1 North Pacific Research Board Project Final Report

1 NORTH PACIFIC RESEARCH BOARD PROJECT FINAL REPORT 2 3 4 5 Seasonal Bioenergetics of Pollock, Cod and Arrowtooth Flounder in the Bering Sea 6 7 8 NPRB Project B54 Final Report 9 10 11 Ron Heintz 12 13 14 Recruitment Energetics and Coastal Assessment Program, Alaska Fisheries Science 15 Center, National Marine Fisheries Service, NOAA, Auke Bay Laboratories, Juneau, AK, 16 99801, USA 17 September 2009

2 Abstract: 3 4 The NPRB funded Bering Sea Integrated Ecosystem Research Program (BSIERP) and NSF funded 5 Bering Ecosystem Study (BEST) provided an unparalleled opportunity to understand how the 6 Bering Sea ecosystem supports the production of commercially valuable species. The seasonal 7 bioenergentics component evaluated the condition of juvenile fish at different ontogenetic 8 stages in an effort to understand the effect of environmental variability on the health of fish 9 populations. The intent was to increase our understanding of the mechanisms underlying 10 recruitment variability in the southeastern Bering Sea so that fishery managers could gain the 11 ability to understand how climate and fishing influence the variability in harvestable biomass. 12 Specifically, the dry mass, lipid, energy density and growth of juvenile gadids and arrowtooth 13 flounder were measured on specimens collected on a series of surveys conducted between May 14 of 2008 and October of 2010. Over 3500 separate analyses were conducted on larvae, young-of- 15 the-year, age-1 juveniles and their zooplankton prey. The results of these analyses were 16 integrated into other findings from the BEST/BSIERP Program to evaluate the health of fish 17 populations under different environmental conditions during different developmental stages. 18 The principle findings of the work are: 19 1. Walleye pollock have a relatively short time window for provisioning themselves with lipid 20 prior to winter after they complete larval development. 21 2. Climatic conditions mediate their ability to provision themselves by influencing the spatial 22 match between young-of-the-year pollock and their prey and the quality of those prey. 23 3. Fish that surpass a critical energy content during this period have the highest probability of 24 surviving winter. This critical level depends on acquiring both a large size and high energy 25 density. 26 4. A mechanistic connection between climate and recruitment has been identified that 27 indicates that under so-called cool conditions in the Bering Sea, high-lipid prey are available 28 to juvenile pollock in fall in sufficient abundance to support rapid growth and energy 29 storage. When these conditions do not exist, recruitment can be expected to be below 30 average. 31 5. Similar results have been found for Pacific cod 32

2 Key Words: 3 pollock, Pacific cod, arrowthooth flounder, juvenile, larvae, energy, lipid, condition, recruitment 4

5 Citation : 6 Heintz, R. (2014) Seasonal Bioenergetics of Pollock, Cod and Arrowtooth Flounder in the 7 Bering Sea. NPRB Final report for project B54. 8

2 Table of Contents

3 Contents

4 Abstract: ...... 2

5 Key Words: ...... 3

6 Citation : ...... 3

7 Table of Contents ...... 4

8 Study Chronology: ...... 9

9 Introduction ...... 9

10 Overall Objectives: ...... 10 11 Chapter 1 - The Effect of Temperature and Nutrition on RNA/DNA Ratio and Growth of 12 Pacific Cod (Gadus macrocephalus) and Walleye Pollock (Theragra chalcogramma) 13 Larvae ...... 13 14 1.1 Abstract: ...... 13 15 1.2 Introduction ...... 14 16 1.3 Materials and Methods ...... 17 17 1.3.1 Pacific Cod Temperature Experiment ...... 18 18 1.3.2 Pacific Cod Starvation Experiment ...... 19 19 1.3.3 Walleye Pollock Temperature Experiment ...... 20 20 1.3.4 Biochemical Analyses ...... 20 21 1.3.5 Statistical Analyses: ...... 21 22 1.3.5.1 Pacific Cod Temperature Experiment ...... 22 23 1.3.5.2 Pacific Cod Starvation Experiment ...... 22 24 1.3.5.3 Pacific Cod Growth Model ...... 23 25 1.3.5.4 Walleye Pollock Temperature Experiment ...... 24 26 1.4 Results ...... 24 27 1.4.1 Pacific Cod Temperature Experiment ...... 24 28 1.4.2 Pacific Cod Starvation Experiment ...... 25 29 1.4.3 Pacific Cod Growth Model Synthesis ...... 26 30 1.4.4 Walleye Pollock Temperature Experiment ...... 26 31 1.5 Discussion ...... 26

1 1.6 Conclusion ...... 32 2 1.7 Acknowledgements ...... 33 3 1.8 Literature Cited ...... 33

4 1.9 Figures and Tables ...... 46 5 Figure. 1.4a,b ...... 49 6 Chapter 2 - Conceptual model of energy allocation in walleye pollock (Theragra 7 chalcogramma) from larvae to age-1 in the southeastern Bering Sea ...... 56 8 2.1 Abstract ...... 56 9 2.2 Keywords: ...... 57 10 2.3 Introduction ...... 57 11 2.3.1 Study region ...... 60 12 2.4 Materials and methods ...... 61 13 2.4.1. Biological sampling ...... 61 14 2.4.2. Chemical Analysis ...... 63 15 2.4.3. Statistical analysis ...... 66 16 2.5. Results ...... 69 17 2.5.1. Biological sampling ...... 69 18 2.5.2. Cohort-specific patterns from age-0 to age- ...... 69 19 2.5.3. Seasonal patterns in energy allocation of age-0 fish ...... 70 20 2.5.4. Seasonal patterns in length of age-0 fish ...... 71 21 2.6 Discussion ...... 71 22 2.7 Acknowledgements ...... 75 23 2.8 References ...... 76 24 2.9 Table legends ...... 84 25 2.10 Figure legends ...... 85 26 Chapter 3 - Correlation between recruitment and fall condition of age-0 pollock 27 (Theragra chalcogramma) from the eastern Bering Sea under varying climate 28 conditions ...... 100 29 3.1 Abstract ...... 100 30 3.2 Keywords: ...... 101 31 3.3 Introduction ...... 102 32 3.4 Materials and methods ...... 105 33 3.4.1 Pollock collection and processing ...... 105 34 3.4.2 Collection and processing of zooplankton ...... 107 35 3.4.3 Data analysis ...... 108

1 3.5 Results ...... 111 2 3.5.1 Climate effects on prey quality...... 111 3 3.5.2 Climate effects on pollock diets ...... 111 4 3.5.3 Climate effects on pre-winter condition ...... 113 5 3..5.4 Influence of pre-winter condition on survival ...... 113 6 3.6. Discussion ...... 114 7 3.6.1 Climate impacts on food abundance and quality ...... 114 8 3.6.2 Importance of winter ...... 115 9 3.7 Conclusion ...... 117 10 3.8 Acknowledgements ...... 117 11 3.9 References ...... 118 12 3.10 Figures ...... 122 13 Chapter 4 - Spatial match-mismatch between juvenile fish and prey provides a 14 mechanism for recruitment variability across contrasting climate conditions in the 15 eastern Bering Sea...... 130 16 4.1 Keywords: ...... 130 17 4.2 Abstract ...... 131 18 4.3 Introduction ...... 132 19 4.4 Materials and methods ...... 134 20 4.4.1 Ethics statement ...... 134 21 4.4.2 Modeling approaches ...... 134 22 4.4.3 Field observations ...... 135 23 4.4.4 Bioenergetics model ...... 139 24 4.5 Results ...... 145 25 4.5.1 Field observations ...... 145 26 4.5.2 Bioenergetics model ...... 148 27 4.5.3 Mechanistic individual-based model ...... 150 28 4.5.4 Spatial comparison of bioenergetics- and IBM-predicted growth ...... 152 29 4.6 Discussion ...... 152 30 4.8 References ...... 158 31 4.9 Figure legends ...... 163 32 4.10 Supporting Information: ...... 178 33 4.10.1 Comparison of observed and predicted prey preferences ...... 178 34 4.10.2 References ...... 180 35 References ...... 192

1 Chapter 5 - Size, diet, and condition of age-0 Pacific cod (Gadus macrocephalus) during 2 warm and cool climate states in the eastern Bering Sea ...... 198 3 5.1 Abstract ...... 198 4 5.2 Introduction ...... 200 5 5.3 Methods ...... 202 6 5.3.1 Field sampling ...... 202 7 5.3.2 Size ...... 204 8 5.3.3. Fish diet...... 205 9 5.3.4. Energy content ...... 207 10 5.4 Results ...... 208 11 5.4.1. Size ...... 208 12 5.4.2. Diet ...... 208 13 5.4.3. Energy content ...... 210 14 5.5. Discussion ...... 210 15 5.6 Conclusion ...... 215 16 5.7 Acknowledgments ...... 215 17 5.8 References ...... 216 18 Chapter 6 - Ecology and taxonomy of the early life stages of arrowtooth flounder 19 (Atheresthes 2 stomias) and Kamchatka flounder (A. evermanni) in the eastern Bering 20 Sea ...... 234 21 6.1 Abstract ...... 234 22 6.2 Introduction ...... 235 23 6.3 Materials and Methods ...... 236 24 6.3.1 Specimen collection ...... 236 25 6.3.2 Genetic identification ...... 236 26 6.3.3 Visual identification ...... 237 27 6.3.4 Distribution ...... 238 28 6.3.5 Measures of Condition ...... 238 29 6.3.6 Biological sampling ...... 238 30 6.3.7 Dry mass ...... 238 31 6.3.8 Estimation of lipid content (%) ...... 238 32 6.3.9 Energy density ...... 239 33 6.3.10 Statistical analysis ...... 239 34 6.4 Results ...... 239 35 6.4.1 Genetics ...... 239 36 6.4.2 Visual identification ...... 240

1 6.4.3 Distribution ...... 241 2 6.4.4 Biological sampling ...... 241 3 6.4.5 Dry mass ...... 242 4 6.4.6 Lipid content ...... 242 5 6.4.7 Energy density ...... 242 6 6.5 Discussion ...... 242 7 6.5.1 Genetics ...... 243 8 6.5.2 Morphology ...... 243 9 6.5.3 Distribution ...... 244 10 6.5.4 Condition ...... 244 11 6.6 Conclusion ...... 245 12 6.7 Acknowledgements ...... 245 13 6.8 Literature Cited ...... 246 14 6.9 Figure and table legends ...... 250

15 Conclusions: ...... 257

16 BSIERP and Bering Sea Project connections: ...... 257

17 Management or policy implications: ...... 258

18 Publications: ...... 258

19 Poster and oral presentations at scientific conferences or seminars ...... 259

20 Outreach ...... 260

21 Acknowledgements ...... 260

22 Literature cited:...... 261 23

2 Study Chronology: 3 4 The Seasonal Bioenergetics Project depended on samples collected during NOAA, BEST 5 and BSIERP surveys beginning in May of 2008 through October of 2010. Samples of 6 larvae, young-of-the-year and age-1+ arrowtooth flounder, pollock, Pacific cod, 7 Kamchatka flounder and various zooplankton species were collected and processed to 8 determine their RNA/DNA energy or lipid content. Sample collection began with the first 9 survey in spring 2008 conducted as part of the North Pacific Climate Regime and 10 Ecosystem Productivity (NPCREP) Program at NOAA’s Alaska Fishery Science Center. It 11 continued with the BEST/BSIERP surveys conducted in 2008, 2009 and 2010, the 2010 12 NPECREP survey and the Bering Aleutian Salmon International Surveys (BASIS) 13 conducted in 2008, 2009 and 2010. Sample processing began shortly after the first 14 survey and continued until June of 2011. Updates were provided to NPRB on a semi- 15 annual basis beginning on September 28, 2008 through April 19, 2013. This is the final 16 report.

18 Introduction 19 20 The work conducted under the Seasonal Bioenergetics Project (B54) was aimed at 21 understanding how juvenile pollock, Pacific cod, and arrowtooth flounder use energy 22 over their first year of life and how that use relates to climatic conditions in the 23 southeastern Bering Sea (SEBS). Juvenile fish at high latitudes need to allocate energy 24 between the conflicting demands of growth and storage during their first year of life. 25 Understanding how they resolve those demands, by measuring their lipid and energy 26 content, provides an index into the relative importance of growth in the near term and 27 starvation in the long term as factors determining their survival. At the start of this 28 project we had relatively little information of this sort for these species, particularly in 29 the SEBS.

31 The basis for this work was a hypothesis (1b), developed as part of the overall 32 framework for the Bering Sea Integrated Ecosystem Research Program. The hypothesis

1 proposed that decreased storm frequency and intensity in the southeastern Bering Sea 2 would produce a series of oceanographic effects leading to a decreased food supply for 3 juvenile fish, which would consequently reduce their nutritional condition and 4 productivity. The measure of nutritional condition invoked was the energy density or 5 energy content per unit mass. Conditions thought to reduce nutritional condition 6 included a shallowing of the stratified layer leading to nutrient depletion throughout the 7 water column and a resulting attenuation of the fall plankton bloom. Limited primary 8 production in fall would constrain the production of prey available for juvenile fish. 9 Consequently, juveniles would be less able to allocate energy to storage lipids in fall, 10 manifesting reduced energy levels in their tissues.

12 Juvenile fish need to meet a variety conflicting life history demands as they develop 13 from newly hatched larvae to newly recruited adults. Early larvae need to develop 14 muscle mass and a digestive apparatus in order to move volitionally through the water 15 and process food into energy and tissue. Late stage larvae are more mobile but still 16 need to complete organ development while increasing in mass and length. Growth for 17 late stage larvae is important because it improves their locomotion, reduces their 18 availability to predators and increases the diversity of their prey field. Post- 19 metamorphic juveniles, commonly referred to as young-of-the-year (YOY) are able to 20 move at will in the water but are still undergoing developmental processes. In addition, 21 they need to start allocating energy to storage to forestall starvation in winter when 22 food supplies will be reduced. Thus we hypothesized that there is a critical period in 23 which juvenile fish must shift from an energy allocation strategy favoring growth to one 24 favoring energy storage and that this period would commence in late summer and early 25 fall after fish have metamorphosed. We also hypothesized that annual changes in 26 energy density during this critical would result from the conditions specified in 27 hypothesis 1b.

29 Overall Objectives: 30

31 The goal of this project was to determine how constraints on energy supplies influence 32 the way in which Pacific cod, pollock and arrowtooth flounder allocate energy to protein 33 (structure) and lipid (storage) and how these allocations relate to each species’ life

1 history. These data were collected to examine hypotheses relating to hypothesis 1b, 2 which relates climatic conditions to the nutritional status of juvenile pollock, Pacific cod 3 and arrowtooth flounder. In order to meet this goal we proposed the following 4 objectives:

5 Objective 1: Characterize seasonal changes in the energy and lipid content of juvenile 6 pollock

7 For this objective pollock were collected on multiple surveys and analyzed to determine 8 their lipid and energy content. Samples included pre-flexion and post-flexion larvae 9 collected on NPCREP (2008, 2010) and BEST surveys (2008, 2009 and 2010). Age-0 and 10 Age-1+ juveniles were collected on BASIS surveys in 2008-2010 and AFSC beam trawl 11 surveys (2008 and 2010). Additional age-1+ juveniles were collected on MACE acoustic 12 surveys (2008, 2009 and 2010). Fish were returned to the laboratory, processed to 13 determine lengths, wet weights, RNA/DNA, moisture contents, energy density and lipid 14 content. Data collected from these analyses and those from previous years were used to 15 write reports describing seasonal changes in the energy allocation strategy of walleye 16 pollock (Siddon et al. 2013), annual variation in the energy content of walleye pollock 17 (Heintz et al. 2013), and growth potential of pollock in warm and cool conditions (Siddon 18 et al. 2014). In addition, data collected from the samples were contributed to a re- 19 evaluation of the Oscillating control hypothesis (Hunt et al. 2011). Finally the findings 20 and conclusions of this project were included in a review of the first year of life for 21 walleye pollock (Duffy Anderson et al. Submitted) and a synthesis of pointing to the 22 importance of winter in the Bering Sea (Sigler et al. Submitted). These data are reported 23 in Chapters 2, 3 and 4.

24 Objective 2: Characterize seasonal changes in the energy and lipid content of juvenile 25 Pacific cod

26 For this objective Pacific cod were collected on multiple surveys and analyzed to 27 determine their lipid and energy content. Samples included pre-flexion and post-flexion 28 larvae, and age 0+ juveniles collected on the same surveys as previously listed for 29 walleye pollock. Much fewer fish were collected, but they were also analyzed to 30 determine their lengths, wet weights, RNA/DNA, moisture contents, energy density and 31 lipid content. Data collected from these analyses were used to write a report describing 32 interannual changes in the size, condition and diet of Pacific cod (Farley et al. Submitted 33 ). This report focuses on age-0 juveniles because too few larvae were collected for a 34 meaningful analysis. These data are reported in Chapter 5.

1 Objective 3: Characterize seasonal changes in the energy and lipid content of juvenile 2 arrowtooth flounder.

3 Fish from the genus Atherestes were collected on the same surveys described for Pacific 4 cod and pollock and returned to the laboratory. Samples included pre-flexion and post- 5 flexion larvae, and age 0+ juveniles. Relatively few fish were collected and they were 6 believed to represent either arrowtooth flounder (ATF) or Kamchatka flounder (KF). The 7 fish were genetically identified to species through collaboration with the 8 ichthyoplankton project. They were also analyzed to determine their lengths, wet 9 weights, RNA/DNA, moisture contents, energy density and lipid content. Data collected 10 from these analyses were used to write a report describing the distribution and seasonal 11 energy allocation of ATF and KF (DeForest et al. Submitted ). These data are reported in 12 Chapter 6.

13 Objective 4: Develop an understanding of how temperature affects the development of 14 juvenile gadids and timing of metamorphosis.

15 Funds awarded for this objective were used to supplement existing studies of Pacific cod 16 (NPRB Project R0605) and ultimately published in a series of papers. In those studies the 17 effects of temperature on growth of Pacific cod were examined in different stages of 18 development and under food limited conditions. As part of these growth studies we 19 examined and related the observed values of growth to observations of nucleotides in 20 pollock and cod tissues. This work formed a chapter in a PhD thesis (Sreenivasan 2011). 21 That work demonstrated that the cost of growth in gadid larvae is reduced at higher 22 temperatures due to increased efficiency of translating DNA into protein. At low 23 temperatures gadids make up for the decreased efficiency of DNA translation by 24 increasing translational capacity, hence RNA/DNA increases at low temperatures. RNA 25 synthesis is metabolically expensive; consequently growth is reduced at low 26 temperatures. Finally, withholding food as one of the feeding treatments in the study 27 provided a lower limit to RNA/DNA that could be used to identify starving fish in the 28 field. These data are reported in Chapter 1.

2 Chapter 1 - The Effect of Temperature and Nutrition on 3 RNA/DNA Ratio and Growth of Pacific Cod (Gadus 4 macrocephalus) and Walleye Pollock (Theragra chalcogramma)

5 Larvae

6 1.1 Abstract: 7 Pacific cod (Gadus macrocephalus) and walleye pollock (Theragra chalcogramma) are

8 among the most economically and ecologically important groundfish species in the

9 North Pacific Ocean. In spite of their importance, little is known about many aspects of

10 their physiology, specifically about larval growth strategies in these fish. Since larval fish

11 growth may determine future growth patterns, which could affect recruitment success,

12 assessments of larval growth strategies might improve predictive growth models. In this

13 study, nucleic acids (RNA and DNA) were used to index early growth in yolk-sac Pacific

14 cod and walleye pollock larvae cultured at two temperatures (5°C and 8°C) and in yolk-

15 sac Pacific cod cultured in two nutritional states (fed and starved). Growth

16 corresponded to changes in RNA/DNA. Growth responses in Pacific cod and walleye

17 pollock larvae were affected by small differences in temperature. Exposure to the lower

18 temperature resulted in higher RNA/DNA in both Pacific cod and walleye pollock larvae.

19 Based on nucleic acid patterns during larval development, it was possible to identify

20 distinct growth stanzas in Pacific cod larvae.

3 1.2 Introduction 4 Pacific cod (Gadus macrocephalus) and walleye pollock (Theragra

5 chalcogramma) are among the most numerous (Mecklenburg et al. 2002) and

6 commercially important groundfish species in the North Pacific Ocean (Jewett 1978,

7 Grant et al. 1987, Stepanenko 1995, Beamish et al. 2004, Bacheler et al. 2010, Hiatt et

8 al. 2010). Distributed from southern California north to the Bering Sea, the Aleutian

9 Islands, the Gulf of Anadyr and the Kurile Islands, the Okhotsk Sea, the Yellow Sea, and

10 the Sea of Japan (Ketchen 1961, Bakkala 1984, Bakkala et al. 1984), Pacific cod are an

11 important upper-trophic level species in subarctic ecosystems (Sakurai and Hattori

12 1996). Walleye pollock also are broadly distributed from the Bering Sea and the Gulf of

13 Alaska to Japan (Bacheler et al. 2010).

14 Assessments of early larval growth and development are, however, limited for

15 both species. Estimating larval growth can be important since growth during this stage

16 could determine future growth (Jonsson et al. 2005, Koedijk et al. 2010 a, b). Fish larvae

17 are especially sensitive to minor changes in ecological factors, including temperature

18 and nutrition (Takatsu et al. 1995), which affect growth. Most growth and recruitment

19 assessments in nature have been feasible only after fish have reached the juvenile or

20 adult stages. Applying most growth indices to individual larvae is not feasible because

21 they lack adequate tissue mass. These limitations are intensified by the fragility of

1 larvae, which makes capture of viable samples difficult, and by associated problems of

2 sample storage and tissue shrinkage.

3 The early development, behavior, and physiology of larval Pacific cod and

4 walleye pollock have been examined in a few studies (Alderdice and Forrester 1971,

5 Sogard and Olla 1996, Theilacker et al. 1996, Laurel et al. 2008, Hurst et al. 2010, Laurel

6 et al. 2010, Laurel et al. 2011). These studies have highlighted the complexities of larval

7 behavior and physiology, especially in the early post-hatch period. Diel migratory

8 behavior in larval Pacific cod can be determined by a combination of factors, including

9 development stage, temperature, and photoperiod (Hurst et al. 2009). Growth in turn

10 can be affected by interacting factors, including prey availability, temperature, and

11 development stage (Laurel et al. 2011).

12 In light of the studies carried out so far, a physiological index of larval growth

13 coupled with energetic assessments in predictive growth models can improve our

14 understanding of natural variation of larval survival and of recruitment. This laboratory

15 study investigated whole-body nucleic acid concentrations as an index of growth in yolk-

16 sac Pacific cod and walleye pollock larvae, assessing the responses of growth and nucleic

17 acids to different temperatures and nutritional states.

18 Larval growth of fish is affected by physiological and morphological changes

19 associated with temperature and nutrition (Blaxter 1992, Pavlov et al. 1994, De March

20 1995, Kamler 2002, Gabillard et al. 2003a, b, Amberg et al. 2008, Kamler 2008, Li and

1 Leatherland 2008, Janhunen et al. 2010, Teletchea and Fontaine 2010). Protein

2 synthesis underlying larval growth is dependent on RNA concentration (translational

3 capacity) and RNA-specific protein synthesis rate (translational efficiency; McMillan and

4 Houlihan 1988, Fraser and Rogers 2007), and the latter is affected by temperature. A

5 change in the relationship between translational efficiency/capacity and growth can

6 affect energy budgets, an important consequence for larvae since they lack energetic

7 reserves (Theilacker et al. 1996).

8 Even though assessing larval growth can be problematic, nucleic acid (R/D) ratios

9 can be used as a sensitive growth index in fish larvae (Buckley 1979, 1984, Robinson and

10 Ware 1988, McLaughlin et al. 1995, Buckley et al. 1999, Weber et al. 2003, Caldarone

11 2005, Stierhoff et al. 2009). While DNA concentrations remain stable (Buckley 1984,

12 McLaughlin et al. 1995), cellular RNA concentrations increase or decrease along with

13 protein synthesis, indicating growth and nutritional condition (Weber et al. 2003).

14 However, temperature affects RNA translational efficiency (Lewis and Driedzic 2007)

15 changing the ratio-growth relationship. This prevents direct R/D comparisons between

16 fish observed at temperature differences >2°C (Buckley et al. 1999), and requires

17 calibration of nucleic acid concentrations with growth rates over a temperature range,

18 that is R/D-Growth-Temperature calibrated models (Buckley 1984). Nucleic acid ratios

19 could also be developed as indicators of catabolic stages in larvae approaching terminal

20 starvation, i.e. a “baseline” ratio.

1 We used R/D ratios to observe the effects of temperature on growth in Pacific

2 cod and walleye pollock larvae, and to measure the effects of different nutritional states

3 (fed and starved) on growth in Pacific cod larvae. The aim of measuring R/D ratios was

4 to estimate larval energetic and growth parameters and incorporate those parameters

5 in a predictive growth model. Our objectives were to a) examine nucleic acid and early

6 growth responses in Pacific cod and walleye pollock larvae at different temperatures, b)

7 examine nucleic acid and growth responses in Pacific cod larvae at different nutritional

8 states, and c) build a predictive larval Pacific cod growth model based on nucleic acids

9 measured over a range of temperatures and different nutritional states.

10 1.3 Materials and Methods 11 This study included two temperature treatment experiments and a starvation

12 experiment. In the temperature treatment experiments, Pacific cod and walleye pollock

13 larvae were fed and cultured at two temperatures; in the starvation experiment, groups

14 of Pacific cod larvae were either fed or starved at a single temperature.

15 Pacific cod larvae used in both experiments were hatched and cultured in the

16 National Oceanic and Atmospheric Administration (NOAA) Alaska Fisheries Science

17 Center (AFSC) Laboratory, Newport, Oregon in April 2008 and May 2009. Eggs were

18 obtained from adults at their spawning grounds in Chiniak Bay, Kodiak Island, Alaska. All

19 mixed gametes were held in incubation trays at 4°C, after which fertilized eggs (24 hours

20 post-fertilization) were shipped in insulated containers to the AFSC. Eggs were held in

1 flow-through plastic trays (~4 liter capacity) and maintained at 4oC until hatch,

2 approximately 19-22 days post-fertilization.

3 Walleye pollock eggs were collected from mature pollock broodstock maintained

4 by the AFSC at the Hatfield Marine Science Center. Eggs were collected using nylon

5 mesh collectors placed on the outflow of broodstock tanks. Larvae hatched in April

6 2008.

7 1.3.1 Pacific Cod Temperature Experiment 8

9 After hatching, larvae (n = 120) were immediately sampled to measure initial

10 condition. Larvae were then randomly assigned to two temperature treatments: cool

11 (5°C) and warm (8°C). Larvae at 0 days post hatch (dph) were in the size range ~4.4-5

12 mm total length. In each treatment, larvae were held in three tanks (100 liter capacity;

13 400 larvae each), with a total of six tanks. All tanks had a flow rate of ~250 mL minute-1

14 seawater and were initially held at 4°C, after which water temperatures were adjusted

15 to experimental conditions over 48 hours. Larvae were held at a 12:12 hour light:dark

16 photoperiod regime (Hurst et al. 2010).

17 Larvae in both treatments were fed on a combined diet of rotifers (Brachionus

18 plicatilis; twice daily) and microparticulate dry food (Otohime A, Marubeni Nisshin Feed

19 Co., Tokyo; two-three times daily). After the initial sampling after hatch, larvae were

20 sampled (10 larvae/tank/treatment) at 23 and 36 dph. At each sampling period, larvae

1 were removed from tanks using droppers, placed in individual 1500 µL microcentrifuge

2 vials on ice, and then stored at -80°C until analysis. All sampled larvae were individually

3 photographed before being placed in vials, to measure total length post-sampling using

4 imaging software. Larval fish lengths were measured to the nearest 0.005mm using NIS

5 Elements D, Nikon Instruments, Melville, NY.

6 Growth rates (length-GRL) were calculated for larvae in both treatments from each

7 sampling date (t2) to the hatch date (t1) using the formula:

8 GRL = (ln l t2 - ln l t1) / (t2 – t1)

9 where l = mean length (mm) and ln = natural logarithm.

10 1.3.2 Pacific Cod Starvation Experiment 11

12 Larvae were initially held at 6.5°C in two common tanks (100 liter capacity) with

13 a flow rate of ~250 mL minute-1 seawater, before transfer to fed and starved treatment

14 tanks (two tanks/treatment; 40 liter capacity) held at 6.5°C. All larvae at 0 dph were in

15 the size range ~5-6.5 mm total length. The feeding regimen was similar to the

16 temperature experiment.

17 During May to June 2009 larvae were sampled across three sampling periods

18 initiated at 0, 9, and 20 dph. The sampling protocol was similar to the temperature

19 experiment. At the start of each sampling, larvae were sampled from the common

20 tanks (10 larvae/tank), and randomly assigned to the fed or starved treatment tanks (60

1 larvae/tank). Larvae in the starved treatment tanks were sampled (10 larvae /tank)

2 when the population in the tanks reached 10% mortality and 50% mortality, with

3 corresponding sampling in the fed treatment. Individual larvae were photographed to

4 measure total length post-sampling using imaging software. The experiment was

5 terminated at 25 dph.

6 Larval fish lengths were measured to the nearest 0.005mm using Image-Pro

7 Plus, Media Cybernetics, Bethesda, MD. Growth rates (length-GRL) were calculated

8 similar to the temperature experiment.

9 1.3.3 Walleye Pollock Temperature Experiment 10 Larvae were cultured in duplicate tanks (n = 200 larvae/tank) at a cool (5°C) and

11 warm (8°C) temperature treatment. All larvae were in the size range ~6-8 mm length

12 and were fed for the duration of the experiment. All experimental conditions and

13 sampling protocols were identical to the Pacific cod temperature experiment, except for

14 the time of sampling (0, 19, and 40 dph). At 0 dph, larvae were sampled from each tank

15 (15 larvae/tank/treatment), followed by sampling at 19 and 40 dph (10

16 larvae/tank/treatment). Growth rates (length-GRL) were calculated similar to the Pacific

17 cod temperature experiment.

18 1.3.4 Biochemical Analyses 19

20 Nucleic acid ratios in both Pacific cod experiments and in the walleye pollock

21 experiment were measured under a one dye-one enzyme (RNase) fluorometric protocol

1 modified from Caldarone et al. (2001). Whole larvae were processed directly in their

2 individual storage vials thus avoiding any tissue loss. Initially, 150 µL 1% sarcosil Tris-

3 EDTA buffer was added to each sample vial. All sample vials were then vortexed for 60

4 minutes. Samples were further diluted with 1350 µL Tris-EDTA buffer, and centrifuged

5 for 15 minutes at 14000g in a Sorvall Legend Micro 21 R refrigerated centrifuge (Thermo

6 Scientific). Supernatants were then treated with 75µL ethidium bromide (5µg ml-1)

7 according to the protocol outlined by Caldarone et al. (2001). Total fluorescence was

8 measured in a Wallac 1420 microplate spectrophotometer (Perkin Elmer, Waltham, MA)

9 at excitation and emission wavelengths of 355nm and 600nm respectively. Samples

10 were treated with RNase, and the resulting reduced fluorescence measured to obtain

11 DNA fluorescence; RNA fluorescence was obtained through subtraction of DNA

12 fluorescence from the total fluorescence. Calibration curves were constructed using

13 serial dilutions of 18s-28s rRNA (Sigma R-0889) and calf thymus DNA (Sigma D-4764)

14 standards. Supernatants for R/D were read on Corning NBS 96-well black flat-bottom

15 microplates (75µL samples). Every R/D sample was run in three individual wells.

16 Fluorescence values for all three wells were read four times, and the coefficient of

17 variation associated with four reads was examined. When the variation was higher than

18 10%, the individual well fluorescence values were examined, and the data point causing

19 the high coefficient of variation was excluded from the analysis.

20 1.3.5 Statistical Analyses: 21

1 In all analyses, differences between groups were considered significant if p ≤

2 0.05

4 1.3.5.1 Pacific Cod Temperature Experiment 5

6 Growth was compared between larvae at 5°C and 8°C. A General Linear Model

7 (GLM) with Tukey’s posthoc test was used to compare length in larvae between

8 treatments over 0-36 dph, with length as the dependent variable, and temperature

9 treatment and sampling day as independent variables. A GLM was used to compare

10 instantaneous growth rates (IGR) in larvae between treatments, with IGR as the

11 dependent variable, and temperature treatment and sampling day as independent

12 variables. A GLM with Tukey’s test was used to compare R/D of larvae between

13 treatments, with R/D as the dependent variable, and temperature treatment and

14 sampling day as independent variables. All statistical analyses were made with Minitab

15 v. 14 (Minitab Inc., State College, Pennsylvania).

16 1.3.5.2 Pacific Cod Starvation Experiment 17

18 Growth was compared between larvae in fed and starved treatments. A GLM

19 was used to compare IGR of fed and starved larvae, with growth rates as the dependent

20 variable, and treatment as well as sampling day as predictor variables. A GLM with

21 Tukey’s test was used to compare R/D of fed and starved larvae over 0-25 dph, with R/D

1 as the dependent variable, and sampling day and treatment as the independent

2 variables

3 The R/D of starved larvae between 10% and 50% mortality was compared within

4 each sampling period using two sample T-tests. The R/D of starved larvae at both 10%

5 and 50% mortality was compared between the last two sampling periods with a two

6 sample T-test.

7 A GLM was used to compare individual nucleic acids of fed and starved larvae

8 over 0-25 dph, with RNA or DNA as the dependent variable, and sampling day and

9 treatment as independent variables. All statistical analyses were made with Minitab v.

10 14 (Minitab Inc., State College, Pennsylvania).

11 1.3.5.3 Pacific Cod Growth Model 12

13 Calibrated growth models were generated through multiple linear regression,

14 describing the relationship between instantaneous growth rates (IGR) and variables

15 including R/D, RNA, DNA, and temperature. A second-order Akaike’s information

16 criterion (AICc) was used to select the best fit model from the model data set

17 (Wagenmakers and Farrell 2004). Estimated growth rates were compared with

18 observed growth rates for all temperature treatments using a two-sample T-test. The

19 regression models were generated by Minitab v. 14 (Minitab Inc., State College,

20 Pennsylvania).

1 Depending on the model, R/D ratios or RNA and DNA from both Pacific cod

2 experiments were combined for the individual growth models in the model data set.

3 Mean tank values for each sampling day (R/D, RNA, DNA) were log-transformed. The

4 growth model only included data from larvae 23 and 36 dph in the temperature

5 experiment and from 9 dph onward (fed fish only) in the terminal starvation

6 experiment. This was due to lack of a correlation between R/D and growth in the early

7 larval stages, possibly due to presence of pre-existing yolk protein.

8 1.3.5.4 Walleye Pollock Temperature Experiment 9

10 Growth was compared between larvae at 5°C and 8°C. All statistical analyses

11 were identical to the Pacific cod temperature experiment and were made with Minitab

12 v. 14 (Minitab Inc., State College, Pennsylvania).

13 1.4 Results

14 1.4.1 Pacific Cod Temperature Experiment

15 Larvae at 8°C were longer than larvae at 5°C at the end of the experiment (p <

16 0.001; Fig 2.1a) with a difference in length between treatments at 23 dph and 36 dph (p

17 < 0.001; Table 2.1). Larvae at 8°C grew faster than larvae at 5°C and had significantly

18 higher instantaneous growth rates over the sampling period (p < 0.001; Table 2.1).

19 Larvae at 5°C had higher R/D than larvae at 8°C at the end of the experiment (Fig. 2.1b),

20 with a difference between temperature treatments at 36 dph (p < 0.001; Table 2.1).

1 1.4.2 Pacific Cod Starvation Experiment

2 Fed larvae had significantly higher instantaneous growth rates (p < 0.001) over 5-

3 25 dph relative to starved larvae (Fig. 2.2a; Table 2.2). Nucleic acid ratios were higher in

4 fed larvae relative to starved larvae over 0-25 dph (p < 0.001; Fig. 2.2b); there were

5 differences in nucleic acid ratios between fed and starved larvae at 5 dph (p = 0.0061),

6 11 dph (p = 0.0101), 15 dph (p = 0.0003), 23 dph (p = 0.0180), and 25 dph (p = 0.0458;

7 Table 2.2).

8 A comparison of R/D between 10% and 50% mortality in starved larvae showed

9 differences over 11-15 dph (p = 0.034) but not over 2-5 dph (p = 0.134) or 23-25 dph (p =

10 0.798). Comparison of R/D at 10% mortality between the second and third sampling

11 periods (11 and 23 dph) showed no difference (p = 0.071). However, R/D differed at

12 50% mortality (p = 0.006) between second and third sampling periods (15 and 25 dph).

13 The RNA and DNA concentrations were higher in fed larvae relative to starved

14 larvae over 0-25 dph (p < 0.001). The RNA concentrations in the fed larvae remained

15 steady until 9 dph, increasing slightly by 11 dph (Fig. 2.3b). From 11-20 dph RNA

16 concentrations remained steady, and increased after 20 dph. In contrast, DNA

17 concentrations in both fed and starved larvae increased until 11 dph (Fig. 2.3c). From

18 11-20 dph, DNA concentrations in fed larvae remained steady, and increased after 20

19 dph.

1 1.4.3 Pacific Cod Growth Model Synthesis

2 A significant positive relationship was found between IGR, log RNA, log DNA, and

3 treatment temperature (p < 0.001; Table 2.3). This growth model was chosen from the

4 model data set based on a comparison of AICc values (AICc = -122; Table 2.4), and

5 incorporated IGR, log RNA, log DNA, and treatment temperature in a multiple linear

6 regression. The response variable was IGR; log RNA, log DNA, and treatment

7 temperature were predictor variables. Estimated and observed growth rates did not

8 differ for fish at 5°C (p = 0.0180), 6.5°C (p = 0.479), and 8°C (p = 0.259)

9 1.4.4 Walleye Pollock Temperature Experiment

10 Larvae at 8°C were longer than larvae at 5°C at the end of the experiment (p <

11 0.001; Fig. 2.4a), with a difference between treatments at 40 dph (p < 0.001; Table 2.5).

12 Larvae at 8°C grew faster than larvae at 5°C and had significantly higher instantaneous

13 growth rates over the sampling period (p < 0.001; Table 2.5). Larvae at 5°C had higher

14 R/D than larvae at 8°C at the end of the experiment (p < 0.001; Fig 2.4b), with a

15 difference between treatments at 0 dph (p = 0.0050) and 40 dph (p < 0.001; Table 2.5).

16 1.5 Discussion 17 Lower R/D coupled with higher growth in larvae at 8°C indicated that

18 translational efficiency defines early growth in Pacific cod and walleye pollock. Early

19 larval fish growth is predominantly due to protein synthesis and deposition (Tong et al.

20 2010), determined by RNA-specific protein synthesis rate (translational efficiency) and

21 RNA concentration (translational capacity; Smith and Ottema 2006, Fraser and Rogers

1 2007). The relative effects of translational efficiency and capacity on growth can be

2 influenced by temperature (Foster et al. 1992, Treberg et al. 2005). Translational

3 efficiency also defines early larval growth in Atlantic herring (Clupea harengus; Houlihan

4 et al. 1995), African catfish (Clarias gariepinus; Smith and Ottema 2006), and rainbow

5 trout (Oncorhynchus mykiss; Peragon et al. 2001).

6 Growth in larval Pacific cod and walleye pollock is affected by small differences

7 in temperature. Exposure to lower temperature reduced translational efficiency in both

8 gadids, resulting in growth driven predominantly by translational capacity. This was

9 suggested by a lack of correspondence between R/D and growth rates. As expected,

10 larvae at 8ºC had higher growth rates and were longer than larvae at 5°C. This higher

11 growth, coupled with lower R/D relative to larvae at 5°C, suggested heightened

12 translational efficiency. The higher R/D at 5ºC suggests a compensatory mechanism to

13 maintain protein synthesis, i.e., heightened translational capacity in lieu of reduced

14 translational efficiency. But with the high costs of RNA and protein synthesis

15 (Bocharova et al. 1992, Houlihan et al. 1995, Smith and Ottema 2006), regardless of cost

16 mitigation strategies (Houlihan et al. 1988, Wieser 1995, Smith et al. 2000, Smith and

17 Ottema 2006), higher translational efficiency is energetically advantageous. The effect

18 of this trade-off between translational efficiency/capacity could be significant as larvae

19 grow and increase in size. Another potential reason for higher R/D at 5°C could be

20 mismatched RNA degradation and synthesis rates, due to a temperature effect on

21 enzymes essential to these processes. In this case, higher R/D would not be an active

1 compensatory measure corresponding to heightened protein synthesis. While protein

2 synthesis rates were not measured in this experiment, Atlantic cod exposed to colder

3 temperatures showed heightened translational capacity active in maintaining protein

4 synthesis (Foster et al. 1992, Treberg et al. 2005), suggesting that the higher R/D in this

5 experiment was a compensatory strategy.

6 The trade-off between translational efficiency/capacity can be examined in

7 terms of the R/D divergence between larvae. The timing of the divergence suggested

8 heightened energetic demands as larvae aged, possibly due to growth and increased

9 protein synthesis, as larvae increased in size and structural complexity. Crucially, the

10 consistently lower growth at 5°C suggested lower translational efficiency prior to the

11 divergence in R/D ratios. However, the lack of an immediate compensatory response

12 could be due to energetic costs involved in heightened translational capacity as well as

13 lower larval energy requirements. As larvae grow, energetic constraints could lead to

14 heightened translational capacity. However, this could be energetically expensive,

15 potentially affecting long-term growth as in Atlantic cod larvae (Koedijk et al. 2010a, b).

16 This experiment highlighted early larval Pacific cod and walleye pollock growth

17 strategies in terms of translational efficiency and capacity. Pacific cod and walleye

18 pollock larvae appear to have physiological compensatory mechanisms to maintain

19 growth trajectories, which could be energetically expensive. Cellular growth indices

20 such as R/D can highlight physiological growth strategies, a prerequisite to further

1 studying any associated metabolic costs which should be accounted for in growth

2 assessments.

3 The terminal fasting experiment examined the growth and R/D relationship in

4 Pacific cod larvae (0-25 dph) based on nutritional state. In contrast to other vertebrates,

5 where post-natal growth is hypertrophic (an increase in muscle fiber size; Smith et al.

6 2000) teleost growth is a combination of hypertrophy and hyperplasia (formation of

7 new muscle fibers). Early teleost development involves hyperplasic organogenesis

8 (Tanaka et al. 1996, Zouiten et al. 2008, Tong et al. 2010) followed by somatic growth

9 which is both hyperplasic and hypertrophic (Westerman and Holt 1994, Tong et al.

10 2010). These development patterns are common across species, shown by a steady

11 decrease in R/D ratios over the immediate post-hatch period (Clemmeson 1987,

12 Puvanendran and Brown 1999, Caldarone et al. 2003).

13 In common with other fish species, R/D in Pacific cod larvae decreased over the

14 initial post-hatch period, and then stabilized. The initial decrease was due to increasing

15 DNA in fed and starved larvae (0-11 dph) coupled with stable (fed group) or reduced

16 (starved group) RNA. The increased DNA content with stable or reduced RNA could

17 correspond to hyperplasic organogenesis. Both RNA and DNA in fed larvae stabilized

18 over 11-20 dph, potentially the transition phase after yolk absorption to complete

19 reliance on exogenous food. Stable R/D in fed larvae implied a lack of growth during the

20 transition to external food, evidenced by a corresponding decrease in growth rate. A

1 similar trend was observed in larval Turbot (Scophthalmus maximus) after yolk depletion

2 (Tong et al. 2010). Laurel et al. (2008) found that Pacific cod larvae had no growth

3 immediately after yolk absorption, while fed Atlantic cod larvae also had the lowest

4 growth and R/D coinciding with yolk absorption (Caldarone et al. 2003). Atlantic cod

5 larvae have also shown closely corresponding transition periods (Yin and Blaxter 1986,

6 Hunt von Herbing et al. 1996, Caldarone et al. 2003) at similar temperatures. While the

7 rate of yolk absorption can be temperature dependent (Laurel et al. 2008), the timing of

8 the transition phase (11-20 dph) integrates the initial increase in DNA (0-11 dph) with

9 yolk absorption.

10 The timing of the transition phase suggested yolk absorption over the preceding

11 0-11 dph, coinciding with the increase in DNA potentially indicating organogenesis.

12 Since organogenesis can utilize lipid and protein components of yolk (Cejas et al. 2004),

13 yolk in Pacific cod larvae appears primarily utilized in organogenesis. While yolk is the

14 most cost-effective source of protein, the energetic cost of protein absorption remains

15 substantial (Smith and Ottema 2006). However, with even higher costs of protein

16 synthesis (Hawkins 1991, Smith and Houlihan 1995), efficient larval growth strategies

17 emphasize absorption of preexisting yolk protein into tissues (Wieser et al. 1988).

18 Similar growth strategies have been observed in larval African catfish (Clarias

19 gariepinus; Smith and Ottema 2006), Atlantic herring (Clupea harengus; Houlihan et al.

20 1995), and nase (Chondrostoma nasus; Houlihan et al. 1992). Yolk absorption coinciding

21 with organogenesis was observed in both starved and fed treatments. This suggested

1 the relative importance of organogenesis, as starved larvae did not divert any yolk

2 towards somatic growth, and had relatively lower growth rates, R/D, and RNA.

3 In conjunction with increases in R/D and RNA after the transition period (20

4 dph), the growth rate in fed larvae increased. This indicated the channeling of resources

5 to somatic growth after organogenesis. In larval Pacific cod, increased growth after the

6 transition phase also coincided with the start of diel vertical migrations (Hurst et al.

7 2009) and heightened responses to prey (Colton and Hurst 2010). In common with

8 other fish species, this growth phase, characterized by increasing RNA, DNA, R/D, and

9 growth rate, is probably a combination of hypertrophy and hyperplasia (Tong et al.

10 2010).

11 Nucleic acid ratios of starved larvae could potentially be used as indicators of

12 terminal starvation, i.e., a “baseline” ratio. Averages of ratios from larvae sampled at

13 10% and 50% starvation over the 2nd and 3rd sampling periods, while preliminary and

14 approximate (~2.5), have promise as a starvation index, but require further validation

15 studies.

16 The Pacific cod growth model incorporated nucleic acids (RNA and DNA) and

17 temperature as variables in predicting growth rate. This model estimated 65.5% of the

2 18 observed variation in growth rate. Based on both adjusted r and AICC values, individual

19 nucleic acids appear to predict early larval growth in Pacific cod better than a model

20 based on R/D ratios. With the lack of correlation between growth and R/D in early stage

1 yolk-sac larvae (Caldarone et al. 2003), this growth model did not include data from 0

2 dph larvae (temperature experiment) or from 0-5 dph larvae (terminal starvation

3 experiment). The ready availability of protein (yolk) in early larval stages could cause

4 the lack of correlation between nucleic acids and growth. This growth model relates

5 nucleic acids to growth and condition of individual larvae and can have applications in

6 management, since growth and condition estimates are required in recruitment and

7 stock assessments. This model is useful in estimating growth in Pacific cod at an early

8 life-stage. Additionally, it enables growth estimation in field-sampled larvae using a

9 single measure, a requirement since it is not realistically possible to periodically sample

10 the same fish in the field to estimate growth.

11 1.6 Conclusion 12 Nucleic acid ratios have potential as a growth index in larval Pacific cod and

13 walleye pollock; this index was sensitive enough to bring out differences in early larval

14 growth responses and physiological strategies. Larval Pacific cod and walleye pollock

15 growth strategies are affected by small differences in temperature, while nutritional

16 state also affected early growth in Pacific cod. Exposure to a lower temperature

17 resulted in higher R/D in both cod and pollock larvae, potentially a compensatory

18 strategy. This could be due to increased energetic demands as larvae increase in size

19 and complexity. These higher ratios were only manifested after a period of consistently

20 lower growth, suggesting costs associated with this strategy.

1 Fed Pacific cod larvae as expected had higher growth relative to starved larvae,

2 shown by R/D and growth rates. Changes in R/D, RNA, and DNA could be used to

3 identify distinct stanzas in early growth of Pacific cod larvae, possibly yolk absorption

4 and organogenesis, exogenous feeding (transition phase), and post-transition growth.

5 Similar growth stages are seen in larvae of other fish species. Yolk appears

6 predominantly utilized towards organogenesis, even in starved fish. While R/D reduced

7 in both treatments over the yolk absorption phase, the R/D in the starved treatment

8 remained lower, indicating that exogenous food supply may be added to energy

9 reserves available to growth despite the presence of yolk. The Pacific cod growth model

10 was able to estimate 65.5% of observed variation in growth rate. This model could have

11 applications in management, since growth and condition estimates are required for

12 recruitment and stock assessments.

13 1.7 Acknowledgements 14 We thank Elaine Caldarone, who was extremely generous with her time in

15 helping to adapt her R/D protocol to this research, as well as Dr. Franz Mueter,

16 University of Alaska Fairbanks, and Dr. Chris Hay-Jahans, University of Alaska Southeast,

17 for their advice on statistical methods. We also thank Dr. Sherry Tamone, University of

18 Alaska Southeast, the Rasmuson Fisheries Foundation for fellowship support awarded to

19 AS, and NOAA, who supported this research.

20 1.8 Literature Cited 21 Alderdice DF, Forrester CR (1971) Effects of salinity, temperature, and dissolved oxygen

1 on early development of the Pacific cod (Gadus macrocephalus). J Fish Res

2 Board Can 28: 883-902.

3 Amberg JJ, Myr C, Kamisaka Y, Jordal A.-E.O, Rust MB, Hardy RW, Koedijk R,

4 Ronnestad I (2008) Expression of the oligopeptide transporter, PepT1, in larval

5 Atlantic cod (Gadus morhua). Comp Biochem Physiol B 150: 177-182.

6 Bacheler NM, Ciannelli L, Bailey KM, Duffy-Anderson JT (2010) Spatial and

7 temporal patterns of walleye pollock (Theragra chalcogramma) spawning in the

8 eastern Bering Sea inferred from egg and larval distributions. Fish Oceanogr 19:

9 107-120.

10 Bakkala RG (1984) Pacific cod of the eastern Bering Sea. Bull Int North Pac Fish Comm

11 42: 157-179.

12 Bakkala RG, Mishima S, Westrheim SJ, Zhang CI, Brown ES (1984) Distribution of

13 Pacific cod (Gadus macrocephalus) in the North Pacific Ocean. Bull Int North Pac

14 Fish Comm 42: 111-115.

15 Beamish RJ, Benson AJ, Sweeting RM, Neville CM (2004) Regimes and the

16 history of the major fisheries off Canada’s west coast. Prog Oceanogr 60: 355-

17 385.

18 Blaxter JHS (1992) The effect of temperature on larval fishes. Neth J Zool 42: 336-357.

1 Bocharova LS, Gordon RY, Arkhipov VI (1992) Uridine uptake and RNA synthesis in

2 the brain of torpid and awakened ground squirrels. Comp Biochem Physiol B

3 101: 189-192.

4 Buckley LJ (1979) Relationships between RNA-DNA ratio, prey density, and growth rate

5 in Atlantic cod (Gadus morhua) larvae. J Fish Res Board Can 36: 1497-1502.

6 Buckley LJ (1984) RNA-DNA ratio: an index of larval fish growth in the sea. Mar Biol

7 80: 291-298.

8 Buckley L, Caldarone E, Ong T.-L (1999) RNA–DNA ratio and other nucleic

9 acid-based indicators for growth and condition of marine fishes. Hydrobiologia

10 401: 265-277.

11 Caldarone EM (2005) Estimating growth in haddock larvae Melanogrammus

12 aeglefinus from RNA:DNA ratios and water temperature. Mar Ecol Prog Ser 293:

13 241-252.

14 Caldarone EM, St. Onge-Burns JM, Buckley LJ (2001) Protocol and guide for

15 estimating nucleic acids in larval fish using a fluorescence microplate reader.

16 Northeast Fisheries Science Center Reference Document 01-11.

17 Caldarone EM, St. Onge-Burns JM, Buckley LJ (2003) Relationship of

1 RNA/DNA ratio and temperature to growth in larvae of Atlantic cod Gadus

2 morhua. Mar Ecol Prog Ser 262: 229-240.

3 Cejas JR, Almansa E, Jerez S, Bolanos A, Felipe B, Lorenzo A (2004) Changes

4 in lipid class and fatty acid composition during development in white seabream

5 (Diplodus sargus) eggs and larvae. Comp Biochem Physiol B: 139: 209-216.

6 Clemmesen CM (1987) Laboratory studies on RNA/DNA ratios of starved and fed

7 herring (Clupea harengus) and turbot (Scophthalmus maximus) larvae. J Cons int

8 Explor Mer 43: 122-128.

9 Colton AR, Hurst TP (2010) Behavioral responses to light gradients, olfactory cues, and

10 prey in larvae of two North Pacific gadids (Gadus macrocephalus and Theragra

11 chalcogramma). Environ Biol Fish 88: 39-49.

12 DeMarch BGE (1995) Effects of incubation temperature on the hatching success of

13 Arctic char eggs. Prog Fish-Cult 57: 132-136.

14 Foster AR, Houlihan DF, Hall SJ, Burren LJ (1992) The effect of temperature

15 acclimation on protein synthesis rates and nucleic acid content of juvenile cod

16 (Gadus morhua L.). Can J Zool 70: 2095-2102.

17 Fraser KPP, Rogers AD (2007) Protein metabolism in marine animals: the

1 underlying mechanism of growth. Adv Mar Biol 52: 267-362.

2 Gabillard J-C, Weil C, Rescan P-Y, Navarro I, Gutierrez J, Le Bail P-Y (2003a)

3 Effects of environmental temperature on IGF1, IGF2, and IGF type I receptor

4 expression in rainbow trout (Oncorhynchus mykiss). Gen Comp Endocrinol 133:

5 233-242.

6 Gabillard J-C, Weil C, Rescan P-Y, Navarro I, Gutierrez J, Le Bail P-Y (2003b)

7 Environmental temperature increases plasma GH levels independently of

8 nutritional status in rainbow trout (Oncorhynchus mykiss). Gen Comp Endocrinol

9 133: 17-26.

10 Grant WS, Zhang CI, Kobayashi T, Stahl G (1987) Lack of genetic stock

11 discretion in Pacific cod (Gadus macrocephalus). Can J Fish Aquat Sci 44: 490-

12 498.

13 Hawkins AJS (1991) Protein turnover: a functional appraisal. Funct Ecol 5:

14 222-223.

15 Hiatt T, Dalton M, Felthoven R, Fissel B, Garber-Yonts B, Haynie A, Kasperski S, Lew

16 D, Package C, Sepez J, Seung C (2010) Stock assessment and fishery evaluation

1 report for the groundfish fisheries of the Gulf of Alaska and Bering Sea/Aleutian

2 Islands area: Economic status of the groundfish fisheries off Alaska, 2009.

3 Published at http://www.afsc.noaa.gov/refm/docs/2010/economic.pdf

4 Houlihan DF, Pedersen BH, Steffensen JF, Brechin J (1995) Protein synthesis,

5 growth, and energetics in larval herring (Clupea harengus) at different feeding

6 regimes. Fish Physiol Biochem14: 195-208.

7 Houlihan DF, Hall SJ, Gray C, Noble BS (1988) Growth rates and protein

8 turnover in Atlantic cod, Gadus morhua. Can J Fish Aquat Sci 45: 951-964.

9 Houlihan DF, Wieser W, Foster A, Brechin J (1992) In vitro protein synthesis

10 rates in larval nase (Chondrostoma nasus L.). Can J Zool 70: 2436-2440.

11 Hurst TP, Cooper DW, Scheingross JS, Seale EM, Laurel BJ, Spencer ML (2009) Effects

12 of ontogeny, temperature, and light on vertical movements of larval Pacific cod

13 (Gadus macrocephalus). Fish Oceanogr 18: 301-311.

14 Hurst TP, Laurel BJ, Cianelli L (2010) Ontogenetic patterns and temperature-dependent

15 growth rates in early life stages of Pacific cod (Gadus macrocephalus). Fish Bull

16 108: 382-392.

1 Hunt von Herbing I, Boutilier RG, Miyake T, Hall BK (1996) Effects of

2 temperature on morphological landmarks critical to growth and survival in larval

3 Atlantic cod (Gadus morhua). Mar Biol 124: 593-606.

4 Janhunen M, Piironen J, Peuhkuri N (2010) Parental effects on embryonic

5 viability and growth in Arctic charr Salvelinus alpinus at two incubation

6 temperatures. J Fish Biol 76: 2558-2570.

7 Jewett SC (1978) Summer food of the Pacific cod, Gadus macrocephalus, near Kodiak

8 Island, Alaska. Fish Bull 76: 700-706.

9 Jonsson N, Jonsson B, Hansen LP (2005) Does climate during embryonic development

10 influence parr growth and age of seaward migration in Atlantic salmon (Salmo

11 salar)? Can J Fish Aquat Sci 62: 2502-2508.

12 Kamler E (2002) Ontogeny of yolk-feeding fish: an ecological perspective. Rev

13 Fish Biol Fish 12: 79-103.

14 Kamler E (2008) Resource allocation in yolk-feeding fish. Rev Fish Biol Fish 18: 143-

15 200.

16 Ketchen KS (1961) Observations on the ecology of the Pacific cod (Gadus

1 macrocephalus) in Canadian waters. J Fish Res Board Can 18: 513-558.

2 Koedijk RM, Le Francois NR, Blier PU, Foss A, Folkvord A, Ditlecadet D, Lamarre SG,

3 Stefansson SO, Imsland AK (2010a) Ontogenetic effects of diet during early

4 development on growth performance, myosin mRNA expression and metabolic

5 enzyme activity in Atlantic cod juveniles reared at different salinities. Comp

6 Biochem Physiol A 156: 102-109.

7 Koedijk RM, Folkvord A, Foss A, Pittman K, Stefansson SO, Handeland S, Imsland AK

8 (2010b) The influence of first-feeding diet on the Atlantic cod Gadus morhua

9 phenotype: survival, development, and long-term consequences for growth. J

10 Fish Biol 77: 1-19.

11 Laurel BJ, Copeman LA, Hurst TP, Parrish CC (2010) The ecological significance of

12 lipid/fatty acid synthesis in developing eggs and newly hatched larvae of Pacific

13 cod (Gadus macrocephalus). Mar Biol 157: 1713-1724.

14 Laurel BJ, Hurst TP, Cianelli L (2011) An experimental examination of temperature

15 interactions in the match-mismatch hypothesis for Pacific cod larvae. Can J Fish

16 Aquat Sci 68: 51-61.

17 Laurel BJ, Hurst TP, Copeman LA, Davis MW (2008) The role of temperature on the

1 growth and survival of early and late hatching Pacific cod larvae (Gadus

2 macrocephalus). J Plankton Res 30: 1051-1060.

3 Lewis JM, Driedzic WR (2007) Tissue specific changes in protein synthesis

4 associated with seasonal metabolic depression and recovery in the north

5 temperate labrid, Tautogolabrus adspersus. Am J Physiol-Reg Integr Comp

6 Physiol 293: R474-R481.

7 Li M, Leatherland J (2008) Temperature and ration effects on components of the

8 IGF system and growth performance of rainbow trout (Oncorhynchus mykiss)

9 during the transition from late stage embryos to early stage juveniles. Gen

10 Comp Endocrinol 155: 668-679.

11 Mecklenburg CW, Mecklenburg TA, Thorsteinson LK (2002) Fishes of Alaska.

12 American Fisheries Society, Maryland.

13 McLaughlin RL, Ferguson MM, Noakes DLG (1995) Concentrations of nucleic

14 acids and protein as indices of nutritional status for recently emerged brook

15 trout (Salvelinus fontinalis). Can J Fish Aquat Sci 52: 848-854.

16 McMillan DN, Houlihan DF (1988) The effect of refeeding on tissue protein

17 synthesis in rainbow trout. Physiol Zool 61: 429-441.

1 Pavlov DA, Mikhaylenko VG, Timeyko VN, Konovalov Ye.S (1994) Reproduction and

2 embryonic larval development of the char, Salvelinus alpinus lepechini, in lakes

3 Onega and Ladoga. J Ichthyol 34: 42-56.

4 Peragon J, Barroso JB, Salguero LG, de la Higuera M, Lupianez JA (2001)

5 Growth, protein-turnover rates and nucleic-acid concentrations in the white

6 muscle of rainbow trout during development. Int J Biochem Cell Biol 33: 1227-

7 1238.

8 Puvanendran V, Brown JA (1999) Foraging, growth and survival of Atlantic cod

9 larvae reared in different prey concentrations. Aquaculture 175: 77-92.

10 Robinson SMC, Ware DM (1988) Ontogenetic development of growth rates in

11 larval Pacific herring, Clupea harengus pallasi, measured with RNA-DNA ratios in

12 the Strait of Georgia, British Columbia. Can J Fish Aquat Sci 45: 1422-1429.

13 Sakurai Y, Hattori T (1996) Reproductive behavior of Pacific cod in captivity.

14 Fish Sci 62: 222-228.

15 Smith RW, Ottema C (2006) Growth, oxygen consumption, and protein and RNA

16 synthesis rates in the yolk sac larvae of the African catfish (Clarias gariepinus).

17 Comp Biochem Physiol A 143: 315-325.

1 Smith RW, Houlihan DF (1995) Protein synthesis and oxygen consumption in

2 fish cells. J Comp Physiol B 165: 93-101.

3 Smith RW, Palmer RM, Houlihan DF (2000) RNA turnover and protein

4 synthesis in fish cells. J Comp Physiol B 170: 135-144.

5 Sogard SM, Olla BL (1996) Diel patterns of behavior in juvenile walleye pollock,

6 Theragra chalcogramma. Environ Biol Fish 47: 379-386.

7 Stepanenko MA (1995) Distribution, behavior, and abundance of Pacific cod, Gadus

8 macrocephalus, in the Bering Sea. J Ichthyol 35: 18-27.

9 Stierhoff KL, Targett TE, Power JH (2009) Hypoxia-induced growth limitation

10 of juvenile fishes in an estuarine nursery: assessment of small-scale temporal

11 dynamics using RNA:DNA. Can J Fish Aquat Sci 66: 1033-1047.

12 Takatsu T, Nakatani T, Mutoh T, Takahashi T (1995) Feeding habits of Pacific

13 cod larvae and juveniles in Mutsu bay, Japan. Fish Sci 61: 415-422.

14 Tanaka M, Kawai S, Seikai T, Burke JS (1996) Development of the digestive

15 organ system in Japanese flounder in relation to metamorphosis and settlement.

16 Mar Freshw Behav Physiol 28: 19-31.

1 Teletchea F, Fontaine P (2010) Comparison of early life-stage strategies in

2 temperate freshwater fish species: trade-offs are directed towards

3 first feeding of larvae in spring and early summer. J Fish Biol 77: 257-278.

4 Theilacker GH, Bailey KM, Canino MF, Porter SM (1996) Variations in larval walleye

5 pollock feeding and condition: A synthesis. Fish Oceanogr 5 (Suppl 1): 112-123.

6 Tong XH, Liu QH, Xu SH, Li J, Xiao ZZ, Ma DY (2010) Changes in RNA, DNA,

7 protein contents and growth of turbot Scophthalmus maximus larvae and

8 juveniles. J Fish Biol 77: 512-525.

9 Treberg JR, Hall JR, Driedzic WR (2005) Enhanced protein synthetic capacity

10 in Atlantic cod (Gadus morhua) is associated with temperature-induced

11 compensatory growth. Am J Physiol-Reg Integr Comp Physiol 288: R205-R211.

12 Wagenmakers EJ, Farrell S (2004) AIC model selection using Akaike weights. Psychon

13 Bull Rev 11: 192-196.

14 Weber LP, Higgins PS, Carlson RI, Janz DM (2003) Development and validation of

15 methods for measuring multiple biochemical indices of condition in juvenile

16 fishes. J Fish Biol 63: 37-658.

1 Westerman M, Holt GJ (1994) RNA:DNA ratio during the critical period and

2 early larval growth of the red drum Sciaenops ocellatus. Mar Biol 121: 1-9.

4 Wieser W (1995) Energetics of fish larvae, the smallest vertebrates. Acta Physiol Scand

5 154: 279-290.

6 Wieser W, Forstner H, Medgyesy N, Hinterleitner S (1988) To switch or not to

7 switch: partitioning of energy between growth and activity in larval cyprinids

8 (Cyprinidae: Teleostei). Funct Ecol 2: 499-507.

9 Yin MC, Blaxter JHS (1986) Morphological changes during growth and starvation of

10 larval cod (Gadus morhua L.) and flounder (Platichthys flesus L.). J Exp Mar Biol

11 Ecol 104: 215-228.

12 Zouiten D, Khemis IB, Besbes R, Cahu C (2008) Ontogeny of the digestive tract

13 of thick lipped grey mullet (Chelon labrosus) larvae reared in “mesocosms”.

14 Aquaculture 279: 166-172.

1 1.9 Figures and Tables 2

3 4 Figure. 1.1a,b: Gadus macrocephalus. Length and RNA/DNA in larvae at 5°C and 8°C. 5 Comparison of length (1a) and RNA/DNA (R/D) ratios (1b) between Pacific cod larvae 6 cultured at 5°C (open symbols) and 8°C (closed symbols). Symbol bars represent

7 standard error. Significant differences (p ≤ 0.05) between treatments represented by

1 2

3 Figure. 1.2a,b: Gadus macrocephalus. Growth rate and RNA/DNA in fed and starved 4 larvae. Comparison of instantaneous growth rates (I.G.R) (2a) and RNA/DNA (R/D) 5 ratios (2b) between fed (closed symbols) and starved (open symbols) Pacific cod larvae. 6 Symbol bars represent standard error. Significant differences (p ≤ 0.05) between fed

7 and fasted treatments represented by

2 3

4 Figure. 1.3a,b,c: Gadus macrocephalus. Nucleic acid trends in fed and starved larvae. 5 Comparison of RNA/DNA (R/D) ratios (2.3a), nucleic acids (RNA and DNA) (2.3b), and 6 DNA (2.3c) between fed (closed symbols) and starved (open symbols) Pacific cod larvae 7 from 0-25 days post hatch. Significant differences (p ≤ 0.05) between fed and starved 8 groups shown by

6 7 Figure. 1.4a,b: Theragra chalcogramma. Length and RNA/DNA in larvae at 5°C and 8°C. 8 Comparison of length (2.4a) and RNA/DNA (R/D) ratios (2.4b) between walleye pollock 9 larvae cultured at 5°C (open symbols) and 8°C (closed symbols). Symbol bars represent 10 standard error. Significant differences (p ≤ 0.05) between treatments represented by

Table 1.1: Gadus macrocephalus. Growth in larvae at 5°C and 8°C. Statistical comparison of total length, instantaneous growth rate (I.G.R.), and RNA/DNA between Pacific cod larvae cultured at two temperature treatments (5°C, 8°C) over 0-36 days post hatch (dph). P-values ≤ 0.05 are statistically significant.

Comparison Tukey posthoc test comparison between 5°C and over 0-36 dph 8°C treatments

0 dph 23 dph 36 dph

Length < 0.001 1.0000 < 0.001 < 0.001

I.G.R. < 0.001

RNA/DNA 0.001 0.9949 1.0000 < 0.001

Table 1.2: Gadus macrocephalus. Growth in fed and starved larvae at 6.5°C. Statistical comparison of instantaneous 3 growth rate (I.G.R.), and RNA/DNA (R/D) between larval Pacific cod cultured at two nutritional states (fed and starved)4 at 6.5°C over 0-25 days post hatch (dph). P-values ≤ 0.05 are statistically significant. 5

6 Between treatment Tukey posthoc test comparison (between treatment comparison for each sampling day) comparison over 0-25 7

dph

0 dph 2 dph 5 dph 9 dph 11 dph 15 dph 20 dph 23 dph 25 dph

I.G.R. < 0.001

R/D < 0.001 1.000 0.2191 0.0061 1.0000 0.0101 0.0003 1.0000 0.0180 0.0458

Table 1.3: Gadus macrocephalus. Growth model regression. Multiple linear regression coefficients describing the relationship between instantaneous growth rates (I.G.R.), log

RNA-a1, log DNA-b1, and temperature-c1 for Pacific cod larvae. Equation is of the form

I.G.R. = a1+b1+ c1+C, where C is a constant. The regression was significant (p < 0.001). SE is standard error.

2 Coefficients a1 b1 c1 C Adjusted r (+/- SE)

Pacific cod 0.002064 0.003922 0.0002160 0.004515 0.655 (0.002950) (0.003243) (0.0004581) (0.006022)

Table 1.4: Gadus macrocephalus. AICC comparison. Model selection by second-order Akaike Information Criterion (AICC) values. Coefficients (± SE) in alternative regression models of growth in larval Pacific cod. In all models, p < 0.001. Log R/D = Log RNA/DNA ratio, T = treatment temperature, K = number of model parameters (including intercept), Δ AIC = difference in AIC with respect to the best fit model.

Model Paarmeters

2 # RNA Log RNA DNA Log DNA Log R/D T Intercept Adjusted r K AICc Δ AIC

1 0.0020 0.0039 0.0002 0.004 0.655 4 -122 0

(0.0029) (0.0032) (0.0004)

2 0.0002 0.0012 0.0002 0.006 0.539 4 -119 3

(0.0002) (0.0013) (0.0005)

3 0.0035 0.0014 -0.002 0.127 3 -114 8

(0.0049) (0.0006)

Table 1.5: Theragra chalcogramma. Growth in larvae at 5°C and 8°C. Statistical comparison of total length, instantaneous growth rate, and RNA/DNA between walleye pollock larvae cultured at two temperature treatments (5°C, 8°C) over 0-40 days post hatch (dph). P-values ≤ 0.05 are statistically significant.

Comparison Tukey posthoc test comparison between 5°C and over 0-40 dph 8°C treatments

0 dph 19 dph 40 dph

Length < 0.001 0.9447 0.0899 < 0.001

I.G.R. < 0.001

R/D < 0.001 0.0050 0.2333 < 0.001

55 56

1 Chapter 2 - Conceptual model of energy allocation in walleye 2 pollock (Theragra chalcogramma) from larvae to age-1 in the 3 southeastern Bering Sea 4 Elizabeth C. Siddona*, Ron Heintzb, Franz J. Muetera

5 Published in Deep Sea Research II (2013)

6 2.1 Abstract 7 Walleye pollock (Theragra chalcogramma) support the largest commercial fishery in the

8 United States and are an ecologically important component of the eastern Bering Sea

9 (EBS) pelagic ecosystem. Alternating climate states influence the recruitment success of

10 walleye pollock through bottom-up control of zooplankton communities. Relating the

11 seasonal progression of energy content and allocation to the distribution and

12 abundance of walleye pollock allows for detection of spatial and temporal trends in fish

13 condition. This provides critical information for predicting overwinter survival and

14 recruitment to age-1 because age-0 walleye pollock rely on energy reserves to survive

15 their first winter. Larval, age-0, and age-1 walleye pollock were collected in the EBS from

16 May to September 2008-2010. Fish condition was determined through quantification of

17 energy density (kJ/g) and proximate composition (% lipid, protein, ash, moisture) with

18 variation in energy density primarily driven by variability in % lipid. Energy densities

19 remained relatively low during the larval phase in early summer, indicating energy

20 allocation to growth and development. Lipid acquisition rates increased rapidly after

21 transformation to the juvenile form, with energy allocation to lipid storage leading to

22 higher energy densities in fall. We propose that this physiologically and ecologically

23 important shift, occurring after the end of larval development (25-40 mm), represents a

56 57

1 short critical period for determining fall condition and overwinter survival of juvenile

2 walleye pollock.

3 2.2 Keywords: 4 Walleye pollock (Theragra chalcogramma), Larval fish, Bering Sea,

6 2.3 Introduction 7 Multiple factors during the early life stages of fishes result in variable recruitment

8 success, including prey availability and environmental conditions (Cushing 1982).

9 Variability in the spatial and temporal overlap of predator and prey (match/mismatch

10 hypothesis; Cushing 1969; 1990), as well as differences in prey quality (Sogard and Olla,

11 2000; Litzow et al., 2006), affect fish growth and energy storage, which may directly

12 affect differences in year-class success of many marine fish species, such as walleye

13 pollock, Theragra chalcogramma (Hunt et al., 2011). In addition, cold water

14 temperatures generally delay ontogenetic development of walleye pollock (Blood 2002;

15 Smart et al., in prep) while also lowering routine metabolic demands (Ciannelli et al.,

16 1998). Such trade-offs affect larval fishes’ ability to achieve sufficient size and energy

17 reserves prior to their first winter (Sogard and Olla, 2000; Heintz and Vollenweider,

18 2010).

19 In high latitude systems, winter is a period of low light, cold temperatures, and reduced prey

20 availability, and is therefore a significant source of mortality and determinant of recruitment success of

21 marine fishes (Hurst 2007). Overwintering survival is likely size-dependent because most sources of

22 mortality tend to select against the smallest individuals (Houde 1987; Bailey and Houde, 1989; Paul and

23 Paul, 1999). Hence, depletion of lipid reserves may be preferred over tissue loss in an effort to maintain

57 58

1 overall body size, which is supported by examples of increased survival for larger fish (Schultz and

2 Conover, 1999; Heintz and Vollenweider, 2010). These ideas are encapsulated in the ‘critical size and

3 period hypothesis’, which emphasizes the importance of increased growth in late summer and fall as

4 indicative of winter survival (Beamish and Mahnken, 2001). Lab studies have experimentally corroborated

5 the importance of size on rates of energy depletion (Schultz et al., 1998), which is proportionally greater

6 in smaller fish (Schultz and Conover, 1999; Kooka et al., 2007). Given the shorter growing season in high

7 latitudes, marine fishes may have adapted to grow particularly fast in response to size-selective winter

8 mortality (Conover 1990).

9 Understanding variability during the early life stages of commercially important

10 species is pivotal to fisheries management in predicting year-class success (Megrey et

11 al., 1996) and subsequent recruitment to the fishery. Numerous studies have

12 established empirical links between recruitment and environmental variability (see

13 Beamish and McFarland, 1989), but incorporating the impacts of climate variability on

14 survival into stock assessments requires knowledge of the mechanistic responses to

15 alternate climate states (Hollowed et al., 2009). Fish condition in fall is increasingly

16 recognized as a predictor of overwintering success for walleye pollock in the eastern

17 Bering Sea (Heintz et al., 2013 and provides a potential early metric of recruitment

18 success to age-1.

19 The Oscillating Control Hypothesis (Hunt et al., 2002; revised in Hunt et al., 2011)

20 provides a theoretical framework within which to predict ecosystem responses to warm

21 and cold regimes in the southeastern Bering Sea (SEBS). In warm regimes with early ice

22 retreat, stratified waters maintain production within the pelagic system (Walsh and

23 McRoy, 1986), which was predicted to result in enhanced survival of species such as

58 59

1 walleye pollock (Hunt and Stabeno, 2002; Mueter et al., 2006; Moss et al., 2009).

2 However, recent data indicate that changes in prey composition and abundance during

3 a warm regime may be detrimental to walleye pollock survival. Specifically, larger

4 zooplankton taxa (e.g., large calanoid copepods and euphausiids) were less abundant

5 during recent warm years, which resulted in reduced growth rates and lipid reserves of

6 young-of-year (YOY) walleye pollock and may have increased their predation risk and

7 decreased their overwinter survival (Coyle et al., 2011). In contrast, higher abundances

8 of larger, lipid-rich zooplankton taxa during cold years, combined with lower metabolic

9 demands, allows YOY walleye pollock to acquire greater lipid reserves by fall, resulting in

10 increased overwinter survival (Hunt et al., 2011).

11 Walleye pollock support the largest commercial fishery in the U.S. and are an

12 important component of the eastern Bering Sea pelagic ecosystem. Walleye pollock are

13 major consumers of zooplankton (Aydin et al., 2007) with pronounced changes in prey

14 preference throughout their early life (Ciannelli et al., 2004; WW Strasburger, unpubl.

15 data). The transition to diel vertical migration and nocturnal feeding, which occurs at

16 approximately 50 mm standard length (SL; Brodeur et al., 2000), coincides with

17 increased gape size and shifts to larger prey (i.e., euphausiids), affecting fish condition

18 (Ciannelli et al., 1998). Walleye pollock are also an important forage species for other

19 predators, including arrowtooth flounder, Atheresthes stomias, Pacific cod, Gadus

20 macrocephalus, skates, flathead sole, Hippoglossoides elassodon, Pacific halibut,

21 Hippoglossus stenolepis, sea birds, and marine mammals (Aydin and Mueter, 2007). In

59 60

1 addition, older age classes exhibit strong cannibalism on YOY walleye pollock

2 (Wespestad and Quinn, 1996), especially in warmer climate regimes (Hunt et al., 2011).

3 Despite the important role of walleye pollock in the Bering Sea, and the

4 relationship between age-0 fall condition and overwinter survival, the seasonal

5 energetic patterns from age-0 to age-1 remain poorly understood. This paper examines

6 the link between larval and juvenile strategies for growth and energy storage in walleye

7 pollock. By maximizing growth and transitioning the larval period rapidly, larvae

8 minimize exposure to increased mortality during this stage (i.e., match/mismatch

9 hypothesis; Cushing 1990). In contrast, overwinter survival is higher in fish that are both

10 larger and have increased lipid reserves, indicating that energy allocation during the

11 age-0 juvenile stage will favor lipid storage while also increasing fish size (i.e., critical

12 size and period hypothesis; Beamish and Mahnken, 2001; Heintz and Vollenweider,

13 2010). We hypothesize that energy allocation strategies will differ seasonally among life

14 stages and we tested this by contrasting body compositions of larval, age-0, and age-1

15 fish. The goals of this study were to (1) describe cohort-specific patterns in fish condition

16 for walleye pollock from age-0 to age-1, (2) describe seasonal patterns in energy

17 allocation during larval and juvenile (age-0) development leading to condition estimates

18 prior to their first winter, and (3) develop a conceptual model of energy allocation with

19 implications for overwinter survival and recruitment to age-1.

20 2.3.1 Study region 21 The SEBS is characterized by a broad continental shelf (> 500 km) with an average depth

22 of about 70 m and supports a highly productive ecosystem owing to on-shelf flow of

60 61

1 nutrient-rich waters. Alternating climate states have resulted in periods of both warm

2 and cold conditions in recent years. The most extensive ice cover and coldest water

3 column temperatures since the early 1970s were observed beginning in 2006 and

4 continued through at least the winter of 2010/11.

5 Current trajectories over the shelf are generally northwestward with the Bering

6 Slope Current flowing along the shelf break and Alaska Coastal Current waters following

7 either the 50 m or 100 m isobaths (Stabeno et al., 2001). The onset and location of

8 fronts affect current trajectories (Kachel et al., 2002) and, therefore, transport pathways

9 of larvae (Duffy-Anderson et al., 2006). The main spawning areas for walleye pollock in

10 the SEBS include the waters off Bogoslof Island, north of Unimak Island and along the

11 Alaska Peninsula, and around the Pribilof Islands (Bacheler et al., 2010); isolated

12 spawning of deeper water aggregations also occurs along the shelf break. Larvae are

13 generally advected northward over the shelf with slope-spawned larvae advected onto

14 the shelf via the Bering Slope Current.

15 2.4 Materials and methods

16 2.4.1. Biological sampling 17 Age-0 and age-1 walleye pollock were collected opportunistically from 13 research

18 cruises conducted in the southeastern Bering Sea between May and September 2008-

19 2010 (Table 1; Fig. 1). The geographic coverage varied across cruises. Sampling for age-0

20 fish approximately reflected the actual distributions, while age-1 fish were

21 predominantly sampled from the shelf break (Fig.1). Gear type, mesh size, and sampling

22 depth also varied across cruises to target the life stages occurring at the time of

61 62

1 sampling (Table 1). Vertically integrated oblique bongo tows were made to a maximum

2 depth of 300 m (or to within 10 m of the substratum). Walleye pollock were also

3 sampled from the drogue net of the MOCNESS, which was open during deployment,

4 thereby providing vertically integrated samples to a maximum depth of 100 m (or to

5 within 10 m of the substratum). The ship speed was monitored and adjusted (1.5-2.5

6 knots) throughout all bongo and MOCNESS tows to maintain a wire angle of 45°. Surface

7 (midwater) trawls were conducted above (below) the pycnocline, as determined by

8 water column profiles of temperature and salinity. In addition, near-bottom fishes were

9 sampled with a beam trawl. Our samples are believed to be broadly representative of

10 age-0 and age-1 walleye pollock in the SEBS during the year of sampling.

11 Bongo and MOCNESS sampling occurred 24 hours a day; vertically integrated

12 tows were conducted to a water depth of 100 m, therefore it was assumed that they

13 were not affected by diel vertical migrations by larval and age-0 juvenile walleye pollock.

14 Sampling occurred in daytime only during BASIS surveys in fall of 2008-2010 (Table 1)

15 using either surface or mid-water rope trawls at depths consistent with observed layers

16 from a hydro-acoustics survey (i.e., above or below the pycnocline, respectively). Fish

17 were predominantly observed and collected from surface waters in fall 2008 and from

18 the surface and mid-water in 2009; in 2010, only walleye pollock collected from surface

19 trawls were used in the analyses in order to facilitate interannual comparisons across

20 2008-2010. This study does not address vertical differences in energy content.

21 After retrieval of the gear, all walleye pollock collected were flash frozen (-80 °C)

22 for later chemical analysis at the Alaska Fisheries Science Center, NOAA (National

62 63

1 Oceanic and Atmospheric Administration) in Juneau, Alaska, USA. Pre-flexion larvae

2 were measured to the nearest 0.01 mm length (standard, fork, or total) and post-flexion

3 larvae were measured to the nearest mm fork length (FL). All lengths were converted to

4 SL using established conversions for walleye pollock preserved by freezing (Buchheister

5 and Wilson, 2005). Fish were classified as either age-0 (<110 mm SL) or age-1 (>120 mm

6 SL and <215 mm SL) based on length-frequency distributions constructed for each

7 cruise. Stomach contents of fish >8 mm SL were removed prior to chemical analysis so

8 as not to affect estimates of fish condition.

9 2.4.2. Chemical Analysis 10 2.4.2.1. Energy content

11 Energy density (ED; kJ/g dry mass) was estimated directly using bomb calorimetry or

12 indirectly from estimates of % lipid (see below). Larvae (<30 mm SL) were dried in a

13 drying oven (60°C) and all data are presented on a dry mass basis. Homogenized tissue

14 was pressed into a pellet form and a Parr Instrument 6725 Semimicro Calorimeter with

15 6772 Precision Thermometer and 1109A Oxygen Bomb was used to measure the energy

16 released from combustion of the sample pellets. The minimum pellet weight was set at

17 0.025g of dry material; samples were composited across stations as needed to attain

18 sufficient dry masses for the pre-flexion larvae collected in 2008-2010. Age-0 and age-1

19 fish (>30 mm SL) were dried to a constant weight at 135°C using a LECO

20 Thermogravimetric Analyzer (TGA) 601 or 701 which provided % moisture values. The

21 dried tissue was homogenized and processed using the bomb calorimeter as described

22 above. Moisture analyses were replicated when sufficient tissue masses were available

63 64

1 to ensure the coefficient of variation was less than 1 standard deviation (SD). When

2 sufficient mass was not available, we relied on the coefficient of variation for a

3 reference material (dried adult walleye pollock homogenate) for duplicate estimates of

4 moisture or energy density processed with each batch (n=15 for moisture; n=17 for

5 energy density) of fish.

6 Quality assurance (QA) samples for the bomb calorimeter included (1) replicate

7 tissue samples to evaluate precision and (2) replicate reference material (benzoic acid

8 standard) to evaluate precision and accuracy. Pre-determined limits for variation

9 observed in QA samples were set, where precision estimates from replicate tissue and

10 reference samples must not vary by more than 1.5 SD or 15% coefficient of variation

11 (CV) and reference samples must not vary by more than 15% CV for accuracy. QA

12 samples did not exceed these limits for any batch of samples used in this study.

13 2.4.2.2. Proximate composition

14 For larvae (<30 mm SL), a Sulfo-phospho-vanillin (SPV) colorimetric analysis (Van Handel

15 1985) was performed to determine % lipid composition. Dried material was sonicated in

16 2:1 (by volume) chloroform:methanol solvent in glass centrifuge tubes for 60 minutes.

17 Washes of 0.88% KCl and 1:1 (by volume) methanol:water were performed on the

18 extracts as in the modified Folch extraction method (Vollenweider et al., 2011).

19 Resulting chloroform extracts were evaporated in a LabConco RapidVap for 30 minutes

20 at 40°C and 250 mbar until reduced to approximately 1 ml in volume. Extracts were

21 evaporated to dryness in 12 mm test tubes on a heating block at 75°C and then allowed

22 to cool. Concentrated sulfuric acid was added to the tubes prior to incubation at 100°C

64 65

1 for 10 minutes with subsequent cooling. SPV reagent (1.2 mg/ml vanillin in 80%

2 phosphoric acid) was added to each tube and allowed to develop for 10 minutes.

3 Absorption was measured on an Agilent 8453 Spectrophotometer at 490nm and

4 extrapolated from species-specific calibration curves determined prior to analysis; %

5 lipid is presented on a dry mass basis. For age-0 and age-1 fish (>30 mm SL), proximate

6 composition analysis was performed as previously described (Vollenweider et al., 2011),

7 with lipid extractions utilizing a Dionex ASE (Accelerated Solvent Extractor) 200 and a

8 modified Folch extraction procedure using a 2:1 (by volume) chloroform:methanol

9 solvent mixture. Analysis was conducted using wet material and estimates of %

10 moisture (see above) were used to convert observations to a dry mass basis presented

11 here.

12 QA samples for the SPV samples included two blank runs to estimate background

13 absorption, two method blank samples containing all analysis reagents but no lipid

14 extract to evaluate contamination and reagent absorption, and two reference samples

15 (adult walleye pollock homogenate) to examine precision and accuracy for each batch of

16 15 samples. Mean background absorbance was subtracted from sample absorbance

17 values. Method blank samples had to be < 10 mg of lipid and walleye pollock reference

18 samples had to vary by < 1 SD and be accurate within 15% of the established lipid value.

19 ASE samples used similar QA criteria except that the method blank samples were

20 allowed to be as high as 0.1 mg of lipid (due to much higher analyzed lipid masses).

21 To compare energy densities of walleye pollock from age-0 to age-1 for the

22 cohort analysis, a linear regression was used to predict energy density estimates from %

65 66

1 lipid values for age-1 fish collected during two cruises (BASIS 2009 and MACE 2009;

2 Table 1). All age-1 fish processed for both energy density and % lipid (n=83) were used

3 to develop a regression relationship, ED = 20.1 + 0.76*% lipid (r2 = 0.79), that was used

4 to predict energy density. The size of the fish used in the regression ranged from 109 -

5 194 mm SL. Direct estimates of energy density from the bomb calorimeter were used

6 when available.

7 2.4.3. Statistical analysis 8 Cohort-specific patterns in energy density from age-0 to age-1 were examined to

9 determine the extent of interannual variation in the seasonal pattern of fish condition

10 using two complete cohorts. The 2008 cohort was sampled as age-0 fish in summer

11 (July) and fall (September) 2008, and as age-1 fish in summer (July) and fall (September)

12 2009. The 2009 cohort was sampled as age-0 fish in summer (June) and fall (September)

13 2009, and as age-1 fish in summer (July) 2010. Fish condition was compared between

14 cohorts within each season (age-0 and age-1 in summer, age-0 in fall) using separate

15 one-way ANOVAs (analysis of variance).

16 Seasonal patterns in energy allocation during larval and early juvenile

17 development (age-0) were analyzed using generalized additive mixed models (GAMM)

18 to identify the seasonal timing and size at which walleye pollock shift energy allocation

19 strategies from growth to lipid storage. These models do not specify a fixed functional

20 form, but rather quantify the relationship between a set of predictors and available

21 estimates of fish condition through non-parametric smooth functions of the predictor

66 67

1 variables. The optimum amount of smoothing was chosen by generalized cross-

2 validation as implemented in the R package ‘mgcv’ (Wood 2006).

3 The sampling area differed among cruises as did the number of fish processed

4 per location, therefore modeling approaches that accounted for spatial patterns and/or

5 included a random station effect were compared using Akaike’s Information Criterion

6 (AIC) (Akaike 1973; Burnham and Anderson, 2002). Fish for which measurements of

7 standard length were not available were removed from models of energy density or %

8 lipid (n=5 and 6, respectively). Based on residual diagnostics, estimates of energy

9 density and % lipid identified as outliers (n=5 and 8, respectively) were removed from

10 further analyses resulting in a total of 247 estimates of energy density and 271

11 estimates of % lipid.

12 Variability in energy density and % lipid over the first summer were modeled as a

13 function of sampling date (Day of year, DOY) to estimate seasonal trends in energy

14 density (% lipid) and as a function of standard length (SL) to estimate changes in energy

15 allocation with fish size. The models also included a spatial smooth term to account for

16 differences in the mean across stations within years to address autocorrelation, as well

17 as a year term to account for possible differences in the average energy density (% lipid)

18 between years. The full model also included station as a random effect to account for

19 variability among stations (ai), in addition to within-station residual variability (εik). The

20 best model for energy density included both a spatial smooth term and a random

21 station effect to account for autocorrelation (see below) while the random station effect

22 was not significant for % lipid and therefore dropped from the best model (Table 2).

67 68

1 Energy density = + f1(DOY) + f2(SL) + f3(latitude, longitude)k + Yk + ai + ik (Eq. 1)

2 2 ai ~N(0, a)

2 3 ik~N(0,  e )  4 where f1, f2, and f3 are smooth functions, Yk is the year-specific intercept for year k, ai is  5 the station-specific, random intercept for station i, and εik is the residual. The random

6 effects ai and residuals εik are assumed to be independent and normally distributed.

7 Residuals and random effects from the models were examined for normality,

8 homoscedasticity, and independence by plotting them against all relevant covariates

9 and by examining spatial patterns in the random station effects by year.

10 Patterns in fish length (mm SL) were also modeled as a function of sampling date

11 (DOY) to estimate seasonal trends in fish growth; the model included a spatial smooth

12 term to account for differences in the mean across stations within years to address

13 autocorrelation and a year term to account for possible differences in the average

14 length between years. The best model also included station as a random effect (Table

15 2).

16 SL = + f1(DOY) + f2(latitude, longitude)k + Yk + ai + ik (Eq. 2)

17 ai ~N(0, )

18 ik~N(0, )

19 where f1 and f2 are smooth functions, Yk is the year-specific intercept for year k, ai is the

20 station-specific, random intercept for station i, and εik is the residual.

68 69

1 2.5. Results

2 2.5.1. Biological sampling 3 A total of 1501 walleye pollock larvae, age-0, and age-1 fish collected from 13 cruises

4 over the 3-year sampling period were measured for standard length with 341 resulting

5 estimates of energy density (kJ/g) from bomb calorimetry (n=257 age-0 including 13

6 composite samples, n=84 age-1) and 423 estimates of % lipid (n=285 age-0 including 41

7 composite samples, n=135 age-1) (Table 1). The overall mean energy density (± SE) of

8 age-0 and age-1 walleye pollock was 22.43 ± 0.11 and 23.67 ± 0.12 (kJ/g dry mass),

9 respectively, and overall mean % lipid (± SE) for age-0 and age-1 fish was 10.58 ± 0.40

10 and 20.53 ± 0.49 (on dry mass basis), respectively.

11 2.5.2. Cohort-specific patterns from age-0 to age-1

12 Age-0 fish from the 2008 and 2009 cohorts had relatively low mean energy densities in

13 early summer with no significant difference between cohorts (1-way ANOVA: F(1,9)=1.07,

14 P=0.33) (Table 1, Fig. 2, Fig. 3a). Corresponding low values for % lipid in 2008 and 2009

15 (Table 1, Fig. 4a) indicate energy allocation to growth and/or development rather than

16 energy storage during the early larval period. There was a short period from early-July

17 to mid-September (DOY: 184-261) during which energy density rapidly increased to

18 higher values in fall 2008 and 2009 (Table 1, Fig. 2). Energy density during NOAA/BASIS

19 2008 was significantly higher than NOAA/EcoFOCI 2008 (1-way ANOVA: F(2,120)=7.04,

20 P<0.01), with NOAA/BASIS 2009 having an intermediate energy density.

21 Energy densities are presumed to decrease over winter when water column

22 temperatures, prey availability, and feeding rates decrease. By the fall of the second

23 year, age-1 juveniles of both the 2008 and 2009 cohorts had regained energy densities

69 70

1 similar to those of age-0 fish the previous fall (Table 1, Fig. 2). However, the 2009 cohort

2 had significantly greater energy density at age-1 than the 2008 cohort (1-way ANOVA:

3 F(2, 85)=13.68, P<0.001; Fig. 2).

4 2.5.3. Seasonal patterns in energy allocation of age-0 fish 5 Energy density and % lipid both varied non-linearly with fish length and season (Fig. 3, a and b; Fig. 4, a

6 and b, respectively) with significant differences between years. Energy density was highest in

7 2008, lowest in 2009, and intermediate in 2010 (Fig. 3c) while % lipid showed a decrease from

8 2008 through 2010 (Fig. 4c). The effect of size is difficult to separate from the seasonal pattern

9 because SL is strongly correlated with sampling date for both energy density (r=0.56) and % lipid

10 (r=0.91), although largely uncorrelated for fish larger than 30 mm SL (energy density r=0.086; %

11 lipid r=-0.35). Therefore, for fish >30 mm SL, we can statistically separate the apparent effects of

12 size and seasonal pattern on fish condition. Residual diagnostics for all models showed no

13 unusual trends and no evidence of remaining spatial autocorrelation.

14 Few estimates of energy density for preflexion larvae (<30 mm SL) were possible

15 because samples had to be composited to acquire sufficient dry mass for analytical

16 processing. Therefore, patterns in energy density for preflexion larvae are uncertain

17 (i.e., Fig. 3b). Preflexion larvae had negative energy density anomalies indicating below-

18 average condition; a period of rapidly increasing energy density occurred between

19 approximately 20-60 mm SL until reaching an asymptote at approximately 70 mm SL

20 (Fig. 3a). The seasonal trend in energy density shows little effect of sampling date,

21 indicating that changes in energy density are primarily driven by changes in fish size as

22 opposed to seasonal timing. Energy density reaches a minimum in early fall

23 (approximately DOY 250; September 6, Fig. 3b).

70 71

1 Percent lipid in fish <30 mm SL showed a decrease with increasing length that may

2 correspond with the transition to exogenous feeding after hatching, which occurs at 3-4 mm SL.

3 In fish >40 mm SL, % lipid increases linearly with increasing size (Fig. 4a). The seasonal trend in %

4 lipid varied non-linearly, although variability in these estimates was high (Fig. 4b). The effect of

5 fish size on patterns of % lipid was greater than the seasonal effect, as indicated by the range in

6 % lipid anomalies between Figures 4a and 4b.

7 2.5.4. Seasonal patterns in length of age-0 fish 8 Walleye pollock length increased slowly during the pre-flexion stage (<30 mm SL), but rapidly

9 after DOY 200 (July 19). Fish lengths reached an asymptote between approximately DOY 235

10 (August 23) and DOY 250 (September 6) before increasing again over the remaining sampling

11 period (Fig. 5). Fish lengths were more variable in fall than in spring and early summer. The

12 asymptote in fish length corresponds to the minimum observed in energy density (Fig. 3b).

13 2.6 Discussion 14 This study provides estimates of fish condition for larval, age-0, and age-1 walleye pollock and

15 suggests how YOY fish allocate resources to growth and energy storage. Patterns in energy

16 allocation differed among larval and juvenile fish. Constraints of the match/mismatch

17 hypothesis (Cushing 1969; 1990), which would favor energy allocation to growth and

18 development in order to escape size-dependent predation, appear limited to larval

19 development. In contrast, age-0 juvenile fish adopt an allocation strategy consistent with the

20 constraints suggested by the critical period hypothesis (Beamish and Mahnken, 2001). Our

21 results suggest a period of rapid growth associated with an increase in energy allocation to

22 storage during late summer. Presumably, differences in prey availability and quality (Coyle et al.,

23 2011) during this period are reflected in differences in energy density (Heintz et al., 2013. This

24 allocation strategy has potentially important consequences for overwinter survival because fish

71 72

1 condition at the end of summer directly relates to observed differences in year-class strength

2 between alternating climate states in the SEBS (Hunt et al., 2011; Heintz et al., 2013). One

3 possible explanation for the apparent link between energy density and recruitment is size-

4 dependent mortality, which contributes to differential overwinter survival (Schultz et al., 1998;

5 Heintz and Vollenweider, 2010). Therefore, we propose that late summer represents a critical

6 period for energy storage in YOY walleye pollock and their subsequent recruitment success.

7 In the recent cold years of 2006-2010, the zooplankton community over the

8 Bering Sea shelf has been dominated by large copepods (i.e., Calanus marshallae) and

9 euphausiids (i.e., Thysanoessa raschii), providing walleye pollock with a lipid-rich diet

10 and sufficient energy reserves for increased overwinter survival. Under warmer

11 conditions (2002-2005), smaller zooplankton taxa were dominant (i.e., Pseudocalanus

12 spp., Acartia spp., Coyle et al., 2011) and the lack of larger prey appeared to have

13 limited growth and energy storage, leading to poor condition and reduced year-class

14 recruitment. The limited availability of large zooplankton coincided with increased rates

15 of cannibalism by older age-classes of walleye pollock, further reducing age-0 survival in

16 warm years (Coyle et al., 2011). Hence, prey quality may be as important as the thermal

17 regime for determining overwinter survival (Hurst 2007), although prey availability and

18 prey quality are closely linked to temperature conditions (Coyle et al., 2011).

19 The 2008 and 2009 cohorts showed similar seasonal patterns of energy

20 allocation for age-0 fish, likely reflecting similar oceanographic conditions and prey

21 resources. The differences observed for age-0 fish during early fall likely reflect the

22 difference in sampling time between cruises and highlight early fall as a period of rapidly

23 increasing energy density in walleye pollock. The 2009 cohort had higher energy

72 73

1 densities by age-1, which could be due to differential overwinter survival, reduced

2 winter energy loss for the 2009 cohort leading to less of an energy deficit in spring 2010,

3 and/or differences in prey availability for age-1 fish during their second summer. The

4 2009 cohort likely experienced greater prey availability and less intra-species

5 competition as the 2009 year-class estimate remains well below the 2008 estimate

6 (Ianelli et al., 2011).

7 Low lipid and energy density levels in spring and early summer indicate larvae

8 preferentially allocate energy to development (i.e., protein synthesis) with little increase in

9 overall fish growth (i.e., tissue accretion) during the preflexion stage (<30 mm SL) until the

10 transformation to the juvenile form. The decrease in lipid content with size as larvae increase

11 from ~5 – 15 mm (Fig. 4a) most likely reflects the transition to exogenous feeding with

12 decreasing energy stores as larvae learn to capture prey. This study was conducted during three

13 cold years in the SEBS (Stabeno et al., in press), therefore delayed development times likely

14 resulted in smaller fish sizes relative to warmer years (Smart et al., in prep). Although the

15 average lipid content of YOY walleye pollock decreased significantly from 2008 to 2010 (Fig. 4c),

16 we did not see a consistent decrease in average energy density over the same time period.

17 However, we do not have estimates of % lipid for all sampling periods (spring, summer, and fall)

18 in all years of the study, therefore we cannot fully address interannual differences in % lipid.

19 During summer, walleye pollock appear to be growing in size while also increasing lipid

20 stores, although patterns in energy allocation are poorly defined due to lack of samples during

21 this period. That said, late July – August seems to be a period when energetic demands are

22 highest. The length at transformation from larval to juvenile form occurs at 25-40 mm SL (Brown

23 et al., 2001) and marks a threshold after which lipid acquisition rates increased, leading to

24 higher energy density in fall as energy was allocated to storage for overwinter survival. Our

73 74

1 samples fall on either side of this size range, supporting the inference that fish below this length

2 range are allocating energy to development and fish above this length range are allocating

3 energy to storage.

4 Early juvenile (age-0) walleye pollock reach an asymptotic length concurrent with a

5 minimum energy density in early fall which may indicate a shift in prey preferences with

6 increasing gape size (i.e., switch to euphausiids) and associated foraging capability. A similar

7 energy minimum was observed in walleye pollock (46 mm SL) near the Pribilof Islands and was

8 attributed to changes in feeding habits (Ciannelli et al., 2002). The energy density of walleye

9 pollock near the Pribilof Islands during 1994-1996 and 1999 showed an asymptote when fish

10 reached 80 mm SL (Ciannelli et al., 2002); the asymptotic size observed in fall 2008-2010 was

11 approximately 60 mm SL. Cold water temperatures observed during the study period likely

12 delayed and/or limited fish growth and development (Smart et al., in prep).

13 While our study focused on seasonal patterns in energy allocation, spatial

14 patterns in the distribution of larvae relative to prey may be equally important in

15 determining recruitment success. Age-0 walleye pollock in the Gulf of Alaska during fall

16 experience spatially variable habitat conditions for growth due to differences in water

17 temperature and prey (Mazur et al., 2007). Because of such differences, both the

18 horizontal and vertical distribution of larvae affects their growth and condition.

19 Hydrography affects larval drift trajectories (Duffy-Anderson et al., 2006) as well as prey

20 resources, which can be concentrated at frontal structures (Ciannelli et al., 2004). Once

21 larvae are capable of diel vertical migration, the vertical position in the water column

22 (i.e., above or below the pycnocline) affects temperature-dependent metabolic rates, as

23 well as trade-offs in feeding rates versus predation risk (Sogard and Olla, 1996). Juvenile

74 75

1 walleye pollock are capable of selecting habitat based on temperature, prey availability,

2 and predator abundance (Kooka et al., 2007). Consequently, we plan to incorporate

3 both local-scale environmental conditions and estimates of prey availability into a

4 bioenergetics model to quantify fine-scale spatial variability in fish condition and growth

5 potential and to support the development of predictive models for recruitment success

6 of walleye pollock in the SEBS.

7 The current study provides a conceptual model of how energy allocation

8 strategies shift in walleye pollock during the larval and juvenile phases. This shift

9 represents adaptations to survival constraints associated with distinct ontogenetic

10 stages. Our results suggest that after a period of enhanced somatic growth until

11 metamorphosis, early juvenile (age-0) fish favor energy storage in late summer/fall. This

12 supports the critical period concept and the hypothesis that overwinter survival is

13 dependent on sufficient storage in the previous growing season and may be an

14 important determinant of recruitment success.

16 2.7 Acknowledgements 17 We thank the officers and crew NOAA’s Miller Freeman and Oscar Dyson, USCG’s Healy,

18 and R/Vs Knorr and Thompson. Funding was provided through the North Pacific

19 Research Board (NPRB) Bering Sea Integrated Ecosystem Research Program (BSIERP) as

20 well as NOAA’s EcoFOCI, MACE, and BASIS programs. We thank three anonymous

21 reviewers for providing helpful comments that improved the manuscript. This research

75 76

1 is contribution EcoFOCI-XXXX to NOAA's Fisheries-Oceanography Coordinated

2 Investigations, NPRB XXXX, and BEST-BSIERP XXXX.

4 2.8 References 5 Akaike H (1973) Information theory as an extension of the maximum likelihood

6 principle. In BN Petrov and F Csaki (eds) Second International Symposium on

7 Information Theory. Akademiai Kiado, Budapest.

9 Aydin K, Mueter FJ (2007) The Bering Sea - A dynamic food web perspective. Deep Sea

10 Res II 54: 2501-2525.

12 Aydin K, Gaichas S, Ortiz I, Kinzey D, Friday N (2007) A comparison of the Bering Sea, Gulf of Alaska, and

13 Aleutian Islands large marine ecosystems through food web modeling. NOAA Tech Memo NMFS-AFSC-

14 178. 298 p.

16 Bacheler NM, Ciannelli L, Bailey KM, Duffy-Anderson JT (2010) Spatial and temporal

17 patterns of walleye pollock (Theragra chalcogramma) spawning in the eastern Bering

18 Sea inferred from egg and larval distributions. Fish Oceanogr 19(2): 107-120.

20 Bailey KM, Houde ED (1989) Predation on early developmental stages of marine fishes

21 and the recruitment problem. Adv Mar Biol 25: 1–83.

76 77

1 Beamish RJ, Mahnken C (2001) A critical size and period hypothesis to explain natural

2 regulation of salmon abundance and the linkage to climate and climate change. Prog

3 Oceanogr 49: 423-437.

5 Beamish RJ, McFarlane GA (1989) Effects of ocean variability on recruitment and an

6 evaluation of parameters used in stock assessment models. Can Spec Pub, Fish Aquat Sci

7 108, 379 p.

9 Blood DM (2002) Low-temperature incubation of walleye pollock (Theragra

10 chalcogramma) eggs from the southeast Bering Sea shelf and Shelikof Strait, Gulf of

11 Alaska. Deep-Sea Res II 49: 6095–6108.

13 Brodeur RD, Wilson MT, Ciannelli L (2000) Spatial and temporal variability in

14 feeding and condition of age-0 walleye pollock (Theragra chalcogramma) in frontal

15 regions of the Bering Sea. ICES J Mar Sci 57: 256–264.

17 Brown AL, Busby MS, Mier KL (2001) Walleye pollock Theragra chalcogramma during

18 transformation from the larval to juvenile stage: otolith and osteological development.

19 Mar Biol 139: 845-851.

77 78

1 Buchheister A, Wilson MT (2005) Shrinkage correction and length conversion equations

2 for Theragra chalcogramma, Mallotus villosus and Thaleichthys pacificus. J Fish Biol 67:

3 541-548.

5 Burnham KP, Anderson DR (2002) Model selection and multi-model inference: a

6 practical information-theoretic approach, Springer-Verlag, New York, NY. 488 p.

8 Ciannelli L, Brodeur RD, Buckley TW (1998) Development and application of a

9 bioenergetics model for juvenile walleye pollock. J Fish Biol 52: 879-898.

11 Ciannelli L, Paul AJ, Brodeur RD (2002) Regional, interannual and size-related variation

12 of age 0 year walleye pollock whole body energy content around the Pribilof Islands,

13 Bering Sea. J Fish Biol 60: 1267-1279.

15 Ciannelli L, Brodeur RD, Napp JM (2004) Foraging impact on zooplankton by age-0

16 walleye pollock (Theragra chalcogramma) around a front in the southeast Bering Sea.

17 Mar Biol 144: 515-526.

19 Conover DO (1990) The relation between capacity for growth and length of growing

20 season: evidence for and implications of countergradient variation. Trans Am Fish Soc

21 119 (3): 416–430.

78 79

1 Coyle KO, Eisner LB, Mueter FJ, Pinchuk AI, Janout MA, Cieciel KD, Farley EV, Andrews

2 AG (2011) Climate change in the southeastern Bering Sea: impacts on pollock stocks and

3 implications for the Oscillating Control Hypothesis. Fish Oceanogr 20(2): 139-156.

5 Cushing, DH (1969) The regularity of the spawning season of some fishes. J Cons Int

6 Explor Mer 33: 81-97.

8 Cushing, DH (1982) Climate and fisheries. Academic Press: London. ISBN 0-12-199720-0.

9 373 p.

11 Cushing, DH (1990) Plankton production and year-class strength in fish populations: an

12 update of the match/mismatch hypothesis. Adv Mar Biol 26: 250-293.

14 Duffy-Anderson JT, Busby MS, Mier KL, Deliyanides CM, Stabeno PJ (2006) Spatial and

15 temporal patterns in summer ichthyoplankton assemblages on the eastern Bering Sea

16 shelf 1996-2000. Fish Oceanogr 15: 80-94.

18 Heintz RA, Vollenweider JJ (2010) Influence of size on the sources of energy consumed

19 by overwintering walleye pollock (Theragra chalcogramma). J Exp Mar Biol Ecol. 393(1-

20 2): 43-50.

79 80

1 Heintz, R.A., Siddon, E.C., Farley, E.V., and Napp, J.M. 2013. Correlation between

2 recruitment and fall condition of age-0 pollock (Theragra chalcogramma) from the

3 eastern Bering Sea under varying climate conditions. Deep-Sea Res. II 94, 150-156.

5 Hollowed AB, Bond NA, Wilderbuer TK, Stockhausen WT, A’mar ZT, Beamish RJ,

6 Overland JE, Schirripa MJ (2009) A framework for modelling fish and shellfish responses

7 to future climate change. ICES J Mar Sci 66 (7): 1584-1594.

9 Houde ED (1987) Fish early life dynamics and recruitment variability. Amer Fish Soc

10 Symp 2: 17–29.

12 Hunt GL Jr, Stabeno PJ (2002) Climate change and the control of energy flow in the

13 southeastern Bering Sea. Prog Oceanogr 55: 5-22.

15 Hunt GL Jr, Stabeno PJ, Walters G, Sinclair E, Brodeur RD, Napp JM, Bond NA (2002)

16 Climate change and control of the southeastern Bering Sea pelagic ecosystem. Deep-Sea

17 Res Part II 49: 5821-5853.

19 Hunt GL Jr, Coyle KO, Eisner L, Farley EV, Heintz R, Mueter FJ, Napp JM, Overland JE,

20 Ressler PH, Salo S, Stabeno PJ (2011) Climate impacts on eastern Bering Sea foodwebs: A

21 synthesis of new data and an assessment of the Oscillating Control Hypothesis. ICES J

22 Mar Sci, doi:10.1093/icesjms/fsr036.

80 81

2 Hurst TP (2007) Causes and consequences of winter mortality in fishes. J Fish Biol 71:

3 315-345.

5 Ianelli JN, Honkalehto T, Barbeaux S, Kotwicki S, Aydin K, Williamson N (2011)

6 Assessment of the walleye pollock stock in the Eastern Bering Sea. In: Stock assessment

7 and fishery evaluation report for the groundfish resources of the Bering Sea/Aleutian

8 Islands regions. North Pacific Fishery Management Council, 605 W. 4th Ave., Suite 306,

9 Anchorage, AK 99501.

11 Kachel NB, Hunt GL Jr, Salo SA, Schumacher JD, Stabeno PJ, Whitledge TE (2002)

12 Characteristics and variability of the inner front of the southeastern Bering Sea. Deep-

13 Sea Res II 49: 5889-5909.

15 Kooka K, Yamamura O, Nishimura A, Hamatsu T, Yanagimoto T (2007) Optimum

16 temperature for growth of juvenile walleye pollock Theragra chalcogramma. J Exp Mar

17 Biol Ecol 347: 69-76.

19 Litzow MA, Bailey KM, Prahl FG, Heintz R (2006) Climate regime shifts and

20 reorganization of fish communities: the essential fatty acid limitation hypothesis. Mar

21 Ecol Prog Ser 315: 1–11.

81 82

1 Mazur MM, Wilson MT, Dougherty AB, Buchheister A, Beauchamp DA (2007)

2 Temperature and prey quality effects on growth of juvenile walleye pollock Theragra

3 chalcogramma (Pallas): a spatially explicit bioenergetics approach. J Fish Biol 70: 816-

4 836.

6 Megrey BA, Hollowed AB, Hare SR, Macklin SA, Stabeno PJ (1996) Contribution of FOCI

7 research to forecast of year class strength of walleye pollock in the Shelikof Strait,

8 Alaska. Fish Oceanogr 5 (S1): 189-203.

10 Moss JH, Farley EV, Feldman AM, Ianelli JN (2009) Spatial distribution, energetic status,

11 and food habits of eastern Bering Sea age-0 walleye pollock. Trans Am Fish Soc 138: 497-

12 505.

14 Mueter FJ, Ladd C, Palmer MC, Norcross BL (2006) Bottom-up and top-down controls of

15 walleye pollock (Theragra chalcogramma) on the Eastern Bering Sea shelf. Prog

16 Oceanogr 68:152–183.

18 Paul AJ, Paul JM (1999) Interannual and regional variations in body length, weight and

19 energy content of age-0 Pacific herring from Prince William Sound, Alaska. J Fish Biol 54:

20 996-1001.

82 83

1 Schultz ET, Conover DO, Ehtisham A (1998) The dead of winter: size-dependent variation

2 and genetic differences in seasonal mortality among Atlantic silverside (Atherinidae:

3 Menidia menidia) from different latitudes. Can J Fish Aquat Sci 55: 1149-1157.

5 Schultz ET, Conover DO (1999) The allometry of energy reserve depletion: test of a

6 mechanism for size-dependent winter mortality.

8 Smart TI, Siddon EC, Duffy-Anderson JT (in prep) Vertical distributions of the early life

9 stages of a common gadid in the eastern Bering Sea (Theragra chalcogramma, walleye

10 pollock). Limnol Oceanogr.

12 Sogard SM, Olla BL (1996) Food deprivation affects vertical distribution and activity of a

13 marine fish in a thermal gradient: potential energy-conserving mechanisms. Mar Ecol

14 Prog Ser 133: 43-55.

16 Sogard SM, Olla BL (2000) Endurance of simulated winter conditions by age-0 walleye

17 pollock: effects of body size, water temperature and energy storage. J Fish Biol 56:1-21.

19 Stabeno PJ, Bond NA, Kachel NB, Salo SA, Schumacher JD (2001) On the temporal

20 variability of the physical environment over the south-eastern Bering Sea. Fish Oceanogr

21 10(1): 81-98.

83 84

1 Stabeno PJ, Moore S, Napp J, Sigler M, Zerbini A (in press) Comparison of warm and cold

2 years on the southeastern Bering Sea shelf. Deep-Sea Res II.

4 Van Handel E (1985) Rapid determination of total lipids in mosquitoes. J Am Mosq

5 Control Assoc 1 (3): 302-304.

7 Vollenweider JJ, Heintz R, Schaufler L, Bradshaw R (2011) Seasonal cycles in whole-body

8 proximate composition and energy content of forage fish vary with water depth. Mar

9 Biol 158: 413-27.

11 Walsh JJ, McRoy CP (1986) Ecosystem analysis in the southeastern Bering Sea. Cont Shelf Res 5: 259–288.

13 Wespestad VG, Quinn II TJ (1996) Importance of cannibalism in the population dynamics

14 of walleye pollock, Theragra chalcogramma. NOAA Tech Rep NMFS, vol. 126: 212–217.

16 Wood SN (2006) Generalized Additive Models: An introduction with R. Chapman &

17 Hall/CRC, Boca Raton, FL. 422 p.

19 2.9 Table legends 20 Table 1. Cruise name, year, dates of sampling, and gear used to collect walleye pollock

21 (Theragra chalcogramma). The mean ( SE, n) energy density, % lipid, and standard

22 length (mm) are shown for each cruise by age class. Energy density values in italics were

84 85

1 predicted from % lipid values based on the regression relationship (ED = 20.1 + 0.76*%

2 lipid).

4 Table 2. Summary of generalized additive mixed model (GAMM) fits for energy density,

5 % lipid, and standard length showing terms, coefficient estimates, standard error (S.E.)

6 for fixed coefficients, degrees of freedom (d.f.; number of parameters for each term in

7 the model, estimated for smooth terms), and P-values. P-values for parametric terms

8 (intercept, year coefficients) based on t-test of the null hypothesis that the coefficient is

9 equal to zero; for smooth terms (fi) based on an approximate F-test (Wood 2006); for

10 random effects term (σa) based on likelihood ratio test. The intercept () corresponds to

11 the 2008 means and the subsequent year effects correspond to the difference between

12 that year's mean and the intercept.

13 2.10 Figure legends 14 Figure 1. Map of the southeastern Bering Sea showing the location of sample collections

15 by age-class and year. Distribution of collections approximates the actual distribution of

16 age-0 and age-1 fish at the time of sampling. Depth contours are shown for the 50m,

17 100m, and 200m isobaths.

19 Figure 2. Plot of energy density (kJ/g dry mass) for the 2008 and 2009 cohorts of walleye

20 pollock (Theragra chalcogramma). Errors bars are plotted as ±1 standard deviation (SD)

21 to show the variability in energy density estimates for each sampling interval. Different

85 86

1 letters indicate significant differences in energy density within season. Note x-axis is

2 consecutive days across the first two years of walleye pollock life.

4 Figure 3. Results from generalized additive mixed model (GAMM) regression analyses

5 showing the relationship between energy density (kJ/g dry mass) and (a) standard

6 length (mm), (b) Day of year, and (c) year for age-0 walleye pollock (Theragra

7 chalcogramma). Dashed lines denote 95% confidence intervals.

9 Figure 4. Results from generalized additive mixed model (GAMM) regression analyses

10 showing the relationship between % lipid (on a dry mass basis) and (a) standard length

11 (mm), (b) Day of year, and (c) year for age-0 walleye pollock (Theragra chalcogramma).

12 Dashed lines denote 95% confidence intervals.

14 Figure 5. Results from generalized additive mixed model (GAMM) regression analyses

15 showing the relationship between standard length (mm) and Day of year for age-0

16 walleye pollock (Theragra chalcogramma). Dashed lines denote 95% confidence

17 intervals.

86 87

1 Table 1.

Year Dates Gear Mesh size (µm) Energy density (kJ/g) % lipid Standard Length (mm)

Age-0 Age-1 Age-0 Age-1 Age-0 Age-1

2008 May 12 – 21 Bongo 505 8.51 5.67

(0.94, 13) (0.07, 274)

2008 July 3 - 17 MOCNESS 505 17.99 11.36 10.61

(0.68, 6) (0.45, 18) (0.14, 168)

2008 September 9 - Beam Trawl 7 mm; 3 mm codend liner 22.36 66.82

20 (0.18, 34) (1.34, 33)

2008 September 7 - Surface 505 23.35 21.49 62.7

30 (0.22, 37) (0.82, 31) (1.25, 102)

2008 June 2 - July 31 Methot Trawl 2x3 mm; 1 mm codend liner 23.13 19.44 132.46

(0.12, 49) (0.52, 49) (1.67, 49)

2009 June 14 - July MOCNESS 505 16.95 8.34 8.79

12 (0.68, 4) (0.26, 89) (0.14, 251)

87 88

2009 September 2 - Midwater 22.81 23.51 17.98 18.8 67.1 155.04

30 (0.14, 50) (0.34, 18) (0.84, 46) (1.66, 18) (1.08, 100) (6.23, 19)

2009 June 9 - August Methot Trawl 2x3 mm; 1 mm codend liner 22.99 17.97 138.45

7 (0.18, 34) (1.0, 34) (3.46, 34)

2010 May 6 - 18 Bongo 505 2.15 5.67

(0.30, 22) (0.05, 164)

2010 June 16 - July MOCNESS 505 15.39 6.0 8.88

14 (0.88, 3) (0.23, 66 (0.14, 135)

2010 September 2 - Beam Trawl 7 mm; 3 mm codend liner 22.43 62.38

15 (0.25, 34) (1.9, 34)

2010 August 16 - Surface 22.64 59.8

September 26 (0.11, 89) (0.76, 104)

2010 June 5 - August Methot Trawl 2x3 mm; 1 mm codend liner 24.41 25.58 148.25

7 (0.17, 34) (0.72, 34) (2.98, 34)

TOTALS 257 84 285 135 1365 136

88 89

1 Table 2.

Energy Density

Term Estimate S.E. d.f. P-value

Intercept () 22.95 0.19 1 <0.001

Y2009 -1.05 0.29 1 <0.001

Y2010 -0.51 0.24 1 0.036

f1(SL) 4.1 <0.001

f2(JD) 2.18 0.21

f3(lat, long)2008 3.1 <0.001

f3(lat, long)2009 2 0.005

f3(lat, long)2010 2 0.18

(ai) 0.46 1 <0.0001

(ik) 0.71 1

% Lipid

Intercept () 14.99 0.65 1 <0.001

Y2009 -4.48 0.81 1 <0.001

Y2010 -7.05 0.87 1 <0.001

f1(SL) 4.18 <0.001

f2(JD) 3.52 <0.001

f3(lat, long)2008 2 <0.001

f3(lat, long)2009 9.53 <0.001

89 90

f3(lat, long)2010 4.36 0.21

(ik) 2.07 1

Standard Length

Intercept () 20.56 1.22 1 <0.001

Y2009 6.61 1.88 1 <0.001

Y2010 2.07 1.66 1 0.21

f1(JD) 6.38 <0.001

f2(lat, long)2008 2 <0.001

f2(lat, long)2009 3.22 <0.001

f2(lat, long)2010 2 0.46

(ai) 5.68 1 <0.0001

(ik) 3.79 1

90 91

2 Figure 1.

91 92

3 Figure 2.

92 93

2 Figure 3a.

93 94

2 Figure 3b.

94 95

2 Figure 3c.

95 96

2 Figure 4a.

96 97

2 Figure 4b.

97 98

2 Figure 4c.

98 99

2 Figure 5.

99 100

2 Chapter 3 - Correlation between recruitment and fall condition 3 of age-0 pollock (Theragra chalcogramma) from the eastern 4 Bering Sea under varying climate conditions 5

6 Ron A. Heintza*, Elizabeth C. Siddonb, Edward V. Farley Jra. and Jeffrey M. Nappc

7 a. NOAA, National Marine Fisheries Service, Alaska Fisheries Science Center, Auke 8 Bay Laboratories, 17109 Pt. Lena Loop Rd., Juneau, AK 99801 USA 9 b. University of Alaska Fairbanks, School of Fisheries and Ocean Sciences, 17101 Pt. 10 Lena Loop Rd. Juneau, AK 99801 USA 11 c. NOAA, National Marine Fisheries Service, Alaska Fisheries Science Center, 7600 12 Sand Point Way NE, Seattle, WA 98115 USA 13 14 Published in Deep Sea Research II (2013)

16 3.1 Abstract 17 Fishery managers require an understanding of how climate influences recruitment if

18 they are to separate the effects of fishing and climate on production. The southeastern

19 Bering Sea offers opportunities to understand climate effects on recruitment because

20 inter-annual oscillations in ice coverage set up warm or cold conditions for juvenile fish

21 production. Depth-averaged temperature anomalies in the Bering Sea indicate the last

22 nine years have included 3 warm (2003-2005), an average (2006), and 5 cold (2007-

23 2011) years. We examined how these climatic states influenced the diet quality and

24 condition (size, energy density and total energy) of young-of-the-year (YOY) pollock

25 (Theragra chalcogramma) in fall. The implications of fall condition were further

26 examined by relating condition prior to winter to the number of age-1 recruits-per-

100 101

1 spawner the following summer (R/S). The percentage of lipid in pollock diets was

2 threefold higher in cold years compared with warm years, but stomach fullness did not

3 vary. Consequently, fish energy densities were 33% higher in cold years (P < 0.001) than

4 in warm years. In contrast, neither fish size (P = 0.666), nor total energy (P = 0.197)

5 varied with climatic condition. However, total energy was significantly (P = 0.007) and

6 positively correlated with R/S (R2= 0.736). We conclude that recruitment to age-1 in the

7 southeastern Bering Sea is improved under environmental conditions that produce

8 large, energy dense YOY pollock in fall.

10 3.2 Keywords: 11 pollock, Bering Sea, recruitment, climate change, winter, prey quality

101 102

2 3.3 Introduction 3 Fishery managers are limited in their ability to predict the number of young fish

4 recruiting into fisheries because recruitment is the product of a linked series of complex

5 and non-linear processes (e.g. Bailey et al. 2005). Most recruitment models rely on the

6 number and fecundity of spawning females to project the future number of recruits.

7 However, offspring must negotiate larval and prolonged juvenile stages prior to

8 recruitment, so there is often little or no correlation between the number of spawners

9 and subsequent recruits (e.g. Zheng, 1996). The problem of predicting recruitment may

10 be exacerbated in the future because climate change will unpredictably alter the

11 present interactions between the biotic and abiotic drivers. Consequently it is important

12 that fishery managers develop a quantifiable and mechanistic understanding of how

13 recruitment processes function under different climate conditions (Hollowed et al.,

14 2009).

15 The southeastern Bering Sea can be viewed as a natural laboratory for understanding

16 climate effects on fisheries recruitment. It occupies a broad, flat continental shelf north

17 of the Aleutian Islands on the western coast of Alaska. In spring, summer, and fall

18 oceanographic fronts bound distinct hydrographic areas known as the inner, middle and

19 outer domains. The southern extent of ice coverage in this area varies annually in

20 response to atmospheric forcing. The sea ice undergoes periods of relatively little

21 coverage (warm years) followed by periods of almost complete coverage (cold years).

22 This oscillation influences the development of oceanographic fronts with a profound

102 103

1 influence on the ecological processes within each domain. These effects of winter ice

2 cover have been incorporated into the Oscillating Control Hypothesis (OCH) (Hunt et al.,

3 2002), which predicts how warming temperatures will affect recruitment of walleye

4 pollock (Theragra chalcogramma) in the southeastern Bering Sea. The pollock harvest

5 there constitutes the largest fishery (by weight) in the United States. Consequently,

6 pollock recruitment in the southeastern Bering Sea is the object of intense interest.

7 A recent revision of the OCH focuses on the relationship between the eastern Bering Sea

8 zooplankton community and the timing of ice retreat in spring (Hunt et al., 2011). Early

9 ice retreat in warm years results in an absence of large crustacean zooplankton over the

10 middle shelf domain and favors communities comprised of small copepods such as

11 Pseudocalanus, Acartia, Oithona and Centropages. Conversely, the zooplankton biomass

12 over the middle shelf is dominated by medium-sized calanoids and euphausiids in cold

13 years (Coyle et al., 2011, Hunt et al., 2011, Ressler et al., 2012, Stabeno et al., 2012). The

14 species composition in cold years has been theorized to supply zooplanktivorous fishes

15 with improved access to lipid and subsequently increased recruitment success (Moss et

16 al., 2009).

17 Improved access to lipid-rich prey is important because young-of-the-year (YOY) pollock

18 face severe energy deficits as they enter winter (e.g. Sogard and Olla, 2000). These

19 energetic deficiencies are believed to account for winter mortality (Hurst 2007), which

20 can significantly reduce year class strength (Farley et al. 2007). High latitude fish need to

21 satisfy those energy deficits with endogenous energy (Post and Parkinson 2001) stored

103 104

1 as lipid. The time window during which pollock can provision themselves with lipid

2 occurs between the completion of metamorphosis in early August and the onset of

3 oceanographic winter (Willson et al., 2011, Siddon et al. 2013). This highlights the

4 importance accessing lipid rich prey in late summer and early fall. Those individuals that

5 can consume high lipid prey during this critical period should survive winter better than

6 individuals consuming leaner prey.

7 The size of fish prior to winter is also an important determinant of winter survival.

8 Observations of size-dependent mortality during winter for many freshwater and

9 marine species have led to the so-called critical size hypothesis (Farley et al. 2007; Hurst

10 2007). Empirical evidence suggests that mortality among high latitude fish in winter

11 depends on size, with larger individuals having the greater chance of survival. Size-

12 dependent mortality during winter has been described for silversides (Menidia menidia)

13 (Conover 1984), Pacific herring (Clupea pallasi) (Norcross et al. 2001), pollock (Sogard

14 and Olla 2000; Heintz and Vollenweider, 2010), pink (Oncorhynchus gorbuscha) (Moss et

15 al. 2005) and sockeye (O. nerka) (Farley et al. 2011) salmon. Larger fish are thought to

16 survive winter better than small fish because they have a greater capacity to store

17 energy (Schultz and Conover 1999). These data indicate that foraging conditions during

18 the critical pre-winter period should be an important determinant to recruitment

19 success because those conditions that allow fish to grow and store lipid will maximize

20 winter survival. Yet, none of these studies has considered food quality as a factor

21 contributing to winter survival.

104 105

1 We examine the hypothesis that YOY pollock are better prepared to survive winter in

2 the eastern Bering Sea when they are able to forage on lipid-rich zooplankton

3 communities characteristic of cold years. The objectives of this study were to

4 understand how climatic conditions in the eastern Bering Sea influence the pre-winter

5 condition of YOY pollock, and how that condition influences their winter survival. We

6 compared the nutritional condition (size, energy density, total energy) of walleye

7 pollock collected from warm and cold years to the percent lipid in their diets. Finally,

8 the importance of fall condition is examined by correlating the pre-winter condition of

9 YOY pollock with their winter survival, expressed as the number of age-1 recruits-per-

10 spawner the following summer.

11 3.4 Materials and methods

12 3.4.1 Pollock collection and processing

13 Pollock were sampled during the Bering Aleutian Salmon International Survey (BASIS)

14 conducted by the National Marine Fisheries Service each year between 2003 and 2011.

15 Years were classified as warm (20003-2005), average (2006) or cold (2007-2011) based

16 on ice coverage in March/April and the depth averaged temperature anomaly between

17 April and October at mooring M2 following Stabeno et al. (2012). BASIS is a fisheries

18 oceanographic survey of the eastern Bering Sea that takes place between mid August

19 and the end of September. Fish were collected by surface trawl during the day in a

20 gridded series of 81 fixed stations (Figure 1; Farley and Moss, 2009). Over the course of

21 the study period the actual number of stations surveyed has varied due to weather and

22 other constraints. However, the core areas where YOY pollock were historically caught

105 106

1 have been sampled in all years. In this report we only consider fish sampled from

2 surface trawls deployed south of 60° north latitude to maintain consistency between

3 years.

4 YOY pollock <100 mm total length, as delineated by Moss et al. (2009), were processed

5 from each station. The total number caught was counted or estimated from subsamples

6 of the total mass. The average weight of the YOY pollock at each station was recorded

7 by dividing the total mass of YOY by the number of YOY in that mass. A subsample of up

8 to 10 fish was retained to estimate diet composition. Another subsample of up to 15 fish

9 was retained and frozen for calorimetry.

10 Fish collected for diet analysis were processed immediately on the vessel to estimate

11 the average diet composition for each station. The stomach contents were removed

12 from each fish, pooled into a common sample and weighed to estimate the average bulk

13 mass of the stomach contents. The average stomach fullness at each station was

14 recorded as the bulk stomach content mass divided by the total fish mass and multiplied

15 by 100. The taxa comprising the bulk were identified to the lowest practical level and

16 weighed. The percentage contribution of each taxon to the bulk weight was recorded.

17 Diet compositions in warm and cold years have been described elsewhere (Moss et al.,

18 2009; Coyle et al., 2011) and are only briefly reviewed here.

19 The energy density (kJ/g wet mass) of the YOY pollock from each cruise (2003-2011) was

20 measured by bomb calorimetry. Samples for bomb calorimetry from a given station

21 were stratified by length. They were dried, pulverized and combusted using a Parr 6725

106 107

1 semi-micro bomb calorimeter. Pulverized homogenates were pressed into pellets

2 weighing at least 25 mg prior to combustion. Calorimetric data collected after 2007

3 were supplemented with energy densities calculated from proximate compositions.

4 Conversions to energy density from proximate composition relied on linear regressions

5 relating percent lipid to energy density in samples for which both analyses had been

6 performed. Details for the calorimetry and proximate analysis along with quality

7 assurance protocols can be found in Siddon et al. (2013).

8 3.4.2 Collection and processing of zooplankton

9 The percent lipid of pollock prey items was estimated from zooplankton samples

10 collected in a warm (2004) and cold (2009) year. Estimates of zooplankton percent lipid

11 for the warm year come from samples collected between 26 July and 19 August on a

12 survey centered near the Pribilof Islands. That survey also included a transect over the

13 shelf break and a grid of stations centered on a mooring (M2) on the continental shelf

14 (56.8° N 164° W) (Hunt et al., 2008, Figure 1). During this survey, zooplankton samples

15 were collected at night using the drogue net (150 µm mesh) of a MOCNESS with a

16 nonfiltering cod end. The MOCNESS was fished obliquely from the surface to near

17 bottom and prey species were sorted under a dissecting microscope on ice and then

18 frozen and transported to the lab in liquid nitrogen. Estimates of zooplankton percent

19 lipid for the cold year (2009) come from samples obtained during the BASIS survey

20 between 30 August and 28 September. During this survey, zooplankton samples were

21 collected using bongo nets (505 µm mesh) (Coyle et al., 2011) fished vertically from at

22 most 100 m depth to the surface during the day. Zooplankton were sorted immediately

107 108

1 after collection and representative individuals of each species were frozen in separate

2 air tight vials and held at < -75° C until they were processed in the laboratory.

3 Prey quality was measured in the laboratory as the percent lipid in the zooplankton wet

4 mass. Zooplankton were gently blotted to remove liquid and weighed to the nearest 10

5 µg. When individuals were too small to weigh individually, the average wet mass was

6 estimated from composited samples; weighed in bulk. Average mass was estimated by

7 dividing the bulk weight by the number of individuals in the composite. Lipid was

8 extracted from each sample using a modified Folch method. The minimum sample size

9 for extraction was 25 mg (wet mass), which often required compositing 3 to 35

10 individuals. Samples were extracted using a solvent mixture of 2:1 chloroform:methanol

11 in a Dionex Accelerated Solvent Extractor (ASE). See Siddon et al. (2013) for details on

12 the extraction protocol and quality assurance procedures. Samples from 2004 that were

13 < 25 mg were extracted in the same solvent mix using hand maceration. The extract was

14 passed over a micro-column of sodium sulfate as a drying agent to remove tissue

15 material. The same quality assurance methods and reference materials were used as

16 those extracted using the ASE to ensure comparability.

17 3.4.3 Data analysis

18 3.4.3.1 Climate effects on prey quality

19 Climate effects on prey quality were examined by comparing the percent lipid of prey

20 sampled from eastern Bering Sea in 2004 and 2009. The percent lipid of five species was

21 measured in both years, Theragra chalcogramma, Neocalanus cristatus, Calanus spp,

108 109

1 Thysanoessa inermis and Thysanoessa raschii. We compared their percent lipid

2 between years using a two-way ANOVA with year, species and their interaction as fixed

3 factors. Assumptions regarding normality (Anderson Darling) and homogeneity of

4 variance (F-test) of the response variables relative to climatic state were tested prior to

5 the analysis.

7 3.4.3.2 Climate effects on pollock diet

8 The diet compositions obtained in each survey year were combined with the percent

9 lipid (wet mass basis) of the prey to estimate the average percent lipid of diets in warm,

10 cold and average years. Percent fullness and dietary lipid of fish collected from the

11 BASIS surveys in warm, cold and average years was compared using a one-way ANOVA.

12 The average percent lipid of the diet (Ist) observed at each station (s) in a given year (t),

13 expressed as the percent of ingested wet mass was calculated as:

15 where Pis is the proportional contribution of prey type i at station s to the total mass

16 ingested and Lic is the percent lipid of prey type i under climate condition c. When only

17 one lipid value was available it was used for both the warm and cold year diets. When

18 diet could only be identified to taxon (i.e. genus, family or order), the overall average

19 lipid value for that taxon was applied. We calculated the average dietary percent lipid

109 110

1 for each year, by weighting Is by the catch at each station in year t (Cts) to account for

2 the highly variable number of fish caught at each station.

4 Estimates of fullness were also weighted by catch prior to analysis.

5 3.4.3.3 Climate effects on pre-winter condition

6 Climate effects on fish condition were examined by comparing the catch-weighted

7 averages across years. Climate responses were examined using a one-way ANOVA.

8 Climate classification was considered a fixed variable. The response variables included

9 catch-weighted average weight, energy density and total energy. The latter response

10 (i.e. kJ per fish) was estimated as the product of the catch-weighted average weight and

11 catch-weighted average energy density for a given year. We used total energy based on

12 the assumption that winter survival is optimized when fish can simultaneously accrete

13 tissue mass and allocate energy to depot lipids. Assumptions regarding normality

14 (Anderson Darling) and homogeneity of variance (F-test) of the response variables

15 relative to climatic state were tested prior to the analysis.

16 2.3.4 Influence of pre-winter condition on survival

17 We used regression analysis to evaluate the strength of the relationship of YOY

18 condition in fall and subsequent survival to age-1. Survival of year class t (St) was

19 indexed as the number of age-1 recruits (Rt+1) per female spawner (Ft) calculated as

110 111

2 using data provided in the Bering Sea Pollock Stock Assessment (Ianelli et al., 2011). The

3 number of recruits- per-spawner (hereafter referred to as survival) was regressed

4 against three indices of condition: catch-weighted average fish mass, catch-weighted

5 energy density and total energy.

6 3.5 Results

7 3.5.1 Climate effects on prey quality

8 We estimated the percent lipid in 22 zooplankton taxa between our 2004 and 2009

9 collections, five of those taxa were sampled in both years (Figure 2). The average

10 percent lipid of zooplankton differed significantly between years (F1,80 = 12.59, p <

11 0.001), and there was no interaction between species and year (F1,80 = 0.12, p = 0.890).

12 Prey species had higher percent lipid in the cold year compared with the warm year.

13 This was true for the euphausiids, Thysanoessa raschii and Thysanoessa inermis, as well

14 as teleosts such as pollock, which all increased in percent lipid by at least 21%. Calanus

15 spp. increased in percent lipid by 37% between the warm and cold year and Neocalanus

16 cristatus increased by 50%.

17 3.5.2 Climate effects on pollock diets

18 YOY pollock consumed a more diverse diet in warm years, but tended to have lower

19 average fullness than in cold years. The diets of the fish sampled in 2004-2007 were

20 previously reported by Moss et al. (2009), while those from 2003-2009 were reported by

111 112

1 Coyle et al. (2011). The diets observed in 2010 and 2011 conformed to the previously

2 described patterns. In general, diets were most diverse in warm years (Figure 3),

3 including many additional prey taxa not observed in cold years. The most frequent taxon

4 in warm years was Pseudocalanus spp., (listed as “Small copepod” in Fig. 3), accounting

5 for up to 60% of the ingested mass. In cold years Pseudocalanus spp. were also relatively

6 frequent, but they made up a smaller proportion of the mass ingested because larger

7 taxa were also ingested. The most frequent items in cold years were Calanus spp., but

8 euphausiids accounted for the majority of the mass. Despite the compositional

9 differences between warm and cold years, stomach fullness remained constant (F2,6 =

10 1.52, p = 0.293). Although, the average fullness increased from 1.7 ± 0.1 % (mean ± 1

11 s.e.) in the warm years to 2.2 ± 0.4% in the cold years.

12 Pollock diets had a greater amount of lipid in cold years due to the larger contributions

13 of Calanus spp and euphausiids. Items typical of warm years, such as small copepods

14 (including Pseudocalanus spp.), were intermediate in percent lipid (Figure 2), but can

15 also be seen in cold year diets. In contrast, Calanus spp. and euphausiids have higher

16 percent lipid (Figure 2), and were abundant in the diets of pollock sampled in cold years

17 (Figure 3). Consequently, pollock diets had significantly more lipid in the cold years than

18 in the warm years (F2,6 = 12.84, p = 0.007) , averaging about threefold greater (Figure 4).

19 The taxa in our collection were sufficient to describe the percent lipid of >98% of the

20 mass observed in pollock stomachs from cold years. We had to supplement our

21 collection of percent lipid with published values (Figure 2) to account for the more

112 113

1 diverse diets observed in warm years. Inclusion of published values allowed us to

2 account for a median of >90% of the mass observed in those stomachs.

3 3.5.3 Climate effects on pre-winter condition

4 Neither the size nor the total energy content of pollock were influenced by climate, but

5 there was an effect of climate on energy density. Pollock were slightly lighter in mass

6 during the warm years averaging 1.97 ± 0.25 g compared with 2.35 ± 0.28 g in the cold

7 years, but these differences were not statistically significant (F2,5 > 0.43 p > 0.666)

8 (Figure 5). In contrast, the catch-weighted average energy density was highly correlated

9 with climatic condition (F2,6 = 35.38, p < 0.001; Figure 5b). Energy density in the cold

10 years (4.90 ± 0.11 kJ/g) was 33% greater than the warm years (3.67 ± 0.04 kJ/g). In

11 contrast to energy density, there was no relation between climate and the total energy

12 content of YOY pollock (F2,6 = 2.16, p = 0.197). However, there was a trend of greater

13 total energy in cold years (11.2 ± 1.5 kJ/fish) compared to warm years (7.3 ± 0.9 kJ/fish).

14 3..5.4 Influence of pre-winter condition on survival

15 Total energy was the best predictor of survival. The product of weight and energy

16 density expressed as total energy was significantly (F1,6 = 16.77, p = 0.006) and positively

17 (r2= 0.736) correlated with recruitment to age-1 (Figure 6). Weight (r2= 0.496) and

2 18 energy density (r = 0.677) were also positively correlated with survival (F1,6 > 5.92, p <

19 0.051), but accounted for less of the variation in survival over the last nine years.

113 114

1 3.6. Discussion

2 These data demonstrate that the ability of YOY pollock to provision themselves prior to

3 winter has direct bearing on their survival to age-1. Climate effects on prey field

4 composition (Coyle et al. 2011) led to the consumption of high lipid diets in cold years.

5 Fish that consumed high lipid diets had higher energy densities as a result of the

6 increased level of lipid in their tissues (Anthony et al. 2000). The effect of consuming

7 high lipid diets on body composition is well known (e.g. Arzel et al., 1994). In larval

8 pollock high lipid diets may also increase survival by reducing activity costs (Davis and

9 Olla, 1992). However, consuming high lipid diets did not guarantee survival in this

10 study, fish also needed to reach sufficient size. This indicates the critical size hypothesis

11 should be refined to reflect both the effects of large size late in the growing season and

12 the availability of high lipid diets. The oceanographic conditions that produce large,

13 energy dense fish are also those conditions likely to lead to increased winter survival of

14 YOY pollock.

15 3.6.1 Climate impacts on food abundance and quality

16 Food limitation is often cited as a mechanism restricting production and recruitment in

17 fish populations, but the data presented here indicate food quality is also important.

18 Zooplankton biomass in cold years was twice that of warm years in the southeastern

19 Bering Sea (Coyle et al., 2011), but pollock catch-per-unit-effort on BASIS surveys was

20 lower in cold years (Hunt et al., 2011). This indicates the per capita availability of prey

21 was likely much higher in cold years. Pollock consumption rates are higher at warm

22 temperatures (Kooka et al. 2007) suggesting lower average stomach fullness in warm

114 115

1 years reflected a density-dependent impact on foraging success. However, pollock in

2 warm years would have had to ingest three times more food in warm years to ingest the

3 same amount of lipid as fish in cold years. It is unlikely that pollock would be able to

4 consume three times as much food in warm years if the food were available because

5 maximum consumption rates only increase about twofold between 3 and 12 °C (Kooka

6 et al., 2007).

7 The processes which led to the increased availability of high quality prey in cold years

8 was augmented by the increased percent lipid found in specific prey taxa. Improved

9 quality in Calanus spp. likely relates to mechanisms associated with their increased

10 abundance such as improved coupling between spring blooms and metamorphosis

11 (Baier and Napp, 2003) or reduced stratification and increased nutrient supply during

12 summer (Coyle et al. 2011). However, it is possible that the climate related changes in

13 food quality we observed are related to seasonal effects on lipid storage. Zooplankton

14 sampled in 2009 were collected as much six weeks later in the year than those in 2004

15 and the percent lipid of the specific prey increased by 20% to 65%. However this

16 seasonal effect, if it exists, is only of marginal importance because the majority of the

17 warm year diets were comprised of leaner prey taxa than in cold years.

18 3.6.2 Importance of winter

19 Larger fish can store proportionately more energy and they use it more slowly than

20 smaller fish. This is because the allometry for lipid storage before winter has a steeper

21 slope than the allometry for lipid depletion during winter (Schultz and Conover, 1999).

22 Therefore, the correlation between pollock condition and their subsequent survival to

115 116

1 age-1 was a product of both their size and energy density, which accounts for the

2 relatively poor survival of the 2007 year class. Fish from the 2007 year class had high

3 energy densities and consumed lipid-rich diets, but were relatively small in size.

4 Consequently, their overall energy content was low at the onset of winter.

5 Conditions during the period in which pollock provision themselves for winter seem to

6 have been more important in determining winter survival, than conditions earlier in

7 summer. In warm years pollock may spawn earlier (Smart et al., 2012) and grow faster if

8 food supplies are not limiting (Kooka et al., 2007). Warm conditions apparently led to

9 improved larval survival and supported the production of YOY pollock as indicated by

10 the catches-per-unit-effort during the late summer and early fall in the southeastern

11 Bering Sea (Hunt et al., 2011). However, density dependence during the provisioning

12 period may have ultimately limited the production of age-1 pollock. These observations

13 suggest winter survival in the Bering Sea determines recruitment to a greater extent

14 than larval production, a process similar to that reported for pollock in the Gulf of Alaska

15 (Bailey, 2000; Cianelli et al., 2005).

16 Moderation of the severity of winters and a warming of the upper water column during

17 winter are probable consequences of changing climate in southeastern Bering Sea. Over

18 the past 30 – 40 years, the stormy spring-winter-fall period has decreased in length

19 resulting in a longer summer stable period and decreased frequency of summer storms

20 (Stabeno et al., 2012). These storms promote new production in the late summer/early

21 fall and may prolong the period during which YOY pollock can find lipid-rich prey items.

116 117

1 In addition, decreased ice cover in the southeastern Bering Sea during warm periods has

2 been associated with a later spring phytoplankton bloom (Hunt et al., 2011; Stabeno et

3 al., 2012). Thus increased water temperatures may present a three-fold challenge to

4 juvenile pollock; 1) low densities of lipid-rich prey in the late summer/early fall result in

5 low lipid reserves, 2) higher temperatures result in higher basal metabolic costs, and 3) a

6 delay in the production cycle the following winter/spring means the period without

7 sufficient prey resources is longer.

8 3.7 Conclusion

9 These observations provide a quantitative basis for predicting recruitment as a function

10 of environmental conditions. Estimates of total energy in fall integrate the effects of

11 climate, diet, growth, prey abundance and quality at the end of the first growing season.

12 The effects of climate on diet quality and condition of YOY pollock are consistent with

13 observations of an inverse relationship between late summer sea surface temperatures

14 and survival (Mueter et al., 2011). These data further demonstrate how fishery

15 managers can take a significant step towards an ecosystem-based approach by linking

16 recruitment models to the environmental and biological parameters responsible for

17 production.

18 3.8 Acknowledgements 19 We thank all of the people involved in collecting and sorting the samples used in this

20 study. This includes the crews of numerous vessels. In addition it includes numerous

21 students that have helped to sort, prepare and process samples in the laboratory.

22 Collection of summer 2004 prey samples was supported by NSF Grant OPP-0327308 (to

117 118

1 G.L. Hunt, Jr.) and NOAA’s North Pacific Climate Regimes and Ecosystem Productivity

2 (NPCREP) research program. This research is contribution NPRB 414, BEST-BSIERP 93,

3 and EcoFOCI-NXXX to NOAA’s NPCREP Program. References to trade names do not imply

4 endorsement by the National Marine Fisheries Service, NOAA. The findings and

5 conclusions in this paper are those of the authors and even though NOAA reviewed the

6 work and paid our salaries the views here do not represent those of NMFS or NOAA.

8 3.9 References 9 Anthony, JA, Roby, DD, Turco, KR (2000) Lipid content and energy density of forage 10 fishes 11 from the northern Gulf of Alaska. J. Mar. Bio. Ecol. 248:53-78. 12 13 Arzel, J, Lopez, FXM, Métailler, R, Stéphan, G, Viau, M, Gandemer, G, Guillaume, J. (1994) Effect 14 of dietary lipid on growth performance and body composition of brown trout (Salmo trutta) 15 reared in seawater. Aquaculture 123:361-375. 16 17 Baier, CT, Napp JM (2003) Climate induced variability in Calanus marshallae populations. J. 18 Plankton Res. 25: 771-782. 19 20 Bailey, KM (2000) Shifting control of recruitment of walleye pollock Theragra chalocgramma 21 after a major climatic and ecosystem change. Mar. Ecol-Prog. Ser. 198: 215-224. 22 23 Bailey, K.M., Ciannelli, L.,Bond, N.A., Belgrano, A., and Stenseth, N.C. (2005). Recruitment of 24 walleye pollock in a physically and biologically complex ecosystem: A new perspective. Prog. 25 Oceanogr. 76:24-42. 26 27 Ciannelli L, Paul AJ, Brodeur RD (2002) Regional, interannual and size-related variation of age 0 28 walleye pollock whole body energy content around the Pribilof Islands, Bering Sea. J. Fish Biol. 29 60: 1267-1279. 30 31 Ciannelli, L, Bailey, KM, Chan, K, Belgrano, A, Stenseth, NC (2005) Climate change 32 causing phase transitions of walleye pollock (Theragra chalcogramma) recruitment 33 dynamics. Proc. Royal. Soc. B 272:1735-1743. 34 35 Conover, DO (1984) Adaptive significance of temperature-dependent sex determination in a 36 fish. Am. Nat. 123:297-313. 37

118 119

1 Coyle KO, Eisner LB, Mueter FJ, Pinchuk AI, Janout MA, Cieciel KD, Farley EV, Andrews AG (2011) 2 Climate change in the southeastern Bering Sea: impacts on pollock stocks and implications for 3 the Oscillating Control Hypothesis. Fish. Oceanogr. 20(2): 139-156. 4 5 Davis, M.W. and Olla, B.L. (1992) Comparison of growth, behavior, and lipid concentrations of 6 walleye pollock Theragra chalcogramma larvae fed lipid-enriched, lipid-deficient and field 7 collected prey. Mar. Ecol-Prog. Ser. 90:23-30. 8 9 Farley EV, Moss, JH, Beamish, RJ (2007) A review of the critical size, critical period 10 hypothesis for 11 juvenile Pacific salmon. North Pacific Anadromous Fish Commission Bulletin 4:311-317. 12 13 Farley EV, Moss JH (2009) Growth rate potential of juvenile chum salmon on the eastern Bering 14 Sea shelf: an assessment of salmon carrying capacity. North Pacific Anadromous Fish 15 Commission Bulletin 5: 265-277. 16 17 Farley, EV, Starovoytov, A, Naydenko, S, Heintz, R, Trudel, M, Guthrie, C, Eisner, L, 18 Guyon JR (2011) Implications of a warming eastern Bering Sea for Bristol Bay sockeye 19 salmon. ICES J. Mar. Sci. 68:1138-1146. 20 21 Foy, RJ, AJ Paul (1999) Winter feeding and changes in somatic energy content of age-0 Pacific 22 herring in Prince William Sound , Alaska. Trans. Am. Fish. Soc. 128: 1193-1200. 23 24 Heintz RA, Vollenweider JJ (2010) Influence of size on the sources of energy consumed by 25 overwintering walleye pollock (Theragra chalcogramma). J. Exp. Mar. Biol. Ecol. 393(1-2): 43-50. 26 27 Hollowed AB, Bond NA, Wilderbuer TK, Stockhausen WT, A’mar ZT, Beamish RJ, Overland JE, 28 Schirripa MJ (2009) A framework for modelling fish and shellfish responses to future climate 29 change. ICES J. Mar. Sci. 66 (7): 1584-1594. 30 31 Hunt GL Jr, Stabeno PJ, Walters G, Sinclair E, Brodeur RD, Napp JM, Bond NA (2002) Climate 32 change and control of the southeastern Bering Sea pelagic ecosystem. Deep-Sea Res. Part II 49: 33 5821-5853. 34 35 Hunt, G. L., Jr., Stabeno, P.J., Strom, S., and Napp, J.M. (2008). Patterns of spatial and temporal 36 variation in the marine ecosystem of the southeastern Bering Sea, with special reference to the 37 Pribilof Domain. Deep Sea Res. II 55:1919-1944. 38 39 Hunt GL Jr, Coyle KO, Eisner L, Farley EV, Heintz R, Mueter FJ, Napp JM, Overland JE, Ressler PH, 40 Salo S, Stabeno PJ (2011) Climate impacts on eastern Bering Sea foodwebs: A synthesis of new 41 data and an assessment of the Oscillating Control Hypothesis. ICES J. Mar. Sci. 42 doi:10.1093/icesjms/fsr036. 43 44 Hurst, TP (2007) Causes and consequences of winter mortality in fishes. J. Fish Bio. 71:315-345. 45

119 120

1 Huss, M, Byström P, Strand Å, Eriksson L, Persson L (2008) Influence of growth history on the 2 accumulation of energy reserves and winter mortality in young fish. Can. J. Fish. Aquat. Sci. 65: 3 2145-2156. 4 5 Ianelli, J.N., Hollowed, A.B., Haynie, A.C., Mueter, F.J., and Bond, N.A. (2011). Evaluating 6 management strategies for eastern Bering Sea walleye pollock (Theragra chalcogramma) in a 7 changing environment. ICES J. Mar. Sci. 68:1297-1304. 8 9 Ianelli JN, Honkalehto T, Barbeaux S, Kotwicki S, Aydin K, Williamson N (2011) Assessment of the 10 walleye pollock stock in the Eastern Bering Sea. In: Stock assessment and fishery evaluation 11 report for the groundfish resources of the Bering Sea/Aleutian Islands regions. North Pacific 12 Fishery Management Council, 605 W. 4th Ave., Suite 306, Anchorage, AK 99501. 13 14 Kooka K, Yamamura O, Nishimura A, Hamatsu T, Yanagimoto T (2007) Optimum temperature for 15 growth of juvenile walleye pollock Theragra chalcogramma. J. Exp. Mar. Biol. Ecol. 347: 69-76. 16 17 Kooka K, Yamamura O (2011) Winter energy allocation and deficit of juvenile walleye pollock 18 Theragra chalcogramma in the Doto area, northern Japan. Env. Biol. Fish doi:10.1007/s10641- 19 011-9957-1. 20 21 Lee, RF, Hagen W, Kattner G (2006) Lipid storage in marine zooplankton. Mar. Ecol- Prog. Ser. 22 307: 273-306. 23 24 Moss, JH, (2005) Evidence for size-selective mortality after the first summer of ocean 25 growth by pink salmon. Trans. Am. Fish. Soc. 134:1313-1322. 26 27 Moss JH, Farley EV, Feldman AM, Ianelli JN (2009) Spatial distribution, energetic status, and food 28 habits of eastern Bering Sea age-0 walleye pollock. Trans. Am. Fish. Soc. 138: 497-505. 29 30 Mueter, FJ, Bond NA, Ianelli JN, Hollwed AB (2011) Expected declines in recruitment of walleye 31 pollock (Theragra chalcogramma) in the eastern Bering Sea under future climate change. ICES J. 32 of Mar. Sci. 68(6): 1284-1296. 33 34 Nomura T, Davis NM (2005) Lipid and moisture content of salmon prey organisms and stomach 35 contents of chum, pink and sockeye salmon in the Bering Sea. North Pacific Anadromous Fish 36 Commission Technical Report Number 6: 59-61. 37 38 Norcross, BL, Brown, ED, Foy, RJ, Frandsen, M, Gay, SM, Kline, TC, Mason, DM Patrick, 39 EV, Paul, AJ, Stokesbury, KDE (2001) A synthesis of the life history and ecology of 40 juvenile Pacific herring in Prince William Sound, Alaska. Fish. Oceanog. 10:42-57. 41 42 Peters, J (2006) Lipids in key copepod species of the Baltic Sea and North Sea – implications for 43 life cycles, trophodynamics and food quality. PhD Dissertation. University of Bremen. 44 45 Post, J. R., and E. A. Parkinson. 2001. Energy allocation strategy in young fish: allometry 46 and survival. Ecology 82:1040-1051. 47

120 121

1 Ressler, P.H., De Robertis, A., Warren, J.D., Smith, J.N., Kotwicki, S. (2012) Developing an 2 acoustic index of euphausiid abundance to undertstand trophic interations in the Bering Sea 3 ecosystem. Deep-Sea Res., II, 65-70:184-195. 4 5 Schultz ET, Conover DO (1999) The allometry of energy reserve depletion: test of a mechanism 6 for size-dependent winter mortality. Oecologia 119: 474-483. 7 8 Siddon, EC, Heintz RA, Mueter FJ (2013) Conceptual model of energy allocation in walleye 9 pollock (Theragra chalcogramma) from larvae to age-1 in the southeastern Bering Sea. Deep 10 Sea Res., II. 11 12 Smart TI, Siddon EC, Duffy-Anderson JT, Horne, JK (2012) Alternating temperature 13 states influence walleye pollock early life stages in the southeastern Bering Sea. Mar. 14 Ecol-Prog. Ser. 455:257-267. 15 16 Sogard, S.M., and Olla, B.L. (2000). Endurance of simulated winter conditions by age-0 walleye 17 pollock: Effects of body size, water temperature and energy stores. J. Fish Biol. 56:1-21. 18 19 Stabeno, P. J., Kachel N. B.. Moore, S. E., Napp, J. M., Sigler, M., Zerbini, A. N., (2012). 20 Comparison of warm and cold years on the southeastern Bering Sea shelf and some implications 21 for the ecosystem. Deep-Sea Res., II, 65-70:31-45. 22 23 Willson, M.T., Mier, K., and Dougherty, A. 2011. The first annulus of otoliths: a tool for studying 24 intra-annual growth and size-at-age variability among walleye pollock (Theragra 25 chalcogramma). Env. Biol. Fish., 92:53-63. 26 27 Yamamoto, T, Teruya K, Hara T, Hokazono H, Hashimoto H, Suzuki N, Iwashita Y, Matsunari H, 28 Furuita H, Mushiake K (2008) Nutrtitional evaluation of live food organisms and commercial dry 29 feeds used for seed production of amberjack Seriola dumerili. Fish. Sci. 74: 1096-1108. 30 31 Zheng, J (1996) Herring stock-recruitment relationships and recruitment patterns in the north 32 Atlantic and northeast Pacific oceans. Fish. Res. 26: 257-277. 33

121 122

2 3.10 Figures 3 Figure 1. Map of the eastern Bering Sea showing BASIS sampling stations (X’s) . 4 Zooplankton were collected at stations identified with filled circles in 2009 and at 5 stations identified by open circles in 2004.

6 Figure 2. Average percent lipid (± 1 s.d.) of prey consumed by YOY pollock in 2004 and 7 2009. Bar colors reflect year the sample was collected, open bars show prey collected in 8 2004, closed bars show prey collected in 2009, gray bars show values taken from other 9 sources: 1. (Nomura and Davis, 2005), 2(Yamamoto et al., 2008),3(Foy and Paul, 1999), 10 4(Lee et al., 2006) and 5(Peters, 2006).

11 Figure 3. Average percent composition (±1 s.d.) of YOY pollock diets in warm (2003- 12 2005) and cold (2007-2011) years. Each prey item is expressed as percent of total mass 13 of prey examined within a year and averaged across the appropriate years. Only those 14 prey items that averaged at least 1% of the mass examined are shown. Taxa have been 15 combined to simplify presentation. For more detailed comparisons of diets in warm and 16 cold years see Moss et al. (2009) and Coyle et al. (2011).

17 Figure 4. Catch-weighted average (± 95% confidence interval) lipid content of YOY 18 pollock diets in warm (2003-2005), average (2006) and cold (2007-2011) years.

19 Figure 5. Catch-weighted average weight (± 1 s.d.) (top panel) and energy density 20 (bottom panel) of YOY pollock sampled from 2003 to 2011.

21 Figure 6. Relationship between catch-averaged weight, energy density and total energy 22 of YOY pollock with survival. Survival is estimated as the number of age-1 recruits per 23 female spawner and was not available for 2011 year class at time of publication.

122 123

1 2

100 50 m m m

Slop e

123 124

1 Figure 1. Eastern Bering Sea showing BASIS sampling stations (X’s). Zooplankton were 2 collected in 2009 at stations identified with filled circles. Zooplankton sampling in 2004 3 was conducted at locations identified by open circles.

Prey Category Gammarid amphipod¹ Thysanoessa longipes Clione limacina Thysanoessa inermis Calanus marshallae Small Copepod Neocalanus plumchrus Themisto libellula Neocalanus cristatus Thysanoessa spinifera hyperiid amphipod Fish laravae (Sebastes) Pseudocalanus sp.⁵ Mysidacea Eucalanus bungii 2009 (Cold) Bivalvia larvae³ Decapoda zoea 2004 (Warm) Cirripedia nauplius³ Literature value Euphausiacea Thysanoessa raschii Ammоdytes hexapterus Metridia pacifica Centropages … Ostracoda³ Cirripedia cyprid Gadus macrocephalus Podon sp. ³ Eurytemora sp.³ Epilabidocera amphitrites³ Limacina helicina Chaetognatha Theragra chalcogramma Fish larvae (pollock) Polychaeta³ Oithona similis² Acartia clausi² Hydromedusae Oikopleura sp.¹ Ctenophore

0 4 8 12 16 20 % Lipid (wet weight) 4

5 Figure 2. Average percent lipid (± 1 s.d.) of prey consumed by YOY pollock in 2004 and 6 2009. Bar colors reflect year the sample was collected, open bars show prey collected in

124 125

1 2004, closed bars show prey collected in 2009, gray bars show values taken from other 2 sources: 1. (Nomura and Davis, 2005), 2(Yamamoto et al., 2008),3(Foy and Paul, 1999), 3 4(Lee et al., 2006) and 5(Peters, 2006).

125 126

Cnidarian Warm Crab zoea/megalopae Cold Shrimp Larval fish Pteropod Chaetognath Squid Small copepod Mysid Other fish Hyperiid amphipod YOY pollock Adult euphausiid Large copepod

60 40 20 0 20 40 60 % of Total Mass

2 Figure 3. Average percent composition (±1 s.e.) of YOY pollock diets in warm (2003- 3 2005) and cold (2007-2011) years. Each prey item is expressed as percent of total mass 4 of prey examined within a year and averaged across the appropriate years. Only those 5 prey items that averaged at least 1% of the mass examined are shown. Taxa have been 6 combined to simplify presentation. For more detailed comparisons of diets in warm and 7 cold years see Moss et al. (2009) and Coyle et al. (2011).

126 127

% Lipid in Diet

0 Warm Average Cold

7 Figure 4. Catch-weighted average (± 95% confidence interval) lipid content of YOY 8 pollock diets in warm (2003-2005), average (2006) and cold (2007-2011) years.

127 128

Weight (g) 4

0 2002 2004 2006 2008 2010 2012

Energy density (kJ/g) 6 Warm Average 5 Cold

3 3 2002 2004 2006 2008 2010 2012 4 Figure 5. Catch-weighted average weight (± 1 s.d.) (top panel) and energy density 5 (bottom panel) of YOY pollock sampled from 2003 to 2011.

128 129

R 2 = 0.496 0 1.5 2.0 2.5 3.0

Weight (g)

e n

w Warm year a 20

p Average year

- r

e Cold year

s t

i 10

- R 2 = 0.677 e

g 0

A 3.5 4.0 4.5 5.0 Energy density (kJ/g)

R 2 = 0.736

0 5.0 7.5 10.0 12.5 15.0 Total Energy (kJ/fish) 5 6

7 Figure 6. Relationship between catch-averaged weight, energy density and total energy 8 of YOY pollock with survival. Survival is estimated as the number of age-1 recruits per 9 female spawner and was not available for 2011 year class at time of publication.

129 130

2 Chapter 4 - Spatial match-mismatch between juvenile fish and 3 prey provides a mechanism for recruitment variability across 4 contrasting climate conditions in the eastern Bering Sea 5 Elizabeth Calvert Siddon1*, Trond Kristiansen2, Franz J. Mueter1, Kirstin K. Holsman3,

6 Ron A. Heintz4, and Edward V. Farley4

7 1University of Alaska Fairbanks, School of Fisheries and Ocean Sciences, Juneau, AK

8 99801 USA

9 2Institute of Marine Research, Bergen, Norway

10 3Joint Institute for the Study of the Atmosphere and Ocean, University of Washington.

11 Alaska Fisheries Science Center, National Marine Fisheries Service, National Oceanic and

12 Atmospheric Administration, 7600 Sand Point Way NE, Seattle, WA 98115 USA

13 4Ted Stevens Marine Research Institute, Alaska Fisheries Science Center, National

14 Marine Fisheries Service, National Oceanic and Atmospheric Administration, 17109 Pt.

15 Lena Loop Road, Juneau, AK 99801 USA

17 Published in PLoS (2014)

19 4.1 Keywords: 20 Growth potential, bioenergetics model, individual-based model, walleye pollock

21 (Theragra chalcogramma), Bering Sea, climate variability, juvenile fish

130 131

1 4.2 Abstract 2 Understanding mechanisms behind variability in early life survival of marine fishes

3 through modeling efforts can improve predictive capabilities for recruitment success

4 under changing climate conditions. Walleye pollock (Theragra chalcogramma) support

5 the largest single-species commercial fishery in the United States and represent an

6 ecologically important component of the Bering Sea ecosystem. Variability in walleye

7 pollock growth and survival is structured in part by climate-driven bottom-up control of

8 zooplankton composition. We used two modeling approaches, informed by

9 observations, to understand the roles of prey quality, prey composition, and water

10 temperature on juvenile walleye pollock growth: (1) a bioenergetics model that included

11 local predator and prey energy densities, and (2) an individual-based model that

12 included a mechanistic feeding component dependent on larval development and

13 behavior, local prey densities and size, and physical oceanographic conditions. Prey

14 composition in late-summer shifted from predominantly smaller copepod species in the

15 warmer 2005 season to larger species in the cooler 2010 season, reflecting differences

16 in zooplankton composition between years. In 2010, the main prey of juvenile walleye

17 pollock were more abundant, had greater biomass, and higher mean energy density,

18 resulting in better growth conditions. Moreover, spatial patterns in prey composition

19 and water temperature lead to areas of enhanced growth, or growth ‘hot spots’, for

20 juvenile walleye pollock and survival may depend on the overlap between fish and these

21 areas. This study provides evidence that a spatial mismatch between juvenile walleye

22 pollock and growth ‘hot spots’ in 2005 contributed to poor recruitment while a higher

131 132

1 degree of overlap in 2010 resulted in improved recruitment. Our results indicate that

2 climate-driven changes in prey composition and quality can impact growth of juvenile

3 walleye pollock, potentially severely affecting recruitment variability.

4 4.3 Introduction 5 The match-mismatch hypothesis [1] proposes that predator survival is dependent on the

6 temporal and spatial overlap with prey resources [2]. Factors affecting temporal

7 overlap, such as climate variability through altered phenology, can lead to changes in

8 survival at critical life stages [3,4]. Temporal variation in spatial patterns of physical or

9 biological conditions may concurrently affect survival. For example, in temperate and

10 sub-arctic marine ecosystems, the timing of the spring bloom varies between years,

11 driven by physical oceanographic conditions that change due to climate variability (e.g.,

12 [5]). These conditions, such as the onset of stratification and light availability, also affect

13 the spatial patterns of zooplankton abundance, which further influences the feeding

14 success of planktivorous fish species. Hence, variability in the spatial overlap of predator

15 and prey, as well as differences in prey quality [6,7], may directly affect differences in

16 year-class success of many marine fish species [8,9].

17 Variability in year-class strength of gadids is often associated with changing

18 physical conditions [10,11]. The eastern Bering Sea (EBS) has experienced multi-year

19 periods of both warm and cold conditions since the turn of the 21st century [12], with

20 cold years having much higher walleye pollock (Theragra chalcogramma) recruitment

21 on average [13]. Changes in zooplankton composition between these periods have been

132 133

1 identified as an important driver of recruitment success for walleye pollock [9,14], but

2 the mechanistic links remain poorly understood.

3 Interannual changes in ocean temperatures [12] and shifts in the spatio-

4 temporal distribution of prey [14] make walleye pollock in the EBS an ideal case study to

5 better understand drivers of recruitment success in sub-arctic marine fish. Larger

6 zooplankton taxa, such as lipid-rich Calanus spp., were less abundant during recent

7 warm years, possibly causing reduced growth rates and subsequent year-class strength

8 of juvenile walleye pollock (hereafter juvenile pollock). In contrast, higher abundances

9 of lipid-rich prey, combined with lower metabolic demands in cold years, may have

10 allowed juvenile pollock to acquire greater lipid reserves by late summer and experience

11 increased overwinter survival [9]. Although the energetic condition of juvenile pollock in

12 late summer is recognized as a predictor of age-1 abundance during the following

13 summer in the EBS [13], the causal mechanism linking differences in prey abundance

14 and quality to walleye pollock survival remains untested.

15 The objectives of this study were to better understand the roles of prey quality,

16 prey composition, and water temperature on juvenile pollock growth through (1)

17 estimating spatial differences in maximum growth potential of juvenile pollock using a

18 bioenergetics modeling approach, (2) comparing maximum growth potential to

19 predicted growth from an individual-based model (IBM), and (3) quantifying the impact

20 of temperature, prey abundance, and prey quality on spatial variability in growth

21 potential.

133 134

1 4.4 Materials and methods

2 4.4.1 Ethics statement 3 Collection of physical and biological oceanographic data and fish samples during the US

4 Bering-Aleutian Salmon International Surveys (BASIS) conducted on the EBS shelf was

5 approved through the National Marine Fisheries Service, Scientific Research Permit

6 numbers 2005-9 and 2010-B1. Collection of biological data in the US Exclusive Economic

7 Zone by federal scientists to support fishery research is granted by the Magnuson -

8 Stevens Fishery Conservation and Management Act.

9 4.4.2 Modeling approaches 10 Two alternative modeling approaches were parameterized based on samples of juvenile

11 pollock, zooplankton, and oceanographic data collected during the BASIS surveys

12 conducted on the EBS shelf from mid-August to October 2005 and 2010 ([15]; Figure

13 S1). We selected 2005 (warm) and 2010 (cold) for our analyses based on data availability

14 and the pronounced contrast in ocean conditions between these years (e.g., depth-

15 averaged temperature anomalies over the middle shelf; [12]). Extensive spatial coverage

16 of the surveys, combined with varying climate conditions between years, provided

17 ample data with which to inform the models and compare differences in predicted

18 growth between a representative warm and cold year in the EBS.

19 Maximum growth potential from a Wisconsin-type bioenergetics model

20 parameterized for juvenile pollock (modified from [16]) was compared with predicted

21 growth from a mechanistic IBM [17]. Comparing model-based predictions of growth

22 allowed for a better understanding of the mechanisms behind temporal and spatial

23 variability in growth patterns and an evaluation of the importance of different model

134 135

1 parameters. Growth (g•g-1•day-1; weight measures refer to wet weight throughout) was

2 estimated for 65 mm standard length (SL; 2.5 g) juvenile pollock, corresponding to the

3 average size of age-0 fish (<100 mm total length; TL) observed in late summer (2005:

4 64.1 ± 6.7 mm SL [mean ± SD] and 1.97 ± 0.93 g; 2010: 64.3 ± 9.2 mm SL and 2.39 ± 0.94

5 g; conversion between TL and SL followed [18]).

6 4.4.3 Field observations 7 Juvenile pollock abundance

8 Juvenile pollock were collected from the EBS shelf (inner domain: 0-50 m isobath,

9 middle domain: 50-100 m isobath, and outer domain: 100-200 m isobath) using a

10 midwater rope trawl following methods described in [19]. Catch per unit effort (CPUE;

11 #•m-2) was calculated as:

ni 12 CPUE i  Eq. 1 di h

13 where ni is the number of fish collected in a given haul i, di is the trawl distance (m)

14 calculated from starting and ending ship position, and h is the horizontal spread of the

15 trawl net (m). Only surface tows at pre-defined stations were used to compute CPUE

16 because midwater tows specifically targeted acoustic sign of walleye pollock.

17 Water temperature

18 Vertical profiles of water temperature were collected at each station sampled for

19 oceanography using a Sea-Bird Electronics (SBE) conductivity-temperature-depth (CTD)

20 profiler SBE-25 (2005) or SBE-911 (2010). The average temperature in the upper 30 m of

21 the water column was used in the bioenergetics model, assuming juvenile pollock

22 collected from surface trawls were concentrated within the upper 30 m [15]. For the

135 136

1 IBM, the water column was divided into 1 m discrete depth bins. For all IBM simulations,

2 the depth of the water column was set to the upper 100 m of all deeper stations (n=9 of

3 116 in 2005, n=27 of 160 in 2010) because MOCNESS data used to develop vertical

4 profiles of zooplankton distribution (see ‘Zooplankton vertical profiles’ below) was

5 limited to 100 m. For stations with missing temperature data (n=1 for 2005), data from

6 the nearest station with similar depth was used. For stations with incomplete

7 temperature profiles (n=1 for 2005), temperatures were linearly interpolated between

8 depths.

9 Zooplankton data

10 Determination of main prey taxa

11 Juvenile pollock (<100 mm TL) collected from both surface (2005 and 2010) and

12 midwater (2010) tows were used in the analysis to characterize diets across the EBS

13 shelf in two contrasting years (n=26 stations in 2005, n=47 stations in 2010 [n=16

14 surface tows, n=31 midwater tows]). Stomach content analyses followed standard

15 methods as described in [19]. Individual prey taxa were allocated a proportional

16 contribution to total stomach contents, as % volume, of each prey taxon to the diet. To

17 compute overall average diet composition, contributions were weighted by the CPUE of

18 juvenile pollock at each station and averaged across stations. All prey taxa of juvenile

19 pollock that cumulatively accounted for at least 90% of the diet by volume and

20 individually accounted for at least 2% of the diet by volume were included in the

21 bioenergetics and IBM models (Table 1). Main prey taxa from either year were included

22 in models for both years for comparing growth across years.

136 137

1 Zooplankton abundance

2 Water-column abundances of small and large zooplankton taxa were estimated from

3 Juday and bongo net samples, respectively, as described in [14]. Small zooplankton

4 representing main prey taxa included Acartia clausi, Acartia spp. (2010 only),

5 Centropages abdominalis, and Pseudocalanus sp. Large zooplankton included Calanus

6 marshallae, Eucalanus bungii, Limacina helicina, Neocalanus cristatus, N. plumchrus

7 (2005 only), Oikopleura sp., Thysanoessa inermis, T. inspinata, and T. raschii.

8 Zooplankton biomass

9 Total sample weights (g) of taxa collected from the Juday net were computed from wet

10 weight tables [20]. Densities (g•m-3) of taxa collected from the bongo net were

11 measured during sample processing at the University of Alaska Fairbanks (2005; [21])

12 and NOAA/NMFS/Alaska Fisheries Science Center (2010). The year-specific average

13 biomass of individuals for the main prey taxa was calculated by dividing the sum of the

14 biomass of all specimens weighed (i.e., subsample) by the total number of specimens

15 subsampled in a given year (Table S1).

16 Zooplankton energy density

17 Taxa-specific energy density (ED; kJ•g-1) values obtained from available zooplankton

18 collections from the EBS during 2004 (warm; no ED data available from 2005) and 2010

19 (cold) were used to estimate average ED values during warm and cold conditions for the

20 main prey taxa. For taxa lacking sufficient information to estimate separate ED values

21 (n=5), a single estimate was used in both years (Table S1). In these cases, only

22 differences in abundance and biomass contributed to differences in average prey energy

137 138

1 between years in the models. A biomass-weighted mean prey ED was calculated for

2 each station and used as input to the bioenergetics model. At each station, the biomass

3 of individual taxa was divided by the total prey biomass, multiplied by the taxa-specific

4 energy density for each year, and summed across all taxa present at a given station.

5 Estimates of ED and % lipid were available for several copepod species (C.

6 marshallae, N. cristatus, and N. plumchrus/flemingeri) from 2010 (see [22] for details on

7 the biochemical processing). A linear regression was developed to predict species

8 specific ED (i ) from % lipid values for other copepod species and/or climate conditions

9 (Table S1), such that:

2 10 i  Li i , where  i ~ N(0, ) Eq. 2

11 where  and  represent the intercept and slope of the regression, respectively, Li is the

12 lipid composition (%) of the individual copepod sample i and i is a residual. The

13 residuals, i, are assumed to be independent and normally distributed with mean 0 and

14 variance 2 ( = 19.3, p = 0.02;  = 0.41, p = 0.07; R2=0.98).

15 Zooplankton vertical profiles

16 To account for diel vertical migrations, taxa-specific vertical profiles for day and night

17 were developed for all main prey taxa as input for the IBM. Vertical profiles were based

18 on summer MOCNESS surveys that provided depth-stratified abundance estimates.

19 MOCNESS data were available for 2004 (warm) and 2009 (cold); these vertical profiles

20 were applied to late-summer model runs for 2005 and 2010, respectively, assuming that

21 the vertical behavior of zooplankton taxa is conserved seasonally and across years

22 within similar oceanographic conditions. To assess the effect of this assumption, a

138 139

1 sensitivity analysis was conducted using constant abundances by depth (see ‘IBM

2 sensitivity analyses’ below).

3 In 2004, vertically stratified MOCNESS samples were collected at 5 daytime and

4 42 nighttime stations over the EBS shelf [21]. In 2009, 7 daytime and 22 nighttime

5 stations were sampled (A. Pinchuk, unpubl. data). Daytime extended from

6 approximately 07:00 (sunrise) to 23:30 (sunset) Alaska Daylight Savings Time during the

7 sampling periods; stations sampled during crepuscular periods were excluded from the

8 analysis. The depth increments of the MOCNESS varied depending on water depth;

9 therefore, data were binned to the finest resolution available (i.e., 5-20 m). Zooplankton

10 abundance was assumed to be uniform within sampling depths and averaged across all

11 daytime and nighttime tows within a given year to obtain four vertical profiles for each

12 taxon (day vs. night, 2004 vs. 2009). Centropages abdominalis were not collected by the

13 MOCNESS and a uniform distribution throughout the water column was applied for both

14 years because their distribution during the 2005 and 2010 BASIS surveys was

15 predominantly at shallow, well-mixed stations of the inner domain. Oikopleura sp. did

16 not occur in daytime tows in 2004; therefore, the 2009 daytime vertical distribution was

17 applied for both 2005 and 2010 model runs. Thysanoessa inspinata were rarely

18 collected by the MOCNESS (n=1 for 2005; n=3 for 2010), therefore an average vertical

19 profile based on all Thysanoessa spp. was applied.

20 4.4.4 Bioenergetics model 21 A bioenergetics model was used to estimate spatially explicit maximum growth

22 potential of juvenile pollock. We used the broadly applied Wisconsin bioenergetics

139 140

1 modeling approach [23,24] that has been adapted for walleye pollock including

2 appropriate model validation ([16, 25]; Table S2). The model estimates temperature-

3 and weight-specific maximum daily (d) consumption for an individual fish at station k in

-1 -1 4 year t (Cmaxd,kt; g•g •d ) as:

 5 Cmaxd,kt  Wd 1 f (Td,kt ) Eq. 3

6 where C maxd ,kt is parameterized from independent laboratory observations of  7 consumption rates for the species, absent competitor or predator interference, and is

8 assumed to scale exponentially with fish weight (W) according to  and  (the allometric

9 intercept and slope of consumption) and thermal experience according to the

10 temperature scaling function f(T) (Table S2).

-1 -1 11 Realized individual daily consumption rates (Cd ,kt ; g•g •d ) based on in situ fish

12 are typically much lower than Cmaxd,kt because inter- and intra-species competition,

13 mismatched prey phenology or distributions, and predator avoidance behaviors by prey

14 species often limit capture and consumption rates [16,26]. The ratio of realized

15 consumption to maximum consumption (i.e., h = Cd,kt Cmaxd,kt ), or the mean relative

16 foraging rate, is a measure of in situ foraging efficiency. The rate h can be estimated

17 using field observations of growth or it can be set to a specific value and used to predict

18 daily growth (Gd,kt ) using the mass balance equation where growth is the difference

19 between energy consumed (Cd,kt ) and energy lost to metabolism and waste

20 ([ | Cd ,kt ,Wd 1,Td ,kt ] ), such that:

21 Gd,kt  (Cd,kt [ | Cd,kt ,Wd1,Td,kt ]) kt Eq. 4

140 141

-1 -1 1 where Gd,kt is the estimated daily specific growth (g•g •d ), Cd,kt is realized

2 consumption (Cd,kt = h ×Cmaxd,kt ), Wd1 is the weight of an individual fish at the start of

3 the simulation day d, Td,kt is the water temperature on simulation day d, and zkt is the

4 ratio of predator energy to prey energy density and is used to convert consumed

5 biomass of prey into predator biomass (for more information see [26]). We used station-

6 specific energy densities for prey (k ) but annual mean predator energy density ( v t) for

7 year t because predator information was not available at all stations.  8 Energy density values for the main prey taxa were used to derive mean station-

9 specific (k) available prey energy density for both years (w kt); diet composition was

10 assumed to be proportional to the relative biomass of zooplankton prey at each station.

11 Individual fish energy density (vi) was determined using biochemical processing (see

12 [22]). At stations where sufficient numbers of juvenile pollock were collected (n=91 in

13 2005 and n=12 in 2010), 2-8 fish were selected to represent the size range at each

14 station. Station-specific mean energy density in a given year ( v kt) was weighted by CPUE

15 and the number of fish processed at each station to calculate the average fish energy

16 density by year ( v t).

17 We ran the model for a single simulation day (i.e., d=1) using base scenario input

18 parameter values (Table 2; see also [16] Tables I and II) that were kept constant across

19 stations and years (i.e., W =2.5 and  =1), were constant across stations but varied by

20 year (i.e., vt ), or varied by station and year (i.e., Tkt and kt ). Because the model is size-

21 specific, running the model for a single simulation day minimized compound errors that

22 can accumulate over multiple simulation days when predicting growth and allowed for a

141 142

1 comparative index of growth across stations. Keeping fish starting weights (W) constant

2 allowed us to evaluate spatial effects of changes in the other parameters; setting =1

3 implies that growth was constrained by physiological processes, but not by prey

4 consumption, hence we evaluated variability in maximum growth potential. Annual

-1 5 average fish energy density was applied across stations in each year (v2005= 3.92 kJ•g ;

-1 6 v2010= 5.29 kJ•g ).

7 Bioenergetics sensitivity analyses

8 Individual input parameters were increased and decreased by 1 standard deviation (SD)

9 and the change in growth relative to maximum predicted growth under the base

10 scenario was recorded. A pooled SD was calculated across stations after removing the

11 annual means. Relative foraging rate ( ) was held at 1 for all sensitivity model runs in

12 order to compare the relative effect of other parameters on maximum growth potential.

13 Station-specific parameters (i.e., Tkt and kt ) were varied to evaluate the relative

14 effect on predicted growth and to examine resulting changes in spatially explicit growth

15 patterns in each year. To evaluate the effect of variability in fish starting weight and

16 energy density (W and vt , respectively) on estimated growth in 2005 and 2010, we

17 used Monte Carlo simulations at a representative station (see Figure S1). A single station  18 was used because mean fish weight and energy density input values did not vary across

19 stations in the model due to data limitations; hence the spatial pattern in estimated

20 growth is not affected by varying these values by a constant amount. The model was run

21 1000 times using parameter values drawn at random from a normal distribution with

142 143

1 the observed mean and SD for each parameter. The resulting distribution of predicted

2 maximum growth potential was examined.

3 Mechanistic individual-based model

4 A mechanistic, depth-stratified IBM was used to predict average growth (g•g-1•d-1) and

5 depth (m) of 100 simulated juvenile pollock by station. The details of the IBM and model

6 validation are described in [17,28]. The IBM was reparameterized for juvenile pollock

7 and forced with input data for water column temperatures and prey availability in 1 m

8 discrete depth bins. Prey abundance (#•m-3) was allocated into depth bins according to

9 vertical profiles of zooplankton distribution and scaled to station depth.

10 The IBM used a mechanistic prey selection component that simulated the

11 feeding behavior of juvenile pollock on zooplankton. The species composition of main

12 prey taxa was based on observations; stage-specific length and width estimates were

13 based on literature values or voucher collections from the EBS (Table S1). Optimal prey

14 size was estimated to be 5-8% of fish length based on larval Atlantic cod research

15 [29,30]; juvenile pollock are predicted to have nearly 100% capture success for prey

16 smaller than 5% of fish length, while the probability of capture success decreases with

17 larger prey [17]. The simulated feeding ecology depended on juvenile pollock

18 development (e.g., swimming speed, gape width, eye sensitivity) and vertical migratory

19 behavior, prey densities and size, as well as light and physical oceanographic conditions

20 (for details see [17]). Gape width was calculated as a function of fish size; conversion

21 between length and weight followed [31]. Juvenile feeding processes were modeled

22 with light-dependent prey encounter rates and prey-capture success (see [29]).

143 144

1 Vertical migratory behavior was modeled assuming that juvenile pollock would

2 seek deeper depths to avoid visual predation risk as long as ingestion rates would

3 sustain metabolism and growth. If not, juvenile fish would seek the euphotic zone

4 where light enhances feeding success, but also increases predation risk. Prey

5 distributions switched between daytime to nighttime profiles when the light level (i.e.,

6 irradiance) reached 1 mol•m-2•s-1 [28]. The cost of vertical migration was included as a

7 maximum of 10% of standard metabolic rates if the fish swims up or down at its

8 maximum velocity, and scaled proportionally for shorter vertical displacements.

9 Swimming velocity was a function of juvenile fish size [32].

10 Gut fullness was estimated based on the amount of prey biomass that was

11 ingested and digested per time step (1 hour) according to the feeding module. Prey

12 biomass flowing through the alimentary system supplied growth up to a maximum

13 growth potential (Cmax; [25]), and standard metabolic cost, egestion, excretion, and

14 specific dynamic action [16] were subtracted. Both maximum growth and metabolic

15 costs were functions of fish weight and water temperature.

16 For all base model scenarios, the starting weight of the fish was held constant

17 across stations, while zooplankton abundance and vertical distribution varied according

18 to observations. Fish starting weight was 2.5 g ± 30% assuming a random uniform

19 distribution around the mean. Year-specific vertical profiles (day and night) for the main

20 prey taxa and station-specific temperature and prey abundance profiles were applied.

21 The model scenarios were run for 72 hours, but only the last 24 hours of the simulations

144 145

1 were used for the analysis to avoid the early part of the simulations that may be unduly

2 influenced by random initial conditions.

3 IBM sensitivity analyses

4 Fish starting weights and the vertical prey profiles were varied and resulting growth and

5 average depth predictions were compared to values under the base model scenario (see

6 [28] for sensitivity of the IBM model to variability in other parameters). To evaluate the

7 effect of fish size separately from the effects of environmental controls, estimated

8 growth based on fish starting weights of 2.0 g ± 30% was compared to the base scenario

9 (2.5 g ± 30%), encompassing the mean weight of juvenile pollock from the BASIS surveys

10 in 2005 (1.97 ± 0.93 g, mean ± SD) and 2010 (2.39 ± 0.94 g, mean ± SD). To test the

11 effect of vertical distributions and diel migrations of prey taxa, model runs assuming a

12 uniform distribution of prey with depth were compared to the base scenario,

13 highlighting the effects of non-uniform zooplankton distribution and diel vertical

14 migrations on juvenile pollock prey selection.

15 4.5 Results

16 4.5.1 Field observations 17 Juvenile pollock abundance

18 Juvenile pollock abundance and distribution had distinct spatial patterns in the surface

19 layer between warm and cold years, with a more northerly distribution in warm years.

20 Specifically, during warm late-summer conditions of 2005 juvenile pollock were

21 distributed over a broad extent of the middle and outer domain, while in the cooler late

22 summer of 2010 fish were concentrated over small regions of the southern shelf and

145 146

1 outer domain (Figure 1, a and b). Abundance also varied between years with higher

2 mean CPUE observed in 2005 as compared to 2010 (CPUE = 0.08 fish•m-2 vs. 0.001

3 fish•m-2, respectively) at positive catch stations.

4 Water temperature

5 The average water temperature in the upper 30 m of the water column during the BASIS

6 survey was 8.8ºC in 2005 and 7.6ºC in 2010 (Figure S2, a and b), while the average

7 temperature below 40 m was 4.5ºC in 2005 and 2.9ºC in 2010 (Figure S2, c and d). The

8 warmest surface temperatures occurred in nearshore waters, although 2005 had warm

9 temperatures over much of the southern shelf. Bottom temperatures reflected the

10 extent of the cold pool (waters <2ºC), which was limited to the northern portion of the

11 study area in 2005 and covered much of the shelf in 2010.

12 Zooplankton

13 Main prey taxa

14 Diets of juvenile pollock shifted from smaller copepod species in the warmer 2005

15 summer season (e.g., Pseudocalanus sp.) to larger species in the cooler 2010 summer

16 season (e.g., N. cristatus). Several large zooplankton species were present in the diets

17 across years, including L. helicina, which was the predominant prey item in both years,

18 as well as C. marshallae and T. raschii. In 2010, the main prey taxa of juvenile pollock

19 collected in surface tows were similar to those from midwater tows, with the exception

20 of E. bungii accounting for 0% and 3% of surface and midwater tows, respectively.

21 Eucalanus bungii was included in further analyses because it represented approximately

22 3% of combined diets by volume (Table 1).

146 147

1 Zooplankton abundance

2 Changes in juvenile pollock diet composition reflect spatial and temporal variability in

3 zooplankton species composition and availability. In 2005, the abundance of available

4 prey was highest in the inner domain and decreased towards the outer domain and

5 northern Bering Sea. The abundance of prey in 2010 was greater in the inner domain; in

6 the southern region of the shelf, abundances decreased towards to the middle and

7 outer domains (Figure 1, c and d). The lowest abundance of zooplankton occurred in

8 areas corresponding to higher concentrations of juvenile pollock predators. The total

9 abundance of zooplankton within the optimal prey size range for 65 mm SL juvenile

10 pollock (species with mean length within 5-8% of fish length) was higher in the

11 northwest region of the study area and over the southern shelf in the outer domain in

12 2005, with lesser overlap with juvenile pollock. In 2010, optimal prey was located across

13 the middle and outer domains with highest abundances in the southern region,

14 mirroring the distribution of juvenile pollock (Figure 1, c and d). Spatial patterns of

15 zooplankton abundance accounting for all taxa <8% of fish length reflected total

16 abundance patterns in both years, indicating that areas of highest zooplankton

17 abundance are driven by small (<5% of fish length) zooplankton taxa.

18 Zooplankton energy density

19 In 2005, available prey energy was highest in the northwest region of the shelf, with low

20 prey energy over most of the shelf south of 60 °N (Figure 1e) where juvenile pollock

21 abundances were higher. In contrast, prey energy was very high across much of the

22 southern shelf in 2010 (Figure 1f), particularly within the cold pool, where juvenile

147 148

1 pollock were more abundant. Spatial patterns in prey energy were similar to spatial

2 patterns of abundance for optimal prey size classes (Figure 1, c and d) because highest

3 energy prey taxa are within 5-8% of fish length.

4 4.5.2 Bioenergetics model 5 Differences in the spatial pattern of maximum growth potential (g•g-1•d-1) of juvenile

6 pollock occurred between a warm and cold year in the EBS (Figure 2, a and b). In 2005,

7 growth potential was highest in the northwest region of the shelf (north of 60 °N) and

8 lowest in the inner domain with one station having negative growth. Gradients in

9 growth potential, from low to high, occurred from the inner to outer domains and from

10 southern to northern regions of the shelf (Figure 2a). In 2010, growth was positive at all

11 stations with highest growth potential over the southern shelf and lower growth

12 predicted in the northeast region (Figure 2b).

13 Bioenergetics sensitivity analyses

14 Effect of water temperature and prey energy density

15 In 2005, increasing temperatures by 1 SD (Figure 2c) resulted in areas of decreased

16 predicted growth at shallow inner domain and southern shelf stations where water

17 temperatures already approached thermal thresholds. Growth could not be estimated

18 at one inner domain station because the increased temperature exceeded 15 ºC, the

19 maximum temperature for consumption (Tcm) in the model. Decreasing water

20 temperatures, resulting in increased growth, had the greatest effect in the same areas

21 (not shown) because temperature-dependent control of growth is magnified where

22 temperatures are close to thermal thresholds. In 2010, the effect of increasing water

148 149

1 temperatures was an order of magnitude less than in 2005 (Table 3), but the spatial

2 patterns were similar with shallow stations in the inner domain being most sensitive, as

3 well as a small area in the outer domain (Figure 2d). Increasing available prey energy

4 resulted in increased predicted growth rates across the region in 2005 (Figure 2e), with

5 weaker effects in the inner domain and northwest region. In 2010, increased prey

6 energy also resulted in elevated growth rates, but the magnitude of change was much

7 lower than in 2005 and the spatial pattern differed; stronger effects occurred in the

8 inner domain and southern region of the outer domain (Figure 2f).

9 Predicted maximum growth potential generally increases with temperature and

10 prey energy until temperature-dependent controls limit growth (Figure 3). Predicted

11 growth is negative when available prey energy cannot meet metabolic demands under

12 increased temperatures. Water temperatures were warmer in 2005, therefore juvenile

13 pollock experienced conditions at or near their metabolic threshold at some stations.

14 Colder water temperatures and higher available prey energy in 2010 resulted in better

15 growing conditions over the shelf.

16 Effect of fish starting weight and fish energy density

17 Increasing fish starting weight resulted in lower predicted growth rates in both years

18 because larger fish have higher metabolic demands (Table 3). Increasing fish energy

19 density had a variable effect across stations in 2005 (not shown). In general, the effect

20 of varying fish energy is dependent on initial fish energy and the relative available prey

21 energy at each station. In 2010, increasing fish energy density resulted in lower

22 predicted growth rates across stations when available prey energy was held constant.

149 150

1 Variability in fish starting weight resulted in a broader distribution of predicted

2 growth rates (2005: 0.002 – 0.109; 2010: 0.017 – 0.170 g•g-1•d-1) than variability in fish

3 energy (2005: 0.007 – 0.013; 2010: 0.022 – 0.036 g•g-1•d-1) from Monte Carlo

4 simulations, indicating that the model was more sensitive to inputs of fish weight. The

5 simulated mean predicted growth rates, when varying fish starting weight or fish

6 energy, were lower and less variable for 2005 (0.012  0.009 [mean  SD] for varying W;

7 0.010  0.001 for varying fish energy) than for 2010 (0.029 0.012 for varying W; 0.027

8  0.002 for varying fish energy).

9 4.5.3 Mechanistic individual-based model 10 Predicted mean growth rates from the IBM were 30% (2005) and 46% (2010) lower than

11 maximum growth potential from the bioenergetics model (Tables 3 and 4) as foraging

12 rates are restricted in the IBM based on stomach fullness and the prey selection module

13 (i.e., capture success). The reduction in growth was greater in 2010, resulting in similar

14 predicted growth rates from the IBM in 2005 and 2010. In addition, predicted growth

15 rates from the IBM have a narrower range than maximum growth potential from the

16 bioenergetics model.

17 In 2005, growth was positive across the region with moderate growth predicted

18 across the southern shelf. North of 60 °N, predicted growth rates decreased from the

19 inner to outer domain (Figure 4a). The average depth (m) of juvenile pollock was 44 m

20 (Table 4), with shallower distributions in the northeast region and deeper distributions

21 in the southern region of the outer domain (Figure 4b). In 2010, growth was positive

22 across the region, with highest predicted growth in the inner domain and areas of lower

150 151

1 growth in the middle domain (Figure 4c). The spatial patterns of average depth of

2 juvenile pollock (Figure 4d) mirrored those of 2005 with a slightly deeper average depth

3 of 47 m (Table 4).

4 IBM sensitivity analyses

5 The effect of smaller fish starting weights on predicted growth was positive across the

6 region, with stronger effects in 2005 than 2010 (Table 4). Similarly, effect strengths

7 varied spatially in both years with areas of higher predicted growth in the middle

8 domain (Figure 4, e and g). In 2005, smaller starting fish weights resulted in shallower

9 depth distributions across the region (mean=-2.6 m; Table 4), with much shallower

10 depths at two stations in the middle domain (Figure 4f). The average change in depth

11 distribution was similar in 2010 (mean=-2.4 m; Table 4), but spatially more variable than

12 in 2005 (Figure 4h).

13 Applying uniform vertical distributions to prey taxa had variable effects on

14 predicted growth rates in both years, with similarly small effect strengths (Table 4).

15 Under uniform prey distributions, modeled fish may move vertically in response to other

16 cues (i.e., predation risk, thermal boundaries) regardless of diel patterns. In 2005,

17 uniform distributions resulted in increased predicted growth rates at several stations in

18 the northern-most region of the shelf (Figure 4i). While the average depth of juvenile

19 pollock was 2.1 m deeper across the region, fish at some of the northern-most stations

20 had shallower depths (Figure 4j). In 2010, strongest effects on growth were observed in

21 the middle domain of the southern shelf, with high spatial variability (Figure 4k).

22 Changes in the depth of fish in response to uniform prey distributions mirrored spatial

151 152

1 patterns in growth effects; stations showing deeper mean depths also resulted in a

2 decrease in growth and vice versa (Figure 4l).

3 4.5.4 Spatial comparison of bioenergetics- and IBM-predicted growth 4 Predicted growth rates from the IBM were within the range of maximum growth

5 potential from the bioenergetics model, but spatial patterns varied due to differences in

6 input parameters of each model. In both years, the bioenergetics model predicted

7 higher growth rates than the IBM over the middle and outer domains. The greatest

8 difference occurred in the northwest region of the shelf in 2005 (Figure 5a) and over the

9 southern region of the middle domain in 2010 (Figure 5b). The IBM predicted higher

10 growth in the shallow, well-mixed inner domain in both years.

11 4.6 Discussion 12 This study demonstrates that warm and cold conditions in the EBS lead to spatial

13 differences in zooplankton species composition, energy content, and abundance, which

14 subsequently affect the feeding ecology and growth of juvenile pollock. Particularly,

15 prey distribution and quality in combination with water temperatures create spatial

16 patterns of increased growth potential (‘hot spots’) that vary with climate conditions.

17 Spatial heterogeneity in growth conditions results from a combination of prey quality

18 and quantity, water temperature, and metabolic costs, which may contribute to size-

19 dependent fish survival and subsequent annual variability in recruitment. We provide

20 evidence that a spatial mismatch between juvenile pollock and growth ‘hot spots’ in

21 2005 is the mechanism that contributed to poor recruitment to age-1 while a higher

22 degree of overlap in 2010 resulted in 42% greater [33] recruitment to age-1.

152 153

1 In the EBS, changes in oceanographic conditions can impact larval and juvenile

2 fish distributions through front formation [34] and subsequent changes in drift

3 trajectories [35]. The resultant variability in fish distributions relative to their prey

4 during late summer and fall may be particularly important because the time period after

5 the completion of larval development and before the onset of winter has been

6 identified as a critical period for energy storage in juvenile pollock [22]. As the spatial

7 distribution of fish, including spawning locations of adult walleye pollock, and

8 zooplankton vary under alternate climate conditions, so do patterns in juvenile fish

9 growth and recruitment success (Figure 6). Here, we find support for the argument that

10 warm years produce smaller, less energy-rich prey and that this reduced prey quality, in

11 combination with higher metabolic demands, results in lower growth of juvenile pollock.

12 Conversely, cold years produce larger, more energy-rich prey which, when combined

13 with lower metabolic demands, are favorable for juvenile pollock growth and survival.

14 Thus, mechanisms responsible for controlling growing conditions during the critical pre-

15 winter period can be linked to variability in recruitment.

16 Projected declines in walleye pollock recruitment under changing climate

17 conditions [11] do not account for adaptive behaviors or changes to phenology that

18 could enable fish to maintain higher growth rates. The sensitivity analyses helped to

19 identify when and where favorable growth conditions may occur under alternate

20 climate conditions. In the bioenergetics model, varying fish size had a stronger effect on

21 growth potential than changes in initial fish energy density. Larger fish have greater

22 capacity for growth due to increased gape size, which allows them to take advantage of

153 154

1 larger, more energy rich prey resources (e.g., euphausiids) prior to winter. The

2 sensitivity analysis of increasing water temperatures showed weaker effects in the cold

3 year of 2010 because fish had a broader range of temperatures over which growth

4 potential was relatively high (Figure 3), including warmer surface waters and a colder

5 refuge in deeper waters that allows fish to conserve energy and avoid predation. In

6 2005, fish were near thermal limits based on temperature-dependent functions in the

7 bioenergetics model; hence further increases in temperature are predicted to result in

8 negative growth. Increasing available prey energy also had a stronger effect in the warm

9 year of 2005 because metabolic demands were greater and mean prey energy density

10 was lower than in 2010.

11 The relative foraging rate was held constant at =1 across all bioenergetics

12 model scenarios, but lower values would better reflect realistic foraging rates and could

13 exacerbate thermal constraints on growth. To maintain positive growth rates at half of

14 all the stations required relative foraging rates of =0.71 in 2005 and =0.57 in 2010.

15 These values correspond to a 29% and 43% reduction in achieved growth relative to

16 maximum growth potential and are similar to the mean differences between growth

17 rates in the bioenergetics and IBM models (i.e., 30% in 2005 and 46% in 2010), providing

18 support of model agreement. A higher relative foraging rate was required in 2005 in

19 order to achieve positive growth at half of all stations, similar to results based on larger

20 juvenile and adult walleye pollock [25], indicating that juvenile pollock growth was more

21 prey limited and constrained by temperature in 2005 than in 2010. Thus, a greater

22 reduction in both achieved growth from the IBM relative to maximum growth potential

154 155

1 and relative foraging rates was observed in 2010 compared to 2005. In 2010,

2 zooplankton abundance was lowest in areas with higher concentrations of juvenile

3 pollock predators, potentially indicating prey limitation. While our study was not

4 designed to explicitly test this question, other research demonstrates that pollock do

5 not have strong top-down control of euphausiid abundance in the eastern Bering Sea (P.

6 Ressler, pers. commun.).

7 The vertical behavior of modeled juvenile pollock in the IBM moderated

8 predicted growth rates leading to differences across domains based on stratification.

9 Smaller (i.e., younger) fish were predicted to move shallower in the water column to

10 improve prey detection, which is dependent on eye development and light availability.

11 Moving into the surface layer also exposed juvenile pollock to higher predation risk

12 because of the stronger light intensity. In the middle and outer domains, once sufficient

13 growth was attained, fish were predicted to move deeper to seek refuge from

14 predation. While the models were run at all stations in both years, observed juvenile

15 pollock abundances were concentrated over the middle and outer domains in 2005 and

16 over small regions of the southern shelf and outer domain in 2010. Few fish were

17 observed in the well-mixed inner domain, possibly due to reduced growth potential

18 based on available prey energy or lack of stratification and predation refuge in deeper

19 waters. Additionally, the inner front, which delineates the stratified middle domain from

20 the well-mixed inner domain [34], may act as a barrier to juvenile pollock distribution

21 [36].

155 156

1 Spatial patterns in juvenile pollock growth differed between models; these

2 differences elucidate underlying mechanisms in feeding potential and ultimately the

3 possible causes for growth ‘hot spots’ and variability in recruitment success between

4 warm and cold climate conditions. The bioenergetics model uses biomass-weighted

5 mean energy density of available prey, assuming fish feed proportional to what is

6 available in the environment. The IBM is length-based and growth is dependent on

7 available prey resources, light conditions, metabolism, development of the fish, and fish

8 behavior. In the middle and outer domains where the water column is stratified, the

9 bioenergetics model predicted higher growth than the IBM; the bioenergetics model

10 allowed fish to feed at maximum consumption while the IBM indicated that fish moved

11 deeper in the water column to conserve energy or avoid predation. In the inner domain,

12 the IBM predicted higher growth; here juvenile pollock may opt to take advantage of

13 available prey and warmer water temperatures to maximize growth because predator

14 avoidance in deeper waters was not an option.

15 Comparing the bioenergetics model and the IBM provided insights that could not

16 be gained by either approach alone. For example, the bioenergetics model highlights the

17 importance of differences in prey energy, a metric not included in the IBM, in

18 determining spatial patterns of growth. On the other hand, the mechanistic feeding

19 behavior implemented in the IBM highlights the role of prey size composition, the

20 vertical distribution of prey, and the tradeoff between predator avoidance and

21 maximizing growth. In practice, data requirements may limit the applicability of the IBM,

22 whereas the bioenergetics model can be applied when less information on prey

156 157

1 resources is available. Future research could benefit from including information on prey

2 energy into IBMs to disentangle not only the importance of prey species and size

3 composition, spatial distribution and abundance, but also prey quality.

4 Warm temperature conditions are predicted to result in reduced prey quality

5 and low energy density of juvenile pollock in late summer [9,13]. Warmer water

6 temperatures are associated with decreased growth [this study], resulting in lower

7 overwinter survival and recruitment to age-1 [33]. The warm years of 2002-2005 had

8 67% lower average recruitment to age-1 relative to the cold years of 2008-2010,

9 although variability during the cold years was very high with strong year classes in 2008

10 and 2010 separated by a weak 2009 cohort [33]. These findings agree with projected

11 declines in recruitment of age-1 walleye pollock [11] under increased summer sea

12 surface temperatures of 2°C predicted by 2050 [37]. Our results corroborate these

13 previous studies and suggest that climate-driven increases in water temperature will

14 have the largest effect on recruitment during anomalously warm years.

15 This study provides evidence that climate-driven changes in prey dynamics can

16 have ecosystem-level consequences via bottom-up control of fish populations in sub-

17 arctic marine ecosystems. This work has improved our understanding of the

18 mechanisms behind recruitment variability, in particular the underlying spatial patterns

19 that drive relationships between prey availability, water temperature, growth, and

20 survival. Our findings inform ongoing discussions of climate effects on predator-prey

21 interactions and recruitment success of marine fishes.

22 4.7 Acknowledgements

157 158

1 We thank NOAA’s Bering Aleutian Salmon International Survey (BASIS) program for data

2 collection and database management as well as the officers and crew of the R/V Oscar

3 Dyson and F/V Sea Storm. The BASIS program is funded in part by the North Pacific

4 Anadromous Fish Commission. MOCNESS data were provided from the PROBES (2004)

5 and BEST-BSIERP (2009) surveys. This is NPRB publication #444 and BEST-BSIERP Bering

6 Sea Project publication #111. T.K. was supported by the Norwegian Research Council

7 project SWIM #13375.

8 4.8 References 9 1. Cushing DH (1990) Plankton production and year-class strength in fish

10 populations – an update of the match mismatch hypothesis. Advances in Marine

11 Biology 26: 249-293.

12 2. Durant JM, Hjermann DO, Ottersen G, Stenseth NC (2007) Climate and the match

13 or mismatch between predator requirements and resource availability. Climate

14 Research 33: 271-283.

15 3. Edwards M, Richardson AJ (2004) Impact of climate change on marine pelagic

16 phenology and trophic mismatch. Nature 430: 881-884.

17 4. Kristiansen T, Drinkwater KF, Lough RG, Sundby S (2011) Recruitment

18 variability in North Atlantic cod and match-mismatch dynamics. PLOS ONE

19 6(3): e17456. doi:10.1371/journal.pone.0017456.

20 5. Hunt GL Jr, Stabeno PJ, Walters G, Sinclair E, Brodeur RD, Napp JM, Bond NA

21 (2002) Climate change and control of the southeastern Bering Sea pelagic

22 ecosystem. Deep-Sea Research II 49: 5821-5853.

158 159

1 6. Sogard SM, Olla BL (2000) Endurance of simulated winter conditions by age-0

2 walleye pollock: effects of body size, water temperature and energy storage.

3 Journal of Fish Biology 56: 1-21.

4 7. Litzow MA, Bailey KM, Prahl FG, Heintz R (2006) Climate regime shifts and

5 reorganization of fish communities: the essential fatty acid limitation hypothesis.

6 Marine Ecology Progress Series 315: 1–11.

7 8. Paul AJ, Paul JM (1998) Comparisons of whole body energy content of captive

8 fasting age zero Alaskan Pacific herring (Clupea pallasi Valenciennes) and cohorts

9 over-wintering in nature. Journal of Experimental Marine Biology and Ecology.

10 226: 75–86.

11 9. Hunt GL Jr, Coyle KO, Eisner L, Farley EV, Heintz R, Mueter FJ, Napp JM, Overland

12 JE, Ressler PH, Salo S, Stabeno PJ (2011) Climate impacts on eastern Bering Sea

13 foodwebs: A synthesis of new data and an assessment of the Oscillating Control

14 Hypothesis. ICES Journal of Marine Sciences 68(6): 1230-1243.

15 10. Beaugrand G, Brander KM, Lindley JA, Souissi S, Reid PC (2003) Plankton effect

16 on cod recruitment in the North Sea. Nature 426: 661-664.

17 11. Mueter FJ, Bond NA, Ianelli JN, Hollowed AB (2011) Expected declines in

18 recruitment of walleye pollock (Theragra chalcogramma) in the eastern Bering

19 Sea under future climate change. ICES Journal of Marine Science 68: 1284–1296.

20 12. Stabeno PJ, Kachel NB, Moore SE, Napp JM. Sigler M, Yamaguchi A, Zerbini AN

21 (2012) Comparison of warm and cold years on the southeastern Bering Sea shelf

22 and some implications for the ecosystem. Deep-Sea Research II 65-70: 31-45.

159 160

1 13. Heintz RA, Siddon EC, Farley EV Jr, Napp JM (In press) Correlation between

2 recruitment and fall condition of age-0 walleye pollock (Theragra

3 chalcogramma) from the eastern Bering Sea under varying climate conditions.

4 Deep-Sea Research II.

5 14. Coyle KO, Eisner LB, Mueter FJ, Pinchuk AI, Janout MA, Cieciel KD, Farley EV,

6 Andrews AG (2011) Climate change in the southeastern Bering Sea: impacts on

7 pollock stocks and implications for the Oscillating Control Hypothesis. Fisheries

8 Oceanography 20(2): 139-156.

9 15. Farley EV, Murphy JM, Wing BW, Moss JH, Middleton A (2005) Distribution,

10 migration pathways, and size of western Alaska juvenile salmon along the

11 eastern Bering Sea shelf. Alaska Fishery Research Bulletin 11: 15-26.

12 16. Ciannelli L, Brodeur RD, Buckley TW (1998) Development and application of a

13 bioenergetics model for juvenile walleye pollock. Journal of Fish Biology 52: 879-

14 898.

15 17. Kristiansen T, Lough RG, Werner FE, Broughton EA, Buckley LJ (2009) Individual-

16 based modeling of feeding ecology and prey selection of larval cod on Georges

17 Bank. Marine Ecolology Progress Series 376: 227-243.

18 18. Buchheister A, Wilson MT (2005) Shrinkage correction and length conversion

19 equations for Theragra chalcogramma, Mallotus villosus, and Thaleichthys

20 pacificus. Journal of Fish Biology 67: 541-548.

21 19. Moss JH, Farley EV Jr, Feldmann AM, Ianelli JN (2009) Spatial distribution,

22 energetic status, and food habits of eastern Bering Sea age-0 walleye pollock.

160 161

1 Transactions of the American Fisheries Society 138: 497-505.

2 20. Volkov AF, Efimkin AY, Kuznetsova NA (2007) Characteristics of the plankton

3 community of the Bering Sea and several regions of the North Pacific in 2002 –

4 2006. Izvestiya Tikhookeanskiy Nauchno-Issledovatelskiy Institut Rybnogo

5 Khozyaystva I Okeanografiy 151: 338–364 (in Russian).

6 21. Coyle KO, Pinchuk AI, Eisner LB, Napp JM (2008) Zooplankton species

7 composition, abundance and biomass on the eastern Bering Sea shelf during

8 summer: the potential role of water column stability and nutrients in structuring

9 the zooplankton community. Deep Sea Research II 55: 1755–1791.

10 22. Siddon EC, Heintz RA, Mueter FJ (2013) Conceptual model of energy allocation in

11 walleye pollock (Theragra chalcogramma) from age-0 to age-1 in the

12 southeastern Bering Sea. Deep-Sea Research II.

13 http://dx.doi.org/10.1016/j.dsr2.2012.12.007.

14 23. Kitchell JF, Stewart DJ, Weininger D (1977) Applications of a bioenergetics model

15 to perch (Perca flavescens) and walleye (Stizostedion vitreum). Journal of the

16 Fisheries Research Board of Canada 34: 1922–1935.

17 24. Ney JJ (1990) Trophic economics in fisheries: assessment of demand–supply

18 relationships between predators and prey. Reviews in Aquatic Sciences 2:55–81.

19 25. Holsman KK, Aydin K (In prep) Comparative field and bioenergetics-based

20 methods for evaluating climate change impacts on the foraging ecology of

21 Alaskan walleye pollock (Theragra chalcogramma), Pacific cod (Gadus

22 macrocephalus), and arrowtooth flounder (Atheresthes stomias).

161 162

1 26. Hanson P, Johnson T, Schindler D, Kitchell J (1997) Fish Bioenergetics 3.0.

2 Madison, WI: University of Wisconsin Sea Grant Institute.

3 27. Kristiansen T, Fiksen Ø, Folkvord A (2007) Modelling feeding, growth, and habitat

4 selection in larval Atlantic cod (Gadus morhua): observations and model

5 predictions in a macrocosm environment. Canadian Journal of Fisheries and

6 Aquatic Sciences 64: 136-151.

7 28. Fiksen O, MacKenzie BR (2002) Process-based models of feeding and prey

8 selection in larval fish. Marine Ecology Progress Series 243: 151–164.

9 29. Munk P (1997) Prey size spectra and prey availability of larval and small juvenile

10 cod. Journal of Fish Biology 51(Suppl A): 340–351.

11 32. Buchheister A, Wilson MT, Foy RJ, Beauchamp DA (2006) Seasonal and

12 geographic variation in condition of juvenile walleye pollock in the western Gulf

13 of Alaska. Transactions of the American Fisheries Society 135: 897-907.

14 33. Peck MA, Buckley LJ, Bengtson DA (2006) Effects of temperature and body size

15 on the swimming speed of larval and juvenile Atlantic cod (Gadus morhua):

16 implications for individual-based modelling. Environmental Biology of Fishes 75:

17 419–429.

18 34. Ianelli JN, Honkalehto T, Barbeaux S, Kotwicki S, Aydin K, Williamson N (2012)

19 Assessment of the walleye pollock stock in the Eastern Bering Sea. In: Stock

20 assessment and fishery evaluation report for the groundfish resources of the

21 Bering Sea/Aleutian Islands regions. North Pacific Fishery Management Council,

22 605 W. 4th Ave., Suite 306, Anchorage, AK 99501. 106 pp.

162 163

1 35. Kachel NB, Hunt GL Jr, Salo SA, Schumacher JD, Stabeno PJ, Whitledge TE (2002)

2 Characteristics and variability of the inner front of the southeastern Bering Sea.

3 Deep-Sea Research II 49: 5889–5909.

4 36. Duffy-Anderson JT, Busby MS, Mier KL, Deliyanides CM, Stabeno PJ (2006) Spatial

5 and temporal patterns in summer ichthyoplankton assemblages on the eastern

6 Bering Sea shelf 1996–2000. Fisheries Oceanography 15: 80–94.

7 37. Logerwell EA, Duffy-Anderson JT, Wilson M, McKelvey D (2010) The influence of

8 pelagic habitat selection and interspecific competition on productivity of juvenile

9 walleye pollock (Theragra chalcogramma) and capelin (Mallotus villosus) in the

10 Gulf of Alaska. Fisheries Oceanography 19: 262-278.

11 38. Hollowed AB, Bond NA, Wilderbuer TK, Stockhausen WT, A'mar ZT, Beamish RJ,

12 Overland JE, Schirripa MJ (2009) A framework for modelling fish and shellfish

13 responses to future climate change. ICES Journal of Marine Science 66: 1584–

14 1594.

17 4.9 Figure legends 18 Figure 1. Log(CPUE) of juvenile walleye pollock collected in surface trawls in 2005 (a)

19 and 2010 (b). Circle size is proportional to catch (#•m-2) at each station; note difference

20 in scale between years. Stations with zero catch () are shown on white background. Log

21 of total zooplankton abundance (#•m-3) for the main prey taxa is shown for 2005 (c) and

22 2010 (d). Circle size is proportional to the abundance of zooplankton within the optimal

163 164

1 size range for 65 mm SL juvenile pollock (5-8% of fish length); note difference in scale

2 between years. Biomass-weighted mean energy density (ED) of available zooplankton

3 prey is shown for 2005 (e) and 2010 (f).

5 Figure 2. Predicted growth (g•g-1•d-1) of juvenile walleye pollock from the bioenergetics

6 model. Top panel (a and b) shows growth under the base model scenarios for 2005 and

7 2010 (W=2.5 g, Temp=average temperature in upper 30 m, h =1.0, k =prey energy

-1 -1 8 density, v2005= 3.92 kJ•g ; v2010= 5.29 kJ•g ). Middle panel (c and d) shows changes in

9 predicted growth when temperature is increased by 1 standard deviation (SD).

10 Predicted growth could not be estimated at one station in 2005 (c) in the inner domain

11 under increased temperatures because the water temperature in the upper 30 m was

12 greater than 15 ºC (Tcm = 15 ºC in the model). Lower panel (e and f) shows changes in

13 predicted growth when prey energy density is increased by 1 SD. Spatial plots of

14 predicted growth when parameters are decreased by 1 SD are not shown, but can be

15 visualized by subtracting the anomalies (lower two panels) from the base scenario plots

16 (top panel).

18 Figure 3. Predicted growth (g•g-1•d-1) of juvenile walleye pollock interpolated over the

19 range of observed temperatures and prey energy density values across both 2005 and

20 2010, providing a continuous scale of growth over a broad range of possible

21 environmental and biological scenarios. The observed fish energy density was higher in

-1 22 2010 (v2010 = 5.29 kJ•g ; used in plot shown); therefore this interpolation demonstrates

164 165

1 the range of predicted growth for fish with high energy density. Temperatures included

2 0-16 °C to show possible range under variable climate conditions. The dashed rectangle

3 encompasses the range of temperatures and prey energy density values observed in

4 2005; solid rectangle encompasses values in 2010. Points are shown for average

5 temperature and prey energy density conditions in 2005 and 2010. Predicted growth

6 above 15 °C was not possible (black) because the bioenergetics model has a

7 temperature threshold of 15 °C.

9 Figure 4. Predicted growth (g•g-1•d-1) and average depth (m) of juvenile walleye pollock

10 from the IBM. Top panel shows growth (a and c) and average depth (b and d) under the

11 base model scenarios for 2005 and 2010 (W=2.5 g, zooplankton prey distributed

12 according to vertical profiles). Middle panel shows changes in predicted growth (e and

13 g) and average depth (f and h) for 2.0 g fish, highlighting the relative importance of fish

14 size (relative to 2.5 g) and water temperature between years. Lower panel shows

15 changes in predicted growth (i and k) and average depth (j and l) when uniform vertical

16 distributions of prey are implemented, highlighting the effect of zooplankton diel

17 vertical distribution and migrations on juvenile walleye pollock prey selection. Negative

18 changes in depth indicate a shallower distribution; positive values indicate a deeper

19 distribution.

21 Figure 5. Difference in predicted growth (g•g-1•d-1) of juvenile walleye pollock between

22 the bioenergetics model and the IBM for 2005 (a) and 2010 (b). Areas of positive

165 166

1 differences indicate where maximum growth potential from the bioenergetics model

2 was higher than predicted growth from the IBM.

4 Figure 6. Conceptual figure of the spatial relationship between juvenile fish abundance

5 (yellow) and zooplankton prey availability (blue). Where these areas overlap (green),

6 juvenile fish are predicted to have higher growth rates and increased survival. Under

7 warm climate conditions, there is reduced spatial overlap between juvenile fish and

8 prey availability, resulting in lower overwinter survival and recruitment success to age-1.

9 In colder conditions, increased spatial overlap between juvenile fish and prey availability

10 results in increased overwinter survival and recruitment to age-1.

12 Figure S1. Eastern Bering Sea with locations of sampling stations at which the

13 bioenergetics model and IBM were run in 2005 (•) and 2010 (). The Monte Carlo

14 Station () is the representative station used for Monte Carlo simulations. Depth

15 contours are shown for the 50 m, 100 m, and 200 m isobaths.

17 Figure S2. Water temperatures interpolated across all stations (•) sampled by the CTD.

18 Top panel shows the mean temperature in the upper 30 m of the water column in 2005

19 (a) and 2010 (b). Bottom panel shows the mean temperature below 40 m in 2005 (c)

20 and 2010 (d).

166 167

1 Table 1. Main prey taxa included in the models for 2005 and 2010. Prey items

2 cumulatively accounting for at least 90% of the diet by % volume and individually

3 accounting for at least 2% of the diet by % volume were included. Prey taxa common to

4 both years are shown in bold.

2005 2010

Taxa Individ Cum Taxa Individ Cum

% Vol % Vol % Vol % Vol

Limacina helicina 26.33 Limacina helicina 35.45

Pseudocalanus sp. 26.04 52.4 Thysanoessa inermis 27.08 62.5

Oikopleura sp. 11.86 64.2 Calanus marshallae 13.87 76.4

Centropages abdominalis 8.98 73.2 Neocalanus cristatus 4.84 81.2

Thysanoessa raschii 8.48 81.7 Thysanoessa inspinata 3.16 84.4

Thysanoessa sp. 4.63 86.3 Thysanoessa raschii 3.09 87.5

Acartia clausi 3.40 89.7 Neocalanus plumchrus* 2.98 90.5

Calanus marshallae 1.71 91.4 Eucalanus bungii 2.95 93.4

5 *Neocalanus plumchrus was not identified in the 2010 bongo data, but did occur in the

6 Juday data (small-mesh; not quantitative for large zooplankton taxa). Due to the

7 absence in the bongo data, N. plumchrus was excluded from further analyses.

167 168

1 Table 2. Parameter definitions and values used in the bioenergetics model to estimate

2 maximum growth potential (g•g-1•d-1) of juvenile walleye pollock. Parameters were

3 used as inputs to the bioenergetics model described in [16].

Paramete Definition (units) Value Reference r C Consumption (g•g-1•d-1)  Relative foraging rate 0-1 a

O2 cal Activity multiplier; convert g O2  g prey 13560 a  Intercept of the allometric function for C 0.119 a  Slope of the allometric function for C -0.46 a

Qc Temperature dependent coefficient 2.6 b

Tco Optimum temperature for consumption 10 b

Tcm Maximum temperature for consumption 15 b -1 -1 R Respiration (g O2•g •day )

Ar Intercept of the allometric function for R 0.0075 b

Br Slope of the allometric function for R -0.251 b

Qr Temperature dependent coefficient 2.6 b

Tro Optimum temperature for respiration 13 b

Trm Maximum temperature for respiration 18 b

Ds Proportion of assimilated energy lost to Specific 0.125 b Dynamic Action

Am Multiplier for active metabolism 2 b F Egestion

Fa Proportion of consumed energy 0.15 b U Excretion

Ua Proportion of assimilated energy 0.11 b 4 a [25]; b [16]

168 169

1 Table 3. Summary of sensitivity analyses for the bioenergetics model in 2005 and 2010

2 showing the minimum (min), mean, and maximum (max) growth potential over all

3 stations. Base values are predicted maximum growth potential (g•g-1•d-1) of juvenile

4 pollock from the base model scenarios (W=2.5 g, Temp=average temperature in upper

-1 -1 5 30 m, h =1.0, k =prey energy density, v2005= 3.92 kJ•g ; v2010= 5.29 kJ•g ). All other

6 values denote the change in growth rate resulting from indicated changes in inputs;

7 therefore (-) effects indicate that varied conditions resulted in lower predicted growth

8 and vice versa. Pooled standard deviations (SDs) for each parameter were calculated

9 across stations after removing the annual means. W and vt are constant values applied

10 across all station, so changes (± 1 SD) act as a scalar and result in similar spatial patterns

11 across the area. Temperature and k vary across stations.

2005 2010

Parameter SD min mean max min mean max

Base -0.0056 0.0146 0.0291 0.0069 0.0172 0.0272

W + 1 SD 0.935 -0.0056 -0.0041 -0.0017 -0.0052 -0.0037 -0.0023

W – 1 SD 0.935 0.0034 0.0076 0.0103 0.0041 0.0068 0.0094

Temp + 1 SD 1.75 -0.0227 -0.0053 0.0017 -0.0071 -0.0007 0.0018

Temp – 1 SD 1.75 -0.0028 0.0018 0.0129 -0.0026 0.0008 0.003

k + 1 SD 497.5 0.0046 0.0061 0.0065 0.0032 0.0044 0.0048

169 170

k – 1 SD 497.5 -0.0065 -0.0061 -0.0046 -0.0048 -0.0044 -0.0032

vt + 1 SD 395.93 -0.0027 -0.0013 0.0005 -0.0019 -0.0012 -0.0005

vt – 1 SD 395.93 -0.0006 0.0016 0.0033 0.0006 0.0014 0.0022

170 171

1 2 Table 4. Summary of sensitivity analyses for the IBM model in 2005 and 2010 showing

3 the minimum (min), mean, and maximum (max) growth potential and depth (m) over all

4 stations. Base values are predicted growth (g•g-1•d-1) and depth (m) of juvenile pollock

5 from the base model scenarios (W=2.5 g, zooplankton prey distributed according to

6 vertical profiles). All other values are predicted changes in growth and depth. Negative

7 changes in depth indicate a shallower distribution; positive values indicate a deeper

8 distribution. Weight is a constant value applied across all station, so varying the

9 parameter acts as a scalar and results in similar spatial patterns across the area. The

10 effect of applying a uniform distribution of zooplankton prey with depth varies across

11 stations.

2005 2010

Parameter min mean max min mean max

Base Growth 0.0062 0.0102 0.0121 0.0055 0.0092 0.0123

Depth 10 44.2 80.9 15 47.4 93

W Growth 0.004 0.0184 0.0512 0.002 0.0068 0.0254

(2.0 g) Depth -30.5 -2.6 0.14 -43.3 -2.4 21.7

Prey Growth -0.0009 0.005 0.0058 -0.0034 0.001 0.0064

distribution Depth -21.4 2.1 15.8 -42.9 -1.8 35.2

(Uniform)

171 172

1 2 Figure 1 3

4 5 6

172 173

1 2 Figure 2. 3 4

5 6 7 8 9 10

173 174

1 2 Figure 3. 3

4 5 6

174 175

1 2 Figure 4 3

4 5 6

175 176

1 2 Figure 5. 3

4 5 6

176 177

1 2 Figure 6. 3 4

5 6

177 178

2 4.10 Supporting Information:

3 4.10.1 Comparison of observed and predicted prey preferences

4 4.10.1.1 Methods 5 Stomach contents of juvenile pollock (<100 mm SL) were identified from selected

6 stations across the eastern Bering Sea shelf to compare observed diet composition,

7 model-predicted diets, and available prey. Chesson’s prey preference index [1] was

8 calculated for the main prey taxa in each year and compared to the individual-based

9 model (IBM) estimates of prey preference at corresponding stations. Chesson’s index

10 (i) for a given prey taxon i is the ratio between ingested prey items (ri) and the

11 frequency of their occurrence in the environment (ni), standardized by the sum of the

12 ratios over all m prey types:

ri n 13 a = i Eq. 1 i m r S j j =1 n j

14 The standardization implies that neutral selection (neu) corresponds to 1/m and a

15 specific prey item or group was actively selected if the index is > aneu, as they appear

16 more frequently in the diet than their abundance in the environment would suggest. To

17 calculate Chesson’s index for observed diets, % volume in the diet was used as a proxy

18 for biomass consumed (ri) relative to prey biomass in the environment (ni).

19 Chesson’s prey preference indices for main prey taxa in each year, based on

20 observed and predicted diets, were compared at each station for which observed diet,

21 predicted diet, and zooplankton composition was available and where at least 90% of

178 179

1 the observed prey taxa (by % volume) were accounted for in the zooplankton data (n=7

2 in 2005; n=9 in 2010). Zooplankton samples that did not include a known prey item

3 were not considered for this analysis because the lack of a known prey item in

4 zooplankton samples collected at the same station suggests that the sample is not

5 representative of prey availability due to small-scale patchiness, or indicates a spatial

6 and/or temporal mismatch between where captured juvenile pollock were foraging and

7 where samples were collected. To compare observed (obs) and predicted (pre) prey

8 preferences, we computed differences between these prey preferences relative to

9 neutral selection:

obs pre 10 i /neu  i /neu Eq. 2

11 and averaged them across all stations within each year, as well as by domain (i.e., inner:  12 0-50 m isobath, middle: 50-100 m isobath, and outer: 100-200 m isobath).

13 4.10.1.2 Results and Discussion 14 Modeled diets from the IBM were comparable to observed diets from the 2005 and

15 2010 surveys (Fig. S1; most differences overlap zero), indicating that the model may

16 adequately capture predator-prey dynamics. Relatively small differences between

17 observed and predicted prey preference were consistent across domains in both years.

18 Limacina helicina, the predominant component of diets across years, was more

19 prevalent in observed diets (except in the inner domain south of 60 ºN in 2005), as was

20 Thysanoessa raschii. Modeled diets, however, consistently overestimated consumption

21 of Calanus marshallae and Eucalanus bungii (2010 only) across domains.

179 180

1 Shifts in C. marshallae abundance between warm and cold years in the EBS have

2 been proposed as a major contributor to differences in juvenile pollock condition and

3 survival [2]. However, observed diets of juvenile pollock (<100 mm SL) in 2005 and 2010

4 do not reflect its relative importance to growth because C. marshallae was less

5 prevalent than expected. Greater prevalence of specific prey items in observed diets

6 (e.g., Centropages abdominalis in 2005) indicates that the IBM model underestimates

7 the ability of juvenile pollock to detect, capture, and ingest that prey item. Alternatively,

8 the prey could have been more abundant in the areas where juvenile pollock were

9 feeding than in the area sampled by the bongo and/or Juday net due to patchiness.

10 Differences between observed and predicted diets may also be explained by prey

11 escape behaviors or size-selectivity by juvenile pollock that is more complex than the

12 prey selection component of the IBM [3]. Juvenile pollock collected in late-summer

13 likely feed more heavily in surface waters during crepuscular or nighttime periods [4],

14 moving deeper during the daytime, while observed diets for this study were sampled

15 from daytime surface hauls. However, the spatial and temporal disconnect between

16 where juvenile pollock feed and were collected for diet analyses likely did not affect our

17 results as previous work has shown that proportional diet compositions do not vary

18 between day and night [4,5] and the IBM integrates predicted diets over 24 hours,

19 encompassing diel vertical patterns.

20 4.10.2 References 21 1. Chesson J (1978) Measuring preference in selective predation. Ecology 59: 211-

22 215.

180 181

1 2. Coyle KO, Eisner LB, Mueter FJ, Pinchuk AI, Janout MA, Cieciel KD, Farley EV,

2 Andrews AG (2011) Climate change in the southeastern Bering Sea: impacts on

3 pollock stocks and implications for the Oscillating Control Hypothesis. Fisheries

4 Oceanography 20(2): 139-156.

6 3. Petrik CM, Kristiansen T, Lough RG, Davis CS (2009) Prey selection of larval

7 haddock and cod on copepods with species-specific behavior: a model-based

8 analysis. Marine Ecology Progress Series 396: 123-143.

9 4. Brodeur RD, Wilson MT, Ciannelli L (2000) Spatial and temporal variability in

10 feeding and condition of age-0 walleye pollock (Theragra chalcogramma) in

11 frontal regions of the Bering Sea. ICES Journal of Marine Science 57: 256–264.

12 5. Schabetsberger R, Brodeur RD, Ciannelli L, Napp JM, Swartzman GL (2000) Diel

13 vertical migration and interaction of zooplankton and juvenile walleye pollock

14 (Theragra chalcogramma) at a frontal region near the Pribilof Islands, Bering Sea.

15 ICES Journal of Marine Science 57: 1283–1295.

181 182

1 Table S1. Stage, sampling gear, length range, width, biomass (g, wet weight), and energy density (kJ•g-1, wet weight) values for the

2 main prey items of juvenile walleye pollock in late summer 2005 and 2010. Biomass estimates were obtained during processing of

3 the zooplankton samples from 2005 (warm) and 2010 (cold) (NA=stage was not collected); energy density values were obtained

4 from zooplankton collected in the eastern Bering Sea during 2004 (warm) and 2010 (cold). Single estimates of energy density (shown

5 in bold) were used when year-specific information was not available for individual taxa.

Species Stage Gear Length Width Warm Cold Warm Cold Comments

range (mm) Biomass Biomass Energy Energy

(mm TL) (g WW) (g WW) Density Density

(kJ•g-1 WW) (kJ•g-1 WW)

Acartia clausi A Juday 0.25 – 1.4a 0.22b 3.5 E-05 3.5 E-05

3.816c 3.816 c

AF Juday 0.8 – 1.4 a 0.29 b NA 4.5 E-05

AM Juday 0.8 – 1.2 a 0.27 b NA 2.5 E-05

Acartia sp. A Juday 0.25 – 0.93 a 0.16 b 1.85 E- 1.9 E-05

182 183

I Juday 0.25 – 0.42 a 0.09 b NA Length range for A.

4.1 E-06 clausi

II Juday 0.42 – 0.51 a 0.12 b NA Length range for A.

9.4 E-06 clausi

III Juday 0.51 – 0.65 a 0.16 b NA Length range for A.

1.7 E-05 clausi

IV Juday 0.65 – 0.76 a 0.19 b 1.0 E-05 Length range for A.

2.5 E-05 clausi

V Juday 0.76 – 0.93 a 0.23 b 2.7 E-05 Length range for A.

4.0E-05 clausi

Calanus AF Bongo 3.2 – 4.2 a 0.78d marshallae 0.003 0.002 5.325f 5.732g

AM Bongo 3.5 – 4 a 1.0e 0.0026 0.0017

183 184

I Bongo 0.5 – 0.7 a 5.25 E- 5.25 E-

0.16 e 05 05

II Bongo 1.2 – 1.5 a 9.19 E- 1.33 E-

0.36 e 05 04

III Bongo 1.6 – 2.3 a 0.68 d 1.9 E-04 2.7 E-04

I-III Bongo 0.5 – 2.3 a 0.37 e 1.13 E-

04 2.2 E-04

IV Bongo 2.3 – 2.6 a 5.82 E- 5.24 E-

0.69 d 04 04

V Bongo 2.8 – 3.8 a 0.73 d 0.002 0.0016

Centropages A Juday 0.3 – 2.1 a 0.36h 7.49 E- 5.73 E- abdominalis 05 05 3.843i 3.843 i

AF Juday 1.6 – 2.1 a 0.53 h 1.61 E-4 1.36 E-4

AM Juday 1.5 a 0.5 h 9.14 E- 1.16 E-4

184 185

C Juday 1.13 a 0.5 h 2.33 E- 2.33 E-

05 05

I Juday 0.3 a 0.16 h 2.92 E-

NA 05

II Juday 0.38 a 0.22 h NA 1.5 E-05

III Juday 0.5 a 0.29 h 1.32 E- 2.82 E-

05 05

I-III Juday 0.33 a 0.17 h NA 9.1 E-06

IV Juday 0.65 a 0.39 h 4.25 E-

NA 05

V Juday 0.85 a 0.5 h 8.55 E- 6.89 E-

05 05

Eucalanus A Bongo 4.8 – 8 a 1.7e 0.0023 8.67 E- 3.916j 4.194k

185 186

bungii 05

AF Bongo 6 – 8 a 1.86 e 0.0086 0.0086

AM Bongo 4.8 – 5.5 a 1.37 e 0.0031 0.0041

I Bongo 1.3 – 1.6 a 0.39 e 4.5 E-05 5.3 E-05

II Bongo 2 – 2.2 a 0.56 e 1.96 E-4 1.56 E-4

III Bongo 2.9 – 3 a 0.79 e 4.92 E-4 2.66 E-4

I-III Bongo 1.3 – 3 a 0.57 e 2.44 E-

04 8.41 E-4

IV Bongo 3.36 – 3.8 a 0.95 e 0.0011 0.0009

V Bongo 4.5 – 5.2 a 1.29 e 0.0027 0.0034

Limacina XS Bongo 0.1 – 0.5l 0.3 l NA 6.93 E- helicina 05 2.51m 2.766g

S Bongo 0.5 – 2 l 1.25 l 9.40 E-

05 1.58 E-4

186 187

M Bongo 2 – 4 l 3 l 3.71 E-4 8.29 E-4

L Bongo 4 - 10 l 7 l 0.0026 0.0045

Neocalanus AF Bongo 8.5 – 10.4a 2.5 e cristatus NA 0.0137 3.253n 3.39g

III Bongo 3.2 a 0.85 e 8.83 E-4 0.0013

IV Bongo 4.9 – 5.3 a 1.36 e 0.0059 0.0025

V Bongo 7.1 – 8.9 a 2.13 e 0.019 0.015

N. plumchrus V Bongo 4.1 – 5.2 a 1.24 e 0.00395 0.0041 4.207o 4.676g

Oikopleura sp. A Bongo 0.1 – 0.6p 0.35 p 1.73 E-4 1.7 E-4 4.076q 4.025r

Pseudocalanus A Juday 0.65 – 1.2s 0.29 d 4.44 E- Length range for P. spp. 05 3.6 E-05 3.951t 3.951 t moultoni

AF Juday 1.05 – 2.27 s 0.42 d 8.27 E- 8.01 E-

05 05

AM Juday 0.91 – 1.74 s 0.35b 4.13 E- 5.42 E-

187 188

05 05

I Juday 0.5 – 0.7 s 0.16 b 5.98 E-

NA 06

II Juday 0.65 – 0.8 s 0.19 b 1.01 E- 1.08 E-

05 05

III Juday 0.8 – 1 s 0.24 b 1.26 E-

05 2.0 E-05

I-III Juday 0.5 – 1 s 0.2 b 1.01 E-

NA 05

IV Juday 1 – 1.2 s 0.3 d 3.06 E- 3.19 E-

05 05

V Juday 1.2 – 1.5 s 0.37 d 8.89 E- 4.93 E-

05 05

II-V Juday 0.65 – 1.5 s 0.29 d 3.55 E- 2.8 E-05

188 189

Thysanoessa A Bongo 10.1 – 29.2u 2.4 d inermis NA 0.083 4.99m 4.99 m

AF Bongo 10.1 – 29.2 u 2.4 d 0.10

AM Bongo 10.1 – 29.2 u 2.4 d 0.069 NA

J Bongo 8.5 – 13.8 u 1.4 d NA 0.011

J (L) Bongo 11.1 – 13.8 u 1.5 d 0.061 0.11

J (S) Bongo 8.5 – 11.1 u 1.2 d 0.011 0.024

T. inspinata J Bongo 12 – 17v 2.2w 0.0012 0.0128 4.99x 4.99x

T. raschii A Bongo 7 – 29.1y 3.3 d 0.006 NA 4.308aa 5.231g

AF Bongo 7 – 29.1 y 3.3 d 0.077 0.0903

AM Bongo 15.3 – 20.2z 2.7 d 0.046 0.088

J Bongo 7.4 – 8.4 z 1.45 d 0.005 0.0117

J (L) Bongo 7.9 – 8.4 z 1.5 d 0.0476 0.0936

189 190

J (S) Bongo 7.4 – 7.9 z 1.4 d 0.0089 0.0059

190 191

1 a [1]; b Estimated width = 26.7% of length (based on Pseudocalanus sp. relationship); c Energy

2 density estimated from % lipid (2.25% wet weight assuming 80% moisture [2]) using the

3 regression relationship: ED = (0.4098•% lipid) + 19.287; d E. Fergusson, NOAA/AFSC,

4 unpublished data; e Estimated width = 26.6% of length (based on C. marshallae relationship); f

5 Energy density estimated from % lipid (10.5%; R. Heintz, NOAA/AFSC, unpublished data); g R.

6 Heintz, NOAA/AFSC, unpublished data; h [3]; i Energy density estimated from % lipid (2.6% wet

7 weight [4]); j Energy density estimated from % lipid (3.55%; R. Heintz, NOAA/AFSC, unpublished

8 data); k Energy density estimated as 7.1% higher in cold years (based on copepod data; [this

9 study]); l C. Stark, UAF, unpublished data; m 2006 collection (R. Heintz, NOAA/AFSC, unpublished

10 data); n Energy density estimated from % lipid (5.85%; R. Heintz, NOAA/AFSC, unpublished

11 data); o Energy density estimated from % lipid (6.83%; R. Heintz, NOAA/AFSC, unpublished

12 data); p Trunk length/width [5]; q Energy density estimated from % lipid of Chaetognatha

13 (2.67%; R. Heintz, NOAA/AFSC, unpublished data); r Energy density estimated from % lipid of

14 Chaetognatha (2.04%; R. Heintz, NOAA/AFSC, unpublished data); s [6]; t Energy density

15 estimated from % lipid (4% wet weight [7]); u Carapace width from [8]; converted to TL using

16 equations from [9]; v Length range of ‘spineless’ T. longipes [10]; w Estimated width as 15% of

17 length; x Used energy density of T. inermis; y Minimum size for T. inermis and maximum size for

18 T. spinifera [8]; converted to TL using equations from [9]; z Minimum size for T. inermis and

19 maximum size for T. spinifera [11]; converted to TL using equations from [9]; aa Energy density

20 estimated as 17.65% higher in cold years (R. Heintz, NOAA/AFSC, unpublished data).

191

192

1 References 2 1. Gardner GA, Szabo I (1982) British Columbia pelagic marine copepoda: an identification

3 manual and annotated bibliography. Canadian Special Publication of Fisheries and

4 Aquatic Sciences 62. 536 p.

5 2. Yamamoto T, Teruya K, Hara T, Hokazono H, Hashimoto H, Suzuki N, Iwashita Y,

6 Matsunari H, Furuita H, Mushiake K (2008) Nutritional evaluation of live food organisms

7 and commercial dry feeds used for seed production of amberjack Seriola dumerili.

8 Fisheries Science 74: 1096-1108.

9 3. Lough RG, Buckley LJ, Werner FE, Quinlan JA, Edwards KP (2005) A general biophysical

10 model of larval cod (Gadus morhua) growth applied to populations on Georges Bank.

11 Fisheries Oceanography 14: 241-262.

12 4. Lee RF, Hagen W, Kattner G (2006) Lipid storage in marine zooplankton. Marine Ecology

13 Progress Series 307: 273-306.

14 5. Tomita M, Ikeda T, Shiga N (1999) Production of Oikopleura longicauda (Tunicata:

15 Appendicularia) in Toyama Bay, southern Japan Sea. Journal of Plankton Research

16 21(12): 2421-2430.

17 6. Frost BW (1989) A taxonomy of the marine calanoid copepod genus Pseudocalanus.

18 Canadian Journal of Zoology 67: 525-551.

19 7. Peters J (2006) Lipids in key copepod species of the Baltic Sea and North Sea –

20 implications for life cycles, trophodynamics and food quality. PhD Dissertation.

21 University of Bremen. 177 p.

192

193

1 8. Pinchuk AI, Coyle KO (2008) Distribution, egg production and growth of euphausiids in

2 the vicinity of the Pribilof Islands, southeastern Bering Sea, August 2004. Deep-Sea

3 Research II 55: 1792-1800.

4 9. Pinchuk AI, Hopcroft RR (2007) Seasonal variations in the growth rates of euphausiids

5 (Thysanoessa inermis, T. spinifera, and Euphausia pacifica) from the northern Gulf of

6 Alaska. Marine Biology 151: 257-269.

7 10. Kathman RD, Austin WC, Saltman JC, Fulton JD (1986) Identification manual to the

8 Mysidacea and Euphausiacea of the northeast Pacific. Canadian Special Publications of

9 Fisheries and Aquatic Sciences 93. 411 p.

10 11. Falk-Petersen S (1985) Growth of the euphausiids Thysanoessa inermis, Thysanoessa

11 raschii, and Meganyctiphanes norvegica in a subarctic fjord, North Conway. Canadian

12 Journal of Fisheries and Aquatic Sciences 42: 14-22.

193

194

1 2 Table S2. Component equations of the bioenergetics model used to estimate maximum growth

3 potential (g•g-1•d-1) of juvenile walleye pollock.

Consumption C = a× W b × f (T)× h

f (T) = V X × e(X ×(1-V ))

V = (Tcm -T)/(Tcm - Tco)

X = (Z 2× (1+(1+ 40/Y)0.5)2 /400

Z = ln(Qc )× (Tcm -Tco )

Y = ln(Qc )× (Tcm - Tco +2)

Br Respiration R = (Ar × W × f (T)× Am × 13560) + (Ds× (C - F))

f (T) = V X × e(X ×(1-V ))

V = (Trm - T)/(Trm - Tro)

X = (Z 2× (1+(1+ 40/Y)0.5)2)/400

Z = ln(Qr )× (Trm - Tro )

Y = ln(Qr )× (Trm -Tro +2)

Egestion F = Fa × C

Excretion U Ua  (C F)

5 

194

195

1 2 Figure S1 3 4 5

195

196

1 2 Figure S2 3

4 5 6

196

197

1 2 Figure S3 3

4 5 6 7

197

198

2 Chapter 5 - Size, diet, and condition of age-0 Pacific cod (Gadus 3 macrocephalus) during warm and cool climate states in the eastern 4 Bering Sea 5 Edward V. Farley, Jr.1*, Ron A. Heintz1, Alex G. Andrews1, Thomas P. Hurst2

6 1Auke Bay Laboratories, Alaska Fisheries Science Center, National Marine Fisheries Service,

7 National Oceanic and Atmospheric Administration, 17109 Point Lena Loop Road, Juneau, AK

8 99801, USA

9 2Resource Assessment and Conservation Engineering Division, Alaska Fisheries Science Center,

10 National Marine Fisheries Service, National Oceanic and Atmospheric Administration, Hatfield

11 Marine Science Center, 2030 SE Marine Science Drive, Newport, OR 97365, USA

12 *Corresponding author: tel: +1 907 789-6085; fax: +1 907 789-6094; e-mail: [email protected]

14 Submitted to Deep Sea Research II (2014)

16 5.1 Abstract 17 The revised Oscillating Control Hypothesis for the Bering Sea suggests that recruitment of

18 groundfish is linked to climatic processes affecting seasonal sea ice that, in turn, drives the

19 quality and quantity of prey available to young fish for growth and energy storage during their

20 critical life history stages. We test this notion for age-0 (juvenile) Pacific cod (Gadus

21 macrocephalus) by examining the variability in size, diet, and energetic condition during warm

22 (2003 to 2005), average (2006), and cool (2007 to 2011) climate states in the eastern Bering

198

199

1 Sea. Juvenile cod stomachs contained high proportions of age-0 walleye pollock (by wet

2 weight) during years with warm sea temperatures with a shift to euphausiids and large

3 copepods during years with cool sea temperatures. Juvenile cod were largest during years with

4 warm sea temperatures and smallest during years with colder sea temperatures. However,

5 energetic status (condition) of juvenile cod was highest during years with cool sea

6 temperatures. This result is likely linked to the shift to high quality, lipid rich prey found in

7 greater abundance on the shelf and in the stomach contents of juvenile cod during cool years.

8 Our examination of juvenile cod size, diet, and energetic status provided results that are

9 coincident with juvenile pollock, suggesting that the common mechanisms regulating gadid

10 recruitment on the eastern Bering Sea shelf are sea temperature, prey quality and quantity, and

11 fitness of gadids prior to winter.

199

200

2 5.2 Introduction 3 The continental shelf of the eastern Bering Sea supports important fisheries for walleye pollock

4 (Gadus chalcogrammus), Pacific cod (G. macrocephalus), flatfish and crab. These populations

5 have undergone large fluctuations due to both natural climate variability and potential

6 anthropogenic forcing (climate warming). The potential for climate warming in the Bering Sea

7 has heightened the need for data to better understand impacts of climate variability on

8 ecologically and commercially important fish stocks. Probably the greatest effect of climate

9 variability on the southeastern Bering Sea ecosystem will be the annual difference in the extent

10 and duration of winter and spring sea ice (Stabeno et al., 2012). These characteristics are

11 believed to affect the quality and quantity of prey resources that, in turn, impact growth,

12 fitness, and survival of young fish (e.g. walleye pollock; Hunt et al., 2011).

13 Understanding how fish allocate energy for growth and lipid storage during critical life history

14 stages is important for assessing climate impact on fish recruitment. For young fish, differential

15 mortality occurs during two critical periods (first feeding and winter) that are related to energy

16 allocation strategies in larval and juvenile fish (Post and Parkinson, 2001). For larval fish,

17 somatic growth is important because smaller fish may be more susceptible to size-selective

18 predation. For juvenile fish, large size and high energy reserves (lipid) are important because

19 they can face severe energy deficits prior to, and during winter (Sogard and Olla, 2000; Heintz

20 and Vollenweider, 2010; Farley et al., 2011). For instance, the energetic status of age-0 Bering

21 Sea walleye pollock in late summer is increasingly recognized as a predictor of age-1 abundance

22 the following summer (Heintz et al., 2013). Therefore, consumption of high quality prey should

200

201

1 improve the growth and nutritional condition in juvenile fish, reduce energy deficits over-

2 winter and lead to increased survival.

3 The oscillating control hypothesis (OCH) that links climate variability to Bering Sea groundfish

4 (e.g. walleye pollock) recruitment (Hunt et al., 2002) was revised (Hunt et al., 2011) based in

5 part on new findings regarding the importance of energetic status to fish survival (Heintz et al.,

6 2013). Originally, the OCH predicted warmer temperatures and early ice retreat on the eastern

7 Bering Sea shelf would lead to increased productivity in the pelagic ecosystem that would

8 improve the chances of successful recruitments for groundfish. However, recent data indicate

9 that changes in prey composition and abundance during periods of anomalously warm

10 temperatures and early ice retreat may have been detrimental to age-0 groundfish survival.

11 Specifically, larger zooplankton taxa, such as lipid-rich Calanus spp., were less abundant during

12 years with anomalously warm temperatures, when reduced lipid reserves in age-0 pollock at

13 the end of summer, led to increased over-winter mortality (Coyle et al., 2011). Thus, the OCH

14 was revised to state that , reduced overwinter mortality for age-0 pollock occurs during years

15 with anomalously cool temperatures because increased numbers of lipid-rich prey, combined

16 with lower metabolic demand, allows age-0 walleye pollock to acquire greater lipid reserves by

17 late summer (Hunt et al., 2011).

18 Pacific cod is a widespread marine species found on the continental shelves throughout the

19 North Pacific and Bering Sea. Pacific cod spawn demersal eggs with larvae rising to the surface

20 waters immediately after hatch (Doyle et al., 2009; Hurst et al., 2009). By July, age-0 cod are

21 known to settle to the bottom and inhabit the demersal, shallow waters of coastal Alaska

201

202

1 (Abookire et al., 2007; Laurel et al., 2007); however, in the Bering Sea, age-0 cod do not appear

2 to be restricted to shallow nearshore habitats, but instead are commonly captured across the

3 broad shelf in both demersal and pelagic trawl surveys (Hurst et al., 2012). Observational data

4 for age-0 Pacific cod and walleye pollock during late summer suggest that their distributions

5 overlap on the eastern Bering Sea shelf, but that pollock distribution tends to be more

6 widespread and variable than Pacific cod (Hurst et al., 2012). While there may be some

7 differences in distribution between age-0 walleye pollock and Pacific cod on the eastern Bering

8 Sea shelf, their recruitment dynamics appear to be synchronous (Mueter et al., 2007),

9 suggesting that the primary ecological drivers (e.g. OCH) of recruitment in these two species are

10 similar in the Bering Sea.

11 In this paper, we describe patterns of interannual variation in size, diet, and energetic status of

12 age-0 Pacific cod in the eastern Bering Sea based on 8 years of fishery-independent survey data.

13 The period examined (2003 to 2011) was characterized by significant variation in thermal

14 regime in the Bering Sea, allowing us to consider the potential effects of climate variability on

15 the biological characteristics of cod that may impact their winter survival. These biological

16 characteristics for cod are discussed in relation to age-0 walleye pollock and their potential for

17 understanding recruitment processes in relation to climate variability.

18 5.3 Methods

19 5.3.1 Field sampling 20 The biological characteristics (size, diet, energy content) of age-0 Pacific cod (hereafter

21 “juvenile cod”) in the eastern Bering Sea were described for 8 cohorts (2003 to 2011; 2009 was

22 omitted due to lack of samples), based on catches from the Bering-Aleutian Salmon

202

203

1 International Survey (BASIS). Surveys were conducted from chartered fishing vessels (38-m FV

2 “Sea Storm” or the 49-m FV “Northwest Explorer”) or the 64-m NOAA ship “Oscar Dyson”. The

3 surveys were conducted during similar time periods (mid-August through late September), with

4 sampling effort beginning in eastern Bristol Bay and moving northwest through the eastern

5 Bering Sea (Fig. 1). Slight changes in the survey locations, sampling dates, and the number of

6 stations sampled in a given year were due to weather conditions and other factors. A summary

7 of annual sampling is provided in Table 1. Fish were collected with a 198-m long midwater rope

8 trawl composed of hexagonal mesh wings and a body fitted with a 1.2-cm mesh codend liner

9 (Farley et al., 2005). Buoys were attached to the headrope to fish near-surface depths (mouth

10 opening of 55-m horizontal by 20-m vertical). The net was towed at speeds from 3.5 to 5.0

11 knots (approx 6.5 – 9.3 km per hour) for 30 min during daylight hours.

12 Several caveats on the use of surface trawl samples for understanding early marine ecology of

13 juvenile cod were explained in Hurst et al. (2012). Briefly, while the survey was not designed to

14 sample 0-group Pacific cod, they are a regular component of the catch. In addition, catch rates

15 of juvenile cod are correlated with those of age-1 fish captured in the Bering Sea bottom trawl

16 survey. There is also little known on when juvenile cod settle to a fully benthic distribution, but

17 during late summer, the surface trawl is catching more juvenile cod than the bottom trawl

18 surveys. Therefore, the assumption is that the catches in the surface trawl survey during late

19 summer adequately represent basin-scale patterns in size, diet, and energy content of juvenile

20 cod.

203

204

1 The contents of the trawl were emptied onto a sorting table on deck, and juvenile cod were

2 sorted from other life stages and species of fish (see Hurst et al., 2012 for details). A subsample

3 of up to 100 fish were individually measured (binned in 5 mm increments during 2003 to 2005

4 and to 1 mm fork length, during 2006 to 2011) to determine average length. The average

5 weight at each station was determined by bulk weighing all measured fish or counting the

6 number of fish in a pre-weighed subsample.

7 Similar to Hurst et al. (2012), we used a subset of core sampling sites that were sampled a

8 minimum of four times between 2004 and 2011 to minimize the potential influence of

9 interannual variation in spatial coverage of sampling effort on diet and caloric content. This

10 resulted in the inclusion of 99 sampling sites in the core sampling area (Fig. 1; Hurst et al. (2012)

11 found 92 core sampling stations within the same area). These core sampling sites were then

12 divided into three domains: inner, middle, and outer based on geographic landmarks and depth

13 contours associated with recognized hydrographic domains (Kinder and Schumacher, 1981;

14 Overland et al., 1999). Data were also categorized into years with anomalously warm sea

15 temperatures and low sea ice extent on the eastern Bering Sea shelf (2003 to 2005) and years

16 with anomalously cool sea temperatures and extensive sea ice extent (2007 to 2011; Fig. 2).

17 While spring sea temperatures were anomalously cold during 2006, sea ice extent and summer

18 temperatures during that year appeared to represent transitional (defined as average)

19 conditions between the two periods (Stabeno et al., 2012).

20 5.3.2 Size 21 Pacific cod size used in the analyses of size, diet, and energetic status was limited to fish less

22 than or equal to 110 mm. In comparison, Hurst et al. (2012) included all fish <= 140 mm in the

204

205

1 analysis for distribution. Prey composition and energetic status are both a function of fish size,

2 therefore we chose the lower size for this analysis to reduce the chance of biasing our analyses

3 on energetic status and diet with fish that could be from an older year class. Analysis of

4 variance (ANOVA, fixed effect) tests were used to examine interannual differences in length of

5 juvenile cod. In this analysis, all juvenile cod less than 110 mm captured in the trawl were

6 included because sample sizes were relatively small. To look for broader effects of climate

7 regime and oceanographic domain on fish length, sizes from each domain (inner, middle, and

8 outer) and climate condition (warm, average, and cool) were pooled across years and analyzed

9 with a 2-way main effects ANOVA. Data were analyzed using Minitab 16 statistical software. If

10 a significant difference (P < 0.05) occurred, Tukey’s multiple comparison test was used to

11 calculate the 95% confidence intervals (α = 0.05) for all pairwise differences between the

12 dependent variable means.

13 5.3.3. Fish diet 14 Food habits followed methods described in Moss et al. (2009) using the standard methods

15 developed at Tikhookeanskiy Nauchno-Issledovatelskiy Institut Rybnogo Khozyaystva I

16 Okeanografiy (TINRO; Pacific Ocean Research Institute of Fisheries and Oceanography; Volkov

17 et al., 2007; Chuchukalo and Volkov, 1986). At each station where juvenile cod were captured,

18 their stomach content was examined by removing and pooling the contents of the entire food

19 bolus from up to 20 randomly selected individuals from the total catch of juvenile cod. Pooled

20 stomach contents were weighed to the nearest 0.001 g, sorted, and identified to the lowest

21 feasible taxonomic group. The contribution of each prey taxa to the diet was determined by

22 dividing the volume occupied by each prey taxa by the total volume of prey contained in the

205

206

1 stomachs and multiplying that proportion by the total weight of the stomach contents. This

2 resulted in 111 diet compositions each representing the average diet of fish collected at a given

3 station among the 99 core sites for this analysis. The number of stations with diet data in a

4 given year ranged from 3 (2004) to 46 (2006) depending on the occurrence of juvenile cod.

5 Differences in dietary composition among climate regimes and domains were analyzed using

6 one-way analysis of similarity (ANOSIM) using a Bray-Curtis similarity matrix constructed from

7 the 111 diet compositions collected in 2003 -2005 (warm years) and 2007-2011 (cool years). To

8 simplify the analysis 51 prey taxa were combined into 17 diet categories (Table 2). Similarity

9 indices (SBC) in the matrix were calculated as:

th 11 Where n1i is the proportion of the i prey class in diet sample 1. . The ANOSIM R statistic

12 indicates the magnitude of the difference in the mean rank similarity for two samples when

13 samples are considered between groups and within groups. If the mean rank difference

14 between two groups is the same as that within groups then R will equal zero; differences are

15 normalized so that R varies between -1 and 1. In an analysis of similarity (ANOSIM) samples are

16 randomized and a distribution of R statistics is generated. The observed R statistic is compared

17 with the randomized distribution to determine if between group similarity is significantly

18 different from the within group similarity. The nine sampling strata considered included each of

19 the three domains (inner, middle and outer) in each of the three periods (warm, average, cool).

20 The similarities between diets were visualized by plotting them in two dimensions following

21 transformation by multidimensional scaling. The contributions of different prey items to the 206

207

1 total dissimilarity between warm and cool diets (pooled across all domains) were evaluated

2 using a similarity percentages (SIMPER) analysis. This permitted identification of those prey

3 items that contributed most to the difference between strata. ANOSIM and SIMPER analysis

4 were both performed with PRIMER software (version 6).

5 5.3.4. Energy content 6 The energy density (kJ/g wet mass) of juvenile cod was determined following methods in Heintz

7 et al. (2013). Random subsamples of 10-20 juvenile cod per station were sealed in plastic bags,

8 frozen at sea, and transported to the Auke Bay Laboratories in Juneau, Alaska, for energy

9 density analysis. In the laboratory, juvenile cod were thawed, and after their stomach contents

10 were removed, they were dried and pulverized to form a homogenate. Samples were

11 processed in the fall of 2012, so they were frozen for different amounts of time. Pulverized

12 homogenates were pressed into pellets weighing at least 25 mg and combusted using a Parr

13 6725 semi-micro bomb calorimeter. Approximately 17 fish from each year were processed.

14 Samples for processing were selected so that the mean size of a sample was similar to the mean

15 size of fish in the catch. Juvenile cod samples were not routinely retained for calorimetry in the

16 early years of the BASIS survey and they were often opportunistic when they were collected.

17 Consequently, energy data were only available from a few years (2005, 2007, 2010 and 2011)

18 and from a limited number of stations in each of those years. Samples processed from 2005 had

19 no station information, in 2007 only two stations were sampled, in 2010 fish from 13 stations

20 were processed and fish from 11 stations were processed in 2012.

21 Energy densities were compared among years by a nested ANOVA. The independent values for

22 the ANOVA were year and stations nested within years. In addition, we regressed the energy

207

208

1 density for each year averaged over the stations against the average energy density of juvenile

2 pollock sampled in the same year as reported in Heintz et al. (2013). The pollock data set has a

3 much better sample coverage, thus the regression was intended to verify that patterns

4 observed in the less well-covered cod data set corresponded with those reported for pollock.

5 All statistical analyses involving cod energy densities were performed on a dry mass basis to

6 remove potential biases from different times in frozen storage. Results of the cod and pollock

7 regression are presented on a wet mass basis for comparability.

8 5.4 Results

9 5.4.1. Size 10 The mean lengths of juvenile cod were significantly (p < 0.001) different among years examined

11 (Table 3; Fig. 3). Mean lengths were largest during 2004 and 2005 and smallest during 2007

12 and 2010. When juvenile cod lengths were pooled by domain and climate condition, lengths

13 were significantly larger in all domains during warm years than during cool years (p < 0.001 Fig.

14 4). During the average year of 2006, lengths of juvenile cod in the inner and middle domains

15 were similar to those observed during the cool years; lengths in the outer domain were

16 intermediate to those observed in the warm and cool conditions. In all three climate

17 conditions, juvenile cod were smaller in the outer domain than they were in the inner and

18 middle domains (p = 0.002).

19 5.4.2. Diet 20 Juvenile cod diets differed among the climate regimes in each of the domains (Fig. 5). Juvenile

21 cod diets in the middle and inner domains in warm years were significantly different from those

22 observed in cool years (Table 4) (R > 0.351; p < 0.01). There were only a few juvenile cod diets

23 available for analysis in the outer domain during warm years, consequently ANOSIM tests 208

209

1 involving the outer domain diets in warm years had few (< 66) possible permutations, obscuring

2 interpretation of the associated R statistic. Pollock and crab zoea accounted for 63% of the

3 consumed mass in juvenile cod diets from the inner domain in warm years. In the average year,

4 pollock were not found in juvenile cod diets in the inner domain, instead crab zoea and

5 euphausiids accounted for 61% of the diet. In the cool years, crab zoea were replaced by large

6 copepods, together with euphausiids they accounted for 55% of the inner domain diets. Hence,

7 diets during the year with average sea temperatures were more similar to the years with cool

8 sea temperatures in the inner domain (R = 0.166, p = 0.074) than during years with warm sea

9 temperatures (R = 0.298; p = 0.057). In the middle domain, pollock and crab zoea accounted

10 for 85% of the diet. The transition in diets from the years with warm to average sea

11 temperatures in the middle domain was characterized by decreased consumption of pollock;

12 euphausiids, pteropods and crab zoea accounted 87% of the diet. In cool years, juvenile cod

13 within the middle domain primarily consumed euphausiids, hyperiid amphipods, crab zoea and

14 large copepods, which accounted for 80% of the diet. Consequently, juvenile cod diets in the

15 middle domain during the average year were equally different from warm (R = 0.20; p = 0.049)

16 and cool years (R = 0.256; p = 0.001). In the outer domain euphausiids, crab zoea, and

17 pteropods were always the most important juvenile cod prey by mass, accounting for at least

18 73% of the diet, regardless of climate. This meant that outer domain diets in the average year

19 were indistinguishable from the warm (R = -0.12; p = 0.685) and the cool year diets (R= 0.009; p

20 = 0.426).

21 The differences in diet between warm and cool years reflected a change from diets dominated

22 by pollock and crab zoea in warm years to diets favoring adult euphausiids, hyperiid

209

210

1 amphipods, large copepods, pteropods, mysids and shrimp in cool years (Fig. 6). Analysis of the

2 contributions of each prey group to the similarity among warm year diets indicated crab zoea

3 and pollock dominated the warm year diets, accounting for > 90% of the average similarity

4 between pairs. In contrast, euphausiids, large copepods, and hyperiid amphipods dominated

5 cool year diets, accounting for >70% of the average similarity among pairs of cool year diets.

6 These same five prey categories accounted for > 66% of the average dissimilarity between

7 warm and cool year diets.

8 5.4.3. Energy content 9 Juvenile cod energy densities differed among years (p < 0.001). Even though energy densities

10 were available for only one year with warm conditions, average energy density (20.29 kJ/g dry

11 mass) observed in 2005, was lower than the averages observed in three cool years ( 21.04 –

12 20.72 kJ/g). Post-hoc analysis indicated the energy densities in each of the cool years differed

13 from the warm year (p < 0.05). The annual variation in energy densities in juvenile Pacific cod

14 was correlated with values previously reported for juvenile pollock in the same four years, but

15 the correlation was marginally significant (P =0.051, R2 = 0.901; Fig. 7).

16 5.5. Discussion 17 Oceanographic conditions in the Bering Sea during warm years led to conditions that favored

18 large size, but low nutritional condition in juvenile cod, whereas the opposite was true for cool

19 years. Oceanographic conditions driving the differences in size and nutritional status of juvenile

20 cod include temperature and the availability of prey. Depth-averaged water temperatures at

21 mooring M2 in the southeastern Bering Sea were near 6° C in summer during the warm years

22 (2000 – 2005) compared with 4° C in the cool years (2007 – 2010; Overland et al., 2012).

210

211

1 Perhaps more importantly, ice retreat from the survey region was delayed in the cool years.

2 Prey fields available to pelagic juveniles in cool years had a higher abundance of large copepods

3 and adult euphausiids compared with warm years (Coyle et al., 2011). Consumption of these

4 abundant and lipid-rich prey led to increased energy densities age-0 pollock (Heintz et al.,

5 2013). Our examination of age-0 Pacific cod on the Bering Sea shelf indicates a similar

6 response.

7 It is important to recognize that size differences among years may not necessarily reflect

8 differences in growth rate. The effects of temperature on spawn timing could contribute

9 significantly to the size differences. For instance, Neidetcher et al. (in press) found that cod

10 spawning occurred earlier in warm (2005) than in cool (2007) years. Walleye pollock also

11 spawned earlier during the years with anomalously warm sea temperatures (Smart et al., 2012).

12 Consequently juvenile pollock appeared in trawls approximately 15 days earlier in warm years

13 than they did in cool years. This early appearance would increase the amount of time in which

14 fish had to grow prior to sampling in BASIS surveys. Furthermore, Farley et al. (2011) ascribed

15 that relatively large size of juvenile Bristol Bay sockeye salmon that were distributed

16 throughout the eastern Bering Sea during anomalously warm years was due to earlier ice break

17 up in the freshwater lakes leading to earlier outmigration to the marine environment, providing

18 more time for growth. Because growth rate of fish is a function of temperature and adequate

19 prey supplies, warmer temperatures may augment these differences in size assuming adequate

20 food supplies were available.

211

212

1 It is interesting that across years, larger body sizes of cod were associated with lower, not

2 higher, energy densities. In both field and laboratory studies favorable growth conditions

3 frequently result in positively correlated size and energy storage (Hurst and Conover, 2003;

4 Sogard and Spencer, 2004). However, the energy density pattern observed in cod appears

5 consistent with that observed in juveniles of at least two other Bering Sea fishes. Energy

6 density of age-0 walleye pollock was also higher in cool years (although body sizes were more

7 similar across climate conditions, Heintz et al., 2013). Juvenile sockeye salmon in cool years had

8 smaller body sizes but higher energy content than in warm years as observed here for juvenile

9 cod (Farley et al., 2011). The consistent pattern of climate effects on size and energy content in

10 these co-occurring species could reflect an adaptation to growth opportunity to reduce

11 predation risk in at larval and juvenile life history stages and one of lipid storage during late

12 summer to increase overwinter survival (Siddon et al., (2013Alternatively, the pattern could be

13 related to the significant differences in juvenile fish diets observed across climate states in the

14 EBS.

15 Increased lipid storage in juvenile cod was coincident with a shift in the quality of prey

16 consumed during warm and cool years. In warm years, juvenile cod consumed prey that

17 averaged between 2% and 4% lipid; whereas,the predominate prey consumed in cool years

18 averaged 4% to 12% lipid (Heintz et al., 2013). These same factors also account for the

19 observation that walleye pollock and juvenile salmon in the eastern Bering Sea, sampled at the

20 same time as the juvenile cod, had higher energy densities in cool years (Heintz et al., 2013).

21 For instance, juvenile (age-0) walleye pollock fed primarily on small zooplankton during warm

22 years and large zooplankton during cool years (Coyle et al., 2011). Juvenile salmon diets were

212

213

1 dominated by age-0 walleye pollock during warm years, but their dietsshifted to large

2 zooplankton during cool years (Coyle et al., 2011; Farley and Trudel, 2009; Farley et al., 2009).

3 Thus, evidence indicates that when abundance of large zooplankton on the eastern Bering Sea

4 shelf is low during warm years (Coyle et al., 2011), apex predators feed on forage fish (mainly

5 age-0-walleye pollock) that were highly abundant within the middle domain during years with

6 anomalously warm spring and summer sea temperatures (Moss et al., 2009; Hollowed et al.,

7 2012). This shift in diet of some larger age-0 and juvenile fish from zooplankton during years

8 with cool sea temperatures to age 0 pollock during years with warm sea temperatures may be

9 one explanation for increased mortality and low recruitment of these fish during years with

10 warm sea temperatures.

11 We found that juvenile cod diets were more diverse during years with cool temperatures than

12 those with warm temperatures. This result is likely a function of the shift in these prey within

13 the eastern Bering Sea ecosystem between warm and cool periods. For instance, age-0 walleye

14 pollock dominated juvenile cod diets during warm years in the inner and middle domain but

15 were absent from the diets in cool years. Age-0 walleye pollock were found to be widely

16 distributed in surface waters during warm years within middle and inner domains of the Bering

17 Sea but declined in abundance in surface waters during cool years (Parker-Stetter et al. 2013;

18 Hollowed et al. 2012). During cool years, hyperiid amphipods and large copepods were some of

19 the dominant prey for juvenile cod found in the middle domain. Recent research on hyperiid

20 amphipds suggests that they were always absent in the inner domain, but their abundance

21 increased in the middle domain (Coyle et al., 2011; Pinchuk et al., 2013). There was a

22 consistent increase in C. marshallae in both the inner and middle domains during years with

213

214

1 cool sea temperatures, but the increase in their abundance was much greater in the middle

2 domain (Coyle et al., 2011). The distribution of juvenile cod was relatively stable and did not

3 shift markedly in response to variation in spring and summer temperatures on the shelf (Hurst

4 et al., 2012), thus the significant differences in diet among domains during cool years was likely

5 related to shifts in prey available to juvenile cod within each domain.

6 Large size and high energy reserves for age-0 fish enable them to withstand severe energy

7 deficits prior to and during winter (Sogard and Olla, 2000; Hurst, 2007). For instance, smaller

8 age-0 pollock from southeast Alaska that lacked sufficient energy to meet metabolic demands

9 during winter had higher size-selective mortality (Heintz and Vollenweider, 2010). Smaller,

10 leaner juvenile Bristol Bay sockeye salmon also had increased mortality during winter (Farley et

11 al., 2011). For age-0 pelagic fish, prey quality and quantity are important because there is a

12 small time window between the completion of metamorphosis and the onset of winter in which

13 juveniles can provision themselves (Siddon et al., 2013). Therefore, consumption of high

14 quality prey should improve fitnessin age-0 fish, reduce their energy deficits over winter and

15 lead to increased survival.

16 Understanding the relationships between climate, ecosystem productivity, and juvenile fish

17 condition are important because condition of juvenile fishes prior to winter is believed to

18 control recruitment in the Bering Sea (e.g. walleye pollock; Heintz et al., 2013). Moreover,

19 models relating environmental factors to survival indicate late summer temperatures account

20 for more variation in recruitment than either the timing of ice retreat or a spring transition

21 index (Mueter et al., 2011). We found a positive correlation between juvenile cod and pollock

214

215

1 energy density. This result, coupled with the understanding that walleye pollock and Pacific cod

2 recruitment appear synchronous (Mueter et al., 2007), suggests that condition of juvenile cod

3 prior to winter may be an important predictor of overwinter survival in the Bering Sea.

4 5.6 Conclusion 5 Our results provide evidence that condition of juvenile gadids prior to winter is a function of

6 prey quality and sea temperature regimes on the eastern Bering Sea shelf. Measures of fish

7 condition during late summer and fall may also be a good predictor of recruitment to age-1 fish.

8 Periods of anomalously warm sea temperatures were associated with decreased nutritional

9 quality of available prey resulting in reduced energetic status of juvenile gadids on the Bering

10 Sea shelf, suggesting increased risk of overwinter mortality. Our examination of juvenile cod

11 size, diet, and energetic status, provided results that are similar to those from studes on

12 juvenile pollock (Moss et al. 2009; Coyle et al. 2011; Hunt et al. 2011; Heintz et al 2013; Siddon

13 et al. 2013), suggesting that the common mechanisms regulating gadid recruitment on the

14 eastern Bering Sea shelf are climate condition, prey quality and quantity, and caloric density of

15 gadids prior to winter.

16 5.7 Acknowledgments 17 We thank all of the people involved in collecting, sorting, and processing the samples used in

18 this study. This includes the crews of the FV Sea Storm, FV Northwest Explorer, and NOAA ship

19 Oscar Dyson. This research is contribution NPRB XXX and BEST-BSIERP XXX. References to

20 trade names does not imply endorsement by the National Marine Fisheries Service, NOAA. The

21 findings and conclusions in this paper are those of the authors do not necessarily represent the

22 views of NMFS or NOAA.

215

216

1 5.8 References 2 Abookire, A.A., Duffy-Anderson, J.T., Jump, C.M. 2007. Habitat associations and diet of young-

3 of-the-year Pacific cod (Gadus macrocephalus) near Kodiak, Alaska. Mar. Biol. 150, 713-

4 726.

5 Chuchukalo, V.I., and Volkov, A.F. 1986. Manual for the study of fish diets. Vladivostok: TINRO,

6 32 p. (in Russian).

7 Coyle, K.O., Eisner, L., Mueter, F., Pinchuk, A., Janout, M., Cieciel, K., Farley, E., and Andrews,

8 A.G. 2011. Climate change in the southeastern Bering Sea: impacts on pollock stocks and

9 implications for the Oscillating Control Hypothesis. Fish. Oceanogr. 20, 139-156.

10 Doyle, M.J., Piquelle, S.J., Mier, K.L., Spillane, M.C., and Bond, N.A. 2009. Larval fish abundance

11 and physical forcing in the Gulf of Alaska, 1981 – 2003. Prog. Oceanogr. 80, 163-187.

12 Farley, E.V., Murphy, J.M., Wing, B.W., Moss, J.H., and Middleton, A. 2005. Distribution,

13 migration pathways, and size of western Alaska juvenile salmon along the eastern

14 Bering Sea Shelf. Alaska Fish. Res. Bull. 11, 15-26.

15 Farley, E.V. Jr., and Trudel, M. 2009. Growth rate potential of juvenile sockeye salmon in

16 warmer and cooler years on the eastern Bering Sea shelf. J. Mar. Biol. Article 640125, 10

17 p.

18 Farley, E.V. Jr., Murphy, J.M., Moss, J., Feldmann, A., and Eisner, L. 2009. Marine ecology of

19 western Alaska juvenile salmon. In: Krueger, C.C., Zimmerman, C.E. (Eds.), Pacific

20 Salmon: Ecology and Management of Western Alaska’s Populations. Am. Fish. Soc.

21 Symp. 70, Bethesda, Maryland.

216

217

1 Farley, E.V., Jr., Starovoytov, A., Naydenko, S., Heintz, R., Trudel, M., Guthrie, C., Eisner, L., and

2 Guyon, J.R. 2011. Implications of a warming eastern Bering Sea for Bristol Bay sockeye

3 salmon. ICES J. Mar. Sci. doi:10.1093/icesjms/fsr021.

4 Heintz, R.A., Siddon, E.C., Farley, E.V., and Napp, J.M. 2013. Correlation between recruitment

5 and fall condition of age-0 pollock (Theragra chalcogramma) from the eastern Bering

6 Sea under varying climate conditions. Deep-Sea Res. II 94, 150-156.

7 Heintz, R.A., and Vollenweider, J.J. 2010. Influence of size on the sources of energy consumed

8 by overwintering walleye pollock (Theragra chalcogramma). J. Exp. Mar. Biol. Ecol.

9 393(1-2), 43-50.

10 Hollowed, A.B., Barbeaux, S., Farley, E., Cokelet, E.D., Kotwicki, S., Ressler, P.H., Spital, C., and

11 Wilson, C. 2012. Effects of climate variations on pelagic ocean habitats and their role in

12 structuring forage fish distributions in the Bering Sea. Deep-Sea Res. II 65-70, 230-250.

13 Hunt, G.L., Jr., Coyle, K.O., Eisner, L.B., Farley, E.V., Heintz, R.A., Mueter, F., Napp, J.M.,

14 Overland, J.E., Ressler, P.H., Salo, S., and Stabeno, P.J. 2011. Climate impacts on eastern

15 Bering Sea foodwebs: a synthesis of new data and an assessment of the Oscillating

16 Control Hypothesis. ICES J. Mar. Sci. doi:10.1093/icesjms/fsr036.

17 Hunt, G.L., Stabeno, P., Walters, G., Sinclair, E., Brodeur, R.D., Napp, J.M., and Bond, N.A. 2002.

18 Climate change and control of the southeastern Bering Sea pelagic ecosystem. Deep-Sea

19 Res. II 49, 5821-5853.

20 Hurst, T.P. 2007. Causes and consequences of winter mortality in fishes. J. Fish Biol. 71, 315-

21 345.

217

218

1 Hurst, T.P. and Conover, D.O. 2003. Seasonal and interannual variation in the allometry of

2 energy allocation in juvenile striped bass. Ecol. 84(12), 3360-3369.

3 Hurst, T.P., Cooper, D.W., Scheingross, J.S., Seale, E.M., Laurel, B.J., and Spencer, M.L. 2009.

4 Effects of ontogeny, temperature, and light on vertical movements of larval Pacific cod

5 (Gadus macrocephalus). Fish. Oceanogr. 18, 301-311.

6 Hurst, T.P., Moss, J.H., and Miller, J.A. 2012. Distributional patterns of 0-group Pacific cod

7 (Gadus macrocephalus) in the eastern Bering Sea under variable recruitment and

8 thermal conditions. ICES J. Mar. Sci. 69(2), 163-174.

9 Kinder, T.H., and Schumacher, J.D. 1981. Hydrographic structure over the continental shelf of

10 the southeastern Bering Sea. In: Hood, D.W., Calder. J.A. (Eds.), The Eastern Bering Sea

11 Shelf, Oceanography and Resources, vol. 1. University of Washington, pp. 31-52.

12 Laurel B.J., Stoner, A.W., Ryer, C.H., Hurst, T.P., and Abookire, A.A. 2007. Comparative habitat

13 associations in juvenile Pacific cod and other gadids using seines, baited cameras and

14 laboratory techniques. J. Exp. Mar. Biol. Ecol. 351, 42-55.

15 Moss, J.H., Farley, Jr, E.V., and Feldmann, A.M. 2009. Spatial distribution, energetic status, and

16 food habits of eastern Bering Sea age-0 walleye pollock. Trans. Am. Fish. Soc. 138, 497-

17 505.

18 Mueter, F.J., Boldt, J.L., Megrey, B.A., and Peterman, R.M. 2007. Recruitment and survival of

19 northeast Pacific Ocean fish stocks: temporal trends, covariation, and regime shifts. Can.

20 J. Fish. Aquat. Sci. 64, 911-927.

218

219

1 Mueter, F.J., Bond, N.A., Ianelli, J.N., Hollowed, A.B. 2011. Expected declines in recruitment of

2 walleye pollock (Theragra chalcogramma) in the eastern Bering Sea under future

3 climate change. ICES J. Mar. Sci. 69(6), 1284-1296.

4 Neidetcher, S.K., Hurst, T.P., Ciannelli, L., and Logerwell, E.A. in press. Spawning phenology and

5 geography of Pacific cod (Gadus macrocephalus) in the eastern Bering Sea. Deep- Sea

6 Res. II

7 Overland, J.E., Salo, S.A., Kantha, L.H., Clayson, C.A. 1999. Thermal stratification and mixing on

8 the Bering Sea shelf. In: Loughlin, T.R, Ohtani, K. (Eds.), Dynamics of the Bering Sea: A

9 Summary of Physical, Chemical, and Biological Characteristics, and a Synopsis of

10 Research on the Bering Sea, University of Alaska Sea Grant, AK-SG-99-03. N. Pac. Mar.

11 Sci. Org. (PICES), pp. 129-146.

12 Overland, J.E., Wang, M., Wood, K.R., Percival, D.B., Bond, N.A. 2012. Recent Bering Sea warm

13 and cold events in a 95-year perspective. Deep-Sea Res. II 65-70, 6-13.

14 Pinchuk, A.I., Coyle, K.O., Farley, E.V., and Renner, H.M. 2013. Emergence of the Arctic Themisto

15 libellula (Amphipoda: Hyperiidae) on the southeastern Bering Sea shelf as a result of the

16 recent cooling, and its potential impact on the pelagic food web. ICES J. Mar. Sci.

17 doi:10.1093/icesjms/fst031.

18 Post, J.R., and Parkinson, E.A. 2001. Energy allocation strategy in young fish: allometry and

19 survival. Ecol. 82(4), 1040-1051.

20 Siddon, E.C., Heintz, R.A., and Mueter, F.J. 2013. Conceptual model of energy allocation in

21 walleye pollock (Theragra chalcogramma) from age-0 to age-1 in the southeastern

22 Bering Sea. Deep-Sea Res. II 94, 140-149.

219

220

1 Smart, T.I., Duffy-Anderson, J.T., Horne, J.K. 2012. Alternating temperature states influence

2 walleye pollock early life stages in the southeastern Bering Sea. Mar. Ecol. Prog. Ser.

3 455, 257-267.

4 Sogard, S.M. 1997. Size-selective mortality in the juvenile stage of teleost fishes: a review. Bull.

5 Mar. Sci. 60(3), 1129-1157.

6 Sogard, S.M., and Olla, B.L. 2000. Endurance of simulated winter conditions by age-0 walleye

7 pollock: effects of body size, water temperature and energy storage. J. Fish Biol. 56, 1-

8 21.

9 Sogard, S.M., and Spencer, M.L. 2004. Energy allocation in juvenile sablefish: effects of

10 temperature, ration and body size. J. Fish. Biol. 64(3), 726-738.

11 Stabeno, P.J., Farley, E.V., Kachel, N.B., Moore, S., Mordy, C.W., Napp, J.M., Overland, J.E.,

12 Pinchuk, A.I. and Sigler, M.F. 2012. A comparison of the physics of the northern and

13 southern shelves of the eastern Bering Sea and some implications for the ecosystem.

14 Deep-Sea Res. II 65-70, 14-30.

15 Volkov, A.F., Efimkin, A.Y., Kuznetsova, N.A. 2007. Results from research on the diets of Pacific

16 salmon in 2002(2003)-2006 under the BASIS program. Izvestia TINRO, 151, 365-402. (in

17 Russian).

220

221

2 Table 1. Sampling summary

Year Vessel(s) Start Date End Date

2003 Sea Storm 31-August 28-September

2004 Sea Storm 14-August 28-September

2005 Sea Storm 14-August 06-October

2006 Sea Storm/Northwest Explorer 16-August 20-September

2007 Sea Storm 15-August 08-October

2008 Oscar Dyson 11-September 27-September

2010 Oscar Dyson 18-August 24-September

2011 Oscar Dyson 23-August 15-September

221

222

2 Table 2. Prey taxa identified in all stomachs analyzed and the associated diet categories used 3 for construction of Bray-Curtis similarity matrix.

Diet categories Prey taxa

Barnacle Cirripedia cyprid

Barnacle Balanidae

Bivalvia Bivalvia

Chaetognath Chaetognatha

Chaetognath Parasagitta elegans

Crab zoea Brachyura

Crab zoea Decapoda

Crab zoea Decapoda larvae

Crab zoea Paguridae

Crab zoea Paguridae zoea

Crab zoea Pagurus sp.

Crab zoea Chionoecetes opilio

Euphausiid Euphausiacea

Euphausiid Euphausiacea eggs

Euphausiid Thysanoessa inermis

Euphausiid Thysanoessa inspinata

Euphausiid Thysanoessa longipes

Euphausiid Thysanoessa raschii

Euphausiid Thysanoessa sp.

Euphausiid Thysanoessa spinifera

222

223

Gammarid Gammaridae

Gammarid Gammaridea

Gelatinous Cnidaria

Gelatinous Cumacea

Hyperiid Hyperia sp.

Hyperiid Themisto libellula

Hyperiid Themisto pacifica

Large copepod Calanus glacialis

Large copepod Calanus marshallae

Large copepod Epilabidocera amphitrites

Large copepod Eucalanus bungii

Large copepod Metridia pacifica

Large copepod Neocalanus cristatus

Large copepod Neocalanus flemingeri

Large copepod Neocalanus plumchrus

Large copepod Neocalanus sp.

Mysid Mysidacea

Other fish Digested fish

Other fish Fish

Pollock Gadus chalcogrammus

Polychaete Polychaeta

Pteropod Limacina helicina

Shrimp Caridea

Small copepod Acartia clausi

223

224

Small copepod Centropages abdominalis

Small copepod Copepoda

Small copepod Oithona similis

Small copepod Pseudocalanus minutus

Small copepod Pseudocalanus newmani

Small copepod Pseudocalanus sp.

224

225

2 Table 3. Sample size (n) for length, diet, and energetic status of juvenile cod sampled in the

3 Bering Sea during August - September 2003 - 2011. The year 2009 is omitted due to low sample

4 size.

Year Length Diet Energy

2003 102 26 0

2004 603 35 0

2005 1,480 50 18

2006 2,918 444 0

2007 587 195 28

2008 421 105

2010 2013 326 18

2011 368 40 18

225

226

2 Table 4. R statistics representing the multivariate distance between Pacific cod diets in pairs of

3 domains from the Bering Sea in warm and cool years. Domains prefixed with W were sampled

4 in warm years (2003-2005) and those prefixed with C were sampled in cool years (2007-2011).

5 * indicates p <0.01, “a” indicates there were < 999 permutations possible.

W-Outer W-Middle W-Inner C-Outer C-Middle

W-Middle 0.091a

W-Inner 0.071a 0.048a

C-Outer -0.151a 0.365* 0.663*

C-Middle 0.009a 0.490* 0.765* 0.019

C-Inner -0.039a 0.351* 0.578* -0.037a 0.238*

226

227

100 m 50 m

2 3 Figure 1. Survey grid (core stations) for the 2003 to 2011 Bering-Aleutian Salmon International 4 Survey. Contour lines are associated with the 50 m and 100 m bottom depths delineating the 5 inner (< 50 m depth), middle (50 m to 100 m depth), and outer (> 100 m depth) domains.

227

228

2 3

4 Figure. 2

228

229

)

m (

e L 75

2003 2004 2005 2006 2007 2008 2010 2011 Year 4 5

6 Figure 3. Mean lengths (mm) and 95% confidence intervals of juvenile cod captured during 7 2003 to 2011 in the eastern Bering Sea.

229

230

) 80

(

h t

g 75

e L 70

60 CI CM CO AI AM AO WI WM WO Climate and Domain 3

4 Figure 4. Mean lengths (mm) and 95% confidence intervals of juvenile cod within cool (C),

5 average (A), and warm (W) climate periods and among domains (inner (I), middle (M), and

6 outer (O)).

230

231

Inner Middle

n e

n -1

R = 0.578; P = 0.003 o

p R = 0.490; P = 0.001

m -2

o -1 0 1 2 3 C

Outer d

n 1 Climate

o c

e Cool

S 0 Average -1 Warm R = -0.151; P = 0.786 -2 -1 0 1 2 3 First Component

2 3

4 Figure 5. MDS plots of juvenile cod diets from warm, average, and cool years in the Bering Sea, 5 showing comparisons among domains. Each point represents the average diet of juvenile cod 6 sampled at a specific station within the domain during the appropriate climatic condition. The 7 MDS model was fit to all domains simultaneously with an overall stress = 0.17. R statistics 8 indicate the distance between warm and cool years in each domain.

231

232

Shrimp Warm years Shrimp Mysiid Mysid Cool yearsWarm years Cool years PteropodPteropod

L. copepod Large copepod Hyperiid amphipod

Hyperiid amphipodAdult euphausiid

Crab zoea Adult euphausiid

Pollock Crab zoea

80 60 40 20 0 20 40 60 Pollock

Percentage by mass -60 - 4040 - 20 0 20 40

Percentage by mass

4 Figure 6. Average diets of juvenile cod sampled in the Bering Sea in warm (2003-2005) and 5 cool (2007-2011) years. Diet categories account for > 90% of the overall dissimilarity between 6 climatic states and are listed from the bottom in order of their contribution to dissimilarity.

232

233

Pollock energy density kJ/g w wt

2010 2011 5

2007 4 R² = 0.7508 2005 3 4 4.2 4.4 4.6 4.8 5 Cod energy density kJ/g w wt

2 Figure 7. Correlation between juvenile cod and pollock energy densities in one warm (2005)

3 and three cool (2007, 2010 and 2011) years. Symbols reflect the mean energy density each

4 year, error bars are ± 1 s.e.. Symbols may obscure error bars.

233

234

2 Chapter 6 - Ecology and taxonomy of the early life stages of 3 arrowtooth flounder (Atheresthes 2 stomias) and Kamchatka flounder 4 (A. evermanni) in the eastern Bering Sea a a b a b a 5 Lisa De Forest , J.T. Duffy-Anderson , R.A. Heintz , A.C. Matarese , E.C. Siddon , T.I. Smart , and I.B. a 6 Spies 7 a 8 Alaska Fisheries Science Center, National Marine Fisheries Service, National Oceanic and Atmospheric 9 Administration, Seattle, WA, USA 98115. bAlaska Fisheries Science Center, Auke Bay Laboratories, National Marine 10 Fisheries Service, Juneau, AK, USA 99801. 11 12 Submitted to DSRII 2014 13 14

15 6.1 Abstract 16 17 Arrowtooth flounder (Atheresthes stomias) and Kamchatka flounder (A. evermanni) are closely 18 related flatfish species that co-occur in the eastern Bering Sea. The inability to distinguish larvae 19 and early juveniles of these species has precluded studies of ecology for the early life stages of 20 both species in the eastern Bering Sea. In this study, we developed a genetic technique to 21 identify the larvae and early juveniles of the two species using mtDNA cytochrome oxidase 22 subunit I (COI). Genetically identified specimens were then examined to determine a visual 23 identification method based on pigment patterns and morphology. Specimens 6.0–12.0 mm SL 24 and ≥ 18.0 mm SL can be identified to the species level, but species identification of individuals 25 12.1–17.9 mm SL by visual means alone remains elusive. The distribution of larvae (< 25.0 mm 26 SL) of both arrowtooth flounder and Kamchatka flounder is similar in the eastern Bering Sea; 27 however, juvenile (≥ 25.0 mm SL) Kamchatka flounder occur closer to the shelf break and in 28 deeper water than juvenile arrowtooth flounder. Condition was measured in larvae and 29 juveniles of each species as analyzing lipid content (%) and energy density (kJ/g dry mass). 30 Kamchatka flounder larvae on average had higher lipid content than arrowtooth flounder 31 larvae, but were also larger on average than arrowtooth flounder larvae in the summer. When 32 corrected for length, both species had similar lipid content in the larval and juvenile stages.

234

235

1 6.2 Introduction

2 Arrowtooth flounder are large (max. size 86 cm total length), predatory flatfish that occur from 3 off the coast of central California, north to the Bering Sea, and west to the Kamchatka 4 Peninsula, Russia (Mecklenburg et al., 2002). Kamchatka flounder are similar in size, but occur 5 primarily in the western Bering Sea with lesser occurrence in the eastern Bering Sea and along 6 the Aleutian Islands (Mecklenburg et al., 2002). In the Bering Sea, arrowtooth flounder have 7 been documented to feed primarily on juvenile and adult walleye pollock (Gadus 8 chalcogrammus), euphausiids, and various shrimps (Yang and Livingston, 1986). They, in turn, 9 are consumed by Alaska skates (Bathyraja parmifera) and sleeper sharks (Somniosus pacificus) 10 as adults, and by Pacific cod (Gadus macrocephalus) and walleye pollock as juveniles (Spies et 11 al., 2011). The diet of Kamchatka flounder is similar and often considered identical to 12 arrowtooth flounder (Yang and Livingston, 1986). Both their predation impact and their role as 13 prey for other organisms indicate that arrowtooth flounder and Kamchatka flounder are 14 important constituents of the Bering Sea ecosystem (Aydin et al., 2007). The directed fishing 15 effort for arrowtooth flounder has increased in recent years, as has the retention of adults 16 caught in other commercial fisheries, possibly in response to an approximate eight-fold increase 17 in adult arrowtooth flounder biomass in the Bering Sea since the early 1980s (Spies et al., 18 2011). Kamchatka flounder also experience some fishing pressure as directed commercial 19 fishing has developed in recent years (Wilderbuer et al., 2011). Despite their many similarities, 20 these two species can be separated as adults by the number of gill rakers on the second upper 21 gill arch (one in Kamchatka flounder, two to three in arrowtooth flounder) and the visibility of 22 the dorsal-most eye from the blind side (top of eye is visible in arrowtooth flounder, no part of 23 eye visible in Kamchatka flounder) (Yang, 1988). Using electrophoretic examiniation of alleles at 24 different loci, Ranck et al. (1986) determined that arrowtooth flounder and Kamchatka flounder 25 were two genetically-distinct species and, based on the distinct distributions of certain alleles, 26 there was no hybridization between the two species. However, adults of arrowtooth flounder 27 and Kamchatka flounder collected in scientific and commercial catches were often recorded as 28 either all arrowtooth flounder or the genus

29 Atheesthes (Yang and Livingston, 1986; Spies et al., 2011). In 1991, adult arrowtooth flounder and 30 Kamchatka flounder individuals began to be separated as distinct species by the Alaska Fisheries 31 Science Center (AFSC) Groundfish Assessment Program (Zimmerman and Goddard, 1996). As of 32 summer 2011, fishing management regulations dictated that arrowtooth flounder and Kamchatka 33 flounder must be separated and recorded as distinct species in commercial fishing hauls due to the 34 emergence of a directed fishery for adult Kamchatka flounder (Wilderbuer et al., 2011). While 35 adults of both species have been well described, the early life stages in the eastern Bering Sea have 36 not been. High recruitment success may be the cause of the increase in population size of 37 arrowtooth flounder in recent years (Spies et al., 2011). Understanding the early life stages is

235

236

1 important to understanding recruitment and the first step to understanding recruitment in 2 arrowtooth flounder is to identify and separate the early life stages from Kamchatka flounder. Our 3 current study had the following objectives: 1) use genetic methods to identify to the species level 4 larval and early juvenile Atheresthes collected in the eastern Bering Sea, 2) use the genetically 5 identified individuals to develop a visual identification method that allows identification of historical 6 and future samples, 3) describe the distribution and abundance of the early life stages of both 7 arrowtooth flounder and Kamchatka flounder, and 4) examine condition as indicated by length 8 adjusted dry mass, lipid content (%), and energy density of larvae and juveniles to determine 9 differences in nutritional quality between species.

10 6.3 Materials and Methods

11 6.3.1 Specimen collection 12 All specimens in this study were either directly collected at sea or were sorted from previously 13 collected preserved samples collected on previous AFSC ichthyoplankton surveys. The 14 specimens collected at sea were obtained on spring and summer cruises from 2006 to 2010 by 15 scientists on AFSC, Bering Ecosystem Study (BEST), and Bering Sea Integrated Ecosystem 16 Research Program (BSIERP) cruises (Table 1). Specimens were collected opportunistically, 17 resulting in varied collection and preservation methods during the study. Methods of collection 18 included using bongo nets (60-cm; 500-µm mesh), Multiple Opening/Closing Net and 2 19 Environmental Sensing System (MOCNESS; 1-m with 500-µm mesh; Wiebe et al., 1976), and 20 surface and midwater trawls. Specimens removed from the bongo net were either placed 21 directly into 100% non denatured ethanol or into 5% formalin after an eyeball was removed. 22 Sampling from the MOCNESS involved removing specimens only from the drogue net, which 23 was open for the duration of the tow; all specimens were immediately frozen at -80º C. The 24 surface and midwater rope trawls sampled above and below the pycnocline; all specimens 25 collected were immediately frozen at -80º C. Specimens sorted from previously collected 26 preserved samples were chosen from cruises occurring from 1994 to 2010 (Table 1). The 27 samples were collected on previous AFSC ichthyoplankton surveys using 60-cm bongo nets, 2 28 MOCNESS tows, Methot trawls (Methot, 1986), Tucker trawls (1-m with 500-µm mesh; Tucker, 29 1951), and a modified bottom trawl (which samples the midwater). Sampling protocol during 30 the cruises, along with sample handling and sorting, follows Matarese et al. (2003). All 31 specimens used in this study were classified as either larvae (3.0–24.9 mm SL) or juveniles (≥ 32 25.0 mm SL) (Matarese et al., 1989).

34 6.3.2 Genetic identification 35 Species determination was based on a restriction enzyme digest that cuts DNA at a specific 36 nucleotide sequence that is present in arrowtooth flounder but not Kamchatka flounder. The

236

237

1 test was based on an initial sequencing of three adult specimens from each species, followed by 2 testing on 20 adult specimens of which the species identification was known. Fin tissue was 3 used for DNA extraction in all adults and was performed using Qiagen DNA extraction kits 4 (Qiagen, Inc. Valencia, CA). For larval and early juveniles, a single eyeball (from either side of 5 head) removed from either fresh, frozen, or 100% non-denatured ethanol preserved specimens 6 was used for DNA extraction. The rest of the specimen body was placed back in its original 7 preservative or, if frozen, photographed for future visual identification work. Larval DNA was 8 extracted from a single eyeball using a standard Chelex DNA extraction protocol. The eyeball 9 was submerged in 150 µl of 10% (w/v) Chelex ® 100 resin (BIO126 RAD Laboratories, Hercules, 10 CA), heated to 60° C for 20 min followed by 103° C for 25 min. A 750 bp segment of cytochrome 11 oxidase subunit I (COI) was amplified using the following primers: COI_RajaF (5’- 12 CCGCTTAACTCTCAGCCATC-3’) and COI_RajaR (5’-TCAGGGTGACCAAAGAATCA-3’; Spies et al., 13 2006). Polymerase chain reactions (PCR) were conducted in a 10 µl volume, with approximately 14 100 ng DNA (1 µl taken from the Chelex supernatant), 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 3.0 15 mM MgCl2, 1.5 mM dNTPs, 0.5 pM of each primer, and 0.5 U Bioline Taq polymerase (Bioline 16 USA, Inc., Boston, MA). PCR amplification using Chelex extractions can fail if the supernatant 17 contains impurities such as proteins. Therefore, failed samples were repeated with a 1:10 135 18 or 1:100 dilution of the DNA, which typically resulted in success. The thermal cycling protocol 19 consisted of 94° C for 2 min, followed by 40 cycles of 94° C (30 seconds), 57°C (30 sec), and 72° 20 C (30 sec). DNA sequencing of amplified DNA from adult specimens was performed using 21 COI_Raja primers at the University of Washington High Throughput Genomics Center 22 (www.htseq.org). Following PCR, DNA was digested using Bmt1 (New England Biolabs, Ipswich, 23 MA) according to manufacturer’s instructions, in a total volume of 10:l with 2:l PCR product 24 and4Uof Bmt1 at 37° C for 1 h. Identification was based on either the presence of a single 750 25 bp fragment (Kamchatka flounder) or two fragments consisting of 450 and 300 bp (arrowtooth 26 flounder).

27 This protocol was performed in a laboratory or at sea using an Agilent 2100 Bioanalyzer with a 28 DNA 1000 kit (Agilent Technologies, Santa Clara, CA) and a BIO-RAD DNA Engine Thermalcycler.

29 6.3.3 Visual identification 30 Specimens identified using genetic methods were examined to develop a visual identification 31 method based on morphological and pigmentation characters. Specimens were examined both 32 with and without knowledge of their genetic identification to determine species-specific 33 characters. Each specimen was critically evaluated with regards to morphology and pigment 34 (such as the number of melanophores in a patch or the general shape and pattern of 35 melanophore patches). Only specimens preserved in the same preservation medium were 36 directly compared to one another. This was to minimize the identification of preservation 37 artifacts as distinguishing species characters.

237

238

1 Morphometric measurements were taken from arrowtooth flounder and Kamchatka flounder 2 larvae chosen from a single cruise. Only specimens initially preserved in formalin were used for 3 morphometric measurements because the other preservation methods (100% ethanol and 4 immediate freezing) shrink tissues and distort morphology. The following measurements were 5 taken: standard length (SL), preanal length (tip of snout to anus), head length, snout length, eye 6 diameter, and body depth (vertical measurement taken at pectoral fin; BD). The measurements 7 were taken using a calibrated digital image analysis system consisting of a video camera 8 attached to a stereo microscope and a computer with image analysis software. All 9 measurements were recorded to the nearest 0.1 mm. A two sample t-test was used to identify 10 statistically significant differences (p < 0.05) between each morphological measurement for the 11 two species. Developmental stage terminology follows Kendall et al. (1984).

12 6.3.4 Distribution 13 Selected larval Atheresthes collected on previous AFSC cruises conducted from 1994 to 2010 14 (Table 1) were identified to the species level using pigmentation and morphological characters 15 identified with the visual identification method. The locations of all specimens identified to 16 species were mapped and a mean distribution based on abundance was calculated for each 17 species using ArcGIS software.

18 6.3.5 Measures of Condition 19 6.3.6 Biological sampling 20 Larval and juvenile Atheresthes were collected from five research surveys conducted in the 21 summer and early fall from 2008 to 2010 (Table 1). All specimens were measured to the nearest 22 0.01 mm SL. Specimens < 30.0 mm SL were identified to species using the genetic methods 23 previously described, while specimens > 30.0 mm SL were identified morphologically using the 24 number of gill rakers on the upper arch of the second gill raker. All specimens had an eyeball 25 removed, even if identified visually. Stomach contents were removed prior to chemical analysis.

27 6.3.7 Dry mass 28 Larvae were individually dried at 60ºC in a drying oven until their weight was stabilized. 29 Juveniles were dried at 135ºC using a LECO Thermogravimetric Analyzer (TGA) 601 or 701, 30 which provided percent moisture values used to convert wet mass to dry mass equivalents.

31 6.3.8 Estimation of lipid content (%) 32 For larvae, a sulfo-phospho-vanillin (SPV) colorimetric analysis (Van Handel, 1985) was 33 performed to determine lipid content (%). Dried material was sonicated in 2:1 (by 196 volume) 34 chloroform:methanol solvent in glass centrifuge tubes for 60 min. Washes of 0.88% KCL and 1:1 35 (by volume) methanol:water were performed on the extracts as in the modified Folch 36 extraction method (Vollenweider et al., 2011). Resulting chloroform extracts were evaporated

238

239

1 in a LabConco RapidVap for 30 min at 40 ºC and 250 mbar 200 until reduced to approximately 1 2 ml in volume. Extracts were evaporated to dryness in 12 mm test tubes on a heating block at 75 3 ºC and then allowed to cool. Concentrated sulfuric acid was added to the tubes prior to 4 incubation at 100 ºC for 10 min with subsequent cooling. The SPV reagent (1.2 mg/ml vanillin in 5 80% phosphoric acid) was added to each tube and allowed to develop for 10 min. Absorption 6 was measured on an Agilent 8453 Spectrophotometer at 490 nm and extrapolated from 7 species-specific calibration curves determined prior to analysis. For juveniles, lipid extraction 8 was performed on dried material using a previously described method (Vollenweider et al., 9 2011) derived from the Folch extraction procedure.

10 6.3.9 Energy density 11 Energy density (kJ/g dry mass) was estimated using bomb calorimetry. Homogenized dry tissue 12 was pressed into a pellet form and a Parr Instrument 6725 Semimicro Calorimeter with 6772 13 Precision Thermometer and 1109A Oxygen Bomb was used to measure the energy released 14 from combustion of the sample pellets. The minimum pellet weight was set at 0.025 g of dry 15 material based on the limits of instrument detection; individual larvae were composited within 16 stations as needed to attain sufficient dry mass.

17 6.3.10 Statistical analysis 18 A general linear model was used to identify differences in the nutritional state of arrowtooth 19 flounder and Kamchatka flounder. Lengths were compared by a nested analysis of variance 20 (ANOVA) to account for differences in the numbers of fish collected. Species was the main 21 factor with life stage nested in species and year nested in life stage. Post-hoc comparisons of 22 species were conducted among the life stages using Bonferonni’s adjusted t value. Analysis of 23 covariance (ANCOVA) was employed to compare dry mass and lipid content between species. 24 This approach permitted comparisons when length distributions differed between species and 25 the responses covaried with length. The ANCOVA used species, year, and their interaction as 26 main factors and length as covariate. Analyses were conducted separately for larvae and 27 juveniles because the response variables were heteroscedastic when pooled across life stages. 28 Variables were transformed prior to analysis using logarithms (base 10), and the assumption 29 that both species had equivalent slopes was tested. Student’s t was used to compare energy 30 densities of larvae and juveniles.

31 6.4 Results

32 6.4.1 Genetics 33 Using mtDNA COI restriction analysis proved successful in identifying Atheresthes larvae and 34 early juveniles to species in the laboratory (n = 337) and at sea (n = 78). Of the 415 total 35 specimens, 165 were identified as arrowtooth flounder, 194 as Kamchatka flounder, and 56

239

240

1 failed to amplify properly resulting in an unknown identity. Size of specimens examined was 2 6.0–38.0 mm SL, with the majority of specimens 6.0–11.5 mm SL.

3 6.4.2 Visual identification 4 Visual examination of genetically-identified specimens was successful in determining key 5 pigmentation differences between arrowtooth flounder and Kamchatka flounder larvae of 6.0– 6 12.0 mm SL and ≥ 18.0 mm SL. These key pigmentation differences are size specific but not 7 stage specific. Specimens 12.1–17.9 mm SL remain indistinguishable at this time because of the 8 small sample size and a high degree of variation in pigmentation patterns.

9 Recently hatched Kamchatka flounder larvae (6.0–7.4 mm SL) can be distinguished from 10 arrowtooth flounder of the same size by the presence of pigment dorsal to the gut at the anus 11 and an anterior dorsal pigment patch located halfway between the anus and the caudal fin (Fig. 12 1a). Arrowtooth flounder begin to develop the same pigment patterns at about 7.5 mm SL, 13 resulting in these characteristics no longer being able to be used. Larvae of 7.5–12.0 mm SL can 14 be identified by two characters. Kamchatka flounder larvae of 7.5–10.5 mm have pigment on 15 the crown of the head, beginning as a few melanophores and developing to numerous 16 melanophores that cover the entire crown of the head by approximately 9.0 mm SL (Fig. 1a). In 17 general, arrowtooth flounder larvae do not begin to develop head pigment prior to 10.5 mm SL, 18 however, a few individuals do develop one or two two small melanophores on the crown of the 19 head beginning at approximately 8.5 mm SL (Fig 1b). Thus, specimens 8.5–10.5 mm SL with only 20 1–3 melanophores on the head cannot be identified to species. In addition to crown pigment, 21 the second character by which Kamchatka flounder larvae 7.5–12.0 mm SL can be identified is 22 the pigment dorsal to the gut that starts at mid gut and extends to the anus (Fig. 1a). 23 Arrowtooth flounder larvae also have pigment dorsal to the gut, but it starts at ¾ gut length 24 and extends to the anus. The combination of these two characters allows larvae 7.5–12.0 mm 25 SL to be identified to species.

26 Few genetically identified specimens were larger than 12.0 mm SL. Identification methods for 27 larger specimens began with the use of adult characters to identify specimens > 25.0 mm SL. 28 These identified specimens were then used as a guide to begin identifying smaller specimens. 29 The adult characters used to determine the larger specimens was the number of gill rakers on 30 the second upper gill arch: Kamchatka flounder have one gill raker, arrowtooth flounder have 31 two or three. While these structures are not fully formed in specimens of 25.0–27.0 mm SL, 32 fleshy protrusions that are the precursors to the gill rakers are evident. After numerous 33 specimens were separated based on gill raker counts, a pattern in the dorsal pigment bands 34 was seen to be different between the two species.

35 The anterior dorsal patch on Kamchatka flounder is generally longer (> 5 melanophores long) 36 and the individual melanophores in the patch are close together, often touching and at times

240

241

1 coalescing. Additionally, Kamchatka flounder develop melanophores between the anterior and 2 posterior bands, which eventually connect the two bands at a smaller size than arrowtooth 3 flounder (beginning as early as approximately 13.0 mm SL, but generally by 18.0 mm SL) (Fig. 4 1a). In contrast, arrowtooth flounder tend to have a short (< 5 melanophores long) anterior 5 dorsal band with melanophores that are widely spaced apart. Melanophores do not develop 6 between the anterior and posterior dorsal bands until approximately 25.0 mm SL (Fig. 1b). The 7 few large larvae and early juveniles used in the genetic technique confirmed the validity of 8 these visual differences.

9 Morphological measurements were taken on 20 arrowtooth flounder larvae and 19 Kamchatka 10 flounder larvae. Specimens ranged in length from 6.3 to 12.6 mm SL. Head length and eye 11 diameter were not statistically different between the two species, but body depth and preanal 12 length were. Larvae of Kamchatka flounder have a stouter body (BD = 12.0 ± 1.1% SL) than 13 arrowtooth flounder larvae (BD = 9.8 ± 1.4% SL; p < 0.05). In addition to a stouter body, 14 Kamchatka flounder larvae also have a slightly longer preanal length than arrowtooth flounder 15 larvae (36.3 ± 1.8% SL vs. 35.1 ± 1.8% SL; p < 0.05).

16 6.4.3 Distribution 17 Examining formalin-preserved specimens from previous AFSC cruises, a total of 1,928 larvae 18 and juveniles were identified to species using pigmentation and morphological characters 19 identified above, with individuals identified to genus because of ambiguous characters or 20 damage. A total of 1,315 arrowtooth flounder larvae and 480 Kamchatka flounder larvae were 21 identified; 125 arrowtooth flounder juveniles and eight Kamchatka flounder juveniles were 22 identified.

23 The majority of larval arrowtooth flounder and Kamchatka flounder were collected along the 24 shelf break in the southeastern Bering Sea between Unimak and Umnak Pass (Fig. 2). There 25 were also individuals collected farther north in Pribilof Canyon along the shelf break near deep 26 water. The calculated mean distribution for both species is similar (Fig. 2). Juveniles of both 27 species were collected primarily north of where the majority of the larvae were collected. The 28 calculated mean distribution of the two species showed that arrowtooth flounder juveniles 29 appear to be located farther east and over shallower waters than Kamchatka flounder juveniles, 30 which have a mean distribution close to Pribilof Canyon and deeper water (Fig. 2). However, 31 fewer Kamchatka flounder juveniles were collected and identified (n = 8) than arrowtooth 32 flounder juveniles (n = 125), so caution should be used when interpreting this result.

33 6.4.4 Biological sampling 34 Length analysis was restricted to larvae collected in 2009 and 2010 and juveniles collected in 35 2010. These were stages and years in which sufficient numbers of specimens were collected. 36 No significant difference was found in the lengths of larvae or juveniles sampled in 2009 and

241

242

1 2010 (F2.154 > 0.77, p = 0.466), and larvae were significantly smaller than juveniles (F2.154 > 1000, 2 p < 0.001). Pairwise comparisons of larvae and juveniles between the two species averaged 3 across years indicated that Kamchatka flounder larvae averaged approximately 25% longer than 4 arrowtooth flounder larvae (t = 4.6, p < 0.001), but no difference was detected among juveniles 5 (t = 0.31, p = 0.991) (Table 2). Differences between the two sets of larvae therefore accounted 6 for the overall difference detected between species (F1.154 = 6.71, p = 0.010).

7 6.4.5 Dry mass 8 The dry weights of larvae did not depend on year nor was there an interaction between year 9 and species (F1.96 < 0.15, p > 0.700). However, larval arrowtooth flounder of a fixed length were 10 approximately 25% lighter than similarly sized Kamchtaka flounder (F1.96 = 4.16, p = 0.044; Fig. 11 3a). No difference in dry mass was detected among juvenile specimens in 2010 (F1.53 = 1.68, p = 12 0.20; Fig 3b). It was not possible to test for year effect due to an inadequate number of samples 13 in 2008 and 2009.

14 6.4.6 Lipid content 15 In 2009 and 2010 the average lipid content of arrowtooth larvae was 7.5 ± 0.3% (n = 39) of dry 16 mass compared with 10.8 ± 0.7% for Kamchatka flounder larvae (n = 11; Fig. 4a). While the lipid 17 content of Kamchatka flounder larvae averaged approximately 30% more than that of 18 arrowtooth flounder larvae, this difference was not significant after controlling for the 19 differences in average length (F1.45 = 3.15, p = 0.083). Additionally, there was not an effect of 20 year or interaction between year and species on larval lipid content (F1.45 < 2.37, p > 0.130). In 21 2010, there was no difference between juveniles (F1.25 = 0.03, p = 0.857). Arrowtooth flounder 22 juveniles averaged 10.2 ± 0.5% lipid (n = 21) in 2010 compared with 10.7 ± 0.7% (n = 7) for 23 Kamchatka flounder juveniles.

24 6.4.7 Energy density 25 Energy density in juveniles was only measured in 2010 and in larvae only in 2009. Only a few 26 larvae could be measured because samples required compositing. There was no statistically 27 significant difference between the energy densities of the two species. Only six measurements 28 of larval energy density were taken and the difference between species was not statistically 29 significant (t = 1.06, p = 0.366) (Fig. 4b). Arrowtooth larvae averaged 19.2 ± 0.8 kJ/g dry mass; 30 Kamchatka flounder larvae averaged 19.8 ± 0.7 kJ/g dry mass. Energy densities between 31 juveniles of the two species were nearly identical (t = 1.19, p = 0.244), averaging 20.5 ± 0.2 kJ/g 32 dry mass for arrowtooth flounder (n = 16) and 20.8 ± 0.2 kJ/g dry mass for Kamchatka flounder 33 (n = 12).

34 6.5 Discussion 35 With the increasing population size of arrowtooth flounder and interest for increased directed 36 fishing effort for both arrowtooth flounder and Kamchtaka flounder, there is a need to fully

242

243

1 identify and study both species in all life stages for management and ecological purposes. 2 Adults of arrowtooth flounder have been studied with regard to their distribution (Zimmerman 3 and Goddard, 1996), predation effects on walleye pollock (Ianelli et al., 2009), and their role as 4 prey for other organisms (Livingston and Jurado Molina, 2000). The early life stages in the Gulf 5 of Alaska have been described by Blood et al. (2007); however, efforts to conclusively identify 6 larvae and juveniles in the Bering Sea have been confounded by the overwhelming physical 7 similarities between arrowtooth and Kamchatka flounder.

8 6.5.1 Genetics 9 Genetic methods have proven to be useful tools in identifying unknown specimens to the 10 species or family level using the mtDNA gene COI (Hebert et al., 2002; Dawnay et al., 2007; 11 Rocha et al., 2007). In Alaskan waters, Spies et al. (2006) also successfully used mtDNA COI to 12 identify 15 different skate species. In this current study we have demonstrated that this genetic 13 technique is valid for identifying specimens of Atheresthes to the species level. In addition, the 14 technique developed in this study proved effective for use both in the laboratory and at sea 15 aboard a research vessel. In particular, this ability to conclusively identify specimens to the 16 species level while at sea is quite useful as it can enable a timely adjustment to sampling 17 location or procedure in order to maximize collection of the target species.

18 6.5.2 Morphology 19 In the marine environment closely related species are often indistinguishable as larvae due to 20 overlap in pigmentation and morphological characters. Prior to this study, all larval and early 21 juvenile Atheresthes collected in the Bering Sea were identified only to the genus level. Despite 22 numerous physical similarities, unique pigmentation and morphological differences were noted 23 between the two Atheresthes species in the size ranges of 6.0–12.0 mm SL and ≥ 18.0 mm SL. In 24 general, it appears that Kamchatka flounder acquire certain pigmentation characters at smaller 25 sizes than arrowtooth flounder, thus the distinguishing characters noted in this paper can only 26 be used for the specific size ranges to which they correspond. Specimens of 12.1–17.9 mm SL 27 are indistinguishable at this time due because of a high degree of variability in pigment 28 amongst individuals and a low number of genetically identified specimens in this size range. In 29 addition to pigment, Kamchatka flounder larvae are stouter than arrowtooth flounder larvae 30 (BD 12.0% SL vs. 9.8% SL) and have a slightly longer snout to anus length (36.3% SL vs. 35.1% 31 SL). These are modest morphological differences and can be difficult to determine but can be 32 helpful when used in conjunction with pigmentation characters to support species 33 identification. Using the pigment characteristics described in this study, two specimens 34 illustrated by Blood et al. (2007; Fig. 13) should be re-examined as both specimens were 35 collected in the Bering Sea. The 13.4 mm SL specimen is most likely Kamchatka flounder, not 36 arrowtooth flounder, based on the large amount of pigment present on the head and the 37 pigment present on the dorsal midline between the two dorsal bands. While arrowtooth

243

244

1 flounder do develop pigment on the crown of the head, the amount of pigment present at this 2 size is more indicative of Kamchatka flounder. In addition, the presence of pigment on the 3 dorsal midline between the two dorsal bands at this size can only be Kamchatka flounder as 4 arrowtooth do not develop this pigment until approximately 25.0 mm SL. The 10.0 mm SL 5 specimen (Blood et al., 2007; Fig. 12) may also be a Kamchatka flounder due to the large 6 amount of pigment located on the crown of the head and within the two dorsal patches. 7 However, complete identification is difficult from the illustrations alone; a close visual 8 examination of the original specimens is needed.

9 6.5.3 Distribution 10 The larval distributions of arrowtooth flounder and Kamchatka flounder in the eastern Bering 11 Sea are similar. Most of the larvae collected were located in the southern part of the eastern 12 Bering Sea in water > 200 m along the shelf break (Fig 2a, b). This may indicate that in the 13 eastern Bering Sea both arrowtooth and Kamchatka flounder adults spawn in deep water, 14 similar to that described for arrowtooth flounder in the Gulf of Alaska by Blood et al. (2007). 15 The distribution of juveniles collected does appear to be different between the two species, but 16 this result is tenuous due to the low number of juvenile Kamchatka flounder collected (n = 8). 17 However, this result may indicate that juvenile Kamchatka flounder are migrating to their adult 18 habitat. In general, the majority of adult Kamchatka flounder are collected in deeper waters 19 and farther to the west than adult arrowtooth flounder, which often occur in shallower waters 20 to the east (Zimmerman and Goddard 1996; Wilderbuer et al., 2011). It is possible that the low 21 number of juvenile Kamchatka flounder collected is due to limited sampling in deeper waters to 22 the west of the shelf break.

23 6.5.4 Condition 24 Differences in the condition between arrowtooth flounder and Kamchatka flounder larvae 25 collected in the summer may relate to differences in timing of development. For years in which 26 sufficient numbers of larvae were collected for analysis, Kamchatka flounder larvae were 27 significantly larger than arrowtooth flounder larvae in the summer. Comparison of the lengths 28 of larvae sampled in the summer indicates that the bulk of arrowtooth flounder larvae were in 29 the preflexion or flexion stages (Blood et al., 2007). Assuming stage designation is similar for 30 Kamchatka flounder, the majority of Kamchatka flounder sampled in the summer were in the 31 flexion and postflexion stages. The difference in developmental stage could account for the 32 differences in the length adjusted dry mass of the two species and the trend towards increased 33 lipid content in larval Kamchatka flounder. Energy allocation strategies can differ among 34 developmental stages during the early life stages. This study suggests that postflexion larvae of 35 Atheresthes may invest more energy accreting tissue mass than earlier stages. Comparison of 36 the lengths of juveniles sampled indicates that the majority of the fish sampled from both 37 species were in the transformation stage (Blood et al., 2007). Hence there were no differences

244

245

1 in condition between juveniles of either species. The results presented here are the first 2 estimates of lipid and energy density in larval and juvenile arrowtooth flounder and Kamchatka 3 flounder. Lipid content values of juveniles in this study are similar to those reported for Atlantic 4 halibut (Hippoglossus hippoglossus) fed a diet that promoted growth and survival (Hamre et al., 5 2002). Energy densities of larvae were similar to those of juveniles and suggests no change in 6 energy allocation strategy between these two stages. This is in contrast to walleye pollock 7 which undergo significant changes in both lipid content and energy density between lengths of 8 20 and 60 mm (Siddon et al., 2012), which is when walleye pollock undergo and complete 9 transformation to the juvenile stage (Brown et al., 2001). The only evidence of a shift in lipid 10 content among both species of Atheresthes sampled here is associated with ontogeny in the 11 larval stages.

12 6.6 Conclusion 13 This study has provided a method to genetically identify larval and early juvenile arrowtooth 14 flounder and Kamchatka flounder both in the field and in the laboratory, and has provided 15 morphological and pigmentation characters to visually identify these two species at small (6.0– 16 12.0 mm SL) and large sizes (≥ 18.0 mm SL). Future sampling effort during June–August when 17 arrowtooth flounder and Kamchatka flounder larvae 12.1–17.9 mm SL are in the water column 18 could aid in completing the visual identification of these two species. Once all arrowtooth 19 flounder and Kamchatka flounder between 6.0 mm SL to the juvenile stage can be identified, all 20 historical samples of Atheresthes collected in 1991–2010 (19 years) by the AFSC from the 21 eastern Bering Sea can be re-identified to the species level. This collection contains 22 approximately 5,065 individuals that, when identified to the species level, will greatly increase 23 our knowledge of distribution and abundance for these two species. Additionally, increased 24 sampling effort in the deeper water off the shelf break during summer and fall would help 25 define the distribution of the larger sizes of these two species and determine if there is a 26 difference in timing of development for larvae in the summer.

27 6.7 Acknowledgements 28 The authors would like to thank all the scientists that collected Atheresthes spp., larvae over the 29 years and the officers and crews of the following ships: NOAA ship Miller Freeman, NOAA ship 30 Oscar Dyson, R/V Thomas G. Thompson, USCGC Healy, and the R/V Knorr. We would also like to 31 thank Dan Cooper for his help with genetic identification of specimens while on the May 2010 32 cruise and for reviewing a draft of this manuscript, Melanie Paquin for her help with genetic 33 identification and sequencing in the laboratory, Ashlee Overdick for her illustrations of both 34 species, and Debbie Blood for also reviewing a draft of this manuscript. This is BEST-BSIERP 35 Bering Sea Project 476 publication number XX. This research is contribution EcoFOCI-XXX to 36 NOAA’s Fisheries-Oceanography Coordinated Investigations. The findings and conclusions in

245

246

1 this manuscript are those of the authors, and do not necessarily represent the views of the 2 National Marine Fisheries Service.

3 6.8 Literature Cited

4 Aydin, K., Gaichas, S., Ortiz, I., Kinzey, D., and Friday, N. 2007. A comparison of the Bering Sea, Gulf

5 of Alaska, and Aleutian Islands large marine ecosystems through food web modeling. U.S. Dept.

6 Commer., NOAA Tech. Memo. NMFS484 AFSC-178, 298 p.

7 Blood, D.M, Matarese, A.C., and Busby, M.S. 2007. Spawning, egg development, and early life

8 history dynamics of arrowtooth flounder (Atheresthes stomias) in the Gulf of Alaska. NOAA

9 Professional Paper NMFS 7, 28 p.

10 Brown, A.L., Busby, M.S., and Mier, K.L. 2001. Walleye pollock Theragra chalcogramma during

11 transformation from the larval to the juvenile stage: otolith and osteological development. Mar.

12 Biol. 139: 845-851 491 Dawnay, N., Ogden, R., McEwing, R., Carvalho, G.R., and Thrope, R.S. 2007.

13 Validation of the barcoding gene COI for use in forensic genetic species identification. Forensic Sci.

14 Int. 173:1-6.

246

247

1 Hebert, P.D.N., Cywinska, A., Ball, S.L., and deWaard, J.R. 2002. Biological identifications

2 through DNA barcodes. Proc. R. Soc. Lond. B. 270: 313-321.

3 Hamre, K., Opstad, I., Espe, M., Solbakken, J., Hemre, G.I., and Pittman, K. 2002. Nutrient

4 composition and metamorphosis success of Atlantic halibut (Hippoglossus hippoglossus, L.)

5 larvae fed natural zooplankton or Artemia. Aquac. Nutr. 8: 139-148

6 Ianelli, J.N., Barbeaux, S., Honkalehto, T., Kotwicki, S., Aydin, K., and Williamson, N. 2009.

7 Assessment of the Walleye Pollock stock in the Eastern Bering Sea. In: Stock assessment and fishery

8 evaluation report for the groundfish resources of the Bering Sea/Aleutian Islands regions. North

9 Pacific Fishery Management Council, 605 W. 4th Ave., Suite 306, Anchorage, AK 99501.

10 Kendall, A.W., Jr., Ahlstrom, E.H., and Moser, H.G. 1984. Early life history stages and their

11 characters. In: Ontogeny and systematic of fishes. (H.G. Moser, W.J. Richards, D.M. Cohen, M.P.

12 Fahay, A.W. Kendall Jr., and S.L. Richardson, eds). P. 11-22. Spec. Publ. 1 Am. Soc. Ichthyol.

13 Herpetol. Allen Press, Lawrence, KS.

14 Livingston, P.A., and Jurado-Molina, J. 2000. A multispecies virtual population analysis of the

15 eastern Bering Sea. ICES J. Mar. Sci. 57: 294-299.

247

248

1 Matarese, A.C., Blood, D.M., Picquelle, S.J., and Benson, J.L. 2003. Atlas of abundance and

2 distribution patterns of ichthyoplankton from the northeast Pacific Ocean and Bering Sea

3 ecosystems based on research conducted by the Alaska Fisheries Science Center (1972-1996).

4 NOAA Prof. Paper NMFS 1, 281 p.

5 Matarese, A.C., Kendall, A.W. Jr., Blood, D.M., and Vinter, B.M. 1989. Laboratory guide to early

6 life history stages of northeast Pacific fishes. U.S Dep. Commer. NOAA Tech. Rep. NMFS 80, 652

7 p.

8 Mecklenburg, C.W., Mecklenburg, T.A., and Thorsteinson, L.K. 2002. Fishes of Alaska. American

9 Fisheries Society, Bethesda, MD, 1116 p.

10 Methot, R.D. 1986. Frame trawl for sampling juvenile fish. Calif. Coop. Oceanic Fish. Invest. Rep.

11 27: 267-278

12 Ranck, C.L., Utter, F.M., Milner, G. B., and Smith, G. B. 1986. Genetic confirmation of specific

13 distinction of arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder, A. evermanni.

14 Fish. Bull. 84: 222-226.

15 Rocha, L.A., Craig, M.T., and Bowen, B.W. 2007. Phylogeny and the conservation of coral reef

16 fishes. Coral Reefs. 26: 501-512.

17 Siddon, E.C., Heintz, R.A., and Mueter, F.J. 2013. Conceptual model of energy allocation in walleye

18 pollock (Theragra chalcogramma) from age-0 to age-1 in the southeastern Bering Sea. Deep-Sea

19 Res. II. 94: 140-149.

248

249

1 Spies, I.B., Gaichas, S., Stevenson, D.E., Orr, J.W., and Canino, M.F. 2006. DNA-based identification

2 of Alaska skates (Amblyraja, Bathyraja and Raja: Rajidae) using cytochrome c oxidase subunit I

3 (coI) variation. J. Fish. Biol. 69(B): 283-292.

4 Spies, I.B., Wilderbuer, T.K., Nichol, D.G., and Aydin, K. 2011. Assessment of the arrowtooth

5 flounder stock in the Bering Sea/Aleutian Islands. North Pacific Management Council, p. 727-

6 744.

7 Tucker. G.H. 1951. Relation of fishes and other organisms to the scattering of underwater sound. J.

8 Mar. Res. 10(2): 215–238.

9 Van Handel, E. 1985. Rapid determination of total lipids in mosquitoes. J. Am. Mosq. Control

10 Assoc. 1 (3): 302-304.

11 Vollenweider, J.J., Heintz, R., Schaufler, L., and Bradshaw, R. 2011. Seasonal cycles in whole-body

12 proximate composition and energy content of forage fish vary with water depth. Mar. Biol. 158,

13 413-27.

14 Wiebe, P.H., Burt, K.H., Boyd, S.H., and Morton, A.W. 1976. A multiple opening/closing net and

15 environmental sensing system for sampling zooplankton. J. Mar. Res. 34: 313–326.

16 Wilderbuer, T.K., Nichol, D.G., and Lauth, R. 2011. Assessment of the Kamchatka flounder stock in

17 the Bering Sea and Aleutian Islands. North Pacific Management Council, p. 795-801.

249

250

1 Yang, M.S. 1988. Morphological differences between two congeneric species of pleuronectid

2 flatfishes: arrowtooth flounder, Atheresthes stomias, and Kamchatka flounder, A. evermanni. Fish.

3 Bull. 86: 608-611.

4 Yang, M.S., and Livingston, P.A. 1986. Food habits and diet overlap of two congeneric species,

5 Atheresthes stomias and Atheresthes evermanni, in the eastern Bering Sea. Fish. Bull. 82: 615-623.

6 Zimmerman, M. and Goddard, P. 1996. Biology and distribution of arrowtooth, Athersthes stomias,

7 and Kamchatka, A, evermanni, flounders in Alaskan waters. Fish. Bull. 94: 358-370.

9 6.9 Figure and table legends 10 Figure 1. Illustrations of eastern Bering Sea a) Kamchatka flounder (Atheresthes evermanni) and 11 b) arrowtooth flounder (A. stomias). Arrows point to key identifying characters between the 12 two species. Illustrations by Ashlee Overdick

2 13 .Figure 2. Distribution of abundance (catch/10m ) for formalin preserved larval and juvenile 14 Atheresthes identified to species using visual identification techniques. a) arrowtooth flounder 15 (ATF: Atheresthes stomias). b) Kamchatka flounder (KF: A. evermanni). In both maps, circles 16 depict larvae (<25 mm SL), triangles depict juveniles (≥25 mm SL), X depicts calculation of mean 17 distribution generated by ArcGIS software of larvae based on abundance, and diamond depicts 18 calculated mean distribution of juveniles based on abundance. The size of circles and triangles 19 is proportional to abundance.

20 Figure 3. Length-dry mass relationships for arrowtooth flounder (Atheresthes stomias) and 21 Kamchatka flounder (Atheresthes evermanni) of larvae (a) and juveniles (b). Lines depict 573 22 the least-squares relationships for all species combined.

23 Figure 4. Measures of condition for arrowtooth flounder (Atheresthes stomias) and Kamchatka 24 flounder (A. evermanni) collected in June/July (larvae) and September (juveniles) 2008–2010. a) 25 % lipid (dry mass) and b) energy density (kJ/g). Sample sizes are indicated above each bar.

250

251

1 Figure 1. a)

8 b)

11 12

13 Figure 2.

251

252

1 a)

12 b)

252

253

1 Figure 3.

253

254

1 Figure 4.

254

255

1 255

256

1 2

256

257

2 Conclusions: 3 4 The primary conclusion from this project is that the energy content of age-0 pollock during their first 5 fall predicts their survival to age-1 (Heintz et al. 2013). Energy content is measured as the total 6 energy content of the fish. It is the product of the energy density of the fish and it’s mass. Fish that 7 are likely to survive winter must achieve a sufficient size and store a sufficient amount of lipid in 8 order to maximize their probability of surviving winter. Their ability to achieve large size depends on 9 water temperature. Development rate is accelerated in warmer water and fish are able to 10 metamorphose sooner than in cool water (Smart et al. 2012). This early transition to the juvenile 11 stage increases the time available for provisioning lipid stores prior to winter (Siddon et al. 2013). 12 While warm conditions might be expected to maximize growth (Hurst et al. 2011) and the 13 provisioning period, prey are less available under warm conditions (Coyle et al. 2011) and their 14 overall lipid content is low (Heintz et al. 2013). In addition, there may be a poor spatial match 15 between pollock and their prey (Siddon et al. 2013) in warm conditions. These observations are in 16 direct contrast to the Oscillating Control Hypothesis (Hunt et al. 2008) and suggest that cool 17 conditions are more likely to support high production of pollock than warm conditions. A similar 18 story may exist for Pacific cod (Farley et al. in press), but there were too few samples to provide 19 adequate detail. 20 21 Analysis of the energy content of arrowtooth flounder and Kamchatka flounder indicated flatfishes 22 have different life history constraints from gadids. Both flatfish species had similar lipid and energy 23 contents through all their developmental stages in contrast to walleye pollock and Pacific cod. 24 However, it is unknown if their energy allocation strategies changes once they settled out of the 25 water column. In addition it appears that arrowtooth and Kamchatka flounder partition their 26 habitats temporally, with the latter species settling out of the water column earlier than the former. 27 28

29 BSIERP and Bering Sea Project connections: 30 31 This project would not have been possible without contributions from the ichthyoplankton and fish 32 projects. Both of those projects conducted surveys and provided this study with laboratory samples. 33 In addition, analyses conducted by those projects were important to understanding the data 34 collected here. The ichthyoplankton study provided data on the distribution and diet of larval fishes 35 and one of the graduate students ended up processing samples for this study and leading 36 manuscripts. The fish study performed a similar role, providing diet and distribution data for juvenile 37 forms. A difficulty associated with the structure of the overall program is that seasonal bioenergetics 38 project was not wholly subsumed into one of the larger study groups (fish or ichthyoplankton). This 39 hampered communication between this project and the others, particularly when they conducted 40 independent meetings. Nevertheless, all the groups were receptive to the efforts of the 41 bionenergetics project and willing to supply any and all requested data. 42 43 The laboratory studies were more ambitious than anticipated and we only were able to complete 44 the RNA/DNA calibration and Pacific cod growth studies with the funds available. There were 257

258

1 insufficient numbers of fish available to conduct energetic studies during the developmental 2 process. However, these data have proven to be useful and provide a basis for understanding 3 observations of RNA/DNA made on field samples. Following this study we acquired three years of 4 RNA/DNA data from the field, this was insufficient for establishing environmental impacts. However, 5 today we now have five years of data and will collect a sixth set in the summer of 2014. This latter 6 set should provide insight into the growth under warm conditions. These latter samples will be the 7 subject of an upcoming manuscript. 8

9 Management or policy implications: 10 11 This study establishes a link between climate and the recruitment of age-0 pollock. It provides a 12 basis for understanding the recruitment process for walleye pollock in the eastern Bering Sea. 13 Beyond simply establishing a correlation, it develops from a theoretical underpinning and 14 demonstrates a way to link environmental variation to biological endpoints. The understanding 15 accrued in this project is important for explaining variability in the biomass of fishable populations. 16 By establishing this link it is now possible to predict how pollock biomass is likely to change in the 17 future in response to a warming climate. Consequently, the data are now considered in the pollock 18 stock assessment (e.g. Ianelli 2012) and NOAA’s annual Alaska Marine Ecosystems Considerations 19 Report for the North Pacific Fisheries Management Council (Zador 2013). 20 21

22 Publications: 23 24 Data or conclusions drawn from this project appear in the following publications. 25 26 De Forest, L., Duffy-Anderson, J.T., Heintz, R. A., Matarese, A. C., Siddon, E.C., Smart, T. I., 27 Spies, I. B. (Submitted). Ecology and taxonomy of the early life stages of arrowtooth 28 flounder (Atheresthes stomias) and Kamchatka flounder (A. evermanni) in the 29 eastern Bering Sea. Deep Sea Research II. 30 31 Duffy-Anderson, J.T., Barbeaux, S., Farley, E., Heintz, R., Horne, J., Parker-Stetter, S. Petrik, 32 C., Siddon, E.C., Smart, T.I. (Submitted). A review and ecological synthesis of the first 33 year of life of walleye pollock (Gadus chalcogrammus) in the eastern Bering Sea and 34 comments on implications for recruitment. Deep Sea Research II. 35 36 Farley, Jr., E. V., Heintz, R. A., Andrews, A. G., Hurst, T. P. (Submitted). Size, diet, and 37 condition of age-0 Pacific cod (Gadus macrocephalus) during warm and cool climate 38 states in the eastern Bering Sea. Deep Sea Research II.

39 Heintz, R.A., Siddon, E.C., Farley, E.V., and Napp, J.M. 2013. Correlation between 40 recruitment and fall condition of age-0 pollock (Theragra chalcogramma) from the 41 eastern Bering Sea under varying climate conditions. Deep-Sea Res. II 94, 150-156. 42

258

259

1 Hunt, G. L., Jr., Stabeno, P. J., Strom, S., and Napp, J. M. 2008. Patterns of spatial and 2 temporal variation in the marine ecosystem of the southeastern Bering Sea, with 3 special reference to the Pribilof Domain. Deep-Sea Research II, 55: 1919-1944. 4 5 Hunt GL Jr, Coyle KO, Eisner L, Farley EV, Heintz R, Mueter FJ, Napp JM, Overland JE, Ressler 6 PH, Salo S, Stabeno PJ (2011) Climate impacts on eastern Bering Sea foodwebs: A 7 synthesis of new data and an assessment of the Oscillating Control Hypothesis. ICES 8 Journal of Marine Sciences 68(6): 1230-1243. 9

10 Siddon EC, Heintz RA, Mueter FJ (2013) Conceptual model of energy allocation in walleye 11 pollock (Theragra chalcogramma) from age-0 to age-1 in the southeastern Bering 12 Sea. Deep-Sea Research II. http://dx.doi.org/10.1016/j.dsr2.2012.12.007.

13 Siddon, EC, Kristiansen T, Mueter FJ, Holsman KK, Heintz RA, Farley EV. 2014. Spatial match- 14 mismatch between juvenile fish and prey provides a mechanism for recruitment 15 variability across contrasting climate conditions in the eastern Bering Sea. PLoS One 16 8(12) doi:10.1371/journal.pone.0084526 17 18 Sigler, M.F., Heintz, R.A., Hunt Jr., G. L., Lomas, M. W., Napp, J. M., Stabeno, P. J. 19 (Submitted). A mid-trophic view of subarctic productivity: Lipid storage, location 20 matters and historical context. Deep Sea Research.

21 22 Sreenivasan, A. (2011) Nucleic acid ratios as an index of growth and nutritional ecology in 23 Pacific cod (Gadus macrocephalus), walleye pollock (Theragra chalcogramma), and 24 Pacific herring (Clupea Pallasii). PhD Dissertation. University of Alaska, Fairbanks, 25 Alaska. 112 p. 26 27

28 Poster and oral presentations at scientific conferences or seminars 29 Oral presentations: 30 Growth and Energy Allocation in Larval Pollock from the Bering Sea. 2009 Larval Fish 31 Conference. Portland, OR. 32 33 Bioenergetic Constraints on Winter Survival in Subadult Walleye Pollock (Theragra 34 chalcogramma). 2010 National Ocean Sciences Meeting. Portland, OR 35 36 37 Climate related changes in the nutritional condition of young-of-the-year pollock 38 (Theragra chalcogramma) from the eastern Bering Sea. 2011. Ecosystem Studies of 39 Subarctic Seas Open Science Meeting. Seattle, WA. 40 41 Climate Effects on the Nutrtional Condition and Productivity of Marine Fish Populations.

259

260

1 2012. Comparative Nutritition Society. Monterey, CA. 2 3 Climate Effects on the Nutrtional Condition and Productivity of Marine Fish Populations. 4 2012. Ecosystem Studies of Subarctic Seas Annual Meeting. Hakodate, JP. 5 6 Posters 7 Differences between observed growth and a physiological growth index (RNA/DNA 8 ratio) in larval Pacific cod (Gadus macrocephalus) at different temperatures. 2009. 9 Alaska Marine Sciences Symposium, Anchorage, AK 10 11 Seasonal bioenergetics of walleye pollock (Theragra chalcogramma) and Pacific cod 12 (Gadus marcroecphalus) in the southeastern Bering Sea. 2009. 2009. Alaska Marine 13 Sciences Symposium, Anchorage, AK 14 15 Seasonal patterns of energy content and allocation in walleye pollock (Theragra 16 chalcogramma). 2011. Alaska Marine Sciences Symposium, Anchorage, AK 17 18 Age-1 walleye pollock in the eastern Bering Sea: distribution, abundance and energy 19 density. 2011. Alaska Marine Sciences Symposium, Anchorage, AK 20 21 Variability in energetic status, diet and size of age 0 Pacific cod during warm and cold 22 climate states in the eastern Bering Sea. 2013. Alaska Marine Sciences Symposium, 23 Anchorage, AK 24 25 An individual-based model for growth of walleye pollock (Theragra chalcogramma) in 26 the eastern Bering Sea: implications for bottom-up control of recruitment success. 2013. 27 Alaska Marine Sciences Symposium, Anchorage, AK 28

29 Outreach 30 Samples and concepts derived from this project formed the basis of several high school science 31 fair projects, a Dartmouth University Steffanson Internship Grant and work conducted by a 32 NOAA Hollings Scholar. One of the science fair projects advanced to the Intel International 33 Science and Engineering Fair. It was titled “Applications of Bioelectrical Impedance Analysis to 34 Predict Energy Content of Walleye Pollock”. The Hollings scholar examined seasonal variation in 35 the energy content of Thysanoessa raschii. The Dartmouth intern conducted a latitudinal 36 analysis of the energy content of capelin. 37

38 Acknowledgements 39 I wish to thank all of the technical staff at the Auke Bay Laboratories for receiving, maintaining 40 and processing the samples. The efforts of Robert Bradshaw, Andrew Eller, Lawrence Schaufler, 41 Ann Robertson, Kevin Heffern, and Elizabeth Parker were particularly helpful. Thanks are due to 42 the staff of the NOAA-AFSC laboratory at the Hatfield Marine Science Laboratory in Newport OR 43 for their help in developing the RNA/DNA studies. 44

260

261

1 Literature cited: 2 De Forest, L., Duffy-Anderson, J.T., Heintz, R. A., Matarese, A. C., Siddon, E.C., Smart, T. 3 I., Spies, I. B. (Submitted). Ecology and taxonomy of the early life stages of 4 arrowtooth flounder (Atheresthes stomias) and Kamchatka flounder (A. 5 evermanni) in the eastern Bering Sea. Deep Sea Research II. 6 7 Duffy-Anderson, J.T., Barbeaux, S., Farley, E., Heintz, R., Horne, J., Parker-Stetter, S. 8 Petrik, C., Siddon, E.C., Smart, T.I. (Submitted). A review and ecological synthesis 9 of the first year of life of walleye pollock (Gadus chalcogrammus) in the eastern 10 Bering Sea and comments on implications for recruitment. Deep Sea Research II. 11 12 Farley, Jr., E. V., Heintz, R. A., Andrews, A. G., Hurst, T. P. (Submitted). Size, diet, and 13 condition of age-0 Pacific cod (Gadus macrocephalus) during warm and cool 14 climate states in the eastern Bering Sea. Deep Sea Research II.

15 Heintz, R.A., Siddon, E.C., Farley, E.V., and Napp, J.M. 2013. Correlation between 16 recruitment and fall condition of age-0 pollock (Theragra chalcogramma) from 17 the eastern Bering Sea under varying climate conditions. Deep-Sea Res. II 94, 18 150-156. 19 20 Hunt GL Jr, Coyle KO, Eisner L, Farley EV, Heintz R, Mueter FJ, Napp JM, Overland JE, 21 Ressler PH, Salo S, Stabeno PJ (2011) Climate impacts on eastern Bering Sea 22 foodwebs: A synthesis of new data and an assessment of the Oscillating Control 23 Hypothesis. ICES Journal of Marine Sciences 68(6): 1230-1243. 24 25 Ianelli JN, Hokalehto T, Barbeaux S, Kotwicki S, Aydin K, Williamson N (2012). Stock 26 Assessment and Fishery Exploitation Report. Chapter 1. Assessment of the 27 walleye pollock stock in the Easter Bering Sea. North Pacific Fishery Management 28 Council. 605 W 4th Ave, Suite 306, Anchorage, AK 99501 29

30 Siddon EC, Heintz RA, Mueter FJ (2013) Conceptual model of energy allocation in 31 walleye pollock (Theragra chalcogramma) from age-0 to age-1 in the 32 southeastern Bering Sea. Deep-Sea Research II. 33 http://dx.doi.org/10.1016/j.dsr2.2012.12.007.

35 Siddon, EC, Kristiansen T, Mueter FJ, Holsman KK, Heintz RA, Farley EV. 2014. Spatial 36 match-mismatch between juvenile fish and prey provides a mechanism for 37 recruitment variability across contrasting climate conditions in the eastern 38 Bering Sea. PLoS One 8(12) doi:10.1371/journal.pone.0084526 39

261

262

1 Sigler, M.F., Heintz, R.A., Hunt Jr., G. L., Lomas, M. W., Napp, J. M., Stabeno, P. J. 2 (Submitted). A mid-trophic view of subarctic productivity: Lipid storage, location 3 matters and historical context. Deep Sea Research.

4 5 Smart, T. I., Duffy-Anderson, J. T., Horne, J. K. 2012. Alternating temperature states 6 influence walleye pollock early life stages in the southeastern Bering Sea. Marine 7 Ecology Progress Series 455:257-267. DOI:10.3354/meps09619 8 9 Sreenivasan, A. (2011) Nucleic acid ratios as an index of growth and nutritional ecology 10 in Pacific cod (Gadus macrocephalus), walleye pollock (Theragra chalcogramma), 11 and Pacific herring (Clupea Pallasii ). PhD Dissertation. University of Alaska, 12 Fairbanks, Alaska. 112 p. 13 14 Zador, S. (ed.), (2013). Ecosystem Considerations 2013, Stock Assessment and Fishery 15 Evaluation Report, North Pacific Fisheries Management Council, 605 W 4th Ave, 16 Suite 306, Anchorage, AK 99501

262