using habitat suitability models to identify essential...

142
USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT FOR THE WINTER FLOUNDER, PSEUDOPLEURONECTES AMERICANUS, IN GREAT BAY ESTUARY, N.H. BY JENNIFER M. WANAT BS in Marine & Freshwater Biology, University of New Hampshire, 1999 THESIS Submitted to the University of New Hampshire In Partial Fulfillment of the Requirements For the Degree of Master of Science In Zoology December, 2002

Upload: others

Post on 31-May-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISHHABITAT FOR THE WINTER FLOUNDER, PSEUDOPLEURONECTES

AMERICANUS, IN GREAT BAY ESTUARY, N.H.

BY

JENNIFER M. WANATBS in Marine & Freshwater Biology, University of New Hampshire, 1999

THESIS

Submitted to the University of New HampshireIn Partial Fulfillment of the Requirements

For the Degree of

Master of ScienceIn

Zoology

December, 2002

Page 2: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

DEDICATION

This thesis is dedicated to my parents, whose sacrifice, patience, and assistance

have made this endeavor possible. Thank you for everything and most importantly, thank

you for sharing your love of the sea and of learning.

Page 3: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

ACKNOWLEDGEMENTS

First, let me acknowledge the tremendous amount of energy that my thesis

advisor, Hunt Howell, has contributed to this project. Thank you for your patience,

advisement, and assistance with this work and with me. It means the world to me.

Second, I’d like to thank my other committee members, Ray Grizzle and David

Berlinsky, for their input. Your contributions to this project have also been extremely

valuable.

Thank you to my labmates who have assisted me along the way with sampling,

lab work, and moral support. You guys are incredible and have made this experience very

special.

Lastly, let me thank my friends and roommates who have made this a fun time

and one that I’ll never forget.

Page 4: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

TABLE OF CONTENTS

DEDICATION……………………………………………………………………………

ACKNOWLEDGEMENTS……………………………………………………………….

LIST OF TABLES………………………………………………………………………..

LIST OF FIGURES………………………………………………………………………

ABSTRACT………………………………………………………………………………

CHAPTER PAGE

INTRODUCTION……………………………………………………………………….

II. TESTING THE USE OF HABITAT SUITABILITY MODELS ASPREDICTORS OF ESSENTIAL FISH HABITAT FOR THE WINTERFLOUNDER IN GREAT BAY ESTUARY, N.H.………………………………………………………………………..

Introduction………………………………………………………………………….

Materials and Methods…………………………………………………………………..

Results…………………………………………………………………………………….

Discussion…………………………………………………………………………………..

II. WINTER FLOUNDER STOMACH CONTENT ANALYSIS AND MACRO-FAUNAL BENTHIC COMMUNITY ANALYSIS……………………………………

Introduction………………………………………………………………………………

Materials and Methods…………………………………………………………………….

Results………………………………………………………………………………………

Discussion…………………………………………………………………………………..

SYNOPSIS………………………………………………………………………………

LIST OF REFERENCES……………………………………………………………….APPENDIX A………………………………………………………………………….

Page 5: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

LIST OF TABLES

Table 1.1 Species Caught ……………………………………………………………….

Table 1.2 Habitat Suitability Index Values by Site and Month…………………………

Table 2.1 Index of Relative Importance: Stomach Contents…………………………..

Table 2.2 Taxa Identified from Benthic Core Samples………………………………….

Page 6: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

LIST OF FIGURES

Figure 1.1 Great Bay Map Estuary…..…………………………………………………….

Figure 1.2 Winter Flounder CPUE (Banner Suitability Index Values)…………………….

Figure 1.3 Winter Flounder CPUE (Brown Suitability Index Values)……………………..

Figure 1.4 Winter Flounder CPUE by Site………………………………………………...

Figure 1.5 Winter Flounder CPUE by Month………………………………………………

Figure 1.6 Winter Flounder CPUE by Temperature………………………………………..

Figure 1.7 Winter Flounder CPUE by Salinity……………………………………………..

Figure 1.8 Winter Flounder CPUE by Depth……………………………………………….

Figure 1.9 Winter Flounder CPUE by Sand Percentage……………………………………

Figure 1.10 Winter Flounder CPUE by Sediment Organic Content……………………….

Figure 1.11 Winter Flounder CPUE by Mysid Shrimp CPUE.…………………………….

Figure 1.12 Winter Flounder CPUE by Crangon Shrimp CPUE…………………………..

Figure 1.13 Winter Flounder CPUE by Green Crab CPUE………………………………..

Figure 1.14 Winter Flounder CPUE by Smooth Flounder CPUE…………………………

Figure 2.1 Great Bay Estuary Map…………………………………………………………

Figure 2.2 Stomach Contents by Different Size Fish Categories…………………………..

Figure 2.3 Stomach Contents by Each Season Sampled……………………………………

Figure 2.4 Stomach Contents by Site……………………………………………………….

Figure 2.5 Abundance of Benthic Community Constituents at Each Site………………….

Figure 2.6 Abundance of Benthic Constituents at Each Site over the Sampling Period…..

Figure 2.7 Abundance of Benthic Constituents over the Sampling Period………………...

Figure 2.8 Community Structure between Sites (October/November 2000)………………

Page 7: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.9 Community Structure between Sites (April/May 2001)………………………

Figure 2.10 Community Structure between Sites (June/July 2001)………………………

Figure 2.11 Community Structure between Sites (August/September 2001)…………….

Figure 2.12 Community Structure between Sites (October/November 2001)……………

Figure 2.13 Community Structure of Site 19 over the Sampling Period………………….

Figure 2.14 Community Structure of Site 23 over the Sampling Period………………….

Figure 2.15 Community Structure of Site 25 over the Sampling Period………………….

Figure 2.16 Community Structure of Site 29 over the Sampling Period………………….

Figure 2.17 Community Structure of Site 35 over the Sampling Period…………………..

Figure 2.18 Community Structure of Site 51 over the Sampling Period…………………...

Figure 2.19 Community Structure of Site 67 over the Sampling Period…………………...

Figure 2.20 Community Structure of Site 73 over the Sampling Period…………………...

INTRODUCTION

Page 8: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

In 1976, the Magnuson-Stevens Fisheries Conservation and Management Act

(MSFCMA) laid the groundwork for changes in fisheries conservation and management

policies nationwide. It created regional fisheries management councils and gave them the

power to set standards, which if followed, would allow stocks to reach levels that would

sustain optimum yield. Optimum yield is defined as the maximum sustainable yield that

would provide the nation with sufficient commercial and recreational benefits (MSFCMA

1976). Maximum sustainable yield could be modified by “any relevant economic, social,

or ecological factor” (MSFCMA 1976). Over the next few years as changes in the

definition of maximum sustainable yield occurred and liberties with ‘economic, social,

and ecological factors’ were taken, it was apparent that an amendment to the original

document was needed.

During the 1980’s, changes were made in the policies to ensure that management

decisions would not lead to over-fishing. For each stock, over-fishing needed to be

quantified, and management plans made, to evaluate the condition of the stock (Fluharty

2000). If a population was over-fished, the council was required to determine the source

of the problem and create a plan to minimize the effects of it, whether the source was

direct or indirect.

In 1986, the National Habitat Conservation Policy was used as a framework to

make changes to the MSFCMA. Councils were required to provide information regarding

the habitat requirements of all regulated fish stocks, and had the ability to recommend

management plans for habitats within their jurisdiction. With greater emphasis placed

upon marine biodiversity during the early 1990’s, fishing was no longer considered a

threat to specific fish species alone, but to other marine species and habitats as well. The

Page 9: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

impacts of fishing gear, bycatch, and discards upon the ecosystems where fish were

caught became a central issue. In addition, there was degradation in coastal habitat

quality due to development, agricultural run-off, and poor waste treatment. It was

determined that these losses in habitat area and quality were having a direct effect on

recruitment to, and growth of, fish stocks.

New legislation, entitled the Sustainable Fisheries Act (SFA), addressed the

concerns of scientists and lawmakers about the importance of sustaining adequate

habitats for fish species. The SFA enabled the management plans of the fisheries councils

to be reviewed and amended to take habitat modification into account. The phrase

“essential fish habitat (EFH)” was born of this document and was defined as “those

waters and substrate necessary to fish for spawning, breeding, feeding, or growth to

maturity (SFA 1996).” Managers were handed the immense job of identifying EFH;

identifying possible actions detrimental to that habitat, possible ways to correct those

actions, and ways to conserve that habitat to sustain fish stocks.

Habitat is essential only if it is limiting to, or necessary for, the activities outlined

by the definition of essential fish habitat. To identify essential habitat, one must first

understand the relationship between a fish and its environment. The concept of

ecophysiology, introduced by F.E.J. Fry in the 1940’s, describes the synergism between

the ecology and physiology of organisms. It explains the process by which a fish changes

its behavior or physiological state in concert with changes in its environment (Rankin and

Jensen 1993 in Yamashita 2001). Since the ocean is a dynamic environment, fish are

exposed to variable conditions. Optimal conditions are those associated with high growth

rates, high survival, and high abundance. Less favorable conditions are those associated

Page 10: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

with the cessation of growth (even weight loss), high mortality, and low densities.

Variation, or degradation, in the condition of an environment will ultimately have an

effect on individual growth, survival, and recruitment to stocks.

Components of the environment that are most important to growth and survival

need to be identified. Fry suggested five categories of ecological factors, or conditions, to

be considered in the context of metabolism: limiting, controlling, masking, directive, and

lethal (Yamashita 2001). Controlling factors are those that dictate the speed of maximum

and maintenance metabolic rate. Two controlling factors are temperature and body size.

Limiting factors are the resources necessary to drive metabolism or limit maximum

metabolic rate. These primarily include food availability and dissolved oxygen. Masking

factors increase baseline metabolism, while directive factors decrease baseline

metabolism. Masking factors include stresses to the fish such as changing salinity.

Directive factors include any environmental factor that elicits a behavioral response that

drives the organism to occupy more favorable conditions (Yamashita 2001, Neill 1994).

Lethal factors include those conditions outside the range that is tolerable to the organism

such as high temperatures.

Neill (1994) suggests that there “must be a continued succession of the proper

time-space series of suitable abiotic environment, sufficient quantity and quality of food,

and tolerable levels of predation” for there to be successful recruitment. Understanding

these ecological factors and selecting the ones that are directing the growth and survival

of fish can quantify the quality of the environment of the fish. Hypotheses can be made

about what areas provide the best conditions and are essential to fish.

Page 11: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Habitat Suitability Models

Habitat Evaluation Procedures (HEP) are used by the US Fish and Wildlife

Service to identify important habitats. HEP are based upon “the fundamental assumption

that habitat quality and quantity can be numerically described (HEP 1980).” They have

been used for years on several endangered and near-endangered species both on land and

in freshwater systems. Recently, with the increased interest in marine habitat assessment,

these evaluation procedures have been considered for use in ocean and estuarine systems.

Data produced can be utilized to compare the quality of habitats or to assess changes in

that habitat over time.

Components of the environment are quantified by a Habitat Suitability Index

(HSI). The values within the index are determined by “the ability of key habitat

components to supply the life requisites of selected species of fish and wildlife.

Evaluation involves using the same key habitat components to compare existing habitat

conditions for the species of interest. Optimum conditions are those associated with the

highest potential densities of the species within a defined area (HEP 1980).”

For each environmental parameter, there is a range of values that is habitable by

the target species. On a scale of one to ten, the most optimal value of the parameter for

that species would be given a value of ten. This would indicate that these values are the

most suitable for growth and survival of that species.

Through experimentation and observation, the optimal range for each

environmental parameter can be determined for the target species. This is compared with

the range of values compiled from field studies of the area under scrutiny. Several

parameters are considered for each habitat, and values are assigned to them depending on

Page 12: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

their relative suitability. By calculating the geometric mean of these values, the overall

value or suitability for the target species can be determined. This data could provide

useful information to fisheries councils about what areas might be essential to stock

maintenance.

Habitat Suitability Indices have been utilized as a means of characterizing and

evaluating the habitats of freshwater and saltwater species of fish. In Florida, a database

of information on the life histories and ecology of fishes and invertebrates is being

constructed for reference in HSI studies (Rubec 1998). Water column and benthic

samples have been collected from estuaries in Florida. These data can be applied to a

gridded map with suitability values given to each block for the target species.

Independent sampling of the specific species can determine the relative accuracy and

predictive value of the model.

Determining the habitat requirements of the target species appears to be key to the

success of the HSI model. A wide spectrum of environmental variables has been utilized

for these studies. In a study of brook trout in the Blue Ridge Province in Missouri, pH,

elevation, presence and absence of competitions (rainbow trout) and predators, and prey

species were used as variables essential to fish productivity (Schmitt 1993). In actuality,

the distribution of brook trout correlated well with some variables but not with others. In

another study, with American shad, temperature and water velocity were determined to

be the key environmental requisites of concern (Ross 1993). Depth, substrate, dissolved

oxygen, salinity, turbidity, submerged aquatic vegetation, and predator composition and

abundance are other environmental variables that have been considered or utilized for

HSI studies.

Page 13: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Those ecological variables that have the greatest effect on the life history and

distribution of the target species should be the variables upon which the most weight is

placed. Data should be available on the ranges of these variables that are most suitable for

the growth and survival of the target species. Sampling of the study area should quantify

these variables to determine means, ranges, and variation over time. Simultaneous

sampling of the target species should be made to verify their presence and provide data

that at the conclusion of the study will enable the accuracy of the model to be determined.

Temperature, salinity, substrate, and depth have been utilized as the backbone of

most HSI studies. The effects of these physical factors on various fish species have been

studied extensively both in the lab and in the field. For HSI studies, these variables are

easily quantified by regimented sampling.

Winter Flounder

The winter flounder, Pseudopleuronectes americanus, is an important

commercially- and recreationally-fished species in the northwest Atlantic. The right-eyed

flounder is one of the most desirable of the flatfishes due to its thick fillet. The Gulf of

Maine stock size reached 26 million in the early eighties, and then proceeded to decline

to approximately 9 million in the early nineties. The stock rebounded to about 13 million

in the mid-nineties, but then declined to 6 million by 1998 (Nitschke 2001). Similarly,

commercial catches have followed this pattern and now remain at less than a 1000 metric

tons a year (data available only to 1998; Nitschke 2001). The range of the winter flounder

extends between Newfoundland and Georgia (Buckley 1989) and is managed as three

stocks: Gulf of Maine, Georges Bank, and Southern New England – Middle Atlantic.

Page 14: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Flounder are typically found in coastal and estuarine waters, although they have been

fished on offshore shoal areas such as Georges Bank. Sexually mature flounder move into

estuaries during the fall, overwinter there, and spawn in the middle to late spring. The

demersal eggs hatch, fish metamorphosis, settle, and remain in the estuary for the first 1-

2 years, after which they move offshore (Klein-MacPhee 1978). After settlement,

flounder are for the most part demersal, spending their time in close association with the

substrate.

The winter flounder has been studied extensively within the field and laboratory.

Data on the effects of major environmental variables suggest optimum conditions under

which winter flounder are the most successful. For these reasons and those listed above,

the winter flounder is a good candidate for a habitat suitability model.

Great Bay Estuary, New Hampshire

Great Bay Estuary in New Hampshire contributes to the natal grounds of winter

flounder from the Gulf of Maine stock. The estuary covers approximately 23 square

kilometers and is fed by seven rivers, draining over 2,330 square kilometers of watershed

(Short et al 1992). The estuary, primarily Portsmouth Harbor, is used for shipping and

commercial fishing boat traffic. Recreational boaters venture further up the Piscataqua

River and into Great Bay. Treated sewage effluent is discharged into the estuary from all

towns surrounding the estuary as well as some industrial pollutants.

The Estuary is very well studied, and data are available for several habitat

characteristics. Because the Bay is important habitat for flounder and it is susceptible to

anthropogenic impacts, it is an appropriate area for study. Great Bay also offers a variety

Page 15: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

of habitat types. There is a large salinity gradient from the mouth of the harbor to the

rivers that supply waters: 5 to 35ppt. There is also a gradient of temperature, which

throughout the year ranges from freezing to 27° Celsius. Depths vary from mudflats

emergent at low tide to parts of the shipping channel that are over 80 feet deep.

Substrates vary from silty-clay and sand to gravel and boulder. Because of this variation

in habitat types, it would seem reasonable that certain areas of the Bay provide better

habitat than others.

A preliminary HSI model was constructed for winter flounder in Great Bay, N.H.,

utilizing temperature, salinity, substrate, and depth as the significant environmental

variables (Banner 1996). However, the predictability of the model was not tested. No fish

sampling was performed in conjunction with the environmental sampling, so it is difficult

to make conclusions about the accuracy, and therefore the usefulness of the model.

In the first chapter the Habitat Suitability Model was tested with

contemporaneous measurement of abiotic conditions and sampling of fish. Each variable

within the HSI was examined for its relationship to fish abundance. In the second chapter

an important biotic variable, prey abundance, was measured and compared to fish

abundance to see if it may be a limiting factor. Predator abundance and competition will

also be discussed as potential directive factors.

Page 16: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

LITERATURE CITED

Banner, A. and G. Hayes. 1996. Mapping Important Habitat of Coastal New Hampshire.Chapter 6: Winter flounder. http://gulfofmaine.org/library/gbay/wfl.htm

Bigelow &Schroeder. 1953. Fishes of the Gulf of Maine. U.S. Fish Wildl.Serv., FishBulletin. 53: 577 pp.

Buckley, J. 1989. Species Profiles: Life histories and environmental requirements ofcoastal fishes and invertebrates (North Atlantic) - winter flounder. U.S. Fish andWildlife Service Biological Report 82(11.87).

Fluharty, D. 2000. Habitat Protection, Ecological Issues, and Implementation of theSustainable Fisheries Act. Ecological Applications 10(2): 325-337.

Habitat Evaluation Procedures: Standards for Development of HSI Models. 1980.http://www.fws.gov/directives/library/hbindex.html#HEP.

Klein-MacPhee, G. 1978. Synopsis of Biological Data for the Winter Flounder,Pseudopleuronectes americanus (Walbaum). NOAA Technical Report NMFSCircular 414: 1-43.

Magnuson Fisheries Management and Conservation Act. 1976.

Neill, W. H., Miller, J.M., van der Veer, H.W., and K. O. Winemiller 1994.Ecophysiology of Marine Fish Recruitment: A Conceptual Framework forUnderstanding Interannual Variability. Netherlands Journal of Sea Research32(2): 135-152.

Nitschke, P., Brown, R., and L. Hendrickson. 2001. Status of Fisheries Resources offNortheastern United States - Winter Flounder.http://www.whoi.edu/sos/spsyn/fldrs/winter/index.html.

Rankin, J.C. and F.B. Jensen. 1993. Fish Ecophysiology: The comparative physiologist’sviewpoint. In: Rankin, J.C., Jensen, F.B. (Eds.) Fish Ecophysiology. Chapman &Hall. London.pp. xvi-xix. In Yamashita, Y., Tanaka, M., and J.M. Miller (2001)Ecophysiology of juvenile flatfish in nursery grounds. J. Sea Res. 45: 205-218.

Ross, R. M. 1993. Habitat use by spawning adult, egg, and American shad in theDelaware River. Rivers 4(3): 227-238.

Rubec et al. 1999. Suitability Modeling to Delineate Habitat Essential to SustainableFisheries. Amer. Fish. Soc. Symposium 22, 108-133.

Rubec, P. J. et al. 1998. Spatial Methods Being Developed in Florida to DetermineEssential Fish Habitat. Fisheries 23(7) 21-25.

Page 17: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Schmitt, C. J. 1993. Habitat Suitability Index Model for brook trout in streams of theSouthern Blue Ridge Province: surrogate variables, model evaluation andsuggested improvements. Biol. Rep. U.S. Fish and Wildlife Service 18.

Short, F.T. (ed.). 1992. The Ecology of Great Bay Estuary, New Hampshire and Maine:an Estuarine Profile and Bibliography. 221pp.

Sustainable Fisheries Act. U.S. Senate 23 May 1996. Report of the Committee onCommerce, Science, and Transportation on S.39: Sustainable Fisheries Act.Report 104-276, 104th Congress, Second Session. US Government Printing,Washington, D.C. U.S.A.

Yamashita, Y., Tanaka, M., and J.M. Miller. 2001 Ecophysiology of juvenile flatfish innursery grounds. J. Sea Res. 45: 205-218.

Page 18: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

CHAPTER I

TESTING THE USE OF HABITAT SUITABILITY MODELS AS PREDICTORS OFESSENTIAL FISH HABITAT FOR THE WINTER FLOUNDER IN GREAT BAY

ESTUARY, N.H.

Introduction

In the introduction, the concept of ecological factors directing metabolic rate, and

therefore growth, was discussed. In this chapter, specific factors were chosen as

components of a habitat suitability model. These factors were chosen by their availability

as archived data, but also for their importance as controlling factors (temperature),

limiting factors (salinity), and driving factors (depth and substrate) of metabolism.

Habitat Suitability Models have value in assessing the quality of a habitat for a

given species and quantifying changes in that habitat over time. Evaluating areas for

‘Essential Fish Habitat’ designation is just one use of habitat models. Models can also be

used to predict the variations in the quality of an area due to environmental changes

(Brown 2000). Maps can be created to evaluate habitats or areas of interest beyond the

scope of sampling efforts (Rubec 1998). Perhaps the most powerful aspect of the habitat

suitability model is the relative ease in creating the model.

Two habitat suitability models have been created for winter flounder in the Gulf

of Maine. Brown et al. (2000) mapped essential fish habitat for the entirety of Casco Bay,

Maine. Banner et al. (1996) concentrated on Great Bay Estuary specifically as part of a

larger project to designate essential habitats for coastal Maine. Both models used

temperature, salinity, depth, and substrate as model components.

Page 19: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

While catch data was used to validate these models, fishing effort was not

contemporaneous with measurement of abiotic variables. Areas of sampling effort

changed over time, so it was difficult to characterize specific sites temporally. In general,

catch data overemphasizes areas of high abundances, because that is where fishing effort

is greatest (Gibson 1994). Often there is little catch data available for areas where it is

thought that there are few fish. Likewise, sampling may not cover areas that represent the

entire range of each component in the model. Variability within the model between areas

should be understood as well as changes in the suitability of that area over time.

In this chapter both suitability models were tested with concurrent sampling of

fish. Sites were chosen to represent a variety of different habitats without prior

knowledge of fish densities at these sites. Each variable was tested against catch to see if

there was a significant relationship. Other potential ecological factors such as the

occurrence of predators and competitive species were also tested.

Materials and Methods

Site Selection

Eight sites were chosen within the Great Bay Estuary System of New Hampshire

(Figure 1.1). Sites were chosen to represent a gradient of four environmental parameters:

temperature, salinity, depth, and substrate. Sites were sampled from September 2000 to

November 2000 and from April 2001 to November 2001.

Fish Collections

Page 20: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Winter flounder were collected monthly from each site. Two ten-minute tows

were conducted with a 1-meter beam trawl (6 mm body and 3 mm liner) and 4.8-meter

otter trawl (2.5cm body and 6 mm codend). The otter trawl was towed at a speed of

approximately 1 knot and the beam trawl at a speed of 2 knots. Trawl catch per unit effort

was standardized to catch per 100 square meters. The length and weight of all winter

flounder were measured. Additional data collected from the trawls included the

enumeration and identification of fish species other than winter flounder and

invertebrates.

Abiotic Variables

Using an Onset HoBo datalogger (part # H08-001-02), temperature was recorded

every hour. The datalogger was anchored above the substrate at the central point of each

site. Salinity was measured at the time of sampling. Sediment particle size and organic

content was determined by the EPA’s Coastal 2000 project (S. Jones pers. comm.) and

archived data of Great Bay (L. Ward pers. comm.). Depth was recorded during each

trawl, and an average for each site and month was calculated.

Habitat Suitability Analysis

Suitability Index values were taken from the models constructed by Banner

(1996) and Brown (2000) [Appendix A]. Suitability values were applied to the

parameters measured during field sampling. The overall habitat suitability was calculated

for each site during each month sampled using both models. Both models were compared

to log-transformed winter flounder catch per unit effort (fish/100m2) using Analysis of

Page 21: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Variance. Catch per unit effort was calculated as the average of both otter trawls and

expressed as fish per 100 m2. Crangon and Mysid shrimp were collected with the beam

trawl, and also standardized to number per 100 m2. Green crabs and smooth flounders

were collected by otter trawl.

Catch per unit effort was compared to each parameter independently to determine

which parameter(s) might be the most important using Analysis of Variance, followed by

a Tukey’s Test when relationships were found to be significant.

Results

The species that were collected in both the beam and otter trawls are summarized

in Table 1.1. Suitability indices (SI) were calculated for each month at each site (Table

1.2). Indices were calculated with both the Brown Habitat Suitability Model and the

Banner Model and are based on a scale of 0 to 10, with a value of 10 indicating optimal

habitat. Values produced with the Brown HSI ranged from 2.24 to 8.41. SI values

produced by the Banner HSI ranged from 3.76 to 10. Neither HSI deemed any site

unsuitable (value = 0) at any time during the year.

Suitability values produced by the Brown suitability index were found to vary

significantly between sites (all months combined, p value = 0.000), but not to vary

significantly by month (all sites combined, p = 0.090). Sites 23, 29, 35, 51, and 73

averaged “low” SIs, ranging from 3 to 4.3, while sites 19, 25, and 67 produced “high” SIs

ranging from 5.8 to 6.2. Although not significant, suitability values were predicted to be

low in July, August, September, and November of 2001, while other months were

relatively higher.

Page 22: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

The Banner Habitat Suitability Model predicted SI values that were significantly

different between sites (p = 0.000) and between months (p = 0.021). Sites 19, 25, 29, and

67 were predicted to have high suitability and Sites 23, 35, 51, and 73 were predicted to

have low suitability. September of 2000 was predicted to have a high suitability as well

as May, September, and October of 2001.

Winter flounder catch per unit effort (CPUE), measured in fish per 100 square

meters, did not significantly increase with an increase in predicted suitability for either

model (Brown: p = 0.542 and Banner: p = 0.096). The Banner Habitat Suitability Model

actually predicted a negative relationship between suitability and catch per unit effort

(Figure 1.2). The Brown Model predicted no significant relationship (Figure 1.3).

Winter Flounder catch per unit effort did not vary significantly between sites

sampled in the survey (Figure 1.4, p = 0.740). However catch per unit effort did vary

significantly between months (Figure 1.5, p = 0.000). November 2000 CPUE was

significantly higher than all other months except October 2000 (p = 0.101). October 2000

CPUE was greater than all months except for November 2000, but only significantly

higher than August 2001 and September 2001. Otherwise there was no significant

difference between months.

Analysis of Variance was performed on components within the habitat suitability

models as well as some other possible ecological factors. These included temperature

(Figure 1.6), salinity (Figure 1.7), depth (Figure 1.8), substrate (Figure 1.9), and loss on

ignition (organic content) (Figure 1.10). Linear regression was used to compare winter

flounder catch per unit effort to Mysid shrimp (prey item) (Figure 1.11), crangon shrimp

Page 23: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

(predator) (Figure 1.12), green crab (predator) (Figure 1.13), and smooth flounder

(competitive species) (Figure 1.14).

Salinity was the only abiotic variable found to have a significant effect on winter

flounder CPUE (Figure 1.7). The category 16–18 ppt was significantly different from all

other categories except the 14-16 and 18-20 categories. All other categories were not

significantly different from each other. There was no significant difference between the

four depth categories (Figure 1.8). There was no significant difference in temperature

categories although there is a general trend of decreasing catch with increasing

temperature (Figure 1.6). There was no significant difference between catches at different

sand percentages (Figure 1.7) or at different organic percentages (Figure 1.10).

Catch per unit effort was positively correlated with green crab abundance

(p=0.000), but no other significant relationship was found with the other biotic variables.

Discussion

Brown (2000) summarizes well the four assumptions made when constructing a

habitat model. The first assumption is that all variables within the model are equally

important to the growth and survival of the organism. As demonstrated by this study, it is

likely that variables within the model have different levels of effect on the winter

flounder. Salinity was found to be the only variable to have a significant effect on winter

flounder catch per unit effort. It appears, although not significant, that other variables

such as temperature might also have an effect on distribution. A supposition is that the

fish will preferentially choose habitats that benefit their overall fitness and catch per unit

effort will reflect this preference.

Page 24: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Brown (2000) suggests two solutions for resolving this issue. First, weighting

certain variables over others may allow the most important variables to have the greatest

bearing on the overall habitat suitability. Likewise, reducing the importance of certain

variables would lessen their effect on the overall habitat suitability. Second, each variable

could be limited in its range of values to reflect its overall importance to that organism.

Variables that have little effect on growth or survival would be limited to a range within

the middle of the scale of values. A third suggestion might be to remove that variable

from the model. A principle components analysis would show how much each

component contributes to the overall variability within the data set. A confidence level

could be determined and variables could be deleted from the model until the confidence

level was reached.

The second assumption is that all variables within the model are independent of

each other. In this study, it was intended for components to represent a gradient of all

possible variable categories. It is impossible to ignore the obvious relationship between

certain variables. For example, during the summer months, temperature tends to increase

as one travels up into the estuary. Water is shallower and there is a smaller body of water

to heat up, while those waters at the coast are constantly fed with cold, ocean water.

Water further in the estuary also tends to be less saline than at the coast due to riverine

input. It is difficult to distinguish the effects of one variable from another as they are

often linked in time and space.

The third assumption concerns the changes in the environment over the time

period sampled (Bailey 1994, Brown 2000). Sampling within this study was conducted

on a monthly basis, so seasonal changes were well documented. Variations on a smaller

Page 25: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

time scale should also be recognized. Tidal changes of temperature and salinity should be

considered and quantified in some manner. Small-scale changes may have an impact on

the stress involved in the acclimation processes. It is reasonable to suspect that a fish that

is residing in waters fluctuating 15 ppt over a tidal cycle would be allocating more energy

to osmoregulation than a fish in water fluctuating a few parts per thousand. Energy spent

on maintaining homeostasis would be taken away from other important processes such as

growth and reproduction. Conversely, components of the model that are not fluctuating

on a monthly or seasonal basis should not have to be sampled as often. Sediment

composition is one of these static elements. However, it should be monitored periodically

to evaluate changes.

Lastly, it is important to consider the availability of certain habitats to the

organism. The fish must have an equal likelihood of occupying all habitats modeled.

Winter flounder are mobile, so it can be assumed that they have equal access to all areas

of the estuary. What may confound this logic is that young-of-the-year fish are very

sedentary after settlement and through the first few months of their lives (Saucerman &

Deegan 1991).

This study showed that these habitat suitability indices do not predict areas

maintaining high densities of flounder. They may predict where the habitats of highest

suitability are found, but they do not necessarily predict where the fish are located. The

reasons for this may lie in the assumptions outlined.

The choice of specific variables is important when creating a Habitat Suitability

Model. As mentioned in the introduction, components of the environment that have the

greatest effect on the organism should be the ones used for modeling. Variables chosen

Page 26: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

for these two models were those that are common in fishery data. However, these four

parameters may not necessarily be the ones that are driving habitat distinctions. Other

factors such as dissolved oxygen, current, submerged aquatic vegetation, and turbidity

may also be important habitat characteristics. Biotic variables such as the presence of

competitive species, predators, and potential prey items should also be considered.

Although they did not seem to have an effect on fish distribution in this study, they might

be of interest in other study areas or for a habitat suitability model using smaller fish

more susceptible to predation.

From the catch data, salinity appears to be the most important factor governing

flounder abundance. One point to make is that depth was not sampled over a great range.

This may be one reason why depth was not found to be significant. The range of

sediment grain sizes may have also represented a smaller range than what is available in

the estuary. Sites were chosen based partly on the accessibility of the trawl gear. Some

areas would have been impossible to trawl due to coarse substrates. The fish that were

sampled, juveniles to adults, may have also found the sediments within the range trawled

acceptable for burying. Smaller, young-of-the-year fish that were not sampled by the gear

may have preferred finer sediments.

Since sampling occurred between April and November, no extremes of

temperature and salinity were measured. Again, the confined range sampled of these

variables may not have eclipsed the values that are considered unsuitable. Winter

flounder are able to maintain themselves over a wide range of environmental conditions.

What the model designates as “fair” habitat may in fact be considered “good” by the fish.

They may be able to occupy areas with conditions that are on the periphery of the

Page 27: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

acceptable range. Each of these “fair” components multiplied in the HSM would produce

a very low overall suitability value. What we predict as low suitability may be in reality

acceptable to the fish.

What may be of more use is using the habitat suitability model to identify areas

that are completely unsuitable for one reason or another. It could also be used to

distinguish areas of potential optimal growth and survival. The Habitat Suitability Model

could be used to identify areas critical to specific life stages or spawning. These areas

could be delineated as “critical” habitat or areas (conditions) that are necessary to fish

reproduction or survival. This habitat could be associated with “critical phase,” or the

time in the fish’s life history when the size of the cohort is determined (Langton 1996).

That would be of more value than describing areas that are “good” or “fair” for juvenile

fish.

One assumption not mentioned above is that flounder populations are at carrying

capacity. That is, the density of flounders would be highest in areas of optimal habitat

and lower in areas that are sub optimal. Catch data from this study indicate that the

flounder population is not at capacity in the Great Bay Estuary System. Flounder catch

did not vary between sites. Catch also averaged well below one fish for every 100 m2

sampled. Compared to other estuaries, this catch per unit effort is very low (Pereira

1999). This suggests that either the flounder population is well below carrying capacity or

Great Bay does not have the habitat components and resources needed to support larger

numbers of fish.

Comparing data accumulated within this study with catch data from 1989 to

1991(Armstrong 1995), values are similar. Likewise, an independent sampling made

Page 28: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

contemporaneously with this study substantiates that catch is low in general (Fairchild

2002). Data from the previous five years show no significant changes in the amount of

catch with the exception of 2000, reflecting a strong year class (NH Fish & Game 2001).

Other than seine survey data, there is no long-term data series that can be used to

compare locations within the estuary over time and space.

Limited catch due to restricted carrying capacity seems unlikely, as the flounder

have demonstrated an exceptional ability to successfully inhabit a wide range of

conditions. They are, however, susceptible to density-dependent processes more than

pelagic fishes. Their environment is limited by surface area, as opposed to volume and

they are relatively limited in their mobility compared to pelagic fishes (Bailey 1994).

Several studies have evaluated the relationship between flatfish and their

environment. A variety of variables have been studied and different conclusions have

been made. In this study Habitat Suitability Indices were used to evaluate optimal habitat

conditions. Analysis of Variance was also used to investigate the relationship between

catch per unit effort and specific habitat components.

Winter flounder habitat requirements have been well documented (Bigelow &

Schroeder 1953, Klein-MacPhee 1978, Buckley 1989, Gibson 1994, Pereira 1999).

Flatfish growth, survival, and distribution have been associated with numerous variables:

temperature (Olla et al. 1969, Targett & McCleave 1974, Everich & Gonzalez 1977,

Casterlin & Reynolds 1982, Guelpen & Davis 1979, Phelan 2000), salinity (Frame 1973,

Burke et al. 1991, Armstrong 1995), depth (Oviatt & Nixon 1973, Abookire & Norcross

1998), sediment grain size (Pearcy 1978, Gibson & Robb 1992, Ansell & Gibson 1993,

Moles & Norcross 1995, Neuman & Able 1998, Abookire & Norcross 1998, Amezcua

Page 29: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

2001, Phelan 2001), sediment organic content (Oviatt & Nixon 1973, Stoner et al. 2001),

dissolved oxygen (Kramer 1987, Bejda et al. 1992, Phelan 2000, Meng 2001), water

current (Marchand 1991), turbidity/light intensity (McCracken 1963, Ansell & Gibson

1993), submerged aquatic vegetation (Sogard 1992, Wennhage & Pihl 1994, Stoner et al.

2001), man-made structures (Able et al. 1999), prey availability (Kennedy & Steele 1971,

Carlson 1997), predator presence (Bailey 1994, Leopold 1998), and competitive species

presence (Armstrong 1995).

In reality, an organism is affected by every aspect of its environment. The most

important components of the environment, or the ones that most influence an organism,

need to be identified and quantified. Recently, there has been a shift from simple single-

variable analyses to complex, multivariate ones. As many variables as possible are tested

simultaneously to elucidate the ones that would have the greatest impact on growth,

survival, or abundance.

Meng el al. (2001) used winter flounder growth rate to assess anthropogenic

influences on habitat quality. Temperature, salinity, dissolved oxygen, and benthic food

items were measured as well as the nitrogen and phosphorus content of the fish. Using

stepwise linear regression it was found that location within the estuary was significant to

growth. Higher growth rates were associated with lower salinities, smaller sizes, and

decreased time at low dissolved oxygen concentrations. It was also shown that the

abundance of benthic species was lower at sites with lower species diversity.

Phelan (2000) calculated instantaneous growth rates for winter flounder held in

cages in three types of estuarine habitats. Temperature and dissolved oxygen were shown

to have the greatest effect on growth rates.

Page 30: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Stoner et al. (2001) utilized generalized additive models to compare

environmental data to winter flounder abundance. Large catches of young-of-the-year

fish were associated with shallow depths, temperature near 22° C and macroalgae.

Abundance of prey items was also correlated with presence of smaller fish, but did not

correlate with abundance of larger fish. Independent sampling showed that the

generalized additive models were relatively useful in predicting relationships between

variables and fish distributions.

Norcross et al. (1997) used Analysis of Variance to study substrate, salinity,

temperature, dissolved oxygen, depth, and distance from shore. In addition, factor

analysis was used to determine which components were most important. It was concluded

that the four species of Pacific pleuronectids partitioned into four habitat types based on

temperature, depth, and substrate.

Szedlmayer (1996) looked at mean depth, temperature mean and range, salinity

mean and range, and percent silt in the sediment to characterize winter flounder habitat.

Similarity matrices were used to compare sites based on these data and the outcome was

plotted with non-metric multidimensional scaling. Percent silt in combination with

salinity was found to have the greatest effect on catch per unit effort.

Walsh (1999) tested salinity, turbidity, depth, distance from marsh edge, benthic

composition, and grain size as potential variables affecting winter flounder distribution.

Factor analysis was performed on a correlation matrix of the variables. Cluster analysis

was performed using the factor scores. It was determined that the eight or nine groups

produced represented eight or nine types of habitats.

Page 31: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

In summary, it is apparent that winter flounder success is related to specific

components of its environment. In each study it was found that variability in growth or

abundance was related to variability in the fish’s environment. However, the discrepancy

in results may suggest several things. First is the spatial variation in areas studied. It

would be reasonable to assume that flounder living in an estuary north of Cape Cod

would be acclimated to a very different environment than those south of the Cape.

Likewise, fish living in different estuaries would be subjected to a different variety of

predators. They would also have a different selection of prey items. Phelan (2000) found

differential growth in three estuaries in New Jersey. This was confirmed by comparing

the microstructure of the otoliths collected from fish in these areas (Sogard & Able 1992)

Second is that an estuary may have certain characteristics that would limit or

overemphasize certain variables. For example, an estuary with high freshwater input

might show that salinity is more important to flounder distribution than temperature.

Estuaries with moderate freshwater input might show temperature as a more important

variable than salinity during summer months. Therefore, limiting or controlling factors

that are important in one area may be completely different in another area. It should not

be assumed that they have similar effects on the organism.

Thirdly, it is important to reiterate the fact that winter flounder are tolerant of a

wide range of conditions. Young-of-the-year fish may be the least tolerant of all life

stages to environmental stresses, and additionally, are more susceptible to predation. It

would be reasonable to put efforts toward understanding settlement patterns and habitat

preferences for these fish. Because flounder are tolerant of many conditions, they might

not be the best species to use to demonstrate preferential habitat use.

Page 32: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Lastly, a distinction should be made between catch per unit effort measurements

and caged-fish growth experiments. Each measurement presents a bias of some kind.

Trawling or seining for fish is only as effective as the gear (Kuipers et al. 1992). In some

cases, efforts are biased by prior knowledge of the area. These data do not provide

information about the health, growth rates, or residence time of fish in that area. They do

provide information about general distributions and abundances for the areas studied.

They might provide information about areas of higher abundances or areas where no fish

were found.

Caging experiments tend to be artificial. Flounder are visual predators and their

success at feeding may decrease within the confines of the cage. Decreased feeding

would translate to decreased growth. Artificially high densities of fish within a cage

would also decrease feeding efficiency. Phelan (2000) also suspected that cages

prevented the fish from avoiding low dissolved oxygen events which would also limit

growth. If it was assumed that these effects were consistent between cages, then

important growth information could be gleaned from these studies. Two areas that have

shown to have similar catch per unit effort may in fact have different characteristics that

allow fish to grow at differential rates.

It would be valuable to use these two methods in conjunction as they complement

each other. Trawling may provide important information about general patterns of fish

distribution, while cage experiments would provide data on how successful fish are

within specific areas. Other methods of sampling such as seining and dive surveys may

also provide useful information for comparison. Understanding the temporal and spatial

dynamics of an area would be the first and most important step. Conducting trawls and

Page 33: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

caging experiments during times when conditions between areas are similar would not

provide useful information.

Predation on small winter flounder by crustaceans is well documented (Pihl & van

der Veer 1992, Keefe & Able 1994). This study shows that the presence of the sand

shrimp, Crangon septemspinosa had no effect on the presence of juvenile winter

flounder. This is not surprising as the fish caught in this study were beyond the size that

is susceptible to predation (Fairchild 2002). It was also found that there was a positive

relationship between winter flounder abundance and the predatory green crab, Carcinus

maenas. This does not indicate a predator-prey interaction but rather two populations that

are successful in the same conditions. Again, fish sampled were most likely beyond the

size range susceptible to predation by the green crab. Avian predators are also a concern

within estuaries. Winter flounder are susceptible to these predators as they move onto the

mud flats to feed (Leopold 1998).

There was also no significant relationship between the abundance of smooth

flounder and winter flounder. This was also not surprising as the catch per unit effort of

both species was very low. Conditions at sites where winter and smooth flounder were

collected were within a range that was tolerable to both (Armstrong 1995). If food were

not limiting, then there would be sufficient conditions to support both.

It is valuable to assess other ecological or physiological effects on the fish.

Quantifying the effects of predation on juvenile fish is still in its infancy. As is assessing

the effects of competitive or sympatric species. As we better understand these

relationships, we can better assess the use of specific habitats by flounder.

Page 34: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.1 Great Bay Estuary System of New Hampshire. Sampling sites indicated.Numbers correspond to sites sampled in the EPA’s Coastal 2000 Project. Map from Shortet al. 1992.

Site 73

Site 67

Site 25

Site 19

Site 23

Site 29

Site 51

Site 35

Page 35: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Table 1.1 Species caught by otter and beam trawl during 2000 and 2001.

common periwinkle (Littorina littorea)Eastern mud snail (Ilyanassa obsoleta)Atlantic dogwelk (Nucella lapillus)mysid shrimp (Mysis sp.)sand shrimp (Crangon septemspinosa)shrimp (Pandalus sp.)jonah crab (Cancer borealis)rock crab (Cancer irroratus)green crab (Carcinus maenus)blue crab (Callinectes sp.)hermit crab (Pagurus sp.)horseshoe crab (Limulus polyphemus)lobster (Homarus americanus)pipefish (Syngnathus fucus)Atlantic silverside (Menidia menidia)rainbow smelt (Osmerus mordax)herring (Clupea harengus)rock gunnel (Pholis gunnellus)cunner (Tautogolabrus adspersus)3-spine stickleback (Gasterosteus aculeatus)4-spine stickleback (Apeltes quadracus)9-spine stickleback (Pungitius pungitius)lumpfish (Cyclopterus lumpus)tomcod (Microgadus tomcod)winter flounder (Pseudopleuronectes americanus)smooth flounder (Liopsetta putnami)windowpane flounder (Scopthalmus aquosus)shorthorn sculpin (Myoxocephalus scorpius)grubby (Myoxocephalus aenaeus)mummichog (Fundulus heteroclitus)hake (Usophysis sp.)eel (Anguilla rostrata)sand lance (Ammodytes americanus)white perch (Morone americanus)alewife (Alosa pseudoharengus)chain pickeral (Esox niger)

Page 36: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Table 1.2 Overall habitat suitability index values for each site and month

September 2000 October 2000 November 2000

Site Brown SI Banner SI Site Brown SI Banner SI Site Brown SI Banner SI19 7.07 8.41 19 7.07 8.41 19 8.41 8.4123 2.66 6.69 23 2.24 6.69 23 2.24 6.6925 7.07 10.00 25 7.07 8.41 25 7.07 8.4129 3.98 7.95 29 3.98 7.95 29 3.98 7.9535 3.98 7.95 35 3.98 6.69 35 3.98 6.6951 3.98 7.95 51 3.98 6.69 51 3.98 6.6967 7.07 10.00 67 7.07 8.41 67 7.07 8.4173 3.98 7.95 73 3.34 6.69 73 3.98 6.69

April 2001 May 2001 June 2001

Site Brown SI Banner SI Site Brown SI Banner SI Site Brown SI Banner SI19 0.00 0.00 19 7.07 10.00 19 7.07 8.4123 3.16 7.95 23 2.24 6.69 23 3.16 7.9525 0.00 0.00 25 7.07 10.00 25 5.95 8.4129 3.98 7.95 29 3.98 7.95 29 4.73 9.4635 3.98 6.69 35 3.34 6.69 35 2.66 4.4751 3.98 6.69 51 4.73 7.95 51 3.98 6.6967 5.95 8.41 67 7.07 10.00 67 7.07 8.4173 3.34 6.69 73 3.98 7.95 73 3.34 6.69

July 2001 August 2001 September 2001

Site Brown SI Banner SI Site Brown SI Banner SI Site Brown SI Banner SI19 4.73 5.62 19 3.16 5.62 19 5.62 10.0023 3.16 7.95 23 2.24 6.69 23 2.66 6.6925 3.98 5.62 25 3.98 8.41 25 4.73 10.0029 4.73 9.46 29 4.73 9.46 29 4.73 9.4635 3.98 7.95 35 3.16 7.95 35 2.66 7.9551 2.24 6.69 51 2.66 6.69 51 3.16 7.9567 3.98 5.62 67 4.73 8.41 67 5.62 10.0073 3.16 4.47 73 2.24 6.69 73 3.98 7.95

October 2001 November 2001

Site Brown SI Banner SI Site Brown SI Banner SI19 8.41 10.00 19 4.73 8.4123 2.66 6.69 23 2.24 3.7625 8.41 10.00 25 4.73 8.4129 4.73 7.95 29 3.98 7.9535 2.24 6.69 35 2.24 6.6951 2.66 6.69 51 4.73 6.6967 8.41 10.00 67 7.07 8.4173 3.98 7.95 73 3.34 6.69

Page 37: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.2 Winter Flounder CPUE compared to Banner Suitability Index values

0

0.1

0.2

0.3

0.4

0.5

0.6

0 2 4 6 8 10

Banner Suitability Index value

CPU

E (

WF/

100

sq.m

.)

Page 38: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.11 Winter flounder CPUE compared to Mysid Shrimp CPUE

y = -0.0089x + 0.0641

R2 = 0.0094

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 0.5 1 1.5 2 2.5 3 3.5Mysid Shrimp CPUE

Win

ter

flou

nder

CPU

E (

WF/

100

sq.m

)

Page 39: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.8 Winter flounder CPUE over depth gradient

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Depth (feet)

CPU

E (

WF/

100s

qm)

Page 40: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.6 Winter Flounder CPUE over temperature gradient.

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Temperature (Celcius)

CPU

E (

WF/

100

sq.m

.)

Page 41: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.7 Winter Flounder CPUE over salinity gradient

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25 30 35 40

Salinity (ppt)

CPU

E (

WF/

100

sq.m

.)

Page 42: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.10 Winter flounder CPUE compared to Organic Composition of the Sediment

y = 0.0027x + 0.0371

R2 = 0.0202

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25

Organic Composition (% loss on ignition)

CPU

E (

WF/

100

sq.m

.)

Page 43: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.9 Winter Flounder CPUE over percent sand composition of sediment

0

0.1

0.2

0.3

0.4

0.5

0.6

0 20 40 60 80 100

% Sand Composition in Sediment

CPU

E (

WF/

100

sq.m

.)

Page 44: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.3 Winter Flounder CPUE compared to predicted suitability index value

0

0.1

0.2

0.3

0.4

0.5

0.6

0 2 4 6 8 10

Brown SI Value

CPU

E (

WF/

100

sq.m

.)

Page 45: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.12 Winter flounder CPUE compared to Crangon Shrimp CPUE.

y = 0.0698x - 0.0057

R2 = 0.2586

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 0.5 1 1.5 2 2.5 3

Crangon Shrimp CPUE (Shrimp/ 100 sq.m)

Win

ter

Flou

nder

CPU

E (

WF/

100

sq.m

)

Page 46: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.13 Winter flounder CPUE compared to Green Crab CPUE

y = 0.1567x + 0.0249

R2 = 0.2004

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 0.2 0.4 0.6 0.8 1 1.2 1.4

GRCRAB

WF

Page 47: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.14 Winter flounder CPUE compared to Smooth flounder CPUE

y = 0.0919x + 0.061

R2 = 0.0026

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Smooth flounder CPUE (SF/ 100 sq.m)

Win

ter

flou

nder

CPU

E (

WF/

100

sq.

m)

Page 48: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.4 Winter flounder CPUE by site (all months combined)

0

0.1

0.2

0.3

0.4

0.5

0.6

0 10 20 30 40 50 60 70 80

Site

CPU

E (

WF/

100

sq.m

.)

Page 49: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT
Page 50: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 1.5 Winter Flounder CPUE by month (all sites combined)

0

0.1

0.2

0.3

0.4

0.5

0.6

Month

CPU

E (

WF/

100

sq. m

.)

Page 51: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

LITERATURE CITED

Able, K.W., Manderson, J.P., Studholme, A.L. 1999. Habitat quality for shallow waterfishes in an urban estuary: The effects of man-made structures on growth. Mar.Ecol. Prog. Ser. 187: 227-235.

Abookire, A. A. and B. L. Norcross 1998. Depth and substrate as determinants ofdistribution of juvenile flathead sole (Hippoglossoides elassodon) and rock sole(Pleuronectes bilineatus), in Kachemak Bay, Alaska. J. Sea. Res. 39: 113-123.

Amezcua, F. and R. D. M. Nash. 2001. Distribution of the order Pleuronectiformes inrelation to the sediment type in the North Irish Sea. J. Sea Res. 45: 293-301.

Ansell, A. D. and R. N. Gibson. 1993. The effect of sand and light on predation ofjuvenile plaice (pleuronectes platessa) by fishes and crusteceans. J. Fish. Biol. 43:837-845.

Armstrong, M.P. 1997. Seasonal and ontogenetic changes in distribution and abundanceof smooth flounder, Pleuronectes putnami, and winter flounder, Pleuronectesamericanus, along estuarine depth and salinity gradients. Fish. Bull. 95: 414-430.

Armstrong, M.P. 1995. A comparative study of the ecology of smooth flounder,Pleuronectes putnami, and winter flounder, Pleuronectes americanus, from GreatBay Estuary, New Hampshire. Ph.D. Dissertation. University of New Hampshire.147p.

Bailey, K.M. 1994. Predation on Juvenile Flatfish and Recruitment Variability. Neth. J.of Sea Res. 32(2): 175-189.

Banner, A., and G. Hayes. 1996. Mapping Important Habitats of Coastal NewHampshire. Chapter 6: Winter flounder.http://gulfofmaine.org/library/gbay/wfl.htm.

Bejda, A. J., Phelan, B.A., and A. L. Studholme. 1992. The effect of dissolved oxygen onthe growth of young-of-the-year winter flounder, Pseudopleuronectesamericanus. Env. Biol. Fish. 34: 321-327.

Bigelow & Schroeder. 1953. Fishes of the Gulf of Maine. U.S. Fish & Wildl. Serv., Fish.Bull. 53: 577p.

Brown, S. K., Buja, K.R., Jury, S.H., Monaco, M.E., and A. Banner. 2000. HabitatSuitability Index Models for Eight Fish and Invertebrate Species in Casco andSheepscot Bays, Maine. N. Amer. J. of Fish Man. 20(2): 408-435.

Buckley, J. 1989. Species Profiles: Life histories and environmental requirements of

Page 52: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

coastal fishes and invertebrates (North Atlantic) - winter flounder. U.S. Fish andWildlife Service Biological Report 82(11.87).

Burke, J. S., Miller, J.M. and D. E. Hoss. 1991. Immigration and Settlement Pattern ofParalichthys dentatus and P. Lethostigma in an Estuarine Nursery Ground, NorthCarolina, U.S.A. Neth. J. of Sea Res. 27(3/4), 393-405.

Carlson, J. K., Randall, T.A., and M. E. Mroczka 1997. Feeding Habits of WinterFlounder (Pleuronectes americanus) in a Habitat Exposed to AnthropogenicDisturbance. J. of Northwest At. Fish. Sci. 21: 65-73.

Casterlin, M. E. and W. W. Reynolds. 1992. Thermoregulatory behavior and diel activityof yearling winter flounder, Pseudopleuronectes americanus (Walbaum). Env.Biol. Fish. 7(2): 177-180.

Everich, D. and J. G. Gonzalez. 1977. Critical Thermal Maxima of Two Species ofEstuarine Fish. Marine Biology 41: 141-145.

Fairchild, E.A. 2002. Winter Flounder Pseudopleuronectes americanus stockenhancement in New Hampshire: developing optimal release strategies. Ph.D.Dissertation, University of New Hampshire. 142pp.

Fairchild, E.A. and W.H. Howell. 2000. Predator-prey size relationship betweenPseudopleuronectes americanus and Carcinus maenas. J. Sea Res. 44(1/2): 81-90.

Frame, D. W. 1973. Biology of Young Winter Flounder Pseudopleuronectes americanus(Walbaum): metabolism under simulated estuarine conditions. Trans. Am. Fish.Soc. 102: 423-430.

Gibson, R. N. 1994. Impact of Habitat Quality and Quantity on the Recruitment ofJuvenile Flatfishes. Neth. J. of Sea Res. 32(2): 191-206.

Gibson, R. N. and L. Robb. 1992. The relationship between body size, sediment grainsize and the burying ability of juvenile plaice, Pleuronectes platessa (L.) J. FishBiol. 40: 771-778.

Grout, D., McBane, C., Patterson, C., Smith, B., Trested, D., and C. Baker. 2001.Programs Improving Management of ASMFC Managed Species in NewHampshire. 2001 Final Report. NMFS Federal Aid Project 3-ACA-071. N.H. Fish& Game Dept. 31p.

Keefe, M. and K. W. Able. 1994. Contributions of Abiotic and Biotic Factors toSettlement in Summer flounder, Paralichthys dentatus. Copeia 2: 458-465.

Kennedy, V. S. and D. H. Steele. 1971. The Winter Flounder (Pseudopleuronectesamericanus) in Long Pond, Conception Bay, Newfoundland. J. Fish. Res. Bd.Canada 28: 1153-1165.

Page 53: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Klein-MacPhee, G. 1978. Synopsis of Biological Data for the Winter Flounder,Pseudopleuronectes americanus (Walbaum). NOAA Technical Repost NMFSCircular 414: 1-43.

Kramer, D. L. 1978. Dissolved oxygen and fish behavior. Environmental Biology ofFishes 18(2): 81-92.

Kuipers, B. R. et al. 1992. Small Trawls in Juvenile Flatfish Research: TheirDevelopment and Efficiency. Neth. J. Sea Res. 29(1-3): 109-117.

Langton, R. W., Steneck, R. S., Gotceitas, V., Juanes, F., and P. Lawton. 1996. TheInterface between Fisheries Research and Habitat Management. N. Am. J. of Fish.Management 16: 1-7.

Leopold, M.F., van Damme, C. and H.W. van der Veer. 1998. Diets of cormorants andthe impact of comorant predation on juvenile flatfish in the Dutch Wadden Sea. J.Sea Res. 40: 93-107.

McCracken, F. D. 1963. Seasonal Movements of the Winter flounder,Pseudopleuronectes americanus (Walbaum), on the Atlantic Coast. J. Fish. Res.Bd. Canada 20(2): 551-583.

Meng, L., Powell, J. C., and B. Taplin. 2001. Using Winter Flounder Growth Rates toAssess Habitat Quality Across an Anthropogenic Gradient in Narragansett Bay,Rhode Island. Estuaries 24(4): 576-584.

Moles, A. and B. L. Norcross. 1995. Sediment Preference in Juvenile Pacific Flatfishes.Neth. J. of Sea Res. 34(1-3): 177-182.

Neuman, M. J. and K. W. Able. 1998. Experimental evidence of sediment preference byearly life history stages of windowpane (Scophthalmus aquosus). J. Sea Res. 40:33-41.

Norcross, B. L., Muter, F., and B. A. Holladay. 1997. Habitat models for juvenilepleuronectids around Kodiak Island, Alaska. Fish. Bull. 95: 504-520.

Olla et al. 1969. Behavior of Winter Flounder in a Natural Habitat. Trans. Amer. Fish.Soc. 4: 717-720.

Oviatt, C. A. and S. W. Nixon. 1973. The Demersal Fish of Narragansett Bay: anAnalysis of Community Structure, Distribution and Abundance. Estuarine andCoastal Marine Science. 1: 361-378.

Pearcy, W. G. 1978. Distribution and Abundance of Small Flatfishes and other DemersalFishes in a Region of Diverse Sediments and Bathymetry off Oregon. FisheryBulletin 76(3): 629-640.

Page 54: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Pereira, J., Goldberg, J. R., Ziskowski, J. J., Berrien, P. L., Morse, W. W., and D. L.Johnson. 1999. Winter Flounder, Pseudopleuronectes americanus, Life Historyand Habitat Characteristics. NOAA Technical Memorandum NMFS-NE-138: 1-39.

Phelan, B. A., Manderson, J. P., Stoner, A.W., and A. J. Bejda. 2001. Size-related shiftsin the habitat associations of young-of-the-year winter flounder(Pseudopleuronectes americanus): field observations and laboratory experimentswith sediments and prey. J. of Exp. Mar. Bio. Eco. 257: 297-315.

Phelan, B. A., Goldberg, R., Bejda, A. J., Pereira, J., Hagan, S., Clark, P., Studholme, A.L., Calabrese, A., and K. W. Able. 2000. Estuarine and habitat-related differnecesin growth rates of young-of-the-year winter flounder (Pseudopleuronectesamericanus) and taugtog (Tautoga onitis) in three northeastern US estuaries. J.Exp. Mar. Bio. Eco. 247: 1-28.

Pihl. L. and H. W. van der Veer. 1992. Importance of Exposure and Habitat Structure forthe Population Density of 0-Group Plaice, Pleuronectes platessa L., in CoastalNursery Areas. Neth. J. of Sea Res. 29(1-3): 145-152.

Rubec, P. J et al. 1998. Spatial Methods Being Developed in Florida to DetermineEssential Fish Habitat. Fisheries. 23(7): 21-25.

Saucerman, S. E. and L. A. Deegan. 1991. Lateral and Cross-Channel Movement ofYoung-of-the-Year Winter Flounder (Pseudopleuronectes americanus) inWaquoit Bay, Massachusetts. Estuaries 14(4): 440-446.

Sogard, S. M. 1992. Variability in growth rates of juvenile fishes in different estuarinehabitats. Mar. Ecol. Prog. Ser. 85: 35-53.

Sogard, S. M., Able, K. W., and S. M. Hagan. 2001. Long-term assessment of settlementand growth of juvenile winter flounder (Pseudopleuronectes americanus) in NewJersey estuaries. J. Sea Res. 45: 189-204.

Stoner, A. W., Manderson J. P., and J. P. Pessutti. 2001. Spatially explicit analysis ofestuarine habitat for juvenile winter flounder: combining generalized additivemodels and geographic information systems. Mar. Ecol. Prog. Ser. 213: 253-271.

Szedlmayer S.T. and K.W. Able 1996. Patterns of Seasonal Availability and Habitat Useby Fishes and Decapod Crustaceans in a Southern New Jersey Estuary. Estuaries19(3): 697-709.

Page 55: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Targett, T. E. and J. D. McCleave. 1974. Summer Abundance of Fishes in a

Maine Tidal Cove With Special Reference to Temperature. Trans. Amer. Fish. Soc. 2:

325-330.

Van Guelpan, L. and C.C. Davis. 1979. Seasonal Movements of the winter flounder,Pseudopleuronectes americanus, in two contrasting inshore locations inNewfoundland. Trans. Am. Fish. Soc. 108: 26-37.

Wennage, H. and L. Pihl. 1994. Substratum Selection by Juvenile Plaice (Pleuronectesplatessa L.): Impact of Benthic Microalgae and Filamentous Macroalgae.Netherlands J. Sea Res. 32(3/4): 343-351.

CHAPTER II

WINTER FLOUNDER STOMACH CONTENT ANALYSIS ANDMACROFAUNAL BENTHIC COMMUNITY ANALYSIS

Introduction

The Sustainable Fisheries Act defines essential fish habitat as “those waters and

substrate necessary to fish for spawning, breeding, feeding, or growth to maturity” (SFA

1996). To maintain any population, females within the population must reproduce and

replace themselves over the span of their lives. First they must reach the age of sexual

maturity and to do this they must grow, maintain health, and avoid mortality. “Three

factors are generally accepted as necessary for successful maintenance of populations

(recruitment) adequate food, refuge from predation, and a benign abiotic environment

(Miller 1991).” The focus of this chapter will be the effects of food availability on growth

and recruitment.

Page 56: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Growth to maturity requires the evolution and use of several physiological and

behavioral adaptations used in counteracting environmental stresses. Physiological

adaptations include such processes as the fish’s ability to thermoregulate and

osmoregulate. Behavioral adaptations include such activities as foraging, predator

avoidance, and spawning. Behavioral adaptations also compensate where physiological

adaptations fall short, such as the ability of a flatfish to bury into a substrate to avoid high

temperature. A fish’s ability to perform these functions will dictate its overall success in

the form of growth and survival.

All adaptations rely on one assumption; that the fish has sufficient energy to

perform them, whether it be to physically move or to drive the membrane potential of a

cell. In this sense, energy acquisition, or food consumption, becomes essential to survival

to maturity and to reproductive success. Growth, metabolism, and excreted products are

equal to the amount and quality of the food that the animal consumes (Eckert 1997). Food

provides not only the energy to drive internal processes, but also supplies the raw

materials needed for growth and repair of the body. In the end, a fish’s diet provides the

necessary components to construct, and the energy to release, the gametes that will go on

to recruit to the population.

After metamorphosis winter flounder are demersal, their diets limited to the

epifauna that reside on the substrate. To some extent they are exposed to emergent

infauna such as the siphons of clams and feeding polychaetes. They are also constrained

in their gape size. Despite these limitations, winter flounder are basically omnivorous and

opportunistic, having been documented to feed on over 260 species of fish, invertebrates,

and algae (Klein-MacPhee 1978). The importance of specific items within the diets of

Page 57: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

adult fish would be difficult to discern for these reasons. Juveniles, however, are very

limited by their gape size and predatory abilities. Their diets contain small, sessile

organisms that they encounter as they swim along the bottom. Juvenile fish are also

limited by their swimming ability. With their decreased range, there is a decrease in the

number and type of prey items available to them. The potential prey taxa are limited by

geographic location, not just on the global scale, but also on a local scale, residing on

different substrates.

Juvenile winter flounder are residents of estuaries for the first 1-2 years of their

life (Klein-MacPhee 1978). Within the estuary, there are specific habitats defined by their

temperature and salinity regimes, substrate type, submerged aquatic vegetation, etc.

These habitats provide homes to specific assemblages of benthic organisms. If winter

flounder are constrained by prey availability and there are different assemblages of

benthic organisms, then it might be reasonable to suspect that some groups provide a

better diet for the flounder than other groups. It might be more advantageous for the

flounder to reside in these areas.

This chapter will address the concept that one habitat might be more essential

than another simply because it provides the fish with a better diet. First the diet or diets of

the juvenile winter flounder within the estuary were characterized through stomach

content analysis. Second, sites within the estuary were evaluated by core samples to see if

they support unique assemblages and if those groups are maintained over time. Lastly,

the core information was examined to see if statements could be made about whether one

site provides a “better” prey assortment than another site.

Page 58: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Materials and Methods

Site Selection

Eight sites were chosen within the Great Bay Estuary System of New Hampshire

(Figure 2.1). Sites were chosen to represent a gradient of four environmental parameters:

temperature, salinity, depth, and substrate. These parameters were chosen based on the

typical ones chosen to create Habitat Suitability Indices. For example, Site 23 is

characterized by a consistent, high salinity; Site 51, a lower salinity with a mild tidal

fluctuation; and Site 73, a low salinity with high tidal fluctuation. These parameters

represent the habitat characteristics that are believed to affect winter flounder distribution

most dramatically. Sites were sampled from September to November 2000 and from

April to November 2001.

Sampling

Winter flounder were collected by a 4.8-m otter trawl (2.5-cm body and 6-mm

codend) and a 1-meter beam trawl (6-mm body and 3-mm body). Each trawl net was

utilized twice for a total of four 10-minute tows during each month of sampling. Each

fish was measured (standard length to the nearest millimeter), placed on ice, and returned

to the lab where they were frozen immediately.

Core samples were taken at each site with an eight-foot long manual corer (core

area = 0.0079 m2) to a depth of 10 cm. Cores (six replicates) were stored in 3.79 liter Zip-

lock‰ bags, placed on ice, and returned to the lab where they were sieved through a 1-

mm mesh sieve. All organisms were fixed in 10% buffered formalin and stored in 40%

ethanol.

Page 59: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Identification

Fish were allowed to thaw before dissection. The stomach was removed ventrally

from the fish from the esophagus to the pyloric region. Contents were identified to the

lowest taxonomic group, counted, and total weight recorded to the nearest 0.001 gram.

Organisms within the core samples were identified to the lowest taxonomic group,

enumerated, and total weight recorded to the nearest 0.001 gram.

Stomach Content Analyses

A total of 178 fish were collected for stomach content analysis (mean length =

129mm, range = 50 to 450mm). Of the fish sampled, 123 fish held identifiable contents,

and these fish were used for the following analyses:

Index of Relative Importance (I.R.I.) (Hyslop 1980)

The index uses three measurements to determine an overall relative importance

of each item within the diet of the flounder. By using three measurements, the index

reduces the amount of bias due to each individual measurement. Measures of numerical

abundance tend to give too much importance to small, abundant species, and less to large,

rare species. Conversely, measures of volume overemphasize large, less abundant species

and under estimate small, abundant species. Again, measures of frequency tend to bias

towards small abundant items. The Index of Relative Importance compensates for these

biases. It is defined as follows:

I.R.I. = (N + V) x F

Page 60: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

where N is equal to the percent numerical abundance (the number of individuals of each

prey species divided by the total number of food items found in all stomachs). V is equal

to the percent volume (the volume of each prey species divided by the total volume of

food from all stomachs). F is equal to the percent frequency of occurrence (the number of

stomachs in which a prey species occurred divided by the total number of stomachs

containing food) (Hyslop 1980).

Size Class Analysis

Fish were divided into five categories to determine if prey selection changes with

an increase in length. The categories were based loosely on size classes related to specific

age classes. Five categories were used: 50-100, 101-150, 151-200, 201-250, and 250+

mm. No fish smaller than 50 mm were used for the analyses as they were rare in

sampling. Eleven species were used to characterize the diets. These eleven were

determined from the I.R.I. Percent numerical abundance was used to calculate the

constituents of the fish size class analysis.

Seasonal Analysis and Site Analysis

Season was considered using three size classes: 50-100, 101-250, 250mm and

greater. Numerical abundance was used as the measure. Site was considered using the

three size classes described above. Numerical abundance was used as the measure.

Page 61: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Core Analysis

Although over 30 species were identified within the cores, only the eleven that

represented the diet of the winter flounder were considered. Because of a realistic overlap

in species composition of the cores and the way sampling was performed, analyses were

confined to grouping analyses such as Cluster Analysis and Multidimensional Scaling

(Field et al. 1982). Although not quantitative, both provide important information about

the similarities and differences of prey composition between sites. Analyses were

performed on numerical abundance data. The average number of a species was calculated

for each month and site. Because the analyses used were comparing abundances across

categories, the data were root-root transformed.

Two types of analyses were performed on the core data. The species composition

of each site was compared to that of each other site. Because there was also a sense that

these sites would change over time, a seasonal component was added. Therefore the

complement of species was compared between sites within the same two-month time

frame. The time frames were Oct-Nov 2000, Apr-May 2000, Jun-Jul 2001, Aug-Sep

2001, and Oct-Nov 2001. Second, the complement of species at each site was

characterized to see how it changed over time. To do this, the numerical abundance of

each prey species for each month was compared at each site.

Cluster Analysis

Cluster Analysis is useful in finding group structure. A cluster is defined as a

group of objects that are close together and are separated from other groups (US Fish &

Wildlife: Multivariate Statistical Analysis). After the data were root-root transformed, a

Page 62: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

similarity index was applied to the matrix. The Bray-Curtis Similarity Equation produces

an output, on a scale of zero to one, which is indicative of the similarities between data

sets (Field et al 1982).

Bray-Curtis1,2 = 2S Max(n1i,n2i)

S n1i + S n2i

Where 1 and 2 indicate the two data sets, “i,” any species within the sample, and “n,” the

abundance of that species. The equation calculates the difference (or the dissimilarity)

between the complement of species found at any 2 sites within a season. In this way all

sites were compared to all other sites. The dissimilarity, or distance between sites was

imputed into a clustering program (Systat 10). Clusters that were closest in distance were

clustered together. The output of the cluster analysis is a dendrogram, connecting groups

together by the similarity distances.

Multidimensional Scaling

Multidimensional Scaling (MDS) is also useful in distinguishing between groups.

Principle Coordinates Analysis takes a set of distances (or a similarity matrix) and finds a

set of coordinates that produce Euclidean distances that are closest to measured similarity

distances. “Stress” is calculated as a measure of the ability of the program to produce

distances that are closest to the actual distances, or a measure of fit. A stress close to zero

is considered excellent. A Shepard diagram is produced which plots the values of the

calculated distances versus the actual distances. Similar to the stress calculation, this is a

way to visualize how well the data were scaled. If the distances were matched well, then

Page 63: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

one would see a set of points which fall on the y = x line. MDS is similar to Principle

Components Analysis in that it finds the dimensions that provide the greatest amount of

variability within the data set. A graph is produced that plots the points in two

dimensions. These graphs can be interpreted by looking for odd points, clusters, or

patterns (U.S. Fish and Wildlife Multivariate Statistical Analyses 2001, Systat 10).

Results

Stomach Content Analysis

Of the 178 fish collected for stomach content analysis, 123 held identifiable

contents. Fifteen prey taxa were identified within the stomachs of the fish. Of these

fifteen, eleven taxa made up 99% of the numerical abundance and 96% of the wet weight.

These eleven items were used for the remainder of the analyses.

Calculation of numerical abundance showed that Amphipodae sp. made up 81%

of the total number of stomach contents found in all stomachs. Capitellidae sp. accounted

for 5.4% of the total, Spionidae sp. for 4.0%, Mya arenaria siphons for 3.4%, and

Anthuridae for 2.9%.

Calculation of abundance by weight showed that Spionidae sp. accounted for 65%

of the total amount of stomach contents. Amphipods followed, comprising 12% of the

total. Others included Mya arenaria siphons (6.2%), Crangon septemspinosa (3.2%), and

Anthuridae sp. (2.9%).

An Index of Relative Importance was calculated using the stomach content data

(Table 2.1). According to the index, Amphipod sp. were determined to be the most

Page 64: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

important (I.R.I. = 3883), followed by Spionidae sp. (I.R.I. = 2312), Mya arenaria

siphons I.R.I. = 124), Anthuridae sp. (I.R.I. = 121), and Capitellidae sp. (I.R.I. = 98).

Percent Numerical Abundance (Fish Size Classes) (Figure 2.2)

Amphipods were determined to be the most important diet item for fish in size

classes 50-100, 101-150, 151-200, and 201-250 mm, comprising more than 75% of the

total number of items found in all stomachs within these size classes.

Anthuridae, Arabellidae, Spionidae, and Capitellidae were also important to fish

within the length class 50 to 100 mm, but made up only approximately 12% of the diet.

Mya arenaria siphons were found in the diets of fish greater than 100mm.

Anthuridae, Myidae, Spionidae, and Capitellidae were important in the diets of fish

between 101 and 250 mm in length, although they make up substantially less of the diet

than the amphipods: 23% in fish 101-150 mm, 15% in fish 151-200 mm, and 21% in fish

201-250 mm in length.

The length classes of fish 201-250 mm and 250+ mm show the greatest variation

of dietary items; 10 and 9 species respectively. This is compared to 5 species (50-100), 7

species (101-150), and 6 species (151-200). The size class 250+ mm shows the greatest

divergence from the other length classes. Spionidae sp. were the dominant prey item at

35% of the numerical abundance. Species that made up the “other” category (those four

species which did not make the top eleven) accounted for 40% of the diet. Amphipods

Page 65: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

made up less than 2% of the diet. Other items included Anthuridae, Crangon

septemspinosa, Carcinus maenas, whole Myidae, Mya arenaria siphons, and Arabellidae.

Percent Numerical Abundance by Season (Figure 2.3)

During the fall of 2000, Amphipods dominated the diets of small (50-100mm

fish) in numerical abundance (50%). They were followed closely by Capitellidae sp.

which made up 40% of the diet. Other species included Spionidae sp. and C .

septemspinosa. Amphipods dominated the diets of fish 101-250 mm in length, making up

97% of the diet. Spionidae sp., Mya arenaria siphons, and Orbiniidae made up the other

3%. Spionidae sp. made up 100% of the diet of fish in the length class 250+ mm.

Amphipods comprised 50% of the diets of fish within the length class of 50-100

mm during the spring of 2001. Arabellidae sp. made up 30%, and Anthuridae and

Spionidae together made up 10% of the total numerical abundance in all stomachs.

Again, amphipods dominated the diets of 101-250 fish with 62% of the total numerical

abundance. Anthuridae made up 13% of the total. The other 25% was comprised of

whole Myidae, Mya arenaria siphons, Tellinidae, Arabellidae, Spionidae, and

Capitellidae. Spionidae sp. made up 65% of diet of fish in the 250+ length category.

Carcinus maenas, Mya arenaria siphons, and other items made up the remaining 35%.

During the summer of 2001, amphipods made up 98% of the diet of small (50-

100mm) fish. Capitellidae sp. made up the other 2%. The diets of fish in the 101-250mm

size class were broken down as follows: 50% amphipods, 20% Mya arenaria siphons,

Page 66: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

15% Capitellidae sp. and the other 15%: Anthuridae, Crangon, Spionidae, Orbiniidae.

The “other” category made up 44% of the diets of the fish in the 250+ length class.

Spionidae sp. made up 32% of the total and Amphipoda, Anthuridae, Crangon, Carcinus

maenas, Mya arenaria, Arabellidae made up the other 24%.

No fish greater than 250 mm were caught in the fall of 2001. Amphipods

accounted for 98% of the diet items found in the stomachs of fish within the 50-100 mm

length class. Spionidae sp. made up the other 2%. Amphipods made up 80% of the diets

of fish 101-250 mm in length. Myidae, Spionidae, Capitellidae, and Orbiniidae made up

the other 20%.

Stomach Content at Each Site (Fish Length Classes: 50-100mm, 101-250, and 250+)

(Figure 2.4)

Data were not collected from fish of the following sites and length classes: Site

19: 50-100 mm and 250+ mm, Site 23: 50-100 mm, Site 35: 250+ mm, Site 67: 50-100

mm and 250+ mm, and Site 73: 250+ mm.

For the length class 50 to 100 mm, Amphipoda were the dominant prey items

(greater than 50%) at sites 25, 29, and 35. They were not dominant at Sites 23, 51, and

73. Within the length class 101 to 250 mm, Amphipoda were dominant at Sites 19, 25,

35, and 51. They were not dominant at Sites 23, 29, 67, and 73. Fish in the 250+ mm size

class fed on a variety of items, but the most notable pattern was the lack of amphipods in

the diet: 0% at Site 23, 0% at Site 25, 4% at Site 29, and 0% at Site 51.

Page 67: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Core Data: Total Percent Average of Core Taxa Over Entire Study Period (Figure

2.5)

Over 30 items were identified within the core samples (Table 2.2). The eleven

items that were identified to be the primary taxa important in the diets of winter flounder

were used in the analyses of the core data. Examining the total average abundance of core

taxa over the entire study period, one sees the same constituents: amphipods, Anthuridae,

Tellinidae, Arabellidae, Spionidae, and Capitellidae. The one notable exception to this is

that no amphipods were collected from sites 67 and 73.

Combined Average Number of Each Taxa per Core at Each Site Over Time.

(Figure 2.6 a-h)

Peaks in abundance were seen during August and November at Sites 19, 25, 23,

and 29. An additional peak was seen at Site 25 in June. Peaks in abundance were seen in

November 2001 at Site 35 and October of 2001 at Site 51. A peak was seen during

September and October 2001 at site 67. Similarly, high abundances were recorded in

August, September, and October 2001 at Site 73.

The general makeup of these sites was very similar with a few notable exceptions.

Sites 25 and 35 contained Orbiniidae. Sites 23, 35, and 51 did not hold high amounts of

Tellinidae. Sites 67 and 73 did not contain amphipods.

Eleven Species by Month (Figure 2.7 a-h)

Many of the items within the cores had fall peaks: Amphipoda, Anthuridae,

Myidae, Tellinidae, Arabellidae, Spionidae, Capitellidae, and Orbinidae. Amphipods also

Page 68: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

had peaks in April and July. Myidae also had a peak in June. Tellinidae also had a peak in

May. Spionidae also had peaks in May and August.

Cluster Analysis and Multidimensional Scaling (MDS): Sites Grouped by Season

October-November 2000 (Figure 2.8 a,b) Sites 67 and 73 were clustered at a

distance of 0.25. Sites 19 and 25 were also grouped at a distance of 0.25. All sites

were grouped at a distance of 0.65. The Multidimensional Scale showed sites 19, 25,

29 grouped together, sites 51 and 35 together, and 73 and 67 together. Stress of the

MDS plot was 0.0494.

April-May 2001 (Figure 2.9 a,b) The cluster analysis grouped sites 67, 73, 29,

25, and 19 at a distance of less than 0.25 and all sites were grouped at a distance of

0.5. The MDS plot placed 23 close to 51, but all others were scattered. Stress was

equal to 0.1348.

June-July 2001 (Figure 2.10 a,b) Sites 51 and 23 were joined at a distance of

0.1 in the cluster analysis. Sites 25, 73, 35, and 19 joined into the cluster at 0.3. All

sites were clustered at 0.5. Sites 23 and 51 were again close in the MDS plot. All

others were scattered. Plot stress was equal to 0.08226.

August-September 2001 (Figure 2.11 a,b) Sites 23 and 51 were joined at 0.15

as well as sites 67 and 73. All sites were joined at 0.25. MDS confirmed this by

Page 69: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

clustering sites 67 and 73 together and 51 and 23 together. All others were scattered.

Stress of the MDS plot was 0.04961.

October-November 2001 (Figure 2.12 a,b) Sites 25 and 67 were joined at a

distance of 0.15 and all others were joined at a distance of 0.4. MDS also plotted 25

and 67 close, but also 19 and 73 were near. All others were scattered. Stress was

equal to 0.08462.

Cluster Analysis and Multidimensional Scaling: Site Over Time (Month)

Site 19 (Figure 2.13 a,b) All joined in cluster analysis at 0.5 except April at 1.0.

Multidimensional Scaling shows that Sep01, Oct01, and Nov01 to be close. Jul01 and

Aug01 are close as well. Stress of the plot is equal to 0.05066.

Site 23 (Figure 2.14 a,b) Aug01, Jul01, Sept01, and Oct01 joined at a distance of

0.5. All were joined at a distance of 0.7 except April at 1.0. MDS showed all the points

scattered with a stress of 0.05044.

Site 25 (Figure 2.15 a,b) All sites were joined in the cluster analysis at a distance

of 0.4 except April which joined at a distance of 1.0. The MDS plot showed Sep00,

Aug01, Nov00, Nov01, May01, Sep01, Jul01, and Oct01 in close proximity to each other.

Jun01 and Oct00 were close to these others and April remained an outlier. Stress of the

plot was 0.05018.

Page 70: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Site 29 (Figure 2.16 a,b) The following pairs of months were joined to each other

at a distance of 0.3 and to the other groups at a distance of 0.6: Apr01/Oct00,

May01/Sep00, Jul01/Oct01, Jun01/Aug01. All months were joined at a distance of 1.0.

MDS showed Jul01, Aug01, Oct01, Sep00, Jun01, and May01 to be grouped together.

Apr01 and Oct00 were also grouped. Sep01 and Nov01 were outliers. Stress of the plot

was 0.05109.

Site 35 (Figure 2.17 a,b) Sep00 and May01 were joined in cluster analysis at a

distance of 0.2. Sep01 and Aug01 were joined at a distance of 0.25. All months were

joined at a distance of 0.55. In MDS May01, Oct00, and Sep00 were clustered together

and all others were scattered. Stress was equal to 0.09851.

Site 51 (Figure 2.18 a,b) Jun01, Apr01, and May01 were joined at a distance

of 0.4 in the Cluster Analysis as well as Jul01, Aug01, Sep01, and Oct01. All months

except Oct00 were joined at a distance of 0.6. In the MDS plot Aug01 and Jul01

were close. May01, Apr01, and Jun01 were also clustered close. Stress of the plot

was equal to 0.02730.

Site 67 (Figure 2.19 a,b) Oct01, Sep01, Oct00 and Apr01 were joined at 0.25.

Aug01 and May01 were also joined at 0.25. These two groups were joined at 0.4. All

months were joined at 0.4. All were joined at a distance of 0.75. MDS clustered

Sep01 and Oct01. Stress was equal to 0.03561.

Site 73 (Figure 2.20 a,b) Oct00, May01, and Aug01were joined at a distance

of 0.25 as were Oct01, Sep01, and Jul01. All months were joined at a distance of

0.55. MDS plotted Oct00 close to May01. Stress of the plot was 0.05422.

Page 71: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Discussion

Stomach Content Analysis

Winter Flounder are sight feeders, acquiring prey items during daylight hours

(Olla et al. 1969). They assume an attack posture with their heads lifted off the substrate,

braced by the rays of their dorsal and ventral fins. They move inshore approximately two

hours after low tide and return to the sublittoral zone two hours before the next low tide

(Buckley 1989, Wells 1973). Pearcy (1962) characterizes winter flounder as being

“euryphagus” and Klein-MacPhee (1978) summarizes the diversity of their diet. They

have been documented to feed on 14 different phyla and over 260 species of vertebrates,

invertebrates, and algae. However, juvenile flounder diet is constrained by their small

gape size and predatory ability (Pearcy 1962).

Over 30 species were identified from the core samples. Of these, only 15 were

found in the stomachs of the winter flounder. The cores may be biased in that they

sample the top 10 cm of sediment, whereas flounder may only be able to capture items

within the top 1-2 centimeters of sediment, thus over-estimating potential prey items. An

argument for sampling by this method is that the corer captures infauna that may be

emergent at some point and thus available to the flounder. Stehlik (2000) concurs that

certain infaunal species may be dietary items, but were not in this study (i.e. Glyceridae).

These items were available in the cores but not found in the stomach contents. This may

have been due to the fish’s inability to capture/dig for them, or it may be due simply to

the low sample size used for the stomach content analysis.

When evaluating stomach contents, it is important to consider the digestion rates

of different items within the gut (Shaw 1992). Macdonald et al. (1984) found that often

Page 72: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

items were passed from the stomach of the winter flounder still recognizable. Polychaetes

had the shortest digestion time at 7 hours (MacDonald 1984). Since sampling in this

study occurred at high tide and the fish feed at high tide, it was assumed that fish had

been feeding recently, and digestion had not progressed. Anecdotically, contents were

typically easy to identify to family.

From the stomach content analysis it appears that fish are selecting items to some

degree. They chose fifteen of the items that were potentially available. The percent of

these items that were found in the stomach contents did not match the percent found in

the cores taken from sites where the fish were collected. Still, this stomach content data

indicates that fish are selecting some specific items that may in fact be the preferred items

within the estuary. Of these 15 prey items, Amphipoda, Spionidae, Capitellidae, Mya

arenaria, and Anthuridae were determined to be numerically and volumetrically

important. These concur with similar studies of flounder diet (Armstrong 1997, Stehlik

2000, Shaw 1992, Pearcy 1962). Core data emphasized the fact that these species are

common in the estuary.

An assumption of this study is that fish are feeding at those sites and are relatively

faithful to those sites. It is important to understand the movements of flounder within the

estuary to understand or qualify different types of habitat. Young-of-the-year fish have

been shown to have high fidelity to mud flat areas and are restricted in their movements

by channels (Saucerman & Deegan 1991). In this study of dispersion, fish were tagged

and released on a mud flat. They were often recaptured within 100m of their release site,

with the greatest distance covered being 550m. It is more difficult to discern the

movements and range of older, larger fish. One tagging study showed fish traveling over

Page 73: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

40 km in one season with a dispersion coefficient of 2.8 km2 per day (Phelan 1992), but

the fish that were tagged were over 180 mm in length and not within an estuarine system.

If the assumption of site fidelity holds true, then another point should be made about the

stomach content analysis. Fish were only caught at eight specific sites. If fish were

relatively faithful to these sites, then the contents would be skewed towards those species

found only at those sites. Caution should be made in statements saying these species

represent the major prey items of juvenile flounder. Tentatively, one can say that they

represent the fish found in these habitats and that they may be representative of juvenile

fish in the estuary.

When the length of the fish was considered, fish showed similar diets through

many of the size classes: 50-100, 101-150, 151-200, 201-250 mm. Fish larger than

250mm showed a strong divergence of diet from the smaller fish. It has been shown that

fish smaller than 50 mm feed mainly on copepods (Stehlik 2000). When they increase in

length past this size, their diet starts to shift towards amphipods and smaller polychaetes.

The diets of these mid-sized fish are comparable to other studies. Armstrong (1995)

found that amphipods and the polychaete family Spionidae were important in the diets of

fish in the size categories of 51-100 and 101-150 mm. Amphipods were also important to

fish within the size class of 151-200 mm, but not as important to fish larger than this.

Carlson (1997) concurs that the amphipod, Ampelisca abdicta is an important prey item

for winter flounder, although it is not clear which size classes were used for this analysis.

Stehlik (2000) also agrees that Ampeliscid amphipods are the most consistent prey for

flounder greater than 50 mm but less than 300mm. After examining the changes in diet

Page 74: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

with increased length, she grouped the fish into the following four length classes based

on dietary shifts that she observed: 15-49, 50-299, 300-349, and 350mm and larger.

Fish larger than 100mm started to feed upon Mya siphons, indicating that the

flounders’ jaw had developed the strength to clip through the tough material of the

siphon. In general there was an increase in the number of prey species with an increase in

size. This the case in other studies, as was an increase in composition of prey consumed

with an increase in size, and an increase in the prey size with an increase in fish size

(Shaw 1992, Armstrong 1995, Stehlik 2000,). Forty percent of the diet of the 250+ mm

category was not used for most of the analyses. These other items had been determined to

be insignificant by IRI to all fish sampled, so they were not used in further analyses. This

was not significant to the analyses as the characterization of stomach contents and cores

was for fish within their first few years. It has been estimated that fish reach lengths of

100 to 150mm within the first year (Bigelow &Schroeder 1953, Pereira 1999). After the

second year of growth, fish average about 200 mm in length and reach 250 mm by their

third year (Berry 1965). This large size class probably represents the length where the

fish have greater mobility, a larger gape, and in general, are better predators. All these

conditions permit the fish to consume a more diverse diet.

After examining the change in diet of fish with change in length, the categories

were rearranged for the next two analyses. The fish were grouped into 3 size classes: 50-

100, 101-250, and 250+ mm. Although the 50-100 mm size class and the 101-250 mm

size class showed similar patterns of numerical stomach contents, they represent different

aged fish. The 50-100 mm group represents the young-of-the-year from about late

summer on. Due to small sample size of smaller fish, the young-of-the-year fish were

Page 75: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

poorly represented in this work. It is still important to make this distinction as they are

the fish that would most likely be released in a stocking program. Again, the 250+ mm

size class was separated due to its divergence from the other size categories in prey diet.

The second analysis examined the changes in diet of the fish by season. The three

length classes were used to characterize diet during the fall of 2000 and the spring,

summer, and fall of 2001. Probably the most important information from this data

analysis is that in some cases the importance of amphipods was over-estimated by the

length class analysis. In the 50-100mm size class it was found that amphipods comprised

50% of the total stomach contents during the fall of 2000 and spring of 2001. During the

summer of 2001 and fall of 2001, they became more important, accounting for 98% of

the numerical abundance.

Interestingly, amphipod abundance varied within the diets of 101-250mm fish, but

not in the same manner as the smaller fish. Abundance within the diet was high in the fall

of 2000 (97%) and also in the fall of 2001 (80%). Abundance was lower in the spring

with amphipods comprising only 62% of the diet. During the summer, they only

accounted for 50% of the diet. Amphipods have a late summer/fall peak and as they

increase in benthic abundance it would seem logical to see an increase in stomach content

abundance.

Small polychaetes such as Capitellidae and Arabellidae were also important to the

diets of the small, 50-100 mm fish. This agrees with the assumption that small fish are

limited to a diet of small, non-motile prey items. Within the diets of the 101-250 mm

sized fish, we see an increase in the variety of species and an increase in their size. Most

notably, Mya arenaria siphons make their way into the diet. Whole bivalves such as Mya

Page 76: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

and Tellinidae also begin to appear. Larger Arthropods such as Anthuridae and Crangon

show up as well as larger polychaetes such as Spionidae and Orbiniidae.

Fish within the large size category showed a wide variety of prey items but most

notable was the lack of Amphipods within the diet. This agrees with the earlier analysis.

Due to greater predatory ability, more items are available to the fish so they are not

reliant on the small, easily accessible items.

Another analysis examined the differences in the dietary composition of fish

collected at the eight different sites in Great Bay Estuary. A note should be made about

the robustness of the data set. Several categories were not represented by the stomach

content analyses. Because of this, only general statements will be made about the site

relationships. Amphipods were dominant prey items at Sites 19, 29, and 35, but were not

at Site 23, 51, and 73 for the small (50-100 mm) fish. For the medium sized fish (101-250

mm), Amphipoda were dominant prey items at Sites 19, 25, 35, and 51 but not dominant

at Sites 23, 29, 67, and 73. What is interesting is the shift of sites 29 and 51 between size

classes. The seasonal analysis indicates that there may be a shift in diet between the 50-

100 and 101-250 mm groups or be somewhat proportional between size classes. If the

amphipods were an important diet component for both size classes, then their abundance

within the diet would be similar. Again this analysis was confounded by season and by

small sample size.

There was a noticeable deficiency of amphipods in the diets of fish greater than

250 mm. None were recorded at sites 23, 25, and 51. Only four percent of the diet of 250

mm fish at site 29 was accounted for by amphipods. Again this concurs with the previous

two analyses that showed that the larger fish’s diet departs from that of the smaller fish.

Page 77: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

One assumption when performing these analyses is that food availability limits

growth. If food were never limiting then it would be easy to predict growth rates of wild

fish from the growth rates of fish reared in the lab at the same temperature (Miller 1991).

Growth is not controlled just by temperature and food availability. Energy is taken from

growth for other activities such as predator avoidance, prey searching, and

osmoregulation. Prey availability may be an important limiting factor, but it is most

likely confounded by these other factors. Fish may be reducing prey populations enough

that they may be limiting their own growth (Shaw 1992). Van der Veer (1993) found that

growth was different between stations of his study and it was positively related with food

abundance. This indicates that in some areas, “growth was not maximal and depends on

food abundance and food composition” (van der Veer 1993). Similarly, it was found

through caging experiments that growth was significantly greater at a site intermediate

within Great Bay Estuary than one located near the mouth, suggesting that areas of the

estuary may be food limiting (Fairchild 2002).

Algae was found in the stomachs of the flounder, although it was not quantified in

this study because of the difficulty in relating stomach content quantities to abundances

found at each site. It has been suggested that algae, in some amount, is important to the

diet of the winter flounder (Keats 1990, Kennedy 1971, Wells 1973). It has also been

suggested that flounder are consuming macroalgae to procure epiphytic organisms such

as copepods and amphipods (Sogard 1992). Amphipod communities have been shown to

be enhanced by the presence of Ulva lactuca and Zostera marina anywhere from 18-

99% (Sogard 1992). Vegetation is an additional element to consider when characterizing

prey habitat. An important point to make is that Ulva lactuca would only provide an

Page 78: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

ephemeral substrate while Zostera marina would be a more stable one. Not only would

the two be difficult to quantify in a specific area, but also to characterize their abundance

and their epiphytic assemblage over time. For these reasons SAV was not collected for

assessment

Core Data Analysis

When the average number of prey taxa were plotted for each site, it is not

surprising to see that each site is similar in composition. The species that were found in

the flounders’ stomachs are ubiquitous and abundant throughout the estuary. What is of

interest is whether a site or habitat maintains a consistent quality and quantity of prey

species. The eleven most important species, determined by stomach content analysis,

were used as a measure of the quality of the habitats chosen for the study. One note to

make is that the characterization of the sites was performed using only these eleven

species. Several others were identified and may have in fact been indicator species or

diagnostic species (Brown 1999). For example, the common blue mussel, Mytilus edulis,

was found at some of the sites. In this study, mussels were not used as an indicator of

good flounder habitat because they were not found in the stomach contents. In this sense,

traditional habitat distinctions were abandoned to define habitat important to flounder. By

reducing the number of species, variability between sites and within the overall data set

may have also been reduced. The analyses have been biased in this manner. The analyses

are useful only in describing the differences in habitat quality for winter flounder based

upon prey abundance in these limited areas.

Page 79: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Although benthic invertebrates are fairly non-motile they do have some

population dynamics and it is important to consider the dynamics of any species over

time. In the second analysis, the average number of each core taxa was plotted against

time. For most of the sites there was a peak in the late summer into the fall. At this time

of the year temperatures are warm and productivity is high. Another important point to

make from this analysis is that amphipods are absent from sites 67 and 73. As the most

important prey item for fish smaller than 250 mm, it is interesting that fish are still found

at these sites and apparently do well foraging on other items.

To emphasize the seasonality of the macrobenthic invertebrates, each species was

plotted by month. Again most of these species had peaks in the late summer into the fall:

Amphipodae, Anthuridae, Myidae, Tellinidae, Arabellidae, Spionidae, Capitelidae, and

Orbiniidae. Amphipods were also abundant in April and July. Spionidae were abundant

in May also. These two may have been the most dominant prey items because they were

consistently abundant in the estuary through the year.

For the Cluster Analysis and Multidimensional Scaling (MDS), the data were

root-root transformed. This reduces the variability of each object in the analysis so when

objects are compared within a data set, one object does not have more “weight” than

others. This is to ensure that one variable, or species, is not driving the distinction

between groups. The data were entered into the Bray-Curtis Equation, which calculates

the dissimilarity between data sets. Ideally, if sites were chosen to show different types of

habitats within the estuary, then most of the sites would be highly dissimilar in species

composition, with a value close to one on a scale of zero to one. Both Cluster Analysis

and Multidimensional Scaling are ways of visualizing these dissimilarities. Cluster

Page 80: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Analysis uses the values from the dissimilarity matrix to produce a dendrogram.

Multidimensional Scaling also uses the dissimilarity matrix to produce a plot that depicts

the groups in space.

Cluster Analysis and MDS are both useful in clustering data to show patterns. The

output can be interpreted in several ways. The purpose of these analyses was to evaluate

two things: the similarity of species composition of a site over time and the similarities

between sites. When evaluating the consistency of a site, a threshold should be

determined. In this case the arbitrary threshold of 75% similarity was considered good.

An argument could be made for a more or less conservative threshold. Consistency, or

similarity, can be used as a measure of stability. More stable communities would be of

more value than sites that fluctuate. After determining the stability of a site over time, the

final step in this analysis is to return to the core data to confirm if the items are indeed

quality prey species.

Similarity between sites can also be measured with these analyses. In this study,

sites were chosen to represent different habitats. In theory, the sites represent different

substrates, temperatures, and salinity regimes. These environmental variables should

support different benthic communities.

Sites were compared using two-month blocks. This was to decrease the seasonal

variability to a minimum. It was not surprising to see sites grouped together that are close

to each other spatially. Sites 73 and 67 were grouped consistently, as were 19 and 25.

What was interesting was that sites 23 and 51 were joined together from April to

September of 2001. They are distant from each other in space and represent very different

salinity and temperature regimes. They may have similar substrates though which may

Page 81: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

strengthen the association (see Chapter 2). These data might be important in

understanding recruitment of these species. Most likely these sites were grouped by

similar salinity regimes, temperatures, and substrates. They might also be subject to other

variables such as eutrophication, changing dissolved oxygen levels, or localized

anthropogenic disturbances. Understanding these processes will help in understanding the

variation in prey community structure.

When each site was compared over time, one general pattern emerged: species

abundance was similar by season. Sites 67 and 73 had the highest level of similarity

between months. Sites 19, 25,29,35 and 51 also had similar constructs although they

varied more than the previous two. Site 23 had the lowest similarity over time. One other

trend was that the month of April was the outlier for most of the sites.

Seasonality is an important consideration when characterizing the benthos of the

estuary. Spring and fall peaks in abundance have been “attributed to two major factors:

reproductive peaks of dominant taxa coincidental with optimal temperatures, and pulses

of food inputs in the form of algal blooms and detritus carried by runoff (Grizzle 1999).”

Many of these numerically dominant taxa are so because they are opportunistic species.

These species are characterized by “the rapid invasion of disturbed areas, being

reproductive year round, brooding young, and releasing juveniles directly (Grizzle

1999).” Carlson (1997) suggests that these opportunistic species may “represent an

enhance forage base for some benthic feeding fishes.” It has been found that an

anthropogenically disturbed area provided higher densities of amphipods than an

adjacent, undisturbed site. Fish caught on both substrates had amphipods as a dominant

prey taxon, indicating that flounder were selecting for amphipods, but fish from the

Page 82: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

disturbed site ate more amphipods than those on the undisturbed substrate (Carlson

1997). The channel adjacent to Site 23 was dredged during the winter of 2000. Mooring

blocks were dragged across the site during the dredging process. This partially explains

the sharp change in the benthic assemblage between the fall of 2000 and the spring of

2001at this site.

Summer declines in the benthic community have been attributed to fish predation

(Grizzle 1999, Armstrong 1997). This may be the case, but from this study, and similar

ones (Fairchild 2002, Armstrong 1995) fish densities were determined to be quite low

and it is unlikely that grazing fish are responsible for reducing benthic abundances so

dramatically.

Understanding the ecology of macrobenthic assemblages is essential to

understanding the ecology of demersal, benthic-feeding fishes. In this work, it has been

shown that different areas do support different assemblages of benthic fauna and that the

abundances of these organisms change over time. Young flounder are selecting certain

species to prey upon, most notably amphipods. When faced with an environment that

does not support these important prey species, the flounder do prey upon other items. In

the context of choosing a site for stock enhancement, surveying the area for potential

prey quality and quantity would be one logical step. Although food availability has been

considered a limiting factor of growth, it is impossible to ignore the other limiting factor

of growth: temperature. It is also impossible to ignore constraining factors such as

salinity, dissolved oxygen, current, substrate type, predators, and competitive species.

These factors must be considered before statements can be made about the quality of each

site as flounder habitat.

Page 83: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Table 2.1 Index of Relative Importance based on the stomach contents of winter floundercaught in Great Bay Estuary.

Table 2.2 Taxa Identified from benthic core samples

AnnelidaC. Polychaeta

F. ArabellidaeF. CapitellidaeF. CirratulidaeF. FlabelligeridaeF. GlyceridaeF. MaldanidaeF. NephtyidaeF. SpionidaeF. OpheliidaeF. OrbiniidaeF. PhyllodocidaeF. SigalionidaeF. Terebellidae

C. OligochaetaArthropoda

C. CrustaceaO. Amphipoda

Amphipoda Anthuridae Crangon Carcinus Myidae Myidae (siphons)Number 1717 61 3 3 26 72Wetwt (g) 9.214 2.257 2.32 1.021 1.403 4.524Frequency 51 25 2 3 9 16%Number 81.1 2.88 0.14 0.14 1.23 3.4%Wetwt 12.54 3.071 3.16 1.39 1.91 6.16% Frequency 41.46 20.33 1.63 2.44 7.32 13.01I.R.I. 3883 121 5 4 23 124

Tellinidae Arabellidae Spionidae Capitellidae Orbidinidae OtherNumber 1 20 84 114 10 6Wetwt (g) 0.034 0.328 48.0633 1.564 0.208 1.185 73.497Frequency 1 5 41 16 7 5%Number 0.05 0.94 3.97 5.39 0.47 0.28%Wetwt 0.05 0.45 65.39 2.13 0.28 1.61% Frequency 0.81 4.07 33.33 13.01 5.69 4.07I.R.I. 0 6 2312 98 4 8

Page 84: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

F. AmpeliscidaeF. Gammaridae

O. IsopodaF. Anthuridae

C. MalacostracaO. MysidaeO. Decapoda

F. CrangonidaeF. Pandalidae

MolluscaC. Bivalva

F. BivalvaF. MesodesmatidaeF. MyidaeF. MytilidaeF. TellinidaeF. Veneidae

C. GastropodaF. LittorinidaeF. Nassariidae

NemerteaHemichordata

Page 85: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.2 Percent Numerical Abundance of prey items found in the stomach contentsfrom five size classes of fish: 50-100, 101-150, 151-200, 201-250, and 250mm andgreater.

Page 86: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

50-1

00

101-

150

151-

200

201-

250

250+

Fish Length (mm)

Perc

ent N

umer

ical

Abu

ndan

ce

Other numberOrbiniidae numberCapitellidae numberSpionidae numberArabellidae numberTellinidae numberMya Siphons numberMyidae numberCarcinus maenas numberCrangon numberAnthuridae numberAmphipoda number

Figure 2.3 Stomach Content Analysis for each season sampled. Fish were divided intosize categories when sample sizes permitted. These three size classes were: 50-100mm,101-250mm, and 250mm and greater.

Page 87: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

0%

20%

40%

60%

80%

100%

50-1

00

101-

250

250+

50-1

00

101-

250

250+

50-1

00

101-

250

250+

50-1

00

101-

250

Fall2000

Fall2000

Fall2000

Spr2001

Spr2001

Spr2001

Sum2001

Sum2001

Sum2001

Fall2001

Fall2001

OtherOrbiniidaeCapitellidaeSpionidaeArabellidaeTellinidaeMya SiphonsMyidaeCarcinus maenasCrangonAnthuridaeAmphipoda

Figure 2.4 Stomach Contents of fish within the following size classes: 50-100, 101-250,and 250+ mm compared between sites.

Page 88: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

0%

20%

40%

60%

80%

100%

101-

250

50-1

00

101-

250

250+

50-1

00

101-

250

250+

50-1

00

101-

250

250+

50-1

00

101-

250

50-1

00

101-

250

250+

101-

250

50-1

00

101-

250

19 23 23 23 25 25 25 29 29 29 35 35 51 51 51 67 73 73

Size Classes of Fish within Sites

Perc

ent N

umer

ical

Abu

ndan

ce

OtherOrbiniidae

CapitellidaeSpionidae

ArabellidaeTellinidae

Mya SiphonsMyidae

Carcinus maenasCrangon

AnthuridaeAmphipoda

Figure 2.5 Average numerical abundance of constituents of each site.

Page 89: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

19 23 25 29 35 51 67 73

Site

Ave

rage

Num

eric

al A

bund

ance

Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Figure 2.6 Average numerical abundance of species found within each core during eachmonth at each site. Benthos not sampled from November 2000 to March 2001.

Page 90: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Site 19

0

1

2

3

4

5

6

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Month

Ave

rage

num

ber

of e

ach

taxa

per

co

re

Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Site 25

0

1

2

3

4

5

6

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Month

Ave

rage

num

ber

of p

rey

taxa

per

cor

e Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Page 91: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Site 23

00.5

11.5

22.5

33.5

44.5

5

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Month

Ave

rage

num

ber

of p

rey

taxa

per

cor

eOrbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Site 29

00.5

11.5

22.5

33.5

44.5

5

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Month

Ave

rage

num

ber

of p

rey

taxa

per

cor

e

Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Page 92: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Site 35

02468

1012141618

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Month

Ave

rage

num

ber

of p

rey

taxa

per

co

re

Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Site 51

0

5

10

15

20

25

Sep-0

0

Oct-00

Nov-0

0

Dec-0

0

Jan-

01

Feb-0

1

Mar-

01

Apr-0

1

May

-01

Jun-

01Ju

l-01

Aug-0

1

Sep-0

1

Oct-01

Month

Ave

rage

num

ber

of p

rey

taxa

per

cor

e Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Page 93: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Site 67

0

1

2

3

4

5

6

7

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Month

Ave

rage

num

ber

of p

rey

taxa

per

cor

e Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Site 73

0

2

4

6

8

10

12

14

16

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Month

Ave

rage

num

ber

of p

rey

taxa

per

cor

e Orbinidae

Capitellidae

Spionidae

Arabellidae

Tellinidae

Large Myiidae

Myidae

Carcinus maenas

Crangon

Anthuridae

Amphipoda

Page 94: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.7 (A. – I.) Abundance of the nine families found in core samples taken fromGreat Bay Estuary. Abundance is equal to the average number found in one core(0.0085m2). Benthos not sampled from November 2000 to March 2001.

Figure 2.7.A Amphipodae

0

0.1

0.2

0.3

0.4

0.5

0.6

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01A

vera

ge N

umbe

r pe

r C

ore

Figure 2.7.B Anthuridae

00.20.40.60.8

1

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01Ave

rage

Num

ber

per

Cor

e

Page 95: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.7.C Myidae

0

0.05

0.1

0.15

0.2Se

p-00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Ave

rage

Num

ber

per

Cor

e

Figure 2.7.D Large Myidae

00.020.040.060.080.1

0.120.140.160.180.2

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Ave

rage

Num

ber

per

Cor

e

Figure 2.7.E Tellinidae

Page 96: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

00.20.40.60.8

11.21.41.6

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01A

vera

ge N

umbe

r pe

r C

ore

Figure 2.7.F Arabellidae

0

0.5

1

1.5

2

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01A

vera

ge N

umbe

r pe

r C

ore

Figure 2.7.G Spionidae

00.20.40.60.8

11.21.41.6

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01A

vera

ge N

umbe

r pe

r C

ore

Figure 2.7.H Capitellidae

Page 97: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

012345

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01

Ave

rage

Num

ber

per

Cor

e

Figure 2.7.I Orbinidae

0

0.1

0.2

0.3

0.4

0.5

0.6

Sep-

00

Oct

-00

Nov

-00

Dec

-00

Jan-

01

Feb-

01

Mar

-01

Apr

-01

May

-01

Jun-

01

Jul-

01

Aug

-01

Sep-

01

Oct

-01

Nov

-01A

vera

ge N

umbe

r pe

r C

ore

Figure 2.8 Cluster Analysis (A.) and Multidimensional Scaling (B.) of the differences inthe benthic community between sites during October and November of 2000. Dataanalyzed using Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------SITE73 SITE67 0.255 2SITE25 SITE19 0.273 2SITE29 SITE25 0.422 3SITE29 SITE23 0.474 4

Page 98: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

SITE29 SITE35 0.480 5SITE29 SITE73 0.481 7SITE51 SITE29 0.652 8

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matrices

Page 99: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Minimizing Kruskal STRESS (form 1) in 2 dimensions

Iteration STRESS--------- ------ 0 0.106366 1 0.066021 2 0.053169 3 0.050448 4 0.049438Stress of final configuration is: 0.04944Proportion of variance (RSQ) is: 0.98815

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SITE19 .55 .10 SITE23 .54 1.00 SITE25 .54 .38 SITE29 .10 .07 SITE35 -.61 .06 SITE51 -1.92 .08 SITE67 .62 -.73 SITE73 .18 -.96

Page 100: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.9 Cluster Analysis (A.) and Multidimensional Scaling (B.) of differences in thebenthic communities found at sites within the bay during April and May of 2001. Dataanalyzed using Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------SITE29 SITE25 0.171 2SITE73 SITE67 0.188 2SITE29 SITE73 0.190 4SITE29 SITE19 0.208 5SITE35 SITE29 0.320 6SITE35 SITE51 0.381 7SITE23 SITE35 0.458 8

Page 101: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic Multidimensional ScalingThe data are analyzed as similaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2 dimensions

Iteration STRESS--------- ------ 0 0.273173 1 0.218781 2 0.201349 3 0.192594 4 0.186429 5 0.182131 6 0.179094 7 0.176548 8 0.174321 9 0.172324 10 0.170445 11 0.168623 12 0.166831 13 0.165082 14 0.163327 15 0.161505 16 0.159450 17 0.157051 18 0.154382 19 0.151418 20 0.148293 21 0.145226 22 0.142424 23 0.140071 24 0.138252 25 0.136940 26 0.136043 27 0.135450 28 0.135067 29 0.134820Stress of final configuration is: 0.13482Proportion of variance (RSQ) is:0.85756

Coordinates in 2 dimensions

Page 102: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Variable Dimension-------- --------- 1 2 SITE19 .96 -.73 SITE23 .00 .20 SITE25 .03 .89 SITE29 -1.31 .45 SITE35 .09 -.78 SITE51 -.21 .16 SITE67 -.77 -.82 SITE73 1.21 .61

Figure 2.10 Cluster Analysis (A.) and Multidimensional Scaling (B.) of differences in thebenthic community found at various sites in the bay during June and July of 2001. Dataanalyzed using Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------SITE51 SITE23 0.101 2SITE73 SITE25 0.290 2SITE35 SITE19 0.291 2SITE35 SITE73 0.308 4SITE35 SITE29 0.368 5SITE51 SITE35 0.394 7SITE67 SITE51 0.495 8

Page 103: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic MultidimensionalScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2dimensions

Iteration STRESS--------- ------ 0 0.210400 1 0.181977 2 0.169191

Page 104: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

3 0.159823 4 0.152146 5 0.145549 6 0.138733 7 0.129328 8 0.117213 9 0.104877 10 0.095407 11 0.090075 12 0.087033 13 0.085195 14 0.083931 15 0.082987 16 0.082261Stress of final configuration is: 0.08226Proportion of variance (RSQ) is:0.95001

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SITE19 1.20 .35 SITE23 -.92 .34 SITE25 1.03 -.69 SITE29 .25 .80 SITE35 .27 .12 SITE51 -.93 .42 SITE67 -1.08 -.80 SITE73 .19 -.54

Page 105: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.11 Cluster Analysis (A.) and Multidimensional Scaling (B.) of differences in thebenthic communities found at various sites in the bay during August and September of2001.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------SITE51 SITE23 0.129 2SITE73 SITE67 0.130 2SITE19 SITE51 0.207 3SITE73 SITE19 0.218 5SITE73 SITE35 0.232 6SITE25 SITE73 0.253 7SITE29 SITE25 0.254 8

Page 106: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2dimensions

Iteration STRESS--------- ------ 0 0.075013 1 0.060642 2 0.055926 3 0.053511 4 0.051914 5 0.050663 6 0.049612Stress of final configuration is: 0.04961Proportion of variance (RSQ) is: 0.98578

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SITE19 .54 .34 SITE23 .06 -.58 SITE25 -.96 1.10 SITE29 -.55 -1.18 SITE35 1.57 .35 SITE51 .18 -.53 SITE67 -.48 .15 SITE73 -.35 .36

Page 107: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.12 October – November 2001: Cluster Analysis and Multidimensional Scaling

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------SITE67 SITE25 0.163 2SITE67 SITE19 0.222 3SITE67 SITE73 0.223 4SITE51 SITE35 0.298 2SITE67 SITE51 0.316 6SITE29 SITE23 0.372 2SITE67 SITE29 0.373 8

Page 108: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1)in 2 dimensions

Iteration STRESS--------- ------ 0 0.121170 1 0.098482 2 0.092063 3 0.088971 4 0.086995 5 0.085634 6 0.084618Stress of final configuration is: 0.08462Proportion of variance (RSQ) is:0.95699

C o o r d i n a t e s i n 2dimensionsVariable Dimension-------- --------- 1 2 SITE19 .70 .38 SITE23 -1.03 .32 SITE25 .60 -.15 SITE29 -.79 1.07 SITE35 -.56 -.51 SITE51 -.80 -.98 SITE67 .84 -.22 SITE73 1.05 .10

Page 109: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.13 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 19. Data analyzed using Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------NOV01 SEP01 0.339 2AUG01 JUL01 0.355 2AUG01 NOV01 0.363 4AUG01 OCT01 0.372 5MAY01 AUG01 0.418 6MAY01 SEP00 0.482 7JUN01 MAY01 0.487 8JUN01 OCT00 0.500 9JUN01 APR01 1.000 10

Page 110: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2 dimensions

Iteration STRESS--------- ------ 0 0.180943 1 0.167239 2 0.161684 3 0.159203 4 0.157669 5 0.156207 6 0.154299 7 0.151572 8 0.147968 9 0.143654 10 0.139174 11 0.135110 12 0.131712 13 0.128844 14 0.126305 15 0.123970 16 0.121757 17 0.119604 18 0.117470 19 0.115411 20 0.113326 21 0.111170 22 0.108947 23 0.106682 24 0.104418 25 0.102201 26 0.100072 27 0.098064 28 0.096199 29 0.094435 30 0.092755 31 0.090941 32 0.088858 33 0.086465 34 0.083755 35 0.080774 36 0.077612

Page 111: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

37 0.074382 38 0.071205 39 0.068189 40 0.065364 41 0.062757 42 0.060395 43 0.058268 44 0.056367 45 0.054670 46 0.053163 47 0.051826 48 0.050661Stress of final configuration is: 0.05066Proportion of variance (RSQ) is: 0.98387

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SEP00 -.83 .71 OCT00 1.04 .89 APR01 -1.66 .02 MAY01 1.14 -.14 JUN01 -.70 -.44 JUL01 .31 -1.03 AUG01 .29 -.74 SEP01 .17 .00 OCT01 .13 .53 NOV01 .11 .20

Figure 2.14 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 23. Data analyzed in Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------AUG01 JUL01 0.320 2SEP01 AUG01 0.402 3

Page 112: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

SEP01 OCT01 0.460 4OCT00 SEP00 0.546 2SEP01 MAY01 0.585 5NOV01 SEP01 0.622 6NOV01 OCT00 0.684 8NOV01 JUN01 0.697 9NOV01 APR01 1.000 10

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2 dimensions

Page 113: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Iteration STRESS--------- ------ 0 0.164069 1 0.137133 2 0.119708 3 0.107850 4 0.099297 5 0.092705 6 0.087160 7 0.082171 8 0.077542 9 0.073184 10 0.069051 11 0.065137 12 0.061635 13 0.058598 14 0.055942 15 0.053712 16 0.051917 17 0.050437Stress of final configuration is: 0.05044Proportion of variance (RSQ) is: 0.98063

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SEP00 -.91 .96 OCT00 -.97 .35 APR01 -.99 -1.07 MAY01 .83 -.82 JUN01 -.06 -.92 JUL01 .50 -.06 AUG01 .07 .65 SEP01 .94 .17 OCT01 .75 .79 NOV01 -.16 -.05

Page 114: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Figure 2.15 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 25. Data analyzed using Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------NOV01 OCT01 0.136 2AUG01 SEP00 0.227 2AUG01 NOV01 0.261 4SEP01 MAY01 0.305 2JUL01 SEP01 0.317 3JUN01 AUG01 0.343 5JUL01 OCT00 0.348 4JUL01 JUN01 0.398 9JUL01 APR01 1.000 10

Page 115: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2 dimensions

Iteration STRESS--------- ------ 0 0.158195 1 0.126621 2 0.111438 3 0.105231 4 0.100318 5 0.095387 6 0.090127 7 0.084396 8 0.078193 9 0.071738 10 0.065543 11 0.060189 12 0.056144 13 0.053431 14 0.051566 15 0.050180Stress of final configuration is: 0.05018Proportion of variance (RSQ) is:0.99485

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2

Page 116: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

SEP00 .28 .58 OCT00 -.01 -.56 APR01 -2.76 .03 MAY01 .43 -.13 JUN01 -.03 .01 JUL01 .38 -.48 AUG01 .28 .31 SEP01 .59 -.23 OCT01 .42 .28 NOV01 .42 .20

Figure 2.16 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 29. Data analyzed using Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------APR01 OCT00 0.210 2MAY01 SEP00 0.254 2OCT01 JUL01 0.306 2AUG01 JUN01 0.318 2OCT01 MAY01 0.322 4OCT01 AUG01 0.353 6OCT01 APR01 0.583 8OCT01 NOV01 0.674 9OCT01 SEP01 1.000 10

Page 117: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in2 dimensions

Iteration STRESS--------- ------ 0 0.127796 1 0.106094 2 0.093739

Page 118: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

3 0.085274 4 0.078417 5 0.072310 6 0.066761 7 0.061822 8 0.057860 9 0.054890 10 0.052706 11 0.051086Stress of final configuration is: 0.05109Proportion of variance (RSQ) is: 0.98693

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SEP00 .19 -.13 OCT00 -.92 -.81 APR01 -.97 -.79 MAY01 .28 -.52 JUN01 .77 -.34 JUL01 .75 .42 AUG01 .46 .01 SEP01 -1.31 .98 OCT01 .90 .04 NOV01 -.14 1.16

Figure 2.17 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 35. Data analyzed using SYSTAT“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------

Page 119: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

MAY01 SEP00 0.185 2SEP01 AUG01 0.223 2MAY01 JUL01 0.319 3SEP01 OCT01 0.321 3MAY01 OCT00 0.333 4SEP01 NOV01 0.382 4MAY01 SEP01 0.428 8JUN01 MAY01 0.482 9APR01 JUN01 0.544 10

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilarities

Page 120: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Fitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2 dimensions

Iteration STRESS--------- ------ 0 0.146471 1 0.122886 2 0.110787 3 0.105184 4 0.102295 5 0.100616 6 0.099626 7 0.098969 8 0.098508Stress of final configuration is: 0.09851Proportion of variance (RSQ) is: 0.93110

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SEP00 -.59 -.33 OCT00 -1.22 -.29 APR01 1.12 -.86 MAY01 -.74 .02 JUN01 .12 -1.03 JUL01 -.97 .58 AUG01 .98 .60 SEP01 .80 .14 OCT01 .01 .31 NOV01 .48 .87

Figure 2.18 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 51. Data analyzed using Systat“ 10.

Page 121: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------MAY01 APR01 0.000 2JUN01 MAY01 0.333 3SEP01 AUG01 0.374 2SEP01 JUL01 0.376 3SEP01 OCT01 0.389 4SEP00 SEP01 0.476 5SEP00 JUN01 0.573 8OCT00 SEP00 1.000 9

Page 122: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2dimensions

Iteration STRESS--------- ------ 0 0.100555 1 0.071853 2 0.059082 3 0.052460 4 0.047843 5 0.044242 6 0.041327 7 0.038794 8 0.036503 9 0.034405 10 0.032467 11 0.030634 12 0.028915 13 0.027303Stress of final configuration is: 0.02730Proportion of variance (RSQ) is: 0.99659

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SEP00 -1.14 .61 OCT00 -1.31 -1.26 APR01 .73 -.45 MAY01 .73 -.45 JUN01 .67 -.65 JUL01 .48 .36 AUG01 .03 .44 SEP01 -.27 .50

Page 123: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

OCT01 .08 .91

Figure 2.19 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 67. Data analyzed in Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------OCT01 SEP01 0.163 2AUG01 MAY01 0.220 2OCT01 OCT00 0.270 3OCT01 APR01 0.272 4SEP00 OCT01 0.306 5SEP00 AUG01 0.398 7JUN01 SEP00 0.566 8JUL01 JUN01 0.746 9

Page 124: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2dimensions

Iteration STRESS--------- ------ 0 0.187505 1 0.148240 2 0.136375 3 0.117919 4 0.092523 5 0.072655

Page 125: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

6 0.060967 7 0.053326 8 0.048045 9 0.044203 10 0.041282 11 0.038985 12 0.037138 13 0.035614Stress of final configuration is: 0.03561Proportion of variance (RSQ) is: 0.99335

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SEP00 .12 .83 OCT00 .24 .21 APR01 .64 -.22 MAY01 .05 -1.26 JUN01 1.12 .78 JUL01 -1.84 .33 AUG01 .31 -.68 SEP01 -.33 .12 OCT01 -.31 -.11

Figure 2.20 Cluster Analysis (A.) and Multidimensional Scaling (B.) of changes in thebenthic community found at Site 73. Data analyzed using Systat“ 10.

A. Cluster AnalysisSingle linkage method (nearest neighbor)

Cluster and Cluster Were joined No. of memberscontaining containing at distance in new cluster------------ ------------ ------------ --------------MAY01 OCT00 0.077 2SEP01 JUL01 0.202 2MAY01 AUG01 0.244 3

Page 126: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

SEP01 OCT01 0.244 3JUN01 APR01 0.273 2MAY01 SEP01 0.301 6JUN01 MAY01 0.372 8JUN01 SEP00 0.546 9

B. Monotonic Multidimensional ScalingThe data are analyzed as dissimilaritiesFitting is split between data matricesMinimizing Kruskal STRESS (form 1) in 2dimensions

Iteration STRESS--------- ------ 0 0.086514 1 0.071397 2 0.062714 3 0.058215

Page 127: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

4 0.055666 5 0.054223Stress of final configuration is: 0.05422Proportion of variance (RSQ) is: 0.98362

Coordinates in 2 dimensionsVariable Dimension-------- --------- 1 2 SEP00 -1.74 -.58 OCT00 .29 .74 APR01 -.35 .10 MAY01 .29 .73 JUN01 -.77 .73 JUL01 .94 -.79 AUG01 .67 .19 SEP01 .64 -.32 OCT01 .04 -.79

LITERATURE CITED

Systat 10 Manual. 2000. SPSS inc.

US Fish and Wildlife Service: Multivariate Statistical Analyses. 2001.

Armstrong, M.P. 1995. A comparative study of the ecology of smooth flounder,Pleuronectes putnami, and winter flounder, Pleuronectes americanus, from GreatBay Estuary, New Hampshire. Ph.D. Dissertation. University of New Hampshire.147p.

Armstrong, M. P. 1997. Seasonal and ontogenetic changes in distribution and abundanceof smooth flounder, Pleuronectes putnami, and winter flounder, Pleuronectesamericanus, along estuarine depth and salinity gradients. Fish. Bull. 95, 414-430.

Bigelow & Schroeder. 1953. Fishes of the Gulf of Maine. U.S. Fish & Wildl. Serv. Fish.Bull. 53: 577p.

Brown, B.A. 1993. Classification System of Marine and Estuarine Habitats in Maine: An

Page 128: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Ecosystem Approach to Habitats. 51p.

Buckley, J. 1989. Species Profiles: Life histories and environmental requirements ofcoastal fishes and invertebrates (North Atlantic) - winter flounder. U.S. Fish andWildlife Service Biological Report 82(11.87).

Carlson, J. K., Randall, T.A., and M. E. Mroczka. 1997. Feeding Habits of WinterFlounder (Pleuronectes americanus) in a Habitat Exposed to AnthropogenicDisturbance. J. Northwest At. Fish. Sci. 21: 65-73.

Choat, J. H. 1982. Fish Feeding and the Structure of Benthic Communities in TemperateWaters. Ann. Rev. Ecol. Syst. 13: 423-49.

Fairchild, E.A. 2002. Winter Flounder Pseudopleuronectes americanus stockenhancement in New Hampshire: developing optimal release strategies. Ph.D.Dissertation, University of New Hampshire. 142p.

Field, J. G., Clarke, K.R., and R. M. Warwick. 1982. A Practical Strategy for AnalysingMultispecies Distribution Patterns. Marine Ecology Progress Series 8: 37-52.

Grizzle, R. E. 1999. Seasonality of Macrofaunal Benthos in New England Estuaries. InReview.

Hyslop, E. J. 1980. Stomach Content Analysis-a review of methods and their application.J. Fish Biol. 17: 411-429.

Keats, D. W. 1990. Food of winter flounder Pseudopleuronectes americanus in a seaurchin dominated community in eastern Newfoundland. Mar. Ecol. Prog. Ser. 60:13-22.

Kennedy, V. S. and D. H. Steele. 1971. The Winter Flounder (Pseudopleuronectesamericanus) in Long Pond, Conception Bay, Newfoundland. J. Fish. Res. Bd.Canada 28: 1153-1165.

Klein-MacPhee, G. 1978. Synopsis of Biological Data for the Winter Flounder,Pseudopleuronectes americanus (Walbaum). NOAA Technical Repost NMFSCircular 414: 1-43.

Macdonald et al. 1984. Fishes, Fish Assemblages, and their Seasonal Movements in theLower Bay of Fundy and Passamaquoddy Bay, Canada. Fishery Bulletin 82(1)121-138.

Martell, D. J. and G. McClelland. 1994. Diets of sympatric flatfishes, Hippoglossoidesplatessoides, Pleuronectes ferrugineus, Pleuronectes americanus, from SableIsland Bank, Canada. Journal of Fish Biology 44: 821-848.

Page 129: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Miller, J. M., Burke, J. S., and G. R. Fitzhugh. 1991. Early Life History Patterns

of Atlantic North American Flatfish: Likely (and Unlikely) Factors Controlling

Recruitment. Neth. J. Sea Res. 27(3/4): 261-275.

Olla et al. 1969. Behavior of Winter Flounder in a Natural Habitat. Trans. Amer. Fish.Soc. 4: 717-720.

Pearcy, W. G. 1962. Ecology of an Estuarine Population of Winter Flounder,Pseudopleuronectes americanus (Walbaum). Bulletin of The BinghamOceanographic Collection 18(1).

Pereira, J. J., Goldberg, R., Ziskowski, J. J., Berrien, P. L., Morse, W. W., and D. L.Johnson. 1999. Winter Flounder, Pseudopleuronectes americanus, Life Historyand Habitat Characteristics. NOAA Technical Memorandum NMFS-NE-138: 1-39.

Peterson, C. H. et al. 2000 Synthesis of Linkages between Benthic and Fish Communitiesas a Key to Protecting Essential Fish Habitat. Bulletin of Marine Science 66(3):759-774.

Phelan, B. A. 1992. Winter Flounder Movements in the Inner New York Bight. Trans.Am. Fish. Soc. 121: 777-784.

Randall, D.J. (ed.) Eckert Animal Physiology: Mechanisms and Adaptations. W.H.Freeman & Company. New York. 1998. pp 627-629.

Saucerman, S. E. and L. A. Deegan. 1991. Lateral and Cross-Channel Movement ofYoung-of-the-Year Winter Flounder (Pseudopleuronectes americanus) inWaquoit Bay, Massachusetts. Estuaries 14(4): 440-446.

Shaw, M. and G. P. Jenkins. 1992. Spatial variation in feeding, prey distribution and foodlimitation of juvenile flounder Rhombosolea tapirina Gunther. J. Exp. Mar. Biol.Ecol. 165: 1-21.

Sogard, S. M. 1992. Variability in growth rates of juvenile fishes in different estuarinehabitats. Mar. Ecol. Prog. Ser. 85: 35-53.

Stehlik, L. L. and C. J. Meise 2000. Diet of Winter Flounder in a New Jersey Estuary:Ontogenetic Change and Spatial Variation. Estuaries 23(3): 381-391.

van der Veer, H. W. and Johannes I. J. Witte. 1993 The 'Maximum growth/optimal foodcondition' hypothesis: a test for 0-group plaice Pleuronectes platessa in the DutchWadden Sea. Mar. Ecol. Prog. Ser. 101: 81-90.

Wells, B., Steele, D.H. and A. V. Tyler. 1973. Intertidal Feeding of Winter Flonders(Pseudopleuronectes americanus) in the Bay of Fundy. J. Fish. Res. Board Can.

Page 130: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

30(9): 1374-1378.

SYNOPSIS

Patterns in Distribution

It has been suggested that “ecological patterns in species distribution and

abundance are linked to habitat characteristics, dispersal mechanisms, colonizing

abilities, gene flow, and genetic structure (Bailey 1997).” In this study, only habitat

characteristics were investigated. Different levels of scale should be considered when

relating an organism to its environment (Menge 1990). Variation in stock size and

distribution is a function of global-scale processes such as El Nino events, meso-scale

processes such as oceanic currents, and small-scale events such as changes in dissolved

oxygen concentrations. Understanding small-scale changes in the context of large-scale

changes is an important step in describing the systems responsible for shaping fish stocks

(Menge 1990).

As mentioned in the first chapter, there is little known about the long-term

changes in the abundance and distribution of winter flounder in Great Bay, N.H. There is

a general sense that the stock has been depleted over time, but there is little evidence that

this is the case. A starting point in evaluating the use of the estuary as important habitat is

quantifying the contribution that Great Bay makes to the Gulf of Maine winter flounder

population. There is no awareness of the magnitude of contribution and therefore it is

difficult to measure changes in this amount over time.

There is also little known about the movements of the adult (spawning)

population within the bay, such as how long they reside there or how far they travel up

Page 131: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

into the estuary. Spawning habitats have been delineated, but not confirmed. These areas

would be important habitat to characterize and map first. A tagging study may provide

useful information about flounder abundance in the bay and how the fish are using this

resource.

Organisms within the Great Bay Estuary System are subjected to strong tidal

currents (Short et al. 1992). It would not be surprising to see that some areas of the

estuary are settled by flounder simply because that is where the currents directed them.

Entrainment of larvae in these currents is not well understood even though the

bathymetry and oceanographic processes of the area are fairly well known. Patterns of

distribution in young-of-the-year fish may be related to these processes rather than in the

small-scale events such as changes in salinity.

Species Richness

It has been suggested that species richness is related to the structural

heterogeneity of a habitat. In one study it was found that the greatest number of species

was associated with a complex substrate (Szedlmayer & Able 1996). Species richness

may be a good indicator of essential habitat, or habitat that is important to a variety of

organisms. Characterizing complexity may be an important for distinguishing important

habitats. It may also give more management power as the habitats that are designated as

“essential” are so to a group of organisms, not just one. In this study, all species collected

were recorded, but there was no evaluation of species richness or habitat complexity in

the areas studied. The number of fish, invertebrates, and benthic fauna that an area would

support would be an indication of how “good” that habitat is for estuarine species.

Page 132: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Evaluation of the structural heterogeneity of the sites and their species richness would be

useful analyses as they evaluate these areas as ecosystems not just as habitats.

Food Limitations

This study indicates that the flounder population is below capacity. There is,

however, the question of whether prey items are a limiting resource in the estuary. It

appears from this study that food is not a limiting factor of flounder growth and survival.

It would be of interest to calculate the amount of resources that a single fish would

require and compare this to the data collected. This would be difficult as the flounder are

in competition with other fish and invertebrate species. The flounder are evenly

distributed through the estuary, so it is apparent that these areas are acceptable habitats.

Prey items did not vary significantly between sites so it would seem that sites, at least in

this study, probably all provide adequate prey availability.

Predation

Predation may be the greatest limiting factor on juvenile flounder in the

bay. It is apparent that young-of-the-year fish attain a size refuge within the first few

months. It would be of interest to study the predation pressure on these fish before and

after this size. Bird predation may provide the greatest pressure on larger juveniles as the

bay lacks large piscivores. Some work has been completed addressing this issue, but

results are still inconclusive (Fairchild 2002). This pressure is difficult to quantify in situ.

Geographic Information Systems

In theory, habitat suitability models provide useful information about how an

organism could use specific habitats. The model assigns a numeric value to an area that is

Page 133: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

indicative of that area’s value to a specific organism. These values can be imputed into a

Geographic Information System (GIS). A GIS not only allows you to visualize abiotic

data layers such as temperature and salinity, but also layers of biological data such as fish

abundance, prey abundance, and bird nesting sites. A GIS allows areas to be shown that

are susceptible to anthropogenic disturbances such as sewage outflow and boating traffic.

A GIS allow one to visualize these data and manipulate them. Overlay analyses could be

performed, such as those used for the habitat suitability models. Cut analyses could also

be used to exclude areas where conditions are completely unsuitable. Proximity analyses

would show areas that are close to point sources of pollution. These data are valuable for

use in management decisions and also enhancement projects. Having these data in a GIS

would allow the user to determine possible release sites that represent the optimal

conditions needed for growth and survival. From a map of the bay, areas that are

completely unsuitable could be cut. Point sources of pollution could be identified. A

buffer could be created around these areas. Areas with preferred sediment types and

abiotic conditions could be identified from what was left. These areas may be considered

first for enhancement projects. They could be the first investigated by cage studies

measuring growth.

References

Page 134: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Able, K.W., Manderson, J.P., and A.L. Studholme. 1999. Habitat quality for

shallow water fishes in an urban estuary: The effects of man-made structures on

growth. Mar. Eco. Prog. Ser. 187:227-239.

Abookire, A.A. and B.L. Norcross. 1998. Depth and substrate as determinants ofdistribution of juvenile flathead sole (Hippoglossoides elassodon) and rock sole(Pleuronectes bilineatus), in Kachemak Bay, Alaska. J. of Sea Res. 39:113-123.

Allen, R.L. and D.M. Baltz. 1997. Distribution and microhabitat use by flatfishes in aLouisiana estuary. Env. Bio. Fish. 50:85-103.

Amezcua, F. and R.D.M. Nash. 2001. Distribution of the order Pleuronectiformes inrelation to the sediment type in the North Irish Sea. J. Sea Res. 45:293-301.

Ansell, A.D. and R.N. Gibson. 1993. The effect of sand and light on predation of juvenileplaice (pleuronectes platessa) by fishes and crusteceans. J. Fish. Biol. 43:837-845.

Armstrong, M.P.1995. A comparitive study fo the ecology of smooth flounder,Pleuronectes putnami, and winter flounder, Pseudopleuronectes americanus,from Great Bay Estuary, New Hampshire. Ph.D. Dissertation. University of NewHampshire. 147p.

Armstrong, M. P.1997. Seasonal and ontogenetic changes in distribution and abundanceof smooth flounder, Pleuronectes putnami, and winter flounder, Pleuronectesamericanus, along estuarine depth and salinity gradients. Fish. Bull. 95:414-430.

Bailey, K.M.1994. Predation on Juvenile Flatfish and Recruitment Variability. Neth. J.Sea. Res. 32(2):175-189.

Bailey, K.M. 1997. Structural dynamics and ecology of flatfish populations. J.

Sea. Res. 37: 269-280.

Banner, A. and G. Hayes. Mapping Important Habitats of Coastal New Hampshire.Chapter 6: Winter flounder. http://gulfofmaine.org/library/gbay/wfl.htm.

Bejda, A.J., Phelan, B.A., and A.L. Studholme. 1992. The effect of dissolved oxygen onthe growth of young-of-the-year winter flounder, Pseudopleuronectesamericanus. Env. Biol. Fish. 34:321-327.

Berry, R.J., Saila, S.B., and D.B. Horton. 1965. Growth Studies of Winter flounder,Pseudopleuronectes americanus (Walbaum), in Rhode Island. Trans. Am. Fish.Soc. 94:259-264.

Bigelow & Schroeder. 1953. Fishes of the Gulf of Maine. U.S. Fish & Wildl. Serv. FishBull. 53: 577p.

Page 135: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Brazner, J.C. and E.W. Beals. 1997. Patterns in Fish Assemblages from coastal wetlandand beach habitats in Green Bay, Lake Michigan: a multivariate analysis ofabiotic and biotic forcing factors. Can. J. Fish. Aq. Sci. 54:1743-1761.

Brown, B.A. 1993. Classification System of Marine and Estuarine Habitats in Maine: AnEcosystem Approach to Habitats.1-51.

Brown, S. K., Buja K.R., Jury, S.H., Monaco, M.E., and A. Banner. 2000. HabitatSuitability Index Models for Eight Fish and Invertebrate Species in Casco andSheepscot Bays, Maine. 1-51.

Buckley, J. 1989. Species Profiles: Life histories and environmental requirements ofcoastal fishes and invertebrates (North Atlantic) - winter flounder. U.S. Fish andWildlife Service Biological Report 82(11.87).

Burke, J.S., Miller, J.M., and D.E. Hoss. 1991. Immigration and Settlement Pattern ofParalichthys dentatus and P. Lethostigma in an Estuarine Nursery Ground, NorthCarolina, U.S.A. Neth. J. Sea Res. 27(3/4):393-405.

Carlson, J.K., Randall, T.A., and M.E. Mroczka. 1997. Feeding Habits of WinterFlounder (Pleuronectes americanus) in a Habitat Exposed to AnthropogenicDisturbance. J. Northwest At. Fish. Sci. 21:65-73.

Casterlin, M.E. and W.W. Reynolds. 1982. Thermoregulatory behavior and diel activityof yearling winter flounder, Pseudopleuronectes americanus (Walbaum). Env.Biol. Fish. 7(2):177-180.

Choat, J.H. 1982. Fish Feeding and the Structure of Benthic Communities in TemperateWaters. Ann. Rev. Ecol. Syst.13: 423-49.

Clarke, K.R. and R.M. Warwick. A taxonomic distinctness index and its statisticalproperties. J. Appl. Eco. 35: 523-531. 1998.

Danila, D.J. and M.J. Kennish. 1981. Tagging Study of Winter Flounder(Pseudopleuronectes americanus) in Barnegat Bay, New Jersey.

Everich, D. and J.G. Gonzalez. 1997. Critical Thermal Maxima of Two Species of

Estuarine Fish. Mar. Biol. 41:141-145.

Fairchild, E.A. 2002. Winter flounder, Pseudopleuronectes americanus, stockenhancement in New Hampshire: developing optimal release strategies. Ph.D.Dissertation, University of New Hampshire. 142p.

Fairchild, E.A. and W.H. Howell. 2000. Predator-prey size relationship betweenPseudopleuronectes americanus and Carcinus maenas. J. Sea Res. 44(1-2):81-90.

Field, J.G., Clarke, K.R., and R. M. Warwick. 1982. A Practical Strategy for Analysing

Page 136: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Multispecies Distribution Patterns. Mar. Ecol. Prog. Ser. 8:37-52.

Fisher, W. and C.S. Toepfer. 1998. Recent Trends in Geographic Information SystemsEducation and Fisheries Research Applications at U.S. Universities. Fisheries23(5):10-13.

Fluharty, D. 2000. Habitat Protection, Ecological Issues, and Implementation of theSustainable Fisheries Act. Ecological Applications 10(2):325-337.

Frame, D.W. 1973. Biology of Young Winter Flounder Pseudopleuronectes americanus(Walbaum); Metabolism Under Simulated Conditions. Trans. Amer. Fish. Soc.2:423-430.

Franz, D.R. and J.T. Tanacredi. 1992. Secondary Production of the Amphipod Ampeliscaabdita (Mills) and Its Importance in the Diet of Juvenile Winter Flounder(Pleuronectes americanus) in Jamaica Bay, New York. Estuaries 15(2):193-203.

Gibson, R.N. 1994. Impact of Habitat Quality and Quantity on the Recruitment ofJuvenile Flatfishes. Neth. J. of Sea Res. 32(2):191-206.

Gibson, R.N. and L. Robb. 1992.The relationship between body size, sediment grain sizeand the burying ability of juvenile plaice, Pleuronectes platessa (L.) J. Fish Biol.40:771-778.

Gordon, W.R. 1994. A Role for Comprehensize Planning, Geographical InformationSystem (GIS) Technologies and Program Evaluation in Aquatic HabitatDevelopment. Bull. Mar. Sci. 55(2-3):995-1013.

Gregr, E.J. and A.W. Trites. 2001. Prediction of critical habitat for five whale species inthe waters of coastal British Columbia. Can. J. Fish. Aquat. Sci. 58:1265-1285.

Grizzle, R.E. Seasonality of macrofaunal benthos in New England estuaries. In press .1999.

Grout, D., McBane, C. Patterson, C., Smith, B., Trested, D. and C. Baker. 2001.Programs Improving Management of ASMFC Managed Species in NewHampshire. 2001 Final Report. NMFS Federal Aid Project 3-ACA-071. N.H. Fishand Game Dept. 31p.

Habitat Evaluation Procedures: Standards for Development of HSI Models.http://www.fws.gov/directives/library/hbindex.html#HEP. 1980.

Hanson, M.J. and S.C. Courtenay. 1996. Seasonal Use of Estuaries by Winter Flounder inthe Southern Gulf of St. Lawrence. Trans. Am. Fish. Soc. 125:705-718.

Hofmann, E.E. and T.M. Powell. 1998. Environmental Variability Effects on MarineFisheries: Four Case Histories. Ecol. App. 8(1):s23-s32.

Page 137: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Howell, P.T. and R.B. Harris. 1999. Juvenile Winter Flounder Distribution by Habitat.Estuaries 22(4):1090-1095.

Huey, R.B. 1991. Physiological Consequences of Habitat Selection. The AmericanNaturalist 137: s91-s115.

Hyslop, E.J. 1980. Stomach Content Analysis-a review of methods and their application.J. Fish Biol. 17:411-429.

Isaak, D.J. and W.A. Hubert. 1997. Integrating New Technologies into Fisheries Science:The Application of Geographic Information Systems. Fisheries 22(1):6-10.

Jager, Z.H., Kleef, L. and P.L. Tydeman. 1993. The distribution of 0-group flatfish inrelation to abiotic factors on the tidal flats in the brackish Dollard (Ems Estuary,Wadden Sea). J. of Fish Biol. 43(Supp. A):31-43.

K.R. Clarke and M. Ainsworth. 1993. A method of linking multivariate communitystructure to environmental variables. Mar. Eco. Prog. Ser. 92:205-219.

Keats, D.W. 1990. Food of winter flounder Pseudopleuronectes americanus in a seaurchin dominated community in eastern Newfoundland. Mar. Eco. Prog. Ser.60:13-22.

Keefe, M. and K.W. Able. 1994. Contributions of Abiotic and Biotic Factors toSettlement in Summer Flounder, Paralichthys dentatus. Copeia 2:458-465.

Keleher, C.J. and F.J. Rahel. 1996. Thermal Limits to Salmonid Distributions in theRocky Mountain Region and Potential Habitat Loss Due to global Warming: AGeographic Information System Approach. Trans. Am. Fish. Soc. 125(1):1-13.

Kennedy, V.S. and D.H. Steele. 1971. The Winter Flounder (Pseudopleuronectesamericanus) in Long Pond, Conception Bay, Newfoundland. J. Fish. Res. Bd.Canada 28:1153-1165.

Klein-MacPhee, G. 1978. Synopsis of Biological Data for the Winter Flounder,Pseudopleuronectes americanus (Walbaum). NOAA Technical Repost NMFSCircular 414:1-43.

Kramer, D.L. 1987. Dissolved oxygen and fish behavior. Env. Biol. Fish. 18(2):81-92.

Kuipers, B.R. et al. 1992. Small Trawls in Juvenile Flatfish Research: Their Developmentand Efficiency. Neth. J. Sea Res. 29(1-3):109-117.

Langton, R.W., Steneck, R.S., Gotceitas, V., Juanes, F., and P. Lawton. 1996. TheInterface between Fisheries Research and Habitat Management. N. Am. J. of Fish.Man.16:1-7.

Laurence, G.C. 1977. A bioenergetic model for the analysis of feeding and survival

Page 138: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

potential of winter flounder, Pseudopleuronectes americanus, larvae during theperiod from hatching to metamorphosis. Fish. Bull. 75(3):529-546.

Leopold, M.F., van Damme, C. and H.W. van der Veer. 1998. Diets of cormorants andimpact of comorant predation on juvenile flatfish in the Dutch Wadden Sea. J. SeaRes. 40:93-107.

Macdonald et al. 1984. Fishes, Fish Assemblages, and their Seasonal Movements in theLower Bay of Fundy and Passamaquoddy Bay, Canada. Fish. Bull. 82(1):121-138.

Magnuson Fisheries Management and Conservation Act. 1976.

Martell, D.J. and G. McClelland. 1994. Diets of sympatric flatfishes, Hippoglossoidesplatessoides, Pleuronectes ferrugineus, Pleuronectes americanus, from SableIsland Bank, Canada. J. Fish. Biol. 44: 821-848.

Mattila, J. and E. Bonsdorff. 1998. Predation by juvenile flounder (Platichthys flesus L.):a test of prey vulnerability, predator preference, switching behavior and functionalresponse. J. Exp. Mar. Bio. Ecol. 227: 221-236.

McCracken, F. D. 1963. Seasonal Movements of the Winter flounder,Pseudopleuronectes americanus (Walbaum), on the Atlantic Coast. J. Fish. Res.Bd. Canada 20(2): 551-583.

Meng, L. and J. Christopher Powell. 1999. Linking Juvenile Fish and Their Habitats: AnExample from Narragansett Bay, Rhode Island. Estuaries 22(4): 905-916.

Meng, L., Powell, J.C., and B.Taplin. 2001. Using Winter Flounder Growth Rates toAssess Habitat Quality Across an Anthropogenic Gradient in Narragansett Bay,Rhode Island. Estuaries 24(4):576-584.

Miller, J.M., Burke, J.S. and G.R. Fitzhugh. 1991. Early Life History Patterns of AtlanticNorth American Flatfish: Likely (and Unlikely) Factors Controlling Recruitment.Neth. J. Sea Res. 27(3/4):261-275.

Moles, A. and B.L. Norcross. 1995. Sediment Preference in Juvenile Pacific Flatfishes.Neth. J. of Sea Res. 34(1-3):177-182.

Neill, W.H., Miller, J.M., van der Veer, H.W., and K.O. Winemiller.1994. Ecophysiologyof Marine Fish Recruitment: A Conceptual Framework for UnderstandingInterannual Variability. Neth. J. Sea Res. 32(2):135-152.

Neuman, M.J. and K.W. Able. 1998. Experimental evidence of sediment preference byearly life history stages of windowpane (Scophthalmus aquosus). J. Sea Res.40:33-41.

Nitschke, P., Brown, R., and L. Hendrickson. Status of Fisheries Resources off

Page 139: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Northeastern United States - Winter Flounder.http://www.whoi.edu/sos/spsyn/fldrs/winter/index.html . 2001.

Norcross, B.L., Muter, F. and B.A. Holladay. 1997. Habitat models for juvenilepleuronectids around Kodiak Island, Alaska. Fishery Bulletin 95:504-520.

Olla et al. 1969. Behavior of Winter Flounder in a Natural Habitat. Trans. Amer. Fish.Soc. 4: 717-720.

Oviatt, C.A. and S.W. Nixon. 1973. The Demersal Fish of Narragansett Bay: an Analysisof Community Structure, Distribution and Abundance. Estuarine and CoastalMarine Science 1:361-378.

Packer, D.B. and T. Hoff. 1999. Life History, Habitat Parameters, and Essential Habitatof Mid-Atlantic Summer Flounder. Amer. Fish. Soc. Symposium 22:76-92.

Pearcy, W.G. 1978. Distribution and Abundance of Small Flatfishes and other DemersalFishes in a Region of Diverse Sediments and Bathymetry off Oregon. FisheryBulletin 76(3): 629-640.

Pearcy, W.G. 1962. Ecology of an Estuarine Population of Winter Flounder,Pseudopleuronectes americanus (Walbaum). Bulletin of The BinghamOceanographic Collection 18(1).

Pereira, J.J., Goldberg, R., Ziskowski, J.J., Berrien, P.L., Morse, W.W., and D.L.Johnson.1999. Winter Flounder, Pseudopleuronectes americanus, Life Historyand Habitat Characteristics. NOAA Technical Memorandum NMFS-NE-138:1-39.

Peterson, C.H. et al. 2000. Synthesis of Linkages between Benthic and Fish Communitiesas a Key to Protecting Essential Fish Habitat. Bull. Mar. Sci. 66(3):759-774.

Phelan, B.A. 1992. Winter Flounder Movements in the Inner New York Bight. Trans.Am. Fish. Soc. 121:777-784.

Phelan, B.A., Manderson, J.P., Stoner, A.W. and A. J. Bejda. 2001. Size-related shifts inthe habitat associations of young-of-the-year winter flounder (Pseudopleuronectesamericanus): field observations and laboratory experiments with sediments andprey. J. of Exp. Mar. Bio. Eco. 257:297-315.

Phelan, B.A., Goldberg, R., Bejda, A.J., Pereira, J., Hagan, S., Clark, P., Studholme,A.L., Calabrese, A. and K.W. Able. 2000. Estuarine and habitat-relateddifferences in growth rates of young-of-the-year winter flounder(Pseudopleuronectes americanus) and taugtog (Tautoga onitis) in threenortheastern U.S. estuaries. J. of Exp. Mar. Bio. Eco. 247:1-28.

Pihl. L. and H.W. van der Veer. 1992. Importance of Exposure and Habitat Structure forthe Population Density of 0-Group Plaice, Pleuronectes platessa L., in Coastal

Page 140: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Nursery Areas. Neth. J. of Sea Res. 29(1-3):145-152.

Randall, D.J. (ed.) Eckert Animal Physiology: Mechanisms and Adaptations. W.H.Freeman & Company. New York. 1998. pp 627-629.

Raz-Guzman, A. and R.E. Grizzle. 2000. Techniques for Quantitative Sampling ofInfauna and Small Epifauna. Chapter 12 in Global Seagrass Research Methods:p18.

Rice, J.C. and A.R. Kronlund. 1997. Community analysis and flatfish: diagnosticpatterns, processes, and inference. J. of Sea Res. 37:301-320.

Rogers, S.I. 1992. Environmental Factors Affecting the Distribution of sole (Solea soleaL.) within a Nursery Area. Neth. J. of Sea. Res. 29(1-3):153-161.

Rosenberg, A., Bigford, T.E., Leathery, S., Hill, R.L., and K. Bickers. 2000. EcosystemApproaches to Fishery Management through Essential Fish Habitat. Bull. Mar.Sci. 66(3):535-542.

Ross, R.M.1993. Habitat use by spawning adult, egg, and American shad in the DelawareRiver. Rivers 4(3):227-238.

Rountree, R.A. and K.W. Able. 1992. Foraging Habits, Growth, and Temporal Patternsof Salt-marsh Creek Habitat Use by Young-of-Year Summer Flounder in NewJersey. Trans. Am. Fish. Soc. 121:765-776.

Rubec et al. 1999. Suitability Modeling to Delineate Habitat Essential to SustainableFisheries. Amer. Fish. Soc. Symposium 22:108-133.

Rubec, P.J. et al. 1998. Spatial Methods Being Developed in Florida to DetermineEssential Fish Habitat. Fisheries 23(7):21-25.

Saucerman, S.E. and L.A. Deegan. 1991. Lateral and Cross-Channel Movement ofYoung-of-the-Year Winter Flounder (Pseudopleuronectes americanus) inWaquoit Bay, Massachusetts. Estuaries. 14(4):440-446.

Schmitt, C.J. 1993. Habitat Suitability Index Model for Brook Trout in streams of theSouthern Blue Ridge Province: surrogate variables, model evaluation andsuggested improvements. Biol. Rep. U.S. Fish and Wildlife Service 18.

Shaw, M. and G.P. Jenkins. 1992. Spatial variation in feeding, prey distribution and foodlimitation of juvenile flounder Rhombosolea tapirina (Gunther.) J. Exp. Mar.Biol. Ecol. 165:1-21.

Short, F.T. (ed.) 1992. The Ecology of Great Bay Estuary, New Hampshire and Maine:an Estuarine Profile and Bibliography. 221p.

Page 141: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

Sogard, S. M. 1992. Variability in growth rates of juvenile fishes in different estuarinehabitats. Mar. Ecol. Prog. Ser. 85:35-53.

Sogard, S.M., Able, K.W., and S.M. Hagan. 2001. Long-term assessment of settlementand growth of juvenile winter flounder (Pseudopleuronectes americanus) in NewJersey estuaries. J. Sea Res. 45:189-204.

Stehlik, L.L. and C.J. Meise. 2000. Diet of Winter Flounder in a New Jersey Estuary:Ontogenetic Change and Spatial Variation. Estuaries 23(3):381-391.

Stoner, A.W., Manderson, J.P., and J.P. Pessutti. 2001. Spatially explicit analysis ofestuarine habitat for juvenile winter flounder: combining generalized additivemodels and geographic information systems. Mar. Ecol. Prog. Ser. 213:253-271.

Sustainable Fisheries Act. U.S. Senate 23 May 1996. Report of the Committee onCommerce, Science, and Transportation on S.39: Sustainable Fisheries Act.Report 104-276, 104th Congress, Second Session. US Government Printing,Washington, D.C. U.S.A.

Swartzman, G. and C. Huang. 1992. Spatial Analysis of Bering Sea Groundfish SurveyData Using Generalized Additive Models. Can. J. Fish. Aq. Sci. 49:1366-1378.

Systat 10 Manual. 2001.

Szedlmayer, S.T. and K.W. Able. 1996. Patterns of Seasonal Availability and HabitatUse by Fishes and Decapod Crustaceans in a Southern New Jersey Estuary.Estuaries 19(3):697-709.

Targett, T.E. and J.D. McCleave. 1974. Summer Abundance of Fishes in a Maine TidalCove With Special Reference to Temperature. Trans. Amer. Fish. Soc. 2:325-330.

US Fish and Wildlife Service: Multivariate Statistical Analyses. 2001.

van der Veer, H.W. and J.I.J. Witte. 1993. The 'Maximum growth/optimal food condition'hypothesis: a test for 0-group plaice Pleuronectes platessa in the Dutch WaddenSea. Mar. Ecol. Prog. Ser.101:81-90.

Van Glupen, L. and C.C. Davis. 1979. Seasonal Movements of the Winter Flounder,Pseudopleuronectes americanus, in Two Contrasting Inshore Locations inNewfoundland. Trans. Am. Fish. Soc. 108:26-37.

Walsh, H.J., Peters, D.S., and D.P. Cyrus. 1999. Habitat Utilization by Small Flatfishes ina North Carolina Estuary. Estuaries. 22(3B):803-813.

Wells, B., Steele, D. H. and A.V. Tyler. 1973. Intertidal Feeding of Winter Flounders(Pseudopleuronectes americanus) in the Bay of Fundy. J. Fish. Res. Board Can.

Page 142: USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL ...ratt.ced.berkeley.edu/readings/GIS_readings/JWThesis.pdf · USING HABITAT SUITABILITY MODELS TO IDENTIFY ESSENTIAL FISH HABITAT

30(9): 1374-1378.

Wennage, H. and L. Pihl. 1994. Substratum Selection by Juvenile Plaice (Peuronectesplatessa L.): Impact of Benthic Microalgae and Filamentous Macroalgae. Neth. J.Sea Res. 32(3/4):343-351.

Werner, E.E., Gilliam, J.F., Hall, D.J. and G.G. Mittelbach. 1983. An Experimental Testof the Effects of Predation Risk on Habitat Use in Fish. Ecology. 64(6):1540-1548.

Yamashita, Y. Otake T. and Yamada H. 2000. Relative contributions from exposedinshore and estuarine nursery grounds to the recruitment of stone flounder,Platichthys bicoloratus, estimated using otolith Sr:Ca ratios. FisheriesOceanography 9(4):316-327.

Yamashita, Y., Tanaka, M. and J.M. Miller. 2001. Ecophysiology of juvenile flatfish innursery grounds. J. Sea Res. 45:205-218.