| 其他摘要 | Damming is one of the most important human activities affecting the river environment. The flow velocity of the river will slow down and nutrient loading will be greater due to the construction of the dam, which is suitable for the growth of phytoplankton. The original river will be an impounded river due to the construction of cascade hydropower and the living conditions of phytoplankton is changed. So we can evaluate the ecological effects of the cascade hydropower development by phytoplankton’s sensitivity to the change of conditions. Since the beginning of 1990s, some researchers use artificial neural networks (ANNs) to simulate the phytoplankton dynamics successfully in the rivers or lakes. But these researches always apply to simulate the dynamics of one or more particular species in single rivers or lakes. Therefore, this paper tries to simulate the dynamics of dominant phytoplankton groups in a whole river-reservoir system. The Wujiang River is a typical dammed river in south-western China and we have done works for years. Six sampling sites were selected in this study, they are the Liuchong River (respecting the original eiver ), the Hongjiadu Reservoir (10 years old), the Dongfeng Reservoir (22 years old), the Liuguang River ( the river affected by damming), the Wujiangdu Reservoir (40 years old) and the Hongfenghu Reservoir (52 years old). Water samples were semimonthly collected from May 2011 to May 2012. Phytoplankton community composition, cell density and relevant physical, chemical parameters were analyzed. Then BPANNs models were constructed and validated base on the data. Finally, the sensitivity analyses were done to discern factors influencing phytoplankton dynamics. Several important conclusions, which can provide a scientific basis for effects on phytoplankton dynamics of the construction of cascade hydropower, have been drawn as follows: 1. In the study areas there were three main algal groups: Chlorophyta, Bacillariophyta, and Cyanophyta. In the Liuchong River, the dominant group was Bacillariophyta, which accounted for 67.2% of phytoplankton total cell density (1.12×106cells/L), and the cell density reached maximum value in August. In the Liuguang River, Bacillariophyta was still the dominant group, but proportion fell to 56.4% of phytoplankton total cell density (0.99×106cells/L). Phytoplankton community composition in the Liuguang River has been changed compared to the original river. Chlorophyta was the dominant group in the reservoirs and that in each reservoir was different due to the ages of the reservoirs. The phytoplankton total cell density was larger with the reservoirs got elder. The phytoplankton total cell density in Hongjiadu Reservoir was 2.13×106cells/L and the dominant group were Chlorophyta (36%) and Bacillariophyta (36%), while that in Hongfenghu Reservoir was 34.88×106cells/L and the dominant group was Chlorophyta (94%) 2. Correlation analysis showed the phytoplankton cell density was significantly and positively correlated with the dissolved silicon (R = 0.639; P < 0.001) in the original Liuchong River. It indicated that dissolved silicon was the restrictive factor for the growth of phytoplankton. In the Hongfenghu Reservoir, phytoplankton cell density was significantly and positively correlated with pH (R = 0.639; P < 0.001). In the other reservoirs, phytoplankton cell density was significantly and positively correlated with the water temperature, which showed that water temperature was the restrictive factor for the growth of phytoplankton. 3. In this study, water temperature, dissolved oxygen, dissolved CO2, total nitrogen, total phosphorus and dissolved silicon were selected as the input parameters, and the phytoplankton cell density was selected as the output parameter. Then the BPANNs models were constructed to simulate the phytoplankton dynamics. The results that the predicted values has well correlation with the measured values showed the models can predict time-series variation of phytoplankton abundance in such complex and nonlinear phenomena in the river-rese |
修改评论