Publication Abstract

Introducing greater ecological realism to fish habitat models

Introducing greater ecological realism to fish habitat models

P.D. Eastwood and G.J. Meaden

Geographical Information Systems (GIS) are now commonly used as a means to construct and visualise models of fish habitats. The process of developing spatial habitat models typically consists of two stages. First, an empirical model of the relationship between the species and its environment is constructed. Second, map layers of the environmental variables, constructed within a GIS environment, are re-coded using the parameters of the empirical model, thus allowing the modelled relationship to be visualised in geographical space. The accuracy of GIS-based fish habitat models is largely dependent on two factors: the quality of the input data and the method used to construct the empirical model. Whilst the former cannot usually be controlled by the model developer, a number of different techniques are available to construct empirical models of the relationship between a fish species and its environment. Choosing an appropriate technique therefore requires careful consideration. This paper argues for the use of quantile regression when constructing fish habitat models. Quantile regression estimates of the upper bounds of species-response distributions offer an ecologically more realistic solution to habitat modelling by providing estimates of the limiting effects of the environment on the distribution and response of a species. To demonstrate this, regression quantiles and GIS procedures were used to model the spatial distribution of sole (Solea solea L.) habitats in the Dover Strait, the passage of water that connects the eastern English Channel to the southern North Sea.

Reference:

P.D. Eastwood, G.J. Meaden. 2005. Introducing greater ecological realism to fish habitat models.

P.D. Eastwood* and G.J. Meaden

p 181-198 in Proceedings of the Second International Symposium on GIS/Spatial Analyses in Fishery and Aquatic Sciences, 2-6 September 2002, Brighton, UK.