KelpPredict: Kelp community structure and function; analysing interactions and developing models along multiple exposure gradients

The Laminaria hyperborea kelp forests and its associated biodiversity are under pressure by kelp trawling, grazing by sea urchins, changing climate (storms) and pollutions (eutrophication). These pressures are related to exposure gradients, e.g. sea urchi n grazing is found in the more sheltered parts of the kelp forests (Sivertsen 2006) while storm disturbances and kelp harvesting mainly takes place in more exposed locations. Both for understanding the kelp forest functions and for evaluations of the ecol ogical effects of the disturbances on these large underwater forests there is need for a model to predict relationships between physical factors and the kelp forest as habitat. We will analyse interactions between kelp (size and biomass), epiphyte (bioma ss, abundance and structure) and faunal (abundance and species dominance) properties and exposure (wave and current) interactions and develop predictive models for kelp forest structure and its associated species along exposure gradients. Data will be sa mpled and models developed for the West coast of Norway. Data on kelp structure and epiphytic and faunal communities will be integrated in ArcGIS and analysed in relations to current and wave exposure models. A set of candidate statistical models will be developed using Generalised Additive Models, and these will be compared using Akaike Information Criterion. Predictive models will be developed and validated using data collected in this study and data received from other studies.