A Geostatistical Approach to Estimate South Atlantic Swordfish Abundance From Commercial Catch Data

Abstract

We estimated relative abundance of South Atlantic Swordfish (Xiphias gladius), a highly migratory marine species, from commercial fishing data using a three-step analysis protocol. First, we modeled catch rate based on variables affecting the catch rate to rule out effects of catchability-related factors. Next, we analyzed residuals to identify autocorrelation. Last, we used an area-weighed recursive algorithm that considered spatial autocorrelation for each year within the analysis period. Commercial South Atlantic swordfish catch data from the Brazilian pelagic longline fleet were analyzed. Swordfish accounted for the highest spatial dependence in distance in 2005 (456 km), 2008 (111 km), 2012 (80 km), and 2014 (443 km), as well as for the lowest in 2011 and 2013 (average = 15.45 km). A downward trend in swordfish abundance between 2010 and 2017 was detected several years earlier than conventional standardized indices, so our proposed index is an alternative, and potentially more accurate, index of swordfish population dynamics in the South Atlantic. Our findings highlight the need to integrate multiple approaches into assessments of the abundance of marine species.