Large mesopelagic fish biomass in the Southern Ocean resolved by acoustic properties

Abstract

The oceanic mesopelagic zone, 200-1000 m below sea level, holds abundant small fishes that play central roles in ecosystem function. Global mesopelagic fish biomass estimates are increasingly derived using active acoustics, where echosounder-generated signals are emitted, reflected by pelagic organisms and detected by transducers on vessels. Previous studies have interpreted a ubiquitous decline in acoustic reflectance towards the Antarctic continent as a reduction in mesopelagic fish biomass. Here, we use empirical data to estimate species-specific acoustic target strength for the dominant mesopelagic fish of the Scotia Sea in the Southern Ocean. We use these data, alongside estimates of fish relative abundance from net surveys, to interpret signals received in acoustic surveys and calculate mesopelagic biomass of the broader Southern Ocean. We estimate the Southern Ocean mesopelagic fish biomass to be approximately 274 million tonnes if Antarctic krill contribute to the acoustic signal, or 570 million tonnes if mesopelagic fish alone are responsible. These quantities are approximately 1.8 and 3.8 times greater than previous net-based biomass estimates. We also show a peak in fish biomass towards the seasonal ice-edge, corresponding to the preferred feeding grounds of penguins and seals, which may be at risk under future climate change scenarios. Our study provides new insights into the abundance and distributions of ecologically significant mesopelagic fish stocks across the Southern Ocean ecosystem.

Keywords: Southern Ocean; acoustics; biomass; mesopelagic fish; target strength.

Figure 1.

(a) Scotia Sea study area, showing sea surface temperature (SST) climatology. Yellow lines are 38 kHz acoustic transects, cyan diamonds indicate net sampling locations used to assess mesopelagic fish community composition and collect morphometric samples. White boxes are CCAMLR subareas 48.2/48.3, for which Scotia Sea fish biomass was calculated. FI, Falkland Islands; SG, South Georgia; SO, South Orkney Islands; SSI, South Sandwich Islands. (b) Predicted log_e NASC for the Scotia Sea used to estimate fish biomass. (c–e) Bar plots illustrating the nonlinear effect of the presence or absence of gas on acoustic estimates of fish biomass. (c) Proportions of fish by swimbladder contents in net samples within 1°C SST groups. Numbers above bars indicate the number of total water column net samples in each group. (d) Acoustic contribution of fish species to mean predicted NASC in SST groups by swimbladder contents, using TS values for median length fish. (e) Relative proportions of gas and non-gas swimbladder fish species contributing to Southern Ocean mean biomass (g m⁻²). Note that the relative proportion of gas-bearing species is highest at warmer temperatures (c) but overall contribute less to biomass estimate (e) because of the nonlinear effect of gas reflection on acoustic backscatter. (Online version in colour.)

Figure 2.

(a) Sea surface temperature climatology (October–April 2005–2017) used to predict NASC, (b) predicted log_e NASC, (c) predicted mesopelagic fish biomass for the Southern Ocean based on median SL fish, assuming that mesopelagic fish are responsible for all of the acoustic backscatter. Scotia Sea CCAMLR sub areas 48.2 and 48.2 are outlined in grey. Circumpolar lines (north to south) indicate mean positions of Antarctic Polar Front (white), Southern Antarctic Circumpolar Current Front (green) and Southern Boundary of Antarctic Circumpolar Current (magenta). (Online version in colour.)