Investigating and interpreting reduced reproductive performance in fish inhabiting streams adjacent to agricultural operations

by

Sandra Marie Brasfield

Masters of Science, Zoology, Oklahoma State University

Bachelor of Science, Biology, Middle Tennessee State University

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

Doctor of Philosophy

In the Graduate Academic Unit of Biology

Supervisor: Kelly Munkittrick, Ph.D, UNBSJ Biology

Examining Board: Jack Terhune, Ph.D, Chair

Jeff Houlahan, Ph.D., UNBSJ Biology

Simon Courtenay, Ph.D., Fisheries and Oceans Canada

Katy Haralampides, Ph.D., UNBF Civil Engineering

External Examiner: D. George Dixon, Ph.D. Biology, University of Waterloo

This thesis is accepted

____________________

Dean of Graduate Studies

THE UNIVERSITY OF NEW BRUNSWICK

JULY, 2007

© Sandra Brasfield 2007

ii

ABSTRACT

Previous studies have indicated that fish in agricultural areas have reduced

proportions of young of the year (YOY) fish, due to either reproductive

dysfunction, increased mortality, or a combination of these factors. The main

objective of this research was to identify the potential mechanisms and timing

associated with the reduced reproductive performance in agricultural areas. Fish

populations were monitored systematically through multiple years to identify peak

mortality and risk periods based on agricultural practice and inputs. In previous

assessments sculpin species have presented conflicting response patterns,

making interpretation difficult. Normal reproductive profiles, growth and mortality

were assessed for reference populations of slimy sculpin (Cottus cognatus) to

identify the most appropriate window to assess reproductive integrity.

Comparisons between spring and fall spawning species in agricultural areas

determined the degree of potential impact of exposure to stressors on differing

reproductive strategies. Finally the integration of this information collected on

episodic mortality, reproductive development and reproductive performance

required some consideration of population level impacts. The tools for population

level ecological risk assessment are poorly developed, and the theoretical aspect

of this thesis focused on trying to integrate the available information, identify the

data gaps, and recommend an overall approach to population risk assessment

that will be based on information gathered regarding non-point, multiple stressor

discharges.

iii

ACKNOWLEDGMENTS

I have many, many people to thank for their support as I pursued this degree, but

head and shoulders above them all is my friend and supervisor, Dr. Kelly

Munkittrick. When my efforts were pitiful and my outlook cloudy, you always

offered sunshine and at least a couple of paths forward. Working with you

continues to be in the top five decisions I have made. My upcoming opportunities

are only possible because of the work I have done, the researcher I have

become since working with you and the CRI. I have improved in ways that are

apparent even to me. And that is really saying something.

The faculty, staff and students of the Canadian Rivers Institute and UNB

Saint John made my research experience one that will never be matched. The

crew and equipment may have got things going, but the people and the places

helped me get the job done. There was always an eager volunteer for the various

pursuits of fish, philosophy, and fun. So many of you have touched my life in

ways I still cannot define. You have set a high standard for the people I will work

with for the rest of my career. In the immortal words of my pal Chris Blanar “It

takes guts to raise the bar when everyone else is playing limbo.” Thank you for

that.

My family may have wondered why I came to Canada, even when they

saw me traveling and working all over the map, they just took it all in stride and

assumed I was getting the education I wanted. They came to visit when they

could, and hopefully understood a little more as to why I could love being in a

place so far from Tennessee. Although they missed me being closer, I was

iv

allowed the freedom to spread my wings and thank goodness for the benefit of

doubt. I would still be somewhere with my head in the clouds if there weren’t for

their rocks in my pockets. I love you all very much.

My coast-to-coast support system has expanded outside of North

America, and I am so thankful for the amazing friends that always had a kind

word or a great joke, and more importantly, faith in me. Even when I was running

low on confidence, you helped me find a way to carry on. Moving to another

country will often provide an opportunity for many people to lose touch and only

strongest friendships are left, thanks for hanging in there until the clock read

00:00.

In Canada, I was fortunate to find kindred spirits for all my varied interests,

including football, music, food and drink, all in the pursuit of fun. I will think of

many of you for the rest of my life and smile, giggle, maybe even blush a little. I

had no idea who I would meet when I moved here, and I couldn’t be more

pleased with the motley crew that amassed, disbanded, and reunited over the

four years I spent with you. You made Canada home, when I was many

thousands of miles from mine, and for that I will always be grateful.

I got somethin' in my heart, I been waitin' to give

I got a life I wanna start, one I been waitin' to live

No more waitin', tonight I feel the light I say the prayer

I open the door, I climb the stairs...

-- Bruce Springsteen

v

TABLE OF CONTENTS

Investigating and interpreting reduced reproductive performance in fish inhabiting

streams adjacent to agricultural operations ..........................................................i

1 GENERAL INTRODUCTION....................................................................... 1

1.1 Overview.................................................................................... 1

1.2 Multiple Stressors and Non-point Source Pollution.................... 2

1.3 Effects Based Assessment ........................................................ 4

1.4 Statement of Problem ................................................................ 5

1.5 General Overview of the Study Area ......................................... 7

1.6 Objectives and outline of thesis ................................................. 8

1.7 References .............................................................................. 16

2 Monitoring of fish populations along a gradient of agricultural inputs in

northwestern New Brunswick, Canada..............................................................22

2.1 Abstract ................................................................................... 22

2.2 Introduction.............................................................................. 23

2.3 Methods................................................................................... 25

2.4 Results..................................................................................... 27

2.5 Discussion ............................................................................... 30

2.6 References .............................................................................. 35

3 Seasonal patterns of energy storage, energy expenditure, and in vitro

gonadal steroidogenic capacity in slimy sculpin (Cottus cognatus) ...................50

3.1 Abstract ................................................................................... 50

vi

3.2 Introduction.............................................................................. 51

3.3 Methods................................................................................... 53

3.4 Results..................................................................................... 56

3.5 Discussion ............................................................................... 59

3.6 References .............................................................................. 64

4 Comparison of spring spawning slimy sculpin (Cottus cognatus) and fall

spawning brook trout (Salvelinus fontinalis) reproductive development in

agricultural regions of the St. John River (New Brunswick, Canada).................73

4.1 Abstract ................................................................................... 73

4.2 Introduction.............................................................................. 74

4.3 Methods................................................................................... 77

4.4 Results..................................................................................... 80

4.5 Discussion ............................................................................... 81

4.6 References .............................................................................. 87

5 Approaching population-level ecological risk assessment from an effects

driven perspective .............................................................................................97

5.1 Abstract ................................................................................... 97

5.2 Introduction.............................................................................. 98

5.3 Background to Agricultural Studies........................................ 100

5.4 Risk Assessment ................................................................... 102

5.5 Population-level assessment of agricultural impacts on fish.. 105

5.6 Population-Level Problem Formulation.................................. 105

5.7 Population-level risk analysis................................................. 108

vii

5.8 Population-level risk characterization .................................... 110

5.9 Additional Uses...................................................................... 111

5.10 Conclusions ........................................................................... 112

5.11 References ............................................................................ 114

6 Conclusions..............................................................................................127

7 VITA ............................................................................................................ 0

viii

LIST OF TABLES Table 1-1 A comparison of stressor-based, effects-based and values-based

approaches to environmental assessment..................................13 Table 1-2 Biological attributes suggested for use in population level

ecological risk assessment by empirical working group of Pellston Workshop on Population Level Ecological Risk Assessment.................................................................................14

Table 2-1 Precipitation data reported as monthly and annual total rainfall (mm) at the St Leonard airport (St. Leonard, NB, station 6256) (http://www.climate.weatheroffice.ec.gc.ca/climateData/monthlydata_e.html) Storms are listed and enumerated based on the total precipitation to indicate storm duration and intensity. The bolded text indicates the months where potatoes are actively in cultivation. ..................................................................40

Table 2-2 Degree-days (sum of mean daily temperature) for water temperatures collected between 27 July and 18 October for the Little River. Temperature data for 2002 was extrapolated from a regression generated by air: water temperature from 3 other years. Air temperature data collected at the St Leonard airport, (St Leonard, NB) . Station ID 6256. http://www.climate.weatheroffice.ec.gc.ca/ .................................41

Table 2-3 Young of the year (YOY) slimy sculpin data collected from monthly monitoring of Little River along a gradient of agricultural inputs (2002-2004). YOY were first caught in nets in August of all years and sampling was conducted in ice- free months only. All fish were measured and released back into collection areas. POP refers to the proportion of the population sample collected represented by YOY. Size of YOY is reported as median size of the sample. K is the condition factor calculated as (weight/length3)*100000, when length is reported in millimeters...................................................42

Table 2-4 Data collected from monthly sampling of adult slimy sculpin (2002-2004). Sampling was conducted post-freshet and in ice free months only. All fish were measured and released back into collection areas. Data are reported as mean (SEM). Letters (a,b) indicate a significant difference in condition (length-weight relationship) between sites in that month only. † indicates presence of interaction in ANCOVA analysis (p<0.05).......................................................................................43

ix

Table 3-1 Monthly values for mean total length and mean body weight of slimy sculpin (Cottus cognatus) collected from May 2003-May 2004. Values are mean ± SEM (N) *indicates significant change from preceding month (p<0.05). Condition factor was calculated as k=(weight/(length)3)*100. * indicates a significant difference (p<0.05) from the preceding month. ..........69

Table 4-1 Length, weight, condition factor, liversomatic index (LSI) and gonadosomatic index (GSI) for slimy sculpin and brook trout collected along the St John River. Data are shown as mean ± standard error..............................................................................90

Table 5-1 Recommended empirical and modeling attributes that should be collected or computed as part of a population level ecological risk assessment (Barnthouse et al. 2007). ................................117

Table 5-2 Available data for brook trout (Salvelinus fontinalis) and slimy sculpin (Cottus cognatus) to be used in population level ecological risk assessment. Efforts were made to utilize information collected in the St John River system with standardized sampling, where possible. ...................................118

x

LIST OF FIGURES

Figure 1-1 Map of (A) North America, (B) New Brunswick and (C) the St John River. sampling locations within the St. John River basin, redrawn from topographical maps obtained from the Province of New Brunswick. Inset maps provided by M.A. Gray and Environment Canada...................................................15

Figure 2-1 Map of (A) North America, (B) New Brunswick and the St John River (C) Little River and sampling sites. These sites were chosen along the Little River based on studies conducted by Gray et al. (2005). Location of Black Brook is indicated by the inset box above lowermost field station. rectangle drawn. ..........44

Figure 2-2 Slimy sculpin population structure for estimated catches of slimy sculpin collected monthly along a gradient of agricultural intensity, A) TenMile (forested), B) Donat (intermediate), C) Dead (Agriculture). Data are shown as estimated catch based on standardized effort across all sites. Data are categorized into bins based on fish total length (mm).....................................45

Figure 2-3 Estimated catch of slimy sculpin along a gradient of agriculture for A) young of the year (YOY), B) adult sculpin >70mm, C) all size classes of slimy sculpin collected monthly.. Data are shown as means of estimated catch based on standardized effort across all sites each month during 2002-2004...................46

Figure 2-4 Relationship between maximum daily temperature and estimated density of all sculpin based on standardized effort.

indicate data from forested sites and ○ indicate data from agricultural sites. .........................................................................47

Figure 2-5 Linear relationship between total rainfall (mm) and percent young of the year (YOY) for sites along the gradient of agriculture. indicate data from forested sites and □ indicate data from agriculture sites. Line shown is for agricultural sites only. ............................................................................................48

Figure 2-6 Relationship between number of summer storms >15 mm in July and August in total precipitation and percent young of the year sculpin. ................................................................................49

Figure 3-1 Monthly changes (mean monthly values) in male slimy sculpin A gonadosomatic index (gonad weight/body weight – gonad weight *100) B gonadal in vitro steroidogenic capacity to produce11-ketotestosterone C gonadal in vitro steroidogenic capacity to produce testosterone. Solid lines represent seasonal changes in the monthly mean. Dashed lines define "shut off" as defined by the maximum production value in the months of minimum steroidogenic capacity. */‡/† indicates significant (p<0.05) change from the preceding month using Mann Whitney nonparametric probabilities. ................................70

xi

Figure 3-2 Monthly changes (mean monthly values) in female slimy sculpin A gonadosomatic index (gonad weight/body weight – gonad weight *100) B gonadal forskolin stimulated in vitro steroidogenic capacity to produce 17β estradiol C gonadal in vitro steroidogenic capacity to produce testosterone. Solid lines represent seasonal changes in the monthly mean. Dashed lines define "shut off" as defined by the maximum production value in the months of minimum steroidogenic capacity. */‡/† indicates significant (p<0.05) change from the preceding month using Mann Whitney nonparametric probabilities. ................................................................................71

Figure 3-3 Monthly changes in energy storage in A male slimy sculpin (Cottus cognatus) hepatosomatic index [liver wt/(body wt-liver wt)]*100 B female sculpin hepatosomatic index from May 2003- 2004. * indicates significant (p<0.05) change from the preceding month using Mann Whitney nonparametric probabilities. ................................................................................72

Figure 4-1 Map of (A) North America, (B) New Brunswick and (C) the St John River (C) Little River and sampling sites. These sites were chosen along the St John River based on studies conducted by Gray et al. (2005). .................................................91

Figure 4-2 Relationship between gonad weight v. body weight (A) and liver weight v. body weight (B) for slimy sculpin collected at 6 sites along the St John River...............................................................92

Figure 4-3 Steroidogenic capacity of (A) testosterone and (B) estradiol by gonadal tissue excised from slimy sculpin collected in April 2004 at agricultural (AG 1-3) and forested (FOR 1-3) sites. Hormone determinations were made using RIA with the incubation media. ........................................................................93

Figure 4-4 Relationship between gonad weight v. length (A) and liver weight v. length (B) for brook trout collected at 4 sites along the St John River.........................................................................95

Figure 4-5 Steroidogenic capacity of (A) testosterone and (B) estradiol by gonadal tissue excised from brook trout collected in September 2004 at agricultural (AG 1 & 3) and forested (FOR 1-2) sites. Hormone determinations were made using RIA with the incubation media............................................................96

Figure 5-1 Comparing assessment endpoints on the individual vs. population level in the overall levels of biological organization. 119

Figure 5-2 Ecological risk assessment framework as defined by the US EPA (1998) and as modified to address population level assessment (Barnthouse et al. 2007)........................................120

Figure 5-3 Conceptual diagram of stressors associated with potato farming practices in northwestern New Brunswick as well as responses documented in previous study (1999-2002) by Gray and colleagues. ................................................................121

xii

Figure 5-4 Relationship between sculpin density (per m2) and maximum mean daily water temperature. Open triangles represent agricultural sites, filled triangles represent forested sites. Graph reprinted with permission from M. Gray..........................122

Figure 5-5 Linear relationship between number of major summer storms (>15 mm total rainfall) and percent young of the year (YOY) for sites along the gradient of agricultural inputs. Data were collected over a period of 1999-2004. .......................................123

Figure 5-6 Map of New Brunswick showing total degree days over 18°C as observed at monitoring stations for the months of July and August. Groups were assigned and lines were drawn in efforts to combine areas with similar values..............................124

Figure 5-7 Map of New Brunswick showing total number of rainfall events exceeding 15 mm of total precipitation as recorded at monitoring stations during the summer months of July and August. Groups were assigned and lines were drawn in efforts to combine areas with similar values..............................125

Figure 5-8 Map of New Brunswick showing overlapping temperature and precipitation data including total number of degree days over 18°C, total rainfall for the summer months of July and August, and number of storms exceeding 10 mm total precipitation. .....126

1

1 GENERAL INTRODUCTION

1.1 Overview

With increasing legislation and mounting popular concern for the

environment, good environmental management has become a priority (Skinner et

al. 1997, Chambers 2002), but little attention is given to non-point sources of

pollution. Intensification of agricultural operations over the last 40 years has

increased the risk of contamination of surface and ground waters by eroded soils,

nutrients, herbicides and pesticides. Subsequently, there has also been a recent

increase in research into the environmental impacts of agricultural activities and

the development of focused environmental management techniques in

agriculture (Chambers et al. 2002).

Following a recent assessment of threats to water quality (Environment

Canada 2001), the Canadian Council of Ministers of the Environment (CCME)

held a series of workshops to explore the effects of agricultural activity on aquatic

ecosystems. The CCME identified and reported several research gaps that

should be addressed by scientists and policy makers (Chambers et al. 2002).

Three key areas of concern identified were soil erosion, nutrient inputs, and

pesticide residues from agricultural runoff. Soil erosion involves mobilization and

deposition of topsoil and can be accelerated by cultivation practices. Eroded

sediments also act as carriers for both nutrients and pesticides. It remains

unclear what determines the bioavailability of these contaminants, and what the

realistic resulting hazards are to aquatic systems. Elevated nutrients such as

2

nitrogen and phosphorus are also of concern, as increased growth of lower

trophic levels can result in eutrophication. These chemical elements are

considered limiting in non-impacted systems, however, agricultural practices

provide a source of nutrients to surface and ground waters, usually from overuse

of fertilizers (Chambers et al. 2002). Finally, pesticides applied to farmed lands

are of concern, as these compounds have become the primary approach to

control weeds, insects and diseases that threaten crop yields (Skinner et al.

1997). Although the water quality guidelines are seldom exceeded during normal

application, extreme weather and poor management practices continue to create

concerns about the consequences of long-term exposure to low levels of

chemicals, as well as synergies between these chemicals in mixtures.

1.2 Multiple Stressors and Non-point Source Pollution

In the mid-1990s, there was a move by multiple regulatory bodies in North

America to modify environmental assessment approaches to address the

multitude of stressors in complex receiving environments (Ferenc and Foran

2000, Munkittrick et al. 2000). This involved a conceptual shift away from single

stressor analyses to consider the importance of multiple stressors and

cumulative impacts. The term multiple stressors originated in the field of

ecotoxicology and historically has been used to denote the interaction among

multiple chemicals. In this study, we present multiple stressors in the context of

both chemical and non-chemical stressors.

3

Estimating effects of multiple stressors to biota can be difficult, especially

when stressors are diffuse in nature and interact on differing temporal and spatial

scales, as in the case of agricultural inputs. Observed effects may be direct, such

as increased mortality or decreased fecundity as a consequence of acute

pesticide exposure, or indirect, as in altered predator prey dynamics as a result

of increased turbidity.

Non-point source pollution can negatively impact aquatic biota by altering

physical habitat, modifying seasonal water flow, altering the food base,

contaminating water with toxic chemicals, and modifying interactions among

organisms (Karr 1999, Potter et al. 2004). However, research on the impacts of

agriculture on ecosystem health have primarily focused on single measures or

stressors such as soil erosion (Chow and Rees 1995, Pimental et al. 1995),

pesticide use (Culliney et al. 1992, Clark et al. 1999, Battaglin and Fairchild,

2002), and contamination of ground water by fertilizer and other agricultural

chemicals (Bouwer 1990, Napier and Brown 1993, Böhlke 2002). While these are

important aspects of ecosystem condition, these measures provide only a partial

picture of impact. Integration of stress effects on ecosystem health may be more

efficiently assessed by addressing cumulative impacts and multiple stressors

(Munkittrick et al. 2000). Resident biota of aquatic ecosystems serve as

continuous monitors of cumulative effects on those systems and are often used

as endpoints in environmental assessments (Munkittrick et al. 2000, Diamond

and Serveiss 2001).

4

1.3 Effects Based Assessment

Despite these recent attempts to modify the traditional risk assessment

process to consider multiple stressors (Munkittrick et al. 2000, Dubé and

Munkittrick 2001), a major deficiency with environmental impact assessment has

been its inability to deal with multiple discharges or complex situations. Non-point

stressors such as agricultural inputs and sediment have been difficult to

characterize. In most situations, discharges from non-point sources are complex

mixtures, the concentrations of toxicants are difficult to characterize, and rates

and timing of discharges are difficult to predict (Landis and Yu 1999). An effects-

driven assessment can be compared and contrasted to stressor- and values

based approaches (Table 1-1). Effects-based assessment uses the performance

of resident fish in a system to determine if and to what degree native fish

populations are stressed by different inputs to the system relative to reference

conditions and/or the level of effects normally observed in similar systems which

are free of those inputs (Munkittrick et al. 2000). The performance of fish

populations can be used to identify where existing conditions are compromising

performance and help to understand the level of stress on a river reach.

Therefore, it is the biological effects of resident fish that effectively drives the

study design and sampling by identifying where performance is affected. Then in

a general sense, if fish are able to grow, reproduce and survive at similar rates to

those in reference conditions, you may conclude that there are no measurable

limitations to performance (Munkittrick et al. 2000). Alternately, if resident fish are

limited in any of the selected performance measures, then you may conclude that

5

there are environmental factors contributing to the reduced performance and

design a detailed monitoring program to identify and study the relative

contribution of possible stressors. This removes a certain level of researcher bias

from dictating where the effects would be assumed to be occurring, and may

possibly identify limiting or enhancing factors that may have been missed or

ignored by the stressor-based approach.

Effects-driven assessment has several advantages over the use of

monitoring of chemical residue levels as a measure of impact. This type of

assessment does not assume that exposure represents a deleterious effect.

Assessment is intrinsically related to a valid ecological endpoint: fish populations

and the ability of fish to grow and reproduce similar to fish at reference sites.

Exposure assessment estimates the magnitude of releases, identifies possible

pathways of exposure and estimates potential exposure. The effects-driven

framework is then used iteratively to design focus follow-up studies on the

aspects of performance that are responding to the stressors within the system.

It is not possible to estimate the impacts of all potential stressors on all biota

at all levels of organization. Thus the focus of the assessment must be narrowed

to receptors that are integrated and can provide an appropriate assessment of

the system.

1.4 Statement of Problem

Agricultural activities have received recent attention because of the

popularity of the endocrine disruptor issue and because of recent fish kill events

6

in Eastern Canada (Cairns 2002). Since 1999, there have been >25 documented

fish kills in Prince Edward Island (PEI) (Cairns 2002). The deaths of thousands of

fish on PEI have been attributed to chemical runoff from potato fields (Gray et al.

2002a, Mutch et al. 2002, Gormley et al. 2005). Investigations are still pending

regarding the potential causes, but based on circumstantial evidence, pesticides,

specifically azinphos-methyl are receiving the blame. Although pesticides are

inherently hazardous, these chemicals are not the only causative factor of fish

kills associated with agricultural runoff. In Atlantic Canada, most crops are

sprayed at least 12 times in the four month-long growing season. Potatoes are

typically grown in a three year rotation, and the more recent fish kills occurred in

the years of 1999 and 2002, indicating that there may be a relationship with the

general practices enlisted by potato farmers or with particular fields adjacent to

streams. Heavy rainstorms in mid to late summer can sometimes coincide with

heavier spraying as the potato vine grows and is more susceptible to peril.

Additional impacts from potato production involve removing large amounts of

organic material from the soil, leaving the soil bare for long periods of time, which

can lead to high erosion risk. Agricultural activities can impair surface and/or

groundwater quality.

Recently completed studies have compared fish performance in 20 New

Brunswick tributaries (Gray and Munkittrick 2005), and have shown that slimy

sculpin reproductive performance was very low at agricultural sites. Young-of-

the-year sculpin were not found at 8 of 10 agricultural sites. Detailed studies

have also documented year class failures in agricultural areas, and changes in

7

growth, fecundity and organ sizes in populations of slimy sculpin (Gray 2003).

Circumstantial evidence indicates that storm events play a role in the year class

failures and extinction events, but there is a need to also assess the causative

factors associated with differences seen in growth, organ size and fecundity.

From this work, it was determined that future studies were needed to tease out

the relative influence of temperature, sedimentation, nutrient and pesticides on

the responses observed in the local fish populations. To address the identified

research gaps, this study aimed to identify the timing of mortality events and

further investigate reductions in fish numbers in agricultural areas. Systematic

sampling of fish densities and size structures through multiple years at sites to

identify peak mortality periods was used to try to identify potential stressors

associated with the suspected periods of mortality. Although this study would aid

in the identification of periods of risk to fish populations, additional information

was required to better characterize the seasonal changes normally exhibited by

sculpin with regard to energy storage and expenditure.

1.5 General Overview of the Study Area

The St. John River watershed from Grand Falls to Hartland is one of the

largest potato-farming regions in eastern Canada. The main study sites used in

the monitoring portion of this project were along the Little River watershed

(uppermost 47° 09’ 95N 67° 40’ 10W to lowermost site 47° 04’ 85N 67° 42’ 95W),

located north (Figure 1-1). The Little River is categorized as a 4th order river,

which briefly defined indicates its position in the hierarchy of tributaries and is

8

meant to serve as way of objectively classifying watercourses (Hynes 2001). This

river originates in a forested landscape and drains predominantly agricultural

lands in its lower reaches. One of the tributaries of this river, Black Brook,

represents one of the most intensely farmed watersheds in Eastern Canada.

Differing agricultural intensities at sites along the Little River reach will provide a

gradient to assess cumulative effects. Previous work in this watershed (Gray

2003) provided a basis for selecting sites along the gradient of potato cultivation

intensity.

For the seasonal profiles, a relatively unimpacted system was necessary

to allow exploration of an ecological or physiological basis for reproductive

changes observed in previous studies with sculpin (Galloway et al. 2003, Gray et

al. 2005). The upper Kennebecasis River (45° 49’ 37”N, 65° 13’ 9”W) in southern

New Brunswick was selected, as it is primarily fed by groundwater and remains

ice free, allowing for continuous monthly collections year round (May 2003-May

2004) (Figure 1-1).

1.6 Objectives and outline of thesis

Fish in agricultural areas have shown reduced proportions of YOY fish, due to

reproductive dysfunction, increased mortality, or a combination of these factors

(Gray 2003). The main objective of the thesis is to identify the potential

mechanisms associated with the reduced reproductive performance in

agricultural areas. Previously, Gray and colleagues (2005) showed reduced

reproductive performance in the agricultural areas of this watershed, but it

9

remains unclear whether these reduced larval densities in summer were a

function of reproductive dysfunction in adult sculpin or difference in apparent

mortality rates in larval fish in forested and agricultural sites along Little River.

To address this, it was necessary to continue and expand monitoring of

fish populations along a gradient of agricultural inputs (Chapter 2 “Monitoring of

fish populations along a gradient of agricultural inputs in Northwestern New

Brunswick, Canada”). This data set allowed for comparisons over time and better

assessment for periods of risk related to agriculture and corresponding sculpin

population responses.

It remains that a major challenge in interpreting monitoring data is a lack

of basic life history information on growth rates, reproductive rates, mobility,

habitat requirements, and longevity. The objective of the seasonal work was to

assess reference populations of slimy sculpin (Cottus cognatus) to identify the

most appropriate window to assess reproductive integrity (Chapter 3 “Seasonal

patterns of energy storage, energy expenditure, and in vitro gonadal

steroidogenic capacity in slimy sculpin (Cottus cognatus)”). The effects-driven

approach recommends ruling out a physiological or ecological basis for the

changes that have been documented in the study species (Munkittrick et al.

2000). To date, no study has addressed how spring-spawning slimy sculpin

function over the winter and the rate at which recrudescence occurs under ice

cover. This chapter presents the seasonal characterization of energy storage,

energy expenditure, and in vitro gonadal hormone production in slimy sculpin,

and describes its reproductive cycle based on these observed patterns.

10

The seasonal study determined that in fact it is possible to measure

physiological endpoints during reproductive development, and collections

optimized the in vitro steroid production assay for this species. This study also

ruled out a seasonal or ecological basis for the depressions that have been

previously documented at agricultural sites. Additional confirmatory studies were

then needed in forested and agricultural areas with both a spring and fall

spawning species, slimy sculpin and brook trout (Salvelinus fontinalis),

respectively. The potato-growing season in northwestern New Brunswick is from

June-October, and may affect sculpin and trout differently as spawning times are

in the spring for sculpin or in the fall for trout. Data indicate that the peak risk

period may occur in late summer, with chemical application coinciding with

increased thunderstorm activity, rather than overwinter or after spawning

mortality. This information was collected on slimy sculpin, a benthic, spring-

spawning species, but additional comparisons included brook trout (Salvelinus

fontinalis), a pelagic, fall-spawning species in agricultural areas to determine the

differential susceptibility relative to reproductive timing (Chapter 4 “ Impacts of

agriculture on fish with differing reproductive strategies: comparing slimy sculpin

(Cottus cognatus) and brook trout (Salvelinus fontinalis)”). Integrating the

knowledge developed on episodic mortality, reproductive development and

reproductive performance for both species allowed for more complete

assessment of impacts related to agriculture. The objective of the final data

chapter was to compare the reproductive functioning of both spring- and fall-

11

spawning species in agricultural areas to assess the degree of potential impact of

exposure to stressors on reproductive function.

Small-bodied species of fish are becoming more widely used in freshwater

assessment programs because of their abundance and the assumptions that

they reflect local environmental conditions because of increased site fidelity. It is

necessary in many areas to focus on population-level assessments using these

small-bodied species, because of low species richness and/or inconsistency

among species in exposure histories. It is necessary to develop a framework for

interpreting population-level life history information on the basis of risk

assessment objectives. The theoretical part of the thesis will focus on trying to

integrate the available information, identify the data gaps, and recommend an

overall approach to population risk assessment that will be based on information

gathered regarding non-point, multiple stressor discharges. A recent Pellston

workshop on population-level risk assessment suggested that empirical data

collected for population level ecological risk assessment should include

information related to population structure, mortality, sex ratios, distribution, and

movement (Table 1-2). Over the past five years, we have been investigating the

performance of fish populations in potato-growing areas of New Brunswick that

exhibit changes in growth, fecundity and size distributions, effects that are

important in the absence of acute mortalities. For example, in these agricultural

areas, YOY sculpin can be 600% larger than conspecifics born in forested areas.

Further exposure assessment will attempt to estimate the magnitude of releases,

identify possible pathways of exposure and estimate potential exposure. In these

12

studies, whole organism and population-level consequences are important for

determining ecological significance and understanding the acceptability of

changes. These studies have been conducted using a small-bodied benthic fish

species, the slimy sculpin (Cottus cognatus) as it is abundant in these areas,

exhibits site fidelity in a small home range, and lives and feeds on the stream

bottom. Chapter 5 (“Developing population-level ecological risk assessment

framework for small freshwater systems using small bodied fish”) will synthesize

the available data in a population level framework suitable for initiating a risk

assessment process, and identify limitations and philosophical challenges to

developing population level risk assessments for small freshwater systems.

13

Table 1-1 A comparison of stressor-based, effects-based and values-based approaches to environmental assessment

Stressor-

based Effects-based Values-based

Focus Stressor-response pathways and valued ecosystem components

Performance indicators of ecosystem status

Ecosystem uses or benefits

Boundaries Related to development

Related to biological components

Related to human uses

Use of existing data

Library searches

Field studies Use and opinion surveys

Endpoints Follow-up requirements

Traditionally very little

Ongoing monitoring and adaptive management

Opinion surveys

Advantages Are often based on previous assessments and experience

Site-specific focus

Focused on user

Disadvantages Ignores unidentified interactions and cumulative effects

Time and expense of baseline monitoring

Not based on ecosystem properties or responses

Question How do I mitigate potentially important impacts?

What are the factors that are limiting energy flow?

How do I protect the uses that are important?

14

Table 1-2 Biological attributes suggested for use in population level ecological risk assessment by empirical working group of Pellston Workshop on Population Level Ecological Risk Assessment. Population parameters computed from population attributes

Attributes of populations computed from individual attributes

Measurable attributes

Population growth rate Abundance Density

Variance of abundance Age/stage structure Age, size, sex

Sex ratio Individual length

Population attractor (k) Recruitment Size

Fecundity

Egg size

Size or age at maturity

Number of viable offspring

Probability of extinction Survivorship Individual weight

Age/stage at death

Timing of mortality

Time to recovery / extinction Biomass Somatic growth rate

Energy storage Liver size

Condition

Density dependence Spatial distribution Movement/dispersal

Habitat preference Home range

Critical patch size Location (specific time)

Diet Stomach contents

15

Figure 1-1 Map of (A) North America, (B) New Brunswick and (C) the St John River. sampling locations within the St. John River basin, redrawn from topographical maps obtained from the Province of New Brunswick. Inset maps provided by M.A. Gray and Environment Canada.

Chapter 3:

Upper Kennebecasis River

45° 49’ 37”N, 65° 13’ 9”W

Chapter 2:

Nonlethal assessment 2002-2004

Little River, Grand Falls, NB

47°06’00N 67°41’10W

47°05’95N 67°42’00W

47°04’85N 67°42’95W Chapter 4:

Species comparison along St John River

47°06’00N 67°41’10W

Black

Outlet

Monquart

Muniac

Shikethawk

N

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

St John River

A

B

C

16

1.7 References

Barrett, JC, GD Grossman, and J Rosenfeld. 1992. Turbidity-induced changes in

reactive distance of rainbow trout. Trans. Am. Fish. Soc. 121: 437-443.

Barnthouse, LW, Munns, WR Jr, Sorenson, MT. 2007. Population-Level

Ecological Risk Assessment. Taylor Francis-CRC Press. 346 p.

Battaglin, W, Fairchild, J. 2002. Potential toxicity of pesticides measured in

Midwestern streams to aquatic organisms. Water Sci. Tech. 45, 95-103.

Böhlke, JK. 2002. Groundwater recharge and agricultural contamination.

Hydrogeology Journal 10,153–179.

Bouwer, H. 1990. Agricultural chemicals and groundwater quality. J. Soil Water

Conserv. 45, 184-189.

Cairns, DK. (Ed.). 2002. Effects of land use practices on fish, shellfish, and their

habitats on Prince Edward Island. Can. Tech. Rep. Fish. Aquat. Sci. No. 2408.

157 pp.

Chambers, PA, J DuPont, KA Schaefer and AT Bielak. 2002. Effects of

agricultural activities on water quality. Canadian Council of Ministers of the

17

Environment, Winnipeg, Manitoba. CCME Linking Water Science to Policy

Workshop Series. Report No. 1.

Chapman, DW. 1988. Critical review of variables used to define effects of fines in

redds of large salmonids. Trans. Am. Fish. Soc. 117:1-21.

Chow, TL, Rees, HW. 1995. Effects of coarse-fragment content and size on soil

erosion under simulated rainfall. Can. J. Soil Sci. 75, 227-232.

Clark, GM, Goolsby, DA, Battaglin, WA. 1999. Seasonal and annual load of

herbicides from the Mississippi River basin to the Gulf of Mexico. Environ. Sci.

Tech. 33, 981-986.

Culliney, TW, Pimentel, D, Pimentel, MH. 1992. Pesticides and natural toxicants

in food. Agriculture Ecosystems Environ. 41, 297-320.

Diamond JM, Serveiss, VB. 2001. Identifying sources of stress to native aquatic

fauna using a watershed ecological risk assessment framework. Environ. Sci.

Technol. 35, 4711-4718.

Dubé M, Munkittrick KR. 2001. Integration of effects-based and stressor-based

approaches into a holistic framework for cumulative effects assessment in

aquatic ecosystems. Human. Ecol. Risk. Assess. 7, 247-258.

18

Environment Canada. 2001. Threats to Sources of Drinking Water and Aquatic

Ecosystem Health in Canada. National Water Research Institute, Burlington,

Ontario. NWRI Scientific Assessment Report Series No. 1. 72 p.

Galloway, B.J., Munkittrick, K.R., Currie, S. Gray, M.A., Curry R.A. Wood, C.S.,

2003. Examination of the responses of slimy sculpin (Cottus cognatus) and

white sucker (Catostomus commersoni) collected on the St. John River

(Canada) downstream of pulp mill, paper mill, and sewage discharges.

Environ. Toxicol. Chem. 22, 2898-2907.

Gormley, KL, Teather, KL, Guignion, DG. 2005. Changes in salmonid

communities associated with pesticide runoff events. Ecotoxicology. 14: 671-

678.

Gray, MA, KL Teather, J Sherry, M McMaster, M Hewitt, and RE Mroz. 2002a.

Potential endocrine disruption in freshwater systems near agricultural areas on

Prince Edward Island. In Effects of land use practices on fish, shellfish, and

their habitats on Prince Edward Island. Cairns D.K. (ed). Can. Tech. Rpt. Fish.

Aquat. Sci. No. 2408. pp. 116-118.

19

Gray, MA, RA Curry and KR Munkittrick. 2002b. Non-lethal sampling methods

for assessing environmental impacts using a small-bodied sentinel fish

species. Water Quality Res J Can 37: 195-211.

Gray, M.A., 2003. Assessing non-point source pollution in agricultural regions of

the upper St. John River basin using the slimy sculpin (Cottus cognatus). PhD

thesis. University of New Brunswick, Fredericton, N.B.

Gray, M.A. Munkittrick, K.R., 2005. An effects-based assessment of slimy sculpin

(Cottus cognatus) populations in agricultural Regions of Northwestern New

Brunswick Water Quality Res. J. Can. 40, 16-27.

Ferenc SA, Foran JA. 2000. Multiple stressors in ecological risk and impact

assessment : approaches to risk estimation. SETAC Press, Pensacola, FL.

264 p.

Karr, JR. 1999. Defining and measuring river health. Freshwater Biol. 41, 221-

234.

Landis, WG and Ming-Ho Yu. 1999. An Introduction to toxicity testing. In

Introduction to Environmental Toxicology: Impacts of chemicals upon

ecological systems. CRC Press, Boca Raton, FL. pp.21-53.

20

Munkittrick, KR, M McMaster, G Van Der Kraak, C Portt, W Gibbons, A Farwell

and M Gray. 2000. Development of Methods for Effects-Based Cumulative

Effects Assessment Using Fish Populations: Moose River Project. SETAC

Press, Pensacola, FL. 236 pp.

Mutch, JP, MA Savard, GRL julien, B MacLean, B Raymond, and J Doull. 2002.

Pesticide monitoring and fish kill investigations on Prince Edward Island, 1994-

1999. In D.K. Cairns (ed.). Effects of land use practices on fish, shellfish, and

their habitats on Prince Edward Island. Can. Tech. Report. Fish. Aquat. Sci.

pp. 94-115.

Napier, TL, Brown, DE. 1993. Factors affecting attitudes toward groundwater

pollution among Ohio farmers. J. Soil Water Conserv. 48, 439-439.

Pimental D, Harvey, C, Resosudarmo, P, Sinclair, K, Kurz, D, McNair, M, Crist,

S, Shpritz, L, Fitton, L, Saffouri, R, Blair, R. 1995. Environmental and

economic costs of soil erosion and conservation benefits. Science. 267, 1117-

1123.

Potter, KM, Cubbage, FW, Blank, GB, Schaberg, RH. 2004. A watershed-scale

model for predicting non-point pollution risk in North Carolina. Environmental

Management. 34, 62–74.

21

Redding, JM, CB Schreck, and FH Everest. 1987. Physiological effects on coho

salmon and steelhead of exposure to suspended solids. Trans. Am. Fish. Soc.

116:737-744.

Skinner, JA, Lewis, KA, Bardon, KS,Tucker, P, Catt, JA, Chambers, BJ. 1997. An

overview of the environmental impact of agriculture in the U.K.. J of Env Man

(1997) 50, 111–128.

Sowden, TK and G Power. 1985. Prediction of rainbow trout embryo survival in

relation to groundwater seepage and particle size of spawning substrate.

Trans. Am. Fish. Soc. 114:804-812.

Waters, TF. 1995. Sediment in streams: sources, biological effects, and control.

Am. Fish. Soc. Monogr. No. 7., Bethesda, MD.

Welch HE, Symons PEK, Narver DW. 1977. Some effects of potato farming and

forest clearcutting on small New Brunswick streams. Technical Report 745.

Fisheries and Marine Service, St. Andrews, NB, Canada.

22

2 Monitoring of fish populations along a gradient of agricultural inputs in

northwestern New Brunswick, Canada

2.1 Abstract Non-point discharges, such as agricultural runoff, are often complex mixtures of

chemical and non-chemical stressors characterized by concentrations of

chemicals which are difficult to distinguish, and rates and timing of discharges

that are difficult to predict. An effects-based approach was used to examine

population-level endpoints including survival and reproduction of slimy sculpin

(Cottus cognatus) in the potato farming belt of New Brunswick. Fish were

nonlethally sampled monthly along a gradient of agricultural intensity over a three

year period. These data were analysed in the context of publicly available

information regarding temperature and precipitation. Results indicate that both

adult and young of the year (YOY) fish are longer and heavier in the downstream

sites as reported in previous studies. Temperature does not appear to be playing

a role in the mortality and changes in population dynamics seen from 2002-2004.

Precipitation, expressed as total rainfall July-August, has a significant negative

relationship with % YOY in the agricultural areas but not in the upstream forested

area. In conclusion, this study provides preliminary evidence that YOY sculpin

may be more impacted in the agricultural areas in years of heavier summer rains.

This study expands the existing knowledge base and development of nonlethal

methods to define cause-effect relationships and an examination of potential

solutions to the issues identified.

23

2.2 Introduction Non-point source pollution can negatively impact aquatic biota by altering

physical habitat, modifying seasonal water flow, altering the systemic food base,

contaminating water with toxic chemicals, and modifying interactions among

organisms (Karr 1999, Potter et al. 2004). Resident biota of aquatic ecosystems

serve as continuous monitors of the cumulative effects of these multiple stressors

on those systems, and are often used as endpoints in environmental

assessments (Munkittrick et al. 2000, Diamond and Serveiss 2001). Research on

the impacts of agriculture on ecosystem health have primarily focused on single

measures or stressors such as soil erosion (Chow and Rees 1995, Pimental et

al. 1995), pesticide use (Culliney et al. 1992, Clark et al. 1999, Battaglin and

Fairchild, 2002), or contamination of ground water by fertilizer and other

agricultural chemicals (Bouwer 1990, Napier and Brown 1993, Böhlke 2002).

While these are important aspects of ecosystem condition, these measures

provide only a partial picture of the integrated impact of agricultural stressors.

Integration of stress effects on ecosystem health may be more efficiently

assessed by addressing the integrated responses of upper ecosystem-level

indicators (Munkittrick et al. 2000).

There have been a number of studies of responses to agricultural

activities along the Little River watershed, located north of Grand Falls, NB,

Canada (Figure 2-1). Recent studies have documented impacts on slimy sculpin

(Cottus cognatus) populations at multiple locations along a gradient within this

watershed (Gray et al. 2002), in a comparative study at multiple rivers with

agricultural gradients (Gray and Munkittrick 2005) and at single locations within

24

21 different watersheds (Welch et al. 1977, Gray et al. 2005). The impacts

included increases in growth and condition; decreases in liver and gonad size,

fecundity, nest density and nest size; and decreases in densities of young-of-the-

year and adult sculpin. This latter study also documented numerous stressors

that may be responsible, including changes in temperature, nutrients and

sedimentation (Gray et al. 2005), but was not successful at teasing out the

relative importance of the various stressors. To address the identified research

gaps, the present study aimed to identify the timing of mortality events and

further investigate reductions in fish numbers in agricultural areas. Systematic

sampling of fish densities and size structure was conducted through multiple

years to identify peak mortality periods in an attempt to identify the timing of

mortality and potential stressors associated with the suspected periods of

mortality.

Growth and mortality are important population-level dynamics influencing

the ecology of fish populations by directly influencing the role of individuals within

the community and interactions among species, especially in size-structured

populations (Werner and Gilliam 1984). Sculpin populations were monitored as

the fish community structure is limited in the potato farming belt of Northern New

Brunswick, with a maximum of three species present at most sites (brook char

(Salvelinus fontinalis), slimy sculpin and brook stickleback (Culea inconstans);

R.A. Curry & K.R. Munkittrick, unpublished data). This project emphasized non-

lethal sampling approaches (as in Gray et al. 2002) because of the design of

repeat sampling over multiple sites, and the limited mobility of sculpin in these

25

areas (Gray et al. 2004, Cunjak et al. 2005, Keeler 2006). The life history of

sculpin also enhances its desirability as a sentinel for these studies. Sculpin are

a spring spawning species and eggs are laid approximately the same time as

potatoes are seeded in late May. These fish are benthic in nature and are in

continuous contact with the stream or lake bottom.

The main objective of this study was to examine whether the previously

observed reductions in density of YOY sculpin in potato-farming areas (Gray et

al. 2003) is related to reproductive dysfunction in adult sculpin or poor survival of

those fish after hatching. Efforts focused on expansion of the existing knowledge

base, the development of methods to define cause-effect relationships and an

examination of potential solutions to the issues identified.

2.3 Methods The main study sites used in this project are along an increasing gradient of

potato farming in the Little River watershed (D/S Ten Mile 47° 09’ 95N 67° 40’

10W; D/S Donat 47° 05’ 95N 67° 42’ 00W; D/S of Dead River (Dead) 47° 04’ 85N

67° 42’ 95W), located north of Grand Falls, New Brunswick, Canada (Figure 2-1).

The Little River is a 4th order river that originates in a forested landscape and

drains predominantly agricultural lands in its lower reaches. It is a tributary of the

St. John River, and the St John watershed from Grand Falls to Hartland is one of

the largest potato-farming regions in eastern Canada. One of the tributaries of

this river, Black Brook, represents one of the most intensely farmed watersheds

in Eastern Canada and is routinely monitored by Agriculture Canada for crop

management and improvement (Chow et al. 2000, Rees et al. 2002). Previous

26

work in this watershed (Gray 2003) provided a basis for selecting sites along the

gradient of potato cultivation intensity.

Slimy sculpin were collected by sampling shallow (approx. 0.50-0.75 m

deep), faster runs and riffles (approx. 1.1-1.5 m/s) with boulder/cobble substrates

using a backpack electrofisher (Smith-Root type VII) and dip nets (6-mm mesh

size). Collections in the first field season targeted the first 100 sculpin caught.

Upon review of the 2002 data, in subsequent collections, sampling continued

until a minimum of 100 adult fish were caught to increase the resolution of size

frequencies in the older age classes after YOY emergence. YOY were still

collected and measured. Barrier nets were not installed, as we have previously

found no significant differences with sculpin collection in open versus closed sites

using one sweep through an area (Gray 2003). Nonlethal sampling of all fish

involved species identification and measurements of fork length, or total length

for sculpin (± 1mm), and weight (± 0.01g). All fish were then released back into

the site from where they were collected. Size frequency data were used to

examine age distributions and condition factors for the fish (Gray et al. 2002).

Temperature recorders (12-bit, Minilog-TR, Vemco Limited, Shad Bay,

NS, Canada) were placed at each site to record hourly water temperatures.

Temperature was recorded beginning in May each year following spring runoff

until mid-October to encompass the period of potential growth for YOY sculpin,

from the time of approximate emergence from the nests to the end of growth for

the first growing season. Degree days were calculated as sum of mean daily

temperature between 27 July-18 October for each reported.

27

Normality and homoscedasticity were assessed by visual examination of

normal probability and residual plots, respectively. YOY were discriminated by

plotting length-frequency distributions for each site. The relationships between

mean daily temperature fluctuation and total rainfall, and YOY body size and

density were assessed using linear regression. Statistical analyses were

completed using Systat© (v. 9, SPSS, Chicago, IL, USA). Length-weight

relationships were analyzed using ANCOVA.

2.4 Results

Physical data including precipitation (Table 2-1) and temperature (Table

2-2) were obtained and presented for each sampling year. Length distribution

data were analyzed monthly for each site by generating frequency histograms

based on estimated catch for a standard amount of electrofishing effort. These

data were assessed visually to determine YOY size and growth and summarized

(Table 2-3).

At all three sites, the YOY size class was easily distinguished, appearing

in August, and increasing in size over time, as described previously (Gray et al.

2002). At the lower agricultural site, August YOY sculpin were significantly

longer than upstream sites in all three years, but had similar condition compared

to reference site fish (Table 2-4).

The proportion of the population composed of YOY increased at the DS

TenMile site between August and September, but remained stable during the rest

of the fall period in both 2002 and 2004. At DS Dead, the lowermost agricultural

28

site, YOY made up a larger portion of the population in 2002 and 2003 compared

to both upstream sites, but not in 2004 (Table 2-3). There were no consistent

differences in condition or fish size early in the year (Table 2-4). On average,

between July and October, average adult fish length and weight increased 8.3%

(4.5) and 16.7% (9.7) upstream, 10.2 (2.1) and 20.7 (13.5) in the middle reach

and 13.3 (3.1) and 37.7 (7.3) downstream (Table 2-4). Temperature units were

on average about 7% warmer downstream compared to the DS TenMile site,

which is approximately 5 km from the lowermost site.

Across all years sampled, the abundance of fish at the forested site was

stable, ranging from 200 to 600 sculpin (per 10000 s electroshocking), with an

29

average (356) similar to that found at the middle reach (349)

(

Figure 2-2). The mainstem site downstream of the agricultural area had

the highest average density (432), and the most variability (Figure 2-3). There

was much less variability at the forested site in the study (Ten Mile) indicating

30

stability of the fish populations

(

31

Figure 2-2). The population in the middle reaches

(

Figure 2-2B) is similar to the forested site, but the agricultural site

demonstrated surges of YOY following emergence in 2002 & 2003, but not in

2004.

32

There was no apparent differences among sites when densities are considered for YOY (Figure 2-3A) or larger fish (fish>70 mm) (Figure 2-3B) and when combined (Figure 2-3C) due to very high variability in total fish caught in the agricultural area. In August 2002, the YOY comprised 72% of all fish collected in the lowest agricultural site (Table 2-3 and

Figure 2-2). Following initial emergence in July, proportion of YOY remained high

at 0.80 and 0.69 for September and October, respectively. At the most upstream

33

forested site, Ten Mile, August YOY proportions were low at 24 %, but upon

subsequent sampling, the proportion more than doubled, suggesting the

presence of late emerging and/or slower growth at this site.

At Ten Mile and Donat, YOY densities were stable or increasing during the

fall, but at the downstream site, the major peak appears usually in August, and

then rapidly declined over the fall. Among larger fish, upstream numbers of

adults are relatively stable year round at upstream sites, and at all sites, large

sculpin (>70 mm) remain relatively constant from November to May, suggesting

that overwinter mortality of larger individuals is not occuring. The more

consistent decrease in density of larger fish is between July and August (2002) or

July and October (2003), suggesting that the peak mortality for both larger and

smaller fish coincides with the period of warmest water temperatures and highest

rainfall events in late summer.

The YOY are larger in the sites downstream of agricultural inputs, with

median sizes of 29 and 30 mm versus 25 mm in the upstream site (Table 2-3). In

2003, monthly sampling was hindered by heavy summer rainfall, as a result YOY

data are only available for August and October. Degree days do not vary along

the gradient; however the most upstream site, Ten Mile, had lower temperatures

across all years (Table 2-2). Maximum daily temperature (27 July-18 Oct) and

fish density were considered (Figure 2-4) and there was no significant

relationship for either forested sites (p=0.49) or agricultural sites (p=0.30).

Precipitation, reported as total rainfall, was lowest in 2002 at 639.2 mm

annually; with 481 mm falling while fields were in use (Table 2-1). As the wettest

34

year of this study, 939.9 mm of rain fell in 2003, 38% of which occurred in July

and August, with a total of 734 mm in the cropping season. The final year, 2004,

was more moderate with regard to precipitation, with 706.3 mm falling annually

and 529 mm during the cropping season. These data were summarized as total

rainfall in late summer (July-August) and regressed against % YOY for both

forested and agricultural samples. Although there was no relationship found for

forested sites (p=0.80), this relationship was significant in agricultural sites

(p=0.039) (Figure 2-5). To follow the significant relationship in precipitation, major

rainfall events were defined as storms that exceeded 15 mm of total precipitation.

Correlating % YOY and number of major storm events resulted in a significant

relationship (Figure 2-6).

2.5 Discussion In the agricultural reach, sculpin showed increased growth, density and

increased variability relative to upstream reaches. Consistent with previous

studies, YOY sculpin downstream of agricultural inputs are longer and heavier

than those of upstream non-agricultural sites (Gray et al., 2002; Gray and

Munkittrick, 2005; Gray et al., 2005). Previous studies indicated that sculpin

populations within the Little River are distinct, based on stable isotopic signatures

for carbon and nitrogen (Gray et al., 2004).The previous studies did not find

major differences between sampling sites on this river (Gray et al., 2002), a

comparison at multiple sites along three rivers (Gray and Munkittrick, 2005), or at

single sites examined on more than 20 different rivers (Gray et al., 2005). The

current study was nondestructive and did not measure internal organ weights,

35

but previous studies (Gray et al. 2005) have shown smaller livers and gonads,

and reduced fecundity in the agricultural reaches. The agricultural site used in

the current study appears to be relatively unimpacted relative to previous reports.

Gray et al. (2005) observed more than 80% of agricultural tributaries did not have

successful reproduction. However the current study was conducted in the

mainstem of Little River and damage may be more extensive in the tributaries.

In comparison to previous studies, this study did not find reduced

abundance of YOY in agricultural areas, but did record increased variability.

Additionally, the previous studies recorded decreased proportion of YOY at

agricultural sites (Gray et al., 2002, 2005). This study showed higher proportions

of YOY in the population for two of the three years.

This study demonstrated a significant negative relationship of rainfall and

proportion of the population comprised of YOY; sculpin year classes are

significantly lower during wetter years. This relationship was not seen at forested

sites. Longer term study in the southeastern United States demonstrated that

over a 10 yr period, abundances of most species either increased or remain

unchanged during low flow and, in fact, mortality from high flow events had a

stronger impact on population size than stresses imposed by low flow (Grossman

et al. 1998).

The data also indicate a significant relationship with the number of major

rain events in July and August (precipitation exceeding 15 mm). In addition, YOY

constituted 30% of the fall population during the three summers with a significant

number of storms >25 mm, compared to 62% for three summer time periods

36

without storms > 25 mm. Research related to soil loss and erosion on these

systems indicates that the majority of soil loss is mobilized by high intensity

thunderstorms during the growing season (Chow et al. 2000, Rees et al. 2002).

These summer thunderstorms coincide with both chemical applications and the

period of major acute fish kill events in potato-growing areas (Cairns et al. 2002,

Hewitt et al. unpublished data).

Temperature is higher in agricultural areas (Gray et al., 2004; Gray and

Munkittrick, 2005), but in this study, the median size of YOY was not different

between sites along the gradient. Gray et al. (2004) reported much higher

temperatures in agricultural sites in this watershed and in those cases, sculpin

were affected or entirely absent. Temperature is recognized as a major

ecological factor affecting the development of freshwater species (e.g. Vannote

and Sweeney 1980) and is thought to influence the density of fish populations

through growth and fecundity (Lobón-Cerviá and Rincón 1998). The relationships

between water temperature, fish growth, and recruitment success have received

considerable attention (Mann et al. 1984). Temperature controls the rate of food

consumption and metabolism, and thus fish growth (Nunn et al. 2003).

Additionally, a recent European study of bullhead (Cottus gobio) suggests that

the distribution of populations and individuals was first structured by the suitability

of physical habitat and hydraulic conditions, and then population dynamics were

mainly governed by the thermal regime (Legalle et al. 2005).

At the two upstream stations, YOY densities are stable or increasing during the fall, but at the downstream site, the major peak appears usually in August, and then shows rapid declines over the fall. Prior to this study, August densities at the same sites were between 400 and 500 / 10,000 s (Gray et al., 2002), very

37

similar to 2002, but lower densities were recorded in 2004. The density of large sculpin was high in 2002, after consecutive dry years

(

Figure 2-2), but 2003 and 2004 had very wet summer storms, and two

years with lower survival characterized by major summer storms (5 in 2003 and 4

in 2004).

38

Among larger fish, upstream numbers of adults are relatively stable year

round at upstream sites, and at all sites, large sculpin (>70 mm) remain relatively

constant from November to May, suggesting that overwinter mortality of larger

individuals is not occuring. The major consistent decrease in density of larger fish

is between July and August (2002) or July and October (2003), suggesting that

the peak mortality for both larger and smaller fish coincides with the period of

warmest water temperatures and highest rainfall events in late summer.

The study has confirmed previous findings of faster growth, and larger

sizes of sculpin in agricultural areas, but found that year class strength and larval

survival were dependent on the number and severity of summer rainfall periods.

Although this study contributes novel information on sculpin population changes

within and between years, more information is needed on the reproductive timing

and population dynamics in an unimpacted system. This would help establish if

there is an ecological basis for the effects that have been documented in

agricultural areas. Although the standardized nonlethal sampling design was

improved over previous studies to isolate periods of risk to fish populations, the

data remain difficult to interpret. Condition factor shows some differences in

energy storage between sites, but no clear patterns emerge. Additional

ecological studies are also needed to better understand the response patterns

observed in slimy sculpin.

Streams in potato dominated landscapes are subjected to a complex array

of stressors associated with increased water temperatures, increased nutrients

from fertilizers, increased sediment loading, increased runoff associated with

39

storm events, and increased chemical exposures associated with pesticide,

fungicide and herbicide applications. The peak risk period appears to be late

summer, and although slimy sculpin are showing higher mortality during these

periods, they may not be the most sensitive fish species for evaluating impacts.

Their spawning period is completed before summer storms, chemical

applications and warm temperatures, and peak exposures to stress occur during

a period of reproductive inactivity. In New Brunswick, other fish species present

are relatively limited, including brook stickleback (Culea inconstans) and brook

trout (Salvelinus fontinalis), and in warmer downstream areas there are also

blacknose dace (Rhinichtys atratulus) (Curry et al. unpublished data). It would

be important to evaluate these other species for potential impacts.

40

2.6 References

Battaglin, W, Fairchild, J. 2002. Potential toxicity of pesticides measured in

Midwestern streams to aquatic organisms. Water Sci. Tech. 45, 95-103.

Böhlke, JK. 2002. Groundwater recharge and agricultural contamination.

Hydrogeology Journal 10,153–179.

Bouwer, H. 1990. Agricultural chemicals and groundwater quality. J. Soil Water

Conserv. 45, 184-189.

Chow, TL, Rees, HW. 1995. Effects of coarse-fragment content and size on soil

erosion under simulated rainfall. Can. J. Soil Sci. 75, 227-232.

Chow, T. L., Rees, H. W. and Monteith, J. 2000. Seasonal distribution of runoff

and soil loss under four tillage treatments in the upper St. John River valley

New Brunswick, Canada. Can. J. Soil Sci. 80: 649–660.

Clark, GM, Goolsby, DA, Battaglin, WA. 1999. Seasonal and annual load of

herbicides from the Mississippi River basin to the Gulf of Mexico. Environ.

Sci. Tech. 33, 981-986.

Culliney, TW, Pimentel, D, Pimentel, MH. 1992. Pesticides and natural toxicants

in food. Agriculture Ecosystems Environ. 41, 297-320.

41

Delong, MD, Brusven, MA. 1998. Macroinvertebrate community structure along

the longitudinal gradient of an agriculturally impacted stream. Environ

Management. 22, 445-457.

Diamond JM, Serveiss, VB. 2001. Identifying sources of stress to native aquatic

fauna using a watershed ecological risk assessment framework. Environ. Sci.

Technol. 35, 4711-4718.

Gray, MA, RA Curry and KR Munkittrick. 2002. Non-lethal sampling methods for

assessing environmental impacts using a small-bodied sentinel fish species.

Water Quality Res J Can 37: 195-211.

Gray, M.A., 2003. Assessing non-point source pollution in agricultural regions of

the upper St. John River basin using the slimy sculpin (Cottus cognatus). PhD

thesis. University of New Brunswick, Fredericton, N.B.

Gray, M.A. Munkittrick, K.R., 2005. An effects-based assessment of slimy sculpin

(Cottus cognatus) populations in agricultural Regions of Northwestern New

Brunswick Water Quality Res. J. Can. 40, 16-27.

42

Grossman, GD, Ratajczak, RE, Crawford, M, Freeman, MC. 1998. Assemblage

organization in stream fishes: effects of environmental variation and

interspecific interactions. Ecol. Monographs. 68, 395-420.

Karr, JR. 1999. Defining and measuring river health. Freshwater Biol. 41, 221-

234.

Landis, WG and Ming-Ho Yu. 1999. An Introduction to toxicity testing. In

Introduction to Environmental Toxicology: Impacts of chemicals upon

ecological systems. CRC Press, Boca Raton, FL. pp.21-53.

Legalle, M. Santoul, F, Figuerola, J, Mastrorillo, S, Ce Re Ghino, R. 2005.

Factors influencing the spatial distribution patterns of the bullhead (Cottus

gobio L., Teleostei Cottidae): a multi-scale study. Biodiversity and

Conservation. 14, 1319–1334.

Lobón-Cerviá J, Rincón, PA. 1998. Environmental determinants of recruitment

and their influence on the population dynamics of stream-living brown trout

Salmo trutta. Oikos, 105 641-646.

Mann, RHK, Mills, CA, Crisp, DT. 1984. Geographical variation in the life-history

of some species of freshwater fish. In Potts, GW, Wootton, RJ (ed.) Fish

Reproduction: Strategies and Tactics pp.171-186. Academic Press, London.

43

Munkittrick, KR, M McMaster, G Van Der Kraak, C Portt, W Gibbons, A Farwell

and M Gray. 2000. Development of Methods for Effects-Based Cumulative

Effects Assessment Using Fish Populations: Moose River Project. SETAC

Press, Pensacola, FL. 236 pp.

Mutch, JP, MA Savard, GRL Julien, B MacLean, B Raymond, and J Doull. 2002.

Pesticide monitoring and fish kill investigations on Prince Edward Island,

1994-1999. In D.K. Cairns (ed.). Effects of land use practices on fish,

shellfish, and their habitats on Prince Edward Island. Can. Tech. Report. Fish.

Aquat. Sci. pp. 94-115.

Napier, TL, Brown, DE. 1993. Factors affecting attitudes toward groundwater

pollution among Ohio farmers. J. Soil Water Conserv. 48, 439-439.

Nunn AD, Cowx, IG, Frear, PA, Harvey, JP. 2003. Is water temperature an

adequate predictor of recruitment success in cyprinid fish populations in

lowland rivers? Freshwater Biol. 48, 579-588.

Pimental D, Harvey, C, Resosudarmo, P, Sinclair, K, Kurz, D, McNair, M, Crist,

S, Shpritz, L, Fitton, L, Saffouri, R, Blair, R. 1995. Environmental and

economic costs of soil erosion and conservation benefits. Science. 267, 1117-

1123.

44

Potter, KM, Cubbage, FW, Blank, GB, Schaberg, RH. 2004. A watershed-scale

model for predicting nonpoint pollution risk in North Carolina. Environmental

Management. 34, 62–74.

Rees, H. W., Chow, T. L., Loro, P. J., Lavoie, J., Monteith, J. O. and Blaauw, A.

2002. Hay mulching to reduce runoff and soil loss under intensive potato

production in northwestern New Brunswick, Canada. Can. J. Soil Sci. 82:

249–258.

Vannote RL, Sweeney, BW. 1980 Geographic analysis of thermal equilibria: a

conceptual model for evaluating the effect of natural and modified thermal

regimes on aquatic insect communities. Amer. Nat. 115, 667–695.

Welch HE, Symons PEK, Narver DW. 1977. Some effects of potato farming and

forest clearcutting on small New Brunswick streams. Technical Report 745.

Fisheries and Marine Service, St. Andrews, NB, Canada.

45

Table 2-1 Precipitation data reported as monthly and annual total rainfall (mm) at the St Leonard airport (St. Leonard, NB, station 6256) (http://www.climate.weatheroffice.ec.gc.ca/climateData/monthlydata_e.html) Storms are listed and enumerated based on the total precipitation to indicate storm duration and intensity. The bolded text indicates the months where potatoes are actively in cultivation. Total Precipitation Storms

>15 mm/>25/>35 mm

2002 2003 2004 2002 2003 2004

January 0.2 0 0

February 18.2 2.2 0

March 22.6 31 23.7

April 59 31.8 66.2 2/0/0 0 1/0/0

May 74 81.8 70.1 0/0/0 0/0/0 0/0/0

June 68.8 103.1 93 0/0/0 2/0/0 3/0/0

July 101.2 225.8 122.3 2/0/0 3/3/3 2/2/1

August 58 135.9 112.5 2/0/0 3/2/0 3/2/0

September 116.8 37.2 80 2/2/0 0/0/0 2/1/0

October 62.2 150.2 51.3 1/0/0 3/2/1 0/0/0

November 31.8 94.2 59.8 1/0/0 4/1/1 3/0/0

December 26.4 46.7 27.4 1/0/0 1/1/1 3/1/0

Total Rainfall 639.2 939.9 706.3

46

Table 2-2 Degree-days (sum of mean daily temperature) for water temperatures collected between 27 July and 18 October for the Little River. Temperature data for 2002 was extrapolated from a regression generated by air: water temperature from 3 other years. Air temperature data collected at the St Leonard airport, (St Leonard, NB) . Station ID 6256. http://www.climate.weatheroffice.ec.gc.ca/

2002 2003 2004

Ten Mile 931.19 939.96 917.54

Donat 997.11 1027.46 965.25

Dead 931.09 1014.83 1024.27

47

Table 2-3 Young of the year (YOY) slimy sculpin data collected from monthly monitoring of Little River along a gradient of agricultural inputs (2002-2004). YOY were first caught in nets in August of all years and sampling was conducted in ice- free months only. All fish were measured and released back into collection areas. POP refers to the proportion of the population sample collected represented by YOY. Size of YOY is reported as median size of the sample. K is the condition factor calculated as (weight/length3)*100000, when length is reported in millimeters.

Ten Mile (Forested) Donat (Intermediate) Dead (Agriculture)

n POP

Median Size (mm)

Body Wt (g) K n POP

Median Size (mm)

Body Wt (g) K n POP

Median Size (mm)

Body Wt (g) K

2002 Aug 21 0.21 25 0.16 1.01 37 0.38 29 0.24 0.93 71 0.72 30 0.17 1.01 Sept 47 0.44 31 0.39 1.15 59 0.61 32 0.37 1.11 80 0.80 34 0.46 1.16 Oct 45 0.45 31 0.38 1.14 47 0.47 36 0.49 1.03 69 0.69 36 0.56 1.08 Nov 47 0.46 34 0.39 0.97 44 0.61 36 0.53 1.07 2003 Aug 47 0.29 27 0.18 0.94 21 0.18 25 0.16 0.97 80 0.45 27 0.26 0.92 Oct 53 0.34 32 0.35 1.02 39 0.31 32 0.34 1.16 88 0.50 36 0.47 0.98 2004 Aug 65 0.43 26 0.14 1.08 20 0.30 27 0.19 1.01 12 0.24 28 0.14 1.07 Sept 72 0.52 31 0.35 1.11 72 0.53 32 0.35 1.05 119 0.59 32 0.36 1.06 Oct 100 0.56 35 0.45 1.05 65 0.40 35 0.45 0.98 74 0.44 35 0.44 0.96

48

Table 2-4 Data collected from monthly sampling of adult slimy sculpin (2002-2004). Sampling was conducted post-freshet and in ice free months only. All fish were measured and released back into collection areas. Data are reported as mean (SEM). Letters (a,b) indicate a significant difference in condition (length-weight relationship) between sites in that month only. † indicates presence of interaction in ANCOVA analysis (p<0.05).

Ten Mile (Forested) Donat (Intermediate) Dead (Agriculture) n length (mm) body wt (g) K n length (mm) body wt (g) K n length (mm) body wt (g) K

2002 July 99 60.13 (1.02) 2.49 (0.14) 1.19 (0.03)a 100 57.54 (0.94) 2.48 (0.12) 1.23 (0.02)a 100 59.45 (1.01) 2.49 (0.14) 1.09 (0.02)b Aug 77 56.70 (0.99) 1.90 (0.11) 0.99 (0.03) 63 59.06 (0.99) 2.17 (0.13 0.99 (0.03) 29 61.52 (1.23) 2.50 (0.16) 1.04 (0.01) Sept † 58 60.41(1.27) 2.29 (0.15) 1.09 (0.02) 38 57.42 (1.26) 2.23 (0.18) 1.10 (0.01) 20 61.60 (1.50) 2.80 (0.23) 1.15 (0.02) Oct 54 58.57 (1.43) 2.29 (0.22) 1.01 (0.01) 54 61.17 (1.00) 2.53 (0.14) 1.05 (0.01) 31 64.45 (1.21) 3.07 (0.17) 1.11 (0.02) Nov 53 60.41 (1.34) 2.43 (0.20) 1.00 (0.01) 28 62.21 (1.41) 2.58 (0.18) 1.03 (0.02)

2003 June 101 49.18 (1.41) 1.68 (0.17) 1.10 (0.02)a 101 48.64 (1.23) 1.58 (0.14 1.14 (0.01)a 99 52.08 (1.15) 1.88 (0.14) 1.18 (0.01)b July † 110 54.21 (1.07) 1.80 (0.11) 0.99 (0.01) 101 52.71 (1.06) 1.49 (0.10) 0.88 (0.01) 105 58.35 (1.25) 2.24 (0.19) 0.94 (0.01) Sept 112 57.52 (0.97) 2.00 (0.12) 0.94 (0.01)a 98 58.18 (1.06) 2.01 (0.12) 0.92 (0.01)b 99 63.99 (1.03) 2.65 (0.15) 0.93 (0.01)a Oct 101 58.79 (0.88) 2.20 (0.12) 1.01 (0.01)a 88 58.79 (0.95) 2.19 (0.12) 1.01 (0.01)a 89 65.58 (0.99) 3.25 (0.17) 1.04 (0.01)b

2004 May 100 40.62 (1.05) 0.83 (0.04) 0.99 (0.02)a 97 48.53 (0.95) 1.49 (0.12) 1.10 (0.02)b 99 43.15 (1.27) 1.06 (0.15) 0.99 (0.02)a June 51 50.04 (1.73) 1.93 (0.23) 1.12 (0.02)a 38 55.42 (1.97) 2.11 (0.21) 1.10 (0.03)a 47 55.23 (2.11) 2.18 (0.30) 1.04 (0.02)b July † 92 53.91 (1.03) 2.00 (0.13) 1.15 (0.02) 91 58.53 (1.41) 2.79 (0.18) 1.19 (0.03) 92 51.88 (0.82) 1.70 (0.07) 1.20 (0.03) Aug † 87 57.09 (0.70) 1.98 (0.08) 1.02 (0.01) 47 62.06 (1.64) 2.80 (0.23) 1.06 (0.01) 39 62.08 (1.78) 2.69 (0.26) 1.07 (0.03) Sept 57 62.26 (1.53) 2.86 (0.29) 1.03 (0.01) 98 62.09 (0.99) 2.70 (0.06) 1.04 (0.02) 103 59.24 (0.95) 2.30 (0.15) 1.01 (0.01) Oct 77 62.39 (1.04) 2.57 (0.17) 0.98 (0.01)a 100 66.39 (1.04) 3.14 (0.17) 0.99 (0.01)a 97 62.05 (0.98) 2.47 (0.15) 0.95 (0.01)b

49

Flow

0k 5k

Little River

TenMile

Donat

Dead

St. John River

Grand Falls Little River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Black Brook

A

B

C

Figure 2-1 Map of (A) North America, (B) New Brunswick and the St John River (C) Little River and sampling sites. These sites were chosen along the Little River based on studies conducted by Gray et al. (2005). Location of Black Brook is indicated by the inset box above lowermost field station. rectangle drawn.

50

Figure 2-2 Slimy sculpin population structure for estimated catches of slimy sculpin collected monthly along a gradient of agricultural intensity, A) TenMile (forested), B) Donat (intermediate), C) Dead (Agriculture). Data are shown as estimated catch based on standardized effort across all sites. Data are categorized into bins based on fish total length (mm).

0

200

400

600

800

1000

1200

1400

1600

1800

2000

J A S O N J J A O M J J A S O

YOY <70 70-75 75-80 >80

0

200

400

600

800

1000

1200

1400

1600

1800

2000

J A S O N J J A O M J J A S O

YOY <70 70-75 75-80 >80

0

200

400

600

800

1000

1200

1400

1600

1800

2000

J A S O N J J A O M J J A S O

YOY <70 70-75 75-80 >80

2002 2003 2004

2002 2003 2004

2002 2003 2004

Total fish length (mm)

Total fish length (mm)

Total fish length (mm)

A

B

C

51

Figure 2-3 Estimated catch of slimy sculpin along a gradient of agriculture for A) young of the year (YOY), B) adult sculpin >70mm, C) all size classes of slimy sculpin collected monthly.. Data are shown as means of estimated catch based on standardized effort across all sites each month during 2002-2004.

0

100

200

300

400

500

600

Estim

ated

Cat

ch o

f YO

Y sc

ulpi

n

A

0

100

200

300

400

500

600

Estim

ated

Cat

ch o

f YO

Y sc

ulpi

n

0

100

200

300

400

500

600

Estim

ated

Cat

ch o

f YO

Y sc

ulpi

n

A

0

20

40

60

80

100

120

140

Estim

ated

Cat

ch S

culp

in >

70 m

m B

0

20

40

60

80

100

120

140

Estim

ated

Cat

ch S

culp

in >

70 m

m

0

20

40

60

80

100

120

140

Estim

ated

Cat

ch S

culp

in >

70 m

m B

0

200

400

600

800

1000

1200

1400

Ten Mile (Forested)

Donat(Intermediate)

Dead(Agriculture)

Estim

ated

cat

ch o

f scu

lpin

(all

size

cla

sses

)

0

200

400

600

800

1000

1200

1400

Ten Mile (Forested)

Donat(Intermediate)

Dead(Agriculture)

Estim

ated

cat

ch o

f scu

lpin

(all

size

cla

sses

)

C

52

y = 55.3x - 574.1R2 = 0.15p=0.487

y = 48.1x - 102.9R2 = 0.07p=0.297

0

200

400

600

800

1000

1200

14 15 16 17 18

Maximum daily temperature 27 July-18 Oct

Estim

ated

Cat

ch o

f Scu

lpin

Figure 2-4 Relationship between maximum daily temperature and estimated density of all sculpin based on standardized effort. indicate data from forested sites and ○ indicate data from agricultural sites.

53

Figure 2-5 Linear relationship between total rainfall **** July and August (mm) and percent young of the year (YOY) for sites along the gradient of agriculture.

indicate data from forested sites and □ indicate data from agriculture sites. Line shown is for agricultural sites only. The relationship between percent YOY and YOY abundance (per 10000 s electrofishing time) has an r2 of 0.98. ***

y = -0.0011x + 0.6289R2 = 0.1803

p=0.039 (0.044)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 50 100 150 200 250 300 350 400

Total Rainfall (mm)

YOY

perc

ent o

f tot

al s

ampl

e po

pula

tion

54

Figure 2-6 Relationship between number of summer storms >15 mm in July and August in total precipitation and percent young of the year sculpin. *** change title

y = -2.7267x + 60.032R2 = 0.2899

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12

Number of Storms >15 mm

Perc

enta

ge Y

oung

of t

he Y

ear (

%)

R2 = 0.32

R2 = 0.83

0200400600800

1000120014001600

0 2 4 6 8 10Total Number Storms >25 mm

YOY

Abu

ndan

ce (f

ish

per 1

0000

s)

55

3 Seasonal patterns of energy storage, energy expenditure, and in vitro

gonadal steroidogenic capacity in slimy sculpin (Cottus cognatus)

3.1 Abstract The objective of this study was to characterize the reproductive seasonality of a

wild population of slimy sculpin (Cottus cognatus), a small-bodied, benthic spring

spawning species. We observed and described the gonadosomatic index,

hepatosomatic index, condition factor, and in vitro gonadal production of estradiol

(E2) and testosterone (T) in females and T and 11-ketotestosterone (11-KT) in

males. Seasonal patterns were observed in the reproduction of both male and

female slimy sculpin, with the males initiating gonadal development before the

females. Following a quiescent period in the summer, female sculpin showed an

increase in gonadal hormone production during the fall, which was associated

with increases in gonad and liver sizes, however males experienced a much

shorter resting phase following spawning in May. Elevated production of both T

and 11-KT appears to both signal the initiation of spermatogenesis in September

and contribute to sperm maturation over the winter. Although hormone

production and precise timing of the spawning period will vary with fluctuation in

environmental factors such as temperature and food availability, this study is

important because it is the first characterization of the seasonal reproductive

pattern in slimy sculpin and it describes the patterns of energy expenditure and

energy storage in this species as it prepares for spawning. Sculpin appear to be

unique compared with other temperate fish species because it has a reproductive

56

pattern that means gonadal maturation can occur in water temperatures of <1 °C

in northern areas of its distribution.

3.2 Introduction Freshwater sculpin are often locally abundant and occupy primarily swift-

water, boulder-gravel river habitat. Sculpin (Cottus spp.) represent an important

ecological component of northern temperate systems, but surprisingly little is

known about their reproductive biology (deVlaming et al. 1984). These fish have

been characterized as rock nesters (Natsumeda 2001), with males maintaining

breeding space under rocks where they mate with females (Goto 1998). They are

usually a spring spawning species in northern temperate waters, and are

relatively unique in that they do not start to accumulate gonadal tissue until the

winter and early spring (Gray 2003). They are complete spawners, with fecundity

ranging from 40-200 depending on their size and age. Freshwater sculpin

reproductive strategies do not appear to be similar to other temperate freshwater

species, however, no comprehensive studies have been conducted.

For an organism to survive, it must be able to convert calories contained in

food into energy, to store part of that energy for subsequent use, and to mobilize

it when necessary (Griffin 2000). Maintenance of energy homeostasis is also

important for regulating body weight, which requires a balance between food

consumption and energy expenditure. The typical diet of slimy sculpin is believed

to consist of small benthic invertebrates and fishes (Scott and Crossman, 1998);

however, several studies have documented finding salmonid eggs in the

stomachs of slimy sculpin (Mirza and Chivers 2002). Spring and early summer

57

spawning species commonly use a reproductive strategy where gonadal tissue

develops in the fall with final maturation as water temperatures rise in the spring,

as with white sucker (Catostomus commersoni) (Scott et al. 1984) or gonadal

development occurs quickly in the spring, as with several species of dace

(Galloway and Munkittrick 2005). Freshwater sculpin species such as the

spoonhead sculpin (Cottus ricei) spawn immediately after ice out in the spring

(Gibbons et al., 1998) and slimy sculpin spawn following the spring freshet. This

strategy requires that the bulk of gonad tissue is generated overwinter and under

ice, when water temperatures are at or below freezing, and activity levels are

usually limited. In this regard, sculpin also present an interesting model for

energy storage, as they are actively synthesizing lipid-rich tissue (gonad) under

ice while food sources are limited.

To ensure that a fish is in breeding condition at the appropriate season,

physiological mechanisms must control the timing of gonad maturation (Wootton

1998). Gonadal steroid hormones are an important factor involved in these

physiological processes, and timely and appropriate changes in gonadal

steroidogenesis are necessary for successful reproduction (Ponthier et al. 1998).

It is not known which hormones are responsible for initiating and sustaining the

processes of sperm and oocyte maturation in sculpin. Previous studies have

shown that female slimy sculpin gonad sizes are very small (<2-5% of body

weight) at the beginning of the ice-cover period, and are then very large (30-

40%) prior to spawning time following the spring freshet (Gray 2003). Hormonal

assessment of reproduction is complicated by the small size of freshwater cottids

58

(usually <<10 g) and low blood volume precludes the use of blood samples for

analysis of circulating sex steroids. A protocol has been developed to measure

the in vitro steroid biosynthetic capability of gonadal tissue as a surrogate for

levels circulating in blood (McMaster et al. 1995). Hormones produced by the

gonads interact to regulate the growth and structural integrity of the reproductive

organs, the production of gametes, the patterns of sexual behavior, the

phenotypic differences between the sexes, and the continuation of the species.

To date, no study has addressed how spring spawning slimy sculpin

function over the winter and the rate at which recrudescence occurs under ice

cover. This study presents the seasonal characterization of energy storage,

energy expenditure, and in vitro gonadal hormone production in slimy sculpin,

and describes its reproductive cycle based on these observed patterns.

3.3 Methods Slimy sculpin were collected from the upper Kennebecasis River (45° 49’

37”N, 65° 13’ 9”W), southern New Brunswick (Canada). This section of the river

is primarily groundwater-fed and remains ice free, allowing for continuous

monthly collections year round (May 2003-May 2004). Slimy sculpin were

collected by sampling faster runs and riffles (approx. 1.1–1.5 m/s) approximately

0.5 to 0.75 m deep with boulder/cobble substrates, between the hours of 10:00

and 13:00. Sculpin were collected with dipnets (1.2 m, 6-mm mesh size) and a

backpack electrofisher unit (Smith-Root type VII). Collections targeted a

minimum of 20 adult males and 20 females per sampling period. Holding time in

59

the cooler did not exceed 4 h, based on results of previous experiments on the

effects of holding time on steroid production (Tetreault 2002). Each adult fish was

rendered unconscious by concussion, followed by spinal severance, and

measured for total length (+0.1 cm), body weight (+0.01 g), gonad weight (+0.001

g), and liver weight (+0.001 g). Gonadosomatic index (GSI) and hepatosomatic

index (HSI) were calculated similarly using the ratio of (organ/(body weight-

organ))*100. Condition factor was calculated as 100000*(body weight/((total

length)^3)) when length is reported in millimetres.

Gonadal tissues were placed in medium 199 (M199; containing Hank’s

salts without bicarbonate; GIBCO, Burlington, ON, Canada which was

supplemented with 25 mM Hepes, 4.0 mM sodium bicarbonate, 0.01%

streptomycin sulfate, and 0.1% bovine serum albumin (pH 7.4)) at 4˚C until

preparation for culture; holding time for gonadal tissue never exceeded 6 h.

Small sections (18-25 mg) of the gonad tissue were placed into M199 in 20-ml

sample tubes on ice. In vitro incubation of the gonadal tissue was conducted in

24-well tissue culture plates (Falcon 3047; Fisher Scientific, Toronto, ON,

Canada). Tissues were subjected to two treatments: basal (M199 alone) or

stimulated (forskolin + M199) (Sigma F6886). Forskolin is a diterpene activator of

the adenylate cyclase pathway which mimics gonadotropin action. This

compound bypasses the gonadotropin receptor which increases cyclic AMP

production and subsequent gonadal steroid production (McMaster et al. 1995).

The level of forskolin-stimulated steroid production provides information

regarding the integrity and maximal capacity of the tissue to produce steroid

60

hormones. Tissues were incubated for 18 h at 16˚C, after which the media from

each of the wells was removed, placed in cryovials, and stored at -80˚C until the

time of analysis.

Hormones were analyzed at the National Water Research Institute

(Burlington, ON). Concentrations of testosterone (T) (both sexes), 17β-estradiol

(E2) (females) and 11-ketotestosterone (11-KT) (males) released into the media

during the incubation period were quantified by radioimmunoassay (RIA) as

described by McMaster et al. (1992). Media samples were assayed in duplicate

at a volume of 100-200 μL for each hormone and values were converted to

correct for size of sub-sample of tissue analyzed, and expressed in pg/mg of

gonadal tissue. For T, E2 and 11-KT, inter-assay variabilities were <10% and

intra-assay variability for each steroid was approximately 5%. T and E2

antibodies were purchased from Medicorp (Prod#07-189016, #07-138016

Montreal, Que., Canada) and radiolabelled T, E2, and 11-KT from Amersham

Pharmacia Biotech (3H-T Prod# TRK 402; 3H-E2 Prod# TRH 322; 3H-11-KT

Prod# TRQ 8945, Baie D’Urfe, Que., Canada). Unlabelled T and E2 were

purchased from Sigma–Aldrich; KT antibody was received from Dr. Glen Van Der

Kraak (University of Guelph, Guelph, ON, Canada) and purchased from Helix

Biotech (0.3024 mg/mL; Vancouver, BC, Canada).

Due to the seasonal aspect of this study, samples in a given month were

only compared to the previous and following month's samples. For example, May

was compared to April and June, but not to other months. Estimates of condition

(weight vs. length), gonad size (gonad weight vs. body weight), and liver size

61

(liver weight vs. body weight) were evaluated using analysis of covariance

(ANCOVA) among adjacent months (SYSTAT v 9.0). Tukey's post hoc test was

then used when p<0.05.

Gonadal tissue weight determined how many replicates were possible,

and average values for in vitro hormone production were obtained from three

replicates when possible. Data are shown for 8-10 fish per sample, to illustrate

the variability as well as the polarization of some individuals. A single mean was

calculated for each month, and we attempted to characterize months when the

fish appeared to be "shut down" or “dormant” with regard to gonadal hormone

production. Due to unequal variances, hormone production data were analyzed

using non parametric comparisons (NCSS© 2004).

3.4 Results Only mature sculpin were kept for analysis during this study. During periods of

low gonadal development, a minimum length of 50 mm was used for sampling to

ensure fish were mature. During periods where gonadal growth was discernable

externally, samples included some smaller, but obviously sexually mature, fish.

Average weights for female fish ranged from 1.7 to 3.3 g, and 2.2 to 5.0 g for

males (Table 3-1). Densities of slimy sculpin in this portion of the Kennebecasis

River can be as high as 8 per m2. There were no detectable differences in size

distributions of the population at this site (n>150) in collections conducted prior to

and following the completion of the study (data not shown).

Male gonad sizes increased rapidly throughout September, reaching their maximum size (2.29 ± 0.07%) in November (

62

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April MayMay June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

Figure 3-1A). Testis size decreased significantly by January (p=0.016), and then continued to gradually decrease until spawning (1.80 ± 0.08%), although the decreases were not statistically significant between adjacent months (p=0.14, 0.8, 0.69, 0.76). Male gonad sizes were at their minimum during July (0.21± 0.04%), however, both T and 11-KT showed significant increases in gonadal in vitro productive capacity during July, compared to June (p=0.028 & p=0.032, respectively) (

63

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April MayMay June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

Figure 3-1). Based on a combination of hormone level and gonadal size, the

reproductive cycle for male sculpin was divided into four “seasons”:

postspawning (June), recrudescence (July-October), maturation (late October to

April) and spawning (May).

64

In vitro steroid synthetic capacity in male testis tissue was highly variable

among individuals, also within a given collection period (i.e. month). Highest

individual values for T were seen during August, immediately preceding the large

increase in gonad size during September and early October. Similarly, males

showed increased synthetic capacity for 11-KT in August, but maximum levels

were not observed until November. This is followed by a significant decrease in

testis weight (p=0.016) and presumably spermatocyte maturation. However,

individual males continued to show low steroid productivity, and there was no

relationship between T production and relative gonad size (r2=0.082), or between

T and 11-KT (r2 =0.40).

In contrast to the male reproductive cycle, female gonad sizes increased

gradually during the winter in preparation for the next spawn (May) (Figure 3-2A).

During the winter, at water temperatures below 1˚C, the ovaries increased in size

from less than 1% in October to more than 12% when water temperatures began

to rise in April. Rapid growth occurred in this tissue between early April and May,

increasing significantly from 12.41 ± .0.54 % to 29.66 ± 0.76 % (p<0.001).

For females, both E2 and T production capacity were diminished during

the collections following the spawning period (Figure 3-2). T capacity remained

low until early February as recrudescence ended (p<0.001). Following this initial

increase, T production remained elevated until spawning. In September, E2

production increased as recrudescence began and stayed elevated in most fish

until late February. E2 production dropped off in the months preceding spawning,

65

with a marked reduction in May (p=0.023). As E2 capacity declined, T production

began to increase starting in early February (p=0.003).

In male sculpin, there was not a significant decrease in condition factor following spawning in May 2003 (p=0.170) (Table 3-1). However, during the season of gonadal growth, there were changes in condition (p<0.0001) (Table 3-1). In November, when both gonad and liver showed significant increases, condition decreased (p<0.0001). Additionally, condition increased more than 10% in the early February collection (p<0.0001), from 0.90 to 1.06 in developing males. Following spawning, liver size in male sculpin was constant, with no detectable change until September (p=0.004) (Figure 3-3). This coincided with a period of rapid testis growth. While testis size reached its maximum in November (

66

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April MayMay June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

Figure 3-1), liver size recovered and began to increase in late fall

(p<0.001), continuing through until March (Figure 3-3).

Unlike males, females exhibited a significant decrease in condition

following the spawn (p=0.022) (Table 3-1). Initially during recrudescence, there

was no change in condition, but condition decreased in late fall (p<0.0001)

67

corresponding with increases in both gonad (p=0.002; Figure 3-2) and liver size

(p=0.031; Figure 3-3). Beginning in February, there was a slight but significant

increase in condition (p<0.0001).

Liver size was dynamic in females, initially increasing after spawn

(p=0.016) and then decreasing until after the onset of recrudescence. Beginning

in October, liver size increased significantly each month, reaching maximum size

in early March. From September to March relative liver size increased by 400%.

During the final phases of gonadal growth, male and female liver sizes dropped

rapidly (p=0.004 & p=0.009, respectively).

3.5 Discussion The assimilation of resources and the management of reserves are

important in the life history of fish and may affect survival and reproductive

success. Energy reserves in the liver, carcass, and body cavity are partitioned to

provide energy requirements for both reproduction and winter maintenance and

survival.

We observed a seasonal pattern in GSI, HSI, condition factor, and

gonadal hormone production in both male and female sculpin. Males have a

much shorter post-spawn/resting period than females, and male recrudescence

begins in July with a significant increase in both 11-KT and T production, but not

in gonad weight. This surge in T production in August may be the cue for the

subsequent increase in testes weight. However, during the period of rapid gonad

growth (Sept-Oct), both T and 11-KT production are near post spawning levels.

68

In November, elevated production of both T and 11-KT appear to trigger a

decrease in testis size, perhaps as gamete maturation progresses. As with the

initiation of recrudescence, the onset of gonadal maturation corresponds with

higher production of both 11-KT and T in the preceding month's collection. Both

11-KT and T production stabilize about 2 months prior to spawning, perhaps

indicating that sperm maturation is complete, and males may be ready to spawn

prior to females.

Decreased steroid production capacity is potentially indicative of reduced

substrate availability, such as cholesterol, or an inhibition or shift in the regulatory

enzymes controlling steroid biosynthesis (McMaster et al. 1995). In some

species, 11-KT is associated with the expression of secondary sex

characteristics, e.g., parental behaviors (Liley et al. 1986, Kindler et al. 1989).

Given the pattern of 11-KT observed, perhaps it also plays a role in the

reproductive behavior in the slimy sculpin. Additionally these data suggest that T

and 11-KT work in conjunction throughout the observed season, however,

additional research is needed to investigate the interplay between these

hormones, as well as the seasonal expression of gonadal enzymes. Stimulation

tests, such as the in vitro assay used in this study, are designed to take

advantage of known endogenous control mechanisms to assess steroidogenic

capacity of the tissue to produce hormone (Griffin 2000). Variability between fish

in vitro steroidogenic capacities in a given month generates questions about

reproductive asynchrony and how microhabitat differences may influence

physiological endpoints. During phases of the reproductive cycle when there are

69

critical changes in gonadal development, the variability in steroidogenesis is high.

This indicates there is not synchrony among fish, or the fish utilize differences in

microhabitat to give themselves benefits in terms of performance. This is a

worthy hypothesis for future study in the slimy sculpin.

For post-spawning females, both T and E2 production appears to be shut

off in August, followed by what may be an increase in aromatase (enzyme that

converts T to E2) in Sept, evidenced by an increase in E2 production after being

"shutdown” from June-August. In February, all the fish have increased E2

production, and T is beginning to increase. Recrudescence in females is initiated

in September and corresponds to a significant increase in E2. T production stays

consistently low until early spring, when there appears to be a shift, perhaps

enzymatic, in gonadal hormone production. In early spring, both T and E2 show

significant increases in hormone production, potentially related to thermal cues

as water temperatures increase. At the time of prespawning, E2 production has

dropped significantly, while T remains elevated. This switch in the steroid

pathway may be to other hormones involved in maturation and ovulation. In white

sucker, during gonadal maturation, E2 production is lessened while 17α-20β-

dihydroxy 4-pregnen-3-one is up-regulated (Van Der Kraak et al. 1992).

In March and April, there is increased energy expenditure (gonads are

growing). The liver appears to be the most dynamic energy source as gonads

develop, although the female sculpin don’t appear to be depleting liver storage as

they build gonad over the winter. This pattern may be indicative of female sculpin

actively feeding during winter for gonadal growth, as liver size increases in a

70

similar pattern until March and the shift into the prespawning stage. Male gonad

growth appears to be fueled by liver stores, and the males could be searching for

nest sites and may not be guarding until May. Maintenance of energy

homeostasis is important for regulating body weight, which requires a balance

between food consumption and energy expenditure. In a previous comparative

study of stable isotopes (Jardine et al. 2005), male slimy sculpin sampled in April

exhibited 15N enrichment, suggesting they may be in poor nutritional condition

prior to spawning in May. Female sculpin in this study did not show significant

15N enrichment suggesting that their prey consumption rates remained constant

throughout the winter. Jardine et al. (2005) suggest this may be a consequence

of cessation or reduction in feeding during the establishment of territories and

nest guarding by males. MacInnis and Corkum (2000) suggest that male sculpin

do not feed while guarding the nest, a contributing factor to the early death of

males after a single spawning cycle. Given this metabolically challenging

environment, is it likely female sculpin have developed a divergent approach to

reproductive and energetic investment.

Assuming that the pattern of allocating time and resources will be related

to the breeding system, we expected to see male sculpin exhibit changes in liver

size and condition during the prespawning period. In this species, males may be

responsible for several of the following behaviors in preparation for spawning:

movement to and choice of a breeding site, preparation of a spawning site,

defence of a spawning site, courtship and parental care.

71

Sculpin have been used recently for freshwater environmental monitoring

related to agriculture (Gray et al. 2002, Gray 2003, Gray et al. 2005), pulp and

paper and sewage effluents (Gibbons et al. 1998, Galloway et al. 2003) as well

as oil sands operations (Tetreault et al. 2003a, Tetreault et al. 2003b). Several

studies have attempted to compare large- and small-bodied fishes to similar

stressors and attribute differences in responses to habitat, life history and

mobility (Gibbons et al. 1998, Galloway et al. 2003). However, sculpin have

consistently presented alternative confounding response patterns and results,

and limited information about their life history and reproductive cycles make the

data difficult to interpret in the context of other species and previous studies.

This study presents the reproductive cycle of both male and female slimy

sculpin which is characterized by distinct seasonal variations in energy storage,

represented by liver size and condition factor, and energy expenditure

represented by gonad size. Data on organ sizes, condition and hormone

production can indicate when males and females are shifting reproductive stages

from pre- and to post-spawning, to recrudescence, and then to final maturation

and will assist in interpreting population data collected in environmental studies.

Sculpin are unique, compared with other temperate fish species, as gonadal

steroid production is more variable and their reproductive pattern reveals that

gonadal maturation can take place in water temperatures below 1 °C in central to

northern Canadian streams.

72

3.6 References deVlaming, V., Fitzgerald, R. Delahunty G. Cech, J.J., Selman, K., Barkley,

M.,1984. Dynamics of oocyte development and related changes in serum

estradiol 17β, yolk precursor, and lipid levels in the Teleostean fish, Leptocottus

armatus. Comp. Biochem. Physiol. 77A, 599-610.

Galloway, B.J., Munkittrick, K.R., Currie, S. Gray, M.A., Curry R.A. Wood, C.S.,

2003. Examination of the responses of slimy sculpin (Cottus cognatus) and white

sucker (Catostomus commersoni) collected on the St. John River (Canada)

downstream of pulp mill, paper mill, and sewage discharges. Environ. Toxicol.

Chem. 22, 2898-2907.

Galloway B.J., Munkittrick, K.R., 2005. Influence of seasonal changes on the

suitability of multiple spawning freshwater fish species for examining reproductive

impacts of stress. Submitted to J Fish Biol.

Gibbons, W.N., Munkittrick K.R., Taylor, W.D., 1998. Monitoring aquatic

environments receiving industrial effluents using small fish species 1: response

of spoonhead sculpin (Cottus ricei) downstream of a bleached-kraft pulp mill.

Environ. Toxicol. Chem. 17, 2227-2237.

Gray, M.A., Curry, R.A. Munkittrick, K.R., 2002. Non-lethal sampling methods for

assessing environmental impacts using a small-bodied sentinel fish species.

Water Quality Res. J. Can. 37, 195-211.

73

Gray, M.A., 2003. Assessing non-point source pollution in agricultural regions of

the upper St. John River basin using the slimy sculpin (Cottus cognatus). PhD

thesis. University of New Brunswick, Fredericton, N.B.

Gray, M.A. Munkittrick, K.R., 2005. An effects-based assessment of slimy

sculpin (Cottus cognatus) populations in agricultural Regions of Northwestern

New Brunswick. Water Quality Res. J. Can. 40, 16-27.

Goto, A., 1998. Life history variation in the fluvial sculpin Cottus nozawae

(Cottidae) along the course of a small mountain stream. Env. Biol. Fishes. 52,

203-212.

Griffin, J.E., 2000. Assessment of Endocrine Function. In: Griffin, J.E., Ojeda,

S.R. (Eds.), Textbook of Endocrine Physiology. Oxford university Press, New

York, pp. 113-127.

Jardine, T.D., Gray, M.A., McWilliam, S.M., Cunjak, R.A., 2005. Stable isotope

variability in tissues of temperate stream fishes. Transactions of the American

Fisheries Society. In press.

Kindler, P.M., Philipp, D.P., Gross, M.R., Bahr, J.M., 1989. Serum 11-

ketotestosterone and testosterone concentrations associated with reproduction in

74

male bluegill (Lepomis macrochirus: Centrarchidae). Gen. Comp. Endocrinol. 75,

446-53.

Liley, N.R., Breton, B., Fostier, A., Tan, E.S., 1986. Endocrine changes

associated with spawning behavior and social stimuli in a wild population of

rainbow trout (Salmo gairdneri). I. Males. Gen Comp Endocrinol. 62, 145-56.

MacInnis, A.J., Corkum, L.D., 2000. Fecundity and reproductive season of the

round goby Neogobius melanostomus in the Upper Detroit River. Trans. Amer.

Fish. Soc. 129,136–144.

McMaster, M.E., Munkittrick, K.R., Van Der Kraak, G.J., 1992. Protocol for

measuring circulating levels of gonadal sex steroids in fish. Can. Tech. Rept.

Fish Aquat. Sci. 1836: 29 p.

McMaster, M.E., Munkittrick, K.R., Jardine, J.J., Robinson, R.D., Van Der Kraak,

G.J., 1995. Protocol for measuring in vitro steroid production by fish gonadal

tissue. Can. Tech. Rept. Fish Aquat. Sci. 1961: 78 p.

Mirza, R.S., Chivers, D.P., 2002. Attraction of slimy sculpins to chemical cues of

brook charr eggs. J. Fish Biol. 61, 532–539.

75

Natsumeda, T., 2001. Space use by the Japanese fluvial sculpin, Cottus pollux,

related to spatio-temporal limitations in nest resources. Env. Biol. Fishes. 62,

393-400.

Ponthier, J.L., Shackleton, C.H.L., Trant, J.M., 1998. Seasonal changes in the

production of two novel and abundant ovarian steroids in the channel catfish

(Ictalurus punctatus). Gen. Comp. Endocrinol. 111, 141-155.

Scott, A.P., MacKenzie, D.S., Stacey, N.E., 1984. Endocrine changes during

natural spawning in the white sucker, Catostomus commersoni. II. Steroid

hormones. Gen. Comp. Endocrinol. 56, 349–359.

Scott, W.B., Crossman, E.J., 1998. Freshwater fishes of Canada. Galt House

Oakville, ON. Canada. 966 p.

Tetreault, G.R. 2002. Monitoring the aquatic environment in the Athabasca Oil

Sands using reproductive responses in small-bodied fish. M. Sc. thesis.

University of Waterloo, Waterloo. Ontario.

Tetreault, G.R., McMaster, M.E., Dixon, D.G., Parrott, J.L., 2003a. Using

reproductive endpoints in small forage fish species to evaluate the effects of

Athabasca Oil Sands activities. Environ Chem. Tox. 22, 2775-2782.

76

Tetreault, G.R., McMaster, M.E., Dixon, D.G., Parrott, J.L., 2003b. Physiological

and biochemical responses of Ontario slimy sculpin (Cottus cognatus) to

sediment from the Athabasca Oil Sands Area. Water Qual. Res. J. Canada. 38,

631-377.

Van der Kraak GJ, KR Munkittrick, ME McMaster, CB Portt and JP Chang.,

1992. Exposure to bleached kraft pulp mill effluent disrupts the pituitary-gonadal

axis of white sucker at multiple sites. Toxicol and Applied Pharmacology 115:

224-233.

Wootton, R.J., 1998. Life-history strategies. In: Wootton, R.J. (Ed.), Ecology of

Teleost Fishes. Fish and Fisheries Series Vol 24. Kluwer Academic, Dordrecht.

pp.259-283.

77

Table 3-1 Monthly values for mean total length and mean body weight of slimy sculpin (Cottus cognatus) collected from May 2003-May 2004. Values are mean ± SEM (N) *indicates significant change from preceding month (p<0.05). Condition factor was calculated as k=(weight/(length)3)*100. * indicates a significant difference (p<0.05) from the preceding month.

Females Males Length (mm) Body Wt (g) Condition

Factor Length (mm) Body Wt (g) Condition Factor

May 61 + 1 (21) 2.63+ 0.24

1.09 ± 0.025 75.67 + 0.33 (3) 5.07 + 0.29

1.17 ± 0.058

June 62+1 (22) 2.47+0.19

0.99 ± 0.015* 67.43 + 1.89 (23) 3.15 + 0.26*

0.98 ± 0.017

July 58+1(21)* 1.94+0.24 * 0.95 ± 0.013 66.95 + 2.45 (21) 3.29 + 0.45 0.99 ± 0.016

August 67+4 (6)* 3.25+0.64* 1.00 ± 0.016 75.17 + 1.74 (6) 3.17 + 0.31 1.03 ± 0.028

September 61+ 1 (22)* 2.18+0.09* 0.96 ± 0.017 65.90 + 1.87 (21)* 3.00 + 0.29* 1.00 ± 0.024

October 61+1 (22) 2.26+0.17 0.96 ± 0.014 67.73 + 2.16 (22) 3.30 + 0.33 0.99 ± 0.017

November 58+1 (23) 1.70+0.10* 0.84 ± 0.015* 68.08 + 1.38 (26) 2.83 + 0.19 0.87 ± 0.01*

January 58+2 (20) 1.87+0.22 0.89 ± 0.013 61.29 + 1.71 (24)* 2.22 + 0.22* 0.90 ± 0.014

55+2 (25) 1.76+0.17 1.01 ± 0.019* 62.17 + 2.09 (18) 2.71 + 0.31 1.06 ± 0.030* February 58+2 (30) 2.12+0.21 1.01 ± 0.014 64.75 + 2.23(24) 2.91 + 0.30 1.02 ± 0.027

March 51+1 (15)* 2.54+0.22 1.03 ± 0.024 53.17 + 2.00 (24)* 3.25 + 0.42 1.07 ± 0.026

April 46+1 (23)* 1.99+0.15* 1.08 ± 0.030 60.10 + 1.54 (30)* 4.56 + 0.38* 1.10 ± 0.022

May 53+1 (55)* 1.95+0.12 1.21 ± 0.01 64.11+ 1.29 (31)* 3.26 + 0.20* 1.19 ± .0.01

78

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

0

0.5

1

1.5

2

2.5

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Maturation

0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April May0

10000

20000

30000

40000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

May June July Aug Sept Oct Nov Jan February March April MayMay June July Aug Sept Oct Nov Jan February March April May

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

40000

50000

In V

itro

11-K

etot

esto

ster

one

Pro

duct

ion

(pg/

ml)

†

‡

†

†

‡

‡

**

**

*

Spawn Spawn

A

B

C

Figure 3-1 Monthly changes (mean monthly values) in male slimy sculpin A gonadosomatic index (gonad weight/body weight – gonad weight *100) B gonadal in vitro steroidogenic capacity to produce11-ketotestosterone C gonadal in vitro steroidogenic capacity to produce testosterone. Solid lines represent seasonal changes in the monthly mean. Dashed lines define "shut off" as defined by the maximum production value in the months of minimum steroidogenic capacity. */‡/† indicates significant (p<0.05) change from the preceding month using Mann Whitney nonparametric probabilities.

79

0

5

10

15

20

25

30

35

40

45

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Pre-spawning

0

5000

10000

15000

20000

25000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

In V

itro

Estra

diol

Pro

duct

ion

(pg/

ml)

‡

‡

‡

†

†

†

* * * * * **

*

* *

Spawn

May June July Aug Sept Oct Nov Jan February March April May

A

B

C

0

5

10

15

20

25

30

35

40

45

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Pre-spawning

0

5000

10000

15000

20000

25000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

In V

itro

Estra

diol

Pro

duct

ion

(pg/

ml)

‡

‡

‡

†

†

†

* * * * * **

*

* *

Spawn

May June July Aug Sept Oct Nov Jan February March April May

0

5

10

15

20

25

30

35

40

45

Gon

ados

omat

ic In

dex

(%)

0

5

10

15

20

25

30

35

40

45

Gon

ados

omat

ic In

dex

(%)

Post-spawning Recrudescence Pre-spawning

0

5000

10000

15000

20000

25000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

0

5000

10000

15000

20000

25000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

0

5000

10000

15000

20000

25000

In V

itro

Test

oste

rone

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

In V

itro

Estra

diol

Pro

duct

ion

(pg/

ml)

0

10000

20000

30000

In V

itro

Estra

diol

Pro

duct

ion

(pg/

ml)

‡

‡

‡

†

†

†

* * * * * **

*

* *

Spawn

May June July Aug Sept Oct Nov Jan February March April May

A

B

C

Figure 3-2 Monthly changes (mean monthly values) in female slimy sculpin A gonadosomatic index (gonad weight/body weight – gonad weight *100) B gonadal forskolin stimulated in vitro steroidogenic capacity to produce 17β estradiol C gonadal in vitro steroidogenic capacity to produce testosterone. Solid lines represent seasonal changes in the monthly mean. Dashed lines define "shut off" as defined by the maximum production value in the months of minimum steroidogenic capacity. */‡/† indicates significant (p<0.05) change from the preceding month using Mann Whitney nonparametric probabilities.

80

May June July Aug Sept Oct Nov Jan February March April May0

1

2

3

4

5

Hep

atos

omat

icIn

dex

(%)

Males

Females*

*

**

**

*

*

*

*

3

May June July Aug Sept Oct Nov Jan February March April May

**

* ***

0

1

2

3

4

5

Hep

atos

omat

icIn

dex

(%)

A

B

May June July Aug Sept Oct Nov Jan February March April May0

1

2

3

4

5

Hep

atos

omat

icIn

dex

(%)

May June July Aug Sept Oct Nov Jan February March April MayMay June July Aug Sept Oct Nov Jan February March April May0

1

2

3

4

5

Hep

atos

omat

icIn

dex

(%)

Males

Females*

*

**

**

*

*

*

*

3

May June July Aug Sept Oct Nov Jan February March April May

**

* ***

0

1

2

3

4

5

Hep

atos

omat

icIn

dex

(%)

3

May June July Aug Sept Oct Nov Jan February March April MayMay June July Aug Sept Oct Nov Jan February March April May

**

* ***

0

1

2

3

4

5

Hep

atos

omat

icIn

dex

(%)

0

1

2

3

4

5

Hep

atos

omat

icIn

dex

(%)

A

B

Figure 3-3 Monthly changes in energy storage in A male slimy sculpin (Cottus cognatus) hepatosomatic index [liver wt/(body wt-liver wt)]*100 B female sculpin hepatosomatic index from May 2003- 2004. * indicates significant (p<0.05) change from the preceding month using Mann Whitney nonparametric probabilities.

81

4 Comparison of spring spawning slimy sculpin (Cottus cognatus) and

fall spawning brook trout (Salvelinus fontinalis) reproductive

development in agricultural regions of the St. John River (New

Brunswick, Canada)

4.1 Abstract Previous research has demonstrated that fish in agricultural areas in New

Brunswick have reduced proportions of YOY fish, due to either reproductive

dysfunction, increased mortality, or a combination of these factors. The objective

of this study was to examine two fish species with different life history strategies

and timing of reproduction. Additional comparisons were needed with a second

species and brook trout (Salvelinus fontinalis), a pelagic, fall spawning species,

were chosen to better determine the differential susceptibility relative to

reproductive timing. Previous studies in these areas indicate that the peak risk

period may occur in late summer, coinciding with trout recrudescence. Pre-

spawning sculpin demonstrated differences in condition factor, LSI, GSI and in

vitro gonadal steroidogenesis among agricultural and forested sites. Trout

responses were highly variable and no consistent patterns were seen in fish

collected at sites near agriculture. It would appear from the data that fall

spawning brook trout are not synchronized for spawning by the same cues as

spring spawning species, thereby making comparisons difficult in this species. In

a study that was designed to identify differences in a reproductive impairment in

fish with differing reproductive strategies, the patterns remain unclear.

82

4.2 Introduction There has been considerable attention given to the impacts of potato

farming on fish populations in Atlantic Canada over the past few years. An

increased frequency of summer fish kills of brook trout (Salvelinus fontinalis) in

potato farming areas of Prince Edward Island has been associated with summer

thunderstorm events (Cairns 2002). In Atlantic Canada, heavy rainstorms in mid

to late summer can coincide with heavier chemical application as the potato vine

grows and is more susceptible to peril (Chow et al. 2000, Rees et al. 2002).

Recently completed studies have compared fish performance in 20 New

Brunswick tributaries (Gray and Munkittrick, 2005), and have shown that slimy

sculpin reproductive performance was very low at agricultural sites. Detailed

studies in northern New Brunswick have demonstrated that slimy sculpin (Cottus

cognatus) in agricultural areas have reduced proportions of young of the year

(YOY) fish, due to either reproductive dysfunction, increased mortality, or a

combination of these factors (Gray et al. 2002, 2004, 2005).

These studies have demonstrated a number of potential causes of stress

in the agricultural areas, including changes in chemicals, and potential

concomitant stress imposed by nutrients, altered habitat, changes in

temperature, runoff, and sediment deposition. The mixtures of chemicals and

non-chemical stressors are complex, difficult to characterize, and the rates and

timing of discharges are difficult to predict (Landis and Yu 1999). Potato

production involves leaving the soil bare for long periods of time, which can lead

to high erosion risk. Sediment is the number one pollutant in streams in North

America (Waters 1995), and can impact fish populations through turbidity,

83

suspended sediments, and deposition of sediments (Sowden and Power 1985,

Redding et al. 1987, Chapman 1988, Barrett et al. 1992).

It is unclear if the decreased abundance of young-of-the-year (YOY) noted

in agricultural areas (Gray and Munkittrick, 2005, Gray et al. 2005) is associated

with altered YOY survival or depressed reproduction. In addition to the difficulties

encountered in trying to separate the potential stressors, there are also

challenges in separating the complex impacts. Detailed studies have

documented year class failures in agricultural areas, and decreases in ovary

size, fecundity, and nest size in populations of slimy sculpin (Gray et al. 2002,

Gray and Munkittrick, 2004). However, a recent study (Chapter 2) has also

identified that the peak mortality period for sculpin occurs in late summer,

coinciding with highest water temperatures and the most intense chemical

applications (>10 applications between July-Sept.). Furthermore, year class

strength in terms of YOY abundance correlates with the number of summer

storms (Chapter 2).

It is difficult to evaluate the contribution of reproductive impairment to the

depressions in YOY because of the life history and reproductive development of

the slimy sculpin. Slimy sculpin spawn in mid May (Keeler and Cunjak 2006),

and hatch in mid to late June (unpubl. data). Gonadal development in females

occurs primarily from November to April (Chapter 3; Brasfield et al., submitted), a

period of no agricultural activity in potato-growing areas, and little pesticide

mobility in the water column (Hewitt et al. 2005). It is not presently known if

pesticides adsorbed to sediments present a route of exposure although extracts

84

from sediments collected from brook trout egg incubators deployed during winter

in Black Brook affected hatching success (Cunjak et al. unpublished). In contrast,

male sculpin undergo most of their gonadal growth between late August and

early October (Chapter 3) and reproductive potential can be assessed in the

early fall.

Previous and continuing studies on the St. John River indicate small-

bodied fish species present more suitable models for effects-driven assessment

(Gray et al. 2002, Galloway et al. 2003, Vallis 2003, Vallieres 2005). Small rivers

in this part of Canada are dominated in colder reaches by a small fish community

of 3 to 4 fish species, including slimy sculpin, brook trout, and brook stickleback

(Culea inconstans) in slower areas, and blacknose dace (Rhinichthys atratulus)

in warmer areas (Curry and Munkittrick, unpublished data). Slimy sculpin have

been utilized as a sentinel species because of their low mobility and high site

fidelity in these systems (Gray et al., 2004; Cunjak et al., 2005). It is possible

that other species may be more sensitive to the impacts of potato farming.

Reproductive development in brook trout occurs during summer period, with

spawning in early fall just after the potato harvest. Brook trout are a fall spawning

species, and therefore are recrudescent during summer periods of increased

runoff events and potentially at greater risk. Comparing the responses of these

two species in agricultural areas may determine the differential susceptibility

relative to timing of reproduction.

This study evaluates the reproductive potential of prespawning slimy

sculpin and brook trout in potato farming and reference areas of Northern New

85

Brunswick. The study focused on whole organism endpoints and in vitro steroid

production. Gonadal steroid hormones are an important regulator of reproductive

development (Ponthier et al. 1998, Wooton, 1998) and laboratory and field

studies have shown that reductions in circulating levels of sex steroids can

indicate exposure to stressors affecting the reproductive system (McMaster et al.

1995). Gonadal steroidogenic capacity can be used for identifying periods of risk

related to agricultural runoff and differential susceptibility of fish populations

warrants comparisons between different species.

4.3 Methods The St. John River from Grand Falls to Woodstock is one of the largest

potato farming regions in eastern Canada. Previous work in this watershed (Gray

et al. 2003) provided a basis for selecting sites along the gradient of potato

cultivation intensity. Fish were collected from the three agricultural and three

reference sites in tributaries along the St. John River, southern New Brunswick

(Canada) during April (prespawning) and September (recrudescence). Paired

forested and agricultural sites were chosen from three different tributaries of the

St John River (Figure 4-1).

Sculpin were collected by sampling run and riffle habitat (approx. 1.1-1.5

m/s) approximately 0.5 to 0.75 m deep with boulder/cobble substrates, while

brook trout were collected from deeper pools and undercut banks. All fish were

collected with dipnets (1.2 m, 6-mm mesh size) and a backpack electrofisher unit

(Smith-Root type VII). Collections targeted a minimum of 10 brook trout adult

females and 20 slimy sculpin adult females per sampling period and site.

86

Guidance for interpreting fish reponses (EEM citation here**) suggests using 20

of each species, however efforts were made to reduce this sample size for trout

because of the recreational importance of the species. Only females were

sampled for this study in an effort to reduce the impact of sampling on

prespawning fish, and because responses of interest were specific to females.

Holding time in the cooler did not exceed 4 h, based on results of previous

experiments on the effects of holding time on steroid production (Tetreault 2002).

Each adult fish was rendered unconscious by concussion, followed by spinal

severance, and measured for total length (± 0.1 cm), body weight (± 0.01 g),

gonad weight (± 0.001 g), and liver weight (± 0.001 g). Gonadosomatic index

(GSI) and hepatosomatic index (HSI) were calculated similarly using the ratio of

(organ/(body weight-organ))*100. Condition factor was calculated as

100000*(body weight/((total length)^3)) when length is reported in millimeters.

Following excision, gonadal tissues were placed in medium 199 (M199;

containing Hank’s salts without bicarbonate; GIBCO, Burlington, ON, Canada

which was supplemented with 25 mM Hepes, 4.0 mM sodium bicarbonate,

0.01% streptomycin sulfate, and 0.1% bovine serum albumin (pH 7.4)) at 4˚C

until preparation for culture; holding time for gonadal tissue never exceeded 6 h.

Small sections (18-25 mg) of the gonad tissue were placed into M199 in 20-ml

sample tubes on ice. In vitro incubation of the gonadal tissue was conducted in

24-well tissue culture plates (Falcon 3047; Fisher Scientific, Toronto, ON,

Canada). Follicles were subjected to two treatments: basal (M199 alone) or

stimulated (forskolin + M199) (Sigma F6886). Forskolin is a diterpene activator of

87

the adenylate cyclase pathway which mimics gonadotropin action. This

compound bypasses the gonadotropin receptor which increases cyclic AMP

production and subsequent gonadal steroid production (McMaster et al. 1995).

The level of forskolin-stimulated steroid production provides information

regarding the integrity and maximal capacity of the tissue to produce steroid

hormones. Follicles were incubated for 18 h at 16˚C, after which the media from

each of the wells was removed, placed in cryovials, and stored at -80˚C until the

time of analysis.

Hormones were analyzed at the National Water Research Institute

(Burlington, ON). Concentrations of testosterone (T) and 17β-estradiol (E2)

released into the media during the incubation period were quantified by RIA as

described by McMaster et al. (1992). Media was assayed in duplicate at a

volume of 200 μL for testosterone and 100 μL for 17β-estradiol and values were

converted to correct for size of sub-sample of tissue analyzed, and expressed in

pg/mg of gonadal tissue. For T and E2, inter-assay variabilities were <10% and

intra-assay variability for each steroid was approximately 5%. T and E2

antibodies were purchased from Medicorp (Prod#07-189016, #07-138016

Montreal, Que., Canada) and radiolabelled T and E2 from Amersham Pharmacia

Biotech (3H-T Prod# TRK 402; 3H-E2 Prod# TRH 322). Unlabelled T and E2 were

purchased from Sigma–Aldrich.

Estimates of condition (weight vs. length), gonad size (gonad weight vs.

body weight), and liver size (liver weight vs. body weight) were evaluated using

analysis of covariance (ANCOVA) among sites (SYSTAT v 9.0). Tukey's post hoc

88

test was used to test for differences between sites when p<0.05. Gonadal tissue

weight determined how many replicates were possible, and average values for in

vitro hormone production were obtained from three replicates when possible.

Data are shown for 8-10 fish per sample. Due to unequal variances, hormone

production data was analyzed using non parametric comparisons (NCSS 2004).

4.4 Results Only adult female fish were sampled in these studies. Efforts were made

to collect 10 mature females of each species at each site. Pre-spawning sculpin

were collected at three agricultural sites in different rivers and three forested sites

in April 2004, while pre-spawning brook trout were only accessible at four of the

six sites sampled in September 2004, perhaps due to the presence of Atlantic

salmon parr in the sites lacking brook trout. No interactions were detected for any

of the endpoints analyzed with treatment (agricultural, reference), thus all sites

were compared to all sites.

Prespawning sculpin exhibited differences in condition factor among sites

(p=0.006), with FOR1 sculpin having higher condition factor relative to AG3

(p=0.007) and FOR2 (p=0.045) (Table 4-1). Energy storage patterns, estimated

based on liver size, were different in fish at the three forested sites (p=0.001).

FOR1 fish had an elevated HSI of 3.94, higher than FOR2 fish with HSI of 2.70

(p=0.001) and FOR3 at 2.94 (p=0.002). Gonad weights were higher at the two

upstream sites, FOR1 and AG1, than in downstream sites, with GSI means of

20.5 and 16.2, respectively (p<0.001). Downstream sites show no difference in

89

energy expenditure

(

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5 6Body Wt (g)

Gon

ad W

eigh

t (g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

0

0.1

0.2

0.3

0 1 2 3 4 5 6Body Wt (g)

Live

r Wei

ght (

g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

p=0.001

FOR1>AG1, FOR2, FOR3

p<0.001

FOR1>AG>all other sites

A

B

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5 6Body Wt (g)

Gon

ad W

eigh

t (g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

0

0.1

0.2

0.3

0 1 2 3 4 5 6Body Wt (g)

Live

r Wei

ght (

g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

p=0.001

FOR1>AG1, FOR2, FOR3

p<0.001

FOR1>AG>all other sites

A

B

Figure 4-2). Gonadal testosterone production was different (p=0.036)

between agricultural and forested areas, however no clear patterns emerged

between sites. Gonadal estradiol production was also different between sites

90

(p=0.037), with AG2 fish producing significantly less than all sites except FOR1

(p=0.032).

There was no difference in condition factor across sites where female

brook trout were collected (p= 0.996,

91

Table 4-1). Energy storage, estimated based on liver size, was also similar

for all sites (p=0.87) sampled. Energy expenditure (GSI) showed no differences

among fish spawning in systems influenced by agriculture (p=0.58). Reproductive

readiness varied within sites, as gonad weights ranging widely (0.05 to 2.5 g) in a

single site. Also, some fish displayed large ripe eggs while others had very small

oocytes (data not shown). Gonadal testosterone production offered ranges of

195-17 470 pg/mg tissue in the two agricultural sites, and 346-16 839 pg/mg

tissue in the forested sites. Variability among fish in a given site ranged several

orders of magnitude. No differences among sites were observed (p=0.603).

Estradiol production was less variable within sites AG1, FOR1, and FOR2,

however AG3 fish expressed steroidogenic capacity ranging from 500-51 907

pg/mg tissue. There were no differences among sites with regard to estradiol

production (p=0.425) in pre-spawning brook trout.

4.5 Discussion

The main objective of this study was to examine potential mechanisms

associated with the reduced reproductive performance in fish near potato

growing areas of New Brunswick, Canada. Previous studies (Gray et al. 2005)

demonstrated that sculpin collected downstream of agricultural inputs had

reductions in gonad size, fecundity, nest size and nest density when compared to

forested reference sites. Similar to previous studies, the sculpin collected in the

Little River showed decreases in liver size and gonad size relative to the

upstream reference site (Gray and Munkittrick 2005, Gray et al. 2005).

92

In this study, impacts on sculpin were not consistent among agricultural

sites. Pre-spawning sculpin demonstrated differences in condition factor, LSI,

GSI and in vitro gonadal steroidogenesis among agricultural and forested sites.

Fish furthest upstream from FOR1 site had increased condition relative to both a

downstream forested site (FOR2) and an agricultural site AG3. This would

indicate that fish were heavier at a given length at the most upstream site, where

ice cover came off at the latest time and water temperatures were lower. Energy

storage patterns, estimated based on liver size, were also different in fish at the

three upstream forested sites, perhaps reflecting different overwintering feeding

strategies. Gonad weights, an indication of energy expenditure, were elevated at

both sites on the tributary furthest upstream, FOR1 and AG1 compared to all

other downstream sites which were all similar. Although previous studies in this

part of New Brunswick demonstrated consistent impacts on reproductive

development at a series of agricultural sites (Gray et al. 2005), more recent

studies have demonstrated that the reproductive impairments are more severe in

summers with lower rainfall (Chapter 2). However, the sculpin collections in this

study occurred in the spring, prior to chemical applications for the current year.

Hormone production by the gonadal tissue was not an endpoint that

allowed for interpretation of impacts, since no differences were detected and

variability spanned orders of magnitude. Additional studies on an unimpacted

system in southern NB indicated that during phases of the reproductive cycle

when there are critical changes in gonadal development, the variability in

steroidogenesis is high (Chapter 3). This may be an indication there is not

93

synchrony among fish, or the fish utilize differences in microhabitat to give

themselves benefits in terms of performance.

This study attempted to further investigate this observation by comparing

the reproductive functioning of spring and fall spawning species in agricultural

areas to assess the degree of potential impact of exposure to stressors on

reproductive function. The potato growing season in northwestern New

Brunswick is typically from June to October (Rees et al. 2002), it was

hypothesized that temporal change in runoff volume and composition would

affect sculpin and trout differently as they spawn in the spring (sculpin) or fall

(trout). In fall 2004, pre-spawning brook trout were collected at four of the six

desired sites where sculpin were collected earlier that same year in April, three

agricultural and three forested in three different tributaries of the St. John River.

Trout responses were highly variable and no consistent patterns were seen in

fish collected at sites near agriculture. It would appear from the fall sampling that

the fall spawning brook trout are not synchronized for spawning by the same

cues as spring spawning species. This makes comparisons among and within

sites difficult as fish are at differing stages of reproductive readiness. Within

treatment variability is higher than expected, and all forested sites are not

showing the same response pattern. In a study that was designed to identify

differences in a reproductive impairment in fish with differing reproductive

strategies, the patterns remain unclear. A study with another salmonid species

indicated environmental variables, such as rainfall and subsequent discharge,

94

affected the availability of habitat, driving the population dynamics and

recruitment (Lobon and Cervia 1999).

Spring spawning species are thought to be synchronized for spawning

with increasing water temperatures, unlike fall spawning species. This is a

potential explanation for the differing reproductive stages. Additionally, brook

trout are highly mobile and migrate for spawning. This may explain why

reproduction in females is unsynchronized, making it more difficult to interpret

responses.

There are many criticisms of current environmental monitoring practices

using a single sentinel species. The main argument is that no two species

occupy the same ecological niche, therefore, no single species should be

expected to act as an indicator for an entire ecosystem (Carignan and Villard

2002). Several species would likely exhibit differential sensitivity to environmental

perturbation and should be monitored in order to identify the causes of change

more precisely and limit errors of interpretation. Additionally, the argument is that

many factors unrelated to the degradation of ecological integrity may affect the

population status of an indicator species. This could complicate the detection and

interpretation of population trends and demonstrates the benefit of having more

than one species in use for decision making. In Canada, for pulp and paper mills

as well as metal mining, fish survey expert working groups recommend using two

species at a minimum to better determine and interpret the potential

environmental effects (Environment Canada, 2005). However, in tributaries such

95

as these in upper St. John River, where fish communities consist of less than five

species, selection of more than one species can be problematic.

In brook trout, the absence of impact may be a result of the increased

rainfall in 2003, the year previous to this prespawning sample collected in

September 2004. Salmonids are much more mobile than cottids and are not as

synchronized for spawning. Brook trout may be originating from the mainstem St.

John River, or resident within the tributary in which they were collected. This

would constitute support for not using trout as indicators, because their

responses would not reflect changes locally, but rather would be more complex

based on where the eggs were laid and where reproductive adults were resident

at the time of recrudescence. It would also be interesting to repeat these

collections in years following dryer conditions. In sculpin, the prespawning period

occurs several months following the summer rainfall and agricultural inputs, so

the exposure period occurs prior to gonadal development (Chapter 3).

This study represents further investigation of previously reported

reproductive impacts on sculpin collected near agricultural operations. Brook

trout were included to assess a larger-bodied, more mobile fall spawning

species. It was our prediction that the two species would have differential

exposure to the stressors based on their life history and their reproductive

strategies. Sculpin and trout were collected prior to their respective spawning

periods to better assess their reproductive integrity, and in fact, the species

exhibited different response patterns, with brook trout having much more

variability in responses. Future studies could address the significance of the

96

variability seen in brook trout, or rather be designed in such a way to minimize

this factor.

97

4.6 References Barrett, JC, GD Grossman, and J Rosenfeld. 1992. Turbidity-induced changes in

reactive distance of rainbow trout. Trans. Am. Fish. Soc. 121: 437-443.

Cairns, DK. (Ed.). 2002. Effects of land use practices on fish, shellfish, and their

habitats on Prince Edward Island. Can. Tech. Rep. Fish. Aquat. Sci. No. 2408.

157 pp.

Carignan, V, and MA Villard. 2002. Selecting indicator species to monitor

ecological integrity: a review. Environ. Monit. Assess. 78: 45-61.

Chambers, PA, J DuPont, KA Schaefer and AT Bielak. 2002. Effects of

agricultural activities on water quality. Canadian Council of Ministers of the

Environment, Winnipeg, Manitoba. CCME Linking Water Science to Policy

Workshop Series. Report No. 1.

Chapman, DW. 1988. Critical review of variables used to define effects of fines in

redds of large salmonids. Trans. Am. Fish. Soc. 117:1-21.

Gray, MA, RA Curry and KR Munkittrick. 2002. Non-lethal sampling methods for

assessing environmental impacts using a small-bodied sentinel fish species.

Water Quality Res J Can 37: 195-211.

98

Landis, WG and Ming-Ho Yu. 1999. An Introduction to toxicity testing. In

Introduction to Environmental Toxicology: Impacts of chemicals upon ecological

systems. CRC Press, Boca Raton, FL. pp.21-53.

Lobón-Cerviá, J, Rincón, PA. 2004. Environmental determinants of recruitment

and their influence on the population dynamics of stream-living brown trout

Salmo trutta. Oikos. 105, 641-646.

Munkittrick, KR, M McMaster, G Van Der Kraak, C Portt, W Gibbons, A Farwell

and M Gray. 2000. Development of Methods for Effects-Based Cumulative

Effects Assessment Using Fish Populations: Moose River Project. SETAC Press,

Pensacola, FL. 236 pp.

Mutch, JP, MA Savard, GRL julien, B MacLean, B Raymond, and J Doull. 2002.

Pesticide monitoring and fish kill investigations on Prince Edward Island, 1994-

1999. In D.K. Cairns (ed.). Effects of land use practices on fish, shellfish, and

their habitats on Prince Edward Island. Can. Tech. Report. Fish. Aquat. Sci. pp.

94-115.

Redding, JM, CB Schreck, and FH Everest. 1987. Physiological effects on coho

salmon and steelhead of exposure to suspended solids. Trans. Am. Fish. Soc.

116:737-744.

99

Sowden, TK and G Power. 1985. Prediction of rainbow trout embryo survival in

relation to groundwater seepage and particle size of spawning substrate. Trans.

Am. Fish. Soc. 114:804-812.

Tetreault, G.R. 2002. Monitoring the aquatic environment in the Athabasca Oil

Sands using reproductive responses in small-bodied fish. M. Sc. thesis.

University of Waterloo, Waterloo. Ontario.

Waters, TF. 1995. Sediment in streams: sources, biological effects, and control.

Am. Fish. Soc. Monogr. No. 7., Bethesda, MD.

100

Table 4-1 Length, weight, condition factor, liversomatic index (LSI) and gonadosomatic index (GSI) for slimy sculpin and brook trout collected along the St John River. Data are shown as mean ± standard error.

Site Total

Length (mm)

Body

Weight (g)

Cond.

Factor

LSI GSI

AG1 60.67±2.59 2.77±0.4 1.17±0.04 3.07±0.23 16.15±1.30

AG2 76.07±2.27 5.53±0.62 1.20±0.03 3.18±0.14 13.51±0.47

AG3 70.75±3.21 4.04±0.59 1.08±0.02 3.21±0.25 12.66±0.63

FOR1 64.31±2.58 3.51±0.41 1.24±0.03 3.94±0.22 20.53±0.79

FOR2 55.50±1.94 1.93±0.21 1.09±0.02 2.71±0.22 11.79±0.73

Slimy

Sculpin

FOR3 67.47±2.59 3.77±0.44 1.15±0.02 2.94±0.12 12.93±0.61

AG1 118.7±1.36 17.6±2.01 1.01±0.03 1.16±0.16 3.98±1.53

AG3 103±7.92 11.14±2.58 0.97±0.05 0.95±0.06 0.19±0.06

FOR1 124.3±2.85 19.9±1.68 1.02±0.04 0.97±0.14 2.12±1.60

Brook

Trout

FOR2 124.4±4.85 20.89±4.00 1.04±0.10 1.32±0.51 2.57±2.14

101

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

A

B

C

FOR2

FOR1

AG1

AG2

FOR3

AG3

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

Grand FallsLittle River

St. John River

Saint John

Fredericton

NEW BRUNSWICK

St. Leonard airport

A

B

C

FOR2

FOR1

AG1

AG2

FOR3

AG3

Figure 4-1 Map of (A) North America, (B) New Brunswick and (C) the St John River (C) Little River and sampling sites. These sites were chosen along the St John River based on studies conducted by Gray et al. (2005).

102

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5 6Body Wt (g)

Gon

ad W

eigh

t (g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

0

0.1

0.2

0.3

0 1 2 3 4 5 6Body Wt (g)

Live

r Wei

ght (

g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

p=0.001

FOR1>AG1, FOR2, FOR3

p<0.001

FOR1>AG>all other sites

A

B

0

0.2

0.4

0.6

0.8

1

0 1 2 3 4 5 6Body Wt (g)

Gon

ad W

eigh

t (g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

0

0.1

0.2

0.3

0 1 2 3 4 5 6Body Wt (g)

Live

r Wei

ght (

g)

AG 1

AG 2

AG 3

FOR 1

FOR 2

FOR 3

p=0.001

FOR1>AG1, FOR2, FOR3

p<0.001

FOR1>AG>all other sites

A

B

Figure 4-2 Relationship between gonad weight v. body weight (A) and liver weight v.

body weight (B) for slimy sculpin collected at 6 sites along the St John River.

103

AG 1AG 2

AG 3FOR 1

FOR 2FOR 3

SITE

0

5000

10000

15000

In v

itro

Tes t

oste

rone

Pr o

duct

ion

(pg /

g tis

s ue)

AG 1AG 2

AG 3FOR 1

FOR 2FOR 3

SITE

0

5000

10000

15000

20000

In v

itro

Est

radi

ol P

r odu

ctio

n (p

g/g

tissu

e)

Figure 4-3 Steroidogenic capacity of (A) testosterone and (B) estradiol by gonadal tissue excised from slimy sculpin collected in April 2004 at agricultural (AG 1-3) and forested (FOR 1-3) sites. Hormone determinations were made using RIA with the incubation media.

B

A

104

105

Figure 4-4 Relationship between gonad weight v. length (A) and liver weight v. length (B) for brook trout collected at 4 sites along the St John River.

0

0.5

1

1.5

2

2.5

3

80 90 100 110 120 130 140 150

Length (mm)

Gon

ad W

t (g)

FOR 1FOR 2AG 1AG 3

p=0.574

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

80 90 100 110 120 130 140 150Length (mm)

Live

r Wt (

g)

FOR 1FOR 2AG 1AG 3

p=0.872

A

B

106

AG 1 AG 3 FOR 1FOR 2SITE

0

5000

10000

15000

20000

In v

itro

test

oste

rone

pro

duct

ion

(pg/

g tis

sue)

AG 1 AG 3 FOR 1FOR 2SITE

0

5000

10000

15000

20000

In v

itro

test

oste

rone

pro

duct

ion

(pg/

g tis

sue)

AG 1 AG 3 FOR 1FOR 2SITE

0

10000

20000

30000

40000

50000

60000

70000

In v

itro

Est

radi

ol P

rodu

ctio

n (p

g/g

tissu

e)

AG 1 AG 3 FOR 1FOR 2SITE

0

10000

20000

30000

40000

50000

60000

70000

In v

itro

Est

radi

ol P

rodu

ctio

n (p

g/g

tissu

e)

Figure 4-5 Steroidogenic capacity of (A) testosterone and (B) estradiol by gonadal tissue excised from brook trout collected in September 2004 at agricultural (AG 1 & 3) and forested (FOR 1-2) sites. Hormone determinations were made using RIA with the incubation media.

A

B

107

5 Approaching population-level ecological risk assessment from an

effects driven perspective

5.1 Abstract

There have been recent attempts to modify the traditional risk assessment

process to consider multiple stressors, and a major deficiency with environmental

impact assessment has been its inability to deal with multiple discharges or

complex situations. From 1999-2004, effects-driven assessments were

conducted to investigate the performance of fish populations in potato-growing

areas of New Brunswick. These studies were conducted using a small-bodied

benthic fish species, the slimy sculpin (Cottus cognatus) as it is abundant in

these areas, exhibits site fidelity in a small home range, and lives and feeds on

the stream bottom. These populations exhibit changes in growth, fecundity and

size distributions–effects that are important in the absence of acute mortalities.

These changes have been associated with higher water temperatures and

shorter-term high rainfall events. This paper examines the data within a

framework for ecological risk assessment, focusing on the ability to conduct

population-level risk assessments. After synthesizing all available information,

high risk periods for sculpin appear to be July and August. Rainfall and

temperature patterns can be used to identify high risk areas for fish populations.

Population level risk assessment holds potential for shifting assessments to

addressing more ecological relevant responses.

108

5.2 Introduction

Ecological risk assessment is considered one of the preferred methods of

applying scientific knowledge to the decision making process. The purpose of

ecological risk assessment is to contribute to the protection and management of

the environment through scientifically credible evaluation of the ecological effects

of human activities (Suter and Barnthouse 1993). The risk assessment process

provides a way to develop, organize and present scientific information so that it is

relevant to environmental decisions. Ecological risk assessments can vary based

on the scenario and the assessor, but usually will contain discrete phases and

goals including problem formulation, analysis, and risk characterization. There

have been attempts to modify the traditional risk assessment process to consider

multiple stressors and population level endpoints (Foran and Ferenc, 1999;

Ferenc and Foran, 2000). These initiatives are meant to address major

deficiencies with environmental risk assessment and its inability to deal with

multiple discharges or complex situations.

In the mid-1990s, there was a move by multiple regulatory bodies to

modify environmental assessment approaches to address the multitude of

stressors in complex receiving environments (Ferenc and Foran 2000,

Munkittrick et al. 2000). This has involved a conceptual shift away from single

stressor analyses to consider the importance of multiple stressors and

cumulative impacts. Despite the development of conceptual models and general

frameworks by both research and regulatory communities, relatively few

109

quantitative assessments of multiple stressors have been conducted (Foran and

Ferenc 1999).

Any assessment must have defined endpoints. Because of redundancy,

feedback loops, and other compensatory mechanisms, many changes can be

documented at one level without affecting higher levels (Suter 1993), and the

significance of an endpoint should be determined by its importance to a higher

level of biological organization. For example, a physiological change may be

significant if it affects the survival, growth, or reproduction of the whole organism.

A recent workshop recognized that despite the many published calls for

assessing risks to populations, the overwhelming majority of assessments of

ecological risks are still based on an individual-level approach (Barnthouse et al.

2007). It was the consensus of these experts that data collected for population

level ecological risk assessment (PLERA) should begin to include information

related to population structure, mortality, sex ratios, distribution, and movement

(Table 5-1). These empirical data often serve as inputs for modelling efforts to

further increase the number and breadth of assessment endpoints that are

included.

From 1999-2004, researchers have been investigating fish responses in

agricultural areas of New Brunswick. Recently completed studies have compared

fish performance in 20 tributaries of the St John River (Gray and Munkittrick

2005), and have shown that reproductive performance was very low at sites

adjacent to intensive potato production. Follow-up studies described in this study

were designed to include nonlethal sampling over discreet time intervals, ruling

110

out an ecological basis for the observed changes, and comparisons of species

with differing life histories. The progression of these studies has an application

towards population-level risk assessment, and this effort will serve to summarize

those data and interpret them in the context of protecting fish at the population

level and conducting a population level risk assessment.

5.3 Background to Agricultural Studies

There are a variety of concomitant stressors associated with intensive

potato production, including increases in water runoff, sediment deposition (Gray

and Munkittrick 2005) and nutrients and pesticides, but the relative roles of the

various stressors in effecting observed changes in size, fecundity, nest size,

gonad size, liver size and fish condition (Gray et al. 2002, 2005) have not been

identified.

It has been established that there are reductions in the density of young-

of-the-year sculpin in agricultural areas of NB (Gray et al. 2002). More recent

work addressed the timing of mortality events and further investigated reductions

in fish numbers in agricultural areas. Systematic sampling of fish densities and

size structures through multiple years at sites to identify peak mortality periods

was used to identify potential stressors associated with the suspected periods of

mortality (Chapter 2). Sculpin mortality appears to be higher during summer

periods, coinciding with high water temperatures and periodic rainfall events.

Circumstantial evidence in other areas of eastern Canada indicates that storm

events may be playing a role in the year class failures and extinction events of

111

fish (Cairns 2002), but there is a demonstrated need to also assess the causative

factors associated with differences seen in growth, organ size and fecundity

(Gray et al. 2005).

Earlier studies documented impacts on reproduction and year class

strengths, and follow-up studies identified that year class strength was affected

by summer rainfall (Chapter 2). There is much higher variability in fish numbers

in the agricultural areas (Chapter 2), however we did not see the same year class

failures associated with wet years (Gray et al. 2005) and thunderstorm events

(Cairns 2002) seen previously. The persistence of sculpin populations despite

poor reproductive performance (Gray and Munkittrick, 2005) suggests that the

persistence of fish populations in agricultural areas cannot be predicted based on

individual performance attributes. Protection of the persistence of populations

will be a complex interaction of the biology of the fish, the timing of chemical

applications, and the interactions of rainfall, flow, temperature and nutrients. It

appears that sculpin populations in agricultural streams persist by having strong

year classes in years with low rainfall during the potato growing season (Chapter

2).

Given the absence of information on the interactions of the various

potential stressors and their consequent impacts on fish performance, it is

difficult to identify areas where agricultural activities put fish populations most at

risk. Although populations are by no means the only entities that can or should

be addressed in ecological risk assessments, the reproducing population is the

lowest level we can meaningfully protect (Suter and Barnthouse 1993; Figure

112

Figure 5-1), and the general rarity of assessments that focus on population

characteristics is quite remarkable. It is simply not possible to estimate the

impacts of all potential stressors on all biota at all levels of biological

organization. Instead the focus should be on the most relevant ecological

receptors for the stressor dynamics or nature of the issue. The organism is the

smallest unit that interacts directly with the environment, and individual-level

responses are currently the best described component of toxicology. For these

reasons, most risk assessment methodologies operate at the individual level.

Moreover, risk assessors continue to struggle in determining how population-

level considerations can be integrated into environmental decision-making.

5.4 Risk Assessment As usually practiced, the ecological risk assessment process focuses on

specific stressors, valued ecosystem components usually identified by key

stakeholders, potential stressor-response pathways associated with known or

previously identified impacts, and potential mitigative approaches to reduce the

risk associated with a development or activity under consideration. The US

EPA’s risk assessment guidelines were deliberately designed to be general and

applicable to the full range of assessment problems by recognizing three classes

of assessments: stressor-based, effects-directed, and values-directed (Harwell

and Gentile 2000). Although the stressor-based risk assessment approach has

helped to reduce the environmental impact of specific stressors, it is lacking

methodology to assess the additive impacts and interaction between effects

113

associated with multiple stressors over space and time (Dubé and Munkittrick

2001).

Quite often field studies involve conducting a retrospective risk

assessment, where a study is initiated because a biological effect has been

noted, and the field study is designed to find the cause. Of the several types of

ecological risk assessments, the retrospective risk assessment is often deemed

the most difficult (Suter 1993). Multiple stressor assessments tend to focus on

larger spatial scales and are concerned with effects at higher levels of

organization: subject areas for which we lack expertise in ecological risk

assessment. Risk-based assessment of multiple stressors presents a situation

with more complexity, and should not be viewed as a sequential set of steps

executed in a well-defined order. Instead, it should be viewed as an iterative

process dominated by problem complexity. Attempts to define the problem,

identify study goals, or recognize the major factors responsible for determining

system characteristics and dynamics will not necessarily be immediately

successful. Estimating effects of multiple stressors to biota is a daunting task

when one considers that stressors are disparate in nature and interact on

differing temporal and spatial scales. Effects may be direct, such as increased

mortality or decreased fecundity, or indirect, as in altered predator and prey

dynamics (Riddell et al. 2005).

A major aim of ecotoxicology is to establish causal linkages between

exposure to a toxicant, or toxicants, and resultant biological effects. In the

laboratory, where investigators study the relationships between exposure to

114

limited numbers of toxicants and alterations in limited numbers of endpoints,

such linkages can be examined exhaustively. Laboratory replication and dose-

response studies allow definitive statements concerning the nature of causal

linkages. However, when these linkages are sought in natural ecosystems, many

other factors may interfere with our ability to make definitive statements

concerning causality. These other factors include the multitude of toxicants often

present in natural systems, non-contaminant stressors (including both

anthropogenic and natural stressors), and the inherently high biological variability

of natural systems.

Limited baseline data availability hinders the ability of stressor-based

approaches to assess the state of the existing environment. An understanding of

the environmental components of the system includes existing stressors, existing

sensitivities to stressor exposure, and natural variability in the system over time

and space (Munkittrick et al. 2000). The risk assessment process also commonly

lacks a post-operational monitoring phase to verify that the mitigative attempts

have successfully eliminated or reduced the risk, or to examine whether

unanticipated interactions have resulted in unpredicted impacts. This will require

the development of a level of commitment to baseline monitoring, science-based

decision-making, and post-operational monitoring that is currently lacking in most

situations.

Non-point sources are more difficult to evaluate because discharges are

complex mixtures, the concentrations of toxicants are variable, and rates of input

and timing of discharges are difficult to predict. Sublethal impacts may already

115

exist within an aquatic system, and further addition of stress, regardless of

magnitude, could have impacts beyond those that would have been observed

through stressor-based approaches. Understanding the capability of a system to

assimilate natural and anthropogenic wastes would be much easier once the

factors limiting ecological performance for a particular system are known and

existing mechanisms of impact are understood.

For the purpose of assessing the impacts of agriculture operations to fish

populations, we chose to adapt the existing ecological risk assessment

framework to better interpret the available information at the population level, and

assess impacts.

5.5 Population-level assessment of agricultural impacts on fish

The risk assessment will rely heavily on the situation of potato farming in

northern New Brunswick, Canada, to illustrate how population-level risk

assessment may be applied to understanding the risks of anthropogenic

developments, in both a retrospective and predictive format. The traditional risk

assessment paradigm consists of specific stages of problem formulation,

analysis, and risk characterization. Recently this has been adapted for population

level endpoints and expanded to refine the interpretation of the collected

information (Figure 5-2).

5.6 Population-Level Problem Formulation

116

In the traditional risk assessment paradigm, the problem formulation stage

involves evaluating goals and selecting assessment endpoints, preparing the

conceptual model, and developing an analysis plan. The problem statement for a

traditional risk assessment usually relates to the potential risk of an exposure or

a stressor to an individual. However, a population can tolerate many individual-

level impacts before a major change occurs at the population level in terms of

recruitment, reproductive rates, mortality or abundance that threatens the

sustainability of the population. As you go up in levels of complexity to a

population from an individual, there are increases in time lag and ecological

relevance, but decreases in reversibility and ability to determine the causes of

changes (Munkittrick et al., 2000). The problem formulation statement for the

population-level risk assessment is going to be influenced by factors such as

abundance, age/stage structure, sex ratio and recruitment, whereas concerns for

individuals revolve around issues such as body burdens, exposure

concentrations, physiological alterations, and acute and chronic toxicity

responses. The population acts as an integrator of exposures, habitat factors

and ecological interactions to give an integrated assessment of the status.

For an agricultural situation, an individual-level risk assessment might

address the potential hazards and risks of a pesticide or herbicide. But at the

population level, the question focuses on whether the fish population can sustain

the cumulative stressors imposed by agriculture. Within the example of potato

farming, potential stressors have been identified and fish responses that have

117

been reported. These will be characterized in the data analysis phase. But the

next step in problem formulation is developing a conceptual model.

The first step in developing the conceptual model is to identify the species

of fish to focus the risk assessment. While the individual level risk assessment

often uses public stakeholder input to identify valued ecosystem components

(VECs), the population level assessment needs to focus on the species that has

the most potential to answer the specific questions. Species selected as VECs

by public stakeholders often do not have ideal characteristics for interpreting the

potential impacts of stressors. In the northern New Brunswick situation, there are

four species present in most tributaries in the agricultural region: brook trout

(Salvelinus fontinalis), slimy sculpin, brook stickleback (Culea inconstans) and

blacknose dace (Rhinichthys atralus). The stickleback are only found in slow

backwater areas and in beaver ponds, and are not widely distributed. The dace

are only found in the lower reaches of the rivers, where water is warmer. Of the

available species, a comparison of species life history characteristics (Table 5-2)

shows that slimy sculpin are less mobile (Gray 2003), more abundant (Curry and

Munkittrick 2005), and are more likely to be able to reflect local conditions.

The reduced proportions of YOY sculpin evident in these agricultural

areas (Gray et al., 2005), are either due to reproductive dysfunction, increased

mortality, or a combination of these factors. Systematic sampling of fish densities

and size structures through multiple years at sites to identify peak mortality

periods was used to try to identify potential stressors associated with the

suspected periods of mortality (Chapter 2). A detailed systematic study of the

118

reproductive development of slimy sculpin (Chapter 3) has ruled out a seasonal

or ecological basis for the depressions that have been previously documented at

agricultural sites.

Over three growing seasons fish densities and size structures have been

monitored non-lethally along an agricultural gradient to identify peak mortality

periods across all size classes (Chapter 2). Data indicate that the peak risk

period may occur in late summer, with chemical applications on the fields

coinciding with increased thunderstorm activity, rather than having major

mortality periods associated with overwinter losses or after-spawning mortality.

The conceptual model includes a variety of potential pathways of stressors

leading to the main responses that have been seen in slimy sculpin populations

(Figure 5-3). The analysis plan is to evaluate the population-level characteristics

of the slimy sculpin in terms of the potential risks in agricultural areas.

5.7 Population-level risk analysis

Following the problem formulation stage, risk assessors evaluate exposure to

stressors and the relationship between stressor levels and ecological effects.

Effects driven assessments focus the discussion on sustainability, acceptability,

and the consequences of additional future changes rather than just on adverse

effects. The framework outlined by Munkittrick and colleagues (2000) is designed

to be used iteratively to design hypotheses to focus follow-up studies on the

aspects of performance that are responding to the stressors within the system.

119

The studies by Gray and Munkittrick (2005) and Gray et al. (2002)

demonstrated that year class failures happened intermittently, consistent with the

periodic thunderstorm-related acute fish kills seen in other potato producing

areas (Cairns, 2002). A detailed study of 20 small streams evaluated the health

and abundance of fish populations, and concluded that temperature was a

dominant factor accounting for the distribution and abundance of sculpin

populations in northern New Brunswick (Figure 5-4) (Gray et al. 2005). Slimy

sculpin populations were absent from streams with a maximum summer water

temperature above 22°C, and were reduced in abundance in a linear fashion in

streams that had a maximum summer water temperatures above 16°C (Gray et

al., 2005). Size of young-of-the-year (YOY) was also related to temperature, with

median sizes of YOY sculpin continuing to increase to water temperatures of

22°C.

The studies conducted by Gray (2003) were conducted during relatively wet

years, while more recent (Chapter 2) studies were conducted in the same area

during relatively dry years. A significant correlation was found between year

class strength and the number of large summer rainfall events in agricultural

areas, but no such relationship was seen in forested areas (Figure 5-5). The

summer period also appears to be the period of highest mortality risk for sculpin

(Chapter 2).

Abundance and size could not be related to sediment deposition although

both sediment deposition and water temperatures were elevated in agricultural

areas (Gray et al., 2005). It is likely that both sediment deposition and pesticide

120

exposures are playing a role, but that the sampling methods employed to sample

sediments and pesticides are not reflecting the real risk to populations. The best

predictors of good sculpin population health are cool water temperatures and

lower numbers of large summer storm events. Increasing risk is associated with

warmer water temperatures and higher summer storm frequencies.

5.8 Population-level risk characterization

Assessors estimate risk through integration of exposure and stressor-response

profiles, describe risk by discussing lines of evidence and determining ecological

adversity. With the information collected as part of this research, combined with

previous and ongoing research, it was attempted to identify areas of concern for

sculpin populations. Rainfall and temperature were identified as important

stressors to sculpin populations. Maps of New Brunswick were constructed

highlighting the St. John River basin and the environmental data collected from

Environment Canada climate monitoring stations

(http://www.climate.weatheroffice.ec.gc.ca/) were overlain.

Contour maps of temperature were drawn, using the total number of

degree days over 18°C to indicate areas of the province where warmer water

temperatures would be expected. It is acknowledged that local groundwater

inputs will alter stream temperatures, but at a coarse level for population-level

risk assessment, this can be used to at least indicate the areas of high risk for

sculpin populations associated with agricultural development. Darker colors

121

indicate areas of the province with higher numbers of degree days over 18°C

(Figure 5-6). Storms exceeding 15 mm total precipitation have shown to be

negatively correlated to % YOY in areas with agricultural practice (Chapter 2),

and these data were also grouped to identify regions of the province receiving

heavy storm activity (Figure 5-7).

Each of these data: temperature, total rainfall for July and August, number

of storms exceeding 10 mm of rainfall, were plotted for the St John River basin.

The combined data identify the Grand Falls area as a concern for sculpin

populations (Figure 5-8). This reaffirms the need to study fish populations in this

farming region, particularly in years with warmer temperatures and heavier

rainfall.

5.9 Additional Uses

Although this is one such use of population-level assessment as it relates

to agricultural stressors, there are many scenarios that lend themselves to use of

this approach. One of the outcomes of this work involved mapping stressors and

overlaying data to better assess areas of risk. This technique could be used for

future monitoring of fish populations in agricultural regions, or even in a larger

context of climate change and further industrial development. There is currently a

larger effort to estimate the assimilative capacity of the St. John River

(Munkittrick et al. NCE), and this type of tool could be of use in overlaying

industrial effluent plumes, critical fish and benthic habitat areas, hydroelectric

operations and resulting flow regimes, etc., in an effort to predict impacts.

122

5.10 Conclusions

This interpretation demonstrates that empirical studies directly measure

population effects and can be used to quantify the causal relationships linking the

stressor to the observed impact. Reproductive and death rates, age structure,

spatial distributions, migration pathways and rates, and habitat utilization are

factors that can be applied in the population-level risk assessment. One of the

more critical pieces of information is the temporal dynamics of the populations

that form the risk assessment. This type of data requires multiple surveys over a

period of time representative of the fluctuations. If the population is known to

have the potential to fluctuate over large numbers very rapidly then a large

number of samples over a short time period will be necessary. The danger is in

selecting a sampling period that does not allow an accurate determination of the

extremes and potential rate of change in the numbers of individuals. Too long a

sampling period will likely underestimate the extremes and the rate of change.

Additional species can provide a more comprehensive picture of the risks

involved. Measurable attributes of individuals should be used to estimate

population parameters. They are also widely used as inputs to population

models, particularly individual-based models. While attributes of individuals can

also be used to evaluate stress to an individual, it is only when aggregated over

multiple individuals that they are useful in population risk assessment. It may be

argued that the only true measures of populations are their total numbers, but

populations and communities can exhibit a graded level of response (Munkittrick

and McCarty 1995).

123

At the population level, the first indicators of stress include changes in

individual growth rates and other performance measures, long before changes in

recruitment or abundance are noticed. For example, individual-level performance

characteristics that can integrate factors that affect key performance

measurements include growth, reproduction, and survival. They are also useful in

determining the mechanism for which the stressor exerts a population-level

impact. These attributes are directly measured on individuals of the assessment

population. Individual level assessments are still important in environmental

decision making, especially as it relates to protected species where the individual

organisms are highly valued and unnecessary deaths of any individuals are

considered a loss to society (Barnthouse et al. 2007).

124

5.11 References

Barnthouse, LW, Munns, WR Jr, Sorenson, MT. 2007. Population-Level

Ecological Risk Assessment. Taylor Francis-CRC Press. 346 p.

Brasfield SM, Tetreault G, McMaster ME, Munkittrick KR. Seasonal

characterization of energy expenditure, energy storage, and in vitro gonadal

steroidogenic capacity in slimy sculpin (Cottus cognatus). Submitted to J. Fish

Biology.

Cairns, DK. (Ed.). 2002. Effects of land use practices on fish, shellfish, and their

habitats on Prince Edward Island. Can. Tech. Rep. Fish. Aquat. Sci. No. 2408.

157 pp.

Curry, R.A. and K.R. Munkittrick. 2005. Fish community responses to multiple

stressors along the Saint John River, New Brunswick, Canada. In (J.N. Rinne,

R. Calamusso, and R. Hughes Eds.) Changes in large river fish assemblages in

the North America: Implications for management and sustainability of native

species. American Fish. Soc. Symp. 45:505-521.

Dubé, M.G. and K.Munkittrick. 2001. Integration of effect-based and stressor-

based approaches into a holistic framework for cumulative effects assessment in

aquatic ecosystems. Human Ecol Risk Assessment 7:247-258.

125

Ferenc, S.A., and Foran, J.A. (eds.), Multiple stressors in ecological risk and

impact assessment: Approach to risk estimation. SETAC Press, Pensacola, FL,

USA. 264 p.

Foran JA, Ferenc, SA, editors. 1999. Multiple stressors in ecological risk and

impact assessment. SETAC Press, Pensacola, FL. USA. 100 p.

Gray, M. A., Curry, R. A. & Munkittrick, K. R. (2002). Non-lethal sampling

methods for assessing environmental impacts using a small-bodied sentinel fish

species. Water Quality Research Journal of Canada. 37, 195-211.

Gray, M. A., (2003). Assessing non-point source pollution in agricultural regions

of the upper St. John River basin using the slimy sculpin (Cottus cognatus). PhD

thesis. University of New Brunswick, Fredericton, N.B.

Gray, M. A. & Munkittrick, K. R., (2005). An effects-based assessment of slimy

sculpin (Cottus cognatus) populations in agricultural Regions of Northwestern

New Brunswick. Water Quality Research Journal of Canada 40, 16-27.

Harwell MA and Gentile JH. 2000. Environmental decision-making for multiple

stressors: framework, tools, case studies, and prospects. In: Ferenc SA and

Foran JA (eds), Multiple Stressors in Ecological Risk and Impact Assessment:

126

Approaches to Risk Estimation. Society of Environmental Toxicology and

Chemistry, Pensacola FL.

Munkittrick, KR, McCarty, LS. 1995. An integrated approach to aquatic

ecosystem health: top-down, bottom-up or middle-out? J. Aqautic Ecosystem

Health 4: 77-90.

Munkittrick, KR, M McMaster, G Van Der Kraak, C Portt, W Gibbons, A Farwell

and M Gray. 2000. Development of Methods for Effects-Based Cumulative

Effects Assessment Using Fish Populations: Moose River Project. SETAC Press,

Pensacola, FL. 236 pp.

Riddell D, Culp JM, Baird DJ. 2005. Sublethal effects of cadmium on prey choice

and capture efficiency in juvenile brook trout (Salvelinus alpinus). Environ.

Toxicol. Chem. 24: 1751-1758.

Suter II, GW. 1993. Introduction to ecological risk assessment. In: Suter II, G.W.

(Ed.) Ecological Risk Assessment, pp. 21-47. Lewis, Boca Raton, Florida.

Suter II, GW, Barnthouse, LW. 1993. Assessment concepts. In: Suter II, G.W.

(Ed.) Ecological Risk Assessment, pp. 21-47. Lewis, Boca Raton, Florida.

127

Table 5-1 Recommended empirical and modeling attributes that should be collected or computed as part of a population level ecological risk assessment (Barnthouse et al. 2007).

Population parameters

computed from population

attributes

Attributes of populations

computed from

individual attributes

Measurable attributes

Population growth rate Abundance Density

Variance of abundance Age/stage structure Age, size, sex

Sex ratio Individual length

Population attractor (k) Recruitment Size

Fecundity

Egg size

Size or age at maturity

Number of viable offspring

Probability of extinction Survivorship Individual weight

Age/stage at death

Timing of mortality

Time to recovery / extinction Biomass Somatic growth rate

Energy storage Liver size

Condition

Density dependence Spatial distribution Movement/dispersal

Habitat preference Home range

Critical patch size Location (specific time)

Diet Stomach contents

128

Table 5-2 Available data for brook trout (Salvelinus fontinalis) and slimy sculpin (Cottus cognatus) to be used in population level ecological risk assessment. Efforts were made to utilize information collected in the St John River system with standardized sampling, where possible. Measurable attributes of

individuals

Available information for Brook

Trout (S. fontinalis)

Available information for

Slimy Sculpin (C. cognatus)

Density 100-200 /100m2

Age, size, sex Up to 5 yrs,

Individual length TL 3-12 cm

Size

Fecundity 100-400 50-300

Egg size

Size or age at maturity 2-3 1-2

Number of viable offspring

Individual weight

Age/stage at death 4-5

Timing of mortality

Somatic growth rate

Liver size

Condition

Movement/dispersal

Home range 10-10,000 m2 1-100 m2

Location (specific time) Deeper pool, embankments Swift water and riffle areas

Stomach contents Invertebrates, small fish invertebrates

129

Biosphere

Ecosystem

Community

Cellular

Tissue

Population

Organ

Individual

Molecular

Feasibility and cost of testing and monitoring

Precision and generality of the results

Suter and Barnthouse 1993

Smallest unit for measuring an integrated response

Transitory

Consumable

Expendable

Smallest ecological unit persistent on scale of decades

Lowest level that we can meaningfully protect

Biosphere

Ecosystem

Community

Cellular

Tissue

Population

Organ

Individual

Molecular

Feasibility and cost of testing and monitoring

Precision and generality of the results

Suter and Barnthouse 1993

Smallest unit for measuring an integrated response

Transitory

Consumable

Expendable

Smallest ecological unit persistent on scale of decades

Lowest level that we can meaningfully protect

Figure 5-1 Comparing assessment endpoints on the individual vs. population level in the overall levels of biological organization.

130

Figure 5-2 Ecological risk assessment framework as defined by the US EPA (1998) and as modified to address population level assessment (Barnthouse et al. 2007).

Problem Formulation

Analysis

Risk Characterization

Collect site information

Identify potential stressors

Select focal species based

on life history

characteristics

Gather species specific

information

Collect stressor attribute data

•Species response •Define stressors of interest

Conduct analysis to identify key

stressors potentially affecting

populations (empirical, modeling)

Conduct risk characterization to

evaluate population-level risks

Risk management/decision process

131

Figure 5-3 Conceptual diagram of stressors associated with potato farming practices in northwestern New Brunswick as well as responses documented in previous study (1999-2002) by Gray and colleagues.

Densities

Survival

Baby Fish

Gonad size

Fecundity

Nest size and density

Liver size

Growth

Responses

Pesticides

Long term alteration flow regime

Sediment

Temperature

Nutrients

Stressors

Episodic events

Pesticides

Temperature

Nutrients

Densities

Survival

Gonad size

YOY

Nest size and density Fecundity

Liver size

Growth

Episodic events

Sediment

Long term alteration flow

Condition

132

Figure 5-4 Relationship between sculpin density (per m2) and maximum mean daily water temperature. Open triangles represent agricultural sites, filled triangles represent forested sites. Graph reprinted with permission from M. Gray.

133

y = -2.7267x + 60.032R2 = 0.2899

0

10

20

30

40

50

60

70

80

0 2 4 6 8 10 12

Number of Storms >15 mm

Perc

enta

ge Y

oung

of t

he Y

ear (

%)

Figure 5-5 Linear relationship between number of major summer storms (>15 mm total rainfall) and percent young of the year (YOY) for sites along the gradient of agricultural inputs. Data were collected over a period of 1999-2004.

134

Figure 5-6 Map of New Brunswick showing total degree days over 18°C as observed at monitoring stations for the months of July and August. Groups were assigned and lines were drawn in efforts to combine areas with similar values.

135

Figure 5-7 Map of New Brunswick showing total number of rainfall events exceeding 15 mm of total precipitation as recorded at monitoring stations during the summer months of July and August. Groups were assigned and lines were drawn in efforts to combine areas with similar values.

136

Figure 5-8 Map of New Brunswick showing overlapping temperature and precipitation data including total number of degree days over 18°C, total rainfall for the summer months of July and August, and number of storms exceeding 10 mm total precipitation.

137

6 Conclusions

Fish in agricultural areas have shown reduced proportions of YOY fish, due to

reproductive dysfunction, increased mortality, or a combination of these factors.

The main objective of the thesis was to identify the potential mechanisms

associated with the reduced reproductive performance in agricultural areas.

Previously, research showed reduced reproductive performance in the

agricultural areas of this watershed, but it remained unclear whether these

reduced larval densities in summer were a function of reproductive dysfunction in

adult sculpin or difference in apparent mortality rates in larval fish in forested and

agricultural sites along Little River.

To address these research needs, monitoring of fish populations along a

gradient of agricultural inputs was continued and expanded to allow monthly

nonlethal samples over three growing seasons (Chapter 2 “Monitoring of fish

populations along a gradient of agricultural inputs in Northwestern New

Brunswick, Canada”). This data set allowed for comparisons over time and better

assessment for periods of risk related to agriculture and corresponding sculpin

population responses. In the agricultural reach, sculpin showed increased

growth, density and increased variability relative to upstream reaches.

Consistent with previous studies, YOY sculpin downstream of agricultural inputs

were longer and heavier than those of upstream non-agricultural sites (Gray et

al., 2002; Gray and Munkittrick, 2005; Gray et al., 2005). Although new

information is presented on sculpin population changes within and between

years, more information was needed on the reproductive timing and population

138

dynamics in an unimpacted system. This would further establish if there is an

ecological basis for the effects that have been documented in agricultural areas.

Although the standardized nonlethal sampling design was improved over

previous studies to isolate periods of risk to fish populations, the data remain

difficult to interpret.

The objective of the seasonal aspect of this thesis was to assess

reference populations of slimy sculpin (Cottus cognatus) to identify the most

appropriate window to assess reproductive integrity (Chapter 3 “Seasonal

patterns of energy storage, energy expenditure, and in vitro gonadal

steroidogenic capacity in slimy sculpin (Cottus cognatus)”). The effects-driven

approach recommends ruling out a physiological or ecological basis for the

changes that have been documented in the study species (Munkittrick et al.

2000). To date, no study has addressed how spring-spawning slimy sculpin

function over the winter and the rate at which recrudescence occurs under ice

cover. The seasonal study determined that in fact it is possible to measure

physiological endpoints during reproductive development, and collections

optimized the in vitro steroid production assay for this species. This study also

ruled out a seasonal or ecological basis for the depressions that have been

previously documented at agricultural sites.

Confirmatory studies were then needed in forested and agricultural areas

with both a spring and fall spawning species, slimy sculpin and brook trout

(Salvelinus fontinalis), respectively. The potato-growing season in northwestern

New Brunswick is from June-October, and may affect sculpin and trout differently

139

as spawning times are in the spring for sculpin or in the fall for trout. Data

indicated that the peak risk period may occur in late summer, with chemical

application coinciding with increased thunderstorm activity, rather than overwinter

or after spawning mortality. This information was collected on slimy sculpin, a

benthic, spring-spawning species, but additional comparisons included brook

trout (Salvelinus fontinalis), a pelagic, fall-spawning species in agricultural areas

to determine the differential susceptibility relative to reproductive timing (Chapter

4 “ Impacts of agriculture on fish with differing reproductive strategies: comparing

slimy sculpin (Cottus cognatus) and brook trout (Salvelinus fontinalis)”). In this

study, impacts on sculpin were not consistent between agricultural sites. Pre-

spawning sculpin demonstrated differences in condition factor, LSI, GSI and in

vitro gonadal steroidogenesis among agricultural and forested sites. Trout

responses were highly variable and no consistent patterns were seen in fish

collected at sites near agriculture. It would appear from the data that fall

spawning brook trout are not synchronized for spawning by the same cues as

spring spawning species. This makes comparisons among and within sites

difficult as fish are at differing stages of reproductive readiness. Within treatment

variability is higher than expected, and all forested sites are not showing the

same response pattern. In a study that was designed to identify differences in a

reproductive impairment in fish with differing reproductive strategies,

unfortunately the patterns remain unclear.

Integrating the knowledge developed on episodic mortality, reproductive

development and reproductive performance for both species allowed for more

140

complete assessment of impacts related to agriculture. Small-bodied species of

fish are becoming more widely used in freshwater assessment programs

because of their abundance and the assumptions that they reflect local

environmental conditions because of increased site fidelity. It has become

necessary in many areas to focus on population-level assessments using these

small-bodied species, because of low species richness and/or inconsistency

among species in exposure histories. With the information collected during this

study, combined with previous and ongoing research, it was our goal to identify

areas of concern for sculpin populations (Chapter 5- “Approaching population-

level ecological risk assessment from an effects driven perspective”.) Rainfall

and temperature were identified as important stressors to sculpin populations in

Chapter 2. Maps of New Brunswick were constructed highlighting the St John

River basin and the environmental data were overlaid. Each of these data:

temperature, total rainfall for July and August, number of storms exceeding 15

mm of rainfall, were plotted for the St John River basin. The combined data

identify the Grand Falls area as a concern for sculpin populations. This reaffirms

the need to study fish populations in this farming region, particularly in years with

warmer temperatures and heavier rainfall.

7 VITA

Candidate's full name: Sandra Marie Brasfield

Universities attended (with dates and degrees obtained):

1995-1999 Middle Tennessee State University BSc Biology, cum laude 1999-2002 Oklahoma State University

MSc Zoology

Publications:

Brasfield SM, Talent LG, Janz DM. 2007. Reproductive and thyroid hormones in captive western fence lizards (Sceloporus occidentalis) following a period of brumation. Accepted by Zoo Biology. Suedel B, Steevens JA, Kennedy AJ, Brasfield SM, Ray GL. 2007. Environmental Consequences of Water Pumped from Greater New Orleans following Hurricane Katrina: Chemical, Toxicological and Infaunal Analysis. Environ. Sci. Technol. 41: 2594-2601. Brasfield SM, Bradham K, Wells JB, Talent LG, Lanno RP, Janz DM. 2004. Evaluation of fence lizard (Sceloporus spp.) eggs as a terrestrial vertebrate model for assessing bioavailability and in ovo effects of soil contaminants. Chemosphere. 54(11): 1643-1651. Brasfield SM, Weber LP, Talent LG, Janz DM. 2002. Dose response and time course relationships of vitellogenin induction in male Western fence lizards (Sceloporus occidentalis) exposed to ethinylestradiol. Environ Toxicol Chem 21(7): 1410-1416. Submitted Gray MA, Keeler RA, Curry RA, Cunjak RA, Clément M, Brasfield SM, Munkittrick KR. 2006. On the biology of the slimy sculpin – a recipe for effective environmental monitoring. Submitted to Transactions of the American Fisheries Society. Brasfield SM, Tetreault G, McMaster ME, Munkittrick KR. Seasonal characterization of energy expenditure, energy storage, and in vitro gonadal steroidogenic capacity in slimy sculpin (Cottus cognatus). Submitted to J. Fish Biology. Brasfield SM, Munkittrick, KR. Monitoring of fish populations along a gradient of agricultural inputs in northwestern New Brunswick, Canada. Submitted Environ. Mon. Assess.

Brasfield SM, Tetreault G, McMaster ME, Munkittrick KR. In prep. Comparison of spring spawning slimy sculpin (Cottus cognatus) and fall spawning brook trout (Salvelinus fontinalis) in agricultural regions of the St John River (New Brunswick, Canada).

Conference Presentations:

Suedel B, Steevens JA, Kennedy AJ, Brasfield SM, Ray GL. Effects of Hurricane Katrina-related Levee Failures on Wetland Sediment" Battelle 4th International Conference on Remediation of Contaminated Sediments in Savannah, GA. 22-23 January 2007.

Brasfield SM, Yoo, LJ, Schlenk, D, Steevens JA. Isolation of Hepatocytes from Fence Lizards for Use as a Screening Tool in Assessing the Estrogenicity of Military Relevant Compounds. Partners in Environmental Technology Technical Symposium & Workshop. 28-30 November 2006.

Suedel B, Steevens JA, Kennedy AJ, Brasfield SM, Ray GL. Environmental consequences of the failure of the New Orleans levee system during Hurricane Katrina: chemical, toxicological, and benthic community analysis. 27th Annual Meeting of SETAC North America, Montreal, PQ, Canada. 5-9 November 2007.

Brasfield SM, Gray MA, Munkittrick KR. Monitoring of fish populations along a gradient of agricultural inputs in New Brunswick, Canada using a nonlethal sampling approach. 27th Annual Meeting of SETAC North America, Montreal, PQ, Canada. 5-9 November 2007. (Invited abstract)

Brasfield SM, Yoo, LJ, Schlenk, D, Steevens JA. Isolation of Hepatocytes from Fence Lizards for Use as a Screening Tool in Assessing the Estrogenicity of Military Relevant Compounds. 27th Annual Meeting of SETAC North America, Montreal, PQ, Canada. 5-9 November 2007.

Gray, MA, Brasfield SM, Keeler R, Curry RA, Cunjak RA, Munkittrick KR. On the ecology of slimy sculpin - a recipe for effective environmental monitoring. 135th Meeting of the American Fisheries Society, Anchorage, AK. 11-15 September 2005.

Brasfield SM, Keeler RA, McMaster ME, Tetreault G, Munkittrick KR. Determining optimal windows for assessing reproductive endpoints in slimy sculpin: what Cottus has taught us. 135th Meeting of the American Fisheries Society, Anchorage, AK. 11-15 September 2005.

Brasfield SM, Keeler R, Gray MA, Munkittrick KR. 2005. Developing population-level ecological risk assessment framework for small freshwater systems using small bodied fish. 10th Annual North Atlantic Chapter of SETAC meeting. 8-10 June 2005. Burlington, VT. USA.

Brasfield, S.M., R. Keeler, M.A. Gray and K.R. Munkittrick. 2004. Investigating

limitations and challenges to developing population-level ecological risk assessments for small freshwater systems. Fourth SETAC World

Congress. 14-18 November 2004. Portland, Oregon, USA (interactive platform).

Brasfield, SM, MA Gray, and KR Munkittrick. 2004. Using small-bodied fish in

effects-based assessments: interpreting non lethal data for use in environmental monitoring studies. 31st Aquatic Toxicity Workshop. 24-27 October 2004. Charlottetown, PEI, Canada.

Brasfield, SM, KR Munkittrick GR Tetreault and ME McMaster. 2004. Use of fish

populations in effects-based assessments: seasonal characterization of reproductive endpoints. North Atlantic Chapter of SETAC. Portsmouth, RI USA.

Brasfield, SM, MA Gray, LM Hewitt, and KR Munkittrick. 2004. Use of fish

populations in an effects based assessment to evaluate non point stressors associated with agriculture. Canadian Conference for Fisheries Research. 8-10 January 2004. St. John’s, NL Canada.

Brasfield, SM, MA Gray, LM Hewitt, and KR Munkittrick. 2003. Use of fish

populations in an effects based assessment to evaluate non point stressors associated with agriculture. 24th Annual SETAC North America Meeting 9-13 November 2003, Austin, TX USA

Brasfield, SM, MA Gray, and KR Munkittrick. 2003. Use of fish populations in an

effects based assessment to evaluate non point stressors associated with agriculture. Society of Environmental Toxicology and Chemistry Asia-Pacific. 29 September-1 October, Christchurch, NZ.

Munkittrick, KR, LM Hewitt, K Teather, D MacLatchy, G Van Der Kraak, S

Brasfield, M Gray, C Jardine and K Gormley. 2003. Quantification of sediment-associated EDSs in agricultural areas, and their potential biological impacts on fish. CNTC abstract CNTC Annual Research Symposium, March 25-26, 2003, Ottawa ON.

Brasfield, SM, Curry RA, Munkittrick KR. 2003. Identification of an upstream

source of contamination on the Saint John River near Clair, NB. Canadian Conference for Fisheries Research. 2-5 January 2003. Ottawa, ON.

Brasfield SM, Weber LP, Talent LG, Janz DM. Dose response and time course

relationships of vitellogenin induction in male Western fence lizards (Sceloporus occidentalis) exposed to ethinylestradiol. 12th Annual SETAC Europe Meeting 12-16 May 2002, Vienna, Austria.

Brasfield SM, Peters LE, Galloway B, Gray MA, Munkittrick, KR. Identification of

an upstream source of contamination near Claire, NB. 2nd Annual Saint John River Meeting. 24 January 2002, Saint John, NB.

Brasfield SM, Weber LP, Talent LG, Janz DM. Is plasma alkaline-labile

phosphate a suitable alternative to measuring vitellogenin? Ozark Prairie Regional Chapter SETAC Meeting, 18-20 May 2001, Stillwater, OK.

Brasfield SM, Eggert SL, Wallace JB. Effects of litter exclusion on headwater

stream metabolism in the southern Appalachian Mountains. Tennessee Academy of Science, November 1998, Cookeville, TN.

Brasfield, S.M., M.E. McMaster, and K.R. Munkittrick. 2004. Investigating

reduced reproductive performance in fish populations in potato growing areas of New Brunswick, Canada. Fourth SETAC World Congress. 14-18 November 2004. Portland, Oregon, USA.

Brasfield, SM, KR Munkittrick, and C Portt. 2004. Examining population-level

responses in small-bodied and short-lived fishes: what Cottus has taught us. 24-27 October 2004. Charlottetown, PEI, Canada.

Gray MA, SM Brasfield, and KR Munkittrick. 2004. The application of effects-

based assessment to study non-point source pollution in agricultural regions. 2nd Annual Canadian Water Network Symposium. 20-22 June. Ottawa, ON.

Brasfield SM, KR Munkittrick, and MA Gray. 2004. Using small-bodied fish in

effects-based assessments: non-lethal assessments downstream of agricultural activity. 2nd Annual Canadian Water Network Symposium. 20-22 June. Ottawa, ON.

Gray MA, SM Brasfield, and KR Munkittrick. 2004. The application of effects-

based assessment to study non-point source pollution in agricultural regions. Society of Experimental Biology. 29 April-2 May. Edinburgh, Scotland.

Chapman, P F; Brasfield, S; Carlsen, T; Elmegaard, N; Landis, W; Moe, S;

Nacci, D; Spromberg, J; Noel, H. 2004. Empirical Approaches to Population Level Ecological Risk Assessment and its Relationship to Mathematical Modeling. SETAC Europe Meeting, 18-22 April 2004, Prague, Czech Republic.

Brasfield, SM, Gray MA, Munkittrick KR. 2004. Non lethal effects-based

assessment of fish populations to evaluate non point stressors associated with agriculture. Environmental Effects Monitoring Symposium. 16-18 February 2004. Fredericton, NB.

Brasfield, SM, McMaster. ME, Portt, C, Munkittrick, KR. 2004. Characterization

of seasonal reproductive endpoints of slimy sculpin (Cottus cognatus) for

use in environmental monitoring and assessment. 16-18 February 2004. Fredericton, NB.

Chapman, P F; Brasfield, S; Carlsen, T; Elmegaard, N; Landis, W; Moe, S;

Nacci, D; Spromberg, J; Noel, H. 2003. Empirical Approaches to Population Level Ecological Risk Assessment and its Relationship to Mathematical Modeling. 24th Annual SETAC North America Meeting 9-13 November 2003, Austin, TX.

Brasfield SM, Haralampides K, Munkittrick KR, Haines C. Quantitative

examination of sediment transport and associated pesticide transport from agricultural areas. 12th Atlantic Region Hydrotechnical Conference of the Canadian Society for Civil Engineering. October 2-4, 2002 Charlottetown, PEI.

Brasfield SM, Bradham K, Wells JB, Talent LG, Lanno RP, Janz DM. Evaluation

of fence lizard (Sceloporus spp.) eggs as a terrestrial vertebrate model for assessing bioavailability and in ovo effects of soil contaminants. 22nd Annual SETAC North America Meeting 11-15 November 2001, Baltimore, MD.

Brasfield, S.M., L.P. Weber, L.G. Talent, and D. M. Janz. "Time course and

dose response relationships of vitellogenin induction in male Western fence lizards (Sceloporus occidentalis) exposed to ethinylestradiol. 21st Annual SETAC Meeting 11-16 November 2000, Nashville, TN.

Brasfield, S.M., L.P. Weber, L.G. Talent, and D. M. Janz. "Time course and

dose response relationships of vitellogenin induction in male Western fence lizards (Sceloporus occidentalis) exposed to ethinylestradiol. Ozark-Prairie Regional SETAC, 18-20 May 2000, Leavenworth, KS.

Brasfield SM, Eggert SL, Wallace JB. Effects of litter exclusion on headwater

stream metabolism in the southern Appalachian Mountains. Undergraduate Research Symposium, April 1999, Murfreesboro, TN.