The potential for acoustic individual identification in mammals

Linhart, Pavel; Mahamoud-Issa, Mathieu; Stowell, Dan; Blumstein, Daniel T.

doi:10.1007/s42991-021-00222-2

Cited by 16 publications

(15 citation statements)

References 131 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Information content of Sarasota dolphin signature whistles, as calculated using Beecher's Information Statistic (Hs), is higher than for any other animal signal studied so far (Linhart et al, 2022). This is not surprising when considering that most nonhuman animals use voice cues, called "byproduct…”

Section: Discussionmentioning

confidence: 91%

“…A variety of metrics have been proposed over the years to quantify individual identity information (individuality) in animal signals. Recently, Linhart and colleagues (Linhart et al, 2019;Linhart et al, 2022) proposed that Beecher's Information Statistic (Beecher, 1989) can serve as a robust and flexible metric for comparing individuality across animal communication systems.…”

Section: Quantification Of Individualitymentioning

confidence: 99%

“…This approach likely does not reflect all information available in the signature whistle (i.e. it does not capture non-contour features, features that are modulated more rapidly than can be sampled using 20 values per whistle, subtle or sequential features), so it is at best a conservative estimate of information content; however, it provides a standardized metric for comparisons to other species (Linhart et al, 2022). After calculating these 21 features (20 contour points and 1 duration) for each loop, measurements were averaged across all loops in a whistle.…”

Section: Quantification Of Individualitymentioning

confidence: 99%

“…The Sarasota Dolphin Whistle Database has enabled a large variety of research projects. One that we present here is a comparison to other animal signals of the potential of signature whistle contours for acoustic individual identification (Linhart et al, 2019;Linhart et al, 2022). To examine this, we use a sample of signature whistles to calculate Beecher's Information Statistic (Hs) as a measure of the amount of information that is theoretically available for differentiating individuals (Beecher, 1989).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

The Sarasota Dolphin Whistle Database: A unique long-term resource for understanding dolphin communication

et al. 2022

View full text Add to dashboard Cite

Common bottlenose dolphins (Tursiops truncatus) produce individually distinctive signature whistles that are learned early in life and that help animals recognize and maintain contact with conspecifics. Signature whistles are the predominant whistle type produced when animals are isolated from conspecifics. Health assessments of dolphins in Sarasota, Florida (USA) provide a unique opportunity to record signature whistles, as dolphins are briefly separated from conspecifics. Recordings were first made in the mid 1970’s, and then nearly annually since 1984. The Sarasota Dolphin Whistle Database (SDWD) now contains 926 recording sessions of 293 individual dolphins, most of known age, sex, and matrilineal relatedness. The longest time span over which an individual has been recorded is 43 years, and 85 individuals have been recorded over a decade or more. Here we describe insights about signature whistle structure revealed by this unique and expansive dataset. Signature whistles of different dolphins show great variety in their fundamental frequency contours. Signature whistle types (with ‘whistle type’ defined as all whistles visually categorized as sharing a particular frequency modulation pattern) can consist of a single stereotyped element, or loop (single-loop whistles), or of multiple stereotyped loops with or without gaps (multi-loop whistles). Multi-loop signature whistle types can also show extensive variation in both number and contour of loops. In addition, fundamental frequency contours of all signature whistle types can be truncated (deletions) or embellished (additions), and other features are also occasionally incorporated. However, even with these variable features, signature whistle types tend to be highly stereotyped and easily distinguishable due to the extensive variability in contours among individuals. In an effort to quantify this individual distinctiveness, and to compare it to other species, we calculated Beecher’s Information Statistic and found it to be higher than for any other animal signal studied so far. Thus, signature whistles have an unusually high capacity to convey information on individual identity. We briefly review the large range of research projects that the SDWD has enabled thus far, and look ahead to its potential to answer a broad suite of questions about dolphin communication.

show abstract

Section: Discussionmentioning

confidence: 91%

Section: Quantification Of Individualitymentioning

confidence: 99%

Section: Quantification Of Individualitymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

The Sarasota Dolphin Whistle Database: A unique long-term resource for understanding dolphin communication

et al. 2022

View full text Add to dashboard Cite

show abstract

“…These parameters are probably sufficiently idiosyncratic to allow an ape to correctly identify the emitter in a real life situation. An algorithm-based automatic classification is of course only partially related to the cognitive task an animal performs when it has to decode the information carried by a conspecific’s vocalization (Fedurek, Zuberbühler & Dahl, 2016; Linhart, et al, 2022), but it can be assumed that individual conspecific individual vocal identification is quite accurate in bonobos. This has been shown in an experimental study using playbacks with a single call type (Keenan, et al, 2016).…”

Section: Resultsmentioning

confidence: 99%

Improving the workflow to crack Small, Unbalanced, Noisy, but Genuine (SUNG) datasets in bioacoustics: the case of bonobo calls

Arnaud

Pellegrino

Keenan

et al. 2022

Preprint

View full text Add to dashboard Cite

Despite the accumulation of data and studies, deciphering animal vocal communication remains highly challenging. While progress has been made with some species for which we now understand the information exchanged through vocal signals, researchers are still left struggling with sparse recordings composing Small, Unbalanced, Noisy, but Genuine (SUNG) datasets. SUNG datasets offer a valuable but distorted vision of communication systems. Adopting the best practices in their analysis is therefore essential to effectively extract the available information and draw reliable conclusions. Here we show that the most recent advances in machine learning applied to a SUNG dataset succeed in unraveling the complex vocal repertoire of the bonobo, and we propose a workflow that can be effective with other animal species. We implement acoustic parameterization in three feature spaces along with three classification algorithms (Support Vector Machine, xgboost, neural networks) and their combination to explore the structure and variability of bonobo calls, as well as the robustness of the individual signature they encode. We underscore how classification performance is affected by the feature set and identify the most informative features. We highlight the need to address data leakage in the evaluation of classification performance to avoid misleading interpretations. Finally, using a Uniform Manifold Approximation and Projection (UMAP), we show that classifiers generate parsimonious data descriptions which help to understand the clustering of the bonobo acoustic space. Our results lead to identifying several practical approaches that are generalizable to any other animal communication system. To improve the reliability and replicability of vocal communication studies with SUNG datasets, we thus recommend: i) comparing several acoustic parameterizations; ii) adopting Support Vector Machines as the baseline classification approach; iii) explicitly evaluating data leakage and possibly implementing a mitigation strategy; iv) visualizing the dataset with UMAPs applied to classifier predictions rather than to raw acoustic features.

show abstract

The value of individual identification in studies of free-living hyenas and aardwolves

Spagnuolo

Lemerle²,

Holekamp

et al. 2022

Mamm Biol

View full text Add to dashboard Cite

From population estimates to social evolution, much of our understanding of the family Hyaenidae is drawn from studies of known individuals. The extant species in this family (spotted hyenas, Crocuta crocuta , brown hyenas, Parahyaena brunnea , striped hyenas, Hyaena hyaena , and aardwolves, Proteles cristata ) are behaviorally diverse, presenting an equally diverse set of logistical constraints on capturing and marking individuals. All these species are individually identifiable by their coat patterns, providing a useful alternative to man-made markings. Many studies have demonstrated the utility of this method in answering a wide range of research questions across all four species, with some employing a creative fusion of techniques. Despite its pervasiveness in basic research on hyenas and aardwolves, individual identification has rarely been applied to the conservation and management of these species. We argue that individual identification using naturally occurring markings in applied research could prove immensely helpful, as this could further improve accuracy of density estimates, reveal characteristics of suitable habitat, identify threats to population persistence, and help to identify individual problem animals. Supplementary Information The online version contains supplementary material available at 10.1007/s42991-022-00309-4.

show abstract

The potential for acoustic individual identification in mammals

Cited by 16 publications

References 131 publications

The Sarasota Dolphin Whistle Database: A unique long-term resource for understanding dolphin communication

The Sarasota Dolphin Whistle Database: A unique long-term resource for understanding dolphin communication

Improving the workflow to crack Small, Unbalanced, Noisy, but Genuine (SUNG) datasets in bioacoustics: the case of bonobo calls

The value of individual identification in studies of free-living hyenas and aardwolves

Contact Info

Product

Resources

About