Linking glucocorticoid variations to monthly and daily behavior in a wild endangered neotropical primate

Analytical and biological validation of a noninvasive measurement of glucocorticoid metabolites in feces of Geoffroy's spider monkeys (Ateles geoffroyi)

We report on an analytical and biological validation of a commercial cortisol enzyme immunoassay (EIA) to measure glucocorticoids (GC) in feces of Geoffroy's spider monkeys (Ateles geoffroyi). Validation of endocrinological methods for each sample matrix and study species is crucial to establish that the methods produce reliable results. For the analytical validation of the EIA, we assessed parallelism, accuracy, and precision. We carried out a biological validation based on three well-studied GC patterns with the following predictions: (1) increased fecal GC metabolite (fGCM) concentrations after veterinary intervention; (2) increased fGCM concentrations during early morning hours; and (3) higher fGCM concentrations during gestation than in other female reproductive states. For the first prediction, we sampled feces of two zoo-housed females 2 days before, the day of, and 2 days after a veterinary intervention. For the second prediction, we analyzed 284 fecal samples collected from 12 wild males using a linear mixed model (LMM). For the third prediction, we analyzed 269 fecal samples of eight wild females using an LMM. Analytical validation revealed that the EIA showed parallelism, was accurate, and precise within each assay. However, there was elevated variation in between-assay precision. The biological validation supported all predictions: (1) the two zoo-housed females showed a substantial increase in fGCM concentrations 2.5 and 11 h after veterinary intervention; (2) there was a negative effect of sample collection time on fGCM concentrations (i.e., higher concentrations during early morning); (3) gestating females had significantly higher fGCM concentrations than lactating females. Thus, we analytically validated the commercial EIA and, despite between-assay variation, we were able to find three biologically relevant GC signals in captive and wild settings, and in males and females. We are therefore confident that the method can be used to noninvasively address behavioral endocrinology questions in Geoffroy's spider monkeys.