OUP user menu

Comparison of Nonmaize Hosts to Support Western Corn Rootworm (Coleoptera: Chrysomelidae) Larval Biology

Thomas L. Clark, Bruce E. Hibbard
DOI: http://dx.doi.org/10.1603/0046-225X-33.3.681 681-689 First published online: 1 June 2004


With the recent commercialization of transgenic rootworm-resistant maize with high levels of antibiosis to larval feeding, the biology of western corn rootworm, Diabrotica virgifera virgifera LeConte, on hosts beyond maize, Zea mays L., has become an important topic for which data are limited. Larval survivorship and growth parameters were monitored on the roots of 29 plant species comprised of maize, maize-field weeds, native prairie grasses, forage grasses, and small grain crops. Data on larval recovery and growth (measured as increases in head capsule width and accumulation of dry weight) were recorded at five samplings (6, 10, 14, 20, and 24 d) after initial infestation of the 29 species. Recovery and growth parameters were analyzed for inter- and intraspecific differences within and among sampling dates. Larvae survived at least 6 d after infestation on 27 species and 24 d on 23 plant species. Larval recovery and growth were impacted by both species and time after infestation. Growth and development of larvae were significantly slower on most plant species beyond maize; however, 18 of the species had larvae develop to the second instar, whereas larvae on 14 species developed to the third instar. Adults were recovered from five plant species in addition to maize. Because rootworm-resistant transgenic maize with high levels of antibiosis has become a part of the agroecosystem, weeds in grassy maize fields as well as adjacent forage grass species may become more important in the western corn rootworm life cycle, particularly because genes conferring resistance to postemergent herbicides such as glyphosate are stacked with transgenic rootworm-resistant maize hybrids.

  • Diabrotica virgifera virgifera
  • alternate hosts
  • resistance management
  • rootworm


THE GENUS Diabrotica is divided into three species groups (virgifera, fucata, and signifera) that can be distinguished by morphology (Wilcox 1972, Krysan and Smith 1987), mitochondrial and nuclear DNA genes (Clark et al. 2001), and biology and life history traits. One of the more intriguing distinctions between the groups is larval host selection. A distinct difference between the species groups is the apparent larval specialization of virgifera group species on the roots of specific grass species (Poaceae), whereas fucata group species are polyphagous on the roots of several plant families (Branson and Krysan 1981, Krysan and Smith 1987). Perhaps one of the more striking examples of purported larval specialization is observed in the western corn rootworm, Diabrotica virgifera virgifera Le-Conte,–maize, Zea maysL., herbivore–host association (Branson and Krysan 1981). This association is important because it may result annually in a $1 billion loss to U.S. maize producers due to control costs and yield losses (Metcalf 1986). In general, damage is caused by young larvae (first and early second instars) feeding on fine roots to older larvae (mid-second to third instars) burrowing into the root core (Apple and Patel 1963) feeding on cortex tissues and excised root sections (Riedell and Kim 1990).

Because damage by western corn rootworm has such a negative impact on maize production, several management strategies have been implemented in attempts to keep populations below economically damaging levels. One strategy is the application of soil insecticides at either planting or first cultivation with the aim of protecting the maize root zone from rootworm feeding, especially in fields where continuous maize is grown (Mayo and Peters 1978). Soil insecticides have been relatively successful in reducing damage compared with untreated fields; however, overuse of this tactic has led to the development of western corn rootworm resistance to organochlorine insecticides in parts of the Corn Belt (Ball and Weekman 1962, Siegfried and Mullin 1989). Related to soil insecticide application is the use of aerial applied insecticides targeting adult populations. The intent of this tactic is to suppress egg deposition by adults (larval population reduction for the following year) (Pruess et al. 1974), prevent clipping of maize silks (Culy et al 1992), and seed quality issues. Problems that are associated with adult control include acceleration of insecticide resistance (Meinke et al. 1998, Miota et al. 1998, Scharf et al. 1999), and unnecessary cost, when directed toward silk clipping, because this problem rarely reaches economically damaging levels in commercial maize production (Capinera et al. 1986). Crop rotation is one of the more effective means of controlling western corn rootworm. However, a behavioral variant that oviposits eggs in soybean, Glycine max (L.) Merrill, fields (impacting maize the following year) has become prevalent in portions of Illinois, Indiana, and Michigan, eliminating the soybean– maize rotation system as a viable management strategy where the variant hasbecome established (Levine and Gray 1996, OʼNeal et al. 1999). Another potential control strategy is the development of rootworm-resistant maize varieties. For example, cultivars with certain degrees of tolerance (Wilson and Peters 1973, Owens et al. 1974, Branson et al. 1986), antibiosis (Assabgui et al. 1995), and unknown mechanisms (Branson et al. 1983, Hibbard et al. 1999) have been identified, but levels of resistance in all categories have been relatively low, limiting the integration of this tactic in commercial production systems. More recently, several companies have developed maize hybrids containing genes that code for production of insecticidal proteins from the soil bacterium Bacillus thuringiensis Berliner (Bt) that have high levels of antibiosis to neonate western corn rootworm. northern corn rootworm, Diabrotica barberi Smith & Lawrence; Mexican corn rootworm, D. virgifera zeae Krysan & Smith; and southern corn rootworm, Diabrotica undecimpunctata howardi Barber. Because maize with strong antibiosis to economically important rootworms is being introduced into the production system for the first time (Monsanto Companyʼs YieldGard rootworm product received registration for commercial sale on 25 February 2003), research on rootworm–host plant interactions has become a critical research area because rootworm larvae may be challenged in a manner unlike any other tactic implemented for rootworm control. Hosts beyond maize may become more important in the rootworm life cycle as transgenics are introduced into the maize production system.

In studies to determine the larval host range of western corn rootworm, Branson and Ortman (1967, 1970) observed larval survival for at least 10 d on 18 of 44 grass species screened, all of which developed to the second instar, whereas no larvae developed on any of the 27 broadleaf species screened. Their studies provided a foundation for exploring the expanded host range of western corn rootworm, particularly regarding the “grasses only” larval host hypothesis of this and other virgifera group species. Despite the importance of those studies, more research is necessary in refining growth and development parameters of rootworm larvae on hosts beyond maize. Such information may have application toward the development of resistance management plans for rootworm-resistant transgenic maize hybrids and provide additional insights into the evolution of virgifera group species with grasses. In this study, our goal was to measure growth parameters of western corn rootworm larvae when placed upon grassy hosts that occur within its geographical range that are considered to be important maize-field weeds, native prairie species, forage species, and small grain crops.

Materials and Methods

Plant Material and Insects

Twenty-eight grass and one broadleaf species obtained from several sources (Table 1) were sown in pots containing a 1:1 soil:peat-based growing medium mixture (Professional General Purpose Growing Medium, Premier Horticulture Inc., Red Hill, PA). The soil portion of the mixture was sterilized using an autoclave (Amsco Eagle Series 2051 Gravity, Continental Equipment Co., Lawrence, KS) at 120°C for 3 h to kill seed contaminants. Drainage openings in each pot were fitted with a fine (114 μm per opening) stainless steel mesh (TWP, Inc. Berkeley, CA) to prevent larval escape. After planting, pots were placed in a greenhouse maintained at 25 ± 2.4°C and a photoperiod of 14:10 (L:D) h. All plants were watered as necessary and fertilized (Scotts Peters Professional 20–20-20, Scotts-Sierra Horticultural Products Corp., Marysville, OH) every 10 d.

View this table:
Table 1.

Host Suitability Experiment

The experiment was designed as a split-plot and evaluated for survival and growth parameters among the 29 selected plant species (the main effect) as well as within each species over five sample dates (the subplot effect). Six pots of each plant species were randomly placed on four separate benches (24 pots total for each species), where each bench represented a replication. The six pots within each replication were randomly assigned a sampling date for larval extraction or adult emergence. Five weeks after planting, each pot was infested with 15 neonate western corn rootworm larvae by gently transferring individuals with a moistened camelʼs-hair brush from an egg dish to a single hole (1 cm in diameter by 3 cm in depth) that was dug in each pot. After infestation, each hole was gently filled with excess soil mixture. Because the pots were placed in proximity to each other, a second series of uninfested pots containing maize were planted to document whether larvae would move from pot to pot. Any larva that seemed to be injured during the infestation process was removed and replaced by a healthy larva. A subsample of 100 neonate larvae was placed in 95% ethanol to serve as a baseline for growth measurements.

At 6, 10, 14, 20, and 24 d after infestation the contents (soil mixture, roots, and larvae) of the randomly assigned pots were placed individually in Tullgren funnels equipped with 60-W light bulbs for extraction of larvae. Collecting jars containing water were placed under the funnels and checked for larvae twice daily for 4d. All collected larvae were transferred from the collecting jars to individually labeled glass scintillation vials containing 95% ethanol. On day 28 after infestation, the sixth and only remaining pots were covered with insect mesh (0.60 by 0.60 mm opening, ECONET L, LS Americas, Co., Charlotte, NC) after trimming the vegetative material to within 2 cm of the soil mixture surface. Each remaining pot was checked daily for adult emergence. All adults collected were placed in individually labeled vials containing 95% ethanol, and adult number was recorded. Each larva from all samples (including the initial neonate subsamples) was measured for head capsule width by using an ocular micrometer (10X/21, Wild Co., Heerbrugg, Switzerland) mounted on a microscope (M3Z, Wild Co.). When head capsule measurements were complete, the ethanol was carefully poured off with the larvae remaining in the vials. The vials (with caps removed) were then placed in a desiccating oven (Thelco model 16, GCA/Precision Scientific Co., Chicago, IL) at 90°C for 48 h to remove excess moisture. Larvae (including the subsample) were then weighed

using an analytical scale (ER-182A, A & D Co., Tokyo, Japan) with total weight for all dried larvae from individual pots being recorded. The average dry weight for each individual larva was then calculated by simple division.

Statistical Analysis

Three variables were measured to determine host suitability of each plant species evaluated. The first was the average change in larval head capsule width. This was determined by subtracting the average head capsule width of the initial neonate larvae subsample from the head capsule widths of the larvae recovered from the individual pots at each sample date. Second, the average change in larval dry weight was determined by subtracting the average dry weight of the neonate subsample from the average dry weight of larvae collected from each sampling date. The third variable was percentage of larval recovery. All data were analyzed for analysis of variance (ANOVA) with the PROC GLM procedure (SAS Institute 1990). Least significant differences (LSDs) (α=0.05) were calculated for intraspecific growth and recovery through time and interspecific growth and recovery at a given time as described by Steel and Torrie (1980) with mean squares, sum of squares, and degrees of freedom parameters from the PROC GLM procedure. Percentage of recovery data were transformed (arcsine) before analysis but are presented as untransformed means (Steel and Torrie 1980). Adult emergence from the plant species was also recorded but not analyzed due to minimal recovery.

Means between plant species within a sampling date (columns) followed by the same lowercase letter are not significantly different at a = 0.05. Means between sampling dates (rows) followed by the same uppercase letter are not significantly different.


Larval Recovery

Larvae survived at least 6 d on 27 of 29 species evaluated and 24d on 23 species (Table 2). The percentage of western corn rootworm larvae recovered from root masses placed in Tullgren funnels was significantly impacted by plant species (F =18.36; df =29, 87; P <0.0001). Time after infestation also significantly impacted recovery rates (F =9.39; df =4, 360; P <0.0001). The interaction of treatment and sample date also significantly impacted larval recovery (F =1.55; df =116, 360; P <0.0013) with the presence of pupae in some of the later sample dates, particularly in the maize treatment, accounting for this significant interaction. Although maize had the highest larval recovery rate with larvae being recovered from all five samplings, 17 species also had larvae recovered at all samplings, two had recovery from four samplings, five from three samplings, one each for two sampling and one sampling(s), and three treatments had no larval recovery (Table 2). The samples taken on days 10,14, and 20 had the highest larval recovery for most species (Table 2). The presence of pupae for the day 24 maize sample reduced the number of individuals collected on that date, because pupae are rarely collected using the Tullgren insect recovery technique.

View this table:
Table 2.

Head Capsule Width and Dry Weight

The subsample of 60 neonates (for growth measurements) had a mean head capsule width (HCW) of 0.220 mm and mean dry weight was 9.298 &mu;-g. Both values were used as a baseline for growth measurements. Plant species significantly impacted change in HCW and accumulation of dry weight (F =15.82; df =29, 87; P <0.0001 and F =9.06; df =29, 87; P <0.0001, respectively (Tables 3 and 4). Significant increases in HCW and mass accumulation were also observed between sampling dates (F =42.13; df =4, 360; P <0.0001 and F =37.26; df =4, 360; P <0.0001). The treatment by date interaction also significantly impacted increase in HCW and mass accumulation (F =2.04; df =116, 360; P <0.0001 and F =1.91; df =116, 360; P <0.0001) because larvae grew at different rates for the different species evaluated. In general, a significant increase in HCW for both intra- and interspecific comparisons represents a molt to an advanced instar (Table 3). On sample day 6, most larvae recovered from maize had reached the second instar, whereas some larvae recovered from western wheatgrass and Rhodes grass had also reached second instar (Table 3). In total, 18 of the 29 infested species evaluated had the majority of larvae develop to at least the second instar (as defined by an increase of 0.124 mm in HCW), whereas 14 of the species had larvae that developed to the third instar (Table 3). For maize, most larvae had reached the third instar by day 14, whereas on most other species larvae did not reach maximum HCW until either day 20 or 24(Texas panicum and wheat were the exceptions) (Table 3). For species with significant intraspecific larval dry weight gains, four species (maize, reed canarygrass, witchgrass, and wild prosso millet) reached maximum weight gain on day 20, whereas other species did not reach maximum weight until day 24(Table 4). In total, 15 of the 29 infested species had statistically significant intraspecific dry weight gain.

View this table:
Table 3.
View this table:
Table 4

Adult Recovery

Adult recovery was minimal with single adults (0.25 ±0.25 per pot) being recovered from large crabgrass, barnyardgrass, sand lovegrass, and western wheatgrass; three adults (0.75 ±0.75 per pot) were recovered from Texas panicum; and 18 adults (4.50 ±1.84 per pot) were recovered from maize.


Several companies have developed transgenic maize hybrids with strong antibiosis properties for neonate western corn rootworm larvae. On 25 February 2003, the first of these products received registration for commercial use by the Environmental Protection Agency. It is a transgene placed in maize hybrids that expresses the Cry3Bb1 protein from Bt developed by Monsanto Co. The product, YieldGard rootworm, is available for sale as a stack with broad-spectrum nonresidual herbicide tolerance (glyphosate) that allows for postemergence treatment of weeds. Interestingly, this product is not effective on larvae that have developed beyond the first instar (FIFRA Scientific Advisory Panel Meeting 2002). Therefore, any larva that initially establishes and develops on a grassy weed beyond maize has the potential to move to rootworm-resistant transgenic corn after the weed is removed (sprayed with a postemergence herbicide such as glyphosate or atrazine) or becomes phenologically unfavorable. Such a scenario is likely, because larvae will move from plant to plant when their food source is destroyed or eliminated (Hibbard et al. 2003). Results of this study show that western corn rootworm larvae can develop significantly on hosts beyond maize. However, it remains unclear as to whether such movements have the potential to enhance resistance by supporting larval survival at sublethal doses (i.e., larvae may gain enough mass by feeding on weed roots to overcome a dose of the transgenic insecticidal protein), enhance susceptibility by supporting larval growth to a mass that renders the transgenic dose as ineffective (i.e., would keep additional susceptible alleles in the population), and/or enhance the possibly economic damage on rootworm-resistant transgenic maize. Further examination is necessary in these areas to answer these important questions.

Although the results of our study corroborate those of previous studies by Branson and Ortman (1967, 1970) that many grassy species beyond maize are able to support the growth of western corn rootworm larvae, our methods and findings differ in several respects. First, Branson and Ortman (1967, 1970) used a 10-d assay of qualitative survivorship to assess whether larvae would survive on growing root tissue in petri dishes. In contrast, our assay was comprised of a time-series removal of larvae from the different plant species growing in a soil medium that allowed for documentation of percentage of survivorship and growth over time. We evaluated 16 species in common with Branson and Ortman (1967, 1970), and we are in disagreement with their assessment of western corn rootworm larval host range for five of the common species. For example, reed canarygrass and barnyard grass were listed as nonhosts by Branson and Ortman (1967, 1970), but our results show that these species were actually among the better hosts of western corn rootworm larva in percentage of recovery and growth parameters (Tables 2–4). The other species in common for which our assessment differs (johnsongrass, orchardgrass, and switchgrass) were what we considered to be marginal hosts with limited recovery and reduced growth, whereas RBranson and Ortman (1967, 1970) listed these as nonhosts. A possible reason for the differing results could be sample size and evaluation time, because Branson and Ortman (1967, 1970) evaluated three to five larvae in six petri dishes (18 –30 total larvae per species), and we evaluated 15 larvae per pot with five sample dates and four replicates (240 total larvae per species). The third primary difference between the studies was our emphasis on grassy weed species (17 of 29 species evaluated), whereas Branson and Ortman (1967, 1970) examined a greater percentage of native prairie and forage species as only six of 44 grass species were weeds.

Many phytophagous insects such as western corn rootworm are either monophagous or oligophagous on a few host plant species, and factors such as genetics, phenology, and environment influence the suitability of a host plant for an insect herbivore (Ehrlich and Raven 1964, Singer and Parmesan 1993). Variation within a host plant species, such as the introduction of a transgene with antibiosis properties, can be a source of heritable change in host selection behavior and possible divergence between insect populations (Futuyma 1983, Jaenike and Holt 1991). Sometimes, such a shift is from a more monophagous feeding habit toward a more oligo- to polyphagous host preference. For example, Harrison (1987) observed that three geographically distinct populations of Colorado potato beetle, Leptinotarsa decemlineata (Say) (a species historically associated with buffalo bur, Solanum rostratum Dunal), had become locally adapted to other solenaceous plants that were native to their respective home ranges. Interestingly, they maintained some preference to their original host, S. rostratum, despite 100 or more generations of isolation. Other studies have documented that acceptance of host plants is a selectable trait that can have implications for host acceptance. For example, heritable differences in host selection behavior by using different phytochemical cues were observed after three generations of divergent selection for the pine engraver, Ips pini (Say), by using phytochemical cues in full-sib breeding lines (Wallin et al. 2002). Whether such a scenario is possible for western corn rootworm larval host selection remains to be seen; however, some evidence suggests that this species does have an ability to adapt to selection pressures that involve host choice. For example, western corn rootworm beetles in parts of the Corn Belt have adapted to crop rotation with a lost or reduced fidelity to maize fields as a preferred ovipositional environment (Spencer et al. 1999).

Beyond promotion of behavioral biotypes, the deployment of an antibiosis factor in a crop plant may have enhanced effectiveness if rare genotypes exist in the population that prefer alternate host plant species, especially if these individuals have enhanced fitness on the alternate host (Gould 1983, Kennedy et al. 1987). When a host switch occurs it is most likely to do so within the same plant family (Mitter et al. 1991) with related or ancestral host plants being colonized (Futuyma and McCafferty 1990, Funk et al. 1995). The introduction of maize with strong antibiosis properties to western corn rootworm represents a source of variation within the host plant (maize) that could be a genetic driver for utilization of other grassy species beyond or in addition to maize. Our results support the potential for alternate host utilization; however, additional studies are necessary to document comparative selection intensity and fitness of western corn rootworm reared on hosts beyond maize in comparison to rootworm-resistant transgenic maize.

In summary, we examined survivorship, growth, and development of the biology western corn rootworm larva on several alternate hosts beyond maize. Our data, although not inclusive of all grassy species for the geographic distribution of western corn rootworm, suggest that any one of several grassy species can have a significant impact on the biology of western corn rootworm larvae. As weed management programs for maize tend toward the use of broad-spectrum, non-residual herbicides such as the use of maize containing glyphosate tolerance genes stacked with rootworm-resistant transgenic, weeds in grassy maize fields as well as adjacent forage grass species may become more important in the life cycle of western corn rootworm. However, additional research is necessary to ascertain the role grassy weeds may have in the western corn rootworm life cycle because intense selection pressures in the form of strong antibiosis is being applied to western corn rootworm populations for the first time in the history of controlling this pest.


We thank Arnulfo Antonio, Charles Thiel, Julie Barry (USDA–ARS, Plant Genetics Research Unit), and Kevin Moore and Yvonne Schweikert (University of Missouri Plant Sciences Unit) for assistance in different phases of this study. We thank Mark Ellersieck (University of Missouri Agriculture Experiment Station) for statistical assistance. We thank Ted Wilson, Matt Higdon (Department of Entomology, University of Missouri), and Larry Darrah (USDA–ARS, Plant Genetics Research Unit) for thoughtful comments on the manuscript. Funding, in part, was provided by CSREES Project Award No. 2001-35316-10000 to B.E.H and supported in part by the Missouri Agricultural Experiment Station.


  • This article reports the results of research only. Mention of a proprietary product does not constitute an endorsement or recommendation for its use by the USDA or the University of Missouri.

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/), which permits non-commercial reuse, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions{at}oup.com

References Cited

View Abstract