Advances in regional-scale predictions of VMEs in the New Zealand region and their use in decision-support tools for spatial management planning.

Spatial management planning for vulnerable marine ecosystems (VME) requires detailed predictions of species presence across broad areas of un-sampled seafloor. We utilised two habitat suitability modelling techniques, boosted regression trees (BRT) and maximum entropy (MaxEnt), to create potential distribution maps for 11 VME indicator taxa in the New Zealand area and adjacent seas. New bathymetry data was combined with existing environmental, chemical and physical data to produce a set of 52 predictor variables for the seafloor. Nine of these variables were selected for use in the models based on low covariance and high explanatory power. Historical biological survey data was used to provide models with absence data (BRT) or target-group background data (MaxEnt). Model agreement was high, with each model predicting areas of suitable habitat both in the vicinity of known VME indicator taxa presence locations as well as across broad regions of un-sampled seafloor where environmental conditions were suitable. Model performance measures, including cross-validation testing of models against sets of spatially independent data, did not clearly indicate a preferred model type across all taxa modelled. Despite its value in spatial management planning, previous habitat suitability modeling efforts have rarely accounted for model precision. In this study we used a bootstrap re-sampling technique to produce precision maps to accompany each habitat suitability map. Because of the similar performance of BRT and MaxEnt in this study, the best approach to incorporating model results into planning, for example using decision-support tools, may be to average predictions from the two techniques and/or select the model with the best performance for each taxon.