieshad increased in three, decreased in six, andremained constant in one. Stocks were described asfragmented and highly variable in year-classstrength (Allen et al. 1993). They concluded thatno undisturbed stream and riparian areasremained to provide a baseline of a “healthy” red-band trout population.

In 1994, data suggested a decline in the desertredband trout populations of Idaho (W. J. Paradis,Idaho Department of Fish and Game, personnalcommunication). Many streams sampled in 1994were dry throughout much of their drainage; only4 of 11 redband trout sites sampled had fish (Allenet al. 1995). Magpie Creek, a tributary of SouthFork Castle Creek, had spawning redband in thespring of 1997. Subsequently, Western WatershedsProject (WWP), BLM, and grazing permitteesnegotiated a legal settlement that included protec-tion for this spawning habitat. Rancher-construct-ed fencing did not follow the flagging of BLM, anda segment of Magpie Creek was reduced to amuddy cattle trail in the summer of 1997.

A recent review of southwestern Idaho redbandpopulations of 32 streams compared surveys with17 different sampling intervals (one to sevenstreams/interval) and concluded that redbandabundance did not differ between the 1970s and1990s (Zoellick et al. 2004). The data, however, sug-gested that population density had decreased at39.5% of the sampling sites with redband and wereextirpated from three sites with prior densitiesmeasured at 12.5, 58, and 92.3 fish/100 m2. Red-band were absent from 19% of the streams duringthe last sampling period and had decreased in den-sity at 37.5% of the sites. Cattle had been removedfrom two creeks (Little Jack and Syrup) and thehabitat had improved, but there had been no

increase in redband density by the last survey.Grazing had been reduced in one drainage, SalmonCreek, and numbers of redband were greater whenlast surveyed. Some sampling sites were dry duringthe last survey periods. The estimated changes orstability of redband populations for the abovecomparisons are uncertain since sampling mayhave occurred during different stages of a wateryear (between mid-May and November) or afterdifferent water years. Initial surveys were in 1977,1980, 1981, or 1982, and the final monitoring couldhave been in 1993, 1994, 1995, 1996, 1997, 2001,2002, or 2003. This is, however, the most reliablecomparative information available.

Thurow et al. (1997) estimated that 69% of thepotential range of these redband populations wasunclassified, and their status was unknown. Theyconcluded that we know less about the current dis-tribution of redband trout than any of the othersalmonids, with the largest area of apparentlyunoccupied habitat in this hydrological unit. Fur-thermore, known or predicted strong populationsaccount for 17% of their potential range. Steelheadwere estimated to occupy 0.6% of their potentialrange and 1.3% of their current range.

Conclusion

Many native trout populations are at risk of extinctionin the next 10–20 years, particularly desert redbandtrout. Those populations are the legacy of millions ofyears of adaptation to local conditions. When aspecies is reduced to isolated remnant populations, itsnear-term extinction is more likely. Redband trouthave suffered fragmentation and isolation of qualityhabitat with declines in abundance and local extinc-tions, while components of the Snake River steelheadESU are near extinction (USDA/USDI 2000). Accord-ing to Allendorf et al. (1980), the rationale of perpet-uating native races centers on the value of perpetuat-ing genetic diversity.

It was the concensus of NMFS scientists andregional fishery biologists that based on availablegenetic information, resident fish should generallybe considered part of the steelhead ESUs (Schmit-ten 1997). Where Snake River redband trout onceshared their gene pool with the listed anadromoussteelhead producing both resident and anadro-

88

JOHNSON AND FITE


mous forms, they now have the potential to pro-vide the native genetic diversity necessary for theSnake River steelhead ESU to survive. Listing theseresident redband trout in desert drainages, wherethey were once sympatric with the listed anadro-mous Snake River basin steelhead, would con-tribute to the maintenance and restoration of theseunique trout and their aquatic ecosystems, as wellas preserve a gene pool that could contribute tofuture steelhead runs (Schmitten 1997).

Extensive cattle grazing has negatively impactedmany streams in southwestern Idaho, althoughsome canyon reaches inaccessible to grazing are ingood condition and producing redband trout. Stateand federal laws are inadequate to promote conser-vation of redband trout in streams bounded by pri-vate lands (S. Yundt, Idaho Department of Fish andGame, personal communication). Faced withpotential listing, IDFG began developing a conser-vation (recovery) plan for redband trout, but theplan was not completed, a decision that was notrelated to any positive change in the status of pop-ulations. Listing will be required to protect thesethreatened ESUs. If protection is enacted andchanges in land use and management are made,improvements in redband habitat can occur. Red-band trout production in five Great Basin studyareas increased by an average of 184% when graz-ing was reduced or eliminated (Bowers et al. 1979).

The status of redband trout stocks has been rec-ognized by public agencies, private individuals,environmental, and professional associations asrequiring the protection provided by an Endan-gered Species Act listing. Public agencies such asthe U.S. Forest Service, BLM, and Idaho Depart-ment of Lands have been unable to effectivelyimplement existing grazing land management pro-cedures or to obtain any meaningful working rela-tionship with the livestock industry to effectreform of present range management practices.The private sector (e.g., WWP, American FisheriesSociety) has attempted without success to breakthe agency impasse in improving public land man-agement to arrest degradation or improve streamhabitat. Aquatic habitat on BLM and U.S. ForestService-administered lands is vital to native fishpopulation survival, and securing and restoringfederal freshwater habitat may be critical to the

short-term persistance of many anadromous pop-ulations (USDA/USDI 2000). Drought and landmanagement practices are threatening these desertredband trout with extinction.

ReferencesAllen, D. B., B. J. Flatter, K. Fite, and S. P. Yundt. 1993. Red-

band trout (Oncorhynchus mykiss) population and habi-tat inventory in Owyhee County, Idaho. Idaho Depart-ment of Fish and Game Report, Bureau of LandManagement, Boise.

Allen, D. B., B. J. Flatter, and K. Fite. 1995. Redband trout(Oncorhynchus mykiss gairdneri) population and habitatsurveys in Jump, Reynolds, and Sheep creeks, and sec-tions of the Owyhee River in Owyhee County, Idaho.Idaho Department of Fish and Game Report, Boise Dis-trict of Bureau of Land Management, Boise.

Allendorf, F. W., D. M. Espeland, and D. T. Scow. 1980. Coex-istence of native and introduced rainbow trout in theKootenai River drainage. Proceedings of the MontanaAcademy of Science 39:28–36.

Armour, D., D. Duff, and W. Elmore. 1994. The effects oflivestock grazing on western riparian and stream ecosys-tems. Fisheries 19(9):9–12.

Behnke, R. J. 1986. Redband trout. Trout 27(4):34–39.Behnke, R. J. 1992. Monograph of the native trouts of the

genus, Salmo of western North America. Funded by U.S.Forest Service, Rocky Mountain Region, Denver.

Belsky, A. J. 1997. A petition for rules to list Great Basin red-band trout (Oncorhynchus mykiss spp.) as threatened orendangered under the Endangered Species Act. OregonNatural Desert Association, Bend.

Belsky, A. J., A. Matzke, and S. Uselman. 1999. Survey of live-stock influences on stream and riparian ecosystems inthe western United States. Journal of Soil and WaterConservation 54(1):429–431.

Berg, W. J. 1987. Genetic characteristics of an enigma: what isthe redband trout? Pages 89–125 in Evolutionary geneticsof rainbow trout, Parasalmo gairdnerii (Richardson). Doc-toral dissertation. University of California, Davis.

Bowers, W., B. Hosford, A. Oakley, and C. Bond. 1979. Wildlifehabitats in managed rangelands—the Great Basin ofsoutheastern Oregon, native trout. U.S. Forest Service,Pacific Northwest Forest and Range Experiment Station,General Technical Report PNW 084, Portland, Oregon.

Coffin, P. D. 1982. A fisheries management plan for the red-band trout in the upper Columbia River basin of ElkoCounty, Nevada. Nevada Department of Wildlife, Feder-al Aid Project F-20–18, Study IX Job No. 1-P-l, Reno.

Currens, K. P., C. B. Schreck, and H. W. Li. 1990. Allozymeand morphological divergence of rainbow trout(Onchorhynchus mykiss) above and below waterfalls inthe Deschutes River, Oregon. Copeia 1990(3):73–74.

Dodd, C. K., Jr., G. E. Drewry, R. M. Nowak, J. M. Sheppard,and J. D. Williams. 1985. Endangered and threatenedwildlife and plants; review of vertebrate wildlife; noticeof review. Federal Register 50(181):37958–37967.

89

DESERT REDBAND TROUT IN SOUTHWESTERN IDAHO


Fleischner, R. L. 1994. Ecological costs of livestock grazing inwestern North America. Conservation Biology (3):629–644.

Grunder, S. 1992. Potential impacts of flares and extensiveground disturbance to fishery resources and habitat.Evaluation of impacts to fish and wildlife resources asso-ciated with development of a bombing range in OwyheeCounty, Idaho. Idaho Department of Fish and Game,Briefing notes, Boise.

Howell, P. 1997. Oregon Chapter workshop raises visibilityof declining redbands. Fisheries 22(3):34–36.

Klar, G. T., C. B. Stalknaker, and T. M. Farley. 1979. Compar-ative physical and physiological performance of rainbowtrout, Salmo gairdneri, of distinct lactate dehydrogenaseB2 phenotypes Comparative Biochemistry and Physiol-ogy 63A:229–235.

Leary, R. F., F. W. Allendorf, and K. L. Knudsen. 1983. Genet-ic analysis of four rainbow trout populations fromOwyhee County. University of Montana, Final Report#8316, Missoula.

Meehan, B. T. 1994. Realm of the redband. The Oregonian(December 26):l, 6.

Moskowitz, D., and G. Rahr. 1994. Oregon Trout–nativetrout report—an analysis of the status and managementof Oregon’s native trout with management recommen-dations for conservation. Prepared by The Native TroutConservation Council for Oregon Trout, Portland.

Roberts, D. 1986. Reclusive redbands. Trout 27(4):12-25.Schmitten, R. A. 1997. Steelhead biology and ecology. Feder-

al Register 62:159(18 August 1997):43941–43953.Thurow, R. F., D. C. Lee, and B. E. Rieman. 1997. Distribu-

tion and status of seven native salmonids in the interiorColumbia River basin and portions of the Klamath Riverand Great basins. North American Journal of FisheriesManagement 17:1094–1110.

USDA (United States Department of Agriculture)/USDI(Department of Interior). 1997. Interior Columbia BasinEcosystem Management Project, Eastside draft environ-mental impact statement, volume 1. U.S. Forest Serviceand U.S. Bureau of Land Management, BLM-OR-WA-PL-96–037+1792, Walla Walla, Washington.

USDA (United States Department of Agriculture)/USDI

(Department of Interior). 2000. Interior Columbia Basinsupplemental draft EIS summary. U.S. Forest Service andU.S. Bureau of Land Management, BLM/OR/WA/Pt-00/018 + 1792, Walla Walla, Washington.

USFWS (U. S. Fish and Wildlife Service). 1995. Endangeredand threatened wildlife and plants; a 90-day finding for apetition to list desert redband trout in the Snake Riverdrainage above Brownlee Dam and below Shoshone Fallsas threatened or endangered. Federal Register 60:87(20September 1995):49819–49821.

Vinson, M. 1991. Distribution, ecology, and population andhabitat status of redband trout in southwestern Idaho.Bureau of Land Management, TFBFO-91-0611), Boise,Idaho.

Wallace, R. L. 1981. Morphological study of native troutpopulations of Owyhee County, Idaho. Final Report.Bureau of Land Management, Contract ID-010-CTO-0002, Boise, Idaho.

Warren, C. D., and F. E. Partridge. 1993. Evaluation of thestatus of bull trout in the Jarbidge River drainage, Idaho.Idaho Bureau of Land Management, Technical BulletinNo. 93–1, Boise.

Williams, R. N., J. E. Carter, D. K. Shiozawa, and R. F. Leary.1991. Taxonomic and genetic analysis of five rainbowtrout populations from southern Idaho. Boise State Uni-versity, Final Report 91–2, Boise, Idaho.

Williams, R. N., and D. K. Shiozawa. 1993. Genetic analysisof rainbow trout from the Big Wood River and TrailCreek, Blaine County, Idaho. Idaho Department of Fishand Game, Project F-73-R-14 Subproject V, Boise.

Wishard, L. N., J. E. Seeb, and F. M. Utter. 1980. Biochemicalgenetic characteristics of native trout populations ofOwyhee County, Idaho. Final Report. Bureau of LandManagement, Contract ID-010-DT9–20, Boise, Idaho.

Wishard, L. N., J. E. Seeb, F. M. Utter, and D. Stefan. 1984. Agenetic investigation of suspected redband trout popula-tions. Copeia 1984(1):120–132.

Zoellick, B. W., P. Olmstead, D. Allen, and B. Flatter. 2004. Along-term comparison of redband trout density and sizestructure in southwestern Idaho. North American Jour-nal of Fisheries Management 25:1179–1190.

90

JOHNSON AND FITE


91

Oregon chub Oregonichthys crameri (Sny-der 1908) are small floodplain minnowsendemic to the Willamette Valley of west-

ern Oregon (Markle et al. 1991). Historically, thisspecies was widely distributed throughout theWillamette Valley (Markle et al. 1991). Studiesconducted in the 1970s and 1980s (Bond 1974;Bond and Long 1984; Markle et al. 1991) foundthe distribution of Oregon chub to be severelyrestricted. The loss of habitat and their restrictedrange led to their listing as endangered under theEndangered Species Act in 1993 (U.S. Fish andWildlife Service 1993).

Oregon chub are small cyprinids that grow to75–80 mm total length (TL), mature at age 2 (>40mm) and are relatively long-lived (up to 9 years).Oregon chub prefer off-channel habitats with min-imal or no current velocity, an abundance of vege-tation, and depositional substrate (Pearsons 1989;Scheerer 2002). They spawn in aquatic vegetationfrom May to July when water temperatures exceed15°C (Scheerer and McDonald 2003).

The Willamette River originates in theCalapooya Mountains of southwestern Oregon

and flows in a northerly direction through theWillamette Valley to the Columbia River, a distanceof 474 km (Sedell and Froggatt 1984). It is a ninth-order channel, drains an area of 29,728 km2, and isthe tenth largest river in the continental UnitedStates in terms of total discharge. The climate inthe Willamette Valley is humid temperate with anannual precipitation of 127 cm.

Historically, Oregon chub thrived in an uncon-strained Willamette River under a hydrologicregime that featured frequent flood events (Bennerand Sedell 1997), which continually created anddestroyed off-channel habitats (Lewin 1978;Dykaar and Wigington 2000). Floods provided themechanism of dispersal and genetic exchangeamong isolated off-channel habitats for Oregonchub populations.

Today, the Willamette River is a highly alteredsystem. In the past 150 years, the channel lengthand complexity of the Willamette River has beendrastically reduced by the construction of 13 majorflood control dams, large scale removal of largewoody debris for navigation, channelization andrevetments, and the drainage of wetlands to

Improved Status of the Endangered Oregon Chub in the Willamette River, Oregon

ABSTRACT Status and trends in the abundance of populations of federally endangered Oregon chub Orego-nichthys crameri, small floodplain minnows endemic to the Willamette Valley of western Oregon, were investigated

by estimating fish abundance and from extensive fish surveys of 650 off-channel habitats from 1991 through 2004.

The recent discovery of previously unknown populations of Oregon chub, some occurring in subbasins where they

were presumed extinct, combined with successful reintroductions into suitable habitats have resulted in the

improved status of this species. In 1991, eight populations of Oregon chub were known to exist. In 2004, we iden-

tified 33 populations of Oregon chub in the Willamette River basin. Ten of these populations, including the two most

abundant populations, were introduced. The status of Oregon chub is approaching the recovery plan goal for down-

listing the species to threatened. Nonnative fishes, which were found to be widespread in off-channel habitats pre-

ferred by Oregon chub, are the largest threat to full recovery and delisting of this species.

Paul D. Scheerer



PAUL D. SCHEERER Oregon Department of Fish and Wildlife, 28655 Highway 34, Corvallis, Oregon 97333, USA; [email protected]


increase the land available for river bottomlandagriculture (Sedell and Froggatt 1984; Benner andSedell 1997). Floods in the winter and springmonths were common prior to the construction ofthe dams (1941–1969), averaging 14 floods abovebank-full per decade from about 1884 through1969 (Corps of Engineers 1970). What was consid-ered a 10-year flood event prior to construction ofthe dams now has a 100-year return interval (Ben-ner and Sedell 1997). Channelization and the con-struction of flood control dams restricts or elimi-nates many of the linkages and interactionsbetween the river and its floodplain (Gabriel 1993)and has been detrimental to native fishes that relyon floodplain habitats (Bayley 1991; Osmundsonand Burnham 1998; Modde et al. 2001).

Introduction of nonnative fishes in WillametteRiver began in the late 1800s (Dimick and Merry-field 1945; Lampman 1946; McIntosh et al. 1989).Nonnative centrarchids and bullhead catfishesAmeiurus spp. have been widely implicated in thedecline of native fishes (Moyle 1976; Lemly 1985;Rinne and Minckley 1991; Newman 1993; Simonand Markle 1999), are common in the WillametteRiver basin and are considered to be the greatest cur-rent threat to the recovery of Oregon chub (U.S. Fishand Wildlife Service 1998; Scheerer 2002).

Markle et al. (1991) found that nonnative fish-es were common in historic Oregon chub habitatsthat no longer contained Oregon chub. Scheerer(2002) found that Oregon chub were absent, or inlow in abundance, when nonnative fishes werepresent in off-channel habitats and described sev-eral Oregon chub populations that declined orwere extirpated when their habitats were invadedby nonnative fishes following flood events or whennonnative fishes were illegally stocked.

This paper describes (1) the current known dis-tribution, status, and trends of Oregon chub pop-ulations and their habitats in the Willamette River,(2) the status of reintroductions, and (3) progresstowards the initial goal of downlisting this speciesfrom endangered to threatened status.

Methods

Oregon chub distributional surveys were conduct-ed by the Oregon Department of Fish and Wildlife

throughout the Willamette Valley from 1991through 2004. A total of 650 off-channel habitatsand small tributaries were sampled in theWillamette River basin.

Fish sampling was conducted using a combina-tion of gear types. A minimum of 20% of the sur-face area of each site was sampled within the rangeof habitat types present at each location. Mosthabitats were sampled using a 1 x 5 m seine (64-mm mesh). In deep sites (greater than 1.5 m max-imum depth) and/or sites where seining was inef-ficient because of large amounts of woody debris,baited minnow traps and dip nets (32-mm mesh)were used. Dipnetting was conducted in shallowshoreline areas and around woody debris. A gillnet (four panels measuring 7.6 m long x 1.8 mdeep, with square mesh sizes of 127, 191, 254, and381 mm) was also used at certain locations to cap-ture larger, mostly nonnative fishes. The gill netwas set for a minimum of 2 h and extended fromthe shore into deeper water. All fish captured wereidentified, counted, and their length recordedwithin 25-mm increments.

Population estimates were obtained for adultOregon chub (greater than 35 mm TL) at selectedlocations between 1992 and 2004. When catch rateswere very low, attempts to estimate abundancewere abandoned. Low numbers of centrarchid fish-es were captured during our sampling efforts com-pared to visual observations of their abundance.No population estimates were obtained for thesespecies. Minnow traps measuring 23 by 46 cm with64-mm mesh were used to capture fish for mark-ing. The traps were baited with a half of a slice ofbread. Minnow traps were regularly spaced at adensity of one trap per 100–250 m2 of surface area,up to 60 traps per site, to ensure that fish weremarked from all locations within the pond orslough. Traps were typically fished for 3–4 h duringthe day. If catch rates were low during the day set,traps were set overnight (up to 18 h). All fish weregiven a partial caudal fin clip before returningthem to the water. Marked fish were distributedthroughout the pond to promote mixing with theunmarked population. Population size was esti-mated using an adjusted Peterson mark–recaptureprocedure (Ricker 1975) from the total number ofmarked fish, and the catch and recaptures from the

92

SCHEERER


last sample date, if more than 1 d of marking wasrequired. The last sample date was separated by aminimum of 2 d from the last date of marking, toallow time for adequate mixing of the marked andunmarked segments of the population. Confidenceintervals were calculated using a Poisson approxi-mation (Ricker 1975). No estimates were made forage-0 fish (less than 35 mm), which were too smallto be efficiently captured in the minnow traps(Scheerer and McDonald 2003). Regeneration ofcaudal fins was rapid following marking, with sub-stantial regeneration noted as early as 3–4 weekspostmarking. In subsequent years, caudal fin clipsfrom the previous years’ marking were almostcompletely regenerated and easily distinguishedfrom new fin clips.

The connectivity of a habitat to the river orreservoir was described for sites containing Ore-gon chub, based on the degree of isolation of theoff-channel habitat from the adjacent water body.Sites with high connectivity had year-round con-nection, had yearly influx of water during the win-ter and/or spring months, or were connected by aculvert to the adjacent river or reservoir. Sites withlow connectivity were isolated from the adjacentriver by impassable culverts, high beaver dams,and/or regulated flows. All sites characterized withlow connectivity remained isolated during two1996 flood events (approximate 20-year recur-rence interval postdams).

Abundance estimates were used to determinethe status of Oregon chub in relation to recoverycriteria set forth in the Oregon Chub RecoveryPlan (U.S. Fish and Wildlife Service 1998). Therecovery plan adopted the following criteria fordownlisting the species from endangered to threat-ened: (1) establish and manage 10 populations ofat least 500 adults each; (2) all populations mustexhibit stable or increasing trends for 5 years; and(3) ensure that at least three populations are locat-ed in each of the three major recovery areas (Mid-dle Fork Willamette, Santiam, and main-stemWillamette subbasins). Delisting will occur whenthere are 20 populations totaling 500 or moreadults, which maintain a stable or increasing trendfor 7 years. At least four populations must be locat-ed in each of the three recovery areas. Managementof these populations must be guaranteed in perpe-

tuity. Abundance trends were defined quantitative-ly as increasing, declining, stable, not declining, orunknown. A linear regression of abundance overtime was calculated for each population for thepast 5 years (2000–2004). When the slope of thisregression was negative and significantly differentfrom zero (P < 0.10), the population was defined asexhibiting a declining trend in abundance. Whenthe slope was positive and significantly differentfrom zero (P < 0.10), the population was defined asexhibiting an increasing trend in abundance. Whenthe slope was not significantly different from zero(P > 0.10), then the coefficient of variation was cal-culated for the abundance estimates for the past 5years. When this coefficient of variation was lessthan 1.0 then the population was defined as stable.Otherwise, the population was defined as notdeclining in abundance. At locations with fewerthan 5 years of data or where no abundance esti-mates were obtained due to low catch rates, theabundance trend was defined as unknown.

Results

In 2004, 33 locations were identified that containedOregon chub in the Willamette basin (Table 1; Fig-ure 1). Eight of these were on the list of 29 histori-cal sites in Oregon State University fish museumrecords, 10 were locations where populations wereintroduced between 1988 and 2004, and 15 werenew populations discovered since 1991 (Figure 2).Distribution included populations located in theSantiam River (6 sites), McKenzie River (3 sites),Coast Fork Willamette River (2 sites), mid-Willamette River (9 sites), and the Middle ForkWillamette River (13 sites) (Figures 1 and 2).

Known chub distribution and status hasimproved substantially since the listing in 1998(Table 2). This change in distribution and status ofOregon chub is likely the result of increased sam-pling effort, rather than range expansion, sincemany of the newly discovered populations werefound in isolated habitats. The number of chubpopulations known in 1998 was 18, compared to33 in 2004. The number of abundant chub popula-tions (500 or more fish) increased from 9 in 1998to 15 in 2004. The number of populations thatmeet the recovery plan criteria (500 or more fish

93

IMPROVED STATUS OF THE ENDANGERED CHUB


94

Tabl

e 1.

Or

egon

chu

b po

pula

tion

abun

danc

e,st

atus

,5-y

ear t

rend

s,si

te c

onne

ctiv

ity,a

nd p

rese

nce

of n

onna

tive

fish.

Mar

k–re

capt

ure

popu

latio

n es

timat

es w

ere

obta

ined

at l

ocat

ions

whe

n po

ssib

le.W

hen

abun

-da

nce

was

low

,the

tota

l num

ber o

f fis

h ca

ptur

ed is

sho

wn

in b

old.

Num

bers

in p

aren

thes

es a

re th

e nu

mbe

r of f

ish

intro

duce

d at

the

site

.In

the

Oreg

on C

hub

Reco

very

Pla

n,th

e M

cKen

zie R

iver

sub

basi

n is

incl

uded

in th

e m

ain-

stem

Will

amet

te R

iver

reco

very

are

a.Si

te lo

catio

ns w

ere

assi

gned

a c

ode

cons

istin

g of

one

or m

ore

lette

rs fo

llow

ed b

y a

num

ber.

The

larg

er th

e po

pula

tion

in e

ach

subb

asin

,the

smal

ler t

he n

umbe

r.Si

tes

in th

e M

iddl

e Fo

rk W

illam

ette

Riv

er d

rain

age

wer

e co

ded

with

the

lette

r “M

,”in

the

Coas

t For

k W

illam

ette

Riv

er b

asin

with

the

lette

r “C,

”in

the

McK

enzie

Riv

er d

rain

age

with

the

lette

r “M

CK,”

in th

e Sa

ntia

m R

iver

dra

inag

e w

ith th

e le

tter “

S,”

and

in th

e m

ain-

stem

and

mid

-Will

amet

te R

iver

trib

utar

ies

with

the

lette

r “W

.”

Site

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Tren

d (5

-yea

r)Co

nnec

tivity

Nonn

ativ

e fis

hes

W1a

(200

)(3

73)4

604,

860

14,0

9026

,240

19,2

7028

,740

25,8

10st

able

low

no

M1a

(500

)48

01,

420

6,31

0

5,03

07,

770

6,37

05,

620

5,85

0st

able

low

no

M2

690

780

3,16

03,

030

3,02

02,

980

2,70

02,

130

1,60

04,

940

stab

lelo

wno

M3a

30

19

2516

04,

580

4,08

02,

410

4,10

04,

780

stab

lelo

wno

M4

1,63

04,

770

3,77

04,

240

3,79

03,

650

2,86

03,

830

2,28

02,

420

2,33

04,

210

stab

lelo

wno

M5

3,01

03,

570

7,14

04,

080

2,83

03,

600

stab

lelo

wno

M6

8,77

07,

540

7,13

04,

470

4,02

04,

440

4,78

0

5,05

03,

380

3,27

03,

650

3,14

0de

clin

ing

low

no

M7

4,01

01,

910

2,01

05,

350

2,72

0 1,

190

3,97

04,

910

2,14

02,

950

stab

lelo

wno

S18,

340

8,66

01,

830

860

360

760

740

1,59

02,

290

incr

easi

nglo

wb

no

M8

1,06

01,

170

1,09

094

061

01,

340

stab

lelo

wno

MCK

112

065

01,

050

unkn

own

low

no

M9c

5915

1,33

083

050

880

1,95

02,

270

870

790

stab

lehi

ghye

s

MCK

2a(3

50)

(150

)470

450

720

unkn

own

low

no

S2a

(85)

(20)

80(7

5)21

0(5

0)32

0(1

58)6

40

(112

)570

incr

easi

nglo

wno

W2

370

600

460

470

520

620

510

730

630

290

230

520

stab

lehi

ghye

s

W3a

(500

) un

know

nlo

wno

W4a

(500

) un

know

nlo

wno

M10

2,78

042

0un

know

nhi

ghye

s

C1a

(400

)42

035

0un

know

nlo

wno

S32

32

013

42

1227

034

0in

crea

sing

high

yes

S45

23

1313

350

220

320

incr

easi

nghi

ghye

s

MCK

394

062

031

0un

know

nhi

ghno

W5a

(50)

5022

0un

know

nlo

wno

(con

tinue

d)

SCHEERER


95


Site

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

Tren

d (5

-yea

r)Co

nnec

tivity

Nonn

ativ

e fis

hes

C216

130

190

unkn

own

high

yes

M11

c78

014

040

920

450

1,13

01,

440

800

460

390

70de

clin

ing

high

yes

W6a

(60)

(4

5)36

01,

750

(49)

670

500

130

70st

able

low

no

S53

41

00

021

un

know

nhi

ghye

s

W7

5un

know

nhi

ghye

s

M12

03

72

12

unkn

own

high

no

W8d

262

221

unkn

own

high

no

M13

48

221

480

140

140

91

1de

clin

ing

high

no

S61,

250

830

320

250

134

122

01

decl

inin

ghi

ghye

s

M14

130

extin

ct?

high

yes

S72

00

29

46

0ex

tinct

?hi

ghno

S85

25

02

03

24

0ex

tinct

?hi

ghye

s

M15

76

12

22

20

extin

ct?

high

yes

M16

30

70

00

82

0ex

tinct

?hi

ghye

s

S9a

(15)

7(2

6)29

00

extin

ct?

low

no

M17

310

220

00

00

0ex

tinct

?hi

ghye

s

M18

60

10

10

00

00

extin

ct?

high

yes

M19

a(5

76)

3,52

05,

610

7,16

03,

490

600

00

00

extin

ct?

high

yes

M20

30

00

00

00

extin

ct?

high

yes

M21

400

00

00

00

0ex

tinct

?hi

ghno

M22

30

00

00

00

0ex

tinct

?hi

ghye

s

C31

20

00

00

00

0ex

tinct

?hi

ghye

s

S10e

20

extin

ct?

high

no

W9f

55

unkn

own

high

yes

M23

a3

(525

)2

00

00

extin

ct?

high

yes

aOr

egon

chu

b re

intro

duct

ion

site

s.b

Conn

ectiv

ity c

hang

ed fr

om h

igh

to lo

w in

200

1 w

hen

a po

rtion

of t

his

habi

tat w

as is

olat

ed (s

cree

ned)

to e

xclu

deno

nnat

ive

fishe

s.No

nnat

ive

fishe

s w

ere

not c

olle

cted

in 2

001–

2004

.c

Nonn

ativ

e fis

hes

have

acc

ess

to th

ese

site

s th

roug

h a

culv

ert c

onne

cted

to a

rese

rvoi

r.No

nnat

ive

fish

wer

e co

llect

edfro

m s

ite M

9 in

199

9,20

03,a

nd 2

004.

Nonn

ativ

e fis

h w

ere

colle

cted

from

site

M11

in 1

996,

2002

,and

200

4.

dAc

cess

was

den

ied

1998

–200

2.e

Site

was

des

troye

d by

199

6 flo

ods

and

no lo

nger

exi

sts.

fAc

cess

was

den

ied

1997

–200

4.

Tabl

e 1.

(co

ntin

ued)


and stable or increasing in abundance for 5 years)increased from 2 to 12 over this same period. Fourof the 12 populations that meet recovery plan cri-teria were reintroductions, and four were popula-tions known to exist historically. In addition, theknown distribution expanded from three to fivemajor subbasins from 1998 to 2004, includingrecent collections in the McKenzie River subbasinand the Coast Fork Willamette River subbasin, lastcollected in 1899 and 1993, respectively.

Oregon chub population abundance estimatesand 5-year trends are shown in Table 1. Abundanceestimates ranged from 40 fish to more than 28,000fish per population. The lower 95% confidencelimits for the estimates were fairly tight, averaging75% of the estimate (95% confidence interval:73–77%; range: 50–94%). Oregon chub were morewidespread and abundant in the Middle ForkWillamette River drainage than in the other

Willamette River subbasins. Nine of the 15 largestpopulations (500 or more fish) were found in theMiddle Fork Willamette River drainage. Of the 12populations that met the recovery plan criteria in2004 (stable or increasing in abundance for 5years), eight were located in the Middle ForkWillamette River subbasin, two were located in theSantiam River subbasin, and two were located inthe mid-Willamette subbasin.

Oregon chub were more abundant at sites wherenonnative fishes were absent (Table 1). In 2004,only 2 of the 15 locations that supported large pop-ulations of Oregon chub (500 or more fish) con-tained nonnative fishes. Conversely, nonnative fish-

96

SCHEERER

Figure 1. Distribution of sampling effort during 1991–2004surveys in the Willamette River basin. Circles with centerdots indicate sites where sampling occurred. Solid circlesindicate the current distribution of Oregon chub.

Figure 2. Distribution of Oregon chub populations in theWillamette River basin in 2004. Open squares representhistorical Oregon chub locations where no Oregon chubcurrently exist. Closed squares represent historical Oregonchub locations where Oregon chub currently exist. Dia-monds represent newly discovered Oregon chub popula-tions (since 1991). Circles with center dots representintroduced Oregon chub populations. Bold dashes represent the locations of major flood control dams.


es were present in many of the locations where Ore-gon chub abundance was low or where chub arepresumed extinct. Eight of the 15 sites supportingsmall chub populations (fewer than 500 fish) con-tained nonnative fishes (53%), and 11 of the 15 siteswhere chub are presumed to be extinct containednonnative fishes (73%). Sites with low connectivity(n = 8) supported larger naturally occurring popu-lations of Oregon chub (P < 0.001) and containedfewer species of nonnative fishes (P < 0.001) thansites with high connectivity (n = 24).

Forty-two percent of all locations sampled inthe Willamette River drainage, and 53% of thelocations where any fish were present containednonnative fishes (Table 3). Nonnative fishes werecollected most frequently from off-channel habi-tats in the main-stem Willamette River (59%),the Coast Fork Willamette River drainage (55%),mid-Willamette River tributaries (43%), SantiamRiver drainage (42%), and the lower WillametteRiver tributaries (37%). In these subbasins, natu-rally occurring chub populations frequentlyoccurred in lower abundance and/or were indecline. The Middle Fork Willamette Riverdrainage, which contained the largest concentra-tion of abundant chub populations, and theMcKenzie River drainage, where two populationswere discovered in 2002, had the lowest occur-

rence of nonnative fishes in off-channel habitats(28% and 22%, respectively).

Nonnative fishes invaded several Oregonchub locations during the course of this study.These include two Santiam River locations (sitesS1 and S7) that were invaded during flooding in1996 and one Middle Fork Willamette Riverlocation (site M19), which was illegally stockedwith largemouth bass Micropterus salmoides in1997 (Table 1). The Oregon chub populationssubsequently declined at these locations. At oneSantiam location (S1), we were successful in iso-lating a portion of the habitat by screening in2000 and 2001. No centrarchids were capturedfrom this isolated habitat in 2001 through 2004,and the chub population has since increased inabundance (Table 1).

Discussion

Oregon chub status and the known distributionhas improved over the past 13 years, resulting fromthe discovery of new populations through exten-sive surveys of off-channel habitats and from theestablishment of new populations through success-ful reintroductions within their historical range.

Determining the distribution of a smallcyprinid, Oregon chub, in large river basin such as

97


Table 2. Status of Oregon chub populations by subbasin in 1998 and 2004. The number of populations includes the total number known ineach subbasin, the number of large populations (500 or more adults), the number of viable populations (i.e., those that meet thecriteria in the recovery plan [500 or more adults and exhibiting a stable or increasing trend for five years]), the number of intro-duced populations (included in the total number of populations), and the number of populations presumed extinct. The Oregonchub recovery plan includes the McKenzie River subbasin in mid-Willamette River subbasin. The totals listed for the mid-Willamette subbasin do not include totals for the McKenzie subbasin.

Middle Fork Coast ForkNumber of Santiam Mid-Willamette McKenzie Willamette Willamette All subbasinspopulations 1998 2004 1998 2004 1998 2004 1998 2004 1998 2004 1998 2004

Total 4 6 2 9 0 3 12 13 0 2 18 33

Large 1 2 1 2 0 2 7 9 0 0 9 15

Viable 0 2 1 2 0 0 1 8 0 0 2 12

Introduceda 0 1 2 5 0 1 4 2 0 1 6 10

Extinctb 0 3 0 0 0 0 4 9 1 1 5 13_________________________a Three introductions failed (sites M19, M23, and S9) and are not included in totals. Total attempted reintroductions is 13.b Failed introductions were not included in the number of presumed extinct populations listed.


the Willamette River, has been a challenge consider-ing that 64% of the watershed is in private owner-ship (Hulse et al. 2002) and historic species recordsare sparse (29 records from 1894 to 1979) (Markleet al. 1991). In 1990, when the species was peti-tioned for listing, only three populations, restrictedto 30 km of the Middle Fork Willamette River, wereknown to exist (Markle and Pearsons 1990). In1998, 12 naturally occurring populations wereidentified, distributed over three major subbasins.Six years later in 2004, 23 naturally occurring pop-ulations were identified, distributed over five majorsubbasins. These data illustrate the importance ofextensive sampling efforts to determine the distri-bution of small rare fishes in large watersheds.

Reintroduction is a common managementaction for endangered species recovery (IUCN1995). We have established 10 introduced popula-tions of Oregon chub in suitable habitats withintheir current range; 4 have been quite successful.We chose locations that had suitable chub habitatwith low connectivity to reduce the risk of invasionby nonnative fishes (U.S. Fish and Wildlife Service1998; Scheerer 2002). To minimize the potentialgenetic consequences of our activities (genetic

drift, bottlenecks, and inbreed-ing) in the absence of geneticdata, we transferred a minimumof 500 fish when we conductedintroductions to establish alarge effective population sizeand we used multiple donorpopulations when possible(Scheerer 2002). When possible,we chose donor populationsthat were large (more than1,000 fish) and always limitedthe number of fish transferredfrom the donor site to 10% ofthe estimated adult populationabundance. Successive annualtransfers were sometimes need-ed to achieve the minimumstocking target of 500 fish.

Metapopulation theory statesthat the probability of extinctionof local populations decreaseswith increased population size

(Hanski and Gilpin 1997). A goal in recovery ofthreatened and endangered fish species is toincrease the population size enough to withstandstochastic fluctuations and maintain sufficientgenetic diversity (Li et al. 1995; Fagan 2002). Ore-gon chub expand rapidly in suitable habitats, as evi-denced by successful reintroductions, achieve largepopulation size at most isolated sites, and tend to bethe numerically dominant fish species in thesehabitats. These species characteristics act to reducethe risk of extinction and facilitate the recovery ofOregon chub.

Fragmentation disrupts patterns of immigra-tion and emigration, which can restrict recolo-nization of habitats and reduce gene flow amongpopulations (Lafferty et al. 1999). Habitat alter-ation in the Willamette River has resulted in frag-mentation of Oregon chub habitat. Since nonna-tive fishes are common in habitats preferred byOregon chub, large chub populations tend to befound in isolated habitats with low connectivity.Exchange among populations is likely minimal ornonexistent because the hydrology of theWillamette River has been severely altered. Conse-quently, reduced gene flow may be a significant

98

SCHEERER

Table 3. The number of locations sampled in subbasins of the Willamette River and the number and percentage of these locations that contained nonnative fishes. LowerWillamette tributaries include all tributaries downstream of the Santiam River.Main-stem Willamette River locations include side channels and backwaters to themain stem. Mid-Willamette tributaries include all tributaries to the Willamette Riverbetween the confluence of the Coast and Middle Forks and the Santiam River, exceptfor the McKenzie River. The total number of sites with fish excludes sites where nofish were collected that likely desiccate annually (isolated wetlands).

Number Locations containing nonnative fishesSubbasin of locations Number Percentage

Lower Willamette tributaries 51 19 37

Main-stem Willamette River 99 58 59

Mid-Willamette tributaries 115 49 43

Santiam River 122 51 42

McKenzie River 54 12 22

Coast Fork Willamette River 84 46 55

Middle Fork Willamette River 125 35 28

All subbasins 650 270 42

Sites with fish 508 270 53


impediment to Oregon chub recovery. Geneticinvestigations are currently being conducted todetermine the levels of genetic diversity withinand among Oregon chub populations and to helpdetermine the genetic relationships among popu-lations relative to their spatial distribution.

Habitat degradation and introduced species,major factors implicated in the decline of Oregonchub (Markle et al. 1991; U.S. Fish and WildlifeService 1998), have been implicated in the declineof native minnows throughout the western UnitedStates (Cross 1976; Kaeding et al. 1990; Blinn et al.1993; Scoppettone 1993; Meng and Moyle 1995;Marsh and Douglas 1997; Sommer et al. 1997;Osmundson and Burnham 1998; Simon andMarkle 1999; Scheerer 2002). Most large westernstreams have been dammed creating permanentlentic environments (reservoirs) and more stableflow regimes than previously existed. These condi-tions tend to favor introduced species that weremost common in lakes and river backwaters intheir native range (Moyle 1986). These introducedfishes may eliminate native fishes directly by preda-tion or indirectly through competition forresources or higher reproductive success (Schoen-herr 1981; Taylor et al. 1984). In the western Unit-ed States, many native minnows are unable toestablish populations in reservoirs or other slack-water habitats, likely due to the presence of nonna-tive centrarchids (Moyle et al. 1986). Rahel (1984)found native cyprinids were absent in lakes thathad centrarchid or northern pike Esox lucius pred-ators. Moyle (1976) documented the extinction ofsix native fish species in Clear Lake, California fol-lowing the introduction of centrarchids. Meng etal. (1994) found native fishes were forced intosmaller, dead-end sloughs when nonnative fisheswere present in the system, and over time, thenatives declined in abundance in these habitats aswell. In the Willamette basin, large populations(500 or more fish) of Oregon chub were found pri-marily in isolated off-channel habitats. Sites withhigh connectivity frequently contained centrar-chids and rarely contained Oregon chub. The fre-quency of occurrence of nonnative predators washighest in the main-stem Willamette River and inflood control reservoirs and was higher in mostdownstream subbasins compared to the subbasins

in the upper watershed. Conversely, naturallyoccurring Oregon chub populations were concen-trated primarily in the upstream subbasins and inisolated habitats where nonnative fish occurrencewas less common. Many of the isolated habitats inthe Middle Fork Willamette River basin are locatedin close proximity to flood control dams. Thesehabitats are less likely to be impacted by floodevents that can transport nonnative fishes intothese habitats. This isolation may account for thelarge concentration of abundant Oregon chubpopulations in this subbasin.

In large alluvial rivers, declines of populationsof native floodplain fishes have been attributed toaltered river-floodplain connectivity and functionas well as impacts from nonnative fishes (Marshand Brooks 1989; Minckley 1982, Modde et al.2001; Moyle 1976; Mueller 1995; Tyus 1987).Floodplain habitats increase the productivity anddiversity of riverine communities (Bayley 1995;Gutreuter et al. 1999; Junk et al. 1989) and canprovide survival and growth advantages to fish(Starrett 1951; Peterson 1982; Tyus 1987; Kwak1988; Matheney and Rabeni 1995; Osmundsonand Burnham 1998; Modde et al. 2001; Sommer etal. 2001). Native fishes are adapted to natural flowregimes in which they evolved. These adaptationsprovide advantages over nonnative fishes duringperiodic flooding (Gido et al. 1997; Meffe 1984;Minckley 1981; Minckley and Meffe 1987;Osmundson and Kaeding 1991). Damming ofrivers has made them more lake-like, thus elimi-nating the adaptive advantages of the native fishesand conferring the advantage to nonnative fishes,like centrarchids. However, in certain situations,flooding has resulted in the invasion of habitats bynonnative fishes with subsequent declines innative fishes (Lafferty et al. 1999; Meffe 1983;Scheerer 2002). Schultz et al. (2003) found theabundance of some nonnative species that inhab-ited reservoirs and creeks increased in thedammed creek following flooding, probably dueto the movement of nonnative fishes out of thereservoir. Stanford and Ward (1986) documentedthe spread of nonnative fishes downstream ofimpoundments in the Colorado River basin. The13 large flood control reservoirs in WillametteRiver basin are both sources of nonnative fishes as

99



well as barriers (physical and biological) to suc-cessful migration by Oregon chub.

The discovery of new populations, some in sub-basins where they were presumed extinct, and theestablishment of additional populations throughreintroductions has substantially improved theknown status of Oregon chub. Meeting the recov-ery criteria to downlist Oregon chub to threatened(i.e., maintaining 10 large populations of Oregonchub that exhibit a stable or increasing trend ofabundance for 5 years) can likely be achieved with-in the next 5–7 years. Future introductions andadditional surveys in previously unsurveyed habi-tats will likely assist in meeting this goal.

Acknowledgments

I am grateful for the financial support from theU.S. Fish and Wildlife Service, U.S. Army Corps ofEngineers, Oregon Department of Transportation,and U.S. Department of Agriculture- Forest Ser-vice. Thanks to the following field assistants: T.Cornwell, P. Kavanagh, S. Loerts, C. Mease, E. Rich-mond, C. Shafer, C. Stein, and J. Stevens. Thanks toJ. Scheurer and an anonymous reviewer for con-structive editorial suggestions.

References

Bayley, P. B. 1991. The flood pulse advantage and the restora-tion of river-floodplain systems. Regulated Rivers:Research and Management 6:75–86.

Bayley, P. B. 1995. Understanding large river-floodplainecosystems. Bioscience 45:153–158.

Benner, P. A., and J. R. Sedell. 1997. Upper Willamette Riverlandscape: a historic perspective. Pages 23–47 in A. Laenenand D. A. Dunnette, editors. River quality: dynamics andrestoration. CRC Press, Boca Raton, Florida.

Blinn, D. W., C. Runck, and D. A. Clark. 1993. Effects ofrainbow trout predation on Little Colorado spinedace.Transactions of the American Fisheries Society122:139–143.

Bond, C. E. 1974. Endangered plants and animals of Oregon.I. Fishes. Oregon State University, Agricultural Experi-ment Station Revised Special Report 205, Corvallis.

Bond, D. D., and J. J. Long. 1984. Status of the Oregon chub,Hybopsis crameri. Proceedings of the Annual Conferenceof the Western Association of Fish and Wildlife Agencies64:483–486.

Corps of Engineers. 1970. Flood plain information,Willamette River, Johnson, Kellog and Mt. Scott Creeks.U.S. Army Engineer District, Portland, Oregon.

Cross, J. N. 1976. Status of the native fish fauna of the Moapa

River (Clark County, Nevada). Transactions of theAmerican Fisheries Society 105:503–508.

Dimick, R. E., and F. Merryfield. 1945. The fishes of theWillamette River system in relation to pollution. OregonState College, Engineering Experiment Station, Bulletin20, Corvallis.

Dykaar, B. B., and P. J. Wigington, Jr. 2000. Floodplain for-mation and cottonwood colonization patterns on theWillamette River, Oregon, USA. Environmental Manage-ment 25:87–104.

Fagan, W. F. 2002. Connectivity, fragmentation, and extinc-tion risk in dendritic metapopulations. Ecology83:3242–3249.

Gabriel, J. T. 1993. A modified synoptic analysis of the statusof Oregon’s Willamette Valley wetlands. M. S. thesis. Ore-gon State University, Corvallis.

Gido, K. B., D. L. Propst, and M. C. Molles. 1997. Spatial andtemporal variation of fish communities in secondarychannels of the San Juan River, New Mexico and Utah.Environmental Biology of Fishes 49:417–434.

Gutreuter, S., A. D. Bartels, K. Irons, and M. B., Sandheinrich.1999. Evaluation of the flood-pulse concept based onstatistical models for growth of selected fishes of theupper Mississippi River system. Canadian Journal ofFisheries and Aquatic Sciences 56:2282–2291.

Hanski, I., and M. E. Gilpin. 1997. Metapopulation biology:ecology, genetics, and evolution. Academic Press, SanDiego, California.

Hulse, D., S. Gregory, and J. Baker, editors. 2002. WillametteRiver basin planning atlas: trajectories of environmentaland ecological change. Oregon State University Press,Corvallis.

IUCN (International Union for the Conservation ofNature). 1995. IUCN/SSC guidelines for re-introduc-tions. In 41st Meeting if the IUCN council, Gland,Switzerland. Available: www.iucn.org/themes/ssc/pubs/policy/reinte.htm (May 1995).

Junk, W. J., P. B. Bayley, and R. E. Sparks. 1989. The floodpulse concept in river-floodplain systems. Pages 110–127in D. P. Dodge, editor. Proceedings of the InternationalLarge River Symposium. Canadian Special Publication ofFisheries and Aquatic Sciences 106:110–127.

Kaeding, L. R., B. D. Burdick, P. A. Schrader, and C. W.McAda. 1990. Temporal and spatial relations between thespawning of humpback chub and roundtail chub in theupper Colorado River. Transactions of the AmericanFisheries Society 119:135–144.

Kwak, T. J. 1988. Lateral movement and use of floodplainhabitat by fishes of the Kankakee River, Illinois. Ameri-can Midland Naturalist 120:241–249.

Lafferty, K. D., C. C. Swift, and R. F. Ambrose. 1999. Post-flood persistence and recolonization of endangered tide-water goby populations. North American Journal ofFisheries Management 19:618–622.

Lampman, B. H. 1946. The coming of the pond fishes, anaccount of the introduction of certain spiny-rayed fish-es, and other exotic species, into the waters of the lowerColumbia River region and the Pacific Coast states. Bin-fords and Mort, Publishers, Portland, Oregon.

100

SCHEERER


Lemly, A. D. 1985. Suppression of native fish populations bygreen sunfish in first-order streams of Piedmont NorthCarolina. Transactions of the American Fisheries Society114:705–712.

Lewin, J. 1978. Floodplain geomorphology. Progress in Phys-ical Geography 2:408–437.

Li, H. W., K. Currens, D. Bottom, S. Clarke, J. Dambacher,C. Frissell, P. Harris, R. M. Hughes, D. McCullough, A.McGie, K. Moore, R. Nawa, and S. Thiele. 1995. Safehavens: refuges and evolutionarily significant units.Pages 371–380 in J. L. Nielsen, editor. Evolution and theaquatic ecosystem: defining unique units in populationconservation. American Fisheries Society, Symposium17, Bethesda, Maryland.

Markle, D. F., and T. N. Pearsons. 1990. Petition to list theOregon chub, Oregonichthys crameri, as an endangeredspecies. Oregon State University, Corvallis, Oregon.

Markle, D. F., T. N. Pearsons, and D. T. Bills. 1991. Naturalhistory of Oregonichthys (Pisces: Cyprinidae), with adescription of a new species from the Umpqua River ofOregon. Copeia 1991(2):277–293.

Marsh, P. C., and J. E. Brooks. 1989. Predation by ictalurid cat-fishes as a deterrent to re-establishment of hatchery-rearedrazorback suckers. Southwestern Naturalist 34(2):188–195.

Marsh, P. C., and M. E. Douglas. 1997. Predation by intro-duced fishes on endangered humpback chub and othernative species in the Little Colorado River, Arizona. Trans-actions of the American Fisheries Society 126:343–346.

Matheney, M. P., and C. F. Rabeni. 1995. Patterns of move-ment and habitat use by northern hog suckers in anOzark stream. Transactions of the American FisheriesSociety 124:886–897.

McIntosh, B. A., S. E. Clarke, and J. R. Sedell. 1989. Summa-ry report for Bureau of Fisheries stream habitat surveys:Willamette River basin 1934–1942. U.S. Department ofEnergy, Bonneville Power Administration, DOE/BP-02246–2, Portland, Oregon.

Meffe, G. K. 1983. Attempted chemical renovation of an Ari-zona springbrook for management of the endangeredSonoran topminnow. North American Journal of Fish-eries Management 3:315–321.

Meffe, G. K. 1984. Effects of abiotic disturbance on coexis-tence of predator-prey fish species. Ecology65:1525–1534.

Meng, L., P. B. Moyle, and B. Herbold. 1994. Changes inabundance and distribution of native and introducedfishes of Suisun Marsh. Transactions of the AmericanFisheries Society 123:498–507.

Meng, L., and P. B. Moyle. 1995. Status of splittail in theSacramento–San Joaquin Estuary. Transactions of theAmerican Fisheries Society 124:538–549.

Minckley, W. L. 1981. Ecological studies of Aravaipa Creek,central Arizona, relative to past, present, and future uses.United States Bureau of Land Management, FinalReport, Safford, Arizona.

Minckley, W. L. 1982. Trophic interrelations among intro-duced fishes in the lower Colorado River, southwesternUnited States. California Fish and Game 68:78–89.

Minckley, W. L., and G. K. Meffe. 1987. Differential selection

for native fishes by flooding in stream communities ofthe American Southwest. Pages 93–104 in W. J. Matthewsand D. C. Heins, editors. Community and evolutionaryecology of North American stream fishes. University ofOklahoma Press, Norman.

Modde, T., R. T. Muth, and G. B. Haines. 2001. Floodplainwetland suitability, access, and potential use by juvenilerazorback suckers in the middle Green River, Utah.Transactions of the American Fisheries Society130:1095–1105.

Moyle, P. B. 1976. Inland fishes of California. University ofCalifornia Press, Berkeley.

Moyle, P. B. 1986. Fish introductions into North America:patterns and ecological impact. Pages 27–43 in H. A.Mooney and H. A. Drake, editors. Ecology of biologicalinvasions of North America and Hawaii. Springer-Ver-lag, New York.

Moyle, P. B., H. W. Li, and B. A. Barton. 1986. The Franken-stein effect: impact of introduced fishes on native fishesin North America. Pages 415–426 in R. H. Stroud, editor.The role of fish culture in fishery management. Ameri-can Fisheries Society, Bethesda, Maryland.

Mueller, G. 1995. A program for maintaining the razorbacksucker in Lake Mohave. Pages 127–135 in H. L.Schramm, Jr. and R. G. Pipers, editors. American Fish-eries Society, Symposium 15, Bethesda, Maryland.

Newman, L. 1993. The most endangered fish in Nevada: theWhite River spinedace (Lepidomeda albivallis). NorthAmerican Native Fishes Association, American Currents,summer issue, 1993, Baltimore, Maryland.

Osmundson, D. B., and K. P. Burnham. 1998. Status andtrends of the endangered Colorado squawfish in theupper Colorado River. Transactions of the AmericanFisheries Society 127:957–970.

Osmundson, D. B., and L. R. Kaeding. 1991. Recommenda-tions for flows in the 15-mile reach during October-Junefor maintenance and enhancement of endangered fishpopulations in the upper Colorado River. U.S. Fish andWildlife Service, Final Report, Grand Junction, Colorado.

Pearsons, T. N. 1989. Ecology and decline of a rare westernminnow: the Oregon chub (Oregonichthys crameri). M.S.thesis. Oregon State University, Corvallis.

Peterson, N. P. 1982. Population characteristics of juvenilecoho salmon (Oncorhynchus kisutch) overwintering inriverine ponds. Canadian Journal of Fisheries andAquatic Sciences 39:1303–1307.

Rahel, F. J. 1984. Factors structuring fish assemblages along abog lake successional gradient. Ecology 65:1276–1289.

Ricker, W. E. 1975. Computation and interpretation of bio-logical statistics of fish populations. Fisheries ResearchBoard of Canada Bulletin 191.

Rinne, J. N., and W. L. Minckley. 1991. Native fishes of aridlands: a dwindling resource of the desert southwest. U.S.Forest Service General Technical Report RM-206.

Scheerer, P. D. 2002. Implications of floodplain isolation andconnectivity on the conservation of an endangered min-now, Oregon chub, in the Willamette River, Oregon.Transactions of the American Fisheries Society131:1070–1080.

101



Scheerer, P. D., and P. J. McDonald. 2003. Age, growth, andtiming of spawning of an endangered minnow, Oregonchub (Oregonichthys crameri), in the Willamette basin,Oregon. Northwestern Naturalist 84:68–79.

Schoenherr, A. A. 1981. The role of competition in thereplacement of native fishes by introduced species. Pages173–203 in R. J. Naiman and D. L. Soltz, editors. Fishes inNorth American deserts. John Wiley Press, New York.

Schultz, A. A., O. E. Maughan, and S. A. Bonar. 2003.Effects of flooding on abundance of native and nonna-tive fishes downstream from a small impoundment.North American Journal of Fisheries Management23:503–511.

Scoppettone, G. G. 1993. Interactions between native andnonnative fishes of the upper Muddy River, Nevada.Transactions of the American Fisheries Society122:599–608.

Sedell, J. R., and J. L. Froggatt. 1984. Importance of stream-side forest to large rivers: the isolation of the WillametteRiver, Oregon, USA, from its floodplain by snagging andstreamside forest removal. Verhandlungen Interna-tionale Vereinigung fur theoretische und angewandteLimnologie 22:1828–1834.

Simon, D. C., and D. F. Markle. 1999. Evidence of a relation-ship between smallmouth bass (Micropterus dolomeiu)and decline of Umpqua chub (Oregonichthys kalawatseti)in the Umpqua basin, Oregon. Northwestern Naturalist80:110–113.

Snyder, J. O. 1908. The fishes of the coastal streams of Ore-gon and northern California. U.S. Bureau of FisheriesBulletin 27:153–189.

Sommer, T. R., R. Baxter, and B. Herbold. 1997. Resilience ofsplittail in the Sacramento–San Joaquin Estuary. Trans-actions of the American Fisheries Society 126:961–976.

Sommer, T. R., M. L. Nobriga, W. C. Harrell, W. Batham, andW. J. Kimmerer. 2001. Floodplain rearing of juvenile Chi-nook salmon: evidence of enhanced growth and survival.Canadian Journal of Fisheries and Aquatic Sciences58:325–333.

Stanford, J. A., and J. V. Ward. 1986. Fish in the Colorado sys-tem. Pages 385–402 in B. R. Davies and K. F. Walker, edi-tors. The ecology of river systems. Kluwer, Hingham,Massachusetts.

Starrett, W. C. 1951. Some factors affecting the abundance ofminnows in the Des Moines River, Iowa. Ecology 32:13–27.

Taylor, J. N., W. R. Courtenay, Jr., and J. A. McCann. 1984.Known impacts of exotic fishes in the continental Unit-ed States. Pages 322–373 in W. R. Courtenay, Jr. and J. R.Stauffer, Jr., editors. Distribution, biology, and manage-ment of exotic fishes. John Hopkins University Press,Baltimore, Maryland.

Tyus, H. M. 1987. Distribution, reproduction, and habitatuse of the razorback sucker in the Green River, Utah,1979–1986. Transactions of the American Fisheries Soci-ety 116:111–116.

U.S. Fish and Wildlife Service. 1993. Endangered and threat-ened wildlife and plants: determination of endangeredstatus for the Oregon chub. Federal Register58:53800–53804.

U.S. Fish and Wildlife Service. 1998. Oregon chub (Orego-nichthys crameri) recovery plan. U.S. Fish and WildlifeService, Portland, Oregon.

102

SCHEERER


103

Roundtail chub Gila robusta are a uniquenative fish in Arizona because they are man-aged as both a sport fish and a Species of

Special Concern (AZGFD 1996). Although not fed-erally listed as threatened or endangered, roundtailchub distribution and abundance is decreasingthroughout its range (Voeltz 2002). Reasons for thedecline include alteration of the historic hydro-graph (dam construction), habitat degradation,and predation by and competition with nonnativefishes (Minckley and Deacon 1991).

Prior to 1998, Arizona Game and Fish Depart-ment had few records of roundtail chub inhabitingthe lower Verde River below Bartlett Dam (Girmen-donk and Young 1997). Similarly, very few roundtailchub had been reported in the lower Salt River belowStewart Mountain Dam (Clarkson 1998). This ledbiologists to believe that the roundtail chub popula-tion in the lower Salt and Verde rivers was somewhatsparse. However, a multi-year research project con-ducted from 1999 to 2000 indicated that roundtailchub were more abundant than previously thought(Bryan and Robinson 2000).

Bryan and Robinson (2000) provided impor-tant baseline information for population size, sizestructure, habitat use, movement, and growth ofroundtail chub in the lower Salt and Verde rivers.They suggested that continued monitoring wasneeded to determine population stability and toidentify factors that affect reproduction andrecruitment. The primary objectives of this studywere to determine the current population size andsize structure of roundtail chub in the lower Saltand Verde rivers, Arizona. We also make compar-isons with results reported by Bryan and Robinson(2000; hereafter referred to as the 2000 study) toevaluate trends in the population status.

MethodsStudy area

The study was conducted in the lower Salt River,between Stewart Mountain Dam and GraniteReef Dam, and the lower Verde River, betweenBartlett Dam and its confluence with the SaltRiver (Figure 1). Both stretches of river are regu-

Roundtail Chub Population Assessment in the Lower Salt and Verde Rivers, Arizona

ABSTRACT Roundtail chub Gila robusta were collected during spring and fall 2003 in the lower Salt and

Verde rivers to determine population size and size structure. We collected and passive integrated transponder-

tagged 262 roundtail chub using a combination of experimental gill nets and canoe electrofishing. The majority of

roundtail chub were collected in the upper 16 km of the study area. Length-frequency distributions indicate that

the population is comprised almost entirely of large adults with minimal recruitment. The estimated population

size of roundtail chub in the lower Salt and Verde rivers during 2003 is 1,657 (95% confidence interval =

1,097–2,742), which represents a 74% decrease from 3 years ago. Based on these results, we conclude that the

roundtail chub population in the lower Salt and Verde rivers is declining rapidly due to low recruitment and high

natural mortality. We recommend that immediate management actions be taken to ensure the persistence of this

population of roundtail chub.

Scott D. Bryan and Matthew W. Hyatt



SCOTT D. BRYAN 1 and MATTHEW W. HYATT Arizona Game and Fish Department; 2221 West Greenway Road; Phoenix, Arizona 85023, USA1Current address: Aquatic Consultants, Inc., 4415 Hawkins Street NE, Suite D, Albuquerque, New Mexico 87109, USA


lated by Salt River Project (SRP) and providehydroelectric power and water to the Phoenixmetropolitan area.

The lower Salt River flows through Sonorandesert scrub and low gradient, tamarisk-mesquiteflats for approximately 21.5 km before reachingGranite Reef Diversion Dam, where it is divertedinto canals for the city of Phoenix. During winter,water is stored in the four reservoirs above StewartMountain Dam, and flows below the dam are held atthe required minimum of 8 ft3/s. In this time of lowflow, fish are restricted (and sometimes stranded) toshallow runs and several large pools (typically < 3 mdeep). During summer, water and electric demandsin the Phoenix area increase, and as a result, flows inthe lower Salt River increase to 800–1,200 ft3/s. Thehigh flows create a number of lateral scour pools,high gradient riffles (rapids), and swift runs.

The lower Verde River flows approximately 41km from Bartlett Dam to the Salt River confluence.The upper portion of the river has a high gradientand is bound by canyon walls. It then slowly opens

into desert scrub brush flatlands as it flows throughthe Tonto National Forest, Fort McDowell YavapaiNation, and Salt River Pima-Maricopa IndianCommunity. During winter, flows typically rangefrom 400 to 1,000 ft3/s, which provide highly vari-able habitat consisting of moderate to deep pools,long runs, and both high and low gradient riffles.Summer flows are held relatively constant, between125 and 250 ft3/s, and the river becomes very shal-low but still includes diverse habitat for fish.

Size structure and population estimate

We sampled the lower Verde River twice in springand twice in fall, 2003. The lower Salt River wassampled three times during 2003, twice in springand once in fall (equipment failure precluded asecond sample during fall). Sites where roundtailchub were collected during the 2000 study weresampled during each trip (fixed sites). In spring2003, we also sampled an equal number of randomsites in each river (44 in the Verde River, 7 in theSalt River). However, since no chub were collectedat random sites during spring sampling, weassumed that fixed sites were representative of thepopulation and sampled only fixed sites in fall.

Fish were collected using a combination of gillnetting and canoe electrofishing. Up to threeexperimental gill nets (2.4 x 45.5 m, with meshsizes of 12.7, 25.4, 38.1, 50.8, 63.5, and 76.2 mm)were set perpendicular to flow and the canoe elec-troshocker (600 V; 6–8 A) was fished over the nets.Nets were not fished longer than 30 min to avoidmortality. When flows were too high or water wastoo shallow to effectively fish gill nets, only thecanoe electroshocker was used.

Upon capture, roundtail chub were measured(total length [TL], ± 1 mm), weighed (±1 g), andscanned for the presence of an internal passive inte-grated transponder (PIT) tag. Unmarked individu-als > 90 mm TL were injected with a PIT tag, andtag numbers of recaptured fishes were recorded.Tags were injected posterior to the pelvic girdle intothe abdominal body cavity. A length-frequency his-togram was generated to evaluate the size structureof the population and identify possible gaps in year-classes. Length and weight data were comparedbetween rivers and years (2000 versus 2003) using t-tests. Proportional data (distribution frequency)

104

BRYAN AND HYATT

Figure 1. Map of the lower Salt and Verde rivers.


was compared between rivers and years using chi-square analysis. Data were considered to be signifi-cantly different when P < 0.05.

Population Estimate

The joint hypergeometric maximum likelihoodestimator (JHE; Bartmann et al. 1987; White andGarrott 1990; Neal et al. 1993) was used to calculatea population estimate from mark–recapture dataof PIT-tagged fish. The JHE is an adaptation of theLincoln-Petersen estimate for closed populations.We considered fish in both the Salt and Verde riversto be from a single population for estimates ofpopulation size because fish can move freelybetween the two rivers (Bryan and Robinson2000). Furthermore, we treated the population ofroundtail chub as a closed population both geo-graphically and demographically because there wasno (or rare) opportunity for immigration or emi-gration and, during our study, we assumed negligi-ble recruitment into the tagged population. Wealso assumed equal mortality for tagged anduntagged fish (Seber 1973).

Only fish tagged in 2003 were used for popula-tion estimates (i.e., chub captured in 2003 thatwere tagged during the 2000 study were consid-ered initial captures for the 2003 estimate; to meetassumptions of the closed population). The pop-ulation estimate was calculated using NORE-MARK software (G. C. White, Colorado StateUniversity, 1996). Confidence intervals weredetermined with the profile likelihood method(Venzon and Moolgavkar 1988).

Results

During 2003, roundtail chub were collected at 34sites on the lower Verde River and two sites on thelower Salt River (Figure 2). In the lower VerdeRiver, 272 roundtail chub were captured; 242 werePIT tagged, 16 were recaptures, 13 were sacrificedfor aging otoliths, and 6 escaped prior to tagging.In the lower Salt River, 25 chub were captured, ofwhich 20 were PIT tagged and 5 were recaptures.

Roundtail chub were more numerous and moreevenly distributed in the upper 16 km of the VerdeRiver, below Bartlett Dam, than in the lower 25 km

(Figure 3). There was an increase in the proportionof chub collected in the upper 16 km of the riverfrom 64% in 2000 to 88% in 2003 (χ2, P = 0.0001).In the Salt River during 2003, roundtail chub weresparsely distributed with fish being captured atonly two sites located 9.7 and 10.7 km downstreamof Stewart Mountain Dam. During 2000, chubwere collected at six sites in the Salt River rangingfrom 5 to 21 km from the dam.

Length-frequency histograms of roundtail chubcollected during spring and fall 2003 in the lowerSalt and Verde rivers (Figure 4) indicate that thispopulation is comprised almost entirely of largefish (>35 cm), with few small fish (<35 cm) and anapparent absence of juveniles (<25 cm).

Adult roundtail chub (>25 cm) in the lowerSalt River had a higher mean weight (t-test, df =290, P = 0.000) and a mean total length greater(t-test, df = 229, P = 0.000) than those collected inthe lower Verde River (Table 1). Within the lowerVerde River, chub collected in the upper 16 kmwere significantly smaller (mean TL = 414 mm)

105

ROUNDTAIL CHUB POPULATION ASSESSMENT

Figure 2. Locations of roundtail chub captured and PITtagged in the lower Salt and Verde rivers, 2003.


than chub collected in thelower reaches of the river(mean TL = 440 mm; t-test, df= 254, P = 0.001).

Roundtail chub collected in2003 had a mean total lengthgreater than those collected inthe 2000 study for both thelower Salt River (Table 1; t-test,df = 73, P = 0.000) and thelower Verde River (t-test, df =1,144, P = 0.000). The largestroundtail chub collected in2003 was caught in the lowerVerde River and measured 552mm and weighed 1,826 g.

In 2003, 262 roundtail chubfrom the lower Salt and Verderivers were marked; 16 ofwhich were recaptured on sub-sequent trips (and two wererecaptured on multiple occa-sions). The number of individ-uals marked in 2003 was 70%less than the number of fishmarked in the 2000 study(Table 2), although effort dif-fered slightly. The modifiedLincoln-Petersen populationestimate for roundtail chub inthe lower Salt and Verde riversfor 2003 was 1,657 individualswith a 95% confidence intervalof 1,097–2,742, which is a 74%decrease from the 2000 esti-mate of 6,424 individuals, witha 95% confidence interval of5,048–8,397 (Table 2).

Discussion

The lower Salt and Verde riverswere sampled at fixed and ran-dom sites throughout theirentire lengths, but the majorityof roundtail chub were collectedin the upper 16 km of the lowerVerde River. This section of the

Num

ber o

f Rou

ndta

il Ch

ub C

olle

cted

Distance from Bartlett Dam (km)

70

60

50

40

30

20

10

0

70

60

50

40

30

20

10

0

0 5 10 15 20 25 30 35 40

0 5 10 15 20 25 30 35 40

A

B

Figure 3. Distribution and abundance of roundtail chub collected in the lower VerdeRiver, Arizona during 1999–2000 (A; Bryan and Robinson 2000) and in 2003 (B).

Figure 4. Length-frequency histograms of roundtail chub collected during the springand fall, 2003 in the lower Salt River (LSR) and lower Verde River (LVR), Arizona.

106

BRYAN AND HYATT


river is confined within steep canyon walls resultingin a high stream gradient, characterized by numer-ous pool-run complexes adjacent to swift movingriffles, which is the preferred habitat of adultroundtail chub (Bestgen and Propst 1989; Karp andTyus 1990; Rinne and Minckley 1991). There are

also several shoreline eddies available in the upper16 km, which is the preferred spawning habitat forroundtail chub (Vanicek and Kramer 1969; Karpand Tyus 1990). Abundance of roundtail chubdecreased abruptly approximately 16 km down-stream of Bartlett Dam as the river becomes widerand shallower with a more moderate gradient, withfew areas of preferred habitat. In addition to moresuitable habitat, roundtail chub may prefer theupper 16 km because the water is cooler and thereare fewer predators and nonnative species than inthe lower reaches (Bryan and Robinson 2000).

Bryan and Robinson (2000) also found a highabundance of chub in the upper 16 km (64%), but itrepresented a significantly lower proportion of thetotal catch than was found in 2003. The increase inthe proportion of roundtail chub in the upper 16 kmfrom 2000 to 2003 may be due to several factors,including an increase in the abundance of nonnativefishes in the lower reaches of the river (personalobservation), which could have resulted in predationon smaller individuals. In addition, chub in the lowerreaches were significantly larger and, therefore, prob-ably older and may be experiencing a higher rate ofnatural mortality than fish in the upper 16 km.

Collection of roundtail chub from the lower SaltRiver proved to be somewhat difficult. Althoughlow flows during winter (<10 ft3/s) generally forcethe fish to congregate in deep pools (which effec-tively limits their movement), sampling techniquesmay have been ineffective. Bryan and Robinson(2000) snorkeled sites in the lower Salt River justafter electrofishing and approximated that theywere collecting only 10% of the roundtail present ateach site. It was thought that using a combinationof gear types (electrofishing over gill nets) wouldincrease the likelihood of capture. However, theelectrofisher was likely ineffective due to the highconductivity water (~1,255 μS/cm). Sampling dur-ing late spring and summer was not possible due toextremely high flows (typically > 1,000 ft3/s). Othermethods, such as trammel nets, hoop nets, seining,and backpack electrofishing were even less effectivefor collecting adult roundtail chub in the lower SaltRiver (Bryan et al. 2000).

Despite difficulties in sampling, the low numberof roundtail chub collected from the lower SaltRiver is probably a reflection of their low abundance

107


Table 1. Length and weight statistics from roundtail chub collectedin the lower Salt and Verde rivers, 2000 and 2003. Stan-dard errors of means are in parentheses.

Salt River Verde River2000 2003 2000 2003

Length n = 50 n = 25 n = 879 n = 267Mean 363 (13) 474 (5) 367 (2) 417 (3)Min 72 430 81 254Max 502 523 492 552

Weight n = 50 n = 20 n = 582 n = 211Mean 568 (48) 1283 (63) 449 (8) 680 (16)Min 3 930 4 158Max 1,406 1,830 1,206 1,865

Table 2. Number of roundtail chub marked and recaptured duringeach sampling trip in 2000 and 2003 in the lower Salt andVerde rivers, with modified Lincoln-Peterson populationestimates and 95% confidence intervals for each year.

Trip # marked # recaptured

2000 (51 sites/trip)

1 178 –

2 162 3

3 40 0

4 208 18

5 155 12

6 135 20

2000 population estimate (95% CI)6,424 (5,048–8,397)

2003

1 (102 sites) 70 –

2 (102 sites) 16 4

3 (51 sites) 92 4

4 (44 sites) 84 8

2003 population estimate (95% CI)1,657 (1,097–2,742)


due to a lack of suitable habitat (dominated by longruns and riffles) and a high abundance of nonnativefishes (Bryan et al. 2000). The same difficulties donot exist when sampling the lower Verde River;water conductivity is approximately 520 μS/cm,which allows for more effective electrofishing(Reynolds 1996).

Length-frequency distributions, can provideinsight into the dynamics of populations and aid inidentifying problems such as year-class failures orlow recruitment, slow growth, or excessive annualmortality (Anderson and Neumann 1996). Thelength-frequency distributions developed fromroundtail chub collected in 2003 indicate that thepopulation is composed almost entirely of largeadults with little or no recruitment. Althoughlength-frequency histograms from 1999 and 2000(Bryan and Robinson 2000) also show a large adultpopulation, there was evidence of reproduction andrecruitment. Chub that apparently hatched in 1998were collected in spring 1999 as 90–120-mm 1 yearolds, and their growth into the adult population canbe seen in samples during summer 1999 and into2000. The absence of juvenile and small adult chubin 2003 suggests that the last significant spawn mayhave occurred in 1998.

The appearance of juveniles in the 1999 samplefollowed a spring flood event in 1998, which lendscredibility to the hypothesis that, for many nativedesert fish species, including roundtail chub, suc-cessful spawning and subsequent recruitment isrelated to the occurrence of significant floodevents (Nesler et al. 1988; Poff and Allan 1995;Brouder 2001). However, it should be noted thatroundtail chub have sustained a viable populationin the lower Salt and Verde rivers despite longperiods (up to 20 years) without a significantflood event. This suggests that flood events are notsolely responsible for successful spawning andrecruitment, at least not at the level required tosustain a population.

Although initial research (Bryan and Robinson2000) indicated that the roundtail chub popula-tion in the lower Salt and Verde rivers was relative-ly large, we found that the size of this populationhas decreased by 74% in just 3 years. Despite thisreduction in population size, the majority of theindividuals we collected were in very good to

excellent physical condition. Therefore, we do notbelieve the decline is a result of food availability.Although a variety of factors probably have causedthis decline (including an increasing number ofnonnative competitors and predators), we believethat the primary cause is diminished recruitmentdue to a lack of significant spikes in flow in the last5 years (Poff and Allan 1995; Rinne and Stefferud1996; Brouder 2001). The last flood eventoccurred in 1998, which coincides with the lastsignificant roundtail chub spawning event in thelower Salt and Verde rivers. That means that thepopulation consists primarily of individuals atleast 5 years old. Assuming a maximum lifeexpectancy of 7–11 years for roundtail chub(Brouder et al. 2000; Vanicek and Kramer 1969), alarge portion of the population in the lower Saltand Verde rivers may be lost to natural mortalityover the next 5 years.

Conclusions and Recommendations

The distribution and abundance of roundtail chubis declining throughout its historic range. In Ari-zona, statewide declines have been attributed toaquifer pumping, stream diversion, alteration ofthe historic hydrograph, and predation by andcompetition with nonnative fishes (AZGFD 1996).As roundtail chub numbers continue to dwindle,each remaining population becomes increasinglyimportant to the survival of the species. Thedecline of the roundtail chub in the lower Salt andVerde rivers raises serious concern that extirpationof this population may be closer to becoming areality, especially considering the short time spanover which this decline has occurred.

Immediate management action needs to betaken to ensure the maintenance of this (and all)roundtail chub population. To reduce the negativeimpacts of competition and predation, the intro-duction of nonnative fishes into roundtail chubhabitat should be carefully evaluated and probablysuspended, especially with regards to predatoryspecies. In the lower Salt and Verde rivers, rainbowtrout Oncorhynchus mykiss stocking is an impor-tant management tool but should be thoroughlyevaluated to determine its economic impact andthe specific impacts to the chub population.

108

BRYAN AND HYATT


In addition, research should be conducted todetermine the relationship between flood events, orspikes in discharge, and successful spawning andrecruitment. The influence of flood events may berelated to (1) timing of the spike, (2) duration of thespike, and/or (3) magnitude of the spike. Eventhough discharge below Bartlett and Stewart Moun-tain dams are regulated by SRP for hydroelectricpower and irrigation, flood events are still realizedand apparently have an effect on chub reproduc-tion. Arizona Game and Fish Department needs toopen dialog with SRP to explore scenarios that willallow for periodic spike flows to enhance roundtailchub reproduction and recruitment.

We also recommend the development andimplementation of a roundtail chub recovery planbased on the following management needs: (1)watershed and streamflow protection, (2)research to determine the mechanisms of disap-pearance of the species, and (3) actions to reducethe effects of nonnative fishes (AZGFD 1996). Wefurther recommend the immediate developmentof a broodstock program to ensure the survival ofat-risk populations such as the population in thelower Salt and Verde rivers. Finally, it is vital thatwe continue to monitor populations of roundtailchub to learn more about the mechanisms gov-erning reproduction and survival.

References

Anderson, R. O., and R. M. Neumann. 1996. Length, weight,and associated structural indices. Pages 447–481 in B. R.Murphy and D. W. Willis, editors. Fisheries techniques, 2ndedition. American Fisheries Society, Bethesda, Maryland.

AZGFD (Arizona Game and Fish Department). 1996.Wildlife of special concern in Arizona. Arizona Gameand Fish Department, Phoenix.

Bartmann, R. M., G. C. White, L. H. Carpenter, and R. A.Garrott. 1987. Aerial mark-recapture estimates of con-fined mule deer in pinion-juniper woodland. Journalof Wildlife Management 51:41–46.

Bestgen, K. R., and D. L. Propst. 1989. Distribution, statusand notes on the ecology of Gila robusta (Cyprinidae) inthe Gila River drainage, New Mexico. Southwestern Nat-uralist 34:402–412.

Brouder, M. J. 2001. Effects of flooding on recruitment ofroundtail chub, Gila robusta, in a southwestern river.Southwestern Naturalist 46:302–310.

Brouder, M. J., D. D. Rogers, and L. D. Avenetti. 2000. Lifehistory and ecology of the roundtail chub Gila robusta,

from two streams in the Verde River basin. ArizonaGame and Fish Department Research Branch TechnicalGuidance Bulletin No. 3, Phoenix.

Bryan, S. D., and A. T. Robinson. 2000. Population character-istics and movement of roundtail chub in the lower Saltand Verde rivers, Arizona. Final report prepared for theU.S. Bureau of Reclamation. Arizona Game and FishDepartment, Phoenix.

Bryan, S. D., A. T. Robinson, and M. J. Fry. 2000. Native-non-native fish interactions in the lower Salt and Verde rivers.Final report prepared for the U.S. Bureau of Reclamation.Arizona Game and Fish Department, Phoenix.

Clarkson, R. W. 1998. Results of fish monitoring of select-ed waters of the Gila River basin, 1995–1996. Reportprepared for the U.S. Fish and Wildlife Service and Ari-zona Game and Fish Department. U.S. Bureau of Recla-mation, Phoenix, Arizona.

Girmendonk, A. L., and K. L. Young. 1997. Status review ofthe roundtail chub (Gila robusta) in the Verde Riverbasin. Arizona Game and Fish Department Nongameand Endangered Wildlife Program Technical Report No.114, Phoenix.

Karp, C. A., and H. M. Tyus. 1990. Behavioral interactionsbetween young Colorado squawfish and six fish species.Copeia 1990:25–34.

Minckley, W. L., and J. E. Deacon. 1991. Battle against extinc-tion: native fish management in the American West. TheUniversity of Arizona Press, Tucson.

Neal, A. K., G. C. White, R. B. Gill, D. F. Reed, and J. H.Olterman. 1993. Evaluation of mark-resight modelassumptions for estimating mountain sheep numbers.Journal of Wildlife Management 57:436–450.

Nesler, T. P., R. T. Muth, and A. F. Wasowicz. 1988. Evidence ofbaseline flow spikes as spawning cues for Coloradosquawfish in Yampa River, Colorado. Pages 69–79 in R. D.Hoyt, editor. 11th annual larval fish conference, AmericanFisheries Society, Symposium 5, Bethesda, Maryland.

Poff, N. L., and J. D. Allen. 1995. Functional organization ofstream fish assemblages in relation to hydrological vari-ability. Ecology 76:606–627.

Reynolds, J. B. 1996. Electrofishing. Pages 221–253 in B.R.Murphy and D.W. Willis, editors. Fisheries techniques,2nd edition. American Fisheries Society, Bethesda,Maryland.

Rinne, J. N., and W. L. Minckley. 1991. Native fishes of aridlands: a dwindling resource of the desert Southwest. U.S.Forest Service, Rocky Mountain Forest and Range Exper-iment Station, General Technical Report RM-206, FortCollins, Colorado.

Rinne, J. N., and J. A. Stefferud. 1996. Relationships of nativefishes and aquatic macrohabitats in the Verde River, Ari-zona. Hydrology and Water Resources in the Southwest26:13–22.

Seber, G. A. F. 1973. Estimation of animal abundance andrelated parameters. Griffin, London.

Vanicek, C. D., and R. H. Kramer. 1969. Life history of theColorado squawfish, Ptychocheilus lucius, and Coloradochub, Gila robusta, in the Green river in Dinosaur

109



110

BRYAN AND HYATT

National Monument, 1964–1966. Transactions of theAmerican Fisheries Society 98:193–208.

Venzon, D. J., and S. H. Moolgavkar. 1988. A method forcomputing profile-likelihood based confidence inter-vals. Applied Statistics 37:87–94.

Voeltz, J. B. 2002. Roundtail chub (Gila robusta) status

survey of the lower Colorado River basin. ArizonaGame and Fish Department, Nongame and Endan-gered Wildlife Program Technical Report 186,Phoenix.

White, G. C., and R. A. Garrott. 1990. Analysis of wildliferadio-tracking data. Academic Press, New York.


113

Members of the Lewis and Clark expedi-tion were given credit for documentingthe occurrence of nearly 400 species of

plants and animals during their Voyage of Discov-ery from 1804 to 1806. The total included 31species of fish (Burroughs 1995), including up to12 species considered unknown to science prior tothat time (Cutright 1969). Almost all the newspecies were found in the Columbia River system.While there is little doubt of the identity of mostfishes for which Lewis and Clark provided detaileddescriptions, other species designations were con-strued by editors of the journals of the expedition,and later by compilers of natural historical facts.For example, Cutright (1969) listed only eight fishspecies from the Pacific Northwest, with four oth-

ers considered questionable. Burroughs (1995)concluded that there was sufficient informationabout salmon, steelhead, and eulachon to assignidentity, but other identification was less certain.

Unlike other biological specimens (e.g., plantsand birds) collected during the expedition, no fish-es were brought back for study. Methods forpreservation were not well established, and storagespace during transit was limited. Indeed, earlyrecords of fishes from the Pacific Northwest regionwere largely based on reconstruction of the anthro-pological record (e.g., Schalk 1977; Hewes 1998;and others). As a result, the definitive list of west-ern fishes encountered by Lewis and Clark has notbeen resolved.

In this article, I review and summarize fish facts

Columbia River Fishes of the Lewis and Clark Expedition

ABSTRACT The Lewis and Clark expedition crossed the Continental Divide in 1805 on the way west to the

Pacific Ocean. Based on journal entries, members of the expedition probably encountered two species of resident

salmonids and four of the six species of anadromous salmonids and steelhead (Family Salmonidae, genus

Oncorhynchus). The salmonid species were called common salmon (now known as Chinook salmon O.tshawytscha), red charr (sockeye salmon O. nerka), white salmon trout (coho salmon [also known as silver salmon]

O. kisutch), salmon trout (steelhead O. mykiss), and spotted trout (cutthroat trout O. clarkii). There was no evidence

of the expedition encountering pink salmon O. gorbuscha, chum salmon O. keta, or species of true char Salvelinusspp. Common fishes procured from Indian tribes living along the lower Columbia River included eulachon Thale-ichthys pacificus and white sturgeon Acipenser transmontanus. The identity of three additional resident freshwa-

ter species is questionable. Available descriptions suggest that what they called mullet were largescale sucker

Catostomus macrocheilus, and that chubb were peamouth Mylocheilus caurinus. The third questionable fish, which

they called bottlenose, was probably mountain whitefish Prosopium williamsoni, although there is no evidence that

the species was observed in the Columbia River drainage. Missing from the species list were more than 20 other

fishes known to Sahaptin-speaking people from the mid-Columbia region. More complete documentation of the

icthyofauna of the Pacific Northwest region did not occur until the latter half of the 19th century. However, journals

from the Lewis and Clark expedition provide the first documentation of Columbia River fishes.

Dennis Dauble



DENNIS DAUBLE Pacific Northwest National Laboratory, Post Office Box 999, Richland, Washington 99352, USA, [email protected]


from primary or unabridged editions of journalskept by members of the Lewis and Clark expedi-tion (e.g., Moulton 1983; Couse 1985) and naturalhistory accounts by Burroughs (1995) andCutright (1969). To clarify taxonomic relationshipsand nomenclature, I examined historical literatureon fishes from the Columbia River system (e.g.,Suckley 1861; Gilbert and Evermann 1894; Scottand Crossman 1973) as well as accounts of NativeAmerican tribes from the Pacific Northwest region(e.g., Hunn 1980; Walker 1998). My objective wasto describe the contribution of the Lewis and Clarkexpedition to our understanding of native fishfauna from freshwaters west of the ContinentalDivide. The geographic extent studied was fromthe headwaters of Lemhi River (Clearwater-SnakeRiver system) downstream to the mouth of theColumbia River near Astoria, Oregon (Figure 1).

Salmonids

Based on journal entries, Lewis and Clark encoun-tered at least two species of resident salmonids andfour of the six species of anadromous salmonidsduring their travels up and down the ColumbiaRiver system (Moulton 1983). Species includedwhat they termed the common salmon (now knownas Chinook salmon Oncorhynchus tshawytscha), redcharr (sockeye salmon O. nerka), white salmontrout (coho salmon [also known as silver salmon]O. kisutch), salmon trout (steelhead O. mykiss), and

spotted trout (cutthroat trout O. clarkii). Althoughthe type species Salmo mykiss or Kamchatka rain-bow trout was described by Walbaum in 1792(based on Russian forms), no resident or anadro-mous forms of what are now considered within thegenus Oncorhynchus had been previously collectedin the United States. The explorers had only limitedunderstanding of salmonid life cycles, includingjuvenile development. As a result, descriptions werebased on adult forms and limited to characteristicssuch as number of fin rays, size of body scales,mouth shape, body size and shape, and coloration.In this section, I review key passages that confirmspecies designations.

Common Salmon

There is little doubt that the common salmon ofLewis and Clark was the Chinook salmon. Pas-sages like “it is this species that extends itself intoall the river and little creeks on this side of theContinent and to which the natives are so muchindebted for their subsistence” (Moulton 1983)indicate that the explorers recognized the exten-sive geographic range as well as the migrationtiming of this species of salmon. The first journalentry on salmon occurred in August 1805 whilethe expedition traveled along the Lemhi River inMontana. Extensive notes were also made ofIndian harvest methods near the confluence ofthe Columbia and Snake rivers. Nearby is the

Hanford Reach, the principalmain-stem spawning area for fallChinook salmon (Dauble andWatson 1997). Other details fromthe explorer’s journals, includingthe relative size of scales, spottingpattern, and body size, provideimportant clues that clearly distin-guish Chinook salmon from othersalmonids present at that time.

White Salmon Trout

The journals of Lewis and Clarkincluded a detailed sketch of onlyone species of salmon, the “whitesalmon trout” or coho salmon. The

114

DAUBLE

Figure 1. Route of the Lewis and Clark expedition and chronology during their travels across the Continental Divide and through the PacificNorthwest.


likeness was drawn from a freshly gigged specimenobtained near their winter camp at Fort Clatsop(14 March 1806). One passage, “the white specieswhich we found below the falls where in excellentorder when the salmon were entirely out of seasonand not fit for use” (Moulton 1983), indicated thatwhite salmon trout spawned later than the com-mon salmon, an observation consistent with theirlife histories (Scott and Crossman 1973). Anotherobservation related to its appearance in smallcreeks in early March, presumably to spawn, is alsoin accord with life history requirements of cohosalmon. Thus, the assertion that white salmontrout were coho salmon appears correct.

Salmon-Trout

Most naturalists agree that Lewis and Clark’ssalmon-trout was an upstream migrating adultsteelhead (e.g., Burroughs 1995; Cutright 1969). Apassage written late October 1805 while at the GreatFalls of the Columbia (Celilo Falls) is consistentwith an encounter with a early winter-run steel-head: “we met this fish of a Silvery white colour onthe belly and sides and a bluish light brown on theback and head” (Moulton 1983). Clark alsodescribed this species as “narrow in proportion totheir length so much so than the salmon and redcharr” (Moulton 1983). Collectively, the exploreraccounts support that salmon-trout were steelheadand not a species of Pacific salmon.

Red Charr

Red charr were first referred to by Lewis and Clarkin November 1806 near Grays Harbor on the lowerColumbia River: “we purchased of the Indians 19red charr which we found to be excellent fish” and“some of them are almost entirely red on their bel-ley and sides” (Moulton 1983). This color patternis limited to either sockeye or coho salmon. That“none of them are variagated with the dark spots”(Moulton 1983) points to sockeye salmon.

Passages provided no further detail on bodyshape, scale size, jaw, or teeth characteristics ofred charr that would help differentiate themfrom other adult salmon. Additionally, journalentries on migration timing and habitat use did

not always comport with the life history of sock-eye salmon. For example, sockeye salmon usual-ly spawn in areas associated with lakes (Foerster1968). However, there were no accessible lakesystems in the lower Columbia River. Also, sock-eye salmon usually spawn from late July toOctober (Burgner 1991) when the journalsdescribed a salmon near spawning. Despite along list of inconsistencies, I am inclined tobelieve that accounts of red charr could havebeen sockeye salmon.

Mountain Trout

A common resident trout, now known as west-slope cutthroat tout, was readily collected by theexpedition in the Lemhi/Salmon River drainage.This fish was similar to what Lewis and Clarkknew as “mountain” or “speckled trout” and wasreadily identified by its spotting pattern, color,and presence of vomer teeth. A second species oftrout was alluded to in late August 1805: “I nowfor the first time saw 10 or a douzen of a whte spe-ceis of trout. They are of a silvery colour except onthe back and head, where they are a bluish cast”(Moulton 1983). These fishes were likely a resi-dent form of rainbow trout (nonanadromous O.mykiss). Another account described a fish trap ona small creek that Indians used to catch “the smallfish which pass down with the stream” (Moulton1983). The second observation may have been inreference to downstream migrant steelhead.

Other Salmonidae

The journals of Lewis and Clark, as well as the notesof other members of their party, provide no recordof either pink salmon O. gorbuscha or chum salmonO. keta. Pink and chum salmon are common tosmaller river systems along the coast of Washingtonand Oregon without substantial runs in the lowerColumbia River and its tributaries (Heard 1991;Salo 1991). Both species have unique coloration andphysical features at maturity but were evidently notpresent in significant enough numbers during thewinter of 1805–1806 to be noted.

No true char, including bull trout Salvelinusconfluentus or Dolly Varden Salvelinus malma,

115

COLUMBIA RIVER FISHES OF THE LEWIS AND CLARK EXPEDITION


were documented in the Columbia River systemduring the expedition. This is somewhat surpris-ing considering that these closely related fisheswere likely present in principal tributaries to theColumbia and Snake rivers. These char shouldhave been readily identified as different fromother trout and salmon based on their distinctspotting pattern.

Eulachon and Sturgeon

Lewis sketched a detailed life-sized likeness of theeulachon (also known as candlefish) Thaleichthyspacificus while at Fort Clatsop in February 1806(Moulton 1983). He also provided sufficient meas-urements of various body parts and counts of finrays to confirm its identity. Members of the expedi-tion were most enamored with this small fish findingit “more delicate and lussious than the white fish ofthe lakes” (Moulton 1983). Eulachon are anadro-mous smelt that occur only along the northernPacific Coast of North America (Scott and Crossman1973), migrating about 100 km up the lower Colum-bia River and its tributaries to spawn. It was highlyprized by Native American tribes because it migrat-ed earlier in the spring than salmon, was highlyabundant, had a high oil content, and was easily cap-tured using long-handled dip nets.

Lewis and Clark procured sturgeon from Indianson more than one occasion during their travel in thelower Columbia River. Journal passages referring tosturgeon were largely devoted to how local Indiantribes prepared them for eating. These accountsmost likely were of white sturgeon Acipenser trans-montanus, the most common sturgeon in theColumbia River (Scott and Crossman 1973).

Questionable Fishes

Lewis and Clark provided several accounts of fish-es in the Columbia River that were confusing tohistorians. Only three fishes had sufficient detail toassign an identity, as described below.

Mullet

The first West Coast entry on mullets was from 16April 1806, “At the rapids (near The Dalles thenatives subsist chiefly on a few white-salmon trout

which they take at this time and considerablequantities of a small indifferent mullet of an inte-rior quality.” References were also made to mulletsin late April while Lewis and Clark where in thecompany of the Wallah Wallah tribe. A detaileddescription of a fishweir and seining methods thelocal Indian tribe used to collect mullets providesstrong evidence that what they called mullets werelargescale sucker Catostomus macrocheilus, congre-gating in the lower Walla Walla River during theirannual spawning migration. Largescale sucker areone of the most abundant fishes in the ColumbiaRiver (Dauble 1986) and were highly valued by theSahaptin-speaking people of the mid-Columbiaregion (Hunn 1980) because they arrived early inthe spring prior to migration of Chinook salmon.

There is one specific reference to sucker by Clarkon 8 May 1806: “with those nets (Skooping net)they take the Suckers,…” referring to the Shoshonetribe in the upper Clearwater drainage. That mulletwere synonymous with suckers is reinforced byJohn Ordway’s description of fishing methods atthe mouth of the Walla Walla River (Moulton1983). Northwest fur trappers of the 18th and 19thcenturies commonly used the term mullet todescribe several species of suckers, including thoseendemic to the eastern United States (Scott andCrossman 1973). Interestingly, both Couse (1985)and Cutright (1969) considered Columbia Rivermullet to be northern pikeminnow Ptychocheilusoregonensis. Their deduction does not bear out.Mullet from this region had to be suckers.

Chubb

A controversial species account, as described byLewis on 26 April 1806, near Columbia River km490, noted how a small Indian boy “caught severalchubbs with a bone (in this form) which he substi-tuted for a hook.” Chub is a common name for sev-eral members of the minnow family, Cyprinidae,including taxa that may have been familiar tomembers of the expedition. Thus, the “chubb” ofLewis and Clark was most likely from the familyCyprinidae. Size characteristics limit the fish to oneof three common species.

A key description from Lewis’s entry “theupper exceeded the under jaw. the latter is trun-cate at the extremity” suggested that the fishes

116

DAUBLE


were either peamouth Mylocheilus caurinus orchiselmouth Acrocheilus alutaceus rather thannorthern pikeminnow, which have a terminalmouth (Table 1). The Oxford Dictionary definestruncate as “ending abruptly as if cut off from thebase or tip.” Peamouth have a small, pointed, anda somewhat oblique mouth. The name chisel-mouth relates to a dense cartilaginous plate ontheir lower jaw that resembles the blade of a car-penter’s chisel. Chiselmouth also have an inferiormouth, more snubbed off than a peamouth, andcloser to truncate. Because chiselmouth feedmainly by scraping algae from rocks and becausepeamouth are omnivorous (Gray and Dauble2001) and more commonly caught by hook-and-line, weight of evidence brings us to what mosthistorians agree to: the “chubb” of Lewis andClark were most likely peamouth.

Bottlenose

Two journal entries referred to a fish known as bot-tlenose. The first entry was made while the partytraveled upstream in the Missouri River drainageeast of the Continental Divide: “The fish of this partof the river are trout and a species of scale fish of awhile [white] colour and a remarkable small longmouth which one of our men inform us are the

same with the species called in the Eastern statesbottlenose” (Moulton 1983). This description mostclosely comports with mountain whitefish Prosopi-um williamsoni, a species not formally describeduntil 1856 by Charles Girard. Evermann and Cox(1896) and Brown (1971) also alluded that bot-tlenose may have been mountain whitefish.

There were no site-specific observations of thebottlenose from the west side of the ContinentalDivide, rather a general reference made by Clarkwhile reconstructing events from the expedition’swinter camp at Fort Clatsop: “…besides the fishof this coast and river already mentioned we havemet with the following Species. viz. … a SpeceisSimilar to one of those noticed on the Missouriwithin the mountain, called in the Eastern states,bottlenose” (Moulton 1983). Thus, whether bot-tlenose were found on the west side of the Conti-nental Divide remains in dispute.

Unknown Fishes

Other accounts of fishes from the Columbia Riverwere fleeting at best. For example, Clark wrote on 25October 1805, while near The Dalles, that “one ofthe guards saw a Drumfish today.” Was this mysteryfish really a freshwater drum Aplodinotus grunniens,

117


Table 1. Taxonomic characteristics of the “chubb” of Lewis and Clark (from Moulton 1983) in comparison with three common Cyprinidaefrom the Columbia River (after Scott and Crossman 1973).

Three common cyprinids

Characteristic Chubb Pikeminnow Chiselmouth Peamouth

Dorsal fin rays 10 9–10 10 8

Pelvic fin rays 10 9 9–10 9–10

Pectoral fin rays 9 15–16 15–18 15–18

Anal fin rays 12 8–9 9–10 8–9

Mouth position “upper exceeded the lower jaw” terminal inferior slightly inferior

Lower jaw “truncate at the extremity” large cartilaginous sheath premaxillary protacile

Caudal peduncle small where tail joined the body” thick very narrow narrow

Pelvic fin origin “equidistant from head” under or in advance of slightly in advance of slightly posterior to(relative to dorsal fin)

Head length small 22–23% of TLa 17–19% of TL 17–18% of TL

Body depth large 15–18% of TL 17–20% of TL 15–18% of TL

a TL = total length.


fish restricted to eastern United States at that time?Drumfish or freshwater drum do make a peculiarnoise, called grunting, croaking, or drumming,which is supposedly caused by forcing air from theswim bladder into one of the lateral horns (Jordanand Gilbert 1882). However, their distinct shape andlong, double dorsal fin with spines should not haveconfused them with any fish present in the lowerColumbia River at that time.

Discussion

Collectively, edited accounts of journals from theLewis and Clark expedition substantiate that fiveOncorhynchus spp. were encountered, includingthree species of Pacific salmon, steelhead/rainbowtrout, and cutthroat trout, during their travelsalong the Columbia and Snake river system. How-ever, the explorers provided too few details on lifehistory and habitat requirements to corroborateeach observation. Uncertainty in species identifica-tion because of variation in color and size is notsurprising because the taxonomy of salmonids wasin considerable flux during most of the 19th centu-ry. For example, Suckley (1861) described 43species of North American salmon and trout, com-pared to the current recognized list of 15 species(Behnke 2002). Thus, it is conceivable that differ-ent members of the family Salmonidae were oftenconfused with variants of each other.

Lewis and Clark provided sufficient detail toconfirm the existence of eulachon and white stur-geon in the lower Columbia River. During thereturn leg of the expedition in spring 1806,peamouth and largescale sucker were encounteredin the mid-Columbia region. There is no evidencethat bottlenose (likely mountain whitefish, notmountain sucker C. platyrhynchus as some histori-ans have alluded to) were collected in the Colum-bia River system by members of the Lewis andClark expedition.

One source of confusion to Lewis and Clark wasthe proper zoological classification of certainmarine mammals: “The Porpus is common on thiscoast and as far up the river as the water is brack-ish. The Indians Sometimes gig them and alwayseat the flesh of this fish when they Can precure it.”That Clark confused the taxonomic relationship

between fishes and finned mammals implies abasic lack of understanding of the Linaaen classifi-cation system in which he had limited training.

Assuredly, many other fish species inhabitedthe Columbia and Snake River systems during theearly 19th century. This fact can be deduced basedon the list of fishes known to Sahaptin-speakingpeople in the mid-Columbia region, a classifica-tion scheme including 20 kinds of fish that corre-sponded to about 30 of the ichthyologist (Hunn1980). Lewis himself noted,“I have no doubt thereare many other speceis of fish, which also exist inthis quarter of different seasons of the year, whichwe have not had the opportunity of seeing”(Moulton 1983).

Based on knowledge of fishes native to theColumbia River system, ethnobiological notes, andobservations of later explorers to this region, addi-tional freshwater species of fish present in the early19th century would have included mountain white-fish; redside shiner Richardsonius balteatus; variousdace Rhinichthys spp., northern pikeminnow, chis-elmouth, various sculpin Cottus spp., sand rollerPercopsis transmontana, Pacific lamprey Lampetratridentata, brook lamprey Lampetra spp; and riverlamprey L. ayresii, bridgelip sucker Catostomuscolumbianus, mountain sucker C. platyrhynchus,and three-spine stickleback Gasterosteus aculeatus.

The freshwater fish community in the mid-Columbia region now includes 44 species of fish(Gray and Dauble 1977). More recently, Wydoskiand Whitney (2003) listed approximately 60 fishesfor the Columbia River system. Thus, the specieslist has doubled in the last 200 years due to humanintervention. Additional shifts in composition andrelative abundance of the fish community can beexpected due to changes in water use, energy devel-opment activities, climate change, and fisheriesmanagement practices.

In summary, Lewis and Clark can be forgivenfor not providing a more complete list of fishesfrom the Columbia River system. Their greatestaccomplishment remains that their travels sparkedthe imagination of the greater American publictowards biota found west of the ContinentalDivide. The era of specializing within the broadfield of biology was only emerging. Not until thelatter half of the 19th century did fisheries scien-

118

DAUBLE


tists such as Girard (1856), Suckley (1861), Jordanand Gilbert (1882), and Gilbert and Evermann(1894) more thoroughly document ichthyofauna ofthe Pacific Northwest.

Acknowledgments

I thank David Geist for helpful review commentsand Georganne O’Connor for editorial assistance.This research was supported in part by the U.S.Department of Energy under contract DE-AC05–76RL01830.

ReferencesBehnke, R. J. 1992. Native trout of western North America.

American Fisheries Society, Monograph 6, Bethesda,Maryland.

Behnke, R. J. 2002. Trout and salmon of North America. TheFree Press, New York.

Brown, C. J. D. 1971. Fishes of Montana. Big Sky Books,Bozeman, Montana.

Burgner, R. L. 1991. Life history of sockeye salmon(Oncorhynchus nerka). Pages 1–117 in C. Groot and L.Marcolis, editors. Pacific salmon life histories. UBCPress, Vancouver.

Burroughs, R. D. 1995. The natural history of the Lewis andClark expedition. Michigan State University, East Lansing.

Couse, E., editor. 1985. History of the expedition under thecommand of Lewis and Clark. Volumes II and III. DoverPublications, Inc., New York.

Cutright, P. 1969. Lewis and Clark: pioneering naturalists.University of Illinois Press, Chicago.

Dauble, D. D. 1986. Life history and ecology of the largescalesucker (Catostomus macrocheilus) in the Columbia River.American Midland Naturalist 116:356–367.

Dauble, D. D., and D. G. Watson. 1997. Status of fall Chinooksalmon populations in the mid-Columbia River. NorthAmerican Journal of Fisheries Management 17:283–300.

Evermann, B. W., and U. O. Cox. 1896. Report upon the fishesof the Missouri River basin. Report of the U.S. Commis-sioner of Fish and Fisheries for 1894, Document number424. Government Printing Office, Washington, D.C.

Foerster, R. E. 1968. The sockeye salmon Oncorhynchusnerka. Fisheries Research Board of Canada Bulletin162:422.

Gilbert, C. H., and B. W. Evermann. 1894. A report uponinvestigations in the Columbia River basin, with descrip-tions of four new species of fishes. Bulletin of the UnitedStates Fish Commission 1:169–207.

Girard, C. 1856. Notice upon the species of the genus Salmo,of authors, observed chiefly in Oregon and California.Proceedings of the Academy of Natural Sciences8:165–168.

Gray, R. H., and D. D. Dauble. 1977. Checklist and relativeabundance of fish species form the Hanford reach of theColumbia River. Northwest Science 51:208–215.

Gray, R. H., and D. D. Dauble. 2001. Some life history char-acteristics of cyprinids in the Hanford Reach, mid-Columbia River. Northwest Science 75:122–136.

Heard, W. R. 1991. Life history of pink salmon(Oncorhynchus gorbuscha). Pages 120–230 in C. Grootand L. Marcolis, editors. Pacific salmon life histories.UBC Press, Vancouver.

Hewes, G. W. 1998. Plateau. Pages 620–640 in D. E Walker,Jr., editor. Handbook of North American Indians. Vol-ume 12. Smithsonian Institution, Washington, D.C.

Hunn, E. S. 1980. Sahaptin fish classification. NorthwestAnthropological Research Notes 14(1):1–19.

Jordan, D. S., and C. H. Gilbert. 1882. Synopsis of the fishesof North America. Government Printing Office, Wash-ington, D.C.

Moulton, G. E., editor. 1983. The Journals of Lewis & ClarkExpedition. Volumes 5–7, 9, and 11. University ofNebraska Press, Lincoln and London.

Schalk, R. F. 1977. The structure of an anadromous fishresource. Pages 207–249 in L. R. Binford, editor. For the-ory building in archaeology. Academic Press, New York.

Salo, E. O. 1991. Life history of chum salmon (Oncorhynchusketa). Pages 232–309 in C. Groot and L. Marcolis, editors.Pacific salmon life histories. UBC Press, Vancouver.

Scott, W. B., and E. J. Crossman. 1973. Freshwater fishes ofCanada. Fisheries Research Board of Canada. Bulletin184, Ottawa.

Suckley, G. 1861. Report upon the fishes collected on the sur-vey. Chapter I. Report upon the Salmonidae. Pages307–339 in Explorations and surveys for a railroad routefrom the Mississippi River to the Pacific Ocean, 1853–55.War Department. Volume XII, Book I. Thomas H. FordPrinter, Washington, D.C.

Walker, D. E., Jr., editor. 1998. Plateau. Volume 12 in W. C.Sturtevant, general editor. Handbook of North AmericanIndians. Smithsonian Institution, Washington, D.C.

Wydoski, R. S., and R. R. Whitney. 2003. Inland fishes ofWashington. University of Washington Press, Seattle.

119



121

Bluehead sucker Catostomus discobolus, flan-nelmouth sucker C. latipinnis, and roundtailchub Gila robusta have declined in Wyoming

and throughout their native ranges. A recent statusreview estimated that flannelmouth sucker androundtail chub each occupy 45% and blueheadsucker occupy 50% of their historic range in theColorado River basin (Bezzerides and Bestgen2002). The authors indicate that reservoir construc-tion and operation and nonnative fish speciesintroductions are likely causes of these declines.Due to these declines, a conservation agreement is

being developed by the states of Arizona, Colorado,Nevada, New Mexico, Utah, and Wyoming toensure the persistence of these three speciesthroughout their native range.

All three are classified by the Wyoming Gameand Fish Department (WGFD) as NSS1 species,indicating that they are rare and their habitat isdeclining or vulnerable. Surveys conducted by Bax-ter (Baxter and Simon 1970) and Wheeler (1997)represent the only drainage-wide surveys conduct-ed in the Green River drainage of Wyoming. Themost recent survey conducted by Wheeler (1997)

Current Distribution of Bluehead Sucker,Flannelmouth Sucker, and Roundtail Chub in

Seven Subdrainages of the Green River, Wyoming

ABSTRACT Little is known about the distribution and status of native fishes in the Green River drainage ofsouthwestern Wyoming, particularly bluehead sucker Catostomus discobolus, flannelmouth sucker C. latipinnis,and roundtail chub Gila robusta. These species face a number of threats to their continued survival in Wyoming,including predation by nonnative fishes and competition and hybridization with nonnative fishes, habitat fragmen-tation (caused by water development and stream dewatering), and habitat destruction (caused by livestock graz-ing and road construction). Due to these threats, this project was undertaken to gather information to guide futuremanagement of these species. Objectives included determining current distribution and abundance of the threespecies, determining the extent of hybridization with nonnative species, and documenting overall species com-position throughout the study area. Sampling began in 2003, and efforts were focused on the eastern portion ofthe Green River watershed between Fontenelle and Flaming Gorge reservoirs. Twenty-three fish species wereidentified in 60 sites; only seven were native. Flannelmouth sucker were widely distributed throughout the studyarea. Bluehead sucker were less abundant and were limited to the main-stem Green River and the upper BigSandy River and upper Little Sandy Creek. Perhaps the biggest threat to these native suckers is hybridization withwidely distributed introduced white sucker Catostomus commersonii. Roundtail chub were not collected during2003 surveys, but still exist in other portions of the drainage in Wyoming. This is an ongoing project, and sam-pling will continue in remaining portions of the drainage in subsequent years.

Curtis J. Gill, Kevin R. Gelwicks, and Robert M. Keith



CURTIS J. GILL Wyoming Game and Fish Department, 351 Astle Avenue, Green River, Wyoming 82935, USA

KEVIN R. GELWICKS Wyoming Game and Fish Department, 528 South Adams Street, Laramie, Wyoming 82070, USA; Corresponding author:[email protected]

ROBERT M. KEITH Wyoming Game and Fish Department, 351 Astle Avenue, Green River, Wyoming 82935, USA


showed that, between 1965 and 1995, all threespecies had declined at three spatial scales (site,stream, and subdrainage) in the Green Riverdrainage of Wyoming. He only found blueheadsucker in the Hams Fork drainage in 1995, andfound roundtail chub in the Blacks Fork, HamsFork, and Little Snake River drainages. Flannel-mouth sucker were documented throughout theGreen River drainage downstream of FontenelleReservoir, including the Little Snake River drainage(Wheeler 1997).

Additional subdrainage surveys have been con-ducted in the Bitter Creek drainage (Carter andHubert 1995) and in the Big Sandy River drainage(Miller 1978). Bluehead sucker and roundtail chubwere not collected in the 1993 Bitter Creek survey,but flannelmouth sucker were collected from a sitenear the Bitter Creek–Green River confluence(Carter and Hubert 1995). Miller (1978) reportedon a survey of the entire Big Sandy River drainagein the 1960s and 1970s. They found roundtail chuband bluehead sucker at one site each in the BigSandy River upstream of Big Sandy Reservoir. Theyalso collected bluehead sucker at one site on LittleSandy Creek. Flannelmouth sucker were found tooccur throughout the Big Sandy River and LittleSandy Creek below the national forest boundary.Locations of various subdrainagescan be found in Figure 1.

Bluehead sucker, flannelmouthsucker, and roundtail chub face anumber of threats to their contin-ued survival in Wyoming. Intro-duced fish species are probably theforemost threat to these threespecies, particularly the threat ofhybridization between nonnativewhite sucker and native blueheadand flannelmouth suckers. Nonna-tive fishes also threaten native fish-es through competition and preda-tion. Additional threats includehabitat fragmentation (caused bywater development and streamdewatering) and habitat destruc-tion (caused by livestock grazingand road construction). Because ofthese threats, additional informa-

tion is needed to guide future management deci-sions for these species.

Previous surveys have documented the historicdistribution of these species, but new informationis necessary to truly assess the status and distribu-tion of these species in Wyoming. Drainage-widesurveys began in the summer of 2003 to documentthe distribution and abundance of bluehead suck-er, flannelmouth sucker, and roundtail chub (targetspecies) in the Green River drainage of Wyoming,document the nonnative and native species com-position, and assess the degree of introgressivehybridization between the three species and intro-duced fish species. This paper presents the resultsof the first season of surveys. Approximately fourfield seasons will be necessary to map distributionsof the target species throughout the study area,with completion of fieldwork targeted for fall of2006. Once drainage-wide surveys are complete,this information will be used to focus furtherresearch and management of these three species.

Study Area

The Colorado River basin (CRB) covers approxi-mately 632,000 km2 of the southwestern UnitedStates. The Green River is the longest tributary to

122

GILL ET AL.

Figure 1. Locations of subdrainages and reaches sampled in 2003 in theGreen River drainage, Wyoming.

Sampling LocationsStreams

Green River Watershed

2003 Sample Area

Big Sandy sub-drainage

Bitter Creek sub-drainage

Eastside Tributaries sub-drainage

Fontenelle Creek sub-drainage

Henrys Fork sub-drainage

Middle Green River sub-drainage

Slate Creek sub-drainage


the Colorado River and covers approximately 18%of the CRB. The Green River begins in the WindRiver range of Wyoming and flows south to itsconfluence with the Colorado River in southeast-ern Utah. This project focuses on the portion ofthe Green River watershed located in southwesternWyoming. This watershed drains an area ofapproximately 44,500 km2 in Wyoming or 7% ofthe CRB.

Land ownership in the watershed is 72% publicand 28% private. The Bureau of Land Managementmanages the majority (75%) of the public land,followed by the U.S. Forest Service, the State ofWyoming, the Bureau of Reclamation, and the U.S.Fish and Wildlife Service at 15.8%, 5.5%, 2.9%, and0.8%, respectively.

Surveys in 2003 focused primarily on the east-ern portion of the middle Green River watershed:the Green River between Fontenelle and FlamingGorge reservoirs, the Big Sandy River subdrainage,the Bitter Creek subdrainage, and the eastside trib-utaries to Flaming Gorge Reservoir (Figure 1).Owing to extreme drought, a number of streams orstream reaches targeted for sampling in 2003 wentdry during the summer. For this reason, three addi-tional subdrainages in the western portion of themiddle Green River drainage were sampled. Theseincluded the Slate Creek, Fontenelle Creek, andHenrys Fork subdrainages (Figure 1).

Methods

Sampling reaches were systematically chosenthroughout the study area. Reaches were spaced at8–16-km intervals on most streams, shorter inter-vals on smaller streams, and longer intervals onlarger streams. In some instances, it was not possi-ble to adhere to these criterion due to access prob-lems. Upstream extent of sampling was mainly dic-tated by upstream extent of wetted stream.However, in the Big Sandy River, Little Sandy Creek,and Henrys Fork, the upstream extent of samplingwas dictated by where the community becamedominated by trout and mottled sculpin Cottusbairdii. This upstream extent was at elevation 2,275m on the Big Sandy River, 2,424 m on Little SandyCreek, and 2,543 m on the Henrys Fork.

Sampling reaches were 200 m long in perennial

portions of streams. Prior to electrofishing, blocknets were placed at the upstream and downstreamboundaries of the sampling reach to prevent fishfrom moving in or out. At times, natural barriers,such as beaver dams, were used as the upstreamboundary of the sampling reach.

Due to the extreme high and low conductivitiesencountered in most streams, the majority ofreaches were sampled with shore-based electrofish-ing equipment consisting of a Coffelt VVP-15,5,000-W generator and either one or two elec-trodes, depending on stream size. A single anodewas used in smaller streams (<4.6 m wide); twoanodes were used in larger stream reaches (>4.6 mwide). When conditions allowed, backpack elec-trofishing equipment was used (e.g., FontenelleCreek). On most occasions, a single electrofishingpass was completed over the entire sampling reachto assess the fish community. If electrofishing effec-tiveness was poor, an additional electrofishing passor a seine haul may have been completed over all ora portion of the reach.

When residual pools were the only habitat typein a section of stream, a bag seine was used to sam-ple the reach. Generally, a seine was hauled throughthe entire residual pool. Depending on the numberof fish collected on the first seine haul, additionalseine hauls may have been completed.

All fish collected were sorted by species andcounted. Individual lengths and weights wererecorded for all target species (although no round-tail chubs were captured in this survey), whitesucker, and hybrids. When a large number of whitesucker were collected, lengths and weights wererecorded for a minimum of 30 individuals, and acount and batch weight was recorded for theremaining white suckers. Counts, length ranges,and batch weights were recorded for all nontargetspecies collected.

Tissue samples were collected from all blueheadsucker and flannelmouth sucker and some hybrids.The samples will be used to evaluate genetic purityand to verify phenotypic identification. A thumb-nail-sized piece of left pelvic fin was clipped andpreserved in a microcentrifuge tube filled with 95%ethanol. In addition, a close-up lateral photo of theentire left side of each fish was taken with a digitalcamera. If fish were too small to provide a sufficient

123

CURRENT DISTRIBUTION OF BLUEHEAD SUCKER, FLANNELMOUTH SUCKER, AND ROUNDTAIL CHUB


sample from a fin, the entire fish was sacrificed andpreserved in a microcentrifuge tube. In reacheswhere large numbers of bluehead sucker, flannel-mouth sucker, and hybrids were collected, a mini-mum of 20 tissue samples were collected per species.

Results

A total of 60 reaches were sampled across the sevensubdrainages in 2003 (Figure 1). Twenty-three fish

species were identified, of which only seven specieswere native to the Green River watershed: blueheadsucker, flannelmouth sucker, Colorado River cut-throat trout Oncorhynchus clarkii, mottled sculpin,mountain sucker Catostomus platyrhynchus,mountain whitefish Prosopium williamsoni, andspeckled dace Rhinichthys osculus (Table 1). Of thetarget species, only bluehead sucker, flannelmouthsucker, and their hybrids with introduced whitesucker were collected in 2003. Bluehead sucker

124

GILL ET AL.

Table 1. Presence (X) of fishes collected in 2003 from seven subdrainages of the Green River Drainage, Wyoming. Refer to Figure 1 forlocations of subdrainages.

Subdrainage a

Common name Scientific name BSR BC b EST FC HF MGR SC

Native species

bluehead sucker Catostomus discobolus X X X

Colorado River cutthroat trout Oncorhynchus clarkii pleuriticus X

flannelmouth sucker C. latipinnis X X X X X

mottled sculpin Cottus bairdi X X X X X

mountain sucker Catostomus platyrhynchus X X X X X X

mountain whitefish Prosopium williamsoni X X X

speckled dace Rhinichthys osculus X X X X X X

Introduced species

burbot Lota lota X

brook trout Salvelinus fontinalis X X X

brown trout Salmo trutta X X X

Bonneville cutthroat trout Oncorhynchus clarkii utah X

common carp Cyprinus carpio X X

fathead minnow Pimephales promelas X X X X X X

kokanee salmon Oncorhynchus nerka X

lake chub Couesius plumbeus X X

longnose dace Rhinichthys cataractae X X

leatherside chub Gila copei X

rainbow trout Oncorhynchus mykiss X X X X

redside shiner Richardsonius balteatus X X X X X X

smallmouth bass Micropterus dolomieu X

Snake River cutthroat trout Oncorhynchus clarkii spp. X X

Utah chub Gila atraria X X X

white sucker Catostomus commersonii X X X X X X X

Hybrids

bluehead x white sucker X

flannelmouth x white sucker X X X X X

a BSR = Big Sandy River, BC = Bitter Creek, EST = Eastside tributaries to Flaming Gorge Reservoir, FC = Fontenelle Creek,HF = Henrys Fork, MGR = Middle Green River, SC = Slate Creek.

b Fish assemblage upstream of Point of Rocks, Wyoming consists of only flannelmouth sucker, mountain sucker, and speckled dace.


were noted in 7 of the 60 reaches (Figure 2); flan-nelmouth sucker were noted in 16 of the 60 reach-es (Figure 3). Hybridization between blueheadsucker and white sucker was identified in 3 of the60 reaches (Figure 2); hybridization between flan-nelmouth sucker and white sucker was identifiedin 11 of the 60 reaches (Figure 3).No roundtail chub were collectedin the seven subdrainages sampled.

Bluehead suckers were collectedin the upper Big Sandy River (aboveBig Sandy Reservoir), upper LittleSandy Creek, and in the main-stemGreen River between FlamingGorge and Fontenelle reservoirs(Figure 2). Bluehead suckers wereextremely rare, especially in themain-stem Green River. Flannel-mouth suckers were relatively com-mon in the main-stem Green Riverand were also collected in the BitterCreek, Big Sandy, and Henrys Forksubdrainages (Figure 3).

Nonnative white sucker werewidely distributed throughout thestudy area (Figure 4), and pheno-typic evidence of hybridizationwith native suckers was widespread(Figures 2 and 3). In the GreenRiver drainage of Wyoming to date,only one subpopulation of blue-head sucker has been identified thatdoes not coexist with nonnativewhite sucker. This sub-populationis located in Ringdahl Reservoir inthe Henrys Fork drainage (WGFDfile information). In addition, 2003surveys only identified one subpop-ulation of flannelmouth sucker thatdoes not coexist with white sucker.This population is in upper BitterCreek, upstream of Point of Rocks,Wyoming. This portion of BitterCreek contains an entirely nativespecies assemblage consisting offlannelmouth sucker, mountainsucker, and speckled dace.

During 2003 surveys, tissue

samples and matching photos were collected from257 native suckers and hybrids (167 flannelmouthsucker, 59 bluehead sucker, 21 flannelmouth x

white sucker hybrids, and 10 bluehead x whitesucker hybrids). These samples and others collect-ed in 2002 will be used to determine the validity of

125


Green River watershed

Bluehead sucker presence

BHS

BHS, BXW

No BHS or BXW present

Streams


Flannelmouth sucker presence

FMS

FMS, FXW

FXW

No FMS or FXW present

Streams

Figure 2. Locations of reaches where bluehead sucker (BHS) or hybrids (BXW)were identified in 2003 in the Green River drainage, Wyoming.

Figure 3. Locations of reaches where flannelmouth sucker (FMS) or hybrids(FXW) were identified in 2003 in the Green River drainage, Wyoming.


phenotypic identification of bluehead sucker, flan-nelmouth sucker, white sucker, and hybrids. Theywill also be used to determine the purity of sub-populations of flannelmouth and bluehead suck-ers, such as those found in Bitter Creek and Ring-dahl Reservoir.

Discussion

Surveys conducted in 2003 documented the distri-bution and abundance of bluehead sucker, flannel-mouth sucker, and roundtail chub in sevensubdrainages of the Green River watershed ofWyoming. Bluehead sucker and flannelmouthsucker were the only target species identified in2003. Bluehead sucker were documented in sevenstream reaches (two subdrainages) and flannel-mouth sucker were documented in 16 streamreaches (five subdrainages).

The collection of bluehead sucker in the BigSandy and Middle Green River subdrainages wasencouraging. Surveys conducted in 1965 and 1995(Wheeler 1997) did not sample the middle reachesof the Green River as we did. The only Green Riversites sampled in these surveys were in the extremeupper and lower reaches of the river. However,

these previous surveys did samplethe Big Sandy River near the GreenRiver confluence, and blueheadsucker were collected here in 1965but not in 1995 (Wheeler 1997).Bluehead sucker are still presentbut rare in the middle Green River.

Bluehead sucker were alsofound in the Big Sandy Riverupstream of Big Sandy Reservoirand in Little Sandy Creek upstreamof the Eden Diversion. Previoussurveys by Baxter and Wheeler didnot sample within these streamreaches (Wheeler 1997). Miller(1978) sampled within these reach-es and noted bluehead sucker inBig Sandy Reservoir, in the upperBig Sandy River above the reser-voir, and in upper Little SandyCreek upstream of the Eden Diver-sion. This survey also noted round-

tail chub in the Big Sandy River upstream of BigSandy Reservoir (Miller 1978). Bluehead sucker arestill present but rare in the upper Big Sandy Riverand upper Little Sandy Creek.

In this survey, roundtail chub were not collect-ed in the Big Sandy River upstream of Big SandyReservoir. However, only five 200-m reaches weresampled in approximately 90 km of available habi-tat in the Big Sandy River, and sampling was notconducted in the 1,012-ha Big Sandy Reservoir.Miller (1978) reports two roundtail chub collectedin a 3.2-km reach of the Big Sandy River sampledin 1972 (which overlaps one 200-m reach sampledin 2003) but did not find any roundtail chub in BigSandy Reservoir in the 1960s or 1970s during rou-tine sampling. If roundtail chub persist in the BigSandy River upstream of Big Sandy Reservoir, theyare extremely rare.

Another notable finding in 2003 was the exis-tence of adult flannelmouth sucker in the upperBitter Creek drainage. These represent the onlypopulation of flannelmouth sucker that is isolatedfrom introduced white sucker. Carter and Hubert(1995) sampled in this area in 1993 but only iden-tified mountain sucker and speckled dace. The Bit-ter Creek fish community upstream of Point of

126

GILL ET AL.


White sucker presence

No WHS present

WHS

Streams

Figure 4. Locations of reaches where white sucker (WHS) were identified in2003 in the Green River drainage, Wyoming.


Rocks is composed only of native species (flannel-mouth sucker, mountain sucker, and speckleddace) and is apparently isolated by a physical barri-er or dewatering. The fish assemblage in thisstream reach is unique and should receive highconservation priority.

The native and nonnative species compositionwas noted for all stream reaches sampled. In total,16 nonnative species and 7 native species were iden-tified in the 60 reaches sampled. It was extremelyrare to find a stream portion that did not containnonnative species. In many reaches, nonnative fishspecies were more prevalent than native species.However, there were instances where only nativespecies were collected (e.g., upper Bitter Creek [asnoted above] and the uppermost reach on the Hen-rys Fork [containing Colorado River cutthroattrout, mottled sculpin and mountain sucker]).

A portion of left pelvic fin was clipped from allbluehead sucker and flannelmouth sucker andsome suspected hybrids to determine the extent ofhybridization occurring between white sucker andflannelmouth and bluehead sucker populations.Phenotypic evidence of hybridization betweenflannelmouth sucker and white sucker was noted at11 of the 60 stream reaches sampled during 2003.Just over 56% of the sites where flannelmouthsucker were found also contained flannelmouth xwhite sucker hybrids. Hybridization between blue-head sucker and white sucker was documented at 3of the 60 stream reaches (Figure 4). Nearly 34% ofthe sites where bluehead sucker were found alsocontained bluehead x white sucker hybrids. Genet-ic analyses of the tissue samples should determinethe accuracy of phenotypic identification of nativesuckers and their hybrids. It will also help deter-mine the extent of introgression with introducedwhite sucker (e.g., whether hybrids are sterile or arebackcrossing with other individuals). If introgres-sion is occurring, management scenarios mayinvolve preserving isolated headwater populations.If introgression is not occurring, management ofconnected populations may be more appropriate.

Roundtail chub were not collected in 2003 sur-veys but were documented in the Blacks Fork Riverby Wheeler (1997) in 1995 and by WGFD person-nel in 2002 and 2003 (WGFD file information). Inaddition, roundtail chub still persist in Boulder,

Burnt, and Willow lakes near Pinedale, Wyoming(WGFD file information). These populations areunique and may represent the highest elevation(~2,300 m) and perhaps the only lentic popula-tions in the CRB. Additional surveys will definedistributions of these known populations and willhopefully document other populations.

Surveys of the Green River drainage ofWyoming will continue through 2006 and will pro-duce a valuable data set that not only describes thecurrent distribution of the three target species, butalso shows the distribution of all fish speciesthroughout the study area. In addition, habitatinformation will offer insights into how reach-scalestream habitat features are related to observed dis-tributions. And finally, tissue samples collectedfrom bluehead sucker, flannelmouth sucker, androundtail chub will allow us to determine purityand conservation priority for various subpopula-tions identified in the drainage. This informationwill be critical in focusing future research and man-agement efforts to conserve these unique species.

Management of bluehead sucker, flannelmouthsucker, and roundtail chub in the Green Riverdrainage of Wyoming will entail identifying strate-gies to reduce threats to these species. Priorityshould be given to conserving pure populations ofbluehead sucker in Ringdahl Reservoir and flannel-mouth sucker in upper Bitter Creek and possiblyestablishing additional refugia by relocating indi-viduals from these populations to other suitablehabitats in the drainage. Reducing the threat ofhybridization with white sucker will be more chal-lenging, given that they are widely distributedthroughout the drainage. Studies will be initiatedin the upper Big Sandy and Little Sandy drainagesin 2005 to investigate differences in spawningmovements and habitat use between white suckerand native bluehead and flannelmouth suckers.This information may provide insights into strate-gies to control white sucker while preserving nativesuckers. Controlling nonnative fishes and reestab-lishing native fish assemblages may be pursued inthe future. Threats posed by habitat fragmentationand destruction have been and continue to beaddressed through cooperation between theWGFD and other federal, state, and local agenciesas well as private landowners. As a signatory of the

127



Range-wide Conservation Agreement for Round-tail Chub, Bluehead Sucker and FlannelmouthSucker, the WGFD is in the process of developing amanagement plan for these species. The informa-tion collected in these ongoing distributional sur-veys will be instrumental in focusing managementefforts to ensure the persistence of these species inWyoming.

Acknowledgments

We thank Shawn Blajszczak, Angela Corbine,Brooks Fost, Tom Hardy, Kevin Magowan, and SueSpong for assistance in the field. We also thankWyoming Game and Fish Department personnelPete Cavalli, Dirk Miller, Kevin Spence, and DavidZafft for their help in various aspects of this project.And finally, we thank two anonymous reviewers forvaluable suggestions that greatly improved thismanuscript. This project was funded by a U.S. Fish

and Wildlife Service State Wildlife Grant, the U.S.Department of Interior Bureau of Reclamation, andthe Wyoming Game and Fish Department.

References

Baxter, G. T., and J. R. Simon. 1970. Fishes of Wyoming.Wyoming Game and Fish Department, Cheyenne.

Bezzerides, N., and K. Bestgen. 2002. Status review of round-tail chub Gila robusta, flannelmouth sucker Catostomuslatipinnis, and bluehead sucker Catostomus discobolus.Colorado State University, Larval Fish Lab Contribution118, Fort Collins.

Carter, B., and W. A. Hubert. 1995. Factors influencing fishassemblages of a high-elevation desert stream system inWyoming. Great Basin Naturalist 55:169–173.

Miller, D. 1978. Comprehensive survey of the Big SandyRiver. Wyoming Game and Fish Department Adminis-trative Report, Project No. 4077–01-6402. Cheyenne.

Wheeler, C. A. 1997. Current distributions and distribu-tional changes of fishes in Wyoming west of the con-tinental divide. Master’s thesis. University ofWyoming, Laramie.

128

GILL ET AL.


129

The bonytail Gila elegans is a native cyprinidthat belongs to a suite of western fishesknown as “big-river” fishes (Minckley 1973;

Minckley and Deacon 1991). This includes thehumpback chub Gila cypha, razorback suckerXyrauchen texanus, and the Colorado pikeminnowPtychocheilus lucius. Bonytail exhibit the character-istics of big-river fish, including embedded scales,a large size (60 cm, 1 kg), an elongate caudalpeduncle, and an enlarged caudal fin. All of thesefish are endemic to the Colorado River drainage, allare endangered, and their populations are marked-ly reduced from historical levels (Minckley et al.2003). In the lower Colorado River basin, wildpopulations of Colorado pikeminnow are extinct,wild bonytail are extremely rare, and humpbackchub and razorback sucker populations are declin-ing (Minckley 1973, 1991). In discussing bonytail,Minckley (1973) in the book Fishes of Arizonastates “This unique species now appears extinct in

the moderate-sized tributaries of the ColoradoRiver, such as the Gila and Little Colorado….Thefish persists in unknown numbers in some main-stream reservoirs of the Colorado River….”Although written three decades ago, these state-ments still ring true with wild bonytail being evencloser to extinction, if not extinct.

Members of this genus (i.e., bonytail, hump-back chub, and roundtail chub G. robusta) haveattracted considerable attention because of theirunique morphological adaptations to a variety ofriverine habitats and because they were and aresympatric in some river systems. Intermediateforms between these species have been reportedadding to the confusion during the early years ofdetermining the taxonomy of these species (Hold-en and Stalnaker 1970; Smith et al. 1979). Howev-er, more recent morphological and genetic analysesillustrate that these taxa do maintain their distinc-tiveness when sympatric (Dowling and DeMarais

A Review of the Distribution and Management ofBonytail in the Lower Colorado River Basin

ABSTRACT The bonytail Gila elegans is endemic to the Colorado River drainage of the American West. In the

lower basin, this unique cyprinid historically occurred in the Colorado River main stem and its tributaries from Glen

Canyon Dam downstream into Mexico. The species is distinct morphologically and genetically although interme-

diate forms have been noted from the upper basin. Today, wild individuals may persist in the lower basin in Lake

Mojave, Arizona–Nevada. Management activities include broodstock development, the development of grow-out

ponds and native fish habitats, and the stocking of bonytail into lakes Mojave and Havasu. Smaller bonytail (<10

cm) were first stocked in Lake Mojave in 1980. Very few of these bonytail have been collected during annual mon-

itoring of the lake during the last 25 years. Today, larger fish (>25 cm) are stocked after being passive integrated

transponder-tagged. Predation, particularly on the smaller fish, is thought to be the main reason for the lack of

recruitment. Bonytail were stocked in Lake Havasu in 1994 using fish greater than 25 cm. To date, 19 fish have

been collected. Although this is a small number of individuals, their recapture is encouraging and suggests that

bonytail can be established by stocking larger individuals. A management plan for the big-river fishes of the lower

Colorado River basin has been completed by a multi-agency committee.

Charles O. Minckley and Mitch S. Thorson



CHARLES O. MINCKLEY and MITCH S. THORSON U.S. Fish and Wildlife Service, Arizona Fishery Resources Office, 60911 Highway 95, Parker,Arizona 85344, USA


1993). It is has also been documented that localintrogression has occurred between these taxa dur-ing past millennia (Dowling and DeMarais 1993).

This paper presents the status of the bonytail inthe lower Colorado River basin with comments onits distribution and management activities direct-ed toward this species. It is a compilation of thework of numerous agencies and individuals duringthe last three decades.

Discussion

The lower Colorado River basin includes the stateof Arizona and begins at Lees Ferry and ends at theSouthern International Border with Mexico, fol-lowing the flow of the Colorado River. The bound-ary between the two basins is based on a jurisdic-tional decision determined by the Colorado RiverCompact of 1922 and not by river connectedness.Historically, the river flowed into the Gulf of Cali-fornia, although today, its waters rarely, if ever,reach the gulf. The major lower basin tributariesinclude the Virgin, Paria, Little Colorado, BillWilliams, and Gila rivers.

Early collections and anecdotal reports before1950 are generally based on fish provided byanglers, including bonytail reported from GrandCanyon in the 1940s (Dill 1944; Euler 1978). Otherearly records in the basin include Lake Mead,where bonytail were collected and observed spawn-ing (Moffett 1943; Wallis 1951). In Lake Mohave,the earliest record is from 1938, with collectingrecords dropping off markedly by the 1950s (Miller1961; Marsh 1996). A similar trend is found in LakeHavasu with few collections after 1950. Early col-lections from the Colorado River downstream ofLake Havasu and the Salt, Gila, Hardy rivers andSalton Sea near Yuma were made between 1894and 1951. Recent information provided by Muellerand Marsh (2002) in their summary of the devel-opment of the lower Colorado River basin over a120-year period indicate that bonytail were com-monly taken by anglers from the Colorado Riverdownstream of Davis Dam prior to 1950.

Little is known about the habitat types pre-ferred by bonytail within the Colorado River.Jonez and Sumner (1954) observed bonytail “insandy areas along the Colorado River.” Minckley

(1991), in discussing habitat use by bonytail, stat-ed “I suspect that the fish characterized valleyreaches and perhaps flatwater areas in canyons.Adult bonytail must have lived in mid-channel,maintaining position with low-amplitude beatsof the caudal fin, as they do in hatchery raceways,and moving from midwater to surface and bot-tom to inspect and feed on drifting particles inthe water column. The potential power indicatedby the morphology would have been used to passthrough zones of turbulence or escape preda-tors.” These observations are reflected by theabove citations, which illustrate few collectionsfrom canyon bound reaches and increase wherethe river left the canyon entering the broad allu-vial valleys downstream. This is where the mean-dering river formed a network of runs, pools, andbraided channels, associated with backwaters thatprovided important nursery areas for native fish.It is here where the bonytail flourished, living ina relatively slow-moving river, punctuated byperiods of massive floods and droughts, alongshorelines scalloped with backwaters and standsof riparian vegetation. Historically, bonytailoccurred in the Little Colorado River (Minckleyet al. 2001) and in the impounded waters of thebasin such as the Salton Sea, where they wereused for food by Native Americans (Wilkie 1980;Gobalet and Wake 2000). Based on telemetrystudies in Lake Mohave and Cibola High LeveePond, an isolated backwater, adult bonytail seekcover during the day and are active during thenight (Marsh 1997b; Mueller et al. 2003). Covercan be deep water areas or, as in the case of highlevee pond, rip-rapped banks.

The marked decline in bonytail populationswas well underway by the mid-1930s and hasaccelerated to present. The main reason for thisdecline is the impact of introduced fishesthrough predation and competition (Pacey andMarsh 2001). The construction of hydroelectricdams and their coldwater releases do adverselyimpact bonytail and other native fishes (Marsh1985), but there is a consensus among lowerbasin researchers that, if no introduced fisheswere present, there would still be viable popula-tions of big-river fish in the lower ColoradoRiver. Today, Lake Mohave is the only site where

130

MINCKLEY AND THORSON


wild bonytail may still occur (Marsh 1997a),whereas introduced populations are present inthe basin as presented below.

Management Actions

The first management action implemented forbonytail was the development of a broodstock.This was accomplished at Willow Beach NationalFish Hatchery in 1980–1981 (Hamman 1982,1985) using a small number of wild-caught fishfrom Lake Mohave. Although the broodstock wasproduced from a small number of individuals, it isgenetically sound (Minckley et al. 1989). Hatcheryproduction has been very successful, and theresulting progeny have been used to provide bony-tail for all of the stockings in the lower basin.Annual trips have been made to Lake Mohavebetween 1980 and 2005 to collect additional wildfish to increase the genetic diversity of the brood-stock. Even one individual would have a markedeffect on broodstock genetics, increasing diversityby some 20% (Hedrick et al. 2000). These effortshave been unsuccessful. Today, the original brood-stock is aging, and a second broodstock is beingdeveloped by Dexter National Fish Hatchery andTechnology Center. To accomplish this, the originalbroodstock was characterized genetically andplaced into five family lots that were used to devel-op the new broodstock over the course of a 4-yearprogram, with only nonsibling matings acceptedinto the future broodstock. The final number ofbroodstock should reflect equal contributions bythe original founding fish.

The second management action was the stock-ing of bonytail into Lake Mohave. This started in1981, when a small number of the F1s werereleased. Subsequent to that, bonytail produced bynatural reproduction in the Dexter broodstockponds were released into Lake Mohave. Between1981 and 2002, 180,129 bonytail were stocked inLake Mohave. Eighty-six percent (N = 155,005) ofthe fish stocked between 1981 and 1995 were lessthan 10 cm. The remaining fish (25,124; 14%)averaged 25 cm and were stocked between 1996and 2002. This program is ongoing.

A second stocking program began in LakeHavasu in 1993. The goal of this project was to

stock 30,000 bonytail greater than 25 cm by 2003and has been achieved. This project was inresponse to a biological opinion issued to theBureau of Land Management in relation to theLake Havasu Fisheries Improvement Project. Ini-tially, isolated backwaters on Lake Havasu wereused to raise bonytail to a larger size for release intothe lake. This proved logistically difficult, resultingin hatchery raised bonytail produced at WillowBeach NFH and its satellite station, Achii Hanyo,on the Colorado River Indian tribal lands beingused to meet the stocking goals. The effort in LakeHavasu is also being supplemented by bonytailraised in golf course ponds in the vicinity of Park-er, Arizona. All of the bonytail being stocked intoLake Havasu are tagged with 125 MHz passive inte-grated transponder (PIT) tags. The goal of both ofthese programs is to develop and maintain popula-tions of large individuals in both Lake Mohave(50,000) and Lake Havasu (30,000).

The third management activity is the develop-ment of isolated backwaters or constructed pondsfor bonytail. These have two purposes: (1) to growout bonytail to a larger size (>35 cm) before release,and (2) to provide a habitat where a self-recruitingpopulation can develop and persist through time.Although not as successful as first anticipated, thegrow-out ponds have resulted in several hundredbonytail being released into the lower ColoradoRiver. One native fish habitat has been developed(High Levee Pond), and it supported populations ofboth bonytail and razorback sucker from 1993through April 2004 when largemouth bass wereintroduced into the system and decimated the pop-ulations of native fish. Based on what has beenlearned, production of larger fish for augmentationneeds to be increased using hatcheries and grow-out ponds. A number of isolated native fish habitatsthat mimic historic backwater conditions also needto be developed to allow this species to persist out-side of a hatchery environment.

Monitoring

Monitoring of the bonytail introductions in LakeMohave have never been part of an agency drivenprogram to determine the status of introductions.However, week-long trips made in the spring

131

A REVIEW OF THE DISTRIBUTION AND MANAGEMENT OF BONYTAIL


occurred during 1980–1999. Between 2000 and2004, funding was received from the U.S. Fish andWildlife Foundation, allowing an increased pres-ence on Lake Mohave to look for bonytail 5weeks/year. The goal of these trips was to collectwild adults for inclusion into the broodstock; how-ever, they also have been used to monitor popula-tions. Trips are made in the spring as this is whenbonytail spawn and collection is more likely. As aresult of these trips, 113 bonytail have been cap-tured between 1980 and 2003. One coded wiretagged individual has been recaptured. The fishwas from a recent stocking and had been in the sys-tem less than 2 months. Another bonytail that wasPIT-tagged as part of a telemetry study in 1994 wascaptured in 2003, but it is unknown if it was wildor a hatchery release.

Monitoring efforts on Lake Havasu were startedin 1996 and a total of 19 bonytail have been cap-tured since then. Ten were collected from thewaters of Bill Williams River National WildlifeRefuge. Additionally, two have been caught andreleased by anglers and seven by other surveysfrom other areas of the lake. Initially, the anglersdid not know what they had captured but laterconfirmed their catch by looking at availablebrochures and wall mounts of bonytail. The sevenother bonytail were captured during surveys; onefish had been at large for 2 years; the others hadbeen at large from 2 to 6 months. Fish that hadbeen in the system a short time were within 8 kmof their release site. The bonytail that had been inthe lake for 2 years was in the Colorado Riverapproximately 56 km from its release site. Thesmall number of returns from both lakes (particu-larly Lake Mohave) is of concern, although it isencouraging to recover stocked bonytail from LakeHavasu so soon after the start of introductions.

The Management Plan for the Big-River Fish-es of the Lower Colorado River Basin has beencompleted by a multi-agency team (USFWS2004). This document provides suggestions onhow to work toward the recovery of bonytail inthe lower basin in a variety of ways, includinggrow-out ponds, native fish habitats, and aug-mentation of larger fish into the lower ColoradoRiver. It is meant to complement the recoverygoals for big-river fishes (USFWS 2002). The

implementation of the Multispecies Conserva-tion Plan in the near future also will assist incontinuing these management actions.

Acknowledgments

This article is a compilation of some three decadesof information developed and published by indi-viduals from the following agencies/institutions.The Arizona Game and Fish Department, ArizonaState University, California Department of Fishand Game, National Park Service, Lake MeadNational Recreational Area, Nevada Department ofWildlife, U.S. Bureau of Land Management, U.S.Bureau of Reclamation, U.S. Fish and Wildlife Ser-vice Region 2, to include Arizona and Nevada Eco-logical Services Offices, Arizona Fishery ResourcesOffice, Dexter National Fish Hatchery and Tech-nology Center, Willow Beach National Fish Hatch-ery, U.S. Geological Survey, and University ofNevada, Las Vegas.

References

Dill, W. A. 1944. The fishery of the lower Colorado River.California Fish and Game 30:209–212.

Dowling, T. E., and B. D. DeMarais. 1993. Evolutionary sig-nificance of introgressive hybridization in cyprinid fish-es. Nature 362:444–446.

Euler, R. C. 1978. Archeological and paleobiological studiesat Stanton’s Cave, Grand Canyon National Park, Arizona– a report of progress. Pages 142–162 in National Geo-graphic Society Research Report, 1978. National Geo-graphic Society, Washington, D.C.

Gobalet, K. W., and T. A. Wake. 2000. Archeological and pale-ontological fish remains from the Salton basin, SouthernCalifornia. The Southwestern Naturalist 45(4).

Hamman, R. L. 1982. Induced spawning and culture ofbonytail chub. Progressive Fish-Culturist 44:201–203.

Hamman, R. L. 1985. Induced spawning of hatchery rearedbonytail. Progressive Fish-Culturist 47:35–37.

Hedrick, P. W., T. E. Dowling, W. L. Minckley, C. A. Tibbets,B. D. Demarias, and P. C. Marsh. 2000. Establishing acaptive broodstock for the endangered bonytail chub(Gila elegans). Journal of Heredity 91:35–39.

Holden, P. B., and C. B. Stalnaker. 1970. Systematic studies ofthe cyprinid genus Gila, in the upper Colorado Riverbasin. Copeia 1970:409–420.

Jonez, A., and R. C. Sumner. 1954. Lakes Mead and Mojaveinvestigations: a comparative study of an establishedreservoir as related to a newly created impoundment.Nevada Fish and Game Commission, Federal Aid toFisheries Restoration Project Completion Report, F-1-R,1–186, Reno.

132

MINCKLEY AND THORSON


Marsh, P. C. 1985. Effects of incubation temperature on sur-vival of embryos of Native Colorado River fishes. TheSouthwestern Naturalist 30:129–140.

Marsh, P. C. 1996. Bonytail Gila elegans stocking records forLake Mohave, Arizona and Nevada. U.S. Bureau of Recla-mation, Boulder City, Nevada.

Marsh, P. C. 1997a. Age estimation from otoliths of bonytailchub Gila elegans from Lake Mohave, Arizona and Neva-da. U.S. Fish and Wildlife Service, Parker, Arizona.

Marsh, P. C. 1997b. Sonic telemetry of bonytail in LakeMohave, Arizona and Nevada. Final Report to U.S. Geo-logical Biological Resources Division, Denver.

Miller, R. R. 1961. Man and the changing fish fauna of theAmerican West. Papers of the Michigan Academy of Sci-ence, Arts, and Letters 46:365–404.

Minckley, C. O., K. W. Gobelet, and K. Hardin. 2001. Report ona collection of fish remains from the Little Colorado RiverCanyon, Arizona. Proceedings of the Desert Fishes Coun-cil, volume 33. Desert Fishes Council, Bishop, California.

Minckley, W. L. 1973. Fishes of Arizona. Arizona Game andFish Department, Phoenix.

Minckley, W. L. 1991. Native fishes of the Grand Canyonregion: an obituary? Pages 124–177 in G. R. Marzolf, edi-tor. Colorado River ecology and dam management.National Academy Press, Washington, D.C.

Minckley, W. L., and J. E. Deacon, editors. 1991. Battleagainst extinction: native fish management in the Amer-ican West. University of Arizona Press, Tucson.

Minckley, W. L., D. G. Buth, and R. L. Mayden. 1989. Originof broodstock and allozyme variation in hatchery-rearedbonytail, an endangered North American cyprinid fish.Transactions of the American Fisheries Society118:139–145.

Minckley, W. L., J. E. Deacon, T. E. Dowling, P. W. Hedrick, P.C. Marsh, W. J. Matthews, and G. Mueller. 2003. A con-servation plan for lower Colorado River native fishes.BioScience 533:219–234.

Moffett, J. W. 1943. A preliminary report on the fishery ofLake Mead. Transactions of the 8th North AmericanWildlife Conference 8:179–186, Washington, D.C.

Mueller, G., and P. C. Marsh. 2002. Lost, a desert river and

its native fishes; a historical perspective of the lowerColorado River. U.S. Government Printing Office, Infor-mation and Technology Report USGS/BRD/ITR-202–0010, Denver.

Mueller, G., J. Carpenter, P. C. Marsh, and C. O. Minckley.2003. Cibola High Levee Pond Annual Report 2003. U.S.Geological Survey, Denver.

Pacey, C. A., and P. C. Marsh. 2001. Resource use by nativeand non-native fishes of the lower Colorado River. Liter-ature review, summary, and assessment of relative rolesof biotic and abiotic factors in management of an imper-iled indigenous icthyofauna. Final Report. Arizona StateUniversity modification number 1 between ArizonaState University and U.S. Bureau of Reclamation. Ari-zona State University, Department of Biology, Tempeand U.S. Bureau of Reclamation, Boulder City, Nevada.

Smith, G. R., R. R. Miller, and W. D. Sable. 1979. Species rela-tionships among fishes of the genus Gila in the upper Col-orado River drainage. Pages 613–623 in R. M. Linn, editor.Proceedings of the First Conference on Scientific Researchin National Parks. U.S. Department of the Interior,National Park Service, National Park Service Transactionsand Proceedings Series Number 5, Washington, D.C.

USFWS (U.S. Fish and Wildlife Service). 2002. Bonytail (Gilaelegans) recovery goals: amendment and supplement tothe Bonytail Chub Recovery Plan. U.S. Fish and WildlifeService, Mountain-Prairie Region (6), Denver.

USFWS (U.S. Fish and Wildlife Service). 2004. ManagementPlan for the Big-River Fishes of the Lower ColoradoRiver Basin: amendment and supplement to the bony-tail, humpback chub, Colorado pikeminnow, and razor-back sucker recovery plans. U.S. Fish and Wildlife Ser-vice, Albuquerque, New Mexico.

Wallis, O. L. 1951. The status of the fish fauna of the LakeMead National Recreation Area, Arizona-Nevada. Trans-actions of the American Fisheries Society 80:84–92.

Wilkie, P. J. 1980. Prehistoric weir fishing on recessionshorelines of Lake Cahuilla, Salton basin, southeasternCalifornia. Pages 101–102 in Proceedings of the DesertFishes Council, volume 11. Desert Fishes Council, Bish-op, California.

133

A REVIEW OF THE DISTRIBUTION AND MANAGEMENT OF BONYTAIL


135

Sculpins Cottus spp. or cottids are an impor-tant component of freshwater ecosystems ofthe Pacific Northwest. In many habitats, they

are often the most abundant type of fish (Bond1963). Because cottids can be difficult to identify tospecies and are not considered commerciallyimportant, they often are not considered in fish-eries research projects. Much of the research oncottids has dealt with their interrelationships withsalmonids. Freshwater cottids primarily consumebenthic macroinvertebrates but in some situationscan be a predator of salmonid eggs (Foote andBrown 1998) or juvenile salmonids (Hunter 1959).The presence of cottids has been shown to causesalmonid fry to emerge earlier from their redds and

increase their downstream movement (Bardonnetand Heland 1994; Gaudin and Caillere 2000; Mirzaet al. 2001). Alternatively, cottids can be an impor-tant forage fish for stream- (Lowry 1966) and lake-dwelling salmonids (Ricker 1960; Heard 1965;Wyman 1975). Cottids may also compete withsalmonids (Holmen et al. 2003; Hesthagen et al.2004) as well as benthic fishes (Baltz et al. 1982;Resetarits 1997) and crayfish (Miller et al. 1992).Cottids may also have other important effects onaquatic ecosystems, primarily through direct pre-dation (Rosenfeld 1999) or indirectly by intimida-tion and reducing the foraging of invertebrates(Konishi et al. 2001; Kuhara et al. 2001).

Studies of cottids in lotic environments show

Distribution and Habitat Use of Cottids in the Lake Washington Basin

ABSTRACT We collected cottids Cottus spp. from a wide variety of habitat types in the Lake Washington

basin to determine their distribution and habitat use. Habitat types included large lowland lakes, riverine habitats,

and off-channel ponds. Cottids were also collected above and below anadromous barriers on the Cedar River, the

main tributary to Lake Washington. In general, the five species of cottids in the Lake Washington basin appeared

to be spatially segregated. Prickly sculpin Cottus asper was the dominant cottid species in benthic areas of the

lowland lakes. Coastrange sculpin C. alecticus primarily inhabited riffles in the lower reaches of riverine systems,

but they were also found along the shoreline of Lake Washington. Riffle sculpin C. gulosus were typically found in

low-velocity areas in the lower Cedar River and Issaquah Creek and were the dominant species in off-channel

habitats. Torrent sculpin C. rhotheus occupied a wide range of habitats and appeared to be the most numerous

cottid in the lower Cedar River and Bear Creek (a Sammamish River tributary). When sympatric with other cottids,

shorthead sculpin C. confusus appeared to primarily inhabit riffles; however, they appeared to occupy a wide

range of habitats when allopatric (above anadromous barriers).

Roger A. Tabor, Kurt L. Fresh, Dwayne K. Paige, Eric J. Warner, and Roger J. Peters



ROGER A. TABOR U.S. Fish and Wildlife Service, Western Washington Fish and Wildlife Office, 510 Desmond Drive SE, Suite 102, Lacey,Washington 98503, USA

KURT L. FRESH National Marine Fisheries Service, Northwest Fisheries Science Center, 2725 Montlake Boulevard East, Seattle, Washington98112, USA

DWAYNE K. PAIGE Seattle Public Utilities, Watershed Management Division, 19901 Cedar Falls Road SE, North Bend, Washington 98045, USA

ERIC J. WARNER Muckleshoot Indian Tribe, 39015 172nd Avenue SE, Auburn, Washington 98002, USA

ROGER J. PETERS U.S. Fish and Wildlife Service, Western Washington Fish and Wildlife Office, 510 Desmond Drive SE, Suite 102, Lacey, Wash-ington 98503, USA


they are commonly segregated spatially by species(Millikan 1968; Finger 1982; Matheson andBrooks 1983; Daniels 1987). For example, pricklysculpin Cottus asper and coastrange sculpin C.alecticus, two common species in the PacificNorthwest, are usually spatially segregated withprickly sculpin inhabiting pools and coastrangesculpin inhabiting riffles (Taylor 1963; Mason andMachodori 1976; Ringstad 1982; White and Har-vey 1999). Cottids are also a common inhabitantof lacustrine environments, but relatively littlework has been done on their use of various lacus-trine habitat types (Rickard 1980; McDonald andHershey 1992). In many lacustrine systems, morethan one species is present (Northcote 1954; Footeand Brown 1998), but little is known about theirhabitat segregation.

The Lake Washington basin offers an ideal area inwhich to study the distribution,habitat use, and segregation of cot-tids because at least five species ofcottids are present and there is awide variety of habitat types avail-able to them. These habitats rangefrom large, deep lakes to low-gradi-ent, warm rivers, to high-gradient,cool streams. In addition, sections ofthe basin are above impassable fallsthat prevent upstream migration;thus, different species assemblagesof cottids may occur throughout thebasin. Cottids appear to be abun-dant throughout the basin and arean integral part of the aquaticecosystem. Eggers et al. (1978) esti-mated that cottids comprised 73%of the fish biomass of Lake Wash-ington. Cottids also appear to be themost numerous fish in the CedarRiver, where they are importantpredators of sockeye salmonOncorhynchus nerka fry (Tabor etal. 1998). Although cottids areabundant, little is known abouttheir ecology in this basin. Theobjective of this study was to pro-vide some basic information oncottid ecology by determining their

distribution and habitat use in a wide variety ofhabitat types within the Lake Washington basin.

Study Site

We divided the Lake Washington basin (Figure 1)into six major areas that included (1) lowlandlakes, Lake Washington and Lake Sammamish; (2)small independent tributaries to Lake Washingtonand Lake Sammamish; (3) Sammamish River, BearCreek, and Little Bear Creek; (4) Issaquah Creekbasin; (5) lower Cedar River basin; and (6) upperCedar River basin. The Cedar River, SammamishRiver, and Issaquah Creek systems are three largesubbasins. The Cedar River subbasin was furtherdivided into an upper and lower subbasin due tothe presence of Cedar Falls, a natural fish barrier.Lake Washington and Lake Sammamish are both

136

TABOR ET AL.

Figure 1. Map of the Lake Washington basin showing the major streams andlakes. Highlighted stream sections are areas sampled for cottids. Lake Wash-ington, Lake Sammamish, and Chester Morse Lake were also sampled for cot-tids. Numbers represent the six major areas that we used to separate thebasin: (1) lowland lakes, (2) small independent tributaries to lowland lakes,(3) Sammamish River basin, (4) Issaquah Creek basin, (5) lower Cedar Riverbasin, and (6) upper Cedar River basin. The location of the basin within Wash-ington State is shown. LWSC = Lake Washington Ship Canal; L.J. Creek =Laughing Jacobs Creek.


large lowland lakes and were grouped together.Additionally, there are several small independenttributaries to Lake Washington and Lake Sam-mamish, which were grouped together becausethey are at similar elevations and have similar habi-tat conditions.

The Lake Washington basin is approximately1,570 km2 and ranges in elevation from 6 to 1,650m. The eastern 14% (by area) of the basin lies with-in the Cascade Range, while the western 86% ispart of the Puget Sound lowlands. Much of thebasin is heavily urbanized with more than a millionpeople inhabiting the basin. Much of the city ofSeattle is within the basin. The basin has under-gone numerous anthropogenic changes over thepast 150 years.

Lake Washington is a large, monomictic lakewith a total surface area of 9,495 ha and a meandepth of 33 m. The lake typically stratifies fromJune through October. Surface water temperatureranges from 4°C to 6°C in winter to more than20°C in summer and may exceed 23°C in someyears. More than 78% of the shoreline is composedof residential land use. The lake drains through theLake Washington Ship Canal (LWSC), a 13.8-km-long waterway that allows navigation between LakeWashington and Puget Sound. Historically, LakeWashington drained at the south end into theBlack River, which flowed into the Duwamish Riverand eventually into Elliot Bay. The original water-shed was approximately 4,250 km2 and includedthe entire Green River watershed and much of theWhite River watershed. In 1912, the water level ofLake Washington was lowered 2.4 m, and the lakewas rerouted through the LWSC. The BallardLocks, located at the downstream end of theLWSC, now controls the lake level. During winter(December to February), the lake level is kept lowat an elevation of 6.1 m. Starting in late February,the lake level is slowly raised from 6.1 m in Januaryto 6.6 m by 1 May and 6.7 m by 1 June.

Lake Sammamish, located just east of LakeWashington, has a surface area of 1,980 ha and amean depth of 17.7 m. Most of the shoreline iscomposed of residential land use. Issaquah Creek isthe major tributary to Lake Sammamish and entersthe lake at the south end. The creek is 28 km longand originates at approximately 730 m elevation.

Surface water temperatures may exceed 17°C dur-ing the summer. The Sammamish River is a 22-kmriver that connects Lake Sammamish and LakeWashington. The river has a very low gradient(<0.1%) and summer water temperature canexceed 23°C. The river has been extensively chan-nelized and dredged. The river is approximately halfas long as it was historically. The largest tributariesof the Sammamish River are Bear Creek and LittleBear Creek. Bear Creek can exceed 18°C in summer,whereas Little Bear Creek rarely exceeds 16°C.

Cedar River is the largest tributary to LakeWashington and enters the lake at the south end(Figure 1). The river originates at approximately1,090 m elevation and, over its 89-km course, falls1,085 m. Historically, the Cedar River flowed intothe Black River near the old outlet of Lake Wash-ington. We divided the Cedar River basin into twosubbasins; above and below Cedar Falls, a naturalfish barrier at river kilometer (rkm) 55. Basestreamflows near the mouth of the Cedar River areapproximately 375 ft3/s in the fall, winter, andspring and 125 ft3/s in the summer. Water temper-atures in the lower Cedar River may exceed 20°Cduring the summer. The gradient of the lowerCedar River ranges from 0.2% to 0.7%. The Lands-burg Diversion Dam, a water diversion structure atrkm 34.9, was built in 1901 and preventedupstream fish movements until a fish ladder systemwas constructed and began operation in fall 2003.The lower Cedar River subbasin includes severaltributaries as well as Walsh Lake, a 28-ha lake(maximum depth, 11 m) located 38 km upstreamfrom Lake Washington. Surface water temperaturesof Walsh Lake may exceed 25°C during the sum-mer. Tributaries, such as Petersen Creek and WalshDiversion Ditch, that originate at small lakes orponds are often warm during the summer and mayexceed 22°C depending on the proximity to thelakes. Other tributaries fed mostly by surfacerunoff (lower Taylor Creek and upper Rock Creek)have summer water temperatures that are usuallyless than 16°C. One tributary, lower Rock Creek, isa spring-fed stream and has cool water tempera-tures throughout the summer (<14°C).

The upper Cedar River subbasin includesChester Morse Lake, Rex River, and the upperreaches of the Cedar River. All of the upper Cedar

137

DISTRIBUTION AND HABITAT USE OF COTTIDS


River basin is part of the Cedar River MunicipalWatershed that is managed by the Seattle PublicUtilities as a major source of drinking water forresidents of Seattle and much of the surroundingarea; thus, the upper watershed is largely undevel-oped. Chester Morse Lake is an enhanced naturallake that is used as a water supply reservoir. Thelake has a surface area of 698 ha and a maximumdepth of 42 m. Typically, the reservoir is at its high-est water level in late spring, and by late summer,the water level has dropped 8.2 m to its lowest level.Maximum surface water temperatures may exceed21°C during the summer. Summer water tempera-tures in the upper Cedar River and Rex River aretypically less than 16°C.

In comparison to other similar-sized basins inthe Pacific Northwest, the Lake Washington basin isinhabited by a relatively large number of fishspecies. Five cottid species are common in thebasin, which include coastrange sculpin, pricklysculpin, riffle sculpin C. gulosus, shorthead sculpinC. confusus, and torrent sculpin C. rhotheus. Besidescottids, there are 20 native species and at least 15introduced species. Anadromous salmonids in theLake Washington basin include sockeye salmon,Chinook salmon O. tshawytscha, coho salmon O.kisutch, steelhead O. mykiss, and cutthroat trout O.clarkii. Sockeye salmon are by far the most abun-dant anadromous salmonid in the basin. Residentsalmonids include rainbow trout (nonanadromousO. mykiss), cutthroat trout, bull trout Salvelinusconfluentus, mountain whitefish Prosopiumwilliamsoni, and pygmy whitefish P. coulterii. Onlythree salmonid species occur above Cedar Falls:rainbow trout, bull trout, and pygmy whitefish.

Methods

Cottids were collected periodically in the LakeWashington basin from 1995 to 2004. Many of thecollections (1995–2001) were conducted in con-junction with a study to determine consumptionrates of juvenile sockeye salmon by predatory fish-es (Cardwell 1998; Tabor et al. 1998; R. A. Tabor,unpublished data). Collections for the predationstudy included samples from the lower CedarRiver (below Landsburg Dam), Lake Washington,Lake Sammamish, Sammamish River, and the

Bear Creek subbasin. To get a more complete pic-ture of cottid distribution, supplemental surveyswere conducted in other areas of the basin, includ-ing upper Cedar River (summer 1998), IssaquahCreek (spring 2001), and small independent trib-utaries to Lake Washington (summer 2003 and fall2004) and Lake Sammamish (spring 1997 and fall2004). Collections at each supplemental site weredone once. Samples from the lower Cedar River,Bear Creek, Sammamish River, and the shorelineof Lake Washington and Lake Sammamish weretaken from February to June when sockeye salmonfry were present. Each of the latter sites was sur-veyed once every 3 to 4 weeks during this timeperiod. The deep benthic areas of Lake Washing-ton were sampled quarterly from spring 1998 towinter 2000.

Because habitat types sampled encompassed awide range of depths and current velocities, a vari-ety of collection methods were needed to capturecottids. Most sampling of lotic habitats was donewith electrofishing equipment (mostly backpackelectrofishing equipment except a boat electrofish-er was used to sample the downstream end of theCedar River and the Sammamish River). Backpackelectrofishing was conducted differently in slowwater and fast water. For slow-water habitats, per-sonnel slowly moved upstream and shocked fishwere collected with dip nets. For fast-water habi-tats, we used block nets that had a rigid metalframe with a 2-m wooden handle so that theycould easily be held in place in swift water. The netswere 74 cm wide and 31.5 cm high with a 4-mmstretch mesh. One or two block nets were placed inthe water, and we shocked an area immediatelyupstream of the nets that was 3 m (upstream fromthe nets) by the width of the block nets. Stunnedfish floated downstream into the block nets. Withblock nets, all size-classes of cottids were captured;however, when shocked fish were visually netted,small cottids less than 50 mm total length (TL)were often overlooked because they are difficult toobserve and our efforts usually focused on captur-ing cottids large enough to consume sockeyesalmon fry.

Along the shoreline of Lake Washington andLake Sammamish, we collected cottids with bothboat and backpack electrofishing equipment.

138

TABOR ET AL.


Shoreline sampling of Lake Washington was con-centrated in the south part of the lake (south of I-90 Bridge). To sample the deep benthic areas ofLake Washington, we used an otter trawl deployedby a 16.8-m seiner. The cod end had a 32-mmstretch mesh. We sampled cottids in four generalareas of Lake Washington: south end, north end,and two areas in the middle. Sampling depthsranged from 10 to 60 m. Cottids were separatedinto three depth categories: 10–20, 20–40, and40–60 m, and lengths were compared with ananalysis of variance (ANOVA) test and a Tukeyhonestly significant difference test. The deep ben-thic areas (18 and 30 m deep) of Chester MorseLake were sampled with baited minnow traps. AtWalsh Lake and off-channel sites in the CedarRiver, cottids were collected by snorkelers whoslowly swam along the shoreline at night and cap-tured cottids with small aquarium nets. Smallheadlamps were used for illumination.

We divided lotic sites into habitat types basedprimarily on water velocities and lacustrine siteswere divided based on depth and substrate. In themain stem of the Cedar River, the largest river sur-veyed, we sampled cottids from three habitattypes: pools, mid-channel areas of riffles, andshoreline areas of riffles. The mid-channel areas ofriffles were characterized by high current velocities(average column velocity greater than 0.3 m/s),whereas the shoreline areas of riffles had low watervelocities and often had some type of structure,woody debris, or riparian vegetation. In otherstreams, shoreline areas of riffles that had lowwater velocities were relatively small and were notsampled. Only two habitat types (pools and riffles)were used in these streams. At each site, the pres-ence of woody debris and riprap was noted. Shore-line sites of Lake Washington and Lake Sam-mamish were separated into three substratecategories: boulders (riprap), cobble/gravel shore-line, or sand/mud shoreline. Deep benthic areaswere separated based only on depth as the sub-strate was generally silt for all sites.

For each collection, we attempted to catch atleast 10 cottids. After capture, cottids were anes-thetized with MS-222 and identified to species andTL was measured. Some specimens from each loca-tion were taken as vouchers and later were exam-

ined in the laboratory under a dissection micro-scope to verify field identification. Additionally,some specimens were sent to C. Hill, Oregon StateUniversity for further verification.

Riffle sculpin and reticulate sculpin C. perplexushave not been clearly separated by existing mor-phometric characteristics (Wydoski and Whitney2003). The two species are usually separated by thepresence (riffle sculpin) or absence (reticulatesculpin) of palatine teeth (Wydoski and Whitney2003). In initial collections of cottids in the CedarRiver, we found both types of sculpin, with andwithout palatine teeth. To help us confirm theiridentification, we had several specimens from theCedar River, as well as other western Washingtonrivers, sent to the University of British Columbiafor genetic analysis (L. Ritchie, University of BritishColumbia, unpublished data) and an examinationof morphometric characteristics (D. McPhail, Uni-versity of British Columbia, unpublished data).Results of both types of analysis indicated thatthere was little difference between individuals.Therefore, we referred to them collectively as onespecies, riffle sculpin. Kinziger et al. (2005) con-ducted genetic analyses of several sculpin speciesand found that riffle sculpin and reticulate sculpinwere unique species. However, the range and distri-bution of both species is in need of careful exami-nation. For example, Wydoski and Whitney (2003)indicated the northern extent of reticulate sculpinis the Green River in Washington (the drainagebasin immediately south of the Cedar River systemand the two river systems were historically con-nected), while Moyle (2002) indicated that reticu-late sculpin only extend north to the lower Colum-bia River. Reticulate sculpin appear to bewidespread in many areas of western Oregon(Bond 1963), but their distribution in Washingtonis not well known.

ResultsDistribution

Lake Washington and Lake Sammamish.—Theonly cottids collected in Lake Washington wereprickly sculpin and coastrange sculpin (Table 1).In Lake Sammamish, we only found pricklysculpin. Because of the small sample size (n = 79),

139



additional sampling of Lake Sammamish needs tobe undertaken.

Small tributaries to Lake Washington and LakeSammamish.—We collected cottids in four smalltributaries to Lake Washington and two to LakeSammamish (Table 2). Prickly sculpin and coas-trange sculpin were the only cottids collected in thelower sections of each tributary, except one torrentsculpin was collected at May Creek. Coastrangesculpin were also found in upstream sites and wereeither allopatric (Laughing Jacobs Creek) or sym-patric with riffle sculpin (May Creek and TibbettsCreek) or torrent sculpin (May Creek).

Lower Cedar River.—Five cottid species (coas-trange sculpin, prickly sculpin, riffle sculpin,shorthead sculpin, and torrent sculpin) were pres-ent in the lower Cedar River subbasin (Table 2).Prickly sculpin occurred primarily in the lowerkilometer of the Cedar River. Coastrange sculpinwere found only in the lower reaches of the CedarRiver and were present up to rkm 22.4. Rifflesculpin were present throughout the lower CedarRiver and were found in all major tributariesexcept lower Rock Creek. Torrent sculpin were col-lected throughout the lower Cedar River, includ-ing all major tributaries. Shorthead sculpin wererare in the lower 20 km of the Cedar River andwere present in only two of the lower Cedar River

tributaries, lower Rock Creek and upper TaylorCreek. Lower Rock Creek is a spring-fed streamthat has cool water temperatures throughout thesummer. Shorthead sculpin was the only cottidspecies above the impassable waterfall at rkm 0.3on upper Taylor Creek.

Upper Cedar River.—Shorthead sculpin werethe only cottid collected above Cedar Falls on theCedar River. A total of 350 cottids were collectedfrom 13 sites that included 7 sites in the main-stem, 3 sites in Chester Morse Lake, and 3 sites intributaries. Shorthead sculpin were found as farupstream as rkm 87 (elevation, 878 m).

Sammamish River and Bear Creek.—All cottidspecies found in the Cedar River except riffle sculpinwere present in the Sammamish River and Bear Creeksubbasin.At both study sites on the Sammamish River(rkm 0.2 and 18.8), we only collected prickly sculpin(n = 107). Similarly to the lower Cedar River, thelower reaches of Bear Creek were dominated by prick-ly sculpin (rkm 0–1.8) and coastrange sculpin (at twosites, rkm 1.8 and 5.8), while the upper reaches weredominated by torrent sculpin.

Issaquah Creek.—All cottid species excepttorrent sculpin were found in the IssaquahCreek subbasin. Similarly to other drainages,prickly sculpin and coastrange sculpin werepresent in the lower reaches but absent in the

140

TABOR ET AL.

Table 1. Cottid species present in lakes and off-channel ponds in the Lake Washington basin, 1995–2004. Off-channel ponds were alllocated within the lower Cedar River subbasin. The number of sites is the number of sample locations where cottids were collect-ed and includes nearshore sites and deep benthic sites (only sampled in Lake Washington and Chester Morse Lake). Numbersampled is the total number of cottids collected in all habitats.

Type Surface Number Number Species presenceLocation area (ha) of sites sampled Coastrange Prickly Riffle Shorthead Torrent

Lakes

Lake Washington 9,495 14 4,010 X X

Lake Sammamish 1,980 7 79 X

Chester Morse Lake 698 5 69 X

Walsh Lake 28 2 2 X

Off-channel ponds

Cavanaugh Pond 3.8 5 405 X X X

Ricardi Alcove <1 1 51 X X

McDaniels Pond <1 1 20 X

Wetland 69 2.2 1 10 X


141


Table 2. Cottid species present in rivers and streams in the Lake Washington basin, 1995–2004. Streams are listed from downstream toupstream. Length is the total length of each stream. The distance to lake is the number of river kilometers to Lake Washington orLake Sammamish. The number of sites is the number of sample locations where cottids were collected. Number sampled is thetotal number of cottids collected in all habitats. At most sites, more than one habitat type (pools and riffles) was sampled.

Subbasin Length Distance to Number Number Species presenceStream (km) lake (km) of sites sampled Coastrange Prickly Riffle Shorthead Torrent

Lake Washington independent streams

May Creek 14 0 3 43 X X X X

Johns Creek 0.5 0 1 26 X X

Thornton Creek 9 0 2 25 X X

Juanita Creek 6 0 1 10 X X

Lake Sammamish independent streams

Laughing Jacobs Creek 5 0 2 61 X X

Tibbetts Creek 7 0 2 46 X X X

Lower Cedar River 55 0 47 10,798 X X X X X

Molasses Creek 3 7 1 18 X X X

Elliot groundwater channel 0.2 10 1 90 X X

Ricardi Creek 2 11 1 17 X

McDaniels Creek 0.4 19 2 102 X X X

Lower Taylor Creek 6 21 1 24 X X

Petersen Creek 4 22 3 92 X X

Lower Rock Creek 6 29 2 139 X X

Walsh Diversion 7 32 2 72 X X

Webster Creek 6 39 2 68 X X

Upper Rock Creek 10 38 3 54 X X

Upper Taylor Creek

below falls 0.3 47 2 51 X X

above falls 12 48 2 64 X

Upper Cedar River 24 55 7 138 X

Rex River 14 64 1 41 X

Eagle Ridge Creek 1 71 2 39 X

South Fork Cedar River 7 82 2 20 X

Sammamish River 22 0 2 107 X

Little Bear Creek 12 9 2 21 X X

Bear Creek 20 20 7 319 X X X

Cottage Lake Creek 11 27 2 104 X

Issaquah Creek 18 0 4 137 X X X X

East Fork Issaquah Creek 12 4 2 72 X X

Holder Creek 10 18 1 19 X

Carey Creek 9 18 1 32 X X


upper reaches. Prickly sculpin were documentedup to rkm 1.8, and coastrange sculpin werefound 4.5 km upstream from the mouth in EastFork Issaquah Creek. Both riffle sculpin andshorthead sculpin were present in mostupstream sites (rkm 6.9 and upstream) and wereabsent at downstream sites. At the uppermostsite of East Fork Issaquah Creek and HolderCreek, only shorthead sculpin were present.Prickly sculpin and coastrange sculpin did notoverlap with riffle sculpin or shorthead sculpinat any study site within this subbasin.

Habitat Use

Lakes.—Nearly all (>99) of the cottids collected inthe deep benthic area of Lake Washington wereprickly sculpin. The remaining cottids were a fewsmall unidentified cottids that were probably adwarf pelagic form of coastrange sculpin. Seventy-five percent of the prickly sculpin collected werebetween 90 and 140 mm TL. Few prickly sculpinless than 90 mm TL were collected. The maximumlength collected was 230 mm TL, and the mini-mum length was 73 mm TL (mean, 120 mm TL).The size of prickly sculpin collected varied betweenthree depth categories (ANOVA, F = 4.7, df = 2,2,946, P = 0.009); however, the only significant dif-ference (Tukey test) was between shallow (10–20 mdeep) areas (mean, 122.6 mm TL) and mediumdepth (20–30 m deep) areas (mean, 118.5 mm TL).

Prickly sculpin were also the dominant cottid col-lected in the nearshore area of Lake Washington andLake Sammamish. Several coastrange sculpin werealso collected in the nearshore area of Lake Washing-ton and represented 21% of the total catch. No coas-trange sculpin were collected in the nearshore area ofLake Sammamish. We collected cottids at five sites inLake Washington with cobble/large gravel substrate,and at four of these sites, 83% of the cottids collected(n = 379) were prickly sculpin. The fifth site was arestoration site where new substrate (cobble/gravel)was added 2 years before we sampled. At this site,with clean substrate, all of the cottids collected werecoastrange sculpin (n = 23). One site was sampledwith sand substrate, and 96% of the cottids wereprickly sculpin (n = 27). Also, the sites in Lake Sam-mamish had mostly sand substrate and only pricklysculpin were collected.

Only shorthead sculpin were present inChester Morse Lake and were collected as deep as30 m. All shorthead sculpin collected in the deepareas of Chester Morse Lake were 60 mm TL (n= 6; mean, 51 mm TL). Shorthead sculpin werecommon along the shoreline and were collectedin water temperature as high as 21.9°C. Mostshorthead sculpin collected along the shoreline ofChester Morse Lake were also relatively small(mean, 51 mm TL). Only 3% of the cottids col-lected were greater than 65 mm TL. The maxi-mum size collected was only 80 mm TL. In con-trast, 44% of the shorthead sculpin collected inlotic environments of the upper Cedar River sub-basin were greater than 65 mm TL, and the maxi-mum size was 116 mm TL.

Off-channel ponds.—Within four off-channelponds in the lower Cedar River, prickly sculpinwere only found at Cavanaugh Pond, which was thelargest and deepest site (maximum depth, 3.1 m)and closest to Lake Washington. They were collect-ed throughout the pond and represented 81% ofthe cottids collected, but only 8% of the cottids col-lected in the outlet channel. Prickly sculpin as largeas 153 mm TL (mean, 94 mm TL) were collected.Riffle sculpin were present in each site and were thedominant cottid except at Cavanaugh Pond. AtCavanaugh Pond, they made up 18% of the cottidscollected, whereas they made up an average of 94%of the cottids at the other three sites. Torrentsculpin were rare at the off-channel sites except atthe mouth of Cavanaugh Pond and Ricardi Pondon the Cedar River. Shorthead sculpin and coas-trange sculpin were absent from each site.

Stream habitats.—Within the lower reaches oflarge and small tributaries of Lake Washington andLake Sammamish, prickly sculpin were the domi-nant cottid in pools (includes convergence pooland other pools) and coastrange sculpin were thedominant cottid in riffles (Figure 2). Large pricklysculpin (>125 mm TL) were especially abundant inriprap areas (2–4 m deep) near the mouth of theCedar River. Several prickly sculpin greater than175 mm TL (maximum, 239 mm TL) were collect-ed in this area. Within the lower 3 km of the CedarRiver, prickly sculpin were common in pools andshoreline areas of riffles but were rare in mid-chan-nel areas of riffles. Of the nine mid-channel riffle

<

142

TABOR ET AL.


sites, prickly sculpin represented only 1% of thecottids collected and were only present in the threesites closest to Lake Washington.

Between rkm 2.5 and 4.5 of the Cedar River,coastrange sculpin made up an average of 38% ofthe cottids collected in slow-water habitats (poolsand shoreline-riffle sites; n = 12), but between rkm7 and 14, they made up only 11% of the cottids.Most coastrange sculpin greater than 100 mm TL(maximum, 126 mm TL) were collected in slow-water habitats. Between rkm 14 and 22, coastrangesculpin were only found in mid-channel riffles sites.Torrent sculpin were present at every pool andshoreline-riffle site in the lower Cedar River as wellas 86% of the mid-channel riffle sites. Most of thesites in the lower Cedar River where torrent sculpinwere rare or absent were in the lower 5 km, whereprickly sculpin and coastrange sculpin were abun-dant. Above rkm 10, torrent sculpin were the domi-nant sculpin in all three habitat types (Figure 3),comprising an average of 69% of the catch in mid-channel areas of riffles, 82% in the shoreline of rif-fles, and 81% in pools. In the Bear Creek system,

upstream of rkm 5, torrent sculpincomprised more than 99% of the cot-tids in both pools and riffles (Figure 4).The largest torrent sculpin collected inthe Lake Washington basin was 150 mmTL. Torrent sculpin greater than 125mm TL were mostly found in slow-water habitats, especially near riprap indeep pools.

Riffle sculpin were commonly col-lected in slow-water habitats but wererarely present in mid-channel areas ofriffles (Figure 3). A total of two rifflesculpin were collected from 35 mid-channel riffle sites in the lower CedarRiver. In contrast, 1,436 riffle sculpinwere collected from 59 of 77 slow-water sites (pools and riffle-shorelinesites) sampled. Additionally, rifflesculpin in Issaquah Creek were onlyfound in pools (Figure 5). The largestriffle sculpin collected was 124 mm TL.The occurrence of riffle sculpin in thelower Cedar River appeared to be relat-ed to the presence of woody debris or

riprap. At slow-water sites with some type of struc-ture (woody debris or riprap), riffle sculpin madeup a significantly higher percentage of the catch(Mann–Whitney U-test = 194; P < 0.001) than atsites without structure. They were present at 84% ofthe sites with structure and made up an average of21% of the catch. At sites without structure, theywere present at 40% of the sites and made up 4% ofthe catch. Conversely, the percentage of torrentsculpin was higher at slow-water sites withoutstructures (92%) than with structures (78%).

Riffle sculpin (n = 225) were collected in seventributaries to the lower Cedar River. In two tribu-taries where we only sampled pool habitat, rifflesculpin comprised all of the cottids collected in onetributary (Ricardi Creek) and 99% in the othertributary (Elliot groundwater channel). In four ofthe five tributaries where pools and riffles weresampled, they were only collected in pools (torrentsculpin were present in both riffles and pools). Inthe one exception, McDaniels Creek, riffle sculpinmade up 98% (n = 102) of the cottids collected (allthe cottids in pool habitat and 93% of the cottids

143


Prickly Coastrange

Perc

ent

Perc

ent

Cedar R. Issaquah Cr. Tibbetts Cr. LaughingJacobs Cr.

Johns Cr.

Cedar R. Issaquah Cr. Tibbetts Cr. LaughingJacobs Cr.

Johns Cr.

Figure 2. Relative catch (percent) of prickly sculpin and coastrangesculpin in the lower reaches of five tributaries to Lake Washington orLake Sammamish. The streams are in order from largest to smallest.


in riffles). This small tributary is the result of arecently formed channel between a newly con-structed groundwater pond and the Cedar River.Thus, other cottid species may not have yet com-pletely colonized this tributary.

In the upper Cedar River subbasin, shortheadsculpin were found in a wide variety of stream habi-tats. Of the seven slow-water sites (six pools and oneriffle-shoreline site) surveyed above Cedar Falls inthe Cedar River and Rex River, shorthead sculpinwere found at each site, whereas they were only

found in 25% of 76 slow-watersites in the lower Cedar River.Above Cedar Falls, shortheadsculpin (>50 mm TL) in poolswere significantly larger thanthose in riffles (Mann–Whit-ney U-test = 4,611, P = <0.001). The largest shortheadsculpin collected was 132 mmTL. At three other locations(two in Issaquah Creek sub-basin and one in the lowerCedar River subbasin) whereshorthead sculpin wereallopatric, they also occupiedall pools and riffles.

DiscussionDistribution

The distribution of pricklysculpin and coastrange sculpinappeared to be substantiallydifferent than the other threecottid species. Prickly sculpinand coastrange sculpin werethe only cottid species presentin the two lowland lakes andwere usually the only speciespresent in the lower reaches oftributary streams to the low-land lakes. In contrast to theother cottid species in LakeWashington basin, they bothhave planktonic larvae. Bothspecies migrate downstream inthe spring to spawn and after

spawning adults and juveniles migrate upstream inthe fall (Morrow 1980). Additionally, both speciesinhabit the estuarine environment. In slow-waterhabitats where prickly sculpin occur, they are usu-ally the dominant cottid species. They grow to asubstantially larger size than other cottids. Pricklysculpin may displace other cottids directly throughpredation or indirectly through risk of predation orinterference competition. Coastrange sculpinappear to occur primarily in habitats where pricklysculpin are not present such as riffles.

144

TABOR ET AL.

Perc

ent

Perc

ent

Perc

ent

Figure 3. Relative catch (percent) of five species of cottids collected in three habitattypes in the mainstem of the lower Cedar River, 1995–2000. The number of sites surveyed is indicated in parentheses.

River kilometer

River kilometer

River kilometer


Riffle sculpin and torrent sculpin appeared tohave a somewhat patchy distribution. Riffle sculpinwere absent in the Bear Creek drainage, and torrentsculpin were absent in the Issaquah Creek drainage.Additionally, riffle sculpin were only found in onetributary to Lake Washington (May Creek) and onetributary to Lake Sammamish (Tibbetts Creek),and torrent sculpin were only found in May Creek.Their patchy distributions may reflect their histor-ical distribution or be the result of recent changesin habitat conditions and, thus, a reduction of theirhistorical range. Increased urbanization often leadsto severe changes in stream habitat conditions suchas higher peak flows, elevated water temperatures,reduction in water quality, and reduction of habitat

quality (i.e., reduction in theamount of woody debris andpools). The effects of highpeak flows may be especiallydeleterious to cottid popula-tions (Erman et al. 1988).Some locations that areextremely urbanized, such asthe upper watershed ofThornton Creek, were devoidof cottids. Records from theUniversity of Washington fishcollections indicate torrentsculpin were present in LakeWashington (1941 and 1958)and Issaquah Creek (1969).Because riffle sculpin and tor-rent sculpin have demersal lar-vae rather than planktonic lar-vae (coastrange sculpin andprickly sculpin), they disperserelatively slowly (Moyle 2002)and thus may be slow to recol-onize former streams, even ifhabitat conditions improve.For torrent sculpin and rifflesculpin, large lakes such asLake Washington and LakeSammamish may serve as abarrier to dispersal to otherstreams due to predation fromprickly sculpin and exotic fish-es such as yellow perch Perca

flavescens and smallmouth bass Micropterusdolomieu.

Shorthead sculpin are usually found at higherelevations and in colder water than other species(Bond 1963; Mongillo and Hallock 1997). At lowerelevations in areas of warmer water temperature,shorthead sculpin are often absent or rare.Although shorthead sculpin are usually found inareas of cold water, they can withstand high watertemperatures. We collected them in Chester MorseLake that had water temperatures up 21.9°C. Bond(1963) also collected shorthead sculpin in areaswith water temperatures up to 22°C and men-tioned another researcher who collected them inareas with water temperatures up to 24°C. There-

145


Perc

ent

Perc

ent

Figure 4. Relative catch (percent) of three species of cottids collected in two habi-tat types in the Bear Creek subbasin, 1998–2001. Distance to Sammamish River isthe number of rkm to the Sammamish River, the mouth of Bear Creek. No riffleswere sampled at the lower end of Bear Creek (rkm 1.5). One site was surveyed ateach location except at rkm 1.5 of Bear Creek, where three sites were surveyed.

Distance from Sammamish R. (km)

Distance from Sammamish R. (km)

Riffles

Pools

Bear Cr. Cottage Lake Cr.

Bear Cr. Cottage Lake Cr.


fore, the distribution of shorthead sculpin mightbe an interrelationship between competition andwater temperature. In areas of warm water temper-ature, shorthead sculpin may be outcompeted byother cottids and may only be present in riffleswith high current velocities where competitionmay be reduced.

The only cottid species above Cedar Falls wasshorthead sculpin. Riffle sculpin and torrent sculpinwere present immediately below the falls but werenot present above the falls even though there wassuitable habitat. The overall fish assemblage aboveCedar Falls is markedly different than the rest of theLake Washington basin. Only four species occurabove the falls: shorthead sculpin, bull trout, rain-bow trout, and pygmy whitefish. Rainbow trout have

been introduced into this systemand whether the current popu-lation represents a native stock isunclear. The other three speciesare considered native. After thePleistocene glaciation, many ofPuget Sound freshwater fishesare believed to have come fromthe Chehalis River, which wasnot glaciated (McPhail 1967).However, shorthead sculpin arepresent in the lower ColumbiaRiver system and Puget Sounddrainages but absent from theChehalis River system. Bailey andBond (1963) and McPhail (1967)suggested that shorthead sculpinentered the Puget Sound areaafter glaciation by crossing overfrom the Yakima River system(an east slope Columbia Rivertributary). Within the upperYakima River basin, bull troutare found in the upper reachesand pygmy whitefish occur inthree lakes (Wydoski and Whit-ney 2003). Therefore, the fishassemblage above Cedar Fallsmay more reflect the fish assem-blage of the upper Yakima Riverbasin than lowland drainages ofthe Puget Sound.

One habitat type that we did not sample wasthe pelagic area. Previous surveys have document-ed that this area is inhabited by a dwarf pelagicform of coastrange sculpin that occurs in LakeWashington and Lake Sammamish (Ikusemiju1967; Larson and Brown 1975; Wydoski and Whit-ney 2003). Besides these lakes, this dwarf pelagicform of coastrange sculpin has only been docu-mented in one other location, Cultus Lake, BritishColumbia (Ricker 1960). These sculpin are close tothe bottom during the day and then move up intothe water column at night. During our daytimebottom trawling, a few small unidentified cottidswere caught, and based on their size and col-oration, they were probably the dwarf pelagicform of coastrange sculpin. These sculpin are

146

TABOR ET AL.

Figure 5. Relative catch (percent) of four species of cottids collected in two habi-tat types in the Issaquah Creek subbasin, April–May 2001. Distance from lake isthe number of rkm to Lake Sammamish, the mouth of Issaquah Creek. One sitewas surveyed at each location. No riffles were present at the lower end ofIssaquah Creek (rkm 0.1). EF Iss. Cr. = East Fork Issaquah Creek

Perc

ent

Perc

ent

Distance from lake (km)

Distance from lake (km)

RifflesIssaquah Cr. EF Iss. Cr. Holder Cr. Carey Cr.

PoolsIssaquah Cr. EF Iss. Cr. Holder Cr. Carey Cr.


morphologically different from coastrangesculpin, but isozyme analyses done in the 1970sindicated that they were similar to coastrangesculpin from the Cedar River. Additional analysesare needed to determine their genetic relationshipto coastrange sculpin.

Habitat segregation

In general, cottids in the Lake Washington basinappeared to be spatially segregated. Pricklysculpin and riffle sculpin inhabited quiet waterswith prickly sculpin inhabiting lakes and lowerreaches of streams and riffle sculpin inhabitingthe upper reaches. Coastrange sculpin and short-head sculpin were common in riffles, but theirdistribution did not overlap significantly. Spatialsegregation of cottids has been documented inother studies. Patten (1971) found a similar trendin the Elokomin River in southwest Washingtonwhere prickly sculpin and coastrange sculpin werethe dominant sculpin in the lower reaches withprickly sculpin inhabiting the pools and coas-trange sculpin inhabiting the riffles. Torrentsculpin inhabited riffles and were the dominantsculpin in upstream locations and reticulatesculpin were primarily found in pools inupstream locations. Moyle (2002) reported that inCentral Valley, California, riffle sculpin typicallyoccupy the cool, upper reaches of streams, whileprickly sculpin occupy the warm, lower reaches.In the Marys River basin in Oregon, torrentsculpin and Paiute sculpin C. beldingii were segre-gated within riffles, and reticulate sculpinoccurred primarily in pools (Finger 1982). Sever-al studies have examined the spatial segregationbetween prickly sculpin and coastrange sculpin inlotic environments and found that prickly sculpininhabit pools and other slow-water habitats andcoastrange sculpin inhabit riffles (Taylor 1963;Mason and Machodori 1976; Ringstad 1982,White and Harvey 1999). In contrast, Brown et al.(1995) found no segregation between pricklysculpin and coastrange sculpin in the Eel Riverbasin, California. Lack of segregation was believedto be due to low cottid densities. Also, deep poolswere not sampled and catch in pools and riffleswas not separated.

The mechanisms that cause spatial segregation

among cottids are poorly understood. Habitatchoice, competition, and predation are often con-sidered the casual mechanisms for spatial segrega-tion. Interference competition has been suggestedto occur in cottids (Finger 1982) while exploitationcompetition has not been documented. Largeprickly sculpin and large torrent sculpin often con-sume cottids (Tabor et al. 1998). Predation effectscan also be caused through direct predation orindirectly through intimidation and forcing small-er fish to avoid a particular habitat type inhabitedby predators. Interrelated with competition andpredation are the effects of various physical vari-ables such as temperature and current velocity.

For the most part, cottids in the Lake Wash-ington basin appeared to be spatially segregatedon a macrohabitat scale; however, there was alsoconsiderable overlap in habitat use in some loca-tions. Torrent sculpin were found in a wide vari-ety of habitats and usually coexisted with at leastone other cottid species. Prickly sculpin andcoastrange sculpin were often found together inthe nearshore area of Lake Washington. We onlyexamined the macrohabitat use (pools and rif-fles), and spatial segregation may have been at amesohabitat or microhabitat scale. Finger (1982)found that torrent sculpin and Piute sculpin useddifferent areas of cobbles in riffles when theycoexisted. In the lower Cedar River subbasin, rif-fle sculpin and torrent sculpin coexist in pools.Riffle sculpin appeared to occupy the deposition-al area (fine substrates and low current velocities)of the pool near the shoreline and were usuallyassociated with woody debris, riprap, or under-cut bank, while torrent sculpin appeared toinhabit the middle of the pool (coarser substratesand higher current velocities) and were oftenaway from cover.

Cottids may also coexist in the same habitatthrough food resource partitioning. We haveexamined the diets of cottids in the Cedar Riverand have not found any noticeable difference in thediet of cottids inhabiting the same habitat type(Tabor, unpublished data). However, we only iden-tified aquatic insects (the principal component oftheir diet) to taxonomic order, and prey size wasnot determined. Torrent sculpin and pricklysculpin have a wider gape than other cottids (Pat-

147



ten 1971) and thus may be able to take advantageof larger prey such as large stoneflies.

The level of complexity or heterogeneity maydetermine if more than one species can coexist inthe same habitat. Riffles, for example, often havea variety of substrate sizes and current velocitiesand may be more complex than many pools. Pageand Schemske (1978) found that in each pool,only one species of darter Etheostoma spp. waspresent, whereas two species were present in rif-fles. Pools had flat stone substrates with little het-erogeneity, while riffles had a variety of sub-strates and current velocities. Finger (1982)suggested that riffles with high embeddedness(reduced heterogeneity) could only support tor-rent sculpin, whereas other riffles that were notembedded (increased heterogeneity) could sup-port both torrent sculpin and Piute sculpin. Inthe nearshore area of Lake Washington, we foundthat prickly sculpin and coastrange sculpin wereable to coexist in cobble substrates, but coastrangesculpin were rare in sandy beach areas, possiblydue to the lack of complexity. We only sampledone sandy site, however, and additional samplingneeds to be undertaken. In the lower Cedar River,slow-water habitats (pools and shoreline of rif-fles) as well as riffles often had more than onespecies. In slow-water habitats with woody debrisor riprap, riffle sculpin appeared to be able tocoexist with torrent sculpin but in areas withoutthese structures they were rare.

The habitat use of shorthead sculpin providesthe best evidence that cottids display interactivesegregation, where a cottid species has a confinedniche when sympatric with other cottid speciesand has a much broader niche when allopatric(Nilsson 1967). When sympatric with other cottids(lower Cedar River), shorthead sculpin appeared toprimarily inhabit riffles; however, when allopatric(upper Cedar River), they appeared to occupy awide range of habitats, including pools and lakes.At sites where shorthead sculpin are allopatric inWashington, they commonly inhabit slow velocityas well as high velocity habitats. In the North ForkSnoqualmie River, shorthead sculpin are allopatricand were documented in riffles and pools as well asbeaver ponds (Sweeney et al. 1981). Hillman(1989) also found shorthead sculpin inhabited

glides and pools in the Wenatchee River where theyare allopatric. In streams of the Olympic Peninsu-la, shorthead sculpin frequently inhabited glidesand pools, especially where other cottid specieswere not present (Mongillo and Hallock 1997). Incontrast, shorthead sculpin were never found inareas with slow current velocities at several collec-tion sites in Oregon (Bond et al. 1988). This mayhave been because only sites where shortheadsculpin were sympatric with other cottids wereincluded in the study.

Riffle sculpin also displayed interactive segrega-tion. For the most part, riffle sculpin were confinedto pools and other low-velocity areas in the LakeWashington basin. However, at one site, whereother cottids were absent, riffle sculpin occupiedboth pools and riffles. In Deer Creek, California,riffle sculpin were the dominant fish in upstreamriffles (Baltz et al. 1982). Additionally, reticulatesculpin have been found in pools and riffles whenallopatric but were confined to pools when torrentsculpin were present (Finger 1982). In most loca-tions in the Lake Washington basin where rifflesculpin occur, they coexist with torrent sculpin;thus, riffle sculpin may not compete well with tor-rent sculpin and perhaps other cottids and theyoccupy habitats where these other cottids are notcommonly found.

Large shorthead sculpin were found in the river-ine environment above Chester Morse Lake butappeared to be absent from the lake itself (bothshoreline and deep areas). In contrast, large pricklysculpin are present in Lake Washington and pricklysculpin in riverine habitats (upstream of conver-gence pool) were generally smaller. Different tech-niques were used to collect cottids in the deep areasof the lakes, but minnow traps were used byRickard (1980) in Lake Washington and the sizedistribution was similar to what we collected withtrawls. Also, the nearshore area was sampled thesame in the two lakes. Prickly sculpin (maximumsize, 239 mm TL) attain a much larger size thanshorthead sculpin (maximum size, 120 mm TL)and thus may grow large enough that most lake-dwelling predators can no longer consume them.Prickly sculpin may also be better able to avoidpredators than shorthead sculpin. The predatorspecies are different between the two lakes; Lake

148

TABOR ET AL.


Washington predators would include primarilynorthern pikeminnow Ptychocheilus oregonensis(Eggers et al. 1978) and cutthroat trout (Nowak etal. 2004), while in Chester Morse Lake, it wouldinclude bull trout and rainbow trout (Wyman1975). Additionally, the annual water level fluctua-tion of Chester Morse Lake is 8.2 m but only 0.6 min Lake Washington, and therefore, the productionof cottids in Chester Morse Lake could be affectedby the water level fluctuations. The ecology andbehavior of shorthead sculpin in lakes is unknownand why few large individuals are present is unclear.

Acknowledgments

We appreciate the help of the many U.S. Fish andWildlife Service (USFWS) employees that assistedwith the field work, which included J. Chan, S.Hager, M. Mizell, H. Gearns, R. Piaskowski, A. Hird,C. McCoy, D. Low, and B. Missildine. B. Footen,Muckleshoot Indian Tribe (MIT), assisted with theLake Washington sampling; A. Cardwell, Universityof Washington conducted much of the Bear Creeksampling; Alex Ottley, MIT, helped with the IssaquahCreek sampling; W. Belknap, Seattle Public Utilities,assisted with the upper Cedar River sampling; and D.Estelle, Washington Department of Fish andWildlife, conducted much of the independent tribu-tary sampling. We thank the crew of the Chasina fortheir help with the Lake Washington trawling. C.Hill, Oregon State University, D. McPhail, and L. Ric-thie, University of British Columbia, helped withcottid identification. S. Dilley, USFWS, helped withthe figures. B. Wunderlich, USFWS, J. Scheurer,National Marine Fisheries Service, and two anony-mous reviewers provided valuable suggestions thatgreatly improved this manuscript.

References

Bailey, R. M., and C. E. Bond. 1963. Four new species offreshwater sculpins, genus Cottus from western NorthAmerica. Occasional Papers of the Museum of Zoology,University of Michigan 634:1–27.

Baltz, D. M., P. B. Moyle, and N. J. Knight. 1982. Competitiveinteractions between benthic stream fishes, riffle sculpin,Cottus gulosus, and speckled dace, Rhinichthys osculus.Canadian Journal of Fisheries and Aquatic Sciences39:1502–1511.

Bardonnet, A., and M. Heland. 1994. The influence of poten-tial predators on the habitat preferenda of emergingbrown trout. Journal of Fish Biology 45(SupplementA):131–142.

Bond, C. E. 1963. Distribution and ecology of freshwatersculpins, genus Cottus, in Oregon. Doctoral dissertation.University of Michigan, Ann Arbor.

Bond, C. E., E. Rexstad, and R. M. Hughes. 1988. Habitat useof twenty-five common species of Oregon freshwaterfishes. Northwest Science 62:223–232.

Brown, L. R., S. A. Matern, and P. B. Moyle. 1995. Compara-tive ecology of prickly sculpin, Cottus asper, and coas-trange sculpin, Cottus aleuticus, in the Eel River, Califor-nia. Environmental Biology of Fishes 42:329–343.

Cardwell, A. 1998. Piscivorous fish feeding habits and pre-dation on sockeye salmon fry in Bear Creek, Washing-ton. Undergraduate thesis. University of Washington,Seattle.

Daniels, R. A. 1987. Comparative life histories and micro-habitat use in three sympatric sculpins (Cottidae: Cottus)in northeastern California. Environmental Biology ofFishes 19:93–110.

Eggers, D. M., N. W. Bartoo, N. A. Rickard, R. E. Nelson, R. C.Wissmar, R. L. Burgner, and H. Devol. 1978. The LakeWashington ecosystem: the perspective from the fishcommunity production and forage base. Journal of theFisheries Research Board of Canada 35:1553–1571.

Erman, D. C., E. D. Andrews, and M. Yoder-Williams. 1988.Effects of winter floods on fishes in the Sierra Nevada.Canadian Journal of Fisheries and Aquatic Sciences45:2195–2200.

Finger, T. R. 1982. Interactive segregation among threespecies of sculpins (Cottus). Copeia 1982:680–694.

Foote, C. J., and G. S. Brown. 1998. Ecological relationshipbetween freshwater sculpins (genus Cottus) and beachspawning sockeye salmon (Oncorhynchus nerka) inIliamna Lake, Alaska. Canadian Journal of Fisheries andAquatic Sciences 55:1524–1533.

Gaudin, P., and L. Caillere. 2000. Experimental study of theinfluence of presence and predation by sculpin, Cottusgobio L., on the drift of emergent brown trout, Salmotrutta L. Archiv fur Hydrobiologie 147:257–271.

Heard, W. R. 1965. Limnetic cottid larvae and their utilizationas food by juvenile sockeye salmon. Transactions of theAmerican Fisheries Society 94:191–193.

Hesthagen, T., R. Saksgard, O. Hegge, B. K. Devro, and J.Skurdal. 2004. Niche overlap between young brown trout(Salmo trutta) and Siberian sculpin (Cottus poecilopus) ina subalpine Norwegian river. Hydrobiologia 521:117–125.

Hillman, T. W. 1989. Nocturnal predation by sculpins on juve-nile Chinook salmon and steelhead. Pages 249–264 inDon Chapman Consultants, Inc. Summer and winterecology of juvenile Chinook salmon and steelhead trout inthe Wenatchee River, Washington. Final Report to ChelanCounty Public Utility District, Wenatchee, Washington.

Holmen, J., E. M. Olsen, and L. A. Vollestad. 2003. Interspe-cific competition between stream-dwelling brown troutand Alpine bullhead. Journal of Fish Biology62:1312–1325.

149



Hunter, J. G. 1959. Survival and production of pink and chumsalmon in a coastal stream. Journal of the FisheriesResearch Board of Canada 16:835–886.

Ikusemiju, K. 1967. The life history and ecology of Cottus sp. inLake Washington. Master’s thesis. University of Washington,Seattle.

Kinziger, A. P., R. M. Wood, and D. A. Neely. 2005. Molecularsystematics of the genus Cottus (Scorpaeniformes: Cotti-dae). Copeia 2005:303–311.

Konishi, M., S. Nakano, and T. Iwata. 2001. Trophic cascadingeffects of predatory fish on leaf litter processing in aJapanese stream. Ecological Research 16:415–422.

Kuhara, N., S. Nakano, and H. Miyasaka. 2001. Alterations in thegrazing activities of cased caddisfly larvae in response tovariations in predation risk and resource level. EcologicalResearch 16:705–714.

Larson, K. W., and G. W. Brown, Jr. 1975. Systematic status ofa midwater population of freshwater sculpin (Cottus)from Lake Washington, Seattle, Washington. Journal ofthe Fisheries Research Board of Canada 32:21–28.

Lowry, G. R. 1966. Production and food of cutthroat trout inthree Oregon coastal streams. Journal of Wildlife Manage-ment 30:754–767.

Mason, J. C., and S. Machodori. 1976. Populations of sym-patric sculpins, Cottus aleuticus and Cottus asper, in fouradjacent salmon-producing coastal streams on VancouverIsland, B.C. Fishery Bulletin 74:131–141.

Matheson, R. E., Jr., and G. R. Brooks, Jr. 1983. Habitat segre-gation between Cottus bairdi and Cottus girardi: an exam-ple of complex inter- and intraspecific resource partition-ing. American Midland Naturalist 110:165–176.

McDonald, M. E., and A. E. Hershey. 1992. Shifts in abun-dance and growth of slimy sculpin in response to changesin the predator population in an Arctic Alaskan lake.Hydrobiologia 240:219–223.

McPhail, J. D. 1967. Distribution of freshwater fishes in west-ern Washington. Northwest Science 41:1–11.

Miller, J. E., J. F. Savino, and R. K. Neely. 1992. Competition for foodbetween crayfish (Orconectes virilis) and the slimy sculpin(Cottus cognatus). Journal of Freshwater Ecology 7:127–136.

Millikan, A. E. 1968. The life history of Cottus asper (Richard-son) and Cottus gulosus (Girardi) in Conner Creek, Wash-ington. Master’s thesis. University of Washington, Seattle.

Mirza, R. S., D. P. Chivers, and J-G. J. Godin. 2001. Brook charralevins alter timing of nest emergence in response tochemical cues from fish predators. Journal of ChemicalEcology 27:1775–1786.

Mongillo, P. E. and M. Hallock. 1997. Distribution and habi-tat of native nongame stream fishes of the Olympic Penin-sula. Washington Department of Fish and Wildlife, Tech-nical Report #FRD 97–05, Olympia.

Morrow, J. E. 1980. The freshwater fishes of Alaska. AlaskaNorthwest Publishing Company, Anchorage.

Moyle, P. B. 2002. Inland fishes of California. University ofCalifornia Press, Berkeley.

Nilsson N. A. 1967. Interactive segregation between fishspecies. Pages 295–313 in S. D. Gerking, editor. The bio-

logical basis of freshwater fish production. Blackwell Sci-entific Publishers, Oxford, England.

Northcote, T. G. 1954. Observations on the comparativeecology of two species of fish, Cottus asper and Cottusrhotheus, in British Columbia. Copeia 54:25–28.

Nowak, G. M., R. A. Tabor, E. J. Warner, K. L. Fresh, and T. P.Quinn. 2004. Ontogenetic shifts in habitat and diet ofcutthroat trout in Lake Washington, Washington. NorthAmerican Journal of Fisheries Management 24:624–635.

Page, L. M., and D. W. Schemske. 1978. The effect of interspe-cific competition on the distribution and size of darters ofthe subgenus Catnotus (Percidae: Etheostoma). Copeia1978:406–412.

Patten, B. G. 1971. Predation by sculpins on fall Chinooksalmon, Oncorhynchus tschawytscha, fry of hatchery ori-gin. U.S. Department of Commerce, National MarineFisheries Service, Special Scientific Report, Fisheries No.621, Washington, D.C.

Resetarits, W. J., Jr. 1997. Interspecific competition and quali-tative competitive asymmetry between benthic streamfish. Oikos 78:429–439.

Rickard, N. A. 1980. Life history and population characteris-tics of the prickly sculpin (Cottus asper Richardson) inLake Washington. Master’s thesis. University of Washing-ton, Seattle.

Ricker, W. E. 1960. A population of dwarf coastrange sculpins(Cottus aleuticus). Journal of the Fisheries Research Boardof Canada 17:929–932.

Ringstad, N. 1982. Carnation Creek watershed project freshwa-ter sculpins: genus Cottus, a review. Pages 219–239 in G. F.Hartman, editor. Proceedings of the Carnation Creek work-shop: a ten-year review, Malaspina College, Nanaimo, B.C.

Rosenfeld, J .S. 1999. The effect of fish predation on benthiccommunity structure in streams. Doctoral dissertation.University of British Columbia, Vancouver.

Sweeney, S. J., K. W. Kurko, and T. C. Juelson. 1981. NorthFork Snoqualmie River basin wildlife study. Final reportof Washington Department of Game to U.S. Army Corpsof Engineers, Seattle.

Tabor, R. A., J. Chan, and S. Hager. 1998. Predation on sock-eye salmon fry by cottids and other predatory fishes inthe Cedar River and southern Lake Washington, 1997.Miscellaneous report. U.S. Fish and Wildlife Service,Western Washington Office, Lacey, Washington.

Taylor, G. D. 1963. Distribution and habitat response of coas-trange sculpin, Cottus aleucticus, and prickly sculpin, C.asper, in the Little Campbell River, British Columbia. Mas-ter’s thesis. University of British Columbia, Vancouver.

White, J. L., and B. C. Harvey. 1999. Habitat separation ofprickly sculpin, Cottus asper, and coastrange sculpin,Cottus aleuticus, in the mainstem Smith River, north-western California. Copeia 1999:371–375.

Wydoski, R. S., and R. R. Whitney. 2003. Inland fishes ofWashington. University of Washington Press, Seattle.

Wyman, K. H., Jr. 1975. Two unfished salmonid populationsin Lake Chester Morse. Master’s thesis. University ofWashington, Seattle.

150

TABOR ET AL.


151

Pacific lamprey Lampetra tridentata are con-sidered primitive and have an ancestry dat-ing back at least 400 million years (Bond

1996). Pacific lamprey are anadromous and para-sitic during their ocean phase. Pacific lampreyspawn and rear in coastal and inland streams fromnorthern Baja California to Alaska (Scott andCrossman 1998; Simpson and Wallace 1982).Ammocoetes (larvae) occupy freshwater streamsubstrates for 4 to 7 years (Beamish and Levings1991) where they filter-feed on plant and animaldetritus (likely desmids, diatoms, and protozoa;Creaser and Hann 1929; Richards and Beamish1981). Ammocoetes undergo transformation(Richards and Beamish 1981) into macropthalmia(juveniles) and migrate to the ocean to begin a par-asitic phase. Following 1 to 2 years parasiticallyfeeding on a variety of fish species and potentiallymammals (Scott and Crossman 1998) in the ocean,Pacific lamprey adults return to spawn in freshwa-ter (Beamish and Levings 1991), surviving only ashort time after spawning.

Pacific lamprey were historically abundant inthe Columbia River basin (Close et al. 1995), per-haps numbering in the millions. Hydroelectricimpacts and alteration of rearing habitat are con-sidered two major factors contributing to Pacificlamprey decline in the Columbia River basin andSnake River subbasin (Jackson et al. 1996). Hydro-electric projects hinder upstream adult passage anddownstream ammocoete and macropthalmia out-migrations and delay downstream movementthrough slack water areas. Significant habitat alter-ation to spawning and rearing streams since 1850has potentially resulted in reduced production inthe Columbia River basin and Snake River sub-basin as well. Jackson et al. (1996) cited mining,logging, irrigation practices, agricultural activities,and streamside riparian habitat destruction as hav-ing potential negative impacts affecting ammo-coete rearing conditions.

Little is known about ammocoete rearing habi-tat requirements and utilization in the ColumbiaRiver basin. Hammond (1979), however, provides

Pacific Lamprey Ammocoete Habitat Utilization in Red River, Idaho

ABSTRACT The Pacific lamprey Lampetra tridentata is a native Snake River basin fish species occupying a

unique ecological niche. The recent decline in numbers of returning Pacific lamprey adults to the Snake River basin

has focused attention on the species. In 2000–2002, we employed electrofishing surveys to determine habitat

utilization and distribution of Pacific lamprey ammocoetes in Red River, South Fork Clearwater River drainage,

Idaho. Ammocoete average densities were 25.7/100 m2 in scour pools, 4.4/100 m2 in riffles, 2.1/100 m2 in rapids,

and 253.3/100 m2 in the one alcove sampled. Ammocoetes were found in water depths ranging from 1.0 cm to

1.0 m; however, the two greatest densities were observed in habitat units with maximum depths greater than 0.50

m. Pacific lamprey ammocoete density decreased with increased velocity and coarse substrate, and increased

with fine and medium substrates and riparian shade.

Christopher W. Claire, Timothy G. Cochnauer, and George W. LaBar



CHRISTOPHER W. CLAIRE and TIMOTHY G. COCHNAUER Idaho Department of Fish and Game, 1540 Warner Avenue, Lewiston, Idaho 83501, USA,[email protected], [email protected]

GEORGE W. LABAR Department of Fish and Wildlife Resources, University of Idaho, Moscow, Idaho 83843, USA, [email protected]


documentation of ammocoete habitat for selectedsites in the Potlatch River, Idaho. Pletcher (1963)provided extensive information on lamprey lifehistory in the Salmon River, British Columbia,including documentation of selected stream habi-tats where ammocoetes were captured. The Clear-water River drainage and Red River in north-cen-tral Idaho currently support a population ofPacific lamprey. Determination of life history andhabitat utilization required extensive examinationof the subbasin’s stream habitats. The objective ofthis study was to determine habitat utilization ofPacific lamprey ammocoetes in response to sixstream parameters: habitat type, habitat unit flowvelocity, depth, substrate, temperature, and ripar-ian canopy in Red River.

Study Area

Red River is a fourth-order tributary watershed con-sisting of approximately 42,000 ha in the South ForkClearwater River drainage (300,440 ha) in north-central Idaho (Figure 1). Red River joins AmericanRiver 8.0 km west of Elk City to form the SouthFork Clearwater River at river kilometer (rkm) 83.0.Red River is the largest tributary to the South ForkClearwater River and contributes approximatelyone-third of the flow of the South Fork ClearwaterRiver. Average maximum flow of 31.9 m3/s occursin May and average minimum flowof 1.6 m3/s occurs in September atrkm 80.0 in the South Fork Clearwa-ter River. Land use in the South ForkClearwater River drainage is pre-dominantly forestry related in theupper basin with livestock grazing inthe middle and lower reaches. His-torically, mining was centered in theupper reaches. Widespread miningfrom the 1860s to the mid-1900soccurred in four tributaries: CrookedRiver, Red River, American River, andNewsome Creek. In the early 1900s,bucket dredge mining occurred inthe upper drainage and the Red Riversubbasin (U.S. Forest Service 1998).Dredging impacted numerousstream reaches by confining the

channel, reducing habitat diversity, eliminatingriparian canopy, and directly discharging sedimentinto Red River.

Information pertaining to anadromous speciespopulations, including Pacific lamprey, in the SouthFork Clearwater River drainage before 1900 is limit-ed. In 1910, Grangeville Electric Light and PowerCompany constructed Harpster Dam at rkm 32.0on the main South Fork Clearwater River. AdultPacific lamprey migration was likely possible, butrestricted, over the dam from 1935 to 1949. Highflows destroyed the fishway in 1949 eliminatingadult salmonid passage until the dam was removedin 1963 (U.S. Forest Service 1998). The impact toPacific lamprey upstream migrants is unknown.

Methods

In 2000, we divided the 41-km length of the main-stem Red River into 1-km sections from mouth toheadwaters. One-hundred-meter stream reacheswere randomly selected from each 1-km section fordetermination of habitat utilization and ammo-coete distribution. In subsequent years, samplingwas focused in the lower Red River as no ammo-coetes were found above rkm 7.5 even though sim-ilar habitat types were found above this point.Stream habitat units were classified as pool, riffle,glide, falls, rapids, cascades, and alcoves using a

152

CLAIRE ET AL.

Figure 1. Location of Pacific lamprey ammocoete habitat utilization studiesin Red River, Idaho, 2000–2002.


modified version of stream classification method-ology utilized in Platts et al. (1983) and Overton etal. (1997). Pools, riffles, and rapids were furtherclassified into subtypes (Table 1). Individual sam-pling units were defined as a single pool, riffle,glide, falls, rapids, cascades, and alcove. Streamhabitat parameters measured in selected habitatunits included wetted width, channel width, habi-tat unit length, maximum depth, average flowvelocity, substrate size percent composition, streamtemperature, and percent riparian canopy cover.

The substrates within habitat units were visuallyclassified using Platts et al. (1983) size classification(Table 2). Water velocities were measured with aGeneral Oceanics blade type flowmeter 1.0 cmabove the substrate (Hammond 1979) at 25% ofdistance from left bank, 25% of distance from rightbank, and at center distance in a habitat unit toobtain average and focal velocities. In instanceswhere a habitat unit failed to span the entire streamwidth, velocities were measured at the same propor-tional distances within the width of that habitatunit. The three measurements within a habitat unitwere averaged. Maximum depth was recorded inindividual habitat units. Individual water velocityand site depth measurements were taken over sub-strates where Pacific lamprey ammocoetes were cap-

tured on emergence. Individual site emergence flowvelocity measurements for each unit were averaged.Substrates yielding Pacific lamprey ammocoeteswere documented. Stream canopy cover (shade) val-ues expressed as percentages were obtained using astandard concave forestry densiometer. Stream tem-peratures were measured at each individual habitatunit location with a hand-held mercury thermome-ter positioned 2.0 cm below the water surface. Singlepoint location substrate temperatures were meas-ured for 10 Red River sites (rkm 1.0 to rkm 10.0) infiner substrates at a depth of 10 cm on 9 August2001. Additional stream temperature data wereobtained at rkm 5.0 utilizing an Onset ComputerCorporation HOBO temperature logger.

Habitat types in the randomly selected 100-mreaches were electroshocked systematically with anEngineering Technical Services ABP-2 elec-troshocker. However, in some instances, not allhabitat units in a randomly selected stream reachwere sampled due to logistics or weather condi-tions. Habitat units were electroshocked from bankto bank working upstream from the lowermostpoint of the habitat unit to the upstream end toensure complete coverage of the unit. Due to theirrarity and the fact that 100-m sections were ran-domly selected, no falls, cascades, typical rapids,rapids with bedrock, plunge pools, dammed pools,or glides were selected for sampling. Effective elec-trofisher settings were optimized, recorded, andrepeated throughout units electrofished to stan-dardize effort. Initially, we conducted several test

153

PACIFIC LAMPREY AMMOCOETE HABITAT UTILIZATION

Table 1. Habitat unit classification for sampling sites in Red River,Idaho.

Habitat type Subtype

FallsCascadesRapids

TypicalBoulder Bedrock

RifflesTypicalPocket-water

GlidePool

Lateral scourStraight scourPlungeDammed

Alcove

Table 2. Substrate classification for sampling sites in Red River,Idaho.

Substrate type Size (mm)

Silt/organic 0.004–0.062

Fine sand 0.062–0.50

Coarse sand 0.50–2.0

Fine gravel 2.0–8.0

Medium gravel 8.0–16.0

Coarse gravel 16.0–64.0

Cobble 64.0–256.0

Small boulder 256.0–512.0

Large boulder >512.0


samplings with multiple passes, but few if anyadditional ammocoetes were captured after thefirst pass. Thereafter, we utilized a one pass tech-nique moving at a slow rate (1.1–3.0 m/min) tomaximize catch. Pacific lamprey ammocoete abun-dance, expressed as ammocoetes/100 m2, was esti-mated by the number of ammocoetes captured bythe area electroshocked.

Pacific lamprey ammocoete habitat preferenceswere determined by analysis of variance (ANOVA).The natural log average density values for individ-ual habitat types were analyzed to determine ifmean densities were different. The density values inlateral scour pool and straight scour pool habitatswere analyzed as a single habitat type (scour pool)due to similarity in the structure of the two habi-tats. Similarly, the typical riffle and riffle withpocket water ammocoete densities were combinedand analyzed as “riffles.” The single alcove habitatunit density was not included in the ammocoetedensity and habitat type relationship becauseANOVA requires an average value.

The ammocoete density relationship to streamhabitat unit parameters (water velocity, maximumdepth, canopy shade, and substrate type) wereassessed through linear regression and modeledwith best fit stepwise multiple regression. Linear

and multiple regression F-tests (α = 0.05) were uti-lized to determine the strength of the relationshipbetween ammocoete density and stream parame-ters. Due to minimal sample size (1), the densitiesobtained in alcove habitat and an irregular ammo-coete density of one outlier scour pool wereexcluded from linear and multiple regressions ofthe velocity, substrate, maximum depth, andcanopy cover parameters.

Initial sampling in Red River indicated ammo-coete density differences were minimal when com-paring individual substrate size-classes (Table 2);therefore, substrates were combined into threeclassifications: “coarse” (large boulder, small boul-der, cobble), “medium” (course gravel, mediumgravel, fine gravel), and “fine” (coarse sand, finesand, silt/organic), and analyzed in relationship toammocoete densities in the corresponding units.

Results

Pacific lamprey ammocoetes were captured acrossthe entire range of Red River habitat types sampled.The greatest density was observed in alcove habitattype (Table 3), and the greatest total number ofammocoetes was captured in scour pools. Ammo-coete densities averaged 952% and 690% greater in

154

CLAIRE ET AL.

Table 3. Habitat utilization of Pacific lamprey ammocoetes in randomly sampled units, Red River and selected units in the South ForkClearwater River, Idaho, 2000–2002.

Total Total Totallamprey area fished time fished Density Catch per unit of effort

Habitat type captured m 2 (min) (Lamprey/100 m 2) (Lamprey/min)

Scour pools 342 1,283.4 1461 25.7 0.20(n = 9) (±87.7) (±0.65)

Riffle 15 603.5 726 4.3 0.05(n = 4) (±11.3) (±0.13)

Riffle with pockets 57 1,269.8 825 4.5 0.06(n = 5) (±12.2) (±0.15)

Rapids with boulders 10 357.3 305 2.1 0.03(n = 3) (±3.6) (±0.04)

Alcove 19 7.5 20 253.3 0.95(n = 1) (na) (na)

Totals 443 3,521.5 3,337 – –

Average 12.6 0.13


scour pool habitat than in rapids and riffle habitats,respectively. Scour pool densities ranged from 0.8 to152.3 ammocoetes/100 m2, riffle densities from 0.0to 14.9 ammocoetes/100 m2, and the rapids from 0.0to 3.3 ammocoetes/100 m2.

Although alcove habitat is rare in Red River(0.56% of total stream habitat from rkm 0.0 torkm 7.0), ammocoete density was greater than inother habitat types. The single alcove sampled at

rkm 0.9 yielded a density of 253.3 ammo-coetes/100 m2.

Ammocoete densities were similar for compara-ble velocity habitat types (Figure 2). Average veloc-ities ranged from 0.47 m/s in riffle habitat to 0.050m/s in the alcove habitat unit (Table 4). Ammocoetedensity decreased with increasing stream flowvelocity in units sampled (R2 = 0.477; Figure 3).The inability to identify exact individual emergence

locations prevented calculation of siteof emergence flow velocity. Maximumdepths for sampled units averagedfrom 0.77 m in scour pool habitat to0.40 m in the alcove habitat. The rela-tionship between ammocoete densityand depth was not significant (Figure4). Coarse substrate averages rangedfrom 69.5% in rapids habitats to32.0% in alcove habitat (Table 4).Medium substrate ranged from 29.3%in riffle habitat to 10.0% in alcovehabitat. Fine substrate averaged from58.0% in alcove habitat to 10.5% inrapids habitat. The greatest averagecanopy shade measured was 33.0% forthe alcove habitat, and the minimumwas 9.8% in riffle habitat (Table 4).

Red River substrates within theknown ammocoete distribution (rkm

155


Figure 2. Pacific lamprey ammocoete average densities (95% confidenceintervals) in Red River, Idaho, 2000–2002. (LSP = Lateral Scour Pool, RIF= Riffle Typical, RIP = Riffle with Pockets, RBB = Rapids with Boulders,ALC = Alcove, [n = 1]).

250

200

150

100

50

0LSP RIF RIP RBB ALC

Habitat Type

Amm

ocoe

tes/

100m

2

Table 4. Pacific lamprey ammocoete density and habitat unit parameter averages (95% confidence intervals in parenthesis), Red River,Idaho, 2000–2002.

VelocityDensity @ substrate Max. depth Substrate (%) Canopy

Habitat unit (Lamprey/100 m 2) (cm/s) (m) Coarse Medium Fine shade (%)

Scour pools 25.7 26 0.77 61.4 22.6 16.0 21.3(n = 9) (±87.7) (±17) (±0.30) (±20.5) (±15.4) (±11.9) (± 26.2)

Riffle 4.3 47 0.60 53.4 29.3 17.3 9.8(n = 4) (±11.3) (±41) (±0.37) (±17.8) (±14.4) (±9.2) (±12.4)

Riffle with pockets 4.5 29 0.70 65.1 21.2 13.7 16.1(n = 5) (±12.2) (±24) (±0.29) (±10.5) (±10.1) (±6.0) (±26.2)

Rapids with boulders 2.1 41 0.62 69.5 20.0 10.5 20.3(n = 3) (±3.6) (±28) (±0.29) (±10.4) (±9.0) (±8.7) (±32.4)

Alcove 253.3 5 0.40 32.0 10.0 58.0 33.0(n = 1) (na) (na) (na) (na) (na) (na) (na)


0.0–7.5) are largely boulder and cobble (Table 2).The ammocoete density decreased with increasingcoarse substrate (R2 = 0.263), increased withincreasing medium substrate (R2 = 0.333), andincreased with increasing fine substrate percentage(R2 = 0.157; Figures 5, 6, and 7).

Red River drains in a northwesterly directionfrom Red Horse Creek (rkm 9.0) to the mouth. Thesolar input reaching the riparian and upslopecanopy in the Red River stream section rkm 0.0–7.5is comparable throughout due to streamflow direc-tion (NNW) and similar upslope topography.Therefore, the resultant relative solar input to thestream is primarily a function of riparian canopy.Ammocoete density in Red River increased with

increasing habitat unit canopy shadepercentage (Figure 8).

Analysis of variance of ammocoetedensities in pool, riffle, and rapids habi-tat in sampled units was insignificant(F = 2.83; P = 0.0854). However, analy-sis of the habitat unit (pool, riffle, andrapids) mean densities with Fisher’sleast significant difference indicated thescour pool mean was modestly differ-ent from the riffle mean (P = 0.0616)and the rapids mean (P = 0.0709).

A linear regression of ammocoetedensities by habitat unit with averageflow velocity yielded a significantresponse (P = 0.0007; Table 5). Linearregression of ammocoete density andcanopy cover (shade) was also signifi-cant (P = 0.0280; Table 5). The coarsesubstrate and medium substrate ammo-coete density responses were significantas well (P = 0.0208, P = 0.0070), respec-tively (Table 5). Multiple regression ofammocoete densities and streamparameters produced a best fit modelwith velocity, coarse substrate, andcanopy cover (P = 0.0080, P = 0.0350,and P = 0.0400 respectively; Table 6).

Temperatures in Red River com-monly reach 20.0°C or higher duringthe summer (Figure 9). Maximumstream temperature obtained at rkm5.0 in 2000 was 26.7°C. Substrate tem-

peratures were an average of 2.2°C (P < 0.05) coolerthan stream temperatures when measured 9 August2001. Stream temperature increased with distancefrom the source. However, daily stream temperaturesin the rkm 0.0–7.5 reach during August and Septem-ber were comparable, with slightly higher tempera-tures near the mouth. The inability to isolate ammo-coetes in reaches with different temperaturesprecluded analysis of ammocoete densities andstream temperature relationship.

Discussion

Red River substrates (rkm 0.0–7.5) were predomi-nantly boulder and cobble in the lower reach; how-

156

CLAIRE ET AL.

10

9

8

7

6

5

4

3

2

1

00.30

Depth (m)

(In) L

ampr

ey/1

00m

2

0.40 0.50 0.60 0.70 0.80 0.90 1.00

Figure 4. Natural logarithm of Pacific lamprey ammocoete densities andhabitat unit maximum depth, Red River, Idaho, 2000–2002.

10

9

8

7

6

5

4

3

2

1

00.0

Flow Velocity (cm/s)

(In) L

ampr

ey/1

00m

2

10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0

Figure 3. Natural logarithm of Pacific lamprey habitat unit ammocoetedensities and average habitat unit flow velocity in Red River, Idaho,2000–2002.


ever, extensive areas of finer substrateswere present. Random selection ofsamples excluded several fine sub-strate dominated scour pool habitatunits from sampling. It is possible thatammocoete densities in finer substratedominated reaches of Red Riverwould differ from the randomly sam-pled sites. In 2002, a pool at rkm 6.0yielded an ammocoete density of152.3/100 m2, which is 480% greaterthan the maximum pool densityobtained in 2000 and 2001.

Studies in the Salmon River ofBritish Columbia and other river sys-tems (Pletcher 1963; Close et al. 1995)have indicated that Pacific lampreyammocoetes prefer low stream veloci-ty habitat. Ammocoete densities inRed River were greater in pool habitatscompared to riffle and rapids habitats,supporting the findings of Pletcher(1963). Even though the ANOVAanalysis of ammocoete densities andhabitat type relationship was not sig-nificant (P = 0.0854), the habitat typeand densities in Red River are modest-ly correlated. Ammocoete densities inRed River scour pool habitat rangedfrom 0.8/100 m2 to 152.3/100 m2. Theammocoete densities in other habitatsalso ranged widely, reducing thepotential to obtain statistical signifi-cance. Stream parameters of substrate,canopy shade, velocity, and depthwere quite variable between and with-in habitat units, thereby influencingammocoete density at any sample site.

The range of habitats sampledincluded scour pools, riffles, rapids,and alcove habitats. Other habitattypes (glides, rapids over bedrock,and dammed pools) were present inRed River but were extremely rare andnot sampled due to random selection.

Pletcher (1963) indicated the abilityof all Pacific lamprey ammocoetes (age0 or greater) to burrow was impacted at

157


6

5

4

3

2

1

040

Coarse Substrate %

(In) L

ampr

ey/1

00m

2

50 60 70 80 90

Figure 5. Natural logarithm of Pacific lamprey ammocoete densities andpercentage of stream habitat unit coarse substrate (large boulder, smallboulder, and cobble), Red River, Idaho, 2000–2002. All sites contain atleast 42% coarse substrate.

6

5

4

3

2

1

00

Medium Substrate %

(In) L

ampr

ey/1

00m

2

5 10 15 20 25 30 35 40

Figure 6. Natural logarithm of Pacific lamprey ammocoete densities andpercentage of stream habitat unit medium substrate (coarse gravel,medium gravel, and fine gravel) in Red River, Idaho, 2000–2002.

6

5

4

3

2

1

00

Fine Substrate %

(In) L

ampr

ey/1

00m

2

5 10 15 20 25 30

Figure 7. Natural logarithm of Pacific lamprey ammocoete densities andpercentage of stream habitat unit fine substrate (coarse sand, fine sand,silt/organic) in Red River, Idaho, 2000–2002.


stream velocities exceeding 31.0 cm/s and burrow-ing time of ammocoetes increased with increases inwater velocity in the Salmon River, British Colum-bia. He noted the burrowing ability of age-0 ammo-

coetes was reduced by small increasesin velocity (15.0 cm/s) and ammo-coetes are believed to seek pools afterhatching in response to reduced veloc-ities. Ammocoete densities in RedRiver decreased with increasing flowvelocity. Ammocoetes were capturedin stream habitats with greater veloci-ties; however, ammocoetes in thesehabitats were found in margin pocketswhere velocity is reduced. Shorelineboulder created calm water pockets inRed River commonly supported high-er ammocoete densities.

Pletcher (1963) observed Pacificlamprey ammocoetes using deeper

water in summer and shallower water in the winterin the Salmon River of British Columbia, Canadaand larger ammocoetes (age 2 or older) used deep-er water in one river system. Ammocoete densities

158

CLAIRE ET AL.

Table 5. Linear regression of Pacific lamprey habitat unit ammocoete densities(ln density = ln (density/m2 x 1,000 + 1)) and stream habitat parameters, Red River, Idaho, 2000–2002.

Source DF Type III SS Mean square F value P value R-square

Velocity 1 22.84545 22.84545 16.48 0.0007 0.4779

Substrate coarse 1 12.57106 12.57106 6.42 0.0208 0.2630

Substrate medium 1 15.94397 15.94397 9.01 0.0070 0.3335

Canopy shade 1 9.18372 9.18372 5.71 0.0280 0.2515

Table 6. Stepwise multiple regression of Pacific lamprey habitat unit ammocoete densities (ln density = ln (density/m2 x 1,000 + 1)) and stream habitat parameters in Red River, Idaho, 2000–2002.

Source DF Type III SS Mean square F, t value P > F, t R-square

Model 3 23.06964 7.68988 8.58* 0.0015* 0.6319

Error 15 13.44007 0.89600

Corrected total 18 36.50972

Velocity 1 3.01** 0.0080** –

Substrate coarse 1 2.30** 0.0350** –

Canopy shade 1 2.25** 0.0400** –

* F value.** t value.

10

9

8

7

6

5

4

3

2

1

00

Stream Canopy Cover (Shade %)

(In) L

ampr

ey/1

00m

2

10 20 30 40 50

Figure 8. Natural logarithm of Pacific lamprey ammocoete densities andpercentage of stream habitat unit riparian canopy shade in Red River,Idaho, 2000–2002.


in Red River were greater in depths less than 0.3 mwhen habitats were sampled in August and Sep-tember. Ammocoete density and stream depth uti-lization relationships in Red River during wintermonths are unknown. Habitat sampling in RedRiver is restricted to minimum flow, ice freemonths (primarily July, August, and September).Larger ammocoetes (>120 mm total length) werecaptured in the range of Red River depths withoutnoticeable patterns. Ammocoete density and habi-tat unit maximum depth relationship in Red Riveris generally weak (R2 = 0.029); however, habitatunit maximum depth reflects the attributes of theentire habitat unit, not the specific locationsammocoetes occupy. The difficulty with visuallyidentifying individual ammocoete emergence sitesprecluded analysis of single ammocoete depth uti-lization in Red River.

Substrate preference of Pacific lamprey ammo-coetes in the Salmon River, British Columbia waspredominantly mud and silt for age-1 ammocoetes(Pletcher 1963). Older ammocoetes were foundpredominantly in sand and leaf substrates. Ham-mond (1979) found most ammocoetes in sand, silt,or clay substrates in the Potlatch River, Idaho.Beamish and Lowartz (1996) found that Americanbrook lamprey L. appendix density was positively

correlated with the amount ofmedium fine sand and organic mat-ter in the substrate. The substratesin the Red River drainage are pre-dominantly gravels, boulders, andsand with lesser amounts of silt andorganic material. Ammocoete densi-ties in Red River were inversely cor-related to substrate size. Ammocoetedensities noticeably increased insites with a greater percentage offiner substrates (sand, silt/organic);however, the regression strengthwith fine substrate was impacteddue to the influence of other streamhabitat parameters (flow velocity,depth, canopy cover, etc.) on ammo-coete habitat selection. Ammocoeteswere captured in substrates of allsize-classes but were captured inmodest numbers from sites with

large and small boulders dominating the substrate.Ammocoetes in Red River commonly emergedfrom substrate gaps in cobble and small boulderareas. Pletcher (1963) indicated that Pacific lam-prey ammocoete burrowing ability is impacted bysubstrate size. Generally finer substrates are con-sidered the preferred Pacific lamprey ammocoetesubstrate (Close et al. 1995; Pletcher 1963; Ham-mond 1979). However, if concealment is providedwith larger substrates and the interstitial spaces aresufficient for ammocoetes to penetrate, or finersubstrate pockets exist in coarse substrate sites, it islikely that they are adequate for limited rearing.

Pacific lamprey ammocoete rearing temperaturerequirements are not fully known, but ammocoetesare generally found in cold waters (Close et al.1995); however, tolerance may exceed 25°C (Mallat1983). In a laboratory environment, Meeuwig et al.(2005) found significant decreased survival duringincubation at water temperature of 22°C than atlower temperatures. They also found that abnormaldevelopment is significantly more pronounced atthe higher water temperatures in initial stages of lar-val development. Holmes and Youson (1998) foundthat the percentage of sea lamprey ammocoetes thatmetamorphosed was maximized with water tem-perature of 21.0°C and inhibited with higher tem-

159


30

25

20

15

10

5

0

Date

Nov.1,

1999

Dec. 1

, 199

9

Jan.

1,20

00

Feb.

1,20

00

Mar.1,

2000

Apr.1,

2000

May 1,

2000

Jun.

1,20

00

Jul. 1

, 200

0

Aug.

1,20

00

Sep.

1,20

00

Tem

pera

ture

°C

Figure 9. Red River daily maximum/minimum stream temperatures at rkm5.0 from 5 November 1999 to 14 September 2000, Red River, Idaho. Tempera-tures ranged from –0.6°C (min) to –0.1°C (max) 21 November 1999 to 15March 2000. Data logger removed for download 10 May 2000–24 May 2000.


peratures. Red River stream temperatures areknown to reach or exceed 26.7°C. Temperaturemonitoring in the Red River subbasin infers thatPacific lamprey ammocoetes and macropthalmiaare capable of surviving with stream temperaturesin excess of 20.0°C; however, the duration ammo-coetes are able to withstand high stream tempera-tures is unknown. Red River substrate temperaturescommonly exceed sea lamprey metamorphosis tem-perature preference of 21°C for a limited time peri-od in July and August; however, it is unknown ifPacific lamprey ammocoetes are negatively impact-ed. Substrate temperatures were cooler than streamtemperatures sampled, but it is unclear whetherPacific lamprey benefit.

Limited riparian canopy cover is often cited as afactor impacting stream production (Jackson et al.1996). Elevated stream temperature due to removalof riparian canopy in coldwater species watershedsresults in limited production of a number of nativeaquatic species (U.S. Forest Service 1998); however,it is unclear if temperature extremes in Red Riverimpact Pacific lamprey population productivity.Ammocoetes are adept at detecting light. Pacificlamprey movements predominantly occurred atnight (presumably due to light or predator avoid-ance) in the Salmon River in British Columbia(Pletcher 1963). Ammocoetes are predominantlycaptured from dusk to sunrise in the rotary screensmolt traps operated in Red River. Maximum streamtemperature tolerances likely limit Pacific lampreyproduction in watersheds where human removal ofriparian canopy results in excessive insolarization ofstreams (Close et al. 1995; Jackson et al. 1996; Jack-son et al. 1997). Low-angle shading was identified asan important parameter for Australian lamprey (alsoknown as pounched lamprey) Geotria australisammocoetes (Potter et al. 1986). The Pacific lampreyammocoete density and riparian canopy relation-ship in Red River was stronger (R2 = 0.251) thancompared to the maximum depth (R2 = 0.029) andfine substrate relationships (R2 = 0.157). Ammo-coetes were captured in greater numbers repeatedlyunder overhanging hardwood riparian vegetation;however, whether the increased densities were inresponse to decreased microhabitat temperatures orlight intensity is unknown.

Pacific lamprey is a species critically linked to the

ecological function of the Snake River and SouthFork Clearwater River biological communities. Theecological interaction of Snake River Pacific lam-prey populations and other riverine species isthought to contribute to Snake River basin overallaquatic productivity. Pacific lamprey ammocoetesprovide Snake River basin white sturgeon Acipensertransmontanus populations with an important foodsource, which potentially contributes to Snake Riverwhite sturgeon population productivity (Galbreath1979). Pacific lamprey adults are a source of marinederived nutrients in the Snake River basin. Aquaticand avian predator utilization of ammocoetes andmacropthalmia potentially results in reduced preda-tion impact to out-migrating juvenile salmon andsteelhead Oncorhynchus mykiss in the lower SnakeRiver migrational corridor. Pacific lamprey, Chi-nook salmon O. tshawytscha, and summer steelheadrear in Snake River basin stream habitats; however,the ecological relationship interactions of the threespecies in the basin are little known. The decline ofPacific lamprey adult upstream migrants to theSnake River basin is undoubtedly in part a functionof instream migration corridor mortality, hydro-electric upstream passage impediments, and rearingstream habitat degradation (Close et al. 1995; Jack-son et al. 1996; Vella et al. 1997). Pacific lampreypersistence in the South Fork Clearwater Riverdrainage following the installation of Harpster Damindicates the resilient and enduring character of thespecies. Knowledge of Columbia River basin Pacificlamprey population ecology, subbasin distributions,and species habitat requirements is currently limit-ed. Increased Pacific lamprey habitat requirementdata will further augment the potential to intensive-ly manage the species. Habitat utilization samplingin the South Fork Clearwater River drainage sug-gests that maintenance of remaining preferred habi-tats in the South Fork Clearwater River and its trib-utaries, including Red River, is paramount to ensurethat rearing conditions are adequate for the speciesto continue to inhabit the drainage.

Acknowledgments

This research was supported by the BonnevillePower Administration (Project Number 2000-02800), the Idaho Department of Fish and Game,

160

CLAIRE ET AL.


and the Cottonwood District Office of the Bureau ofLand Management. We would like to acknowledgeBonneville Power Administration COTR, DeborahDocherty, for her continued support of the project.We would also like to recognize field techniciansAnne Peterson, Rebecca Repp, Terry Douglas, andKyle Steele for their assistance in collection of fielddata. Dale Everson of the University of Idaho pro-vided statistical expertise and advice. Appreciationis also given to two anonymous reviewers for theirpositive critical review of the document. Use oftrade or firm names in this document is for readerinformation only and does not constitute endorse-ment of a product or service by the state of Idaho.

References

Beamish, R. J., and C. D. Levings. 1991. Abundance andfreshwater migrations of the anadromous parasitic lam-prey, Lampetra tridentata, in a tributary of the FraserRiver, British Columbia. Canadian Journal of Fisheriesand Aquatic Sciences 48:1250–1263.

Beamish, F. W. H., and S. Lowartz. 1996. Larval habitat ofAmerican brook lamprey. Canadian Journal of Fisheriesand Aquatic Sciences 53:693–700.

Bond, C. E. 1996. Biology of fishes, 2nd edition. SaundersCollege Publishing, Fort Worth, Texas.

Close, D. A., M. Fitzpatrick, H. Li, B. Parker, D. Hatch, and G.James. 1995. Status report of the Pacific lamprey Lampe-tra tridentata in the Columbia River basin. U.S. Depart-ment of Energy, Bonneville Power Administration, Port-land, Oregon.

Creaser, C. W., and C. S. Hann. 1929. The food of larval lam-preys. Michigan Academy of Sciences 10:433–437.

Galbreath, J. 1979. Columbia River colossus, the white stur-geon. Oregon Wildlife 1979:3–8.

Hammond, R. J. 1979. Larval biology of the Pacific lamprey,Entosphenus tridentate (Gairdner), of the Potlatch River,Idaho. Master’s thesis. University of Idaho, Moscow.

Holmes, J. A., and J. H. Youson. 1998. Extreme and optimaltemperatures for metamorphosis in sea lampreys. Trans-actions of the American Fisheries Society 127:206–211.

Jackson, A. D., P. D. Kissner, D. R. Hatch, B. L. Parker, M. S.Fitzpatrick, D. A. Close, and H. Li. 1996. Pacific lamprey

research and restoration annual report 1996. U.S.Department of Energy, Bonneville Power Administra-tion, Portland, Oregon.

Jackson, A. D., D. R. Hatch, B. L. Parker, D. A. Close, M. S.Fitzpatrick, and H. Li. 1997. Research and RestorationAnnual Report 1997. U.S. Department of Energy, Bon-neville Power Administration, Portland, Oregon.

Mallat, J. 1983. Laboratory growth of larval lampreys Lam-petra entosphenus tridentata Richardson at different foodconcentrations and animal densities. Journal of FisheriesBiology 22:293–301.

Meeuwig, M. H., J. M. Bayer, and J. G. Seelye. 2005. Effects oftemperature on survival and development of early lifestage Pacific and western brook lampreys. Transactionsof the American Fisheries Society 134:19–27.

Overton, K. C., S. P. Wollrab, B. C. Roberts, and M. A. Radko.1997. R1/R4 (Northern Intermountain Regions) fish andfish habitat standard inventory procedures handbook.U.S. Forest Service, General Technical Report INT-GTR-346, Boise, Idaho.

Platts, W. S., W. F. Megahan, and G. W. Minshall. 1983. Meth-ods for evaluating stream riparian and biotic conditions.U.S. Forest Service General Technical Report INT-138,Boise, Idaho.

Pletcher, T. F. 1963. The life history and distribution of lam-preys in the salmon and certain other rivers in BritishColumbia, Canada. Master’s thesis. University of BritishColumbia, Vancouver.

Potter, I. C., R. W. Hilliard, J. S. Bradley, and R. J. McKay.1986. The influence of environmental variables on thedensity of larval lampreys in different seasons. Oecologia70:433–440.

Richards, J. E., and F. W. H. Beamish. 1981. Initiation of feed-ing and salinity tolerance in the Pacific lamprey Lampe-tra tridentata. Marine Biology 63:73–77.

Scott, W. B., and E. J. Crossman. 1998. Freshwater fishes ofCanada. Fisheries Research Board of Canada Bulletin184, Ottawa.

Simpson, J. C., and R. L. Wallace. 1982. Fishes of Idaho. Uni-versity of Idaho Press, Moscow.

U.S. Forest Service (U.S. Department of Agriculture). 1998.South Fork Clearwater River landscape assessment. Vol-ume I. U.S. Forest Service, Nez Perce National Forest,Grangeville, Idaho.

Vella, J. J., L. C. Stuehrenberg, and T. C. Bjornn. 1997. Migra-tion patterns of Pacific lamprey Lampetra tridentata in thelower Columbia River 1997. Annual Report of Research.U.S. Army Corps of Engineers, Portland, Oregon.

161



165

Conservation efforts almost always focus onrare species, particularly those with limitedgeographic distributions or those that face

immediate threat of extinction. Classic examples ofsuch triage conservation include past attempts topreserve rare species like California condorGymnogyps californianus, black-footed ferretsMustela nigripes, and whooping cranes Grus amer-icana, as well as ongoing efforts to protect taxa likegolden toads Bufo periglenes or desert pupfishesCyprinodon diabolis and C. macularius, species thatnow stand at the verge of extinction. Against thisbackdrop, it is not surprising that numericallyabundant, widespread species receive very little

attention from conservation biologists or fisheriesmanagers charged with protecting biologicalresources. Yet, failure to view such species as targetsof conservation may ultimately constitute a greatercrisis than the ongoing loss of already rare anddepleted taxa.

Common, widespread species are prone to con-servation neglect on at least two fronts. First, littleattention is focused on delineating conservationunit boundaries within species that are numerical-ly abundant. The irony of this is that withoutknowledge of the distinct evolutionarily unitswithin common species, such taxa may in fact bemore susceptible to losing unique evolutionary

What the Status of Utah Chub Tells Us about Conserving Common, Widespread Species

ABSTRACT The fundamental charge of conservation biology is to preserve biological diversity. Yet, efforts to

accomplish this goal have focused too narrowly on reversing the slide toward extinction in already threatened or

endangered species. In this review, we argue that conservation biologists and fisheries managers should broad-

en their vision to include efforts to preserve the ecological and evolutionary processes that ultimately give rise to

new biodiversity. Our view is based upon the simple observation that biological diversity is a function of both the

rate at which new taxa originate as well as the rate at which established taxa are lost to extinction. Efforts to stem

extinction that fail to maintain the ecological and evolutionary processes of speciation are ultimately unsustain-

able. We suggest that common, widespread species are particularly important to the origin of new diversity and

argue that conservation biologists should pay particular attention to the evolution of diversity within such species.

We illustrate several key points to this argument using the desert minnow, Utah chub Gila atraria, as a model sys-

tem. In particular, we show that conservation efforts in common species must focus on clearly delineating con-

servation unit boundaries and that particular care should be paid to unique ecological and evolutionary diversity

within such species. We also show the importance of understanding and conserving the range of ecological and

evolutionary interactions that are common hallmarks of abundant and widespread taxa. We conclude our review

by suggesting several specific areas of future research in Utah chub that would help more clearly define conser-

vation and management priorities in this species.

Jerald B. Johnson and Mark C. Belk



JERALD B. JOHNSON Evolutionary Ecology Laboratories, Department of Integrative Biology and Monte L. Bean Life Science Museum, BrighamYoung University, Provo, Utah 84602, USA. Corresponding author: [email protected]

MARK C. BELK Evolutionary Ecology Laboratories, Department of Integrative Biology, Brigham Young University, Provo, Utah 84602, USA

NatAmerFishWAPart2-8 5/25/07 1:48 PM

diversity than their rare, but better-studied coun-terparts. Unfortunately, rare components of evolu-tionary diversity are easily overlooked in commonspecies. Second, very little attention has focused onidentifying and conserving the variety of ecologicaland evolutionary processes that typify geographi-cally widespread, abundant species. Unlikeextremely rare taxa, common species tend to occu-py a variety of ecological niches. Conserving thisecological diversity is important for several rea-sons, particularly when species play different rolesin different communities. Perhaps more impor-tantly, ecologically diverse species tend to havepopulations that evolve under very different selec-tive regimes. Such selection, when coupled withgenetic isolation and additive genetic variation,ultimately gives rise to local adaptation and mayplay a key role in speciation. Stated simply, com-mon species are often key generators of new bio-logical diversity. Yet, the idea of conserving ecolog-ical and evolutionary processes within species hasbeen largely neglected by conservation biologistsand is seldom considered in fisheries management.Such neglect is unfortunate because conservationprograms that focus solely on minimizing the riskof extinction, while failing to protect the ecologicaland evolutionary processes that give rise to newdiversity are in the long run unsustainable.

Western North America fishes face both kinds ofthreats: (1) risk of demographic extinction, and (2)ongoing loss of ecological and evolutionary diversi-ty, including the processes that sustain such diversi-ty. The battle against extinction is well underway formany rare western fishes (Minckley and Deacon1991). Organizations such as the Desert FishesCouncil, the Western Native Fishes Committee ofthe American Fisheries Society Western Division,and several private and public entities are activelyengaged in the recovery of rare and endangered fishspecies in western North America. The EndangeredSpecies Act, International Union for Conservationof Nature and Natural Resources (IUCN) Red List,and several other laws and regulations have beeninstrumental in motivating these efforts. By con-trast, efforts to identify and preserve ecological andevolutionary processes within common, wide-spread fish species are only just beginning, and nolegislation exists to fund or drive this work. Yet, pre-

serving evolutionary diversity with its attendantecological and evolutionary processes within abun-dant species should be a critical component of con-servation management.

In this review, we focus on the desert minnowUtah chub Gila atraria as a case study to explorethe causes and consequences of conservation neg-lect toward common, nongame fish species. Wefirst describe the natural history of Utah chub andreport on the current distribution and status ofnative chub populations. Next, we confront inUtah chub the kind of conservation neglect that istypical of most widespread, numerically abundantspecies: failure to delineate and conserve uniqueevolutionary units, and failure to identify and con-serve the ecological and evolutionary processesthat give rise to new biological diversity. We con-clude our review by examining fundamental gapsin our current knowledge about Utah chub andoutline several promising areas for future researchthat could help guide the conservation and man-agement of common species.

Natural History, Distribution, and Status of Utah chubNatural history and distribution

Utah chub belongs to a clade of cyprinid fishes col-loquially referred to as the “western minnows”(Figure 1). This clade contains several endangeredand threatened species, including humpback chubG. cypha, roundtail chub G. robusta, and tui chubG. bicolor (also known as Siphatales bicolor), as wellas several common species native to western NorthAmerica (Simons and Mayden 1998). Althoughfossil evidence for Utah chub dates only to the LatePleistocene (Smith et al. 1968; Broughton 2000),molecular data clearly show that this species origi-nated several million years ago, sometime duringthe Late Pliocene (Johnson 2002). Hence, likemany common and widespread species, Utah chubhas had a considerable amount of time to diversifyacross its range.

Utah chub is ubiquitous in aquatic environmentsthroughout the Bonneville basin and upper SnakeRiver of Western North America (Figure 2). Thisnative range reveals a tight link between the currentdistribution of chub populations and the recent

166

JOHNSON AND BELK


hydrological and geological history ofthis region. For example, several biolo-gists have postulated that the similarityin fish fauna between the Bonnevillebasin and the upper Snake River is theresult of late Pleistocene flooding ofLake Bonneville into the upper SnakeRiver basin (Hubbs and Miller 1948;Minckley et al. 1986). Geological evi-dence indicates that such a floodingevent occurred around 18,000 yearsbefore present (Oviatt 1997). A recenttest of this idea in Utah chub usingmolecular data to compare populationsfrom these two regions revealed thatBonneville basin chub have in fact beenintroduced to the upper Snake River ona timetable consistent with geologicalpredictions (Johnson 2002). Similarly,patterns of genetic isolation among

Utah chub populations within the Bonneville basinand within the upper Snake River drainage are com-pletely consistent with changing hydrological condi-tions in these regions over the past 30,000 years(Johnson 2002). Hence, the distribution of Utahchub appears to have been largely contingent upongeological events and upon repeated cycles of flood-ing and fragmentation that characterize climatechange in the Great Basin region.

Status of Utah chub

Utah chub has long been viewed as a commonspecies in the Bonneville basin and upper SnakeRiver, both in terms of its abundance and in terms

167

CONSERVING COMMON, WIDESPREAD SPECIES

Figure 2. Natural geographic range of Utah chub throughout the Bonneville basin and upper Snake River drainage of western North America. Circles denote populations of Utah chub sampled for variousstudies referenced in this review (black = Bonneville basin populations,open = Bear River and upper Snake River populations, gray = introduced populations).

Figure 1. Phylogenetic relationships amongmembers of the western clade of NorthAmerican minnows. Strict consensus oftwo shortest trees generated by maximumparsimony (modified from Figure 7 inSimons and Mayden 1998) using mtDNAsequence data from 12-s and 16-s riboso-mal genes. Numbers at nodes show boot-strap values greater than 50% based on1000 bootstrap replicates. Species denotedwith asterisk are currently listed as eitherthreatened or endangered under the U.S.Endangered Species Act.


of its lack of economic and social importance.Unfortunately, a systematic survey of the currentstatus of this species is also lacking (a point dis-cussed in detail below). However, we do have awell-documented set of commentaries on Utahchub over the past 100 plus years, with input fromichthyologists, fisheries managers, scientists, andeven the public at large. Such a history, althoughnot particularly helpful in a scientific context, isrevealing with regard to the evolving perception ofthis species and, in many ways, reflects attitudestoward other common nongame minnows in west-ern North America.

Utah chub was described first by Girard in1856. However, in the decades following, severaladditional populations of Utah chub were mistak-enly also described as new species. In fact, by 1936,Utah chub had been the recipient of no fewer than15 distinct taxonomic names (synonyms; Tanner1936). One reason for this proliferation of mistak-en taxonomy is that many Utah chub populationsare quite distinct from one another with respect toseveral key traits used in alpha taxonomy, includ-ing morphology, coloration, and meristic charac-ters (Figure 3). What this early work indicates isthat Utah chub is a remarkably diverse species andthat attempts to cast the species in general termsare not only biologically inaccurate, but also poten-tially quite misleading. Hence, efforts to describebiology and the status of this speciesmust focus on units smaller than thespecies at large.

Beginning in the early 1900s andcontinuing to present day, Utah chubhas been viewed with some con-tempt by fishers and fisheries biolo-gists. The perceived problem stemsfrom the fact that Utah chub is aremarkably good competitor withexotic fish species such as rainbowtrout Oncorhynchus mykiss, particu-larly in unnatural or disturbed envi-ronments such as reservoirs. Fish-eries managers attempting to stockexotic trout in lakes and reservoirsdiscovered early on that Utah chubcould eventually proliferate to thepoint that exotic trout fisheries were

severely diminished. Consequently, Utah chub waslabeled a “rough fish” or a “trash fish” (Olsen 1959;Gaufin 1964), and considerable effort by fisheriesmanagers has focused for decades on ways to erad-icate this native species from game fish environ-ments. Curiously, Utah chub coexists stably and inmoderate numbers with native cutthroat trout inseveral locations and in fact provides an integralpart of the trout diet in these waters.

Most recently, Utah chub is beginning to beviewed as an important model system for under-standing a variety of important questions in ecolo-gy and evolutionary biology. One feature thatmakes this species so compelling as a scientificmodel is that several populations are completelyisolated from one another in remnant springs,lakes, and rivers that were only recently (~10,000years ago) connected to one another by PleistoceneLake Bonneville. This known time since diver-gence, coupled with different replicated ecologicalenvironments, provides an uncommon set of nat-ural experiments that researchers are only nowbeginning to explore.

In brief, although we still know very little aboutthe status of most Utah chub populations, percep-tions about this species are slowly changing. Thevariety of phenotypic variation in this species thatpuzzled early ichthyologists, and that for severaldecades was largely ignored by both scientists and

168

JOHNSON AND BELK

Figure 3. Photographs showing phenotypic variation among individuals rep-resentative of different Utah chub populations. Individuals shown are fromlocations spanning the natural distribution of this species. Panel a: HeartLake, Wyoming; panel b: Big Spring, Utah; panel c: Fish Springs, Utah; andpanel d: Locomotive Springs, Utah.


fisheries managers, is now becoming the focus ofscientific research and is poised to become a majorfocal point of fisheries management and conserva-tion efforts. As these perceptions change, it will beimportant to extend our understanding about thecauses and consequences of diversity among chubpopulations. Here, we describe two key areas thatwarrant particular attention from conservationmanagers: delineating natural evolutionaryboundaries that can be used to frame conservationand management decisions, and understandingand conserving the variety of ecological and evolu-tionary processes that give rise to continued bio-logical diversity within Utah chub. We considereach of these ideas in turn.

Identifying Conservation Units

Delineating conservation units is one of the firststeps to protecting rare or endangered taxa (seeBelk and Johnson 2007) and several biologists havedebated over the most effective way of doing so(Moritz 1994; Bowen 1998). However, these tech-niques are seldom applied to species that are notalready in decline or at risk of extinction. Yet, with-out knowing the boundaries of units that warrantindependent consideration, even common wide-spread species may be prone to losing unique com-ponents of their evolutionary diversity. Identifyingconservation units typically requires two kinds ofdata (Crandall et al. 2000): (1) information on thestructure and degree of genetic isolation amongpotential units, and (2) information on local adap-tation, phenotypic divergence, and ecological nich-es that render potential units distinct from oneanother. Available evidence in both of these cate-gories suggests that Utah chub is composed of sev-eral distinct evolutionary units that require inde-pendent management consideration.

Genetic structuring in Utah chub

Earliest efforts to identify historic relationshipsamong Utah chub populations using neutralgenetic markers were based on allozyme data(Rosenfeld 1991). This work suggested that severalUtah chub populations were genetically distinctfrom one another and that much of this variationcould be accounted for by geographic isolation due

to climate change in the Late Pleistocene. However,given limitations of the allozyme data, no effortswere made to infer from the genetic data the actu-al mechanisms of reproductive isolation over time.Nonetheless, this pioneering work clearly demon-strated that Utah chub is not one large homoge-neous species, but rather is composed of severalgenetically distinct units.

Recent research using DNA sequence data hasfurther broadened our understanding of Utah chubpopulation histories and provides a more usefulframework for delineating conservation unitboundaries. In comparing mtDNA among Utahchub collected from 16 distinct geographic locationsthroughout the Bonneville basin and upper SnakeRiver, Johnson (2002) identified several key aspectsof historic relationships among Utah chub popula-tions. First, Utah chub is composed of two distinctevolutionary clades that separated from one anoth-er several million years ago—these clades show lev-els of sequence divergence (~3.5%) that approacheslevels typically found among sister species. Oneclade occupies the upper Snake River and Bear Riverdrainages; the second occurs throughout the Bon-neville basin with more minor representation in theupper Snake and Bear rivers. The presence of someBonneville basin haplotypes in the upper SnakeRiver drainage is consistent with a Late Pleistoceneintroduction when Lake Bonneville flooded northinto the Snake River system.

A second important conclusion from Johnson(2002) is that the relationship among populationswithin the Bonneville basin is largely concordantwith the Late Pleistocene history of Lake Bon-neville. For example, nested clade analysis (Tem-pleton 1998) revealed evidence for populationrange expansions as Lake Bonneville filled, long-distance colonization and isolation by distanceamong populations consistent with high waterperiods of the lake, and finally fragmentation coin-cident with the recession of the lake. In fact, theseUtah chub data provide the most compelling bio-logical evidence to date corroborating geologicalhypotheses of the hydrological history of LatePleistocene Lake Bonneville (Oviatt 1997).

Finally, the mtDNA shows that there are severalpopulations with unique haplotypes. Althoughsample sizes are relatively small to make firm rec-

169



ommendations, several distinct groups within theBonneville clade appear to be supported. Theseunits include (1) Fish Springs, (2) LelandHarris/Bishop/Locomotive Springs, (3) Beaver/EastFork Sevier River, and (4) Rush Lake/SpringCreek/Mona Spring. Genetic diversity among Utahchub populations in the Bear River and upperSnake River drainages is much lower than that inthe Bonneville basin, and genetic structuring ismuch less pronounced. However, in both regions,several additional populations should be sampledto provide a more accurate depiction of potentialgenetic unit boundaries.

Ecological diversity and local adaptation in Utah chub

We know much less about ecological diversity andadaptation in Utah chub than we know about genet-ic structuring in this species. Nonetheless, severallines of evidence indicate that Utah chub is highlydiverse both in terms of phenotypic variationamong populations and in terms of the wide rangeof environments where chub populations occur. Ashighlighted in Figure 3, Utah chub collected fromdifferent locations can be strikingly distinct fromone another. Whether this phenotypic variation isgenetically based and whether it is systematicallyassociated with apparent genetic unit boundariesare important questions that will need to beaddressed to delineate clear conservation units.Unfortunately, little research has been completed tounderstand either the proximate or ultimate causesof this phenotypic divergence. However, we haveplanned or initiated several studies to explore evo-lutionary divergence among chub populations,including studies that examine differences in bodyshape, gut length, daily activity patterns, fin lengthpolymorphisms, and tolerance to cestode parasites.The outcome of such studies should help clarifywhere evolutionarily distinct populations exist.

Conserving Ecological and Evolutionary ProcessesEcological interactions in Utah chub

One of the most striking features of Utah chub—anda trait typical of many widespread, abundantspecies—is the broad range of ecological environ-

ments where populations persist. Utah chub occur indesert springs, shallow and deep lakes, small creeksand large rivers, reservoirs, and even low-flow,spring-fed marshes. Given this range of habitats, it isnot surprising that individual Utah chub engage in avariety of ecological interactions. Here, we considerthe ecology of Utah chub at three scales: trophic,community, and ecosystem. However, we expect thata careful examination of Utah chub in other ecologi-cal contexts (e.g., behavioral ecology or populationecology) will also be revealing. Also, it is worth notingthat no systematic investigations of ecological inter-actions have been made and that much of what weare able to say is based on anecdotal information orsmall scale experiments.

Utah chub can be viewed in a trophic context asboth predator and prey. Juvenile chub feed prima-rily on zooplankton, and in some populations,zooplankton remain an integral component of thediet through adulthood (Schneidervin 1985). Infact, some have argued that Utah chub overwhelmsgame fishes by more effectively competing for zoo-plankton resources; however, experimental sup-port for this argument is tenuous (Tuescher andLuecke 1996). Stomach contents of Utah chubreveal aquatic insects, crustaceans, diatoms, plantmaterial, snails, and benthic debris (Rees 1935). Inaddition, we have observed that in several loca-tions, including small desert springs, that Utahchub relies heavily on a diet of plant material. Insuch populations, Utah chub appear to have muchlonger intestinal tracts than their carnivorouscounterparts from other sites, presumably inresponse to this low-energy food source. Overall,Utah chub can be viewed as an omnivore. Howev-er, local variation in diet indicates that the trophicimpact of Utah chub could be quite differentamong geographically distinct sites.

Utah chub is also an important prey species andparasite host in some parts of its range. The mostapparent evidence for predation occurs in largelake habitats (e.g., Bear Lake in Utah and Jacksonand Heart lakes in Wyoming) where cutthroattrout Oncorhynchus clarkii prey extensively onUtah chub. We have observed Utah chub up to 200mm in the stomach contents of large cutthroattrout, suggesting that chub are a key component ofthe trout diet. Such observations also demonstrate

170

JOHNSON AND BELK


that Utah chub remain vulnerable to trout preda-tion until they reach relatively large body sizes. Wehave also observed that Utah chub can be host tothe cestode parasite Ligula intestinalis and thatinfection rates can fluctuate dramatically from yearto year (J. B. Johnson, unpublished data).

The role of Utah chub in community andecosystem processes is poorly understood. Howev-er, in several locations, we suspect that this speciesmight be extremely important in each of these eco-logical contexts. For example, in several desertspring populations, Utah chub is the only fishspecies that persists. Loss of this link in aquatic and(perhaps) terrestrial food webs would likely haveimportant ramifications for the kinds of speciesthat would continue to persist. Interestingly, inselect springs of the Bonneville basin, Utah chubstably coexists with least chub Iotichthys phlegethon-tis, a species in danger of extinction, and there maybe strong, dependent interactions between thesetaxa. Similarly, rare freshwater mussels, such as theCalifornia floater Anodonta californiensis, thatrequire a fish intermediate host to complete theirlife cycle may be completely dependent on Utahchub in desert spring environments where this isthe only fish species present. How these communi-ties evolve and what critical links exist among theconstituent taxa are key questions. In terms ofecosystem processes, the sheer abundance of Utahchub in some locations suggests that this speciescould be a critical player in nutrient processing andtransfer. This may be particularly true in depauper-ate environments, but again, work testing theseideas is lacking.

Evolutionary processes in Utah chub

Understanding the evolutionary processes thatgive rise to novel diversity in Utah chub is criticalto managing this species. Our work examining lifehistory evolution among Utah chub populationsprovides one promising line of research to helpachieve this goal. Fossil evidence confirms thatUtah chub and cutthroat trout coexisted in LakeBonneville during the high water mark of LakeBonneville during the Late Pleistocene (Smith etal. 1968). However, as Lake Bonneville dried, sev-eral chub populations became isolated with notrout predators, and several persisted with preda-

tors. We postulated that these two different selec-tive regimes would result in the evolution ofdivergent life history strategies among popula-tions (Johnson and Belk 1999). Our survey ofUtah chub from these different environmentssupported this hypothesis: Utah chub that havecoevolved with predatory trout grow more rapid-ly as juveniles than their counterparts from pred-ator-free environments; this occurs despite thefact that Utah chub from predator environmentstend to occur at higher latitudes with shortergrowing seasons. Utah chub that coevolved withtrout also show a decreased annual reproductiveinvestment, but an increased reproductive lifes-pan relative to predator-free populations. Suchdifferences are presumably maintained as anadaptation to first outgrow the threat of gape-limited predators and to then extend reproductivelifespan once individuals reach safe size-classes.Such drastic differences in reproductive strategiesamong populations, coupled with genetic isola-tion, could provide the basis for early stages ofspeciation (Schluter 2001).

Other evolutionary processes might also beimportant in Utah chub. We expect that a carefulexamination of Utah chub in different selectiveenvironments will reveal considerable evidencefor adaptive evolution in response to a variety ofselective regimes. Whether any of this local adap-tation contributes to reproductive isolation wouldbe a promising area of research. However, Utahchub is also likely to be subject to the effects ofgenetic drift, at least in small populations withrelatively long periods of isolation from largerstocks. Finally, one remarkable feature of westernminnows is that several species appear to haveoriginated as a consequence of hybridization(Dowling and DeMarias 1993; Dowling and Secor1997). The repeated cycles of isolation and rangeexpansion coincident with the long-term hydrol-ogy of the Bonneville basin, coupled with stronglocal adaptation or evolutionary divergence whenchub populations are isolated, may make Utahchub a likely progenitor of new hybrid species. Acareful look at the Late Pleistocene introductionof Bonneville basin clade Utah chub into theupper Snake River drainage may provide someinsight into this possibility.

171



Gaps in Our Current Knowledge and Future Research

Throughout this review, we have highlightedseveral gaps in our current knowledge of Utahchub. Most prominent among these is the rela-tively poor understanding we have of the evolu-tionary causes and ecological consequences ofdiversity among Utah chub populations.Although Utah chub is common and wide-spread, it is incredibly diverse and we have verylittle understanding as to why. We recommendcontinued effort to broaden our understandingof the mechanisms and consequences of pheno-typic variation among populations. This infor-mation will not only yield considerable insightsinto basic science questions, but will also be use-ful in helping identify biologically meaningfulconservation units. Similarly, efforts focused atunderstanding the nature of ecological interac-tions across Utah chub populations will providea framework to understand the consequences oflosing chubs from different kinds of communi-ties and ecosystems. Finally, as we come tounderstand how Utah chub populations natural-ly diversify, we can better focus our conservationefforts at preserving these evolutionary process-es, thus maintaining the engines for generatingnew biological diversity.

There are also several important applied ques-tions that warrant additional attention. First, whydoes Utah chub coexist stably with native cutthroattrout but easily out-competes exotic rainbow trout?What role do disturbance, diet, nutrient loads, orecological interactions play in explaining thisdemographic disparity? One possibility is that Utahchub might out-compete cutthroat trout whenboth species are at small body sizes, but as cutthroattrout grow larger, they are capable of preying uponUtah chub. Such ontogenetic changes in size-struc-tured interactions are thought to explain coexis-tence in bass–bluegill systems (Werner and Gilliam1984) and could be important here. Answeringsuch basic questions could save fisheries managersmillions of dollars currently spent on poisoningUtah chub from exotic trout fisheries. Second, whatare the impacts of unintentional translocation ofUtah chub among populations? Because Utah chub

is still used as a baitfish, the possibility of introduc-ing chub to nonnative environments is real. Thefact that Utah chub has been spread to several loca-tions outside of its historic range also demonstratesthis ongoing potential. One possible effect ofanthropogenic gene flow would be the dilution oflocal gene pools and the potential homogenizationof current phenotypic diversity in Utah chub. Theecological outcomes of such actions might includedisruption of species interactions or communityorganization. Finally, fisheries managers shouldexplore what effects stocking of nonnative specieshas had on the ecological and evolutionary diversi-ty of native Utah chub populations. Rainbow trout,walleye Sander vitreus (also known as Stizostedionvitreum), smallmouth bass Micropterus dolomieu,and several other exotics now occupy Utah chubenvironments, and we know little of the eventualconsequences to native fish populations. However,the extinction of Utah chub several decades agofrom Utah Lake following the introduction of exot-ic piscivores and the rapid decline of chub in reser-voirs where warmwater predators have been recent-ly introduced suggest that some exotic species canhave strong negative impacts on Utah chub. Of par-ticular concern are Utah chub in desert spring envi-ronments, where warmwater predators have thepotential to drastically diminish chub populations.Removal of exotics from these environmentsshould be a first priority, particularly because theyprovide no real fishery.

Conclusions

Conservation efforts must continue to focus oneliminating extinction risks to rare or threatenedtaxa. However, biologists must also expand their def-inition of conservation to include efforts to main-tain biological diversity within common, wide-spread species. Utah chub provides a case study inwhy this broader vision is necessary. Although thespecies at large is not at risk of extinction, the levelsof biological diversity among populations within thespecies suggests that Utah chub is an important gen-erator of biological diversity. This species alsoappears to be an important store of phenotypic vari-ation, and this reservoir of variability may provideecological and evolutionary resilience against chang-

172

JOHNSON AND BELK


ing environmental conditions that characterize thelong-term hydrological cycles in the Bonnevillebasin and upper Snake River. Yet, Utah chub is notunique. Many ecological communities are com-posed of common, widespread taxa that exhibitsimilar patterns of diversity, and these too shouldwarrant out attention. Fishes such as speckled daceRhinichthys osculus, longnose dace R. cataractae,mountain sucker Catostomus platyrhynchus, andUtah sucker C. ardens are all neglected taxa thatcould be at risk of losing rare components of theirdiversity. Clearly, we still have much to learn abouthow such diversity arises and what efforts will berequired to allow populations to continue to evolve.What is also clear is that several human activitiescould interfere with natural ecological and evolu-tionary processes in Utah chub and that care shouldbe taken to safeguard against such abuses as wecome to better understand the biology of this bio-logically important species.

References

Belk, M. C., and J. B. Johnson. 2007. Biological status ofleatherside chub: a framework for conservation of west-ern freshwater fishes. Pages 67–76 in M. J. Brouder and J.A. Scheurer, editors. Status, distribution, and conserva-tion of native freshwater fishes of western North Ameri-ca: a symposium proceedings. American Fisheries Soci-ety, Symposium 53, Bethesda, Maryland.

Bowen, B. W. 1998. What is wrong with ESUs?: the gapbetween evolutionary theory and conservation princi-ples. Journal of Shellfish Research 17:1355–1358.

Broughton, J. M. 2000. Terminal Pleistocene fish remains fromHomestead Cave, Utah, and implications for fish biogeog-raphy in the Bonneville basin. Copeia 2000:645–656.

Crandall, K. A., O. R. P. Bininda-Emonds, G. M. Mace, and R. K.Wayne. 2000. Considering evolutionary processes in conser-vation biology. Trends in Ecology and Evolution 15:290–295.

Dowling, T. E., and B. D. DeMarias. 1993. The evolutionarysignificance of introgressive hybridization in cyprinidfishes. Nature (London) 362:444–446.

Dowling, T. E., and C. L. Secor. 1997. The role of hybridiza-tion and introgression in the diversification of animals.Annual Review of Ecology and Systematics 28:593–619.

Gaufin, R. F. 1964. Ecology of the Utah chub in Fish Lake,Utah. Master’s thesis. University of Utah, Salt Lake City.

Hubbs, C. L., and R. R. Miller. 1948. The zoological evidence:correlation between fish distribution and hydrographichistory in the desert basins of western United States.Pages 17–166 in The Great Basin, with emphasis on gla-cial and postglacial times. Bulletin of the University ofUtah, Biological Series 10(7), Salt Lake City.

Johnson, J. B. 2002. Evolution after the flood: phylogeogra-phy of the desert fish Utah chub (Gila atraria). Evolution56:948–960.

Johnson, J. B., and M. C. Belk. 1999. Effects of predation onlife-history evolution in Utah chub (Gila atraria). Copeia1999:948–957.

Minckley, W. L., and J. E. Deacon. 1991. The battle againstextinction: native fish management in the AmericanWest. University of Arizona Press, Tucson.

Minckley, W. L., D. A. Hendrickson, and C. E. Bond. 1986.Geography of Western North American freshwater fish-es: description and relationships to intracontinental tec-tonism. Pages 519–614 in C. H. Hocutt and E. O. Wiley,editors. The zoogeography of North American freshwa-ter fishes. Wiley, New York.

Moritz, C. 1994. Defining ‘evolutionarily significant units’for conservation. Trends in Ecology and Evolution10:373–375.

Olsen, H. F. 1959. The biology of the Utah chub, Gila atraria(Girard), of Scofield Reservoir, Utah. Master’s thesis.Utah State University, Logan.

Oviatt, C. G. 1997. Lake Bonneville fluctuations and globalclimate change. Geology 25:155–158.

Rees, H. D. 1935. The feeding habits of the chub, Tigomaatraria. Master’s thesis. University of Utah, Salt LakeCity.

Rosenfeld, M J. 1991. Systematic studies of members of thegenus Gila (Pisces: Cyprinidae) from the Great Basin andColorado River: protein electrophoretic and cytogeneticvariation. Doctoral dissertation. University of Utah, SaltLake City.

Schluter, D. 2001. Ecology and the origin of species. Trendsin Ecology and Evolution 16:372–380.

Schneidervin, R. W. 1985. Zooplankton stocks and thenature of size-selective predation by fishes in FlamingGorge Reservoir, Wyoming-Utah. Master’s thesis. Uni-versity of Wyoming, Laramie.

Simons, A. M., and R. L. Mayden. 1998. Phylogenetic rela-tionships of the western North American Phoxinins(Actinopterygii: Cyprinidae) as inferred form mitochon-drial 12s and 16s ribosomal RNA sequences. MolecularPhylogenetics and Evolution 9:308–329.

Smith, G. R., W. L. Stokes, and K. F. Horn. 1968. Some LatePleistocene fishes of Lake Bonneville. Copeia1968:807–816.

Tanner, V. M. 1936. A study of the fishes of Utah. Proceedingsof the Utah Academy of Arts, Science, and Letters13:155–184.

Templeton, A. R. 1998. Nested clade analysis of phylogeo-graphic data: testing methods about gene flow and pop-ulation history. Molecular Ecology 7:381–397.

Tuescher, D., and C. Luecke. 1996. Competition betweenkokanees and Utah chub in Flaming Gorge Reservoir,Utah-Wyoming. Transactions of the American FisheriesSociety 125:505–511.

Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic nicheand species interactions in size-structured populations.Annual Review of Ecology and Systematics 15:393–425.

173



175

The primary impact on native trout has beenintroduction of nonnative, invasive troutsand, to a lesser degree, land management

activities. Because of the characteristic physiogra-phy, meteorology, and hydrological setting of theSouthwest, the occurrence and degree of intermix-ing of local populations of these species within“disconnected” streams is markedly limited by notonly the natural setting, but anthropogenic influ-ences. The approach to management and recoveryfor the three native trouts has been in headwater,low-order streams and on a stream-by-stream,population-by-population approach. Recently,populations have become increasingly at risk frompost-wildfire impacts in these headwater reaches.The current status of and threats to these threenative salmonids are an expression of time, cli-mate, the landscape, and human-induced changesimposed upon the natural components. Recovery

efforts for southwestern trouts have been and mustcontinue to be approached in the context of bothnatural and anthropogenic influences. Future con-servation activities should be conducted on a land-scape/watershed or river basin scale and with refer-ence to natural and human-induced factors.

Recent interest in native trout management inthe West has escalated because of their continuedreductions in range and numbers and worseningconservation status. Rieman and McIntyre (1993)discussed and delineated demographics and habitatrequirements of bull trout Salvelinus confluentus rel-ative to management and conservation. Conserva-tion assessments for inland cutthroat troutsOncorhynchus clarkii have recently appeared (Duff1996; Young 1996), and a recent symposiumaddressed potamodromy and western salmonids(Gresswell et al. 1997). Population viability analysesare beginning to be addressed because of the con-

Native Southwestern Trouts: Conservation with Reference to Physiography,

Hydrology, Distribution, and Threats

ABSTRACT Three native trouts occur in the southwestern United States. The Rio Grande cutthroat trout

Oncorhynchus clarkii virginalis persists in New Mexico and southern Colorado on the Santa Fe, Carson, and Rio

Grande national forests and private lands. The Gila trout O. gilae and the Apache trout O. gilae apache (also known

as O. apache) occur in isolated headwater streams of the Gila and Little Colorado rivers on the Gila and Apache-

Sitgreaves national forests and Fort Apache Indian Reservation in southwestern New Mexico and east-central Ari-

zona, respectively. For more than two decades, intensive management has been directed at the Apache, Gila, and

Rio Grande cutthroat trouts. Despite the efforts, their decades-long listed status remains unchanged for the Gila

and Apache trouts, and the Rio Grande native is under consideration for listing. The objectives of this paper are to

review the literature and management activities over the past quarter of a century in order to delineate why recov-

ery and conservation have been so difficult for southwestern trout.

John N. Rinne and Bob Calamusso



JOHN N. RINNE U.S. Forest Service, Rocky Mountain Research Station, Southwest Forest Science Complex, 2500 South Pineknoll Drive, Flagstaff,Arizona 86001, USA

BOB CALAMUSSO Tonto National Forest, 2324 East McDowell Road, Phoenix, Arizona 85010, USA


cern of loss of genetic variability with the extirpa-tion of individual, isolated stocks of native inlandtrouts (Propst et al. 1992; Lee and Rieman 1997).

Despite these efforts, the literature includes lit-tle information (see Rinne 1988) on a landscape orwatershed scale approach to recovery of southwest-ern native trout. Because of the cumulative effectsof aridity, physiography, hydrology, and human-induced threats, these inland, rare trout species areequally or more imperiled than salmonids in moremesic landscapes to the north. Apache and Gilatrout have been listed as threatened and endan-gered, respectively, for more than two decades; theRio Grande species is under consideration for list-ing. Historically, these southwestern species weredispersed in a fragmented pattern across the land-scape. More recently, the occupation of the area byEuropeans and associated effects of managementactivities have further disconnected populations ofthese species. The combination of natural factorsand anthropogenic influences in the Southwest hasrendered nonexistent the common connectivity ofstreams that characterizes the more mesic PacificNorthwest and northern Rocky Mountains (Gress-well et al. 1997). Conservation of the southwesterntrout has been difficult, in part, because of the dis-connected and often harsh nature of aquatic habi-tats in which they reside.

The objectives of this paper are to discuss anddelineate (1) the natural setting and climate thatdirectly and indirectly influence southwesterntrouts, (2) historic and present distributions, (3)threats to native trout sustainability, and (4) futureconservation to prevent loss of populations ofthese species and the species themselves.

Physiography, Hydrology, and Meterology

The natural surface hydrology of any region is con-trolled by climate and topography. Such controlhas been documented for fishes and aquatic habi-tats of the Great Basin (Hubbs and Miller 1948).Because of the highly diverse geology, topography,and climate of the Southwest, a highly variable pat-tern of hydrology and drainage basins has devel-oped (Miller 1954; Minckley 1973; Figure 1). Rinne(1995d) described the interactions of climate andtopography that produces a variable pattern of pre-

cipitation and surface hydrology across theMadrean region or “sky island” landscape in south-eastern Arizona. Drainage basins issue in oppositedirections off insular massifs to provide a radialarray of aquatic habitat refugia for native fishes(Figures 2–5). Streams course in three directionsfrom Mount Baldy and Mount Ord in the WhiteMountains of Arizona to form the headwaters ofthe Gila, Little Colorado, and Salt rivers. In NewMexico, the mountain ranges of the southernRocky Mountains likewise dictate drainages to flowin easterly (the Canadian River basin), southeaster-ly (Pecos River drainage), southerly (Rio Grande),northwesterly (San Juan basin), and westerly (Lit-tle Colorado basin) directions. The native trout inthese headwater reaches, nearly connected acrossridges, have no opportunity to mix, being separat-ed by tens to hundreds of kilometers downstream.Further, at downstream junctures or confluences,most streams lie in arid, desert landscapes wherewater temperatures and loss of surface flow arebarriers to both potadromy (Gresswell et al. 1997)and survival of native trout.

Aquatic habitats in the southwestern landscapeare fragmented because of the physiography, cli-mate and meteorology, and resultant hydrology(Rinne and Stefferud 1998; Rinne 2002; Rinne andMiller 2006). The drainage systems inhabited bythe three trout lie largely within an amalgamationof geological features (Figure 1). The landscape inArizona and New Mexico is composed of the Basinand Range Province, the central highlands or cen-tral mountain region, the Colorado and/orMogollon plateaus, and the southern RockyMountains and plains (Nations and Stump 1981;Chronic 1983, 1987; Figure 1). Topographically,the Basin and Range Province is comprised of aparallel interpositioning of mountain ranges andvalley basins coursing northwest to southeast. Thisinterdigitation of mountain ranges and valleyfloors results in creation of the “sky islands” (War-shall 1995) of the Southwest lying within theMadrean Archipelago (Debano et al. 1995). At itsnorthern boundary, the Madrean ecoregion abutsthe Arizona–New Mexico (Central) Highlandsecoregion (Bailey 1995) along a major fault blockknown as the Mogollon Rim. These highlands arethe southern extent of the Colorado Plateau and

176

RINNE AND CALAMUSSO


are often referred to as the Mogollon Plateau. TheMogollon Rim commences in west-central Ari-zona and courses southeasterly into southwesternNew Mexico (Nations and Stump 1981) where thismajor geologic feature loses identity in the GilaMountains of southwestern New Mexico (Figure1). The southern Rocky Mountains terminate inmountainous terrain of southern New Mexico andwestern Texas.

The overriding climate of the southwesternUnited States is arid (Jaeger 1957; Dunbier 1968).The climate and weather of the Southwest are con-trolled by dominant airflow patterns, positioning ofhigh and low pressure areas, and diverse topographyof the region (Green and Sellers 1964). Arizona andNew Mexico are the second and third most aridstates in the United States. (Jaeger 1957). However,the presence of rugged and diverse topographyresults in locally, highly variable climatic and mete-orological conditions. Accordingly, cooler, wetter,and higher elevations areas are surrounded by hot,dry lowlands (Rinne 1995d). As a result, Arizona isone of the most geologically and climatically diverse

areas in the United States of America ranging fromarid and hot in the extreme southwestern region tocold and treeless tundra on the highest peaks in theWhite Mountains of east-central Arizona (Greenand Sellers 1964; Minckley 1973).

In summary, the natural topography andhydrology of the landscape encompassing therange of southwestern trouts control inter-stream mobility, thereby severely limiting con-nectivity of populations and resulting in historicisolation of populations. Stream connectivitythat would facilitate metapopulations did notexist in the river basins of the Southwest to theextent that it occurred historically and even cur-rently for salmonids in more mesic landscapes ofthe Pacific Northwest and northern RockyMountains. In part, historic definition for thethree southwestern trouts is precluded by a lackof collection records. Notwithstanding, drainagepatterns and elevation changes have combinedwith aridity and climatic diversity to render con-nectivity of suitable habitat much reduced fornative trout in the Southwest.

177

NATIVE SOUTHWESTERN TROUTS

Figure 1. Major landscape regions and features of Arizona and New Mexico, southwestern United States.


Distributions

Historic

Reconstruction of indigenous ranges and connec-tivity of southwestern trout populations are difficultbecause of a lack of historic records. Specific distri-butions can be found in Miller (1950, 1972) for Gilaand Apache trout and in Sublette et al. (1990) andCalamusso and Rinne (1996) for Rio Grande cut-throat trout and are not repeated here. The mostlikely historic distributions of the southwesterntrouts are illustrated in Figure 2. By the time the Gilaand Apache trouts were officially described, bothhad declined drastically in range and numbers(USDI 1979, 2004). Introductions of nonnativetrouts and land-management practices were cited aslargely responsible for the declines of both species(Miller 1961; Minckley 1973; Behnke and Zarn1976). The original distribution of the Rio Grandecutthroat trout is not well known (Wernsman 1973;Wallace and Behnke 1974). It is uncertain whetherthe Rio Grande cutthroat trout was native to theCanadian River system (Behnke 1992), but based ona collection of pure Rio Grande cutthroat trout

from Ricardo Creek, a tributary to the CanadianRiver, Las Animas County, Colorado, Behnke (1992)concluded that the Rio Grande cutthroat trout isnative to the Canadian drainage.

Historically, the three native trouts occupyingcoldwater, montane streams had relatively greaterintra-river basin connectivity–the extent of whichwas determined by elevation and water qualityconditions. Based on interviews with long-timeresidents, Gila trout were reported as occurringdownstream to “the mouth of the box canyon”about 10 km northeast of Cliff, New Mexico(Miller 1950). Such distributions severely limitedconnectivity with other populations of Gila troutin the San Francisco and Blue rivers (Figure 2).Populations of “Gila trout-like” fish in the headwa-ters of the Verde River are markedly separated fromthose in the headwaters of the Gila River in south-western New Mexico. Gila trout were probably onlypotentially connected temporally in winter withlowered (12–20ºC) waters temperatures withintributary streams of the Verde, such as Oak andWest Clear creeks. The Black and White river pop-ulations of Apache trout were separated from the

178

RINNE AND CALAMUSSO

Figure 2. Historic and current distributions of Apache, Gila, and Rio Grande cutthroat trouts.


Little Colorado River populations by MountsBaldy and Ord (Figure 3). Apache trout in K. P.Creek were topographically and hydrologically dis-connected by hundreds of stream kilometers,although lying a few kilometers from nearby pop-ulations in the headwaters of the Black, White, andLittle Colorado rivers (Figure 3).

The original distribution of the Rio Grande cut-throat trout is uncertain (Calamusso and Rinne1999); however, of the three species, this species dis-plays the greatest potential for contemporarypotadromy within the Rio Grande basin (Calamus-so and Rinne 1996, 2004; Figure 5). Populations inthe Chama and Jemez rivers to the west of themainstream Rio Grande were connected with theSanta Fe River to the east. From the east, the RioCapulin, Rio Nambe, and other rivers entered theRio Grande mainstream. Evidence of occurrence ofthe Rio Grande cutthroat trout in the Rio Grandemainstream is provided by Sublette et al. (1990).Notwithstanding, headwater populations in thePecos and Canadian River basins were markedlydisconnected from those in the Rio Grande.

Current

Current distributions of the three native troutspecies are shown in Figures 2–5. In 1972, Apachetrout occupied less than 48 km (3%) of its original1,600 km of stream habitat (Harper 1978; USDI2004). Such reduction in range reflects loss of pop-ulations in downstream reaches of the Black andWhite rivers (Figure 2). Rinne (1985a) reportedthat pure populations were extant in less than adozen headwater streams in the White Mountainsof east-central Arizona, but cooperative recoveryefforts have increased that number of streams to 29(Figure 3; USDI 2004).

The Gila trout is now restricted to headwatersabove natural barrier falls or dry streambeds(USDI 1979; Mello and Turner 1980; Rinne 1981a;Figure 4). Such isolation has prevented hybridiza-tion and drastically reduces opportunity forpotadromy (Mello and Turner 1980; Loudenslageret al. 1986). By the early 1970s, Gila trout occurrednaturally in only five headwater streams.

The distribution of Rio Grande cutthroattrout has declined to 5–7% of its former range inthe Rio Grande, Pecos, and Canadian drainages

(Stumpff and Cooper 1996; Figure 2). Presently,Rio Grande cutthroat trout are confined to smallheadwater streams (Figure 5), and occur in theRio Grande drainage of southwest Colorado andin four drainages in New Mexico (Rio Grande,Pecos, Canadian, including the Mora, andTularosa; Hendrickson et al. 1980; Sublette et al.1990). Currently, there are approximately 86known populations of Rio Grande cutthroattrout (50 considered pure) in New Mexico (Cala-musso and Rinne 1996)—45 populations on theCarson National Forest and 41 on the Santa FeNational Forest. Most of these populations arelocated in wilderness areas, in isolated headwaterstreams disconnected from other populations ofcutthroat trout.

179


Figure 3. Streams in the White Mountains of east-centralArizona currently containing and managed for the nativeApache trout. Streams are 1. White River, 2. Elk CanyonC., 3. Deep Cr., 4. Firebox Cr., 5. Little Diamond Cr., 6. Par-adise Cr., 7. Ord Cr., 8. Little Bonita Cr., 9 Crooked Cr., 10.Boggy Cr, (Ft. Apache Indian Res), 11. Flash Cr., 12. SquawCr., 13, Big Bonito Cr., 14, Hurricane Cr., 15. Paddy Cr., 16.Soldier Cr., 17. Wildcat Cr., 18. Hay Cr., 19., Stinky Cr., 20.Home Cr., 21. Snake Cr., 22. Bear Wallow Cr., 23. K. P. Cr.24. Grant Cr., 25. Coleman Cr., 26. Lee Valley Cr., 27. Coy-ote Cr., and 28. Mineral Cr.


Anthropogenic Factors and Threats

Habitat modification and introduction of non-native fishes have reduced connectivity ofstream systems and populations of southwest-ern trout beyond the natural setting. These twoprimary, anthropogenic factors have resulted inmarked decline in range and numbers of nativesouthwestern fishes (Minckley and Deacon1991; Rinne 1991, 1994, 1995c, 2003a; Rinneand Minckley 1991; Rinne and Stefferud 1998).Although these two factors have paralleled eachother in upper elevation, montane streams,introduced salmonids currently remain as thegreater impact of the two (Calamusso andRinne 1999).

Introduced Salmonids

Introduced rainbow trout O. mykiss, browntrout Salmo trutta, and brook trout Salveli-nus fontinalis have markedly impacted thethree native southwestern trouts. Rainbowtrout was first introduced into the waters ofArizona and the lower Colorado River in1900 (Rinne 1995a) and has been stockedfor sportfishing into suitable waters acrossArizona (Minckley 1973). For example,more than 61 million nonnative sport fisheshave been introduced into lakes in the LittleColorado and Black river drainages since1930 (Rinne and Janisch 1995), and an addi-tional 8 million nonnative sport fishes havebeen introduced into these two rivers andtheir tributaries. Most (60%) were rainbowtrout, but cutthroat trout (12%), brooktrout (7%), and brown trout (2%) were alsostocked. As a result, hybridization occurredbetween rainbow and the native Apachetrout (Rinne 1985b; Rinne and Minckley1985; Carmichael et al. 1993, 1995) andresulted in loss of purity of many popula-tions, further fragmentation of populations,and reduction in the potential for connectiv-ity of pure Apache trout populations.

In New Mexico, rainbow trout also wereintroduced widely into the native ranges ofboth the Gila and Rio Grande cutthroattrout by the turn of the century. Rainbow

trout was first introduced in Bluewater Creek (nearGrants, New Mexico, current-day Cibola NationalForest) and Eagle Creek (near Ruidoso, New Mex-ico, current-day Lincoln National Forest) in 1896(Sublette et al. 1990). Subsequently, rainbow trouthave been introduced into all waters of the state.The Gila and Rio Grande cutthroat trouts readilyhybridize with rainbow trout resulting in markedreduction of natural pure populations and muchreduced opportunity for intermixing of popula-tions to occur (Miller 1972; USDI 1979; Rinne1988; Sublette et al. 1990; Calamusso and Rinne1996). Brook trout and brown trout were intro-duced into Arizona in the 1920s (Rinne 1995a,2003b) and New Mexico in the early 1900s (Sub-lette et al. 1990). Both species potentially affect

180

RINNE AND CALAMUSSO

Figure 4. Status and current distributions of populations of Gila trout,Gila National Forest, southwestern New Mexico. Streams that arenative (N), introduced (I), low level cross (LL), pure (P), extinct (Ex)and suspect (S) are: 1. Spruce Cr. (N,P), 2. Sacaton Cr (I, LL), 3. IronCr (N. LL), 4. West Fork of Gila River (S), 5. White Cr. (I, LL), 6. McKen-na Cr. (N,LL), 7. Little Cr. (I, LL), 8. Main Diamond Cr. (N, P), 9. S. Diamond Cr. (Ex), 10. Mcknight Cr. (I, P), and 11. Sheep Corral (I, P).


native trout through competitive interactions forfood, space, and direct predation and replaceApache trout. Relative numbers of brown, brook,and Apache trout in Ord Creek, Arizona (Rinne etal. 1981), and more recent, observed interactions inother streams on the Fort Apache Indian Reserva-tion (Carmichael et al. 1995), illustrate the effectivereplacement of Apache trout by these two species.Similarly, population data where brown trout andGila trout (Mello and Turner 1980) or Rio Grandecutthroat (Calamusso and Rinne 1996, 1999) co-occur, reveal dramatic reductions in both nativesfollowing introductions of brook and brown trout.Table 1 specifically illustrates the negative impactof brown trout on Rio Grande cutthroat in multi-ple streams in northern New Mexico.

The primary modes of interaction between allthree introduced trout and the native trout includecompetition for food and space and direct preda-tion. However, rainbow trout primarily interacts

with and has negatively impacted native south-western trout through introgression and reductionin purity of stocks (Rinne et al. 1985; Rinne 1988).Introgression of rainbow trout genes has been amajor mechanism of impact to all three nativesouthwestern trouts (Rinne 1994). Distributionpatterns of pure Apache trout populations closelyparallel the stocking histories of rainbow trout ontwo major land management areas in the WhiteMountain region, the Fort Apache Indian Reserva-tion, and the Apache-Sitgreaves National Forest(Rinne and Minckley 1985).

Rinne et al. (1981) reported the negative impactby brook and brown trout on the native Apachetrout in Ord Creek, Fort Apache Indian Reservation.These two invasive salmonids, through competitionfor food and space and direct predation, had nearlyextirpated a type locality population of Apachetrout. Rinne (1985a) suggested that the Fort ApacheReservation was the refugium for pure Apache trout;

181


Figure 5. Current distribution of Rio Grande Cutthroattrout in Colorado and New Mexico. Streams are 1. Ritode los Pinos, 2. Rio Puerco, 3. Rito Anastacio, 4. Rio delas Vacas, 5. Rito de las Perchas, 6. Rito Café, 7. Rio dela Cebolla, 8. Rito Resumidderor (Oso Cr.), 9. CanonesCr., 10, Chihuahuenos Cr., 11. East Fork Polvedera Cr.,12. Rio Name, 13, Rio Frijoles, 14. Rio Quemado, 15.Rio Truchas, 16, Dalton Cr., 17. Macho Cr., 18. DoctorCr., 19. Cave Cr., 20. Jack’s Cr., 21. Rito Azul, 22. Ritode los Chima;yosos, 23. Rito Maestas. 24. Riot Murphy,25,. Santiago Cr., 26. Alamitos Cr., 27. Policarpio Cr., 28.Sardinas Canyon, 29. Jaroso/Saloz Cr., 30. PalocientoCr., 31. Rio Frijoles, 32. Luna Cr., 33. Rio de la Olla, 34.Rio Chiquito, 35. Tienditas Cr., 36. Rio Hondo (SFNF),37. San Cristobal Crl, 38. Columbine Cr., 39. SawmillCr., 40. Bitter Cr., 41. Lake Fork Cr., 42. Comanche Cr.,43. Vidal Cr., 44. Little Costilla Cr., 45. LaCueva Cr., 46.Powderhouse Cr., 47. MyCrystal Cr., 48. Tanques Cr.,49. Rio Nutrias Cr., 50. Tio Grand Cr., 51. El Rito, 52.Canjilon Cr., 53. Nabor Cr., 54. Native Lake, 55. Wolf Cr.,56. Rio de las Pinos, 57. Cascade Cr., 58. Osier Cr, 59.Lake Fork Cr., 60. Torcido Cr., 61. Jim Cr. 62. RhodesCr., 63. Rough Cr., 64. Middle Fork San Francisco

Cr/Lake, 65. San Francisco Cr., 66. Bennet Cr., 67. Cave Cr., 68. Unknown Cr., 69. East Fork, Middle Cr., 70 Tueele Cr., 71.Cherry Cr., 72. Little Medano Cr., 73. Medano Cr./Lake, 74. Placer Cr., 75. Sangre de Cristo Cr., 76. Little Ute Cr./Lake, 77.West Indian Cr., 78. McCarty Lake, 79. North Fork Trinchera Cr., 80. South Fork Trinchera Cr., 81. North Fork Vallejos Cr., 82.Vallejos Cr., 83. Alamosito Cr, 84. Torcido Cr., 85. Jaroso Cr., and 86. Cuates Cr.


most streams sampled in the late 1970s did not havebrown trout sympatric with the native Apache trout.A decade later, Carmichael et al. (1995) listed 35streams that contain Apache trout in the WhiteMountains; 25 were found on the reservation and 10on the Apache Sitgreaves National Forest. Of these,only 15 were devoid of brown trout; 3 of these 15contained brook trout. Therefore, almost one-thirdof all streams sampled in 1988 on both land manage-ment areas contained no introduced trout.

The threat of replacement with invasioninto Rio Grande cutthroat trout waters andsubsequent hybridization and competition bynonnative salmonids is ever present (Cala-musso and Rinne 1999). Rainbow and othercutthroat trout taxa have hybridized with theRio Grande cutthroat, resulting in intro-gressed forms (Behnke 1992). By comparison,brown trout and brook trout compete withnative cutthroat trout for food and space.Based on relative abundance data fromstreams (Table 1), introduction of nonnativebrown trout may significantly reduce theabundance pure strains of Rio Grande cut-throat through competitive interactions.

Research suggests that brown trout, throughpredation, are currently the greatest threat to thesustainability of Rio Grande cutthroat wildstocks (Calamusso and Rinne 1999; Table 1).Laboratory studies suggest rapid and efficientpredation of brown trout on Apache trout fry(Rinne 1995b). Up to 60% (10 of 25, two trials)of Apache trout fry (75–100 mm) were con-sumed in a period of 24 h in absence of cover.Despite the presence of artificial cover, 68% and76% loss was recorded within 4 d. Predation bybrown trout on age-0 and age-1 Apache troutappears to be the major mechanism of interac-tion between this introduced and native speciesthat results in reduction and replacement ofnative Apache trout populations. The predatoryeffects of brown trout on Rio Grande cutthroatalso have been demonstrated in controlledexperiments. Preliminary field predation exper-iments reveal that 50–80% of age-0 (<60 mm)cutthroat are consumed by brown trout (>200mm) within 2 d, even in presence of cover (J. N.Rinne, unpublished data).

Land-Management ActivitiesUngulate grazing

Domestic livestock grazing has been a land-useactivity on watersheds containing native troutsince before the 1900s in the Southwest (Scurlock1998). Although most information on grazingeffects on fishes addresses salmonids, studies arefrom the northern Rocky Mountain and Great

182

RINNE AND CALAMUSSO

Table 1. Relative density (D, n/ha) and biomass (B, kg/ha) of Rio Grande cut-throat trout (RGCT) and brown trout (BT) in study reaches in Ameri-can Creek and Rito Cafe, and Rio de las Vacas, Santa Fe NationalForest, 1995.

RGCT BTStudy reach D B D B

American Creek

1 324 9 279 222 702 32 0 03 90 6 90 134 488 13 195 95 300 9 486 126 89 5 714 587 56 5 617 46

Rito Cafe (below barrier)

1 676 14 2,801 2412 400 15 2,300 1823 0 0 3,200 1444 0 0 3,875 96

Rito Cafe (above barrier)

5 2,545 30 91 306 2,608 36 0 07 1,080 22 83 5

Rio de las Vacas (below barrier)

1 0 0 318 182 117 2.7 233 103 0 0 595 604 156 2.4 739 435 0 0 1,468 846 61 0 .55 1,908 119

Rio de las Vacas (above barrier)

1 3,905 133 0 02 2,827 154 0 03 33 86 121 0 4 2,548 121 0 0


Basin regions and little information is available onthe effects of grazing on the habitats of the nativetrout (Rinne 1985a, 1988, 1998, 2000). Less than ahalf dozen studies exist for the Southwest; only twoare published in the open literature and nonespecifically address any of the native trout (Rinne1999). None addresses the potential impact of thisland-use activity on southwestern trout or theinfluence of the large, wild ungulate elk Cervus ela-phus. Because of the lack of information and fac-tors that confound research (Rinne 1998), and the“general consensus” that grazing unequivocallyand negatively affects fishes, much controversy sur-rounds the grazing-fisheries issue (Platts 1991).

A recent study on the effects of ungulate grazingon Apache trout in tributaries of the West ForkBlack River in the Apache Sitgreaves National For-est indicated that most fish were in upstreamreaches of two creeks within pastures grazed byboth ungulates (Table 2). Fish were absent fromheavily impacted, downstream montane meadowreaches of the study area from which both ungu-

lates had been excluded for a year or more. Sam-pling in autumn 1994 indicated a reduction inupstream trout populations compared to a 6-foldto 10-fold increase in trout number per kilometerin the cattle-excluded reaches and 10-fold to 20-fold increases in fish per kilometer in the elk exclo-sures. Trout populations remained low in 1995 inall reaches of stream. By comparison, sampling inautumn 1996, following a summer of severedrought resulted in further, marked reductions inApache trout populations in both creeks.

Results of study in three streams on the WestFork allotment in the White Mountains suggesttrout densities appear unrelated to ungulate graz-ing intensity. Trout populations fluctuated greatlyand were even highest in those study sections thatwere grazed by both ungulates (i.e., common pas-tures, 1993 and 1994). Increases in trout popula-tions in downstream reaches of stream (i.e., cattleand total ungulate exclosures) in 1994 appear morerelated to (1) downstream migration of trout intomeadow reaches, (2) droughts that occurred bothin summer 1993 and 1996, and (3) substrate com-position (Rinne and Neary 1996a) than to actualtrout population increase resulting from habitatimprovement. Although riparian habitats haveimproved in vegetation composition and biomassfrom 1993 to 1996 (A. Medina, Rocky MountainStation, personal communication), fish populationresponse does not parallel these changes. By com-parison, drought in 1993 and 1996 resulted inmarked reductions of Apache trout populations.

Fire management

Only recently have information on the effects of firemanagement on aquatic habitats and fishes began toemerge in the Southwest (Rinne 1996; Rinne andNeary 1996b; Rinne 2004; Carter and Rinne, inpress; Rinne and Carter, in press). Because of theU.S. Forest Service’s fire suppression policy over thepast century, dead and live fuel loading on the land-scape has occurred (Covington and Moore 1992;Rinne and Neary 1996b). The aftermaths of recentwildfires in the Southwest have been devastating tofishes and their habitats (Propst et al. 1992; Rinne1996, 2003c); marked reduction to total loss of fishpopulations has resulted in some headwaterstreams. Turner (1996) reported the loss of the

183


Table 2. Mean number of trout per kilometer in tributaries of the WestFork Black River, autumn, 1993–1996. Numbers in parenthe-ses are number of 40-m sample sections. Common indicatesboth ungulates present, cattle indicates that only elk arepresent, and elk indicates no ungulates present.

StreamTreatment Centerfire Boggy Wildcat

1993Common (10) 240 (6) 330 (8) 12.5 Cattle (13) 7 (14) 4 (2) 0 Elk (6) 0 (7) 0 –

1994Common (3) 160 (8) 180 (8) 30 Cattle (10) 40 (14) 40 (9) 30 Elk (6) 20 (7) 10 –

1995Common (3) 38 (5) 60 (2) 25Cattle (7) 129 (5) 5 (2) 0Elk (10) 33 (5) 0 –

1996Common – (13) 19 –Cattle (6) 6 (6) 4 – Elk (6) 0 (6) 0 –


largest (type locality) population of Gila trout inMain Diamond Creek and severe effects on anotherpopulation in South Diamond Creek resulting froma single wildfire event. More than a half a dozen firesin the past 6 years on the landscape encompassingstreams containing Gila trout illustrate the signifi-cance and chronic nature of this threat. Floods fol-lowing the denudation of watersheds and their toxicnature and heavy sediment loads can be lethal toentire populations (Rinne 1996; Turner 1996). Thewildfire in 1989 on the watershed of Main DiamondCreek nearly extirpated the type locality populationof Gila trout (Propst et al. 1992; Rinne and Neary1996b). Subsequent fires on the watersheds encom-passing streams inhabited by Gila trout have furtherthreatened the current geographic distribution ofthis species (Turner 1996; Propst and Stefferud1997). Expansion of the species’ much reducedrange to larger, downstream waters will address theimminent threat of wildfires and their real andpotential impact on the species in isolated, headwa-ter streams (Rinne 1996; Rinne and Neary 1996b).The Dude, the Divide, and several other wildfiresemphasize the need to align fire and fisheries man-agement in the Southwest by recent fire use wildfiremanagement and control or prescription burns(Rinne and Neary 1996b; Rinne 2003c).

Future Conservation

Management of the three native trout over the pasttwo and a half decades since enactment of the 1973Endangered Species Act has been case-by-case,stream-by-stream, and situation-by-situation. Alarge, landscape-scale approach to management incontext of drainage basin connectivity, threats, andland management activities has not been, andbecause of landscape and stream fragmentationmay not be a viable management option for futureconservation and management of the native south-western trouts (Rinne et al. 2004). A landscapeapproach to sustaining the southwestern trout isextremely difficult because of topography, climate,hydrology, and human-induced factors. Notwith-standing, land and resource managers must be evervigilant of opportunities that embrace utilizationof large aquatic systems or drainage basins to con-serve and sustain these rare southwestern trouts.

Rinne (1988) first suggested the need for managingsouthwestern trout on a large, river drainage orbasin scale. Analyses of population, habitat, anddistributional data for Gila trout by Propst andStefferud (1997) suggest that this is a viable man-agement alternative to sustain this endangeredtrout. We suggest the same for the Apache and RioGrande cutthroat trout. We also suggest that futureconservation of the three southwestern troutshould be conducted (1) in the context of a land-scape/watershed or river basin approach, and (2)relative to physiography, hydrology, current distri-butions, and threats.

Physiography, Hydrology, and Current Distributions

The endangered Gila trout is now restricted to lessthan a dozen natural and introduced populationsin the Gila River headwaters (Figure 4). The great-est opportunity to address this situation and tomanage this endangered species on a larger, riverbasin approach is present in the West, Middle, andEast forks of the Gila River. However, only theWhiskey Creek population is pure in the West Forkdrainage. Although secure above an artificial barri-er, an introduced population in Little Creek hasintrogressed with rainbow trout that was intro-duced from wild stock in McKenna Creek. Similar-ly, the single population in the Middle Fork (IronCreek) is secure above an artificial barrier but isfrom impure McKenna Creek stock.

On the East Fork, the type locality populationin Main Diamond Creek and the population inSouth Diamond Creek were markedly impacted tothe threat of extinction by the Divide Fire in 1989(Propst et al. 1992). Because of this event, the puri-ty and type locality status of Main Diamond Creek,the East Fork stock should be considered of highpriority for management on a watershed scale. Atthe largest scale, a plausible long-term goal wouldbe to manage all three forks of the Gila and theirtributaries for Gila trout, with a barrier positionedbelow their confluence. At this scale, maximumhydrologic connectivity and opportunity forpotadromy of populations could be achieved andthe security and sustainability of this now endan-gered trout almost assured.

184

RINNE AND CALAMUSSO


The primary example of a watershed approach tomanagement of Apache trout is the recent (1996) ren-ovation of the upper West Fork of the Black River onboth the reservation and forest. This cooperative pro-gram among the Arizona Game and Fish Department,the U.S. Forest Service, and Trout Unlimited was initi-ated in 1993 on the West Fork of the Black River(Rinne and Janisch 1995). The overall plan includedrenovation, barrier construction, reintroduction, habi-tat enhancement, and special regulations of 29 km ofstreams and rivers. This program alone will increase by50% the stream kilometers containing pure Apachetrout in the West Fork. Further downstream, renova-tion and barrier construction above the confluence ofthe east and west forks would permit production oflarger-sized (>300 mm total length; see Rinne et al.1979) native trout for anglers in larger, downstreamreaches of the West Fork of the Black River.

The Rio Grande cutthroat trout is the most abun-dant of the three southwestern trouts, and althoughcurrently only a sensitive species (Calamusso andRinne 1996, 2004; Figure 5), it is being considered forlisting. Accordingly, opportunities are greatest forwatershed or river basin management of this unlistedsubspecies of cutthroat trout. One candidate river sys-tem for a watershed approach is the upper Jemez Riverbasin that contains the Rio de las Vacas (Rinne 1988).Here, six streams contain Rio Grande cutthroat, and ifcombined with five additional populations in streamsthat issue off the San Pedro Parks Wilderness, could bemanaged as a unit. A barrier would be required abovethe reservoir on Jemez River, but unfavorable hydrolo-gy (i.e., drying and low-elevation warmwater reachesof stream) would secure populations in five addition-al streams (Rito de los Pinos, Rio Puerco, Canones,Chihuahuenos, and East Fork Polvedera creeks).Opportunities for management to maximize connec-tivity are optimum and should be addressed, priori-tized, and implemented while populations and oppor-tunities are favorable. A fish kill in spring 2006 in theVacas provides opportunity to establish the entireJemez River as a Rio Grande native basin.

Introduced Salmonids

The threat of interactions of introduced rainbow,brown, and brook trout is of paramount impor-tance to conservation and sustainability of the

three native trout. Protecting aquatic habitats fornative trout (Rinne and Turner 1991) has been aprimary activity of both the Apache and Gila troutrecovery plans (USDI 1979, 2004). Restorationefforts in small, first-order, headwater streams havebeen largely reactive and preservation-oriented.Conservation efforts need to move beyond protec-tion to progressive, proactive recovery for the threenative trouts (Johnson and Rinne 1982). Theseaquatic habitats are habitat-limited and perhapssuboptimal for the native trouts (Rinne 1981b) andcertainly have a greater probability of loss as aresult of the constant and increasing threat of wild-fire. Barrier construction to preclude upstreammigration and prevent interactions with nonnativesalmonids has been an instrumental and viablemethod in light of the wide distribution of nonna-tive salmonids in the Southwest. These same, well-refined techniques that have been used for severaldecades in conservation efforts for the three nativetrout will continue to be effective. However, theymust be employed in the context of a landscape orriver-basin scale. A key to the success of restoringthe native trouts to larger watersheds will be recip-rocal designation of often adjacent watershedsexclusively for nonnative salmonid management.The state of Arizona has already adopted thisapproach (Rinne and Janisch 1995). Utilizing lakesfor sport angling (Rinne et al. 1979) will not onlyincrease the opportunity for quality sport fishing,but a positive relationship with and attitude by thepublic toward native trout and their management.

Fire and Grazing Management

Recent setbacks in Gila trout conservation in con-text of fire management and utilization of onlyheadwater streams stresses the need for a larger-scale approach to effectively sustain the threenative southwestern trout. Recovery of this specieshad progressed to near downlisting in 1989 whena series of wildfires severely impacted headwaterpopulations and set the recovery program backseveral decades. Two decades of intensive manage-ment using techniques to protect populationsfrom nonnative salmonids were almost negatedbecause of the failure to corroborate fire and fish-eries management (Rinne and Neary 1996b).

185



Implementation of compatible fire managementin selected river basins or subbasins both on thewatershed and in riparian stream corridors needsto be adopted as soon as possible.

Unlike the direct impact of introducedspecies of salmonids, the effect of ungulate graz-ing on native trout is indirect (Rinne 2000).Habitat degradation has been implicated asresponsible for the decline in distribution andstatus of all three trout (USDI 1979, 2004; Rinne1988). However, little published data are avail-able to substantiate the existence or extent ofgrazing effects on native trout in the Southwest.Based on the salmonid literature (Meehan1991), habitat degradation could result not onlyfrom grazing but logging, mining, and roadbuilding associated with these land managementactivities. These land use activities potentiallyalter structure and function of riparian areas,flow regimes, water quality, and water tempera-ture (Platts 1991). Much remains to be learnedabout the relationship between the native trout,their habitats, and abundance in response to allland-management activities.

Summary and Conclusions

In summary, the combination of natural featuresand conditions across the southwestern landscapeand human-induced impacts have greatly reducedor essentially precluded population connectivity ofsouthwestern native trout populations. Historically,the natural landscape setting of the Southwestgreatly reduced connectivity of native trout habi-tats; anthropogenic effects further fragmentedalready isolated populations. Management and sus-tainability of the three species of southwesterntrout therefore must proceed within the naturaland human-altered framework present. Alternativemanagement options for fire, grazing, and fisheriesmanagement are possible and may help to sustainthe native trout through habitat and populationmanagement. Sustaining native southwesterntrouts will require closely integrated managementand research in the context of “adaptive manage-ment” with opportunity for periodic adjustmentbased upon research and monitoring (Rinne 2000).Renovation and restocking of streams ensure both

security and purity of stocks by removing the mul-tiple impacts of nonnative trouts. Stream renova-tions and restocking of the native trout within his-toric ranges have been and will continue to be,within a landscape approach, an essential compo-nent in the future conservation efforts.

ReferencesBailey, R. G. 1995. Descriptions of the ecoregions of the

United States. U.S. Forest Service, Miscellaneous Publica-tion 1391, Washington, D.C.

Behnke, R. J. 1992. Native trout of western North America.American Fisheries Society, Monograph 6, Bethesda,Maryland.

Behnke, R. J., and M. Zarn. 1976. Biology and managementof threatened and endangered western trout. U.S. ForestService, Rocky Mountain Research Station, GeneralTechnical Report 28, Fort Collins, Colorado.

Calamusso, B., and J. N. Rinne. 1996. Distribution of RioGrande Cutthroat trout (Oncorhynchus clarki virginalis)and its co-occurrence with Rio Grande sucker (Catosto-mus plebeius), and Rio Grande chub (Gila pandora) onthe Carson and Santa Fe national forests. Pages 157–167in D. W. Shaw and D. M. Finch, technical coordinators.Desired future conditions for southwestern riparianecosystems: bringing interests and concerns together.U.S. Forest Service, Rocky Mountain Research Station,General Technical Report 272, Fort Collins, Colorado.

Calamusso, B., and J. N. Rinne. 1999. Native montane fishesof the middle Rio Grande ecosystem: status, threats andconservation. Pages 231–237 in D. M. Finch, J. C. Whit-ney, J. F. Kelly, and S. R. Loftin, technical editors. RioGrande ecosystems: linking land, water, and people:toward a sustainable future for the middle Rio Grandebasin. U.S. Forest Service, Rocky Mountain Research Sta-tion, Proceedings 7, Ogden, Utah.

Calamusso, B., and J. N. Rinne. 2004. Forests, fish and fire:relationships and management implications for fishes inthe southwestern USA. Pages 151–156 in G. J. Scrimgeour,G. Eisler, B. McCullock, U. Silins, and M. Morita, editors.Forest Lands – Fish II, Ecosystem stewardship throughcollaboration. Proceedings of the Forest Lands –Fish IIConference. Edmonton, Alberta, April 26–28, 2004.

Carmichael, G. J., J. N. Hanson, J. R. Novy, K. J. Meyer, and D.C. Morizot. 1995. Apache trout management: culturedfish, genetics, habitat improvements, and regulations.Pages 112–121 in H. L. Schramm, Jr. and R. G. Piper, edi-tors. Uses and effects of cultured fishes in aquatic ecosys-tems. American Fisheries Society, Symposium 15,Bethesda, Maryland.

Carmichael, G. J., J. N. Hanson, M. E. Schmidt, and D. C.Morizot. 1993. Introgression among Apache, cutthroat,and rainbow trout in Arizona. Transactions of the Amer-ican Fisheries Society 122:121–130.

Carter, C. D., and J. N. Rinne. In press. Short-term effects of thePicture Fire on fishes and aquatic habitat. Hydrology and

186

RINNE AND CALAMUSSO


Water Resources in Arizona and the Southwest 35,Flagstaff, Arizona.

Chronic, H. 1983. Roadside geology of Arizona. MountainPress Publishing Company, Missoula, Montana.

Chronic, N. 1987. Roadside geology of New Mexico. Moun-tain Press Publishing Company, Missoula, Montana.

Covington, W. W., and M. M. Moore. 1992. Change in forestconditions and multi-resource yields from ponderosapine forest since European settlement. Journal ofForestry 92:39–47.

Debano, L. F., G. J. Gottfired, R. H. Hamre, C. B. Edminster,P F. Ffolliott, and A. Ortega-Rubio, technical coordina-tors. 1995. Biodiversity and management of the MadreanArchipelago: the sky islands of southwestern UnitedStates and northwestern Mexico. U.S. Forest Service,Rocky Mountain Forest and Range Experiment Station,General Technical Report 264, Fort Collins, Colorado.

Duff, D. A. 1996. Conservation assessment for inland cut-throat trout status and distribution. U.S. Forest Service,Intermountain Region, Ogden, Utah.

Dunbier, R. 1968. The Sonoran Desert. Its geography, econ-omy, and people. University of Arizona Press, Tucson.

Green, C. R., and W. D. Sellers. 1964. Arizona climate. Uni-versity of Arizona Press, Tucson.

Gresswell, R. E., J. Kershner, and J. B. Dunham. 1997. Ecolo-gy and management of potamodromous salmonids.North American Journal of Fisheries Management17(4):1027–1028.

Harper, K. C. 1978. Biology of a southwestern salmonid, Salmoapache (Miller1972). Pages 99–111 in J. R. Moring, editor.Proceedings of the Wild Trout-Catchable Trout Sympo-sium. Oregon Department of Game and Fish, Eugene.

Hendrickson, D. A., W. L. Minckley, R. R. Miller, D. J. Siebert,and P. H. Minckley. 1980. Fishes of the Rio Yaqui basin,Mexico and the United States. Journal Arizona-NevadaAcademy of Science 15(3):65–106.

Hubbs, C. L., and R. R. Miller. 1948. The zoological evidence:correlation between fish distribution and hydrographichistory in the desert basins of the western United States.The Great Basin with emphasis on glacial and postglacialtimes. Bulletin of the University of Utah 38, BiologicalSeries 10:17–166, Salt Lake City.

Jaeger, E. C. 1957. The North American deserts. StanfordUniversity Press, Stanford, California.

Johnson, J. E., and J. N. Rinne. 1982. The Endangered SpeciesAct and southwest fishes. Fisheries 7(2):1–8.

Lee, D. C., and B. E. Rieman. 1997. Population viabilityassessment of salmonids by using probabilistic networks.North American Journal of Fisheries Management17(4):1144–1157.

Loudenslager, E. J., J. N. Rinne, G. A. E. Gall, and R. E. David.1986. Biochemical genetic studies of native Arizona andNew Mexico trout. The Southwestern Naturalist31:221–234.

Mello, K., and P. R. Turner. 1980. Status of Gila trout. U.S.Fish and Wildlife, Endangered Species Report 6, Wash-ington, D.C.

Miller, R. R. 1950. Notes on the cutthroat and rainbow troutswith the description of a new species from the Gila River,

New Mexico. Occasional Papers of the Museum of Zool-ogy University of Michigan 529:1–42.

Miller, R. R. 1954. A drainage map of Arizona. SystematicZoology 3:80–81.

Miller, R. R. 1961. Man and the changing fish fauna of theAmerican Southwest. Papers of the Michigan Academyof Science Arts and Letters 46:365–404.

Miller, R. R. 1972. Classification of the native trouts of Ari-zona with the description of a new species, Salmo apache.Copeia 1972(3):401–422.

Minckley, W. L. 1973. Fishes of Arizona. Arizona Game andFish Department, Phoenix.

Minckley, W. L., and J. E. Deacon. 1991. Battle against extinc-tion: native fish management in the American West. Uni-versity of Arizona Press, Tucson.

Nations, D., and E. Stump. 1981. Geology of Arizona.Kendall/Hunt Publishers, Dubuque, Iowa.

Platts, W. S. 1991. Livestock grazing. Pages 389–423 in W. R.Meehan, editor. Influences of forest and rangeland man-agement on salmonid fishes and their habitats. AmericanFisheries Society, Special Publication 19, Bethesda,Maryland.

Propst, D. L., and J. A. Stefferud. 1997. Populations dynam-ics of Gila trout in the Gila River drainage of southwest-ern United States. Journal of Fish Biology 51:1137–1154.

Propst, D. L., J. A. Stefferud, and P. R. Turner. 1992. Conser-vation and status of Gila trout, Oncorhynchus gilae. TheSouthwestern Naturalist 37(2):117–125.

Rieman, B. E., and J. D. McIntyre. 1993. Demographic andhabitat requirements for conservation of bull trout. U.S.Forest Service, Intermountain Region, General TechnicalReport 302, Ogden, Utah.

Rinne, J. N. 1981a. Problems associated with habitat evalua-tion of an endangered fish in headwater environments.Pages 202–209 in N. B. Armantrout, editor. Proceedingsof the Symposium on Acquisition and Utilization ofAquatic Habitat Inventory Information. American Fish-eries Society, Western Division, Portland, Oregon.

Rinne, J. N. 1981b. Stream habitat improvement and nativesouthwestern trouts. U.S. Forest Service, Rocky Moun-tain Forest and Range Experiment Station, ResearchNote 401, Fort Collins, Colorado.

Rinne, J. N. 1985a. Livestock grazing effects on southwesternstreams: a complex research problem. Pages 295–299 inR. R. Johnson and C. H. Lowe, editors. Riparian ecosys-tems and their management: reconciling conflicting uses.U.S. Forest Service, Rocky Mountain Forest and RangeExperiment Station, General Technical Report 120, FortCollins, Colorado.

Rinne, J. N. 1985b. Variation in Apache trout populations inthe White Mountains, Arizona. North American Journalof Fisheries Management 5:146–158.

Rinne, J. N. 1988. Native southwestern (USA) trouts: status,taxonomy, ecology and conservation. Polish Archives furHydrobiology 34(3–4):305–320.

Rinne, J. N. 1991. An approach to management and conser-vation of a declining regional fish fauna: southwesternUSA. Pages 56–60 in D. Matamura, B. Bobek, Y. Ono, W.Regelin, L. Bartos, and P. R. Ratcliffe, editors. Wildlife

187



conservation: present trends and perspectives for the21st century. Japanese Wildlife Research Center, Tokyo.

Rinne, J. N. 1994. Declining southwestern aquatic habitatsand fishes: are they sustainable? Pages 256–265 in W. W.Covington and L. F. Debano, editors. U.S. Forest Service,Rocky Mountain Forest and Range Experiment Station,General Technical Report 247, Fort Collins, Colorado.

Rinne, J. N. 1995a. The effects of introduced fishes on nativefishes: Arizona, southwestern United States. Pages149–159 in D. P. Philipp, editor. Protection of aquaticdiversity. Proceedings of the World Fisheries Confer-ence, Theme 3. Oxford & IBH Publishing Co. Pvt. Ltd.,New Delhi.

Rinne, J. N. 1995b. Interactions of predation and hydrologyon native southwestern fishes: Little Colorado spinedacein Nutrioso Creek, Arizona. Pages 33–38 in Hydrologyand Water Resources in Arizona and the Southwest, vol-umes 22–25. Proceedings of the 1995 Meetings of theArizona Section, American Water Resources Associationand the Hydrology Section, Arizona-Nevada Academy ofScience, Flagstaff, Arizona.

Rinne, J. N. 1995c. Rio Grande cutthroat trout. Pages 24–27in M. K. Young, technical editor. Conservation assess-ment for inland cutthroat trout. U.S. Forest Service,Rocky Mountain Forest and Range Experiment Station,General Technical Report 256, Fort Collins, Colorado.

Rinne, J. N. 1995d. Sky Island aquatic resources: habitats andrefugia for native fishes. U.S. Forest Service, RockyMountain Forest and Range Experiment Station, Gener-al Technical Report RM-264:251–260, Fort Collins, Col-orado.

Rinne, J. N. 1996. Short-term effects of wildfire on fishes andaquatic macroinvertebrates; southwestern United States.North American Journal of Fisheries Management16:653–658.

Rinne, J. N. 1998. Grazing and fishes in the southwest: con-founding factors for research. Pages 75–84 in D. Potts,editor. Proceedings American Water Resources AgencySpecialty Conference: rangeland management and waterresources. American Water Resources Association, Reno,Nevada.

Rinne, J. N. 1999. Fish and grazing relationships: the factsand some pleas. Fisheries 24(8):12–21.

Rinne, J. N. 2000. Grazing and fish relationships: southwest-ern United States. Pages 329–371 in R. Jemison, C.Raisch, and D. Finch, editors. Ecological and socioeco-nomic aspects of livestock management in the South-west. Elsevier Press, Amsterdam

Rinne, J. N. 2002. Hydrology, geomorphology and manage-ment: implications for sustainability of native southwest-ern fishes. Hydrology and Water Resources of Arizonaand the Southwest 32:45–50.

Rinne, J. N. 2003a. Fish habitats: conservation and manage-ment implications. Pages 277–297 in P.F. Ffolliott, M. B.Baker, L. F. Debano, and D. G. Neary, editors. Ecologyand management of riparian areas in the southwesternUnited States. Lewis Publishers, Boca Raton, Florida.

Rinne, J. N. 2003b. Native and introduced fishes: their status,threat, and conservation. Pages 194–213 in Ffolliott, P. F.,

M. B. Baker, L. F. Debano, and D. G. Neary, editors. Ecol-ogy, hydrology and management of riparian areas in thesouthwestern United States. Lewis Publishers, BocaRaton, Florida.

Rinne, J. N. 2003c. Wildfire in the southwestern UnitedStates: effects on fishes and their habitats. Proceedings ofthe 5th International Wildland Fire Symposium. Ameri-can Meterological Society, Boston.

Rinne J. N. 2004. Forest, fires and fishes: lessons and manage-ment implications from the southwestern USA. Pages151–156 in G. J. Scrimgeour, G. Eisler, B. McCullock, U.Silins, and M. Morita, editors. Forest lands – fish II,Ecosystem stewardship through collaboration. Proceed-ings of the Forest Lands –Fish II Conference. NaturalResources of Canada, Ottawa.

Rinne, J. N., and C. Carter. In press. Short-term effects ofwildfires on fishes in the southwestern United States,2002: management implications. Proceedings of theSymposium on Effects of Fire on Wildfire. U.S. ForestService, Pacific Southwest Research Station, Berkeley,California.

Rinne, J. N., and J. Janisch. 1995. Coldwater fish stocking andnative fishes in Arizona: past, present and future. Pages397–406 in H. L. Schramm, Jr. and R. G. Piper, editors.Uses and effects of cultured fishes in aquatic ecosystems.American Fisheries Society, Special Publication, Bethes-da, Maryland.

Rinne, J. N., and D. Miller. 2006. Hydrology, geomorphol-ogy, and management: implications for sustainability ofsouthwestern native fishes. Reviews in Fishery Science14:91–110.

Rinne, J. N., and W. L. Minckley. 1985. Patterns of variationand distribution in Apache trout (Salmo apache) relativeto co-occurrence with introduced salmonids. Copeia1985:285–292.

Rinne, J. N., and W. L. Minckley. 1991. Native fishes of aridlands: a dwindling natural resource of the desert south-west. U.S. Forest Service, Rocky Mountain Forest andRange Experiment Station, General Technical Report206, Fort Collins, Colorado.

Rinne, J. N., W. L. Minckley, and J. N. Hanson. 1981. Chem-ical treatment of Ord Creek, Apache County, Arizona, tore-establish Arizona trout. Journal of the Arizona-Neva-da Academy of Sciences 16:74–78.

Rinne, J. N., and D. G. Neary. 1996a. Fine sediment and fish:indicators of natural and man-induced landscape distur-bances. Pages IX-1–IX-8 in J. Barnard, technical coordi-nator. Proceedings of the 6th Federal Interagency Sedi-mentation Conference, Volume 2. U.S. GovernmentPrinting Office, Washington, D.C.

Rinne, J. N., and D. G. Neary. 1996b. Fire effects on aquatichabitats and biota in Madrean-type ecosystems: south-western USA. Pages 135–145 in U.S. Forest Service,Rocky Mountain Forest and Range Experiment Station,General Technical Report 289, Fort Collins, Colorado.

Rinne, J. N., L. Riley, B. Bettaso, K. Young, and R. Sorensen.2004. Managing southwestern native and nonnative fish-es: can we mix oil and water and expect a favorable solu-tion. Pages 117–138 in M. J. Nickum, P. M. Mazik, J. G.

188

RINNE AND CALAMUSSO


Nickum, and D. D. MacKinlay, editors. Propagated fishesin resource management. American Fisheries Society,Symposium 44, Bethesda, Maryland.

Rinne, J. N., B. Robertson, R. Major, and K. Harper. 1979.Sportfishing for the native Arizona trout, Salmo apacheMiller in Christmas Tree Lake: a case study. Pages158–164 in W. King, editor. Proceedings of the WildTrout II Symposium. Trout Unlimited, Reston, Virginia.

Rinne, J. N., R. Sorenson, and S. C. Belfit. 1985. An analysisof F1 hybrids between Apache, (Salmo apache) and rain-bow trout, (Salmo gairdneri). Journal of the Arizona-Nevada Academy of Sciences 20:63–69.

Rinne, J. N., and J. A. Stefferud. 1998. Single versus multiplespecies management: native fishes in Arizona. ForestEcology and Management 14(2–3):357–365.

Rinne, J. N., and P. R. Turner. 1991. Reclamation and alter-ation as management techniques, and a review ofmethodology in stream renovation. Pages 219–244 in W.L. Minckley and J. E. Deacon, editors. Battle againstextinction—native fish management in the AmericanWest. University of Arizona Press, Tucson.

Scurlock, D. 1998. From the Rio to the Sierra: an environ-mental history of the middle Rio Grande basin. U.S. For-est Service, Rocky Mountain Forest and Range Experi-ment Station, General Technical Report 5, Fort Collins,Colorado.

Stumpff, B., and J. Cooper. 1996. Rio Grande cutthroat,Oncorhynchus clarki virginalis. Pages 74–86 in D. Duff,editor. Conservation assessment for inland cutthroattrout. U.S. Forest Service, Intermountain Region,Ogden, Utah.

Sublette, J. E., M. D. Hatch, and M. Sublette. 1990. The fish-es of New Mexico. University of New Mexico Press,Albuquerque.

Turner, P. R. 1996. Gila trout recovery: a case study in main-taining biodiversity. New Mexico Journal of Science36:152–167.

USDI (U.S. Department of the Interior). 1979. Gila troutrecovery plan. U.S. Fish and Wildlife Service, Albu-querque, New Mexico.

USDI (U.S. Department of the Interior). 2004. Draft Apachetrout recovery plan. U.S. Fish and Wildlife Service, Albu-querque, New Mexico.

Wallace, R. L., and R. J. Behnke. 1974. Report on recent col-lections of Rio Grande cutthroat trout: evaluation of rel-ative purity of populations in recent collections. Col-orado Cooperative Fisheries Unit, Fort Collins.

Warshall, P. 1995. The Madrean Sky Island Archipelago: aplanetary overview. Pages 6–18 in L. F. DeBano P. F. Ffol-liott, A. Ortega-Rubio, G. J. Gottfried, R. H. Hamre, andC. B. Edminster, technical coordinators. Biodiversity andmanagement of the Madrean Archipelago: the SkyIslands of southwestern United States and northwesternMexico. U.S. Forest Service, Rocky Mountain Forest andRange Experiment Station, General Technical Report264, Fort Collins, Colorado.

Wernsman, G. R. 1973. Systematics of native Colorado cutthroattrout. M.S. thesis. Colorado State University, Fort Collins.

Young, M. K. 1996. Conservation assessment for inland cut-throat trout. U.S. Forest Service, Rocky Mountain Forestand Range Experiment Station, General TechnicalReport 256, Fort Collins, Colorado.

189



191

The bull trout Salvelinus confluentus popula-tions of the upper Tieton River basin aboveTieton Dam, which impounds Rimrock

Lake, are among the many Columbia basin fishpopulations listed as threatened under the federalEndangered Species Act (ESA). The ESA requiresthe U.S. Fish and Wildlife Service (USFWS) tominimize human actions that jeopardize popula-tion persistence and develop a strategy that willrecover the species to a viable population size. Thisdocument describes a life cycle model and simula-tions developed to assess the impacts in terms ofnumbers of individuals in the population from

flushing (entraining) subadult bull trout past theimpassable Tieton Dam and the benefits in termsof individuals in a population of building fish pas-sage facilities on the dam.

Bull trout in Rimrock Lake typically follow anadfluvial life history in which subadults and adultsrear in the reservoir, while spawning and juvenilerearing are confined to two tributaries, IndianCreek and the South Fork Tieton River (Figure 1).Indian Creek is dominated by springs and provides8.2 km of bull trout habitat. South Fork TietonRiver is glacier fed and provides 20 km of bull trouthabitat (MBI 2004). Adult bull trout enter these

Simulation of Human Effects on Bull Trout PopulationDynamics in Rimrock Reservoir, Washington

ABSTRACT A life cycle model was employed to identify the response of an adfluvial bull trout Salvelinusconfluentus population to chronic and catastrophic losses of subadults. The model simulates the bull trout popu-

lation within Rimrock Lake, Washington, a reservoir on the Tieton River impounded by Tieton Dam. Subadult bull

trout are entrained during summer water releases for irrigation, and the dam has no fish passage facilities to

enable those fish to return upstream. Suitable spawning and rearing habitat is primarily upstream of the dam. Tag-

ging studies of adult bull trout passing weirs in the two major tributaries to the reservoir were used to estimate

model parameters for survival, maturity rates, reproductive capacity, and initial abundance. Sampling data and

the deterministic model simulations indicated that the population was capable of rebounding quickly from inter-

mittent catastrophic events. Resilience of the bull trout population resulted from high adult longevity and repeat

spawning. The accumulation of mature adults across multiple age-classes led to egg deposition that fully seed-

ed rearing capacity of the natal tributaries, even when several consecutive broods of juveniles exhibited poor sur-

vival. Catastrophic events simulated to entrain 50% of the subadults every 15 years caused a 40% reduction in

adult abundance within 4 years of the event, followed by a full recovery to maximum production within 9 years.

Even during the low point of adult abundance, 15 times more eggs were deposited than were needed to fully seed

juvenile tributary habitat. Because the population was already producing juveniles at capacity, simulations for

opening fish passage over Rimrock Dam showed that the sustained spawner population would increase by only

two adults. However, simulations for expanded juvenile habitat by 14% indicated that the sustainable adult pop-

ulation would increase by 14%. Thus, habitat improvements that target juvenile rearing capacity appear to have

the greatest potential to increase population size.

Keith Underwood and Steve P. Cramer



KEITH UNDERWOOD S. P. Cramer and Associates, 7802 Onyx Court SW, Lakewood, Washington 98498, USA.Current address: HDR|FishPro, 3780 SE Mile Hill Drive, Port Orchard, Washington 98366, USA; [email protected]

STEVE P. CRAMER Cramer Fish Sciences, 600 NW Fariss Road, Gresham, Oregon 97030, USA


streams during May through August and spawn inAugust through October. Juveniles typically rear2–3 years in their natal area and then migrate to thelake, where they become piscivorous, grow rapidly,and mature at age 5 or 6.

Rimrock Lake is one of five reservoirs for water

storage within the Yakima IrrigationProject along with Clear Lake on theTieton River and Keechelus, Kachess,and Cle Elum lakes on the YakimaRiver (Figure 1). The Tieton Riverenters the Naches River, which is amajor fork of the Yakima River. Rim-rock Lake is managed to provide irriga-tion water and, secondarily, to addressflood control and recreation needs.Since 1981, water stored in the reser-voir has been released predominantlyduring August through October as partof a coordinated plan with other Yaki-ma Irrigation Project reservoirs to sup-ply irrigation water to areas down-stream (USBR 2004; Figure 2).

Area biologists have expressedconcerns that entrainment of bulltrout in reservoir outflow may placethe populations at risk and that lack offish passage at Tieton Dam mayincrease genetic and demographicrisks to the Rimrock populations.

Entrainment studies during the summer of 2000suggest that roughly 90,000 kokanee Oncorhynchusnerka and 145 subadult bull trout were entrained inthe reservoir outflow (Hiebert et al. 2003). James(2000, 2002) estimated from mark–recapture stud-ies that there were an average of 1,700 adult bull

trout leaving the tributaries (postspawn) and reentering the lake. Thedraft USFWS Bull Trout RecoveryPlan suggested that population con-nectivity is paramount to recoveringthe species (USFWS 2002). Since theTieton Dam does not allow entrainedadults to migrate back upstream (nofish ladder), entrained fish cannotreproductively contribute to the Rim-rock population. The bull trout lifecycle model was used to identify theeffect of chronic and catastrophicentrainment with and without adultdam passage back into Rimrock Lake.

Data were available to reconstructestimates of juvenile bull trout popu-lation abundance. The abundance of

192

UNDERWOOD AND CRAMER

Figure 1. Rimrock Lake, Tieton River, and Yakima basin tributaries andreservoirs.

Figure 2. Comparison of stream flows in the Tieton River below Tieton Damduring three periods: reconstructed natural stream flows for 1981–2003,actual mean flows during 1925–1980, and actual mean outflows during1981–2003. U.S. Geological Survey data.


spawning bull trout was estimatedfrom data obtained during the mid-1990s in the two principal tributarieswhere bull trout spawn, Indian Creekand the South Fork Tieton River(James 2000, 2002; P. James, CentralWashington State University, unpub-lished data). The model demonstratesthe effects of factors limiting bull troutpopulation size, survival, and propen-sity to rebound after a catastrophicevent while suffering from chronicanthropogenic impediments to sur-vival (e.g., harvest).

MethodsModel Framework

Bull trout live in a dynamic ecosystemwith multiple environmental factorslimiting survival at each life stage. Adeterministic, density-dependent lifecycle model was employed to predictpopulation size and survival overnumerous generations in the presence ofmultiple limiting factors (Beissinger andMcCullough 2002; Morris and Doak2002). The life cycle model presentedherein is an enhanced version of thepopulation models outlined by Taylor(1981) and Walters (1969), with the addition of repro-duction and recruitment processes (Beamesderfer1991). The model is written in Microsoft Excel, allow-ing users to easily inspect model formulation and alterparameter values to simulate “what if” scenarios.

The life stages acknowledged by the life cyclemodel depended strongly on available data fromthe Rimrock population and literature. Adult markand release, entrainment estimates, and age-classstrength data were sufficient to include the follow-ing life stages in the model: egg to age 1; age 1 toage 2 rearing in the natal stream; subadult rearingin the reservoir; and mature adults that migrate upthe tributaries to spawn and then return to thereservoir shortly after spawning (Figure 3). Juvenilerearing in the natal stream was assumed to lastthrough two summers, with juveniles migrating tothe reservoir in their third spring (age 2).

Once juveniles entered the reservoir, survival wasassumed to be constant (density independent), andgrowth was rapid until maturing at age 5. Maturefish spent the early summer months in their natalstream, August and September on their spawningrun, and reared in the reservoir for the rest of theyear. Survival of mature fish was high and constant,which allowed many fish to spawn multiple times.Length of mature fish increases with age and thenumber of eggs per female increases with length.

In the model, entrained fish are tracked separate-ly from the lake population to simulate effects ofentrainment on the lake population and predict thepotential contribution of the entrained fish to thelake population spawners. Entrained fish wereassumed to represent the two tributary populationsin the same proportions they occur in the reservoir.

Three types of anthropogenic mortality were

193

SIMULATION OF HUMAN EFFECTS ON BULL TROUT POPULATION DYNAMICS

Figure 3. Life cycle model flow chart of Rimrock bull trout population fol-lowed by the model.


assessed independently for subadults and adults.These were entrainment (both chronic and cata-strophic), upstream fish passage, and angler har-vest. The model allowed entrained fish to migrateback over the dam to simulate the installation of afish ladder. We assumed that all entrained fishattempted to migrate back upstream in the spring.Entrained fish that did not migrate back over thedam were able to rear, but not successfully repro-duce in the tailrace. Kalin et al. (2002) determinedthat warmwater temperatures below the damwould prohibit successful egg incubation.

Model Initialization

The first year of the recorded simulation began aftera 30-year (~6 generations) initialization periodbefore anthropogenic causes for mortality wereassessed. Initialization of the model required an esti-mate of adult spawners and contributed eggs. Thesewere based on mark–recapture population estimatesin South Fork Tieton River and Indian Creek during1997 through 1999. The number of age-5 and greateradults varied annually, but when estimates from bothSouth Fork Tieton River and Indian Creek were com-bined, an average of 1,700 adults migrated out of thestreams back into Rimrock Lake in the fall (James2000, 2002; Table 1). The first 5 years (one genera-tion) of the 30-year initiation period required manu-ally seeding the model with eggs until a full genera-tion of fish were producing eggs within the model.The model was initiated with 3,000,000 eggs, asexpected from the number of redds observed annu-ally. After the fifth year, the model estimated thenumber of eggs deposited in redds annually.

Natural SurvivalJuveniles (Egg to Age 1)

Egg to age-1 survival was estimated from valuesobserved in bull trout studies outside of the Yaki-ma basin. In Montana, Fraley and Shepard (1989)reported bull trout egg-to-emergence survival tobe 50% in two different streams. Estimated bulltrout survival ranged from 28% to 86% for age 0to age 1 over 5 years of study in Kemess Creek,British Columbia (Bustard 2001). Conversely,Downs and Jakubowski (2003) estimated a 0.04%to 0.20% egg to out-migrant (ages 1–3) survival inTrestle Creek, Idaho of a Lake Pend Oreille adflu-vial population. Lestelle et al. (2004) conducted aliterate review and determined that 8.5% was anappropriate egg to age-1 survival for modelingpurposes. For simulations, we assumed a constantegg to age-1 survival of 10%; by doing so, wefocused the model on the effect of entrainment.Greater survival rates would make the populationmore robust and resilient to entrainment effects,and overly pessimistic estimates produced insup-portably small simulated populations when com-pared to adult population observed in the streams.The goal of this analysis was to quantify the poten-tial negative effects of entrainment; therefore, wechose a value that appeared to reflect a reasonableassumption. Hence,

where Nt is number of eggs at year, r is survival at10%, and Nt + 1 is the number of age-1 fish.

194


Table 1. Tagging, recovery, and estimated populations for bull trout spawners in Indian Creek, based on postspawn downstreammigrants trapping. Raw data from Paul James, Central Washington State University, Ellensburg.

1994 1995 1996 1997 1998 1999 2000

New tags (#) 47 73 268 161 92 101 73

Recap tags (#) – 4 36 160 110 110 161

Total tags – 77 304 321 202 211 234

Old tags (%) – 5% 12% 50% 55% 52% 69%

Population estimated – 658 ± 202 868 ± 235 658 ± 58 1,067 ± 311 800 ± 259

% population sampled – 12% 35% 49% 19% 26%


Juveniles (Age 1–2)

Studies have demonstrated that survival from age1 (parr) to age 2 (out-migrant) is density depend-ent in adfluvial bull trout populations. Paul(2000) found that density affects survival of juve-nile bull trout by up to 60% across a wide range ofdensities observed over 15 years of an adfluvialbull trout studies in Eunice Creek, Alberta.

Paul (2000) also found a significant negativerelationship between resident juvenile bull troutgrowth and density based on fish size in PrairieCreek (1995–1999), but no relationship to densi-ty based on the number of fish. Consumption rateis an exponential function of body size, with fishconsuming less per body weight as they increasein size (Walters and Post 1993; Post et al. 1999).Our interpretation of Paul’s findings are that theoverall population growth rate is dependant onthe size-class structure of the population, wherethe abundance of larger juveniles has a greatereffect on bull trout growth rate than the abun-dance of smaller juveniles.

Experiments conducted by enclosing streamsections were used to demonstrate that growth ofage-1 bull trout was density dependent (Paul2000). A 42-d experiment did not detect differ-ences in survival among experimental treatments,but there were highly significant (P < 0.01) differ-ences in growth. Further, fish size at the beginningof the experiment was positively correlated withsurvival at the end of the experiment.

Given that age-1 to age-2 survival was densitydependent, we used a Beverton–Holt function todescribe density dependent survival at this lifestage (Beverton and Holt 1957). The function hastwo parameters: (1) maximum survival thatwould occur at the lowest densities; and (2) carry-ing capacity, which represents the maximumnumber of age-2 fish that can be produced(asymptote). The maximum survival at lowestdensities was assumed to be 50% based on sur-vival estimates from Paul (2000) for age-2 bulltrout in Eunice Creek. The age-2 (out-migrantstage) capacity was selected to be the value thatwould result in an equilibrium adult populationof 1,700 adults. The adult abundance of 1,700 isan approximation of the stable spawner escape-

ment observed by (James 2002) during the years1998–2000 in Indian Creek and South ForkTieton River combined. Based on the survivalassumptions presented in this document, the age-2 capacity was estimated to be 7,200 individuals.This estimate appears to be reasonable for thefollowing reason. Underwood and Cramer(2004) determined the maximum and averagenumber of bull trout by age-class and habitatunit type in the Tucannon River and Mill Creek,Washington, using habitat unit multiple passdepletion methods. We applied those maximumand averages observed densities to Indian Creekand South Fork Tieton by habitat unit area asdescribed by MBI (2004). The rearing surfacearea of the tributaries was estimated to be236,957 m2. The maximum age-2 carrying capac-ity was estimated at 12,388 individuals, and theaverage was 4,805 individuals. Based on thisinformation, the 7,200 age-2 population assump-tion appeared to be reasonable.

Density-dependent survival is assumed tooccur during the stream rearing phase and followsthe Beverton and Holt (1957) function:

where α is the maximum survival rate at low den-sity, β is the carrying capacity of age-2 fish, Nt isthe number of age-1 parr, and Nt +1 is the num-ber of resulting age-2 fish.

Subadults (Age 3–4)

Survival was assumed to be constant at 40% peryear for subadults in the lake. Paul (2000) esti-mated annual survival rates of age-3 bull troutbetween 10% and 100% over a 15-year period.The median (as interpreted from a scatter plot)was approximately 40%. Philip Howell (USFWS,LaGrande, Oregon) provided 5 years of unpub-lished passive integrated transponder tag data col-lected in Mill Creek, Washington, which was usedby Underwood and Cramer (2004) to predict age-3 and -4 survival ranging between 35% and 45%.The formula used to estimate the number of eggssurviving to age 1 is also used to estimate age-3and -4 survival, with the exception that r is 40%.

195



Adults (Age 5 and Older)

Data from a 5-year mark and recapture study ofadult bull trout in Indian Creek were used to esti-mate survival between age-classes for the lake pop-ulation. Estimated survival rates from the first yearof tagging to repeat spawning the next year werenear 100% (James 2000, 2002; Table 1). Survivalestimates for subsequent years were highly variable(probably a reflection of small sample size), butgenerally ranged from 45% to more than 100%.Figure 4 aligns the percent of capture by age, incomparison to a 65% survival rate. The two curveshave a strong resemblance, so 65% survival wasassumed in the model. The survival rate of theentrained population was unknown and thereforeassumed to be the same as the lake population.

Length-at-Age Relationship

Length at age was estimated to determine survivalby age and fecundity of females. Tagged and

recaptured bull trout adults in Indian Creek wereused to determine length at age and length of firstyear spawners (James 2002). First year spawnerswere assumed to be age 5, consisting mostly offish greater than 500 mm. Growth incrementsafter first spawning were determined by regres-sion of annual growth increment to fish length.Growth increments were determined for taggedfish recovered in successive years at the IndianCreek weir. The resulting regression formula wasgrowth increment (mm) = –0,1207 (total length)+ 99.984 (n = 196; R2 = 0.34), indicating thatgrowth increments declined as the fish grew inlength. For age-5 fish (>500 mm in length), theaverage annual growth increment was 40 mm.Lengths in years after age 5 were estimated bysequentially adding the growth increment pre-dicted by the regression of spawner size in year ton the growth increment by return in year t + 1.This growth rate and resulting size was applied tothe lake population (Table 2).

Lengths of the entrained popula-tion were believed to differ from theadfluvial lake population; however,entrained fish lengths were not avail-able for fish downstream of the lake.Literature values were compared tobetter understand the relationshipbetween fluvial and adfluvial lengthsat age. Length at age for the fluvial lifehistory strategy was based on averagevalues from the following locationsand reports: McLeod River andEunice Creek, Paul (2000); andAnderson Creek and Sheep Creek,Hunt et al. (1997). Adfluvial length atage was an average of values from thefollowing locations and reports: Flat-head Lake, Leathe and Graham

196


Table 2. Bull trout predicted length at age for fish captured in Indian Creek weir. Raw data from Paul James, Central Washington StateUniversity, Ellensburg.

5 6 7 8 9 10 11 12

Lake 500 540 575 606 634 658 679 698

Entrained 300 350 400 460 520 580 620 660

Figure 4. Mean adult catch composition by age based on bull trout taggedand recaptured at the mouth of Indian Creek during the period 1994–2000and assumed 65% annual survival rate. Raw data from P. James, CentralWashington State University (personal communication).


(1982); North and Middle Fork Flat-head River, Fraley et al. (1981); LakeKoocanusa, May et al. (1979); PriestLake, Bjornn (1961); Lake Billy Chi-nook and Metolius River, Pratt(1991); and Thutade Lake, Bustard(1999). Growth rates differedbetween fluvial and adfluvial fish, sowe assumed that any entrained fishrearing below Tieton Dam wouldgrow at rates typical of fluvial fish(Table 2).

Fecundity

The number of eggs produced perfemale (fecundity) was determinedto increase exponentially with fish length based ona compilation of fecundity and length data fromthe McKenzie Creek, BC; upper Flathead River,Montana; Bull River, Montana; Clark Fork, Mon-tana; and Sun Creek, Oregon (Wallis 1948; Brun-son 1952; McPhail and Murray 1979; Fraley andShepard 1989; McPhail and Baxter 1996). Theregression analysis provided fecundity at a specif-ic length, which was applied to both lake andentrained populations (Figure 5; eggs per female =0.0005 * (fish length)2.4653; N = 19; R2 = 0.8993).The life cycle model tracks number of spawningfemales by age, not by length. Thus, the fecundityequation relied on the length at age reported inTable 2. The model determines the number of eggsper female based on the length and fecunditycurve. Multiplying the number of eggs per femalefor each age-class by the number of females in theage-class and summing across all age-classes pro-vided the total number of eggs deposited.

Percent Female Spawners

The number of spawning females drives the pop-ulation size since population persistence dependson egg production. A single male can fertilize theeggs of many females. Therefore, it is assumedthat the number of returning males does not limitthe production of fertilized eggs. Rather, thenumber of eggs produced per female and thenumber of spawning females defines initial popu-

lation size. The adult males to females ratio wasfrom data collected in South Fork Tieton Riverand Indian Creek (James 2000, 2002). From atotal of 389 fish, 103 were classified as males and286 as females. Thus, roughly 70% of theobserved fish were female and were applied to themodel. The Indian Creek tagging studies suggestthat the age at first spawn was 5 years. IndianCreek ages were estimated from length-at-agerelationships, since age was not known during theIndian Creek study.

The percentage of the Rimrock adult popula-tion that spawn in alternate years was unknown.Based on tag studies conducted on adfluvial bulltrout in Pend Oreille Lake, Idaho, roughly 7% ofthe population was estimated to be alternateyear spawners (Downs and Jakubowski 2003).We assumed that 7% of the Rimrock adult pop-ulation and entrained population spawned inalternate years.

The formula to determine egg deposition is asfollows:

where E is the number of eggs deposited, Ntc isthe number of adults, Np is the percent of avail-able population spawning (inverse is alternatespawners), Fp is the percent of spawning adultsfemale, 0.0005 is the intercept of the length to

197


Figure 5. Bull trout fecundity (eggs per female) related to fish length, basedon data from Pacific Northwest populations.


fecundity regression, Lntc is fish length at age, and2.4653 is the slope of the length regression curve.

Entrainment

Entrainment out of Rimrock Lake downstream ofthe Tieton Dam was simulated in two ways: (1)“chronic entrainment,” a steady annual rate; or(2) “catastrophic entrainment,” an interannualpunctuated event. Chronic entrainment movesreservoir subadults (ages 3–4) and adults (ages5–12) downstream of the dam annually, based onan entrainment rate. Catastrophic entrainmentmoved subadults and adults out of the reservoirdownstream of the dam based on an entrainmentrate and yearly intervals.

Entrainment, either through the chronic orcatastrophic path, was viewed as a loss to the lakepopulation, unless the model was set up to simu-late an upstream fish passage device allowing thefish to reenter the lake population. We assumedthat once fish are entrained, they lived the rest oftheir lives below the dam in the Tieton River. Theentrained population lived a fluvial life historystrategy. The fluvial fish also experienced differ-ential growth rates in relation to the lake popula-tion. The model allowed for separate entrain-ment rates to be applied to subadults (age 3–4),and adults (age 5 and above). Entrainment ofsubadults was believed to be far more likely thanadults because most adults were in their spawn-ing tributaries during mid-August through Sep-tember. Furthermore, adults were not captured inentrainment studies at the Tieton Dam (Hiebertet al. 2003).

The rate of annual chronic entrainment was notknown. However, the 2002 study estimatedentrainment of roughly 145 subadults from Sep-tember through mid-October (Hiebert et al. 2003).This estimate was assumed to be the minimumannual entrainment rate from discharging waterthrough the dam over the entire year. The baselinesimulation scenario used a chronic entrainmentrate at 1.4% for subadults and 0.5% for adults,which resulted in entraining 148 subadults and 9adults annually. The baseline analysis did notinclude catastrophic entrainment.

Rimrock reservoir has been drawn down to

10,000 acre-feet or less six times since 1925, andthe low drawdowns were assumed to contributeto entrainment by aligning midwater column fishdistribution with the dam intake tower. As aresult, a recurrence interval of 15 years wasassumed to approximate the typical occurrencerate of drawdown. The percent of subadults andadults entrained from catastrophic events wasindependent of chronic entrainment. Combined,chronic and catastrophic entrainment rates cap-ture both annual average entrainment and punc-tuated entrainment associated with climatic con-ditions such as drought or high runoff, whichinfluence Rimrock Lake hydrodynamics andmanagement.

Entrained Fish Survival

Entrainment occurs when fish move from RimrockLake to the Tieton River via the outlet works at thedam. Fish were entrained through the dam out ofthe forebay at an intake tower standing 50 ft abovethe lake bed and 150 ft below the lake surface at fullpool. Fish entered the intake tower, moved througha conduit, and were expelled at the base of the damthrough a jet valve. The jet valves blast water hori-zontally to the river surface, dissipating energyprior to reaching the tailrace. Fish enter the outletstructures at a depth of 75–100 ft below the lakesurface. Due to the large change in depth and ele-vation, as fish pass through the outlet structures,entrained fish were exposed to a large pressure dif-ferential. At high lake surface levels, fish in the out-let works pass through the jet valve at more than100 ft/s and experience a pressure difference of upto 80 lb/in2, equivalent to a pressure difference ofroughly 5 atmospheres (Kalin et al. 2002). A pres-sure difference of this magnitude was likely tocause gas bubble trauma in fish passing throughthe outlet works. Gas bubble disease is similar tothe “bends” experienced by human divers who sur-face too quickly. It occurs when the dissolved gasin the fish’s tissues is not in equilibrium with thesurrounding water and effervesces out of solutioncausing nitrogen bubbles in the tissues and blood.These bubbles block blood flow and cause tissuedamage. Based on an analysis of pressure differen-tial from the intake tower to outlet works and per-

198



cent kokanee survival from netting downstream ofthe dam, Kalin et al. (2002) concluded thatentrained fish mortality was near 60% and may beas high as 80% when the effect of mechanicallyinduced injury, such as hitting the jet valve hous-ing, was considered. In the model simulation, weassumed a survival rate of 40% for entrained fish.In addition, the model allows for the simulation offish migrating upstream over a ladder to simulatethe effect of reconnecting the entrained popula-tion with the lake population. We assumed that afish ladder would allow 90% of the entrained pop-ulation to migrate upstream back into the lake.

Harvest

Harvest of bull trout has been prohibited; how-ever, harvest likely occurs due to poaching andincidental take. Rimrock Lake supports a popu-lar kokanee fishery that received significantangling effort. We assumed that the kokaneefishery resulted in an incidental take of 4%, splitequally among adults and subadults (2% each).

Results and DiscussionEntrainment

The baseline simulation chronically entrained1.4% of the subadults and 0.05% adults with 0%catastrophic entrainment and 2% harvest for

subadults and adults. This baseline simulationpredicted entrainment of 9 adults and 148subadults annually. The adult lake population sizewas estimated to be 1,677 individuals. This simu-lation suggested that the population was stableand produced roughly 20 times the number ofeggs necessary to maintain the adult populationsize of 1,677 adults, suggesting that the populationwas resilient to perturbation.

The resiliency of the Rimrock bull trout popu-lation was tested by simulating catastrophicentrainment. Simulations were run with unreason-ably low, pessimistic survival rates in order toemphasize the effect of punctuated events on thepopulation. Simulated entrainment events thatkilled 50% of all subadults (age 3–4) living withinthe reservoir every 15 years caused a 40% reduc-tion in adult abundance within 4 years of the event,followed by a full recovery to maximum produc-tion within 9 years (Figure 6). This simulation didnot entrain adults because the catastrophicentrainment was believed to occur during Augustand September while adults were in the tributariesspawning. Notably, this scenario suggested thatroughly 15 times more eggs were deposited thanwere needed to fully seed the stream with juvenilesduring the lowest population period. Thus, thepopulation remained highly resistant to cata-strophic events even during years maintaining thelowest simulated adult abundance.

199


Figure 6. A 50-year simulation of adult bull trout abundance in Rimrock Lake, assuming that 50% of the subadult popula-tion was lost to entrainment every 15 years. There is no upstream passage for adults.


The simulated population recovered quicklybecause repeat spawners continued to fully seedthe stream to rearing capacity, and juveniles rearingwithin the streams supplied the reservoir withjuveniles during the 2 years following the event.Population recoveries from catastrophic events,such as the one simulated, do occur in nature. Paul(2000) reported from a 15-year study of fluvial bulltrout in Eunice Creek, Alberta, an absence ofspawning adults in 1979–1981 and a correspon-ding absence of age-1 juvenile in 1981–1983. Inspite of these absences, juvenile densities reachedan all-time high in 1985.

An additional simulation was run thatdecreased the time between catastrophic en-trainment events from 15 to 2 years (Figure 7).Losing 50% of the subadults every 2 yearsreduced the population size from an average of1,677 to 628 adults. However, the population wasable to maintain a stable population size, indi-cating an ability to resist extinction underextremely harsh environment conditions causingpoor subadult survival.

The 15-year interval was believed to be the bestrepresentation of the current environment. Every15 years, the lake elevation is drawn down to anactive storage of 10,000 acres. During theseextreme drawdowns, the intake tower is within 75ft of the lake surface and at a lake depth believed to

contain a majority of the fish. However, we doubtthat 50% of the subadult population has ever beenentrained. Hiebert et al. (2003) estimate thatroughly 150 subadults were entrained duringAugust through September of 2002. Steve Hiebert(U.S. Bureau of Reclamation, personal communi-cation) also indicated that entrainment estimatesfor 2003 were similar to those reported for 2002.Based on model simulations, 148 fish represented1.4% of the subadult population. A 50% subadultentrainment rate (a rate we simulated) wouldcause roughly 8,000 subadults to be entrainedevery 15 years. Hence, we conclude that the Rim-rock bull trout population is capable of sustainingitself even at high entrainment rates, and currentlyobserved rates were far below a level that wouldjeopardize the population.

Upstream Fish Passage

The draft USFWS Bull Trout Recovery Plan sug-gested that population connectivity is paramountto recovering the species (USFWS 2002). SinceTieton Dam does not allow entrained adults tomigrate back upstream (no fish ladder), entrainedfish cannot reproductively contribute to the Rim-rock population when they mature. This situationmay negatively affect the reservoir population.

The effect of a hypothetical fish ladder thatallowed entrained adults to migrate back into the

200


Figure 7. A 50-year simulation of adult bull trout abundance in Rimrock Lake, assuming that 50% of the subadult popula-tion was lost to entrainment every 2 years. There is no upstream passage for adults.


reservoir was simulated. The percent of adultsmigrating upstream past the dam was assumed tobe 90% annually. We also assumed that allpostspawn adults migrated back downstream ofthe Tieton Dam and then returned annually untildeath. When the model was set up to represent thebaseline condition, no catastrophic entrainmentoccurred and 40% of the entrained fish survivedpassage through the dam outlet works. Under thisscenario, 31 adults migrated upstream of the dam,less than 1% of the egg production was supportedby the returning entrained fish, and 22 times asmany eggs were produced under this scenario thannecessary to sustain the adult population.

Catastrophic entrainment was then added tothe upstream passage simulation. Under theseconditions, the number of entrained adultsmigrating back upstream over the dam averaged98 adults and fluctuated between 75 and 319adults over the 50-year simulation (Figure 8).These entrained spawning adults contributed toegg production on average 3% ranging between1% and 6%, and egg production far exceeded thenumber necessary to sustain the adult popula-tion. Based on the model simulation, the lakesupports ample adults to sustain the populationand a ladder would not significantly ameliorateentrainment impacts. This is due in part to thelow survival rate of entrained fish and the over-

abundance of spawners and eggs. From a demo-graphic (population size) prospective, the addi-tion of a fish ladder would not significantly ben-efit the population.

Genetic Prospective

The draft USFWS Bull Trout Recovery Plan sug-gested that population connectivity is required inorder to recover the species (USFWS 2002). From agenetics viability prospective, the Tieton Dam doesnot allow entrained adults to migrate backupstream, and therefore, mature entrained fishcannot reproductively contribute to the Rimrockpopulation limiting the effective population size.Furthermore, the recovery plan suggests that fishfrom other in-basin populations are unable tospawn with the Rimrock population, which report-edly limits gene flow.

The accuracy of these theorized negative affectsof Tieton Dam on Rimrock bull trout depends on(1) whether the degree of entrainment reduceseffective population size in the reservoir to a levelthat negatively affects the impounded population,and (2) whether gene flow previously existedacross areas now fragmented by Tieton Dam. IfTieton Dam did not interrupt predam gene flow, assome genetic data suggest, then reconnection ofareas above and below the dam will not affect geneflow (Reiss 2003).

201


Figure 8. A 50-year simulation of adult bull trout abundance in Rimrock Lake, assuming that 50% of the subadult populationevery 15 years were entrained to below the Tieton Dam, of which 40% survived and migrated upstream with a 90% success.

Num

ber

of

adul

t fis

h


Gene Flow and Genetic Variability

The rationale behind population connectivity forbull trout comes from the well-known metapopu-lation concept, dating back to the early 1900s(reviewed in Hanski and Gilpin 1997) and appliedto bull trout by Rieman and Dunham (2000) andRieman and McIntyre (1993). In simple terms, thisconcept suggests that maintenance of metapopula-tion structure depends in part on the ability ofadjacent subpopulations to rebuild each other dur-ing asynchronous periods of low survival and smallpopulation size. The USFWS recovery strategyrelies heavily on this concept to recover the species(USFWS 2002). As a result, the draft recovery planrecommends barrier removal or modification tofacilitate active bull trout migration past barriers.Although this approach may be logically applied tosubpopulations that exhibit gene flow andmetapopulation structure, no such functions areevident in the Rimrock populations.

Recent bull trout genetic, population, and lifehistory studies suggest that metapopulation theorymay be less relevant to some bull trout populations(Reiss 2003; Spruell et al. 2003; Whitesel et al. 2004)than originally proposed (Rieman and McIntyre1993). A number of metapopulation models existthat describe migration and gene flow patternsamong subpopulations (e.g., mainland-island,island, linear, stepping stone, or isolation by dis-tance models; Meffe and Carroll 1993; Hanski andGilpin 1997). Whitesel et al. (2004) suggested thatdifferent metapopulation models likely explainvarying bull trout population structures evidentamong Columbia River basin populations. Due toobserved variation in bull trout population struc-ture, no single specific metapopulation model like-ly describes all bull trout populations. Further-more, estimates from empirical genetic datasuggest very limited immigration occurs amongpopulations. Immigration substantial enough toprovide for gene flow among populations appearsto occur every 50–100 years, rather than the 5–10years typically observed in salmon populations(Bartley and Gall 1990; Reiss 2003; Spruell et al.2003). Therefore, loosely interpreted metapopula-tion structure may exist for some bull trout popu-lations, albeit with very low levels of gene flowamong subpopulations at a time interval that may

limit its short-term (10 years) relevance to improv-ing population viability.

Genetic studies have suggested that individualsin adjacent bull trout populations are more close-ly related to each other than to fish from moregeographically distant populations (Reiss 2003;Spruell et al. 2003). Costello et al. (2003) foundthat genotypes among surveyed populations had ahigh degree of differentiation, whereas genotypeswithin populations showed little variation. Thissuggests that surveyed bull trout populationsexhibited an isolation-by-distance model, onecharacterized by little gene flow across larger spa-tial scales (Costello et al. 2003). Alternatively, highgene flow across larger spatial scales would tend tohomogenize genetic signatures across geography.Whether bull trout historically exhibited lowgenotypic variation, or whether this condition isdue to a recently depressed population size, con-tinues to be debated. However, we believe that theIndian Creek and South Fork Tieton River bulltrout populations historically maintained relative-ly low genotypic variation due to their apparentdifferences in life history strategy and spatialreproductive isolation.

Within Rimrock Reservoir, available geneticdata suggest that gene flow among populationsoccurs but is very limited. Indian Creek and SouthFork Tieton River maintain two genetically distin-guishable populations with little interpopulationgene flow. Reiss (2003) hypothesized that prior todam construction, Indian Creek contained a resi-dent population and South Fork Tieton River con-tained a fluvial population, which implies geo-graphic or spatial reproductive isolation. This isbecause prior to 1925, when Tieton Dam was cre-ated, Indian Creek and South Fork Tieton Riverconverged into a river, not a lake, and Indian Creekis influenced by spring water making conditionsconducive for maintenance of a resident popula-tion. Only since 1925 has a lake been present thatprovides an environment conducive for sympatry(Reiss 2003). Curiously, the lake environment hasbeen available to more than 15 bull trout genera-tions, with little apparent gene flow between popu-lations. Reiss (2003) suggested that spawning sitefidelity was likely a key component to maintaininggenetic distinction between these populations.

202



James (2002) conducted 6 years of adult tag,release, and recapture studies in Indian Creek andSouth Fork Tieton River. During that time, onlyone tagged fish crossed over to the other stream,indicating minimal spawner exchange betweenpopulations. This observation is also supported bythe empirical genetic data.

For the Rimrock population, gene flow does notappear to be required in order to maintain popula-tion-level fitness or survival of bull trout. Reiss(2003) deduced from direct genetic evidence thatthese populations have been genetically isolated fornumerous generations prior to the creation of theTieton Dam. Hence, there is no evidence to indi-cate that opening fish passage at Tieton Damwould affect population genetics or productivityover a 100-year horizon.

Habitat Improvement

Available literature and our model suggests that thepopulation bottleneck occurs during ages 1–2.Juveniles compete for food and space in the con-fines of the South Fork Tieton River and IndianCreek. Once the fish migrate out of the stream andinto the lake at age 2, there appears to be sufficienthabitat to sustain high survival rates (adult survival65%). In the Pend Oreille basin, Downs andJakubowski (2003) determined that the adfluvialpopulation was limited by the age-0 to out-migrant ages (ages 1–3). They discovered that theegg to out-migrant survival in the Pend Oreillebasin ranges between 0.04% and 0.20%. Our sim-ulations suggest survival was 0.05% for the Rim-rock populations.

As a result, we explored the effect of increasinghabitat for the age-2 fish by incrementallyincreasing the stream carryingcapacity by 14%, 42%, and 100%.Each incremental increase was runthrough a 50-year model simula-tion. We discovered that increasingcarry capacity for age-2 fish provid-ed an equal corresponding effect onthe number of adults. Hence, a 14%increase in age-2 abundance causeda 14% increase in adult abundance.Habitat enhancement actions thatincreased age-2 carrying capacity

also appeared to increase population size and via-bility. Increased population size afforded greaterresilience to catastrophic effects because the pop-ulation gain of 14% juveniles, subadults, andadults allowed for greater mortality while sustain-ing sufficient egg production.

Harvest

Bull trout harvest is illegal in Rimrock Lake and itstributaries. Hence, when the model was set up toreflect the baseline simulation, harvest wasassumed to be at 2% for both subadults and adultsfrom incidental hooking mortality and limitedpoaching. The baseline simulation predicted that209 subadults and 34 adults were taken by harvestactivities. Prior to ESA listing, bull trout harvestwas allowed and may have had detrimental effectson the population. We tested the effect of harvestand observed surprising results. Sustained harvestappears to have a strong effect on population size.A readily sustainable adult exploitation rate of 20%reduced the population size by 39% and, at 50%exploitation rate, reduced the adult population sizeby 75% (Table 3).

Bull trout populations appear to be moreresilient to factors that limit juvenile life stages thanlimit adults. The bull trout resist catastrophic envi-ronmental events by geographically segregatingjuvenile and adult life stages, where adults live in thelake and juveniles live in the lake tributaries duringthe most dynamic period of the year (winter andearly spring) when catastrophic events are likely tooccur. Therefore, those catastrophic effects leadingto a greater probability of mortality were likelycaused by habitat variation in the tributaries, ratherthan in the lake environment. Hence, the bull trout

203


Table 3. Simulation outcomes in year 50 for Rimrock bull trout populations subjected toharvest rates of 2%, 20%, and 50% on adults, and 2% on subadults. All otherparameters set to baseline conditions.

Adult Subadult AdultHarvest assumption harvest harvest population size

2% adults, 2% subadults 34 209 1,677

20% adults, 2% subadults 257 202 1,027

50% adults, 2% subadults 409 179 408


population relies on a high adult survival rate dur-ing periods of low juvenile survival in order toreseed the populations. Further, since many spawn-er year-classes exist, the population quicklyrebounds with minimal long-term adult popula-tion size oscillations and inbreeding depression.Factors that lower adult survival appear to under-mine the species strategy for persistence and viabil-ity. Paul et al. (2000) discovered that bull trout werehighly susceptible to capture by anglers, and inmixed-stock fisheries, the combined high catchabil-ity and incidental mortality caused a significantreduction in the population. Limiting or abolishingdirected harvest and minimizing incidental mortal-ity will greatly improve population size and viabili-ty of bull trout. In our modeling exercise, discontin-uing directed harvest from a hypothesized historicrate of 20% caused adult population size to increaseby 39%. Harvest prohibition was instituted in theearly 1990s immediately following the listing of bulltrout as threatened under the ESA. Since the harvestprohibition, the bull trout population appears to berebuilding in Rimrock reservoir. Based on reddcount data collected in Indian Creek, the averagenumber of spawners, while bull trout harvest waslegal, was 44 during the period 1988–1990 (E.Anderson, Washington Department of Fish andWildlife, personal communication). After harvestprohibition was in place for 8 years, the averagenumber of redds was 182 during 1999–2001, sug-gesting that harvest may have played a central rolein limiting the population viability.

Conclusion

Model simulation runs suggest that Rimrock bulltrout are a viable, resilient population, able torebound quickly from intermittent catastrophicconditions as long as (1) conditions allow for theadult population to exceed an effective popula-tion size of 1,000 individuals to meet populationgenetic criteria, and (2) spawners are composedof many age-classes to assure that an adequatenumber of egg are deposited. The population via-bility was high because the number of eggsdeposited annually exceeds age-2 carrying capac-ity by more than 20 times, and multiple spawnerage-classes were present allowing the population

to quickly rebound from catastrophic events. Forexample, a catastrophic event as severe as killing50% of the subadult population every 2 yearsresulted in a reduced population size, but thepopulation remained stable during a 50-year peri-od. A summary of results from nine model runsunder varying environmental conditions comparethe minimum, maximum, and average adult pop-ulation size in the lake of a 50-year simulation(Table 4; Figure 9).

The model simulations suggest that the additionof a fish ladder to the Tieton Dam would not signif-icantly benefit the population. Factors such as thelimited number of entrained fish, the limited sur-vival of entrained fish, the large number of spawn-ing age-classes, and the overseeding of juvenilehabitat suggest that little benefit would be realizedfrom reconnecting the entrained fish with the lakepopulations. A reasonably achievable angler harvestrate of 50% adult had the greatest negative impacton the population, by reducing adult populationsize to 25% of its prior size. Increasing the carryingcapacity of age-2 fish had the greatest populationincrease potential in comparison to the baselinesimulation. When the age-2 carrying capacity wasdoubled, the adult population doubled.

The simulation where 50% of the subadults arecatastrophically entrained every 15 years resultedin an average adult population size of 1,524 indi-vidual. When a hypothetical fish ladder was addedto the simulation, the lake population increased byonly 2 adults to 1,526 individuals. Simulated habi-tat improvements that caused age-2 carryingcapacity to increase 14% (from 7,200 to 8,200individuals), increased the adult population size14% to an average of 1,738 individuals, whichexceeded the baseline estimate of 1,677 adults.Habitat improvements that target juvenile rearingcapacity have the greatest potential to increasepopulation size, thereby increasing both gene vari-ability and population viability. In the case ofRimrock, reconnecting entrained bull trout popu-lations to the lake population via a fish ladderdoes little to benefit the species and is not a viableoption for assuring bull trout rebuild to a pointwhere the ESA standards of recovery are met.Model simulations suggest focusing recoveryefforts on metapopulation connectivity will not

204



greatly benefit the Rimrock population; instead,improving juvenile habitat will increase popula-tion size and viability.

Acknowledgments

This paper would not have existed without thegenerosity of the Roza-Sunnyside Board of Joint

Control, who financially supported this work. Wethank Ray Beamesderfer and Paul Anders for theirinsights and Colin Chapman for draft reviews.

References

Bartley, D. M., and G. A. E. Gall. 1990. Genetic structure andgene flow in Chinook salmon populations of California.

Transactions of the American FisheriesSociety 119:55–71.

Beamesderfer, R. C. 1991. MOCPOP 2.0: a flex-ible system for simulation of age-structuredpopulations and stock-related functions.Oregon Department of Fish and Wildlife,Oregon Department of Fish and WildlifeInformation Report No. 91–4, Clackamas.

Beissinger, S. R., and D. R. McCullough, edi-tors. 2002. Population viability analysis.University of Chicago Press, Chicago.

Beverton, R. J. H., and S. J. Holt. 1957. Onthe dynamics of exploited fish popula-tions. United Kingdom Ministry of Agri-culture, Fisheries and Food FisheryInvestment Series 2, No. 19(2), London.

Bjornn, T. C. 1961. Harvest, age structure,and growth of game fish populationsfrom Priest to Upper Priest lakes. Trans-actions of American Fisheries Society90:27–31.

Brunson, R. B. 1952. Egg counts of Salveli-nus malma from the Clark’s Fork River,Montana. Copeia 3:196–197.

205


Table 4. Description of variables altered for the scenario results reported in Figure 9.

Scenario Catastrophic entrainment

Juvenile Upstream Adult Age-2 carryingDescription Frequency entrainment dam passage harvest capacity

A Base case N/A None None 2% 7,200

B Infrequent high entrainment rate 15 years 50% None 2% 7,200

C Frequent high entrainment rate 2 years 50% none 2% 7,200

D Infrequent high entrainment rate and upstream passage 15 years 50% 90% 2% 7,200

E Frequent high entrainment rate and upstream passage 2 years 50% 90% 2% 7,200

F High harvest rate None None None 50% 7,200

G Moderate harvest rate None None None 20% 7,200

H Low increase to juvenile carrying capacity None None None 2% 8,200

I Moderate increase to juvenile carrying capacity None None None 2% 10,200

Figure 9. Comparison of adult bull trout abundance after 50 years of simu-lation under various management scenarios. Line represents the minimumand maximum number of adults during the simulation.


Bustard, D. 1999. Kemess South Project Fish MonitoringStudies. Prepared for Kemess Mines Inc., Smithers,British Columbia.

Bustard, D. 2001. Kemess South Project Fish MonitoringStudies. Prepared for Kemess Mines Inc., Smithers,British Columbia.

Costello, A. B., T. E. Down, S. M. Pollard, C. J. Pacas, and E.B. Taylor. 2003. The influence of history and contempo-rary stream hydrology on the evolution of genetic diver-sity within species: an examination of microsatelliteDNA variation in bull trout, Salvelinus confluentus(Pisces: Salmonidae). Evolution 57:328–344.

Downs, C. C., and R. Jakubowski. 2003. Lake Pend Oreille /Clark Fork River fishery research and monitoring, 2002progress report. Submitted by Idaho Department ofFish and Game to Avista Corporation, Spokane, Wash-ington.

Fraley, J., D. Read, and P. Graham. 1981. Flathead River fish-ery study: April 1981. Montana Department of Fish,Wildlife and Parks, Kalispell.

Fraley, J. J., and B. B. Shepard. 1989. Life history, ecology andpopulation status of migratory bull trout (Salvelinusconfluentus) in the Flathead Lake river system, Montana.Northwest Science 63:133–143.

Hanski, I. A., and M. E. Gilpin, editors. 1997. Metapopula-tion biology. Academic Science Press, New York.

Hiebert, S., E. Best, J. Sechrist, R. Swartz, D. Moore, J. Wil-son, and S. Kennedy. 2003. Fish entrainment form Rim-rock Reservoir Tieton River, Washington 2002. U.S.Bureau of Reclamation, Denver.

Hunt, C. W., R. Hawryluk, and D. Hildebrandt. 1997. Bulltrout studies in Alberta’s fish management area four.Pages 171–186 in M. C. Mackay, M. K. Brewin, and M.Monita, editors. Friends of the Bull Trout ConferenceProceedings. Trout Unlimited Canada, Calgary, Alberta.

James P. W. 2002. Population status and life history charac-teristics of bull trout in the Yakima River basin: FinalReport. Submitted by Central Washington University tothe U.S. Bureau of Reclamation, Yakima, Washington.

James, P. W. 2000. Status and monitoring of bull trout pop-ulations in the Yakima River basin, 1999 annual report.Submitted by Central Washington University to the U.S.Bureau of Reclamation, Yakima, Washington.

Kalin, A., S. P. Cramer, and C. G. Ackerman. 2002. Biologicalassessment – potential impacts from the Tieton Hydroelec-tric Project on ESA listed bull trout and steelhead in theYakima basin. SP Cramer and Associates, Gresham, Oregon.

Leathe, S. A., and P. J. Graham. 1982. Flathead Lake fish foodhabits study, final report. Montana Department of Fish,Wildlife and Parks, Kalispell.

Lestelle, L. C., G. R. Blair, L. E. Mobrand, and W. E.McConnaha. 2004. Species-habitat rules, bull trout inecosystem diagnosis and treatment. Mobrand Biomet-rics, Inc. Vashon, Washington.

May, B., J. Huston, and S. McMullin. 1979. Lake Koocanusapost-impoundment fisheries study, completion report.Montana Department of Fish and Game, Kalispell.

MBI (Mobrand Biometric Incorporated). 2004. Ecosystemdiagnosis and treatment habitat database for the Yakima

River. Available: http://www.mobrand.com/edt/ (Sep-tember 2005).

McPhail, J. D., and C. Murray. 1979. The early life historyand ecology of Dolly Varden (Salvelinus confluentus) inthe upper Arrow lakes. Report to the British ColumbiaHydro and Power Authority and Kootenay Region Fishand Wildlife, Nelson, BC.

McPhail, J. D., and J. S. Baxter. 1996. A review of bull trout(Salvelinus confluentus) life history and habitat use inrelation to compensation and improvement opportuni-ties. University of British Columbia, Fisheries Manage-ment Report No. 104, Vancouver.

Meffe, G. K., and C. R. Carroll. 1993. Principles in conserva-tion biology. Sinauer Press, Sunderland, Massachusetts.

Morris, W. F., and D. F. Doak. 2002 Quantitative conserva-tion biology, theory and practice of population viabilityanalysis. Sinauer Associates, Inc. Sunderland, Massa-chusetts.

Paul, A. J. 2000. Recruitment dynamics in bull trout (Salveli-nus confluentus): linking theory and data to speciesmanagement. Doctoral dissertation. University of Cal-gary, Calgary, Alberta.

Paul, A. J., J. R. Post, G. L. Sterling, and C. Hunt. 2000. Den-sity-dependent inter-cohort interactions and recruit-ment dynamics: models and a bull trout time series.Canadian Journal of Fish and Aquatic Sciences57:1220–1231.

Post, J. R., E. A. Parkinson, and N. T. Johnston. 1999. Densi-ty-dependent processes in structured fish populations:interaction strengths in whole lake experiments. Ecolog-ical Monographs 69:155–175.

Pratt, K. L. 1991. Bull trout scale analysis - Metolius Riverbasin, final report. U.S. Forest Service, DeschutesNational Forest, Bend, Oregon.

Reiss, K. Y. 2003. Genetic variability within bull trout(Salvelinus confluentus) populations in the Yakima Riverbasin. Master’s thesis. Central Washington University,Ellensburg.

Rieman, B. E., and J. B. Dunham. 2000. Metapopulation andsalmonids: a synthesis of like history patterns andempirical observations. Ecology of Freshwater Fish9:51–64.

Rieman, B. E., and J. D. McIntyre. 1993. Demographic andhabitat requirements for conservation of bull trout. U.S.Forest Service, Intermountain Research Station, Gener-al Technical Report INT-302, Boise, Idaho.

Spruell, P., A. R. Hemmingsen, P. J. Howell, and F. W. Allendorf.2003. Conservation genetics of bill trout: geographic distri-bution of variation at microsatellite loci. ConservationGenetics 4:17–29.

Taylor, M. W. 1981. A generalized inland fishery simulator formanagement biologists. North American Journal of Fish-eries Management 1:60–72.

Underwood, K. D., and S. P. Cramer. 2004. Mill Creek bulltrout life cycle model description. Prepared by SP Cramerand Associates for the Walla Walla Bi-State Habitat Con-servation Plan Coordination Committee, Walla Walla,Washington.

USBR (U.S. Bureau of Reclamation) 2004. Yakima Hydromet

206



ARCHIVE data access. Available: www.usbr.gov/pn/hydromet/yakima/yakwebarcread.html (December 2006).

USFWS (U.S. Fish and Wildlife Service) 2002. Bull trout(Salvelinus confluentus) draft recovery plan. Available:http://ecos.fws.gov/docs/recovery_plans/2002/021129.pdf(September 2005).

Wallis, O. L. 1948. Trout studies and a stream survey of CraterLake National Park, Oregon. Master’s thesis. Oregon StateUniversity, Corvallis.

Walters, C. J., and J. Post. 1993. Density-dependent growth andcompetitive asymmetries in size-structured fish popula-

tions. Transactions of the American Fisheries Society122:34–45.

Walters, C. J. 1969. A generalized computer simulation modelfor fish population studies. Transactions of the AmericanFisheries Society 98:505–512.

Whitesel, T. A., J. Brostrom, T. Cummings, J. Delavergne, W.Fredenberg, H. Schaller, P. Wilson, and G. Zydlewski.2004. Bull trout recovery planning: a review of the sci-ence associated with population structure and size. U.S.Fish and Wildlife Service, Regional Office, Science TeamReport No. 2003–01, Portland, Oregon.

207



american fisheries society · status, distribution, and conservation of native freshwater fishes of...

Documents