Return to METRIA 12

The Acquisition, Maintenance and Breeding Potential of Camellia Germplasm

Clifford R. Parks
Professor of Biology, University of North Carolina at Chapel Hill

A paper from the Proceedings of the 12th Metropolitan Tree Improvement Alliance Conference (METRIA 12), Landscape Plant Symposium: Plant Development And Utilization, held in Asheville, NC, May 23-25, 2002, co-sponsored by the North Carolina State University, North Carolina Division of Forest Resources, USDA Forest Service Southern Region, North Carolina Landscape Association, North Carolina Association of Nurserymen, The Landscape Plant Development Center, The North American Branch of The Maple Society, and The International Ornamental Crabapple Society.

Topics to be discussed

A. Acquisition of germplasm Collections obtained from the wild for specific horticultural objectives (3 examples):

B. Maintenance:

C. Breeding potential (3 examples from Camellia):

Some Sources of Hardiness

Trips in the 1950s and 1960s to collect Camellia japonica from northeastern Japan and Korea represent the first effort to import germplasm from the coldest areas in which Camellia is native. These collections were brought to the U.S. for field testing and breeding. Although plants from northeastern Japan and islands off the coast of Korea near the 38th parallel probably have similar hardiness, the Korean accessions have been much more extensively tested and used in breeding experiments.

Camellia oleifera from the northern end of the range in southeastern China has proven to be very hardy. In 1978 one accession of this species included in a large Camellia plantation at the National Arboretum in Washington, DC was the only survivor after a particularly severe winter. Dr. William Ackerman observed this survival, and initiated a program to breed hardy fall-blooming cultivars by combining this species with C. sasanqua. This program has been very successful. The accession of C. oleifera in the National Arboretum is from the northern end of the range at Lushan, and other accession of this species from that region have also proven to be very hardy.

Camellia tenuivalvis was recently discovered on Longzhoushan Mountain in southern Sichuan near the town of Huili. It is a diploid species closely related to C. saluenensis. It is abundant on the mountain up to a cleared meadow on a ridge at 3200m. It survives in the meadow as shrubs that have sprouted from stumps after deforestation. On the opposite side of the ridge, the mature Lithocarpus forest is intact, and C. tenuivalvis occurs there as a large shrub or small tree under the larger Lithocarpus trees. The flowers in this high elevation population are shades of pink and quite large, and thus the species often makes an excellent display in its natural habitat. It has been very difficult to establish in the eastern United States because it declines in humid, summer heat. We have been successful with several grafts. These flower, and we are able to make hybrids. It is easily hybridized with C. japonica.

Camellia saluenensis (C. pitardii var. pitardii) is widespread in Yunnan and Sichuan. It occurs as shrubs in dry, weedy and deforested fields, but a close check usually shows it to be regenerating from old stumps. Although it survives in dry open fields, it was originally a component in open woodlands. It flowers very freely and can make striking floral displays. It is readily crossed with C. japonica and these hybrids are known as C. X williamsii. These hybrids are a very important group of cultivars in most of the areas where camellias are cultivated. Some of these hybrids make some of our best floral displays in gardens. C. saluenensis was recently found as an understory tree in the forests of Jizushan (Chicken Foot Mountain near Dali, Yunnan). On that mountain it occurs to nearly 3000 m. The accession of C. saluenensis from all localities do not tolerate the hot, humid summers of the eastern United States, and the C. X williamsii hybrids are also not at their best in these climates. In the hybrid group there is variation in resistance to fungal dieback diseases, and thus cultivars can be selected that are more tolerant of hot and humid climates.

A species recently obtained from hills south of the Yangtze River, C. chekiangoleosa, is probably the most closely related species to C. japonica in China. The flowers are larger and have more of an orange cast in the red pigmentation than the Japanese species. It hybridizes readily with other species in Section Camellia. It has been out in test plots for a few years, and has shown no cold injury symptoms so far; however, recent winters have been mild.

Camellia sinensis has been introduced into cultivation from various sources. It occurs over a very wide area in China and Japan, and it is impossible to determine where it was originally native. It has been divided into many different species, but these are all very similar, and difficult to distinguish. Accessions from northern areas, particularly Japan, often have small leaves, are shrubby and are very hardy - perhaps as hardy as any Camellia species. Those from the southern end of the range have large leaves, are more arboreal and are rather cold tender. (There are persistent reports of tea being cultivated in quite cold regions in China, but exact localities for these populations are unknown.)

A species from the hills in Chekiang Province south of Shanghai is C. octopetala. It has cream to yellowish flowers in the autumn, and large leaves with indented veins. It is distantly related to other camellias, and hybrids have not been successfully synthesized yet. It has been growing under garden conditions for several years, and it has survived temperatures down to O°F (-18°C) without apparent injury.

Camellia edithae is only known in cultivation as two double-flowered cultivars. One of these is a deep pink formal that flowers at the end of the flowering season in April. The leaves tend to be more cordate than other Camellia species and are richly textured. It is an excellent garden plant that is hardy at least to O°F (-18°C). This cultivar is very male sterile, and it has not been possible to use it in a crossing program; however, the other cultivar which is peony-flowered can serve as a male parent.

We have recently obtained other species from western China that are ornamental and worth field testing, but no results are available yet. Jimin Gao has collected other high elevation forms from southwestern China which might be sources of hardiness, but these also have not been established in cultivation yet.

The hardiness level required for a given region needs to be carefully determined. Cultivars should be bred and selected for the climate extremes of the region in which they will be grown. In the following outline Camellia hardiness is divided into three temperature zones. Germplasm that might be used to breed cultivars for each zone is listed. Different germplasm should be used to breed cultivars adequately hardy according to the severity of winter cold in each zone.

a. C. oleifera or its hybrids, particularly with C. sasanqua
b. Hardiest selections of C. japonica
c. Hardiest forms of tea
d. High elevation species from western China

a. Many moderately hardy species (C. cuspidata, C. edithae, C. octopetala, etc.), and their hybrids
b. Hardy segregants from combinations between tender species such as C. reticulata and hardy species such as C. japonica or C. sasanqau

a. Most species and hybrids

Yellow Flower-color - Many highly similar species with yellow flowers are native to central and southern Quangxi Province and adjacent areas in south central China. The range of the yellow-flowered species extends into Vietnam where the diversity of the yellow-flowered taxa may be even greater. Near Nanning in central Guangxi we visited steep limestone hills where yellow-flowered species were growing in steep, rocky talus at the base of limestone cliffs. We collected a species named C. longzhouensis and what our guide considered two other species at this site; however, there is every indication these are all one species.

A second locality was visited near Nanning, and there we found a distinctly different yellow camellia, C. nitidissima. Yellow petal-color intensity varied greatly from plant to plant at this site. We also visited the Fangcheng Yellow Camellia Preserve near the Vietnam border in southern Guangxi, and here two different varieties of C. nitidissima were growing, and again there was considerable variability in flower color and size.

Camellia nitidissima can be readily hybridized with C. reticulata, but yellow color is never significantly transferred to the hybrids in these crosses. A small amount of yellow petal-pigmentation can be observed in hybrids between C. nitidissima and C. reticulata - C. japonica hybrids. Hybrids between C. nitidissima and other Camellia species are difficult to synthesize; however, a small number of successful hybrids between C. japonica and C. nitidissima have pale yellow petal-pigmentation. Many other yellow-flowered species are being included in breeding programs now, but hybrids are difficult to synthesize and progress is slow.

The Camellia reticulata Complex - Camellia reticulata is widely cultivated for its large, spectacular blooms. For the most part the plants of this species have an open structure, and they are poorly adapted to cold winters and high summer humidity. It appears to have come into cultivation long ago at religious centers in the hills of northern Yunnan. In the valley of the Jinsha River in northern Yunnan and southern Sichuan, there is a great diversity of Camellia populations related to C. reticulata. In recent years botanist have divided this variation in the Jinshajiang valley into many species, but there is great disagreement as to the number of distinct species. One treatment recognizes 3 species while another lists about 30. Utilizing the natural variation in the Jinsha Valley for breeding, it may be possible to improve the adaptation of C. reticulata for colder and more humid climates.

On the Zixishan (a mountain with ancient temples near Kunming, yunnan, China), there are large populations of the wild forms of the typical hexaploid C. reticulata. There is great variation in flower size and other traits. Some ancient trees, many centuries old, are carefully protected there. These ancient cultivars are like the wild forms except they are double in flower - a trait that may be controlled by a single gene. It is my opinion that these represent the beginnings of cultivation of C. reticulata. Located in the Camellia Temple at Lijiang, perhaps 100 miles further north is an ancient camellia tree that is speculated to be 300 to 500 years old. The flowers and foliage of this tree are essentially identical to the old cultivar trees at Zixi. The great diversity of flower color and foliar morphology in modern cultivars probable comes from crosses between the ancient Zixi cultivars and the diverse populations of the C. reticulata complex.

The Zixishan cultivars stem from the rose-red hexaploids native to that area, but on adjacent mountains at around 7000 feet there are extensive populations of a pale pink to nearly white hexaploid (considered by some to be a different species) that is fully compatible with the red-flowered forms. I have found population of the pink hexaploid occurring to 9000 feet. We are combining these with C. reticulata cultivars with the objective of introducing increased hardiness into C. reticulata. Another source of hardiness is from the hybrids between the hexaploid Kunming reticulatas and the diploid, C. japonica. These hybrids are usually grown in the eastern U.S. in greenhouses for shows. I have found that they segregate for hardiness, and we have selected several that are hardy down to about O°F. The interploid hybrids are rather sterile, and a dead end in breeding while the crosses between the two forms of hexaploids are fully fertile, so we are pursuing the intrahexaploid approach more vigorously.

West of the Jinshajiang Gorge in southern Sichuan, there are many populations of a similar tetraploid with flower colors ranging (in the same populaton) from rose-red to intense red colors. There are many different botanical names applied to different populations in the valley, but they are all very similar. They generally have finer texture than the red hexaploids. East of the Jinshajiang Gorge in southern Sichuan, there are extensive populations of tetraploids that have the same flower colors as the pink hexaploid. Different populations have been given different species names, but they are almost certainly one species. In one mountain population near Xichang, Sichuan there was an extremely variable population that contained some large individuals that were making spectacular floral displays.

In northern Yunnan outside of Hauping near the provincial border with Sichuan there are some populations that are vegetatively very similar to the Zixishan trees but have flower colors much the same as the red tetraploids. These are diploid but have a coarser texture than the tetraploids. In the eastern U. S., these have survived out-of-doors in the recent mild winters and some individuals seem resistant to dieback diseases.

We have collections of all of these forms that are just now beginning to produce blooms. The first few attempts at crossing has produced vigorous seedlings, and we are now setting up a breeding program to "upgrade" C. reticulata for gardens in the eastern U. S. using the rich source of diversity of the "reticulata complex".

Maintenance of Germplasm Collections

Once collections of diverse germplasm has been obtained, it is essential that this resource be preserved for breeding and selection. The need to maintain diversity means that we must maintain large collections. This is essential if the collections are to be used as a foundation for plant breeding. Germplasm collections may be maintained in at least three different ways:

  1. Personal collections maintained by the investigator: Many collectors and breeders maintain their own collections; however, private individuals usually lack the facilities to maintain a large collection, and they can only maintain it for a limited period of time. Many valuable private collection have been lost as their owners found it impossible to maintain them.
  2. Botanic Gardens: Many botanic garden facilities maintain germplasm collections, but since most gardens usually maintain a large number of taxa, they often can only designate a limited amount of space for one plant group. There often is not space for large collections of germplasm from one species or genus. Also, with changes in administration, the botanical focus of a garden may change to the degree that older collections are removed.
  3. Germplasm Repository: These are excellent facilities in which germplasm collections can be maintained for research for an extended period of time. The only limitation is the number of these facilities available.

In the case of Camellia and other slow developing woody plants, it is important that the performance of new selection be field tested along with established cultivars of known response. This step is often bypassed in breeding programs and severe problems may crop up years later as a result. Since field-testing may be beyond the facilities of an individual breeder, public testing facilities should be available.


Camellia germplasm collection have been described, and their use or potential use for specific breeding objectives have been discussed. A outline of some breeding programs with Camellia that involve germplasm recently acquired from the wild condition follows:

  1. Developing winter hardy cultivars
    a. C. oleifera and C. sasanqua Hybrids easily synthesized and fertile; progress has been rapid. (Much research done by Dr. William Ackerman.)
    b. Crosses among wild forms and hardy selections of C. japonica. Hybrids easily synthesized and fertile; progress rapid, but the availability of field trials is a severe limitation.
    c. Species from cold areas, for example, C. tenuivalvis have mostly been easy to hybridize with standard hardier variaties already in cultivation.
    d. Hardy tea accessions are difficult to hybridize with other Camellia species.
  2. Transferring the yellow flower pigmentation from sub-tropical yellow-flowered species to temperate garden Camellia species is difficult. The hybrids are difficult to synthesize, and yellow pigmentation does not readily transfer to hybrids.
  3. Research only initiated: Improving plant form, cold tolerance and disease resistance in Camellia reticulata.

CAMELLIA SPECIES (In order of appearance)

Return to METRIA 12

Web Crafter: Anne S. Napier ~ Email:

Format updated November 3, 2009