Plant diversity in a managed temperate deciduous forest: understorey response to two silvicultural systems.
Disturbances and resource availability are key factors affecting plant diversity in managed forests. As disturbance regimes vary among silvicultural systems and may simultaneously affect different types of resources, effects on biodiversity can be unpredictable. We compared the effects of two silvicultural systems on understorey plant diversity, including species composition, structural attributes and functional organization. One hundred and thirty-five phytosociological relevés were sampled from 27 forest stands managed under either a traditional coppice-with-standards (CWS, n=12) or a 'close-to-nature' selective cutting system (SC, n=15), over similar edaphic conditions. Important environmental factors affecting vegetation were deduced using Ellenberg indicator values. Structural diversity was described using traditional indices of α and β diversity. Guilds were defined within the local pool of species using a set of 14 traits and their relationship with silviculture was assessed using correspondence analysis. Post-logged CWS stands share some compositional and structural characteristics with selectively cut stands, including high species richness and a dominance of early successional species. However the species pool for all coppicing areas was higher than for selectively cut areas, suggesting that the high disturbance frequency occurring in the latter may progressively eliminate the most sensitive species. Functional diversity strongly differs between the two systems. Although it is conserved through the silvicultural cycle in the coppice-with-standards system, some guilds were lacking in selectively cut stands. The most negatively impacted guilds were tree and shrub saplings, prostrated ruderals, shade-tolerant perennials and vernal geophytes. The latter two comprise 'true forest species' which may also be considered as 'coppicing-maintained species'. To reach the same values of guild richness (i.e. number of guilds) or redundancy (i.e. proportion of the maximal species richness within each guild), larger areas were required in SC compared with CWS systems. In the SC system, the high proportion of light reaching the forest floor induced a spectacular spread of blackberries Rubus fruticosus agg., which decreased species richness. It also caused shifts in guild composition: graminoids and ferns grew strongly to the detriment of true forest species. Synthesis and applications. Our results suggest long-term negative effects of selective cutting on both structural and functional plant diversity, compared with coppice-with-standards. Cutting intervals are shorter than recovery times, so that early successional species-dominated communities are maintained. Vernal geophytes and shade-tolerant perennials seem to be limited by the frequency of disturbance rather than by the severity of disturbance. We conclude that, from a biodiversity point of view, this 'close-to-nature' system does not cope with the objective of sustainable forest management. The rotations currently in use do not match natural disturbances very closely and are applied to a managed system rather than a natural forest. Retaining remnants of old coppice woods and extending rotations to at least 50 years are recommended where biodiversity conservation is a goal of forest management.