Abstract # 944 Event # 4:

Scheduled for Thursday, August 18, 2005 10:15 AM-10:30 AM: Session 1 (Parliament Room) Oral Presentation

Quantifying the Vocal Repertoire of Wild Adult Diademed Sifakas (Propithecus diadema diadema) in Madagascar

E. R. Patel1, J. D. Anderson1, M. T. Irwin2 and M. J. Owren1
1Cornell University, Department of Psychology, Psychology of Voice and Sound Research Lab, 211 Uris Hall, Ithaca, NY 14853, USA, 2Stony Brook University, Interdepartmental Doctoral Program in Anthropological Sciences
     There has never been a quantitative vocal repertoire study of any wild diurnal lemur. The vocal repertoire of wild diademed sifakas was documented through all-occurrence audio-recording and behavioral sampling from February through December 2003. Data were collected from sifakas in two habitats in the Tsinjoarivo region of eastern Madagascar, including primary forest (Vatateza) and fragmented forest (Mahatsinjo). Both usage frequency per call type over 4208 focal-animal sampling hours and ad libitum data on associated socio-ecological context were recorded. Temporal, amplitude, and frequency measurements were extracted from 72 calls comprised of 342 total segments using spectrographic, spectral, and waveform analyses. Seven call-types were distinguished, with Kruskal-Wallis (P < 0.001) and post-hoc Mann-Whitney U (P < 0.001) tests confirming significant differences among them on 8 of 10 acoustic variables. Cross-validated discriminant-function analysis successfully classified 84% (P < 0.001) of the vocalizations by call-type. While discriminant-function results thus indicate large differences among call-types, intra-call analyses suggested that all the sounds were composed of some combination of the same two basic components, namely tonal versus atonal segments. The vocal repertoire of these diademed sifakas appears significantly smaller than in ring-tailed (Lemur catta) and ruffed lemurs (Varecia variegata), which may reflect variation in species-typical behavioral ecology or result from differences in the analysis methodologies used in the respective studies. Supported by Margot Marsh Biodiversity Foundation, Primate Conservation, Inc, IPPL
and Stony Brook University.