Improving Stock Assessments for Rockfishes Using Habitat-referenced Acoustic Surveys in the Gulf of Alaska

Principal Investigator: Chris Wilson
Co-Principal Investigators: Darin Jones, Chris Rooper, Paul Spencer, Dana Hanselman, John Heifetz
External Collaborators: Tom Weber (CCOM, University of New Hampshire)

Groundfish stock assessments in the Gulf of Alaska (GOA) rely on a time series of estimated biomass from biennial bottom trawl (BT) surveys. These BT surveys use stratified random designs where densities of fish from survey tows are expanded across stratum areas to compute biomasses (and variances). The expansion of fish densities estimated from trawlable areas to untrawlable areas potentially results in both bias and imprecision in BT survey abundance estimates for two reasons: Fish density differences likely exist for these two habitat types and the area of the two habitat types within each stratum are poorly known. Thus, to accurately assess groundfish populations, reliable estimates are needed of both the extent of untrawlable substrate to better define survey strata, as well as the fish densities in both the trawlable and untrawlable habitats.

We will use acoustic and optic tools in this project to: 1) map a subsample of areas historically designated as trawlable and untrawlable by the BT survey to re-evaluate the trawlability designation based on newly developed acoustic backscatter metrics, and 2) compare the species and abundance of rockfishes within these trawlable and untrawlable areas (Figs 1-2). An innovative acoustic-camera (AC) survey method will be used to provide abundance estimates for the dominant rockfish species in the trawlable and untrawlable habitats. Multibeam data will be collected to more precisely characterize the trawlability of the seafloor using various backscatter metrics. The AC survey data and multibeam data will be collected in subsamples of both habitats throughout much of the GOA as an ancillary project during the summer 2015 GOA-wide biennial AFSC acoustic-trawl pollock survey aboard the NOAA ship Oscar Dyson.

FY2015Project15028_002.png          FY2015Project15028_003.png



Figure 1. Historically classified trawlable (green) and untrawlable (red) bottom trawl (BT) survey grid-cells for a portion of the Gulf of Alaska (GOA) BT survey near Kodiak Island. Lower inset represents proposed mini acoustic-camera survey sampling design of three acoustic transects (black lines) and three arbitrary locations for stereo camera deployments (red Xs) in a single grid-cell. Upper insets are representative images of rockfishes and untrawlable habitat from the 2013 pilot study conducted during the GOA acoustic-trawl survey.  We anticipate that the multibeam data also collected during these mini surveys will more precisely estimate seafloor trawlability compared to historical classifications.



Figure 2. Preliminary results of average acoustic backscatter (sA – at 38-kHz, m2/nmi2) attributed to rockfishes (other than midwater Pacific Ocean perch) based on 16 trawlable and 16 untrawlable grid-cell mini-surveys from a 2013 pilot study suggest higher rockfish densities  in untrawlable sites.


Project #


InPort ID#




Alaska Fisheries Science Center (AFSC)


Publications & Presentations

Media & Other Products