Data from: Optimal multisensory integration
Munoz, Nicole; Blumstein, Daniel (2019), Data from: Optimal multisensory integration, Dryad, Dataset, https://doi.org/10.5061/dryad.747n4dv
Animals are often confronted with potentially informative stimuli from a variety of sensory modalities. While there is a large proximate literature demonstrating multisensory integration, no general framework explains why animals integrate. We developed and tested a quantitative model that explains why multisensory integration is not always adaptive and explain why unimodal decision-making might be favored over multisensory integration. We present our model in terms of a prey that must determine the presence or absence of a predator. A greater chance of encountering a predator, a greater benefit of correctly responding to a predator, a lower benefit of correctly foraging, or a greater uncertainty of the second stimulus favors integration. Uncertainty of the first stimulus may either increase or decreases the favorability of integration. In three field studies, we demonstrate how our model can be empirically tested. We evaluated the model with field studies of yellow-bellied marmots (Marmota flaviventer) by presenting marmots with an olfactory-acoustic predator stimulus at a feed station. We found some support for the model’s prediction that integration is favored when the second stimulus is less noisy. We hope additional predictions of the model will guide future empirical work that seeks to understand the extent to which multimodal integration might be situation dependent. We suggest that the model is generalizable beyond antipredator contexts and can be applied within or between individuals, populations or species.
National Science Foundation, Award: DEB-1119660