Reinforcement in the banded darter Etheostoma zonale: The effect of sex and sympatry on preferences

Abstract Reinforcement occurs when selection against hybrid offspring strengthens behavioral isolation between parental species and may be an important factor in speciation. Theoretical models and experimental evidence indicate that both female and male preferences can be strengthened upon secondary contact via reinforcement. However, the question remains whether this process is more likely to affect the preferences of one sex or the other. Males of polygynous species are often predicted to exhibit weaker preferences than females, potentially limiting the ability for reinforcement to shape male preferences. Yet, in darters (Percidae: Etheostoma), male preference for conspecific mates appears to arise before female preferences during the early stages of allopatric speciation, and research suggests that male, but not female, preferences become reinforced upon secondary contact. In the current study, we aimed to determine whether the geographically widespread darter species Etheostoma zonale exhibits a signature of reinforcement, by comparing the strength of preference for conspecific mates between populations that are sympatric and allopatric with respect to a close congener, E. barrenense. We examined the strength of preference for conspecifics for males and females separately to determine whether the preferences of one or both sexes have been strengthened by reinforcement. Our results show that both sexes of E. zonale from sympatric populations exhibit stronger conspecific preferences than E. zonale from allopatric populations, but that female preferences appear to be more strongly reinforced than male preferences. Results therefore suggest that reinforcement of female preferences may promote behavioral isolation upon secondary contact, even in a genus that is characterized by pervasive male mate choice.

Theoretical models addressing the likelihood of male mate choice suggest that it is evolutionarily constrained relative to female mate choice (Servedio & Lande, 2006). Many evolutionary scenarios model speciation facilitated by the coevolution of female preferences and divergent male traits (e.g., Fisher-Lande, Lande, 1981; mutation-order, Mendelson, Martin, & Flaxman, 2014;reinforcement, Servedio, 2007), whereas models of male mate choice show that male preferences are often selected against (Servedio, 2007;Servedio & Dukas, 2013;Servedio & Lande, 2006), especially if male preferences are for "arbitrary" female traits (i.e., traits that do not directly indicate female fecundity; Servedio & Lande, 2006). However, one scenario in which male mate choice may be maintained is in the case of reinforcement (Servedio, 2007).
Reinforcement is the process whereby natural selection against hybrids leads to strengthened behavioral isolation between species (Dobzhansky, 1937;Noor, 1995;Shaw & Mendelson, 2013), with both theoretical and empirical studies showing that reinforcement can play an important role in the process of speciation (Kirkpatrick, 2001;Matute, 2010;Ortiz-Barrientos, Grealy, & Nosil, 2009;Servedio, 2007;Servedio & Noor, 2003). The classic signature of reinforcement is stronger preferences for conspecifics in populations that are sympatric versus allopatric with respect to a close congener, and many studies investigating reinforcement test for its signature in females (Kelly & Noor, 1996;Kirkpatrick & Servedio, 1999;Lemmon & Lemmon, 2010;Liou & Price, 1994;Servedio & Saetre, 2003). However, empirical support for reinforcement has been found in males for a number of species as well (Gregorio, Berdan, Kozak, & Fuller, 2012;Kronforst, Young, & Gilbert, 2007;Peterson et al., 2005), and Servedio (2007) found that, under some theoretical conditions, selection on male mate choice is more likely to result in reinforcement than selection on female mate choice. The question therefore remains whether reinforcement is more likely to affect preferences in one sex or the other, and if so, under what conditions. An effective way to address this question empirically is with comparative studies that quantify male and female preferences for conspecifics across multiple species pairs within a lineage, but these studies are rare. A recent study by Yukilevich and Peterson (2019), for example, found that both male preference and female preference for conspecific mates are widespread in Drosophila; however, in species that are sympatric with respect to a close congener, female preferences for conspecifics were stronger than male preferences, suggesting that reinforcing selection primarily acts upon female preferences in sympatry. In contrast, in a study of the darter subgenus Oligocephalus, male preferences showed a signature of reinforcement, being stronger for conspecifics in sympatric versus allopatric populations, whereas female preferences for conspecific males were not significant in either sympatric or allopatric populations . Recent findings also suggest that male preferences for conspecific females evolve earlier than female preferences across several allopatric darter species across the subgenera Nanostoma and Ulocentra (clade Simoperca sensu Near et al., 2011) (Mendelson et al., 2018. In the current study, we focused on two sympatric darter species, from a different subgenus, to evaluate whether a signature of reinforcement was present in males and/or females. Given that reinforcement appears to shape male but not female preferences in Oligocephalus, comparing preferences in another darter subgroup may reveal whether this pattern is characteristic of the genus. Specifically, we tested whether (a) preferences for conspecifics are stronger in populations that are sympatric with a close congener than in allopatric populations, that is, the classic signature of reinforcement, and (b) whether one sex exhibits a greater strength of preference for conspecific mates.
Darters (Percidae: Etheostoma) are a species-rich group of freshwater fishes native to North America with over 200 described species (Page & Burr, 2011). Premating barriers are found to evolve more rapidly than other barriers in darters (Martin & Mendelson, 2016;Mendelson, 2003;Mendelson, Imhoff, & Venditti, 2007;Williams & Mendelson, 2014), and diversification in darters is thought to occur primarily in allopatry, with sister species pairs all having allopatric ranges (Near et al., 2011). However, natural hybrids have been reported involving over 25% of darter species (Keck & Near, 2009), and recent phylogenetic analyses find evidence of ancient hybridization and genome-wide introgression between major darter lineages (MacGuigan & Near, 2018). These data suggest that hybridization between darters, both current and historically, is common enough to support a potential role for reinforcement in this system. Darter species Etheostoma zonale and E. barrenense represent some of the most closely related darter species to co-occur without hybridizing in nature (Hubbs, 1955(Hubbs, , 1967Keck & Near, 2009).
Previous work on this species pair shows that crosses between conspecific and heterospecific individuals have similar fertilization and hatching success in the laboratory (Williams & Mendelson, 2014); however, hybrid survivability appears to be reduced relative to that of intraspecific offspring (Williams & Mendelson, 2014), suggesting a degree of postzygotic incompatibility that could promote reinforcement. Male E. zonale and E. barrenense exhibit elaborate and divergent nuptial coloration and pattern, with male E. barrenense displaying primarily red-orange coloration with black blotches fused along the lateral line, and male E. zonale having alternating green and yellow bars along the body (Figure 1). Females of both species display some muted coloration and the patterning that characterizes conspecific males. Previous studies have found that these visual signals are important for mate choice in both females (Williams & Mendelson, 2010) and males (Roberts & Mendelson, 2017).
For this study, we used E. zonale as a focal species since most of its range is allopatric with respect to E. barrenense ( Figure 2). The geographic range of E. barrenense is small and entirely sympatric with E. zonale, whereas E. zonale occurs over a wide range throughout the Mississippi River basin, Ohio River basin, Ozark-Ouachita drainage, and was recently (1950s-1960s) introduced into the Susquehanna River (Kneib, 1972). If reinforcement plays a role in the evolution of mate preferences in E. zonale, we predicted that populations in areas of sympatry with E. barrenense would have stronger preference for conspecific mates than populations allopatric with E. barrenense. Additionally, we measured the strength of preference for conspecifics in both males and females to determine whether reinforcement has sex-specific effects in the focal species.

| Overview
To estimate the strength of preference for conspecific mates, we conducted two types of behavioral experiments. Dichotomous mate choice trials gave individual fish a choice between a conspecific or heterospecific of the opposite sex and did not allow fish to physically interact. Artificial stream assays simulated natural conditions and allowed multiple fish of both sexes and species to freely interact. Details of the experimental designs are provided below. Both trial types were conducted previously for a single sympatric population (Roberts & Mendelson, 2017;Williams & Mendelson, 2010). For the current study, we conducted trials with two allopatric populations and an additional sympatric population of E. zonale with respect to E. barrenense. We combined data from these populations with those from the previously studied sympatric population to test for a signature of reinforcement.

| Fish collection and maintenance
We collected E. zonale by kick seine from two allopatric populations and one sympatric population for the current study, between   Kuehne and Barbour (1983) We transported fish to the University of Maryland Baltimore County and housed them in a recirculating aquarium system (Aquatic Habitats, Inc.). Water temperature, conductivity, and pH for fish housing replicated the natural habitat (temperature = 12°C; conductivity = 550-650 μs; pH = 8.3). We separated individual fish into gravel-lined aquaria that were visually isolated from individuals of the opposite sex, species, and population and maintained a 12:12-hr light/ dark cycle. Fish were fed a diet of live black worms provided once daily.

| Dichotomous choice assays
Dichotomous choice assays followed previously published methods from Roberts and Mendelson (2017) and Williams and Mendelson (2010). Briefly, a 37.9-liter glass "focal" tank (50L × 25W × 30H cm) was positioned between two 9.6-L glass "stimulus" tanks (30L × 15W × 20H cm) so that the long sides of the stimulus tanks were flush against the short side of the focal tank. The focal tank was marked with two 5-cm "association zones" at either end closest to the stimulus tanks. Trials took place between 27 March-2 May 2017 and 27 March-4 April 2018, coinciding with the natural breeding season (Etnier & Starnes, 1993).
The focal and stimulus tanks were lined with gravel of equal heights and filled with equivalent depths of water from the aquarium housing and individually aerated. Each stimulus tank was illuminated with an incandescent light source (GE Crystal Clear, A19, 100W), and a 91 cm full-spectrum light source (Coralife ® F/W/T5 Aqualight, 21 W Colormax™ bulb, 21°W 6700 K bulb) spanned all three tanks. Prior to a trial, an opaque partition was placed between each stimulus tank and the focal tanks, then one E. barrenense was placed into a stimulus tank, one E. zonale of the same sex was placed in the other, and a focal E. zonale (opposite sex of stimulus fish) was introduced into the test tank. Once the focal fish began free-swimming activity, the opaque partitions were removed and acclimation began. Acclimation was complete after the focal fish entered both association zones and subsequently entered the "neutral zone" (i.e., was not in either association zone). Following acclimation, a 15-min observation period began, and time spent in each association zone was recorded using JWatcher™ V1.0 (Blumstein, Evans, & Daniel, 2000). Association time appears to be a reliable indicator of mating preference in several species of fishes (Aspbury & Basolo, 2002;Gonçalves & Oliveira, 2003;Jeswiet & Godin, 2011;Lehtonen & Lindström, 2008), including darters (Martin & Mendelson, 2013;Williams & Mendelson, 2010), meaning that association time in the current study is likely a reliable proxy for mate preference. Between trials, gravel was rinsed and mixed, and water was replaced in the test and stimulus tanks. We alternated the side of the test tank in which the conspecific individual was placed to help control for experimental side bias.
We tested the preferences of male and female E. zonale from the

| Artificial stream assays
Preferences expressed as association time in dichotomous trials may not result in mate choice under more natural, unrestricted

| Statistical analysis
Strength of preference (SOP) was calculated for each individual using the equation: where T C is the total time spent in the conspecific association zone and T H is the total time spent in the heterospecific association zone over the entire 15-min trial. SOP can range from +1 to −1 indicating a complete preference for conspecific or heterospecific individuals, respectively, whereas a score of 0 indicates no preference. This measure of preference represents the proportional reduction in gene flow relative to expectations under random mating (e.g., a value of 0.3 is interpreted as a 30% reduction in gene flow due to behavioral isolation than expected under random mating; Sobel & Chen, 2014 behaviors were summed, for each species separately, to yield the total number of conspecific-directed behaviors and the total number of heterospecific-directed behaviors in each replicate. We took the average of these summed totals across the three replicates per population and applied these values to Equation 2 to calculate a total isolation index for each species and population. We used a generalized least squared (GLS) model to determine the effect of context (sympatry or allopatry), species, and population on the total isolation index calculated for the artificial stream assays for the LC, MF, and FC populations. Using a GLS allowed us to account for heteroskedastic data (Zuur, Ieno, Walker, Saveliev, & Smith, 2009). All E. barrenense were coded as sympatric with respect to E. zonale, while E. zonale were coded as sympatric or allopatric accordingly. Model selection started with a full model, including all possible terms and interactions, and included trial replicate as a random effect. We then sequentially removed nonsignificant terms to select the fixed effect structure of the final model. The final model included only context (sympatry or allopatry) as explanatory variables for total isolation. All analyses were conducted in R (ver. 3.5.0; R Core Development Team, 2015). Post hoc comparisons were calculated using the package "lsmeans."
The results of our final model for the dichotomous choice trials showed that context and the interaction of sex and population were significant factors in predicting strength of preference (Table 2).
Model results indicate that E. zonale from populations sympatric with E. barrenense had significantly greater preferences for conspecific mates than E. zonale from allopatric populations (Table 2; Figure 3a).
A post hoc t test comparing SOP between sympatric and allopatric populations also revealed that SOP was significantly greater for sympatric than for allopatric populations (t = −3.53, p = .0006; Figure 3a).
The final model also showed a small but significant effect of the interaction of population and sex (Table 2; Figure 3b). This appears to arise from a significant difference in SOP between the sympatric populations for female E. zonale (EF-LN, t 134 = 2.40, p = .02). There was no significant difference in SOP between the allopatric popu-

| Artificial stream trials
The final model found a significant effect of context (i.e., sympatry/allopatry) on total isolation in the artificial stream trials (GLS: t = 2.25, p = .04; Figure 5). Results of model validation found no significant effect of species or interactions between context and species, meaning that total isolation was not explained by one species or the other.
Isolation indices for both E. zonale and E. barrenense were all positive, indicating conspecific bias, and ranged from 0.43 to 1.00 (Table 3). Isolation indices were lower for populations of E. zonale allopatric with respect to E. barrenense than for sympatric E. zonale across all recorded behaviors except male-male chases ( Figure 6).
Interspecific spawning was only observed once throughout all trials and was a result of a group spawn in which a male E. zonale from the allopatric MF population spawned alongside a pair of spawning E. barrenense. Thus, the difference in isolation indices between sympatric and allopatric populations is largely accounted for by differences in male solicitation behaviors (both successful

| D ISCUSS I ON
Our aim was to evaluate geographic variation in mate choice in Etheostoma zonale and determine whether it is consistent with a pattern of reinforcement. We also aimed to determine whether a pattern of reinforcement is more evident in males than females, as was shown in a previous study in the Oligocephalus subgenus of Etheostoma . Our results were consistent with a pattern of reinforcement. We found that E. zonale from populations sympatric with the close congener E. barrenense had a greater preference for conspecific mates than E. zonale from allopatric populations when tested in dichotomous choice trials (Figure 3a). Results of our artificial stream assays further support this finding, with geographic context being the only significant predictor of total isolation and overall stronger isolation indices in sympatric versus allopatric populations ( Figure 5). This signature of reinforcement-greater preference for conspecific mates in populations that are sympatric with a close congener compared to allopatric-was evident for both males and females in the dichotomous mate choice trials in our linear model, but was statistically significant only for females when evaluating each sex independently (Figure 4).
Our results therefore suggest that reinforcement acts on both male and female preferences in E. zonale, but they do not support a universally stronger effect of reinforcement on males in the genus Etheostoma.
Previous studies that address reinforcement in both sexes provide conflicting support for whether male or female preferences are more strongly affected. One study in Drosophila found that while male and female preferences were widespread, female preferences were primarily responsible for isolation between species in sympatry (Yukilevich & Peterson, 2019). That result is consistent with basic sexual selection theory, which assumes that females invest more heavily in reproduction than males and would pay a higher cost for maladaptive mate choice (Andersson, 1994). However, in Etheostoma caeruleum and the E. spectabile species complex (all in the darter subgenus Oligocephalus), males have both heightened conspecific mate preference and aggression bias in sympatric as compared to allopatric populations . In this same group, female preference for conspecific males was determined to be insignificant in both sympatric and allopatric populations Moran, Zhou, Catchen, & Fuller, 2017). Combined with an additional study showing maladaptive effects of hybridization in sympatric species pairs in this subgenus (Moran, Zhou, Catchen, & Fuller, 2018), their results suggest that selection is acting on male, but not female, mate choice in sympatric populations. Our results, which taken together suggest that female preferences are more strongly reinforced in sympatric populations, are therefore more consistent with results from Drosophila than with those of a congeneric species group.
Several notable differences between our focal species pair and species tested by  could account for the disparity between the studies. One difference is methodological, in that the dichotomous trials of  allowed unrestricted access among two males and one female. An unrestricted design allows mate choice to be quantified in a more naturalistic setting, but it can mask female preference if male behavior has a greater effect on mating outcomes (Dougherty & Shuker, 2014). Estimating female preference in unrestricted choice trials therefore must control for male behavior, essentially quantifying the degree to which females mate with males more than expected given male courtship intensity (see also Martin & Mendelson, 2016). In the current study, female preference in the dichotomous choice trials, where males  (2010) and for males from the EF population from Roberts and Mendelson (2017) are restricted from interacting with each other and with the focal female, may be less likely confounded by male behavior. Thus, one explanation for the difference between the two studies in the effect of sympatry on female preference is that female preference was measured in different ways.
Another difference between the studies is the biology of the focal species. Whereas males of E. zonale and E. barrenense differ markedly in nuptial coloration-green vertical bars compared to a black horizontal stripe on a vivid red background ( Figure 1)-species used in  are characterized by similar male coloration, that is, alternating red and blue vertical bars (e.g., Page & Burr, 2011). Ichthyologists can identify distinguishing species-specific color features, but the differences are subtle compared to those of the focal species in our study (e.g., Page & Burr, 2011). Interestingly, previous studies of species in the subgenus Oligocephalus have consistently failed to find female preferences for male color patterns either within or between species (Fuller, 2003;Moran et al., 2017;Pyron, 1995;Zhou, Loew, & Fuller, 2015). Rather, isolation is thought to be maintained by male behaviors in Oligocephalus, with males preferring conspecific over heterospecific females Moran et al., 2017;Zhou et al., 2015). There is also evidence that male coloration functions as an aggressive male-male signal in competition over access to females as opposed to being a target of female selection in that group (Moran et al., 2017;Zhou & Fuller, 2016). In contrast, females in the focal species pair have been shown to prefer conspecific over heterospecific nuptial hue (red vs. green), nuptial pattern (bars vs. stripe), and hue and pattern combined (Williams & Mendelson, 2011). Female E. barrenense also prefer one hue of conspecific red over another, providing evidence of female preference for color variation within species as well (Williams, Gumm, & Mendelson, 2013).
Thus, female preference for male nuptial color may be present in some, but not all, species of darters. In these species, female preferences may be shaped by reinforcement. Further, the focal species in our study diverged approximately 6.5 mya (Williams & Mendelson, 2010), whereas E. spectabile and E. caeruleum diverged approximately 22 mya (Near et al., 2011), suggesting that the two studies may represent different time scales of evolutionary change.
One important result of our study was that the strength of preference for conspecifics differed between the two sympatric populations, such that the signature of reinforcement we observed for females in the dichotomous mate choice trials was only evident in   (Hubbs, 1955(Hubbs, , 1967, and a previous study indicates a potential reduction in F1 hybrid survival (Williams & Mendelson, 2014). Our results therefore warrant further investigation into sex-specific fitness costs to hybridization for this species pair.
In addition to dichotomous choice assays, which allowed us to separately test the strength of male and female preferences, artificial stream assays allowed us to observe courtship and spawning behaviors in freely interacting individuals. We found that trials of sympatric populations yielded greater total isolation indices than trials with E. zonale from allopatric populations ( Figure 5). The difference in total isolation between sympatric and allopatric trials appears to be accounted for by differences in male solicitation behaviors. Indeed, interspecific spawning events were rare even in allopatric trials, with only one instance observed across all trials (Table 3; Figure 6). For comparison, isolation indices for the EF population (Williams & Mendelson, 2010) are presented in Table 3. The largest differences between the previously studied and the current sympatric populations are between male-male chase behaviors and unsuccessful male unsuccessful courtship attempts. In the current study, there was stronger conspecific bias (greater I) in male-male aggression and more unsuccessful heterospecific solicitations (lower I) than in the previous study. This difference may be due to differences in experimental design. Most notably, only one species of female was present with males of both species in each of the artificial stream assays in the previous study (Williams & Mendelson, 2010), while the current study had both species of males and females included within each replicate for artificial stream assays. Authors noted that the heterospecific males (relative to the female group) exhibited minimal social interactions compared to the conspecific males, often attempting to escape the stream arena. In the current study, males and females of both species were included in artificial stream assays, and we did not observe either species consistently exhibiting asocial or escape behaviors, which may have increased the likelihood of heterospecific solicitations, albeit unsuccessful, in our experimental design.
In conclusion, we found stronger overall preference for conspecific mates in populations of Etheostoma zonale that are sympatric versus allopatric with respect to E. barrenense. This classic signature of reinforcement suggests that speciation can be completed in darters by selection for mate choice in sympatry. Reinforcement can shape both male and female preferences (Servedio, 2007), but few studies of reinforcement directly compare both sexes, often focusing exclusively on female preferences, especially in sexually dimorphic species. We found that preferences for conspecifics may be strengthened in sympatry for both males and females in the focal species pair due to reinforcement. Female preferences appear to be more strongly reinforced than male preferences, and previous studies of female choice for male nuptial coloration in these species suggest this could be due to reinforcement of female preference for elaborate male ornaments (Williams et al., 2013;Williams & Mendelson, 2011). In other darter species, however, behavioral isolation is thought to be maintained and reinforced primarily through male choice and male competition (Fuller, 2003;Moran et al., 2017;Zhou et al., 2015).
Given that the prevalence of male mate choice is becoming more broadly studied across animal taxa (Edward & Chapman, 2011) and appears pervasive in the sexually dimorphic genus Etheostoma (Ciccotto, Gumm, & Mendelson, 2013;Mendelson et al., 2018;Moran et al., 2017;Roberts & Mendelson, 2017;Zhou et al., 2015), future work should consider how both male and female preferences contribute to speciation in sympatry and allopatry.

ACK N OWLED G M ENTS
We would like to thank S. Hulse, C.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

AUTH O R CO NTR I B UTI O N S
NSR and TCM conceived the study and conducted field collections.
NSR conducted dichotomous choice and artificial stream trials and conducted all statistical analyses. The manuscript was written by NSR with input from TCM.

O PE N R E S E A RCH BA D G E
This article has earned an Open Data Badge for making publicly available the digitally-shareable data necessary to reproduce the reported results. The data is available at https ://doi.org/10.5061/ dryad.18931 zcsg.