Correlation Analysis Based on Grid Regression Model: A Case Study in Beibu Gulf

Using the data from the 4 seasonal survey cruises in Beibu Gulf carried out from 1998-1999, this study analyzed the taxonomic composition, the spatial distribution of the abundance and diversity of phytoplankton as well as the relationship between the abundance and the nutrients such as dissolved inorganic nitrate (DIN), dissolved phosphate (IP) and silicate (Si) by using the methods of GIS interpolation and grid regression analysis model. 382 species of phytoplankton were identified in the study with most of them being diatoms. The abundances in all seasons showed the same spatial distribution with the dense areas focusing on the northern coastal waters of the gulf, while in the southern part and from the gulf mouth to the open sea, the abundances were quite lower. The results of the grid regression analyses on the abundance of phytoplankton with the nutrients showed that in autumn there existed a weak positive relationship with DIN (R2=0.270), while in summer and winter it showed the weak negative relationship with IP (R2=0.304 in summer and 0.327 in winter), and also in summer a medium positive relationship with the variation of Si was found (R2=0.478). The N:P ratio in most of the gulf ranged from 0-15 with the average of about 13.8, and the analysis showed that the impact by the N:P ratio on phytoplankton caused the variation of diversity rather than the abundance.