Search

We note that Murray-Wallace et al. (2013), in their comments above and in their original articles (see references in Slee et al. 2012), provide no detailed stratigraphic column or location for their observations or photographs so it is difficult to relate some of their arguments and observations to our described section. Before sampling, we examined the Mary Ann Bay exposure. At the eastern end of the site (Slee et al., 2012, Figure 2) were slumped deposits with concentrations of reworked shells in greater quantity than were found in undisturbed parts of the section. We avoided these areas when sampling and trust that Murray-Wallace and co-workers adopted the same strategy.

The coastal shelf of the Gulf of Oman experiences periodic upwelling events during the summer months that are driven by the southwest monsoon. It is unclear what role these events play in the spatial and temporal distribution of the region’s fish assemblage. We carried out trials on two different video techniques to characterize the habitat and fish assemblage along the continental shelf margin near Muscat, Oman. Exploratory surveys with a drift stereo-video revealed three main habitat types: Sand, Reef and Megabenthos. Three areas were chosen for additional sampling using stereo-BRUVS (‘baited’ remote underwater-video systems). On two separate occasions (November 2005 and March 2006) replicate stereo-BRUVS were deployed in each area stratified by the main habitat types. For each teleost and elasmobranch species encountered on the video, an estimate of total body length and the relative abundance (MaxNi) was made. The stereo-BRUVS recorded a wide range of demersal and pelagic teleosts including species of conservation interest such as sharks, rays and groupers. The drift stereo-video recorded significantly fewer species than the stereo-BRUVS (N = 15 versus N = 43). Species diversity from the stereo-BRUVS increased by 96% in March 2006 (N = 41) compared to November 2005 (N = 23), a pattern consistent at all three areas. The structure of the overall fish assemblage (using canonical analysis of principal coordinates analysis) was highly variable both in time and space. There was ample evidence of strong habitat associations, particularly with depth and seasonal shifts in abundance and diversity. We argue the upward migration of oxygen-depleted water into the shallow depths during the late monsoon displaces the demersal fish community along this coast.

Quantitative sampling of benthic communities is central to a wide range of ecological research, from understanding spatial distribution and ecology to impact studies. With the need to sample deep as well as shallow regions, limited sampling capabilities of diver-based methods and the expanding footprint of human activity, there is a need for an effective system capable of classifying benthic assemblages and able to monitor potential anthropogenic impacts. Here we describe a remote system capable of collecting benthic photo-quadrats to depths of 100 m. A procedure for the classification of these images into 64 abiotic and biotic categories is also described. During a 64-day sampling program that included sampling at seven locations along 1,200 km of coastline that resulted in the collection of over 9,000 images, only one day of sampling was lost due to equipment malfunction, with 99.5% of points able to be classified to the taxonomic resolution required, demonstrating the reliability and accuracy of this system. Furthermore, the incorporation of differential GPS and ultra-short baseline positioning system allowed collected images to be geo-referenced to within 0.5 m. Such precision allows the system to be used in conjunction with hydroacoustic habitat mapping techniques and potentially for repeated monitoring of areas with a small spatial extent. Development of this system provides a cost-effective means of quantifying benthic assemblages over broad scales.