4.7 Article

Low-cost open-source recorders and ready-to-use machine learning approaches provide effective monitoring of threatened species

Related references

Note: Only part of the references are listed.
Review Multidisciplinary Sciences

Computational bioacoustics with deep learning: a review and roadmap

Dan Stowell

Summary: Animal vocalisations and natural soundscapes provide valuable evidence for studying animal behaviors, populations, and ecosystems. Computational bioacoustics, especially deep learning, has made significant progress recently, benefiting from affordable digital sound recording devices and advancements in informatics. However, there are still unresolved problems in this field, and tasks that have not been fully explored. This paper reviews the latest developments in deep learning for computational bioacoustics and offers a subjective but principled roadmap for future research.

PEERJ (2022)

Article Acoustics

Estimating Florida manatee (Trichechus manatus latirostris) abundance using passive acoustic methods

Athena M. Rycyk et al.

Summary: This study evaluated the effectiveness of using vocal detections to estimate manatee abundance and found that vocalization counts can serve as strong predictors of manatee numbers, particularly in areas with clear water.

JASA EXPRESS LETTERS (2022)

Article Ecology

BirdNET: A deep learning solution for avian diversity monitoring

Stefan Kahl et al.

Summary: Recent advances in deep artificial neural networks have revolutionized the field of bird sound recognition, with models like BirdNET able to accurately identify 984 North American and European bird species. Task-specific model designs and training regimes play a crucial role in audio event recognition, while high temporal resolution of input spectrograms improves classification performance for bird sounds.

ECOLOGICAL INFORMATICS (2021)

Review Ornithology

Estimating bird density using passive acoustic monitoring: a review of methods and suggestions for further research

Cristian Perez-Granados et al.

Summary: Passive acoustic monitoring is a non-invasive tool for automated wildlife monitoring, with potential benefits including reducing biases related to traditional field surveys. While using autonomous recording units (ARUs) to estimate animal density has been traditionally challenging, recent studies have proposed approaches to estimate bird density using ARUs. Many studies have shown that bird estimates obtained from ARUs are consistent with those from human surveyors, and some methods have been successful in reducing biases in acoustic surveys.
Article Ecology

Survey coverage, recording duration and community composition affect observed species richness in passive acoustic surveys

Connor M. Wood et al.

Summary: The study indicates that observed species richness is influenced by survey coverage and recording duration, while species detection probabilities are affected by reducing the number of days of recording and daily recording duration. The impact of survey coverage decreases as recording duration decreases, and rare species are more likely to be underestimated as survey coverage decreases. Increasing recording duration lowers the cost per species observed.

METHODS IN ECOLOGY AND EVOLUTION (2021)

Article Ecology

The Kinabalu Recorder, a new passive acoustic and environmental monitoring recorder

Eva Catharina M. Karlsson et al.

Summary: Low-cost options for long-term acoustic monitoring in terrestrial ecology are increasingly available, but integration with sensors for additional data logging is rare. A hardware system called the Kinabalu Recorder, designed for both acoustic and environmental data logging, offers a cost-effective solution that promotes cross-field collaboration.

METHODS IN ECOLOGY AND EVOLUTION (2021)

Article Multidisciplinary Sciences

Comparing recurrent convolutional neural networks for large scale bird species classification

Gaurav Gupta et al.

Summary: This study presents a deep learning approach for large-scale prediction and analysis of bird acoustics from 100 different bird species. The hybrid modeling approach involving CNN and RNN shows the most accurate model on this dataset, with the best performing model achieving an average accuracy of 67% over 100 bird species and a highest accuracy of 90% for the Red crossbill species. Visual analysis of learned representations reveals intuitive clustering of related bird species.

SCIENTIFIC REPORTS (2021)

Article Zoology

Distance estimation of howling golden jackals (Canis aureus) using relative sound level

Graf Lukas et al.

Summary: The study demonstrates that Golden jackals have a complex repertoire of calls used for different communication types. By using bioacoustic stimulation and passive recording devices, researchers can effectively monitor and estimate the distance and number of responding animals in Golden jackal packs.

MAMMAL RESEARCH (2021)

Article Biodiversity Conservation

Passive Acoustic Monitoring of Chaco Chachalaca (Ortalis canicollis) Over a Year: Vocal Activity Pattern and Monitoring Recommendations

Cristian Perez-Granados et al.

Summary: The declining Neotropical bird, Chaco Chachalaca, was monitored using passive acoustic monitoring and automated signal recognition software for the first time. The study revealed the vocal behavior of the species and proposed an acoustic monitoring protocol. The species was found to be vocally active throughout the day and night, with breeding season occurring in the last months of the dry season in the Brazilian Pantanal.

TROPICAL CONSERVATION SCIENCE (2021)

Article Biodiversity Conservation

Combining bioacoustics and occupancy modelling for improved monitoring of rare breeding bird populations

Carlos Abrahams et al.

ECOLOGICAL INDICATORS (2020)

Article Biodiversity Conservation

Testing a New Passive Acoustic Recording Unit to Monitor Wolves

Shannon M. Barber-Meyer et al.

WILDLIFE SOCIETY BULLETIN (2020)

Review Biology

Terrestrial Passive Acoustic Monitoring: Review and Perspectives

Larissa Sayuri Moreira Sugai et al.

BIOSCIENCE (2019)

Article Biodiversity Conservation

Comparing the sampling performance of sound recorders versus point counts in bird surveys: A meta-analysis

Kevin Darras et al.

JOURNAL OF APPLIED ECOLOGY (2018)

Article Zoology

Automatic recognition of bird individuals on an open set using as- is recordings

Ladislav Ptacek et al.

BIOACOUSTICS-THE INTERNATIONAL JOURNAL OF ANIMAL SOUND AND ITS RECORDING (2016)

Article Ornithology

Call-based identification as a potential tool for monitoring Great Spotted Kiwi

Jennifer M. Dent et al.

EMU-AUSTRAL ORNITHOLOGY (2016)

Article Ecology

Applying bioacoustic methods for long-term monitoring of a nocturnal wetland bird

Karl-Heinz Frommolt et al.

ECOLOGICAL INFORMATICS (2014)

Article Ornithology

Comparison of audio recording system performance for detecting and monitoring songbirds

Robert S. Rempel et al.

JOURNAL OF FIELD ORNITHOLOGY (2013)

Article Ornithology

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

Antonio Celis-Murillo et al.

JOURNAL OF FIELD ORNITHOLOGY (2009)

Review Ecology

Bird song: The interface of evolution and mechanism

J Podos et al.

ANNUAL REVIEW OF ECOLOGY EVOLUTION AND SYSTEMATICS (2004)