A Copula-based Habitat Preference Index in Fish Spatial Population Modelling

The habitat preference and spatial distribution of fish populations may depend on a number of environmental factors. For example, the habitat preference for a given species may depend on sea surface height, sea surface temperature and net primary productivity, among others. These variables are however not independent. We investigate an approach for determining a habitat preference index for spatial modelling of a fish population in a non-homogeneous environment. We use copulas to combine univariate habit preference variable distributions into a multivariate habit index. Such an index may be used to estimate a non-homogeneous spatial distribution of a population, with concentrations of fish in ‘favourable’ areas and avoidance of other less ‘favourable’ areas. The efficiency of copulas lies in the ability to study marginal behaviors separately from global dependence.