Rendall D, Owren M J, Rodman P S
Department of Anthropology, University of California at Davis 95616, USA.
J Acoust Soc Am. 1998 Jan;103(1):602-14. doi: 10.1121/1.421104.
The importance of individual identity and kinship has been demonstrated in the social behavior of many nonhuman primates, with some evidence suggesting that individually distinctive acoustic features are present in their vocalizations as well. In order to systematically test whether acoustic cues to identity are reliably present across the vocal repertoire of rhesus monkeys (Macaca mulatta), we examined coos, grunts, and noisy screams produced by adult females of two free-ranging groups. First, acoustic analyses were used to characterize spectral patterning, the fundamental frequency, and temporal characteristics of these three distinct call types. Vocalizations were then classified by caller identity, based on discriminant function analyses. Results showed that coos (rich, harmonically structured sounds) were markedly more distinctive by caller than were either grunts or noisy screams, and that spectral-patterning measures related to vocal tract filtering effects were the most reliable markers of individual identity. Grunts (pulsed, noisy calls) were classified at lower, but above-chance rates and spectral patterning cues were again critical in this sorting. Noisy screams (continuous, broadband noise bursts that could include a high-frequency, periodic component) could not be reliably sorted by caller. Playback experiments conducted with the screams showed no response differences when listening animals heard vocalizations produced by kin or nonkin individuals. This result was strikingly different from the corresponding outcome of a previous test with coo calls, but consistent with the acoustic analysis. Implications of these findings for vocal production mechanisms in nonhuman primates and previous studies of rhesus monkey vocalizations are discussed.
个体身份和亲属关系的重要性已在许多非人类灵长类动物的社会行为中得到证明，一些证据表明，它们的叫声中也存在个体独特的声学特征。为了系统地测试恒河猴（猕猴）整个发声库中是否可靠地存在身份识别的声学线索，我们研究了两个自由放养群体中成年雌性发出的咕咕声、呼噜声和嘈杂叫声。首先，通过声学分析来描述这三种不同叫声类型的频谱模式、基频和时间特征。然后，基于判别函数分析，根据叫声者的身份对叫声进行分类。结果表明，咕咕声（丰富、谐波结构的声音）在叫声者之间的差异明显大于呼噜声或嘈杂叫声，并且与声道滤波效应相关的频谱模式测量是个体身份最可靠的标志。呼噜声（脉冲式、嘈杂叫声）的分类准确率较低，但高于随机水平，频谱模式线索在这种分类中再次至关重要。嘈杂叫声（连续的宽带噪声爆发，可能包括高频周期性成分）无法可靠地按叫声者进行分类。对这些叫声进行的回放实验表明，当聆听的动物听到亲属或非亲属个体发出的叫声时，没有反应差异。这一结果与之前对咕咕声测试的相应结果截然不同，但与声学分析一致。讨论了这些发现对非人类灵长类动物发声机制以及之前恒河猴发声研究的意义。
