(d) Discussion
Our data support previous observations on the function of push-ups inS. occidentalis . This may be a visual signal displayed towards conspecifics from longer distances and reduces the escalation of conflicts between males differing in status (Sheldahl & Martins 2000). The chromatic properties of the yellow patch significantly differ between the sexes in this population and, specifically, its spectral chroma correlates with the body length of the males (see Megía-Palma et al. 2018b). However, it was not an important predictor of the behavioral traits tested here, suggesting that intrasexual male competition does not influence evolution of the yellow patches in this species. Conversely, an intersexual signal of quality based on the display of yellow patches was previously described in Sceloporus virgatus(Weiss 2006), a closely related species. This suggests that yellow patches in S. occidentalis might also signal individual quality (body size) to conspecifics of both sexes in this population (Megía-Palma et al. 2018b), but not necessarily as an agonistic signal of male aggression (Martins 1993).
The factor analysis of the behavioral traits indicated that Factor 1 correlates with push-ups, lateral compressions, and lunge behavior; three behavioral traits displayed during male agonistic interactions inS. occidentalis . Indeed, the blue ventral patches are displayed towards opponents during the performance of lateral compressions, at a short distance of 1.6 cm between contestants in this study, and this behavior frequently (76.1%) elicited aggressive responses from rivals. Thus, factor 1 will be hereafter referred to as ‘aggressive behavior’.
The comparison of the SELF versus the RIVAL models provided us with a comprehensive view of the relative importance of behavioral traits, color, and infection as predictors of male aggressive behavior during pairwise contests in S. occidentalis . The behaviors summarized in factors 1 (i.e. aggressive behavior) and 3 (i.e. submissive behavior; Table 1) of rival lizards were important predictors, with similar magnitudes, of the aggressive behavior of focal males (see Figs. 1a and 1c). Thus, our analyses suggest that behavioral traits from rivals have an important influence on the aggressive response of focal male lizards, supporting the escalated model of aggression (i.e. Enquist & Leimar 1983) previously proposed for this species (Sheldahl & Martins 2000). However, these relationships were non-linear. Strikingly, this may follow a quadratic pattern with a maximum intensity of escalation at intermediate aggression levels from both contestants, deescalating if the aggression of the rival increases (Fig. 1a). This suggests that push-ups, lateral compression, and lunge behaviors may not necessarily imply physical altercation, but performing these aggressive behaviors is exhausting for the lizards and has likely physiological costs (Schall & Dearing 1987; Marler & Moore 1988). Similarly, an intriguing result is the behavioral pattern observed for focal lizards receiving submissive responses from their opponents; these ‘bully’ males performed the highest intensity of aggressive display with no sign of de-escalation (Fig. 1c). Tail waving (i.e. Factor 4) was also an important predictor of aggression in the RIVAL model, but its relationship with the aggressive behavior from focal males suggested a neutral effect on it (Fig. 1d). This behavior might have a deflective function as proposed for other lizards with similar behavior (Telemeco et al. 2011).
Not all the traits that significantly explained the aggressive behaviors of focal males (i.e. Factor 1) had a positive feedback effect; some behavioral traits likely related with dominance and correlated with Factor 2, such as climbing on top of the rock, licking an opponent’s body, or rubbing the cloaca on the substrate (i.e. marking), reduced the level of aggression of the focal males. This suggests that males may use other cues to assess fighting ability of rivals prior to confrontations, avoiding the costs of fighting with males of higher status (Briffa 2014; Abalos et al. 2016). Chemical scrutiny of the opponent’s scents might be one discrimination mechanism used by the lizards because 40% (487/1217) of the behaviors registered were tongue flicking, which was either performed towards the ground, walls and air, or towards different parts of the opponent’s body (Duvall 1982). This hypothesis is supported by the fact that S. occidentalis can discriminate different individuals by their odor (Duvall 1979). Furthermore, femoral secretions of lizards convey cues of individual quality that may be used by conspecifics to assess the fighting ability of rivals (Moreira et al. 2006; Martín et al. 2007; Martín & López 2007).
In addition, the spectral chroma of the rival’s blue patch had a negative feedback effect on the aggressive escalation of focal males. This suggests that the visual scrutiny of the rival’s blue patches can also dissuade focal males from fighting, confirming that the ventral blue patches of S. occidentalis play the dual role of ornaments and of armaments during social interactions (e.g. Cooper & Burns 1987). This is an interesting finding because when these males were infected by intestinal coccidians of the genus Acroeimeria they exhibited blue patches with lower spectral chroma (Megía-Palma et al. 2018b) and in the present study the males that engaged in fights with lower spectral chroma in their blue patches elicited more intense aggressive behaviors from opponents. Thus, this relationship suggests that intestinal coccidians may ultimately reduce the fitness of infected males through intrasexual competition because infected lizards that receive more aggression may also suffer higher costs of fights, hindering their access to reproduction (Marler & Moore 1989).
Although less informative than behavioral and color traits of the rival, parasites of focal lizards were also important predictors of their own aggressive behavior. Lizards infected by haemococcidians displayed significantly fewer push-ups, lateral compressions, and lunges. In this population, larger males were shown to have a greater prevalence of infection by this parasite (Megía-Palma et al. 2018b). However, the observed effect of this parasite on the behavioral display of the males was independent of lizard body size. The infection byLankesterella might reduce the males’ competitive ability and ultimately hinder their access to reproduction (Schall & Dearing 1987; Schall & Houle 1992). A similar effect was also suggested forPlasmodium mexicanum , a virulent haemosporidian parasite that infects S. occidentalis in other localities (Schall & Dearing 1987). Interestingly, Lankesterella is a parasite that is significantly more frequent in males (Megía-Palma et al. 2018b). In a previous study in this population, we failed to find a significant relationship between color expression in males and Lankesterellainfection (P = 0.06 with the spectral luminance of the blue patch; Megía-Palma et al. 2018b). This does not discard the possibility that females still perceived a difference, because if parasites of the genus Lankesterella hinder the fighting ability of the males, we expect that females could discriminate healthy males, likely based on males winning fights (Schall & Dearing 1987). This might allow for the vertical transmission of genes for parasite resistance into their male offspring (Hamilton & Zuk 1982). These new findings suggest thatLankesterella may be a virulent parasite that plays an important role promoting both inter- and intrasexual selection in this population through its influence on males’ aggressive behavior.
Similarly, males with the higher tick load (> 6 ticks) displayed less push-ups and lateral compressions, and performed fewer lunges. However, this relationship was non-linear. Lizards tended to perform more of these behaviors at low levels of infestation (Fig. 3b). Testosterone may positively influence the aggressive behavior of this lizard species, similar to other closely-related species of the family Phrynosomatidae (Moore 1988; Weiss & Moore 2004), and tick load can co-vary with the concentration of this androgen in S. occidentalis (Pollock et al. 2012). Taken together, this suggests that testosterone may link the aggressive behavior of males to low intensities of tick infestation. However, as commented previously, the opposite was observed at higher intensities of infestation. Ticks draw blood from the lizards and significantly reduce their hematocrit at moderate infestations (> 5 ticks according to Dunlap & Mathies, 1993). As estimated in a previous study in this population, an infestation of only 5 to 7 ticks may provoke, on average, a 10% reduction of hematocrit in the lizards (Megía-Palma et al. 2020). This may impair the rapid oxygen demands of the muscles involved in the sudden movements displayed by the males during agonistic encounters (Schall & Dearing 1987). Lizards with moderate tick infestations, which likely impair the oxygen carrying capacity of the lizards’ blood, may not be able to keep up the displays for an extended duration (Dunlap & Mathies 1993). Thus, our results suggest that moderate infestations by ticks reduce intensity of behavioral display during intrasexual aggressive interactions in this species, supporting that aggressive displays are exhausting for males. This, together with previous results in this population, reaffirms that ticks exert an important selective pressure on the lizards. Tick prevalence and mean intensity of infestation are very high in April, reducing the body mass of some lizards and potentially leading to their death (Megía-Palma et al. 2020), while later in May and June tick abundance is reduced but they affect the behavior of the lizards during male agonistic interactions (this study). This suggests that ticks may exert a selective pressure on this population during April, when massive infestations of up to 61 nymphal ticks per lizard occur (Megía-Palma et al. 2020), and here we revealed a similar pattern that may also contribute to the positive selection of superior competitors or more resistant lizards during the mating season.
Our results confirm the role of the male ventral blue patches as armaments in S. occidentalis , given that its spectral chroma deterred aggressive responses from rivals during pairwise contests. Conversely, males with less chroma received higher intensity of aggression. This suggests that intestinal coccidians affect male agonistic interactions because the males infected by these parasites had less blue chroma (Megía-Palma et al. 2018b). In addition, the negative relationships between infection by haemococcidians and ticks at moderate intensities (> 6), with the aggressive behavior of the males indicate that these parasites reduce the lizards’ ability to fight. This, altogether, suggests a role of parasites in male-male interactions, and in the intrasexual selection of lizards. Parasite resistance might be under selection in this population because the least parasitized males are also the most competitive ones. Mate choice is likely based on males’ behavioral display. Therefore, most aggressive males likely achieve more frequent access to females (Schall & Dearing 1987). We propose here a mechanism mediated by parasites to explain how parasites and male intrasexual competition contribute to the evolution of a structural-based blue patch with dual function in S. occidentalis .