Measuring surface complexity in ecological studies

Habitat complexity is a potential structuring force in benthic communities. Different studies often estimate complexity in different ways, and it is not always clear how precise the separate techniques are. Here we review three methods of estimating surface complexity: stereo photography, profile gauges, and lengths of chain contoured over the substratum. We derived fractal dimensions for the quadrats in the rocky intertidal zone using each technique. Complexity estimates from chains and profile gauges were related, but neither technique was correlated with the results from stereo photographs. Stereo photographs appeared to overestimate complexity on smooth surfaces. The variance of fractal dimension estimates increased nonlinearly with the mean fractal dimension in each quadrat. Recommendations for the number of replicates needed for a reliable estimate of fractal dimension from a quadrat, therefore, vary as a function of surface complexity. Within the range of complexities typically encountered on rocky shores, as few as three profiles or sets of chains can produce relatively reliable estimates of fractal dimension. The most robust and time effective method, however, would be to sample using as many chain profile sets per quadrat as is logistically feasible. Given the changes in precision with surface complexity, comparisons between studies need to take careful note of the number of replicates and the average level of surface complexity. A null result (no relationship between surface complexity and an ecological variable) could be produced by imprecise estimates of surface complexity based on too few replicate measurements per quadrat