With the advent of new technologies to probe brain-behavior relationships deeper than ever before, the field of behavioral neuroscience finds itself at a crossroads for translation. The rich theoretical foundations upon which many behavioral assays are based have largely been lost; and factors known to impact motivated behavior, such as social and environmental challenges, tend to be ignored. Here, the rich neurobiological understanding that has been obtained from captive animals will be leveraged to assess individual differences in motivated behavior that derive from social and/or environmental adversity (or advantage) in a natural environment. Laboratory-based paradigms will be brought to a wild population of ~350 white-faced capuchin monkeys (Cebus imitator) living in a small, tractable forest in Costa Rica (Taboga Forest Reserve). Approximately 70 of these capuchins have been tracked and studied near-daily for the past 7 years, generating a rich life-history database for these individuals. These capuchins are habituated to humans and readily interact with stimuli on experimental platforms. But importantly, they are also exposed to the natural elements – predators, droughts, disease, and social adversaries – without human intervention. This project will lay the groundwork for an Artificial Intelligence (AI) Forest that will transform our ability to study these animals as they develop, live, and die in their natural environment. The AI Forest will be the first of its kind; an experimental forest allowing us to probe laboratory-based behaviors in a wild primate population. Smart testing stations will be placed throughout the forest to assess reward valuation, inhibitory control, and behavioral flexibility. A visual deep-learning algorithm will be harnessed to recognize individuals as they voluntarily approach and perform at testing stations. The AI system will automatically guide capuchins through experiments, allowing them to “self-pace” as they complete each assessment. Data collection at the smart testing stations will be automated, potentially yielding the largest dataset of its kind from a wild primate population. In addition, a vocal deep-learning algorithm will be used to track movements of individuals and groups and to detect their social and environmental challenges (intergroup and predator encounters). The multidimensional behavioral and environmental data will be integrated to examine the impact of environmental adversity (or advantage) on behavioral traits across the lifespan. This AI Forest represents an exciting possibility of what the future of behavioral neuroscience could be – bridging studies of complex behaviors with function (adaptive evolution) – to inform translational science and ultimately enhance our understanding of human health and disease.
0 commit comments