Patricia J. Moore, Department of Biological Sciences, 3.614 Stopford Building, University of Manchester, Oxford Road, Manchester M13 9PT, UK. Tel.: +44 (0)161 275 5128; fax: +44 (0)161 275 3938; e-mail: email@example.com
Selective pressure arising from sperm competition has been predicted to influence evolutionary and behavioural adjustment of ejaculate investment, but also may influence developmental adjustment of ejaculate investment. Immature males able to target resources strategically based on the competitive environment they will experience when they become sexually mature should be at a selective advantage. In our study we investigated how the presence of potential competitors or mates affects ejaculate and testes investment during development in the cockroach Nauphoeta cinerea, a species where males control female remating via their ejaculate size (large spermatophores prevent females from remating and therefore function to avoid sperm competition for males) and females store sperm. Our aim was to determine whether the social environment influences developmental adjustment of ejaculate investment and the relative importance of ejaculate components with different functions; avoidance of or engagement in sperm competition. We conclude that the social environment can influence developmental and behavioural flexibility in specific ejaculate components that may function to avoid or engage in sperm competition.
Females mate with more than one male in many species, which can result in sperm competition. The strength of sperm competition is generally predicted to result in selective pressure on males to increase their ejaculate size (Parker, 1990a,b; Parker et al., 1996, 1997; Parker, 1998). However, ejaculate production may be costly for males due to a tradeoff between energy spent manufacturing sperm and other demands such as growth (Sella & Lorenzi, 2003), somatic maintenance (Van Voorhies, 1992), and finding and obtaining mates (Dewsbury, 1982). Thus, males may have evolved mechanisms to adjust ejaculate investment in order to maximise fertilization opportunities (Wedell et al., 2002).
Evolutionary models predict that variation in the risk and intensity of sperm competition males encounter will result in variation in their investment in ejaculates (Parker, 1998). Both taxonomic and population level variation should result in the adjustment of ejaculate size. Variation in the strength of sperm competition is also associated with different mating systems (Shuster & Wade, 2003) and a number of studies have examined the evolutionary adjustments in ejaculate investment between species in response to the strength of sperm competition (Wedell et al., 2002). Thus, species in which males encounter sperm competition tend to have increased investment in testes (Gage, 1994; Hosken, 1997; Kappeler, 1997; Stockley et al., 1997; Hosken & Ward, 2001). Aside from taxonomic variation, variation in the strength of sperm competition within a population results from variation in the social environment (Wedell et al., 2002; Shuster & Wade, 2003). There is evidence that individual males adjust their ejaculate in response to information about the degree of sperm competition in a particular mating opportunity. Males that experience increased risk of sperm competition, for example in the presence of rival males, invest more in ejaculates than males that experience lower risk (reviewed in Wedell et al., 2002).
In addition to evolutionary and behavioural adjustments to ejaculate investment, males may also make developmental adjustments in investment in testes or ejaculates based on the prospective competitive environment within a particular population (Tan et al., 2004). Such variation in the social environment can lead to ontogenetic plasticity (Moore & Moore, 2003; West-Eberhard, 2003), which will result in phenotypic variation upon which selection can act. Researchers are just beginning to examine the role of social environment relative to the risk of sperm competition on developmental adjustments on ejaculate investment. For example, the presence of potential competitors causes males to increase sperm number in two cricket species, Gryllus bimaculatus and Gryllodes sigillatus, even when the experimental competitor belonged to a different species (Mallard & Barnard, 2003). Larger colony size in cliff swallows is associated with increased testes size (Brown & Brown, 2003), and hermaphroditic leeches make adjustments to gonad volume when raised under different group sizes (Tan et al., 2004). Similarly, high larval density (an indicator of increased sperm competition risk) has been shown to increase mating frequency and testes size in the moth Plodia interpunctella (Gage, 1995) and spermatophore size in the moth Pseudaletia separata (He & Tsubaki, 1992). Such examples are consistent with the prediction that investment in sperm production should respond to information about the sex ratio, and therefore the strength of sperm competition (Harris & Lucas, 2002). All these adjustments will be adaptive if social group size experienced by juveniles is predictive of future sperm competition that males will face.
Males vary investment in testes and ejaculates relative to the risk and intensity of sperm competition, including ontogenetic adjustments in response to social environment, but not all species experience sperm competition. Mounting evidence suggests that there is selection on mechanisms to avoid sperm competition (Simmons, 2001). For example, the rapid evolution of Drosophila seminal proteins, which function to inhibit female remating (Clark, 2002), mating plugs in Lepidoptera which form a physical barrier to remating (Ehrlich & Ehrlich, 1978; Simmons, 2001), or the effect of the whole ejaculate size on female sexual receptivity (Oberhauser, 1989; Kaitala & Wiklund, 1994). Indeed males may facultatively manipulate females by adjusting ejaculate investment relative to the risk of sperm competition. For example, in the moth P. interpunctella, females are inhibited from remating for several hours due to the spermatophore acting as a mating plug (Cook & Gage, 1995). Because of the manipulation of female remating, the presence of rival males does not represent a risk for sperm competition and there is no increase in the ejaculate investment in the presence of rival males in this species. Thus, males may moderate their adjustment in ejaculate investment relative to the risk of sperm competition, which encompasses not only the social environment and presence of rival males, but also their ability to manipulate female remating. Studies investigating the way selection can influence both the number of sperm inseminated and the mechanisms by which males manipulate female remating are needed to broaden our understanding of how sperm competition results in adaptation.
The cockroach Nauphoeta cinerea has been a model system for studying sexual selection and sexual conflict (Moore, 1988; Moore et al., 2001, 2003; Montrose et al., 2004). Males form social dominance hierarchies and dominant males have increased access to females for mating. Females discriminate among males, requiring a shorter courtship period from preferred males. Thus, both mate competition and mate choice operate in N. cinerea. We investigated the influence of the social environment during development on ejaculate investment in N. cinerea. Males manipulate female remating such that males do not experience sperm competition for fertilization of the first clutch of eggs. This occurs when males pass a large spermatophylax (above a threshold size), which stimulates a stretch receptor in the bursa copulatrix, inhibiting female mating receptivity (Roth, 1964; Montrose et al., 2004). However, females can remate after parturition (Moore et al., 2003) and also store sperm between matings, which results in mixed paternity in cockroaches (Cochran, 1979; Woodhead, 1985). Thus, males may experience sperm competition in subsequent clutches of eggs. These characteristics make cockroaches an ideal model to study ejaculate investment, as ejaculate components have different functions: investment in the spermatophylax allows males to avoid sperm competition and investment in sperm allows males to engage in sperm competition. Spermatophore mass is negatively genetically correlated with sperm number in N. cinerea, suggesting that evolution of these traits is constrained (Moore et al., 2004). However, facultative adjustment of each component is possible. Current theory predicts that the presence of rival males should lead to an increase in ejaculate investment (Parker, 1998; Wedell et al., 2002), and increases in the number of potential mates promotes sperm economy (Wedell et al., 2002). However, investment in the spermatophylax by cockroach males to avoid sperm competition may be most important when rival males are present.
Like previous studies (e.g. Gage, 1995; Brown & Brown, 2003; Tan et al., 2004), we examined the effect of social group size individual components of ejaculate investment. In addition, we manipulated group gender composition to examine how potential mates as well as competitors influence ejaculate investment. We tested whether male social environment, reflecting two different levels of future sperm competition risk, during sexual maturation influences either ejaculates or testes investment in N. cinera. Specifically, we examined whether a male's social environment, mediated by chemical signals originating from males (competitors) or females (potential mates), influenced ejaculate components responsible for manipulation of female remating (spermatophore size) and for fertilization (sperm numbers and testes size) in males attaining sexual maturity in their presence. We addressed the following predictions: (1) investment in spermatophylax size should be independent of social environment because of strong selection to avoid sperm competition; (2) when the social environment contains competitors, sperm investment should be relatively high, but relatively low when the environment contains mates but not competitors; (3) testes investment should be high when the social environment contains competitors.
Husbandry and experiments
Mass colonies and all experimental individuals of N. cinerea were housed under standard maintenance conditions of 27 °C with a 12/12 photoperiod and supplied with ad libitum water and rodent chow. Late-instar nymphs were sorted by sex from large mass colonies into 32 × 24 × 10 cm plastic containers before emerging as adults. Nymph colonies were checked daily for newly emerged adults, which were separated individually into 11 × 11 × 3 cm plastic containers.
Focal males were housed under one of three social environments during the 7-day period during which sexual maturation occurs after adult eclosion (Roth, 1964). Males were placed individually into the lower compartment of a container with two 17 × 11 × 4 cm spaces on the day they emerged as sexually immature adults. The upper compartment was filled with either three sexually mature 7-day-old females, 7-day-old males or kept empty. A 5 mm space between the compartments precluded physical contact of any kind whereas the facing panel of each container was perforated with 100 0.5 mm holes allowing air to flow freely between compartments. Thus, the social environment of focal males during sexual development was restricted to chemical interactions from sexually mature females, sexual mature males, or the absence of such signals in an isolation treatment.
To test the effects of differing social environments on ejaculate investment, focal males from each treatment were removed after the 7 days and mated sequentially to two virgin, sexually naïve 7-day-old females. Immediately after mating, females were stored in individual plastic tubes and frozen at −20 °C; focal males were similarly frozen immediately after their second mating. To retrieve spermatophores, females were thawed to room temperature and spermatophores were removed from the bursa after dissection. The weight of the whole spermatophore was measured to the nearest 0.1 mg. Spermatophores in N. cinerea consist of a sperm containing ampulla embedded in a proteinaceous, gelatinous mass called the spermatophylax (Roth, 1964). The sperm ampulla was removed from the spermatophylax and weighed separately to determine the mass of these two components.
To measure sperm number, the whole sperm ampulla from each spermatophore was placed into a 1.5 mL Eppindorf tube containing 200 μL of water. The sperm ampulla was then gently broken using a pestle and the tube was thoroughly mixed for at least 30 s by hand. The sperm mixture was then sampled immediately by taking 10 μL of the sperm mixture and placing in into a second Eppindorf tube containing 490 μL of water and 10 μL of 0.5% Eosin Y dye. The dilute sample tube was then agitated by hand for at least 30 s and eight 10 μL samples were taken and placed onto individually marked microscopic slides (sampling intensity was calibrated for accuracy by exhaustively sampling sperm from several spermatophores to obtain an accuracy curve to calibrate sampling effort). All sperm cells were then counted on each slide for each mating and each male using a light microscope and cell counter. To assess the effects of social environment on testes investment by males, testes size was measured for each male after the second mating. Males were thawed to room temperature and dissected. Testes were then removed and immediately weighed to the nearest 0.1 mg.
Repeated measures analysis of variance (anova) with social environment as a factor were performed on all data for ejaculate characteristics: total spermatophore weight, sperm ampulla weight, spermatophylax weight (total spermatophore weight – sperm ampulla weight), and sperm number. Repeated measures anova results will be presented below in the format of between male results (the overall effect of the social environment factor), the within males results (the effect of the repeated measure, mating order) and the interaction effect between the social environment treatments and mating order. Sperm number data were logarithm transformed to ensure a normal distribution. Tukey tests were used for post hoc comparison of treatment means for testes weight data. All statistical analyses were performed using SYSTAT 10.0.
Effect of social environment on spermatophore size
We found a significant effect of social environment on overall spermatophore size, both first and second spermatophores tended to increase in weight when males were housed in the presence of other individuals, either males or females, than when isolated (between males: F2,33 = 7.69, P < 0.01). As we have observed previously (Montrose et al., 2004), there was a reduction in spermatophore size between first and second matings (within males: F1,33 = 37.31, P < 0.001). There was no interaction effect between social environment and mating order (interaction: F2,33 = 0.15, P = n.s., Fig. 1a).
The pattern of spermatophylax size variation was similar overall to that of total spermatophore size. The spermatophylax produced by males exposed to a social environment that contained either male or females was larger than that produced by isolated males (between males: F2,31 = 7.55, P < 0.01). Again, spermatophylax size decreases from first to second mating (within males: F1,31 = 7.10, P < 0.05). There was no significant interaction effect between social environment and the mate order (interaction: F2,31 = 0.30, P = n.s., Fig. 1b).
Sperm ampulla size did not differ between social environments (between males: F2,31 = 1.58, P = n.s.). As with the spermatophylax, size of the sperm ampulla decreased between the first and second mating (within males: F1,31 = 171.30, P < 0.001). There was no interaction effect between social environment and mate order (interaction: F2,31 = 0.25, P = n.s., Fig. 1c).
Effect of social environment on sperm numbers and testes investment
Although there was no effect on sperm ampulla size, there was a significant effect of social environment on the number of sperm males allocated to their ejaculate (between males: F2,30 = 11.20, P < 0.01). Males invested significantly more sperm during their first mating when the social environment experienced during sexual maturation contained females compared to males housed in odour contact with other males or housed in isolation (Fig. 2). Sperm number in the ejaculates decreased between the first and second mating (within males: F1,30 = 11.20, P < 0.01). In addition, there was no effect of social environment on sperm numbers in the second mating, in which sperm numbers were the same for males in all social conditions. However, there was a significant interaction showing that sperm allocation in successive matings was dependent on the social environment (interaction: F2,30 = 3.42, P < 0.05).
There was no overall difference across treatments in testes size between males raised in a social environment containing males or females compared with isolation (F2,35 = 2.56, P = 0.092, Fig. 3). However, post hoc pairwise comparison of treatment means revealed there is a slight trend in testes size with a slightly smaller investment in testes for the males housed in odour contact with other males compared to isolated males (Tukey: P = 0.053), but not between males housed in odour contact with females and isolated males (Tukey: P = n.s.).
Our results suggest that variation in the allocation of resources to different ejaculate components can arise due to variation in the social environment experienced during sexual development. In N. cinerea, this suggests that differential resource allocation among ejaculate components may result due to pressures to avoid sperm competition by manipulating female remating in addition to pressure to compete against other males for paternity successfully. Our findings demonstrate an ontogenetic adjustment in response to signals in the social environment experienced by males prior to sexual maturity, which could act simultaneously to an evolutionary or behavioural adjustment in resource allocation in response to sperm competition. Resource allocation of the latter two types in response to sperm competition have been relatively well studied (Birkhead & Møller, 1998; Wedell et al., 2002). However studies investigating the evolutionary importance of environmentally induced developmental variation in any context are relatively rare (West-Eberhard, 2003).
The way in which N. cinerea males developmentally adjusted ejaculate investment was different than that predicted under behavioural sperm allocation models through a consideration of engagement in numerical sperm competition. Overall spermatophore size increased when males attained sexual maturity in an environment where they were in chemical contact with either sexually mature males or females. Spermatophore size increase was driven by variation in spermatophylax size. This was contrary to our prediction that spermatophylax size should be relatively constant for males across social environments, given that males manipulate female remating by producing a spermatophore at a threshold size. A possible explanation for the pattern of variation we observed is that males use cues about the likelihood of competitor presence and strategically allocate material to the spermatophylax accordingly. A similar pattern of ejaculate investment has been noted in the moth P. separata where the probability of female remating in this species is negatively correlated with spermatophore size (He & Tsubaki, 1991) and adult males reared in crowded larval conditions produce larger spermatophores as adults (He & Tsubaki, 1992). Our study differs from others in that, rather than increasing investment in components of the ejaculate to engage in sperm competition, N. cinerea appear to invest in the spermatophylax to avoid sperm competition, trading off investment into sperm. For N. cinerea the most costly component of ejaculates may be the nonsperm component of the ejaculate, as is often true for species with substantive nonsperm ejaculate fraction (Simmons & Kvarnemo, 1997). Previous work from our laboratory has demonstrated selective pressure on males to maintain spermatophore size above the threshold to inhibit female remating even in the absence of sperm (Montrose et al., 2004). Evidence of an evolutionary tradeoff among ejaculate components in N. cinerea comes from a strong negative genetic correlation between spermatophore size and sperm number (Moore et al., 2004). Until now most studies have focussed on how sperm competition may influence ejaculate size for numerical sperm competition (Parker, 1998; Wedell et al., 2002). Our findings suggest that pressures other than post-copulatory mate competition can strongly influence ejaculate characteristics, in this case to avoid sperm competition by manipulating female remating.
Some studies have shown that testes investment increases when the environment contains potential competitors (Wedell et al., 2002). For example, intraspecific variation in testes investment has been shown to correlate to increased sperm competition success (Preston et al., 2002). We found a weak trend for males to decrease testes investment during sexual development when their social environment contained adults of either sex. Surprisingly, testes investment was smallest when males were exposed to other males during development. This finding is consistent with the idea that investment in mechanisms to avoid sperm competition (the spermatophylax) is traded off against investment in engaging sperm competition (sperm production capacity). Although we did not measure accessory glands in this study, it is an interesting possibility that male N. cinerea might invest in these structures to bolster spermatophylax production capacity in order to avoid sperm competition when rival males are likely to be present.
Contrary to our expectation, we found sperm number was relatively constant except for the first mating for males in the female environment treatment. This is surprising because sperm allocation theory (Parker, 1998) predicts that males should increase sperm expenditure when the risk of sperm competition is high, such as when there are many competitors in the male social environment. A possible explanation for the pattern of variation in sperm number we observed is that there may be selection maximize fertilization in future clutches. As female N. cinerea store viable sperm throughout their lives (Montrose et al., 2004) and can remate between clutches (Moore et al., 2003), males will encounter sperm competition in subsequent clutches potentially reducing fitness gains for increased sperm investment. Therefore, increased sperm production may be worthwhile to males only when the risk of sperm competition is low. Sperm number for the second mating did not show the same pattern as that of the first mating, being similar in all social environments. This could be due to weak selection on males to produce sperm quickly if mating opportunities are generally rare, and may also reflect the cost of sperm production (Wedell et al., 2002).
Our finding that ejaculate size was adjusted in response to social environment has important implications for evolution of variation in ejaculates not because of sperm competition, in this example, but because of the avoidance of sperm competition by manipulating female remating behaviour. We suggest that males have been selected to facultatively adjust the nonsperm portion of their ejaculate to control female remating behaviour and avoid sperm competition during ontogeny of sexual development. It also appears that males can adjust sperm number independently of other ejaculate components, perhaps to maximize fitness over the lifetime of their mate. Our study provides empirical support for the idea that selection can act on variation in the mechanisms responsible for both engagement in sperm competition and avoidance of sperm competition. Ejaculates can be elaborate structures consisting of integrated components, each with a different function such as sperm transfer, sperm competition and female manipulation (Gillot, 2003). Therefore, to understand the evolution of the ejaculate each of these potential functions should be considered.
This work was supported by the Natural Environmental Research Counsel of the UK. Tamara Montrose provided technical assistance during this study. We thank Allen J. Moore and Per T. Smiseth for helpful discussion and comments on the manuscript. We also thank K. Vahed and an anonymous reviewer for thoughtful reviews of the manuscript.