Search

Nonhuman primate vocalizations have been traditionally described as stereotyped and most likely genetically determined. However, there is increasing evidence of flexibility with a wide variety of species demonstrating the ability to change aspects of their calls, such as note order and phrase duration. We assess patterns of variation in note order and temporal features in wild siamang (Symphalangus syndactylus) phrases. We used three types of analyses on 1,015 ululating scream phrases (US-II), containing 21,609 notes, produced by ten siamang groups at three field sites. First, to assess similarity or differences in the organization of notes in the US-II phrase, we calculated the Levenshtein distance (LD), which quantifies the similarity of sequences of strings through the number of insertions, deletions, and substitutions required to transform one into the other. Second, we used a supervised classification approach to see how well we could assign US-II phrases to their groups and sites. Third, we compared five unsupervised clustering algorithms to investigate the tendency to cluster in US-II phrase types. The note order of the US-II phrase was variable shown by relatively large LD values; there also were significant mean differences between groups. Supervised classification using support vector machine and leave-one-out cross-validation returned 89.5% accuracy for site and 53.1% for group. For unsupervised clustering, the most stable solution based on the Silhouette coefficient returned four unique clusters or classes of the US-II phrase in our dataset. Our findings indicate high levels of intragroup variation in the US-II phrase and are consistent with previous reports on siamang vocalizations being flexible. Quantifying variation in the siamang US-II phrase provides a crucial first step in understanding the evolutionary forces that shaped these signals and the potential information they convey. [ABSTRACT FROM AUTHOR]

Many animals rely on acoustic signals to mediate social interactions with conspecifics. Duets––the alternating vocal exchange between two animals––are of particular interest given the presumed intra- and intergroup communicative functions. Importantly, when there are sex-specific differences in duet contributions, the contribution of each sex may serve different function(s). We investigated variation in male Northern gray gibbon codas from seven sites on Malaysian Borneo using three complementary approaches. First, we used supervised classification to see how well we could classify male gibbon codas to the respective male. Second, we investigated the relative contribution of intramale, intermale, and intersite variance to total variance using a Bayesian multivariate, variance components model. Lastly, we investigated small-scale patterns of variation (<10 km) in male codas from a single site to test two mutually exclusive hypotheses related to small-scale patterns of variation. First, if call features are transmitted from father to offspring, we predicted neighboring males would have codas that were more similar to each other than males at further distances. Alternatively, if males actively differentiate from their neighbors, we predicted to see the opposite pattern. We did not find high levels of vocal individuality in male codas, as individual classification accuracy was relatively low (<63%) and there were no site-level differences in codas. We did not find support for either of our hypotheses regarding small-scale patterns of variation. Taken together, our findings indicate high levels of intraindividual variation in male codas. Future work that explores both the function(s) of the male and female contribution to the duets, along with investigations of heritability of duet features will be informative. [ABSTRACT FROM AUTHOR]

Most pair-living primate species engage in duets, wherein males and females produce coordinated vocalizations. Previous analyses of male gibbon contributions to the duet have shown that calls are individually distinct. Here we investigate variation in the temporal and spectral parameters in the male contribution to the duet, also known as the coda, of wild, nonhabituated male Müller’s Bornean gibbons (Hylobates muelleri), recorded both opportunistically and as a response to playbacks at the Stability of Altered Forest Ecosystems site in Sabah, Malaysia. We used linear discriminant function analysis to estimate the intra- and interindividual variation in 13 spectral and temporal parameters of the vocalizations (N = 337) of 31 male gibbons. To further understand how call features vary within and between individuals we used a multivariate, variance components model to investigate how variance in features was partitioned at these two levels. We could identify males with a 66% accuracy using leave-one-out cross-validation, a relatively low score compared to female Müller’s Bornean gibbons and males of other species. We found that for some features (such as maximum frequency of the notes) most of the variance occurred between males, but for others (specifically total duration of the call and duration of rest in between notes) most of the variance occurred within a single male. Overall, male Müller’s Bornean gibbon coda vocalizations showed greater variability relative to their female counterparts, raising questions about potential differences in the function of the male and female contributions to the duet in Müller’s Bornean gibbons and the gibbon taxon as a whole. [ABSTRACT FROM AUTHOR]