Catchability of the southern rock lobster, Jasus edwardsii

Catchability is estimated indirectly as a 'nuisance' variable in the spatially explicit stock assessment model of the southern rock lobsters Jasus edwardsii in Tasmania, Australia. This study attempted to identify the key mechanisms influencing catchability to enable direct independent estimates of monthly catchability. Seasonal variation in catchability of the southern rock lobster Jasus edwardsii was estimated in a scientific reserve in southeast Tasmania by comparing estimates of lobster density based on direct visual observations underwater with concomitant estimates from trapping surveys. Underwater density estimates of undersized and legal-sized male and female lobsters greater than 80 mm carapace length did not change significantly over the 14 month study period with the exception of undersized males (smaller than 110 mm carapace length). Sex ratios remained constant at approximately 1:1. In marked contrast, catch rates of males and females and the sex ratio of trapped lobsters varied strongly with season, implying that catchability varies seasonally and with sex. Impact of capture on subsequent catchability appeared to be weak, since the ratios of tagged animals in the population observed underwater generally reflected recapture rates of tagged animals in trap catches. Size-specific catchability generally increased with size, but also varied with sex and season. During moulting and mating, size-specific catchability and relative selectivity did not increase, and sometimes decreased for larger animals. The size-frequency distributions of lobsters captured in traps therefore rarely reflected the size-frequency distribution of the population on the ground. Negative associations between small and large lobsters in traps were stronger in winter than in summer, indicating strong behavioural interactions. These interactions can account for the lower catchability of smaller lobsters. Relative selectivity estimates using tag-recapture and size-specific catchability data provided similar results. Seasonal variation in catchability of legal-sized males and females in the scientific reserve was described by modelling the effects of water temperature, moulting and mating. Seasonal changes in water temperature described 63% of the variation of catchability for males, but were a poor predictor of catchability for females outside winter. Both moulting and mating were highly synchronised, although males and females moulted at different times of the year. Gaussian probability density functions were used to represent the timing and intensity of moulting, mating and subsequent compensation periods, and were added to the description of seasonal temperature changes. Four Gaussian functions based on independent biological data considerably improved the model fits for the catchability of males (R2 = 0.83). However, adding a single Gaussian function to the temperature model, representing a combined moulting and mating period based on independent biological data, provided a less adequate description of the variation in catchability of females (R2 = 0.49). Only models unconstrained by the observed timing of these events provided a good fit (R 2 = 0.74). The seasonal catchability models developed for the reserve population were applied to catchability over several years in two commercially fished regions of Tasmania. Catchability was estimated using commercial catch and effort data and fishery-independent estimates of exploitation rates. The seasonal catchability models suggest that similar environmental and physiological processes were the main factors determining seasonal catchability in the two fishery regions, but these factors varied considerably in their relative importance between the two regions. Interannual variation in relative catchability was correlated with density-dependent processes. Full models described 72% of the overall variation in catchability over 6 years in the south and 80% of the variation over 4 years in the north. More work is required before direct estimates of catchability can be included in stock assessment models. In particular, region-specific patterns of seasonal catchability, and the relationship between density-dependent processes and the interarmual variation in catchability need to be determined, before catchability can be reliably predicted in future years and in other regions of Tasmania. Nevertheless, this work has greatly improved our understanding of the processes that apparently underpin seasonal catchability.