Fine roots and above‐ground litterfall play a pivotal role in carbon dynamics in forests. Nonetheless, direct estimation of stocks of fine roots remains methodologically challenging. Models are thus widely used to estimate these stocks and help elucidate drivers of fine root growth and turnover, at a range of scales.
We updated a database of fine root biomass, necromass and production derived from 454 plots across European forests. We then compared fine root biomass and production to estimates obtained from 19 different models. Typical input variables used for the models included climate, net primary production, foliage and above‐ground biomass, leaf area index (LAI), latitude and/or land cover type. We tested whether performance could be improved by fitting new multiple regression models, and explored effects of species composition and sampling method on estimated fine root biomass.
Average fine root biomass was 332 g/m2, and necromass 379 g/m2, for European forests where the average fine root production was 250 g m−2 year−1. Carbon fraction in fine roots averaged 48.4%, and was 1.5% greater in broadleaved species than conifers.
Available models were poor predictors of fine root biomass and production. The best performing models assumed proportionality between above‐ and below‐ground compartments, and used remotely sensed LAI or foliage biomass as key inputs. Model performance was improved by use of multiple regressions, which revealed consistently greater biomass and production in stands dominated by broadleaved species as well as in mixed stands even after accounting for climatic differences.
Synthesis. We assessed the potential of existing models to estimate fine root biomass and production in European forests. We show that recalibration reduces by about 40% errors in estimates currently produced by the best available models, and increases three‐fold explained variation. Our results underline the quantitative significance of fine roots (live and dead) to the global carbon cycle.