Larsen, David P., and A.T. Herlihy. 1998. The dilemma of sampling streams for macroinvertebrate richness. Journal North American Benthological Society 17(3)359-366.
Developing efficient sampling methods to characterize the richness of macroinvertebrate assemblages in streams continues to receive attention, particularly among those charged with characterizing the effects of human disturbances on these assemblages. Richness can be expressed two ways: in terms of the numbers of distinct taxa in a defined area (areal richness), such as the number in a square meter, or in terms of number of distinct taxa tabulated after a specified number of individuals have been enumerated (numerical richness), such as the number of distinct taxa per 300 individuals enumerated. A recent set of articles advocated one perspective or the other, without an apparent synthesis. In this article, we demonstrate that when an adequate amount of material is processed, both measures of richness are highly correlated, consequently provide essentially the same information about taxa richness. At lower levels of effort, the two measures are poorly correlated, consequently provide different information. The lack of correspondence between the two measures at lower count efforts arises because the relative abundance of taxa affects one measure (numerical richness), but not the other (areal richness). We conclude that for many purposes, the use of numerical richness is preferred because of the ease of standardization across numerous studies whose sampling protocols might vary, conditioned on the premise that an adequate amount of material is processed. In this study, processing 300-500 individuals was adequate.