Abstract
Species and hybrids of Tilia are widely used in urban landscapes and as street trees. Planting Tilia species that are less preferred by the Japanese beetle (Popillia japonica Newman) would greatly minimize the need for insecticide sprays, which would in turn reduce environmental contamination, human exposure, and management costs. Thus, a study was initiated to evaluate horticulturally desirable Tilia spp. for feeding preference by the adult Japanese beetle. A secondary objective of the study was to determine if leaf pubescence is a factor in feeding preference. The Asian species of T. petiolaris, T. oliveri, T. chinensis, and T. japonica show promise for partial resistance to Japanese beetle. Three North American species, T. americana, T. heterophylla, and T. caroliniana, also look promising along with the pubescent European species of T. tomentosa and T. tomentosa ‘Erecta’.
Since its introduction into the United States around 1916, the Japanese beetle (Popillia japonica Newman) has become one of the most destructive and difficult pests of nursery crops and landscape plants to manage. Adult beetles feed on approximately 300 species of wild and cultivated plants including, but not limited to, members of Rosaceae, Malvaceae, Vitaceae, Polygonoceae, Aceraceae, Ulmaceae, and Salicaceae (Fleming 1972; Ladd 1987; Spicer et al. 1995). Highly preferred hosts include lindens (Tilia spp.), sassafras (Sassafras albidum [Nuttall] Nees), and purple plum (Prunus cerostifera Erhart), which all may be completely defoliated by adult beetles.
Host plant resistance for P. japonica is poorly documented (Raupp et al. 1992), with susceptible or resistant plant lists being compiled from anecdotal descriptions and field observations (Hawley and Metzger 1940; Fleming 1972; Spicer et al. 1995). In addition, feeding studies conducted by Ladd (1987, 1989) have revealed considerable variation in feeding preference by P. japonica among plant species listed by Fleming (1972) as being either favored hosts, minor hosts, or nonhost species (Spicer et al. 1995).
Information on P. japonica feeding preference among cultivars or varieties of common hosts also is limited. Feeding damage ratings for P. japonica were conducted by Ranney and Walgenbach (1992) on selected cultivars of birch, cherry, and crabapple during one growing season. In addition, Spicer et al. (1995) studied the resistance of 42 Malus cultivars for defoliation by the Japanese beetle over several growing seasons. More recently, Potter et al. (1998) examined the relative susceptibility of cultivars of flowering crabapples, lindens, and roses for defoliation by adult Japanese beetles.
Because of the widespread use of Tilia spp. in urban landscapes and as street trees, planting Tilia species that are less preferred by the Japanese beetle would greatly minimize the need for insecticidal sprays, which would in turn reduce environmental contamination, human exposure, and management costs. With these goals in mind, a study was initiated to evaluate potential horticulturally desirable Tilia spp. for feeding preference by the adult Japanese beetle and to determine if leaf pubescence is a factor in feeding preference.
Materials And Methods
No-Choice Laboratory Feeding Tests
No-choice feeding tests were conducted using field-collected adult male and female Japanese beetles during the 1995 and 1997 growing seasons. In 1995, 16 Tilia genotypes, growing in the Tilia collection at The Morton Arboretum (Lisle, Illinois) were evaluated, with Tilia cordata serving as a reference (see Table 1 for a listing of genotypes tested). In 1997, 3 Tilia genotypes (Table 2) also growing in the Tilia collection were tested. Linden species/hybrids and Sassafras albidum (considered a reference) for both studies ranged in height from 8 to 15 m (26 to 49 ft), with diameters at breast height ranging from 5 to 50 cm. (2 to 20 in.). Also in 1997, 7 additional Tilia genotypes growing in 3.8-L (10-gal.) paper pots in The Morton Arboretum tree nursery, 2.5 cm (1 in.) in diameter, and approximately 2 m (6 ft) tall were tested for feeding preference (refer to Table 3 for a listing of genotypes tested). The selection of Tilia genotypes for all feeding tests was based on their horticultural qualities and potential use in landscapes and urban forests. The number of individual trees tested in each study was based on availability, with 1 to 5 trees randomly selected within a given Tilia genotype.
Mean percent leaf area consumed and mean dried fecal pellet weights of adult Japanese beetles feeding on selected Tilia spp. and cultivars in a no-choice laboratory test (July 1995).
Mean percent leaf area consumed and mean dried fecal pellet weights for adult Japanese beetles feeding on selected Tilia spp. and cultivars growing in the Tilia spp. collection in a nochoice laboratory feeding study (July 1997).
Mean percent leaf area consumed and mean dried fecal pellet weights of adult Japanese beetles feeding on selected Tilia spp. and cultivars growing in 3.8-L (10-gal) pots in a no-choice laboratory feeding study (July 1997).
Leaves were collected randomly from the lower half of the trees at all 4 cardinal directions. The leaf samples included the terminal 31 cm (12 in.) of linden branches. Samples consisted of an equal proportion of actively growing and senescent foliage for each tree. Only fully expanded leaves were used. Leaf samples were taken in this way to compensate for leaf quality within trees. Leaf samples were held in cold storage in plastic bags for a maximum of 2 days. Leaves from each test tree were combined for the trials. Depending on availability, 1 tree (1995) to 5 trees (1997) of each species/hybrid were randomly selected and evaluated. With the exception of the lindens growing in the paper pots (1997 test), all lindens were growing close to each other in the Tilia collection at The Morton Arboretum.
Adult beetles used in the no-choice studies were field collected at The Morton Arboretum and at the Chicago Botanic Garden (Glencoe, Illinois). Upon arrival at the laboratory, beetles were placed in clear plastic cages along with fresh crabapple foliage and held under a photoperiod of 16:8 (L:D) hours at approximately 25°C (77°F).
For the 1995 test, 2 unsexed beetles were placed in each of 10 replicate plastic petri dishes, (0.6 × 10 cm [0.2 × 4 in.]) with foliage from the test linden species/hybrids. Petri dishes were examined daily for evidence of feeding and beetle mortality. Foliage was replaced every 2 days. Beetles that died within the first 3 days were replaced to ensure that healthy beetles had an opportunity to feed on test lindens. The petri dishes were placed in clear plastic bags to prevent drying of the foliage and were held under a 16-hour photoperiod at approximately 25°C (77°F). Condensation of water on the lid of the petri dishes indicated a high relative humidity. One tree of every species/hybrid was tested. After 1 week, the foliage was removed from the dishes, marked with a permanent marker as to genotype and dish number, placed in a plant press, and dried in a herbarium drying oven. After drying, leaf area (nearest mm2) was measured using a Licor meter, after which the remaining leaf area after feeding was measured. The measure of acceptance of each genotype was defined by the mean percent leaf area consumed and mean dried fecal pellet weight. Leaf area consumed was calculated by subtracting leaf area remaining after feeding from initial leaf area before feeding and then dividing this value by the initial leaf area before feeding to obtain a percentage of leaf area consumed. Fecal pellets were removed from each dish, labeled, dried, and then weighed (nearest 0.1 mg). The measure of acceptance of each genotype was defined by the mean percent leaf area consumed and mean dried fecal pellet weight.
In 1997, no-choice laboratory tests were performed as described previously except that only 1 unsexed adult Japanese beetle was placed in the petri dish with the candidate foliage. Because of the laborious task of preparing foliage for leaf area measurements, visual estimates of percent defoliation (nearest 10%) were used. All other methods are as described previously.
Multiple-Choice Laboratory Feeding Study, 1997
One unsexed adult Japanese beetle was placed in each of 10 replicate plastic petri dishes (0.6 × 15 cm [0.2 × 5.9 in.]). Six circular leaf discs, 2.5 cm (1 in.) in diameter, each representing 6 different linden genotype choices, were placed into each dish and randomly arranged around the outside perimeter of the dish. Within a given dish, the beetle had access to all foliage sections. The petri dishes were placed in clear plastic bags to prevent drying of the leaf discs and were held under a 16:8 photoperiod (L:D) at approximately 25°C (77°F). The dishes were examined daily for 7 days. The foliage discs were removed each day from the dishes, replaced, and visually examined for the percent of leaf tissue removed (nearest 10%).
Field Defoliation Study, 1997
After all beetle feeding had ceased, a field defoliation survey was conducted for all Tilia species/hybrids evaluated in the no-choice and multiple-choice tests.Depending on availability, 3 to 5 single tree replicates for each Tilia genotype were visually rated for percent defoliation using the following scale: NF = no feeding, VL = very light feeding (1% to 10%), L = light feeding (11% to 25%), M = moderate feeding (25% to 50%), H = heavy feeding (>50%). Trees were growing at The Morton Arboretum and were 2 to 10 m (6.6 to 10 ft) in height with diameters at breast height of 2.5 to 15 cm (1 to 6 in.). A field defoliation survey was not conducted in 1995 because beetle pressure was not high enough to detect differences.
Statistical Analysis
Measures of preference were subjected to analysis of variance (ANOVA) using species/hybrid as the main effect. Percent defoliation for each species/hybrid was arcsine square-root transformed before analysis to correct for normality. Means of significant effect were compared with a Student-Newman-Keuls multiple comparison test. All data are presented as means ± SEM. Data were analyzed using SigmaStat for Windows (Jandel Scientific 1992).
Results and Discussion
No-Choice Laboratory Feeding Study, 1995
Significant differences were observed for mean percent leaf tissue consumed (F = 155.5; df= 15, 159; P < 0.0001) and mean dried fecal pellet weight (F = 7.8; df = 15, 159; P < 0.0001). Overall, T. platyphyllos, T. × euchlora, T. cordata (reference), T. orbicularis, T. mongolica, T. maximowicziana, and T. amurensis had significantly more leaf tissue removed as compared to the least-preferred species, T. americana (Table 1). Tilia tomentosa ‘Erecta’, T. tomentosa, T. heterophylla, T. caroliniana, T. petiolaris, T. oliveri, T. chinensis, and T. japonica were intermediate, with less than 18% of the leaf area consumed (Table 1). In a recent study, Potter et al. (1998) found that Crimean linden (T. × euchlora) and all 4 cultivars (‘Olympic’, ‘Chancellor’, ‘Greenspire’, and ‘Glenleven’) of littleleaf linden (T. cordata) were severely damaged even during years with less beetle activity. By comparison, T. americana ‘Legend’, T. americana ‘Redmond’, and silver linden, T. tomentosa ‘Sterling’ were found by Potter et al. (1998) to have generally sustained less damage than littleleaf or Crimean lindens. Likewise, Sacksteder and Gerhold (1978) reported that T. cordata ‘Greenspire’ had significantly greater defoliation than T. tomentosa.
Overall, mean dried fecal pellet weights reflected the rankings for the mean percentage of leaf area consumed, with species with a greater percentage of leaf area consumed having greater mean dried fecal pellet weights (Table 1). Tilia platyphyllos, T. cordata, T. mongolica, and T. amurensis had significantly greater dried fecal pellet weights (71 to 80 mg), while T. japonica, T. americana, T. petiolaris, T. oliveri, and T. chinensis had significantly lower mean dried fecal pellet weight (18 to 38 mg). The remaining genotypes were moderately preferred with mean dried fecal pellet weights between 48 and 66.7 mg (Table 1).
No-Choice Laboratory Feeding Study, 1997
Overall, individual trees had no significant effect on mean percentage of leaf tissue consumed (F = 13.4; df = 4, 224; P = 0.11) or mean dried fecal pellet weight (F = 8.5; df = 4, 224; P = 0.08). Therefore, the data for all trees for each species/hybrid were pooled.
In contrast to 1995, no significant difference (F = 13.9; df =3, 119; P = 0.36) was observed for mean percent leaf tissue consumed among the 3 Tilia genotypes and the standard (S. albidum) (Table 2) or for the 7 Tilia genotypes growing in the paper pots in the nursery (F = 12.8; df = 7, 204; P = 0.24) (Table 3). Mean percent leaf tissue consumed in both studies ranged from a low of 44% for T. cordata ‘Prestige’ to 62% for T. cordata ‘Greenspire’ (Table 2 and Table 3). In a direct comparison of like cultivars in our study and in a study by Potter et al. (1998), T. cordata ‘Greenspire’ had the highest degree of leaf tissue consumed (62%). In addition, we found T. tomentosa ‘Sterling’ to have the second highest level of leaf tissue consumed (61%), whereas Potter et al. (1998) found T. tomentosa ‘Sterling’ to have one of the lowest levels of feeding damage. Tilia tomentosa in our study had a lower amount of leaf tissue consumed, which is consistent with Potter et al. (1998). Tilia americana ‘Redmond’ in this study had lower feeding damage, which is also consistent with Potter et al. (1998) for this same cultivar.
Significant differences (F = 36.2; df = 3, 119; P < 0.001) in mean dried fecal pellet weights were observed however, among the 3 Tilia genotypes and the standard (S. albidum) growing in the field (Table 2) and among the 7 genotypes growing in the paper pots (F = 31.5; df = 7, 204; P = 0.001) (Table 3). Tilia americana ‘Redmond’ had significantly greater dried fecal pellet weights of 68.0 mg. (Table 3). Sassafras albidum (standard) and T. tomentosa had significantly lower mean dried fecal pellet weights of 38.1 and 38.4 mg, respectively (Table 2). The remaining Tilia genotypes were moderately preferred with pellet weights of 44.3 to 63.9 mg (Table 2 and Table 3).
Mean dried fecal pellet weight for T. tomentosa was not as consistent between the 1995 and 1997 studies (Table 1 and Table 2).
Multiple-Choice Laboratory Feeding Study, 1997
Overall, results from the multiplechoice study tended to support the preference rankings observed in the no-choice laboratory bioassays and the field defoliation survey. Adult Japanese beetles fed significantly more (F = 24.2; df = 5, 240; P < 0.001) on T. tomentosa ‘Sterling’ as compared to the other 5 Tilia genotypes (Table 4). About 37% of the leaf tissue was removed from the T. tomentosa ‘Sterling’ leaf discs, while only 6% to 17% damage was recorded for the other 5 varieties. These results are consistent with the nochoice laboratory tests: T. tomentosa ‘Sterling’ was second only to T. cordata ‘Greenspire’ in preference, with 61% defoliation (Table 3). In the field defoliation survey, T. tomentosa ‘Sterling’ exhibited only light (L) feeding damage (Table 3). The more mature T. tomentosa trees experienced a moderate (M) feeding rating (Table 2) but only 3% defoliation in the multiple-choice bioassay test (Table 4). The less-preferred genotype of T. cordata ‘Prestige’ exhibited light to moderate (M) feeding damage in the field and had approximately 44% damage in the laboratory bioassays (Table 3). Personal observations by the authors have revealed that Tilia platyphyllos ‘Parade’, T. cordata ‘Prestige’, ‘Sundance’, and T. tomentosa have dense mats of white tomentose and/or very long fine hairs on the lower leaf surface. In contrast, the more highly preferred T. tomentosa ‘Sterling’ has sparse, short stellate (star-shaped) hairs on the lower leaf surface. The differences in pubescence may partially account for the differences in preference observed in the multiple-choice study.
Mean percent leaf tissue consumed by Japanese beetles for multiple-choice studies on selected Tilia spp. and varieties in a multiplechoice test (August 1997).
Field Defoliation Survey, 1997
Overall, rankings for the field defoliation survey were fairly consistent with the no-choice and multiple-choice laboratory feeding studies. Tilia genotypes, such as T. cordata ‘Greenspire’, T. tomentosa, T. tomentosa ‘Sterling’, and T. cordata ‘Fairview’, had light (L) to moderate (M) feeding damage and tended to have higher percentages of leaf area consumed in the no-choice laboratory feeding study. One exception, T. cordata ‘Prestige’, had moderate feeding in the field but had a lower percentage of leaf area consumed (Table 3). Fleming (1972) classifies T. cordata as having moderate feeding and T. tomentosa as general light feeding. Genotypes such as T. amurensis, T. platyphyllos ‘Parade’, ‘Sundance’, and T. tomentosa, which had no (NF) to light (L) feeding, also had lower percentages of leaf area consumed both in the no-choice and multiple-choice laboratory feeding studies (Tables 2 through 4).
Leaf pubescence may play a role in feeding preference of Tilia spp. by adult Japanese beetles. We have observed that the upper leaf surfaces of the Tilia spp. evaluated in the 1997 study were glabrous, while the lower leaf surfaces varied in presence and density of hairs. Tilia genotypes that are glabrous or have very sparse densities of hairs, such as T. cordata ‘Fairview’, T. cordata ‘Greenspire’, and T. tomentosa ‘Sterling’, tend to have more leaf area consumed as compared to T. cordata ‘Prestige’, T. platyphyllos ‘Parade’, and T. tomentosa, which have thick white tomentose mats and/or long and upright hairs. The one exception is T. americana ‘Redmond’, which has a glabrous lower leaf surface but which experienced feeding damage comparable to the highly pubescent genotypes. We have also observed in related feedingpreference studies that elm (Ulmus spp.) genotypes with heavy pubescence (i.e., U. lamellosa, U. macro-carpa) tend to be less preferred by adult Japanese beetles (unpublished data). Undoubtedly, other factors—such as leaf surface chemicals, internal chemicals, or leaf tissue toughness—may also be involved in apparent resistance.
Conclusion
Using mean percent leaf area consumed and mean dried fecal pellet weights as measures of susceptibility, the Asian species of T. petiolaris, T. oliveri, T. chinensis, and T. japonica show promise for partial resistance to Japanese beetles. Three North American species, T. americana, T. heterophylla, and T. caroliniana also show promise along with the pubescent European species of T. tomentosa and T. tomentosa ‘Erecta’. Further studies are needed to determine the factor(s) responsible for apparent resistance and the horticultural qualities of these species for use in the land-scape and as street trees.
Acknowledgements
The authors wish to thank D. Potter, University of Kentucky, for his helpful review of an earlier draft of this manuscript. Special thanks is extended to D. Danielson, The Morton Arboretum, for her assistance in field collection of beetles and laboratory work, and T. Tiddens, the Chicago Botanic Garden, for supplying us with Japanese beetles. Thanks and appreciation is due to M. Miller for her assistance with data entry and manuscript preparation.
- © 1999, International Society of Arboriculture. All rights reserved.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.