Assessment of Trawlable and Untrawlable Seafloor Using Multibeam-Derived Metrics
|Title||Assessment of Trawlable and Untrawlable Seafloor Using Multibeam-Derived Metrics|
|Publication Type||Journal Article|
|Authors||Pirtle, JL, Weber, TC, Wilson, CD, Rooper, CN|
|Journal||Methods in Oceanography|
|Date Published||May 2015|
|Keywords||Acoustic backscatter, Bottom-trawl survey, gulf of alaska, Seafloor Characterization|
Groundfish that associate with rugged seafloor types are difficult to assess with bottom-trawl sampling gear. Simrad ME70 multibeam echosounder (MBES) data and video imagery were collected to characterize trawlable and untrawlable areas, and to ultimately improve efforts to determine habitat-specific groundfish biomass. The data were collected during two acoustic-trawl surveys of the Gulf of Alaska (GOA) during 2011 and 2012 by NOAA Alaska Fisheries Science Center (AFSC) researchers. MBES data were collected continuously along the trackline, which included parallel transects (1–20 nmi spacing) and fine-scale survey locations in 2011. Video data were collected at camera stations using a deployed camera system. Multibeam-derived seafloor metrics were overlaid with the locations of previously conducted AFSC bottom-trawl (BT) survey hauls and 2011 camera stations. Generalized linear models were used to identify the best combination of multibeam metrics to discriminate between trawlable and untrawlable seafloor for the region of overlap between the camera stations or haul paths and the MBES data. The two best models were developed using data collected at camera stations with either oblique incidence backscatter strength (Sb)(Sb) or mosaic SbSb in combination with bathymetric position index and seafloor ruggedness; these described over 54% of the variation between trawlable and untrawlable seafloor types. A map of predicted seafloor trawlability produced from the model using mosaic SbSb and benthic-terrain metrics demonstrated that 58% of the area mapped (View the MathML source5987km2) had ≥50%≥50% probability of being trawlable and 42% of being untrawlable. The model correctly predicted 69% of trawlable and untrawlable haul locations. Successful hauls occurred in areas with 62% probability of being trawlable and gear damage occurred in areas with a 38% probability of being trawlable. This model and map produced from multibeam-derived seafloor metrics may be used to refine seafloor interpretation for the AFSC BT surveys and to advance efforts to develop habitat-specific biomass estimates for GOA groundfish populations.