A simple multivariate technique to improve the design of a sampling strategy for age-based fishery monitoring

Abstract

This paper presents a novel application of multivariate statistics to assist in the design of a sampling strategy for age-based fishery monitoring. Non-metric multidimensional scaling (MDS) and analysis of similarities (ANOSIM) were used to assess spatial and temporal variation in age frequency distributions of sea mullet sampled from an estuarine commercial fishery in New South Wales (NSW), southeastern Australia. Age of fish ranged from 1 to 10 years. A similarity matrix of samples was generated by treating each age class as an individual variable. The age composition of samples was not significantly different between estuaries at scales of 50–150 km, but was often significantly different between estuaries at scales of several 100 km. Seasonal differences in age composition were limited. Results provide a guide to appropriate spatial and temporal levels of sampling for age-based monitoring of the fishery. Methods used in this study could be applied to other fisheries to determine appropriate sampling levels for length or age-based monitoring. Variations in age structure of sea mullet observed in this study also suggest the scales at which biological processes, such as larval dispersal, growth and migration, operate for this species and so will also assist in the design of future sampling to examine these processes in more detail.