Using cultivar mixtures to improve pest control and grain crop production

Final Report for LNE12-320

Project Type: Research and Education
Funds awarded in 2012: $164,919.00
Projected End Date: 12/31/2015
Region: Northeast
State: Pennsylvania
Project Leader:
Expand All

Project Information


Recent ecological research and our preliminary data indicated that intraspecific genetic diversity can be quite valuable for suppressing pest populations and improving primary productivity. Unfortunately, most fields planted with modern crop varieties harbor very little genetic diversity.

Our integrated research and education project explored the pest-management potential of genotypically diverse cultivar mixtures to better resist populations of plant pathogens and insect herbivores to improve crop productivity. Moreover, because cultivar mixtures tend to improve yield over monocultures even in absence of pests, our work also assessed general improvements in crop productivity by increasing genotypic diversity.

Using field experiments at two research centers and farms of three cooperating growers, we quantified production benefits of planting fields containing several different varieties of wheat. At all these sites, we compared insect pests and natural enemies, and plant pathogens among cultivar mixtures and the constituent monocultures.


Conventional field-crop growers rely heavily on pesticides, which can have deleterious effects on human and environmental health. In contrast, organic field-crop growers have few tools that they can use economically to combat pest invasions in their fields. Therefore, designing cropping strategies that are amenable to both conventional and organic production and generate crop fields that are more resilient and better able to resist pest invasion should improve crop productivity for a wide range of farmers.

Recent ecological research,including some of our own, indicate that intraspecific genetic diversity can be quite valuable for suppressing pest populations and improving primary productivity. Unfortunately, most fields planted with modern crop varieties harbor very little genetic diversity. Our integrated research and education project explored the pest-management potential of genotypically diverse cultivar mixtures to better resist populations of insect herbivores and some plant pathogens. Moreover, because cultivar mixtures tend to improve yield over monocultures even in absence of pests, our work also examined general improvements in crop productivity by increasing genotypic diversity.

Using field experiments at two research centers and farms of three cooperating growers, we assessed and quantified production benefits of planting fields containing several different varieties of wheat. At some of these sites, we compared insect and pathogen populations and yield among cultivar mixtures and the constituent monocultures. Field days and other extension programming exposed growers to this farming strategy and generate interest for farmers to try this approach for themselves.

Our project introduced a cropping strategy that promises to generate more productive and pest-resistant crop fields. Increasing crop genotypic diversity in agroecosystems is a simple approach that can be adapted to conventional or organic production with cascading ecological effects that should reduce reliance on pesticides and stabilize, or even improve, yields. Importantly, our research goals squarely aligned with the mission of the SARE program, which is “to advance innovations that improve profitability, stewardship and quality of life by investing in ground breaking research and education.”

Performance Target:

Performance Target:

Thirty conventional and organic farmers will adopt mixtures on at least 300 acres (ten acres per grower) of small grains to decrease insect, disease, and weed problems and improve production. By employing mixtures, growers will increase yield at least 5% compared to typical monoculture yield monocultures, meaning yield increases per acre will range between 2.5 bushels per acre for organic growers that typically yield 50 bu/acre to 4 bu/acre for conventional growers who average 80 bu/acre. Conventional growers will also reduce pesticides costs by $20 per acre, which represents nearly 50% savings on a typical preventative pesticide program that costs approximately $37 per acre, further improving profitability.


1. Through winter meetings (in-person and webinars) and summer field days, 1500 farmers will be exposed to the idea of cultivar mixtures and learn of the yield and pest management benefits of mixing cultivars together. These farmers will receive evaluations to capture the state of their knowledge before and after presentations, and their interest in learning more about farming with cultivar mixtures. Growers with interest will be asked for their contact information (November 2012-March 2013 & November 2013-March 2014).

2. One-hundred-fifty growers expressing interest in learning more about cultivar mixtures will be surveyed to characterize their concerns about and expectations for cultivar mixtures (November 2012-March 2013 & November 2013-March 2014).

3. Fifty of these growers will attend field days featuring the benefits of cultivar mixtures. Portions of these sessions will specifically address the concerns raised by growers in the survey (April-July 2013, April-July 2014).

4. Thirty of these growers will trial cultivar mixtures on their farms, and many will share their experiences with other growers via in-person presentations, webinars, podcasts, and newsletters (October 2013 to end of project and beyond).



Click linked name(s) to expand
  • Dr. John Tooker


Materials and methods:

The research portion of our project was a collaborative effort comprising university-based specialists, county-based extension educators, and growers in Pennsylvania and Maryland.  We focused our intensive research efforts on two university-owned research farms, and establish fewer experimental plantings on three grower’s farms.

Our research followed a hub-and-spoke model; therefore, we used the larger of the two university research farms, Penn State’s Russell E Larson Agricultural Research Center (LARC; Rock Springs, PA) to plant a large-scale, replicated field experiment with all 12 of our treatments (six monoculture and six mixtures), and smaller versions of the experiment were conducted at spoke locations.  One smaller-scale replicated version of the large experiment was established at Penn State’s Southeast Agricultural Research & Extension Center (SAREC; Landisville, PA), where we established six of the treatments (three monocultures and three mixtures). In contrast, the three farmer cooperators hosted smaller research (three treatments; two monocultures, one mixture) efforts that were designed to provide a robust, cross-location assessment of pest control and production potential of cultivar mixtures.

This hub-and-spoke model allowed us to focus our more intensive research effort on the two research farms, and then used more realistic settings of growers’ fields as more than just demonstration sites, but rigorous evaluations of mixtures. 

Hypotheses that we tested:

Cultivar mixtures will increase pest resistance of wheat plantings.

Cultivar mixtures will have lower herbivore and pathogen populations than constituent monocultures.

Because of lower pest populations, conventional growers who scout their fields will use fewer pesticides in cultivar mixtures than in monocultures.

Cultivar mixtures will be more productive than monocultures, significantly increase yields and profitability.

To evaluate the field performance of genotypically diverse cultivar mixtures, we established a replicated experiment at LARC with all 12 treatments. These experiments were repeated for two years, using the same varieties and mixtures each year, but a different field. We established a randomized complete block design with twelve treatments to compare six monoculture varieties and six five-line mixtures (all possible five-line combinations of the six varieties). This experimental design was necessary to understand the value of genotypic diversity per se and specific mixtures, while minimizing sampling effects (one variety driving the effect) that could have accompanied a simpler design. Importantly, this approach provided a powerful statistical comparison of low (30 plots) and high diversity (30 plots). We purchased six varieties of wheat seed from Seedway (Emmaus, PA) and planted these six varieties as monocultures. The varieties were SW20 (i.e., Seedway 20), SW27, SW50, SW56, SY1526, and Branson; these latter two varieties were sourced by Seedway from Syngenta Seeds. From these varieties, we mixed equal amounts of each to create six five-line mixtures, each of which lacked one of the varieties. We identified these mixtures by what they lacked: xSW20 (did not have variety SW20, but did have the other five), xSW27, xSW50, xSW56, xSY1526, and xBranson. We planted these monocultures and mixtures on 24 Oct 2012 and 21 Oct 2013 into a two-acre no-till field that had been in soybeans the previous year. To establish the plots, we made four passes per plot of a 6-ft wide Great Plains (Salina, Kansas) no-till, seed drill (Model 3P606NT), providing 7.5-in row spacing and a target seed population of 1.5 million seeds per acre (20 seed per ft in a 7-in row). We randomly assigned treatments to each plot (24 x 34 ft) and replicated each treatment five times for a total of 60 plots. In April (the spring after planting), when wheat plants were at the stage 3 of growth (tillers formed), we fertilized the plots with 60 units of N and applied 0.5 pt/acre of 2,4-D LVE to manage broadleaf weeds.

Monthly during the growing season, we assessed in each plot herbivore and natural-enemy populations using vacuum samplers (leaf blowers modified to be vacuums; Stihl Inc., Virginia Beach, VA) fitted with nylon mesh sleeves. We sorted the insect-community samples in the lab to species where possible and morphospecies otherwise, and assigned them to a trophic guild (e.g. herbivore, predator, parasitoid, omnivore) and to a pest management role (e.g. important pest, predator of pests, parasitoid of pests). We also assessed in plots predation by using sentinel aphids, which we deployed by glueing five live aphids to 2 x 3-in pieces of index cards. These aphids could survive for 24 h on the cards. Using paper clips, we placed two cards mid-canopy in each plot to make them available to resident predator populations for 3 h, assessing survival at after 1.5 and 3 h. At maturity, we harvested a 4-ft swath from each plot to estimate yield. We also sub-sampled grain of each variety and mixture and had it analyzed for the mycotoxins deoxynivalenol and zearalenone.

Because of space limitations at the second research farm, the version of this experiment established at SAREC was similar, but smaller. From the twelve treatments planted at LARC, we randomly selected six for use at SAREC, three monocultures (xSW27, xSW60, x1526) and three five-line mixtures (SW50, SW56, Branson) and planted on 23 Oct 2012 and 20 Oct 2013 them using a 10 ft John Deere no-till, seed drill with the same spacing and seeding rate in five replicates blocks (30 30-×-30-ft plots). In these smaller plots, we sampled insect populations using sweep nets and predator communities using sentinel aphids.

For our efforts at the three commercial farms, we decreased the experimental complexity even further to strike a balance between sampling effort and logistics for these widely spaced farms. We randomly selected one mixture and two of the constituent monocultures to evaluate their performance in farmers fields (four replicates of each monoculture and the mixture; 40 x 150 ft plots, 12 plots at each of three farms). These plots were planted within a week of the dates on which the experiments were planted on the two research farms. In these plots, we assessed insect populations using sweep nets.

For our extension and education efforts, we took advantage of our hub-and-spoke research design to market our efforts and distribute our findings. The two research farms provided the most complete view of our results and effort, but cooperators’ farms provided a larger-scale view of the treatments and allowed access to larger populations of growers keen on evaluating performance.

Field days at the research and cooperating farms started in the project’s first year to get people engaged in the project from beginning to end. Field days and workshops at all sites allowed growers and other agricultural professionals to gain first-hand experience with selecting varieties, mixing cultivars, establishing and managing mixtures, pest and natural-enemy populations in fields, and the general advantages of cultivar mixtures.

County-based educators facilitated extension presentations (on-farm or otherwise, and webinars) during the year, including the important winter extension season. Farmer collaborators provided first-hand experience and their perspectives on the benefits and challenges of cultivar mixtures at these events.

To expose our work to as many growers and other agricultural professionals as possible, we marketed our project to a broad segment of the farming population across Pennsylvania and Maryland in events hosted by Cooperative Extension, Conservation Districts, and professional groups. At these events, growers were exposed to all aspects of cultivar mixtures from variety selection to harvest.

Research results and discussion:

Milestone 1, associated accomplishments: For the 2012-2013 season, we exposed to our work with cultivar mixtures to 1094 growers at 19 separate presentations. In 2013-2014, we presented the concept and some research results to an additional 930 farmers at 16 extension meetings. In 2014-2015, we raised the issue at seven meeting, reaching 735 growers. Evaluations revealed that few growers had been exposed previously to cultivar mixtures. Unfortunately, very few were willing to share their contact information with us. For growers approached in 2012-2013, many liked the idea, but were not willing to commit to the idea until we had some performance/yield data to share. Even though we had promising yield data from 2012-2013, we still did not have good success in the 2013-2014 or 2014-2015 extension season. Growers acknowledged the potential benefits but preferred to focus on potential logistical challenges associated with buying different varieties of seeds and planting a mixture, seeing these issues as unnecessary complications to their operations. Moreover, the majority of growers we reached were conventional (i.e., not organic) farmers and they were concerned that mixtures could complicate the timing of fungicides if they were necessary. Particularly in moist years, fungicides are often applied at flowering by conventional farmers, and growers felt having more than one wheat variety would widen the flowering window too much, making precisely targeting fungicide applications difficult.

Milestone 2, associated accomplishments: We have been unable to survey growers due to a lack of contact information—too few growers expressed an interested in cultivar mixtures at our meetings. Nevertheless at our project-specific field days, we engaged growers in frank discussions of the benefits and limitations of growing cultivar mixtures. In 2013-2014 and 2014-2015, we addressed these limitations head-on at field days, but still encountered wide spread hesitancy about growing more than one varieties of wheat together—again the primary concern was fungal diseases at heading, specifically wheat scab, and the logistical challenge of spraying field that contain wheat varieties that flower at different times. We were unable to gather enough contact information to survey.

Milestone 3, associated accomplishments: We held field walks at two of our cooperating farms, attracting 63 growers to learn about cultivar mixtures first hand. In 2013, We also had field days at our research farms, attracting an additional 260 growers that heard about cultivar mixtures. In 2014, we had two additional field days with 105 attendees. In these presentations, we addressed growers concerns and addressed them with in-depth discussions.

Milestone 4, associated accomplishments: We succeeded in getting six growers to trial mixtures. Despite the our efforts to reach as many wheat growers as possible, too few growers were interested in committing to growing cultivar mixtures on their farms. This is clearly insufficient. Growers just seem reluctant to complicate their farming practices, even though we can point to production and pest management benefits.

Participation Summary


Educational approach:

Research and extension publications are still in development.

Additional Project Outcomes

Project outcomes:

Impacts of Results/Outcomes

We had no research plots in the ground in 2014-2015, but we wrapped up our research efforts from 2013-2014, when we had plots on three collaborating farms (Mertztown and Holtwood, PA, and Clear Spring, MD) and two Penn State research farms (Centre and Lancaster Counties, PA). In 2015, we finished sorting the previously collected insect samples. We did not analyze the results of our research farm plots because they were small and we expected that many insect species moved freely among them, homogenizing populations. From our statistical analyses of our on-farm samples, consistent with our hypothesis, we learned that genotypic diversity hosted significantly fewer aphids than the monoculture plots. In fact, our monoculture plots hosted twice as many aphids as the diverse plots. It is important to note, however, that our aphid populations were low in all the plots and that sampling aphids with sweep nets is not the best way to assess their populations.  Nevertheless, our aphid data firmly supported our hypothesis that genotypic diversity can suppress aphid populations.

In contrast to aphids, cereal leaf beetle adult and nymphs were equally abundant in mixtures and monocultures in our collaborators fields. This outcome was a bit disappointing because cereal leaf beetles were the most abundant pest species that colonized our plots, and it would be great if their populations were influenced by the diversity we deployed. There is some previous evidence that cereal leaf beetles are sensitive to genotypic diversity, but the main trait to which they are sensitive is leaf pubescence, and our varieties were somewhat uniform for this trait, so the outcome is understandable (we selected our varieties with aphids in mind) and there remains potential to create a mixture that is capable of suppressing populations of cereal leaf beetle.

We hypothesized that natural enemies would be more abundant in the diverse plots, but most species were equally abundant in diverse and monoculture plots. Contrary to our expectations, lady beetles were among those groups equally abundant in diverse and monoculture plots. In our previous research, lady beetles were more abundant in genotypically diverse soybean fields, and in lab experiments lady beetles were more strongly attracted to pots planted with mixtures of wheat varieties than monocultures. Perhaps a more intensive sampling effort would have revealed lady beetles to be more common in diverse plots, or perhaps our field plots were still too small even though they were 1/8 acre, and beetles moved readily between them. Interestingly, parasitic wasps, important natural enemies of many herbivorous insect species, showed a pattern that was opposite of our expectations; monocultures hosted significantly more parasitic wasps than the genotypically plots. This outcome is understandable given that populations of many parasitoid species are density dependent, meaning that they are more abundant where their hosts are more abundant. If the parasitoids we captured were natural enemies of aphids, then perhaps the parasitoid population is just tracking their aphid hosts.

After harvest we also sampled our grain from the plots on the research farm analyzed for alfatoxins, which can be produced by fungal infections of the grain head (i.e., Fusarium head blight, aka head scab). We did not sample the on-farm plots for fungal infections because the growers treated the fields with fungicides to prevent wheat scab infections. Fortunately, none of our 130 samples yield any detectable levels of alfatoxins. While it is frustrating that the relevant pathogen did not infest our plots, it is encouraging that mixing cultivars did not leave the plots more vulnerable to disease as some farmers fear. We will use these data to continue to promote this approach to farming.

As reported in previous years, we found that yields of cultivar mixtures can compete with the yields of the best yielding varieties. In 2012-2013, we found that mixtures of wheat grown in eighth-acre plots on commercial farms improved yield by 6% over monocultures, whereas yields of mixtures and monocultures in small plots on our research farms were equivalent. In 2013-2014, yields of mixtures and monocultures on the three commercial farms were equivalent as were yields of the two treatments from our research farms. These results strongly suggest that mixtures do not impose a yield drag on wheat production in our region. Given the potential pest management benefits, we see advantages that can be provided by mixtures of wheat even in the absence of a yield advantage, which we detect in the first year of our study but not the second.

Economic Analysis

Not applicable

Farmer Adoption

We are aware of six farmers who planted genotypic mixtures following the advice generated by our project. We had difficulty generating excitement among farmers for our approach.

Assessment of Project Approach and Areas of Further Study:

Areas needing additional study

Limitations to farmer adoption appears to be the key detail requiring additional study.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.