The cluster analysis of the 24 modem pollen rain samples shows five main clusters (Figure 5), demonstrating a clear discrimination between pseudomaquis communities (cluster E; samples 1, 2, 21, 22 and 23) and other clusters.
A fourth cluster (cluster 4) includes tightly clustered Paliurus/Rhamnus and Pistacia pollen, plus evergreen Quercus, Cistus, Lonicera and Hypericum clustered at greater distance, reflecting a low elevation group of taxa found in pseudomaquis type vegetation.
In general, samples from acidophilous, ravine and calcicolous beech forests and subalpine calcareous grasslands were placed at the negative side of the axis 1, while samples from pseudomaquis communities and thermophilous beech forests were mostly at the positive part.
Axis 1 separated taxa related to clusters 3 and 4 of the HCA (Figure 6), situated at the positive side, representing thermophilous beech forests and pseudomaquis type vegetation respectively, from the clusters 1 and 2 related with ravine and acidophilous beech forests, which are situated at the negative half.
In fact, those plant communities that develop at lower altitudes in warmer and less moist conditions are those that are represented on the positive side of CA axis 1, in which pseudomaquis samples have higher positive loadings than thermophilous beech forest samples.
Cluster E of the HCA (Figure 5) represents pseudomaquis communities of mixed deciduous and shrubland formations occurring as a result of degradation of deciduous forests in Greece.
Other herbaceous taxa from pseudomaquis samples include moderate amounts of nitrophilous elements suchs as Aster, Cardueae, Cichorioideae, Dipsacus fullonum, Rumex acetosella, Plantago lanceolata and Urtica dioica, as well as spores of some coprophilous non-pollen palynomorphs such as Coniochaeta, Riccia, Sordaria (maximum of 11%) and Sporormiella (Figure 4).
For the Quercus coccifera pseudomaquis communities there is no accepted syntaxonomy in Greece (Dimopoulos & al.