If you are serious about a need for better P and K recommendations, it
is probably better to do a six-inch test for those nutrients.
Soil testing in Western Canada began in a serious way in the 1960s — the same time as fertilizer use began to be a significant part of farm operations. Despite 50 years of experience, there are still a lot of misconceptions about soil testing.
Soil is a complex item. Geology, chemistry, physics and biology interact in complex ways, and our understanding of the whole beast is sometimes sketchy. Testing a soil is not like pulling the dipstick from a crankcase. It is not a matter of measuring a level and “topping it up.”
To have a soil test that works, you need to conduct a large series of experiments over a particular soil climatic zone and for the major crops in that zone. These experiments must be on soils that test high, medium and low by the test we propose to use. If responses to a nutrient are high on a low-testing soil and low (or non-existent) on a high-testing soil, then we have a useful soil test. If a soil test has not been calibrated, it might be based on good theory, but where is the practical result?
Provincial labs located at the soil science departments of the three major Prairie universities performed many of the first soil test experiments. Their soil test based recommendations were vetted by a provincial fertilizer council comprised of university, fertilizer industry and provincial and federal government reps. Soil test benchmark tables were published and were the basis of recommendations for whatever lab operated.
Governments and universities are not much good at running businesses, and soil test labs are a business, so the provincial labs have all disappeared. As private labs have taken over the function, there is no single benchmark source and it is more difficult to evaluate the recommendations made.
The good news is that there is now a cadre of private agrologists who can interpret the results from a lab and prepare a recommendation based on local knowledge and experience. Former students of University of Saskatchewan who are now doing private agronomy have taught me a thing or two based on local experience. Here’s the latest on sample depth.
The original sample depth regime was to take separate samples from zero to six inches, six to 12 inches, and 12 to 24 inches. That was a very good sampling regime. Zero to six was used for phosphorus (P) and potassium (K) recommendations. The sum of the three depths was used to make nitrogen (N) and sulphur (S) recommendations. The limitation of the three sample depths is the practicality of taking the three depths and doing it at enough sites on a quarter section to get a reliable average sample.
So along came the one-foot sample. I was the one who first championed the foot sample in Saskatchewan. It was based on an extensive sampling program where 100 cores of various depths and core size were taken on many fields over various agronomic conditions. That study showed that we were kidding ourselves if we thought much less than 30 cores could be used to get a field average. For N, a two-foot sample was better but a good one-foot sample was considered better than a poor two-foot sample because of insufficient sample numbers.
The one-foot sample is a very crude instrument for making P and K recommendations, but the thinking was that soil testing was mainly done to refine the N rate. P is usually applied at a low rate with the seed and K deficient areas were mostly known, so the importance of soil testing for P and K is not as crucial as N.
But if you are serious about a need for better P and K recommendations, it is probably better to do a six-inch test for those nutrients.
For S, the two-foot sample was originally considered best because S is mobile in the soil. The concept works fine in garden patch agriculture of small plots, but falls down on a field basis. With a two-foot sample, you can get down into subsoil gypsum, which may give you a false impression of S availability. Even a few cores with subsoil gypsum will tell you there is lots of S when much of the field might be deficient. If a one-or two-foot sample tells you S is deficient, then you really do need it. But if that same sample says there is enough, that may not be true over the whole field. For canola, most farmers, including me, just apply S without any reference to a soil test.
J. L. (Les) Henry is a former professor and extension specialist at the University of Saskatchewan. He farms near Dundurn, Sask.