Brose U, Martinez ND, Williams RJ.. Estimating species richness: sensitivity to sample coverage and insensitivity to spatial patterns. Ecology 84: 2364-2377

Ecology (Impact Factor: 4.66). 09/2003; 84(9):2364-2377. DOI: 10.1890/02-0558


The number of species in an area is critical to the development of evolu-tionary and ecological theory from mass extinctions to island biogeography. Still, the factors influencing the accuracy of estimators of species richness are poorly understood. We ex-plored these factors by simulating landscapes that varied in species richness, relative abun-dances, and the spatial distribution. We compared the extrapolations of nine nonparametric estimators and two species accumulation curves under three sampling intensities. Com-munity evenness of species' abundances, sampling intensity, and the level of true species richness significantly influenced bias, precision, and accuracy of the estimations. Perhaps most surprisingly, the effects of gradient strength and spatial autocorrelation type were generally insignificant. The nonparametric estimators were substantially less biased and more precise than the species accumulation curves. Observed species richness was most biased. Community evenness, sampling intensity, and the level of true species richness influenced the performance of the nonparametric estimators indirectly via the fraction of all species found in a sample or ''sample coverage.'' For each particular level of sample coverage, a single estimator was most accurate. Choice of estimator is confounded by a priori uncertainty about the sample coverage. Accordingly, researchers can extrapolate species richness by various estimators and base the estimator choice on the mean estimated sample coverage. Alternatively, the most reliable estimator with respect to community evenness can be chosen. These predictions from our simulations are confirmed in two field studies.

Download full-text


Available from: Neo Martinez,
  • Source
    • "On the other hand, we also investigate one important consideration that is irrespective of reproductive strategy: the effect of spatial autocorrelation on the ability to accurately characterize genetic diversity. Our finding that, in the presence of spatial autocorrelation, non-random sampling methods are biased in estimating genetic diversity agrees with similar studies looking at other diversity indices (Fager 1972; Baltanás 1992; but see Brose et al. 2003). "
    [Show abstract] [Hide abstract]
    ABSTRACT: The interpretation of genetic variation on and among corals on reefs has rarely been contextualized in an explicitly spatial framework where each individual has both genetic and spatial parameters. Here, we explore interactions between sampling effort, sampling design, and the presence of spatial genetic structure (SGS) on the ability to effectively characterize genetic diversity on a coral reef. To do this, we take our dataset of 2352 genotypes (78 unique multi-locus genotypes) of the coral, Pocillopora damicornis, collected from June 2007 to October 2009, a dataset of near-exhaustive sampling of a single patch reef in Kāne‘ohe Bay, Hawai‘i (21°27.462N, 157°48.405W), and subsample from it using three different strategies: (1) random sampling from throughout the reef; (2) saturation sampling all corals from within a pre-defined area of the reef, avoiding the reef’s edge; and (3) sampling all nearest neighbors of a randomly chosen coral while allowing for the inclusion of the reef’s edge. Our results demonstrate appreciable variation (e.g., 0.35–0.46) in estimates of observed heterozygosity (H O) using a typical sample size of 50 and that in the presence of SGS, non-random sampling schemes can give biased estimates of genetic diversity. Furthermore, our results indicate that over 1000 samples (i.e., ~40 % of the total number of colonies) are required to reveal the true pattern of spatial genetic structure at our site. We also demonstrate by rarefaction analysis that the bias in estimating clonal richness (i.e., the proportion of unique genotypes in a given sampling area relative to the total number of samples surveyed) for small sample numbers is due to the predominance of clones (i.e., high level of clonality) and not skew in genet frequency distribution. Overall, we argue that: (1) consideration of sampling design is important in population genetic studies, particularly since non-random sampling in the presence of SGS can give biased estimates of genetic diversity and (2) intense to near-exhaustive sampling schemes may be important for characterizing genetic diversity in highly clonal populations.
    Marine Biology 05/2015; 162(5). DOI:10.1007/s00227-015-2634-8 · 2.39 Impact Factor
  • Source
    • "For instance: if the assemblage is made up almost exclusively by species with high numerical densities, the use of small samples may suffice to estimate total species number (because nearly all species are already represented in small samples and larger samples would hardly add new species to the list). On the other hand, in the case of an area that is inhabited by many species of low abundance, large samples would be needed to encounter a sufficient proportion of the total assemblage, i.e. to reach sufficient " coverage " (Brose et al., 2003). When most species are present in low numbers, the initial part of the species accumulation curve is bound to rise slowly with increasing sample size and a wide range of samples sizes (meaning a high sampling effort) would be needed to reliably construct a species accumulation curve and estimate the total species richness of the assemblage. "

  • Source
    • "Nonetheless, the relative inaccuracies of the eight models we evaluated provide some indication of estimator performance in the context of the present study. The greater inaccuracy of MM may have been due to a poor fit of the model to the data (O'Hara 2005) or the underlying relative abundance distribution (Brose et al. 2003). That the abundance-based CH1 model was statistically more inaccurate than the incidence-based counterpart (CH2) may be attributable to tendencies in the variables that comprise them. "

Show more

Similar Publications