An investigation of variability in krill length and observation duration of scientific observer involving in krill biological data collection in Antarctic krill fishery

The CCAMLR Observer Manual details scientific observation guidelines and contains a set of logbook forms with instructions for recording observation, and krill length distribution and time budget are two key aspects related to Scientific Observation. This document investigates the variability in krill length under 4 sampling strategies, i.e., 50, 100, 200 and 500 krill randomly sampled from the catch in accordance with the requirements in the Manual, in which sampling strategy of 200 krill is required for the current SISO. The relationship between observation duration and sex (proportion of females in the samples) and period (day and night) has been compared for each sampling strategy. A Markov Chain Monte Carlo Sampler for multivariate generalized linear mixed model is used to investigate the effect of the biological parameters, including size, sex, maturity stage, color (feeding intensity), stipulated in the Observer Manual and period on observation duration of scientific observer, and the results indicated that krill length, period and body color affects observation duration. The size distributions are similar generally for the sampling strategies of 50, 100 and 200 krill. Compared to undertake same task during the daytime, observers consumed more time to undertake biological data collection at night. 500 krill sampling has significantly different length distribution from other sampling strategies, which does not mean necessarily difference in length distribution in the catch in nature between them, rather than change in body shape due to long-time exposure in the water-losing environments. The results and relevant recommendations are expected to be used to support the revision and improvement of the current CCAMLR Observer Manual and therefore krill fishery management.