The ecological indication content of Central European village floras

Abstract

The paper is aimed at detecting the main factors determining the composition of Central European village floras and their ecological uniqueness by exploitation of the condensed ecological information of quasi-complete datasets of their spontaneous (and subspontaneous) vascular plants. Data were gained in two districts each in Germany (Pomerania; Thuringia) and the Czech Republic (Bohemia; Moravia) at a N–S gradient of 700 km and altitudes of 0 – 775 m a.s.l. In each district two regions were selected and seven villages sampled in each. The lists of plant taxa of each village resulted in a matrix of 56 villages and 1008 taxa. These presence / absence data were used as input for multivariate analyses: NMDS, cluster analysis and CCA. Ten variables related to climate, position, geology, population and phytodiversity were employed to evaluate their effect on structuring the village floras. NMDS and cluster analysis revealed possibilities of forming of village groups very close to the reality using similarity relations between their distinct, sufficiently heterogeneous village floras and inherent specific ecological information; in both cases the Jaccard-index was preferred. CCA with the eight significant variables explained 18.3 % of the variance of species data and 76.7 % of the variance of species–environment relation, both referred to the first four axes and a sum of canonical eigenvalues of .768. Climatic variables proved to be most important, with July, January and mean annual temperatures, “thermic continentality”, annual precipitation and altitude having strong effects. Of the two variables related to geology, only the base presence in subsoils was included. Hence, natural factors determine the composition of Central European village floras despite all human influence in a scale-dependent way: the larger the study area, the more important are biogeographic parameters; the smaller its extent, the more relevant become landscape-ecological determinants. At the over-regional international scale of the present study the patterns found are distinct and plausible, since they fit in general phytogeographic principles. Therefore, quasi-complete datasets of spontaneous village floras are considered to be solid and multi-purpose ecological indicators, and this novel method is recommended for further testing.