Gross changes in forest area shape the future carbon balance of tropical forests

Bookkeeping models are used to estimate land-use change (LUC) carbon fluxes (ELUC). These models combine time 10 series of areas subject to different LUC types with response curves of carbon pools in ecosystems and harvested products after a unit change of land use. The level of detail of bookkeeping models depends on the number of response curves used for different regions, the carbon pools they represent, and the diversity of LUC types considered. The uncertainty of bookkeeping models arises from data used to define response curves (usually local data) and their representativeness of large regions. Here, we compare biomass recovery curves derived from a recent synthesis of secondary forest plots data by Poorter et al. (2016) 15 with the curves used in bookkeeping models from Houghton (1999) and Hansis et al. (2015) in Latin America. We find that both Houghton (1999) and Hansis et al. (2015) overestimate the long-term (100 years) biomass carbon density of secondary forest, by about 25%. We also show the importance of considering gross forest area change in addition to the net forest area change for estimating regional ELUC. To do so, simulations are constructed with a bookkeeping model calibrated with three different sets of response curves (linear, exponential and logarithmic) to study ELUC created by a pulse of net forest area change, 20 with different gross-to-net forest area change ratios (γ Agross Anet ). Following the initial pulse of forest area change, ELUC is subsequently calculated over 100 years. Considering a region subject to a net gain in forest area during one year, different values of gross forest area changes that sum up to this initial net gain can change the magnitude and even the sign of ELUC with a given time horizon after the initial forest area change. In other words, in the case of a net gain in forest area composed of a large gross loss and a large gross gain, the initial gross loss has an important legacy effect that the system can be a net source 25 of CO2 to the atmosphere. We show the existence of a critical value of γ Agross Anet above which ELUC switches from CO2 sink to source with a given time horizon after the initial forest area change. This critical ratio derived from the structure of the bookkeeping model is compared against real-world high resolution Landsat TM observations of gross forest area change in the Amazon to distinguish areas where current forest land turnover will legate LUC carbon emissions or sinks in 20 years, 50 years and 100 years in the future. 30 Biogeosciences Discuss., https://doi.org/10.5194/bg-2017-291 Manuscript under review for journal Biogeosciences Discussion started: 23 August 2017 c © Author(s) 2017. CC BY 4.0 License.