Feedback from Plant Species Change Amplifies CO 2 Enhancement of Grassland Productivity

TitleFeedback from Plant Species Change Amplifies CO 2 Enhancement of Grassland Productivity
Publication TypeJournal Article
Year of Publication2012
AuthorsPolley, HW, Jin, VL, Fay, PA
JournalGlobal Climate Biology
Start Page2813
Keywordsfire and rangelands, technical reports and journal articles

Dynamic global vegetation models simulate feedbacks of vegetation change on ecosystem processes, but direct, experimental evidence for feedbacks that result from atmospheric CO 2 enrichment is rare. We hypothesized that feedbacks from species change would amplify the initial CO 2 stimulation of aboveground net primary productivity (ANPP) of tallgrass prairie communities. Communities of perennial forb and C 4 grass species were grown for 5 years along a field CO 2 gradient (250-500 microL/L) in central Texas USA on each of three soil types, including upland and lowland clay soils and a sandy soil. CO 2 enrichment increased community ANPP by 0-117% among years and soils and increased the contribution of the tallgrass species Sorghastrum nutans (Indian grass) to community ANPP on each of the three soil types. CO 2-induced changes in ANPP and Sorghastrum abundance were linked. The slope of ANPP-CO 2 regressions increased between initial and final years on the two clay soils because of a positive feedback from the increase in Sorghastrum fraction. This feedback accounted for 30-60% of the CO 2-mediated increase in ANPP on the upland and lowland clay soils during the final 3 years and 1 year of the experiment, respectively. By contrast, species change had little influence on the ANPP-CO 2 response on the sandy soil, possibly because Sorghastrum increased largely at the expense of a functionally similar C 4 grass species. By favoring a mesic C 4 tall grass, CO 2 enrichment approximately doubled the initial enhancement of community ANPP on two clay soils. The CO 2-stimulation of grassland productivity may be significantly underestimated if feedbacks from plant community change are not considered.