I’ve seen too many floods and and too much erosion during my 35 years of farming. Watching a flood always doubles my desire to take action and we have made progress in reducing erosion by 1) installing many miles of terraces and dams and 2) steadily reducing tillage so more crop residue is left on the soil surface. The next logical step would be to eliminate tillage altogether and go to chemical fallow and direct seeding.
I hope to make this switch in the future. However, switching to chemical fallow and direct seeding now will reduce average wheat yields and cause big yield reductions when dry falls cause emergence of direct seeded wheat to be delayed into late October or November. I examine these difficulties and attempt to estimate the size of the yield reduction in
Why I haven’t switched to direct seeding
Without significant rainfall in August and/or September, chemical fallow does not have enough moisture in the top 6” to germinate fall seeded wheat. The summer sun bakes much of the moisture out of the top foot of untilled ground. Hence, wheat seeded on chemical fallow must wait for fall rains. If rains are delayed until October or November, the yield of fall seeded wheat is reduced.
Tilled summerfallow has one big advantage. Properly done, tillage establishes a “moisture line” about 4-5 inches below the surface. Wheat seeded into this moisture will usually germinate, even after long dry periods.
I examined the most recent 30 years of rainfall records from the Experiment Station at Moro, Oregon. I estimated that in 15 of the 30 years enough rainfall occurred in August and September to germinate wheat seeded on chemical fallow in late September/early October (the optimum seeding date). Hence, in about half of the years, wheat seeded on tilled summerfallow and wheat seeded on chemical fallow would emerge at the same time. In 10 of the years of the 30 years, significant rains did not occur until October and the germination of direct seeded wheat would be delayed a month. In five of the 30 years, significant rains did not start until November and the emergence of direct seeded wheat would be delayed by two months.
Several studies have examined the effect of delayed emergence on wheat yields. I discuss three studies in the above article. The studies indicate a yield reduction of about 18% if emergence is delayed a month and a reduction of 40% if emergence is delayed by two months. Hence, I estimate the average yield reduction from switching to direct seeding to be
( ( 0% x 15) + (18% x 10) + (40% x 5))/30 = 13%
Assuming a 50 bushel yield and a $5 per bushel price, a 13% average yield reduction would reduce gross revenue by $32 per acre.
The fall of 2008 was very dry and significant rains didn’t arrive until November. The 2009 wheat yields should highlight the yield reductions caused by late emergence. I took the picture below on April 2, 2009. The wheat on the left was seeded in mid September 2008 on tilled fallow and the wheat on the right is direct seeded on chemical fallow.

To make direct seeding more profitable in the dry areas of the PNW, new wheat varieties must be developed that produce good yields when the crop emerges late in the fall. Finding these new varieties should be a research priority.
In the article, I don’t discuss the differences in production costs between tilled and chemical summerfallow. Tillage can be reduced to one primary tillage plus a rodweeding or two by using a reduced tillage system such as the “undercutter.” Hence, I don’t expect much cost savings when chemical fallow is compared with reduced tillage systems. However, I would be interested in your comments.