Scaling water use of perennial grasses from the plot to Midwestern U.S

Perennial rhizomatous C4 grasses are currently considered, in many regions, one of the most promising vegetation types to supply biomass for bioenergy production. Because one goal of bioenergy production is to benefit the environment, the potential environmental impacts and services that would be associated with large scale production must be investigated. Of particular interest is the impact that altering the composition of vegetation at the landscape scale would have on regional hydrological cycles driven by higher rates of evapotranspiration (ET). To assess this, we implemented micrometeorological measurements using two independent techniques over multiple growing seasons for replicated plots of two perennial rhizomatous grasses, Miscanthus giganteus (miscanthus) and Panicum virgatum (switchgrass), and two traditional crop species, Zea mays (maize) and Glycine Max (soybean), planted throughout Central Illinois. When averaged across the entire growing season, ET for miscanthus was double relative to annual crops, and 140% of P. virgatum ET. The differences between the perennial grasses and annual crops were primarily due to longer growing season associated with the perennial grasses, but Miscanthus also demonstrated higher instantaneous water use. These results, coupled with physiological measurements of these species, were then used to parameterize and validate a dynamic vegetation model, Agro-IBIS, to investigate the large-scale consequences of land-use change on ecosystem hydrology from existing agriculture to various perennial grass production scenarios ranging from 10% to 100%. Results show that uniform production scenarios of less than 25% have little impact on regional hydrology but ‘hotspots' with higher percentage land cover devoted to perennial grasses can have important consequences in localized areas. Given the increasing demand on water, we investigated whether the increased carbon gain associated with bioenergy crops was sufficient to offset the water used, thereby improving water use efficiency. Using in-field measurements coupled with ecosystem modeling as above, we calculated the WUE for Miscanthus and switchgrass, and compared it to maize. Results indicate that production increased more than water use for miscanthus compared with maize, thereby driving higher water use efficiency. Switchgrass, however, showed no improvement in WUE over maize. The results from these experiments show that water use of perennial grasses can impact hydrological cycling at various spatial scales depending on the percent displacement of existing vegetation. Sustainable implementation of perennial grasses as bioenergy feedstocks will vary based on percent land coverage and location, and in the case of miscanthus, can improve the efficiency in which water is used.