Assessing abundance and habitat preferences of Goldcrest <i>Regulus regulus</i> and Firecrest <i>Regulus ignicapilla</i> using passive acoustic monitoring and point-count surveys in temperate forest ecosystem in Poland

AimSpecies distribution models (SDMs) are essential tools in ecology and conservation. However, the scarcity of visual sightings of marine mammals in remote polar areas hinders the effective application of SDMs there. Passive acoustic monitoring (PAM) data provide year‐round information and overcome foul weather limitations faced by visual surveys. However, the use of PAM data in SDMs has been sparse so far. Here, we use PAM‐based SDMs to investigate the spatiotemporal distribution of the critically endangered Antarctic blue whale in the Weddell Sea.LocationThe Weddell Sea.MethodsWe used presence‐only dynamic SDMs employing visual sightings and PAM detections in independent models. We compared the two independent models with a third combined model that integrated both visual and PAM data, aiming at leveraging the advantages of each data type: the extensive spatial extent of visual data and the broader temporal/environmental range of PAM data.ResultsVisual and PAM data prove complementary, as indicated by a low spatial overlap between daily predictions and the low predictability of each model at detections of other data types. Combined data models reproduced suitable habitats as given by both independent models. Visual data models indicate areas close to the sea ice edge (SIE) and with low‐to‐moderate sea ice concentrations (SIC) as suitable, while PAM data models identified suitable habitats at a broader range of distances to SIE and relatively higher SIC.Main ConclusionsThe results demonstrate the potential of PAM data to predict year‐round marine mammal habitat suitability at large spatial scales. We provide reasons for discrepancies between SDMs based on either data type and give methodological recommendations on using PAM data in SDMs. Combining visual and PAM data in future SDMs is promising for studying vocalized animals, particularly when using recent advances in integrated distribution modelling methods.

Passive acoustic monitoring (PAM) has become an important tool for surveying birds, and there is a growing demand for approaches to obtain abundance and behavioural information from PAM recordings. Changes in bird populations have been assessed by counting recorded calls and calculating the vocal activity rate (VAR, i.e. the number of calls per recording time). However, bird calls could be counted in various ways and depending on species traits, these call counts could give us different insights on bird abundance, vocal behaviour and/or habitat use. Our study had two goals: (1) to present and evaluate two new indices based on call counts, the detection rate (DR, i.e. the number of 1‐min recordings in which the presence of a target vocalization is detected) and the maximum count per minute (MAX, i.e. the maximum number of calls found in a 1‐min recording); and (2) to present a conceptual framework showing how the interpretations of VAR, DR and MAX could depend on the index and on species traits. For three Neotropical bird species with distinct traits, we calculated VAR, DR and MAX based on PAM data from 25 sites in the Yucatan Peninsula (Mexico) that varied in their degree of anthropogenic habitat disturbance. We found moderate to high correlations between the indices and higher temporal variability in VAR compared to DR and MAX. We also found different effect sizes of habitat disturbance on the three species and indices. We suggest that DR might be a more reliable index of relative abundance than VAR for species whose calling behaviour exhibits a high cue rate and that MAX may be suitable for estimating family or flock size in gregarious birds. Our findings show the potential usefulness of developing new indices based on call counts to generate ecological hypotheses and assess changes in bird abundance and behaviour.

Passive acoustic monitoring (PAM) data collection has been growing exponentially, resulting in petabytes of data that document ocean soundscapes, how they change over time, and what animals use these ecosystems at varying timescales. Efficiently extracting this critical information and comparing it to other datasets in the context of ecosystem-based management is a Big Data challenge that traditional desktop processing methods cannot address. The curation, management, and dissemination of PAM datasets is another challenge in need of collaborative progress. To meet these exigencies, a multi-agency funded Sound Cooperative (SoundCoop) project is building community-focused, national cyberinfrastructure capability for PAM data to promote improved, scalable and sustainable accessibility and applications for management and science. Driven by partnerships and framed by four case studies, the SoundCoop has established guidance on the standardized processing of sound level metrics using free software toolkits and begun developing core cyberinfrastructure components that future PAM projects can leverage. U.S. and international scientists contributed PAM data collected across 10 long-term monitoring projects to operationalize the production of hybrid-millidecade spectra across a diversity of labs/instruments. Collectively, the contributed data demonstrate the value of standardized processing that enables the creation of comparable results from disparate monitoring efforts.

Technological advances over the last two decades have seen a large uptake in passive acoustic monitoring (PAM) to supplement traditional biodiversity surveys. To address the growing backlog of acoustic recordings these projects create, multispecies acoustic classifiers are now widely used in favour of manually processing data. Despite the uptake of these technologies, the efficacy of these recognisers needs to be evaluated to ensure they are detecting levels of diversity and community structure similar to traditional surveys. This study compared metrics of avian diversity across eastern Australia, between traditional bird surveys (both dawn, and dawn + nocturnal), and PAM processed using BirdNET, a widely used multi-species classifier. On average, PAM and multi-species classifiers returned higher values for all assessed diversity metrics (Species Richness, Chao2 estimators, Petchey's Functional Diversity, Rao's Q and Phylogenetic Distance), however traditional surveys supplemented with nocturnal surveys returned intermediate values. The efficacy of classifiers varied considerably across the study locations, with the number of incorrect identifications in the tropics substantially higher than those in temperate areas. Both methodologies failed to detect certain taxonomic groups, some threatened species, and detected significantly different avian communities. While these results provide a promising outlook for the future of PAM, it underscores the importance of maintaining traditional surveys as part of biodiversity monitoring, and relying on skilled ornithologists to ensure recorded acoustic data is appropriately interrogated. Furthermore, this study cautions against relying solely on automated classifiers in regions where training data for models is depauperate, such as Australia's tropical and subtropical woodlands. • We compared avian diversity predictions between traditional and PAM surveys. • PAM consistently detected more species sooner than traditional surveys. • BirdNET processed PAM returned higher values for all diversity metrics. • Each survey methods detected unique species, and different community composition. • Combined survey approaches should be considered to maximise species detections.

North Atlantic right whale monitoring in Roseway Basin, Canada, is primarily based on short‐term (<14 d) visual surveys conducted during August–September. Variability in survey effort has been the biggest limiting factor to studying changes in the population's occurrence and habitat use. Such efforts could be enhanced considerably using passive acoustic monitoring (PAM). We sought to determine if variation in whale presence, relative abundance, demography, and/or behavior (estimated through visual surveys) could be explained by variation in three right whale call types in this habitat. A generalized linear model was fit to 23 d of concurrent PAM and visual monitoring during four summers within the Roseway Basin Right Whale Critical Habitat boundaries. The model revealed significant positive relationships between relative abundance, call counts and presence of surface‐active group behavior. PAM can refine daily right whale presence estimates. While visual observations (n= 23 d) implied a 40% decline in right whale presence during 2014–2015 relative to 2004–2005, PAM data (n= 211 d) showed right whales were present between 71%–85% of survey days throughout all years analyzed. We demonstrate that PAM is a useful tool to extend periods of right whale monitoring, especially in areas where visual monitoring efforts may be limited.

Integration of passive acoustic monitoring data into OBIS-SEAMAP, a global biogeographic database, to advance spatially-explicit ecological assessments

The urgent need for effective wildlife monitoring solutions in the face of global biodiversity loss has resulted in the emergence of conservation technologies such as passive acoustic monitoring (PAM). While PAM has been extensively used for marine mammals, birds, and bats, its application to primates is limited. Black-and-white ruffed lemurs (Varecia variegata) are a promising species to test PAM with due to their distinctive and loud roar-shrieks. Furthermore, these lemurs are challenging to monitor via traditional methods due to their fragmented and often unpredictable distribution in Madagascar's dense eastern rainforests. Our goal in this study was to develop a machine learning pipeline for automated call detection from PAM data, compare the effectiveness of PAM versus in-person observations, and investigate diel patterns in lemur vocal behavior. We did this study at Mangevo, Ranomafana National Park by concurrently conducting focal follows and deploying autonomous recorders in May-July 2019. We used transfer learning to build a convolutional neural network (optimized for recall) that automated the detection of lemur calls (57-h runtime; recall = 0.94, F1 = 0.70). We found that PAM outperformed in-person observations, saving time, money, and labor while also providing re-analyzable data. Using PAM yielded novel insights into V. variegata diel vocal patterns; we present the first published evidence of nocturnal calling. We developed a graphic user interface and open-sourced data and code, to serve as a resource for primatologists interested in implementing PAM and machine learning. By leveraging the potential of this pipeline, we can address the urgent need for effective primate population surveys to inform conservation strategies.

Passive acoustic monitoring (PAM) allows for cost-effective, unattended and non-invasive acoustic sampling over an extended period of time and is now an invaluable tool for acoustic monitoring of vocally active species. Its application is rapidly growing in studies covering multiple aspects of avian ecology and behaviour, including presence-absence surveys, population density estimations, threatened species monitoring and anthropogenic impacts on populations. However, the potential for information on year-round variation in male and female vocalisations and the factors affecting duetting behaviour to be derived from PAM has never been exploited. In the present study we deployed automatic recording units (ARU) to investigate long-term sex-specific life strategies based on the vocal activity of the Yellow-breasted Boubou Laniarius atroflavus, an Afromontane, duetting songbird. Using automatic detection we showed strong seasonality in singing performance with males producing solo songs at a higher rate during the breeding than non-breeding season whereas female solos peaked at the end of the breeding season. Duets were produced at a relatively stable rate throughout the year except the time encompassing the turn of the rainy and dry seasons when overall vocal activity was at a low level. In general, year-round singing patterns coincided with the rainy and dry seasons at the study site with vocal activity peaking in the dry season and gradually declining with the onset of rainfall. In addition, we found that boubous were slightly more vocally active when morning temperature was higher, especially in the rainy season. Sex-dependent variation in vocal activity in relation to life cycle stage may suggest that differences between males and females are of functional significance. Most likely, the seasonality of male solo songs could be explained on the basis of sexual selection pressure and that male and female joint vocalizations act as a cooperative behaviour playing a role in territory defence against conspecifics. Our PAM-based results provide new and important insights into how male–female solo songs and duet interactions may be related to year-round territoriality. This may help us to better understand the evolutionary significance of duetting. Furthermore, our findings highlight the link between life cycle events of a tropical songbird and seasonal changes in weather conditions. By tracking the effect of weather on vocal activity, PAM might provide an important indication of how changes in climate may affect bird behaviour.

Passive acoustic monitoring, when coupled with automated signal recognition software, allows researchers to perform simultaneous monitoring at large spatial and temporal scales. This technique has been widely used to monitor cetaceans, bats, birds, and anurans but rarely applied to monitor primates. Here, we evaluated the effectiveness of passive acoustic monitoring and automated signal recognition software for detecting the presence and monitoring the roaring behavior of the Black and Gold Howler Monkey (Alouatta caraya) over a complete annual cycle at one site in the Brazilian Pantanal. The diel pattern of roaring activity was unimodal, with high vocal activity around dawn. The howler monkey showed a clear seasonal pattern of roaring activity, with most of the roars detected during the wet season (74.9%, peak activity during November and December). The maximum vocal activity occurred during the period of maximum flowering and fruit production in the study area, suggesting a potential role of roaring in defending major feeding sites, which is in agreement with the findings of previous studies on the species. However, we cannot rule out the possibility that roaring may serve different purposes. Vocal activity was negatively associated with relative air humidity, which might be related to lower vocal activity on wetter and rainy days, while vocal activity was not related to minimum air temperature. Automated signal recognition software allowed us to detect the species in 89% of the recordings in which it was vocally active, but with a reduced time cost, since the time investment for data analyses was 2% of recording time. The good performance of the recognizer might be related to the long and loud roars of the howler monkey. Further research should be performed to evaluate the effectiveness of automated signal recognition for detecting the calls of different species of primates and under different environmental conditions.

Passive acoustic monitoring (PAM) is a promising, if underused, technology for primate conservation. Successful PAM requires an understanding of the target species' vocal activity patterns and the factors that influence them, but this information remains scarce for most vocal primates. This is true for sportive lemurs ( Lepilemur spp.), which are understudied but otherwise excellent candidates for PAM, being highly vocal and threatened. We deployed autonomous audio recorders to measure vocal activity in the Critically Endangered Nosy Be sportive lemur ( Lepilemur tymerlachsoni ), sampling a 4‐h window from twilight each night for two lunar cycles. Our objectives were to identify suitable call types for monitoring, evaluate a user‐friendly automated call detection algorithm, assess temporal variation in vocal activity, and examine how environmental variables and moon illumination influence vocal activity. Automated call detection found an estimated 38% of all target calls but generated a high rate of false positives (96%). Among three call types, “ouah” calls were common and had the highest detection rate (51%), making them suitable target calls. Call rates were highest in the fourth hour following twilight, increased with temperature and moon illumination, and decreased during rainfall. We also observed variation in vocal activity between recording dates and sites, highlighting the need for sufficient temporal and spatial replication. We present recommendations for improving survey design, detection probability, and population inferences from PAM. The recommendations are specific to L . tymerlachsoni and may guide similar work on other sportive lemurs, although species‐specific differences in vocal behavior and ecology must also be considered.

Chaco Chachalaca ( Ortalis canicollis) is a declining Neotropical bird, for which our current knowledge about its natural history is very limited. Here, we evaluated for first time the utility of passive acoustic monitoring, coupled with automated signal recognition software, to monitor the Chaco Chachalaca, described the vocal behavior of the species across the diel and seasonal cycle patterns, and proposed an acoustic monitoring protocol to minimize error in the estimation of the vocal activity rate. We recorded over a complete annual cycle at three sites in the Brazilian Pantanal. The species was detected on 99% of the monitoring days, proving that this technique is a reliable method for detecting the presence of the species. Chaco Chachalaca was vocally active throughout the day and night, but its diel activity pattern peaked between 0500 and 0900. The breeding season of Chaco Chachalaca in the Brazilian Pantanal, based on seasonal changes in vocal activity, seems to occur during the last months of the dry season, with a peak in vocal activity between August and October. Our results could guide future surveys aiming to detect the presence of the species, both using traditional or acoustic surveys, or to evaluate changes in population abundance using passive acoustic monitoring, for which recorders should be left in the field for a minimum period of nine days to obtain a low-error estimate of the vocal activity of the species. Our results suggest that passive acoustic monitoring might be useful, as a complementary tool to field studies, for monitoring other cracids, a family with several threatened species that are reluctant to human presence.

ABSTRACTAutomated detection of acoustic signals is crucial for effective monitoring of sound‐producing animals and their habitats across ecologically relevant spatial and temporal scales. Recent advances in deep learning have made these approaches more accessible. However, few deep learning approaches can be implemented natively in the R programming environment; approaches that run natively in R may be more accessible for ecologists. The “torch for R” ecosystem has made deep learning with convolutional neural networks (CNNs) accessible for R users. Here, we evaluate a workflow for the automated detection and classification of acoustic signals from passive acoustic monitoring (PAM) data. Our specific goals include (1) present a method for automated detection of gibbon calls from PAM data using the “torch for R” ecosystem, (2) conduct a series of benchmarking experiments and compare the results of six CNN architectures; and (3) investigate how well the different architectures perform on data sets of the female calls from two different gibbon species: the northern gray gibbon (Hylobates funereus) and the southern yellow‐cheeked crested gibbon (Nomascus gabriellae). We found that the highest‐performing architecture depended on the species and test data set. We successfully deployed the top‐performing model for each gibbon species to investigate spatial variation in gibbon calling behavior across two grids of autonomous recording units in Danum Valley Conservation Area, Malaysia and Keo Seima Wildlife Sanctuary, Cambodia. The fields of deep learning and automated detection are rapidly evolving, and we provide the methods and data sets as benchmarks for future work.

Passive Acoustic Monitoring (PAM) has emerged as a pivotal technology for wildlife monitoring, generating vast amounts of acoustic data. However, the successful application of machine learning methods for sound event detection in PAM datasets heavily relies on the availability of annotated data, which can be laborious to acquire. In this study, we investigate the effectiveness of transfer learning and active learning techniques to address the data annotation challenge in PAM. Transfer learning allows us to use pre-trained models from related tasks or datasets to bootstrap the learning process for sound event detection. Furthermore, active learning promises strategic selection of the most informative samples for annotation, effectively reducing the annotation cost and improving model performance. We evaluate an approach that combines transfer learning and active learning to efficiently exploit existing annotated data and optimize the annotation process for PAM datasets. Our transfer learning observations show that embeddings produced by BirdNet, a model trained on high signal-to-noise recordings of bird vocalisations, can be effectively used for predicting anurans in PAM data: a linear classifier constructed using these embeddings outperforms the benchmark by 21.7%. Our results indicate that active learning is superior to random sampling, although no clear winner emerges among the strategies employed. The proposed method holds promise for facilitating broader adoption of machine learning techniques in PAM and advancing our understanding of biodiversity dynamics through acoustic data analysis.

Monitoring vulnerable species inhabiting mountain environments is crucial to track population trends and prioritize conservation efforts. However, the challenging nature of these remote areas poses difficulties in implementing effective and consistent monitoring programmes. To address these challenges, we examined the potential of passive acoustic monitoring of a cryptic high mountain bird species, the Rock Ptarmigan Lagopus muta. For 5 months in each of two consecutive years, we deployed 38 autonomous recording units in 10 areas of the Swiss Alps where the species is monitored by a national count monitoring programme. Once the recordings were collected, we built a machine‐learning algorithm to automate call recognition. We focused on studying the species' daily and seasonal calling phenology and relating these to meteorological and climatic data. Rock Ptarmigans were vocally active from March to July, with a peak of activity occurring between mid‐March and late April, 1 or 2 months earlier than the second half of May when the counts of the monitoring programme take place. The calling rate peaked at dawn before dropping rapidly until sunrise. Daily vocal activity demonstrated a consistent association with weather conditions and moon phase, whereas the timing of seasonal vocal activity varied with temperature and snow conditions. We found that the peak of vocal activity occurred when the snowpack was still thick and snow cover was close to 100% but with a local peak of high temperatures. Between our two study years, the peak of vocal activity occurred 30 days later in the colder year, suggesting phenological plasticity in relation to environmental conditions. Passive acoustic monitoring has the potential to complement conventional acoustic counts of cryptic birds by highlighting periods of higher detectability of individuals, and to survey small populations that often remain undetected during single visits. Moreover, our study supports the idea that passive acoustic monitoring can provide valuable data over large spatial and temporal scales, allowing decryption of hidden ecological patterns and assisting in conservation efforts.

Navy training events involving the use of explosives pose a potential threat to marine mammals. This study used passive acoustic and visual monitoring data to evaluate marine mammals’ behavioral responses to noise from explosive events. Monitoring was conducted during five training events in the Virginia Capes (VACAPES) Range Complex during August/September of 2009–2012. Passive acoustic monitoring methods ranged from a single hydrophone to an array of sonobuoys monitored in real time. Visual monitoring effort over the five events totaled approximately 34 h (day before events: 10.1 h; days of events: 22.3 h; day after events: 1.5 h), yielding a total of 27 marine mammal sightings. Approximately 54 h of acoustic data were collected before, during, and after the five events. Behavioral changes were evaluated based on analysis of vocalizations detected before, during, and after explosions and concurrent data from visual sightings. For time periods with both visual and acoustic monitoring data, detection methods were compared to evaluate effectiveness. Continuing use and evaluation of both visual and passive acoustic methods for monitoring of explosive training events will improve our knowledge of potential impact resulting from explosive events and help improve management and conservation of marine mammals.