Open AccessCan we generate robust species distribution models at the scale of the Southern Ocean?
Author(s) -
FabriRuiz Salomé,
Danis Bruno,
David Bruno,
Saucède Thomas
Publication year - 2019
Publication title -
diversity and distributions
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.918
H-Index - 118
eISSN - 1472-4642
pISSN - 1366-9516
DOI - 10.1111/ddi.12835
Subject(s) - biological dispersal , environmental niche modelling , species distribution , ecological niche , sampling (signal processing) , niche , robustness (evolution) , scale (ratio) , biogeography , abiotic component , macroecology , ecology , environmental science , spatial ecology , spatial distribution , physical geography , geography , habitat , computer science , remote sensing , cartography , biology , population , biochemistry , demography , filter (signal processing) , sociology , gene , computer vision
Aim Species distribution modelling ( SDM ) represents a valuable alternative to predict species distribution over vast and remote areas of the ocean. We tested whether reliable SDM s can be generated for benthic marine organisms at the scale of the Southern Ocean. We aimed at identifying the main large‐scale factors that determine the distribution of the selected species. The robustness of SDM s was tested with regards to sampling effort, species niche width and biogeography. Location Southern Ocean. Methods The impact of sampling effort was tested using two sets of data: one set with all presence‐only data available until 2005, and a second set using all data available until 2015 including recent records from campaigns carried out during the Census of Antarctic Marine Life ( CAML ) and the International Polar Year ( IPY ) period (2005–2010). The accuracy of SDM s was tested using a ground‐truthing approach by comparing recent presence/absence data collected during the CAML and IPY period to pre‐ CAML model predictions. Results Our results show the significance of the SDM approach and the role of abiotic factors as important drivers of species distribution at broad spatial scale. The addition of recent data to the models significantly improved the prediction of SDM and changed the respective contributions of environmental predictors. However, the intensity of change varied between models depending on sampling tools, species ecological niche width and biogeographic barriers to dispersal. Main conclusions We highlight the need for new data and the significance of the ground‐truthing approach to test the accuracy of SDM s. We show the importance of data collected through international initiatives, su ch as the CAML and IPY to the improvement of species distribution modelling at broad spatial scales. Finally, we discussed the relevance of SDM as a relevant marine conservation tool particularly in the context of climate change and the definition of Marine Protected Areas.