Tillage and potassium placement effects on potassium use efficiency in a corn-soybean rotation.

Illinois Fertilizer Conference Proceedings
January 25-27, 1999

Home 1999 Index Search

Tillage and Potassium Placement Effects on Potassium
Use Efficiency in a Corn-Soybean Rotation

S.A. Ebelhar and E.C. Varsa¹

Introduction

Potassium (K) fertilizer placement and mixing within the soil Ap horizon can influence its availability for plant uptake. Other factors affecting uptake are root growth of plants, diffusion rate of K to roots, K fixation, environmental conditions, and tillage. Previous work by Kovar and Barber (1987) at Purdue suggested the possibility for optimizing potassium uptake and utilization by corn based on optimal placement and soil mixing of K fertilizers. Fertilizing too large of a soil volume (i.e., broadcasting and mixing in the Ap horizon) could result in the excessive fixation of K in soils that fix K. Fertilizing too small of a soil volume may create a situation where the plant roots, which occupy only about 1 % of the soil volume, do not occupy enough of the fertilized volume for optimal fertilizer utilization.

Farmers in Southern Illinois need to utilize no-tillage and reduced-tillage management practices on highly erodible land. This limits their ability to mix fertilizer throughout the plow layer. For this reason, farmers need to know how to best fertilize no-till areas to enhance optimum K uptake by plants. By comparing different K placement methods we can determine if stratification, low K utilization, and reduced yields would warrant the periodic tillage of these soils.

The objectives of this study are to:

1. Determine the placement method that optimizes potassium utilization for corn and soybeans under no-till and reduced-tillage systems.

2. Determine effects of using K in starter fertilizer.

3. Determine effects of potassium rate and placement on potassium uptake and yields of corn and soybean.

4. Evaluate monthly sampling of selected plots at each location to measure variance among samples and seasonal effects on soil test K values.

5. Determine soil K stratification after three years of K placements.

6. Include a final report at the conclusion of this project to address each of the objectives stated above.

Materials and Methods

A field study was initiated in 1994 at the Dixon Springs Agricultural Center (DSAC) and the SIU Belleville Research Center (BRC). No-till (NT) and chisel tillage (CT) treatments were used in a corn-soybean rotation. Prior to study initiation, soil samples were taken from the top eight inches of soil at each location. The averages are shown in the table below. Lime, at a rate of four tons per acre, was spring-applied to the DSAC site before tillage was performed in 1994. A finely ground lime (54% passed #60 mesh sieve) was chosen for quick reaction.

Soil test levels (0-8-inch samples) at DSAC and BRC at the start of the study:

Location	1994 Crop	pH	P1	K	Ca	Mg	CEC
			lb/acre	lb/acre	lb/acre	lb/acre	meq/100g
DSAC	Corn	4.9	30	192	545	114	4.6
	Soybeans	4.6	26	250	177	125	4.1
BRC	Corn	7.1	56	194	3286	351	10.3
	Soybeans	6.9	74	249	3604	463	11.8

Tillage systems were blocked (whole plots) within each replication. A factorial arrangement of K rates and K placements plus a zero K check (subplots) were randomly placed within each tillage block in a split-split plot design. K rates consisted of 60, 120 and 180 lbs K₂0 per acre. K placement methods consisted of surface broadcast (BC), 8-10-inch surface band (BD) over intended row, surface dribble (DR) 6 inches to the side of the intended row, and starter (ST). The starter treatment placed 30 lbs of K₂O 2 inches to the side and 2 inches below the seed and was applied at planting. The remainder of the K rate with the starter treatment was surface broadcast. These treatments were applied to both corn and soybeans at DSAC and BRC with four replications per crop per location.

The K treatments were applied in solution form by dissolving potash (0-0-60) in water. For the chisel tillage system, fertilizer treatments were applied after chiseling but before secondary tillage to provide some mixing of the fertilizer with the tillage treatment. In 1994, a disk was used to incorporate the fertilizers, whereas in 1995 and 1996 a power tiller was used. This power tiller worked very well at incorporating the bands vertically with little spread horizontally.

1994 Study

Corn (Pioneer 3394 variety) was planted on May 13 at DSAC and May 23 at BRC at a seeding rate of 26,000 seed/acre. The corn had to be replanted at DSAC on May 27 due to poor stands. Soybeans (Pioneer 9443 variety) were planted on May 25 at DSAC and May 26 at BRC at a planting rate of 10 seeds/foot in 30-inch rows (175,000 seeds/acre).

Phosphorus fertilizer was broadcast across all plots prior to planting. Nitrogen was sidedressed on corn at a rate of 150 lbs N/acre at four weeks after planting. Corn ear-leaves were sampled from each plot at mid-Bilking, and the uppermost fully developed trifoliate leaves of soybeans were sampled at early bloom for nutrient analysis. Grain yields, grain moisture, and final plant stands were measured after physiological maturity.

1995 Study

Chisel tillage was performed at DSAC on April 6 and at BRC on June 14. The delay at BRC was due to over 17 inches of precipitation received during May (Figure 1). Corn (Pioneer 3394 variety) was planted on May 26 at DSAC and June 19 at BRC at a seeding rate of 26,000 seeds/acre. Soybeans (Pioneer 9451 variety) were planted on May 31 at DSAC and June 19 at BRC at a planting rate of 10 seeds/foot in 30-inch rows (175,000 seeds/acre). Phosphorus fertilizer was broadcast across all plots prior to planting. Nitrogen was sidedressed on corn at a rate of 150 lbs N/acre at four weeks after planting.

1996 Study

Chisel tillage was performed at DSAC and BRC in mid-May. Some delay was due to excess precipitation received during late April and early May. Magnesium was surface broadcast at DSAC at a rate of 36 lb/acre to alleviate possible low Mg supply in the soil.

Corn (Pioneer 3335 variety) was planted on May 24 at DSAC and May 31 at BRC at a seeding rate of 26,000 seeds/acre. Soybeans (Pioneer 9451 variety) were planted on June 4 at DSAC and May 31 at BRC at a planting rate of 10 seeds/foot in 30-inch rows (175,000 seeds/acre). Both the corn and soybeans were replanted at BRC on June 21 due to poor stands created by excessive rains in early June (Figure 1). Phosphorus fertilizer was broadcast across all plots prior to planting. Nitrogen was sidedressed on corn at a rate of 150 lbs N/acre at four weeks after planting. When starter N was applied, the sidedress N rate was adjusted accordingly.

Samples taken in the studies

In both 1995 and 1996, whole plant corn samples (from check plots and 120 lb K2O rates) were sampled between the 6 and 8 leaf growth stage, and whole soybean plants were sampled at about 12 inches of top growth from selected plots. These plants were weighed in the field, transported to the lab, then dried and weighed again in order to determine a wet weight, dry weight, and % moisture. After drying, plants were ground and analyzed for nutrient concentrations (%). Nutrient contents were determined by multiplying nutrient concentrations (in decimal form) by dry weights. Corn ear-leaves were sampled from each plot at mid-Bilking, and uppermost fully developed trifoliate leaves of soybeans were sampled at mid-bloom for nutrient analysis. Grain yields, grain moisture, and final plant stands were measured after physiological maturity.

Since the start of the study in 1994, soil samples have been taken each month from the 120 lbs K2O/acre plants along with the check plots for each crop at each location. These samples were analyzed for pH, P, K, Ca, and Mg levels.

In April, 1997, soil samples were taken from the 120 lb K2O/acre plots for each tillage and K placement treatment. Soil samples were collected across four two-row transects per plot. The samples were taken every 2 inches along the transects from mid-row to mid-row and at sampling depths of 0-2-inch, 2-4-inch, and 4-8-inch increments. These samples were dried, ground and analyzed for soil test K levels.

Results and Discussion

Because much of the interactions between K rates, K placement, and tillage are not significant, only main effects will be included in the discussion.

Corn Results

Whole Plant Data

At BRC, fertilizer K placement significantly increased whole plant dry weights in 1996, whole plant moisture in both 1995 and 1996, % K both years, and K contents in 1996, above the check treatments (Table 1). There were some significant differences between placement methods for whole plant dry weights and K contents. K contents are determined by multiplying the dry weight by the respective concentration (%). The banded treatment appeared to have the best early growth, even with no-till. In both 1995 and 1996, the broadcast K treatment had the lowest K concentration. This is an indication that the broadcast treatment was not as efficient in getting K into the plant early. By silking, there was no difference in the leaf K concentrations.

At DSAC, differences between placement effects on whole plant moisture, % K, and K contents were small but significantly better than the check treatments in most cases (Table 2). Both the broadcast and banded treatments had slightly lower K contents than dribble or starter treatments.

There were no interactions between tillage and either K rate or K placement.

Ear-leaf data

Increasing K rates from zero to 180 lb K2O/acre significantly increased ear-leaf K concentrations in all six site-years (two locations X three years) (Table 3). In five out of six site-years, the relationship was linear and positive. The other site-year showed a quadratic but still positive relationship. The zero and 60 lb K2O treated corn had leaf K levels below the sufficiency range of 1.8% to 3% at BRC but not at DSAC.

There was an antagonistic relationship between K concentration and Mg concentration (data not shown). The Mg levels were approaching a critical level, especially with high K rates, but no visual Mg deficiency symptoms appeared in the field. At DSAC, Mg levels in the ear-leaves in 1996 were double the levels in 1995 and were above what is normally considered a critical range for ear-leaf Mg. This is attributed to the fertilizer Mg applied to the area in 1996; however, Mg levels still decreased with increasing K rates, but not below critical plant levels.

The effects of K rates were more pronounced than the effects of K placement. In five out of six site-years, placement method increased ear-leaf K levels above the check, but there were no significant differences between placement methods (Table 4). The leaf K levels are within the sufficiency range for corn, except for the checks.

Again, there were no interactions between tillage and either K rate or K placement.

Yield Data

Corn yields were good to excellent in five out of six site-years, even with the late plantings in many circumstances. There were small but significant increases in corn grain yields as K rates increased for three out of six site-years (Table 5). However, when averaged across K rates, K placement had no significant effect on corn yields (Table 6), even though most had greater yields than the check. Only at BRC in 1996 did the method of K placement significantly increase yield over the check plots, but even then there was no significant difference between placement method.

There was little difference between tillage systems and no interactions with either K rate or K placement (Table 7).

Starter nitrogen plus K data

In 1995, a treatment of starter nitrogen plus potassium was added to compare to the starter K treatment without N. At BRC, the nitrogen in the starter failed to significantly increase grain yields or K uptake as measured by %o K in whole plants and ear-leaves (Table 8). There were significant increases in whole plant dry weights and %N. By silking, there were no differences measured in the ear-leaf concentrations of N or K. Late planting in June into warm soils may have diminished effects of starter N in both 1995 and 1996 at BRC.

As at BRC, starter treatment effects at DSAC were small. Only N concentrations in whole plants were significantly higher with N in the starter than without, across both years (Table 9). In 1995, the N in the starter was responsible for increased whole plant dry weight, but by silking the effects of the N in the starter were negative and led to a decrease in yield. Again, the lateness of planting may have played a role in the response to starter N. Tillage failed to play a role at either location across the two years of study.

Soybean Results

At BRC, whole plant dry weights and moisture were relatively unaffected by K placement (Table 10). Moisture contents, K concentration, and K contents were higher with K placement than check plots.

At DSAC, addition of K fertilizer increased whole plant dry weights and moisture, K concentrations, and K contents (Table 11). Differences between placement method were less than the comparison to check plots. In 1995, the broadcast and banded treatments had bigger plants (as measured by dry weights) and higher K contents, with the banded treatment being better yet in 1996. These results are similar to the BRC soybean results, but opposite those with corn at DSAC. This may be an indication that soybeans may be more sensitive to close placement of K fertilizers either with the dribble or starter K treatments.

Trifoliate-leaf data

Increasing K rates increased trifoliate-leaf K concentration, in five out of six site-years (Table 12). At BRC, the response was quadratic in nature, whereas the response at DSAC was linear. Increasing K rates increased leaf K concentrations but decreased Mg concentrations (data not shown).

Each K placement method increased trifoliate-leaf K concentrations above the check plots (Table 13). There was, however, very little difference among the various placement methods for leaf K. There were also few interactions between tillage and K rates or K placement.

Yield Data

Soybean yields were average to high across the three years and two locations of the study (Table 14). The highest soybean yield in many cases occurred at the 60 lb K2O rate and declined at the higher rates, an indication of the sensitivity of soybeans to high rates of fertilizer (salt).

K placement had little or no significant effect on grain yields (Table 15). Again, the importance of K rate (especially above the check treatment) was more critical than the impact of K placement. At DSAC in 1996, the banded treatment had a slightly lower yield than the other placements. Banding increasingly higher K rates decreased yields, especially with CT (Table 16). This may be another indication of salt-type injury to soybeans where high K rates have been applied and incorporated in close proximity to the seed. The very dry August (Figure 1) experienced at DSAC in 1996 may have resulted in the salt injury to soybeans that led to yield losses.

Starter nitrogen plus K data

The N added to the starter increased early-season growth but failed to provide any significant yield benefit. Adding N to the starter treatment at BRC increased whole plant dry weights but not N or K concentrations (Table 17).

At DSAC, there was no significant increase in yields with starter+N treatments and little or no benefit to early-season growth (Table 18). Only whole plant N concentrations were affected and only in 1996.

Monthly Soil Samples

Monthly soil sample K levels at BRC were highly variable but revealed some interesting trends (Figure 2). In June of 1994, soil test K levels for the check plots were 250 lb when averaged across both corn and soybean areas. By September, that level had fallen to under 170. This could be attributed to plant uptake and possibly some fixation by the soil. The broadcast (BDCT) treatment of 120 lb K2O increased the soil test level approximately 30 lb by the June sampling and continued to be higher than the check throughout the growing season. After September 1994, soil test levels began to increase, corresponding to about the time of leaf drop by the crops and at a time when soil moisture was rising. The soil test levels remained high throughout the winter months until April, when there was a dip. Interestingly, April was a very dry month at BRC and K fixation may have occurred. April is also a time of rapid weed growth, which could remove K from the soil. After April, soil tests began to rise, even with the check plots. This shows the seasonality of soil testing for K and is probably a result of K release from previous crop residues breaking down, from fixed K being released, or from weeds after they are killed either mechanically (via tillage) or chemically.

The lines appear to have more spread between them through 1995 and 1996 as more K is applied to the broadcast treatments and no K to the checks. The lowest soil test values usually occurred in September, corresponding to the point of maximum plant uptake. The highest soil test values are difficult to determine because of a great diversity in sample readings.

Soil tests for K at DSAC (Figure 3) showed similar trends with the broadcast line separating from the check line over time and with subsequent application of fertilizer K. Peaks often occurred right after planting even where no fertilizer K was applied, indicating a release of K from plant residues or weeds resulting from tillage or chemical application (for weed control). Soil moisture and temperatures may also play a role in soil K release from clay particles. The valleys were often associated with maximum K uptake by the grain crops being grown.

Final Soil Test K Levels

In April of 1997, soil samples were taken at BRC to determine the effects of placement on soil nutrient stratification to a depth of 8 inches and across a 5-foot transect running perpendicular to the corn rows. The transects started on a mid-row and crossed two corn rows for a total span of 60 inches with sampling points every 2 inches across the transect. Four such transects were taken from each plot, with the soil composited by layer and by position across the transect.

Nutrient stratification was evident with both the chisel and no-tillage systems (Figures 4 and 5). However, no-till showed higher levels of Kin the 0-2-inch depth than chisel, but the chisel treatment had higher K levels in the 2-4-inch depth. This shows the degree of mixing of applied K into the soil with tillage, whereas the no-till system results in higher K levels at the surface.

All of the placement methods-even the broadcast treatment-had higher K levels near the row than mid-row. This is an indication that K is being redeposited near the row after the plants mature. This could be associated with leaching from the plant after physiological maturity, from breakdown and release of K from the plant after harvest, or both. The banded and dribble treatments with no-till show very sharp peaks in the 0-2-inch depths due to lack of incorporation via tillage but may also indicate a less efficient system for K uptake later in the season when surface roots are less active.

Summary

The broadcast K treatment often had the lowest K concentration in whole plants early in the growing season. This is an indication that the broadcast treatment was not as efficient in getting K into the plant early. By silking, there was no difference in the leaf K concentrations for the various placement methods.

Increasing K rates almost always increased ear-leaf concentrations of Kin corn and trifoliate-leaf concentrations in soybeans. Increasing K rates also seemed to increase early plant growth and yields, but more for corn than for soybeans. Soybeans seemed to maximize yields at the 60 lb K2O rate, whereas corn often showed higher yields (although not always significant) with K rates of 120 and 180. Corn usually requires more K than soybeans do, which would explain its responsiveness to K.

With corn, the placement method of K was not that critical (certainly less critical than rate), even for no-till across the six site-years of this study. Some problems occurred in 1996 with the longterm banding of K fertilizers in 10-inch bands over the intended rows of soybeans. This is in contrast to a yield benefit from banding at other locations or crops. Weather problems of a wet spring followed by very dry conditions late in the season may have contributed to the results.

Monthly soil samples for K showed much variability from month to month, but some seasonal trends could be identified based on fertilization, plant uptake, and plant release. Trends associated with K fixation by the soil could not be well identified.

Placement method of fertilizer K and subsequent tillage can have a pronounced effect on spatial patterns of residual soil K. High concentrations of residual K were usually observed with banded or dribbled K placement methods. Redeposition of K from plant tissue also contributed to elevated soil K near planted rows. These bands of increased K should be included in routine field soil sampling for fertility assessment with special consideration in the total soil fertility management program.