The study of vegetation turnover, composition gradients and β-diversity requires systematically collected sampling data from vegetation plots and relevés. Occurrence data from museum and herbarium specimens, however, may be used in cases where such data is scarce.
This study compares phytosociological vegetation data with mainly GBIF-mediated herbarium data from páramos (alpine ecosystems) in Colombia in their ability to describe diversity patterns and species composition gradients.
The authors show that overall species composition differed between the two datasets—perhaps explained by a bias towards charismatic and/or rare species in the herbarium data. So while a large number of species were present in a few locations in the herbarium data—and absent from the plot data—many species present in the plot data were common to a larger number of plots than described by the herbarium data.
Pairwise β-diversity analyses showed no significant differences between the datasets, and among common species, ecological distributions patterns were similar. The authors conclude that while plot data is best for describing local patterns, combining data from the two sources may be efficient at broader scales.