Evaluating the Importance of Regional Characteristics and Land Cover on Water Use Efficiency and Climate Change across Kansas
Climate change predictions have shown that the Central U.S. will experience an increased occurrence of growing season droughts as a result of changes in the magnitude and timing of precipitation events. Precipitation is a significant driver of ecosystem carbon and water cycling. However, there are inherent nonlinearities between water, energy, and carbon cycling in ecosystems. This study explores how water use efficiency (WUE = rate of carbon gained to water lost) varies between sites with differing land covers and uses, precipitation gradients, and microclimates. The eddy-covariance method is employed at four different grassland sites across Kansas to assess the flows of water, energy, and carbon at the land-atmosphere interface. WUE was calculated from the linear regression line between net ecosystem exchange and latent heat flux for annual, monthly, and 5 day periods.
Each site appeared to respond to land cover, land use, and regional characteristics in different manners. Sites experiencing woody encroachment became more WUE while the site that wasn't became less. Burning tends to inhibit woody encroachment, creating a more homogeneous land cover, and resulting in a lower WUE at the site that was burned more frequently. WUE at sites that were consistent in species composition mainly responded to environmental conditions such as precipitation and temperature differences.
A low-dimensional model was also employed in order to evaluate the impact of climate change. Precipitation values for a normal year were used from one of the sites and parameters in the model were adjusted to try to fit the observed values of evapotranspiration, carbon assimilation, and WUE for that year. Precipitation values for a dry year were then used to see how well it represented climate change. The timing and magnitude of rainfall largely impacted the accuracy of the model.
- University of Kansas Field Station (KFS). Located 8 km north of Lawrence, Kansas (39N, 94W). The site is located within a tallgrass prairie deciduous forest ecotonal area. The Ameriflux tower was installed in 2007 on a restored prairie that experiences burning every 3 years and mowing for maintenance. KFS currently contains a mixture of C3 and C4 grasses and is experiencing woody encroachment. Data available from June 2007 to December 2013.
- Land Institute (SLN). Located south of Salina, Kansas (38N, 97W). The tower is located in a kernza field that has had over two decades of sustained breeding work. Kernza is a perennial wheatgrass that is being bred to use as a grain crop. Data available from May 2012 to December 2013.
- Konza Prairie Biological Station (KZU). About 8 km south of Manhattan, KS (39N, 96W) in the Flint Hills Ecoregion. Site KZU is on an annually burned, non-grazed watershed in an upland topographic area. The soils are rocky and thin, of the Florence series. KZU is primarily dominated by native C4 grasses. Data available from August 2006 to December 2012.
- Konza Prairie Biological Station (K4B). Primarily dominated by native C4 grasses, but experiencing woody encroachment by woody C3 species. Soil depth is greater than at KZU and soil is of the Tully series. K4B has experienced controlled burns every 4 years since 1975 but also had unexpected wildfires in 1994 and 2000. Data available from May 2007 to December 2011.
Water Use Efficiency:
Generally, water use efficiency reveals how efficiently carbon is assimilated per unit of water used, but the actual definition can vary depending on the scale. At the leaf level, WUE is defined as the ability of the canopy (or individual photosynthetic organ) to fix gross ecosystem production (GEP) for a given amount of water lost via transpiration (TR) through the stomata. At the ecosystem level, WUE is defined as the ratio between net CO2 fixation of the ecosystem (NEE) and water lost through evapotranspiration (ETR) (Tallec et al., 2013). A common method for calculating WUE is by taking the slope of the regression line between NEE and LE. A conclusion drawn from this definition is that ecosystems with high rates of carbon uptake use more water (Baldocchi et al., 2001).
WUE is a very important variable to consider, especially over grasslands, which are not only among the most important globally for long-term carbon storage, but also among the most responsive of terrestrial ecosystems to inter-annual variability in precipitation (Knapp et al., 2008). At the leaf level, C4 species generally have higher water use efficiencies than that of C3 due to photosynthetic processes, but heterogeneous landscapes and differing microclimates can alter this assumption.
FLUXNET is a global network of over 140 sites with a variety of ecosystems that measures micrometeorological fluxes. The FLUXNET project aims to quantify the spatial and temporal differences in carbon, water, and energy flux densities within and across natural ecosystems and climatic gradients. By using the eddy-covariance method, turbulent fluxes of mass and energy exchange between the surface and atmosphere can be measured. Vertical flux densities of carbon dioxide, latent, and sensible heat are proportional to the mean covariance between vertical velocities and their respective scalar fluctuations. This is known as the Eddy Flux equation. Standard instrumentation at FLUXNET sites include a three-dimensional sonic anemometer for wind velocities and virtual temperature, a fast responding sensor for carbon dioxide and water vapor measurements, and open- and closed-path infrared gas analyzers for scalar concentration fluctuations. These samplings are taken at rates between 10 and 20 Hz to ensure complete sampling of the high-frequency portion of the flux cospectrum (Baldocchi et al., 2001). Additional meteorological measurements such as net radiation, soil moisture, and soil heat flux are taken at half hour averages in order to illustrate the atmospheric conditions when the fluxes are measured. The sensors are typically mounted on small poles, the height of which depends on the height of the vegetation, fetch, range of wind velocity, and the frequency response of the instruments (Baldocchi et al., 2001).
Post-processed and gapfilled data was obtained from the four Ameriflux sites. A unit conversion was done for Latent Heat flux (LE) and Net Ecosystem Exchange (NEE) to convert from units of W/m2 to kgH2O and umol/m2/s to gCarbon. The method for defining water use efficiency was as indicated in Baldocchi et al. (2001) as the slope of the linear regression line between NEE and LE (see Figure 1 for example). Only daytime values (10 AM to 2 PM) were used in this analysis and annual, monthly, and 5 day period WUE's were calculated.
The low-dimensional model from Petrie et al. (2012) was implemented at a daily time step where soil moisture and potential evaporation were the dominant variables used for simulating daily evapotranspiration. Soil moisture was influenced by precipitation inputs at the beginning of the time step and reduced at the end of the time step from the magnitude of the mass fluxes as well as a soil leakage function. This reduced soil moisture value was then used for the next time step. The precipitation from 2009 at KZU was used as the normal year and the model was altered in order to provide the best fit to the observed evapotranspiration. Parameters that were adjusted included wilting point, field capacity (both obtained from observed data), soil active depth, and maximum evapotranspiration. The dry year at KZU (2012) was then input into the model to survey its accuracy.
Results and Discussion
Below is a table showing the meteorological data for all sites
Below are the WUE plots for annual (Figure 2), monthly (Figure 3), and 5 day period (Figues 4-7) WUE's
Ultimately, each site seems to be impacted by regional characteristics, land cover, and land use in different proportions. Since KZU is primarily a C4 grassland, it is most impacted by environmental conditions. Despite K4B being located in the same geographic region as KZU, its surface heterogeneity and species composition largely influences the variability in WUE. KFS likely responded to its land cover, land use, and regional characteristics collaboratively as it experienced higher WUE. Site SLN is difficult to ascertain which variable is impacting WUE at that site the most due to the short record of measurements, but the land cover and land use are most likely the dominant factors due to it being a perennial grain crop. Further studies at SLN would be enlightening to survey, for instance comparing WUE against regular row crops.
Below is the annual precipitation for 2009 (1055 mm) and 2012 (558 mm) at KZU
Below is the comparison of observed and modeled evapotranspiration for both years
Results from the model were interesting to evaluate, especially in first two comparison plots. Surprisingly, the model did better in the drought year for the first half of the year. This could be due to the model responding better with earlier precipitation inputs since 2012 seemed to have more of in the beginning of the year than 2009 did. The sporadic timing and magnitude of precipitation in 2012 likely is the reason the model failed to accurately depict the observed evapotranspiration throughout the whole year though. Because of climate change predictions showing those exact consequences, future studies pursuing this model would be extremely advantageous.
- Madison May
- University of Kansas, entering senior year
- Major: Hydrometeorology
- Minor: Mathematics
- Hometown: Andale, Kansas
- This material is based upon work supported by the National Science Foundation under Award No. EPS-0903806 and matching support from the State of Kansas through Kansas Board of Regents.
- Mentor: Nathaniel Brunsell, Associate Professor, University of Kansas Dept. of Geography-Atmospheric Sciences
- Amber Campbell Hibbs, Program Coordinator Kansas NSF EPSCoR Climate Change and Mitigation Project
- Baldocchi, D., Falge, E., Gu, L., Olson, R., Hollinger, D., Running, S., et al. (2001). FLUXNET: A new tool to study the temporal and spatial variability of ecosystem-scale carbon dioxide, water vapor, and energy flux densities. Bulletin of the American Meteorological Society, 82(11), 2415.
- Brunsell, N. A., Nippert, J. B. & Buck, T. L. (2013). Impacts of seasonality and surface heterogeneity on water-use efficiency in mesic grasslands. Ecohydrology. doi: 10.1002/eco.1455
- Knapp, A. K., Briggs, J. M., Collins, S. L., Archer, S. R., Bret-harte, M. S., Ewers, B. E., et al. (2008). Shrub encroachment in north american grasslands: Shifts in growth form dominance rapidly alters control of ecosystem carbon inputs. Global Change Biology, 14(3), 615-623.
- Petrie, M., Brunsell, N., & Nippert, J. (2012). Climate change alters growing season flux dynamics in mesic grasslands. Theoretical and Applied Climatology, 107(3), 427-440.
- Tallec, T., Beziat, P., Jarosz, N., Rivalland, V., & Ceschia, E. (2013). Crops' water use efficiencies in temperate climate: Comparison of stand, ecosystem and agronomical approaches. Agricultural and Forest Meteorology, 168(0), 69-81.