Nutritional Ecology of Aphaenogaster Ants in Response to Climate Change Katie A

University of Vermont ScholarWorks @ UVM

Graduate College Dissertations and Theses Dissertations and Theses

2018 Nutritional Ecology of Aphaenogaster Ants in Response to Climate Change Katie A. Miller University of Vermont

Follow this and additional works at: https://scholarworks.uvm.edu/graddis Part of the Ecology and Evolutionary Biology Commons

Recommended Citation Miller, Katie A., "Nutritional Ecology of Aphaenogaster Ants in Response to Climate Change" (2018). Graduate College Dissertations and Theses. 899. https://scholarworks.uvm.edu/graddis/899

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

NUTRITIONAL ECOLOGY OF APHAENOGASTER ANTS IN RESPONSE TO CLIMATE CHANGE

A Thesis Presented

Katie Ann Miller

The Faculty of the Graduate College

The University of Vermont

In Partial Fulfillment of the Requirements for the Degree of Master of Science Specializing in Biology

May, 2018

Defense Date: March 21st, 2018 Thesis Examination Committee:

Sara Helms Cahan, Ph.D., Advisor Kimberly Wallin, Ph.D., Chairperson Nicholas Gotelli, Ph.D. Jason Stockwell, Ph.D. Cynthia J. Forehand, Ph.D., Dean of the Graduate College

ABSTRACT

Climate change is predicted to impact organismal nutritional ecology. Increased temperatures can directly accelerate physiological rate processes, which in turn, impact nutritional requirements. Climate change can also impact organisms indirectly by altering the quality and quantity of nutritional resources, creating the potential for nutritional mismatch between what nutrients are available in the environment and what organisms require. Investigation of organismal stoichiometry, particularly the balance of carbon, nitrogen, and phosphorus content of organisms, can help illuminate the extent to which changes in climate may impact organism nutritional ecology. Ants represent an excellent system to examine stoichiometry because they occur across a broad range of environmental conditions and perform important ecosystem services, such as seed dispersal, which may impact ecosystem functioning. In this thesis, I examined how climate variables influence ant stoichiometry across a broad latitudinal gradient in natural populations of three closely-related ant species in the genus Aphaenogaster. In a common garden study, I tested the extent to which such stoichiometric variation was due to plastic or evolved variation. I found significant species-specific differences in how ant stoichiometry responded to climate gradients. The northern species, A. picea contained more C, and less N and P at higher latitudes and elevation, consistent with increased winter lipid storage. In contrast, the more southern species, A. rudis, showed the opposite pattern, which may reflect N and P limitation at southern extremes. Aphaenogaster fulva, whose range is intermediate in latitude and partially overlaps with both congeners, contained more C in environments with more seasonal precipitation. Thus, these species appear to use different nutrient storage strategies in response to the variation in abiotic and trophic conditions across their range. When reared under the same feeding regime and thermal conditions, site-level differences in nitrogen storage between a northern and a southern ant population were retained over time and across years, suggesting that adaptive divergence in elemental composition is at least partially responsible for clinal patterns in the field. To connect latitudinal patterns to temporal changes projected under climate change, I evaluated how increases in temperature impact ant stoichiometry and associated functional traits at the individual and colony level using an experimental field mesocosm experiment at two sites, Harvard Forest (HF) and Duke Forest (DF). I examined how experimental increases in temperature impacted ant body size, colony demography, and nutritional status of two Aphaenogaster ant species. I found that Aphaenogaster ants at the northern site, HF, responded positively to direct increases in temperature, with increases in colony biomass, colony size, total reproductive output, and shifts toward increased nitrogen content with increases in temperature. In contrast, Aphaenogaster ants at the southern site, DF, were generally unaffected by temperature except for a decrease in maximum colony size with increases in temperature. Together, my findings provide evidence that both climate variables and evolutionary history impacts ant stoichiometry, which in turn, may impact ant colony fitness. Examination of the biochemical basis of stoichiometric trait variation is needed to ascertain the role stoichiometry may play in how ant species adapt to changing environmental conditions.

ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Sara Helms Cahan, for her guidance and support throughout the graduate school process. I’d especially like to thank her for helping me to improve my writing and problem-solving skills and for being such a patient and understanding advisor. I would also like to thank my graduate committee, Nicholas

Gotelli, Kimberly Wallin, and Jason Stockwell, for taking the time to provide feedback on my research and help guide me through my graduate career, in addition to providing feedback about my future endeavors. I would like to thank all the high school and undergraduate students who assisted with completing this research, especially Amanda

Meyer, Delaney Smith, Bonnie Kelley, Megna Senthilnathan, Rubén A. García Reyes,

Curtis Provencher, Kenzie Mahoney, Julia Cline, Simone Valcour, Jenna Todero, and

Courtney Hay. I would also like to thank Manisha Patel and Aaron Ellison for assistance with protocol development and completion of the analysis of all phosphorus analyses in this research. In addition to being an extremely huge emotional support during my graduate career, Vanesa Perillo also assisted with completing analysis of soil nutrient composition. I am deeply grateful for her willingness to help me with whatever I needed during this process. I’d especially like to thank Yainna Hernáiz-Hernández (Hall),

Andrew Nguyen, Mike Herrmann, Lucia Orantes, and other past members of the Helms

Cahan lab, for providing emotional and logistical support throughout my graduate career.

I would also like to thank the plethora of graduate students at the University of Vermont, and my “writing-group” group friends (Bailey Bradley, Shannon Smullen, Morgan Paine, and Vanesa Perillo), who provided me with a family away from home. I thank Audrey

Maran and Dr. Shannon Pelini who assisted with my work at Harvard Forest and were a constant source of hilarity, as well as Clint Penick, whose assistance and feedback were vital to the completion of my second chapter and who provided great feedback on my future endeavors. I would also like to thank all members of the “Warm Ant Dimensions” research group. I thank the National Science Foundation (NSF) which helped fund my graduate career through a Graduate Research Fellowship and helped fund my research through a Dimensions of Biodiversity (DEB) grant. I would also like to thank the

University of Vermont for the many grants, both travel and research grants, which helped to maximize my success in graduate school. Finally, I would like to deeply thank my family and friends back in Minnesota, especially my mom (Ann Miller), my dad (Mike

Miller), step-mom (Maggie Miller), my sister (Mary Miller), my brothers (Jacob,

Matthew, and Ben Miller), and my close friends (Jo Jo and David Thomas, and Kristen and Luke Markstrom), for their emotional support and for encouraging me throughout this whole process. You all are amazing!

iii

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS ...... ii

LIST OF TABLES ...... vi

LIST OF FIGURES ...... vii

CHAPTER 1: INTRODUCTION ...... 1

1.1. Background ...... 1

1.2. The Physiology of Organismal Stoichiometry ...... 2

1.3. How Climate Change may impact Organism Stoichiometry and Body Size ...... 5

1.4. Research Objectives and Chapter Outline ...... 8

1.5. Literature Cited ...... 10

CHAPTER 2: GEOGRAPHIC VARIATION IN STOICHIOMETRY OF A SOCIAL INSECT ...... 16

2.1. Abstract ...... 16

2.2. Introduction...... 17

2.3. Materials and Methods ...... 21 2.3.1. Colony and sample collection sites ...... 21 2.3.2. Chemical analyses ...... 22 2.3.3. Climatic predictors ...... 22 2.3.4. Species identification ...... 23 2.3.5. Statistical Analyses ...... 23 2.3.6. Common Garden study ...... 24 2.4.2. Body mass ...... 26

2.4.3. CNP stoichiometry ...... 26 2.4.3. Annual variation and common garden comparison ...... 28

2.5. Discussion ...... 29

2.7. Acknowledgements...... 54

2.8. Literature Cited ...... 54

CHAPTER 3: EFFECTS OF EXPERIMENTAL WARMING ON ANT BODY SIZE, COLONY DEMOGRAPHY, AND NUTRITIONAL STATUS ...... 60

3.1. Abstract ...... 60

3.2. Introduction...... 61

3.3. Materials and Methods ...... 65 3.3.1. Sites, warming chamber construction, and sample collection ...... 65 3.3.2. Trait Measurements ...... 67 3.3.3. Statistical analyses ...... 69

3.4. Results ...... 70

3.5. Discussion ...... 72

3.6. Figures and Tables ...... 80

3.7. Acknowledgements...... 84

3.8. Literature Cited ...... 85

CHAPTER 4: CONCLUSIONS AND FUTURE DIRECTIONS ...... 90

4.2. Literature Cited ...... 95

COMPREHENSIVE BIBLIOGRAPHY ...... 98

APPENDIX A ...... 113

LIST OF TABLES

Table Page

Table 2.2. 1. Table of Hypotheses which may explain variation in ant stoichiometry. GRH stands for Growth Rate Hypothesis. MH stands for Metabolic Hypothesis. SH stands for Storage Hypothesis...... 19

Supplementary Table 2.6. 1. Geographic Variation in Ant Stoichiometry...... 47

Supplementary Table 2.6. 2. Geographic Variation in Ant Stoichiometry – By Species...... 48

Supplementary Table 2.6. 3. Common Garden Statistical Results...... 51

Supplementary Table 2.6. 4. PC loadings for BioClim Variables...... 53

LIST OF FIGURES

Figure Page

Figure 2.6. 1. Sampled locations across eastern U.S. Color gradient represents mean annual temperature (MAT)...... 36

Figure 2.6. 2. Phylogeny of ant colonies collected across the eastern U.S., from 23 sites, using RAD-seq. Colored branches represent different species identified through principle coordinate analysis. Blue branches represent A. picea; green branches represent A. rudis; orange branches represent a hybrid region; and red branches represent A. fulva...... 37

Figure 2.6. 3. Differences in body mass across Aphaenogaster species...... 38

Figure 2.6. 4. Species differences in stoichiometry...... 39

Figure 2.6. 5. Percent Carbon, Nitrogen, and Phosphorus in ant pupae, for each species designation, as a function of PC1, PC2, and PC3...... 40

Figure 2.6. 6. Stoichiometric molar ratios of ant pupae, for each species, as a function of PC1, PC2, and PC3. Colored lines correspond to species identification (A. fulva = pink, A. picea = green, A. rudis = blue)...... 41

Figure 2.6. 7. Elemental composition of ant pupae between timepoints (panels A-C) and between years (panels D-F) for two sites, Neshaminy State Park, PA (NSP) and Yates Mill, NC (YM). Panels A-C show differences in elemental composition within initial field samples versus three months later, under lab conditions. Panels D-F show differences in elemental composition of samples from the field only, between the two collection years of this study...... 42

vii

ratios within initial field samples versus three months later, under lab conditions. Panels D-F show differences in stoichiometric ratios of samples from the field only, between the two collection years of this study...... 43

Figure 3.6. 1. Ant nest box structure ...... 66

Figure 3.6. 2. Colony demography as a function of Temperature - Harvard Forest...... 80

Figure 3.6. 3. Colony demography as a function of Temperature - Duke Forest...... 80

Figure 3.6. 4. Total Protein content in ant pupae. Duke Forest is highlighted in red and Harvard Forest is highlighted in blue...... 81

Figure 3.6. 5. Isotopic composition of ants and soils as a function of temperature. Southern ant (Aphaenogaster rudis) and soil (Soil-DF [Duke Forest]) samples are indicated in red and pink, respectively. Northern ant (Aphaenogaster picea) and soil (Soil-HF [Harvard Forest]) samples are indicated in dark blue and light blue, respectively...... 82

Figure 3.6. 6. Elemental composition and stoichiometry of ants and soils...... 83

Supplementary Figure 2.6. 1. Principle Coordinate Analysis - PCo1~PCo2 ...... 43

Supplementary Figure 2.6. 2. Principle Coordinate Analysis - PCo1~PCo5 ...... 44

Supplementary Figure 2.6. 3. Principle Coordinate Analysis - PCo2~PCo5 ...... 44

Supplementary Figure 2.6. 4. Body mass in ant pupae, for each species designation, as a function of PC1...... 45

Supplementary Figure 2.6. 5. Contour map of PC1...... 45

Supplementary Figure 2.6. 6. Contour map of PC2...... 46

Supplementary Figure 2.6. 7. Contour map of PC3...... 46

viii

CHAPTER 1: INTRODUCTION

1.1. Background

Understanding how nutrient availability and environmental stress may interact to affect organisms is particularly important in light of the increasingly detrimental effects of climate change (Diffenbaugh and Field 2013; Schmitz 2013; Thornton et al. 2014; van de Waal et al. 2010; Cross et al. 2015). Climate change is predicted to increase both global mean temperatures and climate variability and has already had significant detrimental effects on ecosystems (Diffenbaugh and Field 2013; IPCC 2007; Schmitz

2013; Thornton et al. 2014; van de Waal et al. 2010). Climate change can impact organisms directly, via changes in physiological reaction rates (e.g. increased metabolic rates), as well as indirectly through impacts on the abundance and quality of food

(Sardans et al. 2012a, b; Thornton et al. 2014). Thus, shifts in climate are likely to affect the distribution, quality, and quantity of essential nutritional resources for terrestrial consumers, while at the same time changing their nutritional requirements.

Understanding how these shifts will affect the success of organisms is crucial to be able to accurately predict species responses to climate change.

Examination of organismal stoichiometry, particularly the balance of carbon, nitrogen, and phosphorus content, is useful for connecting different levels of biological organization. Elemental stoichiometry influences organism lifestyle and interactions within food webs, and ultimately can affect the structure and function of ecosystems

(Meunier et al. 2017). Stoichiometry can also be strongly influenced by changes in

environmental conditions. Thus, understanding how stoichiometric ratios, along with other important functional traits such as body size, may be impacted by changes in climate may give insight on how climate change may impact the future of ecological communities.

1.2. The Physiology of Organismal Stoichiometry

All organisms need nutrients to perform essential biological functions, such as growth, reproduction, and somatic maintenance. Organisms must balance the demands for these nutrients to perform these functions with the availability of nutrients in their environment. The theory of Biological Stoichiometry, the study of energy and chemical elements in living systems, provides a framework to examine this balance across multiple levels of biological organization (Reiners 1986; Elser et al. 1996; Hessen 1997; Elser et al. 2000). The expression of life history traits such as diapause, growth or reproduction requires the production or use of particular biomolecules, principally proteins, carbohydrates, lipids, and nucleic acids, each which have fixed chemical structures that vary in their elemental composition. Thus, for organisms to produce, store, or use these biomolecules, they need to obtain particular proportions of chemical elements, such as carbon (C), nitrogen (N), and phosphorus (P), from their food (Sterner and Elser 2002).

The relative importance of each element needed depends on the needs of the organism and thus, this may depend on ontogenetic stage (Kay et al. 2006), behavior (Grover et al.

2007), reproductive status (Méndez and Karlsson 2005) and environmental stress (Mas-

Marti et al. 2015).

Physiological processes, such as nutrient acquisition, assimilation, and release, play a significant role in controlling the degree of stoichiometric homeostasis, or balance of multiple chemical elements, within the body over time (Frost et al. 2005).

To maintain stoichiometric homeostasis, organisms can modify their rates of ingestion or selectively feed on food resources of varying nutritional quality (Cease et al. 2016).

For example, many insects such as ants and grasshoppers, have been found to selectively feed to attain a specific intake target for a particular macronutrient (Cease et al. 2016, Cook and Behmer 2010; Cook et al. 2012; Simpson et al. 2004; Raubenheimer and Simpson 1999). When limited by low-quality food, organisms can increase ingestion rates to compensate for nutritional deficiencies, and this has been linked to corresponding shifts in elemental composition (Plath and Boersma 2001). Many ant species have been found to switch food preferences when faced with certain environmental pressures (Cook et al. 2012). When confined to foods of differing nutritional quality, ants exhibit corresponding changes in C:N:P ratios, with higher C:N and C:P ratios when feeding on the carbon-rich diets (Kay et al. 2006; Kay et al. 2012).

In some instances, organisms may over-ingest to obtain limiting nutrients (Pearson et al. 2011). However, ingesting nutrients in excess of requirements may result in severe fitness consequences (Pearson et al. 2011; Raubenheimer et al. 2005; Cook et al. 2010;

Kay et al. 2012). For example, Kay et al. 2012 found that ants restricted to high

Protein:Carbohydrate (P:C) diets exhibited increased storage of uric acid, a nitrogenous waste product that may become toxic at high concentrations. Indeed, these same ant colonies displayed increased mortality on high P:C diets (Kay et al. 2012).

Post-ingestion processes may also be used to maintain balance of elemental composition. Assimilation efficiencies, or the amount of food used for growth and reproduction versus lost through respiration and excretion, may be modified to meet specific elemental demands (Mitra and Flynn 2007; Jochum et al. 2017; Cease et al.

2016). Tropical litter arthropods, for example, display increased assimilation efficiencies with increasing percentage of nitrogen in their food (Jochum et al. 2017).

Furthermore, assimilation efficiencies may differ between ontogenetic stages (Cease et al. 2016). For instance, third instar grasshoppers Schistocerca americana have higher P assimilation efficiency than fifth instar grasshoppers (Cease et al. 2016). When fed high

P diets, in excess of requirements, grasshoppers also exhibit increased excretion of excess P in their frass (Zhang et al. 2014). Thus, organisms may regulate their elemental composition through shifts in the efficiency of digestion (Frost et al. 2005;

Zhang et al. 2014).

The extent to which particular elements are assimilated, stored, or incorporated into new biomass depends on organismal demand for specific life-history traits. For example, a central hypothesis in the theory of Biological Stoichiometry is the Growth

Rate Hypothesis, GRH, which posits that organisms with high growth rates should exhibit high phosphorus content because of increased investment in mRNA, a phosphorus-rich biomolecule that is essential for the synthesis of proteins (Sterner and

Elser 2002; Raven 2012; Elser et al. 2003). Organisms can further regulate their elemental composition through selective storage of particular nutrients for future use

(hereafter called Storage Hypothesis (SH)) (Hahn and Denlinger 2007, King and

MacRae 2015, Bale and Hayward 2010). Elemental composition may also reflect

metabolic constraints. Energy obtained from nutrients is used to fuel metabolic

processes that impact growth, reproduction, and survival. Thus, impacts on metabolic

rates can impact nutrient demand to fuel such processes. Temperature and body size are

two variables that play a large role in determining metabolic rates (Brown et al. 2004,

Allen and Gillooly 2009) and therefore can impact nutrient demand and elemental

composition. For example, increased consumption of carbohydrates, an immediate

source of energy, may be used to fuel higher metabolic rates at higher temperatures

(hereafter called Metabolic Hypothesis) (Rho and Lee 2017). If organismal demands

for nutrients are not met, then severe fitness consequences will result (Leal et al. 2017;

Fagan et al. 2002). Thus, a central challenge for scientists is to understand how

organisms balance demands for nutrients, mediated through life history strategies, in

response to environmental pressures, such as through shifts in temperature.

1.3. How Climate Change may impact Organism Stoichiometry and Body Size

The direct effects of thermal and water stress via climate change affect multiple life history traits, including growth, metabolic rates, behavior, reproduction, and somatic maintenance (Kendrick et al. 2013, Mas-Marti et al. 2015; Lemoine et al. 2013; Brown et al. 2004). Increased food consumption, food processing rates, respiration, and increases in enzyme activity have been linked to increased temperatures (Lemoine and Burkepile

2012; Brown et al. 2004; Dong and Somero 2009). The direct effects of thermal stress on these biological processes may alter organism nutritional requirements, and in turn, change the amount of specific chemical elements that are needed in their diet. As such,

physiological responses to environmental stress may require distinct nutritional resources.

For example, Drosophila ananassae flies reared on high protein diets have higher desiccation and heat shock resistance, whereas flies which were reared on high carbohydrate diets had higher starvation and cold resistance (Sisodia and Singh 2012).

Thus, obtaining particular essential nutrients can help alleviate environmental stress and increase organism fitness (Kendrick et al. 2013).

Climate change can also impact organisms indirectly through decreasing abundance and quality of nutritional resources, which may or may not match shifts in their requirements (Sardans et al. 2012a,b; Thornton et al. 2014). Climate change may affect the total amount of food available by producing offsets in the distributions of predators and their prey, increased species extinctions, and changes in species interactions (Thomas et al. 2004; Parmesan 2006; Beaumont and Hughes 2002). Climate change is expected to shift species distributions because, if possible, organisms are likely to migrate to environments where the climate is more optimal for maximal survival and reproduction (Diamond et al. 2011; Parmesan 2006; Pelini et al. 2011). Some species, however, may not be able to migrate fast enough or not at all and may go extinct, effectively eliminating them as a potential nutritional resource (Diamond et al. 2012;

Parmesan 2006; Pelini et al. 2011). Direct effects of climate change on the quality of nutritional resources may also occur. For example, increasing mean or variability in temperature increased C:N and N:P ratios of plant tissues (Sardans et al. 2012b).

Increases in carbon storage in primary producers can result in decreases in the availability

of nitrogen (N) and phosphorus (P) for higher trophic levels, two elements which are essential for growth and reproduction (Sterner and Elser 2000; Sardans et al. 2012a).

Given these simultaneous shifts in the nutritional inputs and nutritional requirements of organisms under climate change, a nutritional mismatch between what is available in the environment for organisms to consume and what organisms need in terms of nutrition to deal with climate change is possible. Multiple studies have demonstrated that imbalances in elemental composition between consumers and their resources (i.e. stoichiometric mismatches) can have important consequences for organismal performance and behavior, and ultimately can impact ecosystem functioning through shifts in community structure and nutrient cycling (Sterner and Elser 2002; Frainer et al.

2016; Bullejos et al. 2014; Lemoine et al. 2014; Sitters et al. 2015). The degree of stoichiometric mismatch between resources and consumers may be exacerbated due to anthropogenic shifts in nutrient availability and increases in thermal stress due to climate change (Mas-Marti et al. 2015; Hillebrand et al. 2009). Therefore, understanding how both temperature and nutrition may interact to affect whole-organism traits, and ultimately population- and community-level responses, is vital if we are to make accurate predictions of species responses to climate change.

Examination of organism responses across latitudinal gradients can help predict how species may respond to climate change. Life history strategies often vary with latitude (De Frenne et al. 2013, Pelini et al. 2012). Relatively few studies have examined stoichiometric trait variation across latitude, and thus, our understanding of the

mechanisms which may drive such variation is limited to studies at smaller scales (Sun et al. 2013). Examination of stoichiometric trait variation using common garden experiments and analysis of trait variation in natural populations across latitude can be used to partition out the independent plastic effects of the environment versus adaptive effects due to local genetic differentiation among populations. In addition, experiments at different latitudes which are replicated across space and time can allow researchers to isolate independent effects of temperature while controlling for other covarying factors

(De Frenne et al. 2013). Although an increasing number of studies have begun to use these approaches, relatively few studies have simultaneously examined trait variation across latitude in natural populations with common garden approaches and manipulative field experiments (De Frenne et al. 2013). Using these approaches can help unravel how direct and indirect effects of climate change may impact organism stoichiometric trait variation, which ultimately can impact ecosystem functioning.

1.4. Research Objectives and Chapter Outline

I used an ecologically important group of ants from the genus Aphaenogaster to examine how thermal stress affects ant elemental composition and overall ecological success. Specifically, I tested the Growth Rate Hypothesis (GRH), a Metabolic

Hypothesis (MH), and a Storage Hypothesis (SH) to identify mechanisms driving variation in ant elemental composition and stoichiometry across different spatial scales.

To do this, I first examined how natural variation in climatic conditions impacts ant elemental composition and stoichiometry of three Aphaenogaster ant species across the eastern United States (Chapter 2). Using a common-garden approach, I further examined

how variation in ant stoichiometry might be explained by local adaptation to abiotic conditions. For each experiment, I compared changes in ant stoichiometry observed across climate gradients to the predictions. Second, I examined how increases in temperature affect nutrient allocation strategies and performance of Aphaenogaster ants at the colony- and individual-level at different latitudes (Chapter 3). Here, I relate ant elemental composition to other functional traits, namely body size and ant colony demography. Furthermore, I use stable isotopic composition from ants and basal resources (soils) to examine thermally-associated shifts in nutritional inputs. Finally, I review the results from Chapters 2 and 3, pointing out patterns that are consistent across spatial scales and whether such patterns were or were not consistent with my hypotheses.

I then re-examine existing approaches in the field of ecological stoichiometry and climate change research, assess how my research advances the field, and point out future directions. Overall, my work provides a more thorough understanding of how consumers may respond to future climate change.

1.5. Literature Cited

Allen AP, Gillooly JF (2009). Towards an integration of ecological stoichiometry and the metabolic theory of ecology to better understand nutrient cycling. Ecology Letters 12(5):369-384.

Bale JS, Hayward SAL (2010). Insect overwintering in a changing climate. Journal of Experimental Biology 213:980-994.

Beamont LJ, Hughes L (2002). Potential changes in the distributions of latitudinally restricted Australian butterfly species in response to climate change. Global Change Biology 8(10):954-971.

Brown JH, Gillooly JF, Allen AP, Savage VM, West GB (2004). Toward a metabolic theory of ecology. Ecology 85:1771-1789.

Bullejos FJ, Carrillo P, Gorokhova E, Medina-Sanchez JM, Gabrial B, Villar-Argaiz M (2014). Shifts in food quality for herbivorous consumer growth: Multiple golden means in the life history. Ecology 95(5):1272-1284.

Cease AJ, Fay M, Elser JJ, Harrison JF (2016). Dietary phosphate affects food selection, post-injestive phosphorus fate, and performance of a polyphagous herbivore. Journal of Experimental Biology 219:64-72.

Cook SC, Wynalda RA, Gold RE, Behmer ST (2012). Macronutrient regulation in the Rasberry crazy ant (Nylanderia sp. nr. pubens). Insectes Sociaux 59(1):93-100.

Cook SC, Behmer ST (2010). Macronutrient regulation in the tropical terrestrial ant Ectatomma ruidum (Formicidae): A field study in Costa Rica. Biotropica 42(2):135- 139.

Cross WF, Hood JM, Benstead JP, Huryn AD, Nelson D (2015). Interactions between temperature and nutrients across levels of ecological organization. Global Change Biology 21(3):1025-1040.

De Frenne P, Graae BJ, Rodriguez-Sanchez F, Kolb A, Chabrerie O, Decocq G, Kort H, Schrjver A, Diekmann M, Eriksson O, Gruwez R, Hermy M, Lenoir J, Plue J, Coomes DA, Verheyen K (2013). Latitudinal gradients as natural laboratories to infer species’ responses to temperature. Journal of Ecology 101(3):784-795. 10

Diamond SE, Frame AE, Martin RA, Buckley LB (2011). Species’ traits predict phenological responses to climate change in butterflies. Ecology 92(5):1005-1012.

Diamond SE, Sorger DM, Hulcr J, Pelini SL, Del Toro I, Hirsch C, Oberg E, Dunn RR (2012). Who likes it hot? A global analysis of the climatic, ecological, and evolutionary determinants of warming tolerance in ants.

Diffenbaugh NS, Field CB (2013). Changes in ecologically critical terrestrial climate conditions. Science 341(6145):486-492.

Dong Y, Somero GN (2009). Temperature adaptation of cytosolic malate dehydrogenases of limpets (genus Lottia): differences in stability and function due to minor changes in sequence correlate with biogeographic and vertical distributions. Journal of Experimental Biology 212:169-177.

Elser JJ, Dobberfuhl DR, MacKay NA, Schampel JH (1996). Organism size, life history, and N:P stoichiometry. BioScience 46(9):674-684.

Elser JJ, Sterner RW, Gorokhova E, Fagan WF, Markow TA, Cotner JB, Harrison JF, Hobbie SE, Odell GM, Weider LW (2000). Biological stoichiometry from genes to ecosystems. Ecology Letters 32(6):540-550.

Elser JJ, Acharya A, Kyle M, Cotner J, Makino W, Markow T, Watts T, Hobbie S, Fagan W, Schade J, Hood J, Sterner RW (2003). Growth rate-stoichiometry couplings in diverse biota. Ecology Letters 6(10):936-943.

Fagan WF, Siemann E, Mitter C, Denno RF, Huberty AF, Woods HA, Elser JJ (2002). Nitrogen in Insects: Implications for Trophic Complexity and Species Diversification. The American Naturalist 160(6):784-802.

Frainer A, Jabiol J, Gessner MO, Bruder A, Chauvet E, McKie BG (2016). Stoichiometric imbalances between detritus and detritivores are related to shifts in ecosystem functioning. Oikos 125(6):861-871.

Frost PC, Evans-White MA, Finkel ZV, Jensen TC, Matzek V (2005). Are you what you eat? Physiological constraints on organismal stoichiometry in an elementally imbalanced world. Oikos 109(1):18-28.

Grover CD, Kay AD, Monson JA, Marsh TC, Holway DA (2007). Linking nutrition and behavioral dominance: carbohydrate scarcity limits aggression and activity in Argentine ants. Proceedings of the Royal Society B 274:2951-2957.

Hahn DA, Denlinger DL (2007). Meeting the energetic demands of insect diapause: Nutrient storage and utilization. Journal of Insect Physiology 53(8):760-773.

Hessen DO (1997). Stoichiometry in food webs: Lotka revisited. Oikos 79(1):195- 200.

Hillebrand H, Borer ET, Bracken MES, Cardinale BJ, Cebrian J, Cleland EE, Elser JJ, Gruner DS, Harpole WS, Ngai JT, Sandin S, Seabloom EW, Shurin JB, Smith JE, Smith MD (2009). Herbivore metabolism and stoichiometry each constrain herbivory at different organizational scales across ecosystems. Ecology Letters 12(6):516-527.

IPCC (Intergovernmental Panel on Climate Change) (2007). Climate change 2007: the physical science basis. Contribution of Working Group I to the Fourth assessment report of the IPCC. Cambridge University Press, Cambridge, United Kingdom.

Jochum M, Barnes AD, Ott D, Lang B, Klarner B, Farajallah A, Scheu S, Brose U (2017). Decreasing stoichiometric resource quality drives compensatory feeding across trophic levels in tropical litter invertebrate communities. The American Naturalist 190(1):131-143.

Kay AD, Rostampour S, Sterner RW (2006). Ant stoichiometry: elemental homeostasis in stage-structured colonies. Functional Ecology 20(6):1037-1044.

Kay AD, Shik JZ, Van Alst A, Miller KA, Kaspari M (2012). Diet composition does not affect ant colony tempo. Functional Ecology 26(2):317-323.

Kendrick MR, Benstead JP (2013). Temperature and nutrient availability interact to mediate growth and body stoichiometry in a detritivorous stream insect. Freshwater Biology 58(9):1820-1830.

King AM, MacRae TH (2015). Insect heat shock protein during stress and diapause. Annual Review of Entomology 60:59-75.

Leal MC, Seehausen O, Matthews B (2017). The Ecology and Evolution of Stoichiometric Phenotypes. Trends in Ecology & Evolution 32(2):108-117.

Lemoine NP, Burkepile DE (2012). Temperature-induced mismatches between consumption and metabolism reduce consumer fitness. Ecology 93(11):2483-2489.

Lemoine NP, Drews WA, Burkepile DE, Parker JD (2013). Increased temperature alters feeding behavior of a generalist herbivore. Oikos 122(12):1669-1678.

Lemoine NP, Giery ST, Burkepile DE (2014). Differing nutritional constraints of consumers across ecosystems. Oecologia 174(4):1367-1376.

Mas-Marti E, Romani AM, Munoz I (2015). Consequences of warming and resource quality on the stoichiometry and nutrient cycling of a stream shredder. PLoS one 10(3):e0118520.

Mendez M, Karlsson PS (2005). Nutrient stoichiometry in pinguicula vulgaris: nutrient availability, plant size, and reproductive status. Ecology 86(4):982-991.

Meunier C, Boersma M, El-Sabaawi R, Halvorson HM, Herstoff EM, Van de Waal DB, Vogt RJ, and Litchman E (2017). From elements to function: toward unifying ecological stoichiometry and trait-based ecology. Frontiers in Environmental Science 5:art18.

Mitra A, Flynn KJ (2007). Importance of interactions between food quality, quantity, and gut transit time on consumer feeding, growth, and trophic dynamics. The American Naturalist 169(5):632-646.

Parmesan C (2006). Ecological and evolutionary responses to recent climate change. Annual Review of Ecology, Evolution, and Systematics 37(1):637-669.

Pearson REG, Behmer ST, Gruner DS, Denno RF (2011). Effects of diet on performance and nutrient regulation in an omnivours katydid. Ecological Entomology 36(4):471-479.

Pelini SL, Bowles FP, Ellison AM, Gotelli NJ, Sanders NJ, Dunn RR (2011). Heating up the forest: open-top chamber warming manipulation of arthropod communities at Harvard and Duke Forests. Methods in Ecology and Evolution 2(5):534-540. 13

Pelini SL, Diamond SE, MacLean H, Ellison AM, Gotelli NJ, Sanders NJ, Dunn RR (2012). Common garden experiments reveal uncommon responses across temperatures, locations, and species of ants. Ecology and Evolution 2(12):3009-3015.

Plath K, Boersma M (2001). Mineral limitation of zooplankton: stoichiometric constraints and optimal foraging. Ecology 82(5): 1260-1269.

Raubenheimer D, Simpson SJ (1999). Assuaging nutritional complexity: geometrical approach. Proceedings of the Nutrition Society 58(4): 779-789.

Raven JA (2012). Protein turnover and plant RNA and phosphorus requirements in relation to nitrogen fixation. Plant Science 188-189:25-35.

Reiners WA (1986). Complementary Models for Ecosystems. The American Naturalist 127(1):59-73.

Rho MS, Lee KP (2017). Balanced intake of protein and carbohydrate maximizes lifetime reproductive success in the mealworm beetle, Tenebrio molitor (Coleoptera: Tenebrionidae). Journal of Insect Physiology 91-92:93-99.

Sardans J, Rivas-Ubach A, Penuelas J (2012a). The elemental stoichiometry of aquatic and terrestrial ecosystems and its relationships with organismic lifestyle and ecosystem structure and function: a review and perspectives. Biogeochemistry 111(1- 3):1-39.

Sardans J, Rivas-Ubach A, Penuelas J (2012b). The C:N:P stoichiometry of organisms and ecosystems in a changing world: a review and perspectives. Perspectives in Plant Ecology, Evolution, and Systematics 14(1):33-47.

Schmitz (2013). Global climate change and the evolutionary ecology of ecosystem functioning. Annals of the New York Academy of Sciences 110(27):11035-11038.

Simpson SJ, Sibly RM, Lee KP, Behmer ST, Raubenheimer D (2004). Optimal foraging when regulating intake of multiple nutrients. Animal Behavior 68(6):1299- 1311.

Sisodia S, Singh BN (2012). Experimental evidence for nutrition regulated stress resistane in Drosophila ananassae. PLoS one 7(10):e46131. 14

Sitters J, Atkinson CL, Guelzow N, Kelly P, Sullivan L (2015). Spatial stoichiometry: cross-ecosystem material flows and their impact on recipient ecosystems and organisms. Oikos 124(7):920-930.

Sterner RW, Elser JJ (2002). Ecological stoichiometry:the biology of elements form molecules to the biosphere. Princeton University Press.

Sun X, Zhou X, Small GE, Sterner R, Kang HZ, Liu C (2013). Energy storage and C:N:P variation in a holometabolous insect (Curculio davidi Fairmaire) larva across a climate gradient. Journal of Insect Physiology 59(4):408-415.

Thomas CD, Cameron A, Green RE, Bakkenes M, Baeumont LJ, Collingham YC, Erasmus BFN, Ferreira de Siqueira M, Grainger A, Hannah L, Hughes L, Huntley B, van Jaarsveld AS, Midgley GF, Miles L, Ortega-Huerta MA, Peterson AT, Phillips OL, Williams SE (2003). Nature 427:145-148.

Thornton PK, Ericksen PJ, Herrero M, Challinor AJ (2014). Climate variability and vulnerability to climate change: a review. Global Change Biology 20(11):3313-3328. van de Waal DB, Verschoor AM, Verspagen JMH, van Donk E, Huisman J (2010). Climate-driven changes in the ecological stoichiometry of aquatic ecosystems. Frontiers in Ecology and the Environment 8(3):145-152.

Zhang Z, Elser JJ, Cease AJ, Zhang X, Q Yu, Han X, Zhang G (2014). Grasshoppers regulate N:P stoichiometric homeostasis by changing phosphorus contents in their frass. PLoS one 9(8):e103697.

CHAPTER 2: GEOGRAPHIC VARIATION IN STOICHIOMETRY OF A

SOCIAL INSECT

2.1. Abstract

Organismal stoichiometry and elemental composition vary within and between species, with important consequences for ecosystem functioning. However, the proximate and ultimate mechanisms driving this variation, particularly across large climatic gradients, is not well understood. In this study, we investigated spatial patterns of variation in elemental stoichiometry in three closely related, ecologically important ant species in the genus Aphaenogaster whose ranges partially overlap along the eastern seaboard of North America. To identify climatic factors influencing elemental composition, we tested for a relationship between climate variables and composition of

C, N and P and their stoichiometric ratios using data from field-collected worker pupae from 23 sites from Maine to Georgia. To determine whether field variation represented evolved differences or a plastic response to local conditions, colonies from two sites were compared under identical feeding regimes in a common garden experiment. We found significant species-specific differences in how ant stoichiometry responded to climatic gradients, suggesting that these species differ in nutrient storage strategies in response to the local conditions. The northern species, A. picea contained more C, and less N and P at higher latitudes and elevation, consistent with increased winter lipid and carbohydrate storage. In contrast, the more southern species, A. rudis, showed the opposite pattern, which may reflect N and P limitation at southern extremes. Aphaenogaster fulva, whose range is intermediate in latitude and partially overlaps with both congeners, was not

responsive to latitude or elevation; instead, A. fulva contained more C in environments with more seasonal precipitation. Under common garden conditions, lab-reared workers from sites whose N and P compositions differed in the field showed strong, systematic shifts toward increased C:N ratios; however, site-level differences in N were retained, suggesting that adaptive divergence in elemental composition is at least partially responsible for producing clinal patterns in the field. Together, these results further illuminate the role of ecological and evolutionary mechanisms which may impact organism stoichiometry and elemental composition, as we found evidence for adaptive divergence in N content, as well as plastic shifts in elemental composition among

Aphaenogaster ant populations. This study further highlights a need to link stoichiometric traits to functional traits across climatic gradients, both within and between species.

2.2. Introduction

Animal growth, development and reproduction rely on obtaining sufficient amounts of the basic elemental building blocks of biomolecules: carbon, nitrogen, and phosphorus. How organisms invest in different functions in response to environmental and evolutionary pressures determines the stoichiometric ratio of these elements required for optimal performance (Sterner and Elser 2002). A variety of factors can influence elemental composition over an individual's lifetime, including developmental stage (Kay et al. 2006), behavior (Grover et al. 2007), and reproductive status (Méndez and Karlsson

2005). Stoichiometric setpoints can vary considerably across even closely-related taxa

(Hamback et al. 2009, McFeeters and Frost 2011).

Interspecific differences in elemental composition are thought to reflect evolutionary divergence in life history strategies (Elser et al. 2000, Huberty and Denno

2006, Woods et al. 2003, Woods et al. 2004, Jaenike and Markow 2003), morphology

(Gibb et al. 2015, Fagan et al. 2002, Davidson 2005), and resource use (Cross et al.

2003). Three main hypotheses may explain variance in elemental composition within and between species. Each hypothesis represents different life history strategies organisms may use to maximize fitness and further, each hypothesis has different predictions. First, in the Growth Rate Hypothesis (GRH), elemental composition is partially determined by growth rates, where species with higher intrinsic growth rates exhibit increases in P content due to an associated increase in P-rich RNA needed for increased protein synthesis (Elser et al. 2000, Sterner and Elser 2002). Indeed, species with higher intrinsic growth rates tend to have lower C:P and N:P ratios (Elser et al. 2000, Sterner and Elser

2002). Organisms at higher latitudes tend to have higher growth rates because of a shorter growing season (Elser et al. 2000, Sterner and Elser 2002). Thus, under the GRH hypothesis, organisms in locations with a shorter growing season, such as at higher latitudes and elevation, are predicted to have increased N and P storage (Table 2.2.1;

Elser et al. 2000, Sterner and Elser 2002). However, the correlation between growth rates and N and P storage is not always consistent across taxa (Schroder et al. 2015, DeMott and Pape 2005). Deviations from a positive association between N and P storage and growth rates may be due to adaptations to local habitat and environmental conditions. In a second hypothesis, hereafter called the Storage Hypothesis (SH), organism elemental composition may be partially determined by selective storage of specific nutrients to deal with future stressful conditions, such as winter conditions (Hahn and Denlinger 2007).

Under the SH hypothesis, organisms in environments which are more variable should exhibit increases in storage of specific nutrients (Table 2.2.1). In thermally variable environments, for example, organisms may store more carbon in the form of lipids to provide a source of energy throughout winter (Williams et al. 2012). In a third hypothesis, hereafter called the Metabolic Hypothesis (MH), elemental composition may be partially explained by metabolic rates. Increases in temperature has often been associated with increased in metabolic rates and this, in turn, may increase consumption of C which is used to fuel metabolic processes (Rho and Lee 2017). Under the MH hypothesis, organisms would be expected to show decreases in C content with increases in temperature, such as at lower latitudes and elevation (Table 2.2.1).

Table 2.2. 1. Table of Hypotheses which may explain variation in ant stoichiometry. GRH stands for Growth Rate Hypothesis. MH stands for Metabolic Hypothesis. SH stands for Storage Hypothesis.

A promising approach to assess how the interaction between life history and ecological conditions may influence the evolution of elemental composition is to examine patterns of inter- and intraspecific variation in elemental composition across environmental gradients (Leal et al. 2017; El-Sabaawi et al. 2014, Bertram et al. 2008,

Yang et al. 2017). Variation in factors such as temperature, moisture availability, predation and competitive pressures all have the potential to influence the evolution of 19

stoichiometric ratios, both by altering the amount and quality of resource inputs and by imposing divergent selective pressures on relevant aspects of the phenotype. Ranges in the expression of functional traits, such as growth rate, can vary significantly within species, and this has been associated with shifts in elemental composition (El-Sabaawi et al. 2014, Gonzalez et al. 2011). Elemental composition has also been shown to be plastically influenced by environmental factors such as temperature, precipitation, and local nutrient availability (Janssens et al. 2015, Urbina et al. 2015, Jaenike and Markow

2003), and in some cases, may be more important than functional traits for explaining differences in elemental composition (El-Sabaawi et al. 2012). Indeed, an increasing number of studies aim to partition out the independent effects of multiple environmental parameters, such as soil chemistry and climate, on organism stoichiometry (Yang et al.

2015, Hao et al. 2015).

In this study, we tested the extent to which variation in ant elemental composition and stoichiometry could be explained by the Growth Rate Hypothesis, a Metabolic

Hypothesis, and a Storage Hypothesis (Table 2.2.1). To test these hypotheses, we investigated spatial patterns of inter- and intraspecific variation in elemental composition of worker pupae of three closely-related, morphologically and ecologically similar forest ant species, Aphaenogaster picea, A. rudis and A. fulva, whose ranges together extend along the eastern seaboard of North America and thus experience a wide range in climatic conditions. We collected field samples of worker pupae from sites across their latitudinal ranges for C, N and P analysis, and tested whether observed variation could be explained by species, body size, and composite climate variables associated with latitude, elevation,

and precipitation. If any of the composite climate variables significantly predicted elemental composition, the direction of effect informed us of which of your hypotheses may explain such patterns. To determine the extent to which variation observed in the field was caused by evolved differences in stoichiometric setpoints rather than by variation due to plastic responses to local conditions two sites were compared over two successive seasons and offspring reared under uniform laboratory conditions in a common garden experiment.

2.3. Materials and Methods

2.3.1. Colony and sample collection sites

Up to 10 ant pupae per colony were collected from 1-7 field colonies in June to late September 2013 from twenty-three sites spanning across the eastern United States from Georgia to Maine (Figure 1, Appendix Table 1). Across this geographical zone, mean annual temperature (hereafter MAT, °C) ranges from 4.7-17.6 °C. Pupae were chosen because the meconium, which is the collection of the entire stool gained during larval development, is excreted by larvae just before entering the pupal stage, when they do not feed. Thus, recently ingested food is unlikely to have affected our analyses of elemental composition. At each site (Appendix Table 1), colonies were excavated and collected as intact as possible into an aerated container. GPS coordinates, elevation, and a tentative species ID were recorded. Samples were shipped live to North Carolina State

University overnight, where pupae were separated from adults and frozen at -20°C for chemical analysis.

2.3.2. Chemical analyses

Each chemical assay (e.g. C, N, and P analysis) can be performed on one pupa; up to three pupae were used as replicates for each assay. A total of 580 pupae were analyzed.

All pupae were dried at 50oC for 48 h and individually weighed to the nearest 0.01mg prior to analysis. Percent C and N (mg/g) was determined using an elemental analyzer

(Elementar vario MICRO analyzer, Elementar, Germany). Phosphorus concentration

(mg/g) was quantified using acid persulfate digestion followed by colorimetric analysis using a Lachat Quickchem 8000, Flow Injection Analysis, Autoanalyzer (Hach Co.,

Loveland, CO). After pupae were dried and weighed, individual pupae were placed in glass extraction containers, crushed with a glass stirring rod, and 10mL of Persulfate

Digestion stock solution (40mL of 5.6M H2SO4, 2000mL DI H2O, and 16g of Persulfate) was added, while taking care to pipette down the sides of the rod to collect all material in the extraction containers. Glass extraction containers with the crushed ant pupae and

Persulfate Digestion solution were then autoclaved for 30 minutes at 121oC, 15-20 psi.

Digests were then analyzed using an ascorbic acid colorimetric method (Boren 2001).

2.3.3. Climatic predictors

To examine whether elemental composition was reflective of local climatic conditions, we obtained climate data corresponding to each site location from

BioClim.org and identified independent axes of climate variation across the sampling region by performing a principal component analysis (PCA) on all 18 BioClim variables.

PC1, PC2, and PC3 each explained 51.86%, 30.53%, and 7.18% of the total variation in climate, respectively. High PC1 loadings correlated most strongly with increases in

latitude, high PC2 loadings were associated with increases in elevation, and high PC3 loadings correlated with increases in precipitation seasonality (Supplementary Figure

2.6.2 – 2.6.3, Supplementary Table 2.6.3).

2.3.4. Species identification

Because Aphaenogaster species in the rudis-picea species group have relatively few diagnostic characters, display substantial intraspecific variation in body coloration, and may hybridize in parts of their range (DeMarco and Cognato 2015), species identity was confirmed genetically for all colonies using multi-locus SNP genotypes derived from double-digest restriction site-associated DNA sequencing (ddRADseq) according to the methodology in Helms Cahan et al. (2017). A phylogenetic tree was constructed by analyzing 65,266 SNPs in a maximum likelihood framework using GTR + gamma substitution model. Group support for each clade was evaluated with 100 bootstraps. To identify species, we captured independent nodes of the phylogeny using a principal coordinate analysis (Helms Cahan et al. 2017; Diniz-Filho et al. 2017; Paradis et al.

2004).

2.3.5. Statistical Analyses

Principal coordinate axis values for each ant colony were used to identify putative species. Thus, all statistical models analyzed used species ID as input data, rather than principal coordinate axis values. We partitioned out the relative importance of species, body mass and climate with a General Linear Model including species identity, the three climate PC variables and species x climate interactions on ant pupal elemental

composition, with body size included as a covariate (Eqn1), and then selected the best model using an AIC criterion (as in Borer et al. 2015). When interactions were present, each model was split by species and re-analyzed using AIC criterion. All statistical analyses were done in R Studio (Version 1.1.414).

(Eqn1) lm(Percent C ~ PC1 + PC2 + PC3 + species + mass + PC1*species +

PC2*species + PC3*species)

2.3.6. Common Garden study

To test if elemental differences across sites were due to variation in field conditions or intrinsic stoichiometric setpoints, nine queenright field colonies from a northern site (Neshaminy State Park, PA) and a southern site (Yates Mill, NC; Fig. 2.6.1) were collected in 2014. Up to three pupae from each colony were sampled at the time of collection in 2014 to compare elemental composition to that from the initial collections in

2013. Colonies were maintained under identical laboratory conditions within temperature-controlled growth chambers (14h:10h light-dark cycle; ~23oC) for three months to allow a cohort of fully lab-reared offspring to be raised to the pupal stage, which were then collected for analysis. A nested ANOVA (nANOVA; with colony nested within site) was conducted for each element to test for an effect of colony, species, site of origin, rearing context, or their interaction. Body mass was included in all analyses as a covariate.

2.4. Results

2.4.1. Species identification and distributions

Phylogenetic analysis of 81 colonies produced four species groups which included three nominal species: A. picea (n= 38 colonies), A. rudis (n= 31 colonies), and A. fulva

(n= 8 colonies) (Fig. 2.6.2). Aphaenogaster picea was the predominant species at northern latitudes from Pennsylvania to Maine, and was also found at two southern sites,

Blue Ridge Parkway (NC) and in the Great Smoky Mountains (TN), which occurred at a higher elevation than the other southern sites. Colonies of A. rudis were found along the southern edge of Pennsylvania down to southern North Carolina and Tennessee. Despite only collecting a few A. fulva colonies, we were able to capture its wide range, as colonies were collected from as far north as New York and as far south as North

Carolina. A fourth clade, sister to A. picea/A. rudis, consisted of three colonies from one site at the southern edge of Pennsylvania in the overlap zone between the ranges of A. picea and A. rudis. An additional single colony from northeastern New York State also fell out with a basal placement as a single long branch (Fig. 2.6.2). Their geographic origin and basal placement on the phylogeny suggested possible hybrid ancestry; all four colonies were omitted from downstream analyses. Principal Coordinate analysis produced three axes that explained 88.48% of the total genetic variation. Together, these axes clearly resolved A. fulva from A. rudis, A. picea, and the hybrid group (axis 1,

Supplementary Figure 2.6.1); A. picea from A. rudis (axis 2; Fig. Supplementary Figure

2.6.2); and the hybrid group from A. fulva, A. rudis, and A. picea (axis 5, Supplementary

Figure 2.6.3).

2.4.2. Body mass

The three species did not differ significantly in body mass (Species; F2,109=0.424,

P=0.656; Fig. 2.6.3; Supplementary Table 2.6.1), but there was a significant interaction effect between PC1 and species (Supplementary Table 2.6.1; F2,151=21.218, P < 0.0001).

When the results were split by species, only A. fulva displayed a significant relationship between body mass and any of the climate variables, increasing significantly in body mass with increasing latitude (PC1) (F1,9= 40.7542, P < 0.001; Supplementary Figure

2.6.4). Body mass of A. picea and A. rudis was not influenced by any of the PCs

(Supplementary Table 2.6.2).

2.4.3. CNP stoichiometry

Pairwise comparisons did not reveal any significant difference between species for any of the elemental ratios. Although they did not differ in overall body mass, there were significant differences between species in both percent C and percent N (ANOVA;

Percent C, F2,109=8.324, P<0.001; Percent P, F2,109=6.284, P<0.01). Pairwise comparisons revealed that A. picea pupae contained a lower percent C (Tukey’s HSD test: P < 0.001) and a higher percent N (Tukey’s HSD test: P<0.01) than A. rudis, while A. fulva was intermediate and not significantly different from either A. picea or A. rudis (Fig. 2.6.5a- b). Although significant overall differences in percent P were found (ANOVA; Percent P,

F2,109=3.463, P<0.05), pairwise comparisons did not reveal significant differences in percent P between species pairs (Fig. 2.6.4c).

Aphaenogaster picea C:N molar ratio was significantly influenced by all three environmental PCs, while C:P molar ratio showed no significant relationship with any of the variables (Supplementary Table 2.6.2; Fig. 2.6.6). C:N molar ratio increased with increasing latitude (Fig. 2.6.6, PC1; P < 0.05), but decreased with increasing precipitation seasonality (Fig. 2.6.6, PC3; P < 0.05). Aphaenogaster picea C:N molar ratio, however, showed no significant relationship with elevation (Fig. 2.6.6, PC2). Variation in C:N ratios was driven by changes in both percent C and percent N across its range. Percent C increased with increasing latitude (Fig. 2.6.5, PC1; P < 0.01) and elevation (Fig. 2.6.5,

PC2; P < 0.05), and decreased with increasing precipitation seasonality (Fig. 2.6.5, PC3;

P < 0.001). In contrast, percent N decreased with increasing latitude (Fig. 2.6.5, PC1; P

<0.05) and elevation (Fig. 2.6.5, PC2; P <0.05), and increased with increasing precipitation seasonality (Fig. 2.6.5, PC3; P < 0.0001). None of the PCs influenced percent P, although increasing body mass was associated with a decrease in percent P

(Mass; P<0.05).

Aphaenogaster rudis showed contrasting patterns of nutrient composition to that of A. picea (Supplementary Table 2.6.2). Both C:N and C:P molar ratios decreased with increasing latitude (Fig. 2.6.6; PC1; C:N, P<0.01; C:P, P<0.01). C:N ratios also decreased with increasing elevation (Fig. 2.6.6; PC2, P<0.01) and precipitation seasonality (Fig. 2.6.6; PC3, P < 0.05). All three elements changed significantly in response to environmental variation. Percent C decreased with increasing latitude (Fig.

2.6.5; PC1, P < 0.01) and elevation (Fig. 2.6.5; PC2, P < 0.05), while percent N increased with increasing latitude (Fig. 2.6.5; PC1, P < 0.01) and elevation (Fig. 2.6.5; PC2, P <

0.05). Percent P increased with increasing latitude (Fig. 2.6.5; PC1, P < 0.01), but showed no significant relationship with elevation (PC2). None of the three elements were correlated with changes in precipitation seasonality (PC3).

None of the molar ratios in A. fulva showed a significant relationship with any of the variables examined (Supplementary Table 2.6.2, Fig. 2.6.6). When considered individually, however, percent C, N, and P were all significantly associated with precipitation seasonality: percent C increased (Fig. 2.6.5; PC3, P < 0.05), while percent N and P decreased (Fig. 2.6.5; PC3, Percent N, P <0.05; Percent P, P <0.001). Percent P also significantly decreased with elevation (Fig. 2.6.5; PC2, P < 0.01).

2.4.3. Annual variation and common garden comparison

Both A. rudis and A. fulva were sampled from sites NSP and YM in 2013 and

2014. When collections from the two sites were compared across years, all three elements showed a significant site-by-year interaction (Fig. 2.6.7, Supplementary Table 2.6.3).

Although percent C, percent N, and percent P all differed significantly in 2013 between sites (Fig. 2.6.7; Percent C, F1,7=19.8566, P<0.01; Percent N, F1,7=28.2171, P<0.01;

Percent P, F1,7=11.0375, P<0.05), only percent N remained higher at site NSP the following year (Fig. 2.6.7E; F1,24=4.8286, P<0.05), while percent C and P no longer differed (Fig. 2.6.7D,F). Consistent with this pattern, both C:N and C:P ratios differed between sites in 2013 (Fig. 2.6.8; C:N ratio, F1,7=13.99, P<0.01; C:P ratio, F1,7=34.00,

P<0.001), where YM had higher C:N and C:P ratios than NSP, but this pattern disappeared in 2014 (Fig. 2.6.8D-E). Neither site, year, or their interaction significantly

affected N:P ratios (Fig. 2.6.8F). Body mass did not significantly differ between species in 2013, but in 2014, body mass was significantly different between species (Species,

F1,24=10.6912, P<0.01) and between sites (Site, F1,24=8.0282, P<0.01).

All three elements shifted in proportions when colonies were placed under common garden conditions, although overall body mass did not change (Fig. 2.6.7; main effect of treatment: percent C, F1,57 = 20.84, P < 0.001; percent N, F1,57= 33.67, P <

0.0001; percent P, F1,50 = 25.79, P < 0.0001; body mass, F1,50 = 1.16, P = 0.28). Percent

C increased under laboratory conditions while both percent N and percent P decreased, resulting in significantly higher C:N and C:P ratios (Fig. 2.6.8; main effect of treatment,

C:N, F1,57 = 54.62, P < 0.0001; C:P, F1,50 = 11.79, P < 0.01). Only percent N and N:P ratios retained relative differences between the two sites when reared in the laboratory

(Fig. 2.6.7-8; main effect of Site, percent N, F1,57 = 4.87, P < 0.05; N:P, F1,50 = 4.77, P <

0.05).

2.5. Discussion

Organismal stoichiometry can influence and be influenced by a multitude of physiological and ecological processes, which, in turn, can impact the ecological success of organisms and ultimately ecosystem functioning (Carnicer et al. 2012). However, we still have a relatively poor understanding of the mechanisms behind variation in consumer stoichiometry, particularly across large geographic scales, and how such variation may influence and be influenced by evolutionary history and environmental heterogeneity (Sun et al. 2013, Carnicer et al. 2012). Here we demonstrate how

stoichiometry and elemental composition of three ecologically important and closely related ant species (e.g. A. rudis, A. picea, and A. fulva) vary across a large latitudinal gradient and how this variation is explained by both local climate conditions and evolutionary history. We found that each species responded differently to climate.

Aphaenogaster rudis and A. picea showed contrasting patterns of stoichiometry across latitude and elevation, while A. fulva stoichiometry was mostly unresponsive to these factors, but instead showed significant variation across a gradient of precipitation seasonality. Our common garden study further revealed possible evidence of evolutionary divergence in nitrogen storage, as differences in Percent N between a northern and southern site were retained across years and under common garden conditions, regardless of species examined. Together, these results highlight the different roles that environmental heterogeneity and evolutionary history play in determining consumer stoichiometry.

A variety of hypotheses have been proposed to explain inter- and intraspecific variation in organismal stoichiometry and elemental composition. One such hypothesis, the Growth Rate Hypothesis, proposes that organisms with high growth rates should have high P content because of increased investment in P-rich RNA needed for increased protein production under high growth rates (Elser et al., 1996, 2000, 2003; Hessen et al.

2013; Sterner and Elser 2002). Increased growth rates have been linked to shorter season length (Penick et al. 2017; Elser et al. 2000), such as that seen at higher latitudes and elevations. Thus, under GRH, ants at northern latitudes and higher elevations would be expected to have higher N and P content. We did not find support for this prediction for

Aphaenogaster picea, the ant species with the northernmost distribution in this study. In contrast, A. picea displayed increased Percent C and increased C:N ratio with increasing latitude (PC1), as well as increased Percent C with increasing elevation (PC2). These results suggest that A. picea may increase investment in carbon storage over nitrogen storage with increasing latitude. Aphaenogaster picea may be forgoing storage of N and

P in favor of increased storage of C to support cold winters (e.g Storage Hypothesis).

Indeed, many insects store lipids, or C-rich energetic resources, to consume during harsh cold winter conditions (Williams et al. 2012). The pattern for A. picea contrasts with that found in other poikilothermic organisms, which display increased N and P content (RNA and protein) in response to colder temperatures (Woods et al. 2003). It is also possible that A. picea is using up C at lower latitudes because of increased metabolic rates associated with higher temperatures (e.g. Metabolic Hypothesis).

In contrast to A. picea, A. rudis showed opposite patterns of nutrient use across climate gradients. Aphaenogaster rudis stored more N and P with increases in latitude

(PC1) and stored more N with increases in elevation (PC2) (Supplementary Table 2.6.2).

Thus, patterns of elemental composition in A. rudis were consistent with the GRH hypothesis, as increases in percent N and P with increases in latitude and elevation is predicted under GRH. One reason why we may not have seen increased P content at higher elevations in A. rudis, despite seeing increased N content, could be due to increases in the rate of protein production per ribosome, instead of increases in the production of ribosomes (Moody et al. 2017; Farewell and Neidhardt 1998). Increased storage of N and P in colder locations, such as at high latitudes and elevation, may reflect

production and storage of soluble protein content in cells to prevent freezing (Woods et al. 2003).

Another mechanism which might explain the opposite patterns of elemental composition observed in A. picea and A. rudis is soil age (Reich and Oleksyn 2004).

Specifically, the Soil-Substrate-Age Hypothesis (SAH) posits that older soils tend to contain less N and P because of increased weathering which can result in increased leaching of N and P from soils (Reich and Oleksyn 2004). Under the SAH, soils at southern latitudes and lower elevations are likely to contain less N and P because soils at these locations tend to be older because of post-glaciation processes (Reich and Oleksyn

2004). Assuming absence of constant stoichiometric homeostasis, when a particular element is scarce, the amount of that element in organism biomass tends to decrease

(Sterner and Elser 2002; Turner et al. 2017). Therefore, under the SAH, organisms at lower latitudes and elevation are likely to be more responsive to deficiencies in N and P.

In keeping with this hypothesis, changes in storage of individual elements in A. rudis was reflected in corresponding shifts in stoichiometric molar ratios, where C:N ratio decreased with increasing latitude and elevation (Supplementary Table 2.6.2).

Aphaenogaster rudis also displayed decreased C:P ratio with increasing latitude

(Supplementary Table 2.6.2). Given that patterns of elemental composition and stoichiometry of A. rudis was consistent with the GRH and SAH hypotheses, but A. picea elemental composition and stoichiometry was more consistent with the MH and SH hypotheses, it’s possible that different hypotheses may be localized to specific geographic regions.

Aphaenogaster fulva appeared to use a unique nutritional strategy across its range, as it was more responsive to changes in precipitation seasonality. Specifically, A. fulva stored more N and P and less C in response to high precipitation seasonality

(Supplementary Table 2.6.2). Total C, N, and P did not significantly differ across latitude in A. fulva. Furthermore, total C and N also did not significantly differ across elevation

(Supplementary Table 2.6.2); however, A. fulva did store more P at low elevation

(Supplementary Table 2.6.2). These results suggest that A. fulva may be more responsive to changes in precipitation, as A. fulva colonies in locations with high precipitation seasonality occurred in the southeastern United States, where late-summer droughts are common (Hanson and Weltzin 2000). Increases in N and P content may be driven by corresponding increases in the production of heat shock proteins which may function to prevent water loss (Chown et al. 2011).

Comparison of field-caught and common garden-reared ants revealed evidence of evolved differences in ant stoichiometry. Lab-reared ants from the northern site, NSP, and the southern site, YM, maintained significant differences in N and N:P ratio similar to field-caught conspecifics. Workers from NSP had significantly higher percent N and

N:P ratios than those from YM, and this difference between sites was maintained over the three-month common garden rearing period. As there was no significant site x timepoint interaction, this suggests that ants at these two sites may have evolved differences in nitrogen storage. Differences in percent N between sites were also maintained across

years. However, there was a significant site x year interaction for percent N and no significant differences in N:P ratio were seen across years.

In addition to potential evolved differences in nitrogen content between our study populations, we found significant evidence for plasticity in ant stoichiometry. Percent C,

N, and P shifted over the course of the three-month rearing period, where percent C increased, and percent N and P decreased leading to higher C:N and C:P ratios than their field-collected counterparts (Supplementary Table 2.6.3). Furthermore, we found significant site x year interactions for nearly all stoichiometric parameters (e.g. percent C,

N, P, C:N, and C:P ratios), except for N:P ratio. Multiple studies have demonstrated plastic stoichiometric responses to environmental parameters (Sterner et al. 1998; Sterner and Elser 2002; Persson et al. 2010) so it is not surprising that we found this to be true for our ant populations. Given that elemental composition and stoichiometry appear to change between years, it may be possible that the observed patterns between ant stoichiometry and climate are not as strongly correlated as suggested in this study. The range of percent C between 2013 and 2014 for A. rudis colonies (~6.69%) from our common garden experiment was larger than the range in percent C across the full geographic range of A. rudis colonies (~5.14%). The range of percent N and P between

2013 and 2014 (percent N, ~0.69%; percent P, ~0.11%), however, was smaller than the range in percent N and P across the full geographic range of A. rudis colonies (percent N,

~2.69%, percent P, ~0.15%). These results suggest that the patterns observed between carbon content and climate may not hold between years, but that N and P content are likely to be more consistent over time. Further investigation is needed to determine the

extent of variation in elemental composition and stoichiometry across years and how such variation is consistent with changes in environmental conditions.

In conclusion, both environmental and genetic factors appear to influence ant stoichiometry. Given that the ant species examined in this study only diverged relatively recently (DeMarco and Cognato 2015), this study suggests that adaptive shifts in stoichiometric composition may occur rapidly in response to environmental selective pressures. Several previous studies have provided evidence for evolved divergences in organismal stoichiometry in response to selection on traits such as elemental availability

(Turner et al. 2017; Tobler et al. 2015), morphological traits (Leal et al. 2017), temperature adaptation (Schulter et al. 2014), lipid accumulation (Velmurugan et al.

2014), and energy storage and fecundity (Zhang et al. 2016). Although we were not able to explicitly link evolutionary divergence in ant stoichiometry to selection on specific traits in this study, the previous studies mentioned here suggest several possible mechanisms which might drive this divergence, and future studies are needed to link specific functional traits to evolved differences in elemental composition observed in this study, and how this correlates to species-specific responses across environmental gradients.

2.6. Figures and Tables

Figure 2.6. 1. Sampled locations across eastern U.S. Color gradient represents mean annual temperature (MAT).

Figure 2.6. 2. Phylogeny of ant colonies collected across the eastern U.S., from 23 sites, using RAD- seq. Colored branches represent different species identified through principle coordinate analysis. Blue branches represent A. picea; green branches represent A. rudis; orange branches represent a hybrid region; and red branches represent A. fulva.

Figure 2.6. 3. Differences in body mass across Aphaenogaster species.

Figure 2.6. 4. Species differences in stoichiometry.

Figure 2.6. 5. Percent Carbon, Nitrogen, and Phosphorus in ant pupae, for each species designation, as a function of PC1, PC2, and PC3.

Figure 2.6. 8. Stoichiometric ratios of ant pupae between timepoints (panels A-C) and between years (panels D-F) for two sites, Neshaminy State Park, PA (NSP) and Yates Mill, NC (YM). Panels A-C show differences in stoichiometric ratios within initial field samples versus three months later, under lab conditions. Panels D-F show differences in stoichiometric ratios of samples from the field only, between the two collection years of this study.

Supplementary Figure 2.6. 1. Principle Coordinate Analysis - PCo1~PCo2

Supplementary Figure 2.6. 2. Principle Coordinate Analysis - PCo1~PCo5

Supplementary Figure 2.6. 3. Principle Coordinate Analysis - PCo2~PCo5

Supplementary Figure 2.6. 4. Body mass in ant pupae, for each species designation, as a function of PC1.

Supplementary Figure 2.6. 5. Contour map of PC1.

Supplementary Figure 2.6. 6. Contour map of PC2.

Supplementary Figure 2.6. 7. Contour map of PC3.

Supplementary Table 2.6. 1. Geographic Variation in Ant Stoichiometry.

Dependent Independent Variables Variables df F P-value PC1 1 1.318 0.253 Body Mass Species 2 2.125 0.123 PC1*Species 2 21.218 7.57*10^(-9) PC1 1 6.74 0.010844 PC2 1 23.229 5.14*10^(-6) PC3 1 0.877 0.351302 Percent C Mass 1 16.548 9.48*10^(-5) Species 2 3.433 0.03613 PC1*Species 2 9.084 0.000237 PC2*Species 2 13.591 6.01*10^(-6) PC1 1 6.346 0.013363 PC2 1 12.334 0.000672 PC3 1 3.001 0.086306 Percent N Species 2 2.075 0.131012 PC1*Species 2 7.941 0.000633 PC2*Species 2 13.25 7.96*10^(-6) PC3*Species 2 3.131 0.048036 PC1 1 2.043 0.155888 PC2 1 5.659 0.01921 Percent P Mass 1 20.161 1.86*10^(-5) Species 2 6.55 0.002099 PC1*Species 2 7.621 0.000819 PC1 1 8.494 0.005245 PC2 1 16.523 0.000163 PC3 1 0.973 0.328404 C:N ratio Species 2 1.196 0.310566 PC1*Species 2 8.252 0.000772 PC2*Species 2 7.083 0.001904 PC1 1 2.309 0.1423 PC3 1 2.503 0.1273 C:P ratio AVG_Mass 1 1.328 0.261 Species 2 1.103 0.3489 PC1*Species 2 3.666 0.0415 N:P ratio PC2 1 2.464 0.1273

Supplementary Table 2.6. 2. Geographic Variation in Ant Stoichiometry – By Species. Dependent Independent Variables Species Variables df F P-value Estimate R-square PC1 - - - - PC2 - - - - A. fulva 0.4394 PC3 1 7.8375 0.01881 1.782 mass - - - - PC1 1 10.616 0.0018085 0.3449 PC2 1 5.649 0.0205158 0.3516 %C A. picea 0.2427 PC3 1 14.805 0.0002809 -0.6841 mass - - - - PC1 1 8.2977 0.007532 -0.9168 PC2 1 6.8754 0.013975 -1.7169 A. rudis 0.6092 PC3 - - - - mass 1 12.9103 0.001236 21.9658 PC1 - - - - PC2 - - - - A. fulva 0.4803 PC3 1 9.2403 0.01247 -0.7614 mass - - - - PC1 1 6.1252 0.01603 -0.15709 PC2 1 4.2846 0.04257 -0.18357 %N A. picea 0.2746 PC3 1 22.4518 1.27x10^(-5) 0.50508 mass - - - - PC1 1 8.3542 0.007354 0.4515 PC2 1 5.0889 0.032079 0.7249 A. rudis 0.5242 PC3 - - - - mass 1 6.4279 0.017101 -7.6068 PC1 - - - - PC2 1 15.772 0.0032479 -0.15381 A. fulva 0.7383 PC3 1 23.927 0.0008576 -0.12781 mass - - - - PC1 - - - - PC2 - - - - %P A. picea 0.06142 PC3 - - - - mass 1 4.2535 0.04317 -0.2504 PC1 1 9.3426 0.004881 0.05956 PC2 1 3.1096 0.088744 0.06773 A. rudis 0.4526 PC3 - - - - mass 1 6.2075 0.018917 -0.71017

PC1 1 1.3737 0.3619 -0.478 PC2 1 1.1194 0.401 1.5943 A. fulva 0.7743 PC3 1 4.0274 0.1826 1.6284 mass - - - - PC1 1 4.7843 0.03639 0.19765 PC2 1 2.241 0.14451 0.20125 C:N A. picea 0.2106 PC3 1 6.135 0.01891 -0.37839 mass - - - - PC1 1 9.6829 0.006343 -0.4772 PC2 1 10.782 0.004382 -1.1863 A. rudis 0.6504 PC3 1 4.786 0.042951 -0.6427 mass - - - - PC1 - - - - PC2 - - - - *Too small A. fulva PC3 - - - - sample size mass - - - - PC1 - - - - *Only PC2 - - - - intercept PC3 - - - - came out as C:P A. picea significant in Best Mass - - - - Model. PC1 1 93.4449 0.00235 -33.983 PC2 1 1.9688 0.255168 8.502 A. rudis 0.9785 PC3 - - - - mass 1 75.8274 0.003188 909.466 PC1 - - - - *No PC2 - - - - significant A. fulva PC3 - - - - interactions mass - - - - found; N:P PC1 - - - - ratio did not PC2 - - - - differ A. picea between PC3 - - - - N:P species; mass - - - - (models PC1 - - - - split by PC2 - - - - species, for PC3 - - - - A. rudis observation, also were not mass - - - - significant).

PC1 1 40.7542 0.0001277 0.0793 A. fulva PC2 - - - - 0.857 PC3 1 3.8867 0.0801519 0.0468 PC1 - - - - MASS A. picea PC2 1 2.286 0.1354 0.00849 0.03397 PC3 - - - - PC1 - - - - A. rudis PC2 1 3.161 0.08554 -0.03198 0.09533 PC3 - - - -

Supplementary Table 2.6. 3. Common Garden Statistical Results. Dependent Variables Independent Variables df F P-value Species 1 0.8765 0.3531 Site 1 0.1347 0.715 Time Point model Time Point 1 20.84 2.714x10^(-5) Site*Time %C Point 1 1.921 0.1711 Species 1 0.5404 0.4673 Site 1 1.171 0.2868 Year model Year 1 5.408 0.02615 Site*Year 1 5.253 0.02822 Species 1 0.3075 0.5814 Site 1 4.866 0.03144 Time Point model Time Point 1 33.67 3.018x10^(-10) Site*Time %N Point 1 0.04526 0.8323 Species 1 0.2239 0.6391 Site 1 17.08 0.0002211 Year model Year 1 0.6455 0.4273 Site*Year 1 7.116 0.01162 Species 1 2.563 0.1157 Site 1 3.345 0.07339 Time Point model Time Point 1 25.79 5.662x10^(-6) Site*Time %P Point 1 1.183 0.28186 Species 1 0.729 0.3994 Site 1 2.681 0.111 Year model Year 1 0.0623 0.8045 Site*Year 1 13.72 0.0007735 Species 1 0.04706 0.829 Site 1 1.967 0.1662 Time Point model Time Point 1 54.62 7.116x10^(-10) Site*Time C:N Point 1 0.9824 0.3258 Species 1 1.131 0.2951 Site 1 34.94 1.124x10^(-6) Year model Year 1 14.76 0.000508 Site*Year 1 30.46 3.63x10^(-6)

Species 1 0.02264 0.881 Site 1 3.573 0.06453 Time Point model Time Point 1 11.79 0.001204 Site*Time C:P Point 1 0.7499 0.3907 Species 1 2.46 0.1263 Site 1 4.709 0.03731 Year model Year 1 0.09191 0.7637 Site*Year 1 16.41 0.0002912 Species 1 0.09063 0.7646 Site 1 4.773 0.03363 Time Point model Time Point 1 0.003811 0.951 Site*Time N:P Point 1 0.1257 0.7244 Species 1 0.9024 0.349 Site 1 0.6257 0.4346 Year model Year 1 2.725 0.1083 Site*Year 1 1.743 0.1958 Species 1 0.9571 0.332618 Site 1 7.742 0.007592 Time Point model Time Point 1 1.1575 0.287144 Site*Time Mass Point 1 0.7186 0.400652 Species 1 7.1155 0.01175 Site 1 1.1146 0.29875 Year model Year 1 2.0394 0.16267 Site*Year 1 6.9079 0.01292

Supplementary Table 2.6. 4. PC loadings for BioClim Variables. BioClim Variable ID Comp.1 Comp.2 Comp.3 BioClim Variable Defined: bio1 -0.26779 -0.22843 0.023602 Mean Annual Temperature bio2 -0.17281 0.02422 -0.21556 Mean Diurnal Range (DRT) bio3 -0.27094 -0.08523 -0.07462 Isothermality (Bio2/Bio7) bio4 0.273712 0.14422 -0.08074 Temperature Seasonality bio5 -0.2459 -0.27027 -0.01192 Max Temp of warmest Month bio6 -0.26185 -0.23316 0.09093 Min temp of coldest month bio7 0.246497 0.182299 -0.15615 Temperature annual range bio8 -0.09354 -0.2717 -0.35948 Mean temp of wettest quarter bio9 -0.24583 -0.02923 0.089227 Mean temp of driest quarter bio10 -0.25187 -0.26202 0.0077 Mean temp of warmest quarter bio11 -0.27179 -0.21058 0.049791 Mean temp of coldest quarter bio12 -0.2386 0.31845 -0.02294 Annual Precipitation bio13 -0.22338 0.269683 -0.25702 Precipitation of wettest month bio14 -0.21979 0.30231 0.216805 Precipitation of driest month bio15 0.067459 -0.0824 -0.60246 Precipitation Seasonality bio16 -0.21785 0.286215 -0.27969 Precipitation of wettest quarter bio17 -0.23374 0.282695 0.192673 Precipitation of driest quarter Precipitation of warmest bio18 -0.2027 0.208123 -0.38975 quarter bio19 -0.22573 0.305591 0.156791 Precipitation of coldest quarter

2.7. Acknowledgements

I would like to thank the following undergraduate students for their assistance with processing and analyzing ant samples: Jenna Todero, Amanda Meyer, and Kenzie

Mahoney. Members of the “Warm Ant Dimensions” working group provided feedback and assistance throughout the initiation and completion of this work. Manisha Patel and

Aaron Ellison aided with completing phosphorus analyses. This work was funded by an

NSF Dimensions in Biodiversity grant (NSF-1136703) and a NSF Graduate Research

Fellowship to KAM.

2.8. Literature Cited

Bertram SM, Bowen M, Kyle M, Schade JD (2008). Extensive Natural Intraspecific Variation in Stoichiometric (C:N:P) Composition in Two Terrestrial Insect Species. Journal of Insect Science 8(26):1-7.

Boren K (2001). QuikChem method 10-115-01-4-E. Total phosphorus in manual persulfate digests. Lachat Instruments, Loveland, CO.

Borer ET, Lind EM, Ogdahl EJ, Seabloom EW, Tilman D, Montgomery RA, Kinkel LL (2015). Food-web composition and plant diversity control foliar nutrient content and stoichiometry. Journal of Ecology 103:1432-1441.

Carnicer J, Sardans J, Stefanescu C, Ubach A, Bartrons M, Asensio D, Penuelas J (2015). Global biodiversity, stoichiometry and ecosystem function responses to human-induced C-N-P imbalances. Journal of Plant Physiology 172:82-91.

Chown SL, Sorensen JG, Terblanche JS (2011). Water loss in insects: An environmental change perspective. Journal of Insect Physiology 57(8):1070-1084.

Cross WF, Benstead JP, Rosemond AD, Wallace JB (2003). Consumer-resource stoichiometry in detritus-based streams. Ecology Letters 6(8):721-732.

Davidson DW (2005). Ecological stoichiometry of ants in a New World rain forest. Oecologia 142:221-231.

Demarco BB, Cognato AI (2015). Phylogenetic Analysis of Aphaenogaster Supports the Resurrection of Novomessor (Hymenoptera:Formicidae). Annals of the Entomological Society of America 108(2):201-210.

DeMott WR, Pape BJ (2005). Stoichiometry in an ecological context: testing for links between Daphnia P-content, growth rate and habitat preference. Oecologia 142(1):20-27.

Diniz-Filho JAF, Ramos de Sant’Ana CE, Bini LM (1998). An eigenvector method for estimating phylogenetic inertia. Evolution 52(5):1247-1262.

El-Sabaawi RW, Zandona E, Kohler TJ, Marshall MC, Moslemi JM, Travis J, Lopez- Sepulcre A, Ferriere R, Pringle CM, Thomas SA, Reznick DN, Flecker AS (2012). Widespread intraspecific organismal stoichiometry among populations of the Trinidadian guppy. Functional Ecology 26 (3):666-676.

El-Sabaawi RW, Travis J, Zandona E, McIntyre PB, Reznick DN, Flecker A (2014). Intraspecific variability modulates interspecific variability in animal organismal stoichiometry. Ecology and Evolution 4(9):1505-1515.

Elser JJ, Dobberfuhl DR, MacKay NA, Schampel JH (1996). Organism Size, Life History, and N:P Stoichiometry. BioScience 46(9):674-684.

Elser JJ, O’Brien WJ, Dobberfuhl DR, Dowling TE (2000). The evolution of ecosystem processes: growth rate and elemental stoichiometry of a key herbivore in temperate and artic habitats. Journal of Evolutionary Biology 13(5):845-853.

Elser JJ, Acharya K, Kyle M, Cotner J, Makino W, Markow T, Watts T, Hobbie S, Fagan W, Schade J, Hood J, Sterner RW (2003). Growth rate-stoichiometry couplings in diverse biota. Ecology Letters 6(10):936-943.

Farewell A, Neidhardt FC (1998). Effect of Temperature on In Vivo Protein Synthetic Capacity in Escherichia coli. Journal of Bacteriology 180(17):4704-4710.

Gibb H, Stoklosa J, Warton DI, Brown AM, Andrew NR, Cunningham SA (2015). Does morphology predict trophic position and habitat use of ant species and assemblages? Oecologia 177(2):519-531.

González AL, Fariña JM, Kay AD, Pinto R, Marquet PA (2011). Exploring patterns and mechanisms of interspecific and intraspecific variation in body elemental composition of desert consumers. Oikos 120(8):1247-1255.

Hamback PA, Gilbert J, Schneider K, Martinson HM, Kolb G, Fagan WF (2009). Effects of body size, trophic mode and larval habitat on Diptera stoichiometry: a regional comparison. Oikos 118(4):615-623.

Hanson PJ, Weltzin JF (2000). Drought disturbance from climate change: response of United States forests. Science of The Total Environment 262(3):205-220.

Hao Z, Kuang Y, Kang M (2015). Untangling the influence of phylogeny, soil and climate on leaf element concentraitons in a biodiversity hotspot. Functional Ecology 29(2):165-176.

Helms Cahan S, Nguyen AD, Stanton-Geddes J, Penick CA, Hernáiz-Hernández Y, DeMarco BB, Gotelli NJ (2017). Modulation of the heat shock response is associated with acclimation to novel temperatures but not adaptation to climatic variation in the ants Aphaenogaster picea and A. rudis. Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology 204:113-120.

Hessen DO, Elser JJ, Sterner RW, Urabe J (2013). Ecological Stoichiometry: An elemental approach using basic principles. Limnology and Oceanography 58(6):2219- 2236.

Huberty AF, Denno RF (2006). Consequences of nitrogen and phosphorus limitation for the performance of two planthoppers with divergent life-history strategies. Oecologia 149(3):444-455. 56

Jaenike J, Markow TA (2003). Comparative elemental stoichiometry of ecologically diverse Drosophila. Functional Ecology 17(1):115-120.

Janssens L, Van Dievel M, Stoks R (2015). Warming reinforces nonconsumptive predator effects on prey growth, physiology, and body stoichiometry. Ecology 96(12):3270-3280.

Kay AD, Rostampour S, Sterner RW (2006). Ant stoichiometry: elemental homeostasis in stage-structured colonies. Functional Ecology 20(6):1037-1044.

Leal MC, Seehausen O, Matthews B (2017). The Ecology and Evolution of Stoichiometric Phenotypes. Trends in Ecology & Evolution 32(2):108-117.

McFeeters BJ, Frost PC (2011). Temperature and the effects of elemental food quality on Daphnia. Freshwater Biology 56(7):1447-1455.

Méndez M, Karlsson PS (2005). Nutrient stoichiometry in pinguicula vulgaris: nutrient availability, plant size, and reproductive status. Ecology 86(4):982-991.

Moody EK, Rugenski AT, Sabo JL, Turner BL, Elser JJ (2017). Does the growth rate hypothesis apply across temperatures? Variation in growth rate and body phosphorus of neotropical benthic grazers. Frontiers in Environmental Science 5(14).

Paradis E, Claude J, Strimmer K (2004). APE: analyses of phylogenetics and evolution in R language. Bioinformatics 20:289-290.

Penick CA, Diamond SE, Sanders NJ, Dunn RR (2017). Beyond thermal limits: comprehensive metrics of performance identify key axes of thermal adaptation in ants. Functional Ecology 31:1091-1100.

Persson J, Fink P, Goto A, Hood JM, Jones J, Kato S (2010). To be or not to be what you eat: regulation of stoichiometric homeostasis among autotrophs and heterotrophs. Oikos 119(5):741-751.

Reich PB, Oleksyn J (2004). Global patterns of plant leaf N and P in relation to temperature and latitude. PNAS 101(30):11001-11006.

Schluter L, Lohbeck KT, Gutowska MA, Groger JP, Riebesell U, Reusch TB (2014). Adaptation of a globally important coccolithophore to ocean warming and acidification. Nature Climate Change 4:1024-1030.

Schroder B, Steiner N, Merbach I, Schadler M, Filser J (2015). Collembolan reproduction in soils from a long-term fertilization experiment opposes the growth rate hypothesis. European Journal of Soil Biology 68:56-60.

Sterner RW, Clasen J, Lampert W, Weisse T (1998). Carbon:phosphorus stoichiometry and food chain production. Ecology Letters 1(3):146-150.

Sterner RW, Elser JJ (2002). Ecological Stoichiometry: the biology of elements from molecules to the biosphere. Princeton University Press.

Sun X, Zhou X, Small GE, Sterner R, Kang H, Liu C (2013). Energy storage and C:N:P variation in a holometabolous insect (Curculio davidi Fairmaire) larva across a climate gradient. Journal of Insect Physiology 59(4):408-415.

Tobler M, Alba DM, Arias-Rodriguez L, Jeyasingh PD (2015). Using replicated evolution in extremophile fish to understand diversification in elemental composition and nutrient excretion. Freshwater Biology 61(1):158-171.

Turner CB, Wade BD, Meyer JR, Sommerfeld BA, Lenski RE (2017). Evolution of organismal stoichiometry in a long-term experiment with Escherichia coli. Royal Society Open Science 4:170497.

Velmurugan N, Sung M, Yim SS, Park MS, Yang JW, Jeong KJ (2014). Systematically programmed adaptive evolution reveals potential role of carbon and nitrogen pathways during lipid accumulation in Chlamydomonas reinhardtii. Biotechnology for Biofuels 7:117.

Woods HA, Makino W, Cotner JB, Hobbie SE, Harrison JF, Acharya K, Elser JJ (2003). Temperature and the chemical composition of poikilothermic organisms. Functional Ecology 17(2):237-245.

Woods HA, Fagan WF, Elser JJ, Harison JF (2004). Allometric and phylogenetic variation in insect phosphorus content. Functional Ecology 18(1):103-109.

Yang X, Huang Z, Zhang K, Cornelissen JHC (2015). C:N:P stoichiometry of Artemisia species and close relatives across northern China: unravelling effects of climate, soil and taxonomy. Journal of Ecology 103(4):1020-1031.

Yang X, Huang Z, Zhang K, Cornelissen JHC (2017). Taxonomic effect on plant base concentrations and stoichiometry at the tips of the phylogeny prevails over environmental effect along a large scale gradient. Oikos 126(9):1241-1249.

Zhang Q, Xie J, Lyu M, Xiong D, Wang J, Chen Y, Li Y, Wang M, Yang Y (2017). Short-term effects of soil warming and nitrogen addition on the N:P stoichiometry of Cunninghamia lanceolata in subtropical regions. Plant and Soil 411(1-2):395-407.

CHAPTER 3: EFFECTS OF EXPERIMENTAL WARMING ON ANT BODY

SIZE, COLONY DEMOGRAPHY, AND NUTRITIONAL STATUS

3.1. Abstract

Climate change is likely to have different fitness consequences for organisms across latitude. Recent evidence suggests that ectotherms at southern latitudes may be more sensitive to climate change, as they may be closer to their upper thermal limits.

However, few studies have simultaneously evaluated the effects of warming on the nutritional status and functional traits of ectotherm communities at different latitudes experimentally. Using experimental warming chambers in the field at two sites, Harvard

Forest (MA) and Duke Forest (NC), we examined the effects of increasing temperature on body size, colony demography, and nutritional status of two ecologically important ant species from the genus, Aphaenogaster. We found that Aphaenogaster picea ants at the northern latitude site, Harvard Forest, were more responsive to increases in temperature.

Specifically, increases in temperature corresponded to increases in ant colony biomass, colony size, and total reproductive investment. Furthermore, increases in temperature were associated with an increase in nitrogen storage, which may be due to a shift toward a more predacious diet. In contrast to our predictions, Aphaenogaster rudis ants at the southern site, Duke Forest, were largely unresponsive to increases in temperature, as most colony demographic parameters did not change. The only demographic parameter that changed with temperature for A. rudis colonies was maximum colony size, which decreased with increases in temperature. Decreases in maximum colony size may have

been driven by migration out of the warmest chambers. Overall this study sheds more insight on the differential responses of ectotherms to climate change.

3.2. Introduction

Organisms experience wide variation in environmental conditions, and thus, have evolved key functional traits and life history strategies to maximize their success in their native environments. However, organisms may be limited in their ability to adapt to rapidly changing environmental conditions as seen under climate change. In ectotherms, temperature directly determines the rate of physiological processes, including metabolic and growth rates, which can impact behavior, reproduction, and somatic maintenance

(Kendrick and Benstead 2013, Mas-Marti et al. 2015; Lemoine et al. 2013; Brown et al.

2004). Indirect effects of climate change may also produce nutritional deficiencies through decreasing the abundance and quality of nutritional resources, which may occur via offsets in the distributions of predators and their prey, species extinctions and migrations, shifts in species interactions, or through direct effects on size, condition and stoichiometric balance of food resources (Thomas et al. 2004; Parmesan 2006; Beaumont and Hughes 2002, Diamond et al. 2012; Pelini et al. 2011a, Sterner and Elser 2000;

Sardans et al. 2012a). The impact of resource declines may be further exacerbated by increases or shifts in nutrient demand driven by upregulation of energetically costly stress responses (Sisodia and Singh 2012). Indeed, climate change has already had significant detrimental effects on ecosystems through shifts in species distributions and interactions, and increases in species extinctions (Parmesan 2006, Diffenbaugh and Field 2013; IPCC

2007; Schmitz 2013; Thornton et al. 2014; van de Waal et al. 2010).

Although climate change is a global phenomenon, both the extent of temperature increase and the sensitivity of communities to such changes are predicted to vary by latitude (Pelini et al. 2014, Deutsch et al. 2008, Shah et al. 2017). Ectotherms, which regulate their internal body temperatures based on external thermal conditions, can be more sensitive to changes in temperature at lower latitudes (Deutsch et al. 2008; Pelini et al. 2011; Pelini et al. 2014; Shah et al. 2017; Stuble et al. 2013; Diamond et al. 2012;

Huey et al. 2009). Tropical insects, for example, have narrower thermal tolerances and already live close to their thermal optima, suggesting that further warming may result in fitness declines for tropical species as compared to their temperate conspecifics (Deutsch et al. 2008; Sunday et al. 2011). Even within temperate zones, species with narrower thermal tolerances at lower latitudes are more sensitive to warming, in comparison to northern latitude conspecifics (Diamond et al. 2012; Stuble et al. 2013; Creggar et al.

2014).

However, this predicted geographic difference in sensitivity to climate change across latitude may be masked by other factors. For example, the magnitude of temperature increase is predicted to be larger at higher latitudes (IPCC2007). Although some ectotherms at northern latitudes may experience initial increases in fitness due to warming, trade-offs may occur if local adaptation to historical climates produces maladaptation to new climates (Pelini et al. 2009). For example, Pelini et al. 2009 found that despite having increased fitness under warmer conditions during the summer, peripheral populations of the butterfly, Erynnis Propertius, used more energy during the

middle of winter under warming. Geographic differences in thermal sensitivity may also be masked by other physiological and behavioral adaptations. For example, organisms may be able to modify their thermal tolerance through enhanced nutrition (Bujan and

Kaspari 2017; Anderson et al. 2010). Indeed, enhanced carbohydrate nutrition enabled higher thermal tolerance in a dominant tropical canopy ant (Bujan and Kaspari 2017). In general, however, very few studies have explored how nutrition may be used as a mechanism to regulate thermal tolerance, and how this in turn may impact organism fitness (Bujan and Kaspari 2017; Nyamukondiwa and Terblanche 2009). To determine the most useful predictors of species responses to climate change across environmentally heterogeneous environments, assessments of how warming impacts multiple functional traits, particularly nutritional and physiological traits, in conjunction with analysis of warming impacts on estimates of fitness, is needed.

Ants are numerically abundant, ecologically important ectotherms, which play a variety of roles within ecosystems. As ants are strongly responsive to changes in temperature, shifts in their performance under climate change is likely to have cascading effects throughout ecosystems (Dunn et al. 2009). In particular, essential ecosystem services which ants provide, such as seed dispersal and nutrient transport, may be disrupted. Ants at lower latitudes may be closer to their upper thermal limits, and thus may exhibit negative consequences from warming under climate change (Stuble et al.

2013; Diamond et al. 2012a; Diamond et al. 2012b; Pelini et al. 2011). For instance, two large-scale warming experiments at northern (e.g. Harvard Forest) and southern (e.g.

Duke Forest) latitude sites in North America have shown declines in ant worker

abundances and foraging intensities in ants with lower thermal tolerances at the lower latitude site, but not at the higher latitude site (Diamond et al. 2012, Stuble et al. 2013).

Furthermore, for some ant species at the southern site, increases in temperature was associated with decreased thermal niche space, or rather reductions in foraging time, but that this was not the case for ant species at the northern site (Diamond et al. 2013).

However, whether latitudinal differences in foraging may result in differential impacts of warming on ant colony nutrition and individual and colony-level life history traits is unclear.

In this study, we examined how experimental increases in temperature impact body size, colony demography, and nutritional status of two closely related species of forest ant in the genus Aphaenogaster in northern and southern latitude eastern deciduous forests. We used two large-scale field-based warming arrays to heat up the forest floor between 1.5-5.5oC (Pelini et al. 2011), to examine how increases in temperature directly and indirectly impact ant colony investment. Differential investment within ant colonies was examined in three ways: 1) as differences seen in the number of individuals produced of each life stage, 2) as shifts in storage within individuals, as seen through changes in body size or storage of essential nutritional macronutrients (e.g. Percent C, N, and P, or protein storage), and 3) as changes seen in whole ant colony biomass, thus reflecting whole colony investment. Furthermore, we examined whether increases in temperature might impact trophic position using stable isotope analysis to determine if ant colonies might use nutrition to behaviorally adapt to increases in temperature. We predicted declines in colony demographic parameters, such as colony size, at our southern site, but

either no change or weak increases in colony demographic parameters at our northern site. Further, we predicted that ants at our southern site would show stronger negative responses (e.g. steeper slopes) to increases in temperature in terms of their nutritional status.

3.3. Materials and Methods

3.3.1. Sites, warming chamber construction, and sample collection

To examine the effects of warming on ant body size and colony demography, we used a long-term warming experiment at two sites – a low latitude site (Duke Forest in

North Carolina, USA) and a high latitude site (Harvard Forest in Massachusetts, USA).

Experimental warming chambers consisted of 12 open-top chambers, 9 of which were heated between 1.5-5.5 oC above the ambient temperature, at 0.5oC increments, and the other 3 chambers served as controls (Pelini et al. 2011). The chambers were arranged spatially in three blocks, where each block consisted of one control chamber and three heated chambers. Heated chambers within each block were randomly assigned to a low

(1.5, 2.0, 2.5 oC), medium (3, 3.5, 4.0 oC), or high (4.5, 5.0, 5.5 oC) target delta. Using this regression design, as opposed to an ANOVA design, can aid in revealing potential nonlinearities and threshold effects in the relationship between temperature and organism traits and community responses (Pelini et al. 2011). All warming chambers, which were run continuously for five years, were shut off in 2015. All samples were collected at the end of this five-year period.

Figure 3.6. 1. Ant nest box structure

Within each warming chamber, there were eight artificial ant nest boxes (15cm x 15cm), each consisting of a square wooden base and a plastic cover (Figure 3.6.1). Four of these artificial ant nest boxes were placed within the warming chambers in 2010 and another four were added mid-way through the 5-year warming experiment (Diamond et al. 2016). Ant colonies were allowed to colonize these nest boxes and could move freely in and out of the warming chambers. All artificial ant nest boxes were collected at Duke

Forest and Harvard Forest in early summer 2015, at the end of the warming experiment where the chambers were shut off and extensive sampling could be done. Nest box collection was completed with hand trowels, fluon-lined plastic bins, and plastic ziplock bags. In cases where an ant colony was nesting underneath the nest box, care was taken to dig up the entire colony. Sub-samples (N=10) of ants from each colony were flash frozen in liquid nitrogen and then stored at -20oC for nutrient composition and body mass measurements. All remaining ants from each colony were collected and stored in 95%

EtOH. All ant life stages from each colony were sorted, identified, and counted. We chose to focus on two ant species, Aphaenogaster picea at Harvard Forest and

Aphaenogaster rudis at Duke Forest, as these two ant species were the most common among warming chambers at each respective site and are both ecologically important for forests, as they are key seed dispersers.

Samples of soils were taken from each warming chamber at both sites at the end of the warming experiment to determine baseline stable isotope values. Six soil cores were taken from each warming chamber by first pulling back the top leaf litter layer and obtaining the first 10cm of soil. For each chamber, the six soil cores were pooled in a gallon plastic bag and homogenized, after which a sub-sample was taken and frozen at -

20oC until analysis.

3.3.2. Trait Measurements

Colony demographic traits were determined through counting total number of individuals of each life stage, including pupae, workers, and male and female reproductives. Worker-to-brood, worker-to-pupae, and worker-to-larvae ratios were calculated for each colony, by dividing each respective ant life stage by the other.

Average body mass of each caste per colony was determined by first drying between two and ten individuals at 60oC overnight and then weighing each individual to the nearest

0.001mg. Colony biomass was calculated as the average body mass of adult workers multiplied by the number of adult workers in a colony. Total reproductive investment (Ti) and sex ratio investment (SRi) were calculated using the following equations taken from

Helms Cahan and Julian (2010): Ti = #M + #F*C; SRi = #F * C/(#M + #F*C). A

0.7 correction factor, (WF/WM) , represented C to account for the relative cost of males and

females (Boomsma 1989) and the colony-specific mean dry mass for each sex was used for WF and WM.

Total soluble protein content in individual ants was quantified spectrophotometrically at 595nm using a standard Bradford Protein Assay (Bradford

1976). A calibration curve was prepared using bovine serum albumin (Sigma Aldrich), which ranged from 0-25 ug/mL. Two individual ant workers per colony were used for each protein assay.

For stable isotope analysis, ant pupae and soil samples were dried at 60oC for 48 hours. For ant specimens, up to three individuals per colony were each individually weighed using a microbalance to the nearest 0.001mg and placed into a 5 x 9 mm tin capsule (Costech Analytical Technologies Inc., Valencia, CA, USA). Soil samples were prepared for stable isotope analysis by further homogenizing individual soil samples using a mortar and pestle and then weighing between 15mg – 25mg of the homogenized soil into tin capsules. Once weighed, all capsules were then crushed, placed into 96-well plates, and shipped to the Stable Isotope Facility at the University of California, Davis.

Total carbon, total nitrogen, and ratios of heavy to light carbon and nitrogen isotopes (e.g. 13C - 12C and 15N - 14N) were measured using a PDZ Europa ANCA-

GSL elemental analyzer interfaced to a PDZ Europa 20-20 isotope ratio mass spectrometer (Sercon Ltd., Cheshire, UK). Ratios are reported as delta (δ) values, which are calculated using the following formula:

Equation 1. Formula for the calculation of Delta (δ) stable isotope values.

Rsample represents the ratio of heavy to light isotopes in a given sample and Rstandard represents the ratio of heavy to light isotopes within a specific standard (N2 in air and

CO2 in PeeDee belemnite). Delta (δ) values are expressed in permil (‰) units, where

13 higher δ C values in consumers suggests a higher acquisition of resources from C4 plants, and higher δ15N suggests consumers are feeding at a higher trophic level, as δ15N increases with trophic position (Post 2002, Bluthgen et al. 2003). Isotopic enrichment typically ranges between 3-4‰ per trophic level (Post 2002, Bluthgen et al. 2003).

3.3.3. Statistical analyses

To determine the effect of temperature treatment on ant body mass, colony size, and total biomass, independent regressions were constructed in R Studio to determine the importance of both direct and indirect effects within the model. Linear regression was used to examine the effects of temperature and colony size on colony-level traits, including the number of individuals of specific castes, demographic ratios (e.g. worker:larvae ratio, etc.), and estimates of reproductive investment (e.g. Ti and SRi).

Colony average values were used in almost all analyses. To avoid pseudoreplication, regression models included colony ID nested within chamber. In instances where only

one sample was found in a given chamber, such as for the analyses of male and female reproductive pupae body mass, traits were analyzed using mixed models with colony ID included as a random variable instead of using colony averages. Linear regression was used to determine the effect of temperature on total carbon, nitrogen, and C:N molar ratio, as well as δ15N and δ13C. As δ15N baseline values, in soil samples, did not change across warming chambers (Figure 3.6.5), we did not standardize δ15N values of ant pupae based on baseline δ15N values within each chamber.

3.4. Results

Colony Demography and Body Mass

At Harvard Forest, total colony biomass increased within increasing temperature for A. picea (F1,19=5.704, P< 0.05). This relationship was driven by shifts in colony demography. Chamber temperature was positively associated with the mean number of workers (F1,19=7.33, P<0.05). When controlling for worker number, there was a trend of increasing number of developing larvae with temperature, however, this trend was not significant (F1,18 = 3.47, P = 0.08). Nevertheless, chamber temperature did have a significantly positive effect on to the worker:larval ratio (F1,18=5.49, P<0.05) and total reproductive investment (F1,6=10.35 P< 0.05) (Fig. 3.6.2). Maximum colony size did not change with increases in temperature. In addition to the direct effects of temperature, colonies with more workers contained significantly more worker pupae (F1,18 = 4.75, P <

0.05) and larvae (F1,18 = 8.66, P < 0.01). Neither temperature nor colony size affected sex ratio investment (Target delta, F1,6=0.006, P=0.94; colony size, F1,6=0.25, P=0.63), mean worker body mass (Target delta, F1,19=0.34, P=0.57; colony size, F1,19=0.47,

P=0.5), mean pupal body mass (Target Delta, F1,15=0.05, P=0.82; colony size, F1,15=4.53,

P=0.05), or body mass of reproductive females (Target delta, F1,4.8=0.06, P=0.82; colony size, F1,2.6=0.35, P=0.60) or males (Target delta, F1,5.6=1.75, P=0.24; colony size,

F1,4.1=0.02, P=0.90) (Fig. 3.6.2).

In contrast, neither biomass nor its constituent components were affected by chamber temperature for A. rudis at Duke Forest (Fig. 3.6.3). Chamber temperature was not significantly associated with the number of workers, brood, reproductive investment, or sex investment ratio (Fig. 3.6.3). The sole exception was the relationship between chamber temperature and the maximum colony size found, which was significantly negative (F1,10=6.3185, P=0.03), Fig. 3.6.3). Colonies with larger numbers of workers, in contrast, contained significantly more larvae (F1,38 =44.36, P < 0.0001) and higher total reproductive investment (F1,17=12.19, P<0.01). Although worker body mass was positively associated with colony size (F1,39=4.84, P < 0.05), temperature did not have a significant direct effect on body mass of any caste or lifestage (Fig. 3.6.3).

Nutritional Status

At Harvard Forest, A. picea workers contained significantly lower %C

(F1,14=6.50, P<0.05), higher %N (F1,14=5.87, P< 0.05), and lower C:N molar ratio

(F1,14=6.98, P< 0.05) with increasing temperature (Fig. 3.6.6). Despite the increase in nitrogen storage, A. picea workers did not significantly differ in mass-specific protein content across temperature treatments (Fig. 3.6.4). Aphaenogaster picea δ15N was not found to shift with increasing temperature, although there was a slight positive, though

15 non-significant trend (F1,14=2.80, P=0.12; Fig. 3.6.5). No significant change in soil δ N with increasing temperature was evident (Fig. 3.6.5). In contrast, A. picea δ13C significantly increased with increases in temperature (F1,14=5.47, P<0.05; Fig. 3.6.5).

Corresponding increases in δ13C in soils with increasing temperature also occurred

(F1,34=0.20, P=0.02; Fig. 3.6.5). There were no significant changes in %C, %N, or C:N molar ratio of soils (Fig. 3.6.6).

In contrast to A. picea, A. rudis workers at Duke Forest did not significantly differ in %C, %N, or C:N molar ratio (Fig. 3.6.6), δ15N, or δ13C (Fig. 3.6.5) across the experimental temperature treatments, and there was no change in mass-specific protein content across temperature treatments (Fig. 3.6.4). Soil %C and %N both increased with increases in temperature; however, C:N molar ratio was unchanged. A striking difference between A. rudis δ15N and soil δ15N was observed, where A. rudis δ15N was

15 approximately 1‰ lower than soil δ N (F1,104=216.63, P <0.0001; Fig. 3.6.5). In contrast, A. picea δ15N and soil δ15N values were much closer to each other.

3.5. Discussion

Increases in temperature can impact a variety of physiological traits which ultimately can impact fitness and ecological success. This effect is predicted to be most profound in warmer, less thermally variable habitats, where resident species are expected to be closer to their thermal limits and thus are more susceptible to significant fitness declines. The results of this study indicate that temperature affects ant colony demography and nutritional status differently across latitude, but in contrast to

predictions, higher-latitude populations may be more strongly affected by progressive increases in temperature. Despite their chronically warmer environments, the southern species A. rudis was largely unresponsive to increases in temperature, whether measured as colony size, demography or nutritional status. In contrast, the northern species A. picea benefitted from increasing temperatures, with larger colonies, increased reproductive output, and nutritional shifts toward increased nitrogen storage. Together, these results highlight the importance of considering organism nutrition, demography, and latitude, when making predictions of how species may respond to climate warming.

In ectotherms, increases in temperature result in increases in metabolic, growth and development rates up to an optimum, after which thermal extremes result in rapid physiological decline and ultimately death (Diamond et al. 2012; Huey and Kingsolver

1989). Previous studies have suggested that ectotherms at lower latitudes are already close to this thermal optimum, suggesting that further warming will result in declines in fitness (Diamond et al. 2012). We found limited support for this effect in A. rudis. Only the maximum colony size present within a given warming chamber declined with increases in temperature (Figure 3.6.3). Declines in maximum colony size might suggest that although colonization and growth are unaffected directly by temperature, larger colonies may be the most mobile and disproportionately migrate out of the warmest chambers, possibly in response to interspecific competition. Consistent with this pattern,

Diamond et al. (2016) showed declines in occupancy and colonization of A. rudis colonies in the same nest boxes used in this study. Furthermore, the presence of a more thermally tolerant ant species, Crematogaster lineolata, was negatively associated with

A. rudis occupancy and colonization (Diamond et al. 2016). These patterns might suggest that biotic rather than physiological constraints play a larger role in determining ant responses to increased warming at southern, high-temperature range edges, a common pattern predicted for other terrestrial species (Sunday et al. 2012).

Despite exhibiting declines in maximum colony size, direct increases in temperature in the field at Duke Forest did not result in declines in mean colony size, biomass, or body size of workers, pupae, queens, or males (Figure 3.6.3). The apparent lack of corresponding changes in the abundance of several different demographic life stages within these colonies is surprising, given that previous work in these warming chambers showed strong negative responses of ant communities to experimental increases in temperature. For example, Pelini et al. (2011) found declines in diversity, abundance, and foraging activities of several ant species with increases in temperature in the same experimental mesocosms as that studied here. One possible explanation for the lack of change in ant colony demography with increases in temperature could be because brood development at southern latitudes may be less temperature-dependent than foraging behavior. Previous studies have found evidence for a latitudinal trend (e.g. steeper slopes) in the duration and temperature dependence of brood development, where brood development of northern ant populations was more temperature dependent than southern populations (Kipyatkov and Lopatina 2002; Kipyatkov et al. 2004; Kipyatkov et al. 2005). Metrics of thermal performance, such as critical thermal limits (e.g. CTmax), have been linked to the thermal requirements of brood (Penick et al. 2017). Thus, understanding how latitude may impact the temperature dependence of brood

development and how that relationship may change with the age of ant colonies or through nutritional provisioning should provide further insight on how ants may respond to future climate change.

In contrast to the southern latitude site, temperature positively impacted fitness of

A. picea colonies at the northern site. Increases in temperature were associated with higher colony size, and the combination of the direct effects of temperature and indirect effects mediated by colony size led to higher brood production and reproductive output.

This is consistent with expectations of increased performance of northern latitude ectotherm populations with increases in temperature, as they are not as close to their upper thermal limits, and further warming is expected to move them towards a thermal optimum (Stuble et al. 2013).

Central to parent-offspring conflict theory is the prediction that parents should increase the number of offspring but not size of offspring if resources increase, as this maximizes parental fitness (Backus 1993). In contrast, offspring should work to maximize their individual fitness by demanding more resources from parents, resulting in larger offspring sizes (Backus 1993). In social insects such as ants, this conflict manifests in queen-worker conflict over investment decisions for the next generation, where trade- offs in investment in reproduction versus growth can occur (Backus 1993). Along those lines, growing worker and reproductive pupae would be expected to increase in individual size if allocation towards investment was “won” by offspring. In contrast, if allocation towards investment was “won” by the parent, the number of individuals, both

workers and reproductive, would be expected to increase. Our results are consistent with the second expectation, as increases in productivity driven by higher temperatures were mediated solely at the level of increases in numbers, while body sizes of individual offspring were unaffected. However, increases in parental investment with increases in temperature for ant colonies may not be a common pattern across time-scales, as reproductive cycles of a colony can play a large role in influencing allocation patterns

(Ode 2002). For example, food supplementation while queens were laying reproductive brood (early-fed treatment) versus food supplementation while workers were tending reproductive brood (late-fed treatment), led to larger numbers of females in the early-fed treatment but heavier females in the late-fed treatment for the ant Messor pergandei (Ode

2002). Thus, further work is needed to understand how increases in temperature impacts shifts in colony demography over longer time-scales, especially within a growing season.

Two consequences of higher temperatures may be contributing to productivity increases. First, faster development may accelerate the rate of colony growth across the growing season, such that colonies in warmer chambers have added more adults to the workforce from the current year's brood, and begun to rear additional cohorts, by the time of collection. This is supported by a shift towards older demographic life stages, indicated by an increase in the mean number of workers, as well as increases in worker: larvae ratio with increases in temperature. Second, warmer habitats may yield increased food supplies, particularly insect prey that serves as the primary protein resource for larval development (Dussutour and Simpson 2009). Although overall body masses did not change, worker pupae from colonies collected in warmer chambers had a lower C:N

molar ratio than those from cooler chambers, suggesting that greater nitrogenous input from the diet may have occurred. Although δ15N did not significantly change with increases in temperature, there was an increasing trend in δ15N (Figure 3.6.5). Increases in δ15N would indicate a potential shift in trophic level towards increased carnivory

(Bluthgen et al. 2003, Post 2002). Furthermore, prey resources may be more abundant or of higher quality, as increased N-content of plants may scale up to increased N in herbivores and consumers, including ants. Indeed, a previous warming study performed at Harvard Forest found increased N availability and foliar N content in response to soil warming (Butler et al. 2011). It is important to note that isotopic ratios can change systematically for a variety of reasons unrelated to trophic position, including temperature-dependent changes in fractionation rates or the consumption patterns of prey species; however, because δ15N in Harvard Forest soils did not significantly differ among our warming chambers (Figure 3.6.5), it is unlikely that our temperature treatments influenced baseline fractionation rates (δ15N). A more complete field isotope study encompassing multiple species across the food web would help to resolve the nature of shifts in trophic relationships and their interaction with emergence of ecological

"winners" and "losers" under climate change.

Increases in nitrogen storage appeared to not be associated with shifts in protein storage per se, as protein content did not significantly change with increasing temperature. It is possible that the Bradford protein assay, despite being a “method of choice” for protein quantification, may have been unable to detect shifts toward higher-N content proteins with skewed amino acid compositions (Wilder et al. 2015). Other N-rich

biomolecules not measured in this study, such as RNA or chitin, might also have increased in abundance. Increases in RNA content might be expected, as rapidly- growing organisms often have high RNA and N content (Elser et al. 2000) and temperature has been found to impact RNA and nutrient content of organisms (Woods et al. 2003).

Conclusions

In this study, we expanded on previous work conducted within two experimental warming arrays in the field at two latitudes which examined responses of temperate forest ant communities to increases in temperature. Specifically, we examined how temperature impacted colony demography, ant body mass and nutritional status of two closely-related ant species, A. rudis and A. picea, across latitude. Our results partially support previous studies which found that ant species at northern and southern latitudes should differ in the direction of effect of temperature on colony productivity, but suggest that the strongest effect size may occur at northern latitudes. In support of this, we found that maximum colony sizes declined with increasing temperature for our southern ant species, A. rudis.

In contrast, the northern ant species, A. picea, displayed increased mean colony size and reproductive output. Despite these changes in ant colony demography and development, previous studies have found no effect of increases in temperature on ant diversity or community composition at our northern site (Pelini et al. 2014, Pelini et al. 2011). Thus, changes in colony demography with increases in temperature might not scale up to drastic shifts at the community level. However, as Diamond et al. 2016 predicted decreased community stability with increases in temperature over time at both our

northern and southern sites, understanding how these changes in ant colony demography and nutritional status, or lack thereof, may be impacted by species interactions within communities is needed to fully understand species responses to climate change.

3.6. Figures and Tables

Figure 3.6. 2. Colony demography as a function of Temperature - Harvard Forest.

Figure 3.6. 3. Colony demography as a function of Temperature - Duke Forest.

Figure 3.6. 4. Total Protein content in ant pupae. Duke Forest is highlighted in red and Harvard Forest is highlighted in blue.

Figure 3.6. 6. Elemental composition and stoichiometry of ants and soils.

3.7. Acknowledgements

I would like to thank the following undergraduate students for their assistance with processing and analyzing samples: Curtis Provencher, Amanda Meyer, Delaney

Smith, Megna Senthilnathan, and Bonnie Kelley. Members of the “Warm Ant

Dimensions” working group provided feedback and assistance throughout the initiation and completion of this work. This work was funded by an NSF Dimensions in

Biodiversity grant (NSF-1136703) and a NSF Graduate Research Fellowship to KAM.

3.8. Literature Cited

Anderson LH, Kristensen TN, Loeschcke V, Toft S, Mayntz D (2010). Protein and carbohydrate composition of larval food affects tolerance to thermal stress and dessication in adult Drosophila melanogaster. Journal of Insect Physiology 56:336- 340.

Backus VL (1993). Packaging of offspring by nests of the ant, Leptothorax longispinosus: parent-offspring conflict and queen-worker conflict. Oecologia 95(2):283-289.

Beamont LJ, Hughes L (2002). Potential changes in the distributions of latitudinally restricted Australian butterfly species in response to climate change. Global Change Biology 8(10):954-971.

Bluthgen N, Gebauer G, Fiedler K (2003). Disentangling a rainforest food web using stable isotopes: dietary diversity in a species-rich ant community. Oecologia 137(3):426-435.

Boomsma JJ (1989). Sex investment ratios in ants: has female bias been systematically overestimated? American Naturalist 133:517-532.

Bradford MM (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72:248-254.

Brown JH, Gillooly JF, Allen AP, Savage VM, West GB (2004). Toward a metabolic theory of ecology. Ecology 85:1771-1789.

Bujan and Kaspari (2017). Nutrition modifies critical thermal maximum of a dominant canopy ant. Journal of Insect Physiology 102:1-6.

Butler SM, Melillo JM, Johnson JE, Mohan J, Steudler PA, Lux H, Burrows E, Smith RM, Vario CL, Scott L, Hill TD, Aponte N, Bowles F (2011). Soil warming alters nitrogen cycling in a New England forest: implications for ecosystem function and structure. Oecologia 168(3):819-828.

Cregger MA, Sanders NJ, Dunn RR, Classen AT (2014). Microbial communities respond to experimental warming, but site matters. Peer J 2,e358.

Deutsch CA, Tewksbury JJ, Huey RB, Sheldon KS, Ghalambor CK, Haak DC, Martin PR (2008). Impacts of climate warming on terrestrial ectotherms across latitude. PNAS 105(18):6668-6672.

Diamond SE, Nichols LM, McCoy N, Hirsch C, Pelini SL, Sanders NJ, Ellison AM, Gotelli NJ, Dunn RR (2012a). A physiological trait-based approach to predicting the responses to species to experimental climate warming. Ecology 93(11):2305-2312.

Diamond SE, Sorger DM, Hulcr J, Pelini SL, Del Toro I, Hirsch C, Oberg E, Dunn RR (2012b). Who likes it hot? A global analysis of the climatic, ecological, and evolutionary determinants of warming tolerance in ants.

Diamond SE, Nichols LM, Pelini SL, Penick CA, Barber GW, Helms Cahan S, Dunn RR, Ellison AM, Sanders NJ, Gotelli NJ (2016). Climatic warming destabilizes forest ant communities. Science Advances 2(10):e1600842.

Diffenbaugh NS, Field CB (2013). Changes in ecologically critical terrestrial climate conditions. Science 341(6145):486-492.

Dunn RR, Guenard B, Weiser MD, Sanders NJ (2009). Geographic gradients. In: Ant Ecology (eds Lach L, Parr C, Abbott K). Oxford University Press, Oxford, UK.

Dussutour A, Simpson SJ (2009). Communal nutrition in ants. Current Biology 19(9):740-744.

Helms Cahan S, Julian GE (2010). Shift in frequency-dependent selection across the life-cycle in obligately interbreeding harvester ant lineages. Evolutionary Ecology 24:359-374.

Huey RB, Deutsch CA, Tewksbury JJ, Vitt LJ, Hertz PE, Alvarez Perez HJ, Garland T (2009). Why tropical forest lizards are vulnerable to climate warming. Proceedings of the Royal Society of London B: Biological Sciences 276(1664):1939-1948.

Huey RB, Kingsolver JG (1989). Evolution of thermal sensitivity of ectotherm performance. Trends in Ecology and Evolution 4(5):131-135.

Kendrick MR, Benstead JP (2013). Temperature and nutrient availability interact to mediate growth and body stoichiometry in a detritivorous stream insect. Freshwater Biology 58(9):1820-1830.

Kipyatkov VE, Lopatina EB, Imamgaliev AA, Shirokova LA (2004). Effect of temperature on rearing of the first brood by the founder females of the ant Lasius niger (Hymenoptera, Formicidae): latitude-dependent variability of the response norm. Journal of Evolutionary Biochemistry and Physiology 40(2):165-175.

Kipyatkov V, Lopatina E, Imamgaliev A (2005). Duration and thermal reaction norms of development are significantly different in winter and summer brood pupae of the ants Myrmica rubra Linnaeus, 1758 and M. ruginodis Nylander, 1846 (Hymenoptera:Formicidae). Myrmecologische Nachrichten 7:69-76.

Kipyatkov VE, Lopatina EB (2002). Reaction norm in response to temperature may change to adapt rapid brood development to boreal and subarctic climates in Myrmica ants (Hymenoptera: Formicidae). European Journal of Entomology 99:197-208.

Lemoine NP, Drews WA, Burkepile DE, Parker JD (2013). Increased temperature alters feeding behavior of a generalist herbivore. Oikos 122(12):1669-1678.

Mas-Marti E, Romani AM, Munoz I (2015). Consequences of warming and resource quality on the stoichiometry and nutrient cycling of a stream shredder. PLoS one 10(3):e0118520.

Nyamukondiwa C, Terblanche JS (2009). Thermal tolerance in adult Mediterranean and natal fruit flies (Ceratitis captiata and Ceratitis rosa): effects of age, gender and feeding status. Journal of Thermal Biology 34:406-414. 87

Ode PJ, Rissing SW (2002). Resource abundance and sex allocation by queen and workers in the harvester ant, Messor pergandei. Behavioral Ecology and Sociobiology 51(6):548-556.

Parmesan C (2006). Ecological and evolutionary responses to recent climate change. Annual Review of Ecology, Evolution, and Systematics 37(1):637-669.

Pelini SL, Dzurisin JDK, Prior KM, Williams CM, Marsico TD, Sinclair BJ, Hellmann JJ (2009). Translocation experiments with butterflies reveal lmits to enhancement of poleward populations under climate change.

Pelini SL, Diamond SE, Nichols LM, Stuble KL, Ellison AM, Sanders NJ, Gotelli NJ (2014). Geographic differences in effects of experimental warming on ant species diversity and community composition. Ecosphere 5(10):1-12.

Penick CA, Diamond SE, Sanders NJ, Dunn RR (2017). Beyond thermal limits: comprehensive metrics of performance identify key axes of thermal adaptation in ants. Functional Ecology 31(5):1091-1100.

Post DM (2002). Using stable isotopes to estimate trophic position: models, methods, and assumptions. Ecology 83:703-718.

Sardans J, Rivas-Ubach A, Penuelas J (2012). The elemental stoichiometry of aquatic and terrestrial ecosystems and its relationships with organismic lifestyle and ecosystem structure and function: a review and perspectives. Biogeochemistry 111(1- 3):1-39.

Schmitz (2013). Global climate change and the evolutionary ecology of ecosystem functioning. Annals of the New York Academy of Sciences 110(27):11035-11038.

Shah AA, Gill BA, Enalada AC, Flecker AS, Funk WC, Guayasamin JM, Kondratieff BC, Poff NL, Thomas SA, Zamudio KR, Ghalambor CK (2017). Climate variability predicts thermal liits of aquatic insects across elevation and latitude. Functional Ecology 31:2118-2127.

Sisodia S, Singh BN (2012). Experimental evidence for nutrition regulated stress resistane in Drosophila ananassae. PLoS one 7(10):e46131.

Sterner RW, Elser JJ (2002). Ecological stoichiometry: the biology of elements form molecules to the biosphere. Princeton University Press.

Stuble KL, Pelini SL, Diamond SE, Fowler DA, Dunn RR, Sanders NJ (2013). Foraging by forest ants under experimental climatic warming: a test at two sites. Ecology and Evolution 3(3):482-491.

Sunday JM, Bates AE, Dulvey NK (2011). Global analysis of thermal tolerance and latitude in ectotherms. Proceedings of the Royal Society B: Biology 278:1823-1830.

Sunday JM, Bates AE, Dulvey NK (2012). Thermal tolerance and the global redistribution of animals. Nature Climate Change 2:686-690.

Wilder SM, Raubenheimer D, Simpson SJ (2015). Moving beyond body condition indices as an estimate of fitness in ecological and evolutionary studies. Functional Ecology 30:108-115.

Woods HA, Makino W, Cotner JB, Hobbie SE, Harrison JF, Acharya K, Elser JJ (2003). Temperature and the chemical composition of poikilothermic organisms. Functional Ecology 17(2):237-245.

CHAPTER 4: CONCLUSIONS AND FUTURE DIRECTIONS

Climate change is predicted to impact organism nutritional ecology through direct effects on organism physiology and through indirect effects on the quantity and quality of nutritional resources (Sardans et al. 2012a, Sardans et al. 2012b, Thornton et al. 2014,

Cross et al. 2015). Thus, understanding how environmental stress, such as increases in temperature, impact organism nutritional ecology is particularly important, especially given the increasingly detrimental effects observed due to climate change (Diffenbaugh and Field 2013, Schmitz 2013, Thornton et al. 2014, van de Waal et al. 2010, Cross et al.

2015). Analysis of organism stoichiometry allows for connecting different levels of biological organization, as the balance of multiple chemical elements in living systems can impact the function of individuals, communities, and ultimately ecosystems (Sterner and Elser 2002). Thus, examination of elemental stoichiometry can aid in understanding how changes in climate may impact nutritional ecology. However, our understanding of the proximate and ultimate mechanisms driving variation in consumer stoichiometry, however, is still quite limited (Sun et al. 2013). This thesis has provided evidence that both climate and evolutionary history influence ant stoichiometry and this, in turn, can impact ant colony fitness.

Ant stoichiometry of three-closely related ant species from the genus

Aphaenogaster was linked to climate variables across a broad latitudinal gradient along the eastern United States (Chapter 2). Differences in nutritional storage strategies were

observed across the range of Aphaenogaster ant species, as C:N:P stoichiometry in whole ant pupae responded differently to climate variables depending on the species examined.

Specifically, the northernmost species, A. picea, stored more C, and less N and P at higher latitudes and elevation, whereas the southernmost species, A. rudis, displayed the opposite pattern. Increased carbon storage at northern latitudes and higher elevations may reflect increased lipid and carbohydrate storage, as these biomolecules represent a form of energy storage which ants can use to withstand harsh winter conditions (Sorvari et al.

2011). Other cold stress resistance compounds may also reflect differences in carbon storage, as certain carbon-rich carbohydrates such as glycerol can serve as a cryoprotectant to maximize cold-hardiness (Hahn and Denlinger 2007). Increased N and

P storage at southern latitudes and lower elevations may reflect N and P limitation in basal resources. Meta-analyses collected from mid-latitudes to the tropics have found decreases in N and P content of soils and plant leaves with decreasing latitude (Reich and

Oleksyn 2004, Zechmeister-Boltenstern et al. 2015). Changes in the stoichiometry of basal resources can scale up to higher trophic levels and influence insect performance

(Schade et al. 2003). Thus, examination of how clinal differences in soil and plant nutrient composition impacts ant stoichiometry is a fruitful area for future research.

Interestingly, the third species examined, A. fulva, which has a broader geographical range that partially overlaps A. picea and A. rudis, was not found to be responsive to changes in latitude or elevation. Instead, A. fulva responded strongly to changes in precipitation seasonality, where increased precipitation seasonality was associated with increased C storage, suggesting the A. fulva may be more responsive to

drought conditions. Arnan et al. 2014 demonstrated that precipitation seasonality significantly impacted the largest number of functional traits of Mediterranean ants.

Thus, further investigation of how precipitation seasonality impacts ant stoichiometry in conjunction with an analysis of effects on functional traits is needed. As carbon-rich cuticular hydrocarbons play a role in reducing water loss in insects (Chown et al. 2011), future work could also examine variation in cuticular hydrocarbon composition of ants along gradients in precipitation seasonality. Together, these distinct clinal patterns in ant stoichiometry demonstrate that Aphaenogaster ant species can respond plastically to changing environmental conditions.

In a common garden study, I further demonstrated that adaptive divergence in elemental composition is at least partially responsible for the observed clinal patterns in ant stoichiometry in the field. When reared under the same feeding and thermal regimes, differences in N storage between Aphaenogaster ants from a northern and a southern population were retained over time while under lab conditions. Further, differences in nitrogen storage between sites were retained when examining ant stoichiometry between two different years in the field. Adaptive divergence in nitrogen storage may be explained by local adaptation to low-N conditions in basal resources (Fagan et al. 2002), where southern Aphaenogaster ant populations evolved lower dependence on N, which is expected to be limiting at lower latitudes (Reich and Oleksyn 2004). Lower dependence on N as a result of local adaptation may be reflected in shifts in the use and storage of amino acids which vary in N content (Fagan et al. 2002). The emerging field of

“Stoichioproteomics”, which seeks to relate the elemental composition of proteomes and

proteins to environmental and physiological variation in organisms, presents a promising avenue to explore such adaptive divergence in consumer stoichiometry (Gilbert et al.

2013).

To connect latitudinal patterns to temporal changes projected under climate change, I evaluated how increases in temperature impacts ant stoichiometry and associated functional traits at the individual and colony level using an experimental field mesocosm at two sites, Harvard Forest (HF) and Duke Forest (DF) (Chapter 3). Increases in temperature may impact ant stoichiometry through increases in growth and development rates or through indirect effects on food web interactions. Thus, to examine how increases in temperature impacts ant colony development and how such shifts may or may not reflect changing nutrient conditions, I examined how experimental increases in temperature impacted body size, colony demography, and nutritional status of two

Aphaenogaster ant species within this system. I found that ants at the northern site, HF, responded positively to direct increases in temperature, with increases in colony biomass, colony size, total reproductive output, and shifts towards increased nitrogen storage with increases in temperature. Increases in ant colony fitness is expected for ant populations at northern latitudes, as such populations are not as close to their upper thermal limits

(Diamond et al. 2012). Increases in nitrogen storage with increases in temperature may be due to increases in nitrogen availability in basal resources, as soil N was found to increase with increasing temperature in this study. Butler et al. (2011) also found increases in soil and foliar N content with increases in temperature at HF, despite using a different experimental set-up than this study. Although nitrogen is a primary component

of protein, we did not find that temperature impacted mass-specific protein concentrations. Thus, examination of changes in the abundance of other N-rich biomolecules, such as RNA and chitin, with increases in temperature is needed to explain this increase in nitrogen storage. In contrast to Aphaenogaster ants at HF, ants at the southern site, DF, were generally unaffected by temperature, except for a decrease in maximum colony size with increases in temperature. This was surprising, as ants at southern latitudes are predicted to be closer to their upper thermal limits, and thus increases in temperature are predicted to have negative fitness consequences (Diamond et al. 2012). Increased migration of larger, older ant colonies out of warmer chambers may explain the overall lack of change in ant colony demography, body size, and nutritional status observed at DF.

Ultimately, my results demonstrate the role that climate and evolutionary history can play in determining ant stoichiometry. My work further demonstrates how increases in temperature, as expected under climate change, may impact the ecological success of ant colonies differently across latitude and this can relate to changes in ant stoichiometry.

This thesis further highlights the need to examine the biochemical basis of stoichiometric trait variation to ascertain the role stoichiometry may play in how ant species adapt to changing environmental conditions. As changes in the abundance of other biologically relevant elements, such as magnesium and sodium (Sun et al. 2013, Kaspari et al. 2008), within organisms has been found to vary with environmental conditions, examination of shifts in micronutrient composition within ants across climatic gradients should provide for a more holistic understanding of the impacts of climate on ant nutritional ecology.

4.2. Literature Cited

Arnan X, Cerda X, Retana J (2014). Ant functional responses along environmental gradients. Journal of Animal Ecology 83(6):1398-1408

Chown SL, Sorensen JG, Terblanche JS (2011). Water loss in insects: an environmental change perspective. Journal of Insect Physiology 57(8):1070-1084.

Cross WF, Hood JM, Benstead JP, Huryn AD, Nelson D (2015). Interactions between temperature and nutrients across levels of ecological organization. Global Change Biology 21(3):1025-1040.

Diffenbaugh NS, Field CB (2013). Changes in ecologically critical terrestrial climate conditions. Science 341(6145):486-492.

Fagan WF, Siemann E, Mitter C, Denno RF, Huberty AF, Woods HA, Elser JJ (2002). Nitrogen in insects: implications for trophic complexity and species diversification. The American Naturalist 160(6):784-802.

Gilbert JD, Acquisti C, Martinson HM, Elser JJ, Kumar S, Fagan WF (2013). GRASP [Genomic Resource Access for Stoichioproteomics]: comparative explorations of the atomic content of 12 Drosophila proteomes. BMC Genomics 13:599.

Hahn DA, Denlinger DL (2007). Meeting the energetic demands of insect diapause: Nutrient storage and utilization. Journal of Insect Physiology 53(8):760-773.

Kaspari M, Yanoviak SP, Dudley R (2008). On the biogeography of salt limitation: A study of ant communities. PNAS 105(46):17848-17851.

Reich PB, Oleksyn J (2004). Global patterns of plant leaf N and P in relation to temperature and latitude. PNAS 101(30):11001-11006.

Schade JD, Kyle M, Hobbie SE, Fagan WF, Elser JJ (2003). Stoichiometric tracking of soil nutrients by a desert insect herbivore. Ecology Letters 6(2):96-101.

Schmitz (2013). Global climate change and the evolutionary ecology of ecosystem functioning. Annals of the New York Academy of Sciences 110(27):11035-11038.

Sorvari J, Haatanen M, Vesterlund S (2011). Combined effects of overwintering temperature and habitat degradation on the survival of boreal wood ant. Journal of Insect Conservation 15 (5):727-731.

Sterner RW, Elser JJ (2002). Ecological stoichiometry:the biology of elements form molecules to the biosphere. Princeton University Press.

Zechmeister-Boltenstern S, Keiblinger KM, Mooshammer M, Penuelas J, Sardans J, Wanek W (2015). The application of ecological stoichiometry to plant-microbial-soil organic matter transformations. Ecological Monographs 85(2):133-155.

COMPREHENSIVE BIBLIOGRAPHY