| Summary: | 碩士 === 國立臺灣海洋大學 === 環境生物與漁業科學學系 === 93 === In order to probe into the effect of industrial sewage on the distribution of zooplankton in the coastal waters, this thesis was carried out around Da-Lin-Pu waters, Kaohsiung. Some statistical methods, such as cluster analysis and Principle component analysis (PCA), were applied to explain the relationship between industrial sewage and zooplankton. In cluster analysis, we can find that analysis method is not applicable to the data of zooplankton, the reason is that coefficient of variation of Copepoda is excessively large to make divide into four groups by Copepoda only. It is possible to lead to neglect the coefficient of variations of other zooplanktons. PCA can reduce and find out the important independent variable, and not reduce the ability of estimate for dependent variable. So this study uses PCA to probe into the effect of industrial sewage on the distribution of zooplankton in the coastal waters. Results of this study were as follow.
1. In PCA, it divided that sea area into A group belonging to coastal waters, B group belonging to industrial sewage discharging area, and C group belonging to the outer sea area.
2. In PCA, we can find that alkalinity of A group is higher, and A group has higher temperatures, and higher copper and zinc.
3. B group of water area contains more organic pollution, such as the cyanide, phenol, selenium from industrial sewage.
4. C group of water area has the highest amount of dissolved oxygen, it shows that its water quality is comparatively pure.
5. On the other hand, the total of zooplankton is getting the least along A group. This phenomenon in A group may be caused by relatively high temperature, relatively alkaline water quality, copper and zinc heavy metal and oil pollution, etc..
Keywords: Principle component analysis, Zooplankton distribution, Industrial sewage, Kaohsiung waters
|
|---|
