Biology subjects

Tuia, D.

Publications and source records attributed to Tuia, D..

3 recordsLinked to original sources

bioRxiv2025

Multi-scale neural networks enhance species distribution modelling across predictors and taxonomic groups

Ecological processes and patterns are scale-dependent, as reflected in how relationships between species occurrence and the environment vary across scales. In species distribution models (SDMs), spatial scale can strongly affect model predictions and performance. Despite this, the choice of scale is often overlooked. The development of SDMs using deep learning models has enabled spatially structured data, such as patches of environmental raster data, to be considered as model inputs, making the question of scale even more crucial. Here, we evaluate convolutional neural network-based multi-species SDMs considering different scales on a dataset from Switzerland, including more than 8 million observations for 2390 plant and 1006 animal species comprising amphibians, butterflies, beetles, and mammals. We investigate how scale affects model performance and compare single- and multi-scale models. Our results reveal stronger scale effects for remote sensing indices than for bioclimatic and edaphic variables. In single-scale models, we find that plant species perform better with smaller spatial scales, while the opposite is true for animal species. Multi-scale models consistently improve predictive performance and reduce sensitivity to arbitrary scale selection across all modalities and taxonomic groups. The effect of scale is nonetheless smaller than the increase in performance achieved by considering multiple predictor groups. To improve the interpretability of such complex models, we use attribution methods to compute the relative contribution of different scales and predictor groups. Our study demonstrates the importance of accounting for scale and the potential of deep learning methods to integrate ecological complexity with spatial data across scales. HighlightsO_LIDeep learning SDMs can use spatial environmental data as model predictors C_LIO_LIThe spatial extent of the predictors affects model performance C_LIO_LIMulti-scale models improve performance across predictors and taxonomic groups C_LIO_LIMulti-scale modelling reduces sensitivity to arbitrary scale selection C_LIO_LIExplainable AI methods provide contribution scores for scales and predictor groups C_LI

ecology↗

bioRxiv2025

Migrating in a warming world: A deep learning approach to predict pan-American seasonal shifts in the monarch butterfly niche

Climate change is driving biodiversity loss, disrupting ecosystem functioning, and altering species distributions. Migratory species, whose range varies across seasons depending on specific climatic conditions, are particularly sensitive to environmental changes and serve as indicators of ecosystem health. However, current species distribution models often fail to capture the temporal dynamics critical for migratory species, limiting their ability to provide accurate future range estimations. In this study, we address this gap by developing a time-aware deep learning species distribution model for the monarch butterfly (Danaus plexippus), an iconic species for biodiversity conservation. Using monarch occurrence records across the Americas gathered from scientific and citizen science sources, we embed the effect of monthly climatic variables in a sequential framework. We compare the performance of our seasonal model to conventional time-static baselines, showing not only better performance in the present, where the models have been trained and validated, but also in the past. Our findings show that climatic factors such as humidity, temperature, precipitation and cloud coverage strongly influence the ecological niche of the monarch butterfly, with notable seasonal and spatial variability. Applying our model under climate change scenarios, we predict a northwestward shift in the monarch range by the end of the XXI century, with expansion in Canada and significant contraction in California and Central America, key sites for overwintering that also host resident monarch populations. These changes could severely impact the species migratory cycle and population stability. Using Shapley values, an explainable AI technique, we identify the decrease in precipitation and humidity as important environmental drivers responsible for the contraction of overwintering sites. By focusing on a species of high ecological relevance through a time-aware modeling approach, this work brings novel insights for the conservation of migratory species in the face of the challenges posed by climate change.

ecology↗

bioRxiv2023

WildCLIP: Scene and animal attribute retrieval from camera trap data with domain-adapted vision-language models

Wildlife observation with camera traps has great potential for ethology and ecology, as it gathers data non-invasively in an automated way. However, camera traps produce large amounts of uncurated data, which is time-consuming to annotate. Existing methods to label these data automatically commonly use a fixed pre-defined set of distinctive classes and require many labeled examples per class to be trained. Moreover, the attributes of interest are sometimes rare and difficult to find in large data collections. Large pretrained vision-language models, such as Contrastive Language Image Pretraining (CLIP), offer great promises to facilitate the annotation process of camera-trap data. Images can be described with greater detail, the set of classes is not fixed and can be extensible on demand and pretrained models can help to retrieve rare samples. In this work, we explore the potential of CLIP to retrieve images according to environmental and ecological attributes. We create WildCLIP by fine-tuning CLIP on wildlife camera-trap images and to further increase its flexibility, we add an adapter module to better expand to novel attributes in a few-shot manner. We quantify WildCLIPs performance and show that it can retrieve novel attributes in the Snapshot Serengeti dataset. Our findings outline new opportunities to facilitate annotation processes with complex and multi-attribute captions. The code will be made available at https://github.com/amathislab/wildclip.

ecology↗