10The extraction and immunoassay of fecal steroids is an increasingly common technique, used in both captive and field studies to 11 provide an approximation of an animal's circulating concentration of hormones through non-invasive methods. Storage of fecal 12 samples is of critical concern because fecal bacteria metabolize fecal steroids within hours after deposit. Ethanol is often used as a 13 preservative for fecal samples stored for several hours at room temperature. We examined the stability of fecal estrogen (fE) and 14 glucocorticoid (fGC) metabolites from baboon (Papio cynocephalus) samples in a 95% ethanol solution at ambient temperature and 15 at )20°C over the course of six months, to determine the effect of storage on steroid concentrations. As measured by radioim-16 munoassay, fE metabolite concentrations increased by 122% at 90 days and fGC metabolite concentrations increased by 92% at 120 17 days. After peaking, both hormones declined to near initial concentrations by 180 days in ambient temperature samples. In samples 18 stored at sub-zero temperatures, fGC metabolite concentrations showed a similar but dampened pattern, while fE metabolite 19 concentrations exhibited small and variable changes with no consistent trend. We discuss explanations for the dynamic pattern of 20 changing fecal metabolite concentrations and offer practical and analytical guidance to field workers for situations in which ideal 21 conditions for stabilizing hormones are not available. Ó 2002 Elsevier Science (USA). All rights reserved. 22

Environmental stressors impact physiology and behavior in many species of animals. These effects are partly mediated through changing concentrations of glucocorticoids, which also vary with reproductive state and social conditions. Prior research has focused largely on seasonal breeders, but the close temporal linkage between season and reproductive state in these species hinders ability to disentangle environmental effects from those of the animal's reproductive status. Here we assessed the effects of environmental challenges on the fecal glucocorticoid (fGC) levels of non-seasonal breeders, female baboons (Papio cynocephalus) of Amboseli, Kenya. Amboseli is characterized by a long dry season, during which food and water become scarce, and by extreme temperatures above 40°C in the shade during some months of the year. We found that after accounting for female reproductive status and individual variability, females exhibited higher fGC levels during the dry season than during the wet season. Further, during the wet season, fGC levels were higher in months of high average daily maximum temperatures. During the dry season, fGC levels were elevated both in hotter months and in months during which the baboons spent a relatively high proportion of time feeding. In spite of these stressors, female baboons reproduce during all months of the year in Amboseli, unlike most other mammals in this environment. This may be attributable to their extreme adaptability, specifically their diversified diet, and their ability to modify their behavior, including their activity profiles.

One source of both bias and ''noise'' in fecal steroid analysis is temporal change in steroid concentrations resulting from duration or conditions of fecal sample storage. However, no consensus currently exists regarding correct procedures or precautions necessary for fecal sample storage, and conditions vary widely within field endocrinology literature. This study considered the effects of shortterm, weeks-long, storage conditions on quantifiable fecal testosterone (fT), glucocorticoids (fGC), estrogens (fE), and progestagen (fP) metabolite concentrations in wild baboons (Papio cynocephalus). Quadruplicate subsamples of fecal samples (n ¼ 29) collected at Amboseli National Park and its environs were subjected to four different storage conditions prior to lyophilization, in order to determine the effects of storage on subsequent steroid concentrations, as assessed by 125 I radioimmunoassays. As expected, the best alternative to the ''initial condition'' of lyophilization at three days after collection was to freeze fecal samples at )20°C for two weeks prior to lyophilization. This storage method resulted in no significant change from initial steroid concentrations for fE, fT, or fP, although fGC showed a slight but significant decline. Storage for two weeks in a charcoal refrigerator caused a mean increase in all four steroid concentrations. However, the results from this storage condition were robust in terms of practical questions asked of the data: fE and fP values still reflected pregnant versus non-pregnant states in baboon females; a fGC profile constructed by age class resembled that created from the samples from the initial condition, although slightly inflated across age classes; and there were only moderate changes in relative fT concentrations across adult males. Knowledge of the effects of storage upon each steroid analyzed within oneÕs study is a necessary component in determining the optimal compromise for storage protocol in a particular research project.

Large gaps exist in our knowledge about common patterns and variability in the endocrinology of immature nonhuman primates, and even normal hormonal profiles during that life stage are lacking for wild populations. In the present study we present steroid profiles for a wild population of baboons (Papio cynocephalus) from infancy through reproductive maturation, obtained by noninvasive fecal analyses. Fecal concentrations of glucocorticoid (fGC) and testosterone (fT) metabolites for males, and of fGC, estrogen (fE), and progestin (fP) metabolites for females were measured by radioimmunoassay (RIA). In males, infancy was characterized by high and declining levels of fGC and fT, whereas steroid concentrations were low during the juvenile years. During the months immediately prior to testicular enlargement, fT (but not fGC) concentration tended to increase. Males that matured early consistently had higher fT and fGC concentrations than those that matured late, but not significantly so at any age. Individual differences in fT concentrations were stable across ages, and average individual fT and fGC concentrations were positively correlated. For females, high and declining levels of fE characterized infancy, and values increased again after 3.5 years of age, as some females reached menarche by that age. Both fP and fGC were relatively low and constant throughout infancy and the juvenile period. During the months immediately prior to menarche, fGC concentration significantly decreased, while no changes were observed for fE levels. fP exhibited a complicated pattern of decrease that was subsequently followed by a more modest and nonsignificant increase as menarche approached. Early- (EM) and late-maturing (LM) females differed only in fP concentration; the higher fP concentrations in EM females reached significance at 4-4.5 years of age. Maternal rank at offspring conception did not predict concentrations of any hormone for either sex. Our results demonstrate the presence of individual endocrine variability, which could have important consequences for the timing of sexual maturation and subsequently for individual reproductive success. Further evaluation of the factors that affect hormone concentrations during the juvenile and adolescent periods should lead to a better understanding of mechanisms of life-history variability.

No abstract