An assessment of abundance trends and biology of langoustines (Metanephrops mozambicus) and pink prawns (Haliporoides triarthrus) from the deep-water trawl fishery off eastern South Africa.

Abstract

Deep-water trawling (>200 m deep) for crustaceans in the South West Indian Ocean (SWIO) yields catches of several species, including prawns (Haliporoides triarthrus, Aristaeomorpha foliacea, Aristeus antennatus and Aristeus virilis), langoustine (Metanephrops mozambicus), spiny lobster (Palinurus delagoae) and geryonid crab (Chaceon macphersoni). Infrequent deep-water trawling takes place off Tanzania, Kenya and Madagascar; however, well-established fisheries operate off Mozambique and South Africa. Regular trawling off South Africa started in the 1970’s, mainly targeting M .mozambicus and H. triarthrus. Catch and effort data for the South African fishery were regularly recorded in skipper logbooks over a 23 year period (1988 – 2010); this database was obtained from the Department of Agriculture, Forestry and Fisheries (DAFF) in order to assess abundance trends of M. mozambicus and H. triarthrus. Generalized linear models (GLM) were used to quantify the effects of year, month, depth and vessel on catch per unit effort (CPUE). By year, the standardized CPUE of M. mozambicus increased, and three factors (or a combination of them) could explain the trend: reduced effort saturation, improved gear and technology, or an increase in abundance. By month, CPUE peaked in July and was highest between depths of 300 and 399 m. The standardized CPUE of H. triarthrus fluctuated more by year than for M. mozambicus, possibly because it is a shorter-lived and faster growing species. The monthly CPUE peaked in March, and was highest between depths of 400 and 499 m. Totals of 2 033 M. mozambicus (1 041 males and 992 females) and 5 927 H. triarthrus (2 938 males and 2 989 females) were sampled at sea between December 2010 and March 2012, during quarterly trips on-board a fishing trawler. A GLM framework was used to explore their size composition, sex ratio variability, size at maturity and reproductive cycles. Male and female M. mozambicus size distributions were similar, but varied by month and decreased as depth increased. Female H. triarthrus were significant larger than males; size structure varied by month, but showed no change over depth. The sex ratio of M. mozambicus favoured males (1 : 0.89), but was close to parity in all months, except November when males predominated. H. triarthrus exhibited parity (1 : 1.002) with no significant variations in sex ratios by month. The proportion of egg-bearing M. mozambicus in the population declined between March and August (hatching period) and then increased until December (spawning period). The L₅₀ (length at 50% maturity) of M. mozambicus was estimated to be 49.4 mm carapace length (CL), and the smallest and largest observed egg-bearing females were 33.5 and 68.6 mm, respectively. No reproductively active female H. triarthrus were recorded during the sampling period. Growth parameter estimates for M. mozambicus (male and female combined) using Fabens method were K = 0.48 yearˉ¹ and L∞ = 76.4 mm CL. Estimates for the von Bertalanffy growth formula (VBGF) were: K = 0.45 yearˉ¹ and L∞ = 76.4 mm CL. H. triarthrus male and female growth parameter were estimated separately. For males they were K = 0.5 yearˉ¹ and L∞ = 46.6 mm CL using Fabens method, and K = 0.76 yearˉ¹ and L∞ = 46.6 mm CL using the VBGF. For females they were K = 0.3 yearˉ¹ and L∞ = 62.9 mm CL using Fabens method, and K = 0.47 yearˉ¹ and L∞ = 62.9 mm CL using the VBGF. CL to total weight regressions were calculated for both species; no significant differences were found between male and female M. mozambicus, although H. triarthrus females became larger and heavier than males. Comparisons with three earlier studies (Berry, 1969; Berry et al., 1975; Tomalin et al., 1997) revealed no major changes in the biology of either species off eastern South Africa. Stocks appear to be stable at current levels of fishing pressure, although some factors are not yet fully understood. Disturbance caused by continual trawling over a spatially limited fishing ground may affect distribution and abundance patterns, especially in M. mozambicus, which was less abundant in the depth range trawled most frequently. The absence of reproductive H. triarthrus in samples suggests that they occur elsewhere, and there is some evidence of a possible spawning migration northwards to Mozambique; this suggests that H. triarthrus is a shared stock between South Africa and Mozambique. The results from this thesis will add to the knowledge of M. mozambicus and H. triarthrus in the SWIO, and provide a basis for developing sustainable management strategies for the deep-water crustacean trawl fishery off eastern South Africa.