The objective of these studies was to assess the degree to which bean leaf beetle, Cerotoma trifurcata (Forster), will feed on cucurbits. In 2003, we documented an infestation of C. trifurcata in a commercial pumpkin field near Rosemount, MN, USA. To evaluate C. trifurcata feeding on cucurbits, we conducted laboratory no-choice and choice test feeding studies. In the laboratory, C. trifurcata fed most heavily on cotyledon-stage cucumber plants, followed by pumpkin and squash. With soybean plants present, C. trifurcata still fed on cucumber plants. However, C. trifurcata appeared to prefer soybeans until the quality of the soybean plants was diminished through feeding damage. This is the first known report of C. trifurcata feeding on cucurbits. The pest potential of C. trifurcata in cucurbit cropping systems should be further evaluated.
The bean leaf beetle, Cerotoma trifurcata (Forster), is a pest of leguminous crops throughout the eastern United States (Kogan et al. 1980). Adults feed on above ground plant tissue, such as leaves, stems, and pods (Kogan et al. 1980). C. trifurcata can also vector various pathogens, such as bean pod mottle virus, cowpea mosaic virus, and southern bean mosaic virus (Walters 1969). In Minnesota, C. trifurcata completes one generation per year, with peaks in adult abundance occurring in late-May to early-June and late-August (Loughran and Ragsdale 1986). In the Midwestern United States, the abundance of C. trifurcata has increased over the past several years (Bradshaw and Rice 2003). Concurrently, the severity and importance of the been leaf beetle as a pest in leguminous crops has increased (e.g., Hutchison et al. 2002; Hutchison et al. 2003).
Despite being known to feed primarily on legumes, C. trifurcata will also feed on non-legumes. Feeding on non-legume hosts appears to be an early season phenomenon. Helm et al. (1983) documented feeding on wild non-legume plants, such as Urtica dioica, Laportea canadensis, and Euonymous atropurpurea. Metcalf and Metcalf (1993) reported that they may feed on corn, Zea mays. However, Zeiss and Pedigo (1996), under laboratory conditions, observed no feeding on grasses, such as Z. mays, oats, Avena sativa, and wheat, Triticum aestivum. While C. trifurcata may possibly feed on non-leguminous crops (e.g., Metcalf and Metcalf 1993), Tallamy et al. (1991) have shown that cucurbitacins from cucurbits serve as a feeding deterrent.
While monitoring for striped cucumber beetle, Acalymma vittatum (Fabricius), in pumpkin fields near Rosemount, MN, we observed a high infestation of C. trifurcata and feeding damage on cotyledon-stage pumpkin plants. Cucurbits are grown on ca. 1,900 ha in Minnesota, for a total value of $8,477,000 (Hutchison and O'Rourke 2002). Historically, A. vittatum has been the most important pest in Minnesota cucurbits (Ives and Walters 1985; Subramanyam et al. 1993), with the early growth stages of the plants being most vulnerable to attack (Brewer et al. 1987; Burkness and Hutchison 1998). The objective of the studies presented in this paper was to assess the degree to which C. trifurcata feeds on cucurbits.
Materials and Methods
The pumpkin variety, ‘Magic Lantern’, was planted on 25 May 2003 into a 4-ha field near Rosemount, MN. The field was planted with soybeans the year before. Pumpkin plants began emerging on 6 June. Cotyledon stage volunteer soybean plants were also present in the field at this time. On 13 June, we began using whole-plant visual inspections to monitor the field for early season pests. Subsequent whole-plant visual inspections were conducted on 16, 17, 18, 20, 23, and 24 June. On each sample date, 60 to 140 plants were examined, and counts of C. trifurcata were recorded.
No-Choice Feeding Study
To determine if C. trifurcata will feed on cucurbits, we conducted a no-choice feeding study and a choice test feeding study in the laboratory. For the no-choice feeding study, seed from pumpkins (‘Magic Lantern’), slicing cucumbers (‘Marketmore 76’), and squash (‘Turks Turbin’) were planted separately into 3.8 liter pots containing universal soil mix. Only one seed was planted in each pot. The pots were held in a growth chamber at 27° C with a 16:8 (L:D) cycle. When the plants reached the cotyledon stage, a small cage was placed over each potted plant. The cages were made from two-liter clear plastic bottles with the bottoms cut off. The large opening at the bottom of the bottles was buried about 2 cm into the soil. Many pinholes were punched into the bottles to allow ventilation. Each cage was infested with ten C. trifurcata adults, which were collected from a soybean field at the Rosemount Research and Outreach Center, University of Minnesota, Rosemount, MN. Until infestation, C. trifurcata were held in a growth chamber at 27° C with a 16:8 (L:D) cycle, and fed soybean foliage. The cucumbers, pumpkins, and squash were each replicated in six pots. At 72 h post-infestation, the cages were removed, and the plants were assessed for damage. Damage was recorded as follows: percentage of cotyledon area with holes going completely through the cotyledons; percentage of the upper surface of the cotyledons with surface scarring; percentage of the lower surface of the cotyledons with surface scarring; presence or absence of damage to the stem; and presence or absence of damage to the small, folded first true leaf. Data for each injury type were arc-sine square root transformed, and analyzed with analysis of variance and the Ryan-Einot-Gabriel-Welsch multiple range test (SAS, 1995).
Choice Test Feeding Study
For the choice test feeding study, a single seed of a slicing cucumber (‘Marketmore 76’) and a soybean (‘M96-133 151’) were planted into 3.8 liter pots containing universal soil mix. Plants were grown under the conditions described for the no-choice feeding study. When the plants reached the cotyledon stage, a small cage, as in the no-choice feeding study, was placed over the pair of plants within each pot. Each cage was then infested with 10 C. trifurcata adults that were collected from the same location as the beetles used in the no-choice feeding study. Beetles were maintained under the same conditions until infestation as in the no-choice feeding study. This feeding trial was replicated in nine pots. Plants were visually inspected at 4, 8, 18, 24, and 72 h, and the number of beetles per plant and presence or absence of damage to the plants was recorded. At 24 and 72 h post-infestation, plants were more thoroughly assessed for damage. Damage to the cucumber and soybean plants was recorded as in the no-choice feeding study. Data on the abundance of C. trifurcata on plants within the cages were analyzed by time period using a two-sample t-test on the difference between the number of beetles on soybean and cucumber plants. The mean difference was compared to zero, which would indicate no preference. Data on the occurrence of damage on the caged plants was analyzed by time period using a two-sample t-test on the mean difference of arc-sine square root transformed data on the percent of plants with damage for soybean versus cucumber. Statistical analyses were not necessary at 4, 8, and 72 h post-infestation, because there was no variability for the percentage of damaged plants for cucumber and soybean. Data on the damage ratings conducted at 24 and 72 h post-infestation were analyzed using a two-sample t-test on the mean difference of arc-sine square root transformed data on the percent damage to soybean and cucumber plants for each injury type. The mean difference was compared to zero, which would indicate no difference in the amount of damage.
C. trifurcata adults were present on the pumpkin plants with associated damage at each sample date (Fig. 1, 2, 3). From 13 June to 24 June 2003, the abundance of C. trifurcata decreased from 0.36 to 0.03 individuals per plant (Fig. 1). In addition, A. vittatum was present on the plants at each sample (RLK unpublished data).
No-Choice Feeding Study
Cucumber plants were generally more heavily damaged than squash plants, with pumpkin plants having an intermediate amount of damage (Table 1; Fig 4; 5). The percentage of plants with stem damage was significantly greater for cucumber and pumpkin plants compared to squash plants (F = 5.00, df = 2, 15, P = 0.022), while the percentage of plants with damage to the first true leaf did not differ significantly (F = 0.83, df = 2, 15, P = 0.45) (Table 1). Significantly more cotyledon area had distinct holes on cucumber plants than pumpkin or squash plants (F = 11.88, df = 2, 15, P = 0.0008) (Table 1). The percentage of the lower (F = 16.86, df = 2, 15, P = 0.0001) and upper (F = 9.21, df = 2, 15, P = 0.0025) surface of the cotyledons scarred for cucumber and pumpkin plants was significantly greater than that of squash plants (Table 1).
Choice Test Feeding Study
More C. trifurcata were found on soybean than on cucumber (Fig. 6A). The difference in abundance on soybean compared to cucumber plants was significant at all observations, except for 72 h post-infestation (4 h: t = 4.63, df = 8, P = 0.0017; 8 h: t = 5.12, df = 8, P = 0.0009; 18 h: = 6.4, df = 8, P = 0.0002; 24 h: t = 5.38, df = 8, P = 0.0007; 72 h: t = 0.61, df = 8, P = 0.56) (Fig. 6A). The abundance of C. trifurcata on soybeans decreased from 4.11 ± 0.89 (mean ± SEM) individuals per plant at 4 h post-infestation to 1.89 ± 0.48 individuals per plant at 72 h post-infestation (Fig. 6A). Abundance on cucumber ranged from 0.11 ± 0.11 to 0.33 ± 0.24 individuals per plant from 4 to 24 h post-infestation, respectively (Fig. 6A). At 72 h post-infestation, abundance on cucumber reached a maximum of 1.44 ± 0.34 individuals per plant. C. trifurcata will feed on cucumber plants when soybean plants are present (Table 2; Fig. 6B). However, they fed on soybean plants nearly immediately, with 100 % of soybean plants being damaged at 4 h post-infestation (Fig. 6B). Cucumber plants were not damaged so rapidly. Only 22.22 ± 14.70 % of plants were damaged at 18 and 24 h post infestation (Fig. 6B). However, at 72 h post infestation, 100 % of the cucumber plants were damaged (Fig. 6B). The percentage of damaged plants was significantly greater for soybeans than cucumber at all observations, except 72 h post infestation (8 h: t = 5.29, df = 8, P = 0.0007; 18 h: t = 5.29, df = 8, P = 0.0007) (Fig. 6B).
All types of injury, due to C. trifurcata feeding, were more intense on soybean than on cucumber plants (Table 2). At 24 h post-infestation, significantly more soybean than cucumber plants had incurred stem damage (Table 2). The percentage of cotyledon area with distinct holes (t = 3.34, df = 8, P = 0.01) and percentage of the lower surface area of the cotyledon scarred (t = 13.62, df = 8, P < 0.0001) was greater on soybean compared to cucumber plants (Table 2). At 72 h post-infestation, significantly more soybean than cucumber plants had stem damage and damage to the first true leaf (Table 2). The percentage of cotyledon area with distinct holes on soybean and cucumber plants did not differ significantly (t = 1.82, df = 8, P = 0.11) (Table 2). The percentage of surface area scarred on the upper (t = 10.10, df = 8, P < 0.0001) and lower surfaces (t = 21.76, df = 8, P < 0.0001) of the cotyledons was significantly greater on soybean than on cucumber plants (Table 2).
In mid to late-June, C. trifurcata was observed on pumpkin plants at the first true-leaf growth stage with associated damage to the plants in the field (Fig. 2, 3). Adults rarely migrate from one field to another within a growing season (Waldbauer and Kogan 1976), and over winter under stubble in the field or in wooded areas adjacent to soybean fields (Lam and Pedigo 2000). The pumpkin field that we sampled was likely at high risk to being infested by C. trifurcata, because it was planted to soybeans the year before, contained soybean stubble and volunteer soybeans, and was surrounded by wooded habitat on three sides. The observed decline in the abundance of C. trifurcata (Fig. 1) temporally matched the decline of over wintered adults observed by Loughran and Ragsdale (1986), suggesting that the decline was due to natural population dynamics rather than repellency from the pumpkin field.
Our laboratory studies confirmed that C. trifurcata will feed on cucurbits. Damage to cucurbits caused by C. trifurcata appeared similar to damage caused by A. vittatum, with feeding occurring between the veins of the leaves, consuming either a layer of tissue from the top or bottom the leaf; eventually feeding may result in a hole through the leaf (Burkness 1996). Cucumber plants appeared to be more susceptible to C. trifurcata than either pumpkin or squash (Table 1; Fig. 4, 5). Differential susceptibility of the various cucurbits to feeding may have been due to differences in cucurbitacin concentration (e.g., Tallamy et al. 1991) or differences in size and thickness of the leaf tissue (Fig. 4). C. trifurcata fed on cucumber plants, albeit a small amount, when soybean plants were present, which corroborates our observation of their presence on pumpkin plants in the field when volunteer soybean plants were present.
To our knowledge, this is the first report of C. trifurcata feeding on cucumber, squash, or pumpkin. However, their pest potential for cucurbits remains uncertain and needs further investigation. Burkness and Hutchison (1998) found that for cotyledon to first true leaf stage cucumbers, only 10-15 % defoliation was necessary to cause a significant yield loss. In our no-choice feeding study, densities of 10 C. trifurcata per cucumber plant resulted in about 19 % of the leaf area having distinct holes. With maximum densities of 0.36 C. trifurcata per plant observed in the field (Fig. 1), it seems unlikely that feeding will result in economic losses. However, densities of ten C. trifurcata per plant did result in severe damage to some of the cucumber plants, which would likely lead to plant death (Fig. 5). C. trifurcata may be even less of a threat to pumpkin and squash plants, due to the ability of pumpkin and squash plants to tolerate higher levels of defoliation (Hoffmann et al. 2000), and the relatively low levels of feeding on pumpkin and squash plants (Table 1; Fig. 4). Therefore, additional feeding studies, including the relationship between cucurbitacin concentration and feeding damage should be examined.
We thank T. Galvan and K. Bennett for field assistance, Pahl's Market for allowing us to sample their pumpkin field, the laboratory of D. Ragsdale for providing soybean seed, and S. Burkness for assistance with digital images. We also thank R. Venette and R. Moon for assistance with the statistical analyses.