MEPS 435:183-195 (2011)  -  DOI: https://doi.org/10.3354/meps09204

ABSTRACT: Variability at small temporal scales can confound comparisons over longer periods of time but has often been ignored in studies of marine organisms. To determine appropriate methods for sampling fish at multiple temporal scales, we used hierarchical sampling to test hypotheses about temporal and spatial variation of individual species and assemblages of fish sampled with multi-mesh gillnets in shallow and deep habitats of a coastal lake in New South Wales, Australia. The design included temporal scales of 1 wk, 1 and 3 mo and spatial scales of sites (separated by 1 km) and zones (separated by several km). Patterns of temporal variation in univariate and multivariate analyses were complex and inconsistent across the different spatial scales and between habitats. The only general pattern was that spatial variance among replicate gillnet samples in each site was larger than variance at any of the temporal scales examined. Potential mechanisms explaining the observed patterns are likely to be complex and these are considered along with the implications for sampling assemblages of fish. A lack of generality in patterns of temporal variance reinforces the need for pilot studies to identify relevant scales of variation before commencing large-scale, long-term studies of any organism.

KEY WORDS: Estuaries · Fish assemblages · Hierarchical experimental design · Patchiness · Temporal variability · Variance