Most corn and soybean growers are familiar with USDA's National Agricultural Statistics Services corn and soybean yield estimates released in mid-August. Likewise, most know there's a lot that can happen from the middle of August until corn reaches black layer, and there will inevitably be some difference between NASS' August estimates and the final tally in February.

IT PAYS TO WAIT: Waiting until September to count soybean pods and seeds means Extension educator Nathan Mueller can pick up on things he might miss in mid-August. "You don't know how many seeds per pod you have until R6, which is typically at the end of August," he says. "Then you can still lose some pods into early September."

However, Nathan Mueller, University of Nebraska Extension cropping systems educator in Dodge and Washington counties, has been experimenting with a new method that he hopes will tighten the gap between those numbers. For the last two years, Mueller has spent several days in September in corn and soybean fields to try out his new method. Last year, his September estimates came within 2 bushels for USDA's final numbers for corn and soybeans in Dodge County.

In the long term, Mueller hopes by calculating yield estimates in September, he can increase awareness of how earlier-season agronomic issues impacted the crop in the county and educate local growers on any late-season weed, disease or insect pressures going on. It could also help growers use yield estimates for grain storage or marketing purposes to have a rough idea of what local basis might be, or ensure they have enough grain to meet hedging contracts.

"From a marketing standpoint and grain storage standpoint, that information could be beneficial," he says. "The reason I went down that road was to help some of those decisions in September."

Rundown of plan
Here's how Mueller’s new method works:

• Mueller randomly selects fields along a set route (more on that later). He walks past all the end rows and then walks at least an additional 60 feet or more into the field. Then he measures out a thousandth of an acre for corn or ten-thousandth of an acre for soybeans based on row width.

• He pulls back the husks on every fifth ear on dryland cornfields and every sixth ear for irrigated fields, and repeats this step for the adjacent row to ensure the confidence of the estimate and account for factors that impact yield in individual rows. After counting the total number of ears within the two row lengths, as well as kernel rows and kernels within a row, he calculates the average number of kernels per ear and the number of kernels in that thousandth of an acre.

• The next question is: How many kernels per bushel? In years past, the industry standard has been 90,000. But for Mueller, it changes based on the year. So far, he's narrowed it down to an average of 75,000 to 85,000 kernels per bushel for dryland and 70,000 to 75,000 for irrigated corn in Dodge County. Again, he adjusts for seed size at each field. This ten-thousandth-of-an-acre method was suggested by Shaun Casteel at Purdue University, but Mueller has since added some additional yield corrections (moisture, late season, preharvest and harvest losses).

• For soybeans, he counts the total number of pods on all plants in that ten-thousandth of an acre, and counts the average number of seeds per pod based on two representative plants in the sample. He then multiplies the total pods per ten-thousandth of an acre by the average number of seeds per pod and divides by 18 (a factor for 3,000 seeds per pound) to get the yield estimate at 13% moisture.

• For corn, he breaks the ears in half to keep track of growth stages. For corn, preharvest, postharvest and moisture loss is already accounted for; you don't count the kernels at the tip or the base.

• For soybeans, he also monitors growth stages at each field. However, Mueller notes you can't always account for things like late-season loss, preharvest loss, harvest loss and moisture loss in soybeans. So, Mueller accounts for this assuming harvest moisture at 9% and deducting late-season, preharvest and harvest losses of 15% from the final yield.

A different approach
Mueller typically visits 17 cornfields and 17 soybean fields in Dodge County. But they're not just randomly selected fields; they're randomly selected within specific routes to account for different soil types and different weather events that affect yields in different parts of the county.

"When you're walking in the middle of a field, you're making the assumption that what you're seeing is representative of that field. That's why I visit multiple different fields," Mueller says. "For example, because the Platte River bottom makes up about 25% of Dodge County, I want to make sure roughly four of the 17 fields are on the Platte River bottom."

In addition, about 39% of the corn raised in Dodge County is irrigated, so Mueller tries to make sure roughly 39% of the cornfields he visits are irrigated. After tallying up the averages for dryland and irrigated, he applies a weighted average to both to get a closer representation of the county average.

The 2016 corn preharvest yield estimate was 184 bushels per acre — a 13-bushel decline from last year. Meanwhile, the 2016 soybean preharvest yield estimate was 62 bushels per acre — a 4-bushel increase.

Of course, this changes year after year. Over time, Mueller's goal is to have enough data to determine how close he is each year and what needs to be changed to get closer to the farmer survey data used by USDA for the county average. "Part of the goal of the crop tour is about understanding the degree of error in my method. I won't know how close I am until after the fact," he says. "So it will be interesting to see how close I am on year two."

You can get the full results from this year’s Dodge County Crop Yield Tour at croptechcafe.org.