A GIS-based habitat suitability model for commercial yield estimation of Tapes philippinarum in a Mediterranean coastal lagoon (Sacca di Goro, Italy)

Habitat suitability models have been extensively used by conservation planners to estimate the likelihood of occurrence and abundance of threatened wildlife species in terrestrial ecosystems, but they have been rarely applied to aquatic environment. In this study a GIS-based habitat suitability (HS) model has been developed to identify suitable sites for Manila clam (Tapes philippinarum) farming in the Sacca di Goro lagoon and to estimate the expected commercial yield. Habitat suitability and the assessment of potential yield for each specific site of the lagoon are derived on the basis of information on six exogenous parameters, namely sediment type, dissolved oxygen, salinity, hydrodynamism, water depth and chlorophyll “a”. For each of them, a parameter-specific non-linear suitability function has been used to transform parameter values into a quality index normalized between zero (i.e., habitat non suitable to clam rearing) and 1 (habitat best suitable with respect to that specific parameter). Habitat suitability of a specific site within the lagoon is then computed as a weighted geometric mean of the quality indices corresponding to the six measured parameters.Weight has been set according to expert knowledge. Finally, a scaling function derived from field observations is used to transformHSvalues into estimates of annual potential yield (kgm−2 year−1). The model has been applied to the Sacca di Goro lagoon where data on actual yield and the main biogeochemical and hydrodynamic parameters have been gathered in 15 sampling sites. These point data have been interpolated by Kriging so as to derive full maps of each biochemical parameter for the whole lagoon. TheHSmodel has been then applied to derive the thematic maps of suitable sites for clam rearing and the corresponding expected yield. In order to assess the sensitivity of model predictions to errors in weights estimation, we performed a MonteCarlo analysis considering three different assumptions on ranges of variation of weights values. The model has proved to provide managers with sound estimates on yield potential of the different sites of Goro lagoon where T. philippinarum is commercially exploited.We thus claim that this reasonably rapid and cost-effective approach poses the basis for a fair partition of harvesting concessions among competitive users and for a remarkable improvement of transparency in the decision-making process