Studies were conducted across nine U.S. states, over 5 yr, to characterize the efficacy of transgenic corn (Zea mays L.) hybrids producing insecticidal proteins derived from Bacillus thuringiensis (Bt) for control of western corn rootworm, Diabrotica virgifera virgifera Le Conte. Hybrids tested had the same genetic background, contained one of two single events (DAS-59122–7 expressing Cry34Ab1/Cry35Ab1 or MON 88017 expressing Cry3Bb1) or a pyramid consisting of both rootworm-active events (SmartStax traits) and were compared with a non-Bt near isoline. Frequency analyses of root feeding data showed that hybrids containing both events sustained less root damage (0–3 node injury scale) than hybrids containing either event alone. The levels of root protection provided by MON 88017 and DAS-59122-7 were not different from each other. Efficacy was also evaluated based on consistency of protection, based on the proportion of plants with root ratings of either ≤0.25 or <1.00 on the node injury scale. The combination of two modes of action in SmartStax provided greater product consistency over a single mode of action at the 0.25 level and all hybrids producing Bt proteins provided equally high consistency at the 1.00 level. Overall these data show single and multiple mode of action hybrids provided high, consistent protection over the past 5 yr across the trial geography; however, pyramiding the rootworm Bt events provided greater and more consistent root protection. These findings also support that pyramided traits like SmartStax (Cry3Bb1 + Cry34Ab1/Cry35Ab1) remain a viable strategy for delaying resistance to either trait.
The western corn rootworm, Diabrotica virgifera virgifera LeConte, is a significant pest of corn in the United States. Most yield loss from western corn rootworm results from larval feeding on corn roots and when damage is severe results in primary roots failing to grow. This damage is referred to as pruning and inhibits the plant's ability to take up water and nutrients from the soil leading to decreased photosynthetic activity and lower yields (Spike and Tollfeson 1989, Sutter and Gustin 1990, Spike and Tollefson 1991, Dunn and Frommelt 1998). Root pruning can lead to lodging of the corn stalks that can complicate harvest further reducing yield and increasing harvest expense (Sutter and Gustin 1990).
The overwintering stage of western corn rootworm is the egg stage. Eggs hatch in the early spring and larvae start to feed on young corn roots (Chiang 1973). To make management decisions before planting based an action threshold, attempts have been made to scout rootworm adults or eggs in the previous year to predict larval injury in the coming year. The egg stage is very labor intensive to sample and studies have shown egg sampling to be a poor predictor of larval feeding damage (Shaw et al. 1976, Ruesink and Shaw 1983, Hein et al. 1985, Tollefson 1990). Tollefson (1990) found several adult beetle sampling techniques were correlated with larval injury to roots; but the low precision (or high amount of labor to improve precision) made these impractical tools for corn growers. Therefore, to avoid root injury, corn growers usually make management decision before planting with little or no knowledge of the potential for western corn rootworm damage in the field. A common management strategy has been the application of a soil insecticide at planting. Some of the first insecticides used were chlorinated hydrocarbons such as aldrin and heptachlor but resistance developed rapidly to these insecticides (Ball and Weekman 1962). More recently there has been a shift to organophosphate, carbamate, and pyrethroid insecticides. However, the efficacy of soil applied insecticides is dependent on several factors including planting date (May 1980), plating-time conditions (Bergman et al. 1991), and soil moisture (Sutter et al. 1989, Gray et al. 1992). As an example, lack of soil moisture can prevent proper contact between the larvae and the insecticide (Sutter et al. 1990) or high winds at the time of planting can result in poor placement of insecticide granules (Bergman et al. 1991); both situations leading to decreased efficacy of soil applied insecticides. Another common management strategy for western corn rootworm is annual rotation of corn and soybean to disrupt the rootworm's life cycle. Traditionally, western corn rootworm laid eggs almost exclusively in corn (Shaw et al. 1978, Levine and Oloumi–Sadeghi 1991). However, resistance of western corn rootworm to crop rotation has developed by a shift in a portion of the population to individuals that do not have such a strong fidelity to corn and will lay eggs in other crops such as soybeans and therefore survive in the following year when the rotation returns to corn (Levine and Oloumi–Sadeghi 1996, Levine et al. 2002).
Transgenic plants that express insecticidal Bt proteins are a relatively new management strategy that overcome some of the limitations associated with managing field corn insects with synthetic insecticides (Rice 2004). Corn hybrids expressing Bt proteins active against corn rootworm larvae were first developed and available to growers in 2003 with the Environmental Protection Agency (EPA) registration and U.S. Department of Agriculture (USDA) deregulation of event MON 863 expressing Cry3Bb1 (YieldGard Rootworm, Monsanto Co., St. Louis, MO; U.S. EPA 2003, Vaughn et al. 2005) and the subsequent introduction of event MON 88017 expressing Cry3Bb1 protein and the 5-enolpyruvylshikimate-3-phosphate synthase protein from Agrobacterium spp. strain CP4 (CP4 EPSPS), which confers tolerance to glyphosate herbicides (U.S. EPA 2010). In 2005, Dow AgroSciences obtained USDA deregulation and EPA registration for event DAS-59122-7 (that was built in collaboration with Pioneer Hi-Bred International, now known as DuPont Pioneer) expressing Cry34Ab1 and Cry35Ab1, which act together as a binary insecticidal protein (Herculex RW Rootworm Protection, Dow AgroSciences LLC, Indianapolis, IN; Herman et al. 2002, Ellis et al. 2002, U.S. EPA 2005).
Dow AgroSciences and Monsanto Company each combined these events with other events that express Cry proteins providing protection against leaf, stalk, and ear feeding insects. Event MON 89034, YieldGard VT Pro, produces Cry1A.105 and Cry2Ab2 (Monsanto Co.; Drury et al. 2008) and event TC1507, Herculex I Insect Protection, produces Cry1 F (Dow AgroSciences LLC; U.S. EPA 2004, Dow AgroSciences 2007). The stacked products from Monsanto Company and Dow AgroSciences are YieldGard VT Triple Pro (events MON 89034 × MON 88017) and Herculex Xtra (events TC1507 × DAS-59122-7), respectively.
Events MON 89034 and TC1507 and events MON 88017 and DAS59122-7 have been combined through conventional breeding to create SmartStax (Dow AgroSciences LLC, and Monsanto Co.), a trait pyramid (plants with simultaneous expression of two or more insecticidal proteins that target the same insect) (Gould 1986, Sachs et al. 1996). Field corn hybrids containing these six Cry proteins (Cry1A.105, Cry2Ab2, Cry1 F, Cry3Bb1, and Cry34Ab1/Cry35Ab1) broaden the spectrum of activity and aid in resistance management of target pest species (McGaughey and Whalon 1992; Tabashnik 1994; Gould 1998; Storer et al. 2006, 2010). Field corn hybrids containing events MON 89034, TC1507, MON 88017, and DAS-59122-7 also contain transgenes for tolerance to glyphosate and glufosinate ammonium herbicides, providing another tool for managing glyphosate resistant weed species (Green and Owen 2010).
The objective of the following series of trials established in the United States was to compare the efficacy of field corn containing transgenic Cry3Bb1 and Cry34Ab1/Cry35Ab1, alone and in combination, for control of western corn rootworm. These data are essential in understanding product performance and managing potential resistance to Bt proteins used for corn rootworm management.
Materials and Methods
Studies evaluating root injury from western corn rootworm to Bt and non-Bt corn hybrids were conducted from 2007 to 2011 for a total of 39 trials conducted over nine U.S. states (Table 1). Depending on the year, studies compared Bt events, both alone and in combination, and a non-Bt corn hybrid. To minimize the effects of corn genotypes, near-isogenic corn hybrids were evaluated. Hybrids tested included those expressing: 1) Cry34Ab1/Cry35Ab1 (event DAS-59122-7, Herculex Xtra Insect Protection, Dow AgroSciences LLC), 2) Cry3Bb1 (event MON 88017, YieldGard VT Triple Pro, Monsanto Co.), 3) a combination of Cry34Ab1/Cry35Ab1 and Cry3Bb1 (events DAS-59122-7 and MON 88017, SmartStax, Dow AgroSciences LLC, and Monsanto Co.), and 4) a non-Bt hybrid (containing event NK603 conferring only glyphosate herbicide tolerance, Monsanto Co.). The corn hybrids in treatments one and three also contained two herbicide tolerance traits (event NK603 conferring glyphosate tolerance and event TC1507 conferring glufosinate-ammonium herbicide tolerance) and treatment two hybrids only contained the glyphosate herbicide tolerance trait. Seed used for testing was treated only with a commercial fungicide and no preventive treatments of soil or foliar applied insecticides were applied across the test areas.
Treatments, no. of replicates, infestation type, and node-injury scale (NIS) rating in the non-Bt treatment for each trial location from which efficacy against the western corn rootworm was evaluated, 2007–2011
 |
|
Treatments, no. of replicates, infestation type, and node-injury scale (NIS) rating in the non-Bt treatment for each trial location from which efficacy against the western corn rootworm was evaluated, 2007–2011
|
|
All trials were planted on trap crop ground that consisted of late planted corn or pumpkins planted in the previous year to attract ovipositing female western corn rootworm. Some locations were also artificially infested with western corn rootworm eggs at a rate of 1,000 eggs per row foot at the V2–V5 stage of corn during the trial year (Sutter and Branson 1980, Ritchie et al. 1993). In 2007–2009, eggs were obtained from beetles collected in Wisconsin, Illinois, and Indiana and reared in the Dow AgroSciences Insectary in Huxley, IA. Dependent on location, eggs for trials in 2010 and 2011 were obtained from nondiapausing colonies at one of three insectary facilities: 1) USDA Agricultural Research Service, Brookings, SD; 2) Crop Characteristics Inc., Farmington, MN; or 3) French Agricultural Research, Lamberton, MN. Utilizing trap crop ground and artificial infestations helped ensure that larval pressure was high across locations and years. Pressure in the non-Bt treatment (RR-null) ranged from 0.57 to 2.98 (Table 1) on the ISU node-injury scale of 0–3 (Oleson et al. 2005). Treatments in the field were planted in a randomized complete block design with 2–6 replications dependent on trap crop space available at each location (Table 1). In 2007, because of limited seed because the trait combinations were early in development, plots were one row by 2 m in length. Plot size across locations in 2008–2011 was four rows (76.2 cm row centers) by 6 m in length. All studies were maintained using agronomic practices for optimal productivity.
In mid- to late-July at approximately the V16–R1 corn growth stage, 5–10 randomly selected roots were dug from each plot (Ritchie et al. 1993). Each root mass was shaken vigorously upon removal from the soil and rinsed with a garden hose to remove the bulk of the soil around the roots. A pressure washer was then used to remove any remaining soil so rootworm feeding could be seen clearly. Each root was then assigned a rating from 0 to 3 using the ISU node-injury scale (Oleson et al. 2005). A rating of 0.00 indicated no feeding damage, a rating of 1.00 indicated one node of roots or the equivalent of an entire node was eaten (pruned to within 3.8 cm of the stalk), a rating of 2.00 indicated two nodes pruned, and a rating of 3.00 indicated three or more nodes were pruned (the highest rating that can be given). Damage in-between complete nodes pruned was noted as the percentage of the node missing, for example, 1.50 = one and one-half nodes pruned.
Statistical Analysis.
Root rating data were not normally distributed and highly skewed toward the lower end of the 0–3 scale. Transformations were not successful in meeting the model assumptions for a linear mixed model analysis. As an alternative, the multinomial characteristics of the 0–3 scale data in conjunction with biological reasoning related to potential for yield loss was used to categorize the individual root rating data into levels of root protection: (a) 0–0.25, (b) >0.25 to <0.50, (c) 0.50 to <1, (d) 1 to <2, and (e) 2–3. These categories were chosen based on the likelihood that certain levels of root injury may cause varying degrees of yield loss and lodging (Oleson et al. 2005). A 0.25 injury level is a conservative economic injury level (EIL) for western corn rootworm. The amount of yield loss associated with a particular root injury level depends greatly on the environmental conditions experienced by the crop (Oleson et al. 2005). Under high stress conditions such as low moisture, some yield loss can occur with node injury scores as low as 0.25; however, if injury is kept at or below 0.25 yield loss is unlikely under most field conditions (Oleson et al. 2005). Consequently, the first category, 0–0.25, represents injury that is unlikely to result in any significant yield loss or lodging. Injury from >0.25 to <0.50 could cause some yield loss under stressed field conditions but lodging is unlikely while injury from 0.50 to <1 is likely to cause yield loss unless moisture is ample but again, no harvest issues with lodged plants are expected. The likelihood of corn lodging, leading to difficulty with harvest tends to increase as node injury scores reach and exceed 1.00 (Oleson et al. 2005). Therefore, injury from 1 to <2 is likely to cause yield loss even under nonstress conditions and lodging becomes increasingly likely. Injury from 2 to 3 is likely to result in both high levels of yield loss and in almost all cases, severe lodging. To determine which treatment(s) provided the best root protection, the frequency of occurrence of the root rating data in these categories was analyzed using a generalized linear mixed model procedure for a multinomial response distribution with a cumulative logit link function (SAS GLIMMIX, SAS 2008). Odds ratios were compared among treatments to distinguish treatment differences and determine which treatments provided better root protection by having a higher frequency of lower root ratings.
In addition to the frequency analysis of the root rating data, another way to compare treatments is to calculate how consistently a treatment provides a certain level of protection. Consistency is defined as the percentage of times a particular treatment provides root protection at a set level of injury. Two consistency levels were chosen, 1) protection at or below the 0.25 level and 2) protection below the 1.00 level to determine which treatment(s) would provide the highest protection against any yield loss and which treatment(s) would protect against lodging, respectively (Oleson et al. 2005). Therefore, each node-injury score was categorized into two different binomial classifications, 1) 0–0.25 or >0.25 and 2) 0 to <1.00 or >1.00. These binomial data were analyzed using generalized linear mixed procedure for a binomial response distribution with a logit link function (SAS GLIMMIX, SAS 2008). A Tukey–Kramer mean separation test was performed to distinguish treatment differences. Location, replicate within location and treatment by location factors were considered random in all models. All of the analyses were performed using an alpha level of 0.05.
Results and Discussion
The analysis of the frequency of root injury scores models the probability of a reference treatment having higher or lower ordered values in the response profile table (GLMMIX procedure, SAS 2008; Table 2). With the field corn hybrid containing both Cry34Ab1/Cry35Ab1 and Cry3Bb1 (SmartStax traits) set as the reference standard for treatment comparisons, all other treatments had higher ordered values on the response profile table (Table 3; F = 90.28; df = 3, 85; P < 0.0001). In other words, the frequency of higher root injury ratings was greater for all treatments compared with SmartStax (i.e., SmartStax provided superior root protection compared with hybrids with just one Cry protein or no Cry proteins). With the treatment containing no Cry proteins set as the reference standard, all treatments had lower ordered values in the response profile or Cry34Ab1/Cry35Ab1, Cry3Bb1, and Cry34Ab1/Cry35Abl + Cry3Bb1 all provided superior root protection compared with the hybrid with no Cry proteins (Table 3; F = 90.30; df = 3, 85; P < 0.0001). With the treatment containing Cry3Bb1 set as the reference standard, Cry34Ab1/Cry35Ab1 + Cry3Bb1 provided superior root protection but the protection provided by hybrids expressing only Cry34Ab1/Cry35Ab1 was not significantly different from those expressing Cry3Bb1 only (Table 3; F = 90.29; df = 3, 85; P < 0.0001). The combination of two modes of action for the management of western corn rootworm feeding resulted in superior root protection. Based on these trial locations, there was no difference in the level of protection provided by the two single-mode of action treatments (Cry34Ab1/Cry35Ab1 and Cry3Bb1).
Response profile for node-injury score data across locations and years, comparing the percentages of roots for each treatment that fall into a particular root injury range
 |
|
Response profile for node-injury score data across locations and years, comparing the percentages of roots for each treatment that fall into a particular root injury range
|
|
Odds ratio estimates and 95% CL for node-injury score data across locations and years comparing the frequencies in the reference treatment to the other treatments in the trial (GLMMIX procedure, SAS 2008)
 |
|
Odds ratio estimates and 95% CL for node-injury score data across locations and years comparing the frequencies in the reference treatment to the other treatments in the trial (GLMMIX procedure, SAS 2008)
|
|
The 0.25 level of product consistency gives an indication of how well the Bt traits or trait combinations protect against yield loss while the 1.00 level of product consistency indicates how well the treatments protect against lodging. At the 0.25 level, all the Bt-containing hybrids (Cry34Ab1/Cry35Abl, Cry3Bb1, and the combination of Cry34Ab1/Cry35Ab1 + Cry3Bb1) provided a high level of consistency (99.8, 98.5, and 97.5%, respectively; Table 4) and therefore provided high levels of protection against yield loss. SmartStax traits provided greater consistency than hybrids with Cry34Ab1/Cry35Ab1 alone or hybrids with no Cry protein but was not statistically different from those expressing Cry3Bb1 (F = 28.39; df = 3, 66; P < 0.0001). Both Cry3Bb1 and Cry34Ab1/Cry35Ab1 provided greater consistency than the hybrid with no Cry protein but were not significantly different from each other. At the 1.00 level, all Bt-containing hybrids provided a high level of consistency (Table 4), all were significantly more consistent than the hybrid containing no Cry proteins and none was significantly different from any other (F = 35.93; df = 3, 117; P < 0.0001). Therefore, all Bt-containing hybrids provided excellent protection against the potential for lodging. Although consistency for all Bt-containing hybrids was high at both levels, the combination of traits with different modes of action provided measurable improvements in root protection, which is consistent with higher levels of corn rootworm control.
Consistency (±SEM) of western corn rootworm root protection at or below a 0.25 node-injury rating (0.25 consistency) and below a 1.00 node-injury scale rating (1.00 consistency), across locations, 2007–2011
 |
|
Consistency (±SEM) of western corn rootworm root protection at or below a 0.25 node-injury rating (0.25 consistency) and below a 1.00 node-injury scale rating (1.00 consistency), across locations, 2007–2011
|
|
These trials were conducted across nine U.S. states over a 5 yr period and as such represent a wide range of environmental growing conditions, soil types, and general agronomic management practices. These data help characterize the efficacy of multiple Bt events pyramided in a single corn hybrid (event DAS-59122-7 + event MON88017, SmartStax) as well as single Bt event hybrids (event DAS-59122-7 and event MON88017) for control of western corn rootworm. These data show that single and multiple mode(s) of action hybrids have provided high levels of consistent root protection over 5 yr across the trial geography, and pyramiding the rootworm Bt events provided better and more consistent root protection than a single Bt event. Recently, western corn rootworms with reduced susceptibility to Cry3Bb1 corn have been identified in Iowa (Gassmann et al. 2011). Gassmann et al. (2011) confirmed that there was no cross-resistance in this population to Cry34Ab1/Cry35Ab1. The data from our trials shows high efficacy of both single traits and the pyramid across years and multiple locations. These findings also support that pyramided traits like SmartStax (Cry3Bb1 + Cry34Ab1/Cry35Ab1) remain the best strategy for delaying resistance to either trait because the multiple modes of action will provide redundant control of western corn rootworm larvae that survive one of the single protein toxins. Even in locations where a portion of the western corn rootworm population may have resistance to one of the Bt proteins, hybrids with pyramided traits may still be the best strategy for delaying resistance especially compared with only using a single toxin (Gassmann et al. 2011). It is also important to ensure that refuge is properly implemented meaning an appropriately matched hybrid is planted, the correct amount is planted, and it is planted in the right place at the right time. To help ensure this is the case, the concept of refuge-in-the-bag has been pursued and registered by Monsanto as RIB Complete and Dow AgroSciences as Refuge Advanced powered by SmartStax. These products provide convenient, simple solutions that ensure refuge is planted correctly and help delay the development of resistance. In the future it will be important to continue to monitor the western corn rootworm efficacy of both single and multi-trait hybrids to manage resistance and ensure the longevity of these management tools.
Acknowledgments
The authors gratefully acknowledge the efforts of our many colleagues at Dow AgroSciences, Monsanto Company, Colorado State University, Cornell University, Iowa State University, Ohio State University, Purdue University, South Dakota State University, University of Illinois, and University of Nebraska, and private researchers who contributed to this project. We thank Jon Babcock, Randy Smith, Rodney Schultz, Scott Hutchins, Simran Trana, Andrea Borucki, Joe Miller, and Kari Lukasik of Dow AgroSciences for reviewing the manuscript as well as Jarrad Prasifka for his assistance in its preparation.
References Cited
Author notes
Herculex and Refuge Advanced are trademarks of The Dow Chemical Company (“Dow”) or an affiliated company of Dow. SmartStax, YieldGard, YieldGard VT, YieldGard VT Triple Pro, and RIB Complete are registered trademarks of Monsanto Technology LLC. SmartStax multi-event technology developed by Dow AgroSciences and Monsanto.
