Woody species diversity is crucial for the resilience of forests under climate change. The early stages of regeneration, particularly after canopy disturbance, shape the composition of future forest. Light availability, browsing pressure and their interactions should be key drivers of woody species diversity, biomass and density, but are not well understood due to limited experimental setups. We used exclosures in a mixed broadleaf forest with high woody diversity to protect seedlings from browsing by roe deer ( Capreolus capreolus ), the predominant ungulate in Central Europe. In a full‐factorial design, we paired 75 exclosure (fenced) and control (unfenced) plots in both closed canopy stands and experimental gaps. We measured woody regeneration 1 and 4 years after the start of the treatments. Rarefaction‐extrapolation curves revealed that diversity of common (Shannon diversity) woody species was highest in sunny and shaded exclosures (7), compared to sunny (4.5) and shaded (3.5) controls, for species that had outgrown the 130 cm browsing‐susceptible height of roe deer. Stage‐structured matrix models showed that saplings ≤20 cm had the highest probability of reaching >130 cm within 3 years in sunny exclosures (1.51%), followed by sunny controls (0.32%), shaded exclosures (0.27%) and shaded controls (0.07%). Browsing resulted in homogenized woody regeneration, particularly in forest gaps. Increasing light availability did not compensate for diversity loss by browsing. Given the densities of roe deer in the study area—typical for many Central European forests—effective population control or fencing appears essential to maintain high diversity of future forest shaping and forestry relevant woody species, especially in open forests. Synthesis and applications . Our results provide strong evidence that in the current context of increasing tree mortality and subsequent light availability, roe deer act as a keystone species. It is crucial to implement browsing protection measures and control roe deer population before or immediately after canopy disturbance. Otherwise, the rapid growth of a few dominant, browsing‐resistant woody species will significantly reduce the woody plant diversity of future forests.

Forest microclimates, their dependence on forest structure, and their impact on biodiversity are crucial for future forest management under climate change. However, standard approaches for measuring forest microclimates do not capture within-plot heterogeneity, which, according to the habitat heterogeneity hypothesis, is a key driver of local biodiversity. We quantified horizontal and vertical microclimatic heterogeneity within 30 broad-leaved forest plots in Central Europe using a three-dimensional design with high spatial resolution. Moreover, we examined whether microclimatic heterogeneity differs among silvicultural treatments and whether it can be predicted using forest structure indices derived from laser scanning. Additionally, we explored the relationship between microclimatic heterogeneity and biodiversity. In the understory of canopy gaps, warm and cold habitats co-existed in close proximity, leading to a high horizontal microclimatic heterogeneity. In closed stands with high structural complexity, we found steep gradients of increasing temperature and vapor pressure deficit from the ground to the canopy during mid-day. Canopy cover and forest structural complexity were strong indicators of microclimatic heterogeneity. We found positive relationships between herb layer temperature heterogeneity and the diversity of plants, Hymenoptera, and Diptera. Our results demonstrate that single-point measurements fail to capture the substantial microclimatic heterogeneity within plots, potentially misrepresenting the conditions experienced by forest species. However, laser scanning provides reliable indicators for within- plot microclimatic heterogeneity. With canopy gaps featuring high horizontal microclimatic heterogeneity and promoting the biodiversity of several taxonomic groups, we argue that managing forests for maximized temperature buffering should not be the only strategy to conserve forest biodiversity. Highlights: High small-scale horizontal microclimatic heterogeneity in canopy gaps Steep vertical microclimatic gradients in closed-canopy forests Canopy cover and structural complexity: indicators for microclimatic heterogeneity Positive relationship between herb layer temperature heterogeneity and biodiversity