Agronomy Journal 94:51-56 (2002)
© 2002 American Society of Agronomy
SOIL FERTILITY
Long-Term Broadcast and Banded Phosphorus Fertilization of Corn Produced Using Two Tillage Systems
Donald D. Howard*,a,
Michael E. Essingtonb and
Joanne Loganb
a Dep. of Plant and Soil Sci., West Tennessee Exp. Stn., 605 Airways Blvd., Jackson, TN 38301
b Dep. of Plant and Soil Sci., Univ. of Tennessee, P.O. Box 1071, Knoxville, TN 37901-1071
* Corresponding author (dhoward2{at}utk.edu)
Received for publication April 14, 2000.
 |
ABSTRACT
|
|---|
Surface-broadcasting P is a management concern in no-till (NT) corn (Zea mays L.) production. Field research was established in 1983 and continued through 1993 to evaluate broadcast and banded P applications for corn produced utilizing NT and disk-till (DT) systems on a Loring silt loam (Oxyaquic Fragiudalfs). A wheat (Triticum aestivum L.) cover was established each fall. The tillage systems were main plots with a factorial arrangement of broadcast P rates, and the N + P band treatments were subplots. Annual broadcast P rates were 0, 20, 39, and 59 kg ha-1. A 22–20–0 kg ha-1 N + P band was applied after planting. All plots received a total of 168 kg N ha-1 each year. Disk-till yields were higher than NT yields for 5 of the 11 yr and were increased by P rates up to 39 kg ha-1. No-till yields were increased with P rates up to 20 kg ha-1. Yield response to banding N + P was inconsistent with year, tillage, and P rates. Three banded N + P plus broadcast P combinations equal to a broadcast P rate were applied each of the 11 yr. Yields for 5 of the 33 comparisons were increased by the banded plus broadcast P combination while three yields were reduced relative to broadcasting equal P rates. Yields for the other 25 comparisons were similar. Extractable P (EP) was increased with increased broadcast P rates, but the yield response was generally restricted to low EP situations.
Abbreviations: DT, disk-till • EP, extractable phosphorus • NT, no-till
|
INTRODUCTION
|
|---|
FERTILIZER MANAGEMENT of no-till (NT) corn production has been the focus of considerable research over the past decade. Primary management concerns include nutrient availability from surface applications and the need for a subsurface band. Banding a small amount of fertilizer provides a readily available nutrient source during periods of drought when surface root activity and nutrient uptake are restricted. Other concerns include fertilization rate, soil extractable nutrient level (or soil test level), and soil texture.
Surface-broadcasting P has been reported to be equivalent to or better than soil incorporation for increasing corn yields. In Kentucky, Belcher and Ragland (1972) found surface P applications to be as efficient for corn production as a combination of broadcast and banded P. In Georgia, Hargrove (1985) concluded that nutrient uptake by NT corn from surface-applied fertilizers was equal to or greater than nutrient uptake following soil incorporation. In Virginia, Moscheler et al. (1972) reported higher corn yields when P was broadcast than when it was soil-incorporated. Howard and Tyler (1987) found surface P applications to be as effective as subsurface-banded P for NT corn production in Tennessee.
Generally, the recommended nutrient application rate for NT production is comparable to that recommended for conventional tillage. However, Eckert and Johnson (1985) reported the need to broadcast higher P rates for corn planted using conservation tillage in Ohio.
Traditionally, crop response to an applied nutrient is related to the extractable level of that nutrient. As the extractable nutrient level is increased through fertilization, yield increases from subsequent nutrient applications are reduced. The combination of banding a small amount of fertilizer plus broadcasting a larger amount is recommended for crop production on low-testing soils (Ext. Plant and Soil Sci., 2000). The banded nutrients serve as a readily available supply for uptake while the broadcasted nutrients build up the soil nutrient supply. Bordoli and Mallarino (1998) reported that corn response to surface P application in Iowa was limited to very low and low extractable P (EP) soils. In a separate study, Mallarino et al. (1991) reported economically insignificant yield increases from broadcasting P to a soil having a high EP level. However, from a 10-yr Minnesota study, Randall et al. (1997) reported corn yield responses from surface P applications when EP was either medium or high. They also reported reduced yields as EP was reduced by inadequate or zero P applications.
The surface residues associated with NT corn production reduce moisture evaporation and soil temperatures, which in turn may reduce or delay seed germination and early plant growth. Placing a small amount of readily available fertilizer (starter) close to the seed has been used to partially overcome possible detrimental effects of the surface residues. From reviewing starter P research, Randall and Hoeft (1988) concluded that yield responses from band placement (5 by 5 cm) generally occur on low EP soils. In Minnesota, Rehm et al. (1988) reported that banded fertilizers increased yields of conservation-tilled corn on low P-testing soils but not on high EP soils. In Tennessee, infurrow-applied N + P starters increased NT corn yields on a low EP soil (Howard and Mullen, 1991; Howard and Tyler, 1987). Eckert and Johnson (1985) reported higher yields from banded P than from broadcast P on two Ohio soils having a low EP level. However, in Iowa, Bordoli and Mallarino (1998) reported no yield increase due to banded P when compared with broadcast P on low or very low EP soils. In a Louisiana study, Mascagni and Boquet (1996) reported that infurrow starter N + P applications increased NT corn yields in 2 of 3 yr on a high EP soil.
Yield response to banding N + P has been reported to be affected by the cultivar (Mascagni and Boquet, 1996; Gordon et al., 1997; Teare and Wright, 1990). From a study in Louisiana, Mascagni and Boquet (1996) reported that yield increases from an infurrow N + P (11–16–0) application on a high P-testing soil varied with corn hybrid and year. Yields were increased by the starter application in 2 of the 3 yr. In the first year, yields from four of the six hybrids responded to the starter while three of the hybrids responded in the third year. In a 3-yr Kansas study, Gordon et al. (1997) reported a yield increase from three of five hybrids from an N + P starter application to a high EP soil. In Florida, Teare and Wright (1990) reported a consistent yield response to starter fertilizers by some corn hybrids while others either did not respond or the response was negative. In these studies, it is not possible to distinguish if the yield response can be attributed to either N, P, or their combination.
This study was conducted to evaluate broadcasted P rates and a banded N + P application for corn produced on a loess-derived soil using two tillage systems. The study was conducted as part of a continuing effort to develop nutrient recommendations for NT corn production.
|
MATERIALS AND METHODS
|
|---|
Field Establishment Methods
Field research was established in 1983 and continued through 1993 on a loess-derived Loring silt loam (fine-silty, mixed, active, thermic, Oxyaquic Fragiudalfs) at the Milan Experiment Station, Milan, TN (35°59' N, 88°50' W; elevation of 124 m). Average annual rainfall at Milan is 138.4 cm, and average annual temperature is 14.3°C. Initial soil test results showed the soil had 3 mg kg-1 Mehlich-I (Mehlich, 1953) EP and 80 mg kg-1 extractable K. These extractable nutrient levels were categorized as low P and medium K (Ext. Plant and Soil Sci., 2000). A wheat cover was established each fall following corn harvest.
Pioneer brand 3389 corn was planted in 1983 through 1991, and Pioneer brand 3140 corn was planted in 1992 and 1993. The experimental design was a split-plot with treatments replicated four times. The main plot was tillage [disk-till (DT) and NT], and the subplots were a factorial arrangement of broadcast P rates and an N + P band treatment. Broadcast P rates were 0, 20, 39, and 59 kg ha-1. A 22–20–0 kg ha-1 N + P band was applied after planting. The broadcast P treatments were hand-applied using triple superphosphate. The 22–20–0 band, a combination of 11–16–0 and urea ammonium nitrate (32%), was applied in both tillage systems using a four-row injector having a knife configured behind a coulter. The liquid materials were applied with a CO2–pressurized system. The band was applied approximately 5 cm from the row and 5 cm below the surface immediately after planting as a separate operation. The DT plots were disked two or three times to a depth of 
10 cm immediately before planting and after broadcasting P. Nitrogen was broadcast as ammonium nitrate [NH4NO3] (34% N) immediately after planting. When the N + P band was applied, a reduced N rate was broadcast to provide a total N rate of 168 kg ha-1. Potassium was broadcast at 56 kg ha-1 using KCl. Corn was planted early to mid-April in 0.76-m row widths. Individual plots were four rows wide (76-cm spacing) and 9.1 m long. Treatments were maintained on the same plots throughout the 11 yr.
Approximately 2 wk before planting, the wheat cover crop was killed with paraquat [1,1'-dimethyl-4,4'-bipyridinium ion] (0.52 kg a.i. ha-1) containing 0.5% nonionic surfactant. Atrazine [2-chloro-4-ethylamino-6-isopropylamino-s-triazine] (2.2 kg a.i. ha-1) and alachlor [2-chloro-N-(2,6 diethylphenyl)-N-(methoxymethyl) acetamide] (2.2 kg a.i. ha-1) were applied for control of residual weed and seedling johnsongrass [Sorghum halepense (L.) Pers.]. Carbofuran [2,3-dihydro-2, 2-dimethyl-7-benzofuranyl methylcarbamate] (1.12 kg a.i. ha-1) was applied infurrow at planting for insect control.
Measurements
Soil samples (0–15 cm) were collected by hand from the broadcasted P plots in the fall of selected years. A random sampling procedure was used to ensure sample collection from positions both in the row and between the rows. Eight cores (1.9 cm diam.) were collected from each plot and subdivided into 0- to 7.5- and 7.5- to 15-cm depths before chemical analysis. Mehlich-I EP was determined using a 1:5 soil/extractant ratio. Extractable P was determined using the ascorbic acid colorimetric procedure (Watanabe and Olsen, 1965) and a Model 20 Bausch and Lomb spectrophotometer (Rochester, NY).
Corn grain was harvested with a plot combine from the two center plot rows (9 by 1.5 m). Yields were corrected to 155 g kg-1 moisture.
Weather Collection
Daily temperature and precipitation data were collected from the National Weather Service Cooperative Station at Milan. Correlation analysis and mean comparisons were used to quantify the effects of weather on crop response to P fertilization.
Statistical Analysis
The statistical analyses of yield and soil test data were performed using Mixed Model procedures of SAS (SAS Inst., 1997). The Mixed Model procedure provides Type III F-values but does not provide mean square values for each element within the analysis or the error terms. Mean separation was evaluated through a series of pairwise contrasts among all treatments (Saxton, 1998). Probability levels >0.05 were not significant.
|
RESULTS AND DISCUSSION
|
|---|
Weather Data
The maximum and minimum temperature and total precipitation data from April to August of each year showed large variability over the 11-yr study period (Table 1). Overall, it was hotter and drier than the 30-yr average (1961–1990). The hottest and driest season was 1987 when the average maximum temperature was 2.2°C higher than average and the seasonal rainfall was 190 mm below average. In contrast, 1983 had an average minimum temperature 0.7°C below normal and rainfall 100 mm above normal.
Yields
The ANOVA indicates that the main effects of year (Y), tillage (T), broadcast P rate (P), and banded N + P (B) had a highly significant (P < 0.05) effect on corn yield (Table 2). The ANOVA showed that the two-way interactions (P x Y, P x T, B x Y, and B x P) plus one three-way interaction (B x P x Y) affected yield.
View this table:
[in this window]
[in a new window]
|
Table 2. Mixed-model ANOVA for tillage, broadcast, and banded N + P effects on corn yield during 11 production years.
|
|
As expected for rainfed production, corn yields differed each year, ranging between 5 and 10 Mg ha-1. However, crop response could not be correlated with precipitation. Yields were greater for DT production than NT production for 5 of the 11 yr. For these years, DT yields ranged from 0.59 to 1.34 Mg ha-1 higher than the NT yields. Although NT yields were lower, the yield response to applied P was greater for NT production (Table 3). The DT yields were increased 0.44 Mg ha-1 (from 8.10 to 8.54 Mg ha-1, a 5.4% increase) by applying 39 kg P ha-1. No-till yields were increased 0.62 Mg ha-1 (from 7.47 to 8.09 Mg ha-1, an 8.3% increase) by applying 20 kg P ha-1. This interactive effect of P fertilization and tillage on corn yields was consistent over the 11 yr, as the effect of years (P x T x Y) was not significant. It could be speculated that the yield response to the lower fertilization rate (20 kg P ha-1) suggests a greater efficiency for the NT system, which would agree with the findings of others (Belcher and Ragland, 1972; Hargrove, 1985; Howard and Tyler, 1987; Howard and Mullen, 1991) although Eckert and Johnson (1985) noted that higher P rates were needed to produce NT corn yields that were comparable to those obtained when P was incorporated.
View this table:
[in this window]
[in a new window]
|
Table 3. The effect of broadcast P rates on corn yield as affected by no-till (NT) and disk tillage (DT) during the 11-yr study.
|
|
The P x Y interaction shows that the yield effect, averaged across the tillage systems, resulting from increased P rates differed in time (Table 2). Applying P to the low EP soil increased yields in 9 of the 11 yr (Table 4). In 1983 and 1985, yields were increased by broadcasting 20 kg P ha-1 and were reduced with the application of the higher P rate(s). Yields were not increased by P fertilization in either 1987 or 1991.
Banding N + P significantly affected corn yields (Table 2). However, the banding effect on yields was also affected by B x Y, B x P, and B x P x Y interactions. The B x Y interaction resulted when banded N + P increased yields in 5 of the 11 yr (data not presented).
The ANOVA shows that yields from banding N + P were inconsistent across the four broadcast P rates as indicated by the significant B x P interaction (Table 2). Banding N + P increased yields when 0 and 20 kg P ha-1 were broadcast, but yields were increased further by broadcasting 39 and 59 kg P ha-1 (Table 5). The effect of the banded treatment for increasing yields must be evaluated for equal P applications. Banding 20 kg P ha-1 produced yields similar to those produced by broadcasting 20 kg P ha-1. Yields were also comparable for banding plus broadcasting 39 and 59 kg P ha-1 compared with broadcasting at the same rates. These observations suggest broadcast P to be sufficient for corn production on this soil and agree with the findings of Belcher and Ragland (1972) in Kentucky and Bordoli and Mallarino (1998) in Iowa.
However, the banding effect must be evaluated for each year and P rate because of the B x P x Y interaction (Table 2). For the 11 yr, three treatments of banded N + P plus broadcast P were applied at a rate equal to the broadcast P rate each year. Five of these 33 comparisons showed the banded N + P plus broadcast P treatment increased yields (Table 6). In 1985, banding N + P plus broadcasting 20 kg P ha-1 increased yields compared with broadcasting 39 kg P ha-1 (9.43 vs. 8.55 Mg ha-1). This was also observed within the same year for applying 59 kg P ha-1 (9.21 vs. 8.71 Mg ha-1). Banding N + P resulted in higher yields compared with broadcasting 20 kg P ha-1 in 1987, 1990, and 1991. For the remaining 28 combinations, three of the yields were higher when P was broadcast. In 1983, broadcasting 20 kg P ha-1 plus banding N + P resulted in lower yields compared with broadcasting 39 kg P ha-1 (5.26 vs. 5.77 Mg ha-1). Broadcasting 20 kg P ha-1 resulted in higher yields in 1986 (8.71 vs. 8.23 Mg ha-1) and 1993 (7.76 vs. 7.16 Mg ha-1) compared with banding 20 kg P ha-1. Yields from the other 25 combinations were similar. Considering the cost of banding equipment, extra tractor power required, and extra time to fill fertilizer hoppers, etc., the cost benefit of banding N + P (5 by 5 cm) on this low EP soil is questionable.
Extractable Soil Phosphorus
Extractable P, evaluated in 3 of the 11 yr (1984, 1987, and 1988), was increased by broadcast P rate (Table 7). In 1984, following 2 yr of applying either 39 or 59 kg P ha-1, EP within the DT system had been increased to 12 and 11 mg kg-1, which would be categorized as medium testing (Ext. Plant and Soil Sci., 2000). Extractable P levels in the NT system were also increased, but the soil test rating remained low.
View this table:
[in this window]
[in a new window]
|
Table 7. Effect of broadcast P rate on yield and extractable P (EP) in the 0- to 7.5- and 7.5- to 15-cm soil depths as a function of tillage.
|
|
In 1984, yields were increased with the increased broadcast P rates for the DT system. Increasing the broadcast P rate increased DT yields when EP was 11 mg kg-1. Broadcasting 59 kg P ha-1 increased yields (10.07 Mg ha-1) compared with broadcasting 39 kg P ha-1 (9.40 Mg ha-1) when EP was 11 and 12 mg kg-1. Yields from broadcasting either 20 kg P ha-1 (9.42 Mg ha-1) or 39 kg P ha-1 (9.40 Mg ha-1) when EP levels were 6 and 12 mg kg-1, respectively, were not increased compared with the check (9.36 Mg ha-1) having 3 mg kg-1 EP. The EP levels of the check and 20 kg P ha-1 fertilized plots were not significantly different, but both were lower than the EP level when 39 kg P ha-1 was broadcast. No-till yields were reduced by broadcasting 59 kg P ha-1 when the EP was 9 mg kg-1 (7.34 Mg ha-1) compared with the check (8.14 Mg ha-1) with an EP of 4 mg kg-1. Broadcasting 20 kg P ha-1 when EP was 6 mg kg-1 resulted in the highest NT yields (8.78 Mg ha-1). The 6 mg kg-1 EP was higher than the check EP (4 mg kg-1 EP) and equal to EP associated with broadcasting 39 kg P ha-1, but it resulted in lower yields (7.70 Mg ha-1). These observations suggest that Mehlich-I EP may not be a good indicator for predicting corn responses to broadcast P.
Nutrient stratification had occurred in both tillage systems within the 0- to 7.5- and 7.5- to 15-cm soil depths. Tillage associated with the DT system only partially incorporated broadcast P below the 7.5 cm depth.
The 1987 EP levels and yields improved with increased P rates for both tillage systems. Disk-till yields produced by broadcasting either 20 kg P ha-1 (7.40 Mg ha-1) or 39 kg P ha-1 (7.92 Mg ha-1) when EP was 8 mg kg-1 were not increased compared with the check (7.03 Mg ha-1) having 5 mg kg-1 EP. The EP of the check was lower than the EP resulting from broadcasting either 20 or 39 kg P ha-1. No-till yields resulting from broadcasting either 20 kg P ha-1 (6.74 Mg ha-1) or 39 kg P ha-1 (6.99 Mg ha-1) when EP was 6 and 8 mg kg-1, respectively, were not increased compared with the check (6.27 Mg ha-1) having 2 mg kg-1 EP. The EP of the check was lower than that associated with broadcasting 39 kg P ha-1. Again, it appears that Mehlich-I EP may not be a good indicator for predicting corn responses to broadcast P regardless of tillage system.
The 1988 EP levels of the DT system were increased from 4 to 12 mg kg-1 with increased broadcast P rates. Broadcasting 20 kg P ha-1 when EP was 7 mg kg-1 increased DT yields (9.89 Mg ha-1) compared with the check (9.06 Mg ha-1) having 4 mg kg-1 EP. However, the EP levels were not significantly different. Broadcasting either 39 or 59 kg P ha-1 did not increase yields compared with broadcasting 20 kg P ha-1 when the EP ranged from 10 to 12 mg kg-1. The NT yield responses were similar to those of the DT system. Yields were increased from broadcasting 20 kg P ha-1 (8.98 Mg ha-1) when EP was 8 mg kg-1 compared with the check (8.08 Mg ha-1) having 2 mg kg-1 EP. The EP associated with broadcasting 20 kg P ha-1 was higher than the check. However, broadcasting either 39 or 59 kg P ha-1 when EP was 11 and 22 mg kg-1, respectively, did not increase yields compared with broadcasting 20 kg P ha-1 with an EP of 8 mg kg-1. The EP levels associated with broadcasting 20 and 39 kg P ha-1 were not different, but the EP level associated with broadcasting 59 kg ha-1 was higher than that associated with broadcasting 20 kg P ha-1.
These observations show that yield responses from broadcasting P occurred when Mehlich-I EP ranged between 6 and 13 mg kg-1. Evaluating the EP contained in the 0- to 7.5-cm sampling depth shows that yield responses occurred at levels ranging between 4 and 8 mg kg-1 EP. Based on the soil test ratings (0–15 cm depth), three of the responses to broadcasting P occurred when EP was rated low. Two of the responses for medium EP (1987 DT and NT) could have been rated as a response to low EP because there was no yield difference for broadcasting 20 kg P ha-1 when EP was 6 and 8 mg kg-1 and for broadcasting 59 kg P ha-1 when EP was 13 and 9 mg ha-1. Broadcast yield responses being primarily limited to low EP soils agrees with findings of Bordoli and Mallarino (1998).
|
CONCLUSIONS
|
|---|
Broadcasted P rates increased DT and NT corn yields on this loess-derived soil. Over the 11 yr, DT yields were higher than NT yields. In general, DT yields responded to rates of P fertilization up to 39 kg P ha-1 while NT yields responded to P rates up to 20 kg P ha-1. Yield responses to banding N + P (5 by 5 cm) were inconsistent over the years, tillage systems, and broadcast P rates. When P was banded plus broadcast, the yields were similar with yields produced by broadcasting the same total P rate. For the 11 yr, three banded N + P plus broadcast P treatments were applied at a rate equal to the broadcast P rate each year. For the 33 treatments of equal P rates, five of these comparisons showed the banded N + P plus broadcast P treatment to increase yields while three showed that yields were decreased. The other 25 comparisons produced similar yields. This observance of yields from banding N + P makes the practice questionable for corn production on this soil. Extractable P was increased with increased broadcast P rates and was stratified with soil depth within both tillage systems. The relationship between increased yields for either tillage system with broadcast P rates and Mehlich-I EP levels indicates that responses primarily occurred when EP was low. Although both seasonal temperature and precipitation varied during the field study, the climate did not influence crop response to fertilizer P.
|
REFERENCES
|
|---|
- Belcher, C.R., and J.L. Ragland. 1972. Phosphorus adsorption by sod-planted corn. Agron. J. 64:754–756.[Abstract/Free Full Text]
- Bordoli, J.M., and A.P. Mallarino. 1998. Deep and shallow banding of phosphorus and potassium as alternatives to broadcast fertilization for no-till corn. Agron. J. 90:27–33.[Abstract/Free Full Text]
- Eckert, D.J., and J.W. Johnson. 1985. Phosphorus fertilization in no-tillage corn production. Agron. J. 77:789–792.[Abstract/Free Full Text]
- Extension Plant and Soil Science. 2000. Soil fertility and soil testing. In Plant and soil science handbook. Univ. of Tennessee Inst. of Agric., Knoxville.
- Gordon, W.B., D.L. Fjell, and D.A. Whitney. 1997. Corn hybrid response to starter fertilizer in a no-tillage, dryland environment. J. Prod. Agric. 10:401–404.
- Hargrove, W.L. 1985. Influence of tillage on nutrient uptake and yield of corn. Agron. J. 77:763–768.[Abstract/Free Full Text]
- Howard, D.D., and M.D. Mullen. 1991. Evaluation of in-furrow and banded starter N-P-K nutrient combinations for no-tillage corn production on soil fertilized with four rates of phosphate and potash. J. Fert. Issues 8:34–39.
- Howard, D.D., and D.D. Tyler. 1987. Comparison of surface applied rates of phosphorus and potassium and in-furrow fertilizer solution combinations for no-till corn production. J. Fert. Issues 4:48–52.
- Mallarino, A.P., J.R. Webb, and A.M. Blackmer. 1991. Corn and soybean yields during 11 years of phosphorus and potassium fertilization on a high-testing soil. J. Prod. Agric. 4:312–317.
- Mascagni, H.J., Jr., and D.J. Boquet. 1996. Starter fertilizer and planting data effects on corn rotated with cotton. Agron. J. 88:975–982.[Abstract/Free Full Text]
- Mehlich, A. 1953. Determination of P, Ca, Mg, K, Na and NH4. North Carolina Soil Testing Div., Raleigh.
- Moscheler, W.W., G.M. Shear, D.C. Martens, G.D. Jones, and R.R. Wilmouth. 1972. Comparative yield and fertilizer efficiency of no-till and conventionally tilled corn. Agron J. 64:229–231.[Abstract/Free Full Text]
- Randall, G.W., S.D. Evans, and T.K. Iragavarapu. 1997. Long-term P and K applications: II. Effect on corn and soybean yields and plant P and K concentrations. J. Prod. Agric. 10:572–580.
- Randall, G.W., and R.G. Hoeft. 1988. Placement methods for improved efficiency of P and K fertilizers: A review. J. Prod. Agric. 1:70–79.
- Rehm, G.W., S.D. Evans, W.W. Nelson, and G.W. Randall. 1988. Influence of placement of phosphorus and potassium on yield of corn and soybeans. J. Fert. Issues 5:6–13.
- SAS Institute. 1997. SAS/STAT software. Changes in enhancements through release 6.12. SAS Inst., Cary, NC.
- Saxton, A.M. 1998. A macro for converting mean separation output to letter groupings in Proc. Mixed. p. 1243–1246. In Proc. Annu. SAS Users Group Int. Conf., 23rd, Nashville, TN. 22–25 Mar. 1998. SAS Inst., Cary, NC.
- Teare, I.D., and D.L. Wright. 1990. Corn hybrid-starter fertilizer interaction for yield and lodging. Crop Sci. 30:1298–1303.[Abstract/Free Full Text]
- Watanabe, F.S., and S.R. Olsen. 1965. Test of an ascorbic acid method for determining phosphorus in water and NaHCO3 extracts from soil. Soil Sci. Soc. Am. Proc. 29:677–678.
TOP
ABSTRACT
INTRODUCTION
