Energy Flow and Food Web Ecology Along a Hydroperiod Gradient

Tiffany A. Schriever

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Ecology and Evolutionary Biology University of Toronto

Tiffany A. Schriever

Doctor of Philosophy

Ecology and Evolutionary Biology University of Toronto

2012 Abstract

Identifying the ecological mechanisms that determine food web structure is critical for understanding the causes and consequences of diversity. The objective of this thesis was to identify the mechanisms structuring aquatic food webs across environmental gradients from a multi-level perspective (individual to ecosystem) using integrative methodology and field experiments to test classic ecological theory. My results demonstrate support for the dynamic constraints hypothesis, which predicts habitats with greater disturbance should have shorter food chains, but are not consistent with the ecosystem size hypothesis that predicts larger ecosystems have longer food chains. Insect and amphibian richness increased with increasing pond size and hydroperiod, indicating that insertion of new consumers into pond communities was driving variation in food-chain length. A multivariate analysis testing the influence of physicochemical variables on food-web characteristics revealed that hydroperiod and pond area had a strong influence on amphibian and invertebrate assemblages, trophic diversity and 15N range. Food- chain length did not respond strongly to any one variable, but instead responded weakly to multiple environmental variables, suggesting interacting influences on food-web structure.

Conversely, the trophic niche of amphibian larvae was not influenced by pond hydroperiod, but did exhibit ontogenetic diet shifts. Populations of amphibian larvae with broader niche widths ii also had increased individual variation, supporting the niche variation hypothesis. In addition, I assessed whether species diversity influenced the magnitude of cross-habitat resource flow between aquatic and terrestrial habitats via emerging aquatic insects, metamorphosing amphibians, and litter deposition. Deposition into ponds far exceeded carbon exported via insect and amphibian emergences. We found a negative relationship between resource flux and the diversity of amphibians and insects, which contradicts the general pattern of positive biodiversity-ecosystem function relationships. My research strongly suggests environmental variation is a key process in shaping food-web structure and function and that multiple methodologies are needed to understand temporal and spatial dynamics of aquatic ecosystems.

iii

Acknowledgments

My first and biggest thank you goes to Devin Bloom. I can’t possibly express in words how grateful I am for all his love, support, and pep-talks. I surely would not have kept any sliver of sanity or had any fun without him around. He has been my greatest supporter, best friend, and editor for too many years to count. I wish I was half as great of a scientist as he is. Thank you and I love you!

I owe an enormous thank you to my advisor, Prof. Dudley Williams. He took a chance and gave me the opportunity to become part of his legacy. Dudley gave me the freedom to develop as an aquatic biologist through his guidance and financial support through the years. I am grateful for the quick turnaround and thorough edits of manuscripts and positive feedback. I owe him a big thank you because he is responsible for my newfound love of invertebrates. I am saddened that I am one of the last members of the Williams’ lab, but humbled to have been part of it.

I am thankful for my PhD advisory committee, Rudy Boonstra, Bob Murphy, and Marc Cadotte. I appreciate all their helpful comments and discussion throughout the different stages of my program.

A big muddy, pond-smelling hug for all those who labored in the field with me: Jennifer Arblaster, Katherine Bannar-Martin, Devin Bloom, Shauna Bloom, Kristen Brochu, Monica Candelaria, Kirsten Comberford, Mark Conboy, Jake Cowper Szamosi, Klara Jaspers-Fayer, Maria Modanu, Stephen Pynn, David Stitt, Caroline Tucker, and Tristan Willis. I am thankful for Fiona Munro who walked around the ponds numerous times with a heavy Trimble GPS unit and gave me my first introduction to GIS. I loved working in the ponds all day with such wonderful company! Caroline, I’m sorry for the nights lacking sleep and bug bites. Kirsten thanks for pointing out all the spiders even though you were not thrilled to see them – everywhere! Kat, Maria, Kristen, Pynn and Devin, you are all workhorses with strong backs and big smiles. Thank you again and again!

I could not have processed as many samples, sorted as many invertebrates, or dried and ground as many amphibians as I did without the amazing help of: Sarah Booth, Siao Ryan Yang, Ruby Sambi, Gaayathiri Jegatheeswaran, Maryam Mahmood, Gagan Gill, David Stitt, Katie Keogh, Kristen Beck, and Christen Dschankilic. Some of these students stuck with me for multiple years iv and even after graduating! I wish you all the best of luck in your academic and life endeavors. Thank you so much for all your countless hours of unpaid work. Sarah and Jenn you are amazing, intelligent, and creative biologists. I can’t wait to see what you bring to the field. Thank you for letting me practice playing the role of advisor on you!

I want to extend a special thanks to Marc Cadotte and his lab members, Caroline, Lanna, Nick, Kelly, and Robin, for welcoming me into their joyful lab. I enjoyed the meetings, get-togethers, birthday cakes and the candy that was always available. Marc you have been kind, extremely helpful, and a very important mentor to me.

I thank Brian Crother, my master’s supervisor, who continually provides spirited support and encouragement. Nate Lovejoy, thank you for writing a reference letter for my new post-doctoral position. I am very grateful for the Toronto Zoo staff and Malcolm Campbell for providing an enjoyable and enlightening way to spend my teaching hours.

My time at UTSC was made tremendously better by the friendship of Emily, Maria, Kristen, and Megan. My lab mates: Catherine, Kevin, Julie, Sarah, Jessica, Oksana, and Judith provided guidance, support, and good times. I will miss the walks in the valley, lunch on the patio, and your smiles.

I have enjoyed my time in Toronto, although I have a love-hate relationship with Toronto. Toronto has been a fun city to have called home for five years, but I will be happy to leave the traffic behind. I love the friends I’ve met, of whom I will miss greatly. My newly acquired family of Newfoundlanders: Dave, Steve, Kirsten, and Danielle (yes, you are guilty by association) were always there to provide awesome food, wine, laughs, and encouragement. Karen and Hernán I am blessed to have you as friends. Until we all meet again!

Lastly, I am very grateful for a family that has provided support and encouragement through the many tiring years of graduate work. Mom and Dad, South Dakota will always be my home regardless of where I live. I am grateful for the Blooms as they have always supported me, pushed me to be confident, and provided entertainment. My heart is continuously filled with love from my family and for that I thank you all for giving me the strength to pursue my dreams.

If I forgot someone, please forgive me. So many people have helped in some way and I am sincerely grateful to all of you. v

Chapter Acknowledgments

This thesis is comprised of four papers that are either submitted, in review, or in preparation for publication in peer-reviewed scientific journals (Chapters 2 – 5). Co-author Marc Cadotte provided statistical expertise, editorial comments, and consultation on research topic (Chapter 4). Chapter 1 provides the framework for this dissertation research and Chapter 6 summarizes the major findings and importance of this work in the field of Ecology.

Schriever, T.A. and Williams, D. D. Accepted. Ontogenetic and individual diet variation in amphibian larvae across an environmental gradient. Freshwater Biology

Schriever, T.A. and Williams, D. D. In review. Duration matters, size does not in pond food webs: a study of the dynamic constraints and ecosystem size hypotheses. Oecologia

Schriever, T.A., Cadotte, M. W. and Williams, D. D. In preparation. How hydroperiod and species richness affect the balance of resource flows across aquatic-terrestrial habitats. Ecosystems

Schriever, T. A. In preparation. Food webs in relation to environmental and community variation: a multivariate approach.

Table of Contents

Contents

Acknowledgments...... iv

Chapter Acknowledgments ...... vi

Table of Contents ...... vii

List of Tables ...... xi

List of Figures ...... xii

List of Appendices ...... xvi

1 Introduction ...... 1

1.1 Freshwater hydroperiod gradient ...... 2

1.2 Food web ecology ...... 3

1.3 Use of stable isotopes in ecology ...... 4

1.4 Aquatic-terrestrial linkages ...... 4

1.5 Thesis objectives ...... 6

2 Ontogenetic and individual diet variation in amphibian larvae across an environmental gradient ...... 9

2.1 Abstract ...... 10

2.2 Introduction ...... 10

2.3 Methods...... 12

2.3.1 Study area...... 12

2.3.2 Field collecting and laboratory methods ...... 13

2.3.3 Gut content analysis (GCA) ...... 13

2.3.4 Stable isotopes analysis (SIA) ...... 14

2.4 Data analysis ...... 15

2.4.1 Quantifying diet variation and niche width ...... 15

vii

2.4.2 Statistical analyses ...... 18

2.5 Results ...... 19

2.5.1 Niche variation across hydroperiod gradient ...... 19

2.5.2 Diet and trophic niche ...... 20

2.5.3 Trophic position ...... 21

2.5.4 Ontogenetic diet shifts ...... 22

2.6 Discussion ...... 23

2.6.1 Ontogenetic diet shifts ...... 27

2.6.2 Conclusions ...... 28

2.7 Acknowledgements ...... 29

2.8 Figures...... 30

3 Duration matters, size does not in pond food webs: A study of the dynamic constraints and ecosystem size hypotheses ...... 36

3.1 Abstract ...... 37

3.2 Introduction ...... 37

3.3 Methods...... 40

3.3.1 Study sites ...... 40

3.3.2 Field collection of pond communities...... 40

3.3.3 Characterization of ecosystem size and environmental variation ...... 41

3.3.4 Insect community composition ...... 42

3.3.5 Sample preparation and stable isotope analysis ...... 42

3.3.6 Food-chain length and trophic position calculation ...... 43

3.3.7 Data analysis ...... 44

3.4 Results ...... 45

3.4.1 Food-chain length and environmental variation ...... 45

3.4.2 Proximate mechanisms underlying changes in FCL ...... 45

viii

3.5 Discussion ...... 47

3.5.1 Ultimate determinants of FCL ...... 47

3.5.2 Proximal mechanisms underlying FCL variability ...... 49

3.6 Acknowledgments...... 51

3.7 Figures...... 52

4 Reciprocal resource flows in temporary ponds: the importance of community composition and hydroperiod ...... 59

4.1 Abstract ...... 60

4.2 Introduction ...... 60

4.3 Methods...... 63

4.3.1 Study sites and physical habitat ...... 63

4.3.2 Field and laboratory methodology ...... 63

4.3.3 Statistical analyses ...... 65

4.4 Results ...... 66

4.4.1 Timing of resource flows and environmental influences ...... 66

4.4.2 Magnitude of fluxes ...... 66

4.4.3 Diversity-ecosystem function response ...... 67

4.4.4 Influence of pond area and hydroperiod ...... 68

4.5 Discussion ...... 69

4.5.1 Diversity-ecosystem function relationship ...... 71

4.5.2 Conclusions ...... 72

4.6 Acknowledgements ...... 73

4.7 Tables ...... 74

4.8 Figures...... 76

5 Food webs in relation to environmental and species assemblage variation: A multivariate approach ...... 82

5.1 Abstract ...... 83 ix

5.2 Introduction ...... 83

5.3 Methods...... 86

5.3.1 Study sites ...... 86

5.3.2 Field data collection ...... 87

5.3.3 Laboratory and stable isotope analysis ...... 88

5.3.4 Statistical analyses ...... 90

5.4 Results ...... 91

5.4.1 Physical-chemical pond environment ...... 91

5.4.2 Insect and amphibian community-environment relationships ...... 91

5.4.3 Food web-environment relationships ...... 94

5.5 Discussion ...... 96

5.5.1 Summary ...... 98

5.6 Acknowledgments...... 98

5.7 Tables ...... 100

5.8 Figures...... 102

6 Thesis summary ...... 107

6.1 Individual, population and ontogenetic trophic niches of amphibian larvae ...... 107

6.2 Hydroperiod determines food-chain length ...... 108

6.3 Less diverse temporary ponds boast higher subsidy to surrounding terrestrial habitat .108

6.4 Importance of multiple environmental influences on aquatic community and food-web structure ...... 109

7 References ...... 110

8 Appendices ...... 124

List of Tables

Table 4.1 Physical characteristics of the study ponds...... 74

Table 4.2 Summary of flux values for insects and amphibians. Emergent flux per m pond margin (F) takes into consideration the shoreline developmental factor (DL), total flux is the total aquatic to terrestrial flux of carbon (Etotal)...... 75

Table 5.1 Hydroperiod (days with water) of each pond...... 100

Table 5.2 Eigenvalues, percentage of variation explained, and permutation test statistics from CA of the invertebrate and amphibian presence-absence and CCA of invertebrate and amphibian communities and environmental data from ponds in Southeastern Ontario...... 101

List of Figures

2.1 Relationship between diet variation among individuals (V) and population total niche width (TNW). The empirical results are shown by filled symbols and solid regression line. Open symbols and dashed regression line indicate the expected trend under a null model in which diet results by individuals randomly sampling from a limited set of prey from a shared prey distribution...... 30

2.2 Mixing polygon showing 13C and 15N signatures of four food sources and their contribution to the diet of L. sylvaticus (wood frog tadpoles and metamorphs). Histograms represent feasible source contributions from each source to wood frog diet (denoted by picture). Values in boxes are the 1  99th percentile ranges of source distributions with the y-axis scale 0 to 15%...... 31

2.3 Mixing polygon for 13C and 15N signatures of seven food sources for A. laterale larvae. Histograms represent feasible source contributions from each source to Blue spotted salamander diet (denoted by picture). Values in boxes are the 1  99th percentile ranges of source distributions with the y-axis scale 0 to 30%...... 32

2.4 Carbon and nitrogen bi-plot showing A. laterale shown in filled squares (n = 34) and L. sylvaticus grey filled circles (n = 62)...... 33

2.5 Trophic position calculated for L. sylvaticus throughout development using gut contents

(TPGC: open circles and solid line) and stable isotope data (TPSI: filled circles and dotted regression line)...... 34

2.6 Isotopic variation across size range of A. laterale A13C and B) 15N and variation across development for L. sylvaticus C13C and D) 15N. Linear regression line indicates significant positive relationship between 13C and stage for L. sylvaticus and marginally significant relationship between 15N and SVL for A. laterale...... 35

3.1Relationships between FCL (maximum TP) and a) environmental variation (i.e., hydroperiod 2 in days, GLM: FCL = 1.0299 log (hydroperiod) + 1.0237, R = 0.23, P = 0.03), b) ecosystem size

xii

2 (GLM: FCL = 0.5723 log (area) + 1.5265, R = 0.25, P = 0.09), and c) variation in water depth 2 (GLM: FCL = -0.5687CV water depth (m) + 3.7191, R = 0.38, P = 0.05)...... 52

3.2 Regression demonstrating significant relationship between the mean trophic positions of the four common apical predators and the FCL (maximum trophic position) which provides evidence for the omnivory and insertion mechanisms...... 53

3.3 Relationship between number of predatory insect families and ecosystem size (i.e., pond area (m2)...... 54

3.4 Relationship between predatory insect family richness and environmental variation (hydroperiod)...... 55

3.5 Non-linear relationship between breeding amphibian species richness and environmental variation (R2 = 0.57, P = 0.002)...... 56

3.6 Relationship between breeding amphibian species richness and ecosystem size (R2 = 0.55, P = 0.02)...... 57

3.7 Comparison of FCL variability amongst ponds and between sampling years. Intermediate hydroperiod ponds are denoted by Int and a corresponding number. Short3 and Int4 were not sampled in 2008...... 58

4.1 Temporal variation in insect and amphibian emergence throughout the study period. Insect collection was from day 124 to 230 (May 4th to August 18th), amphibian collection occurred between day 171 and 241 (June 20th to August 29th). Each point represents the sum abundance across ponds per collection day. Overall, amphibian peak emergence occurred between days 174 and 176 (23-25 June), while insect peak emergence occurred between days 200 and 202 (19-21 July)...... 76

4.2 Flux (mean and standard error) attributed by each A) and B) insect family and C) amphibian species collected. Odonata is separated from other families for better viewing...... 77

4.3Comparison of total carbon flux from pond-to-land (Etotal) (upper portion of figure) and land- to-pond (Dtotal) (lower portion of figure) attributed from each resource across study ponds...... 78

xiii

4.4 Variation in the magnitude of litter components received in each pond. Litter components are: deciduous leaves, miscellaneous (bud scales, pollen, etc.), aquatic plants (Family Equisetaceae; horsetail), coniferous needles, and woody material (twigs, branches, bark, cones, etc.). Note y-axis is not on the same scale for each pond...... 79

4.5 The effect of A) insect and amphibian diversity and B) species richness on the magnitude of cross habitat energy flow, C) total insect flux (Eitotal) and insect species diversity relationship, and D) insect (open symbols) and amphibian (filled symbols) diversity within each pond on the magnitude of cross habitat energy flow...... 80

4.6 The influence of A) hydroperiod and B) pond size on the flux (g C m-1 yr-1) of three types of spatial subsidies. The regression lines represent the significant relationship between amphibian flux and hydroperiod and amphibian flux and pond area. Flux values were log transformed for better visualization of relationship...... 81

5.1 Canonical correspondence analysis (CCA) ordination biplot showing the association between the presence-absence of invertebrate families to A) four environmental predictors in 2008 (arrows) from seven ponds and B) seven environmental variables (arrows) from nine ponds collected in 2009. The direction of the arrow indicates direction of maximum change and the length is proportional to the rate of change of that variable. See Appendix R for full family names...... 102

5.2 Canonical correspondence analysis (CCA) biplots showing the association of amphibian species presence-absence and A) four environmental predictors (arrows) in 2008 from seven ponds and B) five environmental variables (arrow) from eight ponds collected in 2009. Species placement in biplot is marked with an open circle...... 103

5.3 Ordination plot showing main variation in food-web structure (n = 4 variables, black arrows) in response to environmental variation (n = 4 variables, grey arrows) summarized by redundancy analysis (RDA) for 2008. Sampling sites are denoted by black outlined grey circles. MaxTP is the trophic position of the top predator or food-chain length, TAfoodweb is the total area encompassed by all species in δ13C–δ15N bi-plot space. Plot used scaling = 1 to create a distance biplot where objects approximate their Euclidean distances in the space of response variables. Length of the arrow represents the strength of the gradient. Arrows that are directed in opposite

xiv directions are negatively correlated. The angles between environmental and food-web variables reflect their correlations...... 104

5.4 Redundancy Analysis (RDA) biplot illustrating the relationship between environmental variables (n = 4, grey arrows), food-web metrics (n = 4, black arrows), and ponds (n = 7) in 2009. Black outlined grey dots are sample sites with pond name adjacent to dot. Plot used scaling = 1 to create a distance biplot where objects approximate their Euclidean distances in the space of response variables. Length of the arrow represents the strength of the gradient. Arrows that are directed in opposite directions are negatively correlated. The angles between environmental and food-web variables reflect their correlations...... 105

5.5 RDA biplot displaying the variation in food-web structure (black arrow) in relation to six environmental variables (grey arrows, excluding hydroperiod) in 2009. Plot used scaling = 1 to create a distance biplot where objects approximate their Euclidean distances in the space of response variables. Length of the arrow represents the strength of the gradient. Arrows that are directed in opposite directions are negatively correlated. The angles between environmental and food-web variables reflect their correlations...... 106

List of Appendices

Appendix A Sample sizes per pond for A. laterale...... 124

Appendix B Mean stable isotope values and standard deviation (SD) for basal resources, macroinvertebrates, and amphibian larvae collected from ponds ranging in hydroperiod...... 125

Appendix C Mean relative importance (IRI) and standard error of different diet items for L. sylvaticus (n = 45 tadpoles, n = 21 metamorphs). Diatoms consisting of species in the genera Fragilaria, Synedra, Cymbella, Navicula, Gomphormena, Cyclotella, and Melosira, plant material included pollen and seeds, hyphomycete fungi, and arthropods included chitinous fragments, chironomid larvae, copepods, and largely unidentifiable insect pieces (but some legs and wings)...... 126

Appendix D Relative importance (IRI) of diet items (mean and SE) found in the stomachs of A. laterale (n = 34). Protozoans included primarily ciliates...... 127

Appendix E Mean trophic position (standard deviation) calculated from gut contents (TPGC) and stable istotopes (TPSI) of two amphibian species collected during larval aquatic stages across a hydroperiod gradient. NP denotes ponds were species were not present...... 128

Appendix F Occurrence of breeding amphibians in ponds sampled during 2008 and 2009. If present in only one year, year present is in parenthesis. Short3 and Inter4 were only sampled in 2009...... 129

Appendix G Invertebrate community composition in ponds sampled in 2008. Presence is represented by 1. Inter is abbreviated for intermediate hydroperiod ponds. Predatory families are in bold font...... 130

Appendix H Invertebrate community composition in ponds sampled in 2009. Presence marked by 1. Inter is abbreviated for intermediate hydroperiod ponds. Predatory families are in bold font...... 134

Appendix I Macroinvertebrate length–biomass regressions. Macroinvertebrate taxa list with regression equations and references used to estimate biomass. Biomass was estimated as:

xvi biomass DM = a(length)b, where DM is dry mass (mg), length is total body length (mm), and a and b are constants...... 140

Appendix J Emergent aquatic insects length, mass, and numbers collected...... 143

Appendix K Relationships between larval snout-to-vent length (SVL; mm) and larval ash-free dry mass (AFDM). Predictive equations reported with range of values used to develop relationship and sample size (n). Significance codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’...... 146

Appendix L The correlation between insect and amphibian emergence (no./day) and pond water temperature (◦C). Each point represents the total number of individuals collected in a day. On average 20% of the variance in insect and 18% of the variance in amphibian emergence abundance can be explained by variation in pond water temperature (Spearman correlation coefficient)...... 147

Appendix M Pond water depth declined in each pond during the sampling period. Warmer water temperatures were associated with shallower water (pond water depth and temperature shared an average of 0.54 of the variance; Spearman correlation). The day in the year was a strong predictor of water depth as later in the year the ponds were shallower (Pearson correlation coefficient r(35) = -0.78 no separation of ponds or mean across ponds = -0.92)...... 148

Appendix N Association of ponds based on a CA of insect family presence-absence in 2008. 149

Appendix O Association of ponds based on a CA of amphibian species presence-absence in 2008...... 150

Appendix P Association of ponds based on a correspondence analysis (CA) of invertebrate family presence-absence in 2009...... 151

Appendix Q Correspondence analysis (CA) of amphibian species presence-absence in 2009. . 152

Appendix R Abbreviations used in multivariate analyses, CA and CCA, for invertebrate families. NA represents taxa that were not present in both years of sampling. Abbreviation in parentheses is the code used in 2009, if different from 2008 abbreviation...... 153

xvii 1

1 Introduction

The following introductory sections will provide fundamental information for the main themes and analyses presented in the following chapters.

1.1 Freshwater hydroperiod gradient

Temporary waters are extremely diverse ecosystems spanning small rain pools to large wetlands, such as of the Florida Everglades or the vast llanos lowlands of tropical South America, and may be fresh or saline. I restrict my coverage to temporary water bodies that range from intermittent freshwater woodland ponds to freshwater marshes. Temporary waters are among the most imperiled freshwater ecosystems (Colburn 2004) under threat from habitat destruction and climate change (Williams et al. 2009), which is unfortunate given the relatively high biodiversity in these less-studied systems. Many diverse and abundant assemblages of amphibian larvae reside in freshwater habitats, but these are often less well known compared to the invertebrate or fish biotas (Altig and Johnston 1989). Temporary ponds harbor the highest diversity of tadpoles compared to permanent habitats (Starrett 1973, Halliday 2005). Unfortunately, such habitats are often viewed as expendable, overlooked in policy making, scientific research, and conservation. For example, 50% of historic wetlands in the United States have been destroyed (Dugan 1993) and 72% have been lost in Ontario since 2002 (Federal 2010).

My dissertation examines the spatial and temporal variability in aquatic food webs in freshwater ponds along a hydroperiod gradient. Hydroperiod is the aquatic phase of an ephemeral water body. Lentic (still water) habitats are aligned along a gradient from temporary, to permanent habitats without fishes, to permanent habitats with fishes, and with variation in both abiotic (e.g., drying and associated changes in water chemistry) and biotic (e.g., increasing predation pressure) properties. Although temporary waters are abundant, diverse in physical and biological characteristics, and common globally, little is known about their food web structure especially in the context of their dynamic physical environment and energy connection to the surrounding landscape. Aquatic communities of ephemeral water bodies are constrained by the hydroperiod, where predictable shifts in invertebrate (Brooks 2000, Stoks and McPeek 2003, Williams 2006, Turner and Montgomery 2009) and amphibian (Skelly 1996, Wellborn et al. 1996, Snodgrass et al. 2000, Eason and Fauth 2001, Babbitt et al. 2003) species richness and diversity occur.

Hydroperiod and predation are the main influences that organize aquatic communities at the permanence transition between temporary habitats and fishless permanent habitats (Wellborn et al. 1996). These natural, low-disturbance ponds host a diverse amphibian and invertebrate community (Wiggins et al. 1980, Williams 1997, Brooks 2000, Williams 2006) and provide

3 optimal baseline for comparing trophic structure among ponds. Further, as temporary ponds exist along continua of temporal and spatial scales (Wilbur 1997) the best way to compare the structural and functional complexities of natural communities is via a gradient design.

I studied intermittent woodland freshwater ponds in southeastern Ontario that varied in hydroperiod and size. Some ponds filled predictably from snow melt in early spring and dried in early-to-mid summer, while others dried in late summer but refilled with autumn rains. The ponds that refilled were frozen throughout the winter months and thawed in early spring. The permanent ponds lost water from evapotranspiration during the summer, but never completely dried. All ponds were devoid of fishes. My study ponds were located in or near Queen’s University Biological Station, 50km North of Kingston, Ontario. The area sits on the Frontenac Axis ecoregion containing mixed forest, exposed bedrock outcroppings, and an assortment of aquatic habitats ranging from large lakes, forested swamps, marshes and temporary ponds.

1.2 Food web ecology

Food webs depict trophic interactions among species or individuals in an ecosystem (Post 2002a) and are a central concept in ecology because they give insight into population dynamics, community ecology (Polis and Strong 1996, Wilbur 1997), and ecosystem level processes (DeAngelis 1992). Food webs are spatially and temporally variable (Paine 1980, Warren 1989, Winemiller 1990, 2006), are influenced by the surrounding landscape, and by the movement of nutrients and energy among habitats (Polis et al. 1997, Nowlin et al. 2007). Changes in trophic structure, energy and nutrients, and other abiotic and biotic factors may influence food-web dynamics (Bengtsson and Martinez 1996). A food web is composed of multiple food chains, or linear routes to the top predators. The maximum height of the food web, the number of trophic transfers from the basal resources to the apical predator, is the food-chain length. Food-chain length is an important regulator of community structure and ecosystem function (Hairston et al. 1960, Duffy et al. 2005). Environmental variation can have substantial influences on food web structure (Winemiller 1990, Winemiller and Polis 1996, Sabo and Post 2008, Sabo et al. 2009, Sabo et al. 2010), and thus it is important to understand its influence on food-chain length. Aspects of food chain length, trophic cascades, interaction strength, complexity, food web theory, and food web analysis have played major roles in the field of food web ecology.

1.3 Use of stable isotopes in ecology

Food webs track the flow of energy through a community. Links between resource and consumer can be inferred from direct feeding observations, stomach-contents analysis, and stable isotope analysis. Stable isotopes can provide broad inferences on diet and trophic relationships, and are used to develop models of community trophic structure. They have been increasingly used in ecology to understand food web structure and function following a few seminal works by (Deniro and Epstein 1978, 1981, Minagawa and Wada 1984, Peterson and Fry 1987). Stable isotopes are naturally abundant and can effectively trace energy flow within and between ecosystems. Stable carbon isotope ratios (13C ) measure the derived energy source (Peterson and Fry 1987, Fry 1991) because different energy sources often show distinct 13C signatures with little fractionation up the food chain (Hecky and Hesslein 1995). Nitrogen stable isotope ratios (15N) predictably increase through trophic transfer up the food chain, providing a reliable measure for estimating trophic position (Cabana and Rasmussen 1996, Vander Zanden and Rasmussen 1999) in a food web. Trophic position is a continuous measure of diet, estimating an organism’s place within a food web.

Traditionally, gut contents analysis has been used to describe the diet and trophic position of individuals and populations. However, gut contents studies are time and labour intensive, require large sample sizes, and give only “snapshot” information on what an animal has ingested. Using stable isotopes to define trophic niche is a powerful complement to gut contents analysis (Newsome et al. 2007), because stable isotopes integrate not only what food was ingested but also what resources were assimilated in the tissue over time (Peterson and Fry 1987). Stable isotopes have been widely used in lake and river ecosystem studies, usually in assessing the importance of fishes in these systems. Rarely have stable isotopes been used to investigate the trophic ecology of amphibians and reptiles (Vanderklift and Ponsard 2003, Dalerum and Angerbjorn 2005). By integrating data from stable isotopes and gut contents, it is possible to obtain a detailed description of a species’ trophic niche, which in turn can be used to test hypotheses about the mechanisms driving niche variation.

1.4 Aquatic-terrestrial linkages

The importance of aquatic-terrestrial linkages is well established and has been demonstrated to create resource subsidies between diverse habitat types. For example, oceans subsidize terrestrial

5 primary production through the transport of nutrients via marine mammal carrion, algal wrack, seabird guano (Polis and Hurd 1996, Anderson and Polis 1998), sea turtle egg deposition (Vander Zanden et al. 2012), and spawning migrations of salmon (Gende et al. 2004). Emerging freshwater insects can increase the abundance of the riparian arthropod assemblage (Hoekman et al. 2011, Dreyer et al. 2012). Organisms can also transport contaminants from one habitat to another (O'Toole et al. 2006, Unrine et al. 2007).

Habitats are open to the flow of nutrients, detritus, or organisms, a spatial subsidy that is donor controlled and has the potential to influence population, consumer-resource, food web, and community dynamics of the recipient system (Polis et al. 1997). The magnitude and rate of resource subsidy can influence the variation and strength of trophic cascades (Leroux and Loreau 2008). Woodland ponds are reciprocally linked to the surrounding terrestrial habitat via the transfer of energy out of ponds via insects and amphibians and input from terrestrial carbon sources. Terrestrial primary production, in the form of leaf litter, is a significant allochthonous resource subsidy in woodland ponds, forming the base of the aquatic food web (Batzer and Palik 2007). Intermittent ponds are an integral component of forest ecosystems, especially for organisms with biphasic life cycles, such as amphibians and insects; insects and amphibians have aquatic larval stages that depend on productive temporary water bodies to acquire the necessary resources to grow and metamorphose before entering the terrestrial stage of their life cycle.

There are few studies that have quantified the movements of amphibians and insects from temporary ponds and their effects on terrestrial food webs. Given the extensive occurrence of these water bodies in the global landscape, ignoring their connections to surrounding habitat hinders our understanding of system dynamics. One way to understand this is by examining organisms that link both systems. When salamanders and frogs breed they deposit nutrient- and energy-rich eggs in aquatic habitats, oftentimes newly formed intermittent forest ponds. However, this energy flow is not unidirectional because developing larvae convert aquatic resources into biomass, which is later carried onto land at metamorphosis. Pond-derived energy can be used as an energy subsidy to the terrestrial habitat in three ways: 1) as a prey source for terrestrial vertebrate predators; 2) as a nitrogen and phosphorus source, via excretion, for soil and plant uptake; and 3) after death, as a release of nutrients for decomposers and primary producers. Quantification of these energy transfers is important in recognizing cross-habitat linkages,

6 ecosystem processes, and energy budgets. As far as I know, there are no studies that have quantified the bidirectional movement of resource subsidies from intermittent ponds.

1.5 Thesis objectives

I tested classic ecological hypotheses from a novel approach using an environmental gradient, integrated analyses, and incorporated spatial and temporal variation in complete food webs. I was interested in how environmental variation influences species distributions, community dynamics, and ecosystem structure and function. I studied ecosystem processes from the merged perspective of landscape and food web ecology.

The chapters in this thesis span from the individual to the ecosystem level, and provide insight into the mechanisms driving food web structure, community composition, and habitat connections. Below, I briefly introduce each chapter, the primary objectives and hypotheses tested.

Chapter 2: Feeding ecology, ontogenetic diet shifts, and niche differentiation in amphibian larvae

In this chapter I characterize the trophic niche of larval amphibians, and using stable isotope and gut contents analyses, test whether ontogenetic diet shifts take place. I wanted to determine whether a shift in diet resulted in a change in trophic position and whether diet differed among species, individuals, and ponds across a hydroperiod gradient. I tested whether populations with wider trophic niches would show a higher degree of individual specialization as predicted by the niche variation hypothesis (Bolnick et al. 2003a). This study is one of the few to take a community approach and to quantify the relationship between niche width and specialization for multiple species from different trophic levels.

Chapter 3: Food web ecology of ponds along a hydroperiod gradient

I tested the ecosystem size and dynamic constraints hypotheses, and using carbon and nitrogen stable isotopes, evaluated the proximate mechanisms underlying food chain length variability in nine ponds spanning a hydroperiod gradient. Environmental features of ecosystem size, disturbance, and resource availability are often discussed as potential determinants of food-chain length, with a common focus being to link the proximate mechanisms underlying food web

7 structure to the relationship between food-chain length and environmental factors. My research is novel in its use of a natural environmental gradient to ascertain which environmental factors are important determinants of food-chain length in ephemeral systems. Further, I examined the role of the underlying mechanisms causing variation in food-chain length.

Chapter 4: Linking aquatic and terrestrial ecosystems

In this chapter, I test for the influence of hydrology, ecosystem size, and temperature on the flow of resource subsidies across the aquatic-terrestrial boundary. I quantified the abundance, biomass, and carbon content (energy) of emerging insects and amphibians to measure energy flow leaving ponds. Insect production is generally high in temporary waters and can represent a substantial subsidy to terrestrial habitats (Gratton et al. 2008), but the subsidy from amphibians is unknown. Additionally, since many temporary ponds are heterotrophic systems that rely upon allochthonous inputs of leaf litter as their primary energy source (Bonner et al. 1997, Rubbo and Kiesecker 2004, Rubbo et al. 2006), I quantified litter biomass and carbon inputs into each pond. Given the increasing global loss of amphibian species and wetland habitats, understanding how such ecosystems are linked will provide insight into the possible consequences of severed ecosystem linkages. Compounding this is the stark realization that little is known about wetland biodiversity and ecosystem functioning. Many studies have shown that greater biodiversity results in higher ecosystem functioning (i.e., productivity and energy flow; (Loreau et al. 2001), yet most of their findings come from artificially constructed communities. I present new insights into the biodiversity-ecosystem function relationship in natural communities and how the pattern is influenced by environmental variability.

Chapter 5: A multivariate perspective on food webs in relation to environmental variation

The final chapter expands on chapter 3 by taking a multivariate approach to disentangle the abiotic and biotic factors that determine food-web structure in a dynamic ecosystem.

The final chapter details the invertebrate and amphibian assemblages’ response to a suite of environmental factors, and uses multivariate statistics to investigate how food-web structure responds to environmental variation. Ecologists have long thought environmental factors such as ecosystem size, disturbance, and productivity (i.e., resource availability) determine food-chain length. However, recently the history of community organization has been implicated as an

8 important factor interacting with environmental determinants to cause variation in food-chain length (Post 2002a, Takimoto and Post 2012). Taxonomic diversity and richness of ephemeral ponds often increases with increased hydroperiod (Pechmann et al. 1989, Brooks 2000, Batzer et al. 2004, Williams 2006) and has been linked to changes in the physicochemical environment (Angelibert et al. 2004, Magnusson and Williams 2006). Structural mechanisms such as insertion of new intermediate community members or additions of predators, influence food-chain length (Post and Takimoto 2007). These events occur along the hydroperiod gradient from intermittent to permanent ponds. Therefore, the best way to study the abiotic and biotic influences on food- web structure is by characterizing community patterns across an environmental gradient. Surprisingly, few studies have tested how the biological and physicochemical properties in tandem influence food web structure. Using multivariate statistics, I test the hypothesis that ponds with a longer hydroperiods will be larger, have higher species richness, and have a longer food-chain length and 15N range than ponds with shorter hydroperiods.

2 Ontogenetic and individual diet variation in amphibian larvae across an environmental gradient

2.1 Abstract

Variation among individuals within size or age classes can have profound effects on community dynamics and food web structure. I investigated the potential influence of habitat disturbance on intrapopulation niche variation. Amphibians occupy a range of lentic habitats from short hydroperiod intermittent ponds to long hydroperiod permanent ponds; however, the influence habitat heterogeneity has on trophic niche is poorly understood. I quantified ontogenetic diet variation and individual specialization in wood frog tadpoles (Lithobates sylvaticus) and blue- spotted (Ambystoma laterale) salamander larvae to investigate the influence of hydroperiod on population niche width across a natural hydroperiod gradient using stable isotope and gut contents analyses. In one of the few tests using larval forms, I tested the niche variation hypothesis, which predicts that populations with larger niche widths also have increased individual variation. My results support the niche variation hypothesis indicating that more generalized populations exhibit higher among-individual diet variation. I report gradual changes in the relative importance of diet items, decreased dietary overlap, and increased trophic position in L. sylvaticus throughout development. A. laterale tissues showed increased 15N throughout its larval period. I did not find a relationship between hydroperiod and niche parameters, indicating that niches are conserved across heterogeneous habitats. In contrast to most documented cases, we estimated low levels of individual specialization in amphibian larvae. Amphibians are an important link between aquatic and terrestrial ecosystems, whereby diet shifts can influence food web structure by altering energy flow pathways and the trophic position of higher consumers ultimately changing food-chain length.

2.2 Introduction

The ecological niche concept is defined for and most often applied to species-level resource use (Hutchinson 1957). Further, food webs portray trophic interactions among species within an ecosystem (Winemiller and Polis 1996). However, the realized niche of a species represents the sum of prey selection by individuals within a population. Variation in resource use among individuals within a population arises from ontogenetic diet shifts (Polis 1984, Werner and Gilliam 1984, Post 2003, Frederich et al. 2010), morphological differences (Svanbäck and Eklöv 2002, Bolnick and Paull 2009), and sex (Donnelly 1991, Sandlin and Willig 1993). Trophic

11 changes among individuals can alter food web structure and function, therefore, understanding individual resource use is fundamental to the study of trophic ecology.

Intra- and interspecific competition are often inferred to be selection agents accounting for individual variation (Roughgarden 1972, Bolnick et al. 2003b, Svanbäck and Persson 2004, Svanbäck and Bolnick 2005, 2007). The prediction that a release from interspecific competition facilitates an increase in population niche width because individuals are free to become more specialized (i.e., between-individual variation increases) was formalized as the “niche-variation hypothesis” or NVH (Van Valen 1965). Recently, several authors have found support for Van Valen’s NVH by demonstrating a positive relationship between the degree of individual specialization and population total niche width (Bolnick et al. 2007, Costa et al. 2008, Araújo et al. 2009) and have shown that individual specialization is more pervasive in natural populations than originally thought (Bolnick et al. 2003a). This realization has generated renewed interest in identifying and quantifying individual diet variation. Additionally, increased intraspecific competition can also drive increased individual niche width (Svanbäck and Bolnick 2007, Araújo et al. 2011). Therefore, two possibly opposing processes can produce increased individual specialization.

Environmental heterogeneity can create niche variation within and among populations (Vander Zanden et al. 2000) because interactions among prey availability, competition pressure, and physiological stress can alter foraging success. However, studies focusing on how environmental variation influence niche width is less common than studies that look at competition as driving niche variation. In an aquatic system, habitat duration (i.e., hydroperiod) influences species richness, diversity, community composition, survivorship (Wellborn et al. 1996, Babbitt et al. 2003, Urban 2004, Werner et al. 2007), growth and development (Rowe and Dunson 1995), and competition (Werner and McPeek 1994, Wilbur 1997) which can lead to alterations in trophic structure and food web dynamics (Persson et al. 1996). Hydroperiod exhibits spatial (among ponds) and temporal variation (across years), which can directly cause mortality and limit species colonization. Since the hydroperiod is such a strong structuring force, it may also influence the trophic ecology of resident organisms.

Amphibian larvae occupy a broad range of freshwater lentic habitats from short hydroperiod intermittent ponds to long hydroperiod permanent ponds, and can occupy multiple trophic levels,

12 thus they provide a useful system in which to examine the relationship between habitat duration and trophic niche. Resource partitioning in anuran larvae can be explained by the interaction among habitat factors (e.g., pond duration), predation, and competition (Toft, 1985); however, only a few studies have investigated food resource partitioning in these animals. Tadpoles are generally deemed herbivores (Alford 1999), but have been shown to be omnivorous (Seale 1980, Duellman and Trueb 1994, Stebbins and Cohen 1995) and macrophagous predators (Petranka and Kennedy 1999), whereas salamander larvae are carnivorous and consume an array of small insects (Petranka 1998). I investigated whether these generalist populations consist of specialist individuals with divergent diets.

This study examines the influence of habitat variation on resource use in two common pond inhabitants of northeastern North America: wood frogs (Lithobates sylvaticus LeConte) and blue-spotted salamanders (Ambystoma laterale Hallowell). Using stable isotope and gut contents analyses, I tested: 1) for the presence of ontogenetic diet shifts and 2) individual dietary specialization; 3) whether populations with wider trophic niches would show a higher degree of individual specialization as predicted by the niche variation hypothesis (Bolnick et al., 2003); and 4) the effect of hydroperiod on a species’ trophic niche.

2.3 Methods

2.3.1 Study area

The study sites were seven ponds located at the Queen’s University Biological Station north of Kingston, Ontario, Canada (44.565977 N, -76.324223 W). The area is predominantly mixed coniferous-deciduous forest. These ponds are filled by snow melt in early spring and lose water as the summer progresses, resulting in some ponds drying completely. The hydroperiod of these seven ponds ranged from ‘short’ (64 days, n = 2), ‘intermediate’ (156 to 186 days, n = 3), to ‘long’ (those that did not dry during the study period, ~365 days, n = 2). This gradient spanned intermittent freshwater woodland ponds to permanent freshwater marshes, all of which are naturally fishless with predaceous insects and/or salamander larvae as top predators. Breeding within and among ponds was nearly synchronous for L. sylvaticus and A. laterale, with calling commencing from 17 to 24 April in 2008.

2.3.2 Field collecting and laboratory methods

In order to get a representative sample of developmental stages and sizes, I collected larval amphibians from each pond bi-weekly from May through September 2008 by dip net. Collected individuals were euthanized by immersion in Tricaine methanesulfonate (MS-222) and then frozen until further processing. Species were identified using tadpole (Altig 1970) and larval (Petranka 1998) keys, and developmental stages were determined for tadpoles using Gosner (1960). Tadpoles of L. sylvaticus were present in four of the seven ponds and larval A. laterale were present in all ponds. Each month I collected the following basal resources: detritus, aquatic plants, algae, and seston. The most abundant macroinvertebrates were collected by hand and by dip net from each pond. Detritus was comprised of primarily decaying leaf litter and plant tissue. Seston was acquired by collecting a 1 L water sample at mid-depth in the water column of each pond. I filtered water samples (i.e., seston), which contained zooplankton, bacteria, and phytoplankton onto pre-combusted Whatman glass microfiber GF/F filters. Basal resources were used to identify pond baselines for calculation of trophic position using stable isotopes, to quantify changes in niche width, and to assess ontogenetic changes in diet.

2.3.3 Gut content analysis (GCA)

I analyzed the gut contents of 45 tadpole and 21 metamorph L. sylvaticus and 38 larval A. laterale (for sample size information, see appendix A). I measured gut length (nearest mm, uncoiled in tadpoles) and stomach length for each specimen prior to gut contents examination. Tadpole gut contents from 10-mm sections of the fore-, mid-, and hindgut regions were flushed, and their contents mixed. Salamander stomachs and intestines were cut open and emptied by flushing with distilled water. Macroscopic prey were examined under a dissecting microscope, counted and identified to order or family level for invertebrates. Broad categories were used for detritus, plant particles, algal matter, unidentifiable insect fragments, diatoms, and fungi. The flushed liquid containing small particles was sonicated for 60-90 seconds to loosen and break up clumped material and then filtered through a 45-µm gridded filter. Filters were covered with an inverted petri dish and left to dry for 24 hours before being mounted on slides and cleared with Type A immersion oil. Using a compound microscope at 400× magnification, the first 200 particles seen along a randomly chosen gridline were identified. Particle percentage for each prey item was quantified by dividing the summed particle counts by 200. Subsequently, at 100× magnification, 10 randomly selected fields-of-view were used to estimate the volume occupied

14 by each prey category assuming all items represented 100% volume. The estimated volumes were averaged per prey category to represent total volume percentage. In most cases for salamander larvae all particles were identified and counted prior to filtering contents, therefore I found five fields-of-view were sufficient for volume estimation. If an item occurred both prior to filtering and on the filter, the item was placed onto the filter to assess the volume it would occupy and the initial particle count was adjusted to reflect the addition. This was done to keep the total volume of a particular prey item in the gut unbiased between tadpole and salamander GCA methods.

2.3.4 Stable isotopes analysis (SIA)

The use of stable isotopes to define trophic niche is a powerful complement to gut contents analysis (Newsome et al. 2007) because isotopes data can be used to estimate resources assimilated into the consumer’s tissue over time (Peterson and Fry 1987), rather than merely the resources consumed. Further, among-individual variance in 13C and 15N can be used as an indicator of feeding niche width (Bearhop et al. 2004).

We used the specimens from GCA for SIA (with a few exceptions that did not provide a sufficient amount of dried tissue). In total, I used 44 tadpoles, 18 metamorphs, 34 salamander larvae, and 55 invertebrate samples representing 12 taxa (snails and clams without shells) for SIA. Amphibian samples were single specimens, whereas invertebrate samples were composite samples of 2 to 50 specimens. Prior to SIA, amphibian tissue, basal resources, and invertebrate tissues were rinsed, dried, and ground to a homogeneous powder in a Mini-Beadbeater (BioSpec products Inc., OK, USA). SIA was performed at the Cornell Isotope Laboratory (COIL). Stable isotope values are reported in delta notation as 13C or 15N using the equation 13C or 15N = 13 12 15 14 13 ((Rsample/Rstandard) – 1)  1000, where R is C: C or N: N. The standard for  C is Vienna Pee Dee Belemnite, and the nitrogen standard is atmospheric Nitrogen. COIL calibrates BCBG (plant) and Mink (animal) in-house standards against international reference samples, which is used to determine the accuracy and precision of the instrument during sample runs. Accuracy is measured as the mean difference ± 1 SD between the reference standards (BCBG: 13C = -27.08 ± 0.10, 15N = 0.62 ± 0.16, n = 105; Mink: 13C = -25.35 ± 0.09, 15N = 11.47 ± 0.07, n = 13). Amphibian 13C values were normalized to account for lipid variation in 13C (mean C: N = 4.19) following the recommended equations from Post et al. (2007).

In order to use stable isotopes to measure niche width, prey must be isotopically distinct, basal resource or prey resource isotope signatures must remain relatively stable over time, and an appropriate tissue to account for niche differences should be used (Bearhop et al., 2004). We tested for an effect of temporal changes on 13C and 15N values for each basal resource and found that the isotopic signatures for all resources remained stable over the sampling period (all one-way ANOVAs: P > 0.05) except for FBOM. The latter had a significantly more negative 13C signature in May (13C = -34.66, 15N = -1.23) compared to all other months (R2 = 0.455, 13 15 13 F5,33 = 4.51, P = 0.004, mean  C -32.20, mean  N = -1.18). Detritus was distinct in  C (range = mean ± SD: -27.92 ± 1.30, n = 33) from algae (-35.56 ± 2.94, n = 26), aquatic plants (- 30.09 ± 1.63, n = 12) and FBOM (-32.73 ± 1.73, n = 33). Basal sources were distinct and did not vary temporally; therefore, changes in consumer isotopic signatures are the result of diet shifts. Stable isotope values for basal sources and consumers are supplied in Appendix B.

2.4 Data analysis

2.4.1 Quantifying diet variation and niche width

The index of relative importance (IRI) was used to rank prey importance in each species diet (Piankas et al. 1970):

IRI = (P + V) × F where P is particle percentage (numerical abundance in percent), V is volume percentage, and F is percent frequency of occurrence of a diet item in the sampled species. Values of V, F and P were summed across ponds to get an overall species IRI and summed within ponds to get a pond level IRI per species.

I also used IsoSource, a complementary approach to IRI to calculate the range of all possible food source contributions (i.e., prey items, basal sources) to an amphibians’ diet (i.e., mixture) from δ13C and δ15N (Phillips and Gregg 2003). IsoSource is designed to handle situations where the number of potential sources is greater than n isotopes +1, which is common in natural systems. It has been shown that including more than seven sources in IsoSource increases processing time and produces less resolved results than using fewer sources (Mantel et al. 2004). Therefore, I first ran each model with mean 13C and 15N values for A. laterale mixtures and

16 the first seven food sources based on IRI rank in diet. We have isotope data for 12 of the 27 diet items found in A. laterale stomachs. Sources with minor contributions (<10% maximum feasible contribution) were removed and the analysis was performed again keeping diet items contributing ≥10% and adding diet item not included in the first iteration to bring the total number of sources analyzed to seven. Five iterations were necessary to include all sources. Isotopic fractionation was accounted for by subtracting 2.3‰ from the 15N value of each consumer prior to use in model. The IsoSource program, in small increments (in this case, 1%), iteratively creates combinations of possible source contributions and predicts mixture isotopic signatures. The model then compares the predicted combinations to the observed mixture signatures and if the predicted are equal or fall within a small tolerance range (for this case, 0.1‰) then this combination is stored as a feasible diet solution.

Notably, there are other ways besides diet that can contribute to variation in tissue isotope values. Variation can be caused by many factors, such as incorporation rates into tissues, the environment, the organisms physiology and life history, the tissue analyzed, and quality resources (Boecklen et al. 2011). These factors complicate cross-study comparisons; therefore the solutions found in this study may be unique to these ponds and species.

Using gut contents data, I calculated total niche width (TNW), the within-individual component (WIC), and the between-individual component (BIC) for each species per pond using Roughgarden’s (1979) equation based on the Shannon-Weaver diversity index within the IndSpec1.0 program (Bolnick et al. 2002). Indices to measure niche width have both advantages and disadvantages. Roughgarden’s index has the advantage of quantifying both components contributing to TNW, however it assumes resources are evenly distributed and used equally by consumers (Roughgarden 1974) and can overestimate individual specialization by using the natural log of a proportion (Bolnick et al. 2002). The proportional similarity index does not have this bias. Using IndSpec, I quantified diet specialization among individuals by first calculating the proportional similarity (PSi) between each individual’s diet proportions and the averaged population diet distribution using the equation adapted to individual-level analysis (Bolnick et al. 2002):

PSi  minpij ,qj  j

17 where pij is the frequency of prey item j in individual i's diet, and qj is the frequency of prey j in the population as a whole. PSi varies from near 0 (no overlap) to 1 (total overlap). The average of all proportional similarity values (PSi) in a population reflects the populations’ level of individual specialization (IS). The degree of individual diet specialization (V) is then calculated as V = 1 - IS (Bolnick et al. 2007), where higher values correspond to a higher degree of individual diet specialization. I then used the program IndSpec1.0 to generate 1000 replicate null diet matrices by nonparametric Monte Carlo techniques for each individual by randomly drawing from the population resource distribution (Bolnick et al. 2002, Bolnick et al. 2007); I compared the null distribution against our observed values of V to test whether observed diet specialization differed from random expectations. In two cases, only a single A. laterale specimen was captured from the pond, therefore they were left out of these analyses; consequently sample numbers differed between IRI values and niche measurements.

I calculated trophic position, a continuous measure of diet estimating an organism’s place within the vertical structure of a food web, for each individual using 1) quantitative gut contents data (GC) and 2) tissue nitrogen isotope data (SI):

TPGC  Vi Ti 1 i1 where TPGC = trophic position using GC data, Vi = volume percentage of each diet item and Ti = trophic level of each diet item (Vander Zanden et al. 2000) based on published estimates of trophic level (Merritt and Cummins 1996, Vander Zanden et al. 1997);

15 15 TPSI = λ + ( Nsc -  Nbaseline)/Δ

15 where TPSI is the trophic position of an individual estimated from  N, λ is the trophic level of the baseline (1 for basal resources), 15Nsc is the nitrogen isotope signature of the consumer 15 15 being evaluated,  Nbaseline is the pond specific mean  N signature of basal resources (Winemiller et al. 2007), and Δ is the enrichment in 15N per trophic level. Nitrogen fractionation (Δ) was set at 2.3‰ based on an analysis of aquatic animals (McCutchan et al. 2003).

The isotopic fractionation (also referred to as discrimination) value used to estimate trophic position is very important and can have considerable consequences to interpreting isotope data. I used 2.3‰ because based on previous study results it seemed like a good compromise and robust value for our situation. Carbon and nitrogen show high variation in discrimination dependent on consumer taxonomic group, tissue and diet isotopic ratio (Caut et al. 2009). One of the few studies to determine isotopic fractionation of amphibian larvae found larval wood frog heart tissue relative to rabbit food was 1.98 ± 0.17‰ for 15N (Schiesari et al. 2009). Vanderklift and Ponsard (2003) found the overall mean δ15N enrichment was 2.54‰ (± 0.11‰ SE) representing 134 literature derived estimates from diverse taxa (mammals, birds, invertebrates, spiders and fishes). In addition, they found that organisms consuming detritus had significantly lower estimates of ∆ and that most aquatic organisms are ammonolectic, which averaged 2.0‰ per trophic step. McCutchan et al. (2003) found ∆15N to be significantly lower for consumers raised on invertebrate diets compared to consumers raised on other high-protein diets and aquatic animals had on average a ∆ of 2.3 ± 0.28. In a recent review (Martinez del Rio et al. 2009) advised scientists to use literature derived ∆ values for related species fed on similar diets and measured on the same tissues as the focus group. We used whole amphibian larvae for SIA. Amphibian larvae fed on detritus and invertebrates and they are small, rapid growing ectotherms. They could very well have faster incorporation rates because of higher metabolic rates. McIntyre and Flecker (2006) observed that turnover rates were higher in smaller organisms. Thus, the use of a lower fractionation rate is appropriate. For instance, the isotopic discrimination factor of nitrogen ranged from -0.25 to 1.65‰ in the hatchling turtles’ tissues (Reich et al. 2008). Many researchers use ∆ of 3.4, but Post (2002b) stated using 3.4 is best applied to the entire food web and many species. My study focused on the trophic ecology of two species not the entire community food web. Therefore, I feel 2.3 is a reasonable value to use given wood frogs and salamander larvae feed on primary resources and insects, are fast growing, aquatic, and it is in close agreement to a value estimated in the only known study using our study species (Schiesari et al. 2009).

2.4.2 Statistical analyses

To test for the NVH, I used general linear models (GLM) to test for a relationship between individual niche variation and total niche width using R (R Development Core Team, 2009). I also used GLMs and linear mixed-effects (MEM or hierarchical) models in the R package lme4

19 to test for differences in diet, niche parameters, and 13C and 15N values among hydroperiod categories for each species. Hydroperiod was treated as a fixed effect and individual ponds were considered random effects. I compared the goodness of fit between the GLM and MEM analyses and report the model with the lowest AIC score (in most cases the GLM was the better model). Diet differences between tadpoles and metamorphs were evaluated by t-tests. I tested for the correlation between SVL or developmental stage and diet overlap (PSi) using spearman rank correlation. If ontogenetic shifts occur, I would expect to see a negative correlation between SVL or stage and diet overlap. I also tested for ontogenetic diet shifts in 13C and 15N signatures, gut morphology, and trophic position within each species using GLM. Since each individual had two measurements of trophic position, gut content and stable isotope, paired t-tests were used to test whether the trophic position estimation methods yielded similar results.

2.5 Results

2.5.1 Niche variation across hydroperiod gradient

Individual diet variation (V) increased significantly with population total niche width (observed:

GLM: P = 0.0002, V = 0.308TNW - 0.298, AIC = -29.015, df = 8) to a greater degree than predicted by the null model values (null model: GLM: P = 0.0018, nullV = 0.027TNW + 0.012, AIC = -66.17, df = 8; Fig. 2.1). I ran an ANCOVA to test for differences in slopes and intercepts among regression lines. The results show a significant effect of TNW and model (P < 0.0001) and a significant interaction (P < 0.0001), suggesting that the slope of the regression between TNW and V is different for both models. Therefore, amphibian larvae populations (tadpoles and salamander larvae) with wider niche widths are considered generalists but wider niches were generated because individuals became more specialized. However, the relationship between V and TNW was not significantly influenced by hydroperiod (GLM: long: intermediate P = 0.61, short: intermediate P = 0.22, df = 8, AIC = -28.13) or species identity (GLM: P = 0.42, df = 8, AIC = -28.077). The degree of individual variation was low (V < 0.33) for both species indicating populations are primarily composed of generalist individuals.

The TNW of L. sylvaticus did not differ significantly between tadpoles and metamorphs (two- sample t-test: t(6) = -0.9, P = 0.38). For both species, hydroperiod had little effect on TNW; population niche widths were similar across hydroperiod categories for L. sylvaticus (GLM: P =

0.3, df = 7) and A. laterale (GLM: intermediate to long P = 0.77, intermediate to short P = 0.24, df = 4).

2.5.2 Diet and trophic niche

Tadpoles of L. sylvaticus showed a high degree of diet similarity with most diet items found in all tadpoles from all ponds. In total, 10 main diet categories were found in tadpole and metamorph stomachs (Appendix C). Tadpole diet consisted mainly of non-filamentous algae (IRI = 1207) and detritus (IRI = 1162). Diatoms were ranked third (IRI = 656, mostly from the genera Synedra and Fragilaria) followed by insects (IRI = 335), and plants (IRI = 205). The diet of metamorphs was similar in that detritus ranked highest (IRI = 325). However, insects (IRI = 207) were ranked higher, non-filamentous algae (IRI = 166) and diatoms (IRI = 91) were still high ranking components.

Lithobates sylvaticus had a wide range of isotopic values across ponds (13C = 6.34‰ and 15N = 3.82‰), suggesting generalist feeding, in agreement with gut content findings. Tadpoles and metamorphs had significantly lower mean15N signatures in long hydroperiod ponds (Mixed- effects model: P < 0.0001, df = 60, n = 13) compared to intermediate hydroperiod ponds (n = 49). Although mean carbon use (13C = -32.76) did not differ between intermediate and long hydroperiod ponds (Mixed-effects model: P = 0.25, df = 60). According to the IsoSource mixing model results, L. sylvaticus biomass was supported mostly by detritus (725%) and aquatic plants (1732%; Fig. 2.2). Algae (062%) and seston (057%) showed wider, and hence less precise distributions of source contributions to wood frog biomass.

Overall, 27 different diet items were identified in the stomachs of A. laterale. Arthropods and diatoms contributed the most (IRI = 18, 17, respectively), followed by Chironomidae larvae, plants, algae, detritus, non-filamentous algae, oligochaete worms, amphibian tissue, and Sphaeriidae clams (IRI range 7 – 3, see Appendix D). The IsoSource results suggest freshwater leeches (Hirudinea) and lestid damselfly nymphs were major contributors (Fig. 2.3). Aquatic plants, amphipods, algae, and wood frogs also were present in their diets, however, their importance is less precise (both zero and high values), while dytiscid larvae were minor sources.

Tadpole and metamorph L. sylvaticus had lower 15N values than larval A. laterale, in agreement with our finding that they occupied a lower trophic level (Fig. 2.4). They also had significantly

21 different 13C signatures than A. laterale. There was no clear linear relationship between hydroperiod and isotopic niche, although some differences were observed along the hydroperiod gradient for A. laterale. Ambystoma laterale 15N values were highest in intermediate hydroperiod ponds (GLM: P = 0.04, df = 33, n = 13) compared to short (n = 16) and to long hydroperiod ponds (GLM: P = 0.006, df = 33, n = 5, AIC = 82.225), however, there was a pond effect (Mixed-effects model: P = 0.02, df = 31).

Both species showed differences in the importance of specific diet items across ponds. Non- filamentous algae were the highest ranked component in tadpole guts from Inter1 and Long1 ponds, but ranked second in Inter2 and Inter3 ponds. Diatoms were most important in tadpoles from Inter3 pond whereas detritus was most important in Inter2. Metamorphs from Inter3 and Inter1 ponds showed similar diets with insects, detritus, and non-filamentous algae ranked in the same order in both ponds. Detritus was ranked one for Long1 and Inter2, but rankings of other diet items differed considerably. For A. laterale in Inter2 and Inter3, the index of relative importance indicated chironomid larvae were dominant food items, however in both short hydroperiod ponds it was of low importance.

2.5.3 Trophic position

Lithobates sylvaticus occupied a lower trophic position as herbivorous primary consumers (mean

TPSI ± SD: 2.01 ± 0.22) than A. laterale as omnivorous secondary consumers (2.99 ± 0.27). Regardless of the method used to estimate TP, both species maintained relatively constant trophic positions across hydroperiod categories (GLM: TPSI: A. laterale P = 0.25, df = 31, L. sylvaticus P = 0.3, df = 61; TPGC: L. sylvaticus P = 0.08, df = 61; Appendix E). Ambystoma laterale had significantly lower TPGC in short (GLM: P = 0.02, df = 37, n = 19) compared to intermediate (n = 16) and long (n = 3) hydroperiod ponds, which were not different (GLM: P = 0.4, df = 37). TP estimates using stable isotopes were slightly higher than corresponding TP estimated using gut contents for larval A. laterale (paired t-test: t(31) = -1.836, P = 0.08, TPGC =

2.88, TPSI = 3.00). The TP estimated from tissue stable isotopes was significantly lower than that estimated from gut content data for L. sylvaticus tadpoles (paired t-test: t(40) = -5.307, P < 0.0001) and metamorphs (paired t-test: t(20) = -2.982, P = 0.007; Fig. 2.5). Although I found a difference in trophic position between methods, both methods placed L. sylvaticus at the same trophic level (trophic level 2, primary consumers).

2.5.4 Ontogenetic diet shifts

Ontogenetic diet shifts were evident in L. sylvaticus based on both isotopic signature and diet composition data. Generally, L. sylvaticus exhibited a switch around stages 31 through 33 from reliance on primarily non-filamentous algae, to primarily detritus, followed by an abrupt increase in the ingestion of insects near metamorphic climax. I found a significant negative correlation between developmental stage and diet overlap (PSi) in L. sylvaticus (spearman correlation: df =

64, rs = -0.52, P2-tail = 0.000008), indicating an ontogenetic diet shift in consumed prey. The proportional similarity between the diets of L. sylvaticus individuals was very high in early developmental stages indicating almost complete diet overlap (PSi mean = 0.922, stages 26-31), but decreased throughout tadpole development (0.871, stages 32-41) and metamorphic stages (0.688, stages 42-46). This resulted in a significant decrease in diet overlap with ontogeny (GLM: P <0.0001, df = 65). The trophic position of L. sylvaticus steadily increased throughout development until metamorphosis (GLM: TPGC P <0.0001, df = 65; TPSI P = 0.02, df = 61, Fig.

2.5). TPGC estimates exhibited lower variation during tadpole stages and higher variation during metamorph stages. The opposite pattern (higher variation in tadpole stages) was seen for TPSI.

Salamander development is not measured by stages after limb development. Therefore, I used snout-to-vent length (SVL) as a measure of growth, and time since eggs were laid as the length of the larval period. I did not find evidence for ontogenetic diet shifts in A. laterale using gut contents data as the correlation between snout-to-vent length and diet overlap was not significant

(Spearman correlation: df = 27, rs = -0.1251, P2-tail = 0.5). Changes in trophic position (TPGC and

TPSI) in A. laterale were not significantly related to changes in body size (GLM: A. laterale

TPGC P = 0.06, df = 33, TPSI P = 0.24, df = 27). Individuals of A. laterale ranged from primary to secondary consumers. A. laterale fed on a diverse set of prey but diatoms and arthropods were consistently ranked higher than other prey types throughout ontogeny.

Although, the isotopic values of the amphibian tissues changed across the sampling period, the isotopic signatures of basal resources did not. Tissue 15N increased with increasing size in A. laterale (GLM: P = 0.1, n = 27, AIC: 69.352, R2 = 0.12, Fig. 2.6B). Though variation in carbon isotope assimilation throughout development was evident, the change was not a linear response with development or size (A. laterale P = 0.7, df = 27, AIC: 93.359, Fig. 2.6A). Lithobates sylvaticus became more enriched in 13C (less negative) throughout larval ontogeny, indicating

23 feeding on different resources (GLM: P = 0.02, df = 61, AIC = 224.06, Fig. 2.6C), but 15N did not change substantially over that time (P = 0.7, df = 61, AIC = 162.03, Fig. 2.6D). I divided larval ontogeny into tadpole (stages 28  42) and metamorph (stage 43 – 45) components, and found that 15N increased during tadpole development (P = 0.06, df = 41, AIC = 108.75), but did not change during metamorphic stages (P = 0.16, df = 19, AIC = 52.62). Carbon isotopic signature did not change significantly during tadpole stages (P = 0.4, df = 41, AIC = 157.58) or metamorphic stages (P = 1.0, df = 19, AIC = 71.679). Tadpoles and metamorphs were distinguishable in terms of their 13C values, indicating their diets were significantly different and corroborates the diet change found in gut contents (two-sample t(43) = 2.297, P = 0.026). Mean 15N however, was not different between tadpoles and metamorphs (two-sample t(39) = - 0.774, P = 0.44).

2.6 Discussion

My results show that individual specialization of two amphibian species increased as population trophic niche width (TNW) increased, in agreement with the predictions of the niche variation hypothesis. Contrary to our expectations, however, I did not find evidence that hydroperiod influenced trophic niche width or individual niche variation. A positive relationship between individual specialization and population trophic niche width has been shown in several taxa (Bolnick et al. 2003a, Bolnick et al. 2007, Araújo et al. 2011), but this study is one of the few to take a community approach and quantify the relationship between niche width and specialization for multiple species from different trophic levels. Both Svanbäck et al. (2008) and Araújo et al. (2011) have advocated moving beyond single-species studies of individual specialization in order to observe and understand community-level patterns. My study provides a unique contribution to the understanding of these community-level patterns, because I incorporate ontogenetic changes in diet. I found diet variation occurred throughout ontogeny in A. laterale and L. sylvaticus. Ontogenetic diet shifts are an important aspect not commonly incorporated into food web studies, but could influence our understanding of food web structure and dynamics.

Intraspecific competition (Svanbäck and Persson 2004, Svanbäck and Bolnick 2005, 2007), release from interspecific competition (Bolnick et al. 2010), and predation may influence specialization (Eklöv and Svanbäck 2006). Population niche expansion via increased individual variation is often associated with species-poor, temperate communities and has strong ecological

24 and evolutionary implications (Bolnick et al. 2003a). Such implications of increased individual variation can influence population and community dynamics (Vindenes et al. 2008, Schreiber et al. 2011). Interspecific competition is thought to be low in short hydroperiod ponds (Wellborn et al. 1996), meanwhile intraspecific competition can be high due to high population densities (Toft 1985) especially during pond drying (Rogers and Chalcraft 2008). More permanent ponds (those with longer hydroperiod) have a higher diversity of predators compared to short hydroperiod ponds (Wellborn et al. 1996). Hence, the hydroperiod gradient spanning temporary to permanent ponds sets up a natural experiment to test predictions about niche width. Therefore, short hydroperiod ponds with low interspecific and high intraspecific competition, individual specialization is predicted to be higher than that in long hydroperiod ponds, which have high interspecific and low intraspecific competition.

In the study ponds, the number of coexisting amphibian species increased from two to six and I observed over a fourfold increase in the number of predator insect taxa (mean 2.5 compared with 11.5 predominately predator families, TAS unpublished data) creating a natural progression of increasing interspecific competition and predation risk. I did not use a manipulation approach in this study, which limits the ability to directly measure competition. However, my results are in agreement with my predictions, individual specialization (V) and population total niche width tended to be highest in short hydroperiod ponds, although this trend was not significant. One possible explanation for the lack of significance is limited statistical power due to small sample sizes. Another explanation is that competition and predation interact in complex ways to influence amphibian larval performance thus complicating the strength of the relationship. It is also possible that where only one species is present, diet variation is higher than expected, as seen in lakes containing only northern pike (Beaudoin et al. 1999). Ambystoma laterale was the only amphibian species present in one of the short hydroperiod ponds and was one of only two species (the other Pseudacris crucifer Wied-Neuwid) in the other short hydroperiod pond used in this study. In fact, all the measured variables varied little with hydroperiod, suggesting that hydroperiod may play only a small role in determining diet.

Functional role may be coincident with the extent of individual specialization in these larvae. Ambystoma laterale held the highest trophic position as a secondary consumer and the highest degree of individual specialization (V = 0.28), whereas L. sylvaticus tadpoles were a trophic level below the salamander larvae and exhibited almost no specialization (mean V = 0.09). This

25 study lends support to the view that individual specialization is more prevalent in higher trophic level consumers (Araújo et al. 2011) and generalist predators (Woo et al. 2008) because predators may experience stronger intraspecific competition (Svanbäck and Bolnick 2007) reducing intensity of interspecific competition.

Interestingly, both species of amphibian larva had low levels of individual specialization (average V = 0.24), something which is rarely documented in the literature (Araújo et al. 2011). In an analysis of individual specialization for 142 samples, on average, individuals showed only 47% similarity with their populations (Araújo et al. 2011). In contrast, on average, A. laterale and L. sylvaticus showed at least 73% overlap with their populations’ niche. Because I examined larval forms rather than adults, it is possible that the values from Araújo et al. are not directly comparable to mine. I am aware of only one other study that looked at individual variation in larvae. Zerba and Collins (1992) compared variation using a different method than mine, yet they also found high diet similarity among individual salamander larvae. It is also possible that the occurrence of individual specialization is elevated because of publication bias against low or non-significant levels of individual specialization.

Mechanisms underlying individual specialization are based on trade-offs in behavior, morphology, or physiology that limit an individual from using the full set of available resources (Bolnick et al. 2003). One possible explanation for the observed low levels of specialization is that larvae differ in their physiological abilities to digest resources. Tadpoles differ in gut clearance rates and assimilation of food (Altig and McDearman 1975) and prey resources differ in digestibility (Altig et al. 2007). Tadpoles feed mostly on detritus, which is a poor quality food in comparison to algae (Schiesari 2006) and invertebrates. Behavioral and or morphological constraints could also restrict diet specialization in salamander larvae. For example, as salamander larvae grow their mouth gapes widen, allowing ingestion of larger prey items (Bardwell et al. 2007), therefore, differences in diet can exist along size and/or age gradients. Another possibility is that resources are sufficiently abundant and so there is no need to specialize. These results suggest more research is needed to understand the trade-offs influencing amphibian larval diet choice and how ontogenetic changes may interact with individual specialization.

Gut contents data although an informative tool to describe diet, it has a few problems. A gut contents sample is assumed to represent the general feeding pattern of the individual. However, what it actually describes is a snapshot of the items very recently ingested. For this reason, using gut contents data can give misleading information about the population trophic niche. High individual specialization can result from individuals feeding in a patchy habitat. For instance, if an individual opportunistically fed on a dense aggregation of one prey type just prior to being collected, the data would suggest the individual specializes on that prey type. However, one would have no way of knowing if this is an anomaly or a consistent feeding pattern without sampling this individual over time. It cannot be under emphasized that that gut contents data of an individual is only a snapshot of what an individual recently ate and do not convey long term diet information. I tried to reduce the snapshot problem and get an idea of the consistency of the individual and population niche by sampling multiple individuals within the same developmental stage.

In addition, diet category selection can influence individual specialization estimates. Too coarse grained categories (lumping of diet items) may underestimate specialization. Based on the number of diet categories used in diet studies using amphibian larvae (9 items each for A. laterale-jeffersonianum and A. maculatum, Nyman 1991; 11 items for A. jeffersonianum, Bardwell et al. 2007; 20 items for A. tigrinum nebulosum, Zerba and Collins 1992; 20 diet categories Nannophrys ceylonensis (semi-terrestrial tadpole, Wickramasinghe et al. 2007), the number of categories I used (27 A. laterale and 10 categories L. sylvaticus) seems appropriate. Jenssen (1967) described the diet of Rana clamitans (n = 144) using 34 algal taxa, one diatom, three protozoa, and nine animal groupings, and still concluded tadpoles were indiscriminate, generalist feeders throughout ontogeny. I found low levels of individual diet specialization for both species.

Niche width is a relative measure dependent on resource availability. TNW calculations therefore are not exact estimates and interpretation of the absolute value of TNW should be done with caution. In a relative sense a population with a higher value than another can be said to have a wider niche. I did not measure resource distributions in each pond, however other studies have found proportions of food items in guts are proportionally similar to resource availability in the habitat (Jenssen 1967, Tavares-Cromar and Williams 1997).

2.6.1 Ontogenetic diet shifts

Ontogenetic diet shifts dictate the temporal complexity, structure, and function in food webs and are essential to understanding community dynamics (Polis 1991), yet are rarely incorporated in food web studies. Changes in resource consumption of intermediate level consumers (i.e., anuran tadpoles in our study) can induce changes in higher trophic levels (i.e., salamander larvae and predaceous insects) resulting in a shortening or lengthening effect on food-chains (Post and Takimoto 2007).

Discrete shifts in diet between larval or juvenile and adult life stages are ubiquitous across animal groups (Werner and Gilliam 1984), however, documentation of continuous change through development is lacking for amphibians (Wickramasinghe et al. 2007). My study found ontogenetic changes in diet and morphology, for example, younger tadpoles had longer guts suitable for absorbing nutrients and accommodating large quantities of poor quality food (non- filamentous algae and detritus) and older (later stage) tadpoles developed shorter guts and switched to feeding on insects during metamorphosis. The agreement between diet and gut morphology has been described for many anurans (Toloza and Diamond 1990). Changes in diet were reflected in increases in trophic position. TPSI increased gradually over ontogeny, meanwhile TPGC revealed an abrupt shift from aquatic primary resources to terrestrial insects

(e.g., ants) thereby showing a rapid increase in TP. Lower variation in TPGC during the tadpole stages can be the result of many individuals feeding on basal resources that have the same estimated TP in similar amounts (by volume) in early stages. Whereas, metamorphs showed high among individual variation in gut contents (some guts filled with plant and detritus material and others full of insects). The variation may also be partially driven by sample size. In accordance with these changes in diet with ontogeny, I found that tadpoles of L. sylvaticus shifted from high dietary overlap and low individual diet variation, to lower diet overlap during metamorphic stages. Diet overlap would suggest a potential for high intraspecific competition among tadpoles, however, even with such high overlap the resource may not be limiting. These results suggest that, while in the larval aquatic habitat, conspecific frogs feed heavily on an abundant shared resource (i.e., detritus) but that diet diverges when they begin feeding terrestrially.

In contrast, salamander larvae did not exhibit clear patterns of diet use through ontogeny. My investigation using stable isotopes and gut contents data indicated that trophic position does not

28 increase linearly with increasing body size (SVL) for A. laterale. Salamander larvae were always carnivorous and at the top of the food chain in the study ponds. A small increase in 15N with increasing SVL in A. laterale was observed (Fig. 2.6), but the increase did not produce an increase in trophic position. However, ontogenetic diet shifts do not necessarily coincide with an increase in trophic position, as found by Vander Zanden et al. (2000) who attributed a lack of body size – trophic position relationship in lake trout (Salvelinus namaycush) to a broad range of prey sizes ingested by this species. Diets were diverse in A. laterale throughout their larval periods. Thus, diet diversity or omnivory could lead to a de-coupling of body size and trophic position during the larval period in salamander larvae.

Amphibians are an important link between aquatic and terrestrial ecosystems via metamorphosis and as prey for terrestrial predators, thus changes in one habitat may manifest in another. Given amphibian larvae are abundant in the habitats they occupy, convert energy efficiently providing a high protein food source to predators, and have broad diets, these larvae play an important role in the food web.

Although trophic position did not vary with ontogeny, it did vary among individuals; however, the mean trophic position remained stable across the hydroperiod gradient. This suggests that the trophic or functional role of amphibian larvae is maintained regardless of habitat duration. Notably, trophic position estimated from stable isotope analysis showed more variation than trophic position estimated from gut contents. Gut contents data use a snapshot of information accounting for only recent feeding activity, whereas stable isotopes work on a longer time frame producing a time-integrated, trophic position value (Vander Zanden et al. 1997). However, stable isotopes are not without error and many factors other than diet influence variation in isotopic values (Boecklen et al. 2011), which could explain the incongruence between findings.

Although, TPGC was different from TPSI for L. sylvaticus tadpoles, interpretations of diet and trophic level were consistent. Using both gut contents and stable isotope analyses allowed me to differentiate diet items, assign trophic positions within the food web, and determine which resources were assimilated by consumers over a longer time scale.

2.6.2 Conclusions

Freshwater amphibian assemblage structure is determined by the interaction of interspecific competition, predation, and pond drying (i.e., hydroperiod) (Toft 1985). Amphibian species align

29 themselves along the hydroperiod gradient based on their abilities to withstand in situ conditions. Since hydroperiod is a critical niche axis, I wondered if it has a strong influence on resource use. This study presents a novel test of and support for the niche variation hypothesis. I found the degree of individual specialization increased with increasing population trophic niche width (TNW) when compared across two amphibian species during their larval stages. I document conserved generalist feeding patterns across an environmental gradient. Individual specialization was low for both species and across the hydroperiod gradient. However, diet variation occurred throughout ontogeny in A. laterale and L. sylvaticus, revealing an important aspect not commonly incorporated into food web studies, but which could influence our understanding of food web structure and dynamics.

2.7 Acknowledgements

I thank J. Arblaster, S. Booth, M. Mahmood, G. Jegatheeswaran, M. Sambi, D. Bloom, and S. Bloom for help collecting and processing samples. M. Cadotte provided statistical advice and reviewed an earlier draft. D. Bloom, E. Macleod and C. Tucker provided helpful comments on earlier drafts. Funding was provided by Chicago Herpetological Society (TAS) and an NSERC discovery grant (DDW). Animals were collected under a scientific collector’s permit (OMNR # 1051193) and used in accordance with animal care protocols approved by the University of Toronto and Queen's University Animal Care Committees (# 20007692).

2.8 Figures

2.4 Carbon and nitrogen bi-plot showing A. laterale shown in filled squares (n = 34) and L. sylvaticus grey filled circles (n = 62).

2.5 Trophic position calculated for L. sylvaticus throughout development using gut contents

(TPGC: open circles and solid line) and stable isotope data (TPSI: filled circles and dotted regression line).

3 Duration matters, size does not in pond food webs: A study of the dynamic constraints and ecosystem size hypotheses

3.1 Abstract

Identifying the ecological mechanisms that determine food web structure is critical for understanding the causes and consequences of diversity. Food-chain length (FCL) is a product of both the biotic interactions within a community and the environment, but how environmental variation affects FCL is not well understood. To understand the variation on FCL, I examined how gradients of ecosystem size and environmental variation in ponds spanning hydroperiod and size gradients affected FCL. Using carbon and nitrogen stable isotopes, I found that average FCL was 3.3 and varied by three trophic levels across ponds and years. I show that ponds with shorter hydroperiod have shorter food chains and FCL was not strongly influenced by ecosystem size.

These results demonstrate support for the dynamic constraints hypothesis, which predicts less predictable environments should have shorter food chains. My data are not consistent with the ecosystem size hypothesis prediction of larger ecosystems having longer food chains. Insect and amphibian richness increased with increasing pond size and hydroperiod, indicating that insertion of new species into the pond communities is a driving mechanism causing variation in

FCL. While omnivory could explain variation in FCL, my results show that the incidence of omnivory was similar across the environmental gradients. This represents one of the few empirical studies to link structural changes in the food web with variability in FCL. I show temporal variation and species composition shape pond communities and influence food web structure.

3.2 Introduction

Food webs are useful depictions of consumer-resource interactions allowing different aspects of food web structure to be quantified. One such important measure is food-chain length (FCL),

38 which measures the height of the food web (Post 2002a) or the number of trophic transfers from basal resources to the top predators of a system. FCL is a particularly useful metric describing food web structure because it can be compared across habitats (Post et al. 2000, Vander Zanden and Fetzer 2007) and influences ecosystem functions (Hairston et al. 1960, Duffy et al. 2005) such as: trophic cascades (Carpenter and Kitchell 1993); productivity (Pauly et al. 1998); and regulates bioaccumulation of toxins (Cabana and Rasmussen 1994). Much current research in food web ecology tries to understand how environmental variation correlates with food web structure and function, and more importantly, what mechanistic changes in food web structure influence food-chain length within an ecosystem.

Several hypotheses have been proposed to explain variation in FCL (Pimm 1982, Persson et al. 1996). One is the dynamical constraints hypothesis, where theoretical models predict that habitats subject to disturbance have shorter food chains because longer food chains tend to be less resilient to, and take longer to recover from perturbations than shorter food chains (Pimm and Lawton 1977, Pimm 1982). However, there is only limited empirical evidence that supports the idea that dynamical stability limits FCL (Post 2002a), even though the effects of disturbance on ecosystem properties should be strong in temporally variable habitats (Walters and Post 2008). The ecosystem size hypothesis predicts FCL will be longer in larger ecosystems because of higher species diversity, habitat availability and heterogeneity (Post et al. 2000). Previous tests have both confirmed support for and refuted each of these hypotheses. For example, FCL has been correlated with fish species richness in lakes and ecosystem size of lakes (Vander Zanden et al. 1999, Post et al. 2000), streams (Walters and Post 2008, McHugh et al. 2010), permanent ponds (Doi et al. 2009), river and terrestrial habitats (Schoener 1989, Vander Zanden et al. 1999, Post et al. 2000, Post 2002a, Takimoto et al. 2008, Walters and Post 2008), but not related to disturbance events in streams and terrestrial systems thus refuting the dynamic constraints hypothesis (Walters and Post 2008). A general explanation for the variation in FCL seen in natural ecosystems still eludes ecologists.

Environmental variation can, however, have substantial influences on food web structure (Sabo and Post 2008, Sabo et al. 2009, Sabo et al. 2010) and thus it is important to understand its influence on FCL. Predictable shifts in invertebrate (Brooks 2000, Stoks and McPeek 2003, Williams 2006, Turner and Montgomery 2009) and amphibian (Skelly 1996, Snodgrass et al. 2000, Eason and Fauth 2001, Babbitt et al. 2003) species richness and diversity occur along

39 gradients in hydroperiod (Wellborn et al. 1996). In lakes, studies have shown that changes in FCL are influenced by community composition (Vander Zanden and Rasmussen 1996, Vander Zanden et al. 1999, Lepak et al. 2006); however, it is less clear how community structure influences food web dynamics across a habitat duration gradient, or in temporary waters. Studies comparing food web structure across a habitat duration gradient are few in number (Williams 2006), but have been advocated as a promising area of research effort leading to an understanding of the relationship between dynamics and food web structure (Winemiller and Polis 1996). Temporary water bodies are abundant, heterogeneous in physical and biological characteristics (Blaustein and Schwartz 2001), exhibit temporal and spatial variability, are important in cross-habitat energy flow (Regester et al. 2006), and provide essential habitats for many species of invertebrates, amphibians, and plants, some of which are rare and threatened (Schneider and Frost 1996). However, we know little about wetland food web structure and dynamics (Sabo et al. 2009, Opsahl et al. 2010). The limited knowledge that exists has been derived from experimental manipulation of artificially constructed habitats (Magnusson and Williams 2006). Small ponds and wetlands thus represent an ideal environment in which to study food web ecology.

Underlying mechanisms influencing food web structure are responsible for variation in FCL. Post and Takimoto (2007) determined that variation in FCL is caused by either the addition (or removal) of a top predator(s) (additive mechanism), or changes in the trophic position of top predators. Three mechanisms lead to changes in the trophic position of top predators: 1) addition or removal of intermediate predators (insertion mechanism); 2) changes in trophic omnivory by top predators (omnivory mechanism); and/or 3) changes in the trophic position of intermediate predators (Post and Takimoto 2007). It is important to describe the variability of FCL, but it is also insightful to clarify the mechanisms creating the pattern.

In this study I determine which environmental factors influence FCL in ephemeral systems. To do so, I used carbon (13C) and nitrogen (15N) stable isotopes to characterize the food web structure along a natural aquatic habitat duration gradient and test the strength of the dynamic constraints and ecosystem size hypotheses. I predicted that FCL would increase with increasing permanence (i.e., lengthening in hydroperiod equals less environmental variability and stability) because of increased predator diversity (additive mechanism) and changes in amphibian and insect community composition (insertion mechanism). I predict larger habitats will have longer

40 food-chains relative to smaller ponds, because of their higher species richness and habitat heterogeneity. I examined amphibian and insect community information for each study pond to determine if the three mechanisms were responsible for variability in FCL.

3.3 Methods

3.3.1 Study sites

The study ponds were located at the Queen’s University Biological Station (QUBS), north of Kingston, Ontario, Canada a 3000-ha area of mixed coniferous-deciduous forest and wetlands. I characterized aquatic food webs and examined relationships between FCL and environmental variables for nine natural, un-manipulated ponds varying in hydroperiod and size. These ponds are filled by snow melt in early spring and lose water as the summer progresses, resulting in some ponds drying completely. They ranged from having a hydroperiod categorized as ‘short’ (65  102 days, n = 3), ‘intermediate’ (125  175 days, n = 4), or ‘long’, (i.e., those that did not dry during the study period; ~365 days, n = 2). Seven ponds were used in 2008 (short n = 2, intermediate n = 3, long n = 2) with two other ponds (one short and one intermediate) added to the study in 2009. This spanned the natural gradient from intermittent freshwater woodland ponds to permanent freshwater marshes, common throughout eastern North America. All of the study ponds are naturally fishless with predaceous insects and/or salamander larvae as top predators. In this respect, our study ponds are in stark contrast to most observational food web studies conducted in lakes.

3.3.2 Field collection of pond communities

Ponds have discrete habitat boundaries therefore, the food webs in this study were delimited by the resident aquatic consumers (i.e., larval amphibians and invertebrates), and the production sources and detritus consumed by those consumers. Reptiles, birds, parasites, and other non- aquatic organisms that potentially may have visited the ponds were excluded from my analysis.

I sampled each pond monthly (April to September of 2008 and 2009) to collect food web components for isotopic analysis. Basal organic resources consisted of detritus, aquatic macrophytes (submerged and emergent), filamentous algae, seston, and fine benthic organic matter (FBOM). To comprehensively sample multiple trophic pathways leading to the top consumers, I collected aquatic consumers (i.e., invertebrates and amphibians) and composite

41 samples of detritus, algae, and aquatic plants by hand and using a dip net from several areas within each pond. A 1000 ml water sample was collected from the water column at mid-depth from each pond. FBOM was obtained by dredging a 53 μm mesh net across the bottom of the pond in several randomly chosen locations, the net contents were then emptied onto a sieve tower of decreasing mesh size, and the benthic matter collected on the bottom sieve was rinsed with distilled water into a collection jar. Invertebrates were picked off the coarser sieves as well as collected by hand while dip-netting. Additional collections of amphibian developmental stages were made every two weeks to insure representation of each species’ entire life cycle (i.e., salamander larvae, anuran tadpoles, and metamorphs). Because of the relatively low species richness (four to nine species per pond) I was able to sample the entire amphibian community. Amphibians were euthanized by emersion in Tricaine Methanesulfonate (MS-222), buffered with an equal amount of sodium bicarbonate to a pH of ~7.0, held on ice during field collection, and then frozen until further processing.

3.3.3 Characterization of ecosystem size and environmental variation

Every month from April through August 2009 I mapped each pond by walking the perimeter with a Trimble TSC1 GPS connected to a ProXRS satellite and RTCM receiver. All mapping was completed in one or two consecutive days within each month. The area (m2) of each pond was then calculated in ArcMap. Ecosystem size was determined from the average monthly pond area from the available wetted months. Pond size, shape, and water depth fluctuated throughout the study period. Therefore, I used the coefficient of variation (CV) of water depth to address environmental variation in ecosystem size, because it is a proxy for available habitat. Since area and water depth were measured only in 2009, I tested the ecosystem size hypothesis with 2009 data only. Hydroperiod, the primary structuring agent in these ponds, was measured as the length of the aquatic phase or days from ice-off until each pond dried, as recorded in 2008 and 2009. The CV of water depth and hydroperiod showed a marginally significant negative relationship (GLM: P = 0.07, df = 8) reflecting the codependence of size and hydroperiod. I performed a Mantel test using the ade4 package in R on the spatial location of and the food-chain length measured in each pond. I accepted the null hypothesis that spatial location and FCL are unrelated based on 9999 Monte-Carlo replicates for 2008 (P = 0.16, observed correlation r = 0.3888131) and 2009 (r = -0.001588911, P = 0.31). I checked for temporal autocorrelation by computing a mixed effects model testing the influence of year on FCL. I found that only 0.0000000002% of

42 the total variance of the random effects is attributed to the year effect, therefore, a GLM without the random effect is more appropriate for these data.

3.3.4 Insect community composition

I performed a qualitative survey of aquatic invertebrate community composition for each pond. I collected invertebrates from several random locations throughout each pond with a dip-net, and using a fine-mesh net each month (April – September 2008 and 2009). I sorted major insect groups in the field, placing samples in plastic jars, and holding them on ice until deposited in a freezer. Using Merritt and Cummins (1996) and Marshall (2006) as guides, we identified and enumerated invertebrates to family or genus level under a dissecting microscope. The number of insect families that function as predators per pond were tallied and used to assess proximate mechanisms (Post and Takimoto 2007) underlying variation in FCL.

3.3.5 Sample preparation and stable isotope analysis

FBOM samples were poured into glass scintillation vials, allowed to settle, and the clear liquid was suctioned off using a pipette and discarded. The resultant concentrated FBOM was dried in the vials. Water samples were filtered onto pre-combusted Whatman glass microfiber GF/F filters. Water samples contained zooplankton, bacteria, and phytoplankton so we termed this resource as ‘seston’. Aquatic macrophytes, algae, and detritus samples were rinsed of attached detritus and invertebrates (removal checked under a dissecting microscope). Whole tadpoles and salamander larvae (minus viscera), and whole inverts (snails and clams without shells) were used for stable isotope analysis (SIA). Amphibian samples were single specimens, whereas invertebrate samples were composite samples of 2 to 50 specimens, depending on relative biomass. All samples were dried in a drying oven at ~ 60C for 2-3 days, and then ground to a homogeneous powder in a Mini-Beadbeater (BioSpec products Inc., OK, USA). All SIA was performed at the Cornell Isotope Laboratory (COIL). Stable isotope values are reported in delta 13 15 13 15 notation as  C or  N using the equation  C or  N = ((Rsample/Rstandard) – 1)  1000, where R is 13C: 12C or 15N: 14N. The standard is Vienna Pee Dee Belemnite for 13C and atmospheric Nitrogen for 15N. I adjusted 13C values to account for lipid variation in 13C in amphibian, invertebrate, and plant samples using the recommended equations from Post et al. (2007b). The mean C:N of aquatic consumers was greater than 3.5 (mean amphibian: 2008 C:N = 4.2, 2009 =

3.8; mean invertebrate: 2008 C:N = 5.0, 2009 = 4.9). A total of 182 basal resources, 313 amphibian samples, and 118 invertebrate samples was analyzed for SIA.

3.3.6 Food-chain length and trophic position calculation

I calculated the realized FCL from the carbon and nitrogen stable isotopes (Post 2002a) of each pond community and the realized trophic position of each consumer. Nitrogen stable isotope ratios (15N) predictably increase through trophic transfer up the food chain, providing an estimate of trophic position (Cabana and Rasmussen 1996, Vander Zanden and Rasmussen 1999) in a food web. Trophic position is a continuous measure of diet estimating an organism’s place within a food web. Stable carbon isotope ratios (13C ) measure the derived energy source (Peterson and Fry 1987, Fry 1991) because different energy sources often show distinct 13C signatures with little fractionation up the food chain (Hecky and Hesslein 1995). FCL was the maximum trophic position held by any larval amphibian or insect for each web. The trophic position (TP) of each individual amphibian larvae was calculated and then averaged to obtain species-specific mean trophic positions. I used pond-specific baselines, the mean of all basal resources (Hoeinghaus et al. 2008), to calculate trophic position for each consumer using the 15 15 equation, TP = λ + ( Nsc - Nbaseline)/∆, where λ is the trophic level of the baseline (1 for basal 15 resources),  Nsc is the nitrogen isotope signature of the consumer being evaluated, and ∆ is nitrogen fractionation. I used 2.3‰ for ∆ based on an analysis by McCutchan et al. (2003) of nitrogen fractionation for aquatic animals.

44 trophic step. McCutchan et al. (2003) found ∆15N to be significantly lower for consumers raised on invertebrate diets compared to consumers raised on other high-protein diets and aquatic animals had on average a ∆ of 2.3 ± 0.28. A recent review (Martinez del Rio et al. 2009) advised scientists to use literature derived ∆ values for related species fed on similar diets and measured on the same tissues as the focus group. I used whole amphibian larvae for SIA. Amphibian larvae fed on detritus and invertebrates and they are small, rapid growing ectotherms. They could very well have faster incorporation rates because of higher metabolic rates. McIntyre and Flecker (2006) observed that turnover rates were higher in smaller organisms. Thus, the use of a lower fractionation rate is appropriate. For instance, the isotopic discrimination factor of nitrogen ranged from -0.25 to 1.65‰ in hatchling turtles’ tissues (Reich et al. 2008). Therefore, I feel 2.3 is a reasonable value to use given the taxonomic scope of this study and study organisms feed on primary resources and insects, are fast growing, aquatic, and it is in close agreement to a value estimated in the only known study using one of our study species (Schiesari et al. 2009).

3.3.7 Data analysis

All statistical analyses were performed in R (R Development Core Team 2009). I tested the ecosystem size and dynamic constraints hypotheses using general linear models (GLM), separately regressing FCL against independent variables of size (pond area) and hydroperiod

(days). Ecosystem size and hydroperiod were not significantly correlated (GLM: log10area = 2 0.8972 log10hydroperiod + 0.705, P = 0.20, df = 8, AIC = 15.275, R = 0.22). I investigated the proximate controls of FCL by testing the additive, insertion, and omnivory mechanisms. The most straightforward shift in FCL is caused by the addition of a top predator thereby increasing FCL. To evaluate the additive mechanism I used GLM to test for a relationship between trophic positions of the top predator from each pond and FCL, followed by investigating whether the identity of the top predator was related to pond hydroperiod and size. To address the insertion of new intermediate predators, GLMs were used to look at the relationships between size and hydroperiod and the richness of insect predator families, total invertebrates, breeding amphibians, and all resident amphibians (amphibian richness) in each pond. To support the omnivory mechanism, the degree of trophic omnivory (i.e., feeding on more than one trophic level) by the top predators must change with a corresponding change in FCL. Therefore, I conducted a poisson regression using the coefficients of variation (CV) of 15N from four top predators: Ambystoma laterale larvae (blue-spotted salamander), A. maculatum larvae (spotted

45 salamander), Libellulidae nymphs (dragonflies), and Hirudinea (leeches) to test whether omnivory influenced FCL. Higher variation approximates a higher degree of trophic omnivory. I examined the CV of 15N against hydroperiod and pond size to test for changes in trophic omnivory across the environmental gradients. I also looked at changes in mean trophic positions of common top predators (A. laterale, A. maculatum larvae, nymphs of the families Libellulidae and Lestidae [damselflies]) across the pond size and hydroperiod gradients.

3.4 Results

3.4.1 Food-chain length and environmental variation

Food-chain length averaged 3.3 and varied by up to three trophic levels across ponds and years (range 1.7 to 4.6). Global estimates of average FCL are not available for wetlands, however, our mean FCL value is lower than values reported for streams (3.5) and lakes (~4.0) (Vander Zanden and Fetzer 2007). Food chains were shorter in ponds with shorter hydroperiod, supporting the dynamic constraints hypothesis (GLM: df = 15, R2 = 0.23, P = 0.03, AIC = 28.5; Fig. 3.1). A one unit increase in hydroperiod is equivalent to adding 1 trophic level to the food web (maxTP = 1.0293logduration + 1.025). FCL showed a marginally significant increasing trend with pond size (GLM: df = 8, R2 = 0.25, P = 0.09, AIC = 17.4; Fig. 3.1), but FCL was not related to pond volume (GLM: df = 8, P = 0.18, AIC =17.55). FCL was not influenced by the CV of pond area (df = 8, P = 0.37, AIC = 18.915) however, the variance of pond water depth negatively influenced FCL (df = 8, P = 0.05, AIC = 14.52, Fig. 3.1). This concludes variation in habitat availability influences the FCL of temporary ponds. In a separate post-hoc analysis we looked at the possible interactive effect of hydroperiod and ecosystem size on FCL, but it was not significant (GLM: df = 8, P = 0.99).

3.4.2 Proximate mechanisms underlying changes in FCL

FCL variation can be explained by three potential mechanisms: 1) the addition (or removal) of top predators; or 2) changes in top predator identity; or 3) changes in trophic position of the top predator. Hydroperiod and ecosystem size influenced the identity of top predators, however, identity switches did not substantially change FCL. Ambystoma laterale was the apical predator in 60% of the ponds and was present across the two year sampling period. That said, A. laterale had a significantly higher probability of being the top predator in smaller, shorter-to-intermediate

46 hydroperiod ponds (GLM using logit link with binomial errors, area: df = 8, P = 0.3; hydroperiod: df = 15, P = 0.03; see Appendix F).

The trophic position of all species of top predator significantly increased with increasing FCL (GLM: df = 63, R2 = 0.07, P = 0.02). Similar results were found when using only the four most prominent predators (A. maculatum, A. laterale, libellulid nymphs, and leeches; GLM: df = 38, R2 = 0.25, P = 0.001; Fig. 3.2) indicating either or both the insertion and omnivory mechanisms were at play.

Testing the insertion mechanism, I examined whether insect and amphibian richness changed along the environmental gradients and if this influenced FCL. Insect predator richness significantly increased with increasing pond size (df = 8, R2 = 0.35, P = 0.05; Fig. 3.3) and lengthening hydroperiod (df = 15, R2 = 0.39, P = 0.006; Fig. 3.4). Invertebrate familial richness was positively related to hydroperiod (df = 15, R2 = 0.39, P = 0.006) and pond size (df = 8, R2 = 0.45, P = 0.03; Appendix G, Appendix H). The number of amphibians breeding in each pond was significantly related to hydroperiod (df = 15, R2 = 0.34, P = 0.009, AIC = 71.98), and this was best represented by a quadratic relationship (df = 13, R2 = 0.57, P = 0.002, AIC = 3.93; Fig. 3.5), indicating that more amphibians bred in intermediate hydroperiod ponds (average 6 species) compared to both short (average 1 species) and long hydroperiod ponds (average 5 species). The number of breeding amphibians in a pond increased with pond size (df = 8, R2 = 0.55, P = 0.01; Fig. 3.6).

Variation in 15N was marginally significantly higher in the smallest ponds (GLM poisson: df = 8, Z = -0.795, P = 0.06) and fairly consistent across the hydroperiod gradient (GLM poisson: df = 12, Z = -0.414, P = 0.36), thus no relationship between omnivory and hydroperiod or ecosystem size was found. There was a trend of increased omnivory (higher CV) with decreasing FCL (GLM poisson: df = 15, z = -0.553, P = 0.2), but this relationship was strongly influenced by one pond (Short1 pond had 1 top predator sample thus a CV of 1). Removal of that pond resulted in a trend reversal (i.e., variation increased with increasing FCL; GLM poisson: df = 14, z = 0.623, P = 0.3), therefore, no clear or statistically significant pattern between omnivory and FCL was found. Overall, mean CV in 15N was highest in short hydroperiod ponds (35%), followed by long (23%) and intermediate hydroperiod ponds (14%). Across-pond trophic

47 position variation in all consumers sampled was low both in 2008 (CV = 23%, n = 207) and in 2009 (CV = 31%, n = 224).

FCL variability was also observed between the two sampling years (Fig. 3.7). The removal mechanism is illustrated in Short1 pond where A. laterale larvae held the maximum trophic position (max TP = 3.45) in 2008, but in 2009 no larvae were present in the pond and libellulid dragonfly nymphs (max TP = 1.72) replaced A. laterale as the apical predator. The max TP varied only slightly within the intermediate hydroperiod ponds across years presumably because the top predator was consistently either A. laterale or A. maculatum (range TP: 2008 = 3.2 - 3.5, 2009 = 2.9 - 3.4). The FCL of Long1 pond decreased between years because the trophic position of the top predator (a leech) decreased.

3.5 Discussion

Important environmental features of ecosystem size and dynamical stability are potential determinants of FCL (Pimm 1982). The two associated hypotheses have been tested primarily using lakes, streams, and container systems (Post 2002a, Sabo et al. 2009), but it is not clear how the findings from these systems transfer to wetlands or ponds. Small, isolated wetlands are abundant in most landscapes and harbor rich biotas, and yet their ecology is relatively unknown and understudied compared to other aquatic ecosystems. Ponds are an ideal system to test FCL controls because: 1) they exist along a natural gradient from temporary to permanent; 2) their boundary is easily delineated which is not the case for lotic systems (Post et al. 2007a, Gratton and Vander Zanden 2009); and 3) numerous ponds having similar species composition and physical makeup can be found in a small geographic area (Williams 2006, Sabo et al. 2009). I addressed the ecosystem size and dynamic constraints hypotheses using spatially and temporally variable ponds spanning a natural gradient of size and hydroperiod. Interestingly, the effect of ecosystem size on FCL was minor. My results strongly support the hypothesis that variability in hydroperiod, not ecosystem size, nor their interaction, dictates FCL in this system.

3.5.1 Ultimate determinants of FCL

We found that the variability of the physical environment is a critical influence on FCL (e.g., variation in pond depth vs. average pond depth). FCL ranged three trophic levels from the shortest hydroperiod to the longest hydroperiod ponds, clearly demonstrating that environmental

48 variability plays an important role in determining FCL. My data support the dynamic constraints hypothesis, which predicts that less predictable environments should have shorter FCLs. Disturbance events or other forms of environmental variability have not been found to be overly influential in determining FCL in the few systems in which it has been tested empirically (Townsend et al. 1998, Walters and Post 2008, McHugh et al. 2010). Sabo et al. (2009) suggested a FCL-environmental variation relationship would be found only in studies using a complete gradient including extremes. A study conducted in New Zealand streams ranging from small spring-fed to medium size permanent streams (McHugh et al. 2010) showed that FCL was shorter in habitats more variable in terms of temperature, hydrology and geomorphology. This study and that of McHugh et al. (2010) used naturally variable ecosystems including at least one extreme of the gradient giving a realistic picture of environmental variation effects in contrast to studies that have experimentally simulated a disturbance or were conducted in small container systems.

The ecosystem size hypothesis predicts that larger ecosystems will have longer food chains. Generally, I found the smallest pond had the shortest FCL, but the FCL of the largest pond was not different from the FCL of ponds in the intermediate size range. I conclude that in our study system, overall ecosystem size played a minor role in FCL variability. Often an increase in FCL is associated with increases in ecosystem size because new species are added to the species pool (Cohen and Newman 1992). In the ponds used in this study, the positive relationship between FCL and ecosystem size was fairly strong but was not statistically significant. This is curious given the strong relationship between species richness (insect and breeding amphibian) and pond size. This could be the result of organisms being added but not changing the feeding dynamics within the community. Predators in the study ponds fed on insects and tadpoles, both of which are omnivorous primary to intermediate level consumers, whose diets change little across the hydroperiod and size gradients (TAS unpublished), thus constraining FCL. This conclusion reflects that of McHugh et al. (2010) who found that FCL varied little across stream sizes.

The ecosystem size hypothesis has typically been tested on ecosystems of constant size. In this study, pond size was variable throughout both years, and larger area generally did not translate to a longer permanency. I used mean pond area calculated from monthly pond area measurements. The ponds shrank rapidly and changed shape over this time frame; for instance the difference in size from filling in April to mid-August was 1620.81 m2, or a 39% reduction, for one of the long

49 hydroperiod ponds. Short hydroperiod ponds experienced a 100% reduction in size because they dried completely. Food-chain length was not influenced by variability in pond area, however the variance of pond water depth negatively influenced FCL. In systems (e.g., lakes) where available habitat is unchanging, area and volume are good measures, although in systems like the ones in this study, variability is crucial in understanding food web structure. A measure of variability in ecosystem size is normally not taken into consideration in testing the ecosystem size hypothesis, but it could play a role in determining predator-prey interactions, population density, and competitive interactions which ultimately structure the food web. Relatively, low sample size and clustering of ponds at the tail-ends of the size range could also have contributed to the lack of ecosystem size-FCL fit in my study.

In summary, aquatic community composition and structure respond to environmental variability in habitat duration and available habitat. Hydroperiod is a strong determinant of FCL. Thus, alterations to climate that influence hydroperiod may have a dramatic effect on energy flow and ecosystem productivity.

3.5.2 Proximal mechanisms underlying FCL variability

Proximate structural changes within the food web can either lengthen or shorten FCL (Post and Takimoto 2007). Changes in community composition, for example, the insertion of new intermediate level consumers, may drive variation in FCL across ecosystem size and disturbance gradients. Species composition changes are well known to occur across hydroperiod gradients both for invertebrates (Williams 2006) and amphibians (Babbitt et al. 2003). Lake system studies have routinely found a positive relationship between species richness and FCL (Vander Zanden et al. 1999). In this study, hydroperiod influenced species composition leading to changes in FCL. A study on intermittent streams showed FCL increased with increasing predator species richness (Closs and Lake 1994) demonstrating the additive mechanism. However, my results indicate that species richness was not sufficient to explain variability in FCL across a gradient of ecosystem size. Over the study period, species richness and/or species composition were not static in these habitats. Different species of insects and amphibians entered and left the ponds at various times (i.e., insect emergence, breeding phenology, amphibian metamorphosis), creating fluctuating species richness and composition patterns. Yet, I used a measure of overall species richness that is an integrated rather than snapshot view of community dynamics that captured one

50 possible ecosystem size-FCL relationship in a system where quite possibly, many relationships could exist at different times. Post and Takimoto (2007) have shown that higher species richness can lead to a greater degree of trophic omnivory and shorten FCL or slow its rate of increase. Additional insect species probably did not add new trophic guilds but instead added functional redundancy within the study ponds. Omnivory was relatively constant across the gradients and did not show a strong influence on FCL in this study. Hydroperiod and pond size explained up to or more than half of the variation in insect and amphibian species richness, thus confirming the great importance of these environmental features to the faunal community dynamics in ponds. I suggest that in heterogeneous wetland habitats, synergistic interactions between species richness, degree of trophic omnivory, and functional diversity may dampen or obscure a pattern of increased FCL with increasing ecosystem size.

In the study ponds, FCL increased as the trophic position of top predators increased suggesting both insertion and omnivory mechanisms could have caused FCL variation. Insertion of new predatory insect species increased from short to long hydroperiod ponds (Fig. 3.4), but the number of breeding amphibians was highest in intermediate hydroperiod ponds (Fig. 3.5), a pattern reported by others (Werner et al. 2007). Primary consumer aquatic invertebrates and anuran tadpoles are prey for top predators (Ambystoma larvae and large predaceous insects) thereby shuttling energy gained from primary resources up the food chains. New intermediate level consumers were added to the food web redirecting energy flow patterns leading to top predators and creating variation in FCL across the hydroperiod gradient. Omnivory was constant across ponds, and therefore is ruled out as a driving mechanism behind FCL variability.

This field study using naturally variable aquatic habitats to understand the determinants of FCL, found that hydroperiod variability reduced FCL, but ecosystem size was not highly influential on FCL. This constitutes one of the few studies to empirically support the dynamic constraints hypothesis in real habitats. FCL variability is driven by underlying mechanisms of omnivory, addition of top predators, or changes in the trophic position of top predators (Post and Takimoto 2007). I showed insertion of new invertebrate and amphibian taxa was sufficient to elicit change in FCL across environmental gradients. The interaction of spatially and hydrologically variable habitats coupled with temporally variable amphibian and insect communities combined to form incredibly interesting food-web dynamics that are not easily disentangled.

Although, it has long been known that community composition and diversity are influenced by habitat stability, it has not been shown that a major component of food web structure, the FCL, is also strongly determined by environmental variation. FCL determines energy flow through and across ecosystems. Changes in temperature and precipitation can alter the natural predictability of the wet-dry phases of temporary ponds disrupting breeding and larval development. Therefore, deviations from the typical variability in the physical and chemical properties of the aquatic habitat can cascade across habitat boundaries to alter the magnitude of energy flow. This may have profound influences on temporary pond food webs and the adjacent riparian food web to which it is uniquely coupled. My data show that the less predictable a habitat is, the community is less diverse and the FCL is shorter.

Furthermore, temporary ponds frequently sit at a tipping point between ephemerality and permanency. In the US and Canada, temporary wetlands provide vital breeding habitat for amphibians (Semlitsch and Skelly 2008), but climate change (Magnuson et al. 1997, Pyke 2005), habitat loss, and poor policy regarding wetland importance and protection (Leibowitz 2003, Herbert et al. 2010) are all contributing to changing hydrologic regimes and possible decline of, and change in, faunal composition (Brooks 2009) of these habitats.

3.6 Acknowledgments

We send a special thanks to J. Arblaster, D. Bloom, S. Booth, G. Gill, G. Jegatheeswaran, M. Mahmood, F. Munro, R. Sambi, and S. R. Yang for their laboratory and field assistance. We thank Mike Delong and Marc Cadotte for their extremely insightful comments on manuscript drafts. This research was funded by an NSERC Discovery Grant (D. D. Williams) and a CGCS Graduate Student Research Award (T. A. Schriever). Animals were collected under scientific collector’s permit (# 1051193) approved by the Ontario MNR and used in accordance of animal care protocols approved by the University of Toronto and Queen's University Animal Care Committees (# 20007692).

3.7 Figures

3.1Relationships between FCL (maximum TP) and a) environmental variation (i.e., hydroperiod 2 in days, GLM: FCL = 1.0299 log (hydroperiod) + 1.0237, R = 0.23, P = 0.03), b) ecosystem size 2 (GLM: FCL = 0.5723 log (area) + 1.5265, R = 0.25, P = 0.09), and c) variation in water depth 2 (GLM: FCL = -0.5687CV water depth (m) + 3.7191, R = 0.38, P = 0.05).

3.3 Relationship between number of predatory insect families and ecosystem size (i.e., pond area (m2).

3.4 Relationship between predatory insect family richness and environmental variation (hydroperiod).

3.5 Non-linear relationship between breeding amphibian species richness and environmental variation (R2 = 0.57, P = 0.002).

3.6 Relationship between breeding amphibian species richness and ecosystem size (R2 = 0.55, P = 0.02).

3.7 Comparison of FCL variability amongst ponds and between sampling years. Intermediate hydroperiod ponds are denoted by Int and a corresponding number. Short3 and Int4 were not sampled in 2008.

4 Reciprocal resource flows in temporary ponds: the importance of community composition and hydroperiod

4.1 Abstract

Ecosystem functioning in open habitats is influenced by the flow of nutrients, detritus, and organisms and variation in these flows affects temporal and spatial patterns of community diversity and secondary production. I evaluated the influence of hydroperiod and ecosystem size on the bidirectional flow of energy from intermittent ponds and surrounding forest by quantifying the abundance and biomass of emerging insects and amphibians, and litter deposition. In addition we assessed whether species diversity influenced the magnitude of cross- habitat resource flux. I found spatial and temporal variation in the magnitude, composition, and timing of cross-habitat resource subsidies. Overall, deposition into ponds far exceeded carbon exported via insect and amphibian emergence. I found a negative relationship between resource flux and the diversity of amphibians and insects, which contradicts the general pattern of positive biodiversity-ecosystem function relationships. Different species groups differentially contributed to flux patterns with insects having higher diversity but lower flux compared to amphibians.

Organismal flux varied among ponds with amphibians having the highest flux in short hydroperiods and insect flux was highest from an intermediate hydroperiod pond. This work reveals how variation in pond size and permanence affects species diversity and ecosystem flows. Further, given the general lack of research and conservation prioritization of temporary ponds, uncovering how these ponds contribute to ecosystems is necessary to develop fully integrated management strategies.

4.2 Introduction

Environmental variation is a major force shaping community composition and population persistence and influences ecological processes at multiple scales (Sabo and Post 2008) and has direct consequences for different aspects of biodiversity, including: secondary production, species abundance, richness and diversity. A considerable accumulation of published research

61 has found that greater biodiversity results in higher ecosystem functioning (i.e. productivity and energy flow; (Loreau et al. 2001), yet studies often focus on the causes and consequences of diversity in isolation. For example, publications document the effect of disturbance events on ecological communities (Schriever et al. 2009, Bogan and Lytle 2011) while others measure species richness and community productivity (Tilman et al. 2001, Hooper et al. 2005, Maron et al. 2011). Whereas these types of studies have resulted in informative ecological patterns and processes, the question of how environmental variability affects the biodiversity-ecosystem function relationship remains (Romanuk et al. 2010).

Habitats are invariably open systems linked by the flow of nutrients, detritus, and organisms, all of which have the potential to influence population and community dynamics, food webs, and diversity-stability relationships in the recipient system (Polis et al. 1997, Nowlin et al. 2008, McCoy et al. 2009). Spatial flows may also have a large influence at the meta-ecosystem scale thereby strengthening connections among local ecosystems (Loreau et al. 2003). It is unclear how environmental variability interacts with species richness to influence cross-habitat energy flow. Spatial resource subsidies are spatially and temporally variable, with donor-control originating in one habitat and moving into another (Polis et al. 1997). The flow of resources across habitat boundaries is influenced by the distance separating habitats, habitat size, perimeter-to-area ratio of the focal habitat, adjacent habitat type, and the behavior of fluxes that move between habitats (Cadenasso et al. 2004). Spatial subsides can enter the food web at multiple trophic levels (Polis et al. 1997) and vary in quantity and quality.

Aquatic food webs, especially temporary woodland ponds, are intimately linked with the surrounding terrestrial landscape through a number of energy flow pathways. For example, aquatic systems contribute to the terrestrial food web aquatic insect emergences (Benke 1993, Stagliano et al. 1998, Gratton et al. 2008) and the reciprocal return of terrestrial material to aquatic food webs through litterfall, such as, organic matter deposition in streams (Fisher and Likens 1973, Hutchens and Wallace 2002, Rubbo and Kiesecker 2004, Rubbo et al. 2006), insect infall (periodical cicadas deposition into streams and ponds; (Pray et al. 2009), and nitrogen deposition (geese deposit nitrogen-rich waste into wetlands gained from foraging in agricultural land; (Post et al. 1998, Kitchell et al. 1999). It remains unclear how such resource subsidies influence the diversity-ecosystem function relationship.

Woodland ponds have diverse communities of insects and amphibians. Species in both groups undergo metamorphosis and leave the aquatic habitat and enter the terrestrial realm making them importance sources of energy flow between habitats (Schreiber and Rudolf 2008). These ponds are often relatively small, have high perimeter-to-area ratios, and are abundant. These unique features of pond geometry and community composition make temporary ponds great systems to use for study of cross-habitat energy flow. In an aquatic system, habitat duration (i.e., hydroperiod) influences species richness, diversity, community composition, and survivorship (Wellborn et al. 1996). Freshwater habitats along the hydroperiod gradient range from temporary habitats with relatively few small bodied predators, transitioning to permanent fishless habitats with large invertebrate predators, and at the end of the gradient permanent habitats with fish (Wellborn et al. 1996). I hypothesized the predictable shift in community composition across a hydroperiod gradient will have a noticeable and positive effect on the magnitude and/or composition of the cross-habitat energy flow. There are few studies that have quantified the connection of temporary ponds to the surrounding landscape. The quantity of inputs can provide a value predictor of subsidy effects on the recipient food web dynamics (Huxel and McCann 1998, Takimoto et al. 2002). Given the extensive occurrence of temporary ponds in temperate regions, ignoring their connections to surrounding habitat hinders our ability to adequately conserve the floras and faunas that use these habitats and manage the ecosystem services provided by these systems.

Here I examine the environmental (hydroperiod and ecosystem size) and community (species richness, diversity) influences on the magnitude of energy flow across habitat boundaries (strength of habitat coupling). I hypothesized that pond hydroperiod and size would strongly affect the magnitude of the aquatic-terrestrial linkage. In particular, I predict that longer hydroperiods and larger ecosystem size positively affect the magnitude of amphibian and insect emergence. I tested this hypothesis with a natural experiment using seasonally intermittent freshwater ponds varying in hydroperiod and size. Temporary habitat classification follows Williams (2006). I quantified the aquatic export of resources associated with the emergence of insects and metamorphosis of amphibians leaving ponds and the reciprocal terrestrial input of organic matter deposition of litterfall entering the ponds. I tested the relationship between biodiversity and-ecosystem function (BEF), predicting that lower diversity will result in reduced

63 ecosystem-level processes (Srivastava and Vellend 2005). I predict ponds with higher insect and amphibian diversity will export more energy compared to less diverse ponds.

4.3 Methods

4.3.1 Study sites and physical habitat

I conducted this investigation at the Queen’s Biological Station north of Kingston, Ontario, Canada (44.565977 N, -76.324223 W). The area is predominantly mixed coniferous-deciduous forest and supports numerous types of ponds. Ponds are filled by snow melt in early spring and lose water as the summer progresses, resulting in some ponds drying completely. I used four ponds that varied in hydroperiod and size (hereafter, Short3, Intermediate1, Intermediate3, Long2; Table 4.1). The experimental design took advantage of a natural environmental gradient, capitalizing on spatial and temporal variation, but this resulted in a trade-off in replication. I was most interested in measuring responses to natural environmental variation and laying a foundation for an unknown system. I used similar, non-replicated natural studies to check the generality of our results.

Hydroperiod was measured as the number of days with water. I measured pond area each month by mapping the perimeter of each using a Trimble TSC1 GPS connected to a ProXRS satellite and RTCM receiver. All mapping was completed in one or two consecutive days within each month. The area (m2) of each pond was then calculated in ArcMap10 (Esri, Redlands, CA, USA) and averaged to yield an average pond area for the entire study period. Every month I measured water depth and percent canopy cover measured at the center of each pond. I took four canopy readings facing each cardinal direction counting the number of open squares on a spherical densitometer. The four readings were averaged, multiplied by 4.17, and the product subtracted from 100% to obtain percent canopy density. Data loggers placed at the center of each pond ~ 3cm from pond bottom (StowAway Tidbit; Onset computer, Pocasset, MA, USA) recorded water temperature every 5 hours. These were averaged for a 24-hour period to acquire a daily temperature reading during the period of 7 April to 14 October 2010 or until the pond dried.

4.3.2 Field and laboratory methodology

To quantify emerging aquatic insects I placed three floating emergence traps (total trap area 0.63 m2) in each pond. Traps were deployed from May through August 2010 for approximately 48

64 hours collection periods 2 to 4 times a month (2032.5 total trap hours). Insects were aspirated from the traps and placed into vials containing 80% ethanol. All aquatic insects were counted and identified to family. Insects of non-aquatic origin (e.g. ants) were removed and not used in further analyses. I measured the body length (mm) of a subsample of random individuals in each family using a Nikon SMZ1500 microscope connected to a computer using the NIS Elements D3.1 software (Shinagaw-ku, Tokyo, Japan). I used published length-mass conversions (Appendix I) to estimate dry mass (DM) for each individual and calculated biomass for each family by summing biomass of all constituent individuals (Appendix J). Aquatic insect -2 emergence (Ei; g C m ) was calculated for each pond using dry mass estimates divided by trap area.

To capture metamorphosing amphibians, I encircled each pond with drift fence and sank pitfall traps (5 gal buckets) every 10 m flush with the fence and the ground along the pond side of the fence (Gibbons and Bennett 1974). Each trap contained a wet sponge to provide moisture to sustain trapped amphibians. This method allowed the capture of every amphibian moving out of the ponds, although treefrogs can climb over fences, and therefore their abundance may have been underestimated. Pitfall traps were opened and checked daily during the periods of 20 – 29 June, 7 – 13 and 19 – 22 July, 9 – 12 and 16 – 18 August for all ponds. Intermediate3 pond traps were also open 28 – 29 August. In total traps were open for 99 collection days. All amphibians collected were counted, identified to species, weighed, and the snout-to-vent length (SVL) measured with digital calipers. I euthanized a small sample of individuals for biomass estimation. Euthanized individuals were dried at 80C for 5 to 7 days, weighed (DM), then ashed at 500°C for 1 to 2 hr and reweighed to determine ash-free dry mass (AFDM). I established species-specific length–DM and length–AFDM relationships using linear regression to estimate dry mass and AFDM from the SVL of all individuals collected (Appendix K). These values were summed to provide the total biomass of amphibians from each pond. Amphibian emergence -2 production (Ea; g C m ) was calculated for each pond using dry mass estimates divided by pond area.

Coarse terrestrial particulate organic carbon (CTPOC) in the form of litterfall was collected once a month from two litter traps (0.5 m2 each) placed in each pond from August 29 through November 14, 2010. In the laboratory, bags of collected litter were emptied onto a white enamel tray, sorted into deciduous leaf (mostly Quercus sp., Acer sp., Betula sp.), coniferous needle

(Pinus sp.), woody material (twigs, branches, bark, cones, etc.), and miscellaneous (bud scales, pollen, etc.) components, dried at 55C for at least 96 hours, and weighed (g DM). Dried material was then ashed at 500C for 1 to 2 hours and reweighed to establish AFDM. I quantified -2 litter dry mass deposition for each trap (Dareal; g m ). Total litter deposition for each pond was the sum of litter dry mass collected over the sampling period.

I used Gratton and Vander Zanden’s (2009) measure of flux because it incorporates habitat geometry into the calculation of cross-habitat energy flow. They define “the flux (F) of emerging aquatic insects to land as the amount of insect production annually leaving the body of water per -1 -1 meter of aquatic shoreline (Fi; g C m yr )”. I used this metric of energy flow for amphibians

(Fa) as well since they are equivalent in terms of leaving the aquatic habitat. The calculation was: -2 -1 F  E(A/ p) , where E is the emergence in g C m yr of insects (Ei) or amphibians (Ea), A is the pond area and p is pond perimeter. C represents grams of dry mass carbon. The study ponds

1/ 2 were not perfectly circular therefore, a shoreline development factor ( DL  p /[A ]) corrected for irregularities in pond shape. I used the equation F  E(r / 2DL) , where r is pond radius. Litter input per meter of shoreline (Fl) flux to pond was calculated as Darealr / 2DL . I also used a measure of total flux to estimate the total aquatic to terrestrial carbon flux (Etotal) and total terrestrial to aquatic carbon flux (Dtotal) (Vander Zanden and Gratton 2011). I calculated Eitotal for insects and Eatotal for amphibians separately because I was interested in the magnitude of each subsidy type ( Eitotal  Ei  A; Eatotal  Ea  A ). An Etotal was calculated for each insect family and amphibian species per pond as well as the Etotal for each pond. Litterfall total flux was calculated -1 in a similar fashion (Dtotal = Dareal ×A; g C y ).

4.3.3 Statistical analyses

I used general linear models (GLM) in R (R Development Core Team 2009) to test for an effect of pond size and hydroperiod on the export fluxes of insect (i.e., Eitotal, Fi) and amphibians (i.e.,

Eatotal, Fa), and the input flux of litterfall (i.e, Dtotal, Fl). Ponds vary monthly in size and shape, so I used mean pond area (m2) as calculated from the available wetted months as the pond-size variable. I calculated Friedman rank sum tests, a nonparametric version of one-way ANOVA with repeated measures to test if the distributions of emergence and litterfall are the same across repeated sampling events. I also tested for differences in species richness, diversity, and evenness

66 in emergent amphibian and insect communities. GLMs were used to test the diversity- ecosystem function relationship between amphibian and insect diversity and total flux (Eitotal, Eatotal). I also looked at the relationship between litterfall and pond canopy cover.

4.4 Results

4.4.1 Timing of resource flows and environmental influences

Resource flows were out of phase in that insects emerged prior to amphibian metamorphosis from the ponds (Fig. 4.1) and litterfall input occurred after emergence. Emergent insect export began before and lasted almost as long as the export from amphibians. Insect emergence abundance peaked first in Short3 pond (14–16 May) followed by Intermediate3 (19–21 July) and then Intermediate1 and Long2 (16–18 August). Insect emergence increased over the four-month 2 sampling period (R = 0.1, F48 = 5.34, P = 0.025) with higher emergences just prior to pond drying. Amphibians had two major waves of emergence that were responsible for 43% of the total abundance collected (23–25 June and 8–10 July) which corresponded to the increase in temperature in one pond and the drawdown in another (Appendices L and M). The greatest number of amphibians (n = 154) leaving the ponds occurred on the 24th of June with most individuals originating from Short3 pond (121 wood frog metamorphs, Lithobates sylvaticus, and 1 newt, Notophthalmus viridescens) and the remainder from Intermediate3 (33 wood frog metamorphs). The number of metamorphosing amphibians declined over the sampling period (R2

= 0.09, F98 = 9.331, P = 0.003).

4.4.2 Magnitude of fluxes

-1 I collected 1050 individuals representing a total insect flux (Eitotal) of 3418.85 g C yr from aquatic to terrestrial habitat. This constituted a flux of 498.24 g C m-1 yr-1 or the emergent insect flux per meter of pond margin. Intermediate3 accounted for 95% of the total insect flux with the -2 highest abundance of emerging insects (791 individuals∙ m ), emergence production (Ei = 3.32 g -2 -1 -1 C m ) and flux (Fi = 474.07 g C m yr ) (Table 4.2). High flux was driven by presence of odonate species. In terms of abundance, Chironomidae midges made up 68% of collected individuals, but contributed only 2.4% to total flux (Eitotal) of all insect families (Fig. 4.2a).

Lestid damselflies were important contributors to the total biomass collected (83% or Etotal = 2840.63 g C yr-1) even though they represented only 0.004% of collected individuals (Fig. 4.2b).

Limnephilidae were collected from all ponds and although not abundant (n = 26), they -1 -1 -1 contributed a flux of 25.27 g C m yr or 5% of Etotal (172.16 g C yr ) (Fig. 4.2a). The number of emerging insects from each pond was not homogeneous across sampling events (Friedman rank sum test: χ2 (df = 3, N = 12) = 12.8974, P = 0.005).

The total amphibian flux and flux per meter of pond margin was nearly 6× higher than that from -1 -1 -1 insects (Eatotal = 18,833.9 g C yr ; Fa = 2899.33 g C m yr ) and the emergence 18× higher -2 -2 (mean Ea = 18.03 g C m ) than insect emergence (mean Ei = 1.00 g C m ) (Fig. 4.2c). Total abundance of amphibians emerging (n = 1197) from four ponds ranged from 123 individuals at Long2 to 571 individuals at Short3 between late June and late August 2010. A Friedman test was conducted to evaluate differences in the number of individuals emerging among the ponds across sampling events. The test was significant indicating that the number of emerging amphibians varied among ponds and time steps (χ2 (df = 3, N = 29) = 10.1774, P = 0.017). The pond with the 2 shortest hydroperiod (78 days) had the densest (3 individuals/m ) and highest emergence (Ea = 41.9 g C m-2 yr-1) of amphibians. Emerging wood frog metamorphs (L. sylvaticus) represented

71% of all individuals collected and 64% of the total flux (Eatotal) across ponds to the terrestrial landscape (Fig. 4.2c). Eatotal significantly differed among species (ANCOVA: P = 0.01, df =7, F = 3.66). A post-hoc holm adjustment showed L. sylvaticus - A. maculatum (P = 0.035), L. sylvaticus - Ambystoma sp. (P = 0.016) and N. viridescens - L. sylvaticus (P = 0.016) pairs were significantly different from each other in their contribution to the total flux of energy from aquatic to terrestrial habitat.

Overall, there was a carbon deposition into ponds that exceeded that exported via insect and amphibian emergence (Fig. 4.3). The average net energy gain (Dtotal − Eatotal + Eitotal) across ponds was 41,279.76 g C yr-1 (range: 21019.83 - 88312.94 g C yr-1). Deciduous leaves were the most important litter resource across the ponds (Fig. 4.4). Coniferous needles were of secondary importance to one pond (Short3). Litterfall deposition into ponds was directly related to canopy cover over the pond (GLM, poisson: P <0.0001, df = 3, Dtotal = 9.795 + 0.0159mean percent cover).

4.4.3 Diversity-ecosystem function response

Ponds with both higher amphibian and insect diversity had significantly lower flux per meter of pond margin (P = 0.04, df = 7, AIC = 119.27) and total flux (Etotal, P = 0.06, df = 7, AIC = 149.44; Fig. 4.5a). A similar negative, but not statistically significant association between flux

68 and species richness was seen (P = 0.24, df = 7, AIC = 123.22; Fig. 4.5b). Analyses separating insect and amphibian diversity indicated taxonomic groups had different diversity-ecosystem function relationships (Fig. 4.5). Emergent aquatic insect diversity exhibited a negative, but not significant, relationship with flux (Fi; P = 0.32, df = 3, AIC = 57.29) and total insect flux (Eitotal ; P = 0.32, df = 3, AIC = 72.83). However, insect diversity showed a significant non-linear relationship with Eitotal (P = 0.05, df = 2, AIC = 12.91; Fig. 4.5c). Amphibian flux (Fa) annually leaving the pond per meter of pond margin was not influenced by diversity (GLM: P = 0.62, df =

3, AIC = 63.73). The highest total amphibian flux (Eatotal) was from the least diverse community, but the relationship between Eatotal and diversity was not significant (P = 0.73, df = 3, AIC = 78.26).

Ponds with longer hydroperiod tended to have higher amphibian diversity, however the relationship was not statistically significant (GLM: P = 0.4, df = 3; Fig. 4.5d). Amphibian species richness was not related to hydroperiod (GLM: P = 0.89, df = 3). Amphibian richness and insect richness showed positive trends with increasing pond size (GLM: P = 0.10, df = 3, AIC = 11.55), which was best illustrated with a non-linear quadratic model (P = 0.03, df = 3). Neither a linear nor a quadratic fit explained the relationship between insect species richness and pond area (P = 0.44, df = 3, AIC = 30.41; P = 0.78, AIC = 7.47, df = 3, respectively). Insect richness showed a positive relationship with hydroperiod, but it was not significant (P = 0.13, df = 3, AIC = 26.34). Insect diversity was not related to hydroperiod (P = 0.73, df = 3, AIC = 5.47; Fig. 4.5d) or pond area (P = 0.38, df = 3, AIC = 3.87).

4.4.4 Influence of pond area and hydroperiod

Hydroperiod and pond area had differential effects on resource flows of amphibians and insects. Hydroperiod had a strong negative effect on the total amphibian flux to terrestrial habitat. The longer the hydroperiod the lower the magnitude of amphibian export (GLM: Eatotal; P = 0.08, df =

3, AIC = 71.21; Fa: P = 0.02, df = 3, AIC = 52.0; Fig. 4.6a). Pond area had no effect on the total amphibian flux to the terrestrial ecosystem (P = 0.79, df = 3, AIC = 78.40). Conversely, insect flux showed a strong positive relationship with pond area (GLM: Eitotal P = 0.004, df = 3, AIC =

55.6; Fi P = 0.004, df = 3, AIC = 40.75; Fig. 4.6b) and hydroperiod did not influence emergent insect total flux (GLM: Eitotal P = 0.9, df = 3, AIC = 75.27). Pond hydroperiod and area had no influence on the flux of litter input to ponds (GLM: hydroperiod Fl: P = 0.54, AIC = 78.99, df =

3, Dtotal P = 0.58, AIC = 97.90; Fig. 4.6a; Area Fl: P = 0.50, AIC = 78.78, df = 3, Dtotal P = 0.70, AIC = 98.30, Fig. 4.6b).

4.5 Discussion

Aquatic food webs are not isolated from the surrounding terrestrial landscape, especially for temporary woodland ponds, and are coupled through multiple flow types and from multiple trophic levels. I have shown that amphibians and insects can contribute a substantial amount of carbon to the terrestrial ecosystem, though litterfall input was substantially higher than reciprocal flows to adjacent terrestrial habitat. But these flows also depended on the physical and biological aspects of the ponds themselves. Pond hydroperiod had a strong negative effect on the total amphibian flux to the terrestrial habitat, but did not influence emergent insect total flux. The flux of amphibians was ~30% lower from longer hydroperiod ponds (168 days) compared to the short hydroperiod pond. Larger ponds had significantly higher total insect flux, but amphibian flux did not differ across pond size. Interestingly, passive allochthonous input of litterfall was not altered by pond hydroperiod or area.

It is generally accepted that amphibians and insects account for a large proportion of energy flow within aquatic and terrestrial food webs (Burton and Likens 1975, Pough 1980). However, the transfer of energy through amphibians between aquatic and terrestrial habitats was only recently quantified (Gibbons et al. 2006). Organisms of aquatic origin represent an energy subsidy to the terrestrial habitat in three ways: 1) as a prey source for terrestrial vertebrate and insect predators; 2) as a nitrogen and phosphorus source via excretion for soil and plant uptake; and 3) as a release of nutrients, through death, for decomposers, primary producers, and scavengers. Thus, energy subsidies provided by emerging insects and amphibians could influence recipient food webs by directly and indirectly affecting predators, decomposers, herbivores, and plants.

Few studies have simultaneously quantified multiple types of resource flow from a single habitat, and here I provide a unique picture of the energy exchange between the aquatic and terrestrial ecosystem by quantifying the magnitude of resource flows from obligate habitat couplers (amphibians and emergent aquatic insects) and a passive allochthonous subsidy (litterfall) input across a gradient of environmental variation. Regester et al. (2006) provided one of the first estimates of energy flow for amphibians using ambystomatid salamanders from woodland ponds to the surrounding forest. From one temporary and two permanent natural ponds they collected

662 emerging juvenile salamanders contributing 210 g ash-free dry mass to the adjacent forest ecosystem (no anurans were reported in their study). In comparison, I collected almost twice the number of amphibians (n = 1197) emerging from four ponds, but this represented half as much energy (carbon, 68 g ash-free dry mass). In this study, cross-habitat energy flow was dominated by anurans (64.5% and 35.5% salamanders), which generally weigh less on a per individual basis and thus contribute less energy than salamanders. Other studies like that of Gibbons et al. (2006) found anurans made up ~95% of exported amphibian biomass from one large open canopy pond in South Carolina, USA. Differences in amphibian community composition can have a profound effect on the amount of cross-habitat resource flow and striking consequences for, and influences on, the terrestrial ecosystem. This is another aspect of cross-habitat linkages that is understudied.

Magnitude and functional role of spatial subsidies can differentially impact the recipient food web (Wesner 2010). Some insects (odonates) and amphibians act as predators in the terrestrial food web and can have a high consumptive affect on arthropods as they disperse from the natal pond (McCoy et al. 2009). Wesner (2010) found the majority of aquatic insects emerging from small streams were non-consumers as adults, thus contributing as a major prey subsidy to the terrestrial food web. Intermittent ponds, therefore, have a dual role as providers of resource subsidies to terrestrial predators and a seasonal source of intermediate predators to the terrestrial food web, making a case that these aquatic habitats form multiple connections within discrete food webs.

There are many studies that have focused on one or two insect families, but there are surprisingly few studies that have measured the entire insect community production from wetlands. I collected individuals representing 36 insect families. Insect flux was 2× higher in our study ponds compared to that found by Whiles and Goldowitz (2001) from open canopy wetlands with and without fish in Nebraska. My insect emergence rates were similar for those estimated for streams (Ei = 1.052, range = 0.422 - 3.119 g C m-2 yr-1) and lakes (Ei = 0.318, range = 0.070 - 1.226 g C m-2 yr-1) by Gratton and Vander Zanden (2009). Only three ponds used in the Gratton and Vander Zanden study were similarly sized to our study ponds. Of the three, the largest pond (800 m2) in their study had comparable emergence production rates to a similar size pond (987.48 m2) in this study (our empirical value 3.32 vs. theirs 3.29 g C m-2 yr-1). Therefore, I have validated their model using empirical data from one extreme of the lentic habitat size gradient. My estimates of emergence numbers and magnitude of flux are likely conservative given that

71 odonates are not trapped with the same efficiency as other aquatic insects and that treefrogs can climb drift fences avoiding capture (Gibbons et al. 2006); therefore flux is likely underestimated.

Terrestrial primary production in the form of leaf litter, is a significant resource in woodland ponds forming the base of the aquatic food web (Batzer and Palik 2007). Litterfall is a donor controlled seasonally recurring resource subsidy to woodland ponds. This subsidy showed spatial variation (32.5% difference in Dtotal from lowest input to highest input). I measured almost 8.5× -1 more litter input (Dtotal = 187,371.8 g C yr ) than the combined total export flux of amphibian -1 and insect resources (Etotal = 22,252.75 g C yr ) from four ponds, resulting in an overall net carbon input. These results are consistent with Vander Zanden and Gratton’s (2011) modeled results of lake-to-land flux (Dtotal) that exceed insect emergence (Etotal) in small lakes (< 2000 ha). This may be common to most lakes and streams (Jackson and Fisher 1986, Leroux and Loreau 2008). My estimates of average litterfall input was similar to other published amounts (this study: 188 g C m-2 yr-1, mean 0.70 g C m-2 per day; Batzer and Palik 2007: 200 g m-2, Wallace et al. 1997: ~1.0 g m-2 per day).

4.5.1 Diversity-ecosystem function relationship

I did not find a positive relationship between diversity and ecosystem function; in fact I found, the opposite that communities with high insect or amphibian diversity result in lower amounts of energy flow to the terrestrial ecosystem. Pond communities tended to have higher insect diversity than amphibians, but considerable lower productivity (Etotal and emergent flux; Fig. 4.5a). A positive response is common in primary producers from terrestrial ecosystems (Schlapfer and Schmid 1999, Bouchard et al. 2007), but more complex responses are possible when studying multiple trophic levels of consumer groups (Duffy 2002, Hooper et al. 2005). The amphibian export flux was dominated by one species, wood frog, from one small, intermittent woodland pond with low insect and amphibian species richness. This result is contrary to the sampling effect model which states there is a higher chance of sampling a productive species in a species rich assemblage (Aarssen 1997, Tilman et al. 1997), but is consistent with the dominance effect (Norberg 2004). Wood frogs had higher biomass than the cumulative biomass of a more diverse amphibian community in an intermediate hydroperiod pond, perhaps because its development is best suited to temporary ponds (Paton and Crouch 2002). I speculate that predation, presence of salamander larvae, and insect predators, as well as competition from other larval anurans

72 decreased productivity in other ponds. Further, perhaps the positive relationship between biodiversity and ecosystem function may not be a general pattern for aquatic vertebrates (Cardinale et al. 2006).

Another potential reason why I did not see the expected pattern is because my study was observational and used natural assemblages rather than experimentally manipulating species richness and community composition. Although, this study is more ecologically realistic, many factors that cannot be controlled for in the field may confound the BEF relationship, therefore making it difficult to tease out potential causes of the observed pattern. Schlapfer and Schmid (1999), in a meta-analysis, found less than 25% of the cases with <10 species produced a positive BEF relationship. Amphibian species richness in this region of Canada is generally low, with 8 or fewer amphibian species in the study ponds. Naturally low amphibian species richness might explain the lack of positive BEF relationship. The greatest difference between observed amphibian species richness between any two of our ponds was 3 species. Schlapfer and Schmid (1999) found negative BEF relationships in 67% of cases where difference in species richness between treatments was low. However, this does not explain the lack of relationship for insects, because I observed fairly high insect species richness (up to 25 families) and a difference of 16 families between the lowest and highest species-rich pond. To understand the generality of the BEF relationship requires further research on a diversity of consumer groups and aquatic habitat types.

4.5.2 Conclusions

Intermittent ponds had a tight coupling with the terrestrial landscape in two reciprocal ways: ponds are heterotrophic depending on allochthonous carbon inputs to sustain food web and aquatic emergences of amphibians and insects provide resources to terrestrial vertebrates (e.g., adult chironomids) and become part of the terrestrial food web as consumers (e.g., amphibian juveniles). Cross-habitat resource flows were variable in time, space, and taxonomic group. Environmental variation in hydroperiod had a strong influence on amphibian flux to the adjacent terrestrial habitat, but less influential to insect flux. Diversity was negatively related to ecosystem productivity providing a ‘non-experimental’ view of the BEF relationship. Therefore, landscapes comprised of ponds varying in hydroperiod are essential to conserving insect and amphibian diversity and to maximize aquatic-terrestrial linkages.

4.6 Acknowledgements

Field work could not have been completed without the help of: Katherine Bannar-Martin, Devin Bloom, Kristen Brochu, Kirsten Comberford, Maria Modanu, Stephen Pynn, David Stitt, Caroline Tucker, Mark Conboy, Klara Jaspers-Fayer, Tristan Willis and Monica Candelaria. The tremendous help of Siao Ryan Yang and Ruby Sambi in processing samples is greatly appreciated. We thank Karen Pope for generously donating the emergence traps. This research was supported by an NSERC Discovery Grant awarded to DDW.

4.7 Tables

Table 4.1 Physical characteristics of the study ponds.

Pond

Measurement Short3 Inter1 Inter3 Long2

Mean depth (m) 0.26 0.25 0.47 0.36

Mean water temperature (C) 15 19 18 20

Mean canopy cover (%) 74.5 87.3 31.7 25.2

Mean pond area (m2) 182.68 249.13 987.48 188.52

Hydroperiod (days) 78 168 145 168

Insects Amphibians

Emergent flux† Emergence‡ Total Emergent Emergence‡ Total flux -1 -1 -2 (Fi, g C m yr ) (Ei, g C m ) flux flux† (Fa, (Ea, (Eatotal, (Eitotal, g C m-1 yr-1) g C m-2) g C yr-1) g C yr-1)

Short3 1.23 0.04 7.40 1273.26 41.91 7655.35

Int1 3.58 0.12 30.20 420.11 13.00 3541.57

Int3 474.07 3.32 3281.20 806.01 6.17 5569.05

Long2 19.35 0.53 100.04 399.94 11.06 2067.93

† Eshore in Vander Zanden and Gratton 2011

‡ Eareal in Vander Zanden and Gratton 2011

4.8 Figures

4.2 Flux (mean and standard error) attributed by each A) and B) insect family and C) amphibian species collected. Odonata is separated from other families for better viewing.

4.3Comparison of total carbon flux from pond-to-land (Etotal) (upper portion of figure) and land- to-pond (Dtotal) (lower portion of figure) attributed from each resource across study ponds.

5 Food webs in relation to environmental and species assemblage variation: A multivariate approach

5.1 Abstract

The abiotic environment can have strong influences on the growth, survival, behavior, and ecology of aquatic organisms. Biotic interactions and species life histories interact with abiotic factors to structure the food web. One measure of food-web structure is food-chain length.

Several hypotheses predict a linear relationship between one environmental variable (e.g., disturbance or ecosystem size) and food-chain length. However, many abiotic and biotic variables interact in diverse ways to structure a community, and may affect other measures of food web structure besides food-chain length. This study took a multivariate approach to test the influence of several important environmental variables on four food-web characteristics measured in nine ponds along a hydroperiod gradient over two years. I assessed the relationship between amphibian and invertebrate communities and pond habitat variables in order to understand the underlying food web structure. Hydroperiod and pond area had a strong influence on amphibian and invertebrate communities, trophic diversity and 15N range. The range in 13C values responded strongly to dissolved oxygen. Food-chain length did not respond strongly to any one variable, but instead responded weakly to multiple environmental variables. Invertebrate and amphibian communities were structured by pond hydroperiod which in turn influenced the trophic diversity of the food web. The results of this study suggest food-chain length is not the only food-web characteristic that is influenced by environmental variation and that a multivariate approach may allow us to better understand the dynamics within and across aquatic food webs.

5.2 Introduction

Temporary waters are abundant, diverse in physical, chemical, and biological characteristics, and are located throughout the world (Blaustein and Schwartz 2001) . Temporary waters fall on one extreme end of the habitat duration gradient, with lentic habitats ranging from small intermittent (temporary or episodic) to large permanent habitats (Williams 2006). The

84 hydroperiod (length of aquatic phase) is a major factor determining invertebrate and amphibian species diversity, composition, (Batzer and Wissinger 1996, Schneider and Frost 1996, Wellborn et al. 1996) and breeding success (Batzer and Wissinger 1996, Snodgrass et al. 2000). The hydroperiod gradient consists of temporary to semi-permanent ponds, to permanent ponds without fishes, to permanent ponds with fishes, and thus is characterized by variation in both abiotic (drying and associated changes in water chemistry) and biotic (increasing predation pressure) properties. The physicochemical environment of aquatic habitats is dynamic and can have severe consequences for the survival of its inhabitants. Aside from hydroperiod, water temperature is perhaps the most important abiotic factor in aquatic systems because of its effect on the growth and survival, behavior, food-web structure and ecology of aquatic organisms (Beveridge et al. 2010, Hoekman 2010). A full analysis describing the variation in aquatic communities across a hydroperiod gradient and how it pertains to food-web dynamics has not been conducted.

A current focus in food-web ecology is to understand how environmental variation influences food-web structure and function. Food webs depict consumer-resource (or predator-prey) interactions by characterizing trophic relationships among species or individuals in a particular habitat (Post 2002a). Stable isotopes are frequently used to examine the structure and function of food webs. In particular, carbon and nitrogen stable isotope ratios (13C and 15N, respectively) from members of the community are plotted in bi-variate space to give an overall picture of energy flow from the base (breadth of carbon used) to the top (height of the food-web along the vertical 15N axis). Typically, when food webs are compared across habitats or time simple qualitative methods are used to conclude food-web shifts; however, more quantitative methods are either rarely used, (although they have been recently advocated (Layman et al. 2007, Schmidt et al. 2007, Turner et al. 2010), or methods do not exist.

Food-chain length (FCL) is one common metric used to describe food-web structure. Several hypotheses have been proposed to explain the variability in, and the factors determining, food- chain length (Pimm 1982, Persson et al., 1996). Two such widely discussed, debated, and tested hypotheses are the dynamic constraints and the ecosystem size hypotheses (Post 2002a). The dynamic constraints hypothesis predicts that habitats subject to disturbance have shorter food chains because theoretical models predict longer food chains to be less resilient to disturbance or environmental variation and unattainable in nature (Pimm and Lawton 1977). However, there is

85 only limited empirical evidence that supports the idea that dynamical stability limits food-chain length (Post 2002a, Takimoto and Post 2012). The ecosystem size hypothesis predicts food-chain length will be longer in larger ecosystems because of higher species diversity, habitat availability and heterogeneity (Post et al. 2000), and more available resources (Schoener 1989). These hypotheses all predict a linear relationship between FCL and ecosystem size or disturbance; consequently simple linear models have been used to test them, even though a more complex relationship may be driving FCL. To my knowledge, no study has simultaneously examined the relationship between natural variation in the physicochemical environment (multiple parameters) and changes in food-web structure through a multivariate approach. Using multivariate approaches to study the impact of environmental variation on aquatic food webs is an appropriate direction given the temporal variability of these habitats and because most aquatic organisms are physiologically sensitive to environmental fluctuation. Temporary waters are abundant with diverse physical and biological characteristics, are essential habitats for many species of invertebrate, amphibian, and plant (Schneider and Frost 1996), and are located throughout the world (Blaustein and Schwartz 2001, Williams 2006). Despite this, scientists know far less about the food-web ecology of these smaller water bodies than larger, permanent habitats, such as lakes and rivers. Studies comparing food-web structure across a habitat duration gradient are few in number (Williams 2006), but have been advocated as a promising area of research effort in order to understand the relationship between dynamics and food-web structure (Winemiller and Polis 1996). These habitats are relatively small, therefore allowing numerous ponds to be compared and sampled. For these many reasons, extending testing of hypotheses of FCL and environmental variability to include multiple variables and responses in such ponds is a logical and promising research direction.

Previous tests have used single environmental variables to understand variation in food-chain length. FCL is a useful metric describing food-web structure because it can be compared across habitats (Post et al. 2000, Vander Zanden and Fetzer 2007) and influences ecosystem functions (Hairston et al. 1960, Duffy et al. 2005, Takimoto and Post 2012) such as: trophic cascades (Carpenter and Kitchell 1993); productivity (Pauly et al. 1998); and regulates bioaccumulation of toxins (Cabana et al. 1994). FCL is a summation of all consumer-resource interactions in a food web, yet other food-web properties or measures of food-web structure may also have a relationship to environmental variation. Layman et al. (2007) described a method of using

86 community-wide metrics calculated on stable isotope ratios to quantitatively assess food web changes across time or habitats. Three of Layman et al.’s metrics were used in this study: 13C range, 15N range, and total area of the food web (TAfoodweb), along with the routinely used metric FCL (MaxTP).

In this study, environmental variables in combination with food-web metrics were used to track how and if the structure of food webs changes along a hydroperiod gradient. To do so, FCL and three community-wide metrics developed by Layman et al. (2007): 15N range (represents extent of vertical structure), 13C range (denotes niche diversification at base of food web), and total area (proxy for the total extent of consumer trophic diversity in a food web) were calculated from 13C and 15N values for nine ponds. A redundancy analysis (RDA), with community metrics and FCL as response variables and environmental variables as predictor variables was computed to estimate the environment’s influence on food-web properties. I predicted that ponds with a longer hydroperiod would be associated with larger size, higher species richness, and an increased FCL and 15N range than ponds with a shorter hydroperiod. Community-wide metrics use mean 13C and 15N values from multiple individuals of a species within a defined habitat. I also determined the relationship between invertebrate and amphibian presence and their physicochemical pond environments using Canonical Correspondence Analysis (CCA). The main objective was to extend the traditional single variable, linear regression approach of testing environmental influence on FCL to a multivariate analysis of a few important environmental predictors and food-web response variables and to relate this to community structure. If the ecosystem size and dynamic constraints hypotheses are upheld, then a strong association between area and FCL and hydroperiod and FCL should be found.

5.3 Methods

5.3.1 Study sites

I characterized aquatic food webs, amphibian and invertebrate communities, and examined their relationships with environmental variables for nine natural ponds varying in hydroperiod on the Queen’s University Biological Station (QUBS) property north of Kingston, Ontario, Canada (301km [187miles] Northeast of Toronto). The hydroperiod gradient spanned from intermittent freshwater woodland ponds to near-permanent freshwater marshes (classification of temporary

87 waters follows (Williams 2006). All of these ponds are naturally fishless; therefore, the top predators are expected to be predaceous invertebrates and/or salamander larvae.

5.3.2 Field data collection

StowAway TidbiT temperature loggers (Onset Computer Corporation, Pocasset, MA, USA) were placed in the center of each pond at ~ 3cm from the bottom from the 4 May until 24 Oct. 2008 and from 23 April until 29 Oct. 2009 or until the ponds dried (Table 5.1). Temperature was read automatically and stored every 6 hours in 2008 and every 5 hours in 2009. Monthly sampling of the physicochemical environment in 2008 included measuring pH, dissolved oxygen (mg/L and percent of saturation), and air temperature (C). In addition, I measured overstory density by Spherical Densiometer (Forestry Suppliers, Inc.), water depth (m), and pond area (m2) in 2009. Water-quality measurements were taken by a hand-held Hydrolab Quanta multiparameter probe (Hach Environmental, Loveland, CO). To obtain percent canopy density, I took four canopy readings facing each cardinal direction counting the number of open squares on a spherical densitometer, the four readings were averaged, multiplied by 4.17, and the product subtracted from 100%. Pond perimeter was mapped using a Trimble handheld GPS Pathfinder® and pond area (m2) calculated in ArcMap.

A qualitative survey of the amphibian and invertebrate assemblage in each pond occurred simultaneously with food-web component collection and environmental measurements. I collected invertebrates and amphibians from numerous areas within each pond every two weeks (April – September 2008 and 2009) by dip-netting to comprehensively sample multiple trophic pathways leading to top consumers. Because of the relatively low species richness (four to nine species per pond) I could sample the entire amphibian assemblage. I sorted major invertebrate groups in the field, placing samples in plastic jars, and holding them on ice until deposited in a freezer. Amphibians were euthanized by immersion in Tricaine Methanesulfonate (MS-222) buffered with an equal amount of sodium bicarbonate to a pH of ~7.0. Samples were held on ice during field collection and then frozen until processed.

Food-web resources were collected monthly from each pond for isotopic analysis during the spring, summer, and fall seasons (April to September) of 2008 and 2009. Basal organic resources consisted of detritus, aquatic macrophytes (submerged and emergent), algae, seston, and fine benthic organic matter (FBOM). Seston was acquired by collecting a 1000-mL water sample at

88 mid-depth in the water column from each pond. FBOM was sampled by dredging a 53-μm net across the bottom of the pond in several random locations. The net contents were emptied onto a sieve tower of decreasing mesh size. The benthic matter collected on the 53-μm sieve was rinsed with distilled water into a collection jar. Samples were held on ice during field collection and then frozen until processed.

5.3.3 Laboratory and stable isotope analysis

Seston samples containing zooplankton, bacteria, and phytoplankton were obtained by filtering water samples onto pre-combusted Whatman glass microfiber GF/F filters. FBOM was poured into glass scintillation vials, allowed to settle, and the clear liquid was suctioned off using a pipette and discarded. The resultant concentrated FBOM was dried in scintillation vials. Aquatic macrophytes, algae, and coarse detritus samples were rinsed of attached periphyton or sediment and invertebrates (removal checked under a dissecting microscope). Tadpoles and salamander larvae were identified to species using specialist keys (Altig 1970, Altig and Ireland 1984, Petranka 1998). Using Merritt and Cummins (1996) and Marshall (2006) as guides, I identified and enumerated invertebrates to family or genus level under a dissecting microscope. Whole specimens minus viscera of tadpoles, salamander larvae, and invertebrates (snails and clams without shells) were used for stable isotope analysis (SIA). Replicates of amphibians represent single individuals, whereas replicates of invertebrates were composite samples of 2 to 50 individuals. Tadpoles, salamander larvae, basal resources, snails, and clams were rinsed with distilled water, dried in a drying oven at ~ 60C for 2-3 days, and ground to a homogeneous powder in a Mini-Beadbeater (BioSpec products Inc., OK, USA) for use in SIA. All samples were prepared in the Williams lab at the University of Toronto Scarborough and shipped to the Cornell Isotope Laboratory (COIL) for isotopic analysis. Stable isotope values are reported in 13 15 13 15 delta notation as  C or  N using the equation  C or  N = ((Rsample/Rstandard) – 1)  1000, where R is 13C: 12C or 15N: 14N. The standard is Vienna Pee Dee Belemnite for 13C and atmospheric nitrogen for 15N. Mathematical standardization was used to account for lipid variation in 13C in consumer samples (mean amphibian: 2008 C:N = 4.2, 2009 = 3.8; mean invertebrate: 2008 C:N = 5.0, 2009 = 4.9) using the recommended equations from Post et al. (2007b). We used 15N because it predictably increases through trophic transfer up the food chain, providing an estimate of vertical trophic position (Cabana and Rasmussen 1996, Vander

Zanden and Rasmussen 1999). 13C was used to estimate assimilation of basal energy sources by consumers (Peterson and Fry 1987, Fry 1991), because different energy sources often show distinct 13C signatures with little fractionation up the food chain (Hecky and Hesslein 1995).

Realized food-chain length was estimated as the maximum trophic position obtained by a species from each pond (Post 2002b). Trophic position was calculated for each individual using the mean 15N of basal resources as a baseline (Winemiller et al. 2007).

15 15 TPSI     Nsc  Nbaseline/ n

15 where λ is the trophic level of the baseline (1 for basal resources),  Nsc is the nitrogen isotope 15 signature of the consumer being evaluated and  Nbaseline is the mean nitrogen isotope signature of basal resources (algae, detritus, seston, FBOM, and aquatic plants). I used 2.3‰ for trophic level enrichment in 15N (n) based on a analysis of aquatic animals by McCutchan et al. (2003). Recent reviews of consumer-diet 15N enrichment (Vander Zanden and Rasmussen 2001, McCutchan et al. 2003, Vanderklift and Ponsard 2003, Caut et al. 2009) do not provide estimates for amphibians. Vanderklift and Ponsard (2003) found that organisms consuming detritus had significantly lower estimates of ∆ and that most aquatic organisms are ammonolectic, which averaged 2.0‰ per trophic step. Many researchers use ∆ of 3.4‰ based on Minagawa and Wada (1984) and Post (2002b). I decided on the 2.3‰ n value because it is specific to aquatic animals and is an intermediate value across the range of available estimates.

The community-wide metrics, 15N range, 13C range, and total area encompassed by the consumer food web (TAfoodweb) were calculated for each pond using mean consumer species 15N and 13C values. Total area was calculated using the geometry (Barber et al. 2012) package in R (R Development Core Team 2009) as the convex hull area encompassed by the pond community in isotopic space. The total area metric as proposed by Layman et al. (2007) has been criticized by Hoeinghaus and Zeug (2008) because they did not standardize isotopic values. According to Cornwell et al. (2006) it is not appropriate to calculate convex hull measurements on non-transformed data because traits are measured in different units and often have different variances, which is true for 15N and 13C. Therefore, I log transformed 15N and 13C values for the convex hull calculations. No amphibians were found in Winter1 during the 2009 collection period; therefore TAfoodweb calculated the extent of invertebrate trophic diversity, whereas

90 other TAfoodweb calculations included invertebrates and amphibians. Layman et al. (2007) states the convex hull of the consumer food web approximates the total area of the food web, even though I use the same terminology, I think this metric is a better descriptor of the community trophic space. The metric does not use the resources available to consumers in the food web therefore, is not estimating the entire food-web space.

5.3.4 Statistical analyses

Water temperature values were reduced to mean daily water temperature per pond for the length of the pond hydroperiod. I calculated the coefficient of variation (CV) to describe the dispersion of water temperature and water depth within a pond; the higher the CV, the greater the variability. Amphibian species and invertebrate family data sets were analyzed separately using correspondence analysis on presence-absence data. These communities were compared with the pond environmental data sets using canonical correspondence analysis (CCA). This direct gradient multivariate method summarizes the maximum amount of variation in the community data set while constraining it to axes associated with the environmental data (Jackson and Harvey 1993).

A redundancy analysis (RDA) was performed to test the influence of environmental factors on food web components. RDA is a direct gradient ordination method that can be used to test if species composition is related to a set of measured variables. RDA is similar to principal components analysis because the Euclidean distances among the objects in ordination space are preserved (Legendre and Legendre 1998). In the RDA ordination plot, the axes represent linear combinations of the explanatory variables. The length of the arrow relates to the strength of the relationship with the axis. RDAs were performed separately for 2008 and 2009 using means (or otherwise stated) of each predictor and response variable computed from the entire data set (April to October). The environmental predictor or constraining variables were: mean pH, mean dissolved oxygen (mg/L), hydroperiod (number of days with water), and the CV of water temperature (C). The response variables were food-chain length (maxTP), 15N range, 13C range, and total area encompassed of community trophic space (TAfoodweb) for 2008 and 2009 analyses. An extra RDA was run for 2009 data using three additional environmental variables that were not measured in 2008 (i.e., overstory density (meanCanopy), CV of water depth (m), and mean pond area (m2). The variance inflation factor (VIF), a statistic to detect

91 multicollinearity among independent variables, was calculated for all predictor variables for each year. VIF values were low ( 5), unless otherwise stated, so all variables were kept in further analyses. By using this approach, we show which ponds are characterized by particular environmental variables and whether food webs from a particular pond or group of ponds could be attributed to specific measured community metrics (Gotelli and Ellison 2004). I conducted permutation tests on the CCA and RDA models to test the significance of constraints. Multivariate statistics were computed in R version 2.15 using the vegan (Oksanen et al. 2012) (R Development Core Team 2009). Summary statistics were calculated using MYSYSTAT 12 (SYSTAT 2007).

5.4 Results

5.4.1 Physical-chemical pond environment

The hydroperiod varied between years in the study ponds. All ponds had water with partial ice coverage on the 3rd of April of both years. Link, Blue2, and QUBS had drastically reduced hydroperiods in 2009 compared to 2008 (Table 5.1). For example, Blue2 dried a month earlier in 2009 (6 August) compared to 2008 (between 20 September and 24 October).

5.4.2 Insect and amphibian community-environment relationships

The CA of the invertebrate community in 2008 explained 49% of the total variance in the first two axes (Table 5.2). Ordination plots of invertebrate families contrast drought-tolerant families on the left side of the first axis with short and intermediate hydroperiod ponds and less drought- resistant families with long hydroperiod ponds to the right of the axis (Appendix N). CCA results estimate 70% of the total variation in the invertebrate community in 2008 was explained by the environmental variables. Axis 1 represented a gradient of short hydroperiod ponds (far left side of the ordination) to long hydroperiod ponds (far right, Fig. 5.1A). The two shortest hydroperiod ponds (Winter1 and Winter2) were most closely associated with Anostraca (fairy shrimp), intermediate hydroperiod ponds associated with numerous taxa from beetles (Dytiscidae, Halipidae), dragonflies and damselflies (Libellulidae, Lestidae, Coenagrionidae) and caddisflies (Limnephilidae) to snails (Physidae, Lymnaeidae, Planorbidae) and clams (Sphaeriidae), whereas the two longest hydroperiod ponds were associated with Ceratopogonidae, Mesoveliidae, Nepidae, Carabidae, Hydrophilidae, Chysomelidae, and Hydraenidae (Fig 5.1A). Ponds with low

92 family richness were located to the left of the origin and more taxon rich ponds to the right. The percentage variation explained by Axis 2 was 26% and summarized variation in DO, CV of temperature, and pH, which were positively correlated. Blue2 was characterized by high dissolved oxygen, Gyrinidae beetles and Chaoboridae Phantom Midge larvae which were polarized by Link pond exhibiting low dissolved oxygen, little variation in temperature, low pH and most closely associated with isopods (Asellidae) that are tolerant of low dissolved oxygen. The permutation test for significance based on the CCA was not significant (P = 0.15) which was constrained by four environmental variables.

The CA of the amphibian community in 2008 summarized 75% of total variance in the first two axes (Table 5.2). The ponds were widely separated in the ordination plot suggesting distinct amphibian communities. All species except for one were positioned near the center of the plot because they were associated with more than one pond (Appendix O). The CCA of amphibian species and environmental data explained 53% of the total variability in the amphibian community. The first axis (CCA1) summarized 42% of the variation and was mostly strongly influenced by hydroperiod, meanwhile the other variables had similar influence but of opposite direction on the axis. Dissolved oxygen (DO) was strongly associated with the second axis (CCA2) as was CV of temperature, hydroperiod, and to a lesser degree with pH. Ambystoma maculatum (spotted salamander) was positioned in the origin, because it was found in all ponds except those with the two shortest hydroperiods. Longer hydroperiod ponds, including QUBS, Indian, and P82, had higher species richness and composition than shorter hydroperiod ponds, Winter1 and Winter2. Blue2 was distinctly characterized by high DO, high temperature variation, pH and the only pond to have Notophthalmus viridescens (red-spotted newt; Fig 5.2). The permutation test for significance based on the CCA was not significant (P = 0.84) which was constrained by four environmental variables.

The CA of the invertebrate community in 2009 explained 46% of the total variance in the first two axes (Table 5.2). All three short and one intermediate hydroperiod ponds fall to the left of axis 1 describing a strong hydroperiod gradient (axis 1 explains 28% of variance, Appendix P). I found no discernible pattern in axis 2. Eighty-seven percent of the total variation in the 2009 invertebrate community was explained by the environmental data set, using CCA. In accordance with the CA, all short hydroperiod ponds and one intermediate hydroperiod pond fell on the left of axis 1. Axis 1 described 29% of the variation and represents a gradient from shorter

93 hydroperiod, smaller ponds with dense canopy cover, high variation in water depth and temperature (left of origin in ordination) to longer hydroperiod, larger ponds with more open canopies and less variation in water depth and water temperature (right of origin; Fig. 5.1B). The second CCA axis explained 19% of the variation and was strongly determined by pH. Therefore, there were three main factors controlling aquatic invertebrate communities in 2009, one is related to physical pond properties, pond area and hydroperiod, the second and negatively correlated to the first is related to intrinsic pond conditions, variation in water depth and temperature, and extrinsic conditions (canopy cover) which can influence the intrinsic conditions. The third factor was orthogonal, therefore unrelated to the other two factors, and was influenced by water chemistry, pH and DO. Link did not show similar trends in pond conditions and invertebrate composition. Blue2 and Blue1 were opposite of Winter1 and Winter2 showing that these pond pairs contained different invertebrate fauna. Blue2 was characterized by high dissolved oxygen and accordingly the only pond associated with Ephemeroptera taxa (Potamanthidae, Baetidae, Caenidae, Fig. 5.1B). The permutation test for significance based on the 2009 CCA was not significant (P = 0.5).

The CA of the amphibian community in 2009 summarized 73% of the total variation in the first two axes (Table 5.2). Lithobates sylvaticus, Hyla versicolor, A. maculatum, A. laterale, and Psuedacris crucifer were near the center of the ordination and close to each other suggesting these species often co-occur in ponds (Appendix Q). Blue1, Winter2 and P82 were at opposite ends of the axis and pulled away from the origin, indicating that they had distinct amphibian communities. Notothalamus viridescens is farthest from the origin and positioned near P82 because it was rarely encountered, and P82 was where this salamander was found. No amphibians were found in Winter1 during 2009 so it was not included in the CA and CCA analyses. Using all seven environmental variables in the CCA caused the model to overfit the data. Examination of the VIF values indicated hydroperiod and CV water temperature were highly correlated with the other variables. Given this, I removed these variables from the dataset and re-ran the CCA using the remaining five environmental variables. The CCA of amphibian species and environmental data in 2009 explained approximately 82% of the total variation in the amphibian community. CCA Axis 1 explained 45% of the variation, most of which was associated with opposing responses - variation in water depth and dissolved oxygen. High dissolved oxygen was positively correlated with pH and associated with Bufo americanus and H.

94 versicolor. In contrast, low dissolved oxygen, low pH and high variation in water depth described pond communities containing A. maculatum and L. sylvaticus (left side of ordination). Notothalamus viridescens was associated with one pond in particular, and with higher dissolved oxygen and pH (right side of ordination) (Fig 5.2). The second axis explained the variation in species presence-absence in ponds by a strong negative relationship between pond area and canopy cover (proportion explained was 32%). Winter2 was a small pond with dense canopy cover harboring P. crucifer and A. laterale (lower half of ordination), whereas L. clamitans was found only in larger ponds with open canopies, most often in Indian pond. The permutation test for significance based on the CCA was not significant (P = 0.17) which was constrained by five environmental variables.

5.4.3 Food web-environment relationships

The RDA model indicated 68% of the variation in food-web structure in 2008 was explained by the environmental predictors. The first axis (RDA axis 1) explained 78% of the variability and described a gradient with ponds ranging from short hydroperiod, basic (higher pH), low temperature variability, and low dissolved oxygen (negative side of axis 1) to longer hydroperiod ponds with high variation in temperature, high dissolved oxygen, and high trophic diversity (TAfoodweb) (Fig. 5.3). Trophic diversity was strongly influenced by hydroperiod which is illustrated by the greater number of invertebrate taxa on the right half of the ordination (Fig. 5.1). RDA axis 2 explained 21% of the variation in the food web-environment relationship. Dissolved oxygen and pH were similarly associated with axis 2. The permutation test based all constrained eigenvalues was not significant (P = 0.45). The range in δ15N showed a strong negative response to increasing pH and responded heavily to hydroperiod. Winter2 was distinct from other ponds in terms of food-web structure and response to environmental variables. The length of the hydroperiod was positively correlated with variation in water temperature, which together strongly influenced TAfoodweb. MaxTP (food-chain length) was weakly positively related to Axis1 and positioned near the origin of the ordination plot, indicating either: 1) an overall general response to all variables rather than strongly responding to one environmental variable over another; or 2) a low response to all environmental variables. MaxTP was strongly positively correlated with hydroperiod (r = 0.74) and the total area of community trophic space (TAfoodweb; r = 0.91) which suggests the longer the hydroperiod the longer the food-chain length; with a longer food chain conferring a larger community trophic space or higher trophic

95 diversity. However, MaxTP was correlated to the other food-web metrics (range δ13C r = 0.78, range δ15N r = 0.44) potentially causing the variable to locate at the origin of the ordination.

The reduced RDA model using 2009 data explained 52% of the variation in community-wide metrics by the four environmental variables originally measured (i.e., CV temperature, DO, pH, hydroperiod). The first axis (RDA axis 1) explained 59% of the variation and was mostly summarized by pH and dissolved oxygen in one direction and hydroperiod in the other. MaxTP was weakly positively associated with axis 1, whereas range δ15N and range δ13C were strongly, but oppositely responding to axis 1 (Fig. 5.4). TAfoodweb responded to both axes, however more strongly to axis 2. The second axis (RDA axis 2) explained 40% of the variation and was most strongly associated with variation in water temperature (CVtemp). The range in δ13C had a small angle with the arrow for dissolved oxygen indicating that they were positively correlated. Range in δ13C was most strongly associated with axis 1 or pH. TAfoodweb and range δ15N responded positively to hydroperiod. Short and intermediate hydroperiod ponds had less temperature variation with the exception of Winter2. Longer hydroperiod ponds (QUBS, Indian, and P82, especially) had a strong influence on the range of δ15N and trophic diversity (TAfoodweb), but less affect on food-chain length (MaxTP). The permutation test based all constrained eigenvalues was not significant (P = 0.4).

The RDA model on the full 2009 environmental data set (7 variables) explained 99% of the variation leaving only 1% unexplained by the environmental variables in the model, meaning the model was very close to over fitting the data. Examination of the VIF indicated hydroperiod was highly correlated to other variables. Given this, I removed the variable from the dataset and re- ran the RDA using six environmental variables. The conclusions were similar between the analyses (with and without hydroperiod) because of the similar number of environmental variables to the number of sites. In 2009, 96% of the variation was explained in the RDA model by the first two axes, indicating that food-web variables could be explained mostly by the environmental variables even without hydroperiod (Fig. 5.5). All constrained canonical axes significantly summarized relationships between the food-web metrics and the environmental variables (permutation test P = 0.01). The coefficient of variation in water depth and pH showed a positive association. The first axis explained 68% of the variation and approximated a gradient from ponds characterized by higher dissolved oxygen and temperature variability, dense canopy cover, and a broader range in 13C (left side of origin) most closely associated with Link (mean

96 canopy cover = 88%) and Winter2 (mean DO = 6.35 mg/L). The second axis (RDA axis 2) summarized 28% of the variation and was polarized on one end by pond area and CV water temperature (upper half of ordination plot) and pH, CV water depth, and canopy cover on the other end (lower half of ordination plot). Indian pond was characterized by a large pond size, moderate variation in water temperature, higher dissolved oxygen, and a food web with increased 15N range and trophic diversity. In contrast, Winter1 was the smallest pond with the shortest food-chain length (MaxTP), narrower 13C and 15N ranges, and far less trophic diversity than Indian pond. Food-chain length (MaxTP) was positioned near the center of the ordination at the origin of the environmental vectors.

5.5 Discussion

I found that invertebrate and amphibian communities were structured by the pond hydroperiod which in turn, especially for the invertebrates, influenced the trophic diversity of the food web. Trophic diversity responded strongly to hydroperiod, pond area, and variation in water temperature in both years. Larger ponds with longer hydroperiods conferred higher species richness and trophic diversity. This was best illustrated in 2009 when invertebrate richness ranged from 17 to 41, and amphibian species richness ranged from 3 to 6 from short to longer hydroperiod ponds. 13C range also responded strongly to hydroperiod and variation in temperature in 2008, but was influenced by dissolved oxygen, temperature variation, pH, and canopy cover in 2009. Higher 13C range was seen in Indian pond in 2008, this is a large pond with a long hydroperiod, mostly open canopy, emergent and non-emergent aquatic plants, a large detrital component from leaf fall and aquatic plant senescence, and a rich invertebrate community (Schriever unpublished). The range in δ13C was positively correlated with dissolved oxygen indicating that ponds such as Link, which had high dissolved oxygen and was weakly acidic pH (mean = 6.90), also support diverse basal production sources (range δ13C). Range in 13C was often correlated with total area of community trophic space (TAfoodweb), indicating a strong association and a similar response to environmental variation. A wide 13C range suggests organisms in the community have the opportunity to feed on a diverse set of resources (Layman et al. 2007) and, as shown here this allows for higher trophic diversity and possibly longer food chains.

From this analysis, it is clear that FCL (MaxTP) was influenced by multiple environmental variables and was not strongly responsive to any one particular variable forcing it to the center of the ordinations. This finding has been suggested previously (Post 2002a). However, most studies still test the relationship between FCL and environmental variables one at a time. To list a few studies: ecosystem size (Vander Zanden et al. 1999, Post et al. 2000, Vander Zanden and Fetzer 2007) or environmental disturbance and/or variation (Schriever and Williams in review, (Takimoto et al. 2008, Walters and Post 2008). My results do not directly support the ecosystem size or dynamic constraints hypotheses because FCL was not strongly associated with pond area or hydroperiod. However, other predictors of environmental variation, in particular temperature and depth, were associated with variation in 15N, 13C, and trophic diversity. Longer FCL’s are attributed to larger habitat size through mechanisms of higher species richness, habitat heterogeneity (Post 2002a, Sabo et al. 2009), or changes in the degree of trophic complexity or omnivory (Post and Takimoto 2007). I found trophic diversity positively correlated with pond area; however, FCL (MaxTP) was only weakly correlated. These results give partial support for the productive-space hypothesis which states that ecosystem size allows for a greater FCL through an increase in food availability (Schoener 1989). In a recent meta-analysis, Takimoto and Post (2012) found significant positive mean effects of size and productivity indicating support for the ecosystem size hypothesis, and agreed that the productive-space mechanism could shape the FCL-ecosystem size relationship. In summary, I found other aspects of food- web structure besides FCL (i.e., trophic complexity, range in 15N and13C) are influenced by environmental variation in temperate ponds.

There are a few potential reasons for lack of support for the ecosystem size and dynamic constraints hypotheses. Perhaps the environmental gradients among ponds were not large enough to elicit a change in FCL. However, the study ponds ranged in size from 78 m2 to almost 3500 m2 and differed in hydroperiod by 291 days between the shortest and longest hydroperiod pond. Alternatively, given the disparity in environmental conditions the top predator of the system (usually A. laterale or A. maculatum) was consistent across ponds. Therefore, similarity in the top predator species could have constrained food-chain length amongst the ponds preventing large differences among the habitats under study. Additionally, community-wide metrics of food-web structure were weakly to moderately correlated with FCL, which could decrease the reliability of predicting the relationship between FCL and other environmental predictors.

A consistent pattern was seen when pond communities were compared across years. Variation in water temperature and dissolved oxygen were positively correlated and associated with variation in δ13C. Notothalamus viridescens, Veliidae, Ephemeroptera and Trichoptera taxa were more common in ponds having higher dissolved oxygen. On the other hand, ponds that had an intermediate hydroperiod (Blue1, Blue2, QUBS, Link) showed some similarities in associations with certain environmental variables, most often low temperature variation, dissolved oxygen, and pH, but did not show consistent patterns for food-web variables across years. This result exemplifies the strong environmental variability of intermittent ponds. Magnusson and Williams (2006) found natural temporal variation to be more important than biological factors in shaping the physicochemical environment in four, fishless, intermittent ponds. Therefore, our results suggest a common feature of intermittent systems in that they share strong associations among environmental variables making it more difficult to elucidate food-web structure.

5.5.1 Summary

Most food-web studies test the effect of one environmental variable on FCL. By extending the effects on FCL to a multivariate approach, I was able to evaluate the relative influence of several abiotic variables on multiple food-web components including FCL. I found short hydroperiod ponds are characteristically different in their amphibian and invertebrate communities and differ markedly from longer hydroperiod ponds in terms of food web responses and environmental predictors. Interestingly, some ponds were orthogonal to the main food-web axes demonstrating the complexity of and difficulty in describing food-web dynamics in intermittent habitats. In conclusion, I argue that to better understand diverse environmental influences on food-web dynamics, a multivariate approach including interacting factors should be adopted in theoretical and empirical research.

5.6 Acknowledgments

I thank the individuals who helped collect and process samples; Jennifer Arblaster, Sarah Booth, Maryam Mahmood, Gaayathiri Jegatheeswaran, Ruby Sambi, Gagan Gill, Devin Bloom, Shauna Bloom, Fiona Munro, Mark Conboy, and Jake Cowper Szamosi, the agencies that funded the work; Center for Global Change Science (TAS), Chicago Herpetological Society (TAS), and NSERC Discovery Grant (DDW), and Julie Helson who commented on an earlier manuscript

99 draft. I also thank Don Jackson who introduced me to multivariate statistics and provided valuable feedback on this project.

100

5.7 Tables

Table 5.1 Hydroperiod (days with water) of each pond.

Hydroperiod

(days)

Category Pond 2008 2009 short Winter1 65 74 short Winter2 65 74 short Snake NA 102 intermediate Link 147 126 intermediate QUBS 175 127 intermediate Blue2 175 125 intermediate Blue1 NA 140 long Indian 365 365 long P82 365 365

101

Table 5.2 Eigenvalues, percentage of variation explained, and permutation test statistics from CA of the invertebrate and amphibian presence-absence, and CCA of invertebrate and amphibian communities and environmental data from ponds in Southeastern Ontario.

Eigenvalues

Data set Axis Axis No. % of variation % of Pseudo- P- environmental explained by variation F all value 1 2 variables the explained axes all selected environmental by axes 1 axes variables and 2

Amphibian

CA 2008 0.222 0.098 NA 43 75 NA NA

CCA 2008 0.094 0.082 4 53 79 0.5611 0.84

CA 2009 0.255 0.190 NA 61 73 NA NA

CCA 2009 0.25 0.17 5 49 77 1.7682 0.17

Invertebrate

CA 2008 0.191 0.142 NA 67 49 NA NA

CCA 2008 0.187 0.124 4 70 66 1.1611 0.15

CA 2009 0.220 0.145 NA 80 46 NA NA

CCA 2009 0.202 0.129 7 75 52.5 1.0048 0.4

102

5.8 Figures

103

104

105

106

107

6 Thesis summary

My thesis focused on how environmental variability influences community, ecosystem dynamics and cross-habitat resource flows. This work fills gaps in knowledge on the trophic ecology of amphibian larvae, food-web dynamics of temporary ponds, and spatial subsidies linking aquatic and terrestrial ecosystems. I relate physicochemical variation to aquatic community dynamics and food-web structure, an approach seldom seen in the food-web literature. Further, I present an analysis of food-web structure across a gradient of water body permanence. I conclude that there is a general lack of protection and policy recognition for temporary aquatic habitats that does not match their importance in nature.

6.1 Individual, population and ontogenetic trophic niches of amphibian larvae

Chapter 2 quantifies the degree of individual specialization and diet variation over ontogeny in amphibian larvae across an environmental gradient. Niche variation is shown to be an important aspect in evolution and community ecology, with potentially profound influences on food-web structure. This research is novel in its use of larval amphibians to quantify niche patterns throughout ontogeny and to test the niche variation hypothesis using both stable isotopes and gut contents analyses. It provides a unique perspective of the niche within and among individuals from multiple trophic levels across a natural environmental gradient (pond duration). The results support the niche variation hypothesis indicating that populations with wider trophic niches also have a higher degree of individual specialization. Pond duration, however, did not significantly influence population total niche width or individual niche variation. This is an interesting pattern of conserved generalist feeding across an environmental gradient. Amphibian larvae had low individual variation which is contrary to the high individual specialization values that are commonly published for invertebrate and vertebrate taxa. It is important to document the variation in the degree of individual specialization across species and habitats to determine the incidence of this phenomenon in nature. Thus, I provide support for Van Valen’s niche variation hypothesis, reinforcing the observed pattern in nature but also extending the

108 pattern to include larval forms. The degree of individual specialization changed with ontogeny, an interesting finding needing more investigation. Surprisingly, the diet of most amphibian larvae, especially tadpoles is still poorly known (Altig et al. 2007). I provide significant information describing the functional role, ontogenetic diet shifts, and trophic position of two obligate temporary pond specialists in relation to natural environmental variation.

6.2 Hydroperiod determines food-chain length

Understanding what determines food-chain length in ecological systems is a longstanding focus in ecology. I used stable isotopes to calculate realized food-chain length and trophic position, and to measure omnivory in ponds varying in size and hydroperiod across two years (Chapter 3). I surveyed pond invertebrate and amphibian communities to establish community composition and diversity measures. The results strongly support the hypothesis that disturbance, not ecosystem size, nor their interaction, dictates food-chain length in these systems. My study provides a unique perspective of the complex inter-relationships between community dynamics and food- web structure in both temporary and permanent habitats. This is one of the few studies to empirically support the dynamic constraints hypothesis in real habitats, and its findings have broad implications for the conservation of wetland habitats. Addition of top predators and insertion of intermediate community members along the hydroperiod gradient from temporary to permanent ponds caused an increase in food-chain length, however these mechanisms were not sufficient to illicit a change in food-chain length across a gradient of ecosystem size.

6.3 Less diverse temporary ponds boast higher subsidy to surrounding terrestrial habitat

There has been increased focus in ecology on the importance of cross-habitat energy flow. Most habitats are open to multiple forms of resource subsidies; however most studies quantify only one conduit of energy flow. Although the connections between oceans and land, and streams to riparian habitats have been well established, the importance of wetlands to their surrounding terrestrial landscape has received far less study (Gratton and Vander Zanden 2009). My data (described in Chapter 4) show high insect and amphibian subsidies emerging from ponds, although both of these were lower than the organic carbon flux into the ponds. Measuring one trophic group would have produced a biased representation of energy flow in this system. Short hydroperiod ponds had lower insect and amphibian diversity, but high productivity. This does

109 not follow the generally positive pattern relating higher biodiversity to higher productivity. Insect and amphibian emergences from my ponds were high and similar to those of the few comparative studies available. I found that community composition played a large role in determining the magnitude of resource subsidy to the terrestrial ecosystem. Such research is critical to understanding the factors maintaining aquatic biodiversity to aid future management and conservation plans. If land use (e.g. agriculture practices) was to change in the terrestrial mosaic surrounding my study ponds, I predict that it would affect the productivity, food-web dynamics, and community structure of the local wetlands because of their tight coupling to the surrounding terrestrial habitat via multiple trophic pathways.

6.4 Importance of multiple environmental influences on aquatic community and food-web structure

Food-chain length is variable in time and space, influenced by multiple forms of environmental variation, dietary shifts of consumers, and seasonality. Despite the latter, current hypotheses addressing variation in food-chain length typically invoke a single variable that measures one aspect of the food web (i.e. food-chain length) and one determinant (e.g., disturbance or ecosystem size). Such one-dimensional studies invariably show conflicting relationships (positive, negative, non-significant, or mixed support; (Sabo et al. 2009). My approach was to investigate the relationships between multiple food-web metrics and multiple environmental factors to better describe food-web dynamics (Chapter 5).

A multivariate Redundancy Analysis explained food-web properties given a set of environmental variables across habitats and across years. Different metrics of food-web structure responded to different environmental variables, suggesting a non-uniform food-web response. Ponds with longer hydroperiods had more diverse invertebrate and amphibian communities which corresponded to greater trophic diversity and 15N range. However, food-chain length did not respond strongly to any one variable, but instead responded weakly to a combination of multiple environmental variables. I found the determinants of food-chain length in relatively small, isolated wetlands differed from the determinants in other habitat types (e.g., ecosystem size is a strong determinant of food-chain length in lakes). This reinforces the need for further investigation of the structure and function of temporary water systems.

110

7 References Aarssen, L. W. 1997. High productivity in grassland ecosystems: effected by species diversity or productive species? Oikos 80:183-184. Alford, R. A. 1999. Ecology: Resource Use, Competition, and Predation.in M. R.W. and A. R., editors. Tadpoles: the Biology of Anuran Larvae. University of Chicago Press, Chicago, IL, USA. Altig, R. 1970. A key to the tadpoles of continental United States and Canada. Herpetologica 26:180-207. Altig, R. and P. H. Ireland. 1984. A key to salamander larvae and larviform adults of the United States and Canada. Herpetologica 40:212-218. Altig, R. and G. F. Johnston. 1989. Guilds of anuran larvae: relationships among developmental modes, morphologies, and habitats. Herpetological monographs 3:81-109. Altig, R. and W. McDearman. 1975. Percent assimilation and clearance times of five anuran tadpoles. Herpetologica 31:67-69. Altig, R., M. R. Whiles, and C. L. Taylor. 2007. What do tadpoles really eat? Assessing the trophic status of an understudied and imperiled group of consumers in freshwater habitats. Freshwater Biology 52:386-395. Anderson, W. B. and G. A. Polis. 1998. Marine subsidies of island communities in the Gulf of California: evidence from stable carbon and nitrogen isotopes. Oikos 81:75-80. Angelibert, S., P. Marty, R. Cereghino, and N. Giani. 2004. Seasonal variations in the physical and chemical characteristics of ponds: implications for biodiversity conservation. Aquatic Conservation-Marine and Freshwater Ecosystems 14:439-456. Araújo, M. S., D. I. Bolnick, and C. A. Layman. 2011. The ecological causes of individual specialisation. Ecology Letters 14:948-958. Araújo, M. S., D. I. Bolnick, L. A. Martinelli, A. A. Giaretta, and S. F. dos Reis. 2009. Individual-level diet variation in four species of Brazilian frogs. Journal of Animal Ecology 78:848-856. Babbitt, K. J., M. J. Baber, and T. L. Tarr. 2003. Patterns of larval amphibian distribution along a wetland hydroperiod gradient. Canadian Journal of Zoology-Revue Canadienne De Zoologie 81:1539-1552. Barber, C. B., K. Habel, R. Grasman, R. B. Gramacy, A. Stahel, and D. C. Sterratt. 2012. geometry: Mesh generation and surface tesselation. Bardwell, J. H., C. M. Ritzi, and J. A. Parkhurst. 2007. Dietary selection among different size classes of larval Ambystoma jeffersonianum (Jefferson Salamanders). Northeastern Naturalist 14:293-299.

111

Batzer, D. P. and B. J. Palik. 2007. Variable response by aquatic invertebrates to experimental manipulations of leaf litter input into seasonal woodland ponds. Fundamental and Applied Limnology 168:155-162. Batzer, D. P., B. J. Palik, and R. Buech. 2004. Relationships between environmental characteristics and macroinvertebrate communities in seasonal woodland ponds of Minnesota. Journal of the North American Benthological Society 23:50-68. Batzer, D. P. and S. A. Wissinger. 1996. Ecology of insect communities in nontidal wetlands. Annual Review of Entomology 41:75-100. Bearhop, S., C. E. Adams, S. Waldron, R. A. Fuller, and H. Macleod. 2004. Determining trophic niche width: a novel approach using stable isotope analysis. Journal of Animal Ecology 73:1007-1012. Beaudoin, C. P., W. M. Tonn, E. E. Prepas, and L. I. Wassenaar. 1999. Individual specialization and trophic adaptability of Northern Pike (Esox lucius): an isotope and dietary analysis. Oecologia 120:386-396. Bengtsson, J. and N. D. Martinez, editors. 1996. Cause and effect in food webs: Do generalities exist? Chapman and Hall. Benke, A. C. 1993. Concepts and patterns of invertebrate production in running waters. Pages 15-38 in Congress of the International Assoc. of Theoretical and Applied Limnology, Barcelona, Spain. Beveridge, O. S., S. Humphries, and O. L. Petchey. 2010. The interacting effects of temperature and food chain length on trophic abundance and ecosystem function. Journal of Animal Ecology 79:693-700. Blaustein, L. and S. S. Schwartz. 2001. Why study ecology in temporary pools? Israel Journal of Zoology 47:303-312. Boecklen, W. J., C. T. Yarnes, B. A. Cook, and A. C. James. 2011. On the use of stable isotopes in trophic ecology. Annual Review of Ecology, Evolution, and Systematics 42:411-440. Bogan, M. T. and D. A. Lytle. 2011. Severe drought drives novel community trajectories in desert stream pools. Freshwater Biology 56:2070-2081. Bolnick, D. I., T. Ingram, W. E. Stutz, L. K. Snowberg, O. L. Lau, and J. S. Paull. 2010. Ecological release from interspecific competition leads to decoupled changes in population and individual niche width. Proceedings of the Royal Society B-Biological Sciences 277:1789-1797. Bolnick, D. I. and J. S. Paull. 2009. Morphological and dietary differences between individuals are weakly but positively correlated within a population of threespine stickleback. Evolutionary Ecology Research 11:1217-1233. Bolnick, D. I., R. Svanbäck, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. L. Forister. 2003a. The ecology of individuals: incidence and implications of individual specialization. American Naturalist 161:1-28. Bolnick, D. I., R. Svanbäck, J. A. Fordyce, L. H. Yang, J. M. Davis, C. D. Hulsey, and M. A. Forrister. 2003b. Comparative approaches to intra-population niche variation. Pages

112

1078-1078 in Annual Meeting of the Society for Integrative and Comparative Biology. Soc Integrative Comparative Biology, New Orleans, LA. Bolnick, D. I., R. Svänback, M. S. Araújo, and L. Persson. 2007. Comparative support for the niche variation hypothesis that more generalized populations also are more heterogeneous. Proceedings of the National Academy of Sciences of the United States of America 104:10075-10079. Bolnick, D. I., L. H. Yang, J. A. Fordyce, J. M. Davis, and R. Svanbäck. 2002. Measuring individual-level resource specialization. Ecology 83:2936-2941. Bonner, L. A., W. J. Diehl, and R. Altig. 1997. Physical, chemical and biological dynamics of five temporary dystrophic forest pools in central Mississippi. Hydrobiologia 353:77-89. Bouchard, V., S. D. Frey, J. M. Gilbert, and S. E. Reed. 2007. Effects of macrophyte functional group richness on emergent freshwater wetland functions. Ecology 88:2903-2914. Brooks, R. T. 2000. Annual and seasonal variation and the effects of hydroperiod on benthic macroinvertebrates of seasonal forest ("vernal") ponds in central Massachusetts, USA. Wetlands 20:707-715. Brooks, R. T. 2009. Potential impacts of global climate change on the hydrology and ecology of ephemeral freshwater systems of the forests of the northeastern United States. Climatic Change 95:469-483. Burton, T. M. and G. E. Likens. 1975. Salamander populations and biomass in Hubbard Brook Experimental Forest, New Hampshire. Copeia:541-546. Cabana, G. and J. B. Rasmussen. 1994. Modeling food-chain structure and contaminant bioaccumulation using stable nitrogen isotopes. Nature 372:255-257. Cabana, G. and J. B. Rasmussen. 1996. Comparison of aquatic food chains using nitrogen isotopes. Proceedings of the National Academy of Sciences of the United States of America 93:10844-10847. Cabana, G., A. Tremblay, J. Kalff, and J. B. Rasmussen. 1994. Pelagic food-chain structure in Ontario lakes - A determinant of mercury levels in Lake Trout (Salvelinus namaycush). Canadian Journal of Fisheries and Aquatic Sciences 51:381-389. Cadenasso, M. L., K. C. Weathers, and S. T. A. Pickett. 2004. Integrating food web and landscape ecology: subsidies at the regional scale. Pages 263-267 in G. A. Polis, M. E. Power, and G. R. Huxel, editors. Food Webs at the Landscape Level. The University of Chicago Press, Chicago. Cardinale, B. J., D. S. Srivastava, J. E. Duffy, J. P. Wright, A. L. Downing, M. Sankaran, and C. Jouseau. 2006. Effects of biodiversity on the functioning of trophic groups and ecosystems. Nature 443:989-992. Carpenter, S. R. and J. F. Kitchell, editors. 1993. The Trophic Cascade in Lakes. Cambridge Univ. Press, Cambridge. Caut, S., E. Angulo, and F. Courchamp. 2009. Variation in discrimination factors (Delta N-15 and Delta C-13): the effect of diet isotopic values and applications for diet reconstruction. Journal of Applied Ecology 46:443-453.

113

Closs, G. P. and P. S. Lake. 1994. Spatial and temporal variation in the structure of an intermittent stream food web. Ecological Monographs 64:1-21. Colburn, E. A. 2004. Vernal pools: Natural history and conservation. McDonald & Woodward, Blacksburg. Cornwell, W. K., D. W. Schwilk, and D. D. Ackerly. 2006. A trait-based test for habitat filtering: convex hull volume. Ecology 87:1465–1471. Costa, G. C., D. O. Mesquita, G. R. Colli, and L. J. Vitt. 2008. Niche expansion and the niche variation hypothesis: does the degree of individual variation increase in depauperate assemblages? American Naturalist 172:868-877. Dalerum, F. and A. Angerbjorn. 2005. Resolving temporal variation in vertebrate diets using naturally occurring stable isotopes. Oecologia 144:647-658. DeAngelis, D. L. 1992. Dynamics of Nutrient Cycling and Food Webs. Chapman and Hall, London. Deniro, M. J. and S. Epstein. 1978. Influence of diet on distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta 42:495-506. Deniro, M. J. and S. Epstein. 1981. Influence of diet on the distribution of nitrogen isotopes in animals. Geochimica et Cosmochimica Acta 45:341-351. Doi, H., K.-H. Chang, T. Ando, I. Ninomiya, H. Imai, and S. Nakano. 2009. Resource availability and ecosystem size predict food-chain length in pond ecosystems. Oikos 118:138-144. Donnelly, M. A. 1991. Feeding patterns of the Strawberry poison frog, Dendrobates pumilio pumilio (Anura, Dendrobatidae). Copeia:723-730. Dreyer, J., D. Hoekman, and C. Gratton. 2012. Lake-derived midges increase abundance of shoreline terrestrial arthropods via multiple trophic pathways. Oikos 121:252-258. Duellman, W. E. and L. Trueb. 1994. Biology of amphibians. Johns Hopkins University Press, Baltimore. Duffy, J. E. 2002. Biodiversity and ecosystem function: the consumer connection. Oikos 99:201- 219. Duffy, J. E., J. P. Richardson, and K. E. France. 2005. Ecosystem consequences of diversity depend on food chain length in estuarine vegetation. Ecology Letters 8:301-309. Dugan, P., editor. 1993. Wetlands in Danger: A World Conservation Atlas. Oxford University Press, New York. Eason, G. W. and J. E. Fauth. 2001. Ecological correlates of anuran species richness in temporary pools: A field study in South Carolina, USA. Israel Journal of Zoology 47:347-365. Eklöv, P. and R. Svanbäck. 2006. Predation risk influences adaptive morphological variation in fish populations. American Naturalist 167:440-452. Federal, P. a. T. G. o. C. 2010. Canadian biodiversity: ecosystem status and trends 2010. Canadian Councils of Resource Ministers, Ottawa.

114

Fisher, S. G. and G. E. Likens. 1973. Energy flow in Bear Brook, New Hampshire - Integrative approach to stream ecosystem metabolism. Ecological Monographs 43:421-439. Frederich, B., O. Lehanse, P. Vandewalle, and G. Lepoint. 2010. Trophic niche width, shift, and specialization of Dascyllus aruanus in Toliara Lagoon, Madagascar. Copeia:218-226. Fry, B. 1991. Stable isotope diagrams of freshwater food webs. Ecology 72:2293-2297. Gende, S. M., T. P. Quinn, M. F. Willson, R. Heintz, and T. M. Scott. 2004. Magnitude and fate of salmon-derived nutrients and energy in a coastal stream ecosystem. Journal of Freshwater Ecology 19:149-160. Gibbons, J. W. and D. H. Bennett. 1974. Determination of anuran terrestrial activity patterns by a drift fence method. Copeia 1974:236-243. Gibbons, J. W., C. T. Winne, D. E. Scott, J. D. Willson, X. Glaudas, K. M. Andrews, B. D. Todd, L. A. Fedewa, L. Wilkinson, R. N. Tsaliagos, S. J. Harper, J. L. Greene, T. D. Tuberville, B. S. Metts, M. E. Dorcast, J. P. Nestor, C. A. Young, T. Akre, R. N. Reed, K. A. Buhlmann, J. Norman, D. A. Croshaw, C. Hagen, and B. B. Rothermel. 2006. Remarkable amphibian biomass and abundance in an isolated wetland: implications for wetland conservation. Conservation Biology 20:1457-1465. Gosner, K. 1960. A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica 16:183-190. Gotelli, N. J. and A. M. Ellison. 2004. A Primer of Ecological Statistics. Sinauer Associates, Inc., Sunderland, MA. Gratton, C., J. Donaldson, and M. J. Vander Zanden. 2008. Ecosystem linkages between lakes and the surrounding terrestrial landscape in Northeast Iceland. Ecosystems 11:764-774. Gratton, C. and M. J. Vander Zanden. 2009. Flux of aquatic insect productivity to land: comparison of lentic and lotic ecosystems. Ecology 90:2689-2699. Hairston, N. G., F. E. Smith, and L. B. Slobodkin. 1960. Community structure, population control, and competition. American Naturalist 94:421-425. Halliday, T. 2005. Diverse phenomena influencing amphibian population declines. Pages 3-6 in M. Lannoo, editor. Amphibian declines: the conservation status of United States species. University of California Press, Berkeley. Hecky, R. E. and R. H. Hesslein. 1995. Contributions of benthic algae to lake food webs as revealed by stable isotope analysis. Journal of the North American Benthological Society 14:631-653. Herbert, M. E., P. B. McIntyre, P. J. Doran, J. D. Allan, and A. Abell. 2010. Terrestrial reserve networks do not adequately represent aquatic ecosystems. Conservation Biology 24:1002-1011. Hoeinghaus, D. J., K. O. Winemiller, and A. A. Agostinho. 2008. Hydrogeomorphology and river impoundment affect food-chain length of diverse Neotropical food webs. Oikos 117:984-995. Hoeinghaus, D. J. and S. C. Zeug. 2008. Can stable isotope ratios provide for community-wide measures of trophic structure? Comment. Ecology 89:2353-2357.

115

Hoekman, D. 2010. The effect of temperature on the relative importance of top-down and bottom-up effects. Ecology 91:2819-2825. Hoekman, D., J. Dreyer, R. D. Jackson, P. A. Townsend, and C. Gratton. 2011. Lake to land subsidies: experimental addition of aquatic insects increases terrestrial arthropod densities. Ecology 92:2063-2072. Hooper, D. U., F. S. Chapin, J. J. Ewel, A. Hector, P. Inchausti, S. Lavorel, J. H. Lawton, D. M. Lodge, M. Loreau, S. Naeem, B. Schmid, H. Setala, A. J. Symstad, J. Vandermeer, and D. A. Wardle. 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs 75:3-35. Hutchens, J. J. and J. B. Wallace. 2002. Ecosystem linkages between southern Appalachian headwater streams and their banks: Leaf litter breakdown and invertebrate assemblages. Ecosystems 5:80-91. Hutchinson, G. E. 1957. Population studies - Animal ecology and demography - Concluding remarks. Cold Spring Harbor Symposia on Quantitative Biology 22:415-427. Huxel, G. R. and K. McCann. 1998. Food web stability: the influence of trophic flows across habitats. American Naturalist 152:460-469. Jackson, D. A. and H. H. Harvey. 1993. Fish and benthic invertebrates: community concordance and community-environment relationships. Canadian Journal of Fisheries and Aqautic Sciences 50:2641-2651. Jackson, J. K. and S. G. Fisher. 1986. Secondary production, emergence, and export of aquatic insects of a Sonoran Desert stream. Ecology 67:629-638. Jenssen, T. A. 1967. Food habits of the green frog, Rana clamitans, before and during metamorphosis. Copeia 1:214-218. Kitchell, J. F., D. E. Schindler, B. R. Herwig, D. M. Post, M. H. Olson, and M. Oldham. 1999. Nutrient cycling at the landscape scale: The role of diel foraging migrations by geese at the Bosque del Apache National Wildlife Refuge, New Mexico. Limnology and Oceanography 44:828-836. Layman, C. A., D. A. Arrington, C. G. Montana, and D. M. Post. 2007. Can stable isotope ratios provide for community-wide measures of trophic structure? Ecology 88:42-48. Legendre, L. and P. Legendre. 1998. Numerical Ecology. 2 edition. Elsevier Scientific Publishing Company, Amsterdam. Leibowitz, S. G. 2003. Isolated wetlands and their functions: an ecological perspective. Wetlands 23:517-531. Lepak, J. M., C. E. Kraft, and B. C. Weldel. 2006. Rapid food web recovery in response to removal of an introduced apex predator. Canadian Journal of Fisheries and Aquatic Sciences 63:569-575. Leroux, S. J. and M. Loreau. 2008. Subsidy hypothesis and strength of trophic cascades across ecosystems. Ecology Letters 11:1147-1156. Loreau, M., N. Mouquet, and R. D. Holt. 2003. Meta-ecosystems: a theoretical framework for a spatial ecosystem ecology. Ecology Letters 6:673-679.

116

Loreau, M., S. Naeem, P. Inchausti, J. Bengtsson, J. P. Grime, A. Hector, D. U. Hooper, M. A. Huston, D. Raffaelli, B. Schmid, D. Tilman, and D. A. Wardle. 2001. Ecology - Biodiversity and ecosystem functioning: Current knowledge and future challenges. Science 294:804-808. Magnuson, J. J., K. E. Webster, R. A. Assel, C. J. Bowser, P. J. Dillon, J. G. Eaton, H. E. Evans, E. J. Fee, R. I. Hall, L. R. Mortsch, D. W. Schindler, and F. H. Quinn. 1997. Potential effects of climate changes on aquatic systems: Laurentian Great Lakes and precambrian shield region. Hydrological Processes 11:825-871. Magnusson, A. K. and D. D. Williams. 2006. The roles of natural temporal and spatial variation versus biotic influences in shaping the physicochemical environment of intermittent ponds: a case study. Archiv fur Hydrobiologie 165:537-556. Mantel, S. M. K., M. Salas, and D. Dudgeon. 2004. Foodweb structure in a tropical Asian forest stream. Journal of the North American Benthological Society 23:728-755. Maron, J. L., M. Marler, J. N. Klironomos, and C. C. Cleveland. 2011. Soil fungal pathogens and the relationship between plant diversity and productivity. Ecology Letters 14:36-41. Marshall, S. A. 2006. Insects: The natural history and diversity: with a photographic guide to insects of Eastern North America. Firefly Books Ltd., Richmond Hill, ON. Martinez del Rio, C., N. Wolf, S. A. Carleton, and L. Z. Gannes. 2009. Isotopic ecology ten years after a call for more laboratory experiments. Biological Reviews 84:91-111. McCoy, M. W., M. Barfield, and R. D. Holt. 2009. Predator shadows: complex life histories as generators of spatially patterned indirect interactions across ecosystems. Oikos 118:87- 100. McCutchan, J. H., W. M. Lewis, C. Kendall, and C. C. McGrath. 2003. Variation in trophic shift for stable isotope ratios of carbon, nitrogen, and sulfur. Oikos 102:378-390. McHugh, P. A., A. R. McIntosh, and P. G. Jellyman. 2010. Dual influences of ecosystem size and disturbance on food chain length in streams. Ecology Letters 13:881-890. McIntyre, P. B. and A. S. Flecker. 2006. Rapid turnover of tissue nitrogen of primary consumers in tropical freshwaters. Oecologia 148:12-21. Merritt, R. W. and K. W. Cummins. 1996. An Introduction to the Aquatic Insects of North America. Kendal/Hunt Publishing Company, Dubuque. Minagawa, M. and E. Wada. 1984. Stepwise enrichment of 15N along food chains: further evidenceand the relation between 15N and animal age. Geochimica et Cosmochimica Acta 48:1135-1140. Newsome, S. D., C. M. del Rio, S. Bearhop, and D. L. Phillips. 2007. A niche for isotopic ecology. Frontiers in Ecology and the Environment 5:429-436. Norberg, J. 2004. Biodiversity and ecosystem functioning: a complex adaptive systems approach. Limnology and Oceanography 49:1269-1277. Nowlin, W. H., M. J. Gonzalez, M. J. Vanni, M. H. H. Stevens, M. W. Fields, and J. J. Valentei. 2007. Allochthonous subsidy of periodical cicadas affects the dynamics and stability of pond communities. Ecology 88:2174-2186.

117

Nowlin, W. H., M. J. Vanni, and L. H. Yang. 2008. Comparing resource pulses in aquatic and terrestrial ecosystems. Ecology 89:647-659. O'Toole, S., C. Metcalfe, I. Craine, and M. Gross. 2006. Release of persistent organic contaminants from carcasses of Lake Ontario Chinook salmon (Oncorhynchus tshawytscha). Environmental Pollution 140:102-113. Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, P. R. Minchin, R. B. O'Hara, G. L. Simpson, P. Solymos, M. Henry, H. Stevens, and H. Wagner. 2012. Vegan: Community Ecology Package. Opsahl, S. P., S. W. Golladay, L. L. Smith, and S. E. Allums. 2010. Resource-consumer relationships and baseline stable isotopic signatures of food webs in isolated wetlands. Wetlands 30:1213-1224. Paine, R. T. 1980. Food webs: linkage, interaction strength and community infrastructure. Journal of Animal Ecology 49:666-685. Paton, P. W. C. and W. B. Crouch. 2002. Using the phenology of pond-breeding amphibians to develop conservation strategies. Conservation Biology 16:194-204. Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F. Torres. 1998. Fishing down marine food webs. Science 279:860-863. Pechmann, J. H. K., D. E. Scott, G. J. W., and R. D. Semlitsch. 1989. Influence of wetland hydroperiod on diversity and abundance of metamorphosing juvenile amphibians. Wetlands Ecology and Management 1:3-11. Persson, L., J. Bengtsson, B. A. Menge, and M. E. Power. 1996. Productivity and consumer regulation-concepts, patterns, and mechanisms. Pages 396-434 in G. A. Polis and K. O. Winemiller, editors. Food Webs: Integration of Patterns and Dynamics. Chapman and Hall. Peterson, B. J. and B. Fry. 1987. Stable isotopes in ecosystem studies. Annual Review of Ecology and Systematics 18:293-320. Petranka, J. W. 1998. Salamanders of the United States and Canada. Smithsonian Books. Petranka, J. W. and C. A. Kennedy. 1999. Pond tadpoles with generalized morphology: is it time to reconsider their functional roles in aquatic communities? Oecologia 120:621-631. Phillips, D. L. and J. W. Gregg. 2003. Source partitioning using stable isotopes: coping with too many sources. Oecologia 136:261-269. Piankas, L., M. S. Oliphant, and I. L. K. Iverson. 1970. Food habits of Albacore, Bluefin Tuna, and Bonito in California waters. Fish Bulletin 152:1-105. Pimm, S. L. 1982. Food Webs. Chapman and Hall, London. Pimm, S. L. and J. H. Lawton. 1977. Number of trophic levels in ecological communities. Nature 268:329-331. Polis, G. A. 1984. Age structure component of niche width and intraspecific resource partitioning - Can age-groups function as ecological species. American Naturalist 123:541-564.

118

Polis, G. A. 1991. Complex trophic interactions in deserts- an empirical critique of food-web theory. 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. Polis, G. A. and S. D. Hurd. 1996. Linking marine and terrestrial food webs: allochthonous input from the ocean supports high secondary productivity on small islands and coastal land communities. American Naturalist 147:396-423. Polis, G. A. and D. R. Strong. 1996. Food web complexity and community dynamics. American Naturalist 147:813-846. Post, D. M. 2002a. The long and short of food-chain length. Trends in Ecology & Evolution 17:269-277. Post, D. M. 2002b. Using stable isotopes to estimate trophic position: Models, methods, and assumptions. Ecology 83:703-718. Post, D. M. 2003. Individual variation in the timing of ontogenetic niche shifts in largemouth bass. Ecology 84:1298-1310. Post, D. M., M. W. Doyle, J. L. Sabo, and J. C. Finlay. 2007a. The problem of boundaries in defining ecosystems: A potential landmine for uniting geomorphology and ecology. Geomorphology 89:111-126. Post, D. M., C. A. Layman, D. A. Arrington, G. Takimoto, J. Quattrochi, and C. G. Montana. 2007b. Getting to the fat of the matter: models, methods and assumptions for dealing with lipids in stable isotope analyses. Oecologia 152:179-189. Post, D. M., M. L. Pace, and N. G. Hairston. 2000. Ecosystem size determines food-chain length in lakes. Nature 405:1047-1049. Post, D. M. and G. Takimoto. 2007. Proximate structural mechanisms for variation in food-chain length. Oikos 116:775-782. Post, D. M., J. P. Taylor, J. F. Kitchell, M. H. Olson, D. E. Schindler, and B. R. Herwig. 1998. The role of migratory waterfowl as nutrient vectors in a managed wetland. Conservation Biology 12:910-920. Pough, F. H. 1980. Advantages of ectothermy for tetrapods. American Naturalist 115:92-112. Pray, C. L., W. H. Nowlin, and M. J. Vanni. 2009. Deposition and decomposition of periodical cicadas (Homoptera: Cicadidae: Magicicada) in woodland aquatic ecosystems. Journal of the North American Benthological Society 28:181-195. Pyke, C. R. 2005. Assessing climate change impacts on vernal pool ecosystems and endemic brachiopods. Ecosystems 8:95-105. R Development Core Team. 2009. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Regester, K. J., K. R. Lips, and M. R. Whiles. 2006. Energy flow and subsidies associated with the complex life cycle of ambystomatid salamanders in ponds and adjacent forest in southern Illinois. Oecologia 147:303-314.

119

Reich, K. J., K. A. Bjorndal, and C. Martinez Del Rio. 2008. Effects of growth and tissue type on the kinetics of 13C and 15N incorporation in a rapidly growing ectotherm. Oecologia 155:651–658. Rogers, T. N. and D. R. Chalcraft. 2008. Pond hydroperiod alters the effect of density dependent processes on larval anurans. Canadian Journal of Fisheries and Aquatic Sciences 65:2761–2768. Romanuk, T. N., R. J. Vogt, A. Young, C. Tuck, and M. W. Carscallen. 2010. Maintenance of positive diversity-stability relations along a gradient of environmental stress. Plos One 5:1-9. Roughgarden, J. 1972. Evolution of niche width. American Naturalist 106:683-718. Roughgarden, J. 1974. Niche width - Biogeographic patterns among Anolis lizard populations. American Naturalist 108:429-442. Roughgarden, J. 1979. Theory of Population Genetics and Evolutionary Ecology: An Introduction. Macmillan Publishing Company, New York. Rowe, C. L. and W. A. Dunson. 1995. Impacts of hydroperiod on growth and survival of larval amphibians in temporary ponds of central Pennsylvania, USA. Oecologia 102:397-403. Rubbo, M. J., J. J. Cole, and J. M. Kiesecker. 2006. Terrestrial subsidies of organic carbon support net ecosystem production in temporary forest ponds: Evidence from an ecosystem experiment. Ecosystems 9:1170-1176. Rubbo, M. J. and J. M. Kiesecker. 2004. Leaf litter composition and community structure: Translating regional species changes into local dynamics. Ecology 85:2519-2525. Sabo, J. L., J. C. Finlay, T. Kennedy, and D. M. Post. 2010. The role of discharge variation in scaling of drainage area and food chain length in rivers. Science 330:965-967. Sabo, J. L., J. C. Finlay, and D. M. Post. 2009. Food chains in freshwaters. Annals of the New York Academy of Sciences 1162:187-220. Sabo, J. L. and D. M. Post. 2008. Quantifying periodic, stochastic, and catastrophic environmental variation. Ecological Monographs 78:19-40. Sandlin, E. A. and M. R. Willig. 1993. Effects of age, sex, prior experience, and intraspecific food variation on diet composition of a tropical folivore (Phasmatodea, Phasmatidae). Environmental Entomology 22:625-633. Schiesari, L. 2006. Pond canopy cover: a resource gradient for anuran larvae. Freshwater Biology 51:412–423. Schiesari, L., E. E. Werner, and G. W. Kling. 2009. Carnivory and resource-based niche differentiation in anuran larvae: implications for food web and experimental ecology. Freshwater Biology 54:572-586. Schlapfer, F. and B. Schmid. 1999. Ecosystem effects of biodiversity: a classification of hypotheses and exploration of empirical results. Ecological Applications 9:893-912. Schmidt, S. N., J. D. Olden, C. T. Solomon, and M. J. Vander Zanden. 2007. Quantitative approaches to the analysis of stable isotope food web data. Ecology 88:2793-2802.

120

Schneider, D. W. and T. M. Frost. 1996. Habitat duration and community structure in temporary ponds. Journal of the North American Benthological Society 15:64-86. Schoener, T. W. 1989. Food webs from the small to the large: the Robert H. MacArthur Award Lecture. Ecology 70:1559-1589 Schreiber, S. and V. H. W. Rudolf. 2008. Crossing habitat boundaries: coupling dynamics of ecosystems through complex life cycles. Ecology Letters 11:576-587. Schreiber, S. J., R. Bürger, and D. I. Bolnick. 2011. The community effects of phenotypic and genetic variation within a predator population. Ecology 92 1582-1593. Schriever, T. A., J. Ramspott, B. I. Crother, and C. L. Fontenot. 2009. Effects of hurricanes Ivan, Katrina, and Rita on a southeastern Louisiana herpetofauna. Wetlands 29:112-122. Seale, D. B. 1980. Influence of amphibian larvae on primary production, nutrient flux, and competition in a pond ecosystem. Ecology 61:1531-1550. Semlitsch, R. D. and D. K. Skelly. 2008. Ecology and conservation of pool-breeding amphibians Pages 127–147 in A. J. K. Calhoun and P. G. deMaynadier, editors. Science and conservation of vernal pools in northeastern North America. CRC Press, Boca Raton. Skelly, D. K. 1996. Pond drying, predators, and the distribution of Pseudacris tadpoles. Copeia 1996:599-605. Snodgrass, J. W., M. J. Komoroski, A. L. Bryan, and J. Burger. 2000. Relationships among isolated wetland size, hydroperiod, and amphibian species richness: implications for wetland regulations. Conservation Biology 14:414-419. Srivastava, D. S. and M. Vellend. 2005. Biodiversity-ecosystem function research: is it relevant to conservation? Annual Review of Ecology, Evolution, and Systematics 36:267-294. Stagliano, D. M., A. C. Benke, and D. H. Anderson. 1998. Emergence of aquatic insects from 2 habitats in a small wetland of the southeastern USA: temporal patterns of numbers and biomass. Journal of the North American Benthological Society 17:37-53. Starrett, P. H. 1973. Evolutionary patterns in larval morphology. Pages 251-271 in J. L. Vial, editor. Evolutionary biology of the anurans. Contemporary research on major problems. University of Missouri Press, Columbia Stebbins, R. C. and N. W. Cohen. 1995. A natural history of amphibians. Princeton University Press, Princeton. Stoks, R. and M. A. McPeek. 2003. Predators and life histories shape Lestes damselfly assemblages along a freshwater habitat gradient. Ecology 84:1576-1587. Svanbäck, R. and D. I. Bolnick. 2005. Intraspecific competition affects the strength of individual specialization: an optimal diet theory method. Evolutionary Ecology Research 7:993- 1012. Svanbäck, R. and D. I. Bolnick. 2007. Intraspecific competition drives increased resource use diversity within a natural population. Proceedings of the Royal Society B-Biological Sciences 274:839-844. Svanbäck, R. and P. Eklöv. 2002. Effects of habitat and food resources on morphology and ontogenetic growth trajectories in perch. Oecologia 131:61-70.

121

Svanbäck, R., P. Eklöv, R. Fransson, and K. Holmgren. 2008. Intraspecific competition drives multiple species resource polymorphism in fish communities. Oikos 117:114-124. Svanbäck, R. and L. Persson. 2004. Individual diet specialization, niche width and population dynamics: implications for trophic polymorphisms. Journal of Animal Ecology 73:973- 982. SYSTAT. 2007. MYSTAT. Systat Software, Inc., Chicago, IL. Takimoto, G., T. Iwata, and M. Murakami. 2002. Seasonal subsidy stabilizes food web dynamics: Balance in a heterogeneous landscape. Ecological Research 17:433-439. Takimoto, G. and D. Post. 2012. Environmental determinants of food-chain length: a meta- analysis. Ecological Research:1-7. Takimoto, G., D. A. Spiller, and D. M. Post. 2008. Ecosystem size, but not disturbance determines food-chain length on islands of the Bahamas. Ecology 89:3001-3007. Tavares-Cromar, A. F. and D. D. Williams. 1997. Dietary overlap and coexistence of chironomid larvae in a detritus-based stream. Hydrobiologia 354:67-81. Tilman, D., C. L. Lehman, and K. T. Thomson. 1997. Plant diversity and ecosystem productivity: Theoretical considerations. Proceedings of the National Academy of Sciences of the United States of America 94:1857-1861. Tilman, D., P. B. Reich, J. Knops, D. Wedin, T. Mielke, and C. Lehman. 2001. Diversity and Productivity in a Long-Term Grassland Experiment. Science 294:843-845. Toft, C. A. 1985. Resource partitioning in amphibians and reptiles. Copeia:1-21. Toloza, E. M. and J. M. Diamond. 1990. Ontogenic development of transporter regulation in bullfrog intestine. American Journal of Physiology 258:G770-G773. Townsend, C. R., R. M. Thompson, A. R. McIntosh, C. Kilroy, E. Edwards, and M. R. Scarsbrook. 1998. Disturbance, resource supply, and food-web architecture in streams. Ecology Letters 1:200-209. Turner, A. M. and S. L. Montgomery. 2009. Hydroperiod, predators and the distribution of physid snails across the freshwater habitat gradient. Freshwater Biology 54:1189-1201. Turner, T. F., M. L. Collyer, and T. J. Krabbenhoft. 2010. A general hypothesis-testing framework for stable isotope ratios in ecological studies. Ecology 91:2227-2233. Unrine, J. M., W. A. Hopkins, C. S. Romanek, and B. P. Jackson. 2007. Bioaccumulation of trace elements in omnivorous amphibian larvae: implications for amphibian health and contaminant transport. Environmental Pollution 149:182-192. Urban, M. C. 2004. Disturbance heterogeneity determines freshwater metacommunity structure. Ecology 85:2971-2978. Van Valen, L. 1965. Morphological variation and width of ecological niche. American Naturalist 99:377-390. Vander Zanden, H. B., K. A. Bjorndal, P. W. Inglett, and A. B. Bolten. 2012. Marine-derived nutrients from green turtle nests subsidize terrestrial beach ecosystems. Biotropica 44:294-301.

122

Vander Zanden, M. J., G. Cabana, and J. B. Rasmussen. 1997. Comparing trophic position of freshwater fish calculated using stable nitrogen isotope ratios (delta N-15) and literature dietary data. Canadian Journal of Fisheries and Aquatic Sciences 54:1142-1158. Vander Zanden, M. J. and W. W. Fetzer. 2007. Global patterns of aquatic food chain length. Oikos 116:1378-1388. Vander Zanden, M. J. and C. Gratton. 2011. Blowin' in the wind: reciprocal airborne carbon fluxes between lakes and land. Canadian Journal of Fisheries and Aquatic Sciences 68:170-182. Vander Zanden, M. J. and J. B. Rasmussen. 1996. A trophic position model of pelagic food webs: impact on contaminant bioaccumulation in lake trout. Ecological Monographs 66:451-477. Vander Zanden, M. J. and J. B. Rasmussen. 1999. Primary consumer delta C-13 and delta N-15 and the trophic position of aquatic consumers. Ecology 80:1395-1404. Vander Zanden, M. J. and J. B. Rasmussen. 2001. Variation in delta N-15 and delta C-13 trophic fractionation: Implications for aquatic food web studies. Limnology and Oceanography 46:2061-2066. Vander Zanden, M. J., B. J. Shuter, N. Lester, and J. B. Rasmussen. 1999. Patterns of food chain length in lakes: A stable isotope study. American Naturalist 154:406-416. Vander Zanden, M. J., B. J. Shuter, N. P. Lester, and J. B. Rasmussen. 2000. Within- and among-population variation in the trophic position of a pelagic predator, lake trout (Salvelinus namaycush). Canadian Journal of Fisheries and Aquatic Sciences 57:725-731. Vanderklift, M. A. and S. Ponsard. 2003. Sources of variation in consumer-diet delta N-15 enrichment: a meta-analysis. Oecologia 136:169-182. Vindenes, Y., S. Engen, and B.-E. Saether. 2008. Individual heterogeneity in vital parameters and demographic stochasticity. American Naturalist 171 455-467. Walters, A. W. and D. M. Post. 2008. An experimental disturbance alters fish size structure but not food chain length in streams. Ecology 89:3261-3267. Warren, P. H. 1989. Spatial and temporal variation in the structure of a freshwater food web. Oikos 55:299-311. Wellborn, G. A., D. K. Skelly, and E. E. Werner. 1996. Mechanisms creating community structure across a freshwater habitat gradient. Annual Review of Ecology and Systematics 27:337-363. 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. Werner, E. E. and M. A. McPeek. 1994. Direct and indirect effects of predators on two anuran species along an environmental gradient. Ecology 75:1368-1382. Werner, E. E., D. K. Skelly, R. A. Relyea, and K. L. Yurewicz. 2007. Amphibian species richness across environmental gradients. Oikos 116:1697-1712. Wesner, J. S. 2010. Seasonal variation in the trophic structure of a spatial prey subsidy linking aquatic and terrestrial food webs: adult aquatic insects. Oikos 119:170-178.

123

Whiles, M. R. and B. S. Goldowitz. 2001. Hydrologic influences on insect emergence production from central Platte River wetlands. Ecological Applications 11:1829-1842. Wickramasinghe, D. D., K. L. Oseen, and R. J. Wassersug. 2007. Ontogenetic changes in diet and intestinal morphology in semi-terrestrial tadpoles of Nannophrys ceylonensis (Dicroglossidae). Copeia 2007:1012-1018. Wiggins, G. B., R. J. Mackay, and I. M. Smith. 1980. Evolutionary and ecological strategies of animals in annual temporary pools. Archiv fur Hydrobiologie 58:97–206. Wilbur, H. M. 1997. Experimental ecology of food webs: complex systems in temporary ponds. Ecology 78:2279-2302. Williams, D. D. 1997. Temporary ponds and their invertebrate communities. Aquatic Conservation-Marine and Freshwater Ecosystems 7:105-117. Williams, D. D. 2006. The Biology of Temporary Waters. Oxford University Press, Oxford. Williams, D. D., C. M. Febria, and T. A. Schriever. 2009. Structure and mechanics of intermittent wetland communities: bacteria to anacondas Pages 17-55 in J. R. Herrara, editor. International Wetlands: Ecology, Conservation and Restoration. Nova Science Publishers, Inc., Hauppauge. Winemiller, K. O. 1990. Spatial and temporal variation in tropical fish trophic networks. Ecological Monographs 60:331-367. Winemiller, K. O. 2006. Interplay between scale, resolution, life history and food web properties. Page 277 in N. Rooney, K. S. McCann, and D. L. G. Noakes, editors. Energetics to Ecosystems: The Dynamics and Structure of Ecological Systems. Springer, Dordrecht. Winemiller, K. O., S. Akin, and S. C. Zeug. 2007. Production sources and food web structure of a temperate tidal estuary: integration of dietary and stable isotope data. Marine Ecology- Progress Series 343:63-76. Winemiller, K. O. and G. A. Polis. 1996. Food webs: What can they tell us about the world? . Pages 1-22 in G. A. Polis and K. O. Winemiller, editors. Food Webs: Integration of Patterns and Dynamics. Chapman and Hall, New York. Woo, K. J., K. H. Elliott, M. Davidson, A. J. Gaston, and G. K. Davoren. 2008. Individual specialization in diet by a generalist marine predator reflects specialization in foraging behaviour. Journal of Animal Ecology 77:1082-1091. Zerba, K. E. and J. P. Collins. 1992. Spatial heterogeneity and individual variation in diet of an aquatic top predator. Ecology 73:268-279.

124

8 Appendices

Appendix A Sample sizes per pond for A. laterale.

Pond name Gut content Stable Isotope Hydroperiod sample size sample size category

Short1 7 4 Short Short2 12 12

Inter1 1 1

Inter2 9 6 Intermediate

Inter3 6 6

Long1 1 3 Long Long2 2 2

125

Appendix B Mean stable isotope values and standard deviation (SD) for basal resources, macroinvertebrates, and amphibian larvae collected from ponds ranging in hydroperiod.

SD SD Sample Taxon 13C 13C 15N 15N size

A. laterale -30.75 1.43 2.84 0.82 34

L. sylvaticus -33.29 1.48 0.70 0.86 62

Gammaridae (scuds) -30.30 1.27 2.27 0.80 6

Anax junius nymphs -32.92 1.31 1.53 1.14 3

Lestid nymphs -33.27 1.14 2.04 1.36 4

Libellulid nymphs -31.17 1.25 0.93 1.08 5

Dytiscid larvae -31.93 0.67 0.66 1.47 7

Hirudinea leeches -29.46 4.77 3.53 1.90 3

Culcid larvae -28.42 1.95 -2.21 0.89 2

Limnephilid larvae -32.03 1.19 -0.98 1.13 4

Lymnaeid snails -33.38 2.10 -0.42 1.07 7

Physid snails -30.04 1.23 -0.31 1.30 2

Planorbid snails -31.41 2.41 -1.49 1.33 2

Sphaeriid clams -31.15 2.31 0.16 1.39 10 detritus -27.92 1.30 -0.17 1.37 33 algae -35.56 2.94 0.00 1.32 26 aquatic plants -30.09 1.63 2.46 1.30 12

FBOM -32.73 1.73 1.11 1.01 33 seston -36.56 2.68 0.46 1.16 28

126

127

Appendix D Relative importance (IRI) of diet items (mean and SE) found in the stomachs of A. laterale (n = 34). Protozoans included primarily ciliates.

128

Short hydroperiod Intermediate hydroperiod Long hydroperiod

Species TPGC TPSI TPGC TPSI TPGC TPSI

L. sylvaticus* NP NP 2.25 (0.29) 2.01 (0.23) 2.14 (0.16) 1.98 (0.21)

A. laterale 2.79 (0.32) 3.02 (0.25) 2.98 (0.20) 2.95 (0.32) 2.96 (0.17) 2.96 (0.29)

*includes tadpoles and metamorphs

129

Appendix F Occurrence of breeding amphibians in ponds sampled during 2008 and 2009. If present in only one year, year present is in parenthesis. Short3 and Inter4 were only sampled in 2009.

Species Short1 Short2 Short3 Inter1 Inter2 Inter3 Inter4 Long1 Long2

A. laterale X (2008) X X (2009) X X X X X X

A. maculatum X X X X X X

N. viridescens X (2008) X

H. versicolor X X X X X

P. crucifer X (2008) X X (2009) X (2008) X X X X X

L. sylvaticus X (2009) X X X X X

L. clamitans X (2008) X (2008) X X

130

Appendix G Invertebrate community composition in ponds sampled in 2008. Presence is represented by 1. Inter is abbreviated for intermediate hydroperiod ponds. Predatory families are in bold font.

Phylum (subphylum), Class (subclass) Order (suborder), family Genus species Short1 Short2 Inter1 Inter2 Inter3 Long1 Long2

Annelida, (Oligochaeta) 1 1 1 1 1 1 1

Annelida, Hirudinea Pharyngobdellida, Erpobdellidae Macrobdella decora 1 1 1

Arthropoda (Crustacea), Branchiopoda Anostraca (or Notostraca) 1 1 1

Arthropoda (Crustacea), Branchiopoda Cladocera 1 1 1 1

Arthropoda (Crustacea), Malocostraca Amphipoda, Gammaridae Gammarus 1 1 1 1 1 1

hyalella 1 1 1 1 1

Isopoda, Asellidae Asellus sp. 1 1

Arthropoda (Crustacea), Maxillopoda (Copepoda) 1 1 1 1 1 1 1

Arthropoda, Arachnida Araneae, Pisauridae 1 1 1 1

(Acari) Hydrachnida 1 1 1 1 1 1 1

Arthropoda, Crustacea (Ostracoda) 1 1 1 1 1 1

Arthropoda, Insecta Coleoptera 1 1 1 1 1 1 1

Arthropoda, Insecta Coleoptera, Carabidae 1

Coleoptera, Chrysomelidae 1

Coleoptera, Dryopidae 1 1

Coleoptera, Dytiscidae 1 1 1 1 1 1 1

131

Coleoptera, Dytiscidae Desmopachria 1

Coleoptera, Dytiscidae Acilius sp. 1 1 1 1

Coleoptera, Dytiscidae Dytiscus sp. 1

Coleoptera, Dytiscidae Hygrotus 1

Coleoptera, Dytiscidae Hydrovatus pustulatus 1

Coleoptera, Gyrinidae 1

Coleoptera, Haliplidae 1 1 1 1

Coleoptera, Hydraenidae 1

Coleoptera,Hydrophilidae Hydrochara sp. 1

Coleoptera, Scirtidae 1 1 1 1

Diptera, Ceratopogonidae 1

Diptera, Chaoboridae 1

Diptera, Chironomidae 1 1 1 1 1 1 1

Diptera, Culicidae 1 1 1 1 1 1 1

Diptera, Stratiomyidae 1 1 1

Ephemeroptera, Baetidae 1 1 1

Hemiptera, Belostomatidae Lethocerus sp. 1

Hemiptera, Belostomatidae Belostoma sp. 1 1 1

Hemiptera, Corixidae 1 1 1 1 1

132

Hemiptera, Gerridae 1 1 1

Hemiptera, Mesoveliidae 1 1

Hemiptera, Nepidae Ranatra fusca 1

unknown 1 1

Hemiptera, Notonectidae notonecta sp. 1 1 1 1 1

Hemiptera, Veliidae Microvelia sp. 1 1 1 1 1

Megaloptera 1 1 1 1 1

Odonata (Zygoptera), Lestidae 1 1 1 1

Odonata (Zygoptera), Coenagrionidae 1 1 1 1

Coenagrion or Enallagma sp. 1

Nahalennia sp. 1

Odonata (Anisoptera), Libullulidae 1 1 1 1

Odonata (Anisoptera), Aeshnidae 1 1 1 1

Anax junius 1 1 1 1

Trichoptera, Limnephilidae Limnephilus sp. 1 1 1 1 1

Nemotaulius sp. 1

Mollusca, Bivalvia (Pelecypoda) Veneroida, Sphaeriidae 1 1 1 1 1 1 1

Mollusca, Gastropoda Basommatophora, Planorbidae 1 1 1 1 1

Mollusca, Gastropoda Lymnaeidae 1 1 1 1 1 1

133

Mollusca, Gastropoda Physidae Physa elliptica 1 1 1 1 1

134

Appendix H Invertebrate community composition in ponds sampled in 2009. Presence marked by 1. Inter is abbreviated for intermediate hydroperiod ponds. Predatory families are in bold font.

Phylum (subphylum), Class Genus, Short Short Short Inter Inter Inter Inter Long Long (subclass) Order (suborder), family species 1 2 3 1 2 3 4 1 2

Annelida, (Oligochaeta) 1 1 1

Pharyngobdellida, Annelida, Hirudinea Erpobdellidae Macrobdella decora 1 1

Arthropoda (Crustacea), Branchiopoda Anostraca (or Notostraca) 1 1 1 1 1 1

Cladocera 1 1 1 1 1 1 1 1

Arthropoda (Crustacea), Malocostraca Amphipoda, Gammaridae 1 1 1 1 1 1 1

Isopoda, Asellidae Asellus 1 1

Isopoda, Asellidae Lirceus 1

Arthropoda (Crustacea), Maxillopoda (Copepoda) 1 1 1 1 1 1 1 1

Arthropoda (Crustacea), Ostracoda Ostracoda 1 1 1 1 1 1 1

Arthropoda, Arachnida Araneae 1 1 1 1 1

(Acari) Hydrachnida 1 1 1 1 1 1

Arthropoda, Collembola Poduromorpha 1 1 1 1

Arthropoda, Crustacea unknown 1 1 1

Isopoda, Asellidae 1 `

135

Arthropoda, Insecta Coleoptera 1 1 1 1 1 1 1

Coleoptera, Coccinellidae Chilocorinae 1 1

Coleoptera, Chrysomelidae 1

Coleoptera, Curculionidae 1

Coleoptera, Carbaidae 1 1 1 1 1 1

Coleoptera, Dytiscidae 1 1 1 1 1 1 1 1 1

Acilius 1 1 1 1 1

Cybistrinae 1

Desmopachria 1 1

Dytiscus 1

Dytiscus harrisii 1 1 1

Halipidae 1 1 1

Hydovatus 1

Coleoptera, Elmidae 1 1

Coleoptera, Gyrinidae 1 1 1 1

Dineutus 1 1 1 1 1 1

Gyrinus 1 1

Coleoptera, Haliplidae peltodytes 1 1 1 1

Haliplus 1 1 1

136

Coleoptera, Hydraenidae 1

Coleoptera, Hydrophilidae 1 1 1 1 1 1 1

Hydrochus 1 1 1

Coleoptera, Noteridae 1

Hemiptera 1 1

Hemiptera, Aphididae

Hemiptera, Corixidae 1 1 1 1 1 1 1

Hemiptera, Corixidae Corixini 1 1 1

Hemiptera, Gerridae 1 1 1

Neogerris 1 1 1 1 1

Hemiptera (Heteroptera) Reduviidae 1 1

Hemiptera, Nepidae 1

Ranatra 1 1 1

Nepa 1 1 1 1

Hemiptera, Notonectidae 1 1 1 1

Notonecta 1 1 1 1 1 1 1

Hemiptera, Veliidae 1 1 1 1 1

Hemiptera, Mesoveliidae 1 1 1 1

Hemiptera, Pleidae 1 1

137

Hemiptera, Belostomatidae 1 1 1 1 1

Belostoma 1 1 1 1 1

Hemiptera, Saldidae 1

Hemiptera, Hydrometridae 1

Homoptera 1 1

Megaloptera, Corydalidae 1 1 1

Nigronia 1 1

Chauliodes 1 1 1 1

Trichoptera 1 1 1 1

Trichoptera, Leptoceridae 1 1 1 1

Trichoptera, Limnephilidae 1 1 1 1

Nemotaulius 1 1 1 1 1

Trichoptera, Phryganeidae

Trichoptera, Phryganeidae Fabria 1 1

Ephemeroptera 1 1 1

Ephemeroptera, Caenidae 1

Ephemeroptera, Baetidae 1 1 1 1

Ephemeroptera Potamanthidae Anthopotamus 1

Diptera 1 1 1 1 1 1

138

Diptera, Ceratopogonidae 1 1 1 1

Diptera, Chironomidae 1 1 1 1 1 1 1

Diptera, Chaoboridae Chaoborus 1 1 1 1 1 1

Diptera, Culicidae 1 1 1 1 1 1 1 1 1

Diptera, Dixidae Dixella 1 1 1 1 1

Diptera, Stratiomyidae 1 1

Nemotelus 1 1

odontomyia 1 1

stratiomys 1 1

Diptera, Tipulidae Tipulinae 1 1

Odonata (Anisoptera) 1 1 1 1

Odonata (Anisoptera), Libullulidae 1 1 1 1 1 1

Odonata (Anisoptera), Aeshnidae 1 1 1

Odonata (Anisoptera), Aeshnidae Anax junius 1 1 1 1

Odonata (Zygoptera) 1 1 1 1 1

Odonata (Zygoptera), Lestidae 1 1 1 1 1

Lestes 1 1 1 1 1

Odonata (Zygoptera), Coenagrionidae 1 1 1

Mollusca, Bivalvia (Pelecypoda) Veneroida, Sphaeriidae 1 1 1 1 1 1 1

139

Mollusca, Gastropoda unknown 1 1 1 1 1

Basommatophora, Lymnaeidae 1 1 1

Basommatophora, Planorbidae 1 1 1 1 1 1 1 1

Helisoma 1 1 1 1 1 1

Basommatophora, Physidae 1 1 1

Aplexa 1 1 1 1 1

140

Appendix I Macroinvertebrate length–biomass regressions. Macroinvertebrate taxa list with regression equations and references used to estimate biomass. Biomass was estimated as: biomass DM = a(length)b, where DM is dry mass (mg), length is total body length (mm), and a and b are constants.

Taxon a b Reference

EPHEMEROPTERA

Metretopodidae 0.014 2.49 Sabo, Bastow and Power 2002 equation for Ephemeroptera

Baetidae 0.0053 2.875 Benke et al. 1999

Caenidae 0.0054 2.772 Benke et al. 1999

TRICHOPTERA

Leptoceridae 0.01 2.9 Sabo, Bastow and Power 2002 equation for Trichoptera

Hydroptilidae 0.0062 2.816 McNeely, Finlay, and Power. 2007

Limnephilidae 0.004 2.933 Benke et al. 1999

Phryganeidae 0.0054 2.811 Benke et al. 1999

DIPTERA

Ceratopogonidae 0.0025 2.469 Benke et al. 1999

Chironomidae 0.0018 2.617 Benke et al. 1999

Diptera 0.04 2.26 Sabo, Bastow and Power 2002

Ephydridae 0.0054 2.546 Benke et al. 1999 equation for Empididae

141

Sciaridae 0.0042 2.091 Benke et al. 1999

Simuliidae 0.002 3.011 Benke et al. 1999

Tabanidae 0.005 2.591 Benke et al. 1999

Tipulidae 0.0029 2.681 Benke et al. 1999

Cucilidae 0.032 2.038 Sabo, Bastow and Power 2002

Muscidae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Scathophagidae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Sarcophaginae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Chloropidae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Phoridae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Dixidae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Chaoboridae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Psychodidae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Sciomyzidae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera

Dolichopodidae 0.04 2.26 Sabo, Bastow and Power 2002 equation for Diptera unknown diptera 2.158 -3.374 Sample et al. 1993 equation used for Diptera Brachycera

ODONATA

Lestidae 0.14 2.27 Sabo, Bastow and Power 2002 equation for Odonata

142

Coenagrionidae 0.001 2.672 Sabo, Bastow and Power 2002

Libellulidae 0.0076 2.809 Benke et al. 1999

HYMENOPTERA

Scelionidae 0.56 1.56 Sabo, Bastow and Power 2002 equation for Hymenoptera

Ichneumonidae 0.56 1.56 Sabo, Bastow and Power 2002 equation for Hymenoptera

Eulophidae 0.56 1.56 Sabo, Bastow and Power 2002 equation for Hymenoptera

Mymaridae 0.56 1.56 Sabo, Bastow and Power 2002 equation for Hymenoptera

Braconidae 0.56 1.56 Sabo, Bastow and Power 2002 equation for Hymenoptera

HEMIPTERA

Cicadellidae 0.079 2.229 Sabo, Bastow and Power 2002

Benke, AC, Huryn AD, Smock LA, Wallace JB, 1999. Length-mass relationships for freshwater macroinvertebrates in north america with particular reference to the southeastern united states. Journal of the North American Benthological Society 18, 308-343.

McNeely, C, Finlay JC, Power ME, 2007. Grazer traits, competition, and carbon sources to a headwater-stream food web. Ecology 88, 391-401.

Sabo, JL, Bastow JL, Power ME, 2002. Length-mass relationships for adult aquatic and terrestrial invertebrates in a california watershed. Journal of the North American Benthological Society 21, 336-343.

143

Appendix J Emergent aquatic insects length, mass, and numbers collected.

Length DM mean (mg) DM mean (g) n measured sd DM (g) n collected

Order Family mean ± sd

Diptera Ceratopogonidae 2.10 ± 0.84 0.02 ± 0.02 0.000 18 0.000 21

Diptera Chaoboridae 6.34 ± 0.60 2.63 ± 0.54 0.003 16 0.001 16

Diptera Chironomidae 4.08 ± 1.58 0.09 ± 0.09 0.000 259 0.000 711

Diptera Culcidae 4.94 ± 0.95 0.86 ± 0.40 0.001 26 0.000 71

Diptera Dolichopodidae 3.86 ± 1.54 1.03 ± 1.12 0.001 12 0.001 51

Diptera Ephydridae 3.70 ± 0.89 0.17 ± 0.09 0.000 11 0.000 35

Diptera Muscidae 6.11 ± 1.17 2.51 ± 1.01 0.003 9 0.001 9

Diptera Phoridae 1.90 ± 0.26 0.17 ± 0.05 0.000 2 0.000 2

Diptera Psychodidae 1.47 ± 0.19 0.10 ± 0.03 0.000 46 0.000 68

Diptera Sarcophaginae 3.72 ± 0.78 0.80 ± 0.37 0.001 2 0.000 2

Diptera Scathophagidae 3.37 ± 0.03 0.62 ± 0.01 0.001 2 0.000 3

Diptera Sciaridae 2.98 0.04 0.000 1 . 1

Diptera Sciomyzidae 3.72 ± 1.41 0.92 ± 0.82 0.001 7 0.001 12

Diptera Simulidae 5.13 ± 1.34 0.30 ± 0.22 0.000 2 0.000 2

144

Diptera Tabanidae 9.09 ± 2.02 1.64 ± 1.04 0.002 4 0.001 4

Diptera Tipulidae 2.55 ± 1.23 0.05 ± 0.05 0.000 2 0.000 4

Diptera unknown 1.55 ± 0.14 0.52 ± 0.17 0.001 5 0.000 12

Ephemeroptera Baetidae 8.44 2.44 0.002 1 . 3

Ephemeroptera Caenidae 4.97 ± 3.38 0.72 ± 0.91 0.001 2 0.001 2

Ephemeroptera Metretopodidae 9.27 3.58 0.004 1 . 2

Hemiptera Cicadellidae 4.89 ± 2.37 3.49 ± 3.77 0.003 7 0.004 7

Hymenoptera Braconidae 2.48 ± 0.35 2.33 ± 0.50 0.002 7 0.000 7

Hymenoptera Ichneumonidae 5.17 7.26 0.007 1 . 1

Hymenoptera Mymaridae 1.07 0.62 0.001 1 . 1

Hymenoptera Scelionidae 1.28 ± 0.31 0.83 ± 0.31 0.001 2 0.000 2

Odonata Coenagriiondae 27.65 7.12 0.007 1 . 1

Odonata Lestidae 35.42 ± 0.58 460.29 ± 17.02 0.460 3 0.017 4

Odontata Libellulidae 30.43 ± 4.10 114.11 ± 42.32 0.114 2 0.042 2

Trichoptera Hydroptilidae 6.13 1.02 0.001 1 . 1

Trichoptera Limnephilidae 11.45 ± 2.25 5.57 ± 2.24 0.006 26 0.002 26 biomass unaccounted for:

145

Diptera Dixidae 2

Coleoptera Staphylinidae 1

Hemiptera Aphidoidea, aphids 1

Orthoptera Acrididae 2

Psocodea Psocoptera 1

Coleoptera unknown 1

Homoptera unknown 1

146

Species Predictive equation P n SVL range

A. maculatum g AFDM = -0.058 + 0.006*SVL 0.0448 * 8 21.97-31.00

A. laterale g AFDM = -0.301 + 0.014*SVL 0.000000195 *** 22 23.46-36.34

L. sylvaticus g AFDM = -0.116 + 0.010*SVL 0.0000000000363*** 36 13.04-21.91

N. viridescens g AFDM = -0.157 + 0.010*SVL 0.00516 ** 6 18.13-35.2

P. crucifer g AFDM = -0.040 + 0.005 *SVL 0.00256 ** 23 9.97-12.93

H. versicolor g AFDM = 0.001 + 0.060*SVL 0.811 NS 13 15.11-18.96

Ambystoma g AFDM = -0.214 + 0.012 *SVL 0.00000000023 *** 30 21.97-36.34

147

148

149

Appendix N Association of ponds based on a CA of insect family presence-absence in 2008.

150

Appendix O Association of ponds based on a CA of amphibian species presence-absence in 2008.

151

Appendix P Association of ponds based on a correspondence analysis (CA) of invertebrate family presence-absence in 2009.

152

Appendix Q Correspondence analysis (CA) of amphibian species presence-absence in 2009.

153

Order (suborder), family invert code

(Acari) Hydrachnida acar

Amphipoda, amga

Annelida, (Oligochaeta) oli

Anostraca ano

Araneae pisa

Basommatophora, Planorbidae plan

Cladocera clad

Coleoptera, Carabaidae cara

Coleoptera, Chrysomelidae chry (NA)

Coleoptera, Curculionidae NA (curc)

Coleoptera, Dryopidae dryo (NA)

Coleoptera, Dytiscidae dytis

Coleoptera, Elmidae NA (elmi)

Coleoptera, Gyrinidae gyri

Coleoptera, Haliplidae hali

Coleoptera, Hydraenidae hydra (hydr)

Coleoptera, Noteridae NA (note)

Coleoptera, Scirtidae scir (NA)

Coleoptera, Hydrophiloidea, Hydrophilidae/Hydrochidae hych copepods cope

154

Collembola NA (coll)

Diptera, Ceratopogonidae cera

Diptera, Chaoboridae chao

Diptera, Chironomidae chir

Diptera, Culicidae culi

Diptera, Dixidae NA (didix)

Diptera, Stratiomyidae stra

Diptera, Tipulidae NA (tipu)

Ephemeroptera NA (ephe)

Ephemeroptera, Baetidae baet

Ephemeroptera, Caenidae NA (caen)

Ephemeroptera, Potamanthidae NA (pota)

Hemiptera, Belostomatidae belo

Hemiptera, Corixidae, Corixinae cori

Hemiptera, Gerridae gerr

Hemiptera, Mesoveliidae meso

Hemiptera, Nepidae rafu (nepi)

Hemiptera, Notonectidae noto

Hemiptera, Heteroptera, Reduviidae NA (redu)

Hemiptera, Veliidae micr

Hemiptera, Pleidae NA (plei)

Hemiptera, Saldidae NA (sald)

Isopoda, Asellidae asel

Basommatophora, Lymnaeidae lymn

Megaloptera, Corydalidae mega

155

Odonata (Anisoptera), Aeshnidae aesh

Odonata (Zygoptera), Coenagrionidae coen

Odonata (Anisoptera), Libullulidae libu

Odonata (Zygoptera), Lestidae lest

Ostracoda ostr

Annelida, Hirudinea made

Basommatophora, Physidae phys

Trichoptera, Limnephilidae limn

Trichoptera, Leptoceridae NA (lept)

Trichoptera, Phryganeidae NA (phry)

Veneroida, Sphaeriidae spha