The influence of environmental variability on the catch of chokka, Loligo reynaudii, off the coast of South Africa

Abstract

Globally, cephalopod fisheries are being relied on more heavily due to the depletion of longer-lived teleost species. The South African chokka squid (Loligo reynaudii) fishery is a case in point. Although previously numerous squid were often caught as bycatch, the fishery has officially been in place since 1985. Since the inception of the chokka fishery in South Africa, several studies have investigated the relationship between environmental drivers and annual chokka squid catch, with varying degrees of success. Recently, in 2013, chokka squid catches hit a record low, prompting resurgence in the topic of the squid environment-catch relationship. This study was initiated in an attempt to provide a quantitative relationship between the chokka squid catch and environmental variability, and to build a predictive model that could be used in fisheries management strategies. Historical data were obtained from various sources and included the mean and standard deviation in ocean bottom temperature; the mean and standard deviation in sea surface temperature; the maximum and minimum as well as the variation in wind speed; the mean, predominant and standard deviation in wind direction; the mean and standard deviation in atmospheric pressure; the mean chlorophyll concentration; the number of upwelling events; the hours of easterly winds blowing per day; and two large variation-in-climate indices, namely, the oceanic Nino index and the Antarctic Oscillation index. The monthly catch data were also provided. These data were initially analysed for inter-annual and intra-annual cyclic trends and followed by analysis of the delay in response of catch to the environmental variables, anticipating some impact on the different stages of the chokka life cycle. These lagged data were incorporated into a negative binomial generalised linear model, as well as a generalised additive model, which revealed a strong relationship (r²=0.707) between the catch and environmental variability. The inclusion of all the parameters was necessary; however, the mean bottom temperature and the standard deviation in sea surface temperature were the only parameters that had a significant effect on the catch. These results were used to build a predictive model that indicated that, although the relationship was strong, the ability of the model to predict catch was weak, particularly from the year 2005 onwards.