Combining N-mixture and occupancy analysis offers a more complete picture of carnivore habitat use in Northeastern Türkiye

PLoS ONE, Год журнала: 2025, Номер 20(5), С. e0320768 - e0320768

Опубликована: Май 14, 2025

Occupancy and N-mixture analyses have been successfully used to understand habitat use in various species. However, since these methods fundamentally answer different questions about wildlife distribution, the results from each modelling approach may provide insights into species’ use. In this study, we leveraged data a long-term camera trapping study northeastern Türkiye compare occupancy analyses, with main objective of understanding how can influence our knowledge Specifically, compared preferences for three carnivore species varying baseline abundances. Our evidence that sensitivity environmental anthropogenic factors. Whereas analysis provides relatively broad summary factors affect where or not be located on landscape which areas they more likely over certain time period, degree at individual sites, particular emphasis being able deduce small-scale changes across landscape. Furthermore, while detection probability an has formally as measure site intensity, N-Mixture models offer higher resolution quantity Therefore, two tend investigate spatial scales, when conjunction refined through repeat-survey sampling like trapping.

Язык: Английский

---

Zero‐shot animal behaviour classification with vision‐language foundation models

Methods in Ecology and Evolution, Год журнала: 2025, Номер unknown

Опубликована: Май 20, 2025

Abstract Understanding the behaviour of animals in their natural habitats is critical to ecology and conservation. Camera traps are a powerful tool collect such data with minimal disturbance. They however produce very large quantity images, which can make human‐based annotation cumbersome or even impossible. While automated species identification artificial intelligence has made impressive progress, automatic classification animal behaviours camera trap images remains developing field. Here, we explore potential foundation models, specifically vision‐language models (VLMs), perform this task without need first train model, would require some level annotation. Using two datasets, alpine African fauna, investigate zero‐shot capabilities different kinds recent VLMs predict estimate behaviour‐specific diel activity patterns three ungulate species. By comparing our predictions annotated by participatory science, results show that using these methods, it possible achieve F1‐score as high 86.39% closely align those derived from science (overlap indexes between 84% 90%). These findings demonstrate ecological research. Ecologists encouraged adopt new methods leverage full facilitate studies.

Язык: Английский

---

Brown bear denning habits in northeastern Türkiye

Global Ecology and Conservation, Год журнала: 2024, Номер 54, С. e03156 - e03156

Опубликована: Сен. 6, 2024

Язык: Английский

---

Zero-shot animal behavior classification with image-text foundation models

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Апрель 9, 2024

Abstract 1. Understanding the behavior of animals in their natural habitats is critical to ecology and conservation. Camera traps are a powerful tool collect such data with minimal disturbance. They however produce very large quantity images, which can make human-based annotation cumbersome or even impossible. While automated species identification artificial intelligence has made impressive progress, automatic classification animal behaviors camera trap images remains developing field. 2. Here, we explore potential foundation models, specifically Vision Language Models (VLMs), perform this task without need first train model, would require some level annotation. Using an original dataset alpine fauna annotated by participatory science, investigate zero-shot capabilities different kind recent VLMs predict estimate behavior-specific diel activity patterns three ungulate species. 3. Our results show that using these methods, it possible achieve accuracies over 91% closely align those derived from science (overlap indexes between 84% 90%). 4. These findings demonstrate models vision-language ecological research. Ecologists encouraged adopt new methods leverage full facilitate studies.

Язык: Английский

---

Wolves across borders

Wildlife Biology, Год журнала: 2024, Номер 2024(6)

Опубликована: Ноя. 1, 2024

Язык: Английский

---