[Research on seasonal variation of self-purification ability for small shallow lakes in South Lake Taihu]

Seasonal variations of self-purification ability for small natural shallow lakes in South Lake Taihu were investigated. The results showed that seasonal difference of self-purification of permanganate index, total nitrogen (TN), ammonium (NH4(+)-N), nitrate (NO3(-)-N), total phosphorus (TP), chlorophyll (Chl-a) in small shallow lakes were remarkable. Effects of self-purification were better in spring and winter, and were worse in summer by NH4(+)-N and NO3(-)-N and in autumn by TP and Chl-a. Organic pollution was light, TN and TP pollution were seriously in four seasons. Concentrations of TN and TP brought a well condition to algae growth, and lakes were eutrification easily by the limiting factor of phosphorus. Concentrations of Chl-a were showed that lakes were eutrophic in summer or autumn and mesotrophic in winter or spring. Growth and blooms of phytoplankton impacted water quality and self-purification of lakes. Species and quantity of aquatic plants were the main factors to affect the change of pH and dissolved oxygen (DO), and loss of fertilizer and domestic wastewater were the main reasons for high nitrogen and phosphorus concentrations in lakes. Organic nitrogen was the main portion of TN by using excessive organic fertilizer in summer, NO3(-)-N was the main portion in other seasons. The principal component analysis result showed that the three principal components of self-purification ability were phytoplankton factor (water temperature, pH, permanganate index and Chl-a), farm drainage factor (pH, DO and TN), nutrient factor (TN and TP). The cluster analysis result showed that the water quality of four seasons in 11 sampling sites of three lakes could be divided into two categories: first, in spring, autumn and winter; second, in summer. This was caused by the temperature changes and agricultural drainage. Water temperature and pH were used to calculate the concentrations of permanganate index, TN, TP, Chl-a by linear equations, which improved the quick prediction ability in locale.