Optical Sea Scallop Surveys and Yellowtail Flounder Observations Using the HabCam Imaging System on the U.S. Northeast Continental Shelf

We used the HabCam camera system to conduct scallop surveys for abundance and size and observe yellowtail flounder in relation to substrate in the Nantucket Lightship Closed Area (NLSCA) (June 2009), and in the Southeast Part of Closed Area II (SEPart) and the Western side of the Great South Channel (WGSC) (August 2010). HabCam is a cabled optical imaging system “flown” ~ 3 meters off bottom, from a ship traveling 5 kts, collecting ~6 high resolution still images per second. A track 100 nm long and 259,200 m² is imaged each 24 hours. In each area every 10^th image was manually processed and kriging was used to interpolate from counts along track into 2D estimates of scallop abundances. In the NLSCA (area: 425 nm²) our trackline was a modified spiral with track spacing from 2.6 to 1.3 nm. Trackline length was 348 nm producing 1,235,251 images in three days resulting in an area sampled of 0.187 nm² or 0.57%. Scallop density along tracklines was between 0 and 23 scallops/m² with dense aggregations found at patch scales of 400 and 900m. Total number of scallops in the NLSCA was estimated at 197,545,580 with a mean of 0.187 scallops/m² and a CV of 0.04. The proportion of exploitable (commercially available) scallops was 92%. Yellowtail flounder were sparse. The SEPart survey (area: 870 nm²) was 475 nm in length producing 1,760,000 images in four days with an area sampled of 0.025 nm². Total number of scallops was estimated at 702,851,843 with a mean of 0.21 scallop/m² and a CV of 0.02. 39% were exploitable. Scallops were most abundant in the northeast corner of the area. 251 observations of yellowtail flounder showed a scattering throughout the area with greatest concentrations in the southwest corner. The WGSC survey (area: 450 nm²) was 395 nm in length producing 295,000 images in 3.5 days with an area sampled of 0.09 nm² . Total number of scallops was estimated at 1,271,999,000 with a mean of 1.5 scallop/m² and a CV of 0.03. Scallops were scattered in sandy areas, but a dense patch reaching 18/m² five nm in length was evident between two ridges on sand/gravel substrate in the central western part of the area. 28% were exploitable. Yellowtail flounder were scattered with no obvious aggregations. Initial automated counting and sizing of scallops showed a 68-95% accuracy compared with manual classifications, depending on substrate.