Investigating bird call identification uncertainty using data from processed audio recordings

Mortimer, James A. J.; Greene, Terry

doi:10.20417/nzjecol.41.10

Cited by 19 publications

(20 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Panasonic OverEar (̴ noise control") RP-HC200 headphones were used to listen to the recording. For both sampling approaches, the presence/absence of bird taxa was scored across a set of thirty 10-second snippets or subsamples (Mortimer and Greene 2017). From these data, the number of taxa recorded per five-minute sample (as well as per 10-second subsample) was also calculated.…”

Section: Protocol For Manual Listeningmentioning

confidence: 99%

“…Therefore, the honeyeaters Tūī (Prosthemadera novaeseelandiae) and Bellbird (Anthornis melanura; Family: Meliphagidae), were scored as an aggregate taxon, as were Blackbird (Turdus merula) and Song Thrush (T. philomelos; Family: Turidade), in order to streamline the identification process and to minimize false-positive detections (Mortimer and Greene 2017). For the sake of brevity we refer hereafter to analysis of species' call rates, although recognizing that two of the "species" are actually higher order taxa.…”

Section: Protocol For Manual Listeningmentioning

confidence: 99%

See 1 more Smart Citation

Efficient sampling of avian acoustic recordings: intermittent subsamples improve estimates of single species prevalence and total species richness

Cook¹,

Hartley²

2018

ACE

View full text Add to dashboard Cite

ABSTRACT. Automated sound recording devices have become an important monitoring tool used to estimate species richness and abundance of birds in a variety of ecological and conservation studies. The prevalence of calls detected in a specific time period can be used as an index of relative abundance, to compare between populations. However, the statistical power to infer true differences in abundance between populations is low when detections are highly aggregated in time leading to high variance between samples from the same population. Here, we used two different sampling methods, and used the data from each to calculate species richness and acoustic prevalence of nine bird taxa from a total of 50 sound recordings. The first method simulated typical monitoring techniques used by observers in the field by using a continuous five-minute section of the recording. The second method used the first 10 seconds of each minute to create a composite recording, also of five minutes total duration. There was no difference in the mean prevalence index between methods. The intermittent samples, however, produced prevalence indices with a lower standard deviation (mean difference = 19 %), detected 26% more species per five-minute sample and required 60% less total listening time to detect as many species as the continuous method. The only cost of subsampling from a long recording is the extra digital memory and battery life required to obtain the recordings in the first place. Given that these costs are minor, the intermittent method holds much promise because it detects species more efficiently and provides greater power to detect differences in a species' relative abundance, which in turn should allow for better-informed management regarding population status and trends.Échantillonnage efficace d'enregistrements d'oiseaux : le sous-échantillonnage intermittent améliore les estimations d'occurrence pour une espèce et de la richesse spécifique totale RÉSUMÉ. Les appareils d'enregistrement automatiques sont devenus des outils de suivi importants pour estimer la richessse spécifique et le nombre d'oiseaux dans le cadre de diverses recherches en écologie et conservation. La fréquence des chants détectés dans une période spécifique peut servir d'indice d'abondance relative, afin de comparer des populations. Toutefois, la puissance statistique à révéler les vraies différences d'abondance entre des populations est faible lorsque les détections sont très rapprochées dans le temps, ce qui engendre une grande variance entre les échantillons d'une même population. Dans la présente étude, nous avons utilisé deux méthodes différentes d'échantillonnage et les données provenant de celles-ci pour calculer la richesse spécifique et l'occurrence de neuf taxons d'oiseaux à partir de 50 enregistrements. La première méthode simulait la technique habituellement utilisée par les observateurs sur le terrain, soit l'écoute d'une section de 5 minutes continues de l'enregistrement. La seconde méthode reposait sur les 10 premières secondes de chaque mi...

show abstract

Section: Protocol For Manual Listeningmentioning

confidence: 99%

Section: Protocol For Manual Listeningmentioning

confidence: 99%

Efficient sampling of avian acoustic recordings: intermittent subsamples improve estimates of single species prevalence and total species richness

Cook¹,

Hartley²

2018

ACE

View full text Add to dashboard Cite

show abstract

“…petrels), insects, environmental noise (such as wind, rain, streams and thunder) and anthropogenic noise such as aeroplanes. Many of the recordings were a randomly selected subset of raw recordings from the National Biodiversity Monitoring and Reporting System, for which the details of recorder deployment can be found in (Mortimer & Greene, ). For these recordings we report experiments on filters that are developed for three nocturnal species and two diurnal species: North Island Brown Kiwi ( Apteryx mantelli , Bartlett) a nocturnal flightless bird with sexually dimorphic calls in the frequency range 500–8,000 Hz.…”

Section: Methodsmentioning

confidence: 99%

Wavelet filters for automated recognition of birdsong in long‐time field recordings

et al. 2020

View full text Add to dashboard Cite

Ecoacoustics has the potential to provide a large amount of information about the abundance of many animal species at a relatively low cost. Acoustic recording units are widely used in field data collection, but the facilities to reliably process the data recorded – recognizing calls that are relatively infrequent, and often significantly degraded by noise and distance to the microphone – are not well‐developed yet. We propose a call detection method for continuous field recordings that can be trained quickly and easily on new species, and degrades gracefully with increased noise or distance from the microphone. The method is based on the reconstruction of the sound from a subset of the wavelet nodes (elements in the wavelet packet decomposition tree). It is intended as a preprocessing filter, therefore we aim to minimize false negatives: false positives can be removed in subsequent processing, but missed calls will not be looked at again. We compare our method to standard call detection methods, and also to machine learning methods (using as input features either wavelet energies or Mel‐Frequency Cepstral Coefficients) on real‐world noisy field recordings of six bird species. The results show that our method has higher recall (proportion detected) than the alternative methods: 87% with 85% specificity on >53 hr of test data, resulting in an 80% reduction in the amount of data that needed further verification. It detected >60% of calls that were extremely faint (far away), even with high background noise. This preprocessing method is available in our AviaNZ bioacoustic analysis program and enables the user to significantly reduce the amount of subsequent processing required (whether manual or automatic) to analyse continuous field recordings collected by spatially and temporally large‐scale monitoring of animal species. It can be trained to recognize new species without difficulty, and if several species are sought simultaneously, filters can be run in parallel.

show abstract

“…Moreover, auditory detections are sometimes uncertain (Mortimer and Greene, 2017). When point count observations have corresponding photographic or audio evidence material, the bias between observers can be lessened, but these verification data are seldom available.…”

Section: Point Counts Suffer From a Trade-off Between Observation Timmentioning

confidence: 99%

Autonomous bird sound recording outperforms direct human observation: Synthesis and new evidence

Darras

Batáry

Furnas

et al. 2017

Preprint

View full text Add to dashboard Cite

Abstract1) Autonomous sound recording techniques have gained considerable traction in the last decade, but the question still remains whether they can replace human observation surveys to sample some animal taxa.Especially bird survey methods have been tested using classical point counts and autonomous sound recording techniques. 2) We review the latest information by comparing both survey methods' standardization, verifiability, sampling completeness, data types, compatibility, and practicality by means of a systematic review and a meta-analysis of alpha and gamma species richness levels sampled by both methods across 20 separate studies. 17, 2017; 3) Although sound recording surveys have hitherto not enjoyed the most effective setups, they yield very similar results in terms of alpha and gamma species richness. We also reveal the crucial importance of the microphone (high signal-to-noise ratio) as the sensor that replaces human senses. 4) We discuss key differences between both methods, while richness estimates are closely related and 81% of all species were detected by both methods. Sound recording techniques provide a more powerful and promising tool to monitor birds in a standardized, verifiable, and exhaustive way against the golden standard of point counts. Advantages include the capability of sampling continuously through day or season and of difficult-to-reach regions in an autonomous way, avoidance of observer bias and human disturbance effects and higher detection probability of rare species due to extensive recordings.peer-reviewed) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was . http://dx.doi.org/10.1101/117119 doi: bioRxiv preprint first posted online Mar.2

show abstract

Investigating bird call identification uncertainty using data from processed audio recordings

Cited by 19 publications

References 18 publications

Efficient sampling of avian acoustic recordings: intermittent subsamples improve estimates of single species prevalence and total species richness

Efficient sampling of avian acoustic recordings: intermittent subsamples improve estimates of single species prevalence and total species richness

Wavelet filters for automated recognition of birdsong in long‐time field recordings

Autonomous bird sound recording outperforms direct human observation: Synthesis and new evidence

Contact Info

Product

Resources

About