Southeast Kansas cropland soil test is complex when statistics are involved

Humorist Evan Esar once wrote that “Statistics is the only science that enables different experts using the same figures to draw different conclusions.” What is the average soil test in Southeast Kansas? This simple question, like many simple questions, is in fact quite complex when statistics are involved.

Over the past year, I have had the opportunity to make fertility recommendations on just over 220 soil tests across the four counties (Crawford, Labette, Montgomery, and Wilson) the Wildcat District covers. The vast majority (nearly 200) came from Crawford and Montgomery so this data is definitely skewed toward them.

For the purposes of this data set, I have only included crop ground soil tests. If hay and pasture soil tests were included, I am quite confident the data would be considerably different.

The common soil test includes three major tests, Soil pH, Phosphate parts per million, and Potash parts per million. The K-State Soil Testing Lab uses the Mehlich-3 extraction method for extracting and measuring phosphate. Different labs may or may not use this method so comparing between labs may not be recommended.

The average phosphate level across the 220 samples was 29.4 parts per million. I must admit that I would have assumed a lower level than this without running the data. While the average is the most common measure of central tendency, the median (half of the measurements below and half above) can minimize the outliers in data. For phosphate, the median is 16.6 parts per million, which would be much closer to my assumptions.

The highest sample I reviewed this year had 214 parts per million phosphate while the lowest had only one parts per million. If you break down the data into percentiles, the upper 10 percent of the samples were at 69 parts per million or above while the lowest 10 percent were at 5.9 and below. For the major crops in our area, 20 parts per million is the level in which no phosphorus fertilizer would be recommended on a sufficiency basis. There were 118 samples below 20 parts per million which would have had some phosphorus fertilizer recommended.

For potassium, my assumptions were actually much closer to the results. The average across all samples was 146 parts per million potash. Once again, the median was lower at 123 parts per million and the range was quite broad with a maximum of 571 parts per million and a minimum of 43 parts per million.

For potash, the highest 10 percent of the samples were at 255 parts per million or greater while the lowest 10 percent were 79 parts per million or lower. The critical level for potassium fertilizer to be recommended on a sufficiency basis in our major crops is 130 parts per million and 123 samples fell below that crucial amount.

Finally, pH was also quite interesting and my prior assumptions would be nearly accurate. The average and the median were quite close in for pH at 6.53 and 6.6 respectively. In addition, the range was wide with a maximum of 7.9, which is extremely high for our area, and a minimum of 5.1.

The highest 10 percent of pH readings were at 7.3 and above while the lowest 10 percent fell below 5.69. In pH, the critical value is below 6.5 which would induce a lime recommendation. There were 91 samples which met this criteria.

To try to quantify the “average” soil test in Southeast Kansas, a number of statistical measures were applied. Feel free to draw your own conclusions.

For more information, contact Josh Coltrain, Crop Production Agent, [email protected] or 620-724-8233.