Nitrogen Losses: A Meta-analysis of 4R Nutrient Management in U.S. Corn-Based Systems

IPNI-2014-USA-4RM10

24 Mar 2015

2014 Annual Interpretive Summary


The specific aim of this data analysis project was to determine the impact of 4R N management on nitrous oxide (N2O) and nitrate (NO3) environmental losses relative to corn yield. Beginning in May 2014, published data were collected and summarized on 4R fertilizer N management (right rate, source, timing, and placement) in North American corn-based cropping systems.

We identified 4,400 research papers that mention fertilizer, N, or nutrient management in agriculture, or fertilizer-associated N2O or NO3 losses. After a review of titles and abstracts, the majority of studies were discarded because they were not about cropland, corn, or N losses, were outside North America, or addressed N losses and transport after the field. This triage found 237 studies remaining for further review; only 27 and 22 studies contained N2O and NO3 loss data and also reported corn yield. An additional nine N2O and five NO3 studies reported losses but not yield. The final database includes 417 observations of N2O emissions and 396 observations of NO3 leaching losses.

The limited available data make it difficult to identify possible trade-offs between N2O and NO3 losses. Only one study reported losses of both N2O and NO3. Across studies, the geographies for the N2O and NO3 data rarely overlap, and management practices for both types of field studies are diverse. For example, 60% of NO3 observations, but only 1.4% of the N2O data were from tile-drained fields. Forty percent of the N2O observations, and only 8% of the NO3 observations, were from no-till systems.

Fertilizer N rate and N excess response curves were developed for yield as well as N2O and NO3 losses. Crop yield increases as a function of N fertilizer application rate up to a peak, N2O emissions respond in an exponential fashion, and NO3 losses tend to be linear. Fertilizer source comparisons can be made with about half of the N2O observations, but only one NO3 study examined different sources. Placement or timing was compared in 15 to 19% of cases each (for both types of losses). With very little geographic overlap between N2O and NO3 data, as well as such limited data on placement and timing impacts, the best-calibrated process-based models may be the only somewhat reliable way to compare and contrast loss responses in the foreseeable future. Data gaps and specific research needs were identified.