No abstract

Aim To investigate large-scale patterns of above-ground and below-ground biomass partitioning in grassland ecosystems and to test the isometric theory at the community level.Location Northern China, in diverse grassland types spanning temperate grasslands in arid and semi-arid regions to alpine grasslands on the Tibetan Plateau. MethodsWe investigated above-ground and below-ground biomass in China's grasslands by conducting five consecutive sampling campaigns across the northern part of the country during 2001-05. We then documented the root : shoot ratio (R/S) and its relationship with climatic factors for China's grasslands. We further explored relationships between above-ground and below-ground biomass across different grassland types. ResultsOur results indicated that the overall R/S of China's grasslands was larger than the global average (6.3 vs. 3.7). The R/S for China's grasslands did not show any significant trend with either mean annual temperature or mean annual precipitation. Above-ground biomass was nearly proportional to below-ground biomass with a scaling exponent (the slope of log-log linear relationship between above-ground and below-ground biomass) of 1.02 across various grassland types. The slope did not differ significantly between temperate and alpine grasslands or between steppe and meadow. Main conclusionsOur findings support the isometric theory of above-ground and below-ground biomass partitioning, and suggest that above-ground biomass scales isometrically with below-ground biomass at the community level.

Based on the data from China's second national soil survey and field observations in northwest China, we estimated soil organic carbon (SOC) storage in China and investigated its spatial and vertical distribution. China's SOC storage in a depth of 1 meter was estimated as 69.1 Pg (10 15 g), with an average density of 7.8 kg m -2 . About 48% of the storage was concentrated in the top 30 cm. The SOC density decreased from the southeast to the northwest, and increased from arid to semi-humid zone in northern China and from tropical to cold-temperate zone in the eastern part of the country. The vertical distribution of SOC differed in various climatic zones and biomes; SOC distributed deeper in arid climate and water-limited biomes than in humid climate. An analysis of general linear model suggested that climate, vegetation, and soil texture significantly influenced spatial pattern of SOC, explaining 78.2% of the total variance, and that climate and vegetation interpreted 78.9% of the total variance in the vertical SOC distribution.

AimOur objective was to document the general relationship between plant species richness (SR) and above-ground net primary productivity (ANPP) at different spatial scales and the environmental influence on this relationship.Location Temperate and alpine grasslands of China. MethodsWe investigated SR and ANPP at 321 field sites (1355 plots) across the widely distributed temperate and alpine grasslands of China. Ordinary least squares (OLS) regressions were used to test SR-ANPP relationships among site means. Plot-level data of SR and ANPP were analysed with general linear models (GLMs) and the correlation between SR and ANPP was decomposed into covariance components to test the influence of climatic variables, region, vegetation type and remaining variation among sites on SR, ANPP and their relationship. ResultsWe found positive linear relationships between SR and ANPP among sites in both the alpine and temperate grassland regions and in different grassland vegetation types of these biomes. Environmental gradients such as growing-season precipitation affected both SR and ANPP in parallel. However, after removing the among-site environmental variation, residual SR and ANPP were no longer correlated at the pooled within-site level. Main conclusionsThe positive SR-ANPP relationship across large-scale environmental gradients among sites was most likely the result of climatic variables influencing SR and ANPP in parallel. Our results suggest that in China's natural grasslands there is no direct relationship between SR and ANPP, presumably because the pool of available species for local community assembly is large, in contrast to experiments where species pools are artificially reduced.