| Summary: | The aim of this study is to differentiate old-field plant communities along the abandonment time and/or environmental gradient in the landscape surrounded villages with established Czech settlers in Romanian Banat area conserving traditional agriculture, and to identify site factors which cause plant diversity of particular vegetation types. Study area: Wider territory centered by the village Sfânta Elena, southern Romania ((44°40’ N; 21°43’ E). Methods: We collected 97 phytosociological relevés covered the same number of old-fields in the area and the following habitat parameters were measured: soil pH, available phosphorus, total carbon and nitrogen, Heat Load Index. Software TURBOVEG / JUICE was used to collect and elaborate the data set of relevés. Old-field vegetation was classified into five basic plant communities using TWINSPAN (all the botanical material includes 291 plant species). For each community, we detected diagnostic species according to their fidelity index. The presence of mowing, grazing or burning was registered for recorded stands. Ecological preferences of each community were examined using one-way analysis of variance (ANOVA). Vegetation-environment relationships were analysed using ordination method – Cannonical correspondence analysis (CCA) in CANOCO for Windows (version 4.5) to find the main variability gradients within the dataset. Scatter plot relationships between variables were constructed. Main results and conclusions: Dependence of number of species (alpha diversity) on the abandoned field’s age exhibits an unimodal shape of this relationship with the maximum peak of species diversity in plant stands aged approximately 13 years. The most importnat ecological factors and/or type of management in the relationship to the old-field plant composition show the following significance order: available phosphorus content in the soil (P), total nitrogen content in the soil (N), presence of burning, length of abandonment (old-field age), carbon/nitrogen ratio in the soil (C/N). Other parameters (grazing, mowing, zero management) do not demonstrate effective impact according to our dataset and seem to be equal to the absence of burning.
|
|---|
