No abstract

No abstract

No abstract

This work investigated sediment samples collected from Dapeng Bay and three neighboring rivers (Kaoping River, Tungkang River, and Lingbeng River) in southwestern Taiwan, Republic of China. Multivariate statistical analysis techniques, i.e., factor analysis, cluster analysis, and canonical discriminant analysis were used for the evaluation of spatial variations to determine the types of pollution and to identify pollutant sources from neighboring rivers. Factor analysis results showed that the most important latent factors in Dapeng Bay are soil texture, heavy metals, organic matter, and nutrients factors. Contour maps incorporating the factor scores showed heavy metals accumulate along the lakesides, especially on the southeastern banks of the lakes. A cluster analysis was performed using factor scores computed from these latent factors. We then classified these areas into five distinct classes using sampling stations, and we illustrate that in the three river classes, the sediment properties are influenced by industrial and domestic wastewater and agricultural activities (including livestock rearing and farm activities). However, in Dapeng Bay, the rivers were influenced more by complicated biogeochemical processes; these could be identified as a type of pollution. Canonical discriminant analysis illustrated that two constructed discriminant functions made a marked contribution to most of the discriminant variables, and the significant parameters of porosity and Cd, Cr, Al, and Pb content were combined as the "heavy metal factor". The recognition capacities of the two discriminant functions were 82.6% and 17.4%, respectively. It is also likely that the annual mean of the water exchange rate is insufficient (taking about 7 days to eliminate pollutants) and therefore has significantly influenced the carbon and nutrient biogeochemical processes and budgets in the semi-enclosed ecosystem. Thus, the sediment properties are not similar between the lagoon and the neighboring rivers. Our results yield useful information concerning estuary recovery and water resources management and may be applicable to other basins with similar characteristics that are experiencing similar coastal environmental issues.

No abstract