A sampling plan for phycotoxins surveillance in bivalve mollusc farms along the Santa Catarina coast, Brazil

Bivalve molluscs are filter-feeders and have a wide range of phytoplankton species as main food source. Some of these species are toxic, producing the so-called phycotoxins, that can get concentrated in edible parts of mussels, oysters, scallops, and clams, causing severe intoxication syndromes in consumers. Since 2008, harmful algae and phycotoxins occurrence related to three shellfish poisoning syndromes: Diarrhetic (DSP), Amnesic (ASP), and Paralytic (PSP) shellfish poisonings have been monitored in Santa Catarina State, Brazil. The objectives of this study were: (1) to identify representative sampling areas from which shellfish samples should be collected each round of the monitoring cycle and (2) to define an alternative sampling strategy for phycotoxins detection that requires less samples and/or smaller pool sizes according to different laboratorial methods officially recognized. Using geographic information system, we designed 24 sampling areas. To calculate sample sizes for phycotoxins detection in mollusc soft tissues, we simulated six scenarios with different values of prevalence and test sensitivity. Considering High-Performance Liquid Chromatography with Diode-Array Detection (HPLC-DAD) and mice bioassays, the most realist scenario was the one of ≥20% of prevalence and highly sensitive tests, which resulted in one pool of 20 Perna perna mussels each to detect ASP toxins, two pools of 15 to detect PSP, and two pools of 30 to detect Lipophilic Toxins (DSP + Yessotoxin). With the use of liquid chromatography with mass spectrometry (LC-MS/MS) analysis, only one pool of 15 mussels would be enough for target phycotoxins detection. The strategy of sampling using the defined areas associated with LC-MS/MS analysis requires less samples and a smaller pool size without losing area representativeness and surveillance system sensitivity.