Finding a Replacement for the Eastern Hemlock: Research at the Arnold Arboretum

Peter Del Tredici and Alice Kitajima he hemlock woolly adelgid tsugae, hereafter (AdelgesHWA) is an introduced insect from Asia that was first discovered feeding on eastern (or Canadian) hemlock ( canadensis) in Virginia in the 1950s.1 It did not become a seri- ous problem on the East Coast until the 1980s, when it started killing entire populations of both wild and cultivated trees in the mid-Atlantic region. HWA is now well established in the eastern portion of the range of eastern hemlock, from New Hampshire south to North Carolina,z as well as in most of the range of Carolina hemlock (T. caroliniana). While considerable research has been directed toward developing chemical and biological controls for HWA on eastern and Carolina hemlock,3 relatively little work has been done to determine the resistance of other hemlock spe- cies. In one experiment, McClure’ found that one Japanese species, T. diversifolia (northern Japa- nese hemlock), and two species from western North America, T. heterophvlla and T. mertensiana (western hemlock and mountain This specimen of Tsuga chinensis was collected by E. H Wilson m showed more resistance hemlock) Province, Chma in 1910 It Is now growmg at the Arnold Arboretum. to HWA than Carolina and east- ern hemlock when all tive species were cul- only infrequently in natural populations of T tivated outdoors in Connecticut for one year. diversifolia and T. sieboldii in Japan~ and in a Subsequent fieldwork on native hemlock popu- subspecies of T. chinensis (Chinese hemlock), lations in Asia has shown that HWA occurs in .b These results have been attributed to 34

a combination of host resistance and the pres- second Japanese species, T. sieboldil (southern ence of natural predators. Bentz et al.’ reported Japanese hemlock), showed variable levels of that cultivated specimens of northern Japanese damage. The existing literature on HWA resis- hemlock and Chinese hemlock growing in tance of various hemlock species is summa- close proximity to infected eastern hemlock in rized in Table 1. Washington, D.C., and in Philadelphia, Penn- At the Arnold Arboretum we have been study- sylvania, showed strong resistance to HWA ing the Chinese hemlock as a possible replace- over an eight-year period of exposure, while a ment for eastern hemlock in landscape settings.

Table 1. Comparison of the environmental tolerance factors of various Tsuga species cultivated at the Arnold Arboretum.

Table 2. This hst of Tsuga chinensis accessions received by Arnold Arboretum and ongmatmg m Chma between 1901 and 2002 documents the Arboretum’s persistence m mtroducmg specimens from diverse parts of T. chmensis’s native range. See text for details. 35

Our ten-year research project has focused on three primary goals: reconstructing the history of the introduction of Chinese hemlock into cultivation in North America; documenting its resistance to HWA; and delineating its environ- ’Excellent = outstandmg specimen, good = healthy specimen, no evidence of disease mental tolerances. or physical damage, fair = specimen in decline, evidence of disease or physical damage, poor = spectmen m poor condtuon, more dead branches than Itvtng CHINESE HEMLOCK IN CULTIVATION IN Table 3. Changes m the condition ratmgs of the 1,905 eastern NORTH AMERICA hemlocks (i.e., those mth dbh’s greater than 2 mches (5 cm~J on Hemlock Hill at the Arnold Arboretum between Chinese hemlock is widely distributed at eleva- growmg 1998 and 2002 condition was rated both the tions between 3,282 and 11,487 feet (1,000 and years by same staff members. 3,500 m) in mountainous regions of eastern, central, and southwestern China.8 E. H. Wilson collected and distributed seeds from wild popu- is credited with introducing the species into lations growing at altitudes between 3,282 and cultivation in North America with seed he 8,697 feet (1,000 and 2,650 m) m the provinces collected in Xing Shan, Hubei Province (col- of , Hubei, , Fu~ian, , lection #952), in October 1901, while working and (Table 2).). for the Veitch Nursery Company of Chelsea, England.9 Despite this early introduction, Chi- THE ARBORETUM’S RESEARCH PROJECT nese hemlock remains poorly represented in The north-facing slope called Hemlock Hill North American botanical gardens. One of the consists of approximately 22 acres (10 ha) cov- very few specimens of known provenance was ered with a nearly pure stand of eastern hem- collected by Wilson as a seedling in Fang Xian, lock. Bedrock is close to the surface on much Hubei Province, China, in September 1910. of the hill, and the soils that overlay it are well He sent it to the Arnold Arboretum, where it drained but poor in nutrients. When HWA was arrived in February 1911.’° As of the winter of first discovered there in April of 1997, the hem- 2004, Wilson’s tree was growing under acces- lock population numbered 1,905 individuals sion #17569 and was 49.2 feet (15 m) in height with diameters at breast height greater than with a diameter at breast height of 14.4 inches two inches (five cm). Smce then the pest has (37 cm) and a branch spread of 39.4 feet (12 m). spread rapidly. The entire population was It showed no sign of HWA infestation. labeled, mapped, and qualitatively assessed for Over the years, the Arboretum’s staff has condition and HWA damage during the win- propagated both seedlings and cuttings from ter of 1997-1998. The trees along the base of the Wilson tree and distributed them to vari- Hemlock Hill have been sprayed annually in ous botanical gardens and nurseries throughout the fall with dormant oil since 1997, which has the United States; records show at least sixteen effectively protected them from HWA, but those separate distributions involving twenty-eight in the out-of-reach interior portions have been between 1915 and 1945. Many of the left untreated and are now m a serious state of older Chinese hemlocks now growing in botani- decline. By the winter of 2002-2003, when the cal collections in the United States are direct entire population was recensused, 263 trees had descendants of Wilson’s Hubei seedling. been removed (all were either dead or nearly Apart from Wilson’s collections in 1901 and dead), and those remaining had lost foliage and 1911, wild-collected germplasm of Chinese were in poor health." Table 3 shows the dra- hemlock appears not to have entered North matic decline in the hemlocks’ condition that America again until 1979 and 1980, when vis- occurred between 1998 and 2002 as a result of iting delegations of Chinese botanists presented the HWA infestation. their hosts with seed from the Chinese Acad- For our study of Chinese hemlock’s resis- emy of Forestry. Since then, numerous Ameri- tance to HWA, we used seedlings that were can and European expeditions to China have raised from a seed lot purchased in February 36

’Based on samples of 12 shoots per tree

Table 4 Performance of Tsuga chinensis versus T. canadensls m the Arnold Arboretum study.

1994 and known to have been collected in six topmost shoots on each branch (consisting China in the wild (AA accession #100-94/. After of growth from both 2002 and 2003). The fol- a three-month period of cold stratification, seed lowing was then recorded for all the shoots: was sown m a warm greenhouse. In April 1999, (1) shoot length to the nearest millimeter; (2) when the five-year-old seedlings were between the number of HWA egg sacs found on the 23.4 and 42.9 inches (60 and 110 cm) tall, we undersides of the 2002 shoots; (3) the presence planted 42 of them in scattered light gaps of the or absence of new (2003) growth; and (4) the interior portions of Hemlock Hill, in groups of presence or absence of spider mite damage, three to six individuals. The canopy for all the assessed by looking for the characteristic seedlings consisted of eastern hemlocks that stippling on the upper sides of the 2002 shoots. 12 were badly infected with HWA. At the same The 38 Chinese hemlock seedlings were time, we established a control group by tagging recensused on March 9, 2004, to obtain final 33 seedlings of eastern hemlock that were grow- height measurements for the 2003 growing sea- ing spontaneously on the north-facing slope of son and to learn how many had survived after Hemlock Hill, adjacent to the newly planted an unusually cold winter that saw the tempera- Chinese hemlocks. Unlike the majority (68 per- ture at the Arboretum reaching a low of -8.5 cent) of the Chinese hemlock seedlings, which degrees Fahrenheit (-22.5 degrees centigrade) were growing in light gaps with at least some on January 16. direct sunlight during the day, the majority of RESULTS OF THE STUDY eastern hemlock seedlings (94 percent) were in understory positions that received no direct Results of the study are summarized in Table 4. sunlight at all. The most dramatic finding was the total absence Four years later, on June 25 and 26, 2003, of HWA egg sacs on all of the 38 Chinese hem- we evaluated the growth and extent of HWA lock seedlings, in contrast to a mean of 45.9 egg infestation of the control group and of 38 of the sacs per 12-shoot sample for eastern hemlock. 42 Chinese hemlocks that had been planted in Another indicator of Chinese hemlock’s resis- 1999. By this time the Chinese hemlock seed- tance to HWA was found in the measurement lings had had four years of exposure to HWA of new growth: 100 percent of the terminal buds and the eastern hemlock seedlings six years. on the sampled shoots of Chinese hemlock had We measured the heights of all the seedlings in produced new growth in 2003, compared with both groups and rated their canopy positions as only 45 percent for eastern hemlock. Finally, "sun" (growing in a large canopy gap with mod- the mean shoot length for Chinese hemlock erate amounts of direct sun), "gap" (growing in was 4 inches (10.3 cm), more than twice the a small canopy gap with minimal amounts of 1.9 inches (4.9 cm) recorded for eastern hem- direct sun), or "shade" (growing in complete lock. These results clearly show that Chinese shade). To assess the level of HWA infestation, hemlock possesses a high degree of resistance we selected at random two branches on oppo- to HWA when growing in conditions that are site sides of each tree, taking for evaluation the optimal for infestation of eastern hemlock. 37

, ~~ --- Hemlock Hill as photographed m 1905 by T. E. Marr 38

One of the Tsuga chinensis seedlings (AA #100-94) growmg m a canopy gap on Hemlock Hill at the Arnold Arboretum, photographed in summer 2003. 39

The remeasurement of the 38 Chinese hem- 6 M. E. Montgomery, D. Yao, and H. Wang. 2000. Chmese Coccinellidae for control of the locks on March 9, 2004, showed them to be 177 biological hemlock of native taller than had been the woolly adelgid: Description habitat, percent they previous pp 97-02. In K. A. McManus et al. 1999. summer when their shoot tips were drooping, with an average height of 77.1 inches (197.8 cm) ’ S. E. Bentz, L. G. H. Riedel, M. R. Pooler, and A. M. within a range of 57.3 to 96.9 inches (147.1 to Townsend. 2002. Hybndization and self-compatibillty 248.5 cm). The seedlings had averaged between m controlled pollinations of eastern North American and Asian hemlock (Tsuga) species. Journal of 31.2 and 35.1 inches and 90 in (80 cm) height Arbonculture 28~4/: 200-205. when they were planted out in April 1999; their average growth over four growing seasons was Z.-Y. Wu and P. H. Raven, eds. 1999. Flora of Chma, therefore more than a meter. This is a remark- Vol. 4. Science Press, Beyng, Chma, and Missouri able figure, considering the stressful conditions Botamcal Garden Press, St. Louis, MO. on Hemlock Hill and the minimal aftercare the 9 C. S. Sargent, ed. 1913-1917. Plantae Wilsonianae plants received. It should also be noted that An Enumeration of the Woody Plants Collected m the plants showed very little winter damage at Western Chma for the Arnold Arboretum of Harvard the Years and 1910 the time of the March resurvey, despite the low University During 1907, 1908, by E H. Wilson recorded in 2004. Cambndge University Press, Cambndge, temperatures January MA; A. Rehder. 1940. Manual of Cultivated Trees These results show that Chinese hemlock is and Shrubs, 2nd ed. Macmillan, New York, NY; K. S. fully hardy in USDA Zone 6 and is a suitable Clausen and S. Y. Hu. 1980. Mappmg the collectmg replacement for eastern hemlock in landscape localities of E. H. Wilson m Chma. Arnoldia 40~3/: situations thanks to its 139-145, R. A. Howard. 1980. E. H. Wilson as a relatively rapid growth botamst Arnoldia 102-138. (part I). 40/3/: ~ rate, its tolerance of shade, and its resistance to HWA. ’° Sargent 1913-1917, 3:446; Howard 1980

Endnotes " P. Del Tredici, T. Akin, J. Coop, J. DelRosso, R. Ervm, S. Kelley, A. Kitayma, J. Papargirls, and K. Port. 2003. Hemlock Hill Plant. Arnold ’ R. J. Gouger. 1971. Control of Adelges tsugae on Proposed Management Arboretum Collections internal hemlock m Pennsylvania. Sci Tree Topics 3 1-9. Living Department, report. Jamaica Plain, MA. *. 2 M. S. McClure. 1990. Role of wind, birds, deer, and ’z M. E. 2003. Research scientist, humans m the dispersal of hemlock woolly adelgid Montgomery. (Homoptera: Adelgidae). Environ Entomol 20: 258- USDA Forest Service, Camden, CT. Personal 264, D. A. Ormg, D. R. Foster, and D. L. Mausel. 2002. communication. Landscape patterns of hemlock decline m New England due to the mtroduced hemlock woolly adelgid. Journal Acknowledgments of Biogeography 29:1475-1488. The authors would hke to thank Arnold Arboretum staff 3 McClure. 1995. Managing hemlock woolly adelgid in members Tom Akm, Julie Coop, John DelRosso, Bob Ermn, ornamental landscapes. Bulletm of the Connecticut Susan Kelley, Jim Papargms, and Kyle Port, whose work Agnculture Experiment Station # 925. New Haven, on the Hemlock Hill project made this study possible; CT; McClure, C. A. Cheah, and T. C. Tigner. 2000. Is Mike Montgomery of the USDA Forest Service, Northeast Pseudoscymnus tsugae the solution to the hemlock Region, for providing details on protocols for countmg woolly adelgid problem? An early perspective, pp HWA; and Nathan Havill for his helpful comments on a 89-96. In K. A. McManus, K. S. Shields, and D. R. draft of the manuscript. Souto, eds. Proceedmgs. Symposmm on Sustamable Management of Hemlock Ecosystems in Eastern North America, 22-24 June 1999, Durham, NH. GTR- Peter Del Tredici is a senior research scientist at the NE-267. USDA Forest Service, Northeastern Research Arnold Arboretum. Ahce Kityama now coordinates plant Station, Newtown Square, PA. records at Descanso Gardens in La Canada Flintndge, Cahforma. This article is adapted from "Introduction ’ McClure. 1992. Hemlock woolly adelgid. American and Cultivation of Chinese Hemlock (Tsuga chinensis) Nurseryman 175/6/: 82-89. and Its Resistance to Hemlock Woolly Adelgid (Adelges tsugae)," published in Journal of Arbonculture (2004) 5 McClure et al. 2000. 30~5/: 282-287.