Passive acoustic monitoring in ecology and conservation.

Browning, Ella; Gibb, Rory; Glover‐Kapfer, Paul; Jones, Kate

doi:10.25607/obp-876

Cited by 10 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Passive acoustic monitoring using automated recorders has gained popularity in ecology, encompassing a wide range of species and applications (Sugai et al, 2019). However, the utilization of largescale projects often results in an overwhelming accumulation of data that becomes arduous, if not nearly impossible, to analyse manually (Browning et al, 2017). In response to these challenges, there has been an increase in the development of tools and techniques that help process this data automatically (Priyadarshani et al, 2020;Sueur & Farina, 2015).…”

Section: Contextmentioning

confidence: 99%

ECOGEN: Bird sounds generation using deep learning

Guei,

Christin,

Lecomte

et al. 2023

Methods Ecol Evol

View full text Add to dashboard Cite

Large‐scale acoustic projects generate vast amounts of data that can now be efficiently processed using deep learning tools. However, these tools often face limitations due to sound labeling and imbalanced sampling. Data augmentation can help overcome such challenges, particularly through the generation of synthetic and lifelike sounds. Synthetic samples can be valuable not only for deep learning but also for species with limited available data. Despite advancements in computer power, sound generation remains a time‐consuming process, even requiring a substantial number of samples. We present ECOGEN, a novel deep learning approach designed to generate realistic bird songs for biologists and ecologists. The primary objective of ECOGEN is to enhance the number of samples in under‐represented bird song classes, thereby improving the performance and robustness of classifiers in ecological research.The ECOGEN framework employs spectrograms as a representation of bird songs and leverages proven image generation techniques to create new spectrograms, subsequently converted back to digital audio signals. As a class‐agnostic tool, ECOGEN is applicable to a wide range of biophonic sounds, including mammal and insect calls. We show that adding samples generated by ECOGEN to a bird song classifier improved the classification accuracy by 12% on average and improved results compared with classic data augmentation techniques 80% of the time. Our approach is both fast and efficient, enabling the generation of synthetic bird songs on standard computing resources. By facilitating the creation of synthetic bird songs, ECOGEN can contribute to the conservation of endangered bird species, while providing valuable insights into their vocalizations, behaviours and habitat preferences. Future development of ECOGEN can be easily implemented and could focus on incorporating additional configurable parameters during the generation phase for increased control over the output, catering to the specific needs of biologists.

show abstract

Section: Contextmentioning

confidence: 99%

ECOGEN: Bird sounds generation using deep learning

Guei,

Christin,

Lecomte

et al. 2023

Methods Ecol Evol

View full text Add to dashboard Cite

show abstract

“…Due to the variability in the types of sounds that can occur in underwater environments, processing of acoustic data for biodiversity purposes requires that a wide range of sounds can be accurately detected and differentiated. Although machine learning has expedited the review process, most methods remain either semi-automated and reliant on an experienced acoustic analyst or are limited to detecting specific, stereotyped calls (Browning et al, 2017). To achieve more timely processing of passive acoustic data and inform ecosystem monitoring efforts, techniques for the efficient assessment of continuously collected data are required.…”

Section: Introductionmentioning

confidence: 99%

Acoustic indices respond to specific marine mammal vocalizations and sources of anthropogenic noise

Ferguson¹,

Clayton

Sakai

2023

Front. Mar. Sci.

View full text Add to dashboard Cite

Using passive acoustic methods for biodiversity conservation and effective ecosystem monitoring is hindered by laborious, human-mediated processes of accurately identifying biologic and anthropogenic sounds within large datasets. Soundscape ecology provides a potential means of addressing this need through the use of automated acoustic-based biodiversity indices, which show promise in representing biodiversity in terrestrial environments. However, the direct relationship between specific underwater sounds and acoustic index measurements are largely unexplored. Using passive acoustic data collected from three broadband hydrophones within the Ocean Observatories Initiative’s cabled arrays in the Pacific northwest, we identified periods of vocalizing marine mammals and sources of anthropogenic noise. Automated calculations of seven acoustic indices were compared were compared across biologic and anthropogenic sound type and call parameters. Although several index measurements did not vary significantly, the Acoustic Complexity Index (ACI) measurements increased in response to echolocation clicks from sperm whales (Physeter macrocephalus) and burst pulses originating from unidentified delphinid species. Measurements of the Bioacoustic Index (BI) decreased dramatically in responses sperm whale echolocation clicks, a more obvious trend when loud clicks were parsed from moderate and quite clicks. Correlations coefficient and confidence interval values between ACI and BI measurements and call characteristics from sperm whales indicate a moderate to strong relationship, which was not found in correlations with delphnid calls. A generalized linear mixed-effect model indicated multiple species and sound types contribute significantly to the variation of several index measurements. Noise generated by passing ships consistently resulted in decreased values for the Normalized Difference Soundscape Index (NDSI) and Total Entropy (H) as compared to quiet periods and periods with vocalizing marine mammals. These findings provide information on the relationship between several acoustic indices and specific underwater sounds produced by marine mammals and anthropogenic sources. This ground-truthing endeavor expands the understanding of acoustic indices and their potential use as a tool for conservation and ecosystem health management purposes.

show abstract

“…PAM is one of the newer techniques that allows for recording acoustically active species with limited disturbance to a species' activity and breeding behaviour (Aide et al 2013;Ospina et al 2013;Alvarez-Berríos et al 2016;Browning et al 2017). While PAM has not been intensively used to study bird communities in the wetlands of South Africa, it has been implemented in other ecosystems around the world and proven to yield rich data sets over a relatively short period compared with traditional methods (Stowell et al 2019).…”

Section: Introductionmentioning

confidence: 99%

Passive acoustic monitoring detects new records of globally threatened birds in a high-elevation wetland (Free State, South Africa)

Mosikidi,

Le Maitre,

Steenhuisen

et al. 2023

Bird Conservation International

View full text Add to dashboard Cite

Summary Wetlands act as islands of high biodiversity within the ecological landscape and provide crucial ecosystem services to society. Anthropogenic activities are driving wetland degradation and it has become increasingly rare to find wetlands that do not show signs of biodiversity loss or alteration. The exacerbated loss of biodiversity in wetlands has a negative impact on the local economy and ecosystem services provided by these systems. We responded to the South African National Biodiversity Assessment (NBA) call to document wetland biodiversity against the backdrop of sustained wetland degradation in southern Africa. We monitored the soundscape of a high-elevation wetland in the Golden Gate Highlands National Park (GGHNP) from June 2019 to December 2020 across 24 localities using a rolling grid layout. We detected 35.9% of the avian species previously recorded from ad hoc sightings in the GGHNP of which 68.1% are wetland obligate species. We contributed an additional 10.2% new species records to the avian diversity of the GGHNP, including 24 species that are considered threatened by the International Union for Conservation of Nature (IUCN). Our remote monitoring technique enabled the first ever continuous monitoring using remote acoustic equipment for a high-elevation wetland in South Africa, thus providing a valuable contribution to the NBA call.

show abstract

Passive acoustic monitoring in ecology and conservation.

Cited by 10 publications

References 0 publications

ECOGEN: Bird sounds generation using deep learning

ECOGEN: Bird sounds generation using deep learning

Acoustic indices respond to specific marine mammal vocalizations and sources of anthropogenic noise

Passive acoustic monitoring detects new records of globally threatened birds in a high-elevation wetland (Free State, South Africa)

Contact Info

Product

Resources

About