Abstract. 1. The sampling methodology used in diversity studies may at times define perceived biological patterns. In this study we used urban grassland fragments in the Front Range of northern Colorado, USA, to determine whether grasshoppers exhibit a species–area relationship (SAR) and to examine the potential effects of sampling protocol, namely uniform and proportional sampling, and sampling biases on our ability to detect such a relationship.
2. A uniform sampling protocol was used to survey grasshoppers across 13 urban fragments and then sample-based rarefaction curves were used to determine how the SAR would have been modified had we used proportional sampling. We also used individual-based rarefaction curves to explore the issue of ‘passive’ sampling and used species richness estimators to address the efficacy of our sampling protocols.
3. We found that reductions in habitat area due to urban fragmentation lead to a significant SAR between grasshopper species density and richness and fragment size. Using field data we also show, that relative to a uniform sampling protocol, proportional sampling can lead to similar conclusions about the density of species found within an area with the potential benefit of not over sampling smaller relative to larger areas. Finally, the estimators showed that most of the species in each fragment were detected and the individual-based rarefaction curves showed that the detected SAR was not due to ‘passive’ sampling.