Description: Based on at-sea surveys from Oct-Mar 2006-2013, Inshore species presence/absence was rasterised and modelled with environmental variables; bathymetric depth, seabed slope, sea surface temperature, wind speed, significant wave height, amplitude of tidal surface height, chlorophyll and suspended matter concentrations, minimum distance to mainland, minimum distance to land and minimum distance to colony for breeding species. Ensemble predictions were calculated using a weighted average from four statistical models; generalised linear models, generalised additive models, boost regresssion trees and maximum entropy. Weights were assigned to each model based on its predictive performance. Spatial predictions of species distributions were only considered if they passed a thorough expert review process.