There is increasing concern about plant nutrient losses to the environment. Such losses arise through leaching to the groundwater and emissions to the atmosphere. Mineral fertilizer is one of the sources of plant nutrients used in crop production systems. Applied nutrients that crops fail to take up constitute both a financial loss to the farmer and a considerable economic loss at the national level. With the demand for food and fibres projected to increase, albeit at a decreasing pace, farmers will apply more fertilizer in the attempt to boost agricultural production. The implementation of mitigating measures will probably focus on those nutrient sources that are easiest to regulate, with mineral fertilizer use being a prime target.
Reconciling the goals of ensuring ample food supply, enabling adequate farm income and reducing the adverse environmental impact of increased fertilizer use requires the adoption of sound agricultural practices. The adoption of balanced fertilization would be one of the measures for achieving higher yields without incurring accelerated nitrogen losses to the environment.
In order to develop the measures and implementation programmes for fostering efficient mineral fertilizer use, it is necessary to quantify the nutrient losses involved. The magnitude provides a benchmark of the potential benefits at farm, national and global level. The effective implementation of such programmes should focus on regions where losses are largest.
This report generates global estimates of nitrous oxide (N2O) and nitric oxide (NO) emissions and of ammonia (NH3) volatilization losses. The analysis establishes the relative contribution of the sustained losses from the application of mineral fertilizers and animal manure to croplands and grasslands. The report also identifies the regions where such losses are primarily sustained.
The supply of nitrogen (N) for growth often constrains yields of crops and forage species. Although there are wide differences between various countries and regions, more than 50 percent of the world's major food crops, sugar crops and some cash crops are routinely fertilized and high nitrogen fertilizer application rates are not uncommon. Moreover, fertilizer N is being increasingly applied to grasslands.
At present the global use of mineral fertilizers is 78 million t N per year, and farmers use large quantities of animal manure to fertilize crops and grasslands. The use of N fertilizer is expected to increase, particularly in developing countries, and the production of animal wastes may also continue to grow.
The use of N fertilizer by plants is rather inefficient. Plant uptake is commonly only 50 percent of the N applied. The major cause of this low N use efficiency is the loss of N from the plant-soil system via leaching, runoff, erosion, or by gaseous emissions. The relative importance of each of these pathways varies from site to site and from year to year. In climates where precipitation exceeds evapotranspiration or in irrigated fields leaching losses may be considerable. Water and wind erosion may occur in systems with bare fallow and in irrigation systems where water flows down the slopes from one field to another (e.g. wetland rice fields).
Gaseous loss of N is the dominant mechanism in many agricultural production systems. Loss processes include volatilization, nitrification and denitrification, resulting in the release of NH3, NO, N2O and dinitrogen (N2) to the atmosphere. The figure opposite illustrates the nitrogen cycle.
N2O is one of the so-called greenhouse gases, constituting 6 percent of the anthropogenic greenhouse effect, and contributes to the depletion of stratospheric ozone. Neither the sources nor the causes of the increase in N2O of 0.7 ppbv per year are well known. It is generally accepted that the most important source is natural soils, followed by emissions from the oceans (Table 1), although there is uncertainty regarding the distribution and magnitude of the sources themselves. Identified anthropogenic sources include: agricultural fields amended with N fertilizers, animal manure, aquifers, sewage, industry, automobiles, biomass burning, land clearing, and trash incineration. The contribution of agriculture to the global N2O source is about 35 percent.
Source: OECD (2001): Environmental indicators for Agriculture Volume 3: Methods and Results, Publications Service, Paris, France
NO also plays an important role in atmospheric chemistry as it participates in regulating the oxidant balance of the atmosphere. In the atmosphere, NO oxidizes to nitrogen dioxide (NO2). Re-deposition of NOx (NO and NO2 together) contributes to the acidification and eutrophication of ecosystems. The contribution of food production to the global NO emission may be about 10 percent (Table 1). The estimates for N2O and NO emissions from animal manure and mineral fertilizers are based on the so-called `fertilizer induced emission'. Therefore, they do not represent the total emission, as in their calculation the emission from unfertilized control plots is subtracted from that from the fertilized plot. The product of this report will be the total emission, which is more appropriate for global budget studies.
Food production is also a major source for NH3, contributing more than 50 percent to the global emission. NH3 is an important atmospheric pollutant with a wide variety of impacts. In the atmosphere, NH3 neutralizes a large portion of the acids produced by oxides of sulphur and nitrogen. A large proportion of atmospheric aerosols, acting as cloud condensation nuclei, consists of sulphate neutralized to various degrees by ammonia. Deposition returns most of the NH3 to the surface, where it may contribute to the acidification and eutrophication of ecosystems.
The effects of fertilizer application on the quantities of N2O released from agricultural soil have been the subject for a number of reviews in recent years. Other similar studies have focused on NO. However, because of the extremely high spatial, temporal and interannual variability, considerable uncertainty persists in the estimates of N2O emissions. The available data on NO emissions are more uncertain than those for N2O as fewer studies have focused on NO, mainly because the measurements are intrinsically more demanding. There are many reviews of ammonia volatilization from flooded rice fields, N losses including ammonia volatilization from fertilizers in general, and ammonia losses from grazing systems. Various authors have estimated country emissions on the basis of their measurements. The ECETOC (1994) and Asman (1992) have compiled inventories of European NH3 emissions.
Recent global estimates indicate that the global NH3 loss from mineral N fertilizers is about 9 million t of NH3-N, constituting more than 10 percent of global N fertilizer use (Table 1). However, the uncertainty in these estimates may be 50 percent.
A recent development is the use of simulation models to describe N2O and NO fluxes. Models can form a firm basis for extrapolation by developing an understanding of the mechanisms that regulate processes and their spatial and temporal patterns. Models help to break down a system into its component parts and describe the behaviour of the system through their interaction. In general, trace gas flux models include descriptions of the processes responsible for the cycling of carbon or nitrogen and the fluxes associated with these processes. Various types of models exist, including empirical and process (or mechanistic) models.
An example of a global model describing processes of N2O formation is an expanded version of the Carnegie-Ames-Stanford Biosphere model on a 1° resolution developed by Potter et al. (1996). This ecosystem model simulates nitrogen mineralization from net primary production, N uptake, litterfall and its decomposition. Fluxes of organic matter from litter and soil to microbial pools and from microbial pools back to soil pools occur in proportion to C assimilation rates so that prescribed C:N ratios for the various organic matter pools are maintained. The N2O (and NO) fluxes are a fixed fraction of 2 percent of nitrogen mineralization. However, the uncertainty of the model is unknown as the output has not been validated against flux measurements.
Li and Aber (2000) and Li et al. (1992a, 1992b) developed a field-scale process model to simulate N2O and NOx fluxes from decomposition and denitrification in soils. The model simulated gas fluxes by using soil, climate and management data to drive three submodels (thermal-hydraulic, denitrification and decomposition). The management practices considered include tillage timing and intensity, fertilizer and manure application, irrigation (amount and timing), and crop type and rotation. Comparison of observed flux data with fluxes simulated by models including the field-scale process model reveal major differences in the simulated N gas fluxes from soils. A major problem in developing trace gas flux models is the proper statistical description of soil processes that operate in `hot spots' in field-scale models.
Large-scale patterns of soil gaseous emissions, when aggregated to prolonged (e.g. seasonal or annual) time scales, may have a strong element of predictability. This is because at such spatial and temporal scales integrated gas fluxes may be strongly related to `average' biophysical conditions. Therefore, if estimation of seasonal or annual emissions is the objective, the use of empirical relationships between gas fluxes and environmental and management conditions represents a suitable approach for bridging the gap between site and landscape scales.
Although the approaches to estimating N2O and NO emissions and NH3 volatilization losses are distinct, they are sufficiently similar to appear together in this document.
First, this report analyses the direct emissions of N2O and NO from fertilized crops and grasslands. It does not consider indirect emissions from ground and surface water, or resulting nitrogen leached from agricultural soils. A review of measurement data from the literature forms the basis for calculating the estimates. An extrapolation exercise uses models for N2O and NO emissions that include the major controlling factors. The basis for this is a data set of measurements primarily from the reviewed literature. This report does not consider chemicals such as nitrification inhibitors as their use is still limited on the global scale. It also excludes experiments in grazed grasslands, as that system is not within the scope of this report. Furthermore, measurements of emissions from organic soils were excluded from the analysis. This report summarizes 896 measurements of N2O emissions from 139 studies and 99 measurements for NO emissions from 29 studies.
The experiments represent a range of different techniques for measuring fluxes for different crops and uncropped systems, different soil types, climates, fertilizer types and N application rates, and methods and timing of fertilizer application. Table 2 lists the available information on the environmental conditions, soil properties and measurement techniques for each measurement provided in the literature reports and included in the data set.
The various research papers used different methods to determine soil pH, cation exchange capacity (CEC), carbon content, texture, etc. In addition, they sometimes reported temperature as air temperature, and in other cases as soil temperature. This report ignored these differences and used the reported values as such.
For NH3 losses, the first objective of this report is to summarize the available literature on NH3 volatilization from the application of mineral N fertilizers and animal manure to crops and grasslands in order to assess the factors that regulate NH3 volatilization. A second objective is to describe the relationships found between regulating factors and NH3 volatilization rates in an empirical summary model, and to use this model for a global-scale extrapolation on the basis of national fertilizer statistics and land-use data.
The data set of measurements (primarily from the reviewed literature) contains data for almost 1 900 individual measurements. Table 2 presents details of the available information on the environmental conditions, soil properties and measurement techniques for each measurement provided in the literature.
This report does not use data on the effect of controlled-release and stabilized fertilizers and chemical additives used to reduce NH3 volatilization as their use also is still very limited on the global scale. In addition, the report does not include NH3 losses from stored manure and plants.
The analysis of the complete set of literature data to assess relationships between the various regulating factors and N2O and NO emission rates and NH3 volatilization rates made use of Genstat 5 release 4.1 (PC/Windows NT).
Contrary to many studies using regression analysis, this report used the Residual Maximum Likelihood (REML) directive of Genstat for summarizing the data set and developing models relating gas emissions to controlling factors. REML is appropriate for analyzing unbalanced data sets with missing values. By assuming all factor classes to have an equal number of observations, REML balances the emission for factors not represented by the full range of environmental and management conditions.
Chapter 2 discusses the factors that regulate N2O and NOx emissions and NH3 volatilization. Chapter 3 presents the different measurement techniques used in determining N2O and NOx emissions and NH3 volatilization rates, as it is important to know the characteristics of each technique for the interpretation of flux estimates in comparison with those obtained with other techniques. Chapter 4 analyses the data collected for this study in order to determine the major regulating factors and their effect on N2O and NO emissions and NH3 losses. Chapter 5 extrapolates the results up to global estimates based on fertilizer statistics and geographic data for the regulating factors. Finally, Chapter 6 presents some conclusions.