- Email: [email protected]
Variation in sperm precedence during mating in male flies,Dryomyza anilis
Anim. Behav., 1997, 53, 1233–1240
Variation in sperm precedence during mating in male flies, Dryomyza anilis MERJA OTRONEN Department of Zoology, Uppsala University (Received 14 March 1996; initial acceptance 23 May 1996; final acceptance 30 August 1996; MS. number: 5191)
Abstract. To understand how individual differences in fertilization success arises in the fly Dryomyza anilis, variation in sperm precedence between and within males was studied. In D. anilis, a mating consists of a copulation followed by tapping sequences during which the male taps the external genitalia of the female. These tapping sequences increase last male sperm precedence. Each male in the experiment was repeatedly mated with a set of females. In the first treatment, mating was interrupted after the male had an intromission. In the second treatment, the male was allowed to perform five tapping sequences after an intromission. In the third treatment, the male was allowed to perform an unlimited number of tapping sequences after an intromission until interrupted by female resistance. There were large differences between males in all three treatments as shown by the 95% confidence limits of the mean fertilization success. Repeatabilities of mating time and the number of tapping movements per tapping sequence were high indicating clear differences between males in these components of mating behaviour. Within individual males, the more tappings per sequence, the higher the fertilization success. In matings with an unlimited number of tapping sequences, fertilization success depended on female resistance. Within-male variation in fertilization success was higher in matings with an unlimited number of tapping sequences than in matings interrupted after a copulation. Several components of male mating behaviour thus contribute to the variation in sperm precedence in D. anilis. Female behaviour, in particular, resistance of tapping movements, can also increase variation in sperm ? 1997 The Association for the Study of Animal Behaviour precedence.
Variation in sperm precedence between males of the same species can result in large differences in male reproductive success. As with other sexually selected traits, this variation can be attributed to either male–male competition for fertilization or post-copulatory female choice (Smith 1984; Eberhard 1991; Birkhead & Møller 1993). Intraspecific variation in sperm precedence due to male–male competition can be phenotypic or context related. Phenotypic variation occurs when, for example, large males have higher fertilization success than small ones because they have larger ejaculates (Simmons & Parker 1992; Otronen 1994b) or males develop different mating strategies depending on their rearing conditions (Gage 1995). Context-related variation emphasizes the flexibility of the individual male’s mating strategy, for example, the ejaculate size or fertilization Correspondence: M. Otronen, Department of Zoology, Uppsala University, Villava¨gen 9, S-752 36 Uppsala, Sweden (email: [email protected]). 0003–3472/97/061233+08 $25.00/0/ar960425
success can vary according to the risk of sperm competition (Gage 1991; Gage & Baker 1991; Otronen 1994b; Cook & Gage 1995). Some studies have also shown variation in sperm precedence due to post-copulatory female choice, for example, females adopting different re-mating patterns depending on previous matings (Simmons 1986; Siva-Jothy & Hooper 1995) or storing sperm differently from different males (Ward 1993; Otronen et al., in press). Since several selection pressures are likely to affect sperm precedence patterns simultaneously it is important to examine how different factors interact and contribute to intraspecific variation. In the fly Dryomyza anilis there are several potential causes of individual variation in sperm precedence because of the very distinct postcopulatory mating behaviour (Otronen 1990). Most matings include several copulation and oviposition bouts. A copulation bout consists of an intromission followed by tapping sequences. After
? 1997 The Association for the Study of Animal Behaviour
1233
1234
Animal Behaviour, 53, 6
an intromission, the male withdraws his genitalia but remains mounted. During the following tapping sequences, the male taps the external genitalia of the female with his claspers. Last male fertilization success increases with an increasing number of tapping sequences. Last male sperm precedence is also increased during repeated copulation bouts which alternate with oviposition bouts (Otronen 1994a). As a result, variation in fertilization success between males can arise at different stages of mating. Here, I examine variation in fertilization success between and within males to find out how and at what stage of mating individual differences arise. I concentrate on the variation in sperm precedence during one copulation bout, involving one intromission followed by tapping sequences, that is, I exclude the effect of several copulation bouts. In a mating experiment, I allowed each male to mate with several non-virgin females. In the first treatment, the last mating was interrupted after an intromission, and the males could compete only through successful sperm transfer. In the second treatment, males were allowed five tapping sequences after an intromission and competed both with their ejaculate and the efficiency of their tapping movements. In the third treatment, males were allowed to continue mating until the female started resisting and the pair separated. Thus, the third treatment gave females a chance to influence male–male competition for fertilization by allowing different numbers of tapping sequences for different males. In D. anilis, females resist by bending their abdomen down which effectively prevents male tapping and often results in a clear conflict between the male and the female (Otronen 1989). Of course, more subtle female choice can take place at any stage of mating.
MATERIAL AND METHODS I used sterilized males to examine fertilization success for the last mating male. Experimental males were individually marked and sterilized with gamma radiation (about 300 rad/min). Because the competitive ability of sterile sperm is about the same as in normal males (Otronen 1990), all matings were run in the order normal males followed by one sterile male, thus giving the relative fertilization success for the last mating male.
To examine the variation in fertilization success within males. I allowed each male to mate with 8–11 females. The experiment included three treatments with an increase in the number of factors that might affect last male fertilization success. A different set of males was used in each treatment. In the first treatment with only a copulation, each pair was separated immediately after an intromission when the male withdrew his genitalia. In the second treatment, the mating was interrupted after each male had performed a copulation followed by five tapping sequences. In the third treatment with an unlimited number of tapping sequences, the male was allowed to continue the tapping sequences after a copulation until the female resisted, and the pair separated. If the male remained mounted in spite of strong female resistance I interrupted the mating after 10 min had elapsed from the last tapping sequence. I collected males from the wild and divided them into small, medium sized and large by eye. Males within each size class were then randomly assigned to the three treatments. I used 15 males in each of the first two treatments and 11 in the third. I used fewer males in the third treatment because the matings took much longer and I ran one treatment each day, that is, I performed the three treatments in 3-day cycles. Both normal and irradiated males were kept in cages (50#40#40 cm; 5–10 males per cage). Water, sugar and yeast were available ad libitum. The males were kept in a natural light:dark regime at 22–23)C. I collected females from the wild the day before I used them in the experiment. I used wild females because when collected on arrival at the oviposition site they are likely to contain mature eggs and be non-virgins. To make sure that females contained sperm, I allowed them to copulate with two normal males before the final experimental mating (see Otronen & Siva-Jothy 1991). Experimental matings took place in plastic tubes (5 cm diameter, 10 cm high). After the mating the female was immediately removed into another tube where she laid her eggs on a piece of fish. The eggs were then individually transferred to a piece of wet paper on a dish. Eggs fertilized with normal sperm hatched within 24 h. The eggs that did not hatch gave the fertilization success for the last mated male (the hatching failure among eggs fertilized with normal sperm is less than 1%, Otronen 1994a). The number of successful
Otronen: Variation in sperm precedence in a fly matings per male varied from six to nine because not all females laid eggs. The temperature during the experiments was 20–24)C. To remove the effect of temperature, I calculated residuals for all time variables. The decrease in fertilization success with an increasing number of eggs laid by the female (Otronen 1990) was taken into account by using residual fertilization success when comparing individual males. Repeatability (Falconer 1981) for different components of mating was calculated as shown by Lessells & Boag (1987).
1235
30
(a)
20
10
0
1
2
3
4
5
RESULTS Variation between Treatments I used repeated analysis of variance to test differences in fertilization success between the three treatments and between the consecutive matings by individual males (mating number). The three treatments were included as a between-subjects factor and egg number and male size as covariates. Both multivariate and univariate tests gave similar results and the data set fulfilled the assumptions for the use of univariate results: neither Mauchly’s test of sphericity on the effect of mating number (÷214 =18.681, ) nor Box’s M test for the equality of variance-covariance matrices of treatments was significant (÷242=47.150, ). Only univariate results are given below. Fertilization success in the three treatments was significantly different (arcsine transformed percentages; F2,35 =28.25, P<0.001). Neither of the covariates was significant (egg number t35 = "1.475, ; male wing length: t35 =1.638, ). Mating number was not significant (F5,184 =1.72, ) but there was a significant interaction between mating number and treatment (F10,184 =2.82, P<0.01). This was due to the first treatment where fertilization success decreased in consecutive matings (Fig. 1; the analysis included only the first six matings from each male because repeated measure requires the same number of observations from all subjects). The two other treatments did not show any systematic change in fertilization success in consecutive matings. Variation between Males Fertilization success between individual males varied greatly. The range of mean fertilization
Fertilization success (%)
50
6 (b)
40 30 20 10 0
1
2
3
4
5
60
6 (c)
50 40 30 20 10 0
1
2
4 3 Mating number
5
6
Figure 1. Mean (+) fertilization success in six consecutive matings in the three treatments. (a) Mating was interrupted after copulation; (b) matings with five tapping sequences; (c) matings with an unlimited number of tapping sequences.
success after intromission was 5–25%, after five tapping sequences 9–44%, and after an unlimited number of tapping sequences 22–57%. To compare differences between males, I calculated the mean and 95% confidence limits of the residual fertilization success for each male (Fig. 2). Residual fertilization success was used to remove the effect of egg number. In all three treatments, there were some males with no overlap in
Animal Behaviour, 53, 6
1236 0.4
(a)
0.3 0.2 0.1 0.0 –0.1 –0.2 –0.3
Residual fertilization success
–0.4
6
0.4
7
8
9
10
11
7
8
9
10
11
(b)
0.3 0.2 0.1 0.0 –0.1 –0.2 –0.3 –0.4
6
0.4
(c)
0.3 0.2 0.1
To examine the performance of different sized males, I tested the relationship between male size and mean residual fertilization success in the three treatments with ANCOVA. The three treatments did not differ from each other in residual fertilization success (F2,37 =0.024, ) and male size as a covariate was not significant (F1,37 =3.033, P<0.1). The relationship between male size and some behavioural components of mating was examined with a regression analysis. In matings with only a copulation, the residual copulation time decreased with increasing male size (t13 = "2.263, P<0.05). In matings with five tapping sequences, there was a positive relationship between male size and the number of tapping movements per tapping sequence (t12 =2.625, P<0.05), and a negative relationship between male size and mating time (t12 = "2.271, P<0.05; the whole regression: F2,12 =3.733, P<0.1). For matings with an unlimited number of tapping sequences, I tested male size against the total number of tapping sequences (t7 =1.274, ), the number of tapping sequences before the female resisted (t7 = "0.873, ) and time spent mating (t7 = "0.677, ), but none of the variables was significant (the whole regression: F3,7 =1.923, ).
0.0
Variation within Males
–0.1 –0.2 –0.3 –0.4
6
7
8 9 10 Male wing length (mm)
11
Figure 2. Within-male variation in fertilization success in the three treatments. The figure shows residual fertilization success (see text for explanation). Mean and 95% confidence limits for each male are given. Males are identified by their wing length. (a) Mating was interrupted after copulation; (b) matings with five tapping sequences; (c) matings with an unlimited number of tapping sequences.
confidence limits suggesting significant differences in fertilization success between individual males. However, 99% confidence limits, which provide better protection against experiment-wise error rate, indicated no significant differences between males.
Most components of mating behaviour showed large within-male variation and low repeatability (Table I). The only high repeatabilities were mating time in the treatment with five tapping sequences (Fig. 3) and the number of tapping movements per tapping sequence in the treatments with five or an unlimited number of tapping sequences (Fig. 4). Within-male variation of fertilization success was not related to male size. As the indicator of within-male variation in fertilization success I used the standard deviation of residual fertilization success. Within-male variation in fertilization success was different in the three treatments (ANCOVA: F2,37 =4.151, P<0.05), but male size was not significant (F1,37 =0.063, ). An a posteriori test showed that the treatment with an unlimited number of tapping sequences had a larger within-male variation than that including only a copulation (Tukey test: MD=0.056, P<0.05).
Otronen: Variation in sperm precedence in a fly
1237
Table I. Repeatability of different components of mating Variable Only copulation: Residual fertilization success Residual copulation time Five tapping sequences: Residual fertilization success Residual copulation time No. of tapping movements Residual total time Unlimited number of tapping sequences: Residual fertilization success Residual copulation time No. of tapping movements No. of tapping sequences Residual time per tapping sequence No. of tapping sequences before female resists
F-ratio
P
R
Range
1.220 (14,96) 2.121 (14,96)
<0.02
0.03 0.13
5.3–24.7 42–45
2.625 2.429 22.145 9.405
(14,108) (14,108) (14,108) (14,108)
<0.01 <0.02 <0.001 <0.001
0.17 0.14 0.72 0.51
8.9–43.9 41–61 17–43 497–986
2.548 3.644 9.079 2.838 2.066 2.037
(10,74) (10,74) (10,74) (10,74) (10,74) (10,74)
<0.02 0.001 0.001 <0.01 <0.05 <0.05
0.15 0.26 0.51 0.19 0.12 0.12
22.0–56.9 39–64 13–28 7–18 122–207 7–15
F-ratio (df ) and P refer to one-way ANOVA for differences between males used to calculate repeatability (R). Range gives the highest and lowest mean value for each untransformed variable. Fertilization success is given as a percentage and time variable in s.
Mating time (min)
18 16 14 12 10 8 6
7
8 9 10 Male wing length (mm)
11
Figure 3. Within-male variation in mating time (X&) against male size in matings with five tapping sequences.
To examine how the different components of mating behaviour interact with fertilization success within males, I used multivariate profile analysis. I first ranked experimental matings within males according to fertilization success in the six consecutive trials in the analysis. If a component of mating behaviour interacts with fertilization success, a significant change in that component is expected when analysed in that same order. I tested only for linear trends and when calculating them I used uneven spacing between trials which I determined on the basis of the change in the average fertilization success in the six ordered matings.
I analysed residual copulation time in relation to fertilization success, with the treatment as a factor. There was no significant linear trend in copulation time with fertilization success (F1,38 =1.452, ), nor any significant relationship between copulation time and treatment (F2,38 =0.486, ). Number of tapping movements per tapping sequence in the second and third treatments showed a significant linear trend with fertilization success (F1,24 =7.596, P<0.05), without any significant interaction between the number of tapping movements and treatment (F1,24 =0.501, ), suggesting that in matings resulting in a high fertilization success males performed more tapping movements per tapping sequence than in less successful ones (Fig. 5). Residual mating time in the treatment with five tapping sequences did not show any significant linear trend with fertilization success (F1,14 =1.010, ). In matings with an unlimited number of tapping sequences I tested both the total number of tapping sequences the male performed and the number of tapping sequences before the female resisted. Only the second variable changed significantly in relation to fertilization success (number of tapping sequences: F1,10 =0.892, ; number of tapping sequences before the female resisted: F1,10 =5.069, P<0.05; Fig. 6). Residual time spent per tapping sequence did not have any significant trend with fertilization success (F1,10 =2.130, ).
Animal Behaviour, 53, 6
1238 50
29
(a)
(a)
27 25 23
30
21
Number of tappings per sequence 13 .4 ± 2. 3 23 .0 ± 1. 9 29 .4 ± 2. 1 36 .6 ± 2. 0 47 .6 ± 2. 0 61 .6 ± 3. 9
Number of tappings per sequence
40
19
20
17
10
7
6
30
8
9
10
11
(b)
15
29
(b)
27 25
20
23 21 19
DISCUSSION My results show that there were large differences in the mean fertilization success between males after similar type of matings. As pointed out by Lewis & Austad (1990), differences between males in sperm precedence show that the often large variation in sperm precedence within species is not random but indicates actual differences between males. Variation in sperm precedence between species can be traced back to the underlying sperm precedence mechanism (Parker et al. 1990). The sperm precedence mechanism inherent to the species is also likely to determine how intraspecific variation arises. For example, in some species large males are more successful than small ones because they have a higher displacement rate or larger ejaculates than small males (e.g. Simmons & Parker 1992; Parker & Simmons 1994; Eady 1994). In D. anilis, large males had shorter copulations than small ones but copulation time did not affect
4 ±
5.
6 .8 70
±
4.
9 .6 53
±
2.
8 .9 44
±
2.
8 35
.7
±
2.
4
15 .6
Figure 4. Within-male variation in the number of tapping movements per tapping sequence (X&) against male size in matings with (a) five tapping sequences and (b) an unlimited number of tapping sequences.
17
3.
11
±
8 9 10 Male wing length (mm)
28
7
.3
6
17
10
Average fertilization success Figure 5. The number of tapping movements per tapping sequence (X&) in six matings, ordered, within each male, according to increasing fertilization success. The X-axis shows the average fertilization success (&) in these matings. (a) Matings with five tapping sequences and (b) with an unlimited number of tapping sequences.
fertilization success suggesting that variation in copulation time could be related to differences in sperm transfer. In the yellow dung fly, Scathophaga stercoraria, larger males have higher displacement rates than small ones and therefore need a shorter copulation time to attain the same fertilization success (Parker & Simmons 1994). In D. anilis, males may also compete with ejaculate size (see Otronen, in press) although the number of tapping sequences after a copulation is very important for male fertilization success (Otronen 1990). Large and small males also differed in how long it took males to complete a certain number of tapping sequences but there was no clear relationship between male size and fertilization success. As a result, males of different size might not differ
15
10
5
5. 4 ± .8
.6
±
4. 6 70
2. 9 53
.9
±
2. 8 44
± .7
±
±
35
.6 28
.3 17
2. 8
0 3. 4
Number of tapping sequences
Otronen: Variation in sperm precedence in a fly
Average fertilization success Figure 6. The number of tapping sequences in six matings with an unlimited number of tapping sequences ordered, within each male, according to increasing fertilization success. The total number of tapping sequences before the pair separated (.) and the number of tapping sequences before the female resisted (/) + are shown.
much in fertilization success after a certain number of tapping sequences but small males might have spent twice the time achieving it. Mating time could also affect male reproductive success via the number of females males manage to mate during one visit to an oviposition site. In the treatment with only a copulation, there was a decrease in fertilization success in consecutive matings. Although irradiation may affect sperm quality, it cannot explain the result in this study because the other two treatments did not show any decrease. It may have been that the experimental procedure, where males were forced to separate from the female, affected their future performance during mating. The consistency of fertilization success, in particular, in the second treatment where the number of tapping sequences was controlled, suggests that sperm in irradiated males retained their quality for some weeks after irradiation. Repeatability (Falconer 1981) can be used as an indicator of consistency in individuals’ characteristics (Boake 1989) and in this study I used repeatability to examine stereotypy of male mating behaviour. Repeatability is high if there is little variation within individuals and much variation between them (Boake 1989). In this study, repeatability was high in only two components of mating behaviour: males differed in the number of tapping movements per tapping sequence and
1239
in mating time in matings with five tapping sequences. These variables were also related to male size. Previous experiments on D. anilis, which have concentrated on differences between males, have shown that the number of tapping movements per tapping sequence and time spent mating are not directly related to fertilization success (Otronen 1990). The within-male comparison in this study suggests, however, that the number of tapping movements per tapping sequence is less important per se but within the range of tapping specific for each male, a high number of tapping movements results in a high fertilization success. There may be several reasons for the low repeatability of the other components of mating behaviour (see Boake 1989). For example, there may not have been any large differences between males in the characteristics examined. An alternative explanation for a low repeatability is a large environmental effect which increases variance within individuals. The most important ‘environmental’ factor in this study may have been the mating behaviour and preferences of the female. Female choice can take place at various stages of mating (Birkhead & Møller 1993). In D. anilis, female choice before mating may be restricted because of the male-biased sex ratio and strong male–male competition at oviposition sites (Otronen 1993). However, during mating females could control sperm precedence by allowing a different number of tapping sequences depending on how attractive the male is. Females could influence sperm precedence in the third treatment where males performed tapping sequences as long as female resistance did not stop them. As pointed out earlier, this does not mean that females would not have more subtle post-copulatory female choice before resistance. In the third treatment, the repeatability of tapping sequences was low and within-male variation in fertilization success increased when compared with the first treatment. This within-male variation may have partly resulted from female preference for different males (see Lewis & Austad 1994) as suggested by the positive relationship between male fertilization success and the number of tapping sequences before females resisted. Thus, in D. anilis, females could control male fertilization success by resisting tapping sequences. Although males may continue tapping sequences after female resistance,
Animal Behaviour, 53, 6
1240
their fertilization success may not increase as effectively as before female resistance. In conclusion, male fertilization success in D. anilis is affected by several factors during matings. Factors related to male–male competition for fertilization, such as mating time, the number of tapping sequences and the efficiency of tapping movements, will contribute to the observed differences between and within individual males. Female choice, in particular, female resistance during tapping movements, can also increase variation in sperm precedence.
ACKNOWLEDGMENTS I thank Peder Fiske for commenting on the manuscript and Pekka Rintama¨ki for reading it. Lammi Biological Station provided good facilities during the experiment and Timo Autio at the Department of Radiochemistry, University of Helsinki irradiated the flies. Financial support was received from The Swedish Natural Science Research Council.
REFERENCES Birkhead, T. & Møller, A. P. 1993. Female choice of paternity. Trends Ecol. Evol., 8, 100–104. Boake, C. R. B. 1989. Repeatability: its role in evolutionary studies of mating behaviour. Evol. Ecol., 3, 173–182. Cook, A. & Gage, J. G. 1995. Effects of risk 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. Eady, P. 1994. Intraspecific variation in sperm precedence in the bruchid beetle Callosobruchus maculatus. Ecol. Entomol., 19, 11–16. Eberhard, W. G. 1991. Copulatory courtship and cryptic female choice in insects. Biol. Rev., 66, 1–31. Falconer, D. S. 1981. Introduction to Quantitative Genetics. London: Longman. Gage, M. G. J. 1991. Risk of sperm competition directly affects ejaculate size in the Mediterranean fruit fly. Anim. Behav., 42, 1036–1037. 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. Ser. B, 261, 25–30. Gage, M. J. G. & Baker, R. R. 1991. Ejaculate size varies with socio-sexual situation in an insect. Ecol. Entomol., 16, 331–337.
Lessells, C. M. & Boag, P. T. 1987. Unrepeatable repeatabilities: a common mistake. Auk, 104, 116–121. Lewis, S. M. & Austad, S. N. 1990. Sources of intraspecific variation in sperm precedence in red flour beetles. Am. Nat., 135, 351–359. Lewis, S. M. & Austad, S. N. 1994. Sexual selection in flour beetles: the relationship between sperm precedence and male olfactory attractiveness. Behav. Ecol., 5, 219–224. Otronen, M. 1989. Female mating behaviour and multiple matings in the fly Dryomyza anilis. Behaviour, 111, 77–79. Otronen, M. 1990. Mating behaviour and sperm competition in the fly Dryomyza anilis. Behav. Ecol. Sociobiol., 26, 349–356. Otronen, M. 1993. Male distribution and interactions at female oviposition sites as factors affecting success in the fly Dryomyza anilis (Dryomyzidae). Evol. Ecol., 7, 127–141. Otronen, M. 1994a. Repeated copulations as a strategy to maximize fertilization in the fly Dryomyza anilis (Dryomyzidae). Behav. Ecol., 5, 51–56. Otronen, M. 1994b. Fertilization success in the fly Dryomyza anilis (Dryomyzidae): effects of male size and the mating situation. Behav. Ecol. Sociobiol., 35, 33–38. Otronen, M. In press. Sperm numbers, their storage and usage in the fly Dryomyza anilis. Proc. R. Soc. Lond. Ser. B. Otronen, M. & Siva-Jothy, M. T. 1991. The effect of postcopulatory male behaviour on ejaculate distribution within the female sperm storage organs in the fly Dryomyza anilis. Behav. Ecol. Sociobiol., 29, 33–37. Otronen, M., Requera, P. & Ward P. In press. Female choice of sperm in the yellow dung fly: sperm length and sperm distribution within female sperm storage organs. Ethology. Parker, G. A. & Simmons, L. W. 1994. Evolution of phenotypic optima and copula duration in dungflies. Nature, Lond., 370, 53–56. Parker, G. A., Simmons, L. W. & Kirk, H. 1990. Analysing sperm competition data: simple models for predicting mechanisms. Behav. Ecol. Sociobiol., 27, 55–65. Simmons, L. W. 1986. Female choice in the field cricket Gryllus bimaculatus (De Geer). Anim. Behav., 34, 1463–1470. Simmons, L. W. & Parker, G. A. 1992. Individual variation in sperm competition success of yellow dung flies, Scatophaga stercoraria. Evolution, 46, 366–375. Siva-Jothy, M. T. & Hooper, R. E. 1995. The disposition and genetic diversity of stored sperm in females of the damselfly Calopteryx splendens xanthostoma (Charpentier). Proc. R. Soc. Lond. Ser. B, 259, 313–318. Smith, R. L. (ed.) 1984. Sperm Competition and the Evolution of Animal Mating Systems. New York: Academic Press. Ward, P. I. 1993. Females influence sperm storage and use in the yellow dung fly Scathophaga stercoraria (L.). Behav. Ecol. Sociobiol., 32, 313–319.
Variation in sperm precedence during mating in male flies, Dryomyza anilis MERJA OTRONEN Department of Zoology, Uppsala University (Received 14 March 1996; initial acceptance 23 May 1996; final acceptance 30 August 1996; MS. number: 5191)
Abstract. To understand how individual differences in fertilization success arises in the fly Dryomyza anilis, variation in sperm precedence between and within males was studied. In D. anilis, a mating consists of a copulation followed by tapping sequences during which the male taps the external genitalia of the female. These tapping sequences increase last male sperm precedence. Each male in the experiment was repeatedly mated with a set of females. In the first treatment, mating was interrupted after the male had an intromission. In the second treatment, the male was allowed to perform five tapping sequences after an intromission. In the third treatment, the male was allowed to perform an unlimited number of tapping sequences after an intromission until interrupted by female resistance. There were large differences between males in all three treatments as shown by the 95% confidence limits of the mean fertilization success. Repeatabilities of mating time and the number of tapping movements per tapping sequence were high indicating clear differences between males in these components of mating behaviour. Within individual males, the more tappings per sequence, the higher the fertilization success. In matings with an unlimited number of tapping sequences, fertilization success depended on female resistance. Within-male variation in fertilization success was higher in matings with an unlimited number of tapping sequences than in matings interrupted after a copulation. Several components of male mating behaviour thus contribute to the variation in sperm precedence in D. anilis. Female behaviour, in particular, resistance of tapping movements, can also increase variation in sperm ? 1997 The Association for the Study of Animal Behaviour precedence.
Variation in sperm precedence between males of the same species can result in large differences in male reproductive success. As with other sexually selected traits, this variation can be attributed to either male–male competition for fertilization or post-copulatory female choice (Smith 1984; Eberhard 1991; Birkhead & Møller 1993). Intraspecific variation in sperm precedence due to male–male competition can be phenotypic or context related. Phenotypic variation occurs when, for example, large males have higher fertilization success than small ones because they have larger ejaculates (Simmons & Parker 1992; Otronen 1994b) or males develop different mating strategies depending on their rearing conditions (Gage 1995). Context-related variation emphasizes the flexibility of the individual male’s mating strategy, for example, the ejaculate size or fertilization Correspondence: M. Otronen, Department of Zoology, Uppsala University, Villava¨gen 9, S-752 36 Uppsala, Sweden (email: [email protected]). 0003–3472/97/061233+08 $25.00/0/ar960425
success can vary according to the risk of sperm competition (Gage 1991; Gage & Baker 1991; Otronen 1994b; Cook & Gage 1995). Some studies have also shown variation in sperm precedence due to post-copulatory female choice, for example, females adopting different re-mating patterns depending on previous matings (Simmons 1986; Siva-Jothy & Hooper 1995) or storing sperm differently from different males (Ward 1993; Otronen et al., in press). Since several selection pressures are likely to affect sperm precedence patterns simultaneously it is important to examine how different factors interact and contribute to intraspecific variation. In the fly Dryomyza anilis there are several potential causes of individual variation in sperm precedence because of the very distinct postcopulatory mating behaviour (Otronen 1990). Most matings include several copulation and oviposition bouts. A copulation bout consists of an intromission followed by tapping sequences. After
? 1997 The Association for the Study of Animal Behaviour
1233
1234
Animal Behaviour, 53, 6
an intromission, the male withdraws his genitalia but remains mounted. During the following tapping sequences, the male taps the external genitalia of the female with his claspers. Last male fertilization success increases with an increasing number of tapping sequences. Last male sperm precedence is also increased during repeated copulation bouts which alternate with oviposition bouts (Otronen 1994a). As a result, variation in fertilization success between males can arise at different stages of mating. Here, I examine variation in fertilization success between and within males to find out how and at what stage of mating individual differences arise. I concentrate on the variation in sperm precedence during one copulation bout, involving one intromission followed by tapping sequences, that is, I exclude the effect of several copulation bouts. In a mating experiment, I allowed each male to mate with several non-virgin females. In the first treatment, the last mating was interrupted after an intromission, and the males could compete only through successful sperm transfer. In the second treatment, males were allowed five tapping sequences after an intromission and competed both with their ejaculate and the efficiency of their tapping movements. In the third treatment, males were allowed to continue mating until the female started resisting and the pair separated. Thus, the third treatment gave females a chance to influence male–male competition for fertilization by allowing different numbers of tapping sequences for different males. In D. anilis, females resist by bending their abdomen down which effectively prevents male tapping and often results in a clear conflict between the male and the female (Otronen 1989). Of course, more subtle female choice can take place at any stage of mating.
MATERIAL AND METHODS I used sterilized males to examine fertilization success for the last mating male. Experimental males were individually marked and sterilized with gamma radiation (about 300 rad/min). Because the competitive ability of sterile sperm is about the same as in normal males (Otronen 1990), all matings were run in the order normal males followed by one sterile male, thus giving the relative fertilization success for the last mating male.
To examine the variation in fertilization success within males. I allowed each male to mate with 8–11 females. The experiment included three treatments with an increase in the number of factors that might affect last male fertilization success. A different set of males was used in each treatment. In the first treatment with only a copulation, each pair was separated immediately after an intromission when the male withdrew his genitalia. In the second treatment, the mating was interrupted after each male had performed a copulation followed by five tapping sequences. In the third treatment with an unlimited number of tapping sequences, the male was allowed to continue the tapping sequences after a copulation until the female resisted, and the pair separated. If the male remained mounted in spite of strong female resistance I interrupted the mating after 10 min had elapsed from the last tapping sequence. I collected males from the wild and divided them into small, medium sized and large by eye. Males within each size class were then randomly assigned to the three treatments. I used 15 males in each of the first two treatments and 11 in the third. I used fewer males in the third treatment because the matings took much longer and I ran one treatment each day, that is, I performed the three treatments in 3-day cycles. Both normal and irradiated males were kept in cages (50#40#40 cm; 5–10 males per cage). Water, sugar and yeast were available ad libitum. The males were kept in a natural light:dark regime at 22–23)C. I collected females from the wild the day before I used them in the experiment. I used wild females because when collected on arrival at the oviposition site they are likely to contain mature eggs and be non-virgins. To make sure that females contained sperm, I allowed them to copulate with two normal males before the final experimental mating (see Otronen & Siva-Jothy 1991). Experimental matings took place in plastic tubes (5 cm diameter, 10 cm high). After the mating the female was immediately removed into another tube where she laid her eggs on a piece of fish. The eggs were then individually transferred to a piece of wet paper on a dish. Eggs fertilized with normal sperm hatched within 24 h. The eggs that did not hatch gave the fertilization success for the last mated male (the hatching failure among eggs fertilized with normal sperm is less than 1%, Otronen 1994a). The number of successful
Otronen: Variation in sperm precedence in a fly matings per male varied from six to nine because not all females laid eggs. The temperature during the experiments was 20–24)C. To remove the effect of temperature, I calculated residuals for all time variables. The decrease in fertilization success with an increasing number of eggs laid by the female (Otronen 1990) was taken into account by using residual fertilization success when comparing individual males. Repeatability (Falconer 1981) for different components of mating was calculated as shown by Lessells & Boag (1987).
1235
30
(a)
20
10
0
1
2
3
4
5
RESULTS Variation between Treatments I used repeated analysis of variance to test differences in fertilization success between the three treatments and between the consecutive matings by individual males (mating number). The three treatments were included as a between-subjects factor and egg number and male size as covariates. Both multivariate and univariate tests gave similar results and the data set fulfilled the assumptions for the use of univariate results: neither Mauchly’s test of sphericity on the effect of mating number (÷214 =18.681, ) nor Box’s M test for the equality of variance-covariance matrices of treatments was significant (÷242=47.150, ). Only univariate results are given below. Fertilization success in the three treatments was significantly different (arcsine transformed percentages; F2,35 =28.25, P<0.001). Neither of the covariates was significant (egg number t35 = "1.475, ; male wing length: t35 =1.638, ). Mating number was not significant (F5,184 =1.72, ) but there was a significant interaction between mating number and treatment (F10,184 =2.82, P<0.01). This was due to the first treatment where fertilization success decreased in consecutive matings (Fig. 1; the analysis included only the first six matings from each male because repeated measure requires the same number of observations from all subjects). The two other treatments did not show any systematic change in fertilization success in consecutive matings. Variation between Males Fertilization success between individual males varied greatly. The range of mean fertilization
Fertilization success (%)
50
6 (b)
40 30 20 10 0
1
2
3
4
5
60
6 (c)
50 40 30 20 10 0
1
2
4 3 Mating number
5
6
Figure 1. Mean (+) fertilization success in six consecutive matings in the three treatments. (a) Mating was interrupted after copulation; (b) matings with five tapping sequences; (c) matings with an unlimited number of tapping sequences.
success after intromission was 5–25%, after five tapping sequences 9–44%, and after an unlimited number of tapping sequences 22–57%. To compare differences between males, I calculated the mean and 95% confidence limits of the residual fertilization success for each male (Fig. 2). Residual fertilization success was used to remove the effect of egg number. In all three treatments, there were some males with no overlap in
Animal Behaviour, 53, 6
1236 0.4
(a)
0.3 0.2 0.1 0.0 –0.1 –0.2 –0.3
Residual fertilization success
–0.4
6
0.4
7
8
9
10
11
7
8
9
10
11
(b)
0.3 0.2 0.1 0.0 –0.1 –0.2 –0.3 –0.4
6
0.4
(c)
0.3 0.2 0.1
To examine the performance of different sized males, I tested the relationship between male size and mean residual fertilization success in the three treatments with ANCOVA. The three treatments did not differ from each other in residual fertilization success (F2,37 =0.024, ) and male size as a covariate was not significant (F1,37 =3.033, P<0.1). The relationship between male size and some behavioural components of mating was examined with a regression analysis. In matings with only a copulation, the residual copulation time decreased with increasing male size (t13 = "2.263, P<0.05). In matings with five tapping sequences, there was a positive relationship between male size and the number of tapping movements per tapping sequence (t12 =2.625, P<0.05), and a negative relationship between male size and mating time (t12 = "2.271, P<0.05; the whole regression: F2,12 =3.733, P<0.1). For matings with an unlimited number of tapping sequences, I tested male size against the total number of tapping sequences (t7 =1.274, ), the number of tapping sequences before the female resisted (t7 = "0.873, ) and time spent mating (t7 = "0.677, ), but none of the variables was significant (the whole regression: F3,7 =1.923, ).
0.0
Variation within Males
–0.1 –0.2 –0.3 –0.4
6
7
8 9 10 Male wing length (mm)
11
Figure 2. Within-male variation in fertilization success in the three treatments. The figure shows residual fertilization success (see text for explanation). Mean and 95% confidence limits for each male are given. Males are identified by their wing length. (a) Mating was interrupted after copulation; (b) matings with five tapping sequences; (c) matings with an unlimited number of tapping sequences.
confidence limits suggesting significant differences in fertilization success between individual males. However, 99% confidence limits, which provide better protection against experiment-wise error rate, indicated no significant differences between males.
Most components of mating behaviour showed large within-male variation and low repeatability (Table I). The only high repeatabilities were mating time in the treatment with five tapping sequences (Fig. 3) and the number of tapping movements per tapping sequence in the treatments with five or an unlimited number of tapping sequences (Fig. 4). Within-male variation of fertilization success was not related to male size. As the indicator of within-male variation in fertilization success I used the standard deviation of residual fertilization success. Within-male variation in fertilization success was different in the three treatments (ANCOVA: F2,37 =4.151, P<0.05), but male size was not significant (F1,37 =0.063, ). An a posteriori test showed that the treatment with an unlimited number of tapping sequences had a larger within-male variation than that including only a copulation (Tukey test: MD=0.056, P<0.05).
Otronen: Variation in sperm precedence in a fly
1237
Table I. Repeatability of different components of mating Variable Only copulation: Residual fertilization success Residual copulation time Five tapping sequences: Residual fertilization success Residual copulation time No. of tapping movements Residual total time Unlimited number of tapping sequences: Residual fertilization success Residual copulation time No. of tapping movements No. of tapping sequences Residual time per tapping sequence No. of tapping sequences before female resists
F-ratio
P
R
Range
1.220 (14,96) 2.121 (14,96)
<0.02
0.03 0.13
5.3–24.7 42–45
2.625 2.429 22.145 9.405
(14,108) (14,108) (14,108) (14,108)
<0.01 <0.02 <0.001 <0.001
0.17 0.14 0.72 0.51
8.9–43.9 41–61 17–43 497–986
2.548 3.644 9.079 2.838 2.066 2.037
(10,74) (10,74) (10,74) (10,74) (10,74) (10,74)
<0.02 0.001 0.001 <0.01 <0.05 <0.05
0.15 0.26 0.51 0.19 0.12 0.12
22.0–56.9 39–64 13–28 7–18 122–207 7–15
F-ratio (df ) and P refer to one-way ANOVA for differences between males used to calculate repeatability (R). Range gives the highest and lowest mean value for each untransformed variable. Fertilization success is given as a percentage and time variable in s.
Mating time (min)
18 16 14 12 10 8 6
7
8 9 10 Male wing length (mm)
11
Figure 3. Within-male variation in mating time (X&) against male size in matings with five tapping sequences.
To examine how the different components of mating behaviour interact with fertilization success within males, I used multivariate profile analysis. I first ranked experimental matings within males according to fertilization success in the six consecutive trials in the analysis. If a component of mating behaviour interacts with fertilization success, a significant change in that component is expected when analysed in that same order. I tested only for linear trends and when calculating them I used uneven spacing between trials which I determined on the basis of the change in the average fertilization success in the six ordered matings.
I analysed residual copulation time in relation to fertilization success, with the treatment as a factor. There was no significant linear trend in copulation time with fertilization success (F1,38 =1.452, ), nor any significant relationship between copulation time and treatment (F2,38 =0.486, ). Number of tapping movements per tapping sequence in the second and third treatments showed a significant linear trend with fertilization success (F1,24 =7.596, P<0.05), without any significant interaction between the number of tapping movements and treatment (F1,24 =0.501, ), suggesting that in matings resulting in a high fertilization success males performed more tapping movements per tapping sequence than in less successful ones (Fig. 5). Residual mating time in the treatment with five tapping sequences did not show any significant linear trend with fertilization success (F1,14 =1.010, ). In matings with an unlimited number of tapping sequences I tested both the total number of tapping sequences the male performed and the number of tapping sequences before the female resisted. Only the second variable changed significantly in relation to fertilization success (number of tapping sequences: F1,10 =0.892, ; number of tapping sequences before the female resisted: F1,10 =5.069, P<0.05; Fig. 6). Residual time spent per tapping sequence did not have any significant trend with fertilization success (F1,10 =2.130, ).
Animal Behaviour, 53, 6
1238 50
29
(a)
(a)
27 25 23
30
21
Number of tappings per sequence 13 .4 ± 2. 3 23 .0 ± 1. 9 29 .4 ± 2. 1 36 .6 ± 2. 0 47 .6 ± 2. 0 61 .6 ± 3. 9
Number of tappings per sequence
40
19
20
17
10
7
6
30
8
9
10
11
(b)
15
29
(b)
27 25
20
23 21 19
DISCUSSION My results show that there were large differences in the mean fertilization success between males after similar type of matings. As pointed out by Lewis & Austad (1990), differences between males in sperm precedence show that the often large variation in sperm precedence within species is not random but indicates actual differences between males. Variation in sperm precedence between species can be traced back to the underlying sperm precedence mechanism (Parker et al. 1990). The sperm precedence mechanism inherent to the species is also likely to determine how intraspecific variation arises. For example, in some species large males are more successful than small ones because they have a higher displacement rate or larger ejaculates than small males (e.g. Simmons & Parker 1992; Parker & Simmons 1994; Eady 1994). In D. anilis, large males had shorter copulations than small ones but copulation time did not affect
4 ±
5.
6 .8 70
±
4.
9 .6 53
±
2.
8 .9 44
±
2.
8 35
.7
±
2.
4
15 .6
Figure 4. Within-male variation in the number of tapping movements per tapping sequence (X&) against male size in matings with (a) five tapping sequences and (b) an unlimited number of tapping sequences.
17
3.
11
±
8 9 10 Male wing length (mm)
28
7
.3
6
17
10
Average fertilization success Figure 5. The number of tapping movements per tapping sequence (X&) in six matings, ordered, within each male, according to increasing fertilization success. The X-axis shows the average fertilization success (&) in these matings. (a) Matings with five tapping sequences and (b) with an unlimited number of tapping sequences.
fertilization success suggesting that variation in copulation time could be related to differences in sperm transfer. In the yellow dung fly, Scathophaga stercoraria, larger males have higher displacement rates than small ones and therefore need a shorter copulation time to attain the same fertilization success (Parker & Simmons 1994). In D. anilis, males may also compete with ejaculate size (see Otronen, in press) although the number of tapping sequences after a copulation is very important for male fertilization success (Otronen 1990). Large and small males also differed in how long it took males to complete a certain number of tapping sequences but there was no clear relationship between male size and fertilization success. As a result, males of different size might not differ
15
10
5
5. 4 ± .8
.6
±
4. 6 70
2. 9 53
.9
±
2. 8 44
± .7
±
±
35
.6 28
.3 17
2. 8
0 3. 4
Number of tapping sequences
Otronen: Variation in sperm precedence in a fly
Average fertilization success Figure 6. The number of tapping sequences in six matings with an unlimited number of tapping sequences ordered, within each male, according to increasing fertilization success. The total number of tapping sequences before the pair separated (.) and the number of tapping sequences before the female resisted (/) + are shown.
much in fertilization success after a certain number of tapping sequences but small males might have spent twice the time achieving it. Mating time could also affect male reproductive success via the number of females males manage to mate during one visit to an oviposition site. In the treatment with only a copulation, there was a decrease in fertilization success in consecutive matings. Although irradiation may affect sperm quality, it cannot explain the result in this study because the other two treatments did not show any decrease. It may have been that the experimental procedure, where males were forced to separate from the female, affected their future performance during mating. The consistency of fertilization success, in particular, in the second treatment where the number of tapping sequences was controlled, suggests that sperm in irradiated males retained their quality for some weeks after irradiation. Repeatability (Falconer 1981) can be used as an indicator of consistency in individuals’ characteristics (Boake 1989) and in this study I used repeatability to examine stereotypy of male mating behaviour. Repeatability is high if there is little variation within individuals and much variation between them (Boake 1989). In this study, repeatability was high in only two components of mating behaviour: males differed in the number of tapping movements per tapping sequence and
1239
in mating time in matings with five tapping sequences. These variables were also related to male size. Previous experiments on D. anilis, which have concentrated on differences between males, have shown that the number of tapping movements per tapping sequence and time spent mating are not directly related to fertilization success (Otronen 1990). The within-male comparison in this study suggests, however, that the number of tapping movements per tapping sequence is less important per se but within the range of tapping specific for each male, a high number of tapping movements results in a high fertilization success. There may be several reasons for the low repeatability of the other components of mating behaviour (see Boake 1989). For example, there may not have been any large differences between males in the characteristics examined. An alternative explanation for a low repeatability is a large environmental effect which increases variance within individuals. The most important ‘environmental’ factor in this study may have been the mating behaviour and preferences of the female. Female choice can take place at various stages of mating (Birkhead & Møller 1993). In D. anilis, female choice before mating may be restricted because of the male-biased sex ratio and strong male–male competition at oviposition sites (Otronen 1993). However, during mating females could control sperm precedence by allowing a different number of tapping sequences depending on how attractive the male is. Females could influence sperm precedence in the third treatment where males performed tapping sequences as long as female resistance did not stop them. As pointed out earlier, this does not mean that females would not have more subtle post-copulatory female choice before resistance. In the third treatment, the repeatability of tapping sequences was low and within-male variation in fertilization success increased when compared with the first treatment. This within-male variation may have partly resulted from female preference for different males (see Lewis & Austad 1994) as suggested by the positive relationship between male fertilization success and the number of tapping sequences before females resisted. Thus, in D. anilis, females could control male fertilization success by resisting tapping sequences. Although males may continue tapping sequences after female resistance,
Animal Behaviour, 53, 6
1240
their fertilization success may not increase as effectively as before female resistance. In conclusion, male fertilization success in D. anilis is affected by several factors during matings. Factors related to male–male competition for fertilization, such as mating time, the number of tapping sequences and the efficiency of tapping movements, will contribute to the observed differences between and within individual males. Female choice, in particular, female resistance during tapping movements, can also increase variation in sperm precedence.
ACKNOWLEDGMENTS I thank Peder Fiske for commenting on the manuscript and Pekka Rintama¨ki for reading it. Lammi Biological Station provided good facilities during the experiment and Timo Autio at the Department of Radiochemistry, University of Helsinki irradiated the flies. Financial support was received from The Swedish Natural Science Research Council.
REFERENCES Birkhead, T. & Møller, A. P. 1993. Female choice of paternity. Trends Ecol. Evol., 8, 100–104. Boake, C. R. B. 1989. Repeatability: its role in evolutionary studies of mating behaviour. Evol. Ecol., 3, 173–182. Cook, A. & Gage, J. G. 1995. Effects of risk 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. Eady, P. 1994. Intraspecific variation in sperm precedence in the bruchid beetle Callosobruchus maculatus. Ecol. Entomol., 19, 11–16. Eberhard, W. G. 1991. Copulatory courtship and cryptic female choice in insects. Biol. Rev., 66, 1–31. Falconer, D. S. 1981. Introduction to Quantitative Genetics. London: Longman. Gage, M. G. J. 1991. Risk of sperm competition directly affects ejaculate size in the Mediterranean fruit fly. Anim. Behav., 42, 1036–1037. 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. Ser. B, 261, 25–30. Gage, M. J. G. & Baker, R. R. 1991. Ejaculate size varies with socio-sexual situation in an insect. Ecol. Entomol., 16, 331–337.
Lessells, C. M. & Boag, P. T. 1987. Unrepeatable repeatabilities: a common mistake. Auk, 104, 116–121. Lewis, S. M. & Austad, S. N. 1990. Sources of intraspecific variation in sperm precedence in red flour beetles. Am. Nat., 135, 351–359. Lewis, S. M. & Austad, S. N. 1994. Sexual selection in flour beetles: the relationship between sperm precedence and male olfactory attractiveness. Behav. Ecol., 5, 219–224. Otronen, M. 1989. Female mating behaviour and multiple matings in the fly Dryomyza anilis. Behaviour, 111, 77–79. Otronen, M. 1990. Mating behaviour and sperm competition in the fly Dryomyza anilis. Behav. Ecol. Sociobiol., 26, 349–356. Otronen, M. 1993. Male distribution and interactions at female oviposition sites as factors affecting success in the fly Dryomyza anilis (Dryomyzidae). Evol. Ecol., 7, 127–141. Otronen, M. 1994a. Repeated copulations as a strategy to maximize fertilization in the fly Dryomyza anilis (Dryomyzidae). Behav. Ecol., 5, 51–56. Otronen, M. 1994b. Fertilization success in the fly Dryomyza anilis (Dryomyzidae): effects of male size and the mating situation. Behav. Ecol. Sociobiol., 35, 33–38. Otronen, M. In press. Sperm numbers, their storage and usage in the fly Dryomyza anilis. Proc. R. Soc. Lond. Ser. B. Otronen, M. & Siva-Jothy, M. T. 1991. The effect of postcopulatory male behaviour on ejaculate distribution within the female sperm storage organs in the fly Dryomyza anilis. Behav. Ecol. Sociobiol., 29, 33–37. Otronen, M., Requera, P. & Ward P. In press. Female choice of sperm in the yellow dung fly: sperm length and sperm distribution within female sperm storage organs. Ethology. Parker, G. A. & Simmons, L. W. 1994. Evolution of phenotypic optima and copula duration in dungflies. Nature, Lond., 370, 53–56. Parker, G. A., Simmons, L. W. & Kirk, H. 1990. Analysing sperm competition data: simple models for predicting mechanisms. Behav. Ecol. Sociobiol., 27, 55–65. Simmons, L. W. 1986. Female choice in the field cricket Gryllus bimaculatus (De Geer). Anim. Behav., 34, 1463–1470. Simmons, L. W. & Parker, G. A. 1992. Individual variation in sperm competition success of yellow dung flies, Scatophaga stercoraria. Evolution, 46, 366–375. Siva-Jothy, M. T. & Hooper, R. E. 1995. The disposition and genetic diversity of stored sperm in females of the damselfly Calopteryx splendens xanthostoma (Charpentier). Proc. R. Soc. Lond. Ser. B, 259, 313–318. Smith, R. L. (ed.) 1984. Sperm Competition and the Evolution of Animal Mating Systems. New York: Academic Press. Ward, P. I. 1993. Females influence sperm storage and use in the yellow dung fly Scathophaga stercoraria (L.). Behav. Ecol. Sociobiol., 32, 313–319.