Intra- versus inter-annual variation in counts of reef fishes and interpretations of long-term monitoring studies

Many sampling strategies have been proposed as appropriate for describing spatial patterns in marine organisms. There remain, however, many problems with the description, analysis and interpretation of temporal variation in abundances of organisms. In particular, there is a need to understand temporal error in the estimation of abundance of mobile organisms. In this paper we report estimates of temporal variation in abundances of tropical reef fishes attributable to sampling error at diurnal, daily and 'monthly' scales and compare these to inter-annual variation that might arise from processes such as mortality and recruitment. Uncertainty in estimates taken from the same sites over consecutive days was large for several species and accounted for the majority of error in estimates of abundance within years. Sources of error in estimates of abundance are discussed with consideration of the implications for long-term sampling and monitoring of fish assemblages. Short-term temporal variation must be considered along with spatial variation in the design and interpretation of temporal studies of mobile species.