Issue No. 2, Article 7/April 4, 2008
Thinking About Corn Planting Date and Population
Most of the state has wet soils at present, with no immediate indication that soils will start to dry out soon. March was generally cooler than average, and precipitation has been average to much above average over much of the state. Cool weather in March is never conducive to early drying, and cool temperatures coupled with normal rainfall mean that we may get well into April before much can happen in the fields. There may be scattered areas dry enough to allow fertilizer application. But this is a good time to remember that soil at field capacity is at its most "compactable," and that even flotation tires can do some damage when they carry heavy equipment across soils that are wet.
Corn prices have been quite volatile, and they are likely to remain that way as we move into April with little or no planting done in Illinois. This is a good time to review planting date response to see when yield losses start to accelerate as planting is delayed. We have been conducting research for the past three years at six Illinois locations looking at both planting date and plant population responses. The data will be used to update the responses that we currently use, which are from data generated some 15 years ago at fewer locations.
In these studies, we make the first planting as early as practical without mudding the crop in. The first date has ranged from late March to mid-April. We then plant at intervals of about 3 weeks, for a total of four dates, with the last date in late May. Within each planting date we thin to stands of 20,000 plants to 40,000 or 45,000 plants per acre.
As we expected from previous experience, the effect of planting date varied considerably, with the highest yields sometimes coming from the earliest planting date and in a few cases from the third (mid-May) date. Much of this is due to the rainfall pattern over the entire season, with good rainfall throughout the summer usually meaning highest yields from the earliest plantings. Dry weather in June and early July followed by good rainfall might give an advantage to the third planting. The late May planting has never produced the best yields, because plants that start that late cannot utilize the whole growing season effectively.
We combine data from sites over years to predict effects of planting date on yield. These estimates are the best we have, but they are fuzzy, with the yield from one planting date at a particular location often not very close to the response line based on all the data. For example, dry weather reduced yields considerably at Perry (Pike County) in two of the three years. We combined the data from Perry and Urbana to make the central Illinois response curve, but the curve doesn't fit the data very well in a particular sense, even though it is the average and so the best prediction we've got. In other words, we expect the next trials at these locations to differ from the "average."
At both Monmouth and Urbana, the earliest planting in 2005 was destroyed by the frost the first week of April. The crop was emerged at DeKalb that year as well but was smaller, and it grew back normally. At Monmouth, the stand was completely destroyed, while at Urbana at least half of the plants were killed. Frost that late is unusual, but we also observed that having the growing point underground does not guarantee survival of V2 to V3 corn plants when temperatures dip into the 20s. Most fields that were planted in early April in the northern part of western Illinois that year had to be replanted. The rarity of this type of event means that it probably should not have a lot of influence in altering planting date. We did delete the data from that planting date for these two locations, since including low or zero yields is not realistic when such fields would clearly need to be replanted.
Figure 1 gives the planting date response curves for northern Illinois (Monmouth and DeKalb), central Illinois (Urbana and Perry), and southern Illinois (Brownstown and Dixon Springs) over the three years. We used the same hybrid for the northern and central locations and a later-maturing one in the southern locations. There was no problem with the crop from any of the latest planting dates maturing before frost, though some of the latest plantings might have had more insect or disease pressure than earlier ones.
Figure 1. Planting date responses averaged over two site and three years for each of three regions in Illinois.
The responses shown in Figure 1 are converted to numbers in Table 1, with average yields and average losses in yield per day of delay during each 10-day period from early April to early June. The data show that yield losses do not approach 1 bushel per day until at least 20 days after the optimum planting date, and that planting during the first week of May is expected to yield only 6 to 10 bushels less than planting on the optimum date in central and northern Illinois. Delays in planting tend to be more costly in yield in southern Illinois, where planting the first week of May is a month past the optimum date and brings a yield penalty of about 17 bushels per acre. Still, good yields are certainly possible when planting is delayed, and planting two weeks after the optimum date might well produce higher yields than planting into cool, wet, compacted soils closer to the optimum date.
Negative yield "losses" for the periods in early April mean that yield actually increases as planting is delayed during that period, because the "best" planting date is later than this. Such yield penalties for planting very early are typically not large, but they are common--of 16 site-years (not counting Urbana and Monmouth in 2005 when the first planting failed), the earliest planting yielded more than the second one eight times, the second planting yielded more than the first seven times, and at one site there was no difference. When the weather is consistently warm after the first planting emerges, early planting tends to do best, but if cool or cold temperatures occur after corn has passed the 2- or 3-leaf stage (this would most commonly be the largest, earliest-planted corn), we think there can be temperature-related damage that affects the yield potential. In 2006 at Urbana, where the first planting yielded almost 40 bushels per acre less than the second planting, there was a period in mid-May during which high temperatures averaged less than 50 degrees. We do not know how or where such damage occurs in corn plants.
Plant population. As I have described, this series of trials has included a range of plant populations established by thinning to stand within each planting date. While we found interactions between planting date and plant population at a number of site-years, they were not consistent enough to tell us that plant population ought to be changed depending on when we plant. It might be good to raise seed drop rates slightly when planting early to compensate for lower expected emergence, but if soil conditions are good, this may not be necessary. It is normal to drop 5% to 10% more seed than the desired final stand, with the lower end of this range appropriate if seed quality and soil conditions are good.
Figure 2 shows yield response to plant population averaged over nine northern Illinois locations (DeKalb, Monmouth, and Urbana) and averaged over three southern Illinois locations (Perry, Brownstown, Dixon Springs). Data were averaged over the first three planting dates in each trial. Perry is in west-southwestern Illinois, of course, but it has had yields lowered by dry weather in recent years and has soils similar to other areas in southwestern Illinois, so yields and responses were similar to those in the southern locations. The data from northern locations show increasing yield up to a population of about 37,000 plants per acre, while in southern Illinois yields increased only up to 27,000 plants per acre. There was little or no yield increase at populations above that at the highest yield level. This reinforces an important feature of modern hybrids: yields tend to level off, not drop off, at populations above the point of maximum yield. From a risk management standpoint, this means that it tends to be prudent to plant higher populations to take advantage of good conditions, since the only loss under average or poor conditions is the additional seed cost.
Figure 2. Yield responses to plant population over 9 northern and 9 southern Illinois site-years. Data are averaged over the first three of four planting dates at each site-year
The large increases in both corn price and seed costs in the past year raise questions about using economics to set appropriate plant populations. Table 2 shows the effect of changing corn and seed prices on optimum harvest population based on the northern Illinois response data shown in Figure 2. When corn is bringing $5 a bushel, the economically optimum population drops by about 1,200 plants per acre when the seed cost doubles from $120 to $240 per unit. In practice, such changes should be moderated somewhat by the point made above, that the economic risk of having populations too low for good conditions tend to be greater than the risk of having them too high, even when seed is expensive.
For those who remain conservative on plant populations, the data show that yield losses when populations are 3,000 to 5,000 less than optimum tend to be relatively small. But with less need to make sure populations are not too high, the potential benefit from maintaining lower populations, other than direct savings in seed costs, is also small.--Emerson Nafziger