Mali Matilda, Dell'Anna Maria Michela, Notarnicola Michele, Damiani Leonardo, Mastrorilli Piero
DICATECh, Politecnico di Bari, via Orabona, 4, I-70125, Bari, Italy.
DICATECh, Politecnico di Bari, via Orabona, 4, I-70125, Bari, Italy.
Chemosphere. 2017 Oct;184:784-794. doi: 10.1016/j.chemosphere.2017.06.028. Epub 2017 Jun 13.
Almost all marine coastal ecosystems possess complex structural and dynamic characteristics, which are influenced by anthropogenic causes and natural processes as well. Revealing the impact of sources and factors controlling the spatial distributions of contaminants within highly polluted areas is a fundamental propaedeutic step of their quality evaluation. Combination of different pattern recognition techniques, applied to one of the most polluted Mediterranean coastal basin, resulted in a more reliable hazard assessment. PCA/CA and factorial ANOVA were exploited as complementary techniques for apprehending the impact of multi-sources and multi-factors acting simultaneously and leading to similarities or differences in the spatial contamination pattern. The combination of PCA/CA and factorial ANOVA allowed, on one hand to determine the main processes and factors controlling the contamination trend within different layers and different basins, and, on the other hand, to ascertain possible synergistic effects. This approach showed the significance of a spatially representative overview given by the combination of PCA-CA/ANOVA in inferring the historical anthropogenic sources loading on the area.
几乎所有的海洋沿岸生态系统都具有复杂的结构和动态特征，这些特征同样受到人为因素和自然过程的影响。揭示污染源和控制高污染区域内污染物空间分布的因素的影响，是对其进行质量评估的一个基本预备步骤。将不同的模式识别技术应用于污染最严重的地中海沿岸流域之一，可得出更可靠的危害评估结果。主成分分析/对应分析（PCA/CA）和因子方差分析被用作互补技术，以了解多源和多因素同时作用并导致空间污染模式异同的影响。PCA/CA和因子方差分析的结合，一方面能够确定控制不同层和不同流域内污染趋势的主要过程和因素，另一方面能够确定可能的协同效应。这种方法表明，PCA-CA/ANOVA组合给出的具有空间代表性的概述对于推断该地区历史人为源负荷具有重要意义。
