With soil management zones ranging from 1 to 15 acres in size, pulling 15-20 cores per sample was costly but worth it to Kyle Mehmen, MBS Family Farms, Plainfield, Iowa. Although he also retails fertilizer and chemicals, he didn't offer a soil sampling service. He had complete confidence in C8MP and Dale Thorson, the technician who has long sampled his fields.
"My customers didn't have the same confidence in how their samples were being taken, and we were concerned that something might happen to Dale," says Mehmen.
After looking at options, the solution to him and his customers concerns was to invest in AutoProbe, automated core gathering and soil sample collection technology. The ability to gather as many as 40 cores in as many seconds in each grid while traveling at 5 mph was key.
"We were able to double the number of cores we pulled per sample, map out more of a serpentine pattern within the zone and get as much representation of the zone as we could," he says. "We also have customers on grid where we set up a straight A-B line. Either way, we can repeat it on the same pattern. AutoProbe makes it easier and faster. We had one day where we did 800 acres."
It is the number of cores that can be collected for each sample with the AutoProbe, or its similarly fast competitor Falcon Soil Technology, that makes the machines valuable to Antonio Mallarino, professor, Nutrient Management Research and Extension, Iowa State University. While he is quick to say there's no magic number when collecting cores for a soil sample, he is even quicker to say you can't take too many...because you probably won't.
"There should be as many cores as possible, which is a tough thing to define," admits Mallarino.
Part of the problem is that different nutrients require a different number of cores to achieve a certain degree of precision. Phosphorus and potassium require the most; organic matter the fewest. The field and its inherent variability also play a role. Variability, banding history and no-till can all raise the target number of cores needed.
"With these machines, the number of cores is no longer an issue, whether 20, 30 or 40," says Mallarino.
If doing classic soil sampling by hand, Mallarino recommends collecting at least 12 to 15 cores per sample, mixing the samples in a bucket if needed and pulling a subsample for testing. He strongly advises sampling by grid or a zoning method that includes several information layers (soil survey maps, yield maps, aerial images, EC maps, etc.), not only according to soil survey maps.
"With long histories of fertilizer or manure applications, the phosphorus and potassium variation within map units can be as large as the variation across a field," he warns.
He also advises increasing the frequency of sampling to every two years, questioning the point of having a very dense and precise sampling to assess spatial variation if samples are not taken for four or six years. The greater the variability in the field, the greater the payback from more frequent testing.
"There has to be a balance between dense and frequent sampling so it can be done in a cost-effective way," says Mallarino. "While I'm not saying 36 or 40 cores always are needed, machines that automate core collection make it easier to improve soil sampling and testing."