Development of Sustainable, Cost-Efficient Strategies for
Managing Cotton Insects -- An Interim Report

INTRODUCTION

      In response to this need, the project "Development of Sustainable, Cost-Efficient Strategies for Managing Cotton Insects" was initiated by the Mississippi Agricultural and Forestry Experiment Station in early spring of 1995. Objectives of the project were to: (1) establish a coordinated program of evaluating cotton insect management options in replicated small plots in the Delta and Hill regions of the state; (2) initiate an experimentally sou nd program of evaluating major changes to current recommended strategies in production-level fields; and (3) revitalize a cooperative management-science effort to develop decision-making tools and simulators necessary to introduce new management capabilit ies, guide research projects, and create a proactive approach to developing new cost-effective management options for cotton insects.

      One new management option for control of the tobacco budworm (Heliothis virescens [F.]) and the bollworm (Helicoverpa zea [Bodie]) is the use of transgenic cotton containing the

      Beginning in 1995, Monsanto supplied seed for planting a total of 200 acres of Bt cotton and its conventional parent variety in field-size research plots at five locations in the cotton-growing areas of Mississippi. Dat a from this cooperative study allow a comparison of the relative economic benefits and costs associated with growing transgenic and common conventional cotton varieties.

      The field-size plots were also used in an evaluation of management strategies for controlling the tarnished plant bug (Lygus lineolaris [Palisot de Beauvois]). Conventional tarnished plant bug control recommendat ions were compared to an aggressive early-season strategy. Conventional control consisted of treatments described in the Mississippi Cotton Insect Control Guide. The aggressive strategy included a preventative, early-season treatment of acephate. < p>       These studies served to amass an entomological data set that will benefit future research to analyze pest and beneficial insect trends and to evaluate the efficacy of insect control procedures in the major cotton-growin g areas of Mississippi. Future analysis of the data also will provide information that may lead to an increased understanding of factors associated with dramatic pest population growth changes, such as the one that occurred in the 1995 tobacco budworm epi zootic. Much of this research focused on providing data to modify and develop computerized insect control decision-making aids and to revitalize computer models for testing innovative management strategies in a simulated environment.

      This bulletin describes results obtained during 1995 and 1996 in the field-size research plots at five Mississippi locations. A detailed report of findings for three years (1995-1997) of research in these plots will follow this publication. The follow-up bulletin also will report on systems research to date and on small-plot research that emphasizes pest control strategies and efficient pesticide use at the Delta Branch Experiment Station in Stone ville and the North Mississippi Branch Experiment Station in Holly Springs.

MATERIALS AND METHODS

Methodology Used in Both Years

      Fields were located in Lee, Leflore, Madison, Tallahatchie, and Yazoo counties in 1995, and in Itawamba, Leflore, Madison, Tallahatchie, and Yazoo counties in 1996. The Leflore County site included four field-sized rep licates of each treatment to allow estimation of variability within a single farm relative to variability across regions. All other locations included only one replicate of each treatment. Thus, there were a total of eight replicates at five different loc ations spanning the cotton-growing region of the state. Although an effort was made to locate all treatments within each site in adjacent fields, variations in fields did exist because of location. Fields at both Hill locations were different in slope and soil characteristics and were typically more closely associated with woods and wild flowers than the Delta locations.

      The growers maintained plot maintenance records, including planting date, insecticide application dates and rates, plant growth regulator application information, defoliation and harvest dates, and yield. In Itawamba Co unty, however, yield for each field was estimated based on the area required to produce a single module. The weight of the cotton module was obtained from the gin report. The area required to produce the module was calculated by counting and measuring the rows and converting the row-feet included in the area to acres. Seed cotton obtained from a second picking in Itawamba County was weighed in the boll buggy or cotton wagon by use of truck scales under each wheel, and the weight of lint was computed based on the gin turn out reported from the first picking. Yield data were used to compare total insecticide cost and cost of controlling heliothine larvae and tarnished plant bug based on the estimated cost of insecticides reported in the 1997 Delta Planning Budgets.

1995 Insecticide Management Strategy

      The aggressive tarnished plant bug control strategy consisted of a prophylactic application of acephate (Orthene 90S) at 0.33 pound (active ingredient) per acre at the fourth-true-leaf stage of plant development. A seco nd treatment was applied during the first week of squaring if 2 plant bugs were found per 100 sweeps with a 15-inch diameter sweep net. During the second week of squaring, a third application was used if 4 plant bugs were found per 100 sweeps. Subsequent control of all insects was based on the MCICG. This treatment is designated as GV-ETPBO (Early Tarnished Plant Bug - Orthene).

      At three locations, (Leflore, Yazoo, and Tallahatchie counties), additional fields of conventional cotton were treated with a similar aggressive tarnished plant bug strategy. However, in these treatments a pyrethroid wa s used rather than acephate. This approach was designated as GV-ETPBP (Early Tarnished Plant Bug - Pyrethroid). All treatments at the Leflore County location were replicated four times, and plots within each replicate were adjacent to each other.

1996 Insecticide Management Strategy

      The aggressive tarnished plant bug management strategy for 1996 was similar to that used in 1995. However, unlike the 1995 strategy, follow-up applications were not triggered by tarnished plant bug thresholds. Orthene 9 0S was applied at 0.33 pound (active ingredient) per acre beginning at the 4th true leaf stage. Two subsequent applications were made at 7-day intervals after the prophylactic treatment. This change was made so that other parameters, such as the effect of these sprays on beneficial insects, could still be monitored even if plant bug populations were too low to trigger an application.

      In 1996, the Bt cotton fields at the Itawamba and Tallahatchie locations were separated from the conventional (GV-MCICG) fields by at least 1 mile. Thus, the assumption that each field within these locations is subject to the same insect populations and population densities may be questionable. Fields located at Madison and Yazoo counties were adjacent, and the fields at Leflore County within each replicate were located close to each other.

RESULTS FROM 1995 AND 1996

1995 Insecticide Management Study Results

      Population densities of tarnished plant bugs were low in all locations during 1995. Based on means of the five locations, DPL 5415 and grower's-choice conventional cotton fields managed by the MCICG collectively averaged 0.8 applications of insecticide for tarnished plant bug management. Conventional cotton receiving the aggressive plant bug treatments averaged 1.1 applications, and the NuCotn 33B (BT-MCICG) averaged 1.4 applications. Mean yields across all five locations were similar for both the MCICG tarnished plant bug management strategy (GV-MCICG) and the early-season acephate treatment (GV-ETPBO). GV-MCICG fields yielded 813 pounds of lint, while GV-ETPBO yielded 820 pounds.

      Averaged over all locations, insecticide cost for the transgenic cotton fields was less than half the cost of heliothine control for the conventional varieties managed under MCICG guidelines. Economic benefit was statistically similar (LSD; p = 0.05) for all of the treatments in the study when averaged across all locations. Yield ranged from 19.5 pounds of cotton lint produced per dollar invested in insecticide in the GV-MCICG fields to 25.3 p ounds per dollar invested in the BT-MCICG fields. The DPL5415, GV-ETPBO, and GV-ETPBP fields returned 23, 21.6, and 23 pounds of lint, respectively, per dollar spent on heliothine control.

      Cotton in the Leflore County GV-MCICG and GV-ETPBO treatments was harvested as one field in each replicate, which made it impossible to separate any effects based on yield from the two treatments. The field with these same treatments in Lee County was damaged by hail on July 7, 1995. Although the two treatments were in the same field, the damage in both treatments may not have been equal. In addition, Lee County suffered severely from the tobacco budworm infestation. Unexpectedly high insecticide use in the Midsouth caused a sho rt supply of preferred insecticides for budworm control late in the growing season, compounding the problem in Lee County. This insecticide shortage, combined with the occasional inability to spray at close intervals because of rain or other factors, resu lted in drastically reduced yields in all conventional cotton at the Lee County location.

      Average yield and standard deviation (SD) from GV-MCICG and GV-ETPBO fields at the other four locations were essentially equal. Average yield for GV-MCICG fields was 813 (SD 127.3) pounds of lint per acre; for GV-ETPBO fields, 820 pounds (SD 71.2). When Lee County is included in the overall calculation, the mean yield is reduced to 681 (SD 283.4) pounds of lint for the GV-MCICG treatment and 665 pounds (SD 313.7) for the GV-ETPBO treatment. In comparison, fields in Lefl ore, Tallahatchie, and Yazoo counties that used pyrethroids for plant bug control (GV-ETPBP) averaged 840 (SD 45.3) pounds of lint per acre when averaged by location (three means). The average of the six replicates (four in Leflore County, one each in Yaz oo and Tallahatchie counties) was 824 (SD 128.4) pounds of lint per acre.

      Averaged across locations, mean yield was 833 (SD 166.7) pounds of lint per acre for the BT-MCICG fields and 622 (SD 229.7) pounds for the DPL5415 fields. BT-MCICG and DPL5415 yields averaged by the eight replicates, in stead of by location, resulted in 840 (SD 139.3) and 645 (SD 204.0) pounds of lint cotton per acre, respectively. Yields from four locations (using the mean of the four replicates at Leflore County and excluding the low yield of the Lee County location) f or the BT-MCICG and DPL5415 fields were 886 (SD 135.4) and 714 (SD 117) pounds of lint, respectively. Although results from Lee County appeared to be an outlier, as compared to other locations in the study, it is nonetheless a valid indication of the budw orm problem in this area in 1995 as documented by Luttrell et al., 1997.

      Some of the insect sampling data for 1995 from the five locations are summarized in Table 3. The population density of tarnished plant bugs was low in all fields, but transgenic cotton averaged 0. 48 (SD 0.794) tarnished plant bugs per 25 sweeps season-long over all locations, and conventional cotton averaged 0.23 (SD 1.445). This difference was statistically significant (ANOVA; p = 0.0001). The number of tarnished plant bugs in all fields in the b oll weevil eradication zone, which included Lee County, may have been reduced by the multiple applications of ULV malathion used by the eradication program.

      The ability of Bt cotton to prevent damage from the tobacco budworm is strongly indicated by the damaged square data. Based on the data in Table 3, transgenic cotton sustained a seasonal average of 0.98% damaged squares, fourfold less than the average of all conventional varieties (3.92%). Such a dramatic difference was measured despite the fact that conventional cotton was managed with conventional insecticides according to current Extension recom mendations for lepidopteran pests. Season-long averages of percent damaged squares in conventional cotton varied greatly in each county: Lee, 9.6% (SD 1.40); Madison, 1.5% (SD 0.38); Tallahatchie, 6.8% (SD 1.10); Yazoo, 2.3% (SD 1.13); and Leflore, 0.2% (SD 2.55). These data indicate the severity of the epizootic of tobacco budworm during 1995 in both Lee and Tallahatchie counties.

      Throughout the season in these counties, mean percent of terminals of all varieties with heliothine eggs was 5.72% (SD 3.34). Bt cotton averaged 7.6% (SD 3.09) infested terminals, and the conventional cotton averaged 5. 2% (SD 3.3). These differences were statistically significant (ANOVA, p = 0.018) and may reflect adult mortality resulting from insecticidal sprays in conventional cotton.

1996 Insecticide Management Study Results

      When the cost of the Bollgard^TM license is included in the cost of Bt cotton, 1996 insecticide expenditures for heliothine larvae control was higher in transgenic cotton fields than in conventional cotton fields (Table 4). The average cost for heliothine-directed insecticides was $37.56 for the BT-MCICG fields, $23.14 for the GV-MCICG fields, and $36.04 for the BT-ETPBO fields. These values demonstrate that transgenic cotton may not be as co st effective as conventional cottons when larval densities are very low. It certainly was cost effective under the high tobacco budworm densities seen in 1995.

CONCLUSIONS

      The average number of insecticide applications made per location for tarnished plant bug control during the 2-year period in ETPBO fields was 4.15 applications. Those managed strictly according to the Mississippi Cot ton Insect Control Guide received an average of 1.35 applications. Results of the first 2 years of research indicated that tarnished plant bug population densities were low in all fields in the five locations. Under such conditions, the strategy of ea rly-season scheduled or aggressive management of tarnished plant bugs with applications of Orthene 90S provided no benefit in yield. When the tarnished plant bug population is low, these results indicate no need to alter current management recommendations for this pest.

      There were no statistical differences in yield, total insecticide cost, or heliothine insecticide cost between the MCICG-based transgenic and conventional cotton treatments when summarized over the 2-year period (ANOVA; p = 0.05) (Table 6). These mean values incorporate data from a year of high heliothine populations and a year of relatively low heliothine numbers. Data from individual years suggest that the Bollgard-protected cotton would p rovide a greater return than conventional cotton in years of high-budworm pressure, while the reverse would be true in years of low-budworm pressure.

References

Harris, F.A., R.E. Furr, and D.S. Calhoun. 1996. Cotton insect management in transgenic Bt cotton in the Mississippi Delta, 1992-1995. In: Proc. Beltwide Cotton Conf., National Cotton Council, Memphis, TN. pp. 854-859.

Luttrell, R.G., F.T. Cooke, T.B. Freeland, Jr., J. Gibson, F.A. Harris, D.H. Laughlin, J.T. Reed, and C.E. Watson. 1997. Study of cotton insect control problems in Mississippi during 1995 with special emphasis on boll weevil eradication and tobacco budworm. Mississippi Agricultural and Forestry Experiment Station Technical Bulletin 217. 29 pp.

Luttrell, R.G., V.J. Mascarenhas, J.C. Schneider, C.D. Parker, and P.D. Bullock. 1995. Effect of transgenic cotton expressing endotoxin protein on arthropod populations in Mississippi cotton. pp. 760-763. In: Proc. Beltwide Cotton Conf., Nat ional Cotton Council, Memphis, TN.

Mascarenhas, V.J. 1994. Direct and indirect effects of transgenic cotton expressing the delta-endotoxin of Bacillus thuringiensis on survival of tobacco budworm (Heliothis virescens) and bollworm (Helicoverpa zea). M.S. Thesis, Mississippi State University, 102 pp.

Jenkins, J.N., W.L. Parrott, J.C. McCarty, F.E. Callahan, S.A. Berberich, and W.R. Deaton. 1993. Growth and survival of Heliothis virescens (Lepidoptera: Noctuidae) on transgenic cotton containing a truncated form of the delta endotoxin gene from B acillus thuringiensis. J. Econ. Entomol. 86:181-185.

Acknowledgments