The use of fecal cortisol for enrolment of free-roaming dogs in Singapore to a national rehabilitation-rehoming program: A possible indicator of allostasis

Abstract

Singapore's trap-neuter-release-management program strives to rehome as many free-roaming dogs as possible. However, before rehoming, most free-roaming dogs require rehabilitation and this requires sheltering them for some time. Therefore, an individual's suitability for rehoming hinges on his or her adaptability to shelter conditions. Currently, this is being assessed qualitatively across behavioral contexts. Augmenting the current assessment with a known bioindicator of stress, such as cortisol, would increase its robustness. However, the serum measurement of cortisol can be invasive. Canine fecal samples have presented an attractive alternative for measuring cortisol indirectly in various contexts. Nevertheless, it is unclear whether the resolution of fecal cortisol metabolites (FCM) is sufficient for evaluating Singapore free-roaming dogs’ adaptability to shelter conditions. To elucidate this, we performed linear mixed-effects model on 27 candidate models to compare FCM between pet and recently sheltered dogs, with sex, state of the feces, and time as covariates. Subsequently, we ran generalized mixed-effected model on 12 candidate models to examine the influence of between- and within-subject variations in FCM on the prognosis of sheltered dogs—based on our current assessment. Our analyses showed that FCM levels of recently sheltered dogs were higher than pet individuals. However, unlike several studies, FCM concentrations of recently sheltered dogs continued to rise, indicating that individuals have yet to adapt. Furthermore, the difference in FCM levels between newly sheltered individuals and pet dogs was moderated by sex. Unexpectedly, we found the average FCM levels of dogs with a good prognosis to be higher than those that received a poor prognosis, with sex and state of the fecal samples as important covariates. This may be due to the difference in allostasis between the 2 groups of dogs. Thus, our results showed FCM not only has sufficient resolution but is useful in evaluating dogs’ adaptability to shelter conditions.