Abstract # 3017 Poster # 99:

Scheduled for Saturday, September 17, 2011 07:00 PM-09:00 PM: Session 14 (Salon G (Sixth Floor)) Poster Presentation


M. M. Petracca1 and N. G. Caine2
1University of Nebraska-Lincoln, Department of Psychology University of Nebraska, Lincoln, NE 68588-0308, USA, 2California State University San Marcos
     Vocal signatures may allow individuals to distinguish among conspecifics. At least three New World primate species have individually distinct long calls, but individual differences in alarm calls have not received much attention. We recorded the alarm calls of six captive adult marmosets (Callithrix geoffroyi) in response to a snake model. We then used acoustic software to measure four parameters: duration, and starting, maximum, and ending peak frequency. A discriminant function analysis confirmed that all four parameters differed among the individual marmosets [X2(20)=740.58, p<.001]. A jackknifed classification process was run using sample proportions as prior probabilities. For the 219 alarm calls used, 195 (89.0%) were correctly classified; only 42 (19.25%) would be correctly classified by chance alone [X2(25)=842.74, p<.05]. The stability of the classification process was checked by cross-validation, in which 87.7% (192) of the alarm calls were correctly classified. Our data suggest that an individual’s alarm calls elicited by a snake model are acoustically distinct from those of other individuals. Additional research is needed to determine whether these acoustic differences are actually perceived by the marmosets. If individual recognition of alarm calls is possible, individuals may adjust their response to calls given by certain individuals, especially if the caller tends to alarm call inappropriately. This may have implications for future playback studies using alarm calls as stimuli.