VOCAL COMMUNICATION AND ACOUSTIC MONITORING OF THE KING RAIL (Rallus elegans)

Rare and elusive species present a challenge for researchers , and are consequently often overlooked in evolutionary and behavioral studies. The King Rail , Rallus elegans , is a secretive marsh bird considered globally 'Near Threatened' by Birdlife International. Their simple , unlearned calls provide opportunity to decipher how information is aurally transferred during social interactions while excluding variation from learning and structural complexity. Furthermore , the dense , visually concealing habitat in which the species lives poses pressures on these calls for efficient sound transmission and reveals this system as a potential model of evolution. This thesis combines evolutionary ecology , behavioral observation , field experiments , and applied management approaches to examine behaviors associated with vocal communication in the King Rail and how these can be utilized to monitor populations more efficiently. In order to use calls to research and monitor the King Rail , it is necessary to characterize their full vocal repertoire. Previous accounts for this species are incomplete and inconsistent. In Chapter 1 , I review published findings and integrate these with my own field observations , recordings , and analyses to provide a comprehensive account of structure and function of the vocalizations produced by King Rails. Most calls are variations of a series of repetitive , pulsed notes created by altering peak frequency , bandwidth , amplitude , pulse rate , and note length. In Chapter 2 , I focus on the two most commonly used calls , the grunt (contact and disturbance context) and kek (mating and territorial context) , to determine what information , such as individuality , might be encoded in these signals. While mean peak frequency of grunts increased in stressful contexts and mean peak frequency of kek notes increased over the course of the breeding season , I did not find sufficient evidence to suggest that these calls are individually variable. However , a playback experiment revealed that King Rails were more likely to vocalize and did so sooner in response to a neighbor's grunt rather than a stranger's grunt. This suggests that classifying information is perceived from this call , although it is unclear if King Rails can only distinguish between familiar and unfamiliar individuals or if they can tell caller identity. In Chapter 3 , I explored how vocal behaviors , such as temporal call rate patterns and likelihood of response to callback under certain conditions , could be used for auditory detection of breeding King Rails. I found that weather , number of human observers , and time of day and season did not affect the number of callback survey detections. Using computer-learning software , I created a signal recognizer to find King Rail calls in large recording files and automate the analysis of data from autonomous recording units (ARUs). These analyses revealed a significant seasonal decline and a clear crepuscular diel pattern in King Rail call rate. I also found that grunt call rate , but not kek call rate , was significantly positively correlated to King Rail density. Overall , this thesis provides insight into the behavioral and evolutionary processes shaping the vocal communication system in a species with simple , unlearned calls. It has also generated tools for surveyors and managers to utilize while monitoring and implementing conservation plans for the King Rail