Deep-sea biodiversity at the extremes of the Salas y Gómez and Nazca ridges with implications for conservation

salinity, dissolved oxygen, nitrate, phosphate, and silicate) (S1 Table). The environmental attri- bute data used in the statistical clustering are based on a set of objectively analyzed climatologi- cal fields of in situ long-term historical average data from the 2013 World Ocean Atlas version 2 (https://www.nodc.noaa.gov/OC5/woa13/) and represent the means of the decadal averages from a 57-year climatology of the six parameters. The database is a point-mesh grid that includes over 52 million points that depict the global ocean in x, y, and z dimensions. The data points were statistically clustered based on the six environmental parameters using a k-means clustering algorithm to identify environmentally distinct regions in the water column. The resulting 37 distinct EMUs partition the ocean environment into ecologically relevant regions and are a valuable tool in exploring drivers of biological distributions. This resulting EMU dataset has a spatial resolution of ¼˚ × ¼˚ (~27 km × 27 km at the equator) in the horizontal dimension and was accessed from ESRI EMU online (http://esri. maps.arcgis.com/home/group.html?id=6c78a5125d3244f38d1bc732ef0ee743). EMU numbers, and values for oceanographic variables derived from the World Ocean Atlas 2013 database (temperature, salinity, dissolved oxygen, %O2 saturation, nitrate, silicate, phosphate, chloro- phyll-a) were extracted from the EMU dataset using the Extract Multi Values to Points tool in the ArcGIS Spatial Analyst toolbox. EMUs and associated environmental variables vary across

horizontal space as well as with depth, represented by the EMU dataset as a 3D point grid. Deployments were attributed with the respective EMU at the given sample depth, thus captur- ing depth-specific variations in physicochemical factors. Values for particulate organic carbon (POC) flux to the seafloor were obtained from Lutz et al. [72] via Watling et al. [73]. Seamount data were obtained from the Global Seamount Database [74]. Distance to the nearest seamount and the number of seamounts near sample sites were derived for a 30 km radius around deployment locations using the Buffer tool in ArcGIS Pro. Age of the underly- ing crust [75] was extracted for the seamount nearest each deployment location, using the Extract Multi Values to Points tool, as an indication of seamount age. Bathymetric position index (BPI) is an index of bathymetric heterogeneity that highlights areas of topographic anomalies and is used to identify benthic features (e.g., ridges and valleys) [76]. It is a continu- ous metric of the elevation of a pixel relative to other pixels in a landscape [77]. BPI values at or near zero represent consistent bathyscapes, either flat areas or areas of constant slope, whereas sharp peaks or pits are represented by higher absolute values [77, 78]. BPI was calcu- lated in ArcMap 10.2 using the Benthic Terrain Modeler extension [79]. Input bathymetric data used was Global Multi-Resolution Topography (GMRT) [80] from the Marine Geosci- ence Data System (www.marine-geo.org), with spatial resolution of 3 arcseconds (~90 m). The

GMRT dataset, available through OpenTopography (https://opentopography.org/), is a compi- lation of edited multibeam sonar datasets collected worldwide. The data for our study region includes swaths of mapping data throughout the region, intersecting and surrounding deploy- ment locations. Broad-scale BPI was calculated using an inner radius of 25 cells and an outer radius of 50 cells. Fine-scale BPI was derived with inner and outer radii of 1 and 10 cells, respectively. Both broad- and fine-scale BPI have been shown to be associated with variations in marine faunal communities [76, 81], including those of the deep ocean [45, 82]. Statistical methods Due to the numerical dominance of Amphipoda in our samples, they were excluded from all analyses, except for the comparisons of number of taxa and analyzed separately. Total MaxN per deployment was calculated as the sum of MaxN for all taxa for that deployment. Species diversity was calculated from the Shannon-Weaver Diversity Index [83]:H0 ¼ PðpilnpiÞ, PLOS ONE Deep-sea biodiversity along the Salas y Go´ mez and Nazca ridges PLOS ONE | https://doi.org/10.1371/journal.pone.0253213 June 30, 2021 8 / 27