Using woody genera for phytogeographic regionalization at a medium scale. A case study of Italy

We present a phytogeographic regionalization based on native woody flora, identifying the most useful taxonomic level, geographic variables, and orographic pattern, selecting Italy as a case study. We generated seven distance matrices among the 20 administrative regions, and using Pearson’s correlation coefficients and PCA, we verified whether distances between regions were invariant across the different sampling strategies. Once this invariance was established, we focused on genera representation. We defined two orographic indices and performed Kruskal–Wish multidimensional scaling and K-means clustering to assess Italy’s phytogeographic regionalization. A major North–South and a minor East–West gradient described the relationships between regions. Floristic diversity was strongly correlated with the region’s orography, with hills being the most important orographic feature that increases plant diversity; the effect of the orographic patterns was independent from the geographic clines observed. Despite the coarse scale, our phytogeographic regionalization comprising six clusters (variables = 133 woody genera) was consistent with previous ones based on the endemic flora (variables = 1371 units) or on bioclimatic approaches. In particular, the phytogeographic uniqueness of Northern and peninsular Italy, and of Sardinia Island, was confirmed. Next step will be to test our method at a finer scale.