Effect of land-use legacy on the future carbon sink for the conterminous US

Abstract

Modeling the effects of the terrestrial carbon sink in the future depends upon not just current-day land use and land cover (LULC) but also the legacy of past LULC change (LULCC), which is often not considered. The age distribution of trees in the forest depends upon the history of past disturbances, while the nutrients in the soil depend upon past LULC. Thus, establishing the correct initial state of the vegetation and soil is crucial to model accurately the effect of biogeochemical cycling with environmental change in the future. This study models the effects of LULCC from 1750 to 2014 using the land-use harmonization dataset (LUH2) of land-use transitions with the terrestrial ecosystems model (TEM) for the conterminous US. Modeled LULC include plant functional types (PFTs) of potential vegetation, as well as managed cropland, pastureland, and urban areas. LULCC is treated using a cohort approach, in which a separate cohort occurs every year there is a land-use transition, thereby ensuring proper age structure of forests and regrowth with the correct soil nutrients. From 2000–2014 the modeled net ecosystem productivity (NEP) is 989 TgC yr−1 for the conterminous US but only −15 TgC yr−1 if accounting for carbon lost from land-use transitions and management. The hypothesis is that the initial state of the vegetation and soils significantly affects the future state of the terrestrial carbon sink. In this study, LULC remains constant in the future, with the NCAR CCSM4 RCP8.5 climate used to force the TEM-Hydro model. The following experiments are run from 2015 to 2100, including (a) restarting from existing cohorts in 2014 (RESTART), (b) reinitializing in 2015 based on condensing the cohorts for each PFT into a single cohort (CONDENSED), and (c) restarting from average cohort conditions for each PFT (AVERAGE). The NEP is too low when using condensed cohorts without reinitializing due to a larger increase in heterotrophic respiration (Rh) resulting from the assumption of mature forests. The carbon stocks are larger than using all the cohorts if condensed cohorts are reinitialized due to the assumption of mature, equilibrated forests. Where nitrogen-limited, forest regrowth is enhanced if regrowth starts from more nutrient-rich conditions. Water fluxes are dominated by environmental factors but can be slightly dependent upon the underlying carbon dynamics. It is therefore necessary to account for past disturbances when modeling future changes in carbon dynamics.