OUP user menu

Potential Biomass Reductions to Miscanthus × giganteus by Stem-Boring Caterpillars

J. R. Prasifka, J. D. Bradshaw, M. E. Gray
DOI: http://dx.doi.org/10.1603/EN11254 865-871 First published online: 1 August 2012


Injury from stem-boring caterpillars has been observed on the perennial grass Miscanthus × giganteus Greef and Deuter ex Hodkinson and Renvoize in both its native and introduced ranges. Because some species causing stem injury in the United States have not been identified, potential biomass reductions to M. × giganteus were measured using southwestern corn borer, Diatraea grandiosella Dyar (Crambidae), an insect pest of several related crops within the Andropogoneae. Results indicate D. grandiosella is capable of survival on whorl and stem tissue from hatch to 21 d in the laboratory, and field infestations with third instars support adult development, as exuviae were found during dissection of injured tillers. Relative to uninfested controls, M. × giganteus tillers with stem injury yielded 12-30% less dry mass in four infestations over 2009-2010. As in some D. grandiosella hosts, data indicate decreased susceptibility to stem-boring as tillers increase in size or age. Regressions of residuals (observed − predicted mass) for injured M. × giganteus tillers onto the cumulative length of tunnels per tiller also showed significant negative slopes (i.e., decreasing tiller mass with increasing tunnel length). Although D. grandiosella survival appeared low in both laboratory and field trials, results indicate that M. × giganteus productivity could become limited by other stem-boring caterpillars known to attack Andropogoneae, including the following: Elasmopalpus lignosellus (Zeller) (Pyralidae), Diatraea saccharalis (F.) (Pyralidae), and Eoreuma loftini (Dyar) (Crambidae). For perennial grasses grown exclusively for biomass, certain management strategies for stem borers or other pests may be uneconomical or impractical, suggesting long-term investment in breeding for host plant resistance may be needed.

  • ethanol
  • bioenergy
  • herbivore
  • pest management
  • lesser cornstalk borer

Miscanthus (sensu stricto) includes around a dozen species of perennial grasses native to southeast Asia. In particular, Miscanthus × giganteus Greef and Deuter ex Hodkinson and Renvoize (a sterile hybrid) and Miscanthus sinensis Andersson (a M. × giganteus parent) have been identified as high-yielding sources of biomass to provide energy (direct combustion or conversion to liquid fuel) or inputs to manufacture fiber-based products (Cappelletto et al. 2000, Clifton-Brown et al. 2008). The prospective range of cultivation for Miscanthus spp. includes much of the eastern United States, overlapping with several other biomass crops (USDOE 2006). Miscanthus spp. appear capable of supplying biomass to meet legislative requirements for advanced biofuels (Energy Independence and Security Act of 2007, 42 U.S.C. § 17001) while minimizing competition with food and feed crops; M. × giganteus yields currently compare very favorably with corn (Zea mays L.) and switchgrass (Panicum virgatum L.) (Heaton et al. 2008, Dohleman and Long 2009), but there may be potential for significant improvement because breeding and agronomic practices are still in early stages of development.

Another perceived advantage is the popular belief that Miscanthus spp. could be grown as a pest-free crop; to some degree, this idea has been supported by experience with M. × giganteus in Europe (Lewandowski et al. 2000, Semere and Slater 2007) and the fact that Miscanthus spp. have been used to breed disease resistance into sugarcane (Saccharum spp.) (James 2004). However, observations of damage by stem-boring insects in Asia (Clifton-Brown et al. 2008) and injury by a growing number of insects in the United States (Prasifka et al. 2009, Spencer and Raghu 2009, Bradshaw et al. 2010, Prasifka et al. 2011) indicate that rather than being pest free, the identity of pests or their effects on biomass are unknown. Miscanthus is related to several common food and feed plants in the Andropogoneae (sorghum tribe; Hodkinson et al. 2002), suggesting information on pest management in corn, sorghum (Sorghum bicolor L.), and sugarcane should help direct research and development of Miscanthus spp. in the United States. The first way in which knowledge from sorghum tribe crops can be used is to screen for potential pest species. Notably, all of the insects reported to feed on M. × giganteus in the United States are known pests from corn, sorghum, or sugarcane. A second valuable lesson from sorghum tribe crops is that pest identity and severity will continue to change over time. In recent years, pests such as the Mexican rice borer, Eoreuma loftini (Dyar) (Lepidoptera: Crambidae); Reay-Jones et al. 2008 have undergone range expansions, or evolved to defeat previous methods of management, such as resistance to crop rotation in western corn rootworm Diabrotica virgifera virgifera LeConte (Coleoptera: Chrysomelidae); Levine et al. 2002).

Although Clifton-Brown et al. (2008) noted damage to M. × giganteus by unspecified stem-boring caterpillars in Asia, only two lepidopteran species have been reported feeding on Miscanthus spp. in the United States: fall armyworm, Spodoptera frugiperda (J.E. Smith) (Noctuidae), have been observed feeding on emerging leaves in whorls (Prasifka et al. 2009); and stalk borer, Papaipema nebris (Guenée) (Noctuidae), tunneling within stems (Prasifka et al. 2011). Another unknown species caused significant injury boring into stalks of a first-year planting of M. × giganteus in Kentucky during 2008, but similar symptoms were not observed the next year (J.R.P. and J.D.B., unpublished data). Because P. nebris damage to M. × giganteus appears in a small proportion of tillers and may be restricted to plantings located near (or preceded by) smaller-stemmed hosts (Prasifka et al. 2011), it may not emerge as a problem for Miscanthus spp. However, other stem borers that infest cultivated Andropogoneae, particularly Diatraea spp., occur in the south central United States and have potential to interact with increased plantings of Miscanthus spp. The southwestern corn borer, Diatraea grandiosella Dyar (Crambidae), is of particular interest as a stem-boring caterpillar infesting corn, sorghum, and sugarcane. The primary importance of D. grandiosella is as a pest of corn, with two generations each year in northern parts of its range (e.g., Kansas) and three generations in warmer locations such as Texas and Louisiana (Morrison et al. 1977). In general, first generation larvae affect young corn plants by feeding in the whorl before boring into the stem, leading to shortened internodes and permanent stunting (Williams and Davis 1985). Later in the season, borer feeding may girdle the base of the stem, leading to stalk breakage (Zepp and Keaster 1977, Archer et al. 1987). Crop losses may be limited by planting cultivars with some host plant resistance along with cultural practices such as tillage (to kill overwintering larvae) and early planting. However, if D. grandiosella or other stem borers are capable of developing on Miscanthus spp., the use of early planting or tillage would be limited to the year of establishment because M. × giganteus and M. sinensis are perennials that form a dense mass of rhizomes and roots. Further, although the northern limits to the distribution of D. grandiosella and other southern species are restricted by cold-tolerance (Popham et al. 1991), there appears to be potential for significant range expansion with even modest increases in temperature in the central United States (Diffenbaugh et al. 2008).

Based on the known host range of D. grandiosella across the Andropogoneae, preliminary tests were conducted with newly-hatched (first-instar) larvae. Because initial tests indicated M. × giganteus whorl tissue was adequate to support larval development, subsequent field infestations of M. × giganteus with D. grandiosella were made over 2 yr. Given the unknown potential for D. grandiosella or other stem-boring species to infest plantings of Miscanthus spp. in the future, field trials were intended to generally assess the potential for feeding by stem-boring caterpillars to reduce biomass production by M. × giganteus.

Materials and Methods

Both laboratory and field observations of D. grandiosella were conducted on the ‘Illinois clone’ of M. × giganteus. This clone is the most thoroughly researched M. × giganteus in the United States and is being grown across the country as part of the U.S. Department of Energy evaluation of biomass feedstocks.

Insect Material.

D. grandiosella used in both laboratory- and field-based tests were obtained from a commercial insectary (Benzon Research Inc., Carlisle, PA) where the colony is maintained under warm (29°C), long-day (16-h photoperiod) conditions and larval rearing is accomplished with a wheat germ- and soybean flour-based artificial diet. The original source colony was founded and annually restocked with field-collected D. grandiosella in Mississippi before transfer to the insectary in 2008, after which no field-collected material was added. To examine larval development on M. × giganteus, eggs were received from the insectary and hatched inside an environmental chamber (28°C). For field infestations, D. grandiosella were shipped from the insectary as third instars within cells of artificial diet.

Laboratory Development.

To help assess the relative suitability of M. × giganteus for D. grandiosella larvae, neonates were provided tissue cut from greenhouse-grown tillers that averaged seven collared leaves (stage V7; Moore et al. 1991) at the beginning of the trial. Within 12 h of hatch, groups of five larvae were placed into small (60 by 15 mm) petri dishes (n=10) containing a 55-mm disc of filter paper wetted with 0.15-ml water, along with whorl tissue from 5 cm above the collar of the uppermost emerged leaf. The edge of each dish was sealed with stretched pieces of Parafilm M (Pechiney Plastic Packaging, Chicago, IL) to prevent desiccation. Additional whorl tissue was provided after 3 d. After 6 d, all surviving larvae were transferred in their original groups from small dishes into large (100 by 20 mm) petri dishes containing leaf tissue and a 90-mm disc of filter paper wetted with 0.30-ml water. In addition to whorl tissue, larvae were provided with 5–10 cm of stem tissue below the collar of the uppermost emerged leaf, and dishes were sealed. Additional whorl and stem tissue were added as needed, usually every 2 d. Throughout the trial, survival was assessed whenever plant tissue was added. Though no D. grandiosella cannibalism was observed, at 9 d concerns about potential cannibalism led to the removal of all but the largest larva from each dish. Larval masses were assessed at 9, 15, and 21 d, after which the trial was concluded.

Field Infestations. 2009.

Trials were conducted at sites in Champaign Co., IL, further north than significant numbers of D. grandiosella regularly occur. However, trapping of moths from southern Illinois indicates adults are active during the first or second week of June. Accordingly, in 2009, infestations were made in the third week of the month, on 19 June. In an M. × giganteus plot established in 2005, 120 tillers were identified with vinyl tags and the initial height (from soil level to the uppermost leaf collar) was recorded. Half of the tagged tillers were immediately infested with a single third-instar D. grandiosella placed into the whorl. Because of an unexpected, large amount of precipitation (>3 cm) within 6 h of infestation, the same tillers were infested with two larvae again 24 h later.

Both the control (uninfested) and infested tillers were harvested manually by cutting each stem at ground level on 30 November (within normal harvest period for central Illinois). After transport to the laboratory, the final height of individual tillers (from the base to the uppermost node) was measured. Tillers that showed evidence of D. grandiosella injury to the stem (i.e., visible entry or exit holes) were split and the number, location of holes, and cumulative length of tunnels was recorded. Finally, tillers were cut into sections, placed individually in paper bags, and dried (65°C for 10 d) before the mass of each stem and any remaining leaves was determined.


Observations from 2009 suggested that relatively short tillers (at time of infestation) might be more susceptible to D. grandiosella. Consequently, infestations in 2010 included date as a variable, with infestations made on 27 May, 10 June, and 24 June (early, middle, and late). In a M. × giganteus planting established in 2007, 200 tillers were identified with vinyl tags. As in 2009, initial height was recorded for 50 tillers at the time of each infestation, or for controls (which were used for comparison to each infestation date) on all three dates. Though no significant precipitation was noted around the dates of infestation in 2010, to increase the chances of D. grandiosella establishing on infested tillers, a single larva was placed in each tiller on successive days (i.e., for the early infestation, one larvae on 27 May, one on 28 May).

Relative to 2009, harvest and sample processing in 2010 differed on two points: 1) tillers were cut on 19 August rather than in November; with this early harvest, tillers contain significantly more moisture, but overall dry mass is at its peak (Heaton et al. 2008), as leaves have not fallen or blown off of M. × giganteus stems; and 2) the locations of holes made by D. grandiosella were not recorded.

Statistical Analyses.

For the laboratory development trial, the primary objective was to determine before any field testing whether D. grandiosella larvae were capable of developing when supplied only M. × giganteus tissue. Consequently, other than summary statistics for developing larvae (mean mass ± SE), no additional analysis was conducted. For the 2009 field infestations, a significant proportion of infested M. × giganteus tillers were found to be free of any injury to the stems. Consequently, tillers were categorized by infestation type as control (uninfested), uninjured (but infested), or injured (stem with visible holes). Subsequently, an analysis of covariance (ANCOVA) (PROC GLM; SAS Institute 2007) was used to test for the effects of infestation type (control, uninjured, injured), initial height (covariate) and a possible infestation × height interaction on the dry mass of M. × giganteus tillers. If no significant interaction was detected, the model was run again excluding the interaction term. When an effect of infestation was indicated, pairwise comparisons were made using least-squares estimated means via the Least Significant Difference test. In addition, regression (PROC REG; SAS Institute 2007) was used to estimate the relationship between dry mass and initial height for control tillers. The regression of dry mass on to initial height for uninfested tillers allowed calculation of residuals for injured M. × giganteus tillers (i.e., residual=observed dry mass - expected mass from regression); in this case, residuals represent estimated yield loss in grams of biomass per tiller. Residuals then were regressed onto the cumulative length of tunnels per tiller, to test whether tunnel length could predict the degree of damage to M. × giganteus. Analysis of 2010 field infestations used the same method, but was repeated independently for each infestation date.


Laboratory Development.

For the 9 d in which D. grandiosella were reared in groups, survival was 28% (14 of 50), though at least one larva remained in each of the 10 small petri-dishes. After transfer of the largest single larva in each of the small dishes into larger petri-dishes, mortality continued, with only 40% (four of 10) transferred larvae surviving the period between 9 d and 21 d. However, surviving larvae continued to develop with masses of (mean ± SE) 2.3 ± 0.4 mg (n=10), 9.7 ± 3.6 mg (n=7), and 42.3 ± 12.9 mg (n=4) at 9, 15 and 21 d, respectively.

Field Infestations.

The full ANCOVA for 2009 indicated no significant interaction term (P > 0.05), resulting in a revised model with significant effects of infestation (F=4.23; df=2, 113; P=0.017) and initial height (F=78.48; df=1, 113; P < 0.001) on the dry mass of harvested M. × giganteus tillers. Comparisons of least-squares estimated means found the dry masses of tillers with stem injury 20% less than uninjured (but infested) M. × giganteus, but no statistical difference with tillers designated as (uninfested) controls (Fig. 1). The regression of dry mass onto initial height of control tillers explained nearly one-third of variation in harvested biomass (F=26.70; df=1, 57; P < 0.001); though no interaction was detected statistically, there appeared to be a trend of greater damage from D. grandiosella for the shortest tillers at the time of infestation (Fig. 2).

Fig. 1.

Mean mass (least-squares estimate ± SE) per M. × giganteus tiller by D. grandiosella infestation and stem injury for 2009. Differences indicated by the Least Significant Difference test are shown by different lowercase letters above columns. Sample sizes shown numerically at the base of each column.

Fig. 2.

Mass versus initial height for M. × giganteus tillers by D. grandiosella infestation and injury to stem in 2009. Plotted regression line and statistics provided for uninfested controls.

As in 2009, the analyses for the three 2010 infestation periods showed no significant interaction between infestation and initial height. However, dry mass was significantly influenced by both infestation and initial height for the early ([infestation] F=20.27; df=2, 93; P < 0.001 [initial height] F=68.83; df=1, 93; P < 0.001); middle ([infestation] F=25.04; df=2, 92; P < 0.001 [initial height] F=86.99; df=1, 92; P < 0.001); and late ([infestation] F=4.49; df=2, 93; P=0.014 [initial height] F=71.32; df=1, 93; P < 0.001) infestation periods. Across all three infestation periods, dry mass of tillers with stem injury was ≈ 27% less than controls; for the middle infestation period, estimated dry masses also were lower for uninjured M. × giganteus than control tillers (Fig. 3). Regressions of dry mass onto initial height of control tillers explained approximately half of the variation in harvested biomass at the early (F=60.24; df=1, 45; P < 0.001); middle (F=47.88; df=1, 45; P < 0.001); and late (F=43.13; df=1, 45; P < 0.001) infestation periods (Fig. 4 .).

Fig. 3.

Mean mass (least-squares estimate ± SE) per M. × giganteus tiller by D. grandiosella infestation and stem injury for (A) early, (B) middle, and (C) late infestation periods in 2010. Differences indicated by Least Significant Difference test within an infestation period are shown by different lowercase letters above columns. Sample sizes shown numerically at the base of each column.

Fig. 4.

Mass versus initial height for M. × giganteus tillers by D. grandiosella infestation and injury to stem for (A) early, (B) middle, and (C) late infestation periods in 2010. Plotted regression line and statistics provided for uninfested controls. Note differences in scale of x-axis between infestation periods.

Regressions of residuals for injured M. × giganteus tillers onto the cumulative length of tunnels per tiller showed significant negative slopes (i.e., decreasing tiller mass with increasing tunnel length) in 2009 (F=13.66; df=1, 24; P=0.001) (Fig. 5) and for the early infestation of 2010 (F=5.50; df=1, 25; P=0.027) (Fig. 6 A); however, cumulative length of tunnels within stems did not explain the distribution of residual dry mass during the middle and late infestations in 2010 (Figs. 6 B, C).

Fig. 5.

Residual mass (observed- predicted by control data) for injured M. × giganteus tillers versus cumulative length of tunnels produced by D. grandiosella in 2009. Regression indicates a significant linear relationship between residuals and tunnel length (i.e., increasing damage with increased tunnel length).

Fig. 6.

Residual mass (observed- predicted by control data) for injured M. × giganteus tillers versus cumulative length of tunnels produced by D. grandiosella for (A) early, (B) middle, and (C) late infestation periods in 2010. Regression indicates a significant linear relationship between residuals and tunnel length (i.e., increasing damage with increased tunnel length) for an infestation period.


D. grandiosella was a suitable, if not optimal, species to investigate the potential effects of stem borers on M. × giganteus biomass production. Larvae were capable of survival on whorl and stem tissue from hatch to 21 d in the laboratory, and field infestations of third instars supported development of adults, as exuviae occasionally were found during dissection of injured tillers. Most importantly, third instars placed into whorls significantly reduced M. × giganteus yields (Figs. 1 and 3). Although D. grandiosella survival appeared low in both laboratory and field trials, results indicate that M. × giganteus productivity could become limited by stem borers known to attack Andropogoneae [e.g., E. loftini, Diatraea saccharalis (F.)].

If the colony of D. grandiosella used is representative of natural insect populations, M. × giganteus may have resistance relative to crops for which southwestern corn borer is an established pest. Although yield reductions for the injured M. × giganteus are similar to those observed in other crops, far fewer M. × giganteus stems showed tunneling relative to hand-infested corn, sorghum, or millet (Starks et al. 1982). Acknowledging differences in methodology, comparisons also can be made to growth on resistant or susceptible corn hybrids. Daves et al. (2007) reared larvae on diets with lyophilized whorl tissue, finding larval masses four (resistant) or eight (susceptible) times greater after 14 d as were seen for D. grandiosella fed M. × giganteus for 15 d (with both sets of larvae reared at 28°C); when compared with D. grandiosella in the field, larval masses were two (resistant) or five (susceptible) times greater after 21 d than when southwestern corn borers were provided M. × giganteus in the lab over the same period of time. Also, in M. × giganteus field trials, significant numbers of tillers infested showed no significant damage to stems (Figs. 1 and 3). However, the combined results of lab and field trials are not sufficient to assess the insect resistance of M. × giganteus; to determine relative resistance to D. grandiosella would require direct comparison to other hosts like corn or sorghum by using equivalent methods. Furthermore, the use of a single clone of M. × giganteus does not adequately represent the diversity of Miscanthus spp. being explored for biomass production.

Previous research with D. grandiosella has shown an impact of host plant maturity on both injury and damage. Arbuthnot et al. (1958) found a 10-d delay in infestation date significantly decreased the effect of two or five borers per plant in both field and sweet corn. Williams and Davis (1985) observed that ratings of leaf feeding were greatest for corn infested 4–5 wk after planting; whereas very little leaf injury was observed in M. × giganteus (likely a result of using a small number of third-instar larvae per tiller), data from 2009 and 2010 indicate decreased susceptibility to stem boring as tillers increase in size or age. In particular, less than one fifth of infested tillers showed stem injury for the late infestation in 2010 (Fig. 3 C), though all previous infestations (Figs. 1 ; 3 A, B) collectively show nearly half of all infested tillers with stems bored by D. grandiosella. Also, although the analyses of covariance accepted common slopes for control, uninjured and injured tillers, plots of harvested mass versus initial height (Figs. 2, 4 A) appear to show the greatest reductions in dry mass (relative to the control regression) for injured tillers with low initial heights. Starks et al. (1982) showed the effect of D. grandiosella on yields to be greatest in late planted corn, but observed the opposite trend for millet and sorghum, suggesting that though plant phenology and damage from stem borers may interact, the relationship may not be consistent among biomass crops.

The length of tunneling produced by D. grandiosella and other important stem-borer species [e.g., Ostrinia nubilalis (Hübner)] has been used as an indicator of conventional and transgenic insect resistance (Walker et al. 2000, Daves et al. 2007). Though cumulative tunnel length also may be influenced by D. grandiosella larval population (Davis et al. 1978) or infestation time (Starks et al. 1982), it is reasonable to expect this measure of injury to relate to overall damage (i.e., yield reduction) for infested plants. Regression of residual dry mass (≈ estimated yield reduction per tiller) onto tunnel length in 2009 showed little effect with <10 cm of D. grandiosella tunnels, but ≈40% reduction in yield with >15 cm of tunnels (Fig. 5). In 2010, only one of three infestations (early, Fig. 6 A) showed a significant relationship between residual dry mass and tunnel length. However, further examination of the middle infestation (Fig. 6 B) suggests that excluding a single influential observation, a similar relationship would exist to that found in the early infestation. Consequently, excluding the late infestation (Fig. 6 C, where only eight tillers infested with D. grandiosella showed tunnels), the extent of tunneling in M. × giganteus may also prove useful in any attempts to examine plant resistance in Miscanthus spp.

Based on its response to D. grandiosella, a significant pest of corn, M. × giganteus ' reputation as insect-resistant (Lewandowski et al. 2000, Semere and Slater 2007) may appear justified. But experience with other introduced crops such as soybean [Glycine max (L.) Merr] suggests that with time and expansion of total growing area, significant pest problems likely will emerge (Giesler et al. 2002, Venette and Ragsdale 2004). Evidence that some regional problems are developing for Miscanthus spp. come from recurring infestation of leaf blight in Kentucky (Ahonsi et al. 2010) and reports of transplants killed by stem-boring insects in several locations in Georgia (R. D. Lee, personal communication) and Arkansas (R. Pyter, personal communication); although the identities of the insects in Georgia and Arkansas are not yet known, the damage to stems at or near ground level and the sandy soils at each location seem consistent with lesser cornstalk borer, Elasmopalpus lignosellus (Zeller) (Lepidoptera: Pyralidae)(Leuck 1966). Although few stem-boring lepidopterans or other insects appear to present serious, immediate threats to perennial grasses grown exclusively for biomass, the degree to which insects and other pests limit biomass yield is likely to rise as the area used for their production may increase by orders of magnitude (Perlack et al. 2005); as pest problems do develop, investment in breeding for host plant resistance may be needed, as some basic management strategies for stem-borers or other pests may be uneconomical (pesticides or transgenic resistance) or impractical (modified planting dates, mechanical control by using tillage).


Research funding was provided by the Energy Biosciences Institute. We appreciate assistance from Drew Schlumpf and Chris Rudisill, who helped with establishment and maintenance of research plots. Bret Hash and Kevin Mazur assisted in artificial infestations of M. × giganteus.

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


View Abstract