Evaluating community-wide temporal sampling in passive acoustic monitoring: A comprehensive study of avian vocal patterns in subtropical montane forests.

Background: From passive acoustic monitoring (PAM) recordings, the vocal activity rate (VAR), vocalizations per unit of time, can be calculated and is essential for assessing bird population abundance. However, VAR is subject to influences from a range of factors, including species and environmental conditions. Identifying the optimal sampling design to obtain representative acoustic data for VAR estimation is crucial for research objectives. PAM commonly uses temporal sampling strategies to decrease the volume of recordings and the resources needed for audio data management. Yet, the comprehensive impact of this sampling approach on VAR estimation remains insufficiently explored. Methods: In this study, we used vocalizations extracted from recordings of 12 bird species, taken at 14 PAM stations situated in subtropical montane forests over a four-month period, to assess the impact of temporal sampling on VAR across three distinct scales: short-term periodic, diel, and hourly. For short-term periodic sampling analysis, we employed hierarchical clustering analysis (HCA) and the coefficient of variation (CV). Generalized additive models (GAMs) were utilized for diel sampling analysis, and we determined the average difference in VAR values per minute for the hourly sampling analysis. Results: We identified significant day and species-specific VAR fluctuations. The survey season was divided into five segments; the earliest two showed high variability and are best avoided for surveys. Data from days with heavy rain and strong winds showed reduced VAR values and should be excluded from analysis. Continuous recordings spanning at least seven days, extending to 14 days is optimal for minimizing sampling variance. Morning chorus recordings effectively capture the majority of bird vocalizations, and hourly sampling with frequent, shorter intervals aligns closely with continuous recording outcomes. Conclusions: While our findings are context-specific, they highlight the significance of strategic sampling in avian monitoring, optimizing resource utilization and enhancing the breadth of monitoring efforts.

An acoustic detection dataset of birds (Aves) in montane forests using a deep learning approach.

Background: Long-term monitoring is needed to understand the statuses and trends of wildlife communities in montane forests, such as those in Yushan National Park (YSNP), Taiwan. Integrating passive acoustic monitoring (PAM) with an automated sound identifier, a long-term biodiversity monitoring project containing six PAM stations, was launched in YSNP in January 2020 and is currently ongoing. SILIC, an automated wildlife sound identification model, was used to extract sounds and species information from the recordings collected. Animal vocal activity can reflect their breeding status, behaviour, population, movement and distribution, which may be affected by factors, such as habitat loss, climate change and human activity. This massive amount of wildlife vocalisation dataset can provide essential information for the National Park's headquarters on resource management and decision-making. It can also be valuable for those studying the effects of climate change on animal distribution and behaviour at a regional or global scale. New information: To our best knowledge, this is the first open-access dataset with species occurrence data extracted from sounds in soundscape recordings by artificial intelligence. We obtained seven bird species for the first release, with more bird species and other taxa, such as mammals and frogs, to be updated annually. Raw recordings containing over 1.7 million one-minute recordings collected between the years 2020 and 2021 were analysed and SILIC identified 6,243,820 vocalisations of seven bird species in 439,275 recordings. The automatic detection had a precision of 0.95 and the recall ranged from 0.48 to 0.80. In terms of the balance between precision and recall, we prioritised increasing precision over recall in order to minimise false positive detections. In this dataset, we summarised the count of vocalisations detected per sound class per recording which resulted in 802,670 occurrence records. Unlike data from traditional human observation methods, the number of observations in the Darwin Core "organismQuantity" column refers to the number of vocalisations detected for a specific bird species and cannot be directly linked to the number of individuals.We expect our dataset will be able to help fill the data gaps of fine-scale avian temporal activity patterns in montane forests and contribute to studies concerning the impacts of climate change on montane forest ecosystems on regional or global scales.

Various criteria in the evaluation of biomedical named entity recognition.

BACKGROUND: Text mining in the biomedical domain is receiving increasing attention. A key component of this process is named entity recognition (NER). Generally speaking, two annotated corpora, GENIA and GENETAG, are most frequently used for training and testing biomedical named entity recognition (Bio-NER) systems. JNLPBA and BioCreAtIvE are two major Bio-NER tasks using these corpora. Both tasks take different approaches to corpus annotation and use different matching criteria to evaluate system performance. This paper details these differences and describes alternative criteria. We then examine the impact of different criteria and annotation schemes on system performance by retesting systems participated in the above two tasks. RESULTS: To analyze the difference between JNLPBA's and BioCreAtIvE's evaluation, we conduct Experiment 1 to evaluate the top four JNLPBA systems using BioCreAtIvE's classification scheme. We then compare them with the top four BioCreAtIvE systems. Among them, three systems participated in both tasks, and each has an F-score lower on JNLPBA than on BioCreAtIvE. In Experiment 2, we apply hypothesis testing and correlation coefficient to find alternatives to BioCreAtIvE's evaluation scheme. It shows that right-match and left-match criteria have no significant difference with BioCreAtIvE. In Experiment 3, we propose a customized relaxed-match criterion that uses right match and merges JNLPBA's five NE classes into two, which achieves an F-score of 81.5%. In Experiment 4, we evaluate a range of five matching criteria from loose to strict on the top JNLPBA system and examine the percentage of false negatives. Our experiment gives the relative change in precision, recall and F-score as matching criteria are relaxed. CONCLUSION: In many applications, biomedical NEs could have several acceptable tags, which might just differ in their left or right boundaries. However, most corpora annotate only one of them. In our experiment, we found that right match and left match can be appropriate alternatives to JNLPBA and BioCreAtIvE's matching criteria. In addition, our relaxed-match criterion demonstrates that users can define their own relaxed criteria that correspond more realistically to their application requirements.