CCAMLR

The swept area method was used to estimate biomass density changes in the stocks of five fish species Champsocephalus gunnari, Chaenocephalus aceratus, Pseudochaenichthys georgianus, Notothenia gibberifrons, and Notothenia rossii marmorata in the South Georgia area in the seasons of 1976/77-1986/87. In most of the seasons analysed, the estimates covered a near-bottom layer in about 1/6 of the area of the island shelf.
Assessment results point to considerable variations in fish biomass density, in the studied period. Estimates of fish stocks biomass ranged from 43 to 158 thous.tons. The changes in the biomass level are first of all a result of periodical fluctuations in the biomass density of C.gunnari stock. High biomass density appeared with 2-4 years intervals after recruiting an abundant year-class to the exploited stock. Biomass of other bottom species is more stable, especially in case of N.gibberifrons. Recently, an increase of the C.aceratus stock is observed and, at the same time, a gradual decline of the P.georgianus.

Simple relationships expressing dependence of sampling relative error in biomass estimate (by means of a survey) on statistical characteristics of fish concentration density field under examination and on parameters of survey itself, have been derived with the help of mathematical statistics and methods of calculus of probabilities; biomass estimate is determined as a product of average density for region under examination and area of this region. As to hydroacoustic survey, anisotropy parameter, correlation radius along transects and variation coefficient serve as field characteristics on which error depends, and direction of survey with respect to the axis of the correlation ellipse and frequency of transects (sections) serve as survey parameters. Dependences offered here, can be used in practice both for a posteriori estimation of error made when calculating biomass, and for survey planning on the basis of a priori estimates of statistical characteristics of concentration density field. They might make a basis for the procedure of survey operative control.

A model is set up for the operation (which includes both searching and fishing) of a Japanese krill trawler over a half-month period, and the output is compared to statistics from a sample of data from the commercial fishery. Twenty-four candidate "CPUE" biomass indices are considered, and their performance investigated in response to six different ways in which the overall krill biomass in a 600 n. mile square oceanic sector might decline by 50%. In most cases there is essentially no response of the index, or a response rather smaller in relative magnitude than the biomass decline. Catch statistics collected at present (centred primarily on catch per fishing time) are of low utility in detecting biomass declines. Indices incorporating search time information may give improved performances, but would probably be reliable only for detecting changes in within-concentration krill distribution parameters. Other approaches (e.g. research vessel surveys) need to be considered to monitor changes in the number, distribution and density of krill concentrations. Priority for further analyses should be given to examination of existing scientific krill survey data to improve the krill distribution model underlying this analysis (this might perhaps be appropriately effected by constituting and organising a meeting of a small joint CCAMLR-BIOMASS Working Group). This is because of a major discrepancy between the typical length of Japanese tows (generally through a single swarm) which is much larger than the typical size of krill swarms recorded in scientific surveys; this discrepancy needs to be resolved before greater confidence will warrant being placed in the results from a simulation model of the fishery.

Japanese krill fishing operation was observed on board a commercial vessel and actual operation was described in detail. Fish finder was most useful in detecting swarms while sonar was used to judge the towing direction. Size of fished swarms ranged from 100 m to 10 km with different distribution in different fishing areas. While catch per towing time reflects within-swarm density catch per day was thought to be a good index to measure abundance of concentration. Comparison of such index across time and space may have difficulties in standardization when data from different vessels are to be used.