Comparing four nominal CPUEs indices of swordfish (Xiphias gladius) with longline observer data in the Indian Ocean

The catch per unit effort (CPUE) is an essential statistical indicator of the status of
stocks. In the longline fishery, because of the different statistical methods of catch and
fishing effort, there are many forms for calculating nominal CPUE. Using the swordfish
(Xiphias gladius) of Chinese tuna longline fishery in the Indian Ocean as an example,
we evaluate the performance of four nominal CPUEs of two effort forms (1000 hooks
and 10000 hours) and two catch forms (number and weight) combinations in CPUE
standardization. This study uses 2,305 sets recorded by Chinese observers from 2012
to 2019 and the Tweedie GAM model for comparison. The results show that the
explanatory variables of the best model for the four forms of nominal CPUE are the
same, including year, month, hook type, bait type, longitude, latitude, hooks between
floats (HBF), and sea surface temperature (SST). Those model's explanation rates are
50.1%-53.8%, and the four standardized annual CPUEs have a very similar trend. This
study suggests that the logbook of the tuna longline fishery should prioritize ensuring
accurate records of swordfish numbers and hook numbers each set because they are
easier to obtain and have higher credibility, particularly in developing countries with
limited conditions.