Drivers of tree communities in naturally colonised woodlands

Governments around the world are committed to increasing tree cover to address the climate and biodiversity crises. There is a growing interest in using natural processes, in addition to tree planting, to achieve this; however, evidence to support its use in temperate systems is limited. Here, we aim to identify the drivers of tree‐community development in young naturally colonised woodlands.

At 10 young naturally colonised woodlands (site age range: 9–22 years) across lowland England, we sampled tree size, species and locations, seedling and sapling presence and herbivore pressure.

We found that although highly variable, natural colonisation is likely to create sparse woodlands, in a relatively short time scale (16 years). Models predicted that during the early stages of woodland development, tree density increased with age but declined significantly with distance from a potential seed source. Tree‐community composition analysis identified three distinct tree species communities: willow‐birch, oak‐birch and ash‐field maple‐thorn communities and tree communities were statistically significantly different from the adjacent mature woodlands. As such, an additional function of the existing woodland during the initial colonisation is likely to enhance conditions for tree establishment in the adjacent open habitat (e.g. shelter, shade, more similar soil conditions to woodland). Further, oak was the only common species that displayed dispersal limitation in our sites with reduced colonisation beyond 80 m.

Practical implications : Low‐density woodlands, dominated by pioneer species, will start to develop naturally adjacent to a potential seed source after approximately 16 years (average site stem density 969.95 stems ha ⁻¹ , range between 74.3 to 2641.9 stems ha ⁻¹ ; site age range between 9 and 22 years). Over time, tree density and diversity will increase, but interventions such as supplementary planting or direct seeding, particularly further from a potential seed source, could be implemented to speed up the process. This work demonstrates that natural colonisation can contribute to meeting national woodland creation targets but should be spatially targeted adjacent to mature woodlands and that further uptake of natural colonisation will ensure a diverse mix of woodland at landscape and national scales.