Diel effects on bottom-trawl survey catch rates of shallow- and deep-water Cape hakes Merluccius capensis and M. paradoxus off Namibia, using solar zenith angle

Diel patterns in survey trawl catches for the Cape hakes Merluccius capensis and M. paradoxus off Namibia were studied in order to examine the effect of diel bias on catchability, and its implication for survey abundance estimation and the consistency of the survey time-series. Catch rates (numbers per haul) by species and length from summer demersal biomass surveys conducted during the period 2002–2013 were used, together with a computation of the corresponding light-level data from which the solar zenith angles were obtained. Generalised additive models were fitted to assess the relationship between the catches and a number of explanatory variables. Significant covariates were zenith angle, depth and geographical position. The final models explained 78% and 59% of the variability in catch rates of M. capensis and M. paradoxus, respectively. For M. capensis, the response to zenith angle increased sharply for values above 100°, which represents the time between sunset and sunrise. For M. paradoxus there was a moderate increase in the response to zenith angle during the night. In cases where some fishing took place at night in shallow water, the survey results for M. capensis were more greatly affected than was the case for M. paradoxus, which is related to the different depth preference of the two species. Fishing in depths shallower than 400 m outside daylight hours should therefore be avoided in order to reduce bias and ensure consistency in abundance estimates from surveys.Keywords: Benguela Current system, consistency of survey indices, efficiency of bottom-trawl surveys, negative binomial GAM, transect survey desig