Body plan of consumed organisms influences ecological range of consumers through neural processing bias

The number of different resources used by an organism to feed, reproduce, and survive (ecological range) is a key determinant of community structure and community response to environmental change. It is known empirically that animals can have strong sensory biases, and here we use artificial neural networks to demonstrate how such biases may influence ecological range. We show that networks have strong biases for resources of a certain shape, producing distinct profiles of "cognitive accuracy" with ecological range. Generally, ecological specialization is tolerated for some resource body plans more than others; however, certain combinations of resources actually promote ecological specialization by producing local maxima in cognitive accuracy. Some of the variation in ecological range across consumers may be explained by these interactions between evolving nervous system and resource shape. Ecological patterns at even the largest of scales may be influenced by the miniscule processes of information exchange within networks of neurons.