Abundance and size of the sea scallop population in the Mid-Atlantic Bight
Date
2013
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
The stock of the Mid-Atlantic Bight sea scallop fishery is assessed every year through the use of various dredging and imaging techniques. The sustainability of the fishery depends on the proper setting of the yearly catch limits based on the assessment of the preceding year. Within the past 10 years, digital image surveys have been explored as a potential method to supplement the yearly dredged based surveys. AUVs have been shown to be a successful platform for rapidly and accurately performing seafloor image surveys of benthic habitats. In 2011, a Teledyne-Gavia autonomous underwater vehicle (AUV) with a hull-mounted camera was used to non-invasively optically and acoustically image 313 km of the seafloor within the Mid-Atlantic Bight at a constant altitude of 2 m. Survey transects were completed at 24 open access ground locations and 3 additional locations within the Elephant Trunk Access Area. Trained image analysts, using a scallop counting and sizing algorithm developed for this stock assessment, were able to enumerate and size sea scallops within the collected 250,000 seafloor images, finding that the region had an overall scallop density of 0.027 scallops/m2. Georeferenced data was tagged by the AUV inertial navigation system (INS) to every seafloor image, allowing for unprecedented meter scale spatial analysis of the sea scallop distribution. The relationship between image subsampling and the accuracy of the resulting scallop density was explored via simulations run on the image analysis results. Eight AUV transects were resurveyed by a New Bedford commercial scallop dredge for shell height calibration data and to calculate the harvest efficiency of the dredge (0.60). Image analysis and backscatter data collected by the AUV’s 900 kHz side-scan sonar were used to classify seafloor substrate types. The surveyed scallop strata were classified as 98.6% sandy seafloor with the remaining 1.4% representing intermittent shell hash, mounds, and ripples. The side-scan backscatter data revealed other varied seafloor texture, including escarpments from scallop dredge trawling and wave created sorted bedforms. Seafloor dredge scar area measured from the side-scan backscatter data and National Oceanic and Atmospheric Administration (NOAA) vessel monitoring system (VMS) tracking data were used as a proxy for fishing effort. Increased dredging was found to positively skew shell height distributions.