Centre for Evolutionary Biology, School of Animal Biology (M092), University of Western Australia, Crawley, Western Australia, Australia
Correspondence: Leigh W. Simmons, Centre for Evolutionary Biology, School of Animal Biology (M092), University of Western Australia, Crawley 6009, Western Australia, Australia. Tel.:+61 8 6488 2221; fax: +61 8 6488 1029; e-mail: firstname.lastname@example.org
Recent evidence shows that females exert a post-copulatory fertilization bias in favour of unrelated males to avoid the genetic incompatibilities derived from inbreeding. One of the mechanisms suggested for fertilization biases in insects is female control over transport of sperm to the sperm-storage organs. We investigated post-copulatory inbreeding-avoidance mechanisms in females of the cricket Teleogryllus oceanicus. We assessed the relative contribution of related and unrelated males to the sperm stores of double-mated females. To demonstrate unequivocally that biased sperm storage results from female control rather than cryptic male choice, we manipulated the relatedness of mated males and of males performing post-copulatory mate guarding. Our results show that when guarded by a related male, females store less sperm from their actual mate, irrespective of the relatedness of the mating male. Our data support the notion that inhibition of sperm storage by female crickets can act as a form of cryptic female choice to avoid the severe negative effects of inbreeding.
Genetic benefit models for the evolution of polyandry predict that females gain fitness benefits from choosing males with ‘good’ or ‘compatible’ genotypes through post-copulatory mechanisms of selection (Zeh & Zeh, 1996, 1997; Yasui, 1997; Jennions & Petrie, 2000; Simmons, 2005). Cryptic post-copulatory processes may strongly affect fertilization success and patterns of paternity (Birkhead & Møller, 1998; Simmons, 2001; Birkhead & Pizzari, 2002). Hence, females that exert control over paternity may acquire genetic benefits that enhance offspring viability through better hatching rate, increased survival or improved reproductive success of offspring (Jennions & Petrie, 2000; Simmons, 2005). The proximate mechanisms by which fertilization biases are achieved, however, remain poorly understood (Birkhead & Møller, 1998; Simmons, 2001). Females may bias paternity through discard, selective use or control over transfer and storage of sperm (Eberhard, 1996; Pizzari & Birkhead, 2000; Ward, 2007; Bretman et al., 2009).
Recent studies suggest that post-copulatory mechanisms play an important role in inbreeding avoidance by females. Females of several species are known to exhibit biases in paternity towards unrelated mates when inseminated by both sibling and nonsibling males as shown for the sand lizard (Lacerta agilis, Olsson et al., 1996), two cricket species (Gryllus bimaculatus, Tregenza & Wedell, 2002; Bretman et al., 2004; Teleogryllus oceanicus, Simmons et al., 2006), the house mouse (Mus musculus, Firman & Simmons, 2008), the fruit fly (Drosophila melanogaster, Mack et al., 2002 but see Ala-Honkola et al., 2011), a marsupial (Antechinus agilis, Kraaijeveld-Smit et al., 2002), the alpine newt (Triturus alpestris, Garner & Schmidt, 2003) and a dioecious plant (Silene latifolia, Teixeira et al., 2009).
Fertilization biases may reduce the risk of reproducing with genetically incompatible males, and limit the negative fitness effects of inbreeding (Charlesworth & Charlesworth, 1987; Zeh & Zeh, 1996, 1997; Tregenza & Wedell, 2000). Hence, post-copulatory inbreeding-avoidance mechanisms should evolve when the risk of mating among relatives is high, inbreeding depression is severe, and cues to discriminate relatedness are absent or insufficient to avoid mating (Parker, 1979; Kokko & Ots, 2006). This may explain why inbreeding avoidance is not ubiquitously found, and why several studies have failed to reveal fertilization biases towards unrelated males as for example for the black field cricket (Teleogryllus commodus, Jennions et al., 2004), the mallard duck (Anas platyrhynchos, Denk et al., 2005) and the most recent finding for fruitflies (Ala-Honkola et al., 2011).
Although many of the studies discussed above provide support for post-copulatory inbreeding avoidance, only few provide evidence for the mechanisms responsible for the observed fertilization biases. In the guppy, ovarian fluid affects sperm swimming speed so that sperm of sibling males has lower mobility than sperm of nonsibling males (Gasparini & Pilastro, 2011). In other cases, females appear to regulate the amount of sperm retained when inseminated by a sibling: less sperm reaches the perivitelline layer of the eggs in the fowl (Pizzari et al., 2004), 50% less sperm is found in the oviducts and ovarian cavities of the least killifish (Ala-Honkola et al., 2010), and fewer sperm are stored in the female sperm-storage organ (spermatheca) in an orb-web spider (Welke & Schneider, 2009).
Studies that quantify sperm transfer and sperm storage in crickets suggest that females actively promote or inhibit the transport of sperm into the spermatheca, and that the control of sperm storage may represent one of the major mechanisms through which post-copulatory selection operates (Bretman et al., 2009; Hall et al., 2010). For example, females of the field cricket G. bimaculatus appear to avoid inbreeding by controlling sperm storage; sperm from unrelated males have a higher representation in the female's storage organ compared with those from sibling males, and unrelated males achieve a correspondingly higher paternity share of the offspring (Bretman et al., 2009). However, there is evidence to suggest that male insects may also exercise cryptic choice, ejaculating fewer sperm when mating with sibling females (Lewis & Wedell, 2009). Cryptic male choice could thus account for some of the bias we see in the storage of unrelated male sperm. We still know very little about the proximate mechanisms employed by females in the assessment of mate relatedness, or the physiological mechanisms that underlie preferential storage of the sperm of unrelated males.
Here, we investigate sperm storage in response to inbreeding in the Australian field cricket T. oceanicus. Females from natural populations mate polyandrously and store sperm from multiple males (Simmons & Beveridge, 2010). When mated sequentially first to an unrelated and then to a sibling male, paternity is biased towards the unrelated male, suggesting the use of post-copulatory mechanisms by females to avoid producing inbred offspring (Simmons et al., 2006). However, the precise mechanisms through which paternity biases are achieved in T. oceanicus remain unclear. In crickets, mating and insemination are temporally separated events. During a relatively brief copulation, the male inserts a spermatophore into the female's reproductive tract, and sperm drain from the spermatophore into her spermathecal duct over a period of approximately 40 min (Simmons et al., 2003). While sperm are being transferred from the spermatophore to the female's reproductive tract, the mated male remains with and guards the female (Simmons, 1991; Bussiére et al., 2006). Females often attempt to leave unattractive males and remove their spermatophores before sperm have been transferred, while guarding males try to prevent females from doing so (Simmons, 1986). Thus, guarding can represent a period of sexual conflict over insemination, where males attempt to subvert female mate-choice decisions. However, mate guarding can sometimes be beneficial to females, for example when access to a male's burrow protects them from predation (Rodríguez-Muñoz et al., 2011). Under such circumstances, females may adopt more covert means of controlling sperm transfer than spermatophore removal.
Here, we tested the idea that mate guarding provides females with the opportunity to assess potential sires and to covertly choose among them via differential storage of sperm. Thus, we ask whether females store sperm with regard to the relatedness of their mates, and whether discrimination of male relatedness occurs at mating and/or during the post-copulatory mate guarding period. As we know that paternity is biased towards unrelated males when they are first to mate (Simmons et al., 2006), all females were mated first to a nonsibling male. Subsequently females were mated to either a sibling or a nonsibling male. Following their second mating, females mated to a nonsibling male were guarded either by a sibling or a nonsibling, and vice versa. We adopted a competitive microsatellite polymerase chain reaction (PCR) technique based on genotyping microsatellite markers (Bretman et al., 2009; Bussiére et al., 2010) to assess the relative contribution of nonsibling and sibling males to the female's sperm-storage organ. If females control sperm stores by discrimination at the time of mating or if males exhibit cryptic choice based on female relatedness, we expect unrelated mates to contribute more sperm to the sperm stores. However, if bias in sperm storage is a strictly female process, driven by phenotypic assessment of relatedness during the guarding period, females should store less sperm from the second male to mate when the guard is a sibling.
Materials and methods
Collecting, rearing, and creating family groups
Crickets were collected from a natural population of T. oceanicus found in Carnarvon in North-Western Australia in December 2009. Adult females were brought to the University of Western Australia where they were housed in individual boxes (7 × 7 × 5 cm) in a constant-temperature (25 °C) room with a 12 : 12-hour light : dark cycle. Crickets were supplied with water and cat chow ad libitum, and a pad of moist cotton in which to oviposit. Half-sib families from these field-caught females were reared to adulthood in groups of approximately 25 individuals in 5-L plastic boxes, and the sexes separated at the final instar. To create full-sibling families, an adult female from one family was mated to an adult male from an unrelated family. The offspring from these crosses were raised in full-sibling family groups of approximately 25 individuals in 5-L plastic containers. A large outbred laboratory population of T. oceanicus is maintained at the University of Western Australia from which unrelated crickets could be drawn. Offspring of full-sibling families and from the outbred stock were used for double-mating experiments in October and November 2010.
Males and females were separated at their penultimate nymphal instar and checked daily. Upon their adult moult, individuals were housed individually in boxes (7 × 7 × 5 cm) for 10 days to ensure sexual maturity. To confirm male ability to transfer successfully the spermatophore during mating trials, all males were mated once to a nonexperimental female on the day of the experiment.
During the first mating trial, each experimental virgin female was mated to a nonsibling male (NS). Once individuals were placed together, they were observed closely, and the time to spermatophore attachment noted. In this species, sperm transfer proceeds over a period of 40 min (Simmons et al., 2003). We observed the pair for 40 min after mating, during which a male's post-copulatory guarding behaviour generally prevents the female from dislodging the spermatophore (Simmons et al., 2003). After 40 min, we removed the drained spermatophore from the female with forceps.
Once mated, the first male was replaced with a second male. For their second mating, each female was randomly assigned to one of the following treatments (Fig. 1): (a) females were mated with/and guarded by a nonsibling male (NS/NS); (b) females were mated with a nonsibling and guarded by a sibling male (NS/S); (c) females were mated with a sibling and guarded by a nonsibling male (S/NS); (d) females were mated with/and guarded by a sibling male (S/S).
Matings were observed as described above with the only difference being that after spermatophore attachment, mated males were removed from the female for approximately 1 min. In NS/NS and S/S treatments (where the guarder was also the mating partner), the male that was removed was reintroduced and the mating trial resumed. In NS/S and S/NS treatments (where the guard was a male other than the mating partner) a different male, a sibling or nonsibling, respectively, was introduced to the female. Substitute male guarders had mated to a nonexperimental female immediately prior to their introduction to the experimental female to induce post-copulatory guarding behaviour towards experimental females. Therefore, all males included in our study guarded as if they had mated with the experimental female, positioning themselves close to the female, remaining in physical contact by antennation, or assuming the typical head-butting position (Simmons, 1991; Bussiére et al., 2006). All mating trials were conducted in clear plastic boxes (7 × 7 × 5 cm) that were placed upside down to facilitate movements of individuals inside and outside the box without having to handle crickets directly. Males were frozen at −20 °C immediately after the mating trial. Females were left in their box, provided with cat chow and a Petri dish containing moist cotton wool, and left for 24 h before being frozen.
Extracting DNA and genotyping
To estimate the contribution of sperm to the storage organ of the second male to mate (S2), we genotyped the sperm stored in the female's spermatheca. DNA was extracted from the hind legs of crickets using a rapid salt-extraction method where the tissue was placed in 350 μL of DNA extraction buffer (50 mM Tris-HCl pH 8.0, 100 mM EDTA, 100 mM NaCl, 1% SDS) and 5 μL proteinase K (10.8 mg mL−1). Samples were homogenized, incubated at 56 °C for 30 min and then cooled. After addition of 150 μL 5 M NaCl, samples were vortexed and debris pelleted by centrifugation for 10 min at 18 000 g (at 24 °C). The supernatant was transferred to a clean Eppendorf tube and 500 μL isopropanol added and mixed by inversion. Samples were placed at −20 °C for 30 min. The DNA was then pelleted by centrifugation at 4 °C for 10 min at 13 000 r.p.m. The isopropanol was removed and the pellet washed with 70% ethanol, centrifuged for 5 min at 13 000 r.p.m. (at 4 °C) and dried for 1 h at room temperature (RT). The pellet was resuspended in 50 μL TLE buffer and vortexed. The DNA was then treated with 1 μL of Riboshredder (Epicentre) to remove RNA by incubating at 37 °C for 30 min. Samples were cooled at RT and the DNA concentration was determined using a Nanodrop 1000 (Thermofisher Scientific, Malaga, WA, Australia). DNA was diluted to 30 ng μL−1. The spermatheca was dissected from each female and preserved in 100% ethanol. It was subsequently rehydrated and DNA was extracted from the contents using the QIAamp DNA micro kit (Qiagen, Chadstone Centre, Vic., Australia).
Muscle and spermathecal samples were initially screened using four microsatellite markers (lociTotri9a, 55a,57, 78) that were developed specifically for T. oceanicus (Beveridge & Simmons, 2005). The four loci were combined into one multiplexed PCR each 10 μL reaction containing 1 × PCR buffer (10 mM Tris–HCl pH 8.3, 50 mM KCl) (Invitrogen), 1.5 mM MgCl2, (Invitrogen), 200 μM of each dNTP (Invitrogen), 250 nM of each forward primer, 250 nM of each reverse primer, 0.5 units of Platinum Taq polymerase (Invitrogen) and 1–10 ng DNA. PCR amplification was performed with cycling conditions as follows: 94 °C for 3 min, then 30 cycles of 94 °C for 30 s, 55 °C for 30 s and 72 °C for 1 min, and finally 72 °C for 30 min. Products from each PCR were analysed using an ABI3730 Sequencer, sized using Genescan-500 LIZ internal size standard and genotyped using the software GENEMAPPER3.7 (Applied Biosystems, Mulgrave, Vic., Australia).
We adopted a competitive microsatellite PCR technique, which requires the identification of one unique allele at one locus and the establishment of standard curves derived from mixtures of each pair of males (Bretman et al., 2009). For those samples where a unique allele could not be identified, further genotyping was carried out using additional multiplex PCRs as follows: Multiplex 2 (ToC3-10, ToC3-86, ToC3-94), Multiplex 3 (ToC3-8, ToC3-30, ToC3-93) and Multiplex 4 (Totri54, Totri59). PCR conditions were as described above except for the MgCl2 concentration for multiplex 4, which was 3 mM, and the annealing temperature for multiplex 2 and 3, which was 60 °C (Simmons & Beveridge, 2010). Standard curves were made by mixing DNA extracted from hind tibia muscle of each pair of males so that the second male to mate accounted for 6.25%, 12.5%, 25%, 50%, 75%, 87.5% and 93.7% of the mix.
We identified one locus at which both males had one unique allele (an allele not shared by the other male or by the female). To estimate S2, we compared relative peak areas of the unique alleles calculated from GENEMAPPER's sequenced data to standard curves derived from mixing DNA of each pair of males. The relative peak area of the second male's allele (2nd male allele peak area/1st + 2nd male allele peak area) from sperm genotypes was used to solve the regression equation obtained using standard curves as predicted proportions (R2 range 0.803–0.999).
The S2 data were not normally distributed, and could not be normalized by transformation. We therefore used permutation methods for statistical analysis, using the lmPerm package in R version 2.15.2 (www.R-project.org). The effect of the relatedness of the second male to mate and to guard on S2 was analysed using a two-way permanova of the mate treatment, guard treatment and the interaction between the two treatments. Results are given as average proportion ± SE.
A total of 81 double-mated females were genotyped. DNA was extracted from the spermathecal content of 78 females, due to one case of failed spermatheca dissection and two cases in which DNA from sperm did not amplify. When mismatching between parents and sperm genotypes occurred (N = 7), females were excluded from the analysis. Therefore, we were able to estimate the relative contribution of our focal second males to sperm in the spermatheca (S2) of 71 (16 NS/NS, 14 NS/S, 21 S/NS and 20 S/S) females.
S2 was significantly affected by the relatedness of the male that guarded a female after spermatophore transfer (F1,68 = 3.97, P = 0.029), but was not affected by the relatedness of the mating male (F1,68 = 0.01, P = 0.594) (Fig. 2). The interaction between relatedness of the mated male and the guarding male was not significant (F1,67 = 0.001, P = 0.941) and was removed from the final model. The second male to mate failed to obtain representation in the sperm store (S2 = 0) in 16.9% of the trials (2 NS/NS, 3 NS/S, 3 S/NS and 4 S/S treatment), whereas first males to mate rarely failed to obtain representation (S2 = 1) (1 S/S and 1 NS/S and 1 NS/NS treatment). Reanalysis of the data excluding cases in which the second male failed to obtain representation did not change the outcome of our statistical analyses (effect of guard: F1,56 = 3.92, P = 0.032; effect of guard: F1,56 = 0.16, P = 0.804). The design of our experimental manipulation meant that in some cases, the female was guarded by the male she mated with (NS/NS and S/S) and in others she was guarded by a novel male (NS/S and S/NS). We therefore reanalysed our data adding to the model a factor that categorized guards as being novel or familiar. The main effect of guard relatedness remained significant (F1,68 = 3.90, P = 0.037), and there was no significant effect of guard familiarity on S2 (F1,68 = 0.001, P = 0.863).
We examined the effect of relatedness of mates or post-copulatory guards on female post-copulatory sperm storage in the cricket T. oceanicus. When females were guarded by nonsibling males, the second male to mate (either sibling or nonsibling) had a greater representation of sperm in the spermatheca, suggesting that the relatedness of the guard rather than the relatedness of the mate influenced whether a female stored sperm as they were delivered from the spermatophore.
During post-copulatory mate guarding, males remain in close physical contact with females by mutual antennation, jerking the body back and forward or head-butting (Bussiére et al., 2006; Simmons, 1911). During these interactions, females will be in a position to obtain chemoreceptive information about the genetic relatedness of the guard. The contact pheromonal system of cuticular hydrocarbons has proved to play an important role in mate recognition in many insects, including crickets (Tregenza & Wedell, 1997; Mullen et al., 2007; Thomas & Simmons, 2009). These cuticle compounds reflect genetic relatedness due to their association with underlying genetic variation (e.g. Tsutsui et al., 2003; Dronnet et al., 2006; Van Zweden et al., 2010). In T. oceanicus, quantitative genetic analysis of cuticular hydrocarbon profiles indicate that males have sufficient phenotypic and genetic variation in this trait for females to distinguish related from unrelated individuals (Thomas & Simmons, 2008). Moreover, cuticular hydrocarbons are used during mate choice to distinguish the genetic similarity of potential mating partners, and to preferentially mate with individuals possessing more dissimilar profiles (Thomas & Simmons, 2011). Use of pheromonal cues to distinguish among related partners is also known for the cricket G. bimaculatus, where females exert precopulatory inbreeding avoidance by rejecting matings with related males, and post-copulatory avoidance by removing the spermatophores of related males before sperm has migrated to the female's reproductive tract (Simmons, 1990, 1991).
Our experimental design allows us to conclude unequivocally that the biased storage of sperm from unrelated males is female controlled. Sperm are not transferred directly into the spermatheca by the male, but rather into the spermathecal duct. Migration of sperm into the spermatheca is likely to be regulated through movements of the muscular layer surrounding the entrance of the spermathecal duct to the spermatheca, which may either act as a locking device preventing sperm from entering, or as a support for sperm transport (Essler et al., 1992). Failed sperm transfer by males could account for a lack of representation of sperm in the spermatheca. However, insemination failure should occur with equal frequencies among our mating treatments and cannot explain the associations we have observed with guard, but not mate relatedness. Importantly, exclusion of cases in which an experimental male obtained zero representation in sperm storage did not affect our conclusions. Biased representation of sperm in the sperm stores found in previous studies (Bretman et al., 2009; Bussiére et al., 2010; Hall et al., 2010) could be a manifestation of male cryptic choice. For example, attractive males might have larger ejaculates (Perry & Rowe, 2010), or males may allocate fewer sperm to related females (Lewis & Wedell, 2009). However, male choice cannot explain our findings. Critically, it was the relatedness of the guarding male, not the mating male that influenced the proportion of sperm found in storage. Thus, T. oceanicus females are able to control the storage of sperm as a form of cryptic female choice, based on cues obtained during the mate guarding period. We suggest that the findings from related cricket species, where females store more sperm from unrelated or otherwise more attractive males (Bretman et al., 2009; Bussiére et al., 2010; Hall et al., 2010), are likely to reflect similar processes of female-controlled selective sperm storage.
Biased storage of sperm from unrelated males is expected to generate biased paternity when females fertilize their eggs. Indeed, previous studies of T. oceanicus have demonstrated paternity biases towards unrelated males when they are the first to mate (Simmons et al., 2006). This paternity advantage is congruent with the greater proportion of sperm from first mating nonsibling males found in this study. In the cricket G. bimaculatus, the proportion of sperm stored in the spermatheca that come from unrelated males has been found to predict very strongly the proportion of offspring they sire (Bretman et al., 2009). Previously, Simmons et al. (2003) reported no detectable influence of sperm numbers on competitive fertilization success in T. oceanicus, based on the observation that spermatophore attachment duration was not a significant predictor of paternity. However, we now know that spermatophore attachment and transfer of sperm into the female's spermathecal duct does not guarantee that those sperm will be transported by the female into her storage organ and used for fertilization. In the congeneric T. commodus, the effect of male attractiveness on sperm-storage success was of equal magnitude to the effect of spermatophore attachment duration (Hall et al., 2010), indicating generally that the duration of spermatophore attachment cannot be used as an accurate predictor of paternity.
Mating with close relatives can be costly if it generates inbreeding depression (Keller & Waller, 2002). Indeed, laboratory studies of inbreeding depression in crickets have reported negative effects of sibling mating on a number of important life-history traits (Roff, 1998; Drayton et al., 2010, 2011). In T. oceanicus, inbred sons have been shown to suffer from a 27% reduction in competitive fertilization success, and inbred daughters from a 27% reduction in lifetime fecundity, which will ultimately lead to reduced fitness (Simmons, 2011). Thus, the evolution of inbreeding-avoidance mechanisms makes adaptive sense. However, it is not at all clear how frequently individuals will be faced with such mating decisions in natural populations. A recent study of a natural population of the cricket G. campestris showed that although close relatives did encounter each other, mate and produce offspring, the risk of inbreeding was remarkably low (Bretman et al., 2011). These data call for greater effort in exploring patterns of relatedness and mate choice in natural populations.
In conclusion, we have shown that T. oceanicus females assess mate relatedness during the post-copulatory mate-guarding period, and are more likely to store sperm from unrelated males. Hence, mechanisms of selection that occur at a physiological level in the female reproductive tract can act to reinforce mate-choice decisions achieved by mechanisms of precopulatory mate assessment (Hall et al., 2010; Thomas & Simmons, 2011). Kin-related variation in cuticular hydrocarbon profiles are the most plausible cues utilized during this mate assessment (Thomas & Simmons, 2008, 2011), which itself is likely to have evolved as a form of cryptic female choice to avoid the severe negative effects of inbreeding in this species (Simmons, 2011).
We thank Trine Bilde for comments and discussion on earlier drafts of the manuscript. This research was supported by the Australian Research Council via a Discovery-Project to LWS, and AGSoS Mobility Grant to CT.