Weed Management in Roundup Ready Cotton

Alan York, Extension Weed Scientist
Stanley Culpepper, Extension Associate
and Keith Edmisten, Extension Cotton Specialist
(CCN-98-2B - February 1998)

This Carolina Cotton Note is also available in a PDF format suitable for printing.
File In PDF format
You Will Need Adobe Acrobat Reader to View File

Approximately 70,000 acres of Roundup Ready cotton were grown in North Carolina last year. Although 1997 was not one of our better cotton years, comments from growers concerning Roundup Ready cotton were positive. Seed orders for 1998 were being placed in September and October, indicating grower satisfaction with this new technology was phenomenally high. Demand for Roundup Ready seed will exceed the supply in 1998.

We have recently learned that seed supply in 1998 will be tighter than originally expected. In mid-February, Paymaster recalled all of certain Roundup Ready and Roundup Ready/BollGard lines. Paymaster 1220 RR, Paymaster 1220 RR/BG, Paymaster 1244 RR, DP 5415RR, DP 5690RR, and DP 90RR will be available.

We can't recall another situation where growers have been so willing to make a wholesale switch in the varieties they grow while having so little research data or personal experience with the new varieties. While we expect the currently available Roundup Ready cotton varieties to perform similarly to their non-transformed recurrent parents, one should not assume they are exactly the same.

Fortunately, results to date indicate Roundup Ready varieties are performing well. Five Roundup Ready varieties were entered into North Carolinas Official Variety Tests in 1997. Averaged over four locations, yields of Roundup Ready varieties ranged from 95 to 101% of the mean of all entrees. In Extension tests, we compared eight Roundup Ready varieties to Stoneville 474. Stoneville 474 was selected for comparison because it has the highest three-year average yield in the states Official Variety Test. Averaged over two locations, yield of the eight Roundup Ready varieties ranged from 92 to 115% of the yield of Stoneville 474. These results can be found on pages 24-42 and 118 in  "1998 Cotton Information", available from Extension offices.

Much effort has been devoted to comparing Roundup systems with standard weed control programs. Our research, summarized in Tables 1-4, has shown that weed control, cotton yield, and net returns with Roundup Ready systems are very similar to those with our best conventional systems consisting of at-planting herbicides plus postemergence herbicides.

Roundup Ready technology will have an impact on how growers manage weeds. One of the biggest decisions growers will have to make is whether or not to use soil-applied herbicides in a Roundup Ready system. Some growers have cautiously decided to continue using at least some of their normal preplant or preemergence herbicides. Others have decided to go with a total postemergence program. There is no "one size fits all" answer here. This has to be an individual decision made after consideration of several factors.

Our research has shown that soil-applied herbicides are not absolutely necessary in most cases. A total postemergence program of Roundup alone or Roundup followed by standard directed herbicides has given weed control, cotton yields, and net returns that are statistically equivalent to that achieved with our best traditional programs (Tables 1-4).

Before abandoning soil-applied herbicides, however, growers need to consider several things. First, there are a few weeds that Roundup will not control. Probably of greatest concern across the Southeast is Florida pusley. Florida pusley is shown in the picture below.

Florida pusley is an annual broadleaf weed more commonly found on drier, sandier soils. We have never been concerned about this weed because our normal herbicide programs have controlled it well. Hence, most growers do not know where this weed might be a problem. We know there is some pusley in North Carolina, especially in the southeastern quarter of the state. However, we do not think it is a wide-spread problem.

For all practical purposes, Roundup will not control Florida pusley. Some researchers have reported moderate success if the weed was treated when very small. It is unlikely, however, that growers will treat it small enough to get adequate control.

To make matters worse, none of our other postemergence cotton herbicides will control pusley. Hence, once emerged, there is not much we can do about pusley beyond cultivation. Heavy infestations can be very damaging.

Fortunately, soil-applied herbicides are very effective on Florida pusley. A grower would be well advised to use a soil-applied herbicide on any field where pusley is suspected.

An incorporated application of Prowl or Treflan is the best option to control Florida pusley. Either herbicide gives excellent control at an economical price. If an incorporated treat does not fit into your system, all of the registered preemergence herbicides will give respectable pusley control.

Another weed of concern in eastern North Carolina is dayflower. This weed is typically found on darker, wetter type soils. Infestations are typically spotty rather than uniformly distributed over large areas. Spreading dayflower is shown in the picture below.

Roundup has very limited activity on dayflower. Fortunately, dayflower can be controlled with other herbicides. Although research on dayflower is very limited, we have generally observed that Cotoran preemergence controls it. This would be an option on soils with less than about 4% organic matter.

It appears dayflower can also be controlled with postemergence herbicides. Two species of dayflower can be found in North Carolina. We have no data on control of Asiatic dayflower. In preliminary research in 1997, 3- to 4-inch spreading dayflower was controlled well by Staple, Bladex, directed rates of MSMA, and combinations containing directed rates of MSMA. At this point, we do not know how effective these herbicides are dayflower larger than 3 to 4 inches.

Timing of Roundup application can be critical for some weeds. This includes weeds such as morningglory, prickly sida, smartweed, and hemp sesbania.

Although soil-applied herbicides may not be absolutely necessary in most situations, they can broaden the window of application for Roundup. Timing of the first application of Roundup becomes less critical. This can be important to a grower with large acreages. Soil-applied herbicides also provide some insurance against weathers delays in Roundup application.

An example of the impact of soil-applied herbicides on timing of Roundup can be seen in Tables 3 and 4. This location has a moderate infestation of several broadleaf species and an extremely heavy infestation of crabgrass. Even though weed control in a Roundup-only program was good, delaying Roundup application until the 3-leaf stage of cotton in the absence of soil-applied herbicides allowed the weeds to compete enough to reduce yield. However, as shown in Table 4, initiation of Roundup application at the first leaf stage of cotton removed the weeds before they competed with the cotton, and yields were similar in systems with and without soil-applied herbicides.

Will weeds become resistant to Roundup? This question has been raised by both opponents and proponents of the technology. If we are in a system of continuous Roundup Ready crops and we rely almost exclusively on Roundup, resistance could develop. Resistance has been reported in Australia after long-term use of Roundup.

Although we cannot totally dismiss the possibility of resistance to Roundup, the odds of weeds becoming resistant to Roundup are, in our opinion, considerably less than the chances for resistance to a number of other commonly used herbicides.

The greater concern with Roundup will be weed shifts. A weed shift is a situation where species most susceptible to the herbicide program decline over time while less susceptible species build up. Weed shifts can generally be predicted. If the same or similar herbicides are used repeatedly, we can anticipate an increase in those species less susceptible to the herbicide program.

In the case of Roundup, one might expect a shift toward morningglory, dayflower, and Florida pusley. Morningglory can be controlled by Roundup systems, but this weed is less susceptible to Roundup than most of our other common species. Research at N. C. State by Harold Coble has shown that annual grass and sicklepod populations declined and morningglory populations increased after three years of a continuous Roundup-only program.

The potential for weed shifts or resistance is certainly not unique to Roundup systems. There were numerous examples of weed shifts and resistance long before Roundup Ready technology entered the scene. We can greatly reduce or eliminate the potential for both weed shifts and resistance through good management practices.

One of the key components in a strategy to avoid either weed shifts or weed resistance is rotation of herbicide chemistry among crops in the rotation and/or use of different types of chemistry within a given crop.

If we are rotating cotton and another crop and will be relying heavily on Roundup in the cotton, it might be advisable to use a non-Roundup program on the rotational crop. If we are growing continuous cotton, it would be advisable to integrate some other herbicides into the program. For example, we might want to use Roundup early and then follow with a traditional directed herbicide program. In our research, this type of program has been a little more effective on morningglory than a Roundup-only program.

A second advantage of a program consisting of Roundup early followed by a traditional directed herbicide such as Bladex or Caparol/Cotton-Pro would be some residual activity from the traditional herbicide. This could be important, especially where no residual herbicides are applied at planting.

Timing of over-the-top applications can be critical. The label states that Roundup should be applied over-the-top prior to the five-leaf stage. Two over-the-top applications can be made prior to the five-leaf stage but those applications should be separated by 10 days and two nodes of growth. Growers will have to start early at the cotyledon or one leaf stage if they want to make two over-the-top applications of Roundup. Figure 1 shows a cotton plant in the late stages of the one-leaf stage.

Figure 1. A cotton plant in the late stages of the one-leaf stage.

The plant shown in figure 1 might appear to be four leaf cotton to someone with little knowledge of cotton growth and development. The leaves with the letters A and B are cotyledon leaves and are not counted as true leaves. The leaf with the letter C is the first true leaf. The leaf with the letter D is the second true leaf. When this leaf (D) totally unfurls and is the size of a quarter the cotton will be in the two-leaf stage.

Figure 2 shows a cotton plant in the four-leaf stage. The cotyledonary leaves are below the leaf labeled A and are not shown in the picture. Cotyledon leaves are easy to recognize as the two leaves come of the mainstem directly across from each other. The first true leaf is labeled as A, the second true leaf is labeled as B, the third true leaf is labeled as C and the fourth true leaf is labeled as D. The fifth true leaf is labeled as E. Because the fifth true leaf is not as large as a quarter the cotton is still in the four leaf stage. However this plant is close to being five-leaf cotton. Once the fifth true leaf is as large as a quarter, Roundup applications must be directed or fruit loss can occur.

Figure 2. Cotton in the four-leaf stage.

Cotton develops nodes in a predictable manner. It takes about 50 DD60's for each new node of development. Early in the season in North Carolina this is usually every 3 to 5 days. Growers need to keep in mind this fact in order to plan so that over-the-top Roundup applications can be made prior to the five leaf stage and so that if two over-the-top applications are desired the first one can be applied early enough to make the second application based on the label restrictions listed above. Five day weather forecasts can be used to predict how quickly nodes will develop on the cotton. The formula for predicting DD-60's is shown below.

DD-60 calculation for each day:

[(°F Max + °F Min Temp)/2 ]- 60 = DD-60s

For example, if today's high and low temperatures were 80°F and 60°F, respectively, then the formula would arrive at [(80°F + 60°F)/2]-60 = 10 DD-60s. If temperatures remained like this for 5 days it would take about 5 days to develop a new node on the cotton plant.

Table 1. Late-season weed control and lint yield in Roundup Ready Cotton. Experiment 1, 1996-97. Average of three locations.¹ 

1Cotton planted with conventional tillage. All herbicides broadcast. Plots not cultivated. Lint yields and corresponding net returns are based upon actual lint percentage rather than gin turnout. Yields are approximately 25% greater than if based upon a normal gin turnout. Means within a column followed by the same letter are not different at P = 0.05.

2 Treflan applied preplant incorporated; Cotoran applied preemergence.

3EPOST = early postemergence. Treatments applied to 2- to 4-leaf cotton. Staple and Roundup applied over-the-top; Cotoran + MSMA directed.

4LPOST = late postemergence directed.

Table 2. Late-season weed control and lint yield in Roundup Ready Cotton. Experiment 2, 1996-97. Goldsboro location. Averaged over years.¹

1Prowl broadcast; all other herbicides in 50% band. Plots cultivated twice. Lint yields and corresponding net returns are based upon actual lint percentage rather than gin turnout. Yields are approximately 25% greater than if based upon a normal gin turnout.

2EPOST = early postemergence on 3-leaf cotton. Staple and Roundup applied over-the-top; Cotoran + MSMA directed.

3MPOST = mid postemergence directed to 10- to 12-inch cotton.

Table 3. Late-season weed control and lint yield in Roundup Ready Cotton. Experiment 2, 1996-97. Rocky Mount location. Averaged over years.¹

1Prowl broadcast; all other herbicides in 50% band. Plots cultivated twice. Lint yields and corresponding net returns are based upon actual lint percentage rather than gin turnout. Yields are approximately 25% greater than if based upon a normal gin turnout.

2EPOST = early postemergence on 3-leaf cotton. Staple and Roundup applied over-the-top; Cotoran + MSMA directed.

3MPOST = mid postemergence directed to 10- to 12-inch cotton.

Table 4. Cotton yield and net returns with selected treatments in Experiment 2, 1997.¹

1Prowl broadcast; all other herbicides in 50% band. Plots cultivated twice. Lint yields and corresponding net returns are based upon  actual lint percentage rather than gin turnout. Yields are approximately 25% greater than if based upon a normal gin turnout. Means  within a column followed by the same letter are not different at P=0.05.

2EPOST = early postemergence over-the-top at the 1- or 3-leaf stage.

3MPOST = mid postemergence directed to 10- to 12-inch cotton.

4Layby directed to 25-inch cotton