Spatial Analysis of In-Season Site-Specific Nitrogen Management Effects on Groundwater Nitrate and Agronomic Performance
No Thumbnail Available
Files
Date
2004-10-21
Authors
Journal Title
Series/Report No.
Journal ISSN
Volume Title
Publisher
Abstract
In-season, site-specific (SS) N management based on remote sensing (RS) has been suggested as a way of reducing groundwater NO3-N contamination. In-season N management seeks to match the temporal variability of crop N needs by applying appropriate amounts of N at critical crop growth stages. Site-specific N management attempts to match the spatial variability of crop N requirements by applying appropriate, spatially variable N rates within fields. We evaluated the environmental and agronomic benefits of two in-season, RS-informed N management strategies applied on a uniform field-average (FA) or SS basis. We compared these to current uniform N recommendations based on "Realistic Yield Expectations" (RYE) in a typical coastal plain cropping system. We also sought to understand the spatial and temporal dynamics of shallow groundwater NO3-N. An additional objective was to develop a statistical procedure for the analysis of spatially dense, georeferenced subsample data in randomized complete block designs, a common characteristic of precision agriculture research. The experiment was established in a 12-ha North Carolina field with a 2-yr winter wheat double-crop soybean-corn rotation. The three N management treatments were applied to 0.37 ha plots in a randomized complete block design with 10 replications. Groundwater NO3-N and water table depth were measured every two weeks at 60 well nests (two per plot) sampling 0.9- to 1.8-, 1.8- to 2.7-, and 2.7- to 3.7-m depths from 2001 to 2003. We developed a statistical procedure for selecting an appropriate covariance model in randomized complete block analyses in the presence of spatial correlation. When warranted, incorporating spatial covariance in the statistical analysis provides greater efficiency in estimating treatment effects. Elevations, soil organic matter (SOM), and water table elevations (WTE) were spatial covariates used for explaining NO3-N spatial correlation. Compared to RYE, SS achieved: (i) less groundwater NO3-N by reducing fertilizer N and increasing the harvest N ratio (the ratio of N harvested in grain or forage to the total fertilizer N applied) for wheat in 2001; (ii) increased yield associated with higher N applied and decreased harvest N ratio for corn in 2002; and (iii) increased yield associated with similar fertilizer N and increased harvest N ratio for wheat in 2003. Overall, FA performed similarly to SS for wheat, but differed greatly for corn due to an overapplication of N at tasselling. These results indicate that RS-informed SS and FA might improve groundwater quality with no sacrifice in yield, or increase grain yield with similar fertilizer N compared to RYE-based N recommendations in the Coastal Plain. Mean NO3-N concentrations averaged over sampling depth at each well nest showed clear temporal fluctuations and were positively correlated with WTE. Groundwater NO3-N was frequently spatially correlated and spatial covariance structure changed periodically. The spatial correlation range varied over time from 46 to 551 m, and appeared to follow the trend of the mean water table depth. Blocking alone or together with elevation, SOM, and WTE frequently explained NO3-N spatial correlation. Our data suggest that to assess the environmental efficacy of N management, frequent and periodic monitoring of groundwater NO3-N, especially after significant rainfall, is essential to capture in-season treatment effects. Simultaneous measurement of precipitation and water table depth facilitate understanding of these effects. The traditional sampling of NO3-N only at or after harvest is likely to be insufficient to capture the entirety of treatment effects throughout the growing season. This is especially true in coastal plain and other coarse-textured soils where in-season NO3-N leaching may be pronounced. Our data also suggest that residual effects of differential N management may appear long after N application, even on these coarse-textured soils, indicating a need for longitudinal sampling.
Description
Keywords
remote sensing, precision agriculture, groundwater nitrate, In-season site-specific N management, covariance model selection, spatial analysis
Citation
Degree
PhD
Discipline
Soil Science
