Past cover modulates the intense and spatially structured natural regeneration of woody vegetation in a pastureland

Vegetation natural regeneration after agricultural abandonment is changing the landscape patterns in many areas worldwide. However, the expansion rate, spatio-temporal dynamics, and the role of past vegetation cover in shaping such patterns are still barely quantified in fine and meso scales. Here, we aim to quantify the expansion rate and assess the spatio-temporal patterns and the effects of past cover on natural woody vegetation cover increase. We sampled woodland and shrubland cover from 1992 to 2016 in 30 ha in a formerly managed pastureland in Sicily, Italy. We combined field sampling, GIS tools, and spatial analysis to assess the spatial structure dynamics and test the effects of past cover amount and type and distance from forest or nearest woody patch on the proportional expansion of natural regeneration. After 24 years, woody cover increased 68%, despite the aggregated spatial structure in 1992 remaining almost unchanged in 2016. The past vegetation cover was the best predictor of woody vegetation expansion in two out of three plots. Distance to continuous forest and to the nearest woody patch, as well as cover type, was not relevant. Our study highlights the importance of fine- and meso-scale studies to reveal both the deterministic and stochastic facet of woody vegetation dynamics. Natural regeneration may strongly change landscape patterns even under constant herbivory pressure and long-term deforestation. The detection of cold and hotspots of regeneration provide an important prompt for the design of restoration programs and landscape management.