Sperm blocking is not a male adaptation to sperm competition in a parasitoid wasp

Sperm blocking is not a male adaptation to sperm competition in a parasitoid wasp Abstract The extent to which sperm or ejaculate-derived products from different males interact during sperm competition—from kamikaze sperm to sperm incapacitation—remains controversial. Repeated matings in the parasitoid wasp Nasonia vitripennis lead to a short-term reduction of efficient sperm use by females, which is crucial for a haplodiploid organism when needing to allocate sex adaptively (i.e. by fertilizing eggs to produce daughters). Repeated matings by females in this species, therefore, constrain sex allocation through this “sperm-blocking” effect, eliciting a cost to polyandry. Here, we explore the causes and consequences of sperm blocking and test the hypothesis that it is an ejaculate-related trait associated with sperm competition. First, we show that sperm blocking, which leads to an overproduction of sons, is not correlated with success in either offensive or defensive roles in sperm competition. Then, we show that the extent of sperm blocking is not affected by self–self or kin–kin ejaculate interactions when compared to self vs nonself or kin versus nonkin sperm competition. Our results suggest that sperm blocking is not a sperm competition adaptation, but is instead associated with the mechanics of processing sperm in this species, which are likely shaped by selection on female reproductive morphology for adaptive sex allocation. INTRODUCTION Polyandry, or the multiple mating of females with different males, is now known to be widespread in nature (Pizzari and Wedell 2013; Taylor et al. 2014). Polyandry has also been clearly demonstrated to confer significant benefits in a wide range of taxa (Arnqvist and Nilsson 2000; Slatyer et al. 2012; Taylor et al. 2014). However, mating can also be costly for females (Daly 1978; Boulton and Shuker 2013), leading to sexual conflict between males and females over the mating rate, even if some degree of multiple mating for females is adaptive (Arnqvist and Rowe 2005). In the parasitoid wasp Nasonia vitripennis, females suffer a novel cost from mating multiply, as repeated matings constrain their ability to allocate sex optimally (Boulton and Shuker 2015a). This effect arises from the fact that multiple ejaculates in a female limit the ability of that female to mobilize and use sperm to fertilize eggs, an effect that we have labeled “sperm blocking” (Boulton and Shuker 2015a). Although N. vitripennis are mostly monandrous in the wild, polyandry evolves under laboratory culture conditions (van den Assem and Jachmann 1999; Burton-Chellew et al. 2007) and can provide a fecundity benefit (Boulton and Shuker 2015b). As with all Hymenoptera, N. vitripennis is haplodiploid, and so mated females can choose whether or not to lay a fertilized egg (which develops into a daughter) or an unfertilized egg (which develops into a son). Females allocate sex in line with the predictions of local mate competition (LMC) theory (Hamilton 1967). Under high LMC conditions typically experienced by N. vitripennis in the field, females maximize their fitness by producing highly female-biased broods (Werren 1980, 1983; Shuker and West 2004; Shuker et al. 2005; Burton-Chellew et al. 2008). After a second mating, however, females are temporarily constrained in their ability to produce daughters (for at least 24 h) resulting in a fitness cost for polyandrous mothers (Boulton and Shuker 2015a). Often the costs of mating to females are thought of as arising due to sexual conflict. Males can benefit from increasing the costs of mating to females if this discourages female remating and so reduces the risk of encountering sperm competition (Simmons 2001). However, the sex allocation cost that we see in N. vitripennis does not appear to benefit males. On the contrary, at first glance, it appears to reduce male fitness as females lay fewer fertilized eggs (daughters) and males only pass on their genes through these daughters, in haplodiploids. The potential fitness costs of sperm blocking to both males and females does mean that it represents something of a paradox as to its origin and function. Here, we will consider 2 possible scenarios. First, sperm blocking may be beneficial to males as an adaptation to sperm competition (which is expected to be higher in laboratory strains demonstrating elevated polyandry; van den Assem and Jachmann 1999; Burton-Chellew et al. 2007), with sperm blocking a side-effect of increased paternity share when ejaculates compete. In other words, although sperm blocking reduces the number of daughters produced, sperm blocking may still be beneficial for a male if it increases his share of the remaining (female) offspring. Although it is clear that adaptations to sperm competition, such as physical displacement and increased swimming velocity, can and do occur (Manier et al. 2013), the extent to which sperm or ejaculate-related products directly, i.e. physically, interact has been subject to controversy. Despite hypotheses of kamikaze sperm and ejaculate–ejaculate phenomena such as sperm incapacitation, how common many of these phenotypes are is unclear and remains contentious (Baker and Bellis 1988; Harcourt 1991; Price et al. 1999; Moore et al. 1999; Snook and Hosken 2004; den Boer et al. 2010; Manier et al. 2010). Furthermore, it may be that patterns of sperm precedence that are commonly interpreted as male adaptations actually reflect female physiological characteristics that have been shaped by natural selection to optimize sperm use and storage (Simmons et al. 1999; Hosken and Stockley 2004; Herberstein et al. 2011). This is represented in our second scenario, whereby disrupted sex allocation is nonadaptive and occurs as a result of female morphological and physiological constraints on sperm processing, storage, and usage. The sperm competition scenario generates testable predictions. Sperm blocking may enhance the fertilization success of one male over the other in a number of ways. First, it may reflect a defensive adaptation that reduces the risk of encountering sperm competition for the first male and thus increases his fertilization success (Figure 1A; Simmons 2001). For instance, in some insect species, including some hymenopterans, males transfer mating plugs that obstruct the entry of rival ejaculates into the female reproductive tract (Baer et al. 2001; Mikheyev 2003; Simmons 2001, 2014). Although there is no evidence for a physical mating plug in Nasonia, a form of chemical mating plug or sperm–sperm interaction that obstructs sperm movement after copulation (as seen in some species of crab; Bawab and El-Sherief 1989) may be present. If a defensive adaptation against sperm competition such as this occurs in N. vitripennis, then it may not only obstruct incoming sperm but may also impede the movement of outgoing sperm to be used for fertilization, resulting in sperm blocking (female parasitoid wasps possess only a single spermathecal duct; King 1961). In this scenario, we predict that the severity of the overproduction of sons (i.e. sperm blocking) will be positively correlated with first male fertilization success and negatively correlated with second male fertilization success (Figure 1A). Figure 1 View largeDownload slide Schematic representing predicted results if sperm blocking is (A) a defensive male trait or (B) an offensive male trait. The x-axis represents sperm blocking which is the total number of sons in the first clutch laid. The y-axis is the total number (over all clutches) of eggs fertilized (daughters sired) by the first male (P1 solid line) or the second male (P2 dashed line). We estimated P1 and P2 in experiment 2 using the sterile male technique (see main text for details). Figure 1 View largeDownload slide Schematic representing predicted results if sperm blocking is (A) a defensive male trait or (B) an offensive male trait. The x-axis represents sperm blocking which is the total number of sons in the first clutch laid. The y-axis is the total number (over all clutches) of eggs fertilized (daughters sired) by the first male (P1 solid line) or the second male (P2 dashed line). We estimated P1 and P2 in experiment 2 using the sterile male technique (see main text for details). In contrast, it may be that sperm blocking occurs as a result of an offensive adaptation to sperm competition, increasing the success of a competing male when sperm competition is encountered (Simmons 2001). In this case, we envisage a scenario where the second male’s ejaculate blocks the female spermathecal duct, preventing the first males ejaculate from being used to fertilize eggs and resulting in sperm blocking. Our prediction for this situation is that the degree of sperm blocking will be positively correlated with second male fertilization success and negatively correlated with first male fertilization success (Figure 1B). As a further test of sperm blocking as a male adaptation, if there is some chemically mediated process at work, we might expect that sperm blocking should not occur—or should at least be considerably reduced—when ejaculates from the same male, or from closely related males, come into competition. For instance, seminal fluid proteins can play a role in the incapacitation of rival unrelated sperm. In several polyandrous species of social hymenopteran, males exhibit offensive seminal fluid traits that target the sperm of their rivals (den Boer et al. 2010). This may even be facilitated by haplodiploidy, because individual sperm produced by a single haploid male will, except in the event of mutation, be identical (de la Filia et al. 2015). We, therefore, predict that sperm blocking will be more severe when unrelated ejaculates come into competition. The second possibility is that sperm blocking is not a result of male adaptations to sperm competition, but instead reflects female physiological processes that constrain efficient sperm use. For instance, Price et al (1999) suggested that in Drosophila melanogaster seminal fluid proteins serve to incapacitate previously stored sperm of rival males. However, Snook and Hosken (2004) found evidence to suggest that this apparent male incapacitation actually occurs because females “dump” the sperm from previous ejaculates, which can result in skewed sperm precedence toward either male (see also Manier et al. 2013). Similarly, in N. vitripennis, sperm blocking could occur as a result of sperm dynamics and movement within the female reproductive tract if, for instance, females require time to move sperm from the site of storage to the site of fertilization. Although we do not explicitly test this hypothesis here, we suggest that sperm blocking most probably occurs due to constraints on sperm processing by females, if there is no effect of mating order or male relatedness on the sex ratio. We present the results of a series of experiments to contrast these 2 scenarios. In the first 2 experiments, we evaluate and then use the sterile male technique (Parker 1970; Ramadan and Wong 1989; Siva-Jothy and Tsubaki 1994) to test whether sperm blocking is an offensive or defensive trait that is associated with increased sperm precedence for the first or second male to mate. In the next 2 experiments, we test whether sperm blocking is influenced by the relatedness of a female’s mating partners. Specifically, we ask whether sperm blocking is ameliorated if a female mates twice to the same male (versus 2 different males) or with 2 brothers (vs. 2 unrelated males). METHODS Study species Nasonia vitripennis is a gregarious chalcidoid parasitoid wasp. It attacks dipteran pupae and is an ectoparasitoid, depositing eggs on the surface of the developing larva within the puparium (Whiting 1967). As with all Hymenoptera, N. vitripennis is haplodiploid and females facultatively lay either fertilized eggs that develop as daughters or unfertilized eggs that develop as sons. Females lay multiple eggs on a single host (i.e. they are gregarious) and sibmating at the emergence site is the norm (Boulton et al. 2014). After mating, the winged females disperse to find hosts on which to oviposit (brachypterous males are largely confined to the natal patch; Boulton et al. 2014). The local mating patches that result from this mating system lead to local mate competition (LMC; Hamilton 1967), and this in turn leads to selection for female-biased offspring sex ratios. In the wild, females typically oviposit alone (Grillenberger et al. 2008) resulting in high LMC because males will mate exclusively with their sisters. In this situation, females will maximize their fitness by producing only enough sons to inseminate their daughters. As more females contribute offspring to a mating patch, males can mate with nonsibling females and this increases their reproductive value and so less female-biased sex ratios are favored. Female Nasonia show an impressive ability to allocate sex facultatively in line with the predictions of the LMC theory (Werren 1980; 1983; Shuker and West 2004; Shuker et al. 2005; Burton-Chellew et al. 2008; Martel et al. 2016). The strain of N. vitripennis used for these experiments was the outbred strain HVRx, which was collected from 5 sites in the Netherlands in 2001 (van de Zande et al. 2014), reared on Calliphora vicina pupae as hosts. We reared wasps at 25 °C on a 16:8 h light:dark cycle, which results in a 14-day egg-to-adult cycle. To standardize the rearing environment of focal males and females to be used in the following experiments, we provided virgin and mated females with 3 C. vicina hosts each for 48 h, these females served as grandmothers for the focal individuals used for experiments. We isolated focal females as pupae (based on the presence of wing buds and a visible ovipositor) from the hosts 2 days prior to expected adult eclosion, 12 days after the grandmothers had been provided with hosts to oviposit on. The focal males were sons produced by virgin grandmothers, and these males were left to eclose naturally. First, we will describe our pilot experiment to calculate the lowest effective sterilizing dose for our sterile-male sperm competition experiment. Pilot: determining the lowest effective sterilizing dose We sterilized N. vitripennis males using gamma radiation in order to investigate patterns of sperm precedence. Gamma radiation has been shown to sterilize males of many species effectively without detrimentally affecting sperm morphology (Wishart and Dick 1985). If the dose is sufficiently low, irradiation does not affect sperm motility or other aspects of the ejaculate, but the accumulation of lethal mutations in the sperm render them infertile (Ray 1948; Wishart and Dick 1985). The main reason for using the sterile male technique in N. vitripennis is due to the lack of suitable morphological markers. Although previous studies have used red- or oyster-eyed inbred mutant strains to assay sperm precedence (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977), these traits are recessive and so sperm precedence cannot be assayed in a wild-type background. Although sterilization can result in reduced fertilizing capacity (Simmons 2001), this issue can be reduced through pilot studies that assess the Lowest Effective Sterilizing Dose (LESD), and through randomization of the mating sequence (Simmons 2001). To our knowledge, there are no published studies where gamma radiation has been used to sterilize N. vitripennis for the purposes of assaying sperm precedence (but see Ray 1948 and Saul 1955 for other uses). Therefore, we conducted a pilot experiment to ascertain the LESD by generating a dose-response curve for 4 treatment groups: 80, 100, 120, and 140 Gy of radiation (Ray 1948). The dose was determined by varying the amount of time that a vial containing male N. vitripennis was exposed to a source of gamma radiation. The dose rate was 2.59 Gy/min, so that males that received 80 Gy remained in the radiation chamber for 1854 s, 100 Gy for 2316 s, 120 Gy for 2778 s, and 140 Gy for 3240 s. On the day of emergence, focal males were exposed, in 4 groups of approximately 100 wasps, to a 137Cs source emitting gamma radiation. The source rotates so that the dose rate is constant over space. Males from each treatment (plus untreated controls) were then mated to virgin females either straight after irradiation (day 1) or 24 h later (day 2; N = 25 per treatment per day). All pairings were observed to determine whether successful copulation had occurred, after which the females were provided with hosts that were maintained at 25 °C. After the offspring had emerged and died, the number of sons (unfertilized eggs) and daughters (fertilized eggs) were counted. A mating was considered sterile when 100% of the offspring were male. The dose-response curve (Figure 2) shows the percentage of females from each treatment group that produced any daughters. Figure 2 View largeDownload slide Dose-response curve for males irradiated with 80–140 Gy of gamma radiation (137Cs) and control (untreated) males in the pilot (Error bars = binomial Confidence Intervals). Figure 2 View largeDownload slide Dose-response curve for males irradiated with 80–140 Gy of gamma radiation (137Cs) and control (untreated) males in the pilot (Error bars = binomial Confidence Intervals). Our results suggest that the minimum effective dose was 100 Gy, after which less than 5% of males successfully fertilized eggs. Irradiation did induce sterility, as matings with control males were significantly more likely to result in daughter production than matings with exposed males (Quasibinomial GLM: F4, 178 = 92.77, P < 0.0001). Sterilization was not effective below 100 Gy (Tukey P < 0.007) but fertility was no different when males were treated with 100–140 Gy (100 Gy is the LESD). There was no effect of the day on which mating occurred on fertility (F1, 181 = 0.02, P = 0.89). Finally, if the sperm of irradiated males became less viable over time (Simmons et al. 1996), then the sex ratio produced by females mated on the second day should be more male biased (since unfertilized eggs will develop as males). This was not the case, however, (day: F1,174 = 0.29, P = 0.59; interaction effect treatment*day F4,170 = 0.56, P = 0.69) suggesting that the sperm of irradiated males did not decrease in viability over the experimental period. Experiment 1: assaying sperm precedence using gamma-irradiated males Four hundred virgin females were isolated from a grand-maternal generation. Virgin males were generated from unmated grandmothers and were maintained in groups of brothers. Six stock tubes containing brothers were exposed to 100 Gy of gamma radiation. We used irradiated males from 6 families to standardize competitor identity as much as possible, ideally the same males would have been used repeatedly (García‐González and Evans 2011) but male N. vitripennis do not produce new sperm after emergence (i.e. they are prospermatogenic; Boulton et al. 2014) and using the same males repeatedly would result in mating order effects on fecundity and the sex ratio (Boulton and Shuker 2015b). In this experiment, we investigated whether sperm blocking is associated with patterns of sperm precedence. To do this, focal females and males were assigned to one of the following 7 treatments: (V) virgin female; (N) once-mated to a normal male; (I) once-mated to an irradiated male; (NN) twice-mated to two normal males; (II) twice-mated to two irradiated males; (NI) mated to a normal male followed by an irradiated male; (IN) mated to an irradiated male followed by a normal male (Total N = 348, N = 45–54 across treatments). To help explain the rationale for our interpretation of the following results, our assumptions are as follows. The production of a daughter arises from a fertilized egg, that is, an egg fertilized by the mother using a viable sperm (from a normal ‘N’ male). A male is produced from an unfertilized egg that the mother left unfertilized. The effects of irradiation are manifested through mothers fertilizing eggs with infertile sperm (from an ‘I’ male), which results in the production of inviable embryos that fail to develop. This will lead to a change in the offspring sex ratio (fewer females, so less female-biased or even male-biased sex ratios), as well as a reduction in the total number of adult offspring produced (eggs fertilized by inviable ‘I’ sperm will not develop into adults). In terms of the experiment, females assigned to the twice-mated treatments (II, IN, NI, NN) were mated first on Day 1. All copulations were observed and in order to increase the likelihood of remating, we prevented males from engaging in postcopulatory courtship as this behavior serves to reduce female receptivity to additional matings (van den Assem and Visser 1976). Twenty-four hours after their first mating (on Day 2), these females were presented with a second male. All once-mated females were presented with their first and only male on Day 2. Females were given 3 C. vicina hosts on Day 2 (i.e. all females were given hosts immediately after their final mating) and kept in an incubator at 25 °C. In this experiment, the initial 3 hosts were removed 24 h later and replaced with a single host. This was repeated every day for 3 days such that females surviving until the end of the experiment received a total of 3 hosts in 4 batches. Three hosts were presented in the first batch, to allow host-feeding opportunities as well as to provide oviposition resources. These hosts were maintained at 25 °C and emerging sons and daughters were counted after they had died. We kept experimental females in isolation after host provisioning and checked them every day at 9 am, 12 pm and 5 pm for mortality. After all females had died, we removed and measured their right hind tibia as a proxy for body size (Godfray 1994) in order to assess whether larger females (possibly with larger spermathecae; Martel et al. 2011) suffer sperm blocking to a lesser extent. An Olympus SZX10 microscope with an ocular micrometer was used for all measurements and each tibia was repositioned and remeasured 3 times in order to assess measurement error, which we found to be low based on a high repeatability estimate (Intraclass correlation coefficient: ICC = 0.94 ± 0.006 CI, P < 0.001). We tested whether treatment or host batch influenced the sex ratio (measured throughout as proportion of offspring that are male) that females produced across all treatments using a repeated-measures Quasibinomial GLMM (lme4 (Bates et al. 2015) in R Studio, RStudio, Inc., Boston, MA; R Core Team 2016). Fixed factors were treatment, host batch and their interaction. Female longevity (hours) and hind tibia length (mm) were included as covariates. Female identity nested within host batch was included as a random effect. Pairwise comparisons within these models allowed us to assess the effects of treatment on sperm blocking as well as the effectiveness of irradiation for sterilizing males. If irradiation successfully sterilizes males without impairing their ability to transfer an ejaculate, virgin females should produce significantly larger (all-male) clutches than mated (I or II) females (virgin female clutches are generally the same size as the clutches of once-mated females in N. vitripennis, only all-male clutches; Boulton and Shuker 2015a). We, thus, tested whether females that mated with either one (I) or two (II) irradiated males produced fewer sons (unfertilized eggs) than virgin females using a repeated-measures QuasiPoisson GLMM. To investigate patterns of sperm precedence and mixing, we tested whether daughter production (i.e. eggs fertilized by N males that develop successfully) differed between treatment groups NN, NI and IN over the four host batches using a repeated-measures GLMM with a QuasiPoisson error structure. For this analysis, treatment, host batch and their interaction effect were included in the model as fixed factors, female longevity and hind tibia length as covariates, and female identity nested within host batch as a random effect. The following formula was used to calculate sperm precedence (PN) in treatments NI (P1) and IN (P2) (modified from Boorman and Parker 1976):  PN= x−zp−z where, x = proportion of daughters (fertilized eggs) in NI or IN matings, z = proportion of daughters in II matings (0.018), p = proportion of daughters in N matings (0.814). We considered daughter production in II as z rather than daughter production in I because the reduced daughter production that occurs after repeated matings (sperm blocking) renders this estimate more comparable to the other twice-mated treatments (NI and IN). To calculate p we used the proportion of daughters in N matings to estimate “maximum” paternity by a single male, and NN matings were not used here as paternity could not be assigned to either male (in addition, daughter production in NN matings was reduced by sperm blocking and would underestimate maximum daughter production; see Results below). In the original formula described in Boorman and Parker (1976)x, z, and p are the proportion of eggs fertilized by the sperm of the N or I males and are calculated from the total number of viable offspring that hatch divided by the total number of eggs laid (including inviable eggs). Here, however, we have modified this formula due to constraints imposed by haplodiploidy and the life cycle of Nasonia. We consider the proportion of daughters in a clutch, excluding male offspring from our estimates of sperm precedence. This is because males develop from unfertilized eggs and so are not useful for assigning sperm use to the N male. Additionally, due to the nature of parasitism by N. vitripennis, unfertilized eggs typically cannot be counted. This is because around 40 eggs or so are laid inside the puparium and any eggs that fail to develop are generally destroyed by the larvae that do hatch. As a result, we lack knowledge regarding the number of eggs that failed to develop (those fertilized by the I male). This means that the values of P1 and P2 will be overestimates (see the Supplementary Material, Supplementary Table S1, for calculations of PN from a set of simulated data with and without knowledge of numbers of inviable eggs laid). Although these estimates are sufficient for investigating patterns of sperm precedence within this study, this limitation prevents comparisons of results between studies. Under high LMC (when females oviposit alone), the number of sons in the first host batch is typically around 10%–20% of the total brood size (Werren 1980, 1983). When sperm blocking occurs, however, the proportion of sons in the first host batch increases, as females are unable to mobilize sperm successfully to fertilize eggs. To assess whether sperm blocking in the first clutch influences fertilization success of the N male, we tested whether PN over host batches 2–4 was associated with the number of sons in the first host batch for the treatments IN and NI. We infer P1 from treatment NI and P2 from treatment IN (i.e. the order of the focal male) and predict the following: If sperm blocking is 1) a defensive trait that benefits the first male, there will be a positive correlation between P1 (use of the first male’s sperm in NI) and son production in the first clutch of eggs and a negative correlation for P2 (use of the 2nd male’s sperm in IN; Figure 1A). If, on the other hand, sperm blocking is 2) an offensive trait used by second males to increase their relative paternity, then P2 (use of the 2nd male’s sperm in treatment IN) will be positively correlated and P1 (the first males sperm use in NI) negatively correlated with son production in host batch 1 (sperm blocking; Figure 1B). In other words, if sperm blocking is an offensive sperm competition trait that benefits the second male, high sperm blocking should lead to high P2 (daughter production will be positively correlated with sperm blocking in treatment IN). If it is a defensive trait that benefits the first male, high sperm blocking will lead to higher P1 (daughter production will be positively correlated with sperm blocking in treatment NI). These predictions are outlined in Figure 1. We ran a GLMM with female ID nested within host batch as a random effect, PN was the outcome variable and the predictors were “Number of sons in host batch 1”, as well as treatment, host batch, and the interaction effects. A significant interaction effect between the number of sons and treatment suggests that the relationship between sperm blocking and fertilization success differs depending on the order in which the nonirradiated (N) male has mated (i.e. whether prediction 1 or 2 is supported). Experiment 2: sperm blocking with competition between self versus nonself sperm In our second experiment, we tested whether the presence of ejaculates from 2 different males influences the severity of sperm blocking. We also tested whether, in addition to mating, prolonged exposure to either the same male or 2 different males influenced sperm blocking. This is because previous work had suggested that harassment by males during oviposition could influence sperm use independent of mating (Wylie 1976; Boulton and Shuker 2015b, 2016). To do this, focal females were exposed to a single male, copulation was observed and postcopulatory courtship prevented as before. Females assigned to polyandrous treatments mated again 24 h later. The identity of the male that each female mated with first was recorded and all males were retained. After their first mating, females were randomly assigned to the following treatments, and in each case females were given 3 hosts to oviposit on for 24 h after the final mating, either in the presence of a male or not. The 6 treatments were: 1) Control (C1) with once-mated females given hosts (C. vicina) immediately after mating; 2) control + 24 h (C24) with once-mated females given hosts 24 h after mating; 3) same-male mating (S) with females mating with the same male twice with a 24-h interval between matings; 4) same-male harassment (SH) in which females were exposed to the same male 24 h after the initial mating and the male and female kept together during oviposition; 5) nonself mating (NS) in which females remated with a different once-mated male 24 h after the initial mating; and 6) nonself harassment (NSH) in which females were exposed to a different once-mated male 24 h after the initial mating, with the male and female kept together during oviposition (total N = 196, N = 23–37 across treatments). In order to standardize male mated status in the “same-male” and “nonself” treatments, all second males used had mated once previously. All experimental hosts were removed the following day and incubated at 25 °C. After 2 weeks, the offspring of the focal females emerged from the hosts. After the offspring had died, the number of sons and daughters produced was counted. Experiment 3: sperm blocking with competition between kin versus nonkin sperm In our final experiment, we tested whether the severity of sperm blocking is influenced by male relatedness. The basic design of Experiment 2 was repeated to test how sperm blocking differed when females mated with 2 brothers or 2 unrelated males. Females were randomly assigned to one of the following treatments: 1) Control (C1) with once-mated females given hosts immediately after mating; 2) control + 24 h (C24) with once-mated females given hosts 24 h after mating; 3) kin-mating (K) with females mating with the 2 virgin brothers with a 24-h interval between matings; 4) kin-harassment (KH) in which females were exposed to a virgin brother of their first mate 24 h after the initial mating and where the male and female were kept together during oviposition; 5) nonkin mating (NK) in which females remated with a virgin male that was unrelated to their first mate 24 h after the initial mating; and 6) nonkin harassment (NKH) in which females were exposed to a virgin male that was unrelated to their first mate 24 h after the initial mating, and the male and female were kept together during oviposition (Total N = 206, N = 23–38 across treatments). In Experiments 2 and 3, the sex ratio (proportion of sons) and son and daughter production (total clutch size) was analyzed using GLMMs with a Quasibinomial and Gaussian error structure, respectively, using the package lme4 in R Studio (Bates et al. 2015). Treatment, harassment, and the interaction effect were included as fixed factors. Experiments 2 and 3 were each conducted across 3 blocks, and so experimental block was included as a random effect. RESULTS Experiment 1: assaying sperm precedence using gamma-irradiated males Irradiation was generally effective at preventing daughter production. However, females that mated with 2 irradiated males (II) were more likely to produce some daughters compared to both virgin and I females (Quasibinomial GLMM: F2,544= 95.58, P < 0.0001; Figure 3A). The absolute number of sons produced over the 3 treatments differed significantly. Clutches laid by virgin females are typically comparable in size to mixed-sex clutches (van den Assem 1977) but here virgin females produced around 3 times as many sons as females that mated with 1 or 2 irradiated males (QuasiPoisson GLMM: F3,573= 230.72, P < 0.0001; Figure 3B). This suggests that mated females were being inseminated and using sperm but that the daughters sired by I males made up the putative remaining 60–70% of the clutch that failed to develop. Figure 3 View largeDownload slide (A) Proportion of sons (sex ratio) and (B) total son production by virgin females and females mated to either 1 (I) or 2 (II) irradiated males in experiment 1 (A) Sex ratio (proportion of sons), error bars = binomial confidence Intervals (CIs). Note that the y-axis runs between 0.96 and 1.0 and that in all cases daughter production was extremely low. (B) Absolute number of sons produced. Virgin females (V) produce more sons than females mated with 1 (I) or 2 (II) irradiated males. Error bars = 95% CIs. (A and B) Different lower case letters represent treatment groups that are statistically different (P < 0.05). Figure 3 View largeDownload slide (A) Proportion of sons (sex ratio) and (B) total son production by virgin females and females mated to either 1 (I) or 2 (II) irradiated males in experiment 1 (A) Sex ratio (proportion of sons), error bars = binomial confidence Intervals (CIs). Note that the y-axis runs between 0.96 and 1.0 and that in all cases daughter production was extremely low. (B) Absolute number of sons produced. Virgin females (V) produce more sons than females mated with 1 (I) or 2 (II) irradiated males. Error bars = 95% CIs. (A and B) Different lower case letters represent treatment groups that are statistically different (P < 0.05). Females that mated only once to an unsterilized male (N) produced the most female-biased sex ratios as expected. Twice-mated females all experienced sperm blocking, laying significantly more male-biased sex ratios than ‘N’ females (all pairwise P < 0.001; F3,118 = 25.44, P < 0.0001). The significant difference between the sex ratio produced by N and NN females confirms the problem of sperm blocking for polyandrous females; females that mated twice produced more male-biased sex ratios immediately after mating. Moreover, the sex ratios produced by N and NN females converged in later host batches which demonstrates that the effect of sperm blocking does wear off over time (interaction effect: Quasibinomial GLM: F9,461 = 4.87, P < 0.0001; main effect of treatment F3,467 = 30.65, P < 0.0001; main effect of host batch F3,467 = 9.73, P< 0.0001; See Figure 4). Figure 4 View largeDownload slide Sex ratios (proportion of sons) produced by females that mated with either 1 or 2 unsterilized (N) males in experiment 1 (error bars = binomial CIs). Figure 4 View largeDownload slide Sex ratios (proportion of sons) produced by females that mated with either 1 or 2 unsterilized (N) males in experiment 1 (error bars = binomial CIs). In terms of sperm precedence—and remembering that in haplodiploids such as Nasonia only daughters can tell us about sperm usage—the first male had higher daughter production initially (NI) but sperm mixing occurred in later clutches (Figure 5A). There was a significant effect of treatment on daughter production (F2,371= 35.44, P < 0.0001; Figure 5A), but this was because females that mated to 2 unsterilized males (NN) laid more daughters than females that mated to an unsterilized male either first (NI) or second (IN). In the first host batch, first males (NI) did achieve higher daughter production than second males (IN) but, overall across all host batches, sperm use was not significantly biased towards either male (P1 = 0.76, P2 = 0.64, albeit close to significance: treatment: F1,360= 3.60, P = 0.06). This was partly because daughter production by IN females was equal to that of NI females in host batches 3 and 4 (treatment*host batch interaction effect: F6,367 = 2.37, P = 0.03), suggesting that the second male sperms (viable sperm in treatment IN) are used more in later host batches. There was a significant main effect of host batch on sperm precedence (F3,369 = 10.44, P < 0.0001), but this did not appear to relate to the number of hosts provided per batch (three in host batch 1 and one host in batches 2–4) as the only pairwise significant difference was between host batches 1 and 2 (P < 0.05). Figure 5 View largeDownload slide Daughter production and sperm precedence for twice-mated females in Experiment 1. (A) Fewer daughters (fertilized eggs) were produced by the second male in host batches 1–2 (IN), but there was no significant difference in daughter production by the first (NI) or second (IN) male in host batches 3 and 4. Error bars = 95% CIs. (B) Son production or “sperm blocking” in the first clutch reduced fertilization success in host batches 2–4 for the first (P1: NI) and second (P2: IN) male equally. Figure 5 View largeDownload slide Daughter production and sperm precedence for twice-mated females in Experiment 1. (A) Fewer daughters (fertilized eggs) were produced by the second male in host batches 1–2 (IN), but there was no significant difference in daughter production by the first (NI) or second (IN) male in host batches 3 and 4. Error bars = 95% CIs. (B) Son production or “sperm blocking” in the first clutch reduced fertilization success in host batches 2–4 for the first (P1: NI) and second (P2: IN) male equally. In terms of sperm competition outcomes, more explicitly, sperm blocking was not associated with higher paternity for either the first or the second male. When sperm blocking occurs, more sons are produced in the first host batch. Although “sperm blocking” did reduce the fertilization success of the unsterilized (N) male in host batches 2–4 (β = −0.49, SE = 0.25; F1, 234 = 5.01, P = 0.03), there was no significant interaction between sperm blocking and treatment (F2,231 = 1.77, P = 0.18; Figure 5B) demonstrating that sperm blocking reduces the success of the first and second male equally. We found no evidence that sperm blocking (sex ratio in the first host batch) was related to female size (F1,121 = 1.14, P = 0.29) or longevity (F1,121 = 1.14, P = 0.29). We also found no evidence that female body size interacts with treatment to influence sperm blocking (Treatment*size: F3,118 = 1.29 P = 0.28). Experiment 2: self versus nonself All twice-mated females experienced sperm blocking to the same degree, regardless of whether they mated twice with the same male (S) or with different males (NS; P = 0.99; Figure 6A). Although there was an overall effect of mating treatment on the sex ratio ( F3, 313 = 42.22, P < 0.0001), the only significant pairwise differences were between both controls and twice-mated females (P < 0.0001). There was no effect of harassment (F1, 315 = 0.08, P = 0.78) or any interaction effect between harassment and treatment (F5,311 = 0.72, P = 0.40) on the sex ratio. Figure 6 View largeDownload slide Sex ratio and clutch size for females in Experiments 2 (A and B) and 3 (C and D). White bars = No harassment, gray bars = harassment. (A) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to, the same male twice or 2 different males. (B) The clutch size that females produced was not affected by mating treatment. However, the interaction effect between harassment and treatment was significant. (C) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to 2 brothers (kin) or 2 unrelated males (nonkin). (D) Clutch size was not affected by harassment, the number of matings, or relatedness between the first and second males. Error bars = binomial CIs (for A and C) and 95% CIs (for C and D). Figure 6 View largeDownload slide Sex ratio and clutch size for females in Experiments 2 (A and B) and 3 (C and D). White bars = No harassment, gray bars = harassment. (A) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to, the same male twice or 2 different males. (B) The clutch size that females produced was not affected by mating treatment. However, the interaction effect between harassment and treatment was significant. (C) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to 2 brothers (kin) or 2 unrelated males (nonkin). (D) Clutch size was not affected by harassment, the number of matings, or relatedness between the first and second males. Error bars = binomial CIs (for A and C) and 95% CIs (for C and D). In terms of clutch size (total son and daughter production), there was no significant main effect of harassment (F1, 315 = 0.01, P = 0.91) or treatment (F3, 313 = 2.00, P = 0.11). The interaction effect between treatment and harassment was, however, statistically significant (F5,311 = 5.34, P = 0.02; Figure 6B). When control treatments were removed (no harassment in C1 or C24), the significant interaction effect remained (F3,216 = 5.20, P = 0.02). Prolonged exposure (harassment) reduces offspring production when the first and second males are different individuals (Figure 6B) but clutch sizes were larger when females were exposed to the same male twice. Experiment 3: kin versus nonkin Sperm blocking was not influenced by the relatedness between competing males when females mated twice. Although there was again an overall effect of mating treatment on the sex ratio (F3, 316 =39.75, P < 0.001), the only significant pairwise differences were between the controls and all twice-mated females as before (P < 0.0001). As above, females in the once-mated control treatments laid the most female-biased sex ratios. The sex ratio was less female-biased when females mated twice but relatedness between the males with which a female mated had no effect on the sex ratio (P = 0.99; Figure 6C). There was no effect of harassment on the sex ratio (F1, 318 = 0.04, P = 0.84) nor was there a significant interaction effect between treatment and harassment on the sex ratio produced (F5,314 = 0.02, P = 0.87). Turning again to clutch size, there was no effect of harassment (F1, 318 = 0.79, P = 0.37) and no statistically significant effect of treatment (F3, 316 = 2.56, P = 0.054) on clutch size, nor was there a significant interaction effect between treatment and harassment (F5,314 = 0.05, P = 0.82; Figure 6D). DISCUSSION A diverse range of offensive sperm competition traits have been suggested, and demonstrated, to confer an advantage to males when females mate multiply and sperm competition is encountered (Simmons 2014). However, the parasitoid wasps remain relatively understudied with respect to postcopulatory sexual selection (Boulton et al. 2014). In this study, we investigated the possibility that sperm blocking, that is, the overproduction of unfertilized eggs (sons) that occurs immediately after female remating, might be advantageous to male N. vitripennis when they experience sperm competition. However, it is clear from our results that sperm blocking does not occur as a result of males displacing, blocking, or incapacitating the ejaculates of their rivals. First, in Experiment 1, sperm blocking did not favor the paternity of either the first or second male, and so was not associated with increased paternity in either offensive or defensive sperm competition. Second, in Experiments 2 and 3, sperm blocking was not ameliorated by the presence of a male’s own sperm or the sperm of a brother, and so sperm blocking occurred regardless of sperm identity and origin. In Experiment 1, we found that N. vitripennis males do not actively “block” rival sperm, as sperm blocking reduced paternity for the first and second males equally. We did find, however, that in the first clutch, the first male sired more daughters, but in subsequent clutches second males gained more paternity success, such that the first and second males shared equal paternity overall (P1 = 0.76, P2 = 0.64; note that these values overestimate the true PN and sum to more than 1, see Methods and Supplementary Table S1 in the Supplementary Material). This is not likely to be due to the first male’s ejaculate being fully depleted, because a single mating with a virgin male was sufficient for females to maintain daughter production when provided with up to 24 hosts in a previous study (Boulton and Shuker 2015b). Instead, this pattern is more likely to be due to how sperm is processed and stored. Previous work on N. vitripennis has demonstrated that patterns of sperm use are influenced by mating order and also by the mated status of the male and the timing of the second mating. Holmes (1974) found that if the first male to mate was partially sperm depleted, there was no consistent pattern of use of the first or second male’s sperm (i.e. sperm mixing occurred) and so paternity was more equal than if males were not depleted (in which case paternity was skewed towards the first male). Additionally, van den Assem and Feuth-de-Bruijn (1977) found little sperm mixing when females remated immediately after their initial mating, but sperm mixing increased when there was a delay of 3 days between matings (van den Assem and Feuth-de-Bruijn 1977; van den Assem 1977). The processes that influence the degree of sperm mixing outlined above have also been found to influence sperm blocking. If the second male does not succeed in inseminating the female (Geuverink et al. 2009), or if there is no interval between matings, and so no sperm mixing, van den Assem (1977) found that sperm blocking was less severe (i.e. the sex ratio in the first clutch was less male biased). Superficially, this suggests that sperm blocking might promote sperm mixing, but the current results argue against a causal relationship and show that sperm blocking does not relate to male ejaculate physiology. Nor does sperm blocking appear to be a constraint related to the size of the spermatheca, as we found no effect of female body size (which is associated with larger spermatheca size in another parasiotid, Trichogramma euprodctidis; Martel et al. 2011) on the severity of sperm blocking. Our second and third experiments also do not support the hypothesis that sperm blocking is a targeted male adaptation in N. vitripennis that incapacitates (rather than indiscriminately blocks) unrelated rival male sperm (den Boer et al. 2010). Whether the sperm are one’s own, or that of a brother, or an unrelated competitor, the effect on reduced sperm use by the recipient female is the same. A previous example of apparent “sperm incapacitation” in D. melanogaster (Price et al. 1999) turned out on further investigation not to reflect male ejaculate physiology, but rather female physiological processes (Snook and Hosken 2004; see also Manier et al. 2010). In D. melanogaster, it is the act of copulation itself that leads to females “dumping” sperm from previous ejaculates, resulting in extreme last-male precedence. In D. melanogaster, however, it does appear that some males are better than others at eliciting “sperm dumping”, but we found no evidence that this is the case in N. vitripennis, as increased sperm blocking was not positively associated with sperm precedence. Instead, the patterns of, and correlation between, sperm precedence and sperm blocking exhibited by N. vitripennis in the current study, and in others (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977; van den Assem 1977; Boulton and Shuker 2015b) may relate to the morphology of the spermatheca and the physiology of sperm storage and movement within the female. The single, narrow spermathecal duct of the female’s sperm store is thought to allow the movement of a single sperm to the site of fertilization, limiting wastage of sperm and facilitating precise sex allocation (Flanders 1956; Wilkes 1965). Whilst adaptive in terms of sperm economy, this narrow tube may become congested if the capacity of the spermatheca is exceeded. As such, building on our work and that of earlier authors, we suggest that the phenotype of sperm blocking is generated as follows. During insemination, sperm are deposited at the base of the single, narrow spermathecal duct, from where they rapidly migrate to the rigid spheroid spermathecal capsule. The first sperm reach the capsule after just 1 min (King 1961) but take several hours to quiesce and line up at the capsule mouth, when they are then ready to be used for fertilization (King 1962; Wilkes 1965). The sperm leave the capsule to fertilize the eggs through the same narrow duct. Following a second mating, the presence of incoming sperm in the narrow spermathecal duct prevents sperm required for fertilization from exiting the capsule to some extent (see Wilkes 1965 and Finney et al. 1947 for evidence from other parasitoids). If the first ejaculate is small, or there is sufficient time between matings, then the second ejaculate can enter the spermathecal duct and spermatheca itself, and then sperm mixing can occur, equalizing P1 and P2 (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977; van den Assem 1977). Crucially though, in these cases the second ejaculate is able to enter the spermathecal duct and the incoming ejaculate will temporarily obstruct outgoing sperm, reducing the efficiency of sperm use and generating sperm blocking. The pattern of sperm precedence that we observed in Experiment 1 fits this scenario. The first 2 clutches laid were predominantly sired by the first male to mate, but sperm mixing occurred later. The sperm in the first ejaculate would be the first to line up and quiesce at the mouth of the spermathecal duct, ready to be used for fertilization. The 24-h delay between matings should result in some of the second ejaculate entering the duct, but before the first eggs are fertilized the small capsule will be fairly full. The sperm in the second ejaculate will move into storage slowly, at the same time impeding any outgoing sperm (and resulting in sperm blocking). The space created when the first sperm leaves the capsule will allow more of the second ejaculate to enter and facilitate the mixing of initially stratified sperm, increasing the potential for sperm mixing and, as such, sperm competition over time (Simmons 2001). Taken together then, it seems that the very morphological and physiological machinery that allows coordinated control of fertilization in parasitoid species such as N. vitripennis is also responsible for the disruption of adaptive sex allocation that occurs when ejaculates overlap during sperm processing prior to fertilization. Therefore, control over fertilization, as required by Nasonia because of the facultative nature of sex allocation under LMC, should strongly select against polyandry if sperm blocking arises with repeated matings (Boulton and Shuker 2015a; Ridley 1988). As such, sperm processing may be a major factor constraining female multiple mating in the parasitoids, a group noted for its rather limited level of polyandry (Ridley 1993; Godfray 1994; Boulton et al. 2014). This constraint may be less important when selection on female-biased sex ratios is relaxed, for instance under laboratory conditions, where LMC is reduced and polyandry is freer to evolve (for N. vitripennis see: van den Assem and Jachmann 1999; Burton-Chellew et al. 2007; Boulton and Shuker 2015b). The findings of the current study show that constrained sex allocation does not occur as an evolutionary consequence of polyandry in N. vitripennis (via postcopulatory sexual selection on males), rather selection on sex allocation may limit polyandry in these wasps, and weakening that selection may be a key contributing factor to the evolution of elevated remating in mass culture conditions. When polyandry does increase, such as under mass culture, postcopulatory sexual selection should favor the evolution of large ejaculates in N. vitripennis males as a defense against sperm competitors (Simmons 2001). Larger ejaculates would then occupy more space in the spermathecal duct, thereby obstructing the ejaculates of rivals for longer (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977). However, as discussed, N. vitripennis females appear unable to process such large ejaculates efficiently, and sperm blocking leads to a reduction in fitness via sex allocation disruption. Thus, the evolutionary interests of males and females are brought into conflict over the size of ejaculate that is transferred. In haplodiploids, sexual conflict is more likely to be resolved in favor of females and so any male adaptation, such as large ejaculate size, that harms females is unlikely to persist (de la Filia et al. 2015). Asymmetric sexual conflict such as this may explain why ejaculates transferred by parasitoid males remain relatively small even in more polyandrous species, as well as why the costs of mating (in terms of fecundity and longevity) tend to be low for females (Boulton et al. 2014). The interpretation of the interaction between sperm use, sperm competition, and polyandry that we present here also suggests a different (but not mutually exclusive) hypothesis for the evolution of the postcopulatory courtship seen in Nasonia and in many other parasitoids (Boulton et al. 2014). Such postcopulatory courtship typically reduces female receptivity to subsequent matings, and so has often been interpreted as an adaptation to reduce future sperm competition for a male (i.e. a defensive trait, and a clear example that postcopulatory sexual selection can shape behavior, physiology, and morphology even without double-matings occurring: Dougherty et al. 2016). However, multiple matings not only risk male paternity share, they also reduce the extent to which sperms are used by females. Thus, postcopulatory courtship that reduces female receptivity not only protects paternity, but also protects efficient sperm use by that female. Because haplodiploid males only pass on their genes via daughters, this might be a nontrivial selective force, to sit alongside that of protecting paternity. To conclude, by interpreting our findings and those of earlier studies in the light of the structure and function of the female parasitoid reproductive tract, we have been able to suggest a mechanism for sperm blocking and to understand how sperm dynamics can result in the patterns of sperm precedence and mixing seen in N. vitripennis and, indeed, across other parasitoids more generally. Herberstein et al (2011) advocate this approach, encouraging behavioral ecologists to refer to early morphological descriptions when interpreting data regarding sperm competition and sperm dynamics. The current study demonstrates that this approach can facilitate the understanding of the processes that lead to patterns of sperm precedence, demonstrating that phenomena that appear likely to result from male ejaculate traits, such as sperm blocking (or “sperm incapacitation”; Simmons et al. 1999; Hosken and Stockley 2004) may relate to female physiological characteristics that, in this case, and perhaps many others, are shaped by natural selection on efficient female sperm use and storage. Rather than immediately looking to the more appealing and enigmatic (and often more contentious) processes, such as cryptic female choice (Birkhead 1998) or “kamikaze sperm” and incapacitating seminal fluid proteins (Baker and Bellis 1988; Harcourt 1991; Price et al. 1999; den Boer et al. 2010), it is important that we understand the basic role that female physiology plays in determining the outcomes of sperm competition. In doing so, we can gain a better appreciation of how postcopulatory sexual selection operates and the types of traits that are likely to result in males and females. SUPPLEMENTARY MATERIAL Supplementary data are available at Behavioral Ecology online. FUNDING This work was supported by a NERC Doctoral Training Grant studentship to R.A.B. We are grateful to Professor Andrew Riches for the use of the 137Cs gamma radiation source. Also, many thanks to Justa Heinen-Kay, Jessie Tanner, Rachel Olzer, and Rebecca Ehrlich and to 2 anonymous reviewers for their insightful comments on an earlier version of this manuscript. Conflict of Interest: The authors declare that they have no conflict of interest to report. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Boulton et al (2017). REFERENCES Arnqvist G, Nilsson T. 2000. The evolution of polyandry: multiple mating and female fitness in insects. Anim Behav . 60: 145– 164. Google Scholar CrossRef Search ADS PubMed  Arnqvist G, Rowe L. 2005. Sexual conflict . Princeton (NJ): Princeton University Press. Google Scholar CrossRef Search ADS   Baer B, Morgan ED, Schmid-Hempel P. 2001. A nonspecific fatty acid within the bumblebee mating plug prevents females from remating. Proc Natl Acad Sci USA . 98: 3926– 3928. Google Scholar CrossRef Search ADS PubMed  Baker RR, Bellis MA. 1988. Kamikaze sperm in mammals. Anim Behav . 36: 936– 939. Google Scholar CrossRef Search ADS   Bates D, Maechler M, Bolker B, Walker S. 2015. Fitting Linear Mixed-Effects Models Using lme4. J Stat Softw . 67: 1– 48. Google Scholar CrossRef Search ADS   Bawab FM, El-Sherief SS. 1989. Contributions to the study of the origin, nature and formation of the plug in the spermatheca of the female crab Portunus pelagicus (Linnaeus, 1766) (Decapoda, Brachyura). Crustaceana . 57: 9– 24. Google Scholar CrossRef Search ADS   Birkhead TR. 1998. Cryptic female choice: criteria for establishing female sperm choice. Evolution . 52: 1212– 1218. Google Scholar CrossRef Search ADS PubMed  Boorman E, Parker GA. 1976. Sperm (ejaculate) competition in Drosophila melanogaster, and the reproductive value of females to males in relation to female age and mating status. Ecol Entomol . 1: 145– 155. Google Scholar CrossRef Search ADS   Boulton RA, Shuker DM. 2013. Polyandry. Curr Biol . 23: R1080– R1081. Google Scholar CrossRef Search ADS PubMed  Boulton RA, Collins LA, Shuker DM. 2014. Beyond sex allocation: the role of mating systems in sexual selection in parasitoid wasps. Biol Rev Camb Philos Soc . 90: 599– 627. Google Scholar CrossRef Search ADS PubMed  Boulton RA, Shuker DM. 2015a. A sex allocation cost to polyandry in a parasitoid wasp. Biol Lett . 11: 4– 7. Google Scholar CrossRef Search ADS   Boulton RA, Shuker DM. 2015b. The costs and benefits of multiple mating in a mostly monandrous wasp. Evolution . 69: 939– 949. Google Scholar CrossRef Search ADS   Boulton RA, Shuker DM. 2016. Polyandry is context dependent but not convenient in a mostly monandrous wasp. Anim Behav . 112: 119– 125. Google Scholar CrossRef Search ADS   Boulton RA, Cook N, Green J, Greenway EV (Ginny), Shuker DM. 2017. Sperm blocking is not a male adaptation to sperm competition in a parasitoid wasp. Dryad Digital Repository . http://datadryad.org/resource/doi:10.5061/dryad.23f0c. Burton-Chellew MN, Koevoets T, Grillenberger BK, Sykes EM, Underwood SL, Bijlsma K, Gadau J, van de Z ande L, Beukeboom LW, West SAet al.   2008. Facultative sex ratio adjustment in natural populations of wasps: cues of local mate competition and the precision of adaptation. Am Nat . 172: 393– 404. Google Scholar CrossRef Search ADS PubMed  Burton-Chellew MN, Beukeboom LW, West SA, Shuker DM. 2007. Laboratory evolution of polyandry in the parasitoid wasp Nasonia vitripennis. Anim Behav . 74: 1147– 1154. Google Scholar CrossRef Search ADS   Daly M. 1978. The cost of mating. Am Nat . 112: 771– 774. Google Scholar CrossRef Search ADS   de la Filia AG, Bain SA, Ross L. 2015. Haplodiploidy and the reproductive ecology of Arthropods. Curr Opin Insect Sci . 9: 36– 43. Google Scholar CrossRef Search ADS   den Boer SP, Baer B, Boomsma JJ. 2010. Seminal fluid mediates ejaculate competition in social insects. Science . 327: 1506– 1509. Google Scholar CrossRef Search ADS PubMed  Dougherty LR, Simmons LW, Shuker DM. 2016. Postcopulatory sexual selection when a female mates once. Anim Behav . 116: 13– 16. Google Scholar CrossRef Search ADS   Finney GL, Flanders SE, Smith HS. 1947. Mass culture of Macrocentrus ancylivorus and its host, the potato tuber moth. Hilgardia . 17: 437– 483. Google Scholar CrossRef Search ADS   Flanders SE. 1956. The mechanisms of sex-ratio regulation in the (parasitic) Hymenoptera. Insectes Soc . 3: 325– 334. Google Scholar CrossRef Search ADS   Garcia‐Gonzalez F, Evans JP. 2011. Fertilization success and the estimation of genetic variance in sperm competitiveness. Evolution . 65: 746– 756. Google Scholar CrossRef Search ADS PubMed  Geuverink E, Gerritsma S, Pannebakker BA, Beukeboom LW. 2009. A role for sexual conflict in the evolution of reproductive traits in Nasonia wasps? Anim Biol . 59: 417– 434. Google Scholar CrossRef Search ADS   Godfray HCJ. 1994. Parasitoids: behavioural and evolutionary ecology . Princeton (NJ): Princeton University Press. Grillenberger BK, Koevoets T, Burton-Chellew MN, Sykes EM, Shuker DM, Van de Zande L, Bijlsma R, Gadau J, Beukeboom LW. 2008. Genetic structure of natural Nasonia vitripennis populations: validating assumptions of sex-ratio theory. Mol Ecol . 17: 2854– 2864. Google Scholar CrossRef Search ADS PubMed  Hamilton WD. 1967. Extraordinary sex ratios. A sex-ratio theory for sex linkage and inbreeding has new implications in cytogenetics and entomology. Science . 156: 477– 488. Google Scholar CrossRef Search ADS PubMed  Harcourt AH. 1991. Sperm competition and the evolution of nonfertilizing sperm in mammals. Evolution . 1: 314– 328. Google Scholar CrossRef Search ADS   Herberstein ME, Schneider JM, Uhl G, Michalik P. 2011. Sperm dynamics in spiders. Behav Ecol . 22: 692– 695. Google Scholar CrossRef Search ADS   Holmes HB. 1974. Patterns of sperm competition in Nasonia vitripennis. Can J Genet Cytol . 16: 789– 795. Google Scholar CrossRef Search ADS PubMed  Hosken DJ, Stockley P. 2004. Sexual selection and genital evolution. Trends Ecol Evol . 19: 87– 93. Google Scholar CrossRef Search ADS PubMed  King PE. 1961. A possible method of sex ratio determination in the parasitic hymenopteran Nasonia vitripennis. Nature . 189: 330– 331. Google Scholar CrossRef Search ADS   King PE. 1962. The structure and action of the spermatheca in Nasonia vitripennis (Walker) (Hymenoptera: Pteromalidae). Proc R Ent Soc Lond (A) . 37: 73– 75 Manier MK, Belote JM, Berben KS, Novikov D, Stuart WT, Pitnick S. 2010. Resolving mechanisms of competitive fertilization success in Drosophila melanogaster. Science . 328: 354– 357. Google Scholar CrossRef Search ADS PubMed  Manier MK, Belote JM, Berben KS, Lüpold S, Ala-Honkola O, Collins WF, Pitnick S. 2013. Rapid diversification of sperm precedence traits and processes among three sibling Drosophila species. Evolution . 67: 2348– 2362. Google Scholar CrossRef Search ADS PubMed  Martel V, Darrouzet E, Boivin G. 2011. Phenotypic plasticity in the reproductive traits of a parasitoid. J Insect Physiol . 57: 682– 687. Google Scholar CrossRef Search ADS PubMed  Martel V, Shuker DM, Boulton RA, Damiens D, Boivin G. 2016. Sex allocation and the evolution of insemination capacity under local mate competition. Entomol Exp Appl . 159: 230– 242. Google Scholar CrossRef Search ADS   Mikheyev AS. 2003. Evidence for mating plugs in the fire ant Solenopsis invicta. Insectes Soc . 50: 401– 402. Google Scholar CrossRef Search ADS   Moore HD, Martin M, Birkhead TR. 1999. No evidence for killer sperm or other selective interactions between human spermatozoa in ejaculates of different males in vitro. Proc Biol Sci . 266: 2343– 2350. Google Scholar CrossRef Search ADS PubMed  Parker GA. 1970. Sperm competition and its evolutionary consequences in the insects. Biol Rev Camb Philos Soc . 45: 525– 567. Google Scholar CrossRef Search ADS   Pizzari T, Wedell N. 2013. The polyandry revolution. Philos Trans R Soc Lond B Biol Sci . 368: 1471– 2970. Google Scholar CrossRef Search ADS   Price CS, Dyer KA, Coyne JA. 1999. Sperm competition between Drosophila males involves both displacement and incapacitation. Nature . 400: 449– 452. Google Scholar CrossRef Search ADS PubMed  Ramadan MM, Wong TTY. 1989. Effect of gamma radiation on Biosteres longicaudatus (Ashmead) (Hymenoptera: Braconidae), a Larval Parasitoid of Dacus dorsalis Hendel (Diptera: Tephritidae). Proc Hawaiian Entomol Soc . 29: 111– 113. RAY DT. 1948. Dominant lethals induced by X-rays in sperm of the chalcidoid wasp Nasonia brevicornis Ashmead [=Mormoniella vitripennis (Walker) fide Muesebeck in lit.]. Biol Bull . 95: 257. Google Scholar PubMed  R Core Team. 2016. R: A language and environment for statistical computing . Vienna: R Foundation for statistical computing. Ridley M. 1988. Mating frequency and fecundity in insects. Biol Rev Camb Philos Soc . 63: 509– 549. Google Scholar CrossRef Search ADS   Ridley M. 1993. Clutch size and mating frequency in parasitic hymenoptera. Am Nat . 142: 893– 910. Google Scholar CrossRef Search ADS   SAUL GB 2nd. 1955. The induction by x-rays of recessive lethals in the mature sperm of Mormoniella vitripennis (Walker). Radiat Res . 2: 447– 460. CrossRef Search ADS PubMed  Shuker DM, West SA. 2004. Information constraints and the precision of adaptation: sex ratio manipulation in wasps. Proc Natl Acad Sci USA . 101: 10363– 10367. Google Scholar CrossRef Search ADS PubMed  Shuker DM, Pen I, Duncan AB, Reece SE, West SA. 2005. Sex ratios under asymmetrical local mate competition: theory and a test with parasitoid wasps. Am Nat . 166: 301– 316. Google Scholar CrossRef Search ADS PubMed  Simmons LW. 2001. Sperm competition and its evolutionary consequences in the insects . Chichester: Princeton University Press.#8232; Simmons LW. 2014. Sperm competition. In: Shuker DM, Simmons LW, editors. The Evolution of Insect Mating systems . London: Oxford University press. p. 181– 203. Google Scholar CrossRef Search ADS   Simmons LW, Parker GA, Stockley P. 1999. Sperm displacement in the yellow dung fly, Scatophaga stercoraria: an investigation of male and female processes. Am Nat . 153: 302– 314. Google Scholar CrossRef Search ADS   Simmons LW, Stockley P, Jackson RL, Parker GA. 1996. Sperm competition or sperm selection: no evidence for female influence over paternity in yellow dung flies Scatophaga stercoraria. Behav Ecol Sociobiol . 38: 199– 206. Google Scholar CrossRef Search ADS   Siva‐Jothy MT, Tsubaki Y. 1994. Sperm competition and sperm precedence in the dragonfly Nanophya pygmaea. Physiol Entomol . 19: 363– 366. Google Scholar CrossRef Search ADS   Slatyer RA, Mautz BS, Backwell PRY, Jennions MD. 2012. Estimating genetic benefits of polyandry from experimental studies: a meta-analysis. Biol Rev Camb Philos Soc . 87: 1– 33. Google Scholar CrossRef Search ADS PubMed  Snook RR, Hosken DJ. 2004. Sperm death and dumping in Drosophila. Nature . 428: 939– 941. Google Scholar CrossRef Search ADS PubMed  Taylor ML, Price TAR, Wedell N. 2014. Polyandry in nature: a global analysis. Trends Ecol Evol . 29: 376– 383. Google Scholar CrossRef Search ADS PubMed  van de Zande L, Ferber S, de Haan A, Beukeboom LW, van Heerwaarden J, Pannebakker BA. 2014. Development of a Nasonia vitripennis outbred laboratory population for genetic analysis. Mol Ecol Resour . 14: 578– 87. Google Scholar CrossRef Search ADS PubMed  van den Assem J, Visser J. 1976. Aspects of sexual receptivity in female Nasonia vitripennis. Biol. Behav. 1:37–56. van den Assem J, Feuth-De Bruijn E. 1977. Second matings and their effect on the sex ratio of the offspring in Nasonia vitripennis (Hymenoptera: Pteromalidae). Entomol Exp Appl . 21: 23– 28. Google Scholar CrossRef Search ADS   van den Assem J, Jachmann F. 1999. Changes in male perseverance in courtship and female readiness to mate in a strain of the parasitic wasp Nasonia vitripennis over a period of 20+ years. Neth J Zool . 49: 125– 137. Google Scholar CrossRef Search ADS   van den Assem J. 1977. Note on the ability to fertilize following insemination (with females of Nasonia vitripennis, Hymenoptera: Chalcidoidea). Neth J Zoo . 27: 230– 235. Google Scholar CrossRef Search ADS   Werren JH. 1980. Sex-ratio adaptations to local mate competition in a parasitic wasp. Science . 208: 1157– 1159. Google Scholar CrossRef Search ADS PubMed  Werren JH. 1983. Sex ratio evolution under local mate competition in a parasitic wasp. Evolution . 37: 116– 124. Google Scholar CrossRef Search ADS PubMed  Whiting AR. 1967. The biology of the parasitic wasp Mormoniella vitripennis [= Nasonia brevicornis](Walker). Q Rev Biol . 42: 333– 406. Google Scholar CrossRef Search ADS   Wilkes A. 1965. Sperm transfer and utilization by the arrhenotokous wasp Dahlbominus fuscipennis (Zett.) (Hymenoptera: Eulophidae). Can Entomol . 97: 647– 657. Google Scholar CrossRef Search ADS   Wishart GJ, Dick LA. 1985. Effect of γ-radiation on fowl sperm function in vitro and in vivo. J Reprod Fertil . 75: 617– 622. Google Scholar CrossRef Search ADS PubMed  Wylie HG. 1976. Interference among females of Nasonia vitripennis (Hymenoptera: Pteromalidae) and its effect on sex ratio of their progeny. Can Entomol . 108: 655– 661. Google Scholar CrossRef Search ADS   © The Author(s) 2017. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral Ecology Oxford University Press

Sperm blocking is not a male adaptation to sperm competition in a parasitoid wasp

Loading next page...
 
/lp/ou_press/sperm-blocking-is-not-a-male-adaptation-to-sperm-competition-in-a-iGcd00zBjD
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1045-2249
eISSN
1465-7279
D.O.I.
10.1093/beheco/arx156
Publisher site
See Article on Publisher Site

Abstract

Abstract The extent to which sperm or ejaculate-derived products from different males interact during sperm competition—from kamikaze sperm to sperm incapacitation—remains controversial. Repeated matings in the parasitoid wasp Nasonia vitripennis lead to a short-term reduction of efficient sperm use by females, which is crucial for a haplodiploid organism when needing to allocate sex adaptively (i.e. by fertilizing eggs to produce daughters). Repeated matings by females in this species, therefore, constrain sex allocation through this “sperm-blocking” effect, eliciting a cost to polyandry. Here, we explore the causes and consequences of sperm blocking and test the hypothesis that it is an ejaculate-related trait associated with sperm competition. First, we show that sperm blocking, which leads to an overproduction of sons, is not correlated with success in either offensive or defensive roles in sperm competition. Then, we show that the extent of sperm blocking is not affected by self–self or kin–kin ejaculate interactions when compared to self vs nonself or kin versus nonkin sperm competition. Our results suggest that sperm blocking is not a sperm competition adaptation, but is instead associated with the mechanics of processing sperm in this species, which are likely shaped by selection on female reproductive morphology for adaptive sex allocation. INTRODUCTION Polyandry, or the multiple mating of females with different males, is now known to be widespread in nature (Pizzari and Wedell 2013; Taylor et al. 2014). Polyandry has also been clearly demonstrated to confer significant benefits in a wide range of taxa (Arnqvist and Nilsson 2000; Slatyer et al. 2012; Taylor et al. 2014). However, mating can also be costly for females (Daly 1978; Boulton and Shuker 2013), leading to sexual conflict between males and females over the mating rate, even if some degree of multiple mating for females is adaptive (Arnqvist and Rowe 2005). In the parasitoid wasp Nasonia vitripennis, females suffer a novel cost from mating multiply, as repeated matings constrain their ability to allocate sex optimally (Boulton and Shuker 2015a). This effect arises from the fact that multiple ejaculates in a female limit the ability of that female to mobilize and use sperm to fertilize eggs, an effect that we have labeled “sperm blocking” (Boulton and Shuker 2015a). Although N. vitripennis are mostly monandrous in the wild, polyandry evolves under laboratory culture conditions (van den Assem and Jachmann 1999; Burton-Chellew et al. 2007) and can provide a fecundity benefit (Boulton and Shuker 2015b). As with all Hymenoptera, N. vitripennis is haplodiploid, and so mated females can choose whether or not to lay a fertilized egg (which develops into a daughter) or an unfertilized egg (which develops into a son). Females allocate sex in line with the predictions of local mate competition (LMC) theory (Hamilton 1967). Under high LMC conditions typically experienced by N. vitripennis in the field, females maximize their fitness by producing highly female-biased broods (Werren 1980, 1983; Shuker and West 2004; Shuker et al. 2005; Burton-Chellew et al. 2008). After a second mating, however, females are temporarily constrained in their ability to produce daughters (for at least 24 h) resulting in a fitness cost for polyandrous mothers (Boulton and Shuker 2015a). Often the costs of mating to females are thought of as arising due to sexual conflict. Males can benefit from increasing the costs of mating to females if this discourages female remating and so reduces the risk of encountering sperm competition (Simmons 2001). However, the sex allocation cost that we see in N. vitripennis does not appear to benefit males. On the contrary, at first glance, it appears to reduce male fitness as females lay fewer fertilized eggs (daughters) and males only pass on their genes through these daughters, in haplodiploids. The potential fitness costs of sperm blocking to both males and females does mean that it represents something of a paradox as to its origin and function. Here, we will consider 2 possible scenarios. First, sperm blocking may be beneficial to males as an adaptation to sperm competition (which is expected to be higher in laboratory strains demonstrating elevated polyandry; van den Assem and Jachmann 1999; Burton-Chellew et al. 2007), with sperm blocking a side-effect of increased paternity share when ejaculates compete. In other words, although sperm blocking reduces the number of daughters produced, sperm blocking may still be beneficial for a male if it increases his share of the remaining (female) offspring. Although it is clear that adaptations to sperm competition, such as physical displacement and increased swimming velocity, can and do occur (Manier et al. 2013), the extent to which sperm or ejaculate-related products directly, i.e. physically, interact has been subject to controversy. Despite hypotheses of kamikaze sperm and ejaculate–ejaculate phenomena such as sperm incapacitation, how common many of these phenotypes are is unclear and remains contentious (Baker and Bellis 1988; Harcourt 1991; Price et al. 1999; Moore et al. 1999; Snook and Hosken 2004; den Boer et al. 2010; Manier et al. 2010). Furthermore, it may be that patterns of sperm precedence that are commonly interpreted as male adaptations actually reflect female physiological characteristics that have been shaped by natural selection to optimize sperm use and storage (Simmons et al. 1999; Hosken and Stockley 2004; Herberstein et al. 2011). This is represented in our second scenario, whereby disrupted sex allocation is nonadaptive and occurs as a result of female morphological and physiological constraints on sperm processing, storage, and usage. The sperm competition scenario generates testable predictions. Sperm blocking may enhance the fertilization success of one male over the other in a number of ways. First, it may reflect a defensive adaptation that reduces the risk of encountering sperm competition for the first male and thus increases his fertilization success (Figure 1A; Simmons 2001). For instance, in some insect species, including some hymenopterans, males transfer mating plugs that obstruct the entry of rival ejaculates into the female reproductive tract (Baer et al. 2001; Mikheyev 2003; Simmons 2001, 2014). Although there is no evidence for a physical mating plug in Nasonia, a form of chemical mating plug or sperm–sperm interaction that obstructs sperm movement after copulation (as seen in some species of crab; Bawab and El-Sherief 1989) may be present. If a defensive adaptation against sperm competition such as this occurs in N. vitripennis, then it may not only obstruct incoming sperm but may also impede the movement of outgoing sperm to be used for fertilization, resulting in sperm blocking (female parasitoid wasps possess only a single spermathecal duct; King 1961). In this scenario, we predict that the severity of the overproduction of sons (i.e. sperm blocking) will be positively correlated with first male fertilization success and negatively correlated with second male fertilization success (Figure 1A). Figure 1 View largeDownload slide Schematic representing predicted results if sperm blocking is (A) a defensive male trait or (B) an offensive male trait. The x-axis represents sperm blocking which is the total number of sons in the first clutch laid. The y-axis is the total number (over all clutches) of eggs fertilized (daughters sired) by the first male (P1 solid line) or the second male (P2 dashed line). We estimated P1 and P2 in experiment 2 using the sterile male technique (see main text for details). Figure 1 View largeDownload slide Schematic representing predicted results if sperm blocking is (A) a defensive male trait or (B) an offensive male trait. The x-axis represents sperm blocking which is the total number of sons in the first clutch laid. The y-axis is the total number (over all clutches) of eggs fertilized (daughters sired) by the first male (P1 solid line) or the second male (P2 dashed line). We estimated P1 and P2 in experiment 2 using the sterile male technique (see main text for details). In contrast, it may be that sperm blocking occurs as a result of an offensive adaptation to sperm competition, increasing the success of a competing male when sperm competition is encountered (Simmons 2001). In this case, we envisage a scenario where the second male’s ejaculate blocks the female spermathecal duct, preventing the first males ejaculate from being used to fertilize eggs and resulting in sperm blocking. Our prediction for this situation is that the degree of sperm blocking will be positively correlated with second male fertilization success and negatively correlated with first male fertilization success (Figure 1B). As a further test of sperm blocking as a male adaptation, if there is some chemically mediated process at work, we might expect that sperm blocking should not occur—or should at least be considerably reduced—when ejaculates from the same male, or from closely related males, come into competition. For instance, seminal fluid proteins can play a role in the incapacitation of rival unrelated sperm. In several polyandrous species of social hymenopteran, males exhibit offensive seminal fluid traits that target the sperm of their rivals (den Boer et al. 2010). This may even be facilitated by haplodiploidy, because individual sperm produced by a single haploid male will, except in the event of mutation, be identical (de la Filia et al. 2015). We, therefore, predict that sperm blocking will be more severe when unrelated ejaculates come into competition. The second possibility is that sperm blocking is not a result of male adaptations to sperm competition, but instead reflects female physiological processes that constrain efficient sperm use. For instance, Price et al (1999) suggested that in Drosophila melanogaster seminal fluid proteins serve to incapacitate previously stored sperm of rival males. However, Snook and Hosken (2004) found evidence to suggest that this apparent male incapacitation actually occurs because females “dump” the sperm from previous ejaculates, which can result in skewed sperm precedence toward either male (see also Manier et al. 2013). Similarly, in N. vitripennis, sperm blocking could occur as a result of sperm dynamics and movement within the female reproductive tract if, for instance, females require time to move sperm from the site of storage to the site of fertilization. Although we do not explicitly test this hypothesis here, we suggest that sperm blocking most probably occurs due to constraints on sperm processing by females, if there is no effect of mating order or male relatedness on the sex ratio. We present the results of a series of experiments to contrast these 2 scenarios. In the first 2 experiments, we evaluate and then use the sterile male technique (Parker 1970; Ramadan and Wong 1989; Siva-Jothy and Tsubaki 1994) to test whether sperm blocking is an offensive or defensive trait that is associated with increased sperm precedence for the first or second male to mate. In the next 2 experiments, we test whether sperm blocking is influenced by the relatedness of a female’s mating partners. Specifically, we ask whether sperm blocking is ameliorated if a female mates twice to the same male (versus 2 different males) or with 2 brothers (vs. 2 unrelated males). METHODS Study species Nasonia vitripennis is a gregarious chalcidoid parasitoid wasp. It attacks dipteran pupae and is an ectoparasitoid, depositing eggs on the surface of the developing larva within the puparium (Whiting 1967). As with all Hymenoptera, N. vitripennis is haplodiploid and females facultatively lay either fertilized eggs that develop as daughters or unfertilized eggs that develop as sons. Females lay multiple eggs on a single host (i.e. they are gregarious) and sibmating at the emergence site is the norm (Boulton et al. 2014). After mating, the winged females disperse to find hosts on which to oviposit (brachypterous males are largely confined to the natal patch; Boulton et al. 2014). The local mating patches that result from this mating system lead to local mate competition (LMC; Hamilton 1967), and this in turn leads to selection for female-biased offspring sex ratios. In the wild, females typically oviposit alone (Grillenberger et al. 2008) resulting in high LMC because males will mate exclusively with their sisters. In this situation, females will maximize their fitness by producing only enough sons to inseminate their daughters. As more females contribute offspring to a mating patch, males can mate with nonsibling females and this increases their reproductive value and so less female-biased sex ratios are favored. Female Nasonia show an impressive ability to allocate sex facultatively in line with the predictions of the LMC theory (Werren 1980; 1983; Shuker and West 2004; Shuker et al. 2005; Burton-Chellew et al. 2008; Martel et al. 2016). The strain of N. vitripennis used for these experiments was the outbred strain HVRx, which was collected from 5 sites in the Netherlands in 2001 (van de Zande et al. 2014), reared on Calliphora vicina pupae as hosts. We reared wasps at 25 °C on a 16:8 h light:dark cycle, which results in a 14-day egg-to-adult cycle. To standardize the rearing environment of focal males and females to be used in the following experiments, we provided virgin and mated females with 3 C. vicina hosts each for 48 h, these females served as grandmothers for the focal individuals used for experiments. We isolated focal females as pupae (based on the presence of wing buds and a visible ovipositor) from the hosts 2 days prior to expected adult eclosion, 12 days after the grandmothers had been provided with hosts to oviposit on. The focal males were sons produced by virgin grandmothers, and these males were left to eclose naturally. First, we will describe our pilot experiment to calculate the lowest effective sterilizing dose for our sterile-male sperm competition experiment. Pilot: determining the lowest effective sterilizing dose We sterilized N. vitripennis males using gamma radiation in order to investigate patterns of sperm precedence. Gamma radiation has been shown to sterilize males of many species effectively without detrimentally affecting sperm morphology (Wishart and Dick 1985). If the dose is sufficiently low, irradiation does not affect sperm motility or other aspects of the ejaculate, but the accumulation of lethal mutations in the sperm render them infertile (Ray 1948; Wishart and Dick 1985). The main reason for using the sterile male technique in N. vitripennis is due to the lack of suitable morphological markers. Although previous studies have used red- or oyster-eyed inbred mutant strains to assay sperm precedence (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977), these traits are recessive and so sperm precedence cannot be assayed in a wild-type background. Although sterilization can result in reduced fertilizing capacity (Simmons 2001), this issue can be reduced through pilot studies that assess the Lowest Effective Sterilizing Dose (LESD), and through randomization of the mating sequence (Simmons 2001). To our knowledge, there are no published studies where gamma radiation has been used to sterilize N. vitripennis for the purposes of assaying sperm precedence (but see Ray 1948 and Saul 1955 for other uses). Therefore, we conducted a pilot experiment to ascertain the LESD by generating a dose-response curve for 4 treatment groups: 80, 100, 120, and 140 Gy of radiation (Ray 1948). The dose was determined by varying the amount of time that a vial containing male N. vitripennis was exposed to a source of gamma radiation. The dose rate was 2.59 Gy/min, so that males that received 80 Gy remained in the radiation chamber for 1854 s, 100 Gy for 2316 s, 120 Gy for 2778 s, and 140 Gy for 3240 s. On the day of emergence, focal males were exposed, in 4 groups of approximately 100 wasps, to a 137Cs source emitting gamma radiation. The source rotates so that the dose rate is constant over space. Males from each treatment (plus untreated controls) were then mated to virgin females either straight after irradiation (day 1) or 24 h later (day 2; N = 25 per treatment per day). All pairings were observed to determine whether successful copulation had occurred, after which the females were provided with hosts that were maintained at 25 °C. After the offspring had emerged and died, the number of sons (unfertilized eggs) and daughters (fertilized eggs) were counted. A mating was considered sterile when 100% of the offspring were male. The dose-response curve (Figure 2) shows the percentage of females from each treatment group that produced any daughters. Figure 2 View largeDownload slide Dose-response curve for males irradiated with 80–140 Gy of gamma radiation (137Cs) and control (untreated) males in the pilot (Error bars = binomial Confidence Intervals). Figure 2 View largeDownload slide Dose-response curve for males irradiated with 80–140 Gy of gamma radiation (137Cs) and control (untreated) males in the pilot (Error bars = binomial Confidence Intervals). Our results suggest that the minimum effective dose was 100 Gy, after which less than 5% of males successfully fertilized eggs. Irradiation did induce sterility, as matings with control males were significantly more likely to result in daughter production than matings with exposed males (Quasibinomial GLM: F4, 178 = 92.77, P < 0.0001). Sterilization was not effective below 100 Gy (Tukey P < 0.007) but fertility was no different when males were treated with 100–140 Gy (100 Gy is the LESD). There was no effect of the day on which mating occurred on fertility (F1, 181 = 0.02, P = 0.89). Finally, if the sperm of irradiated males became less viable over time (Simmons et al. 1996), then the sex ratio produced by females mated on the second day should be more male biased (since unfertilized eggs will develop as males). This was not the case, however, (day: F1,174 = 0.29, P = 0.59; interaction effect treatment*day F4,170 = 0.56, P = 0.69) suggesting that the sperm of irradiated males did not decrease in viability over the experimental period. Experiment 1: assaying sperm precedence using gamma-irradiated males Four hundred virgin females were isolated from a grand-maternal generation. Virgin males were generated from unmated grandmothers and were maintained in groups of brothers. Six stock tubes containing brothers were exposed to 100 Gy of gamma radiation. We used irradiated males from 6 families to standardize competitor identity as much as possible, ideally the same males would have been used repeatedly (García‐González and Evans 2011) but male N. vitripennis do not produce new sperm after emergence (i.e. they are prospermatogenic; Boulton et al. 2014) and using the same males repeatedly would result in mating order effects on fecundity and the sex ratio (Boulton and Shuker 2015b). In this experiment, we investigated whether sperm blocking is associated with patterns of sperm precedence. To do this, focal females and males were assigned to one of the following 7 treatments: (V) virgin female; (N) once-mated to a normal male; (I) once-mated to an irradiated male; (NN) twice-mated to two normal males; (II) twice-mated to two irradiated males; (NI) mated to a normal male followed by an irradiated male; (IN) mated to an irradiated male followed by a normal male (Total N = 348, N = 45–54 across treatments). To help explain the rationale for our interpretation of the following results, our assumptions are as follows. The production of a daughter arises from a fertilized egg, that is, an egg fertilized by the mother using a viable sperm (from a normal ‘N’ male). A male is produced from an unfertilized egg that the mother left unfertilized. The effects of irradiation are manifested through mothers fertilizing eggs with infertile sperm (from an ‘I’ male), which results in the production of inviable embryos that fail to develop. This will lead to a change in the offspring sex ratio (fewer females, so less female-biased or even male-biased sex ratios), as well as a reduction in the total number of adult offspring produced (eggs fertilized by inviable ‘I’ sperm will not develop into adults). In terms of the experiment, females assigned to the twice-mated treatments (II, IN, NI, NN) were mated first on Day 1. All copulations were observed and in order to increase the likelihood of remating, we prevented males from engaging in postcopulatory courtship as this behavior serves to reduce female receptivity to additional matings (van den Assem and Visser 1976). Twenty-four hours after their first mating (on Day 2), these females were presented with a second male. All once-mated females were presented with their first and only male on Day 2. Females were given 3 C. vicina hosts on Day 2 (i.e. all females were given hosts immediately after their final mating) and kept in an incubator at 25 °C. In this experiment, the initial 3 hosts were removed 24 h later and replaced with a single host. This was repeated every day for 3 days such that females surviving until the end of the experiment received a total of 3 hosts in 4 batches. Three hosts were presented in the first batch, to allow host-feeding opportunities as well as to provide oviposition resources. These hosts were maintained at 25 °C and emerging sons and daughters were counted after they had died. We kept experimental females in isolation after host provisioning and checked them every day at 9 am, 12 pm and 5 pm for mortality. After all females had died, we removed and measured their right hind tibia as a proxy for body size (Godfray 1994) in order to assess whether larger females (possibly with larger spermathecae; Martel et al. 2011) suffer sperm blocking to a lesser extent. An Olympus SZX10 microscope with an ocular micrometer was used for all measurements and each tibia was repositioned and remeasured 3 times in order to assess measurement error, which we found to be low based on a high repeatability estimate (Intraclass correlation coefficient: ICC = 0.94 ± 0.006 CI, P < 0.001). We tested whether treatment or host batch influenced the sex ratio (measured throughout as proportion of offspring that are male) that females produced across all treatments using a repeated-measures Quasibinomial GLMM (lme4 (Bates et al. 2015) in R Studio, RStudio, Inc., Boston, MA; R Core Team 2016). Fixed factors were treatment, host batch and their interaction. Female longevity (hours) and hind tibia length (mm) were included as covariates. Female identity nested within host batch was included as a random effect. Pairwise comparisons within these models allowed us to assess the effects of treatment on sperm blocking as well as the effectiveness of irradiation for sterilizing males. If irradiation successfully sterilizes males without impairing their ability to transfer an ejaculate, virgin females should produce significantly larger (all-male) clutches than mated (I or II) females (virgin female clutches are generally the same size as the clutches of once-mated females in N. vitripennis, only all-male clutches; Boulton and Shuker 2015a). We, thus, tested whether females that mated with either one (I) or two (II) irradiated males produced fewer sons (unfertilized eggs) than virgin females using a repeated-measures QuasiPoisson GLMM. To investigate patterns of sperm precedence and mixing, we tested whether daughter production (i.e. eggs fertilized by N males that develop successfully) differed between treatment groups NN, NI and IN over the four host batches using a repeated-measures GLMM with a QuasiPoisson error structure. For this analysis, treatment, host batch and their interaction effect were included in the model as fixed factors, female longevity and hind tibia length as covariates, and female identity nested within host batch as a random effect. The following formula was used to calculate sperm precedence (PN) in treatments NI (P1) and IN (P2) (modified from Boorman and Parker 1976):  PN= x−zp−z where, x = proportion of daughters (fertilized eggs) in NI or IN matings, z = proportion of daughters in II matings (0.018), p = proportion of daughters in N matings (0.814). We considered daughter production in II as z rather than daughter production in I because the reduced daughter production that occurs after repeated matings (sperm blocking) renders this estimate more comparable to the other twice-mated treatments (NI and IN). To calculate p we used the proportion of daughters in N matings to estimate “maximum” paternity by a single male, and NN matings were not used here as paternity could not be assigned to either male (in addition, daughter production in NN matings was reduced by sperm blocking and would underestimate maximum daughter production; see Results below). In the original formula described in Boorman and Parker (1976)x, z, and p are the proportion of eggs fertilized by the sperm of the N or I males and are calculated from the total number of viable offspring that hatch divided by the total number of eggs laid (including inviable eggs). Here, however, we have modified this formula due to constraints imposed by haplodiploidy and the life cycle of Nasonia. We consider the proportion of daughters in a clutch, excluding male offspring from our estimates of sperm precedence. This is because males develop from unfertilized eggs and so are not useful for assigning sperm use to the N male. Additionally, due to the nature of parasitism by N. vitripennis, unfertilized eggs typically cannot be counted. This is because around 40 eggs or so are laid inside the puparium and any eggs that fail to develop are generally destroyed by the larvae that do hatch. As a result, we lack knowledge regarding the number of eggs that failed to develop (those fertilized by the I male). This means that the values of P1 and P2 will be overestimates (see the Supplementary Material, Supplementary Table S1, for calculations of PN from a set of simulated data with and without knowledge of numbers of inviable eggs laid). Although these estimates are sufficient for investigating patterns of sperm precedence within this study, this limitation prevents comparisons of results between studies. Under high LMC (when females oviposit alone), the number of sons in the first host batch is typically around 10%–20% of the total brood size (Werren 1980, 1983). When sperm blocking occurs, however, the proportion of sons in the first host batch increases, as females are unable to mobilize sperm successfully to fertilize eggs. To assess whether sperm blocking in the first clutch influences fertilization success of the N male, we tested whether PN over host batches 2–4 was associated with the number of sons in the first host batch for the treatments IN and NI. We infer P1 from treatment NI and P2 from treatment IN (i.e. the order of the focal male) and predict the following: If sperm blocking is 1) a defensive trait that benefits the first male, there will be a positive correlation between P1 (use of the first male’s sperm in NI) and son production in the first clutch of eggs and a negative correlation for P2 (use of the 2nd male’s sperm in IN; Figure 1A). If, on the other hand, sperm blocking is 2) an offensive trait used by second males to increase their relative paternity, then P2 (use of the 2nd male’s sperm in treatment IN) will be positively correlated and P1 (the first males sperm use in NI) negatively correlated with son production in host batch 1 (sperm blocking; Figure 1B). In other words, if sperm blocking is an offensive sperm competition trait that benefits the second male, high sperm blocking should lead to high P2 (daughter production will be positively correlated with sperm blocking in treatment IN). If it is a defensive trait that benefits the first male, high sperm blocking will lead to higher P1 (daughter production will be positively correlated with sperm blocking in treatment NI). These predictions are outlined in Figure 1. We ran a GLMM with female ID nested within host batch as a random effect, PN was the outcome variable and the predictors were “Number of sons in host batch 1”, as well as treatment, host batch, and the interaction effects. A significant interaction effect between the number of sons and treatment suggests that the relationship between sperm blocking and fertilization success differs depending on the order in which the nonirradiated (N) male has mated (i.e. whether prediction 1 or 2 is supported). Experiment 2: sperm blocking with competition between self versus nonself sperm In our second experiment, we tested whether the presence of ejaculates from 2 different males influences the severity of sperm blocking. We also tested whether, in addition to mating, prolonged exposure to either the same male or 2 different males influenced sperm blocking. This is because previous work had suggested that harassment by males during oviposition could influence sperm use independent of mating (Wylie 1976; Boulton and Shuker 2015b, 2016). To do this, focal females were exposed to a single male, copulation was observed and postcopulatory courtship prevented as before. Females assigned to polyandrous treatments mated again 24 h later. The identity of the male that each female mated with first was recorded and all males were retained. After their first mating, females were randomly assigned to the following treatments, and in each case females were given 3 hosts to oviposit on for 24 h after the final mating, either in the presence of a male or not. The 6 treatments were: 1) Control (C1) with once-mated females given hosts (C. vicina) immediately after mating; 2) control + 24 h (C24) with once-mated females given hosts 24 h after mating; 3) same-male mating (S) with females mating with the same male twice with a 24-h interval between matings; 4) same-male harassment (SH) in which females were exposed to the same male 24 h after the initial mating and the male and female kept together during oviposition; 5) nonself mating (NS) in which females remated with a different once-mated male 24 h after the initial mating; and 6) nonself harassment (NSH) in which females were exposed to a different once-mated male 24 h after the initial mating, with the male and female kept together during oviposition (total N = 196, N = 23–37 across treatments). In order to standardize male mated status in the “same-male” and “nonself” treatments, all second males used had mated once previously. All experimental hosts were removed the following day and incubated at 25 °C. After 2 weeks, the offspring of the focal females emerged from the hosts. After the offspring had died, the number of sons and daughters produced was counted. Experiment 3: sperm blocking with competition between kin versus nonkin sperm In our final experiment, we tested whether the severity of sperm blocking is influenced by male relatedness. The basic design of Experiment 2 was repeated to test how sperm blocking differed when females mated with 2 brothers or 2 unrelated males. Females were randomly assigned to one of the following treatments: 1) Control (C1) with once-mated females given hosts immediately after mating; 2) control + 24 h (C24) with once-mated females given hosts 24 h after mating; 3) kin-mating (K) with females mating with the 2 virgin brothers with a 24-h interval between matings; 4) kin-harassment (KH) in which females were exposed to a virgin brother of their first mate 24 h after the initial mating and where the male and female were kept together during oviposition; 5) nonkin mating (NK) in which females remated with a virgin male that was unrelated to their first mate 24 h after the initial mating; and 6) nonkin harassment (NKH) in which females were exposed to a virgin male that was unrelated to their first mate 24 h after the initial mating, and the male and female were kept together during oviposition (Total N = 206, N = 23–38 across treatments). In Experiments 2 and 3, the sex ratio (proportion of sons) and son and daughter production (total clutch size) was analyzed using GLMMs with a Quasibinomial and Gaussian error structure, respectively, using the package lme4 in R Studio (Bates et al. 2015). Treatment, harassment, and the interaction effect were included as fixed factors. Experiments 2 and 3 were each conducted across 3 blocks, and so experimental block was included as a random effect. RESULTS Experiment 1: assaying sperm precedence using gamma-irradiated males Irradiation was generally effective at preventing daughter production. However, females that mated with 2 irradiated males (II) were more likely to produce some daughters compared to both virgin and I females (Quasibinomial GLMM: F2,544= 95.58, P < 0.0001; Figure 3A). The absolute number of sons produced over the 3 treatments differed significantly. Clutches laid by virgin females are typically comparable in size to mixed-sex clutches (van den Assem 1977) but here virgin females produced around 3 times as many sons as females that mated with 1 or 2 irradiated males (QuasiPoisson GLMM: F3,573= 230.72, P < 0.0001; Figure 3B). This suggests that mated females were being inseminated and using sperm but that the daughters sired by I males made up the putative remaining 60–70% of the clutch that failed to develop. Figure 3 View largeDownload slide (A) Proportion of sons (sex ratio) and (B) total son production by virgin females and females mated to either 1 (I) or 2 (II) irradiated males in experiment 1 (A) Sex ratio (proportion of sons), error bars = binomial confidence Intervals (CIs). Note that the y-axis runs between 0.96 and 1.0 and that in all cases daughter production was extremely low. (B) Absolute number of sons produced. Virgin females (V) produce more sons than females mated with 1 (I) or 2 (II) irradiated males. Error bars = 95% CIs. (A and B) Different lower case letters represent treatment groups that are statistically different (P < 0.05). Figure 3 View largeDownload slide (A) Proportion of sons (sex ratio) and (B) total son production by virgin females and females mated to either 1 (I) or 2 (II) irradiated males in experiment 1 (A) Sex ratio (proportion of sons), error bars = binomial confidence Intervals (CIs). Note that the y-axis runs between 0.96 and 1.0 and that in all cases daughter production was extremely low. (B) Absolute number of sons produced. Virgin females (V) produce more sons than females mated with 1 (I) or 2 (II) irradiated males. Error bars = 95% CIs. (A and B) Different lower case letters represent treatment groups that are statistically different (P < 0.05). Females that mated only once to an unsterilized male (N) produced the most female-biased sex ratios as expected. Twice-mated females all experienced sperm blocking, laying significantly more male-biased sex ratios than ‘N’ females (all pairwise P < 0.001; F3,118 = 25.44, P < 0.0001). The significant difference between the sex ratio produced by N and NN females confirms the problem of sperm blocking for polyandrous females; females that mated twice produced more male-biased sex ratios immediately after mating. Moreover, the sex ratios produced by N and NN females converged in later host batches which demonstrates that the effect of sperm blocking does wear off over time (interaction effect: Quasibinomial GLM: F9,461 = 4.87, P < 0.0001; main effect of treatment F3,467 = 30.65, P < 0.0001; main effect of host batch F3,467 = 9.73, P< 0.0001; See Figure 4). Figure 4 View largeDownload slide Sex ratios (proportion of sons) produced by females that mated with either 1 or 2 unsterilized (N) males in experiment 1 (error bars = binomial CIs). Figure 4 View largeDownload slide Sex ratios (proportion of sons) produced by females that mated with either 1 or 2 unsterilized (N) males in experiment 1 (error bars = binomial CIs). In terms of sperm precedence—and remembering that in haplodiploids such as Nasonia only daughters can tell us about sperm usage—the first male had higher daughter production initially (NI) but sperm mixing occurred in later clutches (Figure 5A). There was a significant effect of treatment on daughter production (F2,371= 35.44, P < 0.0001; Figure 5A), but this was because females that mated to 2 unsterilized males (NN) laid more daughters than females that mated to an unsterilized male either first (NI) or second (IN). In the first host batch, first males (NI) did achieve higher daughter production than second males (IN) but, overall across all host batches, sperm use was not significantly biased towards either male (P1 = 0.76, P2 = 0.64, albeit close to significance: treatment: F1,360= 3.60, P = 0.06). This was partly because daughter production by IN females was equal to that of NI females in host batches 3 and 4 (treatment*host batch interaction effect: F6,367 = 2.37, P = 0.03), suggesting that the second male sperms (viable sperm in treatment IN) are used more in later host batches. There was a significant main effect of host batch on sperm precedence (F3,369 = 10.44, P < 0.0001), but this did not appear to relate to the number of hosts provided per batch (three in host batch 1 and one host in batches 2–4) as the only pairwise significant difference was between host batches 1 and 2 (P < 0.05). Figure 5 View largeDownload slide Daughter production and sperm precedence for twice-mated females in Experiment 1. (A) Fewer daughters (fertilized eggs) were produced by the second male in host batches 1–2 (IN), but there was no significant difference in daughter production by the first (NI) or second (IN) male in host batches 3 and 4. Error bars = 95% CIs. (B) Son production or “sperm blocking” in the first clutch reduced fertilization success in host batches 2–4 for the first (P1: NI) and second (P2: IN) male equally. Figure 5 View largeDownload slide Daughter production and sperm precedence for twice-mated females in Experiment 1. (A) Fewer daughters (fertilized eggs) were produced by the second male in host batches 1–2 (IN), but there was no significant difference in daughter production by the first (NI) or second (IN) male in host batches 3 and 4. Error bars = 95% CIs. (B) Son production or “sperm blocking” in the first clutch reduced fertilization success in host batches 2–4 for the first (P1: NI) and second (P2: IN) male equally. In terms of sperm competition outcomes, more explicitly, sperm blocking was not associated with higher paternity for either the first or the second male. When sperm blocking occurs, more sons are produced in the first host batch. Although “sperm blocking” did reduce the fertilization success of the unsterilized (N) male in host batches 2–4 (β = −0.49, SE = 0.25; F1, 234 = 5.01, P = 0.03), there was no significant interaction between sperm blocking and treatment (F2,231 = 1.77, P = 0.18; Figure 5B) demonstrating that sperm blocking reduces the success of the first and second male equally. We found no evidence that sperm blocking (sex ratio in the first host batch) was related to female size (F1,121 = 1.14, P = 0.29) or longevity (F1,121 = 1.14, P = 0.29). We also found no evidence that female body size interacts with treatment to influence sperm blocking (Treatment*size: F3,118 = 1.29 P = 0.28). Experiment 2: self versus nonself All twice-mated females experienced sperm blocking to the same degree, regardless of whether they mated twice with the same male (S) or with different males (NS; P = 0.99; Figure 6A). Although there was an overall effect of mating treatment on the sex ratio ( F3, 313 = 42.22, P < 0.0001), the only significant pairwise differences were between both controls and twice-mated females (P < 0.0001). There was no effect of harassment (F1, 315 = 0.08, P = 0.78) or any interaction effect between harassment and treatment (F5,311 = 0.72, P = 0.40) on the sex ratio. Figure 6 View largeDownload slide Sex ratio and clutch size for females in Experiments 2 (A and B) and 3 (C and D). White bars = No harassment, gray bars = harassment. (A) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to, the same male twice or 2 different males. (B) The clutch size that females produced was not affected by mating treatment. However, the interaction effect between harassment and treatment was significant. (C) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to 2 brothers (kin) or 2 unrelated males (nonkin). (D) Clutch size was not affected by harassment, the number of matings, or relatedness between the first and second males. Error bars = binomial CIs (for A and C) and 95% CIs (for C and D). Figure 6 View largeDownload slide Sex ratio and clutch size for females in Experiments 2 (A and B) and 3 (C and D). White bars = No harassment, gray bars = harassment. (A) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to, the same male twice or 2 different males. (B) The clutch size that females produced was not affected by mating treatment. However, the interaction effect between harassment and treatment was significant. (C) There was no significant difference in the sex ratios produced by females that either mated with or were exposed to 2 brothers (kin) or 2 unrelated males (nonkin). (D) Clutch size was not affected by harassment, the number of matings, or relatedness between the first and second males. Error bars = binomial CIs (for A and C) and 95% CIs (for C and D). In terms of clutch size (total son and daughter production), there was no significant main effect of harassment (F1, 315 = 0.01, P = 0.91) or treatment (F3, 313 = 2.00, P = 0.11). The interaction effect between treatment and harassment was, however, statistically significant (F5,311 = 5.34, P = 0.02; Figure 6B). When control treatments were removed (no harassment in C1 or C24), the significant interaction effect remained (F3,216 = 5.20, P = 0.02). Prolonged exposure (harassment) reduces offspring production when the first and second males are different individuals (Figure 6B) but clutch sizes were larger when females were exposed to the same male twice. Experiment 3: kin versus nonkin Sperm blocking was not influenced by the relatedness between competing males when females mated twice. Although there was again an overall effect of mating treatment on the sex ratio (F3, 316 =39.75, P < 0.001), the only significant pairwise differences were between the controls and all twice-mated females as before (P < 0.0001). As above, females in the once-mated control treatments laid the most female-biased sex ratios. The sex ratio was less female-biased when females mated twice but relatedness between the males with which a female mated had no effect on the sex ratio (P = 0.99; Figure 6C). There was no effect of harassment on the sex ratio (F1, 318 = 0.04, P = 0.84) nor was there a significant interaction effect between treatment and harassment on the sex ratio produced (F5,314 = 0.02, P = 0.87). Turning again to clutch size, there was no effect of harassment (F1, 318 = 0.79, P = 0.37) and no statistically significant effect of treatment (F3, 316 = 2.56, P = 0.054) on clutch size, nor was there a significant interaction effect between treatment and harassment (F5,314 = 0.05, P = 0.82; Figure 6D). DISCUSSION A diverse range of offensive sperm competition traits have been suggested, and demonstrated, to confer an advantage to males when females mate multiply and sperm competition is encountered (Simmons 2014). However, the parasitoid wasps remain relatively understudied with respect to postcopulatory sexual selection (Boulton et al. 2014). In this study, we investigated the possibility that sperm blocking, that is, the overproduction of unfertilized eggs (sons) that occurs immediately after female remating, might be advantageous to male N. vitripennis when they experience sperm competition. However, it is clear from our results that sperm blocking does not occur as a result of males displacing, blocking, or incapacitating the ejaculates of their rivals. First, in Experiment 1, sperm blocking did not favor the paternity of either the first or second male, and so was not associated with increased paternity in either offensive or defensive sperm competition. Second, in Experiments 2 and 3, sperm blocking was not ameliorated by the presence of a male’s own sperm or the sperm of a brother, and so sperm blocking occurred regardless of sperm identity and origin. In Experiment 1, we found that N. vitripennis males do not actively “block” rival sperm, as sperm blocking reduced paternity for the first and second males equally. We did find, however, that in the first clutch, the first male sired more daughters, but in subsequent clutches second males gained more paternity success, such that the first and second males shared equal paternity overall (P1 = 0.76, P2 = 0.64; note that these values overestimate the true PN and sum to more than 1, see Methods and Supplementary Table S1 in the Supplementary Material). This is not likely to be due to the first male’s ejaculate being fully depleted, because a single mating with a virgin male was sufficient for females to maintain daughter production when provided with up to 24 hosts in a previous study (Boulton and Shuker 2015b). Instead, this pattern is more likely to be due to how sperm is processed and stored. Previous work on N. vitripennis has demonstrated that patterns of sperm use are influenced by mating order and also by the mated status of the male and the timing of the second mating. Holmes (1974) found that if the first male to mate was partially sperm depleted, there was no consistent pattern of use of the first or second male’s sperm (i.e. sperm mixing occurred) and so paternity was more equal than if males were not depleted (in which case paternity was skewed towards the first male). Additionally, van den Assem and Feuth-de-Bruijn (1977) found little sperm mixing when females remated immediately after their initial mating, but sperm mixing increased when there was a delay of 3 days between matings (van den Assem and Feuth-de-Bruijn 1977; van den Assem 1977). The processes that influence the degree of sperm mixing outlined above have also been found to influence sperm blocking. If the second male does not succeed in inseminating the female (Geuverink et al. 2009), or if there is no interval between matings, and so no sperm mixing, van den Assem (1977) found that sperm blocking was less severe (i.e. the sex ratio in the first clutch was less male biased). Superficially, this suggests that sperm blocking might promote sperm mixing, but the current results argue against a causal relationship and show that sperm blocking does not relate to male ejaculate physiology. Nor does sperm blocking appear to be a constraint related to the size of the spermatheca, as we found no effect of female body size (which is associated with larger spermatheca size in another parasiotid, Trichogramma euprodctidis; Martel et al. 2011) on the severity of sperm blocking. Our second and third experiments also do not support the hypothesis that sperm blocking is a targeted male adaptation in N. vitripennis that incapacitates (rather than indiscriminately blocks) unrelated rival male sperm (den Boer et al. 2010). Whether the sperm are one’s own, or that of a brother, or an unrelated competitor, the effect on reduced sperm use by the recipient female is the same. A previous example of apparent “sperm incapacitation” in D. melanogaster (Price et al. 1999) turned out on further investigation not to reflect male ejaculate physiology, but rather female physiological processes (Snook and Hosken 2004; see also Manier et al. 2010). In D. melanogaster, it is the act of copulation itself that leads to females “dumping” sperm from previous ejaculates, resulting in extreme last-male precedence. In D. melanogaster, however, it does appear that some males are better than others at eliciting “sperm dumping”, but we found no evidence that this is the case in N. vitripennis, as increased sperm blocking was not positively associated with sperm precedence. Instead, the patterns of, and correlation between, sperm precedence and sperm blocking exhibited by N. vitripennis in the current study, and in others (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977; van den Assem 1977; Boulton and Shuker 2015b) may relate to the morphology of the spermatheca and the physiology of sperm storage and movement within the female. The single, narrow spermathecal duct of the female’s sperm store is thought to allow the movement of a single sperm to the site of fertilization, limiting wastage of sperm and facilitating precise sex allocation (Flanders 1956; Wilkes 1965). Whilst adaptive in terms of sperm economy, this narrow tube may become congested if the capacity of the spermatheca is exceeded. As such, building on our work and that of earlier authors, we suggest that the phenotype of sperm blocking is generated as follows. During insemination, sperm are deposited at the base of the single, narrow spermathecal duct, from where they rapidly migrate to the rigid spheroid spermathecal capsule. The first sperm reach the capsule after just 1 min (King 1961) but take several hours to quiesce and line up at the capsule mouth, when they are then ready to be used for fertilization (King 1962; Wilkes 1965). The sperm leave the capsule to fertilize the eggs through the same narrow duct. Following a second mating, the presence of incoming sperm in the narrow spermathecal duct prevents sperm required for fertilization from exiting the capsule to some extent (see Wilkes 1965 and Finney et al. 1947 for evidence from other parasitoids). If the first ejaculate is small, or there is sufficient time between matings, then the second ejaculate can enter the spermathecal duct and spermatheca itself, and then sperm mixing can occur, equalizing P1 and P2 (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977; van den Assem 1977). Crucially though, in these cases the second ejaculate is able to enter the spermathecal duct and the incoming ejaculate will temporarily obstruct outgoing sperm, reducing the efficiency of sperm use and generating sperm blocking. The pattern of sperm precedence that we observed in Experiment 1 fits this scenario. The first 2 clutches laid were predominantly sired by the first male to mate, but sperm mixing occurred later. The sperm in the first ejaculate would be the first to line up and quiesce at the mouth of the spermathecal duct, ready to be used for fertilization. The 24-h delay between matings should result in some of the second ejaculate entering the duct, but before the first eggs are fertilized the small capsule will be fairly full. The sperm in the second ejaculate will move into storage slowly, at the same time impeding any outgoing sperm (and resulting in sperm blocking). The space created when the first sperm leaves the capsule will allow more of the second ejaculate to enter and facilitate the mixing of initially stratified sperm, increasing the potential for sperm mixing and, as such, sperm competition over time (Simmons 2001). Taken together then, it seems that the very morphological and physiological machinery that allows coordinated control of fertilization in parasitoid species such as N. vitripennis is also responsible for the disruption of adaptive sex allocation that occurs when ejaculates overlap during sperm processing prior to fertilization. Therefore, control over fertilization, as required by Nasonia because of the facultative nature of sex allocation under LMC, should strongly select against polyandry if sperm blocking arises with repeated matings (Boulton and Shuker 2015a; Ridley 1988). As such, sperm processing may be a major factor constraining female multiple mating in the parasitoids, a group noted for its rather limited level of polyandry (Ridley 1993; Godfray 1994; Boulton et al. 2014). This constraint may be less important when selection on female-biased sex ratios is relaxed, for instance under laboratory conditions, where LMC is reduced and polyandry is freer to evolve (for N. vitripennis see: van den Assem and Jachmann 1999; Burton-Chellew et al. 2007; Boulton and Shuker 2015b). The findings of the current study show that constrained sex allocation does not occur as an evolutionary consequence of polyandry in N. vitripennis (via postcopulatory sexual selection on males), rather selection on sex allocation may limit polyandry in these wasps, and weakening that selection may be a key contributing factor to the evolution of elevated remating in mass culture conditions. When polyandry does increase, such as under mass culture, postcopulatory sexual selection should favor the evolution of large ejaculates in N. vitripennis males as a defense against sperm competitors (Simmons 2001). Larger ejaculates would then occupy more space in the spermathecal duct, thereby obstructing the ejaculates of rivals for longer (Holmes 1974; van den Assem and Feuth-de-Bruijn 1977). However, as discussed, N. vitripennis females appear unable to process such large ejaculates efficiently, and sperm blocking leads to a reduction in fitness via sex allocation disruption. Thus, the evolutionary interests of males and females are brought into conflict over the size of ejaculate that is transferred. In haplodiploids, sexual conflict is more likely to be resolved in favor of females and so any male adaptation, such as large ejaculate size, that harms females is unlikely to persist (de la Filia et al. 2015). Asymmetric sexual conflict such as this may explain why ejaculates transferred by parasitoid males remain relatively small even in more polyandrous species, as well as why the costs of mating (in terms of fecundity and longevity) tend to be low for females (Boulton et al. 2014). The interpretation of the interaction between sperm use, sperm competition, and polyandry that we present here also suggests a different (but not mutually exclusive) hypothesis for the evolution of the postcopulatory courtship seen in Nasonia and in many other parasitoids (Boulton et al. 2014). Such postcopulatory courtship typically reduces female receptivity to subsequent matings, and so has often been interpreted as an adaptation to reduce future sperm competition for a male (i.e. a defensive trait, and a clear example that postcopulatory sexual selection can shape behavior, physiology, and morphology even without double-matings occurring: Dougherty et al. 2016). However, multiple matings not only risk male paternity share, they also reduce the extent to which sperms are used by females. Thus, postcopulatory courtship that reduces female receptivity not only protects paternity, but also protects efficient sperm use by that female. Because haplodiploid males only pass on their genes via daughters, this might be a nontrivial selective force, to sit alongside that of protecting paternity. To conclude, by interpreting our findings and those of earlier studies in the light of the structure and function of the female parasitoid reproductive tract, we have been able to suggest a mechanism for sperm blocking and to understand how sperm dynamics can result in the patterns of sperm precedence and mixing seen in N. vitripennis and, indeed, across other parasitoids more generally. Herberstein et al (2011) advocate this approach, encouraging behavioral ecologists to refer to early morphological descriptions when interpreting data regarding sperm competition and sperm dynamics. The current study demonstrates that this approach can facilitate the understanding of the processes that lead to patterns of sperm precedence, demonstrating that phenomena that appear likely to result from male ejaculate traits, such as sperm blocking (or “sperm incapacitation”; Simmons et al. 1999; Hosken and Stockley 2004) may relate to female physiological characteristics that, in this case, and perhaps many others, are shaped by natural selection on efficient female sperm use and storage. Rather than immediately looking to the more appealing and enigmatic (and often more contentious) processes, such as cryptic female choice (Birkhead 1998) or “kamikaze sperm” and incapacitating seminal fluid proteins (Baker and Bellis 1988; Harcourt 1991; Price et al. 1999; den Boer et al. 2010), it is important that we understand the basic role that female physiology plays in determining the outcomes of sperm competition. In doing so, we can gain a better appreciation of how postcopulatory sexual selection operates and the types of traits that are likely to result in males and females. SUPPLEMENTARY MATERIAL Supplementary data are available at Behavioral Ecology online. FUNDING This work was supported by a NERC Doctoral Training Grant studentship to R.A.B. We are grateful to Professor Andrew Riches for the use of the 137Cs gamma radiation source. Also, many thanks to Justa Heinen-Kay, Jessie Tanner, Rachel Olzer, and Rebecca Ehrlich and to 2 anonymous reviewers for their insightful comments on an earlier version of this manuscript. Conflict of Interest: The authors declare that they have no conflict of interest to report. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Boulton et al (2017). REFERENCES Arnqvist G, Nilsson T. 2000. The evolution of polyandry: multiple mating and female fitness in insects. Anim Behav . 60: 145– 164. Google Scholar CrossRef Search ADS PubMed  Arnqvist G, Rowe L. 2005. Sexual conflict . Princeton (NJ): Princeton University Press. Google Scholar CrossRef Search ADS   Baer B, Morgan ED, Schmid-Hempel P. 2001. A nonspecific fatty acid within the bumblebee mating plug prevents females from remating. Proc Natl Acad Sci USA . 98: 3926– 3928. Google Scholar CrossRef Search ADS PubMed  Baker RR, Bellis MA. 1988. Kamikaze sperm in mammals. Anim Behav . 36: 936– 939. Google Scholar CrossRef Search ADS   Bates D, Maechler M, Bolker B, Walker S. 2015. Fitting Linear Mixed-Effects Models Using lme4. J Stat Softw . 67: 1– 48. Google Scholar CrossRef Search ADS   Bawab FM, El-Sherief SS. 1989. Contributions to the study of the origin, nature and formation of the plug in the spermatheca of the female crab Portunus pelagicus (Linnaeus, 1766) (Decapoda, Brachyura). Crustaceana . 57: 9– 24. Google Scholar CrossRef Search ADS   Birkhead TR. 1998. Cryptic female choice: criteria for establishing female sperm choice. Evolution . 52: 1212– 1218. Google Scholar CrossRef Search ADS PubMed  Boorman E, Parker GA. 1976. Sperm (ejaculate) competition in Drosophila melanogaster, and the reproductive value of females to males in relation to female age and mating status. Ecol Entomol . 1: 145– 155. Google Scholar CrossRef Search ADS   Boulton RA, Shuker DM. 2013. Polyandry. Curr Biol . 23: R1080– R1081. Google Scholar CrossRef Search ADS PubMed  Boulton RA, Collins LA, Shuker DM. 2014. Beyond sex allocation: the role of mating systems in sexual selection in parasitoid wasps. Biol Rev Camb Philos Soc . 90: 599– 627. Google Scholar CrossRef Search ADS PubMed  Boulton RA, Shuker DM. 2015a. A sex allocation cost to polyandry in a parasitoid wasp. Biol Lett . 11: 4– 7. Google Scholar CrossRef Search ADS   Boulton RA, Shuker DM. 2015b. The costs and benefits of multiple mating in a mostly monandrous wasp. Evolution . 69: 939– 949. Google Scholar CrossRef Search ADS   Boulton RA, Shuker DM. 2016. Polyandry is context dependent but not convenient in a mostly monandrous wasp. Anim Behav . 112: 119– 125. Google Scholar CrossRef Search ADS   Boulton RA, Cook N, Green J, Greenway EV (Ginny), Shuker DM. 2017. Sperm blocking is not a male adaptation to sperm competition in a parasitoid wasp. Dryad Digital Repository . http://datadryad.org/resource/doi:10.5061/dryad.23f0c. Burton-Chellew MN, Koevoets T, Grillenberger BK, Sykes EM, Underwood SL, Bijlsma K, Gadau J, van de Z ande L, Beukeboom LW, West SAet al.   2008. Facultative sex ratio adjustment in natural populations of wasps: cues of local mate competition and the precision of adaptation. Am Nat . 172: 393– 404. Google Scholar CrossRef Search ADS PubMed  Burton-Chellew MN, Beukeboom LW, West SA, Shuker DM. 2007. Laboratory evolution of polyandry in the parasitoid wasp Nasonia vitripennis. Anim Behav . 74: 1147– 1154. Google Scholar CrossRef Search ADS   Daly M. 1978. The cost of mating. Am Nat . 112: 771– 774. Google Scholar CrossRef Search ADS   de la Filia AG, Bain SA, Ross L. 2015. Haplodiploidy and the reproductive ecology of Arthropods. Curr Opin Insect Sci . 9: 36– 43. Google Scholar CrossRef Search ADS   den Boer SP, Baer B, Boomsma JJ. 2010. Seminal fluid mediates ejaculate competition in social insects. Science . 327: 1506– 1509. Google Scholar CrossRef Search ADS PubMed  Dougherty LR, Simmons LW, Shuker DM. 2016. Postcopulatory sexual selection when a female mates once. Anim Behav . 116: 13– 16. Google Scholar CrossRef Search ADS   Finney GL, Flanders SE, Smith HS. 1947. Mass culture of Macrocentrus ancylivorus and its host, the potato tuber moth. Hilgardia . 17: 437– 483. Google Scholar CrossRef Search ADS   Flanders SE. 1956. The mechanisms of sex-ratio regulation in the (parasitic) Hymenoptera. Insectes Soc . 3: 325– 334. Google Scholar CrossRef Search ADS   Garcia‐Gonzalez F, Evans JP. 2011. Fertilization success and the estimation of genetic variance in sperm competitiveness. Evolution . 65: 746– 756. Google Scholar CrossRef Search ADS PubMed  Geuverink E, Gerritsma S, Pannebakker BA, Beukeboom LW. 2009. A role for sexual conflict in the evolution of reproductive traits in Nasonia wasps? Anim Biol . 59: 417– 434. Google Scholar CrossRef Search ADS   Godfray HCJ. 1994. Parasitoids: behavioural and evolutionary ecology . Princeton (NJ): Princeton University Press. Grillenberger BK, Koevoets T, Burton-Chellew MN, Sykes EM, Shuker DM, Van de Zande L, Bijlsma R, Gadau J, Beukeboom LW. 2008. Genetic structure of natural Nasonia vitripennis populations: validating assumptions of sex-ratio theory. Mol Ecol . 17: 2854– 2864. Google Scholar CrossRef Search ADS PubMed  Hamilton WD. 1967. Extraordinary sex ratios. A sex-ratio theory for sex linkage and inbreeding has new implications in cytogenetics and entomology. Science . 156: 477– 488. Google Scholar CrossRef Search ADS PubMed  Harcourt AH. 1991. Sperm competition and the evolution of nonfertilizing sperm in mammals. Evolution . 1: 314– 328. Google Scholar CrossRef Search ADS   Herberstein ME, Schneider JM, Uhl G, Michalik P. 2011. Sperm dynamics in spiders. Behav Ecol . 22: 692– 695. Google Scholar CrossRef Search ADS   Holmes HB. 1974. Patterns of sperm competition in Nasonia vitripennis. Can J Genet Cytol . 16: 789– 795. Google Scholar CrossRef Search ADS PubMed  Hosken DJ, Stockley P. 2004. Sexual selection and genital evolution. Trends Ecol Evol . 19: 87– 93. Google Scholar CrossRef Search ADS PubMed  King PE. 1961. A possible method of sex ratio determination in the parasitic hymenopteran Nasonia vitripennis. Nature . 189: 330– 331. Google Scholar CrossRef Search ADS   King PE. 1962. The structure and action of the spermatheca in Nasonia vitripennis (Walker) (Hymenoptera: Pteromalidae). Proc R Ent Soc Lond (A) . 37: 73– 75 Manier MK, Belote JM, Berben KS, Novikov D, Stuart WT, Pitnick S. 2010. Resolving mechanisms of competitive fertilization success in Drosophila melanogaster. Science . 328: 354– 357. Google Scholar CrossRef Search ADS PubMed  Manier MK, Belote JM, Berben KS, Lüpold S, Ala-Honkola O, Collins WF, Pitnick S. 2013. Rapid diversification of sperm precedence traits and processes among three sibling Drosophila species. Evolution . 67: 2348– 2362. Google Scholar CrossRef Search ADS PubMed  Martel V, Darrouzet E, Boivin G. 2011. Phenotypic plasticity in the reproductive traits of a parasitoid. J Insect Physiol . 57: 682– 687. Google Scholar CrossRef Search ADS PubMed  Martel V, Shuker DM, Boulton RA, Damiens D, Boivin G. 2016. Sex allocation and the evolution of insemination capacity under local mate competition. Entomol Exp Appl . 159: 230– 242. Google Scholar CrossRef Search ADS   Mikheyev AS. 2003. Evidence for mating plugs in the fire ant Solenopsis invicta. Insectes Soc . 50: 401– 402. Google Scholar CrossRef Search ADS   Moore HD, Martin M, Birkhead TR. 1999. No evidence for killer sperm or other selective interactions between human spermatozoa in ejaculates of different males in vitro. Proc Biol Sci . 266: 2343– 2350. Google Scholar CrossRef Search ADS PubMed  Parker GA. 1970. Sperm competition and its evolutionary consequences in the insects. Biol Rev Camb Philos Soc . 45: 525– 567. Google Scholar CrossRef Search ADS   Pizzari T, Wedell N. 2013. The polyandry revolution. Philos Trans R Soc Lond B Biol Sci . 368: 1471– 2970. Google Scholar CrossRef Search ADS   Price CS, Dyer KA, Coyne JA. 1999. Sperm competition between Drosophila males involves both displacement and incapacitation. Nature . 400: 449– 452. Google Scholar CrossRef Search ADS PubMed  Ramadan MM, Wong TTY. 1989. Effect of gamma radiation on Biosteres longicaudatus (Ashmead) (Hymenoptera: Braconidae), a Larval Parasitoid of Dacus dorsalis Hendel (Diptera: Tephritidae). Proc Hawaiian Entomol Soc . 29: 111– 113. RAY DT. 1948. Dominant lethals induced by X-rays in sperm of the chalcidoid wasp Nasonia brevicornis Ashmead [=Mormoniella vitripennis (Walker) fide Muesebeck in lit.]. Biol Bull . 95: 257. Google Scholar PubMed  R Core Team. 2016. R: A language and environment for statistical computing . Vienna: R Foundation for statistical computing. Ridley M. 1988. Mating frequency and fecundity in insects. Biol Rev Camb Philos Soc . 63: 509– 549. Google Scholar CrossRef Search ADS   Ridley M. 1993. Clutch size and mating frequency in parasitic hymenoptera. Am Nat . 142: 893– 910. Google Scholar CrossRef Search ADS   SAUL GB 2nd. 1955. The induction by x-rays of recessive lethals in the mature sperm of Mormoniella vitripennis (Walker). Radiat Res . 2: 447– 460. CrossRef Search ADS PubMed  Shuker DM, West SA. 2004. Information constraints and the precision of adaptation: sex ratio manipulation in wasps. Proc Natl Acad Sci USA . 101: 10363– 10367. Google Scholar CrossRef Search ADS PubMed  Shuker DM, Pen I, Duncan AB, Reece SE, West SA. 2005. Sex ratios under asymmetrical local mate competition: theory and a test with parasitoid wasps. Am Nat . 166: 301– 316. Google Scholar CrossRef Search ADS PubMed  Simmons LW. 2001. Sperm competition and its evolutionary consequences in the insects . Chichester: Princeton University Press.#8232; Simmons LW. 2014. Sperm competition. In: Shuker DM, Simmons LW, editors. The Evolution of Insect Mating systems . London: Oxford University press. p. 181– 203. Google Scholar CrossRef Search ADS   Simmons LW, Parker GA, Stockley P. 1999. Sperm displacement in the yellow dung fly, Scatophaga stercoraria: an investigation of male and female processes. Am Nat . 153: 302– 314. Google Scholar CrossRef Search ADS   Simmons LW, Stockley P, Jackson RL, Parker GA. 1996. Sperm competition or sperm selection: no evidence for female influence over paternity in yellow dung flies Scatophaga stercoraria. Behav Ecol Sociobiol . 38: 199– 206. Google Scholar CrossRef Search ADS   Siva‐Jothy MT, Tsubaki Y. 1994. Sperm competition and sperm precedence in the dragonfly Nanophya pygmaea. Physiol Entomol . 19: 363– 366. Google Scholar CrossRef Search ADS   Slatyer RA, Mautz BS, Backwell PRY, Jennions MD. 2012. Estimating genetic benefits of polyandry from experimental studies: a meta-analysis. Biol Rev Camb Philos Soc . 87: 1– 33. Google Scholar CrossRef Search ADS PubMed  Snook RR, Hosken DJ. 2004. Sperm death and dumping in Drosophila. Nature . 428: 939– 941. Google Scholar CrossRef Search ADS PubMed  Taylor ML, Price TAR, Wedell N. 2014. Polyandry in nature: a global analysis. Trends Ecol Evol . 29: 376– 383. Google Scholar CrossRef Search ADS PubMed  van de Zande L, Ferber S, de Haan A, Beukeboom LW, van Heerwaarden J, Pannebakker BA. 2014. Development of a Nasonia vitripennis outbred laboratory population for genetic analysis. Mol Ecol Resour . 14: 578– 87. Google Scholar CrossRef Search ADS PubMed  van den Assem J, Visser J. 1976. Aspects of sexual receptivity in female Nasonia vitripennis. Biol. Behav. 1:37–56. van den Assem J, Feuth-De Bruijn E. 1977. Second matings and their effect on the sex ratio of the offspring in Nasonia vitripennis (Hymenoptera: Pteromalidae). Entomol Exp Appl . 21: 23– 28. Google Scholar CrossRef Search ADS   van den Assem J, Jachmann F. 1999. Changes in male perseverance in courtship and female readiness to mate in a strain of the parasitic wasp Nasonia vitripennis over a period of 20+ years. Neth J Zool . 49: 125– 137. Google Scholar CrossRef Search ADS   van den Assem J. 1977. Note on the ability to fertilize following insemination (with females of Nasonia vitripennis, Hymenoptera: Chalcidoidea). Neth J Zoo . 27: 230– 235. Google Scholar CrossRef Search ADS   Werren JH. 1980. Sex-ratio adaptations to local mate competition in a parasitic wasp. Science . 208: 1157– 1159. Google Scholar CrossRef Search ADS PubMed  Werren JH. 1983. Sex ratio evolution under local mate competition in a parasitic wasp. Evolution . 37: 116– 124. Google Scholar CrossRef Search ADS PubMed  Whiting AR. 1967. The biology of the parasitic wasp Mormoniella vitripennis [= Nasonia brevicornis](Walker). Q Rev Biol . 42: 333– 406. Google Scholar CrossRef Search ADS   Wilkes A. 1965. Sperm transfer and utilization by the arrhenotokous wasp Dahlbominus fuscipennis (Zett.) (Hymenoptera: Eulophidae). Can Entomol . 97: 647– 657. Google Scholar CrossRef Search ADS   Wishart GJ, Dick LA. 1985. Effect of γ-radiation on fowl sperm function in vitro and in vivo. J Reprod Fertil . 75: 617– 622. Google Scholar CrossRef Search ADS PubMed  Wylie HG. 1976. Interference among females of Nasonia vitripennis (Hymenoptera: Pteromalidae) and its effect on sex ratio of their progeny. Can Entomol . 108: 655– 661. Google Scholar CrossRef Search ADS   © The Author(s) 2017. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Behavioral EcologyOxford University Press

Published: Jan 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off