Emerson D. Nafziger Eric Adee, Robert E. Dunker, and Lyle E. Paul 1
Recent high “contest” yields have usually been from fields where corn is grown continuously. This has resulted in support for the idea that such yields may be high because corn follows corn. There is no known research data supporting this idea; our recent work in Illinois shows that in direct comparisons, corn following corn produces 7 to 10 percent less yield than corn following soybean. There is some indication that this “penalty” might be less where soil are least limiting to yield. But inputs needed to produce such yields must be investigated and rationalized.
While we have accumulated a considerable amount of data on the nitrogen response of corn following corn compared to that of corn following soybean, we do not know the effects of some of the “high yield practices” used by some producers, including those producers who are attempting to produce yields above 300 bushels per acre. Such practices typically include deep and thorough fall tillage, high N rates, and high plant populations. Most fields also have high to very high P and K levels, usually as a result of high inputs from fertilizer and/or manure over a period of years.
While this set of practices clearly results in high yields, it is possible, or even likely, that some of these practices may contribute little or nothing to yield. This study is designed as a way to isolate the effects of deep tillage, additional plant nutrient supply, and higher plant population, and interactions among these factors, on corn yield at a number of productive sites in Illinois.
Trials were established in 2003 at two locations, and were expanded in 2004 and 2005 to four locations, all University of Illinois Research and Education Centers operated by the Department of Crop Sciences: 1) DeKalb (North Central Illinois), with predominantly Drummer-Flanagan silt loam-silty clay loam soils; 2) Monmouth (West-Northwestern Illinois), with predominantly Tama-Muscatine silt loam soils; 3) Urbana (East Central Illinois), with predominantly Drummer silty clay loam soil; and 4) Orr Center, near Perry in West-Southwestern Illinois, with Clinton-Keomah-Rushville silt loam soils. In 2005, we established two locations at Urbana, in preparation for moving the trial from Site A to Site B.
Treatments were arranged as a 2 x 2 x 2, split-split-plot factorial, in a randomized complete-block design with 4 replications. The previous crop was corn, and in second and subsequent years at each location, treatments were kept in the same plots as in the previous year. Main plots consisted of 1) fall chisel plow following corn, and 2) deep tillage using a modified mini-moldboard or another tillage tool capable of soil disturbance to a depth of about 15 inches.
Subplots consisted of two levels of fertilizer: 1) normal amounts of P and K according to soil test values and 220 lb of N in the spring, and 2) an additional increment of 80 lb P2O5 and 150 lb K2O per acre annually, and an additional 100 lb N in the spring, for a total of 320 lb of N. Sub-subplots consist of two final plant populations – 32,000 and 40,000 per acre, established after emergence following planting of about 45,000 seeds per acre. Hybrids used at the different sites included Pioneer 33P67, Pioneer 34h21, Pioneer 34B24, Pioneer 33N11, and Pioneer 32B83. Sub-subplots were eight (30-inch) rows wide by 60 ft long. Yields within each sub-subplot were taken by machine harvest of the center 4 rows. Tillage could not be varied at Urbana in 2007, so data from that trial were not included in the analysis. However, foliar fungicide application at tasseling was substituted for tillage in the main plots at Urbana, and was applied to splits within population at Perry.
Data were averaged over years using PROC MIXED, taking years, reps, and interactions of years and treatments as random. This is conservative, in that inconsistency among years tends to mean few significant effects. But it is appropriate if we are using the data to predict future effects.
We have now completed five years of this study at Monmouth and Urbana, four years at Perry, and three years at DeKalb. Yields have varied with weather effects at all locations, but on average have been good. Average yields over years and treatments ranged were 180 bu per acre at Monmouth (5 years), 182 at Perry (4 years), 205 at DeKalb (3 years), and 219 at Urbana (4 years). Monmouth and Perry each had dry years with lower yields that brought down the averages.
Using a probability (alpha) level of 0.1, there was no effect of any treatment or interaction on corn yield averaged over three years at DeKalb (Table 1). At Monmouth, deep tillage increased yield significantly, raising the average yield from 173 to 187 bu per acre. A “modified (cut-down) mini-moldboard” plow was used at Monmouth, which might be more effective in encouraging root growth. A similar plow was used at DeKalb, but did not increase yield there.
At Monmouth, using additional fertilizer increased yield significantly, from 175 to 184 bu per acre, but at a cost of more than $35.00 per acre for the additional N alone, the return from the added N was barely more than breakeven, not counting the cost of the additional P and K. Raising the plant population from 32,000 to 40,000 plants per acre resulted in a significant yield decrease at Monmouth, from 183 to 176 bu per acre. The only treatment that significantly affected yield averaged over four years at Urbana was additional fertilizer, which raised the yield from 213 to 225 bu per acre (Table 1). As at Monmouth, this yield increase would barely cover the additional cost of the N.
Raising the plant population from 32 to 40 thousand per acre decreased yield in seven of 17 environments and increased yield in two environments, both at Perry. There were in some environments an interaction between population and tillage or fertilizer, but these effects were not consistent. More fertilizer increased yields in nine of 17 environments, but the overall effect was well below that needed to pay for this input. Additional or deeper tillage increased yield over the years at Monmouth, perhaps due to the use of the modified moldboard there. This same tillage tool did not increase yield over that from the chisel plow at DeKalb, however. The reason for such a favorable response to tillage at one location is not clear, but roots appear to be challenged when corn follows corn in this environment, and it is possible that this tillage tool reduces root restrictions.
This trial has been useful and interesting, in that some of the ways promoted as assuring high yields in continuous corn are not proving to be helpful. In fact, the reduction in yield from increasing plant population in many of the sites here suggests that having populations higher than those generally considered optimum might increase the stress that continuous corn may be more prone to than corn following soybean. Additional fertilizer is occasionally increasing yield, but overall has not been economically beneficial, due primarily to the fact that the “normal” rate of fertilizer, especially N (220 lb N per acre), is already quite high. Additional tillage is showing some benefits, but not in most environments. We are continuing this study for two or three more years, and will switch to comparing strip-till to conventional (fall) tillage in main plots. We will also use hybrids with the rootworm Bt trait in order to see if that reduces root-related problems and if it might increase the response to these inputs.