Statistical Model for Predicting Fish Fingerling Abundance in the Na Thap River of Southern Thailand

Abstract

This study aimed to examine pattern variation and find out the association between fish fingerling abundance and season, location, standing crop and some water quality parameters. Data were collected monthly from 10 sampling sites along the Na Thap River from June 2005 to October 2015. Fish fingerling abundance was transformed using square root to maintain normality distribution. Factor analysis was applied to group 58 species of fish. Multiple regression model was used for investigating the association between fish fingerling and month, year, sampling site, standing crop and some water quality parameters. Fish fingerlings were classified by factor analysis into 3 interpretable factors: saltwater, freshwater and ubiquitous species. The results show that month, year, sampling site, standing crop, salinity, dissolved oxygen and transparency were statistically significant associated with fish fingerling abundance. During dry season, the saltwater fish fingerling showed significantly increased from January to May, whereas freshwater fish presented a maximum peak in rainy season from June to December. This finding confirmed that factor analysis and multiple regression analysis can be used for classifying and clustering fish fingerling abundance in established regulation measures for sustaining fish population management.