Using Rao's quadratic entropy to define environmental heterogeneity priority areas in the European Mediterranean biome

Although environmental heterogeneity (EH) has recently gained increasing attention, its connection to conservation planning still remains limited. We aim to assess the efficiency of Rao's quadratic entropy for measuring EH and conducting prioritization schemes. We focus on EH priority areas in the European Mediterranean biome and we estimate to which extent they are included in national and international protected area networks. We estimate Rao and Shannon EH indices for the bioclimatic and topographic variables of the CHELSA and the ENVIREM datasets. We use systematic conservation planning algorithms, for each dataset and for several moving window extends and we estimate the overlaps among the different schemes. The Rao and Shannon EH indices do not measure EH levels in a similar way and Shannon indices tend to overestimate EH comparing to Rao. The overlaps between the ENVIREM and CHELSA prioritization schemes are moderate (Jaccard index = 0.4-0.6) and not sensitive to the moving window extend. We consider only the priority areas that are identified as such by both independent datasets. One third of the EH priority areas are located in Greece, one forth in Italy and one fifth in France and Spain. Only one third of the EH priority areas are located within nationally designated protected areas and up to half are located in Key Biodiversity Areas and the Natura 2000 network. We conclude that Rao's methodology should be further used for environmental heterogeneity analyses. EH priority conservation areas are only partially included in protected area networks, mostly in mountainous areas.