Crop Sciences logo University of Illinois at Urbana-Champaign logo

Agronomy Day 2009

Home Welcome(Hoeft) Welcome(Dunker) Field Tour Presentations Tent Displays Credit & Thanks Sponsors
Tour A

How much is too much and how little is too little for phosphorus and potassium?

Fabián G. Fernández
Fabián G. Fernández
Assistant Professor Soil Fertility & Plant Nutrition
Department of Crop Sciences
217-333-4426
fernande@illinois.edu

In spring 2008 phosphorus (P) wholesale price in central Florida was $1,070/ton of DAP and potassium (K) wholesale price in the Midwest was $705/ton of potash. In spring 2009, P price was down considerably at $305/ton, while K price was holding steady at $645/ton. The high prices have discouraged producers from applying P and K in preparation for the current growing season, and many are still undecided about how much to apply this coming fall. The most common questions about P and K have recently been: How little can I apply and still get good yields? How much would my soil test levels drop if I don’t apply a full rate, or altogether eliminate the application? Is there a placement technique that would allow me to get the same yield, but with less fertilizer? In this era of high prices and high price volatility, conscientious planning and sound strategies to continue to produce high yields while maintaining profitability is critical.

How much to apply?
This will depend on where your soil test levels are. It can be difficult to know which direction to go if you don’t know where you are. The way to find out your starting point is by collecting soil samples and analyzing them for P and K. In the past two years we conducted a soil fertility survey by sampling almost 600 random locations in more than 50 counties in Illinois. For P, approximately 55% of the soils were above the level where additional P is not recommended; 35% of the fields were below the critical level, or the point at which yields are maximized; and 10% of the samples were at maintenance, or the level at which applications should replenish what is being removed by the crop  (Figure 1). For K, 26% of the fields were above the point at which additional K is not needed to maximize yield; 37% of the soils were below the critical level; and 37% of the samples were at maintenance levels (Figure 2). These data illustrates the importance of knowing your test levels to help you determine whether an application is needed.

How much would the test level drop?
This is again dependent on where your soil test level is, especially for P. The higher the value the more rapid the decline in test levels if no fertilizer is applied. In general soil P levels will drop 9 lb P acre-1 year-1 if no P is added when test levels are very high (around 140 lb P acre-1). If test levels are high (around 70 lb P acre-1), the drop will be approximately 5 lb P acre-1 year-1. If test levels are at the critical level (around 30 to 45 lb P acre-1), the drop will be approximately 2 to 3 lb P acre-1 year-1. Potassium is typically too variable, but in general it can be expected to drop 13 lb K acre-1 for each year that K is not applied. Conversely, P levels will increase on average 1 lb per 9 lb of P2O5 applied and K levels will increase on average 1 lb per 4 lb of K2O applied. All these are approximate values to assist you in making general decisions, but should not be considered a substitute for testing soils periodically to track changes under your specific soil and management.    

Is there a more effective placement technique?
While P and K can be fixed (become unavailable for the crop) by some soils, this is not the case for most soils in Illinois. Thus, limiting soil contact with the fertilizer to reduce the potential for fixation is typically not necessary. Subsurface band applications normally result in increased soil test levels in the site of application. Some have taken this as evidence that subsurface applications are more effective because a maintenance application rate builds up the soil test level. In reality, this build up is because the crop is taking nutrients from the entire soil volume, not only the fertilizer band. Our research has shown no yield improvement for corn (Figure 3) or soybean (Figure 4) with subsurface band applications compared to broadcast applications. 

Distribution of soil phosphorus levels. The horizontal bar marks the range of values at which maintenance applications of phosphorus are recommended.Figure 1: Distribution of soil phosphorus levels. The horizontal bar marks the range of values at which maintenance applications of phosphorus are recommended.
Distribution of soil potassium levels. The horizontal bar marks the range of values at which maintenance applications of potassium are recommended.Figure 2: Distribution of soil potassium levels. The horizontal bar marks the range of values at which maintenance applications of potassium are recommended.
Two-year average corn yield response to phosphorus (P) and potassium (K) application in strip-till deep (6 inch) placement (ST-DP), No-till deep placement (NT-DP) and no-till broadcast (NT-BC). Response averaged across P and K rates. Bars followed by the same letter are not significantly different (p>0.05).Figure 3: Two-year average corn yield response to phosphorus (P) and potassium (K) application in strip-till deep (6 inch) placement (ST-DP), No-till deep placement (NT-DP) and no-till broadcast (NT-BC).  Response averaged across P and K rates. Bars followed by the same letter are not significantly different (p>0.05).
Two-year average soybean yield response to phosphorus (P) and potassium (K) application in strip-till deep (6 inch) placement (ST-DP), No-till deep placement (NT-DP) and no-till broadcast (NT-BC). Response averaged across P and K rates. Bars followed by the same letter are not significantly different (p>0.05).Figure 4: Two-year average soybean yield response to phosphorus (P) and potassium (K) application in strip-till deep (6 inch) placement (ST-DP), No-till deep placement (NT-DP) and no-till broadcast (NT-BC).  Response averaged across P and K rates. Bars followed by the same letter are not significantly different (p>0.05).
change and challenge