Abstract

Finding, assessing, rejecting, and copulating with a mate is assumed to carry fitness costs, particularly for females, that have to be traded off against fitness benefits of mating such as increased fecundity, fertility, longevity, or better quality offspring. Female dung flies, Sepsis cynipsea (Diptera: Sepsidae), typically attempt to dislodge mounted males harassing them by vigorous shaking. Shaking duration has been shown to reflect both direct and indirect female choice in this species. The latter is an expression of the females' general reluctance to mate due to presumed costs of mating. We investigated the costs of copulation in the laboratory. Females were randomly assigned to one of three treatment groups and allowed to copulate either not at all, once, or twice. The males' armored genitalia injured females internally during copula. Injuries were visible as sclerotized scars in the female ovipositor, and their occurrence increased with mating frequency. Presumably due to these injuries, mated females showed higher mortality. This effect was statistically independent from additional costs of reproduction related to oviposition, as copulation also increased lifetime egg production and tended to augment oviposition rate (eggs per day), while fertility (proportion of offspring emerged) was unaffected. We thus found high mortality costs of copulating, indicating substantial sexual conflict, which helps explain female reluctance to mate in this species.

Mating is a costly activity. Over the past decades, a slow paradigm shift has occurred from regarding mating as a process of mutual benefit to one loaded with conflict (Arnqvist, 1989; Eberhard, 1996; Gowaty and Buschhaus, 1998; Holland and Rice, 1998; Parker, 1979; Partridge and Hurst, 1998; Rice, 1996; Rowe et al., 1994; Thornhill and Alcock, 1983). It is now accepted that finding, assessing, rejecting, and copulating with a mate can carry fitness costs, particularly for females, which have to be traded off against fitness benefits of mating such as increased fecundity, fertility, longevity, or better quality offspring (Andersson, 1994; Arnqvist and Nilsson, 2000; Chapman et al., 1998; Johnstone and Keller, 2000). Mating costs include increased predation due to reduced mobility or greater visibility (Arnqvist, 1989; Fairbairn, 1993; Gwynne, 1989; Hosken et al., 1994; Magnhagen, 1991; Moore, 1987; Rowe, 1994; Ward, 1986), disease transmission (Daly, 1978; Kaltz and Schmid, 1995), or loss of energy or foraging time (Bailey et al., 1993; Clutton-Brock and Langley, 1997; Cordts and Partridge, 1996; Watson et al., 1998; Wilcox, 1984). Furthermore, deleterious effects of toxic accessory gland products (Chapman et al., 1995, 1998; Chen, 1984), internal (i.e., genital: Crudgington and Siva-Jothy, 2000; Kummer, 1960; Merritt, 1989; von Helversen and von Helversen, 1991), or external injuries (e.g., wing: Cartar, 1992; body: LeBoeuf and Mesnick, 1991; Leong et al., 1993; Michiels and Newman, 1998) have been reported, all of which may increase mortality. The costs and benefits of mating need to be evaluated as comprehensively as possible to fully understand the mating system of any particular species and the evolution of sexual conflict (Arnqvist and Nilsson, 2000; Eberhard, 1996; Johnstone and Keller, 2000; Parker, 1979). Recent evidence suggests the intriguing possibility that males actively harm females (Crudgington and Siva-Jothy, 2000; Johnstone and Keller 2000; Michiels and Newman, 1998), using armored genitalia to increase their reproductive success (which have been known for quite some time; Eberhard, 1985).

In many species, (presumed) mating costs can result in obvious female reluctance to mate (Crudgington and Siva-Jothy, 2000; Godin and Briggs, 1996; Rowe et al., 1994; Thornhill and Alcock, 1983). This is the case in the dung fly Sepsis cynipsea (Diptera: Sepsidae), in which females shake vigorously, attempting to dislodge males trying to copulate with them (Blanckenhorn et al., 2000; Parker, 1972a,b; Ward, 1983, Ward et al., 1992). Ward et al. (1992) suggested that females may minimize their copulation frequency to reduce the risk of internal injury by the male's armored genitalia (Figure 1a) because internal injuries may cause mortality or decrease fecundity. These are the costs of copulating. Female rejection behavior also may carry costs in terms of increased predation risk, energy expenditure, or wing injuries inflicted by spines on the male's forelegs (Blanckenhorn et al., 1998; Hennig, 1949; Mühlhäuser and Blanckenhorn, 2002; Pont, 1979). These are the costs of assessing and/or rejecting a mate before copulation, which likely vary in space and time. The mating system of S. cynipsea thus suggests substantial sexual conflict (Gowaty and Buschhaus, 1998; Holland and Rice, 1998; Partridge and Hurst, 1998; Rowe et al., 1994). It seems obvious that female reluctance to mate can only evolve if the cost of avoiding matings does not exceed the cost of copulation; otherwise females would try to offset one cost with another (Blanckenhorn et al., 2000). Therefore, both types of cost need to be evaluated in the same currency (Rowe, 1994; Watson et al., 1998).

Figure 1

(a) Male genitalia (aedeagus) with numerous spiny chitinous structures (see enlargement), and (b) sclerotized scar in the female's ovipositor presumably inflicted by these structures.

Figure 1

(a) Male genitalia (aedeagus) with numerous spiny chitinous structures (see enlargement), and (b) sclerotized scar in the female's ovipositor presumably inflicted by these structures.

In this study we experimentally assessed the costs and the benefits of copulating in S. cynipsea. We compared the mortality, fecundity, and fertility of females mated zero, one, and two times in the laboratory. We also screened females specifically for genital injuries in an attempt to link the expected reproductive costs of copulation to these injuries. The costs of avoiding and/or assessing mates were also investigated and are presented in a companion paper (Mühlhäuser and Blanckenhorn, 2002).

METHODS

We performed two laboratory experiments, one to assess female internal injuries and another to assess the reproductive consequences of copulation frequency. The two experiments were performed separately because from preliminary work we knew that dead individuals disintegrate quickly (within 24 h) so that internal injuries are no longer possible to score reliably. We therefore deemed it necessary to experimentally assess these injuries in a standardized manner (i.e., after a predetermined time and not at the end of a female's life).

Female internal injuries as a function of copulation frequency

Females emerging over a period of 3 days from S. cynipsea laboratory stocks originating from our field site in Fehraltorf, near Zürich, Switzerland, were isolated within 24 h of emergence. They were randomly assigned to one of three treatment groups and allowed to copulate either not at all, once, or twice. As even virgins do not mate readily (Blanckenhorn et al., 2000), individual females aged 5-10 days were placed with three to six randomly chosen males from the same population (to increase the likelihood of mating) in a 50-ml glass vial containing one large grain of sugar and a miniature dish (ca. 5 g) of recently defrosted cow dung, their oviposition medium. As soon as copulation was initiated, all other males were removed, and after copulation ended (which lasts about 20 min), the females were kept singly in their vial until they were frozen for later analysis of internal injuries. When no copulation occurred within 90 min, the same females underwent the same procedure once again 2-3 days later. A random subset of the females that copulated once were then subjected to the same procedure 3-5 days later so they could copulate a second time. Females were thus in contact with males for at most two times 90 min, during which they were under constant observation. Females were frozen for later analysis 4-9 days after their last (i.e., first or second) copulation. Virgins of the zero-copulation treatment were never confronted with males and frozen 7-20 days after the first copulation of the females in the other treatments to keep life span approximately equal across all treatment groups. We deemed 4-7 days necessary and sufficient for female internal injuries to heal and develop scars at the climate conditions given below so they could later be scored (see Crudgington and Siva-Jothy, 2000). Females that died within 4 days after copulation were discarded because they could not be reliably scored (see above), as were females preassigned to the copulation treatments that did not copulate at all during two trials (see Mühlhäuser and Blanckenhorn, 2002).

To assess internal injuries, each female was placed on a slide, in a drop of Ringer solution, under a stereo microscope. The protruding end of the ovipositor was grasped with fine tweezers and the whole reproductive tract pulled out. Using a microscope, the ovipositor was then examined for scars. All females were scored blindly (i.e., without knowledge of their copulation frequency) by the same person. We analyzed differences in the extent of internal scars (presence or absence) among the three treatment groups using stepwise (backward elimination) logistic regression with treatment as a fixed factor and female age at freezing as a covariate.

Female reproductive success as a function of copulation frequency

For this separate experiment, three treatment groups (zero, one, and two copulations) were generated as for the internal injury experiment. Again, care was taken to randomize age and size among treatments (see Results). After their last copulation, females were kept singly in their vials until their death, in a climate chamber at 25°C, 50% relative humidity, and 14-h light period. Females of the zero-copulation treatment were held likewise. They had access to ad libitum sugar and a miniature dish of defrosted cow dung twice a week for food, moisture, and oviposition. For each female, we scored her day of death, the total number of eggs laid over her life-time (and per week), and her fertility as the percentage of offspring emerged per week from her eggs, at the environmental conditions given above.

For analysis, we subdivided the reproductive data for all females of all treatment groups into two time periods, the period before the date of the second copulation of the females in the two-copulation treatment group (henceforth the date of second copulation), and the period thereafter. Only the latter period was analyzed in most cases. This was done to control for systematic age differences among the treatment groups, as effects of copulation(s) on reproductive performance necessarily appear later in the females that copulated more often. All females that died before the date of second copulation were discarded, except that their mortality (dead or alive) was related to whether they had copulated once or not, using logistic regression. Females that survived this date were also discarded if they never laid any eggs (n = 14 females, 13%), as they were apparently unable to reproduce due to unknown causes and thus were uninformative with regard to the study questions.

Individual fecundity was expressed both as total eggs laid before/after the date of second copulation and eggs laid per day (i.e., mean oviposition rate: total eggs laid divided by days of life before/after the date of second copulation). Both these estimates of fecundity after the date of second copulation were compared among the three treatment groups using one-way factorial ANOVA (copulation treatment) with days of life before and after the date of second copulation, body size (head width), and fecundity before the date of second copulation as covariates. We did this to control for differences among the females in emergence date and life span, as well as to control for potential costs of reproduction relating to the period before the date of second copulation. In addition, we analyzed lifetime fecundity with total life span and head width as covariates (combining the estimates before and after the date of second copulation).

Fertility, a log[p/ (1 — p)]-transformed proportion p, was expected to decline with age due to sperm depletion or aging and was calculated per week and compared only for the females which copulated once and twice (as fertility of the virgins was necessarily zero). Fertility was regressed on week with copulation treatment as a factor and days of life before and after the date of second copulation, body size (head width), and fecundity before and after the date of second copulation as covariates. As there is no repeated-measures regression, this analysis was additionally performed using the mean fertility per week of all females of a copulation treatment, without covariates, to avoid pseudoreplication. To assess the immediate effect of the second copulation on fertility, we additionally compared only the fertility during the 7 days immediately before and after the date of second copulation, using repeated-measures ANOVA with treatment as the between-subject and week as the repeated factor, and the same covariates as above. Only females that laid eggs during both weeks were included in this analysis.

We analyzed survivorship after the date of second copulation using Cox regression (the appropriate analysis for survivorship curves) with treatment as a fixed discrete factor, and days of life before the date of second copulation, body size (head width), and fecundity as covariates. The latter tests for mortality costs of reproduction and was expressed either as mean oviposition rate or total number of eggs laid during the entire lifetime (thus combining the estimates before and after the date of second copulation). In all analyses, total eggs laid, oviposition rate, and days of life were square-root transformed to equalize variances.

RESULTS

Female internal injuries as a function of copulation frequency

We detected scars in the female ovipositor (Figure 1b), the occurrence of which increased with copulation frequency (Figure 2; logistic regression: χ 2 = 8.77, p =.012, n = 131; the effect of age at freezing was eliminated from the final model). Only 1 of 28 virgins had a scar (for unknown reasons), whereas 6 (of 61) females of the one-copulation and 12 (of 42) of the two-copulation had a scar. This sample includes 37 females in the one-copulation treatment that refused a second copulation (i.e., were not preassigned). However, the result was qualitatively the same when only analyzing those females originally preassigned to the one copulation treatment (χ2 = 6.45, p =.039, n = 94).

Figure 2

Percentage of females internally scarred as a function of copulation frequency.

Figure 2

Percentage of females internally scarred as a function of copulation frequency.

Female reproductive success as a function of copulation frequency

Copulation had an obvious and drastic immediate effect on female mortality. Of the original 43 females preassigned to the zero-copulation treatment and the 125 females preassigned to both copulation treatments, respectively, 8 (18.6%) and 51 (40.8%) died before the date of second copulation. Females that copulated had a 67% greater probability of dying within a week after copulation (logistic regression: χ2 = 5.00, p =.025, odds ratio = 1.67; positive effect of female age: χ2 = 4.42, p =.036, odds ratio = 1.09, n = 168). For the remaining females that survived this date and laid eggs, Cox regression indicates a further effect of copulation on survivorship after the date of second copulation (χ2 = 15.67, p =.004, n = 91; Figure 3, Table 1): relative to virgins, females that copulated once had a 21% (odds ratio = 1.21) and those that copulated twice a 88% greater mortality rate (odds ratio = 1.88). (Note that this analysis does not include the many females that copulated once and died before the date of second copulation.) The same analysis showed (expected) elevated mortality rates when females were older at the date of second copulation (χ2 = 8.78, p =.003, odds ratio = 1.62) and when they had higher lifetime fecundity (χ2 = 21.48, p <.001, odds ratio = 1.15), but there were no effects of body size (χ2 = 0.32, p =.572, odds ratio = 0.98). (Using mean oviposition rate yielded not quite as strong, but qualitatively similar, results, except that the effect of fecundity was no longer significant.)

Figure 3

Proportion of females surviving as a function of time and copulation frequency.

Figure 3

Proportion of females surviving as a function of time and copulation frequency.

Table 1

Mean ± SE reproductive success before and after the date of second copulation and body size by copulation treatment

  Before   After  
aComparing only the 7 days immediately before and after the second copulation and including only females which laid eggs before and after.  
Days of life    
0 copulations (n = 28)   15.54 ± 0.51   27.13 ± 2.38  
1 copulation (n = 41)   14.26 ± 0.61   25.23 ± 2.02  
2 copulations (n = 22)   14.04 ± 0.43   21.85 ± 2.89  
Total eggs laid    
0 copulations (n = 28)   34.31 ± 6.25   72.67 ± 11.21  
1 copulation (n = 41)   54.05 ± 7.78   122.34 ± 16.67  
2 copulations (n = 22)   60.84 ± 6.56   134.97 ± 16.80  
Eggs per day    
0 copulations (n = 28)   2.20 ± 0.42   2.84 ± 0.36  
1 copulation (n = 41)   4.13 ± 0.61   4.84 ± 0.63  
2 copulations (n = 22)   4.51 ± 0.50   6.05 ± 0.66  
Proportion offspring emergeda   
0 copulations (n = 11)   0.00 ± 0.00   0.00 ± 0.00  
1 copulation (n = 30)   0.70 ± 0.05   0.53 ± 0.05  
2 copulations (n = 19)   0.79 ± 0.04   0.71 ± 0.06  
Head width (mm)    
0 copulations (n = 28)   0.92 ± 0.01   —  
1 copulation (n = 41)   0.93 ± 0.01   —  
2 copulations (n = 22)   0.95 ± 0.01   —  
  Before   After  
aComparing only the 7 days immediately before and after the second copulation and including only females which laid eggs before and after.  
Days of life    
0 copulations (n = 28)   15.54 ± 0.51   27.13 ± 2.38  
1 copulation (n = 41)   14.26 ± 0.61   25.23 ± 2.02  
2 copulations (n = 22)   14.04 ± 0.43   21.85 ± 2.89  
Total eggs laid    
0 copulations (n = 28)   34.31 ± 6.25   72.67 ± 11.21  
1 copulation (n = 41)   54.05 ± 7.78   122.34 ± 16.67  
2 copulations (n = 22)   60.84 ± 6.56   134.97 ± 16.80  
Eggs per day    
0 copulations (n = 28)   2.20 ± 0.42   2.84 ± 0.36  
1 copulation (n = 41)   4.13 ± 0.61   4.84 ± 0.63  
2 copulations (n = 22)   4.51 ± 0.50   6.05 ± 0.66  
Proportion offspring emergeda   
0 copulations (n = 11)   0.00 ± 0.00   0.00 ± 0.00  
1 copulation (n = 30)   0.70 ± 0.05   0.53 ± 0.05  
2 copulations (n = 19)   0.79 ± 0.04   0.71 ± 0.06  
Head width (mm)    
0 copulations (n = 28)   0.92 ± 0.01   —  
1 copulation (n = 41)   0.93 ± 0.01   —  
2 copulations (n = 22)   0.95 ± 0.01   —  

Oviposition rate (eggs per day) after the date of second copulation did not decrease with copulation frequency; if anything, oviposition rate showed a tendency to increase (F2,83 = 2.44; p =.093; Table 1). Virgins also laid eggs, though at a reduced rate (planned contrasts; zero vs. one and two copulations: t83 = 4.25, p <.001; one vs. two copulations: t83 = 1.44, p =.154; Table 1). The same analysis showed an expected positive correlation between fecundity before and after the date of second copulation (partial r =.28, p =.005), and a negative effect of age at the date of second copulation on oviposition rate (partial r = -.20, p =.030; all other covariates p > .2). Total eggs laid after the date of second copulation (i.e., residual lifetime reproductive success) increased with copulation treatment (F2,83 = 4.41; p =.015; effect of eggs laid before the date of second copulation: partial r =.22, p =.004; effect of days of life after the date of second copulation: partial r =.70, p <.001; effect of days of life before the date of second copulation: partial r = -.13, p =.066; effect of head width: partial r =.17, p =.091; Table 1). Analogous analysis of the lifetime number of eggs laid, with life span and head width as covariates, yielded qualitatively similar results: egg output increased with copulation treatment (F2,85 = 12.82; p <.001), correlated strongly with life span (partial r =.46, p <.001), and also correlated with head width (partial r =.28, p =.012). The mean ± SE lifetime egg output of all females that laid eggs, including those that died before the date of second copulation, as a function of copulation treatment was 103.84 ± 14.11 (n = 29) for the zero-copulation, 153.40 ± 18.51 (n = 49) for the one-copulation, and 151.68 ± 19.53 (n = 30) for the two-copulation treatment (F2,104 = 3.34; p =.039). This estimate includes the probability of dying after the first copulation, by randomly allocating those females that died after the first copulation to the one or two copulation treatments in proportion to the treatment frequency.

We expected a second copulation to enhance fertility, counteracting sperm depletion (Arnqvist and Nilsson, 2000). This should have resulted in a steeper decline in fertility with time in the one- compared to the two-copulation treatment (i.e., an interaction of treatment with week in our multiple regression). Fertility obviously declined with time (age) due to sperm depletion (raw data: F1,238 = 138.56; mean data: F1,12 = 158.3; both p <.001), but there was no interaction of treatment and time (raw data: F1,238 = 0.18; p =.670; mean data: F1,12 = 0.95; p =.350; Figure 4). However, there was an effect of copulation (raw data: F1,238 = 8.28; p =.004; mean data: F1,12 = 6.24; p =.028), as females that copulated twice were more fertile to begin with, perhaps by chance (Figure 4). Comparing only the weeks immediately before and after the date of second copulation showed no effects of copulation on fertility whatsoever (effect of treatment: F1,42 = 2.52; p =.122; effect of time: F1,42 = 1.93; p =.172; interaction: F1,42 = 0.30; p =.589; all covariates p > .2; Table 1).

Figure 4

Proportion of offspring emerging (fertility) as a function of time for females that copulated once and twice. The vertical line shows approximately when the second copulation occurred.

Figure 4

Proportion of offspring emerging (fertility) as a function of time for females that copulated once and twice. The vertical line shows approximately when the second copulation occurred.

DISCUSSION

Our study revealed drastic costs of mating in S. cynipsea, both in terms of increased mortality immediately after the first copulation and decreased survivorship given a second copulation. These may relate partly to standard costs of reproduction, as we also found increases in fecundity with copulation frequency. Physiological substances transferred by the male during copulation that stimulate oviposition (and sometimes also increase mortality) may be responsible (e.g., Drosophila spp.: Baumann, 1974; Chapman et al., 1995, 1998; Chen, 1984). However, there was an additional, statistically independent effect of copulation on mortality, which we believe is linked to the male's armored genitalia injuring the females. The sclerotized scars in the female ovipositor (Figure 1b) appear to have been caused by the spines of the male's copulating organ, the aedeagus (Figure 1a). Of course, this connection can only be inferred by correlation, as done here and in a recent study by Crudgington and Siva-Jothy (2000). Alternatively, mortality could be a side effect of male substances that reduce female receptivity (Chapman et al., 1995, 1998), but such substances remain to be identified in this species. These mortality costs of copulating are sufficient to explain the strong female reluctance to mate in S. cynipsea (Blanckenhorn et al., 2000; Ward et al., 1992).

We do not know the precise function of the male's armored genitalia. They may function to remove sperm from previous males (e.g., damselfly: Waage, 1979) or to inject accessory gland material (e.g., sheep blowfly: Merritt, 1989), but to date we have no evidence for either phenomenon in S. cynipsea. A possible general function of thorns and spines is to lock the male's genitals within the female tract, making it difficult, if not impossible, for her to eject him (Eberhard, 1985). Sepsis cynipsea males have to twist about 180° to disengage from copula (Parker, 1972a). In our laboratory cultures, we have repeatedly (but rarely) observed dead males permanently attached to the female's abdomen. This suggests copulation is also costly for males, but additionally that females are unable to disengage from copula without the male's assistance. Sepsis cynipsea is thus another case of a growing number of species showing obvious sexual conflict in genital morphology, reproductive physiology, or mating behavior (Chapman et al., 1998; Gowaty and Buschhaus, 1998; Holland and Rice, 1998; Partridge and Hurst, 1998; Rowe et al., 1994).

How can a male gain from actively harming his mate? Three related arguments have been put forward (Crudgington and Siva-Jothy, 2000; Johnstone and Keller, 2000). By making mating costly and females reluctant to mate again, successful males (1) indirectly reduce the risk of sperm competition and thus increase their relative share of paternity, and (2) may indirectly force females to augment oviposition rates because of the imminent threat of her death soon after copulation (provided, of course, the female survives copulation; Polak and Starmer, 1998). (3) Males may also stimulate female ovipoistion and reduce their recipitivity directly using physiological accessory gland stimulants (Chapman et al., 1998; see below). In the first two cases, the male's strategy of producing possibly lethal injuries is the necessary selection pressure to modify female behavior; in the third case female mortality may just be a side effect. Johnstone and Keller (2000) have recently shown theoretically that such spiteful male behavior can evolve if males inflict strongly accelerating costs on females that remate. This might be the case here (see Figure 2), but S. cynipsea females typically copulate too rarely to test this idea quantitatively. Nevertheless, females can apparently avoid internal injuries during copulation, as a majority of them did not show any. Also, the wounds inflicted can effectively heal, although healing will likely carry a cost. Aside from these very general counteradaptations, S. cynipsea females at this point may be able to do little more than shake off males to avoid superfluous matings (Blanckenhorn et al., 2000; Parker, 1972a,b; Ward et al., 1992), but by doing so they incur other potential costs (Mühlhäuser and Blanckenhorn, 2002).

Although our study demonstrated costs of mating in terms of survivorship, there were no such costs in terms of fecundity or fertility. For example, genital injuries may cause oviposition problems and thus reduce fecundity. On the other hand, mating multiply in many species confers benefits in terms of increased fecundity or fertility (Arnqvist and Nilsson, 2000). This may be simply because more sperm allows fertilization of more eggs, but also because male accessory substances may increase female oviposition rate (e.g., Chapman et al., 1995, 1998). Whether male accessory substances are transferred during copula and effective in S. cynipsea remains to be demonstrated. Our study revealed increased fecundity with copulation frequency, but no benefits in terms of fertility. As mentioned above, females might increase their reproductive effort because of the elevated mortality risk after copulation, an effect demonstrated in another fly (Polak and Starmer, 1998).

Because mating confers both costs and benefits, an optimal copulation frequency likely exists for a given species or individual (Arnqvist and Nilsson, 2000), but this will most certainly be different for males and females (Andersson, 1994; Bateman, 1948; Darwin, 1871), generating the mating conflict apparent in this and many other species. Whereas males are generally expected to maximize their number of matings (Andersson, 1994; Bateman, 1948), females need at least one copulation, but perhaps few more. Predicting the optimal copulation frequency for females involves assessing and trading off the costs and benefits of copulating (Arnqvist and Nilsson, 2000). Here we found a decrease in survivorship after the date of second copulation but at the same time an increase in oviposition rate, resulting in overall increased lifetime fecundity (Table 1). However, when accounting for the large increase in the probability of female death after the first copulation, the latter was entirely caused by the reduced fecundity of virgins. (We were surprised to find virgins reproducing at all; apparently they cannot detect their virginity, halt egg production, or resorb unfertilized eggs.) Calculating trade-offs is further complicated because the costs of rejecting and assessing a mate also have to be included (Blanckenhorn et al., 2000; Mühlhäuser and Blanckenhorn, 2002), and all these costs and benefits are difficult to estimate in the same fitness currency (such as the probability of death or lifetime reproductive success). Qualitatively at least, the costs of copulating documented in this study are more obvious than the costs of mating (i.e., shaking) behavior (see Mühlhäuser and Blanckenhorn, 2002) and help explain female reluctance to mate in S. cynipsea and perhaps many other species (Blanckenhorn et al., 2000; Crudgington and Siva-Jothy, 2000; Rowe et al., 1994; Ward et al., 1992).

We thank the Swiss National Science Foundation for funding and A. Bourke, L. Rowe, and an anonymous referee for comments.

References

Andersson M,
1994
.
Sexual selection
 . Princeton, New Jersey: Princeton University Press.
Arnqvist G,
1989
. Multiple mating in a water strider: mutual benefits or inter-sexual conflict?
Anim Behav
 
38
:
749
-756.
Arnqvist G, Nilsson T,
2000
. The evolution of polyandry: multiple mating and female fitness in insects.
Anim Behav
 
60
:
145
-164.
Bailey WJ, Withers PC, Endersby M, Gaull K,
1993
. The energetic costs of calling in the bushcricket Requena verticalis (Orthoptera: Tettigoniidae: Listroscelidinae).
J Exp Biol
 
178
:
21
-37.
Bateman AJ,
1948
. Intra-sexual selection in Drosophila.
Heredity
 
2
:
349
-368.
Baumann H,
1974
. Biological effects of paragonial substances PS-1 and PS-2 in females of Drosophila funebris.
J Insect Physiol
 
20
:
2347
-2362.
Blanckenhorn WU, Mühlhäuser C, Morf C, Reusch T, Reuter M,
2000
. Female choice, female reluctance to mate and sexual selection on body size in the dung fly Sepsis cynipsea.
Ethology
 
106
:
577
-593.
Blanckenhorn WU, Mühlhäuser C, Reusch T,
1998
. Fluctuating asymmetry and sexual selection in the dung fly Sepsis cynipsea—testing the good genes assumptions and predictions.
J Evol Biol
 
11
:
735
-753.
Cartar RV,
1992
. Morphological senescence and longevity: an experiment relating wing wear and life span in foraging wild bumble bees.
J Anim Ecol
 
61
:
225
-231.
Chapman T, Liddle LF, Kalb JM, Wolfner MF, Partridge L,
1995
. Cost of mating in Drosophila melanogaster females is mediated by male accessory gland products.
Nature
 
373
:
241
-244.
Chapman T, Miyatake T, Smith HK, Partridge L,
1998
. Interactions of mating, egg production and death rates in females of the Mediterranean fruit fly, Ceratitis capitata.
Proc R Soc Lond B
 
265
:
1879
-1894.
Chen PS,
1984
. The functional morphology and biochemistry of insect male accessory glands and their secretions.
Annu Rev Entomol
 
29
:
233
-255.
Cordts R, Partridge L,
1996
. Courtship reduces longevity of male Drosophila melanogaster.
Anim Behav
 
52
:
269
-278.
Clutton-Brock T, Langley P,
1997
. Persistent courtship reduces male and female longevity in captive tsetse flies Glossina morsitans morsitans.
Behav Ecol
 
8
:
392
-395.
Crudgington HS, Siva-Jothy MT,
2000
. Genital damage, kicking and early death.
Nature
 
407
:
655
-656.
Daly M,
1978
. The cost of mating.
Am Nat
 
112
:
771
-774.
Eberhard WG,
1985
.
Sexual selection and animal genitalia
 . Cambridge: Harvard University Press.
Eberhard WG,
1996
.
Female control: sexual selection by cryptic female choice
 . Princeton, New Jersey: Princeton University Press.
Fairbairn DJ,
1993
. The costs of loading associated with mate-carrying in the waterstrider Aquarius remigis.
Behav Ecol
 
4
:
224
-231.
Godin JG, Briggs SE,
1996
. Female mate choice under predation risk in the guppy.
Anim Behav
 
51
:
117
-130.
Gowaty PA, Buschhaus N,
1998
. Ultimate causation of aggressive and forced copulation in birds: female resistance, the CODE hypothesis, and social monogamy.
Am Zool
 
38
:
207
-225.
Gwynne DT,
1989
. Does copulation increase the risk of predation?
Trends Ecol Evol
 
4
:
54
-56.
Henning W,
1949
. Sepsidae. In:
Die Fliegen der paläarktischen Region, Band V
  (Lindner W, ed). Stuttgart: Schweizerbart'sche Verlagsbuchhandlung,
1
-90.
Holland B, Rice WR,
1998
. Chase-away sexual selection: antagonistic seduction versus resistance.
Evolution
 
52
:
1
-7.
Hosken DJ, Bailey WJ, O'Shea JE, Roberts JD,
1994
. Localization of insect calls by the bat Nyctophilus geoffroyi (Chiroptera: Vespertilionidae): a laboratory study.
Aust J Zool
 
42
:
177
-184.
Johnstone RA, Keller L,
2000
. How males can gain by harming their mates: sexual conflict, seminal toxins, and the cost of mating.
Am Nat
 
156
:
368
-377.
Kaltz O, Schmid B,
1995
. Plant venereal disease: a model for integrating genetics, ecology and epidemiology.
Trends Ecol Evol
 
10
:
221
-222.
Kummer H,
1960
. Experimentelle Untersuchungen zur Wirkung von Fortpflanzungs-Faktoren auf die Lebensdauer von Drosophila melanogaster-Weibchen.
Z Vgl Physiol
 
43
:
642
-679.
LeBoeuf BJ, Mesnick SL,
1991
. Sexual behavior of male northern elephant seals. I. Lethal injuries to adult females.
Behaviour
 
116
:
142
-162.
Leong KLH, Frey D, Hamaoka D, Honma K,
1993
. Wing damage in overwintering populations of the monarch butterfly at two California sites.
Ann Entomol Soc Am
 
86
:
728
-733.
Magnhagen C,
1991
. Predation risk as a cost of reproduction.
Trends Ecol Evol
 
6
:
183
-185.
Merritt DJ,
1989
. The morphology of the phallosome and accessory gland material transfer during copulation in the blowfly, Lucilia cuprina (Insecta, Diptera).
Zoomorphology
 
108
:
359
-366.
Michiels NK, Newman LJ,
1998
. Sex and violence in hermaphrodites.
Nature
 
391
:
647
.
Moore SM,
1987
. Male-biased mortality in the butterfly Euphydryas editha: a novel cost of mate acquisition.
Am Nat
 
130
:
306
-309.
Mühlhäuser C, Blanckenhorn WU,
2002
. The costs of avoiding matings in the dung fly Sepsis cynipsea.
Behav Ecol
 
13
:
359
-365.
Parker GA,
1972
. Reproductive behaviour of Sepsis cynipsea (L) (Diptera: Sepsidae) I. A preliminary analysis of the reproductive strategy and its associated behavior patterns.
Behaviour
 
41
:
172
-206.
Parker GA,
1972
. Reproductive behaviour of Sepsis cynipsea (L) (Diptera: Sepsidae) II. The significance of the precopulatory passive phase and emigration.
Behaviour
 
41
:
242
-250.
Parker GA,
1979
. Sexual selection and sexual conflict. In:
Sexual selection and reproductive competition in insects
 . (Blum NS, Blum NA, eds). New York: Academic Press,
123
-166.
Partridge L, Hurst LD,
1998
. Sex and conflict.
Science
 
281
:
2003
-2008.
Polak M, Starmer WT,
1998
. Parasite induced risk of mortality elevates reproductive effort in male Drosophila.
Proc R Soc Lond B
 
265
:
2197
-2201.
Pont AC,
1979
. Sepsidae Diptera, Cyclorrhapha, Acalyptrata. In:
Handbook for the identification of British insects
 , vol.
X
, part 5 (Fitton MG, ed). London: Royal Entomological Society;
1
-35.
Rice WR,
1996
. Sexually antagonistic male adaptation triggered by experimental arrest of female evolution.
Nature
 
381
:
232
-234.
Rowe L,
1994
. The costs of mating and mate choice in water striders.
Anim Behav
 
48
:
1049
-1056.
Rowe L, Arnqvist G, Sih A, Krupa JJ,
1994
. Sexual conflict and the evolutionary ecology of mating patterns: water striders as a model system.
Trends Ecol Evol
 
9
:
289
-294.
Thornhill R, Alcock J,
1983
.
The evolution of insect mating systems
 . Cambridge: Harvard University Press.
von Helversen D, von Helversen O,
1991
. Pre-mating sperm removal in the bushcricket Metaplastes ornatus Ramme 1931 (Orthoptera, Tettigonoidea, Phaneropteridae).
Behav Ecol Sociobiol
 
28
:
391
-396.
Waage J,
1979
. Dual function of the damselfly penis: sperm removal and transfer.
Science
 
203
:
916
-918.
Ward PI,
1983
. The effects of size on the mating behavior of the dung fly Sepsis cynipsea.
Behav Ecol Sociobiol
 
13
:
75
-80.
Ward PI,
1986
. A comparative field study of the breeding behaviour of a stream and a pond population of Gammarus pulex (Amphipoda).
Oikos
 
46
:
29
-36.
Ward PI, Hemmi J, Röösli T,
1992
. Sexual conflict in the dung fly Sepsis cynipsea.
Funct Ecol
 
6
:
649
-653.
Watson PJ, Arnqvist G, Stallmann RR,
1998
. Sexual conflict and the energetic costs of mating and mate choice in water striders.
Am Nat
 
151
:
46
-58.
Wilcox RS,
1984
. Male copulatory guarding enhances female foraging in a water strider.
Behav Ecol Sociobiol
 
15
:
171
-174.