Short communicationRestoration treatments to control Molinia arundinacea and woody and alien species encroachment in Calluna vulgaris heathlands at the southern edge of their distribution

- Six restoration treatments were applied in southern Calluna vulgaris heathlands.
- Annual fire shifted the vegetation towards a Molinia arundinacea-dominated community.
- In the short-term, fire treatments triggered a peak in alien species cover.
- Six years after treatments, plant diversity did not differ between treated and control sites.
- Six years after treatments, goat browsing and annual mowing provided the best results.

Throughout the last decades, Calluna vulgaris (L.) Hull heathlands have declined across Europe and nowadays their conservation is particularly challenging at the southern edge of their distribution. In the Nature Reserve of Vauda (north-western Italy), six restoration treatments were applied (extensive annual goat browsing, one-off mowing, annual mowing, one-off fire without and with annual browsing, and annual fire) and their effects on plant diversity and the cover of C. vulgaris, its competitor grass Molinia arundinacea Schrank, woody, and alien species were monitored between 2005 and 2011. In the short-term, most of the treatments changed the vegetation community, reducing C. vulgaris cover according to a gradient of increasing biomass removal. In the mid-term, C. vulgaris, M. arundinacea, woody and alien species cover followed different trajectories according to the treatment and functional group. Annual fire shifted the vegetation towards a M. arundinacea-dominated community, while extensive annual browsing did not affect the heathland community and resulted in the lowest increase in M. arundinacea, which showed a remarkable fitness in these environments. Moreover, annual burning and mowing were effective in reducing woody species encroachment (p < 0.05), and fire treatments triggered a peak in alien species cover (mainly Panicum acuminatum Swartz) in the short-term. Six years after treatment, species richness and Shannon index did not differ between treated and control sites (p > 0.05). In conclusion, these results highlight the need and potential benefit of integrating multiple techniques to preserve C. vulgaris heathlands at their southern edge.