碩士 / 國立臺灣海洋大學 / 環境生物與漁業科學學系 / 100 / The swordfish (Xiphias gladius) is one of the important commercial species of the Taiwanese longline (LL) fishery in the Indian Ocean. In this study, we collected the Taiwanese LL data and environment variables during the period of 1998-2008. The Generalized Additive Models (GAM) was used to explore the correlation between LL catch data of swordfish and oceanic environmental factors. We then use the predict modes of GAM to predict the fishing grounds of swordfish.
High nominal catch per unit effort (CPUE) areas are concentrated in the northwestern and southwestern Indian Ocean where are accounting for 66% and 20% of total catch, receptively. The major fishing season is in the second and third quarters. The results of statistics showed the CPUE were significantly correlated with all the temporal (year and month), spatial (longitude and latitude) and environmental variables (Sea surface temperature (SST), net primary production (NPP), sea surface height anomaly (SSHA), mixed layer depth (MLD), number of hook per basket (NHB)). The high CPUE is associated with 22-23℃ of SST, 200-400 mg C/m-2 d-1 of NPP, 0.6-0.7m of SSHA and around 100m of MLD. The predict models of GAM exhibit the best fishing grounds were located in the northwestern and southwestern Indian Ocean, too.
The variations of catches and distributions of swordfish in relation to climatic index in the Indian Ocean were investigated by Regression Analysis. It suggested that decreases in areas of optimal SST and NPP areas for swordfish during positive Indian Ocean Dipole events would decrease the CPUE in the western Indian Ocean, while an increase in optimal areas would result in an increased CPUE in negative Indian Ocean Dipole events, especially in the seas around the eastern Somalia and northern Madagascar.
Identifer | oai:union.ndltd.org:TW/100NTOU5451043 |
Date | January 2012 |
Creators | Zhao-Yang Chen, 陳昭仰 |
Contributors | Ming-Anne Lee, 李明安 |
Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
Language | zh-TW |
Detected Language | English |
Type | 學位論文 ; thesis |
Format | 98 |
Page generated in 0.0015 seconds