SURVEY DESIGN CONSIDERATIONS FOR DEEP-SEA BENTHIC COMMUNITIES USING ROVS

Recent advances in availability and technical sophistication of manned and unmanned submersibles and other sampling gear have increased scientist's abilities to investigate the biology and ecology of deep-sea organisms. In this paper we discuss several issues (primarily gear type, sample unit, and sample distribution) affecting the design of sampling to assess abundance and distribution of species in deep-sea habitats. Although all sampling gear have advantages and limitations, remotely operated vehicles (ROVs) are one of the most valuable tools for assessing the distribution of many species. The appropriate sample unit for a particular population (e.g., single trawl, single photograph, or video transect), depends upon the focus of the investigation, ability to quantify the spatial dimensions of the area observed, and abundance of the species in question. If possible, samples should be taken randomly in the study area (or randomly through time where appropriate), but compromises must usually be made between statistically rigorous randomized designs and logistic constraints on studies in the deep sea. The objective of any sampling program is to quantify, as accurately as possible, the parameter of interest, with minimal cost. Because the cost of sampling in the deep sea is usually much greater than other habitats, sampling design may open need to be planned more carefully to maximize the information gained.