Abstract
1. In herbaceous communities which host many perennial species, belowground clonal organs and traits remain largely overlooked in ecosystem functioning studies. However, the belowground compartment is expected to play a key role as the greatest proportion of biomass is allocated belowground. Our main goal was to test whether including underexplored clonal traits (in tandem with widely used aboveground traits) improves the ability to predict biomass production and soil carbon in temperate grasslands. We examined the relationships between five plant functional traits (canopy height, specific leaf area [SLA], lateral spread, multiplication rate, persistence of rhizome) and biomass allocation (aboveground, roots, rhizomes) and soil organic carbon (SOC).
2. We studied 52 grassland plots clustered in two Central European regions differing by management intensity. We assessed biomass allocated aboveground and belowground (into roots and rhizomes separately). SOC was measured from averaging five soil cores at each plot. Trait community means were calculated for each trait from species compositions using databases.
3. The most important traits in predicting biomass allocation were canopy height and persistence of rhizome. Results, however, differed for the two regions implying context dependency of the relationships. Persistence of rhizome and canopy height emerged as traits largely affecting ecosystem functioning at the less intensively managed grasslands-likely because of their links to biomass production and plant economics spectrum. At the more intensively managed grasslands, canopy height and SLA were negatively correlated with root biomass, with persistence of rhizome playing a (lesser) role. We found no significant predictors for SOC.
4. Synthesis. This study is the first attempt to integrate belowground clonal and aboveground traits to predict biomass allocation and soil carbon. Findings indicate that the ability of plant functional traits in predicting ecosystem functioning is highly context dependent, varying greatly even within the same community type. Only SLA and, to a lesser extent, persistence of rhizome consistently predicted root biomass in the two regions. We call for a broader implementation of the whole-plant multifunctional approach applied here, ideally extending to other open ecosystems.