Introduction
Almost 50 years ago, C. C. Delwiche () wrote in Scientific American “Of all man's recent interventions in the cycles of nature, the industrial fixation of nitrogen far exceeds all the others in magnitude.” Since then, the climate change impacts of human contributions to the carbon cycle have come under intense scrutiny and debate (IPCC, ). Meanwhile anthropogenic production of reactive nitrogen has continued to increase with associated adverse environmental consequences. As in the case of carbon dioxide (CO2), some impacts of reactive nitrogen are difficult to measure on a short timescale. Reactive nitrogen also has a place in the agricultural revolution, which is analogous to the role of carbon from fossil fuels in the Industrial Revolution. This leads us to ask the question: Is the anthropogenic augmentation of the nitrogen cycle growing to a point where it may have adverse environmental consequences on a global scale, and where the critical role of reactive nitrogen in the agricultural system will make it very difficult to mitigate these consequences? In short, “Is nitrogen the next carbon?” (Figure ).
This paper combines historic data on crop cultivation and fertilizer usage with recent estimates of nitrogen‐fixation rates in order to analyze trends in the anthropogenic production of biologically available nitrogen since 1900. The paper also places these trends in context with recent estimates of natural nitrogen‐fixation and removal of fixed nitrogen by denitrification in terrestrial and aquatic ecosystems.
Trends in Human Nitrogen use
Nitrogen‐rich manure has been used to fertilize crops for millennia—long before the discovery of nitrogen as a distinct chemical element (Columella, ). The role of nitrogen as a crop nutrient was explored scientifically in the 1800s (Boussingault, ), and a market developed for nitrogen‐rich deposits of mineralized guano from South America and the South Pacific islands, reaching about 1 Tg N y−1 by 1900 (Cushman, ). Though small by today's standards, the trade in nitrogen fertilizer was a main object of contention in the 1879 War of the Pacific, between nascent South American nations (Ortega, ). Industrial processes for fixing nitrogen, especially the Haber‐Bosch process, greatly expanded the availability of nitrogen‐based fertilizers in the early 1900s (Smil, ). Nevertheless, use of synthetic fertilizers did not become routine until the mid‐twentieth century. In the U.S., less than 40% of farms reported the use of any synthetic fertilizer in 1939, but more than 60% reported the use of chemical fertilizer in 1954 (Hurley et al., ).
Synthetic nitrogen fertilizer usage grew to about 12 Tg N y−1 in 1960, more than a tenfold increase over mineral nitrogen fertilizer usage prior to the commercialization of the Haber‐Bosch process (1911). Since 1960, synthetic nitrogen fertilizer use has grown almost another tenfold to 110 Tg N y−1 in 2013. Initially, this growth occurred in the U.S. and Europe. Consumption in these regions leveled off in about 1990, and the growth since then has occurred primarily in Asia (IFA, ).
The last half of the 20th century also saw large increases in the cultivation of nitrogen‐fixing crops, especially soybeans. These have been cultivated in Asia for at least 1000 years, but after the nitrogen‐fixing capabilities of the crop were recognized, cultivation in other parts of the world grew rapidly. World soybean production increased by more than a factor of 3 between 1910 and 1960 (Shurtleff and Aoyagi, ); with production in the U.S. exceeding Asian production by 1955 (FAO, ). Since 1960, world production of soybeans has increased by a factor of 10, with the largest share of new production occurring in South America. Soybeans are often grown in rotation with other crops, such as corn. This crop rotation serves to increase the amount of nitrogen available and to break the cycles of crop pests.
Cultivation of other legumes (e.g., peas, beans, lentils, and peanuts) has increased by a factor of about 2.6 since 1960. Total global nitrogen fixation in croplands is estimated at about 43 Tg N y−1, with a range from 30 to 51 Tg N y−1 based on the ranges of nitrogen fixation yields (Herridge et al., ). This estimate includes soybeans, other beans and legumes, and inadvertent nitrogen fixation by cyanobacteria associated with the cultivation of rice and sugar. The estimate does not include legumes in pasturelands or savannas used for grazing.
A substantial portion of the nitrogen is transported and lost to the surrounding environment, through runoff, leaching to groundwater, and emissions of ammonia (NH3), nitrogen oxides (NOX), and other nitrogen compounds (Galloway and Cowling, ; Conant et al., ). NOX emissions from combustion sources also constitute a source of reactive nitrogen. These emissions increased from approximately 5.8 Tg N y−1 in 1910 to approximately 38 Tg N y−1 in 2010. Since the 1990s, NOX emission sources have been subject to pollution controls in the U.S. and Europe, resulting in substantial emission reductions. However, emissions continue to increase in the developing world, especially China (UN, ).
The increased use of synthetic fertilizer, increased cultivation of nitrogen‐fixing crops, and increased emissions of NOX all contribute to increases in the overall level of reactive nitrogen in the environment. The total of these three sources for 2014 is 190 Tg N y−1, with a plausible range from 160 to 210 Tg N y−1. The trajectories of anthropogenic production of reactive nitrogen are shown in Figure , and are compared with the trajectory of CO2 (Boden et al., ; International Energy Agency (IEA), ). The CO2 curve shows a sharp upward inflection in the mid‐1940s. The nitrogen curve lags the CO2 curve by 10–15 years, showing an upward inflection about 1960. The trends for anthropogenic nitrogen and CO2 in the last half of the 20th century are similar. While reported global CO2 emissions have not shown an increase since about 2013, the use of nitrogen fertilizers and the cultivation of nitrogen‐fixing crops are expected to continue to increase in most future scenarios (Bouwman et al., ; Winiwarter et al., , International Energy Agency (IEA), ).
Fig. 2. Trends in anthropogenic reactive nitrogen sources since 1900 compared with the trend in anthropogenic CO2 emissions. Fertilizer trends are adapted from IFA (), Cushman (), and Smil (). NOX trend is adapted from UN (). Trends in nitrogen‐fixing crops are computed by combining nitrogen fixation yields (Herridge et al., ) with crop production statistics (Shurtleff and Aoyagi, ; FAO, ). CO2 trend is adapted from Boden et al. () and IEA ().
Populations of animals raised for food have also increased in recent decades. Animals convert only a fraction of the nitrogen in their feed to meat protein or milk protein for human consumption (
Human Impacts on the Nitrogen Cycle
Figure illustrates the natural nitrogen cycle, showing how anthropogenic inputs of nitrogen compounds contribute to this cycle. Nitrogen‐fixing bacteria convert nitrogen gas (N2) to biologically available nitrogen compounds, which are in turn taken up by plants and incorporated into proteins and other essential molecules. Nitrogen compounds in decomposing plant matter and animal waste are released as ammonium (NH4+), which is oxidized by bacteria, producing nitrate ion (NO3−), which is consumed by other bacteria that perform denitrification. The denitrification reaction completes the cycle by producing N2 gas, but also produces some gaseous nitrous oxide (N2O) and nitric oxide (NO).
Both the use of synthetic nitrogen fertilizer and the anthropogenic cultivation of nitrogen‐fixing crops increase the overall mass of biologically available nitrogen compounds. This nitrogen is intended to remain within the farm system, ultimately for the production of food for humans or for animals raised to feed humans. However, a substantial portion of the reactive nitrogen (∼70%) escapes to the surrounding environment (Galloway and Cowling, ). Fertilizers and residues from nitrogen‐fixing crops, and deposited nitrates all ultimately increase the amount of nitrate processed by denitrifying bacteria, both on farmlands and in surrounding ecosystems.
Fig. 3. The nitrogen cycle. Orange arrow depict anthropogenic additions to the natural cycle.
The magnitudes of fluxes in the natural nitrogen cycle are subject to considerable uncertainty, as stated colorfully in the title of a paper by Burris: “The global nitrogen budget – science or séance?” (Burris, ). Galloway et al. () gave a global estimate of 120 Tg N y−1 for biological nitrogen fixation in terrestrial ecosystems under preindustrial conditions based on a compilation of nitrogen fixation fluxes from various ecosystems. Using nitrogen‐isotope abundances, Vitousek et al. () obtained a considerably lower preindustrial estimate of 58 Tg N y−1 for terrestrial nitrogen fixation, with a possible range from 40 to 100 Tg N y−1. This lower estimate is adopted by Fowler et al. (). Cleveland et al. () estimate natural nitrogen fixation in terrestrial ecosystems at 127 Tg N y−1, based on biogeochemical modeling of the contributions of different terrestrial ecosystems.
Estimates of nitrogen fixation in marine ecosystems range from 121 to 177 Tg N y−1 (Groβkopf et al., ; Jickells et al., ). Lightning strikes are estimated to account for an additional 5.4 Tg N y−1 on a global basis (Galloway et al., ). Combining these estimates with the range of values for terrestrial ecosystems, we estimate the global rate for production of reactive nitrogen at 166–302 Tg N y−1 under preindustrial conditions. Using central estimates for terrestrial nitrogen fixation and marine ecosystems, the overall global preindustrial nitrogen fixation rate would be about 240 Tg N y−1. Table summarizes estimates of anthropogenic and natural fluxes of reactive nitrogen.
Fluxes of Reactive NitrogenSources (Tg N y−1) | Range (for Present Values) | References | |||
1910 | 1960 | Present | |||
Sources | |||||
Anthropogenic | |||||
Synthetic or mined fertilizers | 1 | 12 | 110 | Cushman (), IFA (), Smil () | |
N‐fixing crops | |||||
Soybeans | 0.4 | 1.5 | 25 | Shurtleff and Aoyagi (), FAO (), Herridge et al. () | |
Other beans and legumes | 2.9 | 7.5 | FAO (), Herridge et al. () | ||
Other | 7.4 | 10 | FAO (), Herridge et al. () | ||
Total N‐fixation | 12 | 43 | 30–51 | ||
NOX from combustion | 2.5 | 15 | 38 | UN () | |
Total anthropogenic | ∼4 | 39 | 190 | 160–210 | |
Natural | |||||
Terrestrial N‐fixation | 84 | 40–127 | Galloway et al. (), Vitousek et al. (), Cleveland et al. () | ||
Lightning | 5.4 | Galloway et al. () | |||
Marine N‐fixation | 150 | 121–177 | Galloway et al. (), Groβkopf et al. () | ||
Total natural sources | 240 | 166–310 | |||
Total sources | 430 | ||||
Sinks | |||||
Denitrification | |||||
Agricultural soils | 22–87 | Hofstra and Bouwman () | |||
Total terrestrial | 130 | 58–175 | Seitzinger et al. (), Eugster and Gruber () | ||
Freshwater and groundwater | 120 | 39–216 | Eugster and Gruber () | ||
Marine | 150 | 107–331 | Seitzinger et al. (), Eugster and Gruber (), DeVries et al. () | ||
Total denitrification | 400 | 210–720 | |||
Other sinks | |||||
Terrestrial biomass change | 9 | 9 | Schlesinger () | ||
Marine sediments | 13 | 10–16 | Galloway et al. (), Schlesinger and Bernhardt () | ||
Total sinks | 420 | 220–745 |
Even considering the uncertainty in the magnitude of fluxes in the nitrogen cycle, anthropogenic sources of reactive nitrogen represent a significant perturbation to the terrestrial nitrogen cycle and to the global nitrogen cycle as a whole. The trend for the anthropogenic nitrogen contribution includes synthetic fertilizer, cultivation of nitrogen‐fixing crops, and NOX emissions from fossil fuel combustion (Figure ). Reactive nitrogen in pasturelands may also be increased by anthropogenic activities such as the increased pasturing of food animals.
Fig. 4. Trend in the anthropogenic contribution of new fixed nitrogen relative to the estimated ranges of global nitrogen fixation under preindustrial conditions.
Figure shows that the anthropogenic nitrogen contribution would have surpassed the lower end estimate of preindustrial nitrogen fixation in terrestrial ecosystems by 1960; and would have surpassed the upper end estimate for terrestrial ecosystems by about 1980. Based on the central estimates given in Table , the current anthropogenic contribution accounts for about 70% of total production of reactive nitrogen in terrestrial ecosystems. The anthropogenic contribution is close to our best estimate of overall global nitrogen fixation under preindustrial conditions (240 Tg N y−1), and is increasing across the uncertainty bar for the global estimate. This suggests that the anthropogenic inputs may now account for about half of the total reactive nitrogen flux on Earth, both terrestrial and marine.
Anthropogenic reactive nitrogen produces multiple impacts at local, regional, and global scales. Emissions of NH3 and NOX contribute to the formation of fine particulate matter (PM2.5), which is associated with various adverse human health impacts, including premature death (
Reactive nitrogen fertilizes terrestrial and aquatic ecosystems, which can affect species diversity and can lead to eutrophication of aquatic ecosystems (
Reactive nitrogen compounds can leach into groundwater, contaminating drinking water supplies (
The impacts of reactive nitrogen on the environment have different time scales. Elevated levels of PM2.5 may persist for only a few days, but may also be chronic. Impacts on groundwater, the oceans, and climate have much longer timescales. The lifetime of N2O in the atmosphere is estimated at 121 years (
Fate of Anthropogenic Nitrogen
Most of anthropogenic augmentation of reactive nitrogen is believed to be removed ultimately by denitrification, producing nitrogen gas (N2) and N2O. Some of this denitrification occurs in agricultural soils, with estimates of global flux ranging from 22 to 87 Tg N y−1 (
Estimates of the sinks for reactive nitrogen are also subject to considerable uncertainty. Recent estimates of global denitrification flux are from 58 to 175 Tg N y−1 for terrestrial ecosystems (
Many terrestrial, freshwater, and marine ecosystems are historically nitrogen limited (Elser et al., ). Thus, export of fixed nitrogen from farmlands to these ecosystems can adversely affect biodiversity (Zaehle et al., ; DeVries et al., ). It is possible that the deposition of reactive nitrogen to forested ecosystems may increase the sequestration of carbon in soils and biomass (Pinder et al., ). However, the deposition of reactive nitrogen in natural ecosystems also enhances production of N2O as a byproduct of microbial denitrification reactions. The climate change impacts of the increased N2O may offset the impacts of carbon sequestration (Butterbach‐Bahl et al., ; Zaehle et al., ).
Ice‐core analyses provide a record of the increase in the atmospheric concentration of N2O (Bullister, ), and in the long‐range transport of reactive nitrogen in the form of NO3−(Hastings et al., ). Figure compares trends for N2O and NO3− with trends measured for CO2 (Etheridge et al., ; WMO, ). The increases in ice‐core concentrations of CO2 and N2O are believed to reflect trends in the global concentrations of these gases, since they are well‐mixed in the atmosphere. The ice‐core measurements of NO3− (in Greenland) are believed to be related to an increase in regional transport of NO3−. This increase corresponds with a change in the isotopic composition of nitrate‐N in the ice cores, reflected by a reduction in the abundance of nitrogen‐15 relative to nitrogen‐14 [δ15N].The change in isotopic composition may be indicative of an increased contribution of agricultural sources (Felix and Eliott, ). However, the fractionation of nitrogen isotopes is complex, and the samples from Greenland ice cores may be open to different interpretations (Hastings et al., ).
Fig. 5. Trends in CO2, N2O, NO3−, and isotopic composition of NO3− as reflected in ice cores. Trend lines are smoothed by averaging over 5 years. Trend in CO2 is adapted from Etheridge et al. () and WMO (). N2O is adapted from Bullister (). NO3− and δ15N are adapted from Hastings et al. () and Felix and Eliott ().
There is growing recognition of the impact of anthropogenically produced nitrogen compounds on the nitrogen cycle (Gruber and Galloway, ; Sutton et al., ; U.S. Environmental Protection Agency Science Advisory Board (USEPA), ; Fowler et al., ). Researchers have used the concept of planetary boundaries to evaluate the magnitudes of human impacts consistent with the sustainable maintenance and development of human society (Rockstrom et al., ; DeVries et al., ). Within this construct, the planetary boundary for the anthropogenic contribution to total reactive nitrogen is estimated at 62–82 Tg N y−1, based on the risk of eutrophication of terrestrial and aquatic ecosystems (Steffen et al., ). This threshold was exceeded in the 1970s, and the current anthropogenic contribution is approximately 153 Tg N y−1.
Discussion
Regulatory and voluntary measures have been adopted to address some components of the nitrogen stream. In the U.S., livestock production facilities are required to obtain permits under the Clean Water Act in order to ensure that the surface waters surrounding the operations are not negatively impacted by animal waste. In implementing this permit system, states have identified Best Management Practices (BMPs) for the management of animal waste and for the use of synthetic nitrogen fertilizers.
Nitrogen‐use efficiency has increased in the U.S. and Europe in recent years. Although the trend in nitrogen fertilizer usage in developed nations has been flat since the 1990s, agricultural production continues to increase. This suggests that increases in nitrogen‐use efficiency can abate or perhaps reverse the worldwide increase in nitrogen fertilizer use (Zhang et al., ). Changes in dietary habits and food wastage can also help reduce the need for synthetic nitrogen fertilizers.
NOX emissions from combustion sources have been the target of pollution controls in order to mitigate acid rain and smog in the U.S. These controls have produced a 50% reduction in nationwide NOX emissions since 1996 (U.S. Environmental Protection Agency (USEPA), ). Total deposition of inorganic nitrogen has declined by 19 to 32% since the 1980s (Burns et al., ). This demonstrates that NOX emissions can be reduced without preventing economic growth.
Unfortunately, there is no integrated regulatory approach for control of reactive nitrogen compounds (Aneja et al., , ). In particular, agricultural emissions of reactive nitrogen compounds to the atmosphere are not regulated in the U.S. Because of its contribution to the formation of PM2.5, individual states can regulate NH3 as part of their approaches to meeting PM2.5 standards. However, the U.S. has not adopted a national program for NH3 emissions. NH3 is the largest volume air pollutant for which no national or regional control program has been developed. Thus, wet deposition of NH4+ has increased in much of the U.S. by ∼22% in the last 20 years (Li et al., ). Some European nations have adopted control measures for NH3, which have produced significant reductions in emissions from agriculture (Erisman et al., ). A combination of BMPs and engineered solutions for the management of animal waste and for the use of synthetic nitrogen fertilizers can reduce releases of NH3 and other reactive nitrogen compounds to the natural environment (Galloway et al., ; Erisman et al., ).
Human‐induced changes to the global nitrogen cycle bear a number of similarities to our changes in the global carbon cycle. Fixed N and fossil C have provided great benefits to the human standard of living. The increased use of nitrogen has been critical for increased crop yields and protein production to keep pace with the growing world population. Like the burning of fossil carbon, increased fixation of nitrogen can have adverse environmental consequences at local, regional, and global scales. In addition, our use of both fossil carbon and synthetically fixed nitrogen has grown exponentially in the past 150 years. Anthropogenic production of fixed nitrogen has grown in relation to natural sources, so that the anthropogenic increment is nearly as large as the best estimate of the total natural nitrogen fixation in terrestrial and marine environments.
Some measures for reducing CO2 emissions will reduce releases of fixed nitrogen, and vice versa. For instance, renewable energy sources generally will reduce fuel consumption, thereby reducing NOX emissions. Switching from coal to natural gas also reduces both CO2 and NOX emissions. In addition, the production of synthetic nitrogen fertilizer requires significant fuel consumption, with associated CO2 emissions. Thus, improvements in nitrogen‐use efficiency would reduce CO2 emissions as well as releases of fixed nitrogen.
There are also important differences pertaining to human impacts on the carbon and nitrogen cycles. A significant fraction of anthropogenic CO2 emissions has been taken up by the oceans. In contrast, most anthropogenic fixed nitrogen is believed to be converted to N2. Although some fixed nitrogen may be accumulating in soils and in ocean sediments, the fraction is much smaller than the fraction of carbon taken up by the oceans. This could mean that the nitrogen cycle could possibly recover more quickly from anthropogenic perturbation than the carbon cycle if releases of fixed nitrogen are mitigated. However, demands for increased nitrogen usage will continue as world population and agricultural production continue to rise. Thus, reducing the demands for fixed nitrogen may prove to be more difficult than reducing emissions of CO2.
The impacts of anthropogenic perturbations on the reactive nitrogen cycle are local, regional and global scale. Current estimates indicate that global nitrification fixation is balanced by denitrification, although these budget calculations are subject to large uncertainties. Thus, the large increase in anthropogenic production may be balanced by increased denitrification, and there is no apparent “missing sink” for reactive nitrogen. Nevertheless, if anthropogenic production continues to increase, denitrification processes may not be able to offset the increased production. Continued increases in reactive nitrogen production may accelerate species diversity impacts and N2O production.
Public awareness of the impact of reactive nitrogen is also increasing. We anticipate that reactive nitrogen may be similar to the situation with carbon in another respect. The environmental impacts of anthropogenic reactive nitrogen may become more difficult to rectify as time passes. In the case of carbon, we are accumulating a burden of CO2 that will impact the atmosphere far into the future (IPCC, ). In the case of reactive nitrogen, anthropogenic contributions continue to grow in relation to the natural budget, with uncertain consequences. “Is nitrogen the next carbon?” is a thought provoking question. Mitigating both carbon and nitrogen is a grand challenge.
Acknowledgments
In keeping with AGU policy, the data reported in this paper are all freely available in the published literature. Support for this work was provided by the NASA Earth and Space Science Fellowship (NESSF) program, grant No. NNX15AN15H; and U.S. GFDL‐NOAA project NOAA CPO AC4. We acknowledge North Carolina State University Air Quality Research Group. We thank The Kenan Fund and The Cary Institute of Ecosystem Studies for financial support. We also thank James N. Galloway for all his research efforts to help us understand the global nitrogen cycle.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2017. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
Key Points
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer