Gender in Factors Influencing the Infection of the Beetle, Tenebrio Molitor with the Tapeworm, Hymenolepis Diminuta

GENDER IN FACTORS INFLUENCING THE INFECTION OF THE BEETLE, TENEBRIO MOLITOR WITH THE TAPEWORM, HYMENOLEPIS DIMINUTA

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of The Ohio State University

John Francis Shea, B.S.

*****

The Ohio State University 2003

Dissertation Committee: Approved by Professor Jerry F. Downhower, Adviser ______Professor Peter W. Pappas, Co-Adviser Adviser

Professor Tom Waite ______Co-adviser Professor P. Larry Phelan Evolution, Ecology, & Organismal Biology

ABSTRACT

Parasites that require multiple hosts often employ strategies to increase the

probability of transmission to their next host, including altering host behavior. Alteration

of host behavior may occur in the beetle-tapeworm system that I studied.

Various factors influence the transmission of the tapeworm, Hymenolepis

diminuta to its grain beetle host, Tenebrio molitor. Previous studies suggest the presence

of a beetle attractant present in rat feces containing tapeworm eggs, which increases the

probability of tapeworm transmission to the beetle. The results also suggest that various

factors work to influence beetle foraging behavior and its impact on parasite fitness.

When infected beetles increase their feeding activity, they increase the probability of becoming re-infected. If host resources do not limit parasite growth, then increased

feeding activity may increase parasite fitness under some conditions. To simulate

conditions in which host resources are limited, male and female beetles were starved, and

the weight change and frass production were compared between infected and control

beetles. A second experiment provided male and female beetles with food so that the beetle’s weight change, frass production, and food intake were compared between infected and control beetles. Results show that host resources are not limited, but infected beetles do not feed more than uninfected beetles. Instead, male beetles feed more than females suggesting an explanation behind the higher median load of parasites recovered from male beetles.

The preference of individual male and female beetles was tested for infective

versus uninfective feces under varying conditions. When allowed to feed, females

preferred infective feces while males showed lower activity levels. Further, the results suggest that the beetle attractant is found in infective feces instead of the tapeworm eggs.

Infected hosts often behave differently from uninfected hosts. The preference of infected male and female beetles for infective bait was tested. When allowed to feed, neither male nor female beetles exhibited a preference for infective bait. This was true when beetles were tested individually or in groups. These results suggest that beetles, once infected, lose their preference for infective feces. This preference change may be a host response, parasite adaptation, or both.

The preference of groups of uninfected male and female beetles was tested for infective bait. Males avoided infective feces while females showed no preference. These results suggest that the selective pressure to avoid parasitism is stronger in males than in females, which may reflect the higher reproductive cost of infection for males.

iii

Ad Majorem Dei Gloriam

ACKNOWLEDGMENTS

I thank all of my former teachers and advisors who inspired me to pursue a degree

in biology at Ohio State University, especially Dr. Sonja Teraguchi. I thank my advisor,

Jerry Downhower for his intellectual guidance, insightful advice and enthusiastic support.

This dissertation would not have been possible without his mentorship. I thank Pete

Pappas for his encouragement, sense of humor and for the use of his lab. I also thank him for providing the critters used in this research.

I thank my committee members, Tom Waite and Larry Phelan, for their input and invaluable comments. Tom’s enthusiasm and hard work while bearing physical injury, was a great source of inspiration and encouragement for me. Larry’s provided helpful and thought provoking comments on previous drafts of this dissertation.

I also thank the Culver lab for the use of their balance, computers, software, and other equipment. I especially thank them for the sense of welcome I felt whenever I entered their lab. Thanks also to Cathy Doyle for the use of her laptop and to Kwee Tew for his help in outsmarting Microsoft Word.

Finally, thanks to all of the EEOB graduate faculty, staff, and students who filled my graduate student life with stimulating intelligence and, most importantly, warm friendship. I also thank my mom and Ben Russo for their help in data entry and proofreading.

VITA

October 13, 1974………………………Born – Cleveland, OH

1997……………………………………B.S. Biology, John Carroll University.

1997-present……………………………Graduate Teaching and Research Associate, The Ohio State University

FIELDS OF STUDY

Major Field: Evolution, Ecology, & Organismal Biology

TABLE OF CONTENTS

Page Abstract…………………………………………………………………. ii

Dedication………………………………………………………………. iv

Acknowledgments………………………………………………………. v

Vita……………………………………………………………………… vi

List of Tables……………………………………………………………. ix

List of Figures…………………………………………………………… xi

Chapters:

Introduction……………………………………………………………… 1

1. Feeding activity and infection intensities of male and female beetles (Tenebrio molitor) with cysticercoids of the tapeworm, Hymenolepis diminuta...... ………………………………………………………. 6 Abstract………………………………………………………….. 6 Introduction……………………………………………………… 7 Methods…………………………………………………….....…. 10 Results…………………………………………………………… 13 Discussion……………………………………………………….. 16

2. Parasite-altered preference for infective feces in uninfected individual hosts…………………………………………………………... 40 Abstract…………………………………………………………… 40 Introduction…………………………………………………....…. 41 Methods………………………………………………………….. 45 Results……………………………………………………………. 51 Discussion………………………………………………………… 54

3. Parasite-altered preference for infective feces in infected individual hosts and groups of infected hosts…………………………….. 78 Abstract…………………………………………………………… 78 Introduction………………………………………………………. 79 Methods………………………………………………………….. 82 Results……………………………………………………………. 88

vii

Discussion………………………………………………………… 90

4. Gender differences in susceptibility to infection in groups of uninfected hosts………………………………………………………… 106 Abstract…………………………………………………………… 106 Introduction………………………………………………………. 107 Methods………………………………………………………….. 109 Results……………………………………………………………. 112 Discussion………………………………………………………… 112

Integrative Discussion……………………………………………………. 120

Appendix A Description of beetles used in the feeding experiment…….. 126

Bibliography……………………………………………………………… 128

viii

LIST OF TABLES

Table Page 1.1 Summary of predicted ramifications for transmission dynamics under conditions of both limiting and unlimiting host resources…………………………………………. 24

1.2 Average beetle weights and average total proportional weight change for beetles in the fed and starved experiments.………………………………………………………. 25

1.3 Average frass count for the starved beetle experiment and average frass count and pellet consumption for the fed beetle experiment.………………………………………………… 26

1.4 Observed beetle response to parasitism when host resources are known not to limit parasite growth and development, and its ramifications on parasite and host fitness……………………… 27

1.5 Expected mean beetle weight after adjusting for humidity……….. 28

2.1 Description of the 13 experiments performed…………………….. 61

2.2 Bait preference results from experiments 1 and 2………………… 62

2.3 Bait preference results from experiment 1a………………………. 63

2.4 Bait contact results from experiments 1, 2 and 1a………………... 64

2.5 Bait preference results from experiments 3 and 4………………… 65

2.6 Bait preference results from experiment 5………………………… 66

2.7 Bait preference results from experiments 6 and 7………………… 67

2.8 Bait preference results from experiments 8 and 9………………… 68

2.9 Bait preference results from experiment 10……………………….. 69

2.10 Bait contact results from experiments 3, 4, 5 and 6………………. 70

Table Page

2.11 Bait contact results from experiments 7, 8, 9 and 10……………… 71

2.12 Bait preference results from experiment 11………………………. 72

2.13 Bait contact results from experiments 11, 12 and 13……………… 73

2.14 Bait preference results from experiments 12 and 13……………… 74

3.1 Bait preference results from individual beetle experiment……….. 95

3.2 Bait contact results from individual experiment………………….. 96

3.3 Bait preference data from group beetle experiment………………. 97

3.4 Mixed procedure results of infected beetle preference for infective bait by sex, and for infective bait by sex with cysticercoid number in the group beetle experiment……………… 98

3.5 Mixed procedure results of uninfected beetle preference for infective bait by sex in the group beetle experiment……………… 99

4.1 Mixed procedure results for beetle preference by sex for the bait area, and for the non-bait area in group trials……………. 116

4.2 Total number of beetles at each bait type summed over the entire 60 minute trial with mean number of beetles……………… 117

4.3 Total number of beetles in the non-bait area for each half of the arena summed over the entire 60 minute trial with mean number of beetles…………………………………………… 118

LIST OF FIGURES

Figure Page

1.1 Diagram showing foraging possibilities for a beetle infected with a cysticercoid……………………………………… 29

1.2 Weight gain on day one of starved beetles……………………… 30

1.3 Proportional total weight loss in starved beetles………………… 31

1.4 Frass production in starved beetles……………………………… 32

1.5 Average frass production in starved control and infected females…………………………………………………………… 33

1.6 Frequency distribution of the number of cysticercoids dissected from starved female and male beetles…………………………… 34

1.7 Weight gain in fed beetles on day one………………………….. 35

1.8 Proportional total weight loss in fed beetles…………………….. 36

1.9 Frass production in fed beetles………………………………….. 37

1.10 Pellet consumption in fed beetles……………………………….. 38

1.11 Water weight loss over time in beetles kept at 40% and 90% relative humidities………………………………………………. 39

2.1 Picture of plastic arena used in preference trial experiment…….. 75

2.2 Observed standard deviation for each of the trials showing their relative deviations from expected binomial distributions with indicated k values………………………………………….. 76

2.3 Observed standard deviation for experiments with significant results showing their relative deviations from expected binomial distributions with indicated k values…………………. 77

Figure Page

3.1 Diagram of test arena used in the individual beetle experiment… 100

3.2 Diagram of test arena used in the group beetle experiment…….. 101

3.3 Frequency distribution of the total number of cysticercoids dissected from female and male beetles in the individual beetle experiment………………………………………………………. 102

3.4 Data from individual trials showing the number of infected and re-infected beetles after two exposures to tapeworm eggs……… 103

3.5 Plots of the average number of early stage cysticercoids and stage 5 cysticercoids against its variance for both male and female beetles of the group trial………………………………… 104

3.6 Data from group trials showing the number of infected and re-infected beetles after two exposures to tapeworm eggs……… 105

4.1 Diagram of test arena used in the group beetle experiment…….. 119

xii

INTRODUCTION

Several studies suggest that parasites alter the behavior of their hosts so as to

increase the probability of transmission to the next host (Holmes and Bethel 1972; Bethel

and Holmes 1973; Moore 2002). For example, ants infected with the digenetic

trematode, Dicrocoelium dendriticum will climb atop and cling to blades of grass where

they are more likely to be eaten by grazing sheep, the definite host (Anokhin 1966;

Carney 1969). Also, amphipods infected with encysted (infective) trematodes are more

likely to be crawling on the ground than burrowing underground, making them

susceptible to consumption by birds which feed during the day (McCurdy et al. 1999).

Knowledge about factors that influence transmission probability is important because it allows epidemiologists to model and predict disease virulence (Ewald 1983).

In the tapeworm, Hymenolepis diminuta, adults live in the small intestines of rodents, the definitive host. Eggs pass out of the rodents in their feces. The cysticercoid, or larval stage, occurs in an insect (the intermediate host, most often a beetle) that became infected by feeding upon infective feces. The rodent completes the cycle when it ingests an infected beetle. Thus, the success of this parasite depends on rodent-to-beetle and beetle-to-rodent transmission. Rodent-to-beetle transmission depends upon consumption of the egg stage by the beetle for successful transmission.

The aim of my dissertation is to study how sex, nutritional status, infection status

and the presence of conspecifics influence this rodent-to-beetle transmission of the

2 tapeworm. These factors relate to the beetles’ probability of infection, susceptibility to

infection and cost of infection.

Infection by H. diminuta cysticercoids harms the beetle host in a variety of ways.

Infected male T. molitor exhibit a lowered response to pheromone produced by

uninfected females (Hurd and Parry 1991), while defensive glands of infected T. molitor contain fewer defensive compounds than uninfected controls (Blankespoor et al. 1997).

Fat body glycogen reserves are significantly depleted three days post-infection in male and five days post-infection in female T. molitor (Kearns et al. 1994), suggesting that parasitism has a nutritional effect on the host the extent of which may differ by sex. In terms of reproduction, reduction in host fecundity occurs in both Tribolium confusum

(Keymer 1980, 1981) and Tenebrio molitor (Hurd and Arme 1986). Also, infected males

are less attractive to females, and females mated with infected males produce fewer

offspring than females mated with uninfected males (Worden et al. 2000). When results

from these studies are quantified, they again suggest differential costs to parasitism

between males and females.

When males mated singly with infected females produced an average (n = 30) of

115.2 ovulated eggs, while males mated with uninfected females produced an average (n

= 30) of 127.9 ovulated eggs (Hurd and Arme 1986). Females mated singly with infected

males produce an average (n = 7) of 37 larvae while females mated with uninfected males

produce an average (n = 13) of 63.7 larvae (Worden et al. 2000). Considering that

tapeworm infected males (T. molitor) are less attractive to females (Worden et al. 2000),

these results suggest that the cost to infection is higher in males. Thus, selection to avoid

infective feces may act stronger on males.

2 The reduction in fat body glycogen reserves suggests that parasitism may impact

male and female hosts nutritionally, which is the subject of Chapter 1. Infected hosts

may consume more food than uninfected hosts which will increase the probability of re-

infection. Under some conditions, this may impact the parasite. Alternatively, feeding

activity of infected hosts may not change or decrease, which would not increase the

probability of re-infection. The probability of infection can change in other ways;

specifically, at the passive rodent-to-beetle transmission stage.

Change in probability of rodent-to-beetle transmission was first demonstrated when a group of beetles (T. confusum) were allowed to forage in an arena with two types of food; rat feces from uninfected rats and rat feces from rats infected with H. diminuta

(presumed to contain eggs). The results show that beetles, starved 48 hours prior to the trial, preferred infective feces (Evans et al. 1992). A similar experiment using beetles (T.

molitor) of known sex, age and feeding history showed that female beetles (both fed and

those starved 72 hours prior to the trial) preferred infective feces but starved male beetles preferred infective feces and fed males preferred uninfective feces (Pappas et al. 1995).

These experiments suggest that gender and feeding history influences a beetle’s

preference for infective or uninfective feces.

The mechanism for a beetle’s preference may be the presence of a volatile

attractant in infective feces, which was collected on a solid adsorbent by aspirating fresh

rat feces from infected and uninfected rats (Evans et al. 1998). Preference tests showed

that more beetles (T. confusum) were attracted to volatiles from infective feces. The

volatile attractant has not been identified and may originate from the tapeworm eggs.

Alternatively, the attractant may be present in the infective feces itself if secretions from

3 the adult tapeworm contact feces before defecation to create a beetle attractant. Finally, the attractant may result from a host immunological reaction to the adult tapeworm.

An analysis of beetle (T. confusum) movement showed that starved beetles (of mixed sex and age) made fewer visits to infective bait, but the duration of visits to each bait did not differ when tested individually. When tested in groups during a two hour trial, starved and fed beetles preferred control bait during two time intervals, but showed no preference during the other four time intervals. The authors concluded that beetle behavior appeared “highly heterogeneous, both among individuals and by the same individual over time” (Shostak and Smyth 1998). The presence of behavioral variability among individuals is a necessary component for the co-evolution of host and parasite.

Chapter 2 continues the previous research, but tests bait preference of individual beetles alone and in the presence of another beetle. Selection acts on the individual and the presence of aggregation pheromones may confound beetle behavior. Chapter 2 also tests beetle preference for bait of varying quality while separating olfactory and gustatory cues. Finally, it tests beetle preference in the presence of tapeworm eggs mixed in uninfective feces versus uninfective feces alone to ask if the adult tapeworm plays a role in altering beetle fecal preference.

Beetle behavior may change after infection (Hurd and Fogo 1991; Robb and Reid

1996). Chapter 3 tests the bait preference of infected beetles – both individually and in groups. Its novel approach allows for comparisons between initial and secondary infections. Infected individuals may be more or less susceptible to a second infection.

The chapter also returns to the study of gender biased infection loads.

4 Males and females often differ in their susceptibility and exposure to infection.

They also differ in the cost they pay when parasitized as a consequence of reproductive

and immunological differences. Thus, they may also differ in their ability to avoid

infection. Using unisex groups of starved beetles, Chapter 4 tests the fecal preference of

male and female beetles.

CHAPTER 1

FEEDING ACTIVITY AND INFECTION INTENSITIES OF MALE AND FEMALE BEETLES (TENEBRIO MOLITOR) WITH CYSTICERCOIDS OF THE TAPEWORM, HYMENOLEPIS DIMINUTA

ABSTRACT

Parasitism results in nutritionally-related changes in hosts, often leading to altered feeding behavior. Infected hosts that increase their feeding also increase their probability of re-infection and, depending upon host resource availability, can impact the parasite.

To study this, a beetle (Tenebrio molitor)-tapeworm (Hymenolepis diminuta) system was used. To simulate limited resources, infected and uninfected male and female beetles were individually housed in vials without food. Each beetle’s weight change and frass production were measured over 24 h periods at 3, 7, 12 and 16 days post-infection.

Another experiment housed beetles in vials with food and their weight change, frass production, and food intake were measured. In both experiments, treatment (infection) had no effect on weight change, but males lost more weight and produced more frass than females. Starved, infected females produced more frass than control females. Treatment had no effect on food consumption, but males had a higher food intake than females.

These results suggest that infection status will not alter the probability of re-infection, but males will be more susceptible to infection than females. This greater susceptibility to infection may explain why males had a higher median cysticercoid infection level.

6 INTRODUCTION

Parasitic infections result in nutritionally-related pathology, including altered food

consumption (Thompson 1983). For example, invertebrates often reduce food intake

when infected with certain pathogens, as exemplified by locusts infected with a fungus

(Moore et al. 1992; Seyonum et al. 1994), grasshoppers infected with a microsporidian

(Johnson and Pavlikova 1986), and beet armyworms infected with a fungus (Hung and

Boucias 1992). Enhanced growth occurs in mollusks infected with digenean trematodes

(Sorenson and Minchella 1998) as a result of host castration. Greater weight gain also

occurs in three-spined stickleback fish infected with a cestode and, although not tested,

may result from increased food intake (Arnott et al. 2000). If infection does increase

food intake, then the host would be more susceptible to re-infection, which would have ramifications on the parasite.

The tapeworm, Hymenolepis diminuta, uses a grain beetle, Tenebrio molitor, as

its intermediate host to facilitate transmission to the rat definitive host. Amino acid

concentrations change in T. molitor grain beetles infected with the tapeworm H. diminuta

(Hurd and Arme 1984). In females, these changes include increased concentrations of

isoleucine, leucine, arginine, serine, and threonine, and decreased concentrations of

tyrosine, phenylalanine, praline, and alanine/citrulline in the hemolymph. In males,

concentrations of threonine and glycine increase, while concentrations of histidine and

arginine decline. Fat body glycogen reserves are depleted 3 days post-infection in male

and 5 days post-infection in female T. molitor infected with cysticercoids of H. diminuta

(Kearns et al. 1994). These results suggest that parasitism may have a nutritional effect

7 on the infected beetle the extent of which may differ by sex. Further, this nutritional effect may lead to increased susceptibility to re-infection, which would impact the parasite (Table 1.1).

Beetles risk infection when feeding upon infective rat feces. Coprophagy, the ingestion of feces, can be divided into two types, the ingestion of conspecific feces which occurs in gastropods (Brendelberger 1997) and cockroaches (Kopanic et al. 2001), and the ingestion of heterospecific feces, which occurs in moths feeding on feces from a variety of organisms (Sanchez-Pinero and Perez-Lopez 1998). Although little studied, T. molitor ingests heterospecific feces, which may be a way for the beetle to supplement its diet with water, specific nutrients, or both. For example, rats acquire cystine, thiamine and vitamin K through coprophagy (Soave and Brand 1991). Larvae of T. molitor do require an external source of carnitine, a B vitamin (Chapman 1998), so perhaps larval and adult beetles acquire these nutrients by eating rat feces.

Any change in host feeding behavior could have ramifications on parasite transmission (Table 1.1). First consider ramifications when host stored resources limit parasite growth and development. If host feeding increases, then host susceptibility to further infection also increases. When fed infected terrestrial isopods, the heaviest starling nestlings have the most parasites (Moore and Bell 1983). Such re-infection would be disadvantageous to the parasites, which must compete for limited resources.

Thus, if parasites were capable of manipulating host behavior, we would not expect to see infected beetles increasing their feeding because it could disadvantage the parasite.

Alternatively, if host feeding does not change or is reduced then the probability of re- infection would not change and there would be no impact on the parasite.

8 Now consider what occurs if host resources do not limit parasite growth and development. The probability of re-infection will increase if host feeding increases.

However, this would not disadvantage the parasite. Although the number of parasites increases with each host re-infection, the parasite remains unaffected. In fact, parasite inclusive fitness could increase if the host is re-infected by parasites with a sufficiently high coefficient of relatedness to the original parasites (Fig 1.1), but see Caraco and

Giraldeau (1993). Note that this scenario considers only one host and ignores possible costs to a parasite from being in a dead-end host. Alternatively, if host feeding does not change or is reduced, then the probability of re-infection will not change and there will be no impact on the parasite.

Finally, the number of parasites present in a host can alter the host’s metabolic rate, the host’s resource availability, and ultimately, transmission dynamics. An infection may range from a near fatal dose of parasites to an imperceptibly light infection with density-dependent effects observed in this host-parasite system (Yan et al. 1994; Hurd

1998). Realizing this, I attempted to infect the beetles with a consistent intensity of parasites.

To simulate conditions in which food resources are limited, male and female beetles were starved, and the weight change and frass production of infected and control beetles were compared. A second experiment provided male and female beetles with food pellets so that the beetle’s weight change, frass production, and food intake

(measured as the change in food pellet weight) could be compared between infected and control beetles.

9 METHODS

The “OSU Strain” (Pappas and Leiby 1986) of H. diminuta was maintained in grain beetles (T. molitor) and male Sprague-Dawley rats. Beetles were maintained on wheat bran to which small pieces of potato were added on a regular basis. Pupae were removed and male and female pupae (Bhattacharya et al. 1970) were placed in separate dishes containing wheat bran. Male and female beetles that emerged during a 24 h period were collected (13 beetles of each sex per day), maintained at 26°C until they were 9 days old, and randomly assigned into control or experimental groups (Appendix A). All beetles were marked individually with latex paint, starved for 24 h and then weighed.

Unisex groups of experimental beetles (n♀=13, n♂=13) were allowed to feed on 1.5g of

air-dried apple scrapings mixed with a 0.05 ml solution of water and tapeworm eggs for

24 h (= Day 1). Unisex groups of control beetles (n♀=13, n♂=13) were allowed to feed

on 1.5g of air dried apple scrapings mixed with a 0.05 ml solution of distilled water for

24 hours.

Starved beetles

Each beetle was weighed after feeding on Day 1 and placed in a separate glass

vial, without food or water, and maintained at 26°C and constant humidity (90% R.H.).

A total of 208 beetles was exposed to apple scrapings mixed with either tapeworm eggs

(infected) or water (control) as described above. A total of 131 beetles survived until

Day 17 (infected: n♀=32, n♂=31 and control: n♀=36, n♂=32). All infected beetles were

dissected, and the numbers of cysticercoids were recorded.

At three days post-infection (= Day 3), each beetle was weighed, and the frass present in the vial was counted. On Day 4, each beetle was weighed again, and the frass

10 present in the vial was counted. This process was repeated for days 7-8, 12-13, and 16-

17. Cysticercoids require 12-16 days to become fully infective, so host changes should

be most noticeable during this time (Hurd and Arme 1987; Voge and Heyneman 1957).

Fed beetles

After feeding on Day 1, each beetle was weighed and placed with a weighed

pellet of commercial rat chow in a separate glass vial, which was maintained at 26°C and

approximately 40% R.H. (an experiment at 90% R.H. was unfeasible because of mold

and lack of cohesion of the food pellet). A total of 260 beetles was exposed to apple

scrapings mixed with either tapeworm eggs (infected) or water (control) with a total of

119 beetles surviving until Day 17 (infected: n♀=31, n♂=27 and control: n♀=29, n♂=32).

All infected beetles were dissected, and the numbers of cysticercoids were recorded.

At three days post-infection, each beetle was weighed and the pieces of frass present in the vial were counted. The remaining food was also weighed. Then, the food pellet by itself (without crumbs) was weighed and returned to the vial with the beetle.

After 24 h (Day 4), beetles, pellets, and food crumbs were weighed again, and the frass present in the vial was counted. The pellet was then returned to the vial with the beetle.

This process was repeated for days 7-8, 12-13, and 16-17.

Humidity Differences

Since it was not possible to conduct both experiments at the same humidity, groups of 10 beetles were kept at 40% and 90% R.H.’s for 6 days without food to assess water loss at the two humidities. They were then placed in a drying oven to obtain their

11 dry weights. The rates of water loss were then calculated and compared. This study assumed that infection with H. diminuta did not affect water regulation of T. molitor maintained at 26°C (Granath 1980).

Statistical Analysis

Differences in beetle weight and frass count after each 24 hour period were calculated for each beetle and analyzed for statistical differences between control and infected beetles. Proportional total weight change data were transformed with the arcsine square root. The General Linear Model in Minitab (v. 13.1, State College, PA) was used to measure effects of treatment, sex, and the interaction between sex and treatment. If the treatment effect was significant, the repeated measures analysis program of SAS (v. 8,

Cary, NC) was used to determine the source of the difference with the additional factor of time (days). For starved beetles, an ante-dependence co-variance model (which allows for unequal variances, correlations and covariances such that correlations decrease through time) was fitted to the data to determine significant differences in the least squares means of the relevant combinations of sex, infection status and day effects. The coefficient of linear correlation (r) was used to determine the correlations of weight change and frass count with the number of cysticercoids, and the correlations of food intake with frass production in fed beetles. The Fligner-Policello test (Hollander and

Wolfe 1999) measured differences in the medians between male and female cysticercoid loads without assuming equal variances.

12 RESULTS

The infection protocol resulted in a range of 1-223 cysticercoids and a mean infection of 44.7 cysticercoids per beetle (n = 124). This infection level is higher than

that reported in Rau’s (1979) field study, who found a mean infection of 10.5

cysticercoids per beetle in a natural beetle population, but is comparable to other

experimental studies (Hurd and Arme 1984; Hurd and Fogo 1991). Fed beetle mortality

was 52% (22% control and 30% infected beetles), and there was no treatment effect on

mortality (Appendix A, 2-tailed Fisher’s exact test, p = 0.2041). Starved beetle mortality

was 35% (13.4% control and 21.4% infected beetles), and there was no treatment effect

on mortality at p = 0.05 (Appendix A, 2-tailed Fisher exact test, p = 0.077). To confirm

that treatment had no effect on mortality, the data were re-analyzed using the available

measurements from beetles that did not survive the entire experiment. However, all data

presented here are from the surviving beetles. Additionally, mortality of fed males (73 of

132) did not differ from fed females (55 of 115) prior to day 17 (2-tailed Fisher’s exact

test, p = 0.2531), but fewer starved females (12 of 80) than starved males (58 of 121) died

(2-tailed Fisher’s exact test, p < 0.01) prior to Day 17.

Starved Beetles

Control females weighed an average of 13.0 mg more than infected females

(GLM, p < 0.001 with Bonferroni post hoc comparison) and 12.7 mg more than control

males (GLM, p < 0.001 with Bonferroni post hoc comparison) at the start of the

experiment, but did not differ from infected males (Table 1.2). Males gained

significantly more weight (5.72 mg) than females during the Day 1 feeding period

independent of infection status (Fig. 1.2). While there was no effect of infection on total

13 proportional weight loss from Day 1 to Day 17, males lost significantly more weight than

females during the course of the experiment (Fig. 1.3).

Infected females produced significantly more frass than control females (Table

1.3 and Fig. 1.4). On further analysis, infected females passed significantly more frass on

Days 13-16 than uninfected controls (Fig. 1.5). Also, male controls produced more frass than female controls (Fig. 1.4). There were no differences in frass production between infected and control males. Frass production was unrelated to the number of cysticercoids recovered from infected male (r = 0.219, p = 0.237) and female (r = -0.137, p = 0.456) beetles.

Male beetles had a significantly higher median load (61) of cysticercoids than female (37.5) beetles (Fig. 1.6; Fligner-Policello, n♂ = 31, n♀ = 32, U = 2.39, p = 0.0085).

However, weight gain during infection was independent of the number of cysticercoids

recovered from males (r = 0.169, p = 0.363, n = 31) and females (r = 0.239, p = 0.189, n

= 32). Over the course of the experiment, males with greater numbers of cysticercoids

lost more weight (r = -0.359, p = 0.047, n = 31), but weight loss in females was unrelated

to the number of cysticercoids recovered (r = 0.129, p = 0.48, n = 32).

Fed Beetles

The initial weights of beetles at the start of the experiment in both treatments for

each sex did not significantly differ (Table 1.2, GLM, p > 0.50 for sex and treatment).

Like their starved counterparts, fed males gained significantly more weight (2.71 mg)

than females during the experimental infection period (Fig. 1.7), independent of infection

status. There was no effect of infection status on total proportional weight loss from Day

14 1 to Day 17, but the proportional weight loss in males was greater than in females during

the course of the experiment (Fig. 1.8).

Although there were no differences in frass production between control and

infected beetles, males (infected and control) produced significantly more frass than

females (Table 1.3 and Fig. 1.9). Frass production was unrelated to the number of cysticercoids in females (r = -0.293, p = 0.11, n = 31) but was positively correlated in

males (r = 0.574, p = 0.002, n = 27).

Infection status had no effect on food consumption, but males (infected and

control) consumed more of the food than did females (Tables 1.3 and Fig. 1.10). Frass

production increased with food intake for each treatment in both males (rinfected = 0.641, n

= 27, p < 0.001; rcontrol = 0.649, n = 32, p < 0.001) and females (rinfected = 0.506, n = 31, p

= 0.004; rcontrol = 0.627, n = 29, p < 0.001). For infected males, increased frass

production was correlated with greater total absolute weight loss (rinfected = -0.449, n = 27, p = 0.0019; rcontrol = -0.29, n = 32, p = 0.108).

The median load of cysticercoids recovered from males did not differ from the

median load recovered from females (Fligner-Policello, n♂ = 27, n♀ = 31, U = 0, p >

0.50). Further, weight gain during infection was not related to the number of cysticercoids recovered from females (r = -0.036, n = 31, p = 0.847), but was marginally

correlated in males (r = 0.373, n = 27, p = 0.055). Total weight loss was not related to

the number of cysticercoids for male (r = -0.108, n = 27, p = 0.592) or female (r = 0.33, n

= 31, p = 0.07) beetles.

Humidity differences

15 Beetles kept at 40% R.H. lost water at a rate of 0.12 mg per hour (n = 10, F1, 28 =

11.48, p = 0.002) while beetles kept at 90% R.H. lost water at a rate of 0.05 mg per hour

(n = 10, F1, 28 = 2.02, p = 0.166) over the course of 144 hours (Fig. 1.11).

DISCUSSION

Starved beetles

Infected males do not differ from control males in terms of weight change or frass production, suggesting that host resources do not limit parasite growth in males. Males consistently consume more food on the day of infection than females indicating a higher gut capacity or a faster feeding rate (Fig. 1.2). Males also lose more weight during the course of the experiment, which suggest that males may store fewer energetic resources than females (Fig. 1.3). Also, control males produce more frass than control females

(Fig. 1.4). These observations, in combination with the male’s greater food consumption on the day of infection, suggest that male beetles have a higher metabolic rate. This is consistent with literature indicating invertebrate males such as Daphnia (MacArthur and

Baillie 1926) and houseflies (Edwards 1946) have higher metabolic rates, and predicts that males will have a larger food intake than females.

Mortality between fed males and females does not differ; these data also agree with those of the controls in Soltani’s (1984) study of the effect of pesticide on the longevity of fed T. molitor. The lower mortality in starved females may result from an ability of females to draw upon greater reproductive resources and live longer.

Infection status affects frass production in starved females, but infected females produce more frass than controls (Fig. 1.4). Because metabolic processes result in waste production, a high metabolism will be correlated positively with high fecal production

16 (Downer 1981). Assuming equal assimilation efficiencies, this suggests that infection

increases host metabolic rate. Given that control females have higher initial weights,

they should produce more frass, but do not. This physiological difference is most

apparent on Days 13-16 (Fig. 1.5) during the time of cysticercoid maturation, which

occurs 12-16 days post-infection (Hurd and Arme 1987; Voge and Heyneman 1957).

The difference not only coincides with behavioral changes observed by Hurd and Fogo

(1991), but also with the observed reduction in vitellogenin proteins in the ovaries of infected females and the corresponding increase of these proteins in the hemolymph at 12 days post-infection (Hurd and Arme 1986; Webb and Hurd 1996; Hurd and Webb 1997;

Hurd 1998). Recent evidence shows that the cysticercoids of H. diminuta produce a

chemical that actively suppresses production of vitellogenin, which results in reduced

host fitness (Webb and Hurd 1999). Not only does the parasite benefit nutritionally, but the parasite-induced shift of metabolic resources may have the additional effect of extending the host’s lifespan, resulting in an increased probability of transmission to the rat host (Hurd 1998; Webb and Hurd 1999; Hurd 2001). In addition, there is some evidence that the host acts to reduce fecundity by inhibiting the uptake of yolk protein by

the ovary (Major et al. 1997), which may serve to increase the lifespan of the host and, in

turn, its lifetime reproductive output (Hurd et al. 2001, Hurd 2001). Over the course of

30-40 days, experimentally infected beetles do survive longer, but only after day 12 when

the cysticercoids have matured into infective stages (Hurd et al. 2001), which may

explain why infected beetles were not observed to survive longer in this study.

The statistically significant higher intensity of cysticercoids in starved males (Fig.

1.6) may indicate that males are more susceptible to infection than females or it may

17 result from ingesting more food. Males, regardless of infection status, do gain significantly more weight on Day 0, indicating that they consume more food on that day.

Males that eat more on Day 0 may increase their probability of acquiring more cysticercoids, but this is not supported statistically. No linear correlation between weight gain and number of cysticercoids in male beetles exists (r = 0.169, n = 31, p = 0.363).

The male infection bias may result from a difference in foraging behavior between the sexes (see Chapter 2).

If males have fewer stored energetic resources, then it may explain why we observe a significant relationship between absolute total weight loss and the number of cysticercoids in starved males (r = -0.359, p = 0.047), but not in fed males or in females.

The dual constraints of starvation and infection may have made the relationship more apparent in the sex with the fewer stored resources. Also, frass production does not correlate with number of cysticercoids in either sex. This failure to see a dose response to parasitism may be related to the beetles’ starved conditions.

Fed beetles

Infection had no effect on weight change or frass production. Males lost more weight than females during the course of the experiment (Fig. 1.8) and produced more frass than females (Fig. 1.9), suggesting a higher metabolism than females.

No treatment difference was found in overall food pellet consumption, indicating that beetles did not increase their feeding when infected. However, male beetles consistently consumed more food on the day of infection (Fig. 1.7) and consumed more of the food pellet (Fig. 1.10). Further, there was a marginally significant relationship between food consumed on the day of infection and number of cysticercoids recovered

18 from male beetles (r = 0.373, p = 0.055). If males feed more, then they will have a higher

probability of acquiring additional cysticercoids and becoming re-infected. However, males that feed more may gain other advantages (more mating opportunities, for

example), suggesting an optimal trade-off between risking infection and being well fed.

Since food intake and frass production both indicate metabolic rate and others have used frass production to assess feeding (Thomas et al. 1997), results showing a

significant relationship between the two are not surprising. Further, the significant

correlation between frass production and number of cysticercoids in males (r = 0.574, n =

27, p = 0.002) suggests that male metabolism increases in response to increased loads of parasites. This dose response is not seen in fed females although starved infected females produced significantly more frass than control females. This may be due to a female’s ability to partition energetic resources between reproduction and somatic maintenance.

When infected, this resource partitioning may benefit both the female host and parasite

(Hurd 1998; Hurd et al. 2001; Hurd 2001). Metabolism would increase with infection, but a dose response would only be observed in the sex unable to partition energetic resources.

In infected fed males, increased frass production resulted in greater weight loss (r

= -0.449, n = 27, p = 0.019), but did not do so for control fed males, any of the fed females, or any of the starved beetles. There should be a strong relationship between weight loss and frass production for all beetles regardless of sex or nutritional status.

Perhaps this relationship is less obvious in starved beetles if they respond to starvation by physiologically lowering their metabolic rate. It may also be less obvious in females, which may redirect reproductive resources to mitigate body weight loss. Fed males have

19 the highest metabolic rates (Fig. 1.9), which may explain why frass production is more

apparently related to weight loss in infected fed males, but not in control males.

Experiment and humidity differences

It was not possible to conduct both experiments at the same time and so

comparisons between fed and starved experiments are difficult because they differ not

only in the obvious nutritional status of the beetle, but in two other important ways. First,

fed beetles were, on average, lighter than starved beetles (Table 1.2). This may result

from the fact that the fed experiment was done before the starved experiment and the

average weight of the beetle population shifted. Second, due to constraints involving the

food pellet, the two experiments differed in relative humidity with the starved experiment

being conducted at the higher 90% R.H. Despite these differences, these results warrant

further discussion.

First, all beetles lost weight over the course of the experiment. After pupation and

sclerotization, adult beetles do not undergo further molting or growth resulting in weight

loss. The closely related adults of Tribolium castaneum and Tribolium confusum lost

weight during a 48 hour period while being fed and maintained at 33°C and 70% R.H.

(Gandhi and Goswami 1993).

Second, fed beetles lost 15% more body weight than starved beetles. It is unlikely that the 11% difference in their initial weights had a significant effect on the overall weight change during the 17 days of the experiment. It is more likely that the fed beetles, kept at the lower humidity, suffered a greater water loss than starved beetles. Such a water loss is consistent with findings from Saleem and Shakoori (1986), who found that starved T. castaneum adults kept at lower humidities lost a greater percentage of their

20 body weight compared to beetles kept at higher humidities. In this experiment, when

beetles are kept at 40% R.H., they lose water at a rate of 0.12 mg per hour while beetles

kept at 90% R.H. lose water at a rate of 0.05 mg per hour (Fig. 1.11). This two-fold difference in water loss is consistent with my data. Fed beetles, kept at 40% lost 28% of their body weight after 17 days while starved beetles, kept at 90% lost only 13% of their body weight. When this water loss is corrected for (Table 1.5), starved males lost more weight than females (p = 0.001) and fed males gained less weight than females (p =

0.003). Infection status had no effect on proportional weight change for starved (p =

0.303) or fed (p = 0.598) beetles.

Alternatively, the weight loss difference may be due to a physiological response by beetles to conserve energy when food is not available. Researchers have found that

Tenebrio larvae decrease their metabolism when starved (Buxton 1930), as do crabs

(Vinagre and Da Silva 2002) and amphipods (Hervant et al. 1999).

Overall, I can conclude that infection does result in a higher metabolic rate in starved females and fed males, but does not lead to increased feeding. The failure to observe a treatment effect on frass in starved males may be confounded by physiological responses to starvation. Frass production increased with number of cysticercoids only in fed beetles, presumably because of their nutritional status. This was observed only in fed males because, unlike females, they lack large reproductive energetic resources, which can be partitioned in response to parasitism. Several explanations for the observed increase in metabolism in infected beetles are possible. To establish an infection, the larval tapeworm must penetrate the host gut resulting in epithelial damage (Lethbridge

1971a). Repairing this damage may drain some metabolic resources (Hurd et al. 2001).

21 Also, H. diminuta undergoes several distinct developmental stages within the host’s hemocoel (Voge and Heyneman 1957), which may drain host resources and contribute to increased metabolism. Finding significant frass count differences in starved females during the time of parasite development supports this explanation. Alternatively, parasite presence may elicit an immune response from the host, which in turn, may drain host resources. Hemocytes defend the beetle against the larval tapeworm, but some evidence exists suggesting that the tapeworm can counter this host defense (Ubelaker et al. 1970;

Lackie 1976; Richards and Arme 1985). Finally, the behavior of infected beetles may change such that their activity and metabolism increases.

Because there was no treatment effect on total proportional weight change in both starved and fed beetles, I conclude that host resources do not limit parasite growth in males or females. Finally, males have a higher metabolic rate regardless of their feeding or infection status. These results will have ramifications on the beetle’s feeding behavior and subsequent probability of infection (Table 1.4). Despite the observed increase in frass production in infected female beetles, food consumption did not change. If food consumption does not increase after infection, then the probability of re-infection does not change. When host resources do not limit parasite growth, as it appears in this experiment, then the parasite gains no advantage if the probability of re-infection remains the same. Finally, fed males, with a greater metabolism than females, also consumed more food. This male bias in feeding could explain the results showing a male bias in infection (also, Chapter 3 and Pappas et al. 1995). Also, males that feed more may gain a benefit that outweighs the increased risk of infection. Perhaps well fed males mate more often with females. Female choice and mate experiments involving fed and starved

22 males could help answer this question. Also, further experiments where starved and fed beetles are tested under uniform conditions are needed to explain the conditions under which the male infection bias will occur.

23 Assuming constant conditions and limited host resources If host feeding then the probability Parasite: activity of either sex: of re-infection: increases increases disadvantage

decreases or no change no impact stays same

Assuming constant conditions and unlimited host resources If host feeding then the probability Parasite: Activity of either sex: of re-infection: increases increases may have advantage decreases or no change no impact stays same

Table 1.1: Summary of predicted ramifications for transmission dynamics under conditions of both limiting and unlimiting host resources.

24 Starved Female Control Female Infected Male Control Male Infected N 36 32 32 31

Day 0 124.4 (2.71) 111.4 (3.44) 111.7 (3.9) 119.4 (4.07)

Day 1 132.2 (2.8) 118.6 (3.34) 125.8 (4.08) 131.7 (3.96)

Day 17 107.6 (2.22) 97.1 (3.05) 98.3 (2.94) 103.8 (2.94)

Proportional -0.18 (0.007) -0.18 (0.008) -0.21 (0.007) -0.21 (0.005) weight change Fed N 29 31 32 27

Day 0 103.5 (2.57) 106.2 (3.78) 103.2 (2.88) 104.1 (2.85)

Day 1 111.1 (2.48) 113.7 (3.63) 114 (3.09) 113.8 (3.27)

Day 17 75.8 (2.03) 77.5 (3.15) 73.9 (2.51) 72.3 (2.00)

Proportional -0.32 (0.007) -0.32 (0.009) -0.35 (0.009) -0.36 (0.006) weight change

Table 1.2: Average beetle weights (mg) and average total proportional weight change (Day 17 – Day 1 / Day 1) for beetles in the fed and starved experiments. Standard errors are in parentheses.

25 Starved Female Control Female Infected Male Control Male Infected N 36 32 32 31

Frass count 20.8 (1.05) 27.7 (1.66) 25.4 (1.92) 26.1 (1.64)

Fed N 29 31 32 27

Frass count 175.7 (10.8) 166.5 (11.5) 187 (14.7) 218.9 (16.2)

Pellet 3.91 (0.46) 3.29 (0.55) 4.87 (0.56) 5.19 (0.47) consumption

Table 1.3: Average frass count for the starved beetle experiment and average frass count and pellet consumption (mg) for the fed beetle experiment. Standard errors are given in parentheses.

26 Beetle Food Probability of Parasite Host host consumption re-infection

Males Did not Did not No impact No impact at change change low parasite intensities

Females Did not Did not No impact No impact at change change low parasite intensities

Table 1.4: Observed beetle response to parasitism when host resources are known not to limit parasite growth and development, and its ramifications on the parasite and host.

27 Starved Female Control Female Infected Male Control Male Infected N 36 32 32 31

Expected 113.0 (2.8) 99.4 (3.3) 106.6 (4.1) 112.5 (4.0) weight

Proportional -0.05 (0.01) -0.02 (0.01) -0.08 (0.01) -0.08 (0.01) weight change Fed N 29 31 32 27

Expected 65.0 (2.5) 68.0 (3.6) 67.8 (3.1) 67.7 (3.3) weight

Proportional 0.15 (0.01) 0.14 (0.02) 0.09 (0.02) 0.08 (0.02) weight change

Table 1.5: Average expected beetle weights (mg) and average total proportional weight change (Day 17 – Day 1 / Day 1) for beetles in the fed and starved experiments after adjusting for water loss. Standard errors are in parentheses. Male and female proportional weight change (after an arcsine square root transformation) were significantly different from each other in starved (F1,127 = 12.56, p = 0.001, GLM) and fed (F1, 115 = 9.02, p = 0.003) beetles.

Beetle infected with a cysticercoid

egg with high r egg with low r

Fig. 1.1: Diagram showing foraging possibilities for a beetle infected with a cysticercoid. If the beetle ingests an egg that shares a high coefficient of relatedness with the cysticercoid (egg on the left), then the parasite’s inclusive fitness increases. If the beetle ingests an egg that shares a low coefficient of relatedness with the cysticercoid (egg on the right), then the parasite’s inclusive fitness will not increase. Instead, if host resources limit parasite growth, then the parasite’s fitness may decrease. Note that diagram is not drawn to scale.

29 30

* 20

Weight gain (mg) on Day 1 for starved beetles -10 FC FI MC MI

Figure 1.2: Weight gain (mg) on day one of FC (Female, Control), FI (Female, Infected), MC (Male, Control), MI (Male, Infected) starved beetles. Grey (infected) and white (control) boxplots show median, 10th, 25th, 75th and 90th percentiles and outliers (data from Table 1.2; * = males gained significantly more weight, F = 32.4, p < 0.001, GLM).

30 -0.05

-0.10 * -0.15

-0.20

-0.25

Total proportional weight loss for starved beetles loss for starved weight proportional Total -0.30 FC FI MC MI

Figure 1.3: Proportional total weight loss in starved beetles. Conventions as in Figure 1.2 (data from Table 1.2; * = males lost significantly more weight (after an arcsine square root transformation), F = 19.29, p < 0.001, GLM).

31 60

** * 20 Total frass count of starved beetles Total frass count

0 FC FI MC MI

Figure 1.4: Frass production in starved beetles. Conventions as in Figure 1.2 (data from Table 1.3; * = infected beetles produced more frass than controls, F = 5.8, p = 0.017, GLM; specifically, infected females produced significantly more frass than control females, p = 0.0026 with Bonferroni post hoc comparison. **Also, male controls produced more frass than female controls, p = 0.039 with Bonferroni post hoc comparison).

32 10

8 Infected (N = 32) Control (N = 36) * 6

4 Average Frass Count Average

0 1-3 4 5-6 7 8-11 12 13-16 17 Days

Fig. 1.5: Average frass production in starved females with standard errors and sample sizes for controls (shaded bars) and infected (empty bars). Asterisks indicate a significant difference between control and treatment (repeated measures ANOVA testing the effects of sex, infection status, day and their interactions on frass production. The overall effects of sex, infection status and day were significant at p < 0.001. On days 13-16, infected females produced more frass than control females, d.f. = 141, t-value = -3.29, adjusted p < 0.003 for Bonferroni post hoc comparisons).

33 12

10 Females Males 8

4 Number of cysticercoids of Number

0 1-20 21-40 41-60 61-80 81-100 101-120 121-140 141-160 161-180 181-200 201-220 221-240 Frequency

Fig. 1.6: Frequency distribution of the number of cysticercoids dissected from starved female (empty bars) and male (shaded bars) beetles at the completion of the experiment (Fligner-Policello, n♂ = 31, n♀ = 32, U = 2.39, p = 0.0085).

34 30

15 * 10

0 Weight gain (mg) on Day 1 for fed beetles

-5 FC FI MC MI

Figure 1.7: Weight gain (mg) in fed beetles on day one. Conventions as in Figure 1.2 (data from Table 1.2; * = males gained significantly more weight, F = 7.21, p < 0.008, GLM).

35 -0.2

* -0.3

-0.4 Total proportional weight loss of fed beetles loss weight proportional Total -0.5 FC FI MC MI

Figure 1.8: Proportional total weight loss in fed beetles. Conventions as in Figure 1.2 (data from Table 1.2; * = males lost significantly more weight (after an arcsine square root transformation), F = 21.61, p < 0.001, GLM).

36 400

300

200 *

100 Total frass count of fed beetles

0 FC FI MC MI

Figure 1.9: Frass production in fed beetles. Conventions as in Figure 1.2 (data from Table 1.3; * = males produced more frass than females, F = 5.58, p = 0.020, GLM).

37 14

8 * 6

Total pellet consumption in fed beetles fed Total pellet consumption in -2

-4 FC FI MC MI

Figure 1.10: Pellet consumption (mg) in fed beetles. Conventions as in Figure 1.2 (data from Table 1.3; * = males consumed more than females, F = 7.55, p = 0.007, GLM).

38 90

90% RH: y = -0.05x + 81 R2 = 0.03 80

40% RH: y = -0.12x + 70.8 R2 = 0.27

60 Water weight (mg) weight Water

40%RH 90%RH 50

072144 Time (hours)

Figure 1.11: Water weight (mg) loss over time (hrs) in beetles kept at 40% (filled circles) and 90% (empty circles) relative humidities (n = 10 in all cases).

CHAPTER 2

PARASITE-ALTERED PREFERENCE FOR INFECTIVE FECES IN UNINFECTED INDIVIDUAL HOSTS

ABSTRACT

Parasite life cycles often require multiple hosts and parasite fitness depends on the

consistent transmission to the next host. Many parasites can increase the probability of

being transmitted via a variety of mechanisms. The beetle, Tenebrio molitor, risk infection with cysticercoids when ingesting rat feces containing eggs of the tapeworm,

Hymenolepis diminuta. Previous studies demonstrated that groups of beetles prefer infective feces over uninfective feces due to the presence of a volatile attractant. Since aggregation pheromones may confound beetle behavior, this study tests the fecal preference of individual beetles alone and in the presence of another beetle, while separating olfactory from gustatory cues. When allowed to feed, females prefer infective feces, while males show lower activity levels. These results may explain the findings from other studies observing a higher load of cysticercoids in males. Further, the fecal attraction is for the infective feces and not the egg and suggests an active role for the adult tapeworm.

40 INTRODUCTION

Characteristics of cestode eggs of aquatic hosts appear to increase the probability of their ingestion by the correct intermediate host (Jarecka 1961). For example, heavy

Hymenolepis megalops eggs sink to the bottom of lakes where they are likely to be

consumed by their intermediate host, the benthic ostracode, Cypris pubera. The filiform

eggs of Diorchis spp. become entangled in an aquatic plant where they are more likely to

be ingested by grazing ostracods. The small, thin shelled eggs of Hymenolepis furcifera

and Anomotaenia ciliata can be eaten by their intermediate host, a lithoral species of

Cladocera (Jarecka 1961). The resemblance of Fimbriaria fasciolaris eggs to

filamentous algae increases the likelihood that they will be eaten by benthic amphipods

(Podesta and Holmes 1970). Others have found two morphologically different eggs in

the same tapeworm (Berntzen and Voge 1962) as well as two types of eggs differing in

size (Pappas and Leiby 1986) suggesting an adaptation to two or more different

intermediate hosts (Mackiewicz 1988). However, no one has provided experimental

evidence consistent with the hypothesis that these examples are, in fact, adaptations that

increase parasite transmission to the intermediate host.

Other studies suggest that hosts avoid the consumption of parasites. For example,

reindeer avoid pastures with high densities of dung suggesting a behavioral adaptation

that reduces their risk of infection with Trichostrongyle eggs (Van der Wal et al. 2000).

Because nematode larvae develop in moist areas where dung density is low, the authors

concluded that dung density would be a poor indicator of infection risk. Also, grazing

sheep avoid grazing on patches contaminated with feces (Hutchings et al. 1998).

Because nematode-infected sheep increase their fecal avoidance behavior, the authors

41 suggest that fecal avoidance is a host adaptation that decreases the probability of further

nematode infection. Finally, oystercatchers feed on intermediate-sized cockles instead of large-sized cockles, which would maximize their energy intake. Because cockle size is positively correlated with helminth parasite load, one model suggests that oystercatchers

balance their energy intake against the risk of parasite infection by foraging selectively

on intermediate-sized cockles (Norris 1999).

Host-altered behavior does not always occur in the egg-to-intermediate host stage.

Isopods can serve as the intermediate host for an acanthocephalan parasite,

Plagiorhynchus cylindraeus, when they consume feces from starlings containing the

parasite’s eggs. When presented with starling feces and starling feces hydrated with an

aqueous solution of P. cylindraeus eggs, isopods consumed equivalent amounts of

control and “infected” feces (Moore 1983).

However, in another system, experiments suggest that host behavioral alteration may benefit the parasite. In the tapeworm, Hymenolepis diminuta, adults live in the small intestines of rodents, the definitive host. Eggs pass out of the rodents in their feces. The cysticercoid, or larval stage, occurs in an insect (the intermediate host, most often a

beetle) that becomes infected by feeding upon infective feces. The rodent completes the

cycle when it ingests an infected beetle. Thus, the success of this parasite depends on

rodent-to-beetle and beetle-to-rodent transmission, and the former depends upon

consumption of the egg stage by the beetle. Recent evidence suggests that even this

passive type of transmission involves a mechanism that influences host behavior to

benefit the tapeworm.

42 Three studies tested the fecal preference of groups of beetles by allowing them to forage in an arena with two types of food: rat feces from uninfected rats and rat feces from rats infected with H. diminuta (presumed to contain eggs). Beetles (Tribolium confusum) starved 48 hours prior to the trial preferred infective feces (Evans et al. 1992).

When beetles (Tenebrio molitor) of known sex, age, and feeding history are tested, female beetles (those fed and those starved 72 hours prior to the trial) preferred infective feces. Starved male beetles preferred infective feces while fed males preferred uninfective feces (Pappas et al. 1995). In the third study, an analysis of beetle (T. confusum) movement showed that starved beetles (of mixed sex and age) made fewer visits to infective bait, but the duration of visits to each bait did not differ when tested individually. When tested in groups during a two hour trial, starved and fed beetles preferred control bait during two time intervals, but showed no preference during the other four time intervals. The authors concluded that beetle behavior appeared “highly heterogeneous, both among individuals and by the same individual over time” (Shostak and Smyth 1998). These three experiments demonstrated the importance of sex and feeding history in predicting a beetle’s preference for infective or uninfective feces, but did not address any mechanism for the altered behavior.

The mechanism may be the presence of a volatile attractant in infective feces, which was collected on a solid adsorbent by aspirating fresh rat feces from infected and uninfected rats (Evans et al. 1998). The volatiles were eluted with diethyl ether, and preference tests showed that more beetles were attracted to volatiles from infective feces.

The volatile attractant has not been identified and may originate either from the tapeworm eggs or the infective feces. Secretions from the adult tapeworm may somehow

43 alter feces before defecation and create a beetle attractant that could result in more

infected beetles. If infection harms the beetle, then there should be strong selection for

host-altered behavior that decreases the probability of parasite transmission.

Infection by H. diminuta cysticercoids harms the beetle host in a variety of ways.

Reduction in host fecundity occurs in both T. confusum (Keymer 1980, 1981) and

Tenebrio molitor (Hurd and Arme 1986). Infected male T. molitor exhibit a lower response to pheromone produced by uninfected females (Hurd and Parry 1991).

Defensive glands of infected T. molitor contain fewer defensive compounds than uninfected controls (Blankespoor et al. 1997). Fat body glycogen reserves are significantly depleted 3 days post-infection in male and 5 days post-infection in female T. molitor (Kearns et al. 1994). Finally, infected males are less attractive to females, and females mated with infected males produce fewer offspring than females mated with uninfected males (Worden et al. 2000). Despite the harmful effects of infection, previous studies did not observe beetles consistently avoiding infective feces, which may be due to several reasons.

Olfactory cues are necessary to detect volatiles, but previous experiments (except

Evans et al. 1998) allowed the beetles to feed, providing additional gustatory cues. Also, with the exception of Shostak and Smyth (1998), previous studies tested groups of

beetles, and results may be confounded by influences due to sex and aggregation

pheromones found in these grain beetles (Burkholder and Ma 1985, August 1971,

Tschinkel et al. 1967). Additionally, the volatile attractant may only be released by the

mechanical breaking of the tapeworm egg during the beetle’s feeding. Finally, results

44 from Chapter 1 indicated that males and females, with the same nutritional status, react differently in response to parasitism.

To clarify the various factors influencing beetle preference, individual male and female beetles were tested for fecal preference under various conditions. Olfactory and gustatory cues were separated by enclosing feces in wire cages to test the hypothesis that beetles require both cues to distinguish between baits. Experiments involving preference for tapeworm eggs only were performed to test the hypothesis that the volatile attractant is present in tapeworm eggs. Beetle preference in the presence of other beetles was examined to test the hypothesis that social conditions influence beetle preference.

Response differences between the sexes are analyzed to test that hypothesis that one sex would be more susceptible to infection than the other.

METHODS

Study subjects

The “OSU Strain” (Pappas and Leiby 1986) of H. diminuta was maintained in male Sprague-Dawley rats and beetles (T. molitor). Because this is a highly inbred strain of tapeworm (having survived a severe bottleneck in 1995), I assume it has low genetic variability, which should minimize concern about parasite genetic variation confounding the results of host behavior (but see Meffert 1999). Three male Sprague-Dawley rats were infected with 30 cysticercoids and maintained on commercial rodent chow and water. Three additional male rats, obtained from the same commercial source, of identical age and from the same litter so as to minimize rat genetic variability, were maintained under the same conditions as the infected rats to serve as the source of control

45 (uninfective) feces. Fecal pellets from infected rats were not used prior to 20 days post- infection. Rat cages were checked every 10 minutes on the morning of the trial, and fecal pellets were collected with forceps to minimize contamination by rat urine. After each trial, the infective fecal pellets were examined to verify the presence of H. diminuta eggs.

Beetles were maintained on wheat bran, and small pieces of potato were added to the cultures on a regular basis. Pupae were removed from the cultures, and male and female pupae (Bhattacharya et al. 1970) were placed in separate dishes containing wheat bran. Beetles that emerged during a 24 h period were collected such that a daily cohort of beetles was maintained for each sex.

Experimental protocol

Each preference experiment consisted of 60 trials (except where noted) that tested

30 individual male and female beetles in a small plastic arena divided down the middle

(Fig 2.1). Fresh control and infective fecal pellets were matched according to time of defecation (within one hour of each other) and used immediately as a paired “batch.”

Each pellet was placed in a separate plastic screw top vial, labeled, and covered so that the preference trial was conducted blind. Bait placement, at opposite corners of the arena, was assigned randomly at the start of each trial, and the arena was rotated 90° after each trial. To create a homogenous distribution of eggs, the fecal pellets were mixed with a stirrer for one minute. From these paired batches, feces, matched by weight (20-

35 mg) to within 2 mg, were placed on cover slips and one paired batch of feces was used for no more than four trials. The positions of both baits were alternated in successive trials. Beetles of known sex, age (6-19 days) and feeding history (starved 2-7 days) were tested in a 15 min trial (after a 7 min acclimatization period). I recorded the side to

46 which the beetle first moved, the time spent on each side, the time spent on each type of bait, the number of contacts with each type of bait, and the number of times the midline of the arena was crossed. Trials were conducted under various experimental conditions

(Table 2.1).

In experiment 1, designed to mimic natural conditions of infection via fecal consumption, equal amounts of fresh feces were placed on cover slips to allow the beetles to feed during the trial. Hence, both gustatory and olfactory cues were present. In experiment 2, equal amounts of fresh feces were enclosed in a wire cage that prevented the beetle from feeding. Hence, only olfactory cues were available to the beetle.

To confirm that the amount of feces used in these trials did not overwhelm the beetles' senses and confound their navigation, experiment 1 was repeated using smaller amounts of fresh infective and control feces (10-15 mg each and differing by +/- 1 mg) in an uncovered arena.

In experiments 3-5, beetle preference was tested for infective and control feces, collected 0, 1 or 6 hours after defecation. They were stored frozen at -40°C until the time of the trial. After thawing, equal amounts of infective and control feces were enclosed in wire cages. These experiments simulated natural conditions in which beetles would encounter infective rat feces varying in time after defecation.

Experiments 6 and 7 tested beetle preference for fecal volatiles in the absence of feces. In experiment 6, 10-15 fecal pellets (either infective or control) were soaked in 10-

20 ml of methanol overnight. The solution was decanted, centrifuged at 7,000 RPM for

30 minutes and refrigerated until the trial. Before the trial, a drop of the extract was placed on a glass cover slip and allowed to dry to test beetle preference for extracts from

47 infective or control feces. In experiment 7, volatiles were collected on filter paper that was enclosed for at least three hours in a jar containing 15-20 fecal pellets (either infective or control).

Experiments 8-10 tested the possibility that a volatile attractant was present only in tapeworm eggs by allowing beetles to feed on tapeworm eggs in the absence of infective feces. Tapeworm eggs were recovered from infective feces using a gradient centrifugation technique (Lethbridge 1971b). Experiment 8 tested preference for apple scrapings mixed with tapeworm eggs versus apple scrapings mixed with de-ionized water. Since crushing the eggs might release a volatile attractant, experiment 9 tested apple scrapings mixed with crushed (using a pestle and mortar) tapeworm eggs versus apple scrapings mixed with uncrushed eggs. The presence of crushed eggs was verified microscopically. To represent a more natural condition in which tapeworm eggs would be present in rat feces, experiment 10 tested control feces (that had been frozen) mixed with tapeworm eggs versus control feces mixed with de-ionized water.

Experiments 11-13 tested beetle preference in the presence of other beetles using equal amounts of refrigerated feces as bait. In experiment 11, one beetle was placed in a wire cage containing infective feces while the other cage contained only infective feces.

In experiment 12, one beetle was placed in a cage containing control feces while the other cage contained only control feces. In experiment 13, one beetle was placed in the cage containing control feces and one was placed in the cage containing infective feces. In these experiments, beetles were matched within each trial according to age, sex, size and feeding history.

48 Data analysis

Many feeding preference trials use the amount of food consumed as an indicator

of preference. Many researchers have questioned the assumption of independence

between food types (the consumption of one type of food is dependent on the presence of

others if a preference exists), the proper use of controls, and the correct choice of

statistical test (e.g., Peterson and Renaud 1989; Roa 1992; Manly 1993; Horton 1995).

However, if time spent on food types is the measure of preference, then another issue

must be considered. The proportion of time spent on the side containing the infective

feces was tested with a Wilcoxon Signed Rank test for equality to 50% for all male and

female beetles. This test assumes a binomial distribution where the presence of the beetle

in the next moment (on the infective or uninfective side) is independent of its current

location. However, this assumption is not met because the beetle’s location in the next moment of time depends on its current location. This autocorrelation introduces greater than binomial variance, which can only be accounted for by testing individual beetles in repeated choice experiments. Yet, this was not possible because such tests may result in an infection which could alter subsequent choices. Figures 2.2 and 2.3 shows how the observed probability distributions deviate from an expected binomial distribution in which the beetle would have k hypothetical opportunities to make choices in the 15- minute trial. Thus, large deviations (especially for males) make it difficult to demonstrate small preferences. For example, k = 2 represents a hypothetical situation in which each beetle is making two discrete choices and will result in an expected standard deviation of 0.36.

49 The Fligner-Policello procedure tested for a median difference in the proportion

of time spent on infective and control baits between male and female beetles that

experienced both sides (Hollander & Wolfe 1999). As paired data, the difference between the proportion of time spent on infective and control baits was tested for equality to zero with a Wilcoxon Signed Rank test. Confidence intervals (95% except where indicated) were constructed about each median using a nonlinear interpolation procedure

(Minitab v. 13.3, State College, PA). The Benjamini-Hochberg method (Benjamini and

Hochberg 1994) was used to account for the multiple experiments (13) analyzed.

Significant tests (p < 0.05) remained significant under the following condition: pi <

(iq)/m, where p1 < p2 < … < pm are the ordered p values, q is the assigned False

Discovery Rate (0.05), and m is the number of experiments (13).

Because many beetles spent no time on either infective or control bait, the data

were re-analyzed with the same tests using the data for beetles that contacted both baits,

and significance was adjusted to 0.025 according to the post hoc Bonferroni adjustment.

A two-tailed Fisher exact test (GraphPad, San Diego, CA) was used to evaluate the

association between sex and beetles that contacted both baits and beetles that contacted only one bait. For beetles that contacted only one bait, the same test analyzed the relationship between sex and beetles that contacted infective bait and beetles that contacted control bait. Finally, the number of bait contacts was analyzed for all beetles using the Scheirer-Ray-Hare extension of the Kruskal-Wallis test (with a correction for ties) for ranked data in a 2-way design with treatment and sex as factors (Sokal and Rohlf

1995). Beetles that contacted both baits were analyzed with the same test. Again, significance was adjusted to 0.025 for the question of treatment and to 0.0167 for the

50 question of sex effects (since this question is asked three times with the Fisher Exact test)

according to the post hoc Bonferroni adjustment.

RESULTS

When beetles are allowed to feed on fresh feces (experiment 1), females spent a

significantly greater proportion of time (median difference = 0.22) on the infective bait

than on the control bait (Table 2.2; Wilcoxon Signed Rank, W = 366, p = 0.001). Among

beetles that contacted both baits, females still spent a significantly greater proportion of

time (median difference = 0.18) on the infective bait (W = 309, p = 0.004). Analysis of

beetle contact with baits showed that females contacted both baits more frequently than

males (2-tailed Fisher exact test, p = 0.001). Of the 20 beetles that contacted only one

bait, there was no association between sex and contact with infective or control bait (2-

tailed Fisher exact test, p = 1.0). When all beetles were considered, females contacted

2 either bait more times than males (Table 2.3; Scheirer-Ray-Hare, χ 1 d.f., p = 0.0017).

However, this difference was no longer significant when beetles that contacted both baits were analyzed (Table 2.3; p = 0.629).

In contrast, when beetles were prevented from feeding on fresh feces (experiment

2), no significant time differences between treatments were found (Table 2.2). Male

beetles spent a significantly higher (median difference = 0.22) proportion of time on the

infective bait than females (Fligner-Policello, p = 0.0095), but when beetles that contacted both baits were considered, the median difference between the sexes was no longer significant. Unlike experiment 1, the number of beetle contacts with both baits, or

only one bait did not differ between males and females (2-tailed Fisher exact test, p =

51 0.39). Of the 39 beetles that contacted only one bait, there was no association between

sex and contact with infective or control bait (2-tailed Fisher exact test, p = 0.34). There

was no treatment (trt) or sex difference for number of contacts with either baits when all

beetles were considered (Table 2.3; ptrt = 0.667, psex = 0.142), or when beetles that

contacted both baits were considered (ptrt = 0.362, psex = 0.282).

When the experiment was repeated such that beetles were allowed to feed on

small amounts of control or infective feces in an uncovered arena, no median differences

in treatment or sex existed (Table 2.4). Contact with both baits or only one bait did not

differ between males and females (2-tailed Fisher exact test, p = 0.695). Likewise, of the

beetles that contacted only one bait, there was no association between sex and contact

with infective or control bait (p = 1.00). Finally, there was no treatment or sex difference

for number of contacts with either baits when all beetles were considered (Table 2.3; ptrt

= 0.863, psex = 0.409) and when beetles that contacted both baits were considered (ptrt =

0.174, psex = 0.18).

In experiments 3 through 10, beetles that visited both sides and beetles that

contacted both baits did not differ in the median time spent on infective and control baits

(Tables 2.5-2.9). The frequency of contact with both baits, or only one bait did not differ

between males and females (2-tailed Fisher exact test, all p’s > 0.16), except in

experiment 6 where males did not significantly contact one bait more frequently than

females (p = 0.0819). Likewise, among beetles that contacted only one bait, there was no

association between sex and contact with infective or control bait (all p’s > 0.36).

When all beetles were considered, there was no treatment or sex difference for

number of contacts with either baits except for the following. In experiment 10, females

52 contacted the baits more often than males (Table 2.11; p = 0.0025), but this difference

was no longer significant when beetles that contacted both baits were considered (p =

0.125). When beetles that contacted both baits were considered, there was no treatment or sex difference for number of contacts with either bait except for the following. In experiment 6, females contacted the baits more than did males (Table 2.10; p = 0.011).

In experiment 11, male beetles spent a greater proportion of time (median difference = 0.45) on the cage containing infective bait with another beetle relative to the cage containing infective bait without another beetle (Table 2.12; W = 237, p = 0.003).

However, when only beetles that contacted both baits were considered, the median

difference was no longer significant. There were no median differences between sexes.

Beetles of both sexes contacted the cage containing infective bait with another beetle

more times than the cage containing only infective bait (Table 2.13; p = 0.0066) when all

beetles are considered. When beetles that contacted both baits were considered, females

contacted the baits more than did males (p = 0.012).

Males and females did not express a preference for infective bait and another

beetle enclosed in a cage, or for control bait and another beetle enclosed in a cage (Table

2.14; experiment 12). Likewise, beetles showed no treatment or sex differences when in

an arena with control bait and another beetle enclosed in a cage and control bait enclosed

in a cage without another beetle (Table 2.14; experiment 13). In experiments 12 and 13,

there was no treatment or sex difference for number of contacts with either baits (Table

2.13) when all beetles were considered and when beetles that contacted both baits were

considered. In experiments 11-13, contact with both baits or only one bait did not differ between males and females (2-tailed Fisher exact test, p > 0.05). Likewise, of the beetles

53 that contacted only one bait, there was no association between sex and contact with

infective or control bait (p > 0.05).

DISCUSSION

When allowed to feed, female beetles spend more time on the infective bait than

on the control bait, suggesting a preference for infective feces (Table 2.2). Such a

preference increases the likelihood of ingesting and becoming infected with H. diminuta eggs. If egg to intermediate host transmission is successful, parasite fitness increases.

These results are consistent with previous studies suggesting that feces containing eggs of

H. diminuta are capable of directing the foraging activity of groups of beetles so as to

increase the probability of their transmission (Evans et al. 1992, Pappas et al. 1995).

However, these results refute the findings that starved individual beetles (T. confusum)

made fewer visits to infective baits than to uninfective baits, spent more time than

expected at the control baits and spent less time than expected at the infective baits

(Shostak and Smyth 1998). Calculating an expected time makes assumptions that may not be valid (see Methods) so that measures of preference can significantly alter the results. Additionally, the two beetle species react differently to moist foods. T. molitor seem to prefer moist food (Pappas et al. 1995), whereas members of Tribolium generally prefer dry food (Sokoloff 1974).

Males, on the other hand, do not show any preference for infective feces, suggesting that males are less likely to be infected than females. However, no difference in infection intensities has been found between male and female Tenebrio infected in the lab (Hurd and Arme 1987), or in nature (Rau 1979). Further, results from Chapter 1

54 demonstrate that males, in one experiment, have a higher intensity of cysticercoids than

do females. Differences in activity level may explain this paradox.

Females contact both baits more frequently than males in experiment 1 (Table

2.3) and in experiment 6 (Table 2.10; p = 0.0819). Also, when all beetles are considered, females contact the baits more times than males in experiments 1 and 10 (Tables 2.3 and

2.11). Finally, when only beetles that contact both baits are considered, females contact

the baits more times than males in experiments 6 and 11 (Tables 2.10 and 2.13).

Although the same significant difference is not found in other experiments, its presence in four very disparate experiments suggests that females, in general, have a higher activity level than males. If male activity is lower and individuals feed longer on infective and uninfective feces than females, then cysticercoid intensity should be higher in males than in females. Thus, parasite transmission depends not only on the foraging behavior of T. molitor in the presence of H. diminuta eggs, but also on the beetle’s sex.

Experiment 2 tests the preference of beetles for infective and control feces enclosed in wire cages. Both male and female beetles fail to exhibit a significant preference. Further, contact with both baits or only one bait does not differ between males and females. These data suggest that beetles, relying only upon olfactory cues, cannot distinguish between infective and control feces. Alternatively, beetles may distinguish between baits, but the experimental design cannot measure this because beetles do not stay at the baits when prevented from feeding. Compared with experiment

1 in which beetles were allowed to feed, the results stress the importance of gustatory cues in distinguishing infective and control feces. Infective feces may contain no volatile attractants, or an insufficient amount for the beetles to distinguish between the types of

55 bait based on olfactory cues alone. Alternatively, too much of the volatile attractant may

have hindered the beetle's navigation in the arena.

When experiment 1 is repeated using a smaller amount of feces (see Methods), no

median treatment or sex differences are found (Table 2.4), suggesting that the amount of

feces used is too low for the beetles to distinguish between infective and control feces.

There is no sex or treatment difference in number of bait contacts (Table 2.3), suggesting

that the female's higher activity level occurs only in the presence of detectable levels of

infective feces.

Males from experiment 2 spend more time on the infected cage than do females

(Table 2.2), which underscores the importance of sex when considering parasite

transmission. It is also consistent with bait contact results showing that females have

higher activity levels.

Experiments 3, 4 and 5 simulate conditions in which a beetle would encounter rat feces various times after defecation. When contrasted with the sex difference found in experiment 2, the lack of significant sex and treatment differences in these experiments suggest that, in the wild, the ability of infective feces to direct the foraging activity of beetles would be time limited.

The mechanism responsible for the directed foraging behavior of beetles may be the presence of a volatile attractant found in infective feces (Evans et al. 1998).

However, to collect the attractant, it was necessary to use 30-115 fresh rat fecal pellets.

The probability of encountering such a large concentration of fresh infective fecal pellets in the wild is unknown. In the wild, some defecation spots contain up to 100 fecal pellets from multiple rats deposited at different times. Temperature and food availability also

56 contribute to the wide feces per rat variability (3.8 – 72.7) observed (Calhoun 1963).

Also, in the Evans et al. (1998) experiment, beetle preference for the volatile attractant

was based solely on olfactory cues. Again, beetles are likely to rely upon both olfactory

and gustatory cues in the wild. Further, beetles in the wild are likely to encounter both

fresh and old feces. Finally, the experiment did not distinguish if the volatile attractant

was present in the tapeworm eggs or in the infective feces.

Experiments 6 and 7 test beetle preference for volatiles from infective feces.

Failure to observe any treatment differences (Table 2.7) contrasts with the results from

Evans et al. (1998). One possible explanation is the larger number of fecal pellets (30-

115) that were required to elicit a response in the Evans et al. (1998) experiment as

opposed to the number used in this experiment (15-20 in experiment 6 and 10-20 in

experiment 7). There may simply not have been enough volatiles present in the amount

of feces used.

Experiments 8, 9, and 10 test preference for tapeworm eggs in the absence of

infective feces while allowing the beetles to feed. Again, there is no treatment difference

(Tables 2.8-2.9). When compared to experiment 1, this result suggests that beetles do not

distinguish between presence and absence of tapeworm eggs. Instead, beetles distinguish

between infective and control feces. This may indicate an influence from the adult

tapeworm before defecation. Excretions from the adult tapeworm in the gastrointestinal

tract may contribute to the attractiveness of the infective feces. Alternatively, the volatile

attractant may have been removed from the eggs during the process that separates them

from the infective feces.

57 Previous studies, testing groups of beetles, demonstrated preference for infective

feces (Evans et al. 1992, Pappas et al. 1995, Shostak and Smyth 1998). Beetle preference

may be influenced by sex and aggregation pheromones that are found in grain beetles

(Burkholder and Ma 1985, August 1971, Tschinkel et al. 1967). To examine the effect of

beetle social interaction, experiments 11, 12 and 13 test beetle preference in the presence

of conspecifics. In experiment 11, males spend more time on the cage containing

infective feces and another beetle, than on the cage containing infective feces alone

(Table 2.12), suggesting that male beetles respond to the presence of conspecifics. Both

males and females contact the cage containing infective feces with a beetle more often than the cage containing only infective feces (Table 2.13), suggesting a higher activity level at the cage containing a conspecific. Compared with experiment 11, the lack of significant differences in experiment 13 suggest that the male response to a conspecific may only occur in the presence of infective feces. However, male response to a conspecific is also not observed in experiment 12, which used both infective and control feces. A possible explanation is that the concurrence of two types of feces and two beetles in the same small arena generates too many olfactory cues, which confound the beetle's ability to navigate.

Due to constraints in the experimental design, it was not possible to assume that the beetle’s movements were independent from one moment to the next. In fact, distributions for the proportion of time spent on the infective side deviated from expected binomial distributions in which the beetle would have the hypothetical opportunity to make 15 or 30 choices during the trial (Figs. 2.2 and 2.3). Instead, the observed standard deviation range (0.32 – 0.38) is consistent with a hypothetical situation in which the

58 beetles made a small number of discrete binary choices, or when k = 2. This made

detecting small preferences difficult in most cases, especially in male beetles.

In conclusion, only female beetles held individually show a preference for

infective feces, and they do so only when the feces are fresh and the beetles are permitted

to feed. Although these results suggest that females are more susceptible to infection, this is not observed in the field (Rau 1979). An explanatory factor is the lower activity of

male beetles, which tend to remain near one bait. In nature, rats tend to defecate near

their food and colonies (Emlen et al. 1948; Davis et al. 1948), in areas where their

locomotion is stalled, or in sheltered areas (Calhoun 1963). If males exhibit low activity

levels in nature, then some males should have very heavy infection levels while others should have low infection levels. Females, on the other hand, should have a more homogenous level of infection. Some support for this is found in Chapter 1 showing that starved males have a higher median load of cysticercoids than females.

Also, when presented with the amount of feces used in this experiment, females require gustatory cues to differentiate between infective and control feces. Unlike T. confusum, T. molitor is unable to detect the volatile attractant of Evans et al. (1992) using olfactory cues alone. Further, experiments 1 and 10 suggest this attractant is present in infective feces and not in the tapeworm eggs (assuming that the egg separation process did not remove the attractant from the eggs).

If infection disadvantages the host, then beetles of both sexes should avoid infective feces. However, when allowed to feed, females prefer infective feces while males show no preference. This suggests that females experience less selective pressure to avoid or show no preference for infective feces than males. This may result from

59 differential costs to infection. For example, infected male T. molitor exhibit a lowered response to pheromone produced by uninfected females (Hurd and Parry 1991), while fat body glycogen reserves are significantly depleted 3 days post-infection in male and 5 days post-infection in female T. molitor (Kearns et al. 1994). Finally, infected males are less attractive to females, and females mated with infected males produce fewer offspring than females mated with uninfected males (Worden et al. 2000). Thus, avoiding infective feces may be more strongly selected in males.

60 Experiment Control Bait Infective Bait Allowed to feed? 1. fresh uninfective feces fresh infective feces Yes 1a. 50% less fresh uninfective feces 50% less infective feces Yes 2. fresh uninfective feces fresh infective feces No

3. thawed uninfective feces collected thawed infective feces collected No immediately after defecation immediately after defecation 4. thawed uninfective feces collected thawed infective feces collected No 1 hour after defecation 1 hour after defecation 5. thawed uninfective feces collected thawed infective feces collected No 6 hours after defecation 6 hours after defecation

6. methanol extract of uninfective feces methanol extract of infective feces NA

61 7. filter paper exposed to uninfective feces filter paper exposed to infective feces NA

8. apple scrapings with de-ionized water apple scrapings with tapeworm eggs Yes 9. apple scrapings with de-ionized water apple scrapings with crushed eggs Yes 10. thawed uninfective feces thawed uninfective feces with eggs Yes

11. refrigerated infective feces refrigerated infective feces with beetle No 12. refrigerated uninfective feces with beetle refrigerated infective feces with beetle No 13. refrigerated uninfective feces refrigerated uninfective feces with beetle No

Table 2.1: Description of the 13 experiments performed.

1 Median proportion of time spent by Experiment 1 Females Males All beetles Infected side 0.55 (30, 0.37-0.75) 0.65 (37, 0.33-0.89) Beetles that visited both sides Infective bait 0.60 (29, 0.34-0.66)a 0.35 (33, 0.17-0.70) Control bait 0.20 (29, 0.09-0.36) a 0.36 (33, 0.03-0.50) Both baits contacted Infected side 0.52 (27, 0.36-0.74) 0.57 (19, 0.24-0.74) Infective bait 0.50 (27, 0.31-0.64)a 0.35 (19, 0.22-0.64) Control bait 0.24 (27, 0.11-0.37)a 0.44 (19, 0.30-0.58) Only one bait contacted Infective bait 0.79 (2) 0.76 (12, 0.24-0.89) Control bait 0.80 (1) 0.81 (5, 0.8-0.97)1 193.8% C. I. due to sample size

Experiment 2 Females Males All beetles Infected side 0.36 (30, 0.11-0.64) 0.71 (30, 0.37-0.83) Beetles that visited both sides Infected cage 0 (29, 0-0.21)b 0.63 (28, 0.04-0.76)b Control cage 0.46 (29, 0-0.74) 0.44 (28, 0-0.73) Both cages contacted Infected side 0.60 (6, 0.18-0.84) 0.66 (10, 0.40-0.82) Infected cage 0.45 (6, 0.09-0.85) 0.66 (10, 0.42-0.86) Control cage 0.33 (6, 0.07-0.75) 0.68 (10, 0.34-0.85) Only one cage contacted Infected cage 0.67 (8, 0.31-0.88) 0.77 (11, 0.54-0.90) Control cage 0.79 (12, 0.62-0.91) 0.81 (8, 0.60-0.89)

Table 2.2: In experiment 1, beetles were allowed to feed on fresh infective or control feces; one male beetle did not contact either bait. In experiment 2, beetles were prevented from feeding on fresh infective or control feces enclosed in cages; one male beetle and four female beetles did not contact either cage. Sample sizes and 95% median confidence intervals (except where noted) are in parenthesis. Significant median differences between treatments are indicated by “a” and median differences between sex are indicated by “b”.

62 Median number of bait contacts by Experiment 1 Females Males Infective bait All beetles 5.0 (30, 3.0-6.8)b 2.0 (37, 2-3.9)b Contacted both baits 5.0 (27, 3.0-7.0) 4.0 (19, 2.0-6.4) Control bait All beetles 5.0 (30, 1.2-7.8)b 2.0 (37, 0.1-4.0)b Contacted both baits 6.0 (27, 2.0-8.0) 4.0 (19, 2.9-7.1)

Experiment 2 Infected cage All beetles 0 (30, 0-1.) 1.0 (30, 1.0-2.0) Contacted both cages 1.5 (6, 1-3.6) 2.0 (10, 1-4.3) Control cage All beetles 1 (30, 0-1.0) 1.0 (30, 0-2.0) Contacted both cages 1 (6, 1-3.3) 1.5 (10, 1-3.7)

Experiment 1a Infective bait All beetles 2.0 (15, 0-3.0) 0 (15, 0-1.0) Contacted both baits 3.0 (4, 2.0-3.0)1 2 (3, 1.0-5.0)2 Control bait All beetles 1.0 (15, 0-1.6) 1.0 (15, 0-1.0) Contacted both baits 2.5 (4, 1.0-5.0)1 1.0 (3, 1.0-1.0)2 1 at 88% C.I. due to sample size 2 at 75% C.I. due to sample size

Table 2.3: Median number of bait contacts for all beetles and for beetles that contacted both baits/cages in experiments 1, 2 and when experiment 1 is repeated (1a). See text for descriptions of experiments. Sample sizes and 95% median confidence intervals (except where noted) are in parenthesis. Significant median differences, as determined by the Scheirer-Ray-Hare procedure for ranked data in a 2-way ANOVA design, between treatments are indicated by “a” and median differences between sex are indicated by “b”.

63 Median proportion of time spent by Experiment 1a Females Males All beetles Infected side 0.44 (15, 0.24-0.80) 0.41 (15, 0.09-0.64) Beetles that visited both sides Infective bait 0.23 (15, 0-0.35) 0 (15, 0-0.06) Control bait 0.03 (15, 0-0.39) 0.15 (15, 0-0.31) Both baits visited Infected side 0.36 (4, 0.19-0.44)1 0.45 (3, 0.09-0.68)2 Infective bait 0.26 (4, 0.23-0.36)1 0.02 (3, 0.01-0.08)2 Control bait 0.40 (4, 0.28-0.56)1 0.16 (3, 0.04-0.33)2 Only one bait visited Infective bait 0.51 (4, 0.33-0.75)1 0.56 (4, 0.02-0.78)1 Control bait 0.37 (4, 0.03-0.85)1 0.38 (6, 0.16-0.73) 1 at 88% C.I. due to sample size 2 at 75% C.I. due to sample size

Table 2.4: Experiment 1 was repeated such that beetles were allowed to feed on small amounts (see text) of fresh infective and control feces; two male and three female beetles did not contact either bait. Conventions as in Table 2.2.

64 Median proportion of time spent by Experiment 3 Females Males All beetles Infected side 0.47 (18, 0.23-0.70) 0.60 (19, 0.29-0.90) Beetles that visited both sides Infected cage 0.66 (17, 0.04-0.84) 0.49 (18, 0.11-0.77) Control cage 0.24 (17, 0-0.84) 0 (18, 0-0.54) Both cages visited Infected side 0.51 (9, 0.37-0.61) 0.59 (4, 0.08-0.71)1 Infected cage 0.66 (9, 0.27-0.89) 0.57 (4, 0.39-0.79)1 Control cage 0.61 (9, 0.11-0.91) 0.63 (4, 0.28-0.80)1 Only one cage visited Infected cage 0.84 (4, 0.67-0.92)1 0.74 (11, 0.22-0.91) Control cage 0.88 (4, 0.74-0.95)1 0.65 (4, 0.46-0.90)1 1 at 88% C.I. due to sample size

Experiment 4 Females Males All beetles Infected side 0.43 (20, 0.24-0.74) 0.50 (29, 0.22-0.73) Beetles that visited both sides Infected cage 0.43 (20, 0.03-0.72) 0.15 (25, 0.01-0.51) Control cage 0.31 (20, 0-0.58) 0.38 (25, 0.04-073) Both cages visited Infected side 0.38 (9, 0.26-0.66) 0.32 (10, 0.19-0.58) Infected cage 0.55 (9, 0.07-0.81) 0.38 (10, 0.06-0.64) Control cage 0.55 (9, 0.31-0.66) 0.58 (10, 0.34-0.82) Only one cage visited Infected cage 0.74 (6, 0.30-0.98) 0.53 (7, 0.24-0.78) Control cage 0.80 (4, 0.27-0.87)1 0.72 (7, 0.37-0.95) 1 at 88% C.I. due to sample size

Table 2.5: In experiments 3 and 4, beetles were prevented from feeding by enclosing the bait in a cage. In experiment 3, the arena contained thawed infective and control feces that had been collected and frozen immediately after defecation; one female beetle did not contact either bait. In experiment 4, the arena contained thawed infective and control feces that had been collected and frozen one hour after defecation; one male beetle did not contact either bait. Conventions as in Table 2.2.

65 Median proportion of time spent by Experiment 5 Females Males All beetles Infected side 0.55 (20, 0.21-0.83) Beetles that visited both sides Infected cage 0.69 (18, 0.45-0.82) Control cage 0.16 (18, 0-0.63) Both cages visited Infected side 0.40 (8, 0.13-0.58) Infected cage 0.64 (8, 0.14-0.78) Control cage 0.64 (8, 0.38-0.88) Only one cage visited Infected cage 0.82 (8, 0.71-0.83) Control cage 0.50 (2)

Table 2.6: In experiment 5, beetles were prevented from feeding by enclosing the bait in a cage. The arena contained thawed infective and control feces that had been collected and frozen six hours after defecation; two male beetles did not contact either bait. Conventions as in Table 2.2.

66 Median proportion of time spent by Experiment 6 Females Males All beetles Infected side 0.51 (27, 0.42-0.57) 0.57 (29, 0.49-0.62) Beetles that visited both sides Infective bait 0.11 (27, 0.03-0.17) 0.13 (29, 0.05-0.25) Control bait 0.09 (27, 0.03-0.23) 0.10 (29, 0.03-0.22) Both baits visited Infected side 0.51 (17, 0.42-0.57) 0.56 (23, 0.46-0.63) Infective bait 0.16 (17, 0.11-0.36) 0.17 (23, 0.09-0.35) Control bait 0.16 (17, 0.05-0.29) 0.22 (23, 0.11-0.40) Only one bait visited Infective bait 0.05 (4, 0-0.14)1 0.06 (4, 0.01-0.25)1 Control bait 0.23 (4, 0.03-0.47)1 0.07 (1) 1 at 88% C.I. due to sample size

Experiment 7 Females Males All beetles Infected side 0.52 (29, 0.44-0.59) 0.46 (30, 0.42-0.53) Beetles that visited both sides Infective bait 0.26 (29, 0.18-0.33) 0.23 (30, 0.20-0.30) Control bait 0.19 (29, 0.17-0.26) 0.26 (30, 0.20-0.30) Both baits visited Infected side 0.51 (24, 0.44-0.59) 0.47 (29, 0.43-0.53) Infective bait 0.25 (24, 0.18-0.32) 0.24 (29, 0.20-0.30) Control bait 0.20 (24, 0.18-0.27) 0.26 (29, 0.19-0.30) Only one bait visited Infective bait 0.70 (3, 0.62-0.8)1 none Control bait 0.84 (1) 0.54 (1) 1 at 75% C.I. due to sample size

Table 2.7: In experiment 6, the arena contained methanol extracts from infective and control feces; two female and three male beetles did not contact either bait. In experiment 7, the arena contained filter papers which had been exposed for at least three hours to infective and control feces; one female beetle did not contact either bait. Conventions as in Table 2.2.

67 Median proportion of time spent by Experiment 8 Females Males All beetles Infected side 0.55 (31, 0.37-0.66) 0.54 (30, 0.33-0.75) Beetles that visited both sides Infective bait 0.26 (31, 0.13-0.46) 0.30 (27, 0.17-0.45) Control bait 0.36 (31, 0.22-0.56) 0.51 (27, 0.20-0.70) Both baits visited Infected side 0.55 (25, 0.34-0.65) 0.56 (19, 0.38-0.76) Infective bait 0.29 (25, 0.17-0.48) 0.39 (19, 0.25-0.47) Control bait 0.36 (25, 0.25-0.54) 0.24 (19, 0.18-0.60) Only one bait visited Infective bait 0.71 (2) 0.70 (4, 0.10-0.97) 2 Control bait 0.78 (3, 0.71-0.99)1 0.96 (7, 0.91-0.97) 1 at 75% C.I. due to sample size 2 at 88% C.I. due to sample size

Experiment 9 Females Males All beetles Infected side 0.46 (29, 0.38-0.55) 0.48 (30, 0.28-0.67) Beetles that visited both sides Infective bait 0.36 (25, 0.10-0.46) 0.29 (28, 0.14-0.56) Control bait 0.39 (25, 0.24-0.49) 0.38 (28, 0.27-0.54) Both baits visited Infected side 0.50 (22, 0.40-0.58) 0.54 (21, 0.30-0.66) Infective bait 0.36 (22, 0.16-0.48) 0.30 (21, 0.17-0.57) Control bait 0.37 (22, 0.23-0.50) 0.40 (21, 0.30-0.59) Only one bait visited Infective bait 0.85 (2) 0.82 (3, 0.55-0.88)2 Control bait 0.96 (5, 0.44-0.96)1 0.96 (5, 0.38-0.96)1 1 at 93.8% C.I. due to sample size 2 at 75% C.I. due to sample size

Table 2.8: In experiments 8 and 9, beetles were allowed to feed. In experiment 8, the arena contained apple scrapings mixed with tapeworm eggs and apple scrapings mixed with D. I. water; one female beetle did not contact either bait. In experiment 9, the arena contained apple scrapings mixed with crushed tapeworm eggs and apple scrapings mixed with tapeworm eggs; one male beetle did not contact either bait. Conventions as in Table 2.2.

68 Median proportion of time spent by Experiment 10 Females Males All beetles Infected side 0.55 (30, 0.43-0.68) 0.44 (30, 0.18-0.82) Beetles that visited both sides Infective bait 0.41 (30, 0.25-0.59) 0.30 (29, 0.11-0.42) Control bait 0.31 (30, 0.22-0.48) 0.58 (29, 0.12-0.64) Both baits visited Infected side 0.55 (27, 0.48-0.69) 0.38 (21, 0.17-0.72) Infective bait 0.41 (27, 0.28-0.60) 0.34 (21, 0.15-0.48) Control bait 0.36 (27, 0.22-0.50) 0.60 (21, 0.29-0.71) Only one bait visited Infective bait 0.90 (1) 0.52 (5, 0.23-0.89)1 Control bait 0.28 (2) 0.66 (4, 0-0.94)2 1 at 93.8% C.I. due to sample size 2 at 88% C.I. due to sample size

Table 2.9: In experiment ten, beetles were allowed to feed on control feces mixed with tapeworm eggs and control feces mixed with D. I. water. Conventions as in Table 2.2.

69 Median number of bait contacts by Experiment 3 Females Males Infected cage All beetles 1.0 (30, 0-1.0) 1.0 (30, 0.23-2.0) Contacted both cages 1.0 (9, 1.0-2.0) 2.0 (12, 1.3-3.0) Control cage All beetles 1.0 (30, 0-2.0) 1.0 (30, 0-2.0) Contacted both cages 2.0 (9, 1.0-2.0) 1.5 (12, 1.0-2.7)

Experiment 4 Infected cage All beetles 1.0 (20, 1.0-2.0) 1.0 (29, 0.8-1.0) Contacted both cages 2.0 (9, 1.0-2.0) 1.0 (10, 1.0-1.7) Control cage All beetles 1.0 (20, 0-2.0) 1.0 (29, 0-1.2) Contacted both cages 2.0 (9, 1.2-4.5) 1.5 (10, 1-3.3)

Experiment 5 Infected cage All beetles 2.0, (20, 1.0-2.0) Contacted both cages 3.5 (8, 1.0-5.1) Control cage All beetles 1.0 (20, 0-3.0) Contacted both cages 2.0 (8, 1-4.1)

Experiment 6 Infective bait All beetles 3.0 (27, 1.0-4.0) 1.0 (29, 1.0-4.0) Contacted both baits 4.0 (17, 3.0-7.0)b 2.0 (21, 1.0-4.3)b Control cage All beetles 2.0 (27, 1.0-4.0) 2.0 (29, 1.0-2.0) Contacted both baits 4.0 (17, 2.0-6.0)b 2.0 (21, 1.0-3.0)b

Table 2.10: Median number of bait contacts for all beetles and for beetles that contacted both baits/cages in experiments 3-6. See text for descriptions of experiments. Conventions as in Table 2.3.

70 Median number of bait contacts by Experiment 7 Females Males Infective bait All beetles 7.0 (29, 3.8-8.0) 7.0 (30, 5.0-8.0) Contacted both baits 7.0 (24, 4.0-8.2) 7.0 (29, 5.0-8.0) Control bait All beetles 6.0 (29, 3.0-6.0) 7.0 (30, 5.2-9.8) Contacted both baits 6.0 (24, 4.0-7.0) 7.0 (29, 5.8-10)

Experiment 8 Infective bait All beetles 5.0 (31, 2.0-9.0) 6.0 (30, 1.5-9.0) Contacted both baits 8.0 (25, 3.2-9.0) 9 (19, 7.0-13) Control bait All beetles 5.0 (31, 2.7-6.0) 3.5 (30, 1.0-5.0) Contacted both baits 6.0 (25, 3.0-7.0) 5.0 (19, 3.7-10)

Experiment 9 Infective bait All beetles 4.0 (29, 2.8-5.2) 5.5 (30, 3.0-8.0) Contacted both baits 4.5 (22, 3.0-10.0) 8.0 (21, 4.7-10.3) Control bait All beetles 4.0 (29, 3.0-7.2) 6.5 (30, 4.2-8.8) Contacted both baits 5.5 (22, 4.0-9.0) 8.0 (21, 6.7-11.3)

Experiment 10 Infective bait All beetles 9.5 (30, 8-11)b 4.0 (30, 2.2-5.8)b Contacted both baits 11.0 (27, 8-12) 5.0 (21, 3-11.7) Control bait All beetles 6.5 (30, 4.0-8.8)b 4.0 (30, 2.0-5.8)b Contacted both baits 7.0 (27, 4.0-10.1) 5.0 (21, 3.7-8.0)

Table 2.11: Median number of bait contacts for all beetles and for beetles that contacted both baits in experiments 7-10. See text for descriptions of experiments. Conventions as in Table 2.3.

71 Median proportion of time spent by Experiment 11 Females Males All beetles Side w/ beetle 0.84 (30, 0.42-0.95) 0.81 (30, 0.43-0.92) Beetles that visited both sides Cage w/ beetle 0.76 (27, 0.49-0.87) 0.73 (23, 0.59-0.83)a Cage w/o beetle 0.17 (27, 0-0.74) 0 (23, 0-0.59)a Both cages visited Side w/ beetle 0.46 (10, 0.28-0.77) 0.64 (10, 0.39-0.83) Cage w/ beetle 0.78 (10, 0.49-0.89) 0.68 (10, 0.50-0.86) Cage w/o beetle 0.60 (10, 0.32-0.79) 0.63 (10, 0.27-0.73) Only one cage visited Cage w/ beetle 0.89 (14, 0.75-0.93) 0.80 (13, 0.67-0.93) Cage w/o beetle 0.90 (6, 0.88-0.96) 0.96 (6, 0.78-0.98)

Table 2.12: Beetles were prevented from feeding on refrigerated feces enclosed in wire cages. In experiment 11, the arena consisted of a cage containing infective bait with a beetle and a cage containing infective bait alone; one male beetle did not contact either bait. Conventions as in Table 2.2.

72 Median number of bait contacts by Experiment 11 Females Males Cage with beetle All beetles 1.5 (30, 1.0-2.0)a 1.0 (30, 1.0-2.0)a Contacted both cages 2.0 (10, 1.0-3.7)b 2.0 (10, 1.0-2.3)b Cage without beetle All beetles 1.0 (30, 0-2.0)a 1.0 (30, 0-1.0)a Contacted both cages 3.5 (10, 2.0-5.3)b 1.0 (10, 1.0-2.3)b

Experiment 12 Infected cage with beetle All beetles 2.0 (31, 1.0-2.3) 2.0 (32, 1.0-4.0) Contacted both cages 2.0 (15, 1.0-3.6) 3.0 (15, 2.0-4.6) Control cage with beetle All beetles 1.0 (31, 0-2.0) 2.0 (32, 1.0-2.0) Contacted both cages 2.0 (15, 1.0-4.6) 2.0 (15, 1.4-3.6)

Experiment 13 Cage with beetle All beetles 2.0 (30, 1.0-2.8) 1.0 (30, 1.0-2.0) Contacted both cages 2.0 (14, 1.0-3.0) 2.0 (10, 1.0-3.3) Cage without beetle All beetles 1.0 (30, 1.0-2.0) 1.0 (30, 0-1.0) Contacted both cages 1.0 (14, 1.0-3.0) 1.5 (10, 1.0-3.0)

Table 2.13: Median number of bait contacts for all beetles and for beetles that contacted both cages in experiments 11-13. See text for descriptions of experiments. Conventions as in Table 2.3.

73 Median proportion of time spent by Experiment 12 Females Males All beetles Infected side w/ beetle 0.54 (31, 0.40-0.77) 0.46 (32, 0.22-0.70) Beetles that visited both sides Infected cage w/ beetle 0.56 (29, 0.40-0.69) 0.60 (26, 0.41-0.78) Control cage w/ beetle 0.30 (29, 0-0.65) 0.53 (26, 0.14-0.81) Both cages visited Infected side w/ beetle 0.50 (15, 0.39-0.58) 0.46 (15, 0.30-0.66) Infected cage w/ beetle 0.56 (15, 0.38-0.65) 0.60 (15, 0.42-0.79) Control cage w/ beetle 0.64 (15, 0.30-0.72) 0.65 (15, 0.21-0.86) Only one cage visited Infected cage w/ beetle 0.86 (9, 0.67-0.93) 0.74 (8, 0.62-0.92) Control cage w/ beetle 0.94 (5, 0.77-0.98) 1 0.90 (9, 0.61-0.94) 1 at 93.8% C.I. due to sample size

Experiment 13 All beetles Side w/ beetle 0.60 (30, 0.35-0.81) 0.68 (30, 0.36-0.83) Beetles that visited both sides Cage w/ beetle 0.64 (27, 0.39-0.80 0.67 (26, 0.31-0.79) Cage w/o beetle 0.66 (27, 0.18-0.73) 0.56 (26, 0-0.85) Both cages visited Side w/ beetle 0.53 (14, 0.39-0.74) 0.47 (10, 0.33-0.77) Cage w/ beetle 0.77 (14, 0.59-0.82) 0.71 (10, 0.50-0.80) Cage w/o beetle 0.72 (14, 0.67-0.82) 0.80 (10, 0.55-0.92) Only one cage visited Cage w/ beetle 0.64 (9, 0.39-0.90) 0.83 (12, 0.65-0.93) Cage w/o beetle 0.89 (7, 0.20-0.95) 0.88 (8, 0.82-0.97)

Table 2.14: Beetles were prevented from feeding on refrigerated feces enclosed in wire cages. In experiment 12, the arena consisted of a cage containing infective bait with a beetle and a cage containing control bait with a beetle; two female beetles did not contact either bait. In experiment 13, the arena consisted of a cage containing control bait with a beetle and a cage containing control bait alone. Conventions as in Table 2.2.

Figure 2.1: Picture of plastic arena used in preference trial experiment. Dimensions of the arena were 10.5 X 10.5 cm and each cage was 2.5 X 2.5 X 2.5 cm. Beetles were placed in the middle of the arena and allowed to acclimatize for 7 minutes at which time the glass tube was removed, and the beetle’s movements were recorded for 15 minutes. Position of infective feces, at opposite corners in cages (unless otherwise described), was determined randomly, and the arena was rotated 90° after each trial.

75 Observed and Expected Standard Deviations based on Binomial Distribution

0.45

0.40 overlapping points Æ

0.35

0.30

0.25 Observed, Females K = 30 0.20 K = 15 K = 2 Standard Deviation Standard Observed, Males 0.15

0.10

0.05 12345678910111213

Experiment

Figure 2.2: Observed standard deviation of the proportion of time spent on the infected side for each sex within each trial showing their relative deviations from expected binomial distributions. Indicated k values represent the hypothetical situation in which the beetle would have k opportunities to make choices in the 15 minute trial. Experiment 5 tested males only. Note that the observed standard deviations for females were generally lower than that for males, which had lower activity levels. Also, the two lowest observed standard deviations occur in experiments 6 and 7, which both involved non- food items.

Observed and Expected Standard Deviations based on Binomial Distribution

0.40

0.35

0.30

0.25

0.20

Standard Deviation 0.15

0.10

0.05 abcd

Infected or female Experiment Control or male K = 2 K = 15 K = 30

Figure 2.3: Observed standard deviation of the proportion of time spent on bait in experiments with significant findings. Females that visited both sides in experiment 1 spent more time on the infective bait than on the control bait (a). Females that contacted both baits in experiment 1 spent more time on the infective bait than on the control bait (b). Of beetles that visited both sides in experiment 2, males spent more time on the infected cage than females (c). Males that visited both sides in experiment 11 spent more time on the cage with the beetle than on the cage without a beetle (d). Data from Tables 2.2 and 2.12. Indicated k values represent the hypothetical situation in which the beetle would have k opportunities to make choices in the 15 minute trial.

CHAPTER 3

PARASITE-ALTERED PREFERENCE FOR INFECTIVE FECES IN INFECTED INDIVIUAL HOSTS AND GROUPS OF INFECTED HOSTS

ABSTRACT

Infected hosts often behave differently from uninfected hosts and may influence the probability of host re-infection. The beetle-tapeworm lifecycle provides a convenient system to study this. The beetle, Tenebrio molitor, becomes infected when it ingests rat feces containing the eggs of the tapeworm, Hymenolepis diminuta. Previous studies demonstrated that beetles prefer infective feces over uninfective feces, but these studies involved uninfected beetles. In the field, both uninfected and infected beetles would be present. This study tests the fecal preference of infected beetles in groups and individually. Results suggest that infected beetles show no preference for infective feces, which may be a host adaptation to avoid further infection, parasite manipulation to avoid competition for host resources, or both. Further, once infected, beetles are no more or no less likely to become re-infected than uninfected beetles. An analysis of the mean and variance of infection suggests that some individuals are highly susceptible and some are highly resistant to infection with males being more variable than females. This may explain the higher load of cysticercoids observed in males.

78 INTRODUCTION

Parasites manipulate the behavior of their hosts to increase the probability of transmission to the next host (Holmes and Bethel 1972; Moore 2002). For example, ants infected with the digenetic trematode, Dicrocoelium dendriticum, climb atop and cling to blades of grass where they are more likely to be eaten by sheep, the definite host

(Anokhin 1966; Carney 1969). In another example, when amphipods carrying infective stages of larval trematodes are eaten by a bird, then the bird is infected. During the day, amphipods with the infective stage of the trematode are more likely to be crawling on the ground than to be burrowing underground, making them more susceptible to consumption by birds, which feed during the day (McCurdy et al. 1999). However, the prevalence of the pre-infective trematode stage does not differ between crawling and burrowing amphipods. Also, at night, neither infective nor pre-infective amphipods are associated with increased crawling (McCurdy et al. 1999). Clearly, parasite and host life histories

play an important role in determining if and when parasite manipulation of host behavior

will occur. Such factors need to be considered in other host-parasite systems.

In the tapeworm, Hymenolepis diminuta, adults live in the small intestines of

rodents, the definitive host. Eggs pass out of the rodents in their feces. The cysticercoid,

or larval stage, occurs in an insect (the intermediate host, most often a beetle) that

becomes infected by feeding upon infective feces. The rodent completes the cycle when

it ingests an infected beetle. Thus, the success of this parasite depends on rodent-to-

beetle and beetle-to-rodent transmission, and the former depends upon consumption of

the egg stage by the beetle. Recent evidence suggests that even this passive type of

transmission involves a mechanism that influences host behavior to benefit the tapeworm.

79 Three studies tested the fecal preference of groups of beetles by allowing them to

forage in an arena with two types of food: rat feces from uninfected rats and rat feces

from rats infected with H. diminuta (presumed to contain eggs). Beetles (Tribolium

confusum) starved 48 hours prior to the trial preferred infective feces (Evans et al. 1992).

When beetles (Tenebrio molitor) of known sex, age, and feeding history are tested,

female beetles (those fed and those starved 72 hours prior to the trial) preferred infective

feces. Starved male beetles preferred infective feces, while fed males preferred

uninfective feces (Pappas et al. 1995). In the third study, an analysis of beetle (T.

confusum) movement showed that starved beetles (of mixed sex and age) made fewer

visits to infective bait, but the duration of visits to each bait did not differ when tested

individually. When tested in groups during a two hour trial, starved and fed beetles

preferred control bait during two time intervals, but showed no preference during the

other four time intervals. The authors concluded that beetle behavior appeared “highly

heterogeneous, both among individuals and by the same individual over time” (Shostak

and Smyth 1998). These experiments involved uninfected beetles, but in the field, both

infected and uninfected beetles can be found (Rau 1979). Would infected beetles behave

differently? They may if there are costs associated with infection.

Infection by Hymenolepis diminuta cysticercoids harms the beetle host in several ways. Reduction in host fecundity occurs in both T. confusum (Keymer 1980, 1981) and

T. molitor (Hurd and Arme 1986). Infected male T. molitor exhibit a lowered response to

pheromone produced by uninfected females (Hurd and Parry 1991). Defensive glands of

infected T. molitor contain less defense compounds than uninfected controls

(Blankespoor et al. 1997). Fat body glycogen reserves are significantly depleted 3 days

80 post-infection in male and 5 days post-infection in female T. molitor (Kearns et al. 1994).

Finally, infected males are less attractive to females, and females mated with infected

males produce fewer offspring than females mated with uninfected males (Worden et al.

2000). Thus, we may expect altered behavior in hosts that decreases the probability of

parasite transmission. Yet, based on previous experiments (Evans et al. 1992; Pappas et

al. 1995; Shostak and Smyth 1998), this is not always the case because the gender, social

status, and feeding history of a beetle influence its preference for infective or uninfective

feces.

The behavior of infected hosts may differ depending upon the gender of the host

because costs to infection differ by gender (Worden et al. 2000; Kearns et al. 1994).

Thus, one sex may be under greater selective pressure to avoid infective feces than the

other. Grain beetles produce sex and aggregation pheromones (Burkholder and Ma 1992,

August 1971, Tschinkel et al. 1967), which may influence the behavior of beetles

foraging in groups. Yet, group foraging behavior may be more typical, so both

conditions should be tested. Finally, both starved male and female beetles preferred

infective feces, but fed males avoided infective feces (Pappas et al. 1995), while

individual fed beetles did not differ in number or duration of contacts between infective

and control baits (Shostak and Smyth 1998). Perhaps starved beetles are more susceptible to manipulation by the parasite such that they tend to prefer infective feces, and thus represent a more interesting and, perhaps, more natural condition to test.

To assess the relative importance of infection status this study repeats the preference experiments using infected individual and groups of starved beetles of known sex and age. Infected beetles may act to avoid infective feces if super-infections result in

81 reduced host fitness or death. If heavy infections results in death, then parasites would also benefit when infected beetles avoid or show no preference for infective feces.

Parasites may also benefit when their hosts avoid infective feces if re-infection results in competition with unrelated parasites for limited resources (Chapter 1). Thus, I hypothesize that infected beetles will avoid or show no preference for infective feces. To test this, beetle movement in an arena containing infective and uninfective feces is recorded and analyzed for preference. The novel approach to this study provides me with data to ask if infected beetles are more or less susceptible to a second infection. I hypothesize that there will be no difference in susceptibility to infection. To test this, observations from the study are used to assess infection variance, infection bias, and the relationship between initial and secondary infections.

METHODS

The “OSU Strain” (Pappas and Leiby 1986) of H. diminuta was maintained in male Sprague-Dawley rats and beetles (T. molitor). Because this is a highly inbred strain of tapeworm (having survived a severe bottleneck in 1995), I assume it has low genetic variability, which should minimize concern about parasite genetic variation confounding the results of host behavior (but see Meffert 1999). Three male Sprague-Dawley rats were infected with 30 cysticercoids and maintained on commercial rodent chow and water. Three additional rats, obtained from the same commercial source, of identical age and when possible, from the same litter so as to minimize rat genetic variability, were maintained under the same conditions as the infected rats to serve as the source of control

(uninfective) feces. Fecal pellets from infected rats were not used prior to 20 days post-

82 infection. Rat cages were checked every 10 minutes on the morning of the trials, and

fecal pellets were collected with forceps to minimize contamination by rat urine. After

each trial, the infective fecal pellets were examined to verify the presence of H. diminuta

eggs.

Beetles were maintained on wheat bran, and small pieces of potato were added to

the cultures on a regular basis. Pupae were removed from the cultures, and the male and

female pupae (Bhattacharya et al. 1970) were placed in separate dishes containing wheat bran. Beetles that emerged during a 24 h period were collected such that a daily cohort of

beetles was maintained for both sexes.

Individual trials

Fifteen to 17 days before the trial, male and female beetles (each in groups of ten) were exposed to a consistent concentration of tapeworm eggs (one gram of air dried apple scrapings mixed with a 0.05 ml solution of water and tapeworm eggs) for 24 hours. This concentration resulted in an average infection intensity of 12.5 cysticercoids per beetle (n

= 53, ranging from 1 to 41) and simulated Rau’s (1979) field study who found an average

cysticercoid load of 10.5 per beetle.

The preference experiment consisted of 53 trials that tested 27 male and 26

female infected beetles individually in a small plastic arena divided down the middle (Fig

3.1). Pairs of fresh control and infective fecal pellets were matched according to time of

defecation (within one hour of each other) and used immediately as a paired “batch.”

Each pellet was placed in its own plastic screw top vial, labeled and covered so that the

preference trial was conducted blind. Bait placement was assigned randomly at the start

of each trial, and the arena was rotated 90° after each trial. To create a homogenous

83 distribution of eggs, the fecal pellets were mixed with a stirrer for one minute. From

these paired batches, feces, matched by weight (28-37mg) to within 2 mg, were placed on

cover slips and one paired batch of feces was used for no more than four trials. The

positions of both baits were alternated in successive trials. Beetles of known sex, age

(21-25 days) and feeding history (starved 2-3 days) were tested in a 15 minute trial (after

a 7 minute acclimatization period). I recorded the side to which the beetle first moved, the time spent on each side, the time spent on each type of bait, the number of contacts with each type of bait, and the number of times the midline of the arena was crossed. Six days after the trial, each beetle was dissected and the number of each stage of cysticercoid was recorded. Voge and Heyneman (1957) showed that cysticercoids undergo five morphologically distinct stages while developing inside the beetle. Beetles with one or more stage 5 cysticercoids were considered infected at the time of the trial, and only these were included in the preference analysis. All other beetles, those with no cysticercoids or those with stages one through four cysticercoids only, were considered uninfected at the time of the trial. Thus, infected beetles that consume infective feces during the trial may experience a secondary infection.

Data analysis

The proportion of time spent on the side containing the infective feces was tested with a Wilcoxon Signed Rank test for equality to 50% for all male and female beetles.

This test assumes a binomial distribution where the presence of the beetle in the next moment (on the infective or uninfective side) is independent of current location.

However, this assumption is not met because the beetle’s location in the next moment of time depends on its current location. This autocorrelation introduces greater than normal

84 binomial variance. As evidence, the observed standard deviation for the average

proportion of time spent on the infected side equals 0.42 for the individual trials and 0.28

for the group trials, while the expected deviation (based on a binomial distribution)

equals 0.09 if k = 30, 0.16 if k = 10, and 0.22 if k = 5 where k equals the number of

hypothetical opportunities to make discrete, binary choices during the trial. Thus, large

deviations make it difficult to demonstrate small preferences.

The Fligner-Policello procedure tested for a median difference in the proportion

of time spent on infective and control baits between male and female beetles that

experienced both sides (Hollander and Wolfe 1999). As paired data, the difference

between the proportion of time spent on infective and control baits was tested for equality

to zero with a Wilcoxon Signed Rank test. Confidence intervals (95% except where

indicated) were constructed about each median using a nonlinear interpolation procedure

(Minitab v. 13.3, State College, PA).

Because 15 of 26 males and 17 of 27 females spent no time on either the infective

or control bait, the data were re-analyzed with the same tests using the data for beetles

that contacted both baits, and significance was adjusted to 0.025 according to the post hoc

Bonferroni adjustment. A two-tailed Fisher exact test (GraphPad, San Diego, CA) was used to evaluate the association between sex and beetles that contacted both baits and beetles that contacted only one bait. For beetles that contacted only one bait, the same test analyzed the relationship between sex and beetles that contacted infective bait and beetles that contacted control bait. Finally, the number of bait contacts was analyzed for all beetles using the Scheirer-Ray-Hare extension of the Kruskal-Wallis test (with a correction for ties) for ranked data in a 2-way design with treatment and sex as factors

85 (Sokal and Rohlf 1995). Beetles that contacted both baits were analyzed with the same

test. Again, significance was adjusted to 0.025 for the question of treatment and to

0.0167 for the question of sex effects (because the Fisher Exact test asks this question

three times) according to the post hoc Bonferroni adjustment. Also for beetles that

contacted both sides, the proportion of time spent on the infected cage was plotted against

the number of cysticercoids. Finally, the Fligner-Policello procedure (Hollander and

Wolfe 1999) tested for differences in the medians between male and female cysticercoid

loads without assuming equal variances.

The data were then divided on the basis of the beetles’ infection status: those

with successful initial infections (defined as those with stage five cysticercoids), those

with successful secondary infections (defined as those with stages one through four

cysticercoids), those with both, and those with neither. Their relative numbers were

analyzed with a 2-tailed Fisher exact test.

Group trials

Seven to nine days before the trial, male and female beetles (in groups of ten)

were allowed to feed on one gram of air dried apple scrapings mixed with a 0.01 ml

solution of water and tapeworm eggs for 24 hours. Immediately before the trial, each

beetle was marked with latex-based paint so individuals could be tracked through the

trial. No food was provided 20-35 hours before the beetles’ respective trials. Eleven

groups of males and females (due to mortality, there were between eight and ten beetles

per group) were used in the preference trial. Each group was placed under a glass bowl in the center of a plastic arena (that was scrubbed twice and rinsed in hot water for at

least 30 minutes before each trial) under red light conditions. Beetles were provided with

86 two types of bait – uninfective (control) or infective feces. The control feces were

randomly positioned either in areas 1 and 3, or 2 and 4 (Fig. 3.2). The infective feces

were then positioned in the other areas. The bait areas were large enough for all beetles

to occur simultaneously. A paper towel was used to mash the fecal pellets on the squares

to prevent their displacement by beetles during the trial. After 15 minutes, the bowl was

removed and the beetles’ movements were videotaped for one hour under red light

conditions. The video was played back and the square that each individual beetle was on was recorded at one minute intervals as an occurrence. Six days after the preference trial, each beetle was dissected and the number of each stage of cysticercoid was recorded.

Beetles with no cysticercoids or those with stages one through four cysticercoids were considered uninfected at the time of the trial.

Data analysis

Preference was measured as the difference between the number of occurrences at infective and uninfective baits and determined with the Mixed procedure in SAS (v.8,

Cary, NC) with sex and trials as class variables and sex nested in trial as a random variable. The response variable was defined as (number of occurrences at the infective

bait) – (number of occurrences at the control bait). The analysis was repeated after

including number of cysticercoids in the model. The Fligner-Policello procedure

(Hollander and Wolfe 1999) tested for differences in the medians between male and

female cysticercoid loads without assuming equal variances. For each trial, the average cysticercoid infection level (both early stage and stage 5 cysticercoids) was plotted against its variance, and the slope was compared to one (Minitab v. 13.3, State College,

PA) to assess its similarity to a Poisson distribution. The data were then divided on the

87 basis of the beetles’ infection status: those with successful initial infections (defined as

those with stage five cysticercoids), those with successful secondary infections (defined

as those with stages one through four cysticercoids), those with both, and those with

neither. Their relative numbers were analyzed with a 2-tailed Fisher exact test.

RESULTS

Individual trials

Of beetles that experienced both sides, neither females nor males spent a greater proportion of time on the infective feces than on the control feces (Table 3.1; Wilcoxon

Signed Rank, W♀ = 179, p = 0.667, W♂ = 174, p = 0.127). Further, males did not differ

significantly from females in the proportion of time spent on the control bait (Table 3.1;

Fligner-Policello, p = 0.3907) or infective bait (p = 0.0568) after a post hoc Bonferroni

adjustment. Of beetles that contacted both baits, neither females nor males spent a greater proportion of time on the infective feces than on the control feces (W♀ = 24, n =

11, p = 0.45; W♂ = 19, n = 9, p = 0.722). Further, males did not differ from females in

the proportion of time spent on the control bait (p = 0.374) or infective bait (p = 0.228).

There were no treatment or sex differences for number of contacts with either baits when all beetles were considered, or when beetles that contacted both baits were considered

(Table 3.2; Scheirer-Ray-Hare, p > 0.3 for all parameters).

The median load of stage 5 cysticercoids for males (median = 14) was significantly higher than the median load for female (median = 6) beetles (Fligner-

Policello, n♂ = 27, n♀ = 26, U = 2.26, p = 0.0119). The median for each sex remained the

same when data for beetles with early stage cysticercoids were added to those of beetles

88 with the stage 5 cysticercoids so that males still had the higher total median load of cysticercoids (Fig. 3.3; Fligner-Policello, U = 2.078, p = 0.0189).

The number of stage five cysticercoids was not correlated with the proportion of time beetles spent on the infected cage for either males (r = 0.06, power = 0.049, p =

0.808) or females (r = 0.094, power = 0.067, p = 0.655). Further, the number of early- stage cysticercoids was not correlated with the number of stage five cysticercoids for either male (r = -0.12, power = 0.0052, p = 0.547) or female (r = -0.25, power = 0.0007, p

= 0.213) beetles.

The number of beetles with stage five cysticercoids was independent of the number of beetles with early cysticercoids stages (Fig. 3.4) for females (2-tailed Fisher exact test, n = 29, p = 1.0), and for males (2-tailed Fisher exact test, n = 29, p = 0.069) at p = 0.05.

Group trials

Preference for infective feces between infected male and female beetles did not differ (Table 3.4; mixed procedure, p = 0.626). The average difference in the total number of occurrences (infective bait occurrence – uninfective bait occurrence) of infected beetles did not differ from zero for females (-2.73, p = 0.626) and males (-5.68, p

= 0.332). Preference for infective feces between infected male and female beetles was unrelated to the number of stage five cysticercoids in each beetle (p = 0.704). Again, the average difference in the total number of occurrences of infected beetles did not differ from zero for females (-2.26, p = 0.712) or males (-5.34, p = 0.385). Of the 35 uninfected females in 11 trials, the average difference in the total number of occurrences did not differ from zero (-4.7, n = 11, p = 0.641). Of the 15 uninfected males in eight trials, the

89 average difference in the total number of occurrences did not differ from zero (8.37, n =

8, p = 0.285).

A plot of the average number of early stage cysticercoids against its variance

resulted in a slope significantly greater than one for both male (b = 6.99, p < 0.001) and

female (b = 3.93, p < 0.001) beetles (Fig. 3.5). The same was true for a plot of stage 5

cysticercoids for males (b = 9.94, p < 0.001) and females (b = 3.0, p = 0.0042). When

compared, the slope for males was steeper for early stage cysticercoids (F1,18 = 33.3, power = 0.045, p < 0.0001) and stage 5 cysticercoids (F1, 18 = 7.36, power = 0.702, p =

0.0143).

Median stage 5 and total cysticercoid distributions between males and females did not differ (Fligner-Policello, U = 0, p > 0.15). There was no correlation between the number of early-stage cysticercoids and number of stage five cysticercoids for either male (r = -0.02, power = 0.016, p = 0.848) or female (r = 0.015, power = 0.033, p =

0.905) beetles.

The number of beetles with stage five cysticercoids was independent of the number of beetles with early cysticercoid stages (Fig. 3.6) for females (2-tailed Fisher exact test, n = 105, p = 0.408), and for males (2-tailed Fisher exact test, n = 105, p =

0.409).

DISCUSSION

Bait preference

Infected beetles of either sex do not prefer one type of bait over another when tested individually. Additionally, infected beetles of either sex do not exhibit a

90 preference for either the infective or uninfective baits when tested in groups (Table 3.4).

A smaller sample size, uninfected male beetles behave in the opposite (but not significantly different) direction of infected beetles, which is consistent with the results of previous studies (Chapter 2, Pappas et al. 1995; Evans et al. 1992). These previous studies demonstrated that uninfected beetles show a preference for infective feces and that sex plays a role in the beetle’s preference. Combined, these results suggest that beetles, once infected, lose their fecal preference. This may be an adaptation on the part of the tapeworm to reduce competition for host resources. Once established in their hosts, some of the larvae of certain parasitoids develop into a “soldier caste”, which eliminate other parasites that would compete for host resources (Cruz 1981; Cruz et al.

1990; Utsunomiya and Iwabuchi 2002). However, there is no evidence that host resources limit parasite growth and development at the infection intensity levels studied in Chapter 1. The lack of fecal preference in infected beetles may be a host adaptation to prevent a lethal infection level. Hosts can alter their behavior to avoid parasites (Hart and

Hart 1994; Hart 1997; Moore 2002) or to mitigate the costs of parasitism once infected such as in behavioral fever (Boorstein and Ewald 1987; Adamo 1998). Both parasite manipulation and host adaptation may be occurring if the interests of host and parasite do not conflict. Further, data from the individual trials suggest that at the level of infection achieved in this study, the proportion of time spent on the infective bait is not dose- dependent.

91 Infection differences

Studying the behavior and bait preference of infected hosts allows for the study of

the infection itself. Specifically, observations concerning infection variance, infection

bias, and the relationship between initial and challenge infections become possible.

If cysticercoid infection follows a Poisson distribution (Rau 1979), then a plot of

mean infection against its variance should generate a slope equal to one. However, plots of both initial and secondary infection generated slopes that are significantly greater than

one. This is true for both males and females, suggesting that some beetles are highly

susceptible to infection while others are highly resistant to infection. Further, the male

slope was significantly steeper than the female slope for both initial and secondary

infections. Previous results suggest that male foraging activity could lead to infection

levels that are extremely high in some individuals and extremely low in others (Chapter

2). When allowed to feed, male activity level was lower than females, suggesting that

males spend a greater amount of time feeding on infective and uninfective feces. If true,

then females should have a more homogenous distribution of cysticercoids than males.

This is supported by the highly variable male distribution (Fig. 3.5) and the male’s higher

median load of cysticercoids (Fig. 3.4). However, there is no difference in bait contact

between males and females so activity level cannot explain the highly variable

distributions in the infection data or the male infection bias. Males do feed more than

females (Chapter 1) and if males differ in their susceptibility to infection, then natural

selection can favor increased feeding activity in those males that are highly resistant to

infection. This may also explain the large variation observed in the male infection

distribution (Fig. 3.5) as well as the male infection bias.

92 A similar male infection bias was found in the experiment using starved beetles, but not in the experiment using fed beetles (Chapter 1), but this may be explained by the differences in beetle mortality between the two experiments. When beetles are fed after being exposed to infective rat feces for one hour, more cysticercoids are recovered from male than from female beetles (Pappas et al. 1995). Yet, males of the group trials in this experiment do not show the infection bias observed in the individual trials despite nearly identical infection procedures. Further, previous studies failed to find a difference in infection intensity between the sexes in Tenebrio beetles infected by feeding on gravid proglottids (Hurd and Arme 1987) or in natural populations (Rau 1979). The presence of conspecifics may influence male behavior since they spend more time on a cage containing infective feces and another beetle than on a cage containing infective feces only (Chapter 2). Additionally, both males and females displayed higher activity levels around the cage containing infective feces with a conspecific. The interaction of beetle activity levels with the presence of conspecifics, and the nutritional status of males and females may explain why male infection bias is present or absent in the previously described studies.

Because data were available on challenge infections, I examined the relationship between initial and secondary infections. It is conceivable that infected beetles may be resistant to secondary infections. This does not seem to be the case for beetles of the group experiment or with females of the individual trials. Neither the number of cysticercoids or simple presence of cysticercoids could be correlated between initial and secondary infections. This lack of correlation suggests that beetles are unable to initiate an immune response to re-infection during the course of the trial. With males of the

93 individual trials, the secondary infection is independent of the initial infection at p =

0.069, which is also consistent with the idea that beetles, once infected, are not any more or any less resistant to further infection. However, because conditions for the initial and secondary infection differ, it is difficult to interpret the result from the individual trials.

In conclusion, infected male and female beetles, both in groups and individually, show no preference for infective bait, which is contrary to previous studies suggesting that some uninfected beetles do prefer infective feces (Evans et al. 1992; Pappas et al.

1995; Shostak and Smyth 1998; Chapter 2). The parasite may manipulate the host to prevent competition for limited host resources with other parasites. However, host resources do not limit parasite growth under the infection intensity levels tested in

Chapter 1. The host may avoid re-infection so as to prevent potentially lethal loads of parasites. Both parasite manipulation and host adaptation may explain the lack of preference for infective feces by infected beetles since a lack of preference will not increase the probability of re-infection. The male infection bias (observed in Pappas et al. 1995; Chapter 1) observed in this experiment could not be explained by differences in activity levels, but is consistent with the highly variable infection distribution observed in males. Although the initial and secondary infection protocols differed, the lack of correlation between initial and secondary infection levels suggests that infected beetles are no more or no less resistant to re-infection.

Median proportion of time spent by Individual beetle experiment Females Males All beetles Infected side 0.62 (26, 0.13-0.89) 0.89 (27, 0.12-0.98)

Beetles that visited both sides Infective bait 0.28 (25, 0.03-0.59) 0.72 (22, 0.32-0.88) Control bait 0.14 (25, 0-0.51) 0.10 (22, 0-0.61)

Both baits visited Infected side 0.30 (11, 0.08-0.80) 0.30 (9, 0.06-0.80) Infective bait 0.28 (11, 0.07-0.48) 0.37 (9, 0.02-0.85) Control bait 0.43 (11, 0.13-0.80) 0.57 (9, 0.18-0.77)

Only one bait visisted Infective bait 0.70 (9, 0.25-0.89) 0.89 (12, 0.68-0.93) Control bait 0.47 (6, 0.21-0.90) 0.86 (5, 0.74-0.96)*

* 93.8% C. I. due to sample size

Table 3.1: Individual beetles, after infection, were allowed to feed on fresh infective or control feces; one female beetle did not contact either bait. Sample sizes and 95% median confidence intervals (except where noted) are in parenthesis. There were no significant median differences between treatments or between sexes. All p’s > 0.05 after a post hoc Bonferroni adjustment.

Median number of bait contacts by Females Males Infective bait All beetles 2.0 (26, 1.0-3.0) 2.0 (27, 1.0-3.0) Contacted both baits 2.0 (11, 1.0-6.0) 2.0 (9, 1.0-8.4)

Control bait All beetles 1.0 (26, 0-4.0) 1.0 (27, 0-3.0) Contacted both baits 4.0 (11, 1.0-7.0) 4.0 (9, 1.2-6.8)

Table 3.2: Median number of bait contacts for all beetles and for beetles that contacted both baits/cages in the individual beetle experiment. Sample sizes and 95% median confidence intervals (except where noted) are in parenthesis. All p’s > 0.3 after a post hoc Bonferroni adjustment.

96 Females Males Trial Uninfective Infective Difference (n, s.d.) Uninfective Infective Difference (n, s.d.) 1. 19.8 1.6 -18.2 (9, 11.1) 11.9 2.8 -9.1 (10, 18.2) 2. 4.4 32.5 28.1 (8, 15.8) 21.0 8.9 -12.1 (8, 18.9) 3. 14.6 5.4 -9.1 (7, 15.5) 8.3 26.4 18.1 (9, 33.2) 4. 21.1 2.9 -18.3 (8, 23.2) 26.8 7.4 -19.3 (9, 18.3) 5. 28.8 7.5 -21.3 (6, 24) 31.6 7.9 -23.8 (8, 24.5) 6. 9.5 14.2 4.7 (6, 19.7) 19.6 6.3 -13.3 (7, 21.5) 7. 14.8 11.0 -3.8 (4, 27.5) 2.2 19.6 17.3 (9, 13.3) 8. 14.2 15.7 1.5 (6, 32.5) 22.7 2.3 -20.3 (6, 28.0) 9. 23.7 16.8 -6.8 (6, 42.4) 20.1 10.2 -9.9 (9, 32.7) 10. 21.7 28.8 7.2 (6, 49.6) 20.3 19.7 -0.7 (6, 40.5) 11. 30.0 0.5 -29.5 (4, 19.8) 1.3 41.4 40.1 (9, 15.2)

Table 3.3: For group trials, the mean number of occurrences of infected beetles at each type of bait in each of 11 trials as well as the difference between the occurrences with sample sizes and standard deviations in parenthesis.

Effect Mean difference estimate1 Std Error t-value Pr>|t| Sex -2.95 5.51 -0.50 0.626 Males -5.68 5.72 -0.99 0.332 Females -1.55 5.81 -0.27 0.626

Effect Mean difference estimate1 Std Error t-value Pr>|t| Sex -3.07 7.67 -0.39 0.704 Males -5.34 6.01 -0.89 0.385 Females -2.26 6.05 -0.37 0.712 Cysticercoid -0.09 0.49 -0.19 0.852 number 1For males and females, the mean difference equals total number of occurrences at the infective bait minus total number of occurrences at the control bait such that a positive value indicates preference for infective feces.

Table 3.4: Results of the mixed procedure (SAS v.8) of infected beetle preference for infective bait by sex (top), and for infective bait by sex with cysticercoid number (bottom) in the group trials. Estimates of the difference of the mean number of occurrences (infected – uninfected) for males and females are provided. (data from Table 3.3; n♂ = 11, n♀ = 11, d.f. = 20 in all cases except for cysticercoid number which = 137 ).

98 Effect Estimate Std Error d.f. t-value Pr>|t| Sex -13.1 12.46 17 1.05 0.309 Males 8.37 7.59 17 1.10 0.285 Females -4.70 9.88 17 -0.48 0.641

Table 3.5: Results of the mixed procedure (SAS v.8) of uninfected beetle preference for infective bait by sex in the group trials. Estimates of the difference of the mean number of occurrences (infected – uninfected) for males and females are provided.

Arena = 10.5 X 10.5 cm

Cover slip = 2.2 X 2.2 cm

Figure 3.1: Diagram of plastic arena used in preference trial experiment. Dimensions of the arena were 10.5 X 10.5 cm and each cover slip was 2.2 X 2.2 cm. Beetles were held in the middle of the arena in a glass tube and allowed to acclimatize for 7 minutes at which time they were released, and the beetle’s movements were recorded for 15 minutes. Position of infective feces on the cover slips was determined randomly and the arena was rotated 90° after each trial.

100

4 75 mm 2 25 mm square

64 mm dia.

254 mm

Figure 3.2: Diagram of the test arena for groups of beetles. Alternative baits were placed in adjacent bait areas. Position of the control feces was determined randomly (e.g. either areas 1 and 3 or 2 and 4). Each bait area was 25 mm square. The diameter of the center circle was 64 mm and the distance from its center to the edge of a square was 75 mm.

101 12

10 Females Males

6 Frequency 4

0 1-5 6-10 11-15 16-20 21-25 26-30 31-35 36-40 41-45 Number of cysticercoids

Fig. 3.3: Frequency distribution of the total number of cysticercoids dissected from female (empty bars) and male (shaded bars) beetles at the completion of the individual beetle experiment. Males have the greater load of cysticercoids (Fligner-Policello test, n♂ = 27, n♀ = 26, p = 0.0189).

102 Male beetles from individual trials

Secondary infection: 20 Successful infection N o in fe ctio n

10 621 Number of Number beetles

5 20

0 No initial infection Successful initial infection

Female beetles from individual trials

20 Secondary infection: 18 Successful infection 16 No infection

12 10 818 8

6 Number of Number beetles

4 12 2

0 No initial infection Successful initial infection

Figure 3.4: Data from individual trials showing the number of infected and re-infected beetles after two exposures to tapeworm eggs. There was no relationship between the initial and secondary infections for females (2-tailed Fisher exact test, p = 1.0) or for males (p = 0.069) at the 0.05 alpha level.

103 50

Males: y = 7.0x - 2.8 40 R 2 = 0.86 Fem ales: y = 3.9x - 0.15 R 2 = 0.83 30

20 Variancein infection level 10 Females Males

0 01234567 Average number of early stage cysticercoids

100

Males: y = 9.9x - 25.8 80 R 2 = 0.95 Fem ales: y = 3.0x - 2.9 R 2 = 0.74 60

40 Variance in infectionlevel 20 Females Males

0 024681012 Average number of stage 5 cysticercoids

Figure 3.5: Plots of the average number of early stage cysticercoids (top) and stage 5 cysticercoids (bottom) against its variance for both male (empty circles) and female (filled circles) beetles of the group trial. All slopes were significantly different from one (t(.975, 9) > 2.26, p < 0.01), and male slopes were steeper than female slopes (p’s < 0.05).

104 Male beetles from group trials

Secondary infection: 40 Successful infection N o infection

20 48 42 Number of beetles

10 6 9

0 No initial infection Successful initial infection

Female beetles from group trials

40 35 35 Secondary infection: Successful infection 30 N o infection

14 21

Number of beetles of Number 10

0 No initial infection Successful initial infection

Figure 3.6: Data from group trials showing the number of infected and re-infected beetles after two exposures to tapeworm eggs. There was no relationship between initial and secondary infections for females (2-tailed Fisher exact test, p = 0.408) or for males (p = 0.409).

105

CHAPTER 4

GENDER DIFFERENCES IN SUSCEPTIBILITY TO INFECTION IN GROUPS OF UNINFECTED HOSTS

ABSTRACT

Males and females often differ in their susceptibility and exposure to infection.

Thus, they may also differ in their ability to avoid infection. The beetle, Tenebrio

molitor, risks infection with cysticercoids when ingesting rat feces containing eggs of the tapeworm, Hymenolepis diminuta. Previous studies demonstrated that beetles sometimes

prefer infective feces. Based on these results, I hypothesize that the tapeworm will

manipulate the foraging behavior of unisex groups of starved male and female beetles such that they will prefer infective feces. To test this, beetle movement in an arena containing infective and uninfective feces was recorded at one-minute intervals for one hour. Males and females did not differ in their bait preference, beetle movement or in

their group foraging behavior. However, the hypothesis was partially refuted because

more male beetles occurred on the uninfective bait than on the infective bait. This

suggests that male beetles avoid infective feces, which decreases their probability of infection with H. diminuta.

106 INTRODUCTION

In general, male vertebrates tend to show a greater parasite occurrence and load

than females (Bundy 1988; Zuk 1990; Zuk and McKean 1996; Poulin 1996; Schalk and

Forbes 1997), but the same pattern is not observed as commonly in invertebrates

(Sheridan et al. 2000). However, there are some exceptions in which experimentally

infected arthropods resulted in a male infection bias (Pappas et al. 1995; Gray 1998;

Wedekind and Jakobsen 1998). This male infection bias may be linked to differences in infection susceptibility that are immunological, differences in infection exposure that are behavioral, or some combination. For example, female scorpionflies have greater immune function than males (Kurtz et al. 2000; Kurtz and Sauer 2001). Experimental evidence suggests that there is a trade-off between immune function and reproductive activity in male fruitflies (McKean and Nunney 2001) and in male crickets (Adamo et al.

2001). Behavioral differences between the sexes may also lead to an infection bias due to differences in exposure to parasites. For example, sex differences in parasitism in threespine sticklebacks can be traced to dietary differences (Reimchen and Nosil 2001).

Female damselflies are more heavily infected with gregarines because they consume more food (Hecker et al. 2002). Also, the male bias in tapeworm-infected beetles may be due to gender differences in the foraging behavior of beetles (Chapters 1 and 2). Finally, sexual differences in susceptibility and exposure to infection may both operate if the same physiological differences lead to behavioral differences that make one sex more likely to be infected than the other (Klein 2000). If susceptibility and exposure to infection differs by sex, then does avoidance of infection also differ by sex? The beetle- tapeworm system lends itself to the study of this question.

107 Adults of the tapeworm Hymenolepis diminuta, live in the small intestines of

rodents, the definitive host. Eggs pass out of the rodents in their feces. The cysticercoid,

or larval stage, occurs in an insect (the intermediate host, most often a beetle) that

becomes infected by feeding upon infective feces. The rodent completes the cycle when

it ingests an infected beetle. Thus, the success of this parasite depends on rodent-to-

beetle and beetle-to-rodent transmission, and the former depends upon consumption of the egg stage by the beetle. Recent evidence suggests that even this passive type of

transmission involves a mechanism that influences host behavior to benefit the tapeworm.

Three prior studies have tested the fecal preference of groups of beetles by

allowing them to forage in an arena with two types of food, rat feces from uninfected rats

and rat feces from rats infected with H. diminuta (presumed to contain eggs). In one

study, groups of beetles (Tribolium confusum) of mixed sex, and starved 48 hours prior to

the trial preferred infective feces (Evans et al. 1992). In another study, groups of starved

and fed beetles (T. confusum) preferred control bait during two time intervals of a two-

hour trial, but showed no preference during the other four time intervals. The authors

concluded that beetle behavior appeared “highly heterogeneous, both among individuals

and by the same individual over time” (Shostak and Smyth 1998).

In contrast to the first two studies, a third study examined unisex groups of beetles

(Tenebrio molitor), which allowed for gender comparisons while eliminating potentially

confounding interactions between the sexes during the preference trial. This study found

that females, whether fed or starved 72 hours prior to the trial, preferred infective feces

whereas starved male beetles preferred infective feces, and fed males preferred

108 uninfective feces (Pappas et al. 1995). These three experiments demonstrate that sex and

feeding history may alter beetle’s preference for infective or uninfective feces.

To assess the relative importance of sex, this study tests the fecal preference of groups of starved beetles of known sex and age using the same experimental protocol of

Pappas et al. 1995, but with a different statistical analysis. Based on the results of their experiment (Pappas et al. 1995), I hypothesize that groups of male and female beetles will prefer infective feces. To test this, beetle movement in an arena containing infective and uninfective feces was recorded at one-minute intervals for one hour.

METHODS

Study subjects

The “OSU Strain” (Pappas and Leiby 1986) of Hymenolepis diminuta was maintained in male Sprague-Dawley rats and beetles (T. molitor). Because this is a highly inbred strain of tapeworm (having survived a severe bottleneck in 1995), I assume it has low genetic variability, which should minimize concern about parasite genetic variation confounding the results of host behavior (but see Meffert 1999). Three male

Sprague-Dawley rats were infected with 30 cysticercoids and maintained on commercial rodent chow and water. Three additional rats, obtained from the same commercial source, of identical age and from the same litter so as to minimize rat genetic variability, were maintained under the same conditions as the infected rats to serve as the source of control (uninfective) feces. Fecal pellets from infected rats were not used prior to 20 days post-infection. On the morning of the trials, rat cages were checked every 10 minutes, and fecal pellets were collected with forceps to minimize contamination by rat

109 urine. After each trial, the infective fecal pellets were examined to verify the presence of

H. diminuta eggs.

Beetles were maintained on wheat bran, and small pieces of potato were added to

the cultures on a regular basis. Pupae were removed from the cultures, and male and

female pupae (Bhattacharya et al. 1970) were placed in separate dishes containing wheat bran. Beetles that emerged during a 24 h period were collected such that a daily cohort of

beetles was maintained for both sexes.

Experimental Protocol

Before the trial (7-10 days), male and female beetles were placed in unisex groups of ten. All beetles were starved two-three days before their respective trials. Twelve

groups of each sex were used in the preference trial in an alternating sequence. Each

group was placed under a glass bowl in the center of a plastic arena (that was bleached

twice and rinsed in hot water for at least 30 minutes before each trial) under red light

conditions. Beetles were provided with two types of bait: uninfective (control) or

infective feces both of which had been collected within 10 minutes of defecation.

Control feces were positioned randomly either in areas 1 and 3 or 2 and 4 (Fig. 4.1).

Infective feces were then positioned in the other areas. A paper towel was used to mash

the fecal pellets on the squares to prevent their displacement by beetles during the trial.

After 15 minutes, the bowl was removed and the beetles’ movements were videotaped for one hour under red light conditions. The video was played back and the total number of

beetles observed at each bait area and side was recorded at one-minute intervals.

110 Data Analysis

The data analysis does not set the null hypothesis equal to 50%. This would

assume a binomial distribution where the presence of the beetle in the next moment (on

the infective or uninfective side) is independent of current location. However, this

assumption is not met because the beetle’s location in the next moment of time depends

on its current location. This autocorrelation introduces greater than normal binomial

variance. As evidence, the observed standard deviation for the average proportion of

time spent on the infected side equals 0.42 for the individual trials and 0.28 for the group

trials, while the expected deviation (based on a binomial distribution) equals 0.09 if k =

30, 0.16 if k = 10, and 0.22 if k = 5 where k equals the number of hypothetical opportunities to make discrete, binary choices during the trial. Thus, large deviations make it difficult to demonstrate small preferences.

Preference was measured as the difference between the total number of beetles at infective and control baits. The total number of beetles on the infective or control bait was subtracted from the total number of beetles in the infective or control side to equal the number of beetles in the infective or control non-bait area. The total number of beetles at the bait areas and in the non-bait areas were analyzed with the Mixed procedure in SAS (v.8, Cary, NC) with sex and trial as class variables and sex nested in trial as a

random variable. The response variable was defined as (number of beetles at control

bait/non-bait) – (number of beetles at infected bait/non-bait).

The maximum proportion of female and male beetles occurring at the bait areas

for each 60-min trial was plotted against the average proportion of beetles at the bait

areas (after an arcsine square root transformation for both proportions), and the slopes

111 were compared. The same was done for the maximum proportion of female and male

beetles occurring in the non-bait areas.

RESULTS

The total number of beetles did not differ between males and females in bait (p =

0.159) or non-bait regions (Table 4.1; Mixed procedure, p = 0.873). However, more

males occurred at the uninfective bait than at the infective bait (1.58, p = 0.008), while

females did not differ in the number of occurrences at each type of bait (0.46, p = 0.412).

The total number of beetles in the non-bait regions did not differ between infected and

control halves of the arena for males (Table 4.1; 0.30, p = 0.292) or for females (0.24, p =

0.403).

In a plot of the maximum number of beetles at each bait against the average

number of beetles at each bait (data from Table 4.2), the slopes did not differ between

males and females at the infective (F1, 20 = 0.045, p = 0.834) or the control baits (F1, 20 =

2.75, p = 0.113). In a plot of the maximum number of beetles in each non-bait area against the average number of beetles in each non-bait area (data from Table 4.3), the slopes did not differ between males and females at the infective (F1, 20 = 0.211, p = 0.651)

or the control non-bait areas (F1, 20 = 0.748, p = 0.397).

DISCUSSION

These data indicate that male beetles avoid infective feces while females show no

preference. Beetles that avoid infective feces would be able to avoid the various costs

associated with infection and these costs may be sex-specific. Infected male T. molitor

112 exhibit a lowered response to pheromone produced by uninfected females (Hurd and

Parry 1991), while fat body glycogen reserves are significantly depleted three days post- infection in male and five days post-infection in female T. molitor (Kearns et al. 1994).

Male Tribolium castaneum infected with H. diminuta sire 14-22% fewer offspring than uninfected males (Pai and Yan 2003). Two experiments examined the consequences of infection on fertility. In the first, males mated singly with infected females produced an average (n = 30) of 115.2 ovulated eggs, while males mated with uninfected females produced an average (n = 30) of 127.9 ovulated eggs (Hurd and Arme 1986). In the second, females mated singly with infected males produce an average (n = 7) of 37 larvae while females mated with uninfected males produce an average (n = 13) of 63.7 larvae

(Worden et al. 2000). Further, tapeworm-infected males (T. molitor) are less attractive to females (Worden et al. 2000). Thus, infection in males not only results in less reproductive success, but infected males are less likely to mate with a female. These costs to infection suggest that selection to avoid infective feces may act stronger on males.

Males consume more food than females (Chapter 1), which increases their exposure to parasites and thus their probability of infection. Given the previously described costs of infection to males and their greater exposure to infection, avoiding infective feces would be adaptive for a male host. This may be especially true if males are more susceptible to infection as some experimental infections indicate (Chapters 1, 3,

Pappas et al. 1995). Females, on the other hand, do not differ in their foraging behavior, suggesting that they are unable to discriminate between the bait types when in groups.

When tested individually, females preferred infective bait (Chapter 2). Although females

113 share some of the same consequences of infection as males, infection results in greater

reproductive costs to males, which may explain the preference differences observed between males and females.

Host movement is often restricted in the presence of parasites. For example, reindeer avoid pastures with high densities of dung suggesting a behavioral adaptation that reduces their risk of infection by Trichostrongyle nematode eggs (Van der Wal et al.

2000). Also, grazing sheep will avoid grazing on patches contaminated with feces

(Hutchings et al. 1998). In my study, male and female beetles do not differ in the number of occurrences in the non-bait regions suggesting that beetle movement is not restricted by the presence of feces or tapeworm eggs. This lends support to the contention that male beetles avoid infective feces rather than feces in general.

Grain beetles produce sex and aggregation pheromones (Burkholder and Ma

1992, August 1971, Tschinkel et al. 1967), which may influence the behavior of beetles foraging in groups. Male and females do not differ in the maximum number of occurrences at either the infective or control baits, indicating that both sexes behave similarly when foraging in unisex groups. Although this minimizes concern about the confounding influence of social interactions, it does not represent conditions found in the wild (Rau 1979). Studies with T. confusum found that beetles in mixed-sex groups preferred infective feces (Evans et al. 1992). Starved beetles tested individually differed in some respects to beetles tested in groups (Shostak and Smyth 1998), but this may be an artifact of the how preference was measured and analyzed. Starved females tested individually preferred infective feces while males showed no preference (Chapter 2).

114 When contrasted with the results of this study, it suggests that social interactions do

influence beetle fecal preference.

Statistical analysis of these trials did not assume that the beetle’s movements were

independent from one moment to the next. In fact, distributions for the proportion of time spent on the infective side deviated from the expected binomial distribution. Thus, a null hypothesis based upon random chance or the area of the arena is not valid. Instead, my analysis compares time spent between infective and control baits as well as between male and female beetles.

Overall, this study demonstrates that starved males, when tested in groups, avoid infective feces. There would be strong selection for such a host adaptation if the costs of infection were high. Females, on the other hand, failed to discriminate between infective and uninfective feces. Selection pressure on the host to avoid infective feces may be balanced by selection pressure on the parasite to manipulate the host to prefer infective feces. Future studies need to address the influence of feeding status and the presence of conspecifics on beetle preference for infective bait.

115 Effect Mean difference estimate1 Std Error t-value Pr>|t| Sex -1.12 0.77 -1.46 0.159 Males 1.58 0.54 2.90 0.008 Females 0.46 0.54 0.84 0.412

Effect Mean difference estimate1 Std Error t-value Pr>|t| Sex -0.06 0.40 -0.16 0.873 Males 0.30 0.28 1.08 0.292 Females 0.24 0.28 0.85 0.403

1For males and females, the mean difference equals total number of beetles at control bait/non-bait area minus total number of beetles at infected bait/non-bait area such that a positive value indicates preference for uninfective feces.

Table 4.1: Results of the mixed procedure (SAS v.8) for beetle preference by sex for the bait area (top), and for the non-bait (beetles on side – beetles on bait) area (bottom) in group trials (n♂ = 12, n♀ = 12, d.f. = 22 in all cases). Data from Tables 4.2 and 4.3.

116 Males Females Trial Infected Control Infected Control 1. 34 (0-3) 299 (0-8) 26 (0-4) 22 (0-1) 2. 3 (0-1) 0 (0-0) 54 (0-3) 12 (0-2) 3. 36 (0-3) 243 (0-7) 22 (0-3) 45 (0-3) 4. 33 (0-3) 198 (0-7) 24 (0-3) 161 (0-8) 5. 73 (0-7) 173 (0-7) 36 (0-2) 71 (0-5) 6. 42 (0-4) 245 (0-8) 18 (0-2) 38 (0-2) 7. 192 (0-7) 34 (0-2) 9 (0-1) 90 (0-5) 8. 10 (0-2) 276 (0-8) 43 (0-3) 272 (0-9) 9. 34 (0-3) 1 (0-1) 202 (0-6) 71 (0-4) 10. 2 (0-1) 8 (0-6) 19 (0-2) 37 (0-3) 11. 141 (0-5) 182 (0-6) 20 (0-2) 36 (0-2) 12. 27 (0-2) 104 (0-6) 69 (0-8) 15 (0-2) Mean 52.3 (16.7) 146.9 (32.6) 45.2 (16.7) 72.5 (21.7)

Table 4.2: Total number of beetles at each bait type (infected or control) summed over the entire 60 minute trial (range in parenthesis) with mean number of beetles (standard errors in parenthesis).

117 Males Females Trail Infected Control Infected Control 1. 150 (0-8) 117 (0-5) 254 (0-10) 298 (0-9) 2. 238 (1-10) 359 (0-9) 265 (1-8) 269 (1-8) 3. 179 (0-7) 142 (0-6) 244 (0-7) 289 (2-9) 4. 163 (0-6) 206 (0-7) 208 (0-8) 207 (0-7) 5. 164 (0-8) 190 (0-7) 215 (0-8) 278 (1-8) 6. 158 (0-7) 155 (0-5) 301 (0-10) 243 (0-9) 7. 208 (0-8) 166 (0-9) 204 (0-7) 297 (2-9) 8. 136 (0-6) 178 (0-8) 124 (0-5) 161 (0-8) 9. 206 (1-6) 359 (4-8) 160 (0-7) 167 (0-6) 10. 324 (1-9) 266 (1-9) 315 (3-8) 229 (0-7) 11. 153 (0-7) 126 (0-6) 265 (0-9) 279 (0-10) 12. 217 (0-7) 252 (0-9) 253 (1-8) 263 (1-9) Mean 191.3 (15.1) 209.6 (24) 234 (16) 248.3 (13.9)

Table 4.3: Total number of beetles in the non-bait area for each half of the arena (infected or control) summed over the entire 60 minute trial (range in parenthesis) with mean number of beetles (standard errors in parenthesis).

118

4 75 mm 2 25 mm square

64 mm dia.

254 mm

Figure 4.1: Diagram of the test arena for groups of beetles. Alternative baits were placed in adjacent bait areas. Position of the control feces was determined randomly (e.g. either areas 1 and 3 or 2 and 4). Each bait area was 25 mm square. The diameter of the center circle was 64 mm and the distance from its center to the edge of a square was 75 mm.

119

INTEGRATIVE DISCUSSION

Altered host behavior can be a consequence of a host adaptation that benefits the

host. It may also be a by-product of infection. Finally, host altered behavior may result

from a parasite manipulation that benefits the parasite (Poulin 1994; Moore 2002).

Typically, the parasite manipulates the host to facilitate its transmission to the next host.

An increase in the probability of transmission or a parasite can contribute to the increase

of that parasite’s virulence (Ewald 1983). Thus, studying the factors that influence

parasite transmission has practical application in the study of disease. Further, many of

these factors differ in males and females because the costs to parasitism differ between

genders.

When considering the factors that influence the transmission of the tapeworm H. diminuta to its beetle intermediate host, several observations and conclusions can be made. From Chapter 1, host resources do not limit parasite growth and development in males and females and so, under some conditions, host re-infection could benefit the parasite. However, the results suggest that infection has no effect on feeding activity, which will not alter the probability of re-infection. Instead, male beetles feed more than females. This male feeding bias may benefit males if well fed males gain some benefit over starved males. For example, well fed males may acquire more mating opportunities than starved males. This suggests an optimal trade-off between risking re-infection and being well fed. Also, the male feeding bias suggests an explanation behind the higher median load of parasites recovered from male beetles (Chapter 1).

120 A higher median load of parasites in males is also observed in the experiment

reported in Chapter 3. The increased feeding activity of males may partially explain this

infection bias. However, in Chapter 2, females prefer infective feces over uninfected feces while males show lower activity levels when allowed to feed. The lower male activity level will result in a differential exposure to tapeworm eggs across all males resulting in a more clumped distribution of parasites with many males being heavily

infected and many males being uninfected. This clumped distribution is interesting

because it implies that some individual males are much more (or much less) susceptible

to infection than others, while females vary less in their susceptibility to infection. This

is different from simply asking if males are more susceptible to infection than females.

In addition, a clumped distribution deviates from a normal distribution such that a proper

comparison of male and female parasite loads requires a non-parametric test that does not

assume equal variances (see Fligner-Policello test in Hollander and Wolfe 1999). Thus,

the commonly used measure of parasitism, intensity (the mean number of parasites per

host), fails to accurately describe highly variable distribution observed in male infections.

Given that natural selection acts on variation, studies of infection bias (see meta-analyses

in Poulin 1996 and Sheridan et al. 2000) should consider a measure of variation.

The experimental design in Chapter 3 allows for a measure of initial and

secondary infection variance and its analysis suggests that some beetles are highly

susceptible to infection while others are highly resistant. Since the beetles were

experimentally infected in the initial infection, they all had equal exposure to the

tapeworm eggs. The secondary infection could be explained by the activity levels of the

beetles since infection depends upon the foraging behavior of the beetles in an arena.

121 However, this explanation is not supported statistically. Further, the variation in the infection distribution is statistically greater in males than in females. Combined, these results suggest variation in susceptibility to infection within sexes as well as greater susceptibility to infection in males. If males vary in susceptibility to infection (Yan et al.

1994), then natural selection can favor increased feeding activity in those males that are

highly resistant to parasites. However, results from Chapter 1 do not support this because

increased feeding on the day of infection resulted in greater numbers of cysticercoids in

males (r = 0.373, p = 0.055). Finally, the larger variation observed in males may play an

important role in sexual selection in these beetles.

From Chapter 3, when allowed to feed individually or in groups after being

infected, neither male nor female beetles exhibit a preference for infective bait.

Preference for infective bait increases the probability of rodent-to-beetle transmission.

These results suggest that beetles, once infected, lose their preference for infective feces.

This lack of preference benefits the host if increased loads of parasites reduce host

fitness. Or, the lack of preference benefits the parasite if increased loads of parasites

decrease parasite fitness due to competition for limited resources or host death. Chapter

1 results suggest that, under the infection intensities used, host resources do not limit

parasite development. Alternatively, the lack of preference benefits host and parasite if

both are interested in avoiding re-infection.

Beetles do not discriminate between control feces mixed with tapeworm eggs and

control feces when allowed to feed (Chapter 2). However, female beetles prefer infective

feces when allowed to feed in another experiment (Chapter 2). Combined, the results

suggest that the beetle attractant is found in infective feces instead of tapeworm eggs.

122 The adult tapeworm may secrete the volatile attractant observed by Evans et al. (1998) on

infective feces before the rat defecates. When control feces is mixed with ground up

tapeworm tissue to create an artificial bait, fed and starved beetles did discriminate

between control and infective artificial bait (Shostak and Smyth 1998), which partially

supports the contention that the attractant is associated with the tapeworm. However, their experiment did not separate the tapeworm tissue from the eggs, leaving the question of the attractant’s origin unanswered. If this attractant increases the likelihood of egg-to- beetle transmission and is released by the adult tapeworm, then this is an interesting case of parental care by a tapeworm. Unlike the Evans et al. (1998) experiment, however, there was no evidence that this attractant is volatile. Female beetles discriminate between infective and uninfective baits only when allowed to feed. This contrasts with the results of Evans et al. (1998), who found that Tribolium confusum responded to a volatile

attractant isolated from infective feces. Thus, T. molitor may require gustatory cues to

discriminate between fecal types.

When allowed to feed on ample sizes of bait, groups of uninfected male beetles

avoid infective feces. This avoidance behavior decreases the probability of infection.

Females show no preference for infective or uninfective feces when allowed to feed in

groups. Males and females may differ in their parasite avoidance behavior because the

costs to parasitism differ between genders. For example, the results from two studies suggest that infected males suffer a greater reduction in fertility than females (Worden et al. 2000; Hurd and Arme 1986). Further, compared to females, males are more likely to be exposed to infection because they consume more food (Chapter 1) and are more

123 susceptible to infection once exposed (Chapters 1 and 3). Combined, these observations

suggest that selection to avoid infective feces may act stronger on males.

Finally, the behavior of beetles foraging in groups may be influenced by sex and

aggregation pheromones produced by grain beetles such as T. molitor (Burkholder and

Ma 1992, August 1971, Tschinkel et al. 1967). On the other hand, grain beetles are

normally observed to occur in groups in the wild (Rau 1979). In Chapter 4, male and

female beetles do not differ in the number of occurrences in the non-bait regions

suggesting that beetle movement is not restricted by the presence of feces or tapeworm eggs for both sexes. Further, male and females do not differ in the maximum number of

occurrences at either the infective or control baits again indicating that both sexes behave

similarly when foraging in unisex groups. When the fecal preference of groups of beetles

is compared to individual beetles, there are differences. Starved females tested individually prefer infective feces while males show no preference (Chapter 2). Starved females tested in groups show no preference while males prefer uninfective feces

(Chapter 4). Combined, these results suggest that social context does influence beetle fecal preference. Tribolium confusum, on the other hand, show similar behavior when tested individually or in groups suggesting that their behavior was independent of beetle density (Shostak and Smyth 1998). This contradiction may result from differences in the two species’ social structure and size, or from differences in the studies’ measure and analysis of preference, or some combination. For example, if a social hierarchy exists, then beetles may be avoiding one or two individuals with high social status instead of

“preferring” a particular type of bait.

124 These studies raise several questions for further research. First, a more

comprehensive study can address the various factors involved in male infection bias.

Such factors as beetle activity level, humidity, infection status, and infection medium (rat

feces or apple scrapings) need more study. Second, an experimental design, that recognizes the problem of greater than binomial variance in preferences tests should be

used to test fecal preference. Third, experiments that compare the behavior of beetles

infected with immature and mature (infective) cysticercoids are required to better

understand the how the host-parasite relationship influences the probability of

transmission. Fourth, a more comprehensive experiment involving the simultaneous use of starved and fed, male and female, and groups and individual beetles can show how these various factors interact to influence beetle fecal preference.

Finally, an understanding of the possible tradeoffs between risking infection and being well-fed is necessary. Male beetles suffer physiological costs (Hurd and Parry

1991; Kearns et al. 1994) and greater reproductive costs than females when infected

(Hurd and Arme 1986; Worden et al. 2000). Thus, selection to avoid infection should act stronger on males than on females. Yet, males feed more than females which increases their risk of infection (Chapter 1). On the other hand, males also avoid infective feces

(Chapter 4). Thus, males may balance the risk of infection against the cost of starvation.

Future experiments could measure the number of offspring males sire under various conditions. It would be necessary to compare the mating success of infected fed, uninfected fed, infected starved, and uninfected starved males to understand the adaptive significance of the male feeding bias and how it relates to the male infection bias.

125

Appendix A. Description of beetles used in the starved experiment.

For each trial, the sex (F = female or M = male), the treatment (I = infected or C = control), the number of beetles used at the start of the trial and the number of beetles that died is given. Beetles were dropped from the analysis due to experimental error or if used in another experiment. Beetles exposed to tapeworm eggs, but uninfected were also not analyzed. Both cases are reported. Thus, from 208 beetles, 201 were included in the starved experiment and after 35% mortality, a total of 131 live beetles were analyzed. From 293 beetles, 275 were included in the fed experiment and after a 54% mortality, 126 live beetles were analyzed. Trials were conducted at random and no time order should be inferred from the numbering scheme. ______Trial Sex Status Attempted Died Dropped Uninfected Analyzed 1 F C 13 0 1 0 12 2 F C 13 2 0 0 11 3 F C 13 0 0 0 13 4 F I 13 2 0 0 11 5 F I 13 3 1 0 9 6 F I 13 4 0 0 9 7* F I 5 1 0 1 3 8 M C 13 3 0 0 10 9 M C 13 7 0 0 6 10 M C 13 7 0 0 6 11 M C 13 6 0 0 7 12* M C 8 2 3 0 3 13 M I 13 5 0 0 8 14 M I 13 12 0 0 1 15 M I 13 8 0 0 5 16 M I 13 3 0 0 10 17 M I 13 5 1 0 7 ______Totals 70 6 1 131 208

* 5 beetles of each sex were randomly removed from two unisex cohorts of 13 beetles and were used to form a female infected (7) and male control (12) trial.

126

Description of beetles used in the fed experiment.

Trial Sex Status Attempted Died Dropped Uninfected Analyzed 1 F C 13 11 0 - 2 2 F C 13 2 0 - 11 3 F C 13 5 0 - 8 4 F C 13 4 1 - 8 5 F I 13 9 0 1 3 6 F I 13 7 0 0 6 7 F I 13 10 0 0 3 8 F I 13 1 0 0 12 9 F I 13 6 0 0 7 10 M C 13 10 0 - 3 11 M C 13 8 0 - 5 12 M C 13 2 0 - 11 13 M C 13 8 0 - 5 14 M C 13 5 0 - 8 15 M I 13 11 1 0 1 16 M I 13 7 0 0 6 17 M I 13 6 0 0 7 18 M I 13 7 0 1 5 19 M I 13 4 0 2 7 20 M I 13 5 7 0 1 Totals 128 9 4 119 260

127

BIBLIOGRAPHY

Adamo, S. A. 1998. The specificity of behavioral fever in the cricket Acheta domesticus. Journal of Parasitology 84: 529-533.

Adamo, S. A., Jensen, M. and Younger, M. 2001. Changes in lifetime immunocompetence in male and female Gryllus texensis (formerly G. integer): trade-offs between immunity and reproduction. Animal Behaviour 62: 417-425.

Anokhin, I. A. 1966. Daily rhythm in ants infected with metacercariae of Dicrocoelium lanceatum. Doklady Akad. Nauk SSSR, Biological Sciences. 166: 93-94.

Arnott, S. A., Barber, I. and Huntingford, F. A. 2000. Parasite-associated growth enhancement in a fish-cestode system. Proceedings of the Royal Society of London, Series B 267: 657-663.

August, C. J. 1971. The role of male and female pheromones in the mating behavior of Tenebrio molitor. Journal of Insect Physiology 17: 739-751.

Benjamini, Y., and Hochberg, Y. 1994. Controlling the False Discovery Rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society, Series B, 57 (1): 289-300.

Berntzen, A. K. and Voge, M. 1962. In vitro hatching of oncospheres of Hymenolepis nana and Hymenolepis citelli (Cestoda: Cyclophyllidea). Journal of Parasitology 48: 110-119.

Bethel, W. M. and Holmes, J. C. 1973. Altered evasive behaviour and response to light in amphipods harbouring acanthocephalan cystacanths. Journal of Parasitology 59: 945-956.

Bhattacharya, A. K., Ameel, J. J. and Waldebauer, G. P. 1970. A method for sexing living pupal and adult yellow mealworms. Annals of the Entomological Society of America 63: 1783.

Blankespoor, C. L., P. W. Pappas and T. Eisner. 1997. Impairment of the chemical defence of the beetle, Tenebrio molitor, by metacestodes (cysticercoids) of the tapeworm, Hymenolepis diminuta. Parasitology 115: 105-110.

128 Boorstein, S. M. and Ewald, P. W. 1987. Costs and benefits of behavioral fever in Melanoplus sanguinipes infected by Nosema acridophagus. Physiological Zoology 60: 586-595.

Brendelberger, H. 1997. Coprophagy: a supplementary food source for two freshwater gastropods? Freshwater Biology 38: 145-157.

Bundy, D. A. P. 1988. Gender-dependent patterns of infection and disease. Parasitology Today 4: 186-189.

Burkholder, W. E. and Ma, M. 1985. Pheromone for monitoring and control of stored product insects. Annual Review of Entomology. 30: 257-272.

Buxton, P. A. 1930. Evaporation from the meal-worm (Tenebrio: Coleoptera) and atmospheric humidity. Proceedings of the Royal Society of London, Series B 106: 560-577.

Calhoun, J. B. 1963. The ecology and sociology of the Norway rat. U.S. Department of Health, Education, and Welfare. Public Health Service Publication 1008: 1-288.

Caraco, T. and Giraldeau, L. A. 1993. Genetic relatedness and group-size in an aggregation economy. Evolutionary Ecology 7 (4): 429-438.

Carney, W. P. 1969. Behavioral and morphological changes in carpenter ants harboring decrocoeliid metacercariae. American Midland Naturalist 82: 605-611.

Chapman, R. F. 1998. The Insects: Structure and Function. Cambridge: University Press.

Cruz, Y. P. 1981. A sterile defender morph in a polyembryonic Hymenopterous parasite. Nature 294: 446-447.

Cruz, Y.P., Oelhaf, R. C. and Jockusch, E.L. 1990. Polymorphic precocious larvae in the polyembryonic parasitoid Copidosomopsis-tanytmema (Hymenoptera, Encyrtidae). Annals of the Entmological Society of America 83(3): 549-554.

Davis, D. E., Emlen, J.T. and Stokes, A. W. 1948. Studies on home range in the brown rat. Journal of Mammalogy 29(3): 207-225.

Downer, R. G. H. 1981. Energy Metabolism in Insects. New York: Plenum Press.

Edwards, G. A. 1946. The influence of temperature upon the oxygen consumption of several arthropods. Journal of Cellular and Comparative Physiology 27: 53-64.

Emlen, J. T., Stokes, A. W., and Winsor, C. P. 1948. The rate of recovery of decimated populations of brown rats in nature. Ecology 29 (2): 133-145.

129

Evans, W.S., Hardy, M.C., Singh, R., Moodie, G.E., and Cote, J.J. 1992. Effect of the rat tapeworm, Hymenolepis diminuta, on the coprophagic activity of its intermediate host, Tribolium confusum. Canadian Journal of Zoology 70: 2311-2314.

Evans, W.S., Wong, A., Hardy, M., Currie, R.W., and Vanderwel, D. 1998. Evidence that the factor used by the tapeworm, Hymenolepis diminuta, to direct the foraging behavior of its intermediate host, Tribolium confusum, is a volatile attractant. Journal of Parasitology 84(6): 1098-1101.

Ewald, P. W. 1983. Host-parasite relations, vectors, and the evolution of disease severity. Annual Review of Ecology and Systematics 14: 465-485.

Gandhi, R. S. and Goswami, S. L. 1993. Comparative growth in Tribolium castaneum and Tribolium confusum. Journal of Dairying, Foods and Home Science 12: 191-194.

Granath, W. O. 1980. Hymenolepis diminuta (Cestoda): effects of parasite density and temperature on the water balance of Tenebrio molitor and subsequent infectivity of the cysticercoids. Journal of Invertebrate Pathology 36: 235-239.

GraphPad Quick Calc calculator. 2003. San Diego, CA. http://graphpad.com/quickcalcs/contingency1.cfm

Gray, D. A. 1998. Sex differences in susceptibility of house crickets, Acheta domesticus, to experimental infection with Serratia liquefaciens. Journal of Invertebrate Pathology 71: 288-289.

Hart, B.L. 1997. Effects of hormones on behavioral defenses against parasites. In Parasites and pathogens: effects on host hormones and behavior (ed. N. E. Beckage). Pp 210-230. Chapman and Hall: New York.

Hart, B.L. and Hart, L. A. 1994. Fly switching by Asian elephants: tool use to control parasites. Animal Behaviour 48: 35-45.

Hecker, K. R., Forbes, M. R. and Leonard, N. J. 2002. Parasitism of damselflies (Enallagma boreale) by gregarines: sex biases and relations to adult survivorship. Canadian Journal of Zoology 80: 162-168.

Hervant, F., Mathieu, J., and Barré, H. 1999. Comparative study on the metabolic responses of subterranean and surface-dwelling amphipods to long-term starvation and subsequent refeeding. Journal of Experimental Biology 202: 3587-3595.

Hollander, M. and Wolfe, D. A. 1999. Nonparametric Statistical Methods. 2nd ed. New York: John Wiley & Sons, Inc.

130

Holmes, J. C. and Bethel, W. M. 1972. Modification of intermediate host behavior by parasites. In Behavioral Aspects of Parasite Transmission (eds. E. U. Canning and C. A. Wright). pp. 123-49. Academic Press, New York.

Horton, D. H. 1995. Statistical considerations in the design and analysis of paired- choice assays. Entomological Society of America 24: 179-192.

Hung, S. Y. and Boucias, D. G. 1992. Influence of Beauveria bassiana on the cellular defence response of the beet armyworm, Spodoptera exigua. Journal of Invertebrate Pathology 60: 152-158.

Hurd, H. 1998. Parasite manipulation of insect reproduction: who benefits? Parasitology 116: S13-S21.

Hurd, H. 2001. Host fecundity reduction: a strategy for damage limitation? TRENDS in Parasitology 117: 363-368.

Hurd, H. and Arme, C. 1984. Tenebrio molitor (Coleoptera): effect of metacestodes of Hymenolepis diminuta (Cestoda) on haemolymph amino acids. Parasitology 89: 245-251.

Hurd, H. and Arme, C. 1986. Hymenolepis diminuta: The effects of metacestodes upon egg production and viability in the intermediate host Tenebrio molitor. Journal of Invertebrate Pathology 47: 227-234.

Hurd, H. and Arme, C. 1987. Hymenolepis diminuta (Cestoda): the role of intermediate host sex in the establishment, growth and development of metacestodes in Tenebrio molitor (Coleoptera). Helminthologia 24: 23-31.

Hurd, H. and Fogo, S. 1991. Changes induced by Hymenolepis diminuta in the behavior of the intermediate host, Tenebrio molitor. Canadian Journal of Zoology 69: 2291-2294.

Hurd, H. and Parry, G. 1991. Metacestode-induced depression of the production of, and response to, sex pheromone in the intermediate host Tenebrio molitor. Journal of Invertebrate Pathology 58: 82-87.

Hurd, H., Warr, E. and Polwart, A. 2001. A parasite that increases host lifespan. Proceedings of the Royal Society of London: Biological Series 268: 1749-1753.

Hurd, H. and Webb, T. 1997. The role of endocrinological versus nutritional influences in mediating reproductive changes in insect hosts and insect vectors. In Parasites and pathogens: effects on host hormones and behavior (N. E. Beckage, ed.), pp. 179-197. New York: Chapman & Hall.

131 Hutchings, M. R., I. Kryriazakis, D. H. Anderson, I. J. Gordon and Coop, R. L. 1998. Behavioural strategies used by parasitized and non-parasitized sheep to avoid ingestion of gastro-intestinal nematodes associated with faeces. Animal Science 67: 97-106.

Jarecka, L. 1961. Morphological adaptations of tapeworm eggs and their importance in the life cycles. Acta Parasitologica Polonica 9: 409-426.

Johnson, D. L. and Pavlikova, E. 1986. Reduction of consumption by grasshoppers (Orthoptera: Acrididae) infected with Nosema locustae Canning (Microsporidia: Nosematidae). Journal of Invertebrate Pathology 48: 232-238.

Kearns, J., Hurd, H., and Pullin, A. S. 1994. The effect of metacestodes of Hymenolepis diminuta, on storage and circulating carbohydrates in the intermediate host, Tenebrio molitor. Parasitology 108: 473-478.

Keymer, A. E. 1980. The influence of Hymenolepis diminuta on the survival and fecundity of the intermediate host, Tribolium confusum. Parasitology 81: 405- 421.

Keymer, A.E. 1981. Population dynamics of Hymenolepis diminuta in the intermediate host. Journal of Animal Ecology 50: 941-950.

Klein, S. L. 2000. The effects of hormones on sex differences in infection: from genes to behavior. Neuroscience and Biobehavioral Reviews 24: 627-638.

Kopanic, R. J., Holbrook, G. L., Sevala, V. and Schal, C. 2001. An adaptive benefit of facultative coprophagy in the German cockroach Blattella germanica. Ecological Entomology 26: 154-162

Kurtz, J, Wiesner, A., Götz, P., Sauer, K. P. 2000. Gender differences and individual variation in the immune system of the scorpionfly Panorpa vulgaris (Insecta: Mecoptera). Developmental and Comparative Immunology 24: 1-12.

Kurtz, J. and Sauer, K. P. 2001. Gender differences in phenoloxidase activity of Panorpa vulgaris hemocytes. Journal of Invertebrate Pathology 78: 53-55.

Lackie, A.M. 1976. Evasion of the haemocytic defence reaction of certain insects by larvae of Hymenolepis diminuta (Cestoda). Parasitology 73: 97-107.

Lethbridge, R. C. 1971a. The hatching of Hymenolepis diminuta on the survival and fecundity of the intermediate host Tribolium confusum. Parasitology 64: 405- 412.

132 Lethbridge, R. C. 1971b. An improved gradient centrifugation technique for the separation of Hymenolepis diminuta eggs from rat feces. Journal of Parasitology 57: 1140-1142.

MacArthur, J. W. and Baillie, W. H. T. 1926. Sex differences in mortality and metabolic activity in Daphnia magna. Science 64: 229-230.

Mackiewicz, J. S. 1988. Cestode transmission patterns. Journal of Parasitology 74(1): 60-71.

Major, M., Webb, T. J. and Hurd, H. 1997. Haemolymph from female beetles infected with Hymenolepis diminuta metacestodes retards the development of ovarian follicles in recipient Tenebrio molitor (Coleoptera). Parasitology 114: 175-181.

McCurdy, D. G., Forbes, M. R., & Boates, J. S. 1999. Testing alternative hypotheses for variation in amphipod behavior and life history in relation to parasitism. International Journal for Parasitology. 29: 1001-1009.

Manly, B. F. J. 1993. Comments on design and analysis of multiple-choice feeding- preference experiments. Oecologia 93: 149-152.

McKean, K. A. and Nunney, L. 2001. Increased sexual activity reduces male immune function in Drosophila melanogaster. Proceedings of the National Academy of the Sciences of the United States of America 98: 7904-7909.

Meffert, L. M. 1999. The assumptions of captive breeding strategies – that founder events reduce genetic variation – may not always be correct. BioScience 49: 701-711.

Minitab v.13.1 2000. State College, PA.

Moore, D., Reed, M., LePatourel, G., Abraham, Y. J. and Prior, C. 1992. Reduction of feeding by the desert locust Schistocerca gregaria, after infection with Metarhizium flavoviride. Journal of Invertebrate Pathology 60: 304-307.

Moore, J. 1983. Altered behavior in cockroaches (Periplaneta americana) infected with an Archiacanthocephalan, Moniliformis moniliformis. Journal of Parasitology. 69: 1174-1176.

Moore, J. 2002. Parasites and the Behavior of Animals. Oxford: University Press.

Moore, J. and Bell, D. H. 1983. Pathology (?) of Plagiorhynchus cylindraceus in the starling, Sturnus vulgaris. Journal of Parasitology 69: 387-390.

133 Norris, K. 1999. A trade-off between energy intake and exposure to parasites in oystercatchers feeding on a bivalve mollusk. Proceedings of the Royal Society of London. 1429: 1703-1709.

Pai, A and Yan, G. 2003. Effects of tapeworm infection on male reproductive success and mating vigor in the red flour beetle, Tribolium castaneum. Journal of Parasitology 89: 516-521.

Pappas, P. W. and Leiby, D. A. 1986. Variation in the sizes of eggs and oncospheres and the numbers and distributions of testes in the tapeworm, Hymenolepis diminuta. Journal of Parasitology 72: 383-391.

Pappas, P.W., Marschall, E. A., Morrison, S.E., Durka, G.M., and Daniel, C.S. 1995. Increased coprophagic activity of the beetle, Tenebrio molitor, on feces containing eggs of the tapeworm, Hymenolepis diminuta. International Journal for Parasitology 25: 1179-1184.

Peterson, C. H. and Renaud, P. E. 1989. Analysis of feeding preference experiments. Oecologia 80: 82-86.

Podesta, R. B. and Holmes, J. C. 1970. Hymenolepidid cysticercoids in Hyalella azteca of Cooking Lake, Alberta: life cycles and descriptions of four new species. Journal of Parasitology 56: 1124-1134.

Poulin, R. 1994. Parasite manipulation of host behavior: should hosts always lose? Oikos 70(3): 479-484.

Poulin, R. 1996. Sexual inequalities in helminth infections: a cost of being a male? American Naturalist 147: 287-295.

Rau, M. E. 1979. The frequency distribution of Hymenolepis diminuta cysticercoids in natural, sympatric populations of Tenebrio molitor and T. obscurus. International Journal for Parasitology 9: 85-87.

Reimchen, T. E. and Nosil, P. 2001. Ecological causes of sex-biased parasitism in threespine stickleback. Biological Journal of the Linnean Society 73: 51-63.

Richards, K. S. and Arme, C. 1985. Phagocytosis of microvilli of the metacestodes of Hymenolepis diminuta by Tenebrio molitor haemocytes Parasitology 90: 365- 374.

Roa, R. 1992. Design and analysis of multiple-choice feeding-preference experiments. Oecologia 89: 509-515.

134 Robb, T. and Reid, M.L. 1996. Parasite-induced changes in the behavior of cestode- infected beetles: adaptation or simple pathology? Canadian Journal of Zoology 74: 1268-74.

Saleem, M.A. and Shakoori, A. R. 1986. The effect of seven different relative humidities on survival and body weight loss of starved Tribolium castaneum adults. Pakistan Journal of Zoology 18: 395-402.

Sanchez-Pinero, F. and Perez-Lopez, F. J. 1998. Coprophagy in Lepidoptera: observational and experimental evidence in the pyralid moth Aglossa pinguinalis. Journal of Zoology 244: 357-362.

SAS/STAT® User’s Guide, 8.0 edition, 2001. SAS Inst., Inc, Cary, NC.

Seyoum, E., Moore, D. and Charnley, A. K. 1994. Reduction in flight activity and food consumption by the desert locust Schistocerca gregaria, Forskal (Orth.; Cyrtacanthacrinae), after infection with Metarhizium flavoviride. Journal of Applied Entomology 118: 310-315.

Schalk, G. and Forbes, M. R. 1997. Male biases in parasitism of mammals: effects of study type, host, age, and parasite taxon. Oikos 78: 67-74.

Sheridan, L. A. D., Poulin, R., Ward, D. F. and Zuk, M. 2000. Sex differences in parasitic infections among arthropod hosts: is there a male bias? Oikos 88: 327- 334.

Shostak, A. W. and K. A. Smyth. 1998. Activity of flour beetles (Tribolium confusum) in the presence of feces from rats infected with rat tapeworm (Hymenolepis diminuta). Canadian Journal of Zoology 76: 1472-1479.

Soave, O. and Brand, D. 1991. Coprophagy in animals: a review. Cornell Veterinarian 81 (4): 357-364.

Sokal, R. R. and F. J. Rohlf. 1995. Biometry: The Principles and Practice of Statistics in Biological Research. 3rd ed. New York: W. H. Freemand and Company.

Sokoloff, A. 1974. The biology of Tribolium. Vol. 2. Oxford: Clarendon Press.

Soltani, N. 1984. Effects of ingested diflubenzuron on the longevity and peritrophic membrane of adult mealworms (Tenebrio molitor L.). Pesticide Science 15: 221-225.

Sorensen, R. E. and Minchella, D. J. 1998. Parasite influences on host life history: Echinostoma revolutum parasitism of Lymnaea elodes snails. Oecologia 115: 118-195.

135 Thomas, M. B., Blanford, S. and Lomer, C. J. 1997. Reduction of feeding by the variegated grasshopper, Zonocerus variegatus, following infection by the fungal pathogen, Metarhizium flavoviride. Biocontrol Science and Technology 7: 327- 334.

Thompson, S. N. 1983. Biochemical and physiological effects of metazoan endoparasites on their host species. Comparative Biochemistry and Physiology 74B(2): 181-211.

Tschinkel, W., Willson, C., and Bern, H.. 1967. Sex pheromone of the mealworm beetle (Tenebrio molitor). Journal of Experimental Zoology 164: 81-86.

Ubelaker, J. E., Cooper, N. B. and Allison, V. F. 1970. Possible defensive mechanism of Hymenolepis diminuta cysticercoids to hemocytes of the beetle Tribolium confusum. Journal of Invertebrate Pathology 16: 310-312.

Utsunomiya, A. and Iwabuchi, K. 2002. Interspecific competition between the polyembryonic wasp Copidosoma floridanum and the gregarious endoparasitoid Glyptapanteles pallipes. Entomologia Experimentalis Et Applicata 104: 353- 362.

Van der Wal, R., J. Irvine, A. Stien, N. Shepherd and S. D. Albon. 2000. Faecal avoidance and the risk of infection by nematodes in a natural population of reindeer. Oecologica 124: 19-25.

Vinagre, A. S. and DaSilva, S. M. 2002. Effects of fasting and refeeding on metabolic processes in the crab Chasmagnathus granulata (Dana, 1851). Canadian Journal of Zoology 80: 1413-1421.

Voge, M. and Heyneman, D. 1957. Development of Hymenolepis nana and Hymenolepis diminuta (Cestoda: Hymenolepididae) in the intermediate host Tribolium confusum. University of California Publications in Zoology 59: 549- 580

Webb, T. J. and Hurd, H. 1996. Hymenolepis diminuta: metacestodes-induced reduction in the synthesis of the yolk protein, vitellogenin, in the fat body of Tenebrio molitor. Parasitology 112: 429-436.

Webb, T. J. and Hurd, H. 1999. Direct manipulation of insect reproduction by agents of parasite origin. Proceedings of the Royal Society of London: Biological Series 266: 1537-1541.

Wedekind, C. and Jakobsen, P. J. 1998. Male-biased susceptibility to helminth infection: an experimental test with a copepod. Oikos 81: 458-462.

136 Worden, B. D., Parker, P. G. and Pappas, P. W. 2000. Parasites reduce the attractiveness and reproductive success in male grain beetles. Animal Behaviour 59: 543-50.

Yan, G., Stevens, L. and Schall, J. J. 1994. Behavioral changes in Tribolium beetles infected with a tapeworm: variation in effects between beetle species and among genetic strains. The American Naturalist 143(5): 830-847.

Zuk, M. 1990. Reproductive strategies and disease susceptibility – an evolutionary viewpoint. Parasitology Today 6: 231-233.

Zuk, M. and McKean, K. A. 1996. Sex differences in parasitic infections: patterns and processes. International Journal for Parasitology 26: 1009-1024.

137