Finescale species associations in alvar limestone grasslands

Abstract

To examine the importance of positive or negative interactions of plant species on a micro‐scale, species associations were studied in 38 1 m2 ‐plots in alvar limestone grasslands on Öland, Sweden. Each plot was assumed to be homogeneous as to its environmental conditions and spatial distribution of dominant species, and was divided into 25 sub‐plots of 100 cm2 size arranged in a regular grid. Presence‐absence data of rooted vascular plants were recorded for each sub‐plot. Significant positive or negative associations were observed in all plots. The number of significant interactions was positively related to the within‐plot variation in soil depth. In seven and nine cases, respectively, there were more positive or negative associations than expected on the basis of a random distribution of species over the sub‐plots. The ratio of the number of positive associations to the potential number of positive associations increased with increasing species richness. There was a significant spatial autocorrelation in species composition in most plots, but the coefficients of determination were on average very low, indicating a minor importance of spatial dependence in the data set. Correspondingly, the ratio of observed to potential species associations was unrelated to the extent of spatial autocorrelation. About 2/3 of all species encountered were at least once involved in a significant interaction. We found differences in the proportion of positive to negative associations between species with different habitat requirements and attributes. In general, species with predominantly negative associations were mostly relatively tall and competitive, long‐lived hemicryptophytes characteristic of Festuco‐Brometea and more mesotrophic grasslands, for example Filipendula vulgaris and Achillea millefolium. In contrast, species with largely positive associations were mostly smaller and shortlived stress‐tolerators and stress‐tolerant ruderals, preferring more shallow and exposed soils, such as Cerastium semidecandrum and Trifolium arvense. While the negative associations prevailing among the dominant species in the alvar grasslands are likely to be the outcome of interspecific competition for space and resources, the positive associations may be the result of facilitative relations between the dominant and sub‐ordinate species.