Résumé
The common octopus (Octopus vulgaris) is an important fishery resource in Morocco. The most important stock of octopus in the Moroccan commercial fishing ground is the stock of Dakhla. Using two approaches for the stock assessment of octopus, Direct and Indirect approaches, the INRH have however concentrated many efforts to develop the first one whish is based on the trawl-scientific surveys. According to the results of the surveys carried out from 1998 to now, we evaluate this approach and show some of its advantages and some of its limits. The geostatistical methods are the main tools used in this study. Among the advantages, we present the results of the spatial analyses, the mapping by kriging and the variance estimation of the abundance estimates. In the other hand, among the limits, using simulations, we show the level of fluctuation inherent to the sampling process. According to the sampling fluctuations, we show the difficulty to distinguish between the fluctuations intrinsic to the variable and those due to the sampling design. In particular, the year-to-year fluctuations observed at the beginning of the time series can be interpreted as sampling variations rather than changes in the fish population. This issue and other perspectives are discussed.