Sperm competition risk generates phenotypic plasticity in ovum fertilizability

Renée C. Firman, Leigh W. Simmons


Theory predicts that sperm competition will generate sexual conflict that favours increased ovum defences against polyspermy. A recent study on house mice has shown that ovum resistance to fertilization coevolves in response to increased sperm fertilizing capacity. However, the capacity for the female gamete to adjust its fertilizability as a strategic response to sperm competition risk has never, to our knowledge, been studied. We sourced house mice (Mus domesticus) from natural populations that differ in the level of sperm competition and sperm fertilizing capacity, and manipulated the social experience of females during their sexual development to simulate conditions of either a future ‘risk’ or ‘no risk’ of sperm competition. Consistent with coevolutionary predictions, we found lower fertilization rates in ova produced by females from a high sperm competition population compared with ova from a low sperm competition population, indicating that these populations are divergent in the fertilizability of their ova. More importantly, females exposed to a ‘risk’ of sperm competition produced ova that had greater resistance to fertilization than ova produced by females reared in an environment with ‘no risk’. Consequently, we show that variation in sperm competition risk during development generates phenotypic plasticity in ova fertilizability, which allows females to prepare for prevailing conditions during their reproductive life.

1. Introduction

Studies of postcopulatory sexual selection have focused on the evolutionary and facultative responses of the ejaculate to sperm competition, and mechanisms by which females influence the transfer and storage of sperm [1,2]. Much less attention has been given to the evolutionary consequences of postcopulatory sexual selection for ovum form and function.

Selection upon the ejaculate occurs when the sperm of rival males are forced to compete for fertilizations [3]. Male evolutionary adaptations to sperm competition have been well documented, and studies of experimental evolution have shown that sperm competition selects for ejaculates with high sperm densities and enhanced competitive ability [46]. Indeed, competitive conditions are expected to favour sperm that rapidly penetrate the ovum. However, a detrimental outcome of increased sperm fertilizing capacity is an elevation in the frequency of polyspermy [7]. Although the frequency of polyspermy in natural fertilizations has received little attention, it has been shown to be present even when sperm are limited [8,9]. As a response to this fatal threat, ova are expected to counteradapt by decreasing their fertilizability (ova defensiveness) [10], leading to a coevolutionary arms race between the gametes [11]. Evolutionary responses in ovum resistance to fertilization have been explored both among and within Mus species. A comparative study of three species of Mus that had evolved under different levels of sperm competition reported asymmetries in interspecific fertilization rates and provided evidence that sexual conflict at the gametic level has the potential to promote prezygotic reproductive barriers [12]. Indeed, ovum defences could account for faster conspecific fertilization rates, compared with heterospecific rates, in the naturally hybridizing species Mus domesticus and Mus musculus [13]. More recently, a study of experimental evolution has shown that female house mice (M. domesticus) that had evolved under a polyandrous mating regime produced ova that were more resistant to fertilization compared with females that had evolved under a monogamous regime [14]. These investigations have provided compelling evidence for the evolution of ovum resistance to fertilization in response to sperm competition.

Soliciting copulations from multiple males can also be beneficial to females by allowing them to ‘choose’ between males in the postcopulatory arena; ‘cryptic female choice’ [1]. Although isolating the subtle processes of prezygotic sperm choice from sperm competition is difficult [15], sperm selection is expected to generate a fitness advantage by enabling females to skew paternity toward males of high genetic quality [1,15]. Consequently, females may use mechanisms of ovum defence to filter male quality and ensure that only the highest quality sperm penetrate the outer coat of the ovum, the zona pellucida (Zp), and fuse with the nucleus.

Phenotypic plasticity is defined as the capacity of a single genotype to exhibit variable phenotypes in different environments [16]. Phenotypic plasticity is ubiquitous in nature and has been demonstrated to occur in a large number of traits across a diverse range of taxa (reviewed in [17,18]). Anticipatory plastic responses have been shown to be stimulated by an individual's interaction with cues that reflect future environmental or social conditions [18]. Plastic responses are usually highly adaptive and may influence an individual's survival [1921] and/or reproductive success [2225]. For example, social cues that are indicative of an elevated risk of sperm competition have been shown to generate adaptive responses in the ejaculate. When presented with a risk of sperm competition, males have been shown to make both immediate adjustments in the number of ejaculated sperm [22,2629], as well as long-term, developmental changes in sperm production [24,3032]. There is also evidence of phenotypic plasticity in sperm morphology. Thus, studies have shown that sperm size can be adjusted under different social scenarios [3335]. The influence of the social environment on the ovum phenotype has not before been investigated. In this study, we addressed whether female house mice (M. domesticus) can phenotypically adjust the fertilizability of their ova to variation in the sperm competition environment.

Population density varies considerably in natural house mouse populations and has been shown to be predictive of the level of sperm competition as measured by the incidence of multiply sired litters [36,37]. We sourced house mice from two island populations that have rates of multiply sired litters that represent a low (17% multiple paternity) and a high (71% multiple paternity) level of sperm competition [37]. As predicted by sperm competition theory, males from the high sperm competition population produce more sperm and better quality sperm compared with males from the low sperm competition population [32]. Moreover, males from these populations were found to increase the fertilizing potential of their sperm when reared under a risk of sperm competition [32]. Here, we exposed maturing females to different sperm competition environments by creating variation in their perception of male density. Olfaction is the primary sensory modality in house mice, and conspecifics are recognized by individually distinct scent signals [38]. We therefore manipulated the social experience of developing females via exposure to male odours and direct encounters with conspecifics, and thus created local environments in which females perceived either a ‘risk’ or ‘no risk’ of sperm competition. We retrieved ova of females from the low and high sperm competition populations that had matured under either a ‘risk’ or ‘no risk’ of sperm competition, and quantified ovum resistance to fertilization by using in vitro fertilization (IVF) techniques. Our IVF assays produced a pattern of fertilization rates among females from the different populations that is consistent with the theoretical prediction that sperm competition should generate divergence in ovum defensiveness via sexual conflict at the gametic level [11]. More importantly, we found that females exposed to a ‘risk’ of sperm competition produced more defensive ova than females that were reared under ‘no risk’ of sperm competition. Thus, our investigation provides evidence of phenotypic plasticity in a female trait in response to the risk of sperm competition.

2. Material and methods

(a) Experimental animals

Sexually mature male (20) and female (20) house mice were trapped on Whitlock Island (30°19′ S, 114°59′ E) and Rat Island (28°42′ S, 113°47′ E) located off the coast of Western Australia, and brought to the University of Western Australia. We selected these two populations because they are known to differ in the frequency of multiple paternity [37]. In the Whitlock Island population, 17% of litters are of mixed paternity, whereas in the Rat Island population 71% of litters are of mixed paternity. Associated with this difference in the level of sperm competition, males on Rat Island have relatively larger testes than males on Whitlock Island [37]. The mice were bred under common-environment conditions for four generations. Mice were maintained in a constant temperature room (CTR; 24°C) on a reversed light–dark cycle, and food and water were provided ad libitum. During breeding, the females were paired with a single male until they were seen to be pregnant upon which males were removed. Mice of the fifth generation were used as focal individuals in this experiment.

(b) Social environment manipulation

House mice can recognize and identify conspecifics via individual scent signals [38]. Specifically, male scent has been shown to influence female reproductive development [39] and function [40,41]. Thus, we exposed developing females to the soiled chaff of one to 10 males and created variation in their perception of male density, and thus the sperm competition environment. From the time that they were weaned (three weeks old) until they reached sexual maturity, ‘risk’ females were maintained in a CTR that also housed caged males (see the electronic supplementary material, figure S1). These females received soiled chaff from multiple males and ‘encountered’ sexually mature males. Thus, each week ‘risk’ females received approximately 15 g of a mixture of soiled chaff that consisted of 30 g of bedding from each of the boxes of 10 sexually mature males. ‘Encounters’ were conducted once a fortnight. The focal females were placed in large, opaque plastic tubs (49 × 74 × 41 cm) containing two individually caged mature males. For 30 min, the females roamed freely inside the tub, which allowed them to receive olfactory, auditory and visual (but not tactile) stimuli from the encounter animals. Each time the focal females encountered two different males.

By contrast, ‘no risk’ females were maintained in a CTR that contained only females. Focal females received soiled chaff from a single male and they ‘encountered’ sexually mature females. Once every two weeks, females in the ‘no risk’ treatment received soiled chaff (15 g) from the same mature male that was housed in a different CTR. Every other week ‘no risk’ female chaff (15 g) was transferred from the back to the front of their own box. During encounters, ‘no risk’ females were exposed to two mature females. The ‘risk’ and ‘no risk’ individuals were switched between CTRs after four weeks of social manipulation.

The IVF assays commenced after eight weeks of social manipulation. One replicate assay for each of the four treatment groups was performed each week for four weeks. Consequently, the age of the females at the time of ova donation ranged from 11 to 14 weeks. The females received the social manipulation regime until they were sacrificed for ova donation. Ova were mixed with the sperm of males from the same population that had been reared under standard colony conditions. A total of 16 IVF assays were performed.

(c) In vitro fertilization

The protocol for the IVF assays has been described previously [12,42]. Briefly, adult females were induced to superovulate with 5IU pregnant mare's serum gonadotropin and 5IU human chorionic gonadotropin that were administered 48-h apart. The female reproductive tracts were removed and cumulus–oocyte complexes (COCs) were released from the ampullae. Epididymides were removed from males and placed in a 1 ml drop of human tubal fluid (HTF) and incubated to allow sperm to disperse into the medium [12,42]. Following an initial 10 min incubation, sperm concentration and sperm motility was quantified using a CEROS computer-assisted sperm analysis system (v. 10, Hamilton and Thorne Research) using standard mouse parameters. On average, the sperm traits did not differ significantly among the small sample of males used in the IVF assays (see the electronic supplementary material, tables S1 and S2).

The IVF assays were performed in 500 µl drops of HTF under mineral oil. The COCs from four females of the population were pooled in a single assay. Previous IVF experiments both within and between Mus species produced higher fertilization rates of conspecific ova–sperm combinations compared with heterospecific combinations [12], and trial assays performed on wild-derived, laboratory populations of mice at sperm concentrations of 2.0 × 106 motile sperm ml−1 provided evidence of a ceiling effect with fertilization rates close to 100%. Thus, to ensure that higher fertilization rates across the different treatments would be detectable, the IVF assays were mixed with a final concentration of 1.5 × 106 motile sperm ml−1. The gametes were coincubated for 4 h at 37°C under 5% CO2/air, which allowed the sperm to penetrate the ova, fuse with the oolema, decondense and form a pronucleus [12,42]. After the incubation period, the ova were washed in 100 µl drops of HTF to remove remaining cumulus cells and/or attached sperm. The oocytes where then stained with a DNA fluorochrome (Hoechst 33342) and viewed under a Zeiss Axio Imager.A1 fluorescent microscope. Fertilization rate was determined as the number of ova with stained pronuclei over the total number of mature ova. The number of ova scored and the proportion of ova fertilized in each IVF assay are presented in the electronic supplementary material, table S3.

3. Results

We analysed IVF rates among the treatment groups using a generalized linear mixed model (GLMM) fit by the Laplace approximation using the ‘lme4’ library in the R-statistical analysis package [43], which is one of the recommended maximum-likelihood estimation methods for generalized mixed-effects models [44]. Fixed effects in the GLMM were population, treatment and their interaction, while assay ID was modelled as a random effect. The interaction term was non-significant (p = 0.976), and therefore removed from the model (table 1). We found that females which had matured under a perceived risk of sperm competition had increased ova defensiveness (38% of ova fertilized) compared with females that had matured in a social environment with no risk of sperm competition (73% of ova fertilized; table 1 and figure 1). Our statistical analysis also revealed that there was a significant difference between populations in fertilization rate (table 1 and figure 1). Under identical fertilization conditions, we found reduced IVF rates in ova derived from the high sperm competition population (44% of ova fertilized) compared with the low sperm competition population (67% of ova fertilized).

View this table:
Table 1.

Fertilization rates of ova of female house mice from two populations that matured under a ‘risk’ or ‘no risk’ of sperm competition. (A GLMM fit by the Laplace approximation (R) revealed that both population and sperm competition risk (treatment) had a significant effect on IVF rates in house mice.)

Figure 1.

Phenotypic plasticity in ova fertilizability. Fertilization rates of ova donated by females from a low-level or high-level sperm competition (SC) population that matured under a ‘risk’ or ‘no risk’ of sperm competition. (Online version in colour.)

4. Discussion

Phenotypic plasticity is widespread in nature, and plastic responses have been shown to include changes in behavioural, physiological, morphological and life-history traits [17,18]. Individuals may be sensitive to cues or stimuli that reflect future social conditions, and thus make ontogenic adjustments to enhance specific traits that will be adaptive during their reproductive life [18]. Phenotypic plasticity in a female trait in response to sperm competition has not before, to our knowledge, been studied. Here, we used IVF to explore whether females had the capacity to respond to increased fertilization effort among males by increasing their resistance to fertilization. We sourced house mice from two natural populations that differed in the strength of selection via sperm competition and tested whether there were adjustments in ovum fertilizability in response to cues that corresponded to different sperm competition environments. We found that females exposed to a ‘risk’ of sperm competition produced more defensive ova compared with the ova of females that were reared under ‘no risk’ of sperm competition. We thereby provide, to our knowledge, the first demonstration of phenotypic plasticity in the female gamete in response to the sperm competition environment.

Sexual conflict over the postmating interests of males and females is widespread in nature [45] and arises at the gametic level when the sperm of rival males compete for fertilizations [11]. Selection for increased sperm fertilizing potential is predicted to generate an elevation in the frequency of the fatal condition of polyspermy [7]. Sexual conflict theory proposes that an increase in male sperm competitiveness will provoke a counteradaptation in females of increased ova resistance to fertilization [11]. Until recently, empirical support for this sexual conflict paradigm was limited to just two studies. A comparative study on free-spawning invertebrates reported a correlation between sperm density (a proxy for the strength of sperm competition) and ova resistance to fertilization [10]. Similarly, ova fertilizability was found to be correlated with relative testes size among species of Mus [12]. The first direct line of evidence that sperm competition can generate sexual conflict at the gametic level, and create asymmetries in fertilization rates among populations, comes from a study of experimental evolution on house mice (M. domesticus) [14]. Females that had evolved with sperm competition were found to produce ova with increased resistance to fertilization compared with the ova of females that had evolved under a monogamous regime [14]. The natural mouse populations that we used in this study are likely to differ in many respects other than the level of sperm competition. However, it is interesting to note that the pattern of divergence in IVF rates observed between these two populations of mice is consistent with the theoretical prediction that sperm competition should generate sexual conflict at the gametic level, leading to an evolutionary divergence in ova defensiveness among populations and species [11].

A number of theoretical models have been developed to predict how males will strategically adjust the ejaculate in response to sperm competition risk [46,47], and empirical studies of rodents have provided evidence for phenotypic plasticity in both sperm production [24,32] and expenditure [28]. Previously, we found that males from our high sperm competition population produced more sperm and better quality sperm than did males from the low sperm competition population [32]. Moreover, males reared under a perceived risk of sperm competition produced more aggressive sperm compared with males reared under no risk [32]. Adaptive adjustments in the ejaculate would enhance male fitness in competitive situations [24,29] but are also likely to lead to an increased risk of polyspermic fertilizations [10]. In the rat Rattus norvegicus, increased numbers of sperm at the fertilization site will generate elevated rates of polyspermy [48]. For mice (M. musculus), it is known that polyspermy does occur under natural mating conditions (approx. 2%) [49], however the relationship between the number/quality of naturally inseminated sperm and the rate of polyspermy is yet to be determined. Although IVF conditions are likely to produce inflated rates of polyspermy compared with natural conditions, it has been shown that the in vitro rate of polyspermy increases as sperm density increases [50]. Consequently, we might expect females to adjust the fertilizability of their ova in accordance with expected sperm densities at the site of fertilization.

We documented variation in ova fertilizability between females that were exposed to different perceived male densities during development and for the first time, to our knowledge, provide evidence of phenotypic plasticity in a female trait in response to the sperm competition environment. It could be argued that the observed result is attributable to sperm limitation in the ‘no risk’ treatment and the production of more fertilizable ova. However, it is important to note that the IVF rates in the ‘no risk’ treatment were comparable to those that were obtained during trial assays when animals were reared under standard colony conditions (approx. 80%) [14]. Consequently, the rates of fertilization in the ‘risk’ treatment are representative of a reduction in fertilization success and show that female house mice respond to an elevated risk of sperm competition by increasing their ova resistance to fertilization. The response of increased ovum defences among females parallels our finding that males from these populations increase their fertilization efficiency when reared under a perceived risk of sperm competition [32]. Thus, in house mice we have found that both male (sperm) and female (ova) responses to sperm competition risk align with the sexual conflict hypothesis [11].

An alternative though not mutually exclusive interpretation of our data, is that females may moderate ova fertilizability under different risks of sperm competition as a mechanism of cryptic female choice [1]. When male density is high, and females have the opportunity to mate with multiple partners, they may benefit from elevating ova resistance to fertilization to ensure that only the best quality sperm, with the greatest fertilizing potential, are able to penetrate the Zp. We are unable to separate processes of sperm selection from sperm competition in our study. Future research will endeavour to determine whether cryptic female choice at the gametic level selects for enhanced offspring fitness. Regardless of the process(es) accounting for the observed pattern of fertilization, we have shown that females are sensitive to cues of sperm competition risk and have the ability to adjust the degree of ovum resistance to fertilization accordingly. Indeed, phenotypic plasticity in ova defensiveness during development would allow females to prepare for conditions during their reproductive life and potentially reduce the likelihood of polyspermy in an environment where the risk of sperm competition is high.

The plastic responses we have documented in ova fertilizability could be explained by physical, biochemical and/or molecular mechanisms of the ovum and its associated extracellular matrix. The cumulus oophorus is a group of closely associated granulosa cumulus cells that support the maturing oocyte prior to ovulation [51]. Following ovulation, the cumulus oophorus surrounds the oocyte as a cloud of cumulus cells, which facilitates the transfer of the oocyte into the oviduct [52] and participates in the passage of sperm to the surface of the oocyte, the Zp [53]. Sperm are required to move through the extracellular matrix of the cumulus oophorus to achieve successful fertilization. It has been suggested that the cumulus oophorus attracts sperm via chemotaxis [54] and directs the few sperm that reach the ampulla to the immediate vicinity of the oocyte [55]. Thus, the cumulus oophorus has an important role in controlling the number of sperm that interact with the Zp. Consequently, variation in the number or density of cumulus cells around the oocyte may account for the plastic response in ovum defences that we have observed in house mice. A greater concentration of cumulus cells, or more compacted cumuli, around the oocyte may create a physical barrier that inhibits the sperms’ passage to the oocyte. In this case, it could be that females exposed to a risk of sperm competition made strategic adjustments in the density of cumuli that surrounded the ooctyes, restricting the ease at which the sperm reach the Zp, and thus increasing ovum resistance to fertilization. Alternatively, considering that the oophurus physiochemical environment controls the number of sperm that are attracted to the Zp [54], it might be that fewer cumulus cells around the oocytes resulted in reduced fertilization rates of females reared under a risk of sperm competition.

Following successful fertilization, cortical granules (membrane bound organelles located in the cortex of unfertilized oocytes) undergo exocytosis to release their contents into the perivitelline space [56]. The released cortical granule proteins are responsible for blocking polyspermy by modifying the Zp [57]. Although other biochemical and structural modifications of the Zp following cortical granule exocytosis have been inferred, only a change in a Zp glycoprotein (Zp2) has been experimentally described [58]. The ‘zona hardening’ inhibits additional sperm from binding to the Zp and prevents already bound sperm from penetrating the ovum [59,60]. It is possible therefore, that females might prepare for a future risk of sperm competition by increasing the speed or efficiency of the Zp's block. In this case, variation in the expression of the Zp genes may account for differences in ova resistance to fertilization, offering an alternative potential mechanism for strategic adjustments in ovum defences of females reared in different social environments. The mechanism/s by which mouse ova regulate the establishment of the membrane block is largely unknown, but has been suggested to be the culmination of multiple postfertilization events that together modify the ovum's receptivity to sperm and is established in approximately the same time frame as the Zp block [61]. Certainly, responses in the cumulus oophorus, the Zp and the plasma membrane that account for plasticity in ova fertilizability need not be mutually exclusive. In order to gain insights into these potential mechanisms of ova defensiveness, our future research will explore expression levels in the Zp genes and compare fertilization rates of cumulus-enclosed versus cumulus-denuded oocytes among females that have prepared for different sperm competition environments.

In conclusion, it is well known that males show phenotypic plasticity in ejaculate investment in response to their perception of sperm competition risk, and it has long been recognized that male responses to sperm competition can be costly for females, for example, if overly aggressive ejaculates result in polyspermy and ova mortality. Here, we provide, to our knowledge, the very first account of adaptive plasticity in a female trait in response to the risk of sperm competition. We suggest that strategic adjustments in the physical microenvironment of the cumulus oophorus and/or variation in the expression of glycoprotein/s associated with the Zp genes could account for variation in ova fertilizability. Importantly, this research has shown that variation in sperm competition risk during development generates plastic responses in the ovum that would allow females to prepare for prevailing conditions during their reproductive life.

This research was approved by the UWA animal ethics committee (07/100/607).

Funding statement

This research was funded by the Australian Research Council.


We are grateful to E. Roldan and M. Gomendio from the National Museum of Natural Sciences (Spain) for IVF training. We also thank T. Friend, T. Button, S. Stankowski and C. McHarrie for assistance with field collections; F. Simmons and S. Lobind for assistance with the social manipulation regime and animal husbandry; and B. Buzatto for statistical assistance.

  • Received August 13, 2013.
  • Accepted September 19, 2013.


View Abstract