Effects of reproductive condition and dominance rank on cortisol responsiveness to stress in free-ranging female rhesus macaques

Abstract

The hypothalamic-pituitary-adrenal (HPA) axis modulates individuals' physiological responses to social stress, which is an inevitable aspect of the daily lives of group-living animals. Previous nonhuman primate studies have reported that sex, age, rank, and reproductive condition influence cortisol levels under stressful conditions. In this study we investigated cortisol responses to stress among 70 multiparous, free-ranging female rhesus macaques (Macaca mulatta) on the island of Cayo Santiago, PR. Plasma cortisol samples were collected in two consecutive years under similar conditions. Twenty-two females were sampled both years, and most of those females were lactating in only one of the years. Individual differences in cortisol levels were stable across years, even though reproductive condition changed for most individuals. No relationship was found between age or social rank and cortisol levels. Of the females that changed reproductive conditions, cortisol levels were higher when they were lactating than when they were cycling, and the amount of change in cortisol from cycling to lactating was greatest for low-ranking individuals. Heightened reactivity to stress during lactation may be the result of concerns about infant safety, and such concerns may be higher among low-ranking mothers than among higher ranking mothers. Psychosocial stress and hyperactivation of the HPA axis during lactation can suppress immune function and increase vulnerability to infectious diseases, thus explaining why adult females in the free-ranging rhesus macaque population on Cayo Santiago have a higher probability of mortality during the birth season than during the mating season.

Figure 1

Correlation between 2008 cortisol levels and 2007 cortisol levels.

Figure 2

Mean (±S.E.M.) change in cortisol levels from the cycling condition to the lactating condition. The z-score values are used here because, as described in the text, the 2007 and 2008 mean cortisol values differed significantly, even though there were no obvious differences in the 2007 and 2008 samples. Thus, z-scores were necessary to compare values collected in separate years.