Applying the nonparametric bootstrap technique to estimate the Patagonian hake (Merluccius hubbsi) biomass using the Daily Egg Production Method

PÁJARO, M.; HANSEN, J.; MACCHI, G. J.; EHRLICH, M.; MACHINANDIARENA, L.; CADAVEIRA, M.

Vigo

Conferencia; Fish Reproduction and Fisheries Conference; 2011

Instituto Investigaciones Marinas (CSIC)

Argentine hake (Merluccius hubbsi) is the most important fish resource in Argentina, where landings of the species were 277,000 t in 2010. The Patagonian hake stock inhabits a wide latitudinal area in the Argentine Sea, south of 41º S. During the peak of its spawning season, from December to January, fish are aggregated within an area of about 20,000 nm2. A cruise for the assessment of the spawning biomass of that stock by the swept area method was carried out in January 2005. For the first time, the assessment was also performed by applying the Daily Egg Production Method (DEPM). The main aim of this work was estimating the mean biomass and building confidence intervals by a nonparametric bootstrapping. Resampling methods are becoming increasingly popular as statistical tools, as they are generally very robust. Argentine hake spawning mainly occurred in stratified waters, but eggs were placed along the water column. Therefore, it was difficult to determine the right depth of spawning, and consequently the temperature at which egg developed. For this reason, three scenarios were considered: eggs placed at 20 m depth (i.e., above the thermocline), at 50 m (below), and between 10 and 70 m depth. In the DEPM, biomass is derived from estimates of the daily egg production (Po). Po estimates, depending on the assumption made about the temperatures of development, ranged between 1,300 and 1,800 eggs*10 m-2 (CV: 36-53%). The mean proportion of females was 0.487 (CV: 8.6%), mean spawning frequency was 0.155 (CV: 19%) and mean relative fecundity was 435 (CV: 3.4%) oocytes per gram of fish. Spawning biomass in the three scenarios ranged 262,000 - 359,000 t (CV: 46.8-64.0%). The coefficients of variation were rather high, mainly reflecting imprecision about the estimation of both egg mortality rate and egg production. The above mentioned range of DEPM estimates of spawning stock biomass was somewhat lower that of the traditional, swept area method.