F.G. Fernández, K. D. Greer, C. E. Byers, B. S. Farmaha, and R.E. Dunker 1
Application of phosphorus (P) and potassium (K) are required to attain high quality yields in corn (Zea mays L.). The application of these fertilizers represents a multi-billion dollar investment for US farmers. Thus improving management practices that optimize returns on fertilizer investment can be of major importance. In recent years, with the adoption of conservation tillage systems, such as no-till (NT) and strip-till (ST), there has been a renewed interest on the efficient placement of P and K fertilizers. Broadcast application of P and K under NT results in higher content of these nutrients in the surface compared to the subsurface layer of the soil (vertical stratification) while deep band applications in ST result in enrichment of these nutrients in the band and depletion between bands (horizontal stratification) (Buah et al., 2000; Holanda et al., 1998; Rehm, et al., 2002).
Corn roots takes up P and K from the soil mostly by diffusion (Barber, 1995). In order for this process to occur, the root system needs to be active and there needs to be water available in the soil for these nutrients to diffuse to root surfaces. Stratification forces P and K uptake to be more dependent on the ability of plant roots to exploit the volume of soil with elevated concentrations of these nutrients and on the characteristics of that portion of the soil. Under vertical stratification in rainfed conditions, inadequate soil water supply can diminish P and K availability at the soil surface layer. On the other hand, deep band placement of P and K can take advantage of potentially higher water content present at lower depths. However, without sufficient rainfall, subsurface soils could eventually become dry and low rainfall might re-wet the surface soil layer, but may not be sufficient to penetrate to the subsurface soil layer. In addition, the ability of the crop to take up nutrients from the subsurface can be reduced because the fibrous root system of corn is more prolific in the surface compared to the subsurface soil layer, and soil oxygen levels needed for nutrient uptake decline rapidly with increased soil depth. In light of all these factors, the question then becomes: What placement technique of P and K constitutes the best management practice for conservation tillage rainfed-corn?
Despite the importance of all the factors mentioned above for adequate P and K nutrition, seldom are P, K, and water status monitored at the different rooting depths of the soil through the growing season; nor is this information coupled with the corresponding root data necessary to adequately quantify plant P and K availability. The amount of root surface area and soil water availability determine by in large how much of the applied P and K fertilizers could be taken up by corn. Unfortunately, these factors that determine P and K availability have not been properly identified in fields under conservation tillage. This lack of knowledge and the need to improve nutrient use efficiency by truly understanding when, where, and under what field conditions P and K are being taken up by corn, constitutes the basis of this research.
This study was started in the spring of 2007 at the Crop Sciences Research & Education Center in Champaign on a field that remained untilled after the previous corn crop. The study site is mostly on a Flanagan silt loam soil (Fine, smectitic, mesic Aquic Argiudolls) and small portions of the field is in a Drummer silty clay loam soil (Fine-silty, mixed, superactive, mesic Typic Endoaquolls). At the start of the experiment the soil had the following characteristics: CEC of 14.2 cmol (+) kg-1, 3.6% organic matter, a pH of 5.7, P concentration of 16.2 mg kg-1, K concentration of 185 mg kg-1,Ca concentration of 1564 mg kg-1, and Mg concentrations 253 mg kg-1. The field was divided to establish a corn-soybean rotation system with both crops present each year. Annual applications of all possible combinations of four P levels (0, 25, 50, 75 lb P2O5 acre-1) and four K levels (0, 45, 90, 180 lb K2O acre-1) were applied either by broadcast or deep band under the row (30 inches apart and 6 inches below the soil surface). The treatments will remain in the same 20 x 75 ft. plot during successive years. The experimental design was a split-split-plot arrangement of treatments in a randomized complete block design with 3 replications. Tillage/placement (no-till with broadcast application (NT-BC), no-till with deep placement application (NT-DP), and strip-till with deep placement application (ST-DP)) was the main plot, P treatment was the subplot and K treatments was the sub-subplot. Application of P and K treatments in NT-DP was done with a low disturbance knife to minimize tillage effects. Fertilizer was applied early in the spring of 2007. Urea ammonium nitrate at the rate of 180 lb N acre-1 was applied on the corn plots with a sprayer prior to planting. Corn hybrid Pioneer 34N40 was planted on May 23, 2007.
Three composite soil samples were collected from the center of each plot prior to fertilizer application from the 0-2, 2-4, 4-7, and 7-20 inch soil depth increment. Soil samples were collected in the same fashion after harvest except this time two separate samples (in the crop row and between rows) were collected. From selected plots (P-K rates: 0-0, 75-0, 0-180, 75-180, and 50-90 lb acre-1 of P2O5 and K2O, respectively) soil, shoot, and root samples were collected at R1 and R3 corn development stages. Each plot was divided into two halves and samples were kept separate for each half to produce two sub-samples per plot. Three composite soil samples were collected from the 0-2, 2-4, 4-8, and 8-16 inch depth increments from the crop row and between the rows and kept separate for P and K analyses. Similar procedures were followed for root samples. A five-plant composite sample was harvested by cutting plants at the soil surface. Leaves were detached from the rest of the above ground plant materials to determine leaf area and the leaf and remaining plant materials were dried to determine leaf and total above ground dry matter accumulation. Gravimetric soil water (Θg) content in replication 2 was measured at least once weekly during the season by extracting three composite soil cores from the 0-2, 2-4, 4-8, and 8-16 inch depth increments from the crop row and between the rows and kept separate from plots with the P-K rates 0-0 and 75-180 lb acre-1 P2O5-K2O. Additionally in replication 2, volumetric water content was measured continuously using data loggers and soil water probes installed at the 0-2, 2-4, 4-6, 6-8, and 8-16 inch soil depth increment in plots with the P-K rate75-180 lb acre-1 P2O5-K2O. Yield component parameters were measured by hand harvesting 10 plants from the selected plots for intensive measurements on October 2. Grain yield was measured by machine harvest in the center of the middle two rows of each plot on October 10, 2007.
Soil P was determined by Bray and Kurtz P-1 test (Frank et al., 1998), soil K by the ammonium acetate extraction procedure (Warncke and Brown, 1998). Root samples were cleaned by elutriation and further cleaned by hand and analyzed for physical parameters by scanning with Win-RHIZO software and an Epson Expression 10000XL scanner (Régent Instruments Inc.). Tissue samples were analyzed for nutrient content following the official methods of analysis of AOAC International (Horwutz, 2000). Grain protein and oil contents were measured by near-infrared reflectance spectroscopy. Data was analyzed by the GLM procedure (SAS Institute, 2004). Results of significance are at p<0.05 unless otherwise indicated.
Grain Yield and Composition
A combination of late planting and dry conditions early in the growing season negatively impacted corn yields in 2007. Individual plot yields varied from 87 to 166 bu acre-1 with a median value for all plots of 123 bu acre-1. Corn grain yields were not significantly different for the three tillage-placement methods or for the four K fertilizer rates, but were affected by P fertilizer rate (Table 1). When averaged over all other variables, the 75 lb P2O5 acre-1 produced a yield increase of 11 bu acre-1 over the control (no P added). The grain yield response is in agreement with the current P and K recommendations for Illinois (Hoeft and Peck, 2002). The starting soil P levels were about 8 lb P2O5 acre-1 below the minimum to produce optimum yield, thus it was anticipated that application of P would enhanced grain yield. On the other hand, starting soil K levels were about 70 lb K2O acre-1 above the minimum required for optimum yield, thus yield increase with additional K fertilization was not expected. The significant 2-way interactions between tillage-placement and P rate indicated a slight yield advantage for the 75 lb P2O5 acre-1 rate in the NT-BC compared to the NT-DP (10 bu acre-1) and ST-DP (7 bu. acre-1). In other words, applying 75 P2O5 acre-1 by broadcast increased yield slightly compared to deep placement of P. The significant 2-way interaction between tillage-placement and K rate indicated a slight yield increase for NT-BC and ST-DP compared to NT-DP when K is applied at the 45 lb K2O acre-1 rate (20 bu acre-1 yield increase) and 180 lb K2O acre-1 rate (16 bu acre-1 yield increase). However, since this yield increase was not observed for the intermediate rate of 90 lb K2O acre-1, the interpretation of this 2-way interaction can be considered inconsistent. Samples for yield component analysis have been collected but no information is available at this time since not all samples have been processed.
Harvested grain oil, protein, and starch contents were not influenced by any of the treatment variables. Mean values of grain composition were: oil 4%, protein 9%, and starch 71.5%.
Soil P and K levels
Soil P and K measurements presented here are preliminary since not all samples have been processed yet to perform the necessary statistical analyses. All samples are currently being analyzed and will be reported along with the findings from next year. Phosphorous concentrations in the soil prior to treatment application were highly stratified with highest concentrations in the surface (Figure 1a). There was approximately a 50% decline in concentration with each successive soil layer moving down the soil profile. Potassium was also stratified with two fold higher concentrations in the top 2 inches of the soil compared to the 2-4 soil depth increment (Figure 1b). Potassium concentrations at the 2 to 20 inch depth increment were fairly uniform. Similar vertical stratification of soil P and K levels were observed for the in-season samplings and the fall sampling (data not shown).
Statistical analysis of main effects on soil P and K levels on the top 16 inches of soil depth, averaged over two in-season samplings (R1 and R3 corn developmental stages), are presented in Table 2. Soil K levels were significantly impacted by tillage-placement. While it is too early to draw any certain conclusion from these preliminary results, a possible explanation of the fact that strip-tillage showed lower K concentrations than the no-tillage treatments is that strip-tillage increased the amount of soil disturbance compared to the other treatments. This disturbance, followed by dry weather conditions, could have caused some K fixation (due to soil dryness) which would have caused a decline in soil K levels measured by the ammonium acetate extraction. Application of P increased both P and K soil levels compared to the control (no P added). On the other hand, application of K increased soil K levels compared to the check (no K added), but no effect was observed for soil P levels. The only significant 2-way interaction was the tillage-placement treatment-by-P rate for both soil P and K levels. The interaction indicated that a slight increase in soil P levels (5 lb P2O5 acre-1) occurred when 75 lb P2O5 acre-1 rate was applied in the NT-BC treatment compared to no P application. This interaction for soil K levels indicated that deep placing P increased soil K levels by 27 lb K2O acre-1 compared to no P application.
Soil Water Content
Since crops obtain P and K from the soil solution, measurement of Θg is important to help determine nutrient availability. The Θg content measurements collected in the corn row and the precipitation measured during the growing season are shown in Figure 2. Statistical analysis of Θg content showed no effect by either the tillage-placement or the highest and lowest P-K combination variables. Therefore, this graph shows the mean values across those treatment variables. Through the growing season, Θg measurements were significantly influenced by the date of measurement (which is related to precipitation events) and soil depth increments. The pattern of Θg content depletion and recharging were fairly consistent through the growing season for all soil depth increments. Water content decreased from the deepest soil depth increment measured to the soil surface. In all cases, except for the August 17 measurement, after rain events totaling more than 1 inch of precipitation, this pattern was inverted to decreasing Θg content from the soil surface to the deepest soil depth increment. Finally, across the growing season, the place of measurement (in the corn row or in-between rows) significantly influenced Θg content at the p<0.01 level. Over the growing season and treatment variables, the in-row Θg measurement was 0.008, 0.007, 0.011, and 0.010 greater than the in-between rows measurement for the 0 to 2, 2 to 4, 4 to 8, and 8 to 16 inch depth increment, respectively. Figure 3 shows the Θg content in-row and in-between rows across the season averaged over all treatment variables and soil depth increments. Early in the growing season, Θg content was slightly less in the row compared to the in-between rows position. This is likely because roots of the young corn plant are exploiting water from a small portion of the soil volume near the location where the seed was planted. Another possible explanation is that early in the growing season the in-row position could have more surface evaporation as the crop residue from the previous crop was removed, or at least disturbed, in a narrow band during planting operations. Greater depletion of soil water in the between rows position compared to the in-row position occurred later in the growing season likely because of greater root activity in that portion of the soil volume.
Above-Ground Vegetative Measurements
Leaf area indexes (LAI) and above-ground dry matter accumulation were measured at the R1 and R3 development stages. Statistical analysis of main effects for factorial combination of highest and lowest P and K rates showed significant differences only for dry matter accumulation at the R3 development stage in response to tillage (p<0.05) and K rate (p<0.1) (Figure 4). Across all other variables, application of 180 lb K2O acre-1 increased dry matter accumulation by 1,084 lb acre-1 compared to no K application; and the NT-BC treatment increased dry matter accumulation by 2,665 lb acre-1 compared to the NT-DP treatment.
Root Measurements
The results on root measurements are preliminary since not all samples have been processed to perform the necessary statistical analyses. All soil and organic debris have been removed from root samples. Root parameters are currently being quantified and will be reported along with the findings from next year. None of the main factors (Tillage-placement, P rate, or K rate) influenced any of the root parameters measured from a limited subset of samples collected from the 0-0 and 75-180 P-K rate treatments. Root length density (RLD), which is a measurement of root length per given volume of soil, was significantly influenced by soil depth increment (p<0.01). There was approximately a two-fold step increase for each successive depth increment measured from the 8-16 to the 0-2 in soil layer (Figure 5). Root surface area (RSA) was also greatest at the 0-2 in soil depth increment (data not shown). For the top 16 inches of soil depth, the in-row position had a RSA of 41.2 cm2 compared to 29.8 cm2 for the between-rows position indicating that there is an overall greater potential for nutrient and water absorption by the root system in the row position. While RLD were not significantly different in-row or in-between rows positions, significant mean root diameter (MRD) differences of 0.27 mm in the row compared to 0.23 mm in between rows positions were observed (p<0.001). Thus the greater RSA observed at the in-row position was related to larger overall root diameters and not to greater root lengths. Larger RSA and greater water availability later in the growing season at the in-row position could indicate that banding of fertilizers at the crop row position could be a beneficial practice to increase crop-nutrient uptake.
Table 1. Grain yield as affected by tillage-placement and factorial P and K rate of application.
1F.G. Fernández is Assistant Professor, K.D. Greer is Research Specialist, C.E. Byers and B. S. Farmaha are Research Assistants, and R.E. Dunker is Agronomists, All with the Dept. of Crop Sciences, Univ. of Illinois, Urbana, IL.
Barber, S.A. 1995. Soil nutrient bioavailability: A mechanistic approach. 2nd ed. Wiley & Sons, New York.
Buah, S.S.J., T.A. Polito, and R. Killorn. 2000. No-tillage soybean response to banded and broadcast and direct and residual fertilizer phosphorous and potassium applications. Agron. J. 92: 657-662.
Frank, K, D. Beegle, and J. Denning. 1998. Phosphorus. p. 21-29. In: J.R. Brown (ed.) Recommended chemical soil test procedures for the North Central Region. North Central Reg. Publ. 221(rev.). Univ. of Missouri, Columbia.
Hoeft, R.G., and T.R. Peck. 2002. Soil testing and fertility. In Illinois Agronomy Handbook, 23rd ed. College of Agriculture, Consumer and Environmental Sciences, Dept. of Crop Sciences, University of Illinois Extension, University of Illinois.
Holanda, F.S.R., D.B. Mengel, M.B. Paula, J.G. Carvaho, and J.C. Bertoni. 1998. Influence of crop rotations and tillage systems on phosphorous and potassium stratification and root distribution in soil profile. Commun. Soil Sci. Plant Anal. 29: 2383-2394.
Horwutz, W. (ed.) 2000. Official methods of analysis of AOAC International. 17th ed. Vol. 1. AOAC Int., Gaithersburg, MD.
Rehm, G.W., A.P. Mallarino, K. Reid, and J. Lamb. 2002. Soil sampling for variable rate fertilizer and lime application [on line]. Available at: http://www.extension.umn.edu/distribution/cropsystems/DC7647.html (Verified 21 Dec. 2007).
SAS Institute. 2004. SAS/STAT user’s guide. Version 9.1 ed. SAS Inst., Cary, NC.
Warncke, D., and J.R. Brown. 1998. Potassium and other basic cations. p. 31-33. In: J.R. Brown (ed.) Recommended chemical soil test procedures for the North Central Region. North Central Reg. Publ. 221(rev.). Univ. of Missouri, Columbia.