TY - JOUR T1 - Modelling spatial distribution of epibenthic communities in the Gulf of St. Lawrence (Canada) A1 - Moritz,Charlotte A1 - Levesque,Melanie A1 - Gravel,Dominique A1 - Vaz,Sandrine A1 - Archambault,Diane A1 - Archambault,Philippe AD - Univ Quebec, Inst Sci Mer, Rimouski, PQ G5L 3A1, Canada. AD - Univ Quebec, Rimouski, PQ G5L 3A1, Canada. AD - French Res Inst Exploitat Sea, IFREMER, Channel & North Sea Fisheries Unit, F-62321 Boulogne Sur Mer, France. AD - Fisheries & Oceans Canada, Inst Maurice Lamontagne, Mont Joli, PQ G5H 3Z4, Canada. UR - https://doi.org/10.1016/j.seares.2012.10.009 DO - 10.1016/j.seares.2012.10.009 KW - Biodiversity KW - Epibenthic Communities KW - Estuary and Northern Gulf of St. Lawrence KW - Generalized Linear Model KW - Community Distribution Model KW - Redundancy Analysis N2 - Correlative habitat models using relationships between marine organisms and their surrounding environment can be used to predict species distribution, and the results can assist management of human activities sharing the marine space (e.g. fisheries, MPAs, tourism). Here, epi-benthic megafauna was sampled at 755 stations in the Lower Estuary and Northern Gulf of St. Lawrence (EGSL) each summer between 2006 and 2009. We combined various types of multivariate analyses to 1) describe the structure and spatial distribution of benthic communities, 2) analyse the relationship between these communities and environmental parameters, and subsequently 3) build a community distribution model to predict the spatial distribution of the communities, creating community distribution maps covering the entire area to be used for marine management and conservation. We identified distinct benthic communities in the study area that closely correlate with the 200 m depth contour and with major environmental variables. A redundancy analysis revealed that communities were associated with depth, oxygen saturation, temperature, bottom current, seabed uniformity, distance to coast and type of sediment. Together these environmental descriptors explained 38% of the variation in megafaunal community composition. The environmental variables were used to build a community distribution model using generalized linear models to predict high and low suitability zones of each community in the EGSL. (C) 2012 Elsevier B.V. All rights reserved. Y1 - 2013/04 PB - Elsevier Science Bv JF - Journal Of Sea Research SN - 1385-1101 VL - 78 SP - 75 EP - 84 ID - 24972 ER -