Connecticut Chestnut Research: Breeding and Biological Control

PP085 (8/07R)

By Dr. Sandra L. Anagnostakis
Department of Plant Pathology and Ecology
The Connecticut Agricultural Experiment Station
123 Huntington Street
P. O. Box 1106
New Haven, CT 06504-1106

Telephone: (203) 974-8489 Fax: (203) 974-8502

Connecticut's Early Forests

Connecticut was heavily forested in 1600, but by the early 1800's forest covered only about 20% of the state.  Now trees have grown up on land no longer farmed, and we are back to about 60% tree cover.  In 1910, when Chestnut Blight Disease started killing our chestnut trees, half of the standing timber was chestnut and there were about 130 million mature American chestnut trees in the state.  Chestnut was the only wood used for telephone poles, and most of the railroad ties were chestnut.  The trees were tall and straight, and after clear-cutting they easily out-competed the other hardwoods to dominate the forests, making pure stands.  Castanea dentata is native from southern Maine to northern Georgia, all along the Appalachian mountain range.  Ink Disease, caused by Phytophthora cinnamomi, killed trees in the southern coastal part of the range in the early 1800's, and is still present there (5).  We are fortunate that this organism cannot over-winter in Connecticut, and so is rarely a problem here.  The chestnut blight fungus (now called Cryphonectria parasitica (Murr.) Barr) came into the U.S. on Japanese chestnuts, first imported in 1876, and was spread around by mail order (Japanese trees were available from many catalogs) and by everything walking across the cankers (5).  This disease reduced the American chestnuts to understory shrubs, which die, sprout from the base, die and sprout again.  This fungus is now present throughout the original range of C. dentata, and has spread to many of the mid-western places where chestnuts were planted.

Experiment Station History

The Connecticut Agricultural Experiment Station was founded in 1875.  Offices and laboratories are in the city of New Haven on the former estate of Eli Whitney, the inventor.  We have a 50-acre field station in the central Connecticut valley in Windsor, and a 75 acre farm in Hamden, where annual and perennial plantings can be made for experimentation.  Our first plant pathologist, George Clinton, studied the progress of chestnut blight disease, through our native chestnuts.  The pathogen was described, and the species named (Endothia parasitica, Anderson and Anderson), by another of our scientists, Paul J. Anderson.

Chestnut breeding work was begun early in the U.S., but the only program that has continued without interruption is that at The Connecticut Agricultural Experiment Station.  In 1930, Arthur Graves made his first crosses of American and Japanese chestnut, and began a long collaboration with geneticist Donald Jones.  Graves gave us about 9 acres of his land in Hamden, CT with plantings of species and hybrids, to insure the continuation of Connecticut's chestnut breeding program.  Graves' students Hans Nienstaedt and Richard Jaynes made many of the hybrids that we still use today.

My early work at The Experiment Station included studies of the basic genetics of the blight fungus (and the system of vegetative compatibility that restricts hyphal fusion and the transfer of biocontrol viruses from one strain to another), and tests of extracellular enzymes produced by this fungus (1, 2, 13).

Chestnut Breeding in Connecticut

Our breeding plan was first based simply on making hybrids of blight resistant Asian trees with susceptible American trees and testing the hybrids for resistance to Chestnut Blight Disease (4).  When it became clear that at least two genes were responsible for this resistance, we began a back-cross breeding program based on the plan of Charles Burnham (7).  Asian trees are crossed with American trees, and the hybrids (partially blight resistant) are crossed to American trees again.  If there are two resistance genes, one out of four of the progeny from these back-crosses have one copy of both resistance genes, giving them partial resistance.  If there are three genes for resistance, one out of eight of the progeny will have one copy of all three resistance genes.  these trees with partial blight resistance are crossed again to American.  This repeated back-crossing increases the percentage of American genes in the hybrids, and selecting for partial resistance insures passage of the resistance genes.  A final cross of two trees with partial resistance should result in one of sixteen trees having two copies of two resistance genes (or one of sixty four trees having two copies of three resistance genes), which will make them fully resistant to the chestnut blight fungus (7).  We bag female flowers in late June to protect them from pollen, and put selected pollen on the flowers in July and cover them back up again.  Many hybrids are male sterile; the catkins form but the flowers never bloom to produce pollen.  This is only seen in interspecific hybrid trees, but is a feature valued by nut growers who want to plant orchards of male sterile trees with a few pollen producing trees for yields of nuts which are uniform.

We have what is probably the finest collection of species and hybrids of chestnut in the world for use in this breeding program, with all seven species represented.  The breeding work has been greatly helped by a genetic map prepared by Thomas Kubisiak (11).  He has molecular (RAPD) markers for the chromosomal regions (“genes”) associated with resistance to chestnut blight, which could make selection much easier in the future.  Trees of two kinds are being chosen:  for timber (tall and straight, with little energy put into forming nuts) and for nut production (short and spreading with maximum energy put into forming large, good tasting nuts).   Both kinds of trees must have resistance to Chestnut Blight Disease and be well adapted to our climate.  We are starting to select our trees for resistance to Ink Disease, caused by the root pathogen Phytophthora.

Biological Control of Chestnut Blight Disease

In the late 1950’s, a chestnut recovery phenomenon was discovered and studied by Jean Grente in France.  He called the system “hypovirulence,” because the chestnut blight fungus that he isolated had less than normal ability to kill chestnut trees, and the “fungal sickness” could spread.  We found that this is due to a viral parasite of the C. parasitica (9, 5) that is transferred from strain when the hyphae fuse.  The genes of three kinds of these (dsRNA) viruses have now been sequenced, and the viruses placed in the genus Hypovirus by Bradley Hillman and his collaborators (10).  In Connecticut, hypovirulence can keep trees alive in the forest and orchards, and we have studied the many kinds of creatures that move both killing and curing strains of the fungus from tree to tree (3, 5).  There has not been a general recovery of forest chestnuts in Connecticut, as seen in Italy, and our new efforts to improve hypovirulence for biological control of Chestnut Blight Disease include use of genetically engineered strains with the genes of the Hypovirus added to the genome of the fungus.  This work was done by Donald Nuss at the University of Maryland, using Connecticut strains of C. parasitica (8).  We obtained permission from USDA/Plant Quarantine to test these strains in the forest in 1994.  We first made a single release, using plug inoculations and by painting conidia onto the surface of cankers, and we can now say that transgenic strains survive, make spores, transfer virus to new cankers, and initiate cankers (6). The sexual spores that develop on the killing strains in the plot after mating with the new Hypovirulent strains, may carry the genes for Hypovirus.  the virus gene packet segregates as a single gene on the fungal chromosome, and also makes viruses that go into the cytoplasm and can be transferred to other strains by hyphal fusion.

We started testing a population replacement experiment with transgenic strains in 1997, spraying chestnut sprouts at least twice a year for four years with conidia from three transgenic strains, and spraying chestnuts in the control plot with water.  We still cut back competing trees to allow the chestnuts to have the best chance to grow, and monitor the trees to see whether we have effected a stable biological control.  For the last two years we have recovered none of these transgenic strains from the two test plots, but hypovirulence viruses have been found in isolates from many cankers.  Soon we will begin looking for evidence that a biological control can spread through the forest around the plots.

Synthesis of Breeding and Biological Control

American chestnuts

Chestnut seeds of four kinds of hybrids were planted at the Connecticut State Nursery in the spring of 1998.  the resulting 500 chestnut trees were lifted in the spring of 2000 and 100 of them were planted in a clear-cut in Prospect, CT on land owned by the Town of Prospect and managed by the Connecticut Water Company, and 25 were planted in a clear-cut at Sessions Woods Wildlife Area in Burlington.  These are being evaluated for survival under forest competition conditions.  Of the remaining seedlings, the 200 best were planted in an orchard at our substation in Windsor.  The Windsor trees with the best timber form, hardiness, and blight resistance have been selected over the years since planting, and the rest have been removed.  There were 50 trees left in 2006 and their offspring were planted in 2007 in the Goodwin State Forest, in the Farmington Town Forest, at Windsor, and on private land in CT and New York.  We will continue selecting trees from these breeding lines for the best trees to plant in our northern forests.

Ozark chinquapin

Ozark chinquapins (Castanea ozarkensis) are timber trees found on the Ozark Plateau in Oklahoma and Arkansas.  They have a single nut in each bur, as do the Allegheny chinquapins (C. pumila) and Chinese chinquapins (C. Henryi).  Chestnut blight disease has recently reached their native range, and the combination of this disease with forest fire damage and land disturbance seriously threatens their survival.  We have started crossing Ozark chinquapins with Chinese chinquapins and Japanese chestnuts to produce blight resistant trees, and these will be back-crossed to C. ozarkensis to produce trees that will have a better chance of surviving in their habitat.

Project logic

The crosses that have produced blight-resistant trees for timber have, by necessity, used a rather narrow genetic base, even though different trees were used as parents in each generation.  Since many of the native populations of American chestnuts in the Connecticut continue to sprout, by using our biological control, we will be able to keep them alive and flowering.  The same is true of the Ozark chinquapins in Oklahoma and Arkansas.  Now, if we plant resistant trees in the forests where native trees survive, crosses of the native trees with the resistant trees will incorporate blight resistance and all of the native genetic diversity into the future generations.  The first generation offspring will be intermediate in resistance, but in subsequent generations trees with full resistance will be produced.

The Next Problem

Of course, no project is ever quite "finished."  The oriental Chestnut Gall Wasp, Dryocosmus kuriphilus, was introduced into the U.S. in 1974 by a grower who evaded plant quarantine (12).  The insect lays its eggs in leaf and flower buds, resulting in defoliated trees with no flowers.  Entomologist Jerry Payne chronicled the devastation of orchards of Chinese chestnut trees planted in the state of Georgia.  We have reports of infestations throughout Alabama, North Carolina, and into Tennessee, and now in Columbus, Ohio.  Unfortunately, now that the insect has reached the part of Tennessee where most of the mail-order companies get their chestnut trees for retail sale, it is possible that gall wasp will be inadvertently shipped all over the United States, just as the blight fungus was.  Our breeding work must now include selection for resistance to Oriental Chestnut Gall Wasp.  Jerry Payne has observed that American and Chinese chinquapins (Castanea pumila, C. ozarkensis, and C. Henryi), resist infestation, and some cultivars of C. crenata have some resistance.  Once again our collection of species and hybrids is being used to make new progeny for testing in North Carolina where the insect is now endemic.  Preliminary results are encouraging.  Of 93 trees planted in 1995, there were 53 that survived the droughts, deer, rabbits, and weed competition for 12 years.  Among the survivors, 11 had no wasp galls and 25 had few galls.  We hope to understand how resistance is inherited, and will incorporate this resistance into our trees as quickly as possible.

Since we now live in a world where travel and transport of pests and pathogens is all too easy, global communication and cooperation is our hope for the future.


1.      Anagnostakis, S. L.  1977.  Vegetative incompatibility in Endothia parasitica.  Experimental Mycology  1:306-316.

2.      Anagnostakis, S. L.  1988.  Cryphonectria parasitica, cause of chestnut blight.  p123-136  IN:  Advances in Plant Pathology, vol. 6, Genetics of Plant Pathogenic Fungi, G. S. Sidhu, D. S. Ingram, and P. H. Williams, eds., Academic Press, New York

3.      Anagnostakis, S. L.  1990.  Improved chestnut tree condition maintained in two Connecticut plots after treatments with hypovirulent strains of the chestnut blight fungus.  Forest Science 36. 113-124.

4.      Anagnostakis, S. L.  1992.  Measuring resistance of chestnut trees to chestnut blight.  Canadian Journal of Forest Research 22:568-571.

5.      Anagnostakis, S. L.  1995.  The Pathogens and Pests of Chestnuts.  p125-145 IN:  Advances in Botanical Research, vol. 21, J. H. Andrews and I Tommerup, eds., Academic Press, New York. 

6.      Anagnostakis, S. L.  2001.  The effect of multiple importations of pests and pathogens on a native tree.  Biological Invasions 3:245-254.

7.      Anagnostakis, S. L., Chen, B., Geletka, L. M., and Nuss, D. L.  1998.  Hypovirus transmission to ascospore progeny by field-released transgenic hypovirulent strains of Cryphonectria parasitica.  Phytopathology 88:598-604.

8.      Burnham, C. R.  1988.  The restoration of the American chestnut.  American Scientist 76:478-487.

9.      Choi, G. H. and Nuss, D. L.  1992.  Hypovirulence of chestnut blight fungus conferred by an infectious viral cDNA.  Science 257:800-803.

10.  Grente, J.  1965.  Les formes Hypovirulentes d'Endothia parasitica et les espoirs de lutte contre le chancre du châtaignier.  Académie d’Agriculture de France, Extrait du Proces-verbal de la Séance.   51:1033-1037.

11.  Hillman, B. I., Fulbright, D. W., Nuss, D. L., Van Alfen, N. K.  1994.  Hypoviridae.  In. “Virus Taxonomy: Sixth Report of the International Committeefor the Taxonomy of Viruses”  (F. A. Murphy, C. M. Fauquet, D. H. L. Bishop, S. A. Ghabrial, A. W. Jarvis, G. P. Martelli, M. P. Mayo, and M. D. Summers, eds.).  Springer-Verlag, Wein, New York, 580 pp.

12.  Kubisiak, T. L., F. V. Hebard, C. D. Nelson, J. Zhang, R. Bernatzky, H. Huang, S. L. Anagnostakis, and R. L. Doudrick.  1997.  Mapping resistance to blight in an interspecific  cross in the genus Castanea using morphological, isozyme, RFLP, and RAPD markers.  Phytopathology  87:751-759.

13.  Payne, J. A., Green, R. A., and Lester, D. D.  1976.  New nut pest: an oriental chestnut gall wasp in North America. Annual Report of the Northern Nut Growers Association  67:83-86.

14.  Puhalla, J. E. and Anagnostakis, S. L.  1971.  Genetics and nutritional requirements of Endothia parasitica.  Phytopathology  61:169-173.