Valorization of the Natural Phosphate of Tilemsi (PNT) for the Development of the Agricultural Productions in Mali

In Mali, the deficiency of agricultural soils in phosphorus is one of the limiting factors in crop production. They respond well to phosphorus fertilization, but the high price of imported chemical fertilizers limits their use by the producers. However, Mali has large deposits of natural phosphate in the Tilemsi Valley (PNT) in northern Mali. Faced with this situation, the Government encouraged the use of local fertilizing resources including the PNT. Despite the satisfactory results of several years of research and popularization, the PNT was poorly adopted by the producers because of its powdery appearance, which made it difficult to apply in the field, its low solubility in the soil, which made its effect noticeable on crops in the second year of its application and its brown color reminiscent of that of earth made the producers believe that it had no fertilizing value. Faced with these problems relating to the physical appearance and the low solubility of the PNT, the laboratory work of GREAT QUEST Fertilizer SA resulted in the development of simple formulations based on enriched granular PNT 27% P2O5 (average content) and 35% P2O5 (high content) for direct application and NPK complex formulations based on granulated enriched PNT 35% P2O5 mixed mainly with urea, potassium, and sometimes with other nutriments like sulfur and boron (NPK 15-15-15; NPKSB 14-18-18-6-2) with a solubility in citric acid of 71.1%. The results of four years of experimentation of enriched granular PNT in the research stations and in the peasant environment showed a greater or equal effect than the other popular fertilizers. This project aims to improve phosphate nutrition and crop production by using granulated formulations of enriched PNT 27% P2O5 and 35% P2O5.


Introduction
Phosphates exist in two forms: the soluble form (superphosphates) and the insoluble form (natural phosphates).
Superphosphates are synthetic phosphorus fertilizers. They are mainly composed of hydrated mono-calcium phosphate [Ca (H2PO4) 2, H2O] and generally contain 43-50% of P 2 O 5 (Budavari, 1996). Superphosphates can be ordinary or triple depending on their P 2 O 5 composition. They are applied to the soil and used in soil fertilization as a source of phosphorus directly available to plants (Tisdale and Nelson, 1985). Natural phosphates (PN) are an inexpensive option for soil fertilization, particularly in countries where raw material is available (Truong et al., 1993). However, these PNs have a very low solubility. But some soil properties may affect the solubility of natural phosphates such as pH, reaction time or Ca ++ ion absorption capacity, Nahas (1996). Solubility also depends on the degree of fineness of the phosphate and its origin (GREAT QUEST, 2011). PNs of sedimentary origin may be more soluble than phosphates of woody or metamorphic origin (Hammond et al., 1986). Among the West African NPs, that of Tilemsi (PNT), in Mali, occupies a good place. It is one of the best natural phosphates and most receptive (Truong et al., 1978). PN deposits in the Tilemsi Valley have been detected since 1930 but, for several reasons, including differences in quality assessment, they have not been exploited (SONAREM, 1988). However, SONAREM, for agronomic testing purposes, produced a certain amount of PN between the 1960s and 1970s. These tests yielded some encouraging results and, starting in 1976, the production of PNT was undertaken. The deposit in operation is that of Tamaguilet, a residual plateau located in the valley of Tilemsi in northern Mali, some 120 km from Bourem. Its geographical coordinates are O ° 15 'East longitude and 17 ° 38' North latitude. The deposits of this PN are estimated at between 20 and 25 million tons, of which 2 to 5 million are located on the edges of the plateau and can be mined outdoors (Bathiono et al., 1997). PNT is rich in phosphorus. 100 kg PNT contain 25 to 30% P 2 O 5 , 42% calcium and also 0.0081% iron and aluminum oxides (Jenny, 1973). Following their work to characterize and classify tricalcic natural phosphates from West Africa, Truong et al. (1978) have shown that PNT has chemical and mineralogical characteristics that theoretically rank it among the best natural phosphates in Africa. Indeed, the PNT is the only one to have solubility in formic acid, higher than the limit of 55% retained by the European Commission for the direct employment in agriculture (Pieri, 1989). These same authors found that, among the phosphates of different solubility originating from Anecho in Togo, Kodjari in Burkina Faso, Tahoua in Niger, Tilemsi in Mali and Matam in Senegal, only the phosphates of Tahoua and Tilemsi are suitable for direct application. Both of these phosphates have solubility in citric acid of 3.2 to 5% (average solubility). For the PN to be able to serve as a source of phosphorus and to optimize the growth of plants, it is necessary that its conditions of use, in direct application or in acidulated form, favor its dissolution (Bationo et al., 1998). These conditions are physical, chemical and biological (Morel, 1996). Thus the following factors may affect the agronomic efficiency of phosphates: (1) The reactivity of phosphates, the chemical composition and the particle size determine their reactivity. Phosphates of sedimentary origin are generally more reactive and therefore better intended for direct application (PNT is that). The work of Lehr and McClellan (1972) and Chien (1977) have shown that the smaller the particle size, the greater the contact between phosphates and the soil and the higher the solubilization rate; (2) soil properties, according to FAO (2004) studies, for a phosphate to be agronomically effective, it must not only be dissolved but also available to plants. The properties of the soil that favor the dissolution of phosphates are the pH (less than 5.5), the low concentration of the Ca ++ ion in soil solution, the low soil level in P and the high content of organic matter. With respect to cation exchange capacity, recent studies by Perrott (2003) have suggested that a high level of exchangeable magnesium (Mg) in the soil solution can accelerate the dissolution of phosphates; (3) Climatic conditions, it was reported by Weil et al. (1994) that rain is the most important climatic factor influencing the dissolution of phosphates and their agronomic efficiency. Increasing the amount of soil water by rain or irrigation water increases the dissolution of phosphates. Hinsinger and Gilkes (1997) reported that the maximum agronomic effectiveness of phosphates with crops partially reflects the acidifying nature of soils and the high root density. The high root density facilitates the intensive exploitation of a larger soil volume for phosphorus because of the presence of a high number of fine roots per unit volume of soil; (4) Soil management practices and fertilizers, according to FAO (2004) studies, 4 management practices influence the agronomic efficiency of phosphates: the placement of phosphate relative to the plant, the dose of application, application period and liming. To maximize the agronomic efficiency of phosphates, they must be spread evenly over the soil surface and then incorporated at a depth of 10 to 15 cm. The incorporation of phosphates on the soil facilitates greater dissolution by improving contact between soil and phosphate particles. It also improves the uptake of P from the large volume of phosphorus-enriched soil and gives the roots greater chance of being in contact with the dissolved phosphate particles. According to FAO (2004), the average relative agronomic efficiency of phosphates is about 80%; (5) the activity of soil microorganisms, for phosphate rock to serve as a source of phosphorus and to optimize plant growth, conditions of use must stimulate microbial activity (Germida and Jansen, 1993). The texture and structure of the soil act on the microbial activity either directly or indirectly. Thus, in sufficiently moist sandy soil, there is a rapid spread of microbial activity while in clay soil, clay forms with organic substances organo-mineral complexes in which these substances become less accessible to microorganisms causing a slowdown in microbial activity (Morel, 1996). Each microbial species has specific pH limits between which it is active. A relatively low pH of the soil and mainly rhizosphere promotes the activity of microorganisms dissolving inorganic phosphates. Some researchers (Hedley et al., 1990, Doumbia et al., 1993, Toro et al., 1998 have shown that stimulation of microbial activity leads to a decrease of exchangeable calcium in the soil and favors the dissolution of natural phosphates. Regarding biological factors, the solubilization of natural phosphates must be a function of the microbial species and cultivars used Traoré (2010Traoré ( , 2012Traoré ( , 2014. The work of Flash et al., 1987 showed that sorghum mobilizes natural phosphates more efficiently than maize. According to Richardson (2001), some plants release into the rhizosphere organic acids capable of mobilizing phosphorus from soils rich in iron and aluminum phosphates. According to Boiffin and Sebillotte (1977), the incorporation of very high C/N residues into the soil can lead to phosphate deficiencies in plants by acting on the growth of soil microorganisms, which are the key to improving soil nutrients. efficiency of natural phosphates. Thus, the increasingly important role played by microorganisms in agriculture has led to their use in many parts of the world (Johri et al., 1999, Nautiyal, 1999. Komy (2005) has shown that the combination of the inoculation of a nitrogenfixing bacterium (Azospirillum spp.) With that of a solubilizing bacterium of inorganic phosphates significantly improves plant growth. Since 1977, the Institute of Rural Economy (IER) of Mali and its partners have conducted several studies on the agronomic value of the PNT in controlled conditions and in peasant environment in the different agroecological zones and cropping systems of Mali. The results of these studies showed that the agronomic efficiency of the NTP was highly dependent on rainfall. This work has shown that the PNT incorporated in the soil at the rate of 300 kg/ha has a maximum effect in the second year and minimum effects in the first and third years. SAFGRAD, (1985), Bagayoko and Coulibaly's (1995) research results indicated that the low efficacy of PNT was mainly related to its low solubility in soil. Thus research has been conducted to improve its physical appearance and solubility in soil by Sanogo et al., (1978); IFDC-IER (1982,1989); IER-FED (1986); LUXCONSULT SA (1985) and Samaké (1987). Despite the satisfactory results of research and extension efforts, the PNT has been poorly adopted by producers because of its powder form that made it difficult to apply in the field, its brown color reminiscent of that of the soil, suggesting that the PNT has no fertilizing value and its low solubility in direct application because its effect was observed only at the second year of its use. Faced with these problems related to the physical and chemical aspects of the PNT, GREAT QUEST Fertilizer SA, after having developped its metallurgical process to enrich the PNT that is based on operations of washing and screening the phosphate that reduce mainly the minerals oxides which consequently ameliorate the solubility and its granulation process that allows mixing phosphate with other nutriments in one granule , had interred in collaboration with the IER to implement a program of agronomical trials to use its granular formulations based on enriched natural phosphate. 27% P 2 O 5 enriched PNT (medium grade), 35% P 2 O 5 (high grade) for direct application and PNT-enriched 35% P 2 O 5 mixed with urea, potassium, sulfur and boron (NPK 15-15-15, NPKSB 14-18-18-6-2) as normal fertilizer. This project falls within the framework of a valorization of the PNT for the increase of the agricultural production in Mali.

Experimental sites:
The enriched granular PNT fertilizer formulations were tested for two (2) years in the seven (7) agronomic research stations and tested in a peasant environment for two (2) years in seven (  7)

Plant material:
it consisted of all the most commonly used popular varieties of cereals, cotton and seed legumes (peanut, cowpea) given in Table 1 below.  15 15. The bacterium M4 is a Bacillus subtilis which is a bacterial strain recognized for the production of the substances favoring growth and production of the plant.

Management of tests and tests:
The sowing of maize, millet, sorghum, was made on line with the aid of a graduated rope at a rate of 3 seeds per hill, at distances of 80 cm x 50 cm, 1 m x 1 m and 75 cm x 50 cm respectively, thinned at 2 plants per plant. Fertilization consisted on the base application of the enriched PNT and the popularized fertilizer according to Table 2. The urea supplement (150 kg/ha) was provided in two fractionations with 50 kg/ha at seeding and 100 kg/ha at steam elongation for maize, 50 kg/ha for sorghum and millet one month after the first application. 2 to 3 weedings at the intervals of 15 days. The 3rd weeding was done to recover the urea after application. Irrigated rice, lowland rice and submersion rice were transplanted on line using a graduated rope at a spacing of 20 cm x 20 cm for irrigated rice, lowland rice and submersion or 25 cm x 25 cm and transplanted using a clump of 3 to 4 plants per hole from a nursery of plants aged 15 to 21 days. The amount of seed for the nursery was 40 kg/ha. As for rainfed rice and wheat, sowing was done on line using a graduated rope at 25 cm x 25 cm spacing and at a rate of 3 to 4 seeds per hill. The quantity of seed was 50 kg/ha. For fertilization, phosphate fertilizers were used as bottom fertilizer. The urea supplement was provided at a rate of 200 kg/ha in 2 fractions: early tillering and panicle initiation for irrigated rice; 100 kg/ha in 2 fractions: early tillering and panicle initiation for rainfed and lowland rice: 150 kg/ha for wheat: early tillering and panicle initiation. The plots were maintained for the first weeding 15 days after transplanting and the others on demand for irrigated rice; 7 to 10 days after sowing and others on demand for lowland rice and rainfed rice. Seeding of cotton, peanut and cowpea was done in rows using a graduated rope at 80 cm x 30 cm spacings for cotton, 40 cm x 150 cm for peanut and 75 cm x 50 cm for cowpea with a 2-seedling per hill 15 days after sowing for cotton and cowpea and 1 seedling per hill for peanut. Phosphate fertilizers are applied as base fertilizer and the supplement urea 50 kg/ha was brought in early bloom cotton. The plots received 2 to 3 weeds at 15 day intervals after sowing.
NB: Enriched granular PNT, DAP, NPK and NPKSB formulations are provided as seedbed fertilizer (10-15 days after sowing); The system was to put the fertilizer in a furrow traced along and 5 cm from the crop line and then close.

Data Collection:
Soil samples from all IER agronomic research stations were collected the first year prior to planting for physico-chemical analysis by the Soil-Water-Plant Laboratory in Sotuba. The agronomic data collected included the following variables: Number of plants after thining, Vigor of plants (1)(2)(3)(4)(5), plants height (m), number of days to 50% bloom, number of plants harvested/plot, number of ears harvested/plot, number of pods/plant, ear aspect (length and size), pod or grain yield (kg/ha), effects related to biotic (diseases, insects, Striga) and abiotic (climate) factors.

Data Analysis:
The data collected were analyzed using the MSTATC statistical software using the analysis of variance method, and the ranking of averages with the Duncan test at the 5% threshold. To get an idea of the economic profitability of using the tested fertilizers, a partial economic analysis of the results was made using market prices for the concerned crops and the fertilizers used. The parameters considered for this analysis were: production or yield (kg/ha); the value of production = yield kg/ha x price (FCFA/kg of grain) of the crop; fertilizer cost/ha = CFAF/kg of fertilizer) x quantity of fertilizer used (kg/ha); net profit/ha = value of production (FCFA/ha) -fertilizer cost (FCFA/ha); difference in gain vs popular fertilizer = benefit from PNT -benefit from popular fertilizer.

Pysico-chemical analysis of the research stations soils
The results of soils analysis have shown that the IER stations and substations used in the different regions for agronomic trials have acid soils of pH 5.5 to 4. 8 or even lower in some cases. They are very poor in organic matter, in total nitrogen and in available phosphorus. Only cation exchange capacities (CEC) for exchangeable calcium (Ca), magnesium (Mg) and potassium (K) are at acceptable levels (Table 3). Thus, given the unfavorable conditions of these soils, the production of a food or industrial crop requires the application of phosphate fertilizers to correct phosphorus deficiencies and urea fertilizers to correct nitrogen deficiency. This clearly justifies the use of new formulations based on enriched PNT plus (+) complementary doses of urea.   (Figures 2 and 3). In general, the rainfall pattern in the area was characterized by a moderate frequency and intensity of rainfall with a marked deficit in May and June, a good distribution of rains from July to September with an early stop towards the end leading to premature drying of the plants ( Figure 2).

Effect of Fertilizer Formulations on Station Crops Across Regions
In the first year, enriched granular PNT formulations were tested on the different cereal crops (maize, millet, sorghum, rice (rainfed, lowland and irrigated), cotton and seeded legumes (peanut and cowpea). The statistical analysis of the yield data did not reveal any significant differences between the treatments for all the crops, and the yields did not vary significantly with the addition of sulfur and M4 bacteria to the trials in the first year ( Table 4). The yields obtained by fertilizer formulation are very close to the grand averages for all the crops, explaining a great similarity between the PNT formulations compared to the popularized ones with coefficients of variation ranging from 9.15% to 19.32%. On the other hand, the statistical analysis showed significant differences between the different sites indicating a local effect for the formulations and the crops. Regarding Table 5, the statistical analysis showed significant differences between fertilizer formulations for sorghum and millet. For sorghum, the best yields were obtained by NPK fertilizer 15 15

Sites & annees
formulations with 2850 kg/ha and a ppds of 457 kg/ha with coefficients of variation of 4.65% to 16, 30% (Table 5). Another statistical difference was observed for the rest of the cultures, however significant differences were observed between the sites or localities. The equal behavior of these formulations with that of the popularized fertilizers (DAP, NPK) and their superiority to the PNT powder in the majority of the trials and On-farm tests was an indication of their good performance under the Malian conditions. In the second year of experimentation, as the effect of sulfur and that of the bacterium M4 were not noticeable were removed from the experiment. Thus enriched granular PNT formulations were tested as in the previous year to determine their effect on the same main crops in Mali. Statistical analysis of yield data (Table 6), showed significant differences between fertilizer formulations for maize, sorghum and irrigated rice. For maize, the best yields were obtained by the popular fertilizer NPK 15 15 (Table 6). Another statistical difference was observed for the rest of the crops. Significant differences were observed between sites or localities ( Table 6) NS = not significant at the 5% threshold; CV = coefficient of variation; ** = highly significant; + = only one locality; ppds = smallest significant difference. NS = not significant at the 5% threshold; S = significant at the 5% threshold; CV = coefficient of variation; ** = highly significant; + = only one locality; ppds = smallest significant difference. NS = not significant at the 5% threshold; S = significant at the 5% threshold; CV = coefficient of variation; ** = highly significant; + = only one locality; ppds = smallest significant difference.
The Statistical analysis of the On-Farm yield data (Table 7) showed significant differences between fertilizer formulations for sorghum. The best yields were obtained by the popularized NPK 15 15 15, 100 kg/ha, 35% NPK, 100 kg/ha, 35% P 2 O 5 with respectively 1388 kg/ha and 1383 kg/ha; a smallest significant difference (ppds) of 127.41 kg/ha with coefficients of variation varying between 5% and 25.60%. Another statistical difference has been observed for other crops. But significant differences were observed between the different localities or extension services. The same is true for Table 8, which contains data collected on the legumes used (peanut and cowpea) with coefficients of variation of 8.95% and 9.14%.  The last year of the technology development focused on On-Farm demonstrations of the effect of enriched granular PNT formulations in comparison with popular fertilizers. The obtained results not presented here revealed a similarity of the effect of the enriched PNT with that of the popularized fertilizers for all the main crops used of Mali namely: cereals: maize, millet, sorghum, rice (irrigated, rain-fed, lowland and submersion), cotton and grain legumes (peanut and cowpea). Results from the first and the second years of field testing and demonstration confirmed those of the two years of agricultural research stations.

Discussion
The equality between the behavior of the enriched PNT formulations and that of the popularized fertilizers from the first year of use is a strong signal of the advantage of the use of the enriched PNT in the Kayes region compared to its powdered form which effect was noticeable only the second year of its application. This phenomenon has been illustrated by the similar results obtained on sorghum, cowpea and groundnuts in the agronomic research stations and in pre-extension trials in the peasant environments. Authors like Hisinger, (1997); Chien, showed that the large-scale use of enriched PNT is linked to its economic profitability and its competitiveness in the market. cropping seasons confirmed those obtained at the research station with a similar effect of the enriched PNT to that of the popular control fertilizers. Partial comparative and economic analysis carried out on the agronomic results made it possible to highlight the economic profitability for each type of crop for the benefit of the producers. Given the observed performance, the enriched PNT product could be the solution to the large-scale use of the phosphate rock in Mali and in the sub-region with the following advantages: (1) the enriched PNT improves agricultural production considerably; (2) contains calcium, about 40%, which has a soil-letting power to fight against acidity, unlike the popularized fertilizers which acidify it with the excess of H + ions; (3) maintains moisture in the soil longer than conventional fertilizer, resulting in crops ripening with 75% of green leaves compared to conventional fertilizers where crops retain only 25%. This has the advantage of increasing the producers' resilience to the adverse effects of climate change; (4) loosens the soil by facilitating maintenance work; (5) simple formulations of enriched PNT (27% P 2 O 5 and 35% P 2 O 5 ) can be used in organic agriculture because do not contain chemicals; (6) It is better suited for crop rotation systems; (7) the posted price of the enriched PNT is half that of the popularized fertilizers making it a competitive product on the Malian market. The enriched PNT can be an effective way to enrich the soils with phosphorus, knowing that the soils of Mali and the sub-region have a proven deficiency in this element.