Predicting large-scale habitat suitability for cetaceans off Namibia using MinxEnt

Knowledge of the occurrence and distribution of cetaceans is particularly important for conservation and management, but is still limited within Namibian waters. We collated 3211 cetacean records from the Namibian Exclusive Economic Zone (EEZ) for the period 2008 to 2016 and applied the principle of minimum cross entropy (MinxEnt) to predict habitat suitability. MinxEnt is a generalised form of maximum entropy modelling that allows incorporation of additional information such as sampling bias. The habitat suitability of 9 cetacean species or species groups (5 odontocete species, 2 mysticete species and 2 taxonomic groups: pilot whales Globicephalus spp. and balaenopterids Balaenopteridae spp.) were predicted per season, in relation to environmental variables likely to drive cetacean presence: sea surface temperature, chlorophyll a concentration, water depth or distance to shore, seabed slope and habitat complexity. The environmental variable which most frequently influenced habitat suitability was depth, which was the main environmental driver for bottlenose dolphin Tursiops truncatus, humpback Megaptera novaeangliae and southern right whales Eubalaena australis. Further, Heaviside's dolphin Cephalorhynchus heavisidii habitat was best predicted by distance to shore in all seasons, while common dolphin Delphinus delphis and the balaenopterid group habitats were best predicted by habitat complexity, and sperm whale Physeter macrocephalus habitats by chlorophyll a concentration. We identify distinct spatial patterns in habitat suitability for different species and provide baseline maps which can be used by managers of wildlife resources.