| 英文摘要 |
Gneralized additive models (GAMs) were applied to examine the relative influence of various factors on fishery performance of white marlin and blue marlin for the Tawianese longline fisheries operated in the Atlantic Ocean. Commercial fisheries data for the analysis consisted of a 5 year (2002-2006) time series. GAM analyses revealed the relative imiportance of the variables in expaining the variance in CPUE. For white marlin, by decreasing importance, the variables ranked:(1) SST, (2) Latitude, (3) SST frontal energy, (4) Change of SST frontal energy, (5) Change of SST, (6) longitude, and (7) Bathymetry. For Blue marlin, the variables ranked: (1) Latitude,(2) Change of SST frontal energy, (3) longitude, (4) SST, (5) Bathymetry, (6) Change of SST, and (7) SST frontal energy. |