FOREST ENTOMOLOGY Survival and Development of Lymantria mathura (Lepidoptera: Lymantriidae) on North American, Asian, and European Tree Species
MARINA A. ZLOTINA,J VICTOR C. MASTRO,2 DAVID E. LEONARD,2 AND JOSEPH S. ELKINTON1
Otis Methods Development Center, USDA-APHIS-PPQ, Otis ANGB, MA 02542-5008 Downloaded from https://academic.oup.com/jee/article/91/5/1162/919949 by UMass Amherst Libraries user on 05 March 2021
J. Econ.Entomol.91(5): 1162-1166(1998) ABSTRACT The moth Lymantria mathura Moore is a polyphagous defoliator of forest and fruit trees in the Russian Far East, Japan, India, and China, and it would cause serious economic consequences if established in North America. To project the host range of this insect if accidentally introduced into North America, we examined survival and development of 1st instal'Sof L. mathura on 24 broadleafed and conifer tree species from North America, Europe, and Asia. L. mathura performed well on hosts in the family Fagaceae regardless of the species origin. Survival and rate of development on European beech, Fagus grandifolia Ehrh., and American beech, Fagus syLvatica L., was equivalent to survival on oriental oak, QlIerCllS variabilis Blume. Performance of larvae was also high on white oak, QlIerCllS alba L., and chestnut oak, QuerCllS priJlllS L. Suitability of other oaks, such as red oak, Quercl/s rubra L., black oak, QuerCllS velutina Lamarck, and pin oak, QuerCllS pall/stris Muenchh., was intermediate and equivalent to suitability of species from the families Juglandaceae, Betulaceae, and Oleaceae. Survival was poor on hosts from Aceraceae and Pinaceae. Comparison of host preferences and larval survival of L. mathura and of the gypsy moth, Lymantria dispar (L.), indicates differences between these species when tested on the same or congeneric host plants.
KEY WORDS Lymantria mathura, host plant suitability, survival, rate of development
EGGMASSES OF the moth Lymantria mathura Moore and The wide host range and distribution of L. mathura the Asian form of the gypsy moth, Lymantria dispar in temperate forests from the Russian Far East to Japan (L.), have been intercepted on ships coming to the suggest that it could establish and achieve pest status Pacific Northwest from ports in the Russian Far East if accidentally introduced into North America. To (Mudge and Johnson 1992, Wallner 1993, Mastro predict its potential for establishment in North Amer- 1995). Asian gypsy moth was found in 1992 in several ica, we examined survival and development of neo- localities adjacent to ports in the Pacific Northwest. nates on 19 North American hardwood and conifer Federal, state, and Canadian provincial agencies re- species. Three Asian and 2 European species were sponded to these findings with eradication programs included for comparison. and other measures to prevent additional introduc- tions (Gibbons 1992, Wallner et al. 1995). It is not Materials and Methods known if L. mathura has become established in North America, but sensitive methods of survey, such as We used cultures of L. rnathum originating from the pheromone traps, are not available. egg masses collected from the following 2 sources: Lymantria mathura is a native polyphagous defoli- infested Russian ships in port areas of the Pacific ator of orchards and hardwood forests in the Russian Northwest in 1991 and in Kavalerovo, Primorskyi Re- Far East, India, northern China, and Japan (Nishigaya gion of the Russian Far East in 1993.The cultures were 1918, Kozhanchikov 1950, Roonwal1953, Wallner et maintained on artificial diet (Bell et at 1981) in al. 1995). A host plant list based on bibliographical USDA-APHIS quarantine facilities at Otis Plant Pro- citations compiled in 1995 by P. Schaefer (U.S. Agri- tection Center (Otis ANGB, MA). Normally, egg cultural Research Service, Newark, DE, personal com- masses were removed from cold storage (5-7°C) after munication) includes 29 species from 13 families. L. 150d and placed in a rearing room at 24°C, =41 % RH, mathura females oviposit both on hardwoods and co- and a photoperiod of 16:8 (L:D) h. nifers (Clausen 1931,Chelysheva and Orlov 1986), but For the host suitability test we used 2-d-old L. there are no reports that larvae feed on conifer foliage. rnathura larvae. Neonates of L.mathura usually cluster on or near the egg mass, imbibe water, but do not feed for the first several days (Kozhanchikov 1950, Anon- ymous 1992). After hatching, we held neonates for 2 d I Department of Entomology,Universityof Massachusetts,Am- in covered 180-ml (6-oz) plastic containers with wa- herst, MA,01003. 20tis Plant Protection Center, USDA-APHIS-PPQ,Otis ANGB, ter-moistened cotton wicks. We randomly selected a MA.02542 sample of 150 neonates to test for each plant species. October 1998 ZLOTINA ET AL.: SURVNAL OF L. mathum ON DIFFERENT TREES 1163
Tnblt~ 1. Trt~t"Hpt'dt"8 (f'riled as potential hosts for L. mathu.ra of a bag were gathered around the main stem, and se- cured with tape. S('it'lltinc name Common name Source We chose only plants with young (=2 wk old after FAGACEAE bud break) and tender foliage in an attempt to keep ()'/I'rrlls ullm I~ White oak Arnold all replicates as uniforn1 as possible. We transferred Arboretum neonates to foliage and placed plants in an environ- (). ",Iwa L. Northern Red oak Arnold Arbor'etum mental chamber with the same conditions as those (). /lrilllls 1.. Chestnut oak Arnold described for the rearing room. We checked the plants Arboretum after 7 d and recorded larval survival, developmental Downloaded from https://academic.oup.com/jee/article/91/5/1162/919949 by UMass Amherst Libraries user on 05 March 2021 (). ('ariaWis BI. (A) Oriental oak Arnold Arboretum parameters, and weight of individual larvae. ().•. dlltilla Lum. Black oak Arnold Data Analysis. We used a chi-square test of heter- Arboretum ogeneity for detecting differences in numbers of lar- (). 1111111.~tris Mut'llchh. Pin oak Arnold vae surviving between host plants. The average pro- Arboretum portion of larvae surviving was calculated from each l"lg,", grcllldifil/ia Ehrh. American becch Warren Co. Nul's. F. s!lll'atiCII 1.. (~:) European beech Musser Forests tree replicate as a ratio of live larvae found at the end JUGLANDACEAE of the 7-d period to the number of larvae used to JlIgIIll"' d/len'u L Butternut Musser Forests initiate the test. Because of the differences in number ]. Iligra I. Black walnut Musser Forests ]. IIIll1ldshl/limMaxim. (A) Manchurian walnut Schumacher Co. of replicates, we did not use analysis of variance BET1.TLACEAE (ANOVA) and pairwise comparison of means for dif- AI,lIts g/l/tillOSll (I..) (E) European alder Musser Forests ferences in the proportion surviving. The proportion CIl/7'iflf"' eam/illillllll Walt. American hornbeam Warrell Co. Nul's. of larvae attaining 2nd and 3rd instar were calculated OLEACEAE as a ratio oflarvae present in each instar divided by the Frtrxinu,<.; 1Ilne';Cllllll L. White ash Musser Forests E Pelllls!l/l'IlIIim ~lllrsh. Green ash Musser Forests total number of surviving larvae. Rate of larval devel- ACERACEA~; opment was compared as numbers attaining the 2nd An .•.saCclUlI1/1/1I\!ttrsh. Sugar maple Musser Forests (and 3rd if applicable) instar by using PROC ANOVA PINACEAE procedure, Scheffe option in MEANS statement in Pi'IWi strullll.~ L. Eastern \Vhite pine Musser Forests AMI'S IIIIL"lIIlI'a (L.) Mill. Bakun fir Musser Forests SAS (SAS Institute 1988). To test differences in larval A. mllmlor (Cord. lIlld White fir Musser Forests weight gain between hosts, we used Kruskal-Wallis CI'>II(I.) Lind!. I-way nonparametric analysis of variance and com- A.fru.,,·,; (Pursh) PoiI' Fraser fir Musser Forests Lmix I/l/idlla (Du Roi) Tamarack Schumacher Co. parison of mean ranks in statistics (Statistix 1992) K. Koch because population variances were heterogeneous P;("¥.'lJ lIt11Igen .."i EIl~('lm, Blue spruce Musser Forests (Bartlett test of equal variances: t' = 351.92,df = 11, PSl'l/dotsuga ml'IlZil'Sii Douglas-fir Schumacher Co. P = 0.000). (mirb.) Franco
A. Asian: E. European: others are native species. Results The overall survival of L. mathura on Fagaceae was By using a CAHN scale (CAHN, Cerritos, CA) we higher than on any other family (Table 2). The highest recorded weights (in milligrams) for 40- 65 individual mean ± SD proportion (0.90 ± 0.13) oflive larvae was larvae of each cohort placed on a host species. on oriental oak, Quercus variabilis Blume. High sur- The trees (Table 1) for the host suitability test were vival was also found on American beech, Fagus gran- grown in a greenhouse from acorns (Arnold Arbore- difalia Ehrh., and European beech, F. sylvatica L. Lar- tUlll, Boston, MA) or seeds (Schumacher, Sandwich, val survival on other host plants in the genus Quercus MA). Several species were obtained for tests as 1- or varied between 0.79 ± 0.20 Q. alba L.) to 0.59 ± 0.36 2-yr-old seedlings (Musser Forests, IN, PA, and War- (Q. palustris Muenchh.). ren County Nursery, McMinnville, TN). A single plant Survival on hardwood trees in other families varied was considered as a replicate. The number of larvae considerably within a family. For example, within the per replicate was dependent on plant size and amount Betulaceae, mean propOltion survival on American of folia!!:enecessary to sustain larvae. The number of hornbeam, Carpinus caroliniana Walt., was 0.51 ± 0.46, replicates was determined by the number of plants whereas no larvae survived on European alder, Alnus available for the test. glutinasa (L.) Gaertn. Within the Juglandaceae, sur- Before placing larvae on foliage,the soilin the planting vival differed from 0.34 ± 0.34for Manchurian walnut, pots Wll~covered with a circular piece of filter paper cut ]uglans mandshurica Maxim., to 0.09 ± 0.20 for black to the center to accept the stem. The paper was then walnut, J. nigra L. Within the Oleaceae, survival varied taped around the bottom of the seedling to prevent from 0.51 ± 0.22 for white ash, Fraxinus americana L., escape of larvae. Small plants were enclosed in a paper to 0.25 ± 0.29 for green ash, F. pennsylvanica Marsh. cylinder (8.5cm dimneter lU1d17cm high) with top and Only 1 larva suvrvived on Acer saccharum Marsh bottom removed. The cylinder was pushed down over (Aceraceae) . the edges of the pot and secured with masking tape. The The species of conifer (Pinaceae) on which L. top ofthe cylinder wascovered with fine nylon mesh and mathura neonates had the highest survival was Dou- held in place with a mbber band. For larger trees, we glas-fir,Pseudatsuga menziesii Franco (Tab Ie 2). When covered each with a bag of fine nylon mesh. The edges newly hatched larvae were reared on Douglas-fir be- 1164 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 91, no. 5
Table 2. Number of L. m.athura larvae tested, proportion of larvae surviving, and instor distribution on different tree Kpt>cit'"K
Proportion of surviving larvae attaining Larvae Number Proportion surviving Host species instar class by end of the test tested of trees (mean ± SO) 2 inst",- 3 instal"
FAGACEAE QlIercus alba 150 10 0.79:!: 0.20 0.81 :!: 0.241 0 Q.mbm 150 10 0.60:!: 0.13 0.06:!: 0.13 0
Q. prinlls 150 10 0.77 ± 0.32 0.37:!: 0.18 0.05:!: 0.17 Downloaded from https://academic.oup.com/jee/article/91/5/1162/919949 by UMass Amherst Libraries user on 05 March 2021 Q. variabilis (A) 150 10 0.90:!: 0.13 0.88 ± 0.19 0.04 ± 0.11 Q. velutina 117 9 0.62 ± 0.30 0.10:!: 0.16 0 Q. palustns 149 10 0.60 ± 0.36 0 0 Fagus grarulifOlia 150 2 0.86 ± 0.07 0.96:!: 0.05 0.48:':: 0.36 F. syloatica (E) 127 8 0.85 ± 0.07 1.00 ± 0.00 0.66 ± 0.27 JUGLANOACEAE JuglallS cinerea 150 9 0.16 ± 0.21 0.01 ± 0.03 0 J. nigra 155 9 0.09 ± 0.20 0 0 J. mandshurica (A) 135 8 0.34 ± 0.34 0.06 ± 0.18 0 BETULACEAE Alnus glutillDsa (E) 150 10 0 0 0 Car7Jinus caroliniana 150 2 0.51 ± 0.46 0.21 ± 0.16 0 OLEACEAE Fraxltlus americana 150 7 0.51 :!: 0.22 0 0 F lIennsylvallica 190 7 0.25:!: 0.29 0 0 ACERACEAE AceI' saccharum 150 5 < 0.01 0 0 PINACEAE Pinus stmblls 150 7 0.13 ± 0.20 0 0 Abies balsamea 150 7 0 0 0 A. concolor 150 5 0.01 ± 0.Q2 0 0 Afraseri 150 6 0.13 ± 0.23 0 0 Lmi.t laricina 90 6 0.12 ± 0.30 0 0 Picea pllngellS 150 10 < 0.01 0.02 0 0 Pseudotsllga menziesii 150 3 0.33 ± 0.27 0 0 yond 7 d, only 2 larvae (1.5%) pupated. However, We measured larval weight gain on the 15 hosts on when Douglas-fir was offered to 20 second instars, 16 which at least 25 larvae survived (Table 3). Kruskal- (80%) pupated. Survival rates of neonate larvae reared Wallis nonparametric comparison of mean ranks on other conifer species were low, ranging from 0.13 ± (Statistix 1992) identified 6 groups within which the 0.23 for Fraser fir, Abiesfraseri (Pursh) Poir, to 0.01 ± means were not significantly different from 1 another 0.02 for white fir, A. concolor (Gord. & Glend.) Lindl. (Kruskal-Wallis statistic, using chi-squared approxi- Only I larva survived on blue spruce, Piceae pungens mation, H = 389.44; P < 0.01; critical Z = 3.37. Para- Engelm., and none survived on balsam fir, Abies bal- metric AOV applied to ranks F = 115.85;df = 11,545; samea (L.) Mill. P < 0.01). Greatest weight gain was in species of Fagus. The rate of development was more rapid when Weight gain in Quercus was intermediate and variable. larvae fed on species of Fagaceae than on species in other families (Table 2). Development was fastest on the 2 species of Fagus. After 7 d on F. sylvatica, all larvae had molted to the 2nd instar and 66% had Table 3. Weight gain of L. nUlthum larvae on difTerent hoot. molted to 3rd instar, whereas on F. grandifolia 96 and Number of Mean weip;ht p;ain Host species 48% of larvae had molted to the 2nd and 3rd instar, larvae (g) :':: SO respectively. Development on species of Quercus was FAGACEAE variable, ranging from 88to 0%oflarvae molting to 2nd Fagus sylvatica 55 12.64 ± 7.39a* instar on Q. variabilis and Q. palustris, respectively. F. grandifolia 61 8.85:' 4.26a Development was faster on species in the white oak QuerCl/s variabilis 37 8.71 :t: 7.41" group (Q. alba, Q. prinus L., Q. variabilis) than on Q. alba 59 1.92 ± 1.53b Q. prilltlS 60 1.80 :+: 1.75bc species of red oak group (Q. velutina Lamarck, Q. nlbra 45 1.00 ± 1.04bcd Q. rubra L., Q. palustris) (F = 42.70; df = 9,66; P < Q. palustris 40 0.57 ± 0.52bcde 0.01). The rate of development on hosts in families Q. velutina 41 0.49 ± 0.66cd",f other than Fagaceae was variable (Table 2). On Amer- BETULACEAE Carpinus caroliniana 50 2.10 ± 2.17b ican hornbeam, Carpinus caroliniana Walt. (Betu- OLEACEAE laceae), 21% oflarvae molted to the 2nd instar, which Fraxinus pennsylvanica 25 0.29 ± 0.42def is significantly more than on Q. velutina and Q. rubra F. americana 44 0.04:' 0.06f (F = 13.90; df = 3, 27; P < 0.01). The rate of larval PINACEAE PseudotslIgil menziesii 40 0.09 :t: 0.09",f development on species of Juglandaceae was lower than on Fagaceae and some Betulaceae (c. carolini- • Means followed by the same lowercase letter are not sip;nificantly ana) but higher than on Oleaceae. different (P> 0.05). October 1998 ZLOTINA ET AL.: SURVIVAL OF L. mathura ON DIFFERENT TREES 1165
The lowest weight gain was in families Oleaceae and (1994) terminated the experiment the 1st larva from Pinaceae. each cohort molted to 2nd instar, whereas our tests lasted 7 d. There were differences between L. mathura and Discussion Asian L. dispar when results on congeneric plant spe- Lymalltria mathllra is a serious pest of hardwoods, cies were compared with results of Baranchikov et al. and outbreaks have been reported from Japan (Nishi- (1994). In our test, 12% of L. mathura survived when gaya 1918), India (RoonwaI1953, 1954). and the Rus- fed Larix lanGina (Du Roi) K. Koch, whereas 80% of Downloaded from https://academic.oup.com/jee/article/91/5/1162/919949 by UMass Amherst Libraries user on 05 March 2021 sian Far East (Kozhanchikov 1950, Chelysheva and Asian gypsy moth larvae survived on L. sibirica Orlov 1986, Maslov et al. 1988). Outbreaks in the (Muenchh.) Ledeb. When we reared L. mathura on RussilUlFar East are often concurrent with those of Pinus strobus L..13% oflarvae survived, whereas 86.7% the Asian gypsy moth (Anonymous 1992). Hosts in- of Asian L. dispar survived on P. sylvestris L. clude species in Salicaceae (willows), Rosaceae (ap- Survival on congeneric plant species was also dif- ple, pear, cherry), Betulaceae (birches), Fagaceae ferent for our studies with L. mathura compared with (oaks, beeches), Anacardiaceae (mango), Combreta- those of Miller et aI. (1991a. b) with North American ceae, Dipterocarpaceae, Sapindaceae, Ulmaceae L. dispar. No L. mathura survived on European alder. (elm), and other broadleaftrees (Kozhanchikov 1950, A. glutinosa, compared with 97% survival of L. dispar Wallner et aI. 1995;P. Schaefer, USDA-ARS, personal on red alder, A. rubra Bong (Miller et aI.1991a). In our communication). Feeding in spring is initiated on study, 22% of L. mathura survived on Douglas-fir, buds, increasing the level of damage (Clausen 1931, whereas survival of Ist-instar L. dispar ranged from 0 Kozhanchikov 1950). During outbreaks in the Russian to 84% (Miller et al. 1991b). depending on tempera- Far East, Q. mongo/ica Fisch. et Turcz., J. mandshurica, ture and foliage age. Miller et al. (1991b) placed lar- and species of Ulmus, Betu/a, and Populus are defoli- vae, <8 h after eclosion, on sprigs of foliage, identified ated (BlU'