In 1987 there was great excitement about the breeding program created when The American Chestnut Foundation (ACF) was founded. Dr. Larry Inman who received his PhD degree in plant breeding and genetics under Dr. Charles R. Burnham at the University of Minnesota in 1957, wrote a paper in volume 2 issue 1 of the Journal of the American Chestnut Foundation which proposed a strategy for capturing the genes of local trees.
The proposed strategy has two parts. One is the preservation of the American chestnut germplasm adapted to different geographic zones. The other is the introduction of blight-resistance into each. Begin by using American chestnut survivors in each zone for crosses with Chinese-American chestnut hybrids. First, divide the natural range of the American chestnut into zones, primarily on the basis of climate, latitude, and altitude. Preserving a representative sample of germplasm from each zone is essential. It must be available for the later generations of breeding. All or most of the trees within each zone must be included, since the collections must be large enough to include the adapted ecotypes in each zone.
Dr. Larry Inman - 1987 Journal of ACF
This approach was gradually built into the TACF state chapter system. In 1991 Dr. Fred Hebard wrote that the minimum objective was to produce progeny from twenty trees to capture the needed diversity from any one region. This was further expanded upon by Dr. Albert Ellingboe in his article in the TACF Journal V8 no.1 1994 in which he is quoted ...
Whether trees adapted to one geographic area will perform well in another area is an interesting question. Currently much of ACF's backcrossing program involves crossing to American chestnuts within the vicinity of the Wagner Research Farm in Meadowview Virginia. But the Foundation has many members in different regions who wish to be physically involved in the breeding program, and we plan to make pollen available from the backcrossing program for the purpose of crossing onto American chestnuts in different parts of the country. Pollen from resistant plants (of some generation of the backcrossing program) could be used to pollinate surviving American chestnuts in New York, or Georgia, or Kentucky, and other areas. Crosses made with pollen from the breeding program with regionally adapted American trees should lead to the production of regionally adapted populations of blight-resistant trees.
Dr. Albert Ellingboe - Professor, Plant Pathology and Genetics University of Wisconsin at Madison
In this same issue of the Journal, Dr. Fred Hebard goes on to detail in the article entitled "The American Chestnut Foundation Breeding Plan: Beginning and Intermediate Steps" the requirements of a regional diversity program and provides the basis for some of the metrics used to drive the process.
Our current breeding program focuses on American trees in the vicinity of Meadowview, Virginia, but our goal is to restore the American chestnut throughout its native range. Thus, to preserve adaptations to local conditions, we hope to replicate at least part of the Meadowview breeding effort every few hundred miles from Maine to Georgia. Alternatively, we could breed trees adapted to local conditions by backcrossing highly blight-resistant BC3-F2 trees from Meadowview into local populations followed by a large intercross generation. However, this might require long-term testing to select trees adapted to the local conditions. A few additional backcrosses to locally adapted American chestnut trees prior to intercrossing is a more rapid, but more labor intensive means of achieving this goal.
Dr. Fred Hebard - 1994 Journal ACF - Chief Pathologist ACF
The idea was that each state could best identify their specific and unique ecosystems, and would be in the best position to identify trees from those areas, that could then be pollinated, or contribute pollen, to preserve their genes. Local people could monitor trees for the stage of flowering required for successful pollination, and be ready to perform the pollinations. The trees would then be grown out in the environment suited to where they had competed successfully ecologically. Dr. Hebard provides an outline of the number of trees one would need to plant to have a reasonable confidence that some progeny would show resistance and survive to the next generation.
This approach has been easy to facilitate for State Chapters in the southern region where they can simply take pollen from American chestnut trees to the Meadowview orchards to pollinate trees that have been selected for their attributes of resistance and timber form. It has been more challenging for chapters in the Western Appalachia (Ohio) and North East since by the time pollen is available, the TACF Meadowview trees are too advanced in flowering to be pollinated. For this reason, the New England Chapters, including Connecticut, have chosen to pollinate their own American chestnut parents at the final back-cross stage, using advanced breeding pollen from TACF Meadowview. The result of that pollination is a crop of nuts, with half their genes from the local "pure American chestnut" parent, and about 15/16ths (92.5 percent) American chestnut genes. The balance of the genes being donated from the Asian ancestor which showed strong resistance from the blight. The progeny (nuts) of these controlled pollinations are then planted out in local orchards - usually near where the American chestnut parents were found - and grown to a size (about five years) where they can be innoculated with a virulent form of the blight to identify which are most resistant.
It was mentioned above that Dr. Hebard had outlined an approach for the number of lines needed from a geographic region - twenty. He also outlined the number of trees that would need to be planted to have a high confidence that a representative number with resistance would reach maturity. The result is that we plant about one hundred twenty-five trees per identified pure American chestnut parent. Some of these are controls (pure American, pure "resistant" Chinese, and F1 hybrids) for blight testing when we screen for resistance - usually five to six years after planting. By experience we have found that four lines is about the maximum an orchard manager can successfully manage. Four lines minimizes risk of incompatibility, and also reduces the risk of catastrophic loss in the event of fire, vandalism or other disaster befalling the orchard. This leads to a typical orchard planting of about five-hundred trees covering about one and a half acres.
The model for three genes controlling resistance indicates that out of those five-hundred trees, "conceptually" about twelve percent will show resistance. These sixty or so remaining trees, all of which "conceptually" have half of their genes (except those for resistance) from the local Connecticut sources, are then inter-crossed (controlled or open cross pollination) and their nuts collected for planting in a seed orchard. It should be noted, that the inter-crossing does not change the percentages of American chestnut genes (93.5 percent), nor does it change the percentage of Connecticut genes (50 percent). Rather, that first inter-crossing increases the level of intermediate resistance, and introduces a very small percentage of high degree of resistance.
The selection for resistance of inter-crossed trees will focus on a blend of that small percentage that show a high degree of resistance, with an effort to capture the diversity and local Connecticut genes introduced through the back-cross breeding. It might take several generations of back-crossing to CT trees to recover trees highly adapted to CT; this depends on how local adaptation is achieved, whether through the simple addition of genes for adaptation or the construction of elaborate “adaptive gene complexes.” This might require larger populations of F2s to allow for this. Selection also would occur at F3. This would mean fewer of our B3F3s would survive, but we plan on recurrent selection at that point anyway.
So here we are (or rather we will be in less than ten years) the ultimate product of the vision that Dr. Charles Burnham and Dr. Lawrence Inman conceived in the nineteen eighties. The progeny of an open pollination inter-cross of these highly selected BC3F2 trees (with 50 percent Connecticut genes) will be resistant to blight, and potentially ecologically competitive. In addition, these trees (progeny) will be fully compatible to cross pollinate with any remaining native American chestnuts that might be induced to flower through forest management or chance. The product of such a cross would be even higher percentage Connecticut native - nearing seventy-five percent - but we don't expect early generations of those crosses to have the blight resistance enabling a chance of reaching sexual maturity. The product of a cross between parents having high and low resistance would be intermediate at best.
However, the product of our inter-crosses, BC3F3 material, should be of fifty percent Connecticut origin, and would breed true (reasonably homozygous) in natural settings for resistance. This is the product that we envision will be ecologically competitive in the Connecticut Forests. We are very excited that TACF has this material available now from both the TACF Meadowview and Penn State University programs. But we are even more excited that we'll have material of local Connecticut origin available in the next decade.