Biology subjects

Chaine, A. S.

Publications and source records attributed to Chaine, A. S..

4 recordsLinked to original sources

bioRxiv2026

Towards a general Detector of terrestrial Arthropods in Natural backgrounds

Widespread arthropod declines pose risks to ecosystem functioning and agriculture. Assessing this decline or potential remediation implies the need for standardized and scalable population monitoring. Image-based methods, including camera traps and citizen science programs, are increasingly used, but the volume of data collected requires automated analysis. Robust arthropod detection is essential for individual counting or fine-grained classification, yet current datasets and algorithms do not address the vast morphological diversity across arthropod species and often overlook the variety of photographic contexts, such as differences in background, lighting, and image composition, in which arthropods are captured. To address this gap, we developed an arthropod detection dataset, covering all terrestrial families present in France with available validated images on the iNaturalist platform (749 families). To achieve this, we employed an iterative workflow in which a YOLOv11 model pre-annotated images -- using one representative species per family-- followed by manual correction and model retraining. Repeating this process progressively reduced annotation effort and improved model accuracy. The final outcome consists of a publicly available curated detection dataset and a robust arthropod detector for natural background scenes. The detector achieves an F1-score of 0.91, demonstrating strong performance despite substantial interspecific morphological variation and heterogeneity in photographic contexts. We further demonstrated the taxonomical universality of the model showing high F1-score and IoU averaged at the class (0.79, 0.85) and order level (0.82, 0.86) and also a good detection generalizability (F1-score>0.90, IoU>0.83) on species, genera and families never encountered by the model during training. Finally, we show how this model can be improved to generalize to new datasets using data augmentation, complementary training data or fine-tuning and increase detection of small objects. In particular, we report performance of the improved models on three use cases largely used in non lethal insect monitoring: (i) diurnal pollinator monitoring through citizen science or (ii) flower and nocturnal insects monitoring through smartphone time-lapse of a UV-illuminated white panel. These results mark an important step toward automated analysis of arthropod images in natural contexts, from both large-scale automated monitoring approaches or from citizen science monitoring programs.

ecology↗

bioRxiv2024

Tit wit: environmental and genetic drivers of cognitive variation along an urbanization gradient

Cognitive abilities can promote acclimation and adaptation to life in cities. However, the genetic versus environmental drivers of cognition have rarely been studied in the wild and there exists a major gap concerning the role of cognition in adaptation to novel urban contexts. To address this, we evaluate cognitive variation in wild great tits (Parus major; N = 393) along an urban gradient, and decipher the genetic basis of this variation using a combination of a common garden experiment, quantitative genetic analyses, and genome-wide association studies. Specifically, we measure inhibitory control abilities which affect how animals respond to novel resources and challenges. We find that wild urban and forest tits do not clearly differ in inhibitory control performance (number of errors or the latency to escape) during a motor detour task; a result that was consistent in birds from urban and forest origins reared in a common garden (N = 73). Cognitive performance was repeatable (R = 0.35 - 0.38) and showed low to moderate heritability in the wild (h2= 0.16 - 0.28 using social and genomic pedigrees). We identified five SNPs that were significantly associated with the number of errors during the task, explaining 21% of the cognitive variation. These SNPs are linked to genes related to serotonergic and dopaminergic systems that are known to play important roles in cognition. Altogether, our study finds limited evidence that inhibitory control abilities have evolved under novel urban contexts, yet reveals a genetic basis of this cognitive trait in great tits.

animal behavior and cognition↗

bioRxiv2024

Density-dependent network structuring within and across wild animal systems

High population density should drive individuals to more frequently share space and interact, producing better-connected spatial and social networks [1-4]. Although this theory is fundamental to our understanding of disease dynamics [2,5-8], it remains unconfirmed how local density generally drives individuals positions within their networks, which reduces our ability to understand and predict density-dependent processes [4,9,10]. Here we provide the first general evidence that density drives greater network connectedness at fine spatiotemporal scales, at the scale of individuals within wild animal populations. We analysed 36 datasets of simultaneous spatial and social behaviour in >58,000 individual animals, spanning 30 species of fish, reptiles, birds, mammals, and insects. 80% of systems exhibited strong positive relationships between local density and network centrality. However, >80% of relationships were nonlinear and 75% became shallower at higher values, signifying that demographic and behavioural processes counteract densitys effects, thereby producing saturating trends [11-15]. Densitys effect was much stronger and less saturating for spatial than social networks, such that individuals become disproportionately spatially connected rather than socially at higher densities. Consequently, ecological processes that depend on spatial connections (e.g. indirect pathogen transmission, resource competition, and territory formation) are likely more density-dependent than those involving social interactions (e.g. direct pathogen transmission, aggression, and social learning). These findings reveal fundamental ecological rules governing societal structuring, with widespread implications. Identifying scaling rules based on processes that generalise across systems, such as these patterns of density dependence, might provide the ability to predict network structures in novel systems.

ecology↗

bioRxiv2021

Social migratory connectivity: do birds that socialize in winter breed together?

Researching the complete life cycles of migratory animals is essential for understanding conservation and population dynamics. Many studies focus on the breeding season, but surviving winter is equally important. Living in groups during winter can play a vital role as social connections within groups can provide many benefits such as protection from predators and increased access to resources. However, it is often unknown how social connections change across seasons in migratory animals. We focus on social connections in a migratory bird and ask whether social connections in winter continue during breeding. Golden-crowned sparrows have distinct, stable winter communities which include both site and group fidelity across years: birds almost always rejoin the same social community each year. If these birds have social connectivity across migration, we would expect individuals that associate in winter would also associate together on their breeding grounds. Our small-scale GPS tagging study combined with intensive social behavior data revealed that sparrows in the same tightly-knit winter community migrated to highly disparate locations during summer, showing that social connections in winter do not continue in summer. This suggests that golden-crowned sparrows have entirely separate social structures across seasons and that long-term social memories allow them to reform stable groups each winter.

animal behavior and cognition↗