Conditional Generative Adversarial Network for the simulation of seabird trajectories based on environmental data
Abstract
Seabirds are impacted worldwide by ongoing and emerging threats, such as habitat loss, unsustainable fishing and climate change. Within this context, models able to simulate and predict seabird behaviour depending on their environment would be valuable for decision making. In this study, we present the Spatial GAN (Generative Adversarial Network), able to simulate seabird foraging trajectories based on a spatial environmental condition. It is illustrated using GPS-tracking for Northern Gannets (Morus bassanus) from five different colonies. We demonstrate that the model simulates central-place foraging trajectories coherent with field data for each colony. The ability of the Spatial GAN to simulate trajectories with realistic global characteristics of length and duration consistently outperforms those of other existing models, including classic approaches such as HMM-informed Random Walks, and deep learning approaches such as VAE and the GAN without spatiality. The Spatial GAN also correctly simulates trajectories for birds from sites not included in the training dataset. While it may simulate erroneous trajectories on land when bathymetric configurations are new, the inclusion of only 10 to 20 GPS-trajectories from a new site greatly improves model simulations. As such, the architecture of the Spatial GAN we developed is a good starting point for a trajectory simulator using environmental maps as input data. The specificity of this approach compared to existing ones is its capacity to approximate the relationship between seabird movements and the environment structure by itself, without any explicit constraint.
Related
- Hi-cGAN: Prediction of Hi-C interaction matrices with conditional generative adversarial networks
- Generative Adversarial Networks for Omics Data: An Overview and Future Directions for Livestock Applications
- DHAFGan: A dense hybrid attention fusion generative adversarial network for infrared and visible image fusion
- ScaffoldForge: Bioinspired generative design of 3D-printed bone scaffolds via conditional diffusion for large segmental bone defect reconstruction
- The Youth-GEMs project: genetic and environmental interactions underlying trajectories of youth mental health
BibTeX
@article{benboudjema2026conditional,
title = {{Conditional Generative Adversarial Network for the simulation of seabird trajectories based on environmental data}},
author = {Sophie Benboudjema and Amédée Roy and Géraldine Nogaro and Philip Atkinson and Bethany Clark and Ronan Fablet and Roland Gauvain and David Grémillet and Jude Lane and Steve Votier and Sophie Lanco},
journal = {Ecological Informatics},
year = {2026},
month = oct,
doi = {10.1016/j.ecoinf.2026.103994},
url = {https://hal.science/hal-05770050v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Conditional Generative Adversarial Network for the simulation of seabird trajectories based on environmental data
AU - Sophie Benboudjema
AU - Amédée Roy
AU - Géraldine Nogaro
AU - Philip Atkinson
AU - Bethany Clark
AU - Ronan Fablet
AU - Roland Gauvain
AU - David Grémillet
AU - Jude Lane
AU - Steve Votier
AU - Sophie Lanco
PY - 2026
DA - 2026/10/02
JO - Ecological Informatics
DO - 10.1016/j.ecoinf.2026.103994
UR - https://hal.science/hal-05770050v1
AB - Seabirds are impacted worldwide by ongoing and emerging threats, such as habitat loss, unsustainable fishing and climate change. Within this context, models able to simulate and predict seabird behaviour depending on their environment would be valuable for decision making. In this study, we present the Spatial GAN (Generative Adversarial Network), able to simulate seabird foraging trajectories based on a spatial environmental condition. It is illustrated using GPS-tracking for Northern Gannets (Morus bassanus) from five different colonies. We demonstrate that the model simulates central-place foraging trajectories coherent with field data for each colony. The ability of the Spatial GAN to simulate trajectories with realistic global characteristics of length and duration consistently outperforms those of other existing models, including classic approaches such as HMM-informed Random Walks, and deep learning approaches such as VAE and the GAN without spatiality. The Spatial GAN also correctly simulates trajectories for birds from sites not included in the training dataset. While it may simulate erroneous trajectories on land when bathymetric configurations are new, the inclusion of only 10 to 20 GPS-trajectories from a new site greatly improves model simulations. As such, the architecture of the Spatial GAN we developed is a good starting point for a trajectory simulator using environmental maps as input data. The specificity of this approach compared to existing ones is its capacity to approximate the relationship between seabird movements and the environment structure by itself, without any explicit constraint.
SN - 1574-9541
ER -