Sexual dimorphism in an adaptive radiation: Does intersexual niche differentiation result in ecological character displacement?

Abstract Evolutionary radiations are one plausible explanation for the rich biodiversity on Earth. Adaptive radiations are the most studied form of evolutionary radiations, and ecological opportunity has been identified as one factor permitting them. Competition among individuals is supposedly highest in populations of conspecifics. Divergent modes of resource use might minimize trophic overlap, and thus intersexual competition, resulting in ecological character displacement between sexes. However, the role of intersexual differentiation in speciation processes is insufficiently studied. The few studies available suggest that intersexual niche differentiation exists in adaptive radiations, but their role within the radiation, and the extent of differentiation within the organism itself, remains largely unexplored. Here, we test the hypothesis that multiple morphological structures are affected by intersexual niche differentiation in “roundfin” Telmatherina, the first case where intersexual niche differentiation was demonstrated in an adaptive fish radiation. We show that sexes of two of the three morphospecies differ in several structural components of the head, all of these are likely adaptive. Sexual dimorphism is linked to the respective morphospecies‐specific ecology and affects several axes of variation. Trait variation translates into different feeding modes, processing types, and habitat usages that add to interspecific variation in all three morphospecies. Intrasexual selection, that is, male–male competition, may contribute to variation in some of the traits, but appears unlikely in internal structures, which are invisible to other individuals. We conclude that intersexual variation adds to the adaptive diversity of roundfins and might play a key role in minimizing intersexual competition in emerging radiations.

Previous studies on nototheniids, centrarchids, and cottid fishes have shown that buccal cavity size can differ according to prey size and mobility (Carlig et al., 2018;Carroll et al., 2004;Cook, 1996).
The buccal cavity is crucial for generating a suction pressure that draws prey items through the mouth opening, and its diameter limits the maximum prey size (Carlig et al., 2018;Carroll et al., 2004;Cook, 1996;Mihalitsis & Bellwood, 2017). A large buccal cavity is advantageous for suction feeders feeding on elusive prey because it can produce a higher pressure gradient, which is effective over distance. Although these adaptive patterns have been identified in several fish radiations, they have mainly been investigated on an interspecific level. However, in radiations where intersexual niche differentiation is documented, these patterns might also be detectable on an intersexual level.
"Roundfin" Telmatherina are a small monophyletic group within the radiation of sailfin silversides and are endemic to ancient Lake Matano located in the central highlands of Sulawesi ( Figure 1) (Herder, Nolte, et al., 2006;Herder et al., , 2008von Rintelen et al., 2012). Roundfins consist of three morphospecies, Telmatherina antoniae "small", Telmatherina antoniae "large", and Telmatherina prognatha (Kottelat, 1991). According to distancebased divergence estimates and molecular clock analyses, the initial divergence of these morphospecies occurred around 1 My (Stelbrink et al., 2014). However, AFLP genotyping revealed that reproductive isolation among the three morphospecies is substantial but incomplete (Herder, Nolte, et al., 2006;Herder Pfaender & Schliewen, 2008;. Roundfins show significant interspecific differences in body and head shape, and a pronounced sexual dimorphism (Pfaender et al., 2011;Wasiljew et al., 2020). All three morphospecies occupy different microhabitats and specific trophic niches ( Figure 1) (Herder et al., 2008;Pfaender et al., 2011): T. antoniae "small" is a pelagic, predominantly planktivorous suction feeder; T. antoniae "large" is a predominantly benthic, mollusk-eating suction feeder; and T. prognatha is a semipelagic, mainly fish eating ram feeder. Roundfin Telmatherina were also the first case where intersexual niche differentiation was demonstrated in an adaptive fish radiation (Pfaender et al., 2011). Two of the three morphospecies (T. antoniae "large" and "small") show intersexual niche differentiation in trophic ecology, while T. prognatha does not (Pfaender et al., 2011). Male T. antoniae "small" take a significantly higher portion of terrestrial insects compared with females, which are more specialized on copepods. Male T. antoniae "large" consume a higher percentage of mollusks than females, which take a higher amount of terrestrial insects ( Figure 1) (Pfaender et al., 2011). Other dietary components are rather negligible in these two morphospecies (Pfaender et al., 2011). However, these intersexual niche differences have not been linked to morphological structures relevant for prey processing and habitat usage so far.
Here, we test the hypothesis that multiple morphological structures are affected by intersexual niche differentiation in roundfin Telmatherina. Variation in three structural components-the opercle, the pharyngeal jaw, and the buccal cavity-which are directly linked to prey capture, prey processing, and habitat usage in fishes, was studied with µ-CT Imaging. This technique enables one to investigate particularly small-scaled variation of internal structures with great detail and precision without damaging the samples (Adams et al., 2004;Kaliontzopoulou, 2011;Wake, 2012;Wasiljew et al., 2020). Interspecific and intersexual variation was analyzed using different three-dimensional morphometric approaches ranging from classical measurements of distances to landmark-free geometric, morphometric analyses. We hypothesized that the opercle, the pharyngeal jaw, and the buccal cavity are adaptive in roundfins, with specific adaptations to resource use in the respective species and sexes. Further, we predicted that the degree of intersexual variation should coincide with the degree of intersexual niche differentiation in each morphospecies.

| Material and µ-CT imaging
The present study was based on formalin-fixated roundfin Telmatherina specimens that were available from collection material.
These were obtained in the dry season of 2002 from three locations around Lake Matano's shoreline, using gill nets ( Figure 1). Since the specimens used in this study were gathered from museum collection material, no living animals were sampled, killed, harmed, or treated in any other way for this paper.
The skulls of 13 specimens of each morphospecies T. antoniae "small", T. antoniae "large", and T. prognatha were used for 3D µ-CT analyses. µ-CT scanning was performed with Skyscan 1272 and Skyscan 1173 scanners (Bruker). All specimens were scanned in

| Classical and geometric morphometrics
To identify variation in the opercle bone and the pharyngeal jaw, linear morphometric measurements and geometric morphometric analyses were conducted based on surface-rendered 3D models created by the software Checkpoint ver. 17.04.21 (Stratovan Corporation) out of µ-CT tiff image stacks. The following traits of the cranial skeleton were quantified by linear measurements: skull length, left opercle height, left opercle length, left opercle circumference, left opercle surface area, lower right pharyngeal jaw length, lower right pharyngeal jaw width, lower right pharyngeal jaw height, F I G U R E 1 Indonesia, Sulawesi, and Lake Matano with the three endemic roundfin morphospecies T. antoniae "small", T. antoniae "large", and T. prognatha. Adult, reproducing males and females are pictured with key aspects of their ecological differentiation. Nutrition size corresponds to the respective relevance in diet composition. Map by T. von Rintelen, modified (with permission). This figure has been designed using resources from Freepik.com and lower right pharyngeal jaw circumference. The number of teeth on the right lower pharyngeal jaw was counted. All measurements were carried out with the software Checkpoint.
In order to test for interspecific and intersexual shape differences, 14 landmarks were placed at homologous points on the pharyngeal jaws of the 30 unstained µ-CT scanned specimens ( Figure 2a). The outline shape and circumference of the left opercle ( Figure 2b) and the pharyngeal jaw were analyzed with 80 semilandmarks. Patches were used to measure the surface area of the opercle in order to quantify its overall size between species and sexes.

| Buccal cavity measurements
All classical morphometric measurements and geometric morphometric analyses of the buccal cavity were based on surface-rendered 3D models created by Amira ver. 6.5.0 by Thermo Fisher Scientific (Stalling et al., 2005) out of µ-CT tiff image stacks. In order to quantify interspecific and intersexual variation in buccal cavity size and shape, the 30 stained specimens were used for creating volumerendered models of the cranial region with the software Amira.
Surface-rendered models of the buccal cavity were created with the help of the semiautomatic segmentation tool of Amira (Figure 2c,d).
Background artifacts were removed by applying the "remove islands" and "fill holes" options of Amira on the segmented 3D model. The length, width, height, and volume of the buccal cavity were measured for every prior-stained specimen. Two female specimens of T. prognatha were removed from the analysis because they showed deformations of the buccal cavity due to a slightly opened mouth.
Due to the rather featureless structure of the buccal cavity, surface scans were used instead of landmarks to describe its shape. A landmark-free shape analysis of the buccal cavity was performed by the Generalized Procrustes Surface Analysis (GPSA) software package in Java executable (version 20200722 provided by B. J. Pomidor upon personal request) (Pomidor et al., 2016;Slice, 2013). Surface renders were superimposed through iterative closest point (ICP) algorithm. After the superimposition, the homologous point coordinates were subjected to dimension reduction and the principal axis scores were calculated for the further analysis (Pomidor et al., 2016).

| Statistical analyses
Bivariate linear models were performed for the absolute measurements of the opercle, pharyngeal jaw, and buccal cavity in order to control for size in each trait. The absolute measurements of the opercle and the buccal cavity were regressed with skull length. The absolute measurements of the pharyngeal jaw were regressed with overall pharyngeal jaw width. The absolute number of pharyngeal teeth was regressed with pharyngeal jaw circumference. In order to test for significant differences between species and sexes, the resulting residuals of each bivariate linear model were used to perform one-way ANOVAs with Tukey's pairwise tests in the software PAST ver. 3.22 (Hammer et al., 2001). Species and sexes were tested simultaneously, resulting in six groups per model and trait.
The shape data of the opercle and pharyngeal jaw were analyzed thin-plate spline interpolation performed in the software PAST ver.
The resulting scores of the PCA and EFA in the three first axes were used to perform a MANOVA and a Tukey's pairwise test with PAST in order to test for significant differences between species and in the dataset. In order to test for intersexual shape differences, this procedure was repeated for each species dataset individually, instead of a pooled-species dataset. This was done to avoid the disproportionate influence of more variable species on the principal axes over the less variable ones.

| RE SULTS
The skulls of T. antoniae "large" (mean: 18.18 mm) and T. prognatha Size variances for different structures and parameters were unequally portioned between species and sexes. All the absolute and the majority of relative size measurements were more divergent interspecifically than intersexually. The exceptions were the relative height and surface area of the opercle, the relative circumference of the pharyngeal jaw, and the relative height of the buccal cavity.
In these parameters, intersexual variance exceeded the variance among species. Detailed size variance ratios are provided in Dryad.
All analyzed structures differed significantly in at least one size parameter between morphospecies. Interspecific size variation was most distinct for the pharyngeal jaw, while intersexual size variation was most distinct for the opercle. Shape variation was most distinct for the buccal cavity among species and sexes. Intersexual size differences were significant in both T. antoniae morphospecies but not in T. prognatha. Intersexual differences in shape were present in all three morphospecies.
No significant differences in opercle size were identified between the latter two morphospecies. The shown values of relative length and circumference also reflect the patterns for the remaining not visualized parameters. In contrast, the morphospaces of the opercle shape EFA showed a large overlap of all three morphospecies without any significant differences in shape within the first three axes ( Figure 5a). However, the opercle outline of T. antoniae "small" could be distinguished from the other morphospecies by its round shape (F = 4.62; p = <.05). Telmatherina antoniae "large" and T. prognatha shared a rather triangular-shaped opercle ( Figure 6).
This discrepancy between size and shape was also detected among sexes. Male and female T. antoniae "small" and T. prognatha did not differ in relative or absolute opercle size (Figures 3a,b and 4a

| Pharyngeal jaw
Relative pharyngeal jaw size differed significantly in height, width, and relative number of teeth (Q = 19.25; p = <.05) but not in length or circumference between species (Figure 3c,d).
The shown values of relative width and number of teeth were the most discriminative parameters. In line with the proportion of mollusks in its diet (Figure 1), T. antoniae "large" was characterized by a sturdy pharyngeal jaw with a high relative number of teeth. As predicted for a piscivorous predator (Figure 1

| Buccal cavity
Morphospecies differed in relative buccal cavity volume, but not in any of the relative linear measurements (Figure 3e Intersexual variation in relative and absolute buccal cavity size was significant within T. antoniae "small" and T. antoniae "large". Compared with females, relative buccal cavity height was higher in male T. antoniae "small", consistent with the higher percentage of copepods in their diet (Figure 1), but lower in male T. antoniae all morphospecies (Figure 5c), but shape variation was only significant in T. antoniae "small" and "large" (F = 13.95; p = <.05).
In order to identify ecological-based intersexual divergent selection, Among Lake Matano's roundfins, sexual dimorphism in ecologically relevant traits is most pronounced in T. antoniae "large". Males have relatively larger opercles (Figure 3a,b), a lower number of teeth, more sturdy pharyngeal jaws (Figures 3d and 5b), and flatter buccal cavities than females (Figures 3e,f and 5c). This coincides with sexspecific trophic profiles: Males feed more pronounced on mollusks than females, while females feed to a higher extent on insects than males (Pfaender et al., 2011). The conspicuously sturdy pharyngeal jaws of males with a low number of large teeth (Figures 3c,d and   5b) are considered advantageous for dealing with hard-shelled prey (Burress, 2016;Burress et al., 2016Burress et al., , 2018Hellig et al., 2010), while a large buccal cavity, as it occurs in female T. antoniae "large", is considered advantageous for catching elusive prey via a suction feeding mode (Carroll et al., 2004). The relatively larger opercle of male T. antoniae "large" is discussed to be an adaptation to benthic suction feeding (Wilson, Colombo, et al., 2013;Wilson et al., 2015;Wilson, Furrer, et al., 2013), enhancing active ventilation of the gills, which is considered advantageous for reduced motility in static waters (Kimmel et al., 2008;Wilson, Colombo, et al., 2013;Wilson, Furrer, et al., 2013). This fits the more benthic lifestyle of male T. antoniae "large", compared with female conspecifics (Pfaender et al., 2011), and matches findings of an adaptive sexual dimorphism in sticklebacks, which likewise differ in the use of both benthic versus limnetic habitats and opercle traits (Albert et al., 2008;McGee & Wainwright, 2013).
In contrast to T. antoniae "large", male and female T. antoniae "small" differ in size and shape of the buccal cavity (Figures 3e   and 5c), but not in opercle or pharyngeal jaw traits (Figures 3a-d   and 5a,b). Males have significantly higher and deeper buccal cavities than females (Figures 3e and 5c). Again, these findings match the sex-specific trophic niches: Male T. antoniae "small" feed to a larger extent on insects than females, which are predominantly zooplanktivores (Pfaender et al., 2011). Both prey types differ significantly in size (Pfaender et al., 2011;Wainwright & Bellwood, 2002), which matches the differences detected in buccal cavity size and shape (Figures 3e and 5c). The diameter of the buccal cavity generally limits both the maximum prey size and the suction pressure (Carlig et al., 2018;Carroll et al., 2004;Cook, 1996;Mihalitsis & Bellwood, 2017). Therefore, large buccal cavities are advantageous for fish species feeding on large prey items (Carroll et al., 2004;Mihalitsis & Bellwood, 2017). Male T. antoniae "small" might benefit from a larger buccal cavity compared with females since they take a higher percentage of insects, which are substantially larger than zooplankton (Pavlov & Kasumyan, 2002;Pfaender et al., 2010Pfaender et al., , 2011Wainwright & Bellwood, 2002). The studies of Herler et al. (2010) and Ronco et al. (2019) reported sexual dimorphism in the buccal cavity of mouth-brooding cichlid fishes. However, they linked intersexual variation to parental care but not to different trophic niches. The present case is hence, to the best of our knowledge, the first study documenting sexual dimorphism in buccal cavity size and shape in a non-mouth-brooding fish radiation. This finding might support the ecological relevance of the buccal cavity in fish radiations. The absence of intersexual variation in opercle and pharyngeal jaw traits (Figures 3a-d and 5a,b) may be explained by the generally similar requirements for taking insects and zooplankton (Pavlov & Kasumyan, 2002;Pfaender et al., 2010;Wainwright & Bellwood, 2002): Both prey types occur in the pelagic zone and share a similar texture (Pfaender et al., 2010(Pfaender et al., , 2011Wainwright & Bellwood, 2002). Thus, this trophic niche partitioning most likely does not affect intersexual variation in the opercle and pharyngeal jaw, which are linked to habitat usage and prey processing (Burress, 2016;Burress et al., 2016Burress et al., , 2018Hellig et al., 2010;Kimmel et al., 2008;Wilson, Colombo, et al., 2013;Wilson et al., 2015;Wilson, Furrer, et al., 2013).
Conspicuously, the present study did not reveal indications for morphological differentiation among male and female T. prognatha (Figures 3 and 5). However, morphology also meets predictions derived from trophic ecology in this species (Pfaender et al., 2011). The absence of intersexual variation appears plausible since both sexes share similar trophic and habitat niches (Pfaender et al., 2011).

| Does intersexual niche differentiation result in ecological character displacement?
The degree of sexual dimorphism detected here largely meets predictions derived from niche segregation in male and female roundfins (Figures 1, 3, and 5). It ranges from the absence of differences in ecologically relevant traits in T. prognatha, the species lacking intersexual niche differentiation, to T. antoniae "large", where both intersexual trophic niches and trait segregation are most pronounced (Pfaender et al., 2011).
Likewise, sexual dimorphism in the internal structures pharyngeal jaw and buccal cavity has probably evolved under ecological selection pressure, since these structures are of relevance for food acquisition (buccal cavity) and processing (pharyngeal jaw).
In contrast to the majority of morphological traits analyzed in roundfins so far (Herder et al., 2008;Pfaender et al., 2011), both are nonvisible and thus unlikely to serve in signaling for potential mates or competitive males (Bolnick & Doebeli, 2003;Ronco et al., 2019). Nevertheless, we cannot exclude the possibility that sexual selection affects these structures in a nonvisual way, that is, through other signaling pathways or through allometric effects of sexual size dimorphism. However, as patterns of size and shape variation in the opercle, pharyngeal jaw, and buccal cavity detected in sticklebacks, cichlids, and other fish radiations (Albert et al., 2008;Burress, 2016;Burress et al., 2016Burress et al., , 2018McGee & Wainwright, 2013) are also present in roundfins, it seems plausible that the identified variation is predominantly a result of ecological- corresponds to different diet compositions between males and females (Carroll et al., 2004;Cook, 1996;Mihalitsis & Bellwood, 2017).

| CON CLUS IONS
We demonstrate that roundfin morphospecies and sexes differ significantly in multiple ecologically relevant traits affecting prey capture, prey processing, and habitat use. As interspecific and intersexual variations meet patterns of niche differentiation reported in roundfins (Herder et al., 2008;Pfaender et al., 2011), these differentiations are likely adaptations to different ecological niches. Since the analyzed structural components are ecologically relevant and, in the case of the pharyngeal jaw and the buccal cavity, are invisible for other individuals, it seems unlikely that sexual selection is responsible for the intersexual variation documented here (Bolnick & Doebeli, 2003;De Lisle, 2019;De Lisle & Rowe, 2017;Ronco et al., 2019). It rather appears plausible that ecological-based intersexual divergent selection is the main driver for the revealed intersexual variation in roundfins. The intersexual ecological character displacement in T. antoniae "small" and T. antoniae "large" likely minimizes trophic and habitat overlap and thus intersexual competition for ecological resources.
Intersexual morphological differentiation adds to the adaptive diversity of roundfin Telmatherina and might play a key role in minimizing intersexual competition in emerging radiations. Further research is also needed on other systems to deepen our knowledge of the role of intersexual niche differentiation in speciation processes.

ACK N OWLED G M ENTS
We thank the Heinrich-Böll-Stiftung e.V. for funding the corresponding author B. D. Wasiljew via a PhD scholarship. The study strongly benefited from previous work funded by the Deutsche Forschungsgemeinschaft (HE 5707/2-1). We are thankful to C. Koch, P. Rühr, J. Vehof, and S. Wesel for technical support and advice concerning 3D imaging techniques. D. Samadi contributed by performing some of the measurements and parts of the landmark placement.
The late R. K. Hadiaty enabled earlier fieldwork for obtaining the focal specimens in Indonesia. The manuscript greatly benefited from discussions and constructive suggestions by J. Flury, L. Hilgers, J.

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