The DFA correctly classified 62% of the contact calls (10 subjects) and 80.9% of the ecstatic display songs (seven subjects) according to the correct emitter, showing that acoustic cues to individuality encoded in both vocal types remained unchanged over four consecutive breeding seasons. Two separate leave-one-out cross-validated Discriminant Function Analyses (DFA) were then performed, generating the discriminant functions from the vocalisations collected in 2017 to classify those recorded in 2020. For each vocalisation, we measured 14 spectral and temporal acoustic parameters related to both source and filter components. Contact calls and ecstatic display songs were recorded from an ex situ colony in 2017 and in 2020. In this study, the stability over time of two discrete vocal types was investigated in the African penguin (Spheniscus demersus), a monogamous and territorial seabird. The stability of individual acoustic features is fundamental in social species, and more importantly in monogamous and territorial species, showing long-term fidelity both to the partner and the breeding site. We propose a potential divergence of alarm calls in Asian pikas to high-frequency whistles (> 20 kHz in Daurian pikas) and in American pikas to low-frequency emissions (0.4–1.3 kHz in Ochotona princeps) during the evolutionary radiation of pikas at the center of the origin of lagomorphs in East Asia and their subsequent geographic dispersal. We discuss the relationship between vocal traits, individuality, vocal production mechanisms, and functions, of pika alarm calls. Nonlinear vocal phenomena (biphonations) only were detected in one individual. The accuracy of classifying alarm calls to correct callers with discriminant function analysis (DFA) was 93.71% for the manually measured set of 12 acoustic variables and 95.43% for the semiautomatically measured set of 12 acoustic variables in both cases exceeding the level of chance (17.28% or 17.33%, respectively). Call duration was very short (0.057 s on average).
The alarm calls of most (32 of the 35) callers started in the ultrasonic range at 22.41 kHz on average and rapidly decreased to 3.88 kHz on average at call end. We recorded the alarm calls produced toward a surrogate predator (researcher), slowly moving (0.5–1 km/h) between densely distributed colonies. In this study, we show that alarm calls of Daurian pikas, Ochotona dauurica (Pallas, 1776), encode information about caller identity. However, we believe that after overcoming the issues outlined above, AIID can quickly become a widespread and valuable tool in field research and conservation of mammals and other animals.Ĭolonial lagomorphs warn conspecifics of potential danger with alarm calls encoding information about attributes of presumptive predators as well as the caller.
Unfortunately, further progress in this area is currently hindered by the lack of appropriate publicly available datasets. Automation of AIID could be achieved with the use of advanced machine learning techniques inspired by those used in human speaker recognition or tailored to specific challenges of animal AIID. A major obstacle for widespread utilization of AIID is the absence of tools integrating all AIID subtasks within a single package. In species with weaker acoustic signatures, AIID could still be a valuable tool once its limitations are fully acknowledged. We found the greatest potential for AIID (characterized by strong and stable acoustic signatures) was in Cetacea and Primates (including humans). We reviewed studies of individual variation in mammalian vocalizations as well as pilot studies using acoustic identification to census mammals and birds. Strong and stable acoustic signatures are necessary for successful AIID. By conducting such work, we will also improve our understanding of identity signals in general. We present a pipeline of steps for successful AIID in a given species. Therefore, acoustic identification of individuals (AIID) has been repeatedly suggested as a non-invasive and labor efficient alternative to mark-recapture identification methods. Many studies have revealed that animal vocalizations, including those from mammals, are individually distinctive.
0 kommentar(er)
