We applied our model (Larjavaara & Muller-Landau, 2012) to quantify the potential impacts of increasing temperatures on old-growth forest biomass (in preparation) and noted two errors caused by negligence in our original analysis. First, estimated potential evapotranspiration was computed erroneously and some plots were therefore incorrectly considered arid [11 in gross primary productivity (GPP) and two in total maintenance cost (TMC)] or humid [five in aboveground biomass (AGB)]. The corrected datasets are available in Appendix S1 in Supporting Information. Second, the acclimation term (1 − m|ΔT|) in Equation 1 of the original article was accidentally treated improperly and the parameterization of GPP was therefore carried out incorrectly. In particular, the equation to calculate this term for every temporal step included the term max(0, 1 − m|ΔT|), which was intended merely to prevent negative values, but as coded instead resulted in the maximum of (1 − m|ΔT|) over all time steps being used for every time step.
The best-fit parameter values for g changed to 4.26 × 10–5 (from 3.6 × 10–5), for c to 4.68 × 10–3 (from 4.8 × 10–3) and for h to 1.67 (from 1.66), while the remaining five parameters were unchanged (they all had their best fits at one of the bounds of the allowed ranges). The temperature dependence of GPP remained the same and that of TMC changed minutely (as h changed only a little). As expected, proper inclusion of the acclimation term in GPP parameterization improved the fit to GPP, with the r2 increasing from 57 to 68% (Appendix S2). In contrast, the r2 for the fit to TMC decreased from 79 to 73% due to the change in the datasets included (Appendix S2). Previously our model explained 50% of the variation in the old-growth AGB in an independent validation dataset. With the correction of the errors in the analysis, this percentage increased to 58% (Appendix S3). The revised sensitivity analysis (Appendix S4), like the original, revealed no substantial weakening of the explanatory power of our model for any of the tested variables. Correcting the errors increased the explanatory power of our model and strengthens further our conclusion that our simple model explains the key mechanisms of climate impacts on AGBmax.