12.5.1 Soil Sampling

Taking a representative soil sample is both the first step and the largest source of error in soil fertility evaluation, because of the magnitude of extrapolation of the analytical results. When a 5-g subsample taken from a 500-g composite sample of a 2.5-ha field at 15-cm depth is used for phosphorus determination, it represents one-billionth (10^–9) of the total soil volume for which the analysis is made (Perur et al. Reference Perur, Subramanian, Muhr and Ray1974).

Soil scientists assume that the distribution of printed instructions and diagrams on how to take soil samples is sufficient to be assured of representative specimens. Unfortunately, this is not always the case. A representative soil sample is a composite of 10–20 samples of the root zone of a field with no major variation in slope, drainage, color or past fertilizer history. For deep-rooted plants, I recommend taking a 0–20-cm and 20–50-cm sample, and for the deepest rooted ones, a 50–100-cm and 100–300-cm sample. Non-representative areas, such as fence rows, termite mounds, those where straw has been burned and manure piles must be avoided. Appropriate information is also needed, including the name and address of the farmer, georeference, previous crop, and fertilizer practices. This is best provided with a web-based questionnaire in the local language that can be sent electronically to a central laboratory. An example for maize-growing in Mindanao, Philippines, has been successfully developed by the International Rice Research Institute (IRRI 2011).

For established pastures or permanent crops where fertilizers are not likely to be incorporated, sampling the top 20 cm is usually sufficient. For nitrate analysis, the 20–50-cm subsoil is usually sampled, and nitrate sampling tools and procedures exist. Subsoil samples, to at least 50 cm and beyond, are important in oxidic subsoils for estimates of anion exchange capacity, gypsum requirements and deep SOC.

Another important question is where to sample when phosphorus has been applied in bands. The answer is between the bands, if their locations are known. The same is true for conservation tillage. A third consideration is the time of the year. Soil samples should be taken a long time before planting so that the results will be available when the decision is made as to how much fertilizer is to be applied. In ustic soil moisture regimes, this means sampling during the dry season, when upward ion movement takes place. This may change some of the soil test results, principally in regard to potassium.

How often a field should be sampled depends on the intensity of fertilizer use and the economic value of the crop. For average management intensity, once every 3 years has been recommended by the International Fertilizer Evaluation and Improvement Program (ISFEIP; 1967–1974), while for very intensively managed areas, annual sampling is necessary. In my opinion, deep soil sampling in oxidic soils should be carried out once every 5 years, or at a changing point in a rotation. Deep sampling and related shallower sampling can estimate changes in soil carbon at that time.

As soil testing is a service program, the time between soil sampling and its analysis should be minimized. As emphasized in the first edition, delays in shipping, power failures and the absence of some reagents in many tropical laboratories are unfortunately very common. Several studies have been made on the effect of time in storage and drying on analytical results. This is still valid, four decades later, in much of the tropics.

After arrival at the laboratory, the soil sample is dried, then ground, either by pounding or by electric-powered grinders, passed through a 2-mm sieve, and assigned a laboratory number or stored until ready for analysis. Bouldin et al. (Reference Bouldin, Greweling and Lui1971) studied the effect of drying versus keeping several Oxisols and Ultisols moist for a period of months. Drying decreased the exchangeable aluminum ion (Al³⁺) content by 20 percent and increased the exchangeable ammonium (NH₄⁺) content when both were extracted by 1 M KCl. No changes in pH, available phosphorus, calcium or magnesium were observed.

12.5.2 Tackling Soil Variability

Spatial variability in soil properties is always an issue because soil is far from a homogeneous system and bears the footprints of generations of farmers who add temporal variability. The question is how wide-ranging is the variability. I have seen thousands of hectares of very uniform soils that follow a predictable pattern due to landscape position in Oxisols of South America, where large-scale farming is practiced. The uniformity of these soils results from the deep, weathered sediments that they are derived from; variability of these soils is minimal, and is found mainly along slopes. Local extension workers recommend sampling every tens or hundreds of hectares, every few years.

The mixing and leveling of flooded and puddled rice soils for decades or centuries in Asia acts essentially like the blender we use in our kitchens, making a pretty uniform slurry. Variability is thus minimized. At the other extreme, in smallholder farms in Africa, the human footprint is enormous. Lots of natural spatial variability exists in a mosaic of soils due to the topography and because the soils are derived from different kinds of rock, and also from management, as illustrated in Fig. 12.5. When people farm 1 or 2 hectares, they get to know their fields very well, and such knowledge helps in determining where to sample.

Fig. 12.5 In-field variability in a maize field in Koraro, Ethiopia.

Photo courtesy of Ray Weil

Fields close to the household (infields) are usually more fertile than those farther away (outfields) because smallholder households dump garbage and kitchens ashes, and their domestic animals often urinate and defecate in these nearby fields. This results in a fertility gradient, illustrated in Fig. 12.6, using a portable pH meter.

Fig. 12.6 Home garden soils are almost always much more fertile than soils in the larger fields away from home. Mbola, Central Tanzania.

Courtesy of Ray Weil

12.5.3 Laboratory Analysis

The most widely used way to assess the status of a nutrient in a soil is by an extraction in a wet chemistry laboratory. Soils are shaken with an extractant, which is supposed to indicate in a few minutes the amount of a nutrient a typical crop would take up from the soil over its lifetime. These values are variously known as available phosphorus, potassium, etc., usually including the name of the procedure’s author (Olsen, for example). Water is also used as an extractant, and resin, to capture ions in a soil slurry. All these extracted products are artifacts (like humic or fulvic acids), which makes them difficult to relate to reflectance spectrometry on the whole soil, but when they correlate with plant growth, they become useful metrics. Because of its mobility in soils, there are no similar soil tests for nitrate nitrogen (nitrate-N).

The wet chemistry laboratory is the backbone of a soil testing program. Unlike research laboratories, service laboratories must be geared to handle large numbers of samples, rapidly and accurately. A complete system of semi-automated apparatus was developed by the ISFEIP for tropical laboratories in the 1970s and 1980s. This system is summarized below, abstracted from the first edition of this book, when ISFEIP was being rapidly scaled up in Latin America and India. Other models are now in place.

Soil samples are measured volumetrically, eliminating the time-consuming weighing process, and then placed in multiple-unit trays, where the extracting and diluting solutions are added or transferred by specially designed diluter-dispensers. Other apparatus transfers the aliquots to spectrophotometers and pH-measuring units automatically. Shaking, stirring and cleaning apparatus are also semiautomatic and capable of handling 30 units at a time. A single technician can run 100 soil samples a day, measuring 10 determinations per sample. The pieces of apparatus are based on manual operation and use a minimum of electrical power. Expensive electronic equipment, such as atomic absorption spectrophotometers, is used only during the last stages. An additional advantage is that the same units can be used for plant tissue as for soil analysis.

The bulk of the work is done with two extractants: the dilute double acid (Mehlich) or the modified Olsen extractant for determining available phosphorus, potassium, calcium, magnesium, sodium, iron, manganese, zinc and copper; and the 1 M KCl extraction for aluminum. Routine tests for nitrogen, boron, sulfur and molybdenum could not be adapted to the system.

12.5.4 Soil Test Correlation

A soil test value per se is worthless; it is an empirical number that may or may not indirectly reflect nutrient availability, and as mentioned before, what is measured is an artifact of the extraction procedure. Soil test values become useful only when they are correlated with crop responses. Such correlations are usually conducted at two levels: an exploratory one in the greenhouse with a large number of widely diverse soils, and a more definite one in the field with fewer, but carefully selected, soils. Correlations can be done in many ways, using fairly complicated formulas, but in my view the simplest and most practical one is to plot relative yields versus the soil test levels as a scatter diagram. Then simply use a transparent overlay with a cross, placing it where the number of points in the upper left quadrant and in the lower right quadrant are at the minimum. This Cate–Nelson method (Cate and Nelson Reference Cate and Nelson1971) gives the soil critical level, as illustrated in Fig. 12.7. (I am sure a software program has been developed to do the same.)

Fig. 12.7 The Cate–Nelson method for determining the critical soil test level. In this case the critical level is 6 ppm available phosphorus by the Bray 1 method, for sugar cane, in the State of Pernambuco, Brazil. Each dot represents a field plot

(ISFEIP 1967)

This concept of a “critical value” for a given soil and farming system has important practical implications for efficient phosphorus use. Maintaining the soil at or close to the critical value has important benefits to the farmer (in terms of economic return) and to the environment, in terms of reducing the risk of phosphorus transfers to surface waters (Syers et al. Reference Syers, Johnston and Curtin2008).

The Cate–Nelson method also tells us the points with a high probability of a phosphorus response (the lower left quadrant), and those with a low probabiliy of phosphorus response (in the upper right quadrant). It does not tell you how much fertilizer to add, but it tells you whether you need it or not. This could be improved with better estimates of uncertainty.

If the soil in question is above the critical level, the decision is clearly not to apply that particular nutrient element to that crop. If the soil is below the critical level, then the decision becomes how much to apply. This requires yield response field trials. There are two main types of models to express yield response curves: the classic curvilinear model and the simpler linear reponse and plateau model.

12.5.5 Curvilinear Models

These classic models are based on the law of diminishing returns, where curvilinear functions (quadratic, square-root, logarithmic, Mitscherlich, etc.) are fitted to the yield response data. Statistical techniques determine which function fits the data best by providing the highest coefficient of determination (R²). The optimum fertilizer rate is determined at that point where marginal revenue equals the marginal cost (i.e. the point where the price of the last yield increment equals the cost of the last increment of fertilizer). The point can be determined mathematically or graphically by drawing a price:cost ratio line, expressed in the agronomic equivalents in the yield response diagram. The optimum yield is then determined at the point where a tangent of the price:cost ratio line intersects the response curve. When more than one element is deficient, regression models take this into account, as well as the interactions between these elements. The recommended rates are determined by solving simultaneous equations.

Curvilinear models are also used in a different manner to produce one function that takes into account soil test levels as variables, as well as other variables related to soil, climate and management properties, in an attempt to account for the uncontrolled variables. A yield equation relating potato yields in Peru needed 27 variables (Ryan and Perrin Reference Ryan and Perrin1973). Optimum fertilizer levels are calculated on the basis of levels of crop prices, fertilizer costs, expected rainfall and soil properties.

Such complex models are effective when there is adequate information about the variables involved and when prices are stable. They fail in the tropics where there are insufficient data to quantify all variables. The use of complex models in tropical regions is limited to after-the-fact analysis in areas with detailed information; they do not serve successfully as predictive tools. In addition, Anderson and Nelson (Reference Anderson and Nelson1975) found that quadratic models are biased when there is a marked response to the first fertilizer increment, followed by little or no response at higher rates. In such cases, optimum fertilizer recommendations became unrealistically high.

12.5.6 The Linear Response and Plateau Model

A series of studies conducted in England by Boyd (Reference Boyd1970) and in the United States by Bartholomew (Reference Bartholomew1972) summarized many fertilizer response functions over the world and concluded separately that in most instances fertilizer response curves can be characterized by a sharp linear increase followed by a flat horizontal line. Bob Cate and Larry Nelson in North Carolina then developed the Linear Response and Plateau model (LRP), which is also based on Liebig’s law of the minimum, and is a logical extension of the Cate–Nelson correlation model (Waugh et al. Reference Waugh, Cate, Nelson, Manzano, Bornemisza and Alvarado1975). Fertilizer response from a field, or a group of fields, is represented in this model by two straight lines for each nutrient. The first line represents the relatively steep response of an added nutrient until it ceases to be a major limiting factor. This is followed by a line showing a flat plateau, where further additions no longer increase yields. The fertilizer response curve so constructed consists of two main points. The “threshold yield” is the yield at the zero level of the nutrient in question, but not of all nutrients. The “plateau yield” is the yield at the point where the nutrient ceases to be a limiting factor; it is not the maximum yield because other factors may still limit yields. The fertilizer rate needed to reach the plateau yield is the recommended rate for the particular nutrient.

The comparison between the two approaches is shown in Fig. 12.8 for the same data set. The dotted lines indicate how the fertilizer recommendations are arrived at. The LRP model has only one optimum point, independent of cost and prices. The curvilinear model shows an optimum point based on the particular price:cost ratio at the time the experiments were conducted.

Fig. 12.8 Comparison between the LRP model and the quadratic model using the same field data from potatoes in Peru. The recommended rate is much lower using the first model.

Adapted from Waugh et al. (Reference von Liebig1973)

It is very important to note the wide differences in recommended rates between the two methods. The LRP model recommended a lower fertilizer rate (75 kg N/ha) to reach a yield plateau of about 19 t/ha of potatoes. This rate in effect provides nearly maximum yields, while preserving an efficient return per unit of fertilizer, because it is still along the increasing slope. The quadratic model in this case more than doubles the recommendation (160 kg N/ha) in order to obtain only an additional 1 t/ha of yield. This is in the relatively flat part of the curve, where variability is quite high. Small changes in yields result in large changes in recommended rates. In other comparisons, however, the difference in recommended rates might not be as large as in this example.

Which approach is right? I tried it myself with a set of field trials on the response to sulfur-coated urea by wetland rice in the Coast of Peru, where I used the quadratic method (Sanchez et al. Reference Waugh, Cate and Nelson1973). By using the graphic technique, plateau yields and recommended rates were computed. The average nitrogen recommendation in this very high-yielding environment was 224 kg N/ha according to the quadratic model, and 170 kg N/ha with the LRP model. The average gross return for fertilization at the recommended rates was not statistically significant between the two models. The net return per dollar invested in fertilizer was superior with the LRP model.

The LRP model is much simpler, and focuses on the range that provides the most bang for the buck, the linear portion at the threshold yield point. The quadratic models, in their search for the optimum value along the flat part of the response curve, are the least cost-effective. While the science is correct in their case (fertilizer response curves are indeed curves), the practicalities are not, because there is wide variability in the actual results under field conditions. As Bob Cate famously said “If you want to go (from North Carolina) to Tokyo you go via the Great Circle route, but if you want to go from this building to the next, you go in a straight line.” So let’s be as simple and practical as we can.

It is heartening to know that the LRP model is now widely used in US agriculture (Beegle Reference Beegle, Sims and Sharpley2005).

12.6.1 Spectrometry

The analysis of wavelengths of light is one of the simplest in physics, and not until the start of this century has it been applied to soil science at scale. Spectrometry has for decades been used as a standard in various industries, such as pharmaceuticals and paints. The near- and mid-infrared wavelengths can be correlated with many, but not all soil properties. In the tropics, the seminal papers by Keith Shepherd and Markus Walsh (Shepherd and Walsh Reference Shepherd and Walsh2002, Reference Shepherd and Walsh2007) showed the value of then near-infrared rudimentary spectrometers that can estimate many soil properties, just by passing light though a soil sample and making the necessary correlations (Fig. 12.9). The main properties that can be determined at present are: sand, silt, clay, organic carbon, exchangeable bases, cation exchange capacity, phosphorus retention and some soil minerals. Where spectrometry does not (yet) work is available phosphorus, potassium and sulfur, very critical soil fertility parameters. Spectrometry is also helpful in determining organic input quality and carbon mineralization rate (Shepherd et al. Reference Shepherd, Vanlauwe, Gachengo and Palm2005).

Fig. 12.9 Keith Shepherd (top left) using a portable spectrometer in the field in Kenya, and a technician using it in his Nairobi laboratory (bottom left). The resulting soil spectral signature (top right), and the close correlation between the predicted value by spectroscopy and the “actual” value of SOC from a wet chemistry laboratory (bottom right).

12.6.2 Digital Soil Mapping

Traditional soil maps are composed of polygons (mapping units) and their legends, showing mostly soil classification terms that are of very limited information value to anyone other than a pedologist. Such maps are static and tell you nothing about soil properties. Polygon maps assume that soils are uniform within the mapping unit, which is seldom the case. Interpretations made from them in terms of soil properties and agronomy are of course no better than the map.

To address some of these limitations, a digital soil-mapping working group of the International Union of Soil Sciences (IUSS) started work in 2006, and has since produced valuable information that can be found in books arising from their conferences (Lagacherie et al. Reference Lagacherie, McBratney and Voltz2007, Hartemink et al. Reference Hartemink, McBratney and Mendonça-Santos2008, Boettinger et al. Reference Boettinger, Howell, Moore, Hartemink and Kienast-Brown2010). In 2008, a group of soil scientists, agronomists and ecologists decided to produce a digital map of soil properties and soil functions of the world (Sanchez et al. Reference Sanchez, Ahamed, Carré, Hartemink, Hempel, Huising, Lagacherie, McBratney, McKenzie, Ml, Minasny, Montanarella, Okoth, Palm, Sachs, Shepherd, Vågen, Vanlauwe, Walsh, Winowiecki and Zhang2009, Fisher Reference Fisher2012), and the African portion of that is in progress.

Soil properties are estimated quantitatively with a statistical inference system using spatial (kriging, co-kriging) or non-spatial (regressions, neural networks) techniques, or both, and expressed as their probability of occurrence with uncertainty estimates. They are derived using quantitative relationships between the punctual soil measurements and the spatially continuous soil covariates. This results in maps of soil properties such as the ones selected by the Global Soil Map Consortium as their minimum data set, i.e. clay, organic carbon, bulk density, pH, effective cation exchange capacity, electrical conductivity and bulk density – bulk density is needed to express carbon and nutrients on a mass basis (t/ha) for biogeochemical modeling (Sanchez et al. Reference Sanchez, Ahamed, Carré, Hartemink, Hempel, Huising, Lagacherie, McBratney, McKenzie, Ml, Minasny, Montanarella, Okoth, Palm, Sachs, Shepherd, Vågen, Vanlauwe, Walsh, Winowiecki and Zhang2009). These are the first two actions in the process for developing the digital soil map, as shown in Fig. 12.10. Such properties are mapped on a pixel basis with a resolution as small as 30 × 30 m, which is great for smallholder farms.

Fig. 12.10 Conceptual flow of the soil digital map process.

Chart based on text by Sanchez et al. (Reference Sanchez, Ahamed, Carré, Hartemink, Hempel, Huising, Lagacherie, McBratney, McKenzie, Ml, Minasny, Montanarella, Okoth, Palm, Sachs, Shepherd, Vågen, Vanlauwe, Walsh, Winowiecki and Zhang2009)

The spatially inferred soil properties are then used to predict more difficult-to-measure soil functions (action 3, Fig. 12.10), such as available soil water storage, carbon density and phosphorus fixation, using pedotransfer functions. Rather than calibrating only individual soil properties to infrared spectra, multivariate associations are used to identify soil functions. The overall uncertainty of the prediction is assessed by combining the uncertainties of the input data with the uncertainties of the spatial inference and those of the soil functions, using global sensitivity analysis.

The key is to connect all these soil functions with fertilizer response or other georeferenced field data from well-conducted agronomic experiments (legacy data) as well as with digital maps of population density, roads, distance to markets, internet access and other socioeconomic digital maps (action 4, Fig. 12.10). Analyses of decisions need to be made, connecting the maps with actual on-the-ground experience. Then the final two actions (management recommendations and reaching the end user) can be achieved. Web-based digital soil maps can be made interactive. This is very much work in progress.

12.6.3 Bringing the Laboratory to the Field

Although soil nutrient replenishment in sub-Saharan Africa is widely recognized as a critical biophysical entry point to an agricultural transformation, the actual application of soil science is still minimal, and frustratingly so, in large part because of the delay in obtaining the necessary information from soils sampled in farmers’ fields and sent to a wet-chemistry, conventional laboratory. The data often take a year to come back (if at all) and the paper-based process is prone to errors of recording and transcribing, as well as delays in transmission, interpretation and visualization. Worse still, the data are usually hard to interpret, largely irrelevant and often of questionable quality. The poor infrastructure of Africa and the budget limitations of national institutions are the main reasons used to rationalize this useless exercise. This is not a problem in the more advanced tropical countries, but it is also a problem in many areas outside Africa as well.

The way to get around this is to get the data in situ and send it electronically to the central laboratory where senior people or algorithms make the recommendations and send it back electronically to the extension agent or farmer. The problem is that, up until recently, field soil test kits have proven to be highly inaccurate and not sufficiently precise (Vanlauwe Reference Vanlauwe2008). Weil (Reference Weil2011) of the University of Maryland and Columbia University has been developing a wet-chemistry “SoilDoc,” which has been proven to work well and is highly correlated with central wet-chemistry laboratories (Fig. 12.11). It enables extension workers to make on-the-spot diagnoses of soil constraints, allowing targeted recommendations to advise farmers in real time (Sanchez et al. Reference Sanchez, Weil and Rose2013).

Fig. 12.11 SoilDoc, a “lab-in-the-box” field test kit. Ray Weil is entering the data in a smart phone that transmits it to the cloud or to a central location. Arusha, Tanzania, September 2012.

This tool uses state-of-the-art battery-powered miniaturized instruments, originally created for other purposes such as the beer and swimming-pool testing industries. The soil fertility parameters analyzed include soil pH, electrical conductivity (indicative of general fertility levels as well as salinity issues), biologically active organic carbon, and 0.01 M calcium chloride (CaCl₂)-extractable nitrate-N, sulfate-S, phosphate-P, and potassium using specific sensors. Additionally, the kit includes tools to measure soil physical properties such as surface sealing strength, plow pans and volumetric soil water content. Soil texture is estimated by feel. SoilDoc essentially measures what a conventional wet-chemistry laboratory does at a similar level of accuracy (Fig. 12.12) plus biologically active SOC and several physical properties (Weil Reference Weil2011, Sanchez et al. Reference Sanchez, Weil and Rose2013). Like digital soil mapping, this too is very much work in progress.

Fig. 12.12 Correlations for pH (top), nitrate-N (middle), and available phosphorus (bottom), determinations by the SoilDoc kit and a wet-chemistry laboratory.

Reproduced with permission from Ray Weil, University of Maryland

At the time of this writing, we need to use the field-based wet chemistry of SoilDoc and similar systems in conjunction with near-infrared spectrometers, because spectrometry cannot determine available soil test levels since they are artifacts of the extractants used. Wavelength data cannot detect something that does not exist in the soil. To repeat, both approaches are works in progress and their combined use could result in positive synergies.

The combination of these approaches, as well as the rapidly developing soil scanning by satellites, shows tremendous potential for faster and more effective ways of evaluating soil fertility.

13.4.1 Rhizobia–Legume Symbiosis

Rhizobial symbiosis with legumes takes place with bacteria that induce nodulation and infect the roots of thousands of species of the subfamilies Papilionoideae and Mimosoideae of the Leguminosae family. A third Leguminosae subfamily, the Cesalpinioideae, has only a few species that fix nitrogen. All legumes have high nitrogen content in the leaves, whether they fix nitrogen or not. Emphasis on nitrogen fixation is now all over the tropics, with many active research programs and review books such as Ladha et al. (Reference Ladha, George and Bohlool1992a), Mulungoy et al. (Reference Mulongoy, Gueye and Spencer1992) and Giller (Reference Giller2001).

Rhizobial strains are classified as promiscuous or specific. Promiscuous strains are rhizobia that enter symbiosis with many legume species, while the specific ones only do it with one or a few (Sanginga et al. Reference Sanginga, Abiadoo, Dashiell, Carsky and Okogun1996,Reference Sanginga, Dashiell, Okogun and Thottappily1997, Mpepereki et al. Reference Mpepereki, Javaheri, Davis and Giller2000). Table 13.3 lists the main rhizobial species important in the tropics and their legume hosts, showing some promiscuous species. Many of them were first reported as recently as the 1990s and many of them come from China. Bradyrhizobium is a genus created to accommodate slow-growing strains in association with soybeans as well as cowpea (Giller Reference Giller2001).

13.4.2 Cyanobacterial Symbiosis

Blue-green algae, or cyanobacteria, are bacteria that have the chlorophyll-a pigment. Producing oxygen (O₂) from photosynthesis damages nitrogenase activity, but cyanobacteria are able to separate photosynthesis from nitrogenase activity in their cells. The nitrogen-fixing symbiosis occurs in cavities in the upper leaf surface of the floating Azolla anabaena fern by the cyanobacteria Anabaena azollae and Nostoc species. (Table 13.4). The Azolla species supplies the carbon and the bacteria the nitrogen. Cyanobacteria associate with fungi forming lichens on rocks, starting the rock weathering process often. Many other cyanobacteria are free-living.

13.4.3 Frankia Symbiosis

Actinomycetes are filamentous bacteria with hyphae similar to those of fungi. The genus Frankia (Fig. 13.6) is the only one capable of forming symbiotic nitrogen-fixing nodules with gymnosperms, of which Casuarina, a tropical pine, is the most important one in the tropics (Giller Reference Giller2001), followed by alder (Alnus spp.) in the tropical highlands.

Fig. 13.6 Frankia nodules on Alnus acuminata roots in an Andisol of Costa Rica.

13.4.4 Free-Living Bacteria

Free-living bacteria in the soil solution around the plant rhizosphere include Azotobacter and Beijerinckia, which have the ability to fix nitrogen aerobically, while Derxia and Clostridium do it anaerobically (Table 13.4). Since they do not undergo symbiosis with a plant, their energy requirements must come from a rhizosphere that is rich in soluble carbon.

13.4.5 Endophytic Bacteria

These are bacteria that are capable of nitrogen fixation, found within tissues internal to the epidermis of non-legume plants (Giller Reference Giller2001). They have received much attention in the tropics since the seminal work of Döbereiner and colleagues in Brazil, in grasses and sugar cane (Döbereiner Reference Döbereiner1961, Reference Döbereiner1968, Döbereiner et al. Reference Döbereiner, Day and Dart1972, Urquiaga et al. Reference 369Urquiaga, Cruz and Boddey1992, Boddey Reference 366Boddey1995, Boddey and Döbereiner Reference Boddey and Döbereiner1995, Urquiaga Reference Urquiaga, Moniz, Furlani, Schaffert, Fageria, Rosolem and Cantanarella1997, Urquiaga et al. Reference Urquiaga, Alves, Soares and Boddey2008). Endophytic bacteria can fix nitrogen only in micro-anaerobic niches because they have no way to protect their nitrogenase enzyme from O₂ damage. Acetobacter diazotrophicus and Herbaspirilum spp. have been found in the roots and aerial tissues of sugar cane (Boddey et al. Reference Boddey and Döbereiner1995) and important pasture grass species of the genera Brachiaria and Paspalum (Döbereiner et al. Reference Döbereiner, Day and Dart1972, Boddey and Victoria Reference Boddey and Victoria1986). These authors indicated that sugar cane was bred under low levels of nitrogen fertilization in Brazil where endophytic nitrogen fixation is believed to account for about 50 percent of the nitrogen uptake of the first harvest of sugar cane.

Flooded rice also harbors many endophytic bacteria of the genera Azospirillum, Acetobacter, Azotobacter, Beijerinckia, Burkholderia, Herbaspirilum and Serratia (Döbereiner et al. Reference Döbereiner, Baldani, Reis, Fendrik, del Gallo, Vanderleyden and de Zamarocy1995, Boddey et al. Reference Boddey and Döbereiner1995, Verma et al. Reference Verma, Ladha and Tripathi2001, Tan et al. Reference Tan, Hurek, Prasad, Ladha and Reinhold-Hurek2001, Gyaneshwar et al. Reference Gyaneshwar, James, Natarajan, Reddy, Reinhold-Hurek and Ladha2001, Muñiz et al. Reference Muñiz, Rives, Castro, Sanzo, Socorro, Alves y and Urquiaga2008). In low soil organic matter (SOM) Aqualfs of Cuba, Muñiz et al. (Reference Muñiz, Rives, Castro, Sanzo, Socorro, Alves y and Urquiaga2008) showed that biological nitrogen fixation contributes 28–32 percent of the nitrogen requirements of flooded rice genotypes yielding around 4 t/ha, and that the process is related to the high population of endophytic microorganisms in soils with low SOM content (< 0.5 percent SOC). Low SOC means low nitrogen mineralization, and we know that high inorganic nitrogen contents decrease biological nitrogen fixation.

13.4.6 Inoculation

Inoculation with appropriate rhizobia began over 100 years ago, mainly focused on soybeans and tropical pasture legumes. In both cases, most cultivars require highly specific rhizobial strains since soybeans were introduced from China, and most inoculation efforts are in the United States, Brazil and Argentina. Likewise, most tropical pasture legumes originate from Latin America and require inoculation when transported to Australia (Andrew and Kamprath Reference Andrew and Kamprath1978, Wilson Reference Wilson1978, Giller Reference Giller2001).

Giller (Reference Giller2001) indicated three situations where inoculation is needed: (1) where indigenous rhizobia are less effective than selected strains for a particular legume species; (2) where rhizobia that are compatible with the legume are not present; and (3) where the population of compatible rhizobia is too small to give sufficiently rapid nodulation – less than 20–50 cells per gram of soil.

This is an area of much research and application in the tropics, as commercial inocula are routinely produced in Latin America for large-scale commercial soybean production, and also in many tropical countries for smallholder farmers in Africa. The basic method is to add pure cultures of the desired rhizobial strain to peat, which serves as the carrier and also prevents desiccation of the bacteria. Small quantities of phosphate rock, lime and molybdenum are often added. This mixture is kept in plastic bags under refrigeration. Inoculants are mixed with the legume seed prior to planting, as shown in Fig. 13.7, where a group of African scientists are watching a demonstration.

Fig. 13.7 Ready-to-mix commercial inoculum with soybean seeds (top), and scientists learning (bottom). Muhoho, western Kenya.

Rhizobia can survive in a free-living state for over 30 years, even in the absence of legumes (Giller Reference Giller2001), so the question of whether successive inoculations are necessary for subsequent grain legume crops of the same species depends on the visual evidence of healthy nodules in the current crop.

Inoculation with other nitrogen-fixing bacteria is different. Anabaena azollae is an obligate, meaning it is never separated from the host Azolla species, hence no inoculation is necessary (Ladha and Watanabe Reference Ladha and Watanabe1982). A similar mechanism may take place in the Frankia–gymnosperm symbiosis. Endophytic bacteria in cereals and grasses have no true symbiosis with the host plant (JK Ladha, personal communication, 2012).

Inoculating with free-living nitrogen-fixing bacteria is a different story and depends largely on the supply of energy as dissolved carbon in the rhizosphere of the crop, and also energy produced from rhizosphere exudation. Some positive plant growth responses take place, but these are largely attributed to bacteria that produce growth-stimulating substances such as indole acetic acid and other hormones, but not much to nitrogen fixation, which contributes perhaps 5 kg N/ha per year (Giller Reference Giller2001). The amounts of nitrogen fixed are larger in flooded anaerobic soils grown to rice, which have copious amounts of straw as a carbon source, with the flooded soil acting as a carbon sink (Roger and Ladha Reference Roger and Ladha1992).

Scientists have dreamed for decades about cereals that could fix N₂ from the atmosphere, just like legumes. Giller (Reference Giller2001) points to two research avenues: (1) to introduce the nitrogenase enzyme, so the cereal can fix nitrogen directly; and (2) to manipulate the cereal genome so it can nodulate with rhizobia. For rice, the International Rice Research Institute initiated a Global Frontier project on nitrogen fixation, with an aim to assess the potential of rice to fix its own nitrogen, similarly to legumes. Through this project, scientists have shown that rice possesses a genetic mechanism for nodulation, as in legumes, and hence the potential exists (Ladha and Reddy Reference Ladha and Reddy2003). It was, however, recognized that the project is a long-term one. Unfortunately, it was stopped about 7 years after its initiation due to lack of continuing funding, but there seems to be renewed interest now. There’s no significant progress yet, but let’s keep researching.

13.4.7 Amounts Fixed

Rhizobia

The quantities of nitrogen fixed by different crops in symbiotic relationships with rhizobia are shown in Table 13.5. It can be seen that the bigger the plant species are, and the longer they live, the larger are the amounts of nitrogen fixed. Grain legumes fix the smallest amounts, and trees the largest. Grain legumes, the earliest maturing group, fix on average around 60–100 kg N/ha per crop. Among the grain legumes, the longer-lived soybean, peanut and pigeon pea (all of which undergo symbiosis with Bradyrhizobium strains) fix higher amounts of nitrogen than the earlier maturing common bean. A rule of thumb is that, under optimum conditions, grain legumes fix 1–2 kg N/ha per day (Giller Reference Giller2001).

Table 13.5 Nitrogen-fixation estimates (above-ground) by tropical grain legumes and green manures. Selected from tables by Giller (Reference Giller2001) and J. K. Ladha (personal communication, 2012).

Species	Nitrogen fixed		Time period	Country in which the work was done
	(kg N/ha)	Nitrogen uptake from biological nitrogen fixation (%)	(days unless otherwise indicated)
Grain legumes (sole crops):
Phaseolus vulgaris (common bean)	18–36	32–47	56	Colombia
	74–91	43–52	74	Kenya
Vigna unguiculata (cowpea)	9–51	32–74	100	Brazil
	66–120	54–70	57	Nigeria
Glycine max (soybean)	85–154	70–80	110	Brazil
	149–176	69–74	70–84	Philippines
Arachis hypogea (peanut, groundnut)	68–116	54–78	110	Brazil
	30–131	59–87	-	India
Cicer arietinum (chickpea)	35–80	66–96	-	Nepal
	150–200	72–77	106–119	Thailand
Cajanus cajan (pigeon pea)	67–85	63–81	170	Australia
	13–163	42–85	120	Malawi
Pasture legumes, green manures:
Aeschynomene afraspera	105–145	68–77	56	Philippines
Arachis pintoy	1–7	68–82	12 weeks	Colombia
Calopogonium mucunoides	1–76	3–79	180	Côte d’Ivoire
	136–182	-	1 year	Western Samoa
Centrosema pubescens	3–104	48–79	180	Côte d’Ivoire
	80–280		1 year	Various
Lablab purpureus	7–70	35–76	180	Côte d’Ivoire
Macroptilium atropurpureum (siratro)	91–132	69–74	190–195	Philippines
Mucuna pruriens cv. utilis	18–213	60–83	180	Côte d’Ivoire
Pueraria phaseoloides (kudzu)	6–53	41–61	180	Côte d’Ivoire
	115	88	17 weeks	Colombia
Stylosanthes capitata	38	87	17 weeks	Colombia
Stylosanthes guianensis	141–179	73–88	16 months	Brazil
	76–102	68–79	16 months	Brazil
Stylosanthes macrocephala	68–89	74–79	16 months	Brazil
Stylosanthes scabra	22–40	52–70	16 months	Brazil
Sesbania rostrata	83–109	36–51	60	Senegal
	70–458	36–94	45–65	Philippines
Zornia glabra	61	88	17 weeks	Colombia
Trees and shrubs:
Calliandra calothyrsus	1026	24–84	27 months	Kenya
Crotalaria grahamiana	116–162	36–80	6–9 months	Kenya
Crotalaria juncea	29–142	59–86	180	Côte d’Ivoire
Gliricidia sepium	1063	56–89	27 months	Kenya
Leucaena leucocephala	224–274	56	-	Prunings
Tephrosia vogelii	12–45	37–78	184–220	Malawi
	98–124	58–73	6 months	Kenya

Most pasture legumes (grown without grazing) and green manures at 6 months of age fix nitrogen in the same range as grain legumes (around 60–120 kg N/ha), ranging from 1 kg N/ha to 458 kg N/ha, with the stem-nodulating Sesbania rostrata fixing the highest amounts in 2 months (Ladha et al. Reference Ladha, Pareek and Becker1992b). Lastly, trees and shrubs growing for 6 months to over 2 years fix much larger quantities of nitrogen, some reaching over 1000 kg N/ha in 2 years. In all cases there is large variability among and within species, making it difficult to predict the amounts of nitrogen fixed. This is because that depends on many other factors, particularly management and phosphorus status.

What is somewhat less variable is the proportion of a plant’s nitrogen uptake coming from nitrogen fixation in legumes, which ranges from 30 percent to 90 percent. This means that nitrogen-fixing legumes take 10–70 percent of their nitrogen from nitrate and ammonium ions in the soil solution. Given the high energy requirements of nitrogen fixation, legumes prefer to take the easy way out if inorganic sources are readily available (Silsbury Reference Silsbury1977). When the soil has high levels of inorganic nitrogen ions, the nitrogen-fixing mechanism can shut down. This is why grain legumes should only receive minimal nitrogen fertilization (say 5–10 kg N/ha) as a starter fertilizer. In the ustic tropics, the Birch effect may act like a starter fertilizer.

There are three principal factors affecting the quantities of nitrogen that are fixed in a cropping system as opposed to the single crop estimates in Table 13.5. Giller (Reference Giller2001) listed them as follows:

• Probably the most important one is the amount of land with legumes or other nitrogen-fixing plants, as obvious as it sounds. This is notoriously evident in Malawi, where farmers often claim they have a maize–pigeon pea intercropped system, while in fact it is only true in some corners of their fields.

• The ability of the nitrogen-fixing plants to establish symbiosis, which is often restricted by environmental or management factors (see Fig. 13.8).

• The ability of the established symbioses to fix nitrogen, which is determined by the genetic potential of the bacteria, the plant and the symbiosis itself.

Fig. 13.8 Peas (Pisum sativum) in the highlands of Ruhiira, Uganda, showing very uneven growth. In the front, stunted, chlorotic, poorly nodulated plants, in contrast with good growth and nodulation in back because weeds and crop residues are customarily removed from the center and piled along the field boundaries. Extension agent Richard Gumisiriza.

Photo and description courtesy of Ray Weil

13.4.8 Transfer of Fixed Nitrogen to Other Plants

It is commonly assumed that nitrogen fixed via biological nitrogen fixation can be used by the next crop or by a cereal intercropped with the legume, thus “increasing soil fertility.” Such generalizations fall apart with actual data, but there are still major contributions. Let’s summarize what we know.

The processes by which nitrogen from nitrogen-fixing plants can be made available to others were reviewed by Giller (Reference Giller2001). He found just three that have major importance:

• Root and nodule senescence, decomposition and mineralization;

• Mineralization of cut or senesced above-ground plant parts; and

• Consumption by grazing animals or insects and return in excreta.

Since then, a synergy between roots and mycorrhizal hyphae has shown promise in increased nitrogen fixation at the greenhouse scale (Chalk et al. Reference Chalk, Souza, Urquiaga, Alves and Boddey2006), but not at the field scale. Grain legumes are greedy, and put most of the nitrogen they take up in the seed (see Table 13.6), leaving little in the stover and roots. Whatever effects grain legumes have on soil properties and subsequent crops depend solely on the crop residues (stover) and how they are managed. Of particular interest are the promiscuous soybean varieties (Sanginga et al. Reference Sanginga, Abiadoo, Dashiell, Carsky and Okogun1996, Reference Sanginga, Dashiell, Okogun and Thottappily1997, Mpepereki et al. Reference Mpepereki, Javaheri, Davis and Giller2000) because they also produce large quantities of stover, in addition to high grain yields. Table 13.6 also shows that cowpeas, pigeon peas and peanuts can produce large quantities of stover, but apart from the promiscuous soybeans, most soybean cultivars leave very little stover after harvest, just the stems. Harvesting of peanuts by hand often involves compete removal of the plant, and the stover usually goes to feed livestock, so in this case they contribute nothing. Common beans produce 0.1–7.5 t/ha of stover, so the high-end cases represent a significant contribution.

Assuming all the stover is returned to the soil and all grain is removed, the nitrogen contribution of grain legumes to the soil (net input from nitrogen fixation) ranges from –11 to –37 kg N/ha per crop to a positive 10 to 136 kg N/ha per crop. Negative values occur when grain legumes remove more nitrogen from the soil to the grain than they fix, a fact that is not commonly realized. The positive values indicate the actual contribution to subsequent nitrogen fertility, the highest being soybean, cowpea, peanut and pigeon pea.

The recovery of stover nitrogen by the subsequent crop is generally quite low, 4–26 percent, in comparison with fertilizer nitrogen recovery (30–70 percent, see Section 13.9.3), with only one impressive value of 58 percent with mung beans. Translating this contribution into nitrogen fertilizer equivalents, grain legumes can provide from 0 to 205 kg N/ha, a very wide range. So the choice of grain legume cultivar, the rhizobial strain and management can make a huge difference, from negligible to contributions above 100 kg N/ha, which is what a good subsequent crop of maize normally needs.

When grain legumes are intercropped with a cereal there is no immediate benefit to the companion crop because there is little decomposition of the stover or roots in growing plants. The real effect (if any) is on the subsequent crop, after the grain legume has been harvested, and provided that the stover is added to the soil. The residual effects are the increases in inorganic nitrogen after the legume stover mineralizes. Figure 13.9 shows that maize yields almost doubled when it was preceded by a Bambara nut crop (Vigna subterranea, a grain legume), and also almost doubled when preceded by peanut, with the stover incorporated in the soil in both cases, in comparison with maize being preceded by a maize crop, all with no fertilizer nitrogen applied. This is the real residual effect in this subhumid environment with one crop per year. When fertilizer nitrogen was applied, the effect of the legume grain crops disappeared. In this case, it does not make sense to apply nitrogen fertilizer at all, letting the grain legume residual effect produce 6–8 t/ha of maize in Zimbabwe.

Fig. 13.9 Maize yields without nitrogen fertilizer applications increased drastically when two grain legumes preceded the current crop in Marondera, Zimbabwe, but the effect of the legumes pretty much disappeared when nitrogen fertilizers were added.

Drawn from Mukurumbira (Reference Mukurumbira1985)

Okito et al. (Reference Okito, Alves, Alves, Urquiaga and Boddey2004), in a longer-term field trial in Brazil, quantified the contribution of peanuts and mucuna using the ¹⁵N natural abundance technique, finding that peanut contributed 41 kg N/ha to the subsequent maize crop and mucuna, a green manure, 60 kg N/ha. But Okito and colleagues found that the peanut–maize rotation eventually depletes soil nitrogen reserves (just as Giller warned about some grain legumes), while the mucuna–maize sequence builds up soil nitrogen.

13.4.9 Overall Contribution of Biological Nitrogen Fixation to Food Production

The magnitude of nitrogen fixation in agricultural systems worldwide will be about 50 Tg N per year (Table 13.1). It is a small fraction in relation to the total agricultural land of the world, 4937 million hectares (Table 2.10), representing an average of 6.5 kg N/ha per year (including all land with and without nitrogen-fixers). It was not possible for me to calculate the percentage of agricultural land in the world where nitrogen fixation takes place, and thus a quantitative estimate of the contribution of biological nitrogen fixation to world or tropical food production.

The potential contribution of biological nitrogen fixation is usually based on experimental plot data, which provide a rosy picture. Giller (Reference Giller2001) indicates that legumes can fix 60 kg N/ha under experimental conditions, but I feel the reality in much of the tropics is a fraction of that because of management limitations of germplasm, rhizobial strains, phosphorus fertilization, rainfall variability and particularly low plant populations. As Norman Borlaug famously said “you can’t eat potential.” There are major exceptions, such as high-yielding soybean cultivation and tropical plantations that have an effective, well-managed, legume understorey, such as rubber and oil palm, which will be described in Chapter 19.

Nevertheless, it is well accepted that free-living bacteria add nitrogen to the SON pool, which certainly helps long-term sustainability (Bohlool et al. Reference Bohlool, Ladha, Garrity and George1992, Giller Reference Giller2001).

One country has provided sound data. Brazil stands alone as not only the tropical but probably the world leader in using biological nitrogen fixation. Based on sound science, Brazil uses very little nitrogen fertilizer (10 kg N/ha) in relation to its overall planted area. It devotes 13 million hectares to soybean production, with a national average yield of 2.4 t/ha (Alves et al. Reference Alves, Boddey and Urquiaga2003). It uses essentially no nitrogen fertilizer on soybeans. Urquiaga et al. (Reference Urquiaga, Alves, Soares and Boddey2008) estimated that 3.8 Tg N was fixed by improved strains of Bradyrhizobium and related species, worth about US $4 billion in nitrogen fertilizer equivalent every year (Fig. 13.10). The little nitrogen fertilizer used in sugar cane, with world-class yields, is presumably also due to nitrogen fixation, although whether it’s due to endophytic nitrogen fixation or another mechanism that remains to be determined.

Fig. 13.10 Alfredo Lopes, one of the leaders in the development of the Cerrado, in the first soybean crop planted in Fazenda Nova Vida, near Brasília in 1975. These Oxisols received lime, phosphorus and the rhizobial inoculum. Such practices are worth about US $4 billion per year in nitrogen fertilizer equivalents.

Photo courtesy of Stan Buol, North Carolina State University

13.5.1 The Haber–Bosch Ammonia Synthesis

It was not until the development of the Haber–Bosch industrial process of ammonia synthesis in the early 1900s that mineral nitrogen fertilizers became to be used at scale. This is the number one way of transferring N₂ into N_r in agriculture, with an estimated 100 Tg N per year in 1990. This compares with 32 Tg per year from agricultural biological nitrogen fixation (Table 13.1). Tilman et al. (Reference Tilman, Cassman, Matson, Naylor and Polasky2002) found a linear relationship between global fertilizer use and global cereal grain production from 1940 to 2000. They did not find such close relationships between phosphorus fertilizer or irrigation and global cereal grain production.

The Haber–Bosch process has been labeled “the world’s greatest fix (Leigh Reference Leigh2004).” Smil (Reference Smil2004) calls it the most important technical invention of the twentieth century. Is this hyperbole or an exaggeration? No. While electricity, cars, airplanes, medical sciences, the space program, the rule of law and the internet have done much to improve the quality of life, it is clear that the world could not support today’s population without mineral nitrogen fertilizers. Vaclav Smil (Reference Smil2004) estimated that about 40 percent of the world’s dietary protein supply in the mid 1990s came from the Haber–Bosch process, considering both the high dietary protein consumption of richer countries (100 g/person per day) and low consumption in the poorer counties of Africa, Asia and Latin America (66 g/person per day). Smil calculated that nitrogen-fertilizer-free practices, along with fisheries and grazing, would feed about 3.2 billion people today or 42 percent of the 7.5 billion population in 2017. Such fertilizer-free practices would feed only one-third of the world population of 9.5 billion people by the year 2050.

Fig. 13.11 A modern Haber–Bosch plant.

A Haber–Bosch plant is shown in Fig. 13.1. Unlike biological nitrogen fixation, which operates at normal pressure and temperatures, the Haber–Bosch process breaks the N₂ molecule by brute force, using high temperatures (up to 400 °C) and high pressures (up to 35 000 kPa or 350 atmospheres), and no O₂, which is poison to nitrogen fixation (Leigh Reference Leigh2004) just as it is to biological nitrogen fixation’s nitrogenase enzyme. The endless supply of N₂ in the air is made to react with the explosive hydrogen (H₂) gas. Natural gas (mainly methane [CH₄]) and steam (water [H₂O] gas) combine to produce H₂ at high temperatures and pressures. Methane is the most economical feedstock because it has the highest H:C ratio of all hydrocarbons. The reaction is helped by iron (magnetite) and other catalysts (Smil Reference Smil2004).

The main reactions are:

CO is then removed by the “water gas shift reaction:”

Then N₂ from the air is converted into ammonia:

Modern plants produce from 300 to 1000 tons of NH₃ per day. Ammonia gas (12–18 percent NH₃) is cooled and kept as a liquid at –33 °C, just below the boiling point of NH₃ (Smil Reference Smil2004). Finally, NH₃ is converted into actual fertilizers by reactions shown in Table 13.7.

About 80 percent of the product of Haber–Bosch process is used to make fertilizer and the remaining 20 percent to make explosives, plastics, nylon, plexiglass, polyurethane, feed supplements, glues and other synthetic products, accounting for 23 Tg N in 2005 (Galloway et al. Reference Galloway, Townsend, Erisman, Bekunda, Cai, Freney, Martinelli, Seitzinger and Sutton2008). Fertilizer use has driven a global change in the nitrogen cycle. China and India account for 80 percent of the total increase in nitrogen fertilizer use since 2000.

Today, synthetic nitrogen fertilizers supply approximately 45 percent of the total nitrogen input for global food production (Ladha et al. Reference Ladha, Reddy, Padre and van Kessel2011). The rest is supplied mostly by soil nitrogen mineralization, but also by biological nitrogen fixation and atmospheric deposition.

Fritz Haber and Carl Bosch were awarded the Nobel Prize in chemistry, Haber in 1918, and Bosch in 1931. More detailed descriptions of the process and the fascinating history behind it are found in books by Leigh (Reference Leigh2004) and Smil (Reference Smil2004).

13.5.2 Nitrogen Fertilizers

Out of the Haber–Bosch process, several fertilizers were produced (Table 13.7), starting with ammonium sulfate, by Imperial Chemical Industries in 1923. The last fertilizer released for large-scale application was sulfur-coated urea, by the US National Fertilizer Development Center in Muscle Shoals, Alabama, in the 1960s. It is disheartening to realize that no new nitrogen fertilizer for agricultural use has been developed and marketed in the last 50 years, and probably longer since sulfur-coated urea is not used at large scale because of its cost. Many “boutique” nitrogen fertilizers are on the market, some with unsubstantiated claims, but these are mainly devoted to golf courses, backyard gardens and nurseries. A significant modification is the compression of urea into super-granules or briquettes, which seems effective in flooded rice culture. The development of the fertilizer industry is still very much mid-twentieth century, and is yet to take advantage of breakthroughs such as nanotechnology. Hopefully, I will stand corrected in the near future. Currently, there is more emphasis on blends that incorporate sulfur, zinc and iron with the major elements nitrogen, phosphorus and potassium.

The amounts of fertilizers applied per hectare of agricultural land varies widely between countries. Some countries apply too much (Netherlands, Vietnam, Japan, UK, China, France), some reasonable amounts (Brazil, United States, India, South Africa, Cuba) and others apply too little, mostly African countries (Fig. 13.12).

Fig. 13.12 National fertilizer applications are very uneven.

Source: FAOSTAT 2002 data

The Africa Fertilizer Summit, held in Abuja, Nigeria, in 2006, committed African countries to increase their annual fertilizer use from 8 kg NPK/ha to 50 kg NPK/ha by 2015. Malawi and Zambia have almost attained that target, Malawi reaching 48 kg/ha NPK and Zambia 43 kg/ha NPK (Peter Crauford, personal communication, 2017).

Nitrogen fertilizer management practices depend on many factors. Some are specific to the fertilizer. Others depend on germplasm and soil and climate conditions, as well as on the agronomic management. Wetland rice is a special case and will be discussed in Chapter 17. Let’s start with the real game changer, germplasm.

13.5.3 Green Revolution Germplasm

The main impact of the Green Revolution of the 1960s–1980s was to drastically change the plant type: new varieties of rice and wheat put more nitrogen into the grain when nitrogen fertilizers were applied, thus making grain to straw ratios of above 1 instead of ~0.5 (Chapter 2). Most traditional varieties responded to nitrogen fertilization by producing more leaf, lodging (falling over on the ground) and losing yield, while the new ones increased grain yields drastically by remaining short-statured and erect. Figure 13.13 shows the results of a typical trial.

Fig. 13.13 One of the early Asian Green Revolution trials showing the different responses of the short-statured Sonora 64 wheat variety from Mexico versus a local one in Pantnagar, India (Sharma et al. Reference Sharma, Misra, Wright and Krantz1970).

The main nitrogen fertilizer management practices are to determine the right source, the right rate of application, the right placement and the right timing. This well-known paradigm has been made popular as the 4Rs (Roberts Reference Roberts2009).

13.5.4 Nitrogen Fertilizer Sources

Much research has been conducted to compare urea, ammonium sulfate and other nitrogen sources on corn, rice, wheat and sorghum in the tropics. For most crops, the overwhelming evidence indicates that there are no differences between urea and ammonium sulfate or other ammonium fertilizers. In instances where ammonium sulfate was superior to urea, the effect was due to sulfur deficiency or volatilization losses of surface-applied urea in high-pH soils. Where urea was superior, differences were due to the acidifying effect of large, long-term applications of ammonium sulfate in already acid soils. The bulk of the evidence indicates few differences in ammonium sources when properly applied, including DAP. Nitrate sources are usually inferior to ammonium sources under conditions favoring leaching or denitrification.

13.5.5 Rates of Application

Choosing the correct rate of application is one of the two most important fertilizer management practices, the other being timing of application. The most common point of departure for determining the rate (kg N/ha) is the existence of nitrogen response curves, which are available in most tropical agricultural regions. The zero fertilizer rate (the control) gives an idea of how much nitrogen a cultivar can obtain from the remaining mineral nitrogen from SOM mineralization, manure applications, legumes in rotation and atmospheric deposition.

There are numerous ways to determine rates, ranging from the empirical, using a response curve like the graphs of Fig. 13.14 combined with local experience, to sophisticated mass balance models and geographical information systems and remote sensing algorithms, with equipment that can be tractor-mounted (Meisinger et al. Reference Meisinger, Schepers, Raun, Schepers and Raun2008). The latter is great for precision agriculture, but for most small- and medium-sized tropical farms I would use the Linear Response and Plateau model (LRP), reproduced as Figure 13.14. Such a rate must also guarantee the maintenance of SOM using the fertilizer nitrogen that is immobilized by bacteria and eventually converted to SOM.

Fig. 13.14 Comparison between the LRP and the quadratic models, using the same field data from potatoes in Peru. The recommended rate is much lower using LRP (Waugh et al. Reference Waugh, Cate and Nelson1973).

Repeated from Fig. 12.8

13.5.6 Fertilizer Placement

Nitrogen fertilizer placement is determined by seed spacing. Plant population itself affects the shape of the response curve of improved germplasm. At low levels of applied nitrogen, a population of 30 000 plants per hectare is optimum for maize, whereas at higher nitrogen rates with the best hybrids, populations nowadays are 100 000 plants per hectare, and above. Figure 13.15 shows the relationship between nitrogen rate and population on maize yields, in an old experiment in Guadalajara, Mexico. To achieve high yields, both high populations and high nitrogen rates are needed. Sub-optimal plant populations are unfortunately very common in smallholder tropical farms without mechanization, and when a poor quality of seed is used. This is partly because the spacing between rows is not constant, or because too many seeds are placed in the same planting hole and then have to be thinned, wasting a valuable resource. In very poor areas, maize and other seeds are broadcast over the soil surface, guaranteeing very low yields. In such cases, the first task of an extension worker is to demonstrate how to plant in rows at the right population. Individual maize seeds are planted with close spacing, about 10–25 cm within the row and 75 cm between rows. This is called the Sasakawa method in parts of Africa, and is an effective way to increase maize yields in comparison with the common 90 × 90-cm spacing (Nziguheba et al. Reference Nziguheba, Palm, Berhe, Denning, Dicko, Diouf, Diru, Flor, Frimpong, Harawa, Kaya, Manumbu, McArthur, Mutuo, Ndiaye, Niang, Nkhoma, Nyadzi, Sachs, Sullivan, Teklu, Tobe and Sanchez2010). Other crops have their particular spacing requirements.

Fig. 13.15 Increasing population and nitrogen applications increase maize yields in Mexico. The optimum population for each nitrogen rate is indicated by an arrow (Laird and Lizárraga Reference Laird and Lizárraga1959).

Seasonal rainfall also exerts a marked influence on maize responses to nitrogen. Figure 13.16 illustrates the lower responses obtained when either excess moisture or drought occurred in a series of maize experiments in Mexico. Also, as discussed before in Chapter 6, short-term droughts during the rainy season (veranicos) can cause crop failures if they happen during the reproductive stage of crops. Such short-term droughts may be predictable in places where topography and soil texture varies, as in southern Malawi.

Fig. 13.16 Excess moisture or drought decreases maize response to nitrogen fertilizers in Mexico (Rockefeller Foundation 1963–64).

13.5.7 Timing of Applications

The timing of nitrogen applications for most crops consists of a basal application at planting and a topdressing application just before the crop’s grand period of growth. Pre-planting applications are usually determined by the need for phosphorus fertilizers, which is critical at early growth stages. But unlike phosphorus, nitrogen can be rapidly lost by leaching while the seed germinates and establishes a root system. Incorporating the fertilizer close to the seed, but with soil in between to avoid fertilizer burn (due to high salinity), is the best way. Due to the danger of leaching, normally only one-third of the nitrogen rate is applied at planting and the rest at topdressing time, when the plant is at about knee height. It is applied close to the maize rows and normally before the second weeding so the topdressing is also incorporated into the soil. It is seldom profitable, however, to have more than two split applications.

Fox et al. (Reference Fox, Talleyrand and Bouldin1974) summarized nitrogen responses to maize and sorghum in Oxisols and Ultisols in udic areas of Puerto Rico, and obtained nitrogen fertilizer recoveries of 51 percent with one side-dressed application, but only 33 percent when the same rate was incorporated into the soil surface at planting. In semiarid Zimbabwe, Piha (Reference Piha1993) found split applications better for sorghum, perhaps because of the unpredictability of rainfall events.

If there is likely to be a drought, farmers may minimize their losses by lowering the rate or avoiding the topdressing application. Also, the topdressing application can be fine-tuned with nitrate sensors or chlorophyll meters (Meisinger et al. Reference Meisinger, Schepers, Raun, Schepers and Raun2008).

Fertilizer applications should not be done after the crop has flowered because there will be little grain yield response. This unfortunately happens in many parts of Africa where subsidized fertilizer arrives too late but farmers still apply it after maize has tasseled. Avoiding this is another simple extension message in areas with no prior experience with mineral fertilizers.

13.6.1 SON Pools and Nitrogen Mineralization

The pools that account for most nitrogen mineralization are the metabolic nitrogen pool and the active SON pool (which includes microbial biomass), while the structural nitrogen pool and the slow and passive SON pools (Fig. 13.4) have limited roles in nitrogen mineralization

Organic resources or organic inputs, as explained in Chapter 11, are partitioned into metabolic or structural nitrogen pools (Fig. 13.4) according to their nitrogen and lignin contents. In the case of manures, where lignins have largely been digested by ruminant animals, the only indicator of quality (in this case mineralization versus immobilization) is nitrogen content, where the critical level seems to be 1.25 percent N (Murwira and Mukurumbira Reference Murwira and Mukurumbira1984), half of the critical level for plant materials (2.5 percent N). C:N ratios do not work because the carbon content of manures varies widely as explained in Chapter 11.

In addition to input quality as defined by the metabolic and structural nitrogen pools, the dominant factor affecting nitrogen mineralization rates in the tropics is topsoil moisture content. Calder (Reference Calder1957) and Semb and Robinson (Reference Semb and Robinson1969) found that mineralization may take place at moisture tensions higher than –1500 kPa, the “wilting point.” The water content present at high tensions in clayey, variable-charge soils, although unavailable to plants, seems to be available to mineralizing microorganisms. Both these pioneering studies took place in East Africa, probably on soils high in iron and aluminum oxides or allophane.

Nitrogen mineralization rates also depend on temperature. Low temperatures can be limiting in the tropical highlands, but high temperatures in tropical lowland areas seldom limit nitrogen mineralization per se. On the contrary, Mosier et al. (Reference Mosier, Wassmann, Verchot, King and Palm2004) indicated that the production of trace nitrogen gases – a product of initial mineralization and subsequently of water-filled pores – increases within the range of 20–40 °C in the tropics.

Between these extremes of moisture and temperature content, most tropical soils undergo several periods of alternate wetting and drying. Nitrogen mineralization is likely to be faster under alternate wetting and drying than under “optimum” moisture conditions. But these dynamics are short term, which results in a decline of the total SON (Ladha et al. Reference Ladha, Reddy, Padre and van Kessel2011).

13.6.2 Immobilization

Immobilization, the opposite process to mineralization, is where a high diversity of bacteria and other microorganisms convert NO₂^– or NH₄⁺ into organic forms of SON. Immobilization occurs primarily in the active nitrogen pool, initially by the microbial biomass fraction of that pool. Manipulating the balance between nitrogen mineralization and immobilization can be a key management practice. The NO₂^– or NH₄⁺ ions from fertilizers are immobilized and then subsequently mineralized as described later.

The balance between mineralization and immobilization is illustrated in Fig. 13.17, with 2.5 percent N being the critical level.

Fig. 13.17 Whether nitrogen is mineralized or immobilized from plant materials is determined by the nitrogen concentration of the materials and modified by high lignin or polyphenol contents. The regression equation is for all materials. Black squares or circles represent materials with > 2.5 percent N. White squares or circles are those with < 2.5 percent N. Squares represent materials with < 15 percent lignin and < 4 percent polyphenol. Circles represent materials with > 15 percent lignin or > 4 percent polyphenol.

Data are from 11 studies put together by Palm et al. (Reference Palm, Gachengo, Delve, Cadisch and Giller2001). Permission granted by Agriculture, Ecosystems and Environment

13.6.3 Influence of the Core Properties

The three core soil properties also play important roles. Sandy topsoils have low levels of SON, mostly in the active pool where SON can be mineralized rapidly, releasing inorganic nitrogen to plants. Clayey soils have a higher SON content and generally have more labile carbon substrates for mineralization to proceed. Mineralogy also plays a role, particularly allophane, whereby Bornemisza and Pineda (Reference Bornemisza and Pineda1969) found that nitrogen mineralization is inversely proportional to its allophane content in Andisols. Mineralization at very high soil moisture tensions has been recorded in variable-charge oxidic and allophanic soils, as mentioned before, but it is not known whether the same happens in permanent-charge soils.

13.7.1 Urea Hydrolysis

Urea is the most commonly used inorganic nitrogen source in the tropics. Its popularity is partly due to its high nitrogen content (46 percent), low unit cost and its availability in the world market. When applied to a moist soil, urea is hydrolyzed into ammonium carbonate [(NH₄)₂CO₃] by the enzyme urease in the following way:

$\mathrm{CO}{\left({\mathrm{NH}}_2\right)}_2+{\mathrm{H}}_2\mathrm{O}\stackrel{\text{Urease}}{\to }{\left({\mathrm{NH}}_4\right)}_2{\mathrm{CO}}_3$(13.12)

(NH₄)₂CO₃ dissociates into the NH₄⁺ and CO₃^– ions in the presence of water. Before hydrolysis, urea is as mobile as NO₃^– and may be leached to down below the root zone with heavy rainfall if soil structure permits. Tamini and Kanehiro (Reference Tamini and Kanehiro1962) showed that urea hydrolysis proceeds at about the same speed in the tropics as in the temperate region, and is complete within 1 to 4 days. In flooded soils, DeLaune and Patrick (Reference Delaune and Patrick1970) found that the rate of urea hydrolysis is similar to that in well-drained soils.

13.7.2 Ammonia Volatilization

At soil pH values higher than 7, the NH₄⁺ ions can be converted to NH₃ (ammonia gas) and lost to the atmosphere.

Volatilization losses of NH₃ were first recognized in the tropics by Jewitt (Reference Jewitt1942), working with high-pH Vertisols in the Sudan. Broadcast nitrogen applications to the soil surface are very common in the tropics; therefore, volatilization losses can be of practical importance in high-pH soils, particularly when high nitrogen rates of urea are applied (Fig. 13.18). Shankaracharya and Mehta (Reference Shankaracharya and Mehta1971), working with a loamy sand of pH 7.1 in Gujarat, India, measured field volatilization losses of 4 percent when 28 kg N/ha was applied as urea to the soil surface. When the rate was increased to 277 kg N/ha, volatilization losses increased to 44 percent. Such high rates are common in areas where high-yielding cereals are planted. They also become the effective rate when fertilizers are point-placed around the planting hole in smallholder farming systems.

Fig. 13.18 A wasteful practice: broadcasting urea on Inceptisols of pH 8.3 in Koraro, Ethiopia.

Urea volatilization losses can be reduced if the material is placed below the soil surface before hydrolysis. This can be accomplished by incorporation, deep placement, or simply by moving the freshly applied urea down with irrigation water or rainfall. Table 13.8 shows the reduction in volatilization losses when irrigation followed a surface application. In the presence of moisture, volatilization of ammonia does not take place. This table also indicates that urea volatilization losses are essentially eliminated by incorporating the material to about 5 cm depth. The practical implication is that urea should be incorporated into the soil if it is to be applied to a dry calcareous soil.

13.7.3 Ammonium Fixation

NH₄⁺ ions can be trapped between sheets of 2:1 layer-silicate minerals when they expand and contract with wetting and drying. This process is limited to clayey topsoils with smectitic, illitic or vermiculitic mineralogy, all of which have permanent charge, and does not happen in the more common kaolinitic, siliceous or oxidic mineralogies. Nevertheless, it is important in Vertisols of the tropics, although little work has been done on these. A good review with a temperate-region focus is that of Kissel et al. (Reference Kissel, Cabrera, Paramisavam, Schepers and Raun2008).

13.7.4 Immobilization of Fertilizer NH₄⁺ and NO₃^–

Contrary to common agronomic knowledge, many, if not most, of the NH₄⁺ and NO₃^– ions released by fertilizers are immobilized by microbes before they are taken up by crop roots (Myrold and Bottomley Reference Myrold, Bottomley, Schepers and Raun2008, Mulvaney et al. Reference Mulvaney, Khan and Ellsworth2009). Microbes are at near-nitrogen-starvation levels, and their large numbers in a rhizosphere that is rich in labile carbon impedes much of the direct uptake of fertilizer nitrogen by plant roots (Vitousek et al. Reference Vitousek, Haettenschweiler, Olander and Allison2002). Much of this immobilized nitrogen is mineralized quickly because many microbes have very high turnover rates (hours, days), and then it is quickly taken up by the crop roots. Vanlauwe et al. (Reference Vanlauwe, Diels, Aihou, Iwuafor, Lyasse, Sanginga, Merckx, Vanlauwe, Diels, Sanginga and Merckx2002) cited two laboratory experiments in which 50 ppm of fertilizer nitrogen was immobilized during the first 25 days after application, and mineralized 40–50 days afterwards. Also, much of a high nitrogen fertilizer application rate (168 kg N/ha) to Mollisols in Kansas was immobilized, released and taken up by maize plants 50–120 days afterwards (Brady and Weil Reference Brady and Weil2008). Such processes may improve the synchrony between nutrient supply and plant demand and reduce nitrate leaching and N₂O losses. It is one reason why the recovery of fertilizer nitrogen by crops is seldom higher than 50–70 percent. More on this issue later.

The above paragraph is full of the phrase “much of,” showing the need for quantification. How much fertilizer nitrogen is immobilized? For how long? How much is released under different soil conditions in tropical agricultural systems? What are the turnover times? How much of the fertilizer nitrogen is building SOM via this mechanism?

Microbial immobilization of fertilizer nitrogen may be a good thing for basal nitrogen applications, which are in danger of loss because the crop seedlings do not have much of a root system at their early stages of growth. However, not having much of a root system means little rhizosphere and probably less free-living bacteria to immobilize the basal fertilizer applied. I did not find any data to confirm this, one way or the other.

13.8.1 Nitrate Accumulation during the Dry Season

The accumulation of NO₃^– ions in the topsoil during the dry season can be explained by the upward movement of NO₃^– with the evaporation stream and the existence of nitrification at soil moisture tensions of –1500 kPa to –8000 kPa (Semb and Robinson Reference Semb and Robinson1969). Although the topsoil may be drier than those tensions indicate, the subsoil may have moisture to support mineralization. Since most of the water movement during the dry season is upward, NO₃^– ions previously present or recently mineralized in the subsoil may move up and accumulate in the topsoil. Wetselaar (Reference Wetselaar1961), working with an Alfisol from northern Australia, found evidence of dramatic nitrate build-ups in the top 5 cm. He explained that NO₃^– ions were mineralized in the subsoil where adequate moisture existed during the dry season, and accumulated just below a soil surface crust where capillary conductivity is broken.

Wild (Reference Wild1972aReference Wildb) followed the nitrate content of a profile in northern Nigeria for 2 years. His results, shown in Fig. 13.19, indicate an upward movement of NO₃^– ions during the dry season. This figure suggests that nitrate was leached into the subsoil during the previous rainy season and moved upward during the dry season. Hardy’s original data (Table 13.9) from a study in Trinidad shows nitrate levels increase in the topsoil (A horizon) during the dry season.

13.8.2 The Birch Effect: Nitrogen Flushes at the Onset of the Rainy Season

Within a few days after the first heavy rains, dramatic increases in inorganic nitrogen take place in the subhumid and semiarid tropics. In the field, such increases may range from 23 kg N/ha to 121 kg N/ha within 10 days after the onset of the rains (Semb and Robinson Reference Semb and Robinson1969). The sharpness of the peaks is directly proportional to the duration and intensity of the preceding dry period. These sharp increases are accompanied by similarly sharp decreases, caused by rapid leaching or plant uptake at the start of the rainy season (Fig. 13.20). Note that these changes are less pronounced in the “short rains” in the bimodal climate of the forest zone of Ghana, shown in this figure. Semb and Robinson’s work in East Africa provided clear evidence of NO₃^– ions moving into the subsoil after such flushes.

Fig. 13.20 Seasonal pattern of NO₃^–-N fluctuation in the top 10 cm of a cultivated Alfisol in Ghana with a bimodal udic soil moisture regime.

Adapted from Greenland (Reference Greenland1958)

These short-term peaks were first described by Hardy in 1946 in Trinidad. Subsequent work in Africa by Birch and other workers substantiated their existence in a wide range of soil conditions (Birch Reference Birch1958, Reference Birch1960aReference Birchb, Reference Birch1964). The flushes are called the Birch effect because of the popularity of Birch’s articles (Chapters 6 and 7).

Birch (Reference Birch1958, Reference Birch1960aReference Birchb, Reference Birch1964) found that drying promotes faster carbon than nitrogen mineralization, thus reducing the C:N ratios. He also found that the critical C:N ratio above which mineralization stops is higher under alternate wetting and drying than under constant soil moisture. For example, an organic material containing 1.5 percent N was mineralized under alternate wetting and drying but was immobilized under constant moisture. Much of the microbial population is dead by the end of a strong > 6 month dry season. Active microbial populations build up rapidly when the rains start, using the dead microbial biomass as a substrate.

This is a very important process that has been largely ignored in the tropics outside of Africa and in the world in general. Gutschick (Reference Gutschick1981) noted that many crops were domesticated from weedy plants, and are adapted to exploit flushes of nitrogen. But for them to take advantage of the short-lived Birch effect, plants need to have a sufficiently developed root system to start with or to be capable of developing one within the first month or so into the rainy season. Ratoon crops of sugar cane are especially suited to take advantage of the Birch effect because they have most of the live root system of the previously cut crop already in the ground – as are natural systems, permanent pastures and tree crops.

This is why I think the Birch effect could explain part of the dilemma of changing soil nitrogen availability during what is supposed to be endophytic nitrogen fixation. Data are needed, particularly by South American scientists. We need to get a handle on how important, or not, this Birch effect is, where sugar cane is grown in ustic soil moisture regimes in Brazil.

The Birch effect is very important in the ustic soil moisture regime of the subhumid and semiarid tropics. It is probably of low importance in the udic, humid tropics due to the lack of a sufficiently prolonged dry season to kill and desiccate a sufficient microbial biomass so it can serve as a substrate for new microbes to rapidly grow at the onset of the rains. It is also unlikely to be important in the arid tropics because insufficient microbial biomass may accumulate during the previous very short rainy season. It will be unimportant in aquic soil moisture regimes, and probably less important in the highland tropics (isomesic and isofrigid soil temperature regimes) because of the constantly low soil temperatures.

The flushes of nitrogen mineralization at the start of the growing season are unlikely to take place in temperate-region soils, characterized by cold, wet springs preceded by winters where much denitrification takes place. In summary, the Birch effect is an important phenomenon in the nitrogen cycle only in ustic soil moisture regimes with isohyperthermic, hyperthermic, isothermic and thermic soil moisture regimes, which cover much of the tropics and some of the subtropics. This flush probably explains why new grass growth appears so green at the beginning of the rainy season in areas that have undergone no burning during the dry season. Where burning does take place, ash deposition, particularly in phosphorus-deficient soils, is a major reason for the greening. The relative effects of the flush of nitrogen mineralization and ash deposition need to be determined.

The Birch effect has been found by Fierer and Schimel (Reference Fierer and Schimel2002), Miller et al. (Reference Miller, Schimel, Meixner, Sickman and Melack2005) and Placella et al. (Reference Placella, Brodieb and Firestone2012) in temperate xeric soil moisture regimes of Mediterranean climates using up-to-date methodologies. The cause is the same, a long, dry summer followed by a wet winter. It is interesting that tropical soil scientists were over 60 years ahead in reporting this important phenomenon. Part of the reason is that many young scientists nowadays only pay attention to the literature that is available on the internet, and much of this old stuff is not digitized.

13.8.3 Downward Nitrate Movement During the Rainy Season

As the rainy season progresses, the inorganic nitrogen supply is reduced by plant uptake and leaching. Hardy (Reference Hardy1946) showed the effect of crop uptake in decreasing NO₃^– content in Table 13.9. Leaching depends on a series of soil factors.

Going back to Fig. 13.19, Wild (Reference Wild1972aReference Wildb) found that peak NO₃^– concentration in the profile gradually moves down as the rainy season progresses. At the beginning of the rainy season (June 3) most of the nitrates were concentrated in the top 15 cm. Three weeks later, after the flush, leaching started and the peak concentration moved down to between 15 cm and 30 cm. By August 11, well into the rainy season, the peak concentration was at 30–45 cm, and 2 weeks later it was at 45–60 cm.

In Wild’s experiment, NO₃^– ions moved at the approximate rate of 0.5 mm per mm of rain. The topsoil of this Alfisol is probably loamy, with clay increasing with depth. This movement is considerably slower than in sandy soils. Terry and McCants (Reference Terry and McCants1970), for example, found that nitrate moved at a rate of 1–5 mm per mm of rain in sandy Ultisols of North Carolina.

In well-aggregated Oxisols (not the loamy Alfisol of Figure 13.19), both downward movement of rainwater through large channels and slow lateral diffusion through the microaggregates occurs. Consequently, although many of the well-aggregated Oxisols act like sands in terms of water retention, their leaching losses of mineralized soil nitrogen are bound to be slower than those in sands because the NO₃^– ions mineralized inside the aggregates have to move through the micropores and out into the macropores before they can be susceptible to leaching. If the subsoil has some degree of anion exchange capacity, as shown in Chapter 8, nitrate movement will be reduced.

Wild (Reference Wild1972aReference Wildb) also found that the total nitrate content in the top 120 cm of the Fig. 13.19 Alfisol did not change significantly during the year, except at the end of the rainy season, when larger leaching losses took place because of a drop in the valley water table. In his 2 years of observations, therefore, most of the inorganic nitrogen never left the soil profile.

Fluctuations are also minimal if actively growing crops absorb nitrate quickly. Hardy’s data in Table 13.9 show that a soil under pasture had lower nitrate content than under maize because of year-round nitrogen uptake by pastures.

It is important to realize that the inorganic nitrogen fluctuation pattern just described is quite different from the one commonly found in much of the temperate region. The growing season in the humid (udic) temperate region begins with a cold, wet soil, low in inorganic nitrogen because of the slow mineralization rates during fall and winter as well as high denitrification losses. As temperatures warm up gradually in the spring, there is a correspondingly gradual increase in organic nitrogen mineralization. Inorganic nitrogen increases slowly, without flushes, because of the high soil moisture status. During the summer, mineralization rates can be very high, and the inorganic nitrogen content will depend on crop uptake, leaching and denitrification. Most of the nitrogen management practices used in the temperate region have been developed to fit such a pattern. Nitrogen management practices in the tropics should give more consideration to the Birch effect pattern.

13.9.1 Classifying Organic Fertilizers

Ten classes of organic fertilizers are described in Table 13.10. The first four are leafy plant biomass, classified according to Palm et al. (Reference Palm, Gachengo, Delve, Cadisch and Giller2001), and the remaining are animal manures and composts that I am adding, based on Chapter 11, the work from Zimbabwe and the Sahel on manures, and my limited experience with composts. The high percentage of phosphorus criterion is designed not to add too much phosphorus to a crop, which can happen with poultry manures. The polyphenol requirement of Palm et al. (Reference Palm, Gachengo, Delve, Cadisch and Giller2001) was dropped because experience indicated it is of little value.

Class 1 plant biomass inputs are almost equivalent to nitrogen fertilizers in terms of mineral nitrogen release. Low-quality (class 4) plant biomass inputs immobilize mineral nitrogen but do not mineralize it fast enough to be synchronized with crop demand (Palm et al. Reference Palm, Myers, Nandwa, Buresh, Sanchez and Calhoun1997). Classes 2 and 6 may immobilize nitrogen from fertilizers and mineralize it slowly enough to increase crop yields, thus improving synchrony and reducing leaching – according to Vanlauwe et al. (Reference Vanlauwe, Diels, Aihou, Iwuafor, Lyasse, Sanginga, Merckx, Vanlauwe, Diels, Sanginga and Merckx2002) and Chivenge et al. (Reference Chivenge, Vanlauwe, Gentile, Wangechi, Mugendi, van Kessel and Six2009) for plant biomass inputs; and Pieri (Reference Pieri1992) and Mugwira and Murwira (Reference Mugwira, Murwira, Waddington, Murwira, Kumwenda, Hikwa and Tagwira1998) for manures. They may also decrease NO and N₂O emissions.

All organic fertilizers contain the other 13 essential nutrients besides nitrogen, and their application often prevents the development of other nutrient deficiencies. Combining does not necessarily mean mixing. Organic fertilizers could be applied before planting followed by topdressing with mineral nitrogen fertilizers.

Chivenge et al. (Reference Chivenge, Vanlauwe and Six2011) conducted a meta-analysis of 104 experiments that compared plant biomass inputs (classes 1 to 4) with mineral nitrogen fertilizers, mixed or applied separately, using rainfed maize in 12 countries of tropical Africa. They measured crop responses; agronomic efficiency (AE_N), in kilograms of additional grain per kilogram of nitrogen applied, either as mineral or organic fertilizers; changes in SOC content; and residual yield effects, in experiments lasting at least 5 years. Control yields ranged from 0 t/ha to 7 t/ha, covering a wide range of initial soil nitrogen fertility conditions. Their main results were as follows:

• Maize yield responses were higher with the combined organic and nitrogen fertilizer applications (114 percent response over the control), while nitrogen fertilizers produced an 84 percent response, and class 1 to 4 of organics alone produced a 68 percent response. The authors attribute this to a temporary immobilization of nitrogen from fertilizers by the organic inputs, which may have improved synchrony.

• Classes 1 and 2 of organics increased yields by about 90 percent (similar to nitrogen fertilizers); class 3 organics increased yields by 50 percent, while the lowest quality class 4 barely increased yields by about 7 percent. Crop yield responses definitely were higher with higher organic input quality, confirming the Palm et al. (Reference Palm, Gachengo, Delve, Cadisch and Giller2001) model, although the separation between classes 1 and 2 was minimal.

• Class 4 organics alone depressed yields in sandy soils but, when combined with nitrogen fertilizers, yield responses were much larger in sandy than in clayey soils. Chivenge et al. (Reference Chivenge, Vanlauwe and Six2011) believe this combination, applied every year, could increase sustainability in sandy soils. This augurs well for the Sahel, where most of the soils are sandy and most biomass that can be added is of low quality. With class 4, low-quality organics, temporary immobilization of the nitrogen from fertilizer and subsequent release may proceed too slowly to benefit the current crop but will be beneficial in the long term.

• The trend in agronomic use efficiency was different in the mineral–organic combinations compared to mineral fertilizers alone. The AE_N value for the combination was only 14 ± 2 kg grain/kg N, far below that of nitrogen fertilizers alone (23 ± 4 kg grain/kg N), and almost the same as with organics alone, 13 ± 3 kg grain/kg N. So, nitrogen fertilizers in this meta-analysis were almost twice as efficient as classes 1 to 4 of organics, but the combination provided higher yield responses, which could be due to the addition of other nutrients in the organics.

• Soil texture had additional effects. Clayey soils were doubly superior in agronomic use efficiency (AE_N = 21 percent) to loamy (AE_N = 8 percent) and sandy soils (AE_N = 10 percent). Most of the soils that had low control yields, but also greater yield responses, were sandy soils.

• The lower the fertilizer nitrogen rate the larger the beneficial effect of combining it with organic biomass resources.

• High-quality organic fertilizers produce longer residual effects. This contradicts the common belief that the lower-quality organic fertilizers may have longer residual effects because of their slow rate of decomposition.

The above results indicate that combining plant biomass resources with nitrogen fertilizers is beneficial, and more so if the organics are of high quality. The combination will also increase SON, which will improve nutrient storage. The latter is particularly the case in sandy soils since it is most likely that the increase in SON will be in the active SON pool where nitrogen will be rapidly mineralized, improving soil fertility but not total SON content.

In a previous study, Chivenge et al. (Reference Chivenge, Vanlauwe, Gentile, Wangechi, Mugendi, van Kessel and Six2009) compared classes 1 to 5 of organics, which included high-quality cattle manure in two soils near Embu, Kenya, over a 10-year period. Class 1 (Tithonia diversifolia) and class 5 (high-quality manure) produced the highest maize yields at both sites. Classes 2, 3 and 4 produced progressively lower yields. Class 5 cattle manure was as effective as the class 1 leafy organics.

As mentioned before, the two newly introduced animal manure classes in Table 13.10 are based primarily on the work of Zimbabwean and Sahelian soil scientists. In Zimbabwe, Mugwira and Murwira (Reference Mugwira, Murwira, Waddington, Murwira, Kumwenda, Hikwa and Tagwira1998) reviewed cattle manure research since 1913 in sandy soils of communal areas and with large cattle populations, which I classify as class 6 organic fertilizers. These authors concluded that “these studies indicated that manure alone generally produced low yields and needed supplementing with inorganic fertilizers, particularly N, for optimum yields.” They further stated that “these benefits derive primarily from the ability of cattle manure to supply other nutrients. Manure application to granitic sands [sandy Alfisols or Ultisols] has been shown to overcome or prevent deficiencies of nutrients including S, Mg, Zn and B, and to enhance soil available N, P and K.”

Summarizing 30 years of research in Francophone West Africa, in mostly long-term field experiments on sandy Alfisols of the Sahel, and using cattle manure similar to the Zimbabwe situation (class 6), Pieri (Reference Pieri1992) concluded the following:

• Where animal traction is used, manure becomes a resource. The best results were obtained when ~ 5 t/ha per year of manure were combined with annual applications of NPK (nitrogen, phosphorus and potassium) fertilizers, eliminating yield declines for many years. Manure applications often eliminated the need for liming to counteract secondary soil acidity in these sandy soils. Apparently, the calcium and magnesium content in the manure had some liming effect, although no nutrient composition data were provided.

• In very sandy, unbuffered soils (> 90 percent sand), even a 10 t/ha manure application could not prevent the loss of SOC with cultivation, which proceeds at a very high rate – 6 percent per year in the hot, semiarid climate of Bambey, Senegal – suggesting that the active SOC level could not be maintained.

• On better textured soils (> 10 percent silt and clay), as in Saria, Burkina Faso, the decomposition rate is reduced to 2 percent with mineral fertilizer plus manure, and decomposition rates of SOC become lower by adding a grain legume in rotation.

• The best results were invariably obtained with the combination of cattle manure and NPK fertilizers. Generally, crop yields increased from values as low as 0.4 t/ha to 2–3 t/ha.

13.9.2 Importance of Knowing Nutrient Composition

Comparisons of cattle manure versus mineral fertilizers in the tropics have been carried out principally in India and Africa (Pieri Reference Pieri1992). The main problem is that the nutrient contents of the manure must be known for comparisons to be valid, but that seldom happens. The large difference in nitrogen lost during manure storage (Chapter 11) further complicates the issue if manure sampling is done shortly after excretion.

Several long-term comparisons in Africa did take into account the nutrient composition of the animal manure employed (Djokoto and Stephens Reference Djokoto and Stephens1961, Stephens Reference Stephens1969, Heathcote Reference Heathcote1970). They show that crop responses to animal manures can be explained in terms of their nutrient composition, particularly the potassium and phosphorus content. For example, Stephens (Reference Stephens1969) found, in 18 long-term experiments in Uganda, that manure applications were superior to mineral fertilizers. Manure provided 75 kg K/ha annually, whereas the fertilizer used supplied only 25 kg K/ha to potassium-deficient soils. Obviously, the difference was due to nutrient composition.

Nevertheless, it is possible to generalize that the combined use of organic and mineral nutrient inputs often produces higher long-term yields than in either full organic or full mineral applications (Greenland Reference Greenland, Leigh and Johnston1994, Swift et al. Reference Swift, Seward, Frost, Qureshi, Muchena, Leigh and Johnston1994, Palm and Swift Reference Palm and Swift2002, Sanginga and Woomer Reference Sanginga and Woomer2009). Livestock manures (classes 5 and 6) often buffer the soil against acidification by continued use of acid-forming nitrogen fertilizers (Pieri Reference Pieri1992). Class 6 manures often do not add enough nutrients, including nitrogen (Murwira and Mukurumbira Reference Murwira and Mukurumbira1984).

The ratio of mineral to organic nutrient inputs usually increases with agricultural intensification. Therefore, identification of the threshold amounts of organic inputs necessary to maintain non-nitrogen functions should be considered in the integrated management of soil fertility (Palm and Swift Reference Palm and Swift2002).

13.9.3 Dynamics of Mineral and Organic Fertilizers with Time

When I was a graduate student, attending International Rice Research Institute (IRRI) seminars in the Philippines, I repeatedly heard Felix Ponnamperuma, Akira Tanaka, Richard Bradfield and SK DeDatta saying that about 50 percent of the nitrogen used by rice came from fertilizers. I wondered, where does the other 50 percent come from? But I was too overwhelmed by these world-class scientists to dare ask them such a seemingly stupid question. It has taken decades for me to answer that question. Advice to present day students: Ask.

Because of the immobilization of fertilizer nitrogen and leaching and denitrification losses, the recovery of applied fertilizers by the crop to which it is applied is generally about 30–50 percent, which is known as percentage of fertilizer recovery, the most commonly used fertilizer efficiency metric. So what happens to the rest?

Dourado Neto et al. (Reference Dourado Neto, Powlson, Abu Bakar, Bacchi, Basanta, thi Cong, Keerthisinghe, Ismaili, Rahman, Reichardt, Safwar, Sangakkara, Timm, Wang, Zagal and van Kessel2010), supported by the International Atomic Energy Agency, conducted a coordinated set of field trials in five tropical countries (Bangladesh, Brazil, Malaysia, Sri Lanka and Vietnam) plus four others (China, Chile, Egypt and Morocco), using the local cropping systems and both ¹⁵N-labeled fertilizer and ¹⁵N-labeled crop residues that were generated from the first cropping season. The results are summarized in Table 13.11.

In spite of the wide array of soils, cropping systems and climates involved, only an average of 30 percent (ranging from 7 percent to 58 percent in the individual country trials – data not shown) of the ¹⁵N fertilizer was recovered by the first crop. The rest of the ¹⁵N fertilizer was found immobilized in the SON (35 percent) and lost by leaching or gaseous losses (36 percent). So, we have a first quantification of the amount of nitrogen fertilizer immobilization.

The ¹⁵N-labeled crop residues contributed only 7 percent of the original crop uptake; 65 percent was immobilized into SON, and 29 percent was lost as leaching and gaseous losses – a similar amount to losses from the ¹⁵N fertilizer.

With subsequent crops, the crop recovery from the original ¹⁵N fertilizer increased by about 4.8 percent with the second crop, with no significant residual effects. Afterwards, according to the literature review by Dourado Neto et al. (Reference Dourado Neto, Powlson, Abu Bakar, Bacchi, Basanta, thi Cong, Keerthisinghe, Ismaili, Rahman, Reichardt, Safwar, Sangakkara, Timm, Wang, Zagal and van Kessel2010), the residual effect of nitrogen fertilizers usually stops at the second crop. They conclusively proved that.

The amount of ¹⁵N fertilizer immobilized by SON decreased gradually from 35 to 16 percent at the fifth crop, indicating that SON was being mineralized and used by the plant. But nitrogen losses also increased, suggesting that some of the SOM mineralized was lost, and not taken up by the crop (Table 13.11).

The treatments receiving crop residues fared worse than those with ¹⁵N fertilizer. Recovery of ¹⁵N crop residues was only 7 percent by the first crop but increased with time to 11–15 percent, indicating the slow mineralization of nitrogen in crop residues. However, 39 percent of the ¹⁵N nitrogen in crop residue was still present in the SOM after 5 years, that is twice the amount from ¹⁵N fertilizer nitrogen, which presumably may enter the slow SON pool. The overall losses were similar in both ¹⁵N sources after 5 years.

Using unlabeled materials (not shown in Table 13.11), Dourado Neto et al. (Reference Dourado Neto, Powlson, Abu Bakar, Bacchi, Basanta, thi Cong, Keerthisinghe, Ismaili, Rahman, Reichardt, Safwar, Sangakkara, Timm, Wang, Zagal and van Kessel2010) concluded that 79 percent of all the nitrogen uptake by the five crops originated from SOM, either nitrogen immobilized before this study or during it.

This excellent study showed that the residual effect of fertilizer nitrogen recovery by the crop is short-lived, and that most of the nitrogen taken up by crops came from the mineralization of SON. The authors concluded that the main residual effect of fertilizer nitrogen is to replenish SON pools. This study provides excellent support of the role of nitrogen inputs for maintaining SOM levels as a major component of sustainability.

13.10.1 Can Mineral Fertilizers Increase SOC?

The answer in most cases is yes, provided that carbon is also added, for instance, by incorporating crop residues or adding organic inputs of differing qualities.

The SOC equilibrium content is a function of the following simple formulas, copied here from Chapter 11, and developed by Greenland and Nye (Reference Greenland and Nye1959). It shows these relationships:

where C is the SOC stock in equilibrium (t/ha or kg C/m²); b is the annual amount of fresh organic inputs added to the soil (t C/ha per year); m is the fraction of fresh organic annual input that enters the SOC pools, i.e. the fraction of total fresh organic matter that is not respired in the initial decay of litter (fraction per year); a is the annual addition of SOC (t C/ha per year); and k is the annual decomposition rate of SOC (fraction per year). Nitrogen fertilization can increase the value of b, change the value of m and decrease the value of k.

Increasing the Annual Amount of Fresh Organic Inputs (b)

When crop yields increase in response to mineral nitrogen fertilization, they add more above-ground crop residues as well as roots to the soil. Unfortunately, in many tropical farming systems, crop residues are removed and used for animal feed, cooking fuel and even fencing. Crop residues are often piled up and burned in rice cropping systems, increasing soil variability. Since 1 ton of additional grain generally produces another 1–2 tons of above-ground crop residues, the return of above-ground crop residues to the soil is an important contribution. But if crop residues are removed, fertilizer applications will have to rely on root decomposition to increase SOC. Roots at harvest time constitute 11–22 percent of the total dry mass production, which includes 5–21 percent of the nutrients that are accumulated by grain crops (Sanchez et al. Reference Sanchez, Palm, Szott, Cuevas, Lal, Coleman, Oades and Uehara1989). This may not be sufficient b input to increase SOC, and in some cases may cause SOC depletion.

Long-term research on Inceptisols and Alfisols in semiarid India shows that the increase in SOC by mineral fertilization was a function of the carbon inputs added by the different rotations and soils, which in turn depended on the proper fertilizer combination. Figure 13.21 shows that insufficient organic inputs resulted in SOC depletion, but adequate inputs increased the SOC. The carbon input required to increase SOC varied with soil texture, being about 6 t C/ha per year in a sandy loam in Barrackpore, about 3 t C/ha per year in the sandy clay loam in Ranchi and about 1.5 t C/ha per year in the clayey soil in Akola. This shows once again that it is harder to increase SOC in sandy soils because of the greater accessibility by decomposing microbes, due to larger pores. The correlation between carbon inputs added and SOC changes shown in Fig. 13.21 shows the importance of the b factor in the Greenland and Nye equation and its relationship with texture, one of the core soil properties.

Fig. 13.21 The total above- and below-ground organic carbon inputs (b value) determined whether nitrogen fertilization added or decreased SOC in (a) the top 30 cm of a sandy loam Inceptisol for 29 years, (b) a sandy clay loam Alfisol for 30 years, and (c) a clayey Vertisol for 14 years, in long-term trials in semiarid India (Manna et al. Reference Manna, Swarup, Wanjari, Ravankar, Mishra, Saha, Singh, Sahi and Sarap2005).

Reproduced with permission from Field Crops Research

Manna et al. (Reference Manna, Swarup, Wanjari, Ravankar, Mishra, Saha, Singh, Sahi and Sarap2005) also found large declines in crop yields and topsoil SOC when fertilization included only nitrogen and phosphorus, but not when NPK fertilizer or NPK fertilizer plus manure were added, indicating the need for potassium and probably micronutrients contained in the manure. It is not reasonable to expect nitrogen fertilizers to increase SOC if the soil is deficient in another element, unless this is corrected.

Decreasing the SOC Decomposition Rate (k)

Mineral fertilizer use can also reduce the value of k by keeping a lower soil temperature under a complete soil cover, provided by a vigorously growing crop. A long-term study of tea soils in India, illustrated in Fig. 13.22, shows that higher SON equilibrium levels were obtained when tea was fertilized annually with a 120–25–36 kg/ha rate of nitrogen, phosphorus and potassium. The unfertilized plots kept losing organic nitrogen at an annual rate of 1.5 percent, with correspondingly lower total nitrogen content as equilibrium was approached (Gokhale Reference Gokhale1959). In contrast, the fertilized plots had a decomposition rate of 1.0 percent. Tea is a perennial crop with year-round ground cover, so the effect will be less marked with annual crops that provide a ground cover for only a few months per year.

Fig. 13.22 Mineral fertilization increases the equilibrium SOM content (expressed as SON) in a long-term experiment in a tea plantation in Assam, India (Gokhale Reference Gokhale1959). Annual fertilizer rates are equivalent to 120–25–36 kg/ha of nitrogen, phosphorus and potassium.

Reproduced by permission from Field Crops Research

13.10.2 Denitrification, NH₃, NO and N₂O Emissions

The leaky pipe concept illustrated in Fig. 13.3 shows that various gaseous losses can occur upon the application of mineral and organic fertilizers. One is NH₃ volatilization, which occurs largely with organic manures (see Chapter 11), but also from surface-applied urea in high-pH soils, as described above. NH₃ is not a greenhouse gas, but it binds with particulate matter in aerosols, which in fact has the effect of decreasing atmospheric warming.

Denitrification is the microbial reduction of NO₃^– to N₂, which includes the production of N₂O and NO as intermediate compounds. The determinants of denitrification are the availability of NO₃^– (the electron acceptor), labile carbon (the electron donor that provides energy to the microbes), and O_2, the competing electron acceptor (Del Grosso et al. Reference Del Grosso, Parton, Mosier, Ojima, Kulmala and Phongpan2000). Biological denitrification is performed by a highly biodiverse, adaptable and widely distributed group of aerobic, heterotrophic bacteria that are also facultative anaerobes (Mosier et al. Reference Mosier, Wassmann, Verchot, King and Palm2004, Coyne Reference Coyne, Schepers and Raun2008, Baggs Reference Baggs2011). The reaction progresses as follows:

Full denitrification is about the only way for N_r to be converted back to its original inert form (N₂) by agriculture. Denitrification from the world’s terrestrial ecosystems is expected to reach 98 Tg N by 2050, at a time that the Haber–Bosch process is expected to reach 165 Tg N (Table 13.1). About 80 percent of the nitrogen fixed by the Haber–Bosch process goes into nitrogen fertilizer production (132 Tg N). Therefore, terrestrial denitrification returns back to the atmosphere 72 percent of the N₂ that was taken out to produce these fertilizers, bringing an additional consideration to the global nitrogen cycle. Natural systems, particularly wetlands, are obviously responsible for a large proportion of terrestrial denitrification and have nothing to do with fertilized agriculture. The nitrogen cycle starts and ends with N₂.

Figure 13.23 shows the relative importance of the emissions of the three gases at different levels of water-filled pore space, a good indication of oxidation–reduction status of the soil. The importance of nitrification increases as pores are better aerated, while the importance of denitrification increases as pores get more waterlogged. Denitrification of inorganic nitrogen can happen simultaneously with nitrification in nearby microsites, where one is under anaerobic conditions and an adjacent microsite is well-aerated. Often, the anaerobic microsites are at the center of microaggregates. Both N₂O and NO can be emitted as intermediate products of both nitrification and denitrification. The end product of denitrification (N₂) does not have an environmental effect, although it represents a loss of nitrates from the soil.

Fig. 13.23 As soil pores get filled with water the distribution of NO, N₂O and N₂ gas emissions varies.

Adapted from Davidson (Reference Davidson, Rogers and Whitman1991)

Field capacity (–10 kPa of soil moisture tension in most tropical soils) often occurs at about 60 percent of the water-filled pore space. Davidson et al. (Reference Davidson, Keller, Erickson, Verchot and Veldkamp2000) indicate that, at this tension, soil micropores are water-filled, which permits microbial activity without water stress, while soil macropores are air-filled, which permits relatively good aeration of the bulk of the soil, providing an equal production of NO and N₂O.

13.10.3 Leaching to Groundwater

The magnitude of leaching losses depends on texture and hydraulic conductivity, the stage of crop growth, the nitrogen application rates and anion exchange capacity, all of which determine the amount of NO₃^– that can be leached below the soil during high-rainfall periods. Temporary immobilization of applied nitrogen also helps to retard leaching. Incorporating urea with low-quality (class 3) maize stover has delayed leaching losses in a lysimeter experiment. Vanlauwe et al. (Reference Vanlauwe, Diels, Aihou, Iwuafor, Lyasse, Sanginga, Merckx, Vanlauwe, Diels, Sanginga and Merckx2002) attributed this to immobilization by the maize stover, but when this organic input was applied as mulch on the soil surface before urea application, no retardation of leaching happened.

The main point (again) is to avoid excessive nitrogen application rates by using the LRP model to determine the appropriate application rate. This is the case regardless of whether the nitrogen comes from organic or mineral fertilizers because no differences in leaching losses have been found between organic or conventional farms, aiming at a similar crop yield level (Cassman et al. Reference Cassman, Dobermann, Walters and Yang2003). The leaching losses of NO₃^–, NH₄⁺ and DON can cause eutrophication if they reach estuaries in large amounts. About 15–30 percent of nitrogen from NO₃^– leached from farmland is denitrified in estuaries; the rest goes to the sea (Matson et al. Reference Matson, Naylor and Ortiz-Monasterio1998; Pamela Matson, personal communication, 2013).

Looking at inorganic nitrogen leaching at the watershed scale instead of at the field scale makes eminent sense, and some work has been done in tropical rainforest watersheds (Groffman et al. Reference Groffman, McDowell, Myers and Merriam2001, McDowell et al. Reference McDowell, Bowden and Asbury1992). The use of vegetative filter strips, such as a riparian forest to take up NO₃^– or NH₄⁺ ions, or a wetland to denitrify them, may “mop up” excess nitrogen in the landscape before entering water bodies.

13.11.1 Nitrogen Use Efficiency

The simplest evaluation of fertilizer nitrogen efficiency is nitrogen use efficiency (NUE), commonly referred to as partial factor productivity (Ladha et al. Reference Ladha, Pathak, Krupnik, Six and van Kessel2005). NUE is the ratio of yield over fertilizer applied:

$\mathrm{NUE}=\frac{\mathrm{Y}\left(\mathrm{kg}/\mathrm{ha}\text{of economic yield}\right)}{\mathrm{N}\left(\mathrm{kg}\mathrm{N}/\mathrm{ha}\right)}$(13.15)

NUE is an empirical value where the denominator N includes the nitrogen released by SOM mineralization, biological nitrogen fixation and atmospheric deposition, as well as nitrogen from fertilizer, but usually only the latter is considered.

NUE values can be misleading because the ratio of yield to nitrogen does not tell us about the actual quantities of nitrogen and yield we are dealing with. It may imply to some that we are using fertilizers less efficiently, but that is not always true. NUE values are highest at low rates of applied nitrogen. Pre-Green Revolution world cereal NUE values averaged 0.75 in 1960 (when cereal yields and nitrogen rates were both low), and decreased to about 0.22 in 1998 (Tilman et al. Reference Tilman, Cassman, Matson, Naylor and Polasky2002). Nitrogen response curves are curves with decreasing yield increases per unit of fertilizer added until they flatten out (see the right-hand graph of Fig. 13.14). They tell us there is “more bang for the buck” in the more linear portion of the response curve, which is another reason why I prefer the LRP model (the left-hand graph).

NUE values in the linear part of the LRP model are similar but not the same because the control with no fertilizer already produced a yield which represents the nitrogen input of SOM mineralization, biological nitrogen fixation and atmospheric deposition. Nevertheless, if the potato farmer chooses to apply only 30 kg N/ha she (as most tropical farmers are) will get a yield of 12 t/ha with an NUE of 0.40 tons of potatoes per kilogram of nitrogen fertilizer added (see Fig. 13.14). If she chooses to apply the recommended rate by the LRP model (75 kg N/ha) she will get 18.5 t/ha of potatoes with an NUE of 0.25. If, overriding my recommendation, she chooses the point in the quadratic model where marginal revenue equals marginal cost, the yield will be 19 t/ha and the nitrogen rate 160 kg N/ha, giving her an NUE of 0.19. Obviously, I would apply the 75 kg N/ha rate even though the NUE is not the highest. Therefore, NUE is not recommended as a useful metric.

Cassman et al. (Reference Cassman, Dobermann, Walters and Yang2003) disaggregated NUE into three more useful parameters: agronomic efficiency, fertilizer recovery and physiological efficiency.

13.11.2 Agronomic Efficiency

Agronomic efficiency (AE_N) differs from NUE because it deals with yield response produced by the nitrogen fertilizer only, while NUE includes the total grain produced regardless of the origin of the nitrogen. Therefore:

${\mathrm{AE}}_{\mathrm{N}}=\frac{\mathrm{Yf}-\mathrm{Yc}\left(\mathrm{kg}/\mathrm{ha}\text{of economic yield}\right)}{\mathrm{F}\left(\mathrm{kg}\mathrm{N}/\mathrm{ha}\right)}$(13.16)

where Yf is the yield with fertilizer, Yc the yield of the control plot and F is the fertilizer nitrogen applied.

The terms NUE and AE_N are frequently used synonymously in the literature. Also, the definitions of NUE and AE_N varied in two major review articles (Cassman et al. Reference Cassman, Dobermann, Walters and Yang2003, Ladha et al. Reference Ladha, Pathak, Krupnik, Six and van Kessel2005). NUE is often used without an actual definition and, as mentioned before, it can be misleading.

The AE_N value is also related to value in terms of cost:benefit ratios, an important economic indicator of fertilizer profitability. In economic terms, AE_N is the partial factor productivity of nitrogen fertilization.

AE_N values for cereals globally average 37 kg grain/kg N (Cassman et al. Reference Cassman, Dobermann, Walters and Yang2003). However, in Malawi, which uses one subsidized blanket fertilizer rate recommendation for the whole country, an estimate of 12 kg grain/kg N was obtained (Denning et al. Reference Denning, Kabambe, Sanchez, Malik, Flor, Harawa, Nkhoma, Zamba, Banda, Magombo, Keating, Wangila and Sachs2009), and Chisanga (Reference Chisanga2008) reported a value of 14 kg grain/kg N. These AE_N values are not purely due to nitrogen in this case because the Malawi subsidy included phosphorus and sulfur in the basal fertilizer plus improved maize seed.

Higher AE_N values occur on the linear part of the response curve, while values are lower when nitrogen rates are excessively high, for instance in recently cultivated soils with high SOM mineralization where there is no nitrogen response (AE_N = 0), or when high nitrogen rates are applied as organic inputs, which illogically is not counted as fertilizer in most agronomic efficiency studies.

Chivenge et al. (Reference Chivenge, Vanlauwe and Six2011) calculated AE_N values more broadly in a meta-analysis comparing organic and mineral fertilizer inputs, so the terms Yf and F in the above equation could be the nitrogen fertilizer applied alone, the organic nitrogen input applied alone, or the total nitrogen applied by the combination of both. All values were within the range normally found in Africa. The mean AE_N for organic nitrogen alone was 13 ± 3 kg/kg N and the combined values 14 ± 2 kg/kg N, both much below that of nitrogen fertilizers alone, 23 ± 4 kg/kg N as previously indicated. So, nitrogen fertilizers in this meta-analysis were almost twice as efficient as organic nitrogen fertilizers. There were, however, large differences in AE_N values between organic quality classes and soil texture, shown in Table 13.14.

In maize-growing areas in the United States, the AE_N value has increased from 42 kg/kg N in 1980 to 57 kg/kg N in 2000, which, according to the Cassman et al. (Reference Cassman, Dobermann, Walters and Yang2003) review, was due to improved stress tolerance of new cultivars, higher plant densities, conservation tillage, improved genetically modified germplasm and better nitrogen management. They also cite that rice AE_N values increased from 57 kg/kg N to 75 kg/kg N between 1961 and 1985 in Japan. These increases are easy to understand and represent the trend of more grain for less fertilizer. They represent a different picture, with basically the same US data, than that produced from NUE values, as reviewed by Tilman et al. (Reference Tilman, Cassman, Matson, Naylor and Polasky2002) and others.

Can something like this be achieved in the tropics? Certainly, this is no problem in highly mechanized, modern tropical agriculture with precision farming, which mimics much of what the United States does. But what about the smallholder farmer?

Vanlauwe et al. (Reference Vanlauwe, Kihara, Chivenge, Pypers, Coe and Six2011) did a meta-analysis of 90 smallholder farm experiments with maize in Africa and obtained the AE_N ranges shown in Table 13.15. Fertilizer nitrogen in researcher-led trials was slightly more efficient than farmer-led trials. When hybrids were incorporated into the researcher-led trials, AE_N values increased further, but additional organic inputs did not have much effect. Standard deviations were all high. Upper-quartile AE_N values with hybrids and full integrated soil fertility management (ISFM) were within the range of the temperate literature, indicating that, with good management, efficiencies can be comparable to the temperate region.

Vanlauwe et al. (Reference Vanlauwe, Kihara, Chivenge, Pypers, Coe and Six2011) also observed the following:

• Hybrid maize (AE_N = 26) was superior to open-pollinated varieties (AE_N = 17).

• Mixing fertilizer and two types of organic resources (classes 2 and 5) substantially increased AE_N values.

• Infields (those close to the homestead) had values (AE_N = 31) almost double those of outfields (AE_N = 17). Infields receive more organic waste and ash and act like responsive soils, while outfields resemble the less responsive soils.

Another significant development has taken place across tropical Asia and China using a technique called site-specific management in rice. Dobermann et al. (Reference Dobermann, Wit, Dawe, Abdulrachman, Gines, Nagarajan, Satawathananont, Son, Tan, Wang, Chien, Thoa, Phung, Stalin, Muthukrishnan, Ravi, Babu, Chatuporn, Sookthongsa, Sun, Fu, Simbahan and Adviento2002) were able to increase AE_N values from 11 kg/kg N to 16 kg/kg N, using a lower fertilizer rate, in 179 sites in eight key irrigated rice domains. This involved a large extension effort in the Philippines, China, India, Thailand, Indonesia and Vietnam. It is interesting to note that these field AE_N values are lower than those obtained by farmer-managed maize in Africa (shown in Tables 13.14 and 13.15), and a fraction of what Japan obtains with rice.

Pixley and Banziger (2001, cited in Vanlauwe et al. Reference Vanlauwe, Kihara, Chivenge, Pypers, Coe and Six2011) found that replanting maize seed resulted in a yield loss of 32 percent for hybrids, but only 5 percent for open-pollinated varieties (OPVs). The advantages of hybrids are well understood by smallholder farmers. In the Malawi fertilizer and seed subsidy program, farmers were given a choice between 2 kg of hybrid seed or 3 kg of improved OPVs. About 76 percent of them chose the hybrid seeds even though they were aware that they should not replant them (Denning et al. Reference Denning, Kabambe, Sanchez, Malik, Flor, Harawa, Nkhoma, Zamba, Banda, Magombo, Keating, Wangila and Sachs2009).

What is becoming clear is that nitrogen fertilizer has no chance of being efficient unless hybrids or improved varieties are used, and in many instances organic inputs need to be added. With such management, AE_N values in smallholder tropical situations can approximate those obtained in large-scale mechanized farming in North America and Europe or, who knows, even higher.

13.11.3 Fertilizer Recovery

The most popular way to express the efficiency of nitrogen fertilization is using the fertilizer recovery (RE_N), expressed as a percentage. By knowing the nitrogen uptake at a certain rate and the uptake without added nitrogen, the percentage recovery can be calculated as follows:

$\mathrm{REN}=\frac{\mathrm{Uf}-\mathrm{Uc}}{\mathrm{F}}\times 100$(13.17)

where Uf is the nitrogen uptake of the whole above-ground plant biomass with the fertilizer rate F, and Uc is the nitrogen uptake of the whole above-ground plant in the control plot.

RE_N values are usually determined from research-station data, with or without the use of ¹⁵N. Without isotopes, it is often termed the “N-difference method” or “apparent fertilizer recovery,” and it is the most commonly reported method. Recovery values are about 7–11 percent higher with the N-difference method than with the ¹⁵N isotope dilution method, which is obviously more accurate (Ladha et al. Reference Ladha, Pathak, Krupnik, Six and van Kessel2005). Fertilizer recovery is sometimes calculated by the nitrogen uptake in the grain, which does not make sense because the nitrogen in other plant parts is needed to produce that grain.

RE_N values in rice, maize and wheat range from 20 percent to 90 percent for the first crop, with 1–7 percent recovery in subsequent crops (Ladha et al. Reference Ladha, Pathak, Krupnik, Six and van Kessel2005). This lack of a significant residual effect is shown also with ¹⁵N labeling in the seminal study by Dourado Neto et al. (Reference Dourado Neto, Powlson, Abu Bakar, Bacchi, Basanta, thi Cong, Keerthisinghe, Ismaili, Rahman, Reichardt, Safwar, Sangakkara, Timm, Wang, Zagal and van Kessel2010), presented in Table 13.11.

The range in RE_N values for cereals in developed countries quoted by Ladha et al. is 50–55 percent, while in developing countries the estimate is 20–30 percent for rainfed cereals and 30–40 percent for irrigated cereals. More than 40 years ago, Jones (Reference Jones1973) obtained a recovery of 70 percent for maize in an Alfisol of Samaru, northern Nigeria, under rainfed conditions and no leaching. So, it can be done in the tropics.

But is the unrecovered fertilizer (100 – RE_N) lost? Obviously not, since much of it is immobilized into SOM. Only a portion of the applied nitrogen fertilizer is really lost by the two main mechanisms we know, which are leaching to groundwater and gas emissions of NO, N₂O via denitrification to the atmosphere. A third possible mechanism is runoff and erosion from the soil surface after a broadcast nitrogen application.

The unrecovered nitrogen or “missing nitrogen,” therefore, includes both storage and loss mechanisms. Groffman (Reference Groffman, Schepers and Raun2008) raises the question of whether or not nitrogen sequestered in SOM could reach nitrogen saturation just like sequestered carbon can reach carbon saturation levels (Chapter 11). Groffman also pointed out that relatively little is quantitatively known about SOM storage; given that the C:N ratio of slow SOM is pretty stable, you can’t be sequestering more nitrogen if you are not sequestering more carbon.

The Dourado Neto et al. (Reference Dourado Neto, Powlson, Abu Bakar, Bacchi, Basanta, thi Cong, Keerthisinghe, Ismaili, Rahman, Reichardt, Safwar, Sangakkara, Timm, Wang, Zagal and van Kessel2010) data shown in Table 13.11 indicate that 42 percent of added ¹⁵N fertilizer was stored in SOM during the first crop. Part of this newly immobilized SOM was mineralized during the five subsequent crops, with that stored in SOM going down to 18 percent. The percent immobilized from the ¹⁵N crop residues was higher, 62 percent, decreasing to 40 percent in five years.

I believe this positive effect should be considered in the efficiency of fertilizer nitrogen utilization, and I propose a new term:

where RS_N is the percent of fertilizer nitrogen immobilized in SOM, RE_N is the same percent recovery of fertilizer N in the plant, and RES_N is the percent recovery in the crop plus that stored in the soil (this needs a better name!). Then the losses L_N will be calculated as the difference:

If both leaching (L_L) and gaseous losses (L_NO, L_N2O and L_N2) are measured or estimated at the same site, then

13.11.4 Physiological Efficiency

The third parameter, physiological efficiency (PE_N), reflects the crop cultivar’s internal efficiency and capacity to produce high grain:straw ratios, as in the Green Revolution varieties. It may also reflect a larger or deeper root system that captures more mineral nitrogen in the soil or that better tolerates adverse factors such as drought, acidity or low available phosphorus. Breeding research is underway on these aspects together with increasing yield potential. This may result in the next big jump in yields since the Green Revolution. The formula, with all parameters defined earlier, is:

${\mathrm{PE}}_{\mathrm{N}}=\frac{\mathrm{Yf}-\mathrm{Yc}}{\mathrm{Uf}-\mathrm{Uc}}$(13.21)

PE_N values range from 29 kg to 53 kg of grain increase per kg N taken up (Ladha et al. Reference Ladha, Pathak, Krupnik, Six and van Kessel2005).

Jain and Kumar (Reference Jain and Kumar2011) edited a book reviewing molecular approaches to improve nitrogen use efficiency, such as post-translational regulation of nitrate reductase, nitrate sensing and signaling, and other aspects related to PE_N. This is frontier stuff and may pay off well in the future.

14.1.1 The Phosphorus Cycle at the Geologic Time Scale

A closed cycle (Fig. 14.2) exists only at a geologic time scale (10⁷–10⁸ years). Phosphorus constitutes 0.1 percent of the Earth’s crust; 95 percent of this phosphorus is apatite (calcium phosphate), its three main forms being fluorapatite [Ca₁₀(PO₄)₆F₂], hydroxyapatite [Ca₁₀(PO₄)₆(OH)₂], and chlorapatite [Ca₁₀(PO₄)₆Cl₂]. Marine sediments containing 4 × 10⁶ Pg (billion tons) of phosphorus occupy vast coastal areas, for example from the coast of Florida to the Grand Banks of Canada in the Atlantic Ocean (22 Pg P), off the coast of southwest Africa and other continental shelves. Estimates of the US coastal resources at different depths by Stewart et al. (Reference Stewart, Hammond, van Kauwenbergh, Sims and Sharpley2005) add up to 727 Pg P.

Fig. 14.2 The global phosphorus cycle in geologic time. Boxes are stocks in Tg (million tons) or Pg (billion tons) of phosphorus (P) or phosphate rock (PR), while fluxes (in red) are in Tg P or Tg PR per year.

Calculated from data by Smil (Reference Smil2000), Stewart et al. (Reference Stewart, Hammond, van Kauwenbergh, Sims and Sharpley2005), Pierzynski et al. (Reference Pierzynski, McDowell, Sims, Sims and Sharpley2005), MacDonald et al. (Reference MacDonald, Bennett, Potter and Ramankutty2011) and other sources. Phosphate rock is calculated as 13.5 percent P

Tectonic uplift in geologic time raises apatite-rich phosphate rock to land and continental shelves at an annual rate of 111–185 million tons (Tg) of phosphate rock per year. Interestingly, this is within the range of the world’s annual extraction rate of phosphate rock, currently estimated at 140 Tg of phosphate rock per year (Fig. 14.2). Strictly speaking, phosphorus is a renewable resource at the geologic time scale, something I have not read in the literature, because tectonic uplift inputs are more or less balanced by phosphate rock mining extraction. This may be of little practical value because of the unknown location of these new uplifted deposits (presumably in the coastal continental shelves), their accessibility and cost of extraction, and whether or not sea level rise caused by global warming would adversely affect phosphate rock extraction.

Another, albeit small, input comes from the excreta and bones deposited on arid tropical islands where sea birds choose to roost. The product is called biogenic phosphate rock, or guano. Biogenic phosphate rocks are apatites of very high solubility and can be used for direct application to the soil, as the Incas did for centuries from the Peruvian guano islands. Some tropical Pacific and Indian ocean islands covered by guano are sinking into the sea as phosphate rock is harvested. Most notable is the island state of Nauru. The current global reserves of biogenic phosphate rock are unknown, but a relevant one, the Minjingu phosphate rock deposit of northern Tanzania, has an estimated reserve of 8–10 Tg of phosphate rock (van Kauwenbergh Reference van Kauwenbergh2010). This is a small deposit, compared to sedimentary deposits.

Approximately 112 Tg of phosphate rock are currently converted into phosphorus fertilizers every year. An additional 28 Tg of phosphate rock are annually converted into detergents, soft drinks, livestock salt licks and other industrial uses of phosphoric acid, adding up to 140 Tg of phosphate rock per year. About 1.5 Tg of phosphate rock per year are directly applied to the soil, including much of the biogenic phosphate rock.

The main process used for converting phosphate rock into soluble phosphorus fertilizers, the wet phosphoric acid process, is very polluting. The process leaves a byproduct called phosphogypsum, some of which often ends up in surface waters, causing eutrophication.

Phosphorus applied to agricultural soils from fertilizers was 14.2 Tg P in 2000 and the manure applied in that year contained 9.6 Tg P (MacDonald et al. Reference MacDonald, Bennett, Potter and Ramankutty2011). Phosphorus fertilizer use increased to 18 Tg P per year by 2007 (Sattari et al. Reference Sattari, Bouwman, Giller and van Ittersum2012).

Soil phoshorus stocks are considerable (~ 50 Pg P in the top 50 cm) and constitutes the largest phosphorus stock in terrestrial systems, of which over 90 percent is inorganic. This is roughly an average stock in the world’s soils of 3570 kg P/ha of which about 400 kg P/ha is organic phosphorus (Smil Reference Smil2000). The world’s croplands (arable and permanent crops), total ~1.5 billion hectares in 2013 (Table 2.10); they have a stock of 5–6 Pg P or about 10–12 percent of the total global soil stock. Terrestrial plant stocks are relatively small, about 0.6 Pg P.

Unlike other element cycles – as mentioned above – there is no major atmospheric component of the phosphorus cycle. However, particulate phosphorus is thrown to the air by wind erosion and biomass burning, but returns to the soil, usually at another location, and can be deposited in the oceans as well. These fluxes are small (1–2 Tg P per year).

The two main loss pathways from the soil are harvest removals and runoff and erosion. Large parts of harvest removals turn into sewage. The fluxes of phosphorus in runoff and erosion and sewage are the main loss pathways from terrestrial to aquatic environments. These two pathways carry 30–35 Tg P per year to freshwater and eventually the oceans, where they accumulate as sediments, and together with dead marine biomass a flux of 20–30 Tg P per year are buried in the continental shelves of oceans, adding to the immense stock of 4 × 10⁶ Pg P. Some of this stock eventually becomes phosphate rock. In geologic time, it is uplifted tectonically to the Earth’s surface, closing the phosphorus cycle.

14.1.2 The Phosphorus Cycle at a Human Time Scale

There is no closed phosphorus cycle at the human time scale (10²–10³ years), because what we have is one source (phosphate rock) and one sink (marine sediments) without a feedback loop to make it a cycle (Fig. 14.3). Unlike nitrogen and potassium, much of the phosphorus that is added to the soil tends to remain there, often building up large residual stocks. Except in very sandy soils there usually is no leaching of the phosphate anions, with the exception of movement just below the topsoil by tillage or other disturbances (see for example Beck and Sanchez Reference Beck and Sanchez1996, Garcia-Montiel et al. Reference Garcia-Montiel, Neill, Melillo, Thomas, Steudler and Cerri2000). Agronomically, the largest agricultural area in the tropics where phosphorus leaching is an important factor is the northern Sahel, because of the predominance of coarse sandy soils (Chapter 4). If the subsoil is clayey (having an argillic horizon), like many sandy Alfisols in that region, the phosphorus leached can be sorbed by the subsoils, which often have reddish colors indicative of iron oxides, giving them phosphorus sorption capacity.

Fig. 14.3 Soil phosphorus stocks and fluxes in the world’s cropland (crops plus cultivated forages) at about the year 2000. The main flux processes are color-coded as shown below the graph. Doubled-pointed arrows indicate opposite processes are happening. Single arrows indicate only one flux process. Stocks are also color-coded. Green for organic phosphorus (Po) and blue for inorganic phosphorus (Pi). Hedley fractions are indicated in italics for each relevant stock. Please note that no arrows have been drawn from the organic and inorganic stocks to erosion and runoff for clarity, but these fluxes are an important part of the system. Many unknown fluxes remain.

Compiled from several sources, particularly Hedley et al. (Reference Hedley, Stewart and Chauhan1982a), Parton et al. (Reference Parton, Sanford, Sanchez, Stewart, Coleman, Oades and Uehara1989), Smil (Reference Smil2000), MacDonald et al. (Reference MacDonald, Bennett, Potter and Ramankutty2011) and Sattari et al. (Reference Sattari, Bouwman, Giller and van Ittersum2012)

Dissolved organic phosphorus, however, has been found to move down and out of the profile in loamy, permanent-charge Alfisols of the long-term plots in Rothamsted, UK, from plots that had accumulated high levels of available phosphorus in the A horizon (Syers et al. Reference Syers, Johnston and Curtin2008). Rothamsted soils don’t have much phosphorus sorption, but Olibone and Rosolem (Reference Olibone, Rosolem and Alves2006) detected movement of dissolved organic phosphorus down to 40 cm in a clayey Oxisol of Brazil, which sorbs phosphorus strongly. There are no clear conclusions, but this is an area for future research.

Between the sources (phosphorus fertilizers 18 Tg P per year, plus 10 Tg P per year from manure phosphorus and 2 Tg P/ha from weathering of primary minerals) and the sink (surface waters) there are many phosphorus fractions or pools, six inorganic (in blue) and eight organic (in green) plus three combined ones (brown background) in Fig. 14.3. These stocks are connected by several opposite fluxes: dissolution and precipitation; sorption and desorption; mineralization and immobilization; decomposition of organic inputs and plant uptake. In addition, there are several one-way physical transfers: crop residue returns, manures, runoff and erosion, and sewage to surface waters.

Harvest removal by crops and cultivated pastures was 12.3 Tg P per year in the year 2000 (MacDonald et al. Reference MacDonald, Bennett, Potter and Ramankutty2011). Runoff and subsequent erosion of phosphorus are the main transfer mechanisms to water bodies, which Smil estimates to be 14 Tg P per year in cultivated land. Since phosphorus does not move much in the soil (except in very sandy soils) leaching losses are usually negligible. Much of the runoff and erosion losses happen during intense rainfall, particularly upon converting phosphorus-rich farmland to urban expansion and road development.

The phosphorus consumed as food from crop harvest removals and livestock products eventually winds up in cemeteries and urban sewer systems, the latter receiving 2–4 Tg P per year worldwide. Organized urban sewer systems are the norm in rich countries and in some cities of tropical Asia and Latin America. The exceptions are mainly in tropical Africa and remote areas of the other regions where sewage systems are seldom organized and are mostly non-point. Also interesting is Japan, where Smil observed that only half its population was connected to sewer systems at the end of the twentieth century. I can testify to that when visiting friends in Hokkaido in the 1980s.

Urban sewer systems also receive domestic waste and industrial detergents (with high phosphorus contents), which raise the total flux to 5 Tg P per year globally (Smil Reference Smil2000). Treatment of urban sewage decreases some of these phosphorus inputs by sedimentation and other means, and frequently biosolids are recycled to crop fields if they do not contain heavy metals.

14.1.3 Will the World Run Out of Phosphorus?

Because phosphorus is a finite resource at the human time scale, a key question to our survival is how long phosphate rock resources will last. Such a question has cropped up about once a decade during my professional career, with advocates estimating that we would run out of phosphorus by the year 2000, 2033, 2100 or similar times. It is an important question because phosphorus is one of the two plant nutrient elements where supplies are not only finite, but they are used at increasing rates that are highly correlated with the increasing demand for food (Stewart et al. Reference Stewart, Hammond, van Kauwenbergh, Sims and Sharpley2005). (The other major nutrient element is potassium where the supplies are also finite, but they are so large that it is not a concern.)

The best estimate I am aware of describes phosphate rock supply in terms of its price, as it will only be mined if it is economical. Stewart et al. (Reference Stewart, Hammond, van Kauwenbergh, Sims and Sharpley2005) calculated that at the global price US $40/Tg P of phosphate rock of the known above-ground reserves as of 2002, 12 Pg of phosphate rock would last for about 90 years, in other words, to the end of the twenty-first century. With an increase in phosphate rock price to about US $100/Tg P, the known above-ground phosphate rock reserves would almost quadruple, to 47 Pg of phosphate rock, which would last for 343 years (Stewart et al. Reference Stewart, Hammond, van Kauwenbergh, Sims and Sharpley2005). Subsequent estimates (van Kauwenbergh Reference van Kauwenbergh2010, Malingreau et al. Reference Malingreau, Eva and Maggio2012) agree with Stewart’s assessment. These estimates are alarming for centuries to come, but not for the century we live in. Such estimates assume the same level of technologies in phosphate rock beneficiation and phosphorus fertilizer use as well as current efficiencies. This should not make us complacent or give us an excuse not to be more efficient and recycle better. Ways of increasing the efficiency of phosphorus fertilizer and manure are presented towards the end of this chapter.

We do not have a mechanism to fully recycle the P used at a human time scale. It is unacceptable in virtually all cultures to recycle our ancestors’ bones, but something can be done about solid waste, livestock bones, erosion and runoff control.

I asked the International Fertilizer Industry Association (IFA) if new extraction technologies from coastal marine sediments are being developed. Michel Prud’homme (personal communication, 2012) indicated that there are a few ongoing projects investigating the economic feasibility of dredging shallow ocean sediments off the coasts of Namibia and New Zealand, but this is work in progress. A promising effort is that found off the coast of North Carolina (Christopher Lambe, personal communication, 2013).

The following sections describe in more detail the stocks and fluxes of phosphorus in the soil that are illustrated in Fig. 14.3, organized in terms of total, inorganic and organic phosphorus stocks.

14.3.1 Primary Minerals

The primary phosphorus soil minerals (mainly apatites) are present in the sand and silt fractions. They are slowly dissolved by the soil’s natural acidity and organic acid secretions from microorganisms, roots and mycorrhizae at the rate of 2 Tg P per year in cultivated soils (Smil Reference Smil2000). The resulting phosphate anions are short-lived in the soil solution because they are rapidly sorbed as secondary phosphorus minerals, or immobilized by microorganisms in soil organic matter (SOM). This is the origin of soil organic phosphorus (SOP).

14.3.2 Secondary Minerals

The secondary phosphorus minerals form in the soil as clay particles. They are grouped into three active fractions and one recalcitrant fraction in terms of their solubility, according to the classic method of Chang and Jackson (Reference Chang and Jackson1957). It is noteworthy that Chang and Jackson included two soils from the tropics out of the four soils used to develop their fractionation method. The active inorganic phosphorus fractions are grouped as calcium (Ca)-bonded phosphates, which are more prevalent in neutral to calcareous soils, and aluminum (Al)-bonded phosphates and iron (Fe)-bonded phosphates, both of which are most common in variable-charge acid soils. Calcium phosphates are more soluble than aluminum phosphates, which, in turn, are more soluble than iron phosphates. Calcium phosphates are present as discrete particles, while aluminum phosphates and iron phosphates are present as films adsorbed on clay surfaces.

The recalcitrant fractions are called occluded phosphorus and reductant-soluble phosphorus in the Chang and Jackson (Reference Chang and Jackson1957) method and as residual inorganic phosphorus extracted by concentrated sulfuric acid in the Hedley sequential fractionation (Hedley et al. Reference Hedley, Stewart and Chauhan1982a). Most of the total phosphorus found in Moura’s Brazilian Eutrustox was also in this fraction. In aerobic soils, occluded and reductant-soluble inorganic phosphorus act pretty much the same and are normally lumped together as occluded phosphorus. They often constitute the largest inorganic phosphorus fraction in the Chang and Jackson method shown in Tables 14.1, 14.2 and 14.3.

The transformation of one fraction of secondary inorganic phosphorus into another is controlled mainly by soil pH (Guo et al. Reference Guo, Yost, Hue, Evensen and Silva2000). As soils become more acid, the activity of iron and aluminum increases and the relatively soluble calcium phosphates are converted into less soluble aluminum and iron phosphates. These processes are slow enough to permit considerable quantities of calcium phosphates to be present in acid soils with pH values below 5.5, as shown in Table 14.2, with some even in a soil of pH 3.9, as will be shown later.

14.3.3 Sorbed Phosphorus and Available Phosphorus

Because of the importance of these two fractions, they will be discussed separately.

14.3.4 Soil Solution Phosphorus

Two phosphate anions exist in the soil solution: the dihydrogen phosphate anion (H₂PO₄^–), more abundant in soils < pH 7, and the hydrogen phosphate ion (HPO₄^2-), most abundant in soils > pH 7.2. Both anions move by the slow diffusion process as opposed to the fast mass flow process of many other nutrient ions, and both are readily taken up by plants.

The dynamics of soil solution phosphorus are very fast. Fluxes shown in Fig. 14.4 that add phosphorus to the soil solution – dissolution of phosphorus fertilizers, dissolution of secondary minerals, desorption of sorbed phosphorus and mineralization of organic phosphorus – vie constantly with depletion fluxes such as phosphorus sorption, immobilization by microorganisms and plant uptake. Plants accumulate in their tissues 1000–5000 mg/kg P (0.1–0.5 percent on a dry mass basis) from a soil solution of 0.2 mg/L P or less. Syers et al. (Reference Syers, Johnston and Curtin2008) estimated that when crops are having a rapid rate of uptake (0.3–0.5 kg P/ha per day), the average root hair cylinder needs to have its surrounding soil solution replaced 10–20 times per day. This varies with the discontinuity and tortuosity of the water films that contain the soil solution. Even at the wilting point, (–1500 kPa of soil moisture tension), water films around clays can be continuous. Other actors also vie for soil solution phosphorus; weeds and microorganisms want to take it up, iron and aluminum oxide surfaces sorb it. It’s a tough neighborhood there.

Fig. 14.4 Representation of the movement of phosphorus from a triple superphosphate granule to a well-granulated soil by mass flow and diffusion, indicating the precipitation and adsorption zones, as pH and phosphorus concentration changes.

From Hedley and McLaughlin (Reference Hedley, McLaughlin, JT Sims and Sharpley2005). Reproduced with permission from the American Society of Agronomy

14.4.1 Composition

Most SOP compounds have phosphorus bonded to oxygen in ester bonds (–C–O–P), unlike in soil organic nitrogen (SON) compounds where the nitrogen is bonded to carbon (–C–N–). More than half of the organic phosphorus is composed of monoesters, mainly inositol phosphate (phytate); about 5 percent is in phospholipids and 2 percent in nucleic acids, but a large amount is yet to be chemically identified (Stewart and Tiessen Reference Stewart and Tiessen1987, Majid et al. Reference Majid, Tiessen, Condron and Piccolo1996, Condron and Tiessen Reference Condron, Tiessen, Turner, Frossard and Baldwin2005). They are probably located in the slow and passive SOP pools shown in Fig. 14.3.

Phosphate ester bonds can be readily broken by one extracellular enzyme, phosphatase, which is produced by microorganisms, roots and mycorrhizae, while the release of nitrogen from its carbon bond requires the coordinated action of several enzymes (Vitousek et al. Reference Vitousek, Haettenschweiler, Olander and Allison2002). Because of this, Vitousek and co-workers assert that the organic component of the phosphorus cycle is more flexible than the nitrogen cycle. When fertilizing annual crops, only 3–10 percent of the applied phosphorus becomes organic phosphorus. Pastures convert more (5–30 percent) and perennial crops, agroforestry and forestry plantations the most (17–44 percent).

14.4.2 The Hedley Fractionation

A method of sequential phosphorus fractionation was developed by Hedley et al. (Reference Hedley, Stewart and Chauhan1982aReference Hedley, White and Nyeb), which has increasingly become the standard for soil phosphorus fractionation with minor modifications. It separates soil phosphorus into five inorganic fractions (Pi), three organic fractions (Po) and one residual fraction, largely inorganic. Increasingly strong reagents extract phosphorus fractions that are believed to be less available to plants. The organic Po fractions were determined by measuring total phosphorus in each extract and subtracting the Pi already determined. The version that I used (Beck and Sanchez Reference Beck and Sanchez1994) starts with an anion resin extraction (called resin Pi), a fraction that is used as a soil test in acid soils of Brazil (van Raij and Quaggio Reference 414van Raij and Quaggio1990). Then comes the 0.5 M NaHCO₃ extraction (called bicarb Pi and bicarb Po), which is the Olsen phosphorus soil test. This is then followed by a 0.1 M NaOH extraction (called NaOH Pi and Po), another 0.1 M NaOH extraction followed by sonication (called sonic Pi and Po), then a 1 M HCl (called HCl Pi) extraction and finally a concentrated H₂SO₄ digestion (called residual phosphorus). Regardless of the size of the residual phosphorus, it represents the most recalcitrant forms of soil phosphorus that have little to do with soil fertility. However Guo et al. (Reference Guo, Yost, Hue, Evensen and Silva2000) observed that the residual phosphorus fraction decreased with exhaustive cropping in soils with permanent charge while there was no change in soils with variable charge.

Cross and Schlesinger (Reference Cross and Schlesinger1995) stated that one advantage of the Hedley scheme over previous methods is that the same soil sample is sequentially treated with the various reagents. As a result, one can establish the proportion of the different fraction in each sample. The largest disadvantage is that residual phosphorus often accounts for the largest proportion of total phosphorus, and we don’t really know whether it is organic or inorganic, or both. This was improved by Tiessen and Moir (Reference Tiessen, Moir and Carter1993) by adding a hot concentrated HCl extraction that separates the residual Po from residual Pi, but it is not commonly used. The organic SOP pools are based on modeling, while the inorganic SOP ones are measured, so they are fractions of the total inorganic phosphorus content of a soil.

Some scientists see sequential extractions as potentially misleading, because they may redistribute the different phosphorus species (Roel Merckx, personal communication, 2013). Perhaps what Guo et al. (Reference Guo, Yost, Hue, Evensen and Silva2000) found is such a case, but the differences due to mineralogy are noteworthy.

Below I have attempted to call each Hedley fraction by what I feel is the most appropriate functional name, and relate them to the organic phosphorus pools of the Century model (Parton et al. Reference Parton, Schimel, Cole and Ojima1987) as well as the Chang and Jackson (Reference Chang and Jackson1957) inorganic phosphorus fractions. They are:

• Soil solution Pi: resin Pi, capturing phosphate anions in the soil solution.

• Labile Pi: available Pi, or Olsen soil test available phosphorus (bicarb Pi). When another soil test is more effective in a region, such as Mehlich or Bray, it can be used instead. This is why I call this fraction labile and not bicarb.

• Active Po: labile Po. Clearly, this is because it contributes directly to plant available phosphorus by rapid mineralization.

• Sorbed Pi: The NaOH-extractable Pi fraction, phosphorus sorbed on the surfaces of iron and aluminum oxides (Garcia-Montiel et al. Reference Garcia-Montiel, Neill, Melillo, Thomas, Steudler and Cerri2000).

• Slow Po: NaOH-extractable Po.

• Primary weatherable phosphorus minerals: sonic Pi.

• Passive Po: sonic Po + residual Po: extracted with hot, concentrated HCl (Tiessen and Moir Reference Tiessen, Moir and Carter1993) if measured.

• Calcium-bonded Pi (Ca-Pi): dilute HCl-extractable Pi. May reflect the phosphorus fertilizer that has not reacted, as well as any apatite present.

• Residual phosphorus: H₂SO₄-extractable Pi: Mostly occluded and reductant soluble Pi.

Needless to say, this classification is likely to require verification. All the Hedley fractions are incorporated in Fig. 14.3.

14.4.3 Organic Inputs, Metabolic and Structural Phosphorus Fractions

The organic inputs or organic resources include the additions to the soil of above- and below-ground plant litter, crop residues, animal manures, green manures, sewage and cadavers of animals that died in the soil. For nitrogen, the plant resources are split in terms of their quality according to their lignin:nitrogen ratios (Chapters 12 and 13). This is not directly applicable to phosphorus (Palm and Sanchez Reference Palm and Sanchez1991). For phosphorus, the main quality parameter is the C:P ratio, which Parton et al. (Reference Parton, Sanford, Sanchez, Stewart, Coleman, Oades and Uehara1989) used to split metabolic and structural phosphorus fractions.

Metabolic phosphorus comes from the decomposition of plant material that has a C:P ratio lower than 200, constituting the high-quality resource that is mineralized quickly and goes directly to the available phosphorus and soil solution phosphorus fractions or enters the active SOP pool (Nziguheba and Bünemann Reference Nziguheba, Bünemann, Turner, Frossard and Baldwin2005). The microbial biomass included in the metabolic phosphorus pool also immobilizes soil solution phosphorus; hence, the two-way flux in Fig. 14.3.

The structural phosphorus results from organic materials with C:P ratio of above 200, which is converted into the slow SOP pool and is then mineralized slowly into available phosphorus and soil solution phosphorus. This flux is one way, because there will be little immobilization by the slow SOP pool. Nziguheba (Reference Nziguheba and Bationo2007) confirmed these limits with African plant materials and related them to the percentage phosphorus content of plants.Footnote ² The organic inputs that result in net mineralization had more than 0.25 percent P, while the ones that resulted in immobilization of phosphorus had less than that amount. Included in the mineralizing group is Tithonia diversifolia with 0.4 percent P in its tissue. Soil moisture and temperature affect the rate of these decomposition, mineralization and immobilization fluxes.

Terrestrial plants take up about 70–100 Tg P per year, all from the soil. Of that amount, Smil (Reference Smil2000) estimates that 18–22 Tg P are annually harvested, and thus removed in crops and livestock products. It is surprising that a major proportion of these removals by humans are recycled back to the soil. Smil (Reference Smil2000) estimates that the actual crop residue returns (1–2 Tg P per year) plus animal manure applications (6–10 Tg P per year) recycle back 7–10 Tg P per year, though the manure portion usually ends up in another field or in feedlots. Childers et al. (Reference Childers, Corman, Edwards and Elser2011) believe that the proportion recycled is only about 25 percent.

14.4.4 Active SOP Fraction

This is the first true organic phosphorus pool in Fig. 14.3, which is equivalent to the Labile Po fraction in the Hedley fractionation. It consists of microbial biomass and organic phosphorus that is not physically protected or chemically recalcitrant as described in Chapter 11. The C:P ratio of this organic phosphorus is about 30–80. This is the fraction recognized by Hedley and co-workers to be the organic phosphorus fraction most responsible for supplying phosphorus to plants, and, as in the case of nitrogen, the main organic one responsible for soil fertility provision. The microbial biomass included in the active pool also immobilizes available or soil solution phosphorus, hence the two-way flux in Fig. 14.3.

14.4.5 Slow and Passive SOP Fractions

Part of the active SOP pool that is not mineralized or taken up by plants decomposes into the slow SOP pool, with a C:P ratio of 90–200, and into the passive SOP pool with a C:P ratio of 20–200 (Parton et al. Reference Parton, Schimel, Cole and Ojima1987, Reference Parton, Sanford, Sanchez, Stewart, Coleman, Oades and Uehara1989). The nice thing about the Hedley fractionation is that the NaOH Po fraction and the Sonic Po fractions represent well the slow and the passive pools, respectively. What is not clear is whether the residual phosphorus fraction, extracted by concentrated sulfuric acid in the Hedley scheme, has any organic phosphorus present. Therefore, a question mark flux is placed in Fig. 14.3 between the passive SOP pool and the residual phosphorus fraction. As in the case of nitrogen, the slow and passive SOP pools will mineralize very slowly and most of the phosphorus will remain in the aggregates, either protected by coatings or in very chemically recalcitrant forms. They contribute little to soil fertility but much to aggregate stability and related soil physical properties. Again, the results of Guo et al. (Reference Guo, Yost, Hue, Evensen and Silva2000) suggest that there is a difference between soils with permanent charge and those with variable charge with regard to the plant availability of the residual phosphorus fraction.

14.5.1 Precipitation from the Dissolution of Phosphorus Fertilizers

Precipitation is the first process that occurs to the phosphorus that has been added to soil as a fertilizer (Fig. 14.4). When a triple superphosphate granule containing mainly monocalcium phosphate [Ca(H₂PO₄)₂·H₂O] is added to an acid mineral soil, water moves into the hygroscopic granule, dissolves it and forms dicalcium phosphate (CaHPO₄·2H₂O) and phosphoric acid, which produces a solution saturated with P, as indicated in the Eq. 14.1.

In a dry soil, this solution is very acid (pH 1.5) and moves outward from the granule by mass flow to about 10 mm, releasing silica (SiO₂), and iron, aluminum and manganese (Mn) ions. As the granule gets wetter the pH increases. The residue is the insoluble CaHPO₄·2H₂O, and the soluble phosphate ions precipitate with iron and aluminum oxides, represented as Al(OH)₂⁺ in the following equation:

In moist soils, the pH increases, releasing aluminum and iron compounds and precipitating them as sparingly soluble phosphates.

These precipitates are some of the secondary phosphorus minerals previously described – crystalline calcium, iron and aluminum phosphates, amorphous iron and aluminum phosphates and iron or aluminum organophosphates. As these precipitates dissolve in an acid medium the H₂PO₄^– ions are sorbed by the aluminum and iron oxides on the clay surfaces and are gradually released, as will be explained later.

It is important to realize that these reactions are not acid-forming and therefore soluble phosphorus fertilizers do not produce secondary acidity as some nitrogen fertilizers do (Chapter 9). Also, it is important to recognize that the H₂PO₄^– and HPO₄^2– ions the plants take up do not come directly from the water-soluble fertilizer but from fertilizer reaction products (White Reference White2006).

In acid soils that contain significant quantities of exchangeable aluminum ions (Al³⁺), phosphorus fertilizers can also be precipitated by exchangeable Al³⁺(Coleman et al. Reference Coleman, Thorup and Jackson1960). Exchangeable Al³⁺, however, must first be displaced into the soil solution by the calcium contained in fertilizers and then hydrolyzed into Al(OH)²⁺ or Al(OH)₂⁺ before being able to precipitate H₂PO₄^–. About 102 mg/kg P as H₂PO₄^– can be precipitated by 1 cmol_c/kg of exchangeable Al³⁺ after being hydrolyzed.

Diammonium phosphate (DAP) is the most common phosphorus fertilizer used in the tropics. It is important to contrast the different reaction that occurs when a DAP granule enters contact with water in the soil, as compared with TSP, described above.

The resulting phosphorus-saturated solution has a pH of 10.6 (Ganga Hettiarachchi, personal communication, 2013) totally the opposite from the strongly, short-lived, acid solution (pH 1.5) produced by the superphosphates. DAP causes less dissolution of iron and aluminum (Hedley and McLaughlin Reference Hedley, McLaughlin, JT Sims and Sharpley2005) and many of the phosphate ions remain in solution and are subject to sorption.

In alkaline soils (> pH 7), phosphate anions are precipitated by Ca²⁺ and Mg²⁺ cations, as relatively insoluble compounds, which are slowly dissolved by organic acids.

14.5.2 Adsorption

In acid soils (< pH 7) phosphate ions in a soil solution that is not saturated with phosphorus enter into ligand exchange reactions with OH^– ions on the surface of iron and aluminum oxide particles or films. Ligand exchange occurs when an ion or molecule binds with metal ions, such as Fe³⁺, Al³⁺, Ca²⁺, Mg²⁺, K⁺, etc. In acid soils, aluminum and iron are the most abundant metals on the surface of clay particles. Oxide surfaces may exhibit net negative, net positive or net zero charge, as discussed in Chapter 8. Equation 14.4 illustrates a ligand exchange reaction at an acid pH value above the zero point of charge, which is the case in most topsoils:

(14.4)

Above the zero point of charge, the sorbed phosphate ion displaces hydroxide (OH^–) ions, and an increase in pH may take place. (This is sometimes referred to as “liming with phosphorus,” which does not make agronomic or economic sense.)

As Fig. 14.4 illustrates, the phosphate ions do not penetrate much into the micropores located inside macroaggregates, because of their slow diffusion and the discontinuities of the soil solution. In fact, most adsorbed phosphorus is on the surface of macroaggregates, the water films of which are in contact with the soil solution containing phosphate anions. Linquist et al. (Reference Linquist, Singleton, Yost and Cassman1997) found that phosphate ions were sorbed in a 188-μm layer around the macroaggregates of a clayey, oxidic Hawaiian Ultisol. Likewise, the precipitation products are also likely to be at the surface of macroaggregates unless they were present when microaggregates coalesced into macroaggregates (see Chapter 11). However, in Linquist et al.’s soil, 88 percent of the aggregates were macroaggregates (> 250 μm). The few microaggregates fully adsorbed all the phosphate throughout their mass, since most of them had a diameter smaller than 188 μm.

There is usually less adsorption in alkaline soils because they have low amounts of iron and aluminum oxide coats.

14.5.3 Absorption

It is common for well-aggregated soils that are high in iron and aluminum oxides to retain over 90 percent of the phosphate applied within the first hour of contact by adsorption or precipitation. Diffusion from this 0.2-mm layer into the rest of the macroaggregate (absorption) is a slow process, requiring at least 2 years to reach equilibrium (Linquist et al. Reference Linquist, Singleton, Yost and Cassman1997). Barrow (Reference Barrow1974) suggested that these slower absorption reactions are a consequence of the disruption of hydroxy-aluminum gels and the displacement of the silicate layer in the crystal structure by phosphate ions penetrating the mineral matrix. Some scientists suggest that the slow reaction is the formation of a second bond of a phosphorus molecule to an oxide surface (Russell Yost, personal communication, 2013).

At high phosphorus concentrations in the soil solution, H₂PO₄^– is adsorbed in new sites arising from slow absorption (Rajan and Perrot Reference Rajan and Perrott1975), meaning that high phosphorus fertilization rates open new phosphorus sorption sites. This is particularly the case with Andisols, where allophane (a hydroxy-aluminum gel) is dominant.

14.5.4 Estimating Phosphorus Sorption in the Laboratory

Most soils retain some amount of phosphorus, which has been estimated by a variety of methods involving widely diverging amounts of phosphorus added and extracted.

Perhaps the most widely used method is an adaptation of the Freundlich and Langmuir sorption isotherms – originally developed to describe the adsorption of a monomolecular layer of gas on a solid surface. Beckwith (Reference Beckwith1965) suggested that the data be plotted in terms of straightforward quantity–intensity relationships. The amount of phosphorus sorbed by the soil surfaces (P_S) is expressed in mg/kg P, kg P/ha or μmol P/cm² as the quantity factor plotted on the y axis; the phosphorus concentration in the soil solution (P_L) is expressed in mg/L P or μmol P/cm³ as the intensity factor, plotted on the x axis at a logarithmic scale.

There are some conceptual problems with the use of sorption isotherms. Most methodologies focus on adsorption and ignore the absorption and precipitation processes, the latter being particularly important in calcareous soils. Pierzynski et al. (Reference Pierzynski, McDowell, Sims, Sims and Sharpley2005) emphasize that the parameters obtained with such equations are empirical artifacts and do not physically exist in soils. This is the third time in this book that we realize that empirical artifacts are used as practical tools of soil science (the previous ones being humic and fulvic acids in Chapter 11, and available nutrients from soil tests in Chapter 12). Barrow (Reference Barrow1985) observed that the term “isotherm” is inappropriate because it implies that temperature is the only factor affecting the shape of the curves. He suggested the more straightforward term “sorption curve,” which I prefer, to make terminology clearer with less jargon.

Fox and Kamprath (Reference Fox and Kamprath1970) developed the most widely used analytical procedure for developing phosphorus sorption curves. A known amount of soil is equilibrated with different concentrations of Ca(H₂PO₄)₂·H₂O dissolved in 0.01M CaCl₂ and shaken for 30 minutes, twice a day for 6 days, after which the phosphorus remaining in solution (P_L) is assumed to be in equilibrium with the sorbed P_S. A 0.01M CaCl₂ solution has been a reasonable approximation of the soil solution from the analytical standpoint (Rajan and Fox Reference Rajan and Fox1972). The amount of P_S sorbed by the soil that results in a P_L soil solution concentration of 0.2 mg/L (0.2 ppm) was then considered the agronomically relevant estimate of phosphorus sorption. The Fox–Kamprath method still dominates tropical soil science, but it fell out of favor in the soil chemistry community for some years. Pierzynski et al. (Reference Pierzynski, McDowell, Sims, Sims and Sharpley2005) indicate that it has come back in vogue, because the sorption curves accommodate the high phosphorus levels present in manures, organic wastes and heavily fertilized soils (1–8 mg/L P_L).

This method, however, does not differentiate between the fast adsorption and the slow absorption processes, which can continue to decrease P_L for weeks or months (Barrow Reference Barrow1985).

Figure 14.5 shows several representative phosphorus sorption curves encompassing a wide range of soils (all topsoils). As a group, Andisols and related andic soils, high in allophane or imogolite, are the highest phosphorus retainers with about 500 mg/kg of added phosphorus to reach 0.01 mg/L P_L. Clayey, oxidic soils, including kaolinitic soils with oxide coatings, mainly Oxisols, Ultisols and rhodic Alfisols, follow, with values ranging from 100 mg/kg P_S to 400 mg/kg P_S to reach 0.01 mg/L P_L. Coarse-textured Ultisols retain little phosphorus. Mollisols with permanent-charge minerals sorb very little phosphorus, as illustrated by the Haplustoll from Hawaii, with 72 percent clay.

Fig. 14.5 Examples of phosphorus sorption curves determined by the Fox and Kamprath (Reference Fox and Kamprath1970) method. The general critical level for crops is 0.2 mg/L (0.2 ppm) P_L in the x axis (right dotted line). Instead I propose that those soils with a P_L of 0.01 mg/L or less be considered as those with high phosphorus sorption capacity (left dotted line).

Updated from Sanchez and Uehara (Reference Sanchez, Uehara, Khasawneh, Sample and Kamprath1980)

14.5.5 Proposed New Limit for Soils with High Phosphorus Sorption Capacity

The critical level of 0.02 mg P/L as the best estimate of what plants need is a gross overstatement when compared with critical soil test levels, which, as described later, represent a well-correlated empirical estimate with plant uptake. Table 14.5 shows the overestimate, where critical soil test levels on the order of 150 mg/kg (ppm) of available phosphorus corresponds to a soil solution level of 0.2 mg/L P_L, as originally suggested by Fox and Kamprath (Reference Fox and Kamprath1970). In contrast, a 0.012–0.006 mg/L P_L relates to critical soil test levels orders of magnitude lower (3–30 mg/kg of available phosphorus), which is the range encountered in practice (8–30 mg/kg P).

Therefore I propose that soils with P_L of 0.01 mg/L or less be considered to have high phosphorus sorption capacity. This is a more realistic P_L critical level relative to the amount of P_S sorbed to provide that critical soil test value. A value of 1 mg/kg P_S sorbed is roughly equivalent to 2 kg P/ha in most soils (100–500 kg P/ha) except in Andisols, where because of low bulk density the conversion factor would be about 1.4 (70–350 kg P/ha).

14.5.6 Estimating High Phosphorus Sorption in the Field

Luckily, a complex metric like high phosphorus sorption can be estimated in the field for topsoils using the FCC system created by Buol et al. (Reference Buol, Sanchez, Cate, Granger, Bornemisza and Alvarado1975), described in Chapter 5. High phosphorus sorption is identified by two FCC attributes as show in Table 14.6.

Except for the first two methods that are used for the i attribute, and one of those used for the x attribute, all are field tests. In addition the x attribute can be quantitatively detected by near-infrared spectroscopy, making this technique very useful, as shown in Chapter 5. Unfortunately, this technique cannot currently be used for identifying soil solution P_L or in soil tests for available phosphorus.

14.5.7 Geographical Distribution of Soils with a High Phosphorus Sorption Capacity

The tabular distribution of soils with high phosphorus sorption capacity by tropical geographic region is shown in Table 14.7. About half of these soils are in tropical America, covering 32 percent of the region, followed by tropical Asia, with a regional prevalence of 30 percent, mainly in the upland areas with Oxisols, Ultisols and Andisols. Tropical Africa has a low prevalence of soils with high phosphorus sorption capacity, and they occur mostly in the Lake Victoria Basin. The only biome in the world with more than 20 percent of such soils is the tropical–subtropical moist broadleaved forest.

14.5.8 Influence of the Core Soil Properties on Phosphorus Sorption

Reminding ourselves that in phosphorus sorption we are dealing with the topsoil, the influences of the three core properties are considered below.