关键词: 长江口, 提升回归树(BRT)模型, 广义加性模型(GAM), 鱼类群落多样性

Abstract: Dominant plant species were collected from the deepwater area and shallowwater area of Caohai Lake, a National Nature Reserve of China. The distribution of mercury and methyl mercury in aquatic plants were analyzed, and the risk of mercury pollution was evaluated. Theresults showed that the total mercury concentrations in those species ranged 5.18-40.06 ng·g^-1 with an average of 12.78 ng·g^-1, which was comparable to the findings from other naturalreserves. The mercury concentration was higher in stems and leaves ofAlternanthera philoxeroides, Najas marina, and Polygonum hydropiper. Methyl mercury concentrations in those species ranged from 0.04 to 2.47 ng·g^-1, with an average of 0.34 ng·g^-1. Methyl mercury concentrations in leaves and stems were positively correlated with total mercury in those tissues (r=0.38,n=46,P<0.01). The concentrations of total mercury and methyl mercury in plants from the shallowwater area were significantly higher than those from the deepwater area. Such a pattern was consistent with the distribution of methyl mercury in sediments, but was opposite to that of total mercury in sediments. Our results suggested that methylation and the bioavailability of mercury in sediments from the shallowwater area were higher than those from the deepwater area in Caohai Lake. The difference in organic matter concentrations of the sediments and the wetdry alternation (flooding recedingflooding) in the surrounding shallow marshes would associate with the relatively higher concentrations of mercury and methyl mercury in aquatic plants in the shallowwater area. The underlying mechanism is needed to be uncovered in further studies.

Key words: generalized additive model, fish community diversity, Yangtze River Estuary, boosted regression tree model