Evaluating landscape quality with vegetation naturalness maps: an index and some inferences

Question: Can landscape quality be evaluated and compared with a single numerical value using vegetation maps? Location: Northern Apennines (Italy), ca. 44 degrees N, 10-11 degrees E. Methods: Seven phytosociological vegetation maps (1 : 25 000), which correspond to man's different impact on mountain landscapes, were considered. Syntaxa were classified into five degrees of naturalness: urbanized, agricultural, semi-natural, sub-natural, and natural. Vegetation maps showing naturalness were derived in a vectorial GIS. The degrees of naturalness were ordered according to increasing naturalness. If c(i) is the cumulative relative value of every mapped area of the degrees of naturalness, the sum of these cumulative values A = Sigma c(i) is is a measure of vegetation artificiality. Its maximum value is A(max) = n-1. The Index of Vegetation Naturalness IVN = 1-A /A(max), ranging from 0 to 1. Our IVN is an extension of the ILC by Pizzolotto & Brandmayr (1996) due to the ordinal character of the vegetation classification into degrees of naturalness. The maps of vegetation naturalness were also analysed by two known metrics for the evaluation of landscape quality: TECI (Total Edge Contrast Index) and MSI (Mean Shape Index). Results: The case studies show that IVN is linearly correlated with decreasing area of urbanized and agricultural vegetation types as well as with increasing area of the highest degree of naturalness. Conclusions: IVN can be joined with the TECI for the evaluation of naturalness of landscapes. TECI can supply additional information about the importance of landscape ecotones. Our case studies suggest that an urbanized landscape should correspond to IVN values lower than 0.20. A natural landscape will have IVN values higher than 0.80.