Home | Past Issues

Issue No. 22, Article 1/September 2, 2004

Transgenic Corn Rootworm Hybrid Stumbles in Urbana Experiment; Some Producers Also Report Severe Lodging with YieldGard Rootworm Hybrids in Commercial Fields

On February 25, 2003, Monsanto Company received a conditional registration under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) section 3(c)(7)(C) from the U.S. Environmental Protection Agency for the sale of YieldGard Rootworm-protected corn (MON 863, Cry3Bb1). The U.S. Environmental Protection Agency judged that the use of transgenic hybrids for corn rootworm larval control would serve the public interest by replacing or reducing the use of soil insecticides (particularly organophosphates) deemed to be of greater risk (environmental, human health, and safety). In addition, the use of transgenic hybrids for corn rootworm control was judged to have "comparable or greater efficacy than current chemical alternatives." As the 2004 growing season revealed, this recombinant technology is still relatively new with respect to corn rootworm control, and we have much to learn.

As we stated last year in an article in issue no. 1 of the Bulletin (March 20, 2003), producers should anticipate finding numerous corn rootworm adults in their fields of YieldGard Rootworm corn. These adults are immigrants from other cornfields as well as survivors from YieldGard Rootworm corn. Different from the poor survival of European corn borers on Bt hybrids used to control them (hybrids resulting from high-dose transgenic events), some rootworms survive after feeding on the roots of current transgenic hybrids for control of corn rootworm larvae. Why? The MON 863 event is a non-high-dose product. In August 2002, the EPA's FIFRA Scientific Advisory Panel believed it was not necessary to distinguish between low- and moderate-dose events. Regardless of the precise characterization of the YieldGard Rootworm event (low dose vs. moderate dose), it remains certain that corn rootworm survivorship is evident. In 2001, we placed adult emergence cages over transgenic plants (MON 863) that were planted on May 10. The cages were checked for adult emergence every 2 days. A total of 96 emergence cages were used in this replicated split-plot experiment. Based on the number of adults emerging into these cages, we projected that 2,450 male western corn rootworm adults would emerge per acre from the MON 863-event corn. The number of females emerging per acre from the transgenic corn was 25,320. Overall, we projected that 27,770 adults would have emerged from 1 acre of transgenic corn (MON 863, Cry3Bb1), and 91% of these were females. Emergence of males and females also from the transgenic corn rootworm hybrid was delayed when compared with emergence from the non-transgenic isoline. Delayed emergence and survivorship data such as these have been reported by other entomologists from land-grant institutions. Although MON 863 is a non-high-dose event, root protection afforded by this technology in university research plots had been exceptional.

Adult corn rootworm emergence cage..

In 2003, we established three corn rootworm product efficacy trials in trap-crop settings located near DeKalb, Monmouth, and Urbana. Average root-injury ratings (1-to-6 scale) in our untreated check plots (DeKalb 60-15 hybrid) were 4.06 (one node destroyed), 5.75 (2-3/4 nodes destroyed), and 5.45 (2-1/2 nodes destroyed), respectively. Planting dates in 2003 for DeKalb, Monmouth, and Urbana were May 28, May 16, and May 13, respectively. In 2003, corn rootworm larvae began hatching by May 29 across much of central Illinois. Despite the intensity of larval feeding, the DeKalb YieldGard rootworm hybrid offered excellent root protection at each experimental site: at DeKalb, the average root rating was 1.45; at Monmouth, 2.05; and at Urbana, 1.35.

In 2004, we again established corn rootworm product efficacy plots in our three standard experiments in trap-crop settings. Our trap crop consists of late-planted, mixed-hybrid corn interplanted with pumpkins. Planting dates for DeKalb, Monmouth, and Urbana were April 28, April 27, and April 19, respectively. These planting dates were considerably earlier in 2004 than the corresponding planting dates in 2003: 1 month earlier in DeKalb, 19 days earlier in Monmouth, and 24 days earlier in Urbana. Corn rootworm larval hatch in central Illinois was estimated to be about 1 week to 10 days earlier in 2004 than in 2003. In 2004, we planted a Golden Harvest (H-8588 RW) YieldGard Rootworm hybrid and its isoline (Golden Harvest H-8799). Root injury (1-to-6 scale) in the untreated checks was severe at each location and averaged 5.0 (2 nodes destroyed), 5.75 (2 3/4 nodes destroyed), and 5.8 (nearly 3 nodes destroyed) at DeKalb, Monmouth, and Urbana, respectively. Root evaluations were conducted for each of these locations on July 21, July 15, and July 10, respectively. Average root ratings for the Golden Harvest YieldGard Rootworm hybrid at DeKalb, Monmouth, and Urbana were 2.35, 1.80, and 3.15 (pruning occurred, considerably less than 1 node), respectively. The level of root injury to the YieldGard Rootworm hybrid in the Urbana experiment was greater than anticipated. Plants from our trials near DeKalb, Monmouth, and Urbana were checked (by Monsanto personnel) for expression of the Cry3Bb1 protein, and the results were positive (i.e., the roots were expressing the protein).

On July 13, a severe storm swept across central, north-central, and east-central counties of Illinois. High winds and torrential rains characterized this impressive storm. Within a few days, we began to receive telephone and e-mail inquiries concerning severely lodged corn, including concerns from some producers who had planted YieldGard Rootworm hybrids. We visited some of these producers' fields. In some cases, root injury was greater than expected. In others, the level of root pruning was minimal in theYieldGard Rootworm fields, yet the fields were severely lodged and harvestability was seriously compromised. Not long after these observations from commercial cornfields became more common, Ron Estes (insect management and insecticide evaluation program manager, Department of Crop Sciences, University of Illinois) noted that the YieldGard Rootworm hybrid (Golden Harvest H-8588RW) in our Urbana experiment was severely lodged. At that point, we elected to take more root-injury and lodging evaluations from this treatment at all three of our experiments.

Lodged YieldGard Rootworm plants, Urbana experiment, August 2004.

During the first week of August, we extracted 40 roots (10 per replication) from the YieldGard Rootworm plots in the Urbana experiment and rated them for larval damage (1-to-6 root-rating scale). Slightly more than 3 weeks had elapsed since our original ratings on July 10. Average root ratings per replication were 3.6, 3.4, 3.6, and 3.8 (overall average = 3.6). The range in root injury was 3 to 4 (60% of the roots were rated 4.0). None of the roots was assigned a root rating of 1 or 2 during the root evaluations in August. To quantify the level of pruning on these roots more precisely, we also used the 0-to-3 rating scale for node injury developed at Iowa State University. Root-injury ratings per replication were 1.43 (nearly 1-1/2 nodes destroyed), 1.08 (1 node destroyed), 1.64 (slightly more than 1-1/2 nodes destroyed), and 1.24 (slightly more than 1 node destroyed). The node-injury rating for YieldGard Rootworm corn in the Urbana experiment was 1.35 (1-1/3 nodes destroyed). The level of injury we observed in August was much greater than what we observed in July.

Injured YieldGard Rootworm roots, Urbana experiment, August 2004.

In addition to the root evaluations, we measured the level of lodging that occurred in the YieldGard Rootworm corn. The percentage of plants lodged in each of the replications (four-row plots) was 40%, 79%, 61%, and 39%.

Because of our experience with YieldGard Rootworm corn in our Urbana experiment, we took additional root ratings (August 12) and lodging counts in our DeKalb and Monmouth experiments. Average root ratings (1-to-6 scale) in the DeKalb experiment (by replication) were 2.6, 2.3, 2.3, and 2.2 (overall average = 2.35). The range in root ratings in our DeKalb experiment was 2 to 5. None of the YieldGard Rootworm corn plants was lodged in the DeKalb study. Average root ratings (1-to-6 scale) in the Monmouth experiment (by replication) were 2.5, 2, 2.7, and 2.2 (overall average = 2.35). The range in root ratings was 2 to 4. The percentage of plants lodged in each replication (four-row plots) was 29%, 3%, 9%, and 0%.

Many questions have arisen from producers regarding performance issues relative to some YieldGard Rootworm hybrids and severe lodging that has occurred in certain commercial cornfields. The observations from our experiments will hopefully begin to help us find answers to many of these perplexing questions. Most notably, why was the level of injury in our Urbana experiment much greater than we had anticipated? The following set of questions and answers is designed to help improve our understanding of this issue.

Has resistance toYieldGard Rootworm hybrids already occurred?

We don't believe so. Keep in mind that YieldGard Rootworm hybrids were released commercially for the first time in 2003. It does not seem probable that rootworm resistance to the Bt protein has occurred this quickly. The FIFRA Scientific Advisory Panel (August 2002) suggested that the likelihood of resistance development within the first 3 years of commercialization is unlikely, regardless of refuge size. In April 2003, the U.S. EPA issued a document titled "Event Mon863 Bt Cry3Bb1 Corn Biopesticide Registration Action Document". The authors clearly point out many of the assumptions underlying model predictions for potential resistance development. If these models have underestimated the actual initial resistance allele frequency, then resistance development could proceed more rapidly. We simply don't have the required data to answer this question.

What steps should be taken if unexpected damage or suspected resistance does occur?

The FIFRA Scientific Advisory Panel (August 2002) specifically established four steps a registrant should follow to confirm whether resistance may have developed:

  1. "Request the grower check planting records."
  2. "Rule out damage from nontarget insects, weather, or other environmental factors."
  3. "Conduct tests to verify MON 863 was planted and that the correct percentage of plants are expressing."
  4. "If plants are MON 863 and damage approaching a 0.5 (node-injury scale) is found on any expressing plant, evaluate roots from the corresponding refuge."

The following paragraph from the April 2003 EPA document (page IID 11) is of particular interest with respect to resistance confirmation.

"Resistance should be confirmed by a standard diet bioassay or evaluation of root node injury. An insect diet bioassay with the Cry3Bb1 protein that results in a LC50 that exceeds the upper limit of the 95% confidence interval of the LC50 established from baseline measurements of susceptible populations could be used to confirm resistance. Alternatively, resistance may be confirmed when one or more root nodes of at least 50% of Cry3Bb1 plants grown in the laboratory are destroyed. [Recall that 60% of the roots evaluated from the YieldGard rootworm treatment in Urbana rated 4.0 (1 node destroyed)]. A discriminating concentration bioassay may also be used to confirm resistance; however, this method may take a long time to develop. The August 2002 SAP also recommended investigating the potential of using samples of populations surviving on Bt corn or an evaluation of larval root tunneling to confirm resistance."

Resistance confirmation must be reported to the U.S. EPA within 30 days.

What other explanations may account for the loss of performance with the YieldGard Rootworm hybrid in the Urbana experiment?

In 2003 (root rating in the untreated check = 5.45) and 2004 (root rating in the untreated check = 5.8), we had severe corn rootworm larval injury in our Urbana trials. So the explanation for the reduced efficacy of the YieldGard Rootworm hybrid in 2004 cannot be blamed solely on severe corn rootworm larval pressure. However, considerable differences in planting dates occurred between the two growing seasons. In 2003, the Urbana experiment was planted on May 13 (late by today's standards). In 2004, the Urbana study was planted on April 19, more than 3 weeks earlier than the 2003 planting date. We, as well as other researchers, have determined that development and emergence of rootworms are delayed by the MON 863 event. Is root protection compromised significantly when certain factors converge, such as (1) early planting (early to mid-April), (2) delayed corn rootworm development resulting from larvae feeding on MON 863 hybrids, and (3) intense larval pressure (root ratings in the untreated check of two or more nodes of roots destroyed)? Might these factors explain the compromised efficacy in our Urbana experiment as well as in some producers' fields?

Other questions to ponder:

Does the expression of the Cry3Bb1 protein diminish as the season progresses, compromising root protection in late July and early August?

Are there critical differences in expression of the Cry3Bb1 protein across hybrids?

Are some hybrids with less than satisfactory root characteristics poor candidates for the YieldGard Rootworm technology?

As indicated at the outset of this article, we still have much to learn about transgenic hybrids and their role in an overall corn rootworm management program. One thing seems certainthere continues to be no silver bullet that can be directed at this impressive and resilient insect pest of corn.--Mike Gray and Kevin Steffey

Kevin Steffey
Mike Gray

Click here for a print-friendly version of this article

Return to table of contents