Agroecosystem Landscape Design for Food and Energy Production Assisted by Modeling
Open Access
- Author:
- Rozum, Rachel
- Graduate Program:
- Ecology (PHD)
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 18, 2022
- Committee Members:
- Chaopeng Shen, Outside Unit & Field Member
Erica Smithwick, Major Field Member
Jonathan Duncan, Major Field Member
Armen Kemanian, Chair & Dissertation Advisor
Jared Ali, Professor in Charge/Director of Graduate Studies - Keywords:
- Agroecosystem modeling
Nitrogen cycling
landscape design
bioenergy - Abstract:
- Industrial agriculture can involve tradeoffs between crop yield and nutrient pollution. Proposed landscape design strategies seek to maximize crop yield while minimizing pollution, but evaluating the potential results of these strategies is not straightforward, especially when they may be intertwined with economic, geographic, and regulatory considerations. Furthermore, in-situ evaluation of landscape designs is not possible at watershed or statewide scales prior to implementation. This limitation can be at least partially overcome by using in-silico models to make agronomic predictions. In this dissertation, the agroecosystem model Cycles is used to study management alternatives in a landscape design framework in two regions. The first region is the Sugar Valley watershed in central Pennsylvania, where we focus on how agricultural management can mitigate the effects of supply chain disruptions and to evaluate the sensitivity of this region to availability of labor and agrochemicals. The second area is much larger in scale: the state of Iowa, which is a region with one the largest, continuous, and most productive soils for corn and soybean in the world. Here, we evaluate the impacts of several complementary landscape design parameters, including corn stover removal rate, tillage, nutrient amendment application timing, cover crop planting, and perennialization. These practices tilt current system towards higher intensification (corn stover removal) or production relaxation (perennialization), but in all cases involve substantial changes compared with current conditions. In chapter 1 of this dissertation, we provide an introduction and background about nutrient modeling in agronomic systems. We introduce Cycles and discuss its key properties. We also describe the importance of properly accounting for how animal agriculture constrains manure distribution and therefore affects crop production, pollution, and landscape design. Chapter 2 discusses work submitted to the journal Ecosphere that models, at the field scale, how supply and demand pressures influence nutrient pollution and yield in Sugar Valley, PA. In chapter 3, we model approximately 871,000 fields across the entire state of Iowa using Cycles. We consider combinatorial changes in tillage, corn residue removal, fertilizer application type and timing, and winter cover with 48 management scenarios, and discuss their benefits and drawbacks across several metrics, including crop yield, soil organic carbon sequestration, nitrate leaching, ammonia volatilization, and nitrous oxide emissions. In chapter 4, we model the feasibility of widespread switchgrass planting in Iowa as a bioenergy feedstock and nitrogen pollution mitigation measure. We consider redistributing manure that is no longer necessary in fields that are switched from corn to switchgrass. Finally, chapter 5 summarizes our results and discusses possible future work. The results presented in these chapters underscore inherent tradeoffs between productivity, nitrogen pollution, and carbon sequestration when developing landscape designs, and how field level decisions couple to dynamic regional and global factors.