Regional Modelling of Grassland Biogeochemistry

Network News Spring 1988, Vol. 3 No. 1
Network News

Regional modelling is an essential step in.scaling plot measurements of biogeochemical cycling to global scales for use in coupled atmosphere/biosphere studies. We have developed a model for carbon (C), nitrogen (N), phosphorus (F) and sulfur (5) biogeochemistry for the U.S. Central Grassland region based on laboratory, field and modelling studies. Model results showed geographic patterns of cycling rates, plant production, and element storage to be a complex function of the interaction of climatic and soil properties. The driving variables for the model include soil texture (sand, silt and clay content), monthly precipitation, average monthly maximum and minimum air temperature, and plant lignin content.

The model was validated by comparing simulated soil C, N, F, and aboveground production to observed values for a large number of sites (16 to 56) in the Great Plains, which span large temperatures (4 to 22 C mean annual temperature) and moisture gradients (30 to 120 cm annual precipitation).

The model has been adapted for the four grassland LTER sites (Konza, Central Plains Experimental Range (CPER), Jornada, and Cedar Creek) and has been used extensively to simulate the interactive impact of fire and grazing at the Konza Tallgrass prairie and CPER sites.

The model (CENTURY) simulates soil organic matter (SOM) dynamics in natural and agroecosystems and represents the dynamics of C, N, P and S in the soil-plant system using a monthly time step. The plant production submodel simulates the dynamics of C, N, P and S in the live and dead aboveground plant material, live and dead roots, and the structural and metabolic surface, and soil residue pools. In the SOM submodel, plant residues are decomposed by microbes and the resulting microbial products become the substrates for SOM formation. We divide SOM into three fractions:

  1. An active soil fraction consisting of live microbes and microbial- products (1 to 4-y turnover  time)
  2. A protected fraction  that is more resistant to  decomposition (20 to 40-y  turnover time) as a result of  physical or chemical protection
  3. A fraction that is  physically protected or chemically  resistant and has a long  turnover time (800-1200  y)

The plant residue is divided  into structural (2 to 5-y turnover  time) and metabolic (.1 to  1 y turnover time) pools as a  function of the lignin to N ratio  of the residue. Decomposition  of each of the state variables  is calculated by multiplying  the decay rate specified for  each variable times the combined  effect of soil moisture  and temperature. The decay  rate of the structural material  is also a function of the lignin  content of the structural pool,  and the active SCM decay rate is a function of the soil  content of silt plus clay (low values for high silt plus clay soils).

The microbial respiration loss for each carbon flow is fixed for all the flows except the active SCM, which varies with the soil content of silt plus clay (decreasing with high silt plus clay content). A more detailed description of the model is presented by Parton et al. (1987 a, b). Regional patterns of ecosystem properties for grasslands in the Central grasslands region have been simulated (Parton et al., in press) and include: above- and belowground production and decomposition, soil organic C, N, P, and S levels, and N, P, and S mineralization rates. Figure 1 shows the simulated regional patterns for aboveground plant production and soil C levels (0- 20cm).

Aboveground production increases form west to east following the dominant precipitation trend. Soil organic C reflects the contrasting patterns for production and decomposition, and thus increase from the southwest to northeast. Soil organic C and N levels vary with soil texture and increase by 50% as you decrease the sand content from 75% to 25%.The results of this modelling work suggest a general approach to regional modelling of biogeochemical cycling. First, driving variables need to be identified, using laboratory or comparative field studies. Second, relationships between driving variables and rates of processes must be quantified. Third, the geography of the driving variables must be determined.

Finally, the regional simulation can be developed. By identifying environmental gradients in variables that have large effects on processes, much of the spatial variability in rates of processes may be explained. For example, a key driving variable with high spatial variance from our study is soil texture, which has virtually no temporal variance, but controls much landscape- and region-scale variation in productivity and soil organic matter accumulation. This approach provides a mechanism for extrapolating from process measurements typically conducted in the laboratory, or using field experiments to thousands or millions of hectares in area. Models like the CENTURY model are mechanistic enough to simulate effects of changing climate or atmospheric chemistry and will be an essential tool in the study of global change.

Literature Cited

Cole, C. V., D. Ojima, W. J. Peterson, J. W. B. Stewart and D. S. Schimel. Modeling land use effect on soil organic matter dynamics in the central grassland region of the U.S. Plant and Soil (in press).

Parton, W. J., J. W. B. Stewart and C. V. Cole. 1987a. Dynamics of C, N, P and S in grassland soils: A model. Biogeochemistry 5:109-131.

Parton, W. J., D. S. Schimel, C. V. Cole and D. Ojima. 1987b. Analysis of factors controlling soil organic levels of grasslands in the Great Plains. Soil Sci. Soc. Am. J. 51:1173-1179.

For additional information, contact Dr. William Parton, Department of Range Science & Natural Resource Ecology Lab, Colorado State University, Fort Collins, CO 80523.