Individual variation and repeatability of testosterone levels

According to the Challenge Hypothesis (Wingfield et al. 1990, American Naturalist 136:829-846) high levels of testosterone are expected to facilitate aggressive behavior, but suppress parental behavior. Recent work by the Grindstaff lab has demonstrated that individual bluebirds in our population exhibit repeatable parental and nest defense behavior. It has been hypothesized that repeatability of hormone levels might underlie repeatable behavior, and given the predictions of the Challenge Hypothesis, one might expect repeatability of testosterone to be a mechanism that maintains repeatable behaviors in our population. To test this hypothess, I conducted gonadotropin-releasing hormone (GnRH) challenges at multiple time points (within parental and aggressive contexts). GnRH challenges are bioassays that measure an individual's natural ability to elevate testosterone, and essentially provide several measures of testosterone (initial levels, GnRH-induced levels, and the difference between these two). Individuals vary among one another in initial levels of testosterone as well as their response to GnRH. Individuals are also repeatable across behavioral contexts in initial testosterone levels (males) and GnRH-induced testosterone levels (males and females). Surprisingly, parental and aggressive behavior were not related to initial testosterone levels, GnRH-induced testosterone levels, or the change in testosterone. Considering sources of individual variation in testosterone levels might elucidate why some species have evolved a decoupling of testosterone, parental care, and aggression.