Predictors of species richness in the deep-benthic fauna of the northern Gulf of Mexico

Species richness in macrofauna and megafauna collected with box cores and trawls from 35 standard stations over a depth range of 175-3720 m in the northern Gulf of Mexico was examined in terms of two primary questions: (1) are observed patterns random? and (2) if not, what environmental factors might account for the patterns? A null model tested whether richness vs. depth distributions were random. Groups with species that had broad vertical depth ranges fit the null model better than groups with small ranges, but for almost all phyla a non-random pattern was indicated. With randomness as a proximal explanation ruled out, further examination of the relationship between richness and environmental factors was justified. A generalized linear model (GLM) showed that a suite of 18 factors categorized as food-related, habitat-related, pollution-related and location-related were significantly related to richness patterns, but that different mixes of factors applied to different phyla. No one factor accounted for any observed patterns. Thus, each taxonomic group needs to be examined individually, and no generally applicable explanation for the causes of richness patterns may exist. Nonetheless, mapping richness itself indicates valuable areas in the Mississippi Trough that must receive special consideration and possible protection. (C) 2008 Elsevier Ltd. All rights reserved.