Do diverse overstoreys induce diverse understoreys? Lessons learnt from an experimental-observational platform in Finland

The understorey fulfils many important ecosystem services, such as mediation of carbon dynamics, provision of habitats, and it contains most of the plant diversity in forest ecosystems. Changes in the overstorey diversity may affect understorey diversity as trees have a species-specific impact on resource availability and soil conditions that influence the understorey. In an attempt to disentangle the overstorey-understorey diversity relations, we combined the strengths of an experimental and observational approach in the boreal vegetation zone in Finland. The Satakunta tree diversity experiment was planted in 1999 using a pool of five tree species and four species richness levels (1, 2, 3, 5 species per plot) (only the first three levels were analysed in this study). Each level was replicated with different species composition to avoid complete dilution, which allows the separation of identity and diversity effects. Understorey surveys were performed in three subplots per plot in 2003 (young phase) and 2011 (established phase). In the full-grown forest surrounding Joensuu, a similar design was used in an observational plot-based study with three tree species richness levels (1, 2, 3 species per plot), containing trees from a pool of three species. The understorey was surveyed in 2012, also in three subplots per plot. We unravelled the relations between (1) tree species richness and (2) understorey composition, diversity, compositional dissimilarity within and between plots and temporal turnover, and searched for tree species identity effects. Tree species richness had a significant influence on the understorey composition in the established phase of the experiment. In contrast with the expectations, plot-level understorey diversity showed no significant differences between the tree species richness levels, neither at the experiment nor at the full-grown forest. At the established phase of the experiment, interplot compositional dissimilarity was significantly higher for monocultures than for mixtures. Monocultures have distinct influences on resources and soil conditions, leading to larger differences with other plots, while mixtures often share the same tree species or species with similar environmental impact. Tree species identity effects were present in monocultures but were predominantly tempered in mixed stands due to stronger dilution. Neither research approach found a clear relation between tree species richness and understorey diversity. The presence of tree species identity effects may partially have skewed the diversity relations. However, results on interplot compositional dissimilarity indicated that creating mixtures using a chessboard pattern of monocultures may positively influence understorey diversity at the forest level. (C) 2014 Elsevier B.V. All rights reserved.