Relationships between near-bottom dissolved oxygen and sediment profile camera measures

The United States Environmental Protection Agency (U.S. EPA) and other environmental authorities regulate concentrations of dissolved oxygen (DO) as a measure of nutrient-related eutrophication in estuarine and coastal waters. However, in situ DO concentrations are extremely variable, and their characterization requires an extensive sampling program to provide data over meaningful scales of time and space. In contrast, benthic faunal communities integrate the impacts of low DO over time, and can be rapidly assessed using benthic imaging. The goal of this study was to quantify the relationships between near-bottom dissolved oxygen and measures derived from benthic imaging with a sediment profile camera. We monitored three stations in Narragansett Bay (Rhode Island, USA) for DO and other water quality parameters 15–20 cm above the sediment surface on 15-minute intervals between July and November 2002, and regularly sampled these stations with a sediment profile camera throughout this time period. These soft-sediment stations encompassed several DO environments. We tested for relationships between near-bottom DO and several camera measures, including Nilsson and Rosenberg's Benthic Habitat Quality (BHQ) index, the apparent Redox Potential Discontinuity (aRPD) depth, and various faunal features that can be identified in sediment profile images. Camera measures were examined against a variety of methods of characterizing DO (including mean DO, and the percent of time under various DO thresholds), over a span of time scales from 1 day to 49 days. The best relationship (highest r 2) between near-bottom DO and BHQ was found when DO was evaluated as the percent of time under a hypoxic threshold of 2.6 mg l − 1 over a 28-day time scale (by examining DO records over the 28 days preceding each camera deployment). We found that, over several benthic settings, the BHQ index was successful at identifying environments that had experienced relatively high or low DO over the preceding four weeks. Our sediment profile data showed more variability with DO in the intermediate values of BHQ. We conclude that sediment profile camera measures correlate to DO in areas where low DO is the primary stressor, integrate DO over ecologically relevant time scales, and enable sampling over spatial scales that are meaningful for mapping by virtue of rapid deployment and analysis. We submit that sediment profile camera imagery is a useful assessment and mapping tool for environmental manage