Monitoring floristic diversity in managed forests. Effects of sampling designs

Forest managers already monitor several aspects of forest stands, such as dendrometric parameters or site type parameters. As far as floristic diversity is concerned, they often ask for distribution maps of species, species richness, or community assemblages. We carried out a project in the State Forest of Tronçais (France) to test whether small floristic quadrats (2m²) integrated to current tree stands inventory sampling designs could help to monitor floristic diversity. Three types of sampling designs were tested, with a twofold objective: (1) how can the sampling intensity be reduced without adding too much noise in biodiversity assessment? For this part, we used regular grids of plots, of different densities; (2) how can particular habitats be sampled? For this purpose, we used (a) irregular samplings (consisting of a 150-by-50-m-interval grid and additional points located on particular microhabitats (ditches, road embankments, windfall mounds, forest edges, etc.); (b) timed-inventory walks on 20 ha compartments. For the studied area, the preliminary results indicate that systematic 50 x 150 m grids using small floristic quadrats may be appropriate to (1) monitor and map local biodiversity measurements and relate them to ecological parameters; (2) monitor species frequencies and map their distribution, but only for the most common species. However, small floristic quadrats are not representative enough to assess total floristic diversity and to provide analysable data on infrequent species. Irregular samplings provide better assessment for gamma species richness, but they are too time-consuming. For this particular purpose, timed-inventory walks are cost efficient and appear to be complementary to regular grids of plots. Their repeatability remains to be tested however.