Fertilizer Management and Environmental Factors Drive N₂O and NO₃ Losses in Corn: A Meta‐Analysis

Abstract

Core Ideas Systematic review and meta‐analysis demonstrate key factors for reducing agricultural N losses. Nitrification inhibitors and side‐dress fertilizer N each reduce N2O losses by ∼30%. Temperature controls N2O emissions and precipitation controls NO3 leaching losses. Higher levels of soil carbon reduce NO3 losses, but increase N2O emissions. Lack of simultaneous data for N2Oand NO3 impedes understanding of tradeoffs and synergies. Effective management of nitrogen (N) in agricultural landscapes must account for how nitrate (NO3) leaching and nitrous oxide (N2O) emissions respond to local field‐scale management and to broader environmental drivers such as climate and soil. We assembled a comprehensive database of fertilizer management studies with data on N2O (417 observations, 27 studies) and NO3 (388 observations, 25 studies) losses associated with 4R fertilizer N management in North American corn‐cropping systems. Only one study measured both losses, and studies of N2O and NO3 differed by location, time period, and management practices. Meta‐analysis of side‐by‐side comparisons found significant yield‐scaled N2O emission reductions when SUPERU replaced urea or UAN, and when urea replaced anhydrous ammonia. Hierarchical regression models found near‐equivalent magnitude effects on N2O emissions of 1°C rise in average July temperature (+), increase in soil C by 10 g kg–1 (+), nitrification inhibitors (–), side‐dressed fertilizer timing (–), broadcast fertilizer (–), and 100 kg N ha–1 decrease in fertilizer rate (–). Average NO3 leaching response to 100 kg N ha–1 reduction in fertilizer rate (–) were comparable to effects of 100 mm less annual precipitation (–), 10 g kg–1 more soil C (–), or replacing continuous corn with corn‐soybean rotations (–). The large effects of climate and soil, and the potential for opposite reactions to some management changes, indicate that more simultaneous measurements of N2O and NO3 losses are needed to understand their joint responses to management and environmental factors, and how these shape tradeoffs or synergies in pathways of N loss.

Meta‐Analysis Constrained by Data: Recommendations to Improve Relevance of Nutrient Management Research

“…Because all confounding factors are not likely known before a study, it is not possible to know whether this assumption is valid or not. Multiple statistical methods estimating the same effect (Nummer, 2016;Qian and Harmel, 2016), and alternatively specified models within one method (Eagle et al, 2017), address this issue as well as other statistical assumptions, like normality and homogeneity of variances. When these alternative methods result in similar effect estimates, more confidence can be placed in the statistical outcome and inferences.…”

“…Published information was particularly lacking in drainage P studies (Christianson et al, 2016) and in Midwestern water quality studies testing enhanced-efficiency N fertilizers (Cook et al, 2015). Although the lack of direct side-by-side comparisons may be partially addressed with alternative modeling methods that can use data across multiple locations while correcting for other factors (Eagle et al, 2017;Qian and Harmel, 2016), this is only possible when data have been collected and reported on specific fertilizer management practices.…”

“…For example, Eagle et al (2017) found only one study that reported both N 2 O and NO 3 losses, making it difficult to identify management synergies and tradeoffs between these two environmentally sensitive loss pathways.…”

“…To test alternative fertilizer management practices, however, many studies use a locally standard rate (or source, placement, timing) as their "control." In fertilizer N source comparisons, some studies tested a new fertilizer source against locally standard sources or used different timing or rates with the new source than with the control (Eagle et al, 2017). Although this may make sense from an agronomic practice point of view, the variability in these standards complicates the comparison across studies and environmental conditions.…”

“…As an example of the value of reported data for meta-analysis, Fig. 2 illustrates the total number of studies that measured and reported N 2 O or NO 3 losses in North American corn-system fertilizer management studies, separated into different publication time periods (Eagle et al, 2017). Data from studies that measured only short-term losses or reported concentrations or fluxes without cumulative totals could not be used in the meta-analysis, and data that were reported as multi-treatment or multi-year averages posed other problems.…”