摘要

Reliable catch statistics are essential for assessing fishing impacts on individual stocks. However, fisheries that capture a mixture of stocks or species for which catch statistics are not disaggregated pose a challenge. Nonetheless, catch composition can be inferred given information on fishing date and location and a prevalent role of season and habitat in structuring fish assemblage composition. Here, a harmonic regression model for multinomial data, intended to predict the species composition of catches based on season and depth, is developed using bottom-trawl survey data. Model development was motivated by the need to quantify catches of individual skate (Rajidae) species in fisheries for which landing and discard data are only reliable at the family level. The model was validated by applying it to flatfishes (Pleuronectidae), whose catches are generally reliably and consistently disaggregated by species. The predicted species composition of flatfish matched the composition observed in fishery catches well. The present approach should be applicable to other well-surveyed ecosystems where assemblage composition is structured by one or more key environmental variables of known spatial distribution.

  • 出版日期2013-2

全文