Projecting the long-term biogeochemical impacts of a diverse agroforestry system in the Midwest

Annual, monoculture cropping systems have become the standard agricultural model in the Midwestern US. Unintended consequences of these systems include surface and groundwater pollution, greenhouse gas emissions, loss of biodiversity, and soil erosion. Diverse agroforestry (DA) systems dominated by fruit and nut trees/shrubs have been proposed as an agricultural model for the Midwestern US that can restore ecosystem services while simultaneously providing economically viable and industrially relevant staple food crops. A DA system including six species of fruit and nut crops was established on long-time conventional agricultural land at the University of Illinois at Urbana-Champaign in 2012, with the conventional corn-soybean rotation (CSR) as a control. Initial field measurements of the nitrogen and water cycles during the first two years of transition have indicated a significant decrease in N losses and modification of the seasonal evapotranspiration (ET) pattern. While these early results suggest that the land use transition from CSR to DA can have positive biogeochemical consequences, models must be utilized to make long-term biogeochemical projections in agroforestry systems.

Initial field measurements of plant phenology, net N₂O flux, nitrate leaching, soil respiration, and soil moisture were used to parameterize the DA system within the DayCENT biogeochemical model as the “savanna” ecosystem type. The model was validated with an independent subset of field measurements and then run to project biogeochemical cycling in the DA system for 25 years past establishment. Model results show that N losses via N₂O emission or nitrate leaching reach a minimum within the first 5 years and then maintain this tight cycle into the future. While early ET field measurements revealed similar magnitudes between the DA and CSR systems, modeled ET continued to increase for the DA system throughout the projected time since the trees would continue to grow larger. These modeling results illustrate the potential long-term biogeochemical impacts that can be generated by a land-use transition to a diverse agroforestry system in the Midwest.