Search

Aims: The study aims at putting together and comparing different types of rock vegetation classified to different syntaxa and usually studied separately. The particular objectives of the study were: 1) to identify and classify the vegetation units of rock communities; 2) to explain floristic differences between phytosociological units with different habitat conditions. Study area: The study area lies at the northern border of the Italian Peninsula, between 43°40' and 44°40' N and between 9° and 11° E. It is characterized by a great variety of rocky habitats and represents a centre of endemism and a zone of phytogeographic transition between the Alps and the Apennines. Methods: The study was based on a dataset including 693 phytosociological relevés that were classified according to the recent updatings of the Braun-Blanquet method. During the classification, relevés were assigned to phytosociological classes based on the cover-abundance prevalence of diagnostic species, taking into account the current literature. The definitive classification into associations was carried out through cluster analysis. The analysis of floristic variation within the rock vegetation was performed through Principal Component Analysis, adopting Ellenberg indicator values as indirect environmental factors. Results: We identified 36 basic vegetation units (35 of them evaluated as associations) belonging to four phytosociological classes (Crithmo-Staticetea, Adiantetea, Asplenietea trichomanis and Polypodietea). All these vegetation units were characterized by their floristic composition, structure, syntaxonomy, habitat and distribution. Conclusions: More than half of the identified associations and a new suballiance Primulenion apenninae (within the alliance Androsacion vandellii) were newly described. Two alliances (Moehringion muscosae and Hymenophyllion tunbrigensis) were first documented as part of the Italian vegetation. Indirect gradient analysis indicated that, within the four classes, the floristic composition varied along different environmental gradients and that temperature can be identified as the factor most strongly correlated with the variation of rock-face vegetation. [ABSTRACT FROM AUTHOR]