Towards an optimal sampling strategy for Alexandrium catenella (Dinophyceae) benthic resting cysts

The study proposes methodological developments to optimize sampling strategy of resting cysts of Alexandrium catenella to estimate their abundance with a predefined error. This work also aims to provide information on spatial distribution of resting cysts in sediments. The distribution mode of A. catenella resting cysts related to the abundance variability was studied through sediment cores sampling on four different spatial scales and using Ludox CLX gradient density method. The quantification method underestimates by a factor of 2 the resting cysts abundance in one gram of sediment. Application of Taylor's power law allowed us to define a compromise between sampling effort and abundance estimation error. In the case of A. catenella resting cysts from Thau lagoon, the optimal sampling strategy consists of sampling 10 stations on a surface of 2 km(2) for a given coefficient of variability (C) of 15%, sampling 3 sediment cores at each station (C = 30%) and counting only one replicate by core (C = 18%). Results related to the application of Taylor's power law are closely dependent on resting cyst density and aggregation in a given sediment. In our area, A. catenella resting cysts are mainly observed in the upper 3 cm of sediment. Horizontally, their heterogeneity is lower on 10 cm(2) surface and tends to stabilize itself beyond a surface of 10 m(2). Each author has to carry out this pre-sampling effort for his own resting cysts-fonning species, in his own area, in order to increase accuracy of resting cyst mapping. (C) 2007 Elsevier B.V. All rights reserved.