Variance and cost-benefit analyses to determine optimal duration of tows and levels of replication for sampling relative abundances of species using demersal trawling

Abstract The effects on catch per unit effort (CPUE) of trawlable species due to the duration of tows and temporal and spatial heterogeneity were examined using demersal trawling in the prawn grounds off the east coast of New South Wales, Australia. Standard trawls in a triple-rig configuration were used in a nested experimental design to compare catch rates in tows of various durations, during the day and night, in shallow and deep fishing grounds. Significant effects from all these factors were detected, but the kinds and degrees of these effects differed among variables. To determine the degree of spatial and short-term temporal variability in CPUE, a pilot survey was completed where replicate tows were done at various locations on replicate days (school prawn grounds) and nights (king prawn grounds). Cost-benefit analyses determined optimal CPUE estimates of species given (i) the time available to survey a location during any sample period, and (ii) the sizes of variances among tows and days/nights. The consequences of this replication on the sizes of standard errors in subsequent sampling were estimated. A uniform and optimal methodology was developed which is currently being used in surveys of the New South Wales prawn trawl grounds.