| 英文摘要 |
Pacific saury fishery is one of the major deep sea fisheries in Taiwan. Taiwan is the 2nd-largest saury harvesting country, which production and value of the saury in 2009 were about 104,000 mt and 2.75 billion NTD, respectively. Recently, the ecosystem-based fisheries exploitation, assessment, and management have been increasingly adopted in the world. A good understanding of migratory geographical characteristics and fishery oceanography of marine fish and their relationship with the fish population density and distribution is thus critical. This project studies the saury resources in the north Pacific. In order to examine effects of various fishery oceanographic variables on the spatial variations of the migratory saury fishing stock, a generalized linear model (GLM) is developed and spatial mappings is applied with geographic information system (GIS) in this study. The questions to be addressed are: (1) evaluating effects of various fishery oceanographic variables on the CPUE variation and distribution of the saury stock with GLM, (2) understanding the spatial relationships between main fishery oceanographic variables and the CPUE variation and distribution of the saury stock with GIS, and (3) continuously comparing the spatial abundant and distribution of the saury stock in 2010 with historical data. Specifically, the hands-on job includes: (1) Dataset preparation: Saury fishery data and fishery oceanographic data are arranged into cellular units of per 0.5 geographic degree in a day. There are 8 fishery oceanographic variables, including water temperature, sea surface temperature, chlorophyll-a concentration, net primary production, mixed layer depth, southern oscillation index, ocean wind vector, and bathymetric depth. (2) GLM and GIS: Dependent variable is CPUE and independent variables are year, month, latitude, longitude, and the above 8 fishery oceanographic variables in GLM. Values of spatial contours for the main fishery oceanographic variables are estimated with Kriging grid method. Maps of spatial distributions of the saury CPUE and the main fishery oceanographic variables are created and plied with ArcGIS. (3) Historical comparison: The saury CPUE value and distribution in 2010 are compared with the historical data. Variations and trends of saury catches between 2010 and historical data are compared among the top 4 saury harvesting countries, Japan, Taiwan, South Korea, and Russia, in the world. |