Cold Hardiness Evaluation of Cornus kousa Provenances

Anthony S. Aiello, Morris Arboretum of the University of Pennsylvania, USA

Plant Exploration Background
For the past quarter-century one of the primary missions of the Morris Arboretum has been domestic and international plant exploration and introduction. The goals of this plant exploration program are to: In pursuit of these goals, in the late 1970s the Arboretum identified regions around the world with climates similar to that of Philadelphia and we have been systematically targeting these areas for plant exploration and collection. As a result, since 1979 we have participated in 18 plant collecting trips, to the following areas: These collecting trips have resulted in a living collection that contains over 4,600 plants of wild-collected and documented origin, representing 936 taxa. Our collection has broad holdings in woody plants suitable for the Mid-Atlantic region of the United States and particular strengths in conifers, Hamamelis, Acer, Magnolia, Ilex, and Quercus.

It is within this context of plant exploration that the Morris Arboretum holds a collection of Cornus kousa from known provenances across the natural range of the species in Japan, Korea, and China. Anecdotal information indicates that the Cornus kousa from Korea grows in areas with extremely cold winters. With this in mind, we decided to examine the cold hardiness of Korean kousa dogwoods and a study was begun in late 2000 using freezing tests to examine the lowest survival temperature of stem tissue of plants grown from seed collected in Japan, Korea, and China. The purpose of this paper is to provide background on the introduction of Cornus kousa into the United States and to present the results of three years of freezing studies on the trees of know wild-collected origin.

Cornus Kousa
Cornus kousa is a small flowering tree widely-used throughout much of the eastern and southern United States. It provides a beautiful counterpart to our native flowering dogwood, Cornus florida, with both species being wonderful additions to any garden. Both trees exhibit showy flower bracts, with C. florida blooming before its leaves in late-April to early-May, and C. kousa blooming after its leaves have emerged, usually in June. Because of their differences in blooming time, both are very useful garden plants and E. H. Wilson wrote that, “…Although these dogwoods of North America and the Orient are close relatives they are very dissimilar as garden plants and … there is room for both and no necessity for invidious comparisons” (Wilson 1926). Wilson went on to write of Cornus kousa var. chinensis that, “…some experts acclaim this the finest gift of China to western gardens…certainly it ranks high in the realm of beauty among hardy flowering trees” (Wilson 1926). Wilson’s comments have stood the test of time, with more recent authors extolling the beauty and usefulness of kousa dogwood (Jaynes et al. 1993, Orton 1993).

Scientific and popular literature is replete with information on the problems with the native dogwood, including dogwood anthracnose (Discula destructiva) and powdery mildew (Microsphaera pulchura) (Santamour et al. 1990, Ranney et al. 1995). With a combination of environmental and pathological factors leading to ‘dogwood decline’ in our native dogwoods in the past few decades, there has been increased interest in evaluating and utilizing kousa dogwood for disease resistance and plant (Orton 1985, Ranney et al. 1995). The disease and insect resistance of kousa dogwood led to a hybridization program using C. florida and C. kousa, resulting in the development of the Rutgers University Stellar series of hybrids, such as Aurora®, Constellation®, and Galaxy® (Orton 1990).

Despite the great focus and large number of cultivated varieties of these two highly ornamental species and their hybrids (Jaynes et al. 1993, Santamour 1985), little if any work has been done on increasing their cold-hardiness. Flowering dogwood is grown generally into USDA hardiness zone 5b and kousa dogwood grown occasionally in zone 5a (Flint 1997). Targeted plant collecting holds great potential for expanding hardiness but prior to the 1980s introductions of Cornus kousa from known locations were limited.

Kousa dogwood was first sent to the United States from Japan in 1861 by George Rogers Hall, who was living in Yokohama, Japan; Hall’s shipment of seed was grown in the Boston garden of Francis Parkman in Jamaica Plain, near the present day site of the Arnold Arboretum (Spongberg 1990). The end of the 19th century brought further introductions of kousa dogwood from Japan into the United States; kousa dogwood was offered for sale in the 1890s by Parsons & Sons, Company of Flushing in Long Island, New York, and the Yokohama Nursery Company of Japan. And, the National Plant Germplasm System’s (NPGS) online Genetic Resources Information Network (GRIN; shows additional introductions of kousa dogwood from Japan in the early 1900s.

In 1907 Cornus kousa var. chinensis was first introduced from Hubei Province in central China to the United States by E.H. Wilson of the Arnold Arboretum (Wilson 1916). More recent introductions from China include those from the 1980 Sino-American Botanical Expedition (SABE) to Hubei Province. This introduction of Chinese kousa dogwood (SABE 1316) is known to be held at 7 institutions (Dosmann and Tredici 2003). Cornus kousa var. chinensis was also collected on the 1996 North America China Plant Exploration Consortium (NACPEC) expedition to Shaanxi and Gansu (QLG 026 & 246). A number of young trees from the NACPEC collections be found growing at the Morris, Morton, Holden, and U.S. National Arboreta, the Chicago Botanic Garden, and Longwood Gardens.

As far as can be determined, there are no known records of Cornus kousa being introduced from Korea in the first half of the 20th century. Neither J.G. Jack nor E.H. Wilson, both of the Arnold Arboretum, collected kousa dogwood on their respective 1905 and 1917 trips to Korea. The first known introduction of Cornus kousa from Korea into the U.S. came from Drs. Richard Lighty and Edward Corbett’s highly productive 1965 expedition (P.I. #316695 and #317223). So far I have determine that only three known trees remain from this trip: one at Dr. Lighty’s residence, and one each at the U.S. National and the Arnold Arboreta (National Arboretum #30230 and Arnold Arboretum #1195-68). In the 1980’s teams of plant collectors from American institutions made several trips to South Korea. Among their numerous collections were several of Cornus kousa, including those shown in Table 1. Additional plants of Korean origin can be found at the Morton, National, and Holden Arboreta, Botanic Garden of the University of Copenhagen, and the Sir Harold Hillier Gardens and Arboretum.

Table 1. Accessions of Cornus kousa planted at the Morris Arboretum and used in this experiment. All accessions are of wild-collected and documented origin.
TaxonMorris Arboretum
Accession Number
Cornus kousa86-007*B
1984 Expedition to Korea, Northwestern Coast
Korea, Kyonggi-Do
KNW 243
Cornus kousa86-016*A
1984 Expedition to Korea, Northwestern Coast
Korea, Kyonggi-Do
KNW 278
Cornus kousa86-115*B
1984 Expedition to Korea, Northwestern Coast
Korea, Kyonggi-Do
KNW 1003
Cornus kousa88-043*A
1987 Warner and Howick Expedition to Japan
Japan, Honshu (Aomori)
WH 645
Cornus kousavar. chinensis83-042*A
1980 Sino-American Botanical Expedition
China, Hubei
SABE 1316

The collection of Cornus kousa at the Morris Arboretum was assembled to represent the diversity of the species and to evaluate performance among known provenances of kousa dogwood. Given the limited introduction of kousa dogwood from the northern extremes of its range, the lack of previous testing of kousa for cold hardiness, and the information on potential hardiness of the Korean plants, we decided to compare cold hardiness among plants of known origin from China, Korea, and Japan. Knowing that the Bernheim Arboretum had amassed a field trial of a large number of Cornus kousa cultivars a research project was initiated in the winter of 2001-2002 to examine potential differences in cold hardiness of provenances of Cornus kousa held at the Morris Arboretum by conducting freezing tests of twig samples.

Materials and Methods
Thirteen Cornus kousa plants representing five accessions growing at the Morris Arboretum were selected to represent plants collected across the species’ native range (Table 1). Fifty-five terminal 8 to 10 cm long stem samples for each accession were collected and processed on three dates (December, January, and March) each in 2001-2002, 2002-2003, and 2003-2004.

Stems samples from the Morris Arboretum’s trees were mailed overnight to the Bernheim Arboretum for processing. Samples were subjected to the following controlled freezing protocol. Stems were placed into polyethylene bags and were suspended in a microprocessor controlled (Honeywell, Fort Washington, PA) freezer (model 40-9.4; Scientemp, Adrian, MI). The chamber temperature was then decreased to -5°C over 8h and then decreased at a rate of 4°C/h (Haynes et al., 1992) to a minimum test temperature of -39°C. Five replicate bags were removed from the chamber at 3°C intervals and were placed immediately in a walk-in cooler and held at 5°C for 24h. Stems were then incubated in the sample bags for 7 days at approximately 21°C before evaluation. Temperature was monitored by an Omega HH506R Digital Thermocouple Thermometer (Omega Engineering Inc., Stamford, CT). Air circulation inside the chamber was provided by an internal fan. Control stem samples were prepared as described above and held at 2°C.

Following incubation, samples were sectioned longitudinally through the terminal 4cm of the stem. Tissues were evaluated visually for damage with stems showing any oxidative browning in the vascular region considered damaged (Stergios and Howell 1973). Tissues not injured by the freezing treatments remained green throughout the vascular region and were thus rated as alive. The number of stems damaged at each temperature was recorded, and from these data a percent survival was determined. From this percentage, a lowest survival temperature (LST) was determined as the lowest temperature at which 100% of the stems remained uninjured (Sakai et al. 1986). Data were analyzed by using the Statistical Analysis System (SAS Institute, Cary, NC). Analysis was performed using the General Linear Model (GLM). Means of lowest survival temperatures were separated using the Duncan’s multiple range test.

Analysis of our results indicated no differences in lowest survival temperature (LST) for the year or month that data were collected. Our analysis indicated differences in LST of accessions from different countries (Table 2). There were no significant differences in lowest survival temperature between plants from Japan (-26.3°C) and China (-29.8°C). Plants from Korea showed significantly lower LST (-34.2°C ) than plants from either China or Japan (Table 2).

Table 2. Results of lowest survival temperature (LST) analysis by country (p=0.001).

Further analysis of the data indicated differences in LST among the Korean accessions (Table 3). Accession #86-007 (-36.4°C) showed significantly lower LST than #86-115 (-32.3°C) but showed no difference compared to #86-016 (-34.2°C). There was not a significant statistical difference between accession #86-016 and #86-115 from Korea (Table 3).

Table 3. Results of lowest survival temperature (LST) analysis of Korean accessions (p= 0.01).
Accession #
86-007-36.4°C24 B

The results from this experiment indicate greater lowest survival temperature of plants of Cornus kousa from Korea compared to plants of either Chinese or Japanese origin. Plants from China were collected at ~32° north latitude, those form Korea from 38° north latitude and those from Japan at 41º north latitude. The southern origin of the Chinese kousa dogwood most likely explains the lack of hardiness found in these plants. Latitudinal differences may not account for the differences in hardiness between plants from Japan and Korea. Although the Japanese plants were collected from the northern-most location of any of the plants in this study, they were collected from an area with a well-tempered maritime climate.

Reports from the 1984 plant expedition to northwest Korea noted that Cornus kousa was found growing in areas where minimum winter temperatures routinely reach -30 to -35°C (Paul Meyer, personal communication). Among the Korean accessions, there is overlap in the lowest survival temperature, most likely representing the natural variation among the native populations. The evidence presented in this paper supports the field observations that Cornus kousa of Korean origin holds potential for greater cold-hardiness and for possibly extending the useful northern limit where this species can be grown.

An interesting sidelight from this investigation concerns the commercially available cultivar Cornus kousa ‘Little Beauty’, selected by the late J.C. Raulston at the North Carolina State University Arboretum (J.C. Raulston Arboretum). Background research found that the original plant of ‘Little Beauty’ was grown from seed that Raulston received from the National Arboretum in April, 1986 (Raulston 1993). It turns out that ‘Little Beauty’ was selected from combined seed lots from the 1984 Korea collecting expedition distributed by the National Arboretum (NA #55043 [KNW 243] and NA#55081 [KNW 278], see Table 1). ‘Little Beauty’ was not included in these freezing tests experiments and while no evidence indicates that this cultivar possesses exceptional cold hardiness, future field evaluations will include ‘Little Beauty’ as a comparison.

It is the hope of this author that this study provides insight into potentially greater cold hardiness of Cornus kousa. Currently kousa dogwood is grown to only a limited extent in zone 5, and is seen rarely in Indianapolis, Chicago, or north. Plants able to withstand temperatures of -35°C (~-30°F), that is hardiness Zone 4a, would represent a significant improvement in the hardiness of this species. To this end, we have selected the most attractive plant from each of our three Korean accessions for further evaluation. The next steps in this project are to propagate and grow plants to sufficient size for distribution to arboreta and nurseries throughout USDA zones 4 and 5 for field evaluation. It is my strong hope that our plant exploration efforts from 20 years ago will increase the useful range of this horticulturally desirable species.

Literature Cited

Dosmann, M.S. and P. Del Tredici. 2003. Plant introduction, distribution, and survival: a case study of the 1980 Sino-American Botanical Expedition. BioScience 53:588-597.

Haynes, C.L., O.M. Lindstrom, and M.A. Dirr. 1992. Cooling and warming effects on the cold hardiness estimations of three woody taxa. HortScience 27:1308-1309.

Flint, H.L. 1997. Landscape Plants for Eastern North America. 2nd Edition. John Wiley and Sons, New York.

Jaynes, R.A., A.J. Brand, and J. Arnow. 1993. Kousa Dogwood. Amer. Nurs. 178:40-47.

Orton, E.R., Jr. 1985. Interspecific hybridization among Cornus florida, C. kousa, and C. nuttallii. Proc. Intl. Plant Prop. Soc. 35:655-661.

Orton, E.R., Jr. 1990. New Plant Forum. Proc. Intl. Plant Prop. Soc. 40:635-636. Orton, E.R., Jr. 1993. A family of dogwoods: Three showy relatives welcome spring with clouds of bloom. Fine Gardening March/April:35-39.

Ranney, T.G., L.F. Grand, and J.L. Knighten. 1995. Susceptibility of cultivrrs and hybrids of kousa dogwood to dogwood anthracnose and powdery mildew. J. Arboric. 21:11-16.

Raulston, J.C. 1993. The Chronicles of the NCSU Arboretum pp. 139-154 (#14, July 1986).

Sakai, A., L. Fuchigami, and C.J. Weiser. 1986. Cold hardiness in the genus Rhododendron. J. Amer. Soc. Hort. Sci. 111:273-280.

Santamour, F.S., Jr. and A.J. McArdle. 1985. Cultivar checklists of the large-bracted dogwoods: Cornus florida, C. kousa, and C. nuttallii. J. Arboric. 11:29-36.

Santamour, F.S., Jr., A.J. McArdle, and P.V. Strider. 1990. Susceptibility of flowering dogwood of various provenances to dogwood anthracnose. Plant Disease 73:590-591.

Spongberg, S.A. 1990. A Reunion of Trees. Harvard University Press, Cambridge, Massachusetts.

Stergios, B.G. and G.S. Howell, Jr. 1973. Evaluation of viability tests for cold stressed plants. J. Amer. Soc. Hort. Sci. 98:325-330.

Wilson, E.H. 1916. Plantae Wilsonianae. Vol. II. The University Press, Cambridge, Massachusetts.

Wilson, E.H. 1926. Aristocrats of the Garden. Vol. II. The Stratford Company, Boston.


Figures 1, 2, & 3. Cornus kousa (Accession #86-007*E) in flower at the Morris Arboretum.