CONSEQUENCES OF ENVIRONMENTAL HETEROGENEITY ON REPRODUCTIVE OUTPUT IN THE LEAF-FOOTED CACTUS BUG, FEMORATA (: )

By

LAUREN ANNE CIRINO

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2020

© 2020 Lauren Anne Cirino

To my family

ACKNOWLEDGMENTS

I am so lucky to have the family and friends that I do. I would first like to thank my wonderful family. My mom and dad, Janice and Charles Cirino, taught me from a young age to never give up on my dreams. My brother, Todd Cirino, has always encouraged me to pursue my science dreams even though I had already set out on another career path. This degree is one step in the direction of that big dream, and I could not have achieved it without them. This degree is not only mine, it is theirs too.

Forming friendships through science has been such a wonderful and supportive experience. I would like to thank my amazing friends who have stuck by me throughout the years. I am appreciative of your love, support, and encouragement. I would like to specifically thank Drs. Deanna Colton, Susan Lad, Michala Stock, Amanda Friend, and

Ginny Greenway for their love and support. I am so grateful to have such strong and supportive women in my life. This degree would have been far more challenging without your encouragement.

I am also grateful for the mentorship and guidance I have received from Christine

Miller throughout this adventure of graduate school. Without her support, patience, and encouragement, this degree would have been far more arduous. I would also like to thank my committee members – Patricia Moore, Colette St. Mary, Lisa Taylor, and Todd

Palmer. I appreciate your advice and valuable contributions to my research. I would also like to thank my funding sources that contributed to the success of this research –

University of Florida Graduate Research Fellowship and the National Science

Foundation (IOS-1553100 to Christine Miller). Finally, I would like to thank the

Behavior Society and the Society for Integrative and Comparative Biology for awarding me grant money to share the results of my dissertation at their conferences.

4

Many undergraduate students have made this dissertation research and other research endeavors possible during my years as a graduate student at the University of

Florida. I would like to thank Skyler Brandfon, Tyler Campbell, Devin Fabian, Stella

Fedele, Daniela Gomez, Amberlika Guruvadoo, Kaylin Kleckner, Brandon Latchman,

Haley Lenga, Meredith Lilley, Amun Majeed, Gagan Midathala, Elizette Rodriguez,

Bryanna Sharot, Kayli Sieber, Ebony Taylor, Joshua Vildor, Kathleen Wang, and

Maxwell Woolridge. Their smart, dedicated, and diverse experiences have truly enriched my life and made my dissertation less stressful. Mentoring them was one of the highlights of my time at the University of Florida. I would also like to thank the funding sources that funded a few of my undergraduate researchers including the

Undergraduate Scholars Research program (Haley Lenga), the McNair Scholars program (Ebony Taylor and Daniela Gomez), and Animal Behavior Society (Daniela

Gomez). Funding the next generation of scientists is crucial for increasing diversity and inclusion.

Finally, I would like to thank all of the Miller Lab members who have been incredibly supportive. I thank Zachary Emberts, Michael Forthman, Sara Zlotnik, Tamsin

Woodman, Ginny Greenway, Daniela Wilner, Ummat Somjee, and Pablo Allen. These fantastic lab mates have brainstormed ideas, helped troubleshoot projects, revised my writing, edited my slides, and helped shape me as a scientist among other things. I am so thankful that they were my lab mates and even more grateful they are my friends.

5

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 9

LIST OF FIGURES ...... 10

ABSTRACT ...... 11

CHAPTER

1 INTRODUCTION ...... 13

2 SEASONAL NUTRITION AFFECTS FEMALE REPRODUCTIVE ANATOMY BUT NOT FEMALE MATING BEHAVIOR ACROSS A LONG BREEDING SEASON ...... 19

Background ...... 19 Methods ...... 22 Rearing and Diet Manipulations ...... 22 Mating Behavior Trials ...... 25 Female Anatomy ...... 26 Statistical Analyses ...... 26 Juvenile diet – week 1 females ...... 26 Adult diet switch and age – week 2 and 3 females ...... 27 Results ...... 28 Juvenile Diet – Week 1 Females ...... 28 Adult Diet Switch and Age – Week 2 and 3 Females ...... 28 Discussion ...... 29

3 MATERNAL BODY SIZE, BUT NOT DIET, AFFECTS EGG SIZE IN THE LEAF- FOOTED CACTUS BUG, NARNIA FEMORATA ...... 42

Background ...... 42 Methods ...... 46 Insect Husbandry ...... 46 Insect Rearing ...... 47 Statistical Analyses ...... 49 Results ...... 49 Discussion ...... 50

4 SEASONAL CHANGES IN DIET QUALITY PARTIALLY RESCUE LONG TERM FEMALE REPRODUCTIVE SUCCESS ...... 56

Background ...... 56

6

Methods ...... 59 Experimental Design ...... 59 Insect rearing ...... 59 Juvenile survivorship and development ...... 60 Adult longevity and fecundity ...... 61 Statistical Analyses ...... 62 Juvenile survivorship and development ...... 62 Overall female reproductive success ...... 63 Components of female reproductive success ...... 63 Speed of adult diet rescue ...... 64 Results ...... 65 Juvenile Survival and Development ...... 65 Overall Female Reproductive Success ...... 65 Components of Female Reproductive Success ...... 65 Female longevity ...... 65 Female fecundity ...... 66 Speed of Adult Diet Rescue ...... 66 Discussion ...... 67

5 MALES WITH DAMAGED WEAPONS PRODUCE MORE OFFSPRING THAN INTACT MALES IN NON-COMPETITIVE ENVIRONMENTS ...... 77

Background ...... 77 Methods ...... 80 Experimental Design ...... 80 Insect husbandry and rearing ...... 80 Treatments ...... 81 Experiment 5.1: 24-hour mate switch ...... 81 Experiment 5.2: Visual confirmation of copulations ...... 82 Photographing and measurement protocol ...... 83 Statistical Analysis ...... 83 Body size PCAs ...... 83 Experiment 5.1: 24-hour mate switch ...... 84 Experiment 5.2: Visual confirmation of copulations ...... 85 Ethical Note ...... 85 Results ...... 86 Body Size PCAs ...... 86 Experiment 5.1: 24-Hour Mate Switch ...... 86 Overall male reproductive success is enhanced for small, weapon- damaged males ...... 86 Weapon-damaged males that mate with large females have higher reproductive success ...... 87 Experiment 5.2: Visual Confirmation of Copulations ...... 87 Larger intact males are more likely to copulate with many females ...... 87 Discussion ...... 87

6 CONCLUSION ...... 96

7

APPENDIX SUPPLEMENTARY FIGURES AND TABLES ...... 101

LIST OF REFERENCES ...... 103

BIOGRAPHICAL SKETCH ...... 118

8

LIST OF TABLES

Table page

2-1 Diet manipulations for both juvenile and adult Narnia femorata...... 37

2-2 Analyses examining the behavior and anatomy of week 1 females...... 38

2-3 Analyses examining the behavior and anatomy of sexually mature Narnia femorata ...... 39

4-1 Diet treatments for juvenile and adult Narnia femorata ...... 72

5-1 Analysis of overall male reproductive success...... 92

5-2 Analysis of the components of male reproductive success ...... 93

5-3 Analysis of visual confirmation of copulations...... 94

A-1 Factor loadings by trait for all Principal Component Analyses in Chapter 5 ..... 101

A-2 Correlation matrix for the PCAs in Chapter 5, Experiment 1...... 101

A-3 Correlation matrix for the PCAs in Chapter 5, Experiment 2 ...... 102

A-4 Total variance explained by PCA in Chapter 5 ...... 102

A-5 Regression of PC1 values in Chapter 5 ...... 102

9

LIST OF FIGURES

Figure page

2-1 Narnia femorata feeding and mating ...... 36

2-2 Opuntia mesacantha ssp. lata fruit abundance and quality in 2014-2015 ...... 37

2-3 Flow chart of treatments and analyses ...... 38

2-4 Narnia femorata females dissected from each diet treatment ...... 39

2-5 Anatomy of newly eclosed adult females is affected by diet ...... 40

2-6 Female mating behavior of sexually mature Narnia femorata ...... 40

2-7 Anatomy of sexually mature (week 2 and 3) adult females ...... 41

3-1 Adult and first instar Narnia femorata ...... 53

3-2 Narnia femorata egg clutch ...... 53

3-3 Effect of juvenile diet on body size ...... 54

3-4 Body size and diet both separately affect whether or not females lay eggs ...... 54

3-5 Effect of body size and diet on egg volume ...... 55

4-1 Adult feeding and recently hatched juvenile Narnia femorata ...... 72

4-2 The effect of juvenile diet on juvenile survivorship and development ...... 73

4-3 Overall female reproductive success ...... 74

4-4 The effect of diet on adult longevity and fecundity over time ...... 75

4-5 Speed of adult diet rescue ...... 76

5-1 Juvenile and adult Narnia femorata with damaged weapons ...... 92

5-2 Overall male reproductive success ...... 93

5-3 Components of male reproductive success ...... 94

5-4 Visual confirmation of copulations ...... 95

10

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

CONSEQUENCES OF ENVIRONMENTAL HETEROGENEITY ON REPRODUCTIVE OUTPUT IN THE LEAF-FOOTED CACTUS BUG, NARNIA FEMORATA (HEMIPTERA: COREIDAE)

By

Lauren A. Cirino

August 2020

Chair: Christine W. Miller Major: Entomology and Nematology

Reproduction is central to the fitness of . And, while it is so important, ecological factors routinely compromise reproductive output. Here, I show that natural environmental changes in diet can have important effects on reproductive traits. I examined reproductive traits under semi-natural conditions using the leaf-footed cactus bug, Narnia femorata. N. femorata has a seasonal host plant diet of prickly-pear cactus.

Cactus fruit blooms, ripens, and disappears over the year. Since N. femorata breed throughout the year, they can be exposed to varying nutrition that can affect reproduction. Thus, this species is well suited to investigate the effects of dynamic environmental conditions on reproductive traits.

I found that improved adult nutrition helps females reproductively recover from poor juvenile diets. I found substantial effects of nutrition at key life stages on female anatomy, longevity, and fecundity. I also show that other factors (time of year and female body size) can also impact female reproductive traits and these factors may contribute to the large variation in reproductive output that we observe in wild populations.

11

Further, these can perform an anti-predatory behavior, autotomy, in which they can drop a limb to escape predation. If males lose a hind leg during development, they grow large testes as adults. However, hind legs are used as weapons to gain access to mates. If males lose a hind leg, than their weapon is damaged and they are likely to lose contests for mates. I found that, in a non- competitive context, weapon-damaged males have higher fertilization success than intact males when they mate with large females. Large weapon-damaged males also mate with fewer females than intact males. Together, these results suggest that some weapon-damaged males may be able to reproductively compensate for pre-copulatory loss though post-copulatory means.

12

CHAPTER 1 INTRODUCTION

Having a suitable phenotype for the environment is crucial to survival and reproduction. Therefore, heritable traits that maximize survival and reproduction under local environmental conditions are more likely to persist in a population, leading to evolutionary change. However, habitats are incredibly diverse and dynamic across space and time. Change can occur quickly at times, altering the direction or intensity of selection, and slowing or speeding up the evolutionary process. Animals must endure changes to their habitats such as changes to food, and changes in food quality and quantity is arguably one of the more important changes. It is important that we examine the trait expression in animals across relevant and dynamic conditions because ultimately it is this phenotypic expression that is exposed to selection.

Reproduction is central to the fitness of animals. In many sexually reproducing animals both females and males contribute gametes to produce offspring. Pre- copulatory mating behavior, number of gametes, and duration of reproductive life are just some of the traits involved in reproductive success. Under ideal conditions, females and males may be able to enhance all reproductive traits and increase their reproductive success. However, optimal environmental conditions are rare and animals are often exposed to poor conditions due to seasonal changes, resource competition, and predator intensity. Poor conditions can increase the resource investment to one trait at the cost of another (i.e. tradeoff). In the last two decades, traits important to reproduction have been the focus of many studies of trait covariances (Mautz et al.

2013, Evans and Garcia-Gonzalez 2016), with theory predicting that greater values of one trait should be associated with reduced values of another trait (a negative

13

covariance). Yet, empirical studies have found a mix of positive and negative covariances, or no association among reproductive traits (Mautz et al. 2013, Evans and

Garcia-Gonzalez 2016, Simmons et al. 2017). Why are we getting contradicting results?

One reason for this ambiguity may be because environmental conditions vary.

Resource allocation decisions must be made for individuals to survive and reproduce under all types of environments. Investigating allocation patterns in reproductive traits under varying natural environmental conditions may reveal the complexity of allocation decisions that organisms make between these important life history traits.

I empirically explored how natural environmental conditions affect reproductive output in my dissertation research. First, I investigated how natural seasonal nutrition at important life stages affect phenotypic correlations among female reproductive traits. I thoroughly examined female reproduction by investigating numerous reproductive traits that relate to female pre- and post-copulatory success (Chapters 2 – 4). I used the leaf- footed cactus bug, Narnia femorata Stål (Hemiptera: Coreidae) to examine these traits.

N. femorata feeds on a seasonal diet that consists of prickly-pear cactus, Opuntia mesacantha ssp. lata (Small) Majure (Cactaceae). Cacti bloom in late April – May each year, and across the landscape they progress in ripening, with the first ripe fruits appearing in July, and some not appearing until autumn. Across this time, insects continue to breed, with at least three generations per year, April – December. These changes mean that insects are exposed to varying nutritional conditions at different life stages.

In my first chapter, I examine the effects of multiple ecological factors on female mating behavior and reproductive anatomy. Breeding seasons are long in many taxa

14

(e.g. Mammals: Rosen 2009, Aves: Blomberg et al. 2013, Reptiles: Hall et al. 2018,

Insects: Uyi et al. 2018). Along with changes in nutrition, other abiotic factors (e.g. temperature and photoperiod) can simultaneously change and alter female reproduction. I found that female reproductive anatomy can partially recover from poor juvenile diets through improved nutrition at adulthood. Variable fruit diets had no effect on the anatomical traits that I measured here, but the time of year was important to female reproductive anatomy. Females that became adults later in the season had reduced reproductive activity (i.e. smaller ovaries and less likely to have eggs in their oviduct), likely preparing their bodies for overwintering as many other animals do

(Tauber et al. 1986, Ultsch 1989, Paul et al. 2008, McAllan and Geiser 2014).

Surprisingly, mating behavior did not appear to be affected by female diet. Instead, female age was more important for females’ to accept a mate and is consistent with female mating behavior research across taxa (Gray 1999, Kodric-Brown and Nicoletto

2001, Moore and Moore 2001, Richard et al. 2005, Barrett et al. 2009b, Anjos-Duarte et al. 2011, Wilner et al. 2020).

Environmental factors, such as nutrition, experienced by females can impact their offspring by way of maternal effects. Maternal effects are when the maternal phenotype affect the offspring phenotype. One likely ubiquitous examples of a maternal effect is egg provisioning. Offspring can benefit from greater resources provisioned to their eggs as large eggs are more likely to hatch and produce large offspring that develop quickly and survive to adulthood (Reavey 1992, Fox and Dingle 1994, Fox and Mousseau

1996). In Chapter 3, I investigate how maternal traits, diet and body size, affect egg size. These maternal traits are known to effect egg size in insects (Fox and Czesak

15

2000). Yet, I find that maternal body size, not diet, affect egg size. This is an intriguing result since diet affects so many other reproductive traits in N. femorata (Chapters 2 and 4). It is possible that females in this species stop producing eggs instead of producing tiny eggs that are less likely to be viable.

In Chapter 4, I examine the effects of host plant diet on female longevity and fecundity. I also investigate the extent to which a seasonal switch to high quality diet can improve these traits when females were raised on low-quality diets. I find that females can quickly, but partially, recover from a poor early life nutritional environment.

Overall reproductive success partially recovers from a low-quality diet as it improves at adulthood. Surviving females fed a low-quality diet and switched to a high-quality diet at adulthood showed patterns of egg production akin to those females that were fed high quality diets throughout their lifetime.

I show that the nutritional environment is clearly important to female reproduction

(Chapters 2 – 4). Yet, sexually reproducing animals need gametes from both females and males to reproduce. If females cannot obtain any or enough ejaculate from males, their overall reproductive output may be compromised. Thus, I also examined factors that affected male reproductive output in this dissertation.

Animals, specifically herbivores, must cope with more than just nutritional changes to their habitats over time. Predatory pressures can also vary across space and time. In Chapter 5, I examined how a natural anti-predatory behavior, called autotomy, can influence pre- and post-copulatory male mating and fertilization success with multiple females. N. femorata can perform autotomy behavior, dropping a limb, to escape predation or entrapment. Males that exert this behavior with one of their

16

enlarged hind legs during development grow larger testes when they become sexually mature adults (Joseph et al. 2018, Miller et al. 2019). Unfortunately, losing one hind leg due to autotomy produces an incomplete weapon (i.e. a damaged weapon) that makes winning male-male contests challenging (Emberts et al. 2018). Males can still gain mating opportunities without an intact weapon (Emberts et al. 2018), but we do not know the reproductive consequences for males that are weapon damaged. I found that weapon-damaged males produce more offspring with larger females than intact males in non-competitive mating conditions. Additionally, large weapon-damaged males are less likely to mate with multiple females compared to large intact males. I discuss the broader implications of weapon-damage in Chapter 5.

Together, my experiments in this dissertation help elucidate how resources are divided among important reproductive traits and enhance our understanding of the environmental factors that affect vital traits for reproduction and fertilization success.

These experiments have also provided us with insight into how populations can cope with changing environments.

The value of using natural diets is immense, but it does come with some challenges for the experimenter. In this case, diet treatments cannot be easily categorized. In Chapters 2, 3, and 4 my goal was to manipulate the natural diet, and to do so I categorized fruit based on the month that the cactus was harvested. For clarity, I specify in each chapter exactly when in the season the insects were raised and the state of the cactus fruit. In some chapters, I reference the cactus fruit as “older” and

“newer” fruit (Chapters 2 – 3), referring to fruit collected the year before or the current year. In Chapter 4, I used early-season harvested cactus fruit. Therefore, I reference

17

this cactus fruit as “unripe,” collected early in the growing season (April – July 2018), and “ripe,” collected at the end of the previous year’s growing season (December –

January 2018). My research reveals the complexity and importance of seasonal diet to the reproductive ecology of N. femorata and provides powerful understanding into natural phenotypic variation that can help us understand natural effects of seasonal nutrition on female reproduction.

18

CHAPTER 2 SEASONAL NUTRITION AFFECTS FEMALE REPRODUCTIVE ANATOMY BUT NOT FEMALE MATING BEHAVIOR ACROSS A LONG BREEDING SEASON

Background

The nutrition available for animals is rarely consistent in nature. Nutritional changes have well-established effects on female reproduction across taxa including, for example, reptiles (Mugabo et al. 2010), mammals (Gardner et al. 2009, Cook et al.

2013, Gese et al. 2016), and insects (Wheeler 1996, Bauerfeind et al. 2007, Geister et al. 2008, Bong et al. 2014). When breeding seasons are long and span across multiple seasons (e.g. Mammals: Rosen 2009, Aves: Blomberg et al. 2013, Reptiles: Hall et al.

2018, Insects: Uyi et al. 2018), food quantity and quality can change. For example, many herbivores feed on plants that flower in the spring and change in quality by soon bearing fruit that will gradually ripen by the autumn. Animals that feed and breed on the same plant species must cope with the concurrent changes in nutrition due to the cascade of the plant’s phenological changes. These animal populations can provide excellent insights on the extent to which life history traits can change and the shifts in allocation that can occur with dietary changes and other seasonal factors.

Plant phenology co-occurs with changes in temperature, photoperiod, and other abiotic factors. Females can use abiotic cues to adjust reproductive investment even with consistent nutrition (Tauber et al. 1986, Ultsch 1989, Paul et al. 2008, McAllan and

Geiser 2014). Thus, it is particularly informative to study the effects of nutrition on trait allocation in a context where abiotic factors also fluctuate over time and with the season.

I studied the leaf-footed cactus bug, Narnia femorata Stål (Hemiptera: Coreidae), a species that breeds over a long reproductive season while its host plant experiences

19

major phenological changes. I used this opportunity to experimentally study changes in reproductive trait allocation and expression related to dietary change. Further, I kept the adult insects in a greenhouse, which resulted in some adjustment of photoperiod, temperature, and other abiotic factors while also letting ambient, seasonal conditions shine through.

N. femorata is a hemimetabolous species with distinct juvenile and adult stages.

Juveniles develop into adults in about nine weeks (Vessels et al. 2013) and adults likely survive several weeks upon adult eclosion in the wild (Cirino and Miller 2017). Multiple generations of N. femorata occur in North Central Florida from spring through autumn

(Cirino and Miller 2017). Juveniles and adults live together on prickly-pear cactus

(Opuntia mesacantha ssp. lata) and feed primarily on the fruits (Figure 2-1A – B).

Prickly-pear fruit nutrient composition changes seasonally (as seen in a related species

Barbera et al. 1992). The cactus fruits become abundant in the spring and mature through the summer and autumn months (Cirino and Miller 2017, Figure 2-2). The amount of fruit on pads declines in autumn (Cirino and Miller 2017, Figure 2-2) due to fruit decomposition (L.A. Cirino, personal observation) and competitive herbivory on the cactus fruit (Janzen 1986). Juvenile N. femorata may hatch early in the season and develop on newly formed cactus fruit or on cactus fruit that ripens as the season progresses. Juveniles can also hatch and develop towards the end of the breeding season when cactus fruit is older. If juvenile N. femorata hatch late in the breeding season, they are more at risk to hatching on a cactus patch that will lose fruit to herbivory by other animals or decomposition. Juveniles do not have fully formed wings during these young life stages which restricts their foraging to the cactus patch in which

20

they hatch and nearby cactus patches. Winged adults are more mobile and can fly to fruited cactus patches and feed. Therefore, both cactus fruit quality and quantity can change for N. femorata throughout the long breeding season. Nutrition is extremely important to trait expression in N. femorata. Males that are fed newly formed cactus fruit can grow large, but have tiny testes for their body size (Sasson et al. 2016) compared to their older cactus fruit fed counterparts. Males that are fed a cactus pad without any fruit are small, but they grow larger testes relative to their body size (Sasson et al. 2016). As the breeding season progresses, body size increases and female readiness to mate increases (Figure 2-1C – D, Cirino and Miller 2017). Further, females prefer males that were raised on older fruit cactus (Addesso et al. 2014, Gillespie et al. 2014; L.A. Cirino, unpublished data). Thus, behavior and anatomy have been demonstrated to be condition-dependent in N. femorata making this species well suited for this study.

My goal was to examine how plant phenology and the time in the breeding season affected female mating behavior and anatomy. To do this, I first examined how seasonal diet and seasonal timing affected female behavior and anatomy. Then, I investigated how a seasonal switch in host plant diet affected these traits. I tested these objectives across a long breeding season and at three distinct time points in the adult stage: the first week that females were in their adult form and the second and third week when females became sexually mature. I raised juveniles on three cactus-pad diets: newer fruit, older fruit, and cactus pad without fruit (i.e. restricted; Table 2-1). These diets represent females raised during the spring, autumn, or late autumn months, respectively (Figure 2-2). A subset of females raised on newer fruit and the restricted diet were switched to older fruit at adulthood to simulate a seasonal switch in the late

21

summer (i.e. fruit switch) or late autumn (i.e. restricted switch; Table 2-1, Figure 2-2). I measured female receptivity, body mass, ovary mass, and noted whether or not females had eggs present within their reproductive tract (i.e. oviducts) during the first, second, and third week of their adult lives. I predicted that females fed a diet with fruit would have larger ovary size, body size, and more likely to have eggs present in their oviducts when compared to the females that had experienced food restriction. I predicted that female mating behavior and reproductive anatomy would be correlated as is common in some heteropterans species (Adams 2000, Castañé et al. 2007, Oku et al. 2010). I also expected female reproductive anatomy to decrease in size towards the end of the breeding season as many other insect species do when they enter a period of dormancy under unfavorable environmental conditions (Tauber et al. 1986). Finally, I tested whether female receptivity and anatomy would be permanently affected by a poor juvenile diet or could be improved by a higher quality adult diet.

Methods

Insect Rearing and Diet Manipulations

I collected all cactus pads and wild-caught parental pairs for this experiment from

Starke, Florida (29.9804° N, 81.9848° W). I collected cactus pads bearing older fruit in

December 2017 and January 2018. I collected cactus pads bearing newer fruit pads and cactus pads without any fruit in the months of March through September 2018. I individually planted all cactus pads in deli cups with top soil, kept in a greenhouse, and checked daily for signs of deterioration.

I created parental pairs using a combination of wild-caught (N=88) and laboratory-reared (N=32) insects, kept in deli cups with a cactus pad with older fruit. N. femorata eggs have been observed in the wild on the underside of cactus spines, on

22

pine needles that have fallen on cactus patches from long-leaf pine trees, and on the cactus pads (L.A. Cirino, personal observation). Thus, for ease of egg extraction, a pine needle was used as an oviposition substrate. I checked parental pair cups for eggs daily and moved egg clutches immediately into separate deli cups with cactus pads with fruit for hatching. Group size is important for early juvenile survival (Allen and Miller 2020), so I kept 1st-3rd instars in groups of 5-12 individuals. This group size also reflects common group sizes found in nature (Cirino and Miller 2017; L.A. Cirino, personal observation). These three developmental stages, together, last approximately 5 weeks

(Vessels et al. 2013), and I kept these groups of young offspring (1st-3rd instars), along with the parental pairs, in temperature-controlled incubators (Percival Scientific; www.percival-scientific.com; 14:10 L:D and 26ᵒC).

Ovaries drastically increase in size starting in the fourth instar of juvenile development in hemipterans (Wick and Bonhag 1955), so I manipulated diet at this stage. I isolated juveniles into their own deli cup with a potted cactus pad and randomly assigned them to a juvenile diet: newer fruit, older fruit, or just the cactus pad without fruit (i.e. restricted diet; Table 2-1). Then, I moved the 4th instars to a greenhouse for the rest of their development (14:10 L:D and 21-32ᵒC). Previous research has shown that N. femorata juveniles take about 2 weeks to develop per instar in each later instar (4th and

5th instar) when fed a high-quality diet in a 25ᵒC rearing environment (Vessels et al.

2013). However, a recent study shows that development time changes based on diet

(Chapter 4). Thus, I checked all individually housed juveniles for survival, adult eclosion, and for signs of deterioration in cactus pads daily. I changed cactus pads if they showed signs of decay. I determined the sex of N. femorata at adulthood, and I randomly

23

assigned females to either remain on their juvenile diet (newer fruit N=69 or restricted diet N=30; Table 2-1; Figure 2-3) or switched to the older cactus fruit diet (fruit switch

N=41, restricted switch N=13; Table 2-1; Figure 2-3). If females were raised on older cactus fruit diets as juveniles, they remained on this diet as adults (N=134, Table 2-1;

Figure 2-3). All adult females remained in the same greenhouse as the 4th and 5th instar juveniles until they were entered into mating behavior trials.

I placed females into mating behavior trials then euthanized them for anatomical assays at weeks 1, 2, and 3 post adult eclosion (Figure 2-3). I chose these time points

(i.e. female age) to capture possible correlations between female mating behavior and anatomy in each treatment. I chose the first week of adult life (days 1-6 from adult eclosion) to examine female reproductive traits when females have just developed into their adult form. Therefore, I only examined the effects of the three juvenile diets on female reproductive traits at this time point (restricted N=13, newer fruit N=21, and older fruit diets N=40; Table 2-1). I chose both 2- and 3-weeks (days 12-16 and 19-24 from adult eclosion, respectively) because these are the time points between which most insects in this species are sexually mature (C.W. Miller, unpublished data). I continued to randomly assign females to these diet and age treatments across a long breeding season from the beginning of June through November 2018. The range of average daily ambient temperatures early in the breeding period was 6.78ᵒC (Lusher et al. 2020). After

October 12, 2018 and through November, the temperature fluctuated more drastically with a temperature range of 21.94ᵒC (Lusher et al. 2020). The amount of natural light dropped below 12 hours after September 27, 2018 (Thorsen 2020), although the greenhouse artificial fluorescent lights (Sylvania Octron/Eco 3500K 32W) remained on a

24

14:10 light/dark cycle. To account for these seasonal abiotic changes, I included the date that females were paired with males in the behavior trials (Mating Behavior Trials) in the analyses described below (Statistical Analysis). This date was when females were brought indoors for mating trials (Mating Behavior Trials) from the greenhouse and then euthanized. Thus, females experienced the greenhouse conditions and seasonal abiotic conditions as described above prior to this date.

Mating Behavior Trials

Mating behavior trials occurred during the week that the females were scheduled to be euthanized (week 1, 2, or 3 post adult eclosion). I entered mating pairs into ‘no choice’ mating behavior trials where one female can either accept one male’s mating attempt (e.g. receptive) or refuse (e.g. not receptive). These mating scenarios are ecologically relevant since one female can often encounter one male on a cactus pad in the wild (L.A. Cirino, personal observation). I raised all male partners on older fruit cactus pads and randomly assigned unrelated males to females. I placed males in the females’ cactus pad deli cups. I observed up to 15 pairs simultaneously and whether or not a female accepted a mating was noted during a 3-hour behavior trial window. This species is well-suited for concurrent behavioral observations because N. femorata have long periods of stasis and periods of slow ritualistic behavior easy to observe.

Specifically, mating behavior begins when the male approaches a female, dorsoventrally mounts the female, tilts the body to make genital contact (terminal abdominal segments touch), and then proceeds with intromission (Figure 2-1C – D). I marked a female as receptive when the male turned at least 90ᵒ from her body into a mating position (Figure 2-1D). Once this occurred, observations for that pair ended.

Unreceptive females walked or ran from males, kicked males away, or prevented males

25

from mating by closing their genital plates. I entered females into a new behavior trial, up to three times during the same week, if males did not attempt to mate with a female during the 3-hour mating trial. Females were excluded from analysis if no mating attempt occurred in three behavior trial attempts.

Female Anatomy

I euthanized females at the conclusion of the 3-hour behavior trials for anatomical assays. Hemipteran females can produce oocytes in the ovaries regardless of mating status (Gordon and Bandal 1967, Adams 2000) and store eggs in the oviducts awaiting fertilization, so I dissected all females and counted the fully formed eggs in the oviducts (Figure 2-4). I then removed the ovaries from the body and placed the ovaries and the bodies, without the oviduct eggs or the ovaries, in microtubes with 70% EtOH. I kept these tissues refrigerated until it was time to be weighed. I constructed and weighed foil boats, to the nearest microgram, using a microbalance (Mettler Toledo

XP6: Columbus, OH, USA). I then placed female bodies and ovaries into the foil boats and dried them in a drying oven at 60ᵒC for 72 hours. I weighed the foil boat with the tissue sample inside at the end of the 72-hour period to obtain the total dry weight. I calculated the mass of the ovaries and bodies by subtracting the foil weight from the total dry weight (foil plus body or ovaries).

Statistical Analyses

Juvenile diet – week 1 females

My first goal was to examine how juvenile diet and time in breeding season influenced female mating behavior and anatomy (Figure 2-3). To do this, I constructed three models for analysis: one binomial (logit link) generalized linear model (GLM) and two linear models (LMs) with juvenile diet (newer fruit, older fruit, and restricted diets;

26

Table 2-1) and Julian date (i.e. date when females were paired with males) as my explanatory variables. (1) Female receptivity (Y/N), (2) female body mass, and (3) ovary mass were included as three separate response variables. My sample size for this analysis was lower because males had to mount females in order to ascertain female receptivity (newer fruit N=10, older fruit N=13, restricted N=4). Ovary mass had to be log transformed to meet the assumption of normality for the LMs. The two-way interaction between juvenile diet and season was originally included in all three models, but was removed because it was well-beyond the level of statistical significance (p>0.15, Gotelli and Ellison 2013).

Adult diet switch and age – week 2 and 3 females

My next goal was to investigate the effects of switching females to a different diet upon adulthood, age, and time in breeding season on female mating behavior and anatomy (Figure 2-3). I constructed a GLM assuming a binomial distribution and logit link function to analyze female receptivity. Since I could only use females that were first mounted by males to assess female receptivity, my sample sizes were lower for this analysis (newer fruit N=41, fruit switch N=35, older fruit N=80, restricted switch N=4, and restricted N=8). Diet (Table 2-1), age (week 2 and 3), and Julian date were the explanatory variables and female receptivity (Y/N) was the response variable.

Finally, I constructed two separate LMs and one GLM (binomial distribution and logit link function) to investigate female anatomy. I again included diet, age, and Julian date as our explanatory variables and (1) female body mass, (2) ovary mass, and (3) the presence of oviduct eggs (Y/N) were the three separate response variables. Female body mass once again met the assumption of normality for LMs and ovary mass was log transformed to meet this assumption. All two-way interactions were originally

27

included in all of the models but were sequentially removed (stepwise reduction) since they were well-beyond the level of statistical significance (p>0.15). The two-way interaction between Julian date and age was kept in the linear model for ovary mass because it did not meet this standard, although it was still not statistically significant

(Table 2-3). All analyses in this manuscript were performed in R v3.6.2 (R Core Team

2019).

Results

Juvenile Diet – Week 1 Females

I found that none of the females fed the restricted diet as juveniles (0 out of 4) accepted a mating attempt during the first week of adult life; whereas, 7 out of 10 (70%) of the females raised on the newer fruit and 5 out of 13 (38.5%) of the females raised on older fruit diets did accept a mating attempt. When I compared females raised on the older versus newer fruits, I did not find any differences in female receptivity due to juvenile diet nor time in breeding season (Table 2-2). Females fed cactus fruit of any kind as juveniles had greater body mass (linear model F=11.132, df=2, p≤0.001) and ovary mass (linear model F=5.770, df=2, p=0.005) than those that were diet restricted during the juvenile stage (Table 2-2; Figure 2-5). No females produced fully formed eggs in their oviducts in the first week of adult life. However, some females had late stage oocytes in their ovaries (12/74 or 16.2%). These females were fed fruit as juveniles (i.e. older fruit or newer fruit). Females in the fruit-fed juvenile diet groups produced an average of one late stage oocyte in the ovary (Figure 2-4).

Adult Diet Switch and Age – Week 2 and 3 Females

Of those females that were mounted by males, week 3 females were more likely to accept a mating attempt (70/86 or 81.4%; Wald χ2=5.483, df=1, p=0.019) and females

28

that were paired with males earlier in the breeding season were also more likely to accept a mating attempt (Wald χ2=4.234, df=1, p=0.040; Table 2-3, Figure 2-6).

Females that were fed a lifetime of any fruit type had the largest body mass (linear model F=20.898, df=4, p≤0.001) and ovary mass (linear model, F=12.440, df=4, p≤0.001) compared to females that experienced a lifetime of food restriction (Table 2-3,

Figure 2-7). Those females that were fed the restricted switch diet were able to regain body mass, ovary mass, and start producing fully formed eggs that they stored in the oviduct compared to the restricted diet females (Table 2-3, Figure 2-7). Food restricted females never had eggs present in their oviducts (Wald χ2=12.000, df=4, p=0.017; Table

2-3, Figure 2-7). Finally, females euthanized at the end of the breeding season also had smaller ovaries (linear model F=4.575, df=1, p=0.034) and were less likely to have fully formed eggs in their oviducts (Wald χ2=21.058, df=1, p≤0.001; Table 2-3, Figure 2-7).

Discussion

Narnia femorata feed on a dynamic diet. Cactus fruit ripens from the spring through autumn and multiple generations must develop on this changing resource.

Further, the cactus fruits can be removed from the plant, leaving juveniles stranded without fruit for part or all of their development (Figure 2-2). Previous studies have shown that these changes profoundly affect insect phenotypes (Addesso et al. 2012,

Gillespie et al. 2014, Sasson et al. 2016, Miller et al. 2016, Cirino and Miller 2017). My goal in this study was to understand how these seasonal changes in diet as well as the time in breeding season affected the mating behavior and anatomy of N. femorata at important time points in adult life. I found that females that were food restricted (i.e. only had a cactus pad to feed on without any fruit) during the juvenile stage that were then

29

switched to a well-fed diet (i.e. cactus pad with older fruit) at adulthood experienced partial reproductive recovery compared to females that continued to be food restricted during the adult stage. I also found that older females (3 weeks post adult eclosion) were more likely to accept a male mating attempt than younger females (2 weeks post adult eclosion). These results suggest that females can reproductively recover from juvenile nutritional restriction, mate, and potentially produce offspring. However, if females became adults later in the breeding season when temperature fluctuated more drastically, females grew smaller ovaries and were less likely to have fully formed eggs in their oviducts compared to those females that became adults earlier in the breeding season. Therefore, reproductive compensatory growth can occur if females get a poor nutritional start in life; however, abiotic seasonal cues may prevent females from fully realizing this reproductive recovery.

As I predicted, late-breeding females, regardless of diet, had smaller ovaries and were less likely to have fully formed eggs present in the reproductive tract (i.e. oviducts) than those early-emerging females. Challenging dietary conditions often occur at the end of a breeding season when seasons start to change from good to poor. However, changes in photoperiod and/or temperature are more reliable indicators to animals that seasons are changing and more challenging conditions are to come. Many insects

(Tauber et al. 1986) and vertebrate taxa (Ultsch 1989, Paul et al. 2008, McAllan and

Geiser 2014) use these cues to regulate reproduction and prepare their bodies for the winter. My results are consistent with this research as time in breeding season (i.e.

Julian date) had no effect on female body size but female ovary mass and the probability of having fully formed eggs in the oviducts declined among late-breeding

30

females compared to those that bred early in the breeding season. My findings suggest that females may be shutting down their reproductive machinery in preparation for overwintering (e.g. quiescence) at the end of the breeding season. Females also appear to alter their behavior at the end of the breeding season and are less likely to accept a mating at this time. However, some females still accept mates, which means they may be storing sperm over the winter that can be used once they exit this dormant phase and start reproducing in the spring (Roth and Reinhardt 2003, Golec and Hu 2015).

Surprisingly, I found that female age and time in season, but not seasonal diet, influenced the likelihood that females would accept a mating attempt. Females in this study do not appear to adjust mating decisions based on the diet they have consumed and this result is consistent with other insect research (Barrett et al. 2009b, Wilner et al.

2020). Rather, females become more willing to mate when they get older. Females across taxa are known to be more receptive and less choosy as they age (Gray 1999,

Kodric-Brown and Nicoletto 2001, Moore and Moore 2001, Richard et al. 2005, Anjos-

Duarte et al. 2011), and my results are consistent with this research. I also did not detect differences in female mating behavior when females were fed different quality diets (i.e. older fruit, switched fruit, or newer fruit diets) as found in another experiment

(Wilder and Rypstra 2008). While I show compelling evidence for age effects on mating decisions for those females fed different quality diets (i.e. older cactus fruit versus newer cactus fruit), mating behavior patterns are less clear between females fed different quantity diets (i.e. restricted and restricted switch versus all other diets). Due to a combination of low female survival and low male mating effort, I was unable to capture clear patterns of female mating behavior under the most challenging nutritional

31

conditions and this should be examined in the future. Many studies have found condition-dependent responses in female mating behavior (Hingle et al. 2001, Hunt et al. 2005, Fisher and Rosenthal 2006, Judge et al. 2014, Kunz and Uhl 2015,

Richardson and Smiseth 2019), and I would expect this same result under the most nutritionally-stressed environments (i.e. cactus pads without fruit). My low sample sizes for the more restricted diets hint that male mating effort may be affected by female nutritional stress as found in other research (Barry and Wilder 2013) and males may be less willing to attempt a mating with females that are food restricted. However, male mating decisions may be undeterred if females only experience a brief period of food restriction (Wilner et al. 2020) and are able to obtain a superior diet upon reaching adulthood as our females fed the restricted switch diet experienced. Since females are able to partially reproductively recover from a food restricted period, the potential fitness consequences are less dire and males may still mate.

The mating environment in which females encounter males may affect the likelihood that females will accept a mate (Jennions and Petrie 1997). Female mating behaviors are known to vary under different encounter environments in N. femorata

(Gillespie et al. 2014). I tested N. femorata female mating behavior in the cactus environment which females were fed as adults (i.e. adult cactus diet). The mating environments in our study may have shaped female mating decisions since females fed a high-quality diet are known to be more likely to accept mating attempts in low-quality environments versus high-quality environments (Gillespie et al. 2014). It is entirely possible that females may face mating decisions under different environmental conditions from those conditions that they were raised under since the juvenile stage of

32

N. femorata lasts about nine weeks, on average (Vessels et al. 2013), and cactus fruit ripens over that time (Cirino and Miller 2017, Figure 2-2). Therefore, future research should examine how the encounter environment may interact with female diet when females make mating decisions. These studies will help us understand how dynamic environments play a role in female receptivity which may influence female reproductive output and overall fitness.

Similar to female mating decisions, I unexpectedly did not detect nutritional effects on female reproductive anatomy among females fed different seasonal fruit. I found that N. femorata females raised on newly harvested cactus fruit were just as large and had similar sized ovaries to those females raised on older fruit and the switched fruit diet. Female reproductive anatomy was only affected by nutrition when females experienced at least one period of food restriction (i.e. restricted or restricted switch diets). These results suggest that having any type of cactus fruit, in addition to the cactus pad (i.e. diet quantity), is crucial to female anatomy and that cactus fruit quality may not be as essential to the expression of these female reproductive traits. Previous research has shown that N. femorata males raised on newly harvested cactus fruit had small testes relative to their body size (Sasson et al. 2016). One reason for the conflicting results might be related to the season with which these studies were conducted. Sasson et al. (2016) manipulated diet early in the growing season when the newly harvested cactus fruit was still unripe. In my study, diet was manipulated using newly harvested cactus from the early to mid-growing season (i.e. March through

September) and females were fed these diets across a long experimental period (i.e. 7- months). During this long season, nutritional quality likely changed in the harvested

33

cactus as the early fruits ripened (e.g. Barbera et al. 1992). Thus, instead of providing females with newly harvested unripe fruit as the Sasson et al. 2016 study did, the newly harvested cactus I used may have become ripe over the experimental period. This change in quality may have allowed females to access important nutrients or more adequate proportions of nutrients that they could use to allocate to somatic maintenance and storage as well as reproduction.

I was also surprised that I did not detect a correlation between behavior and anatomy due to diet as I had predicted since mating behavior and ovarian development for some heteropterans is highly correlated (Adams 2000, Castañé et al. 2007, Oku et al. 2010). This result might illustrate the consequences some males may face when choosing female mates. If males mate with females that are not reproductively ready

(i.e. reproductive machinery is shut down due to a restricted diet), they may not gain the immediate fitness benefits of mating. Females that cannot produce eggs may be unable to produce many offspring until females obtain the nutritional resources required to upregulate reproductive function (Wittmeyer et al. 2001, Barrett et al. 2009a, Plesnar-

Bielak et al. 2017) or exit hibernation (Kott et al. 2000, Akiyama et al. 2011). In both of these scenarios, male reproductive success may be reduced due to the amount of time females spend reproductively inactive. Males in these scenarios are especially disadvantaged since mating is more costly for males than originally theorized (Tang-

Martinez 2016).

Animals must survive and reproduce in complex and seasonally changing environments. I found that female N. femorata are able to partially reproductively recover from a nutritionally poor early life when they access higher quality adult

34

nutrition. However, other seasonal abiotic cues, such as temperature, may prohibit females from attaining partial reproductive compensation. This pattern suggests that both host plant phenology and the time in the breeding season promote the population fluctuations we see in nature for N. femorata (Cirino and Miller 2017). Fruit is abundant in the spring and summer months; however, cactus fruit rapidly declines in the autumn months due to competitive herbivory (Janzen 1986) and fruit decomposition (L.A. Cirino, personal observation). Additionally, abiotic factors, such as temperature, fluctuates more drastically in the autumn months and can act as an important signal for females to shut down reproduction and prepare for the winter months (Insects: Tauber et al. 1986,

Vertebrates: Ultsch 1989, Paul et al. 2008, McAllan and Geiser 2014). Scarce food resources exposes juveniles to nutritionally-stressed environments in which they must grow and develop. Increased foraging capabilities in the adult stage (i.e. winged adults) can help females access nutrient-rich food and facilitate reproductive recovery which may help them regain some reproductive output. However, even if females are able to recover from a nutritionally poor early start in life, other abiotic factors such as temperature may prohibit them from being reproductively successful. Therefore, females subject to these difficult nutritional conditions may still be able to produce offspring and increase fitness through adult nutrition, but this is only more likely to occur early in the breeding season.

35

Figure 2-1. Narnia femorata feeding and mating. (A) Juveniles and (B) adults feeding on newer cactus fruit. (C) Male mounting a female and (D) a female accepting the mating attempt. Photo A courtesy of author. Photos B-D courtesy of Dr. Christine Miller.

36

Figure 2-2. Opuntia mesacantha ssp. lata fruit abundance and quality in 2014-2015. Data was collected from 20 cactus patches at Ordway-Swisher Biological Field Station in North Central Florida. Cactus flowers and green colored fruit were classified as “newer” fruit and red colored fruit was classified as “older” fruit. Figure modified with permission from Cirino and Miller 2017.

Table 2-1. Diet manipulations for both juvenile and adult Narnia femorata. Treatment terms are used throughout the manuscript to indicate the diets that females were fed. Sample sizes for anatomical analyses by week are also listed here. Sample Sample Sample Juvenile diet of a Adult diet of a Treatment Size: Size: Size: cactus pad with cactus pad with Term Week 1 Week 2 Week 3 No fruit No fruit Restricted 13 9 8 Newer fruit Newer fruit Newer fruit 21 22 26 Newer fruit Older fruit Fruit Switch N/A 19 22 Older fruit Older fruit Older fruit 40 46 48 No fruit Older fruit Restricted Switch N/A 5 8

37

Figure 2-3. Flow chart of treatments and analyses. The gray arrow represents female Narnia femorata life stages. Three types of cactus diets were provided to late stage juveniles. Subsets of females raised on restricted and new fruit diets were switched to older fruit diets (i.e. fruit switch and restricted switch) making five total diets for analysis at weeks 2 and 3. Analyses were split into two groups: juvenile diet and lifetime diet manipulation. Response variables are numbered in the order in which they are described in the main text.

Table 2-2. Analyses examining the behavior and anatomy of week 1 females. Results from one generalized linear model (GLM; mating behavior) and two separate linear models (LMs; anatomy) that examine how juvenile diet affects female mating behavior and anatomy. Mating Behavior Anatomy Receptivity Body Ovary Source df (Y/N), Mass, Mass, Wald χ 2 F-value F-value Juvenile Diet 2 2.148 11.132*** 5.770** Julian Date 1 1.039 0.036 0.821 *p≤0.05, **p≤0.01, ***p≤0.001

38

Table 2-3. Analyses examining the behavior and anatomy of sexually mature Narnia femorata. Results from two separate generalized linear models‡ (GLMs) and two separate linear models† (LMs) that examine how diet, age, and time in breeding season (i.e. Julian date) affect female mating behavior and anatomy. Mating Behavior Anatomy Receptivity Body Ovary Oviduct Eggs Source df (Y/N) ‡, Mass†, Mass†, (Y/N) ‡, Wald χ2 F-value F-value Wald χ 2 Diet (juvenile + adult) 4 4.192 20.898*** 12.440*** 12.000* Age 1 5.483* 2.087 2.784 0.301 Julian Date 1 4.234* 1.607 4.575* 21.058*** Age * Julian Date 1 N/A N/A 2.767 N/A *p≤0.05, **p≤0.01, ***p≤0.001

Figure 2-4. Narnia femorata females dissected from each diet treatment. Narnia femorata in the (A) restricted, (B) restricted switch, (C) newer fruit, (D) fruit switched, and (E) older fruit diets. Ovaries (OV), late stage oocytes (LSO), and oviduct eggs (OE) are labeled above. All females pictured were in the 3- week age group. Photos courtesy of the author.

39

Figure 2-5. Anatomy of newly eclosed adult females is affected by diet. Feeding on a cactus fruit diet of any kind as juveniles increases (A) body mass (linear model F=11.132, df=2, p≤0.001) and (B) ovary mass (linear model F=5.770, df=2, p=0.005) during the first week of adulthood compared to their restricted diet counterparts (restricted N=13, older fruit N=40, newer fruit N=21).

Figure 2-6. Female mating behavior of sexually mature Narnia femorata. Of those females that were mounted by males, (A) females that were paired earlier in the in breeding season were more likely to accept a mate (Wald χ2=4.234, df=1, p=0.040). (B) They were also more likely to accept a mate when they were older (3 weeks; Wald χ2=5.483, df=1, p=0.019). Sample sizes for females that were 2 weeks old N=82 and 3 weeks old N=86.

40

Figure 2-7. Anatomy of sexually mature (week 2 and 3) adult females. Feeding on a lifetime of cactus fruit of any kind increases (B) body mass (linear model F=20.898, df=4, p≤0.001), (D) ovary mass (linear model F=12.440, df=4, p≤0.001), and (F) the likelihood that females will have eggs present in their oviduct (Wald χ2=12.000, df=4, p=0.017) compared to females that have a period of food restriction (restricted N=17, restricted switch N=13, older fruit N=94, fruit switch N=41, newer fruit N=48). Time in breeding season has no effect on (A) body mass, but both reproductive measures (i.e. (C) ovary mass and (E) oviduct eggs) of female anatomy declined later in the breeding season (Ovary mass: LM, F=4.575, df=1, p=0.034; Oviduct eggs: Wald χ2=21.058, df=1, p≤0.001).

41

CHAPTER 3 MATERNAL BODY SIZE, BUT NOT DIET, AFFECTS EGG SIZE IN THE LEAF- FOOTED CACTUS BUG, NARNIA FEMORATA

Background

Maternal effects occur when offspring phenotype is influenced by the maternal phenotype (Bernardo 1996). Mothers can influence their offspring’s phenotype by adjusting the amount of yolk provided to their eggs, choosing a suitable oviposition site, and providing maternal care (Mousseau and Dingle 1991, Bernardo 1996). Maternal effects are important because they are likely ubiquitous in nature and can have fitness consequences for offspring (Fox and Czesak 2000).

Egg size is one of the primary ways in which insect mothers can influence their offspring (Mousseau and Dingle 1991). Provisioning eggs with more resources can provide offspring with an early life advantage and this is particularly helpful when females do not provide any type of maternal care (e.g. Mas and Kolliker 2008). When females increase egg size, they help promote the growth and development of the embryo before hatching and will help their offspring survive the early and delicate stages of life (Fox and Czesak 2000). Females that lay large eggs help boost juvenile development, increase juvenile survival and, in some cases, help juveniles become larger adults (Reavey 1992, Fox and Dingle 1994, Fox and Mousseau 1996), making egg size an extremely important fitness trait.

Egg size can be influenced by maternal traits such as body size and diet in insects (Fox and Czesak 2000). Body size is correlated with egg size across many insect taxa (Larsson 1989, McLain and Mallard 1991, Fox and Czesak 2000, Kudo

2001). Larger mothers typically produce larger eggs than their smaller counterparts (Hill and Pierce 1989, Larsson 1989, McLain and Mallard 1991, Fox and Czesak 2000, Kudo

42

2001, Kojima 2015, Yanagi and Tuda 2012); in some cases because their reproductive tract is larger (Yanagi and Tuda 2012). Large female body size is often related to a high-quality diet in insects (Nylin and Gotthard 1998); therefore, diet can indirectly affect egg size. However, female diet can also directly influence egg size through the nutrient reserves mothers have available to draw upon when producing eggs (Boggs 1997,

Fischer et al. 2004, Fox and Czesak 2000, O'Brien et al. 2004, Bauerfeind and Fischer

2005). Females fed high-quality or more food lay larger eggs than those females that are underfed or are provided low-quality food (Boggs 1997, Fischer et al. 2004, Fox and

Czesak 2000, O'Brien et al. 2004, Bauerfeind and Fischer 2005, but see Kojima 2015).

Much of the work examining the effects of maternal traits on egg size have been carried out in insects (Fox and Czesak 2000, Hemiptera: Larsson 1989, McLain and Mallard

1991, Kudo 2001, Newcombe et al. 2015, Sisterson et al. 2018, Coleoptera: Fox and

Dingle 1994, Kyneb and Toft 2006, Kojima 2015, Lepidoptera: Bauerfeind et al. 2007,

Hill and Pierce 1989) that have offspring that feed throughout their juvenile lives (Insect herbivores: Hochuli 2001). Yet, some insect species have a non-feeding first instar stage of development that likely rely heavily on egg resources to sustain them

(Hemiptera: Bowling 1980, Todd 1989, Vessels et al. 2013). It is important that we investigate the influence of maternal body size and diet on egg size in species that have an early non-feeding stage in development since they are likely to be heavily dependent on maternally provisioned egg resources during this stage.

The leaf-footed cactus bug, Narnia femorata Stål (Heteroptera: Coreidae), is well suited to investigate how maternal diet and body size affect egg size because they feed on a diet that varies in quality across time (Cirino and Miller 2017) and body size varies

43

based on diet which is only marginally heritable in this species (Miller et al. 2016). First instar juveniles live in a non-feeding stage of development for about 5 days after hatching (Vessels et al. 2013). These juveniles are presumably entirely reliant on the resources they received from the egg to survive this instar, making egg size an important maternal effect for offspring in this species. Beyond the first stage of development, juveniles and adults feed on the pads and fruit of prickly-pear cactus plants, Opuntia mesacantha spp. lata (Small) Majure 2014 (Cactaceae), in North

Central Florida. The cactus fruit quality changes seasonally where newly formed fruit is available early, starts to mature mid-season, and becomes fully mature later in the season (Cirino and Miller 2017, Figure 2-2). Since the juvenile stage lasts approximately nine weeks (Vessels et al. 2013), juveniles may grow and develop during the early season where cactus fruit is newly formed, later in the season when cactus fruit is older, or in the middle of the season where cactus fruit is ripening (Cirino and Miller 2017).

Thus, some N. femorata females may experience a switch in diet over their lifetime as newer fruit matures into older fruit. N. femorata males grow smaller reproductive organs when raised on a diet of newly formed fruit when compared to males raised on older fruit (Sasson et al. 2016). Females are less likely to produce eggs when raised on the newly formed fruit compared to females that are raised on the older fruit (Chapter 4); therefore, for the purposes of this study, we will refer to the newer fruit as a lower quality diet and the older fruit as a higher quality diet.

Previous work has revealed that body size in N. femorata females is related to the timing of egg laying (Wilner et al. 2020) but not the overall number of eggs laid

(Allen et al. 2018); whereas, female diet affects the number of eggs that females lay in a

44

lifetime (Chapter 4). Females that lay fewer eggs and are known to increase the size and resources allocated to those eggs (i.e. trade-off) across many animal taxa (Dani and Kodandaramaiah 2017). Therefore, N. femorata is an excellent system to examine the effects of body size and diet on egg size in a species that rely on female egg provisioning early in juvenile development.

Here, my aim was to examine the direct and indirect effects (via body size) of maternal diet on egg size in N. femorata. To do this, I raised juveniles through adulthood on their host plant diet that varied in fruit quality: newer and older fruit. I then maintained adult females on the same host plant diet (i.e. newer fruit females or older fruit females) or switched female host plant diet (i.e. switched fruit females) at adulthood to create three diet treatments that varied in fruit quality. I paired females with fruit- raised males, allowed females to lay eggs (Figure 3-1), and then measured egg volume.

I predicted that smaller females would produce smaller eggs since body size can be a morphological constraint on egg size (Bauerfeind and Fischer 2007, Yanagi and Tuda

2012) and previous studies have found this result in other species (Larsson 1989,

McLain and Mallard 1991, Fox and Czesak 2000, Kojima 2015). I also predicted that females fed an improved diet at adulthood would have smaller eggs than those females that were raised on a lifetime of older fruit since body size is highly influenced by juvenile diet and only marginally influenced by heritability in this species (Miller et al.

2016) and body size can constrain egg size (Bauerfeind and Fischer 2007, Yanagi and

Tuda 2012).

45

Methods

Insect Husbandry

I collected adult N. femorata (N=88) and prickly-pear cactus (Opuntia mesacantha ssp. lata) from Starke, Florida (29.9804° N, 81.9848° W). Cactus pads bearing older fruit were collected in December 2017 and January 2018 and potted in 32- ounce clear plastic deli cups (Fabri-Kal Alur RD32) using top soil and maintained in a greenhouse until we were ready to place bugs on them in the summer. I collected cactus pads bearing newer fruit from March through September 2018 and planted them using the same protocol as the winter-collected cactus above. I haphazardly paired wild- caught adults and laboratory-reared adult N. femorata (N=32) to generate 60 parental pairs that produced females for this experiment. I kept all parental pairs in the deli cups with the potted cactus pad and an older fruit from our winter harvest. N. femorata females are known to lay eggs on the underside of cactus spines, pine needles, and on the pads of the cactus in North Central Florida (L. A. Cirino, personal observation).

Therefore, a pine needle acted as the egg-laying substrate which could be easily removed and separated from the females’ deli cups. I checked parental pairs daily for eggs and isolated these eggs into their own deli cup with an older-fruit cactus pad to hatch. I kept juvenile 1st through 3rd instars in groups sizes from 5 to 12 individuals since group size is important to survival during these early juvenile stages (Allen and Miller

2020). Juveniles require about 5 weeks to develop through the first three instars of the juvenile stage (Vessels et al. 2013). In addition to the parental pairs, I kept these early- stage juveniles in temperature-controlled incubators (Percival Scientific; www.percival- scientific.com; 14:10 L:D and 26ᵒC) during this development time.

46

Insect Rearing

I isolated juveniles within 24 hours of becoming a 4th instar, individually housed them in their own deli cup, and randomly assigned them to a juvenile diet consisting of a cactus pad with a fruit (i.e. older or newer). Juveniles were isolated at this developmental stage since hemipteran ovary development begins to increase considerably at this time (Wick and Bonhag 1955). I then transferred these juveniles to a greenhouse (14:10 L/D) and checked them daily for adult eclosion and cactus pad condition (approximately 4 more weeks, Vessels et al. 2013). If a cactus pad showed signs of deterioration (e.g. brown spots, desiccation), I replaced it. I determined sex once insects eclosed into the adult form and randomly assigned females on the newer- fruit diet to either remain on their juvenile diet (i.e. newer fruit; N=24) or switched them to the older-fruit diet (i.e. switch fruit; N=18). All females that were fed the older-fruit diet as juveniles remained on this diet as adults (i.e. older fruit; N=27). Females included in this experiment were offspring from 29 of the original 60 parental pairs that we formed and were randomly assigned to the diet treatments described above. Therefore, some females came from the same parental pairs.

I kept adult females in the greenhouse until 14 days post-adult eclosion. At this time, I transferred females to a rearing room (14:10 L/D and 60% humidity) where I paired them with an older fruit-raised, non-sibling male. I kept most females with a male for four days with the exception of two pairs that were together for two days and one pair that was together for five days. Since these three pairs all experienced hatching success (indicating successful fertilization), I retained them in our analyses. I kept the mating pairs in the female’s deli cup that included the female’s adult diet and a pine needle for oviposition. Females that are 14 to 18 days post-adult eclosion are more

47

sexually receptive to a male mating partner than younger females regardless of female diet (Chapter 2). I then removed males from the female deli cups and monitored females daily for egg laying and survival. N. femorata females can lay unfertilized eggs and these eggs are usually laid in an unorganized fashion in groups of less than 5 eggs

(L.A. Cirino, personal observation). Therefore, I waited for each female to lay a clutch of

10 or more eggs to ensure I measured eggs that were likely fertilized. Once eggs were laid, I immediately removed them, photographed them using a digital camera (Canon

EOS 50D, Canon, Tokyo, Japan), and then placed them into their own deli cup for hatching. I chose to measure eggs laid early in the female reproductive period because egg size is known to decrease later in life among other insect species possibly eroding the effects that body size and diet might have (Fox and Czesak 2000). Females in this experiment either naturally expired or were euthanized in a -20ᵒC freezer. I then photographed their bodies, ventral side down, to procure linear pronotum width (PW) measurements (a proxy for body size) obtained using ImageJ software (v1.42d,

Abràmoff et al. 2004).

Egg Volume. I photographed the cuboid-shaped eggs that each female produced twice (Figure 3-2) to obtain three sides of the eggs for volume measurements

(length, width, and height). Prior research in rove beetles, Tachyporus hypnorum, demonstrated that egg volume was correlated with multiple offspring performance traits

(Kyneb and Toft 2006). Thus, I used egg volume here as my measurement of egg size.

I procured length, width, and height measurements of each egg per clutch using ImageJ software. I calculated volume (length x width x height) for each egg and used the average volume per egg for the following analyses.

48

Statistical Analyses

My primary objective was to examine direct and indirect (i.e. via body size) effects of maternal diet on egg volume for eggs produced early in adulthood. I focused on the first large clutch of eggs that females laid. My first step was to establish the role of juvenile diet in female adult body size in this experiment. I first constructed a linear model (LM) with pronotum width (PW; proxy for body size) as the response variable and juvenile diet (older and newer fruit) as the explanatory variable. Second, I constructed a generalized linear model (GLM) with eggs laid (Y/N) as the response variable and lifetime diet (older, newer, switched) and PW as the explanatory variables.

Many females (20/69 or 29%) in this experiment did not lay any eggs. Thus, I could only examine the effect of diet and body size on egg volume for females that did lay eggs. I constructed a linear mixed model (LMM) to address our original objective. Of those females that laid eggs (newer fruit: N=9, switched fruit: N=16; older fruit: N=24), lifetime diet and PW were included as explanatory variables and egg volume was the response variable. I included female ID as a random factor. I originally included the two- way interaction between lifetime diet and PW in both the GLM and the LMM to rule out the interaction between the two. It was removed since it was far above the level of statistical significance (p>0.15, Gotelli and Ellison 2013). All analyses in this manuscript were performed in R v3.6.2 (R Core Team 2019).

Results

Juveniles that were fed the newer fruit became smaller adult females compared to those that were fed older fruit (linear model F=6.802, df=1, p=0.011; Figure 3-3).

Further, body size and diet both separately affected whether or not females laid eggs.

Smaller females (Wald χ2=11.307, df=1, p=0.001) and females that were fed the newer

49

fruit (Wald χ2=10.095, df=2, p=0.006) were less likely to lay any eggs at all (Figure 3-4).

In fact, 15 out of 20 of females (75%) that did not produce any eggs were fed a lifetime of newer fruit and all 20 females were small (Figure 3-4). There were no direct effects of diet on egg volume of those females that did lay eggs (LMM, F=0.425, df=2, p=0.657;

Figure 3-5); however, there were indirect effects of diet (via body size) on egg volume.

Larger females produced larger volume eggs than smaller females (LMM, Wald

F=4.152, df=1, p=0.047; Figure 3-5).

Discussion

I found that diet had only indirect effects via maternal body size on the volume of eggs that mothers produced. My findings are consistent with the broader literature indicating that large females tend to lay larger eggs than small females within many insect species, suggesting that egg size is subject to morphological constraints

(Bauerfeind and Fischer 2007, Fox and Czesak 2000, Yanagi and Tuda 2012). My results show that juvenile and adult diet did not directly affect egg size; yet they did play a direct role on whether or not eggs were produced and an indirect role in egg size by mediating the size adult females reached. One potential explanation for the absence of a direct effect of diet on egg size is that there may be a size threshold for egg viability and/or survival through the non-feeding stage of juvenile development in this species.

The smallest females or those raised on the poorest diets may simply cease egg production rather than producing tiny eggs that result in tiny offspring (Fox and Czesak

2000) with a potentially small chance of surviving the initial non-feeding instar of juvenile development. All females that failed to produce any eggs were small and 75% of those females were raised on a lifetime diet of newer fruit. Of those females that laid eggs,

50

small females produced small eggs. This suggests that there is a body size threshold that females must attain to produce eggs (Hatle et al. 2006). Females that were able to accumulate enough resources from their diet during the juvenile stage, regardless of quality, to produce eggs were likely restricted in the size of eggs they could produce based on morphological constraints. Thus, female diet plays a role in body size which then, in turn, affects egg size.

Here, I have addressed one element of egg size – egg volume. Although egg volume measurements have been used in previous research to quantify egg size in insects (Kyneb and Toft 2006), this metric may not fully capture the amount of resources that females partitioned to eggs. Egg volume may reflect the size of the female reproductive tract, while mass may better reveal maternal investment to eggs and offspring. Eggs in this species are cuboid in shape and highly structured due to the formation of the chorion during egg development. These egg traits may make detecting a reduction in resources in eggs more difficult. Measuring egg mass may also help reveal provisioning differences between unfertilized and fertilized eggs given that females in this species can lay partial or full clutches of unfertilized eggs. Using egg mass to gauge maternal investment may reveal if females provision these eggs differently and ultimately help explain why some eggs fail to hatch.

This study provides an example of how maternal effects (i.e. egg size) vary directly by body size and indirectly by diet in a species with a non-feeding first instar of development. Interestingly, some marine invertebrates and reptiles that also endure a non-feeding stage early in development have larger eggs compared to closely related species that begin feeding immediately (Strathmann 1985, Nagle et al. 1998, Morafka et

51

al. 2000). I found that egg size, an important maternal trait in many insect species

(Mousseau and Dingle 1991), was positively associated with maternal body size. This and other work on N. femorata established that female diet is associated with egg production (Chapter 4). Thus, females that are fed a high-quality diet during their juvenile stage are more likely to grow into large females and produce large eggs. Since egg size is an important maternal effect that can influence offspring performance across many insect taxa (Lepidoptera: Bauerfeind and Fischer 2007, Coleoptera: Fox 1994,

Fox and Dingle 1994, Fox and Mousseau 1996, Kojima 2015, Kyneb and Toft 2006,

Hemiptera: Kudo 2001,Larsson 1989, McLain and Mallard 1991, Newcombe et al.

2015), future work should assess offspring performance traits such as following the development and survival of offspring through the juvenile stages in N. femorata.

52

Figure 3-1. Adult and first instar Narnia femorata. (A) Female Narnia femorata laying eggs and (B) Narnia femorata first instars and the eggs from which they hatched. Photos courtesy of Dr. Christine Miller.

Figure 3-2. Narnia femorata egg clutch. An example of an egg clutch laid on a pine needle from which egg volume was obtained. The egg clutch was photographed from the (A) top and the (B) side. Photos courtesy of Skyler Brandfon.

53

Figure 3-3: Effect of juvenile diet on body size. Females fed older fruit as juveniles grew to be larger adults than those fed newer fruit (F=6.802, df=1, p=0.011). Means ±1 SE (* denotes significance).

Figure 3-4: Body size and diet both separately affect whether or not females lay eggs. (A) Females that were fed a lifetime of newer fruit were less likely to lay eggs (Wald χ2=10.095, df=2, p=0.006) than the other diet fed females and (B) larger females were more likely to lay eggs than smaller females (Wald χ2=11.307, df=1, p=0.001). Means ±1 SE in graph A (* denotes significance).

54

Figure 3-5. Effect of body size and diet on egg volume. Of those females that laid eggs, (A) maternal diet did not directly affect egg volume (F=0.425, df=2, p=0.657), but (B) large females laid larger eggs (F=4.152, df=1, p=0.047). Means ±1 SE in graph A. Graph B depicts the average egg volume per clutch for each female. Thus, one data point represents one female’s average egg volume. This method was used for visualization purposes only.

55

CHAPTER 4 SEASONAL CHANGES IN DIET QUALITY PARTIALLY RESCUE LONG TERM FEMALE REPRODUCTIVE SUCCESS

Background

Animals live and feed in dynamic environments. Seasonal changes in nutrition can affect the physical condition and trait expression of the animals that consume these resources (Natuhara 1983, Hunter and McNeil 1997, Jablonski et al. 2000, du Dot et al.

2008, Blanco et al. 2014, Gese et al. 2016, Hall et al. 2018, Eyck et al. 2019).

Fluctuating resources are notably challenging for animals that are relatively long lived and are subject to seasonal changes in nutrition as they grow and develop (e.g.

Mammals: Rosen 2009, Aves: Blomberg et al. 2013, Reptiles: Hall et al. 2018, Insects:

Uyi et al. 2018). Some individuals may develop under optimal nutritional conditions, but for many growth, development, and reproduction occur under suboptimal nutrition in at least one life stage (Pechenik et al. 2002, Koski et al. 2006, Taborsky 2006, Eyck et al.

2019). It is therefore important to consider how variable nutritional environments across multiple life stages can affect life history traits.

Poor nutrition in early life stages can have strong negative effects on female reproductive traits (Wheeler 1996, Birkhead et al. 1999, Metcalfe and Monaghan 2001,

Gardner et al. 2009, Vaiserman 2014). Yet, few studies have examined the extent to which females can overcome poor early life diets through improvements in nutrition during adulthood (Wittmeyer et al. 2001, Mevi-Schutz and Erhardt 2005, Barrett et al.

2009a, Krause et al. 2017, Plesnar-Bielak et al. 2017). Existing work has shown that females can partially overcome a poor early life diet through a high-quality adult diet by increasing ovary mass (Wittmeyer et al. 2001, Barrett et al. 2009a) and egg production

(Wittmeyer et al. 2001, Mevi-Schutz and Erhardt 2005, Plesnar-Bielak et al. 2017).

56

However, adult survival is not rescued by adult diet (Barrett et al. 2009a, Krause et al.

2017). This research on the impacts of variable nutrition on female life history traits have used artificial (Barrett et al. 2009a, Plesnar-Bielak et al. 2017) or partially artificial

(Wittmeyer et al. 2001, Mevi-Schutz and Erhardt 2005, Krause et al. 2017) diets to understand how high-quality adult diets can help females recover from poor early life nutrition. Although these studies have been valuable, we are missing studies that tease apart the effects of juvenile and adult diets that use the actual foods the organisms consume in nature. Using ecologically relevant diets is important because animals may be better equipped to handle changes in their host plant diets rather than artificially made diets due to adaptations that these animals may have accumulated throughout their evolutionary history with this food.

Herbivores that feed on a single plant species throughout the year provide an excellent opportunity to investigate how seasonal changes in wild host plants affect female reproductive traits. Some herbivores use the same plants as they flower and fruit, during which time the nutritional quality often change dramatically (Haukioja et al.

1978, Miller 2008, Uyi et al. 2018). One such species is Narnia femorata Stål

(Hemiptera: Coreidae), the leaf-footed cactus bug. This bug is a specialist herbivore that lives and feeds on seasonally dynamic prickly-pear cactus (Figure 2-2, Cirino and Miller

2017). In North Central Florida, these insects feed on the pads and fruit of Opuntia mesacantha ssp. lata (Baranowski and Slater 1986). Previous studies have shown that

N. femorata raised on cactus pads with unripe fruit grow smaller bodies (Miller et al.

2016, Sasson et al. 2016) and males grow tiny testes in relation to body size compared to those raised on ripe cactus fruit when males reach sexual maturity (Sasson et al.

57

2016). This research suggests that unripe cactus fruit is a lower quality diet than ripe cactus fruit. N. femorata is a hemimetabolous insect that is relatively long lived and has five distinct juvenile stages of development that last about two months (Vessels et al.

2013). Interestingly, N. femorata can develop and reproduce during periods of the year when only unripe fruit is present (spring, Figure 4-1A), only ripe fruit is present (autumn,

Figure 4-1B), or when the nutritional environment changes from unripe to ripe cactus fruit (summer, Figure 2-2). Thus, females of this species can develop and reproduce during periods of either low, high, or changing diet quality.

Here, I provide an experimental study using natural host plant diets to test the ability of individuals to overcome poor early life conditions through improvements in nutrition during adulthood. I addressed two major questions: (1) What are the impacts of different host plant diets on female development, adult survival, and fecundity? (2) To what extent can a switch to a high-quality diet at the adult stage improve adult survival and fecundity when females have developed on lower quality diets? I raised N. femorata on two different host plant quality treatments: ripe (high-quality diet) and unripe cactus fruit (low-quality diet). A subset of the insects raised on unripe cactus fruit diets were switched to ripe cactus fruit diets once juveniles became adults. This diet switch simulated the diet that females would experience in the summer. The rest of the insects were maintained on their juvenile diet as adults, which simulated either a spring (unripe) or an autumn (ripe) diet. I tracked juvenile and adult survivorship, recorded juvenile development time, and quantified long-term egg number and hatching success

(fecundity). I predicted that females fed the lower quality diet during the juvenile stage would have lower survivorship and slower development to adulthood. I tested whether

58

adult survivorship and fecundity would be permanently affected by poor juvenile diet or could be rescued by higher quality diet at adulthood. We used this species’ actual food that they consume in the wild because we wanted to understand how life history traits are expressed under natural dietary conditions.

Methods

Experimental Design

Insect rearing

I used a mix of wild-caught (N=88) and laboratory-reared (N=32) N. femorata as parental pairs for this experiment. I collected both wild-caught parents and prickly-pear cacti from Starke, Florida, USA (29.9804° N, 81.9848° W) in the spring through early summer (March-June 2018), when the fruit on the plants are green and unripe. I harvested cacti bearing ripe fruits in the late autumn through winter of the previous year

(December 2017 – January 2018), planted, and kept in a greenhouse until the experiment began. The cactus used in this chapter was harvested early in the cactus growing season and a subset of the cactus we used for chapters 2 and 3. I haphazardly created 60 parental pairs and placed each into their own deli cup with a potted ripe- fruited cactus pad. A pine needle was used as an oviposition substrate for ease of egg transfer and because eggs of N. femorata have been found on fallen pine needles in the wild (L.A. Cirino, personal observation). I immediately separated resulting egg clutches from the parents and kept the egg clutches in their own deli cup with a ripe-fruited cactus pad to hatch and develop. I ensured that groups of 1st-3rd instar juveniles ranged from five to twelve individuals. This group size is common in the wild (Cirino and Miller

2017, L.A. Cirino personal observation) and developing in groups has a positive effect on juvenile survivorship during early life stages (Allen and Miller 2020). If clutch size

59

was too large, I split the group within the range specified. If the clutch size was too small, I combined full siblings from another clutch. I kept parental pairs and 1st-3rd instar juveniles in incubators (Percival Scientific I-30VL; www.percival-scientific.com, 14:10

L/D photoperiod at 26ᵒC) and I checked them daily for food quality and the presence of

4th instar juveniles.

Juvenile survivorship and development

Juveniles take about 5 weeks to mature to the penultimate (4th) instar and about another 4 weeks to mature from the 4th instar to the adult stage when fed a high-quality diet and raised in 25ᵒC incubators (Vessels et al. 2013). Ovary development increases in size considerably at the 4th instar stage of development in hemipterans (Wick and

Bonhag 1955), so I assigned juveniles to different diets at this time. I isolated 4th instar juveniles into deli cups within 0 to 24 hours after eclosion, randomly assigned juveniles to either a ripe-fruit (N=323) or an unripe-fruit (N=307) cactus diet treatment

(www.random.org; Table 4-1) and transferred these individuals to a greenhouse (21-

32ᵒC, artificial light 14:10 L/D) until they were sexually mature adults. I checked each individually housed juvenile, along with its potted cactus pad with fruit, daily for survival, adult eclosion, and to ensure cactus pads and fruit were not deteriorating. I replaced cactus pads and fruits if there were visible signs of decay (e.g. brown spots, shriveled fruit). I placed 4th instar juveniles in the greenhouse to develop into sexually mature females. Female insects lived in the greenhouse between May 9, 2018 and August 12,

2018. During this time, natural light fluctuated by approximately 46 minutes with natural light never decreasing below 13 hours (Thorsen 2020). Daily average ambient temperatures fluctuated 6.89ᵒC with the highest ambient temperature of 28.71ᵒC and the lowest 21.82ᵒC (Lusher et al. 2020). Greenhouse artificial fluorescent lights (Sylvania

60

Octron/Eco 3500K 32W) keeps females reproductive in the greenhouse year-round (L.

A. Cirino, personal observation), but seasonal effects of temperature and/or light clearly impact them beyond what we can easily control (Chapter 2).

Patterns of adult survivorship and fecundity

I determined sex upon adult eclosion and a randomly selected subsample of females from the unripe-fruit diet were switched to a ripe-fruit diet (Switch N=20; Table

4-1). The rest of the females fed unripe fruit remained on this juvenile diet (N=24; Table

4-1). All females fed ripe fruit remained on their juvenile diets (N=27; Table 4-1). All females in this experiment came from 27 parental pairs; therefore, some females in this experiment were full siblings. However, females were randomly assigned to diet treatments and full siblings were nearly evenly distributed across diets. I checked all adult females daily for mortality. Once females were 14 days post adult eclosion (sexual maturity in this species; Figure 2-6), I transferred them to a rearing room (14:10 L/D,

26ᵒC, 60% humidity). I paired females with an unrelated male of similar age that was raised on a ripe-fruit diet. I placed males into deli cups with females for four days along with a pine needle as an oviposition substrate. Previous work suggests that four days should be sufficient time for lifetime egg fertilization (Allen et al. 2018). After the 4-day mating period, I removed the male and checked females daily from this point onward

(18 days post adult eclosion) for egg laying and survival. If eggs were present, I removed each clutch and placed it into its own separate deli cup for hatching. I recorded hatching success by clutch (Y/N) after two weeks (Figure 4-1C), which is sufficient time for fertilized eggs to hatch in this species (Vessels et al. 2013). I continued to check for eggs until female death, which I recorded, or until the female had been alive for 102 days post adult eclosion. I selected this adult age because it is likely beyond typical

61

female survival in the wild (Cirino and Miller 2017) and well beyond peak reproductive output of 35 days on a high-quality diet (Allen et al. 2018). Further, females fed a high- quality diet are known to produce greater than 80% of their total eggs in the first 70 days of their adult life (Allen et al. 2018). Since some females in this study also consumed low-quality diets, they may have needed time to accumulate enough resources to produce eggs, which may delay egg laying (e.g. Valicente and O’Neil 1995). Thus, I continued to track eggs along with survival beyond the 70 day point to understand how the patterns of fecundity and survival are affected by other natural dietary conditions.

Statistical Analyses

Juvenile survivorship and development

My first goal was to understand the effects of juvenile diet on juvenile survival and development. To understand how diet affected juvenile survival (4th instar to adulthood), I ran a generalized linear model (GLM) with juvenile diet (ripe or unripe fruit) as the explanatory variable and survival to adulthood (Y/N) as the response variable. I assumed a binomial distribution with a logit link function. I included both male and female insects in this juvenile survival model as sex cannot be visually determined in N. femorata during juvenile development.

I next examined whether diet affected development time starting in the 4th instar stage of development in juvenile females that survived to adulthood. I constructed a parametric proportional hazards model (time-to-event) with a lognormal distribution (due to over-dispersion) and a log link function. Diet was the explanatory variable and juvenile development time (4th instar to adult eclosion) was the response variable.

62

Overall female reproductive success

My next goal was to examine the effects of diet on total reproductive success. I considered all adult females including those that died before reproductive age. I constructed a GLM assuming a Poisson distribution with a log link function. Diet (i.e. ripe, unripe, or switch; Table 4-1) was the explanatory variable and total egg number was the response variable. I then used the multicomp package in R with single-step adjusted p-values to perform a Tukey’s post-hoc test (Hothorn et al. 2008). Next, I constructed a GLM assuming a binomial distribution with logit link function to evaluate the probability of each egg hatching. I removed all females that did not lay eggs for this analysis (ripe: N=26, unripe: N=10, switch: N=15). Again, diet was included as the explanatory variable and a matrix of hatched and unhatched eggs was used as the response variable.

Components of female reproductive success

Patterns of Adult Female Survivorship. Next, I separately evaluated how two factors, adult survival and fecundity, contribute to overall female reproductive success that I identified above. First, I constructed a GLM assuming a Poisson distribution with a log link function to investigate the effects of adult survivorship on reproductive success.

The total number of days survived was the explanatory variable and the total number of eggs laid was the response variable. Next, I investigated the effects of diet on adult survival by constructing a Kaplan-Meier survival model with a Tarone-Ware test to estimate survival curves. Diet was included as the explanatory variable and survivorship

(adult eclosion to natural death or the end of the experiment) as the response variable.

Data was right-censored since some females were alive at the end of our study (102 days post adult eclosion).

63

Patterns of Adult Female Fecundity. I then examined the effects of diet on the number of eggs that females produced, focusing on the females that were alive in a treatment at a given week throughout the duration of the experiment. I constructed a zero-inflated generalized linear mixed model with a Poisson distribution (glmmTMB package in R; Brooks et al. 2017). Diet, week, and the interaction between these two variables were included as explanatory factors and the number of eggs laid was the response variable. Since the probability of any eggs hatching did not differ between diet treatments, the number of eggs laid was used to understand the patterns of female fecundity. Female ID was included as the random effect. Week and diet were included as the zero-inflated factors (i.e. the probability of a female producing zero eggs was assumed to vary by week and by diet) and female ID was again included as a random effect for this part of the model. Multiple zero-inflated GLMM models were compared to one another using AICtab (bbmle in R; Bolker 2020) and AIC scores were compared.

The most parsimonious model (described above) was identified by lowest AIC score and reported in the results section.

Speed of adult diet rescue

My final goal was to investigate how quickly the switched diet treatment reproductively “rescued” females that were fed an unripe fruit juvenile diet. To this end, I examined if diet affected the time it takes for females to begin laying eggs. I removed all females that did not lay eggs for this analysis (ripe: N=26, unripe: N=10, switch: N=15).

As with juvenile development, I again constructed a parametric survival regression model (time-to-event, lognormal distribution, and log link function). Diet was included as an explanatory variable and time (adult days) to first oviposition event was the response

64

variable. All analyses in this manuscript were performed in R v3.6.2 (R Core Team

2019).

Results

Juvenile Survival and Development

I found that juveniles that were raised on ripe fruit diets were more likely to survive from the 4th instar to adulthood than those raised on unripe fruit diets (GLM,

χ2=13.009, df=1, p<0.001; Figure 4-2). Of those juvenile females that survived to adulthood, those raised on unripe fruit took, on average, 12.5% longer to develop from

4th instar juveniles to adults than females raised on a ripe fruit diet (time-to-event,

χ2=12.485, df=1, p<0.001; Figure 4-2).

Overall Female Reproductive Success

When I examined overall female reproductive success, females that were fed a lifetime of ripe fruit had the highest reproductive output of all three groups (GLM,

F=8.385, df=2, p<0.001; Figure 4-3). Females that were fed the switched diet had lower lifetime reproductive output than the ripe fruit diet group (Tukey’s test Z= -12.51, adjusted p<0.0001), but greater reproductive output than the unripe fruit diet (Figure 4-

3; Tukey’s test Z= -17.800, adjusted p<0.0001). Further, 80% or more of all eggs hatched, on average, regardless of diet treatment for the females that laid eggs (GLM,

χ2=0.748, df=2, p=0.688). Thus, if females laid eggs, they were likely to hatch.

Components of Female Reproductive Success

Patterns of adult female survivorship

I next addressed the question if the aforementioned patterns in reproductive success may be due to a difference in adult survival for insects raised on different diets.

I first tested the effects of female survival on reproductive success. I found that egg

65

number and the number of days survived were correlated. The number of eggs laid was higher as females lived longer (GLM, F=29.698, df=1, p< 0.001). Next, I tested the effects of diet on female survival. I found that females that were fed the unripe fruit diet had a lower adult survivorship than those females fed ripe fruit diet (Tarone-Ware

χ2=6.263, df=2, p=0.044; Figure 4-4). Females that were fed the switched diet did not differ in adult survival from either of the other two diet treatments.

Patterns of adult female fecundity

I then examined if the overall patterns of reproductive success were due to differences in female diets. Thus, I analyzed a subset of my data to only include those females that survived to produce eggs in a treatment at a given week throughout the duration of the experiment. I found that females that were fed the ripe fruit diet and the switched diet produced more eggs early in the reproduction period than those females that were fed the unripe fruit diet (GLMM, Diet*Week: χ 2=52.792, df=22, p<0.001, Diet:

χ2=0.507, df=2, p=0.776, Week: χ 2=111.285, df=11, p<0.001; Figure 4-4). However, switch diet females did not sustain this egg production and after the first week, egg production began to decline compared to the females fed ripe fruit (Figure 4-4). Finally, all females produced similar numbers of eggs at and beyond 4 weeks of reproductive life (week 6 in Figure 4-4B).

Speed of Adult Diet Rescue

Finally, I took a closer look at how long it took females to lay their first clutch of eggs. I found that the amount of time it took females to oviposit for the first time did not differ between the ripe fruit diet and the switched diet treatments (Figure 4-5). Females that were fed unripe fruit only suffered a delay in reproductive activity (time-to-event,

χ2=11.200, df=2, p<0.050; Figure 4-5). Females fed the unripe fruit diet took 4.5 times

66

longer, on average, to lay eggs than those females fed ripe fruit adult diets (i.e. ripe fruit diet and switched diet females). Additionally, ripe fruit diet females and switch diet females laid between 38-42% of their total eggs in the first two weeks of reproduction.

Females fed the unripe fruit diet only laid about 14% of their total eggs in that same timeframe.

Discussion

I exposed females to diets typical of three points within the breeding season to examine effects of seasonal diet on female reproductive success. I found that a seasonal improvement in dietary conditions can partially rescue reproductive success.

Survivors that were switched to a high-quality diet upon reaching adulthood (typical of summer) showed patterns of egg production akin to those that experienced a high- quality diet during the first week of the reproductive period (typical of autumn). After this time point, egg production declined for switch diet females compared to females who had a lifetime of high-quality diet. I also found that females fed low-quality diets (typical of spring) had lower juvenile survival, development, and reproductive success when compared to females that were fed high-quality diets (typical of autumn). Specifically, juvenile and adult survival declined and juvenile development slowed for females that were fed the low-quality diet relative to females that were fed the high-quality diet. Thus, juvenile diet still has lasting impacts on overall female reproductive success that an improved adult diet could not fully overcome.

I found that when diet quality improved at adulthood, N. fermorata females experienced a partial rescue in reproductive success. This was due to an increase egg laying and adult survival. The egg laying patterns I found here are consistent with existing research (Wittmeyer et al. 2001, Mevi-Schutz and Erhardt 2005, Plesnar-Bielak

67

et al. 2017), but patterns in adult female suvivorship were not consistent with studies that used artificial or partially artificial diets (Barrett et al. 2009a, Krause et al. 2017). I did not detect long-lasting effects of juvenile diet on adult survival in my study as Barret et al. 2009a and Krause et al. 2017 did. Conflicting results across this study and the others may result from many factors. First, the natural diet of unripe cactus fruit may simply not be as poor as diets typically used in artificial-diet studies. Second, a diet poor for reproduction does not necessarily mean that it is bad for survival; for example, animals in many species actively switch diets during reproduction because different nutrients are needed for different functions (e.g. Pierotti and Annett 1990). Third, patterns may be the result of the adaptations that exist between N. femorata and their natural diets. Females that were fed host plant diets may have evolved ways in which to boost both adult survival and fecundity based on the evolutionary ties to these diets, a trait that females fed an artificial diet in at least one life stage could not perform (Barrett et al. 2009a, Krause et al. 2017, Wittmeyer et al. 2001, Plesnar-Bielak et al. 2017).

Thus, we should be careful when drawing conclusions about life history strategies of females from experiments that use artificial diets since the underlying causes of enhanced reproductive success may differ between these diet types.

Females fed unripe-fruit diets had reduced reproductive success relative to females fed ripe fruit during the adult stage (i.e. ripe and switch diets). A combination of reduced adult survival and decreased female fecundity contributed to of the reduction in reproductive success. Additionally, females that remained alive and consumed the unripe fruit diet (41.7% survival) took 4.5 times longer to start producing eggs compared to those that consumed the switch diet (75% survival) and the ripe fruit (96.3% survival)

68

diet. This delay in reproduction could be devastating for female fitness in this species since N. femorata females likely live up to two months in the wild (Cirino and Miller

2017). It takes females about two weeks of their adult life to become sexually mature

(Figure 2-6) leaving only a short reproductive window for females to reproduce. Future research should investigate the foraging behavior of adults early in adulthood to see if they actively seek out ripe-fruit diet during their reproductive period.

My research suggests that maternal choice of oviposition site is crucial to overall reproductive success. If mothers lay eggs on low-quality cactus patches, juveniles may be restricted to feeding on low-quality cactus fruit since juveniles are limited in mobility at this stage (i.e. do not have wings). In my study, females may have been reluctant to lay eggs on low-quality resources because it is suboptimal for juvenile growth and development, as our results reveal here. I could not disentangle low-quality nutrition with low-quality oviposition sites in this study as N. femorata live, feed, and oviposit on the same cactus pads. Therefore, females that were given this low-quality diet and oviposition site may have withheld offspring production relative to those females feeding and living on higher quality diets as adults. Further research should be carried out to unravel the degree to which cactus quality cues affects oviposition site choice.

Reproductive success may have also been lower in unripe-fruit fed versus the ripe-fruit fed females because lower quality diets may have delayed sexual maturation.

Females were provided a male mate 14-days post adult eclosion for four days. The patterns seen here from the unripe fruit females could suggest that females did not mate during this 4-day window (Aubret et al. 2002, Medeiros-Santana and Zucoloto

2016). However, recent evidence shows that female receptivity in N. femorata does not

69

differ between diet treatments within a shorter mating window (3-hours during daylight hours) when they encounter their first male mate (Figure 2-6). As my experiment allowed for a much longer period for females to interact with males (i.e. four days), I assume that female receptivity did not differ for females in our experiment either.

Additionally, N. femorata females can lay unfertilized eggs. If unripe-fed females only experienced a sexual maturity delay, then we would expect that all females would produce eggs during their lifetime and hatching success would differ among treatments.

Instead, I found a difference in egg production and no differences in hatching success.

Thus, the production of the eggs themselves may have been too energetically demanding for females that were fed a low-quality diet and females may have to wait to acquire enough resources before they can begin to produce eggs regardless of mating status.

My results show that females can quickly, but partially, reproductively recover from poor early life nutrition through improved diets during the adult stage. Seasonal plants change in quality over time and space where some plants may flower and bear unripe fruit earlier than others. Unripe fruit ripens slowly over time and some ripen quicker than others. Thus, herbivores may experience different nutrition depending on the plant on which they feed. Juvenile insects usually live and feed on the plant in which they hatch since they are wingless and their mobility is restricted to walking at this stage. Walking to another plant for high-quality nutrition can be dangerous since hidden predators may compromise juvenile survival. Thus, juveniles might be better off feeding on the plant in which they have hatched and waiting to become winged adults. As adults, females are able to fly to find better quality food and feed to quickly accumulate

70

enough resources to begin reproduction. My study suggests that life history data on females who are fed artificial laboratory-made diets are comparable to females that experience ecologically relevant diets when we examine overall reproductive success.

Females are able to partially recover from a poor early life diet with an improved diet at adulthood. However, we should be cautious when drawing conclusions about how females achieve this overall reproductive recovery. In this study, females were able to partially recover through both means of increased adult survival and increase fecundity.

However, this is not consistent with other studies examining compensatory or catch-up growth. Therefore, using ecologically relevant diets can provide us with more reliable life history information we can use to better understand our natural world.

71

Figure 4-1. Adult feeding and recently hatched juvenile Narnia femorata. Adult N. femorata feeding on (A) unripe fruit and (B) ripe fruit. (C) Newly hatched first instar N. femorata offspring and the eggs from which they hatched. Photos courtesy of Dr. Christine Miller.

Table 4-1. Diet treatments for juvenile and adult Narnia femorata. Spring season females feed on unripe fruited cactus diets throughout their lives; whereas, females that hatch in autumn experience a lifetime of ripe fruited cactus. Those females that hatch during the summer encounter a diet change due to fruit ripening during this time. Juvenile diet of Adult diet of Treatment Sample Seasonal cactus pads with: cactus pads with: Term Size occurrence Ripe fruit Ripe fruit Ripe 27 Autumn Unripe fruit Unripe fruit Unripe 24 Spring Unripe fruit Ripe fruit Switch 20 Summer

72

Figure 4-2. The effect of juvenile diet on juvenile survivorship and development. (A) Juveniles fed a ripe-fruit diet were more likely to survive to adulthood than those fed an unripe-fruit diet (GLM, χ 2=13.009, df=1, p<0.001). Mean ±1 SE bars (* denotes significance). (B) Of those juveniles that survived to adulthood and were female, ripe fruit fed juveniles (blue line) developed more quickly than those fed unripe fruit (green line; time-to-event, χ 2=12.485, df=1, p<0.001).

73

Figure 4-3. Overall female reproductive success. Adult females in all diet treatments, regardless of survival to reproductive age, are represented here. Females fed ripe-fruit diets had greater overall reproductive success than do the switched diet or the females fed an unripe-fruit diet (GLM, F=8.385, df=2, p<0.001). Mean ±1 SE bars. Letters denote statistical significance at the p<0.05 level from Tukey’s post hoc analysis.

74

Figure 4-4. The effect of diet on adult longevity and fecundity over time. (A) Females in the unripe-fruit diet treatment (green line) had lower adult survivorship than those females raised on the ripe-fruit diets (blue line). Females in the switch diet treatment (yellow line) did not differ in survivorship relative to the other groups (Tarone-Ware χ2=6.263, df=2, p=0.044). (B) The arrow indicates when I began collecting data on female fecundity (18 days post adult eclosion). Females fed the switch diet were able to produce a similar amount of eggs as the female fed a lifetime of ripe fruit during the first week of the reproduction period. However, this reproductive catch-up declines in the subsequent weeks (GLMM, Diet*Week: χ 2=52.792, df=22, p<0.001, Diet: χ 2=0.507, df=2, p=0.776, Week: χ 2=111.285, df=11, p<0.001). Mean ±1 SE bars.

75

Figure 4-5. Speed of adult diet rescue. Females fed the switched diet treatment (yellow line) laid their first clutch of eggs just as quickly as females fed ripe-fruit diets (blue line). Females fed unripe-fruit diets (green line) took longer to oviposit than the other two diet treatments (time-to-event, χ2=11.200, df=2, p<0.050).

76

CHAPTER 5 MALES WITH DAMAGED WEAPONS PRODUCE MORE OFFSPRING THAN INTACT MALES IN NON-COMPETITIVE ENVIRONMENTS

Background

Sexual selection has produced a multitude of elaborate weapons throughout the animal kingdom (Emlen 2008). Males that have well-developed weapons are more likely to displace other males and secure access to females (Clutton-Brock 1982, Sneddon et al. 1997, Karino et al. 2005, Small et al. 2009, Suzaki et al. 2015). Such access provides more mating opportunities, which may ultimately increase male reproductive success and fitness. Yet, not every male within a population can possess the most effective weaponry. Those males that grow large weapons may experience weapon damage, which can impede their ability to oust other males. Males that are weapon damaged may have few mating encounters that are comparable to males with small or absent weapons.

Weapon damage is surprisingly common in nature (Mattlin 1978, Berzins and

Caldwell 1983, Umbers et al. 2012, Jennings et al. 2017, Lane and Briffa 2017, Wong et al. 2018). For example, natural rates of weapon damage can reach 30% in male white- tailed deer, Odocoileus virginianus (Karns and Ditchkoff 2012) and 82% in male tule elk,

Cervus elaphus nannodes (Johnson et al. 2005). Males may sustain weapon damage due to intraspecific competition (Mattlin 1978, Clutton-Brock 1982, Siva-Jothy 1987,

Jennings et al. 2017), predator interactions (Hoadley 1937), disease (Adamo et al.

1995), or developmental irregularities (Emberts et al. 2016). Damage to a sexually selected weapon includes, for example, antler tines that are twisted or crushed when in velvet (Fletcher et al. 2016), sharp mammalian teeth that are knocked out (Paterson

1973, Mattlin 1978, Packer 1979), or insect horns that are not able to fully expand

77

during the final molt to adulthood, a problem that can occur with many adult structures

(Robinson et al. 1991, Maginnis 2006, 2008). A compromised weapon typically results in males being less successful in aggressive pre-copulatory competitions (Mattlin 1978,

O'Neill and Cobb 1979, Berzins and Caldwell 1983, Yasuda et al. 2011, Emberts et al.

2018). Thus, just like the minor dung beetle Onthophagus acuminatus that lack horns and cannot effectively guard a tunnel to defend a mate (Emlen 1997), those males with weapon damage may have fewer mating opportunities. When they do mate, increased fertilization ability may help offset the negative fitness consequences of a damaged weapon. However, few studies have previously investigated how weapon-damage affects males’ ability to achieve fertilization success (Joseph et al. 2018).

Narnia femorata, the leaf-footed cactus bug, engages in resource defense polygyny where males fight in contest competitions and can mate multiply. Males become dominant when they win intraspecific competitions and gain access to high quality cactus territories that attract female mates (Proctor et al. 2012, Nolen et al.

2017). Males use their enlarged hind legs to kick, squeeze, and grapple with each other to establish dominance (Proctor et al. 2012). Yet, males of this species can suffer from autotomy (i.e. limb loss) if their hind leg gets trapped in a fouled molt or caught by a predator (Emberts et al. 2016). This produces an incomplete weapon as males use both hind legs to wrap and squeeze their opponent in contest competitions (Figure 5-1,

Proctor et al. 2012). Previous work has shown that males are five times less likely to win intraspecific competitions if they have one missing hind leg (Emberts et al. 2018).

Autotomy is common in the wild with 12% of males of this species missing a hind limb

(Emberts et al. 2016). Interestingly, testes grow 15-39% larger by adulthood if males

78

experience autotomy of a single hind leg during development (Joseph et al. 2018, Miller et al., manuscript in review, Miller et al. 2019). The increased allocation to testes suggests that autotomized males may be capable of enhanced fertilization capacity, at least in contexts without pre-copulatory competition (Joseph et al. 2018).

In this study, I hypothesized that males with weapon damage during development should develop into adult males capable of greater fertilization success than intact males. I provided males with four female partners to challenge male sperm reserves. I predicted that in a non-competitive context, and sequentially paired with multiple females, males with developmental weapon damage of a single hind limb should produce more offspring. Further, I anticipated that autotomized males would mate more readily with females because these males have larger testes that likely produce more sperm and are less likely to become sperm limited (Puurtinen and

Fromhage 2017), leading to a greater number of female mates during our observation window.

To address these predictions, I conducted two experiments. In the first experiment, I provided damaged and intact males with four previously unmated females consecutively for 24-hours per female. The goal here was to compare the reproductive success of intact and weapon-damaged males in this context. In the second experiment, I observed copulations of weapon-damaged and intact males when provided multiple females sequentially. My aim here was to investigate if weapon- damaged males mated with more females than intact males.

79

Methods

Experimental Design

Insect husbandry and rearing

Adult N. femorata and cactus (Opuntia mesacantha ssp. lata) were collected in

North Central Florida, USA between July and September of 2018. These adults were paired and placed in a deli cup with a fruited cactus pad planted in soil. A pine needle was provided for egg laying. Pairs were left to mate and lay eggs freely. Offspring from these 21 mating pairs were used in the experiments.

Newly hatched juveniles were separated from their parents and either kept in incubators (Percival Scientific I-30VL; www.percival-scientific.com) or a temperature- controlled rearing room to complete early juvenile development (26ᵒC, 14:10 light/dark photoperiod). A range of five to twelve juveniles were raised from first instar through third instar in deli cups that contained a fruited cactus pad potted in soil. We raised insects in groups based on their natural aggregation propensity in groups of this size range (Cirino and Miller 2017, L.A.Cirino, personal observation) and because an increase in survivorship comes with group rearing (Allen and Miller 2020). Males were randomly assigned to treatments groups and males from all juvenile group sizes were represented in each treatment group. Cups were checked daily for cactus quality and insect maturation. Juveniles in their penultimate (4th) instar were randomly selected from both rearing locations (incubators and rearing room) for autotomy and separated into their own deli cup with a fruited cactus pad to complete development. Fourth instar juveniles were chosen for treatment because heteropterans experience the largest gonad development beginning at this stage (Economopoulos and Gordon 1971, Dumser and Davey 1974) and to match previous protocol (Joseph et al 2018, Miller et al. 2019).

80

Treatments

Every insect was randomly assigned (using www.random.org) to either the weapon-damaged or intact (i.e. control) treatment. For the weapon-damaged treatment,

I induced hind limb loss (i.e. autotomy), so that the complete weapon (considered here to include both hind legs used to grapple and squeeze opponents) was no longer intact

(Figure 5-1). I gently held the insect down and grasped the left hind femur with reverse action forceps until the insect naturally dropped the hind limb (Emberts et al. 2016).

Insects of this species cannot regenerate entire limbs (Emberts et al. 2017; Emberts et al. 2016). Thus, loss of a hind limb causes permanent weapon damage. I simulated handling stress in the intact treatment by using closed forceps to lift the left hind limb weapon while holding the body down until the insect lowered that hind limb. Treated juveniles were housed alone and placed in a greenhouse (21-32ᵒC, 14:10 L/D) to develop into sexually mature adults. Sex is difficult to determine before adult eclosion so both males and females, damaged and intact, were produced. All males and only the intact females were used in the subsequent experiments.

Experiment 5.1: 24-hour mate switch

My goal here was to examine whether males with damaged weapons were able to produce more offspring than intact males when placed in a non-competitive mating environment. Males that were at least 28 days post adult eclosion (n = 38) were assigned 4 different non-sibling, previously unmated females (n = 152) for mating one per day over 4 consecutive days. All females were at least 14 days post adult eclosion and thus sexually mature (L.A. Cirino, unpublished data). Mating pairs were maintained in a rearing room (26ᵒC, 14:10 L/D). Males were sequentially paired with each assigned

81

female partner for 24 hours. After the end of the four-day period, males were removed from the last female partner’s cup and immediately frozen. Females were kept in their individual fruited cactus pad containers to lay eggs for two weeks which is about 67% of their reproductive lifespan in wild populations in North Central Florida (Cirino and Miller

2017, L.A. Cirino, unpublished data). Females were then frozen and their eggs were left to hatch for another two weeks, which is sufficient time for the fertilized eggs to hatch in this species (Vessels et al. 2013). The numbers of eggs laid and hatched were then quantified. All insects were photographed and measured.

Experiment 5.2: Visual confirmation of copulations

My next goal was to test our prediction that weapon-damaged males would mate with more females than intact males when in non-competitive mating environments.

Behavior trials were conducted in a fully lit 26ᵒC room with a separate set of insects.

Males (n = 29) were haphazardly paired with a non-sibling unmated female in a deli cup with a fruited cactus pad. No more than ten pairs were viewed by a single observer during each 8-hour trial and deli cups were consistently scanned for mating behavior over this time. A pair was considered copulating when the male dorsoventrally mounted a female, touched the terminal ends of the abdomen together (i.e. genital contact), and turned at least 90 degrees from her body (Figure 5-1C). Females were immediately removed after one copulation when the pair naturally separated. A new non-sibling female was then put in its place. This procedure was repeated until males either mated with four females or the 8-hour observation period ended. Males were frozen immediately after behavior trials for later photographing and measuring (see below) and total copulations were tallied for each male.

82

Photographing and measurement protocol

Egg and hatchling production can vary widely in N. femorata with females laying anywhere from 0 to 278 eggs in a month with up to 91% of those eggs hatching (Wilner et al. 2020). Body size is known to be tied to the number of eggs females can produce in N. femorata (Miller et al. 2013) and other insects (Bonduriansky 2001; Honěk 1993).

Therefore, I measured females then used a composite metric of female body size as a covariate in relevant analyses (see below). Male body size was also considered because female N. femorata mate more readily with larger males (Gillespie et al. 2014).

Since male body size and testes size are highly correlated (Greenway et al. 2019), larger males may produce more sperm and fertilize more eggs (e.g. Kant et al. 2012).

I photographed all frozen specimens using a digital camera (Canon EOS 50D,

Canon, Tokyo, Japan). These images were used to procure linear body size measurements including pronotum width, body length, hind femur length, hind femur width, mid femur length, and mid femur width measurements using ImageJ software

(v1.42d, Abràmoff et al. 2004). These measurements were used in the principal component analyses (PCAs) described below.

Statistical Analysis

Body size PCAs

The six linear body size measurements were loaded into a Principal Component

Analysis (PCA). I ran one for each sex separately in Experiment 5.1 and another for the separate group of males used in Experiment 2. I ran all PCAs using a correlation matrix to reduce our body size measurements down to one composite measure (Tables A-1 –

A-4). I then ran a regression of PC1 values versus pronotum width (PW) measurements, which is commonly used as a proxy for body size in hemipterans. I

83

found that these values are highly correlated (Table A-5). Additionally, PC1 scores explained 81-92% of the variance for all PCAs; therefore, we used PC1 scores for both male and female body size in the following analyses.

Experiment 5.1: 24-hour mate switch

Overall male reproductive success. A primary objective in this study was to understand the effects of weapon damage on overall male reproductive success in a non-competitive setting. To this end, I added up the live hatchlings produced by all four females that spent time with each male. I used a generalized linear model (GLM) with a negative binomial distribution, due to over-dispersion of the data, and log link function to analyze whether weapon damage (i.e. autotomy) affected total live hatchling number.

Male body size (PC1) was used as a covariate and the interaction between weapon damage and male PC1 was also included in the model to test if effects of weapon damage impacted larger males more than small, or vice-versa.

Components of male reproductive success. Next, I took an in-depth look at a range of factors that may contribute to offspring numbers in this study. Breaking up the analyses in a stepwise reduction fashion allowed me to investigate factors that may contribute to offspring production seen in the pooled analysis above (Gotelli and Ellison

2013). I first tested whether weapon-damaged males stimulated higher levels of egg production (Y/N) across their four mates using a Generalized Linear Mixed Model

(GLMM) assuming a binomial distribution with a logit link function. Male weapon damage (Y/N) and mate order (1, 2, 3, 4) were the explanatory variables and egg production (Y/N) was the response variable. Female body size (female PC1) and male body size (male PC1) were used as covariates and male ID was included as a random effect. My next step was to test the potential effects on hatching success (Y/N) for only

84

those females that produced eggs. I constructed a second GLMM that was identical to the first model, but with hatching success (Y/N) as the response variable.

My final goal for this first experiment was to look at only those females that produced live offspring to understand factors contributing to the number of offspring produced. I constructed a third GLMM with a Poisson distribution and a log link function with the number of live hatchlings as the response variable. Male ID was treated again as the random effect. Weapon damage and mate order were the explanatory variables and female PC1 and male PC1 were included as covariates. For all three models, all two-way interactions were sequentially removed if well-beyond the level of statistical significance (P > 0.09) using stepwise reduction.

Experiment 5.2: Visual confirmation of copulations

My purpose here was to understand the effects of weapon damage on male mating success. I constructed a GLM with a multinomial distribution and cumulative logit link function to examine the number of females with which males from each treatment mated. I included weapon damage as an explanatory variable and male PC1 as a covariate. Mate number (1, 2, 3, or 4) was included as an ordinal response variable. I only included males that had mated at least once to ensure they were reproductively mature and ready to mate. All analyses in this manuscript were performed in IBM SPSS

Statistics v.24.

Ethical Note

No licenses were required to carry out experimentation on these insects.

Nevertheless, all insects were treated as humanely as possible by disinfecting the area where autotomy occurred before treating them and providing insects with as optimal as

85

possible natural living conditions. I also humanely euthanized these insects before processing them for final data collection.

Results

Body Size PCAs

I ran a body size PCA separately for each sex in Experiment 5.1 (24-hour mate switch) and found only one significant principal component (PC1), which explained 92% of the variation in the morphological data for males. PC1 scores for females explained

89% of the variation in the morphological data. In Experiment 5.2 (Visual confirmation of copulations), I ran a PCA for a separate group of males and found that the first principal component (PC1) explained 87% of the variation in the morphological data (Tables A-1

– A-5).

Experiment 5.1: 24-Hour Mate Switch

Overall male reproductive success is enhanced for small, weapon-damaged males

First, I tested if males experiencing weapon-damage would have higher overall reproductive success when mating multiply in a non-competitive environment. I found that small weapon-damaged males (i.e. those that experienced developmental autotomy) produced more offspring in total across all four female mates than small intact males (χ2=3.854, df=1, p=0.050, Figure 5-2, Table 5-1). I did not find this pattern for larger males (Figure 5-2). Small weapon-damaged males (lowest body size quartile) had over double the number of total average live hatchlings when compared to small intact males (weapon-damaged = 33.6 average live hatchlings; intact = 15.8 average live hatchlings).

86

Weapon-damaged males that mate with large females have higher reproductive success

I then explored the components of offspring production that could contribute to the differences of male reproductive success between groups. As is typical in insects, larger females were more likely to lay at least one egg (F=14.038, df=1, p≤0.001, Table

5-2, Bonduriansky 2001; Honěk 1993). I found that male weapon damage did not affect the likelihood of female egg laying (F=1.790, df=1, p=0.183, Table 5-2). However, females were more likely to have hatching success when they mated with a weapon- damaged male (F=5.255, df=1, p=0.025, Table 5-2, Figure 5-3). Next, I examined the effects on the number of live hatchlings produced. I found that weapon-damaged males produced more offspring than intact males when they were paired with larger females

(F=26.347, df=1, p≤0.001, Table 5-2, Figure 5-3). However, I did not find evidence of this pattern when males mated with smaller females (Figure 5-3).

Experiment 5.2: Visual Confirmation of Copulations

Larger intact males are more likely to copulate with many females

I found that large males that had a fully intact weapon (i.e. two hind legs) were more likely to mate with many females compared to large males with damaged weapons (χ2=4.269, df=1, p=0.039, Table 5-3, Figure 5-4). In fact, all the largest intact males mated with multiple females. Smaller males, regardless of treatment, had reduced instances of mating; many small males (84.6%) mated only once or twice.

Discussion

Hunters and naturalists alike know the ubiquity of damage to the horns and antlers of weapons that are under sexual selection. In this study, I focused on the hind leg weapons of leaf-footed bugs to experimentally study the effects of weapon damage

87

on male reproductive success. I found that males that experienced weapon damage during development had increased reproductive success in multi-mating, non- competitive environments. Weapon-damaged males were more likely to successfully reproduce than intact males when males mated with large females (Figure 5-3b).

Further, relative to intact males, weapon-damaged males produced a greater number of offspring when provided larger female mates (Figure 5-3c). Large invertebrate females are known to produce more eggs than smaller females (Honěk 1993, Bonduriansky

2001, Miller et al. 2013). These fecund females may be benefitting from mating with weapon-damaged males that have larger testes (Joseph et al. 2018). Contrary to my expectations, I found that large weapon-damaged males mated with fewer females than large intact males (Figure 5-4). Overall, our results suggest that some weapon- damaged males can produce more offspring even when mating events are few.

Trauma, such as weapon damage, may quickly change male competitive abilities and status across taxa. Males with weapon damage may pursue reproductive tactics that are more effective for their physical state (Gross 1996). The patterns we see in this study suggest that N. femorata may invest more in each mating, consistent with a reproductive tactic switch due to a conditional strategy. Indeed, game theory posits that males should change reproductive tactics when their status changes and the ability to achieve reproductive success through the original tactic is low (Gross 1996). Such changes can mean that, in some situations, compromised males may be just as effective as or more effective at fertilizing eggs than typical males (Gage et al. 1995,

Tomkins and Simmons 2000, Young et al. 2013). My results suggest a potential similarity between weapon-damaged males and males that have become subordinate

88

after losing competitions with other males (Moore et al. 1995, Clark 1997, Simmons et al. 1999, Yamane et al. 2010, Painting and Holwell 2014, Suzaki et al. 2015, Wong et al. 2018, Van den Beuken et al. 2019). Interestingly, in this study, males did not directly interact with any other males after autotomizing a limb, thus the patterns I have documented were not triggered by fight loss and defeat. Instead, they may have been due to changes in male anatomy and the increased growth of testes after autotomy

(Somjee et al. 2017, Joseph et al. 2018, Miller et al. 2019), an ability to self-assess a decline in competitive ability, the responses of females to having an autotomized mate, or all of the above.

In my study, males experienced weapon damage during juvenile development.

Early life weapon damage causes reallocation of resources from the weapon towards the testes in N. femorata (Joseph et al. 2018, Miller et al. 2019). This increased investment in the testes and other physiological changes during development may promote the switch between male fertilization tactics. However, weapon damage can occur at various stages across an animal’s lifetime (Berzins and Caldwell 1983, Wong et al. 2018). If damage occurs at a time where the development of the testes is already complete, we might not see the patterns documented here. Indeed, weapon damage in adult male cervids does not confer a decrease in dominance status or mating success

(Johnson et al. 2007, Jennings et al. 2017). Of course, the importance of timing to changes in fertilization ability will be due to the mechanisms involved. For example, if self-assessment of male-male competitive ability can occur at any life stage, fertilization tactics might change immediately upon autotomy. Further, if the patterns witnessed

89

here are due to female responses to males missing a hind limb, then timing of the autotomy should not change our findings.

Males live in a socially dynamic world and the presence of another male will often change reproductive outcomes (Andersson 1994). I expect the results documented here to be a “best case” scenario for males with damaged weapons because males did not have to overcome another male to access a female. Experimental research on consequences of weapon damage under competitive scenarios is scarce but shows that intact males are more likely to have an advantage over weapon-damaged males in physical contests before mating (Berzins and Caldwell 1983, Emberts et al. 2018).

Depending on factors such as population and resource density, weapon-damaged males may only obtain a few copulations in their lifetime. My simplification of the complexities of the natural world has demonstrated that males with weapon damage can, in non-competitive circumstances, have a reproductive advantage. Previous research has also shown that in one-on-one competitive scenarios with a female present, weapon-damaged males have decreased fighting ability and a reduction in mating success (Emberts et al. 2018). However, some males were still able to access females (Emberts et al. 2018). One next step in this species and across taxa is to investigate the reproductive consequences of weapon damage under more realistic scenarios where the behavior and fertization success of multiple males and females can be tracked. In addition, it will be fascinating to see if weapon damage leads to changes in pre-copulatory reproductive behaviors, such as attempting to sneak copulations, and the role of damage itself versus subordinate status in any changes.

90

My study is one of the first to examine the effects of developmental weapon damage on male reproductive and mating success. I found that weapon-damaged males, relative to intact males, have increased reproductive output when mating with larger females. I also found that more mating events are likely not the means by which damaged males increase fertilization success. It is possible that these males may be increasing allocation to ejaculates, but the mechanism is not yet known. This study is an important first step in understanding the implications of male weapon damage on reproductive behaviors and fitness.

The weapon damage I examined here is easy to spot and to study. However, it is important to recognize that weapon quality is not always easily visible to human observers. Like other animal tissues, the weapons of sexual selection can be compromised in many ways. For example, they may grow large, but lack strength (Miller et al. 2020, manuscript in preparation), they may be compromised by disease (Adamo et al. 1995), or they may accumulate stress fractures from combat (Lane and Briffa

2017). Few studies have examined the extent to which males can assess their own weapon damage and whether they may alter resource allocation and reproductive behaviors subsequently. Assessing the consequences of weapon damage is an intriguing path for future research, potentially informing our understanding of animal anatomy and behavior more generally.

91

Figure 5-1. Juvenile and adult Narnia femorata with damaged weapons. (A) Intact (left) and weapon-damaged (right) juveniles. (B) Adult Narnia femorata male with a damaged weapon. (C) Weapon-damaged male mating with a female. The white arrows denote the missing hind limbs that produced incomplete weapons. Photos courtesy of author.

Table 5-1. Analysis of overall male reproductive success. Results from the generalized linear model (GLM) that address overall male reproductive success in the 24- hour mate switch experiment (Exp. 1). Source Wald χ2 Weapon damage 2.443 ♂ body size (PC1) 0.506 Weapon damage*♂ body size (PC1) 3.854* All df = 1 *p≤0.05

92

Figure 5-2. Overall male reproductive success. Small males that had weapon damage (solid line) yielded greater live hatchling numbers than their intact counterparts (dashed line, χ2=3.854, df=1, p=0.050). Each point on the graph represents the total offspring per male from his four pooled mates. This graph was generated from the model summarized in Table 1.

Table 5-2. Analysis of the components of male reproductive success. Results from three separate generalized linear mixed models (GLMMs) that address the components of offspring production leading to increased reproductive success in the 24-hour mate switch experiment (Exp. 5.1). Live Eggs Eggs hatchling laid hatched number (Y/N), (Y/N), (count), Source F-value F-value F-value Weapon damage 1.790 5.255* 1.381 Mate order 2.807 1.169 2.980 Female body size (PC1) 14.038*** 0.878 2.245 Male body size (PC1) 1.981 1.649 0.039 Weapon damage*Female body size (PC1) N/A N/A 26.347*** All df = 1 *p≤0.05, **p≤0.01, ***p≤0.001

93

Figure 5-3. Components of male reproductive success. (A) Weapon damage does not affect the likelihood of egg laying (F=1.790, df=1, p=0.183; NS = not significant); however, (B) weapon-damaged males are more likely to have eggs that will hatch (F=5.255, df=1, p=0.025; * indicates significant result). Graphs A and B were generated from the model summarized in Table 2 - estimated marginal means (±SE) from the Eggs laid and Eggs hatch GLMM. (C) If eggs did hatch, weapon-damaged males (solid line) produce more live hatchlings when mated with large females when compared to intact males (dashed line, F=26.347, df=1, p≤0.001). Graph C was generated from the raw data.

Table 5-3. Analysis of visual confirmation of copulations. Results from the generalized linear model (GLM) that address the number of mating events males attain in the visual confirmation of mating events experiment (Exp. 5.2). Source Wald χ 2 Weapon damage 2.330 ♂ body size (PC1) 2.036 Weapon damage*♂ body size (PC1) 4.269* All df = 1 *p≤0.05

94

Figure 5-4. Visual confirmation of copulations. Large intact males obtain more copulations than large weapon-damaged males and smaller males (χ2=4.269, df=1, p=0.039). Weapon-damaged males are depicted with a solid line while intact males are depicted with a dashed line. This graph was generated from the raw data.

95

CHAPTER 6 CONCLUSION

Environmental heterogeneity poses a ubiquitous challenge for animals. Here, I examined how two aspects of the environment, nutritional changes and predator pressures, may alter individual investment in multiple reproductive traits. Past research on the impacts of nutrition (e.g. Wittmeyer et al. 2001, Barrett et al. 2009a, Krause et al.

2017, Plesnar-Bielak et al. 2017) and predator pressures (Simmons and Emlen 2006) on resource allocation patterns of animals typically uses artificial or partially artificial experimental manipulations. Although these studies have uncovered important information about the life history patterns of animals, their use of artificial environmental conditions made it unclear whether or not these patterns could be extended to patterns in natural environments. My research has begun to elucidate the resource allocation patterns that animals follow in the wild under natural and ecologically relevant environmental conditions.

The harshest nutritional environments in which females feed on resources with very low nutrient content (e.g. Barbera et al. 1992), have the greatest consequences for female reproduction in Narnia femorata. This includes low body and ovary mass and the inability to produce eggs (Chapter 2 – 3). Cactus fruit diet deteriorates over time when temperature and photoperiod change late in the year. At this time, N. femorata females reduce investment into reproduction and prepare their bodies for winter (Chapter 2).

Yet, food restriction can also occur early in the season. Habitat fragmentation, competitive herbivory, and seasonal changes can expose animals to poor nutritional conditions and force animals to increase foraging. Juvenile N. femorata may be limited to foraging on the cactus patch where they hatched since they do not possess wings

96

during this stage. Thus, they may suffer the initial reproductive consequences of a poor nutritional environment. As adults, N. femorata have fully formed wings and females are able to search for better quality food. I have shown that access to improved diets upon adulthood partially rescues reproductive success (Chapters 2 and 4). Females are able to begin reproductive activity without delay and survive just as long as those females that experienced superior nutrition during the juvenile stage (Chapter 4). Interestingly, I did not detect differences in egg size based on maternal diet (Chapter 3) and this suggests that females cannot adjust egg size based on diet but instead may adjust other reproductive traits to cope with poor dietary environments.

Seasonally changing nutrition can be a predictable environmental cue which animals can anticipate. Predator pressure, on the other hand, is more unpredictable.

Some animals have evolved autotomy to escape predator entrapment (Emberts et al.

2019). However, the resulting fitness costs of this behavior may be severe for males. In

N. femorata, when males drop a hind limb to escape a predator, they damage the weapon that they use to compete in male-male competitions for access to females. An intact weapon is composed of the two enlarged hind legs that are used to squeeze, grapple, and kick an opponent (Proctor et al. 2012, Nolen et al. 2017). When a male is missing part of that weapon (i.e. hind leg autotomy), the weapon is rendered ineffective

(Emberts et al. 2018). Thus, access to females declines for weapon-damaged males

(Emberts et al. 2018). Males that experience weapon damage during development invest more heavily into testes size (Joseph et al. 2018, Miller et al. 2019). I show that weapon-damaged males can increase fertilization success when given the opportunity to mate with large females (Chapter 5). Thus, under scenarios where no other males

97

are present, weapon-damaged males are able to fertilize more offspring than intact males when mating with large females. These scenarios are not uncommon for N. femorata (Cirino and Miller 2017). Population size fluctuates throughout the year (Cirino and Miller 2017) which makes access to mates more challenging at certain times of the year than others. Therefore, mating success of weapon-damaged males fluctuate over time.

My research provides us with a better understanding of how seasonal host plant diets at important life stages can affect female reproduction. Improved diets at adulthood can help start female reproductive activity (Chapter 2 and 4), allow females to produce oocytes in the ovary (Chapter 2), and store eggs in the oviduct (Chapter 2).

While improved reproductive anatomy is important for reproduction, it is also important to understand how mating behavior is affected by dietary restriction in the juvenile stage. I did not detect differences in female receptivity among any of my diet treatments once females were sexually mature. This result is consistent with other N. femorata research that showed that a period of adult diet restriction imposed on females also does not affect their receptivity (Wilner et al. 2020). Diet quality and quantity is essential to reproductive traits as I show in Chapters 2 and 4. Although I did not find evidence that maternal diet affects egg size (Chapter 3), I did show that larger females produced larger eggs. This egg size investment may help N. femorata offspring endure their non- feeding stage of development, but we have yet to learn how long lasting this maternal effect is, especially when offspring experience poor dietary conditions. Since egg size is one of the primary ways in which mothers can influence their offspring (i.e. maternal effect), examining the effect of egg size on offspring performance under different

98

offspring nutritional conditions may provide us with a better understanding of the fitness constraints that a seasonally dynamic environment can impose.

Male reproductive allocation patterns may differ from females (Miller et al. 2019).

In chapter 5, I show that weapon-damaged males can increase fertilization success when they have access to multiple large females. This scenario is not always common in nature. We are missing studies that examine how often weapon-damaged males are able to mate when many, better-equipped males are present. After all, when population numbers are high, there is likely a greater chance that weapon-damaged males may interact with other males and struggle to obtain mates (Emberts et al. 2018). It is important to continue investigating the mating and reproductive consequences of weapon-damaged males under different population densities and in freely interacting environments, so we can better understand the reproductive constraints of this weapon- testes tradeoff (Joseph et al. 2018, Miller et al. 2019).

I detected increased fertilization success for weapon-damaged males under the best possible nutritional conditions for N. femorata. Yet, environments rarely stay consistent. Resource acquisition by animals can vary tremendously for many reasons, including seasonal changes in food availability and quality (Feeny 1970, Awmack and

Leather 2002) and competitive herbivory (Jia et al. 2018). Thus, there are likely times when an entire population is flush with food and tradeoffs are minimized (Chapter 5,

Zera et al. 1998, Cahenzli and Erhardt 2012), and other times when resources are scarce and tradeoffs are more extreme (Tatar and Carey 1995, Naya et al. 2007,

Colasurdo et al. 2009, King et al. 2011). Importantly, however, most studies have focused on the effects of resource availability on tradeoffs without considering resource

99

quality. Indeed, it is obvious that not all foods provide the same nutrition to animals.

Animals may have abundant food sources and yet still be missing essential nutrients.

Empirical studies that have manipulated the ratios of proteins-to-carbohydrates, for example, have found evidence of tradeoffs between reproduction and somatic growth

(Naya et al. 2007, Colasurdo et al. 2009) and reproduction and survivorship (Sentinella et al. 2013, Gray et al. 2018, Jang and Lee 2018). Thus, it is entirely possible that resource quality could change the very nature of allocation tradeoffs and, in effect, potentially change how selection is acting on life history traits.

In conclusion, I have found that multiple dynamic environmental factors can affect a suite of reproductive traits in both females and males. We need to continue investigating the relationship between variable environments and reproduction to more accurately predict the evolutionary trajectories of populations and their persistence as the environment continues to change.

100

APPENDIX SUPPLEMENTARY FIGURES AND TABLES

Table A-1. Factor loadings by trait for all Principal Component Analyses in Chapter 5. All eigenvalues and percent variation explained are represented in Table A-4. Factor loadings Factor loadings for Factor loadings for for PC1 (male – PC1 (females – PC1 (males – Exp.1) Exp.1) Exp.2) Hind femur width (HFW) 0.952 0.935 0.933 Hind femur length (HFL) 0.970 0.975 0.616 Mid femur width (MFW) 0.913 0.852 0.909 Mid femur length (MFL) 0.971 0.965 0.953 Pronotum width (PW) 0.967 0.957 0.899 Body length (BL) 0.975 0.975 0.958

Table A-2. Correlation matrix for the PCAs in Chapter 5, Experiment 1. Correlation matrix for the PCA for females (green) and males (gray) in Experiment 5.1: 24-hour mate switch. Abbreviations used are from Table A-1. Correlation HFW HFL MFW MFL PW BL HFW 1.000 0.903 0.766 0.860 0.876 0.885 HFL 0.905 1.00 0.788 0.971 0.911 0.939 MFW 0.892 0.828 1.000 0.772 0.752 0.779 MFL 0.876 0.972 0.866 1.000 0.906 0.945 PW 0.903 0.920 0.828 0.923 1.000 0.963 BL 0.900 0.943 0.841 0.939 0.977 1.000

101

Table A-3. Correlation matrix for the PCAs in Chapter 5, Experiment 2. Correlation matrix for the PCA for males in Experiment 5.2: Visual confirmation of mating events. Abbreviations used are from Table A-1. Correlation HFW HFL MFW MFL PW BL HFW 1.000 HFL 0.443 1.000 MFW 0.839 0.492 1.000 MFL 0.858 0.562 0.879 1.000 PW 0.873 0.405 0.728 0.804 1.000 BL 0.861 0.550 0.830 0.909 0.884 1.000

Table A-4. Total variance explained by PCA in Chapter 5. Total variance explained by PCA for males and females in Experiment 5.1: 24-hour mate switch and males in Experiment 5.2: Visual confirmation of mating events. Exp. 1 males Exp. 1 females Exp. 2. males % of Cumulative % of Cumulative % of Cumulative Component Eigenvalue Eigenvalue Eigenvalue variance variance variance variance variance variance 1 5.507 91.778 91.778 5.348 89.137 89.137 4.712 80.628 80.628 2 0.239 3.982 95.760 0.320 5.334 94.471 0.704 12.040 92.668 3 0.125 2.087 97.847 0.157 2.618 97.089 0.284 4.862 97.530 4 0.094 1.575 99.422 0.121 2.024 99.113 0.144 2.469 99.999

Table A-5. Regression of PC1 values in Chapter 5. Regression of PC1 values, generated from the three separate PCAs in this experiment, versus pronotum width (PW; a proxy of body size). Experiment and sex R2 p-value Exp. 1 males 0.9341 <0.001 Exp. 1 females 0.9159 <0.001 Exp. 2 males 0.8089 <0.001

102

LIST OF REFERENCES

Abràmoff, M. D., P. J. Magalhães, and S. J. Ram. 2004. Image processing with ImageJ. Biophotonics Int. 11: 36-42.

Adamo, S. A., D. Robert, and R. R. Hoy. 1995. Effects of a tachinid parasitoid, Ormia ochracea, on the behavior and reproduction of its male and female field cricket hosts (Gryllus spp). J Insect Physiol. 41: 269-277.

Adams, T. S. 2000. Effect of diet and mating status on ovarian development in a predaceous stink bug Perillus bioculatus (Hemiptera: Pentatomidae). Ann Entomol Soc Am. 93: 529-535.

Addesso, K. R., K. A. Short, A. J. Moore, and C. W. Miller. 2014. Context-dependent female mate preferences in leaf-footed cactus bugs. Behaviour. 151: 479-492.

Akiyama, S., Y. Iwao, and I. Miura. 2011. Evidence for True Fall-mating in Japanese Newt Cynops pyrrhogaster. Zool Sci. 28: 758-763.

Allen, P., and C. Miller. 2020. The hidden cost of group living for aggregating juveniles in a sexually dimorphic species. Biol J Linn Soc. DOI: 10.1093/biolinnean/blaa090.

Allen, P. E., A. G. Dale, S. I. Diyaljee, N. J. Ector, D. Petit-Bois, J. T. Quinn, A. C. Ranieri, J. A. Sanchez, H. M. Smith, D. X. Tran, A. M. Winsor, and C. W. Miller. 2018. One and Done: Long-term sperm storage in the cactus-feeding bug, Narnia femorata (Hemiptera: Coreidae). Ann Entomol Soc Am. 111(5): 1-7.

Andersson, M. B. 1994. Sexual selection. Princeton, NJ: Princeton University Press.

Anjos-Duarte, C. S., A. M. Costa, and I. S. Joachim-Bravo. 2011. Influence of Female Age on Variation of Mate Choice Behavior in Mediterranean Fruit Fly (Diptera: Tephritidae). J Insect Physiol. 24: 11-21.

Aubret, F., X. Bonnet, R. Shine, and O. Lourdais. 2002. Fat is sexy for females but not males: The influence of body reserves on reproduction in snakes (Vipera aspis). Horm Behav. 42: 135-147.

Awmack, C. S., and S. R. Leather. 2002. Host plant quality and fecundity in herbivorous insects. Annu Rev Entomol. 47: 817-844.

Baranowski, R.M.; Slater, J.A. 1986. Coreidae of Florida (Hemiptera: Heteroptera). In of Florida and Neighboring Land Areas; Florida Department of Agriculture and Consumer Services, Division of Plant Industry: Gainesville, FL, USA.

103

Barbera, G., F. Carimi, P. Inglese, and M. Panno. 1992. Physical, morphological and chemcial changes during fruit development and ripening in three cultivars of prickly pear, Opuntia ficus-indica (L.) Miller. J Hortic Sci. 67: 307-312.

Barrett, E., J. Hunt, A. Moore, and P. Moore. 2009a. Separate and combined effects of nutrition during juvenile and sexual development on female life-history trajectories: the thrifty phenotype in a cockroach. Proc R Soc B. 276: 3257-3264.

Barrett, E. L. B., A. J. Moore, and P. J. Moore. 2009b. Diet and social conditions during sexual maturation have unpredictable influences on female life history trade-offs. J Evol Biol. 22: 571-581.

Barry, K. L., and S. M. Wilder. 2013. Macronutrient intake affects reproduction of a predatory insect. Oikos. 122: 1058-1064.

Bauerfeind, S. S., and K. Fischer. 2005. Effects of food stress and density in different life stages on reproduction in a butterfly. Oikos. 111: 514-524.

Bauerfeind, S. S., and K. Fischer. 2007. Maternal body size as an evolutionary constraint on egg size in a butterfly. Evolution. 61: 2374-2385.

Bauerfeind, S. S., K. Fischer, S. Hartstein, S. Janowitz, and D. Martin-Creuzburg. 2007. Effects of adult nutrition on female reproduction in a fruit-feeding butterfly: The role of fruit decay and dietary lipids. J Insect Physiol. 53: 964-973.

Bernardo, J. 1996. Maternal effects in animal ecology. Am Zool. 36: 83-105.

Berzins, I. K., and R. L. Caldwell. 1983. The effect of injury on the agonistic behavior of the Stomatopod, Gonodactylus bredini (Manning). Mar Behav Physiol. 10: 83-96.

Birkhead, T. R., F. Fletcher, and E. J. Pellatt. 1999. Nestling diet, secondary sexual traits and fitness in the zebra finch. Proc R Soc B. 266: 385-390.

Blanco, G., P. Laiolo, and J. A. Fargallo. 2014. Linking environmental stress, feeding- shifts and the 'island syndrome': a nutritional challenge hypothesis. Popul Ecol. 56: 203-216.

Blomberg, E. J., J. S. Sedinger, D. V. Nonne, and M. T. Atamian. 2013. Seasonal reproductive costs contribute to reduced survival of female greater sage-grouse. J Avian Biol. 44: 149-158.

Boggs, C. L. 1997. Dynamics of reproductive allocation from juvenile and adult feeding: Radiotracer studies. Ecology. 78: 192-202.

Bolker, B. M. 2020. bbmle: Tools for General Maximum Likelihood Estimation.

104

Bonduriansky, R. 2001. The evolution of male mate choice in insects: a synthesis of ideas and evidence. Biol Rev. 76: 305-339.

Bong, L. J., K. B. Neoh, C. Y. Lee, and Z. Jaal. 2014. Effect of Diet Quality on Survival and Reproduction of Adult Paederus fuscipes (Coleoptera: Staphylinidae). J Med Entomol. 51: 752-759.

Bowling, C. C. 1980. The Stylet Sheath as an Indicator of Feeding Activity by the Southern Green Stink Bug on Soybeans. J Econ Entomol. 73(1): 1-3.

Brooks, M. E., K. Kristensen, K. J. van Benthem, A. Magnusson, C. W. Berg, A. Nielsen, H. J. Skaug, M. Machler, B. M. Bolker. 2017. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R journal. 9(2): 378-400.

Cahenzli, F., and A. Erhardt. 2012. Nectar sugars enhance fitness in male Coenonympha pamphilus butterflies by increasing longevity or realized reproduction. Oikos. 121: 1417-1423.

Castañé, C., Ò. Alomar, J. Riudavets, and C. Gemeno. 2007. Reproductive biology of the predator Macrolophus caliginosus: Effect of age on sexual maturation and mating. Biol Control. 43: 278-286.

Cirino, L. A., and C. W. Miller. 2017. Seasonal Effects on the Population, Morphology and Reproductive Behavior of Narnia femorata (Hemiptera: Coreidae). Insects. 8.

Clark, R. A. 1997. Dimorphic males display alternative reproductive strategies in the marine amphipod Jassa marmorata Holmes (Corophioidea: Ischyroceridae). Ethology. 103: 531-553.

Clutton-Brock, T. H. 1982. The functions of antlers. Behaviour. 79(2-4): 108-124.

Colasurdo, N., Y. Gelinas, and E. Despland. 2009. Larval nutrition affects life history traits in a capital breeding moth. J Exp Biol. 212: 1794-1800.

Cook, R. C., J. G. Cook, D. J. Vales, B. K. Johnson, S. M. McCorquodale, L. A. Shipley, R. A. Riggs, L. L. Irwin, S. L. Murphie, B. L. Murphie, K. A. Schoenecker, F. Geyer, P. B. Hall, R. D. Spencer, D. A. Immell, D. H. Jackson, B. L. Tiller, P. J. Miller, and L. Schmitz. 2013. Regional and Seasonal Patterns of Nutritional Condition and Reproduction in Elk. Wildl Monogr. 184(1): 1-45.

Dani, K. G. S., and U. Kodandaramaiah. 2017. Plant and Animal Reproductive Strategies: Lessons from Offspring Size and Number Tradeoffs. Front Ecol Evol. 5: 38.

105

Dow, H., and S. Fredga. 1984. Factors affecting reproductive output of the goldeneye duck Bucephala clangula. J Anim Ecol. 53: 679-692. du Dot, T. J., D. A. S. Rosen, and A. W. Trites. 2008. Steller sea lions show diet- dependent changes in body composition during nutritional stress and recover more easily from mass loss in winter than in summer. J Exp Mar Biol Ecol. 367: 1-10.

Dumser, J. B., and K. G. Davey. 1974. Endocrinological and other factors influencing testis development in Rhodnius prolixus. Can J Zool. 52: 1011-1022.

Economopoulos, A. P., and H. T. Gordon. 1971. Growth and differentiation of testes in large milkweed bug, Oncopeltus fasciatus (Dallas). J Exp Zool. 177: 391-405.

Emberts, Z., C. St Mary, and C. Miller. 2016. Coreidae (Insecta: Hemiptera) Limb Loss and Autotomy. Ann Entomol Soc Am. 109: 678-683.

Emberts, Z., I. Escalante, and P. W. Bateman. 2019. The ecology and evolution of autotomy. Biol Rev Camb Philos Soc. 94(6): 1881-1896.

Emberts, Z., C. M. St Mary, T. J. Herrington, and C. W. Miller. 2018. Males missing their sexually selected weapon have decreased fighting ability and mating success in a competitive environment. Behav Ecol Sociobiol. 72(5): 81.

Emlen, D. 1997. Alternative reproductive tactics and male-dimorphism in the horned beetle Onthophagus acuminatus (Coleoptera: Scarabaeidae). Behav Ecol Sociobiol. 41: 335-341.

Emlen, D. 2008. The Evolution of Animal Weapons. Ann Rev Ecol Evol Syst. 39: 387- 413.

Evans, J. P., and F. Garcia-Gonzalez. 2016. The total opportunity for sexual selection and the integration of pre- and post-mating episodes of sexual selection in a complex world. J Evol Biol. 29: 2338-2361.

Eyck, H. J. F., K. L. Buchanan, O. L. Crino, and T. S. Jessop. 2019. Effects of developmental stress on animal phenotype and performance: a quantitative review. Biol Rev. 94: 1143-1160.

Feeny, P. 1970. Seasonal changes in oak leaf tannins and nutrients as a cause of spring feeding by winter moth caterpillars. Ecology. 51: 565-581.

Fischer, K., D. M. O'Brien, and C. L. Boggs. 2004. Allocation of larval and adult resources to reproduction in a fruit-feeding butterfly. Funct Ecol. 18: 656-663.

106

Fisher, H. S., and G. G. Rosenthal. 2006. Hungry females show stronger mating preferences. Behav Ecol. 17: 979-981.

Fletcher, J., A. Foster, P. Goddard, and A. McSloy. 2016. Managing antler problems in deer. In Pract. 38(10): 513-519.

Fox, C. W. 1994. The influence of egg size on offspring performance in the seed beetle, Callosobruchus maculatus. Oikos. 71: 321-325.

Fox, C. W., and H. Dingle. 1994. Dietary mediation of maternal age effects on offspring performance in a seed beetle (Coleoptera: Bruchidae). Funct Ecol. 8: 600-606.

Fox, C. W., and T. A. Mousseau. 1996. Larval host plant affects fitness consequences of egg size variation in the seed beetle Stator limbatus. Oecologia. 107: 541-548.

Fox, C. W., and M. E. Czesak. 2000. Evolutionary ecology of progeny size in arthropods. Ann Rev Entomol. 45: 341-369.

Gage, M. J. G., P. Stockley, and G. A. Parker. 1995. Effects of alternative male mating strategies on characteristics of sperm production in the Atlantic salmon (Salmo salar): Theoretical and empirical investigations. Philos Trans R Soc B. 350: 391- 399.

Gardner, D. S., S. E. Ozanne, and K. D. Sinclair. 2009. Effect of the early-life nutritional environment on fecundity and fertility of mammals. Philos Trans R Soc B. 364: 3419-3427.

Geister, T. L., M. W. Lorenz, K. H. Hoffmann, and K. Fischer. 2008. Adult nutrition and butterfly fitness: effects of diet quality on reproductive output, egg composition, and egg hatching success. Front Zool. 5(1): 10.

Gese, E. M., B. M. Roberts, and F. F. Knowlton. 2016. Nutritional effects on reproductive performance of captive adult female coyotes (Canis latrans). Anim Reprod Sci. 165: 69-75.

Gillespie, S. R., M. S. Tudor, A. J. Moore, and C. W. Miller. 2014. Sexual selection is influenced by both developmental and adult environments. Evolution. 68: 3421- 3432.

Golec, J. R., and X. P. Hu. 2015. Preoverwintering Copulation and Female Ratio Bias: Life History Characteristics Contributing to the Invasiveness and Rapid Spread of Megacopta cribraria (Heteroptera: Plataspidae). Environ Entomol. 44: 411-417.

Gordon, H. T., and S. K. Bandal. 1967. Effect of mating on egg production by large milkweed bug Oncopeltus fasciatus (Hemiptera: Lygaeidae). Ann Entomol Soc Am. 60(5): 1099-1102.

107

Gotelli, N. J., and A. M. Ellison. 2013. A primer of ecological statistics, 2nd ed. Sunderland, MA: Sinauer Associates, Inc.

Gray, D. A. 1999. Intrinsic factors affecting female choice in house crickets: Time cost, female age, nutritional condition, body size, and size-relative reproductive investment. J Insect Behav. 12: 691-700.

Gray, L. J., S. J. Simpson, and M. Polak. 2018. Fruit flies may face a nutrient-dependent life-history trade-off between secondary sexual trait quality, survival and developmental rate. J Insect Physiol. 104: 60-70.

Greenway E.G., Cirino L.A., Wilner D., Somjee U., Anagnostou M.E., Hepple R.T., Miller C.W. 2019. Extreme variation in testes size in an insect is linked to recent mating activity. J Evol Biol. DOI: 10.1111/jeb.13574

Gross, M. R. 1996. Alternative reproductive strategies and tactics: Diversity within sexes. Trends Ecol Evol. 11: 92-98.

Hall, J. M., A. Buckelew, M. Lovern, S. M. Secor, and D. A. Warner. 2018. Seasonal Shifts in Reproduction Depend on Prey Availability for an Income Breeder. Physiol Biochem Zool. 91: 1129-1147.

Hatle, J. D., Waskey Jr, T. Juliano, S. A. 2006. Plasticity of grasshopper vitellogenin production in response to diet is primarily a result of changes in fat body mass. J Comp Physiol B. 176: 27–34

Haukioja, E., P. Niemela, L. Isoiivari, H. Ojala, and E. M. Aro. 1978. Birch leaves as a resource for herbivores. 1. Variation in suitiability of leaves. Rep Kevo Subarctic Res Stat. 14: 5-12.

Hill, C. J., and N. E. Pierce. 1989. The effect of adult diet on the biology of butterflies. 1. The common imperial blue, Jalmenus evagoras. Oecologia. 81: 249-257.

Hingle, A., K. Fowler, and A. Pomiankowski. 2001. The effect of transient food stress on female mate preference in the stalk–eyed fly Cyrtodiopsis dalmanni. Proc R Soc B. 268: 1239-1244.

Hoadley, L. 1937. Autotomy in the brachyuran, Uca pugnax. Biol Bull. 73: 155-163.

Hochuli, D. F. 2001. Insect herbivory and ontogeny: How do growth and development influence feeding behaviour, morphology and host use? Austral Ecol. 26: 563- 570.

Honěk, A. 1993. Intraspecific variation in body size and fecundity in insects - a general relationship. Oikos. 66: 483-492.

108

Hothorn, T., F. Bretz, and P. Westfall. 2008. Simultaneous Inference in General Parametric Models. Biom J. 50: 346-363.

Hunt, J., R. Brooks, and M. D. Jennions. 2005. Female mate choice as a condition- dependent life-history trait. Am Nat. 166: 79-92.

Hunter, M. D., and J. N. McNeil. 1997. Host-plant quality influences diapause and voltinism in a polyphagous insect herbivore. Ecology. 78: 977-986.

Jablonski, N. G., M. J. Whitfort, N. Roberts-Smith, and Q. Q. Xu. 2000. The influence of life history and diet on the distribution of catarrhine primates during the Pleistocene in eastern Asia. J Hum Evol. 39: 131-157.

Jang, T., and K. P. Lee. 2018. Comparing the impacts of macronutrients on life-history traits in larval and adult Drosophila melanogaster. the use of nutritional geometry and chemically defined diets. J Exp Biol. 221(21). DOI: jeb181115.

Janzen, D. H. 1986. Chihuahuan desert nopaleras - defaunated big mammal vegetation. Ann Rev Ecol Syst. 17: 595-636.

Jennings, D. J., R. J. Boys, and M. P. Gammell. 2017. Weapon damage is associated with contest dynamics but not mating success in fallow deer (Dama dama). Biol Lett. 13(11): 20170565.

Jennions, M. D., and M. Petrie. 1997. Variation in mate choice and mating preferences: A review of causes and consequences. Biol Rev. 72: 283-327.

Jia, S. H., X. G. Wang, Z. Q. Yuan, F. Lin, J. Ye, Z. Q. Hao, and M. S. Luskin. 2018. Global signal of top-down control of terrestrial plant communities by herbivores. Proc Natl Acad Sci USA. 115: 6237-6242.

Johnson, H. E., V. C. Bleich, and P. R. Krausman. 2005. Antler breakage in tule elk, Owens Valley, California. J Wildl Manage. 69: 1747-1752.

Johnson, H. E., V. C. Bleich, P. R. Krausman, and J. L. Koprowski. 2007. Effects of antler breakage on mating behavior in male tule elk (Cervus elaphus nannodes). Eur J Wildl Res. 53: 9-15.

Joseph, P. N., Z. Emberts, D. A. Sasson, and C. W. Miller. 2018. Males that drop a sexually selected weapon grow larger testes. Evolution. 72: 113–122.

Judge, K. A., J. J. Ting, and D. T. Gwynne. 2014. Condition dependence of female choosiness in a field cricket. J Evol Biol. 27: 2529-2540.

Kant R., Minor M.A., Trewick S.A., Sandanayaka W.R.M. 2012. Body size and fitness relation in male and female Diaeretiella rapae. BioControl. 57(6):759-766.

109

Karino, K., H. Niiyama, and M. Chiba. 2005. Horn length is the determining factor in the outcomes of escalated fights among male Japanese horned beetles, Allomyrina dichotoma L. (Coleoptera: Scarabaeidae). J Insect Behav. 18: 805-815.

Karns, G. R., and S. S. Ditchkoff. 2012. Antler breakage patterns in white-tailed deer. Proc Ann Conf Southeast Assoc Fish Wildl Agencies. 66: 114-119.

King, E., D. Roff, and D. Fairbairn. 2011. The evolutionary genetics of acquisition and allocation in the wing dimorphic cricket, Gryllus firmus. Evolution. 65: 2273-2285.

Kodric-Brown, A., and P. F. Nicoletto. 2001. Age and experience affect female choice in the guppy (Poecilia reticulata). Am Nat. 157: 316-323.

Kojima, W. 2015. Variation in body size in the giant rhinoceros beetle Trypoxylus dichotomus is mediated by maternal effects on egg size. Ecol Entomol. 40: 420- 427.

Koski, M., W. K. Breteler, N. Schogt, S. Gonzalez, and H. H. Jakobsen. 2006. Life- stage-specific differences in exploitation of food mixtures: diet mixing enhances copepod egg production but not juvenile development. J Plankton Res. 28: 919- 936.

Kott, P., S. Roth, and K. Reinhardt. 2000. Hibernation mortality and sperm survival during dormancy in female Nabidae (Heteroptera: Nabidae). Opusc Zool Flumin. 182: 1-6.

Krause, E. T., O. Kruger, and H. Schielzeth. 2017. Long-term effects of early nutrition and environmental matching on developmental and personality traits in zebra finches. Anim Behav. 128: 103-115.

Kudo, S. 2001. Intraclutch egg-size variation in acanthosomatid bugs: adaptive allocation of maternal investment? Oikos. 92: 208-214.

Kunz, K., and G. Uhl. 2015. Short-Term Nutritional Limitation Affects Mating Behaviour and Reproductive Output in Dwarf Spiders. Ethology. 121: 874-881.

Kyneb, A., and S. Toft. 2006. Effects of maternal diet quality on offspring performance in the rove beetle Tachyporus hypnorum. Ecol Entomol. 31: 322-330.

Lane, S., and M. Briffa. 2017. The price of attack: rethinking damage costs in animal contests. Anim Behav. 126: 23-29.

Larsson, F. K. 1989. Female longevity and body size as predictors of fecundity and egg length in Graphosoma lineatum L. (Hemiptera: Pentatomidae). Dtsch Entomol Z. 36: 329-334.

110

Lusher, R., L. Treadaway, L. Staudt, D. Jia, S. Cox. 2018. FAWN report. The Florida Automated Weather Network (FAWN). 29 June 2020. https://fawn.ifas.ufl.edu/data/

Maginnis, T. L. 2006. Leg regeneration stunts wing growth and hinders flight performance in a stick insect (Sipyloidea sipylus). Proc R Soc B. 273: 1811-1814.

Maginnis, T. L. 2008. Autotomy in a stick insect (Insecta: Phasmida): Predation versus molting. Fla Entomol. 91: 126-127.

Majure, L. C. 2014. Typifications and a nomenclatural change in some eastern North American Opuntia (Cactaceae). Phytoneuron. 106: 1-2.

Mas, F., and M. Kolliker. 2008. Maternal care and offspring begging in social insects: chemical signalling, hormonal regulation and evolution. Anim Behav. 76: 1121- 1131.

Mattlin, R. H. 1978. Population biology, thermoregulation and site preference of the New Zealand fur seal, Arctocephalus forsteri (Lesson, 1828), on the Open Bay Islands, New Zealand. (Ph.D. diss.) University of Canterbury, Christchurch, New Zealand.

Mautz, B. S., A. P. Møller, and M. D. Jennions. 2013. Do male secondary sexual characters signal ejaculate quality? A meta-analysis. Biol Rev. 88: 669-682.

McAllan, B. M., and F. Geiser. 2014. Torpor during Reproduction in Mammals and Birds: Dealing with an Energetic Conundrum. Integr Comp Biol. 54: 516-532.

McLain, D. K., and S. D. Mallard. 1991. Sources and adaptive consequences of egg size variation in Nezara viridula (Hemiptera: Pentatomidae). Psyche. 98: 135- 164.

Medeiros-Santana, L., and F. S. Zucoloto. 2016. Sexual receptivity of Anastrepha obliqua (Diptera: Tephritidae) wild females affected by nutrition and egg load. Entomol Sci. 19: 233-238.

Metcalfe, N. B., and P. Monaghan. 2001. Compensation for a bad start: grow now, pay later? Trends Ecol. Evol. 16: 254-260.

Mevi-Schutz, J., and A. Erhardt. 2005. Amino acids in nectar enhance butterfly fecundity: A long-awaited link. Am Nat. 165: 411-419.

Miller, C. W. 2008. Seasonal effects on offspring reproductive traits through maternal oviposition behavior. Behav Ecol. 19: 1297-1304.

111

Miller, C. W., R. J. Fletcher, and S. R. Gillespie. 2013. Conspecific and heterospecific cues override resource quality to influence offspring production. PlosOne. 8: 1-7.

Miller, C. W., G. C. McDonald, and A. J. Moore. 2016. The tale of the shrinking weapon: seasonal changes in nutrition affect weapon size and sexual dimorphism, but not contemporary evolution. J Evol Biol. 29: 2266-2275.

Miller, C. W., P. N. Joseph, R. M. Kilner, and Z. Emberts. 2019. A weapons-testes trade-off in males is amplified in female traits. Proc R Soc B. 286: 20190906.

Moore, A. J., N. L. Reagan, and K. F. Haynes. 1995. Conditional signalling strategies: effects of ontogeny, social experience and social status on the pheromonal signal of male cockroaches. Anim Behav. 50: 191-202.

Moore, P. J., and A. J. Moore. 2001. Reproductive aging and mating: The ticking of the biological clock in female cockroaches. Proc Natl Acad Sci USA. 98: 9171-9176.

Morafka, D. J., E. K. Spangenberg, and V. A. Lance. 2000. Neonatology of reptiles. Herpetol Monogr. 14: 353-370.

Mousseau, T. A., and H. Dingle. 1991. Maternal effects in insect life histories. Ann Rev Entomol. 36: 511-534.

Mugabo, M., O. Marquis, S. Perret, and J. F. Le Galliard. 2010. Immediate and delayed life history effects caused by food deprivation early in life in a short-lived lizard. J Evol Biol. 23: 1886-1898.

Nagle, R. D., V. J. Burke, and J. D. Congdon. 1998. Egg components and hatchling lipid reserves: Parental investment in kinosternid turtles from the southeastern United States. Comp Biochem Physiol B.120: 145-152.

Natuhara, Y. 1983. The influences of food and photoperiod on flight activity and reproduction of the bean bug, Riptortus clavatus Thungberg (Heteroptera: Coreidae). Appl Entomol Zool. 18: 392-400.

Naya, D., M. Lardies, and F. Bozinovic. 2007. The effect of diet quality on physiological and life-history traits in the harvestman Pachylus paesslerik. J Insect Physiol. 53: 132-138.

Newcombe, D., P. J. Moore, and A. J. Moore. 2015. The role of maternal effects in adaptation to different diets. Biol J Linn Soc. 114: 202-211.

Nolen, Z. J., P. E. Allen, and C. W. Miller. 2017. Seasonal resource value and male size influence male aggressive interactions in the leaf footed cactus bug, Narnia femorata. Behav Processes 138: 1-6.

112

Nylin, S., and K. Gotthard. 1998. Plasticity in life-history traits. Ann Rev Entomol. 43: 63- 83.

O'Brien, D. M., C. L. Boggs, and M. L. Fogel. 2004. Making eggs from nectar: the role of life history and dietary carbon turnover in butterfly reproductive resource allocation. Oikos. 105: 279-291.

O'Neill, D. J., and J. S. Cobb. 1979. Some factors influencing the outcome of shelter competition in lobsters (Homarius americanus). Mar Behav Physiol. 6: 33-45.

Oku, K., Y. Okutani-Akamatsu, and T. Watanabe. 2010. Effects of Female Age and Ovarian Development on Mating Behavior in Stenotus rubrovittatus (Heteroptera: Miridae). Ann Entomol Soc Am. 103: 802-805.

Packer, C. 1979. Male dominance and reproductive activity in Papio anubis. Anim Behav. 27: 37-45.

Painting, C. J., and G. I. Holwell. 2014. Flexible alternative mating tactics by New Zealand giraffe weevils. Behav Ecol. 25: 1409-1416.

Paterson, J. D. 1973. Ecologically differentiated patterns of aggressive and sexual behavior in two troops of Ugandan baboons, Papio anubis. Am J Phys Anthropol. 38: 641-647.

Paul, M. J., I. Zucker, and W. J. Schwartz. 2008. Tracking the seasons: the internal calendars of vertebrates. Philos Trans R Soc B. 363: 341-361.

Pechenik, J. A., J. N. Jarrett, and J. Rooney. 2002. Relationships between larval nutritional experience, larval growth rates, juvenile growth rates, and juvenile feeding rates in the prosobranch gastropod Crepidula fornicata. J Exp Mar Biol Ecol. 280: 63-78.

Pierotti, R. C. A. Annett. 1990. Diet and reproductive output in seasbirds – food choices by individual, free-living animals can affect survival of offspring. BioScience. 40(8): 568-574.

Plesnar-Bielak, A., K. R. Woch, M. A. Malszycki, A. T. H. Alkhawlany, A. Holysz, J. F. A. Correia, N. Turk, M. Ugrin, P. Kramarz, and Z. M. Prokop. 2017. Larval and adult nutrition effects on reproductive traits in the red flour beetle. J Zool. 302: 79-87.

Proctor, D. S., A. J. Moore, and C. W. Miller. 2012. The form of sexual selection arising from male-male competition depends on the presence of females in the social environment. J Evol Biol. 25: 803-812.

Puurtinen, M., L. Fromhage. 2017. Evolution of male and female choice in polyandrous systems. Proc R Soc B. 284: 20162174

113

R Core Team. 2019. R: A Language and Environment for Statistical Computing. Retrieved from http://www.r-project.org.

Reavey, D. 1992. Egg size, first instar behavior and the ecology of Lepidoptera. J Zool. 227: 277-297.

Richard, M., J. Lecomte, M. de Fraipont, and J. Clobert. 2005. Age-specific mating strategies and reproductive senescence. Mol Ecol. 14: 3147-3155.

Richardson, J., and P. T. Smiseth. 2019. Nutrition during sexual maturation and at the time of mating affects mating behaviour in both sexes of a burying beetle. Anim Behav. 151: 77-85.

Robinson, J. V., L. R. Shaffer, D. D. Hagemeier, and N. J. Smatresk. 1991. The ecological role of caudal lamellae loss in the larval dameslfly, Ischnura posita (Hagen) (Odonata: Zygoptera). Oecologia. 87: 1-7.

Rosen, D. A. S. 2009. Steller sea lions Eumetopias jubatus and nutritional stress: evidence from captive studies. Mamm Rev. 39: 284-306.

Roth, S., and K. Reinhardt. 2003. Facultative sperm storage in response to nutritional status in a female insect. Proc R Soc B. 270: S54-S56.

Sasson, D. A., P. R. Munoz, S. A. Gezan, and C. W. Miller. 2016. Resource quality affects weapon and testis size and the ability of these traits to respond to selection in the leaf-footed cactus bug, Narnia femorata. Ecol Evol. 6: 2098-2108.

Sentinella, A. T., A. J. Crean, and R. Bonduriansky. 2013. Dietary protein mediates a trade-off between larval survival and the development of male secondary sexual traits. Funct Ecol. 27: 1134-1144.

Simmons, L. W., and D. J. Emlen. 2006. Evolutionary trade-off between weapons and testes. Proc Natl Acad Sci USA. 103: 16346-16351.

Simmons, L. W., J. L. Tomkins, and J. Hunt. 1999. Sperm competition games played by dimorphic male beetles. Proc R Soc B. 266: 145-150.

Simmons, L. W., S. Lüpold, and J. L. Fitzpatrick. 2017. Evolutionary Trade-Off between Secondary Sexual Traits and Ejaculates. Trends Ecol Evol. 32: 964-976.

Sisterson, M. S., R. Krugner, C. M. Wallis, and D. C. Stenger. 2018. Effects of Energy Reserves and Diet on Glassy-Winged Sharpshooter Egg Maturation. J Econ Entomol. 111: 159-169.

114

Siva-Jothy, M. T. 1987. Mate securing tactics and the cost of fighting in the Japanese horned beetle, Allomyrina dichotoma L. (Scarabaeidae). J Ethol. 5: 165-172.

Small, J., S. Cotton, K. Fowler, and A. Pomiankowski. 2009. Male eyespan and resource ownership affect contest outcome in the stalk-eyed fly, Teleopsis dalmanni. Anim Behav. 78: 1213-1220.

Sneddon, L. U., F. A. Huntingford, and A. C. Taylor. 1997. Weapon size versus body size as a predictor of winning in fights between shore crabs, Carcinus maenas (L.). Behav Ecol Sociobiol. 41: 237-242.

Somjee, U., C. W. Miller, N. J. Tatarnic, and L. W. Simmons. 2017. Experimental manipulation reveals a trade-off between weapons and testes. J Evol Biol. 31(1): 57-65.

Strathmann, R. R. 1985. Feeding and nonfeeding larval development and life-history evolution in marine invertebrates. Ann Rev Ecol Syst. 16: 339-361.

Suzaki, Y., M. Katsuki, T. Miyatake, and Y. Okada. 2015. Relationships among male sexually selected traits in the bean bug, Riptortus pedestris (Heteroptera: Alydidae). Entomol Sci. 18: 278-282.

Taborsky, B. 2006. The influence of juvenile and adult environments on life-history trajectories. Proc R Soc B. 273: 741-750.

Tang-Martinez, Z. 2016. Rethinking Bateman's Principles: Challenging Persistent Myths of Sexually Reluctant Females and Promiscuous Males. J Sex Res. 53: 532-559.

Tatar, M., and J. Carey. 1995. Nutrition mediates resproductive trade-offs with age- specific mortality in the beetle Callosobruchus maculatus. Ecology. 76: 2066- 2073.

Tauber, M. J., C. A. Tauber, and S. Masaki. 1986. Seasonal adaptations of insects. New York, NY: Oxford University Press.

Thorsen, S. 2020, July 08. Gainesville, Florida, USA — Sunrise, Sunset, and Daylength, 2018. Time and Date AS. https://www.timeanddate.com/sun/usa/gainesville.

Todd, J. W. 1989. Ecology and behavior of Nezara viridula. Ann Rev Entomol. 34: 273- 292.

Tomkins, J. L., and L. W. Simmons. 2000. Sperm competition games played by dimorphic male beetles: fertilization gains with equal mating access. Proc R Soc B. 267: 1547-1553.

115

Ultsch, G. R. 1989. Ecology and physiology of hibernation and overwintering among freshwater fishes, turtles, and snakes. Biol Rev. 64: 435-516.

Umbers, K. D. L., N. J. Tatarnic, G. I. Holwell, and M. E. Herberstein. 2012. Ferocious Fighting between Male Grasshoppers. PlosOne. 7: e49600.

Uyi, O. O., C. Zachariades, L. U. Heshula, and M. P. Hill. 2018. Developmental and reproductive performance of a specialist herbivore depend on seasonality of, and light conditions experienced by, the host plant. PlosOne. 13(1): 1-16.

Vaiserman, A. M. 2014. Early-life nutritional programming of longevity. J Dev Orig Health Dis. 5: 325-338.

Valicente, F. H., R. J. O’Neil. 1995. Effects of host plants and feeding regimes on selected life history characteristics of Podisus maculiventris (Say) (Heteroptera: Pentatomidae). Biol. Control. 5: 449-461.

Van den Beuken, T. P. G., C. C. Duinmeijer, and I. M. Smallegange. 2019. Costs of weaponry: Unarmed males sire more offspring than armed males in a male- dimorphic mite. J Evol Biol. 32: 153-162.

Vessels, H. K., C. S. Bundy, and J. E. McPherson. 2013. Life History and Laboratory Rearing of Narnia femorata (Hemiptera: Heteroptera: Coreidae) With Descriptions of Immature Stages. Ann Entomol Soc Am. 106: 575-585.

Wheeler, D. 1996. The Role of Nourishment in Oogenesis. Ann Rev Entomol. 41: 407- 431.

Wick, J. R., and P. F. Bonhag. 1955. Postemhryonic development of the ovary of Oncopeltus fasciatus (Dallas). J Morphol. 96: 31-60.

Wilder, S. M., and A. L. Rypstra. 2008. Diet quality affects mating behaviour and egg production in a wolf spider. Anim Behav. 76: 439-445.

Wilner, D., E. V. Greenway, L. A. Cirino, and C. W. Miller. 2020. Long-term reproductive success is predicted by sexual behavior and impaired by temporary nutritional stress during sexual maturation. Behav Processes. 175: 104122.

Wittmeyer, J. L., T. A. Coudron, and T. S. Adams. 2001. Ovarian development, fertility and fecundity in Podisus maculiventris Say (Heteroptera: Pentatomidae): an analysis of the impact of nymphal, adult, male and female nutritional source on reproduction. Invertebr Reprod Dev. 39: 9-20.

Wong, D. M., L. S. Chou, S. F. Shiao, and A. Bain. 2018. Body size and injury severity associated with mating strategies in male Philotrypesis taida fig wasps. Taiwania. 63: 227-231.

116

Yamane, T., K. Okada, S. Nakayama, and T. Miyatake. 2010. Dispersal and ejaculatory strategies associated with exaggeration of weapon in an armed beetle. Proc R Soc B. 277: 1705-1710.

Yanagi, S., and M. Tuda. 2012. Female size constrains egg size via the influence of reproductive organ size and resource storage in the seed beetle Callosobruchus chinensis. J Insect Physiol. 58: 1432-1437.

Yasuda, C., Y. Suzuki, and S. Wada. 2011. Function of the major cheliped in male-male competition in the hermit crab Pagurus nigrofascia. Mar Biol. 158: 2327-2334.

Young, B., D. V. Conti, and M. D. Dean. 2013. Sneaker "jack" males outcompete dominant "hooknose" males under sperm competition in Chinook salmon (Oncorhynchus tshawytscha). Ecol Evol. 3: 4987-4997.

Zera, A., J. Potts, and K. Kobus. 1998. The physiology of life-history trade-offs: Experimental analysis of a hormonally induced life-history trade-off in Gryllus assimilis. Am Nat. 152: 7-23.

117

BIOGRAPHICAL SKETCH

Lauren Anne Cirino was born in Alexandria, Virginia in 1984. She was raised in northern Virginia where she attended grade school. During those formative years,

Lauren became interested in science. In high school, Lauren became fascinated with the field of evolutionary biology. She knew from that point on that she wanted to focus her studies in this field. Lauren received her B.S. in biology from Clemson University.

While there, she investigated the effect of male display and species on female mate choice in sailfin mollies. After graduation, Lauren decided to begin teaching science.

She taught secondary school life science for seven years and realized her passion and love of sharing her knowledge of science with others. She continued to share her passion for science as a graduate student as she incorporated and taught undergraduate students entomology, evolutionary ecology, animal behavior, and the research process. She looks forward to continuing to combine her two passions in the future: research and teaching. Lauren received her Master of Science degree from the

University of Florida in the summer of 2016 and her Doctor of Philosophy in the summer of 2020.

118