Site-specific nutrient management (SSNM) provides an approach for "feeding" crops with nutrients as and when needed. The concept was first developed for irrigated rice in Asia and has been well documented at the SSNM web site of the Irrigated Rice Research Consortium including a complete list of publications (IRRI, 2007). Since its conception and development in the late 1990s (Dobermann and White, 1999; Dobermann et al., 2002), SSNM has gained popularity among rice farmers, researchers, and policy makers in Asia because it offers nutrient management options that are scientifically sound yet easy to develop, communicate, and apply by extension staff and farmers.

The principles of SSNM are generic and applicable to other crops. In the presentation below we apply the principles of SSNM developed for rice (Buresh and Witt, 2007; IRRI, 2007) to cereal crops in Asia. We then summarize some of the specific recommendations from an SSNM approach for tropical maize currently under development in a collaborative project between the IPNI-IPI Southeast Asia Program and research institutes in the region.

Generic principles of SSNM for cereal crops in Asia
In 2006, rice, maize, and wheat in Asia was harvested on about 281.5 million ha with a total cereal production of 1,050 million mt (FAO, 2007). The largest area cropped with these cereals is found in China (28%) and India (28%) with other countries each having less than 5% share of the total cereal area. The contribution of irrigated or favorable rain-fed environments to total production is significant. Irrigated rice systems provide 75% of the world's rice supply on only about one half of the world's rice land. Adequate fertilizers use in the favorable environments of Asia is therefore essential for global cereal production as yields in the absence of fertilizers are typically insufficient to meet food needs and achieve highest profit for farmers (Dawe, 2004; Moya et al., 2004; Buresh and Witt, 2008).

Existing fertilizer recommendations for rice, maize and wheat in Asia often consist of one predetermined rate of nitrogen (N), phosphorus (P), and potassium (K) for vast areas of crop production. Such recommendations assume that the need of a crop for nutrients is constant over time and over large areas. But the growth and needs of a crop for supplemental nutrients can vary greatly among fields, seasons, and years as a result of differences in crop-growing conditions, crop and soil management, and climate. Hence, the management of nutrients for cereals in Asia requires new approaches, which enables adjustments in applying N, P, and K to accommodate the field-specific needs of the crop for supplemental nutrients.

SSNM strives to enable farmers to adjust fertilizer use dynamically to make up the deficit in nutrient needs between that required by a high-yielding crop and nutrient supply from naturally occurring indigenous sources (i.e. soil, crop residues, manures, and irrigation water). The SSNM approach does not specifically aim either to reduce or to increase fertilizer use. Instead, it aims to apply nutrients at optimal rates and times in order to achieve high yield and high efficiency of nutrient use by the crop, leading to a high cash value of the harvest per unit of fertilizer invested.

Seven questions on SSNM

The following seven questions provide guidance in the development of SSNM recommendations (see also Fig. 1 for the experimental design of SSNM):

1. How large is the yield response (ΔY) to the application of fertilizer N, P, and K in an average season?

The yield response largely determines the total requirement for fertilizer N, P, and K to meet the crop's nutrient demand for a high yield at maximum economic return. Fertilizer N, P, and K are applied to supplement the nutrients from indigenous sources and achieve the yield target. The quantity of required fertilizer is determined by the deficit between the crop's total needs for nutrients - as determined by the yield target - and the supply of these nutrients from indigenous sources - as determined by the nutrient-limited yield.

The attainable yield is the grain yield for a crop grown in farmers' fields with good management practices and without nutrient limitation to yield.

The yield target is the average attainable yield across several seasons.

The nutrient limited yield is the grain yield for a crop not fertilized with the nutrient of interest but with good management and ample supply of all other nutrients from indigenous sources or fertilizer. The nutrient limited yield is an indirect measurement of the soil indigenous nutrient supply. The SSNM approach promotes the optimal use of available indigenous nutrients coming from the soil, organic amendments, crop residue, manure, and irrigation water.

High yield is a major factor contributing to high profit so that we often use the term yield target in our communication when we want to express the aim of high yield with maximum economic return.

However, yield can vary substantially in farmers' fields despite sufficient nutrient supply and good crop management from i) field-to-field, for example because of small scale variation in soil moisture, and ii) from season-to-season, for example, because of seasonal differences in climatic conditions.

Setting a yield target is only one component in developing a meaningful fertilizer recommendation, and possibly not the most important, because not all factors contributing to high yield can be controlled in every field and/or every season.

Yield responses to the application of fertilizer N, P, and K are highly variable among fields and/or seasons. The SSNM strategy for nitrogen with total N rate, split N applications, and dynamic N management using the leaf color chart (LCC) provide assurance that additional yield can be attained in years more favorable than the average.

Likewise, the SSNM strategy for P and K aims at achieving at least 1-2 mt/ha additional grain yield, if conditions for the year are favorable for markedly higher than average yields.

High, sustainable yield relates not only to the variable, short-term need for nutrients in a given season depending on the expected yield response to fertilizer application, but also the steady, long-term need for nutrient application particularly of P and K to avoid soil nutrient depletion.

2. What is the agronomic efficiency of fertilizer N (AEN) achieved in farmers' fields?

The agronomic efficiency is an indicator used for both estimating total fertilizer N needs and optimizing N management. AEN is the yield increase per unit fertilizer N applied. It is calculated as the attainable yield minus the nutrient limited yield divided by the amount of fertilizer N applied. The agronomic efficiency of nitrogen should be estimated experimentally in a few, representative field trials. Benchmark values for AEN exist for many rice, maize, and wheat growing environments. Low AEN compared to these benchmark values indicate either sub-optimal N management or yield limiting constraints other than N. Higher than benchmark AEN may indicate insufficient N supply to meet the crop's need for nitrogen to achieve high yield at maximum economic return.

3. What is the crop's nutrient need during the growing season?

The required fertilizer N is distributed in several applications during the crop growing season. This is particularly so in the tropics, to meet the crop's need for supplemental N. Fertilizer P and K are applied in sufficient amounts early in the season to overcome deficiencies and maintain soil fertility. Fertilizer K is often applied in two split applications early and near mid-season.

4. How variable is the yield response to fertilizer N, P, and K application among fields and within a few seasons?

Large variation in climate or spatial variability in soil nutrient supply determines the need for in-season or site-specific adjustments of nutrient management including the assessment of the crop's need for N using tools like the leaf color chart (LCC).

5. What is the nutrient removal with harvested products?

Grain yields are specific for location and season - depending upon climate, rice cultivar, and crop management. The amount of nutrients taken up by a crop is directly related to yield. The yield target therefore indicates the total amount of nutrients that must be taken up by the crop. Yield target and residue management therefore largely determine the nutrient removal with grain and/or straw and the need for maintenance application of P and K.

6. How does the attainable yield relate to the yield potential?

The yield potential is the theoretical maximum yield determined by germplasm and climate. Relating the attainable yield estimated in farmers' fields to potential yield provides an enhanced understanding of existing yield gaps and opportunities for improvement.

7. Are any other constraints to high yield expected?

High and profitable yield is only achieved when best management practices are followed including optimal planting densities and sufficient supply of all macro- and micro-nutrients either from indigenous or fertilizer sources.

SSNM for tropical maize in Southeast Asia
The following guidelines for the development of SSNM recommendations for maize are from a work in progress. Researchers have developed and successfully evaluated the concept in a collaborative regional project between the IPNI-IPI Southeast Asia Program and research institutes in Indonesia, the Philippines, and Vietnam in 2004-2007. Participatory evaluation of the presented approach is on-going with more than 300 farmers at 19 sites in these countries.

The following crucial information is needed to develop an SSNM recommendation for maize:

• An economic yield target attainable at optimal planting density in farmers' fields with good management.
• Actual yield responses to fertilizer N, P, and K.
• The expected agronomic efficiency of N.

Estimating total fertilizer requirements

The suggested total fertilizer requirements for N, P, and K are provided in Tables 1-3. Fertilizer N rates are estimated depending on the expected grain yield response to fertilizer N application and the expected agronomic N efficiency (Table 1). Note that the table is based on the assumption that the agronomic efficiency of fertilizer N is linked to the yield response to fertilizer N application that can be achieved depending on climate, bio-physical growing conditions, and management. Table 1 further provides guidelines for the timing and splitting of fertilizer N application including in-season adjustments using the standard 4-panel LCC of IRRI.

Fertilizer P and K requirements provided in Table 2 and Table 3 are estimated depending on an attainable target yield and the expected grain yield response to fertilizer application. The SSNM approach advocates sufficient use of fertilizer P and K to both overcome P and K deficiencies and avoid the mining of soil P and K. The determination of fertilizer P and K requirements for maize follow in essence an approach developed for rice (Witt and Dobermann, 2004), which maintains the scientific principles of the underlying QUEFTS model (Janssen et al., 1990; Witt et al., 1999).

Meeting the crop demand for nitrogen at critical growth stages

The demand of maize for N is strongly related to growth stage with a window for N application between crop establishment and tasseling stage (Fig. 2). In order to achieve high yield, maize plants require sufficient N in early growth to promote general shoot development, during the formation of kernel rows per ear beginning with the 5-leaf stage (V5). Likewise N is required at subsequent growth stages leading to the determination of kernels per row before tasseling (VT), and during ripening stages to enhance grain filling. The supply of N from soil and organic sources is seldom adequate for high yield, and supplemental N is typically essential for higher profit from maize fields. The SSNM approach enables farmers to apply fertilizer N in several, usually two to three doses to ensure the supply of sufficient N is synchronized with the crop need for N. An additional late N application before tasseling is recommended when high yields are expected or when N deficiency is observed as determined using a leaf color chart.

Optimize nutrient use efficiencies

Site-specific nutrient management in maize calls for flexible N management strategies that allow adjustments in N rates according to rainfall events and plant N demand using LCC. The LCC was developed for rice (Balasubramanian et al., 1999; Witt et al., 2005) and is also suitable for maize as indicated by spectral reflectance measurements performed on rice and maize leaves (Witt et al., 2004). Detailed experiments with several maize varieties conducted at the Cereals Research Institute in Maros, South Sulawesi, Indonesia, in 2005-2006 showed that yield losses of more than 20% can be expected when LCC readings consistently fall below the color of panel four (S. Saenong, personal comment). The LCC is now being evaluated together with farmers to fine-tune N management in participatory trials with maize. The amount of fertilizer N at critical growth stages is adjusted depending on leaf color which serves as an indicator of the plant N status.

Guidelines on LCC use in maize are provided in Table 1. The time for N fertilization is pre-set at critical growth stages with adjustments in rain-fed environments to ensure sufficient soil moisture. Farmers then adjust the dose of N upward or downward based on the leaf color. The effective management of N requires adequate planting densities, good crop management, and sufficient supply of P, K, and other macro- and micro-nutrients to achieve high and profitable yield.

Planting density
Yield is closely related to planting density if other factors, such as water and nutrients, are not limiting. Under tropical conditions, there are limitations to increasing planting density because of the low yield potential due to high temperatures (compared to temperate maize) and the increased susceptibility of the crop to pests and diseases when temperatures, rainfall, and humidity are high.

Most favorable planting densities for high yield in the tropics are probably in the range of 65,000 to 75,000 plants/ha with one seed per hole to achieve a uniform crop stand. A population of less than 65,000 plants/ha is not advisable because a 10% loss of plants is not uncommon under rain-fed field conditions and the planting density at harvest should be at least 60,000 plants/ha to achieve high yield. Planting more than 75,000 seeds/ha will not increase yield unless growing conditions are very favorable with a yield potential of >13 mt/ha. In drought-prone environments, planting density should not be more than 75,000 plants/ha.

The distance between rows should be narrow and just wide enough to allow field operations while plant spacing within the row should be wide to minimize plant competition for light, water, and nutrients. The optimal combination of row and within-row spacing should create a favorable microclimate in the canopy reducing the risk for pests and diseases. Fig. 3 shows the relationship between row and within-row spacing for different planting densities.

References

• Balasubramanian, V., Morales, A.C., Cruz,. R.T., and S. Abdulrachman. 1999. On-farm adaptation of knowledge-intensive nitrogen management technologies for rice systems. Nutr. Cycl. Agroecosyst. 53:59-69.
• Buresh, R.J., and C. Witt. 2007. Research Findings: III The principles of Site- Specific Nutrient Management [online]. Available at http://www.ipipotash.org/e-ifc/2006-10/research3.php (last update 2007.
• Buresh, R.J., and C. Witt. 2008. Balancing fertilizer use and profit in Asia's irrigated rice systems [in press]. Better Crops. 92(1).
• Dawe, D. 2004. Trends in sustainability and farm level productivity in intensive Asian rice-based cropping systems. In: Dobermann A, Witt C, and D. Dawe, editors. Increasing productivity of intensive rice systems through site-specific nutrient management. Enfield, NH (USA) and Los Baños (Philippines): Science Publishers, Inc., and International Rice Research Institute (IRRI).
• Dobermann, A., and P.F. White. 1999. Strategies for nutrient management in irrigated and rain-fed lowland rice systems. Nutr. Cycl. Agroecosyst. 53:1-18.
• Dobermann. A., Witt, C., and D. Dawe. 2002. Performance of site-specific nutrient management in intensive rice cropping systems in Asia. Better Crops Int. 16 (1):25-30.
• FAO. 2007. FAOSTAT [online]. In: www.fao.org. Available at http://faostat.fao.org (last update 2007; accessed 19 Nov. 2007). Rome, Italy: Food and Agriculture Organization (FAO).
• IRRI. 2007. Site-specific nutrient management helps rice farmers and the environment [online]. Available at http://www.irri.org/irrc/ssnmrice (last update 2007; accessed 05 Dec. 2007).
• Janssen, B.H., Guiking, F.C.T., van der Eijk, D., Smaling, E.M.A., Wolf, J., and H. van Reuler. 1990. A system for quantitative evaluation of the fertility of tropical soils (QUEFTS). Geoderma. 46:299-318.
• Moya, P.F., Dawe, D., Pabale, D., Tiongco, M., Chien, N.V., Devarajan, S., Djatiharti, A., Lai, N.X., Niyomvit, L., Ping, X.H., Redondo, G., and P. Wardana. 2004. The economics of intensively irrigated rice in Asia. In: Dobermann, A., Witt, C., and D. Dawe, editors. Increasing productivity of intensive rice systems through site-specific nutrient management. Enfield, NH (USA) and Los Baños (Philippines): Science Publishers, Inc., and International Rice Research Institute.
• Witt, C., and A. Dobermann. 2004. Towards a decision support system for site-specific nutrient management. In: Dobermann, A., Witt, C., and D. Dawe, editors. Increasing productivity of intensive rice systems through site-specific nutrient management. Enfield, NH (USA) and Los Baños (Philippines): Science Publishers, Inc., and International Rice Research Institute (IRRI).
• Witt, C., Dobermann, A., Abdulrachman, S., Gines, H.C., Wang, G.H., Nagarajan, R., Satawatananont, S., Son, T.T., Tan, P.S., Le Van Tiem, Simbahan, G.C., and D.C. Olk. 1999. Internal nutrient efficiencies in irrigated lowland rice of tropical and subtropical Asia. Field Crops Res. 63:113-138.
• Witt, C., Pasuquin, J.M.C.A., and R. Mutters. 2004. Spectral reflectance of rice and maize leaves and leaf color charts for N management [online]. In: New directions for a diverse planet: Proceedings of the 4th International Crop Science Congress Brisbane, Australia, 26 Sep - 1 Oct 2004. Available at www.cropscience.org.au/ icsc2004 (last update 2004; accessed 15 July 2005).
• Witt, C., Pasuquin, J.M.C.A., Mutters, R., and R.J. Buresh. 2005. New leaf color chart for effective nitrogen management in rice. Better Crops. 89 (1):36-39.