A soybean farmer in northeast Louisiana who lets 4 bushels per acre fall behind the combine, with a written limit of 2.2, gives up $28,350 on 1,500 acres (example price: $10.50 a bushel). Write the loss you accept before harvest starts. Then count the beans on the ground behind the combine, field by field, and mark each field that came in over.
Harvest loss here means grain that ends up on the ground instead of in the bin: beans the header never gathered and beans thrown out the back of the machine. Without a written limit, one person calls the loss normal, another calls it too much, and nobody settles it.
How much grain is too much to leave behind a combine?
A University of Georgia Extension bulletin says harvest losses of 10 percent or more are not unusual, when they should be in the 2 to 4 percent range. On a field expected to yield 55 bushels per acre, 2 percent is 1.1 bushels per acre and 4 percent is 2.2. The example farm writes 2.2 as its limit for the year.
Other published tolerances sit lower. Embrapa, Brazil’s public farm-research agency, sets the tolerable soybean loss at 60 kilograms per hectare, about 0.9 bushel per acre. A talk to Aapresid, the Argentine no-till growers association, gives 4 percent of yield, and 105 kilograms per hectare for yields above 2,500, about 1.6 bushels per acre on the example field.
Each figure comes from a different country and a different set of machines, so none of them is yours. In the first year, write one as your limit and date it. Later, revise it with your own counts, always before harvest.
How do you count the loss in a field?
Lay out a frame that encloses 10 square feet across the swath, the width the combine cuts, and count the beans inside it. A 20-foot swath needs a frame 6 inches long: 10 divided by 20 is 0.5 foot. The Georgia bulletin lists the length for each common swath.
Stop the combine at least 100 feet from the row end without clearing it, drop the frame behind the machine, and count. For soybeans, about 40 beans in 10 square feet equal 1 bushel per acre, a figure that varies slightly among varieties.
Count at five spots at least. Embrapa asks for five points or more in each field because one count does not describe it. A Brazilian study of one soybean harvest found total loss reached 387.3 kilograms per hectare at 9 p.m., about 5.8 bushels per acre, and rose 121 percent between the first and last day. Write the date and the hour on each count.
Did the header or the machine lose the grain?
A field total does not say which part of the combine lost the beans. The header cuts and gathers the plants and drops beans on the ground ahead of the machine. The parts inside, which thresh, separate and clean, lose them out the back.
The Georgia bulletin builds the split into the count. After the count behind the machine, back up no more than 20 feet and count a second frame just ahead of the header. That ground was cut but the rear of the combine never crossed it, so the count holds the beans that fell before the combine arrived plus the header’s loss. The difference between the two counts is what the inside of the machine lost.
On the example field, 100 beans ahead and 160 behind leave 60 beans, or 1.5 bushels per acre, lost inside the machine. The other 2.5 bushels per acre, 62 percent of the 4.0, came at the header or before it.
In a talk in Argentina, the header accounted for 70 to 80 percent of the soybean loss in a normal year. A survey in Maripá, Paraná, cites Embrapa work at 80 to 85 percent. The example field sits below both, which sends the operator and the agronomist to look inside the machine.
Is a loss over the limit the exception or the rule?
It depends on the region and the machine, and only a count tells you. In Maripá, Paraná, a survey of 39 combines on 25 farms found that 79 percent lost more than 60 kilograms per hectare, with an average of 98.5 kilograms, about 1.5 bushels per acre.
In the Mid-North of Mato Grosso, a survey of 16 combines found 69 percent at or under that level. The authors tie the machines above it to poor maintenance.
Each survey covers one region and a few seasons. They are a reason to count your own fields, not numbers to copy.
What goes on the harvest sheet before the first field?
One sheet with a row per field. The limit and its date go in before the combine starts; the counts go in at each field.
| Column | What it holds | What goes wrong without it |
|---|---|---|
| Field | The name or number of one field | Two fields blur, and a bad one hides in the average |
| Limit and date | The loss accepted, in bushels per acre, with the date and the name of who set it | A limit written after the result excuses any number |
| Counted loss | The average of at least five counts, in bushels per acre, with date and hour | “It looked fine” stands in for a number |
| Where it came from | At or before the header, or inside the machine | A field over the limit with no clue where to look |
| Over or under | A mark on every field above the limit | Every field reads the same |
A field over the limit gives you a place to look: that field, that pass, that part of the machine. Next season you examine the one or two marked fields instead of arguing about the whole crop.
Setting the combine and cutting the loss stay with the farmer and the agronomist; a separate article covers what the agronomist decides. If you already record each field’s yield, year by year, the counted loss belongs beside it, because it explains part of a thin year.
Measuring against a limit written beforehand is part of farm management. The production section of the library holds the other field records this one sits beside.