Final Report for GS03-025
Presence/absence (sequential binomial) sampling methods have been developed that can provide quick and easy estimates of both greenbug and parasitoid densities in the wheat field. Specifically these methods were designated “Glance ‘n go” and “Glance ‘n go plus parasitoids” respectively. These sampling methods were evaluated over the 2003-2004 growing season in 10 cooperator supplied wheat fields. In each field four 30 x 30 m plots were established. Each plot randomly received one of four common management treatments: 1. Glance ‘n go (sequential binomial sampling for greenbug densities), 2. Glance ‘n go plus parasitoids (sequential binomial sampling for both greenbug and their parasitoid densities), 3. Prophylactic insecticide applications, 4. No insecticide applications. Whether data were analyzed for all ten fields, with a covariant for cheat Bromus secalinus L. competition, or by only analyzing data from those fields where an insecticide treatment was recommended and made on the Glance ‘n go managed plots, there were no significant differences in the grain yield or one hundred seed weights among treatments. This research indicates that Glance ‘n go and Glance ‘n go plus parasitoids management practices result in grain yields that are not statistically significant, than prophylactic insecticide applications. The management schemes are simple to apply and should save the wheat producer money since the producer will make fewer unnecessary insecticide applications.
Wheat, Triticum aestivum L., is a multipurpose cereal crop that is grown in the Southern Great Plains for grain production, forage production or a combination of the two (Thompson 1990). Annually, some 6 to 7 million acres are planted in Oklahoma (Krenzer et al. 1999). In 1998, 6.4 million acres of winter wheat were planted in Oklahoma, of which about 4.3 million acres were harvested for grain with an average yield of 34 bushels per acre (Krenzer et al. 1999).
Wheat is attacked by many herbivores, including aphids that can inflict losses (Royer et al. 1998). Aphids are of particular interest because they have been observed to damage wheat from plant emergence to heading. Aphids reproduce rapidly, and are often not detected by farmers until their populations reach deleterious levels. These aphid pests include greenbug Schizaphis graminum (Rondani), Russian wheat aphid (Diuraphis noxia Mordviko), bird cherry-oat aphid (Rhopalosiphum padi L.), English grain aphid (Sitobion avenae Fabricius), and corn leaf aphid (Rhopalosiphum maidis Fitch) (Royer et al. 1998). Arguably the most important of these aphid pests is the greenbug.
Greenbugs infest a wide variety of crops and wild hosts throughout the central United States, feeding on over 70 graminaceous species many of which serve as secondary hosts when winter wheat and other grain crops are not present (Michels 1986). First reported in the United States as an agronomic pest of wheat in 1882 (Pfadt 1962), greenbugs can reach tremendous population levels in a short period of time (Starks and Burton 1977). Outbreaks occur in Oklahoma almost every year, and statewide infestations are reported about every 5-10 years (Starks and Burton, 1977). When population levels surpass economic injury levels, greenbug feeding reduces yield and crop quality (Elliott et al. 1994, Kindler et al. 2002). In Oklahoma, losses range from $0.5 to $135 million annually, though much of the losses are due to the expense of insecticide use (Starks and Burton 1977, Webster 1995).
Greenbugs are attacked by a number of predators and parasitoids (Royer et al. 1998). One of the most important examples of these natural enemies in the Southern Great Plains is Lysiphlebus testaceipes Cresson (Hymenoptera: Aphidiidae) (Kring and Gilstrap 1983). Lysiphlebus testaceipes is a common parasitoid of cereal aphids throughout temperate regions of North and South America and has great potential for destroying large numbers of greenbugs (Pergande 1902, Sekhar 1957, Wood and Chada 1969, Krombein et al. 1979, Salto et al. 1983, Jones 2001, Giles et al. 2003). Their effects on greenbug populations can be dramatic. Female L. testaceipes oviposits in all life stages of the greenbug. When parasitized as adults, greenbugs stop reproducing about three days after being parasitized by L. testaceipes (Spencer 1926, Eikenbary and Rogers 1974).
Because grain yield losses are directly related to greenbug population levels, a population assessment (sampling) is required to estimate the potential for economic losses, and whether insecticides are cost effective to apply (Royer et al. 1998, Kindler et al. 2002, Elliott et al 2003a,b). Historically, in Oklahoma, greenbug infestations have been estimated by relatively laborious methods that estimate greenbug populations per 0.3m of crop furrow, or on a per tiller basis (Royer et al. 1998). A simplified sequential binomial sampling system for classifying greenbug densities has been developed (Royer et al. 2002). Coined “Glance ‘n go,” this method involves looking at randomly selected tillers and noting the presence or absence of greenbugs on each tiller. Because the proportion of tillers that are infested accurately corresponds with greenbug density (Giles et al. 2000), samplers can quickly classify high or low greenbug populations through sequential sampling. The important goal of Glance ‘n go sampling is not to determine the exact greenbug density in the field, but rather to classify the likely density as being above or below the economic threshold (ET). The economic threshold is the density at which control measures should be taken to prevent an increasing pest population from reaching the economic injury level (EIL), at which significant losses to the producer occur (Stern et al. 1959). Due to the simplicity, timesaving, and ease of using Glance ‘n go, producers could be more likely to sample for greenbugs and make educated decisions about insecticide applications.
The Glance ‘n go method of greenbug sampling, while simple and easy to use, deals primarily with classifying greenbug population densities as being high or low and does not include sampling for important effects of natural enemies such as parasitoids. To remedy this deficiency, a complimentary sequential binomial sampling method for classifying parasitoid densities in winter wheat fields was been developed. Giles et al. (2003) demonstrated that when the proportion of tillers with parasitized aphids that are pupating (mummies) exceeds 10%, the actual percent parasitism could always be classified as being above 20%. Based on research by Jones (2001), when this level of parasitism is reached, greenbug population densities will quickly decline. Being able to simultaneously predict parasitoid efficacy for greenbug control during greenbug sampling would allow producers to reduce the misuse of pesticides, thus increasing producer profits, reduce the incidence of greenbug resistance to insecticides and reduce negative impacts of pesticides on the environment.
Glance ‘n go and Glance ‘n go plus parasitoids are simple, easy to use sampling methods. However, because these methods are new and their effectiveness has yet to be validated, few producers have utilized them. Our objective was to utilize these sampling methods in various locations from north to south in Oklahoma during the 2003-2004 winter wheat-growing season and compare the management implications of Glance ‘n go and Glance ‘n go plus parasitoids. This was accomplished by determining yield of wheat fields sampled and managed by Glance ‘n go, Glance ‘n go plus parasitoids, with wheat grown without greenbugs (prophylactic insecticide applications), and with wheat that is not treated with insecticide under any circumstances (no insecticide).
A total of ten research fields were secured in Caddo (Apache, Boone, and Jackson), Canadian (El Reno and Yukon), Grady (Chickasha), Kay (Quance #1 and Quance #2) and Payne (Kastl #1 and Kastl #2) counties in Oklahoma during the winter wheat-growing season of 2003-2004. Each site was located such that it was a minimum of 1 km away from any other research site. At each site, four 30 by 30m plots were established with a 10m buffer zone between plots. Each plot was sampled and subsequently managed using one of four aphid management scenarios. These were; Glance ‘n go (Royer et al. 2002), Glance ‘n go along with sequential binomial sampling for parasitoids (Giles et al. 2003), insecticide applications once a month (prophylactic insecticide applications), and no insecticide applications. When a management strategy dictated that insecticides were needed, chlorpyrifos (Lorsban® 4E, Dow AgroSciences) was applied at a rate of 1.6 liters per hectare. Imidacloprid (Provada®, Bayer AG, Leverkusen, Germany) was also applied at a rate of 335g per hectare on those plots that received regular monthly insecticide applications. Applications were made using a custom built, carbon dioxide powered, bicycle wheeled sprayer (3.3m boom).
Each plot was sampled every thirty days to classify aphid and parasitoid densities. Sampling was conducted as often as necessary when greenbug populations neared economic thresholds. Sampling was accomplished by following Glance ‘n go guidelines (Royer et al. 2002). These guidelines were determined before each field trip by accessing the Greenbug Management Decision Support System web-site (http://www.pswcrl.ars.usda.gov/gbweb/index.htm) and entering the date, field location, estimated expense of spraying insecticide and the estimated ultimate value of the wheat crop. For decision purposes, we chose to use $7.00/acre as the expense of spraying and $3.00 per bushel as the anticipated value of the crop. Sampling and subsequent management continued throughout the growing season until wheat reached the boot stage (Feekes stage10-10.1).
Just before harvest, samples were taken by clipping 0.3m of linear crop row at ground level, at ten randomly selected locations throughout each plot. The samples were placed seed-head first into brown paper bags, staple sealed, returned to the laboratory, and threshed using a single head thresher (Precision Machine Company Inc., Lincoln NE). For each sample, the number of tillers, seed heads, and the number of stems of cheat, Bromus secalinus L. were recorded. Total grain weight and 100 seed weights were determined on a digital scale (Sartorius, Göttingen, Germany).
We analyzed the data three times using SAS PROC MIXED (SAS Institute 1999). The first analysis was done using all data from all 10 fields. Next, on some of the plots, because cheat compromised the yield results. In order to compensate for the presence of cheat, raw yields and 100 seed weights were analyzed to include a covariant for the amount of cheat that competed with the wheat during the crop season. Finally, the data for those fields where a decision was made to spray insecticide on one or more of the Glance ‘n go or Glance ‘n go plus parasitoids managed plots (Apache, Boone, Chickasha, and Jackson fields) were analyzed separately.
When data were analyzed for all ten fields, the four management practices resulted in mean yields that were not significantly different from one another (F 3,27.2= 0.37, P = 0.78). A mean yield of 22.03g ± 0.89 (mean ± SE) per 0.3m of crop row was recorded for Glance ‘n go managed plots. Plots managed by Glance ‘n go plus parasitoids produced a mean yield were of 21.40g ± 0.81 per 0.3m of crop row. Unsprayed plots had a mean yield of 21.46g ± 0.98 per 0.3m of crop row and the sprayed plots had a mean yield of 22.35g ± 0.94 per 0.3m of crop row.
Mean one hundred seed weights were also not significantly different from one another (F 3/27.1= 0.52, P = 0.67). Mean one hundred seed weights were 3.142g ± 0.039 for Glance ‘n go plots, 3.107g ± 0.043 for Glance ‘n go plus parasitoids managed plots. Unsprayed plots had a mean one hundred seed weight of 3.125g ± 0.034 and 3.195g ± 0.039 for sprayed plots.
Data analyzed with a covariant for cheat competition for all ten fields also showed no significant differences in mean yield (F 3/27.4= 0.60, P = 0.62) or one hundred seed weights yield (F 3/27.1= 0.53, P = 0.67) between treatments. Adjusted mean yield estimates were 22.14g ± 0.89 per 0.3m of crop row for Glance ‘n go managed plots. 21.16g ± 0.81 per 0.3m of crop row for Glance ‘n go plus parasitoids managed plots. Unsprayed plots yielded 21.43g ± 0.98 per 0.3m of crop row and sprayed plots yielded 22.40g ± 0.94 per 0.3m of crop row. Adjusted mean one hundred seed weights were 3.141g ± 0.039, 3.108g ± 0.043, 3.125g ± 0.034, and 3.197g ± 0.042 for Glance ‘n go, Glance ‘n go plus parasitoids, non-sprayed and sprayed plots respectively.
Greenbug infestations ranged from minor to major in a general trend from north to south across Oklahoma. Decision thresholds were surpassed in four of the ten fields that were monitored (Apache, Boone, Chickasha, and Jackson fields). Those plots with greenbug populations that surpassed Glance ‘n go decision thresholds were immediately treated with chlorpyrifos. Plots managed with Glance ‘n go plus parasitoids exceeded treatment thresholds, but in every case parasitism was sufficient to overturn the treatment decision.
Data analyzed from these four fields showed no significant differences in yield (F 3/9.0= 0.35, P = 0.79) or one hundred seed weight (F3/9.1 = 0.27, P = 0.84) among treatments. Mean yields were 20.02g ± 1.11, 18.82g ± 0.92, 19.26g ± 1.25, and 18.73g ± 1.07, for Glance ‘n go, Glance ‘n go plus parasitoids, non-sprayed and sprayed plots respectively. Mean one hundred seed weights were 2.969g ± 0.058, 3.043g ± 0.077, 2.952g ± 0.054, and 3.041g ± 0.074 for Glance ‘n go, Glance ‘n go plus parasitoids, non-sprayed and sprayed plots respectively.
The common result to all of the analyses is that there were no significant differences in yield or one hundred seed weights among treatments. Had there not been large numbers of parasitoids present in the fields, we would have expected a significant yield loss on the unsprayed plots. Previous research has determined that wheat infested with greenbugs can exhibit yield losses as much as 14.5 kg/ha (0.22 bu/ac) for each greenbug per tiller during years with average precipitation and losses of 34.3 kg/ha (0.51 bu/ac) under severe drought conditions (Kindler et al. 2002, 2003). We encountered greenbug intensities as high as 14.25 greenbugs per tiller on the unsprayed plots. This would translate into a yield loss of approximately 206.6 kg/ha (3.14 bu/ac) under average precipitation conditions and could reach losses of 489 kg/ha (7.27 bu/ac) under drought conditions. However the high parasitism rates encountered on the Glance ‘n go plus parasitoid managed plots was also encountered on the unsprayed plots as well. This caused the treatment effect on those plots to be virtually indistinguishable from the Glance ‘n go plus parasitoids managed plots.
Though we discounted results from the unsprayed plots, it was still possible to determine that management of winter wheat using Glance ‘n go or Glance ‘n go plus parasitoids resulted in savings to the farmer. These savings were primarily due to the farmer making fewer unnecessary insecticide applications. Previous research has indicated that the primary loss due to greenbug infestations on wheat is due to unnecessary insecticide applications (Starks and Burton 1977, Webster 1995). Unnecessary applications of insecticide were also seen in this experiment. Glance ‘n go management guidelines recommended that insecticide applications be made on plots at Apache, Boone, Chickasha, and Jackson. Yet, the yields for the Glance ‘n go and the Glance ‘n go managed plots at these fields, were not statistically different from each other (20.02g vs. 18.82g per 0.3m of crop row, P = 0.42).
It should be cautioned that this experiment spanned only one growing season with a little below average precipitation. Other growing seasons might bring about conditions that cause these management practices to fail. However this experiment along with the work of Royer et al. (2002) and Giles et al. (2003), suggest that these management systems are robust enough to be used as greenbug control management practices for winter wheat in the Southern Great Plains.
Educational & Outreach Activities
Jones, D.B. 2004. Biology of Cereal Aphid Parasites in Winter Wheat: What Do These Wasps Do All Winter, and what should we do About It? 50 min presentation: Central Plains Wheat Production Workshop. Garden City KS, March 25.
Jones, D.B., K.L. Giles, N. C. Elliott, and T. A. Royer. 2005. Integrating Effects of Natural Enemies into Winter Wheat Greenbug Management. 10 min presentation: Southwestern Branch of the Entomological Society of America. Albuquerque NM, March 1.
Jones, D.B., K.L. Giles, N.C. Elliott, and T.A. Royer. Integrating Effects of Natural Enemies into Winter Wheat Greenbug Management. Planned as part of a larger paper reviewing integrated pest management of greenbug in wheat in Oklahoma. This paper is planned to be submitted to the American Entomologist.
This research was aimed at providing validation for the Glance ‘n go and the Glance ‘n go plus parasitoids management systems in winter wheat in Oklahoma. Additionally, we intended to demonstrate the method to several of the leading wheat producers in Oklahoma whose influence could sway others to adopt the practices. We also planned to publish our results and make presentations about the work. To date we have made presentations at several venues such as the Entomological Society of America meeting, and the Central Plains Wheat Production Workshop; publications are forthcoming. We have indeed provided validation for both management systems and our outreach should prove to be successful since our producer/cooperators included two members of the Oklahoma Wheat Commission (Paul Jackson and Leroy Quance). These gentlemen have influence on the wheat growing industry that stretches far outside of Oklahoma. Other cooperators were Brad Tipton who is an extension agent in Canadian county Oklahoma in addition to farmers such as Dennis Kastl, Don Bornemann, and Ken Mach. Through their testimonials to other farmers, our publications and presentations, these management strategies have a good chance of being adopted widely throughout the Southern Great Plains.
Because greenbug losses are frequently because of unnecessary applications, a simple economic benefit can be calculated from the data in the experiment. If we assume that the cost of insecticide applications are $7.00 per acre, then the monetary benefit to the farmer would be $7.00 per acre during the 2003-2004 growing season. This is because we were able to show that the mean yield on the plots treated with prophylactic insecticide applications was statistically insignificant from those managed by Glance ‘n go plus parasitoids. None of the Glance ‘n go plus parasitoid managed plots required insecticide treatments because the parasitoid population density was high enough to prevent the greenbug population from reaching the economic injury level. Management of greenbug infestations by Glance ‘n go or Glance ‘n go plus parasitoids has another outstanding benefit that can save the farmer money. Because the management system is so easy to use, and sampling takes very little time, producers are willing to monitor crops regularly ensuring that increasing aphid populations are managed before they reach outbreak status and destroy crops. This should minimize losses due to greenbug feeding to only the expense of insecticide applications when they are called for.
Determination of farmer adoption was not a part of this work. However, a related separately funded study is underway to determine farmer opinions and practices. Results of this study are yet to be determined.
Areas needing additional study
This sampling and management system should be followed over a number of growing seasons to determine if there are situations where it might fail to make correct greenbug management decisions.