Differences in species richness and life-history traits between grazed andabandoned grasslands in southern Sweden

Disturbance has profound effects on plant community composition. This paper deals with the influence of grazing on species richness and proportions of life-history attributes of grassland vegetation at six spatial scales (0.0 01-1000 m(2)) in two provinces of southern Sweden. The study comprised 33 d ry grassland sites, including 22 grazed and 11 abandoned localities, and 28 sites of coastal brackish meadows, divided into five management types (Fro m "heavily grazed" to "abandoned since long time"). In general grazed sites were species-richer than abandoned sites; especiall y at small plot sizes. However, there was a steeper increase in species num ber rewards larger plot sizes in the abandoned sites. Heavy grazing in the coastal meadows resulted in a comparatively low number of species, corrobor ating the intermediate disturbance hypothesis. The analysis of life-history traits indicated the importance of taxonomic g roup, canopy structure, height, regenerative strategy and, in particular, l ife form. Leaf anatomy and seed dispersal seemed to be less important. The responses to grazing as regards species traits differed somewhat between gr assland types. Grazed sites generally had high proportions of legumes, ther ophytes, species with basal position of leaves and with regeneration by mea ns of a persistent seed bank. Abandonment of grazing favoured monocots, geo phytes, species with vegetative regeneration and (partly) leafy canopy stru cture. Some differences between grazed and abandoned sites were confined to either the smallest or largest plot sizes, indicating different responses of matrix and interstitial species. Various positive associations (attribut e syndromes) or negative associations between individual traits were identi fied. There was, for example, a positive link between the attributes "geoph ytes" and "ability of vegetative regeneration". The recognition of such lin ks is important to avoid misinterpreting certain attributes as functional a daptations to grazing while they are only positively correlated to other at tributes of larger significance.