Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
5743619 | Ecological Engineering | 2017 | 10 Pages |
Increasing anthropogenic interventions involving utilization and conservation of vegetated ecosystems has attracted considerable attention focused on large-scale assessments of the roles played by human-domination. The effects of specific anthropogenic measures have generally been ignored by large-scale assessments, which has probably resulted in irrational regulations and imprecise understanding of these specific measures. This paper aims to reveal the effect of a grassland restoration measure-fencing in northern China-as well as the spatial patterns of fencing efficiency. Spatially continuous Normalized Difference Vegetation Index (NDVI) data based on remote sensing was used to detect grassland vegetation changes during fencing periods. Using subsequent processes to smooth the impact of precipitation from the current year and by setting thresholds for identifying changed/unchanged NDVI, the spatial vegetation changes were converted into two groups of statistical data: the mean NDVI-increase value in each sample region and the pixel areas with different grassland change types in each sample region. The precipitation lag effect on NDVI increases was assessed by multiple comparison tests, and regions with a significant precipitation lag effect were removed from the fencing effect assessment. Finally, the areas with different grassland change types were related to changes in fenced area using regression analyses under different precipitation gradients. The results indicated that the precipitation lag effect caused by legacy moisture significantly affected the spatio-temporal vegetation changes. By excluding the regions with legacy moisture, the increase in fenced area facilitated an expansion of improved grassland and a reduction in degraded grassland. Fencing efficiency was maximized in the 250-300-mm precipitation zone, where the expansion rate of improved grassland area reached 0.70 (R2 = 0.54, p < 0.01). The second-best and least effective zones lay in the mean annual precipitation (MAP) >300-mm and MAP <250 mm zones, respectively. We conclude that a wetter climate does not automatically result in a better fencing effect, and legacy moisture will lead to an overestimation of the fencing effect. A rational fencing implementation should consider the resilience and degradation degree of steppes.