Rockfish Research Projects: Assessment of Rockfishes in Untrawlable Areas

The Resource Assessment and Conservation Engineering Division at the NMFS Alaska Fisheries Science Center (AFSC) conducts bottom trawl surveys in Alaska each year to estimate the abundance of important groundfish species. The survey nets used for these bottom trawl surveys are limited in their ability to sample rocky areas, since the nets are easily snagged and torn on rough rocky substrate (Fig. 1). A different method is therefore needed to survey the abundance of fishes in these rocky areas.

In 2008, researchers from the Groundfish Assessment Team and the Midwater Assessment and Conservation Engineering Program began a research program to apply new acoustic-optic methods to assess rockfishes in untrawlable areas. This research was jointly funded by the AFSC and a grant from the North Pacific Research Board and was conducted at the Snakehead Bank in the central Gulf of Alaska (Fig. 2). The first step was to identify and estimate the area that was untrawlable in the study site using a fisheries multibeam echosounder (SIMRAD ME 70). Then we measured the amount of acoustic backscatter from fishes in the water column in untrawlable areas. Then we partitioned this backscatter into different rockfish species using optical methods and a ruggedized bottom trawl to estimate the species composition and height of the fish off the seafloor. Next we estimated the length of each species of rockfish in untrawlable areas. Finally we combined the acoustic, species and length data to produce biomass estimates for major rockfish species in the untrawlable area.

Estimating the untrawlable areas

An important part of the acoustic-optic assessment research is finding methods that can easily identify and map untrawlable areas. Researchers at the Center for Coastal and Ocean Mapping at the University of New Hampshire analyzed data collected from the fisheries multibeam echosounder (ME70) installed on the NOAA vessel Oscar Dyson. Seafloor depth and multibeam backscatter were collected from the ME70 along a survey trackline (Fig. 3).

For comparison, trawlability along the survey path was estimated from the underwater video collected from a remote operated vehicle (ROV) and a stereo drop camera (SDC) (Fig 4). This allowed us to groundtruth the estimates of trawlability from the ME70. It was discovered that oblique angle backscatter (the backscatter at an incidence angle with the seafloor from 30 to 60°) provided the best discrimination between trawlable and untrawlable areas. This finding allowed the compilation of a trawlability map of the study area. Further research on mapping trawlability has continued in more recent years as researchers continue to seek low-cost methods for mapping trawlability for AFSC bottom trawl surveys throughout Alaska.

Measuring the acoustic backscatter

Fisheries acoustics uses the level of sound backscatter (through echo integration) to estimate fish biomass in the water column. Acoustic-trawl surveys have been commonly applied to midwater schooling fish such as walleye pollock (Theragra chalcogramma). Echo integration techniques are well established and during this study standard protocols for data collection and analysis were used. To estimate the biomass of rockfish at the Snakehead Bank, researchers collected acoustic data over a series of parallel transects using a scientific echosounder operating at 5 frequencies aboard the NOAA vessel Oscar Dyson In total 8 passes (4 during daytime and 4 during nighttime) were completed along the parallel transects.

The acoustics indicated that most rockfish backscatter (95%) was observed in the untrawlable regions of the study area. There were no noticeable differences in the amount of rockfish backscatter observed during daytime versus nighttime passes. Most of the rockfish backscatter was within 2 m of the seafloor during both daytime and nighttime. Very near the seafloor, it is sometimes difficult to distinguish the acoustic echoes from fish from the dominating echo of the substrate, this region is called the acoustic dead zone. In the Snakehead Bank study, the height of the acoustic dead zone (from the seafloor) was estimated at 0.75 m.

Finding the dominant rockfish species

For midwater fish species, midwater trawls are usually used to estimate species and size composition of target fishes. For the Snakehead Bank study, midwater trawls could not be used to verify targets, so optical solutions were explored. These optical solutions were a stereo drop camera (SDC) and a remote operated vehicle (ROV). For comparison, we also used a ruggedized bottom trawl that employed large rockhopper gear allowing it to be towed over rough seafloor areas. These tools were deployed in areas where the acoustic data indicated there were schools of rockfish.

Fourteen species of rockfish were identified in the study area. The majority of rockfish identified were dusky rockfish, followed by harlequin rockfish, northern rockfish and Pacific ocean perch. Almost all of the Pacific ocean perch were found during one SDC deployment that was in deeper water and slightly outside the core survey area.

Using the SDC, researchers were also able to measure the distance of fish from the seafloor (Fig. 5). The only species of fish that were not found disproportionately in the acoustic dead zone were dusky rockfish and northern rockfish. Some Pacific ocean perch and harlequin rockfish were also observed above the acoustic dead zone. The species composition near the seafloor included a more diverse assemblage than was observed in the water column (14 species versus 4 species).

Based on these results, the acoustic backscatter data collected from the NOAA vessel Oscar Dyson was estimated to be comprised of dusky rockfish, northern rockfish and harlequin rockfish in proportion to their abundance above the acoustic dead zone (> 0.75 m off the seafloor).

Measuring the rockfish lengths

Measuring fish lengths is a crucial step in estimating biomass using acoustics. Since the size of a fish’s swim bladder determines how much sound it will reflect, measuring a fish’s length allows the conversion of an individual fish size to an individual fish echo. Length-weight relationships then allow the conversion to biomass (or total weight).

We compared three methods for determining fish length; stereo image analysis for video captured from the SDC, parallel lasers mounted on the ROV and a measuring board for trawl captured animals. Stereo image analysis is much like human vision, in that it takes two synchronized pictures of the same scene and then uses triangulation to determine the size of objects viewed in both images. On the ROV parallel lasers (20 cm) apart allowed the analyst to estimate fish lengths using the known distance. The measurements from the optic methods were compared to those from fish that were captured in the ruggedized bottom trawl.

The lengths for the most abundant species were found to be similar among the three methods (Fig. 6). However, more small fish were measured with the optic methods, pointing to a possible bias in the trawl captured individuals. It is likely that the smallest individuals (such as small harlequin rockfish) were able to escape the trawl through the meshes and were not captured.

Combining the data to estimate biomass

Biomass estimates were computed for the dominant species that were present in the water column (dusky rockfish, northern rockfish and harlequin rockfish). The biomass estimate incorporated the total estimated acoustic backscatter and the mean length for each species. A target strength versus length relationship from the literature was used to compute the strength of an individual echo from a rockfish. The total backscatter divided by the individual echo strength gave an estimate of the total number of fish. Using the average size and a length weight relationship, the number of fish was converted to biomass. Mean biomass estimates resulted in 2,350 t for dusky rockfish, 331 t for northern rockfish, and 137 t for harlequin rockfish. A similar acoustic-optic methodology has been used to estimate the biomass of northern rockfish in another area at Zhemchug Canyon in the eastern Bering Sea.

Contact

For more details on the results summarized above, see the documents below or contact Chris Rooper at chris.rooper@noaa.gov or 1-206-526-4689.