Everson et. al. (1988), Foote et. al. (1990), and Everson et. al. (1990) are all based on the same experiment and report new calculations of the target strength of krill (Euphausia superba) that propose questioning previous measurements of krill target strength and estimates of abundance derived from them. Their results can be shown to have significant weaknesses in both data used and assumptions made in calculating target strength. Whether the results of these papers are valid is critically important, given that they are being cited as a basis for increasing the estimates of krill standing stock in the Antarctic. The weaknesses in their data stem from the methods used to calculate target strength from the data and fall into three categories: 1) Errors in measurements and assumptions; 2) density dependent effects; and 3) compounding of diel effects. The possible bias in the calculations of target strength due to these weaknesses indicate an estimated 3-5 dB error due to unaccounted for directivity effects as a result of items under category 1; an estimated 2-3 dB bias due to systematic errors under category 2; and an estimated 3-4 db error due to category 3. Thus there is reason to believe that there is 8-12 db of possible error in the calculations. This is most if not all of the difference cited between their target strengths and those of other investigators at 120 kHz. These target strength calculations seem to be flawed (no correction for directivity and behavior), contain a significant density dependent bias (correlation of target strength of an individual krill on the number of krill used to measure it), and contain or are likely to contain an error due to diel behavior effects
Abstract:
The homogeneity of krill length composition in commercial trawl catches (hauls) was investigated by comparing length compositions from different parts of the same haul. There was no significant difference (P>0.05) in 53 out of 60 hauls. Therefore it is concluded that sampling from one part of the haul provides a reasonable estimate of length composition in the entire haul as a whole. In contrast to small variations in length composition within the same haul, variation between different hauls was most noticeable in areas where krill length compositions were bimodal.