Fronto-striatal oscillations predict vocal output in bats

Abstract

Abbreviations AUC: area under the curve; BF: best frequency; CN: caudate nucleus; d: Cliff's Delta; dVS: difference in vector strength, FAF: frontal auditory field; IQR: inter-quartile range; LFP: local field potential; Med: median; PFC: prefrontal cortex; PSTH: peri-stimulus time histogram; SVM: support vector machine; VS: vector strength SummaryThe ability to vocalize is ubiquitous in vertebrates, but neural networks leading to vocalization production remain poorly understood. Here we performed simultaneous, large scale, neuronal recordings in the frontal cortex and dorsal striatum (caudate nucleus) during the production of echolocation and non-echolocation calls in bats. This approach allows to assess the general aspects underlying vocalization production in mammals and the unique evolutionary adaptations of bat echolocation. Our findings show that distinct intra-areal brain rhythms in the beta (12-30 Hz) and gamma (30-80 Hz) bands of the local field potential can be used to predict the bats' vocal output and that phase locking between spikes and field potentials occurs prior vocalization production. Moreover, the fronto-striatal network is differentially coupled in the theta-band during the production of echolocation and nonecholocation calls. Overall, our results present evidence for fronto-striatal network oscillations in motor action prediction in mammals.