Contents lists available at ScienceDirect Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul Benthic diversity patterns and predictors: A study case with inferences for conservation Paolo Vassallo a , Chiara Paoli a , Stefano Aliani b , Silvia Cocito c , Carla Morri a , Carlo Nike Bianchi a, ⁎ a DiSTAV (Department of Earth, Environmental and Life Sciences), University of Genoa, Corso Europa 26, I-16132 Genova, Italy b ISMAR (Institute of Marine Sciences), CNR, Forte Santa Teresa, I-19036 Pozzuolo di Lerici, SP, Italy c ENEA (Italian Agency for New Technologies, Energy and Sustainable Economic Development), Marine Environment Research Centre, I-19100 La Spezia, Italy ARTICLEINFO Keywords: Macrobenthos Diversity Species richness Equitability Hot spots Marine protected areas Mediterranean Sea ABSTRACT Understanding which drivers cause diversity patterns is a key issue in conservation. Here we applied a spatially explicit model to predict marine benthic diversity patterns according to environmental factors in the NW Mediterranean Sea. While most conservation-oriented diversity studies consider species richness only and ne- glect equitability, we measured separately species richness, equitability, and ‘overall’ diversity (i.e., the Shannon-WienerH′function)onadatasetof890benthicspecies×209samples.Diversityvalueswerepredicted by means of Random Forest regression, on the basis of 10 factors: depth, distance from the coast, distance from the shelf break, latitude, sea-foor slope, sediment grain size, sediment sorting, distance from harbours and marinas, distance from rivers, and sampling gear. Predictions by Random Forests were accurate, the main predictors being latitude, sediment grain size, depth and distance from the coast. Based on predicted values, diversity hotspots were identifed as those localities where indices were in the 15% top segment of ranked values. Only a minority of the diversity hotspots was included within the boundaries of the protection institutes establishedintheregion.Marineprotectedareasareoftencreatedinsitesharbouringimportantcoastalhabitats, which risks neglecting the diversity hidden in the sedimentary seafoor. We suggest that marine protected areas should accommodate portions of sedimentary habitat within their boundaries to improve diversity conservation. 1. Introduction Diversity, here primarily intended as the variety of species within a biotic community, holds a paramount role for ecological stability and ecosystem functioning (Dufy, 2009; Tilman et al., 2014) and provides goods and services essential to human welfare, making human life both possible and worth living (Haines-Young and Potschin, 2010; Bennett et al., 2015). Its importance is nowadays fully recognized by scientists, administrators and laymen, resulting in initiatives to conserve diversity (Dietz and Adger, 2003; Turak et al., 2017). The early approaches to conservation concentrated on a restricted number of charismatic spe- cies (Boudouresque and Bianchi, 2013; Veríssimo et al., 2017), but in the last decades awareness has grown that whole communities must be protected, if diversity is to be maintained (Franklin, 1993; Yamaura et al., 2018). Diversity is afected by many drivers, either natural or anthro- pogenic (van der Linden et al., 2016). Understanding which drivers cause diversity patterns under diferent circumstances is both a major theme in ecology and a key issue in conservation science (Price and Schmitz, 2016; Ramos et al., 2018). However, most efort to date has been devoted to terrestrial ecosystems (Franklin, 2010), whereas pat- terns of marine diversity are less known (Roy and Witman, 2009; Peterson and Herkül, 2019). Diferences in diversity between land and sea are obvious and dramatic (Boudouresque et al., 2014; Miller and Wiens, 2017), but diversity role and problems are the same (Bianchi and Morri, 2000; Worm et al., 2003). The increasingly urgent need to prioritize regions for conservation requires the availability of diversity maps that cover large sea areas (Connolly, 2009). However, traditional sampling-point-wise feld work is not suitable for covering extensive areas in high detail (Huang et al., 2011). Spatial predictive modelling using diversity data from sampling points and environmental data layers covering the whole study area is a way to create diversity maps for large spatial extents based on diversity-environment relationships (Peterson and Herkül, 2019). In recent years, several studies have attempted to predict marine diversity using spatially explicit models. Pittman et al. (2009) used lidar-derived bathymetry to predict fsh and coral diversity at Puerto Rico and were capable of explaining a large proportion (65% and 64%, https://doi.org/10.1016/j.marpolbul.2019.110748 Received 18 March 2019; Received in revised form 8 September 2019; Accepted 18 November 2019 ⁎ Corresponding author. E-mail address: carlo.nike.bianchi@unige.it (C.N. Bianchi). Marine Pollution Bulletin xxx (xxxx) xxxx 0025-326X/ © 2019 Elsevier Ltd. All rights reserved. Please cite this article as: Paolo Vassallo, et al., Marine Pollution Bulletin, https://doi.org/10.1016/j.marpolbul.2019.110748