The acorn woodpecker (Melanerpes formicivorus) is a highly social species with the cooperative breeding system. In California, acorn woodpeckers live in stable family groups of up to fifteen individuals, including breeders and helpers with both sexes and all ages. Due to the complexity of their social system, there might be strong selection pressure for an equally complex communication system. In particular, the ability of discrimination between conspecific individuals or groups of individuals may be favored by selection as the coordination of many different social interactions requires individuals to be capable of recognizing each other. In this dissertation I focused on the vocal communication system in acorn woodpeckers and aimed to investigate their abilities of discriminating vocalizations from different individuals or different groups.

First, I examined the presence of vocal individuality in acorn woodpeckers as it was the prerequisite of individual recognition. I applied Hidden Markov Models (HMMs) for the identification of vocal individuality in acorn woodpecker. Through the waka call type, the HMMs could discriminate 11 individual acorn woodpeckers with 94.1% accuracy, indicating that individual acorn woodpeckers have unique, recognizable vocal signatures. Further, playback experiments were conducted to test if acorn woodpeckers used waka calls as recognition cues during their social activities. I found that acorn woodpeckers could discriminate waka calls from its group members and non-members. The mechanism of membership recognition may be based on their individual recognition abilities through waka calls.

Another direction of this research was to provide an efficient and convenient method of vocal individual recognition that could be used in further studies of bioacoustics and animal behaviors. I used HMMs, which is a popular tool for human speech recognition, as the basic method for recognizing isolated waka calls and continuous recordings collected in the field. The intrinsic characteristics of HMMs and their high recognition accuracy suggested that HMMs could be a promising tool for automatically monitoring vocal activities in many animal species.

Topics in this dissertation are arranged as follows: Chapter 1 introduces the background and the objectives of the dissertation. Chapter 2 investigates the presence of vocal individuality in acorn woodpeckers using HMMs. Chapter 3 extends the use of HMMs from isolated calls to continuous recordings collected in the field. Chapter 4 is devoted to explore the function of vocal signals in social interactions among free-ranging acorn woodpeckers. Chapter 5 summarizes the study presented and suggests the possible directions for future work.