Baseline fertility recommendations have been around for years and have stayed the same over time. Keep a certain pH; 1.2 pounds nitrogen per bushel for corn; ranges of phosphorus and potassium that are considered low, medium or high by looking at the number on a soil test; and “removal rates” based on element levels per bushel.
Many of these needs were based on “efficiency factors” that may have estimated your crop could find 70% of available nitrogen, 30% of available phosphorus and 40% of available potassium.
These were decent recommendations for their time, but very general in nature, and led to blanket applications of nutrients that could mean overapplication based on unrealistic yield goals and non-representative soil tests. Over the years, farmers got the impression that more fertilizer equaled more yield potential.
Fast forward to 2010. Things have changed in a hurry in agriculture. Many of the input costs have increased by two to 10 times. Equipment has gotten much better, especially planters. Hybrids and varieties have many advances, and seed treatments, along with the planter improvements, have made near-perfect stands possible.
Weed-control options are many, so competition for nutrients should not be a problem. GPS has brought evaluation tools to field level for every farmer that wants to learn. You can now know more about your own performance than ever before and make better decisions for nutrient management.
Let’s talk a little bit about soil testing. This is an excellent tool with many uses, but widely misunderstood. When grid sampling came out in the 1980’s, it was thought to be state of the art. (This was before invention of the yield monitor.) The good part of grid sampling was it got farmers to soil test. Many were doing very little testing at the time, and what was done was not representative, perhaps one sample for 40 acres.
One bad point of grid sampling was that it preceded digitized soil surveys, so it ignored soil types. It cost a lot of money to sample on 2.5-acre grids, so most farmers were asked to live off that one test for 5 years, a risky move with fertilizer dollars.
This system ignores two of the most important elements of a representative soil test: The farmer and the local agronomist or CCA. No one gets asked where to sample except the computer. Give it a boundary and it will tell you where to drive the ATV to for sample No. 1.
And once the ATV is parked, the sampler takes five or six samples, puts it in a bag and sends it to the lab. The lab runs the analysis and magically, when returned to the software, those results now represent a 2.5-acre grid.
But, was that grid adequately represented? No! A 10-by-10-foot area was sampled, and there are 1,090 of those in 2.5 acres. And then, besides being non-representative, variable-rate nutrients and lime are applied, creating more variability than was present to begin with — depending on the mathematical equations chosen in the software.
This type of nutrient management was widely used as a marketing program by fertilizer dealers to differentiate them from their competition, justify variable-rate technology equipment and develop another business enterprise. Could variability be shown? You bet! Could the products be varied? You bet! Could results be analyzed and best management practices that were more economically and environmentally sound be established? No. Another case of “more is considered better.”
Now, let’s look at the actual soil test. Guidelines assumed all soil tests are created equal, and they are not. Ten different labs can give 10 different results, depending on the methodology used at the laboratory. Not wrong, but different. The challenge for the agronomist and CCA is to correlate consistent lab data, from well-represented samples, to crop growth. But, it’s extremely difficult to compare one lab’s results to another.
The most variability probably comes at the field level and the person pulling the sample. Close-core inspection, so the area sampled is representative, is a must. You can mix two soil types together in the same bag and the lab will return an analysis to you, but it may well represent something that is not even present in your field. You make an application based on it, get no results and create variability.
I’ve also seen samples taken by people that didn’t realize the depth needed to be strictly controlled at 6.67 inches, unless otherwise directed to the laboratory. Analysis equipment only can measure parts per million, and the industry assumes one acre of soil 6.67 inches deep weighs 2 million pounds, thus the “pounds-per-acre” figure on the test.
Does it really mean pounds per acre? No, it’s just a number to correlate to crop growth. If the sample depth is anything different than 6.67 inches, and not noted to the lab, poor representation results.
Good representative soil samples on a regular basis — at least every other year — can help you build a good database. This will allow you to verify the accuracy of your tests, and enable good fertility and lime recommendations.
I’d suggest you use a laboratory that participates in the North American Proficiency Testing Program. These laboratories are checked for accuracy regularly and scored accordingly. If the lab you use doesn’t participate in the independent rating program, ask them why not.
Huge dollars in the farmer’s budget are riding on sound fertility recommendations every year. Those recommendations are also the key to environmental stewardship.