LAUSR

Abstract Patterns of phylogenetic diversity and structure of flora in refugial areas may be very complex due to habitat heterogeneity and legacies of various historical processes. Revealing these patterns can help understand the nature of the refugia. Here we study phylogenetic patterns of the flora in one of the main putative Pleistocene refugia of temperate forest biota in Europe through a community-level analysis for the main types of forests. For this purpose, a database of vegetation plots from deciduous and mountain coniferous forests of northern and central Greece was created. The plots were classified into floristically and ecologically interpretable community types. Faith’s phylogenetic diversity, mean phylogenetic distance and mean nearest taxon distance, as well as their standardized equivalents, were used to characterize the mean phylogenetic diversity and structure of each community type. All analyses were applied both for all the vascular plant taxa and for angiosperm taxa only. We related the measures of phylogenetic diversity and structure to ecological factors and compared them among community types. The 25 identified community types differed in their phylogenetic diversity and structure. The observed patterns were significantly different when old evolutionary lineages, such as gymnosperms and ferns, were taken into account. A random or overdispersed phylogenetic structure was observed for most of the community types, while only two oak-dominated community types were clustered. The highest phylogenetic diversity and a notably high proportion of plots with an overdispersed phylogenetic structure were observed in two mesic ravine forest community types (one of these transitional to beech forests). Phylogenetic metrics correlated with ecological conditions, in particular, overdispersion was more common in shaded, cool and moist forests. The high phylogenetic diversity and overdispersed structure of the two ravine community types can be related to the putative refugial role of these forest habitats.