Swept area estimates of Namibian orange roughy (Hoplostethusatlanticus) biomass were recently obtained from commercial catch rate data and applied to temporally stationary strata. This methodology resulted in enormous biomass estimates, especially in the first few years. However, analysis of spatial and temporal patterns in commercial catch rates indicates that there is considerable spatial and temporal heterogeneity in fish biomass and sampling coverage. Keeping the strata temporally stationary results in the extrapolation of estimated extremely high densities in very well-sampled areas to large poorly sampled areas in some years (especially the first few years). It does not appear that such extrapolations are justified and could result in large over-estimates of biomass in early years. This paper outlines and applies an alternative swept area method that uses non-stationary strata for each of the four Namibian orange roughy stocks. Three types of strata are identified, as before: one with very high densities of aggregating orange roughy, one with intermediate densities and another with relatively low densities. Strata were constructed separately for each year on each of the four major fishing grounds. We demonstrate the utility of the approach by showing how its results can be used to evaluate the plausibility of alternative hypotheses for a large apparent drop in aggregating mature biomass seen in the two other orange roughy abundance series. Resulting biomass estimates on each ground are all lower than those obtained previously and no longer show this catastrophic drop in relative biomass that could not be easily explained by catch removal alone. However, on all fishing grounds, the biomass estimates in, and area of, high-density strata were the highest in 1997. Moreover, the proportion of fish in high density to low density strata is also highest in 1997 and then dropped dramatically in the next 2 years. These patterns are not entirely inconsistent with the hypothesis that fishing on aggregations has tended to disturb and disperse them and reduce the catchability of aggregating orange roughy. The large aggregations in 1997, compared to earlier years, could also suggest that the intensity of aggregation on the fishing grounds may be intermittent and independent of previous years of fishing effort.
Kirchner, C.H.; McAllister, M.K. A new improved method to compute swept area estimates of biomass from commercial catch rate data: application to Namibian orange roughy (Hoplostethus atlanticus). Fisheries Research (2002) 56 (1) 69-88. [DOI: 10.1016/S0165-7836(01)00311-3]