Welcome! This project recently migrated onto Zooniverse’s new architecture. For details, see here.
Results
Scientific publications
Dog invasions in protected areas: A case study using camera trapping, citizen science and artificial intelligence
Abstract
Domestic dogs, Canis familiaris, wandering into natural habitats poses a grave threat to wildlife, increasing predation pressure and disease risk and disrupting the ecological balance within ecosystems. This study examines the presence of dogs in a European Protected Area (PA), Doñana National Park (SW Spain), where their access is strictly restricted, and explores how dog presence relates to potential access points. We utilised classifications provided by citizen science and artificial intelligence, subsequently validated by experts, to detect dogs within 5200,000 photos taken by 60 camera traps randomly deployed across the PA from October 2020 to January 2024. We discovered 33 dogs, primarily in groups of 2–5 individuals, recorded across 31 detection events at 22 camera locations. Dogs were detected ranging from 10 to 42 km2 (Minimum Convex Polygon) within the PA. The detection probability of dogs increased by 0.22 log odds per kilometre closer to a village (corresponding to an increase from 0.5 to approximately 0.55) bordering the PA and exceeded 0.9 near it. Our data revealed three types of dogs wandering within the PA: dogs accompanying poachers, free-roaming dogs living in nearby human settlements, and stray dogs, most likely relying on the PA resources. Urgent actions are needed in Doñana as dogs pose severe threats to endangered species like the Iberian lynx Lynx pardinus (six adult female lynx documented killed by dogs). We recommend raising awareness among local authorities of free-roaming dogs, particularly in settlements close to PAs, where their presence should be banned. Regularly monitoring dog presence within PAs is crucial to prevent invasions and their associated impacts. Our findings underscore the importance of using camera traps and integrating artificial intelligence with citizen science to monitor invasive species effectively.
DOI: https://doi.org/10.1016/j.gecco.2024.e03109.
Essential tools but overlooked bias: artificial intelligence and citizen science classification affect camera trap data
Abstract
Camera trapping generates vast image datasets requiring classification before downstream ecological inference, yet the influence of classification errors on subsequent analyses is often overlooked. Classification performance can vary widely depending on the classification method (e.g. citizen science vs. AI), species, illumination conditions (diurnal vs. nocturnal), and other contextual factors. We compared a citizen-science classification method to two AI classifiers (EfficientNet and DeepFaune) using an expert-labelled hold-out of 51,588 images across seven classes (“empty,” “human,” “cervid,” “wild boar,” “red fox,” “leporid,” and “European badger”) captured day and night. For each class and method, we quantified precision (accuracy of positive predictions) and recall (ability to detect all positive instances), then fitted single-season occupancy models to the classified data and compared estimates against expert-derived benchmarks. Finally, we conducted a large-scale simulation to investigate how true occupancy, detection probability, and classification performance (recall and precision) collectively influence the accuracy (root mean square error – RMSE) of occupancy estimates. Citizen scientists exhibited consistently high precision but more variable recall. The AI classifiers outperformed the citizen-science method in recall for several species, including wild boar, leporid, and European badger. Both approaches performed worse on nocturnal images and showed reduced precision for night-time “empty” images. Bias in occupancy estimates differed across species, methods, and space – the AI-based estimates were generally more biased, with both the magnitude and direction of bias varying spatially, especially for rarer species such as leporids. In our simulation study, precision emerged as the strongest predictor of occupancy model accuracy, with lower precision substantially increasing RMSE. Lower occupancy rates increased RMSE, and precision regulated the impact of detection probability: at low precision, higher detection probability worsened errors; at high precision, RMSE remained low – or even decreased – as detection probability rose. Although AI classifiers offer unmatched processing speed, our findings show that citizen science can reduce classification errors. Moreover, low precision and poor recall, especially for rare or nocturnal species, can substantially bias occupancy models. Based on our results, we recommend improving precision and accounting for classification quality and uncertainty to ensure robust inference from camera trap data.