Renée C. Firman, Centre for Evolutionary Biology, School of Animal Biology (M092), University of Western Australia, 35 Stirling Hwy, Crawley, Western Australia 6009, Australia. Tel.: +61 8 6488 2699; fax: +61 8 6488 1029; e-mail: email@example.com
Sperm show a remarkable degree of variation in size, shape and complexity. Murine rodents exhibit a sperm head morphology that differs greatly from the ovoid shape that is characteristic of most mammals. These rodents have sperm that bear one or more apical hooks, the function of which is currently surrounded by much controversy. It has been suggested that the hook serves to facilitate the formation of sperm groups, which in some species exhibit relatively faster velocities than single cells and thus, may provide an advantage when ejaculates are competing for fertilisations. In support of this hypothesis, a comparative study reported a positive association between the strength of sperm competition (estimated from testes size) and the curvature of the sperm hook amongst 37 murine species. Here, we assessed whether sperm competition influences sperm hookedness at the intra-specific level. Following 16 generations of selection, we used geometric morphometry (GM) to describe sperm head morphology in selection lines of house mice evolving with (polygamous) and without (monogamous) sperm competition. Although the GM analysis returned two relative warps that described variation in the curvature of the sperm hook, we found no evidence of divergence between the selection lines. Thus, we can conclude that sperm competition does not influence the degree of sperm hookedness in house mice.
Theory assumes that sperm competition conforms to the raffle principal and predicts that males should increase their expenditure on sperm with increasing strength of selection (Parker, 1990). Indeed, across different invertebrate taxa, testes size and sperm number have been shown to diverge between populations that have been forced to mate monogamously and those left to mate under a natural polygamous regime (Hosken & Ward, 2001; Pitnick et al., 2001; Simmons & García-González, 2008). For vertebrate species, sperm velocity is an important determinant of fertilization success (Gage et al., 2004; Malo et al., 2005) and there is an increasing amount of research that shows that sperm competition can influence the evolution of sperm swimming performance (Fitzpatrick et al., 2009). For example, among selection lines of house mice, males evolving with sperm competition produced ejaculates with more sperm, faster sperm and greater proportions of motile sperm, compared to males evolving without sperm competition (Firman & Simmons, 2010a).
It is less clear what role sperm competition plays in the evolution of sperm morphology. Among mammals, it is assumed that flagellum length is positively correlated with sperm swimming speed and therefore sperm competition promotes the evolution of longer sperm (Gomendio & Roldan, 1991). Indeed, mammalian sperm show a large degree of variation, not only in size (Cummins & Woodall, 1985) but also in the shape and complexity of the sperm head (e.g. see Breed, 2004; Meisner et al., 2005). However, the macroevolutionary association between sperm competition and sperm length is not always evident at the species level (Schulte-Hostedde & Millar, 2004; Malo et al., 2006; Firman & Simmons, 2010b). Thus, it has been suggested that evolution of sperm size in mammals might be influenced by the need to overcome hydrodynamic drag associated with the complexity of the sperm head (Firman & Simmons, 2010b).
Many rodents belonging to the Murinae sub-family possess a falciform sperm head with one or more apical hook/s. The functional significance of the curved, apical hook is currently not known, but it has been suggested to facilitate the formation of motile sperm groups. In some murine species, grouped sperm have been shown to have faster swimming velocities than individual sperm (Apodemus sylvaticus, Moore et al., 2002; Rattus norvegicus, Immler et al., 2007; Peromyscus maniculatus, Fisher & Hoekstra, 2010), which has lead to the assumption that sperm groups may be advantageous in sperm competition (Moore et al., 2002). Consequently, it was suggested that the apical sperm hook may be an evolutionary product of sperm competition (Moore et al., 2002). Support for this hypothesis comes from a comparative study that reported a positive correlation between testes size, a proxy for the strength of selection under sperm competition, and curvature of the sperm hook (Immler et al., 2007). To date, empirical tests of this hypothesis are limited. Here, we utilized established monogamous and polygamous selection lines of house mice (Firman & Simmons, 2010a) to explore how sperm competition influences sperm morphology and more specifically, whether sperm competition selects for increased sperm hookedness in this murine rodent.
Materials and methods
Source population and experimental evolution
A detailed description of the source population and the mating design for the selection regimes are provided elsewhere (Firman & Simmons, 2010a). Briefly, the selection lines were established from a colony of wild-derived mice held at the Animal Resources Centre (Murdoch, WA) (Firman & Simmons, 2010a). We established four monogamous and four polygamous lines with 18 males and 18 females in each. Subsequently, 18 males and 18 females contributed to each generation. That is, every fecund female contributed a son and daughter to the next generation. All selection on juvenile fitness was removed by providing food ad libitum and separate housing to gestating and nursing females. In the monogamous lines, we eliminated almost all selection on adult fitness by ensuring that every male and female pair contributed one son and one daughter to the next generation (Shabalina et al., 1997). Although greatly relaxed, selection may not have been completely eliminated because sometimes, a pair did not mate or produce offspring. In the polygamous lines, adult females had equal fitness (two offspring) and adult males had equal mating success but not equal fertilization success due to sperm competition and/or cryptic female choice. Thus, natural selection was diminished in both selection regimes and only postcopulatory sexual selection on males operated in the polygamous lines. Following eight generations of experimental evolution, we performed sperm assays on males from the selection lines and observed divergence in sperm number and quality. Males from the polygamous lines produced more sperm, had a greater proportion of motile sperm and sperm with greater swimming velocities than males from the monogamous lines, as would be expected following selection via sperm competition (see Firman & Simmons, 2010a). We used a randomly selected sample of three monogamous and three polygamous lines for the current study.
Laboratory selection lines are typically small, insular populations, the genetic diversity of which is expected to decrease across multiple generations of selection. As such, inbreeding depression can have detrimental consequences for the fitness and persistence of laboratory lines as it accrues across successive generations (although see Snook et al., 2009). Spermatogenesis is a complex process requiring precise genetic and physiological control (Gage et al., 2006). Processes that disrupt developmental pathways, such as inbreeding, may affect the production of morphologically normal sperm (Gage et al., 2006; Fitzpatrick & Evans, 2009). Postcopulatory sexual selection has been suggested to reduce the detrimental effects of inbreeding by the removal of deleterious mutations (Radwan et al., 2004). In our study of experimental evolution, postcopulatory sexual selection operated in the polygamous lines, but not in the monogamous lines. To ensure that sperm hook morphology in the monogamous and polygamous selection lines was not influenced by differential levels of inbreeding, and to expose any potential effects of inbreeding on sperm morphology in general, we included consanguineous matings in our experimental design. We used animals from generation 16 of our selection lines. At 3 weeks of age, 10 males and 10 females from three monogamous and three polygamous lines were separated from their siblings and housed individually until they reached sexual maturity. We established five full-sibling mating pairs and five nonsibling mating pairs for each selection line and produced inbred and outbred offspring respectively (Fig. 1).
Once paired, females were checked daily for the presence of a mating plug, which is indicative of a successful mating (Rugh, 1968). Once a mating plug was observed, females were transferred to a clean box and provided with shredded paper for nesting. Those pairs that did not mate within 7 days were separated, rested for 7 days, and then re-paired. If these pairs failed to mate within another 7 days, they were replaced with different animals.
Nests were checked for pups 19 days from mating. Pups were weaned from their mother 21 days after birth, at which time males were placed in individual boxes. Randomly chosen female offspring were housed in groups (3/box) and maintained in the same room as the maturing males. All animals were maintained in a room under natural light/dark cycle and with the temperature regulated at approximately 20 °C. Food and water was provided ad libitum.
The sperm quality assays were conducted as previously detailed (Firman & Simmons, 2010a). At 11 weeks of age, males were euthanased by lethal injection. The caudal region of each epididymis was repeatedly cut with fine scissors and placed in 1 mL of culture medium (Murase & Roldan, 1996). The sperm solution was incubated at 37 °C for 90 min. Following this step, a 10 μL aliquot of sperm was loaded into a haemocytometer, viewed at ×100 magnification, and scanned for motility parameters. Five scans per sample were conducted. We used standard mouse protocols for Ceros computer assisted sperm analysis (v10; Hamilton and Thorne Research, Beverley, MA, USA), which provided measures of (i) sperm concentration (×106 mL−1), (ii) percentage motile, progressive and rapid sperm and (iii) average path (VAP), straight line and curvilinear (VCL) velocities (for more details see Firman & Simmons, 2010a).
Sperm head morphology
Following activation, 100 μL of the sperm solution was fixed in a 4% formaldehyde solution. Sperm smears were prepared on slides and stained with Coomassie brilliant blue (Firman & Simmons, 2010a). The slides were viewed under oil with a ×63 objective using an AxioImager.A1 microscope (Zeiss, Germany). Images of 10 sperm/sample were taken with AxioCam MRc5 (Zeiss), using the imaging program AxioVision 4.7. Geometric morphometric (GM) analysis was then used to quantify variation in sperm head size and shape (Zelditch et al., 2004; Firman & Simmons, 2009). We placed 13 landmarks around the perimeter of the sperm heads using tpsDig 2w32 (http://life.bio.sunysb.edu/morph; F. James Rolpf, Department of Ecology and Evolution, Stony Brook University, Stony Brook, NY), four of which were assigned as fixed landmarks and nine that were assigned as semi-sliders (Firman & Simmons, 2009). We used the software package tpsrelww32 to generate a centroid size score and relative warp scores, which described variation in sperm head morphology.
Due to four complete reproductive failures, we obtained data from a total of 56 families, two of which consisted of only female offspring. Thus, to assess differences in sperm quality we used 13 inbred and 14 outbred litters from the monogamous lines, and 12 inbred and 15 outbred litters from the polygamous lines. Family averages were calculated for sibling males and used in the analyses. To account for the fact that families from replicate lines were not statistically independent, we used nested anovas with replicate selection line nested within selection history. All means are presented ± 1 SE.
Preliminary anovas revealed that there was no effect of breeding treatment across any of the traits that were measured (Table S1). That is, there were no differences in sperm quality scores (Table S2), or the size and shape of the sperm head (Table S3) between the inbred and outbred offspring. Thus, we pooled data across the breeding treatments and included only selection history and replicate line as terms in the subsequent analyses.
To summarize the variation among the seven highly correlated sperm traits, we performed a principal components analysis (PCA). The PCA produced two components with eigenvalues > 1 (Table 1). The first principal component (PC1) was loaded most heavily by the velocity traits and to a lesser extent by motility traits. The second principal component (PC2) was loaded most heavily by percentage motile sperm and percentage progressive sperm, and to a lesser extent by percentage rapid sperm, average path velocity, straight line velocity, curvilinear velocity and sperm concentration (Table 1). There was a significant effect of selection history on PC2 (Table 2). Thus, males from the polygamous lines typically had larger PC2 scores (0.60 ± 0.3) than males from the monogamous lines (−0.62 ± 0.2), which corresponded to higher sperm numbers and greater proportions of motile, rapid and progressive sperm and faster swimming sperm (Tables 1–3).
Table 1. Means (± 1SE) of sperm traits of house mice from monogamous and polygamous selection lines.
Sperm number (×106 mL−1)
5.8 ± 0.7
7.2 ± 0.6
% motile sperm
26.6 ± 2.0
31.6 ± 3.0
% progressive sperm
5.6 ± 1.5
6.1 ± 1.4
% rapid sperm
13.5 ± 1.5
15.2 ± 1.4
VAP (μm s−1)
51.7 ± 2.5
65.1 ± 2.9
VSL (μm s−1)
34.9 ± 1.7
45.6 ± 2.0
VCL (μm s−1)
106.1 ± 6.0
120.3 ± 4.0
Table 2. Principal component analysis of sperm traits in house mice.
% motile sperm
% progressive sperm
% rapid sperm
Proportion of total
Table 3. anova of sperm quality scores (PC1, PC2) of house mice from monogamous and polygamous selection lines. We used anovas with replicate selection line nested within selection history as a random effect.
Significant values (P < 0.05) are given in bold.
Sperm head morphology
The GM analysis returned centroid size scores and relative warps that described variation in sperm head morphology. The first relative warp (RW1) had an eigenvalue of 1.27 and explained approximately 34% of the variation. RW1 described variation in the sperm head hook, with the most negative RW1 values representing a more flattened hook shape and the most positive RW1 values representing a more reflected hook shape (Fig. 2). The second relative warp (RW2), which had an eigenvalue of 1.02 and explained a further 21% of the variation, also described differences in sperm hook shape. Negative RW2 values described a narrowing of the sperm head and a shorter hook, whereas positive RW2 values represented a widening of the head and a longer sperm hook (Fig. 2). Nested anovas revealed that there was no effect of selection history on sperm head size (centroid scores) or sperm head shape (RW1, RW2) (Table 4).
Table 4. anova of centroid size and sperm head morphology scores (RW1, RW2) house mice from monogamous and polygamous selection lines. We used anovas with replicate selection line nested within selection history as a random effect.
Significant values (P < 0.05) are given in bold.
Previously, we reinstated sperm competition in a mouse population that had a long history of enforced monogamy and observed divergence in male and female fitness (Firman & Simmons, 2010a). Following eight generations of selection, we found that males from the polygamous lines evolved more sperm with better motility compared to males evolving under a monogamous regime. Here, at generation 16, we replicated these findings. Males from the polygamous lines had more sperm with significantly better sperm motility/velocity compared to males from the monogamous lines. It is now widely accepted that sperm competition selects for increased testes size and/or efficiency and high sperm numbers (Hosken et al., 2001; Pitnick et al., 2001) and there is an increasing amount of research showing that sperm competition can influence the evolution of sperm performance and sperm morphology (Stockley et al., 1997; Byrne et al., 2003; Ramm et al., 2005; Gomendio et al., 2006; Immler et al., 2007). We have provided evidence that sperm competition is a potent evolutionary force that selects for higher sperm numbers and improved sperm quality in house mice.
Due to the complex process of spermatogenesis, it has been suggested that sperm may be highly susceptible to inbreeding depression (Gage et al., 2006). Here, we found no effects of inbreeding depression on the sperm traits we measured. Therefore, we can conclude that a single generation of inbreeding does not influence sperm number, sperm quality, or sperm morphology in house mice. Indeed, current evidence for mammal species suggests that inbreeding needs to reach an extreme level to impair sperm quality (Gomendio et al., 2000; Fitzpatrick & Evans, 2009).
The adaptive significance of the murine rodent sperm hook is currently a controversial topic amongst evolutionary sperm biologists (Moore et al., 2002; Immler et al., 2007; Firman & Simmons, 2009; Fisher & Hoekstra, 2010). It has been proposed that the sperm hook may facilitate the formation of sperm groups, which have been shown to exhibit faster swimming velocities than single cells in both polygamous (Apodemus sylvaticus,Moore et al., 2002; Rattus norvegicus, Immler et al., 2007; Peromyscus maniculatus, Fisher & Hoekstra, 2010) and monogamous (Peromyscus polionotus, Fisher & Hoekstra, 2010) murine rodents. A comparative study reported a positive correlation between the degree of sperm hookedness and relative testes size (Immler et al., 2007). Here, we used monogamous and polygamous selection lines of house mice that had been evolving for 16 generations to test whether the degree of sperm hookedness diverges under selection from sperm competition. Our analysis revealed that selection history did not explain any of the variation in either of the two relative warps that described sperm hook shape. Thus, although mice sperm can form motile groups (Immler et al., 2007; Firman & Simmons, 2009), the shape of the hook had not diverged between populations evolving with and without sperm competition. This result is consistent with our previous finding that sperm competition risk did not correlate with the degree of sperm hookedness among island populations of house mice that differ in the strength of selection from sperm competition (Firman & Simmons, 2009). We therefore conclude that sperm competition does not influence the morphology of the sperm hook in house mice. Indeed, motile sperm groups exhibit slower swimming velocities than single sperm cells (Immler et al., 2007) suggesting that sperm groups in house mice seem to be an incidental by-product of sperm adhering together and attaining progressive motility (Firman & Simmons, 2009). Aggregated sperm undergo the acrosome reaction and lose their fertilizing potential (Moore et al., 2002; Fisher & Hoekstra, 2010), in which case selection should act against the formation of sperm groups. In addition, motile sperm groups form in the ejaculates of both monogamous and polygamous species (Fisher & Hoekstra, 2010). Thus, the adaptive significance of the murine sperm hook may have little to do with sperm competition. Future research should explore alternative hypotheses for the evolution of the sperm hook (for example see Firman & Simmons, 2009).
We thank M. Beveridge for performing the sperm hook GM analysis and T. Stewart for preparing the culture media. This work was supported by the Australian Research Council and approved by the UWA animal ethics committee (approval number: 3/100/607).