Coupling Echosounding and Baited Stereo-Videography to estimate pelagic fish biomass
Jessica Meeuwig| 01 December 2021
New publication
Acoustic surveys and video-based sampling are two non-destructive methods for counting fish and estimating their biomass. In our new paper, published in Ecosystems, we twin these two methods for the first time in order to estimate fish biomass in the top 200 m across 118,000 km2 of the Chagos Archipelago. Fundamental to achieving this outcome was combining the scale of acoustics – some 3,000 km of acoustic transects – with species identifications and abundance derived from 546 baited remote underwater video systems (BRUVS) samples.
The last estimate, in 2016, was 4.08 million tonnes of fish in the Archipelago. Combining these two methods gave ecologically credible patterns across hours and years, and from metres to 1000’s of kilometres, providing a marked improvement on the use of acoustic data alone. The Chagos Archipelago is part of the British Indian Ocean Territory Marine Protected Area, established in 2010, and this twinned methodology provides a significant step forward to monitor fish biomass within the park.
Citation: Letessier TB, Proud R, Meeuwig JJ, Cox MJ, Hosegood PJ, Brierley AS. 2021. Estimating pelagic fish biomass in a tropical seascape using echosounding and baited stereo-videography. Ecosystems.
01
Acoustic and BRUVS sampling
Coupled acoustic (red lines) and mid-water baited remote underwater video systems (open black circles) sampling activity within the British Indian Ocean Territory fisheries exclusion zone (inset), in November 2012, January 2015 and February 2016 (bottom panels, in chronological order, showing 500 m isobaths). Labels denote sampling sites, defined by BRUVS clusters, at North Western Station (NW), Peros-Banhos and Salomon atolls (PBSa), Great Chagos Banks (GCB), Marline Mount (MaMo), Egmont atoll (Egm), and Sandes Swart seamount (SaSw).
02
Echo intensity to biomass
Steps required for converting echo intensity to biomass using baited camera measurements: a Acoustic intensity distribution (left pane) and typical echogram (right pane; n pings = 200; depth range = 800 m), computed as nautical-area scattering coefficient (NASC, m2 nmi−2), averaged into 20 pings by 200 m depth cells. b Fish fork length distribution (left pane), derived from stereo baited remote underwater videos systems (illustrated by a typical frame, right pane). c) Weight distributions calculated from weight-length relationship (right pane). d Fish biomass distribution (left pane), calculated by converting NASC observations to biomass using acoustic conversion factor, TSkg (right pane).
03
Biomass density
Estimating pelagic fish biomass density in the Chagos Archipelago. Generalised Additive Model partial plots (top two panel rows), predictions of pelagic fish biomass density (for the year 2015, at midday, bottom left panel), and confidence interval (95% CI, bottom right panel) across the Chagos Archipelago. Centroids of BRUVS cluster sites are marked by black spots. Mean fish biomass density is 279.68 (197.87, 400.15) g m−2, yielding 33.09 (23.41, 47.35) million tonnes of fish for the year 2015.
04
Residuals of biomass density per site
Smooth functions (black line) showing mean Pearson residuals from the fish biomass density GAM predictions, colour coded for site.
