Small‐scale plant species turnover in a limestone grassland: the carousel model and some comments on the niche concept

Abstract. This study reports on small‐scale changes in the distribution of plant species in a 2.5 m 2 plot of grazed, species‐rich Veronica spicata ‐ Avenula pratensis grassland on shallow, dry, nutrient‐poor soil in the Great Alvar area (Stora Alvaret) of southern Öland, southeastern Sweden. Multivari‐ate analysis of 0.001 m 2 and 0.25 m 2 quadrats within the plot showed that there is little floristic variation without any trend in the plot. Average species richness varied little throughout the study period from 1986 to 1991 with 1986 averages of 7.0 on 0.001m 2 , 16.3 on 0.01 m 2 , and 26.1 on 0.25 m 2 . On 0.001 m 2 the highest species number found was 12, on 0.01 m 2 , 27. However, cumulative species richness, i.e. species number in the first year plus new species appearing in later years (averaged over 40 quadrats) increased over the same period, on 0.001 m 2 from 7.0 in 1986 to 14.9 in 1991, and on 0.01 m 2 from 16.3 to 24.1. Cumulative frequency, i.e. number of quadrats out of 40 where a species occurred in the first year or/and in later years, increased as well; the number of cumulatively highly frequent (> 80%) species increased from 5 in 1986 to 18 in 1991. Species mobility on the scale of the small quadrats used implies both appearance and disappearance from these quadrats. Using six examples, species mobility is shown to vary individualistically, both in rate and degree. We postulate that in homogeneous, grazed, nutrient‐ and water‐deficient environments many species can reach virtually all microsites within the plot, which we express through the idea of the ‘carousel model’. We also question the usefulness of the niche concept and re‐interpret it by stating that all species of this plant community have the same habitat niche, while most of them are short‐lived and have the same regeneration niche. The essential variation amongst the species is their individual ability to establish or re‐establish by making use of favourable conditions appearing in microsites in an unknown, complex spatio‐temporal pattern.