Predicting species identity of bumblebees through analysis of flight buzzing sounds

We present a study of buzzing sounds of several common species of bumblebees, with the focus on automatic classification of bumblebee species and types. Such classification is useful for bumblebee monitoring, which is important in view of evaluating the quality of their living environment and protecting the biodiversity of these important pollinators. We analysed natural buzzing frequencies for queens and workers of 12 species. In addition, we analysed changes in buzzing of Bombus hypnorum worker for different types of behaviour. We developed a bumblebee classification application using machine learning algorithms. We extracted audio features from sound recordings using a large feature library. We used the best features to train a classification model, with Random Forest proving to be the best training algorithm on the testing set of samples. The web and mobile application also allows expert users to upload new recordings that can be later used to improve the classification model and expand it to include more species.     

Detecting tropical nocturnal birds using automated audio recordings

ABSTRACT Nocturnal bird assemblages are poorly known in most tropical locations, and information about their presence and behavior is often limited to the results of dawn or dusk surveys. We investigated the use of manual‐ and automatic‐detection methods to identify nocturnal birds in acoustic recordings made at Soberania National Park, Republic of Panama. Five nocturnal species were detected in dusk recordings, and a sixth species (Great Potoo, Nyctibius grandis) was detected only after dark. Automatic data template detectors (DTD's) were developed and used to detect Crested Owls (Lophostrix cristata), Black‐and‐White Owls (Ciccaba nigrolineata), Vermiculated Screech‐Owls (Megasops guatemalae), and Great Potoos. Manual analysis of 300 h of overnight recordings allowed us to quantify DTD performance. Sensitivity, the proportion of known calls of target species identified by DTDs, ranged from 0.17 for Black‐and‐White Owls to 0.79 for Vermiculated Screech‐Owls. Positive predictive value, the proportion of detected sounds that corresponded to the target species, ranged from 0.39 for Black‐and‐White Owls to 0.60 for Crested Owls. Our results demonstrate that a combination of manual and automated analysis of audio recordings can provide a verifiable, systematic method to determine the presence of nocturnal birds in tropical forests, investigate temporal activity, and calculate detection probability.     

Adaptive frequency scaled wavelet packet decomposition for frog call classification

Environmental changes have put great pressure on biological systems leading to the rapid decline of biodiversity. To monitor this change and protect biodiversity, animal vocalizations have been widely explored by the aid of deploying acoustic sensors in the field. Consequently, large volumes of acoustic data are collected. However, traditional manual methods that require ecologists to physically visit sites to collect biodiversity data are both costly and time consuming. Therefore it is essential to develop new semi-automated and automated methods to identify species in automated audio recordings. In this study, a novel feature extraction method based on wavelet packet decomposition is proposed for frog call classification. After syllable segmentation, the advertisement call of each frog syllable is represented by a spectral peak track, from which track duration, dominant frequency and oscillation rate are calculated. Then, a k-means clustering algorithm is applied to the dominant frequency, and the centroids of clustering results are used to generate the frequency scale for wavelet packet decomposition (WPD). Next, a new feature set named adaptive frequency scaled wavelet packet decomposition sub-band cepstral coefficients is extracted by performing WPD on the windowed frog calls. Furthermore, the statistics of all feature vectors over each windowed signal are calculated for producing the final feature set. Finally, two well-known classifiers, a k-nearest neighbour classifier and a support vector machine classifier, are used for classification. In our experiments, we use two different datasets from Queensland, Australia (18 frog species from commercial recordings and field recordings of 8 frog species from James Cook University recordings). The weighted classification accuracy with our proposed method is 99.5% and 97.4% for 18 frog species and 8 frog species respectively, which outperforms all other comparable methods.     

Territorial dynamics of Mexican Ant‐thrushes Formicarius moniliger revealed by individual recognition of their songs

The ability to monitor interactions between individuals over time can provide us with information on life histories, mating systems, behavioural interactions between individuals and ecological interactions with the environment. Tracking individuals over time has traditionally been a time‐ and often a cost‐intensive exercise, and certain types of animals are particularly hard to monitor. Here we use canonical discriminant analysis (CDA) to identify individual Mexican Ant‐thrushes using data extracted with a semi‐automated procedure from song recordings. We test the ability of CDA to identify individuals over time, using recordings obtained over a 4‐year period. CDA correctly identified songs of 12 individual birds 93.3% of the time from recordings in one year (2009), while including songs of 18 individuals as training data. Predicting singers in one year using recordings from other years indicated some instances of variation, with correct classification in the range of 67–88%; one individual was responsible for the great majority (66%) of classification errors. We produce temporal maps of the study plot showing that considerably more information was provided by identifying individuals from their songs than by ringing and re‐sighting colour‐ringed individuals. The spatial data show site fidelity in males, but medium‐term pair bonds and an apparently large number of female floaters. Recordings can be used to monitor intra‐ and intersexual interactions of animals, their movements over time, their interactions with the environment and their population dynamics.     

Sex differences in the vocal repertoire of adult red‐capped mangabeys (Cercocebus torquatus): a multi‐level acoustic analysis

Sex differences in the vocal behavior of nonhuman primates can take various forms: sex‐specific call types, differential production of shared call types, or sex discrepancy in phonation. Also, a growing literature is evidencing that systematically analyzing the vocal repertoires of primates at the call level might lead to underestimating their communicative abilities. Here, we present an extensive multi‐level analysis of the still unknown vocal repertoire of adult red‐capped mangabeys (Cercocebus torquatus), with a special emphasis on sex differences. We collected recordings from seven adult males and seven adult females housed in captivity. We present a structurally‐based classification of mangabey calls that we cross‐validated by an analysis of the associated contexts of emission. We found 12 sound units (including six sex‐specific) that were concatenated to form eight call types (including four sex‐specific), which were produced either singularly or in sequences composed of one (“repetition”) or several (“combination”) call types. We extracted organizational principles that ruled call composition and calling patterns. This revealed a high degree of potentially meaningful variability in terms of semantics and syntax. Male–female discrepancy in terms of phonation could be related to morphological dimorphism and would enable listeners to behave appropriately according to the sex of the caller. Sex differences in repertoire size, structural gradation, and call usage could reflect specificities of male–female social roles. We discuss the pertinence of these sex differences according to social system and habitat quality. Am. J. Primatol. 72:360–375, 2010. © 2010 Wiley‐Liss, Inc.     

Species‐level image classification with convolutional neural network enables insect identification from habitus images

1. Changes in insect biomass, abundance, and diversity are challenging to track at sufficient spatial, temporal, and taxonomic resolution. Camera traps can capture habitus images of ground‐dwelling insects. However, currently sampling involves manually detecting and identifying specimens. Here, we test whether a convolutional neural network (CNN) can classify habitus images of ground beetles to species level, and estimate how correct classification relates to body size, number of species inside genera, and species identity.
2. We created an image database of 65,841 museum specimens comprising 361 carabid beetle species from the British Isles and fine‐tuned the parameters of a pretrained CNN from a training dataset. By summing up class confidence values within genus, tribe, and subfamily and setting a confidence threshold, we trade‐off between classification accuracy, precision, and recall and taxonomic resolution.
3. The CNN classified 51.9% of 19,164 test images correctly to species level and 74.9% to genus level. Average classification recall on species level was 50.7%. Applying a threshold of 0.5 increased the average classification recall to 74.6% at the expense of taxonomic resolution. Higher top value from the output layer and larger sized species were more often classified correctly, as were images of species in genera with few species.
4. Fine‐tuning enabled us to classify images with a high mean recall for the whole test dataset to species or higher taxonomic levels, however, with high variability. This indicates that some species are more difficult to identify because of properties such as their body size or the number of related species.
5. Together, species‐level image classification of arthropods from museum collections and ecological monitoring can substantially increase the amount of occurrence data that can feasibly be collected. These tools thus provide new opportunities in understanding and predicting ecological responses to environmental change.

     

RECOGNIZING DINOFLAGELLATE SPECIES USING ITS rDNA SEQUENCES1

Dinoflagellate taxonomy is based primarily on morphology and morphometric data that can be difficult to obtain. In contrast, molecular data can be rapidly and cost‐effectively acquired, which has led to a rapid accumulation of sequence data in GenBank. Currently there are no systematic criteria for utilizing taxonomically unassigned sequence data to identify putative species that could in turn serve as a basis for testable hypotheses concerning the taxonomy, diversity, distribution, and toxicity of these organisms. The goal of this research was to evaluate whether simple, uncorrected genetic distances (p) calculated using ITS1/5.8S/ITS2 (ITS region) rDNA sequences could be used to develop criteria for recognizing putative species before formal morphological evaluation and classification. The current analysis used sequences from 81 dinoflagellate species belonging to 14 genera. For this diverse assemblage of dinoflagellate species, the within‐species genetic distances between ITS region copies (p=0.000–0.021 substitutions per site) were consistently less than those observed between species (p=0.042–0.580). Our results indicate that a between‐species uncorrected genetic distance of p≥0.04 could be used to delineate most free‐living dinoflagellate species. Recently evolved species, however, may have ITS p values <0.04 and would require more extensive morphological and genetic analyses to resolve. For most species, the sequence of the dominant ITS region allele has the potential to serve as a unique species‐specific “DNA barcode” that could be used for the rapid identification of dinoflagellates in field and laboratory studies.     

Using soundscape recordings to estimate bird species abundance, richness, and composition

ABSTRACT Point counts are the most frequently used technique for sampling bird populations and communities, but have well‐known limitations such as inter‐ and intraobserver errors and limited availability of expert field observers. The use of acoustic recordings to survey birds offers solutions to these limitations. We designed a Soundscape Recording System (SRS) that combines a four‐channel, discrete microphone system with a quadraphonic playback system for surveying bird communities. We compared the effectiveness of SRS and point counts for estimating species abundance, richness, and composition of riparian breeding birds in California by comparing data collected simultaneously using both methods. We used the temporal‐removal method to estimate individual bird detection probabilities and species abundances using the program MARK. Akaike's Information Criterion provided strong evidence that detection probabilities differed between the two survey methods and among the 10 most common species. The probability of detecting birds was higher when listening to SRS recordings in the laboratory than during the field survey. Additionally, SRS data demonstrated a better fit to the temporal‐removal model assumptions and yielded more reliable estimates of detection probability and abundance than point‐count data. Our results demonstrate how the perceptual constraints of observers can affect temporal detection patterns during point counts and thus influence abundance estimates derived from time‐of‐detection approaches. We used a closed‐population capture–recapture approach to calculate jackknife estimates of species richness and average species detection probabilities for SRS and point counts using the program CAPTURE. SRS and point counts had similar species richness and detection probabilities. However, the methods differed in the composition of species detected based on Jaccard's similarity index. Most individuals (83%) detected during point counts vocalized at least once during the survey period and were available for detection using a purely acoustic technique, such as SRS. SRS provides an effective method for surveying bird communities, particularly when most species are detected by sound. SRS can eliminate or minimize observer biases, produce permanent records of surveys, and resolve problems associated with the limited availability of expert field observers.     

Rapid mitochondrial genome sequencing based on Oxford Nanopore Sequencing and a proxy for vertebrate species identification

Molecular information is crucial for species identification when facing challenging morphology‐based specimen identifications. The use of DNA barcodes partially solves this problem, but in some cases when PCR is not an option (i.e., primers are not available, problems in reaction standardization), amplification‐free approaches could be an optimal alternative. Recent advances in DNA sequencing, like the MinION device from Oxford Nanopore Technologies (ONT), allow to obtain genomic data with low laboratory and technical requirements, and at a relatively low cost. In this study, we explore ONT sequencing for molecular species identification from a total DNA sample obtained from a neotropical rodent and we also test the technology for complete mitochondrial genome reconstruction via genome skimming. We were able to obtain “de novo” the complete mitogenome of a specimen from the genus Melanomys (Cricetidae: Sigmodontinae) with average depth coverage of 78X using ONT‐only data and by combining multiple assembly routines. Our pipeline for an automated species identification was able to identify the sample using unassembled sequence data (raw) in a reasonable computing time, which was substantially reduced when a priori information related to the organism identity was known. Our findings suggest ONT sequencing as a suitable candidate to solve species identification problems in metazoan nonmodel organisms and generate complete mtDNA datasets.     

Designing better frog call recognition models

Advances in bioacoustic technology, such as the use of automatic recording devices, allow wildlife monitoring at large spatial scales. However, such technology can produce enormous amounts of audio data that must be processed and analyzed. One potential solution to this problem is the use of automated sound recognition tools, but we lack a general framework for developing and validating these tools. Recognizers are computer models of an animal sound assembled from “training data” (i.e., actual samples of vocalizations). The settings of variables used to create recognizers can impact performance, and the use of different settings can result in large differences in error rates that can be exploited for different monitoring objectives. We used Song Scope (Wildlife Acoustics Inc.) to build recognizers and vocalizations of the wood frog (Lithobates sylvaticus) to test how different settings and amounts of training data influence recognizer performance. Performance was evaluated using precision (the probability of a recognizer match being a true match) and sensitivity (the proportion of vocalizations detected) based on a receiver operating characteristic (ROC) curve‐determined score threshold. Evaluations were conducted using recordings not used to build the recognizer. Wood frog recognizer performance was sensitive to setting changes in four out of nine variables, and small improvements were achieved by using additional training data from different sites and from the same recording, but not from different recordings from the same site. Overall, the effect of changes to variable settings was much greater than the effect of increasing training data. Additionally, by testing the performance of the recognizer on vocalizations not used to build the recognizer, we discovered that Type I error rates appear idiosyncratic and do not recommend extrapolation from training to new data, whereas Type II errors showed more consistency and extrapolation can be justified. Optimizing variable settings on independent recordings led to a better match between recognizer performance and monitoring objectives. We provide general recommendations for application of this methodology with other species and make some suggestions for improvements.     

基于多源遥感数据的植物物种分类与识别: 研究进展与展望

物种分类与识别是生物多样性监测的基础, 明确物种的类别及其分布是解决几乎所有生态学问题的前提。为深入了解基于多源遥感数据的植物物种分类与识别相关研究的发展现状和存在的问题, 本文对2000年以来该领域的研究进行了总结分析, 发现: 当前大多数研究集中在欧洲和北美地区的温带或北方森林以及南非的热带稀树草原; 使用最多的遥感数据是机载高光谱数据, 而激光雷达作为补充数据, 通过单木分割及提供单木的三维垂直结构信息, 显著提高了分类精度; 支持向量机和随机森林作为应用最广的非参数分类算法, 平均分类精度达80%; 随着计算机技术及机器学习领域的不断成熟, 人工神经网络在物种识别领域得以迅速发展。基于此, 本文对目前基于遥感数据的植物物种分类与识别中在分类对象复杂性、多源遥感数据整合、植物物候与纹理特征整合和分类算法技术等方面面临的挑战进行了总结, 并建议通过整合多时相监测数据、高光谱和激光雷达数据、短波红外等特定波谱信息、采用深度学习等方法来提高分类精度。     

Are vocal characteristics related to leadership patterns in mixed‐species bird flocks?

What structures the organization of mixed‐species bird flocks, so that some ‘nuclear’ species lead the flocks, and others follow? Previous research has shown that species actively listen to each other, and that leaders are gregarious; such gregarious species tend to make contact calls and hence may be vocally conspicuous. Here we investigated whether vocal characteristics are associated with leadership, using a global dataset of mixed‐species flock studies and recordings from sound archives. We first asked whether leaders are different from following or occasional species in flocks in the proportion of the recordings that contain calls (n = 58 flock studies, 145 species), and especially alarm calls (n = 111 species). We found that leaders tended to have a higher proportion of their vocalizations that were classified as calls than occasional species, and both leaders and following species had a significantly higher proportion of their calls rated as alarms compared to occasional species. Next, we investigated the acoustic characteristics of flock participants’ calls, hypothesizing that leaders would make more calls, and have less silence on the recordings. We also hypothesized that leaders’ calls would be simple acoustically, as contact calls tend to be, and thus similar to each other, as well as being detectable, in being low frequency and with high frequence bandwidth. The analysis (n = 45 species, 169 recordings) found that only one of these predictions was supported: leading species were less often silent than following or occasional species. Unexpectedly, leaders’ calls were less similar to each other than occasional species. The greater amount of information available and the greater variety of that information support the hypothesis that leadership in flocks is related to vocal communication. We highlight the use of sound archives to ask questions about behavioral and community ecology, while acknowledging some limitations of such studies.     

Contributions of MIR to soundscape ecology. Part 3: Tagging and classifying audio features using a multi-labeling k-nearest neighbor approach

Scientists are using acoustic monitoring to assess the impact of altered soundscapes on wildlife communities and human systems. In the soundscape ecology field, monitoring and analyses approaches rely on the interdisciplinary intersection of ecology, acoustics, and computer science. Combining theory and practice of each field in the context of Knowledge Discovery in Databases (KDD), soundscape ecologists provide innovative monitoring solutions for ecologically-driven research questions. We propose a soundscape content analysis framework for improved knowledge outcome with assistance of the new multi-label (ML) concept.Here, we investigated the effectiveness of a ML k-nearest neighbor algorithm (ML-kNN) for labeling concurrent soundscape components within a single recording. We manually labeled 1200 field recordings for the presence of soundscape components and extracted ecological acoustic features, audio profile features, and Gaussian-mixture model features for each recording. Then, we tested the ML-kNN algorithm accuracy with well-established metrics adapted to ML learning.We found that seventeen unique acoustic features could predict a set of biophonic, geophonic, and anthrophonic labels for a single field recording with average precision of 0.767. However, certain labels were predicted incorrectly depending on the time of day and co-occurrence of that label with another label, suggesting further refinement is needed to improve the accuracy of predicted labels.Overall, this ML classification approach could enable researchers to label field recordings more quickly and generate an “alert” system for monitoring changes in a specific sound class. Ultimately, the adaptation of the ML algorithm may provide soundscape ecologists with new metadata labels that are searchable in large databases of soundscape field recordings.     

The application of rapid evaporative ionization mass spectrometry in the analysis of Drosophila species—a potential new tool in entomology

There is increasing emphasis on the use of new analytical approaches in subject analysis and classification, particularly in respect to minimal sample preparation. Here, we demonstrate that rapid evaporative ionization mass spectrometry (REIMS), a method that captures metabolite mass spectra after rapid combustive degradation of an intact biological specimen, generates informative mass spectra from several arthropods, and more specifically, is capable of discerning differences between species and sex of several adult Drosophila species. A model including five Drosophila species, built using pattern recognition, achieves high correct classification rates (over 90%) using test datasets and is able to resolve closely related species. The ease of discrimination of male and female specimens also demonstrates that sex-specific differences reside in the REIMS metabolite patterns, whether analysed across all five species or specifically for D. melanogaster. Further, the same approach can correctly discriminate and assign Drosophila species at the larval stage, where these are morphologically highly similar or identical. REIMS offers a novel approach to insect typing and analysis, requiring a few seconds of data acquisition per sample and has considerable potential as a new tool for the field biologist.     

Acoustic localization of terrestrial wildlife: Current practices and future opportunities

Autonomous acoustic recorders are an increasingly popular method for low‐disturbance, large‐scale monitoring of sound‐producing animals, such as birds, anurans, bats, and other mammals. A specialized use of autonomous recording units (ARUs) is acoustic localization, in which a vocalizing animal is located spatially, usually by quantifying the time delay of arrival of its sound at an array of time‐synchronized microphones. To describe trends in the literature, identify considerations for field biologists who wish to use these systems, and suggest advancements that will improve the field of acoustic localization, we comprehensively review published applications of wildlife localization in terrestrial environments. We describe the wide variety of methods used to complete the five steps of acoustic localization: (1) define the research question, (2) obtain or build a time‐synchronizing microphone array, (3) deploy the array to record sounds in the field, (4) process recordings captured in the field, and (5) determine animal location using position estimation algorithms. We find eight general purposes in ecology and animal behavior for localization systems: assessing individual animals' positions or movements, localizing multiple individuals simultaneously to study their interactions, determining animals' individual identities, quantifying sound amplitude or directionality, selecting subsets of sounds for further acoustic analysis, calculating species abundance, inferring territory boundaries or habitat use, and separating animal sounds from background noise to improve species classification. We find that the labor‐intensive steps of processing recordings and estimating animal positions have not yet been automated. In the near future, we expect that increased availability of recording hardware, development of automated and open‐source localization software, and improvement of automated sound classification algorithms will broaden the use of acoustic localization. With these three advances, ecologists will be better able to embrace acoustic localization, enabling low‐disturbance, large‐scale collection of animal position data.     

Species-specific detection and quantification of common barnacle larvae from the Japanese coast using quantitative real-time PCR

Species-specific detection and quantification methods for barnacle larvae using quantitative real-time polymerase chain reaction (qPCR) were developed. Species-specific primers for qPCR were designed for 13 barnacle species in the mitochondrial 12S ribosomal RNA gene region. Primer specificity was examined by PCR using template DNA extracted from each of the 13 barnacle species, other unidentified barnacle species, and field collected zooplankton samples. The resulting PCR products comprised single bands following agarose gel electrophoresis when the templates corresponded to primers. The amplifications were highly species-specific even for the field plankton samples. The field plankton samples were subjected to qPCR assay. The calculated DNA contents for each barnacle species were closely correlated with the number of larvae measured by microscopic examination. The method could be applied to quantify barnacle larvae in natural plankton samples.     

Guidelines for a priori grouping of species in hierarchical community models

Recent methodological advances permit the estimation of species richness and occurrences for rare species by linking species‐level occurrence models at the community level. The value of such methods is underscored by the ability to examine the influence of landscape heterogeneity on species assemblages at large spatial scales. A salient advantage of community‐level approaches is that parameter estimates for data‐poor species are more precise as the estimation process “borrows” from data‐rich species. However, this analytical benefit raises a question about the degree to which inferences are dependent on the implicit assumption of relatedness among species. Here, we assess the sensitivity of community/group‐level metrics, and individual‐level species inferences given various classification schemes for grouping species assemblages using multispecies occurrence models. We explore the implications of these groupings on parameter estimates for avian communities in two ecosystems: tropical forests in Puerto Rico and temperate forests in northeastern United States. We report on the classification performance and extent of variability in occurrence probabilities and species richness estimates that can be observed depending on the classification scheme used. We found estimates of species richness to be most precise and to have the best predictive performance when all of the data were grouped at a single community level. Community/group‐level parameters appear to be heavily influenced by the grouping criteria, but were not driven strictly by total number of detections for species. We found different grouping schemes can provide an opportunity to identify unique assemblage responses that would not have been found if all of the species were analyzed together. We suggest three guidelines: (1) classification schemes should be determined based on study objectives; (2) model selection should be used to quantitatively compare different classification approaches; and (3) sensitivity of results to different classification approaches should be assessed. These guidelines should help researchers apply hierarchical community models in the most effective manner.