Jonathan L. Watkins

Acoustic discrimination of Southern Ocean zooplankton

Acoustic surveys in the vicinity of the sub-Antarctic island of South Georgia during a period of exceptionally calm weather revealed the existence of a number of horizontally extensive yet vertically discrete scattering layers in the upper 250 m of the water column. These layers were fished with a Longhurst-Hardy plankton recorder (LHPR) and a multiple-opening 8 m2 rectangular mid-water trawl (RMT8). Analysis of catches suggested that each scattering layer was composed predominantly of a single species (biovolume>95%) of either the euphausiids Euphausia frigida or Thysanoessa macrura, the hyperiid amphipod Themisto gaudichaudii, or the eucalaniid copepod Rhincalanus gigas. Instrumentation on the nets allowed their trajectories to be reconstructed precisely, and thus catch data to be related directly to the corresponding acoustic signals. Discriminant function analysis of differences between mean volume backscattering strength at 38, 120 and 200 kHz separated echoes originating from each of the dominant scattering layers, and other signals identified as originating from Antarctic krill (Euphausia superba), with an overall correct classification rate of 77%. Using echo intensity data alone, gathered using hardware commonly employed for fishery acoustics, it is therefore possible to discriminate in situ between several zooplanktonic taxa, taxa which in some instances exhibit similar gross morphological characteristics and have overlapping length– frequency distributions. Acoustic signals from the mysid Antarctomysis maxima could also be discriminated once information on target distribution was considered, highlighting the value of incorporating multiple descriptors of echo characteristics into signal identification procedures. The ability to discriminate acoustically between zooplankton taxa could be applied to provide improved acoustic estimates of species abundance, and to enhance field studies of zooplankton ecology, distribution and species interactions.

An assessment of the utility of an acoustic Doppler current profiler for biomass estimation

Acoustic Doppler current profilers (ADCPs) are being used increasingly to derive estimates of zooplankton and micronekton abundance. The absence of a practical means for users to perform ADCP calibration, however, makes the quantitative value of this approach questionable. In an attempt to evaluate ADCP- derived volume backscattering strength (S-c) data, and hence to assess the utility of such measurements for biomass estimation, a regularly calibrated Simrad EK500 scientific echo-sounder (operating at 38, 120 and 200 kHz) and an RDI narrowband ADCP (153.6 kHz) were similarly configured and run in synchrony on a transect in the Southern Ocean. Data were collected by both instruments from congruent depth (4 m) and time (2 min) bins in order to allow direct comparison of numerous discrete values without the need for further signal averaging. Echoes were recorded from the Antarctic krill, Euphausia super ba, in deep- ocean and on-shelf locations during day and night. ADCP-derived volume back scattering strength data from shallow, evenly distributed krill targets were well correlated with equivalent data from the EK500 (r(2) > 0.98), and the offsets between instruments conformed to those predicted for their respective operating frequencies by a theoretical model of sound scattering by krill (H-0: S-upsilon 153.6 kHz = S-upsilon 120 kHz + 2.3 dB, t(0.05(2),155) = 1.98, t = 0.74, P = 0.46). Data from deeper, more irregularly distributed targets differed significantly (P < 0.001). We conclude that under some ideal, but probably rare, circumstances data from the ADCP can be used to derive biomass estimates. The numerous uncertainties surrounding ADCP calibration and the current practical impossibility for users to monitor system performance should, however, preclude these instruments from being used as a matter of course to determine abundance estimates, a task that we believe should remain firmly within the domain of a well calibrated scientific echo sounder

An assessment of the biological and acoustic characteristics of swarms of Antarctic krill

Measurements on krill from distinct swarms and echo-sounder measurements of the same swarms have been analysed using multivariate methods. The analysis has addressed the problems of whether krill swarms in a small area may be of a limited number of types, and whether the biological characteristics of the krill could be related to the echo-sounder measurements of the swarm. The main sources of biological variation between krill in different swarms seemed to be related to size and maturity. Krill swarms did not appear to form groups based on biological similarity, and swarms caught in the same haul were as different as swarms caught in different hauls. However, swarms could be divided into groups on the basis of time of day and swarm depth. The biological features (size, maturity, sex ratio and moult stage) of krill in the swarms were not related to echo-sounder observations (horizontal and vertical extent, depth and density of swarms), although some degree of separation between swarms containing large and small krill may be possible if the shape of the swarms is considered. Information on biological characteristics of krill in swarms must still be collected using nets.

Distribution, Biomass and Demography of Antarctic Krill, Euphausia superba

There are increasing concerns over potential long-term changes in krill distribution and abundance, determining whether such changes are related to climate change and/or harvesting and how they might be distinguished from multi-scale variability. With nearly a century of observations on krill distribution, the general patterns of occurrence were determined over 50 years ago, however the last decade has seen an important consolidation of details of seasonal and inter-annual distributions, although some potentially important krill habitats such as the Bellingshausen, Amundsen and Ross Seas are still poorly sampled. Moreover the unexplored perennial multi-year pack-ice regions still await survey by remote under-ice samplers. Krill consistently occur in the upper 200 m of the ocean but various recent observations show that a substantial proportion of the population may be found below this pelagic zone. To further understand variability and change there has been a focus on the importance of historic data with the compilation of comprehensive databases, re-analysis of large scale synoptic surveys and the generation of smaller scale time series that now span several decades. In particular the compilation of 90 years of net samples has provided a key resource both in showing changes in large-scale abundance over the last 50 years and together with acoustic data in generating modern estimates of overall krill abundance and biomass. Meanwhile, the increasing availability of ocean circulation models together with remotely sensed data, has enabled the exploration of large-scale teleconnections at both the regional (Scotia Sea) and circum-polar scale and at the seasonal to decadal time scales.