Predicted increased P relative to N growth limitation of dry grasslands under soil acidification and alkalinization is ameliorated by increased precipitation

Abstract

Global change potentially affects soil pH and thereby ecological stoichiometry. However, it remains unclear how changes in soil pH affect the total nitrogen (N) and phosphorus (P) stoichiometry of soils and vegetations, which are of critical importance in understanding ecosystem nutrient cycling. Based on a two-year study for the effects of soil acid and alkali additions on soil and vegetational N:P stoichiometry at nine sites representing three grassland types (i.e., desert, typical and meadow grasslands with increasing precipitation availability), we found that across all grasslands, soil acidification did not affect soil [N], [P] or their stoichiometry, while soil alkalization reduced soil [N] and thus N:P ratio. In desert and typical grasslands, soil acidification showed no effects on vegetation [N], but reduced vegetation [P], resulting in increased vegetation N:P ratios. In meadow grasslands, soil acidification increased both vegetation [N] and [P], leading to little effect on vegetation N:P ratios. Soil alkalization increased vegetation [N] in desert and typical grasslands while reduced vegetation [P] in desert grasslands. Soil alkalization increased vegetation N:P ratios in desert and typical grasslands. However, vegetation N:P ratios in meadow grasslands were not affected by soil alkalization. Under soil acidification, precipitation was positively correlated with vegetation [P] and negatively correlated with the vegetation N:P ratios. Standardized major axis regression showed that both soil acidification and alkalization induced faster accumulation of vegetation P than N. Pearson correlation and structural equation modeling revealed that soil and vegetation [P] were responsible for spatial differences in vegetation N:P stoichiometry under altered soil pH. These results indicate that precipitation modulates vegetation N:P stoichiometry under and possibly via soil pH changes in temperate grasslands. Dryer grasslands showed greater alterations of vegetation N:P ratios after soil pH changed than wetter grasslands. We thus conclude that unproductive and dry grasslands may be sensitive to altered soil pH. This study reveals that precipitation and soil water were critical to predict grassland geochemical cycle in response to altered soil acidity/alkalinity.