Modelling and investigation of the role of anthropogenic noise on the collective behaviour of marine animals

Abstract

From flocks of birds to schools of fish, animals moving in a collective group has fascinated the world for centuries. The displays of interesting and dynamic behaviour has been captured numerically and analytically, providing key insights into the mechanisms behind different behaviours. Many marine animals move as a collective group to enhance their chance of survival, however their environment is constantly changing. With a vast increase in the amount of anthropogenic noise they are subjected to, along with the industrialisation of vast areas in our seas and oceans, the impact this may have on their survival has become an important question for industry and scientists alike.

In this thesis, we propose an individual based model describing a group of interacting particles, whose movements change in response to their neighbours and external factors, such as noise. The group is subject to three main social interactions: repulsion (where they will attempt to move away from any nearby neighbours), orientation (where they will align with neighbours), and attraction (where they will move towards neighbours). We conduct a numerical study to determine the influence each of these rules, as well as various noise sources, has on the aggregates that are formed in a homogenous domain. Further, we identify the existence of phase transitions between the different aggregate types, of both first and second order. These transitions occur in the same model type, and are obtained by varying the behavioural strengths. We then consider a heterogeneous domain containing an anthropogenic structure, and determine how its presence can disrupt the short term movements of various fish group types (with varying degrees of alignment). Further, we consider how the presence of such a structure, along with a noise source, can lead to a long term delay in their movements. In particular, we consider groups who are attempting to reach a spawning ground, and determine the decrease in the length of time they are able to spend in this region. Finally, we introduce a population dynamics model which describes four fish populations: herring and their spawn, cod (who prey upon the herring) and haddock (who prey upon the herring spawn). We study this model analytically to determine how the predator-prey interactions influence the long term dynamics of the system, with particular focus on the offspring population, since this will determine the success of spawning. We then combine this model with our original IBM and conduct a numerical study to determine how multiple factors can contribute to a decrease in the expected number of spawn generated in one spawning cycle; namely, the presence of an anthropogenic structure, a noise source, the need for movements, and predator-prey interactions. To this end, we capture some of the factors which may be contributing to decreases observed in fish stocks.

Date of Award	2021
Original language	English
Awarding Institution	University of Dundee
Sponsors	Engineering and Physical Sciences Research Council
Supervisor	Raluca Eftimie (Supervisor) & Terence Dawson (Supervisor)