Protein content of spermatophores and male investment strategies in nectar-feeding butterflies

PROTEIN CONTENT OF SPERMATOPHORES A N D M A L E INVESTMENT STRATEGIES IN N E C T A R - F E E D I N G BUTTERFLIES by C H A R L E N E J E A N HIGGINS B . S c , Simon Fraser University, 1987 M.Sc., The University of British Columbia, 1991 A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES Department of Zoology We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH C O L U M B I A July 1996 © Charlene Jean Higgins In presenting this thesis degree in partial fulfilment of the requirements for an advanced at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes department or by his or her representatives. may be granted It is by the head of my understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT The objective of this thesis was to investigate how male investment in ejaculates, both in terms of nutritional quality and sperm content, varies with the mating system in nectar-feeding butterflies. Protein content was determined using a dye-binding protein assay (Bio-Rad), and used as a measure of ejaculate quality. A comparative study on 11 species of butterflies from 2 genera was conducted to examine how male nutrient investment varied with female mating frequency (polyandry). Male ability to produce more than one large nutritious ejaculate was evaluated using three species of pierid that varied in the degree of polyandry. The cost of ejaculate production, in terms of recuperation time, was investigated in Pieris napi and Pieris rapae, two polyandrous species of pierid. Lastly, the effect of male body size on sperm precedence was examined using P. napi. I found that relative to males in monandrous systems, males in polyandrous systems transferred larger first ejaculates that contained proportionally more protein. Furthermore, the degree of polyandry had a substantial effect on the reproductive performance of males. M y results suggest that male capacity to produce large, nutritious ejaculates is limited in monandrous species, and that males in polyandrous systems are better adapted to mating more than once. The interval between first and second matings by P. napi and P. rapae males had a strong effect on the size and protein content of second ejaculates. Copulation durations were longer and ejaculates smallest in matings involving recently mated males, suggesting that ejaculates are costly for males to produce. In virgin matings by Pieris napi males, ejaculate mass was positively correlated with male body mass but protein content was not. The effect of male body size was investigated in doubly mated P. napi females using normal and irradiated males. Relative male size had a significant effect on paternity. Larger males obtained the majority of fertilizations regardless of female mating status (virgin or nonvirgin). The effect of male body size on the proportion of offspring sired supports the hypothesis that sperm competition has played a major role in the evolution of ejaculate size. T A B L E OF CONTENTS ABSTRACT ii T A B L E OF CONTENTS iii LIST OF T A B L E S iv LIST OF FIGURES v PREFACE vi ACKNOWLEDGEMENTS vii INTRODUCTION , General Ecology of the Butterflies Used in This Study Technique Used to Analyze Protein Content of Spermatophores 1 9 10 C H A P T E R 1 PROTEIN CONTENT OF SPERMATOPHORES IN R E L A T I O N TO THE M A T I N G S Y S T E M . 12 Introduction 12 Materials and Methods 13 Results 16 Discussion 22 C H A P T E R 2 M A L E INVESTMENT IN SUCCESSIVE E J A C U L A T E S IN R E L A T I O N TO THE M A T I N G S Y S T E M Introduction. Materials and Methods Results Discussion 26 26 28 29 33 C H A P T E R 3 EFFECT OF M A L E M A T I N G HISTORY A N D B O D Y SIZE O N E J A C U L A T E SIZE A N D Q U A L I T Y IN TWO P O L Y G A M O U S BUTTERFLIES Introduction Materials and Methods Results Discussion 39 39 40 42 49 C H A P T E R 4 EFFECT OF M A L E B O D Y SIZE O N S P E R M P R E C E D E N C E IN PIERISNAPI L Introduction Materials and Methods Results Discussion 53 53 54 57 64 CONCLUSIONS 68 BIBLIOGRAPHY 74 APPENDIX 1 81 LIST OF T A B L E S Table 1.1. Body mass of males at eclosion, mass and protein content of first ejaculates in relation to the degree of female polyandry 17 Table 2.1. Eclosion mass of males and females 30 Table 3.1. Comparison of ejaculates produced by males remated the same day (0), the next day (1), two days later (2), three days later (3) 43 Table 3.2. Comparison of first and second ejaculates produced by large and small P. napi males 47 Table 4.1. Mean percentage hatch and fecundity of females used in each treatment.... 58 Table 4.2. The effect of mating order on paternity 59 LIST OF FIGURES Figure 1.1. Wet mass of ejaculates transferred by virgin males in relation to male body mass at eclosion for nine species of pierid and two species of satyrid 18 Figure 1.2. Amount of protein in ejaculates transferred by virgin males in relation to a) male body mass at eclosion and b) ejaculate mass 19 Figure 1.3. The relationship between residual protein mass in ejaculates transferred by virgin males and the degree of female polyandry 20 Figure 1.4. Relative size of ejaculate transferred in relation to the degree of female polyandry 21 Figure 2.1. The effect of male mating history on reproductive investment in three successive ejaculates, in terms of a) ejaculate mass relative to male body mass.(%), and b) ejaculate protein mass relative to ejaculate mass (%), transferred by males of P. brassicae, A. crataegi and P. rapae in their first three matings 31 Figure 2.2. Total investment by individual male P. brassicae, A. crataegi and P. rapae mating 1 to 3 times 34 Figure 2.3. The relationship between protein mass in ejaculates and copula duration in first, second and third matings of transferred by individual males of P. brassicae, A. crataegi and P. rapae 35 Figure 3.1. The relationship between protein content (mg) of first ejaculates transferred by P. napi males and female body mass at eclosion 44 Figure 3.2. The effect of male mating history on the mass, protein content and proportion of protein (% wet weight) of ejaculates produced by Pieris napi and P. rapae males 45 Figure 3.3. Protein content (a) and relative ejaculate mass (b), transferred by small and large males in their first and second matings 48 Figure 4.1. Proportion of offspring sired by the second male (P2) in relation to male 61 body size in doubly mated P. napi females Figure 4.2. Proportion of offspring sired by the second male (P2) as a function of relative male body size 62 vi PREFACE Three of the chapters of this thesis have been previously published or are in press. Chapter 1 appears elsewhere as: Bissoondath CJ and Wiklund C (1995) Protein content of spermatophores in relation to monandry/polyandry in butterflies. Behavioral Ecology and Sociobiology 37 : 365-371, Chapter 2 as: Bissoondath CJ and Wiklund C (1996) Male butterfly investment in successive ejaculates. Behavioral Ecology and Sociobiology (in press), and Chapter 3 as: Bissoondath CJ and Wiklund C (1996) Effect of male mating history and body size on ejaculate size and quality in two polyandrous butterflies, Pieris napi and P. rapae (Lepidoptera: Pieridae). Functional Ecology (in press). ACKNOWLEDGEMENTS I first wish to thank the two people that gave me the opportunity to do this. I'd like to thank Dr. Geoff Scudder, for asking me to be his Ph.D student (I'm sure he's still wondering why he did that), and for generously taking care of all my needs (from office whoas to flying me all over the place for conferences). I am most grateful to Dr. Christer Wiklund, at Stockholm University, Sweden for giving me the opportunity of a lifetime by inviting me to come work in his lab. It was an honour to work and learn from such a great butterfly naturalist and evolutionary ecologist. Simply amazing!!! Next I must thank Nina Wedell for her direction, thoughts and discussions, tuning me in and teaching me protein analysis. Thanks dear!! I won't ever forget how much fun we had or how hard we worked. I would also like to thank Bengt Karlsson, Soren Nylin and P-O Wickman for their thoughts and comments on my ideas and manuscipts, the secretary Pia Widenstrand and Annika Giertz for taking care of my office and lab needs in Sweden, and last but not least, Hasse Termin for getting me interested in Swedish football. Back in Vancouver, I'd like to thank Launi Lucas, Diane Mellor and Karen Needham for keeping me informed and all their help while I was out of the country, Allistar Blachford for all the years of computer know hows and last but not least my committee. Dr. Judy Myers, thank for your friendship and guidance. We've done a couple of thesis' together haven't we! Dr. Marty Adamson for adding some fun and unpredictability to my meetings and beer sessions, Dr. Harold Kasinsky and Dr. K i m Cheng for their alternate views. Thank you all. Now most importantly I must thank my friends and family that supported me and always believed I could do this. First my dear, dear friend Johnny Morgan. The one who has been and is always there for me. To my mom, dad, sisters, my inlaws and Claire who took care of my dear angel while I had to work on this; I could not, and would not, have finished this without you. Thank you all dearly and completely for everything. Most of all I would like to thank Rohan Bissoondath for his love and kindness. And finally, my dear, sweet angel, my daughter Alexandra who shows me every day THERE'S W A Y M O R E TO LIFE!!!! This study was supported by a University Graduate Fellowship from the University of British Columbia to C.J. Bissoondath, a Natural Sciences and Engineering Council of Canada Research Grant to Dr. G.G.E. Scudder, and a Swedish Natural Science Grant-to Dr. C. Wiklund. To all sources I am grateful. INTRODUCTION Reproductive effort is the proportion of an organism's total available energy used in reproduction (Thornhill and Alcock 1983). In some species, in addition to sperm, males provide females with nutrients and other substances prior to, during or shortly after mating. Referred to as nuptial gifts, these gifts may take the form of a prey item, a male-produced glandular product, such as salivary secretions and spermatophores, or in some cases part of, or the male himself (Thornhill 1976, Thornhill and Alcock 1983, Elgar 1992). Reproductive effort has been further divided into mating effort and paternal investment in an attempt to determine the adaptive significance of male reproductive investment when males provide females with more than just sperm during courtship or copulation (Low 1978, Simmons and Parker 1989). There is considerable controversy in the literature concerning the function of nuptial gifts, and thus, the selective forces responsible for these male traits and the behaviours involved in its use (Wicker 1985; Gwynne 1984, Simmons and Parker 1989, Wedell 1994). Nuptial gifts may be the result of natural selection when they increase the total number or quality of eggs a female will produce, or the result of sexual selection when they increase the number of times a male mates, or the number of eggs he fertilizes (Alexander and Borgia 1979; Simmons and Parker 1989). These two functions need not be mutually exclusive as males can benefit by both increasing a female's reproductive output and decreasing a female's receptivity to subsequent mating. A l l male Lepidoptera produce a spermatophore during copulation. The spermatophore is produced directly in the bursa copulatrix of females by the sequential transfer of sperm and accessory gland substances (Khalifa 1950; Drummond 1984). Its ultimate shape is determined in part by the size of the bursa of the female, and in part by the shape of the cuticular simplex and aedeagus of the male (Drummond 1984). In addition to sperm, spermatophores contain mostly water (50-86%), proteins, hydrocarbons, glycerids, phospholipids, sterols and chitin (Boggs 1981a; Drummond 1984; Marshall 1985). After mating, sperm exit the spermatophore and travel up the female spermathecal duct to the spermatheca where they are stored. The spermatophore generally collapses soon after the sperm exit, and depending on the species, is either degraded or absorbed in the bursa. In most butterflies, either the empty spermatophore or the collum of the spermatophore remains in the bursa and can be used to determine the mating frequencies of females (Drummond 1984). However, such measures of the degree of polyandry (female mating frequency) must be considered as estimates, as they vary among generations, with female age and population density. 2 In principle the benefits of remating should outweigh the costs. Thus the degree of polyandry should reflect a balance of the costs and benefits associated with a mating system. Furthermore, unless there are distinct advantages to multiple mating, the loss of time incurred by extra matings has been shown to be sufficient to select for nonreceptivity in mated females (Parker 1970). Female butterflies have a refractory period after mating, during which they reject courting males. Svard and Wiklund (1986) have suggested that because copulation can be quite time consuming, and the number of eggs laid is likely dependent on the time allocated to searching for oviposition sites and egg laying, females need a refractory period between matings. There are several possible advantages to multiple mating, and although discussed with reference to Lepidoptera, these benefits also apply to other systems. First, females may mate repeatedly to ensure an adequate sperm supply, although one mating generally furnishes a female with enough sperm to fertilize all her eggs (Sims 1979; Lederhouse 1981; Rutowski 1984). Very little is know about factors that may affect sperm content of spermatophores. Recently Gage and Cook (1994) showed that diet affects spermatogenesis in a polyandrous moth and Gage (1995) found that larval rearing density affected the number of sperm transferred. Although it is generally assumed that stored sperm remain viable, Tschudi-Rein and Benz (1990) found that after 10 days only inactive spermatozoa were found in the spermatheca of Pieris brassicae. Butterflies that lay eggs singly, such as many of the pierids, must spend time searching for oviposition sites, and likely have to mate throughout their life to replenish sperm. In addition to supplying sperm, multiple mating may provide females male-derived resources that enhance female survivorship and reproduction. Radiotracer studies have shown male-derived amino acids in the soma and eggs of female butterflies within 24 hours after mating (Boggs and Gilbert 1979; Boggs and Watt 1981; Boggs 1981a, 1990). Whether or not these male-derived nutrients increase female reproductive output is somewhat controversial and species-specific, with some studies finding that female reproductive output was enhanced (Boggs 1981a; Boggs and Watt 1981; Oberhauser 1989; Watanabe 1988; Watanabe and Ando 1993, 1994; Wiklund et al. 1993) and others finding no effect (Jones et al. 1986; Svard and Wiklund 1988, 1991). Females may mate multiply to increase the genetic diversity of the progeny they produce. This may be valuable for females that experience variable and unpredictable environments. Maximal genetic diversity would be obtained if sperm from all matings mixed in the spermatheca, which does not appear to be true for several species of Lepidoptera (Drummond 1984; Gwynne 1984). However, the genetic benefits of polyandry are difficult to verify, and genetic diversity resulting from repeat mating has been argued to be too small to explain high mating frequencies by females (Williams 1975). Lastly, females may mate multiply to minimize the loss of time and energy required to resist courting males. Population density and sex ratio may affect the efficacy of female mate-refusal behaviours. However, females of most species control mating frequency and are able to reject and avoid courting males (Forsberg 1988; Svard and Wiklund 1989), although once mating has been initiated, it appears that males decide when copulation terminates (Wickman 1985; Svard and Wiklund 1986; Wiklund etal. 1993). Longevity and reproduction in butterflies and other herbivorous insects is often constrained by protein (Engleman 1970; Wiggelsworth 1972; Marshall 1982). Although nectar contains energy rich carbohydrates, it contains very little protein (Baker and Baker 1973; Watt et al. 1974). Thus nectar-feeding butterflies must obtain most of the protein needed for somatic maintenance and reproduction during the larval-feeding stage. As a result, males are possibly limited in their ability to replace nutrients used for ejaculate production. Moreover, nutrients allocated to reproduction may decrease the pool of nutrients available for somatic maintenance. In some nectar-feeding butterflies, once females have depleted their nitrogenous reserves they cease to reproduce even with unlimited amounts of sucrose in their diet (Boggs 1986; Karlsson 1989). Although butterfly spermatophores contain a complex of nutrients (Marshall 1982, 1985), protein content of ejaculates can be used as a measure of nutritional quality and male investment because protein is likely a valuable resource to both sexes of nectar-feeding butterflies. Only two other studies to date, one using butterflies (Marshall 1985) and the other using bushcrickets (Wedell 1994), have quantified the protein content of spermatophores. A l l others have used nitrogen content or spermatophore mass as a measure of male investment. Although nitrogen and protein content of ejaculatels should be positively correlated, measuring the protein content of ejaculates is a better measure of male reproductive investment because nitrogenous products contained in the ejaculate may be as reproductively important to males and females as protein (such as other non-protein nitrogen containing compounds). Lepidoptera are good models to use to query the function of male nuptial gifts and the evolution of multiple mating by females because males allocate nutrients to reproduction and female mating systems vary from strict monogamy to strong polyandry (Ehrlich and Ehrlich 1978; Drummond 1984; Svard and Wiklund 1989; Wiklund and Forsberg 1991). Female mating frequency is one of the main factors controlling the level of sperm competition (Parker 1970) and paternal certainty. Consequently, the sperm and/or nutrient content, and the size of ejaculate transferred may vary with the likelihood females will remate. Using both comparative and experimental approaches, the general objective of my thesis was to determine how male investment in ejaculates (in terms of size and protein content) varies with the mating system in order to assess: 1) some costs and benefits associated with multiple mating to both sexes and 2) the effect of sperm competition on the role of the ejaculate. This thesis is organized into 4 parts; the first three examine the protein content of spermatophores in relation to the degree of polyandry, male mating history and male size; while the last chapter examines the effect of male body size on sperm competition. To facilitate the integration of these chapters, the first three of which have been published or are in press, each is introduced below in an integrative format. Chapter 1 Protein Content of Spermatophores In Relation To The Mating System Several hypotheses have been proposed concerning evolution of the nutritional quality of ejaculates. Marshall (1982) predicted that the nutrient content of spermatophores should vary depending on the degree of paternal certainty, the probability of finding mates, and a male's reproductive gain as a function of their investment. He suggested that in systems where willing mates are hard to find and where paternal certainty is high, as for example in monandrous species, investment of nutrients should be high. Conversely when mating success is high, males should invest less in each mating to maintain their ability to mate repeatedly. This hypothesis suggests that male-derived nutrients evolved as paternal investment with males "rewarding" females by increasing their investment in systems where females provide high paternal certainty to the donor. In monandrous species, nutrients contained in ejaculates will likely only be used to benefit the donor's offspring. In polyandrous species, where it is likely females will remate, a male's certainty of paternity maybe lower. A paternal investment function predicts that males in monandrous systems should transfer spermatophores containing relatively more nutrients than those produced by males in polyandrous systems (Boggs 1981a; Gwynne 1984). Moreover, if males of different species allocate the same proportion of resources into reproduction regardless of mating system, males in monandrous species would be expected to transfer more nutritious ejaculates than males in polyandrous species, simply because they have a lower average lifetime number of matings. Alternatively, in polyandrous systems, sperm competition may select for male adaptations that maximize their fertilization success. The duration of a female's postmating refractory period has been shown to increase with the size of spermatophore transferred (Oberhauser 1988,1989; Wiklund and Kaitala 1995). Therefore, because of the increased risk of sperm competition, if spermatophores influence a male's fertilization success, ejaculate size should increase with the degree of polyandry, whereas nutrient content may not. Walker (1980) suggested that females use multiple mating as a strategy to acquire nutrients. Since then, several other studies have proposed that polyandry evolved to maximize accumulation of male-derived nutrients by females (Kaitala and Wiklund 1994; Leimar et al. 1994; Karlsson 1995). Depending on the balance between nutrient availability and the nutritional requirements of the female, selection should favour repeat matings by females when males transfer nutrients that enhance her reproductive output and/or survival. Hence, this hypothesis predicts that spermatophores produced by polyandrous males should contain more protein than those of monandrous males. Thus there should be a positive relationship between the degree of female polyandry and protein content of the ejaculate. I used a comparative approach to assess the relationship between female mating frequency and male nutrient investment in an attempt to determine some of the selective pressures responsible for the role of the ejaculate in butterflies that may be limited in their access, as . adults, to nutrients needed for reproduction. To do this I collected data on ejaculate size and protein content for 9 of the 12 Swedish pierid butterflies and 2 species of satyrid butterflies. Chapter 2 Effect of the Mating System on the Reproductive Performance of Males Evidence suggests that spermatophores are physiologically costly and time consuming for males to produce. In several species of Lepidoptera, it has been shown that copulation duration increases and ejaculate size decreases the shorter the interval between matings (Rutowski 1979; Sims 1979; Boggs 1981a; Sviird 1985; Svard and Wiklund 1986, 1989; Oberhauser 1988, 1992; He and Tsubaki 1992; Royer and McNeil 1993; Chapter 3). The degree of polyandry likely affects the mating frequency of both sexes. As a result, male nutrient investment in ejaculates and their ability to produce more than one ejaculate may be influenced by the mating system. In monandrous systems, because females generally mate only once, the number of times a male will have the opportunity to remate is low. If there is little selection on males in monandrous systems to produce a second large ejaculate, males may invest heavily in the first ejaculate and less in subsequent ejaculates. Additionally, at least for nectar-feeding butterflies, males may have a limited amount of nutrients to allocate to ejaculate production. Therefore, if all males have the same amount of resources to invest in reproduction, investment by males in first ejaculates should be higher than that by males in polyandrous systems because the average number of matings by males in polyandrous systems can be much higher than that for males in monandrous systems. Thus, males in polyandrous systems may have to invest less in each mating in order to maintain their ability to mate repeatedly (Marshall 1982). However, relative to males in monandrous systems, it has been shown that males in polyandrous systems invest proportionally more resources in reproduction (Karlsson 1995). Karlsson (1995) also found that polyandrous females allocated less of their own resources to reproduction than monandrous females. High investment by males and low investment by females in polyandrous systems suggests that females in polyandrous systems may depend more on male-derived resources in order to attain their reproductive potential. In monandrous systems, females mate on average only once, thus their fitness may not be as dependent on male-derived resources as long as variability among males in ejaculate quality is small. In polyandrous systems, if females use mating to acquire male-derived nutrients, males should be able to produce produce more than one high quality ejaculate. The objective of chapter 2 was to ascertain how male investment in successive ejaculates varies with the mating system. I used 3 closely related species of pierid that ranged in female mating frequency from highly polyandrous to essentially monandrous. A l l are members of the subfamily Pierinae, with the most polyandrous species, Pieris rapae L., and the most monandrous species, P. brassicae L., sharing the same larval host plants (largely crops of Brassica spp.). Based on spermtophore counts of wild-caught females, the degree of polyandry for the black-veined white, Aporia crataegi L . , is intermediate between the two other species. Chapter 3 Effect Of Male Mating History And Body Size On Ejaculate Size And Quality In Two Polygamous Butterflies Among several species of Lepidoptera, the size of ejaculate transferred in virgin matings is generally correlated with male body size (Boggs 1981a; Rutowski et al. 1983; Svard and Wiklund 1986, 1989; Oberhauser 1988; Royer and McNeil 1993; Wiklund and Kaitala 1995), and the duration of the female's post-mating refractory period increases with increasing ejaculate size (Obara et al. 1975; Sugawara 1979; Rutowski 1980; Rutowski et al. 1981; Oberhauser 1989; Wiklund and Kaitala 1995). Moreover, in some polyandrous species, in twice-mated females, the larger male sired the majority of offspring produced (LaMunyon and Eisner 1993; Oberhauser unpub.). Therefore, males benefit in two ways from the production of a large ejaculate. In polyandrous systems where females mate repeatedly, because male reproductive success may be influenced by the size of ejaculate transferred, males may have the ability to produce more than one large ejaculate. There are several ways males may allocate their reproductive resources to a mating event. Male nutrient investment in successive ejaculates may be related to the function of the ejaculate. In polyandrous systems, i f the probability of obtaining future mates is high, males may have to invest less in each mating in order to maintain their ability to mate repeatedly (Marshall 1982). In most butterflies, the egg-laying rate of females, and thus her reproductive value decreases over time. Thus, males may invest heavily in their first mating, early in the season when more virgins are available, and less in subsequent ones (Svard and Wiklund 1986). Alternatively, males may invest maximally in all matings, producing an ejaculate as large possible to: 1) prolong the duration of the female's postmating refractory period and augment their success in sperm competition, and 2) possibly increase the reproductive value of older females when matings are scarce (Wiklund et al. 1993). Several studies have used spermatophore mass to estimate male investment in ejaculates (Boggs 1981a; Svard 1985; Oberhauser 1989; Svard and Wiklund 1989, 1991; Wiklund et al. 1993). Although spermatophore mass may not be an appropriate measure of male nutrient investment, few studies have attempted to determine the nutritional quality of spermatophores (for exception see Marshall 1985; Marshall and McNeil 1989; Oberhauser 1992; for Orthoptera see Wedell 1994). Therefore to determine how males, that are likely to mate more than once, allocate nutrients to a mating event, I used two of the most polyandrous species in the family Pieridae, Pieris napi L . and P. rapae L . . More specifically, because of the benefits accrued to males that produce large spermatophores, the objective of chapter 3 was to determine: 1) the benefits (in terms of the nutrient content (protein)) and costs to females of receiving spermatophores from males of different mating histories and sizes, and 2) mating costs to males, concerning the recovery time needed to produce spermatophores of similar size and quality to that transferred in their first mating. Chapter 4 Effect of Male Body Size on Sperm Precedence in Pieris napi L. When females mate several times, sperm from different males can compete directly to fertilize eggs (Parker 1970). Several studies in the last decade have revealed an increasing diversity of taxa in which sperm competition has had a significant effect on the evolution of mating systems (Parker 1970; see Smith 1984). In Lepidoptera, the majority of studies have concluded that the last male to mate generally fertilizes the majority of eggs, although in most there is considerable, unexplained variation around the proportion of offspring sired by the last mate (P2) (Labine 1966; Brower 1975; Walker 1980; Drummond 1984). Furthermore, most of what is known about sperm utilization in Lepidoptera is based on circumstantial information obtained from biological control studies (ie., sterile male programs) on pest species (Drummond 1984), in which proper controls and reciprocal crosses were likely not done. Across taxa, intraspecific variation around the mean P2 value exists in nearly all studies on sperm competition, yet few studies have examined the possible causes of this variation (for exceptions see Lewis and Austad 1990; Simmons and Parker 1992). Intraspecific variation in the proportion of offspring sired may be due to at least three things; random variability, differential competitive ability of sperm from different males, or female choice. Among males, individual variation in P2 may arise where males vary in their abilities to transfer and/or displace rival sperm, or if sperm of different genotypes differ in their competitive abilities (Simmons and Parker 1992). In most Lepidoptera, the duration of the female post-mating refractory period is positively correlated with ejaculate size, which in general, is dependent on male body size (Svard and Wiklund 1986, 1989; Oberhauser 1988, 1989; Kaitala and Wiklund 1995). Both the mechanical stimulation of the bursae caused by the physical presence of a spermatophore and the quantity of viable sperm in the spermatheca have been shown to influence the duration of female refractory behaviour (Labine 1964; Taylor 1967, Leopold 1976; Sugawara 1979). Thus large males may have a two-fold advantage over small males if both the female's postmating refractory period and a male's fertilization success are positively correlated with the size of ejaculate transferred. Very little is known about the effect of sperm competition on male investment in ejaculates (see Wedell 1991, 1994 for Orthoptera and LaMunyon and Eisner 1994, Cook and Gage 1995 and Gage 1995 for Lepidoptera). In some butterfly species, nutrients transferred in the ejaculate increase the longevity and reproductive output of the female (Boggs 1981a; Boggs and Watt 1981; Oberhauser 1989; Watanabe 1988; Watanabe and Ando 1993, 1994; Wiklund et al. 1993). Male investment of this type is considered paternal investment (Gwynne 1984; Simmons and Parker 1989) and as such should be associated with a high confidence of paternity and last male sperm precedence (Gwynne 1984). Furthermore, when females benefit from mating repeatedly, to encourage males to remate, selection should favour last male sperm precedence (Parker 1970, Walker 1980). However, in several species of Lepidoptera, the size of ejaculate transferred influences the duration of the female's postmating refractory period (Obara et al. 1975; Sugawara 1979; Rutowski 1980; Rutowski etal. 1981; Oberhauser 1989; Wiklund and Kaitala 1995 ) and thus likely increases the number of number of oocytes fertilized before the female remates. In this way the ejaculate functions as male mating effort (Simmons and Parker 1989). When the ejaculate functions as mating effort, P2 (the proportion of offspring sired by the last male to mate) may vary with ejaculate size. A positive relationship between ejaculate size and the degree of sperm precedence would corroborate a mating effort function. Alternatively, when ejaculates function as paternal investment, the last male to mate should sire the majority of offspring, regardless of the size of ejaculate transferred. These two different outcomes, regarding the paternity of the last male to mate, allow one to distinguish between a mating effort and paternal investment function. Using male body size as an indicator of the size of ejaculate transferred, the polyandrous green-veined white butterfly, Pieris napi, was used to examine the effect of ejaculate size on sperm precedence. General Ecology of the Butterflies Used in This Study I have used Swedish butterflies in the families Pieridae and Satyridae in these studies. Most of the butterflies belong to the family Pieridae. Two species of satyrid butterflies were used in the comparative study (Chapter 1). Members of the family Pieridae are generally sunshine lovers. Female mating frequency within the family ranges from essentially monandrous to highly polyandrous (Svard and Wiklund 1989; Wiklund and Forsberg 1991). The larvae of most white pierids feed on various cruciferous plants (Family Brassicaceae), while the yellow pierids nearly all use plants of the pea family (Family Fabaceae), except for Gonepteryx rhamni L . whose larvae feed on buckthorn (Family Rhamnaceae) and Aporia crataegi L . whose larvae feed on blackthorn, hawthorn and other rosaceous shrubs (Family Rosaceae). Some pierids overwinter as larvae and others as pupae. By contrast, satyrids are generally forest-dwelling and the larvae feed on grasses or sedges. Pararge aegeria L. is completely confined to forests while Lasiommata megera L. prefers grassy places with areas of bare rock. As a rule females are monandrous (Ehrlich and Ehrlich 1978; Svard and 10 Wiklund 1989). Swedish satyrines generally overwinter in the larval stage, although Pararge aegeria hibernate as pupae. The two most studied species in this thesis, the green-veined white, Peiris napi, and the small white, Pieris rapae , have similar life histories. Both belong to the subfamily Pierinae and spend the winter in the pupal stage with first generation (diapause) adults emerging in Sweden in May/June. Eggs are laid singly on host plant leaves throughout the female flight period and second generation directly-developed adults eclose in July/August. Larvae of P. rapae feed on wild and cultivated brassicaes, while those of P. napi feed on crucifers and are rarely found on brassicae plants. Males and females mate shortly after eclosion, and remate several times throughout their lifetime. Technique Used to Analyze Protein Content of Spermatophores Once copulation had terminated, females were immediately frozen. At the end of the day, the bursa copulatrix containing the spermatophore was dissected out of females. After quickly scraping away any somatic tissues, and leaving the appendix bursae when attached externally (in all species except P. napi) the bursa copulatrix was weighed on a Cahn electrobalance to the nearest 0.0001 g (wet weight). Once weighed the bursa was put into an Eppendorph tube and stored in a freezer at -20 °C until protein analysis. For the standard assay procedure, 0.2 - 0.3 ml. of 0.1% Triton-X solution was added to each Eppendorf tube containing one pierid ejaculate. The amount of solubolizing solution used depended on the mass of the ejaculate. Because of the small size of satyrid spermatophores, a microassay procedure had to be used. For this technique, 1 ml of 0.1% Triton-X was added to each tube containing a satyrid ejaculate. Once the Triton-X was added, ejaculates were completely macerated within the Eppendorf tube, and left for 5 days in a cold room at 4°C. The trays used to hold the tubes were completely covered with tinfoil to prohibit light from potentially damaging the proteins while they solubilized. Storage in the cold room allowed the protein time to solubilize and yet not break down prior to analysis. On the day of analysis, using a Biofuge 13 centrifuge, samples were spun in the Eppendorf tubes at 3000 rpm at 17 ^ C for 5 minutes, after which the aliquot was removed and placed in a test tube. Samples from the aliquot were subsequently taken and used for the protein assay. Protein concentration of spermatophores was determined using a dye-binding protein assay (Bio-Rad). For this procedure, an acidic dye, Coomassie Brilliant Blue, was added to 11 the protein solution, and subsequently measured at 595 nm using a spectrophotometer. The absorbance maximum for the acidic dye shifts from 465nm to 595 nm when binding to protein occurs. Colour changes in the dye, assessed using a UV-160 spectrophotometer, are associated with different concentrations of protein. Comparison with a standard curve provides a relative measurement of protein concentration. Standard curves were constructed using dilutions of bovine serum albumin (BSA). For the standard assay procedure 6 dilutions from a 1.4 mg/ml stock solution of B S A were used to construct the standard curve: 0.2, 0.4, 0.6, 0.8, 1.0, 1.2, and 1.4 mg/ml, the last point representing a sample from the stock solution (Appendix 1). For all pierid species except Leptidea sinapis L . and Anthocharis cardamines L., protein concentration was determined for two replicates from each sample and the mean taken. Because of the small size of ejaculates produced by L. sinapis and A. cardamines, there was only enough sample for one run. Satyrid spermatophores were analyzed using a microassay procedure. For this procedure, to construct the standard curve, 4 dilutions were made from a 50/ig/ml stock solution of B S A : 1, 5, 10 and 15jug/ml (Appendix 1). Most satyrid samples had to be diluted prior to analysis. This enabled me to get two replicates for each sample from which a mean was obtained. 12 C H A P T E R 1 PROTEIN CONTENT OF SPERMATOPHORES IN R E L A T I O N T O THE MATING SYSTEM Introduction In insects, there is a diversity of ways in which males provide females with nutrients during courtship and copulation. Males may provide females with prey or a food item, or produce their own nuptial gift from accessory gland secretions (Thornhill 1976; Gywnne 1984). Male accessory gland products, such as spermatophores, are found in a diverse array of taxa (Smith 1984) and may be either eaten or absorbed internally (Thornhill 1976a; Drummond 1984). In the Lepidoptera, males produce a spermatophore during copulation that contains both sperm and accessory substances. Nutrients, such as amino acids, contained in these packaged ejaculates have been found in both the eggs and soma of females (Boggs and Gilbert 1979; Boggs 1981a; Boggs and Watt 1981; Wiklund etal. 1993). Furthermore, singly mated females exhibit a higher reproductive output after receiving a larger ejaculate than females receiving a smaller one (Rutowski et al. 1987); females mating more than once exhibit a higher lifetime reproductive output and longevity than females mating only once (Watanabe 1988; Oberhauser 1989; Watanabe and Ando 1993; Wiklund etal. 1993). These findings suggest that male-derived accessory products possibly provide females with an additional source of nutrients, which may increase their reproductive output and/or longevity. Evidence suggests that spermatophores are physiologically costly for males to produce (Boggs 1981a; Rutowski etal. 1983; Svard and Wiklund 1986; Oberhauser 1988; Royer and McNeil 1993; Gage and Cook 1994; Kaitala and Wiklund 1995; Wiklund and Kaitala 1995). Male mating costs are indicated by the fact that: 1) recently mated males transfer much smaller ejaculates than ejaculates transferred by virgin males, and 2) the duration of copulation increases the shorter the interval between matings (Svard and Wiklund 1986, 1989; Kaitala and Wiklund 1995). Svard and Wiklund (1986) found that both the mass of accessory material and number of sperm transferred in the ejaculate were highest in the first mating compared to subsequent matings. Gage and Cook (1994) showed that male Lepidoptera developed on a low protein diet, with intense competition for food, suffered resource restrictions for spermatogenesis and, as adults, transferred reduced sperm numbers to females at mating. Therefore it seems that in some species, males make a considerable investment in the production of a spermatophore. 13 Nutrients donated by males to females at mating represent investment by the male in reproduction. Consequently, the nutritional composition of spermatophores may vary depending on the mating system. There are two hypotheses concerning evolution of the nutrient content of ejaculates. Marshall (1982) predicted that the nutrient content of spermatophores should vary depending on the degree of paternal certainty, the probability of finding mates, and a male's reproductive gain as a function of their investment. He suggested that in systems where willing mates are hard to find and where paternal certainty is high, as for example in monandrous species, investment of nutrients should be high. Conversely when mating success is high, males should invest less to maintain their ability to mate repeatedly. Thus, one hypothesis argues that male-derived nutrients evolved as paternal investment with males "rewarding" females by increasing their investment in systems where females provide high paternal certainty to the donor. Here, one would predict that males in monandrous systems would transfer spermatophores containing relatively more nutrients than those produced by males in polyandrous systems. Furthermore, if males of different species allocate the same proportion of resources into reproduction regardless of mating system, males in monandrous species would be expected to transfer more nutritious ejaculates than males in polyandrous species, simply because they have a lower average lifetime number of matings. The other hypothesis argues that polyandry evolved to maximize transfer of malederived nutrients to females, and is based on the several lines of evidence suggesting that females in polyandrous systems may actually forage for matings (Kaitala and Wiklund 1994; Leimar et al. 1994; Karlsson 1995). Hence, this hypothesis yields the prediction that spermatophores of polyandrous males should contain more protein than those of monandrous males, and that there should be a positive relationship between the degree of femalepolyandry and protein content of the ejaculate. To distinguish between these two hypotheses, I used protein content of ejaculates to determine how male nutrient investment varied interspecifically with female mating frequency. Although ejaculates contain other nutrients such as water, hydrocarbons and other lipids (Marshall 1985) that may also benefit the female, I chose protein content as a measure of ejaculate quality because of its reproductive importance and limited availability to nectarfeeding butterflies. Materials and Methods Butterflies used in this study were laboratory-reared offspring from females wildcaught in different parts of Sweden. A l l , except Aporia crataegi L., Pontia daplidice L . and 14 Lasiommata megera L., were diapause-generation offspring which had overwintered as pupae in the laboratory. A. crataegi overwintered as larvae, P. daplidice and L. megera individuals were directly-developed progeny produced by females collected in the wild that same year. To minimize the variation in male body size within each species, only males of average size were used, except for Colias palaeno L., Gonepteryx rhamni L . and Leptidea sinapis L. when few individuals were available. Likewise, generally only females of mean body size were used. Adults were weighed on an Cahn automatic electrobalance the day of eclosion and marked individually on their wings with a Staedler permanent pen. On the day matings were to be performed, males were introduced first into the 0.8 x 0.8 x 0.5 m mating cages and allowed to feed for one-half hour before virgin females were added. Cages contained a maximum of 20 butterflies and were not crowded. Mating cages contained species-specific host plants and either wild flowers or a potted Chrysanthemum sp. Into both nectaring sources, a 25% sugar solution was applied with a dropper into the middle of the flowers twice a day; first thing in the morning when the lights were turned on and again in the early afternoon. The majority of pierid matings were carried out in the laboratory in cages placed behind large windows, 2 m below 400 W HQIL lamps which were on from 0800 to 1800 h. Some pierid species will not mate readily under laboratory conditions. These species were placed in cages housed in greenhouses on the roof of the Department of Zoology at Stockholm University. The mating cages were set up exactly as those described and used in the laboratory. Butterflies were put into these cages following the same procedures as that used for laboratory matings. Only those matings which took place between 0800 and 1800h were included. Mating cages were inspected every 15 minutes for mating pairs. Upon discovery, pairs in copula were transferred to a 0.5 L plastic cup to ensure they were not disturbed by conspecifics in the cage. Immediately upon separation, both male and female were put into the freezer. Satyrid matings followed the same marking and removal procedures as that used for pierid matings. Matings were carried out in 0.5 x 0.5 x 0.5 m cages heated and lit by 100 W incandescent lamps. As for pierid matings, lamps were on between 0800 and 1800 h. Mating cages were checked every 10 minutes because satyrids mate for a shorter time. Once all matings for the day had been completed, individual pairs were removed from the freezer. Males were weighed on an automatic electrobalance and the bursa copulatrix containing the ejaculate was dissected out of females. The bursa was weighed to the nearest 0.0001 g. on the electrobalance and immediately put into individual plastic Eppendorf tubes 15 where they were stored in a freezer at -20 ° C until protein analysis. In butterflies the mass of the empty bursa is negligible (Marshall 1985) and thus likely has an insignificant effect on the actual mass or protein content of the ejaculate. Throughout this thesis, all references to the protein content and mass of the spermatophore include the wall of the bursa and will henceforth be referred to as the ejaculate. Protein concentration of spermatophores was determined using the dye-binding protein assay (Bio-Rad) described in the general introduction. Colour changes in the dye, assessed using a U V - 160 colour spectrophotometer, are associated with different concentrations of protein. From this, the actual amount of protein (mg) contained in the ejaculate was estimated. For all pierid species except Leptidea sinapis and Anthocharis cardamines, protein concentration was determined for two replicates from each sample and the mean taken. Because of the small size of ejaculates produced by these two species, there was only enough sample for one run. For the microassay procedure, most Satyrid samples had to be diluted prior to analysis. This allowed me to obtain a mean protein concentration from two replicates for each sample. For each species, the degree of polyandry refers to the average number of spermatophores found in wild-caught Swedish butterfly females as measured by Wiklund and Forsberg (1991). Because ejaculate weight and protein content (mg) were correlated with male body weight, residuals from least squares regression were used when necessary to remove the effects of male size. To control for shared ancestry, I used a phylogenetic analysis (Garland et al. 1993) to examine the relationship between the degree of polyandry and protein content of ejaculates. For this analysis, I used the phylogeny suggested by Geiger (1981) for European pierids, and for the satyrids the two species were assumed to be monophyletic. Assuming these phylogenies, ten independent contrasts for each residual protein mass (y) and degree of phylogeny (x) were extracted (Garland et al. 1993). A linear regression forced through the origin for these contrasts was then performed. A l l statistics were calculated using SYSTAT. Bartlett's test was used to determine heteroscadasticity and Lillifors test to assess normality (p > 0.05). 16 Results The protein content for all species examined are presented as means and SEs in Table 1.1. In all figures, each data point represents a mean for that species. Protein content of ejaculates from a male's first mating varied from 0.35 - 0.02 mg. and 3.83 - 2.23 % (measured as percent wet weight) between species. Across species there was a strong positive relationship between male body mass (eclosion weight) and ejaculate mass (r = 0.858, p = 0.001, n = 11; Fig. 1.1). The amount of protein (mg) found in spermatophores was dependent on both male body mass (r = 0.813, p = 0.002, n = 11) and ejaculate mass (r = 0.946, p < 0.001, n = 11; Fig. 1.2a and 1.2b respectively). The relationship between protein content of ejaculates and the degree of female polyandry was examined using residuals from the regression of protein content (mg) on male body mass (mg) because both protein content and ejaculate mass are correlated with male body size. There was a significant relationship between protein content and polyandry; first spermatophores produced by polyandrous species contained significantly more protein than those of relatively monandrous species (r = 0.729, p = 0.011, n = 11; Fig. 1.3). To control for the effects of shared ancestry, I performed a phylogenetic regression (cf. Harvey and Pagel 1991). Assuming a phylogeny for European pierids suggested by Geiger (1981), and that the two satyrids were monophyletic, ten independent contrasts for each residual protein mass (y) and degree of polyandry (x) were extracted (Garland et al. 1993). The regression through the origin for these contrasts was highly significant (t = 5.68, p < 0.001). Thus, it appears that males allocate proportionately more protein to their ejaculates with increasing degree of polyandry. In terms of investment per unit body weight (ejaculate mass / eclosion mass), males belonging to polyandrous species produced relatively larger ejaculates than monandrous species (r = 0.660, p = 0.027, n = 11) (Fig. 1.4), with Parage aegeria investing the least (0.59 %) and Pieris napi investing the most (11.4 %) (Table 1.1). Recently, Karlsson (1995) has showed that controlling for the effects of shared ancestry, males in polyandrous species allocated proportionally more of their larval resources to reproduction than males in monandrous species. There was a significant relationship across species between the absolute amount of protein (mg) transferred and the duration of copulation (r = 0.653, p = 0.029, n = 11). However, copulation duration had no effect on either the absolute or relative size of ejaculate transferred (p > 0.05). 17 'S.S 00 — c-g s o *•* - .5 f - WD — ^ ON co CN ^ NO © O CN r f CN CN O O -H -H +1 -H -H -H r - oo r-» NO ©g O O H O © d © © T f co ON oo « r-; fl 1 ^ ft ^ <~ 2 73 •© OA '3D S •> o CU V ci -H © -H -H O CO CN uo oq 00 CO CO CN CN o c i O <n CN o CN CN c i 6 d o d -H +1 -H ON — CN CN O O -H +1 .3 CN NO c> -H —< O c i ci 00 o ON o CO T f CO CN CO d 88 O O o c i CO CN ON CN O O O O O O d -H -H -H -H -H -H +1 -H q CN o CO NO 00 rO r- co NO ON 00 CN Tf' d -H -H CN 00 ON O CN* CO* 00 00 NO d d -H NO in < n d o o d d o o CO — d d -H -H T f CN ON — O CO* 3 83 £ 9i r- o CN ON CO CO « If 1 Ci O 8 .' -H -H co ON c« ON CN CN « fl W *3? o c i 00 'w s o c i NO CM 5 o O m ON -i f i ••o CO CN CO — NO o -H -H -H -H -H e s CO CO - CO CN CN CN cs cn <s cn CO ON B ° — S 2 2 "5 g OJD S d d d d -H -H +l -H -H -H -H +1 NO NO T f T> m m NO NO ON ON O 00 — NO r - « NO CN U0 O CN CN ^ - Ci CN -H +1 -H +1 -H <N S 3 N O o oo °1 . oo co r~ 2 S ON ^ IS r-» N O 3 .a CN 00 NO in ~ Tf* r T f T f m oo -* m NO NO in •+H -H m q -H 00 t~- m NO co 00 CO CN ON > n T f ON CN CN CN -H -H CO d d -H T f a ® J« d — ^ °. r- r- r- <M O ov 8 CN r CN CN CN o O oq CO CN d — s in ON NO m '35 ° o g O •< in r- o — CN CO CO CO r - CN ~ <-« »-< oi p CN oq 3= ^ O o .B -§ + fl. ' 3 ** •S|~ ^ 8 'S § « sv *i 3^ •Sm 3 < * .2. 2 (J .2, •> LZT s" c a> o *< 2 O CU H 37 S ON ON ON C « CU e3 - i — Q « 18 151 I 101 C/J cd • 3 8, 5H ST 11 10 0 0 100 150 50 Male Body Mass (mg) 200 Figure 1.1. Wet mass Of ejaculates transferred by virgin males inrelationto male body mass at eclosion for nine species of pierid and two species of satyrid (r = 0.86; p = 0.001, n = 11). Numbersreferto species in Table 1.1. 0,4 T S °' 3 2 3 .2 "3 dt uT •4 • 0,2 H c a • • 5 • g • ~* o a. o.i H 11 • 10 0,050 100 150 200 Male Body Mass (mg) 0,4 T 00 6 0,3 H • 7 2 3 • J2 "3 8 0,2 H a a '3 o 2 Ui « 4 • 5 0.H V 1 • 10 o.o- —i 5 10 15 Ejaculate Mass (mg) Figure 1.2. Amount of protein in ejaculates transferred by virgin males in relation to a) male body mass at eclosion (r = 0.81; p = 0.002; n = 11), and b) ejaculate mass (r = 0.95; p < 0.001; n = 11). Numbers refer to species in Table 1.1. 0,2 n 1 e/5 0,1 H a 2 a- 7 o,o H 9 11 <u 05 -o,H -0,2 10 — 1 1,0 1 1,2 1 1 1,4 1 1 1,6 1 1 1,8 1 1 2,0 1 1 2,2 Degree of Polyandry Figure 1.3. The relationship between residual protein mass in ejaculates transferred by virgin males (i.e. controlling for male body mass) and the degree of female polyandry, measured as the average number of spermatophores found in the bursa copulatrix of wild-caught females (r = 0. 73; p = 0.011, n = 11). Numbers refer to species in Table 1.1. 21 12" 1 • 10C/5 CC £ =3 O cd 8" 4 • 2 • 6H 4H 11 > •Z3 2 in 10 0 r- 1.0 1,2 • i 1,4 • 1 1,6 • — — i 1,8 1 1 2,0 1 1 2,2 Degree of Polyandry Figure 1.4. Relative size of ejaculate transferred (measured as ejaculate mass / male eclosion mass) in relation to the degree of polyandry (measured as the average number of spermatophores found in wild-caught females) (r = 0.66, p = 0.027, n = 11). Numbers refer to species in Table 1.1. 22 Discussion Male Lepidoptera are in a position to make parental investments, via a spermatophore, because proteins and other nutrients that are available during the larvalfeeding stage are difficult to obtain during the adult nectar-feeding stage (Marshall 1982). Marshall (1982) predicted that in monandrous systems, where willing mates are hard to find and because of high paternal certainty, investment of nutrients should be high. Conversely when mating success is high, males should invest less to maintain their ability to mate repeatedly. His prediction is based on the assumption that-males mating repeatedly must invest less in each mating to maintain their ability to mate. I found a strong correlation across species between male body mass and ejaculate mass (Fig. 1.1). Several other studies have found a strong correlation between male body size and ejaculate mass (Rutowski et al. 1983; Svard and Wiklund 1986, 1989; Oberhauser 1988; Royer and McNeil 1993; Wiklund and Kaitala 1994). Moreover, I found that the protein content of ejaculates was positively correlated with both male body mass and ejaculate mass (Fig. 1.2a, 1.2b). Interestingly across several species of bush crickets, Wedell (1994) also found a positive relationship between protein content and ejaculate size. However, contrary to Marshall's (1982) predictions, not only did males in polyandrous species transfer relatively larger ejaculates (Fig. 1.4), their ejaculates also contained proportionally more protein than ejaculates produced by monandrous species (Fig. 1.3). Karlsson (1994) found that while male butterflies allocated more resources to reproduction with increasing polyandry, females showed the opposite pattern. In females, proportional abdomen mass at eclosion decreased with the degree of polyandry. Female "deinvestment" with increasing polyandry suggests that adult females of polyandrous species may have a higher expected nutrient income, and is consistent with the idea that females can benefit from male nutrient donations at mating (Boggs 1990, Wiklund et al. 1993, Leimar et al. 1994). M y finding that the protein content of ejaculates increases with the degree of polyandry corroborates this hypothesis, and additionally suggests that females in polyandrous systems use mating to acquire male-derived nutrients. Males belonging to polyandrous species exhibit a number of adaptations to their mating system. Relative to males in monandrous species, males in polyandrous species (1) allocate proportionally more resources into reproduction, measured as abdomen relative to total body mass (Karlsson 1994), (2) transfer proportionally larger ejaculates (Svard and Wiklund 1989; Karlsson 1994, this study), and (3) are able to produce second ejaculates as 23 large as the first one sooner (Svard and Wiklund 1989, Chapter 2). Here I demonstrate a fourth male adaptation, with males in polyandrous systems transferring relatively more protein in their ejaculates than monandrous males. Males in polyandrous species do not appear to make a trade-off between ejaculate size and nutrient content (in terms of protein). However, protein is not the only nutrient contained in spermatophores. Spermatophores also contain about 80 % water (Boggs 1981a), lipids and hydrocarbons (Marshall 1985), all of which are potentially important and of benefit to the female. Therefore, it is possible that males belonging to monandrous species transfer more of one of these other nutrients. Rutowski et al. (1983) found that males across 4 families of butterflies transferred proportionally similar masses of ejaculate material. They concluded that nutritional investments made by males were similar across species. My results show that, in terms of the actual amount of protein transferred, this is not the case. There is a wide variation both in the amount of protein and the relative size of ejaculate transferred between species (Table 1.1). However, the relative amount of protein transferred (% wet wgt.) is similar across species, which suggests there may be a minimum amount of protein transferred regardless of mating system. Proteins may also be involved in other roles, such as sperm activation and the stimulation of oviposition, which may set a limit to the minimum amount that must be transferred (Oberhauser 1992). Interestingly, copulation duration also influenced protein content of ejaculates. This suggests that it takes longer for males to mobilize larger quantities of protein. As found by Svard and Wiklund (1986) and Karlsson (1995), I found that the relative ejaculate mass transferred by males to females at mating increased with increasing degree of polyandry (Fig. 1.4). In terms of the number of eggs a given male will fertilize there are two possible benefits associated with investment in a large ejaculate. In several polyandrous species, ejaculate mass is positively correlated with the length of a female's postmating refractory period (Labine 1964; Obara et al. 1975; Sugawara 1979; Rutowski et al. 1981; Oberhauser 1989; Wiklund and Kaitala 1995). As a result more of their sperm will be used to fertilize the eggs laid before the female remates. Secondly, larger males transfer ejaculates containing a greater quantity of material, which may contain more nutrients that females can use increase their reproductive output (Boggs and Gilbert 1979; Rutowski et al. 1987, Watanabe 1988; Svard and Wiklund 1989). In P. napi there is evidence that larger males have a higher fertilization success regardless of mating order as long as a female's previous or subsequent mates transfer smaller ejaculates (Chapter 4). Taken together these facts suggest that in polyandrous systems, the high remating frequency of females may select for large male size. Males in polyandrous species appear to invest more in reproduction than males belonging to monandrous species. Consequently, males in polyandrous species should be selected to produce larger ejaculates that protect their investment because the risk of sperm competition increases with female mating frequency. If male-derived nutrients functioned solely as paternal investment, protein content should have been highest in monandrous species where the donor can expect a high confidence of paternity (Boggs 1981a, Marshall 1982). However, I found that both the relative protein content and ejaculate size increased with the degree of polyandry. These results support the hypothesis that male-derived nutrients also function as mating effort in polyandrous systems, with the ejaculate designed to delay female remating. Therefore in some species of butterflies, ejaculates apparently play a dual role, functioning both as paternal investment by increasing the reproductive output of females and as mating effort by increasing a male's fertilization success the larger the ejaculate transferred (Simmons and Parker 1989). In this study I compared the amount of protein females receive in one mating. Therefore, from the female's perspective, the amount of protein received in one ejaculate is a conservative estimate as to the total amount of male-derived proteins females in polyandrous systems receive in their lifetime. Most female P. napi mate at least twice with some mating up to 5 times (Wiklund and Forsberg 1991). Thus, the actual amount of protein females in polyandrous systems receive in their lifetime will be much greater than that received by females in monandrous systems. Svard and Wiklund (1988, 1991) proposed that extent to which reproduction will be affected by male-derived nutrients will depend on (1) the degree of female polyandry, (2) the mass of ejaculate transferred, and (3) female longevity. Boggs (1990) predicted an inverse relationship between the importance of protein and the availability and utilization of proteins acquired by females in their diet. As adults, the species studied here feed predominantly on nectar, which contains very little protein (Baker and Baker 1973). Hence male-derived proteins should be of great importance to females. Polyandry would be selected for when males transfer nutrients via ejaculates that can be used by females to increase their lifetime reproductive output, whereas female monandry would be selected for when males do not transfer any nutrients or when females will not live long to benefit from remating (Wiklund and Forsberg 1991). Male-derived nutrients, such as protein, make it advantageous for females to remate. In some polyandrous species, multiply mated females had a higher 25 reproductive output than females allowed to mate only once (Rutowski et al. 1987; Watanabe 1988; Watanabe and Ando 1993; Wiklund et dl. 1993). When ejaculates contain nutrients used in reproduction, selection should increase a female's ability to obtain these nutrients. Females in polyandrous systems can increase the amount of protein they obtain in their lifetime by remating. From the female's perspective, I suggest that ejaculates function as an additional nutrient source in nectar-feeding butterflies. It is apparent that males in polyandrous systems make a substantial investment in reproduction (Watanabe 1988; Oberhauser 1989; Watanabe and Ando 1993; Wiklund et al. 1993). The investment, both in terms of male body weight and protein content, demonstrates how the ejaculate functions as paternal investment. At the same time, because males in polyandrous systems also produce relatively larger spermatophores, affecting both female mating frequency and the number of progeny a male can hope to sire, the spermatophore also functions as male mating effort. As shown by Wiklund and Forsberg (1991) relative male size increases with polyandry in butterflies. Leimar et al. (1994) argue that male nuptial gifts may reduce the value of large size in females and increase the value for males, assuming that large nuptial gifts can be traded for more offspring. Leimar et al. (1994) further suggest that food variability, causing some males to have much to provide and some females to be in great need, would be conducive to the evolution of such a mating system. Once such a giftgiving system has originated, males that are able to provide gifts of extra high quality will be selected for. In this way selection for high paternal investment will go hand in hand with selection for large size ejaculates that postpone future matings by the female as long as possible. Hence, the association between large ejaculate size and polyandry can be understood in terms of paternal investment and mating effort acting in unison. 26 C H A P T E R 2 M A L E INVESTMENT IN SUCCESSIVE E J A C U L A T E S IN R E L A T I O N TO THE M A T I N G S Y S T E M Introduction Male Lepidoptera transfer a spermatophore to females during copulation that contains both sperm and accessory gland products. Accessory gland products contain nutrients that, in some species, have been found to increase a female's reproductive output and longevity (Boggs 1981a; Boggs and Watt 1981; Oberhauser 1989; Watanabe 1988; Wiklund et al. 1993; Watanabe and Ando 1993, 1994). However, male investment in these packaged ejaculates is physiologically costly and time consuming. Not only do recently mated males transfer smaller ejaculates than males mating for the first time, copulations often last for much longer (Boggs 1981a; Rutowski etal. 1983; Svard and Wiklund 1986; Oberhauser 1988; Royer and McNeil 1993, Chapter 3). It takes time for males to replenish accessory gland materials (Rutowski 1979; Svard 1985; Svard and Wiklund 1986, 1989; Oberhauser 1988; Chapter 3). The degree of polyandry (female mating frequency) likely affects the mating frequency of both sexes. Consequently, male nutrient investment in ejaculates and their ability to produce more than one ejaculate may be influenced by the mating system. In butterflies, the male ejaculate has at least three effects; it i) contains sperm that are capable of fertilizing the eggs of the male's mate, (ii) influences the refractory period of the female, i.e. the duration of the period that the female is unreceptive to courting males, and iii) contains nutrients that are used by the female for somatic maintenance and for increasing her fecundity. Therefore, male investment in successive ejaculates may vary with the mating system. In monandrous systems, male opportunity for remating is low so there is likely to be weak selection pressure on male capability to produce large second ejaculates. Thus males in monandrous systems may produce a large first ejaculate but small successive ejaculates. For males in polyandrous systems the situation is different, with most females remating during their lifetime. Svard and Wiklund (1989) predicted that because males in polyandrous species on average mate more often than male in monandrous systems, they should have a higher capacity for producing many ejaculates. In general, the duration of a female's refractory period is positively correlated with the mass of ejaculate received (Labine 1964; Obara et al. 1975; Sugawara 1979; Rutowski 1980; Rutowski et al. 1981; Oberhauser 1989; Wiklund and Kaitala 1995). Prolonging the period females are unreceptive to courting males increases the amount of time over which a female lays eggs fertilized by the investing 27 male. Furthermore, the size of ejaculate transferred may influence a male's fertilization success when the female remates (Chapter 4). Because a male's fertilization success in polyandrous species may be influenced by the size of ejaculate transferred, there should be strong selection on males to produce several large ejaculates. In polyandrous species , male-derived nutrients may have a greater impact on the reproductive potential of females than female body size (Oberhauser 1989, 1996; Wiklund and Kaitala 1995). Males that contribute nutrients that benefit the female or her offspring are investing paternally (Simmons and Parker 1989). When male-derived nutrients enhance female reproductive output, males may also benefit from transferring a large ejaculate in terms of number of offspring sired. If across species, all butterfly males have the same amount of resources to invest in ejaculates, investment in first ejaculates by males in monandrous systems, where both males and females have an expected lifetime number of matings close to one, should be considerably larger that those transferred by males in polyandrous systems because the expected number of matings for males and females can be much higher (cf. Rutowski et al. 1983). When reproductive resources are limited (Svard 1985, Oberhauser 1988) and mating success is high, as in polyandrous systems, males may have to invest less in each mating in order to maintain their ability to mate more than once (Marshall 1982). The objective of this chapter was to investigate male reproductive investment in successive ejaculates, in terms of both mass and quality, to determine 1) how ejaculate quality (measured as protein content) changes in successive ejaculates and 2) if males in polyandrous systems are capable of maintaining production of larger ejaculates relative to males in more monandrous systems. I used three species of butterfly in the family Pieridae to compare the mass and protein content of ejaculates produced in the first versus subsequent matings. Female mating frequency varied from relatively polyandrous (small cabbage white, Pieris rapae L. (2.13)), to relatively monandrous (large cabbage white, P. brassicae L . (1.28)). The degree of polyandry for the black-veined white, Aporia crataegi L . (1.70), is intermediate between the two other species. For each species, the number in parentheses is an estimate of the degree of polyandry (female mating frequency), obtained by calculating the mean number of spermatophores found in the bursa copulatrix of wild-caught female Swedish butterflies (Wiklund and Forsberg 1991). As adults, the butterflies in this study feed primarily on nectar. Thus for reasons stated in the general introduction and Chapter 1, because of its limited availability and 28 potential reproductive benefit to females (cf. Boggs 1981b), protein content was used as a measure of ejaculate quality . Materials and Methods Butterflies used in this study were diapause-generation offspring produced by females wild-caught in different parts of Sweden. Pieris rapae and P. brassicae larvae were reared on Alliaria petiolata L . (Cruciferae), and overwintered as pupae in the laboratory. Aporia crataegi larvae were reared on Sorbus aucuparia L . (Rosaceae) and overwintered as larvae in the laboratory. Upon eclosion, adults were weighed on a Cahn electrobalance and marked individually on their wings with a Staedler permanent pen. Adults eclosed within three days of each other, and were held in a cold room at 4°C until the start of the experiment. Experiments were started within a day of the last adults to eclose. To minimize variation in male body size within a species, only males of mean body size were used (Table 2.1). Although I tried to use only females of mean body size, female body size was more variable because of the large number of females needed for this experiment (Table 2.1). Pieris rapae and P. brassicae matings took place in 0.8 x 0.8 x 0.5 m mating cages placed behind large windows, 2 m below 400 W HQIL lamps which were on from 0800 to 1800 h. Matings involving Aporia crataegi were carried out in cages of the same size housed in greenhouses on the roof of the Department of Zoology at Stockholm University because this species does not mate readily under laboratory conditions. The same protocol was followed as that used for laboratory matings. Males were introduced first into the mating cages and allowed to feed for one half-hour before virgin females were added. Each cage contained species-specific host plants and either wild flowers or a potted Chrysanthemum sp. A 25% sugar solution was applied with a dropper into the middle of the flowers twice a day, first thing in the morning when the lights were turned on or, for cages on the roof, when the butterflies were first introduced into the mating cages, and again in the early afternoon. Cages contained a maximum of 20 butterflies and were not crowded. Males were given the opportunity to mate every second day. I chose to remate males every second day because this is the minimum time needed by the most polyandrous species, P. rapae, to produce a second ejaculate of similar size and quality to that transferred in their first mating (Chapter 3). Only those males remating two days after their first mating were given the opportunity to remate for a third time, two days later. I ended the experiment after the third mating, because three is just above the number of times most P. rapae females mate 29 in their lifetime in the wild. Cages were inspected every 15 minutes for mating pairs. Upon discovery, copulating pairs were transferred to 0.5 L plastic cups to ensure that they were not disturbed by conspecifics in the cage. Upon separation, females were immediately frozen. Males were weighed, and transferred to holding cages containing host plant leaves and a nectaring source for the day between matings. After mating, A. crataegi males were brought down from the roof and put into holding cages in the laboratory. At the end of the day, females were removed from the freezer and the bursae copulatrix, containing the ejaculate, removed and weighed immediately on a Cahn electrobalance to the nearest 0.0001 g. The bursae were then transferred to individual plastic Eppendorph tubes and immediately put into the freezer, where they were stored at -20°C until protein analysis. A l l statistics were calculated using SYSTAT. Bartlett's test was used to determine heteroscadasticity and Lillifors test to assess normality. Non-parametric tests were used for non-normal data (Lillifors, p < 0.05) and for interspecific comparisons because of the disproportional nature of the data. Non-parametric tests were used only after the appropriate data transformations (ie. arcsine and natural logarithmic transformations) were found not to improve the normality of the data, assessed using Lillfors test (p>0.05). Results Virgin Matings Male eclosion mass differed significantly among all three species (Kruskall Wallis H = 33.85, df = 2, n= 40, p < 0.001) (Table 2.1). Across species, there was a significant difference in the mass (KW H = 23.13, df = 3, n = 39, p < 0.001) and protein content (mg) ( A N O V A F = 40.24, df = 2, n = 39, p < 0.001) of ejaculates produced by males mating for the first time. However intraspecifically, in first matings, there was no relationship between ejaculate mass and male eclosion mass (P. rapae r = 0.35, n = 11, p = 0.30; A. crataegi r = 0.02, n = 18, p = 0.93; P. brassicae r = 0.35, n = 10, p = 0.33), because body size was controlled for, and only males of mean eclosion mass were used (Table 2.1). There was no relationship between ejaculate mass and protein content (mg) of ejaculates produced by P. rapae and P. brassicae males mating for the first time (P. rapae r = 0.17, n = 11, p = 0.63; P. brassicae r = 0.34, n = 10, p = 0.33). However, in first matings, protein content (mg) increased with the size of ejaculate transferred by A. crataegi males (r = 0.68, n = 18, p = 0.002). Thus although males did not vary much in mass, some male A. crataegi appear to 30 Table 2.1. Eclosion mass of males and females. A l l values are listed as means ± SE. Numbers in parentheses represent samples sizes. Species Pieris brassicae Aporia crataegi Pieris rapae Male Eclosion Female Eclosion Mass (mg) Mass (mg) 166.77 ± 2 . 1 7 185.96 ± 3 . 1 8 (10) (19) 114.99 ±2.62 161.99 ± 5 . 5 8 (18) (31) 77.86+ 1.76 80.24 ± 1.06 (11) (31) 31 Vi Vi 2 QJ w > •»M Mating Number Vi 2 31 m e 2 • A. crataegi • P. brassicae O P. rapae X O Mating Number Figure 2.1. The effect of male mating history on reproductive investment in three successive ejaculates, in terms of a) ejaculate mass relative to male body mass (%), and b) ejaculate protein mass relative to ejaculate mass (%), transferred by males of P. brassicae, A. crataegi and P. rapae in their first three matings. Values shown for each species represent the mean 32 invest more, than other males of the same body mass. To compare male investment in successive ejaculates across species I used relative ejaculate mass (ejaculate mass / eclosion mass) and relative protein content (protein content (mg) / ejaculate mass) to control for differences among species in male body and ejaculate size. There was a significant difference among the three species in the relative size of ejaculate transferred in matings 1-3 (Mating 1 K W H = 24.09, df = 2, n = 39, p < 0.001; Mating 2 K W H = 11.65, df = 2, n = 27, p = 0.003; Mating 3 K W H = 8.30, df = 2, n = 15, p = 0.016). Although A. crataegi invested the most in first matings (10.9 %), P. rapae produced relatively larger ejaculates in second and third matings; with P. brassicae making the relatively smallest investment in all three matings (Fig 2.1a). Relative ejaculate mass decreased in the second and third matings by A. crataegi and P. brassicae males, whereas P. rapae males invested a constant proportion of their body mass (8.0 %) in all three matings (Fig. 2.1a). Relative protein content (%) (protein content of ejaculates (mg) / ejaculate mass (mg)) differed among species only in the first mating, with P. rapae producing ejaculates containing proportionally the most protein (3.2%) and P. brassicae the least (2.3%) (KW H = 17.13, df = 2, n = 39, p < 0.001). Protein content (%) decreased in second matings, but did not differ significantly among species in second and third matings (Fig. 2.1b) (Mating 2 K W H = 4.93,df=2, n= 15, p = 0.08; Mating 3 K W H = 2.36, d f = 2 , n = 15, p = 0.31). In all three species, protein content (%) was higher in ejaculates transferred in third matings than those in second matings (Fig. 2.1b) For each species the total investment by individual males, mating 1 to 3 times, was added to obtain the cumulative relative investment in ejaculates in terms of body mass (ejaculate mass / eclosion mass) and protein content (protein content of ejaculates (mg) / ejaculate mass) in terms of ejaculate mass. Both cumulative relative ejaculate mass ( A N O V A F = 18.6, df = 2, n = 40, p < 0.001) and protein investment (KW H = 18.5, df = 2, n = 40, p < 0.001) differed significantly between all three species, and increased with the degree of polyandry (Fig. 2.2a and 2.2b respectively). Most male P. brassicae and A. crataegi were reluctant to mate more than twice. Within species, sample sizes for the second and third matings were less than the number of males that mated for the first time as the number of males willing to remate decreased. Male capability to remate varied with the degree of polyandry; in P. rapae all of the 11 males 33 mated at least twice and 9 mated 3 times; in A. crataegi 10 out of 18 males mated at least twice and 3 mated 3 times; in P. brassicae 6 out of 10 males mated at least twice and 3 mated 3 times. A l l male P. rapae and A. crataegi given the opportunity mated for the first time; 3/13 male P. brassicae given the opportunity did not mate at all. Within each species, males that mated three times did not differ in size from males that mated only once (P. rapae t= -0.48, df = 18, p = 0.63; Separate variance t-test A. crataegi t= -0.96, df = 2.4, p = 0.42; P. brassicae t =-0.05, df= 3.1, p = 0.96). In the most polyandrous species, P. rapae, there was no effect of mating history on the mass or protein content (mg) of ejaculates (Fig. 2.1). P. rapae males transferred the same mass of ejaculate protein in all three matings although copula duration did increase slightly (Repeated measures A N O V A F = 0.42, df = 2, n = 9, p = 0.66; Fig. 2.3a). Ejaculate mass decreased with mating frequency for P. brassicae and A. crataegi (Fig. 2.1a), but the actual protein content (mg) of these ejaculates remained relatively constant in P. brassicae (Friedman test statistic = 0.00, df = 2, n = 2, p = 1.0) and A. crataegi (Friedman test statistic = 4.67, df = 2, n = 3, p = 0.09) in second and third matings, although males took much longer to complete copulation (Fig. 2.3b, c). Discussion Male ability to produce more than one large, nutritious ejaculates was correlated with the propensity of females to remate. M y results show that ejaculates contained 2.3 - 3.6 % protein and that, given a limited time between matings, only males in the most polyandrous system, P. rapae, were able to produce three successive ejaculates of similar mass (Fig. 2.1). Furthermore, the cumulative relative mass of ejaculate protein increased with the degree of polyandry (Fig. 2.3). In six other species of butterfly, Svard and Wiklund (1989) also found that only the polyandrous species produced subsequent ejaculates, similar in mass and sperm content to that transferred by males in their first mating. Males in polyandrous systems may be able to produce more than one high quality ejaculate because they invest proportionally more of their resources, relative to their body mass, in reproduction than males in monandrous species (Karlsson 1995). As found for other monandrous species (Svard 1985; Svard and Wiklund 1986, 1989), in subsequent matings P. brassicae and A. crataegi males never produced another ejaculate as large as that transferred by males mating for the first time (Fig. 2.1). M y results suggest that male capacity to produce large, nutritious ejaculates is limited in monandrous species. Both P. brassicae and A. crataegi males produced their largest ejaculates, containing the most protein (mg) (absolute content), in their first mating. a> > to I E •3 u 2.5 Degree of Polyandry •1.0 1.5 2.0 2.5 Degree of Polyandry Figure 2.2. Total investment by individual male P. brassicae, A. crataegi and P. rapae mating 1 to 3 times, measured as the cumulative investment in a) ejaculates (ejaculate mass (mg) / eclosion mass (mg) x 100) relative to male body mass b) protein (protein content (mg) / ejaculate mass (mg) x 100) relative to ejaculate mass, in relation to the degree of polyandry (measured as the mean number of spermatophores found in wild-caught females of the three species). 0 . 4 E ^ 0.3r- c c o u 0.2 h c 4-* 2 o.o O 5 0 Copula l O O Duration 1 5 0 (min) bo cu c o u a 3: 2 o.o O l O O Copula 2 0 0 3 0 0 Duration 4 0 0 5 0 0 (min) 0.5 o.4 h 0.3 h 0.2 h c *0) -| C O u c cu e OH O . l h o.o O 5 0 Copula l O O I S O Duration 2 0 0 2SO (min) Figure 2.3. The relationship between protein mass (mg) in ejaculates and copula duration in first, second and third matings transferred by individual males of P. brassicae, A. crataegi and P. rapae. 36 Thus in systems where females generally mate only once, males invest heavily in first matings. However, males in polyandrous systems have the ability to recuperate quickly from a mating event and thus appear to be better adapted to mating repeatedly. In several species of polyandrous butterfly, the duration of the female's refractory period is positively correlated with the mass of ejaculate received (Labine 1964; Obara et al. 1975; Sugawara 1979; Rutowski 1980; Rutowski etal. 1981; Oberhauser 1989; Wiklund and Kaitala 1995). The risk of sperm competition increases with female mating frequency. By transferring a large ejaculate, males in polyandrous systems can increase their fertilization success both in sperm competition (Chapter 4) and by prolonging the duration of the female refractory period , although the benefits to males arising from delaying females remating will depend on the duration of the delay, and her daily fecundity (Oberhauser 1989). However, the production of an ejaculate does not come without cost to even the most polyandrous species. In all species, copulation durations were longer in second and third matings (Fig. 2.3), which suggests that copulation may deplete a male's immediate access to accessory gland products, such as protein (Oberhauser 1988). Rutowski et al. (1983) concluded that males in several species of butterflies invest similar proportions of their body mass. However, I found a wide variation in both the mass and protein content of ejaculates. The mass of ejaculates transferred in first matings ranged from 10.9% (A. crataegi) to 7.6% (P. brassicae). Protein content of first ejaculates ranged from (3.2%), found in P. rapae ejaculates, to 2.3% in P. brassicae ejaculates. Oberhauser (1992) proposed that upon remating males should transfer undigestible ejaculates containing more water and less nutrients to decrease the costs of remating and yet function to delay females from remating. Although the second and third ejaculates transferred by P. brassicae and A. crataegi males were smaller, the composition (in terms of proportion of protein) remained relatively constant across matings (Fig. 2.1b). These results suggest that a minimum amount of protein may have to be transferred, which may be the case if proteins are involved in other functions such sperm activation and oviposition (cf. Oberhauser 1992, 1996). Protein donation may also be viewed as a female means of preventing cheating, whereby fitness gains for male providing 'subpar' ejaculates are reduced. Recent evidence demonstrates that relative to males in more monandrous systems, males in polyandrous systems i) transfer relatively larger first ejaculates (Svard and Wiklund 1989; Chapter 1), ii) transfer ejaculates relatively richer in protein (Chapter 1), iii) recuperate faster and are able to transfer a second ejaculate as large as the first sooner (Svard and 37 Wiklund 1989), and iv) allocate proportionally more resources to reproduction, measured as abdomen relative to body mass (Karlsson 1995). Here I additionally demonstrate that polyandrous males are better able to i) maintain relative ejaculate mass (whereas males in the less polyandrous systems were unable to transfer a second ejaculate as large as the first) (Fig. 2.1), ii) donate a larger cumulative mass of nutrients relative to body mass (Fig. 2.2) over a number of matings. Males in polyandrous systems seem to have a number of adaptations that enhance their mating capabilities. In view of the fact that the most polyandrous species in this study, P. rapae , and the most monandrous species, P. brassicae, are closely related and basically share the same larval host plants (largely crops of Brassica spp. in the wild, and both were reared on Alliaria petiolata in my experiments), the evidence suggests that these different adaptations in mating and nuptial gift capacity are causally linked to the mating system. Males are investing paternally when ejaculates contain nutrients that benefit females or their offspring (Simmons and Parker 1989). Assuming that ejaculates function mainly as paternal investment and because paternal certainty is higher in monandrous species, Boggs (1981a) and Marshall (1982) suggested that the protein content of ejaculates should be highest in monandrous mating systems. However I found that in terms of both cumulative ejaculate mass and protein content, P. rapae males, representing the most polyandrous species, invested the most and P. brassicae, the most monandrous species, invested the least (Fig. 2.2). If females in polyandrous systems use mating to acquire male-derived nutrients (Walker 1980; Kaitala and Wiklund 1994), spermatophores transferred by males in polyandrous systems should contain more nutrients than those transferred by males in monandrous systems. In monandrous systems, where females mate on average only once, their fitness should not be as dependent on male-derived resources. Females in polyandrous systems may "expect" and depend more on male-derived resources in order to realize their reproductive potential (Oberhauser 1989, 1996; Leimar etal. 1994; Wiklund and Kaitala 1995; Karlsson 1995). In this study males in the most polyandrous species were able to produce more than one high quality ejaculate suggesting that females can use mating to procure male-derived resources (Walker 1980; Kaitala and Wiklund 1994). In polyandrous systems, although there is selection on males to transfer large ejaculates, females may have evolved to capitalize on this investment. Male ability to remate is correlated with the degree of polyandry, in spite of the fact that males in most polyandrous species transferred larger, higher quality ejaculates than males in monandrous species. Boggs (1995) suggests that from the male's perspective 38 nuptial gifts can function as both resource and mating investments, but to females, their sole function is as a nutrient resource. Nuptial gifts can function as both paternal investment and mating effort (influencing a male's fertilization success) as a result of selection on gift quality once male nutrients contribute significantly to female fitness (Simmons and Parker 1989). When male-derived nutrients are used to enhance female reproductive output, a reduction in nutrient quality may not be possible for males because of the effect on offspring fitness (Simmons and Parker 1989; Gwynne 1990; Simmons and Bailey 1990). Males in polyandrous systems appear to be able to produce more than one high quality ejaculate. Consequently polyandrous females can increase their access to these nutritional resources by remating. Males will benefit most by either producing undigestible ejaculates that will still function to delay female remating, or immediate use of donated nutrients, thereby minimizing the probability their resources are used to benefit another male's offspring when the female remates (Parker and Simmons 1989). Thus, there may be conflict between the sexes (Parker 1979) over female remating intervals, resulting from selection on females to acquire male-derived nutrients, and selection on males to minimize female remating in order to maximize their fertilization success. In Lepidoptera, ejaculates may: 1) affect a male's fertilization success, and 2) contain nutrients that allow females to realize their reproductive potential. In polyandrous species, selection on both sexes to maximize these benefits may explain why males have the ability to produce a number of large, nutritious, ejaculates. 39 CHAPTER 3 EFFECT OF M A L E M A T I N G HISTORY A N D B O D Y SIZE O N E J A C U L A T E SIZE A N D Q U A L I T Y IN TWO P O L Y G A M O U S BUTTERFLIES Introduction Female insects invest a significant proportion of nutritional resources in reproduction. However, males may also invest nutrients in reproduction. Male Lepidoptera transfer a spermatophore to females during copulation that contains sperm and accessory gland products. Because nutrients contained in spermatophores have been found in the eggs and soma of females (Boggs and Gilbert 1979; Boggs 1981a; Boggs and Watt 1981; Greenfield 1982; Boggs 1990; Wiklund etal. 1993), lepidopteran mating systems provide an opportunity to quantify male nutrient investment. In several species of Lepidoptera, the size of ejaculate size transferred in first matings is correlated with male body size (Boggs 1981a; Rutowski et al. 1983; Svard and Wiklund 1986, 1989; Oberhauser 1988; Royer and McNeil 1993; Wiklund and Kaitala 1995), and the length of female refractory period increases with increasing ejaculate size (Obara et al. 1975; Sugawara 1979; Rutowski 1980; Rutowski et al. 1981; Oberhauser 1989; Wiklund and Kaitala 1995). Moreover, in twice-mated Pieris napi females, the male transferring the larger ejaculate sires the majority offspring, regardless if mated first or second (Chapter 4). Consequently, males that transfer large ejaculates appear to be rewarded with increased paternity. Furthermore, because ejaculate size is correlated with male body size, large males may have an advantage over small males because females delay remating for longer when a large ejaculate is transferred. However, it is not known how the nutritional composition of ejaculates changes with ejaculate size. Several studies have used spermatophore mass to estimate the effect of male history on male investment in reproduction (Rutowski 1979; Sims 1979; Boggs 1981a; Rutowski et al. 1983; Svard 1985; Svard and Wiklund 1986, 1989; Oberhauser 1989). Although spermatophore mass does not reveal male nutrient investment in reproduction, few studies have attempted to determine the nutritional quality of spermatophores (for Lepidoptera, see Marshall 1985; Marshall and McNeil 1989; Oberhauser 1992 for exception; for Orthoptera see Wedell 1994). Therefore, to determine how males allocate nutrients to a mating event I chose two of the most polyandrous species in the family Pieridae, Pieris napi L . and Pieris rapae L. and examined the effect of male mating history on ejaculate size and quality. More specifically, because of the benefits accrued to males that produce large spermatophores, the 40 objective of this study was to determine: 1) the benefits and costs (in terms of the nutrient content (protein)) to females of receiving spermatophores from males of different mating histories and sizes, and 2) mating costs to males, concerning the recovery time needed to produce spermatophores of similar size and quality to that transferred in their first mating. Protein content of ejaculates was used as a measure of ejaculate quality for reasons stated in the general introduction and chapter 1. Materials and Methods Butterflies used in experiments were diapause-generation offspring produced by Pieris napi or Pieris rapae females wild-caught near Stockholm, Sweden. Larvae were reared in pairs, in 0.5 L plastic cups on their natural host plant, Alliaria petiolata (Cruciferae) in environmental chambers maintained at 23 and a 20h daylength. Offspring were overwintered as pupae in the laboratory. Males and females were weighed on an Cahn automatic electrobalance the day of eclosion and marked individually on their wings with a Staedler permanent pen. Butterflies were held in a cold room at 4^C until the start of the experiment. Experiments were started within 3 days of the first males and females to eclose. Matings were carried out in 0.8 x 0.8 x 0.5 m mating cages placed 2 m below 400 W HQIL lamps switched on from 0800 to 1800 h. Cages contained species-specific host plant leaves and a nectaring source, either wild flowers or a potted Chrysanthemum spp.. A 25% sugar solution was applied with a dropper into the middle of the flowers twice a day; first thing in the morning when the lights were turned on and again in the early afternoon. Males were introduced first into mating cages and allowed to feed for 1/2 hour after the lights were turned on and before virgin females were added. Cages were inspected every 15 minutes for mating pairs. Upon discovery, pairs in copula were put into 0.5 L plastic cups'so as not to be disturbed by conspecifics in the cage. Immediately upon separation, females were frozen. Males were weighed and transferred to holding cages where they were fed ad libitum until the day they were to be remated. At the end of the day, females were removed from the freezer, and the bursa copulatrix containing the ejaculate dissected out. After quickly scrapping away any somatic tissues, the bursa was weighted to the nearest 0.0001 g. on an electrobalance and put into individual plastic Eppendorph tubes. The tubes containing the ejaculate were immediately put into a freezer where they were stored at -20°C until protein analysis. Protein content of 41 ejaculates was estimated using the protein assay (Bio-Rad) described in the general introduction. Effect of Mating History on Protein Content To minimize the effect of male body size on ejaculate size and protein content, only males of mean eclosion mass were used (P. napi = 69.39 + 0.72, n=37; P. rapae = 75.63 ± 0.67, n=29). Because so many females were needed for this experiment, female eclosion mass was more variable as I was unable to exclude large females. To examine the effect of mating history on protein content of ejaculates, in both species each male was mated once and then remated either: the same day (Treatment 0), one day later (Treatment 1), two days (Treatment 2), or 3 days later (Treatment 3). Virgin females were used for all matings. Because few males are willing to mate twice in one day, all males mated for the first time were given the opportunity to remate the same day as their first mating. Males not remating the same day were then randomly assigned to one of the three remaining treatment groups. These males were held in mating cages without females and fed ad libitum until the day they were to be remated. On the days males were remated, females were added after the lights were turned on and males had fed for 1/2 hour. Effect of Male Size on Protein Content To examine the effect of male body size on protein content and recovery rate of ejaculates, P. napi males were sorted into size categories as pupae, reweighed and marked individually the day of eclosion. Eclosed adults were held in the cold room at 4°C until the start of the experiment. The experiment was started within 7 days of the first butterflies to eclose. Virgin females were added to mating cages containing a) large males (eclosion mass = 82.41 + 0.55 mg) or b) small males (eclosion mass = 55.04 + 0.81). Most P. napi males are willing to remate after 24 hours (pers. observ.). Therefore after mating once, males were held for one day and allowed to feed ad libitum. The following morning, virgin females were added to mating cages and males allowed to remate. A l l statistics were calculated using SYSTAT. Bartlett's test was used to determine heteroscadasticity and Lillifors test to assess normality. Non-parametric tests were used for non-normal data (Lillifors, p < 0.05) and for interspecific comparisons because of the disproportional nature of the data. Non-parametric tests were used only after the appropriate data transformations (ie. arcsine and natural logarithmic transformations) were found not to improve the normality of the data, assessed using Lillfors test (p>0.05). 42 Results Male body size did not differ among treatments (day) in P. napi ( A N O V A F =1.83, df 3,29, p > 0.10), but did in P. rapae ( A N O V A F = 3.69, df 3,21, p = 0.028), and thus was analyzed as a covariate when necessary. Among males in different treatments (day), analysis of variance revealed no difference in ejaculate mass [ A N O V A P. napi F(day) = 0.134, df 3,29; A N C O V A P. rapae F(day) = 2.45, df 3,20; F(male size) = 2.53, df 1,20; all values p > 0.10] or protein content (mg) [P. napi A N O V A F(day) = 0.16, df 3,29; P. rapae A N C O V A F(day) = 1.34, df 3,20; F(male size) = 1.35, df 1,20; all values p > 0.10) of first ejaculates. Female eclosion mass was significantly and positively correlated with protein content of ejaculates transferred by P. napi males mating for the first time (Fig. 3.1). Female eclosion mass had no effect on ejaculate size in either species, or protein content of ejaculates produced by P. rapae males in their first or second mating. Effect of Male Mating History on Ejaculate Mass and Protein Content Repeated-measures A N O V A revealed that the number of days between matings significantly affected both the mass and protein content of second ejaculates in both species (Table 3.1, Fig. 3.2). Ejaculates produced by males remated the same day were the smallest and contained the least amount of protein. The amount of protein and size of ejaculate increased as the number of days between matings increased (Fig. 3.2). Males remated three days after their first mating produced larger ejaculates that contained absolutely more protein. A l l males, except those remated the same day, transferred approximately equal proportions of protein (Table 3.1). The proportion of protein was significantly higher in ejaculates produced by males that mated twice in one day (Table 3.1). Mating history significantly affected the relative investment (defined as ejaculate mass / eclosion mass) in second ejaculates ( A N O V A Pieris napi F = 43.64, df 3, 27, p > 0.001), P. rapae F = 63.58, df 3, 21, p < 0.001). P. napi and P. rapae males delivered first ejaculates that corresponded to an average of 11.4 ± 0.3 % (n = 37) and 8.3 ± 0.3 % (n = 29) of their eclosion mass respectively. Relative investment was lowest in males remated the same day, and increased as the number of days between matings increased (Table 3.1). P. napi males held for two and three days after their first mating delivered ejaculates that represented 13.0 % of their eclosion weight, much more than that invested by males mating for the first time (Table 3.1). The interval between matings affected investment in ejaculates 43 Table 3 . 1 . C o m p a r i s o n o f ejaculates p r o d u c e d b y males remated the same d a y ( 0 ) , the next day ( 1 ) , t w o days later (2), three days later (3). A l l values are means + 1 S E . A D u n n e t t test was used t o detect differences a m o n g the treatments (day remated) f o r a l l parameters except m a t i n g duration. A W i l c o x o n paired-sample test was used t o test f o r s i g n i f i c a n t differences between the first vs. second c o p u l a t i o n f o r each treatment. Letters denote differences • between first and second matings, p < 0.05; w h e n letters are the same p > 0 . 1 . * p = 0.07. Relative ejaculate mass = ejaculate mass ( m g ) / eclosion mass ( m g ) x 100. Second M a t i n g First Mating 0 7.86 + 2.81 + 5.51 + 8.95 + 0.19 0.84° 0.31 0.29 * 9.16± 0.51 0.12 ± 0.02b 0.16 ± 0.00 0.21 ± 0.24 ± 0.01 0.01 1 3 2 Pieris napi Ejaculate M a s s ( m g ) a d a c Protein Content (mg) 0.22 ± Percent Protein 2.78 ± 4.56 ± 3.01 ± 2.32 ± 2.60 ± 0.06 0.58 0.23 0.10 0.08 0.00 a a c b a R e l a t i v e Ejaculate M a s s (%) 11.4 + 0 . 3 Mating Duration (min.) 82.11 ± 3.33 699.25 ± 178.1 ± 261.94 * 14.23 n 37 4 10 a a 4.2 ± 1 . 3 d a 7.8 + 0.4° a a 13.4±0.4 117.11 ± 6.08 c d a a 13.0 + 0 . 7 118.63 ± 6.73 d d 8 9 Pieris rapae Ejaculate Mass ( m g ) 6.31 ± 2.82 ± 4.52 ± 6.56 ± 6.40 ± 0.20 0.12 0.09 0.28 0.32 a b c 0.21 ± 3.40 ± 3.01 ± 3.36 + 0.15 0.19 0.50 0.21 ± 0.13 ± 0.15 ± 0.01 0.01° 0.01 Percent Protein 3.31 ± 4.50 ± 0.08 0.33b a b a 0.19 ± 0.01 P r o t e i n Content (mg) a a 0.02 a c a a a a R e l a t i v e Ejaculate Mass (%) 8.3±0.3 Mating Duration (min.) 58.59 ± 413.86 ± 91.33 + 77.88 + 77.5 ± 3.11 34.19 5.67 5.07 12.89 n 29 a a 3.7 ± 0 . 2 7 b b 6.2 + 0 . 2 6 c C 8.3±0.3 8 d a 8.3 ± 0 . 4 4 a a d 44 E 3 0.30 0.2 6 h u 5T tt 0.22H c c 70 80 90 Female Body Mass (mg) Figure 3.1. The relationship between protein content (mg) of first ejaculates transferred by P. napi males and female body mass at eclosion (r = 0.38, p = 0.022, n = 37). 45 P. rapae P. napi 10 lO- 8H oo 00 6i VI 3 S a 3 « i iff O 1st Mating • 2nd Mating a) i • 0 1 ' 1 2 I ' 0 • i • 1 2 •3 * 2 uT z d) O 1st Mating • 2nd Mating I 3 0 3 0 No. Days between Matings T — 41 1 2 3 No. Days between Matings r No. Days between Matings 1 2 3 No. Days between Matings Figure 3.2. The effect of male mating history on the a, d) mass, b, e) protein content and c, f) proportion of protein (•% wet weight) of ejaculates produced by Pieris napi and P. rapae males respectively. Males were mated for the first time (open circles) and then remated either the same day (0), the next day (1), two days later (2) or three days later (3) (closed circles). A l l values shown are means ± 1SE. differently in P. rapae males. P. rapae males held for two and three days delivered ejaculates corresponding to the same proportion of their eclosion mass (8.3 %) as males mating for the first time. The number of days between matings significantly affected copulation duration of second matings (P. napi, Kruskall-Wallis A N O V A H = 18.80, df = 3, p < 0.001; P. rapae, Kruskall-Wallis A N O V A H = 15.45, df = 3, p = 0.001). In both species, copulations involving males remated the same day lasted the longest (Table 3.1). Mating duration decreased the longer the interval between matings. By day 3, copulation durations did not differ significantly from first matings for P. rapae (Table 3.1). However, copulations by P. napi males remated three days after their first mating were still significantly longer than first matings (Table 3.1). Effect of Male Body Size on Ejaculate Size and Quality Large P. napi males mating for the first time produced significantly larger ejaculates than first ejaculates produced by small males (Table 3.2). There was no difference in the protein content (mg) of first ejaculates (Separate variance t-test T = 1.83, df = 9.3, p = 0.10) or in the copulation duration of first matings (Mann-Whitney U T = 29.5, n = 18, p = 0.33) between large and small males (Table 3.2). When males were remated two days later, there was no affect of male size or copulation duration (analyzed as a covariate) on either ejaculate mass or protein content of second ejaculates (Fig. 3.3). The mass of the second ejaculate (analyzed as a second covariate) also had no affect on the protein content of second ejaculates (Fig. 3.3). In both cases, there was no interaction between the treatment and covariate (p>0.1). However, there was a negative relationship between male eclosion mass and relative ejaculate size (ejaculate mass / eclosion mass); relative to their body size, small males produced larger ejaculates in their first and second matings (Table 3.2; Fig. 3.3b). Moreover, the protein content of ejaculates produced by small males mating for the first time was much more variable than that produced by large males, with some ejaculates containing as much protein as that contained in ejaculates produced by larger males (Fig. 3a). Thus ejaculates from small males contain more protein relative to the size of ejaculate transferred. Within each size class, males remated two days later remained in copula for much longer, and transferred significantly smaller ejaculates that contained significantly less protein than males mating for the first time (Table 3.2). 47 Table 3.2. Comparison of first and second ejaculates produced by large and small P. napi males. 1 = First Mating, 2 = Second Mating. Males were remated two days after their first mating. Statistical comparisons were performed using a t-test (separate variances) between groups and a paired-sample t-test within each size class. Relative ejaculate mass = ejaculate mass (mg) / male eclosion mass (mg) x 100. *Wilcoxon paired-sample test used. 1 2 T df P 84.0 ± 8 . 3 8.9 ± 0 . 4 10.9 ± 0 . 5 0.22 203.9 ± 3 0 . 0 4.6 ± 0 . 3 5.6 ± 0 . 4 0.14 -3.57 13.48 13.65 -2.67 8 8 8 0.01 <0.01 <0.01 <0.01 117.8 7.7 ± 0.4 14.0 ± 0 . 7 0.19 ± 0 . 0 1 194.8 4.3 ± 0.2 7.8 ± 0 . 3 0.14 ± 0 . 0 1 2.10 8.86 8.52 4.01 7 7 7 0.04 <0.01 <0.01 <0.01 2.31 -3.75 -4.04 16 14.8 14.9 0.03 <0.01 <0.01 Large Males Mating Duration 1 vs. 2 (min.) Ejaculate Mass 1 vs. 2 (mg) Relative Ejaculate Mass (%) Protein Content 1 vs. 2 (mg)* Small Males Mating Duration 1 vs. 2 (min.)* Ejaculate Mass 1 vs. 2 (mg) Relative Ejaculate Mass (%) Protein Content 1 vs. 2 (mg) Large vs Small Males Ejaculate Mass 1 Relative Ejaculate Mass 1 Relative Ejaculate Mass 2 0.25 OO a 0.20 3 n u A w 0.15 c C/5 C/J c I 0.10 0.05 0.00 • Large males O Small Males a) 1 1 Mating 0.16 • 2 Number • on 0.12 0.08 - ^ 3 21 0.04 b) 0.00 • 1 Mating 2 Number Figure 3.3. Protein content (a) and relative ejaculate mass (ejaculate mass / male body mass x 100) (b), transferred by small males (open circles) and large males (closed circles) in their first and second matings. Males were remated two days after their first mating. Analysis of covariance revealed no significant effect of male body size or copulation duration on the mass [Ffbody size) = 0.70, df 1,14; F(copula duration) = 0.43, df 1,14; p > 0.42 for both variables] or protein content of second ejaculates [F(body size) = 0.24, df 1,13; F(copula duration) = 0.24, df 1,13; p > 0.63 for both variables; F(ejaculate mass) = 4.07, df 1,13, p = 0.065]. All values shown are means ± 1SE. 49 Discussion Ejaculates transferred by male Lepidoptera contain protein, which is beneficial to females. However, ejaculates are physiologically costly for males to produce as evidenced by the fact that in Pieris napi and P. rapae, copulation durations were longer and ejaculates smaller in matings involving recently mated males. The extent to which copulation duration, ejaculate mass and protein content changed was influenced by the interval between matings. M y results suggest that copulation depletes a male's reserves of ejaculate constituents. Others have found similar effects of male mating history on ejaculate mass and copulation duration in several species of Lepidoptera (Rutowski 1979; Sims 1979; Boggs 1981a; Svard 1985; Svard and Wiklund 1986, 1989; Oberhauser 1988, 1992; He and Tsubaki 1992; Royer and McNeil 1993). In a less polyandrous species, Papilio machaon (mean number spermatophores / female 1.16, Svard and Wiklund 1989), the mass of spermatophores delivered by males upon remating was always less than the first, regardless of the number of days between matings (Svard and Wiklund 1986). Svard and Wiklund (1986) found that the mass of accessory substances was mainly responsible for this difference. In a monandrous species, Pararge aegeria (mean number spermatophores / female 1.00, Svard and Wiklund 1989), Svard (1985) found that it took at least 6 days for males to make an investment similar to that of first matings. However, males belonging to the two polyandrous species studied here (average number of spermatophores / female 2.08, Svard & Wiklund 1989) have the ability to completely recover, gradually producing spermatophores of similar size and quality to that transferred in first matings after only two days. Oberhauser (1992) found a similar pattern of depletion and recovery for both mass and nitrogen content of monarch butterfly ejaculates. The increase in both ejaculate mass and protein content over time suggests that males need time to mobilize and/or replenish ejaculate constituents, corroborating the hypothesis that ejaculates are costly for males to produce (Dewsbury 1982; Svard 1985; Fitzpatrick and McNeil 1989). Interestingly, in both species, although males remated the same day transferred smaller spermatophores, they contained proportionally more protein, which decreased gradually to that found in ejaculates transferred in first matings (Fig. 3.2). From the female's perspective this means that spermatophores from recently mated males, and those produced by relatively small males, contain more protein that would be expected on the basis of spermatophore mass alone. The increase protein content of ejaculates produced by recently mated P. napi and P. rapae males is opposite to that found in the monarch butterfly, {Danaus plexippus), where Oberhauser (1992) showed that recently mated males delivered 50 ejaculates that were both smaller and contained proportionally less nitrogen. Although nitrogen is not the same as protein, it is unclear as to why investment patterns between these three polyandrous species differed. Rutowski (1979) found that male Pieris protodice were reluctant to mate twice in one day and suggests that males are behaving 'honestly', with male courtship persistence being positively correlated with the size of ejaculate a male can deliver. This term is slightly misleading as the only time P. napi and P. rapae males appeared to behave 'honestly' was when they had mated previously that same day (only 4/18 P. napi and 7/17 P. rapae remated twice in one day). However other than the same day, in the remaining treatments (13), given the opportunity, only 6/37 and 4/29 male P. napi and P.rapae didn't remate. As shown here, and in other species (Svard 1985; Oberhauser 1988; Svard and Wiklund 1989), males do not delay remating until they can produce large ejaculates, which they should if they are behaving honestly. This conclusion is based on the assumption that male propensity to remate in the laboratory is similar to that in the wild. Although mating cages were not crowded, it is possible that butterfly densities or experimental conditions may have made males more likely to remate. Nutrients accrued through copulation can be used by females for egg production and/or somatic maintenance (Boggs and Gilbert 1979; Boggs 1981a; Boggs and Watt 1981; Boggs 1990; Wiklund et al. 1993). In some species, the reproductive performance of females is dependent on male-derived nutrients, as evidenced by the fact that multiply mated females have a higher reproductive output and/or longevity than females allowed to mate only once (Watanabe 1988; Oberhauser 1989; Wiklund et al. 1993; Watanabe and Ando 1993). In fact, male-derived nutrients may have a greater impact on the reproductive potential of females than female body size, which is only weakly positively correlated with fecundity in P. napi (Wiklund & Kaitala 1995). Furthermore, I found that female size had a positive effect on the amount of protein contained in ejaculates transferred by P. napi males mating for the first time (Fig. 3.1). Ejaculate mass was not affected, which suggests that males may be providing more resources to larger females. However for females, repeat mating can be costly. Males that have mated previously the same day or the day before remain in copula for much longer, and transfer smaller spermatophores that contain less protein (mg) (Table 3.1). Lengthy copulations prohibit females from foraging and/or oviposition. Thus one would expect selection to favour females that discriminate against recently mated males. However, P. napi females show no 51 preference for virgin males (Kaitala and Wiklund 1995). This lack of discrimination suggests that females can not detect a male's mating history. In several species, ejaculate mass is positively correlated with both male body size (Boggs 1981a; Rutowski etal. 1983; Svard and Wiklund 1986, 1989; Jones etal. 1986; Oberhauser 1988) and the duration of female unreceptivity (Labine 1964; Obara et al. 1975; Sugawara 1979; Rutowski 1980; Rutowski etal. 1981; Oberhauser 1989; Wiklund and Kaitala 1995). The positive correlation between male eclosion mass and ejaculate mass suggests that larval reserves may influence the potential size of ejaculate a male can produce (Boggs 1981b). Wickman and Karlsson (1989) suggested that when reserves collected by larvae limit reproduction, the proportional increase of reserves with body size should be paralleled by an increase in reproductive effort, possibly resulting in increased ejaculate production by larger males. In P. napi, I found a significant relationship between male eclosion weight and mass of first ejaculates. However, there was no effect of body size on either the absolute mass or protein content of second ejaculates produced by males remated two days later. In fact, small males transferred proportionally larger ejaculates in their first and second matings. A negative correlation between male eclosion mass and relative ejaculate mass has also been observed for first matings in P. protodice (Rutowski 1984) and in another study of P. napi (Wiklund and Kaitala 1995). I found that the protein content of ejaculates produced by small males mating for the first time was much more variable than that produced by large males (Fig. 3.3a), with some ejaculates containing as much protein as that contained in ejaculates produced by larger males. Hence, male reproductive investment does not necessarily increase with male size or ejaculate size. This may be one reason why some studies (Greenfield 1982; Jones et al. 1986; Svard and Wiklund 1988; Royer and McNeil 1993; Oberhauser 1989) have found no correlation between female reproductive output and male ejaculate mass or body size. In this study I have shown how male mating history and body size affect ejaculate quality. M y results demonstrate that mating can be costly to both sexes. Females mating to recently mated males remain in copula for extended periods of time, and receive small ejaculates. For males, copulation depletes a male's immediate access to accessory gland products, such as protein. It takes time for males to recover from a mating event. Males in the two polyandrous species studied here seem to need only two days between matings to recover. By the second day males appeared to have fully "recovered", producing ejaculates of the same mass and containing the same amount of protein as spermatophores produced by males mating for the first time. Furthermore, because males do not delay remating until they can produce a large ejaculate, males may maximize their fitness by remating as often as possible (Darwin 1871; Bateman 1948; Oberhauser 1988). In P. napi, small males produce ejaculates containing as much protein (mg) as ejaculates produced by much larger males, and on the basis of size, proportionally more protein (%). In terms of access to protein, there is potentially no selective advantage gained by females that prefer to mate with relatively large males. Moreover, although male size influences the mass of ejaculates produced by virgin males, size seems to have no effect on subsequent ejaculates. Thus there seems to be little reason why females should discriminate among males on the basis of size. In fact, being choosy may result in lost mating opportunities and in turn, a loss of access to additional resources. Thus, the greatest cost of not being choosy may be in terms of time, which although substantial i f the male has mated recently, may be partially offset by procured nutritional benefits. Although male mass affects the size of ejaculate transferred by virgin males, it is male mating history that greatly affects the protein content of ejaculates. The lack of evidence for active female choice suggests that recently mated males are able to successfully conceal their mating history. 53 C H A P T E R 4 EFFECT OF M A L E B O D Y SIZE O N S P E R M P R E C E D E N C E IN PIERIS NAPI L . Introduction When females mate multiply, sperm from different males can be in direct competition to fertilize eggs (Parker 1970). Several studies in the last decade have revealed an increasing diversity of taxa in which sperm competition has had significant affect on the evolution of mating systems (Parker 1970; see Smith 1984). Polyandry is common in Lepidoptera, and although in general the last male sires the majority of offspring ( P 2 ) , in many species there is considerable individual variation among males in the proportion of offspring sired by the last mate (Labine 1966; Brower 1975; Walker 1980; Drummond 1984). In classic studies of sperm competition, the affect of mating order on the outcome of sperm competition was realized, but recently, the influence of male size has been shown to be important as well, with large males having an advantage over smaller males in sperm competition in insect groups as diverse as beetles (Lewis and Austad 1990), flies (Simmons and Parker 1992), and arctiid moths (LaMunyon and Eisner 1993). Male Lepiodoptera produce a spermatophore during copulation that contains both sperm and accessory gland products. Nutrients passed to the female during copulation may represent a sizable and costly investment by the male (Svard and Wiklund 1986, 1989; Chapter 2 and 3). When ejaculates contain nutrients as well as sperm, they may function as paternal investment and/or male mating effort (Simmons and Parker 1989). In some polyandrous butterflies, spermatophores function as both paternal investment and male mating effort (Chapter 1). Spermatophores represent male mating effort because they affect the duration of the female refractory period following copulation which increases with ejaculate size (Labine 1964; Obara et al. 1975; Sugawara 1979; Rutowski 1980; Rutowski et al. 1981; Oberhauser 1989; Kaitala and Wiklund 1994; Wiklund and Kaitala 1995), resulting in sperm from males transferring larger ejaculates fertilizing a greater proportion of the available eggs. In some species of butterflies, spermatophores function as paternal investment because nutrients contained in these packaged ejaculates are used by females to increase their longevity and lifetime reproductive output (Boggs and Gilbert 1979; Boggs 1981a; Boggs and Watt 1981; Watanabe 1988; Oberhauser 1989; Boggs 1990; Wiklund etal. 1993; Watanabe and Ando 1993). Although several studies have examined how ejaculates affect female remating frequency and reproductive output, few have examined the affect of sperm competition on males investment in ejaculates (for exceptions see LaMunyon and Eisner 1993; Gage 1994, 1995, Cook and Gage 1995). 54 In most invertebrates, females are usually larger than males, but in butterflies sexual size dimorphism increases with the degree of polyandry in the families Pieridae and Satyridae, with males being larger than females in three of the four most polyandrous species (Wiklund and Forsberg 1991). Thus sexual size dimorphism appears to be reversed to that found in general for insects. The rationale used to explain why, across species, relative male size increases with female polygamy has been coupled to the positive association between ejaculate size and male size found in several butterfly species (Svard 1985; Svard and Wiklund 1986; Oberhauser 1988; Wiklund and Kaitala 1995). Because large ejaculates are beneficial both from a mating effort function (delaying female remating) and from a paternal investment function (increasing female reproductive output of offspring sired by the donating male), large males have a two-fold advantage over smaller males. In this chapter, I investigate if large males have yet a third advantage over smaller males, by assessing the influence of male size on the outcome of sperm competition. To determine the effect of male body size on sperm precedence, I examined the pattern of sperm utilization in doubly mated females of the green-veined white butterfly, Pieris napi L.. Sperm competition should be common in this strongly polyandrous species, as females store sperm and mate at least twice, and up to five times in their lifetime (Forsberg and Wiklund 1989; Kaitala and Wiklund 1994). The size of ejaculate transferred by virgin males is strongly correlated with male body mass (Wiklund and Kaitala 1995). Therefore I used male body mass as an indicator of the size of ejaculate transferred. In this study sperm precedence is defined as the nonrandom differential fertilization success among mating males and designates the male that obtains the majority of matings (Parker 1970; Lewis and Austad 1990). Materials and Methods The experiment was conducted twice, once in 1992 and once in 1993. In both years offspring were produced by Pieris napi L . females wild-caught in the vicinity of Stockholm, Sweden. Larvae were reared on their natural host plant, Alliaria petiolata L. (Cruciferae). In 1992,1 used diapause-generation butterflies that had overwintered as pupae in the laboratory. In 1993 the experiment was repeated with reciprocal crosses and control groups using directly-developed P. napi. In 1992, adult males were sterilized with 40 krads of gamma radiation using a Gammacell 1000. In 1993, the dose was reduced to 25 krads of gamma radiation to increase the probability that sperm from normal and irradiated males 55 were equally competitive. Experiments showed that dosages lower than 25 krads did not result in sufficient male sterility. The pattern of sperm utilization was investigated in doubly mated females using the standard sterile male technique of Parker (1970). Females were mated to two males, either large (L) or small (S), normal (N) or sterilized by irradiation (I) in consecutive order. In 1992 the mean size of large males was 74.0 mg, that of small males 59.0 mg, and that of females, 74.3 mg; in 1993 the corresponding values were 66.8 mg for large males, 53.3 mg for small males and 64.1 mg for females. Males from the two size classes were randomly assigned to normal or irradiated treatments. Females were randomly assigned to treatment and control groups. In 1992, 5 mating sequences were used, 3 controls (LILI, SISN, LILN) and 2 treatments (LISN, SILN) (Table 4.1). In 1993, eight mating sequences were used, 4 controls (II, N N , L I L N , SISN), and 4 treatments (LISN, LNSI, SILN, SNLI) (Table 4.1). The proportion of eggs fertilized by the normal male (x) was estimated from the proportion of eggs hatching (a) using the mean viabilities in normal 'NN' (b) and sterile TF (c) matings using the formula x = (a-c) / (b-c) (Sillen-Tullberg 1981). The proportion of progeny sired by each male could then be estimated. Values for paternity greater than one were adjusted to one, and negative values were adjusted to zero. This correction factor could not be used in 1992 because all control matings were not done. However, 40 krads appeared to result in 100% male sterility as none of the eggs produced by females doubly mated to irradiated males hatched. Thus, in 1992, the proportion of fertilizations obtained by the 'normal' male was estimated from the number of eggs that hatched. 1992 On the day of eclosion, males were weighed, separated into size classes based on their eclosion weight and numbered on their wings with a Staedler permanent pen. Pupae were not weighed. Females were also weighed the day of eclosion and numbered. To minimize possible differences in fecundity related to female size, only females of mean eclosion mass (74.3 mg.) were used. Butterflies eclosed within 7 days of the start of the experiment and were held in a cold room at 4 °C until they were used for mating. A l l males used for matings were virgins. Virgin males were irradiated with 40 krads of gamma radiation the day females were mated for the first time, and two days later when they were mated for the second time. Males were irradiated first thing in the morning, just prior to being put into mating cages. Lights were turned on at 0800 h and males allowed to feed for 1/2 hr. before females were added. 56 1993 In 1993, individuals were sexed and weighed as pupae, and as newly eclosed adults. Males were separated into size categories based on their pupal weights. Although both eclosion weight and pupal weight are correlated with adult body mass, pupal mass is more strongly correlated with adult wing length. Only females of mean pupal mass (145.0 ± 1 . 5 mg) were used to minimize possible differences in fecundity associated with female size. Adults eclosed over a 9 day period and were held in a cold room at 4°C until they were used for matings. On the day of eclosion, adults were numbered on their wings with a Staedler permanent pen. The experiment was.started within a day of the last adults to eclose. Virgin males were sterilized with 25 krads of gamma radiation first thing in the morning the day females were to be mated for the first time, and again two days later when females were mated for the second time. Lights were turned on at 0800 h and males allowed to feed for 1/2 hour in mating cages before females were introduced. Experimental Procedures Matings took place in ten 0.8 x 0.8 x 0.5 m mating cages placed behind large windows, 2 m. below 400 W HQIL lamps which were on from 0800 to 1800 h. Separate cages were used for each mating sequence so that neither large and small nor normal and irradiated males were put into the same cage. Each cage contained host plant leaves and either wild flowers or a potted Chrysanthemum sp. Into both nectaring sources, a 25% sugar solution was applied with a dropper into the middle of the flowers twice a day; first thing in the morning when the lights were turned on, and again in the early afternoon. Cages contained 1-2 more males than females up to a maximum of 20 butterflies, and were not crowded. Mating cages were checked every 15 min. for pairs in copula, which were transferred to 0.5 L plastic cups so as not to be disturbed by conspecifics in the cage. Upon separation, males and females were transferred to single sex holding cages containing host plant leaves and a nectaring source. To increase the likelihood females would remate, females were held for 24 hrs. in holding cages before they were put back into experimental mating cages and allowed to mate for the second time. Only those females that mated the first day were included in the experiment and given the opportunity to remate two days later. Females were given unlimited access to males until they remated. The experiment was stopped 9-12 days after females were first given the opportunity to remate, and the majority of females had remated. By this time remaining females were relatively old and likely not going to remate (Wiklund et al. 1993). In both years, only virgin males were used for first matings, and second matings in 1992. In 1993, because of the size of the experiment, although the majority of males used for second matings were virgins some non-virgin males 5 7 were needed to maintain at least a 1:1 sex ratio in the mating cages. To control for possible reduction in sterility, non-virgin males were used only in normal matings; all irradiated males were virgins. Non-virgin males were held in holding cages and fed ad libitum for at least 48 hrs, and not longer than 72 hours, before being used. By this time, P. napi males produce a second ejaculate of similar mass and quality to that transferred by males mating for the first time (Chapter 3). Doubly mated females were maintained in individual egg laying cages which contained a nectaring source and a host plant leaf, Armoracia rusticana L., for egg laying. Host plant leaves were replaced daily to minimize incidental loss or cannibalism among different aged larvae during development. After the eggs were counted, leaves were kept in a walk-in environmental chamber at 20 °C and 22h day length. Larvae were counted and removed within a day of hatching, 4 days after the eggs were laid. Leaves were rechecked for larvae 3 days later to ensure no larvae had been missed. In 1992 females were maintained in egg-laying cages until they laid between 100-150 eggs or died, after which they were frozen. The amount of time females were kept in egg-laying cages was increased in 1993 to investigate if sperm utilization changed over time. In 1993 females were housed in egglaying cages until they laid 200 eggs or died, after which they were frozen. In both years, at the end of the experiment, the bursa copulatrix was dissected out of all females and the number of spermatophores counted to ensure that all doubly mated females contained two spermatophores. Only data obtained from doubly mated females, laying at least 100 eggs, and containing two spermatophores were included in the analysis. Because of the difference in generations and irradiation dosages, data from the two years were not pooled. A l l statistics were calculated using SYSTAT. Non-parametric tests were used for non-normal data (Lillifors, p < 0.05). A separate variances t-test was used to determine the difference between group means when sample sizes differed. In this study sperm precedence refers to the male that obtained the majority of fertilizations and P2 specifically refers to the fertilizations obtained by the second male. Results Average second-male sperm precedence, measured by mean P2, was 64 % (n = 14) in 1992 and 67 % (n = 33) in 1993. Within treatments, second-male sperm precedence varied more in 1993 than in 1992 possibly owing to the lower irradiation dose used in 1993. Table 4.1. Mean percentage hatch and fecundity of females used in each treatment (L = large male, S = small male, I = irradiated, N = normal); n = sample size; ± 1SE provided for normally distributed data. Adult males were irradiated with 40 Krads in 1992 and 25 Krads of gamma radiation in 1993. In 1993, for all 'IN' and 'NI matings the percent of eggs fertilized by the 'N' male (x) was estimated from the proportion of eggs hatching (a) using the mean viabilities in 'NN' (b) and 'IP matings (c) using the formula x = (a-c) / (b-c) (SillenTullberg 1981). In 1992 irradiation results in 100% sterility, thus % hatch are those sired by 'normal' male. * mean sterility (LILI + SISI), ** mean egg viability for SNSN males only. % Hatch No. Eggs Laid n LILI 0.4 ± 0.4 137.5 ± 2 . 5 2 LILN 82.7 ± 5 . 8 141.0 ± 16.6 4 LISN 3.9 ± 3 . 9 105.3 ± 0 . 9 3 SILN 83.4 ± 12.2 124.0 ± 16.0 5 SISN 67.2 ± 8.2 140.0 ± 6 . 0 , 2 Treatment 1992 1993 II* 5.0 124.8 ±25.0 7 NN** 69.8 149.5 ± 30.5 4 LILN 63.9 172.3 ± 13.8 4 LISN 53.6 188.7 ± 11.6 11 LNSI 58.7 •211.6 ± 16.7 5 SILN 92.6 168.8 ± 15.7 6 SNLI 0 174.6 ±25.5 4 SISN 66.7 141.3 ±44.9 3 Table 4.2. The effect of mating order on paternity. A male was considered the principle sire he fathered > 60% of the offspring. (L = large male, S = small male, I = irradiated, N = normal). First Male Second Male No Principle Sire LILN 0 4 0 LISN 3 0 0 SILN 1 4 0 SISN 0 1 1 LILN 1 3 0 LISN 3 5 3 LNSI 3 2 0 SILN 0 6 0 SNLI 0 4 0 SISN 1 2 0 Treatment 1992 1993 60 1992 Thirty-two females mated at least once. Seventeen females mated twice. Only 1 doubly mated female contained one spermatophore and was excluded from analysis. Fifteen females mated once, but did not remate. Of those, 4 mated with a small male and 11 (73%) mated with a large male first. Seventy-nine percent of all doubly mated females (11/14) mated for the second time within 2-5 days. Three females remated 8-11 days later; two of which mated first with a large male. There was no difference in fecundity among females in the control or treatment groups ( A N O V A F = 0.90, df = 4, p > 0.05) (Table 4.1). The eclosion mass of large and small males differed significantly (Table 4.1; MannWhitney U = 180.0, p < 0.001, n = 28). However, within size categories there was no difference in size between normal and irradiated males (p > 0.40 for both categories). Mating duration did not differ between males mated first or second (t = -1.29, df = 24, p >0.05), and had no affect on P 2 (p > 0.05 for both mating 1 and mating 2). However, large males copulated for significantly longer than small males (r = 0.63, p = 0.001, n = 24). The proportion of offspring sired by the second male to mate, P 2 , could be determined for all double matings, except those in the control group, LILI. In 1992, P 2 was equal to the proportion of larvae to hatch because males mated last were all 'normal'. The interval between the first and second mating had no effect on P 2 (p > 0.05). However, male body size and mating sequence had a significant effect on P 2 . In 9 out of 14 double matings, the second male sired at least 60 % of the offspring (Table 4.2). In 4 double matings the first male obtained the majority of fertilizations, in 3 of which females mated to a large male first. The second male obtained the majority of fertilizations when the two males were the same size (Table 4.1; L I L N and SISN). In double matings involving a large and small male, small males fertilized a much smaller percentage (4%) of the brood when mated last than when a larger male mated last (83%) ( P 2 LISN vs SILN; Separate variance t-test t =- 6.21, df = 4.7, p = 0.002) (Table 4.1). Large males obtained significantly more fertilizations than small males regardless of if they mated first or last (Fig. 4.1a; A N O V A F = 13.07, df = 3, p = 0.001). The effect of male size on paternity is further demonstrated by the significant negative relationship between P 2 and the relative size of the first and second mate (Fig. 4.2a; slope = -2.43, r = 0.83, n = 11, p = 0.002). Second mates obtained fewer fertilizations the larger the size of the first mate. 61 1992 CN 0.0 llln llsn siln slsn Mating Order 1993 1.0 ~i 1 r 0.9 0.8 0.7 0.6 PM m j -r 0.5 0.4 0.3 0.2 0. 1 0.0 J llln L llsn lnsl J siln slsn L snll Mating Order Figure 4.1. Proportion of offspring sired by the second male (P2) in relation to male body size in doubly mated P. napi females a) diapause generation; b) direct-developed generation. Male categories; li = large irradiated, In = large normal male, si = small irradiated, sn = small normal. 1992 0, f N 0.8 0.9 1.0 l . l w 1.2 W 1.3 Relative Male Size (Ml / M2) 1993 o.o 0.50 0.75 LOOT 1.25 ~ 1.50 Relative Male Size (Ml / M2) Figure 4.2. Proportion of offspring sired by the second male (P2) as a function of relative male body size a) relative male size = Eclosion weight Male 1 / Eclosion weight Male 2. Linear regression: y = -2.43x + 2.93, r = 0.83, n = 11, p = 0.002; b) relative male size = Pupal weight Male 1 / Pupal weight Male 2. Spearman rank correlation coefficient: r = -0.24, p > 0.05, n = 33. s 63 1993 A total of 100 females mated at least once; 65 mated twice and 35 mated once. O f those females that mated only once, 60 % (21/35) mated with a large male, and 14/35 mated with a small male. A l l females mated for the first time with virgin males; 9/65 females mated for the second time with a non-virgin male. Eight doubly mated females were excluded from analysis because they contained only 1 spermatophore (note: all 8 females excluded mated both times with virgin males). Thirteen doubly mated females were excluded because they laid less than 15 eggs before they died. There was no difference in fecundity among doubly mated females in the treatment and control groups ( A N O V A F = 1.62, df = 8, p = 0.15) (Table 4.1). There was no correlation between P2 and the interval between the two matings (Spearman rank correlation coefficient: r = -0.012, n = 33, p > s 0.05). However, females mated to small males first remated sooner than females mated to large males first. Ninety-one percent of females (30/33) remated 2-4 days after their first mating. Three females remated 5-8 days later, and all of these mated with a large male first. The pupal mass of large and small males differed significantly (Mann-Whitney U= 6935, p < 0.001, n = 168). Within the two size categories, there was no significant difference in the pupal mass of normal and irradiated males (p > 0.05). There was no correlation between copulation duration and P2 (copulation 1 Spearman rank correlation r = -0.22; s copulation 2 r = 0.27, n = 33; p > 0.05 for both first and second matings) or male pupal mass s and mating duration (r = 0.075, n = 33, p > 0.05). However, second matings (x* = 116.89 ± s 7.25 min.) were significantly longer than first matings (x = 96.07 ± 3.54 min.) (t = -2.24, df = 90, p = 0.03). The proportion of offspring sired by each male was determined for all double matings except those in the control groups; ie. sterile-sterile (II) and normal-normal (NN) matings. Sperm utilization did not seem to change over time. Among females, there was only a mean difference of 1.32 + 0.7 % in egg viability between eggs laid 5 and 10 days after oviposition. None of the females mated to a large normal mate remated (n = 12). Thus the egg viability of control normal-normal matings could only be determined for small-small normal matings. Mean egg viability for N N matings was 70% (SNSN only, n = 4), and 4.8% in the II matings (mean value for LILI + SISI, n = 8) (Table 4.1). Overall the second male fertilized 67% (n = 33) of the brood. In 30 out of 33 double matings, either the first or second male obtained > 60% of the fertilizations (Table 4.2); in 73% (22/30), the second male sired > 60% of the offspring; in 8 matings the first male sired > 60% of the offspring. When the first male fertilized the majority of eggs, 6 out of 8 females mated with a large male first and a small 64 male second; the remaining 2 females mated with males of the same size (LILN and SISN). Females that had P 2 values of 40 - 54 % (n = 3) all had mated with a large male first and a small male second. There was a significant effect of treatment on paternity (LISN, LNSI, SJJLN, S N L I ; Kruskall Wallis H = 9.52, df = 3, p = 0.023). In general, large males fertilized the majority of eggs, although small males had a higher fertilization success when they mated last. Large males mating first obtained almost 60% of the fertilizations, whereas small males mating first fertilized < 10% of the eggs (Table 4.1). Variation in P 2 increased when a small male mated last, or the two males did not differ in size (Fig. 4.1b). There was no difference in P 2 within groups that had small males mating last (LISN or LNSI) or large males mating last (SILN or SNLI) (Fig. 4.1b) (p > 0.10 for both categories). However, comparing the P 2 values of those groups in which a large male mated last (SILN and SNLI) to those groups that had small males mating last (LISN and LNSI), revealed a significant effect of male body size on the degree of second-male sperm precedence ( P 2 ) (Fig. 4.1b; Mann-Whitney U = 33.5, p = 0.01, n = 26). P 2 was significantly lower when small males mated last (ie. in treatments L I S N and LNSI). Thus the fertilization success of larger males was enhanced when they mated last. Although there was no significant correlation between P 2 and relative male size, (Spearman rank correlation r = -0.24, p > 0.05, n = 33), the degree of second-male sperm precedence s was very low when the first male was much larger (Fig. 4.2b). Discussion The range in P 2 values varied substantially in both years (1992 mean P 2 = 0.64, range = 0.0 - 1.0, n = 16; 1993 mean P 2 = 0.67, range = 0.0 - 1.0, n = 46) and as suggested by Lewis and Austad (1990), say little about the sperm precedence pattern. I found that male body size had a significant affect on paternity in doubly mated P. napi females. Larger body size was associated with greater sperm precedence regardless of female mating status (first or second mating) (Fig. 4.1), or the number of days between matings. The degree of secondmale sperm precedence ( P 2 ) was also affected by relative male size (Fig. 4.2). In general, second males obtained fewer fertilizations the larger the size of the first male. Hence, male butterflies derive three advantages from being large through their transfer of large ejaculates. First in terms of male mating effort, large ejaculates decrease the propensity of females to remate in faculatively multiply mating females, which likely allows more of the donors' sperm to be used for fertilization before the female remates. Second, 65 nutrients contained in ejaculates are used by females in offspring production, and as such function as paternal investment. Larger ejaculates increase the interval between matings, allowing most of the nutrients contributed by the donor to be used in the production of his offspring before the female remates, decreasing the probability his investment will be cuckolded by a future mate. In this study I have found yet a third advantage. Under conditions of sperm competition, large males have a higher fertilization success. Furthermore, my results show why male mating success may not be a good predictor of fertilization success. For many species of butterflies, ejaculate size is correlated with male size and affected by mating history, with recently mated males transferring smaller ejaculates than virgin males (Labine 1964; Sugawara 1979; Svard and Wiklund 1986, 1989; Oberhauser 1988, 1989; Kaitala and Wiklund 1995, Chapter 3). Small males and males that have mated recently may suffer reduced reproductive success for two reasons; 1) females are more likely to remate when they receive a small ejaculates and 2) in terms of the proportion of offspring sired, males that transfer small ejaculates will suffer from a reduced ability to compete in sperm competition if the female remates to a larger or virgin male. A couple of hypotheses have been invoked to explain the evolution of large male ejaculates. The first, associated with the mating effort function, proposes that in polyandrous systems sperm competition selectively favours males that transfer large ejaculates containing large quantities of sperm, that also effectively delay female remating (Svard and Wiklund 1989; Wiklund and Forsberg 1991). The second hypothesis is associated with the paternal investment function, where males that are most effective in providing females with a valuable nuptial gift are selectively favoured (Walker 1980; Gwynne 1984; Leimar et al. 1994). In a comparative study of 8 pierid and 8 satyrid butterfly species, Leimar et al. (1994) found a positive correlation between relative ejaculate mass transferred by males arid variance in female wing length (which was used as measure of the likelihood that females suffer unpredictable food availability during their development). Leimar et al. (1994) suggested that the paternal investment function may be important in species where the food supply for larvae is unpredictable, and thus where adult size at eclosion may be largely determined by stochastic environmental factors. Here females that have suffered food shortage as larvae, and subsequently metamorphosed at a small size, may be particularly in need of nuptial gifts provided by males. Interestingly, in their study, P. napi females exhibited the strongest variation in size and P. napi males transferred the second largest relative ejaculate out of the 16 species examined. These two hypotheses for 66 the evolution of large male ejaculates, and hence large male size, are not mutually exclusive and can obviously act together. It is difficult to interpret the effect of copulation duration on sperm precedence. Data obtained in 1992 showed that larger males copulated for a longer time than smaller males. In 1993 I found that second males (i.e. those mated to non-virgin females) copulated for a longer time than males mated to virgins (i.e. those males mated first). These two results may related to the amount of sperm transferred. There is some evidence that males may be able to assess female mating status (Wiklund and Forsberg 1985; Cook and Gage 1995). In the Indian meal moth, Plodia interpunctella Hiibner, Cook and Gage (1995) found that males responed to the presence of sperm already in storage by increasing the numbers of eupyrene sperm ejaculated. Thus males may vary the amount of sperm transferred during copulation in response to the assessed risks of sperm competition (Gage 1995, Cook and Gage 1995), and it may take longer to transfer more sperm. Three hypotheses have been proposed to explain the underlying selective forces behind observed sperm precedence patterns; 1) direct competition among sperm for access to eggs (Parker 1970), 2) female choice resulting from preferential sperm use by females (Knowlton and Greenwell 1984; Eberhard 1985) and 3) sperm precedence as a secondary consequence of the morphology of the sperm-storage organ (Walker 1980). Walker's more proximate hypothesis suggests that in species with elongate spermathecae, such as lepidoptera, sperm precedence results from the sequential packing of successive ejaculates, such that sperm from the most recent ejaculate fertilizes most of the eggs laid subsequently. However, in my study this cannot account for the high paternity obtained by large males mating first. LaMunyon and Eisner (1993) also found that in doubly mated females of an arctiid moth, most offspring were sired by the larger male, and they suggested that females exercise active sperm preference. M y findings support the sperm competition and/or female choice hypotheses. Both hypotheses offer explanations for the observed body-size advantage. Parker (1982, 1990) predicted that sperm competition selects for increased investment in sperm when mechanisms of sperm competition follow a "raffle principle", whereby a male's fertilization success is directly proportional to the number of his sperm in the "fertilization set" (those used to fertilize the eggs) relative to those of other males. The fair raffle model predicts a decline in P2 when the number of sperm inseminated by the second male is reduced or when the number of sperm inseminated by the first is increased (Parker et al. 1990). This model seems 67 to describe the sperm precedence pattern found in this study; second mates obtained fewer fertilizations the larger the size of the first mate (Fig. 4.2). Interestingly, Simmons and Parker (1992) also found this same relationship in dung flies; relative male body size influenced a male's success in sperm competition. Although in Lepidoptera, it is plausible that females differentially utilize sperm, my results and those of LaMunyon and Eisner (1993) could be the result of a much more basic process; larger ejaculates may contain more sperm, which in turn may determine fertilization success. Cook and Gage (1995) found that sperm numbers increased with ejaculate size in the in the moth Plodia interpunctella. However because sperm numbers were not counted in this study, exactly how sperm precedence is achieved in P. napi remains unclear. Although sperm competition may have influenced ejaculate size, when females benefit from male-derived nutrients, selection on females to 'encourage' males to mate may be responsible for promoting some degree of sperm displacement (Walker 1980). Average second male sperm precedence for both generations, as measured by overall mean P2, was 66%. Small males had a higher fertilization success when they mated after a large male than when they mated before. Thus in the wild it will generally benefit males to mate, even with non-virgin females. In some species, female reproductive output is influenced by their ability to obtain male-derived resources (Watanabe 1988; Oberhauser 1989; Wiklund et al. 1993; Watanabe and Ando 1993, 1994). In these species, male ejaculates function as paternal investment (Gwynne 1984; Simmons and Parker 1989) and as a result, selection should favour last male sperm precedence (Walker 1980; Gwynne 1984). In this study I have shown that for males, it pays to produce a large ejaculate; males producing a large spermatophore may have a selective advantage in sperm competition over those transferring a smaller ejaculate. Although I showed that for doubly mated females, large males have a reproductive advantage, I do not know how his paternity will be affected if the female remates. However, it does appear that a future mate will obtain some fertilizations, the extent of which may depend on the size of ejaculate transferred. The negative relationship between P2 and the relative size of the second mate suggests that sperm precedence may be determined by the number of sperm transferred. Thus it appears that intrasexual selection on males has influenced the evolution of ejaculate size and that selection on females may be responsible for the overall advantage to mating last. 68 CONCLUSIONS Male Lepidoptera transfer an ejaculate to females during copulation that contains both sperm and nutrient-rich accessory substances. Male investment of this type has previously been regarded as paternal investment in that it has been demonstrated that male-derived nutrients are used by females to increase their reproductive output and longevity (Boggs and Gilbert 1979; Rutowski etal. 1987; Oberhauser 1989; Watanabe 1988; Wiklund etal. 1993). However, as first realized by Darwin (1871), competition among males for the opportunity to mate generates strong, evolutionary pressures on characteristics that contribute to a male's mating success, which can continue after insemination (Parker 1970). In the context of these ideas, my research suggests that both forces have worked together in the evolution of ejaculate size and nutrient content in nectar-feeding butterflies. First of all, it is apparent that males in polyandrous systems make a substantial investment in ejaculates (Chapter 1, 2 and 3). At the same time, because males in polyandrous systems also produce relatively larger spermatophores, affecting both female mating frequency and the number of progeny a male can hope to sire (Chapter 4), the spermatophore influences a male's fertilization success, and also functions as male mating effort. In nectar-feeding butterflies, males are in a position to invest in reproduction, via a spermatophore, because proteins and other nutrients that are available during the larvalfeeding stage are difficult to obtain as adults (cf. Marshall 1982). Boggs (1990) predicted an inverse relationship between the importance of protein and the availability and utilization of proteins acquired by females in their diet. As adults, the species studied in this thesis feed predominantly on nectar, which contains very little protein (Baker and Baker 1973). Hence male-derived proteins should be of great importance to females. Polyandry would be selected for when males transfer nutrients via ejaculates that can be used by females to increase their lifetime reproductive output, whereas female monandry would be selected for when males do not transfer any nutrients or when females will not live long to benefit from remating (Wiklund and Forsberg 1991). Furthermore, if male-derived nutrients functioned solely as paternal investment, protein content should be highest in monandrous species where the donor has a high confidence of paternity (Boggs 1981a, Marshall 1982, Gwynne 1984). However, if females in polyandrous systems use mating to acquire male-derived nutrients (Walker 1980; Kaitala and Wiklund 1994), spermatophores transferred by males in polyandrous systems should contain more nutrients than those transferred by males in 69 monandrous systems. In monandrous systems, where females mate on average only once, their fitness should not be as dependent on male-derived resources. Females in polyandrous systems may "expect" and depend more on male-derived resources in order to realize their reproductive potential (Oberhauser 1989, 1996; Leimar et al. 1994; Wiklund and Kaitala 1995; Karlsson 1995). Both relative protein content and ejaculate size increased with the degree of polyandry (Chapter 1). Thus females in polyandrous systems can use mating to access and increase the amount of protein they obtain in their lifetime. The actual amount of protein females in polyandrous systems receive in their lifetime will be much greater than that received by females in monandrous systems (Chapter 1 and 2). Boggs (1990) suggested that in species where availability of protein-rich food is variable, or limits female fecundity, it pays both sexes to allocate protein into egg production. In polyandrous systems, there is selection on males to transfer large ejaculates, and females may have evolved to capitalize on this investment. Rutowski et al. (1983) found that males across 4 families of butterflies transferred proportionally similar masses of ejaculate material. They concluded that nutritional investments made by males were similar across species. However, in terms of the actual amount of protein transferred, this is not the case. I found a wide variation both in the amount of protein and the relative size of ejaculate transferred between species (Chapter 1 and 2). The relative amount of protein transferred (% wet wgt.) is similar across species, suggesting that a minimum amount of protein must be transferred regardless of mating system (Chapter 1), especially if proteins are involved in other functions such sperm activation and oviposition (cf. Oberhauser 1992, 1996). The degree of polyandry had a substantial effect on a male's ability to produce more than one large, nutritious ejaculate (Chapter 2). Relative to males in more monandrous systems, males in polyandrous systems i) transfer relatively larger first ejaculates (Chapter 1) , ii) transfer ejaculates relatively richer in protein (Chapter 1), iii) recuperate faster and are able to transfer a second ejaculate as large as the first sooner (Chapter 2), iv) are better able to maintain relative ejaculate mass over several matings (whereas males in the less polyandrous systems were unable to transfer a second ejaculate as large as the first) (Chapter 2) , ii) maintain copula duration (Chapter 2) and iii) donate a larger total amount of nutrients over several matings (Chapter 2). Thus, males in polyandrous systems are better adapted to mating repeatedly, and seem to invest more resources in reproduction than males in relatively monandrous systems. 70 Male capacity to produce more than one large, nutritious ejaculate is limited in monandrous species. In these systems, where females generally mate only once, males invest heavily in first matings (Chapter 2). Conversely, males in polyandrous systems have the ability to recuperate quickly from a mating event and thus appear to be better adapted to mating repeatedly. Male ability to remate increased with the degree of polyandry, in spite of the fact that males in most polyandrous species transferred larger, higher quality ejaculates than males in monandrous species (Chapter 1 and 2). Once male nutrients contribute significantly to female reproductive output, a reduction in nutrient quality may not be possible because of a possible negative affect on offspring fitness (Simmons and Parker 1989; Gwynne 1990; Simmons and Bailey 1990). In polyandrous systems, males would benefit most by either producing undigestible ejaculates that will still function to delay female remating, or immediate use of donated nutrients, thereby minimizing the probability their resources are used to benefit another male's offspring when the female remates (Parker and Simmons 1989). Consequently, there may be conflict between the sexes (Parker 1979) over female remating intervals, resulting from selection on females to acquire of male-derived nutrients, and selection on males to minimize female remating in order to maximize their fertilization success. In several species, ejaculate mass is positively correlated with both male body size (Boggs 1981a; Rutowski etal. 1983; Svard and Wiklund 1986, 1989; Jones etal. 1986; Oberhauser 1988) and the duration of female unreceptivity (Labine 1964; Obara et al. 1975; Sugawara 1979; Rutowski 1980, Rutowski etal. 1981; Oberhauser 1989; Wiklund and Kaitala 1995). The positive correlation between male eclosion mass and ejaculate mass suggests that larval reserves may influence the potential size of ejaculate a male can produce (Boggs 1981b). Wickman & Karlsson (1989) suggested that when reserves collected by larvae limit reproduction, the proportional increase of reserves with body size should be paralleled by an increase in reproductive effort, possibly resulting in increased ejaculate production by larger males. In P. napi, although I found a significant relationship between male eclosion mass and the mass of ejaculates transferred by males mating for the first time, there was no effect of body size on either the mass or absolute amount of protein (mg) in second ejaculates produced by males remated two days later (Chapter 3). In fact, small males transferred proportionally larger ejaculates in their first and second matings. Furthermore, the protein content of ejaculates produced by small males mating for the first time was much more variable than that produced by large males, with some ejaculates containing as much protein as that contained in ejaculates produced by larger males (Chapter 3). Thus the 71 nutrient content of spermatophores does not necessarily increase with male body size or ejaculate size. This may be one reason why some studies that have looked for an effect of ejaculate mass or male body mass (Greenfield 1982; Jones et al. 1986; Svard and Wiklund 1988; Royer and McNeil 1993; Oberhauser 1989) have found no affect of either of these variables on female reproductive output. Mating can be costly to both sexes. Although females may obtain nutritional benefits by mating, females mating to recently mated males remain in copula for extended periods of time, and receive small ejaculates (Chapter 2 and 3). Lengthy copulations prohibit females from foraging and/or oviposition, which may be especially costly to short-lived species. For males, investment in these nutritious, packaged ejaculates is physiologically costly and time consuming. In two polyandrous species of butterfly, Pieris napi and P. rapae, copulation durations were longer and ejaculates smaller in matings involving recently mated males (Chapter 3). It takes time for males to recover from a mating event. The extent to which copulation duration, ejaculate mass and protein content changed was influenced by the interval between matings, suggesting that copulation depletes a male's immediate access to accessory gland products, such as protein. It is male mating history and not necessarily male body mass that greatly affects the protein content of ejaculates. In several species of Lepidoptera, others have found similar effects of male mating history on ejaculate mass and copulation duration (Rutowski 1979; Sims 1979; Boggs 1981a; Svard 1985; Sviird and Wiklund 1986, 1989; Oberhauser 1988, 1992; He and Tsubaki 1992; Royer and McNeil 1993). The increase in both ejaculate mass and protein content over time (Chapter 3), suggests that males are time-constrained, needing time to mobilize and/or replenish ejaculate constituents, and corroborating the hypothesis that ejaculates are costly for males to produce (Dewsbury 1982; Drummond 1984; Svard 1985; Fitzpatrick and McNeil 1989). Males did not delay remating until they could produce a large ejaculate, suggesting that males in polyandrous systems may maximize their fitness by remating as often as possible (Darwin 1871; Bateman 1948; Oberhauser 1988). In classic studies of sperm competition, the effect of mating order on the outcome of sperm competition was realized, but recently, the influence of male size has been shown to be important as well, with large males have an advantage over smaller males in sperm competition in insect groups as diverse as beetles (Lewis and Austad 1990), flies (Simmons and Parker 1992), and arctiid moths (LaMunyon and Eisner 1993). In addition to this list, Chapter 4 suggests that individual variation in male fertilization success is related to male body size the polyandrous butterfly, Pieris napi. Larger body size was associated with 72 greater sperm precedence regardless of female mating status (first or second mating), or the number of days between mating (Chapter 4). The degree of second-male sperm precedence (P2) was also affected by relative male size. In general, second males obtained fewer fertilizations the larger the size of the first mate (Chapter 4). Interestingly, this same relationship between male body size and P2 was also demonstrated in a species of dung fly (Simmons and Parker 1992). Thus, the affect of sperm competition on male investment in ejaculates appears to be a strong, prevailing force as further demonstrated by several interspecific studies, ranging from humans to birds, showing that higher levels of polyandry are associated with male characteristics promoting increased investment in sperm production (Harcourt etal. 1981; Cartar 1985; Kenagy and Trombulak 1986; Moller 1988a, b; Ginsberg and Rubenstein 1990, Cook and Gage 1995). Relative ejaculate mass increased with the degree of polyandry (Chapter 1). In terms of the number of eggs a given male will fertilize there are three possible benefits associated with large male size and thus the production of a large ejaculate. In several polyandrous species, ejaculate mass is positively correlated with the length of a female's postmating refractory period (Labine 1964; Obara et al. 1975; Sugawara 1979; Rutowski et al. 1981; Oberhauser 1989; Wiklund and Kaitala 1995), resulting in more of a donor's sperm being used to fertilize the eggs laid before the female remates. Secondly, ejaculates contain nutrients that females can use increase their reproductive output (Boggs and Gilbert 1979; Rutowski et al. 1987, Watanabe 1988; Svard and Wiklund 1989). Thirdly, larger males have a higher fertilization success, possibly because they transfer larger spermatophores containing more sperm (Chapter 4). These findings suggest that in polyandrous systems, the risk of sperm competition has influenced the evolution of ejaculate size. By transferring a large ejaculate, males in polyandrous systems can increase their fertilization success both in sperm competition (Chapter 4) and by prolonging the duration of the female refractory period, although the benefits to males arising from delaying females remating will depend on the duration of the delay, and her daily fecundity (Oberhauser 1989). My results suggest that male mating success may not be a good predictor of fertilization success. For many species of butterflies, ejaculate size is correlated with male size and affected by mating history, with recently mated males transferring smaller ejaculates that virgin males (Labine 1964; Sugawara 1979; Svard and Wiklund 1986, 1989; Oberhauser 1988, 1989; Kaitala and Wiklund 1995, Chapter 3). Small males and males that have mated recently may suffer reduced reproductive success for two reasons; 1) females are more likely to remate when they receive a small ejaculates and 2) in terms of the proportion of offspring 73 sired, males that transfer small ejaculates will suffer from a reduced ability to compete in sperm competition if the female remates to a relatively larger or virgin male. Although sperm competition may have influenced ejaculate size, when females benefit from male-derived nutrients, selection on females to 'encourage' males to mate may be responsible for promoting some degree of sperm displacement (Walker 1980). When female reproductive output is influenced by their ability to obtain male-derived resources (Watanabe 1988; Oberhauser 1989; Wiklund etal. 1993; Watanabe and Ando 1993, 1994), male ejaculates function as paternal investment (Gwynne 1984; Simmons and Parker 1989) and as a result, selection should favour last male sperm precedence (Walker 1980; Gwynne 1984). Average second male sperm precedence, measured by overall P2, was 66% in Pieris napi. Thus in the wild it will generally benefit males to remate, even with non-virgin females. Therefore in polyandrous, nectar-feeding butterflies, intrasexual selection on males may have influenced the evolution of ejaculate size but selection on females to obtain malederived nutrients may be responsible for the overall advantage to mating last. Selection on males to maximize their fertilization success may have lead to the evolution of large ejaculates in polyandrous systems. However, polyandry would be selected for when males transfer nutritious ejaculate to females at mating which females use to enhance their reproductive output, whereas monandry would be favoured if females do not obtain nutritional benefits. Across several species of butterfly, nutrient content and size of ejaculate produced increased with the degree of polyandry (Chapter 1). Moreover, in some polyandrous species of butterflies, the reproductive performance of females is dependent on male-derived nutrients, as evidenced by the fact that multiply mated females have a higher reproductive output and/or longevity than females allowed to mate only once (Watanabe 1988; Oberhauser 1989; Wiklund etal. 1993; Watanabe and Ando 1993, 1994). In these species, females may be dependent on male-derived resources to realize their reproductive potential (Leimar et al. 1994; Karlsson 1995). This dependence may have resulted from selection on females to use the nourishment contained within these large spermatophores in an effort to reduce their own reproductive costs. When male fertilization success is influenced by the size of ejaculate transferred, males may not be able to "cheat" by providing females with ejaculates of lower nutritional quality (Simmons and Parker 1989). In this way, male reproductive success and female reproductive output will both be enhanced by the transfer of a large, nutritious ejaculate. 74 BIBLIOGRAPHY Alexander RD, Borgia G (1979) On the origin and basis of the male-female phenomenon. In: Blum, M S and Blum, N A (eds) Sexual Selection and Reproductive Competition in Insects. Academic Press, New York, pp 417-440 Baker H G , Baker I (1973) Amino-acids in nectar and their evolutionary significance. Nature 241 : 543-545 Bateman A J (1948) Intra-sexual selection in Drosophila. Heredity 2 : 349-368 Boggs C L (1981a) Selection pressures affecting male nutrient investment at mating in Heliconiine butterflies. Evolution 35 : 931-940 Boggs C L (1981b) Nutritional and life-history determinants of resource allocation in holometabolous insects. A m Nat 117 : 692-709 Boggs C L (1986) Reproductive strategies of female butterflies: variation in and constraints on fecundity. Ecol Entomol 11:7-15 Boggs C L (1990) A general model of the role of male-donated nutrients in female insects' reproduction. A m Nat 136 : 598-617 Boggs C L (1995) Male nuptial gifts: phenotypic consequences and evolutionary implications. In: Leather, SR, Hardie J (eds) Insect reproduction. C R C Press, Boca Raton, pp 215242 Boggs C L , Gilbert L E (1979) Male contribution to egg production in butterflies: evidence for transfer of nutrients at mating. Science 206 : 83-84 Boggs C L , Watt W B (1981) Population structure of Pierid butterflies IV. Genetic and physiological investment in offspring by male Colias. Oecologia 50 : 320-324 Brower JH (1975) Sperm precedence in the Indian meal moth, Plodia interpunctella. Ann Entomol Soc A m 68 : 78-80 Cartar R V (1985) Testes size in sandpipers. Naturwissenschaften 2 : 157-158 Cook PA, Gage M J G (1995) Effects of risks of sperm competition on the numbers of eupyrene and apyrene sperm ejaculated by the moth Plodia interpunctella (Lepidoptera: Pyralidae). Behav Ecol Sociobiol 36 : 261-268 Darwin, C (1871) The Descent of Man, and Selection in Relation to Sex. London: John Murray Dewsbury D (1982) Ejaculate cost and male choice. A m Nat 119 : 601-610 Drummond B A III (1984) Multiple mating and sperm competition in the Lepidoptera. In: Smith R L (ed) Sperm competition and the evolution of animal mating systems. Academic Press, New York, pp 291-370 Eberhard, W G (1985) Sexual Selection and Animal Genitalia. Harvard University Press, Cambridge, Mass 75 Ehrlich A H , Ehrlich PR (1978) Reproductive strategies in the butterflies. I. Mating frequency, plugging and egg number. J Kans Entomol Soc 51 : 666-697 Elgar M A (1992) Sexual cannibalism in spiders and other vertebrates. In: Elgar M A , Crespi BJ (eds) Cannibalism: Ecology and Evolution Among Diverse Taxa. Oxford University Press, pp 129-156 Engelmann F (1970) The Physiology of Insect Reproduction. Pergamon Press, Oxford. Fitzpatrick, S M , McNeil JN (1989). Lifetime mating potential and reproductive success in males of the true armyworm, Pseudaletia unipuncta (Haw.) (Lepidoptera: Noctuidae). Funct Ecol 3 : 37-44 Forsberg J (1988) Reproductive Biology of Some Pierid Butterflies. Ph.D. Thesis, Stockholm University, Sweden. Forsberg J, Wiklund C (1989) Mating in the afternoon: time-saving in courtship and remating by females of a polyandrous butterfly Pieris napi L . Behav Ecol Sociobiol 25 : 349356 Gage M J G (1994) Associations between body size, mating pattern, testis size and sperm lengths across butterflies. Proc R Soc Lond B 258 : 247-254 Gage M J G (1995) Continuous variation in reproductive strategy as an adaptive response to population density in the moth Plodia interpunctella Proc R Soc Lond B 261 : 25-30 Gage M J G , Cook P A (1994) Sperm size or numbers? Effects of nutritional stress upon eupyrene and apyrene sperm production strategies in the moth Plodia interpunctella (Lepidoptera: Pyralidae). Funct Ecol 8 : 594-599 Garland T Jr, Dickerman A W , Janis C M , Jones J A (1993) Phylogenetic analysis of covariance by computer simulations. Syst Biol 42 : 265-292 Geiger H J (1981) Enzyme electrophoresis studies of the genetic relationship of Pierid butterflies. I. European taxa. J Res Lepid 19:181-195 Gilbert L E (1972) Pollen feeding and reproductive biology of Heliconius butterflies. Proc Natl Acad Sci. U S A 69 : 1403-1407 Ginsberg JR, Rubenstein DJ (1990) Sperm competition and variation in zebra mating behaviour. Behav Ecol Sociobiol 26 : 427-434 Greenfield M D (1982) The question of paternal investment in Lepidoptera: male-contributed protein in Plodia interpunctella. Int J Invert Reprod 5 : 323-330 Gwynne DT (1984) Male mating effort, confidence of paternity, and insect sperm competition. In: Smith, R.L. (ed.). Sperm Competition and the Evolution of Animal Mating Systems. Academic Press, London, pp. 117-150 Gwynne DT (1990) Testing parental investment and the control of sexual selection in katydids: the operational sex ratio. A m Nat 136 : 474-484 Harcourt A H , Harvey PH, Larson SG, Short R V (1981) Testes weight, body weight and breeding system in primates. Nature 293 : 55-57 76 Harvey PH, Pagel M D (1991) The Comparative Method in Evolutionary Biology. Oxford University Press, Oxford He Y , Tsubaki Y(1992). Variation in spermatophore size in the army worm, Pseudaletia separata (Lepidoptera: Noctuidae) in relation to rearing density. App Entomol and Zool 27 : 39-45 Jones K N , Odendaal FJ, Ehrlich PR (1986) Evidence against the spermatophore as a paternal investment in the checkerspot butterflies {Euphydryas: Nymphalidae). A m Midi Nat 116 : 1-6 Kaitala A , Wiklund C (1994) Polyandrous female butterflies forage for matings. Behav Ecol Sociobiol 35 : 385-388 Kaitala A , Wiklund C (1995) Female mate choice and mating costs in the polyandrous butterfly Pieris napi (Lepidoptera: Pieridae). J Insect Behav 8 : 355-363 Karlsson B (1989) Fecundity in Butterflies. Adaptations and Constraints. Ph.D. Thesis, Stockholm University, Sweden Karlsson B (1995) Resource allocation and mating systems in butterflies. Evolution 49 : 955961 Karlsson B , Wickman P-0 (1990) Increase in reproductive effort as explained by body size and resource allocation in the speckled wood butterfly, Pararge aegeria (L.). Funct Ecol 4 : 609-617 Kenagy GJ, Trombulak SC (1986) Size and function of mammalian testes in relation to body size. J Mammal 67 : 1-22 Khalifa A (1950) Spermatophore production in Galleria mellonella L . (Lepidoptera). Proc. R. Entomol. Soc. Lond. A . 25 : 33-42 Knowlton N , Greenwell SR (1984) Male sperm competition avoidance mechanisms: the influence of female interests. In: Smith, R.L. (ed.). Sperm Competition and the Evolution of Animal Mating Systems. Academic Press, London, pp. 61-84 Labine P A (1964) Population biology of the butterfly Euphydryas editha. I. Barriers to multiple inseminations. Evolution 18 : 335-336 Labine P A (1966) The population biology of the butterfly, Euphydryas editha. IV. Sperm precedence - A preliminary report. Evolution 20 : 580-586 Labine P A (1968) The population biology of the butterfly Euphydryas editha. Oviposition and its relation to patterns in other butterflies. Evolution 22 : 799-805 LaMunyon C W , Eisner T (1993) Postcopulatory sexual selection in an arctiid moth (Utetheisa ornatrix). Proc Natl Acad Sci U S A 90 : 4689-4692 Lederhouse R C (1981) The effect of female mating frequency on egg fertility in the black swallowtail, Papilio polyxenes asterius (Papilionidae). J Lepid Soc 35 : 266-277 Leimar OB, Karlsson B , Wiklund C (1994) Unpredictable food and sexual size dimorphism in insects. Proc R Soc London B 258 : 121-125 77 Leopold R A (1976) The role of male accessory glands in insect reproduction. Ann Rev Entomol 21 : 199-221 Lewis S M , Austad S N (1990) Sources of intraspecific variation in sperm precedence in red flour beetles. A m Nat 135 : 351-359 Marshall L D (1982) Male nutrient investment in the Lepidoptera: What nutrients should males invest? A m Nat 120 : 273-279 Marshall L D (1985) Protein and lipid composition of Colias philodice and C. eurytheme spermatophores and their changes over time (Pieridae). J Res Lep 24 : 21-30 Marshall L D , McNeil JN (1989) Spermatophore mass as an estimate of male nutrient investment: a closer look in Pseudaletia unipuncta (Haworth) (Lepidoptera: Noctuidae). Funct Ecol 3 : 605-612 Mefller A P (1988a) Ejaculate quality, testes quality and sperm competition in primates. J Human Evol 17 : 89-502 Moller A P (1988b) Testes size, ejaculate quality and sperm competition in birds. Biol J Linn Soc 33 : 273-283 Obara Y , Tateda H, Kuwabara M (1975) Mating behaviour of the cabbage white butterfly, Pieris rapae crucivora Boisduval. V . Copulatory stimuli inducing changes of female response patterns. Zool Mag 84 : 71-76 Oberhauser KS (1988) Male monarch butterfly spermatophore mass and mating strategies. Anim Behav 36 : 384-1388 Oberhauser KS (1989) Effects of spermatophores on male and female monarch butterfly reproductive success. Behav Ecol Sociobiol 25 : 237-246 Oberhauser K S (1992) Rate of ejaculate breakdown and intermating intervals in monarch butterflies. Behav Ecol Sociobiol 31 : 367-373 Oberhauser KS (1996) The relationship between mating and oogenesis in monarch butterflies. J Insect Behav (in press) Parker G A (1970) Sperm competition and its evolutionary consequences in the insects. Biol Rev 45 : 525-567 Parker G A (1979) Sexual selection and sexual conflict. In: Blum M S , Blum, N A (eds) Sexual Selection and Reproductive Competition in Insects. Academic Press, New York, pp 123-166 Parker G A (1982) Why are there so many tiny sperm? Sperm competition and the maintenance of the two sexes. J Theor Biol 96 : 281-294 Parker G A (1984) Sperm competition and the evolution of animal mating strategies. In: Smith, R.L. (ed.). Sperm Competition and the Evolution of Animal Mating Systems. Academic Press, London, pp. 1-60 Parker G A (1990) Sperm competition games: raffles and roles. Proc R Soc Lond B 242 : 120126 78 Parker G A , Simmons L (1989) Nuptial feeding in insects: theoretical models of male and female interests. Ethology 82 : 3-26 Parker G A , Simmons L W , Kirk H (1990) Analyzing sperm competition data: simple models for predicting mechanisms. Behav Ecol Sociobiol 27 : 55-65 Pliske T E (1975) Courtship behavior of the monarch butterfly, Danaus plexippus L . Ann Entomol Soc A m 68 : 143-151 Royer L , McNeil J N (1993) Male investment in the European corn borer, Ostrinia nubilalis (Lepidoptera: Pyralidae): impact on female longevity and reproductive performance. Funct Ecol 7 : 209-215 Rutowski R L (1979) The butterfly as an honest salesman. Anim Behav 7 : 1269-1270 Rutowski R L (1980) Courtship solicitation by females of the checkered white butterfly, Pieris protodice. Behav Ecol Sociobiol 7 : 113-117 Rutowski R L (1984) Production and use of secretions passed by males at copulation in Pieris protodice (Lepidoptera, Pieridae). Psyche 91 : 141-152 Rutowski R L , Gilchrist G W , Terkanian B (1987) Female butterflies mated with recently mated males show reduced reproductive output. Behav Ecol Sociobiol 20 : 319-322 Rutowski R L , Long C E , Marshall L D , Vetter RS (1981) Courtship solicitation by Colias females (Lepidoptera: Pieridae). A m Midi Nat 105 : 334-340 Rutowski R L , Newton M , Schaefer J (1983) Interspecific variation in the size of the nutrient investment made by male butterflies during copulation. Evolution 37 : 708-713 Sillen-Tullberg B (1981) Prolonged copulation: A male "postcopulatory" strategy in a promiscuous species, Lygaeus equestris (Heteroptera: Lygaenidae). Behav Ecol Sociobiol 9 : 283-289 Simmons L W , Bailey W J (1990) Resource influenced sex role of Zaprochilinae tettigoniids (Orthoptera: Tettigoniidae). Evolution 44 : 1853-1868 Simmons L W , Parker G A (1989) Nuptial feeding in insects: Mating effort versus paternal investment. Ethology 81 : 332-343 Simmons L W , Parker G A (1992) Individual variation in sperm competition success of yellow dung flies, Scatophaga stercoraria. Evolution 46 : 366-375. Sims SR (1979) Aspects of mating frequency and reproductive maturity in Papilio zelicaon. A m Midi Nat 102:36-50 Smith R L (1984) (ed.) Sperm Competition and the Evolution of Animal Mating Systems. Academic Press, London, U . K . Sugawara T (1979) Stretch reception in the bursa copulatrix of the butterfly, Pieris rapae crucivora, and its role in behaviour. J Comp Physiol A l 30: 191-199 Svard L (1985) Paternal investment in a monandrous butterfly, Pararge aegeria. Oikos 45 : 66-70 79 Svard L , Wiklund C (1986) Different ejaculate strategies in first versus subsequent matings in the swallowtail butterfly Papilio machaon L.. Behav Ecol Sociobiol 18 : 325-330 Svard L, Wiklund C (1988) Fecundity, egg weight, and longevity in relation to multiple matings in females of the monarch butterfly. Behav Ecol Sociobiol 23 : 39-43 Svard L, Wiklund C (1989) Mass and production rate of ejaculates in relation to monandry/polyandry in butterflies. Behav Ecol Sociobiol 24 : 395-402 Svard L , Wiklund C (1991) The effect of ejaculate mass on female reproductive output in the European swallowtail butterfly, Papilio machaon (L.) (Lepidoptera: Papilionidae). J Insect Behav 4 : 33-41 S Y S T A T : (1992) Statistics, Version 5.2 Edition. Evanston, IL: SYSTAT, Inc., 724 pp Taylor OR Jr (1967) Relationship of multiple mating to fertility in Atteva punctella (Lepidoptera: Yponomeutidae). Ann Entomol Soc A m 60 : 583-590 Thornhill R (1976a) Sexual selection and paternal investment in insects. A m Nat 110 : 153163 Thornhill R (1976b) Sexual selection and nuptial feeding behavior in Bittacus apicalis (Insecta: Mecoptera). A m Nat 110 : 529-548 Thornhill R, Alcock J (1983) The Evolution of Insect Mating Systems. Cambridge, Massachusetts: Harvard University Press Tschudi-Rein K, Benz G (1990) Mechanisms of sperm transfer in female Pieris brassicae (Lepidoptera: Pieridae). Ann Entomol Soc A m 83 : 1158-1164 Walker W F (1980) Sperm utilization strategies in nonsocial insects. A m Nat 115: 780-799 Watanabe M (1988) Multiple matings increase the fecundity of the yellow swallowtail butterfly, Papilio xuthus L., in summer generations. J Insect Behav 1 : 17-30 Watanabe M , Ando S (1993) Influence of mating frequency on lifetime fecundity in wild females of the small white Pieris rapae (Lepidoptera, Pieridae). Jap J Ent 61 : 691696 Watanabe M , Ando S (1994) Egg load in wild females of the small white Pieris rapae crucivora (Lepidoptera, Pieridae) in relation to mating frequency. Jpn J Ent 62 : 293297 Watt W B , Hoch PC, M i l l SG (1974) Nectar resource use by Colias butterflies. Chemical and visual aspects. Oecologia 14 : 353-374 Wedell N (1991) Sperm competition selects for nuptial feeding in a bushcricket. Evolution 45 : 1975-1978 Wedell N (1993) Spermatophore size in bushcrickets: comparative evidence for nuptial gifts as a sperm protection device. Evolution 47 : 1203-1212 Wedell N (1994) Variation in nuptial gift quality in bushcrickets (Orthoptera: Tettigoniidae). Behav Ecol 5 : 418-425 Wedell N , Arak A (1989) The wartbiter spermatophore and its effect on female reproductive output (Orthoptera: Tettigoniidae, Decticus verrucivorus ). Behav Ecol Sociobiol 24 117-125 Wickman, P-O (1985) Male determined mating duration in butterflies? J Lep Soc 39 : 184185 Wickman, P-O, Karlsson B (1989). Abdomen size, body size and the reproductive effort of insects. Oikos 56 : 209-214 Wigglesworth V B (1972) The Principles of Insect Physiology. 7th ed. Chapman and Hall, London Wiklund C (1977) Courtship behaviour in relation to female monogamy in Leptidea sinapis. Oikos 29 : 275-283 Wiklund C (1982) Behavioural shift from courtship solicitation to mate avoidance in female ringlet butterflies (Aphantopus hyperanthus) after copulation. Anim Behav 30 : 790793 Wiklund C, Forsberg J (1985) Courtship and mate discrimination between virgin and mated females in the orange tip butterfly Anthocharis cardamines. Anim Behav 34 : 328332 Wiklund C, Karlsson B , Forsberg J (1987) Adaptive versus constraint explanations for eggto-body size relationships in two butterfly families. A m Nat 130 : 828-838 Wiklund C, Forsberg J (1991) Sexual size dimorphism in relation to female polygamy and protandry in butterflies: a comparative study of Swedish Peiridae and Satyridae. Oikos 60: 373-381 Wiklund C, Nylin S, Forsberg J (1991) Sex-related variation in growth rate as a result of selection for large size and protandry in a bivoltine butterfly, Pieris napi. Oikos 60 : 241-250 Wiklund C, Kaitala A , Lindfors V, Abenius J (1993) Polyandry and its effect on female reproduction in the green-veined white butterfly (Pieris napi L.). Behav Ecol Sociobiol 33 : 25-33 Wiklund C, Kaitala A (1995) Sexual selection for large male size in a polyandrous butterfly: the effect of body size on male versus female reproductive success in Pieris napi. Behav Ecol 6:6-13 Williams G C (1975) Sex and Evolution. Princeton Univ Press, Princeton, NJ. APPENDIX 1 Samples of the standard curves used for the a) standard assay and b) microassay. In both cases bovine serum albumin was used as the standard. a) WORKING CU RUE C=K*ABS+B 0 .0 C 0 N C. c K= 1 . 9 7 ~ B=- 9 .9943 + 1 . 40A . 290 <.A/D I U . > I + 0 . 20A A 5/0 1 b ) • 5 9 5 . 0NM '? 3 MHMJMililfaHBlM: 1.4000 0 . 2 890 / D I U. C=K*A6S+! K= 95.696 -0.064ft B= 13.725 + 0 . 04A 0.020 C A / D I U .. ) - 0 . 14A X 0 13^35 8^27 .a '?3 CONC. 3 .0000 D I U. 5 9 5 . 3N11 15.000 0 .017A