Generative Adversarial Networks (GAN) for the simulation of central-place foraging trajectories

1. Miniature electronic device such as GPS have enabled ecologists to document relatively large amount of animal trajectories. Modeling such trajectories may attempt (1) to explain mechanisms underlying observed behaviors and (2) to elucidate ecological processes at the population scale by simulating multiple trajectories. Existing approaches to animal movement modeling mainly addressed the first objective and they are yet soon limited when used for simulation. Individual-based models based on ad-hoc formulation and empirical parametrization lack of generability, while state-space models and stochastic differential equations models, based on rigorous statistical inference, consist in 1st order Markovian models calibrated at the local scale which can lead to overly simplistic description of trajectories. 2. We introduce a 'state-of-the-art' tool from artificial intelligence - Generative Adversarial Networks (GAN) - for the simulation of animal trajectories. GAN consist in a pair of deep neural networks that aim at capturing the data distribution of some experimental dataset, and that enable the generation of new instances of data that share statistical similarity. In this study, we aim on one hand to identify relevant deep networks architecture for simulating central-place foraging trajectories and on the second hand to evaluate GAN benefits over classical methods such as state-switching Hidden Markov Models (HMM). 3. We demonstrate the outstanding ability of GAN to simulate 'realistic' seabirds foraging trajectories. In particular, we show that deep convolutional networks are more efficient than LSTM networks and that GAN-derived synthetic trajectories reproduce better the Fourier spectral density of observed trajectories than those simulated using HMM. Therefore, unlike HMM, GAN capture the variability of large-scale descriptive statistics such as foraging trips distance, duration and tortuosity. 4. GAN offer a relevant alternative to existing approaches to modeling animal movement since it is calibrated to reproduce multiple scales at the same time, thus freeing ecologists from the assumption of first-order markovianity. GAN also provide an ultra-flexible and robust framework that could further take environmental conditions, social interactions or even bio-energetics model into account and tackle a wide range of key challenges in movement ecology.

Hunter-gatherer foraging networks promote information transmission

Central-place foraging, where foragers return to a central location (or home-base), is a key feature of hunter-gatherer social organization. Although why or when our ancestors started returning to “home-bases” remains unclear, it likely had significant implications for many aspects of hominin evolution. For example, central-place foraging, by changing hunter-gatherers’ use of space and mobility, could have altered social networks and increased opportunities for information exchange. We evaluated whether central-place foraging patterns facilitate information transmission and considered the potential roles of environmental conditions and mobility strategies. We built an agent-based central-place foraging model where agents move according to a simple optimal foraging rule, and can encounter other agents as they move across the environment. They either forage close to their home-bases within a given radius or move their home-bases to new areas. We analyzed the interaction networks arising across different environments and mobility strategies. We found that, at intermediate levels of environmental heterogeneity and mobility, central-place foraging increased global and local network efficiencies as well as the rate of contagion-based information transmission (simple and complex). Our findings suggest that the combination of foraging and movement strategies, as well as the underlying environmental conditions that characterized early human societies, may have been a crucial precursor in our species’ unique capacity to innovate, accumulate and rely on complex culture.

Tree Swallow selection for wetlands in agricultural landscapes predicted by central-place foraging theory

Millions of wetland basins, embedded in croplands and grasslands, are biodiversity hotspots in North America’s Prairie Pothole Region, but prairie wetlands continue to be degraded and drained, primarily for agricultural activities. Aerial insectivorous swallows are known to forage over water, but it is unclear whether swallows exhibit greater selection for wetlands relative to other habitats in croplands and grasslands. Central-place foraging theory suggests that habitat selectivity should increase with traveling distance from a central place, such that foragers compensate for traveling costs by selecting more profitable foraging habitat. Using global positioning system (GPS) tags, we evaluated habitat selection by female Tree Swallows (Tachycineta bicolor) at 4 sites containing wetlands and where terrestrial land cover was dominated by grasslands (grass, herbaceous cover) and/or cultivated cropland. We also used sweep-net transects to assess the abundance and biomass of flying insects in different habitats available to swallows (wetland pond margins, grassy field margins, and representative uplands). As expected for a central-place forager, GPS-tagged swallows selected more for wetland ponds (disproportionate to availability), and appeared to increasingly select for wetlands with increasing distance from their nests. On cropland-dominated sites, insect abundance and biomass tended to be higher in pond margins or grassy field margins compared to cropped uplands, while abundance and biomass were more uniform among sampled habitats at sites dominated by grass and herbaceous cover. Swallow habitat selection was not clearly explained by the distribution of sampled insects among habitats; however, traditional terrestrial sampling methods may not adequately reflect prey distribution and availability to aerially foraging swallows. Overall, our results underscore the importance of protecting and enhancing prairie wetlands and other non-crop habitats in agricultural landscapes, given their disproportionate use and capacity to support breeding swallow and insect populations.