Natural revegetation has dominated annual runoff reduction since the Grain for Green Program began in the Jing River Basin, Northwest China

The Grain for Green Program (GFGP) has successfully increased the vegetation cover in northern China over the last two decades through large-scale natural and artificial revegetation approaches. Unfortunately, an obvious runoff reduction has been observed in many basins, especially on the semiarid Loess Plateau. Past studies generally attributed this runoff reduction to the artificial revegetation of the GFGP, i.e., the land cover type change from cropland to forest, since there was no significant annual precipitation decrease. Such cognition has led to widespread concern about the hydrological effects of the GFGP. In fact, in addition to artificial revegetation, natural revegetation relies on natural processes of vegetation succession without land cover type changes directed by humans. Therefore, it is necessary to separate and quantify the contributions of natural and artificial revegetation to runoff reduction to assess the GFGP and optimize revegetation in the future. In this study, the contributions of natural and artificial revegetation to annual runoff change after the implementation of the GFGP were quantified in fifteen subbasins of the Jing River Basin in the central Loess Plateau. The mean annual runoff of the Jing River Basin decreased by 11.8 mm after the GFGP, although the climate (precipitation and potential evapotranspiration) inversely increased the mean annual runoff by 6.6 mm. This is due to that the direct effect of climate change was much less than the ecohydrological effect of revegetation on runoff reduction, i.e., the revegetation caused the mean annual runoff reduction of 18.4 mm. Natural revegetation (50.5% of basin area) was the dominant reason for runoff reduction rather than artificial revegetation (9% of basin area). The natural revegetated area with vegetation quality improvement, i.e., an NDVI increase without land cover type change, contributed to a runoff reduction of 15.5 mm, accounting for 131.1% of the total runoff reduction after the GFGP, and the grassland of natural revegetation was the main vegetation type, with runoff reductions of 9.8 mm. However, artificial revegetation by area increase and quality improvement, i.e., NDVI increase, contributed to a runoff reduction of 8.4 mm, accounting for 49.6% of the total runoff reduction. These results demonstrated that the GFGP needs to be properly adjusted, i.e., natural revegetated areas should be opened in a timely manner and properly utilized, and improving vegetation quality improvement rather than increasing vegetation area would be a priority management measure in artificial revegetated areas.