Chapter 5
Factors affecting the agronomic effectiveness of
phosphate rocks, with a case-study analysis

Chapter 4 reviewed several approaches and methods for the evaluation of phosphate rock (PRs) for direct application. It showed that the final step of such evaluation is to determine their agronomic effectiveness in the field because of the need to integrate a range of factors and interactions in practical farming conditions. The information obtained from the field-evaluation programme should be interpreted in order to provide recommendations on the direct application of PRs for the farmers. This entails a correct understanding of the influence of the various factors that affect the dissolution and the agronomic effectiveness of PRs. The main factors are: reactivity of PRs, soil properties, climate conditions (especially rainfall), and types of crops grown.

This chapter examines the influence of these factors on the agronomic effectiveness of PRs. In addition, it presents case studies on the use of PRs in selected countries representing different geographical regions in order to illustrate how these factors and their interactions influence the agronomic performance of PRs. The countries include: Mali and Madagascar (Africa), India (South Asia), Indonesia (Southeast Asia), New Zealand (Oceania), and Venezuela and Brazil (Latin America). Although there is considerable variation in soil types, agroclimate conditions and crops grown within and across regions and countries, the fundamental factors affecting the agronomic effectiveness of PRs remain the same. As several comprehensive reviews on the use of PRs have already been published (Khasawneh and Doll, 1978; Hammond et al., 1986b; Bolan et al., 1990; Sale and Mokwunye, 1993; Rajan et al., 1996; Chien, 2003), this chapter focuses on the analysis of the significance of the main factors mentioned above.

Factors affecting the agronomic effectiveness of phosphate rocks

Reactivity of PRs

The reactivity of PRs is a measure of the rate of dissolution of PRs under standard laboratory conditions or in a given soil and under given field conditions (Rajan et al., 1996). It excludes the changes in the rate of dissolution caused by varying soil properties and by plant effects. The chemical composition and particle size of PRs determines their reactivity. PRs of sedimentary origin are generally most reactive and, therefore, suitable for direct application.

The chemical properties that influence the reactivity of PRs are phosphate crystal (apatite) structure and the presence of accessory materials, especially calcium carbonate (Chapters 3 and 8). Increasing the substitution of carbonate to phosphate in the crystal structure generally increases the reactivity of PRs. This substitution results in decreased cell a dimensions and also in a weakening of the apatite crystal structure (Lehr and McClellan, 1972; Chien 1977). The most reactive PRs are those with a molar PO₄:CO₃ ratio of 3.5-5.

Calcium carbonate is the most abundant accessory mineral in PRs. As calcium carbonate is more soluble than the most chemically reactive phosphate mineral (Silverman et al., 1952), its dissolution increases the calcium (Ca) concentration and pH at the phosphate mineral surface. Thus, it is not surprising that accessory calcium carbonate can reduce the rate of PR dissolution in some soils (Anderson et al., 1985; Robinson et al., 1992). However, under field conditions, leaching and plant uptake may remove Ca ions. The magnitude of the removal by leaching may vary according to soil and climate conditions and to the mode of PR application. For surface applied fertilizer, the calcium carbonate effect can be minimal even if its content is high, as evidenced for Chatham Rise PR from New Zealand, which contained 27 percent free calcium carbonate (Rajan, 1987). On the other hand, for incorporated PR, more than 15 percent of free calcium carbonate may lessen PR effectiveness in a limed alkaline soil (Habib et al., 1999).

As PRs are relatively insoluble materials, their particle size has an important bearing on their rate of dissolution in soil. The finer the particle size, the greater is the degree of contact between PR and soil and, therefore, the higher is the rate of PR dissolution. Moreover, the increase in the number of PR particles per unit weight of PR applied increases the chances of root hairs intercepting PR particles. Thus, application of PR as finely ground materials (usually less than 0.15 mm) enhances both the rate of dissolution of PRs and the uptake of PR-phosphorus in a given soil. On the negative side, because of their dusty nature, the application of finely ground materials is fraught with practical difficulties.

Among the various methods for measuring the reactivity of PRs (Chapters 3 and 4), a rapid test to determine the reactivity of PRs is to extract PRs with dilute chemical solutions, e.g. 2- percent citric acid (CA), 2-percent formic acid (FA) or neutral ammonium citrate (NAC). The phosphorus (P) extracted is usually expressed as a percentage of total P. In general, the greater the chemical extractability of PRs in these solutions, the higher their reactivity and, therefore, their agronomic effectiveness (Chapter 4). The level of chemical extractability needed for PRs to be agronomically effective varies with soil and climate conditions, especially rainfall. In New Zealand and Australia, the recommended extractability level for permanent pasture is 30 percent of total P soluble in 2-percent CA, whereas in the European Community (EC) it is 55 percent soluble in 2-percent FA for crops. The extractable P values should always be considered together with the total P of PRs.

While chemically extractable P is a good indicator of PR reactivity, some PR deposits may give low indices but still be agronomically effective. For example, although Mussoorie PR from India has a CA-extractable P of 8 percent of total P, it has been found to be agronomically as effective as superphosphate in some soils. The enhanced agronomic effectiveness is attributed to the oxidation of iron sulphide to sulphuric acid and to the localized acidulation of the PR. The iron sulphide is present in intimate association with the apatite crystals. In addition, the PR contains organic carbon (1.14 percent) in the mineralogical composition. This is likely to improve its internal porosity and, therefore, the dissolution of PR-phosphorus (PPCL, 1982).

Soil properties

For a given PR to be agronomically effective, the PR should not only dissolve, but the dissolved PR should also be available to plants. The soil properties that favour the dissolution of PR are low pH (less than 5.5), low solution concentration of Ca ions, low P fertility levels and high organic-matter content.

Soil acidity

The dissolution of PR may be expressed by the equation:

Although the above reaction is for a fluorapatite PR, it applies to other members of the apatite minerals including reactive PRs (RPRs). As indicated in the above equation, the dissolution of PR results in the release of hydroxyl ions into solution. Neutralization of the hydroxyl ions released by soil acidity enables the PR dissolution process to continue. In the case of PRs where phosphate has been substituted with carbonate ions, hydrogen ions may also be needed to neutralize hydroxyl ions formed on the release of carbonate ions into solution (Chien, 1977b). Each carbonate ion (CO₃^2-) joins with two hydrogen ions and forms one water molecule and carbon dioxide gas. Thus, an adequate supply of hydrogen ions is of primary importance for the continued dissolution of PR (Chapter 4).

Indicators of hydrogen ion supply are soil pH and titratable acidity. Soil pH shows the magnitude of hydrogen ion supply at a given time, whereas titratable acidity indicates the supply of hydrogen ions in the longer term. A linear positive relationship has been reported between initial pH and titratable acidity in Australian soils (Kanabo and Gilkes, 1987). As a simple guideline, the use of PRs, depending on their reactivity, is generally recommended in soils with a pH of 5.5 or less. The dissolution of PR diminishes with increasing pH up to 5.5 but the decline is more rapid above this pH level (Bolan and Hedley, 1990; and Chapter 4). When considering a large number of soils, titratable acidity may be a better indicator of PR dissolution (Babare et al., 1997).

Cation exchange capacity, and exchangeable calcium and magnesium

For continual dissolution of PR, it is important that the other major reaction product, the Ca ion, be removed or that its concentration in soil solution be maintained at a lower level than that in the film surrounding the dissolving PR particle. It is possible to achieve these outcomes if there are adequate soil cation exchange sites available to adsorb the Ca ions released from the PR, or if Ca is leached away from the site of PR dissolution. A measure of the cation exchange sites available for Ca adsorption is the difference between the cation exchange capacity of soils and the exchangeable Ca (Bolan et al., 1990; Robinson and Syers, 1991).

Recent studies suggest that high exchangeable magnesium (Mg) in soils may enhance PR dissolution (Perrott, 2003). Theory would suggest that, as Mg is held by soils more strongly than Ca, the presence of Mg on the soil exchange sites can block adsorption of Ca released on dissolution of PR and thereby facilitate its removal from the soil-fertilizer system. This will have the effect of enhanced PR dissolution. In soils with a low pH (less than 5.5), the exchangeable Ca and Mg will invariably be low (low base saturation) and, therefore, there will be low solution concentrations of these ions.

The cation exchange capacity of soils is also related closely to soil texture. Sandy soils usually have a low cation exchange capacity and, therefore, do not provide an adequate sink for Ca released from PR. This would lead to a reduction in PR dissolution and in agronomic effectiveness. The other two scenarios occur in areas of sufficient rainfall. The first is where the released Ca may be removed from near the PR particles, with a positive effect on PR dissolution and on agronomic effectiveness. The second is where excess rainfall may lead to leaching of P below the rooting zone of crops and reduce the agronomic effectiveness of PRs.

However, because of their slow-release nature, PRs are likely to be more beneficial under such circumstances than water-soluble fertilizers (Bolland et al., 1995).

Soil solution P concentration and P retention capacity

As the P concentration in soil solutions is usually very low (0.05-0.5 mg/litre), it has little influence on the dissolution of PR. Nevertheless, there have been reports that the greater the P sorption capacity of soils, which results in depletion of soil solution P, the greater the dissolution of PR (Chien et al., 1980a). It is not P adsorption capacity per se that affects PR dissolution but the number of sites available to adsorb the P released from PR and, therefore, maintain a lower P concentration in solution near PR particles.

When considering a large number of soils, the variation in the rate of PR dissolution in soils can be better explained by taking into account the P sorption capacity of soils in addition to the titratable acidity (Babare et al., 1997). Although increased P sorption capacity may favour PR dissolution, its availability will depend on soil P status and the amount of PR added.

When low amounts of PR are added to severely P-deficient soils, the soils adsorb tightly almost all of the dissolved P with very little increase in soil solution P. This results in very little increase in crop production (Zone A in Figure 12). At higher levels of P application, as the solution P increases above the threshold concentration for net P uptake by plants, crop yield rises steeply (Zone B in Figure 12) (Rajan, 1973; Fox et al., 1986). Soils of medium P status are likely to be in the region corresponding to the starting point of Zone B. In this case, the dissolved PR is likely to translate into crop yield. Thus, soils should preferably have a ‘medium or above’ level of P fertility to bring about immediate benefit from the application of PR at the maintenance rate. The maintenance rate is equivalent to the amount of P removed by crop produce. In such soils, the plant-available P can serve as starter P for crop establishment and early growth, which in turn assists the roots in utilizing PR more effectively. This is similar to the effect of water-soluble P on PR effectiveness (Chapter 9). Zone C in Figure 12 represents the yield plateau attained at high levels of PR application.

FIGURE 12
Schematic representation of a crop response curve on adding PR to a severely P-deficient soil

Soil organic matter

Another soil property that increases PR dissolution and its availability to plants is soil organic matter (Johnston, 1954b; Chien et al., 1990b). This seems to arise from: (i) the high cation exchange capacity of organic matter; (ii) the formation of Ca-organic-matter complexes; and (iii) organic acids dissolving PR and blocking soil P sorption sites.

The cation exchange capacity of organic matter is greater than that for clay minerals. Depending on their clay content, the cation exchange capacity of mineral soils may range from a few to 60 cmol/kg, whereas that of organic matter may exceed 200 cmol/kg (Helling et al., 1964). The high cation exchange capacity of organic matter means increased Ca retention capacity of soils, which leads to enhanced PR dissolution. Humic and fulvic fractions of organic matter form complexes with Ca (Schnitzer and Skinner, 1969), which can also reduce Ca concentration in solution, so leading to enhanced PR dissolution. The organic matter content of tropical soils is generally less than 2 percent.

When arable crops are harvested, a large proportion of the root residues and in some cases part of the above-ground portions are left behind in the soil. The decomposition of plant residues in soil results in the production of numerous organic acids, such as oxalic, citric and tartaric acids (Chapter 9). These acids can be expected to dissolve PR by supplying the hydrogen ions needed to neutralize the hydroxyl ions produced when PR dissolves and by forming complexes with cations, especially the Ca from PRs. The organic ions and humus can also reduce the P sorption capacity of soils by blocking P sorption sites and by forming complexes with iron and aluminium hydrous oxides, leading to increased P concentration in solution (Manickam, 1993).

Logic would dictate that incorporation of PR during cultivation between crops would benefit the farmers most. Such a practice would allow the decomposing plant residues, and any animal litter that might have been applied, to enhance the release of PR. The early application of PR would also allow time for the reaction of PR with the soil and the release of some P before the next crop is established.

Climate conditions

Rainfall is the most important climate factor that influences PR dissolution and its agronomic effectiveness. Increased soil water brought about by rainfall or irrigation increases PR dissolution (Weil et al., 1994). The process is affected by speedy neutralization of the hydroxyl ions released and removal of Ca and other reaction products from the area adjacent to PR particles. Adequate water supply will encourage plant growth and P uptake by plants, so leading to the increased agronomic effectiveness of PRs. For surface applied PRs, experience in Australia and New Zealand would indicate an annual rainfall requirement of at least 850 mm for PRs to be agronomically similar to water-soluble fertilizers (Hedley and Bolan, 1997; Sale et al., 1997b). However, the rainfall requirement does depend on soil properties.

Chien et al. (1980b) reported that temperature has negligible or no influence on the solubility of PRs within a range of 5-35°C and, therefore, on its agronomic effectiveness.

Crop species

Plant species differ in their P uptake demand and pattern as well as in their ability to absorb soil solution P (Helyar, 1998; Baligar, 2001). Moreover, plant species show differences in their ability to access sparingly forms of P that are unavailable to other plants (Hocking et al., 1997; Hinsinger, 1998; Hocking, 2001). Among these, some plants can dissolve and take up the products from PR dissolution (Hinsinger and Gilkes, 1997). For example, perennial pastures, tree crops and plantation crops require a steady supply of P over an extended time span. Because PRs in soil dissolve gradually and supply P at a steady rate, increasing amounts of PRs are being applied as phosphate fertilizers for the above-mentioned crops (Ling et al., 1990; Pushparajah et al., 1990; Chew et al., 1992). The high agronomic effectiveness of PRs realized with these crops reflects partly the acidic nature of the soils and the high root density. High root density facilitates the intensive exploration of a large soil volume for P because of the presence of a large number of fine roots per unit of soil volume.

Legumes are particularly suited for the use of PRs. They are effective in dissolving PR and in absorbing its dissolution products because of their demand for Ca and the acidifying effect of nitrogen (N) fixation in the soil near the root system (rhizosphere) (Ankomah et al., 1995; Kamh et al., 1999). This effect can be utilized to improve the P nutrition of a companion crop (intercropping) or that of the subsequent crop in a rotation (Horst and Waschkies, 1987; Vanlauwe et al., 2000).

Some plant species (e.g. rapeseed, lupines and pigeon pea) have been studied because of their ability to secrete organic acids that result in an enhanced dissolution of PR (Jones, 1998; Hoffland, 1992; Adams and Pate, 1992; Ae et al., 1990; Montenegro and Zapata, 2002). Recent studies (Chien, 2003) indicate that reactive PRs may have potential applications even in alkaline soils with organic-acid secreting crops such as rapeseed (canola) (Chapter 9).

Crops that possess high Ca uptake capacity are more suited for PR use. In this respect, finger millet is most suited for PR use, followed by pearl millet and maize (Flach et al., 1987).

Management practices

Four important management practices that can influence the agronomic effectiveness of PRs are: the placement of PR material in relation to the plants; the rate of application; the timing of application; and lime application.

PR placement

In order to achieve the maximum agronomic effectiveness from PRs, the material should preferably be broadcast and incorporated uniformly into the surface soil to the required depth. The depth of incorporation for seasonal crops may be of the order of 100-150 mm. Incorporation facilitates greater dissolution of PR by increasing contact between the soil and PR particles. It also enhances plant absorption of P by providing a greater volume of P-enriched soil. In addition, there is a greater likelihood that a root will encounter a dissolving PR particle.

Rate of PR application

The decision on the rate of PR application needs to be based on the soil P status as indicated by soil testing (Chapter 6; Perrott et al., 1993; Perrott and Wise, 2000), and the expected rate of dissolution of PR and its availability to plants (Chapter 9; Perrott et al., 1996; Rajan et al., 1996). The soil testing method to be used would depend on whether the phosphate fertilizer applied previously was a water-soluble form or a PR. Some general guidelines are:

• PR application is likely to be beneficial in soils of medium P status. In such soils, the minimum rate of application should be such that the expected amount of P dissolved from PR is no less than the amount of P removed from the site as farm produce plus the amount of phosphate retained by soils in a form that is not available to crops under near maximum production levels. This is often called the ‘maintenance P application’ rate. In terms of absolute amounts, soils with high P retention capacities will require greater rates of P fertilizer application than soils of lower P retention. This allows for the P retained by soils in a form that is ‘not available’ to plants.

• In soils of low P status there are two choices: (i) bring the fertility level to medium P status by applying water-soluble fertilizers and follow this by the application of PR; (ii) incorporate large applications of PR (500-1 000 kg/ha) followed by a regular maintenance application of P.

Timing of PR application

In very acid soils (pH less than 5.5) with a high P retention capacity, the incorporation of PR close to planting time is recommended in order to minimize conversion of dissolved P to plant ‘unavailable’ forms (Chien et al., 1990b). However, in less acid soils (pH of about 5.5-6) with a low P retention capacity, incorporation of PR 4-8 weeks ahead of planting is preferred. This allows time for some dissolution of PR and its subsequent availability to plants. Laboratory experiments have shown that it may take 4-8 weeks for PRs to reach their maximum solubility (Barnes and Kamprath, 1975).

The use of PR for flooded rice requires special attention because soil pH generally increases upon flooding. For this reason, it is advisable to apply PR to the soil about two weeks before flooding (Hellums, 1991).

Lime application

Incorporating lime has an adverse effect on PR dissolution in soil because it increases the Ca concentration in solution and reduces soil acidity. However, liming may increase the availability to crops of dissolved P by increasing soil pH and reducing aluminium (Al) toxicity. In view of the above effects, where liming is to be done in order to raise soil pH values to 5.5, it can be applied at the same time as the PR application, but preferably not as an admixture with PR.

This can eliminate Al toxicity while still encouraging PR dissolution. Where the soil pH is to be raised above pH 5.5, liming should preferably be done about six months after the incorporation of PR so that PR dissolution is not reduced drastically.

Case studies

Mali

The agriculture sector in Mali involves 80 percent of the population. However, less than 10 percent of the 2.7 million ha of cultivated land receives fertilizers. The country imports 70 000 tonnes of fertilizer per year. This fertilizer is applied mainly on cash crops such as cotton, rice and groundnut. This amount is less than 15 percent of the nutrients exported by the crops. The authorities have long tried to improve the situation, particularly by using local PR.

Tilemsi PR production

Mali has a relatively large phosphate ore deposit of sedimentary origin, located in the Tilemsi Valley about 120 km north of Bourem. Klockner Industrie Anlagen GMBH (1968) undertook a detailed study for its exploitation. The reserve was estimated to be 20 million tonnes with an average P₂O₅ content of 27-28 percent. The PR is easy to extract with a thin overburden. The exploitation started in 1976 as part of a Malian-German technical cooperation project. This involved the installation at Bourem of: an electricity generating unit; grinding, cycloning and bagging machines; and storage facilities. The production capacity was 36 000 tonnes/year. Actual production was about 1 000 tonnes/year in the 1970s, 3 000 tonnes/year in the 1980s and 10 000 tonnes/year in the 1990s. Production peaked in 1991 at 18 560 tonnes of PR. Since then, production has become inconsistent as a result of political unrest in the mine area.

At an output rate of 10 000 tonnes/year, the estimated production cost was US$20.6/tonne of raw material arriving at Bourem. This increased to US$78.5/tonne for ground and bagged product ex-factory, and then increased to US$157/tonne delivered in Sikasso, the main utilization zone, 1 300 km to the south. The estimated price of triple superphosphate (TSP) delivered in Sikasso was US$273/tonne, which is US$0.60/kg of P₂O₅ compared with US$0.54/kg of P₂O₅ for the Tilemsi PR. This difference is not large (Truong and Fayard, 1993).

A cost analysis shows that the cost of grinding and bagging (US$57.9/tonne) at Bourem is too high. Bourem is located in the desert, so all the components for production including fuel oil, bags, spare parts, maintenance, and the workforce, are very expensive. The transfer of the factory to Segou or Koulikoro 1 100 km to the south could generate important economies. Connection to the national hydroelectricity grid is possible, and many facilities of a big city are available, notably the security needed to ensure regular production. The transport cost is also very high (US$78.5/tonne). The traditional dugouts (capacity 10-20 tonnes) can sail year round on the Niger River. Almost 95 percent of the transported goods flow from south to north, and only 5 percent in the reverse direction. Thus, it would be possible to take advantage of this return freight availability to transport Tilemsi PR at a lower cost.

Tilemsi PR evaluation and use

Tilemsi PR is a medium-reactive rock, with a total P₂O₅ of 29 percent, of which 61 percent is soluble in FA. Therefore, it is suitable for direct application. Since 1977, many studies have been conducted in controlled and natural conditions covering the main pedoclimatic zones and cropping systems in the country. The results have shown that the effectiveness of Tilemsi PR is very dependent on rainfall distribution. The relative agronomic effectiveness (RAE) averages about 80 percent compared with TSP.

Research institutions recommend an annual application of Tilemsi PR of 100-200 kg/ha, or 300-400 kg/ha for a rotation. The PR should be applied on the fallow and incorporated through late ploughing at the end of the rainy season. The Government of Mali has tried to promote the use of Tilemsi PR through national and provincial extension services, and through development companies for cotton, groundnut, rice and sugar-cane crops. The only company to sell significant quantities (about 5 000-10 000 tonnes/year) of Tilemsi PR to farmers has been the Malian Company for the Development of Textiles (CMDT). The CMDT plays an important technical, social and economic role in the region.

In the period 1982-87, the Government of Mali conducted a fertilizer project through the Institut d’Economie Rurale and the International Fertilizer Development Center (IFDC) in the five main pedoecological regions, namely: Mopti (millet), Kayes (groundnut), Segou (millet), Koulikoro (sorghum) and Sikasso (cotton and maize). Henao and Baanante (1999) made a comprehensive analysis of the results, including an economic evaluation. Table 20 presents a summary of the agronomic results. In medium to long-term experiments, Tilemsi PR is practically equivalent to TSP per unit of P₂O₅.

TABLE 20
Yield of millet, groundnut, sorghum, cotton and maize with Tilemsi PR and TSP, Mali, 1982-87

Madagascar

The fertilizer situation in Madagascar is a cause for serious concern. The total consumption is about 15 000 tonnes/year for 1.7 million ha of cultivated land. The export of nutrients by crops amounts to 205 000 tonnes (N + P₂O₅ + K₂O). This means that the replacement rate covers less than 4 percent of the nutrients removed. Therefore, heavy soil mining is in progress.

In Madagascar, PR deposits are limited, with estimated reserves of 600 000 tonnes. These are in the form of guano materials deposited on coral reefs, which are distributed over the island archipelago known as the Barren Islands (Ratsimbazafy, 1975). More than half of these deposits (312 000 tonnes) are located on the island of Andrano. Because these deposits are located on coral reefs, the mining requires special efforts in order to preserve the ecology of the archipelago. This limits mining to 10 000 tonnes/year and requires replacement of the mined PR with the same quantity of soil, which has to be transported from the main island to the coral reefs.

The Barren Island PRs are reactive. For example, the FA solubilities (percentage of total P₂O₅) of the Andrano, Androtra and Morombe PRs all exceed 70 percent. The agronomic effectiveness of these PRs was further confirmed by a pot experiment where the three PRs were applied at the rate of 100 mg of P per kilogram of soil in 100 g of an acid (pH of 4.3) Andosol soil from Madagascar. Three monthly harvests of the test plant Agrostis sp. were taken. The availability coefficients for these PRs, defined as: ((P uptake [PR] - P uptake [control])/(P uptake [TSP] - P uptake [control])) x 100, were all in excess of 100 after one harvest, and over the three harvests. These results show that the Barren Islands PRs are very reactive, and that they are equivalent or superior to TSP in terms of agronomic effectiveness (Truong et al., 1982).

Two long-term field trials were conducted on the High Plateau of Madagascar, where the soil pH is 4.5 and the annual rainfall 1 200 mm. These involved growing maize for eight years and paddy rice for nine years. The PR was applied at a rate of 400 kg of P₂O₅ per hectare for maize and 300 kg of P₂O₅ per hectare for rice. There were marked responses in yield above the control treatment over time. These amounted to 25 000 kg of maize over eight years and 5 020 kg rice over nine years (IFDC-CIRAD-ICRAF-NORAGRIC, 1996). Thus, the reactivity of PR and favourable soil and climate conditions resulted in an effective performance by the PR.

Cost of production

Madagascar has no facilities for fertilizer production, and investment of capital and management staff in a small factory is not profitable. It is better to rent out the services of an existing dolomite factory at Antsirabe for grinding and bagging the PR. Hence, the cost of production will include the mining of PR from Andrano Island, transport by beacher (boat landing directly on the beach) to the port of Morondava, return transport of soil to Andrano, transport by truck from Morondava to Antsirabe, and grinding and bagging at the factory. The estimated ex-factory cost is US$102/tonne (US$0.51/kg P₂O₅) compared with US$418/tonne (US$0.92/kg P₂O₅) for TSP. These figures highlight the cost-effectiveness of the local Andrano PR compared with imported water-soluble P fertilizer.

India

In 1990/91, India imported fertilizer raw materials at a cost of US$338 million and manufactured fertilizers to the value of US$608 million (Srinivasan, 1994). Most of the fertilizer used in the country is in the form of N, and growth in the consumption of phosphate fertilizers has not kept pace with that of N fertilizers. Consequently, there seems to be a nutrient use imbalance. It is estimated that 46 percent of Indian soils are low in available P, 52 percent are of medium P status and 2 percent are of high P status (Tandon, 1987). Thus, there is a need to increase P fertilizer application in order to achieve higher productivity.

India imports about 70 percent of the PR necessary for phosphate fertilizer production and all of the elemental sulphur (S), mainly for use in the phosphate industry (Tandon, 1991). There are sizeable PR deposits in different parts of the country including: Mussoorie PR (Uttar Pradesh), Purulia PR (West Bengal), Jhabua PR (Madya Pradesh), Singhbhum PR (Bihar) and Kasipatnam PR (Andra Pradesh). Although the total estimated reserve is 130 million tonnes, about 60 percent of the deposits are of low grade and unsuitable for the manufacture of single or triple superphosphate (Jaggi, 1986). Of the Indian PRs, the deposit at Mussoorie (estimated reserve 45 million tonnes) and possibly that at Purulia (10 million tonnes) are considered useful for direct application. A low-reactive igneous PR from Rajasthan State (Jhamar-kotra PR) that is probably not suitable for direct application is also being marketed by Rajasthan State Mines and Minerals Ltd. under the product name Raji Phos. The estimated total PR reserve is 77 million tonnes with P contents ranging from 5 to 16 percent. In 1998, the PR applied directly amounted to about 11 000 tonnes of P, or 0.6 percent of the total P consumption of 1.8 million tonnes of P.

In India, acid soils occupy an estimated 49 million ha of agricultural land. Soil pH is less than 5.5 in 29 million ha of this land, with values of 5.6-6.5 in the rest of the area (Tandon, 1987). Almost 70 percent of India’s cropland is rainfed and may not be suitable for the direct application of PR. As a first approximation, assuming a P requirement of 30 kg/ha/year, then the demand for PR-phosphorus in soils with a pH of less than 5.5 is 234 000 tonnes of P per year, or 2.6 million tonnes of PR per year if added in the form of local Mussoorie PR.

Numerous field experiments have shown that the agronomic effectiveness of Mussoorie PR could be equal or similar to that of soluble P fertilizers in soils with a pH of less than 5.5 with plantation, leguminous, rice and maize crops provided there is adequate soil water (Tandon, 1987; Poojari et al., 1988). In soils with higher pH values, PRs may need to be applied as partially acidulated phosphate rock (PAPR) (Basak et al., 1988; Chien and Hammond 1988) (Figure 13) or as mixtures with water-soluble fertilizers (Singaram et al., 1995).

The most widely used Indian PR, Mussoorie PR, contains 8-9 percent total P, 1.14 percent of organic matter and 4 percent of sulphide sulphur. The CA-extractable P of finely ground material (particle size less than 0.15 mm) is less than 10 percent, which makes it about one-third as reactive as unground North Carolina PR. It is claimed that the oxidation of sulphide sulphur to sulphuric acid and the subsequent reaction of the acid on the francolite would increase the dissolution of Mussoorie PR and thus enhance its agronomic effectiveness. It has been demonstrated that oxidation of S in PR-sulphur granules increases the dissolution of PR (Rajan et al., 1983). However, the PR contains 15 percent of free carbonates, which may counter the effect of oxidation of S. In addition, Mussoorie PR contains organic carbon (1.14 percent) in the mineralogical composition, which is likely to improve its internal porosity and, therefore, the dissolution of PR-phosphorus.

FIGURE 13
Effect of partial acidulation of Mussoorie PR on yields of rice (soil pH 7.9) and wheat (soil pH 6.0)

Source: Chien and Hammond, 1988.

Field evaluation of Mussoorie PR

Mussoorie PR was agronomically as effective as single superphosphate (SSP) in a field trial with paddy rice on an acid (pH of about 5.0) hill soil of Uttar Pradesh (Table 21) (Mishra, 1975). In another trial conducted on an acid soil of higher pH (pH 5.8), Mussoorie PR gave 95 percent of maize yield relative to SSP application whereas a 1:1 mixture was as good as SSP (PPCL, 1980). Based on the cost of Mussoorie PR being 54 percent per unit of P as that in SSP, there is a saving of 46 percent on the cost of fertilization of paddy crop.

Mussoorie PR also proved effective on soils with a high pH provided there was an adequate supply of irrigation water (Singaram et al., 1995). This conclusion is based on an experiment conducted over three seasons to study the response of crops to current application as well as the residual effect of the P fertilizers. The experimental soil was a calcareous, reddish-brown, clay loam (Typic Ustropept) containing kaolinite and montmorillonite clays. The soil pH in water was 8.02, and it contained less than 1 percent of organic matter. Treatments included SSP, Mussoorie PR and a physical blend of SSP and Mussoorie PR applied at three rates of application, and a nil-P control. Three crops were grown in succession: finger millet, maize and black gram. Fertilizers were applied for finger millet and maize but not for black gram. All crops received irrigation as required.

TABLE 21
Yield of rice and maize with Mussoorie PR, SSP and a 1:1 mixture of Mussoorie PR-SSP

From the yield response curves, the substitution values (SVs) of the test fertilizers were calculated as the amount of total P applied as SSP that is required to produce 90 percent of maximum yield, divided by total P of the test fertilizer that is required to produce the same yield. Thus, a ratio of less than one indicates that the test fertilizer is less effective than SSP. The fertilizer substitution values for finger millet were 0.42 for Mussoorie PR and 0.68 for Mussoorie PR-SSP. Thus, the P supplying value of Mussoorie PR was 42 percent that of SSP. However, on application of fertilizers for maize, the respective values were 1.25 and 1.39. The residual effect of previously unreacted PR from the first application may have contributed to the substitution value exceeding one. Black-gram yields at the P application rate that gave 90 percent of maximum yield with a fresh application of SSP were 0.74 tonnes for the original SSP application, 0.74 tonnes for Mussoorie PR-SSP and 0.82 tonnes for Mussoorie PR.

Economic analysis (Table 22) shows that the net returns for the application of Mussoorie PR and Mussoorie PR-SSP were marginally better than for SSP (Singaram et al., 1995).

TABLE 22
Net returns on P fertilizers to finger millet and maize, and black gram grown on the residual effect

The above calculations do not take into account the nutritional value of the sulphate component of SSP. The value attributed to S will depend on the sulphur status of soils. Not all soils need S application. Application of PR for the first time may result in a yield penalty for the first crop. Thereafter, with applications on following crops there may be greater ongoing savings. However, the financial benefits are not particularly large for the farmers. Nevertheless, the use of local PRs can result in a substantial financial benefit for the country as a whole. This results from the saving of foreign currency as the local PRs act as import replacements. Therefore, there may need to be a subsidy to encourage farmers to use PRs.

Indonesia

Indonesia has long applied PR to plantation crops such as rubber, oil-palm, coconut, coffee, cocoa and tea. The estimated demand is 500 000 tonnes/year but the consumption is about 50 000-100 000 tonnes/year, with most of the PR coming from Jordan, Tunisia, Algeria, Morocco and Christmas Island.

Local PR resources consist of numerous cave deposits of phosphatic guano and phosphatized limestone. The Directorate of Mineral Resources estimated the total reserves at 700 000 tonnes (Harjanto, 1986). The quality of these PRs is very good with the total P₂O₅ concentration ranging from 28 to 39 percent and the FA-extractable P ranging from 54-80 percent of total P₂O₅. Therefore, these PRs are suitable for direct application.

A consortium studied the phosphate deposits in the Caimis and Tuban regions in West Java (BPPT-BRGM-CIRAD-TECHNIFERT-SPIE BATIGNOLLES, 1989). The consortium reported PR reserves of an estimated 3 million tonnes in the Ciamis region of West Java. In fact, there are many newly formed types of deposits that are solubilized and recrystalized in limestone. The phosphates also occur as impregnations of crandallite and whittokite in clay soils, which are erosion products of the same limestone. Samples from this survey have been used in pot and field experiments.

Performance of the Ciamis PRs

A pot experiment was conducted to compare three PRs (Malang, Bluri and Senori) from the Ciamis region with TSP as the reference fertilizer, all applied at 100 mg of P per kilogram of soil. The treatments included a control without P. The soil was a Podzolic soil from Singkut, Sumatra, with a pH of 4.5. There were five replications. Two test plants (Agrostis and soybean) were grown sequentially in pots containing 150 g of soil in order to measure the direct and residual effects respectively. The RAEs (RAE = [(yield from PR - yield from control)/(yield from TSP - yield from control)] x 100) of the three PRs were all in excess of 100 for the Agrostis yields while the Malang and Bluri PRs had RAEs in excess of 100 for the soybean yields. A separate incubation study was also performed with the same PRs in the same soil. The PRs increased available P to the same level as TSP. They also increased the exchangeable Ca and depressed the exchangeable Al to a greater extent than TSP. Thus, the Ciamis PRs were as effective as TSP in these studies.

A two-year field trial with five consecutive field crops was carried out on an abandoned Ultisol (pH 4.3) in South Kalimantan. The objective was to evaluate the performance of Ciamis PR against PR sources from North and West Africa, using a water-soluble local superphosphate (SP 36) as a reference fertilizer (Sri Adiningsih and Nassir, 2001). The P fertilizers were applied at a single rate of 300 kg of P₂O₅ per hectare and a control without P was included. The Ciamis PR and the highly reactive Gafsa PR from Tunisia were similar in terms of agronomic effectiveness to the water-soluble P fertilizer for the first crop and for the following four crops (Table 23). The performance of the other PRs improved with successive crops.

A simple economic analysis indicated that a large application of P fertilizer increased the net income for the farmer. The increase was US$1 050 for the water-soluble superphosphate, and US$1 264 for the ground PRs (Sri Adiningsih and Nassir, 2001). This demonstrates how the local Ciamis PRs can be used advantageously for food crop production, particularly in the reclaiming of degraded soils abandoned by the farmers.

A pilot plant was set up at Ciamis with a capacity of 10 000 tonnes/year. In 1999, the estimated production cost including mining, grinding, drying and bagging was US$50/tonne of commercial product containing 25 percent P₂O₅ (US$0.20/kg P₂O₅). The market prices of water-soluble phosphate (SP 36 with 36 percent P₂O₅) and imported PR (Gafsa with 31 percent P₂O₅) were US$171 and 114/tonne (US$0.47 and 0.38/kg P₂O₅), respectively. A joint venture has been agreed with a Japanese company to initiate exploitation of the mine. The demand is increasing but the problem is to ensure a regular supply of the commercial product.

TABLE 23
Relative agronomic effectiveness of PRs for crops on a Typic Hapludult, Pelaihari, Kalimantan

New Zealand

Pastoral farming based on ryegrass/white-clover permanent pastures constitutes about 90 percent of agriculture in New Zealand for the production of dairy, sheep, beef and deer products. Whereas the pasture plants rely mainly on the atmospheric N fixed by clover plants, P is applied as fertilizers. Permanent pastures are particularly suited for the use of PRs because: (i) they need a steady supply of P over a long period; (ii) they possess a high root density; and (iii) legumes (clover in this case) are efficient users of PR because of their affinity for Ca and the acidifying effect of N fixation on rhizosphere. In addition, most New Zealand soils are slightly acidic (pH 5-6) with adequate moisture regimes (more than 850 mm rainfall per year), both of which favour PR dissolution (Hedley and Bolan, 2003). About 10 percent of P fertilizers are applied as reactive PRs (about 142 000 tonnes of RPR), and PRs are reported to be gaining about 1 percent of the phosphate market each year (Quin and Scott, 2003). As PR application is allowed under ‘organic farming’ practices, there is an incentive to use PR in some situations.

The first series of trials to evaluate PRs as P fertilizers began in 1932 (Hedley and Bolan, 1997). Since the mid-1970s, systematic studies have included numerous field trials, greenhouse experiments and laboratory tests (Hedley and Bolan, 2003).

The results from the field trials show that there might be a delay before PRs begin to be effective, and it may take 4-6 annual applications before reactive PRs become as effective as TSP. In these trials, the PRs were surface applied and not incorporated into the soil. The time delay can be attributed to the time required for the PR to be incorporated into the soil through the effect of worm activity, rainfall, etc. and for PR-phosphorus to dissolve. Moreover, as the concentration of P around PR particles is about 4 mg/litre (Watkinson, 1994b), it will take longer than for SSP for the flux of P into the bulk soil for absorption of P by plants. Studies have shown that even after four annual applications of North Carolina PR, P concentration in solution does not increase significantly below 20 mm soil depth in a highly P retentive soil (Rajan, 2002).

The decline in absolute dry-matter production in the first few years could be substantial in soils of low P fertility. However, in soils of above-medium fertility, the decline may be less than 10 percent (Quin and Scott, 2003). This is because the increase in pasture production resulting from P fertilizer application can be expected to be small in these sites. In New Zealand, RPR is applied increasingly with water-soluble P in order to enhance its agronomic effectiveness (Chapter 9).

Currently, one kilogram of P in RPR costs US$0.87, whereas it costs US$1.07 as SSP if the soil does not require S. Thus, RPR phosphate is 20 percent cheaper. A promising strategy is to encourage application of RPR with soluble P, which is likely to result in the same level of agronomic effectiveness as the application of soluble P (Chapter 9). The savings on using RPR will be less if the value of S in the fertilizers is accounted for. Thus, the cost-effectiveness of RPR relative to soluble fertilizers will depend on the S requirement of soils in addition to environmental and management factors (Sinclair et al., 1993a; Rajan et al., 1996). A specific incentive for RPR use in New Zealand is its permissibility for use in organic farming.

Latin America

Direct application of PR has received much attention in Latin America in the past 20 years. Experiments have evaluated the agronomic and economic potential of indigenous PRs found in each country. The main objective has been to determine whether the local PRs could be used after grinding, or modified to produce PAPR, or subjected to high temperatures to produce thermophosphate, in order to reduce countries’ dependency on imported water-soluble P fertilizers (Casanova, 1995). The results have shown that PR use is advisable from some sources and under certain conditions (Casanova, 1998; Lopes, 1998; Besoain et al., 1999).

The savannahs located in the tropics and subtropics are the main agricultural frontier of Latin America. The soils, mainly Ultisols and Oxisols, are highly weathered, of low fertility, very acidic and with a high P-fixing capacity (Van Uexkull and Mutert, 1995). Therefore, the use of P fertilizer is important to maintaining and increasing agricultural productivity in these soils (Casanova, 1992). This situation, together with the large PR reserves (especially in Venezuela and Brazil) and the high costs of imported P fertilizers, has promoted a diversification in the production of P fertilizers using PR sources in these countries (Casanova, 1995). Several strategies for a more rational use of P fertilizers have been proposed. The following case studies on Venezuela and Brazil illustrate how local PRs can be modified and used effectively by farmers.

Venezuela

With reserves of 2 650 million tonnes, Venezuela has the third largest PR reserves in Latin America (after Mexico and Peru). The total P₂O₅ content ranges from 20 to 27 percent and the P₂O₅ solubility in 2-percent CA ranges from low to medium, using the solubility criteria reported by Hammond and Leon (1983).

Management guidelines for the direct application of ground or modified PRs have been developed over the last 20 years for annual and permanent crops for different soil and climate conditions (Casanova, 1992; Casanova et al., 1993). The P requirements for these crops, together with the methods and rates of application and their residual effects, have been determined using laboratory, greenhouse, field experiments and commercial evaluations of these P sources (Casanova et al., 2002b). This work has provided the basis for the opening of a fertilizer plant for producing granular PAPR with a capacity of 150 000 tonnes/year. The PR is partially acidulated with locally produced 60-75-percent sulphuric acid (Chapter 10).

A number of factors support the use of PAPR from such a plant in Venezuela. About 75 percent of the soils in the country are acid and the climate conditions are mainly tropical with an average rainfall of about 1 000 mm/year. There are 1 million ha of annual crops and 6.5 million ha of permanent crops (Comerma and Paredes, 1978).

A recent commercial evaluation of the agronomic and economic results from the use of PAPR in the rehabilitation of degraded pastures in Monagas State has found this approach to be very successful. The evaluation involved 30 farms using a dual-purpose milk and meat production system based on highly degraded Brachiaria brizantha pastures. The pasture degradation resulted from overgrazing and low fertilizer input on soils that were very deficient in N and P. The degraded pasture grew to heights of 20-25 cm and covered 30-40 percent of the soil surface. The dry-matter yields were 300 kg/ha/harvest and this forage contained 2-3 percent of crude protein and was of low digestibility. The average milk production per animal was 3-5 litres/d.

These degraded pastures were rehabilitated with a per-hectare application of 200 kg of PAPR and 100 kg of urea mixed with 3 kg/ha of Stylosanthes capitata seed at the beginning of the rainy season. Table 24 shows the results obtained with this approach. The crude-protein concentration of the pasture increased to 6 percent, the P, Ca, Mg, iron (Fe), copper (Cu), zinc (Zn) and manganese (Mn) concentrations increased to sufficiency levels, while per-animal milk production increased to 8 litres/day in the 100-animal herd. After one year, the animals had increased in weight from 350 kg on the traditionally managed degraded pastures to 470 kg for the PAPR-urea-Stylosanthes management system. The increased profits were an additional US$90/day from the milk production and US$15 600 in meat production for the 100-animal herd.

TABLE 24

Comparison between a rehabilitated pasture using the PAPR-urea-Stylosanthes capitata treatment and a traditional degraded pasture, Monagas State, Venezuela

Rehabilitating the degraded pasture by the traditional method using soil tillage, fertilization with soluble fertilizer, herbicides, planting new seed, and a delay until the pasture covered all the soil, cost US$462/ha compared with US$77/ha for the PAPR treatment (Table 24). Thus, all 33 farmers increased their pasture, milk, meat and profitability in a sustainable production system with the PAPR technology.

The social impact is that farmers now have phosphate fertilizer available in more than 175 fertilizer dealers across the country, at a competitive price of US$49/tonne compared with the highly water-soluble, imported TSP (US$148/tonne). Both fertilizers are similar in terms of agronomic and economic efficiency.

Brazil

Brazil has PR reserves of 376 million tonnes, which have a total P₂O₅ content range of 24-38 percent. However, the P₂O₅ solubility in 2-percent CA for the mainly igneous PRs is low (2.6-4.8 percent). Brazil uses about 4 600 tonnes of P per year (almost 3 percent of world consumption). The approaches taken in Brazil have been to either import water-soluble P fertilizers or reactive PRs, or treat indigenous PRs at high temperature in a mixture with basic slags in order to produce thermophosphate (Table 25).

TABLE 25
AEI of phosphate fertilizers in a clayey Oxisol of central Brazil, based on P uptake data over five years with annual crops, followed by three years with Andropogon pasture

*AEI% = [(crop yield using fertilizer (PR) - crop yield of control plot)/(crop yield of reference fertilizer (TSP) - crop yield of control plot)] x 100.

** Brazilian PRs.

A study by Lopes (1998) reports an experiment conducted on a clayey Oxisol with a pH of 4.2 that had received 2.2 tonnes of lime per hectare. The different P fertilizers, including local PRs (Araxa, Patos and Catalao), were evaluated at 200 and 800 kg of P₂O₅ per hectare and were broadcast in the first year. The results (Table 25) show that the Brazilian PRs have a low agronomic efficiency index (AEI). In five years of annual crops, the AEI was less than 50 percent for all PRs at both rates of P₂O₅, except for Patos PR at 800 kg/ha. With Andropogon grass, grown for three years after the annual crops, the AEI was considerable higher, except for Catalao PR. The high AEI for the Gafsa PR is a good indication that this product is an effective source of P in these acid soils.

Lopes (1998) also describes technologies, developed from a series of laboratory studies, glasshouse work and field experiments, that have brought millions of hectares of unproductive land into highly productive farmland. The best strategies developed to build up capital P in these soils are:

• In order to achieve adequate yields in four years, a P application of 1.5-2 kg/ha for each 1 percent of clay is broadcast as water-soluble P fertilizers, thermophosphate or highly reactive PR according to crop needs. This strategy was widely adopted and resulted in large areas of land being brought into crop production in the 1970s and 1980s. The strategy was combined with subsidies and adequate credit funding at low interest rates. The approach is in use today, without subsidies.

• In order to achieve adequate yields in six years, annual applications of mainly soluble P fertilizers, but also thermophosphate and highly reactive PR, are banded according to crops needs. In addition, a small excess (8-10 kg of P per hectare) is applied every year. The subsequent tillage operations and the residual effect of fertilizer P lead to a gradual build-up of P. Resource-poor farmers with insufficient funding for large amounts of P as a capital investment have tended to use this option, which is more consistent with the present agricultural policy in Brazil.

The study by Lopes (1998) points out that these management strategies, developed in the ‘Cerrado’ region of Brazil, have resulted in millions of hectares being developed for croplivestock production systems. It further remarks that these results demonstrate that the highly weathered, acid, nutrient-depleted soils of the tropics can be as productive as the better soils of the world with the use of this appropriate technology.