A holistic approach to developing new corn germplasm that confers multiple insect resistance in various plant tissues at different growth stages was examined. Eight corn germplasm lines were examined for their foliar resistance to fall armyworm [Spodoptera frugiperda (J. E. Smith) (Lepidoptera: Noctuidae)] and natural enemy attraction at V6–V8 (or 6–8 leaf) stages in 2008 and 2009. Four corn germplasm lines with known levels of resistance to root- and ear-feeding insects [‘CRW3(S1)C6’, ‘B37*H84’, ‘SIM6’ and ‘EPM6’], and four germplasm entries with different levels of S. frugiperda resistance (‘Mp708’, ‘Ab24E’, ‘FAW7061’ and ‘FAW7111’) were evaluated in the study. All plants were manually infested with 15–20 neonate S. frugiperda larvae per plant, and injury was rated 7 and 14 d after infestation. Based on cluster analysis of S. frugiperda injury rating and predator survey data, ‘Mp708’ and ‘FAW7061’ were the most resistant, whereas ‘Ab24E’ and ‘EPM6’ were the most susceptible to fall armyworm feeding. The western corn rootworm-resistant ‘CRW3(S1)C6’ showed resistance to S. frugiperda feeding. Surveys for the diversity and abundance of predators of S. frugiperda in each experimental plot were also conducted 7 d after infestation. ‘CRW3(S1)C6’ and ‘Ab24E’ had the highest and lowest predator abundance, respectively. However, there was no direct correlation between S. frugiperda injury ratings and predator abundance. The current study demonstrated the feasibility of developing foliage-, root-, and ear-feeding insect-resistant germplasm covering multiple corn growth stages. In addition, the possibility of utilizing plant volatiles to attract predators, and reduce pest populations and crop damage is discussed.
The fall armyworm, Spodoptera frugiperda (J. E. Smith) (Lepidoptera: Noctuidae) is an important crop pest in the U.S. where it causes significant economic loss on numerous crops annually (Nagoshi 2009). Across the southeastern states, S. frugiperda is an important pest of corn and, in fact, is the most important whorl-feeding insect pest, especially in late-planted corn (Davis et al. 1996). Resistance to S. frugiperda has been studied extensively, and a series of corn germplasm lines conferring S. frugiperda resistance have been developed at Mississippi State, MS (Brooks et al. 2007), and Tifton, GA (Wiseman et al. 1996) for the southern states.
Recent research efforts have been devoted to identifying and developing corn germplasm that confers resistance to multiple insect pests at various crop growth stages. The possibility of developing resistance to multiple whorl- and ear-feeding insect species has been examined for major pests in the Midwest (Wilson et al. 1995a; Abel et al. 2000a; Abel et al. 2000b). Wilson et al. (1995a) evaluated 11 maize accessions from Peru that were previously found to be resistant to leaf feeding by first-generation European corn borer, Ostrinia nubilalis (Hübner) (Lepidoptera: Crambidae). That field evaluation identified new genetic resources for multiple insect resistance, including resistance to stalk boring by second-generation O. nubilalis, the sugarcane borer, Diatraea saccharalis (F.) (Lepidoptera: Crambidae), and the southwestern corn borer, Diatraea grandiosella Dyar (Lepidoptera: Crambidae); foliar feeding by S. frugiperda; root feeding by the western corn rootworm, Diabrotica virgifera virgifera LeConte (Coleoptera: Chrysomelidae); and ear-feeding by the corn earworm, Helicoverpa zea (Boddie) (Lepidoptera: Noctuidae). Abel et al. (2000a) examined another 15 experimental maize lines against 4 lepidopteran pests (H. zea, S. frugiperda, D. grandiosella, and D. Saccharalis) in the midwest and southern states. The 15 lines were developed from crosses between the Peruvian maize lines and the U.S. midwest corn belt adapted inbred lines. Four inbred lines (i.e., ‘100-R-3’, ‘116-B-10’, ‘81-9-B’, and ‘107-8-7’) were identified as new resources for developing breeding populations against each of the 4 insect pests (Abel et al. 2000a). Abel et al. (2000b) also evaluated those same 15 experimental maize lines against O. nubilalis, and D. virgifera virgifera in the midwest region. All 15 experimental lines showed resistance to leaf feeding by O. nubilalis, and 11 of them showed resistance to leaf sheath and collar feeding by O. nubilalis. None of the lines showed any antixenosis (or non-preference) with respect to O. nubilalis oviposition or D. virgifera virgifera root feeding. Although DIMBOA (2,4-dihydroxy-7-methoxy-1,4-benzoxazin-3-one) is well documented to confer resistance to foliar feeding by first generation O. nubilalis, all of these lines had low levels of DIMBOA, suggesting that this compound might not be directly involved in resistance to O. nubilalis in these lines.
High levels of corn silk maysin have been considered an important phenotypic trait that confers resistance to ear-feeding corn earworm (Wilson et al. 1995b; Widstrom & Snook 2001; Ni et al. 2008). After examining 94 CIMMYT corn inbred lines with varying levels of silk maysin, Ni et al. (2008) determined that 10 of them conferred resistance to multiple ear-feeding insects in the southeastern Coastal Plain region. In addition, Ni et al. (2007) examined another 10 corn inbreds and 10 experimental hybrids, and identified 2 inbreds and 2 hybrids showing resistance to multiple ear-feeding insects, including H. zea, maize weevil, Sitophilus zeamais (Motschulsky) (Coleoptera: Curculionidae), brown stink bug [Euschistus servus (Say)], and southern green stink bug [Nezara viridula (L.)] (Heteroptera: Pentatomidae).
In general, limited progress has been made on developing corn inbred lines showing resistance to both whorl- and ear-colonizing insects and diseases. After evaluating corn germplasm resistance to multiple ear-feeding insects (Ni et al. 2007, 2008), we expanded recent evaluations to including multiple insect resistance/susceptibility over varying corn plant tissues (root, leaf, and ear) throughout different growth stages (i.e., vegetative versus reproductive growth).
The objective of the present study was to determine whether corn germplasm lines resistant to ear- and root-feeding insects would confer resistance to foliar feeding by S. frugiperda at vegetative growth stages. Resistance was assessed by visual ratings of S. frugiperda feeding injury and natural enemy profiles at the whorl stage in 8 selected corn inbred lines that possessed known levels of resistance/susceptibility to root-feeding D. virgifera virgifera and ear-feeding H. zea.
MATERIALS AND METHODS
Plants and Insects
The 8 corn germplasm lines known to be resistant and/or susceptible to at least 1 of 3 different insects (D. virgifera virgifera, H. zea, and S. frugiperda) are listed in Table 1. The 2 newly-selected inbred lines ‘FAW7061’ and ‘FAW7111’ were derived from the ‘GT-FAWCC(C5)’ population (Wiseman et al. 1996) after being self-pollinated for 6 generations. Spodoptera frugiperda neonate larvae used in this study in 2008 were from a laboratory colony maintained in the insectary at the Crop Protection and Management Unit, USDA-ARS, Tifton, Georgia. In 2009, the neonate larvae used for manual infestation were from the Corn Host Plant Resistance Research Unit, USDA-ARS, Mississippi State, Mississippi. The fall armyworm colonies at both locations originated from field-collected insects, and have been maintained on a pinto bean diet (Lynch et al. 1989) for over 10 yr with frequent fusion with field-collected insects.
Manual S. frugiperda Infestation and Injury Rating
The experimental plants used in this field study were infested individually with 15–20 S. frugiperda neonate larvae when the plants were at the 6-leaf (or V6) stage using the protocol previously described by Davis et al. (1996). The levels of insect injury were rated using the mean injury level of all 15–20 plants per experimental plot (5 × 1 m2) 7 and 14 d after infestation using a scale of 1–9 (see Davis et al. 1992 and Smith et al. 1994). Briefly, 1 = no damage or few pinholes; 2 = few short holes (also known as shot holes) on several leaves; 3 = short holes on several leaves; 4 = several leaves with short holes and a few long lesions; 5 = several holes with long lesions; 6 = several leaves with lesions < 2.5 cm; 7 = long lesions common on one half of the leaves; 8 = long lesions common on one half to two thirds of leaves; and 9 = most leaves with long lesions. The S. frugiperda injury rating was conducted without information about germplasm entry assigned for an experimental plot to avoid biased ratings for any of the germplasm entries. The insect injury ratings were recorded per experimental plot based on overall visual assessment of S. frugiperda injury under the field conditions.
EIGHT GERMPLASM LINES USED IN THE EVALUATION OF S. FRUGIPERDA RESISTANCE.
Predator Survey Protocols
Both predator types and/or species from each experimental plot were recorded 7 d after S. frugiperda infestation, because predators might be differentially attracted to the corn germplasm lines and they were abundant in the corn fields when the plants were at 6–9 leaf stages (V6–V9). Data were collected by careful field counts of all predators on every plant per experimental plot (5 × 1 m2) with as little disturbance to the predator activities and plants as possible. Field counts of predators have been described as the best sampling method of choice for examining predators in sweet corn fields (Musser at al. 2004). No parasitoids were observed during the visual surveys in both years. The predator survey was conducted within 24 h after the 7 d S. frugiperda injury rating in both years. The plant parts examined for predators included whorls, leaf blades, leaf sheaths and stalks to detect accessible fast-moving predators (like Orius insidiosis and Geocoris spp.) during the survey. The survey of predators in all experimental plots was conducted diurnally between 1000 h and 1700 h EDT when the predators were active and the plants were without dew. As previously described for the S. frugiperda injury rating, the predator survey was also conducted without knowledge of the germplasm entry to avoid biasing observations pertaining to a given experimental plot or germplasm entry.
Experimental Design and Data Analysis
The experiment utilized a randomized complete block design with the 8 corn germplasm lines as treatments, and four replications as the blocking factor to minimize the influence of soil and other environmental factors on plant development. The 8 corn germplasm entries were planted adjacent to one another without buffer rows (zones) to determine their attraction to predators where choices were provided, and to avoid dilution of predator attraction by planting extra rows of corn plants as buffer areas. Although there were no buffer areas between experimental plots (5 ×1 m2 with 15–20 plants), the edge of the field was surrounded by a border row of the commercial corn hybrid ‘DK6410’ to reduce edge effect on natural infestation of fall armyworm and predator distribution in the experimental plots. Each experiment conducted in 2008 and 2009 was considered a separate trial. The data on feeding injury by S. frugiperda were analyzed using analysis of variance (PROC MIXED procedure) followed by Fisher's Protected LSD test (α = 0.05) (SAS Institute 2003). The data for predator diversity and abundance were analyzed using cluster analysis (PROC CLUSTER procedure) (SAS Institute 2003). Briefly, cluster analysis is a generic term for many techniques that have the common goal to determine whether a multivariate data set contains distinct groups or clusters, and if so, finding which of the observations belong in the same cluster (Der & Everitt 2009). The correlation between S. frugiperda injury ratings and predator profiles was also assessed using the PROC CORR procedure (SAS Institute 2003).
Fall Armyworm Injury Ratings
Leaf injury ratings were significantly different among the 8 germplasm lines 7 d and 14 d after the infestation (Table 2). The 7 d ratings were not different between 2008 and 2009, but the 14 d injury ratings varied significantly between the 2 years. The germplasm entry × year interaction significantly affected both 7 d and 14 d injury ratings (Table 2), which reflected the variation caused by weather conditions from year to year for such field studies. Because the entry × year interaction was significant, the 7 d and 14 d rating data from 2008 and 2009 were presented separately (Figs. 1A to 1D). When compared to the S. frugiperda injury ratings of the susceptible control ‘Ab24E’, the resistant control (Mp708’) showed lower ratings in all evaluations, except at 7 d post-infestation in 2009 (Fig. 1C).
ANALYSIS OF VARIANCE TABLE OF S. FRUGIPERDA INJURY RATINGS AND PREDATOR ABUNDANCE ON EIGHT CORN GERMPLASM LINES.*
While S. frugiperda injury ratings varied between 2008 and 2009 at the 7 d post-infestation evaluations (Figs. 1A and 1C), the 14 d post infestation ratings were relatively consistent (Figs. 1B and 1D). Four germplasm lines, i.e., ‘CRW3(S1)C6’, ‘FAW7061’, ‘FAW7111’, and ‘Mp708’ had significantly lower S. frugiperda injury ratings than the susceptible control, ‘Ab24E’. In particular, the lower S. frugiperda injury ratings were recorded on ‘CRW3(S1)C6’, ‘FAW7061’, and ‘FAW7111’ 14 d post-infestation than the susceptible control, ‘Ab24E’ in 2008 (Fig. 1B), but not at the 7 d post-infestation assessment in 2008 (Fig. 1A). Also, S. frugiperda injury ratings on ‘CRW3(S1)C6’, ‘FAW7111’, and ‘Mp708’ at the 14 d post-infestation evaluation in 2009 (Fig. 1D) were lower than the 7 d post-infestation evaluation (Fig. 1C). The reduced S. frugiperda injury ratings on the 14 d evaluation compared to the 7 d evaluation may have been related to predation of the S. frugiperda larvae by predators in the experimental plots. Thus, the predator diversity and abundance were further examined in these experimental plots 7d after the infestation.
Ten predator species were recorded in the experimental plots (Table 3). The 5 lady beetle species (Coleoptera: Coccinellidae) included the convergent lady beetle, Hippodamia convergens Guérin-Meneville, the pink spotted lady beetle, Coleomegilla maculata (De Geer), the multicolored Asian lady beetle, Harmonia axyridis (Pallas), the seven-spotted lady beetle, Coccinella septempunctata L., and a Scymnus sp. In addition, hooded (or flower) beetles, Notoxus spp. (Coleoptera: Anthicidae) was also recorded in the experimental plots. Earwigs (Dermaptera) were not differentiated by species when recorded during the surveys, but later identified as Labidura riparia (Pallas) (Labiduridae) and Doru taeniatum (Dohrn) (Forficulidae). Three taxa of heteropteran predators were also recorded in the experimental plots including the insidious flower bug, Orius insidiosus (Say) (Heteroptera: Anthocoridae), the big-eyed bug, Geocoris spp. (Heteroptera: Geocoridae), and the damsel bugs, Nabis spp. (Heteroptera: Nabidae).
Coleomegilla maculata was the most abundant predator observed, whereas C. septempunctata was the least abundant species (Table 3). Because both predator taxa and the number of each taxon were equally important in evaluating the attraction of the corn germplasm lines to different types of predators, cluster analysis was utilized for assessing predator diversity and abundance on the 8 corn germplasm lines. The number of C. septempunctata, and earwigs were significantly different among the 8 germplasm lines, whereas other predators were equally abundant among the germplasm lines (Table 2). Three predators (C. maculata, H. axyridis, and Geocoris spp.) and the total number of predators were significantly different between 2008 and 2009 (Table 2). Also, the number of C. septempunctata (C7) and Notoxus spp. was influenced by the corn germplasm line × year interaction (Table 2).
PREDATOR ABUNDANCE (MEAN ± SEM) AND DIVERSITY PER 5 M2 (15–20 PLANTS) FROM EIGHT CORN GERMPLASM LINES IN 2008 AND 2009 (N = 4).*
Identification of Spodoptera frugiperda Resistance Using Injury Rating and Predator Survey Data
From cluster analysis of the corn germplasm lines, 4 clusters were extracted each with an eigenvalue >1 (ranging between 1.2 and 3.8), which contributed to 87% of the total variance. Cluster analysis using the combined S. frugiperda injury rating and predator data (Fig. 2) aligned with previous identification of S. frugiperda resistance using only S. frugiperda injury rating data as shown in Figs. 1A to 1D. ‘Mp708’ and ‘FAW7061’ were in the same cluster (Fig. 2), which were S. frugiperda resistant. Rootworm-resistant ‘CRW3(S1)C6’ showed S. frugiperda resistance (Figs. 1A to 1D). In addition, ‘CRW3(S1)C6’ was separated from the other 7 germplasm lines (Fig. 2) because the most predators were observed on ‘CRW3(S1)C6’, as shown in Table 3, and ‘CRW3(S1)C6’ also showed S. frugiperda injury (Figs. 1A to 1D). In particular, more earwigs and C. septempunctata were also recorded on the western corn rootworm resistant line, ‘CRW3(S1)C6’, than on the other 7 germplasm lines (Tables 2, and 3). In contrast, ‘EPM6’ and the susceptible control, ‘Ab24E’, had the highest S. frugiperda injury ratings (Figs. 1A to 1D) and the fewest predators were recorded in this cluster in both years (Table 3 and Fig. 2).
Correlation between S. frugiperda Injury Ratings and Predator Abundance
Based on the combined two-year data of the 8 germplasm lines, the 2 (7d and 14d) ratings of S. frugiperda injury was positively correlated, whereas the S. frugiperda injury ratings were negatively correlated to the number of hooded beetles (Table 4). The 14 d S. frugiperda injury ratings were positively correlated to C. maculata. The correlation coefficients among the 10 predator species varied (Table 4). The total number of predators was positively correlated to C. maculata, C. septempunctata, H. axyridis, Geocoris spp., and O. insidiosus (Table 4), but not to the others. These 5 species were the most common predators in the experimental plots.
Positive and negative correlations were detected among predators (Table 4). Coleomegilla maculata abundance was positively correlated to C. septempunctata, Nabids, and Geocoris spp. In addition, C. septempunctata was also positively correlated with O. insidiosus. Notoxus spp. were positively correlated with Scymnus spp. and ear-wigs. Cluster analysis among the diversity and abundance of predators showed one main cluster with an eigenvalue >1 (i.e., 6.8) that contributed to 85% of the variation. This cluster analysis showed that the most abundant predators across the 8 corn germplasm lines at 7 d after infestation were C. maculata and O. insidiosus in the same cluster (Fig. 3), whereas the least abundant predators in the same cluster were H. convergens, C. septempunctata, Scymnus sp., Nabis spp., and earwigs (Fig. 3). The abundance of the other 8 predator species varied significantly between the 2 years. In particular, O. insidiosus was abundant in 2008, but less so in 2009 (Table 3). In contrast, all species of earwigs and the hooded beetles were more abundant in 2009 than in 2008 (Table 3).
The current study demonstrated that the D. virgifera virgifera-resistant corn germplasm line, ‘CRW3(S1)C6’ conferred fall armyworm resistance. Previous reports on multiple insect resistance were mainly limited to similar plant tissues, such as multiple leaf-feeding insects (Wilson et al. 1995a; Abel 2000a), and multiple ear-feeding insects and ear-colonizing diseases (Ni et al. 2007; Ni et al. 2008). The present study also showed that the 2 newly-developed partial inbred lines, i.e., ‘FAW7061’ and ‘FAW7111’ derived from a previously released population, ‘GT-FAWCC(C5)’, were resistant to S. frugiperda feeding compared to the resistant ‘Mp708’ and the susceptible control, ‘Ab24E’, although ‘FAW7061, had less S. frugiperda injury than ‘FAW7111’. In particular, the rootworm resistance, i.e., ‘CRW3(S1)C6’ will be useful in developing S. frugiperda and multiple insect pest resistance in new corn germplasm adapted to the southern U.S. states, although the phenotypic traits and underlying resistance mechanisms need further elucidation.
CORRELATION COEFFICIENT BETWEEN S. FRUGIPERDA INJURY RATINGS AND PREDATOR ABUNDANCE ON EIGHT CORN GERMPLASM LINES (N = 63)*.
The predator profiles varied among the 8 germplasm entries at whorl stage under the field conditions in the present study. ‘CRW3(S1)C6’ had the largest number of predators at the whorl stage, whereas ‘Ab24E’ had the fewest. Phenotypic traits, e.g., flowering time, leaf color, and leaf trichome density, may interfere in plant attraction to predators, but the impact might be limited in this study, which was only conducted at the V6–9 (6–9 leaf) stages of vegetative growth. For instance, although ‘CRW3(S1)C6’ was phenologically an early-flowering line compared to the other entries in Tifton (Ni, unpublished data), any influence of flowering time would be minimal because the ratings and predator survey was conducted before tasseling.
Prey species of the observed predators were not apparent. A total of 10 different predator species were recorded without any apparent infestations of either aphids or spider mites. Besides the manual infestation of corn plants with S. frugiperda neonate larvae, the only abundant herbivores in the experimental plots were thrips (Ni, personal observation). Predation efficacies of all 10 predator species on thrips are not well known, although some are noteworthy thrips predators, e.g., O. insidiosus (Dicke & Jarvis 1962). Coleomegilla maculata and O. insidiosus were the most abundant species observed in 2008 and 2009, while C. septempunctata was the least common species on the 8 corn germplasm lines. Similarly, Hoballah et al. (2004) noted that both C. maculata and O. insidiosus were abundant on corn plants at 4–5 leaf stages (V4 to V5) between Jan and Feb 2000 in Mexico, and Sueldo et al. (2010) reported that earwigs were effective predators for fall armyworm larvae in Argentina. Our findings indicated that predators are common at the whorl stage in corn fields. The abundant natural enemies recorded on the corn plants might have been attracted to either constitutive corn plant volatiles or to the corn plant volatiles synthesized in response to S. frugiperda-injury, because our sampling was conducted 7 d after the manual insect infestation with the S. frugiperda neonates. Several natural enemies, i.e., C. septempunctata and earwigs, exhibited differential responses to the corn germplasm lines, suggesting possible germplasm-specific interactions. The chemical ecology and general significance of these phenomena observed in the field should be further elucidated.
Differential responses of natural enemies to corn plants could be further examined and utilized as a favorable trait in corn breeding programs intended to reduce foliar injury by S. frugiperda and other pests (particularly aphids) in corn at vegetative growth stages. This ecologically-based resistance has been termed pseudo-resistance (Painter 1951; Panda & Khush 1995). In recent years, a number of studies have demonstrated host plant volatile-mediated insect herbivore-natural enemy interactions (De Moraes et al. 2001; Ryan 2001; Ode 2006; Smith 2010; Hare 2011). Diel pattern of plant volatile profiles may differentially serve to recruit natural enemies diurnally and repel pest oviposition nocturnally (Ryan 2001). The utilization of natural enemies as an extension of conventional (or constitutive) plant defenses against insect herbivory still needs to be further examined and elaborated (Ode 2006). At the same time, Hare (2011) also pointed out that variations in plant volatile blends might be influenced by both abiotic and biotic factors under field conditions, which would in turn alter the tri-trophic interactions among host plants, herbivores, and natural enemies. It is necessary to utilize the techniques from evolutionary quantitative genetics to test the hypotheses related to volatile production of plants in response to herbivory damage under natural or field conditions (Hare 2011).
Predators in this study were surveyed on only 1 date each year, and during a restricted time of the day, using methods similar to those described by Musser et al. (2003). Additional samples over multiple dates and times of day and night would add additional insights into the plant-pest-natural enemy relationships. Further in-depth ecological studies are needed to decipher the roles of predators in crop pest suppression in agricultural ecosystems (Furlong & Zalucki 2010). Understanding these ecologically-based dynamics in host plant-pest-natural enemy interactions could lead to the utilization of plant volatile-mediated insect ecology (or natural enemy attraction) to reduce pest populations and, in turn, to reduce crop losses by insect herbivory and mycotoxin contamination. It is likely that the similarity of S. frugiperda injury ratings on ‘CRW3(S1)C6’ between 7 d and 14 d was the result of predation of S. frugiperda larvae by the predators that were abundant on this line. Utilizing ecological genetics of corn plants to reduce yield and quality losses from and insects and diseases by reducing insect herbivory and attracting natural enemies could be one of the effective tactics of corn breeding program in the long-term. The present study serves as a baseline for our corn breeding program to further examine multiple insect resistance, including foliar-, root-, and ear-feeding insects at various growth stages, and mycotoxin reduction in the southeastern Coastal Plain region of the U.S.
Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. We thank Russell Scarborough and Walter Burgess of the Crop Protection and Management Research Unit, USDA-ARS, Tifton, Georgia, Charles Mullis of the Crop Genetics and Breeding Research Unit, USDA-ARS, Tifton, Georgia, and Susan Wolf of the Corn Host Plant Resistance Research Unit, USDA-ARS, Mississippi State, Mississippi corn for their assistance in performing these trials. We also express our gratitude to T. E. Cottrell, USDA-ARS, Byron, Georgia and Y. C. Zhu, USDA-ARS, Stoneville, Mississippi for their critical reviews of the earlier version of the manuscript.