Assessing the effect of landuse in beech forests ­ an example from central Apeninnes, Italy.

In the mountain belt of the Marches region (Central Apennines, Italy), coppice management is applied to 70% of beech (Fagus sylvatica) woodlands. This method has long played an important socio-economic role, providing firewood and charcoal. Currently 4% of present beech woodlands are being converted to high forest, while in 21%, silviculture is being abandoned entirely. We aimed to answer these questions: (1) How does traditional management (short rotation coppice) affect beech forest plant diversity at different environmental conditions? (2) How is plant diversity influenced by an extension of the rotation cycle or by the abandonment of coppicing? Previous investigations on plant diversity relations in coppice woods focused on limited areas and small numbers of stands, while it would be much more useful for management applications to have diversity assessments on the scale of the landscape. After preliminary selection of a number of forest stands, stratified sampling was conducted according to 200 m elevation classes (starting from 1000 m a.s.l.), bedrock types (limestone and sandstone), and stand age (10-year classes, since the last coppicing), to identify 61 sites (0.3 to 35 ha) that would represent the landscape level heterogeneity. Within each stratum a proportional number of stands were randomly selected among those included in the forestry registers. Orthophotomaps and related cadastral information were derived using an Information System. Elevation was used as a proxy variable accounting for the main beech forest types reported in the regional forest inventory, which also stated soil and stand conditions. The beech forests belong to Cardamino kitaibelii-Fagetum and Lathyro veneti-Fagetum associations on neutro-basic limestone and Solidagini-Fagetum on acidic sandstone. Eighty plots of 20×20 m were assessed by phytosociological records, within the selected forests stands. Species were grouped into phytocoenological groups. Our results suggest that high spatiotemporal variability of forest management accompanied with high landscape-scale heterogeneity and complexity maintain rich regional species pool. On both bedrocks, in all elevation classes, the age after coppicing explained the majority of observed diversity changes at the stand level: forest specialists increased while non-forest species decreased. A relatively long time after coppicing (40–60 years), the contribution of the beech specialist species doubled, while non-forest and species from anthropogenic habitats decreased by about 50%. On sandstone, gap species also diminished over a long period, and the beech stands experienced stronger changes over time. We conclude that an extension of coppice rotation cycle or abandonment would result in lower local species richness but with higher contribution of beech forest specialists.