Nitrogen mineralization drives the response of forest productivity to soil warming: Modelling in ecosys vs. measurements from the Harvard soil heating experiment

Effects of climate warming on ecosystem productivity are strongly influenced by those on mineralization and uptake of key soil nutrients, particularly N. Models used to project these effects must therefore include a comprehensive and fully coupled N cycle which has undergone rigorous testing. Key hypotheses governing the response of the N cycle to warming in the terrestrial ecosystem model ecosys were tested against changes in ecosystem C and N stocks measured during 7 years of 5°C soil heating in the Harvard soil heating experiment. These hypotheses enabled the model to simulate gains in plant C stocks that rose from 50 to 200gCm−2y−1 driven by increased N mineralization and uptake as the soil heating experiment progressed. However these gains were offset by continuing losses of soil C stocks from 125 to 250gCm−2y−1 driven by increased heterotrophic respiration, so that total C stocks changed little with soil heating. Both gains and losses in the model were consistent with those measured in the soil heating experiment. The changes in N cycling with soil heating on which these model results were based were further corroborated by comparing modelled vs. measured increases in soil N mineralization rates, foliar N contents and reductions in root phytomass, and by comparing modelled increases in CO2 fluxes and net primary productivity with those reported in meta-analyses of warming effects on ecosystem productivity. However when the model was run under gradual climate warming by 5°C per century, these modelled changes in N cycling drove gains in ecosystem C stocks that rose gradually to ca. 165gCm−2y−1 after 100 years. The greater gain in ecosystem C modelled under gradual climate warming vs. sudden soil heating was attributed to much smaller losses in soil C stocks.