An implicit measure of olfactory performance for non-human primates reveals aversive and pleasant odor conditioning

We have little understanding of how odorants are processed in neural networks of the primate brain. Because chemo-stimuli are harder to control than physical stimuli (e.g. vision, audition), such research was limited by the temporal resolution, accuracy, and reliability of olfactometers (odor producing machines). Recent advances were able to create olfactometers that overcome these limitations, allowing their use together with neuroimaging techniques in humans. From the behavioral point of view, olfaction research requires a behavioral measure that can be used to quantify olfactory performance. This becomes a real problem when working with animals, where, unlike humans, explicit measures are harder to obtain. Furthermore, because odorants are powerful primitive reinforcers, such implicit measures can be beneficial to use in learning paradigms. Here we describe an olfactometer suitable for use in non-human primates, and an end-port design that allows the accurate measure of real-time respiratory modulations that are elicited in response to odor presentation. We demonstrate that this implicit measure is differentially modulated when experiencing pleasant or aversive odors. We then present an experimental paradigm in which monkeys learn to associate tones with odors, and show that the time delay from the conditioned stimuli to the next breath can be used to measure learning and memory expression in this paradigm. Using this construct, we reveal olfactory performance during acquisition and extinction of odor conditioning. These techniques can be used in electrophysiological recordings from relevant brain areas to shed light on neural networks involved in odor processing and reinforcement-learning.