CPUE standardization of black marlin (Makaira indica) caught by Taiwanese large-scale longline fishery in the Indian Ocean

In this study, the delta-linear models with different assumptions of error distribution were adopted to conduct the CPUE standardization of black marlin caught by the Taiwanese large-scale longline fishery in the Indian Ocean for 1979-2020 and 2005-2020. The groups of data sets derived from cluster analysis based on species compositions were incorporated in the models as a covariate for explaining the target. The results indicate that the targeting effects (clusters) provided most significant contributions to the explanation of the variance of CPUE for the models with positive catches, while the catch probability might be mainly influenced by the position of fishing operations. The standardized CPUE series obtained from different model assumptions revealed quite similar trends for all model except for delta-lognormal model. For 1979-2020, CPUE trends were similar for the northern areas (NW and NE) and they fluctuated before early 1990s, gradually declined until late 2000s, increased until mid-2010s, then substantially decreased again, and reveals an increasing trend in recent years. For 2005-2020, the trends of CPUE for the northern areas (NW and NE) also revealed similar patterns, CPUE increased from 2013 to 2016, decreased until 2018, and increased in recent years.