Present address: Institute of Integrative Biology, CHN K 16, ETH Zurich, Universitaetsstrasse 16,8092 Zurich, Switzerland.
Natural levels of polyandry: differential sperm storage and temporal changes in sperm competition intensity in wild yellow dung flies
Article first published online: 9 MAY 2011
© 2011 The Authors. Functional Ecology © 2011 British Ecological Society
Volume 25, Issue 5, pages 1079–1090, October 2011
How to Cite
Demont, M., Buser, C. C., Martin, O. Y. and Bussière, L. F. (2011), Natural levels of polyandry: differential sperm storage and temporal changes in sperm competition intensity in wild yellow dung flies. Functional Ecology, 25: 1079–1090. doi: 10.1111/j.1365-2435.2011.01861.x
- Issue published online: 22 SEP 2011
- Article first published online: 9 MAY 2011
- Received 11 January 2011; accepted 7 April 2011 Handling Editor : Sara Lewis
- cryptic female choice;
- post-copulatory sexual selection;
- sperm selection
1. Polyandry is common in insects. Nevertheless, the evolutionary causes and consequences of this phenomenon remain contentious, in part because of a lack of information about natural mating rates and the fact that most post-copulatory processes are hidden from view within female reproductive tracts.
2. We captured wild female yellow dung flies (Scathophaga stercoraria) over the whole spring season and genotyped the sperm from their spermathecae to obtain information on sperm transfer, sperm storage and natural levels of polyandry for this model species of post-copulatory sexual selection research.
3. On average, females stored sperm from a minimum of 2·47 males (based on the most conservative estimate). Incorporating knowledge of population allele frequencies yielded a slightly higher estimate of 3·33 mates per female.
4. Sperm storage and therefore sperm competition intensity showed high temporal variation. The proportion of multiply mated females (i.e. females with sperm from ≥2 males within their sperm stores) and the absolute number of ejaculates detected within females increased strongly over the spring season before sharply decreasing as midsummer approached.
5. Interestingly, we detected a positive relationship between the number of stored ejaculates and females’ wing injuries, suggesting that mating not only causes measurable cumulative damage to wild females but also provides a potential mechanism by which males may be able to assess the intensity of sperm competition within a female.
6. Our study found no evidence for intraejaculate sperm sorting, but importantly, the number of ejaculates in storage differed amongst the three sperm storage organs (spermathecae) of female yellow dung flies. Different sperm mixtures across the spermathecae could enable females to bias paternity towards certain males if females can selectively use sperm from a certain spermatheca at the time of fertilization.
Polyandry (females mating with more than one male) is surprisingly common in the animal kingdom. Nevertheless, the evolutionary causes and far-reaching consequences of polyandry remain debated (Arnqvist & Nilsson 2000; Jennions & Petrie 2000; Arnqvist & Rowe 2005; Simmons 2005; Evans & Simmons 2008). Polyandry is especially puzzling if there are no obvious direct benefits associated with mating, e.g. replenishing sperm stores or the acquisition of food from mates. In such cases, repeated mating by females might arise via a number of alternative non-adaptive [e.g. correlated response to sexual selection on multiple mating by males (Halliday & Arnold 1987)] or adaptive mechanisms (e.g. the acquisition of indirect benefits through mating males with high quality or compatible genes). The relative importance of these alternatives is currently unknown both in general and for many specific examples of polyandry. Further, the evolutionary interests of the sexes concerning mating frequency diverge dramatically, and numerous adaptations for persistence in males and resistance in females to force or avoid copulations are known (Chapman et al. 2003; Arnqvist & Rowe 2005). In the extreme, the female response to male persistence traits may simply be to mate to avoid additional harassment costs (i.e. convenience polyandry: Thornhill & Alcock 1983).
Whilst many important laboratory studies have attempted to clarify the forces acting on female mating rates (Tregenza & Wedell 2002; Martin & Hosken 2003), extrapolating results to wild populations remains difficult. This is partly because we often do not know whether laboratory conditions reflect the situation in wild populations (Bretman & Tregenza 2005; Simmons, Beveridge & Kennington 2007). This information gap is a considerable barrier to progress in the study of polyandry and its consequences and could lead to misinterpretation or inappropriate extrapolation of laboratory data. To circumvent such mistakes, we need detailed documentation of natural levels of polyandry in wild populations (Bretman & Tregenza 2005; Simmons, Beveridge & Kennington 2007; Frentiu & Chenoweth 2008; Simmons & Beveridge 2010), ideally assessing spatial and/or temporal variation in mating patterns and any ecological and evolutionary factors shaping selection on female remating rates in the field (Wilson 2009).
Assessing the degree of polyandry by directly observing mating in the field poses a challenge, especially for small and mobile species such as insects. One approach is to genotype sperm within female sperm stores to assess the number of mates (Chapuisat 1998; Krieger & Keller 2000). As copulations may not always result in successful sperm transfer, sperm from recent mates may have displaced sperm from previous males, and females may differentially retain or eject sperm from different males (Pizzari & Birkhead 2000; Garcia-Gonzalez 2004; Pizzari, Lovlie & Cornwallis 2004); this estimate of female mating frequency (genetic mating frequency) may underestimate the actual mating frequency in the field (social mating frequency). Nevertheless, estimates obtained by genotyping stored sperm from females are a good measure of the minimum degree of polyandry prevalent in the wild and a parameter that is probably more important for male sexual behaviour than the social mating frequency of females.
In addition to polyandry, sperm storage is the second prerequisite for certain mechanisms of cryptic female choice (e.g. sperm selection) and sperm competition. Therefore, thorough studies of polyandry in wild populations should be coupled with precise investigations of sperm storage patterns. Sperm competition is a particularly potent evolutionary force, shaping many aspects of male behaviour and anatomy (Parker 1970c; Simmons 2001). Females represent the arena in which sperm competition occurs, making the interdependence of post-copulatory mechanisms involved in sperm competition and cryptic female choice obvious. Numerous mechanisms affecting these processes have been identified (Eberhard 1996; Simmons 2001). However, the degree to which females influence sperm transfer, storage and sorting, and ultimately the ability to select particular sperm for fertilization, remains unclear (Birkhead & Pizzari 2002). Consequently, the relative contributions of sperm competition and cryptic female choice to differential fertilization success are largely unknown (Snook 2005). A more comprehensive understanding of processes occurring during and after copulations is one necessary step in unravelling the mysteries of post-copulatory processes (Hall et al. 2010; Manier et al. 2010). As for polyandry studies, once again we need this information from wild populations as well as from laboratory settings.
Varying levels of polyandry affect the number of ejaculates competing within females for fertilization, and also how much ejaculate from each male is present in the contest (Wedell, Gage & Parker 2002). Several studies have shown that males adjust their behaviour according to sperm competition risk (indicated by the level of polyandry). For example, when subject to higher risks, elephant seals (Mirounga angustirostris) are more aggressive towards rival males (Leboeuf & Peterson 1969), and flour beetles (Tenebrio molitor) increase their mate guarding (Gage & Baker 1991). Additionally, males may respond at the gametic level (Wedell, Gage & Parker 2002; Pizzari et al. 2003). Comparative studies have consistently shown positive relationships between the degree of polyandry (an index of sperm competition) and relative testis size (a standard index of investment in sperm) amongst related taxa (Gage 1994; Hosken 1997). Within species, males adjust ejaculate expenditure during a particular mating event according to cues arising from conspecifics (males and females). Previous studies provide evidence that increased risk of sperm competition displayed by the presence of rival males resulted in increased ejaculate size (Gage 1991; Pound & Gage 2004). Males also exhibit strategic ejaculation according to indicators of female mating status and/or quality (Wedell 1998; Martin & Hosken 2002). Exactly how males detect female mating status (e.g. virgin vs. mated) and/or the number of ejaculates stored by females is often unclear (Engqvist 2007; but see Carazo et al. 2004; Thomas & Simmons 2009). Importantly, the relationship between sperm competition risk (the probability that a female will mate with more than one male) and intensity (the number of males involved in sperm competition) is not straightforward (Engqvist & Reinhold 2005). For example, although there may be few males present at mating sites (i.e. low risk), females might have mated previously and stored sperm from several males (i.e. high intensity). This illustrates that cues arising from other males (e.g. operational sex ratio) and cues arising from the female (e.g. mating status) could affect males very differently. Just like research on polyandry in general, empirical research on strategic sperm allocation (a consequence of varying levels of polyandry) suffers a bias towards laboratory studies. Data from wild populations that directly assess the number of males involved in sperm competition are needed to help test predictions derived from theoretical models on the evolution of male sperm expenditure.
Yellow dung flies are found throughout the northern hemisphere (Parker 1970a), and phenologies differ according to altitude and latitude (Blanckenhorn & Demont 2004; Demont & Blanckenhorn 2008). In the lowlands of Switzerland (e.g. our study site), adult flies are typically present from late March or beginning of April to mid-June and from early September to mid-November (Blanckenhorn 1998). The spring and autumn seasons each consist at most of two generations (Blanckenhorn 1998). Taking into account temperature-dependent development and maturation, sexually mature dung flies of the second spring generation are expected on pastures in mid-May (Blanckenhorn et al. 2001). Copulations usually take place on dung pats or in the grass nearby. Directly after copulation, females lay eggs on the dung, where larvae hatch and develop. During oviposition, the male guards the female to prevent copulation with other males. The operational sex ratio is male-biased, and there is strong pre-copulatory male–male competition. A comprehensive study by Jann, Blanckenhorn & Ward (2000) revealed that in two consecutive years, the number of males per dung pat decreased, whilst the operational sex ratio increased over the spring season, i.e. more females per male were available in late spring. Additionally, the intensity of sexual selection favouring large males was strong in both years (0·499 ± 0·053 and 0·510 ± 0·051), and as expected increased with male density, and therefore decreased over the spring season (Jann, Blanckenhorn & Ward 2000; cf. seasonal fluctuations in sexual selection in a spider: Kasumovic et al. 2008).
Male–male competition seems to drive pre-copulatory sexual selection in S. stercoraria, and intra-sexual selection is also important for post-copulatory sperm transfer, storage and fertilization [e.g. copula duration and male size are both correlated with the number of sperm transferred (Parker & Simmons 1994, 2000)]. Nevertheless, theoretical models and data also support a significant female influence on sperm storage, displacement and utilization. In particular, the complex reproductive morphology of females may facilitate sperm choice. This morphology features a single spermatheca on one side of the body (the singlet) and two spermathecae on the other side (the doublet), each supplied with its own duct. There is some evidence for sperm selection based on male phosphoglucomutase (PGM) genotype (Ward 2000), and recent work using microsatellite competitive PCR demonstrated that the proportions of rival ejaculates differ between spermathecae (Bussiere et al. 2010). Such differences in sperm storage across spermathecae could enable females to bias paternity towards certain males, provided females are able to differentially use sperm from the three spermathecae for fertilization.
The present study was conducted to gather information on sperm storage patterns and natural levels of polyandry. Specifically, we captured wild dung fly females over the whole spring season and genotyped sperm from their spermathecae to address five questions: (i) How many ejaculates compete within the female sperm storage organs? (ii) Do patterns of sperm storage show temporal variation (i.e. does the proportion of multiply mated females and number of stored ejaculates change over time)? (iii) Does intraejaculate sperm sorting occur in female yellow dung flies? (iv) Does the number of stored ejaculates differ amongst spermathecae? (v) Does female phenotype (e.g. size and wing injuries) covary with patterns of sperm storage?
Materials and methods
We sampled dung flies from a pasture in Fehraltorf, Switzerland (8·55°E, 47·37°N), over virtually the whole spring season 2006. We sampled 92 females, each associated with a copulating male, on several dates: seven copulating pairs on 24th April (day 114); 22 on 25th April (day 115); 17 on 20th May (day 140); 13 on 25th May (day 145); 12 on 14th June (day 165); six on 16th June (day 167); and 15 on 22nd June (day 173). We ascertained that females were copulating by ensuring that genital contact was occurring [cf. males engage in prolonged post-copulatory guarding; (Parker 1970b)]. If pairs within vials were still copulating 15 min after capture, flies were separated to avoid unnaturally extended copulations in the absence of male–male competition and takeovers. All individuals were transported to the laboratory and frozen at −80 °C for later dissection, measurement and microsatellite genotyping.
Dissections and morphological measurements
We extracted sperm from the spermathecae using a method originally developed for Anopheles gambiae (Tripet et al. 2001). First, we separated the abdomen of the dung fly females from the rest of the body (this was immediately refrozen at −80 °C for subsequent processing). The abdomens were stored for 48 h in 70% ethanol (Tripet et al. 2001). Under a binocular microscope (Leica MZ-12; Leica Microsystems GmbH, Wetzlar, Germany), we then carefully removed the posterior part of the female reproductive tract (common oviduct, spermathecae, spermathecal ducts, accessory glands and the bursa copulatrix) by grasping the genital valves with forceps and tearing them from the abdomen. Next, we isolated the three spermathecae and transferred them individually to a drop of water. For every female, we could easily distinguish the singlet from the middle and outer doublet spermathecae (Hosken, Meyer & Ward 1999). We removed all tissue surrounding the spermatheca and carefully applied pressure to the spermathecal capsule to break open the spermatheca. As storage in 70% ethanol caused the ejaculate in the spermatheca to coagulate, we were able to remove a sperm pellet from each spermatheca (cf. Bussiere et al. 2010). The three sperm pellets from each female, each originating from a different spermatheca, were transferred to 180 μL of buffer solution (ATL buffer from the QIAamp® DNA Micro Kit; Qiagen, Hombrechtikon, Switzerland) and stored at −80 °C for subsequent DNA extraction.
From the rest of the females’ body (never stored in ethanol) and the males, we assessed several morphological characteristics. We measured left or right hind tibia length of all animals as an index of body size. In addition, we counted female left wing injuries according to previous work on this species by Burkhard, Ward & Blanckenhorn (2002). Wing injuries were classified according to three categories: tears, notches and large missing areas [for details see (Burkhard, Ward & Blanckenhorn 2002)]. Burkhard, Ward & Blanckenhorn (2002) suggested (but did not explicitly test) that tears and notches, i.e. small wing injuries, reflect regular wear, whilst large missing areas reflect intra- and/or interspecific interactions. In the present study, we adopted these suggestions and did not try to distinguish between alternative explanations for large missing areas (i.e. inter- vs. intraspecific interactions) nor between alternative explanations for intraspecific interactions (i.e. intersexual vs. intrasexual). We measured wing injuries of the left wing, because on average the number of injuries does not differ between wings (Burkhard, Ward & Blanckenhorn 2002).
Extraction, amplification and analysis of DNA
We performed DNA extraction from sperm pellets according to Bussiere et al. (2010) using a kit designed for small amounts of DNA sample (QIAamp® DNA Micro Kit; Qiagen AG, Hombrechtikon, Switzerland). We added carrier RNA to buffer AL (1 μL dissolved carrier RNA in 200 μL buffer AL) and used the minimum recommended amount of elution buffer AE (20 μL) to retain the highest possible concentration of sperm DNA. As described by Bussiere et al. (2010), we used the QIAGEN® Multiplex PCR Kit to simultaneously amplify four microsatellite loci: SsCa17, SsCa24, SsCa26 (Garner et al. 2000) and SsCa30 (Demont et al. 2008). Total PCR volume for the sperm pellets was 30 μL (cf. Bussiere et al. 2010 used only 24 μL): 5 μL DNA template, 15 μL QIAGEN Multiplex PCR Master Mix, 7 μL distilled water and 3 μL microsatellite primer mix (100 μm). Cycling conditions for the sperm were as follows: 95 °C for 15 min, then 30 cycles of 94 °C for 30 s, 60 °C for 3 min and 72 °C for 45 s and finally 60 °C for 30 min. These cycling conditions did not produce large stutter peaks for any of the four applied markers.
We used a Chelex extraction method to extract DNA from the heads of all flies. Cropped heads were transferred into 96-well PCR plates kept on ice. We pipetted 100 μL of 6% Chelex suspension (Chelex 100®, Na+-form, particle size 50–100 mesh, Fluka, Buchs, Switzerland) into each well using wide-ended tips. Afterwards, we covered the plate with a plastic mat, carefully shook it and spun down the heads to ensure that the sample was covered in liquid. We used a thermocycler to incubate the plates for 60 min at 55 °C, boil for 9 min at 100 °C and then cool down to 20 °C. After removing samples from the thermocycler, we again carefully shook and spun them down, before the plate was stored at 4 °C for 10–20 h, and the samples were frozen at −20 °C for at least 24 h before using for subsequent DNA amplification. DNA template amount (1 μL), total PCR volume (6 μL) and cycling parameters (number of cycles: 27) were the same as described by Bussiere et al. (2010).
PCR products from sperm and heads were separated on a capillary sequencer (Applied Biosystems 3730 DNA Analyzer, Applied Biosystems, Rotkreuz, Switzerland), and outputs analysed using Applied Biosystems GeneMapper® software. Head genotypes were simple to score, but sperm samples were more challenging because of the number of alleles present. To avoid artificial inflation of our estimate of the number of males present in the spermathecae, we ignored very small peaks on either side of a large peak because they potentially represent stutter peaks. In the paragraphs below, we describe two different procedures for estimating the number of males.
First, the most conservative method counts alleles and divides by two. To do so, we first checked for maternal contamination, and when maternal alleles were present in the allele array, we discounted them. We then identified alleles from the last male in the array and subtracted those from the total. We then divided the remaining alleles by two, because every male could potentially be heterozygous, rounding up when there were an odd number of remaining alleles. Our estimate of the minimum number of mates for this focal female was this resulting number [i.e. remaining alleles divided by two, rounded up, plus 1 (i.e. last male)]. Note that incorporating the last male genotype can improve this conservative estimate of the minimum number of mates compared with pure allele counting when the last male is homozygous: an array of four alleles including a homozygous last male gives a minimum estimate of three males, whilst pure allele counting and dividing by two would have produced a minimum estimate of only two males. We thus obtained four estimates (from four loci) of the minimum number of males present in a spermatheca and used the greatest number of them as our estimate of the minimum number of males present in a spermatheca.
Our second method applied the probabilistic technique described by Bretman & Tregenza (2005) to estimate the number of males within each spermatheca. This technique uses population allele frequencies to determine the probability of observing a certain array of alleles when different numbers of males contribute. The probability of not observing an allele is Pnot observed = [1-f(a)]t, where f(a) is the allele frequency and t the number of attempts at observing the allele, which is twice the number of males. The probability of observing an allele is Pobserved = 1–Pnot observed. The probability of obtaining the observed array of alleles is calculated as the product of Pobserved for alleles present in the array and Pnot observed for those alleles in the population that are not present in the observed array. We implemented this formula in a short program in matlab version 7.8 (MathWorks, Bern, Switzerland) and calculated the probability of receiving the observed array of alleles if a female mated with 1–50 males, representing t = 2 to 100 (Bretman & Tregenza 2005). We did this for all four loci for all spermathecae. The number of attempts with the highest probability indicates the most likely number of males that generate the observed array. For our mixed model analyses, we used the estimate derived from the most polymorphic locus (not necessarily the same locus for all spermathecae).
Finally, we checked the sperm from the last male for the occurrence of intraejaculate sperm sorting, i.e. whether the different sperm from a heterozygous last male was found in the same spermatheca or in different spermathecae.
We analysed the influence of season and other variables on the binary response variable female multiple mating status (yes or no, i.e. if females had sperm from a single male or from several males within their sperm stores) with generalized linear models in R version 2.6.2 using the glm function from the stats package (R Development Core Team, 2008). We preferred analyses with a binary response variable over analyses with proportion data, because we had unique values of different explanatory variables for every individual case (Crawley 2007). Generalized linear models were fitted with binomial errors and logit link function. The explanatory variables included day in the spring season when flies were caught, female size, last male size, tears and notches combined (as an index of female age), and large wing injuries. We started model simplification with a maximal model that included higher powers of the explanatory variables and all two-way interactions. We performed model simplification based on information-theoretic approaches Akaike Information Criterion (AIC) and by using deletion tests (chi-squared tests).
We analysed the influence of season and other variables on sperm storage (e.g. number of males detected within sperm stores) with linear mixed models in R version 2.6.2 (R Development Core Team, 2008) using the lme function from the nlme package (Pinheiro et al., 2008). Linear models are preferred over generalized linear models when the variance increases with the mean on the original scale of measurement (Crawley 2007). We therefore log10 transformed the response variable instead of using generalized linear mixed models with Poisson errors and log link. The explanatory variables included day in the spring season when females were caught, the spermathecal identity (i.e. whether the focal spermatheca was the singlet, middle doublet or outer doublet), female size, last male size, tears and notches combined, and large wing injuries. As nonparametric smoothers in generalized additive models clearly indicated curvature in the relationship between number of males detected in the sperm stores and day in the year when flies were caught (data not shown), we also included higher powers of day in the year (day2 and day3) as explanatory variables in our models. The random effects were the spermatheca nested within female. We performed model selection based on information-theoretic approaches (AIC) and by using hypothesis tests (i.e. testing simpler nested models against more complex models: likelihood ratio tests). Model selection and final model fitting was performed for both estimates (minimum estimate and the probabilistic estimate) of the response variable. Generalized linear models and linear mixed models were fitted by maximum likelihood during the process of model selection, whilst the final model was fitted by restricted maximum likelihood.
From the total 92 females collected, four females were not included in the final analyses: one female escaped before being frozen, one female had no sperm in her spermathecae and two females had four spermathecae (so would have complicated analysis of spermathecal identity). The occurrence of only two females with four spermathecae in our sample (2/92 = 2·2%) is considerably lower than previously reported for the same population [ca. 10%, see (Ward 2000)].
Last male sperm and intraejaculate sperm sorting
The fate of last male sperm was examined in 88 females (3 × 88 = 264 spermathecae). Although all males had copulated for ≥15 min with the female, in four cases we were not able to detect alleles from the last male in any of the three spermathecae. This finding could indicate that successful copulation does not always imply successful sperm transfer in the field (see Discussion). In the remaining 84 cases where we could detect sperm from the last male, the sperm was almost always present in all spermathecae (81 females). In three females, the sperm from the last male was absent in one spermatheca (the singlet in one female and the outer doublet in the other two females). Further, we found no indication that females are able to distinguish and separate sperm from one particular male, and the two types of sperm produced by a heterozygous last male were always found within the same spermatheca or were always absent from a particular spermatheca.
Natural level of polyandry
In 72 females (81·8% of females), we detected sperm from two or more males within the sperm stores. The remaining 16 females had sperm from only one male stored in their spermathecae: in 14 females this sperm belonged to the last male, and in two cases this stored sperm was from another male (i.e. not from the male that was captured together with the female). On average, females stored sperm from 2·47 ± 0·13 (mean ± SE) males for the minimum estimate (range: 1–6), or 3·33 ± 0·24 based on the probabilistic estimate (range: 1–11).
From the 264 genotyped spermathecae (from 88 females), 10 spermathecae provided ambiguous arrays of alleles (i.e. very weak peaks). As the replicated PCR runs of these spermathecae resulted in the same weak unreadable electropherograms, these 10 spermathecae obtained an NA (i.e. not available) in our mixed model analyses in R. Note that a missing value for a certain spermatheca did not imply an unavailable estimate for the other two spermathecae or the female as a whole; hence, our sample sizes can differ across analyses.
Incidence of multiply mated females
We analysed the incidence of multiply mated females (i.e. sperm from two or more males within their spermathecae) with generalized linear models with binary response variable, binomial errors and logit link. The summary of the best model in terms of the AIC is given in Table 1. Only the quadratic and cubic term of day significantly influenced the incidence of multiply mated females (Table 1). The proportion of multiply mated females sharply increased at the beginning of the spring season, stayed on a high level (>90%) until mid-June (day 167) and then decreased to c. 60% at the last sampling day (Fig. 1). The remaining terms (female size, last male size, tears and notches, large wing injuries and all interactions) did not explain variation in multiple mating (Table 1).
|d.f.||Deviance||Residual d.f.||Residual deviance||P value|
|Day of the year||1||1·510||84||81·126||0·21915|
|Last male size||1||0·193||82||80·648||0·66083|
|Tears and notches||1||1·247||81||79·401||0·26414|
|Large wing injuries||1||1·222||80||78·179||0·26902|
|(Day of the year)2||1||7·938||79||70·241||0·00484*|
|(Day of the year)3||1||4·793||78||65·449||0·02858*|
|Day of the year × large wing injuries||1||3·705||77||61·744||0·05425|
Sperm storage, number of males and wing injuries
We analysed sperm storage patterns (i.e. log10-transformed number of males detected within the spermathecae) with linear mixed models in R. The summary of the best model in terms of the AIC for minimum and probabilistic estimates of number of males detected is given in Table 2. The models presented in Table 2 were achieved by removing interactions with P > 0·30, and none of the comparisons of models with progressively simplified fixed effects yielded a significant contribution of a single interaction term (all likelihood ratio tests: P > 0·18). Removal of the spermatheca X female size interaction from the model summarized in Table 2 caused a slight increase in the AIC value. Results from analyses with the probabilistic estimate of number of males detected within the spermathecae of females (Bretman & Tregenza 2005) were not qualitatively different from the analyses with the minimum estimate, with one small exception: the term spermatheca in the final model was marginally non-significant (Table 2).
|Source||Numerator d.f.||Denominator d.f.||ME as response||PE as response|
|F value||P value||F value||P value|
|Day of the year||1||77||14·7012||0·0003*||13·3769||0·0005*|
|Last male size||1||77||1·9736||0·1641||2·5835||0·1121|
|Tears and notches||1||77||0·0978||0·7553||0·0961||0·7574|
|Large wing injuries||1||77||12·5939||0·0007*||13·4472||0·0004*|
|(Day of the year)2||1||77||10·0765||0·0022*||11·9623||0·0009*|
|(Day of the year)3||1||77||11·9123||0·0009*||8·7569||0·0041*|
|Spermatheca × female size||2||151||0·9893||0·3742||1·2696||0·2839|
|Spermatheca × last male size||2||151||3·3522||0·0376*||4·3933||0·0140*|
|Spermatheca × tears and notches||2||151||2·7401||0·0678||1·5110||0·2240|
|Last male size × large wing injuries||1||77||1·9745||0·1640||2·1774||0·1441|
Day of the year, large wing injuries and spermathecal identity (in the more conservative model) significantly influenced observed sperm storage patterns (Table 2). Additionally, the effect of spermathecal identity depended on the size of the last male in both models (significant spermatheca X male size interaction term; see Table 2). The three significant effects of day (day in the year, quadratic term and cubic term) highlight that sperm storage patterns and hence probably also sperm competition intensity strongly vary within the spring season (Table 2, Fig. 2). The number of males detected within the sperm stores of females continuously increases from April (day 114) until mid-June (day 165), before abruptly decreasing in the last 2 weeks dung flies are present at our sampling site in spring (Fig. 2). Sperm storage inferred from including population allele frequencies (i.e. the probabilistic estimate) unsurprisingly produced higher estimates of the number of males within the sperm stores of females, but showed the same seasonal pattern as the minimum estimates (Fig. 2). The significant effect of spermatheca from the minimum estimate model (marginally non-significant in the probabilistic model) indicates a consistently lower number of sperm from different males for the singlet spermatheca (s1) compared to middle doublet (s2) and outer doublet spermathecae (s3) (Table 2, Fig. 3). Again, patterns inferred from both estimates were practically identical, although inclusion of population allele frequencies causes the difference between middle doublet and outer doublet spermathecae to disappear (Fig. 3). In addition, the number of sperm from different males found within the spermathecae significantly increased with increasing number of large wing injuries (Table 2, Fig. 4). The significant interaction between spermatheca and size of the last male indicates an effect of the size of the last male on the sperm from different males detectable within the spermathecae: the larger the last male, the fewer males are detected within each spermatheca, but this decrease occurs differently in the three spermathecae (strongest decrease in the singlet spermatheca, Fig. 5). The size of the female and tears and notches (i.e. small wing injuries) did not contribute to variance in sperm storage patterns.
Wing injuries changed considerably in abundance throughout the spring season (Fig. 6). Small injuries like tears and notches were more numerous than large wing injuries and peaked in the middle (day 140) and at the end (day 173) of the sampling period (Fig. 6). In contrast, large wing injuries exhibited only one peak in mid-June (day 165: Fig. 6).
Our study provides estimates of the prevailing levels of polyandry and temporal changes in sperm competition intensity for a natural population of yellow dung flies. Field data in this context are still scarce (cf. Bretman & Tregenza 2005; Simmons, Beveridge & Kennington 2007; Frentiu & Chenoweth 2008; Simmons & Beveridge 2010). To our knowledge, only one other study has documented temporal changes in sperm competition intensity in a natural population of non-social insects (Simmons, Beveridge & Kennington 2007). Interestingly, large wing injuries explain the number of ejaculates stored within female spermathecae, providing a cue males could potentially use to assess sperm competition intensity before or during matings. Furthermore, we demonstrate that the number of ejaculates in storage differs amongst spermathecae. Such differential sperm storage could represent a basis for female influence on paternity at the time of fertilization, by differentially utilizing sperm from the three spermathecae (Hellriegel & Ward 1998).
Level of polyandry and temporal changes in sperm competition intensity
Our study revealed high levels of polyandry in a natural population of yellow dung flies: 81·8% of females stored sperm from two or more males within their sperm stores. On average, 2·47 or 3·33 ejaculates compete within the sperm storage organs of wild flies, based on the minimum or probabilistic estimate, respectively. Studies investigating direct or indirect benefits of polyandry in this species (Hosken et al. 2003; Tregenza et al. 2003) and studies investigating evolutionary responses to polyandry via experimental evolution (Hosken 2001; Hosken, Garner & Ward 2001; Hosken & Ward 2001) have used laboratory experiments in which females were mated to two or three males. Our results suggest that investigating the causes and consequences of polyandry in dung flies with double or triple matings is a reasonable starting point, because on average two to four ejaculates are found simultaneously within the spermathecae of females. However, there is also substantial variation in the number of ejaculates stored, and this number (which reflects sperm competition intensity) exhibits strong temporal variability. The most likely explanation for this temporal pattern is that females mate repeatedly with different males and continuously accumulate sperm from different males as they become older. The decline in the second half of June could arise because only young flies from the second spring generation (who have mated fewer times) are still present on cow pats. Surprisingly, our study detected an increase in the number of stored ejaculates until mid-June, but the second spring generation adults are assumed to be on the pasture from the mid-May onwards (Blanckenhorn et al. 2001). Our data on minor damage to wings would support this assumption, as the frequency of damage decreases sharply in the second half of May. If the number of sperm from different males stored by females is assumed to primarily reflect female age, then arrival of new young females of the second spring generation would actually cause a decrease in the number of ejaculates competing within females from mid-May (i.e. earlier than observed). This is because from mid-May samples would consist of old females from the first spring generation and young females from the second generation. However, the sample from 14th June only comprised females that had stored sperm from three or more males. The relatively small sample size on this date (n = 12) could explain the observed pattern if the ratio of old to young females was biased towards old females, and we therefore collected only old females by chance. However, the occurrence of females with relatively few tears and notches in this sample contradicts this scenario. The pattern of tears and notches indicates that old and young females were present mid-June and that all of them had already mated many times.
At the end of the spring season, only young, or alternatively females that had not yet mated with many males, were found on and around cow pats. Assuming that tears and notches reflect regular wear (so are a useful indicator of age) and large wing injuries rather indicate inter- or intraspecific (intersexual or intrasexual) interactions [so may indicate ‘activity’, flies spend time on or near the dung pats, where copulations and ovipositions took place and not in the near forest in a kind of quiescence (cf. Blanckenhorn et al. 2001)] as proposed by Burkhard, Ward & Blanckenhorn (2002), then the last females present on cow pats at the end of the spring season are old flies that have not been very active (cf. Fig. 6). At present, these are only speculations, but the possibility that sperm storage patterns (e.g. number of mates) are determined to a greater extent by activity rather than directly by age (tears and notches) is a fascinating idea that requires further investigation. In this context, a precise determination of the different sources of wing injuries (i.e. inter- vs. intraspecific interactions, and within the latter also intersexual vs. intrasexual) would provide a useful tool for future studies.
Temporal changes in the number of ejaculates represented within females and associated changes in sperm competition intensity have three important implications. First, our average of 2·47 or 3·33 different ejaculates within the spermathecae arises from genotyping both young and old females. Consequently, young females (or alternatively ‘less active’ females) may represent a lower than average level of sperm competition intensity, and older females a higher level. Our data also demonstrated that females mate with up to six or 11 males in the field based on our two different estimates. This is in close agreement with an earlier study, which reported that females maximally produce seven clutches of eggs in the field (Gibbons 1987). This is a much higher level of polyandry than commonly applied in laboratory investigations in yellow dung flies. Future research should investigate costs and benefits of polyandrous behaviour when females are mated across the whole range of multiple matings observed in natural populations. This could perhaps be complemented with experimental evolution applying different polyandry levels [previous work of this type contrasted monogamy vs. one level of polyandry with three males; (Hosken 2001; Hosken, Garner & Ward 2001; Hosken & Ward 2001)].
Second, increased sperm competition intensity as the spring season advances contrasts sharply with pre-copulatory sexual selection patterns. Density measured as the number of males per pat is highest at the beginning of spring and then decreases significantly over the spring season. The operational sex ratio (number of females divided by the number of males at the mating site) is likewise lowest at the beginning of spring and then increases significantly over the spring season, indicating that more females per male are available in late spring (Jann, Blanckenhorn & Ward 2000). Jann, Blanckenhorn & Ward (2000) additionally showed that pre-copulatory male–male competition increased with competitor density and consequently decreased over time in spring. In contrast, here we show that sperm competition intensity increased over time in spring before declining abruptly at the end of the season. Therefore, dung flies are confronted with high levels of pre-copulatory male–male competition (based on the pattern described by Jann, Blanckenhorn & Ward (2000) and personal observations in subsequent years) and low levels of sperm competition (this study) early in spring, and exactly the opposite pattern late in spring. Standardized estimates of post-copulatory sexual selection intensity for male paternity success have not been conducted in yellow dung flies, and this might best be achieved by also investigating temporal patterns of variation. Temporal variation in the intensity and form of pre-copulatory and post-copulatory sexual selection could contribute to the maintenance of genetic diversity in this species.
Third, the pronounced temporal changes in the number of ejaculates present within the spermathecae of females could greatly influence sperm investment by males. Several studies have reported evidence for males ejaculating strategically depending on the risk or intensity of sperm competition or depending on the female ‘quality’ or ‘condition’ (Wedell, Gage & Parker 2002). In some of these studies, scientists were also able to uncover how males detect female mating status or sperm competition intensity (Carazo et al. 2004; Thomas & Simmons 2009). In yellow dung flies, female size influences investment in ejaculate size: males copulate longer with larger females (Parker et al. 1999). However, no study has yet specifically investigated strategic sperm allocation in yellow dung flies according to sperm competition intensity or the presence of rival males. Our data clearly indicate that sperm competition intensities change strongly over the season. Furthermore, we reveal a link between the number of large wing injuries and the number of ejaculates the female had stored within their spermathecae, potentially providing males with a cue to assess sperm competition intensity. It would be interesting to manipulate female mating history and wing injuries independently of each other and to investigate how males respond to them (e.g. number of sperm transferred). Additionally, it would be interesting to see whether the seasonal changes in sperm competition intensity are reflected in seasonal changes in investment in reproductive tissue (e.g. testis) or copula duration.
Biases in sperm sorting
The process by which a female has a mixture of sperm from different males in her sperm storage organ(s) and selectively uses sperm from a particular male at fertilization is referred to as sperm selection (Simmons 2001). It is the most cryptic and controversial mechanism of female choice, in part because few studies have documented evidence in favour of it. Ward (2000) found indications that S. stercoraria females are able to select sperm on the basis of their PGM genotype. Matching offspring PGM genotype with prevailing environmental conditions has the potential to increase offspring performance. However, not all predictions from an adaptive sperm selection scenario were supported in that experiment (Ward 2000), raising questions about the precise extent or context in which females can exert sperm selection. A recent population genetics study contributed to this controversy by showing that PGM is in fact neutral (i.e. not under selection) in S. stercoraria (Demont et al. 2008). Nevertheless, differential sperm storage across the spermathecae as found here has been suggested to be a necessary prerequisite for females to exert sperm selection (Hellriegel & Ward 1998; Hellriegel & Bernasconi 2000). If sperm from different males is stored in different proportions across the spermathecae, then females may subtly influence paternity by preferentially utilizing sperm from a certain spermatheca for fertilization. Laboratory experiments clearly indicate that sperm is stored in different proportions across the spermathecae and that time elapsed between copulations might be critically important in determining sperm sorting (Bussiere et al. 2010). In agreement with these laboratory findings, the present study, as well as another field study (Demont 2010), showed that the number of ejaculates stored also differs amongst spermathecae in wild flies. Taking into account sperm storage patterns from the last male and results from Bussiere et al. (2010), it seems that sperm from the last male is usually stored in different proportions across the spermathecae and that these sperm are subsequently displaced by further mates. Different proportions of sperm in the spermathecae that are differentially displaced by subsequent sperm result in the sperm storage pattern described in the present study: different number of ejaculates present in the spermathecae. Spermathecae consisting of unequal sperm mixtures could indeed enable females to bias paternity towards certain males. But further studies are needed to investigate when and to what degree this is possible (cf. Demont 2010).
The significant interaction between spermathecal identity and the size of the last male indicated that females with larger last mates stored fewer competing ejaculates and that this effect was strongest in the singlet spermatheca. The decrease in the number of ejaculates present within the spermathecae could be explained by the previously documented higher rates of sperm transfer of large males (Parker & Simmons 1994, 2000). The subtle difference in slope could arise because spermathecae differ in size: the singlet spermatheca is on average the smallest one (Demont 2010) and hence showed the highest displacement rate in the present study.
Last male sperm and intraejaculate sperm sorting
In four of the 88 females, we could not detect sperm of the last male in storage, although pairs had copulated for at least 15 min. This could indicate that in nature not all copulations lead to successful sperm transfer. Indeed, rates of non-sperm representation owing to insemination failures or other reasons may be high across insects (Garcia-Gonzalez 2004). One possibility is that sperm from the last male had not yet reached the spermathecae because of the abbreviated copulations. Copulations normally last around 35 min, although copula duration decreases with repeated matings (Parker 1970d). P2 (the proportion of paternity assigned to the second of two copulating males) increases with copula duration (Parker 1970d). Similarly, S2 (the proportion of stored sperm assigned to the second of two copulating males) in the spermathecae increases with copula duration (Simmons, Parker & Stockley 1999). However, mean S2 in the spermathecae after 15 min is only about 35% and associated with considerable variation [see Fig. 3b in (Simmons, Parker & Stockley 1999)], emphasizing the possibility that in certain copulations no sperm of the second male (or last male) is found in the spermathecae after c. 15 min. An ongoing study investigating sperm storage patterns in singly mated females where copulations have been interrupted after 20 min confirms that in some spermathecae no sperm is present (C. Wüst, M. Demont, C. Buser, and L. F. Bussière, unpublished data). Unfortunately, in the present study, we are unable to distinguish between failed inseminations, successful inseminations in which the sperm did not yet reach the spermathecae, or cryptic female choice against certain sperm.
As virtually all captured males were heterozygous at one or more loci, we were able to investigate whether intraejaculate sperm sorting occurs (i.e. whether different alleles from the same male are stored in different spermathecae). Intraejaculate sperm sorting could facilitate sperm selection, which can be beneficial in certain situations, for example as a means of producing offspring of one particular sex (Simmons 2001). However, in the present study, we found absolutely no evidence for intraejaculate sperm sorting in yellow dung flies: both alleles of a heterozygous male were always either present in or absent from a particular spermatheca. As the applied microsatellite loci are not located on either sex chromosome (Garner et al. 2000; Demont et al. 2008), this conclusion only refers to autosomes in yellow dung flies. The possibility of intraejaculate sperm sorting and selection based on sex chromosomes remains to be evaluated in this species.
Our study provides essential estimates of levels of polyandry and temporal changes in sperm competition intensity for a natural population of yellow dung flies. Our data also demonstrated a specific and simple visual cue (large wing injuries) through which males could assess prevailing sperm competition intensity. We additionally showed that the number of ejaculates in storage differs amongst spermathecae. Field data as presented here could establish the basis for subsequent detailed studies on sperm storage and utilization, strategic ejaculation, cues indicating sperm competition intensity, and comparison between pre- and post-copulatory sexual selection. More generally, such data can help improve laboratory experiments for investigating polyandry and associated aspects of sexual selection. Summarized, field data on multiple mating, sperm storage, post-copulatory processes and paternity are not just a welcome complement to laboratory data, but crucial in order to acquire a full understanding of sexual selection.
We thank Christian Wüst for providing us with a MATLAB program to obtain our probabilistic estimates. Thanks also to Sarah Ravaioli for helping with dissections, and Erik Postma, Wolf Blanckenhorn and Andy Hector for discussions on statistical modelling. Funding was provided by the University of Zurich and a Claraz-Stiftung grant to MD. OYM is funded by the Swiss National Science Foundation (SNF).
- 2000) The evolution of polyandry: multiple mating and female fitness in insects. Animal Behaviour, 60, 145–164. & (
- 2005) Sexual Conflict. Princeton University Press, Princeton, New Jersey. & (
- 2002) Postcopulatory sexual selection. Nature Reviews Genetics, 3, 262–273. & (
- 1998) Altitudinal differentiation in the diapause response of two species of dung flies. Ecological Entomology, 23, 1–8. (
- 2004) Bergmann and converse Bergmann latitudinal clines in arthropods: Two ends of a continuum? Integrative and Comparative Biology, 44, 413–424. & (
- 2001) Summer decline in populations of the yellow dung fly: diapause or quiescence? Physiological Entomology, 26, 260–265. , , & (
- 2005) Measuring polyandry in wild populations: a case study using promiscuous crickets. Molecular Ecology, 14, 2169–2179. & (
- 2002) Using age grading by wing injuries to estimate size-dependent adult survivorship in the field: a case study of the yellow dung fly Scathophaga stercoraria. Ecological Entomology, 27, 514–520. , & (
- 2010) The assessment of insemination success in yellow dung flies using competitive PCR. Molecular Ecology Resources, 10, 292–303. , , , & (
- 2004) Chemosensory cues allow male Tenebrio molitor beetles to assess the reproductive status of potential mates. Animal Behaviour, 68, 123–129. , , & (
- 2003) Sexual conflict. Trends in Ecology and Evolution, 18, 41–47. , , & (
- 1998) Mating frequency of ant queens with alternative dispersal strategies, as revealed by microsatellite analysis of sperm. Molecular Ecology, 7, 1097–1105. (
- 2007) The R Book. John Wiley & Sons Ltd, West Sussex, England. (
- 2010) Polyandry and Postcopulatory Sexual Selection in Yellow Dung Flies. PhD thesis, University of Zurich, Zurich. (
- 2008) Genetic differentiation in diapause response along a latitudinal cline in European yellow dung fly populations. Ecological Entomology, 33, 197–201. & (
- 2008) Molecular and quantitative genetic differentiation across Europe in yellow dung flies. Journal of Evolutionary Biology, 21, 1492–1503. , , & (
- 1996) Female Control: Sexual Selection by Cryptic Female Choice. Princeton University Press, Princeton, New Jersey. (
- 2007) Male scorpionflies assess the amount of rival sperm transferred by females’ previous mates. Evolution, 61, 1489–1494. (
- 2005) Pitfalls in experiments testing predictions from sperm competition theory. Journal of Evolutionary Biology, 18, 116–123. & (
- 2008) The genetic basis of traits regulating sperm competition and polyandry: can selection favour the evolution of good- and sexy-sperm? Genetica, 134, 5–19. & (
- 2008) Polyandry and paternity skew in natural and experimental populations of Drosophila serrata. Molecular Ecology, 17, 1589–1596. & (
- 1991) Risk of sperm competition directly affects ejaculate size in the mediterranean fruit-fly. Animal Behaviour, 42, 1036–1037. (
- 1994) Associations between body-size, mating pattern, testis size and sperm lengths across butterflies. Proceedings of the Royal Society of London Series B-Biological Sciences, 258, 247–254. (
- 1991) Ejaculate size varies with sociosexual situation in an insect. Ecological Entomology, 16, 331–337. & (
- 2004) Infertile matings and sperm competition: the effect of “nonsperm representation” on intraspecific variation in sperm precedence patterns. American Naturalist, 164, 457–472. (
- 2000) Polymorphic DNA microsatellites identified in the yellow dung fly (Scathophaga stercoraria). Molecular Ecology, 9, 2207–2208. , , , , , & (
- 1987) The causes of seasonal changes in numbers of the yellow Dung Fly, Scathophaga stercoraria (Diptera: Scathophagidae). Ecological Entomology, 12, 173–185. (
- 2010) Competitive PCR reveals the complexity of postcopulatory sexual selection in Teleogryllus commodus. Molecular Ecology, 19, 610–619. , , , & (
- 1987) Multiple mating by females – a perspective from quantitative genetics. Animal Behaviour, 35, 939–941. & (
- 2000) Female-mediated differential sperm storage in a fly with complex spermathecae, Scatophaga stercoraria. Animal Behaviour, 59, 311–317. & (
- 1998) Complex female reproductive tract morphology: its possible use in postcopulatory female choice. Journal of Theoretical Biology, 190, 179–186. & (
- 1997) Sperm competition in bats. Proceedings of the Royal Society of London Series B-Biological Sciences, 264, 385–392. (
- 2001) Sex and death: microevolutionary trade-offs between reproductive and immune investment in dung flies. Current Biology, 11, R379–R380. (
- 2001) Sexual conflict selects for male and female reproductive characters. Current Biology, 11, 489–493. , & (
- 1999) Internal female reproductive anatomy and genital interactions during copula in the yellow dung fly, Scathophaga stercoraria (Diptera: Scathophagidae). Canadian Journal of Zoology-Revue Canadienne De Zoologie, 77, 1975–1983. , & (
- 2001) Experimental evidence for testis size evolution via sperm competition. Ecology Letters, 4, 10–13. & (
- 2003) Superior sperm competitors sire higher-quality young. Proceedings of the Royal Society of London Series B-Biological Sciences, 270, 1933–1938. , , , & (
- 2000) Temporal and microspatial variation in the intensities of natural and sexual selection in the yellow dung fly Scathophaga stercoraria. Journal of Evolutionary Biology, 13, 927–938. , & (
- 2000) Why do females mate multiply? A review of the genetic benefits. Biological Reviews, 75, 21–64. & (
- 2008) Spatial and temporal demographic variation drives within-season fluctuations in sexual selection. Evolution, 62, 2316–2325. , , & (
- 2000) Mating frequency and genetic structure of the Argentine ant Linepithema humile. Molecular Ecology, 9, 119–126. & (
- 1969) Social status and mating activity in elephant seals. Science, 163, 91–93. & (
- 2010) Resolving mechanisms of competitive fertilization success in Drosophila melanogaster. Science, 328, 354–357. , , , , & (
- 2002) Strategic ejaculation in the common dung fly Sepsis cynipsea. Animal Behaviour, 63, 541–546. & (
- 2003) The evolution of reproductive isolation through sexual conflict. Nature, 423, 979–982. & (
- 1970a) Reproductive behaviour and nature of sexual selection in Scatophaga stercoraria L. (Diptera Scatophagidae) .1. diurnal and seasonal changes in population density around site of mating and oviposition. Journal of Animal Ecology, 39, 185–204. (
- 1970b) Reproductive behaviour and nature of sexual selection in Scatophaga stercoraria L (Diptera Scatophagidae) .2. Fertilization rate and spatial and temporal relationships of each sex around site of mating and oviposition. Journal of Animal Ecology, 39, 205–228. (
- 1970c) Sperm competition and its evolutionary consequences in insects. Biological Reviews of the Cambridge Philosophical Society, 45, 525–567. (
- 1970d) Sperm competition and its evolutionary effect on copula duration in the fly Scatophaga stercoraria. Journal of Insect Physiology, 16, 1301–1328. (
- 1994) Evolution of phenotypic optima and copula duration in dungflies. Nature, 370, 53–56. & (
- 2000) Optimal copula duration in yellow dung flies: ejaculatory duct dimensions and size-dependent sperm displacement. Evolution, 54, 924–935. & (
- 1999) Optimal copula duration in yellow dung flies: effects of female size and egg content. Animal Behaviour, 57, 795–805. , , , & (
- the R Development Core Team (2008) nlme: linear and Nonlinear Mixed Effects Models. , , , &
- 2000) Female feral fowl eject sperm of subdominant males. Nature, 405, 787–789. & (
- 2004) Sex-specific, counteracting responses to inbreeding in a bird. Proceedings of the Royal Society B Biological Sciences, 271, 2115–2121. , & (
- 2003) Sophisticated sperm allocation in male fowl. Nature, 426, 70–74. , , , & (
- 2004) Prudent sperm allocation in Norway rats, Rattus norvegicus: a mammalian model of adaptive ejaculate adjustment. Animal Behaviour, 68, 819–823. & (
- R Development Core Team (2008) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.
- 2001) Sperm Competition and its Evolutionary Consequences in the Insects. Princeton University Press, Princeton, New Jersey. (
- 2005) The evolution of polyandry: sperm competition, sperm selection, and offspring viability. Annual Review of Ecology Evolution and Systematics, 36, 125–146. (
- 2010) The strength of postcopulatory sexual selection within natural populations of field crickets. Behavioral Ecology, 21, 1179–1185. & (
- 2007) Polyandry in the wild: temporal changes in female mating frequency and sperm competition intensity in natural populations of the tettigoniid Requena verticalis. Molecular Ecology, 16, 4613–4623. , & (
- 1999) Sperm displacement in the yellow dung fly, Scatophaga stercoraria: an investigation of male and female processes. American Naturalist, 153, 302–314. , & (
- 2005) Sperm in competition: not playing by the numbers. Trends in Ecology & Evolution, 20, 46–53. (
- 2009) Male-derived cuticular hydrocarbons signal sperm competition intensity and affect ejaculate expenditure in crickets. Proceedings of the Royal Society B-Biological Sciences, 276, 383–388. & (
- 1983) The Evolution of Insect Mating Systems. Harvard University Press, Cambridge, Mass. & (
- 2002) Polyandrous females avoid costs of inbreeding. Nature, 415, 71–73. & (
- 2003) Maternal effects on offspring depend on female mating pattern and offspring environment in yellow dung flies. Evolution, 57, 297–304. , , & (
- 2001) DNA analysis of transferred sperm reveals significant levels of gene flow between molecular forms of Anopheles gambiae. Molecular Ecology, 10, 1725–1732. , , , , & (
- 2000) Cryptic female choice in the yellow dung fly Scathophaga stercoraria (L.). Evolution, 54, 1680–1686. (
- 1998) Sperm protection and mate assessment in the bushcricket Coptaspis sp. 2. Animal Behaviour, 56, 357–363. (
- 2002) Sperm competition, male prudence and sperm-limited females. Trends in Ecology & Evolution, 17, 313–320. , & (
- 2009) Opening Pandora’s box: comparative studies of genetic mating systems reveal reproductive complexity. Molecular Ecology, 18, 1307–1309. (