INDEXES OF LARVAL BLUEFIN TUNA, THUNNUS-THYNNUS, ABUNDANCE IN THE GULF-OF-MEXICO - MODELING VARIABILITY IN GROWTH, MORTALITY, AND GEAR SELECTIVITY

Most indices of stock size for Atlantic bluefin tuna are fishery depen dent, and thus do not benefit from statistical design. Alternatively, we derived a fishery independent index of western Atlantic stock size from ichthyoplankton surveys conducted by the National Marine Fisherie s Service. The larval abundance indices developed from these surveys h ave been used to corroborate trends in fishery dependent estimates of stock size, as well as to tune the virtual population analysis. Estima tes of average annual larval abundance at first daily increment format ion per 100 m(2) sampled by oblique bongo tows were used to index tota l annual larval abundance. A model describing the observed mean trend in larva size at otolith daily increment count was developed to estima te a probability of age at length matrix for ageing the captured larva e. Daily loss rates (Z) were estimated through regression analysis of the larval catch curves. Effects of mesh size changes (333 mu m or 505 mu m) during the time series of sampling were incorporated into the e stimates. Uncertainty in the index values for the various components o f the estimator is incorporated through the delta method. Zero catch i nformation from the sampling is incorporated through application of Pe nnington's transform. Finally, we tested several different methods of calculating the index to evaluate the sensitivity of the results to di fferent assumptions.