In the early days of fertilizer use on the Canadian Prairies phosphorus was the only game in town- mostly 11-48-0. “If in doubt use 11-48,” was the chant.
Much work was done to find the amounts to use for various crops and soils and the best way to put it on. We thought we knew it all. But, as so often happens we really had it all wrong and a major change in thinking has happened in the past few years.
We did articles on some of the new ideas in March and April, 2015, issues of Grainews for those who care to look back. In this piece we will try to sum it all up and look to the future.
The old story
The early work showed clearly that the only way to apply P was to put it with the seed. Countless experiments were done to find the most economic rate for each crop and soil situation. Eventually soil testing came along so more accurate recommendations were possible.
The old story was that “a little dab’ll do ya” and phosphorus should be placed with the seed. High tech experiments using radioactively labelled P32 fertilizer showed clearly that only about 30 per cent of the actual fertilizer P ended up in the crop. The old story said that the other 70 per cent of the fertilizer P got “sucked up” by the soil, never more to be accessed by a plant root. Phosphorus fertilizer does not “gas off” and leaching is usually not an issue so it must be sucked up in stubborn soil minerals that hang on to it all.
So, on and on we went with more on more experiments. Why was flax so dumb that it would not respond much to P fertilizer? Endless laboratory fractionations were made and complicated mathematical models derived so we could get to the bottom of what was going on.
Conceptual models like the one shown below were used and combined with detailed laboratory fractionations. The idea was to use mathematical models so the entire picture of how Mother Nature operated could be deduced. Good luck!
In the process, several very good field experiments were done in Saskatchewan, Manitoba and North Dakota with very large rates of one-time broadcast applications of P. Those experiments showed clearly that residual responses were real. But work continued to refine the model so every situation could be dealt with. Many papers were published and many conferences were held to deal with ever more minute details of “the model.”
But what about the farmer?
The new story?
The new story for me started just a few years ago when Renaud Lemke of Agriculture and Agri-Food Canada Swift Current took the simple long term N/P experiments from Swift Current and did some simple arithmetic.
His math showed that 98 per cent of the fertilizer P that had been added over the years could be accounted for by the extra P in the wheat the farmer hauled off to the elevator.
The icing on the cake was a piece by Johnny Johnston and Paul Poulton of the famous Rothamsted Research Center near London,UK, and Paul Fixen of the International Plant Nutrient Institute working in the U.S. “ Better Crops with Plant Food . Vol. 98, 2014, Pages 22-24.
They came up with the simple Four Pool concept as shown in Figure 2 (see below).
This direct quote from the above piece says it all:
“Data from vastly different soils located on two continents, and from both controlled experiments in England and derived state-wide aggregated data in the U.S., were merged to evaluate P use efficiency. The data suggest that there is an underlying “simple rule” for the behaviour of plant-available soil P in these soils, which can be related to a four-pools concept of inorganic soil P.”
I have added a few notes to Figure 2 to explain how it works. Pool 1 is very small and interacts on a short-term basis with Pool 2. Pools 1 and 2 together are what we measure when a sample is sent of to a lab to get “available P.” It is the current account in a money analogy. Pool 3 is the operative savings account, which transfers to the current account in agronomic time. The fourth Pool is the basic soil P minerals that react in geologic time.
So, the conceptual model can be made much simpler and can be used to give the farmer what is needed.
It all boils down to simple math. Take off more P than you add and the soil test level goes down and so does crop yield. Add more P than you take off and the soil test P goes up and so do future crop yields. The soil test is what will keep us honest. The P soil test is to a farmer as the Hemoglobin A1C blood test is to a diabetic.
Manitoba has taken the lead in applying the new story and have worked out the numbers for the P that is taken off the field with various crops. John Heard from Manitoba Agriculture, Don Flaten from the University of Manitoba, and Cindy Grant from AAFC Brandon (now retired) were the folks that put it together into new recommendations for fertilizer P.
What is needed now is a good summary table of all the various P soil tests that are in common use and what the numbers are for Very Low, Low, Medium, High and Very High for each of those tests.
It is possible to apply one large P rate to last several years, and broadcast is OK as long as it does not wash off the surface in a big rain. Banding is OK, but rates with seed are limited. New information is coming out on that front also. Side- or mid-row band at higher rates is good but at very wide spacing the distribution can be a problem, and a problem for future soil samples.
Soil test P varies greatly on rolling land and hilltops (eroded knolls ) are notoriously deficient. Variable rate should fit the bill for P provided you can get the right soil samples to get the right answer for each zone in a field.