Spatial and temporal variability of Antarctic krill: implications for stock assessment

• Published in: ICES Journal of Marine Science
• Main Authors: McClatchie, Sam, Greene, Charles H., Macaulay, Michael C., Sturley, Darrell R.M.
• Format: Text
• Language: English
• Published: Oxford University Press 1994
• Subjects: Articles; Antarctic; Elephant Island; Antarc*; Antarctic Krill
• Online Access: http://icesjms.oxfordjournals.org/cgi/content/short/51/1/11; https://doi.org/10.1006/jmsc.1994.1002

Acoustic estimates of Antarctic krill biomass measured in four surveys around Elephant Island during the American Antarctic Marine Living Resources 1990 (AMLR90) cruises were analyzed to examine spatial and temporal variability in absolute biomass. Each survey took 5-6 days to complete, and survey mid-points were separated by 15-17 days. The depth-integrated biomass data were highly positively skewed. The biomass frequency distributions diverged progressively over time from that measured in the first survey, with greatest change occurring between survey 2 and 3. The coefficients of variation over each survey for 30 s length records (order 100m) ranged from 85 to 247%, indicating intense patchiness. Analysis of the fractal dimension of krill biomass revealed patterns that were consistent between surveys and indicated that a horizontal integration interval smaller than 100 m would be necessary to resolve much of the structure in krill spatial variability. In addition to spatial patchiness we found strong temporal variability between surveys at scales on the order of 2 weeks. Our findings indicate that the krill distributions around Elephant Island are highly variable. We suggest that data assimilation models, incorporating repeated fine-scale sampling of physical variables and acoustic back-scattering, be employed to quantify changes in krill distribution and abundance through time in regions characterized by such high variability. We conclude: (1) that sampling with resolutions finer than 100 m is necessary to determine the characteristic scales of spatial patchiness in krill distributions: and (2) the application of biophysical data assimilation models would facilitate more accurate stock assessment of krill. We suggest that traditional survey methods could be improved by introducing long-range moored sonar arrays or towed bodies to operate conjunction with ship surveys. We present estimates of the feasibility of detecting krill aggregations at up to 1-2 km range using a side-looking sonar operating at 20-40 kHz.