Discussion
Overall the results supported the previous finding of redundancy of cues of D. melanogaster male rival presence. However, this redundancy may be incomplete due to differences in the ways these cue components are perceived and processed. Unexpectedly, no fitness benefits of extending mating duration in response to rivals were observed. This suggests that there may not be a simple, direct relationship between behavioural investment in mating and fitness consequences.
Redundancy of cues of D. melanogaster rival presence
The results supported the previous finding that removing one cue of rival presence does not affect the ability of male D. melanogaster to detect rivals and respond to them by extending their subsequent mating duration (Bretman et al. 2011b). Males exposed to a rival with the auditory cue removed showed equivalent mating duration to males housed with rivals with all cues intact, and both groups of males mated for significantly longer than males that had not encountered a competitor. This supports the conclusion that alternative cue combinations elicit equivalent behavioural responses. However, males that were exposed to rivals with all cues available were significantly slower to mate than males housed alone, while males who were exposed to rivals with the auditory cue removed showed intermediate mating latency. This suggested that the cues by which males detect competition may not be entirely redundant, and that removing one affects a male’s ability to respond to rivals by extending mating latency. However, we note that the extension of mating latency by males exposed to rivals may not show strong repeatability (Bretman et al. 2009; Bretman et al. 2013a; Bretman et al. 2013b). Overall, the results support the idea that there is at least partial redundancy in how cues indicating the presence of rivals are processed by the receiving male (Bretman et al. 2011b).
The way in which multiple cues are perceived and processed is likely to be related to social learning, whether these cues are redundant or confer information about different components of the social environment. Learning relies on cues being perceived, stored and compared to new environmental information (Dukas 2008; Bailey and Zuk 2009). Understanding which cues are important for influencing social behaviour, and how they lead to a behavioural outcome, may in turn increase understanding of the processes underlying social learning. A form of long-term memory has been found to be necessary for male D. melanogaster males to respond to rivals by adjusting mating duration (Rouse et al. 2018). It has been suggested that the timing of this response is important, on the basis that a minimum period of exposure to rivals of 24 h is required to elicit a response (Bretman et al. 2010), which then persists for 12 h (Rouse and Bretman 2016). Males may be required to remember their recent social environment in order to determine whether the cues of competition have persisted for long enough to be representative of the general level of sperm competition (Rouse et al. 2018). The mechanisms by which long-term memory facilitates responses to rivals by male D. melanogaster are localised to the mushroom bodies, highlighting the importance of olfactory cues (Rouse et al. 2018). Olfactory stimuli have also been found to be of particular importance for learning in the nematode Caenorhabditis elegans , in which learned associations with a number of aversive odours are formed at varying speeds (Choi et al. 2018). Therefore, although the multiple cues of D. melanogaster rival presence appear to be redundant in eliciting a longer mating response by males, there may be underlying differences related to how these cues are processed, the associative memories they form and the speed with which this happens. Indeed, the removal of auditory or olfactory cues of rivals have been found to extend the time taken for a male D. melanogaster to respond (Rouse and Bretman 2016). Increasing understanding of the role of social cues in learning may shed further light on the mechanisms by which reproductive plasticity is achieved (Rouse et al. 2018).
Fitness effects of the extension of male D. melanogaster mating duration following detection of rivals
There was no evidence that the extension in mating duration following detection of competitors led to any immediate reproductive fitness benefits, either in the absence or presence of sperm competition. Neither males exposed to all rival cues, nor those for which auditory cues were removed, showed an increase in the number of offspring they fathered when the female was singly mated, compared to males experiencing no cues of competition. Males exposed to rivals, with all cues intact or with the auditory cue removed, also did not increase the proportion of paternity they achieved under sperm competition. In fact, in the absence of sperm competition, there was a trend for males not exposed to a rival to father more offspring than males previously exposed to rivals. In the first set of experiments, males in the - all treatment also achieved greater success in sperm competition than those in the + all treatment. Furthermore, the longer mating demonstrated by males exposed to rivals did not reduce female receptivity to remating.
The finding that exposure to rival males (either with all cues intact or with one cue removed) and the associated longer-mating phenotype did not result in any apparent increases in reproductive fitness for D. melanogaster males was unexpected. Males exposed to cues of competition did not father higher numbers of offspring or reduce female receptivity to remating. This is inconsistent with previous findings (Bretman et al. 2010; Bretman et al. 2011b). There is evidence that the increase in the number of offspring fathered following longer mating occurs via increased transfer of two key seminal fluid proteins, sex peptide and ovulin, which increase female egg production and decrease receptivity to remating (Chapman et al. 2003; Chapman and Davies 2004; Wigby et al. 2009b). Neither of these effects were observed in the current study. Bretman et al. (2012) did find evidence that the behavioural response of longer mating duration can become decoupled from offspring production; however, this only occurred when males experienced a period without rival exposure prior to mating. Furthermore, Bretman et al. (2012) found evidence of males continuing to increase offspring production after mating duration was decreased, rather than of longer mating duration that did not correspond to fitness benefits. Nevertheless, Hopkins et al. (2019) found that sperm transfer and seminal fluid protein (SFP) transfer peak at different intensities of male-male competition, with the amount of SFPs transferred generally increasing with the level of competition. Additionally, the composition of SFPs in the ejaculate can change with the intensity of competition. These studies demonstrate that there may not be a simple relationship between level of competition, behavioural response and reproductive success.
One possible explanation for the absence of an increase in reproductive success among males exposed to competitors is that aggressive interactions with rivals led to the treatment males sustaining harm, reducing their condition and thus their ability to increase their ejaculate investment in response to competition. Nandy et al. (2016) found the expression of male-male aggression to be a key component of the cost of reproduction and a driver of decreased longevity under starvation in D. melanogaster . It has been proposed that aggression between males can impose costs via injury and energy expenditure (Bretman et al. 2013b), ultimately reducing life span (Gaskin et al. 2002; Costa et al. 2010). Males who suffer these costs from aggressive interactions during rival exposure may be less able to subsequently increase their investment in their ejaculate, negating the usual fitness benefits of extending mating duration. However, it has been argued that male-male aggression is a minor contributor to costs of reproduction (Bretman et al. 2013b; Leech et al. 2017). This is based on the findings that males housed with a rival sustained no more wing damage than males housed alone (Bretman et al. 2013b). Although social contact between male D. melanogaster does reduce lifespan, this could not be explained in any signature of behavioural differences between males (Leech et al. 2017). Moreover, male Drosophilaaggression has been found to decline with prolonged exposure to the male-specific pheromone 11-cis­ -vaccenyl acetate (cVA), suggesting that continuous exposure to rivals may reduce aggressive behaviour. Thus, males housed with rivals may not be engaged in high frequencies of aggressive encounters, reducing the likelihood that they would sustain harm during treatment that would decrease their reproductive success.
Male competitive success can respond to various features of the social environment in addition to the presence of competitors, including female condition and female mating status (Lewis and Iannini 1995; Bonduriansky 2001; Friberg 2006). The ejaculate investment of male D. melanogaster in this study may respond to these other variables, masking responses to the presence of competitors. This may explain why there was no elevation in offspring production from longer matings following exposure to rivals. For example, all females in this experiment were virgins prior to mating with the treatment males. Friberg (2006) found that males increased their investment in reproduction, leading to reduced female remating, when they perceived females to have previously mated. The virgin status of females in the current study may have cued to males the low probability of sperm competition, confounding the effects of the prior exposure to rivals. An alternative explanation for the uniformity in reproductive success across treatment groups is that all males were also virgins prior to the experimental mating. In polyandrous butterfly species, the first ejaculate a male produces is larger and contains more protein than subsequent ejaculates (Bissoondath and Wiklund 1996; Hughes et al. 2000). The male D. melanogaster tested here had not encountered a female since reaching reproductive maturity. Due to the high variation in the reproductive success of males (Bateman 1948) and the very high potential fitness cost of never mating at all, it may be beneficial for a male to invest heavily in the first reproductive encounter, whether competition is detected or not. This too may have obscured the differences between treatment groups in reproductive success.
An alternative explanation for the apparent lack of fitness benefits of extending mating duration in response to rivals is that longer matings conferred benefits to males in the form of increased sperm displacement, which was not measured in this study. Reproductive success under sperm competition was only measured in terms of sperm defensiveness. However, previous research has found extended mating and increased reproductive success to follow exposure to rivals whether the focal male was the first or second to mate with a female. Another possible ‘hidden’ fitness benefit of extending mating duration is the delaying of female remating up to 24 h. Females were isolated for 24 h following the first mating, thus their receptivity to remating during this window was not measured. Reduced receptivity during the first 24 h after mating could contribute to the adaptive value of increasing mating duration following rival exposure, despite the apparent lack of increase in offspring production. Nevertheless, previous studies have found that the amount of SFPs and sperm transferred by males can vary differentially with the degree of male-male competition, and that behavioural responses to rivals can be decoupled from fitness benefits (Bretman et al. 2012; Hopkins et al. 2019). Furthermore, a recent study on D. melanogaster similarly found that longer matings by males exposed to competitors did not correspond to increased paternity share (Marie-Orleach et al. 2020). Together with our results that extended mating duration did not have any observed fitness benefits, these results suggest that the relationship between cues of competition, behaviour and reproductive success may not be as simple or direct as previously thought. This opens further questions on how sensory cues are processed to infer the intensity, as well as risk, of sperm competition, and whether redundancy among cues persists at varying degrees of competition.