The Phosphorus (P) and Potassium (K) Soil Testing and Fertilizer Recommendation System in Ontario

Agdex#:	553
Publication Date:	04/97
Order#:	97-005
Last Reviewed:	09/97
History:
Written by:	Murray Miller - University of Guelph; Keith Reid - Soil Fertility Specialist/OMAF; Ken Stevenson - Ridgetown College of Agricultural Technology

Table of Contents

1. Introduction
2. Method of Extraction
3. Fertilizer Requirement
4. Summary

Introduction

The Ontario soil testing and fertilizer recommendation system for Phosphorus (P) and Potassium (K) has a scientific basis as strong as any soil testing program in North America and stronger than most. Since its beginning in the 1940s, the system has been continuously updated as new extraction methods, new hybrids and new management practices have become available.

The Ontario Soil Management Research and Services Committee (OSMRSC) is responsible for the plant nutrient and limestone recommendations for all crops grown in Ontario. The OSMRSC is made up of research, extension, grower and fertilizer industry representatives who meet annually to consider changes in fertilizer recommendations. Suggested changes in fertilizer recommendations are considered first by the appropriate subcommittees and are then passed on to OSMRSC for final approval and inclusion in the OMAF crop recommendation bulletins (e.g., Publication 296 for Field Crops, Publication 363 for Vegetable Crops). In addition, the approved changes are incorporated into the recommendations from the OMAF accredited soil test laboratories.

There are two components that are vital to the development of an effective soil testing service:

1. The method of extraction of available nutrients
2. The response to fertilizer in relation to the measured available nutrient

| Top of Page |

Method of Extraction

An extraction method must be selected which gives the best indication of the amount of nutrient in the soil that is available to crops for the soils in the region in which the test is to be used. This can best be done by collecting a large number of soils representative of those in the region and growing crops on these soils in the greenhouse. This eliminates the effects of climatic variation from site to site. The amount of nutrient absorbed by the crop is then correlated to the amount extracted by different methods to determine the extractant which is the best indicator of the available nutrient across the range of soils.

Phosphorus

Phosphorus exists in the soil largely as P adsorbed on iron and aluminum oxides or in association with calcium. This P is in equilibrium with P in solution. Phosphorus also occurs in organic forms and may be released by microbial activity. To be effective, an extractant must remove a constant proportion of the phosphorus that is available to plants from different soils. Soil pH and the presence of CaC0₃ (lime) in the soil have a major influence on this relationship.

When the soil testing program was first established, the Bray P2 extractant (.05 N NH₄F and 0.1 N HC1) was used. However, it was found that this extractant was not providing accurate results on high pH soils. In 1967, several soil extractants were evaluated against the P absorbed by a crop of wheat and two crops of red clover from 96 soils ranging in pH from below 6 to 8.0. The NaHC0₃ extraction developed by Olsen for calcareous soils was found to account for 89% of the variability in P absorption whereas the Bray P2 test accounted for only 73%. The Olsen test was superior to all other extractants tested at all pH ranges. This test was introduced into the laboratory in 1969. The samples of soil from fertilizer response trials conducted earlier had been retained and were analysed using the NaHC0₃ extractant and the fertilizer requirement tables adjusted as required.

A further evaluation of soil P extractants was conducted in 1988. The P extracted by five extractants was related to that absorbed by corn from 88 Ontario soils. The results are presented in Table 1. It is clear that the NaHC0₃ extractant is a better test than the Bray P1 for soils in all pH ranges.

Although the Bray P2, AB-DTPA and Mehlich 3 were superior to the NaHC0₃ on low pH soils, over all soils the NaHC0₃ test was best. Thus, phosphorus requirements based on other extractants are much less reliable than those based on the NaHC0₃ test which is the test that has been calibrated on Ontario soils.

Note: r2 is the proportion of the variability in P absorption that is accounted for by the amount of P extracted by the extractant.

Potassium

Ammonium acetate (NH₄OAc) is the most commonly used extractant for available potassium. It extracts the potassium in solution plus that held on the soil cation exchange complex. This extractant has been used in Ontario since the soil testing program began. It is the test against which fertilizer response data have been calibrated. Other extractants such as the Mehlich 3 have been found to be as effective as NH₄0Ac for potassium. However, because the Mehlich 3 is not as good as NaHC0₃ for phosphorus, it has not been introduced into the Ontario soil testing program. 

Figure 1A. Potassium Absorbed by Alfalfa from 67 Ontario Soils in the Greenhouse as Influenced by Ammonium Acetate Extractable K (NH₄ OAc - K) and a)

Figure1B. Cation Exchange Capacity (CEC) or b) % K Saturation 

Source: M. Sc. Thesis. Liangxue Liu. University of Guelph, 1989
Note: Proportion of variation in K uptake accounted for (R2) by NH₄0Ac-K alone = 0.849, R2 for NH₄0Ac-K+CEC = 0.850, R2 for NH₄0Ac-K+%Ksat = 0.850 

Some soil testing programs use the cation exchange capacity (CEC) to modify the potassium requirement at a given NH₄OAc- extractable value. New York State decreases the potassium fertilizer requirement as CEC increases; whereas, Ohio State increases the fertilizer required with increasing CEC. Studies in Ontario have shown clearly that neither CEC nor % K saturation (% of cation exchange capacity occupied by K) significantly improve the prediction of available K (see Figure 1). Because of these results, CEC is not included in the Ontario soil testing program.

| Top of Page |

Fertilizer Requirement

The second step in developing a soil testing system is to calibrate the selected soil extractant against response to added fertilizer over a wide range of soils in field trials. There are two approaches, however, on how these calibration data are used. 

In some American states, and with many private laboratory recommendations, the calibration data are used to identify the soil test levels where no further response to added fertilizer is expected. Fertilizer is recommended in this system to build up the soil to this "critical" soil test level, or to maintain the soil test at or above this level. This "Build-Up and Maintenance" system will tend to give higher fertilizer recommendations than are required for maximum crop yields in any given year. 

The Ontario fertilizer recommendations use the calibration data to identify the most economic fertilizer application rate (MER) at any soil test level. This rate will be less than that required to produce the maximum crop yield but will produce the most economic yield (MEY). Beyond this point the cost of the additional fertilizer to produce extra yield is greater than the value of the extra yield. The precise rates can be calculated for any combination of crop value and fertilizer price, but the variation in MER is quite small unless crop species or fertilizer costs change markedly. On soils testing low in available P or K, the MER will result in an increase in soil test P and K although at a slower rate than with the "Build-Up and Maintenance" system.

Between 1954 and 1960 major effort was devoted to calibrating the soil extractants in use at that time (Bray P2 for phosphorus and NH₄OAC for K). Field trials were conducted across Ontario for major crops. The data from these trials were used to establish fertilizer requirement tables. Requirement tables were also established for crops not included in the test based on general understanding of the response compared to those crops for which data were available. Since their initial establishment, the fertilizer requirement tables have been adjusted on numerous occasions based on information from response trials using current hybrids and management practices. Trials were conducted on alfalfa-based forages in southern, central and eastern Ontario between 1980 and 1986. Twenty-one sites included phosphorus fertilizer rates and 40 sites included potassium rates. Premier cultivars from the recommended list were used under a three-cut management system at most sites. Earlier studies had determined the most economic fertilizer rates for forage establishment so, in these trials, fertilizer rates were applied in the fall of the establishment year. In addition to these trials, experiments were conducted to compare the response of trefoil, red clover and timothy to that of alfalfa. Based on results of these trials, the fertilizer requirement tables for forages were adjusted.

As indicated above, the fertilizer requirement tables used in the Ontario Soil Testing Service have been adjusted frequently since they were originally established. During the past 10 years or so, several sets of experiments have been conducted to check the validity of the P and K recommendations for corn and soybeans. The results are summarized in the following sections.

Source: Unpublished Data. Stevenson, C. K. Ridgetown College of Agricultural Technology

Source: Unpublished Data. Stevenson, C. K. Ridgetown College of Agricultural Technology

A set of experiments was conducted by Philom Bios and Dow Elanco with corn in 1993 and 1994 in Oxford, Waterloo, Huron, Perth and Kent counties on fields for which soil P tests indicated a requirement of from 20 to 70 kg P205/ha. Fertilizer P was banded at rates of 0 to 60 kg P205/ha at 10 kg/ha intervals. The maximum economic yields and the fertilizer P required to achieve them were calculated from regression analysis of the data. The results are shown in Table 3A and Table 3B. On 12 of the 15 sites, the P required for MEY was less than that recommended, whereas, on the remaining 3 sites it was more. On average, the amount recommended was more than double that required (41 vs. 18 kg P205/ha). A soybean crop was included in the 1994 Dow trials. As shown in Table 3B, there was a response to fertilizer P on only three of the sites, and at none of the sites was the most economic rate as high as that recommended.

Seven corn trials were conducted by the Department of Land Resource Science between 1989 and 1991 on sites with a low to medium P test. Fertilizer P was banded at rates to provide one-half, one and two times that recommended by soil test. The results are presented in Table 4. On none of the seven sites was the yield increased significantly by applying two times the recommended rate. On five of the seven sites, the recommended rate gave the highest yield while on the two sites in 1990 there was no yield response in spite of recommendations of 50 and 90 kg P205/ha.

Source: Unpublished Data. Labrecque, D. Philom Bios. Inc. and Dow Elanco Canada, Inc.

Source: Unpublished Data. Labrecque, D. Philom Bios. Inc. and Dow Elanco Canada, Inc.

The most recent trials were conducted in 1993 and 1994 at the Elora Research Station. Corn and soybeans were grown on a set of plots previously used for soil test calibration that ranged in soil P test from 3 to 35 ppm. Figure 2 shows the response of corn and soybeans averaged over the two years. There was no increase in corn yield as soil test values increased beyond about 10 ppm. In 1994, a band application of 20 kg P205/ha applied to plots with a soil P test of about 17 ppm did not increase yields. Soybean yields were not influenced at all by increasing soil P test even when the lowest soil test was 3 ppm.

Figure 2. Yield of Corn and Soybeans as Influenced by NaHC0₃ - Extractable P Levels – Average of 1993 and 1994

Source: Unpublished Data. Lauzon, J. D. and M. H. Miller. Department of Land Resource Science, University of Guelph.
Note: Proportion of variation in yield accounted for by soil test (R2) for corn = 0.77, for soybeans = 0.10 

These results clearly indicate that P recommendations based on the NaHC0₃ extraction are at least sufficient for maximum economic yield with currently used hybrids of both corn and soybeans. Although the yield levels varied from trial to trial, the recommended rate of P is adequate even for very high yields. This is evident from the results of the RCAT trials as well as from the Guelph trial at Woodstock in 1991.

Potassium

The corn and soybean trials conducted by RCAT included potassium as a variable as well as phosphorus. The results are presented in Table 5A and Table 5B. With corn the K2O recommended by soil test was sufficient for MEY in 19 out of 22 trials. In 10 of the 22 trials, more K2O was recommended that required for MEY. Averaged over the 22 trials, the recommended rate was two times that required. With soybeans, the K2O recommended by soil test was sufficient for MEY in 16 out of 19 trials. In 11 of the 19 trials, more K2O was recommended than required for MEY. Averaged over the 19 trials, the recommended rate was three times that required.

Source: Unpublished Data. Stevenson, C.K. Ridgetown College of Agricultural Technology.

Source: Unpublished Data. Stevenson, C.K. Ridgetown College of Agricultural Technology.

| Top of Page |

Summary

 The fertilizer recommendations from the Ontario Soil Testing Service are based on very extensive research in both selection of the most appropriate extractant and in calibration of that extractant against fertilizer response in the field. The P and K extractants currently in use are the only ones that have been properly calibrated for Ontario. Recommendations based on other extractants will not give as reliable a fertilizer recommendation. Current P and K recommendations for corn and soybeans are on average greater than that required for most economic yield. This is due, at least in part, to the fact that some P and K is recommended, for maintenance purposes, at soil test levels at which responses have not usually been found.

No soil testing system is perfect. There will always be years and/or soil conditions where crops give a response to rates of fertilizer greater than recommended. As shown in recent trials the frequency of this occurrence is low, but farmers are encouraged to do some testing on their own farms by leaving strips where either more or less P and K than that recommended is applied.

Murray Miller is with the Department of Land Resource Science, University of Guelph. Keith Reid is Soil Fertility Specialist, Agriculture and Rural Division, OMAF. Ken Stevenson is with the Ridgetown College of Agricultural Technology.

| Top of Page |