Quantifying environmental impacts of agricultural nitrogen management with soil block mesocosms

Abstract

In this dissertation, the effects of agricultural management on nitrogen (N) dynamics in a continuous corn cropping system are evaluated using newly designed “soil block” mesocosms. Quantifying and understanding patterns of N cycling and losses from agricultural fields is crucial to reduce costly N losses of high environmental and economic impact. The Midwestern US produces a remarkable amount of the world’s corn at the expense of large greenhouse gas emissions and nitrate pollution. Measurements of agricultural N dynamics examined include ammonia oxidizing bacteria (AOB), N removed in grain harvest, nitrous oxide (N2O) emissions, and leached nitrate-nitrogen (NO3--N) in water drained from the mesocosms. The first chapter broadly introduces limitations of previous methods examining agricultural N dynamics including: obstacles to measure drainage volume and composition; limitations of many current studies evaluating AOB as key mediators in the soil of the N cycle and subsequent N losses and finally, limitations of simple models to estimate soil N2O emissions. The second chapter evaluates the ability of in-field mesocosms containing intact soil profiles to reasonably replicate upper soil (0–10 cm) moisture conditions, crop (i.e. corn) yields, and drainage nitrate concentrations relative to nearby field plots with traditional tile drainage. Drainage volumes from the mesocosms are also evaluated against traditional tile-drained fields and as well as watersheds to compare how the mesocosm drainage compares with that of larger scaled field studies. The mesocosms enabled more consistent and rapid measurements of drainage than tile-drained field plots and watersheds to evaluate management effects on immediate loss from the soil profile while maintaining key N cycling conditions comparable to nearby fields. The third chapter investigates how AOB community composition and how specific amoA members are related to management (soil type, N application rate, the presence or absence of a microbial inoculum) and soil N cycling parameters (e.g. soil inorganic N). To investigate this, a diverse set of 78 primers are utilized on soils collected from the mesocosms. It was found that individual AOB members can have diverse responses to different N cycling parameters and may be able to function as biomarkers for past soil N transformations to help track N leakage from agricultural fields. The fourth chapter investigates how well N2O emission measurements from the mesocosms align with accessible models used to estimate N2O emissions for greenhouse gas inventories and under what conditions the models can estimate measured emissions. Here it was found that the accuracy of models can vary among years and N application rate. In years with higher surface soil moisture (27% by volume at 0-10 cm) after N application, the Intergovernmental Panel on Climate Change (IPCC) model estimating N2O with linear emission factors designed for wet climates was closest to measured N2O regardless of the N application rate. In years with low soil moisture (15-19%) the model closest to measured N2O emissions varied by N application rate. Estimates from nonlinear models based on inorganic fertilizer N inputs or estimates from a partial N balance method were closest to measurements. Under some scenarios model-estimated N2O was markedly close to measurements (within 0.003 kg N ha-1), however models generally lacked the ability to predict how N2O emissions changed with subtle changes in synthetic N application rate (differences of 45 kg N ha-1 or less). Generally, this dissertation concludes that the soil block mesocosms provide a robust way to evaluate nutrient management of agricultural systems and can function well to complement new methodologies to evaluate diverse aspects of management impacts.