Survey of Cotton Weeds and Weed Control Practices in Arizona Upland Cotton Fields
William B. McCloskey, Plant Sciences Department
The distribution of weed species and the herbicides and cultural practices used to control weeds in Arizona cotton fields were surveyed in 1995 and 1996. The most common weeds were purple nutsedge, bermudagrass, annual morningglory, Palmer amarnath, Wright groundcherry, common purselane, yellow nutsedge and Johnsongrass. The average statewide cost for hand weeding in 1995 was reported as $27.87 per acre in addition to other weed control costs. Statewide, most growers used preemergence herbicides before or at planting and used pre- and post-emergence herbicides later in the season. Most of these applications were broadcast applications suggesting that many of the postemergence herbicide applications were layby applications. Preemergence herbicides (usually applied preplant incorporated) such as Treflan, Prowl, and Prometryn were more commonly used than postemergence herbicides. Statewide, few growers banded preemergence herbicides or used electro-hydraulic quick-hitch guidance systems and in-row weeding tools with their cultivators.
Cotton is one of the most important crops grown in Arizona with more than 166,700 hectares (412,000 acres) and 144,000 hectares (355,900 acres) harvested in 1995 and 1996, respectively (Arizona Agricultural Statistics Service, 1995, 1996). The economic value of the 1995 and 1996 Arizona cotton crop was 335 and 296 million dollars, respectively (Arizona Agricultural Statistics Service, 1995, 1996). Herbicides were applied to over 89% of the Arizona cotton acreage in 1995 and over 75% of the cotton acreage in 1996 (National Agricultural Statistics Service, 1995, 1996). Weed competition with cotton reduces seed cotton yields and thus reduces farm income. In addition, cultural practices, including hand weeding, and herbicides used to control weeds cost money and further reduce economic returns to growers.
Weed surveys are useful for determining the occurrence and relative importance of weed species in crop production systems (Frick and Thomas, 1992; McCully etal., 1991; Thomas, 1985). Documenting the weed species present in cotton fields and the herbicides and cultural practices used to control those weeds allows comparisons with past and future surveys. These comparisons can help elucidate the effect of new weed control technologies on farming practices, document weed species shifts in response to new weed control technologies, and document the development of herbicide resistant weeds. Documenting the relative importance of weed species also facilitates the establishment of priorities for research and Extension activities. A recent weed survey was conducted in Arizona cotton fields in the fall of 1989 (Heathman, 1990) with the results shown in Table 1and Table 2.
The development of Staple® (pyrithiobac-sodium) and transgenic cotton varieties resistant to Buctril® (bromoxynil) and Roundup Ultra® (glyphosate) allow the application of these herbicides over-the-top of small seedling cotton. These herbicide applications control broadleaf weeds during an early period of crop growth when it was previously difficult to achieve effective broadleaf weed control. Roundup Ultra also controls grass weeds and suppresses purple and yellow nutsedge. These herbicides combined with grass herbicides such as Select® (clethodim), Fusilade DX® (fluazifop-p-butyl), and Poast® (sethoxydim) make it possible to pursue a postemergence herbicide strategy for weed control prior to layby and reduce or eliminate the use of preemergence herbicides prior to or at planting. In addition, a recent focus on electro-hydraulic, articulated, quick-hitch guidance systems and improved cultivators by the University of Arizona Cooperative Extension Service has set the stage for the adoption of precision cultivation. Thus, the objectives of this project were to survey grower practices related to weed control and to survey the weeds present in cotton fields prior to the adoption and use of new weed control technologies.
Materials and Methods
Cotton fields were surveyed for weeds in 1995 and 1996 late in the season near harvest, using the quantitative survey method of Thomas (1985). The 82 (1995) and 83 (1996) fields surveyed were randomly selected by the Arizona Office of the National Agricultural Statistics Service (NASS). An inverted "W" pattern was used to systematically walk each sample field. Each leg or transect on the inverted "W" contained 5 approximately equally spaced quadrats for a total of 20 quadrats per field (Figure 1). Each quadrat consisted of 2 adjacent cotton rows by 10 feet of row. With a statewide average row spacing of about 38 inches, an average of 116 m2 were surveyed in each field. All weeds in each quadrat were identified, counted, and recorded for subsequent data entry and analysis. Five measures of weed abundance were calculated. Frequency (F) was calculated as the percentage of the total number of fields surveyed in which a species occurred in at least one quadrat (Thomas, 1985).
where Fk is the frequency for species k, Yi is the presence (1) or absence (0) of species k in field i, and n is the number of fields surveyed. Field uniformity (FU) was calculated as the percentage of the total number of quadrats sampled in which a species occurred (Thomas, 1985).
where Fuk is the field uniformity for species k, Xij is the presence (1) or absence (0) of species k in quadrate j in field i, and n is the number of fields surveyed. The density (D) of each species in a field was calculated by summing the number of plants in all quadrates and dividing by the area of 20 quadrats.
where Dki = density (in numbers m-2) of species k in field i, Zj is the number of plants of a species in quadrat j, and Ai is the area in m2 of 20 quadrats in field i. Mean field density (MFD) was the mean number of plants m-2 for each species averaged over all fields sampled.
where MFDk is the mean field density of species k, Dki is the density (in numbers m-2) of species k in field i, and n is the number of fields surveyed. Mean occurrence field density (MOFD) was the mean number of plants m-2 for a weed species averaged over only the fields in which that species occurred.
where MOFDk is the mean occurrence density of species k, Dki is the density (in numbers m-2) of species k in field i, n is the number of fields surveyed, and a is the number of fields from which species k is absent. Relative abundance (RA) was calculated by assuming that the frequency, field uniformity, and mean field density measures were equally important in describing the relative importance of a weed species. The relative frequency (RF), relative field uniformity (RFU), and relative mean field density (RMFD) were calculated by dividing the parameter by the sum of the values for that parameter for all species and multiplying by 100. For example, the relative frequency for species k (RFk) was calculated using:
The relative abundance of species k (RAk) was calculated as the sum of relative frequency, relative field uniformity, and relative mean field density for that species:
A maximum relative abundance value of 300 would have been possible if only one species were found in all of the fields that were surveyed.
The Arizona Office of the NASS summarized 1995 pesticide use data from 55,244 use records contained in 34,128 Arizona Department of Agriculture (ADA) 1080 forms. The 1996 pesticide use summary was based on 41,650 use records. The ADA 1080 form requires applicators to provide information on the pesticide applied including but not limited to: brand name, EPA registration number, rate, total chemical, total acres, label restrictions, days to harvest, section, township, range, and application method. All commercial applicators who apply pesticides are required to submit a 1080 form. Since virtually all pesticide applications made by air are made by commercial applicators the all aerial applications were reported to the ADA. In addition, applications of products on the Arizona Department of Environmental Quality's (DEQ) "Groundwater Protection List" must be recorded on ADA form 1080 and reported. [This list includes the following herbicides registered for use in cotton: Dual (alachlor), Bladex (cyanazine), Karmex (diuron), Eptam (EPTC), Poast (fluazifop-p-butyl), Cotoran (fluometuron), Roundup Ultra (glyphosate), Zorial (norflurazon), Caparol (prometryn), Poast (sethoxydim).] Some common herbicides such as Prowl (pendimethalin) and Treflan (trifluralin) are not on this list. Private applicators (i.e., growers who apply a pesticide on their own field) are not required to submit a 1080 form to the ADA, however, they are required to keep a record of the application.
Results and Discussion
The 15 most common weeds in the 1995 and 1996 surveys as judged by the parameter relative abundance (RA) are listed in Table 3 and Table 4, respectively. The mature weeds counted in these surveys were weeds that were not controlled by the cultural practices and herbicides used by growers. Thirteen weeds were common to both of the surveys and the three most common weeds in each survey were the same. However, the frequency (F), field uniformity (FU), and mean field density (MFD) varied considerably between years and thus the relative abundance varied between years for all of the common species listed. This large amount of variation is probably due to the small amount of area surveyed compared to the total area in a field and the small number of fields surveyed compared to the total number of cotton fields in the state. In addition, the distribution of weed species in cotton fields across the state is affected by many non-random factors related to farming practices, soil characteristics, and biological and environmental factors. Thus, each random selection of fields for a survey is likely to generate different results depending on where fields are located.
The fields surveyed in 1995 and 1996 were located predominately in the low desert areas of La Paz, Maricopa, Pinal, and Yuma Counties and the results of these surveys were similar to a survey of low desert areas conducted in 1989 (Table 1) with 13 of the most common species in 1989 appearing among the 15 most common weeds in 1995 and 1996. This suggests that few changes in farming practices or the way herbicides were used occurred between the two sets of surveys. The recalculated MOFD and MFD data from 1989 survey differ significantly from the MOFD and MFD data calculated for the 1995 and 1996 surveys. The reason for this is unknown but may be an artifact of different survey and calculation methodologies. Despite the variability inherent in the weed surveys, we believe they will be useful in monitoring weed species changes caused by the adoption of new weed control technologies and changes in herbicide use. The ranking of relative abundance also provides information useful in establishing research and Extension priorities.
The herbicides reported on the ADA 1080 form in 1995 and 1996 are listed in Table 5 and Table 6, respectively. These chemical use lists give some indication of the relative importance of the various herbicides in cotton production. However, many herbicide applications are not reported to the ADA because the herbicide is not on the Arizona DEQ's "Groundwater Protection List" or the application is made by a private applicator (i.e., a farmer applying herbicide to a field he is farming). (All herbicide applications made by commercial applicators must be reported on ADA 1080 forms). Three soil-applied herbicides, prometryn, pendimethalin, and trifluralin, accounted for over 75% of the total acres treated and over 66% of the total amount of active ingredients applied (Table 5 and Table 6). Prometryn, which is tank mixed with pendimethalin or trifluralin prior to or at planting, or used as a post-directed spray, or used as a layby treatment, was the most commonly reported herbicide on ADA form 1080. However, this may be an artifact caused by the presence of prometryn on the State's "Groundwater Protection List" that requires reporting prometryn applications and the absence of trifluralin and pendimethalin from this list. The six most widely used herbicides accounted for about 90 percent of both the acres treated and the total amount of active ingredient applied. Only one of these herbicides in each year, MSMA (1995) or clethodim (1996), were postemergence foliar applied herbicides. Soil applied herbicides accounted for about 92% and 87% of both the treated acres and active ingredient applied in 1995 and 1996, respectively. Herbicides that are strictly postemergence herbicides (i.e.,MSMA, fluazifop-p-butyl, clethodim, sethoxydim and glyphosate) accounted for about 7.5% and 12.7% of both the treated acres and active ingredient applied in 1995, and 1996, respectively. Both pyrothiobac-sodium (Staple) and bromoxynil (Buctril) were reported on ADA 1080 forms for the first time in 1996 indicating the initial adoption of new postemergence herbicide technology involving transgenic cotton varieties and early season over-the-top herbicide applications.
The Arizona Office of the National Agricultural Statistics Service (NASS) interviewed Arizona cotton farmers regarding the pesticides and cultural practices they used to produce cotton in 1995 and 1996. The herbicide applications that were reported in these interviews were listed in Table 7 (National Agricultural Statistics Service, 1995, 1996). The herbicide use reported in interviews with cotton growers was similar to that reported on the ADA 1080 forms in that preemergence herbicides such as trifluralin, pendimethalin, and prometryn were more widely used than postemergence herbicides. However, the sample size or number of reports for each herbicide in the NASS interviews was small compared to the ADA 1080 data. For example, in the 1996 NASS survey, the herbicides and number of reports for each herbicide were: trifluralin (35), prometryn (29), pendimethalin (13), diuron (3), and cyanazine (5) (personnel communication, Will Sherman). Statewide results of the NASS survey were pooled with other state surveys in order to describe national cotton herbicide use and the sample size was determined for this purpose. The small sample size and subsequent extrapolation to statewide totals, however, resulted in a highly variable description of the herbicides used in Arizona cotton production including under reporting of less commonly used herbicides. Thus the NASS survey results in Table 7 are not as reliable as the ADA 1080 data in Table 5 and Table 6. The discrepancies and differences between the two surveys suggests that anything less than complete reporting of all pesticide applications will result in data that can be difficult to interpret. Larger sample sizes in the weed surveys and grower interviews would also improve the quality of the results. These data (Table 5, Table 6, and Table 7) should be viewed as rough approximations because some herbicides (including prometryn, cyanazine, diuron, oxyfluorfen, and fluometuron) can be used as both preemergence, soil-applied herbicides or as postemergence, foliar-applied, post-directed, herbicides. In addition, many herbicides applications are made by growers and these applications are not currently reported to the ADA.
The Arizona Office of the NASS, in collaboration with the University of Arizona, conducted a telephone survey of 249 randomly selected growers who farmed 139,697 acres of cotton in 1995 (out of a statewide total of 412,100 acres). Grower responses provided a description of weed control practices used in the 1995 Arizona cotton crop. The average statewide cost for hand weeding was reported as $27.87 per acre in addition to other weed control costs. Statewide, most growers used preemergence herbicides before or at planting and pre- and post-emergence herbicides later in the season. Most of these applications were broadcast applications suggesting that many of the postemergence herbicide applications were layby applications. The practice of banding postemergence herbicides will probably increase in the future with increasing use of Staple, Roundup Ultra, and Buctril. Statewide, few growers band preemergence herbicides or use electro-hydraulic quick-hitch guidance systems and in-row weeding tools with their cultivators. (For a more detailed analysis of cotton grower practices in 1995 see the article titled "IPM/BMP Practices in Arizona Cotton" in this Cotton Report.)
A similar survey of 76 growers conducted by the Arizona Office of the NASS in 1996 found that only 1 of 76 fields was planted to a herbicide resistant variety (BXN57, BXN58). Respondents indicated that scouting is a commonly used practice in Arizona cotton production. Of the 76 reports summarized, 63 scouted for weeds, 68 for insects, and 63 for diseases. Most commonly, the operator or a family member did the scouting (24/63). Other methods of scouting were by crop consultants (18/63) and dealers (17/63). On the question of keeping records of the scouting results, 7 of 76 kept records of weed distribution, but 24 of 76 recorded insect findings. Preemergence herbicides were applied by 57 of the 76 reporting growers. The decision to apply the product was based primarily on experience from previous years (53/76). Some growers applied herbicide based on the recommendation of crop consultants (18/76) and a few (9/76) based on field mapping data. Only 31 of 76 growers reported applications of postemergence herbicides. Some (17/76) made the application as a routine treatment, others (18/76) applied the product based on the density of weeds, and a few (8/76) relied on the advice of crop consultants. All 76 growers reported using cultivation to control weeds. The earliest cultivation date reported was March 25 and the latest was August 27. Dates were asked for 3 cultivations. The average dates were May 1 (first cultivation), May 25 (second cultivation), and June 10 (third cultivation). Some 46 of 56 responding growers indicated that a moldboard plow was used on the selected field. When asked the primary outside source of information on pest management, 40 of 76 indicated crop consultants, 19 picked farm supplier or chemical dealer, and 7 relied on Cooperative Extension information. Fifty six of 76 growers had received training and were certified to apply restricted use pesticides.
In summary, the various surveys provided a description of weed species distribution, the herbicides used, and cultural practices used for cotton weed control prior to the widespread adoption and use of new herbicide technologies such as Staple herbicide, transgenic cotton varieties resistant to Buctril and Roundup Ultra, and precision guided cultivation with in-row weeding. These baseline data will allow the documentation of changes in Arizona cotton production practices in the future.
The authors wish to thank all of the Arizona Agricultural Statistics Service enumerators who walked cotton fields under sometimes harsh conditions to survey weed populations in Arizona. This work was partially funded by a grant from the USDA NAPIAP (95-EPIX-1-0195).
This is a part of publication
AZ1006: "Cotton: A College
of Agriculture Report," 1998, College of
Agriculture, The University of Arizona, Tucson, Arizona,
85721. Any products, services, or organizations that are
mentioned, shown, or indirectly implied in this
publication do not imply endorsement by The University of
Arizona. The University is an Equal
Opportunity/Affirmative Action Employer.