Kefyalew Girma^1*, Clinton Mack¹, Randy Taylor², John Solie², and William Raun¹

 

¹Department of Plant and Soil Sciences, Oklahoma State University, 368 Agricultural Hall, Stillwater, Oklahoma, 74078, USA

²Department of Biosystems and Agricultural Engineering, Oklahoma State University, 111 Agricultural Hall, Stillwater, Oklahoma, 74078, USA

* Corresponding author

 

Summary

Field-average based recommendations have been a common practice for nitrogen (N) rate recommendations. The problem is that even when the same amount of N fertilizer is applied at the recommended rate, yields differ from year to year, the result being poor use efficiency. Reduced fertilizer use efficiency can result in unnecessary costs to producers, especially with a likely price of US$ 1.10 per kilogram anticipated within the next few years, largely due to rising oil and natural gas prices. It has been estimated that by the year 2025 the consumption of N fertilizer will increase 60-90%, with two-thirds of this being applied in the developing world. In this light, methods that increase N use efficiency, farmer profitability, and that reduce environmental impact are no longer simply commendable, but required in developing countries. This review assesses the temporal variability in winter wheat with respect to N nutrition and presents improved  methods, including the calibration stamp, N-rich and ramped calibration strips to overcome this variability. These methods can be used either as visual tools or along with NDVI handheld sensors for accuracy. Unlike the calibration stamp and N-rich strip, the ramped calibration strip method is designed to include more pre-plant N rates in strips to improve top-dress N estimation. These methods have mainly been conceptualized and tested in the USA and Mexico. However, the simplicity of these technologies suggests that they can be readily applied by many farmers in both developed and developing countries.

 

Keywords: Wheat, Nitrogen, temporal Variability, calibration stamp, N-rich strip, ramped calibration strips

 

Introduction

 

Field-average based recommendations have been the common basis for N fertilizer recommendations in wheat (Triticum aestivum L.). The problem is that even when the same amount of N fertilizer is applied at the recommended rate, yields differ from year to year and as a result, poor use efficiency. Reduced fertilizer use efficiency can incur unnecessary costs to producers especially with a price of US$ 1.10 per kilogram anticipated within the next few years, largely due to rising oil and natural gas prices (Raun 2006a). It has been estimated that by the year 2025 the consumption of N fertilizer will increase 60-90%, with two-thirds of this being applied in the developing world (Ortiz-monasterio 2002). In this light, methods that increase N use efficiency, farmer profitability, and that reduce environmental impact are no longer simply commendable, but required in developing countries.

 

Usually, farmers apply large quantities of N every season, often in excess of crop demand. However, even if farmers attempt to apply the correct rate preplant, the fertilizer may not be efficiently used by the crop due to changes in growing conditions. Among these, weather factors in a given area, especially temperature and precipitation (Johnson and Raun 2003), determine N availability and uptake. In seasons when precipitation is sufficient, the crop is likely to lose N from the system through leaching. Several researchers assessed the relationship of N leaching to moisture and precipitation (Raun and Johnson 1999). High temperatures and humidity enhance volatilization of NH₃ from both the soil and the plant (Lees et al. 2000; Kanampiu et al. 1997; Francis et al. 1993). Lees et al. (2000) reported estimated plant N loses of up to 42 kg N/ha, and that losses increased with higher N applications. This suggests that the nutrient use efficiency of winter wheat is largely a function of temporal variability.

 

The uniform application of all or most of the N fertilizer preplant, as if the field has a homogenous fertility level and the same yield potential, leads to poor resource use and efficiency. Results of 30 years of continuous wheat research at the North Central Research Station at Lahoma, Oklahoma show that when fields were fertilized for the average maximum yield (33 g N per kg of wheat grain), 60% of the time predicted yield will be off by at least 10% from the actual yield (Figure 1). Additionally, because available non-fertilizer N varies from year to year, the average rate will be correct only 30% of the time (Figure 2); 37 % of the time levels will be below what is required by at least 22.5 kg N/ha; and 33% of the time application levels will be excessive (22.5 to 90 kg N/ha) (average = 43 kg N/ha). This shows the need to reevaluate the midseason fertilizer N rate every year (Girma et al. 2006).

 

New Methods to Manage Temporal Variability

Given the scope of the problem, in many parts of the world farmers have been trying to minimize preplant N rates and instead top-dress while the crop is growing. However, no method had been developed to help determine whether the crop is responsive to the top-dress rate. This review assesses the temporal variability in winter wheat with respect to N nutrition and presents some methods, including calibration stamp, N-rich and ramped calibration strips to overcome this variability while taking into account the N responsiveness of a field. This will help farmers decide whether to apply top-dress N and how much to apply. These methods have mainly been conceptualized and tested in developed countries. However, the type of technology used suggests that they can be readily applied by many farmers in developing countries.

 

The calibration stamp

The calibration stamp consists of a simple apparatus and procedure that can be easily used by farmers to determine midseason N rates based on visual observation, without an optical sensor or other instrumentation. The details of the engineering of the technology are presented elsewhere (Raun et al. 2005a). A calibration stamp has nine grids, each 1 m², whereby each of the four corners receives no fertilizer N (Figure 3). Rates of 22, 45, 67, 90, and 112 kg N/ha occupy the other five 1 m² areas within the 9 m² grid (termed as an N-rate calibration stamp). Calibration stamps should be applied preplant or soon thereafter and superimposed on top of the farmer’s practice. By midseason, differences between the 1 m² N rate areas can be visualized and a field-specific top-dress N rate prescribed by choosing the lowest N rate for which no visual differences were observed between it and the highest rate. Calibration stamps applied preplant or soon after planting can assist in providing a visual interpretation of N mineralization + atmospheric N deposition from planting to the time midseason N is applied, and improve determination of top-dress N rates (Raun et al. 2005). On-farm evaluation of this method showed that farmers can save as much as 99 kg/ha in N fertilizer and gain US$ 44 kg/ha. The only limitation to the calibration stamp was at times it was difficult to visualize differences, and poor representation of the field due to the small size of the calibration stamp. To over come this, the N-rich strip was proposed.

 

The N-rich strip

The N-rich strip program contains a few simple steps. Farmers need to set up 4-5 strips along the length of the field, each receiving different N rates (determined based on yield potential, Table 1) applied preplant or soon after planting using granular or liquid N fertilizer. Then, the farmer waits until the winter wheat crop reaches Feekes growth stage 5-7 or some time between February and March and subsequently evaluates, either visually or with a handheld sensor (Figure 4), whether any of the strips are visually different from the rest of the field (biomass and color). In the former case, the lowest rate for which the difference was apparent should be applied. If no difference can be observed between the N-rich strip and conventional farmer’s practice, it is unlikely that there will be a benefit from midseason fertilizer N.  

 

In the latter, the N recommendation is based on in-season predicted yield potential and the N responsiveness of the field. In-season yield potential can be estimated using the Normalized Difference Vegetation Index (NDVI = (near infrared reflectance – red reflectance)/ (near infrared reflectance + red reflectance)) readings (value output from the sensors) from the N-rich strip compared to the farmer practice, and knowing the growing degree days (GDD) determined by days from planting to sensing (Lukina et al. 2001). For readings collected between January and March (regardless of when the wheat was planted), it is possible to estimate the obtainable wheat grain yield with and without additional fertilizer N. Independent of the biomass produced per day, the N responsiveness or response index, RI (Mullen et al. 2003; Johnson et al. 2003) of the field is determined as the ratio of the NDVI of the N-rich plot and the NDVI of the farmer plot (NDVI_N-rich/NDVI_farmer). Detailed equations are presented in Raun (2006b). These values are used to accurately predict both the yield and the need for additional N. With the handheld sensor, a farmer can make a more accurate decision than solely using visual evidence because the sensor can accurately predict the yield potential of a field and its responsiveness to N fertilizer (Hodgen et al. 2005; Ball and Mosali 2005).

 

In 2005, 10 N-rich strip and calibration stamp tests were conducted in Oklahoma. A comparison of N-rich strip and calibration stamp methods is presented in Figure 5 for selected fields. Accordingly, the sensor based N-rich N recommendation consistently recommended the lowest rate of N in all fields. This method resulted in improved nitrogen use efficiency (NUE) and an additional profit of US$ 25 per hectare to farmers (data not shown).

 

Table 2 shows top-dress N rates recommended for three locations in Oklahoma using NDVI measurement. The results show how powerful this decision support tool is in identifying the correct rate for each field. The only issue with the N-rich strip was the small number of N-rates, which makes it difficult for farmers to interpolate intermediate optimum N rates. To unravel this problem the ramped calibration strips was introduced.  

 

The ramped calibration strips

The ramped calibration strips applicator (Figure 6) applies up to 16 different N rates (3 m or 6 m intervals) over 48 to 96 m (actual rates and distances can be adjusted depending on the crop). Figure 7 shows 15 rates of N applied preplant using the ramp applicator at Lake Carl Blackwell, Oklahoma. The ramped calibration strips are used in the same way as N-rich strips to prescribe N, either visually or with handheld NDVI sensors. It is important to note that the ramped calibration strips are applied on top of the farmer practice. Assuming that we can catch up and/or achieve maximum yields from the midseason N application, and assuming that yield potentials are not severely restricted by excess early season N stress, the ramp interpolated rate is the quantity a farmer would need to apply on the rest of the field to achieve the same visual or sensor-based response ( Raun 2006a). Without the sensor, farmers can simply walk from the part of the ramp that received the lowest N rate to the other end and stop where they no longer see any differences between that rate and the next higher rate (biomass or intensity of green).

 

If farmers are using the sensors to determine their top-dress N rates, they need to mark the start and end of the ramp N-reference strip (preplant), and collect sensor data using the handheld sensor, walking at a constant speed over the length of the ramp. Farmers can take the NDVI measurements from ramped calibration strips, automatically read the file, and compute the optimum N rate. We recommend the use of the sensor simply because the sensor can pick up differences that our eyes cannot see. Whether determined visually or with a handheld sensor, the point where no differences exist between the rate at which sensing is being done and the immediate next rate is the top-dress N rate. The ramped calibration strips will allow producers to refine midseason fertilizer N rates by knowing exactly what the RI is and by integrating the area under the curve over the distance included within the ramp (currently set at 46 m, but it can be altered to increase the rate change intervals).

 

Summary

The technologies discussed can help refine top-dress N rates by managing temporal variability. These simple temporal variability-based N recommendation tools avoid the need to run soil analyses and account for any in-season N available from rainfall, mineralization, legume rotation, and manure application. These methods have been mainly conceptualized and tested in the USA and Mexico. However, the type of technology used suggests that they can be readily applied by many farmers in both developed and developing countries especially in large-scale production and high yielding environments. Currently they are being evaluated in Argentina, Australia, China, India, and Turkey.

 

These technologies can be used as decision tools, either visually or using handheld NDVI sensors. The cost of the sensor will have an impact as to which farmers will have access to the technology. The sensor has been evaluated mostly by large-scale farmers in the developing world. These tools do not address spatial variability, which is as important as temporal variability. To overcome this, a Variable Rate Applicator (VRT) was developed especially for high yielding environments (Raun et al. 2004, 2005b; Stone et al. 2005). This technology has several components that are well documented in the literature (Lukina et al. 2001; Raun et al. 2002).   

 

References

BALL, J. & MOSALI, J. (2005). Nitrogen-rich strips - A new way to determine N

            fertilizer requirements. In Ag News & Views. Ardmore, OK : The Samuel Roberts

            Noble Foundation, Inc. [online] http://www.noble.org/Ag/Soils/N-RichStrips.

FRANCIS, D.D., SCHEPERS, J.S. & VIGIL, M.F.(1993). Post-anthesis nitrogen loss from corn. Agronomy Journal 85:659-63.

GIRMA,  K.,  MARTIN, K.L., ANDERSON, R.H., ARNALL, D.B., BRIXEY, K.   D., CASILLAS, M.A., CHUNG, B., DOBEY, B.C., KAMENIDOU, S.K.,

            KARIUKI, S.K., KATSALIROU, E.E., MORRIS, J. C., MOSS, J.Q.,

            ROHLA, C.T., SUDBURY, B.J., TUBANA,  B.S. & RAUN, W.R. (2006).

            Mid-season prediction of wheat-grain yield potential using plant, soil, and sensor

            measurements. Journal of Plant Nutrition 29:873–897.

HODGEN, P.J., RAUN,W.R., JOHNSON, G.V., TEAL, R.K., FREEMAN, K.W., K.B.,

            BRIXEY, K.L., MARTIN, J.B., SOLIE, AND STONE, M.L. (2005). Relationship     between response indices measured in-season and at harvest in winter wheat.

            Journal of Plant Nutrition 28:221–235.

JOHNSON, G.V. & RAUN, W.R. (2003). Nitrogen response index as a guide to fertilizer management. Journal of Plant Nutrition 26: 249-62.

KANAMPIU, F.K., RAUN,W.R. & JOHNSON, G.V. (1997). Effect of nitrogen rate on plant nitrogen loss in winter wheat varieties. Journal of Plant Nutrition 20:389-404.

LEES, H.L., RAUN,W.R. & JOHNSON, G.V. (2000). Increased plant N loss with increasing nitrogen applied in winter wheat observed with 15N. Journal of Plant Nutrition 23:219-30.

LUKINA, E.V., FREEMAN, K.W., WYNN, K.J., THOMASON, W.E., MULLEN, R.W., KLATT,A.R., JOHNSON, G.V., ELLIOTT, R.L., STONE, M.L., SOLIE, J.B. & RAUN, W.R. (2001). Nitrogen fertilization optimization algorithm based on in-season estimates of yield and plant nitrogen uptake. Journal of Plant Nutrition 24:885-98.

MULLEN, R.W., FREEMAN, K.W. RAUN, W.R., JOHNSON, G.V., STONE, M.L. & SOLIE, J.B. (2003). Identifying an in-season response index and midseason fertilizer N. Agronomy Journal 95:347-51.

ORTIZ-MONASTERIO, J.I. (2002). Nitrogen management in irrigated spring wheat.

In Bread Wheat Improvement and Production. FAO Plant Production and Protection Series No. 30 (Eds. B. Curtis, S. Rajaram & H. Gomez Macpherson), pp. 433-452. Rome, Italy: FAO.

RAUN, W.R. (2006a). Predicting the potential response to applied N using N-rich

strips or the new ramped calibration strips. [Online]   http://nue.okstate.edu/Index_RI.htm

RAUN, W.R. (2006b). Library of yield prediction equations. [Online] http://nue.okstate.edu/Yield_Potential.htm.

RAUN, W.R. & JOHNSON, G.V. (1999). Improving nitrogen use efficiency for cereal

            production. Agronomy Journal 91:357-363.

RAUN, W.R., JOHNSON, G.V., SOLIE, J.B. & STONE, M.L. (2004). A process for in-season fertilizer nitrogen application based on predicted yield potential. U.S. Patent. 10,195,138. (CIP of U.S. Patent 6,601,341).

RAUN, W.R., SOLIE, J.B., STONE, M.L., ZAVODNY, D.L., MARTIN, K.L. & FREEMAN, K.W. (2005a). Automated calibration stamp technology for improved in-season nitrogen fertilization. Agronomy Journal 97:338-42.

 

RAUN, W.R., JOHNSON, G.V., SOLIE, J.B. & STONE, M.L. (2005b). A process for in-season fertilizer nitrogen application based on predicted yield potential. U.S. Patent. 6,880,291.

RAUN, W.R., SOLIE, J.B., JOHNSON, G.V., STONE, M.L., MULLEN, R.W., FREEMAN, K.W., THOMASON, W.E. & LUKINA, E.V. (2002). Improving nitrogen use efficiency in cereal grain production with optical sensing and variable rate application. Agronomy Journal 94:815-20.

STONE, M.L., NEEDHAM, D., SOLIE, J.B., RAUN, W.R. & JOHNSON, G.V. (2005). Optical spectral reflectance sensor and controller. U.S. Patent. 6,855,933, Cont. of 6,596,996.