Habitat and landscape outweigh stand age as drivers of species composition in managed temperate forests dominated by Quercus robur L. and Quercus petraea (Matt.) Liebl

Key Message Managed oak stands with Quercus robur L. and Quercus petraea (Matt.) Liebl. supported a range of natural forest communities differing in species composition despite regular forest management. Habitat was the most influential driver shaping species composition, followed closely by landscape, whereas stand age showed a relatively small contribution. Soil fertility and the proportions of oak stands and deciduous stands in the research plot surroundings were the strongest independent determinants of species composition. Context Oak forests dominated by Quercus robur L. and Q. petraea (Matt.) Liebl. represent various types of natural forests and provide high-quality timber. To maintain the natural resources of managed oak forests, it is essential to understand the factors that shape their species composition. Aims In this study, we evaluated: (1) the variability of forest types across managed oak stands; and (2) the influence of stand age, habitat and landscape variables on species composition. Methods We surveyed all vascular plants across all forest layers on 100 circular plots ( r = 10 m, 314 m 2 ) located in oak-dominated stands aged 41–180 years (the Bory Stobrawskie Forests, SW Poland). Plots were assigned to forest types using the Two-Way Indicator Species Analysis, then groups of coexisting species were identified within the forest types. Environmental determinants of species composition were identified using Canonical Correspondence Analysis and variation partitioning. Results We identified four forest types differing in structure and coexisting species groups. In the Canonical Correspondence Analysis, explanatory variables constrained 21.5% of the total species variation (adjusted 10.7%). Variation partitioning showed that habitat variables explained ~ 33% and landscape variables ~ 31.5% of the constrained fraction, with a substantial shared habitat–landscape component (~ 26%). Stand age contributed ~ 4% and its partial effect (~ 1.5%) was not significant after Bonferroni correction (p adj = 0.104). Among single variables, soil fertility, the proportion of oak stands, and deciduous stands in the research plot surroundings had the strongest unique effects on species composition. Conclusions The studied managed oak stands differ in species composition and correspond to natural forest communities. Species composition was shaped primarily by habitat and landscape variables rather than stand age, with soil fertility and the proportions of oak stands and deciduous stands in the plot surroundings being the strongest independent determinants.