Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

50 year trends in nitrogen use efficiency of world cropping systems: the relationship between yield and nitrogen input to cropland

50 year trends in nitrogen use efficiency of world cropping systems: the relationship between... Nitrogen (N) is crucial for crop productivity. However, nowadays more than half of the N added to cropland is lost to the environment, wasting the resource, producing threats to air, water, soil and biodiversity, and generating greenhouse gas emissions. Based on FAO data, we have reconstructed the trajectory followed, in the past 50 years, by 124 countries in terms of crop yield and total nitrogen inputs to cropland (manure, synthetic fertilizer, symbiotic fixation and atmospheric deposition). During the last five decades, the response of agricultural systems to increased nitrogen fertilization has evolved differently in the different world countries. While some countries have improved their agro-environmental performances, in others the increased fertilization has produced low agronomical benefits and higher environmental losses. Our data also suggest that, in general, those countries using a higher proportion of N inputs from symbiotic N fixation rather than from synthetic fertilizer have a better N use efficiency. S Online supplementary data available from stacks.iop.org/ERL/9/105011/mmedia Keywords: nitrogen use efficiency, country and global scales, cropping systems, crop yields, nitrogen pollution 1. Introduction most crucial (Tilman et al 2002, Mueller et al 2012, Sinclair and Rufty 2012). The flipside of the coin, however, is an Although malnutrition has not receded in absolute terms, increased alteration of surface and groundwater resources, world agriculture, in the past half century, has succeeded in coastal eutrophication, air pollution and increased greenhouse increasing its production of vegetal proteins by a factor of 3 gas emission (Billen et al 2013, Sutton et al 2013). From this (Lassaletta et al 2014a). This has been made possible by perspective, very different situations exist, linked to the dis- parity of cropping system development in the countries and changes in cropping systems generally referred to as the regions of the world (Billen et al 2014). Green Revolution, based on the adoption of improved crop It is the purpose of this paper to describe these issues, varieties, use of pesticides, and increased application of based on an original analysis of the data available in the FAO synthetic fertilizers, among which nitrogen was by far the data base since 1961 (www.faostat.fao.org). Our approach is based on the calculation of the various components of the Content from this work may be used under the terms of the arable soil budget of 124 countries and, most importantly, on Creative Commons Attribution 3.0 licence. Any further the description of the trajectory drawn from 1961 to 2009 by distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. these countries in terms of their total crop production (Y, 1748-9326/14/105011+09$33.00 1 © 2014 IOP Publishing Ltd Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al –1 –1 expressed in harvested vegetal protein, kgN ha yr ) and the retained. This procedure allowed avoiding discrepancies in total N inputs onto cropland (F), excluding permanent the FAO data base. (See S1 for details). grassland, in the form of synthetic fertilizers, manure, sym- Total fertilization of cropland was defined as the total N biotic fixation and atmospheric deposition input in the form of synthetic fertilizers, symbiotic N fixation, –1 –1 manure application and atmospheric deposition onto crop- (also in kgN ha yr ). This approach differs from, and is land, excluding permanent grassland. The reason for focusing complementary to the Net Anthropogenic Nitrogen Input or our analysis on cropland is that the fate of the agricultural NANI approach (Howarth et al 2012). Our aim is to analyze surplus (excess N input over N export by plant harvest) cropping systems and to evaluate the excess N application on strongly differs between cropland and permanent grassland, arable land, the most sensitive components of agricultural particularly with respect to the relative proportions of NH systems, while the NANI approach deals holistically with the volatilization, denitrification, leaching and storage in the soil complete N cycle at the country scale, including the livestock organic pool (Velthof et al 2009, Billen et al 2013). Note that compartment and the effect of agricultural commodity trade temporary grassland (e.g. the FAOstat crop category ‘Grasses (Swaney et al 2012, Lassaletta et al 2014b). Nes for forage;Sil’), included within crop rotations, are con- Using a different approach, Conant et al (2013) have sidered as cropland. Yearly data on synthetic N fertilizer recently created a global soil N input database that enables application, under different N forms, for the entire period evaluation of trends in nitrogen use and recovery by country were obtained from the Resources module of the FAOstat for a number of important crops over the last 40 years. Their database. Countries with more than 15 missing annual data data show that differences in efficiency of N use between were removed from the analysis. Occasional gaps were filled OECD and other countries have persisted and exhibit no sign with data from the International Fertilizer Industry Associa- of convergence. In this paper we use the concept of the yield- tion (www.fertilizer.org/) if available and if not, by using fertilization relationship in an original way compared to the figures of the closest years. FAO data on annual per country concept commonly used, relating here the mean yield inte- synthetic fertilizer use refer to total use in agriculture and is grated over the entire crop rotation to the total fertilization of provided without distinction between arable and grassland. the cropland soils of a given territory. While the yield-ferti- We therefore had to subtract from these figures the proportion lization relationship is normally used in conventional agr- used for grassland fertilization, which in some European onomy as a tool to predict the yield increase of a given crop countries such as Ireland and the Netherlands accounts for a that could be expected from increasing fertilization in a given significant proportion. We have estimated the proportion of pedo-climatic context, we consider the integrative values of Y synthetic fertilizers to grasslands at the country scales pro- and F as overall indicators of the agronomical and environ- cessing the information compiled from different sources mental performances of a cropping system: the Y/F ratio is a (Richard 1951, Power and Alessi 1971, Anonymous 1992, measure of its nitrogen use efficiency (NUE), while the F-Y FAO 2006, Heffer 2013) (see S1 for details). difference is the regional N surplus (or N balance) repre- To estimate the crop biological nitrogen fixation by fix- senting the potential for hydrological or gaseous losses of ing crops included in the FAOstat database we used a yield- nitrogen to the environment. based approach, assuming that crop yield is the factor that best aggregates variables associated with crop, soil and cli- matic conditions including available N, soil moisture, vigor of stand, and other management factors influencing N fixation: 2. Methods Nfixed=⁎ %Ndfa ⁎ BGN , NHI Based on FAO data, we have reconstructed the trajectory followed by 124 countries in the past 50 years, in terms of where %Ndfa is the percentage of N uptake derived from N –1 –1 crop yield (Y refers to harvested crop part and is expressed in fixation, Y is the yield (expressed in kgN ha yr ), NHI is the –1 –1 kgN ha yr ) and total nitrogen inputs to cropland (F, sum N harvest index, defined as the ratio of the harvested material of nitrogen in manure, synthetic fertilizer, symbiotic fixation to the total above-ground N production, and BGN is a mul- –1 –1 and atmospheric deposition, in kgN ha yr ) for the tiplicative factor expressing the total N fixation including 1961–2009 period. Together these countries represent 99.2% below-ground contributions associated with roots, nodules of the world population and 99.6% of the cropland surface in and rhizo-deposition via exudates and decaying root cells and 2009 (see supplementary material S1, available at stacks.iop. hyphae. These parameters have been obtained from different org/ERL/9/105011/mmedia, for detailed methodology). sources (Herridge et al 2008, Salvagiotti et al 2008, Laberge Total annual crop production by each country was cal- et al 2009, Kombiok and Buah 2013, Álvarez et al 2014, culated taking into account the yearly harvested yield of 178 Anglade et al under review). We applied a regional %Ndfa primary crops and their N content, as reported in Lassaletta for soybean N fixation. For sugar cane, rice, paddy and forage et al (2014a). The cropland surface was estimated by sum- products, we applied a constant rate of biological fixation per ming up the surfaces of all individual crops. Only in the cases hectare, as suggested by Herridge et al (2008) (see S1 for where this sum was higher than the stated value of the ‘arable details). land and permanent crops’ surface area provided by the To estimate the animal excretion factors, we have fol- FAOstat resources module, the latest surface area was lowed the methodology of Sheldrick et al (2003) that assumes 2 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al that excretion rates, within a given livestock category, are the fact that, in the long run, harvest cannot exceed N resti- proportional to the slaughtered animal weights. We have tutions to the soil, and that the effect of low fertilization calculated different ratios for dairy and for other cattle stocks in strongly N-limited systems is characterized by a NUE close using the dairy stocks provided in the FAOstat ‘livestock to 1. The third property expresses the classical law of primary’ module. These stocks have been subtracted from the diminishing return and the fact that, in constant technical- total cattle stock to estimate non-dairy cattle. As a result, a agronomical context, some other limiting factor will always particular excretion factor has been applied to each type of impose a ceiling to production at saturating N availability. animal, country and year. The proportion of N excreted that is Two mathematical functions with only one parameter finally used as manure applied onto cropland was taken from obey both conditions: a hyperbolic function of the form the estimates of Sheldrick et al (2003) at the regional level for Y = Ymax*F/(F + Ymax) [1] and a negative exponential func- each type of animal. It was considered that 30% of the tion such as Y = Ymax [1 − exp(−F/Ymax)] [2]. We observed available manure is lost during management and storage that the former generally provides the best fit to the data. In before reaching the crop, as proposed by Oenema et al (2007) both cases the parameter Ymax represents the yield value for Europe and close to the value estimated by Liu et al reached at saturating N fertilization, as well as the value of (2010) at the world scale. We finally discount the amount of fertilization at which a definite fraction of this maximum yield N that is applied to permanent grasslands by applying the is reached (this fraction being 0.5 in the case of relation [1] or proportions provided by regions, and in some cases at the 0.63 for relation [2]). Over the 1961–2009 period, certain country scale, by Liu et al (2010) (see S1 for details). countries that we will call ‘type I’, such as China, Egypt and Deposition of oxidized and reduced nitrogen compounds India, present a simple trajectory with regularly increasing onto croplands was calculated from the GlobalNEWS data- fertilization and gradual reduction in the crop yield response, base (Seitzinger et al 2010) by extrapolating linearly between following a consistent and unique Y versus F relationship available years. The atmospheric deposition data used in (figure 1(a)). GlobalNEWS are derived for the year 2000 from Dentener Other countries (called ‘type II’), such as the USA, Brazil et al (2006) and previous figures were obtained by scaling and Bangladesh, display a historical trajectory with first a deposition fields for this year following Bouwman et al regularly increasing fertilization and yield, fitting the Y versus (2009). We calculated the input of N per ha (yearly national F relationship with a definite Ymax, then a turning point with average) and we applied this input per ha into the surface of a shift of the trajectory to another relationship with a sig- cropland considered in this study (see S1 for details). nificantly higher Ymax. This likely reflects improved agro- nomical practices in terms of production factors other than nitrogen, together with the pursuit of increasing fertilization. The turning point seems to have occurred in the 1980s or later 3. Results and discussion depending on the country (figure 1(b)). The case of the USA, for example, is consistent with a slowdown in the increase of 3.1. Y versus F trajectories of world agricultures synthetic fertilizers inputs from the 1980s parallel to a mod- The trajectory followed from 1961 to 2009 by a number of erate increase in the yields of the most important crops countries in terms of crop yield and total N inputs into (Howarth et al 2002, Alston et al 2010). cropland is shown in figure 1. The results for all countries of In most European countries (see the example of France, the FAO data base are provided in supplementary material the Netherlands and Greece in figure 1(c)), the trajectory also (S2). The Y versus F trajectory drawn by most countries shows a bi-phasic pattern, describing a regular increase in shows, at least for periods of several decades, a distinct cur- both fertilization and yield during the 1960–1975 period, vilinear relationship. Linear trajectories, like those described followed by a shift towards improved yields without further by de Wit (1992) for individual crops were seldom observed. increasing fertilization and even decreasing fertilization from Several mathematical formulations of the yield-fertilization the 1980s on (‘type III’). The case of the Netherlands is the relationship in a given pedo-climatic and technical-agrono- most spectacular, as in this country, which has always used mical context have been proposed in the agronomical litera- very high rates of fertilization, the level applied in recent ture, most of them involving negative exponential functions years has been reduced to the same as in the 1960s with, (Llewely and Featherstone 1997, Harmsen 2000). Nijland however, doubled yields. This trend is related to the reduction et al (2008) proposed to integrate the production functions of of N inputs prescribed by European environmental policies Liebig, Mitscherlich and Liebscher (de Wit 1992) into one and regulations (van Grinsven et al 2012), which interestingly system model based on Michaelis−Menten hyperbolic rela- did not prevent significant yield increases thanks to a better N tionships. Because we are expressing both output and input in management. Note, however, that despite the increase of –1 –1 exactly the same unit (kgN ha yr ) and because we are NUE and decrease in N surpluses, the nitrogen surplus looking for a simple long-term integrative theoretical rela- emitted to the environment in many cases remains much tionship, we decided to make use of the simplest possible higher than that of other countries belonging to types I and II. function obeying the three following properties: (i) the Finally, there is a small group of countries, such as function intercept should be zero; (ii) the slope of the function Morocco, Benin and Nigeria, whose trajectory does not dis- should be 1 at low fertilization; (iii) the function should reach play any consistent Y versus F relationship (type IV). These a plateau at high fertilization. The first two properties reflect countries have always very low inputs and yields. Very often, 3 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Figure 1. Examples of trajectories followed by countries in the Y versus F diagram. (a) Examples of type I trajectories. (b) Examples of type II trajectories. (c) Examples of type III trajectories. (d) Examples of type IV trajectories. R is the coefficient of determination, defined as: 2 ⎡ 2 2⎤ R =− 1 ∑(obs − calc ) /∑(obs —meanobs) where obs are the observed values of yield, calc the yield value calculated with the ii i i i ⎣ ⎦ relationship and meanobs is the average value of the observed yields over the period considered. Negative values of R indicate poor fit of the relationship on the observed values. This is often the case for the most recent period of type III trajectories because of still evolving agronomical conditions. 4 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Figure 2. Past and current agricultural potential of world countries, in terms of maximum protein yield of cropping systems (Ymax). their trajectory in the Y versus F diagram crosses the 1:1 line, indicating higher yield than fertilization. High inter-annual variation in the agricultural performance observed in some of these countries could be explained by weather phenomena, such as persistent water droughts, socio-political questions, or sometimes could be even an artefact due to the poor quality of our estimates of total nitrogen inputs to agricultural soil: in particular, in those countries where shifting agriculture is practiced, the fertilization of agricultural soil by the nitrogen stock accumulated in forested soil during the fallow period is not taken into account in our input estimations. However the ‘negative’ nitrogen balance displayed in the Y versus F dia- gram can also represent the signature of an unsustainable nitrogen mining of agricultural soils. For type I to III countries, we were able to define the Ymax values providing the best fit of the hyperbolic rela- tionship [1] to the points corresponding to the 1961–1980 period or later, and another Ymax for the most recent 10–15 years. The two Ymax values obtained characterize the past and current agricultural potential respectively, defined as the protein yield that could be obtained from cropland at a maximum N fertilization rate, with the corresponding crop- ping practices (figure 2). Comparison of the two periods shows a significant increase of Ymax in 45 countries (type II and III trajectories). For a large number (55) of countries, however, nearly the same parameter value or Ymax holds over the 50-year period (type I trajectory), as is the case for China, Egypt, Turkey, Chile, India and a many others. Possible N mining is indicated by higher crop yield than fertilization for 18 countries such as Canada, Morocco, Algeria, Iraq and Mozambique in the 1960–1980 period (see S2 for the com- plete list). In recent years, N mining continues in 10 African Figure 3. 50 years trends in nitrogen use efficiency of the cropping countries, as well as in Former Soviet Union countries, system of a number of countries. Afghanistan and Paraguay. N mining has been observed in Argentina for the entire studied period. This result is coherent with that recently reported by Álvarez et al (2014) that apparently wide margins for improvement through better indicate a budget of the copping system in the pampean fertilization practices, including an increasing use of legumes agroecosystems, which only becomes positive when includ- in crop rotations (Vanlauwe et al 2014). However, imbalances ing pasture lands. The severe problem of nutrient mining and with other nutrients such as P could limit yield responses to N loss of soil fertility in African countries has been frequently addition (van der Velde et al 2014). In the Former Soviet highlighted (Vitousek et al 2009, Liu et al 2010). In these Union, after the abrupt changes which occurred from 1989, countries yields are among the lowest in the world but have crops may have benefitted from nutrient legacies. The results 5 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Figure 4. (a) Distribution of the share of symbiotic fixation and synthetic fertilizers in total N inputs to cropland by countries in 2000–2009. (b) Observed relationship between NUE and the proportion of symbiotic fixation, or of synthetic fertilizers in total N inputs to cropland in the period 2000–2009. of our calculations for this country, however, might also be S2-1). This indicator is high (>0.75) in North America, affected by recent and poorly documented changes like Australia, most European and many sub-Saharan countries, massive land abandonment not fully documented by the FAO indicating margins for increasing yields by increasing N fer- (Schierhorn et al 2013). tilization. It is low (<0.3), on the other hand, in countries like China, India and Pakistan, as well as in a number of Central American and North African countries, indicating no benefit 3.2. Agronomical performances: trends in N use efficiency and in terms of yield to be expected from simple increase of N N-based yield gap fertilization in these regions in the absence of radical agro- The above-described trajectories can be translated in terms of nomical improvement of the cropping system. George (2014) changes in the NUE of the cropping system in the different has analyzed why crop yields in many developing countries countries (figure 3). Type I countries display a regularly cannot easily respond to increased inputs due to poor agro- decreasing trend of NUE. During the same period a similar nomical practices. drop for the phosphorus use efficiency (PUE) has been reported for China (Sattari et al 2014). In type II countries, the 3.3. Environmental performances: NUE and N losses shift in the trajectory toward an improved Ymax results in the stabilization or in the increase of NUE. In type III countries, The data we have assembled can also be used to estimate the the reduction of N inputs in recent years with no drop in yield contribution of agriculture in the different world countries to corresponds to increasing NUE. environmental nitrogen contamination, using the nitrogen To characterize the performance of agriculture of a given surplus (F-Y) as an indicator of potential losses. While in territory, van Ittersum et al (1997, 2013) introduced the grassland this surplus is generally stored in the soil organic concept of yield gap, defined as the difference between the matter pool, in the case of cropland, most of it is leached actual farmers’ yield of a particular cultivar and the potential quickly as nitrate, emitted as NH or denitrified as di-nitrogen, yield which could be obtained in the same place in the and nitrous oxide as a by-product, thus contributing to the absence of limitation by nutrient and water and with efficient nitrogen cascade of environmental contamination (Galloway control of pests and diseases. Following the same line of et al 2003). The data thus show the global distribution of reasoning, but applied to the overall production of the crop- environmental N losses from agricultural soils (figure S2-2). –1 –1 ping systems of each country in terms of proteins, we cal- Losses are over 50 kgN ha yr in most of Europe, the Middle culated (Ymax-Y)/Ymax as a dimensionless indicator of the East, the USA and Central America, India and China. –1 –1 degree of N limitation of current agricultural yields (figure They remain on average below 25 kgN ha yr in most 6 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al fertilizers in total fertilization. This higher efficiency of cropping systems relying largely on biological N fixation is observed for the largest soybean producers of South America as noted by Liu et al (2010) as well as for less productive countries in Africa and Asia with significant production of rice, groundnuts and beans. The higher NUE associated to nitrogen fixation is likely explained by a higher efficiency in the incorporation by legumes of their self-supplied nitrogen (Herridge and Peoples 1990). Also, the increase of the price of synthetic fertilizers might have encouraged the use of other sources of N in a most efficient way. 3.4. Global trends At the global scale, lumping together all cropping systems of the world, a type II Y/F trajectory is observed, with a shift during the 1980s from one Y/F relationship characterized by a –1 –1 Ymax of ∼70 kgN ha yr to an improved one with Ymax of –1 –1 110 kgN ha yr (figure 5(a)). The overall observed global trend is a distinct decrease of NUE in the 1961–1980 period (from 68% to 45%), followed by a stabilization during the last 30 years around 47% (figure 5(b)). The share of the different sources of N in the total inputs to cropland, depicted in figure 5(c), change considerably during the last 50 years, with synthetic fertilizers now being the largest source. Despite that the total rate of N excreted by livestock is equivalent to synthetic fertilizer application, the manure, rarely used effi- ciently, finally reaches the crops at a much lower rate which nowadays is slightly lower than crop biological N- fixation (figure 5(c)). 4. Conclusions Currently, only 47% of the reactive nitrogen added globally onto cropland is converted into harvested products, compared to 68% in the early 1960s, while synthetic N fertilizer input Figure 5. The global cropping system. (a) Trajectory followed by increased by a factor of 9 over the same period. This means global world cropping systems in the Y versus F diagram (Y: crop –1 –1 that more than half the nitrogen used for crop fertilization is yield in protein harvested, kgN ha yr ; F: total N inputs to the –1 –1 currently lost into the environment. Even though a significant cropland soil, kgN ha yr ). (b) Trends in nitrogen use efficiency of the global cropping system. (c) Evolution of the components of the improvement in NUE occurred in many countries after the global cropping system budget. 1980s, the present results suggest that a further increase of nitrogen fertilization would result in a disproportionately low increase of crop production with further environmental sub-Saharan Africa, the Former Soviet Union countries, and alterations, unless cropping systems improve their efficiency Australia. High surplus values are associated with low NUE substantially. In that respect, improvement of agronomical (figure S2-2). practices and development and proper application of envir- Total fertilization, as discussed above, is mainly the sum onmental policies have been demonstrated to be efficient of synthetic fertilizers, manure application and symbiotic strategies. A better integration of crop and livestock systems nitrogen fixation. Although not negligible, atmospheric N can also contribute to increasing NUE at the local and global deposition generally contributes a much smaller share. The scale (Herrero et al 2010, Lassaletta et al 2014a, Bonaudo proportion of the three former N inputs to overall fertilization et al 2014, Soussana and Lemaire, 2014). Moreover, our data varies a great deal among the different world cropping sys- suggest that an increase in the contribution of symbiotic N tems, as shown in figure 4(a). Our data show that NUE is fixation would result in increasing NUE. Peoples et al (2009) generally higher (and the N surplus relatively lower) for have stressed that the potential of symbiotic nitrogen fixation agricultural systems with higher proportion of N inputs is currently largely underexploited, given that very few derived from symbiotic N fixation (figure 4(b)). Conversely, countries have a fraction of arable land devoted to legume NUE is generally lower for a higher proportion of synthetic crops greater than a few percent. Increased areas of legumes 7 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al might be achieved by including more leguminous crops in Bouwman A F, Beusen A H W and Billen G 2009 Human alteration of the global nitrogen and phosphorus soil balances for the rotations, or by the introduction of short-duration legume period 1970–2050 Glob. Biogeochem. Cycles 23 GB0A04 green manures or ‘catch crops’ (Blesh and Drinkwater 2013). Conant R T, Berdanier A B and Grace P R 2013 Patterns and trends By evidencing the long-term response of N inputs to the in nitrogen use and nitrogen recovery efficiency in world soil in terms of production and potential losses to the envir- agriculture Global Biogeochem. Cycles 27 558–66 onment, this paper provides a summarized and comprehen- Dentener F et al 2006 Nitrogen and sulfur deposition on regional and global scales: a multimodel evaluation Global Biogeochem. sive diagnosis of the effective changes in agronomical and Cycles 20 B4003 environmental performances of the cropping systems of 124 de Wit C T 1992 Resource use efficiency in agriculture Agric. Syst. countries of the world. 40 125–51 FAO 2006 Fertilizer and Plant nutrition Bulletin n°17 (downloadable from the FAO Fertstat website)(Rome: FAO) Galloway J N et al 2003 The nitrogen cascade BioScience 53 341–56 George T 2014 Why crop yields in developing countries have not Acknowledgments kept pace with advances in agronomy Glob. Food Sec. 3 49–58 Harmsen K 2000 A modified Mitscerlich equation for rainfed crop This study was partly supported by the SeasERA EMoSEM production in semi-arid areas: 1. Theory Netherlands J. Agric. project (ANR-12-SEAS-0005-01). We wish to thank the Sci. 48 237–50 Heffer P 2013 Assessment of Fertilizer Use by Crop at the Global FIRE (Fédération Ile de France de Recherche en Environne- Level. 2010–2011/10. International Fertilizer Industry ment, CNRS and UPMC). We thank Javier Castrillo who Association (Paris: IFA) developed several computer routines for the data management Herrero M et al 2010 Smart investments in sustainable food and Angel Udias for his help in preparing the figures. We are production: revisiting mixed crop-livestock systems Science sincerely grateful to two anonymous reviewers for their 327 822–5 Herridge D F and Peoples M B 1990 Ureide assay for measuring detailed and constructive revision. We thank Augustin del nitrogen-fixation by nodulated soybean calibrated by n-15 Prado, Eduardo Aguilera, Guillermo Pardo, Fernando methods Plant Physiol. 93 495–503 Estellés, Alberto Sanz-Cobena and Dennis Swaney for help- Herridge D F, Peoples M B and Boddey R M 2008 Global inputs of ful suggestions and comments. We also thank to the Scientific biological nitrogen fixation in agricultural systems Plant Soil Network on GHG mitigation from the agroforestry sector in 311 1–18 Howarth R W, Boyer E W, Pabich W J and Galloway J N 2002 Spain (REMEDIA) for providing an space for discussion and Nitrogen use in the United States from 1961–2000 and science developement. The views expressed in this document potential future trends Ambio 31 88–96 are purely those of the authors and may not in any circum- Howarth R, Swaney D, Billen G, Garnier J, Hong B, Humborg C, stances be regarded as stating an official position of the Johnes P, Mörth C-M and Marino R 2012 Nitrogen fluxes from European Commission. the landscape are controlled by net anthropogenic nitrogen inputs and by climate Front Ecol. Environ. 10 37–43 Kombiok J and Buah S 2013 Tillage depth effects on nodulation, nitrogen fixation and yield of three soybean varieties in the Northern Savanna zone of Ghana Afr. J. Agric. Res. 8 2340–5 References Laberge G, Franke A C, Ambus P and Hogh-Jensen H 2009 Nitrogen rhizodeposition from soybean (Glycine max) and its impact on nutrient budgets in two contrasting environments of Alston J, Babcock B and Pardey P 2010 The Shifting Patterns of the Guinean savannah zone of Nigeria Nutr. Cycl. Agroecosys Agricultural Production and Productivity Worldwide (Iowa: 84 49–58 Iowa State University, Ames) Lassaletta L, Billen G, Grizzetti B, Garnier J, Leach A M and Álvarez R, Steinbach H S and De Paepe J L 2014 A regional audit of Galloway J N 2014a Food and feed trade as a driver in the nitrogen fluxes in pampean agroecosystems Agric. Ecosyst. global nitrogen cycle: 50-year trends Biogeochemistry 118 Environ. 184 1–8 225–41 Anglade J, Billen G and Garnier J (under review). Relationships to Lassaletta L, Billen G, Romero E, Garnier J and Aguilera E 2014b estimate N fixation in legumes: incidence for N balance of How changes in diet and trade patterns have shaped the N organic legume-based cropping systems in Europe Ecosphere cycle at the national scale: Spain (1961–2009) Reg. Environ. Anonymous 1992 Fertilizer use by Crop, IFA-FAO (downloadable Change 14 785–97 from the FAO Fertstat website)(Rome: FAO) Llewely R V and Featherstone A M 1997 A comparison of crop Billen G, Garnier J and Lassaletta L 2013 The nitrogen cascade from production functions using simulated data for irrigated corn in agricultural soils to the sea: modelling N transfers at regional western Kansas Agric. Syst. 54 521–38 watershed and global scales Phil. Trans. R. Soc. B 368 Liu J G et al 2010 A high-resolution assessment on global nitrogen flows in cropland Proc. Natl. Acad. Sci. USA 107 8035–40 Billen G, Garnier J and Lassaletta L 2014 A biogeochemical view of Mueller N D, Gerber J S, Johnston M, Ray D K, Ramankutty N and the global agro-food system: nitrogen flows associated with Foley J A 2012 Closing yield gaps through nutrient and water protein production, consumption and trade Glob. Food management Nature 490 254–7 Security doi:10.1016/j.gfs.2014.08.003 Nijlandl G O, Schouls J and Goudriaan J 2008 Integrating the Blesh J and Drinkwater L E 2013 The impact of nitrogen source and production functions of Liebig, Michaelis–Menten, crop rotation on nitrogen mass balances in the Mississippi Mitscherlich and Liebscher into one system dynamics model River Basin Ecol. Appl. 23 1017–35 Wageningen J. Life Sci. NJAS 55 199–224 Bonaudo T, Bendahan A B, Sabatier R, Ryschawy J, Bellon S, Oenema O, Oudendag D and Velthof G L 2007 Nutrient losses from Leger F, Magda D and Tichit M 2014 A groecological manure management in the European Union Livest Sci. 112 principles for the redesign of integrated crop–livestock systems 261–72 Eur. J. Agron. 57 43–51 8 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Peoples M et al 2009 The contributions of nitrogen-fixing crop Sutton M A et al 2013 Our Nutrient World. The Challenge to legumes to the productivity of agricultural systems Symbiosis Produce More Food and Energy with Less Pollution 48 1–17 (Edinburgh: CEH and UNEP) Power J F and Alessi J 1971 Nitrogen fertilization of semiarid Swaney D P, Hong B, Ti C, Howarth R W and Humborg C 2012 Net grassland: plant growth and soil mineral N levels Agron. J. 64 anthropogenic nitrogen inputs to watersheds and riverine N 277–80 export to coastal waters: a brief overview Curr. Opin. Environ. Richard H 1951 Physionomie de la fertilisation en France (Paris: La Sust. 4 203–11 Maison rustique) Tilman D, Cassman K G, Matson P A, Naylor R and Polasky S 2002 Salvagiotti F, Cassman K G, Specht J E, Walters D T, Weiss A and Agricultural sustainability and intensive production practices Dobermann A 2008 Nitrogen uptake, fixation and response to Nature 418 671–7 fertilizer N in soybeans: a review Field Crop. Res. 108 1–13 van der Velde M et al 2014 African crop yield reductions due to Sattari S Z, van Ittersum M K, Giller K E, Zhang F and increasingly unbalanced Nitrogen and Phosphorus Bouwman A F 2014 Key role of China and its agriculture in consumption Glob. Change Biol. 20 1278–88 global sustainable phosphorus management Environ. Res. Lett. van Grinsven H J M et al 2012 Management, regulation and 9 054003 environmental impacts of nitrogen fertilization in northwestern Schierhorn F et al 2013 Post-Soviet cropland abandonment and Europe under the nitrates directive; a benchmark study carbon sequestration in European Russia, Ukraine, and Belarus Biogeosciences 9 5143–60 Glob. Biogeochem. Cycles 27 1175–85 van Ittersum M K et al 2013 Yield gap analysis with local to global Seitzinger S P et al 2010 Global river nutrient export: a scenario relevance-a review Field Crop. Res. 143 4–17 analysis of past and future trends Glob. Biogeochem. Cycles 24 van Ittersum M K and Rabbinge R 1997 Concepts in production GB0A08 ecology for analysis and quantification of agricultural input- Sinclair T R and Rufty T W 2012 Nitrogen and water resources output combinations Field Crop. Res. 52 197–208 commonly limit crop yield increases, not necessarily plant Vanlauwe B, Wendt J, Giller K E, Corbeels M, Gerard B and Nolte C genetics Glob. Food Sec. 1 94–8 2014 A fourth principle is required to define conservation Sheldrick W, Syers J K and Lingard J 2003 Contribution of livestock agriculture in sub-Saharan Africa: the appropriate use of excreta to nutrient balances Nutr. Cycl. Agroecosyst. 66 fertilizer to enhance crop productivity Field Crop. Res. 155 10–3 119–31 Velthof G L et al 2009 Integrated assessment of nitrogen losses from Soussana J-F and Lemaire G 2014 Coupling carbon and nitrogen agriculture in EU-27 using MITERRA-EUROPE J. Environ. cycles for environmentally sustainable intensification of Qual. 38 402–17 grasslands and crop-livestock systems Agric. Ecosyst. Environ. Vitousek P M et al 2009 Nutrient imbalances in agricultural 190 9–17 development Science 324 1519–20 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Environmental Research Letters IOP Publishing

50 year trends in nitrogen use efficiency of world cropping systems: the relationship between yield and nitrogen input to cropland

Loading next page...
 
/lp/iop-publishing/50-year-trends-in-nitrogen-use-efficiency-of-world-cropping-systems-vApN2Ioqsp

References (61)

Copyright
Copyright © 2014 IOP Publishing Ltd
eISSN
1748-9326
DOI
10.1088/1748-9326/9/10/105011
Publisher site
See Article on Publisher Site

Abstract

Nitrogen (N) is crucial for crop productivity. However, nowadays more than half of the N added to cropland is lost to the environment, wasting the resource, producing threats to air, water, soil and biodiversity, and generating greenhouse gas emissions. Based on FAO data, we have reconstructed the trajectory followed, in the past 50 years, by 124 countries in terms of crop yield and total nitrogen inputs to cropland (manure, synthetic fertilizer, symbiotic fixation and atmospheric deposition). During the last five decades, the response of agricultural systems to increased nitrogen fertilization has evolved differently in the different world countries. While some countries have improved their agro-environmental performances, in others the increased fertilization has produced low agronomical benefits and higher environmental losses. Our data also suggest that, in general, those countries using a higher proportion of N inputs from symbiotic N fixation rather than from synthetic fertilizer have a better N use efficiency. S Online supplementary data available from stacks.iop.org/ERL/9/105011/mmedia Keywords: nitrogen use efficiency, country and global scales, cropping systems, crop yields, nitrogen pollution 1. Introduction most crucial (Tilman et al 2002, Mueller et al 2012, Sinclair and Rufty 2012). The flipside of the coin, however, is an Although malnutrition has not receded in absolute terms, increased alteration of surface and groundwater resources, world agriculture, in the past half century, has succeeded in coastal eutrophication, air pollution and increased greenhouse increasing its production of vegetal proteins by a factor of 3 gas emission (Billen et al 2013, Sutton et al 2013). From this (Lassaletta et al 2014a). This has been made possible by perspective, very different situations exist, linked to the dis- parity of cropping system development in the countries and changes in cropping systems generally referred to as the regions of the world (Billen et al 2014). Green Revolution, based on the adoption of improved crop It is the purpose of this paper to describe these issues, varieties, use of pesticides, and increased application of based on an original analysis of the data available in the FAO synthetic fertilizers, among which nitrogen was by far the data base since 1961 (www.faostat.fao.org). Our approach is based on the calculation of the various components of the Content from this work may be used under the terms of the arable soil budget of 124 countries and, most importantly, on Creative Commons Attribution 3.0 licence. Any further the description of the trajectory drawn from 1961 to 2009 by distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. these countries in terms of their total crop production (Y, 1748-9326/14/105011+09$33.00 1 © 2014 IOP Publishing Ltd Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al –1 –1 expressed in harvested vegetal protein, kgN ha yr ) and the retained. This procedure allowed avoiding discrepancies in total N inputs onto cropland (F), excluding permanent the FAO data base. (See S1 for details). grassland, in the form of synthetic fertilizers, manure, sym- Total fertilization of cropland was defined as the total N biotic fixation and atmospheric deposition input in the form of synthetic fertilizers, symbiotic N fixation, –1 –1 manure application and atmospheric deposition onto crop- (also in kgN ha yr ). This approach differs from, and is land, excluding permanent grassland. The reason for focusing complementary to the Net Anthropogenic Nitrogen Input or our analysis on cropland is that the fate of the agricultural NANI approach (Howarth et al 2012). Our aim is to analyze surplus (excess N input over N export by plant harvest) cropping systems and to evaluate the excess N application on strongly differs between cropland and permanent grassland, arable land, the most sensitive components of agricultural particularly with respect to the relative proportions of NH systems, while the NANI approach deals holistically with the volatilization, denitrification, leaching and storage in the soil complete N cycle at the country scale, including the livestock organic pool (Velthof et al 2009, Billen et al 2013). Note that compartment and the effect of agricultural commodity trade temporary grassland (e.g. the FAOstat crop category ‘Grasses (Swaney et al 2012, Lassaletta et al 2014b). Nes for forage;Sil’), included within crop rotations, are con- Using a different approach, Conant et al (2013) have sidered as cropland. Yearly data on synthetic N fertilizer recently created a global soil N input database that enables application, under different N forms, for the entire period evaluation of trends in nitrogen use and recovery by country were obtained from the Resources module of the FAOstat for a number of important crops over the last 40 years. Their database. Countries with more than 15 missing annual data data show that differences in efficiency of N use between were removed from the analysis. Occasional gaps were filled OECD and other countries have persisted and exhibit no sign with data from the International Fertilizer Industry Associa- of convergence. In this paper we use the concept of the yield- tion (www.fertilizer.org/) if available and if not, by using fertilization relationship in an original way compared to the figures of the closest years. FAO data on annual per country concept commonly used, relating here the mean yield inte- synthetic fertilizer use refer to total use in agriculture and is grated over the entire crop rotation to the total fertilization of provided without distinction between arable and grassland. the cropland soils of a given territory. While the yield-ferti- We therefore had to subtract from these figures the proportion lization relationship is normally used in conventional agr- used for grassland fertilization, which in some European onomy as a tool to predict the yield increase of a given crop countries such as Ireland and the Netherlands accounts for a that could be expected from increasing fertilization in a given significant proportion. We have estimated the proportion of pedo-climatic context, we consider the integrative values of Y synthetic fertilizers to grasslands at the country scales pro- and F as overall indicators of the agronomical and environ- cessing the information compiled from different sources mental performances of a cropping system: the Y/F ratio is a (Richard 1951, Power and Alessi 1971, Anonymous 1992, measure of its nitrogen use efficiency (NUE), while the F-Y FAO 2006, Heffer 2013) (see S1 for details). difference is the regional N surplus (or N balance) repre- To estimate the crop biological nitrogen fixation by fix- senting the potential for hydrological or gaseous losses of ing crops included in the FAOstat database we used a yield- nitrogen to the environment. based approach, assuming that crop yield is the factor that best aggregates variables associated with crop, soil and cli- matic conditions including available N, soil moisture, vigor of stand, and other management factors influencing N fixation: 2. Methods Nfixed=⁎ %Ndfa ⁎ BGN , NHI Based on FAO data, we have reconstructed the trajectory followed by 124 countries in the past 50 years, in terms of where %Ndfa is the percentage of N uptake derived from N –1 –1 crop yield (Y refers to harvested crop part and is expressed in fixation, Y is the yield (expressed in kgN ha yr ), NHI is the –1 –1 kgN ha yr ) and total nitrogen inputs to cropland (F, sum N harvest index, defined as the ratio of the harvested material of nitrogen in manure, synthetic fertilizer, symbiotic fixation to the total above-ground N production, and BGN is a mul- –1 –1 and atmospheric deposition, in kgN ha yr ) for the tiplicative factor expressing the total N fixation including 1961–2009 period. Together these countries represent 99.2% below-ground contributions associated with roots, nodules of the world population and 99.6% of the cropland surface in and rhizo-deposition via exudates and decaying root cells and 2009 (see supplementary material S1, available at stacks.iop. hyphae. These parameters have been obtained from different org/ERL/9/105011/mmedia, for detailed methodology). sources (Herridge et al 2008, Salvagiotti et al 2008, Laberge Total annual crop production by each country was cal- et al 2009, Kombiok and Buah 2013, Álvarez et al 2014, culated taking into account the yearly harvested yield of 178 Anglade et al under review). We applied a regional %Ndfa primary crops and their N content, as reported in Lassaletta for soybean N fixation. For sugar cane, rice, paddy and forage et al (2014a). The cropland surface was estimated by sum- products, we applied a constant rate of biological fixation per ming up the surfaces of all individual crops. Only in the cases hectare, as suggested by Herridge et al (2008) (see S1 for where this sum was higher than the stated value of the ‘arable details). land and permanent crops’ surface area provided by the To estimate the animal excretion factors, we have fol- FAOstat resources module, the latest surface area was lowed the methodology of Sheldrick et al (2003) that assumes 2 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al that excretion rates, within a given livestock category, are the fact that, in the long run, harvest cannot exceed N resti- proportional to the slaughtered animal weights. We have tutions to the soil, and that the effect of low fertilization calculated different ratios for dairy and for other cattle stocks in strongly N-limited systems is characterized by a NUE close using the dairy stocks provided in the FAOstat ‘livestock to 1. The third property expresses the classical law of primary’ module. These stocks have been subtracted from the diminishing return and the fact that, in constant technical- total cattle stock to estimate non-dairy cattle. As a result, a agronomical context, some other limiting factor will always particular excretion factor has been applied to each type of impose a ceiling to production at saturating N availability. animal, country and year. The proportion of N excreted that is Two mathematical functions with only one parameter finally used as manure applied onto cropland was taken from obey both conditions: a hyperbolic function of the form the estimates of Sheldrick et al (2003) at the regional level for Y = Ymax*F/(F + Ymax) [1] and a negative exponential func- each type of animal. It was considered that 30% of the tion such as Y = Ymax [1 − exp(−F/Ymax)] [2]. We observed available manure is lost during management and storage that the former generally provides the best fit to the data. In before reaching the crop, as proposed by Oenema et al (2007) both cases the parameter Ymax represents the yield value for Europe and close to the value estimated by Liu et al reached at saturating N fertilization, as well as the value of (2010) at the world scale. We finally discount the amount of fertilization at which a definite fraction of this maximum yield N that is applied to permanent grasslands by applying the is reached (this fraction being 0.5 in the case of relation [1] or proportions provided by regions, and in some cases at the 0.63 for relation [2]). Over the 1961–2009 period, certain country scale, by Liu et al (2010) (see S1 for details). countries that we will call ‘type I’, such as China, Egypt and Deposition of oxidized and reduced nitrogen compounds India, present a simple trajectory with regularly increasing onto croplands was calculated from the GlobalNEWS data- fertilization and gradual reduction in the crop yield response, base (Seitzinger et al 2010) by extrapolating linearly between following a consistent and unique Y versus F relationship available years. The atmospheric deposition data used in (figure 1(a)). GlobalNEWS are derived for the year 2000 from Dentener Other countries (called ‘type II’), such as the USA, Brazil et al (2006) and previous figures were obtained by scaling and Bangladesh, display a historical trajectory with first a deposition fields for this year following Bouwman et al regularly increasing fertilization and yield, fitting the Y versus (2009). We calculated the input of N per ha (yearly national F relationship with a definite Ymax, then a turning point with average) and we applied this input per ha into the surface of a shift of the trajectory to another relationship with a sig- cropland considered in this study (see S1 for details). nificantly higher Ymax. This likely reflects improved agro- nomical practices in terms of production factors other than nitrogen, together with the pursuit of increasing fertilization. The turning point seems to have occurred in the 1980s or later 3. Results and discussion depending on the country (figure 1(b)). The case of the USA, for example, is consistent with a slowdown in the increase of 3.1. Y versus F trajectories of world agricultures synthetic fertilizers inputs from the 1980s parallel to a mod- The trajectory followed from 1961 to 2009 by a number of erate increase in the yields of the most important crops countries in terms of crop yield and total N inputs into (Howarth et al 2002, Alston et al 2010). cropland is shown in figure 1. The results for all countries of In most European countries (see the example of France, the FAO data base are provided in supplementary material the Netherlands and Greece in figure 1(c)), the trajectory also (S2). The Y versus F trajectory drawn by most countries shows a bi-phasic pattern, describing a regular increase in shows, at least for periods of several decades, a distinct cur- both fertilization and yield during the 1960–1975 period, vilinear relationship. Linear trajectories, like those described followed by a shift towards improved yields without further by de Wit (1992) for individual crops were seldom observed. increasing fertilization and even decreasing fertilization from Several mathematical formulations of the yield-fertilization the 1980s on (‘type III’). The case of the Netherlands is the relationship in a given pedo-climatic and technical-agrono- most spectacular, as in this country, which has always used mical context have been proposed in the agronomical litera- very high rates of fertilization, the level applied in recent ture, most of them involving negative exponential functions years has been reduced to the same as in the 1960s with, (Llewely and Featherstone 1997, Harmsen 2000). Nijland however, doubled yields. This trend is related to the reduction et al (2008) proposed to integrate the production functions of of N inputs prescribed by European environmental policies Liebig, Mitscherlich and Liebscher (de Wit 1992) into one and regulations (van Grinsven et al 2012), which interestingly system model based on Michaelis−Menten hyperbolic rela- did not prevent significant yield increases thanks to a better N tionships. Because we are expressing both output and input in management. Note, however, that despite the increase of –1 –1 exactly the same unit (kgN ha yr ) and because we are NUE and decrease in N surpluses, the nitrogen surplus looking for a simple long-term integrative theoretical rela- emitted to the environment in many cases remains much tionship, we decided to make use of the simplest possible higher than that of other countries belonging to types I and II. function obeying the three following properties: (i) the Finally, there is a small group of countries, such as function intercept should be zero; (ii) the slope of the function Morocco, Benin and Nigeria, whose trajectory does not dis- should be 1 at low fertilization; (iii) the function should reach play any consistent Y versus F relationship (type IV). These a plateau at high fertilization. The first two properties reflect countries have always very low inputs and yields. Very often, 3 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Figure 1. Examples of trajectories followed by countries in the Y versus F diagram. (a) Examples of type I trajectories. (b) Examples of type II trajectories. (c) Examples of type III trajectories. (d) Examples of type IV trajectories. R is the coefficient of determination, defined as: 2 ⎡ 2 2⎤ R =− 1 ∑(obs − calc ) /∑(obs —meanobs) where obs are the observed values of yield, calc the yield value calculated with the ii i i i ⎣ ⎦ relationship and meanobs is the average value of the observed yields over the period considered. Negative values of R indicate poor fit of the relationship on the observed values. This is often the case for the most recent period of type III trajectories because of still evolving agronomical conditions. 4 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Figure 2. Past and current agricultural potential of world countries, in terms of maximum protein yield of cropping systems (Ymax). their trajectory in the Y versus F diagram crosses the 1:1 line, indicating higher yield than fertilization. High inter-annual variation in the agricultural performance observed in some of these countries could be explained by weather phenomena, such as persistent water droughts, socio-political questions, or sometimes could be even an artefact due to the poor quality of our estimates of total nitrogen inputs to agricultural soil: in particular, in those countries where shifting agriculture is practiced, the fertilization of agricultural soil by the nitrogen stock accumulated in forested soil during the fallow period is not taken into account in our input estimations. However the ‘negative’ nitrogen balance displayed in the Y versus F dia- gram can also represent the signature of an unsustainable nitrogen mining of agricultural soils. For type I to III countries, we were able to define the Ymax values providing the best fit of the hyperbolic rela- tionship [1] to the points corresponding to the 1961–1980 period or later, and another Ymax for the most recent 10–15 years. The two Ymax values obtained characterize the past and current agricultural potential respectively, defined as the protein yield that could be obtained from cropland at a maximum N fertilization rate, with the corresponding crop- ping practices (figure 2). Comparison of the two periods shows a significant increase of Ymax in 45 countries (type II and III trajectories). For a large number (55) of countries, however, nearly the same parameter value or Ymax holds over the 50-year period (type I trajectory), as is the case for China, Egypt, Turkey, Chile, India and a many others. Possible N mining is indicated by higher crop yield than fertilization for 18 countries such as Canada, Morocco, Algeria, Iraq and Mozambique in the 1960–1980 period (see S2 for the com- plete list). In recent years, N mining continues in 10 African Figure 3. 50 years trends in nitrogen use efficiency of the cropping countries, as well as in Former Soviet Union countries, system of a number of countries. Afghanistan and Paraguay. N mining has been observed in Argentina for the entire studied period. This result is coherent with that recently reported by Álvarez et al (2014) that apparently wide margins for improvement through better indicate a budget of the copping system in the pampean fertilization practices, including an increasing use of legumes agroecosystems, which only becomes positive when includ- in crop rotations (Vanlauwe et al 2014). However, imbalances ing pasture lands. The severe problem of nutrient mining and with other nutrients such as P could limit yield responses to N loss of soil fertility in African countries has been frequently addition (van der Velde et al 2014). In the Former Soviet highlighted (Vitousek et al 2009, Liu et al 2010). In these Union, after the abrupt changes which occurred from 1989, countries yields are among the lowest in the world but have crops may have benefitted from nutrient legacies. The results 5 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Figure 4. (a) Distribution of the share of symbiotic fixation and synthetic fertilizers in total N inputs to cropland by countries in 2000–2009. (b) Observed relationship between NUE and the proportion of symbiotic fixation, or of synthetic fertilizers in total N inputs to cropland in the period 2000–2009. of our calculations for this country, however, might also be S2-1). This indicator is high (>0.75) in North America, affected by recent and poorly documented changes like Australia, most European and many sub-Saharan countries, massive land abandonment not fully documented by the FAO indicating margins for increasing yields by increasing N fer- (Schierhorn et al 2013). tilization. It is low (<0.3), on the other hand, in countries like China, India and Pakistan, as well as in a number of Central American and North African countries, indicating no benefit 3.2. Agronomical performances: trends in N use efficiency and in terms of yield to be expected from simple increase of N N-based yield gap fertilization in these regions in the absence of radical agro- The above-described trajectories can be translated in terms of nomical improvement of the cropping system. George (2014) changes in the NUE of the cropping system in the different has analyzed why crop yields in many developing countries countries (figure 3). Type I countries display a regularly cannot easily respond to increased inputs due to poor agro- decreasing trend of NUE. During the same period a similar nomical practices. drop for the phosphorus use efficiency (PUE) has been reported for China (Sattari et al 2014). In type II countries, the 3.3. Environmental performances: NUE and N losses shift in the trajectory toward an improved Ymax results in the stabilization or in the increase of NUE. In type III countries, The data we have assembled can also be used to estimate the the reduction of N inputs in recent years with no drop in yield contribution of agriculture in the different world countries to corresponds to increasing NUE. environmental nitrogen contamination, using the nitrogen To characterize the performance of agriculture of a given surplus (F-Y) as an indicator of potential losses. While in territory, van Ittersum et al (1997, 2013) introduced the grassland this surplus is generally stored in the soil organic concept of yield gap, defined as the difference between the matter pool, in the case of cropland, most of it is leached actual farmers’ yield of a particular cultivar and the potential quickly as nitrate, emitted as NH or denitrified as di-nitrogen, yield which could be obtained in the same place in the and nitrous oxide as a by-product, thus contributing to the absence of limitation by nutrient and water and with efficient nitrogen cascade of environmental contamination (Galloway control of pests and diseases. Following the same line of et al 2003). The data thus show the global distribution of reasoning, but applied to the overall production of the crop- environmental N losses from agricultural soils (figure S2-2). –1 –1 ping systems of each country in terms of proteins, we cal- Losses are over 50 kgN ha yr in most of Europe, the Middle culated (Ymax-Y)/Ymax as a dimensionless indicator of the East, the USA and Central America, India and China. –1 –1 degree of N limitation of current agricultural yields (figure They remain on average below 25 kgN ha yr in most 6 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al fertilizers in total fertilization. This higher efficiency of cropping systems relying largely on biological N fixation is observed for the largest soybean producers of South America as noted by Liu et al (2010) as well as for less productive countries in Africa and Asia with significant production of rice, groundnuts and beans. The higher NUE associated to nitrogen fixation is likely explained by a higher efficiency in the incorporation by legumes of their self-supplied nitrogen (Herridge and Peoples 1990). Also, the increase of the price of synthetic fertilizers might have encouraged the use of other sources of N in a most efficient way. 3.4. Global trends At the global scale, lumping together all cropping systems of the world, a type II Y/F trajectory is observed, with a shift during the 1980s from one Y/F relationship characterized by a –1 –1 Ymax of ∼70 kgN ha yr to an improved one with Ymax of –1 –1 110 kgN ha yr (figure 5(a)). The overall observed global trend is a distinct decrease of NUE in the 1961–1980 period (from 68% to 45%), followed by a stabilization during the last 30 years around 47% (figure 5(b)). The share of the different sources of N in the total inputs to cropland, depicted in figure 5(c), change considerably during the last 50 years, with synthetic fertilizers now being the largest source. Despite that the total rate of N excreted by livestock is equivalent to synthetic fertilizer application, the manure, rarely used effi- ciently, finally reaches the crops at a much lower rate which nowadays is slightly lower than crop biological N- fixation (figure 5(c)). 4. Conclusions Currently, only 47% of the reactive nitrogen added globally onto cropland is converted into harvested products, compared to 68% in the early 1960s, while synthetic N fertilizer input Figure 5. The global cropping system. (a) Trajectory followed by increased by a factor of 9 over the same period. This means global world cropping systems in the Y versus F diagram (Y: crop –1 –1 that more than half the nitrogen used for crop fertilization is yield in protein harvested, kgN ha yr ; F: total N inputs to the –1 –1 currently lost into the environment. Even though a significant cropland soil, kgN ha yr ). (b) Trends in nitrogen use efficiency of the global cropping system. (c) Evolution of the components of the improvement in NUE occurred in many countries after the global cropping system budget. 1980s, the present results suggest that a further increase of nitrogen fertilization would result in a disproportionately low increase of crop production with further environmental sub-Saharan Africa, the Former Soviet Union countries, and alterations, unless cropping systems improve their efficiency Australia. High surplus values are associated with low NUE substantially. In that respect, improvement of agronomical (figure S2-2). practices and development and proper application of envir- Total fertilization, as discussed above, is mainly the sum onmental policies have been demonstrated to be efficient of synthetic fertilizers, manure application and symbiotic strategies. A better integration of crop and livestock systems nitrogen fixation. Although not negligible, atmospheric N can also contribute to increasing NUE at the local and global deposition generally contributes a much smaller share. The scale (Herrero et al 2010, Lassaletta et al 2014a, Bonaudo proportion of the three former N inputs to overall fertilization et al 2014, Soussana and Lemaire, 2014). Moreover, our data varies a great deal among the different world cropping sys- suggest that an increase in the contribution of symbiotic N tems, as shown in figure 4(a). Our data show that NUE is fixation would result in increasing NUE. Peoples et al (2009) generally higher (and the N surplus relatively lower) for have stressed that the potential of symbiotic nitrogen fixation agricultural systems with higher proportion of N inputs is currently largely underexploited, given that very few derived from symbiotic N fixation (figure 4(b)). Conversely, countries have a fraction of arable land devoted to legume NUE is generally lower for a higher proportion of synthetic crops greater than a few percent. Increased areas of legumes 7 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al might be achieved by including more leguminous crops in Bouwman A F, Beusen A H W and Billen G 2009 Human alteration of the global nitrogen and phosphorus soil balances for the rotations, or by the introduction of short-duration legume period 1970–2050 Glob. Biogeochem. Cycles 23 GB0A04 green manures or ‘catch crops’ (Blesh and Drinkwater 2013). Conant R T, Berdanier A B and Grace P R 2013 Patterns and trends By evidencing the long-term response of N inputs to the in nitrogen use and nitrogen recovery efficiency in world soil in terms of production and potential losses to the envir- agriculture Global Biogeochem. Cycles 27 558–66 onment, this paper provides a summarized and comprehen- Dentener F et al 2006 Nitrogen and sulfur deposition on regional and global scales: a multimodel evaluation Global Biogeochem. sive diagnosis of the effective changes in agronomical and Cycles 20 B4003 environmental performances of the cropping systems of 124 de Wit C T 1992 Resource use efficiency in agriculture Agric. Syst. countries of the world. 40 125–51 FAO 2006 Fertilizer and Plant nutrition Bulletin n°17 (downloadable from the FAO Fertstat website)(Rome: FAO) Galloway J N et al 2003 The nitrogen cascade BioScience 53 341–56 George T 2014 Why crop yields in developing countries have not Acknowledgments kept pace with advances in agronomy Glob. Food Sec. 3 49–58 Harmsen K 2000 A modified Mitscerlich equation for rainfed crop This study was partly supported by the SeasERA EMoSEM production in semi-arid areas: 1. Theory Netherlands J. Agric. project (ANR-12-SEAS-0005-01). We wish to thank the Sci. 48 237–50 Heffer P 2013 Assessment of Fertilizer Use by Crop at the Global FIRE (Fédération Ile de France de Recherche en Environne- Level. 2010–2011/10. International Fertilizer Industry ment, CNRS and UPMC). We thank Javier Castrillo who Association (Paris: IFA) developed several computer routines for the data management Herrero M et al 2010 Smart investments in sustainable food and Angel Udias for his help in preparing the figures. We are production: revisiting mixed crop-livestock systems Science sincerely grateful to two anonymous reviewers for their 327 822–5 Herridge D F and Peoples M B 1990 Ureide assay for measuring detailed and constructive revision. We thank Augustin del nitrogen-fixation by nodulated soybean calibrated by n-15 Prado, Eduardo Aguilera, Guillermo Pardo, Fernando methods Plant Physiol. 93 495–503 Estellés, Alberto Sanz-Cobena and Dennis Swaney for help- Herridge D F, Peoples M B and Boddey R M 2008 Global inputs of ful suggestions and comments. We also thank to the Scientific biological nitrogen fixation in agricultural systems Plant Soil Network on GHG mitigation from the agroforestry sector in 311 1–18 Howarth R W, Boyer E W, Pabich W J and Galloway J N 2002 Spain (REMEDIA) for providing an space for discussion and Nitrogen use in the United States from 1961–2000 and science developement. The views expressed in this document potential future trends Ambio 31 88–96 are purely those of the authors and may not in any circum- Howarth R, Swaney D, Billen G, Garnier J, Hong B, Humborg C, stances be regarded as stating an official position of the Johnes P, Mörth C-M and Marino R 2012 Nitrogen fluxes from European Commission. the landscape are controlled by net anthropogenic nitrogen inputs and by climate Front Ecol. Environ. 10 37–43 Kombiok J and Buah S 2013 Tillage depth effects on nodulation, nitrogen fixation and yield of three soybean varieties in the Northern Savanna zone of Ghana Afr. J. Agric. Res. 8 2340–5 References Laberge G, Franke A C, Ambus P and Hogh-Jensen H 2009 Nitrogen rhizodeposition from soybean (Glycine max) and its impact on nutrient budgets in two contrasting environments of Alston J, Babcock B and Pardey P 2010 The Shifting Patterns of the Guinean savannah zone of Nigeria Nutr. Cycl. Agroecosys Agricultural Production and Productivity Worldwide (Iowa: 84 49–58 Iowa State University, Ames) Lassaletta L, Billen G, Grizzetti B, Garnier J, Leach A M and Álvarez R, Steinbach H S and De Paepe J L 2014 A regional audit of Galloway J N 2014a Food and feed trade as a driver in the nitrogen fluxes in pampean agroecosystems Agric. Ecosyst. global nitrogen cycle: 50-year trends Biogeochemistry 118 Environ. 184 1–8 225–41 Anglade J, Billen G and Garnier J (under review). Relationships to Lassaletta L, Billen G, Romero E, Garnier J and Aguilera E 2014b estimate N fixation in legumes: incidence for N balance of How changes in diet and trade patterns have shaped the N organic legume-based cropping systems in Europe Ecosphere cycle at the national scale: Spain (1961–2009) Reg. Environ. Anonymous 1992 Fertilizer use by Crop, IFA-FAO (downloadable Change 14 785–97 from the FAO Fertstat website)(Rome: FAO) Llewely R V and Featherstone A M 1997 A comparison of crop Billen G, Garnier J and Lassaletta L 2013 The nitrogen cascade from production functions using simulated data for irrigated corn in agricultural soils to the sea: modelling N transfers at regional western Kansas Agric. Syst. 54 521–38 watershed and global scales Phil. Trans. R. Soc. B 368 Liu J G et al 2010 A high-resolution assessment on global nitrogen flows in cropland Proc. Natl. Acad. Sci. USA 107 8035–40 Billen G, Garnier J and Lassaletta L 2014 A biogeochemical view of Mueller N D, Gerber J S, Johnston M, Ray D K, Ramankutty N and the global agro-food system: nitrogen flows associated with Foley J A 2012 Closing yield gaps through nutrient and water protein production, consumption and trade Glob. Food management Nature 490 254–7 Security doi:10.1016/j.gfs.2014.08.003 Nijlandl G O, Schouls J and Goudriaan J 2008 Integrating the Blesh J and Drinkwater L E 2013 The impact of nitrogen source and production functions of Liebig, Michaelis–Menten, crop rotation on nitrogen mass balances in the Mississippi Mitscherlich and Liebscher into one system dynamics model River Basin Ecol. Appl. 23 1017–35 Wageningen J. Life Sci. NJAS 55 199–224 Bonaudo T, Bendahan A B, Sabatier R, Ryschawy J, Bellon S, Oenema O, Oudendag D and Velthof G L 2007 Nutrient losses from Leger F, Magda D and Tichit M 2014 A groecological manure management in the European Union Livest Sci. 112 principles for the redesign of integrated crop–livestock systems 261–72 Eur. J. Agron. 57 43–51 8 Environ. Res. Lett. 9 (2014) 105011 L Lassaletta et al Peoples M et al 2009 The contributions of nitrogen-fixing crop Sutton M A et al 2013 Our Nutrient World. The Challenge to legumes to the productivity of agricultural systems Symbiosis Produce More Food and Energy with Less Pollution 48 1–17 (Edinburgh: CEH and UNEP) Power J F and Alessi J 1971 Nitrogen fertilization of semiarid Swaney D P, Hong B, Ti C, Howarth R W and Humborg C 2012 Net grassland: plant growth and soil mineral N levels Agron. J. 64 anthropogenic nitrogen inputs to watersheds and riverine N 277–80 export to coastal waters: a brief overview Curr. Opin. Environ. Richard H 1951 Physionomie de la fertilisation en France (Paris: La Sust. 4 203–11 Maison rustique) Tilman D, Cassman K G, Matson P A, Naylor R and Polasky S 2002 Salvagiotti F, Cassman K G, Specht J E, Walters D T, Weiss A and Agricultural sustainability and intensive production practices Dobermann A 2008 Nitrogen uptake, fixation and response to Nature 418 671–7 fertilizer N in soybeans: a review Field Crop. Res. 108 1–13 van der Velde M et al 2014 African crop yield reductions due to Sattari S Z, van Ittersum M K, Giller K E, Zhang F and increasingly unbalanced Nitrogen and Phosphorus Bouwman A F 2014 Key role of China and its agriculture in consumption Glob. Change Biol. 20 1278–88 global sustainable phosphorus management Environ. Res. Lett. van Grinsven H J M et al 2012 Management, regulation and 9 054003 environmental impacts of nitrogen fertilization in northwestern Schierhorn F et al 2013 Post-Soviet cropland abandonment and Europe under the nitrates directive; a benchmark study carbon sequestration in European Russia, Ukraine, and Belarus Biogeosciences 9 5143–60 Glob. Biogeochem. Cycles 27 1175–85 van Ittersum M K et al 2013 Yield gap analysis with local to global Seitzinger S P et al 2010 Global river nutrient export: a scenario relevance-a review Field Crop. Res. 143 4–17 analysis of past and future trends Glob. Biogeochem. Cycles 24 van Ittersum M K and Rabbinge R 1997 Concepts in production GB0A08 ecology for analysis and quantification of agricultural input- Sinclair T R and Rufty T W 2012 Nitrogen and water resources output combinations Field Crop. Res. 52 197–208 commonly limit crop yield increases, not necessarily plant Vanlauwe B, Wendt J, Giller K E, Corbeels M, Gerard B and Nolte C genetics Glob. Food Sec. 1 94–8 2014 A fourth principle is required to define conservation Sheldrick W, Syers J K and Lingard J 2003 Contribution of livestock agriculture in sub-Saharan Africa: the appropriate use of excreta to nutrient balances Nutr. Cycl. Agroecosyst. 66 fertilizer to enhance crop productivity Field Crop. Res. 155 10–3 119–31 Velthof G L et al 2009 Integrated assessment of nitrogen losses from Soussana J-F and Lemaire G 2014 Coupling carbon and nitrogen agriculture in EU-27 using MITERRA-EUROPE J. Environ. cycles for environmentally sustainable intensification of Qual. 38 402–17 grasslands and crop-livestock systems Agric. Ecosyst. Environ. Vitousek P M et al 2009 Nutrient imbalances in agricultural 190 9–17 development Science 324 1519–20

Journal

Environmental Research LettersIOP Publishing

Published: Oct 1, 2014

There are no references for this article.