Gross primary production of a semiarid grassland is enhanced by six years of exposure to elevated atmospheric CO2, warming, and irrigation.

Abstract:

The goal of this study was to quantify interannual variation of gross primary production (GPP) and evaluate potential drivers of GPP with global change using the Prairie Heating and CO₂ Enrichment (PHACE) experiment in semiarid grassland in southeastern Wyoming. PHACE consists of the treatments: control, warming only, elevated CO₂ (eCO₂) only, eCO₂ and warming, and irrigation only. We expected that GPP would be most strongly influenced by interannual variability in precipitation under all PHACE treatments, soil water availability under eCO₂, and nitrogen availability. GPP data were obtained from paired measurements of net ecosystem exchange (NEE) and ecosystem respiration (R_eco; GPP = R_eco – NEE) made on 2-4 week intervals over six growing seasons (2007-2012). Soil temperature (T), soil water content (SWC), vapor pressure deficit (VPD), and photosynthetically active radiation (PAR) were continuously recorded at the plot (T, SWC) and site (VPD, PAR) scales. Annual, plot-level aboveground plant nitrogen content (N) was measured during peak biomass. We fit a non-linear light-response model to the GPP data within a Bayesian framework, and modeled the maximum GPP rate (G_max) and canopy light-use efficiency (Q) as functions of N and current and antecedent SWC, T, and VPD. The model fit the GPP data well (R² = 0.64), and regardless of the PHACE treatment the most important drivers of GPP were N (for G_max), VPD (G_max and Q), antecedent T (G_max), and antecedent VPD (Q). Model simulations predicted that annual GPP increased on average by about 16% with eCO₂, 14% with warming, 12% with eCO₂ and warming, and 23% with irrigation. For four of the six years, annual GPP was significantly affected by either eCO₂ alone or when combined with warming. The increase in annual GPP under irrigation was similar to the increase under eCO₂ during a dry year (2012), but irrigation stimulated GPP to a greater degree than eCO₂ during wet years (2008, 2009). Hence, increases in GPP under eCO₂ appear to be indirectly due to increases in SWC, especially under dry conditions. These results suggest that future climate scenarios will lead to more productive grasslands in semiarid regions, but the overall response of the C cycle and the potential for these systems to sequester greater C will depend on the magnitude and direction of both the R_eco and GPP responses.