Detection of individual differences encoded in sequential variations of elements in zebra finch songs

Zebra finches sing individually unique songs and recognize conspecific songs and individual identities in songs. Their songs comprise several syllables/elements that share acoustic features within the species, with unique sequential arrangements. However, the neuronal mechanisms underlying the detection of individual differences and species specificity have yet to be elucidated. Herein, we examined the neuronal auditory responsiveness of neurons in higher auditory area, the caudal nidopallium (NCM), to songs and their elements in male zebra finches to understand the mechanism for detecting species and individual identities in zebra finch songs. We found that various adult male zebra finch songs share acoustically similar song elements but differ in their sequential arrangement between individuals. The broader spiking (BS) neurons in the NCM detected only a small subset of zebra finch songs, whereas NCM BS neurons, as a neuronal ensemble, responded to all zebra finch songs. Notably, distinct combinations of BS neurons responded to each of the 18 presented songs in one bird. Subsets of NCM BS neurons were sensitive to sequential arrangements of species-specific elements, which dramatically increasing the capacity for song variation with a limited number of species-specific elements. The naïve Bayes decoder analysis further showed that the response of sequence-sensitive BS neurons increased the accuracy of song stimulus predictions based on the response strength of neuronal ensembles. Our results suggest the neuronal mechanisms that NCM neurons as an ensemble decode the individual identities of songs, while each neuron detects a small subset of song elements and their sequential arrangement.Significance statement Zebra finches develop unique songs by learning from tutors. Various zebra finch songs consist of repeats of species-specific syllable elements that differ in their sequential arrangements. In vivo, single-unit electrophysiological recordings from neurons in the zebra finch's higher auditory area (caudal nidopallium [NCM]) revealed that each broad-spiking (BS) NCM neuron responded to a small subset of the zebra finch songs. However, a NCM neuronal ensemble detected all the songs. Some NCM BS neurons responded sensitively to sequential song element arrangement, which increased the prediction accuracy in the naïve Bayes decoder analysis. These findings suggest a neuronal mechanism for discriminating individual song variations in NCM neuronal ensembles, in which each neuron detects small subsets of song elements and their sequential arrangements.