The use of video surveys, a Geographic Information System and sonar backscatter data to study faunal community dynamics at Juan de Fuca Ridge hydrothermal vents

The study of benthic faunal dynamics in the deep sea is often hampered by technical limitations that include our inability to accurately reconstruct or view areas of seafloor larger than what can be seen through submersible view ports or camera lenses. Manual mapping and image mosaicking have been used with some success to analyse organism distribution within vent sites and at the scale of hydrothermal fields (Juniper et al. 1998). In both cases, video or photo transects serve as the primary source of information. Image mosaicking, in principle, is less subjective than mapping by hand, but the present state of the art requires precise control of camera distance and viewing angle to avoid spatial distortion as images are adjusted and fitted into the mosaic (Grehan & Juniper 1996). The reality of operating submersibles over the rugged volcanic and sulphide edifice terrains where hydrothermal vents are found does not always permit collection of imagery suitable for mosaicking, and so manual mapping from video records is still used in some applications (Sarrazin et al. 1997). Both techniques yield useful maps that can be used to elucidate relationships between vent organism distribution and habitat features. Time-series mapping with either method can be used to study organism growth, faunal succession, and the response of species and communities to environmental change. Cartographic studies of vent fauna can be further developed by integration of map data into a Geographic Information System (GIS), where a range of powerful, automated spatial analysis tools can be used to explore ecological questions. This paper reports on a combined hand mapping and GIS approach that we have been using to study early faunal community dynamics at Cloud Vent, a new hydrothermal site created by a January 1998 eruption on Axial Volcano, on the Juan de Fuca Ridge. We also report on preliminary results of organism and substratum differentiation in sonar backscatter signals, and consider their complementarity with visual imagery. Sonar backscatter analysis offers a potential for automated, swath mapping of habitat over large areas of seafloor, at scales that go well beyond what can be practicably achieved with visual imagery. Seafloor geological and biological features influence the strength and temporal distribution of the acoustic signal that is reflected back to the detector (the first echo), according to their respective densities and textures.