Avaliação direta da abundância de recursos pesqueiros demersais: aplicabilidade, desempenho e custo benefício da implementação deste método sobre dados gerados por observadores a bordo da frota comercial de arrasto do sudeste e sul do Brasil

An accurate knowledge of the total fish stock biomass is crucial for supporting management measures for the sustainable use of these natural resources worldwide. This information has also been of use in assessing the direct impacts of climate changes on natural fish populations, and in estimating the level of degradation of the marine habitats submitted to anthropic action. Thus, scientific programs aimed at estimating, though direct methods, the biomass of valuable resources that are targets of fishing activity worldwide, have been nanced, year after year, by research bodies and organizations that promote the fishing industry at national and international levels. Due to the high costs involved, countries with relatively poor EEZs and/or smaller fishing industries can barely justify maintaining such programs on a routine basis, which can compromise the process of managing their shery resources. A potentially less costly solution is to use capture data generated by trained observers on board commercial shing trawlers. This procedure can lead to doubts regarding the fundamental concepts of sampling theory, particularly when dealing with an activity that has the sole purpose of maximize prots. This work therefore aims to identify the possible biases originating from different sampling designs, and proposes a model designed to represent the trends in the fishing fleet in an environment of virtual populations aimed at simulating real resources and targets of the commercial bottom trawl eet. Each sampling design was submitted to two expansion methods: one based on the classic method of sampling design, and the other based on the application of geostatistical models. The results of the virtual environment simulations showed that the global biomass estimates of each virtual population tended to be overestimated when using the classic method (average difference between real and estimated biomass > 20%), whereas the geostatistical models tended to correct the biases inherent to onboard sampling. The geostatistical models also demonstrated, on the whole, biomass estimates that were very close to the real volumes for each resource (average difference < 20%). Finally, the latter method was also applied to the real data collected onboard commercial trawlers, resulting in robust estimates for each of the three fish species studied in this phase of the work: Urophycis mystacea (2001 16,402 t/2002 23,816 t), Merluccius hubbsi (2001 267,712 t / 2002 233,107 t) and Lophius gastrophysus (2001 51,947 t/2002 16,988 t). The study also analyzed the cost-benefit ratio of carrying out abundance estimates based on scientic survey cruises, of shing boat yields, and of maintaining a program of onboard observers. These comparisons showed that the cost of maintaining a program of onboard observers (R$ 94,413.47) is up to 990% lower than the worst case scenario of carrying out a scientic survey cruise (R$ 935,185.01), while the best case scenario, based on the leasing of a shing vessel for this purpose, had a total cost of around R$ 365,151.25. The application of the method developed to the historical data, generated by commercial trawlers, could act as spatial and temporal reference points that could contribute to the processes of environmental, economic and social management of this shing activity in Brazil