Effects of rooted and shoot sampling schemes on species richness and species-area curves

Background&Aims:There are two different approaches to record presences of species in a plot: rooted (the plant roots in the sampling unit) and shoot (overlapping with the sampling unit) presence. However it was demonstrated theoretically that the different recording scheme swill lead to differences in richness counts and different shapes of species-area relationships (SARs), this aspect is not generally acknowledged in ecology. The aims of our study are to quantify how big the differences of rooted vs. shoot sampling in vascular plant species richness counts are at different spatial scale and analyze how these two sampling approaches influence the shape of SARs. Material & Methods: We used the plots of the BiodivERsA project SIGNAL in semi-natural grasslands of six Eurasian countries. In each study site we established six blocks of 2.80 m x 0.40 m, subdivided into 448 micro-quadrats of 25 cm2. We recorded the vascular plant species composition in each of these micro-quadrats also with rooted and shoot sampling scheme. We calculated species richness values for ten different sampling unit sizes. To describe the rate of difference between the species richness obtained by the different sampling methods linear mixed-effect models were used. For detecting differences between rooted and shoot presence methods SAR models were applied based on species richness values. Main results & Interpretations: The effect of the different sampling scheme on species richness was scale dependent however effect size varied between different grassland types. Species richness recorded with shoot presence method was higher in every scale in the case of all sites, nevertheless the amount of relative differences became smaller with increasing plot size on every site. On the basis of our results we can confirm the statement that in case of comparing SARs from different studies it is essential to know which was used since if the method is uncertain, the comparison will not be responsible.