OOS 19-5 - Agro-ecological transitions and hysteresis: Combining experiment with theory

Wednesday, August 9, 2017: 9:20 AM
Portland Blrm 253, Oregon Convention Center
Theresa Wei Ying Ong, Ecology and Evolutionary Biology, University of Michigan, Ann Arbor, MI and John H. Vandermeer, Department of Ecology and Evolutionary Biology, University of Michigan, Ann Arbor, MI
Background/Question/Methods

Returning urban lands to a state capable of supporting urban agriculture is a fundamentally different process than the conversion of forested lands or prairies into conventional agriculture. Soils are well-known to undergo hysteresis, a phenomenon where changing conditions in a forward direction cannot predict the state of the system when those same conditions are experienced in the reverse direction. Here we combine theoretical and empirical approaches to test whether changes to agricultural management can effect hysteresis. First we examine how correlations between parameters in simple nutrient-soil feedback models influence hysteresis. Then in a greenhouse experiment, we transition trays of corn seedlings from a completely organic fertilizer regimen to a completely inorganic fertilizer regimen and back, all while measuring effects on state variables including yield, biomass, water loss and nutrient leaching. We also test for the presence of unstable points within hysteretic curves by perturbing the system and monitoring it for sudden changes to the state variable.

Results/Conclusions

We find that increasing management intensity in simple theoretical models can result in a variety of hysteretic patterns including some with hidden stable states. We found several of these same patterns in our empirical results. When the system was perturbed at mid-intensive levels, many state variables experienced sudden shifts that were consistent with expectations for unstable points. Management-driven hysteresis appears to be both common and complex, particularly when two or more driver variables are correlated. Our results imply that past management conditions can have long-term effects on agricultural outcomes. Thus, the ability for soils to retain water, support high crop yields and reduce nutrient leaching cannot be predicted solely from current management practices. For urban gardens, soil impaction and contamination resulting from previous land-use is anticipated to make transitions to viable, productive farms much more difficult than competing conventional farms that begin with better soil conditions.