Quantifying interaction in moving animals

We have limited understanding of how a disease might spread in an animal population or how a population functions depending on the efficiency of information transmission between individuals. To address this gap a theory on the dynamics of information transmission, or transmission theory in brief, will be developed. By accounting for the movement of the animals and the transmission characteristics of the information being exchange, it is possible to predict how the information propagates through the population, being something harmful like an infection, or useful like knowledge about food resources. To do that a null model of information transmission for pairs of animals will be constructed and used to interpret ant movement data in a laboratory environment as well as bat and zebra movement data in the wild. In the ant case, transmission theory will be employed to explain how the frequency of interaction between individual ants regulate cyclic activity in the nest. In the bat case, it will be used to test whether roosts act as central location where information is transmitted to other individuals. In the zebra case, transmission theory will serve the purpose of predicting the spatial spread of anthrax infection in the population.

Biological organisms interact in space and time by signalling and responding to each other's presence. Understanding the spatio-temporal mechanisms with which interactions occur is key to explain group level processes: from the spread of an infection in an animal population or the transmission of alarm signals in a human crowd to the collective migration of epithelial cells healing a wound. Despite the ubiquity of interaction processes between moving entities, our ability to quantify them is limited. The development of a general theory of organism interaction has eluded most efforts because of (i) semantic issues (what is an `interaction'?), (ii) analytic intractability (lack of `null' models), and (iii) estimation difficulties (parameters extraction from individual trajectories). This proposal tackles all of the above three challenges, creates a publicly available software to estimate interactions from movement tracks of individuals, and tests its applicability in an animal behavioural context with movement data on three taxa: Egyptian fruit bats (Rousettus aegyptiacus) from the Hula Valley in Israel; zebras (Equus quagga) from Etosha National Park in Namibia; and Leptothorax ants in a laboratory nest.

(1) The present proposal aims to create a novel theory with which to interpret animal interaction data. From the empirical side, the results of the modelling study will be of direct use in animal ecology and conservation. In the long term the outcome of this study will provide new practical methodology to infer how frequently individuals in the same or different populations interact. It will thus be of use to help maintain biodiversity, as well as to develop models for disease propagation in spatially heterogeneous animal communities. Applications of the new analytical techniques will also help in quantify the ecological consequences of efforts to cull wildlife to control epidemic diseases.
(2) Applied mathematicians will benefit from advances in modelling multi-agent interactions and further from a rigorous methodology to construct interaction network from movement data of multiple individuals. In the long run by devising general principles that link information transmission capability between individuals and their spatial positioning it will be possible to devise optimal strategies to control a swarm of robots with specific searching or navigation tasks.
(3) Dissemination of the findings will be accomplished by either of the project partners presenting at international conference but also through the planned international workshop From social and epidemic networks to the dynamics of information transmission, to be organised in Bristol, at the end of the project. The workshop will be a combination of community steering and dissemination of the novel techniques.
(4) The general public will be kept informed of the outcomes of the work in several ways. Key papers will be published in publicly available journals (such as PLoS) as well as by maintaining a webpage of the activities of the project to ensure that the results are widely disseminated for the best public impact.
(5) A general theory that quantifies interaction between biological organisms is of relevance to applications beyond the animal behavioural context. By adapting the theory and data analysis to account for interactions between other biological organisms, the methodologies that will be developed here could spun a series of novel developments in other biological fields, e.g. in medical relevant disciplines.