QUESTIONS : What drives canopy gap formation in regenerating coastal dune forest? Does canopy gap size frequency distribution differ between new and old‐growth forests? Can canopy gaps divert regenerating trajectories?
LOCATION : Rehabilitating coastal dune forest, KwaZulu‐Natal, South Africa.
METHODS : We mapped canopy gaps in regenerating dune forest patches of varying age, which develop after seeding of the pioneer Vachellia kosiensis, and a reference forest by means of unsupervised classification of multi‐spectral satellite images. We tested if gap formation can be explained by abiotic (exposure to winds) and/or biotic (tree density at early stages) variables. We calculated the scaling exponent of a power‐law model to quantify if gap size frequency differed between new and old‐growth forests. Finally, we measured canopy openness, tree height and number of stems to validate canopy gap classification and assess the consequences of canopy gaps on regenerating trajectories.
RESULTS : Exposure to winds and tree density at early stages were both significant predictors of gap presence. Gaps were more likely to be present along dune ridges and areas with low tree density than valleys and areas with high tree density at early stages. Large gaps were common in both new and old‐growth forests. The scaling exponent was positively related to regeneration age, indicating a reduction in gap size as the forest aged. Areas with open canopies had shorter individual trees that were more likely to be multi‐stemmed than areas with closed canopies.
CONCLUSION : Canopy gaps are an important component of new and old‐growth coastal dune forests dynamics. Although gaps are filled during forest development, changes in individual life‐history strategies and morphology may alter regeneration trajectories. However, because their effect is partly controlled by the exposure to wind and tree density at early stages, it can also be manipulated by management to ensure nucleation, which could accelerate forest recovery.
Appendix S1. Overview of the study area.
Appendix S2. Mixture distribution of canopy openness.
Appendix S3. NDVI values at 10 years since rehabilitation.