Dunn, Christopher P.
Loehle, C.

Species-area parameter estimation: testing the null model of lack of relationship

Date created
The species-area curve from island biogeography theory has been applied to many non-oceanic situations. We study problems of statistical estimation for this curve using the flora of upland and lowland forest islands in south-eastern Wisconsin. These forest data do not fit the hypothetical curve; in both cases the slope does not differ significantly from zero. Explanations for this result are offered and are explored using Monte Carlo sampling of several ranges of island size. The zero slopes of the field data are judged to be 'real' in that they are not likely to be subsamples of a larger population with wider range of island sizes and non-zero slope. Factors other than size, such as disturbance and forest edge,probably contribute to the total number of plant species in these islands.We conclude that sampling and statistical problems need to be considered when applying the species-area relationship.
Alternate Title

Volume, Page Number
15, 721-728