Repeatable intra-individual variation in plasma testosterone concentration and its sex-specific link to aggression in a social lizard

Abstract

Individual hormone profiles can be important generators of phenotypic variation. Despite this, work on the consequences of hormone profiles has traditionally ignored the large inter-individual variation within natural populations. However, recent research has advocated the need to explicitly consider this variation and address its consequences for selection. One of the key steps in this process is examining repeatability in hormone profiles and their links to behavioral traits under selection. In this study we show that individuals within a free-ranging population of the Australian lizard Egernia whitii exhibit temporal repeatability in their circulating baseline testosterone concentrations as well as their aggressive response towards conspecific intruders. Furthermore, we show significant, sex-specific links between testosterone and aggression. Specifically, testosterone and aggression is negatively linked in males, while there is no relationship in females. As conspecific aggression has significant consequences for fitness-related traits (parental care, mating strategies) in this species, inter-individual variation in testosterone concentrations, through their effects on aggression, could have important implications for individual fitness. We discuss the potential causes and consequences of hormonal repeatability as well as provide explanations for its sex-specific links with aggression. Specifically, we suggest that these patterns are the result of alternative hormonal pathways governing aggression within Egernia and may indicate a decoupling of aggression and testosterone across the sexes.

Introduction

Individuals of the same species and sex may differ profoundly in both hormone profiles and hormone-related phenotypic traits. Although individual variation in circulating hormone concentrations has received increased attention since the postulation of the challenge hypothesis (Wingfield et al., 1990), these differences are usually studied in terms of environmental variation, especially social stimulation (Oliveira, 2004). Inter-individual variation of hormone levels characterizing individuals has so far rarely been studied, let alone in relation to inter-individual variation in behavior, despite repeated pleas for this (Zera et al., 2007, Wingfield et al., 2008, Williams, 2008) and its potential relevance to the field of animal personalities (Sih et al., 2004).

A crucial first step to understanding the consequences of inter-individual variation in hormonal profiles is to identify the extent to which individual profiles are repeatable across time and/or contexts (i.e., understanding intra-individual variation; Kempenaers et al., 2008, Williams, 2008). Identifying intra-individual repeatability in hormone profiles is crucial for two reasons. Firstly, it represents an initial step in disentangling the sources of variation in hormone profiles between individuals, i.e., the extent to which inter-individual variation reflects differences in underlying quality, as the result of heritable genetic variation or developmental programming (Schwabl, 1993, King et al., 2004), or differences in the social and/or non-social environment (Oliveira, 2004). Secondly, once repeatable individual differences in particular hormones or hormone profiles are identified, it allows one to further examine the potential links with variation in functionally significant hormone-dependent traits. For example, hormone repeatability after a response to hormone treatments in the dark-eyed junco (Junco hyemalis) is linked to variation in aggression which has significant implications for individual fitness via its effects on parental and territorial behavior (Jawor et al., 2006, McGlothlin et al., 2007). Combined, this approach allows for a better understanding of the role of the endocrine system in the organization and evolution of integrated phenotypic packages, or suites of traits (e.g., behavioral syndromes/animal personalities; Sih et al., 2004, Adkins-Regan, 2005, Kempenaers et al., 2008, Williams, 2008, Wingfield et al., 2008). Establishing a proximate basis for animal personality is crucial to our understanding of how these suites of traits are generated and the extent to which they can be uncoupled (Sih et al., 2004, Stamps and Groothuis, 2010).

Egernia whitii is a social lizard species in which aggression, an ecologically important behavioral trait, exhibits strong repeatability within and between breeding seasons (Sinn et al., 2008, While et al., 2009a). Furthermore, an individual's aggressive phenotype is linked to a number of fitness-related traits, including the proportion of extra-pair offspring (While et al., 2009a) and offspring survival (Sinn et al., 2008). Crucial to our understanding of the causes and consequences of this behavioral repeatability is knowledge of its proximate mechanisms (Sih et al., 2004). Therefore, the aim of this study is (i) to document repeatability of variation in baseline testosterone concentrations, a key mediating hormone of aggression within many vertebrate systems (e.g., Wingfield et al., 1987) and (ii) to assess the extent to which aggression and testosterone are linked in both males and females.