The indices that are used to assess species diversity are sensitive to sampling eff ort, meaning that engineers have to standardize the variability of diversity indices by determining their response to sampling effort. However, it is often difficult for engineers to decide whether individual-based or sample-based sampling eff ort is most suitable for their tasks. In addition, the sample size needed for reliably calculating the diversity indices based on individuals in a community of marine invertebrates remains unclear because we do not understand the extent to which sampling eff ort contributes to the accuracy of diversity indices.
In this study, we fi rst showed that the meaning of the two types of sampling eff ort diff ers when the temporal and spatial distributions of species in a community are heterogeneous. We compared the diversity indices defi ned by using the two types of sampling eff ort by analyzing model communities, and our analysis indicates that there are both merits and demerits in the definitions and analytical techniques of each sampling eff ort. However, using indices defi ned by both types of sampling eff ort is eff ective for clarifying the spatial heterogeneity of species diversity, which is an important concept when attempting to identify communities that require conservation management. Second, our analysis of model communities reveals the importance of sampling dominant species completely; by analyzing the data of a real community, we determined a rule-of-thumb for setting sample size for the reliable assessment of marine invertebrate populations. Our fi ndings should help engineers assessing species diversity in port and coastal areas to choose the most suitable sampling effort, thus increasing the reliability of observations of marine invertebrate populations.

Key words:Diversity index, Sampling effort, Spatial heterogeneity, Dominant species, Rarespecies