Applying a non-parametric bootstrap technique to assess the accuracy of biomass estimates of Argentine Anchovy (Engraulis anchoita) Northern stock (34° S - 41° S) with the Daily Egg Production Method. 1993-2008 Period

Abstract

Engraulis anchoita Northern stock reproductive biomass (SSB) during the 1993-2008 period was estimated applying the Daily Egg Production Method (DEPM) and accuracy of estimates assessed using a bootstrap technique. The basic data were collected in different research surveys carried out in late Spring, at the peak of the spawning season. The aim of this work was to build reliable confidence intervals around the estimates of the mean reproductive biomass applying a non-parametric bootstrap method. Resampling methods are becoming increasingly popular as statistical tools as they are robust and do not require the assumption of any particular distribution of the basic data. This is especially relevant when assessing the DEPM reproductive variables that have an a priori unknown distribution. In the SSB estimates corresponding to the 1993-2008 period, of 0.67-3.67 million tons, coefficients of variation that ranged 22-73% were observed. At each loop of the bootstrap an embryonic mortality rate value was calculated fitting an exponential decay model to eggs data derived from randomly chosen plankton stations. Likewise, the reproductive parameters were estimated based on fish from fishing trawls randomly extracted. The SSB estimates calculated with the bootstrap technique and the traditional DEPM showed similar results, with a 0.8-1.4 ratio range.