INTRODUCTION
Nitrogen (N) remains the most limiting and crucial plant nutrient in crop production since the establishment of its essentiality in 1872 by G. K. Rutherford, a chemist from Scotland (Fageria et al., Reference Fageria, Baligar and Jones1997). World food production has increased since 1950s with the introduction and use of inorganic fertilizers, especially N (Fageria and Baligar, Reference Fageria, Baligar and Rengel2003; Follett, Reference Follett2001). However, excess application of N has been demonstrated to contribute to average leaching losses between 25% and 50% of N applied in some cropping systems (McNeal and Pratt, Reference McNeal, Pratt and Pratt1978), which ends up in surface and ground water.
High levels of nutrients, especially nitrates and phosphates, in water bodies have led to eutrophication, contributing to excess plant growth, which depletes oxygen in the water. Other pathways where N is lost include crop harvest (Foth and Ellis, Reference Foth and Ellis1988; Kuangfei et al., Reference Kuangfei, Yaling, Xuefeng and Pastore1999), plant loss (Harper and Sharp, Reference Harper and Sharp1995; Kanampiu et al., Reference Kanampiu, Raun and Johnson1997), volatilization (Xing and Zhu, Reference Xing and Zhu2000), denitrification (Davidson, Reference Davidson1992; De Klein and Van Logtestijn, Reference De Klein and Van Logtestijn1994) and surface runoff (Yu-Hua et al., Reference Yu-Hua, Bin, Lin-Zhang, Shi-Xue and Zhao-Liang2007).
Corn only takes up limited amounts of N prior to the four-leaf stage and starts accumulating substantial N 40 days after emergence (Sawyer et al., Reference Sawyer, Nafziger, Randall, Bundy, Rehm and Joern2006). In maize, this corresponds to V8 to V10 maize growth stages. At V8 to V10, macro-nutrient and micronutrient deficiencies appear the most and maize begins a steady and rapid increase in both nutrients and dry matter accumulation.
Nitrate N (NO3−-N) is the most mobile N source in the soil and mainly taken up by crops through mass flow and diffusion (Barber, Reference Barber1995). Available soil moisture and diffusion potential plays a great role in NO3−-N mobility and subsequent uptake by crops. Consequently, high moisture and diffusion potentials can result in an increase in N movement in the soil (NaNagara et al., Reference NaNagara, Philips and Leggett1975). Available soil moisture and how far N fertilizer is placed away from the row will be the determining factors for crop N uptake. For example, where moisture is a limiting factor in arid to semi-arid regions, midseason N application (V8 to V10) needs to be synchronized with correct application distance to enhance uptake of applied N.
Low moisture regimes in some environments can hinder N movement in the soil. Inadequate soil moisture can also affect physiological development of maize by hindering the development of the roots system (Shoup and Janssen, Reference Shoup and Janssen2009). Roots system plays a crucial role in the uptake of water and mineral nutrients by plants (Gregory, Reference Gregory2006). In corn, the fine roots are the major sites of water and nutrients uptake into mature root systems (Varney and Canny, Reference Varney and Canny1993), and these are concentrated at the basal region of a growing maize plant. As a result, through fine branch roots, substantial nitrate uptakes have been found in basal regions of axile roots of mature plants in the soil (Reidenbach and Horst, Reference Reidenbach and Horst1997). These findings suggest that planting closer to nitrogen bands can enhance the uptake of N by the crops with a poorly developed roots system.
Studies conducted by Edmonds (Reference Edmonds2007) on midseason application of N applied to every other row indicated that rows that did not have midseason N application had lower yields and did not benefit from midseason N application of the adjacent row. This finding demonstrated that in an environment where moisture is limiting, mass flow might not be enough to move midseason applied N to great distances in a single growing season (76 cm distance between rows). Elsewhere, Vyn and West (Reference Vyn and West2008) established that planting maize using real-time kinematics (RTK) 13 cm away from the pre-plant band of urea ammonia nitrate (UAN) increased yields. Their study showed that maize planted directly over the 10 cm deep band had a higher N concentration, although in most cases with lower yields. Low yields could have been due to local toxicity from direct contact between the seed and fertilizer at planting. These findings identified the need for continued work in N placement distance from the maize row and how it impacts maize grain yield and nitrogen use efficiency (NUE).
OBJECTIVES
The objective of these field experiments was to evaluate the effect of side dress N fertilizer application rates (0, 45, 90 and 134 kg N ha−1) and application distance from the maize row (0, 10, 20 and 30 cm) on maize grain yield and NUE.
MATERIALS AND METHODS
Site description
Field experiments were established in 2009 and 2010 at Haskell, Hennessey, Efaw and Lake Carl Blackwell (LCB), Oklahoma, USA. The soil at Haskell is a Taloka silt loam (fine, mixed, active, thermic Mollic Albaqualfs) and at Hennessey, it is a Bethany silt loam (fine, mixed, superactive, thermic Pachic Paleustolls). The LCB site is predominantly characterized by Pulaski fine sandy loam (coarse-loamy, mixed, superactive, non-acid, thermic Udic Ustifluvent) and Port silt loam (fine-silty, mixed, superactive, thermic Cumulic Haplustolls). The Efaw experimental site is composed of mainly Pulaski fine sandy loam (coarse-loamy, mixed, superactive, non-acid, thermic Udic Ustifluvents) and Grainola silty clay loam (fine, mixed, active, thermic Udertic Haplustalfs). The climatic conditions for each study site during 2009 and 2010 cropping seasons are presented in Figures 1 and 2.
Figure 1. Rainfall and temperature distribution at Haskell, Hennessey and LCB, Oklahoma, 2009.
Figure 2. Rainfall and temperature distribution at LCB and Efaw, Oklahoma, 2010.
Experiment and management
Before planting each experiment, soil samples (0–15 cm) were collected for site characterization. From each plot 15 cores were collected, mixed well and a composite sample obtained. Samples were dried and passed through a 2 mm sieve. Soil pH was determined by adding 15 ml of water (H2O) into 15 g of soil, shaken well and left to equilibrate for 1 hour. The reading was then taken using a pH meter. Nitrate N was determined by adding 25 ml of calcium sulfate into 10 g soil. The suspension was then shaken for 30 minutes, filtered and analysed on flow injection analyser using cadmium reduction chemistry. Ammonium N was determine by adding 15 ml of 2 N KCL solution into already weighed 1.5 g of soil, shaken for 15 minutes on oscillating shaker and filtered immediately. Into a glass tube, 5 ml of the filtrate was pipetted and analysed by flow injection for ammonium N. Total N and percentage carbon was determined using dry combustion LECO analyser. The available potassium (K) and phosphorus (P) was analysed by adding 20 ml Mehlich-3 into 2 g, 2 mm dry sieved soil, shaken for 5 minutes, filtered and analysed on Inductively Coupled Plasma (ICP). The results for the analysis are presented in Table 1.
Table 1. Soil chemical properties determined from initial soil samples (0–15 cm) at four locations, Oklahoma.
PH: 1:1 soil:water; K and P – Mehlich-3; NH4-N and NO3–-N M KCL, total N and organic C-dry combustion.
*Data were not determined.
Throughout, a randomized complete block design with three replications was used. Each plot size was 20 m long and four rows wide with row spacing of 76 cm. Depending on the maize variety used, planting rate varied for each site as indicated in Table 3. Treatments for each plot for all locations and years were administered according to the treatment structure (Table 2). Planting was done using a John Deere Maxemerge 2, four-row vacuum planter. Experiments were rain-fed but experiments at the LCB site were irrigated (15 cm depth of water twice a week) when needed to supplement rainfall. Additional information about the experiments is presented in Table 3.
Table 2. Treatment structure and description of the trials conducted at Haskell, Hennessey, Lake Carl Blackwell and Efaw, Oklahoma, 2009–2010.
*Midseason N was applied between V8 and V10 maize growth stages.
†A stream of urea ammonium nitrate (UAN) was applied at varying distances away from the maize row.
Table 3. Field trial information for all the sites, 2009–2010.
At planting, treatments 2 through 5 did not receive any prep-plant N (Table 2). The objective of these four treatments was to establish how varying placement distance of side dress N away from the maize row was going to affect grain yields and NUE, only at low N rate (45 kg N ha−1) and with 0 kg N ha−1 prep-plant. Past researches have established that split application of N fertilizer leads to increased NUE (Martin et al., Reference Martin, Loudon, Ritchie and Werner1994; Ritter et al., Reference Ritter, Scarborough and Chirnside1993; Westermann and Crothers, Reference Westermann and Crothers1993). Therefore, treatments 6 through 17 received pre-plant UAN (28% N) fertilizer at 45 kg N ha−1 and liquid UAN side dress N fertilizer applied to the soil surface in a continuous stream at rates of 45, 90 and 134 kg N ha−1 at 0, 10, 20 and 30 cm placement distance away from the maize row.
At maturity, the two middle rows of each plot were harvested using a Massey Ferguson 8XP self-propelled, which was equipped with an automated weighing system (HarvestMaster Inc, 1994). Grain yields for all the treatments were calculated and expressed at 15.5% moisture levels.
Data management and analysis
Grain yield data obtained from all locations and years were separately analysed using the GLM procedure using SAS version 9.1 (SAS Institute, 2003) to determine treatment effects. Means were separated using Fishers-protected LSD and non-orthogonal, single-degree-of-freedom contrasts. NUE of the applied nutrients was calculated using the difference method (Varvel and Peterson, Reference Varvel and Peterson1990) described below:
Percentage NUE (% NUE) = ((NF) – (NC)/R) × 100,
where
NF = total N uptake in maize grain yield from N fertilized plots;
NC = total N uptake in maize grain yield from unfertilized plots;
R = rate of fertilizer N applied.
RESULTS
Grain Yield
Grain yields were obtained for LCB in 2009 and 2010. In 2009 results indicated that N rate, application distance away from maize row (with or without prep-plant) and interaction between the two factors did not significantly (p < 0.05) affect maize grain yield (Table 4). In addition, treatments with 0 kg N ha−1 prep-plant and 45 kg N ha−1 side dress N did not record significantly (p < 0.05) different maize grain yields from those that received 45 kg N ha−1 prep-plant and 45, 90 or 134 kg N ha−1 side dress N. However, better maize grain yield was obtained when 45 kg N ha−1 prep-plant and 90 or 134 kg N ha−1 side dress was applied. Varying side dress N application distance of low side dress N rate (45 kg N ha−1) showed inconsistent results with prep-plant N (45 kg N ha−1) applied (Figure 3). Applying 90 and 134 kg N ha−1 side dress N at 0 and 10 cm away from the row resulted to high maize grain yield ranging from 1816 to 2265 kg ha−1 (Figure 3 and Table 4). Further increase of the application distance away from the maize row led to a decline in maize grain yield.
Table 4. Grain yield means as affected by placement distance away from the maize row of side dress N application in 2009 and 2010 at Haskell, Hennessey, Efaw and Lake Carl Blackwell, Oklahoma.
†, *, **, ***: Significant at the 0.1, 0.05, 0.01 and 0.001 probability levels respectively.
§SED: Standard error of the difference between two equally replicated means.
Figure 3. Grain yield as affected by side dress N at 0, 45, 90 and 134 kg N ha−1 applied at 0, 10, 20 and 30 cm away from the maize row, LCB, Oklahoma, 2009.
In 2010, there was a significant (p < 0.0001) effect of treatment on maize grain yield (Table 4). Treatments 2 through 17 resulted to higher maize grain yield, compared with that of the control (treatment 1). Treatments that received 0 kg N ha−1 prep-plant and 45 kg N ha−1 side dress N resulted to low maize grain yield, compared with those that received prep-plant N (45 kg N ha−1). At 0 kg N ha−1 prep-plant and 45 kg N ha−1 side dress N, maize grain yields consistently reduced with increase in side dress N application distance away from the maize row (Table 4). The maize grain yield ranged from 4872 kg ha−1 (0 cm application distance) to 2554 kg ha−1(30 cm application distance). Treatments that received 45 kg N ha−1 prep-plant showed an overall linear increase in maize grain yields with applied side dress N (Figure 4). Application of 45 kg N ha−1 prep-plant followed by 90 or 134 kg N ha−1 side dress N, applied 10 cm away from the maize row, gave the highest maize grain yields of 8221 kg ha−1 and 8038 kg ha−1 respectively (Table 4). But generally, varying the application distance did not contribute to any significant differences in maize grain yields (p < 0.05) regardless of side dress N rate applied. The interaction between the two factors was also not significant (p > 0.05).
Figure 4. Grain yield as affected by side dress N at 0, 45, 90 and 134 kg N ha−1 applied at 0, 10, 20 and 30 cm away from the maize row, LCB, Oklahoma, 2010.
At Haskell, in 2009, the results obtained indicated that there was no significant effect of treatment on maize grain yield (Table 4). Treatments with 0 kg N ha−1 prep-plant and 45 kg N ha−1 side dress N did not significantly differ from the rest of the treatments. Overall poor N response was recorded. Application of side dress N, application distance away from the maize raw and interaction between the two factors did not contribute to any significant (p < 0.05) maize grain yield differences among treatments. However, it was apparent that at low N rate (45 kg N ha−1) maize grain yield increased with close application of side dress N (Figure 5). Maize grain yield declined to as low as 4705 kg ha−1 when side dress N was applied 30 cm away from the maize row (Table 4). At high N rates side dress N, application distance away from the maize row showed inconsistent trend in maize grain yields (Figure 5).
Figure 5. Grain yield as affected by side dress N at 0, 45, 90 and 134 kg N ha−1 applied at 0, 10, 20 and 30 cm away from the maize row, Haskell, Oklahoma, 2009.
The experiment at Hennessey in 2009 gave the lowest maize grain yields compared with all the cropping seasons and locations. Overall, treatments were not significant (p < 0.05) (Table 4) and poor response to side dress N application was recorded (Figure 6). Maize grain yields obtained from treatments that received prep-plant N and those that did not were not significantly (p < 0.05) different from each other. Varying the application distance away from the maize row did not significantly (p < 0.05) influence maize grain yields even for the treatments that did not receive prep-plant N (treatments 2 through 5). However, at low side dress N rate (45 kg N ha−1), maize grain yield reduced with increase in the application distance when prep-plant N was applied (treatments 6 through 9). Applying side dress N at 0 and 10 cm away from the maize row resulted to better maize grain yield (Figure 6). At low side dress N, 1435 kg ha−1 and 1637 kg ha−1 was recorded when side dress N was applied 0 and 10 cm away from the maize row respectively (Table 4). High N rates (90 and 134 kg N ha−1) led to inconsistent results regardless of how far the side dress N was applied from the maize row.
Figure 6. Grain yield as affected by side dress N at 0, 45, 90 and 134 kg N ha−1 applied at 0, 10, 20 and 30 cm away from the maize row, Hennessey, Oklahoma, 2009.
At Efaw, treatments significantly (p < 0.01) affected maize grain yield, contributing to higher maize grain yields compared with that of the control. Treatments that received prep-plant N gave significantly higher yields, compared with treatments 2 through 5 (no prep-plant N applied). A positive response to side dress N application was recorded (Figure 7). However, distance the side dress N was applied away from the maize row did not significantly (p < 0.05) affect maize grain yields even when low side dress N was applied with no prep-plant application. Higher N rates (90 and 134 kg N ha−1) gave better results when side dress N was applied closer to maize row. Maize grain yield was the highest when side dress N was placed 0 to 10 cm for 90 kg N ha−1 and 0 cm for 134 kg N ha−1 side dress N application (Table 4).
Figure 7. Grain yield as affected by side dress N at 0, 45, 90 and 134 kg N ha−1 applied at 0, 10, 20 and 30 cm away from the maize row, Efaw, Oklahoma, 2010.
Nitrogen Use Efficiency
Nitrogen use efficiency results for LCB, Hennessey and Efaw are presented in Tables 5 and 6 for 2009 and 2010 respectively. At LCB in 2009, application of 0 kg N ha−1 prep-plant plus 45 kg N ha−1 side dress N overall contributed to high NUE compared with the rest of the treatments regardless of the side dress N application distance. Lack of prep-plant N application led to a better response to application distance. The highest NUE (35%) was recorded when 45 kg N ha−1 side dress N was applied 0 cm away from the maize row with zero prep-plant. Treatments with prep-plant N applied indicated that NUE decreased with side dress N rate to as low as 7% (45 kg N ha−1 prep-plant + 134 kg N ha−1 side dress N). Application distance and the interaction between the two had no significant effect on NUE and no consistent trend was recorded.
Table 5. Grain nitrogen uptake means and percentage NUE as affected by placement distance away from the maize row of side dress N application in 2009 at Hennessey, and Lake Carl Blackwell, Oklahoma.
†, *, **, ***: Significant at the 0.1, 0.05, 0.01 and 0.001 probability levels respectively.
§SED: Standard error of the difference between two equally replicated means.
Na: Not applicable.
Table 6. Grain nitrogen uptake means and percentage NUE as affected by placement distance away from the maize row of side dress N application in 2010 at Efaw, and Lake Carl Blackwell, Oklahoma.
†, *, **, ***: Significant at the 0.1, 0.05, 0.01 and 0.001 probability levels respectively.
§SED: Standard error of the difference between two equally replicated means.
na: Not applicable.
In 2010, 0 kg N ha−1 prep-plant N plus 45 kg N ha−1 side dress N contributed to significantly (p < 0.05) lower NUE. Application of prep-plant lead to an increase in NUE, which ranged from 44 to 99% across side dress N rates applied. Increasing N rates significantly (p < 0.01) lead to a decline in NUE. Application distance and interaction between N rate and application distance did not significantly (p < 0.05) affect NUE regardless of whether prep-plant was applied or not. Generally, the highest NUE of up to 99% was obtained with 45 kg N ha−1 pre-plant and 45 kg N ha−1 side dress N, applied 10 or 20 cm away from the maize row.
At Hennessey in 2009, results of application of zero prep-plant and 45 kg N ha−1 side dress (low N rate) did not significantly (p < 0.05) influence NUE. However, with zero prep-plant, applying side dress N at the base of the maize row gave the highest NUE (23%). The factorial treatments (6 through 17) showed that side dress N, application distance and the interaction between the two factors did not significantly (p < 0.05) affect NUE with or without prep-plant. Nonetheless, NUE consistently declined with increasing side dress N rate. Overall, NUE increased when low side dress N rate (45 kg N ha−1) was applied closer to the maize row.
At Efaw, application of 0 kg N ha−1 prep-plant and 45 kg N ha−1 side dress N did not significantly (p < 0.05) affect NUE. Overall, with prep-plant applied was followed by side dress N, NUE reduced with an increase in side dress N rate. Application distance and the interaction term were not significant (p < 0.05) whether prep-plant was applied or not. However, application of side dress N at the base (0 cm) of the maize row consistently gave the highest NUE across all the side dress N rates applied. Applying 45 kg N ha−1 prep-plant followed by 45 kg N ha−1 side dress N contributed to the highest NUE of 80% when side dress N was applied 0 cm away from the maize row.
DISCUSSION
The findings presented in this experiment varied with location and cropping season. This outcome was mainly attributed to spatial and temporal variability that exists in any particular study location (Solie et al., Reference Solie, Raun and Stone1999). Soil moisture directly relates to the rainfall distribution of a given location. In N fertilizer management, it is already known that soil moisture is crucial in N mobility, especially NO3−-N, and plays a great role in uptake of N by the crops, and overall NUE (NaNagara et al., Reference NaNagara, Philips and Leggett1975). This is because N in the soil moves mainly through mass flow and diffusion to plant roots (Barber, Reference Barber1995). Temperature as well directly influences soil moisture loss and nutrients uptake through evapotranspiration process (Pregitzer and King, Reference Pregitzer, King and BassiriRad2005). In this study, according to the Oklahoma Mesonet weather station (www.mesonet.org), mean monthly air temperatures were the highest for all sites in June and July, 2009 and 2010 (Figure 1 and 2). The rainfall distribution for Haskell, Hennessey and LCB in 2009 was low for these two months excluding LCB in July (Figure 1). In 2010, Efaw and LCB recorded high mean rainfall distribution for June and July (Figure 2). This indicated that soil moisture could have been a problem in Haskell and Hennessey, but not Efaw and LCB experimental sites. It is also important to remember that LCB experiments were irrigated whenever it was necessary.
Side dress N was applied between June and July for all seasons and sites (Table 3), which corresponded to V8 to V10 maize development stage. At V8 to V10 demand for nutrients and water by maize is relatively high. Moisture and nutrients deficiency at these stages leads to poor growth and development of the ears and eventually reduced maize grain yields.
High temperatures and low rainfall distribution at Haskell and Hennessey could have resulted to low soil moisture and generally recorded poor N response. At both sites, varying side dress N application distance away from the maize row did not significantly affect maize grain yield regardless of whether prep-plant N was applied or not. But at lower N rates (45 kg N ha−1), maize grain yield was high when side dress N was placed 0 to 20 cm and 0 cm away from the maize row for Haskell and Hennessey respectively. The same thing applied for NUE at Hennessey, where NUE only increased when side dress was applied at base (0 cm) of maize row.
Since N is taken up by plants through mass flow and diffusion (Barber, Reference Barber1995), inadequate soil moisture could have contributed to further applied side dress N failing to reach plant roots on time before it was lost through ammonia volatilization. High losses of N through ammonium volatilization have been reported to occur when N fertilizer is applied to soils with low soil moisture and high air temperature (>50 °F) (Sommer et al., Reference Sommer, Olesen and Christensen1991). The decline in NUE with increased side dress N rate emphasized the important of avoiding excess N fertilizer application in maize production.
Availability of adequate moisture at Efaw and LCB (except in 2009) could have contributed to the positive N response recorded. Although application distance of side dress N away from the maize row did not significantly affect maize grain yield regardless of whether prep-plant was applied or not, slightly higher maize grain yield was obtained when side dress N was applied 0 to 10 cm away from the maize row at high N rates (90 and 134 kg N ha−1). For the two sites, high NUE was recorded when side dress was placed 0 to 20 cm away from the maize row. The decrease in root densities as the distance from the maize row increased and high concentration of maize root mass at 0 to 10 cm soil layer, located directly under the plant (Mengel and Barber, Reference Mengel and Barber1974), could have contributed to the increase in NUE and maize grain yield when side dress N was applied closer to the maize row. At both sites, NUE declined with increased side dress N, mainly due to over application. The increased NUE with zero prep-plant at LCB in 2009 emphasized how results can vary in the same site from one year to another due to spatial and temporal field visibilities, especially when dealing with N fertilizer.
CONCLUSIONS
Soil moisture plays a crucial role in N mobility, especially NO3−-N and uptake of N by the maize. Therefore, poor N response, low maize grain yield and NUE were recorded when side dress N was applied at locations with high temperature and low rainfall distribution. Adequate soil moisture contributed to positive N response, high maize grain yields and enhanced NUE. Across sites and seasons, side dress N application distance away from the maize row did not influence maize grain yield and NUE substantially even in cases where no prep-plant was applied. However, in moisture limiting environments (Haskell and Hennessey) high maize grain yields and NUE were achieved when side dress N was applied 0 to 20 cm away from the maize row especially at low N rates (45 kg N ha−1). Environments with adequate rainfall distribution (Efaw and LCB) generally had better maize grain yields and slightly higher NUE when high side dress N rates (90 and 134 kg N ha−1) were applied between 0 cm and 20 cm away from the maize row. Irrespective of how far the side dress N was applied, increasing side dress N to 134 kg N ha−1 lead to a general decline in mean NUE to as low as 4%, 26%, 45%, 43% at Hennessey, Efaw, LCB (2009), and LCB (2010) respectively.
