https://orcid.org/0000-0003-1643-6522
Abstract
The operational sex ratio (OSR) is a key component influencing the magnitude of sexual selection driving the evolution of male sexual traits, but males often also retain the ability to plastically modulate trait expression depending on the current environment. Here we employed an experimental evolution approach to determine whether the OSR affects the evolution of male calling effort in decorated crickets, a costly sexual trait, and whether plasticity in calling effort is altered by the OSR under which males have evolved. Calling effort of males from 2 selection regimes maintained at different OSRs over 18–20 generations (male vs. female biased) was recorded at 2 different levels of perceived competition, in the absence of rivals or in the presence of an experimentally muted competitor. The effect of the OSR on the evolution of male calling effort was modest, and in the opposite direction predicted by theory. Instead, the immediate competitive environment strongly influenced male calling effort as males called more in the presence of a rival, revealing considerable plasticity in this trait. This increased calling effort came at a cost, however, as males confined with a muted rival experienced significantly higher mortality.
Introduction
Sexual selection describes the variation in reproductive success that arises via differences in mating success within a sex and typically acts more strongly on males than on females (Darwin, 1871). In most animal taxa, male reproductive success depends critically on the number of mating partners that a male obtains (Bateman, 1948). Traits that promote a male’s mating success (i.e., secondary sexual characters) can evolve through either of two distinct mechanisms: (a) intrasexual selection, which favors traits that enhance a male’s ability to compete directly with rival males in obtaining access to receptive females (e.g., in insects [Chechi et al., 2022]; mammals [La Boeuf, 1974]; and fish [Dijkstra et al., 2009]) and (b) intersexual selection, which promotes traits that increase the likelihood that males will be chosen as mates (e.g., in insects [Friberg & Arnqvist, 2003]; mammals [Clutton-Brock & McAuliffe, 2009]; and fish [Godin & Briggs, 1996]). Both types of sexual selection can be important concurrent selective pressures within a population (Fitze et al., 2008), but the intensity of sexual selection can also vary across environments (Candolin et al., 2007; Jann et al., 2000). Therefore, a deeper understanding of the evolution of sexually selected traits requires investigation of the critical factors thought to underlie this environmental variation in sexual selection. This is especially important given that sexual selection in populations can influence important life-history traits, such as immunity and pathogen resistance (McKean & Nunney, 2008; McNamara et al., 2013), as well as population adaptation and persistence (Candolin & Heuschele, 2008; Parrett et al., 2019).
Numerous ecological factors can influence the direction and intensity of sexual selection and concomitantly the evolution of secondary sexual characters. For example, variation in habitat altitude can create gradients in the intensity of sexual selection and the evolution of secondary sexual traits in some bird species (Badyaev & Ghalambor, 2001; Snell-Rood & Badyaev, 2008). Other factors such as predation risk (Endler, 1995; Kelly & Godin, 2001) and food availability (Gwynne & Simmons, 1990) have also been shown to alter the intensity of sexual selection. The structure of the mating system has been identified as a key factor influencing the magnitude of sexual selection because of variation in the extent to which individuals are able to exert control over access to potential mates (Emlen & Oring, 1977). The operational sex ratio (OSR), the number of sexually active males to sexually receptive females within a population, is often used as a proxy for this environmental component. Theory predicts that when the OSR is skewed, the more abundant sex should experience greater intrasexual competition, whereas the less abundant sex should exhibit greater choosiness (Emlen & Oring, 1977; Janicke & Morrow, 2018). Previous empirical work has examined how the OSR drives selection for traits associated with mating (Debuse et al., 1999; Jann et al., 2000; Souroukis & Cade, 1993). In addition to affecting trait evolution, some behavioral traits related to mating are not constant but can instead change based on the current OSR in the environment. For example, investment into behaviors such as intrasexual aggression and courtship displays can depend on the OSR (Grant & Foam, 2002; Souroukis & Cade, 1993).
Many traits related to mating and reproduction are not fixed but instead exhibit phenotypic plasticity, in which their expression varies depending on the environment (Agrawal, 2001; Bretman et al., 2016). Various ecological factors can lead to a change in reproductive behavior, including predation risk (Fowler-Finn & Hebets, 2011), climate (Noonan et al., 2018), and the number of competitors (Bretman et al., 2016). The OSR is a fundamental feature of populations that can also lead to differential expression of male sexual traits, such as courtship displays (Grant & Foam, 2002). However, different OSRs may also alter the evolution of behavioral plasticity (Weir et al., 2011). Theory predicts that in persistent, stressful (i.e., extreme) environments, there may be selection for continuous expression of traits that lead to decreased environmental sensitivity, or genetic assimilation (Badyaev, 2005; Levis & Pfennig, 2016). As such, we might expect that males subject to intense directional sexual selection for multiple generations, such as occurs at male-biased OSRs, might lose the ability to modulate the expression of secondary sexual traits because of strong selection on amplified and continued mating effort needed to compete successfully with other males, attract females, or both. While some studies have shown that the OSR drives selection for traits associated with mating (Debuse et al., 1999; Jann et al., 2000; Kvarnemo et al., 1995; Souroukis & Cade, 1993), as well as the plasticity observed in the expression of these traits (Grant & Foam, 2002; Souroukis & Cade, 1993), few studies have employed an experimental evolution approach to test whether the OSR drives the evolution of sexually selected traits and behavioral plasticity in their expression (Dore et al., 2021; House et al., 2019; Maggu et al., 2021; Michalczyk et al., 2011; Palopoli et al., 2015).
Crickets represent an ideal model system with which to investigate the relationship between the OSR and the intensity of sexual selection acting on male secondary sexual traits and plasticity. They have been shown to vary in the sex ratio of adults in wild populations, which can lead to variation in life history traits (Rodríguez-Muñoz et al., 2019). Males exhibit an obvious and readily quantifiable sexual trait to attract mates, acoustic signaling, which is essential for obtaining copulations with females (Sakaluk, 1987; Zuk & Simmons, 1997). Calling occurs when males stridulate, rubbing their forewings to create “chirps” with a characteristic carrier frequency and stereotypic temporal pattern (Huber & Thorson, 1985). Previous studies have shown that females choose mates based on the temporal structure of the calling song produced by males, as well as overall calling effort (Wagner & Reiser, 2000; Walker, 1957); females typically prefer males that call for longer durations (Hedrick, 1986). Although vital to a male’s mating success, calling is also a costly trait. Calling can comprise 56% of an individual’s daily respiration (Prestwich & Walker, 1981), and males who heavily invest in these displays often experience decreased longevity (Callander et al., 2013; Hunt et al., 2004). Previous research has also demonstrated that genetically distinct inbred lines of the decorated cricket, Gryllodes sigillatus, display significant variation in calling effort (Archer et al., 2012), suggesting that it is heritable and, thus, potentially responsive to sexual selection. However, male secondary sexual traits are not necessarily fixed and may be expressed plastically. In a study of field crickets, Gryllus pennsylvanicus, Souroukis and Cade (1993) found that males modulate reproductive behaviors such as calling, searching, and male–male fighting in response to varying OSRs. Environmentally dependent behavioral plasticity in long-range calling effort has also been demonstrated in G. sigillatus, with older male crickets modulating calling based on different levels of perceived infection risk (Duffield et al., 2018). This suggests that calling effort can be flexibly modulated in response to other environmental factors in addition to variation in the OSR.
If, as theory suggests, the OSR influences the intensity of sexual selection, we would expect calling effort of male crickets to vary in accordance with the OSR in the population within which they have evolved. Specifically, we would predict that the increased sexual selection that ensues within a male-biased OSR should favor males who invest more in calling effort (Figure 1). This strategy would increase the likelihood of copulation in the face of more severe competition for females. In contrast, when the OSR is female biased, sexual selection on males should be relaxed, and we would predict a lower calling effort and a more temporally balanced level of calling over the lifetime (Figure 1). This is because the mating success of males from female-biased lines is not limited by the availability of females, and thus, competition among males should be greatly reduced. An important caveat here, however, is that the OSR can be positively related to the opportunity for selection even in the absence of sexual selection (Fairbairn & Wilby, 2001; Klug et al., 2010); nevertheless, a recent meta-analysis has revealed that the OSR is predictive of sexual selection on males and the direction of sexual selection with respect to sex differences across a wide range of animal taxa (Janicke & Morrow, 2018).
Hypothetical evolutionary outcomes for age-specific male investment in calling effort in response to varying operational sex ratio (OSR). Males from male-biased lines (high male:female ratio, solid red) should exhibit higher calling effort shortly after their final eclosion to adulthood at a cost to their longevity. Males from female-biased lines (dashed gray) should exhibit more balanced calling over their lifetime and have increased longevity when compared with males from male-biased lines.
With respect to behavioral plasticity in calling effort, theory predicts that in a male-biased OSR, selection will favor maximal calling effort that increases a male’s competitiveness, ultimately leading to decreased environmental sensitivity or genetic assimilation (Badyaev, 2005; Levis & Pfennig, 2016). In contrast, males reared in female-biased populations should retain the ability to modulate the expression of behavioral traits because of the absence of continuous, intense selection for increased expression of these traits. This dichotomy in the level of plasticity ought to be most apparent when comparing calling effort in males from male-biased and female-biased OSRs subject to varying levels of perceived competition. Males from male-biased populations should always demonstrate higher calling effort than female-biased lines, but those from female-biased lines should have the ability to modulate their calling to a greater degree when faced with a competitor. Thus, female-biased males should have greater sensitivity to environments with varying levels of the perceived threat of competition and be able to alter their calling effort to a greater degree accordingly.
In this study, we investigated whether the OSR affects the evolution of calling effort and age-specific schedule of calling in male decorated crickets, and whether behavioral plasticity in calling is altered by the OSR under which males have evolved. To do this, we leveraged experimental evolution lines developed in our laboratory via the imposition of two different selection regimes: one imposing a severe male-biased sex ratio, and the other imposing a severe female-biased sex ratio. Using a custom-built sound monitoring array, we quantified calling effort and the behavioral plasticity of males from two OSR regimes under varying levels of perceived competition. We predicted that the calling effort of male decorated crickets from male-biased lines would be greater than that of males from female-biased lines. In addition, we predicted that males from female-biased lines would demonstrate a greater degree of plasticity in calling effort when compared with males from male-biased lines.
Methods
Study animals
Outbred G. sigillatus and those from our experimental evolution lines were descended from approximately 500 adults collected at Las Cruces, New Mexico, in 2001 (Ivy & Sakaluk, 2005). Groups of approximately 500 newly hatched nymphs were randomly allocated to 6-L plastic bins provisioned with egg carton to increase rearing surface, finely ground cat food (Purina Complete Cat Chow) provided ad libitum, and water in vials plugged with moist cotton. This process ensured gene flow each generation, thus promoting the maintenance of genetic variation within the mass colony. After approximately 3 weeks, nymphs were transferred to 19-L bins and given whole Purina Complete Cat Chow, Envigo 2018 CM Teklad Certified Global 18% protein rodent diet pellets and water as described earlier. Upon the imaginal molt, crickets were transferred to 62.5 L plastic bins and provisioned as stated earlier except that they were also provided with a moistened cotton pad in a small plastic storage container (118 ml) to provide a suitable substrate for oviposition. All crickets were reared at constant temperature (30 ± 2 °C) and reversed photoperiod (14:10 hr light:dark cycle).
Four replicate lines were established under each of two OSR regimes for a total of eight replicate lines: male biased and female biased. In male-biased lines, populations were initiated with 250 males and 50 females (5:1 OSR), leading to intensified sexual selection on males. Female-biased lines were initiated with 50 males and 250 females (1:5 OSR), resulting in relaxed sexual selection on males. These lines were initiated by manipulating the adult sex ratio (male or female biased) in paired replicate populations. To establish the first pair of replicate populations, we collected approximately 4,000 newly hatched nymphs from the mass colony. Half of these nymphs were allocated at random to a 110-L plastic container to serve as the starting generation for the male-biased population and the remaining half to a second 110-L container to serve as the starting generation for the female-biased population. Each container was provided with food and water ad libitum, as well as an abundance of cardboard egg cartons for shelter as described earlier for the mass colony. This procedure was repeated each day for six consecutive days to produce six paired replicates of each selection regime (i.e., one paired replicate per day), of which we used four replicate lines from each regime in the present study.
During the first 10 generations, each line was sorted into single-sex containers when last-instar nymphs were detected, and after the required number of these nymphs had become adult, the selection regimes were established. For the final 10 generations, containers with nymphs were checked daily for the presence of adults. Once adults were observed, they were checked every other day, and adults added to the new selection line at the prescribed adult sex ratio until the target population size was achieved. Adults for each line were placed in a separate 62.5-L container and provisioned with food and water ad libitum, plus an abundance of cardboard egg cartons. We also provided each population with two oviposition containers replaced weekly. When eggs hatched, nymphs were collected at random from each population to establish the next generation (as described earlier). The adult population was then killed by placing the container at −20 °C overnight.
Males used in this experiment were recorded in blocks over three generations (F18–F20) to secure an adequate sample size. We recorded a total of 447 males over the course of the study (F18: n = 70, F19: n = 153, and F20: n = 224); the distribution of subjects over the various experimental treatments is described in more detail below. Experimental males were removed from their respective selection lines within 1–2 days of their adult molt to control for mating status and age. Males were then housed in individual plastic containers (450 ml) and provisioned with a small section of egg carton as refuge, food, and water. Adult males typically begin calling at around 4–5 days after adult eclosion and mating does not commence until males can call; acoustic signaling is essential for successful copulation (Burpee & Sakaluk, 1993). Hence, song recording was initiated when the experimental males were 4–7 days old. To assess the influence of OSR on plasticity in male calling effort, long-range calling was assessed within two different competitive regimes. In one regime, calling effort of solitary males was measured to determine differences in calling effort in the absence of competition. In a second regime, calling effort of focal males was quantified in the presence of a competitor. In the latter regime, a focal male from one of the selection regimes was confined with a nonexperimental outbred reference competitor. To prevent confounding calling measurements by the nonexperimental competitor with the focal male, we experimentally muted outbred males by surgically removing one of their forewings.
Quantifying calling effort and survival
To quantify the calling effort of males, we used a custom-built high-throughput sound monitoring array (Bertram & Johnson, 1998) in which each individual box (250 ml) housing a male had a lid-mounted microphone (C1163, Dick Smith Electronics). These boxes were then placed within individual Styrofoam enclosures to prevent cross-talk between containers. Males were placed in their respective containers 1 day before recording commenced, either by themselves or with a muted competitor, to allow for their acclimation to the recording environment. Recordings began at 12:00 PM and ended at 10:00 AM the next day so that males could be checked for adequate provisions and to confirm their survival. If a muted outbred reference competitor had died, it was subsequently replaced before recording resumed. Recordings were made over a 3-week period in blocks of 4 days each week to acquire sufficient calling data to resolve any treatment effects, but also to investigate how calling effort varies with male age. This meant that male recordings took place on days 5–8, 12–15, and 19–22 of adulthood, ±2 days for each of the recording periods. The sound monitoring array sampled microphones throughout the sample period every 2 s. Based on the binary output resulting from this protocol, the proportion of time spent calling was calculated for each male each day and averaged within each 4-day block. After the end of each 4-day recording block, males were returned to their containers until the start of the next recording week. The 3-week recording period was based on the longevity of male G. sigillatus under natural conditions in an outdoor enclosure (Sakaluk et al., 2002), which is approximately 3 weeks (Sakaluk, unpublished data); this aligns closely with the natural lifespan of other gryllids for which this has been determined (Rodríguez-Muñoz et al., 2010; Zajitschek et al., 2009). After the end of the 3-week recording period, focal males were returned to their individual containers and checked daily for their continued survival. Upon their death, the pronotum width of experimental males was measured as a proxy for structural body size using a stereomicroscope (Nikon SMZ800) equipped with a digital camera and imaging software (Nikon NIS-Elements Documentation v 4.20). A total of 447 males were sampled, 226 from male-biased lines (112 solitary, 114 with a competitor) and 221 from female-biased lines (113 solitary, 108 with a competitor).
Statistical analysis
All statistical analyses were performed using R version 4.2.1 for Mac. Linear mixed effects models and generalized linear mixed models using repeated measurements were fit with the lme4 package (Bates et al., 2015) and glmmTMB (Brooks et al., 2017), respectively. For survival, Mixed Effects Cox Proportional Hazards models were fit with the package coxme (Therneau et al., 2015). Potential distributions of each response variable were examined for model fit and adherence to model assumptions. We used the package emmeans to produce Estimated Marginal Means with confidence intervals (Lenth et al., 2020). For all analyses, the colony line was included as a random effect, and in the repeated measures analysis, individual was also included as a random effect.
Calling effort was analyzed with a generalized linear mixed model with repeated measurements, with initial fixed effects of OSR, competitive regime, male age, and their two- and three-way interactions. Generational cohort and body size were also included in the model. The model used an ordered beta distribution (ordbeta), which allows for continuous (i.e., proportion) data in a closed interval from 0 to 1 (Kubinec, 2022). For survival, initial fixed effects were OSR, competitive regime, their interaction, body size, and cohort. For the analysis of the effect of regime on male adult body size, OSR and cohort were included as fixed effects in a linear mixed-effects model. Models were compared and simplified using likelihood ratio tests and AIC. The statistics of terms dropped from the models were taken from the step before their removal.
Results
Calling effort
There was a significant effect of OSR on male calling effort. Males from female-biased lines called at higher rates than those from male-biased lines (Table 1; Figure 2). There was an even stronger effect of the competitive regime on male calling effort, with males under the increased perceived competition treatment calling at significantly higher rates than solitary males (Figure 2). There was, however, no significant interaction between OSR and competitive environment in their effects on calling effort, nor was the three-way interaction with male age significant (Table 1). There was also no effect of male age on male calling effort (Table 1; Figure 2). Levels of calling effort varied across the three cohorts (Table 1), but this was not directional, with calling effort in F19 being lower than in F18 and F20, and the effects of OSR, competitive regime, and male age were not contingent on cohort (individual results not shown, p > .12).
Effects of OSR, competitive environment, and age on male calling effort.
| Response measured | df | χ2 | p |
|---|---|---|---|
| Calling effort | |||
| OSR | 1 | 7.64 | .0057* |
| Competitive environment | 1 | 261.68 | <.0001* |
| Age | 2 | 3.65 | .1614 |
| Cohort | 2 | 43.62 | <.0001* |
| Body size | 1 | 0.047 | .8287 |
| OSR × Environment | 1 | 0.17 | .6844 |
| OSR × Age | 2 | 1.96 | .3744 |
| Environment × Age | 2 | 5.53 | .0629 |
| OSR × Environment × Age | 2 | 1.25 | .5353 |
| Response measured | df | χ2 | p |
|---|---|---|---|
| Calling effort | |||
| OSR | 1 | 7.64 | .0057* |
| Competitive environment | 1 | 261.68 | <.0001* |
| Age | 2 | 3.65 | .1614 |
| Cohort | 2 | 43.62 | <.0001* |
| Body size | 1 | 0.047 | .8287 |
| OSR × Environment | 1 | 0.17 | .6844 |
| OSR × Age | 2 | 1.96 | .3744 |
| Environment × Age | 2 | 5.53 | .0629 |
| OSR × Environment × Age | 2 | 1.25 | .5353 |
Note. Bolded terms were retained in the final model.
*Significant effects (α = 0.05).
Effects of OSR, competitive environment, and age on male calling effort.
| Response measured | df | χ2 | p |
|---|---|---|---|
| Calling effort | |||
| OSR | 1 | 7.64 | .0057* |
| Competitive environment | 1 | 261.68 | <.0001* |
| Age | 2 | 3.65 | .1614 |
| Cohort | 2 | 43.62 | <.0001* |
| Body size | 1 | 0.047 | .8287 |
| OSR × Environment | 1 | 0.17 | .6844 |
| OSR × Age | 2 | 1.96 | .3744 |
| Environment × Age | 2 | 5.53 | .0629 |
| OSR × Environment × Age | 2 | 1.25 | .5353 |
| Response measured | df | χ2 | p |
|---|---|---|---|
| Calling effort | |||
| OSR | 1 | 7.64 | .0057* |
| Competitive environment | 1 | 261.68 | <.0001* |
| Age | 2 | 3.65 | .1614 |
| Cohort | 2 | 43.62 | <.0001* |
| Body size | 1 | 0.047 | .8287 |
| OSR × Environment | 1 | 0.17 | .6844 |
| OSR × Age | 2 | 1.96 | .3744 |
| Environment × Age | 2 | 5.53 | .0629 |
| OSR × Environment × Age | 2 | 1.25 | .5353 |
Note. Bolded terms were retained in the final model.
*Significant effects (α = 0.05).
Effects of operational sex ratio and competitive regime on male calling effort. (A) Box plots depicting the proportion of time spent calling in male decorated crickets (Gryllodes sigillatus) over each of three consecutive weeks. Dark horizontal lines within each box indicate the median, the box the interquartile range, and the whiskers the maximum and minimum of the distributions, with outliers shown as individual points. (B) Effect of competitive regime on the proportion of time spent calling (estimated marginal means with 95% confidence intervals). (C) Effect of operational sex ratio (OSR) on the proportion of time spent calling (estimated marginal means with 95% confidence intervals).
Survival
There was no significant effect of the OSR on male survival (χ21 = 1.10, p = .29). There was, however, a significant effect of the competitive environment on male survival (χ21 = 13.07, p < .0003; Figure 3). Males under higher perceived competition exhibited significantly lower survival than solitary males (Figure 3). There was no significant interaction between OSR and competitive environment in their influence on male survival (χ21 = 0.02, p = .88), nor was there an effect of body size (χ21 = 0.53, p = .47). Male calling was significantly associated with the likelihood of mortality within both competitive regimes, with increased calling effort predictive of a linear increase in male mortality (F1 = 9.64, p = .002, Figure 4); the slope of this relationship did not, however, differ between the two regimes (F1 = 0.03, p = .86).
Survival plots of male decorated crickets under varying competitive regimes. (A) Estimated survival hazard of males based on their competitive regime. Points represent the estimated marginal means with bars representing 95% confidence intervals. Sample sizes are provided below confidence intervals. (B) Effect of competitive regime on male survival over time.
The effect of calling effort on the predicted survival of male decorated crickets. Males housed with a muted competitor are shown in steel gray and solitary males are depicted in red. Green dashed lines represent regression lines for both treatments.
Body size
Male body size did not differ between the two OSR selection regimes (F1 = 1.00, p = .36). The mean pronotum length of males in the male-biased lines was 2.08 ± 0.01 mm (mean ± SE) and in the female-biased lines 2.06 ± 0.01 mm.
Discussion
Theory predicts that under a skewed OSR, the more abundant sex should experience more intense sexual selection (Emlen & Oring, 1977; Janicke & Morrow, 2018), and thus we predicted that males from male-biased lines would invest more in calling effort than males from female-biased lines. The OSR had a significant effect on male calling effort, but in the opposite direction to that predicted, with males from female-biased lines exhibiting higher calling effort than those from male-biased lines. Calling effort was surprisingly consistent over the lifetime of the male regardless of the selective regime under which males had evolved. However, the competitive social environment in which males were recorded had a marked effect on male calling effort, with males confined with muted rivals singing approximately three times as much as solitary males. The degree of plasticity in calling was not, however, dependent on the selection regime. We had predicted that males from the male-biased regime, presumably under strong sexual selection for maximum calling effort, would exhibit decreased environmental sensitivity to the presence of a competitor through genetic assimilation, whereas males from female-biased lines would retain their ability to modulate calling effort to a greater degree (Badyaev, 2005; Levis & Pfennig, 2016). Instead, the lack of an interaction between OSR and competitive regime is inconsistent with the existence of any such genetic assimilation, or loss of plasticity through genetic drift, in male calling effort. Indeed, our data suggest that males from a male-biased regime might still benefit from plasticity even if population densities varied only slightly, as the increase in calling effort leads to increased mortality. However, the loss of plasticity is likely to be a slow process, and the time course of the experiment (18–20 generations), coupled with relatively small effective population sizes, may have precluded the detection of any genetic assimilation.
Our results showed that, contrary to predictions, a male-biased OSR resulted in a lower level of male calling effort compared with males from the female-biased selection regime. Why might this be the case? First, it must be acknowledged that manipulation of the mating system influences more than just the intensity of sexual selection, including levels of intra- and interlocus sexual conflict, indirect benefits of mate choice, and patterns of assortative mating, among other evolutionary forces (Rowe & Rundle, 2021). Nonetheless, other studies employing a similar experimental evolution approach in varying the OSR have reported the predicted increase in male competitiveness accruing to a male-biased OSR. In Drosophila melanogaster, for example, males from a female-biased selective regime fought less often in food patches than males from a male-biased OSR (Bath et al., 2021). In Tribolium castaneum, individuals from male-biased OSRs had increased reproductive success compared with individuals from females-biased OSRs when competing for females against rival control males (Michalczyk et al., 2011). Despite studies such as these, a meta-analysis of the effect of OSR on male competitiveness has revealed a more nuanced pattern, as male aggression initially increases with an increase in the OSR, but then decreases at more extreme OSRs, perhaps due to diminishing returns of aggression as rivals become more numerous (Weir et al., 2011).
The relationship between the OSR and sexual selection is determined by the extent to which males are able to monopolize females (Emlen & Oring, 1977), but the increased costs of monopolizing mates in a male-biased OSR may favor alternative tactics for securing copulations (Janicke & Morrow, 2018). In this regard, we note that not only did the OSR differ between our two selective regimes, but so too did the density of males in the replicate lines. It may be that higher male calling effort is favored with an increase in the density of male competitors, but only up to a critical threshold, beyond which a decrease in calling is favored if females are likely to passively encounter noncalling rivals as often as they do stationary calling males. In this kind of competitive social environment, selection might favor greater investment by males in more short-range sexual signals, and an obvious candidate would be close-range courtship song, produced by males when females are in close proximity (Alexander, 1961). However, a previous study has revealed that the structure of male courtship song is unrelated to male condition as manipulated by male diet (Gray & Eckhardt, 2001). A more promising candidate in this regard might be cuticular hydrocarbons, lipid compounds found on the cuticle that facilitate sex recognition in crickets (Ryan & Sakaluk, 2009; Tregenza & Wedell, 1997) and known to influence female mate choice in G. sigillatus (Steiger et al., 2015). Another possibility, not mutually exclusive, is that other unmeasured components of male song reflective of male vigor and known to influence male mating success were favored by the male-biased selective regime, including song amplitude, chirp rate, chirp duration, and carrier frequency (Bertram et al., 2022). Attesting to this possibility in G. sigillatus, Froome (2022) found a significant interaction between male size and age in their influence on male song structure; smaller males called more but had increased interpulse and interchirp durations compared with larger males, indicative of reduced vigor. In the current study, however, body size had no significant effect on male calling effort in either selection regime.
In addition to unmeasured aspects of male song, sexual selection imposed by our selection regimes may have targeted other traits important to male reproductive success, the investment in which tradeoff with resources devoted to male calling effort. In male dung beetles, Onthophagus nigriventris, for example, a trade-off occurs between investment in pre- and postcopulatory sexual traits, such that males without horns, an important precopulatory weapon in the context of male–male competition, allocate more resources into testes growth, affording them an advantage in postcopulatory sperm competition (Simmons & Emlen, 2006). In G. sigillatus, males provide females with a nuptial gift at mating, a gelatinous addition to the male’s spermatophore that the female consumes after mating (Sakaluk, 1984, 1987). Previous studies have shown that the size of the gift has a profound influence on the number of sperm transferred by the male (Sakaluk, 1984), and by extension, his paternity (Sakaluk, 1986; Sakaluk & Eggert, 1996). Additionally, the chemical composition of the gift, specifically, its free amino acid content, influences whether the female accepts or discards the gift (Gershman et al., 2012). In support of the possibility that varying OSR might have influenced some component of male nuptial gifts, a companion study on our experimental evolution lines has revealed that males from male-biased lines synthesize larger food gifts with amino acid profiles that are more likely to elicit a female feeding response compared with males from female-biased lines (Burns-Dunn et al. in review). Similarly, Simmons and Lovegrove (2017) showed that when the apparent risk of sperm competition was experimentally increased in male field crickets, Teleogryllus oceanicus, so too did their expression of several seminal fluid protein genes that affect the quality of males’ ejaculates. Finally, male–male competition in G. sigillatus, as in other crickets, does not end with the transfer of the spermatophore as males guard females after mating to prevent courtship attempts by rival males (Frankino & Sakaluk, 1994; Sakaluk, 1991). It may be that changes in mate guarding intensity are selected for under these varying OSR regimes. In fact, a meta-analysis conducted by Weir et al. (2011) revealed a significant increase in the duration of postcopulatory guarding with an increase in male bias of the OSR.
In contrast to the relatively modest influence of the OSR selection regime on male calling effort, the immediate social competitive environment of the male had a profound influence on male calling effort, with males housed with a muted competitor increasing their calling effort threefold compared with that of solitary males, demonstrating high levels of plasticity in this trait. This aligns with observations in other cricket mating traits associated with male–male competition, including intrasexual aggression (Souroukis & Cade, 1993), sperm allocation (Schaus & Sakaluk, 2001), the composition of the ejaculate (Simmons & Lovegrove, 2017), and calling effort in other cricket species (Noguera, 2019). In fact, the very plasticity that we observed might have attenuated any sexual selection on male calling effort imposed by varying the OSR. Males in the male-biased regime could easily have responded to the increased competitive social environment by ramping up their calling effort compared with males held in the female-biased selective regime, hindering selection for genetic changes underlying variation in calling effort. Indeed, it has been suggested that high levels of plasticity may reduce the likelihood of genetic changes (Price et al., 2003) even when, as appears to be the case in G. sigillatus, calling effort is highly heritable (Archer et al., 2012).
The increased calling effort of males in the presence of rivals did not, however, come without substantial costs, as males confined with a muted rival experienced significantly higher mortality than solitary males. This is consistent with previous studies and supports the hypothesis that increased investment in sexual signaling often comes at the expense of the resources available for maintenance (Callander et al., 2013; Hunt et al., 2004; Noguera, 2019), although this is likely to depend on the availability of resources (Hunt et al., 2004). Highlighting the connection between male calling effort and mortality risk, this pattern was also evident at the individual level within both social competitive environments. Although the two environments differed in overall risk, the rate at which mortality risk increased with calling effort appeared homogeneous across the two environments.
There was no effect of age on male calling effort in either selective regime, in contrast to other studies in G. sigillatus that found an increase in male calling effort with age (Archer et al., 2012; Froome, 2022). This difference is not as surprising as it might appear, as previous research on our cricket population has demonstrated that age-specific variation in male calling effort can be contingent on extrinsic factors, such as disease risk or diet, that similarly affect an individual’s residual reproductive value (Duffield et al., 2018, 2020). Studies of other crickets have shown that calling effort can decrease with age (Bertram et al., 2022; Fitzsimmons & Bertram, 2011), increase with age (Bertram et al., 2022; Kuriwada & Kasuya, 2011), or, as in the present study, be unrelated to male age. This interspecific variation in calling effort, along with the intraspecific variation noted earlier, suggests that the effects of age on calling effort are complex and likely influenced by a myriad of both intrinsic and extrinsic factors (Duffield et al., 2017, 2018, 2019, 2020).
In conclusion, despite the imposition of two vastly different selection regimes over 18+ generations, the effect of the OSR on the evolution of male calling effort was relatively modest, and in the opposite direction predicted by theory. Instead, the immediate competitive environment of the male strongly influenced male calling effort, revealing considerable plasticity in this trait. This plasticity aligns well with that observed in other aspects of male competitiveness in crickets, including the propensity to engage in intrasexual aggression (Souroukis & Cade, 1993), sperm allocation (Gray & Simmons, 2013; Schaus & Sakaluk, 2001), and the composition of the ejaculate (Simmons & Lovegrove, 2017). This plasticity did not come without a cost, however, as increased calling effort resulted in an increased mortality risk. An important objective moving forward, then, is to determine the extent to which this plasticity itself can evolve and to identify those elements of the social environment that alter the breadth of this plasticity.
Data availability
The data and code are available at Dryad: https://doi.org/10.5061/dryad.h9w0vt4pw
Author contributions
J.T.M.: conceptualization, investigation, formal analysis, methodology, funding acquisition, data curation, writing—original draft preparation, writing—review & editing, visualization. B.F.: investigation, writing—review & editing. W.K.: investigation, writing—review & editing. J.H.: conceptualization, methodology, resources, funding acquisition, writing—review & editing. B.M.S.: conceptualization, supervision, project administration, formal analysis, methodology, data curation, funding acquisition, writing—review & editing, visualization. S.K.S.: conceptualization, supervision, project administration, formal analysis, methodology, funding acquisition, writing—review & editing.
Funding
This research was funded by Weigel grants from the Beta Lambda Chapter of the Phi Sigma Biological Honor Society, the Theodore J. Cohn grant from the Orthopterists’ Society, and the E. L. Mockford and C. F. Thompson Summer Research Fellowship to J.T.M., a grant from the Undergraduate Research Support Program at Illinois State University to W.K., a grant from the Australian Research Council (DP180101708) to J.H., and a grant from the National Science Foundation (IOS 16–54028) to S.K.S., B.M.S., and J.H.
Conflict of interest: The authors declare no conflict of interest.
Acknowledgments
We thank Matt Dugas, Kristin Duffield, and Pirmin Nietlisbach for insightful discussions.
