Do environmental predictors affect the regeneration capacity of sandy habitats? A country-wide survey from Hungary

European countries are far from the 15% restoration target to be reached by 2020, partly due to the lack of large-scale studies at the national level that would help prioritise restoration efforts. We investigated the regeneration capacity and the determining environmental factors at the national level for three Pannonian sandy habitat types. The analysis was based on the Hungarian Vegetation Mapping database that includes a three-level regeneration capacity estimate of semi-natural habitats based on expert judgments after local vegetation mapping. We have selected fifteen environmental predictors that could possibly influence regeneration, including proxies for landscape naturalness, landscape context and abiotic factors. Using the decision tree method, we found that the local regeneration of open and closed steppes is primarily determined by habitat naturalness. For juniper-poplar stands the seasonality of the precipitation is the most important predictor and Natural Capital Index of sandy habitats, as a proxy for landscape naturalness, is the second. In case of neighbouring areas and abandoned fields, the regeneration is primarily affected by the sand content of the soil and the total local extent of habitats. Furthermore, grasslands and agricultural areas represent a potential for regeneration after abandonment. Our results show that in addition to habitat adequacy, proxies for landscape naturalness are the most important predictors of regeneration capacity. This implies that the future dynamics of habitats – and consequently ecosystem health and integrity – are determined primarily by the conservation of remnant natural and semi-natural areas and active restoration to increase the area or improve the state of semi-natural habitats.