Evaluation of Mysis Partial Diel Vertical Migration Brian Patrick O'malley University of Vermont

Home , Coregonus fontanae, Fish migration, Leptodiaptomus sicilis, Mysidacea, Mysis, Mysis diluviana

University of Vermont ScholarWorks @ UVM

Graduate College Dissertations and Theses Dissertations and Theses

2019 Evaluation of Mysis partial diel vertical migration Brian Patrick O'Malley University of Vermont

Follow this and additional works at: https://scholarworks.uvm.edu/graddis Part of the Natural Resources Management and Policy Commons

Recommended Citation O'Malley, Brian Patrick, "Evaluation of Mysis partial diel vertical migration" (2019). Graduate College Dissertations and Theses. 996. https://scholarworks.uvm.edu/graddis/996

This Dissertation is brought to you for free and open access by the Dissertations and Theses at ScholarWorks @ UVM. It has been accepted for inclusion in Graduate College Dissertations and Theses by an authorized administrator of ScholarWorks @ UVM. For more information, please contact [email protected].

EVALUATION OF MYSIS PARTIAL DIEL VERTICAL MIGRATION

A Dissertation Presented

Brian Patrick O’Malley

The Faculty of the Graduate College

The University of Vermont

In Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy Specializing in Natural Resources

January, 2019

Defense Date: October 23, 2018 Dissertation Examination Committee:

Jason D. Stockwell, Ph.D., Advisor Sara Helms Cahan, Ph.D., Chairperson Sture Hansson, Ph.D. J. Ellen Marsden, Ph.D. Joe Roman, Ph.D. Cynthia J. Forehand, Ph.D., Dean of the Graduate College

ABSTRACT

Mass animal migrations represent large movements of biomass, energy, and nutrients with predictable patterns and important ecosystem-level consequences. Diel vertical migration (DVM) in aquatic systems, the daily movement of organisms from deeper depths during the day to shallower depths in the water column at night, is widespread in freshwater and marine systems. Recent studies, however, suggest partial migration behavior, whereby only some portion of a population migrates, is the rule rather than the exception in a range of migratory fauna, including those that undergo DVM. Hypotheses to explain why partial migrations occur complicate traditional views on DVM and challenge conventional theories. I address intraspecific variation in DVM behavior of an aquatic omnivore, Mysis diluviana, to test several long-standing assumptions about benthic-pelagic DVM in Mysis. I evaluated the extent of partial DVM and several potential drivers within a Lake Champlain Mysis population. I used traditional net-based field observations, a novel deep-water video camera system, and a laboratory experiment, to compare distributions, demographics, abundance estimates, hunger-satiation state, and feeding behavior, of migrant and non-migrant Mysis across multiple seasons, habitats, and different times of the day. Findings from my dissertation suggest Mysis partial DVM is common, and is associated with body size and demographic differences among individuals. Partial DVM behavior, however, did not correspond to strong differences in feeding preference or hunger-satiation state of individuals. My results contribute toward a more comprehensive understanding of migration theory and mysid biology, by including the often overlooked, but important, benthic habitat component of DVM studies, and fills in several ecological knowledge gaps regarding a key omnivore in many deep lake food webs across North America where Mysis serve as both predators and prey to many organisms.

CITATIONS

Material from this dissertation has been published in the following form:

O’Malley, B. P., Hansson, S. and Stockwell, J. D.. (2018). Evidence for a size-structured explanation of partial diel vertical migration in mysids. Journal of Plankton Research 40: 66-76.

O’Malley, B. P., Dillon, R. A., Paddock, R. W., Hansson, S. and Stockwell, J. D.. (In Press). An underwater video system to assess abundance and behavior of epibenthic Mysis. Limnology and Oceanography: Methods. DOI: 10.1002/lom3.10289

ACKNOWLEDGEMENTS

I would like to thank my advisor, Jason Stockwell, for the opportunity to conduct graduate research at the Rubenstein Ecosystem Science Laboratory. I thank you for the motivation, guidance, and writing tips you have instilled upon me, and look forward to future collaborations and beyond.

Perhaps the most crucial element to accomplishing my field work has been captain Steve Cluett. Steve’s wisdom, expertise, and enthusiasm, was instrumental for all aspects of my project. Along with Steve, deckhands Krista Hoffsis and Brad Roy were also critical. I’ll treasure our conversations aboard the R/V Melosira while sipping coffee and looking back at Burlington on those long nights of Mysis sampling.

I would like to acknowledge several of my past supervisors and colleagues who continue to motivate me throughout my journey to pursue science as a career.

Conversations with David Bunnell formed the impetus for my interest in Mysis. Mentors from my Lake Ontario stint (Lars Rudstam, Jim Watkins, and Brian Weidel), were influential in shaping my career path. Great after-work discussions with friends Bryan

Maitland and Elliot Jackson, while fishing or talking zooplankton, were pivotal in my motivation to pursue graduate studies. Additionally, I thank members of my committee

(Sure Hansson, Sara Helms Cahan, Ellen Marsden, and Joe Roman) for their help in shaping my project. I’ve learned so much from collaborating with all of you, and I will continue to benefit from all these past experiences.

I thank the support of my parents Mark and Ewa O’Malley as they opened the initial door to research by supporting me as I pursued my undergraduate degree. That iii initial launch into undergraduate research and study was how I discovered a career in aquatic research, and ultimately found my niche in large lake research. None of that would have been possible without their support.

I thank Rebecca Dillon for her efforts to get the Mysis video camera system up and running, which helped pave the way for a concept to become a reality. Additionally, several undergraduates: Pascal Wilkins, Kevin Melman, Hannah Lister, Grace Scorpio,

Jeni Knight, and Jessica Griffin; provided substantial help and provided me with opportunities to learn about myself and hone my skills as a mentor. I hope the research experiences were equally rewarding for you as they were for me.

Funding was provided by the Great Lakes Fishery Commission with the assistance of Senator Patrick Leahy. Additional funding was provided by a University of

Vermont Chrysalis Award, and a scholarship from the International Association for Great

Lakes Research.

Lastly, and most importantly, I would like to thank my girlfriend Nikki Saavedra for her love and support during our long-distance lifestyle these past several years while we worked on our degrees. Your support, patience, and laughter throughout this process has been amazing. I look forward to starting our next chapter in life together.

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS ...... iii

LIST OF TABLES ...... viii

LIST OF FIGURES ...... x

CHAPTER 1: DIEL VERTICAL MIGRATION AND MYSID ECOLOGY FROM A PARTIAL MIGRATION PERSPECTIVE ...... 1

Diel vertical migration: examples, hypotheses, and conceptual framework ...... 3

Mysis ecology: distribution and vertical migration behavior ...... 6

Trophic role of Mysis in food webs ...... 12

The need for a partial DVM approach to mysid ecology ...... 15

Lake Champlain as a study system ...... 17

Dissertation outline ...... 19

References ...... 22

CHAPTER 2: EVIDENCE FOR A SIZE-STRUCTURED EXPLANATION OF PARTIAL DIEL VERTICAL MIGRATION IN MYSIDS...... 34

Abstract ...... 35

Introduction ...... 36

Methods...... 39

Results ...... 43

Discussion ...... 49

Conclusions ...... 55

Acknowledgements ...... 55

References ...... 57

CHAPTER 3: AN UNDERWATER VIDEO SYSTEM TO ASSESS ABUNDANCE AND BEHAVIOR OF EPIBENTHIC MYSIS ...... 65

Abstract ...... 66

Introduction ...... 67

Materials and Methods ...... 72

Description of system ...... 72 Spectral intensity and behavior experiment ...... 73 Field deployments and video analysis ...... 74 Field-based comparisons ...... 79 Results ...... 81

Spectral intensity and behavior experiment ...... 81 Field deployments and video analysis ...... 83 Field-based comparisons ...... 85 Discussion ...... 89

Acknowledgements ...... 95

References ...... 96

CHAPTER 4: DIEL FEEDING BEHAVIOR IN A PARTIALLY MIGRANT MYSIS POPULATION: A BENTHIC-PELAGIC COMPARISON ...... 103

Abstract ...... 104

Introduction ...... 106

Methods...... 110

Sample collections ...... 110 Sample analysis ...... 112 Results ...... 117

Discussion ...... 129

Conclusion ...... 135

Acknowledgements ...... 136 vi References ...... 137

CHAPTER 5: COMPARATIVE EVALUATION OF MIGRANT AND NON- MIGRANT MYSIS FEEDING PREFERENCES ...... 147

Abstract ...... 148

Introduction ...... 150

Methods...... 153

Field sampling ...... 153 Experimental setup ...... 154 Data analysis ...... 157 Results ...... 157

Discussion ...... 159

Acknowledgements ...... 164

References ...... 165

COMPREHENSIVE REFERENCES ...... 169

vii LIST OF TABLES

Page

Table 1.1: Compilation of studies that document partial migration in Mysis spp. (M. diluviana, M. mixta, M. relicta, and M. salemaai) at night. Studies that did not report quantitative values of Mysis spp. on bottom at night were only included if they reported presence/absence...... 11 Table 2.1: Number of Mysis collected during descent/retrieval of the benthic sled versus number collected in standard five-minute benthic sled tow, and the percent of pelagic Mysis in benthic tows in Lake Champlain, 2015. Values are the mean of two replicates...... 42 Table 2.2: Number of juvenile, female, male, and total mysids counted from night- benthic and night-pelagic samples by month pooled across both sites, and P- values of chi-squared tests comparing demographic proportions...... 49 Table 3.1: Qualitative tradeoffs with various sampling methods used to estimate benthic mysid abundance...... 71 Table 3.2: Date, time (EDT/EST), and nautical sunset times of benthic video camera deployments at the same 60-m location in Lake Champlain 2014-15. Deployment date/time refers to the time of day when the camera touched- down. Nautical sunset indicates the first time-interval counted, and marks the start of the night time period class...... 78 Table 4.1: One-way ANOVA results of the number of prey for each category in Mysis stomach contents over a diel cycle in Lake Champlain. F values that were significant following a Bonferroni correction (α = 0.05/12) for multiple comparisons are denoted by asterisks...... 125 Table 4.2: Mean size of Mysis and number of prey items, and the minimum and maximum number observed, in Mysis stomach contents collected at different times of day and habitat (benthic/pelagic). Grey denotes samples collected at night (between sunset-sunrise). Bold values denote months when significant differences in mean number of prey over a diel period were detected after Bonferroni correction. Superscripts refer to Tukey’s HSD post-hoc comparison for those cases...... 126 Supplemental Table 4.3: Sample collection times of four separate trips. Sunset/sunrise times are listed below each date in parentheses. Grey shading denotes samples collected during night. Time refers to the midpoint during the time interval of each sample...... 144 Supplemental Table 4.4: Zooplankton taxa, mean length (L in mm), and length-weight equations used to estimate dry biomass (W in dry µg). Equations follow the formula ln(W) = lnA + B • ln(L)...... 145 Table 5.1: Mean length of Mysis used in feeding experiment...... 158

viii Table 5.2: Two-way ANOVA results of Mysis predation rates (No. Daphnia hr-1)...... 159

LIST OF FIGURES

Page

Figure 1.1: (Left) Total or normal DVM where the entire population migrates near surface at night to feed. (Right) Partial DVM with some portion not migrating. Gradients of predation risk and growth potential are scaled from high (red) to low (blue). Predation risk varies with light (daily) while growth potential does not change within a day...... 4 Figure 2.1: Map of sampling sites by bottom depth for Mysis collections in Lake Champlain, USA, in June-November 2015...... 40 Figure 2.2: Monthly diel comparison of mean pelagic density of Mysis (top) estimated from vertical net tows from 2 m above lake bottom to surface tows in Lake Champlain 2015, and mean benthic Mysis density (bottom) estimated from benthic sled tows at 60- (left) and 100-m site sites (right). Open bars are day samples and filled bars are night samples (+ SD)...... 44 Figure 2.3: Mean (± SD) body length (mm) of pelagic (grey circles) and benthic- caught (black circles) Mysis at 60- and 100-m sites in Lake Champlain 2015 as a function of month sampled. Note that mesh size (1 mm) was consistent between the two gear types. Insufficient numbers of Mysis were captured in the pelagic habitat during the day at the 60-m site to compare mean length of day-benthic to day-pelagic individuals...... 46 Figure 2.4: Frequency of Mysis by size class pooled across all months (June-November 2015) in night-pelagic and night-benthic samples at 60- (top) and 100-m (bottom) sites in Lake Champlain...... 47 Figure 2.5: Percent composition of juvenile, female, and male Mysis collected in night- pelagic (top) and night-benthic samples (bottom) at 60- (left) and 100-m (right) sites in Lake Champlain...... 48 Supplemental Figure 2.6: Difference between mean lengths of benthic and pelagic Mysis collected at night by site (60 m = triangles; 100 m = circles) as a function of day of year. Blue line indicates linear model and grey ribbon represents 95% confidence intervals...... 63 Supplemental Figure 2.7: Water temperature isopleths at 60-m (left) and 100-m (right) where Mysis sampling took place in Lake Champlain...... 64 Figure 3.1: Benthic camera system during deployment. Labeled components are: (A) housing for battery pack; (B) housing with DVR; (C) red LED light; (D) video camera; (E) red lasers. Base of frame is 1.2 x 1.2 m. (photo credit: Maureen Walsh)...... 73

Figure 3.2: Map of benthic sled transects, stationary Mysis camera deployments, and locations of point sampling with the GoPro mounted mini-lander in Lake Champlain...... 75 Figure 3.3: Relative spectral intensity of red LED light (dashed line) and lasers (solid line) used on the camera system. Spectral sensitivity curve of Mysis eye (dotted line) from Gal et al. (1999) included for comparison...... 81 Figure 3.4: Percentage of Mysis observed under the red light half of cooler in the laboratory experiment. Percentages are the average of two observations every ten minutes during the first hour (circle), third hour (triangle), and seventh hour (square) of experiment...... 82 Figure 3.5: Coefficient of variation for Mysis benthic density as a function of binned time intervals for 1-hour video segment from June (circles) and October (triangles) 2015. For each average, N ranged from 60 for the 1-min interval to 2 for the 30-min interval...... 83 Figure 3.6: Mysis density (No. m-2) estimated from benthic video at 10-minute intervals for seven separate deployments in 2015 (black points) and one deployment in 2014 (blue triangles) at the same 60-m site in Lake Champlain. Grey boxes represent night defined as 1-hr post-sunrise to 1-hr pre-sunrise. Times are in EDT except for November which is EST. Moon phase and weather conditions are shown...... 84 Figure 3.7: Mean Mysis density (No. m-2) estimated from benthic video during the night vs day at the same 60-m site in Lake Champlain 2015...... 86 Figure 3.8: Synchronous change in orientation of Mysis with direction of bottom current on July 17th 2015 over a 6-hour period. (Left) Screenshots of Mysis at 06:00 (top), 10:00 (middle), and 12:00 EDT (bottom). (Right) Circular histograms with measurements of the direction individual mysids were facing in the corresponding image to the left...... 87 Figure 3.9: Mysis mean video density (No. m-2) observed as a function of mean benthic sled density (No. m-2) collected during the same 1-hr time window...... 88 Figure 3.10: (A) Observed Mysis density (No. m-2) by the mini-lander at the time of landing as a function of distance (m) from the stationary benthic Mysis video camera. Circles represent July and triangles refer to October 2017 sampling. (B) Comparison of mean density (± SD) observed on mini-lander (x-axis) to stationary benthic Mysis video camera (y-axis)...... 88 Supplemental Figure 3.11: Screenshot from deep-water camera system deployed at 60 m bottom depth in Lake Champlain, Vermont, USA. Date/time stamp on the video follow eastern time (24:00 hr MM/DD/YY format)...... 101 Supplemental Figure 3.12: Screenshot of slimy sculpin (Cottus cognatus) interacting with Mysis from deep-water camera system deployed at 60 m bottom depth in Lake Champlain, Vermont, USA. Date/time stamp on the video follow eastern time (24:00 hr MM/DD/YY format)...... 102 xi Figure 4.1: Local time of pelagic (P, red triangles) and benthic (B, black circles) sample collections during four separate trips on Lake Champlain. Sample times represent the midpoint of the time interval in which the sample was collected. Grey box is sunset to sunrise (see Supplemental Table 4.3 for exact time values)...... 111 Figure 4.2: (A) Mysis cephalic region after separation from the body with thoracic carapace removed. (B) Close-up view of Mysis cardiac stomach (circled region) (C-G) Gut scale ranking system for classifying gut fullness of Mysis stomach contents: (C) 0 = empty; (D) 1 = trace food up to one-quarter full; (E) 2 = half full; (F) 3 = three-quarters full; (G) 4 = fully gorged...... 115 Figure 4.3: Vertical profiles of temperature and chlorophyll-a fluorescence at the 100- m sampling site in Lake Champlain in June 2016, and April, July, and October 2017...... 118 Figure 4.4: Crustacean zooplankton biomass (mg/m3) and % composition in 60 m to surface vertical net tows at the 100-m site in Lake Champlain in June 2016 and July-October 2017...... 119 Figure 4.5: Correlation between number of total zooplankton prey (copepods, cladocerans, and rotifers) in Mysis gut contents (ZP) and gut fullness rank score from Lake Champlain in June 2016 and July-October 2017...... 120 Figure 4.6: Temporal changes in stomach fullness index (SFI) of benthic- (black circles) and pelagic-caught (red triangles) mysids by month from Lake Champlain. Grey box is sunset to sunrise...... 121 Figure 4.7: Variation in detritus scores of pelagic and benthic Mysis by time of day in June 2016 and July and October 2017 in Lake Champlain...... 123 Figure 4.8: Mean (± SE) detritus rank of stomach contents from benthic- and pelagic- caught Mysis by time of day. Points labelled with the same letter denote no significant difference among time of day (Dunn test with Bonferroni correction). Letters are located above for benthic and below for pelagic samples. * indicates excluded estimate due to small sample size...... 123 Figure 4.9: Frequency of occurrence of prey items by habitat-time sampled (BD = benthic-day; BN = benthic-night; PN = pelagic-night). Prey items with values of 0% are not shown as symbols. Pelagic prey items are ordered by decreasing body size (green font), from Keratella to Senecella; last four prey items of list (Dinoflagellate.cyst-Detritus.4.only) are strictly benthic prey (brown font)...... 129 Supplemental Figure 4.10: Variation in the gut fullness ranks assigned to Mysis diets in pelagic and benthic habitats collected at different times over a diel cycle from 100-m site in Lake Champlain in June 2016 and April, July, and October 2017. No data were available from the pelagic habitat in April...... 146 Figure 5.1: Schematic of Mysis feeding experiment. For each food treatment (Daphnia-only, Daphnia-detritus, and Detritus-only), a single Mysis was

xii placed in each of 10 1-L jars and allowed to forage on the available food over the allotted time under darkness...... 155 Figure 5.2: Observed ingestion rates of Mysis on Daphnia as a function of Mysis body size...... 159

xiii

CHAPTER 1: DIEL VERTICAL MIGRATION AND MYSID ECOLOGY FROM A

PARTIAL MIGRATION PERSPECTIVE

The wide range of behaviors and spatio-temporal scales used by a wide range of organisms and across all types of habitat is a remarkable feature of the natural world.

From the cross-hemisphere seasonal migrations of Arctic terns (Egevang et al. 2010) to the daily rhythmic vertical migrations by freshwater and marine fauna (e.g., Guisande et al. 1991; Hays 2003; Cartamil and Lowe 2004), migration represents a dynamic array of tradeoffs and choices to hedge bets for survival and reproduction of individuals.

Migration also connects disparate habitats through movement and exchange of energy and nutrient subsidies (Gende et al. 2002; Bauer and Hoye 2014; Darnis et al. 2017).

Consequently, animal migration profoundly influences biogeochemical cycles and food web structure (Longhurst and Harrison 1988; Polis et al. 1997; Jónasdóttir et al. 2015;

Oliver et al. 2015). Understanding how organisms interact within and among landscapes and the decision-making processes governing their behavior are critical for predicting future responses to human-related impacts on ecosystems, including climate change, habitat fragmentation, and species removal.

Migration is the result of evolution working to optimize an animal’s fitness in response to conflicting biotic and abiotic pressures (Milner-Gulland et al. 2011). More recently, the concept of partial migration, where some portion of a population does not migrate, has been recognized as a ubiquitous feature across a range of migratory birds, fishes, and mammals (Chapman et al. 2011, 2012; Mehner and Kasprzak 2011). Recent

work on partial migration shows alternative outcomes and impacts on ecosystem structure and function that are inconsistent with assumptions of whole-population movement (Broderson et al. 2008a, 2011; Chapman et al. 2012). For example, the timing and magnitude of partial migration by planktivorous fish influences alternative stable states in lakes (Broderson et al. 2008a). Therefore, partial migration challenges conventional theories of migration because opposing behaviors are unlikely explained by the same drivers.

Identifying the factors associated with partial migrations and the magnitude or variation associated with these movements is essential for predicting how organisms will respond to ecosystem perturbations. Individual tradeoffs between body-condition and predation risk can act as drivers of partial migrations, among other potential factors

(Chapman et al. 2011). By tracking individual fish over several months, Broderson et al.

(2008b) showed that fish with low energy stores remained in high-risk, high-reward settings while individuals with sufficient energy stores migrated into low-risk, low- reward habitat. The findings of Broderson et al. (2008b) provide valuable insights for how individuals might respond to a future disturbance. For example, if a future disturbance creates limited food availability or increased competition for food availability in a fish population, then under the body-condition hypothesis one might expect more fish with poor-condition and subsequently more of the population to forego migration away from the high-risk habitat. Researchers and managers could incorporate knowledge about drivers of animal migrations at the individual level to make informed predictions about how best to manage populations and evaluate ecosystem consequences of shifts in migration patterns.

Diel vertical migration: examples, hypotheses, and conceptual framework

Diel vertical migration (DVM) in freshwater and marine systems is considered the largest synchronized mass movement of organisms in the world (Hays 2003). The typical pattern of DVM, also commonly referred to as normal DVM, is the movement of organisms from deeper depths where they reside during the day, up to shallower depths after sunset where they remain throughout the night, followed by descent back down at sunrise. Organisms across the globe, ranging from algae up to apex predators, partake in

DVM (e.g., Cullen 1985; Guisande et al. 1991; Stockwell et al. 2010). Large fluxes in biomass distribution brought about by DVM are important to predict to accurately quantify spatial subsidies of energy or biomass across habitat boundaries (Schindler and

Scheuerell 2002; Haupt et al. 2009; Sierszen et al. 2014). DVM in one species can influence the migration behavior of another species, which in turn can influence the next trophic level and so on, thereby leading to cascading effects of DVM behavior on distributions throughout the food web (Bollens et al. 2010). Understanding the vertical movements and factors that drive DVM in one species may in turn help with understanding the behavior and diel distribution patterns of other species within a system

(Bollens et al. 1992; Eshenroder and Burnham-Curtis 1999; Williamson et al. 2011).

Abrupt vertical gradients in temperature, light, food, and predators (Fig. 1 Left) provide a template to conceptualize energetic tradeoffs in growth and survival with respect to habitat (Loose and Dawidowicz 1994; Boscarino et al. 2009; Mehner 2012). A substantial body of evidence suggests DVM has evolved as a mechanism to minimize the ratio of predation risk () to growth (G) potential (sensu Werner et al. 1983), a framework shared across marine and freshwater settings (Gliwicz and Pijanowska 1988; Lampert 1993;

Hays 2003). DVM theory predicts that /G will be lowest at night in food-rich surface waters when visual predators are at a disadvantage (low-risk, high-reward), and lowest during the day down deep where food may be less available but visual predators are still at a disadvantage (low-risk, low-reward; Fig. 1 Left). Species known to exhibit DVM have also demonstrated the capacity to undergo reverse migration (e.g., predators migrating down in the water column at night, Jensen et al. 2011) or partial migration

(Mehner 2014). Deviations from the normal pattern of DVM highlight the flexibility of the behavior and the importance of considering individual variability in unraveling migration decision-making processes.

DVM is a theory-rich field with many hypotheses proposed to explain the adaptive value of DVM behavior. The proximate cause that triggers DVM is generally accepted to be diel change in light intensity (Cohen and Forward 2009). Conversely, ultimate causes for the adaptive significance of DVM vary substantially. Mangel and

Clark (1988) compiled thirteen different hypotheses from the literature for DVM, that 4

range from tactics to minimize exposure to harmful UV radiation (Hairston 1976) to increased energetic efficiency (McLaren 1974). Many of the proposed hypotheses are not mutually exclusive making it difficult to tease apart the exact mechanisms behind DVM.

Furthermore, while proximate factors can be tested with controlled experiments or observations (Cohen and Forward 2009), the evolution of ultimate causes responsible for behavior are more difficult or impossible to test because of difficulties comparing fitness consequences of alternative phenotypes at present to past forms (Reeve and Sherman

1993; Mehner et al. 2007; Eshenroder and Burnham-Curtis 2001). Nonetheless, the most widely accepted explanation for DVM in lower trophic-level organisms is that the behavior has evolved as predator-avoidance mechanism (Zaret and Suffren 1976; Gliwicz

1986; Hays 2003; Ringelberg 2010). Justification for the predator-avoidance hypothesis comes from quantifying tradeoffs through modelling of predation-risk and growth potential with respect to vertical gradients of predators, prey, and abiotic conditions

(Zaret and Suffren 1976; Loose and Dawidowicz 1994; Jensen et al. 2006). Field observations from systems with fluctuations in predator density also indicate support for predator-avoidance. For example, migration amplitude of zooplankton increased when predator densities or concentrations of predator-derived kairomones increase suggesting that DVM is sensitive to predation risk (Ringelberg et al. 1991; von Elert and Pohnert

2000). Similarly, in a study of lakes with and without fish predators, the cyclopoid copepod Cyclops absorssum only performed DVM in lakes where fish were present

(Gliwicz 1986).

In the case of meroplanktonic organisms (benthic-pelagic life history) that undergo DVM, such as freshwater mysids (Mauchline 1980), a recurring question about

their ecology is: Does DVM represent a daily routine-behavior experienced by all individuals? A key assumption with DVM is that organisms performing this behavior would prefer to be at shallow depths under ideal circumstances because of the greater growth and reproductive advantages associated with shallow depths; they only migrate because of predators (Kerfoot 1985; Figure 1.1). However, for organisms adapted to feeding on benthic resources, how certain are we that this assumption is valid to explain

DVM in these organisms? For instance, could the preferred strategy of meroplankton actually be to stay benthic, and that migration into the pelagia is actually the less profitable behavior? Similarly, the questions of can meroplanktonic organisms meet energetic requirements by feeding solely on benthic resources, is rarely evaluated in species that undergo DVM between pelagic and benthic habitats. Assimilation efficiency and food quality of detritus is assumed low relative to planktonic resources (Lasenby and

Langford 1973), however this remains untested in mysids.

Mysis ecology: distribution and vertical migration behavior

Members of the Mysis relicta species complex are considered glacial relicts distributed throughout temperate and boreal zones of North America and Europe (Ricker

1959; Väinölä et al. 1994). They can have large ecological impacts on lakes and reservoirs when introduced (Lasenby et al. 1986; Spencer et al. 1991). In temperate regions of North America, the M. relicta complex is restricted to a single species1 mm Mysis diluviana (formerly M. relicta, hereafter referred to as Mysis). Mysis is native to deep glacial lakes mainly east of the Rockies including the Laurentian Great Lakes, the Finger

Lakes in NY, and several lakes throughout the Canadian provinces of Ontario and

Quebec. In Vermont, Mysis has been reported from lakes Dunmore, Champlain, 6

Memphremagog, and St. Catherine (Gutowski 1978). In the western and some eastern locations of the U.S., Mysis was intentionally introduced via stocking in lakes and reservoirs (and spread to other water bodies) with the intention of bolstering fish populations (Brown 1998; Lasenby et al. 1986; Martinez and Bergersen 1989; Johnson et al. 2018). The practice of stocking Mysis, however, led to undesired outcomes by causing declines in cladoceran zooplankton and reduced growth rates of kokanee salmon

Oncorhynchus nerka demonstrating the need to understand the ecological roles of Mysis on food webs (Lasenby et al. 1986; Nesler and Bergersen 1991; Spencer et al. 1991;

Rudstam and Johannsson 2009; Walsh et al. 2012). In native environments, such as the

Laurentian Great Lakes, M. diluviana can be abundant and potentially consume more zooplankton than fish (Gal et al. 2006; Bunnell et al. 2011). Mysis accounts for approximately 30% of pelagic crustacean biomass in offshore Lake Ontario (Watkins et al. 2015) and recent lake-wide surveys of Lake Superior estimated a population of 9.9 trillion individuals (Pratt et al. 2016). Both the Lake Ontario and Superior estimates are based on night-pelagic sampling of the population during DVM and could be conservative if pelagic individuals are only part of the total population. Therefore, if even a small proportion of the population does not migrate at night, the non-migrant portion could still represent a substantial biomass component in many lakes as evidenced by their importance in food web models as both predators of plankton and a key prey species of most fish species (Gamble et al. 2011a,b; Isaac et al. 2012; Kitchell et al. 2000; Rogers et al. 2014; Stewart and Sprules 2011).

Mysis exhibits extensive DVM, capable of migrating > 100 m in a single night

(Beeton 1960). Based on conventional DVM theory, the ultimate cause of Mysis DVM is

considered to be predator avoidance from fish (Beeton 1960; Eshenroder and Burnham-

Curtis 1999; Boscarino et al. 2009) with light as the proximate factor triggering present- day migrations (Beeton 1960; Teraguchi 1969; Boscarino et al. 2009, 2010). Previous work on Lake Ontario suggests that light and temperature are more suitable for predicting the pelagic depth distribution of Mysis densities than predators or food availability

(Boscarino et al. 2009), but this model does not include benthic-dwelling mysids and therefore may only be useful for predicting the distribution of a select subset of individuals. During nightly migrations, Mysis vertical distributions generally show peak densities in the meta- or upper hypolimnion, remaining in water temperatures < 10°C

(Boscarino et al. 2009). Lab experiments have shown that mortality generally increases at temperatures > 12°C (Rudstam et al. 1999). Lipid accumulation rates are higher at cooler water temperatures if adequate food supply is available (Chess and Stanford 1999) suggesting energetic and survival advantages associated with remaining at depth. Mysis feed by filtering or using raptorial tactics to grab prey (Mauchline 1980). They have relatively large compound eyes that can be used to sense light and visually locate prey unlike other zooplankton (e.g., Daphnia) which have less developed eyes not capable of perceiving prey but rather function only to sense changes in lighting (Ringelberg 2010).

Experimental evidence suggests that foraging rates of Mysis are higher at intermediate water temperatures and light levels, relative to conditions normally experienced at deeper lake depths (Ramcharan and Sprules 1986, Rudstam et al. 1999). Ramcharan and Sprules

(1986) observed higher predation rates on zooplankton at low light levels than complete darkness, and proposed Mysis use visual feeding. Feeding rates follow a Gaussian distribution, increasing at intermediate water temperatures typically experienced in the

lower metalimnion with peak feeding rate at 9°C (Rudstam et al. 1999; Gal et al. 2004).

Thus, Mysis may feed more efficiently on zooplankton prey by migrating into sub- thermocline habitat where light levels and temperature may be higher than near the lake bottom, along with greater zooplankton densities.

Sediments in aquatic habitats serve as a sink for settled organic material and support active areas of nutrient recycling (Baker et al. 1991, den Heyer and Kalff 1998).

Benthic invertebrates play a key role in the resuspension and conversion of particulate carbon into alternative forms via bioturbation, ingestion, excretion, and egestion (Baker et al. 1991; Covich et al. 1999; Holker et al. 2015). Under DVM scenarios, mysids can feed on benthic resources during the day, and transport benthic nutrients back into the water column at night via excretion and egestion during pelagic foraging at night (Van

Duyn-Henderson and Lasenby 1986; Chipps and Bennett 2000; Sierszen et al. 2011).

Also, during night pelagic foraging, mysids can contribute to nutrient recycling within the upper layers of the water column or repackage nutrients into fast-sinking fecal pellets that are expedited to much deeper depths (Madeira et al. 1982; Chipps and Bennett 2000;

Linden and Kuosa 2004; Caldwell 2010). Nutrient recycling is probably higher at night when mysids are pelagic because egestion rates increase with temperature and feeding activity (Murtaugh 1984), although results from Lake Superior suggest Mysis excretion/egestion has little effect on algal production relative to the nutrient regeneration of zooplankton (Oliver et al. 2015).

Apart from describing Mysis DVM, several studies have also reported Mysis on or near the lake bottom at night (Table 1.1; Bowers 1988; Shea and Makarewicz 1989), with some studies proposing that populations actually follow a bimodal vertical distribution at

night (Morgan 1980; Bowers 1988; Rudstam et al. 1989). Therefore, Mysis display partial diel vertical migration with some individuals migrating and some remaining resident

(Figure 1.1 Right). Observations of partial DVM in mysids raise several questions about the theory of DVM behavior which for invertebrates has largely been described as a population-level movement driven by predation-risk/growth tradeoffs. Conventional theory predicts foraging potential to be higher near surface than at depth, however the basis for this assumption comes from studies of holoplankton which only use pelagic habitat. However, meroplanktonic organisms such as Mysis can not only occupy both habitats, but can also feed successfully in both habitats. As a result of a meroplanktonic niche, Mysis likely have a greater growth potential at depth than strictly holoplanktonic species, and therefore may not conform to existing constructs for quantifying growth potential advantages of DVM. Mysis escape tactics from predators can vary between benthic vs. pelagic habitat (Bowers et al. 1990) leading to habitat-specific differences in predation vulnerability regardless of light or predator densities. Mysis propels away from predators with greater acceleration and maximum speed in benthic habitat than in the water column by pushing-off the sediment surface (Bowers et al. 1990). In summary, the biological differences in Mysis relative to holoplankton and consequences of benthic and pelagic strategies create opportunities to challenge unifying constructs of DVM theory that to date have been largely based on assumptions about vertical gradients relevant to animals that rely solely on pelagic habitat throughout ontogeny.

Species Site Depth Season†, Year % On Method‡ Source (m) Bottom M. relicta (diluviana) L. Superior 250 Sum, 1986 50% Submarine; Pel Bowers (1988) M. relicta L. Snasavatnet 48 Spr-Aut, 1984 23-70% Ben; Pel Moen & Langeland (1989) (Norway) M. relicta L. Jonsvatn (Finland) 10-80 Spr-Win, 1986-87 7-84%* Ben; Pel Naesje et al. (2003) M. mixta Baltic Sea 28-40 Sum-Aut, 1985-86 30% Ben; Pel Rudstam et al. (1989) M. relicta (diluviana) L. Ontario 35-100 Spr-Aut, 1984 < 5% Ben; Pel Shea & Makarewicz (1989) M. relicta L. Breiter & 14-40 Spr-Aut, 2001-04 Present Scuba; Pel Waterstraat et al. (2005) Schmaler Luzin (Germany) 11 M. relicta (diluviana) L. Ontario 35-75 Spr-Aut, 1984 Present Predator diet Brandt (1986)

M. relicta (diluviana) L. Ontario 125 Spr-Aut, 1995 Present Ben; Pel Johannsson et al. (2001, 2003) M. mixta; M. Baltic Sea 20-40 Spr-Aut,1985-86; 15% Camera A. Staaf & S. Hansson relicta/salemaii 1996-97 (unpublished data, Stockholm Univ.) M. relicta; M. salemaii Baltic Sea 30-35 Aut, 2008 Present Ben Ogonowski et al. (2013) M. diluviana L. Champlain 70-120 Aut, 2014 Present Ben Euclide et al. (2017) M. diluviana L. Champlain 60-100 Spr-Aut, 2015 3-46% Ben; Pel O’Malley et al. (2018) * in terms of biomass † Spr = Spring (April-June), Sum = Summer (July-August), Aut = Autumn (September-November), Win = Winter (December-March) ‡ ‘Ben’ refers to benthic sled, beam trawl, or epibenthic sledge; ‘Pel’ refers to pelagic net towed vertically or horizontally.

1 Trophic role of Mysis in food webs

Mysis has been characterized as an opportunistic omnivore that consumes a wide

range of benthic and pelagic prey resources such as zooplankton, phytoplankton,

amphipods, insects, detritus, pollen (Rybock 1978; Grossnickle 1982; Johannsson et al.

2001, 2003; Caldwell et al. 2016), and even other mysids (McWilliam 1970; Nordin et al.

2008). One consistent finding is that Mysis target cladoceran prey when available (e.g.,

Daphnia spp. and Bosmina spp.; Cooper and Goldman 1980; Bowers and Vanderploeg

1982; O’Malley and Bunnell 2014) which Mysis only has access to during migration in

the epi- and metalimnetic zones during summer-fall months. However, studies that

identified preference for cladoceran prey have been restricted to analysis of pelagic-

caught Mysis during DVM or lab feeding experiments in which only pelagic zooplankton

were available, and failed to test preference given the true range of resources actually

available to a Mysis in a lake. A more comprehensive test of diet selectivity would

analyze preference in the presence of benthic and pelagic resources. Mysis alter their diet

when presented with changes in prey availability by incorporating larger-sized, non-

native predatory cladocerans (i.e. the fish-hook water flea, Cercopagis pengoi, and the

spiny water flea, Bythotrephes longimanus) in lakes where they occur (Nordin et al.

2008; O’Malley and Bunnell 2014; O’Malley et al. 2017). Techniques used to describe

mysid diets include: stomach analyses (Viherluoto et al. 2000; Nordin et al. 2008;

O’Malley and Bunnell 2014), stable isotope estimation coupled with mixing models

(Johannsson et al. 2001; Sierszen et al. 2011), estimated clearance rates in contained

environments (Bowers and Vanderploeg 1982), gut fluorescence (Grossnickle 1979;

O’Malley et al. 2017) and genetic analyses (Gorokhova 2009).

France (2012) suggested that Mysis tend to be more herbivorous in low productivity environments, omnivorous at intermediate levels, and mostly zooplanktivorous in more productive lakes. Other work has indicated that the extent of

Mysis omnivory varies within a lake primarily by body size and season (Branstrator et al.

2000; Johannsson et al. 2003; O’Malley and Bunnell 2014). Overall the degree of herbivory in Mysis varies with season, mysid size, and likely the environment. In Lake

Ontario, grazing on pelagic phytoplankton appears to be an important feeding strategy of

Mysis during spring when phytoplankton constitute up to 50% of the overall diet, while in summer through fall, mysids shifted towards a greater reliance on zooplankton and rarely fed on phytoplankton (Johannsson et al. 2001). While spring may be an important grazing period, Mysis also feeds on diatoms and other phytoplankton outside of spring potentially using algae present in the deep chlorophyll layer after the onset of thermal stratification

(Bowers and Grossnickle 1978; Grossnickle 1979; O’Malley and Bunnell 2014). Mysids select large-sized diatoms that would otherwise be inedible to smaller-sized zooplankton that can only feed on smaller forms of phytoplankton (Bowers and Grossnickle 1982;

Grossnickle 1982). Ingestion and assimilation efficiency of large diatoms by Mysis is <

100% because they break apart diatoms with their mandibles before ingestion (review by

Takahashi 2004), which benefits copepods that can utilize these smaller fragmented diatoms as they settle out of the upper water column, creating an indirect benefit/service of Mysis grazing to primary consumers (Bowers and Grossnickle 1982).

Empirical data describing the benthic and pelagic contributions to Mysis diet are rare. Most studies have focused on pelagic aspects of Mysis foraging by studying diet of individuals collected during nocturnal migration into the water column (e.g., Whall and

Lasenby 2009; O’Malley and Bunnell 2014; Hyrcik et al. 2015). Mysids are also capable of consuming benthic invertebrates and detritus during the day when they move downwards to avoid light or possibly when some mysids remain on or near the bottom at night (Bowers 1988; Shea and Makarewicz 1989; Johannsson et al. 2001, 2003). Such a dynamic strategy is important to understand and predict to quantify energy flow in deepwater habitats given the relatively large biomass and food web importance of Mysis in lakes where they occur. Mysis have been hypothesized to consume a higher proportion of benthic resources to compensate for low pelagic production during periods of low plankton abundance (i.e., winter, Patwa et al. 2007). Consistent with the hypothesis that seasonality dictates where in a lake Mysis feed, a higher proportion of Mysis are assumed to remain on the bottom during fall through spring, when food availability is low compared to summer (Johannsson 1995; Patwa et al. 2007). Mysis can be a predator on benthic invertebrates such as the amphipod Diporeia (Parker 1980), but the potential impacts of Mysis predation on benthic invertebrate fauna have only recently been discussed (Stewart and Sprules 2011). For example, Stewart and Sprules (2011) suggested that Mysis was likely a main predator on Diporeia in Lake Ontario during the

1990s given the high biomass of Mysis relative to fish in offshore Lake Ontario, and that the overlap of these two invertebrates could have been even greater as water clarity increased, forcing more mysids to reside near bottom during the day (Stewart and

Rudstam 2012). Similar realizations may occur in other systems as well if quantitative estimates of non-migratory mysids at night become available to the research community.

The need for a partial DVM approach to mysid ecology

Understanding the population dynamics, behavior, and trophic ecology of key organisms in food webs is important for depicting energy flow. In food web models, a species’ ecological variation is often ignored or compressed into a single niche for simplification (Pimm 1982; Pimm et al. 1991; Polis 1991). In nature, however, this is not always the case, and in fact, numerous examples of intraspecific variation suggest the role of a species can be far more complex and dynamic (Bolnick et al. 2003; Bolnick et al. 2011). In Lake Michigan, Mysis was ranked as the most important keystone functional group in a recent ecosystem model (Rogers et al. 2014), highlighting the importance to consider the prevalence of alternative behaviors are and the extent to which such alternative behaviors modify our perspective compared to a single niche model.

In the case of Mysis, a key forage species for Great Lakes fishes (Gamble et al.

2011a,b), biological monitoring programs have been designed to assume individuals migrate into the pelagia at night. Consequently, Mysis assessment protocols rely solely on nocturnal pelagic sampling, which fails to account for non-migrant benthic individuals.

Despite decades of research documenting the profound pelagic effects of mysids on food webs (Lasenby et al. 1986; Rudstam and Johannsson 2009), very little if any effort has been made to develop predictive models and sound theories exist to explain why some mysids remain on bottom at night. If a substantial portion of the population remains benthic at night, then Mysis may have equally important effects on benthic ecosystems as they do on pelagic. Assumptions about DVM theory as a whole-population movement 15

and a tradeoff between growth/predation-risk is likely limiting empirical evaluation of these critical assumptions for deep-water organisms. For example, recent food web modelling highlighted the potential for Mysis predation as a cause for declines in Lake

Ontario’s Diporeia population (Stewart and Sprules 2011; Stewart and Rudstam 2012).

Equally concerning is whether Mysis may be experiencing increased predation pressure from fish in the Great Lakes, in response to Diporeia collapse, because both Mysis and

Diporeia were historically major prey items of most Great Lakes fishes (Gamble et al.

2011a,b; Isaac et al. 2012; Bunnell et al. 2015). As a result, the Great Lakes Fisheries

Commission and other agencies have set research priorities to understand the dynamics of

Mysis and monitor their abundance over time throughout the Great Lakes (GLFC 2015;

Jude et al. 2018), including the adaptive capacity of key organisms to respond to changing conditions (McMeans et al. 2016).

From a lake management perspective, if the proportion of Mysis occupying benthic habitat at night is low, then pelagic estimates are appropriate for use in management decisions requiring estimates of Mysis abundance and current pelagic sampling is sufficient. If benthic contributions are high, then current Mysis biomass and production estimates are biased low and potentially misleading. A re-evaluation of profundal food web structure and function and outcomes of management actions would then be necessary to include more realistic estimates of Mysis biomass, distribution, and top-down effects on benthic organisms. Moreover, if partial DVM is related to, for example, seasonality, demographics, feeding, or environmental conditions, then sampling protocols can be better defined to account for these drivers of Mysis behavior. From a theoretical perspective, study of partial DVM in Mysis will address several assumptions

surrounding DVM theory by evaluating classic views of DVM based on whole- population rather than individual processes that might explain intraspecific variability in migration behavior.

Variation in DVM behavior may be a result of a population with specialists opposed to generalists. Individual specialization has been revealed for many populations that were assumed to be generalists (Bolnick et al. 2003), leading to substantial progress in population and community ecology (Bolnick et al. 2003, 2011; Violle et al. 2012).

Specialization can occur in many forms and one widely considered explanation is a competitive advantage for different resources (Darimont et al. 2007; Svanback and

Bolnick 2007; Araújo et al. 2011). Diets of benthic- and pelagic-caught Mysis at night have not been thoroughly considered, but one hypothesis is that partial migration is related to specialized feeding (Chapman et al. 2011 and references therein). A study from

Lake Ontario noted some differences in diet between benthic-caught mysids just before dawn and pelagic-caught mysids at night from a 125-m station (Johannsson et al. 2001).

Main findings were that benthic individuals contained more amphipod remains and diatoms in their gut contents, and pelagic individuals had higher frequency of zooplankton. In summary, partial migration could be due to different feeding behavior, but as pointed out by Chapman et al. (2011), this hypothesis has not been widely tested.

Lake Champlain as a study system

Lake Champlain is a 170-km long, deep lake (max. depth = 122 m) located along the borders of New York, Vermont, and Quebec. A majority of the lake by surface area is classified as oligo-mesotrophic, though the lake is segmented into five distinct segments

(South Lake, Main Lake, Mallets Bay, Northeast Arm, Missiquoi Bay) that range from 17

oligotrophic to eutrophic (Facey et al. 2012; Xu et al. 2015). Mysis are found in the Main

Lake basin where most deep-water habitat occurs (Gutowski 1978; Ball et al. 2015).

Overall, benthic invertebrate biomass in Lake Champlain is low compared to the Great lakes and other large lakes of the northeast (Myer and Gruendling 1979; Knight et al.

2018). The nearshore benthic region contains invasive zebra mussels (Dreissena polymorpha) that were first reported in the South Lake in 1993 (Marsden et al. 2013).

The benthic amphipod Diporeia, which was historically abundant in the Great Lakes and often co-occurs with Mysis in many deep lakes (Dadswell 1974; Nalepa et al. 2009), is rare in Lake Champlain and has not been a prominent member of the benthos dating back to historical records from the 1960s (Dermott et al. 2006; Knight et al. 2018). Quagga mussels (D. rostriformis bugensis), which are similar to zebra mussels, and tend to dominate the profundal benthos in lakes where they have invaded, have not yet been detected in Lake Champlain but could invade the lake with potentially large ecosystem- level effects (Marsden and Hauser 2009; Marsden et al. 2013; Knight et al. 2018).

Since 1992, Mysis abundance has been monitored at deep stations as part of the

Lake Champlain Long-Term Biological Monitoring Project (LTMP). The LTMP dataset is based on pelagic vertical tows during the day to generate density estimates. Day- pelagic tows, however, are rarely used for Mysis monitoring programs elsewhere because

Mysis are light-sensitive and population trends derived from day-pelagic tows could be biased from variability in light intensity or changes in water clarity, in addition to substantially underestimating true abundance when animals are concentrated on the bottom during the day. Ball et al. (2015) analyzed trends in mysid abundance from the

LTMP dataset from 1992 to 2008 and identified a drastic decline in pelagic-day Mysis

densities after 1995. Mechanisms for the decline are not well understood. One hypothesis for the decline is that declines in rotifer abundance, a known prey of Mysis, observed over the same time period created food limitation for Mysis (Mihuc and Recknagel 2018).

However, rotifers are not perceived as a main prey item for Mysis in Champlain or other lakes (O’Malley and Bunnell 2014; Hrycik et al. 2015), therefore food limitation due to rotifer declines remains a correlative rather than mechanistic explanation. An alternative hypothesis is the decline is not actually a change in abundance but rather an artifact due to a change in vertical distribution if water clarity has increased. Changes in water clarity of the main lake were not evident over the same time-period (Smeltzer et al. 2012), suggesting that the population decline is real and not just an artifact of a change in daytime vertical distribution (Ball et al. 2015). Similar long-term data on night-pelagic density would be valuable to compare to day-derived population trends and evaluate changes in vertical distribution. However, historical Mysis abundance estimates at night are unavailable and a comparison of day vs night pelagic density estimates has not been evaluated in Lake Champlain.

Dissertation outline

In the following dissertation chapters, I evaluate several aspects of partial migration in Mysis diluviana from a Lake Champlain population. My aim is to better define the role of partial vertical migration to understand the benthic and pelagic distribution and trophic roles of Mysis in lake food webs, and thus contribute to the field of migration ecology and specifically partial DVM. In the body of my dissertation I address several research questions including: 1) Do demographic differences exist among migrant and non-migrant individuals? 2) Does traditional net-based sampling adequately 19

sample the benthic region or are new camera-based techniques more effective? 3) Does feeding behavior vary over a diel cycle and do diets differ among migrants and non- migrants? and 4) Are migratory mysids more efficient predators on pelagic zooplankton prey than benthic non-migrants? Collectively, the following chapters synthesize several new aspects of partial diel vertical migration and mysid biology.

The second chapter, “Evidence for a size-structured explanation of partial diel vertical migration in mysids”, addresses questions about demographic differences between migrants and non-migrants, and potential implications for seasonal abundance estimates in population assessment programs. Most monitoring programs rely solely on pelagic sampling at night to estimate mysid stock abundance, growth rates, and assess long term trends in abundance. However, to date the benthic portion of a population is most often ignored and assumed negligible in food web models. If large amounts of individuals remain close to the lake bottom at night and not susceptible to pelagic sampling methods, then such behavior could lead to major bias in studies that assume pelagic abundance is representative of the whole-lake population. Accounting for the proportion of the population that does night migrate in terms of abundance, biomass, size structure, and sex ratio will lead to more realistic depictions of Mysis population dynamics in food web models.

In the third chapter, “An underwater video system to assess abundance and behavior of epibentic Mysis”, a novel underwater video system is described and applied to observe the abundance and behavior of Mysis near the lake bottom over the diel cycle.

The camera system is a self-contained, stationary unit that was deployed multiple times at a 60-m site in Lake Champlain. Gear comparisons between the camera, benthic sled, and

a more portable video camera device were used to confirm findings that Mysis were present on the lake bottom throughout the night, and showed little difference in benthic density, suggesting that Mysis are continuously present near the bottom, throughout the night. Moreover, density estimates from camera footage were higher than those from concurrent benthic sled tows, suggesting benthic sled tows may underestimate Mysis density.

In the fourth chapter, “Diel feeding behavior of a partially migrant population: A benthic-pelagic comparison”, I used fine-scale sampling over multiple 24-hour periods to test the hunger-satiation hypothesis for partial DVM. Stomach fullness and composition was compared in mysid diets sampled from benthic and pelagic habitats at night, and benthic habitat throughout the day. In general, Mysis diets were rarely empty, regardless of time of day or habitat, suggesting they feed continuously and do not exhibit a diel pattern in gut fullness related to the hunger-satiation hypothesis.

In the fifth chapter, “Comparative evaluation of migrant and non-migrant Mysis feeding preferences”, a lab experiment was used to test the roles of prey availability on foraging of benthic vs pelagic caught Mysis. I tested the prediction that Mysis caught in the pelagia at night (migrants) prefer pelagic food, whereas benthic Mysis (non-migrants) prefer benthic food. Feeding rates and stomach contents of Lake Champlain Mysis were compared from individuals exposed to one of three treatments: pelagic food only, benthic food only, or both. Contrary to my prediction, I found that even in the presence of

Daphnia (preferred zooplankton prey), Mysis still readily consumed sedimented detritus in treatments where Mysis were offered both.

References Araújo, M. S., D. I. Bolnick, and C. A. Layman. 2011. The ecological causes of individual specialisation. Ecology Letters 14:948-958. Baker, J. E., S. J. Eisenreich, and B. J. Eadie. 1991. Sediment trap fluxes and benthic recycling of organic carbon, polycyclic aromatic hydrocarbons, and polychlorobiphenyl congeners in Lake Superior. Environmental Science and Technology 25:500-509. Ball, S. C., T. B. Mihuc, L. W. Myers, and J. D. Stockwell. 2015. Ten-fold decline in Mysis diluviana in Lake Champlain between 1975 and 2012. Journal of Great Lakes Research 41:502-509. Bauer, S. and B. J. Hoye. 2014. Migratory animals couple biodiversity and ecosystem functioning worldwide. Science 344:1242552. Beeton, A. M. 1960. The vertical migration of Mysis relicta in Lakes Huron and Michigan. Journal of the Fisheries Board of Canada 17:517-539. Bollens, S. M., B. W. Frost, D. S. Thoreson, and S. J. Watts. 1992. Diel vertical migration in zooplankton: field evidence in support of the predator avoidance hypothesis. Hydrobiologia 234:33-39. Bollens, S. M., G. Rollwagen-Bollens, J. A. Quenette, and A. B. Bochdansky. 2010. Cascading migrations and implications for vertical fluxes in pelagic ecosystems. Journal of Plankton Research 33:349-355. Bolnick, D. I., R. Svanbäck, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L. Forister. 2003. The ecology of individuals: incidence and implications of individual specialization. The American Naturalist 161:1-28. Bolnick, D. I., P. Amarasekare, M. S. Araújo, R. Bürger, J. M. Levine, M. Novak, V. H. Rudolf, S. J. Schreiber, M. C. Urban, and D. A. Vasseur. 2011. Why intraspecific trait variation matters in community ecology. Trends in Ecology & Evolution 26:183-192. Boscarino, B. T., L. G. Rudstam, J. L. Eillenberger, and R. O’Gorman. 2009. Importance of light, temperature, zooplankton and fish in predicting the nighttime vertical distribution of Mysis diluviana. Aquatic Biology 5:263-279. Boscarino, B. T., L. G. Rudstam, M. A. Minson, and E. E. Freund. 2010. Laboratory- derived light and temperature preferences of juvenile mysid shrimp, Mysis diluviana. Journal of Great Lakes Research 36:699-706. Bowers, J., and H. Vanderploeg. 1982. In situ predatory behavior of Mysis relicta in Lake Michigan. Hydrobiologia 93:121-131.

Bowers, J. A. 1988. Diel vertical migration of the opossum shrimp Mysis relicta in Lake Superior: Observations and sampling from the Johnson-Sea-Link II submersible. Bulletin of Marine Science 43:730-738. Bowers, J. A., W. E. Cooper, and D. J. Hall. 1990. Midwater and epibenthic behaviors of Mysis relicta Loven: Observations from the Johnson-Sea-Link II submersible in Lake Superior and from a remotely operated vehicle in northern Lake Michigan. Journal of Plankton Research 12:1279-1286. Brandt, S. B. 1986. Ontogenetic shifts in habitat, diet, and diel-feeding periodicity of slimy sculpin in Lake Ontario. Transactions of the American Fisheries Society 115:711-715. Branstrator, D. K., G. Cabana, A. Mazumder, and J. B. Rasmussen. 2000. Measuring life‐ history omnivory in the opossum shrimp, Mysis relicta, with stable nitrogen isotopes. Limnology and Oceanography 45:463-467. Brodersen, J., E. Adahl, C. Brönmark, and L. A. Hansson. 2008a. Ecosystem effects of partial fish migration in lakes. Oikos 117:40-46. Brodersen, J., P. A. Nilsson, L.-A. Hansson, C. Skov, and C. Brönmark. 2008b. Condition-dependent individual decision-making determines cyprinid partial migration. Ecology 89:1195-1200. Brodersen, J., A. Nicolle, P. A. Nilsson, C. Skov, C. Brönmark, and L. A. Hansson. 2011. Interplay between temperature, fish partial migration and trophic dynamics. Oikos 120:1838-1846. Brown, R. H. 1998. Abundance and distribution of Mysis relicta in Moosehead Lake, Maine. Fishery Progress Report Series No. 98-2. Main Department of Inland Fisheries and Wildlife. Bunnell, D. B., B. M. Davis, D. M. Warner, M. A. Chriscinske, and E. F. Roseman. 2011. Planktivory in the changing Lake Huron zooplankton community: Bythotrephes consumption exceeds that of Mysis and fish. Freshwater Biology 56:1281-1296. Bunnell, D. B., B. M. Davis, M. A. Chriscinske, K. M. Keeler, and J. G. Mychek-Londer. 2015. Diet shifts by planktivorous and benthivorous fishes in northern Lake Michigan in response to ecosystem changes. Journal of Great Lakes Research 41:161-171. Caldwell, T. J. 2010. The role of the non-native freshwater opossum shrimp Mysis diluviana in the nutrient and food web dynamics of Lake Pend Oreille, Idaho. Doctoral dissertation, University of Idaho.

Caldwell, T. J., F. M. Wilhelm, and A. M. Dux. 2016. A non-native pelagic macroinvertebrate alters population dynamics of herbivorous zooplankton in a large deep lake. Canadian Journal of Fisheries and Aquatic Sciences. 73:832-843 Cartamil, D. P., and C. G. Lowe. 2004. Diel movement patterns of ocean sunfish Mola mola off southern California. Marine Ecology Progress Series 266:245-253. Chapman, B. B., C. Brönmark, J. Å. Nilsson, and L. A. Hansson. 2011. The ecology and evolution of partial migration. Oikos 120:1764-1775. Chapman, B., C. Skov, K. Hulthén, J. Brodersen, P. A. Nilsson, L. A. Hansson, and C. Brönmark. 2012. Partial migration in fishes: definitions, methodologies and taxonomic distribution. Journal of Fish Biology 81:479-499. Chess, D. W., and J. A. Stanford. 1999. Experimental effects of temperature and prey assemblage on growth and lipid accumulation by Mysis relicta Loven. Hydrobiologia 412:155-164. Chipps, S. R., and D. H. Bennett. 2000. Zooplanktivory and nutrient regeneration by invertebrate (Mysis relicta) and vertebrate (Oncorhynchus nerka) planktivores: implications for trophic interactions in oligotrophic lakes. Transactions of the American Fisheries Society 129:569-583. Cohen, J. H., and R. B. Forward. 2009. Zooplankton diel vertical migration - a review of proximate control. Oceanography and Marine Biology: An Annual Review 47:77- 109. Cooper, S. D., and C. R. Goldman. 1980. Opossum shrimp (Mysis relicta) predation on zooplankton. Canadian Journal of Fisheries and Aquatic Sciences 37:909-919. Covich, A. P., M. A. Palmer, and T. A. Crowl. 1999. The role of benthic invertebrate species in freshwater ecosystems: zoobenthic species influence energy flows and nutrient cycling. BioScience 49:119-127. Cullen, J. J. 1985. Diel vertical migration by dinoflagellates: Roles of carbohydrate metabolism and behavioral flexibility. Contributions in Marine Science 27:135- 152. Dadswell, M. J. 1974. Distribution, ecology and postglacial dispersal of certain crustaceans and fishes in eastern North America. National Museum of Natural Sciences Publication of Theses in Zoology 11:1–110. Darimont, C. T., C. P. Paquet, and T. E. Reimchen. 2007. Stable isotopic niche predicts fitness of prey in a wolf–deer system. Biological Journal of the Linnean Society 90:125-137.

Darnis, G., L. Hobbs, M. Geoffroy, J. C. Grenvald, P. E. Renaud, J. Berge, A. Wold. 2017. From polar night to midnight sun: Diel vertical migration, metabolism and biogeochemical role of zooplankton in a high Arctic fjord (Kongsfjorden, Svalbard). Limnology and Oceanography. 62:1586-1605. den Heyer, C., and J. Kalff. 1998. Organic matter mineralization rates in sediments: A within-and among-lake study. Limnology and Oceanography 43:695-705. Egevang, C., I. J. Stenhouse, R. A. Phillips, A. Petersen, J. W. Fox, and J. R. Silk. 2010. Tracking of Arctic terns Sterna paradisaea reveals longest animal migration. Proceedings of the National Academy of Sciences 107:2078-2081. Eshenroder, R. L., and M. K. Burnham-Curtis. 1999. Species succession and sustainability of the Great Lakes fish community. Pages 145-184. in Great Lakes fishery policy and management: A binational perspective. Michigan State University Press. Eshenroder, R. L., and M. K. Burnham-Curtis. 2001. Comment: Vertical migration of bloaters. Transactions of the American Fisheries Society 130:166-167. Euclide, P. T., S. Hansson, and J. D. Stockwell. 2017. Partial diel vertical migration in an omnivorous macroinvertebrate, Mysis diluviana. Hydrobiologia 787:387-396. France, R. L. 2012. Omnivory, vertical food‐web structure and system productivity: stable isotope analysis of freshwater planktonic food webs. Freshwater Biology 57:787-794. Gal, G., L. G. Rudstam, and O. E. Johannsson. 2004. Predicting Mysis relicta vertical distribution in Lake Ontario. Archiv für Hydrobiologie 159:1-23. Gal, G., L. G. Rudstam, E. L. Mills, J. R. Lantry, O. E. Johannsson, and C. H. Greene. 2006. Mysid and fish zooplanktivory in Lake Ontario: quantification of direct and indirect effects. Canadian Journal of Fisheries and Aquatic Sciences 63:2734- 2747. Gamble, A. E., T. R. Hrabik, J. D. Stockwell, and D. L. Yule. 2011a. Trophic connections in Lake Superior Part I: the offshore fish community. Journal of Great Lakes Research 37:541-549. Gamble, A. E., T. R. Hrabik, D. L. Yule, and J. D. Stockwell. 2011b. Trophic connections in Lake Superior Part II: the nearshore fish community. Journal of Great Lakes Research 37:550-560. Gende, S. M., R. T. Edwards, M. F. Willson, and M. S. Wipfli. 2002. Pacific salmon in aquatic and terrestrial ecosystems. BioScience 52:917-928.

GLFC 2015. Basin-wide research priorities - http://www.glfc.org/research/Basin_Wide_Priorities.pdf Gliwicz, M. Z. 1986. Predation and the evolution of vertical migration in zooplankton. Nature 320:746-748. Gliwicz, M. Z., and J. Pijanowska. 1988. Effect of predation and resource depth distribution on vertical migration of zooplankton. Bulletin of Marine Science 43:695-709. Gorokhova, E. 2009. Toxic cyanobacteria Nodularia spumigena in the diet of Baltic mysids: Evidence from molecular diet analysis. Harmful Algae 8:264-272. Grossnickle, N. E. 1979. Nocturnal feeding patterns of Mysis relicta in Lake Michigan, based on gut content fluorescence. Limnology and Oceanography 24:777-780. Grossnickle, N. E. 1982. Feeding habits of Mysis relicta - an overview. Hydrobiologia 93:101-107. Guisande, C., A. Duncan, and W. Lampert. 1991. Trade-offs in Daphnia vertical migration strategies. Oecologia 87:357-359. Gutowski T. 1978. The ecology of Mysis relicta (Loven) in Lake Champlain. MS thesis. University of Vermont. Hairston, N. 1976. Photoprotection by carotenoid pigments in the copepod Diaptomus nevadensis. Proceedings of the National Academy of Sciences 73:971-974. Haupt, F., M. Stockenreiter, M. Baumgartner, M. Boersma, and H. Stibor. 2009. Daphnia diel vertical migration: implications beyond zooplankton. Journal of Plankton Research 5:515-524. Hays, G. C. 2003. A review of the adaptive significance and ecosystem consequences of zooplankton diel vertical migrations. Pages 163-170 in Migrations and Dispersal of Marine Organisms. Springer. Hölker, F., M. J. Vanni, J. J. Kuiper, C. Meile, H.-P. Grossart, P. Stief, R. Adrian, A. Lorke, O. Dellwig, and A. Brand. 2015. Tube‐dwelling invertebrates: tiny ecosystem engineers have large effects in lake ecosystems. Ecological Monographs 85:333-351. Hrycik, A. R., P. W. Simonin, L. G. Rudstam, D. L. Parrish, B. Pientka, and T. B. Mihuc. 2015. Mysis zooplanktivory in Lake Champlain: A bioenergetics analysis. Journal of Great Lakes Research 41:492-501. Isaac, E. J., T. R. Hrabik, J. D. Stockwell, and A. E. Gamble. 2012. Prey selection by the Lake Superior fish community. Journal of Great Lakes Research 38:326-335.

Jensen, O. P., T. R. Hrabik, S. J. Martell, C. J. Walters, and J. F. Kitchell. 2006. Diel vertical migration in the Lake Superior pelagic community. II. Modeling tradeoffs at an intermediate trophic level. Canadian Journal of Fisheries and Aquatic Sciences 63:2296-2307. Jensen, O., S. Hansson, T. Didrikas, J. D. Stockwell, T. R. Hrabik, T. Axenrot, and J. F. Kitchell. 2011. Foraging, bioenergetic and predation constraints on diel vertical migration: field observations and modelling of reverse migration by young‐of‐the‐ year herring Clupea harengus. Journal of Fish Biology 78:449-465. Johannsson, O. E., L. G. Rudstam, and D. C. Lasenby. 1994. Mysis relicta: assessment of metalimnetic feeding and implications for competition with fish in Lakes Ontario and Michigan. Canadian Journal of Fisheries and Aquatic Sciences 51:2591-2602. Johannsson, O. E. 1995. Response of Mysis relicta population dynamics and productivity to spatial and seasonal gradients in Lake Ontario. Canadian Journal of Fisheries and Aquatic Sciences 52:1509-1522. Johannsson, O. E., M. F. Leggett, L. G. Rudstam, M. R. Servos, M. A. Mohammadian, G. Gal, R. M. Dermott, and R. H. Hesslein. 2001. Diet of Mysis relicta in Lake Ontario as revealed by stable isotope and gut content analysis. Canadian Journal of Fisheries and Aquatic Sciences 58:1975-1986. Johannsson, O., L. Rudstam, G. Gal, and E. Mills. 2003. Mysis relicta in Lake Ontario: population dynamics, trophic linkages and further questions. State of Lake Ontario (SOLO) - Past, present and future. Ecovision World Monograph Series:257-287. Johnson, B. M., W. M. Pate, D. B. Silver, and J. L. Sharp. 2018. Invasion success and population characteristics of the opossum shrimp, Mysis diluviana, in Wyoming, USA. Aquatic Invasions 13:409-420. Jónasdóttir, S. H., A. W. Visser, K. Richardson, and M. R. Heath. 2015. Seasonal copepod lipid pump promotes carbon sequestration in the deep North Atlantic. Proceedings of the National Academy of Sciences 112:12122-12126. Jude, D. J., L. G. Rudstam, T. J. Holda, J. M. Watkins, P. T. Euclide, and M. D. Balcer. 2018. Trends in Mysis diluviana abundance in the Great Lakes, 2006- 2016. Journal of Great Lakes Research 44:590-599. Kerfoot, W. C. 1985. Adaptive value of vertical migration: comments on the predation hypothesis and some alternatives. Contributions in Marine Science 27:91-113. Kitchell, J. F., S. P. Cox, C. J. Harvey, T. B. Johnson, D. M. Mason, K. K. Schoen, K. Aydin, C. Bronte, M. Ebener, and M. Hansen. 2000. Sustainability of the Lake Superior fish community: interactions in a food web context. Ecosystems 3:545- 560. 27

Lampert, W. 1993. Ultimate causes of diel vertical migration of zooplankton: new evidence for the predator-avoidance hypothesis. Archiv für Hydrobiologie– Monographische Beiträge 39:79-88. Lasenby, D. C., and R. R. Langford. 1973. Feeding and assimilation of Mysis relicta. Limnology and Oceanography 18:280-285. Lasenby, D., T. Northcote, and M. Fürst. 1986. Theory, practice, and effects of Mysis relicta introductions to North American and Scandinavian lakes. Canadian Journal of Fisheries and Aquatic Sciences 43:1277-1284. Lindén, E., and H. Kuosa. 2004. Effects of grazing and excretion by pelagic mysids (Mysis spp.) on the size structure and biomass of the phytoplankton community. In Biology of the Baltic Sea (pp. 73-78). Springer, Dordrecht. Longhurst, A. R., and W. G. Harrison. 1988. Vertical nitrogen flux from the oceanic photic zone by diel migrant zooplankton and nekton. Deep Sea Research Part A. Oceanographic Research Papers 35:881-889. Loose, C. J., and P. Dawidowicz. 1994. Trade-offs in diel vertical migration by zooplankton: the costs of predator avoidance. Ecology 75: 2255-2263. Madeira, P. T., A. S. Brooks, and D. B. Seale, D. B. 1982. Excretion of total phosphorus, dissolved reactive phosphorus, ammonia, and urea by Lake Michigan Mysis relicta. In Ecology of Mysidacea (pp. 145-154). Springer, Dordrecht. Mangel, M., and C. W. Clark. 1988. Dynamic modeling in behavioral ecology. Princeton University Press. Marsden, J. E., and M. Hauser. 2009. Exotic species in Lake Champlain. Journal of Great Lakes Research 35:250-265. Marsden, J. E., P. Stangel, and A. D. Shambaugh. 2013. Influence of Environmental Factors on Zebra Mussel Population Expansion in Lake Champlain, 1994– 2010. In Quagga and Zebra Mussels: Biology, Impacts, and Control 33:34-49. Martinez, P. J., and E. P. Bergersen. 1989. Proposed biological management of Mysis relicta in Colorado lakes and reservoirs. North American Journal of Fisheries Management 9:1-11. Mauchline, J. 1980. The biology of mysids and euphausiids. Academic Press, London, UK. McLaren, I. A. 1974. Demographic strategy of vertical migration by a marine copepod. American Naturalist:91-102.

McMeans, B. C., K. S. McCann, T. D. Tunney, A. T. Fisk, A. M. Muir, N. Lester, B. Shuter, and N. Rooney. 2016. The adaptive capacity of lake food webs: from individuals to ecosystems. Ecological Monographs 86:4-19. McWilliam, P. S. 1970. Seasonal changes in abundance and reproduction in the opossum shrimp, Mysis relicta Lovén in Lake Michigan. MS thesis. University of Sydney, Australia. Mehner, T., P. Kasprzak, and F. Hölker. 2007. Exploring ultimate hypotheses to predict diel vertical migrations in coregonid fish. Canadian Journal of Fisheries and Aquatic Sciences 64:874-886. Mehner, T., and P. Kasprzak. 2011. Partial diel vertical migrations in pelagic fish. Journal of Animal Ecology 80:761-770. Mehner, T. 2012. Diel vertical migration of freshwater fishes–proximate triggers, ultimate causes and research perspectives. Freshwater Biology 57:1342-1359. Mehner, T. 2014. Partial diel vertical migration of sympatric vendace (Coregonus albula) and Fontane cisco (Coregonus fontanae) is driven by density dependence. Canadian Journal of Fisheries and Aquatic Sciences 72:116-124. Mihuc, T. B., and F. Recknagel. 2018. Lessons from Bioinvasion of Lake Champlain, USA. In Ecological Informatics. Springer. pp. 401-414. Milner-Gulland, E.J., J.M. Fryxell, and A.R.E Sinclair (Eds.). 2011. Animal migration: a synthesis. Oxford: Oxford University Press. Moen, V., and A. Langeland. 1989. Diurnal vertical and seasonal horizontal distribution patterns of Mysis relicta in a large Norwegian lake. Journal of Plankton Research 11:729-745. Morgan, M. D. 1980. Life history characteristics of two introduced populations of Mysis relicta. Ecology:551-561. Murtaugh, P. A. 1984. Variable gut residence time: problems in inferring feeding rate from stomach fullness of a mysid crustacean. Canadian Journal of Fisheries and Aquatic Sciences 41:1287-1293. Myer, G.E., and G. K. Gruendling. 1979. Limnology of Lake Champlain. Lake Champlain Basin Study. New England River Basin Commission, Burlington, VT. Naesje, R., A. Jensen, and O. Sandlund. 2003. Life history, habitat utilisation, and biomass of introduced Mysis relicta. Limnologica-Ecology and Management of Inland Waters 33:244-257.

Nesler, T., and E. Bergersen. 1991. Mysids and their impacts on fisheries: an introduction to the 1988 mysid-fisheries symposium. Pages 1-4 in American Fisheries Society Symposium. Nordin, L. J., M. T. Arts, O. E. Johannsson, and W. D Taylor. 2008. An evaluation of the diet of Mysis relicta using gut contents and fatty acid profiles in lakes with and without the invader Bythotrephes longimanus (Onychopoda, Cercopagidae). Aquatic Ecology 42:421-436. Ogonowski, M., J. Duberg, S. Hansson, and E. Gorokhova. 2013. Behavioral, ecological and genetic differentiation in an open environment—a study of a mysid population in the Baltic Sea. PloS one 8:e57210. Oliver, S. K., D. K. Branstrator, T. R Hrabik, S. J. Guildford, and R. E. Hecky. 2015. Nutrient excretion by crustacean zooplankton in the deep chlorophyll layer of Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 72: 390-399. O'Malley, B. P., and D. B. Bunnell. 2014. Diet of Mysis diluviana reveals seasonal patterns of omnivory and consumption of invasive species in offshore Lake Michigan. Journal of Plankton Research 36:989-1002. O’Malley, B. P., L. G. Rudstam, J. M. Watkins, T. J. Holda, and B. C. Weidel. 2017. Effects of food web changes on Mysis diluviana diet in Lake Ontario. Journal of Great Lakes Research 43:813-822. O'Malley, B. P., S. Hansson, and J. D. Stockwell. 2018. Evidence for a size-structured explanation of partial diel vertical migration in mysids. Journal of Plankton Research 40:66-76. Parker, J. I. 1980. Predation by Mysis relicta on Pontoporeia hoyi: a food chain link of potential importance in the Great Lakes. Journal of Great Lakes Research 6:164- 166. Patwa, Z., R. Christensen, D. C. Lasenby, E. Webster, and D. Mackay. 2007. An exploration of the role of mysids in benthic‐pelagic coupling and biomagnification using a dynamic bioaccumulation model. Environmental Toxicology and Chemistry 26:1224-1232. Pimm, S. L. 1982. Food webs. Springer. Pimm, S. L., J. H. Lawton, and J. E. Cohen. 1991. Food web patterns and their consequences. Nature 350:669-674. Polis, G. A. 1991. Complex trophic interactions in deserts: an empirical critique of food- web theory. The American Naturalist 138:123-155.

Polis, G. A., W. B. Anderson, and R. D. Holt. 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Review of Ecology and Systematics 28: 289-316. Pratt, T. C., O. T. Gorman, W. P. Mattes, J. T. Myers, H. R. Quinlan, D. R. Schreiner, M. J. Seider, S. P. Sitar, D. L. Yule, and P. M. Yurista. 2016. The state of Lake Superior in 2011 [online]. Available from: http://www.glfc.org/pubs/SpecialPubs/Sp16_01.pdf [accessed 21 April 2016]. Ramcharan, C. W., and W. G. Sprules. 1986. Visual predation in Mysis relicta Lovén. Limnology and Oceanography 31:414-420. Reeve, H. K., and P. W. Sherman. 1993. Adaptation and the goals of evolutionary research. Quarterly Review of Biology:1-32. Ricker, K. E. 1959. The origin of two glacial relict crustaceans in North America, as related to Pleistocene glaciation. Canadian Journal of Zoology 37:871-893. Ringelberg, J. 1991. A mechanism of predator-mediated induction of diel vertical migration in Daphnia hyalina. Journal of Plankton Research 13:83-89. Ringelberg, J. 2010. Diel vertical migration of zooplankton in lakes and oceans. Springer. Rogers, M. W., D. B. Bunnell, C. P. Madenjian, and D. M. Warner. 2014. Lake Michigan offshore ecosystem structure and food web changes from 1987 to 2008. Canadian Journal of Fisheries and Aquatic Sciences 71:1072-1086. Rudstam, L. G., K. Danielsson, S. Hansson, and S. Johansson. 1989. Diel vertical migration and feeding patterns of Mysis mixta (Crustacea, Mysidacea) in the Baltic Sea. Marine Biology 101:43-52. Rudstam, L. G., A. L. Hetherington, and A. M. Mohammadian. 1999. Effect of temperature on feeding and survival of Mysis relicta. Journal of Great Lakes Research 25:363-371. Rudstam, L. G., and O. E. Johannsson. 2009. Introduction to advances in the ecology of freshwater mysids. Aquatic Biology 5:246-248. Rybock, J. T. 1978. Mysis relicta Loven in Lake Tahoe: vertical distribution and nocturnal predation. MS thesis. University of Nevada. Shea, M. A., and J. C. Makarewicz. 1989. Production, biomass, and trophic interactions of Mysis relicta in Lake Ontario. Journal of Great Lakes Research 15:223-232. Schindler, D. E., and M. D. Scheuerell. 2002. Habitat coupling in lake ecosystems. Oikos 98:177-189.

Sierszen, M. E., J. R. Kelly, T. D. Corry, J. V. Scharold, and P. M. Yurista. 2011. Benthic and pelagic contributions to Mysis nutrition across Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 68:1051-1063. Sierszen, M. E., T. R. Hrabik, J. D. Stockwell, A. M. Cotter, J. C. Hoffman, and D. L. Yule. 2014. Depth gradients in food‐web processes linking habitats in large lakes: Lake Superior as an exemplar ecosystem. Freshwater Biology 59:2122-2136. Smeltzer, E., A. D. Shambaugh, and P. Stangel. 2011. Environmental change in Lake Champlain revealed by long-term monitoring. Journal of Great Lakes Research 38:6-18. Spencer, C. N., B. R. McClelland, and J. A. Stanford. 1991. Shrimp stocking, salmon collapse, and eagle displacement. BioScience 41:14-21. Stewart, T. J., and W. G. Sprules. 2011. Carbon-based balanced trophic structure and flows in the offshore Lake Ontario food web before (1987–1991) and after (2001– 2005) invasion-induced ecosystem change. Ecological Modelling 222:692-708. Stewart, T. J., and L. G. Rudstam. 2012. Did Mysis have a role in the decline of Diporeia? GLFC Project Completion Report. Stockwell, J. D., T. R. Hrabik, O. P. Jensen, D. L. Yule, and M. Balge. 2010. Empirical evaluation of predator-driven diel vertical migration in Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 67:473-485. Svanbäck, R., and D. I. Bolnick. 2007. Intraspecific competition drives increased resource use diversity within a natural population. Proceedings of the Royal Society of London B: Biological Sciences 274:839-844. Takahashi, K. 2004. Feeding ecology of mysids in freshwater and coastal marine habitats: A review. Bulletin of Plankton Society of Japan 51:46-72. Teraguchi, M. 1969. Diel vertical migration of Mysis relicta (Loven) in Green Lake, Wisconsin. PhD Dissertation. University of Wisconsin, Madison. Van Duyn-Henderson, J. V., and D. C. Lasenby. 1986. Zinc and cadmium transport by the vertically migrating opossum shrimp, Mysis relicta. Canadian Journal of Fisheries and Aquatic Sciences 43:1726-1732. Väinölä, R., B. J. Riddoch, R. D. Ward, and R. I. Jones. 1994. Genetic zoogeography of the Mysis relicta species group (Crustacea: Mysidacea) in northern Europe and North America. Canadian Journal of Fisheries and Aquatic Sciences 51:1490- 1505.

Viherluoto, M., H. Kuosa, J. Flinkman, and M. Viitasalo. 2000. Food utilisation of pelagic mysids, Mysis mixta and M. relicta, during their growing season in the northern Baltic Sea. Marine Biology 136:553-559. Violle, C., B. J. Enquist, B. J. McGill, L. Jiang, C. H. Albert, C. Hulshof, V. Jung, and J. Messier. 2012. The return of the variance: intraspecific variability in community ecology. Trends in Ecology & Evolution 27:244-252. von Elert, E., and G. Pohnert. 2000. Predator specificity of kairomones in diel vertical migration of Daphnia: a chemical approach. Oikos 88:119-128. Walsh, M. G., B. T. Boscarino, J. Marty, and O. E. Johannsson. 2012. Mysis diluviana and Hemimysis anomala: Reviewing the roles of a native and invasive mysid in the Laurentian Great Lakes region. Journal of Great Lakes Research 38:1-6. Waterstraat, A., M. Krappe, P. Riel, and M. Rumpf. 2005. Habitat shifts of Mysis relicta (Decapoda, Mysidacea) in the lakes Breiter and Schmaler Luzin (NE Germany). Crustaceana 78:685-699. Watkins, J. M., L. G. Rudstam, M. J. Connerton, T. Schaner, P. G. Rudstam, and K. L. Bowen. 2015. Abundance and spatial distribution of Mysis diluviana in Lake Ontario in 2008 estimated with 120 kHz hydroacoustic surveys and net tows. Aquatic Ecosystem Health & Management 18:63-75. Werner, E. E., J. F. Gilliam, D. J. Hall, and G. G. Mittelbach. 1983. An experimental test of the effects of predation risk on habitat use in fish. Ecology 64:1540-1548. Whall, J. D., and D. C. Lasenby. 2009. Differences in the trophic role of Mysis diluviana in two intermontane lakes. Aquatic Biology 5:281-292. Williamson, C. E., J. M. Fischer, S. M. Bollens, E. P., Overholt, and J. K. Breckenridge. 2011. Toward a more comprehensive theory of zooplankton diel vertical migration: integrating ultraviolet radiation and water transparency into the biotic paradigm. Limnology and Oceanography, 56:1603-1623. Zaret, T. M., and J. S. Suffern. 1976. Vertical migration in zooplankton as a predator avoidance mechanism. Limnology and Oceanography 21:804-813.

CHAPTER 2: EVIDENCE FOR A SIZE-STRUCTURED EXPLANATION OF

PARTIAL DIEL VERTICAL MIGRATION IN MYSIDS

Brian P. O’Malley1,*, Sture Hansson2, Jason D. Stockwell1

1 Rubenstein Ecosystem Science Laboratory, University of Vermont, 3 College Street,

Burlington, VT 05401, USA.

2Department of Ecology, Environment and Plant Sciences, Stockholm University, SE-106

91, Stockholm, Sweden

*corresponding author: Tel. 1 + (413) 231-4907; Email address: [email protected]

Running head: Mysis partial DVM

Abstract

Mysids are known for benthic-pelagic diel vertical migration (DVM), where the population is benthic by day and pelagic by night. However, historical and recent observations in members of the Mysis relicta complex suggests populations exhibit partial DVM, with some remaining benthic at night. We used pelagic net and benthic sled tows to assess diel habitat use by Mysis diluviana at two stations (60 and 100 m deep) in

Lake Champlain, USA, during June-November 2015. At both stations, mysids were on the bottom both day and night, but the extent of pelagic habitat use by Mysis varied by site depth. At 60-m, pelagic densities were an order of magnitude lower during the day compared to at night, indicative of benthic-pelagic DVM. Contrary to expectations, we found no diel difference between pelagic and benthic sled density estimates at 100-m, suggesting an equal number of Mysis are benthic day and night, and an equal number are pelagic day and night at deeper sites. Mean body length of benthic-caught mysids was greater than pelagic-caught individuals, a pattern that was evident both day and night at

100-m. Our findings indicate Mysis partial DVM is common across seasons and influenced by body size and depth.

Keywords: Mysis diluviana, diel vertical migration, partial migration, body size, Lake

Champlain, depth

Introduction

Diel vertical migration (DVM), the movement of organisms between deep and shallow habitats from day to night, is a widespread behavior in aquatic organisms ranging from plankton to large vertebrates across marine and freshwater settings (e.g. Cullen,

1985; Guisande et al., 1991; Stockwell et al., 2010). Lakes and oceans present natural frameworks for conceptualizing and measuring tradeoffs in migratory behavior due to the strong vertical gradients of abiotic (temperature, light) and biotic variables (prey, predators) (Werner et al., 1983; Boscarino et al., 2009; Vanderploeg et al., 2015). The most common hypothesis for the evolution of DVM in invertebrates is that DVM represents a strategy to minimize an individual’s predation-risk (µ) relative to its growth potential (g) (i.e. µ/g ratio) to optimize energetic gains relative to survival probability

(Zaret and Suffren, 1976; Loose and Dawidowicz, 1994; Hays, 2003).

Support for the predator avoidance hypothesis in pelagic plankton has come from models (Fiksen and Carlotti, 1998; De Robertis, 2002), experiments (Johnsen and

Jakobsen, 1987; Lampert, 2005), and field observations (Ringelberg, 1991; Bollens et al.,

1992). However, DVM behavior can vary within and among populations (Gliwicz, 1986;

Ringelberg, 1991; Ringelberg et al., 1997). For example, species known to exhibit DVM also demonstrate the capacity to undergo reverse migration (i.e. migrating down in the water column at night, Jensen et al., 2011) or partial migration (Mehner, 2012), further highlighting the importance of considering intraspecific variation in unraveling migration decision-making processes. While variability in DVM behavior has been recognized

(Bollens and Frost, 1989; Hays et al., 2001; Mehner and Kazprak, 2011), the assumption

remains that organisms partaking in DVM represent most of the population (Ringelberg,

2010).

Developments in DVM theory for zooplankton have come mainly through studies of holoplankton (Hays, 2003; Pearre, 2003; Lampert, 2005; Cohen and Forward, 2009;

Ringelberg, 2010). Meroplankton, however, also undergo substantial DVM, shifting from benthic to pelagic habitats and consuming both benthic and pelagic resources

(Johannsson et al., 2001; Jumars, 2007; Schmidt et al., 2011; Sierszen et al., 2011). In freshwater lakes, meroplankton in the Mysis relicta complex exhibit some of the most extensive DVM, migrating hundreds of meters in a single night (Beeton, 1960;

Ahrenstroff et al., 2011). If Mysis follow classic DVM theory, then the population should migrate into the water column at night to secure necessary resources for growth while minimizing predation risk from visual predators. Consistent with theory, one assumption for why Mysis perform DVM is that Mysis require high-energy zooplankton prey, in comparison to just feeding on presumed low-energy benthic resources. However, benthic resources can account for a substantial portion of Mysis’ diet (Johannsson et al., 2001;

Sierszen et al., 2011), which challenges the conventional template of a “one-way” vertical gradient in food potential used in DVM theory (Lampert, 1989, 1993;

Ringelberg, 2010; Mehner, 2012). The benthic environment has been largely ignored when addressing µ/g tradeoffs hypothesized to dictate Mysis DVM behavior, likely because of inherent assumptions with the predator avoidance hypothesis for DVM and vertical structure of food in the pelagic environment (Boscarino et al., 2009; Ahrenstoff et al., 2011). Equally plausible, yet untested, is the hypothesis that Mysis prefer to remain benthic and only migrate into the water column when required.

Previous studies on diel benthic-pelagic habitat use by Mysis suggest they exhibit partial DVM, whereby portions of the population occur in both habitats at night (Bowers,

1988; Moen and Langeland, 1989; Rudstam et al., 1989). Existing hypotheses to explain partial migrations in invertebrates, however, have not yet been widely studied (Chapman et al., 2011). Insights on partial DVM in other aquatic species provide a starting point to test for mechanisms in Mysis partial DVM. Variability in DVM within a population could be related to individual body condition (Hays et al., 2001). The fasting endurance hypothesis predicts that smaller-sized individuals deplete resource stores faster, and therefore are more likely to migrate than larger individuals that use resources more efficiently (Chapman et al., 2011). Size-dependent predation vulnerability could also explain potential size differences in migrants and non-migrants (Hays, 1995; Hansson and Hylander, 2009; Skov et al., 2011). Larger individuals in a population remain in habitats with low risk and low growth potential compared to smaller individuals (Hays,

1995). Under the size predation hypothesis, larger Mysis are more likely to be found on or near the lake bottom at night to reduce risk of predation compared to smaller Mysis.

Mysis are also cannibalistic (Nordin et al., 2008) and juveniles can detect and avoid chemical cues of larger mysids (Quirt and Lasenby, 2002). Therefore, an alternative hypothesis is that DVM is largely a juvenile tactic to avoid cannibalism by benthic adults.

In Lake Champlain, USA, non-migrant mysids (i.e. benthic-caught at night) collected in late-autumn were larger on average than migrants (Euclide et al., 2017). Because seasonal dynamics of pelagic production, light levels, and thermal stratification could create conditions where migration is more or less profitable for some size classes at

different times of year, length data on migrant and non-migrant Mysis from other times of the year are needed to test if body size differences are maintained across seasons.

Here we further test the hypothesis that invertebrates exhibit partial DVM by exploring partial DVM of Mysis diluviana in Lake Champlain, USA. We used monthly sampling of benthic and pelagic habitats to identify if Mysis were present on the bottom and in the water column at night across seasons (June - November 2015), and to test for differences in day and night density estimates and demographic data (body size, sex, and life-stage) within and among each habitat type. Because conventional DVM theory predicts that Mysis should migrate from the bottom, where they reside during day, up into the water column at night (Grossnickle and Morgan, 1979), we predicted that day-pelagic densities would be less than night-pelagic densities. Similarly, we expected patterns in day to night benthic densities to be opposite to pelagic densities – higher densities of benthic Mysis during the day than at night. Using individual body size and demographic data, we evaluated if differences existed between migrants and non-migrants across months. We predicted that non-migrants would be larger and consist of a higher proportion of females than migrants across all seasons. Based on Mysis life-history, we predicted differences in mean body size of migrants and non-migrants would decline from spring to autumn, due to the maturation of a mostly juvenile population in spring to an adult dominated population by late-autumn.

Methods

Mysis diluviana was sampled monthly from June to November 2015 in the main basin of Lake Champlain, USA. Collections were made during the day and night from pelagic and benthic habitats at two sites with bottom depths of 60 and 100 m (Figure 2.1). 39

Visual inspection of substrate indicated fine sediment of similar color at both sites. The main basin of Lake Champlain is considered mesotrophic with secchi depth ranging from

4-6 m (Smeltzer et al., 2012; Xu et al., 2015). Day samples were collected at least two hours before sunset, and night samples at least one hour after sunset between 21:00-

01:00. Sampling concluded several hours prior to sunrise. Temperature profiles were collected using a Seabird CTD.

Figure 2.1: Map of sampling sites by bottom depth for Mysis collections in Lake Champlain, USA, in June-November 2015.

For each site-month, three replicate pelagic samples were collected using a 1 m-2 square-frame net (1000-µm mesh, 250-µm cod-end) towed vertically through the water column from 2 m above bottom to the surface, following standard collection procedures for estimating pelagic Mysis density in Lake Champlain and other North American lakes

(Ball et al., 2015; EPA, 2015). Day and night pelagic densities (individuals m-2) for each site-month were estimated as the mean of the three replicates. Only two replicates were collected for pelagic sampling at the 60-m site in July during the day. Day and night benthic densities at each site-month were estimated by the mean of two replicate samples.

Benthic samples were collected using a benthic sled (opening 0.46 m x 0.26 m; 1000-µm mesh, 250-µm cod-end) towed for 5 minutes along bottom at a vessel speed of 0.8 m sec-

1. Upon retrieval, contents of each replicate were rinsed into a separate jar. Start and end coordinates were used to estimate tow distance and vessel speed for benthic transects

(mean distance 0.20 km). Beginning in August, a depth sensor (TDR-MK9, Wildlife

Computers) was attached to the benthic sled to measure contact time with the bottom.

The sensor showed that at vessel speeds of 0.8 m sec-1 the sled remained in contact with the lake bottom for the duration of the benthic tow. Area swept for each benthic sled tow was estimated as the distance traveled multiplied by the sled opening width, and was used to estimate benthic density (individuals m-2). All samples were stored in ethanol for laboratory processing. Nets were not equipped with flowmeters. Due to the large mesh size of each gear and previous observations, we assumed 100% filtration efficiency when calculating densities.

Because the benthic sled was open on deployment and retrieval and could potentially catch pelagic mysids, additional tows were taken to check how many mysids

were sampled in the water column by the sled. Immediately after benthic sample collections in August and September (at night), the benthic sled was deployed using the same amount of cable as standard tows until it reached bottom, then was immediately retrieved. Any mysids collected were assumed to be pelagic-caught (sensu Rudstam et al., 1989) and considered a maximum estimate since some benthic mysids could still be collected just as the sled reached the bottom. When the mean of two reps was subtracted from the total catch of benthic sled samples, the difference suggested that only a small percentage of mysids in benthic samples were captured during descent or ascent (Table

2.1, mean = 5.5%). Because the number of individuals collected was small and likely even lower during day sampling, we did not adjust total catch of benthic samples for amount captured in the water column.

Table 2.1: Number of Mysis collected during descent/retrieval of the benthic sled versus number collected in standard five-minute benthic sled tow, and the percent of pelagic Mysis in benthic tows in Lake Champlain, 2015. Values are the mean of two replicates.

Month Site depth Descent/retrieval 5-min. benthic Percent (m) sled tow August 60 36.5 360.5 10.1% 100 49.0 1107.5 4.4%

September 60 54.0 1144.5 4.7% 100 45.0 1760 2.6%

Mysids were counted and measured from the tip of the rostrum to the tip of the telson at 10X under an Olympus SZX dissecting scope equipped with a calibrated digitizing tablet. Sex was determined on mysids > 10 mm. Mysids < 10 mm were classified as juveniles. Benthic sled catches were occasionally high (exceeding 4,000

individuals), thus length measurements were only taken on a subset of 200 randomly selected individuals per sample.

To test for presence of partial DVM in Lake Champlain, we assessed monthly estimates of Mysis in the water column and on the bottom, both day and night. During the day, more mysids should be found on the bottom as migrants return from nightly foraging. Therefore, we predicted benthic densities from benthic sled tows would be higher during the day than at night. We used paired t-tests to evaluate the hypothesis that day density is greater than night density for sled density estimates. Day-night pelagic sampling should mirror changes observed in the benthos because our pelagic net samples were towed through nearly the entire water column, from 2 m above the bottom to surface. We tested if night-pelagic density was greater than day-pelagic density using paired t-tests.

Mean length of benthic and pelagic individuals collected at night and during the day were compared using paired t-tests across months for each site, both within (e.g. benthic day vs. benthic night) and across habitat types (e.g. benthic day vs. pelagic day).

Because day-pelagic Mysis were rare at the 60-m station, we only compared day-night lengths from our 100-m station. Differences in the proportion of juveniles (< 10 mm), females, and males between night-benthic and night-pelagic samples were assessed using chi-squared tests. For all statistical tests, α = 0.05 was used as a cutoff for significance.

Results

We found evidence of partial DVM by Mysis at both sites in Lake Champlain across all sampled months. Mysids were abundant on the bottom at night each month sampled (Figure 2.2 bottom). Day-night comparisons of benthic sled data revealed 43

different results depending on site depth (Figure 2.2 bottom). Consistent with predictions about Mysis DVM, mean (± SD) benthic sled density during the day at 60 m (18.5 ± 13.5

-2 -2 Mysis m ) was higher than at night (9.7 ± 7.4 Mysis m ; paired t-test: t5 = 2.83, P <

0.05). However, at 100 m, no difference was observed between mean day (19.3 ± 11.9) and mean night (17.6 ± 12.6) benthic sled densities (paired t-test: t5 = 0.43, P > 0.05).

Day-night comparisons of pelagic densities (Figure 2.2 top) yielded similar findings to benthic sled results – no difference at 100 m (paired t-test: t5 = -0.92, P > 0.05) between pelagic-day (70.5 ± 44.3) and pelagic-night (105.0 ± 91.5), but a significant difference at

60 m (paired t-test: t5 = -7.35, P < 0.01) with an order of magnitude more pelagic Mysis detected at night (67.8 ± 23.6) than during the day (5.0 ± 5.7).

Figure 2.2: Monthly diel comparison of mean pelagic density of Mysis (top) estimated from vertical net tows from 2 m above lake bottom to surface tows in Lake Champlain 2015, and mean benthic Mysis density (bottom) estimated from benthic sled tows at 60- (left) and 100-m site sites (right). Open bars are day samples and filled bars are night samples (+ SD). 44

Mean body length of Mysis differed between benthic and pelagic individuals at night (Figure 2.3). At the 60-m site (Figure 2.3 Left), mean (± SD) length of benthic- caught Mysis (13.4 ± 0.9 mm) was greater than pelagic-caught individuals (10.2 ± 0.7 mm; paired t-test t5 = 6.24, P < 0.01). Similarly, at the 100-m site (Figure 2.3 Mid), mean length of benthic-caught individuals (15.8 ± 0.6 mm) was greater than pelagic-caught at night (10.7 ± 2.0 mm; t5 = 5.09, P < 0.01). Overall, the difference between mean length of night-benthic and night-pelagic individuals decreased as a function of sampling date (P

< 0.01, R2 = 0.54) from a difference of 6.4 mm in June to 1.4 mm in November

(Supplemental Figure 2.6). Pooling lengths across all months, size-frequency distributions were skewed towards smaller individuals in night-pelagic samples (Figure

2.4). More than 60% to 80% of Mysis encountered in night-benthic sampling, at 60 and

100-m respectively, were individuals > 12 mm in length (Figure 2.4).

Mean body length also differed between benthic and pelagic individuals during the day at the 100-m site (Figure 2.3 Right). Consistent with our observations at night, mean length of day-benthic individuals (15.6 ± 0.8 mm) was greater than day-pelagic

(10.1 ± 3.1 mm; t5 = 3.59, P < 0.05). Mean length of pelagic-caught mysids was similar between day and night collections (day= 10.1 ± 3.1 mm; night = 10.7 ± 2.0 mm; t5 = -

0.75, P > 0.05). Similarly, we found no difference in mean length of benthic-caught mysids between day and night at 100-m (day = 15.6 ± 0.8 mm; night = 15.8 ± 0.6 mm; t5

= -0.75, P > 0.05). Because too few individuals were collected in day-pelagic samples at

60-m, mean body length from day to night was only compared for benthic sled samples from this site. Mean body length in benthic sled samples at 60-m was lower during the

day compared to at night (day = 12.5 ± 1.2 mm; night = 13.4 ± 0.9 mm; t5 = -3.35, P <

0.05).

Figure 2.3: Mean (± SD) body length (mm) of pelagic (grey circles) and benthic-caught (black circles) Mysis at 60- and 100-m sites in Lake Champlain 2015 as a function of month sampled. Note that mesh size (1 mm) was consistent between the two gear types. Insufficient numbers of Mysis were captured in the pelagic habitat during the day at the 60-m site to compare mean length of day-benthic to day-pelagic individuals.

Figure 2.4: Frequency of Mysis by size class pooled across all months (June-November 2015) in night-pelagic and night-benthic samples at 60- (top) and 100-m (bottom) sites in Lake Champlain.

Chi-squared analyses revealed significant differences in life-stage composition

(juveniles, females, males) between night-benthic and night-pelagic samples for each month pooled across both sites (Table 2.2). Night-pelagic samples comprised mostly juvenile mysids (mean across sites = 49.9 ± 16.0 %; Fig. 5), followed by mature females

(29.9 ± 8.4%) and mature males (20.2 ± 12.3%). Night-benthic sled samples comprised mostly mature female mysids (63.8 ± 13.1%), followed by mature males (26.9 ± 11.4%) and juveniles (9.4 ± 7.0%; Fig. 5). The proportion of mature males in pelagic samples

increased from June to November at 100-m, but at 60-m the difference in males from

June to November was much smaller (Figure 2.5).

Figure 2.5: Percent composition of juvenile, female, and male Mysis collected in night- pelagic (top) and night-benthic samples (bottom) at 60- (left) and 100-m (right) sites in Lake Champlain.

Table 2.2: Number of juvenile, female, male, and total mysids counted from night- benthic and night-pelagic samples by month pooled across both sites, and P-values of chi- squared tests comparing demographic proportions.

Month Habitat Juveniles Males Females M: F Total χ2 p-value June Benthic 47 89 664 0.13 800 Pelagic 339 47 73 0.64 459 < 0.001 July Benthic 104 123 573 0.21 800 Pelagic 159 41 106 0.39 306 < 0.001 August Benthic 48 246 506 0.49 800 Pelagic 606 213 307 0.69 1126 < 0.001 September Benthic 44 293 463 0.63 800 Pelagic 162 65 94 0.69 321 < 0.001 October Benthic 84 275 441 0.62 800 Pelagic 168 68 142 0.48 378 < 0.001 November Benthic 123 264 413 0.64 800 Pelagic 187 314 200 1.57 701 < 0.001

Discussion

Overall, we found consistent evidence of partial DVM by Mysis across all sampling dates. Mysis occurred in the water column and on the bottom, each night, which agrees with similar studies on the vertical distribution of Mysis spp. in other systems

(Moen and Langeland 1989, Rudstam et al. 1989). Our results suggest that Mysis were not only present on the bottom at night, but that benthic individuals were also larger on average than pelagic individuals, suggesting that variability in Mysis DVM behavior can at least be partly explained by body size. The difference in average length between benthic and pelagic Mysis at night agreed with previous findings in Lake Champlain

(Euclide et al. 2017). In addition, our seasonal sampling revealed the difference in mean

length between benthic and pelagic Mysis at night decreased from June to November, and mean body size of pelagic and benthic Mysis did not change from day to night within a month. Differences in male to female ratios between migrants and non-migrants were evident, with proportionally more females found in night-benthic samples than night- pelagic samples. Length and life-stage differences are consistent with several hypotheses for partial DVM (Chapman et al., 2011), but difficult to tease apart given the current state of knowledge.

The difference in mean body size observed between benthic and pelagic individuals at night became less pronounced as the season progressed, likely reflecting the maturation of spring recruits in pelagic samples to adult stages. Mysis typically undergo breeding in late autumn through winter, with spring and summer cohorts produced annually (Mauchline, 1980). In Lake Champlain, previous Mysis studies using pelagic-day and pelagic-night sampling found distinct juvenile cohorts in spring and summer, transitioning to mostly large individuals in autumn (Ball et al., 2015; Hrycik et al., 2015). Our day and night pelagic sampling revealed a similar life history progression, with high proportion of juveniles in spring-summer months. In contrast, our benthic sampling routinely showed high proportions of adult size classes were present even during summer months.

The mean length of benthic Mysis was consistently greater at our deep site compared to the shallow site. Similar increases in Mysis size with bathymetric depth have been reported from the Great Lakes (Johannsson, 1995; Pothoven et al., 2000). Variation in body size as a function of bottom depth could be related to horizontal migration or differences in growth and maturity. Horizontal migration of Mysis has been reported from

other systems, although interpretation of such patterns seems to vary by system. For example, Moen and Langeland (1989) reported Mysis horizontal distributions follow a seasonal pattern with large Mysis occupying shallow waters in spring, likely to release their young in these shallow waters, then move back to deeper waters. In contrast, horizontal migration reported by Morgan and Threkheld (1982) indicated that large Mysis remained deep year-round, and only small Mysis disperse into shallow depth zones. In total, seasonal differences in body length with bathymetric depth in Lake Champlain were apparent in our sampling and suggest that larger Mysis remain in deep waters throughout the year while smaller sizes were found at our shallower site.

Body size differences between benthic and pelagic individuals at night could relate to multiple hypotheses for partial migration behavior in populations. Several studies have found evidence for size structured vertical distribution of pelagic Mysis with smaller individuals higher up in the water column (Boscarino et al., 2009; Ogonowski et al., 2013). Based on our size-frequency data, we found that the size-structured vertical distribution of mysids continues from pelagic all the way to benthic habitat as well, with larger individuals occurring on bottom. Experimental work suggests juvenile Mysis can tolerate higher temperatures and that adult mysids tend to remain below 6°C during DVM with increased mortality at temperatures > 12°C (Rudstam et al., 1999; Boscarino et al.,

2009, 2010). If temperature preferences of Lake Ontario Mysis are similar in Lake

Champlain populations, then adult Mysis in Lake Champlain could potentially be seeking benthic habitat as a thermal refuge during autumn months when hypolimnetic temperatures exceed 6°C (Supplemental Figure 2.7). Thermal refuge, however, would not

explain size-dependent partial DVM in other seasons when hypolimnetic temperatures are cooler and large individuals were still common on bottom.

Resource and habitat preferences of individuals can shift with ontogeny from maximizing somatic growth when young to maximizing reproductive output as adults.

2008) suggesting growth advantages exist for pelagic juveniles. Mysis may feed more efficiently on plankton prey by migrating into sub-thermocline habitat where light levels, temperature, and prey densities are higher than near the lake bottom. However, lipid accumulation rates are greater at cooler water temperatures (Chess and Stanford, 1999) suggesting alternative growth and survival advantages associated with remaining at depth if adequate food supply (detritus and benthic invertebrates; Parker 1980) is available, especially for adult size classes. Moreover, building sufficient lipid reserves is often critical for winter survival and increased reproductive success (Hagen et al., 1996), therefore remaining benthic could be an energetic strategy for adults to increase reproductive success rather than to increase somatic growth. To evaluate growth-related hypotheses, a comparison of growth rates by different size classes on strict diets of pelagic or benthic food resources is needed to fully explore Mysis life-history tradeoffs in resource use.

Our data suggest that during day, the proportion of benthic and pelagic individuals varies by site depth. At our shallow site, a high degree of benthic-pelagic

DVM was evident – that is, pelagic Mysis were rare during day compared to night.

However, at our deeper site, we did not find a difference in day to night pelagic density suggesting that a similar number of Mysis were pelagic regardless of the time of day.

Such depth-specific differences in DVM patterns could have important consequences for understanding Mysis DVM – the benthic-pelagic function of Mysis is misrepresented if a substantial amount remains pelagic during the day. Previous work at 100-m in Lake

Champlain detected Mysis as far as 40 m off bottom during the day, with no apparent differences in day-night pelagic densities (Ball et al. 2015; Hrycik et al. 2015). At extremely deep sites, benthic-pelagic DVM might not be energetically feasible for Mysis if distance from the thermocline to bottom exceeds some energetic threshold (Morgan et al., 1978). Instead, migrating from deep to shallow pelagic habitat might be advantageous for some individuals, while remaining benthic is better for others. Therefore, at deep sites, a working hypothesis for no difference between day-night pelagic density is that pelagic Mysis during day are the same group of pelagic individuals observed at night.

Likewise, no day-night difference in benthic Mysis density suggests potential for a benthic population, mainly composed of adult stages that likely use benthic food resources to a much greater extent than pelagic Mysis. Previous work that identified differences in C: N ratios between benthic and pelagic Mysis at night (Euclide et al.,

2017) supports our working hypothesis. Future studies that test if day-suspending Mysis are the same pelagic individuals at night could fill in key assumptions about the benthic- pelagic nutrient flux by Mysis DVM.

Our sampling method used different gears, which is not necessarily ideal for comparing across habitat types that vary in structure and dimensions. Several previous

studies have reported biases depending on gear types or dimensions (net width, mesh, tow speed), making comparisons across gear types difficult (De Bernardi, 1984; Pepin and Shears, 1997). Repeated sampling with the same gear type is useful for diel comparisons. Thus, to control for net or habitat effects, we restricted our day-night comparison to within gear types. In comparing benthic-pelagic net-based estimates, we controlled for net mesh effects by using the same mesh size for both the benthic sled and pelagic net. The net opening of the benthic sled (0.12 m2), however, was smaller compared to the 1 m2 pelagic net opening. If nets were size-selective based on their mouth area then we would have expected to see a higher proportion of larger individuals in our pelagic net, which was not the case. Studies on benthic/pelagic gear comparisons are rare, but a recent comparison between large and small diameter nets used to sample pelagic Mysis revealed no difference in catch rates or composition (Silver et al., 2016).

Measuring the efficiency with which each gear type samples its respective habitat would greatly improve our ability to estimate the percent on bottom, ultimately leading to refined theories on partial migration behavior and more holistic population assessments.

Mysid abundance is monitored in many lakes using pelagic sampling at night.

Recent pelagic surveys estimated Mysis accounts for approximately 30% of pelagic crustacean biomass in offshore Lake Ontario (Watkins et al. 2015), and in Lake Superior, a population of 9.9 trillion individuals was estimated (Pratt et al. 2016). Both the Lake

Ontario and Superior estimates are based on night-pelagic sampling of the population during DVM, and could be conservative estimates if migration is only part of the picture as we found in Lake Champlain. If even a small proportion of a Mysis population does not migrate at night, that proportion still represents a substantial biomass component in

many lakes as evidenced by their importance in food web models as both predators of plankton and a key prey to most fish species (Kitchell et al., 2000; Stewart and Sprules,

2011; Isaac et al., 2012). Further evaluation of the factors predicting the proportion of

Mysis populations remaining benthic at night, when pelagic surveys are performed, will allow researchers to better account for sampling biases associated with pelagic methods that miss benthic individuals.

Conclusions

In summary, our results represent a step towards evaluating Mysis DVM in the context of partial migration by including benthic distributions with the oft-studied pelagic distributions. Our findings suggest Mysis undergo an ontogenetic shift in DVM behavior from using the pelagic environment as juveniles to a greater reliance on benthic habitat with increasing size. Further, our results suggest depth-specific differences in DVM behavior should be considered when constructing conceptual frameworks of Mysis DVM.

We suggest that bottom depth and other factors such as water transparency and light, interact to influence the extent to which Mysis populations exhibit complete benthic- pelagic DVM. Improvements in benthic and pelagic sampling methods that incorporate a vertically stratified sampling approach, and simultaneous sampling of benthic and pelagic habitats, will ultimately lead to greater insights regarding Mysis DVM behavior and improved estimates of population size and biomass.

Acknowledgements

We thank Captain Steve Cluett and Deckhand Krista Hoffsis for their technical expertise in field sampling aboard the R/V Melosira, and Jessie Gordon, Kevin Melman, 55

and Pascal Wilkins for help with sample processing. This project was supported by the

Great Lakes Fishery Commission with the assistance of Senator Patrick Leahy.

References

Ahrenstorff, T. D., Hrabik, T. R., Stockwell, J. D., Yule, D. L. and Sass, G. G. (2011) Seasonally dynamic diel vertical migrations of Mysis diluviana, coregonine fishes, and siscowet lake trout in the pelagia of western Lake Superior. Trans. Am. Fish. Soc., 140, 1504-1520.

Ball, S. C., Mihuc, T. B., Myers, L. W. and Stockwell, J. D. (2015) Ten-fold decline in Mysis diluviana in Lake Champlain between 1975 and 2012. J. Great Lakes Res., 41, 502-509.

Beeton, A. M. (1960) The vertical migration of Mysis relicta in Lakes Huron and Michigan. J. Fish. Board Can., 17, 517-539.

Bollens, S. M. and Frost, B. W. (1989) Zooplanktivorous fish and variable diel vertical migration in the marine planktonic copepod Calanus pacificus. Limnol. Oceanogr., 34, 1072-1083.

Bollens, S. M., Frost, B. W., Thoreson, D. S. and Watts, S. J. (1992) Diel vertical migration in zooplankton: field evidence in support of the predator avoidance hypothesis. Hydrobiologia, 234, 33-39.

Boscarino, B. T., Rudstam, L. G., Eillenberger, J. L. and O’Gorman, R. (2009) Importance of light, temperature, zooplankton and fish in predicting the nighttime vertical distribution of Mysis diluviana. Aquat. Biol., 5, 263-279.

Boscarino, B. T., Rudstam, L. G., Minson, M. A. and Freund, E. E. (2010) Laboratory- derived light and temperature preferences of juvenile mysid shrimp, Mysis diluviana. J. Great Lakes Res., 36, 699-706.

Bowers, J. A. (1988) Diel vertical migration of the opossum shrimp Mysis relicta in Lake Superior: Observations and sampling from the Johnson-Sea-Link II submersible. Bull. Mar. Sci., 43, 730-738.

Chapman, B. B., Brönmark, C., Nilsson, J. Å. and Hansson, L. A. (2011) The ecology and evolution of partial migration. Oikos, 120, 1764-1775.

Chess, D. W. and Stanford, J. A. (1999) Experimental effects of temperature and prey assemblage on growth and lipid accumulation by Mysis relicta loven. Hydrobiologia, 412, 155-164.

Cohen, J. H. and Forward, R. B. (2009) Zooplankton diel vertical migration - a review of proximate control. Oceanography and Marine Biology: An Annual Review, 47, 77-109.

Cullen, J. J. (1985) Diel vertical migration by dinoflagellates: Roles of carbohydrate metabolism and behavioral flexibility. Contrib. Mar. Sci, 27, 135-152. 57

De Bernardi, R. (1984) Methods of estimating zooplankton abundance. In A Manual on Methods for the Assessment of Secondary Productivity in Fresh Waters.

De Robertis, A. (2002) Size-dependent visual predation risk and the timing of vertical migration: an optimization model. Limnol. Oceanogr., 47, 925-933.

EPA Great Lakes National Program Office (2015) Standard operating procedure for Mysis sample collection and preservation. LG 409, Revision 01, April 2015, Chicago ,IL, USA [online]. Available from: http://www.epa.gov/sites/production/files/2017-01/documents/sop-for-mysis- sample-collection-and-preservation-201504-8pp.pdf [accessed 10 September 2017].

Euclide, P. T., Hansson, S. and Stockwell, J. D. (2017) Partial diel vertical migration in an omnivorous macroinvertebrate, Mysis diluviana. Hydrobiologia, 787, 387-396.

Fiksen, Ø. and Carlotti, F. (1998) A model of optimal life history and diel vertical migration in Calanus finmarchicus. Sarsia, 83, 129-147.

Gliwicz, M. Z. (1986) Predation and the evolution of vertical migration in zooplankton. Nature, 320, 746-748.

Grossnickle, N. E. and Morgan, M. D. (1979) Density estimates of Mysis relicta in Lake Michigan. J. Fish. Board Can., 36, 694-698.

Guisande, C., Duncan, A. and Lampert, W. (1991) Trade-offs in Daphnia vertical migration strategies. Oecologia, 87, 357-359.

Hagen, W., Van Vleet, E. S. and Kattner, G. (1996) Seasonal lipid storage as overwintering strategy of Antarctic krill. Mar. Ecol. Prog. Ser., 134, 85-89.

Hansson, L.-A. and Hylander, S. (2009) Size-structured risk assessments govern Daphnia migration. Proc. R. Soc. Lond. B: Biol. Sci., 276, 331-336.

Hays, G. C. (1995) Ontogenetic and seasonal variation in the diel vertical migration of the copepods Metridia lucens and Metridia longa. Limnol. Oceanogr., 40, 1461- 1465.

Hays, G., Kennedy, H. and Frost, B. (2001) Individual variability in diel vertical migration of a marine copepod: why some individuals remain at depth when others migrate. Limnol. Oceanogr., 46, 2050-2054.

Hays, G. C. (2003) A review of the adaptive significance and ecosystem consequences of zooplankton diel vertical migrations. Migrations and Dispersal of Marine Organisms. Springer, pp. 163-170.

Hrycik, A. R., Simonin, P. W., Rudstam, L. G., Parrish, D. L., Pientka, B. and Mihuc, T. B. (2015) Mysis zooplanktivory in Lake Champlain: A bioenergetics analysis. J. Great Lakes Res., 41, 492-501.

Isaac, E. J., Hrabik, T. R., Stockwell, J. D. and Gamble, A. E. (2012) Prey selection by the Lake Superior fish community. J. Great Lakes Res., 38, 326-335.

Jensen, O., Hansson, S., Didrikas, T., Stockwell, J., Hrabik, T., Axenrot, T. and Kitchell, J. (2011) Foraging, bioenergetic and predation constraints on diel vertical migration: field observations and modelling of reverse migration by young‐of‐the‐ year herring Clupea harengus. J. Fish Biol., 78, 449-465.

Johannsson, O. E. (1995) Response of Mysis relicta population dynamics and productivity to spatial and seasonal gradients in Lake Ontario. Can. J. Fish. Aquat. Sci., 52, 1509-1522.

Johannsson, O. E., Bowen, K. L., Wood, C. M., Smith, R. W., Chu, C., Rudstam, L. G. and Boscarino, B. (2008) Relating nucleic acid and protein indices to growth in Mysis relicta: ration, cycling temperature, and metabolism. Aquat. Biol., 4, 33-46.

Johannsson, O. E., Leggett, M. F., Rudstam, L. G., Servos, M. R., Mohammadian, M. A., Gal, G., Dermott, R. M. and Hesslein, R. H. (2001) Diet of Mysis relicta in Lake Ontario as revealed by stable isotope and gut content analysis. Can. J. Fish. Aquat. Sci., 58, 1975-1986.

Johnsen, G. H. and Jakobsen, P. J. (1987) The effect of food limitation on vertical migration in Daphnia longispina. Limnol. Oceanogr., 32, 873-880.

Jumars, P. A. (2007) Habitat coupling by mid-latitude, subtidal, marine mysids: import- subsidised omnivores. In: R. Gibson, R. Atkinson and J. Gordon (eds) Oceanography and Marine Biology: An annual review. Vol. 45. Taylor & Francis, Boca Raton, Florida, USA, pp. 89-138.

Kitchell, J. F., Cox, S. P., Harvey, C. J., Johnson, T. B., Mason, D. M., Schoen, K. K., Aydin, K., Bronte, C., Ebener, M. and Hansen, M. (2000) Sustainability of the Lake Superior fish community: interactions in a food web context. Ecosystems, 3, 545-560.

Lampert, W. (1989) The adaptive significance of diel vertical migration of zooplankton. Functional Ecology, 3, 21-27.

Lampert, W. (1993) Ultimate causes of diel vertical migration of zooplankton: new evidence for the predator-avoidance hypothesis. Arch. Hydrobiol. Beih. Ergebn. Limnol., 39, 79-88.

Lampert, W. (2005) Vertical distribution of zooplankton: density dependence and evidence for an ideal free distribution with costs. BMC Biol., 3, 10. 59

Loose, C. J. and Dawidowicz, P. (1994) Trade-offs in diel vertical migration by zooplankton: the costs of predator avoidance. Ecology, 75, 2255-2263.

Mauchline, J. (1980) The biology of mysids and euphausiids. Adv. Mar. Biol., 18, 1-681.

Mehner, T. (2012) Diel vertical migration of freshwater fishes–proximate triggers, ultimate causes and research perspectives. Freshwater Biol., 57, 1342-1359.

Mehner, T. and Kasprzak, P. (2011) Partial diel vertical migrations in pelagic fish. J. Anim. Ecol., 80, 761-770.

Moen, V. and Langeland, A. (1989) Diurnal vertical and seasonal horizontal distribution patterns of Mysis relicta in a large Norwegian lake. J. Plankton Res., 11, 729-745.

Morgan, M. D. and Threlkeld, S. T. (1982) Size dependent horizontal migration of Mysis relicta. Hydrobiologia, 93, 63-68.

Morgan, M. D., Threlkeld, S. T. and Goldman, C. R. (1978) Impact of the introduction of kokanee (Oncorhynchus nerka) and opossum shrimp (Mysis relicta) on a subalpine lake. J. Fish. Board Can., 35, 1572-1579.

Nordin, L., Arts, M., Johannsson, O. and Taylor, W. (2008) An evaluation of the diet of Mysis relicta using gut contents and fatty acid profiles in lakes with and without the invader Bythotrephes longimanus (Onychopoda, Cercopagidae). Aquat. Ecol., 42, 421-436.

Ogonowski, M., Hansson, S. and Duberg, J. (2013) Status and vertical size-distributions of a pelagic mysid community in the northern Baltic proper. Boreal Environ. Res., 18, 1-19.

Parker, J. I. (1980) Predation by Mysis relicta on Pontoporeia hoyi: a food chain link of potential importance in the Great Lakes. J. Great Lakes Res., 6, 164-166.

Pearre, S. (2003) Eat and run? The hunger/satiation hypothesis in vertical migration: history, evidence and consequences. Biological Reviews, 78, 1-79.

Pepin, P. and Shears, T. (1997) Variability and capture efficiency of bongo and Tucker trawl samplers in the collection of ichthyoplankton and other macrozooplankton. Can. J. Fish. Aquat. Sci., 54, 765-773.

Pothoven, S. A., Fahnenstiel, G. L., Vanderploeg, H. A. and Luttenton, M. (2000) Population dynamics of Mysis relicta in southeastern Lake Michigan, 1995–1998. J. Great Lakes Res., 26, 357-365.

Pratt, T.C., Gorman, O.T., Mattes, W.P., Myers, J.T., Quinlan, H.R., Schreiner, D.R., Seider, M.J., Sitar, S.P. et al. (2016) The state of Lake Superior in 2011 [online].

Available from: http://www.glfc.org/pubs/SpecialPubs/Sp16_01.pdf [accessed 21 April 2016].

Quirt, J. and Lasenby, D. (2002) Cannibalism and ontogenetic changes in the response of the freshwater shrimp Mysis relicta to chemical cues from conspecific predators. Can. J. Zool., 80, 1022-1025.

Ramcharan, C. W. and Sprules, W. G. (1986) Visual predation in Mysis relicta Lovén. Limnol. Oceanogr., 31, 414-420.

Ringelberg, J. (1991) A mechanism of predator-mediated induction of diel vertical migration in Daphnia hyalina. J. Plankton Res., 13, 83-89.

Ringelberg, J. (2010) Diel vertical migration of zooplankton in lakes and oceans: causal explanations and adaptive significances. Vol., Springer Science.

Ringelberg, J., Flik, B. J. G., Aaned, D. K. and Van Gool, E. (1997) Amplitude of diel vertical migration (DVM) is a function of ﬁsh biomass, a hypothesis. Arch. Hydrobiol. Beih. Ergebn. Limnol., 49, 71-78.

Rudstam, L., Danielsson, K., Hansson, S. and Johansson, S. (1989) Diel vertical migration and feeding patterns of Mysis mixta (Crustacea, Mysidacea) in the Baltic Sea. Mar. Biol., 101, 43-52.

Rudstam, L. G., Hetherington, A. L. and Mohammadian, A. M. (1999) Effect of temperature on feeding and survival of Mysis relicta. J. Great Lakes Res., 25, 363-371.

Schmidt, K., Atkinson, A., Steigenberger, S., Fielding, S., Lindsay, M., Pond, D. W., Tarling, G. A., Klevjer, T. A., Allen, C. S. and Nicol, S. (2011) Seabed foraging by Antarctic krill: Implications for stock assessment, bentho‐pelagic coupling, and the vertical transfer of iron. Limnol. Oceanogr., 56, 1411-1428.

Sierszen, M. E., Kelly, J. R., Corry, T. D., Scharold, J. V., Yurista, P. M. and Sprules, G. (2011) Benthic and pelagic contributions to Mysis nutrition across Lake Superior. Can. J. Fish. Aquat. Sci., 68, 1051-1063.

Silver, D. B., Johnson, B. M., Pate, W. M., Christianson, K. R., Tipton, J., Sherwood, J., Smith, B. and Hao, Y. (2016) Effect of net size on estimates of abundance, size, age and sex ratio of Mysis diluviana. J. Great Lakes Res., 42, 731-737.

Skov, C., Baktoft, H., Brodersen, J., Brönmark, C., Chapman, B. B., Hansson, L.-A. and Nilsson, P. A. (2011) Sizing up your enemy: individual predation vulnerability predicts migratory probability. Proc. R. Soc. B., 278, 1414-1418.

Smeltzer, E., Shambaugh, A. D. and Stangel, P. (2012) Environmental change in Lake Champlain revealed by long-term monitoring. J. Great Lakes Res., 38, 6-18. 61

Stewart, T. J. and Sprules, W. G. (2011) Carbon-based balanced trophic structure and flows in the offshore Lake Ontario food web before (1987–1991) and after (2001– 2005) invasion-induced ecosystem change. Ecol. Modell., 222, 692-708.

Stockwell, J. D., Hrabik, T. R., Jensen, O. P., Yule, D. L. and Balge, M. (2010) Empirical evaluation of predator-driven diel vertical migration in Lake Superior. Can. J. Fish. Aquat. Sci., 67, 473-485.

Vanderploeg, H. A., Pothoven, S. A., Krueger, D., Mason, D. M., Liebig, J. R., Cavaletto, J. F., Ruberg, S. A., Lang, G. A. and Ptáčníková, R. (2015) Spatial and predatory interactions of visually preying nonindigenous zooplankton and fish in Lake Michigan during midsummer. J. Great Lakes Res., 41, 125-142.

Watkins, J. M., Rudstam, L. G., Connerton, M. J., Schaner, T., Rudstam, P. G. and Bowen, K. L. (2015) Abundance and spatial distribution of Mysis diluviana in Lake Ontario in 2008 estimated with 120 kHz hydroacoustic surveys and net tows. Aquat. Ecosyst. Health Manage., 18, 63-75.

Werner, E. E., Gilliam, J. F., Hall, D. J. and Mittelbach, G. G. (1983) An experimental test of the effects of predation risk on habitat use in fish. Ecology, 64, 1540-1548.

Xu, Y., Schroth, A. W. and Rizzo, D. M. (2015) Developing a 21st Century framework for lake‐specific eutrophication assessment using quantile regression. Limnol. Oceanogr.: Methods, 13, 237-249.

Zaret, T. M. and Suffern, J. S. (1976) Vertical migration in zooplankton as a predator avoidance mechanism. Limnol. Oceanogr., 21, 804-813.

Supplemental Figure 2.6: Difference between mean lengths of benthic and pelagic Mysis collected at night by site (60 m = triangles; 100 m = circles) as a function of day of year. Blue line indicates linear model and grey ribbon represents 95% confidence intervals.

Supplemental Figure 2.7: Water temperature isopleths at 60-m (left) and 100-m (right) where Mysis sampling took place in Lake Champlain.

CHAPTER 3: AN UNDERWATER VIDEO SYSTEM TO ASSESS ABUNDANCE

AND BEHAVIOR OF EPIBENTHIC MYSIS

Brian P. O’Malley1, 4, *, Rebecca A. Dillon1, 5, Robert W. Paddock2, Sture Hansson3, Jason

D. Stockwell1

1 Rubenstein Ecosystem Science Laboratory, University of Vermont, Burlington, VT,

USA

2 School of Freshwater Sciences, University of Wisconsin-Milwaukee, 600 E Greenfield

Avenue, Milwaukee, WI, USA

3 Department of Ecology, Environment and Plant Sciences, Stockholm University, SE-

106 91, Stockholm, Sweden

4 Present address: USGS Great Lakes Science Center, Lake Ontario Biological Station,

Oswego, NY, USA

5 Present address: Aquatic Ecology Laboratory, Department of Evolution, Ecology, and

Organismal Biology, The Ohio State University, 1314 Kinnear Road, Columbus, OH,

USA

*Corresponding author: [email protected]

Abstract

The application of remote video technologies can provide alternative views of in situ behavior and distributions of aquatic organisms that might be missed with traditional net-based techniques. We describe a remote benthic video camera system designed to quantify epibenthic density of the macroinvertebrate Mysis diluviana. We deployed the camera multiple times during the day and night at a 60-m depth site in Lake Champlain and quantified Mysis density from the footage using basic methods and readily available software. Density estimates from the video were on average 43 times higher than concurrent estimates from benthic sled tows, suggesting sleds may be inefficient at sampling mysids. Deployment caused initial scattering of individuals, resulting in low densities immediately after deployment that slowly increased. On some occasions, Mysis densities on video fluctuated greatly over several hours, consistent with organisms that have a patchy distribution on the lake bottom. The camera system provided novel insights on behavior and distribution of Mysis on benthic habitats, demonstrating potential for use as a tool for studying partial diel vertical migration and predator-prey interactions.

Keywords: mysid, macroinvertebrate, benthic density, video camera, benthic sled, DVM

Introduction

Accurate estimates of abundances are crucial in ecological studies, but in practice can be very difficult to obtain. Abundance assessments require knowledge about when and where the target species occurs within an ecosystem to ensure an adequate sampling protocol. Consideration of behavior is especially important for species with spatially and/or temporally variable distributions, e.g., large seasonal fluctuations in biomass (e.g.,

Daphnia spp.; Straile et al. 2012) or those that aggregate in swarms, in particular if aggregations vary over the diel cycle (e.g., Euphausia spp.; Everson and Bone 1986). The methods used to collect data are important to consider because different approaches can provide different results due to gear biases (e.g., Madenjian and Jude 1985; Pepin and

Shears 1997; Masson et al. 2004; Gorbatenko and Dolganova 2007). Comparing alternative methods is often necessary to evaluate which gear will produce the most accurate data or when a combination of different sampling techniques is required to achieve the complete picture (e.g., Stockwell et al. 2010).

In plankton sampling, a range of instruments and methods have been used to estimate densities, with most relying on physical collections with nets (De Bernardi 1984;

Wiebe and Benfield 2003; Brenke 2005). Plankton sampling equipment varies in size, shape, overall design, and towing speed/direction, leading to several potential biases when comparing data derived from different methods (De Bernardi 1984; Wiebe and

Benfield 2003; Wiebe et al. 2015). For example, some plankton nets are more efficient at capturing mobile taxa than pump samplers (Masson et al. 2004). Additionally, habitat structure can impose sampling limitations that lead to knowledge gaps regarding distributions if a habitat type is impossible to sample (creating a black box) or

unknowingly alters sampling efficiency (leading to assumptions). Habitats that are not sampled or are assumed to not affect sampling efficiencies, can hinder our ability to accurately depict the role of species within landscapes.

In recent years, more sophisticated instrumentation has provided new views on plankton ecology and created opportunities to revisit assumptions about distributions and behavior (Davis et al. 1996, 2004; Grosjean et al. 2004; Herman et al. 2004; Yurista et al.

2009). For plankton that undergo diel vertical migration (DVM), the evolution of underwater technologies has advanced understanding of this behavior (Ohman et al.

2013; Bianchi and Mislan 2016). In strictly pelagic species (i.e., holoplankton) that exhibit DVM between shallower and deeper parts of the water column, DVM theory has benefited from instruments that provide high-resolution vertical profiles of organisms

(e.g., laser optical plankton counters, hydroacoustics). Coupling such high-resolution observations with other profiling instruments (e.g., CTD, fluorometers) has provided opportunities to test hypotheses about the mechanisms driving DVM behavior (Boscarino et al. 2009; Cohen and Forward 2009). However, species that migrate between benthic and pelagic habitats (i.e., meroplankton) also exhibit DVM, and present methodological challenges for tracking diel behavior because pelagic equipment cannot sample individuals when they are benthic (Grossnickle and Morgan 1979; Malley and Reynolds

1979; Schmidt et al. 2011). Rather, separate gear types are typically needed to target individuals in either the pelagic or benthic habitats.

Partial migration is a result of individual variability in migratory behavior, where not all individuals in a population migrate (Chapman et al. 2011). Partial DVM has been the term used to describe invertebrates that demonstrate variability in their DVM

behavior within a population (Frost and Bollens 1992; Euclide et al. 2017).

Consequently, population estimates of meroplanktonic species that exhibit partial DVM need to account for both migrant and non-migrant individuals. A benthic sled is often used in conjunction with pelagic sampling gear to sample these populations. Such combined estimates, however, are difficult to obtain and interpret (Moen and Langeland

1989; Rudstam et al. 1989) as benthic sled tows are prone to snagging on the bottom if substrate is uneven or contains obstacles (e.g., cobble, boulders, logs, etc.). This can lead to damaged equipment, safety issues, and reduced repeatability of benthic sled sampling relative to pelagic methods that use nets towed vertically through the water column

(Gregg 1976).

Benthic sled sampling can also involve tradeoffs. Setting the mouth opening close to the sediment surface is typically desirable to catch specimens but runs the risk of filling the sled with sediment; while setting the mouth too high reduces the probability of clogging but may pass over specimens closer to or on the bottom. Similarly, benthic sleds need to be towed at slow speeds to remain in contact with the sediment surface when towed to prevent lifting off the bottom (Bergersen and Maiolie 1981). Furthermore, sleds may scatter individuals directly in the tow path because cables used to tow the sled disturb the area at some distance in front of the sled, leading to questions regarding the efficiency or reproducibility of active benthic gear relative to passive observational techniques for slow swimming animals that may be able to avoid the sled.

We describe a portable, deep-water video system, for use near the lake bottom that is capable of filming epibenthic organisms for up to 24 continuous hours and is deployable from a reasonably small boat. The system provides opportunities to

continuously assess abundances and behavior and study diel differences in partial migratory populations. We applied our video system to Mysis diluviana (hereafter Mysis) in Lake Champlain, where only a single mysid species exists and exhibits partial DVM

(Euclide et al. 2017; O’Malley et al. 2018). The efficacy of measuring benthic Mysis with continuous video was evaluated by comparing density estimates generated from video data to those from concurrent benthic sled tows day and night, and point sampling with a smaller portable camera. We demonstrate how this simple camera system can provide new views of Mysis behavior and distribution at depth, and can be used as a tool in other applications requiring remotely-placed technology. In comparison to previously described techniques (Table 3.1), our method represents a passive rather than active method for estimating density from continuous observation at a fixed site.

Table 3.1: Qualitative tradeoffs with various sampling methods used to estimate benthic mysid abundance.

Gear Type Use in lake Advantages Disadvantages studies Benthic sled Most common Able to sweep large areas along transects Prone to snagging, limited to smooth and collect specimens substrates

Benthic grab Common Able to be deployed by hand, cost effective Ineffective with hard substrates; small sampler (e.g. mouth opening and area sampled reduces Ponar) detection probability for sparse organisms

Manned Rare Can provide detailed observational accounts Requires light for illumination, large submersible of behavior at great depths research vessel, and is typically cost-

prohibitive

ROV Unknown Might provide detailed accounts of behavior Requires umbilical cord; limited depth

Camera equipped Rare Allows for simultaneous comparison Substantial image processing required; light to sled between methods from camera could lead to avoidance

SCUBA Not Common Direct observation yields rapid results Limited to shallow depths and would require light unless during the day at shallow depths

Materials and Methods

Description of system

We designed and built a self-contained, battery-powered video system to record

within 1 m of the lake bottom for at least 24 hours without need for a research vessel to

remain on station. Video camera components were required to withstand depths of at

least 400 m and cold temperatures associated with the hypolimnion (~ 4°C). Additionally,

we intended for the entire unit to be heavy enough to remain fixed at depth during

turbulent conditions without lifting off bottom or creating vibrations in the footage. The

entire unit weighed 56 kg in the air.

The camera system (Figure 3.1) consisted of a black and white fixed focus UWC-

300 HD video camera (Outland Technology, Slidell, LA), a DVR-SD (533 MHz;

Seaviewer Cameras, Tampa, FL) to store recordings, a red LED light for illumination,

and two red lasers positioned 10 cm apart for scale. All components were powered by a

battery pack composed of three 14.8-volt, 20.8-amp lithium-ion batteries connected in

parallel, which allowed for at least 24 hours of continuous recording. The battery pack

and DVR were stored in separate custom-machined aluminum housings, while the light,

lasers, and camera came equipped in their own external housings. All parts were mounted

to a frame made of steel tubing (1.2 m × 1.2 m base; Figure 3.1). Lenses of the camera,

light, and lasers were positioned perpendicular to the lake bottom and 0.5 m from the

bottom of the frame. Upon connecting with the battery, the camera, DVR, light, and

lasers turn on and recording begins and runs until the battery is drained or manually

disconnected after retrieval. A rope attached atop the frame tethered to a surface buoy was used to deploy and retrieve the unit.

Figure 3.1: Benthic camera system during deployment. Labeled components are: (A) housing for battery pack; (B) housing with DVR; (C) red LED light; (D) video camera; (E) red lasers. Base of frame is 1.2 x 1.2 m. (photo credit: Maureen Walsh).

Spectral intensity and behavior experiment

Mysis prefers cold-dark conditions and low light levels, but is not sensitive to red wavelengths (Gal et al. 1999; Boscarino et al. 2009). To assess if the red LED light could 73

attract/repel Mysis, we measured the spectral properties of the lights and laser relative to

Mysis vision, and performed a behavioral experiment. Spectral intensity of the red LED light and lasers was measured using a calibrated spectrophotometer and compared with a spectral sensitivity curve of M. diluviana eyes from Cayuga Lake, New York (formerly

M. relicta; Gal et al. 1999). Live Mysis were caught at night using a benthic sled towed along the bottom nearby the camera site, and stored in a 50 x 30 x 40 cm cooler in a dark

5C walk-in refrigerator for two days. For the experiment, the red light from the video system was positioned over one-half of the cooler at the same height as assembled on the steel frame. A piece of cardboard was placed vertically above the middle of the cooler to block the red light from directly reaching half of the cooler. Another piece of cardboard was placed over the top half of the cooler that did not have the light. Sixty-four Mysis were allowed 10 minutes to adjust to the new conditions. Every 10 minutes the number of

Mysis on the side of the cooler with the red light was counted by two individuals for an hour. The two observations for each 10 min were averaged. A total of 3 hours were observed, at 12:00, 15:00, and 19:00 EST. The number of Mysis on the side without the light was calculated as the total number of Mysis in the cooler (N = 64) minus the average number of Mysis on the light side, and compared to an expected equal distribution using chi-squared analysis.

Field deployments and video analysis

The camera system was deployed once in October 2014 and seven times during

June-November 2015 at a 60-m deep site in Lake Champlain (Figure 3.2; Table 3.2). The

2014 deployment was from a slightly larger steel frame (2.2 × 2.2 m base). The lake bottom at this location is composed of fine silt with dark color. Benthic invertebrate 74

biomass is relatively low at this depth in Lake Champlain, relative to other deep lakes in the region (Knight et al. 2018). On each occasion the camera was deployed during the day and retrieved within 1-3 days, dependent on weather conditions. Upon retrieval, video files were downloaded and backed-up to a portable hard drive. Two additional deployments occurred in 2017 as part of a gear comparison (see Field-based comparison).

Figure 3.2: Map of benthic sled transects, stationary Mysis camera deployments, and locations of point sampling with the GoPro mounted mini-lander in Lake Champlain.

Still images were rendered from videos at 10-min intervals for counting purposes using Adobe Premier Pro v9.0 or Quicktime Player on a Dell Optiplex PC with Dell HD monitor. Images were imported to ImageJ64 (Schneider et al. 2012), and each mysid per image was marked with a dot and then counted. On a subset of images, the horizontal direction of up to 100 individuals was also recorded using the segment tool in ImageJ64 to examine the relative orientation of Mysis. Video was played back 10 seconds before or after the desired screenshot time to help distinguish Mysis. Field of view was determined in ImageJ64 using the known distance of the lasers (10 cm). The area of lake bottom in the field of view was 0.12 m2 for the frame used in 2015, and 0.35 m2 for the 2014 frame.

Counts per 10-min interval were scaled to areal densities (No. m-2) by dividing the number counted by viewing area (0.12 m2 or 0.35 m2). After reviewing the video files, a

“landing effect” was apparent during the first 6-8 hours of video recording where densities were initially low before increasing. Because we always deployed the camera during the day, we ignored observations from deployment leading up to nautical sunset per each deployment in our analysis, therefore allowing adequate time for mysids to redistribute themselves after the landing disturbance. For each deployment in 2015, the mean densities of Mysis at night (between 1-hr post-sunset to 1-hr pre-sunrise; Table 3.2) were compared to mean densities the following day (beginning 1-hr post-sunrise) using a paired t-test. Our 2014 deployment resulted in only one 10-minute interval that occurred the following day because of early retrieval before 24 hours in the field, therefore we did not include this date when comparing mean densities because the mean day-density would only be based on a single observation. All analyses and figures were carried out in

R (R Core Team 2015).

We evaluated if our choice of counting interval (10 min) influenced hourly density estimates by comparing the coefficient of variation generated at different intervals over an hour. For two separate hours of video (one from June-night and one from October-night in 2015) we repeated the counting procedure at 1, 5, 10, 20, and 30 minute intervals, resulting in N = 60 and N = 2 observations hr-1 at 1-min and 30-min intervals, respectively. We plotted the coefficients of variation for each hourly density estimate to visually assess whether variability was substantially higher or lower at 10- minute counting intervals relative to other sampling frequencies.

Table 3.2: Date, time (EDT/EST), and nautical sunset times of benthic video camera deployments at the same 60-m location in Lake Champlain 2014-15. Deployment date/time refers to the time of day when the camera touched-down. Nautical sunset indicates the first time-interval counted, and marks the start of the night time period class.

Year Date/time deployed – retrieved Time from deployment to initial count Nautical sunset (Start count)

2014 20 Oct 13:47 – 21 Oct 09:15 5 h 13 m 20 Oct 19:00

2015 11 Jun 10:02 – 12 Jun 15:19 11 h 28 m 11 Jun 21:30

16 Jul 15:00 – 17 Jul 12:14 6 h 30 m 16 Jul 21:30

04 Aug 13:35 – 05 Aug 11:41 7 h 45 m 04 Aug 21:20

78 12 Aug 14:30 – 13 Aug 15:25 7 h 0 m 12 Aug 21:30

07 Oct 10:18 – 08 Oct 09:52 9 h 2 m 07 Oct 19:20

20 Oct 14:04 – 21 Oct 22:13 4 h 56 m 20 Oct 19:00

10 Nov 08:15 – 11 Nov 20:22 9 h 15 m 10 Nov 17:30

1 Field-based comparisons

Benthic video density estimates were compared to density estimates derived from

concurrent benthic sled samples at the same location (in 2015 only) to assess the relative

efficiency of each gear. For 5 out of 8 camera deployments, we collected mysids during

the day and night with a benthic sled (opening 0.46 m × 0.26 m; 1000-µm mesh, 250-µm

cod-end) towed for 5 minutes along bottom at a vessel speed of approximately 0.8 m s-1.

The bottom of the net opening was 0.15 m from the bottom of the benthic sled skis to

reduce dredging of and subsequent clogging by sediment. Two replicates samples were

collected on each sampling event. Contents were preserved in ethanol and all mysids

were counted per sample. Total counts were converted to areal density estimates (Mysis

m-2) by dividing the number of mysids per sled tow by area swept (distance traveled ×

sled opening width [0.46 m]). These densities were compared to average video-derived

densities from the same 1-hr time window using linear regression and a paired t-test.

Upon reviewing the video and benthic sled data, large discrepancies in density

estimates were apparent between the two. Because benthic sleds are thought to have low

capture efficiency relative to alternative passive methods of observation (e.g, Bergersen

and Maiolie 1981) we designed a lightweight mini-lander to provide point sample

estimates of Mysis density around the stationary Mysis video camera system during

deployment as an additional means of comparison. We considered that Mysis could be

attracted to the camera structure due to potential for reef effects of introducing a structure

into the benthic landscape, therefore we predicted that if the camera system was attracting

Mysis, then we would expect to see higher densities of Mysis adjacent to the camera, and

lower densities further from the camera location. The mini-lander frame was constructed from steel rebar with a smaller footprint (50 x 50 cm) to minimize disturbance and allow for immediate counts upon landing. A GoPro camera and dive flashlight (Sea Dragon

Mini 600 LED SeaLife Inc., Moorestown, NJ) fitted with a red filter were mounted to the mini-lander at similar height as the stationary video camera. Field testing was carried out in July and October 2017, at two times of day during each month (Figure 3.2). First, we deployed the stationary Mysis camera and a concrete mooring placed upwind of the camera to control our location during point sampling. On each occasion the camera and mooring were set 1-2 hours before sunset. Point sampling started at least 12 hours after initial deployment. A line from the bow of the vessel to the mooring was used to position the mini-lander close to the stationary camera. For each point sample, the mini-lander was lowered by hand until the unit reached bottom, and allowed to rest for 20 seconds before being raised approximately 5 meters off bottom. Once up, more bow line was paid out to distance the boat before lowering the mini lander again and repeating the process.

Coordinates and time were recorded each time the mini-lander touched down to calculate distance from the camera lander using the VTrack package in R (Campbell et al. 2012).

In total, the mini lander touched down 63 times at distances of 14-165 m from the stationary camera. We counted the number of Mysis in view of the camera right as the mini-lander touched down. An image of a ruler before deployment was used to scale images and measure the viewing area (range: 0.2 – 0.35 m2) in Image J64. Densities observed on the mini-lander were compared to those from the stationary camera for the same 1 hour interval.

Results

Spectral intensity and behavior experiment

Spectrophotometric measurements of the red LED light and lasers showed different spectral intensity curves with little if any overlap with sensitivity of Mysis eyes

(Figure 3.3). The laser had very narrow wavelength intensity, with a peak at 648 nm, while the LED light had a broader curve with a peak at 631 nm. The spectral sensitivity curve of Mysis eyes, with a peak at 520 nm (Gal et al. 1999), had limited overlap with the system’s red LED and no overlap with the laser. The Mysis eye sensitivity curve intersects the red LED light spectral intensity curve at roughly 1/9th peak intensity, which falls between the 607-608 nm wavelengths (Figure 3.3).

Figure 3.3: Relative spectral intensity of red LED light (dashed line) and lasers (solid line) used on the camera system. Spectral sensitivity curve of Mysis eye (dotted line) from Gal et al. (1999) included for comparison.

Only 6 of the 18 observations in the behavior experiment exceeded 50% (max.

54%) of Mysis in the red-light side of the cooler (Figure 3.4). On average, slightly fewer

Mysis were found on the bright side of the cooler under the red light than under the dark side, but overall, the distribution of individuals was not significantly different from an equal distribution throughout the experiment (χ2 = 14.03, d.f. = 17, P = 0.66).

More counting intervals per hour generally led to lower variability in hourly estimates of Mysis density observed on camera for two separate hours of video (Figure

3.5). Coefficients of variation (CV) did not drastically change between counting at 5, 10, or 20-minute intervals, therefore, 10-min counting intervals served as a sufficient compromise between replication and processing time. One exception to our prediction was in the October file when we observed a surprisingly low CV at 30 min. intervals, though this was likely just by chance.

Figure 3.4: Percentage of Mysis observed under the red light half of cooler in the laboratory experiment. Percentages are the average of two observations every ten minutes during the first hour (circle), third hour (triangle), and seventh hour (square) of experiment.

Figure 3.5: Coefficient of variation for Mysis benthic density as a function of binned time intervals for 1-hour video segment from June (circles) and October (triangles) 2015. For each average, N ranged from 60 for the 1-min interval to 2 for the 30-min interval.

Field deployments and video analysis

Overall, Mysis was observed on video at all times, day and night, though densities within video files fluctuated considerably over the course of a day (Figure 3.6;

Supplemental Figure 3.11). No noticeable vibrations or lifting events were seen in the footage. Maximum densities occurred during night for 2 out of 8 sampling events. On most dates, Mysis densities steadily increased or steadily decreased over periods of several hours. At times, Mysis densities exceeded 2000 m-2. Overall mean density observed during the day was not significantly higher than at night (paired t-test: t6 =

1.885, P = 0.109; Figure 3.7).

Individuals were observed to align themselves parallel to one another on occasion, facing the same direction (Figure 3.8). The alignment appears to be in response to the direction of benthic water currents as mysids apparently face into a current based on our 83

observations of drifting particles in view. Analysis of video footage from July 17th 2015 showed a shift in the orientation angle of mysids on camera over the course of several hours as the current apparently shifted direction (Figure 3.8).

Figure 3.6: Mysis density (No. m-2) estimated from benthic video at 10-minute intervals for seven separate deployments in 2015 (black points) and one deployment in 2014 (blue triangles) at the same 60-m site in Lake Champlain. Grey boxes represent night defined as 1-hr post-sunrise to 1-hr pre-sunrise. Times are in EDT except for November which is EST. Moon phase and weather conditions are shown.

Field-based comparisons

Comparisons of density estimates from different gear types revealed a difference between the video and benthic sled. Mean (± SD) video density (755.0 ± 645.3 m-2) was significantly greater than concurrent mean density from the benthic sled (17.7 ± 14.9; paired t-test: t9 = -3.66, P = 0.005). Video and benthic sled estimates were positively correlated but this relationship was not significant (Figure 3.9; R = 0.574; P = 0.083), likely due to low sample size which reduced the probability of detecting a significant relationship (type II error).

Comparison between the stationary video and our mini-lander point sampling approach provided mixed results. Point sampling at varying distances around the stationary camera did not support our prediction of higher values nearest the camera site if attraction to the structure was real. Distance from the stationary camera to the mini- lander was a poor predictor of the amount of Mysis observed on the mini-lander (Figure

3.10a, P = 0.807, R2 < 0.01). Mean density observed under the stationary camera system over the same time period as point sampling however, was higher than in our point sample estimates when distance from the camera was ignored (Figure 3.10b, paired t-test: t3 = 3.87, P = 0.03). Density from the stationary camera during our testing in 2017 was far lower for both camera methods compared to density estimates observed under the stationary camera for several of the deployments in 2015 (Figure 3.6).

Figure 3.7: Mean Mysis density (No. m-2) estimated from benthic video during the night vs day at the same 60-m site in Lake Champlain 2015.

Figure 3.8: Synchronous change in orientation of Mysis with direction of bottom current on July 17th 2015 over a 6-hour period. (Left) Screenshots of Mysis at 06:00 (top), 10:00 (middle), and 12:00 EDT (bottom). (Right) Circular histograms with measurements of the direction individual mysids were facing in the corresponding image to the left.

Figure 3.9: Mysis mean video density (No. m-2) observed as a function of mean benthic sled density (No. m-2) collected during the same 1-hr time window.

Figure 3.10: (A) Observed Mysis density (No. m-2) by the mini-lander at the time of landing as a function of distance (m) from the stationary benthic Mysis video camera. Circles represent July and triangles refer to October 2017 sampling. (B) Comparison of mean density (± SD) observed on mini-lander (x-axis) to stationary benthic Mysis video camera (y-axis).

Discussion

A non-migrating benthic portion of Mysis that remains on or near the bottom of deep lakes, and thus unavailable to pelagic gear at night when Mysis populations are sampled, could present complications for sampling programs and mass-balance food web models, especially if the non-migrants represent a large proportion of the population biomass or a disproportionate fraction of particular life-stages (e.g., adult females;

Euclide et al. 2017; O’Malley et al. 2018). Our results confirm the presence of Mysis in close contact with the sediment surface throughout the night each month from spring to autumn, confirming that Mysis partial DVM is common in Lake Champlain (Euclide et al. 2017; O’Malley et al. 2018). Our results also suggest that Mysis densities may be much higher on benthic habitats than previously estimated using benthic sleds. Although benthic density estimates from our videos are high, they are within range of areal density estimates reported elsewhere in the literature for M. diluviana, regardless of habitat or sampling method. Maximum reported densities of Mysis collected with pelagic gear range from 1,500-1,950 Mysis m-2 (Reiman and Falter 1981; Ball et al. 2015), but densities of < 1,000 m-2 are much more common in the literature (Lasenby 1991;

Rudstam 2009; Rudstam and Johannsson 2009). Therefore, densities observed on video are on the high end of maximum reported densities for the species, especially considering the camera only sampled mysids near the bottom.

Previous studies that used benthic sled tows compared to pelagic net sampling to estimate Mysis densities have suggested benthic sleds are less efficient at capturing mysids (Grossnickle and Morgan, 1979; Shea and Makarewicz et al. 1989). Furthermore, comparisons of benthic sled tows to other benthic techniques also suggest benthic sleds

may underestimate benthic densities. For example, Bergersen and Maiolie (1981) found that camera-generated density estimates from the front of a benthic sled were 2.4 times higher than estimates based on mysids captured in the benthic sled tows. We reached a similar conclusion, that benthic Mysis are inefficiently sampled via benthic sled, albeit we found a much greater difference – benthic sled densities were on average 43 times lower than densities from our video sampling. We did not find a strong relationship between video and benthic sled density estimates, likely due to limited sample size and high variance. Video also revealed dense aggregations of Mysis at times exceeding 2000 m-2, whereas, the maximum density reported by benthic sled (50 m-2) was over an order of magnitude lower. Expeditions from a manned submersible at bottom depths exceeding

250 m in Lake Superior indicated a substantial number of mysids may remain close to the bottom at night (Bowers 1988). Given the logistical and financial support required to operate ROVs or manned submersibles, simple and remote methods as the camera system in this study are may be essential to gather observational data in deep benthic habitats.

Application of remotely-placed video systems in deep-water habitats is more common in marine than freshwater studies, and the continued development of such methods in freshwater could yield new insights for benthic or benthopelagic fauna. The camera resolution combined with red lighting described in our system is not ideal for species level identification of small organisms such as mysids. However, separation of different mysid species was not an objective when designing this system because in Lake

Champlain and other deep North American lakes only one deep-water mysid species is known to occur. Therefore, use of such a camera system should be planned around pre-

defined study goals with prior knowledge of deep-water taxa, and possibly their shapes and swimming behaviors if recording multiple species is of interest.

Previous work on mysid benthic distributions suggest they exhibit patchy distribution (Mauchline 1971; Bergersen and Maiolie 1981), which is clearly confirmed by the variation in densities observed in our camera footage, sometimes with periods of increasing or decreasing density occurring over several hours. Deployment of the camera resulted in an initial scattering of individuals, which resulted in low densities immediately after landing that gradually increased over time, but did not always level-off at the same maximum density after a standard amount of time. The surprisingly high densities observed at times under our camera led us to evaluate the potential for our camera system to act as a Mysis attractant. Although artificial lighting can attract some underwater organisms (Doherty 1987; Stoner et al. 2008; McConnell et al. 2010), we demonstrated that our choice of red light should not have been an issue for Mysis, an extremely photophobic species in freshwater lakes. Collectively, the wavelengths of the light and lasers, biological properties of Mysis vision, cooler experiment, and fluctuations in densities observed under the stationary camera all suggest the system is not attracting

Mysis. Standard shipboard protocols call for only red light on research vessels when sampling Mysis at night. If the camera was an attractant for Mysis, we expected to see abundance values decrease as the mini-lander was used further away from the stationary camera location, and higher values nearest the stationary camera. Our follow-up work with the mini-lander around the stationary camera did not show any effect on densities with distance from the main stationary camera. Density estimates were higher under the stationary camera but the range of density estimates for these tests was very limited

compared to the range observed when routine monthly sampling was performed in 2015.

Although the mini-lander comparison with the stationary camera was inconclusive, a preponderance of evidence suggests attraction to the camera is not an issue.

In situ observations from our camera system also provided insights into Mysis behavior on the bottom. Previous observations from manned submersibles in the Great

Lakes reported that benthic Mysis appeared to align themselves parallel to one another and face the same direction (Robertson et al. 1968; Bowers et al. 1990). We also observed this behavior from videos, and show that over a course of several hours Mysis shift direction, possibly in response to changes in the direction of oncoming currents.

Orienting to the current could prevent displacement by strong currents. Remaining tight to the sediment surface may be advantageous if feeding conditions at the sediment surface are more profitable than being off-bottom. Regardless of the reason why mysids align themselves, knowledge of such in situ behavior was only possible through direct observation, highlighting the potential for video to provide novel insights on the behavior of deep-water organisms (Bicknell et al. 2016). Our video files also indicated potential to quantify predator-prey interactions in situ. Occasionally slimy sculpin (Cottus cognatus) were seen to strike at individual Mysis or burrow into the sediment likely as a sit-and- wait tactic (Supplemental Figure 3.12). Lab experiments suggest slimy sculpins have a low capture efficiency towards Mysis prey (Hondorp 2006). Our video system offers an alternative field-based approach that could be used to quantify predation metrics (strikes, captures, reaction distance, etc.) that could benefit development of sculpin-Mysis foraging models.

Contrary to classic expectations about DVM behavior that day benthic densities would be greater than night benthic densities, we were unable to detect a difference between day and night densities. Such a result could be an artifact of sampling a relatively small area (0.12 m2), rather than integrating over a broader spatial range which could contribute to high variance in average estimates. As a tradeoff to our limited spatial coverage, night and day density estimates were generated from observations at equal intervals collected throughout the night and day, rather than just a snapshot at midnight and midday. Thus, the high number of observations per day and night are likely still sufficient to allow any apparent difference in diel densities to be revealed, which we did not detect.

Video data are rapidly generated yet automated processing of raw video files into data points is still in early phases of development (Pimm et al. 2015). Manual counting from images is a time-intensive process, especially when numerous individuals occur in the field of view. Our sensitivity analysis suggests counting Mysis at intervals anywhere from 1 to 30 minutes per hour provides similar density estimates. We chose 10-minute intervals as a tradeoff between processing time and number of observations. Automated image recognition software packages and improved image resolution would likely help speed up the process.

Insights gained from field testing the camera system and analyzing associated data will hopefully aid researchers to design or improve upon remote video systems for use in deep lakes. One design element to improve the video camera system is to configure a timer that turns the camera on/off at programmed intervals rather than continuous recording. Such capability would allow the battery to last longer to extend deployment

times from one day to potentially weeks. Recording at coarse time intervals over a longer time-period may be advantageous by allowing researchers to capture day-to-day variability without having to retrieve and recharge the battery. Similarly, wasted battery life during the initial hours after placing the camera on bottom (i.e., the landing effect) could be overcome by programmable start times to delay recording during the landing effect period.

Our experience suggests an opportunity exists for remote video cameras to provide alternative perspectives of animal behavior and population assessments in habitats (particularly in deep lake systems) that are not easily sampled by traditional methods. Remote camera lander systems similar to ours have been used as early as the

1960s in marine systems (Ewing et al. 1967), yet over 50 years since their inception, such systems are only starting to be adopted for remote use in deep lakes. In large lakes research, camera technologies are not widely incorporated in monitoring programs

(Twiss and Stryszowska 2016) and a recent study in Lake Baikal using underwater video documented for the first time that the pelagic amphipod Macrohectopus branickii occurs on substrate where previously it was only considered to be pelagic (Karnaukhov et al.

2016), similar to our novel findings for Mysis in Lake Champlain. In the Laurentian Great

Lakes where invasive dreissenids now dominate benthic invertebrate biomass, camera equipped sleds have recently been incorporated to assess dreissenid patchiness and cover more area than traditional benthic grab surveys (Karateyev et al. 2018). Thus, despite similar research goals across freshwater and marine systems, the freshwater research community seems to be lagging in application of video technologies to investigate behavior and distributions of benthic invertebrates compared to marine research

communities, and will undoubtedly benefit from development and application of deep- water video technology.

Acknowledgements

We thank captain Steve Cluett R/V Melosira for troubleshooting electrical issues and Mysis sampling, and Ellis Loew for providing spectral intensity measurements of the light and lasers. Krista Hoffsis, Kevin Melman, Brady Roy, and Pascal Wilkins and provided helpful assistance with deploying/retrieving the video system. Greg Barske and

Randy Metzger (UW-Milwaukee School of Freshwater Sciences Instrument Shop) fabricated the video system housings, Dylan Olson and John Janssen (UW-Milwaukee

SFS) helped with assembly and testing, and Tom Manley (Middlebury College) provided the initial steel frame. Discussions with Lars Rudstam (Cornell University) helped to spark the initial idea for the camera system. This project was supported by the Great

Lakes Fishery Commission with the assistance of Senator Patrick Leahy’s office, the

University of Vermont’s Office of Undergraduate Research, and the Kate Svitek

Memorial Foundation to RAD.

References

Ball, S. C., T. B. Mihuc, L. W. Myers, and J. D. Stockwell. 2015. Ten-fold decline in Mysis diluviana in Lake Champlain between 1975 and 2012. Journal of Great Lakes Research 41: 502-509.

Bergersen, E. P., and M. Maiolie. 1981. Mysis population estimates: a photographic technique. Applied Sciences Branch, Division of Research, Engineering and Research Center, US Department of the Interior, Bureau of Reclamation.

Bianchi, D., and K. Mislan. 2016. Global patterns of diel vertical migration times and velocities from acoustic data. Limnology and Oceanography 61: 353-364.

Bicknell, A. W., B. J. Godley, E. V. Sheehan, S. C. Votier, and M. J. Witt. 2016. Camera technology for monitoring marine biodiversity and human impact. Frontiers in Ecology and the Environment 14: 424-432.

Boscarino, B. T., L. G. Rudstam, J. L. Eillenberger, and R. O’Gorman. 2009. Importance of light, temperature, zooplankton and fish in predicting the nighttime vertical distribution of Mysis diluviana. Aquatic Biology 5: 263-279.

Bowers, J. A., W. E. Cooper, and D. J. Hall. 1990. Midwater and epibenthic behaviors of Mysis relicta Loven: Observations from the Johnson-Sea-Link II submersible in Lake Superior and from a remotely operated vehicle in northern Lake Michigan. Journal of Plankton Research 12: 1279-1286.

Brenke, N. 2005. An epibenthic sledge for operations on marine soft bottom and bedrock. Marine Technology Society Journal 39: 10-21.

Campbell, H. A., Watts, M. E., Dwyer, R. G., Franklin, C. E. 2012. V-Track: software for analysing and visualising animal movement from acoustic telemetry detections. Marine and Freshwater Research 63: 815-820

Chapman, B. B., C. Brönmark, J. Å. Nilsson, and L. A. Hansson. 2011. The ecology and evolution of partial migration. Oikos 120: 1764-1775.

Cohen, J. H., and R. B. Forward. 2009. Zooplankton diel vertical migration - a review of proximate control. Oceanography and Marine Biology: An Annual Review 47: 77-109.

Davis, C. S., S. M. Gallager, M. Marra, and W. K. Stewart. 1996. Rapid visualization of plankton abundance and taxonomic composition using the Video Plankton Recorder. Deep Sea Research Part II: Topical Studies in Oceanography 43: 1947- 1970.

Davis, C. S., Q. Hu, S. M. Gallager, X. Tang, and C. J. Ashjian. 2004. Real-time observation of taxa-specific plankton distributions: an optical sampling method. Marine Ecology Progress Series 284: 77-96.

De Bernardi, R. 1984. Methods of estimating zooplankton abundance. In J. A. Downing and F. H. Rigler [eds.], A Manual on Methods for the Assessment of Secondary Productivity in Fresh Waters. Blackwell Scientific Publications.

Doherty, P. J. 1987. Light-traps: selective but useful devices for quantifying the distributions and abundances of larval fishes. Bulletin of marine science 41:423- 431.

Euclide, P. T., S. Hansson, and J. D. Stockwell. 2017. Partial diel vertical migration in an omnivorous macroinvertebrate, Mysis diluviana. Hydrobiologia 787: 387-396.

Everson, I., and D. Bone. 1986. Effectiveness of the RMT8 system for sampling krill (Euphausia superba) swarms. Polar Biology 6: 83-90.

Ewing, M., J. L. Worzel, and A. C. Vine. 1967. Early development of ocean-bottom photography at Woods Hole Oceanographic Institution and Lamont Geological Observatory. The John Hopkins Oceanographic Studies.

Frost, B. W., and S. M. Bollens. 1992. Variability of diel vertical migration in the marine planktonic copepod Pseudocalanus newmani in relation to its predators. Canadian Journal of Fisheries and Aquatic Sciences. 49: 1137-1141.

Gal, G., E. R. Loew, L. G. Rudstam, and A. M. Mohammadian. 1999. Light and diel vertical migration: spectral sensitivity and light avoidance by Mysis relicta. Canadian Journal of Fisheries and Aquatic Sciences 56: 311-322.

Gorbatenko, K., and N. Dolganova. 2007. Comparing the catch efficiency with different types of plankton nets in the high production zones of the Pacific Ocean. Oceanology 47: 205-212. 97

Gregg, R. E. 1976. The ecology of Mysis relicta in Twin Lakes, Colorado. MS thesis. Colorado State University.

Grossnickle, N. E., and M. D. Morgan. 1979. Density estimates of Mysis relicta in Lake Michigan. Journal of the Fisheries Board of Canada 36: 694-698.

Herman, A., B. Beanlands, and E. Phillips. 2004. The next generation of optical plankton counter: the laser-OPC. Journal of Plankton Research 26:1135-1145.

Hondorp, D. W. 2006. Factors influencing diet and prey selection of the sculpin Cottus cognatus and Myoxocephalus thompsonii. PhD Dissertation, University of Michigan, Ann Arbor, MI.

Karatayev, A. Y., K. Mehler, L. E. Burlakova, E. K. Hinchey, and G. J. Warren. 2018. Benthic video image analysis facilitates monitoring of Dreissena populations across spatial scales. Journal of Great Lakes Research 44:629-638.

Karnaukhov, D. Y. and others 2016. Behaviour of Lake Baikal amphipods as a part of the night migratory complex in the Kluevka settlement region (south-eastern Baikal). Crustaceana 89: 419-430.

Knight, J. C., B. P. O'Malley, and J. D. Stockwell. 2018. Lake Champlain offshore benthic invertebrate community before and after zebra mussel invasion. Journal of Great Lakes Research 44:283-288.

Lasenby, D. C. 1991. Comments on the roles of native and introduced Mysis relicta in aquatic ecosystems, p. 17-22. American Fisheries Society Symposium.

Madenjian, C., and D. J. Jude. 1985. Comparison of sleds versus plankton nets for sampling fish larvae and eggs. Hydrobiologia 124: 275-281.

Malley, D., and J. Reynolds. 1979. Sampling strategies and life history of non-insectan freshwater invertebrates. Journal of the Fisheries Research Board of Canada 36:311-318.

Masson, S., B. Pinel-Alloul, G. Méthot, and N. Richard. 2004. Comparison of nets and pump sampling gears to assess zooplankton vertical distribution in stratified lakes. Journal of Plankton Research 26: 1199-1206.

Mauchline, J. 1971. Seasonal occurrence of mysids (Crustacea) and evidence of social behavior. Journal of the Marine Biological Association of the United Kingdom 51: 809-825.

McConnell, A., R. Routledge, and B. Connors. 2010. Effect of artificial light on marine invertebrate and fish abundance in an area of salmon farming. Marine Ecology Progress Series 419:147-156.

Moen, V., A. Langeland. 1989. Diurnal vertical and seasonal horizontal distribution patterns of Mysis relicta in a large Norwegian lake. Journal of Plankton Research 11: 729-745.

Ohman, M. D. and others 2013. Autonomous ocean measurements in the California Current Ecosystem. Oceanography 26: 18-25.

O'Malley, B. P., S. Hansson, and J. D. Stockwell. 2018. Evidence for a size-structured explanation of partial diel vertical migration in mysids. Journal of Plankton Research 40: 66-76.

Pepin, P., and T. Shears. 1997. Variability and capture efficiency of bongo and Tucker trawl samplers in the collection of ichthyoplankton and other macrozooplankton. Canadian Journal of Fisheries and Aquatic Sciences 54: 765-773.

Pimm, S. L. and others 2015. Emerging technologies to conserve biodiversity. Trends in Ecology and Evolution 30: 685-696.

Rieman, B. E., and C. M. Falter. 1981. Effects of the establishment of Mysis relicta on the macrozooplankton of a large lake. Transactions of the American Fisheries Society 110: 613-620.

Robertson, A., C. F. Powers, and R. F. Anderson. 1968. Direct observations on Mysis relicta from a submarine. Limnology and Oceanography 13: 700-702.

Rudstam, L., K. Danielsson, S. Hansson, and S. Johansson. 1989. Diel vertical migration and feeding patterns of Mysis mixta (Crustacea, Mysidacea) in the Baltic Sea. Marine Biology 101: 43-52.

Rudstam, L. G. 2009. Other Zooplankton. In G. E. Likens [Ed.] Encyclopedia of Inland Waters. Volume 3. Oxford.

Rudstam, L. G., and O. E. Johannsson. 2009. Advances in the ecology of freshwater mysids. Aquatic Biology 5: 246-248.

Schmidt, K. and others 2011. Seabed foraging by Antarctic krill: Implications for stock assessment, bentho‐pelagic coupling, and the vertical transfer of iron. Limnology and Oceanography 56: 1411-1428.

Schneider, C. A., W. S. Rasband, and K. W. Eliceiri. 2012. NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9: 671.

Shea, M. A., and J. C. Makarewicz. 1989. Production, biomass, and trophic interactions of Mysis relicta in Lake Ontario. Journal of Great Lakes Research 15:223-232.

Stockwell, J. D., T. R. Hrabik, O. P. Jensen, D. L. Yule, and M. Balge. 2010. Empirical evaluation of predator-driven diel vertical migration in Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 67: 473-485.

Stoner, A. W., C. H. Ryer, S. J. Parker, P. J. Auster, and W. W. Wakefield. 2008. Evaluating the role of fish behavior in surveys conducted with underwater vehicles. Canadian Journal of Fisheries and Aquatic Sciences 65:1230-1243.

Straile, D., R. Adrian, and D. E. Schindler. 2012. Uniform temperature dependency in the phenology of a keystone herbivore in lakes of the Northern Hemisphere. Plos One 7: e45497.

Twiss, M. R., and K. M. Stryszowska. 2016. State of emerging technologies for assessing aquatic condition in the Great Lakes-St. Lawrence River system. Journal of Great Lakes Research 42: 1470-1477.

Wiebe, P. H., D. Allison, M. Kennedy, and G. Moncoiffé. 2015. A vocabulary for the configuration of net tows for collecting plankton and micronekton. Journal of Plankton Research 37: 21-27.

Wiebe, P. H., and M. C. Benfield. 2003. From the Hensen net toward four-dimensional biological oceanography. Progress in Oceanography 56: 7-136.

Yurista, P. M., J. R. Kelly, and S. E. Miller. 2009. Lake Superior zooplankton biomass: alternate estimates from a probability-based net survey and spatially extensive LOPC surveys. Journal of Great Lakes Research 35: 337-346.

100

Supplemental Figure 3.11: Screenshot from deep-water camera system deployed at 60 m bottom depth in Lake Champlain, Vermont, USA. Date/time stamp on the video follow eastern time (24:00 hr MM/DD/YY format).

101

Supplemental Figure 3.12: Screenshot of slimy sculpin (Cottus cognatus) interacting with Mysis from deep-water camera system deployed at 60 m bottom depth in Lake Champlain, Vermont, USA. Date/time stamp on the video follow eastern time (24:00 hr MM/DD/YY format).

102

CHAPTER 4: DIEL FEEDING BEHAVIOR IN A PARTIALLY MIGRANT MYSIS

POPULATION: A BENTHIC-PELAGIC COMPARISON

Brian P. O’Malley1, 2, *, and Jason D. Stockwell1

1 Rubenstein Ecosystem Science Laboratory, University of Vermont, Burlington, VT,

USA

2 Current address: USGS Great Lakes Science Center, Lake Ontario Biological Station,

Oswego, NY, USA

*Corresponding author: [email protected]

103

Abstract

Populations that exhibit partial migration include migrants and non-migrants. For benthopelagic organisms that exhibit partial diel vertical migration (PDVM) in aquatic systems, migrants and non-migrants spend different amounts of time in benthic and pelagic foraging arenas over a diel cycle. For example, mysids exhibit PDVM and can feed on both benthic and pelagic resources. Migratory individuals are assumed to undergo vertical migration at night to access pelagic food when predation risk is lower in that arena than during the day. However, feeding behavior of non-migrant benthic individuals is not well understood. One hypothesis to explain individual variability in

DVM behavior is the hunger-satiation state of individuals (hunger-satiation (HS) hypothesis), which predicts that migration is driven by hunger and non-migration is a response to satiation. We assessed diel feeding patterns of benthic- and pelagic-caught

Mysis over a 24-hr cycle from a single site in Lake Champlain on multiple occasions to evaluate if variation in migration behavior was consistent with predictions of the HS hypothesis. Stomach fullness and diet composition revealed little diel difference in stomach contents between time of day or between benthic and pelagic individuals at night. Pelagic individuals were found to consistently have higher stomach fullness shortly after sunset compared to closer to midnight. Non-migrant benthic individuals at night and benthic-caught individuals during the day had similar amounts of detritus in their stomachs. High stomach fullness and high levels of zooplankton prey in benthic-caught stomachs suggest Mysis actively feed when they are benthic, regardless of time of day.

Our results suggest that variation in migration behavior in Mysis is not likely due to

104

hunger-driven behavior, and highlights the dynamic roles of PDVM in habitat and resource use by Mysis in deep lakes.

Keywords: Mysis diluviana, hunger-satiation hypothesis, diel vertical migration, feeding rhythm, benthopelagic, stomach fullness index, partial migration

105

Introduction

Diel vertical migration (DVM) is a widespread behavior of many aquatic organisms (Hutchinson 1967; Hays 2003; Mehner 2012). For organisms that exhibit normal DVM, that is reside in deeper water during the day and migrate upwards at night, many exhibit diel patterns of increased feeding at night at shallower depths compared to the day (Haney and Hall 1975; Ringelberg 2010). While temporal change in light intensity is widely recognized as the proximate cue that triggers DVM (Cohen and

Forward 2009), the vertical distribution of food and internal state of the organism (sensu

Clarke and Mangel 2000) likely also influence depth selection of individuals in lakes and may help explain any deviation from the normal DVM pattern within a population. For example, several krill species (e.g., Euphausia lucens, Meganyctiphanes norvegica, and

Thysanoessa raschi) contain higher amounts of food in their stomachs at night, shortly after they migrate into shallower food-rich waters, followed by midnight sinking behavior after intense feeding (Pearre 1979; Simmard et al. 1986; Gibbons 1993). Midnight sinking varies from normal DVM because downward movement of individuals is not in response to light. The hunger-satiation hypothesis for DVM (HS, Pearre 2003) attempts to explain midnight sinking in organisms that exhibit DVM. The HS hypothesis predicts that individuals migrate to food-rich habitat to feed only when hungry and only return to deeper, food-deplete habitat when satiated. Therefore, individuals ascending to and at shallow depths are expected to be less full than descending individuals.

Not all organisms that use DVM are dependent on pelagic resources to meet nutritional demands (Macquart-Moulin et al. 1987; Wurtsbaugh and Neverman 1988).

Contrary to expectations about foraging gains associated with DVM, some organisms

106

feed primarily on or near the lake bottom during the day and ascend into the water column at night for reasons other than foraging (Wurtsbaugh and Neverman 1988; Levy

1990; Mehner 2012; Harrison et al. 2013). For example, larval Bear Lake sculpin (Cottus extensus) consume mostly benthic invertebrate prey during the day and migrate into the water column at night, where they do not feed (Wurtsbaugh and Neverman 1988;

Neverman and Wurtsbaugh 1992; 1994). Alternative forms of DVM, where substantial foraging occurs in the benthic rather than pelagic arena and are unaligned with feeding opportunity at night, can lead to increased growth by using warmer pelagic habitat to enhance digestion (Wurtsbaugh and Neverman 1988). Consequently, for some benthopelagic organisms which readily consume benthic prey, benthic habitat may be quantitatively more important for prey consumption over the diel cycle. In this context, benthic-pelagic DVM could be used as strategy to exploit benthic resources efficiently or increase bioenergetic efficiency (Ohman 1990; Sims et al. 2006; Mehner et al. 2010), contrary to the more classic model described for planktonic animals to exploit pelagic resources at night (Ringelberg 2010).

Mysids are a group of omnivorous aquatic invertebrates that undergo extensive benthic-pelagic DVM. Members of the Mysis relicta species complex are found throughout northern North America and Europe, and are generally thought to reside on or near the lake bottom during the day and migrate into the water column closer to the surface at night (Mauchline 1980; Väinölä et al. 1994; Rudstam and Johansson 2009).

Mysids feed on detritus, phytoplankton, zooplankton, and benthic invertebrates

(Grossnickle 1982; Takahashi 2004), and through DVM, serve as a trophic linkage for energy flux between benthic and pelagic habitats (Gamble et al. 2011a,b; Isaac et al.

107

2012; Sierszen et al. 2014). In North America, the group is represented by a single freshwater species, M. diluviana (Covich et al. 2009), with one-way migrations of > 100 m reported from deep lakes because of their photophobicity coupled with high water clarity in lakes where they are typically found (Beeton 1960; Rybock 1978; Gal et al.

1999).

Although pelagic zooplankton is often assumed to be the main food source of

Mysis (e.g., Hrycik et al. 2015; Caldwell et al. 2016), bioenergetic and stable isotope mixing models both suggest that consumption of zooplankton alone does not fulfill their energetic requirements, and as a result, mysids likely use other resources to meet daily demand (Johannsson et al. 1994; Johannsson et al. 2001, 2003; Sierszen et al. 2011). The relative contribution of detrital resources to Mysis’ diet is not well known and represents a key knowledge gap to better understand the role of Mysis in nutrient sources and recycling in food webs. For example, in oligotrophic Char Lake, where mysids reportedly fed on shallow, moss-like substrate, mostly inorganic material was observed in stomach contents, in contrast to the more typical zooplankton-based diet (Lasenby and Langford

1973). Compared to the typical zooplankton-based diet reported in the literature, observations from Char Lake highlight the dynamic feeding and habitat use demonstrated by mysids (see also Morgan and Threkheld 1982; Devlin et al. 2016). Furthermore, not all mysids partake in DVM, termed partial DVM (PDVM), where some individuals within the population vertically migrate while others remain benthic at night (Bowers

1988; Moen and Langeland 1989; Euclide et al. 2017; O’Malley et al. 2018). Populations that exhibit PDVM comprise a portion of individuals which have disproportional foraging access to benthic and pelagic arenas at night (Ahrens et al. 2012). Migratory

108

Mysis have greater access to zooplankton prey than non-migrants, and in particular cladocerans, for which Mysis demonstrate strong selectivity (Grossnickle 1978; Bowers and Vanderploeg 1982; Vanderploeg et al. 1982; O’Malley and Bunnell 2014).

Conversely, non-migratory benthic Mysis have greater access to sedimented detritus, harpacticoid copepods, and benthic invertebrates. The diets of pelagic-caught Mysis have been studied on many occasions (e.g., Grossnickle 1979, 1982; Langeland 1988; Chess and Stanford 1998; Whall and Lasenby 2009; Caldwell et al. 2016; O’Malley et al. 2017), but diets of benthic-caught specimens remain a knowledge gap to understand the link between resource use and variability in migration behavior in mysids. Whether mysids that remain benthic at night feed continuously throughout the day and night is largely unknown but seems plausible because of their relatively small gut volume, broad diet range, and appreciable gut evacuation rate at low temperatures experienced near the lake bottom (Chipps 1998). Given that mysids can consume benthic material and some individuals do remain benthic, the contribution of benthic resources to Mysis diet, and thus support to higher trophic levels and benthic nutrient recycling, could be underestimated when research is limited to just the pelagic component of populations.

We tested the HS hypothesis to explain PDVM in a M. diluviana population.

Under scenarios of PDVM and consistent with predictions of the HS hypothesis, individuals that migrate were expected to be hungry (i.e., empty or low stomach fullness), whereas non-migratory individuals collected on the bottom were expected to be satiated

(i.e., relatively higher amounts of food in their stomach compared to migrants). Because

Mysis are omnivorous, we predicted no difference in stomach fullness of benthic-caught

Mysis across the diel cycle because of continuous access to food supply (i.e., detritus).

109

Similarly, for non-migrants captured near bottom at night we predicted diets would be mostly full and similar in composition to those collected during the day, to reflect the benthic environment, but different than diets of migratory individuals collected in the pelagia at night.

Methods

Sample collections

Mysid feeding patterns were assessed at a 100-m deep site in the main basin of

Lake Champlain (44° 28’ N; 73° 18’ W) on four separate occasions: once in 2016 (14-15

June), and three times in 2017 (13-14 April; 18-19 July; 19-20 October). Each occasion consisted of 5-6 sample collection intervals over a 24-hr period (Figure 4.1;

Supplemental Table 4.3). Pelagic mysid samples were only collected at night; benthic mysids were collected all times sampled. Sample collections followed protocols described in O’Malley et al. (2018). Briefly, benthic mysids were collected with a benthic sled (mesh: 1000 µm) towed for 5 minutes along the bottom. Pelagic mysids were collected with a 1-m2 plankton net of similar mesh size towed vertically from 5 m above the bottom to surface. Pelagic mysids were rare in April catches and were not analyzed.

Sample events in June, July, and October contained two separate times the benthic and pelagic regions were compared at night (post-sunset, and midnight). After each sample collection, contents were rinsed into a jar, narcotized with alka-seltzer, and preserved in ethanol for diet analysis.

110

Figure 4.1: Local time of pelagic (P, red triangles) and benthic (B, black circles) sample collections during four separate trips on Lake Champlain. Sample times represent the midpoint of the time interval in which the sample was collected. Grey box is sunset to sunrise (see Supplemental Table 4.3 for exact time values).

Vertical profiles and pelagic zooplankton biomass were measured to characterize pelagic habitat and prey availability across months. Temperature and chlorophyll-a fluorescence profiles were measured with a SeaBird SBE19 CTD equipped with a

WetLabs fluorometer during the day. Crustacean zooplankton biomass and species composition were quantified to assess the prey base available to pelagic mysids.

Zooplankton net tows were collected with a 0.5-m diameter 150-µm mesh net towed vertically from 60 m to surface because most pelagic M. diluviana occupy depths < 60 m after vertical migration (Boscarino et al. 2009). Zooplankton tows were collected during sunset, once per month, prior to the start of night Mysis sampling, except in April when

111

zooplankton were not sampled. Upon collection, zooplankton were rinsed into a jar, narcotized, and preserved in ethanol.

Sample analysis

Zooplankton samples were processed under an Olympus SZX12 dissecting scope equipped with a drawing attachment and digitizing tablet for length measurements. At least 150 individuals per sample were counted, measured, and identified in a counting wheel from a known volume of subsample. Adult copepods and cladocerans were identified to species or genus, while immature copepods were classified as cyclopoid or calanoid copepodids, except for immature Epischura. Copepods were measured from the anterior edge of the cephalosome to the distal end of the caudal rami; cladocerans were measured from the top of the head to the base of the caudal spine or carapace. Nauplii were not included in counts because the mesh size we used results in underestimates

(Evans and Sell 1985). Individual lengths were converted to dry mass using published length:mass equations (Bottrell et al. 1976; Burgess et al. 2015; Supplemental Table 4.4).

Zooplankton dry biomass density (mg/m3) was calculated from volume filtered assuming net filtration efficiency of 1.

Mysis stomach fullness was scored on a semi-quantitative scale derived from methods used in euphausid diet studies (Nakagawa et al. 2003; Ponomareva 1971;

Simmard et al. 1986) because of the many similarities in biology and feeding between mysids and euphasids (Mauchline 1980; Rudstam and Johannsson 2009). Direct examination of the stomach contents prior to tearing open the stomach was possible because Mysis stomach tissue is thin and transparent (Figure 4.2a-b). Individual mysids were photographed under an Olympus SZX9 dissecting scope with a digital camera for 112

body length measurements in ImageJ. Because Mysis diet varies with body size, only adult size-classes (≥ 10 mm) were targeted (Branstrator et al. 2000; Johannsson et al.

2003). Sizes were similar across samples overall except for some night-pelagic samples when slightly smaller individuals were used for analyses because large animals were rare in catches. Image analysis occurred prior to stomach dissection to avoid bias from disturbance. Stomach fullness was scored by the same individual with the following criteria: 0 = empty; 1 = trace food up to one quarter full; 2 = more than one quarter to half full; 3 = more than half up to three quarter full; 4 = more than three quarter full to gorged

(Figure 4.2c-g). A stomach fullness index (SFI) was calculated according to Nakagawa et al. (2003) as the mean score among individuals per sample, where the mid-point gut percentage fullness value of each fullness class (i.e. class 0 = 0; class 1 = 12.5; class 2 =

37.5; class 3 = 62.5; class 4 = 87.5) was used:

4 퐹푚푖 SFI = ∑(퐹푝 ⋅ ) 푖 100 푖=0 where Fpi is the percentage of total Mysis examined with fullness class i (%) and Fmi is the median of each fullness class i.

After scoring a stomach for fullness, stomach contents from a subset of those adults (up to 10 per sample) were analyzed for diet composition and prey counts.

Stomach contents were removed from an individual stomach, placed in a drop of reverse osmosis water, and spread on a Sedgewick Rafter cell for identification and enumeration of ingested prey under an Olympus IX inverted microscope. Ingested rotifer lorica were mostly intact and were identified to genus level, whereas intact crustacean zooplankton prey were rare in diets. Therefore, we relied on diagnostic prey remains to quantify

113

crustaceans that were fragmented (e.g., mandibles, tail spines, rostra, postabdominal claws, and caudal rami). Multiple body parts for a single prey taxon were often available in a gut. For those instances, we relied on structures that yielded the highest number of prey in the diet after dividing by two for parts that occur naturally in pairs (e.g. mandibles). Zooplankton remains were identified according to O’Malley and Bunnell

(2014) and unpublished images of dissected zooplankton. Because detritus is not possible to quantify in stomach contents by counting, we instead relied on a qualitative scoring method. Relative abundance of detritus in a diet was ranked from 0 to 4 similar to protocol of Johannsson et al. (2001), with a value of 0 for no detectable amount, 1 representing trace detritus, and 4 indicating a very high amount. All mysid stomachs were analyzed by the same person. Archived mysids that were experimentally fed sediment in the laboratory were used as a training set to differentiate between detritus and plankton remains in stomach contents. Other non-zooplankton prey items that were easily recognizable in stomach contents (e.g., pollen, ostracods, and dinoflagellate cysts) and have been reported from mysid diet studies elsewhere were also counted (Viherluoto et al. 2000; Johannsson et al. 2003; Caldwell et al. 2016).

114

Figure 4.2: (A) Mysis cephalic region after separation from the body with thoracic carapace removed. (B) Close-up view of Mysis cardiac stomach (circled region) (C-G) Gut scale ranking system for classifying gut fullness of Mysis stomach contents: (C) 0 = empty; (D) 1 = trace food up to one-quarter full; (E) 2 = half full; (F) 3 = three-quarters full; (G) 4 = fully gorged.

115

Our stomach fullness scale only accounted for the approximate volume and not the composition. To evaluate if our assigned fullness score equated to more zooplankton prey ingested, we first tested for a correlation between the number of zooplankton in a stomach as a function of fullness score. For stomachs that were classified as full (fullness rank 4), we expected that those stomachs would either contain a high amount of zooplankton or a high amount of detritus (detritus rank 4), but not both. Stomach fullness was assigned visually prior to dissecting the stomach and assigning a detritus score or counting the number of prey. To test for the possibility that a stomach could be full and have a high amount of detritus or zooplankton independent of each other, we tested for an inverse relationship between the amount of zooplankton and detritus for only those diets that were classified as full. For each model, we also evaluated with and without including habitat as a parameter and used ΔAIC to compare models and assess if habitat improved the model.

Gut fullness and detritus rank scores were compared using a Kruskal-Wallis rank sum test for each sample month to test for variation in scores across time of day. Diel change in SFI was evaluated by visual inspection for increasing or decreasing trends in

SFI values. Diet similarity by abundance of zooplankton and other non-detrital prey items counted in diets (i.e., daphnid ephippia, dinoflagellate cysts, ostracods, and pollen) was evaluated using ANOSIM between pelagic and benthic night-samples to test whether diet composition varied between migrant and non-migrants. ANOSIM produces a global R- statistic, which is calculated from the ratio of between-group and within-group dissimilarities, that ranges from 0 (low similarity) to 1 (high similarity; Clarke and

Warwick 1994). Calculations were carried out in the R package vegan using the anosim

116

function with a Bray dissimilarity matrix and 999 permutations excluding diets recorded as empty (Okansen et al. 2018).

To evaluate if reduced access to zooplankton prey in benthic-caught individuals would result in lower average number of prey in stomach contents compared to diets of pelagic-caught individuals at night, we used one-way ANOVA with a Bonferroni correction for multiple comparisons to compare the number of prey observed in gut contents for select prey categories (cyclopoids, calanoids, cladocerans, and rotifers) in each month. We assumed zooplankton availability was lower in benthic habitat, and higher in pelagic habitat. We kept each month separate for analysis because zooplankton abundance and species composition fluctuate seasonally (Sommer et al. 1986) and thus conditions were expected to vary enough to warrant separate analyses. Frequency of occurrence for prey items in diets by day and night benthic samples, and night-pelagic samples was calculated and graphed by prey type to identify if prey defined as benthic or pelagic origin occurred at relatively higher frequency in diets consistent with expectations of the habitat from which the diet was sampled. The percent occurrence of high detritus in stomachs (detritus rank 4) was included to assess if high detritus was more common in benthic individuals.

Results

The water column was thermally stratified during three of four sampling months

(Figure 4.3). Temperature and chlorophyll concentrations were uniformly distributed throughout the water column in April. The strongest temperature gradient observed was in July and the deepest thermocline depth occurred in October. Chlorophyll profiles

117

showed sub-surface chlorophyll maxima at temperatures of 11-14 °C June-October

(Figure 4.3).

Crustacean zooplankton composition varied from mostly cyclopoid biomass

(80%) in June, to a greater representation of cladocerans (daphnids and bosminids) in

July (46%; Figure 4.4). Calanoids accounted for a greater amount of the community in

October (49%) relative to June (19%) or July (18%). Total zooplankton biomass density was highest in the month of July (60.8 mg/m3), and lower and more similar in June (9.0 mg/m3) and October (19.0 mg/m3).

Figure 4.3: Vertical profiles of temperature and chlorophyll-a fluorescence at the 100-m sampling site in Lake Champlain in June 2016, and April, July, and October 2017.

118

Figure 4.4: Crustacean zooplankton biomass (mg/m3) and % composition in 60 m to surface vertical net tows at the 100-m site in Lake Champlain in June 2016 and July- October 2017.

A total of 754 mysids stomachs were visually examined for fullness. Of those, the diets of 231 mysids were dissected and examined microscopically. The most frequently encountered stomach fullness category was the maximum value 4 (N = 459 diets), followed by 3 (N = 200), 2 (N = 50), and 1 (N = 26). Nineteen stomachs were visually classified as empty (7 day-benthic, 7 night-benthic, and 5 night-pelagic). Eight of the stomachs classified as empty were also dissected and contained zero prey except for one of the diets which yielded one prey part only visible after dissection.

Stomach fullness rankings were positively correlated with the number of zooplankton in Mysis stomach contents, though only a small amount of the variation was explained by linear regression (Figure 4.5, N = 231; P < 0.01; R2 = 0.12). The inclusion of habitat to the model explained more variation (P < 0.01; R2 = 0.19) and resulted in a more parsimonious model than stomach fullness rank alone (ΔAIC = 19.3). In stomachs classified as mostly full (rank 4) and those that were also examined for diet composition

(N = 122), we observed a weak inverse relationship between detritus score and the number of zooplankton in a stomach, as expected. Detritus scores and habitat as 119

predictors of the number of zooplankton stomach contents led to a better model compared to just detritus scores (P < 0.01; R2 = 0.12; ΔAIC = 7.5). The detritus scores alone explained almost none of the variability (P < 0.01; adj. R2 = 0.05).

Figure 4.5: Correlation between number of total zooplankton prey (copepods, cladocerans, and rotifers) in Mysis gut contents (ZP) and gut fullness rank score from Lake Champlain in June 2016 and July-October 2017.

Stomach fullness scores showed little variation with time of day. Benthic SFI values ranged from 57.5 to 82.5 and did not show clear trends with time of day (mean =

72.3 ± 5.6 SD, Figure 4.6; Supplemental Figure 4.10). Pelagic SFI values varied more within a night than did benthic samples, and always showed the same pattern where higher values were first observed shortly after sunset and lower values occurred in samples close to midnight (Figure 4.6). Empty diets were more frequent in night-pelagic than day-benthic diets collected close to midnight in all three months (Supplemental

120

Figure 4.10). The distribution of stomach fullness scores was not significantly different with time of collection in April (Kruskal Wallis, H = 1.43, d.f. = 4, P = 0.84), June (H =

10.92, d.f. = 7, P = 0.14), and October (H = 6.73, d.f. = 7, P = 0.46). A significant difference in gut fullness scores over the diel cycle was only detected in July samples (H

= 21.06, d.f. = 7, P < 0.01). Post-hoc comparison indicated the July post-sunset and afternoon benthic sled samples were significantly different from each other (Dunn test, Z

= 0.24, P-value < 0.01) but were not significantly different from the other July samples.

Figure 4.6: Temporal changes in stomach fullness index (SFI) of benthic- (black circles) and pelagic-caught (red triangles) mysids by month from Lake Champlain. Grey box is sunset to sunrise.

Detritus scores were mostly consistent across the diel cycle for benthic-caught mysids, though significant pairwise differences were detected across the diel cycle in each of the three months (Figures 4.7 and 4.8). In June and July, low detritus scores were more common in night-pelagic diets than in October. In June (Kruskal Wallis, H = 21.91,

121

d.f. = 7, P < 0.01), detritus scores in diets of individuals collected in night-pelagic samples were only significantly lower than benthic diets collected in the late-morning the following day (Dunn test, Z = -3.14, Bonferroni adj. P < 0.05), but were not different among pairwise comparisons of benthic diets from pre-sunset through dawn (adj. P >

0.05). A similar pattern was observed in July (Kruskal Wallis, H = 33.26, d.f. = 7, P <

0.01), with significantly lower detritus scores in night-pelagic diets shortly after sunset, compared to benthic diets collected late-morning (Z = -3.67, adj. P < 0.01) and afternoon

(Z = -4.61, adj. P < 0.01) the following day. July night-pelagic diets near midnight were not significantly different from other observations in that month. In October (Kruskal

Wallis, H = 14.94, d.f. = 7, P < 0.05), detritus scores appeared higher overall relative to other months. Detritus scores of October night-pelagic stomachs collected close to midnight were lower compared to those from benthic-day individuals at pre-sunset (Z = -

3.34, adj. P < 0.05) and mid-morning (Z = -3.16, adj. P < 0.05).

Diet composition by number of prey items observed in night-collected stomachs revealed no significant differences between benthic- and pelagic-caught mysids in June

(ANOSIM, N = 37, R = 0.035, P = 0.13) and July (N = 32, R = 0.047, P = 0.18). October diets were different between night-benthic and night-pelagic mysids, but were still similar overall as indicated by a low R statistic (N = 38, R = 0.115, P < 0.01).

122

Figure 4.7: Variation in detritus scores of pelagic and benthic Mysis by time of day in June 2016 and July and October 2017 in Lake Champlain.

Figure 4.8: Mean (± SE) detritus rank of stomach contents from benthic- and pelagic- caught Mysis by time of day. Points labelled with the same letter denote no significant difference among time of day (Dunn test with Bonferroni correction). Letters are located above for benthic and below for pelagic samples. * indicates excluded estimate due to small sample size. 123

Comparison of the number of zooplankton prey in stomach contents over the diel cycle indicated significant differences in 3 of 12 cases (Table 4.1). Significant differences in the number of cyclopoids in diets were observed in July and October, and in the number of calanoids in diets in October (Table 4.2). No significant differences in the number of cladocerans or rotifers in diets through time were detected out of the three months. In July, more cyclopoids were present in day-benthic diets around mid-morning after night sampling, a trend that was not evident when other prey types were considered.

In October, the number of calanoids in stomach contents was significantly higher in night-pelagic diets closer to midnight than during the day for benthic-caught individuals, and the same was true for cyclopoids. The number of calanoids in night-pelagic diets closer to midnight was not significantly different from the amount observed in night- pelagic diets collected a few hours earlier just after sunset, but was greater than in both night-benthic diet sample collections. In contrast, the number of cyclopoids in night- pelagic stomachs collected closer to midnight was significantly greater than just after sunset for both benthic and pelagic samples.

124

Table 4.1: One-way ANOVA results of the number of prey for each category in Mysis stomach contents over a diel cycle in Lake Champlain. F values that were significant following a Bonferroni correction (α = 0.05/12) for multiple comparisons are denoted by asterisks.

Date Sampled Prey Category d.f. F P 14/15 June Calanoid 7 1.45 0.20 Cyclopoid 7 1.16 0.34 Cladocera 7 2.61 0.02 Rotifer 7 2.71 0.02 18/19 July Calanoid 7 1.29 0.27 Cyclopoid 7 3.49 < 0.01* Cladocera 7 0.61 0.75 Rotifer 7 0.70 0.67 19/20 October Calanoid 7 4.27 < 0.01* Cyclopoid 7 4.35 < 0.01* Cladocera 7 1.48 0.19 Rotifer 7 0.87 0.54

125

Table 4.2: Mean size of Mysis and number of prey items, and the minimum and maximum number observed, in Mysis stomach contents collected at different times of day and habitat (benthic/pelagic). Grey denotes samples collected at night (between sunset-sunrise). Bold values denote months when significant differences in mean number of prey over a diel period were detected after Bonferroni correction. Superscripts refer to Tukey’s HSD post-hoc comparison for those cases.

Mysid length (mm) Prey Category

Date Sampled Time (hr) Gear Mean SD N Calanoid Cladocera Cyclopoid Rotifer Other

14/15 June 14:00 Benthic 14.7 1.1 10 2.7 (0-14) 3.2 (2 - 6) 5.4 (1-19) 0.6 (0-3) 7.1 (0-15)

18:00 Benthic 14.2 1.2 10 0.2 (0 - 2) 1.3 (0 - 3) 4.0 (0-10) 0.4 (0-2) 3.4 (0 - 6)

22:00 Benthic 16.6 2.4 10 0.8 (0 - 2) 1.9 (1 - 3) 6.4 (1-17) 0.5 (0-2) 7.7 (0-24)

126 22:00 Pelagic 14.8 1.3 10 1.8 (0 - 4) 5.3 (0-13) 8.1 (1-16) 2.2 (0-4) 10.8 (2-23)

00:00 Benthic 14.6 0.9 10 1.0 (0 - 3) 2.5 (0 - 6) 5.5 (2 - 8) 2.2 (0-8) 9.0 (1-23)

00:00 Pelagic 11.6 2.2 7 2.9 (0-14) 4.7 (0-12) 3.0 (0-11) 0.9 (0-4) 5.1 (0-16)

06:00 Benthic 14.9 1.5 10 0.9 (0 - 3) 3.3 (1 - 7) 5.4 (2-15) 0.9 (0-2) 1.3 (0-26)

10:00 Benthic 15.8 1.5 10 1.7 (0 - 4) 3.5 (0-10) 4.9 (0 - 7) 0.7 (0-2) 1.1 (0-21)

18/19 July 19:00 Benthic 14.8 1.0 10 1.9 (0 - 5) 12.1 (2-23) 2.4b (1 - 6) 0.3 (0-2) 2.0 (0 - 5)

22:00 Benthic 14.7 1.1 10 1.9 (0 - 6) 11.6 (2-23) 3.0b (0 - 5) 0.8 (0-3) 1.2 (0 - 5)

22:00 Pelagic 11.3 0.5 7 1.0 (0 - 2) 14.1 (8-20) 2.3b (1 - 4) 0.3 (0-1) 0.0 (0 - 0)

00:00 Benthic 14.5 2.2 10 0.8 (0 - 2) 10.5 (0-18) 2.3b (0 - 5) 0.2 (0-1) 1.7 (0 - 4)

00:00 Pelagic 16.4 2.3 7 1.4 (0 - 3) 12.0 (0-22) 2.0b (0 - 5) 0.7 (0-3) 0.0 (0 - 0)

07:00 Benthic 14.7 1.2 10 0.8 (0 - 1) 8.8 (0-18) 3.4ab (0 - 7) 0.5 (0-2) 1.3 (0 - 6)

10:00 Benthic 14.5 2.6 10 1.3 (0 - 4) 10.7 (1-18) 6.5a (2-18) 0.6 (0-2) 3.0 (0 - 7)

13:00 Benthic 16.0 1.2 10 1.3 (0 - 3) 15.0 (3-46) 4.0ab (2 - 7) 0.3 (0-2) 1.4 (0 - 3)

19/20 October 17:00 Benthic 14.8 1.4 10 2.0b (0-10) 11.6 (4-26) 4.1b (0-23) 0.5 (0-2) 4.9 (1 - 13)

19:00 Pelagic 16.5 0.9 10 3.3ab (0-16) 13.7 (1-47) 4.5b (0-11) 0.9 (0-4) 3.8 (0 - 12)

20:00 Benthic 15.5 1.8 10 1.6b (0 - 8) 10.1 (4-24) 4.0b (0-16) 1.2 (0-2) 4.8 (0 - 13)

127 22:00 Benthic 16.6 1.4 10 1.4b (0 - 5) 14.0 (1-44) 6.1ab (1-19) 1.0 (0-4) 8.7 (1-15)

22:00 Pelagic 16.2 1.3 10 8.1a (0-26) 22.2 (0-73) 15.6a (0-41) 0.7 (0-2) 3.5 (0 - 14)

07:00 Benthic 14.9 1.2 10 0.6b (0 - 2) 7.8 (4-14) 1.7b (0 - 3) 0.5 (0-3) 5.6 (1 - 11)

10:00 Benthic 15.5 1.0 10 0.7b (0 - 3) 8.1 (1-13) 2.1b (1 - 4) 0.8 (0-3) 4.1 (0 - 8)

14:00 Benthic 17.0 2.2 10 0.4b (0 - 2) 11.3 (2-44) 2.3b (1 - 9) 0.3 (0-1) 4.7 (0 - 11)

1 The occurrence of pelagic prey in stomachs examined did not vary considerably

with habitat or with day/night collections as predicted. Bosminids, daphnids, and

cyclopoids were the most frequently observed zooplankton prey taxa across all months

and were consistent between benthic and pelagic diets within a month (Figure 4.9). Prey

that were frequently observed in night-pelagic diets were also common in benthic diets.

In June and July, prey items designated as benthic prey were observed in benthic-caught

diets both day and night but not in pelagic-caught diets. High detritus (rank 4 exclusively)

occurred in approximately 50% of benthic day-night diets in June and July, and far less

common in night-pelagic diets those months. By October, the pattern was mixed with

high detritus found in 50% of night pelagic, and 60-80% of benthic individuals examined.

Ostracods were present in 2.5% of benthic-day and 10% of benthic-night individuals

examined in June and July. Harpacticoids were the only other benthic prey type observed

in July of benthic-caught diets but were infrequent. In October, resting stages of prey in

the forms of dinoflagellate cysts and daphnid ephippia were frequent in benthic and

pelagic-caught diets. Daphnid ephippia were more common in benthic-day (60%) and

benthic-night diets (65%), than in pelagic-night diets (40%). Occurrence of dinoflagellate

cysts in October diets were not drastically different across benthic and pelagic samples

(60-70%).

128

Figure 4.9: Frequency of occurrence of prey items by habitat-time sampled (BD = benthic-day; BN = benthic-night; PN = pelagic-night). Prey items with values of 0% are not shown as symbols. Pelagic prey items are ordered by decreasing body size (green font), from Keratella to Senecella; last four prey items of list (Dinoflagellate.cyst- Detritus.4.only) are strictly benthic prey (brown font).

Discussion

We evaluated predictions of the HS hypothesis to explain variability in PDVM behavior in a Mysis population. We predicted that if migration was driven by hunger, then mysids with less full stomachs should make up the migrant, pelagic portion of the population at the start of the night, while satiated, non-migrant benthic individuals would make up the benthic portion throughout the night. A second prediction was that we would see declining stomach fullness from morning to evening. Our stomach fullness results indicated that over a diel cycle, Mysis stomachs were rarely empty, regardless of habitat or time of day from which they were collected. More than half of all stomachs examined 129

were classified as full, suggesting that mysids likely exhibit continuous and opportunistic feeding without a clear diel shift in terms of stomach fullness.

Multiple hypotheses have been proposed that could explain PDVM of Mysis. The hypotheses include differences in demographics, body size, body condition, genetics, and different feeding behavior among migrants and non-migrants (Ogonowski et al. 2013;

Euclide et al. 2017; O’Malley et al. 2018). Though not mutually exclusive from each other, the potential hypotheses that have been proposed and evaluated in some instances are by no means a comprehensive list. Given the current state of knowledge in the field of partial migration, and especially in invertebrates that have been understudied (Chapman et al. 2011), much work still remains to rule out certain hypotheses. In the current study, we rejected the HS hypothesis to explain PDVM in Lake Champlain Mysis. Previous studies have found support for body size and condition related hypotheses (Euclide et al.

2017; O’Malley et al. 2018) within the Lake Champlain population. A comprehensive assessment that evaluates the degree of support among multiple competing hypotheses for Mysis PDVM is needed as a next-step towards explaining PDVM which will undoubtedly benefit the field of partial migration (Chapman et al. 2011; Betini et al.

2017).

Diet descriptions of benthic-caught Mysis are not as common in the literature as pelagic-caught specimens. A brief literature search of 21 M. diluviana/relicta diet studies found only 30% of them examined benthic-caught individuals (JDS, unpublished data).

More time within a day is likely spent on or near the bottom during summer months in the northern hemisphere, when primary and secondary production are typically greatest, because of short nights. Inferred diet from stable isotope mixing models of night-pelagic

130

Mysis from Lake Superior suggested that benthic and pelagic contributions to diet were similar, despite the shorter amount of time available for Mysis to forage in the pelagic zone at night compared to the day during the summer (Sierszen et al. 2011). Such disproportionate nutritional contributions relative to the time presumably spent in each habitat could reflect tradeoffs between food quantity and quality in energy assimilation.

Sedimented detritus is widely available to benthic Mysis during the day and night, but of low energy value compared to pelagic zooplankton (Cummins and Wuycheck 1971), and consequently may serve as a subsistence food resource for individuals that remain near the lake bottom. Abstinence from consuming benthic food while ingested material from a previous night pelagic foray is slowly digested is a possibility. However, mysids have several biological constraints including the small volume of their cardiac stomach.

Further, gut evacuation rate of 4.6 hours has been reported for M. diluviana at cold hypolimnetic temperatures (4°C) with faster rates at higher temperatures (Chipps 1998).

Thus, we would have expected a high frequency of empty guts for benthic-caught Mysis late in the day prior to sunset if pelagic feeding was solely responsible for maintaining food in the stomach. Our results clearly demonstrate Mysis fed continuously over the diel period and use both benthic and pelagic foraging arenas.

The SFI provided semi-quantitative data on feeding history. We found a positive, albeit somewhat weak, relationship between the fullness index and the amount of zooplankton in Mysis stomach contents. High variation in the number of zooplankton per individual was not surprising in stomachs classified as full because animals were gathered from multiple habitats and times of day. Adding habitat as a variable improved the amount variability explained, but only slightly. We anticipated that individuals with

131

full guts and low numbers of zooplankton would have high relative scores of detritus to compensate for low amounts of zooplankton. However, we only observed a weak inverse relationship between the amount of detritus and the number of zooplankton in stomach contents. We did not record the relative amount of phytoplankton in stomach contents.

Mysis transition from primarily herbivorous as juveniles to more carnivorous as adults

(Branstrator et al. 2000), and phytoplankton is not considered a major diet component of adult-sized mysids except possibly in spring (Johannsson et al. 2001, 2003; O’Malley et al. 2017). Our diet data on large-sized Mysis, along with previous observations that large mysids account for the majority of the non-migrant portion (Euclide et al. 2017;

O’Malley et al. 2018), suggest that benthic detrivory might also be an important ontogenetic diet shift to consider. Accounting for phytoplankton and detritus through qualitative (Johannsson et al. 2001) or quantitative (Grossnickle 1979; O’Malley et al.

2017) metrics could provide a more complete picture in future studies that seek to combine gut fullness and diet composition measurements from the same individual to assess feeding. Our results support the need to include benthic Mysis in future studies to better understand drivers and impacts of PDVM on trophodynamics of deep lakes.

Contrary to the expectation that zooplankton prey would be more abundant in diets of night-pelagic or dawn-benthic individuals than those from just before sunset, we only detected diel differences in the amount of zooplankton in Mysis stomachs in a few instances. In October, results were in-line with predictions about pelagic foraging; numerically more calanoids and cyclopoids were observed in stomach contents of night- pelagic caught individuals close to midnight than other times of day. We did not find a similar result for cladocerans or rotifers. We were unable to detect any diel difference,

132

although zooplankton, and in particular cladocerans and rotifers typically occur at higher densities in epi/meta-limnetic waters at night than near the lake bottom (Chess and

Stanford 1998; Nowicki et al. 2017; Watkins et al. 2017), and would be more available to pelagic mysids at night. One caveat with our study and the use of diagnostic prey remains to quantity zooplankton prey from Mysis stomach contents is that variability in digestion times for zooplankton hard structures could bias our snapshots of diet composition through time if some parts degrade more rapidly than others or accumulate in the stomach over time (e.g., Parker et al. 2001). In mysid diet studies, zooplankton mandibles are often used because mysids shred their prey and these represent the most rigid structures with high concentrations of silica to withstand constant grinding. Mandibles can also persist for longer than other structures in guts of starved mysids (Rybock 1978; Murtaugh

1984). Nonetheless, if substantial feeding activity on zooplankton occurs at night we still expected to see a decline in the number of zooplankton in stomachs and the relative fullness given the long window of opportunity for daytime digestion to occur (Rudstam et al. 1989). The lack of diel difference suggests Mysis capture zooplankton during the day perhaps by suspending off the lake bottom. Mysis can occur off the bottom during the day if adequate low-light habitat is available (Robertson et al. 1968; O’Malley et al., 2018).

However, those found suspended off the bottom during the day in Lake Champlain tend to be mostly juvenile mysids that have higher light and temperature tolerance than the adults considered in our study (O’Malley et al. 2018). A second possibility is that mysids ingest zooplankton remains that accumulate at the sediment. Research experiments that address benthic foraging behavior of mysids are needed. We suggest that fine-scale videography could provide vital details on how Mysis forage at the sediment interface.

133

Stomach fullness showed a consistent pattern in pelagic Mysis collected just after sunset and then again around midnight. Each month, fullness was higher in the earlier post-sunset collection compared to the midnight collection. We expected that Mysis stomachs would be less full shortly after sunset and then increase by midnight after ample opportunity to forage in the pelagic zone. Others have shown Mysis migrate up with low amounts of food in their stomach that increases when Mysis are sampled several hours later in the night (Grossnickle 1979; Rudstam et al. 1989). We are unsure why our observations deviated from the expected pattern. One possibility is that Mysis in Lake

Champlain left the bottom with full stomachs. A second possibility is that feeding took place earlier during vertical ascent and was more rapid than expected in pelagic individuals. Such a possibility, coupled with our observations that stomachs of benthic- caught Mysis contained high amounts of food just before sunset, suggests Mysis likely did not have to start from empty when foraging pelagically at night. We were unable to assess if pelagic Mysis exhibited midnight sinking with our limited night sampling.

Extended pelagic sampling that includes collections between midnight and pre-dawn and with finer depth resolution (e.g., acoustics; depth-stratified tows) than our whole-column tows is needed to evaluate potential for midnight sinking behavior.

One limitation of using diet tracers in stomach contents to identify benthic or pelagic feeding is that the approach requires different prey types to only occur in one habitat but not both (Pearre 2003). Although we did not measure the relative abundance of each prey type from benthic and pelagic habitats, we assumed detritus, ostracods, harpacticoids, and resting eggs would be representative of benthic feeding because these groups occur primarily in lake benthos (Vihurelouto et al. 2000; Dussart and Defaye

134

2001; Covich et al. 2009). We acknowledge that crustacean groups such as harpacticoids and ostracods can swim and occur in the pelagia. However, they are generally rare in pelagic waters compared to cladocerans, calanoids, and cyclopoids (Balcer et al. 1984;

Chess and Stanford 1998; Rudstam et al. 1989). Similar to our findings, Mysis diets reported from other studies that investigated benthic diets also reported surprisingly low amounts of these benthic prey groups (i.e., ostracods, harpacticoids, and resting eggs), despite their high abundance in benthos (Rudstam et al. 1989). Johannsson et al. (2001,

2003) found body parts of the benthic amphipod Diporeia in more benthic- than pelagic- caught Mysis diets, and amphipod interactions could be an important benthic energy source to consider because of the high energetic value of Diporeia (Parker 1980; Stewart and Sprules 2011). We did not find amphipods in diets likely because they are rare in profundal waters of Lake Champlain (Dermott et al. 2006; Knight et al. 2018). In summary, stomach content analysis suggests Mysis do not feed heavily on zoobenthos in

Lake Champlain. Instead, Mysis likely use sedimented detritus when benthic, regardless of time of day.

Conclusion

Our findings suggest a lack of support for the HS hypothesis to explain PDVM in mysids because stomach contents were rarely empty and did not display a strong diel pattern consistent with predictions of the HS hypothesis. We did not find strong evidence to suggest that adult mysids mainly prey on zooplankton because of high amounts of detritus observed in stomach contents, especially in benthic non-migrant individuals.

Overall, we found more evidence to reject than support the HS hypothesis for PDVM because of Mysis’ ability to feed on benthic and pelagic food. Therefore, other drivers 135

likely dictate PDVM. Future studies that seek to evaluate connections between foraging and PDVM should also consider specialization or tradeoffs in predation as those were not examined in the present study. We propose that non-migrant benthic individuals could rely on benthic detritus to a much greater extent than migratory individuals because

Mysis feed continuously. In turn, diet and distribution data presented in this study could be useful to understand the roles of partial migration behavior in food web energy dynamics and source contributions that support fish communities reliant on Mysis in systems such as Great lakes where native fish restoration efforts are underway

(Eshenroder and Burnham-Curtis 1999, Hondorp et al. 2005; Eshenroder 2008;

Zimmerman and Krueger 2009). Aspects of pelagic foraging have long been the focus of

Mysis ecology studies, and our analyses point to a need to better understand the benthic ecology of Mysis, including the importance of detritus to Mysis energetics, which by extension could fill in knowledge gaps about the basal energy sources that support both pelagic and demersal fish communities.

Acknowledgements

We thank captain Steve Cluett and deckhand Brad Roy of the R/V Melosira for their help with diel Mysis sampling. This project was supported by the Great Lakes

Fishery Commission with the assistance of Senator Patrick Leahy, with additional support from a University of Vermont Chrysalis Award and a scholarship from the

International Association for Great Lakes Research to BPO.

136

References

Ahrens, R. N., C. J. Walters, and V. Christensen. 2012. Foraging arena theory. Fish and Fisheries 13:41-59. Balcer, M. D., N. L. Korda, and S. I. Dodson. 1984. Zooplankton of the Great Lakes: A guide to the identification and ecology of the common crustacean species. University of Wisconsin Press. Beeton, A. M. 1960. The vertical migration of Mysis relicta in Lakes Huron and Michigan. Journal of the Fisheries Board of Canada 17:517-539. Betini, G. S., T. Avgar, and J. M. Fryxell. 2017. Why are we not evaluating multiple competing hypotheses in ecology and evolution? Royal Society Open Science 4:160756. Boscarino, B. T., L. G. Rudstam, J. L. Eillenberger, and R. O’Gorman. 2009. Importance of light, temperature, zooplankton and fish in predicting the nighttime vertical distribution of Mysis diluviana. Aquatic Biology 5:263-279. Bottrell, H. H., A. Duncan, M. Z. Gliwicz, E. Grygierek, A. Herzig, A. Hillbricht- Ilkowska, H. Kurasawa, P. Larsson, T. Weglenska. 1976. A review of some problems in zooplankton production studies. Norwegian Journal of Zoology. 24: 419-456. Bowers, J. A. 1988. Diel vertical migration of the opossum shrimp Mysis relicta in Lake Superior: observations and sampling from the Johnson-Sea-Link II submersible. Bulletin of Marine Science 43: 730-738. Bowers, J. A., and H. A. Vanderploeg. 1982. In situ predatory behavior of Mysis relicta in Lake Michigan. Hydrobiologia 9:121-131. Branstrator, D. K., G. Cabana, A. Mazumder, and J. B. Rasmussen. 2000. Measuring life‐ history omnivory in the opossum shrimp, Mysis relicta, with stable nitrogen isotopes. Limnology and Oceanography 45:463-467. Burgess, S., E. W. Jackson, L. Schwarzman, N. Gezon, and J. T. Lehman. 2015. Improved estimates of calanoid copepod biomass in the St. Lawrence Great Lakes. Journal of Great Lakes Research 41:484-491. Caldwell, T. J., F. M. Wilhelm, and A. M. Dux. 2016. A non-native pelagic macroinvertebrate alters population dynamics of herbivorous zooplankton in a large deep lake. Canadian Journal of Fisheries and Aquatic Sciences 73:832-843. Chapman, B. B., C. Brönmark, J. Å. Nilsson, and L. A. Hansson, L. A. 2011. The ecology and evolution of partial migration. Oikos 120:1764-1775.

137

Chess, D. W., and J. A. Stanford. 1998. Comparative energetics and life cycle of the opossum shrimp (Mysis relicta) in native and non-native environments. Freshwater Biology 40:783-794. Chipps, S. R. 1998. Temperature-dependent consumption and gut-residence time in the opossum shrimp Mysis relicta. Journal of Plankton Research 20:2401-2411. Clark, C. W., and M. Mangel. 2000. Dynamic state variable models in ecology: methods and applications. Oxford University Press. Clarke, K. R., and R. M. Warwick. 1994. Similarity-based testing for community pattern: the two-way layout with no replication. Marine Biology 118:167-176. Cohen, J. H., and R. B. Forward. 2009. Zooplankton diel vertical migration - a review of proximate control. Oceanography and Marine Biology: An Annual Review 47:77- 109. Covich, A. P., J. H. Thorp, and D. C. Rogers. 2009. Introduction to the subphylum Crustacea. In Ecology and Classification of North American Freshwater Invertebrates – 3rd Edition. pp. 695-723. Cummins, K. W. and J. C. Wuycheck. 1971. Caloric equivalents for investigations in ecological energetics. Internationale Vereinigung für Theoretische und Angewandte Limnologie: Mitteilungen 18:1-158. Dermott, R. M., R. Bonnell, P. Jarvis. 2006. Changes in abundance of deep water amphipod Diporeia (Pontoporeiidae) in eastern North American lakes with or without Dreissena mussels. Canadian Technical Report of Fisheries and Aquatic Aciences. pp. 88. Dussart, B. H., and D. Defaye. 2001. Introduction to the Copepoda. In Guide to the identification of the microinvertebrates of the continental waters of the world, 2nd edition, No. 16. Backhuys Publishers, Leiden. 344 p. Eshenroder, R. L., and M. K. Burnham-Curtis. 1999. Species succession and sustainability of the Great Lakes fish community. In Great Lakes Fishery Policy and Management: A Binational Perspective. Michigan State University Press. Eshenroder, R. L. 2008. Differentiation of deep-water lake charr Salvelinus namaycush in North American lakes. Environmental Biology of Fishes 83:77-90. Euclide, P. T., S. Hansson, and J. D. Stockwell. 2017. Partial diel vertical migration in an omnivorous macroinvertebrate, Mysis diluviana. Hydrobiologia 787:387-396. Evans, M. S., and D. W. Sell. 1985. Mesh size and collection characteristics of 50-cm diameter conical plankton nets. Hydrobiologia 122:97-104.

138

Gal, G., E. R. Loew, L. G. Rudstam, and M. A. Mohammadian. 1999. Light and diel vertical migration: spectral sensitivity and light avoidance by Mysis relicta. Canadian Journal of Fisheries and Aquatic Sciences 56:311-322. Gamble, A. E., T. R. Hrabik, J. D. Stockwell, and D. L. Yule. 2011a. Trophic connections in Lake Superior Part I: the offshore fish community. Journal of Great Lakes Research 37:541-549. Gamble, A. E., T. R. Hrabik, D. L. Yule, and J. D. Stockwell. 2011. Trophic connections in Lake Superior Part II: The nearshore fish community. Journal of Great Lakes Research 37:550-560. Gibbons, M. J. 1993. Vertical migration and feeding of Euphausia lucens at two 72 h stations in the southern Benguela upwelling region. Marine Biology 116:257-268. Grossnickle, N. E. 1979. Nocturnal feeding patterns of Mysis relicta in Lake Michigan, based on gut content fluorescence. Limnology and Oceanography 24:777-780. Grossnickle, N. E. 1982. Feeding habits of Mysis relicta - an overview. Hydrobiologia 93:101-107. Harrison, P. M., L. F. G. Gutowsky, E. G. Martins, D. A. Patterson, A. Leake, S. J. Cooke, and M. Power. 2013. Diel vertical migration of adult burbot: a dynamic trade-off among feeding opportunity, predation avoidance, and bioenergetic gain. Canadian Journal of Fisheries and Aquatic Sciences 70:1765-1774. Hondorp, D. W., S. A. Pothoven, and S. B. Brandt. 2005. Influence of Diporeia density on diet composition, relative abundance, and energy density of planktivorous fishes in southeast Lake Michigan. Transactions of the American Fisheries Society 134:588-601. Hrycik, A. R., P. W. Simonin, L. G. Rudstam, D. L. Parrish, B. Pientka, and T. B. Mihuc. 2015. Mysis zooplanktivory in Lake Champlain: A bioenergetics analysis. Journal of Great Lakes Research 41:492-501. Hutchinson, G. E. 1967. A Treatise on Limnology: II. Introduction to lake biology and their limnoplankton. John Wiley & Sons Publishing. Isaac, E. J., T. R. Hrabik, J. D. Stockwell, and A. E. Gamble. 2012. Prey selection by the Lake Superior fish community. Journal of Great Lakes Research 38:326-335. Langeland, A. 1988. Decreased zooplankton density in a mountain lake resulting from predation by recently introduced Mysis relicta. Verhandlungen des Internationalen Verein Limnologie 23: 419-429. Lasenby, D. C., and R. R. Langford. 1973. Feeding and assimilation of Mysis relicta. Limnology and Oceanography 18:280-285.

139

Levy, D. A. 1990. Reciprocal diel vertical migration behavior in planktivores and zooplankton in British Columbia lakes. Canadian Journal of Fisheries and Aquatic Sciences 47:1755-1764. Johannsson, O. E., L. G. Rudstam, and D. C. Lasenby. 1994. Mysis relicta: assessment of metalimnetic feeding and implications for competition with fish in Lakes Ontario and Michigan. Canadian Journal of Fisheries and Aquatic Sciences 51:2591-2602. Johannsson, O. E., M. F. Leggett, L. G. Rudstam, M. R. Servos, M. A. Mohammadian, G. Gal, R. M. Dermott, and R. H. Hesslein. 2001. Diet of Mysis relicta in Lake Ontario as revealed by stable isotope and gut content analysis. Canadian Journal of Fisheries and Aquatic Sciences 58:1975-1986. Johannsson, O. E., L. G. Rudstam, G. Gal, and E. Mills. 2003. Mysis relicta in Lake Ontario: population dynamics, trophic linkages and further questions. State of Lake Ontario (SOLO) - Past, present and future. Ecovision World Monograph Series:257-287. Knight, J. C., B. P. O'Malley, and J. D. Stockwell. 2018. Lake Champlain offshore benthic invertebrate community before and after zebra mussel invasion. Journal of Great Lakes Research 44:283-288. Macquart-Moulin, C., A. Bourdillon, F. Cubizolles, F. Passelaigue, and R. Rasoanarivo. 1987. Un cas type de migration verticale ‘retard’chez l'amphipode Ampelisca typica. Journal of Plankton Research 9:785-809. Mauchline, J. 1980. The biology of mysids and euphausiids. Academic Press, London, UK. Mehner, T., S. Busch, I. P. Helland, M. Emmrich, and J. Freyhof. 2010. Temperature‐ related nocturnal vertical segregation of coexisting coregonids. Ecology of Freshwater Fish 19:408-419. Mehner, T. 2012. Diel vertical migration of freshwater fishes – proximate triggers, ultimate causes and research perspectives. Freshwater Biology 57:1342-1359. Moen, V., and A. Langeland. 1989. Diurnal vertical and seasonal horizontal distribution patterns of Mysis relicta in a large Norwegian lake. Journal of Plankton Research 11:729-745. Morgan, M. D., and S. T. Threlkeld. 1982. Size dependent horizontal migration of Mysis relicta. Hydrobiologia 93:63-68. Murtaugh, P. A. 1984. Variable gut residence time: problems in inferring feeding rate from stomach fullness of a mysid crustacean. Canadian Journal of Fisheries and Aquatic Sciences 41:1287-1293.

140

Nakagawa, Y., Y. Endo, and H. Sugisaki. 2003. Feeding rhythm and vertical migration of the euphausiid Euphausia pacifica in coastal waters of north-eastern Japan during fall. Journal of Plankton Research 25:633-644. Neverman, D., and W. A. Wurtsbaugh. 1992. Visual feeding by juvenile Bear Lake sculpin. Transactions of the American Fisheries Society 121: 395-398. Neverman, D., and W. A. Wurtsbaugh. 1994. The thermoregulatory function of diel vertical migration for a juvenile fish, Cottus extensus. Oecologia 98:247-256. Nowicki, C. J., D. B. Bunnell, P. M. Armenio, D. M. Warner, H. A. Vanderploeg, J. F. Cavaletto, C. M. Mayer, and J. V. Adams. 2017. Biotic and abiotic factors influencing zooplankton vertical distribution in Lake Huron. Journal of Great Lakes Research 43:1044-1054. Ogonowski, M., J. Duberg, S. Hansson, and E. Gorokhova. 2013. Behavioral, ecological and genetic differentiation in an open environment—a study of a mysid population in the Baltic Sea. PloS one 8:e57210. Ohman, M. D. 1990. The demographic benefits of diel vertical migration by zooplankton. Ecological Monographs 60:257-281. Oksanen J., F. G. Blanchet, M. Friendly, R. Kindt, P. Legendre, D. McGlinn, P. R. Minchin, R. B. O'Hara, G. L. Simpson, P. Solymos, M. Stevens, E. Szoecs, and H. Wagner. 2018. vegan: Community Ecology Package. R package version 2.5-1. https://CRAN.R-project.org/package=vegan. O'Malley, B. P., and D. B. Bunnell. 2014. Diet of Mysis diluviana reveals seasonal patterns of omnivory and consumption of invasive species in offshore Lake Michigan. Journal of Plankton Research 36:989-1002. O’Malley, B. P., L. G. Rudstam, J. M. Watkins, T. J. Holda, and B. C. Weidel. 2017. Effects of food web changes on Mysis diluviana diet in Lake Ontario. Journal of Great Lakes Research 43:813-822. O'Malley, B. P., S. Hansson, and J. D. Stockwell. 2018. Evidence for a size-structured explanation of partial diel vertical migration in mysids. Journal of Plankton Research 40:66-76. Onsrud, M. S., and S. Kaartvedt. 1998. Diel vertical migration of the krill Meganyctiphanes norvegica in relation to physical environment, food and predators. Marine Ecology Progress Series 171:209-219. Parker, J. I. 1980. Predation by Mysis relicta on Pontoporeia hoyi: a food chain link of potential importance in the Great Lakes. Journal of Great Lakes Research 6:164- 166.

141

Parker, S. L., L. G. Rudstam, E. L. Mills, and D. W. Einhouse. 2001. Retention of Bythotrephes spines in the stomachs of eastern Lake Erie rainbow smelt. Transactions of the American Fisheries Society 130:988-994. Pearre, S. 1979. Problems of detection and interpretation of vertical migration. Journal of Plankton Research 1:29-44. Pearre, S. 2003. Eat and run? The hunger/satiation hypothesis in vertical migration: history, evidence and consequences. Biological Reviews 78:1-79. Ponomareva, L. A. 1971. Circadian migrations and feeding rhythms of some Indian Ocean euphausiid species. Oceanology 11:226–231. Ringelberg, J. 2010. Diel vertical migration of zooplankton in lakes and oceans. Springer. Robertson, A., C. F. Powers, and R. F. Anderson. 1968. Direct observations on Mysis relicta from a submarine. Limnology and Oceanography 13:700-702. Rudstam, L. G., K. Danielsson, S. Hansson, and S. Johansson. 1989. Diel vertical migration and feeding patterns of Mysis mixta (Crustacea, Mysidacea) in the Baltic Sea. Marine Biology 101:43-52. Rudstam, L. G., and O. E. Johannsson. 2009. Advances in the ecology of freshwater mysids. Aquatic Biology 5:246-248. Rybock, J. T. 1978. Mysis relicta Loven in Lake Tahoe: vertical distribution and nocturnal predation. PhD Dissertation. University of Nevada. Sierszen, M. E., J. R. Kelly, T. D. Corry, J. V. Scharold, P. M. Yurista, and G. Sprules. 2011. Benthic and pelagic contributions to Mysis nutrition across Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 68:1051-1063. Simmard, Y., G. Lacroix, and L. Legendre. 1986. Diel vertical migrations and nocturnal feeding of a dense coastal krill scattering layer (Thysanoessa raschi and Meganyctiphanes norvegica) in stratified surface waters. Marine Biology 91:93- 105. Sims, D. W., V. J. Wearmouth, E. J. Southall, J. M. Hill, P. Moore, K. Rawlinson, et al. 2006. Hunt warm, rest cool: bioenergetic strategy underlying diel vertical migration of a benthic shark. Journal of Animal Ecology 75:176-190. Stewart, T. J., and W. G. Sprules. 2011. Carbon-based balanced trophic structure and flows in the offshore Lake Ontario food web before (1987–1991) and after (2001– 2005) invasion-induced ecosystem change. Ecological Modelling 222:692-708. Takahashi, K. 2004. Feeding ecology of mysids in freshwater and coastal marine habitats: A review. Bulletin of Plankton Society of Japan 51:46-72.

142

Väinölä, R., B. J. Riddoch, R. D. Ward, and R. I. Jones. 1994. Genetic zoogeography of the Mysis relicta species group (Crustacea: Mysidacea) in northern Europe and North America. Canadian Journal of Fisheries and Aquatic Sciences 51:1490- 1505. Vanderploeg, H. A., J. A. Bowers, O. Chapelski, and H. K. Soo. 1982. Measuring in situ predation by Mysis relicta and observations on under-dispersed microdistributions of zooplankton. Hydrobiologia 93:109-119 (1982). Viherluoto, M., H. Kuosa, J. Flinkman, and M. Viitasalo. 2000. Food utilisation of pelagic mysids, Mysis mixta and M. relicta, during their growing season in the northern Baltic Sea. Marine Biology 136:553-559. Watkins, J. M., P. D. Collingsworth, N. E. Saavedra, B. P. O’Malley, and L. G. Rudstam. 2017. Fine-scale zooplankton diel vertical migration revealed by traditional net sampling and a Laser Optical Plankton Counter (LOPC) in Lake Ontario. Journal of Great Lakes Research 43:804-812. Whall, J. D., and D. C. Lasenby. 2009. Differences in the trophic role of Mysis diluviana in two intermontane lakes. Aquatic Biology 5: 281-292. Wurtsbaugh, W. A., and D. Neverman. 1988. Post-feeding thermotaxis and daily vertical migration in a larval fish. Nature 333:846-848.

143

Supplemental Table 4.3: Sample collection times of four separate trips. Sunset/sunrise times are listed below each date in parentheses. Grey shading denotes samples collected during night. Time refers to the midpoint during the time interval of each sample.

Date Benthic Pelagic (sunset; sunrise) 13/14 April 2017 18:05 (19:35; 06:10) 21:28 00:41 06:43 12:43 14/15 June 2016 14:58 (20:39; 05:08) 18:41 22:07 22:26 00:30 01:05 6:39 10:28 18/19 July 2017 19:43 (20:32; 05:26) 21:57 22:26 0:39 1:07 07:02 10:58 14:02 19/20 October 2017 17:27 (18:01; 07:15) 20:20 19:55 22:30 22:52 07:00 10:08 14:35

144

Supplemental Table 4.4: Zooplankton taxa, mean length (L in mm), and length-weight equations used to estimate dry biomass (W in dry µg). Equations follow the formula ln(W) = lnA + B • ln(L). Major Group Genus/species L (mm) lnA B Source Burgess et al. Calanoid Calanoid copepodid 0.81 1.59 2.59 (2015) Burgess et al. Epischura copepodid 1.05 1.59 2.59 (2015) Burgess et al. Epischura lacustris 1.71 1.59 2.59 (2015) Burgess et al. Leptodiaptomus minutus 0.90 1.59 2.59 (2015) Burgess et al. Leptodiaptomus sicilis 1.20 1.59 2.59 (2015) Burgess et al. Limnocalanus macrurus 2.24 1.59 2.59 (2015) Burgess et al. Senecella calanoides 2.59 1.59 2.59 (2015) Skistodiaptomus Burgess et al. oregonensis 1.18 1.59 2.59 (2015) Bottrell et al. Cladocera Bosmina longirostris 0.36 3.09 3.04 (1976) Bottrell et al. Ceriodaphnia spp. 0.41 2.562 3.34 (1976) Bottrell et al. Daphnia mendotae 1.08 1.468 2.83 (1976) Bottrell et al. Daphnia retrocurva 0.87 1.468 2.83 (1976) Bottrell et al. Eubosmina coregoni 0.44 3.09 3.04 (1976) Bottrell et al. Cyclopoid Acanthocyclops spp. 0.92 1.953 2.4 (1976) Bottrell et al. Cyclopoid copepodid 0.75 1.953 2.4 (1976) Bottrell et al. Diacyclops thomasi 1.00 1.953 2.4 (1976) Bottrell et al. Mesocyclops edax 0.80 1.953 2.4 (1976) Bottrell et al. Tropocyclops prasinus 0.57 1.953 2.4 (1976)

145

Supplemental Figure 4.10: Variation in the gut fullness ranks assigned to Mysis diets in pelagic and benthic habitats collected at different times over a diel cycle from 100-m site in Lake Champlain in June 2016 and April, July, and October 2017. No data were available from the pelagic habitat in April.

146

CHAPTER 5: COMPARATIVE EVALUATION OF MIGRANT AND NON-

MIGRANT MYSIS FEEDING PREFERENCES

Brian P. O’Malley1, 2, *, Jessica Griffin1, 3, and Jason D. Stockwell1

1 Rubenstein Ecosystem Science Laboratory, University of Vermont, Burlington, VT,

USA

2 Current address: USGS Great Lakes Science Center, Lake Ontario Biological Station,

Oswego, NY, USA

3 Current address: Department of Biology, San Diego State University, San Diego,

California, USA

*Corresponding author: [email protected]

147

Abstract

We conducted a feeding experiment to evaluate predictions of the trophic polymorphism hypothesis (TPH) for partially migrant Mysis. Specifically, we compared feeding rates in a Mysis population from Lake Champlain that exhibits partial diel vertical migration, where one portion of the population remains benthic (non-migrant) at night and the other migrates into pelagic (migrant) waters. Migrants and non-migrants experience different foraging arenas (benthic or pelagic) at night, and we hypothesized individuals are more efficient foragers in their habitat of choice. Experimental food treatments included a known preferred pelagic prey (Daphnia), a presumed less desirable benthic resource (detritus), and a combination of both. We tested the hypothesis that migrant (i.e., pelagic-caught) Mysis prefer Daphnia over detritus and therefore exhibit higher predation rates on Daphnia than non-migrant (i.e., benthic-caught) Mysis, which were hypothesized to prefer detritus over Daphnia, consistent with different diet optima that would reflect their observed migration behavior in the field. Consistent with assumptions of zooplankton as preferred prey opposed to detritus, we evaluated if

Daphnia presence would result in a lower amount of detritus consumed compared when only detritus was offered. Experimental results did not support the TPH for partial DVM.

Similar predation rates on Daphnia were observed among migrant and non-migrant

Mysis. However, contrary to our hypothesis, a significant effect of food treatment, but not habitat-origin, was observed with more Daphnia consumed with the inclusion of detritus in experimental arenas. Our findings highlight the omnivorous component of Mysis diet, and suggest Mysis readily consume detritus even in the presence of zooplankton. Overall,

148

results do not support the TPH to explain Mysis partial DVM as similar feeding rates were observed in migrant and non-migrant individuals.

Keywords: partial migration, detritus, trophic polymorphism hypothesis, omnivory

149

Introduction

Intraspecific niche variation within populations is common and has been documented for a wide range of animals, and can generally be predicted by three main effects: sexual dimorphism, ontogenetic shifts, and resource polymorphism

(Roughgarden 1972; Bolnick et al. 2003). In migratory species, the term partial migration is used to describe populations that exhibit variability in migration behavior among individuals. Specifically, partial migration describes instances when some individuals migrate while others remain resident (Chapman et al. 2011, 2012). Several hypotheses have been proposed to explain the ecological and evolutionary drivers of partial migration although much work still remains to evaluate partial migration hypotheses for a range of taxa and ecosystems. Most empirical evaluations of partial migration to date have focused on larger vertebrate fauna (e.g., birds, mammals, and fish; see Chapman et al. 2011). However, partial migration is also evident in invertebrate fauna, but has received less attention. The drivers and consequences of invertebrate partial migration remain largely unknown.

Mysids are a group of aquatic invertebrates found in lakes, estuaries, and oceans, and most mysid species exhibit diel vertical migration (DVM; Mauchline 1980; Jumars

2007). The typical form of mysid DVM is described as a population-level movement where individuals vertically migrate from deep-water habitat in the day, to shallower sub- surface waters at night (Beeton 1960; Haskell and Stanford 2006; Rudstam 2009). DVM in lower trophic level organisms including mysids has generally been accepted as a predator-avoidance strategy, in which animals use DVM to avoid high predation risk from visual predators during the day, when light intensity is high, by moving downward

150

to darker habitat. They then migrate upwards at night to capitalize on feeding and high growth opportunities at shallower depths under darkness when risk is lower (Lampert

1993; Hays 2003; Boscarino et al. 2009; Ringelberg 2010). Mysids are omnivores that consume pelagic phyto- and zooplankton at night during their ascent portion of DVM

(Grossnickle 1982; Takahashi 2004). In contrast to strictly planktonic organisms, mysids can also feed on sedimented detritus in benthic habitat, which they presumably access during the day (Lasenby and Langford 1973; Lasenby and Shi 2004; Sierszen et al.

2011). Morphological and behavioral adaptations to feed on detritus have been described for some mysid taxa, in addition to filter or raptorial feeding behavior used to ingest planktonic organisms in the water column (DeGraeve and Reynolds 1975; Grossnickle

1982; Ramcharan and Sprules 1986). For example, Hemimysis larmonae uses a modified version of filter feeding in which the animal orients its head down at the sediment surface to lift surficial detritus into suspension with their thoracic appendages (Cannon and

Manton 1927). Consequently, mysids demonstrate multiple forms of feeding within species.

In members of the Mysis relicta group (M. diluviana, M. relicta, and M. salemaai;

Audzijonytė and Väinölä 2005), partial DVM occurs within populations in freshwater

(Euclide et al. 2017; O’Malley et al. 2018) and marine systems (Rudstam et al. 1989;

Ogonowski et al. 2013), where some individuals remain benthic at night while others vertically migrate. Partial DVM in mysids goes against conventional DVM theory because individuals that remain on bottom forego access to presumably high concentrations of better-quality prey despite lower predation risk in surface waters at night compared to day. Given that mysids occupy both pelagic and benthic habitats at

151

night, and can feed on resources from both (Johannsson et al. 2003; Lasenby and Shi

2004; Sierszen et al. 2011), one potential hypothesis for Mysis partial DVM is individuals exhibit different feeding preferences within a population. The trophic polymorphism hypothesis (TPH) for partial migration (Chapman et al. 2011) in mysids predicts that some individuals are specialized feeders on either pelagic plankton or benthic resources, consistent with their migration behavior. Support for TPH has been identified in some partially migrant fish species that exhibit specialization on different food types (Svanback et al. 2008; Chapman et al. 2012). In North American lakes, where M. diluviana has largely been described as a predator on zooplankton (Cooper and Goldman 1980; Bowers and Vanderploeg 1982; Hrycik et al. 2015), with a preference for cladocerans over other prey (Grossnickle 1982; O’Malley and Bunnell 2014), the importance or preference of sedimented detritus as a food resource for mysids that remain on the bottom (non- migrants) of the lake at night is unknown. Feeding behavior of M. diluviana on detritus has largely been undescribed to date because most of the focus on their ecological role in food webs has been on pelagic behavior and competitive implications of introductions in non-native environments for fisheries management (Lasenby et al. 1986; Spencer et al.

1991; Ellis et al. 2011).

In this study, we used an experimental approach to evaluate predictions of the

TPH as a potential mechanism to explain partial DVM in a Mysis population. We hypothesized that migrant Mysis found in pelagic waters at night would be more efficient predators and exhibit greater preference for pelagic zooplankton prey compared to non- migrant benthic Mysis. Similarly, we hypothesized that non-migrant benthic Mysis would exhibit greater preference for detritus over Daphnia compared to migrant Mysis when

152

both food types are present, and therefore the presence of detritus would release Daphnia from predation by non-migrant Mysis under experimental conditions. Given mysid preference for Daphnia, we expected that both groups of mysids would feed less on detritus in the presence of Daphnia relative to scenarios when detritus was the only food option.

Methods

Feeding experiments were performed at the Rubenstein Ecosystem Science

Laboratory with M. diluviana (hereafter Mysis) from Lake Champlain. The experimental approach followed a common-garden style experiment to compare behavior of migrant and non-migrant Mysis exposed to the same conditions. Because Mysis are sensitive to light, and even brief exposure to bright lights can damage their eyes (Feldmann et al.

2010), all field collections and experimental procedures that involved Mysis were performed under darkness with limited exposure to only dim red light. Each treatment level used one individual mysid per container as a replicate because mysids feed less when more than one individual is held in experimental containers (Hamrén and Hansson

1999) and they can be cannabilistic (Quirt and Lasenby 2002; Nordin et al. 2008).

Field sampling

Live benthic- and pelagic-caught mysids were collected at night on 11 Oct 2016 from a 100-m site in Lake Champlain (44° 28’ N; 73° 18’ W). Pelagic Mysis were collected with a 1-m2 square framed net towed vertically from 80 m depth to surface.

Benthic Mysis were collected using a benthic sled towed for 5 min along bottom. Upon retrieval of each net or sled tow, organisms were concentrated to the cod-end and

153

immediately submerged into separate coolers filled with chilled lake water. Sufficient numbers of individuals were collected in a single 5-min benthic sled tow, whereas multiple replicate vertical tows were required to collect sufficient numbers of pelagic individuals. Coolers were transported back to the laboratory within 2 hours of collection and placed in a dark walk-in refrigerator maintained at 5°C.

Sedimented lake detritus, used as the benthic food resource for the experiment, was obtained by scraping the top 1-2 cm of surface sediment from Ponar grab samples.

Ponar samples were collected from a nearby 60-m site during sunset to prevent disturbing benthic Mysis at the 100-m site. Previous work has shown detritus organic matter content and benthic invertebrate communities are similar between 60 and 100 m depths in Lake

Champlain (Knight et al. 2018; J. C. Knight unpublished data). All detritus was combined in a single jar, and kept on ice in a dark cooler until transport to the walk-in refrigerator upon return from field sampling.

Experimental setup

The experiment comprised 60 1-L glass jars as foraging arenas for individual

Mysis (Figure 5.1). All jars were setup on a table in a walk-in refrigerator (held at 5°C) with replicates for each treatment level arranged in a mixed pattern. Each jar was filled with 600 mL of chilled dechlorinated water obtained from the municipal water supply sourced from Lake Champlain. Ten individuals of each habitat-origin (i.e., benthic- or pelagic-caught) were exposed to one of three food treatment levels, for a total of 30 individuals per habitat-origin. The three food treatment levels included Daphnia-only, detritus-only, and Daphnia plus detritus (Daphnia-detritus). The scraped detritus samples were homogenized in a slurry with a small amount of dechlorinated water, then 20 grams 154

(wet weight) of the slurry was placed in each of the 1-L jars with 600 mL of dechlorinated water for the detritus-only and Daphnia-detritus treatments. The slurry was added just after arrival to the lab following Mysis collections to allow maximum time to settle on the bottom of jars before the experiment commenced.

Figure 5.1: Schematic of Mysis feeding experiment. For each food treatment (Daphnia- only, Daphnia-detritus, and Detritus-only), a single Mysis was placed in each of 10 1-L jars and allowed to forage on the available food over the allotted time under darkness.

A lab culture of Daphnia magna was used for the Daphnia-only and the Daphnia- detritus treatments. Mysis select for Daphnia over other zooplankton and Daphnia represents a major proportion of the diets of pelagic-caught Mysis in summer-fall months

(Grossnickle 1982; O’Malley and Bunnell 2014; O’Malley et al. 2017). D. magna can reach larger body size than the primary native Daphnia spp. (D. mendotae, and D. retrocurva) that are available to Mysis in Lake Champlain (Balcer et al. 1984; Mihuc et al. 2012). However, even large-sized D. magna still fall within the size range of prey consumed by adult Mysis in lakes (e.g., Bythotrephes – Nordin et al. 2008). We used 30

Daphnia per replicate for each treatment level that included Daphnia. A pipette was used

155

to transfer groups of 30 Daphnia to individual 20 mL scintillation vials filled with dechlorinated water until the start of the experiment. Daphnia vials were moved into the walk-in refrigerator 1 hr before the start of the experiment to acclimate the Daphnia before introducing them to each foraging arena.

After an initial 12 hours of acclimation in the original coolers from field collections in the walk-in refrigerator, large Mysis (> 10 mm) were individually transferred, using a table spoon, into separate 100-ml jars filled with chilled dechlorinated tap water to evacuate their gut contents. Only large Mysis were used to ensure all individuals tested could feed effectively on D. magna (Ramcharan and Sprules 1986).

After 8 hrs, mysids from 100-mL jars were checked to see if they were alive and swimming, then placed into respective 1-L foraging arenas along with the 100 mL of water resulting in a final volume of 700 mL of water per foraging arena for the experiment. Scintillation vials with Daphnia were added at the same time for treatments with Daphnia. For the treatments with Daphnia, the final volume was estimated to be

720-725 mL because of the water added from the scintillation vials and two rinses with a squirt bottle to ensure all Daphnia were transferred from the vial. Additional control foraging arenas with Daphnia-detritus treatments, but no Mysis, were used to assess our ability to recover Daphnia from Daphnia-detritus trials (N = 8). The experiment ran for

11 hours overnight in darkness in the walk-in refrigerator, from approximately sunset to sunrise.

At the end of the experiment, jars were narcotized with alka-seltzer tablets to end feeding. Next, remaining Daphnia and Mysis from each foraging arena were passed through a 150 µm sieve and preserved in vials with ethanol for counting. Daphnia were

156

counted and Mysis were measured for body length using a dissecting microscope equipped with a drawing attachment and digitizing tablet for length measurements.

Missing Daphnia were assumed to have been consumed by Mysis. Stomach contents of

Mysis from detritus-only and Daphnia-detritus treatments were dissected and assessed for detritus under an inverted microscope. The relative amount of detritus in stomach contents was scored following a semi-quantitative scale (O’Malley and Stockwell in review) that ranges from 0 (low) to 4 (high amount).

Data analysis

Ingestion rates of Daphnia prey were compared using a two-way ANOVA with habitat-origin (benthic- and pelagic-caught) and treatment level (Daphnia-only and

Daphnia-detritus) as factors, followed by Tukey’s HSD comparison if a significant difference was detected. For detritus consumption we used a Wilcoxon signed rank analysis to compare detritus scores and habitat-origin treatment levels (detritus-only and

Daphnia-detritus). Because body size of Mysis was not measured prior to the start of the experiment, and only measured post-mortem, we evaluated potential for body size to influence ingestion rates within the dataset by linear regression of ingestion rate as a function of body size. All statistical analyses were carried out in R (R Core Team 2017).

Results

Mysis body sizes ranged from 11.5 to 22.2 mm, the mean length was similar across all treatments (F5,53 = 2.03, P > 0.05, Table 5.1). One Mysis that originated from pelagic habitat and was used in the detritus-only treatment type could not be located at the end of the experiment and the replicate was disregarded from further analysis. Out of

157

the 8 jars used as controls to assess recovery efficiency of Daphnia in the Daphnia- detritus treatment, 7 jars yielded full recovery of 30 Daphnia per jar, and 1 jar yielded 28

Daphnia.

Table 5.1: Mean length of Mysis used in feeding experiment.

Origin Treatment N Mean length (mm ± SD) Benthic Daphnia-only 10 17.6 ± 2.0 Daphnia-detritus 10 17.1 ± 1.3 Detritus-only 10 16.8 ± 1.8 Pelagic Daphnia-only 10 17.9 ± 2.8 Daphnia-detritus 10 16.2 ± 2.7 Detritus-only 9 19.2 ± 2.4

Mean ingestion rate of Daphnia by Mysis (Daphnia hr -1) was significantly affected by feeding treatment type (two-way ANOVA, F1,36 = 13.40, P < 0.01, Table 5.2) but not by habitat-origin (benthic- or pelagic-caught) from which the Mysis were collected (F1,36 = 1.67, P > 0.05). No significant interaction was evident between habitat- origin and Daphnia treatment level (F1,36 = 0.83, P > 0.05). Post-hoc comparison of feeding treatments showed mean ingestion rate (Daphnia hr -1) in the Daphnia-only treatment (N = 20, mean = 0.39 ± 0.08 SE) was significantly lower than in the Daphnia- detritus treatment (P < 0.001, N = 20, mean = 0.85 ± 0.09). Ingestion rates were not significantly influenced by Mysis body size over the range of animal sized used in the experiment (Figure 5.2, P > 0.05).

Comparison of detritus scores in Mysis stomach contents of individuals exposed to the detritus-only treatment revealed similar scores between benthic- (mean = 2.4 ± 0.4

SE) and pelagic-caught Mysis (3.0 ± 0.5, Wilcoxon signed-rank test, W = 33, P > 0.05).

Similarly, in the Daphnia-detritus treatment, detritus scores of stomach contents between 158

benthic- (2.4 ± 0.5) and pelagic-caught Mysis (1.9 ± 0.4) were not significantly different

(W = 60.5, P > 0.05).

Table 5.2: Two-way ANOVA results of Mysis predation rates (No. Daphnia hr-1).

Factor d.f. Sum-of-Squares F-value P-value habitat-origin 1 0.27 1.67 0.21 treatment 1 2.15 13.40 < 0.01 habitat-origin: treatment 1 0.01 0.05 0.83 error 36 5.78

Figure 5.2: Observed ingestion rates of Mysis on Daphnia as a function of Mysis body size.

Discussion

We tested predictions of the TPH for partial DVM in a Lake Champlain Mysis population using an experimental approach. A substantial body of literature on pelagic foraging behavior of Mysis has emphasized the importance of pelagic resources to Mysis, although relatively much less is known about the role of benthic feeding. To our knowledge, studies have yet to assess the relative preference for benthic detritus over

159

other food types available to partially migrant mysids that exhibit different behavior, and thus have different temporal exposure to benthic or pelagic prey dependent on their migration behavior. Experimental results confirmed the assumption that Mysis will eat detritus, and contrary to a common expectation that when offered a choice between

Daphnia or detritus, Mysis continue to feed on detritus even in the presence of their supposedly preferred prey Daphnia. Overall, the use of detritus as a resource by Mysis when Daphnia is available supports studies that have suggested Mysis are omnivorous

(Grossnickle 1982; Johannsson et al. 2003; O’Malley and Bunnell 2014), as opposed to obligate carnivores, and agrees with recent work that suggests benthic feeding may represent a significant contribution to their diet (Sierszen et al. 2011; O’Malley and

Stockwell in review). Our results did not support our original hypotheses based on predictions of the TPH that benthic, non-migrant mysids would feed more extensively on detritus, and pelagic, migrant Mysis would feed more on Daphnia. Similarly, alterations to feeding treatments by presenting both groups of Mysis with both detritus and Daphnia did not release Daphnia from predation in either case.

Increased ingestion rate of Daphnia in the presence of detritus may suggest that the benthic detritus creates a preferred feeding environment for mysids, in which a mysid more likely to feed or sediment leads to increased feeding efficiency. Survival of

Neomysis intermedia fed plankton in experimental chambers increased when mud or sand was placed on the bottom to provide a substrate (Murano 1966). However, multiple studies on predation rates of M. diluviana on zooplankton prey have been performed in the absence of substrate (e.g., Cooper and Goldman 1980, 1982; Ramcharan and Sprules

1986; Rudstam et al. 1999; Chipps and Bennett 2002), suggesting mysids may still feed

160

at comparable rates in bare experimental containers compared to in situ rates (Bowers and Vanderploeg 1982). Impacts of detritus presence in experimental chambers on survival or performance of M. diluviana have not been examined. Alternatively, the sediment may provide a staging zone in small foraging arenas from which M. diluviana can attack prey more effectively. Mysis diluviana has been observed to launch off the sediment by reflexing their abdominal segment against the substrate, which results in a greater acceleration speed than observed in the water column (Bowers et al. 1990). To avoid confounding effects of substrate on predation rates, additional control experiments possibly with inorganic or synthetic substrate with similar composition to sedimented detritus could be used to test for effects of detritus on zooplanktivory rates in mysids.

Our feeding experiment was run in darkness, and could have could have affected the ability to detect an effect of habitat origin on ingestion rates if migratory pelagic

Mysis are better adapted to use light to seek out prey than non-migrant Mysis. Mysis have relatively large compound eyes likely used to perceive prey under low light conditions

(Ringelberg 2010). Previous experimental work showed Mysis exhibit increased predation rates on Daphnia in the presence of low light levels compared to complete darkness (Ramcharan and Sprules 1986). Additional experiments that incorporate effects of low light levels typically experienced by Mysis during DVM into the upper hypo- and lower metalimnion at night (Beeton 1960; Lehman et al. 1990; Gal et al. 1999; Boscarino et al. 2009) are needed to explore possible interactions between light and food type on the usefulness of TPH to explain Mysis partial DVM.

Stable isotope and stomach content analyses of field-collected individuals could provide further evaluation of the TPH in situ. Our study focused on an experiment to

161

evaluate rates and preferences among individuals, whereas stable isotopes can provide a long-term picture of feeding history if benthic and pelagic resources vary in isotopic signatures. In the Baltic Sea, Ogonowski et al. (2013) found significant differences in carbon (δ13C) and nitrogen (δ15N) isotope ratios in mysids from benthic and pelagic habitat, suggesting apparent differences in resource use between groups. In Lake

Champlain, previous investigation however did not find a significant difference in δ13C or

δ15N of migrant and non-migrant Mysis (Euclide et al. 2017). Collections in the Lake

Champlain study were limited to the month of November, a time of year when thermal stratification breaks down and the water column becomes well-mixed compared to summer periods when thermal gradients limit mixing. One might expect substantial differences between pelagic and benthic stable isotope baselines during summer when thermal gradients are stronger. The application of stable isotope ratios to discriminate trophic differences, at least for δ15N, however, can also be limited because δ15N and trophic level can be independent of each other (Flynn et al. 2018). Nitrogen isotope ratios

(δ15N) of benthic material generally become enriched with depth as sinking detritus particles undergo bacterial colonization, while δ15N of planktonic material remains constant across depths (Sierszen et al. 2006; 2011), biasing detection of trophic differences if not accounted for in models. Gut content data provides a snapshot of recently ingested food, and comparisons of benthic- and pelagic-caught mysids in Lake

Champlain indicated higher amounts of detritus in diets of benthic-caught mysids, although diet differences between groups were not as strong as expected if migrants and non-migrants exclusively fed on different resource pools (O’Malley and Stockwell in review). Moreover, both detritus and plankton occur in migrant and non-migrant gut

162

contents (O’Malley and Stockwell in review), which agrees with our experimental results and suggests only limited evidence for trophic differences to explain partial DVM in mysids.

We recognize that our classification of migrant and non-migrant mysids relies only on the habitat from which the individuals were captured, and may not be representative of long-term behavior of those individuals if migration behavior varies day-to-day among individuals. Current limitations in tracking technology for small organisms such as mysids that undergo large vertical changes in depth precludes inference on whether mysids collected in from a particular habitat at night is indicative of long-term migration behavior. Developments and advances in such technologies will undoubtedly benefit the partial migration studies of small aquatic invertebrates including mysids. Nonetheless, experiments and field studies that compare traits of individuals from different habitats in the interim of tracking advances, will provide a foundation for further hypotheses, in addition to insights on food web implications of partial DVM.

In conclusion, results from our study do not suggest support for differences in feeding preference to explain partial DVM in Mysis from Lake Champlain, nor do previous field observations from the same system (Euclide et al. 2017; O’Malley and

Stockwell in review). Although support for the TPH in partially migrant mysids from other systems has been identified (Ogonowski et al. 2013), the lack of support in Lake

Champlain suggests that drivers of mysid partial DVM could vary across systems or among species. While previous work in Lake Champlain has identified body size differences among migrant and non-migrants (Euclide et al. 2017; O’Malley et al. 2018), we controlled for this factor by only using Mysis of similar size in our experiment to limit

163

ontogenetic effects of feeding (Werner and Gilliam 1984), and to maintain focus on variation in feeding of mysids of similar body size. The lack of difference in Mysis of different habitat origins, and the significant effect of feeding treatment when detritus was introduced, does call into question whether Mysis may prefer to be benthic, perhaps by taking advantage of a more favorable environment when adequate food supply is available to attack prey or one with less predation risk. However, the potential for substrate effects to entice Mysis to stay benthic and not migrate into the pelagic environment requires further evaluation.

Acknowledgements

We thank Steve Cluett and Brad Roy for help with field collections and Hannah

Lister, and Sture Hansson for help executing the experiment. Funding was provided by the Great Lakes Fishery Commission with the assistance of Senator Patrick Leahy, a

University of Vermont Chrysalis Award, and a scholarship from the International

Association for Great Lakes Research to BPO. Partial support was also provided by the

National Science Foundation Research Experiences for Undergraduates (REU) Program

(DBI-1358838).

164

References Audzijonytė, A., Väinölä, R., 2005. Diversity and distributions of circumpolar fresh-and brackish-water Mysis (Crustacea: Mysida): descriptions of M. relicta Lovén, 1862, M. salemaai n. sp., M. segerstralei n. sp. and M. diluviana n. sp., based on molecular and morphological characters. Hydrobiol. 544, 89-141. Balcer, M.D., Korda, N.L., Dodson, S.I., 1984. Zooplankton of the Great Lakes: a guide to the identification and ecology of the common crustacean species. Univ. of Wisconsin Press. Beeton, A.M., 1960. The vertical migration of Mysis relicta in Lakes Huron and Michigan. J. Fish. Res. Board Can. 17, 517-539. Bolnick, D.I., Svanbäck, R., Fordyce, J.A., Yang, L.H., Davis, J.M., Hulsey, C.D., Forister, M.L., 2003. The ecology of individuals: incidence and implications of individual specialization. Am. Nat. 161, 1-28. Boscarino, B.T., Rudstam, L.G., Eillenberger, J.J., O’Gorman, R., 2009. Importance of light, temperature, zooplankton and fish in predicting the nighttime vertical distribution of Mysis diluviana. Aquat. Biol. 5, 263-279. Bowers, J.A., Vanderploeg, H.A., 1982. In situ predatory behavior of Mysis relicta in Lake Michigan. Hydrobiol. 93, 121-131. Bowers, J.A., Cooper, W.E., Hall, D.J., 1990. Midwater and epibenthic behaviors of Mysis relicta Loven: Observations from the Johnson-Sea-Link II submersible in Lake Superior and from a remotely operated vehicle in northern Lake Michigan. J. Plankton Res. 12, 1279-1286. Cannon, H.G., Manton, S.M., 1927. On the feeding mechanism of a mysid crustacean, Hemimysis lamornae. Trans. R. Soc. Edin. 55, 219-253. Chapman, B.B., Brönmark, C., Nilsson, J.Å., Hansson, L.A., 2011. The ecology and evolution of partial migration. Oikos 120, 1764-1775. Chapman, B.B., Hulthén, K., Brodersen, J., Nilsson, P.A., Skov, C., Hansson, L.A., Brönmark, C., 2012. Partial migration in fishes: causes and consequences. J. Fish Biol. 81, 456-478. Chipps, S.R., Bennett, D.H., 2002. Evaluation of a Mysis bioenergetics model. J. Plankton Res. 24, 77-82. Cooper, S.D., Goldman, C.R., 1980. Opossum shrimp (Mysis relicta) predation on zooplankton. Can. J. Fish. Aquat. Sci. 37, 909-919. Cooper, S.D., Goldman, C.R., 1982. Environmental factors affecting predation rates of Mysis relicta. Can. J. Fish. Aquat. Sci 39, 203-208. DeGraeve, G.M., Reynolds, J.B., 1975. Feeding behavior and temperature and light tolerance of Mysis relicta in the laboratory. Trans. Am. Fish. Soc. 104, 394-397.

165

Ellis, B.K., Stanford, J.A., Goodman, D., Stafford, C.P., Gustafson, D.L., Beauchamp, D.A., Chess, D.W. et al., 2011. Long-term effects of a trophic cascade in a large lake ecosystem. Proc. Natl. Acad. Sci. 108, 1070-1075. Euclide, P.T., Hansson, S., Stockwell, J.D., 2017. Partial diel vertical migration in an omnivorous macroinvertebrate, Mysis diluviana. Hydrobiol. 787, 387-396. Feldmann, T., Yakovleva, M., Lindström, M., Donner, K., Ostrovsky, M., 2010. Eye adaptation to different light environments in two populations of Mysis relicta: a comparative study of carotenoids and retinoids. J. Crustacean Biol. 30, 636-642. Flynn, K.J., Mitra, A., Bode, A., 2018. Toward a mechanistic understanding of trophic structure: inferences from simulating stable isotope ratios. Mar. Biol. 165, 147. Gal, G., Loew, E.R., Rudstam, L.G., Mohammadian, M.A., 1999. Light and diel vertical migration: spectral sensitivity and light avoidance by Mysis relicta. Can. J. Fish. Aquat. Sci. 56, 311-322. Grossnickle, N. E., 1982. Feeding habits of Mysis relicta- an overview. Hydrobiol. 93, 101-107. Hamrén, U., Hansson, S., 1999. A mysid shrimp (Mysis mixta) is able to detect the odour of its predator (Clupea harengus). Ophelia 51, 187-191. Haskell, C.A., Stanford, J.A., 2006. Ecology of an estuarine mysid shrimp in the Columbia River (USA). River Res. Appl. 22, 739-753. Hays, G.C., 2003. A review of the adaptive significance and ecosystem consequences of zooplankton diel vertical migrations. Hydrobiol. 503, 163-170. Hrycik, A.R., Simonin, P.W., Rudstam, L.G., Parrish, D.L., Pientka, B., Mihuc, T.B., 2015. Mysis zooplanktivory in Lake Champlain: A bioenergetics analysis. J. Great Lakes Res. 41, 492-501. Johannsson, O.E., Rudstam, L.G., Gal, G., Mills, E.L., 2003. Mysis relicta in Lake Ontario: population dynamics, trophic linkages and further questions, in: Munawar, M. (Ed.), State of Lake Ontario (SOLO) - Past, present and future. Backhuys Publishers, Leiden, The Netherlands, pp. 257-287. Jumars, P.A., 2007. Habitat coupling by mid-latitude, subtidal, marine mysids: import- subsidised omnivores. in: Gibson, R., Atkinson, R., Gordon, J. (Eds.), Oceanography and Marine Biology: An Annual Review, Volume 45. CRC Press, pp. 89-138. Knight, J.C., O'Malley, B.P., Stockwell, J.D., 2018. Lake Champlain offshore benthic invertebrate community before and after zebra mussel invasion. J. Great Lakes Res. 44, 283-288. Lampert, W., 1993. Ultimate causes of diel vertical migration of zooplankton: new evidence for the predator-avoidance hypothesis. Archiv. Hydrobiol. –Monogr. Beit. 39, 79-88.

166

Lasenby, D.C., Langford, R.R., 1973. Feeding and assimilation of Mysis relicta. Limnol. Oceanogr. 18, 280-285. Lasenby, D.C., Northcote, T.G., Fürst, M., 1986. Theory, practice, and effects of Mysis relicta introductions to North American and Scandinavian lakes. Can. J. Fish. Aquat. Sci. 43, 1277-1284. Lasenby, D., Shi, Y., 2004. Changes in the elemental composition of the stomach contents of the opossum shrimp Mysis relicta during diel vertical migration. Can. J. Zool. 82, 525-528. Lehman, J.T., Bowers, J.A., Gensemer, R.W., Warren, G.J., Branstrator, D.K., 1990. Mysis relicta in Lake Michigan: abundances and relationships with their potential prey, Daphnia. Can. J. Fish. Aquat. Sci. 47, 977-983. Mauchline, J., 1980. The biology of mysids and euphausiids. Academic Press, London, UK. Mihuc, T.B., Dunlap, F., Binggeli, C., Myers, L., Pershyn, C., Groves, A., Waring, A., 2012. Long-term patterns in Lake Champlain's zooplankton: 1992–2010. J. Great Lakes Res. 38, 49-57. Murano, M., 1966. Culture of Isaza-Ami Neomysis intermedia. Info. Bull. Planktonology Japan 13, 65-68. Nordin, L.J., Arts, M.T., Johannsson, O.E., Taylor, W.D., 2008. An evaluation of the diet of Mysis relicta using gut contents and fatty acid profiles in lakes with and without the invader Bythotrephes longimanus (Onychopoda, Cercopagidae). Aquat. Ecol. 42, 421-436. Ogonowski, M., Duberg, J., Hansson, S., Gorokhova, E., 2013. Behavioral, ecological and genetic differentiation in an open environment—a study of a mysid population in the Baltic Sea. PloS one 8, e57210. O'Malley, B.P., Bunnell, D.B., 2014. Diet of Mysis diluviana reveals seasonal patterns of omnivory and consumption of invasive species offshore Lake Michigan. J. Plankton Res. 36, 989-1002. O’Malley, B.P., Rudstam, L.G., Watkins, J.M., Holda, T.J., Weidel, B.C., 2017. Effects of food web changes on Mysis diluviana diet in Lake Ontario. J. Great Lakes Res. 43, 813-822. O'Malley, B.P., Hansson, S., Stockwell, J.D., 2018. Evidence for a size-structured explanation of partial diel vertical migration in mysids. J. Plankton Res. 40, 66- 76. O’Malley, B.P., Stockwell, J.D., In review. Diel feeding behavior in a partially migrant Mysis population: A benthic-pelagic comparison. Food Webs. Quirt, J., Lasenby, D., 2002. Cannibalism and ontogenetic changes in the response of the freshwater shrimp Mysis relicta to chemical cues from conspecific predators. Can. J. Zool. 80, 1022-1025.

167

R Core Team., 2017. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R- project.org/. Ramcharan, C.W., Sprules, W.G., 1986. Visual predation in Mysis relicta Lovén. Limnol. Oceanogr. 31, 414-420. Ringelberg, J., 2010. Diel vertical migration of zooplankton in lakes and oceans. Springer. Roughgarden, J., 1972. Evolution of niche width. Am. Nat. 106, 683-718. Rudstam, L., Danielsson, K., Hansson, S., Johansson, S., 1989. Diel vertical migration and feeding patterns of Mysis mixta (Crustacea, Mysidacea) in the Baltic Sea. Mar. Biol. 101, 43-52. Rudstam, L.G., Hetherington, A.L., Mohammadian, A.M., 1999. Effect of temperature on feeding and survival of Mysis relicta. J. Great Lakes Res. 25, 363-371. Rudstam, L.G., 2009. Other Zooplankton, in: Likens, G.E. (Ed.) Encyclopedia of Inland Waters, Volume 3. Oxford: Elsevier, pp. 667-677. Sierszen, M.E., Peterson, G.S., Scharold, J.V., 2006. Depth-specific patterns in benthic planktonic food web relationships in Lake Superior. Can. J. Fish. Aquat. Sci. 63, 1496-1503. Sierszen, M.E., Kelly, J.R., Corry, T.D., Scharold, J.V., Yurista, P.M., 2011. Benthic and pelagic contributions to Mysis nutrition across Lake Superior. Can. J. Fish. Aquat. Sci. 68, 1051-1063. Spencer, C.N., McClelland, B.R., Stanford, J.A., 1991. Shrimp stocking, salmon collapse, and eagle displacement. BioScience 41, 14-21. Svanbäck, R., Eklöv, P., Fransson, R., Holmgren, K., 2008. Intraspecific competition drives multiple species resource polymorphism in fish communities. Oikos 117:114-124. Takahashi, K., 2004. Feeding ecology of mysids in freshwater and coastal marine habitats: A review. Bull. Plankton Soc. Japan 51, 46-72. Werner, E.E., Gilliam, J.F., 1984. The ontogenetic niche and species interactions in size- structured populations. Annu. Rev. Ecol. Syst. 15, 393-425.

168

COMPREHENSIVE REFERENCES

Ahrens, R. N., C. J. Walters, and V. Christensen. 2012. Foraging arena theory. Fish and Fisheries 13:41-59. Ahrenstorff, T. D., T.R. Hrabik, J. D. Stockwell, D. L. Yule, and G. G. Sass. 2011. Seasonally dynamic diel vertical migrations of Mysis diluviana, coregonine fishes, and siscowet lake trout in the pelagia of western Lake Superior. Transactions of the American Fisheries Society 140:1504-1520. Araújo, M. S., D. I. Bolnick, and C. A. Layman. 2011. The ecological causes of individual specialisation. Ecology Letters 14:948-958. Audzijonytė, A., and R. Väinölä. 2005. Diversity and distributions of circumpolar fresh- and brackish-water Mysis (Crustacea: Mysida): descriptions of M. relicta Lovén, 1862, M. salemaai n. sp., M. segerstralei n. sp. and M. diluviana n. sp., based on molecular and morphological characters. Hydrobiologia 544:89-141. Baker, J. E., S. J. Eisenreich, and B. J. Eadie. 1991. Sediment trap fluxes and benthic recycling of organic carbon, polycyclic aromatic hydrocarbons, and polychlorobiphenyl congeners in Lake Superior. Environmental Science and Technology 25:500-509. Balcer, M. D., N. L. Korda, and S. I. Dodson. 1984. Zooplankton of the Great Lakes: A guide to the identification and ecology of the common crustacean species. University of Wisconsin Press. Ball, S. C., T. B. Mihuc, L. W. Myers, and J. D. Stockwell. 2015. Ten-fold decline in Mysis diluviana in Lake Champlain between 1975 and 2012. Journal of Great Lakes Research 41:502-509. Bauer, S. and B. J. Hoye. 2014. Migratory animals couple biodiversity and ecosystem functioning worldwide. Science 344:1242552. Beeton, A. M. 1960. The vertical migration of Mysis relicta in Lakes Huron and Michigan. Journal of the Fisheries Board of Canada 17:517-539. Bergersen, E. P., and M. Maiolie. 1981. Mysis population estimates: a photographic technique. Applied Sciences Branch, Division of Research, Engineering and Research Center, US Department of the Interior, Bureau of Reclamation. Betini, G. S., T. Avgar, and J. M. Fryxell. 2017. Why are we not evaluating multiple competing hypotheses in ecology and evolution? Royal Society Open Science 4:160756. Bianchi, D., and K. Mislan. 2016. Global patterns of diel vertical migration times and velocities from acoustic data. Limnology and Oceanography 61: 353-364. Bicknell, A. W., B. J. Godley, E. V. Sheehan, S. C. Votier, and M. J. Witt. 2016. Camera technology for monitoring marine biodiversity and human impact. Frontiers in Ecology and the Environment 14: 424-432.

169

Bollens, S. M. and B. W. Frost. 1989. Zooplanktivorous fish and variable diel vertical migration in the marine planktonic copepod Calanus pacificus. Limnology and Oceanography 34:1072-1083. Bollens, S. M., B. W. Frost, D. S. Thoreson, and S. J. Watts. 1992. Diel vertical migration in zooplankton: field evidence in support of the predator avoidance hypothesis. Hydrobiologia 234:33-39. Bollens, S. M., G. Rollwagen-Bollens, J. A. Quenette, and A. B. Bochdansky. 2010. Cascading migrations and implications for vertical fluxes in pelagic ecosystems. Journal of Plankton Research 33:349-355. Bolnick, D. I., R. Svanbäck, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L. Forister. 2003. The ecology of individuals: incidence and implications of individual specialization. The American Naturalist 161:1-28. Bolnick, D. I., P. Amarasekare, M. S. Araújo, R. Bürger, J. M. Levine, M. Novak, V. H. Rudolf, S. J. Schreiber, M. C. Urban, and D. A. Vasseur. 2011. Why intraspecific trait variation matters in community ecology. Trends in Ecology & Evolution 26:183-192. Boscarino, B. T., L. G. Rudstam, J. L. Eillenberger, and R. O’Gorman. 2009. Importance of light, temperature, zooplankton and fish in predicting the nighttime vertical distribution of Mysis diluviana. Aquatic Biology 5:263-279. Boscarino, B. T., L. G. Rudstam, M. A. Minson, and E. E. Freund. 2010. Laboratory- derived light and temperature preferences of juvenile mysid shrimp, Mysis diluviana. Journal of Great Lakes Research 36:699-706. Bottrell, H. H., A. Duncan, M. Z. Gliwicz, E. Grygierek, A. Herzig, A. Hillbricht- Ilkowska, H. Kurasawa, P. Larsson, T. Weglenska. 1976. A review of some problems in zooplankton production studies. Norwegian Journal of Zoology. 24: 419-456. Bowers, J. A. 1988. Diel vertical migration of the opossum shrimp Mysis relicta in Lake Superior: Observations and sampling from the Johnson-Sea-Link II submersible. Bulletin of Marine Science 43:730-738. Bowers, J., and H. Vanderploeg. 1982. In situ predatory behavior of Mysis relicta in Lake Michigan. Hydrobiologia 93:121-131. Bowers, J. A., W. E. Cooper, and D. J. Hall. 1990. Midwater and epibenthic behaviors of Mysis relicta Loven: Observations from the Johnson-Sea-Link II submersible in Lake Superior and from a remotely operated vehicle in northern Lake Michigan. Journal of Plankton Research 12:1279-1286. Brandt, S. B. 1986. Ontogenetic shifts in habitat, diet, and diel-feeding periodicity of slimy sculpin in Lake Ontario. Transactions of the American Fisheries Society 115:711-715.

170

Branstrator, D. K., G. Cabana, A. Mazumder, and J. B. Rasmussen. 2000. Measuring life‐ history omnivory in the opossum shrimp, Mysis relicta, with stable nitrogen isotopes. Limnology and Oceanography 45:463-467. Brenke, N. 2005. An epibenthic sledge for operations on marine soft bottom and bedrock. Marine Technology Society Journal 39: 10-21. Brodersen, J., E. Adahl, C. Brönmark, and L. A. Hansson. 2008a. Ecosystem effects of partial fish migration in lakes. Oikos 117:40-46. Brodersen, J., P. A. Nilsson, L.-A. Hansson, C. Skov, and C. Brönmark. 2008b. Condition-dependent individual decision-making determines cyprinid partial migration. Ecology 89:1195-1200. Brodersen, J., A. Nicolle, P. A. Nilsson, C. Skov, C. Brönmark, and L. A. Hansson. 2011. Interplay between temperature, fish partial migration and trophic dynamics. Oikos 120:1838-1846. Brown, R. H. 1998. Abundance and distribution of Mysis relicta in Moosehead Lake, Maine. Fishery Progress Report Series No. 98-2. Main Department of Inland Fisheries and Wildlife. Bunnell, D. B., B. M. Davis, D. M. Warner, M. A. Chriscinske, and E. F. Roseman. 2011. Planktivory in the changing Lake Huron zooplankton community: Bythotrephes consumption exceeds that of Mysis and fish. Freshwater Biology 56:1281-1296. Bunnell, D. B., B. M. Davis, M. A. Chriscinske, K. M. Keeler, and J. G. Mychek-Londer. 2015. Diet shifts by planktivorous and benthivorous fishes in northern Lake Michigan in response to ecosystem changes. Journal of Great Lakes Research 41:161-171. Burgess, S., E. W. Jackson, L. Schwarzman, N. Gezon, and J. T. Lehman. 2015. Improved estimates of calanoid copepod biomass in the St. Lawrence Great Lakes. Journal of Great Lakes Research 41:484-491. Caldwell, T. J. 2010. The role of the non-native freshwater opossum shrimp Mysis diluviana in the nutrient and food web dynamics of Lake Pend Oreille, Idaho. Doctoral dissertation. University of Idaho. Caldwell, T. J., F. M. Wilhelm, and A. M. Dux. 2016. A non-native pelagic macroinvertebrate alters population dynamics of herbivorous zooplankton in a large deep lake. Canadian Journal of Fisheries and Aquatic Sciences. 73:832-843 Campbell, H. A., M. E. Watts, R. G. Dwyer, C. E. Franklin. 2012. V-Track: software for analysing and visualising animal movement from acoustic telemetry detections. Marine and Freshwater Research 63: 815-820 Cannon, H. G., and S. M. Manton. 1927. On the feeding mechanism of a mysid crustacean, Hemimysis lamornae. Transactions of The Royal Society of Edinburgh 55:219-253. Cartamil, D. P., and C. G. Lowe. 2004. Diel movement patterns of ocean sunfish Mola mola off southern California. Marine Ecology Progress Series 266:245-253. 171

Chapman, B. B., C. Brönmark, J. Å. Nilsson, and L. A. Hansson. 2011. The ecology and evolution of partial migration. Oikos 120:1764-1775. Chapman, B., C. Skov, K. Hulthén, J. Brodersen, P. A. Nilsson, L. A. Hansson, and C. Brönmark. 2012. Partial migration in fishes: definitions, methodologies and taxonomic distribution. Journal of Fish Biology 81:479-499. Chess, D. W., and J. A. Stanford. 1998. Comparative energetics and life cycle of the opossum shrimp (Mysis relicta) in native and non-native environments. Freshwater Biology 40:783-794. Chess, D. W., and J. A. Stanford. 1999. Experimental effects of temperature and prey assemblage on growth and lipid accumulation by Mysis relicta Loven. Hydrobiologia 412:155-164. Chipps, S. R. 1998. Temperature-dependent consumption and gut-residence time in the opossum shrimp Mysis relicta. Journal of Plankton Research 20:2401-2411. Chipps, S. R., and D. H. Bennett. 2000. Zooplanktivory and nutrient regeneration by invertebrate (Mysis relicta) and vertebrate (Oncorhynchus nerka) planktivores: implications for trophic interactions in oligotrophic lakes. Transactions of the American Fisheries Society 129:569-583. Chipps, S. R., and D. H. Bennett. 2002. Evaluation of a Mysis bioenergetics model. Journal of Plankton Research 24:77-82. Clark, C. W., and M. Mangel. 2000. Dynamic state variable models in ecology: methods and applications. Oxford University Press. Clarke, K. R., and R. M. Warwick. 1994. Similarity-based testing for community pattern: the two-way layout with no replication. Marine Biology 118:167-176. Cohen, J. H., and R. B. Forward. 2009. Zooplankton diel vertical migration - a review of proximate control. Oceanography and Marine Biology: An Annual Review 47:77- 109. Cooper, S. D., and C. R. Goldman. 1980. Opossum shrimp (Mysis relicta) predation on zooplankton. Canadian Journal of Fisheries and Aquatic Sciences 37:909-919. Cooper, S. D., and C. R. Goldman. 1982. Environmental factors affecting predation rates of Mysis relicta. Canadian Journal of Fisheries and Aquatic Sciences 39:203-208. Covich, A. P., M. A. Palmer, and T. A. Crowl. 1999. The role of benthic invertebrate species in freshwater ecosystems: zoobenthic species influence energy flows and nutrient cycling. BioScience 49:119-127. Covich, A. P., J. H. Thorp, and D. C. Rogers. 2009. Introduction to the subphylum Crustacea. In Ecology and Classification of North American Freshwater Invertebrates – 3rd Edition. pp. 695-723. Cullen, J. J. 1985. Diel vertical migration by dinoflagellates: roles of carbohydrate metabolism and behavioral flexibility. Contributions in Marine Science 27:135- 152.

172

Cummins, K. W. and J. C. Wuycheck. 1971. Caloric equivalents for investigations in ecological energetics. Internationale Vereinigung für Theoretische und Angewandte Limnologie: Mitteilungen 18:1-158. Dadswell, M. J. 1974. Distribution, ecology and postglacial dispersal of certain crustaceans and fishes in eastern North America. National Museum of Natural Sciences Publication of Theses in Zoology 11:1–110. Darimont, C. T., C. P. Paquet, and T. E. Reimchen. 2007. Stable isotopic niche predicts fitness of prey in a wolf–deer system. Biological Journal of the Linnean Society 90:125-137. Darnis, G., L. Hobbs, M. Geoffroy, J. C. Grenvald, P. E. Renaud, J. Berge, A. Wold. 2017. From polar night to midnight sun: Diel vertical migration, metabolism and biogeochemical role of zooplankton in a high Arctic fjord (Kongsfjorden, Svalbard). Limnology and Oceanography. 62:1586-1605. Davis, C. S., S. M. Gallager, M. Marra, and W. K. Stewart. 1996. Rapid visualization of plankton abundance and taxonomic composition using the Video Plankton Recorder. Deep Sea Research Part II: Topical Studies in Oceanography 43: 1947- 1970. Davis, C. S., Q. Hu, S. M. Gallager, X. Tang, and C. J. Ashjian. 2004. Real-time observation of taxa-specific plankton distributions: an optical sampling method. Marine Ecology Progress Series 284: 77-96. De Bernardi, R. 1984. Methods of estimating zooplankton abundance. In J. A. Downing and F. H. Rigler [eds.], A Manual on Methods for the Assessment of Secondary Productivity in Fresh Waters. Blackwell Scientific Publications. De Robertis, A. 2002. Size-dependent visual predation risk and the timing of vertical migration: an optimization model. Limnology and Oceanography 47:925-933. DeGraeve, G. M., and J. B. Reynolds. 1975. Feeding behavior and temperature and light tolerance of Mysis relicta in the laboratory. Transactions of the American Fisheries Society 104:394-397. den Heyer, C., and J. Kalff. 1998. Organic matter mineralization rates in sediments: A within-and among-lake study. Limnology and Oceanography 43:695-705. Dermott, R. M., R. Bonnell, P. Jarvis. 2006. Changes in abundance of deep water amphipod Diporeia (Pontoporeiidae) in eastern North American lakes with or without Dreissena mussels. Canadian Technical Report of Fisheries and Aquatic Sciences. pp. 88. Doherty, P. J. 1987. Light-traps: selective but useful devices for quantifying the distributions and abundances of larval fishes. Bulletin of Marine Science 41:423- 431. Dussart, B. H., and D. Defaye. 2001. Introduction to the Copepoda. In Guide to the identification of the microinvertebrates of the continental waters of the world, 2nd edition, No. 16. Backhuys Publishers, Leiden. 344 p. 173

Egevang, C., I. J. Stenhouse, R. A. Phillips, A. Petersen, J. W. Fox, and J. R. Silk. 2010. Tracking of Arctic terns Sterna paradisaea reveals longest animal migration. Proceedings of the National Academy of Sciences 107:2078-2081. Ellis, B. K., J. A. Stanford, D. Goodman, C. P. Stafford, D. L. Gustafson, D. A. Beauchamp, D. W. Chess, et al. 2011. Long-term effects of a trophic cascade in a large lake ecosystem. Proceedings of the National Academy of Sciences 108:1070-1075. EPA Great Lakes National Program Office. 2015. Standard operating procedure for Mysis sample collection and preservation. LG 409, Revision 01, April 2015, Chicago ,IL, USA [online]. Available from: http://www.epa.gov/sites/production/files/2017-01/documents/sop-for-mysis- sample-collection-and-preservation-201504-8pp.pdf [accessed 10 September 2017]. Eshenroder, R. L. 2008. Differentiation of deep-water lake charr Salvelinus namaycush in North American lakes. Environmental Biology of Fishes 83:77-90. Eshenroder, R. L., and M. K. Burnham-Curtis. 1999. Species succession and sustainability of the Great Lakes fish community. In W. W. Taylor [Ed.] Great Lakes fishery policy and management: A binational perspective. Michigan State University Press. Pages 145-184. Eshenroder, R. L., and M. K. Burnham-Curtis. 2001. Comment: Vertical migration of bloaters. Transactions of the American Fisheries Society 130:166-167. Euclide, P. T., S. Hansson, and J. D. Stockwell. 2017. Partial diel vertical migration in an omnivorous macroinvertebrate, Mysis diluviana. Hydrobiologia 787:387-396. Evans, M. S., and D. W. Sell. 1985. Mesh size and collection characteristics of 50-cm diameter conical plankton nets. Hydrobiologia 122:97-104. Everson, I., and D. Bone. 1986. Effectiveness of the RMT8 system for sampling krill (Euphausia superba) swarms. Polar Biology 6: 83-90. Ewing, M., J. L. Worzel, and A. C. Vine. 1967. Early development of ocean-bottom photography at Woods Hole Oceanographic Institution and Lamont Geological Observatory. The John Hopkins Oceanographic Studies. Feldmann, T., M. Yakovleva, M. Lindström, K. Donner, and M. Ostrovsky. 2010. Eye adaptation to different light environments in two populations of Mysis relicta: a comparative study of carotenoids and retinoids. Journal of Crustacean Biology 30:636-642. Fiksen, Ø., and F. Carlotti. 1998. A model of optimal life history and diel vertical migration in Calanus finmarchicus. Sarsia 83:129-147. Flynn, K. J., A. Mitra, A. Bode. 2018. Toward a mechanistic understanding of trophic structure: inferences from simulating stable isotope ratios. Marine Biology 165: 147.

174

France, R. L. 2012. Omnivory, vertical food‐web structure and system productivity: stable isotope analysis of freshwater planktonic food webs. Freshwater Biology 57:787-794. Frost, B. W., and S. M. Bollens. 1992. Variability of diel vertical migration in the marine planktonic copepod Pseudocalanus newmani in relation to its predators. Canadian Journal of Fisheries and Aquatic Sciences. 49: 1137-1141. Gal, G., E. R. Loew, L. G. Rudstam, and M. A. Mohammadian. 1999. Light and diel vertical migration: spectral sensitivity and light avoidance by Mysis relicta. Canadian Journal of Fisheries and Aquatic Sciences 56:311-322. Gal, G., L. G. Rudstam, and O. E. Johannsson. 2004. Predicting Mysis relicta vertical distribution in Lake Ontario. Archiv für Hydrobiologie 159:1-23. Gal, G., L. G. Rudstam, E. L. Mills, J. R. Lantry, O. E. Johannsson, and C. H. Greene. 2006. Mysid and fish zooplanktivory in Lake Ontario: quantification of direct and indirect effects. Canadian Journal of Fisheries and Aquatic Sciences 63:2734- 2747. Gamble, A. E., T. R. Hrabik, J. D. Stockwell, and D. L. Yule. 2011a. Trophic connections in Lake Superior Part I: the offshore fish community. Journal of Great Lakes Research 37:541-549. Gamble, A. E., T. R. Hrabik, D. L. Yule, and J. D. Stockwell. 2011b. Trophic connections in Lake Superior Part II: the nearshore fish community. Journal of Great Lakes Research 37:550-560. Gende, S. M., R. T. Edwards, M. F. Willson, and M. S. Wipfli. 2002. Pacific salmon in aquatic and terrestrial ecosystems. BioScience 52:917-928. Gibbons, M. J. 1993. Vertical migration and feeding of Euphausia lucens at two 72 h stations in the southern Benguela upwelling region. Marine Biology 116:257-268. GLFC 2015. Basin-wide research priorities - http://www.glfc.org/research/Basin_Wide_Priorities.pdf Gliwicz, M. Z. 1986. Predation and the evolution of vertical migration in zooplankton. Nature 320:746-748. Gliwicz, M. Z., and J. Pijanowska. 1988. Effect of predation and resource depth distribution on vertical migration of zooplankton. Bulletin of Marine Science 43:695-709. Gorbatenko, K., and N. Dolganova. 2007. Comparing the catch efficiency with different types of plankton nets in the high production zones of the Pacific Ocean. Oceanology 47: 205-212. Gorokhova, E. 2009. Toxic cyanobacteria Nodularia spumigena in the diet of Baltic mysids: Evidence from molecular diet analysis. Harmful Algae 8:264-272. Gregg, R. E. 1976. The ecology of Mysis relicta in Twin Lakes, Colorado. MS thesis. Colorado State University.

175

Grosjean, P., M. Picheral, C. Warembourg, and G. Gorsky. 2004. Enumeration, measurement, and identification of net zooplankton samples using the ZOOSCAN digital imaging system. ICES Journal of Marine Science 61: 518-525. Grossnickle, N. E. 1979. Nocturnal feeding patterns of Mysis relicta in Lake Michigan, based on gut content fluorescence. Limnology and Oceanography 24:777-780. Grossnickle, N. E. 1982. Feeding habits of Mysis relicta - an overview. Hydrobiologia 93:101-107. Grossnickle, N. E., and M. D. Morgan. 1979. Density estimates of Mysis relicta in Lake Michigan. Journal of the Fisheries Board of Canada 36: 694-698. Guisande, C., A. Duncan, and W. Lampert. 1991. Trade-offs in Daphnia vertical migration strategies. Oecologia 87:357-359. Gutowski T. 1978. The ecology of Mysis relicta (Loven) in Lake Champlain. MS thesis. University of Vermont. Hagen, W., E. S. Van Vleet, and G. Kattner. 1996. Seasonal lipid storage as overwintering strategy of Antarctic krill. Marine Ecology Progress Series 134:85- 89. Hairston, N. 1976. Photoprotection by carotenoid pigments in the copepod Diaptomus nevadensis. Proceedings of the National Academy of Sciences 73:971-974. Hamrén, U., and S. Hansson. 1999. A mysid shrimp (Mysis mixta) is able to detect the odour of its predator (Clupea harengus). Ophelia 51:187-191. Hansson, L.-A. and S. Hylander. 2009. Size-structured risk assessments govern Daphnia migration. Proceedings of the Royal Society of London B: Biological Sciences 276:331-336. Harrison, P. M., L. F. G. Gutowsky, E. G. Martins, D. A. Patterson, A. Leake, S. J. Cooke, and M. Power. 2013. Diel vertical migration of adult burbot: a dynamic trade-off among feeding opportunity, predation avoidance, and bioenergetic gain. Canadian Journal of Fisheries and Aquatic Sciences 70:1765-1774. Haskell, C. A., and J. A. Stanford, J. A. 2006. Ecology of an estuarine mysid shrimp in the Columbia River (USA). River Research and Applications 22:739-753. Haupt, F., M. Stockenreiter, M. Baumgartner, M. Boersma, and H. Stibor. 2009. Daphnia diel vertical migration: implications beyond zooplankton. Journal of Plankton Research 5:515-524. Hays, G. C. 1995. Ontogenetic and seasonal variation in the diel vertical migration of the copepods Metridia lucens and Metridia longa. Limnology and Oceanography 40:1461-1465. Hays, G. C. 2003. A review of the adaptive significance and ecosystem consequences of zooplankton diel vertical migrations. Pages 163-170 in Migrations and Dispersal of Marine Organisms. Springer.

176

Hays, G., H. Kennedy, and B. Frost. 2001. Individual variability in diel vertical migration of a marine copepod: why some individuals remain at depth when others migrate. Limnology and Oceanography 46:2050-2054. Herman, A., B. Beanlands, and E. Phillips. 2004. The next generation of optical plankton counter: the laser-OPC. Journal of Plankton Research 26:1135-1145. Hölker, F., M. J. Vanni, J. J. Kuiper, C. Meile, H.-P. Grossart, P. Stief, R. Adrian, A. Lorke, O. Dellwig, and A. Brand. 2015. Tube‐dwelling invertebrates: tiny ecosystem engineers have large effects in lake ecosystems. Ecological Monographs 85:333-351. Hondorp, D. W. 2006. Factors influencing diet and prey selection of the sculpin Cottus cognatus and Myoxocephalus thompsonii. PhD Dissertation, University of Michigan, Ann Arbor, MI. Hondorp, D. W., S. A. Pothoven, and S. B. Brandt. 2005. Influence of Diporeia density on diet composition, relative abundance, and energy density of planktivorous fishes in southeast Lake Michigan. Transactions of the American Fisheries Society 134:588-601. Hrycik, A. R., P. W. Simonin, L. G. Rudstam, D. L. Parrish, B. Pientka, and T. B. Mihuc. 2015. Mysis zooplanktivory in Lake Champlain: A bioenergetics analysis. Journal of Great Lakes Research 41:492-501. Hutchinson, G. E. 1967. A Treatise on Limnology: II. Introduction to lake biology and their limnoplankton. John Wiley & Sons Publishing. Isaac, E. J., T. R. Hrabik, J. D. Stockwell, and A. E. Gamble. 2012. Prey selection by the Lake Superior fish community. Journal of Great Lakes Research 38:326-335. Jensen, O. P., T. R. Hrabik, S. J. Martell, C. J. Walters, and J. F. Kitchell. 2006. Diel vertical migration in the Lake Superior pelagic community. II. Modeling tradeoffs at an intermediate trophic level. Canadian Journal of Fisheries and Aquatic Sciences 63:2296-2307. Jensen, O., S. Hansson, T. Didrikas, J. D. Stockwell, T. R. Hrabik, T. Axenrot, and J. F. Kitchell. 2011. Foraging, bioenergetic and predation constraints on diel vertical migration: field observations and modelling of reverse migration by young‐of‐the‐ year herring Clupea harengus. Journal of Fish Biology 78:449-465. Johannsson, O. E. 1995. Response of Mysis relicta population dynamics and productivity to spatial and seasonal gradients in Lake Ontario. Canadian Journal of Fisheries and Aquatic Sciences 52:1509-1522. Johannsson, O. E., L. G. Rudstam, and D. C. Lasenby. 1994. Mysis relicta: assessment of metalimnetic feeding and implications for competition with fish in Lakes Ontario and Michigan. Canadian Journal of Fisheries and Aquatic Sciences 51:2591-2602. Johannsson, O. E., M. F. Leggett, L. G. Rudstam, M. R. Servos, M. A. Mohammadian, G. Gal, R. M. Dermott, and R. H. Hesslein. 2001. Diet of Mysis relicta in Lake

177

Ontario as revealed by stable isotope and gut content analysis. Canadian Journal of Fisheries and Aquatic Sciences 58:1975-1986. Johannsson, O. E., L. G. Rudstam, G. Gal, and E. Mills. 2003. Mysis relicta in Lake Ontario: population dynamics, trophic linkages and further questions. State of Lake Ontario (SOLO) - Past, present and future. Ecovision World Monograph Series:257-287. Johannsson, O. E., K. L. Bowen, C. M. Wood, R. W. Smith, C. Chu, L. G. Rudstam, and B. T. Boscarino. 2008. Relating nucleic acid and protein indices to growth in Mysis relicta: ration, cycling temperature, and metabolism. Aquatic Biology 4:33- 46. Johnsen, G. H. and P. J. Jakobsen. 1987. The effect of food limitation on vertical migration in Daphnia longispina. Limnology and Oceanography 32:873-880. Johnson, B. M., W. M. Pate, D. B. Silver, and J. L. Sharp. 2018. Invasion success and population characteristics of the opossum shrimp, Mysis diluviana, in Wyoming, USA. Aquatic Invasions 13:409-420. Jónasdóttir, S. H., A. W. Visser, K. Richardson, and M. R. Heath. 2015. Seasonal copepod lipid pump promotes carbon sequestration in the deep North Atlantic. Proceedings of the National Academy of Sciences 112:12122-12126. Jude, D. J., L. G. Rudstam, T. J. Holda, J. M. Watkins, P. T. Euclide, and M. D. Balcer. 2018. Trends in Mysis diluviana abundance in the Great Lakes, 2006- 2016. Journal of Great Lakes Research 44:590-599. Jumars, P. A. 2007. Habitat coupling by mid-latitude, subtidal, marine mysids: import- subsidised omnivores. In R. Gibson, R. Atkinson and J. Gordon [Eds.] Oceanography and Marine Biology: An annual review 45:89-138. Karatayev, A. Y., K. Mehler, L. E. Burlakova, E. K. Hinchey, and G. J. Warren. 2018. Benthic video image analysis facilitates monitoring of Dreissena populations across spatial scales. Journal of Great Lakes Research 44:629-638. Karnaukhov, D. Y. and others 2016. Behaviour of Lake Baikal amphipods as a part of the night migratory complex in the Kluevka settlement region (south-eastern Baikal). Crustaceana 89: 419-430. Kerfoot, W. C. 1985. Adaptive value of vertical migration: comments on the predation hypothesis and some alternatives. Contributions in Marine Science 27:91-113. Kitchell, J. F., S. P. Cox, C. J. Harvey, T. B. Johnson, D. M. Mason, K. K. Schoen, K. Aydin, C. Bronte, M. Ebener, and M. Hansen. 2000. Sustainability of the Lake Superior fish community: interactions in a food web context. Ecosystems 3:545- 560. Knight, J. C., B. P. O'Malley, and J. D. Stockwell. 2018. Lake Champlain offshore benthic invertebrate community before and after zebra mussel invasion. Journal of Great Lakes Research 44:283-288.

178

Lampert, W. 1989. The adaptive significance of diel vertical migration of zooplankton. Functional Ecology 3:21-27. Lampert, W. 1993. Ultimate causes of diel vertical migration of zooplankton: new evidence for the predator-avoidance hypothesis. Archiv für Hydrobiologie– Monographische Beiträge 39:79-88. Lampert, W. 2005. Vertical distribution of zooplankton: density dependence and evidence for an ideal free distribution with costs. BMC biology 3:10. Langeland, A. 1988. Decreased zooplankton density in a mountain lake resulting from predation by recently introduced Mysis relicta. Verhandlungen des Internationalen Verein Limnologie 23: 419-429. Lasenby, D. C. 1991. Comments on the roles of native and introduced Mysis relicta in aquatic ecosystems. American Fisheries Society Symposium 9:17-22. Lasenby, D. C., and R. R. Langford. 1973. Feeding and assimilation of Mysis relicta. Limnology and Oceanography 18:280-285. Lasenby, D., T. Northcote, and M. Fürst. 1986. Theory, practice, and effects of Mysis relicta introductions to North American and Scandinavian lakes. Canadian Journal of Fisheries and Aquatic Sciences 43:1277-1284. Lasenby, D., and Y. Shi. 2004. Changes in the elemental composition of the stomach contents of the opossum shrimp Mysis relicta during diel vertical migration. Canadian Journal of Zoology 82:525-528. Lehman, J. T., J. A. Bowers, R. W. Gensemer, G. J. Warren, and D. K. Branstrator. 1990. Mysis relicta in Lake Michigan: abundances and relationships with their potential prey, Daphnia. Canadian Journal of Fisheries and Aquatic Sciences 47:977-983. Levy, D. A. 1990. Reciprocal diel vertical migration behavior in planktivores and zooplankton in British Columbia lakes. Canadian Journal of Fisheries and Aquatic Sciences 47:1755-1764. Lindén, E., and H. Kuosa. 2004. Effects of grazing and excretion by pelagic mysids (Mysis spp.) on the size structure and biomass of the phytoplankton community. In Biology of the Baltic Sea. Springer, Dordrecht. Pages: 73-78. Longhurst, A. R., and W. G. Harrison. 1988. Vertical nitrogen flux from the oceanic photic zone by diel migrant zooplankton and nekton. Deep Sea Research Part A. Oceanographic Research Papers 35:881-889. Loose, C. J., and P. Dawidowicz. 1994. Trade-offs in diel vertical migration by zooplankton: the costs of predator avoidance. Ecology 75: 2255-2263. Macquart-Moulin, C., A. Bourdillon, F. Cubizolles, F. Passelaigue, and R. Rasoanarivo. 1987. Un cas type de migration verticale ‘retard’chez l'amphipode Ampelisca typica. Journal of Plankton Research 9:785-809.

179

Madeira, P. T., A. S. Brooks, and D. B. Seale, D. B. 1982. Excretion of total phosphorus, dissolved reactive phosphorus, ammonia, and urea by Lake Michigan Mysis relicta. In Ecology of Mysidacea. Springer, Dordrecht. Pages: 145-154. Madenjian, C., and D. J. Jude. 1985. Comparison of sleds versus plankton nets for sampling fish larvae and eggs. Hydrobiologia 124: 275-281. Malley, D., and J. Reynolds. 1979. Sampling strategies and life history of non-insectan freshwater invertebrates. Journal of the Fisheries Research Board of Canada 36:311-318. Mangel, M., and C. W. Clark. 1988. Dynamic modeling in behavioral ecology. Princeton University Press. Marsden, J. E., and M. Hauser. 2009. Exotic species in Lake Champlain. Journal of Great Lakes Research 35:250-265. Marsden, J. E., P. Stangel, and A. D. Shambaugh. 2013. Influence of Environmental Factors on Zebra Mussel Population Expansion in Lake Champlain, 1994– 2010. In Quagga and Zebra Mussels: Biology, Impacts, and Control 33:34-49. Martinez, P. J., and E. P. Bergersen. 1989. Proposed biological management of Mysis relicta in Colorado lakes and reservoirs. North American Journal of Fisheries Management 9:1-11. Masson, S., B. Pinel-Alloul, G. Méthot, and N. Richard. 2004. Comparison of nets and pump sampling gears to assess zooplankton vertical distribution in stratified lakes. Journal of Plankton Research 26: 1199-1206. Mauchline, J. 1971. Seasonal occurrence of mysids (Crustacea) and evidence of social behavior. Journal of the Marine Biological Association of the United Kingdom 51: 809-825. Mauchline, J. 1980. The biology of mysids and euphausiids. Academic Press, London, UK. McConnell, A., R. Routledge, and B. Connors. 2010. Effect of artificial light on marine invertebrate and fish abundance in an area of salmon farming. Marine Ecology Progress Series 419:147-156. McLaren, I. A. 1974. Demographic strategy of vertical migration by a marine copepod. American Naturalist:91-102. McMeans, B. C., K. S. McCann, T. D. Tunney, A. T. Fisk, A. M. Muir, N. Lester, B. Shuter, and N. Rooney. 2016. The adaptive capacity of lake food webs: from individuals to ecosystems. Ecological Monographs 86:4-19. McWilliam, P. S. 1970. Seasonal changes in abundance and reproduction in the opossum shrimp, Mysis relicta Lovén in Lake Michigan. MS thesis. University of Sydney, Australia. Mehner, T. 2012. Diel vertical migration of freshwater fishes – proximate triggers, ultimate causes and research perspectives. Freshwater Biology 57:1342-1359.

180

Mehner, T. 2014. Partial diel vertical migration of sympatric vendace (Coregonus albula) and Fontane cisco (Coregonus fontanae) is driven by density dependence. Canadian Journal of Fisheries and Aquatic Sciences 72:116-124. Mehner, T., P. Kasprzak, and F. Hölker. 2007. Exploring ultimate hypotheses to predict diel vertical migrations in coregonid fish. Canadian Journal of Fisheries and Aquatic Sciences 64:874-886. Mehner, T., S. Busch, I. P. Helland, M. Emmrich, and J. Freyhof. 2010. Temperature‐ related nocturnal vertical segregation of coexisting coregonids. Ecology of Freshwater Fish 19:408-419. Mehner, T., and P. Kasprzak. 2011. Partial diel vertical migrations in pelagic fish. Journal of Animal Ecology 80:761-770. Mihuc, T. B., F. Dunlap, C. Binggeli, L. Myers, C. Pershyn, A. Groves, and A. Waring. 2012. Long-term patterns in Lake Champlain's zooplankton: 1992–2010. Journal of Great Lakes Research 38:49-57. Mihuc, T. B., and F. Recknagel. 2018. Lessons from Bioinvasion of Lake Champlain, USA. In Ecological Informatics. Springer. pp. 401-414. Milner-Gulland, E.J., J.M. Fryxell, and A.R.E Sinclair (Eds.). 2011. Animal migration: a synthesis. Oxford: Oxford University Press. Moen, V., and A. Langeland. 1989. Diurnal vertical and seasonal horizontal distribution patterns of Mysis relicta in a large Norwegian lake. Journal of Plankton Research 11:729-745. Morgan, M. D. 1980. Life history characteristics of two introduced populations of Mysis relicta. Ecology:551-561. Morgan, M. D., S. T. Threlkeld, and C. R. Goldman. 1978. Impact of the introduction of kokanee (Oncorhynchus nerka) and opossum shrimp (Mysis relicta) on a subalpine lake. Journal of the Fisheries Research Board of Canada 35:1572-1579. Morgan, M. D., and S. T. Threlkeld. 1982. Size dependent horizontal migration of Mysis relicta. Hydrobiologia 93:63-68. Murano, M. 1966. Culture of Isaza-Ami Neomysis intermedia. Information Bulletin of Planktology in Japan 13:65-68. Murtaugh, P. A. 1984. Variable gut residence time: problems in inferring feeding rate from stomach fullness of a mysid crustacean. Canadian Journal of Fisheries and Aquatic Sciences 41:1287-1293. Myer, G.E., and G. K. Gruendling. 1979. Limnology of Lake Champlain. Lake Champlain Basin Study. New England River Basin Commission, Burlington, VT. Naesje, R., A. Jensen, and O. Sandlund. 2003. Life history, habitat utilisation, and biomass of introduced Mysis relicta. Limnologica-Ecology and Management of Inland Waters 33:244-257.

181

Nakagawa, Y., Y. Endo, and H. Sugisaki. 2003. Feeding rhythm and vertical migration of the euphausiid Euphausia pacifica in coastal waters of north-eastern Japan during fall. Journal of Plankton Research 25:633-644. Nesler, T., and E. Bergersen. 1991. Mysids and their impacts on fisheries: an introduction to the 1988 mysid-fisheries symposium. In American Fisheries Society Symposium. Pages 1-4. Neverman, D., and W. A. Wurtsbaugh. 1992. Visual feeding by juvenile Bear Lake sculpin. Transactions of the American Fisheries Society 121: 395-398. Neverman, D., and W. A. Wurtsbaugh. 1994. The thermoregulatory function of diel vertical migration for a juvenile fish, Cottus extensus. Oecologia 98:247-256. Nordin, L. J., M. T. Arts, O. E. Johannsson, and W. D Taylor. 2008. An evaluation of the diet of Mysis relicta using gut contents and fatty acid profiles in lakes with and without the invader Bythotrephes longimanus (Onychopoda, Cercopagidae). Aquatic Ecology 42:421-436. Nowicki, C. J., D. B. Bunnell, P. M. Armenio, D. M. Warner, H. A. Vanderploeg, J. F. Cavaletto, C. M. Mayer, and J. V. Adams. 2017. Biotic and abiotic factors influencing zooplankton vertical distribution in Lake Huron. Journal of Great Lakes Research 43:1044-1054. Ogonowski, M., S. Hansson, and J. Duberg. 2013a. Status and vertical size-distributions of a pelagic mysid community in the northern Baltic proper. Boreal Environment Research 18:1-19. Ogonowski, M., J. Duberg, S. Hansson, and E. Gorokhova. 2013b. Behavioral, ecological and genetic differentiation in an open environment - a study of a mysid population in the Baltic Sea. PloS one 8:e57210. Ohman, M. D. 1990. The demographic benefits of diel vertical migration by zooplankton. Ecological Monographs 60:257-281. Ohman, M. D. and others 2013. Autonomous ocean measurements in the California Current Ecosystem. Oceanography 26: 18-25. Oksanen J., F. G. Blanchet, M. Friendly, R. Kindt, P. Legendre, D. McGlinn, P. R. Minchin, R. B. O'Hara, G. L. Simpson, P. Solymos, M. Stevens, E. Szoecs, and H. Wagner. 2018. vegan: Community Ecology Package. R package version 2.5-1. https://CRAN.R-project.org/package=vegan. Oliver, S. K., D. K. Branstrator, T. R Hrabik, S. J. Guildford, and R. E. Hecky. 2015. Nutrient excretion by crustacean zooplankton in the deep chlorophyll layer of Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 72: 390-399. O'Malley, B. P., and D. B. Bunnell. 2014. Diet of Mysis diluviana reveals seasonal patterns of omnivory and consumption of invasive species in offshore Lake Michigan. Journal of Plankton Research 36:989-1002.

182

O’Malley, B. P., L. G. Rudstam, J. M. Watkins, T. J. Holda, and B. C. Weidel. 2017. Effects of food web changes on Mysis diluviana diet in Lake Ontario. Journal of Great Lakes Research 43:813-822. O'Malley, B. P., S. Hansson, and J. D. Stockwell. 2018. Evidence for a size-structured explanation of partial diel vertical migration in mysids. Journal of Plankton Research 40:66-76. O’Malley, B. P., and J. D. Stockwell. In review. Diel feeding behavior in a partially migrant Mysis population: A benthic-pelagic comparison. Food Webs. Onsrud, M. S., and S. Kaartvedt. 1998. Diel vertical migration of the krill Meganyctiphanes norvegica in relation to physical environment, food and predators. Marine Ecology Progress Series 171:209-219. Parker, J. I. 1980. Predation by Mysis relicta on Pontoporeia hoyi: a food chain link of potential importance in the Great Lakes. Journal of Great Lakes Research 6:164- 166. Parker, S. L., L. G. Rudstam, E. L. Mills, and D. W. Einhouse. 2001. Retention of Bythotrephes spines in the stomachs of eastern Lake Erie rainbow smelt. Transactions of the American Fisheries Society 130:988-994. Patwa, Z., R. Christensen, D. C. Lasenby, E. Webster, and D. Mackay. 2007. An exploration of the role of mysids in benthic‐pelagic coupling and biomagnification using a dynamic bioaccumulation model. Environmental Toxicology and Chemistry 26:1224-1232. Pearre, S. 1979. Problems of detection and interpretation of vertical migration. Journal of Plankton Research 1:29-44. Pearre, S. 2003. Eat and run? The hunger/satiation hypothesis in vertical migration: history, evidence and consequences. Biological Reviews 78:1-79. Pepin, P., and T. Shears. 1997. Variability and capture efficiency of bongo and Tucker trawl samplers in the collection of ichthyoplankton and other macrozooplankton. Canadian Journal of Fisheries and Aquatic Sciences 54: 765-773. Pimm, S. L. 1982. Food webs. Springer. Pimm, S. L., J. H. Lawton, and J. E. Cohen. 1991. Food web patterns and their consequences. Nature 350:669-674. Pimm, S. L. and others 2015. Emerging technologies to conserve biodiversity. Trends in Ecology and Evolution 30: 685-696. Polis, G. A. 1991. Complex trophic interactions in deserts: an empirical critique of food- web theory. The American Naturalist 138:123-155. Polis, G. A., W. B. Anderson, and R. D. Holt. 1997. Toward an integration of landscape and food web ecology: the dynamics of spatially subsidized food webs. Annual Review of Ecology and Systematics 28: 289-316.

183

Ponomareva, L. A. 1971. Circadian migrations and feeding rhythms of some Indian Ocean euphausiid species. Oceanology 11:226–231. Pothoven, S. A., G. L. Fahnenstiel, H. A. Vanderploeg, and M. Luttenton. 2000. Population dynamics of Mysis relicta in southeastern Lake Michigan, 1995–1998. Journal of Great Lakes Research 26:357-365. Pratt, T. C., O. T. Gorman, W. P. Mattes, J. T. Myers, H. R. Quinlan, D. R. Schreiner, M. J. Seider, S. P. Sitar, D. L. Yule, and P. M. Yurista. 2016. The state of Lake Superior in 2011 [online]. Available from: http://www.glfc.org/pubs/SpecialPubs/Sp16_01.pdf [accessed 21 April 2016]. Quirt, J., and D. Lasenby. 2002. Cannibalism and ontogenetic changes in the response of the freshwater shrimp Mysis relicta to chemical cues from conspecific predators. Canadian Journal of Zoology 80:1022-1025. R Core Team. 2017. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R- project.org/. Ramcharan, C. W., and W. G. Sprules. 1986. Visual predation in Mysis relicta Lovén. Limnology and Oceanography 31:414-420. Reeve, H. K., and P. W. Sherman. 1993. Adaptation and the goals of evolutionary research. Quarterly Review of Biology:1-32. Ricker, K. E. 1959. The origin of two glacial relict crustaceans in North America, as related to Pleistocene glaciation. Canadian Journal of Zoology 37:871-893. Rieman, B. E., and C. M. Falter. 1981. Effects of the establishment of Mysis relicta on the macrozooplankton of a large lake. Transactions of the American Fisheries Society 110: 613-620. Ringelberg, J. 1991. A mechanism of predator-mediated induction of diel vertical migration in Daphnia hyalina. Journal of Plankton Research 13:83-89. Ringelberg, J. 2010. Diel vertical migration of zooplankton in lakes and oceans. Springer. Ringelberg, J., B. J. G. Flik, D. K. Aaned, and E. Van Gool. 1997. Amplitude of diel vertical migration (DVM) is a function of ﬁsh biomass, a hypothesis. Archiv für Hydrobiologie–Monographische Beiträge 49:71-78. Robertson, A., C. F. Powers, and R. F. Anderson. 1968. Direct observations on Mysis relicta from a submarine. Limnology and Oceanography 13:700-702. Rogers, M. W., D. B. Bunnell, C. P. Madenjian, and D. M. Warner. 2014. Lake Michigan offshore ecosystem structure and food web changes from 1987 to 2008. Canadian Journal of Fisheries and Aquatic Sciences 71:1072-1086. Roughgarden, J. 1972. Evolution of niche width. The American Naturalist 106:683-718. Rudstam, L. G. 2009. Other Zooplankton. In G. E. Likens [Ed.] Encyclopedia of Inland Waters. Volume 3. Oxford.

184

Rudstam, L. G., K. Danielsson, S. Hansson, and S. Johansson. 1989. Diel vertical migration and feeding patterns of Mysis mixta (Crustacea, Mysidacea) in the Baltic Sea. Marine Biology 101:43-52. Rudstam, L. G., A. L. Hetherington, and A. M. Mohammadian. 1999. Effect of temperature on feeding and survival of Mysis relicta. Journal of Great Lakes Research 25:363-371. Rudstam, L. G., and O. E. Johannsson. 2009. Advances in the ecology of freshwater mysids. Aquatic Biology 5:246-248. Rybock, J. T. 1978. Mysis relicta Loven in Lake Tahoe: vertical distribution and nocturnal predation. PhD Dissertation. University of Nevada. Schindler, D. E., and M. D. Scheuerell. 2002. Habitat coupling in lake ecosystems. Oikos 98:177-189. Schmidt, K. and others 2011. Seabed foraging by Antarctic krill: Implications for stock assessment, bentho‐pelagic coupling, and the vertical transfer of iron. Limnology and Oceanography 56: 1411-1428. Schneider, C. A., W. S. Rasband, and K. W. Eliceiri. 2012. NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9: 671. Shea, M. A., and J. C. Makarewicz. 1989. Production, biomass, and trophic interactions of Mysis relicta in Lake Ontario. Journal of Great Lakes Research 15:223-232. Sierszen, M. E., G. S. Peterson, J. V. Scharold. 2006. Depth-specific patterns in benthic planktonic food web relationships in Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 63:1496-1503. Sierszen, M. E., J. R. Kelly, T. D. Corry, J. V. Scharold, P. M. Yurista, and G. Sprules. 2011. Benthic and pelagic contributions to Mysis nutrition across Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 68:1051-1063. Sierszen, M. E., T. R. Hrabik, J. D. Stockwell, A. M. Cotter, J. C. Hoffman, and D. L. Yule. 2014. Depth gradients in food‐web processes linking habitats in large lakes: Lake Superior as an exemplar ecosystem. Freshwater Biology 59:2122-2136. Silver, D. B., B. M. Johnson, W. M. Pate, K. R. Christianson, J. Tipton, J. Sherwood, B. Smith, and Y. Hao. 2016. Effect of net size on estimates of abundance, size, age and sex ratio of Mysis diluviana. Journal of Great Lakes Research 42:731-737. Simmard, Y., G. Lacroix, and L. Legendre. 1986. Diel vertical migrations and nocturnal feeding of a dense coastal krill scattering layer (Thysanoessa raschi and Meganyctiphanes norvegica) in stratified surface waters. Marine Biology 91:93- 105. Sims, D. W., V. J. Wearmouth, E. J. Southall, J. M. Hill, P. Moore, K. Rawlinson, et al. 2006. Hunt warm, rest cool: bioenergetic strategy underlying diel vertical migration of a benthic shark. Journal of Animal Ecology 75:176-190.

185

Skov, C., H. Baktoft, J. Brodersen, C. Brönmark, B. B. Chapman, L.-A. Hansson, and P. A. Nilsson. 2011. Sizing up your enemy: individual predation vulnerability predicts migratory probability. Proceedings of the Royal Society of London B: Biological Sciences 278:1414-1418. Smeltzer, E., A. D. Shambaugh, and P. Stangel. 2011. Environmental change in Lake Champlain revealed by long-term monitoring. Journal of Great Lakes Research 38:6-18. Spencer, C. N., B. R. McClelland, and J. A. Stanford. 1991. Shrimp stocking, salmon collapse, and eagle displacement. BioScience 41:14-21. Stewart, T. J., and W. G. Sprules. 2011. Carbon-based balanced trophic structure and flows in the offshore Lake Ontario food web before (1987–1991) and after (2001– 2005) invasion-induced ecosystem change. Ecological Modelling 222:692-708. Stewart, T. J., and L. G. Rudstam. 2012. Did Mysis have a role in the decline of Diporeia? GLFC Project Completion Report. Stockwell, J. D., T. R. Hrabik, O. P. Jensen, D. L. Yule, and M. Balge. 2010. Empirical evaluation of predator-driven diel vertical migration in Lake Superior. Canadian Journal of Fisheries and Aquatic Sciences 67:473-485. Stoner, A. W., C. H. Ryer, S. J. Parker, P. J. Auster, and W. W. Wakefield. 2008. Evaluating the role of fish behavior in surveys conducted with underwater vehicles. Canadian Journal of Fisheries and Aquatic Sciences 65:1230-1243. Straile, D., R. Adrian, and D. E. Schindler. 2012. Uniform temperature dependency in the phenology of a keystone herbivore in lakes of the Northern Hemisphere. PloS ONE 7: e45497. Svanbäck, R., and D. I. Bolnick. 2007. Intraspecific competition drives increased resource use diversity within a natural population. Proceedings of the Royal Society of London B: Biological Sciences 274:839-844. Svanbäck, R., P. Eklöv, R. Fransson, and K. Holmgren. 2008. Intraspecific competition drives multiple species resource polymorphism in fish communities. Oikos 117:114-124. Takahashi, K. 2004. Feeding ecology of mysids in freshwater and coastal marine habitats: A review. Bulletin of Plankton Society of Japan 51:46-72. Teraguchi, M. 1969. Diel vertical migration of Mysis relicta (Loven) in Green Lake, Wisconsin. PhD Dissertation. University of Wisconsin, Madison. Twiss, M. R., and K. M. Stryszowska. 2016. State of emerging technologies for assessing aquatic condition in the Great Lakes-St. Lawrence River system. Journal of Great Lakes Research 42: 1470-1477. Väinölä, R., B. J. Riddoch, R. D. Ward, and R. I. Jones. 1994. Genetic zoogeography of the Mysis relicta species group (Crustacea: Mysidacea) in northern Europe and North America. Canadian Journal of Fisheries and Aquatic Sciences 51:1490- 1505. 186

Van Duyn-Henderson, J. V., and D. C. Lasenby. 1986. Zinc and cadmium transport by the vertically migrating opossum shrimp, Mysis relicta. Canadian Journal of Fisheries and Aquatic Sciences 43:1726-1732. Vanderploeg, H. A., J. A. Bowers, O. Chapelski, and H. K. Soo. 1982. Measuring in situ predation by Mysis relicta and observations on under-dispersed microdistributions of zooplankton. Hydrobiologia 93:109-119 (1982). Vanderploeg, H. A., S. A. Pothoven, D. Krueger, D. M. Mason, J. R. Liebig, J. F. Cavaletto, S. A. Ruberg, G. A. Lang, and R. Ptáčníková. 2015. Spatial and predatory interactions of visually preying nonindigenous zooplankton and fish in Lake Michigan during midsummer. Journal of Great Lakes Research 41:125-142. Viherluoto, M., H. Kuosa, J. Flinkman, and M. Viitasalo. 2000. Food utilisation of pelagic mysids, Mysis mixta and M. relicta, during their growing season in the northern Baltic Sea. Marine Biology 136:553-559. Violle, C., B. J. Enquist, B. J. McGill, L. Jiang, C. H. Albert, C. Hulshof, V. Jung, and J. Messier. 2012. The return of the variance: intraspecific variability in community ecology. Trends in Ecology & Evolution 27:244-252. von Elert, E., and G. Pohnert. 2000. Predator specificity of kairomones in diel vertical migration of Daphnia: a chemical approach. Oikos 88:119-128. Walsh, M. G., B. T. Boscarino, J. Marty, and O. E. Johannsson. 2012. Mysis diluviana and Hemimysis anomala: Reviewing the roles of a native and invasive mysid in the Laurentian Great Lakes region. Journal of Great Lakes Research 38:1-6. Waterstraat, A., M. Krappe, P. Riel, and M. Rumpf. 2005. Habitat shifts of Mysis relicta (Decapoda, Mysidacea) in the lakes Breiter and Schmaler Luzin (NE Germany). Crustaceana 78:685-699. Watkins, J. M., L. G. Rudstam, M. J. Connerton, T. Schaner, P. G. Rudstam, and K. L. Bowen. 2015. Abundance and spatial distribution of Mysis diluviana in Lake Ontario in 2008 estimated with 120 kHz hydroacoustic surveys and net tows. Aquatic Ecosystem Health & Management 18:63-75. Watkins, J. M., P. D. Collingsworth, N. E. Saavedra, B. P. O’Malley, and L. G. Rudstam. 2017. Fine-scale zooplankton diel vertical migration revealed by traditional net sampling and a Laser Optical Plankton Counter (LOPC) in Lake Ontario. Journal of Great Lakes Research 43:804-812. Werner, E. E., J. F. Gilliam, D. J. Hall, and G. G. Mittelbach. 1983. An experimental test of the effects of predation risk on habitat use in fish. Ecology 64:1540-1548. Werner, E. E., and J. F. Gilliam. 1984. The ontogenetic niche and species interactions in size-structured populations. Annual review of ecology and systematics 15:393- 425. Whall, J. D., and D. C. Lasenby. 2009. Differences in the trophic role of Mysis diluviana in two intermontane lakes. Aquatic Biology 5: 281-292.

187

Wiebe, P. H., and M. C. Benfield. 2003. From the Hensen net toward four-dimensional biological oceanography. Progress in Oceanography 56: 7-136. Wiebe, P. H., D. Allison, M. Kennedy, and G. Moncoiffé. 2015. A vocabulary for the configuration of net tows for collecting plankton and micronekton. Journal of Plankton Research 37: 21-27. Williamson, C. E., J. M. Fischer, S. M. Bollens, E. P., Overholt, and J. K. Breckenridge. 2011. Toward a more comprehensive theory of zooplankton diel vertical migration: integrating ultraviolet radiation and water transparency into the biotic paradigm. Limnology and Oceanography, 56:1603-1623. Wurtsbaugh, W. A., and D. Neverman. 1988. Post-feeding thermotaxis and daily vertical migration in a larval fish. Nature 333:846-848. Xu, Y., A. W. Schroth, and D. M. Rizzo. 2015. Developing a 21st Century framework for lake‐specific eutrophication assessment using quantile regression. Limnology and Oceanography: Methods 13:237-249. Yurista, P. M., J. R. Kelly, and S. E. Miller. 2009. Lake Superior zooplankton biomass: alternate estimates from a probability-based net survey and spatially extensive LOPC surveys. Journal of Great Lakes Research 35: 337-346. Zaret, T. M., and J. S. Suffern. 1976. Vertical migration in zooplankton as a predator avoidance mechanism. Limnology and Oceanography 21:804-813.

188