4.3 Density effects on above- and belowground biomass allocation and allometric relationships
Our greenhouse manipulation experiment revealed three trends in the RMF and R/S with increasing density among the studied species: statistically insignificant changes (C. glaucum , L. chinensis andS. grandis , Figure 4b, h, l), an increasing trend (C. squarrosa , Figure 4f) and fluctuating patterns (S. viridis andM. sativa , Figure 4d, j). This result suggests that density has different effects on biomass allocation across species. For C. glaucum , L. chinensis and S. grandis , the density-induced above- and belowground competition was symmetrical (Weiner & Thomas, 1986). For C. squarrosa , the RMF and R/S increased with increasing density, indicating that competition was more intense belowground than aboveground at higher densities; relatedly, a higher proportion of biomass must be allocated to roots to achieve more efficient nutrient and water uptake (Berendse & Möller, 2009). The relative intensity of above- and belowground competition in S. viridis and M. sativa varied with density. For example, M. sativa had the highest R/S at a moderate density (D3, 225 plant m-2). This density may be the threshold value for above- and belowground biomass allocation in M. sativa ; when planting densities are close to this value, this species is prone to allocate more biomass to its roots, which is not conducive to increasing forage yields.
Our results further showed that planting density significantly regulated the allometric scaling exponents of shoot biomass against root biomass for C. glaucum , M. sativa and S. grandis (Table 1; Figure 5). These patterns do not support the ontogenetically fixed scaling exponents predicted by allometric partitioning theory, which suggests that scaling relationships are insensitive to biotic or abiotic factors (McCarthy & Enquist, 2007; Müller et al., 2000; Weiner, 2004). Rather, the allometric relationships in this study were altered by the planting density. This inconsistency is likely due to that these three species do not possess the typical fractal structure (e.g., S. grandis has an expanded tiller node rather than a fractal structure) that is the basis for the vascular plant allometry model (Wang et al., 2010; West et al., 1999). In contrast, for S. viridis , C.squarrosa and L. chinensis , the planting density had statistically insignificant effects on the allometric scaling exponents of shoot biomass against root biomass; for M. sativa,the scaling exponents gradually decreased with increasing density (Table 1; Figure 5). This difference suggests that the effects of planting density on plant allometric relationships between above- and belowground biomass are species-specific. The stability or instability of the scaling exponent response to planting density might be attributed to the balanced or unbalanced limitations on light and nutrients under different competitive conditions (Poorter et al., 2015). Overall, the allometric partitioning theory derived from global data analyses of vascular plants is not applicable to the studied herbaceous plants, and the effects of planting density on allometric relationships should be considered when using predefined scaling relationships to quantify the root biomass of herbaceous plants (Yan et al., 2019).