Global patterns in marine biodiversity In: World Ocean Assessment 1
P Snelgrove, EV Berghe, P Miloslavich, P Archambault, N Bailly, A Brandt, A Bucklin, M Clark, F Dahdouh-Guebas, P Halpin, R Hopcroft, K Kaschner, B Lascelles, LA Levin, S Menden-Duer, A Metaxas, D Obura, RR Reeves, T Rynearson, K Stocks, M Tarzia, DP Tittensor, V Tunnicliffe, B Wallace, R Wanless, T Webb, P Bernal, J Rice, A Rosenberg.
Marine environments encompass some of the most diverse ecosystems on Earth. For
example, marine habitats harbour 28 animal phyla and 13 of these are endemic to
marine systems. In contrast, terrestrial environments contain 11 animal phyla, of which
only one is endemic. The relative strength and importance of drivers of broad-scale
diversity patterns vary among taxa and habitats, though in the upper ocean the
temperature appears to be consistently linked to biodiversity across taxa (Tittensor et
al. 2010). These drivers of pattern have inspired efforts to describe biogeographical
provinces (e.g. the recent effort by Spalding et al., 2013)) that divide the ocean into
distinct regions characterized by distinct biogeochemical and physical combinations).
Biogeographers such as Briggs (1974) examined broad-scale pattern in marine
environments in historical treatises and although many of the patterns described
therein hold true today, the volume and diversity of data available to address the
question have increased substantially in recent decades. We therefore focus our chapter
on more recent analyses that build on those early perspectives. The International
Census of Marine Life programme that ran from 2000-2010 provided significant new
data and analyses of such patterns that continue to emerge today (McIntyre, 2011;
Snelgrove, 2010). Indeed, many of our co-authors were part of that initiative and that
influence is evident in the summary below. In the few years since that programme
ended, some new perspectives have emerged which we include where space permits,
noting that we cannot be exhaustive in coverage and also that the large data sets
necessary to infer broad-scale patterns do not accumulate quickly.