The application of general linear modelling methods to estimate trends in abundance of the hake and rock lobster stocks off South Africa

• Main Author: Glazer, Jean Patricia
• Other Authors: Butterworth, Doug S
• Format: Dissertation
• Language: English
• Published: University of Cape Town 2016
• Subjects: Mathematics and Applied Mathematics
• Online Access: http://hdl.handle.net/11427/16395

Bibliography: pages 162-178. === The two species of Cape hake, Merluccius capensis (shallow-water hake) and M paradoxus (deepwater hake), fomi the mainstay of the bottom trawl industry off South Africa and constitute the country's most valuable fishery. It is therefore important that the status of this resource be assessed regularly to ensure that exploitation is at a sustainable level. The two Cape hake species are morphologically similar and no distinction is made between them in commercial catch statistics. Consequently, for assessment purposes, the Cape hakes are treated as a single species. It is assumed that two stocks of Cape hake exist, one off the West Coast and another off the South Coast of South Africa. Central to the assessments of these stocks are the catch per unit effort (CPUE) data because it assumed that CPUE is proportional to abundance. The nominal CPUE (hake catch divided by actual time trawled) for both the West and South Coast stocks has shown a steady growth over the period 1978 - 1994, increasing at a rate of 3.8% per annum on the West Coast and 4.2% per annum on the South Coast. The bulk of this thesis is concerned with determining whether these increases in CPUE are the result of an increasing biomass, or are rather, in part, the result of improved vessel efficiency due to technological advancement or of changes in fishing strategy. The existing CPUE time series had previously been standardised by means of applying power factors which were crudely estimated in the early 1970s and which are likely inappropriate for the current fishing fleets. These CPUE series have therefore been re-standardised by applying the internationally accepted approach of General Linear Modelling (GLM).