Analysing animal movement is essential for understanding processes such as the dynamics and spatial distribution of populations and has strong implications for the design and management of natural reserves. In tropical coastal ecosystems the movement patterns of fishes are particularly important as many fish species undertake diel migrations to utilize resources from different habitats. These small-scale movement patterns play an important role in the energy transfer between habitats and connect coastal systems like seagrass beds and coral reefs. To identify essential habitat properties for key species and anticipate their behavioural responses to changing environmental conditions is therefore critical to successful conservation.
Despite the relevance of fish movements, guidance mechanisms for fish moving between habitats are not well understood and studies investigating possible environmental drivers of movement patterns are still scarce. In this study we thus aim to broaden our knowledge of potential causal mechanisms and spatiotemporal patterns of reef fish movements and space use. To this end, we first investigate a tropical reef system to identify a suitable model organism (the diurnal parrotfish Chlorurus sordidus) and its susceptibility to natural cycles. We then simulate its movement decision-making by linking it with two main functional landscape features (food availability and predation risk) in a novel approach combining individual-based modelling (IBM) with potential field methods. In our IBM a model fish can evaluate (via motivation-specific weighing factors) the perceived (via a perception range) risk and benefits of the surrounding landscape features and adapt its movement direction and speed accordingly taking into account its internal state (energy budget). By explicitly integrating a fish's perceptional abilities into the movement decision-making process our model allows us to evaluate how a fish may perceive (energetic) costs and benefits of habitat features and how much impact relatively sophisticated behavioural rules have on predictions of overall population dynamics and viability in fragmented landscapes.
Our model may further assist in determining how specific landscapes are (ecologically) connected. This organism-based emergent property of the landscape, also referred to as (landscape) connectivity, is often considered to have strong implication for MPA site selection.
Results of our model simulations show that individual movement patterns and the resulting spatial distributions of the population are more irregularly distributed among coral reef patches the more the coral reef habitat becomes fragmented and reduced. The spatial configuration of the underlying seascape thus influences the spatial exploitation of microhabitats, which may have far-reaching consequences on the ecosystem level depending on the functional role of the species under consideration.
Based on our findings and its ability to provide detailed population dynamics over long time periods (years) and at a high spatial (1m2) and temporal resolution (up to 1 s) we believe our model can provide valuable insights into the spatio-temporal variability of local herbivore fish populations. Moreover, the integration of potential field methods into IBMs seems a promising strategy to represent the complexity of dynamic decision-making of animals in applied models. Also, by being easily adaptable to different species and habitat settings as well as extendable with additional modules the model can readily be adjusted to specific questions and study systems. It may therefore provide a basic framework to process and summarize, visualize and analyse fish movement data and predict potential consequences of changing habitat structures. Eventually, the gained information may help to design effective reserves and efficiently manage and protect reef fish populations.
How do fragmented seascapes influence movement behaviour of reef fishes?