Development And Application Of A Bioenergetics Model For Lake Washington Prickly Sculpin (Cottus asper)

by

Jamal Hasan Henry Moss

A thesis submitted in partial fulfillment of the requirements for the degree of

Master of Science

University of Washington

2001

Program Authorized to Offer Degree: School of Aquatic and Fishery Sciences

University of Washington Graduate School

This is to certify that I have examined this copy of a master’s thesis by

Jamal Hasan Henry Moss

and have found that it is complete and satisfactory in all respects, and that any and all revisions required by the final

examining committee have been made.

___________________________________________________ Dave Beauchamp

___________________________________________________ Reginald Reisenbichler

___________________________________________________

Walton Dickhoff

Date: ______________________

In presenting this thesis in partial fulfillment of the requirements for a Master’s degree at the University of Washington, I agree that the Library shall make its copies freely available for inspection. I further agree that extensive copying of this thesis is allowable only for scholarly purposes, consistent with "fair use" as prescribed in the U.S. Copyright Law. Any other reproduction for any purposes or by any means shall not be allowed without my written permission.

Signature___________________________________________________

Date_______________________________________________________

TABLE OF CONTENTS

List of Figures..........................................................................................…. iiList of Tables............................................................................................... iiiAcknowledgements..................................................................................…. ivDedication…............................................................................................…. v Chapter I: Model Development..................................................……… 1 Introduction………………………………… 1 Methods....................................................................…………... 4 Model Development.............................................................… 7 Sensitivity Analysis.............................................................…. 13 Results..............................................................………………... 13 Discussion..............................................................……………. 15 Figures.....................................................................………………………. 19Tables.....................................................................……………………….. 27 Chapter II: Model Application..........................................................…… 31 Introduction...........................................................…………… 31Study Area.............................................................………… 34Methods....................................................…………………...... 35Results.............................................................……………… 40 Discussion............................................................…………….. 43 References.............................................................................. 47 Figures.....................................................................……………………… 55Tables.....................................................................……………………….. 59

ii

LIST OF FIGURES

Number Page1. Consumption (weighted and unweighted)………………19

2. Temperature Dependence on Consumption…………….20

3. Specific Consumption…………………………………...21

4. Specific Consumption at 18ºC…………………………..22

5. Specific Respiration……………………………………..23

6. Specific Consumption for Two Species of Sculpin……..24

7. Proportion of Cmax for Two Species of Sculpin…………25

8. Specific Respiration for Two Species of Sculpin……….26

9. Individual Daily Ration…………………………………55

10. Growth Trajectories for Age 1-6 Sculpin……………….56

11. Growth Efficiency………………………………………57

12. Annual Consumption…………………………………...58

iii

LIST OF TABLES

Number Page

1. Energy Densities…………………………………………...27

2. Temperature Used In Sensitivity Analysis.………………...27

3. Proportion of Prey In Diet…………………………………28

4. Parameter Values…………………………………………..29

5. Parameter Sensitivity………………………………………30

6. Age Specific Abundance.…………………………………..59

7. Thermal Experience.……………………………………….59

8. Juvenile Sockeye Salmon Predation Limitations.………….60

9. Juvenile Chinook Salmon Predation Limitations.………….60

10. Prey Biomass Consumed…………………………………..61

11. Numerical Estimates of Predation……………...…………..62

iv

Acknowledgements

Laboratory space required for this work was made available through Reg Reisenbichler

and Jim Winton, U.S.G.S. Western Fisheries Research Center, Seattle, Washington. The

Washington State Cooperative Fish and Wildlife Research Unit provided laboratory

equipment and King County Water Quality provided funds for additional materials. I

offer my sincere gratitude to Jeffrey Elisoff and Jennifer McIntyre for assisting with

laboratory work, and to Mike Mazur for providing insight. The guidance provided by

Walt Dickhoff and Reg Reisenbichler was of paramount importance in the development

and organization of this thesis, and I thank both of them for their efforts. I thank Tom

Sibley for giving me an opportunity to study at this institution, and praise Dave

Beauchamp for his tireless patience and selfless devotion to education.

v

Dedication

To my grandfather, Henry Lee Moss Sr.

Chapter I

Model Development Introduction

Bioenergetics modeling is based on the laws of thermal dynamics where the

energy consumed by a fish must balance the energy lost through physiological processes

and growth. The terms of the energy budget can be measured directly in laboratory

experiments designed to characterize metabolism, maximum consumption rate, and

energy lost through waste. This modeling approach allows the user to account for energy

costs experienced by the fish and solve for the level of consumption consistent with the

observed growth which integrates the array of environmental conditions experienced by

the fish. Estimates of prey consumption are generated using estimates of predator

growth, diet composition, thermal experience, predator energy density, and prey energy

density. Estimated consumption can then be used to quantify important temporal, spatial,

and size-specific interactions within an aquatic community. Consumption requirements

and growth efficiency relative to ingested material can be simulated to determine

bottlenecks in growth due to prey availability or quality and thermal regime. The

bioenergetics modeling method is based on rates of energy intake, transformation, loss,

and use by the species of interest (Brett and Groves 1979), and provides a theoretical

framework for relating growth and feeding rates to ontogeny and environmental

conditions (Allen and Wootton 1982).

2

The Wisconsin bioenergetics model (Hanson et al. 1997) is based on a mass-

balance equation where consumption (C) equals the net gain or loss of weight (G), plus

metabolic costs (M), plus waste loss (W) accrued over a particular period. For each term

of this equation there are formulae that describe these processes in units of energy

(Joules/day), with the form and parameterization of each term specific to the physiology

or ontogenetic stage of a species. Previous analyses have indicated that functions

describing the effect of body mass and temperature on maximum consumption and

respiration contain the most sensitive parameters (Bartell et al. 1986). Lab experiments

should be designed to measure rates of consumption and respiration for various sizes of

fish at different temperatures to parameterize the size and temperature-dependant

functions. However, parameters are often borrowed from other species resulting in

reduced accuracy and confidence in this approach. It is important that parameters be

developed for a wide range of fish with careful attention to the experimental design of

maximum feeding and respiration experiments to increase the accuracy of bioenergetics

models.

Bioenergetics models have been constructed for aquatic organisms ranging from

omnivorous minnows (Schindler et al. 1993) to hyper-predaceous sea lampreys (Kitchell

1990). These models are commonly used to estimate consumption rates or growth (Rice

and Cochran 1984, Stewart and Ibarra 1991). Model predictions of consumption have

agreed with independent field estimates of consumption (estimates within +/- 10%) for

largemouth bass (Rice and Cochran 1984) and salmonids (Beauchamp et al. 1989,

3

Brodeur et al. 1992); however, bioenergetics models have not performed as well for

species modeled with parameters borrowed from other species (often from different

orders or families) (Post 1990, Boisclair and Leggett 1989, Wahl and Stein 1988).

Sculpin (Cottidae spp.) are commonly abundant in many freshwater, estuarine,

and marine assemblages, occupy multiple trophic levels, and represent important

components of many food webs. Prickly sculpin (Cottus asper) are known to feed on

other sculpins, juvenile salmonids, and aquatic invertebrates (Tabor et al. 1998, Jones

1986, Rickard 1980, Ikusemiju 1975), and are an important food source for some

piscivorous fish (Wydoski and Whitney 1979). Prickly sculpin in Lake Washington have

been suspected of limiting the abundance of sockeye salmon (Tabor et al. 1998), a

commercially, recreationally, and culturally important species.

The objectives of this study were to parameterize a bioenergetics model for

prickly sculpin within the framework of the Wisconsin Bioenergetics Model 3.0 (Hanson

et al. 1997), evaluate the sensitivity of the parameters with individual parameter

perturbations of (+/-) 10% (Kitchell et al. 1977, Bartell et al. 1986), and evaluate the

generality of this model by comparing maximum consumption and respiration of prickly

sculpin to that of coastrange sculpin (Cottus aleuticus) over a limited range of

temperatures and body sizes.

4

Methods

Prickly sculpins were collected from Lake Washington and transported to USGS

Western Fisheries Research Center, Seattle, Washington, where their identification was

confirmed (Wydoski and Whitney 1979). The fish were sorted into six size classes: 2-

3g, 4-5g, 8-9g, 11-13g, 15-18g, and 20-30g. Prickly sculpin were acclimated to one of

the nine temperatures used for feeding and respiration experiments: 6º, 9º, 10º, 12º, 15º,

16.5º, 18º, 21º, and 24ºC. A natural photoperiod was provided with a skylight to avoid

stress associated with abrupt changes in light level from artificial lights.

Consumption Experiments

Feeding and respiration experiments were performed on prickly sculpins to

estimate key parameters used in the Wisconsin bioenergetics model 3.0 (Hanson et al.

1997). Preliminary maximum consumption (Cmax) experiments revealed the importance

of feeding live prey, thus prickly sculpins were fed a diet of white clouds (Tanichthys

albonubes), guppies (Lebistes reticulata), goldfish (Carassius auratus), or a combination

of the three. The smallest size class was fed white clouds exclusively, and the largest size

class was fed a combination of guppies and goldfish. Two prickly sculpins from each size

class were weighed and placed in a 2-liter non-toxic plastic feeding chamber immersed in

a temperature-controlled water bath. Prickly sculpins were held without food at each

experimental temperature for 72h before a 24h feeding trial. Fish were fed four pre-

weighed offerings of excess food over the 24h feeding trials. Unused prey were removed

and reweighed after each trial. Identical experiments were performed on a subset of

5

coastrange sculpin so that a comparison in feeding response between the two species

could be made. No aggressive interactions between paired fish were observed, and both

were examined for bulging stomachs after feeding trials to ensure that each sculpin had

eaten. Cmax was expressed as g food/(g body weight*day).

Respiration Experiments

Standard metabolism, the metabolism of a fasting fish at rest, defines the baseline

or lower bound for metabolic costs (Cech et al. 1985), and is therefore an important

variable in the bioenergetics budget. Standard metabolism is measured by converting the

amount of oxygen consumed over a particular period into calories or Joules with the

oxycaloric average of 3.24 cal/mg O2 (Elliott and Davidson 1975). Routine metabolism

was measured for six size classes of prickly sculpin at nine temperatures. Closed, static,

respirometers were constructed from canning jars, aquarium tubing, stopcock valves and

a YSI-52 dissolved oxygen (DO) meter (Edwards and Cech 1990). The YSI-52 DO

probe was fit into the center of a canning jar lid and made watertight with O-rings and

aquarium seal. Two smaller holes were drilled into the lid, and stopcock valves were

sealed in place and attached to a temperature-controlled water supply. A rubber gasket

was added to the respirometer lid to ensure water tightness.

The DO probe required flow across the membrane for proper operation, so a

magnetic stir bar was added to gently circulate water throughout the chamber. Since the

magnetic stir bar could potentially increase stress and elevate respiration rate, a cover was

6

designed to minimize this effect. I accounted for the oxygen depleted by microbes in the

chamber by subtracting the amount of oxygen lost in trials without fish (blanks), from the

amount of oxygen lost during trials with fish. Oxygen loss as a result of bacteria on the

fish was considered trivial and was ignored.

Prickly sculpins were held at the experimental temperature for a minimum of 3

days before respiration trials. The respirometers were submerged in a temperature-

controlled water bath, and prickly sculpins were allowed to adjust to the respiration

chamber during a 30-minute acclimation period prior to a trial. Oxygenated water was

circulated through the chamber via aquarium tubing and stopcock valves during the

acclimation period. The valves were closed after 30 minutes, and measurements of

dissolved oxygen and temperature were recorded every 5 minutes for 30 minutes. DO

was not allowed to fall below 3mg/l during a trial to ensure that oxygen levels were

measured accurately and that DO did not decline to levels that would stress the fish.

Bomb Calorimetry

Energy densities (Joules/g wet weight) for prickly sculpin and prey fish (0.5-2.5g)

were determined with bomb calorimetric methods. Live sculpins and prey fish were

transported from the USGS Western Fisheries Research Center (Seattle, WA) to the

University of Washington, pithed, blotted, weighed, dried in an oven, reweighed, ground

in a blender, pelletized, and combusted in a Parr 1425 semi-micro bomb calorimeter.

7

Model Development

The Wisconsin bioenergetics model (Hanson et al. 1997) is an energy balance

equation wherein consumption C must equal metabolic costs M, waste W, and growth

(somatic and gonadal) G.

C = M + W + G

Consumption and metabolism vary as functions of body mass and temperature,

and waste varies as a function of consumption. In the most common application, the

model estimates consumption (C) by iteratively fitting the proportion (P) of maximum

consumption (Cmax) required to satisfy the change in body mass observed over a specified

time interval for fish of a given size under a specified temperature regime (Hanson et al.

1997).

C = Cmax * P * F(T)

A series of feeding-to-satiation experiments over a range of temperatures and

body masses were used to parameterize the equations for Cmax. Maximum consumption

is typically a power function of body mass and has a dome-shaped response to

temperature (Hanson et al. 1997). These different response curves complicate the

parameterization of Cmax to body mass and temperature. To simplify the

parameterization, I first identified the temperatures where Cmax was highest for most sizes

8

of prickly sculpin, and then parameterized the weight-dependence function for

consumption at that temperature. For any other temperature, the weight-dependent Cmax

would be reduced by the temperature-dependant function of consumption that varies from

0.0 to 1.0

Because of the important relationship between the mass of prey consumed and the

total energy consumed, it was important to determine whether the consumption functions

in the Wisconsin bioenergetics model should be parameterized using prey mass or total

energy as the response variable. The Wisconsin model uses the mass of food, but

accounts for total energy contained in a given mass of food via the energy density, and

proportional contribution of each prey category in the diet. If consumption is limited by

energy rather than the mass of food consumed, then the model parameterization would be

affected by the quality of food (energy density) used for consumption experiments to

establish Cmax. This evaluation was based on the fit and shape of Cmax functions for

weight and temperature-dependence.

Cmax = CA * W CB

C specific consumption rate (g*g-1*d-1)

Cmax maximum specific feeding rate (g*g-1*d-1)

P proportion of maximum consumption

9

F(T) temperature dependence function

T water temperature (ºC)

W fish mass (g)

CA intercept of the allometric mass function (g*g-1*d-1)

CB slope of the allometric mass function

Absolute consumption (g*d-1) increased with increasing body size at all

temperatures (Figure 1a). This relationship deteriorated when consumption was adjusted

to the mass of prey at a standardized energy density of 4185 J/g (Figure 1b), thus

establishing that prey mass, rather than energy, is the proper currency for developing the

consumption functions.

The response of specific consumption to temperature was characteristic of a

response for cool- and cold-water species (Thornton and Lessem 1978) (Figure 2). The

Thornton-Lessem curve is the combination of two sigmoid curves, with one curve fit to

the portion of the data where consumption increases with temperature, and the other

curve where consumption decreases with temperature. Temperature dependence of Cmax

was estimated over seven temperatures using 7g prickly sculpins. The 7g prickly

sculpins were used to estimate temperature-dependent consumption parameters because

they were just large enough to be relatively insensitive to the body-size effect on

consumption. The temperature at which 98% of maximum consumption occurred (CTO

10

& CTM) was estimated to be 18ºC. We designated CQ = 6º as the temperature at which

consumption was a small fraction (CK1 = 0.1075) of peak consumption. Our warmest

experimental temperature was designated as CTL = 24ºC, a temperature warmer than

CTO and CTM where consumption was some reduced fraction (CK4 = 0.8129) of peak

consumption. The effect of weight on specific consumption (g prey consumed/(g

sculpin*day)) at 6º, 12º, 18º, and 21 ºC suggested a decreasing power function (Figures 3

and 4). Since peak consumption occurred at 18ºC, the weight-dependence at 18ºC was

used for estimating the final values of the intercept CA = 0.2324 and CB = -0.514 of the

decaying power function (r2 = 0.8702, p = 0.004)

Cmax = 0.2324 * W –0.514

Respiration

Respiration was used to describe metabolic costs as a function of body mass and

temperature. Metabolism is influenced by activity rate and specific dynamic action

(SDA). The total metabolic rate of a fish is estimated by adding SDA to respiration costs

(R).

R = RA * W RB * e RQ*T

S = SDA * (C-F)

11

R specific rate of respiration (gO2*g-1*d-1)

W fish mass (g)

RA intercept of the allometric mass function (g*g-1*d-1)

RB slope of the allometric mass function

RQ rate at which function increases over low temperatures

T water temperature (Cº)

S assimilated energy lost to specific dynamic action

SDA specific dynamic action (proportion of assimilated energy lost to digestion)

C specific consumption rate (g*g-1*d-1)

F specific egestion rate (g*g-1*d-1)

Specific respiration (gO2*g-1*d-1) declined as a decaying power function of body

mass, but increased exponentially as a function of temperature (Figure 5). A multiple

regression was performed on the natural log of specific respiration versus the natural log

of weight, and of temperature to obtain parameters RA, RB, and RQ (R2 = 0.7730, p <

0.001)

.

R = 0.0021 * W(-0.124) * e(0.0616) * T

The proportion of energy lost to SDA generally ranges between 0.17 and 0.18 (Hanson et

al. 1997), so I used an average of 0.175 in this model.

12

Egestion and Excretion

From average rates of waste losses W in the literature (Hanson et al. 1997), I

assumed that egestion F was as a constant portion of consumption (FA = 0.16), and

excretion computed as a constant proportion (UA = 0.10) of assimilated energy

(consumption – egestion).

W = 0.16 * C + 0.10 (C - 0.16 * C)

W = F + U

F = FA * C

U = UA * (C - F)

F specific egestion rate (g*g-1*d-1)

FA a constant proportion of consumption lost as feces

U specific excretion (g*g-1*d-1)

UA a constant proportion of assimilated energy lost as urine

Calorimetry

Calometric estimations of prey fish indicated that goldfish contained the highest

water content and lowest energy density, whereas guppies had the lowest water content

and highest energy density (Table 1). The energy density was 4151 Joules/g for age 1

prickly sculpin, and 4532 Joules/g for age 2-6 (Table 1). The mean energy density for

13

age 1 prickly sculpin was 4317 Joules/g, age 2-6 prickly sculpin 4532, white clouds 4014

J/g, guppies 6430 J/g, and goldfish 2679 J/g (Table 1).

Sensitivity Analysis

The sensitivity of prickly sculpin physiological parameters to estimates of annual

consumption was evaluated by changing parameters +/-10% or +/- 1 ºC for temperature

parameters. One parameter was changed during each simulation, while the others were

maintained at their nominal values (Bartell et al. 1986). The model was then allowed to

estimate the proportion of maximum consumption (P-value), and calculate the weight (g)

of prey consumed over a 1-year simulation for a 3 year-old sculpin with an initial weight

of 15.7g and final weight of 26.1g. The seasonal thermal experience (Table 2) and diet

(Table 3) used in the simulation reflected conditions believed to be experienced by the

age 3 cohort in Lake Washington based on life history information from Rickard (1980).

Results

Sensitivity Analysis

The nominal values for all physiological parameters (Table 4) fell within the

range of values reported for other species (Hanson et al. 1997). The most sensitive

parameters in the model were those associated with the weight and temperature

dependence of respiration (Table 5). A 10% parameter perturbation of RA yielded a

change of 9.26% or -9.06% in annual consumption and a 10% perturbation of RB yielded

14

a change of 4.04% or -3.82% in annual consumption (Table 5). A 10% perturbation in

RQ yielded a change of 5.19% or -4.87% in annual prey consumption (Table 5).

Parameters associated with the effect of size and temperature on consumption were

relatively insensitive, with changes in annual prey consumption ranging from 0 to 2.86%.

CQ was the most sensitive of the parameters associated with consumption, with annual

prey consumption ranging from 2.86 to -1.84% (Table 5). P-values were very high, often

exceeding 1.0 (Table 5).

Energy Budget

The annual energy budget for prickly sculpin (Joules/year) allocated more energy

to respiration costs and less to waste and growth when compared to fishes growing at a

“typical” rate (Brett and Groves 1979).

Consumption = Respiration + Waste + Growth

General Carnivore: 100 = 44 + 27 + 29

Prickly Sculpin: 100 = 59 + 24 + 17

The growth efficiency (mass assimilated in growth/mass consumed) for age 1 prickly

sculpin experiencing a diet and thermal experience typical for Lake Washington, and

growing from 1.0g to 7.2g was 17.2%. The resulting annual energy budget (Joules) for

age 1 prickly was as follows:

15

Consumption = Respiration + Waste + Growth

49743 = 29038 + 12137 + 8567

Physiological Comparisons Between Prickly and Coastrange Sculpin

When comparisons were possible, maximum consumption was higher for

coastrange sculpin than for prickly sculpin of the same size at 6ºC, but was very similar at

higher temperatures (Figure 6). Coastrange sculpin had a higher tolerance to temperature

than did prickly sculpin, and consumed more food than did prickly sculpin at cooler

temperatures. Coastrange sculpin consumed the largest proportion of food at the highest

experimental temperature (24ºC), therefore, the temperature at which maximum

consumption occurs for coastrange sculpin could not be estimated (Figure 7). Specific

respiration values for coastrange sculpin were also higher than those for prickly sculpin

(Figure 8).

Discussion

The characteristic relationship for consumption in relation to weight and

temperature deteriorated when prey was standardized to 4185 J/g, indicating that weight

of food consumed has a more dominant effect than does quality of food when

constructing the model. This suggests that stretch receptors play the most important role

in prickly sculpin feeding, and that food mass, and not energy consumed is the correct

currency for constructing the model. This highlights the importance of using accurate

estimates of prey energy density when applying the model to a population in order to

16

predict consumption. Proportions of maximum consumption (P-values) can fluctuate for

the same mass of food consumed if energy densities differ significantly. The wide

variety of prey found in Lake Washington prickly sculpin diet is characteristic of an

opportunistic feeder, and suggests the need for additional work to determine these

relationships.

Resting metabolism was substituted for routine metabolism in the model because

prickly sculpins were generally observed lying motionless on substrate in their natural

environment and in holding tanks. However, previous studies have shown that sculpins

migrate to regions where food is more abundant on a seasonal basis (Ruzycki and

Wurtzbaugh 1999), and various life stages of some sculpin species undergo diel vertical

migration (Ikusemiju 1975, Neverman and Wurtzbaugh 1994). Further studies describing

sculpin activity are needed in order to include this component in the model. Standard

error associated with respiration and consumption increased with fish size because of

greater differences in body weight within the larger size classes. Oxygen consumption

was most variable at 24°C for all size classes, and is probably an effect of holding prickly

sculpins at a near fatal temperature.

After comparing empirically estimated prickly sculpin energetic values to values

estimated for other sculpin species, prickly sculpin appeared to be less energy dense than

other sculpin. Energy density values reported in the literature for other Cottids ranged

from 5,209 Joules/g (Cottus perplexus; Brocksen et al. 1968) to 5,736 J/g (Cottus

17

cognatus; Rottiers and Tucker 1982). We found the energy density of prickly sculpin to

vary with body weight, with an average energy density of 4151 J/g for age 1 fish, and

4532 J/g for age 2-6 fish.

Similarities in responses of prickly sculpin and coastrange sculpin suggest that

our model may be used for other Cottids, however, there were differences between the

two species that should be considered when making parameter substitutions. Coastrange

sculpin had higher metabolic costs than did prickly sculpin, which may be somewhat

compensated for by a higher temperature dependence, especially at low temperatures.

These differences can be explained by differences in the life history and the preferred

habitat of the two species. Both coastrange and prickly sculpin are found in the coastal

region from California to Alaska, however, the two species are usually ecologically

separated, with the coastrange sculpin more commonly occupying streams with moderate

current and the prickly sculpin preferring lakes or slow moving stream habitats (Wydoski

and Whitney 1979). An ability to consume more prey over a wider range of temperatures

would be consistent with the relatively higher currents and wider temperature fluctuations

expected in habitats occupied by coastrange sculpin.

Responses of consumption and respiration to changes in body mass and

temperature were similar to those for other cool and cold-water species, with all

estimated parameter values falling within the range of values reported for other fishes

(Hanson et al. 1997). This consistency suggests that the parameters for prickly sculpin

18

are reasonable. My experimentally derived physiological parameters for consumption

and respiration of prickly sculpin consumption provide an initial template for

characterizing the physiology of a widely abundant and important genus of fish in

northern temperate waters.

19

Figures

W eight (g )

0 5 10 15 20 25 30

Con

sum

ptio

n (g

/d)

0

1

2

3

6 (C ) 9 (C ) 12 (C ) 15 (C ) 18 (C ) 21 (C ) 24 (C )

A

W eight (g)

0 5 10 15 20 25 30

Con

sum

ptio

n (g

/d)

0 .0

0.2

0.4

0.6

0.8

1.0

6 (C ) 9 (C ) 12 (C ) 15 (C ) 18 (C ) 21 (C ) 24 (C )

B

Figure 1. Absolute consumption (g prey consumed/day) by prickly sculpins (a). Absolute consumption (g prey consumed/day) weighted by caloric content to 4185 J/g (b).

20

Temperature (C)

0 2 4 6 8 10 12 14 16 18 20 22 24 26

Prop

ortio

n of

Cm

ax

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Figure 2. Proportion of maximum consumption in relation to temperature (ºC) for 7g prickly sculpin.

21

Weight (g)

0 5 10 15 20 25 30

Spec

ific

Con

sum

ptio

n (g

/(g*d

))

0.00

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

18ºC

21ºC

12ºC

6ºC

Figure 3. Specific consumption (g prey consumed/(g sculpin*day)) as a function of weight (g) at 12, 15, 16.5, and 18ºC.

22

Weight (g)

0 5 10 15 20 25 30

Spec

ific

Con

sum

ptio

n (g

/(g*d

))

0.00

0.02

0.04

0.06

0.08

0.10

0.12

Figure 4. Specific consumption (g prey consumed/(g sculpin*day)) in relation to body weight (g) for prickly sculpin at 18ºC (r2 = 0.8702).

23

Temperature (C)

4 6 8 10 12 14 16 18 20 22 24 26

Spec

ific

Res

pira

tion

g O

xyge

n/(g

*d)

0.000

0.002

0.004

0.006

0.008

0.010

0.012

0.014

2.73 (g) 4.62 (g) 6.48 (g) 12.82 (g) 15.88 (g) 25.77 (g)

Figure 5. Specific respiration (g O2 consumed/(g sculpin*day)) vs. Temperature (ºC) for six size classes of prickly sculpin (multiple regression r2 = 0.7730).

24

W e ig h t (g )

0 1 2 3 4 5 6 7 80 .0

0 .1

0 .2

0 .3

0 .4

Spe

cific

Con

sum

ptio

n g/

(g*d

)

0 .0

0 .1

0 .2

0 .3

0 .4

0 .0

0 .1

0 .2

0 .3

0 .4 6 ºC

1 8 ºC

2 4 ºC

c oas tran g ep ric k ly

Figure 6. Specific consumption (g/(g*d)) coastrange sculpin and prickly sculpin at 6º, 18º, and 24ºC. Error bars represent 1 standard error.

25

Temperature (C)

4 6 8 10 12 14 16 18 20 22 24 26

Prop

ortio

n of

Cm

ax

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

Coastrange Prickly

Figure 7. Proportion of maximum consumption for 4.5g coastrange sculpins and prickly sculpins relative to temperature (ºC). Error bars represent 1 standard error.

26

Weight (g)

0 5 10 15 20 25

Spec

ific

Res

pira

tion

(g O

2 /(g*

d))

0.000

0.002

0.004

0.006

0.008

0.010

0.012

Coastrange 10ºC Prickly 9ºC Coastrange 22ºC Prickly 21ºC

Figure 8. Specific respiration (gO2/(g*d)) for coastrange and prickly sculpin at 9-10º and 20-21º C.

27

Tables Table 1. Mean energy density (J/g) and dry/wet ratio of prickly sculpin and prey fish used in feeding experiments. Because of their small size, multiple guppies and white clouds were homogenized and analyzed as composite samples in the bomb calorimeter (**).

Fish Wet body mass (g)

Dry/wet ratio

N Wet wt. energy density (J/g)

Prickly (age 1) 3.31 0.22 32 4317 Prickly (age 2-6) 18.9 0.22 32 4532 Goldfish 0.96 0.17 6 2679 Guppy 0.14 0.27 ** 6430 W. Cloud 0.03 0.20 ** 4014

Table 2. Temperature values for age 3 prickly sculpin used in sensitivity analysis.

Month Temperature (ºC) January 8.25 February 7.00 March 7.00 April 8.25 May 11.00 June 13.75 July 7.50

August 7.50 September 7.50

October 7.50 November 8.00 December 9.00

28

Table 3. Proportion of prey in Lake Washington prickly sculpin diet categorized by sculpin age and season (Rickard 1980). Miscellaneous invertebrate category includes Amphipoda, Isopoda, Decapoda, and Diptera pupae.

Age Season Fish 4500 J/g

Neomysis 3642 J/g

Chironomid 4428 J/g

Trichoptera 4429 J/g

Mollusca 2500 J/g

M. Invert. 4428 J/g

1 Winter 0.00 0.23 0.69 0.00 0.00 0.08 Spring 0.00 0.04 0.87 0.00 0.00 0.09 Summer 0.00 0.06 0.88 0.00 0.00 0.06 Autumn 0.00 0.13 0.77 0.00 0.00 0.10

2-3 Winter 0.05 0.34 0.30 0.05 0.08 0.08 Spring 0.03 0.06 0.38 0.28 0.07 0.18 Summer 0.05 0.12 0.50 0.14 0.03 0.16 Autumn 0.03 0.18 0.31 0.03 0.15 0.20

4-6 Winter 0.23 0.34 0.09 0.01 0.05 0.29 Spring 0.20 0.09 0.16 0.09 0.06 0.40 Summer 0.29 0.15 0.18 0.04 0.02 0.32 Autumn 0.46 0.14 0.07 0.01 0.08 0.25

29

Table 4. Parameter values for prickly sculpin estimated from consumption and respiration experiments.

Functions used in Wisconsin Model

Parameter description

Nominal value

Consumption

CA Intercept: Cmax (g · g-1 · d-1) 0.2325 CB Coefficient: Cmax versus weight -0.514 CQ Temperature for K1 (°C) 6

CTO Temperature for K2 (°C) 18 CTM Temperature for K3 (°C) 18 CTL Temperature for K4 (°C) 24 CK1 Proportion of peak Cmax at CQ 0.1075 CK4 Proportion of peak Cmax at CTL 0.8129

Respiration

RA Intercept: R (g O2/d) 0.0021 RB Coefficient: R versus weight -0.1240 RQ Coefficient: R versus temperature 0.0616

SDA Proportion: assimilated energy lost to digestion 0.175 Egestion and Excretion

FA Constant proportion of consumption 0.16 UA Constant proportion of assimilated energy 0.10

30

Table 5. Sensitivity analysis of the prickly sculpin model for annual consumption of an age 3 sculpin growing from 15.7g to 26.1g with the thermal experience presented in Table 2. For the nominal run, total consumption was 113.98g and P-value = 0.95017. % Difference in consumption Parameter Pertubation (+) Pertubation (-) Pertubation (+) P-value (-) P-value

CA +/- 10% 0.00 0.00 0.97 1.19 CB +/- 10% 0.52 -0.54 1.26 0.91 CQ +/- 1ºC 2.86 -1.84 1.39 0.86

CTO +/- 1ºC -0.01 -0.11 1.14 1.00 CTM +/- 1ºC 0.01 -0.01 1.07 1.07 CTL +/- 1ºC -0.01 0.01 1.07 1.07 CK1 +/- 10% -0.48 0.55 1.02 1.13 CK4 +/- 10% -0.01 0.00 1.07 1.07 RA +/- 10% 9.26 -9.06 1.16 0.98 RB +/- 10% -3.82 4.04 1.03 1.11 RQ +/- 10% 5.19 -4.87 1.12 1.02

SDA +/- 10% 2.47 -4.10 1.09 1.02

31

Chapter II

Model Application

Introduction

Sculpins (Cottidae spp.) represent a significant component of freshwater fish

fauna in the Pacific Northwest (Bond 1963), and feed at multiple trophic levels (Rickard

1980, Jones 1986, Tabor et al. 1998). In Lake Washington, prickly sculpin (Cottus

asper) represent approximately 80% of the total fish biomass, and are both predators and

prey (Eggers et al. 1978, Rickard 1980, Tabor et al. 1998). Therefore, prickly sculpins

should influence trophic interactions in the lake. Prickly sculpins cannibalize smaller

prickly sculpins and feed on other fishes, insects, crustaceans, and other invertebrates

(Rickard 1980). Consumption of juvenile sockeye salmon by prickly sculpins has been

observed in the Cedar River and throughout Lake Washington (Eric Warner,

Muckleshoot Tribe and Roger Tabor, U.S. Fish and Wildlife, personal communication),

and is a topic of special concern to fisheries managers.

Predation plays an important role in regulating prey populations in aquatic

systems (Post et al. 1998, Elser et al. 1995). In Lake Washington predation was

implicated as a possible cause for reduced survival of juvenile sockeye salmon and

longfin smelt during the 1970s and 1980s (Eggers et al. 1978, Swartzman and

32

Beauchamp 1990, Beauchamp 1994). Although sculpin are primarily benthic, predation

on mobile prey may occur during various seasons and times of day. Sockeye salmon fry

enter Lake Washington as early as January and generally migrate to sea in May of the

following year. After sockeye salmon fry leave the Cedar River, they are present in

nearshore areas for a short time before moving offshore. Martz et al. (1996) reported that

some fry remain in littoral areas during winter and spring. Sockeye salmon presmolts

congregate along the shore before leaving the lake in spring (Martz et al. 1996) and could

be vulnerable to large prickly sculpin. Juvenile and larval fish in benthic regions could

also be particularly vulnerable to prickly sculpin predation. Age 3-6 (total length (TL) =

110-165mm, weight = 15.7-59.7g) prickly sculpin display a seasonal offshore migration

during the summer and fall (Rickard 1980), which should increase the amount of

predation on organisms inhabiting the profundal zone. Juvenile sockeye salmon move

offshore after entering the lake during May-July (Martz et al. 1996), and exhibit diel

vertical migrations throughout the year (Woodey 1972, Eggers et al. 1978). Prickly

sculpin predation may significantly affect the survival of juvenile sockeye in the

profundal and slope zones throughout the year, and fry and presmolts in nearshore

regions during winter and spring.

Juvenile chinook salmon use the lake as a migratory corridor to Puget Sound, and

rear in the littoral zone of Lake Washington for up to 6 months (Tabor and Piaskowski

2001). Even though juvenile chinook salmon spend less time in Lake Washington than

juvenile sockeye salmon, predation by prickly sculpin may have detrimental effects on

33

their population because of their closer association with littoral regions in spring and

summer. Bioenergetically based food web models coupled with direct sampling for diet,

growth, size structure, thermal experience, and abundance can be used to quantify

temporal and size-specific consumption (Ney 1990). The bioenergetics modeling

approach quantifies trophic interactions by estimating the biomass of prey consumed by

individuals or populations of predators (Ney 1993). If the size structure of prey in the

diet is known, then the biomass of prey consumed can be converted into numerical losses

of prey. Bioenergetically based studies have proven useful for quantifying the

consumption demand of specific populations of fish (Stewart et al. 1981, Brandt and

Hartman 1993, Ney 1993). Specific applications have been used to predict the effect of

increased predation (LaBar 1993, Negus 1995), decreased predation (Rand and Stewart

1998), and energy conversion and gross production rates (Rand and Stewart 1998).

Bioenergetics modeling based on existing information for diet, distribution, and growth

of prickly sculpin in Lake Washington can provide estimates of predation rates on

juvenile sockeye salmon and other important prey species.

The primary objective of this study was to apply a bioenergetics model to

estimate consumption rates by prickly sculpin for their major prey in Lake Washington.

The second objective was to investigate the amount of juvenile salmon that might be

consumed by prickly sculpin by combining literature information with model results.

These simulations will provide an exploratory examination into whether prickly sculpins

are capable of imposing significant mortality on juvenile salmon or on other key species

34

in Lake Washington (e.g. Neomysis, larval fishes, and self regulation through

cannibalism). The third objective was to use this model to contrast simulated growth

under epilimnetic conditions, hypolimnetic conditions, and a thermal experience inferred

from onshore and offshore migrations to investigate the effects of spatial and temporal

distribution on growth.

Study Area

With a total surface area of 54.4 square kilometers, Lake Washington is the

second largest lake in Washington State (Comita and Anderson 1959). The lake is 32.2

km in length with an average width of 2.5 kilometers. Lake Washington is an oligo-

mesotrophic, monomictic lake with a mean depth of 33 m and a maximum depth of 66 m.

The shallowest regions are located at the north, south, and southeastern ends, with the

deepest regions located in the center. The Vashon ice sheet formed the steep marginal

slopes (Gould and Budinger 1958) that run longitudinally along the narrow body of

water. Water temperatures drop to 7°C in winter, and full water column mixing occurs

from December through April. Temperatures range from 7-22ºC in the epilimnion and 7-

9ºC in the hypoliminon (King County unpublished data). The lake has four primary

benthic habitat types: mud, sand, gravel, and silt.

35

Methods

Consumption and growth efficiencies for prickly sculpin were estimated with a

bioenergetics model (version 3.0 Hanson et al. 1997) parameterized for prickly sculpin.

Diet, growth, and distribution data were obtained from a life history study of Lake

Washington prickly sculpin (Rickard 1980). Physiological parameters used in the

bioenergetics model were derived from experiments performed on prickly sculpins from

Lake Washington (Chapter I). Consumption was fitted to annual growth estimates (Table

6) using seasonal and size specific proportions of prey in the diet (Table 3) from Rickard

(1980). Simulations began with January 1 and continued for 365 days for each size/age

class. Spawning losses of 1.7% (age 1), 5.0% (age 2), and 5.4% (age 3-6) of total body

weight were simulated as a single step occurring on June 1. Age-0 prickly sculpin were

not modeled because of insufficient data. Simulations were run with the proportion of

maximum daily consumption (P-value) fitted to annual growth increments for prickly

sculpin by the model. The biomass of each prey consumed was reported in the model

output.

I applied the bioenergetics model to life history and diet information from the mid

1970s (Rickard 1980) and to more recent diet information from the southern region of

Lake Washington (Tabor et al. 1998), where most sockeye salmon fry enter the lake

during winter and spring. Thermal experience was inferred from depth distributions of

prickly sculpin and thermal profiles from the lake. I conservatively assumed that prickly

sculpins did not prey on juvenile salmon larger than 50% of their body length (Juanes

36

1994, Persson 2000, Jones 1986). The potential magnitude of losses on prey items of

particular interest (sockeye salmon, chinook salmon, sculpin, mysids, and chironomid

pupae) were estimated per 1000 prickly sculpins, which incorporates the size structure of

the population reported by Rickard (1980). Annual survival rate for sculpin was assumed

to be 24% or 34% (Rickard 1980). Annual growth estimates were entered into the model,

and seasonal growth was interpolated by the Wisconsin Model version 3.0 (Hanson et al.

1997).

Rickard (1980) collected prickly sculpins from each of four major bottom habitats

from December 4, 1974 to December 5, 1975 with baited minnow traps. Smaller prickly

sculpins (ages 1-2, TL 40-100mm) remained in littoral habitats throughout the year,

whereas larger (TL >100mm) individuals moved into profundal habitats after spawning

in spring, and remained there until thermal destratification in November. Prickly sculpins

spawned in littoral regions during April-July, with peak spawning during the first week of

June. Most ripe prickly sculpins were found in the littoral zone, while only a small

fraction with a high gonad index (g gonad/g body weight) were found in the profundal

zone. Fifty percent of females and 12.5% of males were sexually mature by age 1, and

94.3% of females and 19.2% of males were sexually mature by age 2. All prickly

sculpins were sexually mature by age 3. The reported percent of body weight allocated to

gonadal material was 7.8% for females, 3.1% for males, and the average percentage of

body mass lost due to spawning for both sexes was 5.4%.

37

Thermal experience of prickly sculpins was inferred from temporal and spatial

movements (Rickard 1980) and monthly thermal profiles in Lake Washington. Each age

cohort was modeled with epilimnetic temperatures, hypolimnetic temperatures, and a

thermal experience inferred from seasonal onshore and offshore migrations. For age 3-6

prickly sculpins, thermal experience was modeled using epilimnetic temperatures from

November-June and hypolimnetic temperatures from July-October. Age 1-2 prickly

sculpins remained in the littoral zone and experienced epilimnetic temperatures

throughout the year (Table 7).

Rickard (1980) provided size-specific diet information and seasonal diet

information pooled over all sizes of prickly sculpin. Seasonal proportions of prey pooled

over all age classes were multiplied by size-specific proportions of prey. The resulting

values were scaled such that seasonal and size-specific proportions of prey in diet

summed to one (Table 3). Sculpin became piscivorous at age 2-3 (4% diet) and by age 4-

6, fish composed 30% of their diet. Out of 348 prickly sculpin stomachs, all but one of

the fish prey were prickly sculpins (Rickard 1980). Tabor et al. (1998) found five

sockeye salmon fry in the stomach contents of 262 prickly sculpin ≥50mm from littoral

samples collected in ≤10m of water along the southern end of Lake Washington during

peak fry migration in March-April. Juvenile sockeye salmon fry composed less than 5%

of the fish fraction of the diet reported by Tabor et al. (1998), and composed 0% of the

fish fraction of the diet by Rickard (1980). I assumed that juvenile sockeye salmon

composed 1% of the fish fraction of prickly sculpin diet based on this evidence.

38

Energy densities for prickly sculpin were measured using a Parr 1425 Semi-micro

Calorimeter, and values for prey items were obtained from Cummins and Wuycheck

(1971). Energetic values (J/g wet weight) for prickly sculpin used in the model were

4151 J/g for age 1 (1-7.1g), and 4532 J/g for age 2-6 (7.2-59.7g) prickly sculpin. Prey

energy densities (J/g wet weight) used in the model are as follows: Neomysis (3642 J/g),

Mollusca (2500 J/g), Chironomidae pupae (4428 J/g), Trichoptera (4429 J/g), fish (4500

J/g), and miscellaneous invertebrates (4428 J/g).

Rickard (1980) determined size at age by examining 1,108 otoliths (Table 6).

Weight (W) was calculated from total length (TL)(mm) by the equation (n = 2,156; r2 =

0.967) (Rickard 1980):

W = 2.63*10-6*TL 3.3185

I used seasonal growth increments generated by the model to estimate seasonal lengths

because Rickard (1980) only provided annual weight by age. Seasonal lengths were then

used to determine which age classes of prickly sculpin were capable of preying upon

juvenile salmon on a seasonal basis by conservatively assuming that prickly sculpin could

consume salmonid prey up to 50% of their own body length. If a sculpin lost weight

during a season, the length estimate from the previous season was substituted.

39

To determine seasonal susceptibility to predation, seasonal changes in size of

juvenile sockeye salmon (Woodey 1972, Chigbu and Sibley 1994) and chinook salmon

(M. Koehler, University of Washington, unpublished data), were compared to the

seasonal length of each age class of prickly sculpin. In functional response experiments

(Jones 1986), prickly sculpin 120-190mm TL each consumed 2-10 chum salmon fry 35-

44mm FL, and prickly sculpins ≥50mm consumed sockeye salmon fry ≥20mm FL in the

Cedar River (Tabor et al. 1998). I determined that age 2-6 prickly sculpins could prey

upon juvenile sockeye salmon during spring, and age 5-6 could effectively prey upon

juvenile sockeye salmon during the fall (Table 8). Age 4-6 prickly sculpins were able to

prey upon juvenile chinook salmon in the spring (Table 9). Seasonal weights for juvenile

sockeye and geometric mean weights for Neomysis were calculated from seasonal

weights reported in Chigbu and Sibley (1994) and Chigbu et al. (1998).

Initial body weights (g) (Rickard 1980) and P-values generated from runs using

nominal observed thermal experience for age 1-6 prickly sculpins were used to estimate

monthly and final weight for individuals experiencing epilimnetic temperatures and

hypolimnetic temperatures. Prey energy density and percent composition in the diet

remained the same in these simulations. Growth efficiency (g prey consumed/g change

in body weight) and percent body weight consumed daily (g consumed/g body weight)

was examined using prey consumption estimated by the model, and body weight

estimates provided by Rickard (1980).

40

Results

Consumption Rate Simulations

Individual daily ration of prey ranged from 0.04 g/day (age 1) in winter (January-

March) to 0.96 g/day (age 6) in autumn (October-December) (Figure 9). Daily ration

increased with age and increased seasonally from winter through autumn. Similarly,

piscivory increased seasonally from winter through autumn for age 2 and older prickly

sculpin. Age 1 prickly sculpins primarily ate chironomids, Neomysis, and miscellaneous

invertebrates, but not fish (Figure 9). The model estimated annual consumption per 1,000

age 1-6 prickly sculpin based on annual survival rates of 24% and 34%; these simulations

resulted in estimates of 2.88-5.16 kg of fish, 7.14-8.75 kg of Neomysis, 36.37-36.71 kg of

chironomid pupae, 2.90-4.04 kg of Trichoptera, 2.45-3.51 kg of mollusca, and 8.46-11.14

kg of miscellaneous invertebrates being consumed annually per 1,000 prickly sculpin

(Table 10).

Numerical predation losses of juvenile sockeye salmon, chinook salmon, prickly

sculpin, and Neomysis, were simulated per 1,000 prickly sculpin. The estimates of

sockeye salmon losses were based on the assumption that juvenile sockeye salmon

represented 1% of the fish in the diet (most realistic scenario) (Tabor et al. 1998, Rickard

1980). The model estimated 1,000 prickly sculpin (>47mm TL, age 1-6) could

potentially consume 34-17 juvenile sockeye salmon (25-45 mm) in the spring, and 0-1

juvenile sockeye salmon in the summer if they composed 1% of the fish fraction of the

41

diet (Table 11). If juvenile chinook salmon represented 1% of fish in the diet during

periods when they were vulnerable to predation in spring, 1-2 juvenile chinook salmon

would be consumed per 1,000 prickly sculpin annually (Table 11). If other sculpin

composed 99% (most realistic scenario) of the fish in the diet, then 5,764-10,310 sculpin

(0.5g) would be consumed annually per 1,000 prickly sculpin. The model predicted that

4.4-5.3 million Neomysis are consumed per 1,000 prickly sculpin annually (Table 11).

Annual juvenile sockeye salmon consumption estimates for 1,000 age 1 and older

prickly sculpin were expanded to the entire population (94-250 million) of prickly sculpin

in Lake Washington (Rickard 1980). These estimates expanded to 1.60-8.75 million

juvenile sockeye salmon being consumed annually, and accounted for 10.8-59.1% of the

fry population (0.73-27.1 million) (Seiler and Kishimoto 1997; Washington Department

of Fish and Wildlife, unpublished data). Annual juvenile chinook salmon consumption

estimates for a sub-population of 1,000 prickly sculpin were expanded to the entire

population of prickly sculpin in Lake Washington. These estimates expanded to 94-500

thousand juvenile chinook salmon consumed annually, and accounted for 4.9-26.3% of

the population (1.9 million) (Washington Department of Fish and Wildlife unpublished

data). If 99% of the fish consumed by age 1-6 prickly sculpin were other age 1-6 prickly

sculpin, the model predicted that they would consume 88-107% of their own biomass.

42

Simulations of Seasonal, Depth-Specific Growth and Efficiency

Thermal experience provided an explanation for seasonal offshore and onshore

migrations by age 3-6 prickly sculpins. When modeled with their observed thermal

experience (nominal simulations) and a P-value (constant ration) from the nominal

observed thermal experience, age 3-6 prickly sculpins achieved greater weight by the end

of the year than they would have by remaining in the epilimnion or hypolimnion

throughout the year. Age 1-2 prickly sculpin had larger final weights when modeled with

a hypolimnetic thermal experience and a P-value (constant ration) from nominal observed

temperatures (Figure 10). Age 1-2 prickly sculpins lost weight in July and August, and

age 4-5 lost weight in May when modeled with a nominal observed thermal experience.

Growth efficiency, and consumption rate (Figures 11 and 12) were modeled for

each age class using epilimnetic temperatures, hypolimnetic temperatures, and a thermal

experience inferred from seasonal onshore-offshore migrations. Annual growth

efficiency declined markedly with age for ages 1-3, and gradually for ages 3-6 (Figure

11). Prickly sculpin modeled with hypolimnetic temperatures grew most efficiently

(Figure 11). Size dependent growth efficiency ranged from 53% (age 1) to 26% (age 6)

when modeled with hypolimnetic temperatures, and from 15% (age 1) to 5% (age 6)

when modeled with epilimnetic temperatures (Figure 11). When prickly sculpin were

modeled with temperatures inferred from onshore and offshore migrations, growth

efficiency was lowest for age 2 fish. Size dependent estimates of daily consumption

ranged from 12.1% (age 1) to 1.9% (age 6) of body weight per day for fish experiencing

43

epilimnetic temperatures, and from 3.4% to 0.4% for fish experiencing hypolimnetic

temperatures. When prickly sculpin were modeled with temperatures inferred from

onshore and offshore migrations, average daily consumption ranged from 12.1% (age 1)

to 1.4% (age 6). For all ages of prickly sculpin, the least amount of prey biomass was

consumed when they were modeled with hypolimnetic temperatures, and the greatest

amount when they were modeled with epilimnetic temperatures (Figure 12).

Discussion

Although juvenile salmon were not a major prey item for prickly sculpin (less

than 0.03% of the diet overall), predation losses imposed by prickly sculpin could be

potentially large (4.9-26.3% of juvenile chinook population and 10.8-59.1% of juvenile

sockeye population). Juvenile salmon are likely to compose less than 1% of the fish

fraction of prickly sculpin diet in Lake Washington; however, this is speculation, and

their dietary contribution needs to be investigated further. Consumption of juvenile

sockeye salmon by prickly sculpin was not reported by Rickard (1980); however,

significant levels of predation is thought to occur in the lower Cedar River and southern

end of Lake Washington (Tabor et al. 1998). If juvenile salmon are closely associated

with the bottom of the lake, either along the littoral region or along deeper slopes, their

risk of being eaten will be higher, and effective predator avoidance will be important for

survival.

44

Prickly sculpin should feed most heavily on juvenile salmon when they are

present in high densities relative to alternate prey (Jones 1986). As alternative prey were

experimentally added to a feeding area, predation was significantly reduced, suggesting

that salmon fry were either not a preferred prey of prickly sculpin or that alternative prey

effectively inhibited feeding rates on fry. Large aggregations of presmolts could trigger

an accelerated feeding response; however, size dependent relationships will limit prickly

sculpin predation on juvenile salmon.

Fish predation is a major factor controlling Neomysis abundance in aquatic

systems (Chigbu et al. 1998, McDonald et al. 1990, Kjelberg et al. 1991). Chigbu et al.

(1998) examined the spatial and temporal movements of Neomysis mercedis in Lake

Washington, where Neomysis occupy shallower areas during winter and early spring, and

move into deeper waters during summer as water temperature increases. A similar

pattern of migration to deeper waters was noted for older prickly sculpin (Rickard 1980).

Neomysis mercedis represented the largest proportion of prickly sculpin diet by weight

during the winter, and model simulations indicate that between 3.9-4.5 million Neomysis

mercedis were consumed annually per 1,000 prickly sculpin. Neomysis prey upon

Daphnia (Murtaugh 1981), which are a staple prey of sockeye salmon fry. Predation on

Neomysis by prickly sculpin may relieve some predation on Daphina and thus indirectly

improve the food supply for juvenile sockeye salmon

45

Seasonal onshore and offshore movements by prickly sculpin have been linked to

spawning and water temperature (Rickard 1980). Between December and June, the

coolest bottom temperatures in the littoral and profundal zones were 7°C, and

temperatures were warmest in the summer when littoral bottom temperatures increased to

19.5°C. As littoral temperature increased in July, abundance of older prickly sculpin

increased in the profundal zone. Bottom temperature and catch per unit effort (CPUE)

were negatively correlated between December and March. Sand and silt habitats, which

were mostly associated with nearshore regions, were most populated during the winter

months.

An ontogenetic habitat shift for age-0 sculpin (Cottus extensus) from the

profundal zone to the littoral zone in Bear Lake, Utah was attributed to the energetic

benefit from behavioral thermoregulation (Wurtsbaugh and Neverman 1988, Ruzycki and

Wurtsbaugh 1999). The profundal zone in Bear Lake was the less energetically favorable

habitat because of limited food availability and cooler temperatures. As shown through

otolith analysis, juvenile Bear Lake sculpin increased their growth rate in warmer water.

It has been suggested that juvenile Bear Lake sculpin inhabiting the cool profundal zone

migrate into the water column during the night to take advantage of the relatively warmer

metalimnion. Applying the model to Lake Washington prickly sculpin life history data

provided insight into the energetic advantage of habitat selection and migration.

Consumption (% body weight) was lowest, and growth efficiency greatest, for

individuals occupying cooler hypolimnetic waters.

46

Growth simulations modeled with P-values from nominal observed thermal

experience, and epilimnetic thermal experience or hypolimnetic thermal experience

showed that age 3-6 prickly sculpins would lose large amounts of weight if they were to

remained in the epilimnion during the summer and autumn. These results, combined

with onshore and offshore movements by prickly sculpins, suggest that temperature

inluenced distribution. Large fluctuations in the habitat occupied during fall months may

have been due to the redistribution of larger individuals into littoral regions, or may be

related to prickly sculpins switching prey type. The primary prey in summer was

chironomid pupae, and was fish in the fall. Prickly sculpin were the predominant fish

found in prickly sculpin diet, suggesting that cannibalism may be an important self-

regulating mechanism for this population.

Model simulations are only as accurate as the information used to construct them,

and can be useful tools for exploring the potential importance of various trophic

interactions. These simulations place bounds around the potential magnitude of

interactions and can identify important sources of uncertainty. More accurate estimates

of Lake Washington prickly sculpin thermal experience and proportion of juvenile

salmon in diet is needed to minimize extrapolation error in this model and gain better

resolution in consumption estimates.

47

References

Allen, J.R.M. and R.J. Wootton. 1982. Age, growth, and rate of food consumption in an

upland population of the three-spined stickleback (Gasterosteus aculeatus).

Journal of Fish Biology 21:95-106.

Bartell, S.M., J.E. Breck, and R.H. Gardner. 1986. Individual parameter pertubation

and error analysis of fish bioenergetics models. Canadian Journal of Fisheries

and Aquatic Sciences 43:10-168.

Beauchamp, D.A. 1994. Spatial and temporal dynamics of piscivory: Implications for

food web stability and the transparency of Lake Washington. Lake and Reservoir

Management 9:151-154.

Beauchamp, D.A., D.J. Stewart, and G.L. Thomas. 1989. Corroboration of a

bioenergetics model for sockeye salmon. Transactions of the American Fisheries

Society 118:597-607.

Bond, C.E. 1963. Distribution and ecology of freshwater sculpins, genus Cottus, in

Oregon. Ph.D. Dissertation, University of Michigan.

Boisclair, D. and W.C. Leggett. 1989. The importance of activity in bioenergetics

models applied to actively foraging fishes. Canadian Journal of Fisheries and

Aquatic Sciences 46:1859-1867.

Brandt, S.B. and K.J. Hartman. 1993. Innovative approaches with bioenergetics models:

Future applications to fish ecology and management. Transactions of the

American Fisheries Society 122:731-735.

48

Brett, J.R. and T.D.D. Groves. 1979. Physiological energetics. In Fish Physiology,

volume 8 pages 279-352. Edited by W.S. Hoar, D.J. Randall, and J. R. Brett.

Academic press, New York.

Brocksen, R.W, G.E. Davis, and C.E. Warren. 1968. Competition, food consumption,

and production of sculpins and trout in laboratory stream communities. Journal of

Wildlife Management 32(1): 51-75.

Brodeur, R.D., R.C. Frances, and W.G. Pearcy. 1992. Food-consumption of juvenile

coho (Oncorhyncus kisutch) and chinook salmon (Oncorhyncus tshawytscha) on

the continental shelf off Washington and Oregon. Canadian Journal of Fisheries

and Aquatic Sciences 49:1670-1685.

Cech, J.J., Jr., M.J. Massingill, B. Vondrscek, and A.L. Linden. 1985. Respiratory

metabolism of mosquito fish, Gambusia affinis: Effects of temperature, dissolved

oxygen, and sex difference. Environmental Biology of Fishes 13:297-307.

Chigbu, P., T.H. Sibley, and D.A. Beauchamp. 1998. Abundance and distribution of

Neomysis mercedis and a major predator, longfin smelt (Spirinchus thaleichthys)

in Lake Washington. Hydrobiologia 386:167-182.

Chigbu, P. and T.H. Sibley. 1994. Feeding behavior of longfin smelt (Spirinchus

thaleichthys) may affect water quality and salmon production in Lake

Washington. Lake and Reservoir Management 9:145-148.

Comita, G.W. and G.C. Anderson. 1959. The seasonal development of a population of

Diaptomous ashlandi and related phytoplankton cycles in Lake Washington.

Limnology and Oceanography 4:37-52.

49

Cummins, K.W. and J.C. Wuycheck. 1971. Caloric equivalents for investigations in

ecological energetics. Mitteilungen Internationale Vereeinigung Für Theoretische

und Angewandte Limnologie 18.

Edwards, D.G. and J.J. Cech Jr. 1990. Aquatic and aerial metabolism of juvenile

monkeyface prickleback (Cebidichthys violaceus) an intertidal fish of California.

Journal of Comparative Physiology 96:61-65.

Eggers, D.M., N.W. Bartoo, N.A. Rickard, R.E. Nelson, R.C. Wissmar, R.L. Burgner,

and A.H. Devol. 1978. The Lake Washington ecosystem: The perspective from

the fish community production and forage base. Fisheries Research Board of

Canada 35:1553-1571.

Elliott, J.M. and W. Davidson. 1975. Energy equivalents of oxygen consumption in

animal energetics. Oecologia (Berlin) 19:195-201.

Elser, J.J., C. Luecke, M.T. Brett, and C.R. Goldman. 1995. Effects of food web

compensation after manipulation of rainbow trout in an oligotrophic lake.

Ecology 76:52-69.

Gould, H.R. and R.F. Budinger. 1958. Control of sediments and bottom configuration

by convection currents, Lake Washington, Washington. Journal of Marine

Research 17:183-198.

Hanson, P.C., T.B. Johnson, D.E. Schindler, and J.K. Kitchell. 1997. Fish Bioenergetics

3.0. University of Wisconsin-Madison Center for Limnology and University of

Wisconsin Sea Grant Institute.

Ikusemiju, K. 1975. Aspects of the ecology and life history of the sculpin Cottus

50

aleuticus (Gilbert) in Lake Washington. Journal of Fish Biology 7:235-245.

Jones, M.L. 1986. The influence of natural predation on the population dynamics of

pacific salmon. PhD Dissertation, University of British Columbia.

Juanes, F. 1994. What determines prey size selectivity in piscivorous fishes? In Theory

and Application In Fish Feeding Ecology pp. 79-100. Edited by D.J. Stouder,

K.L. Fresh, and R.J. Feller. Belle W. Baruch Library In Marine Science number

18, University of South Carolina. Carolina Press, Columbia, South Carolina.

Kitchell, J.F., 1990. The scope for mortality caused by sea lamprey. Transactions of the

American Fisheries Society 119:642-648.

Kitchell, J.F., D.J. Stewart, and D. Weininger. 1977. Applications of a bioenergetics

model to perch (Perca flavescens) and walleye (Sitzostedion vitreum). Fisheries

Research Board of Canada 34:1922-1935

Kjellberg, G., D.O. Hessen and J.P. Nilssen. 1991. Life history, growth, and production

of Mysis relicta in Lake Michigan: Abundances and relationships with their

potential prey. Canadian Journal of Fisheries and Aquatic Sciences 47:977-983.

Labar, G.W. 1993. Use of bioenergetics models to predict the effect of increased lake

trout predation on rainbow smelt following sea lamprey control. Transactions of

the American Fisheries Society 122:942-950.

Martz, M., F. Goetz, J. Dillon, and T. Shaw. 1996. Lake Washington ecological studies.

Study element II: Early lake life history of sockeye salmon (Onchorynchus nerka)

in Lake Washington. U.S. Army Corps of Engineers. Seattle District 21 p and

App. Septenmber 1996.

51

McDonald, M.E., L.B. Crowder, and S.B. Brandt. 1990. Changes in Mysis and

Pontoporeia populations in southern Lake Michigan: A response to shifts in the

fish community. Limnology and Oceanography 35:220-227.

Murtaugh, P.A. 1981. Selective predation by Neomysis mercedis in Lake Washington.

Limnology and Oceanography 26:445-453.

Negus, M.T. 1995. Bioenergetics modeling as a salmonine management tool applied

to Minnesota waters of Lake Superior. American Journal of Fisheries

Management 15:60-78.

Neverman, D. and W.A. Wurtzbaugh. 1994. The thermoregulatory function of diel

vertical migration for a juvenile fish, Cottus extensus. Oecologia 98:247-25.

Ney, J.J. 1990. Trophic economics in fisheries: Assessment of demand-supply

relationships between predators and prey. Reviews in Aquatic Science 2:55-81.

Ney, J.J. 1993. Bioenergetics modeling today: Growing pains on the cutting edge.

Transactions of the American Fisheries Society 122:736-748.

Persson, L., P. Byström, and E. Wahlström. 2000. Cannibalism and competition in

eurasion perch: Population dynamics of an ontogenetic omnivore. Ecology

81(4):1058-1071.

Post, D.M., J.F. Kitchell, and J.R. Hodgson. 1998. Interactions among adult

demography, spawning date, growth rate, predation, overwinter mortality, and the

recruitment of largemouth bass in a northern lake. Canadian Journal of Fisheries

and Aquatic Sciences 55:2588-2600.

52

Post, J.R. 1990. Metabolic allometry of larval and juvenile yellow perch (Perca

Flavescens) – Insitu estimates and bioenergetic models. Canadian Journal of

Fisheries and Aquatic Sciences 47:554-560.

Rand, P.S. and D.J. Stewart. 1998. Prey fish exploitation, salmonine production, and

pelagic food web efficiency in Lake Ontario. Canadian Journal of Fisheries and

Aquatic Sciences 55:318-327.

Rand, P.S. and D.J. Stewart. 1998. Dynamics of salmonine diets and foraging in

Lake Ontario, 1983-1993: A test of a bioenergetics model prediction. Canadian

Journal of Fisheries and Aquatic Sciences 55:307-317.

Rice, J.A., and P.A. Cochran. 1984. Independent evaluation of a bioenergetics model for

largemouth bass. Ecology 63:732-739.

Rickard, N.A. 1980. Life history and population characteristics of the prickly sculpin

(Cottus asper Richardson) in Lake Washington. Master of Science thesis,

University of Washington.

Rottiers, D.V. and R.M. Tucker. 1982. Proximate composition and caloric content of

eight Lake Michigan fishes. U.S. fish and Wildlife Service Technical Papers

108:1-8.

Ruzycki, J.R. and W.A. Wurtsbaugh. 1999. Ontogenetic habitat shifts of Bear Lake

sculpin. Transactions of the American Fisheries Soceity 128:1201-1212.

Schindler, D.E., J.F. Kitchell, X. He, S.R. Carpenter, J.R. Hodgeson and K.L.

53

Cottingham. 1993. Food web structure and phosphorus cycling in lakes.

Transactions of the American Fisheries Society. 122:756-772

Seiler, D. and L. Kishimoto. 1997. 1997 Cedar River sockeye salmon fry production

evaluation. Annual report to King County Water and Land Resources Division.

Stewart, D.J., J.F. Kitchell, and L.B. Crowder. 1981. Fishes and their salmonid

predators in Lake Michigan. American Fisheries Society 110:751-763.

Stewart, D.J. and M. Ibarra. 1991. Predation and production by salmonine fishes in

Lake Michigan, 1978-88. Canadian Journal of Fisheries and Aquatic Sciences

48:909-922.

Swartzman, G.L. and D.A. Beauchamp. 1990. Simulation of the effect of rainbow

trout introduction in Lake Washington. Transactions of the American Fisheries

Society 119:122-134.

Tabor, R., J. Chan, and S. Hager. 1998. Predation of sockeye salmon fry by Cottids and

other predatory fishes in the Cedar River and Southern Lake Washington. U.S.

Fish and Wildlife Service, Western Washington Office, Aquatic Resources

Division.

Tabor, R.J. and R.M. Piaskowski. 2001. Nearshore habitat use by juvenile chinook

salmon in lentic systems of the Lake Washington Basin. Annual report 2001.

U.S. Fish and Wildlife Service.

Thornton, K.W. and A.S. Lessem. 1978. A temperature algorithm for modifying

biological rates. Transactions of the American Fisheries Society 107:284-287.

54

Wahl, D.H. and R.A. Stein. 1988. Selective predation by three escocids: The role of

prey behavior and morphology. Transactions of the American Fisheries Society

117:142-151.

Woodey, J.C. 1972. Distribution, feeding, and growth of juvenile sockeye salmon in

Lake Washington. Ph.D. Dissertation, University of Washington.

Wurtsbaugh, W.A. and D. Neverman. 1998. Post-feeding thermotaxis and daily vertical

migration in a larval fish. Nature (London) 333:846-848.

Wydoski, R.S. and R.R. Whitney. 1979. Inland fishes of Washington. University of

Washington Press. Seattle, WA.

55

Figures

Season

Win Spr Sum Aut Win Spr Sum Aut Win Spr Sum Aut Win Spr Sum Aut Win Spr Sum Aut Win Spr Sum Aut

Indi

vidu

al D

aily

Rat

ion

(g/d

ay)

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0Fish Neomysis Chironomid Misc. Invertebrates

Age 5

Age 6

Age 4

Age 3

Age 2

Age 1

Figure 9. Individual daily ration for Lake Washington prickly sculpin partitioned by major prey groups. Ration is reported by age and season.

56

Month

Jan Feb Mar AprMayJun Jul AugSep Oct NovDec15

18

21

24

27

Wei

ght (

g)

6

9

12

15

18

0

2

4

6

8

Month

Jan Feb Mar Apr MayJun Jul AugSep Oct NovDec50

55

60

65

70

75

80

85

Epilimnetic TemperatureNominal TemperatureHypolimnetic Temperature

36

40

44

48

52

56

60

64

24

27

30

33

36

39

42

45

Age 4

Age 6Age 3

Age 5

Age 1

Age 2

Figure 10. Growth trajectories for age 1-6 prickly sculpin experiencing epilimnetic temperatures, nominal observed temperatures, and hypolimnetic temperatures. Hypolimnetic and epilimnetic growth trajectories are modeled with fixed P-values from nominal observed simulations.

57

Age

0 1 2 3 4 5 6 7

Gro

wth

effi

cenc

y

0.0

0.1

0.2

0.3

0.4

0.5

0.6

Epilimnetic TemperatureNominal TemperaturesHypolimnetic Temperatures

Figure 11. Growth efficiencies (growth/consumption) for prickly sculpin modeled with epilimnetic temperatures, hypolimnetic temperatures, and nominal observed temperatures.

58

Age

0 1 2 3 4 5 6 7

Annu

al C

onsu

mpt

ion

(g)

0

50

100

150

200

250

300

350

400Epilimnetic TemperaturesNominal TemperaturesHypolimnetic Temperatures

Figure 12. Annual consumption (g) for prickly sculpin modeled with nominal observed P-values and epilimnetic temperatures, hypolimnetic temperatures, and nominal observed temperatures.

59

Tables Table 6. Length/weight and age class abundance for 1,000 age 1-6 prickly sculpin in (Rickard 1980). Model 1 had an annual survival rate of 34% (Z = 1.0788) and model 2 had an annual survival rate of 24% (Z = 1.4271).

Sculpin abundance

Age Length (mm) Weight (g) Model 1 Model 2 1 47.4 1.0 661 7602 86.6 7.2 225 1823 110.0 15.7 76 434 128.4 26.1 26 115 146.9 40.7 9 36 165.2 59.7 3 1

Total 1000 1000 Table 7. Average monthly epilimnetic and hypolimnetic temperatures (C°) for Lake Washington. Age 1-2 prickly sculpin experience epilimnetic temperatures throughout the year, and age 3-6 prickly sculpin experience epilimnetic and hypolimnetic temperatures.

Average Temperature °C

Month

Epilimnion Age 3-6 Thermal

Experience

Hypolimnion January 8.3 = 8.3 7.8 February 7.0 = 7.0 6.8 March 7.0 = 7.0 6.8 April 8.3 = 8.3 6.8 May 11.0 = 11.0 7.0 June 13.8 = 13.8 7.3 July 17.0 7.5 = 7.5

August 21.5 7.5 = 7.5 September 20.5 7.5 = 7.5

October 17.5 7.5 = 7.5 November 13.0 8.0 = 8.0 December 9.0 9.0 = 8.8

60

Table 8. Seasonal size relationships between juvenile sockeye salmon and each age class of prickly sculpin. Predation capability was determined by 50% prey/predator length ratios based on seasonal lengths of juvenile sockeye salmon (Woody 1972, Chigbu and Sibley 1994) and prickly sculpin (1980). Asterisks denote seasons and ages when prickly sculpin were capable of preying upon juvenile sockeye salmon.

Total length Fork length

Prickly sculpin age class juvenile Season 1 2 3 4 5 6 sockeye Spring 60.2 *96.2 *115.6 *133.4 *151.6 *168.8 35.2 Summer 58.6 83.4 95.4 108.8 *123.0 *135.8 74.7 Autumn 75.4 102.6 124.8 142.2 160.0 176.0 93.6 Winter 84.6 107.4 125.6 143.8 161.8 177.4 107.0

Table 9. Seasonal size relationships between juvenile sockeye salmon and each age class of prickly sculpin. Predation capability was determined by 50% prey/predator length ratios based on seasonal lengths of juvenile Chinook salmon (Koehler unpublished data) and prickly sculpin (Rickard 1980). Asterisks denote seasons and ages when prickly sculpin were capable of preying upon juvenile chinook salmon.

Total length Fork length

Prickly sculpin age class Juvenile Season 1 2 3 4 5 6 chinook Spring 60.2 96.2 115.6 133.4 *151.6 *168.8 54.3 Summer 58.6 83.4 95.4 108.8 123.0 135.8 101.9

61

Table 10. Annual consumption of Neomysis mercedis, Trichoptera, Chironomidae pupae, and fish by prickly sculpin. Consumption estimates are based on age-specific abundance per 1000 fish (g consumed/1000 prickly sculpin).

Annual Consumption (g) estimated by model 1 (34% survival) Age Fish Neomysis Chironomid Trichoptera Mollusca Misc. Inverts

1 0 2,644 23,136 0 0 2,644 2 2,023 3,596 9,215 2,697 2,247 4,495 3 688 1,375 3,439 1,070 841 1,681 4 1,455 676 546 156 260 1,377 5 689 318 256 80 115 654 6 300 138 114 36 51 285

Total 5,155 8,747 36,706 4,039 3,514 11,136

Annual Consumption (g) estimated by model 2 (24% survival) Age Fish Neomysis Chironomid Trichoptera Mollusca Misc. Inverts

1 0 3,041 26,605 0 0 3,041 2 1,642 2,919 7,480 2,189 1,824 3,649 3 394 788 1,970 613 482 963 4 588 273 221 63 105 557 5 197 91 73 23 33 187 6 61 28 23 7 10 58

Total 2,882 7,140 36,372 2,895 2,454 8,455

62

Table 11. Number of prey consumed annually by prickly sculpin using two selectivity models. Consumption estimates are based on age class abundance per 1,000 fish (Table 6). Seasonal juvenile sockeye salmon and juvenile chinook salmon fork lengths were compared to seasonal sculpin total lengths to determine which sculpin age classes were able to prey upon them using a 50% prey/predator length ratio (Table 8 and 9). Consumption of other prickly sculpins is modeled using a 0.5g prickly sculpin as prey. Predation rates on juvenile sockeye salmon and juvenile chinook salmon are reported assuming that they represent 1% of the fish fraction of diet reported in Rickard (1980), and prickly sculpin cannibalism reported assuming that they consume 99% of the fish fraction of diet reported by Rickard (1980).

Annual survival rate of 34% (Selectivity Model 1) Age Class

Sockeye (1%)

Spring

Sockeye (1%)

Summer

Chinook (1%)

Spring

P. Sculpin (99%)

All Seasons

Neomysis

All Seasons 1 0 0 NA 0 1,612,247

2–3 14 0 NA 5,422 3,031,346 4-6 20 1 2 4,888 690,037

Total 34 1 2 10,310 5,333,630

Annual survival rate of 24% (Selectivity Model 2)

Age Class

Sockeye (1%)

Spring

Sockeye (1%)

Summer

Chinook (1%)

Spring

P. Sculpin (99%)

All Seasons

Neomysis

All Seasons 1 0 0 NA 0 1,854,031

2–3 10 0 NA 4,072 2,260,412 4-6 7 0 1 1,692 238,932

Total 17 0 1 5,764 4,353,375