The factors influencing plant species diversity in secondary and ancient forests can differ. Apart from environmental variability caused by natural conditions, secondary forests are influenced by historical factors (previous human activity). However, the effect of historical factors on vegetation is not fully understood. Secondary forests that have developed in abandoned villages in the Doupovské hory mountains, Czech Republic, were surveyed and compared with ancient forests in an attempt to determine the effect of historical factors and separate it from that of natural gradients. The results show that secondary forests in abandoned villages form a unique type of vegetation that differs from ancient forests mainly in the presence of species indicating a high nutrient content and high pH of the soils. This indicates that the previous high nutrient input in the villages still influences the soils and causes the differences. Variability of village forests is influenced mainly by a gradient in the available phosphorus content of the soils, soil moisture (approximated by wetness index) and organic matter content. The pattern in the phosphorus content and pH indicate a different intensity of historical influence in the centre compared to the periphery of the villages. Vegetation variability is modified by former land-use and village structure. The effect of historical factors is relatively strong and cannot be explained by coincidental initial conditions.
A phytosociological study of the West Carpathian mesic hay meadows and pastures (order Arrhenatheretalia elatioris) was performed and is the first unified investigation into the vegetation diversity in the area, which is situated in three countries (Slovakia, Czech Republic and Poland). Because of the differences in the current classification systems used in different countries it was not possible to make a single selection of the Arrhenatheretalia relevés from the databases, so a data set containing relevés originally assigned to three orders encompassing this vegetation in hay meadows and pastures in the area (Arrhenatheretalia elatioris, Molinietalia and Nardetalia strictae) was established. This data set was classified using cluster analysis. Only the cluster corresponding to the order Arrhenatheretalia elatioris at the level of three clusters was further classified in the same way as the whole data set. The ecological interpretation of the classification was based on altitude, Ellenberg indicator values and geological bedrock. The clusters were also compared with the syntaxonomical assignment of the relevés by their authors. The classification at the level of 12 clusters reflected the most widespread vegetation types of mesicmeadows and pastures recorded in the area. The vegetation of extensive pastures, corresponding to the association Anthoxantho odorati-Agrostietum tenuis, seemed to be more similar in floristic composition to the mesic meadows of Arrhenatherion elatioris than to the intensive pastures of Cynosurion cristati, where it was traditionally classified, which has important conservation consequences because of the different position of these units in conservation systems such as Natura 2000. Higher altitude meadows were divided into four vegetation types including meadows corresponding to the association Gladiolo imbricati-Agrostietum capillaris, which is a frequent community in the Polish Carpathians that does not occur in the other regions. Montane meadows currently classified in Polygono bistortae-Trisetion flavescentis were less clearly distinguished, probably because of their patchy distribution in theWest Carpathians. The differences in vegetation diversity of meadows and pastures between particular countries were confirmed, with Gladiolo imbricati-Agrostietum capillaris occurring predominantly in the northern part of the West Carpathians and Anthoxantho odorati-Agrostietum tenuis virtually absent here. The ecological determinants of variation in montane meadows are discussed.