Males and females share many traits. However, their adaptive optima can be inherently different as a result of sex-specific selective pressures (Arnqvist and Locke, 2005). Sex differences in the costs and benefits of basic life history traits (Bateman, 1948; Clutton-Brock, 1989) may lead to sexual conflicts, as traits that are beneficial for one sex can be harmful to the other (Chapman et al., 2003; Chippindale et al., 2001). Among the proximate mechanisms underlying the differences between the sexes are circulating steroid hormones that mediate the development of suites of coordinated traits. Testosterone is essential in developmental, structural, and physiological processes in both sexes (Staub and De Beer, 1997; Wingfield et al., 2001). It has a crucial role in both male and female fetuses, regulating growth, maintenance and function of reproductive tissues, development of the brain, gonads, and regulation of receptors (Zambrano et al., 2014). However, its optimal circulating concentration may differ between the sexes, as well as its effects on morphology, physiology and behavior (Mills et al., 2012). In adult males, testosterone has been linked to functional reproductive traits, such as the age of sexual maturity, testes size, sperm characteristics, sexual signals such as antlers, ornamentation, and body size, and behavioral traits such as territoriality, dominance, and competition, and ultimately reproductive success (Adkins-Regan, 2005; Folstad and Karter, 1992; Mills et al., 2009; Preston et al., 2012; Thompson et al., 2012). Since testosterone is involved in the development of male secondary sex characteristics and traits expressing male fertility status, high male testosterone may be favored by sexual selection (Hillgarth et al., 1997), despite the major fitness costs involved with high circulating testosterone for both sexes (Buchanan et al., 2003; Folstad and Karter, 1992; Ketterson et al., 2005; Ketterson and Nolan, 1992; Muehlenbein and Bribiescas, 2005; Wingfield et al., 2001, Wingfield et al., 1990). In females, although testosterone is also essential in basic processes, such as maintaining age-related fertility (Walters et al., 2008), high levels may be deleterious to fitness and reproduction (Cain and Ketterson, 2013; Clotfelter et al., 2004; Gerlach and Ketterson, 2013; McGlothlin et al., 2004; O'Neal et al., 2008; Rosvall, 2013). For example, high testosterone in females has been linked to adverse reproductive outcomes, including an increased risk of infertility and cycle disruption, fetal loss, preterm birth, low birth weight, and a high incidence of ovarian dysfunction (Balen et al., 1995; Fuller et al., 1970; James, 2015; Smith et al., 1979; Steinberger et al., 1981), as well as inferior social and copulation status in some species (Koren et al., 2019; Koren and Geffen, 2009). In both sexes, testosterone is often described as a physiological mediator of the trade-offs between reproduction, growth, survival and parental effort (Hau, 2007; Ketterson and Nolan, 1999, Ketterson and Nolan, 1992; McGlothlin et al., 2007), which are central to life-history (Roff, 1992; Stearns, 1998).

Selection can act on a trait if it varies among individuals, has a significant genetic base, and affects fitness (Falconer and Mackay, 1996). Testosterone is involved in mediating fitness-related traits in both males and females (Hau, 2007; Ketterson et al., 2005; Mills et al., 2009). Although it varies across life-history stages and environmental and social conditions (Wingfield et al., 1990), testosterone shows at least moderately high degree of intra-individual repeatability in some systems (Kraus et al., 2020; Liening et al., 2010; Mutwill et al., 2021, but see Kralj-Fišer et al., 2007; Pavitt et al., 2015). Due to the influence of selection on the genetic architecture underlying testosterone, it is anticipated that genetic factors contribute to its variability, as indicated by heritability estimates (Boake, 2002). Accordingly, a significant heritability estimates of individual testosterone levels are found in both males and females (e.g., Coviello et al., 2011; Flynn et al., 2021; Grotzinger et al., 2018a; King et al., 2004; Kuijper et al., 2007; Mills et al., 2009; Pavitt et al., 2014; Ruth et al., 2020).

Given the differential effects of testosterone levels on fitness in males and females, selection is expected to favor higher testosterone levels in males, and lower levels in females (Møller et al., 2005). Thus, sexual conflict would ensue if high testosterone (i.e., high fitness) fathers would have high testosterone (i.e., low fitness) daughters, possibly as part of an evolutionary arms race between the sexes over reproduction (Mills et al., 2012). Consequently, assuming sex differences in the effects of high testosterone on fitness outcomes, selection should favor a differential genetic architecture between the sexes. Surprisingly, only a few studies assessed testosterone cross-sex heritability, with conflicting findings that might be attributed to species, age, or methodological differences (Hoekstra et al., 2006; Pavitt et al., 2014; Ruth et al., 2020). A sex-specific architecture of the trait may allow each sex to reach its optimum without negatively impacting the other, thus alleviating sexual conflict.

Quantitative genetics can be used to assess genetic divergence and the potential for sex-specific trait evolution. A trait shared by males and females can evolve in a sex-specific manner through low additive genetic covariance between the sexes, sex differences in additive genetic variances, or both (Cheverud et al., 1985; Lande, 1980; Lynch and Walsh, 1998). The cross-sex genetic correlation between shared male and female traits (rMF) measures the extent of similarity between additive alleles when expressed in both sexes (Bonduriansky and Rowe, 2005; Lande, 1980). When rMF is close to unity, a shared trait is assumed to be controlled by a common genetic architecture. This might constrain one or both sexes from reaching their optimum despite sex-specific selection (Lande, 1980). In contrast, when rMF approaches zero, the sexes differ in genetic architecture for the shared trait or differ in expression of alleles related to this trait, allowing them to evolve independently towards trait optimality (Cox et al., 2017; Lande, 1980). To date, relatively few studies employed a quantitative genetic approach to understand whether and how the genetic architecture for testosterone levels differs between the sexes. While there are a few studies on testosterone heritability in neonates, in red deer (Cervus elaphus) and humans (Caramaschi et al., 2012; Pavitt et al., 2014; Sakai et al., 1991), to the best of our knowledge, there are no studies examining testosterone heritability in utero, despite the appreciated influence of in utero testosterone exposure on both fetal development and fitness in adulthood (e.g., reviews by Ryan and Vandenbergh, 2002; Zambrano et al., 2014). Moreover, no studies examined the cross-sex genetic correlation of in utero testosterone levels.

Here, we used a quantitative genetic approach to examine the genetic component of in utero accumulated testosterone levels and test whether it differs between males and females. Specifically, we assessed i) heritability of testosterone between parent and offspring within the same sex (i.e., mother-daughter and father-son); ii) heritability between parent and offspring of the opposite sex (i.e., mother-son and father-daughter); and iii) cross-sex genetic correlation in free-ranging nutrias (Myocastor coypus). The nutria is a large semi-aquatic rodent native to South America that has a high reproductive potential, including post-gestational estrus, large litters, and year-round receptivity, making it a successful invasive species facing extensive eradication and control efforts worldwide (Carter and Leonard, 2002). The long gestation of this polygynous rodent (approximately 4.5 months) makes it an interesting model for studying the effects of hormones on fetal development (Fishman et al., 2019). The nutria also has a high prevalence of multiple paternity (i.e., fetuses of the same litter are sired by more than one father) (Fishman et al., 2023). That, together with its large litter size, places it as an interesting model for examining fetal testosterone heritability.

We collected pregnant nutria carcasses obtained during eradication efforts at Agamon Hula Park in Israel and quantified their hair testosterone levels. Steroids (e.g., testosterone), are stable in hair, allowing quantification in ancient specimens, including extinct species and archeological samples (Koren et al., 2018; Webb et al., 2010). Unlike saliva and blood concentrations, which are confounded by diurnal variation or hormonal reactivity, hair-testing reflects long-term integrated steroid levels accumulated at the time of growth (Grotzinger et al., 2018a; Koren et al., 2002). Furthermore, while most of circulating testosterone is bound to sex hormone binding globulins (SHBG), and thus may not be bioavailable (Dunn et al., 1981), hair testosterone levels reflect free, unbound, and bioavailable testosterone (Chan et al., 2014; Russell et al., 2012; Slezak et al., 2017; Stalder and Kirschbaum, 2012). In the nutria, hair follicles appear at 85–90 days of gestation (around the beginning of the last trimester of their ∼132 day-long pregnancy; Felipe and Masson, 2008; Sone et al., 2008), and have a sufficient amount of hair for testosterone quantification around the last month of gestation, representing accumulating levels of testosterone in the last trimester of pregnancy (Fishman et al., 2019).

Male nutria fetuses have significantly higher testosterone levels than female fetuses (Fishman et al., 2019). Male fetuses related to fathers monopolizing most or all of the litter paternity have higher testosterone levels than those related to the rare father (the father siring commonly only one fetus in the multiple paternity litter), while female fetuses' testosterone levels does not show such relationship (Fishman et al., 2023). These findings are in line with studies that observed higher reproductive success for differentially larger and more aggressive male nutria (Guichón et al., 2003a, Guichón et al., 2003b; Túnez et al., 2009). In adults, male body size is strongly associated with testosterone levels, however, this association does not exist in females (Fishman et al., 2023). Moreover, higher testosterone levels in nutria females are related to lower fitness (Fishman et al., 2018a). Based on these past findings, we hypothesized that there might be sex-differences in the genetic architecture for testosterone levels in the nutria, reducing the potential sexual conflict.