We have all tried to estimate soybean yields; and in doing so, engaged in a hair pulling exercise. (I’ve just tried a little too often.) But as USDA’s crop enumerators are currently measuring soybean yields in your neighborhood, their process is based on pod weight and pod count in a set area. That’s different than either the pods and nodes formula or the wild guess formula that may be popular in your county. But with soybean yields so hard to estimate on August 1, USDA last week revealed some important statistics.

The Soybean Objective Yield Survey for 1992-2006 was released by the National Agricultural Statistics Service to explain its process, ahead of this year’s crop estimates season. Currently, 11 states are in the survey, although Arkansas has been in and out, and recent additions are Kansas, and the Dakotas. Seven core states (IA, IL, IN, MO, MN, NE, and OH) that have been surveyed for the past 15 years accounted for 71% of the 2006 crop.

USDA says its enumerators follow these procedures:
1. Enumerators will randomly select two independently located plots, each containing two parallel 3.5-foot sections of row located within a soybean field. (18 square feet.)
2. Counts, measurements, and observations of plant characteristics are made within these plots during the monthly survey periods.
3. Just before harvest, both units are hand harvested by an enumerator and weighed and a sample of pods is sent to a NASS laboratory where moisture content and pod weight are measured.
4. A final gross yield is computed from the number of pods, average weight per pod, and row space width, and that yield is measured as bushels of soybeans per acre at 12.5% moisture.
5. Harvest loss is measured in separate units located near the monthly yield plots.
6. State statistics are produced from the objective data, and regional estimates are made from the state’s harvested acreage.

USDA says in the past 15 years has resulted in changes in pod numbers and row widths. In the 7 major states, production was 46% higher in 2006 than in 1992, because of a 16% higher yield and more acres. For that period the average annual increase was .36 bushels per acre. Harvested acres went up 26%.

The pod count has increased over the 15 years similar to yields. For the 7 major states the pod count has climbed 21% to 1,791 per 18 square feet. The highest was in 2005 with 1,851 pods. Different states have different pod counts. IL leads all with an average of 1,780 over the 15 year period. MO is next with 1,704, and MN has the least with 1,483 per 18 square feet.

The pod weight has declined over the past 15 years, and the average of 0.296 grams per pod is 4% less than what it was in 1992. The low point of the period was in 2003 when drought hampered seed development. If that year is taken out of the average, there has been a slight increase over the 15 year period, but it is less than 1%.

Row widths, as expected have decreased over the past 15 years, and stood at 18.5 inches in 2006. That is 19% under the 1992 average. During the same time, pod count increased in the 18 square feet, since there were more soybean plants. In 2006, 44% of the samples had a row width between 10 and 18.5 inches, compared to only 10% of the samples in 1992.

Summary:
The USDA process of estimating soybean yields is more scientific than most farmer efforts, primarily because of the need for accurate scales and the ability to reconcile for moisture. Enumerators have found in the past 15 years, in the major producing states, that row widths have fallen, pod counts have increased, but pod weights have changed little.