Jump to Main Content
Validation of plant diversity indicators in semi-natural grasslands
- Öster, Mathias, Persson, Kill, Eriksson, Ove
- Agriculture, ecosystems & environment 2008 v.125 no.1-4 pp. 65-72
- grasslands, botanical composition, species diversity, indicator species, geographical variation, vegetation, land management, grazing management, model validation, surveys, precision, Sweden
- The use of plant biodiversity indicators in a nationwide survey of semi-natural grasslands in Sweden was examined by comparing results from the survey with additional in-depth studies of plant diversity at the same 30 sites in southern Sweden. Additional grasslands, that were not subjected to the nationwide survey (i.e. rejected after an initial site selection), were also investigated in order to examine the quality of the initial site selection and to assess to what extent they harboured species rich plant communities. Results showed that grasslands that were not included in the nationwide survey were generally smaller in size and that they contained significantly lower plant diversity than sites that were included by the survey. However, some indicator species, including indicators for poor management, were abundant in both types of sites. Biodiversity indicators correctly indicated both plant species richness and plant species density when using data from the in-depth study. However, by comparing indicator species found by the nationwide survey and by the in-depth study, it was shown that the survey overlooked 42% of all indicator species occurrences, which removed the significant relationship between indicator richness and total plant species richness. Furthermore, a null model showed that the chosen indicator species did not perform significantly better than species chosen at random from the available species pool. The conclusion was that validation of indicators is crucial, because even though real correlations exist between taxa or between a subset of species and overall diversity, poor precision of surveys might make these indicators useless. This also suggests that the effort put into searching for the indicator species may have to be so high, that it may be more efficient to go directly into assessing the biological values they are supposed to indicate.