ABSTRACT
The lively debates sparked largely by Grassle & Maciolek’s (1992) study point to several obvious shortcomings in understanding deep-sea diversity. First and foremost, sampling efforts of the intensity and quality of their study are needed elsewhere in the deep sea in order to explore the generality of high deep-sea diversity. Some of these sampling efforts must be directed towards the vast areas of the South and equatorial Pacific where limited, but very diverse, samples suggest that the North Atlantic may actually be relatively speciespoor by deep-sea standards (Poore & Wilson 1993). As coverage of previously unsampled areas is achieved, we should be able to determine whether the proportion of undescribed species is closer to 50% or 0.5% on a global basis. Such sampling would considerably improve on one approach to estimating total species number. One other ideal, though perhaps unattainable goal, is to sample an area of the deep sea well enough for species versus abundance curves to approach asymptotes at multiple scales (e.g. km scale to 100s of kms). This would provide a second means to improve estimates of total deep-sea species richness. In terms of taxa, there is a pressing need to improve estimates for meiofaunal species number; Lambshead’s (1993) estimate of 108 species of nematodes swamps macrofaunal estimates and its validity and generality must therefore be tested. Bacteria and Archaea also represent an unknown, but potentially enormous, species pool as indicated by the planktonic species data of Giovannoni et al. (1990). Because essentially all marine taxa are more species-rich in the benthos than in the water column, because microbial habitat complexity is much greater in sediments than in the water column, and because the bulk of marine microbes actually reside in the sediments, a vast diversity of microbial species is likely to be discovered in deep-sea (and shallow-water) sediments (see also Fuhrman & Davis 1997). For all marine habitats, including shallow-water systems, it is necessary to move beyond the previous reliance on samples from the North Atlantic if better estimates of global species numbers and pattern are to be achieved. Better evaluation of diversity patterns will also lead to greater understanding of diversity maintenance because current broad-scale comparisons are based on a very limited number of data points (e.g. Rex et al. 1993, Poore & Wilson 1993).
