Above- and belowground biomass allocation and its regulation by plant density in six common grassland species in China

Above- and belowground biomass allocation is an essential plant functional trait that reflects plant survival strategies and affects belowground carbon pool estimation in grasslands. However, due to the difficulty of distinguishing living and dead roots, estimation of biomass allocation from field-based studies currently show large uncertainties. In addition, the dependence of biomass allocation on plant species, functional type as well as plant density remains poorly addressed. Here, we conducted greenhouse manipulation experiments to study above- and belowground biomass allocation and its density regulation for six common grassland species with different functional types (i.e., C-3 vs C-4; annuals vs perennials) from temperate China. To explore the density regulation on the biomass allocation, we used five density levels: 25, 100, 225, 400, and 625 plant m(-2). We found that mean root to shoot ratio (R/S) values ranged from 0.04 to 0.92 across the six species, much lower than those obtained in previous field studies. We also found much lower R/S values in annuals than in perennials (C. glaucum and S. viridis vs C. squarrosa, L. chinensis, M. sativa and S. grandis) and in C-4 plants than in C-3 plants (C. squarrosa vs L. chinensis, M. sativa and S. grandis). In addition to S. grandis, plant density had significant effects on the shoot and root biomass fraction and R/S for the other five species. Plant density also affected the allometric relationships between above- and belowground biomass significantly. Our results suggest that R/S values obtained from field investigations may be severely overestimated and that R/S values vary largely across species with different functional types. Our findings provide novel insights into approximating the difficult-to-measure belowground living biomass in grasslands, and highlight that species composition and intraspecific competition will regulate belowground carbon estimation.