Leaved Cultivars of Woody Landscape Plants to the Japanese Beetle

PEST MANAGEMENT

HORTSCIENCE 37(2):362Ð366. 2002. purple/green leaves, whereas 25 of the 26 most resistant cultivars had completely green leaves (Spicer et al., 1995). The extent to which this Susceptibility of Purple- Versus Green- apparent color preference may apply to other woody plant species is unknown. leaved Cultivars of Woody Landscape The present study tested the hypothesis, across a range of plant genera and species, that Plants to the Japanese Beetle purple cultivars of woody landscape plants generally sustain more feeding damage from W. Jack Rowe, II1, Daniel A. Potter1,3, and Robert E. McNiel2 Japanese beetles than do green-leaved ones. We also sought to identify species and culti- University of Kentucky, Lexington, KY 40546-0091 vars of purple-leaved woody plants that could Additional index words. integrated pest management, plant resistance, Prunus, Acer, Popillia be substituted for highly susceptible plant materials in areas where Japanese beetles are japonica abundant. Abstract. Twenty-six purple- or green-leaved cultivars representing 12 species of woody landscape plants were evaluated in the field for defoliation by Japanese beetles (Popillia japonica Newman) over three growing seasons. We further evaluated the hypothesis that, Materials and Methods within closely-related plants, purple cultivars generally are preferred over green ones by comparing beetles’ consumption of foliage in laboratory choice tests and their orientation Twenty-six purple- or green-leaved culti- to painted silk tree models baited with Japanese beetle lures. Cultivars of Prunus cerasifera vars (Table 1) representing 12 species of woody Ehrh. and hybrids of that species [e.g., Prunus ×cistena (Hansen) Koehne, Prunus landscape plants were evaluated in the field ×blireiana André] were more heavily damaged than nearly all other plants tested. Among for susceptibility to Japanese beetles over three maples, Acer palmatum Thunb. ‘Bloodgood’ and A. platanoides L. ‘Deborah’ and ‘Fairview’ growing seasons. Green and purple cultivars were especially susceptible. None of the cultivars of Berberis thunbergii DC, Cercis of the same species were included for com- canadensis L., Cotinus coggygria Scop., or Fagus sylvatica L. were heavily damaged, parison. Some species had cultivars with vari- regardless of foliage color. In the choice tests, purple Norway maples were preferred over able foliage color. For example, A. platanoides green ones in three of four comparisons, but preference varied within the other plant ‘Fairview’ and ‘Deborah’ have purple juve- genera. In fact, more beetles oriented to green-leaved tree models than to purple ones. Our nile foliage that changes to green as the leaves results indicate that within a genus, purple-leaved plants do not necessarily sustain more mature. In contrast, leaves of Prunus virginiana damage than green-leaved ones. Widespread use of certain purple-leaved cultivars of ‘Canada Red’ change from green to purple as generally susceptible plant species probably contributes to the perception that purple- they mature. Purple- and green-leaved culti- leaved plants, overall, are preferred. Purple-leaved cultivars of redbud, European beech, vars of Cotinus coggygria, Fagus sylvatica, smoketree, and barberry, or the purple-leaved Prunus virginiana L. ‘Canada Red’ or Berberis thunbergii, and Cercis canadensis, Malus ×hybrida Lemoine ‘Jomarie’ may be suitable substitutes for more susceptible species listed by Fleming (1972) as resistant to purple-leaved plants in landscapes where Japanese beetles are a concern. Japanese beetles, also were included. The main field site was on a Maury silt loam soil (pH 6.4) at the Univ. of Kentucky The Japanese beetle, Popillia japonica Rankings of various plant species as to the South Research Farm, Lexington (U.S. Dept. Newman, an introduced species, is a destruc- extent that they are fed upon by Japanese Agr. Plant Hardiness Zone 6). Plants were tive pest of nursery and landscape plants in the beetles have been compiled from general field obtained as bare-root whips or balled-and- eastern United States (Fleming, 1972; Potter observations (Fleming, 1972) or from labora- burlapped stock from Schmidt Nurseries (Bor- and Held, 2002). The adults feed on 300 tory feeding trials with leaf disks (Ladd, 1987, ing, Ore.), Snow Hill Nursery (Shelbyville, species of wild and cultivated plants, espe- 1989). Within some generally susceptible plant Ky.), and Hillenmeyer’s Nursery (Lexington, cially among the Rosaceae, Malvaceae, genera and species, certain cultivars are more Ky.). The trees and shrubs were planted in Vitaceae, Polygnoceae, Aceraceae, Ulmaceae, resistant than others. Replicated field trials or holes (45 cm diameter; 2 m spacing) augured and Salicaceae (Fleming, 1972; Ladd, 1987). laboratory assays have documented signifi- into plowed and disked soil on 27 Apr. 1993. Highly preferred hosts such as lindens (Tilia cant variation in resistance among cultivars or Plot design was a randomized complete block sp.), sassafras [Sassafras albidum (Nutt.) varieties of birch (Betula sp.), flowering with six replications. All plants were fertilized Nees.], roses (Rosa sp.), Norway maples (Acer crabapples (Malus sp.), flowering cherry with ammonium nitrate (48.8 kg N/ha) in early platanoides), and purple-leaved or cherry plum (Prunus sp.), elm (Ulmus sp.), and linden May, following leaf flush. The field was tilled (Prunus cerasifera) often are severely defoli- (Tilia sp.) (Fulcher et al., 1998; Miller and as necessary to control weeds. ated. The greatest damage results from the Ware, 1999; Miller et al., 1999; Patton et al., Field ratings. Three independent observ- beetles’ tendency to aggregate on individual 1997; Potter et al., 1998; Ranney and ers visually rated each tree or shrub for Japa- plants, a response mediated by attraction to Walgenbach, 1992; Spicer et al., 1995). Sig- nese beetle feeding damage in early August, feeding-induced plant volatiles (Loughrin et nificant resistance has not, however, been found soon after peak beetle flight, in 1993, 1994, al., 1995, 1996). in common garden roses (Potter et al., 1998). and 1995. Estimates of defoliation (nearest We have observed that certain landscape 10%) were averaged among observers to pro- Received for publication 7 May 2001. Accepted for plants having deep red or purplish foliage, vide a single value for each plant. Percentages publication 17 Oct. 2001. This is paper number 01- e.g., ‘Crimson King’ Norway maple, purple- were adjusted using an arcsine square root 08-62 of the Kentucky Agricultural Experiment leaved plums, purple sand cherry (Prunus transformation (Steel and Torrie, 1980), then Station, Lexington. Research supported in part by ×cistena), and ‘Bloodgood’ Japanese maple differences among species or cultivars were USDAÐSRIPM grants 94-34103-0185 and 96- (Acer palmatum), seem to be especially sus- evaluated by two-way analysis of variance 34103-3346, and by a grant from the Horticultural ceptible to Japanese beetles. In field evalua- (ANOVA) using Statistix for Windows (Ana- Research Institute. The authors thank P.G. Spicer tions of 42 flowering crabapple cultivars, Spicer lytical Software 1996). and D.W. Held for technical assistance. et al. (1995) found that purple-leaved variet- Instead of comparing all possible combi- 1Dept. of Entomology, S-225 Agric. Sci. Bldg. N., Univ. of Kentucky, Lexington, KY 40546. ies, and those with purplish juvenile foliage nations of means, the hypothesis that purple- 2Dept. of Horticulture, N-318 Agric. Sci. Bldg. N., that becomes green when mature, tended to be leaved cultivars sustained more damage than Univ. of Kentucky, Lexington, KY 40546. fed upon more heavily than were cultivars green-leaved ones was tested within selected 3To whom reprint requests should be addressed with completely green leaves. Indeed, eight of species. Dunnett’s test (Steel and Torrie, 1980), ([email protected]). the 10 most susceptible cultivars had purple or which compares treatment means vs. a stan-

362 HORTSCIENCE, VOL. 37(2), APRIL 2002

p362-366 362 3/26/02, 2:02 AM Table 1. Species and cultivars of woody landscape plants with green or purple foliage that were evaluated purchased from a local retailer. The artificial for susceptibility to defoliation by the Japanese beetle. trees were set in concrete in 1.4-L black plastic Species/cultivar Foliage colorz Common or trade name pots. Fresh, fully expanded flowering crabapple Acer palmatum Green Japanese maple leaves (Malus sp.) were harvested from Acer palmatum Atropurpureum Purple Purple Japanese maple ‘David’, a green-leaved cultivar, or ‘Royalty’, Acer palmatum Bloodgood Purple Bloodgood Japanese maple a highly susceptible purple-leaved cultivar Acer platanoides Cleveland Green Cleveland Norway maple (Spicer et al., 1995). The leaves were scanned Acer platanoides Crimson King Purple Crimson King Norway maple under daylight with a MacBeth Color-I elec- Acer platanoides Crimson Sentry Purple Crimson Sentry Norway maple tronic spectrophotometer (Kollmorgen Corp., Acer platanoides Deborah Purple/green Deborah Norway maple Newburg, N.Y.), then oil-based enamel paints Acer platanoides L. Fairview Purple/green Fairview Norway maple (Porter Paints, Lexington, Ky.) were mixed to Berberis thunbergii Atropurpurea Purple Purple Japanese barberry match the spectral reflectance of the green or Berberis thunbergii Crimson Pygmy Purple Crimson Pygmy Japanese barberry Berberis thunbergii Kobold Green Kobold Japanese barberry purple cultivars. These were used to paint both Berberis thunbergii DC Rosy Glow Purple Rosy Glow Japanese barberry surfaces of all leaves of a particular tree. The Cercis canadensis L. Green Redbud paint was applied 4 weeks before the first field Cercis canadensis Forest Pansy Purple Forest Pansy redbud trial to ensure that it was thoroughly dry and Cotinus coggygria Green Smoketree any volatile solvents had dissipated. Cotinus coggygria Atropurpurea Purple Purple smoketree Within each trial, a green and purple tree Fagus sylvatica Green European beech were positioned 2 m apart in an open turf area, Fagus sylvatica Riversii Purple Riversii European beech × perpendicular to prevailing wind and 10 m Malus hybrida Jomarie Purple Jomarie crabapple upwind of a plot (15 × 15 m) of cultivated Prunus cerasifera Frankthrees Purple Frankthrees flowering plum Prunus cerasifera Thundercloud Purple Thundercloud flowering plum wildflowers with abundant Japanese beetles. Prunus cerasifera Newport Purple Newport flowering plum One half of a commercial Japanese beetle Prunus ×blireiana Purple Blireiana plum floral lure, as described earlier, was hung from Prunus ×cistena Purple Purple-leaf sand cherry a paper clip in the upper one-third of the Prunus serrulata Royal Burgundy Purple Royal Burgundy Oriental cherry canopy of each tree. The models were baited at Prunus virginiana Canada Red Green/purple Canada Red chokecherry 1:00 PM and then two observers recorded and aWhere two colors are indicated, they refer to juvenile and mature leaves, respectively. removed all beetles that landed on each tree during the subsequent hour. Tree positions dard while controlling experimentwise error (mg) was determined by a weight/area conver- and lures were reversed every 10 min during a rate (α = 0.05), was used for comparisons sion factor calculated from standards of each trial. Trials were conducted on 20 June, 6 July, within Norway and Japanese maples. Pre- plant cultivar. and 16 July 1993. Fresh baits were used for planned paired t tests (P = 0.05) were used to Unless indicated otherwise, foliage used in each trial. Counts from each trial were ana- compare green and purple cultivars of Cotinus preference tests came from green- or purple- lyzed by Chi-square for significant departure and Fagus. Data are reported as non-trans- leaved tree cultivars planted at the main field from a 50:50 ratio landing on each foliage formed means (±SE). site. However, to allow a greater range of color. Preference tests. Choice tests were con- comparisons, we also harvested leaves from ducted using combinations of purple and green additional tree cultivars planted at two other Results and Discussion cultivars of the same species. Beetles were local sites (see below). Trees from those sites collected with standard Japanese beetle traps were similar in size to those in our main Field ratings. Japanese beetles did more baited with floral lures (3:7:3 blend of planting, and were maintained in similar man- damage to cultivars of Prunus cerasifera and phenethyl propionate, eugenol, and geraniol; ner. No insecticides had been applied to any of to hybrids of that species (Prunus ×cistena, Trécé, Salinas, Calif.). Freshly-caught beetles the plants used for the feeding assays. Prunus ×blireiana), than to nearly all other were held overnight, without food, before be- Many cultivars of A. platanoides have been tree and shrub species (Table 2). Overall defo- ing used in tests. Each choice test arena was a released, so we randomly paired each of the liation, averaged across all three years, was 470-mL plastic tub (Solo Inc., Baltimore, Md.) four purple-leaved cultivars from the experi- 63.8% for P. cerasifera cultivars, vs. 43 and with paraffin (2.5 cm deep) and a moistened mental plot with a different green-leaved cul- 35% for P. ×cistena and P. ×blireiana, respec- filter paper disk in the bottom. Leaves used in tivar. Foliage of ‘Cleveland’ was obtained tively. In contrast, P. virginiana ‘Canada Red’, these tests were collected in mid-morning, from trees in the main plot. Leaves of three another purple-leaved species, sustained only placed in sealed plastic bags in a cooler, and other green cultivars, ‘Columnare’, ‘Emerald 6.9% average defoliation. Among the maples, used within 1Ð2 h of harvest. Leaves from a Queen’, or ‘Emerald Luster’, were harvested Acer palmatum ‘Bloodgood’, and A. different tree were used for each replicate. from trees at Hillenmeyer’s Nursery, Lexing- platanoides ‘Deborah’ and ‘Fairview’ were Leaf disks (23 mm diameter, 415 mm2 area) ton, Ky. Similarly, leaves of purple-leaved especially susceptible (32.2%, 31.7%, and were cut with a cork borer and mounted on #3 ‘Forest Pansy’ redbud in the main plot were 19.0% average defoliation, respectively). None insect pins between bits (4 mm2) of acetate paired with foliage of native, green-leaved C. of the woody plants other than Prunus and film to secure them during the beetles’ feeding. canadensis planted at Hillenmeyer’s Nursery. Acer species sustained, on average, more than For choice tests, two leaf disks of each Because the purple-leaved Prunus sp. culti- 10% defoliation (Table 2). Berberis thunbergii cultivar being compared were pinned alter- vars had no green counterpart in the main field was particularly resistant, likely because it nately around the inner wall of the arena. plot, foliage for comparisons was harvested contains berberine, a toxic alkaloid that at Three female beetles were placed inside and from a green-leaved cherry (P. avium ‘Stella’) naturally occurring levels caused complete allowed to feed until, by visual estimate, 75% and plum (P. domestica ‘Bluefree’) cultivar mortality of another generalist, Spodoptera of the preferred food was consumed (gener- that were planted in nursery plots at the Univ. littoralis (Lepidoptera: Noctuidae) (Krug and ally <6 h). Each comparison was replicated six of Kentucky’s Spindletop Research Farm, near Proksch, 1993). times. During assays, the test arenas were kept Lexington. Each comparison was replicated Preplanned comparisons were conducted at ambient laboratory temperature (25 ¡C) six times, with leaves from different trees used within each maple species to test whether under fluorescent lights. At termination, re- for each replicate. purple-leaved cultivars were preferred over maining portions of the leaf disks were mea- Attraction to green or purple-hued leaves. green-leaved ones. Although there were over- sured on an electronic area meter (Delta T To test the hypothesis that Japanese beetles are all differences in defoliation among A. Devices, Cambridge, England) and subtracted attracted more to purple-hued foliage than to palmatum cultivars in 1993 (F = 9.36; df = from the original area of the disk. To account green foliage, nearly identical silk Ficus tree 2,10; P < 0.01), only one of the two purple for differences in leaf thickness, amount eaten models, 1.8 m tall with 650 ± 5 leaves, were cultivars (‘Bloodgood’) sustained significantly

HORTSCIENCE, VOL. 37(2), APRIL 2002 363

p362-366 363 3/26/02, 2:02 AM PEST MANAGEMENT

Table 2. Field evaluations of cumulative Japanese beetle feeding damage to green- or purple-leaved cultivars There also were no differences in defoliation of woody landscape plants by the first week of Aug. 1993Ð95. between green- or purple-leaved cultivars of

z Cotinus and Fagus (paired t tests, P > 0.05), all % defoliation (mean ± SE) on of which had relatively little feeding. Two Species/cultivar Colory 10 Aug. 1993 10 Aug. 1994 7 Aug. 1995 other purple-leaved trees, Cercis canadensis Acer palmatum G 14.9 ± 3.7 11.7 ± 1.1 3.2 ± 0.7 × A. palmatum Atropurpureum P 27.8 ± 12.3 17.2 ± 4.1 6.0 ± 1.6 ‘Forest Pansy’ and Malus hybrida ‘Jomarie’, A. palmatum Bloodgood P 67.2 ± 7.6* 20.6 ± 6.6 8.7 ± 3.0 also sustained little feeding damage (Table 2). Acer platanoides Cleveland G 9.9 ± 5.6 10.0 ± 0.0 6.7 ± 1.2 Preference tests. In paired choice tests A. platanoides Crimson King P 16.6 ± 2.9 10.7 ± 0.7 3.9 ± 1.1 with green- or purple-leaved Norway maples, A. platanoides Crimson Sentry P 10.5 ± 2.0 10.0 ± 0.0 2.4 ± 1.1 Japanese beetles fed significantly more on the A. platanoides Deborah P/G 47.3 ± 14.8* 15.9 ± 2.7* 32.0 ± 7.9* purple cultivar in three of the four compari- A. platanoides Fairview P/G 29.4 ± 7.9x 11.1 ± 0.7 16.7 ± 2.7x sons (Fig. 1). When foliage from the green- Berberis thunbergii Atropurpurea P 0 0 1.8 ± 1.3 leaved plum P. domestica ‘Bluefree’ was paired B. thunbergii Cherry Bomb P 0 0 0.5 ± 0.3 with each of three purple P. cerasifera culti- B. thunbergii Crimson Pygmy P 0 0 0.2 ± 0.2 B. thunberigii Kobold G 0 0 0.8 ± 0.8 vars, however, only ‘Frankthrees’ was signifi- B. thunbergii Rosy Glow P 0 0 1.6 ± 0.9 cantly preferred over the green cultivar (Fig. Cercis canadensis Forest Pansy P 2.2 ± 1.6 10.0 ± 0.0 0.2 ± 0.2 2). In the tests with green and purple cultivars Cotinus coggygria G 6.7 ± 6.7 3.3 ± 2.1 2.5 ± 1.2 of five other plant species, the beetles fed C. coggygria Atropurpurea P 0 8.3 ± 1.7 1.7 ± 1.1 significantly more on the purple cultivar only Fagus sylvatica G 0 NA 2.5 ± 1.2 with cherry (P. serrulata vs. P. avium). In fact, F. sylvatica Riversii P 3.3 ± 0.9 11.7 ± 1.7 2.4 ± 0.9 there was more feeding on the green beech × Malus hybrida Jomarie P 8.3 ± 1.1 12.2 ± 2.2 7.2 ± 1.6 cultivar than on the purple one (Fig. 3). Al- Prunus cerasifera Frankthrees P 91.6 ± 3.6 57.7 ± 10.5 78.3 ± 3.9 though most differences were not statistically P. cerasifera Thundercloud P 61.6 ± 11.9 23.9 ± 1.0 41.1 ± 3.3 P. cerasifera Newport P 89.4 ± 0.6 58.3 ± 9.9 72.0 ± 2.3 significant, there was a trend for greater feed- Prunus ×blireiana P 39.9 ± 6.7 29.4 ± 3.0 35.6 ± 3.3 ing on the purple cultivar in 11 of the 12 Prunus ×cistena P 27.8 ± 8.0 69.4 ± 10.4 31.7 ± 4.1 aforementioned comparisons. Prunus serrulata Royal Burgundy P 22.2 ± 3.3 10.0 ± 0.0 17.8 ± 1.6 Attraction to green or purple-hued leaves. Prunus virginiana Canada Red G/P 6.6 ± 1.4 10.8 ± 0.8 3.2 ± 1.1 Significantly more beetles were attracted to zBased on average ratings of three independent observers. Differences among cultivars are significant within the green-leaved tree model than to the purple- each year (two-way ANOVA on arcsin square root-transformed percentages: F = 32.5; df = 27, 129; P < leaved model in each trial (Fig. 4). This is 0.001 for 1993; F = 42.7; df = 27, 118; P < 0.001 for 1994; F = 71.7, df = 27, 121; P < 0.001 for 1995). opposite what would be expected if Japanese y G, green foliage; P, purple foliage; P/G, foliage is purple when juvenile, green when mature. beetles are indeed disproportionately attracted x, *Significantly different from the corresponding green cultivar of the same species at experimentwise error rate (α) 0.06 or 0.05, respectively; ANO VA; Dunnett’s test for comparison of treatments vs. a standard. to purple-hued foliage. Differences in num- bers of beetles responding in each trial reflect seasonal variability in beetle activity. Color vision is well-documented in plant- feeding insects, and many species respond to optical host-plant cues such as spectral qual- ity, size, and shape as a component of their host-selection behavior (Prokopy and Owens, 1983; Schoonhoven et al., 1998). Indeed, a number of day-foraging insects show visu- ally-based landing in response to artificial leaves that have been painted to simulate host- plant leaf reflectance profiles (Prokopy and Owens, 1983; Schoonhoven et al., 1998). The spectral sensitivity of insect compound eyes ranges from 350Ð650 nm (near ultraviolet to red) and thus includes shorter wavelengths than are perceived by the human eye (Prokopy and Owens, 1983; Schoonhoven et al., 1998). Perception of wavelengths within this range varies by species (Prokopy and Owens, 1983). That P. japonica responds to color and hue is supported by studies showing that trap color affects their capture rates in traps baited with floral-type lures (Fleming, 1976). However, as nothing is known about their optical sensi- tivity, we cannot know the extent to which the beetles perceived our tree models as actual trees. Given the species’ broad dietary range, Fig. 1. Japanese beetles’ feeding on green- vs. purple-leaved cultivars of Norway maple (Acer platanoides) in choice tests. Means marked by an asterisk differ significantly from the corresponding mean (paired and the similarity in foliar spectral quality t tests, P < 0.05). among most plants (Prokopy and Owens, 1983), it seems clear that even if color is a more damage than did the green-leaved stan- 20; P < 0.05). Cultivar Deborah, which has factor in host choice, other factors including dard. There were no overall differences, or purple juvenile leaves that become green when constitutive or feeding-induced plant volatiles differences between green- or purple-leaved mature, was more heavily damaged than (Loughrin et al., 1995, 1996), feeding stimu- cultivars of Japanese maples, in 1994 or 1995. ‘Cleveland’, the green-leaved standard. How- lants, or antifeedants such as prunasin in P. Damage to A. platanoides differed signifi- ever, the two entirely purple cultivars, ‘Crim- virginiana ‘Canada Red’ (Patton et al., 1997) cantly among cultivars in each year (F = 4.02, son King’ and Crimson Sentry’, sustained or foliar phenolics in Malus sp. (Fulcher et al., 6.38, 13.41 for 1993Ð95, respectively; df = 4, similar levels of defoliation as ‘Cleveland’. 1998) may override it. It also should be noted

364 HORTSCIENCE, VOL. 37(2), APRIL 2002

p362-366 364 3/26/02, 2:05 AM In the field test, there was in fact considerable variation in defoliation among purple-leaved cultivars within certain species (e.g., A. palmatum, A. platanoides), and cultivars of other species sustained little feeding damage, regardless of leaf color. The perception that purple-leaved plants, as a group, are preferred by Japanese beetles perhaps reflects the fact that some of the most popular, or at least readily available, purple- leaved woody plants (e.g., ‘Crimson King’ and ‘Crimson Sentry’ Norway maples, ‘Bloodgood’ Japanese maple, Prunus cerasifera and its hybrids) belong to generally susceptible plant species. Additionally, such plants are often focal points in ornamental landscapes. This study indicates that purple cultivars of redbud, European beech, smoketree, or barberry, or purple-leaved P. virginiana ‘Canada Red’ or Malus ×hybrida ‘Jomarie’ could be substituted for more sus- ceptible hosts of the Japanese beetle in land- scapes where purple-leaved plants are desired.

Literature Cited Analytical Software. 1996. Statistix for Windows. Fig. 2. Japanese beetles’ feeding on three purple-leaved Prunus cerasifera cultivars each compared to green- Analytical Software, Tallahassee, Fla. leaved P. domestica ‘Bluefree’ in choice tests. Means marked by an asterisk differ significantly from Dirr, M.A. 1990. Manual of woody landscape plants: corresponding mean (paired t tests, P < 0.05). their identification, ornamental characteristics, culture, propagation, and uses. 4th ed. Stipes, Champaign, Ill. Fleming, W.E. 1972. Biology of the Japanese beetle. U.S. Dept. Agr. Tech Bul. 1449. Fleming, W.E. 1976. Integrating control of the Japa- nese beetle—A historical review. U.S. Dept. Agr. Tech Bul. 1545. Fulcher, A.F., T.G. Ranney, J.D. Burton, J.F. Walgenbach, and D.A. Danehower. 1998. Role of foliar phenolics in host plant resistance of Malus taxa to adult Japanese beetles. Hort- Science 33:862Ð865. Krug, E. and P. Proksch. 1993. Influence of dietary alkaloids on survival and growth of Spodoptera littoralis. Biochem. System. Ecol. 21:749Ð756. Ladd, T.L., Jr. 1987. Japanese beetle (Coleoptera: Scarabaeidae): influence of favored food plants on feeding response. J. Econ. Entomol. 80:1014Ð 1017. Ladd, T.L., Jr. 1989. Japanese beetle (Coleoptera: Scarabaeidae): feeding by adults on minor host and nonhost plants. J. Econ. Entomol. 82:1616Ð 1619. Loughrin, J.H., D.A. Potter, and T.R. Hamilton- Kemp. 1995. Volatile compounds induced by herbivory act as aggregation kairomones for the Japanese beetle (Popillia japonica Newman). J. Chem. Ecol. 21:1457Ð1467. Fig. 3. Choice tests comparing Japanese beetles’ consumption of foliage from purple- vs. green-leaved Loughrin, J.H., D. A. Potter, T.R. Hamilton-Kemp, cultivars of various woody plant species. See text for cultivars that were compared. Means marked by and M.E. Byers. 1996. Role of feeding-induced an asterisk differ significantly from corresponding mean (paired t tests, P < 0.05). plant volatiles in aggregative behavior of the Japanese beetle (Coleoptera: Scarabaeidae). Environ. Entomol. 25:1188Ð1191. that many purple-leaved cultivars of a given insect defoliator. For example, A. platanoides Miller, F., S. Jerdan, and G. Ware. 1999. Feeding species often share a common parent; e.g., ‘Deborah’, which we found to be susceptible preference of adult Japanese beetles (Coleoptera: Malus pumila ‘Niedzwetzkyana’ and P. to Japanese beetles, was the most resistant of Scarabaeidae) for Asian elm species and their cerasifera ‘Atropurpurea’ are parents of many several Norway maple cultivars tested with hybrids. J. Econ. Entomol. 92:421Ð426. of the purple-leaved crabapples and plums, gypsy moth caterpillars (Peterson and Smitley, Miller, F. and G. Ware. 1999. Feeding preference respectively (Dirr, 1980). Thus, susceptibility 1991). for selected Tilia spp. and cultivars by the adult Japanese beetle (Coleoptera: Scarabaeidae). J. of purple-leaved cultivars within these genera Although we found trends for greater feed- Arboric. 25:168Ð174. may driven by other traits that are linked or ing by P. japonica on purple-leaved cultivars Patton, C.A., T.G. Ranney, J.D. Burton, and J.F. shared by these plants, independent of foliage of some plants (e.g., choice tests with A. Walgenbach. 1997. Natural pest resistance of color. platanoides), this study does not broadly sup- Prunus taxa to feeding by adult Japanese beetles: Relative preferences for red or green-leaved port the hypothesis that purple-leaved plants Role of endogenous allelochemicals in host plant cultivars may also vary with the species of are more susceptible than green-leaved ones. resistance. J. Amer. Soc. Hort. Sci. 122: 668Ð672.

HORTSCIENCE, VOL. 37(2), APRIL 2002 365

p362-366 365 3/26/02, 2:08 AM PEST MANAGEMENT

Peterson, N.C. and D.R. Smitley. 1991. Susceptibil- ity of selected shade and flowering trees to gypsy moth (Lepidoptera: Lymantriidae). J. Econ. Entomol. 84:587Ð592. Potter, D.A., P.G. Spicer, D.W. Held, and R.E. McNiel. 1998. Relative susceptibility of culti- vars of flowering crabapple, linden, and rose to defoliation by Japanese beetles. J. Environ. Hort. 16:105Ð111. Potter, D.A. and D.W. Held. 2002. Biology and management of the Japanese beetle. Annu. Rev. Entomol. 47:175Ð205. Prokopy, R.J. and E.D. Owens. 1983. Visual detec- tion of plants by herbivorous insects. Annu. Rev. Entomol. 28:337Ð64. Ranney, T.G. and J.F. Walgenbach. 1992. Feeding preferences of Japanese beetles for taxa of birch, cherry, and crabapple. J. Environ. Hort. 10:177Ð 180. Schoonhoven, L.M., T. Jermy, and J.J.A. van Loon. 1998. Insect-plant biology. From physiology to evolution. Chapman & Hall, London, U.K. Spicer, P.G., D.A. Potter, and R.G. McNiel. 1995. Resistance of crabapple (Malus spp.) cultivars to defoliation by the Japanese beetle (Coleoptera: Scarabaeidae). J. Econ. Entomol. 88:979Ð985. Fig. 4. Number of Japanese beetles landing on silk tree models that had been painted to match the spectral Steel, R.G.D. and J.H. Torrie. 1980. Principles and reflectance of purple or green crabapple foliage and baited with floral lures. More beetles oriented to the procedures of statistics, a biometrical approach, green model than to the purple one in each trial (Chi-square tests, P < 0.05). 2nd ed. McGraw Hill, N.Y.

366 HORTSCIENCE, VOL. 37(2), APRIL 2002

p362-366 366 3/26/02, 2:10 AM