The hollow-shaped species abundance distribution (SAD) and its allied rank abundance distribution (RAD)—showing that abundance is unevenly distributed among species—are some of the most studied patterns in ecology. To explain the nature of abundance inequality, I developed a novel framework identifying environmental favorability, which controls the balance between reproduction and immigration, as the ultimate source and species stress tolerance as a proximate factor. Thus, under harsh conditions, only a few tolerant species can reproduce, while some sensitive species can be present in low numbers due to chance immigration. This would lead to high abundance inequality between the two groups of species. Under benign conditions, both groups can reproduce and give rise to higher abundance equality. To test these ideas, I examined the variability in the parameters of a Poisson lognormal fit of the SAD and a square root fit of the RAD in diatom and fish communities across US streams. Indeed, as environmental favorability increased, more sensitive forms were able to establish large populations, diminishing the abundance disparity between locally common and rare species. Finally, it was demonstrated that in diatoms, the RAD belonged to the same family of relationships as those of population density with body size and regional distribution.
Diatom data
The file “Diatom data” contains 3 Excel spreadsheets, i.e. “Raw data,” “Community data,” and “Species data.” The Raw data spreadsheet shows the population density, measured in cells/cm^2, and the population density rank for all species (BioDataTaxonName) across 3213 diatom samples, listed by their sample identification. The diatom samples were collected from 1580 distinct stream localities, listed by their station identification. The Community data spreadsheet contains relevant information for the diatom samples, including latitude, longitude, date of collection, species richness, evenness, total density (cells/cm^2), statistics of the Poisson lognormal fit to the sample species abundance distribution (parameters mu and sigma, gof (goodness of fit), and p (proportion of species revealed by the sample)), statistics of the square root fit to the sample rank-abundance distribution (intercept a0, slope a, and their standard errors and 95% confidence intervals), ln Tolerant:Sensitive ratio, and density and richness of the three diatom guilds (low profile, motile, and high profile). Environmental data were available for a subset of 2949 samples. The Species data spreadsheet lists the 1777 diatom species (BioDataTaxonName) encountered in this study with guild designation whenever possible and occurrence across 1582 stream localities.
Fish data
The file “Fish data” contains 3 Excel spreadsheets, i.e. “Raw data,” “Community data,” and “Species data.” The Raw data spreadsheet shows the abundance, measured as number of individuals, and the abundance rank for all species (BioDataTaxonName) across 761 fish samples, listed by their sample identification. The fish samples were collected from 399 distinct stream localities, listed by their station identification. The Community data spreadsheet contains relevant information for the fish samples, including latitude, longitude, date of collection, species richness, evenness, total abundance (number of individuals), statistics of the Poisson lognormal fit to the sample species abundance distribution (parameters mu and sigma, gof (goodness of fit), and p (proportion of species revealed by the sample)), statistics of the square root fit to the sample rank-abundance distribution (intercept a0, slope a, and their standard errors and 95% confidence intervals), ln Tolerant:Intolerant ratio, and abundance and richness of the three fish tolerance groups (tolerant, intolerant, and moderate). Environmental data were available for a subset of 732 samples. The Species data spreadsheet lists the 460 fish species (BioDataTaxonName) encountered in this study with tolerance group designation whenever known and occurrence across 1005 stream localities.