摘要

O fator de condi o de peixes, expressa a rela o peso-comprimento e 谷 aceita como um indicador da avalia o do ※bem-estar§ do animal; quanto maior o peso, melhor deve ser sua condi o de sobreviv那ncia. Este trabalho objetivou ajustar um modelo Gama, da fam赤liaexponencial com fun o de liga o pot那ncia, a dados de medidas de comprimento e peso de peixes da esp谷cie Trachydoras paraguayensis, observados na bacia do Rio Paran芍, Paran芍-Brasil. As estimativas dos parametros do modelo foram obtidos por dois processos distintos: pelom谷todo cl芍ssico da m芍xima verossimilhan a e pelo m谷todo Bayesiano (MCMC). Observou-se que todas as estimativas Bayesianas (m谷dias 角 posteriori) para os parametros de loca o foram similares as estimativas cl芍ssicas, contudo o parametro de dispers o 谷 inferior ao obtido pela a estimativa cl芍ssica, al谷m de fornecer um erro padr o 30% menor. The factor of fish condition expresses the weight-length relationship and it is accepted as an indicator for the evaluation of the animal %26quot;well-being%26quot;; as larger the weight, best should be your survival condition. This work aims to fit a Gamma model, from the exponential family with power link function, to data of length measurements and weight of fish from the Trachydoras paraguayensis species, observed in the basin of the Paran芍 River, Paran芍-Brazil. The estimates of the parameters of the model were obtained by two different processes: by a classic method (maximum likelihood) and by Bayesian method (MCMC). It was observed that the Bayesians estimates (posterior mean) of the parameters of interest were similar to the classic estimates, the dispersion parameter, it underestimated the classic estimate, however the dispersion parameter is inferior to the obtained by the classic estimate, besides supplying a standard mistake 30% smaller.

全文