Main content area

Inter-Annual Precipitation Variability Decreases Switchgrass Productivity from Arid to Mesic Environments

Reichmann, Lara G., Collins, Harold P., Jin, Virginia L., Johnson, Mari-Vaughn V., Kiniry, Jim R., Mitchell, Robert B., Polley, H. Wayne, Fay, Philip A.
BioEnergy research 2018 v.11 no.3 pp. 614-622
C4 plants, Panicum virgatum, biofuels, biomass, climate models, cultivars, ecotypes, energy crops, feedstocks, fertilizers, growing season, highlands, meta-analysis, perennials, plant available water, primary productivity, rain, soil types, temporal variation, North America
Cellulosic biofuels are an important source of renewable biomass within the alternative energy portfolio. Switchgrass (Panicum virgatum L.), a perennial C₄ grass native to North America, is widely studied as a biofuel feedstock for its consistently high yields and minimal input requirements. The influences of precipitation amount and temporal variability on the fertilizer response of switchgrass productivity are not fully understood. Moreover, global climate models predict changes in rainfall patterns towards lower and increasingly variable soil water availability in several productive areas worldwide, which may impact net primary production of biofuel crops. We conducted a meta-analysis of aboveground net primary production of switchgrass from 48 publications encompassing 82 different locations, 11 soil types, 52 switchgrass cultivars, fertilizer inputs between 0 to 896 kg N ha⁻¹ year⁻¹, and 1 to 6 years of annual productivity measures repeated on the same stand. Productivity of the lowland ecotype doubled with N rates > 131 kg N ha⁻¹ year⁻¹, but upland ecotype productivity increased only by 50%. Results showed an optimum N rate of 30 to 60 kg N ha⁻¹ year⁻¹ for both ecotypes, after which biomass gain per unit of N added decreased. Growing season precipitation (GSPPT) and inter-annual precipitation variability (inter-PPTvar) affected both ecotypes similarly. Long-term mean annual precipitation (MAP) differentially affected lowland and upland productivity, depending on the N level. Productivity responses to MAP and GSPPT were similar for both upland and lowland ecotypes at none or low N rates. When N increased beyond 60 kg N ha⁻¹ year⁻¹, lowland cultivars had a greater growth response to MAP than uplands. Productivity increased with increasing GSPPT and MAP and had a positive linear response to MAP ranging from 600 to 1200 mm year⁻¹. One third of the variability in switchgrass production was accounted for by inter-PPTvar. After accounting for MAP, sites with higher inter-PPTvar had lower switchgrass productivity than sites with lower inter-PPTvar. Increased inter-annual variation in precipitation reduced production of both ecotypes. Predicted changes in the amount and timing of precipitation thus likely will exert greater influence on production of upland than lowland ecotypes of switchgrass.