Satyrization in Drosophila fruitflies

Abstract The satyr of Greek mythology was half‐man, half‐goat, with an animal persona signifying immoderate sexual appetites. In biology, satyrization is the disruption of reproduction in matings between closely related species. Interestingly, its effects are often reciprocally asymmetric, manifesting more strongly in one direction of heterospecific mating than the other. Heterospecific matings are well known to result in female fitness costs due to the production of sterile or inviable hybrid offspring and can also occur due to reduced female sexual receptivity, lowering the likelihood of any subsequent conspecific matings. Here we investigated the costs and mechanisms of satyrization in the Drosophila melanogaster species subgroup of fruitflies. The results showed that D. simulans females experienced higher fitness costs from a loss of remating opportunities due to significantly reduced post‐mating sexual receptivity than did D. melanogaster females, as a result of reciprocal heterospecific matings. Reciprocal tests of the effects of male reproductive accessory gland protein (Acp) injections on female receptivity in pairwise comparisons between D. melanogaster and five other species within the melanogaster species subgroup revealed significant post‐mating receptivity asymmetries. This was due to variation in the effects of heterospecific Acps within species with which D. melanogaster can mate, and significant but nonasymmetric Acp effects in species with which it cannot. We conclude that asymmetric satyrization due to post‐mating effects of Acps may be common among diverging and hybridising species. The findings are of interest in understanding the evolution of reproductive isolation and species divergence.


| INTRODUC TI ON
Reproductive interference occurs when the courtship and copulation of one species is interrupted or disturbed by another (Gröning & Hochkirch, 2008). It has been observed across many taxa (de Bruyn et al., 2008;Landolt & Heath, 1987;Seehausen et al., 1997;Shuker & Burdfield-Steel, 2017) and can take many forms, including signal blocking, heterospecific rivalry and heterospecific mating (Gröning & Hochkirch, 2008). In insects and other animals, reproductive interference is often referred to as satyrization (Ribeiro & Spielman, 1986). The effects of satyrization can be symmetric or asymmetric, depending on the frequency of heterospecific mating, degree of reproductive incompatibility and strength of post-mating effects. Asymmetric satyrization influences the level of interspecific competition between species that hybrid mate, with greater asymmetry increasing the probability of competitive exclusion (Kishi & Nakazawa, 2013). This is an important consequence of heterospecific mating and is of interest in understanding reinforcement and species divergence (Matute, 2010) as well as in practical applications of satyrization as a method of insect control (Kishi & Nakazawa, 2013).
Satyrization can occur before and after mating. Asymmetries in premating satyrization costs arise when the probability of reciprocal heterospecific matings differs, due to divergent and incomplete mate recognition barriers, facilitating heterospecific mating in one direction at higher frequency than the other. Fitness effects primarily arise as opportunity for remating, energetic or mating trauma costs (Yassin & David, 2016).
Heterospecific matings are well known to result in the production of infertile or inviable hybrid offspring (Coyne & Orr, 1989, 1997Turissini et al., 2018). They can also result in the inhibition of sexual receptivity in heterospecific females, leading to fewer rematings with conspecific males. Seminal fluid proteins (Sfps) govern the extent to which heterospecifically mated females increase their egg production, decrease their subsequent receptivity and store or release sperm (Chapman, 2001;Rubinstein & Wolfner, 2013;Sepil et al., 2019;Sirot et al., 2014). As such, Sfps, including their major constituents, the accessory gland proteins (Acps), are predicted to be key determinants of the magnitude and asymmetry of post-mating satyrization effects. Sfps represent a diverse cocktail of proteins that form the nonsperm part of the male ejaculate of most species of insects and other animals. There are >200 Sfps in D. melanogaster (Mueller et al., 2005;Findlay et al., 2008;Findlay et al., 2009;Sirot, LaFlamme, et al., 2009;Sepil et al., 2019) that influence many post-mating behavioural and physiological responses, such as ovulation, sperm storage and mating receptivity (Chapman et al., 2003;Chapman & Davies, 2004;Hollis et al., 2019;Liu & Kubli, 2003;Rubinstein & Wolfner, 2013).
Approximately 10% of the genes encoding Sfps evolve rapidly (Haerty et al., 2007;Mueller et al., 2005;Swanson & Vacquier, 2002). Though many D. melanogaster Sfps are orthologous to those found in other species within the Drosophila melanogaster species subgroup, others are species-specific (Findlay et al., 2008). As a result of this rapid evolution, Sfps may quickly become incompatible across diverging species, facilitating reproductive isolation (Andrés et al., 2008;van Doorn et al., 2009;Goenaga et al., 2015). Therefore, Sfps are expected to have variable heterospecific effects (Dapper & Wade, 2016;Tsuda & Aigaki, 2016) and could contribute to significant post-mating satyrization. Lineage-specific differences in the rate of evolutionary change of Sfps vs. their receptors in females could generate significant asymmetries indicative of satyrization (Ahmed-Braimah et al., 2017). Sfps with functional effects in the heterospecific context would render females refractory to further matings with conspecifics and induce costs in terms of 'time out' of the mating pool and through the production of infertile or sterile offspring.
Reproductive incompatibilities may also be impacted and potentially ameliorated, by conspecific sperm precedence (Castillo & Moyle, 2019;Price, 1997;Turissini et al., 2018). Several species within the D. melanogaster species subgroup exhibit conspecific sperm precedence, that is, in situations in which females are carrying sperm from both conspecific and heterospecific males, conspecific sperm will be preferentially used to fertilize eggs. While this phenomenon may reduce costs of satyrization through lower production of infertile/sterile hybrid offspring, it does not reduce conspecific mating opportunities lost to heterospecific matings, which are predicted to be significant and contribute to competitive exclusion (Noriyuki et al., 2012). Such costs are predicted to lead to selection for reinforcement to avoid such heterospecific matings (Matute, 2010).
As yet, neither the frequency of asymmetric satyrization, nor the post-mating mechanisms underlying it, are fully resolved.
Potential markers of satyrization include differences in incomplete mate recognition and Sfps that show variable functional effects in heterospecific mating. Both of these effects are reported in natural populations of Aedes mosquitoes, which are vectors of harmful diseases such as Dengue, Zika and Yellow Fever (Alto et al., 2014;Hugo et al., 2019;Johnson et al., 2002). Ae. aegypti females will readily mate with Ae. albopictus males, whereas the reciprocal mating does not occur. Hence, Ae. aegypti females frequently receive Sfps from Ae. albopictus males, causing an increase in the production of infertile eggs and rendering Ae. aegypti females less willing to mate with conspecifics. Therefore, Ae. aegypti (but not Ae. albopictus) females can suffer significant costs from asymmetric satyrization. This is thought to be a major contributor to the observation that Ae. albopictus replaces Ae. aegypti via competitive exclusion in areas of sympatry (Tripet et al., 2011). Ae. albopictus is a less competent vector of Dengue, Zika and Yellow Fever than Ae. aegypti (Alto et al., 2014;Hugo et al., 2019;Johnson et al., 2002). Therefore, in this context, satyrization is of interest for insect control.
There is much interest in the relative contribution of premating and post-mating processes to divergence in sympatry vs. allopatry (Matute, 2010). The underlying processes involved include those that lead to heterospecific matings (Turissini et al., 2018), the actions of Sfps (Sepil et al., 2019) and the relative rates of divergence of reproductive genes (Hollis et al., 2019). Overall, it is increasingly realized that post-mating prezygotic processes can play an important role in initiating and driving reproductive isolation in all settings (Matute, 2010). Here, we build upon this recent interest by investigating these mechanisms in the context of satyrization. We investigated satyrization costs and mechanisms in experimentally tractable Drosophila fruiflies, with a primary focus on the effects of Acps. Our aim was to test the hypothesis that there are significant costs due to asymmetric satyrization, explore whether satyrization is asymmetric across a group of closely related species, and examine the role of Acps in this phenomenon. Previous work investigating satyrization in Drosophila has demonstrated that conspecific mating costs, in the form of physical trauma, are often amplified in heterospecific matings (Yassin & David, 2016). There is also is an extensive body of research into heterospecific matings specifically between D. melanogaster and D. simulans (e.g. Coyne & Orr, 1989, 1997. All hybrid progeny from D. melanogaster x D. simulans matings are sterile or infertile with differences in the frequency and consequences of reciprocal hybridizations reported. We first tested for asymmetries in the frequency and post-mating satyrization effects of reciprocal heterospecific matings between D. melanogaster and D. simulans, to estimate satyrization under our experimental conditions. We then tested for asymmetric satyrization in post-mating responses across the D. melanogaster species subgroup. To do this, we documented female receptivity to mating after injections of conspecific or heterospecific Acps, vs. a saline control, in comparisons between D. melanogaster and five other members of the D. melanogaster species subgroup (Obbard et al., 2012). We used the frequency of copulations as a metric for sexual receptivity, measuring the difference in the number of copulations and speed of copulation onset between treatments. As satyrization includes both a premating and post-mating component, we included three species with which D. melanogaster can physically copulate (D. simulans, D. sechellia, D. teissieri) and two with which it cannot (D. erecta and D. yakuba) (Turissini et al., 2018). 'Post-mating' here refers to the inducement of physiological changes through the effect of Acps by injection into the abdomen, in the absence of actual mating. This allowed us to demonstrate the strength of post-mating satyrization and test whether asymmetry in post-mating satyrization is restricted to species that exhibit complete premating barriers which prevent heterospecific mating.

| Fly culturing and collection
Unless stated otherwise, Drosophila eggs were collected by placing a red grape juice agar plate (275 ml H 2 O, 12.5 g agar, 250 ml red grape juice, 10.5 ml 10% w/v Nipagin solution) into population cages containing the appropriate species. D. melanogaster was cultured in population cages containing overlapping generations at 25°C and 60% RH on a 12 hr:12 hr light:dark cycle. The cages hold 12 bottles each containing 70 ml of sugar yeast agar (SYA) medium (30 ml After 24 hr, first-instar larvae of each species were picked from the plates and placed 100 per vial (75 x 25 mm), each containing 7 ml SYA. This procedure standardized the larval development across and within species and minimized any environmentally induced variation in body size. Virgin adult females and males were collected using ice anaesthesia and separated by sex. The sex-segregated flies were then stored, 10 per vial for 3-6 days until their use in experiments.

| Frequency of heterospecific and conspecific matings between D. melanogaster and D. simulans
Adult D. melanogaster (Dahomey) and D. simulans (National Drosophila Species Stock Center (DSSC)) wildtype flies were allocated at random to one of the four following experimental treatments: D. simu- Figure S1). One male and one female from each species were gently aspirated into a vial within 2 hr after lights on and were continuously observed for 3 hr, during which spot checks were also performed every 20 min to score courtship and copulation frequency. The mating duration of

| Effects of hetero-and conspecific matings on female remating receptivity in D. melanogaster and D. simulans
D. melanogaster and D. simulans were collected as stated above and adults each aspirated into a vial with a conspecific or heterospecific male that had been placed in the vial 24 hr earlier (Experiment 1B, Figure S1). At 9:00 on the first day, pairs were continuously observed for 3 hr and mating latency and mating duration were recorded.
After matings ended, males were immediately removed, and females retained in their vials for 24 hr. Unmated females were discarded. At 13:00 the next day, 24 hr after the previously mated females had finished mating, the females were transferred into a new vial containing a conspecific male and were observed for 3 hr to test for post-mating receptivity. As before, mating latency and mating duration were recorded. No matings were observed between D. melanogaster (♀) x D. simulans (♂). Therefore, no females from this treatment were available for remating tests. Excess heterospecific pairs were set up to ensure sufficient mated females for rematings. The sample size set up for each treatment in each experiment and the number and percentage of pairs that mated are given in Table S1.  (Turissini et al., 2018).

| Effects of reciprocal Acp injections between
To generate Sfp-mediated post-mating physiological effects, Acps were injected into females of each species. Acps were extracted from the entirety of the accessory gland, but did not include proteins from the ejaculatory duct (see dissection details, below).
Male Acp donors, for tests with D. melanogaster x D. simulans/ D. erecta/ D. yakuba males, were collected within 24 hr of eclosion to standardize male age and stored 10 per vial for at least 48 hr to replenish Acps. Thus, the extracted Acps were from fully rested, sexually mature males, and of comparable status and volume across the different species tested. In tests with D. melanogaster x D. teissieri/ D. sechellia, it was found that D. teissieri and D. sechellia showed low fecundity on egg collection plates and suffered high mortality at 25°C. Therefore, flies for these two experiments were cultivated in food vials for 8 hr and 16 hr laying periods at 22°C under 12 hr:12 hr light:dark cycle, 60% RH. Egg-laying vials were set up, each containing 8 females and two males of the respective species (and 4 females and 1 male for D. melanogaster to control egg density across species). Adults were first placed into vials for an 8 hr egg-laying period and then immediately transferred to new vials for 16 hr to lay eggs.
Adult flies were removed after the egg-laying period and the eggs from both oviposition collections placed at 22°C to develop to adult emergence, after which the males were collected and kept in single sex groups of 10 males for at least 48 hr to replenish Acps.
To prepare Acps for injection into females, 90-120 pairs of accessory glands were dissected from 2-to 4-day-old males of each species, separated from the ejaculatory duct, and placed into a microcentrifuge tube containing 1xPBS (phosphate-buffered saline) at a concentration of 3 accessory gland pairs/μl of 1xPBS.
These were stored at −20°C. The day before the injection experiment, the accessory gland pairs were sonicated in 1xPBS with 5x one second pulses and centrifuged at 12,000 g for 15 min at 4°C.
The supernatant was placed into a new microcentrifuge tube and stored at −20°C.
Virgin females for injection were collected in the same way as the Acp donor males for each respective species and given 2-6 days to sexually mature before injection. On the day of the injection experiments, virgin females were anaesthetized on CO 2 and injected with 0.1 μl of either 1xPBS, 0.1 μl of conspecific Acps or 0.1μl of heterospecific Acps. Acps were injected directly into the abdomen of each female (Tsuda & Aigaki, 2016). The volume of fluid injected represents 0.3-0.5 of an accessory gland equivalent and is comparable to the amount of Sfps received in a normal mating . Immediately after injections, each female was placed into a separate vial containing yeast paste (to promote mating) and placed at 25°C (for experiments using D. simulans, D. yakuba and D. erecta) or 22°C (for experiments using D. sechellia and D. teissieri) for 24 hr. Eighty females per treatment were initially injected in each experiment to ensure a sufficient sample size for the subsequent mating assay (Table S2). Twentyfour hours post-injection, a conspecific male was placed into each vial containing a surviving female. Pairs were observed for 3 hr (4 hr for the D. melanogaster x D. sechellia/ D. teissieri experiments conducted at 22°C). Introduction of the male, mating start and mating finish times were recorded to assess the number of matings, mating latency and mating duration.

| Statistical analysis
Copulation frequency and mating latency data were analysed by performing a Kruskal-Wallis test followed by Dunn's post hoc analysis to test for significant differences between treatments.
Differences in the number of matings and rematings, and in post-Acp injection survival, were analysed used a chi-square test.
Differences in female mating receptivity following Acp injection were analysed using a Cox proportional hazards model. A generalized linear model (GLM) was used to test for interaction effects between injection treatments and species of the injected female, with significant differences in the effects of the reciprocal Sfps being indicative of satyrization asymmetry. All analyses were carried out in R v3.2.2 (R Core Team, 2012).
During the second mating, when all females were paired with a conspecific male, all three treatments had a relatively low remating rate with no significant difference between them (χ 2 2 = 5.63, p = .06). There were also no significant differences in mating latency between any of the treatments (H 2 = 2.38; p = .305), demonstrating that the post-mating refractory effect induced by D. melanogaster males was similar in conspecific D. melanogaster and heterospecific D. simulans females. Hence, heterospecifically mated D. simulans females showed a significantly reduced propensity to remate, leading to a potentially costly period of elevated production of sterile or inviable offspring production. As the heterospecific matings were unidirectional, only D. simulans incurred this post-mating cost.

| Effects of reciprocal Acp receipt across the melanogaster species subgroup
Overall, significant asymmetries in female receptivity were seen fol-  Figure S2). However, the Acps from these three species either had no, or a significantly weaker, effect than D. melanogaster Acps on receptivity in the reciprocal tests in D. melanogaster females (Figure 2). In contrast, no significant asymmetries in female receptivity were seen in reciprocal Acp injections between D. melanogaster and D. erecta or D. yakuba (Figure 3).
In these species, the Acps significantly reduced female receptivity equally in conspecific and heterospecific comparisons. Asymmetries Mating latency (mins) during the first (red) and second (blue) matings between D. melanogaster and D. simulans, tested at 25°C. X-axis labels describe the treatments in the first mating. All mated females from the first mating were mated with a conspecific male for the second mating regardless of the species of the male from the first mating. The sample size set up for each treatment and the number and percentage that mated is shown in Table S2. Box plots show the median, 25%-75% IQ range, whiskers (1.5 x IQR) and outliers. Different letters indicate statistically significant differences between groups (p < .05)

(b)
F I G U R E 2 Asymmetrical post-mating responses between members of the D. melanogaster species subgroup. Shown is the Cox proportional hazards model of females that mated over the 3 hr mating assay period, 24 hr following injection with either saline (red), D. melanogasterAcps (blue) or D. simulans (a), D. sechellia (b) and D. teissieri (c) Acps (black). Asymmetry is revealed by a comparison of the left and right panels. Shown in the shaded areas are the 95% confidence intervals for each treatment, asterisks indicate significant differences between treatments connected by black lines (

F I G U R E 3
Symmetrical post-mating responses between members of the D. melanogaster species subgroup. Shown is the Cox proportional hazards model of females that mated over the 3 hr mating assay period, 24 hr following injection with either saline (red), D. melanogasterAcps (blue) or D. erecta (a), and D. yakuba (b) Acps (black). Shown in the shaded areas are the 95% confidence intervals for each treatment, asterisks indicate significant differences between treatments connected by black lines ( In general, saline injections were less harmful to female survival than either con-or heterospecific Acp injections. D. melanogaster females were resistant to most injections of conspecific and heterospecific Acps with no significant differences between Acp and saline injections in any of the injection experiments except for D. melanogaster x D. simulans, where there was significantly lower mortality following saline injections compared to both con-and heterospecific Acps (χ 2 2 = 33.25; p = 6.016e-08). D. yakuba and D. teissieri were particularly sensitive to Acp injections, with females suffering significantly higher mortality when injected with Acps from both con-and heterospecific Acps compared to the saline control (D. yakuba: χ 2 2 = 39.37; p = 2.824e-9. D. teissieri: χ 2 2 = 20.32; p = 3.862e-05) (see Supporting information for a full breakdown of injection mortality).

| D ISCUSS I ON
Our results show significant costs of satyrization for D. simulans females that mated with D. melanogaster males, which were not observed in the reciprocal cross. D. simulans females mated at a reasonable frequency with D. melanogaster males, producing offspring with zero fitness, and showed significant reluctance to remate. In a natural setting, this may result in the female spending a significant time out of the mating pool -though any costs would be tempered by conspecific sperm precedence (Price, 1997). We examined the contribution of post-mating effects to satyrization, by using Acp in- Stronger asymmetries in the fitness effects of heterospecific matings can facilitate competitive exclusion between two species (Kishi & Nakazawa, 2013). The frequency of heterospecific matings can play a significant role in this process (Matute, 2010). Our results supported the extensive previous evidence for asymmetric premating satyrization between D. melanogaster and D. simulans (Barbash, 2010;Barker, 1962;Coyne & Orr, 1989, 1997Moulin et al., 2004;Sperlich, 1962;Sturtevant, 1920;Turissini et al., 2018).
Heterospecific matings occurred unidirectionally, with D. melanogaster males mating infrequently with D. simulans females but with the reciprocal cross occurring at zero frequency. Therefore, D. simulans females that mated with D. melanogaster males incurred significant fitness costs in terms of the production of inviable or sterile hybrid offspring (Barbash, 2010) and reduced willingness to remate with conspecifics and thus receive conspecific sperm. Conspecific matings were significantly more frequent and were shorter to initiate than heterospecific matings between D. melanogaster and D. simulans. This is consistent with reports that incomplete mate recognition contributes to hybridizations between these species and suggests mate recognition control by females (Barbash, 2010). Almost all conspecific pairs mated and some pairs mated several times. D. simulans (♀) x D. melanogaster (♂) pairs mated more frequently than the reciprocal cross which was not observed at all in the mating tests performed here. However, even the most frequent heterospecific matings only occurred at about a third as often as for conspecifics.
This provides evidence for premating satyrization -in addition, the presence of unidirectional heterospecific mating (and associated post-mating effects described below) resulted in females of only one species suffering fitness costs of heterospecific mating. Some previous studies have observed that heterospecific matings between D. melanogaster females and D. simulans males are more frequent than the reciprocal (Moulin et al., 2004;Sperlich, 1962;Sturtevant, 1920).
Our results contrast with this observation, but are in agreement with other reports of exclusive, unidirectional heterospecific mating between D. melanogaster males and D. simulans females (Barker, 1962).
The pattern of unidirectionality in matings between D. melanogaster x D. simulans thus appears to be strain dependent and should be in- The results suggest that, in addition to any direct ecological competition when in sympatry, either of D. melanogaster or D. simulans could be at a potential disadvantage from asymmetric satyrization effects. This is dependent upon the direction of asymmetry which varies across different strains, at least in terms of premating effects (Barker, 1962;Moulin et al., 2004;Sperlich, 1962;Sturtevant, 1920)). Costs of satyrization will be diminished if there is strong conspecific sperm precedence (Castillo & Moyle, 2019;Price, 1997;Turissini et al., 2018 Overall, asymmetry in post-mating effects was found only in different species which can engage in heterospecific mating (Turissini et al., 2018) suggesting that asymmetries occurred between species that are more closely related (Balakrishnan et al., 2009;Miller et al., 2019;Moulin et al., 2004;Sato et al., 2015;Schwarz & McPheron, 2007).
D. yakuba and D. erecta are more phylogenetically distant to D. melanogaster than are D. simulans and D. sechellia, although D. teissieri seems to lie between D. erecta and D. yakuba (Obbard et al., 2012). That asymmetric satyrization occurred in all of the most closely related members tested could suggest that it is widespread. In areas in which closely related species have overlapping ranges, satyrization could shape interactions between closely related sympatric species.
Why there might be a link between the ability to hybridize and asymmetrical post-mating effects of Acps is not yet known, but two possibilities are described below: 1. Evolution of resistance to costly heterospecific matings. Diverged species have generally evolved complete premating barriers which can take the form of behavioural or mechanical premating isolation mechanisms (Ehrman, 1964;Matute, 2010). However, it is also possible that Sfps might, in part, be shaped by selection to reduce the compatibility of interspecific matings, prior to the evolution of complete premating isolation (Billeter & Wolfner, 2018). D. melanogaster and D. yakuba/ D. erecta are highly diverged and show strong premating barriers, which prevent the occurrence of heterospecific matings (Turissini et al., 2018). However, we found that Acps remained functional and induce strong physiological responses similar to those of conspecifics in these species. This indicates that Acps in these species have not been shaped by selection for mating incompatibilities and that premating barriers in these species evolved rapidly and prior to any divergence in Acp functions.
Increasing species divergence is expected to result in degraded interspecific Acp functions over time (Orr, 1996). The finding of a degree of conservation in the re-mating inhibitory functions between Acps of species as widely diverged as D. melanogaster and D. yakuba/ D. erecta suggests the possibility of evolutionary constraints on at least some Acps and their receptors.
2. Consequences of sexual conflict in the D. melanogaster species subgroup. Sfps across a wide variety of taxa evolve rapidly which may be a result of strong or conversely even excessively relaxed selection (Dapper & Wade, 2020;Findlay et al., 2014). In the D. melanogaster species subgroup, it has been hypothesized that sexual conflict can promote the rapid evolution of Sfps (Findlay & Swanson, 2010;Hollis et al., 2019;Minekawa et al., 2018;Pitnick et al., 2001;Sirot et al., 2014Sirot et al., , 2015. The Sfps of Drosophila spp.
have multiple functions, but high apparent functional redundancy, which may prevent females from easily evolving resistance to Sfps with manipulative effects (Chapman, 2008(Chapman, , 2018. However, as a side effect, this may also predispose Sfps to retain their ability to effect post-mating responses in heterospecific females. It is also possible that the degree of any such redundancy is itself variable across the species tested in this study, which might contribute towards the asymmetric satyrization observed. The production of many different types of Sfps per function is likely to be costly and might also trade off against other traits. For example, D. sechellia are endemic to the Seychelles and exhibit relatively low genetic diversity and a small effective population size (David & Capy, 1982;Legrand et al., 2009). D. simulans appears to have fewer Sfps than are found in D. melanogaster (Findlay et al., 2008). This suggests that either D. simulans has shed redundant Sfps or D. melanogaster has evolved novel Sfps.
The observed asymmetries suggest that Acps are evolving faster in some lineages than others but that Acp receptors in these rapidly evolving species have broad-scale specificity. Consequently, these receptors may retain the ability to bind and be activated by less rapidly evolving Acps, resulting in asymmetric effects in reciprocal matings.

| CON CLUS IONS
Here we have found significant asymmetrical satyrization within a single clade of Drosophila fruitflies. This work builds upon studies in other Diptera species (Tripet et al., 2011;Turissini et al., 2018;Yassin & David, 2016), to demonstrate that satyrization is present within members of the D. melanogaster species subgroup and quantify the pre and post-mating costs. Drosophila exhibit variable premating barriers, with biased heterospecific mating frequency, and significant asymmetries in the post-mating effects of Acps. This is evidence that asymmetric satyrization is likely much more widespread than has been originally thought and is likely to be an important yet underappreciated factor in speciation, sexual selection and interspecific competition.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
TC, SL, LA and MIT devised the experiments, SL conducted the research, collected and analysed the data; SL and TC wrote the paper; all authors contributed to the final draft.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/jeb.13733.

DATA AVA I L A B I L I T Y S TAT E M E N T
The raw data are deposited in the DRYAD data depository, https:// doi.org/10.5061/dryad.0zpc8 66wc