An investigation of spatial-temporal diel changes in Loligo reynaudii catch rates in the commercial squid jig fishery of South Africa

Botes, Peet

An investigation of spatial-temporal diel changes in Loligo reynaudii catch rates in the commercial squid jig fishery of South Africa

Thesis / Dissertation

2025

Publisher

University of Cape Town

Department

Department of Biological Sciences

Faculty

Faculty of Science

Abstract

Loligo reynaudii (chokka) squid has been jigged commercially on the South African South-southeast coast since 1982. Starting in 2019, a chokka-directed hydroacoustic survey programme has formed part of scientific efforts to best inform the management of this fishery, mainly focussed on surveying at daytime during October-November, on the main fishing grounds between Plettenberg Bay and Port Alfred. However, it has so far remained unclear whether this programme may be improved in terms of spatial-temporal focus, to better survey acoustically detectable aggregations of adult squid throughout the year. Considering that commercial jig catches similarly depend on the prevalence, intensity and duration of aggregations; and assuming standardised Catch Per Unit Effort (CPUE) to be a representative measure of local abundance and proportional representation of regional population distribution: this study employed statistical modelling to specifically investigate whether, on the commercial squid jig fishing grounds of South Africa, there is: (1) an effect of diel period on chokka CPUE; (2) seasonal variation in any effects of diel period on CPUE; and (3) spatial variation in any effects of diel period on CPUE between seasons and years. The data used were DFFE (South African Department of Forestry, Fisheries and the Environment) commercial session-level squid jig catch logbook records. A tweedie generalized additive model was applied following an iterative semi-inductive modelling approach, for an optimal account of data non-linearity and non-normality when estimating CPUE as a function of relevant space-time variables. The data were limited, inter alia, to 2006-2022 sessions conducted purely at day/night between 20-29°E. The final model explained 24.4 % deviance and all terms were statistically significant in their effect on the response (p<0.05). There were three main findings when estimating relevant terms' marginal effects. Firstly, chokka concentrations are greater at day than night. Secondly, there is clear seasonality in the aggregated proportion with a single greater daytime and lesser nighttime peak in annual spawning concentrations, around October-December. Thirdly, CPUEs involved varying spatial-temporal complexity while being greater at day than night across most of the fishing grounds for most of the average year. The main takeaway was a dismissal of any notions that acoustically surveyable, large aggregations of squid may likely be found consistently in areas and at times other than when-where acoustic efforts have typically been expended in late spring so far, on inshore spawning grounds targeted by the commercial jigging fleet. Recommendations for future research include accounting for session-level variations in relevant oceanographic parameters when modelling session-level jig CPUE.