The first part of this document contains an analysis of data pertaining to the Soviet krill fleet. The data base consists of the records of 12 different cruises by 8 different research vessels between 1981 and 1984. The data are analyzed according to operational characteristics of the fishing process such as trawl duration, krill catch, or between trawl movement. Correlation analyses are presented as a means of understanding the within trawl and between trawl features of the operation. The data support the notion of a “patches within patches” model for the distribution of krill in the southern oceans.

The second part of this document contains the development and use of a simulation model of a Southern Ocean krill fleet. The objective of the work is to answer questions such as: what information do catch and effort data provide about krill abundance or how easily can significant changes in krill biomass be detected? The krill distributional model begins with individual krill which are assumed to aggregate into swarms of krill. The swarms then aggregate into concentrations, which are the foci for the fishing operation. Parameters of the model are motivated by study of the literature and FIBEX results. A model is developed for a survey vessel that does no fishing, but simply locates concentrations of krill for the fishing fleet. The fishery model involves finding concentrations, finding swarms within concentrations and fishing individual swarms. Wherever possible, operational data from Part I are used to provide distributions in Part II. General considerations about the theory of abundance indices for pelagic, schooling species are discussed. In particular, the importance of the time spent searching for swarms is highlighted. A theory for detecting changes in krill biomass is developed. Forty-four different abundance indices are considered and their effectiveness in detecting changes in krill biomass is studied. The best indices involve two separate measures: one in which survey vessel discoveries are used to track the number of concentrations and a measure of the form catch/swarm/search-time to track swarm density within concentrations and krill density within swarms. Operational recommendations are given: (i) I propose an experiment in which survey and fishing vessels operate simultaneously but independently in the same region, (ii) I recommend that fishing vessels begin to indicate in their log books the amount of between trawl time spent searching, (iii) I propose that CCAMLR consider sending a Ph.D. level modeller to sea in order to develop a truly operational model of the fishing process, and (iv) I propose abundance indices that could be used to track krill biomass.