96th ESA Annual Meeting (August 7 -- 12, 2011)

PS 65-123 - Use of an ecosystem process model to evaluate the effects of residue removal on long-term soil fertility in Upper Midwest cropping systems

Thursday, August 11, 2011
Exhibit Hall 3, Austin Convention Center
Natalie Dawn Hunt, Department of Forest & Wildlife Ecology, University of Wisconsin Madison, Madison, WI
Background/Question/Methods

The removal of crop residues from maize agricultural systems could provide a potential source of bioenergy feedstock for the Upper Midwest.  These crop residues come from harvesting corn-cobs or removing corn stover, which is typically either left on the soil surface for decomposition, or sold on the market.  The extraction of such biomass from an agricultural system removes additional carbon and nutrients from the system as well.  The aim of this research is to explore the long-term impacts of removing crop residues on long-term soil productivity. 

Estimation of these impacts involves the validation and application of the ecosystem process model AGRO-BGC.  Simulations were completed for three maize agricultural systems across the Upper Midwest, located near Madison, WI, Arlington, WI and Champaign, IL.  Model outputs of leaf area index (LAI), aboveground net primary productivity (ANPP), soil organic carbon content, net ecosystem productivity, and phenology were validated against measurements at each field site.  Experimental simulations included crop management practices such as fertilizer application, no tillage, grain harvest, and varying amounts of residue removal.

Results/Conclusions

Preliminary results show agreement between observed and simulated values of LAI, ANPP, and annual phenology.  For the Arlington simulation, we found reasonable agreement in LAI values between no-till simulations and field measurements, with all measurements falling within 30% of observed values.  AGRO-BGC tended to underestimate ANPP in simulations where no fertilizer was applied, and overestimated ANPP when fertilizer was applied.  Annual phenology was simulated over four growing seasons, and we observed reasonable agreement between expected and observed daily LAI.  Simulated leaf development shows agreement with expected maize plant phenology, with low values at planting followed by increasing leaf canopy area, until maximizing at plant maturation, then decreasing with leaf senescence.  Simulations with increased amounts of residue removal showed a decrease in soil organic carbon over fifty years.  We observed a range of 4 to 30 percent decrease with increased residue removal of 0 to 100% by mass.  Preliminary findings suggest the importance of incorporating conservation practices such as no till and conservative crop residue removal in order to maintain and improve the soil organic matter and soil carbon content of agricultural systems.