Statistical analysis of nutrient data quality (nitrate and phosphate), applied to useful predictor models in the northwestern Mediterranean Sea

A dataset comprising 12 oceanographic cruises (from 1979 to 1983), covering two regions in the Northwestern Mediterranean Sea (Gulf of Lyon and Catalan Sea), was statistically analysed for nitrate and phosphate relationships with silicate, salinity and depth. This analysis provided a preliminary assessment of the data’s quality, as well as a predictor model of these nutrients below the surface layer. Results from the statistical analysis showed no significant difference (P = 0.05), between the regression equations (nitrate-depth and phosphate-salinity) of most cruises in the Gulf of Lyon, indicating that these relationships do not change in this area, particularly in sum-mer and autumn. A significant relationship between nitrate and phosphate with salinity and depth (multiple regression) was observed, suggesting that nitrate and phosphate distribution in the intermediate level are significantly related to the mixture of the water masses and the degra-dation of organic matter. The phosphate data showed a wide variance and a bias, probably due to procedural problems in the chemical analysis. Below the Intermediate Levantine Water, results from the ANOVA showed no significant variation of the phosphate and nitrate concentrations in the water column. However, a spatial and temporal variation was observed in this level