Grassland salinization causes considerable changes to soil and vegetation, which can lead to changes in soil organic carbon (C) and total nitrogen (N). These changes have complex causal relationships. A significant correlation between soil organic C and total N and any soil or vegetation property does not necessarily imply a significant direct effect of the property on soil organic C and total N. In this study, a field survey was conducted to investigate the changes in soil organic C and total N in grassland along a salinity gradient in Hexi corridor, China, and the direct and indirect effects of soil and vegetation properties on both stocks were quantified using a path analysis approach. Significant decrease in soil organic C and total N contents were observed with increasing salinity. Both had significant positive correlations with the Normalized Difference Vegetation Index (NDVI), soil water, and fine particles (silt+clay) content (p<0.01) and significant negative correlations with soil EC, and sand content (p<0.01). NDVI, fine particles content and soil water content had positive direct effects on soil organic C and total N stocks. Soil EC affected soil organic C and total N stocks mainly through its indirect negative effect on NDVI, soil texture, and water content. NDVI, soil texture, and moisture also indirectly affected soil organic C and total N stocks via changes in each other. These indirect effects augmented each other, although in some cases indirect effects worked in opposing directions.
Pollen limitation of plant reproduction occurs across Angiosperms,particularly those in patched habitats.We investigated the relationship between pollen limitation and patch variables (patch size,visitation frequency) in the desert plant Hedysarum scoparium (Fabaceae),which is an important xerophyte in the arid and semi-arid regions of Northwest China and can grow well as a pioneer plant in shifting sand dunes.We observed insect visitation to H.scoparium over two flowering seasons and estimated pollen limitation using fruit set and seed production.Our results indicate that fruit set and seed production increased significantly with pollen supplementation compared with open pollination.Hedysarum scoparium was pollinated by over 8 species of bees,with 88.4% of visits made by introduced honeybees (Apis mellifera).Bee visitation varied significantly among the patches of habitats,but not associated with patch size of habitat.In general,pollen limitation occurred more strongly during fruit set than during seed production.The patches that received higher rates of pollinator visits were less pollen limited for fruit set.Pollen limitation for seed production,however,was not associated with pollinator visitation frequency.We conclude that pollen limitation in H.scoparium was caused by more than one reason,not just pollinator visits.