Forest canopy structure is closely related to species diversity, crown packing efficiency, and ecological processes, while influencing ecosystem functions. However, most existing structural diversity indices only consider internal heterogeneity within the canopy but often neglect the heterogeneity of the external canopy. While, the external canopy heterogeneity can increase the exchange surface area between canopy and external atmosphere, capture more light, which is closely linked to tree growth and ecosystem functions. Comprehensively and accurately estimating canopy structural diversity is crucial in understanding the underlying mechanisms of productivity change. Therefore, in this study, we apply a topographic index – canopy surface rugosity (CSR) – to capture canopy external heterogeneity, and use crown complementary index (CCI) and other canopy structural variables to quantify canopy internal heterogeneity. We find that CSR varies significantly among quadrats (20 ​m ​× ​20 ​m) and it decreases with species diversity and wood net primary productivity (NPPwood). Species diversity increases NPPwood directly and indirectly through CSR and CCI. Functional diversity increases NPPwood through CSR and CCI. Community weighted mean of light demand decreases with NPPwood through CCI. The mediating effect of CSR on the relationship between species (functional) diversity and NPPwood is greater than the mediating effect of CCI. We find soil fertility positive effects on NPPwood mainly through species richness and CSR. Conversely, water availability affects NPPwood by a direct positive effect and indirect negative effects through species richness, CSR and CCI. Overall, our CSR and CCI provide strong support for a space-based niche partitioning mechanism regulating the relationship between species diversity and NPPwood. Additionally, CSR is sensitive to gap dynamics, suggesting mechanistic relationships between local disturbance, species diversity, and NPPwood.

In Central Europe, anthropogenic coniferous monocultures have been subject to conversion to more diverse mixed forests since the 1990s, however, they are still abundant across many forest landscapes. Artificial and natural tree regeneration both play a key role during conversion by determining the species composition and structure of the future forests. Many abiotic and biotic factors can potentially influence the regeneration process and its specific combinations or interactions may be different among tree species and its developmental stages. Here, we aimed to identify and quantify the effect of the most important drivers on the density of the most abundant regenerating tree species (i.e., Norway spruce and European beech), as well as on species and structural diversity of the tree regeneration. We studied tree regeneration in four former monospecific coniferous stand types (i.e., Norway spruce, Scots pine, European larch, and Douglas fir) in Southwest Germany that have been under conversion to mixed forests since the 1990s. We sampled tree regeneration in four growth height classes together with a variety of potentially influencing factors on 108 sampling plots and applied multivariate analyses. We identified light availability in the understorey, stand structural attributes, browsing pressure, and diaspore source abundance as the most important factors for the density and diversity of tree regeneration. Particularly, we revealed species-specific differences in drivers of regeneration density. While spruce profited from increasing light availability and decreasing stand basal area, beech benefited either from a minor reduction or more strikingly from an increase in overstorey density. Increasing diaspore source abundance positively and a high browsing pressure negatively affected both species equally. Our results suggest that humus and topsoil properties were modified during conversion, probably due to changes in tree species composition and silvicultural activities. The species and structural diversity of the tree regeneration benefitted from increasing light availability, decreasing stand basal area, and a low to moderate browsing pressure. We conclude that forest managers may carefully equilibrate among the regulation of overstorey cover, stand basal area, and browsing pressure to fulfil the objectives of forest conversion, i.e., establishing and safeguarding a diverse tree regeneration to promote the development of mature mixed forests in the future.