Methodological uncertainties in estimating carbon storage in temperate forests and grasslands

Carbon sequestration is an essential ecosystem service (ES) for climate change mitigation. For reasons of simplicity this ES is often quantified considering carbon storage in four carbon pools: aboveground biomass, belowground biomass, dead organic matter and soil organic carbon. Indicators of these four pools are estimated by modelling, reference values, or field methods and data processing of different complexity levels which requires comparing estimations. In order to facilitate the assessment of carbon pools, e.g. in environmental impact assessment, a fast, reliable and easily applicable method is required. First, using a systematic literature review we identified frequently used field methods for estimating carbon pools for forests and grasslands, two ecosystems playing a key role in global climate regulation. Second, from this review we developed field methods for indicators of each carbon pool - aboveground biomass, belowground biomass, soil organic carbon and dead organic matter - in both ecosystem types. We applied these methods in a set of forest and grassland plots in the Grenoble region (France) and asked i) how comparable and consistent are alternative methods for each carbon pool? ii) what is the variability of estimates between these methods? and iii) which level of simplicity has an acceptable level of uncertainty? Thereby, we conducted for the first time method comparisons for all four carbon pools. We based our method comparisons on the quality of the linear relationships between methods and their level of accuracy relatively to the chosen reference methods (the method assumed to be the closest to the actual carbon stock). For most carbon pools - e.g. aboveground biomass and soil organic carbon, both major carbon stocks - selected alternative methods were comparable and consistent with the reference method. Third, we built on these results to suggest easy and quick field methods for each carbon pool in each ecosystem type with accuracy levels between 10 and 20%. We provide guidelines together with associated uncertainty levels to scientists and practitioners aiming to estimate the ecosystem service of global climate regulation from carbon stocks in terrestrial ecosystems. The guidelines also allow adjusting method selection to human, knowledge and financial resources available in the study context.