77th Annual Report
of the
Northern Nut Growers Association

The Dunstan American X Chinese Hybrid Chestnut,
Castanea dentata X C. mollissima

Chestnut Hill Nursery, Inc.
Route 1, Box 341
Alachua, FL 32615

Dr. Robert T. Dunstan has made many contributions to American horticulture over the last 50 years. He was the first person to successfully hybridize French wine grapes with American muscadines (Vitis vinifera X V. rotundifolia), a feat that eluded researchers for over 100 years (Dunstan 1962). He has worked with many tree species, including jujubes, hicans, and Oriental persimmons, and his work has been published widely, ranging from The Journal of Heredity (Dunstan 1962) and The American Horticultural Magazine (Dunstan 1962) to the Northern Nut Growers Association (NNGA) Journal and Pomona. A professor of linguistics in North Carolina, Dunstan performed his work as a labor of love, an amateur in the true meaning of the word. One of his most important accomplishments has been the breeding of an American X Chinese hybrid chestnut tree with outstanding characteristics.


In 1953, Dunstan received scions of a remarkable American chestnut, Castanea dentata (Marsh) Borkh., growing near Salem, OH, from James Carpentar, a NNGA member and nut tree enthusiast. Carpentar had discovered the tree while pheasant hunting one fall day. The tree was a large, mature individual, in excellent health and loaded with burrs, in a large stand of chestnuts in which all of the other trees were either dead or dying of the blight. No cankers were visible on the entire tree. Carpentar was very impressed with the health of the tree and the quality of its nuts, and he obtained both nuts and scionwood. He then worked with a local plant pathologist to inoculate the tree with active blight and spores over several years, but all inoculations failed to induce any canker formation of the tree whatsoever. It was believed by Mr. Carpentar that the tree had some degree of inherent blight resistance. Unfortunately, the original tree is no longer standing. It never succumbed to blight infection, but ironically was cut with the rest of the grove in the 1960s. Mr. Carpentar has since died, and neither Dr. Dunstan nor the author has been able to locate the pathologist or anyone else with knowledge of the tree.

Realizing the potential of a tree with such exceptional qualities, Dunstan decided to use the trees in a hybridization program. He grafted the scions onto Japanese chestnut (C. crenata Sieb. & Zucc.) rootstock, and the trees grew vigorously and began to bear very early. A thorough search was undertaken and no other chestnuts were located within pollination distance except Chinese chestnut (C. mollissima B1.) growing on his own land, a composite tree top worked with three USDA released selections: Kuling, Meiling, and Nanking. The Chinese cultivars had been obtained from the USDA Research Station at Beltsville, MD.

Seven nuts were produced by the first cross between the Carpentar tree and the Chinese composite. These were planted and by 1962, six of these were bearing. One of these F1 trees was chosen as female parent for the second generation on the basis of clearest evidence of hybridity by leaf shape, tree form, and nut characteristics. This tree was then crossed to both the American and Chinese parent trees, and open pollinated with the other five F1 trees.

In spring 1963, nuts from the 2nd cross were planted in the nursery. In December of that year, these seedlings and more nuts similarly pollinated were brought to Alachua in north-central Florida and planted, where there exists today a grove of 60 trees.


The second generation trees are in excellent health and produce large quantities of nuts each fall. Though some crowding has occurred, some of the trees have reached over 40 feet in height and 12 inches in diameter. Individual trees bear as much as 60 pounds of nuts per tree. As predicted by Rutter and Burnham (1983), the second generation of a Chinese X American hybrid crossed to both parents will produce predominately trees similar in form to the original parent trees. Thirty-eight percent of the Dunstans are American-like, straight-boled timber type tree form, 34% are spreading or multiple boled as with Chinese tree form, and 28% are of intermediate tree form (Fig. 1).

Fig. 1. Dr. Dunstan standing beside an
American form second generation hybrid.

Leaf form is similarly variable: some trees produce leaves distinctly like the American and Chinese parent trees, but most generally appear larger, thicker, and a darker, more lustrous green than in either parent. Leaf pubescence is variable; some trees are highly pubescent and some are completely glabrous. This characteristic follows the general correlation of glabrous to American tree form and pubescent to Chinese form (C. R. Burnham, personal communication). Several trees produce leaves very similar to the narrow and elongate leaf with the acute base of the American species (Fig. 2).

The segregation of characteristics in nut production also shows similarities to both parent species (Fig. 2). The size of the nuts produced in the second generation ranges from 18- 50/lb. The three largest of the Dunstan second generation hybrids produce nuts that weigh 18, 22, and 28 per pound. Nut shape varies from an American-like elongated nut with a long, hairy apical tip (3/4" wide X 5/8" thick X 1 1/8" long) to large, rounded and wide (1 3/8" wide X 1 1/8" thick X 1 1/4" long) Chinese-like nut. Nuts produced from all trees are sweet; none are bland, bready or bitter. Sweetness varies somewhat between varieties. Three of the largest nutted varieties are remarkably sweet, even eaten raw and uncured. It is these three varieties, having the combination of producing heavy crops of large and very sweet nuts and upright spreading (intermediate) tree form, that are considered the outstanding cultivars produced by Dr. Dunstan.

First Patented Chestnut

The Dunstan Hybrid Chestnut 'REVIVAL' is the first chestnut in America to be granted a U.S. Plant Patent. The REVIVAL Chestnut, a third generation cross between two of the best nutted varieties of Dunstan Hybrids, has excellent characteristics and holds Plant Patent No. 5537. These characteristics include an upright growth habit, with the branches spreading in the upper regions of the tree, vigorous growth, and dark, lustrous green foliage. The tree is over 25' tall in its tenth season of growth from seed (Fig. 3).

Fig. 2. Second generation Dunstan hybrid varieties with nuts, leaves, and twigs showing both parental characteristics.

The REVIVAL Chestnut bears consistent, yearly crops (8 lbs. in its fifth bearing season) of extremely large (18/lb. or 25 g. each) nuts (Fig. 4). These nuts are a dark, reddish-chocolate brown in color, and come 2 - 3 per burr. The nuts have an easily removed pellicle, even fresh, and are very sweet tasting, even raw and uncured. The burrs split while still on the tree, and a high percentage of the nuts fall free of the burrs during harvest. Harvest period is short, lasting from 15 September through 1 October, with a peak drop around 25 September here in Florida. The tree responds well to shaking, as a large percentage of the nuts ripen near the same date, and the tree form allows easy removal of the burrs from the tree. Craft compatability on Chinese chestnut rootstock has been very good, with high percentage of takes in the nursery, and continued survival in the test orchards. In all respects the REVIVAL Chestnut is an outstanding cultivar.

Blight Resistance

Neither the first, second or third generation trees have shown any sign of blight infection, either in the orchards at Alachua or at any planting site around the nation. The chestnut blight, Endothia parasitica (Murr.) Anderson, occurs throughout north Florida, where it lives on downy chinquapin, C. alnifolia var. floridana, and on several species of oaks (i.e., Quercus virginiana and Q. falcata), which act as highly resistant hosts to the fungus. Infected chinquapins occur within several miles of the planting site, and both of these oaks occur next to the planting site. The grove is situated atop the highest hill around for several square miles, exposed to winds and fungal spores from all sides. On several occasions, trees have been injured, such as from sunburn, but all lesions have healed completely without showing any evidence of blight intrusion. In no instance have the trees been affected with any evident disease, nor been sprayed for any reason.

Fig.3. Twenty five foot tall Revival chestnut
showing upright spreading form.

In light of the possibility that the trees had escaped natural inoculation, a series of tests were initiated to substantiate the trees' resistance (Michelini 1983). The tests were conducted.by Stan Michelini, a Florida nurseryman, following the instructions of Dr. Ed Barnard, pathologist for the Florida Division of Plant Industry (DPI) at Gainesville.

An isolate of Endothia parasitica was obtained from a girdling canker on a downy chinquapin growing locally. Subcultures of the isolate were grown in the lab. In July 1978, twenty-four one-year-old container grown grafted trees representing four of the Dunstan hybrid varieties were inoculated with plugs of agar containing mycelia and spores of the chestnut blight, and sealed with grafting tape. After 8 weeks, the wrapping tapes were removed, showing the wounds still moist, and there were no visible signs of canker initiation or darkening of the wood.

The trees were grown in the nursery with overhead irrigation and fertilization. After six months, two samples from each cultivar were inspected by DPI. The blight was isolated from only one cultivar, but lacked characteristic canker development. After 14 months, apparently healed wound sites from the four varieties were again inspected by DPI. The fungus was isolated from the four trees and regrown on agar, confirming the fungus had survived the incubation period, but no canker formation was present. Dr. Barnard found that the fungus had survived and had made only shallow penetration into the wood, with no girdling effect from canker growth. All of the other trees showed little to no evidence of canker formation; on some, the inoculation sites were indistinguishable from other bark variations.

All trees were planted and continued to grow. In 1982, one tree died, and upon inspection, it was found that the Chinese rootstock below the graft union had been completely girdled by the fungus, and supported abundant pycnidial growth. Though the scion died from the loss of the rootstock, callus tissue at the graft union resisted spread of the blight into the scion, and no infection was evident on the scion above the graft union. This illustrates the virulence of the strain of blight used in the experiment, and the relative resistance of the Dunstan hybrid chestnut to fungal infection. In similar inoculation experiments performed by the USDA on American chestnuts, no trees survived (Anon. 1917). There have never been any reports of tree death due to blight of any Dunstan Hybrids currently being grown around the nation. Though at no time can the trees be considered completely resistant to the blight, it can be stated that the trees maintain a high degree of resistance in a number of environments.

Further Testing

A planting of several hundred second and third generation grafted trees and seedlings is currently being evaluated for vigor, form, and nut production in the Chestnut Hill orchards in north central Florida. Some of these trees are now over 15 years old and show very promising characteristics, better even than the parent trees.

The Dunstan Hybrids are currently being grown at over 700 locations around the nation, some for more than 15 years. Reports from sales of nursery trees have shown that the trees have a high percentage of survival and grow excellently in climates from Connecticut, central New York, Pennsylvania, Illinois, Michigan, and Ohio, to North Carolina, central Florida and Texas, and on the Pacific Coast from northwestern Washington to central California. Growers also report that the trees are bearing consistently throughout much of this range.

Chestnut Hill Nursery, Inc., is also working with the Connecticut Aglicultural Exp. Station, West Virginia University, N. E. Forest Exp. Station in Delaware, OH, Clemson University, University of California at Davis, University of Minnesota, and the Pennsylvania Bureau of Forestry at Middletown, to further test the blight resistance, growth, and productivity of the Dunstan hybrid chestnuts in differing environments and blight populations.

Fig. 4. Nuts of the Revival chestnut (18/lb. or 25 g. each)


In the nursery, trees are both grown from seed and grafted onto Chinese chestnut rootstock. Since there have not been enough Dunstan Hybrid seed to produce all of our own rootstock, a number of different propagation methods have been tried, with varying degrees of success. Bench grafting (whip and tongue) on transplanted rootstock during the dormant season produces 50% takes but is subject to late spring frost damage. Take % increases to 80% or better if grafted later in the season, as the buds begin to swell. Field grafting (whip and tongue) at leaf out on one year seedling rootstock produces a much larger tree than on transplanted stock, and a high percentage of takes. Spring chip budding, ineffectual on dormant stock (field or bench), produces 80% or better takes if field grafted after leaf out, and will produce an excellent tree if the rootstock is in good condition. Chip budding also maximizes use of budwood, and is quicker and less expensive to accomplish. A good budder can chip 2000 trees per day with two tapers.

The best tree that can currently be produced is one that has been chip budded in the early fall on stock in the field (grown from seed planted the previous spring). In Florida these trees are budded in September and early October, after growth has slowed down for the season. The rootstock averages 1/4" caliper. Trees done by this method produce 80 - 95% takes, depending on variety, and produce the largest, most vigorous tree the following year. Chip budded trees will grow 3 - 4 times as much as a bench grafted tree in the same amount of time, on the same size rootstock.

Fall chip budded trees produce the largest trees for the following reasons: 1) there is no transplant shock at the same time as healing of the graft union, 2) the bud has six months to heal in thus forming a much stronger union with the stock, 3) the bud does not have to initiate growth soon after callusing, and 4) the trees often continue to grow up until late October here in Florida, thus producing a stronger stock than one that has been transplanted in the early spring just before or after grafting. In addition, many areas are subject to late spring freezes that can kill newly formed callus tissue and leaves. Chip buds will stay dormant and can be forced by topping the stock after the last date of frost in the spring, thus eliminating potentially devastating freeze damage.

Trees are fertilized with soluble 20-10-20 twice a month during the spring and early summer. Weed control is accomplished by between the row cultivation with a small harrow, the use of Surflan preemergent herbicide applied after seed germination and during the following dormant season, and once during the summer months if conditions warrant. Harvest is during the short dormant season from December through February.

It is felt that a hybrid chestnut combining blight resistance with good form and quality nut production is important commercially. Public acceptance of seedling Chinese chestnuts has been poor, but cultivars of a chestnut with good characteristics may fit the niche left unfilled since the loss of the American species. There is a demand for high quality- chestnut trees in the market today.

Literature Cited

Anonymous. 1917. Endothia parasitica and related species. U. S. Dept. Agr. Bull.
No. 380.
Dunstan, R. T. 1962.
Some fertile hybrids of bunch and muscadine grapes. J.
Heredity. LIII 6: 299-303.
Dunstan, R. T. 1962.
New grapes for old: California moves east. The American
Horticultural Magazine: 168-171.
Michelini, S. 1983.
Genetic resistance of hybrid chestnut to blight, Endothia
parasitica. Proc. Florida. State Hort. Soc. 96: 239-240.
Rutter, P. A., and C. A. Burnham. 1983.
The Minnesota Chestnut Program&emdash;New
promise for breeding a blight-resistant American chestnut. Ann. Report of
Northern Nut Growers Assoc. 74: 81-90.