Nocturnal changes in the mean length of a euphausiid population: vertical migration, net avoidance, or experimental error?

The lengths of 9 421 Antarctic krill, Euphausia superba Dana, from 144 net samples from 4 depth ranges at the same location (north east of South Georgia) were measured over a period of 6 d. While the mean lengths showed differences between each depth range in each time period of each day, the only consistent difference occurred in the period 00.00 to 04.00 hrs local time, when the mean lengths of krill were up to 2 mm greater than those found at other times. While vertical migration may have contributed to this nocturnal change in mean length, differences in measurement techniques between observers were also implicated. Controlled laboratory experiments confirmed that there were significant differences between measurements of the same krill analysed by six different observers while there were no significant differences in replicate measurements made by a single observer. These results have important implications for future studies involving length and length-frequency analysis, and suitable precautions and experimental protocols are suggested.