Illinois
Fertilizer Conference Proceedings
January 26-28, 2006
| Home | 2006 Index | Search |
| Home | 2006 Index | Search |
E.D. Nafziger, Eric Adee, R.E. Dunker, and L.E. Paul
1
Recent high yields recorded by Francis Childs in Iowa and by others 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. On the most productive fields, continuous corn can certainly produce high yields, 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 2004 at at four University of Illinois Research and Education Centers operated by the Department of Crop Sciences:
Treatments were a 2 × 2 × 2, split-split-plot factorial design, arranged 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 N. Sub–subplots consist of two final plant populations – 32,000 and 40,000 per acre, established after emergence following planting about 45,000 seeds per acre. Sub-subplots were at least 8 (30–inch) rows wide by 60 ft long. Yields within each sub–subplot will be taken by machine harvest of the center 4 rows.
The trial at Urbana in 2004 did not include the plant population variable, and the DeKalb location was lost to water damage in 2004. The 2003 cropping season was average at Monmouth, where yields averaged 179, and outstanding at Urbana, where the average trial yield was 241 bu/acre (Table 1). The 2004 season was very favorable at most locations, with high yields in most areas, and relatively high yields for corn following corn compared to corn following soybean seen in other trials. Yields averaged 200, 238, and 224 bu/acre at Monmouth, Urbana, and Perry, respectively. In 2005, dry weather affected yields at most locations, and yields ranged from only 101 bu/acre at Monmouth to 209 bu/acre at the DeKalb.
Deep tillage increased yield significantly in both 2004 and 2005 at Monmouth, where a "modified (cut-down) mini-moldboard" plow was used (Table 1). A similar plow was used at DeKalb in 2005, but did not increase yield there. There was no significant tillage effect at any of the other sites. Across the ten site-years, deep tillage increased yield by an average of 5 bu/acre. Extra fertilizer increased yield significantly at five of the ten locations, and by an average of 10 bu/acre over all site-years (Table 1). The interaction between tillage and fertilizer was significant at only one location - Perry in 2005, where the yield increase from extra fertilizer was large when tillage was by chisel plow, but not when deeper tillage was used.
Raising the plant population from 32 to 40 thousand per acre decreased yield at five of nine sites, and increased yield at only one site (Table 1). Response to raising the population ranged from a decrease of 16 bu to an increase of 9 bu/acre, with an average response over nine locations being a loss of 6 bu/acre. There was an interaction between tillage and population at four of the locations; in 2003, there was less loss of yield from high populations with deep tillage than with chisel, while at two locations in 2005 (Urbana B and DeKalb), there was more loss of yield from high populations under deep tillage than with normal tillage. There was an interaction between fertilizer rate and population at three sites, and all showed less loss in yield from high population under the higher fertilizer rate.
Interactions among the management variables averaged across ten site-years are shown in (Fig. 1). The largest loss in yield from high plant population came under normal tillage and fertilizer rates, while raising the fertilizer rate under normal (chisel) tillage "protected" yields from loss due to high plant population. The negative effect of higher population was more consistent, but smaller, with deep tillage. The inconsistency among site-years to date means that we cannot predict that these effects and interactions will occur consistently in the future; that is, considering site-years as random effects makes most main effects and interactions non-significant. It is clear from these results that simply raising inputs of tillage, fertilizer rates, and plant populations is not a "formula" for high yields in continuous corn, even though some of these inputs might return additional yield in some environments. This is true even if such inputs are not required to pay for themselves, as is usually the case for "contest" fields. In commercial production fields where such inputs have to pay for themselves, the inconsistency we have found is an even larger barrier to increased acceptance. This research is ongoing.
1 E.D. Nafziger is Professor, E. Adee is Principal Research Specialist, and Robert Dunker and Lyle Paul are Agronomists, Dep. of Crop Sciences, Univ. of Illinois, Urbana, IL.
