The Dynamic keyhole key model of coexistence to explain diversity of plants in limestone and other grasslands

How can up to 40 plant species coexist on 1 m(2) of grassland? Taking limestone grasslands of northern Switzer land as an example, coexistence is described and analysed with the Dynamic keyhole-key model. This model is a comprehensive formalization of the observation that different species occupy different microsites within an ecosystem. Where species alpha-diversity (keys) and microsite diversity (keyholes) match, coexistence can occur. Microsite diversity is mostly biogenic. For example, Microtus arvalis (common vole) creates (1) nutrient-rich microsites (excreta) where Dactylis glomerata and Galium album grow preferentially; (2) gaps which are colonized by Arabis hirsuta and Myosotis arvensis; and (3) microsites where this small rodent mainly feeds and where species such as Primula veris s.l. and Viola hii m are found because they are not eaten. Changing weather conditions, an example of abiogenic microsite diversity, permit the coexistence of species with high water-stress tolerance, such as Salvia pratensis and Plantago lanceolata, and low tolerance, such as Dactylis glomerata and Trifolium pratense. Other cases of the matching of plant species and microsite diversity are given. The generality of the model is investigated by testing its predictions about species coexistence in other grassland types in northern Switzerland. In fertilized meadows with 35 - 40 species / 50 m(2), uniform fertilizer application, frequent mowing, poor conditions for small rodents and other factors reduce the diversity of microsites and thus the possibility for the coexistence of different species. In xeric grasslands the diversity is only 31 - 36 species / 50 m(2). This may be explained by both the historic and the actual biogeographic situation which results in a small species pool for these grasslands. Old non-fertilized city lawns have only 8 - 36 species / 50 m(2). Homogeneous soil, frequent mowing and other factors drastically reduce microsite diversity. Only 20 - 50 species of the Swiss flora can stand frequent mowing and even fewer species are able to immigrate into the lawn 'islands' in the cities. Thus both species pool and microsite diversity are very small. These examples show that the keyhole-key model allows the under standing of species diversity and coexistence in these grassland types and probably also in other ecosystems. Comparisons between the model and other models of coexistence and its limitations and potentials are discussed.