Measuring salivary analytes from free-ranging monkeys

Studies of large free-ranging mammals have been revolutionized by non-invasive methods for assessing physiology, which usually involve the measurement of fecal or urinary biomarkers. However, such techniques are limited by numerous factors. To expand the range of physiological variables measurable non-invasively from free-ranging primates, we developed techniques for sampling monkey saliva by offering monkeys ropes with oral swabs sewn on the ends. We evaluated different attractants for encouraging individuals to offer samples, and proportions of individuals in different age/sex categories willing to give samples. We tested the saliva samples we obtained in three commercially available assays: cortisol, salivary alpha amylase, and secretory immunoglobulin A. We show that habituated free-ranging rhesus macaques will give saliva samples voluntarily without training, with 100% of infants, and over 50% of adults willing to chew on collection devices. Our field methods are robust even for analytes that show poor recovery from cotton, and/or that have concentrations dependent on salivary flow rate. We validated the cortisol and SAA assays for use in rhesus macaques by showing aspects of analytical validation, such as that samples dilute linearly and in parallel to assay standards. We also found that values measured correlated with biologically meaningful characteristics of sampled individuals (age and dominance rank). The SIgA assay tested did not react to samples. Given the wide range of analytes measurable in saliva but not in feces or urine, our methods considerably improve our ability to study physiological aspects of the behavior and ecology of free-ranging primates, and are also potentially adaptable to other mammalian taxa.

Research Highlights
► We give methods for sampling saliva from free-ranging monkeys without capture.
► Free-ranging rhesus macaques gave saliva samples voluntarily without training.
► We validated cortisol and salivary alpha amylase (SAA) assays analytically.
► Samples did not respond to a human secretory immunoglobulin A (SIgA) assay.
► Values correlated with animal age (SAA) and dominance rank (cortisol).