Paleo-MARSPEC: gridded ocean climate layers for the mid-Holocene and Last Glacial Maximum

Ecological Archives E095-149


  • Corresponding Editor: W. K. Michener.


Understanding how species have responded to climate change during recent glacial cycles has the potential to inform our understanding of how species will respond to future climate change over the next century. Glacial refugia and historical range shifts over geological time have traditionally been inferred by examining the fossil record or the distributions of sister species using phylogeographic approaches and, more recently, with species distribution models. Species distribution modeling has become a popular tool for projecting range expansions and contractions under climate-change scenarios in terrestrial ecosystems; however, such models are less commonly applied to the study of marine ecosystems despite a similar need for understanding species' vulnerabilities to fluctuating climates. Here I present gridded climatologies for the world ocean representing the annual mean, range, variance, and extremes in sea surface temperature and salinity for the mid-Holocene (6 ka) and the Last Glacial Maximum (LGM; 21 ka) obtained from coupled ocean–atmosphere general circulation models available through the second phase of the paleoclimate modeling intercomparison project (PMIP2). Furthermore, I present geophysical layers representing bathymetry, distance to shore, and six measures of topographic complexity during the LGM, when sea level was approximately 120 meters lower than today. All data layers are global in spatial extent, are downscaled to a 5-arc-minute spatial resolution, and are provided in ESRI raster grid format suitable for analysis in ArcGIS or in R using the raster package. These paleoclimatic data layers complement the modern bioclimatic and geophysical data layers contained in the MARSPEC database and should facilitate further studies on the historical response of marine species to late-Quaternary climate change, including but not limited to inferences based on species distribution models.