Genetic Variation in American chestnut populations in New York State

Ryan MacFee

The American chestnut (Castanea dentata (Marsh.) Borkh.) dominated the eastern deciduous forest comprising as much as 25% of the mature trees and often occurring in pure stands. Its natural range extended north to Vermont, Maine and southern Ontario, west to southeastern Michigan, south to Mississippi and Alabama, and covered over 200 million acres of land. In New York State, chestnut could be found most everywhere except for conifer forest areas at high altitudes in the Adirondack Mountains. The American chestnut tree had profound ecological and economical importance. The agrarian societies of the Appalachian Mountains relied heavily on chestnut for many reasons. The easily worked chestnut wood had many uses including lumber for heavy construction, shingles, paneling, furniture, musical instruments, pulpwood, and fuel. Domesticated animals, such as pigs, were fattened for the winter on fallen chestnuts. The chestnuts were also heavily used by wildlife including squirrels, blue jays, wild turkey, and deer.
Chestnut blight (Cryphonectria parasitica (Murr.) Barr) was first found in the United States in 1904 on trees growing in the Bronx Zoological Park in New York City. Chestnut blight is a fungal pathogen that enters through wounds in the chestnut bark and establishes a canker (FIGURE 1), which eventually girdles the tree killing the branch or stem above the infection site. The blight reached the extreme limits of the chestnut range in 40 years and by 1950 occupied the entire natural range and most of the chestnut trees were dead or dying. The American chestnut survives in forests today as an understory species because of prolific stump sprouting of existing trees (FIGURES 2,3). However, few trees reach flowering age before succumbing to the blight. This lack of genetic crossing has prevented gene flow in the surviving population.
Genetic diversity within a species is a prerequisite for future evolution as there are uncertainties in future environmental conditions and trait values. Natural selection occurs in association with environmental variables, mating systems, gene flow, and genetic drift to act on existing genetic material. A low level of genetic diversity exposes a species to a risk of extinction because of a lack of adaptability. On a practical basis, the assessment of genetic diversity is also important to identify individuals to control a breeding program, identify genotypes for breeding, and for predicting potential genetic gain in a breeding program.
The New York State Chapter of the American Chestnut Foundation has established several restoration orchards within New York State. These orchards serve as important seed sources and a necessary step between breeding and reforestation. (FIGURES 4,5)
During the summer of 2000, I visited five orchards owned and/or maintained by the New York State Chapter of the American Chestnut Foundation (Buffalo, NY) and inventoried each orchard for overall health and quality. All of the trees in these orchards originated from natural chestnut trees in New York State. I took several pieces of data about the physical characters describing the orchard including slope, aspect, size, number of trees and families, elevation, and latitude and longitude. I also took several measurements for each tree within the orchard including form, health, trueness-to-type, height, and diameter at breast height (dbh) (FIGURE 6). I also visited several dozen natural chestnut trees within New York State. These trees are rare enough to have escaped destruction by the blight or have matured enough to produce seed (FIGURE 7). These trees are also identified as "parent trees" by the ACF Chapter in breeding programs and nut collection and distribution. These genotypes therefore have the potential to be dispersed throughout the state and may be interbreeding with other natural trees or in other ACFChapter orchards. I also took several physical characteristics of each of these natural trees including latitude and longitude, dbh, estimated height, and number of seeds harvested. From the latitude longitude data, I created a GIS map of orchard and natural tree locations.
I obtained a total of 110 leaf samples from every visited natural tree and from trees within each orchard. I extracted the DNA each leaf tissue using the DNeasyä Plant Mini Kit (Qiagen, Valencia, CA). Randomly amplified polymorphic DNA (RAPD) primers were used in conjunction with the polymerase chain reaction (PCR) to create a unique genetic fingerprint for each individual tree. RAPD's create a DNA polymorphism based on the amplification of random DNA segments with single primers of arbitrary nucleotide sequence (Williams et. al. 1990). This procedure is similar to those used in humans to determine paternity. It produces a 'DNA fingerprint' for each individual. I used a total of six primers were used for each amplification procedure producing 21 DNA bands. The amplified pieces of DNA can be separated on an agarose gel, stained using ethidium bromide, and visualized under UV light. I determined the size of each band for each primer in each individual from the gel photographs. The presence or absence of a band each individual was used to determine the frequency of that particular band in either the orchard or natural population.
I grouped all samples collected from an orchard together to form one large orchard population for statistical analysis. The nut exchange program of the ACF has ensured that the orchards visited are rather uniform and contained a significant number of the same American chestnut families. The natural population of trees divided into three distinct geographical regions: east, central, and west. I performed several statistical analyses to test for differences within those regions. American chestnut trees are too widely scattered and rare to assume that these are breeding populations, but pre-blight differences may have existed.
I used the following statistical tests to estimate genetic diversity and differences between orchard and natural populations and between east, central, and west regions. I used an analysis of molecular variance (AMOVA) to examine variance components: (i) within and between natural and orchard populations and (ii) within and between west, central, and east regions (WINAMOVA ver 1.55, Excoffier et. al. 1992). The purpose of an AMOVA analysis is to show what percentage of the genetic diversity is found within populations and/or between populations. I also calculated Shannon's diversity estimate for each population as: H0 = -åpi log2 Pi where pi is the frequency of a given RAPD fragment. The calculated numerical value will provide an estimate of total diversity where higher values represent greater genetic diversity. In addition, the between-population variance component was calculated as (Hsp -Hpop)/Hsp, where Hsp is the total diversity in all populations and Hpop the average within-population diversity. The within population variance proportion was calculated as Hpop/Hsp. I calculated Nei's (1978) unbiased genetic distance between populations for variance between natural and orchard populations and between east, central, and west regions. The purpose of this test was to determine how genetically similar the populations were to one another. A large genetic distance between two 0populations would imply that they are genetically similar to one another. This is the condition I was hoping to see between orchard and natural populations as similar levels of geneticdiversity are needed in orchard populations compared to natural populations for successful reforestation.
The results I obtained indicate that American chestnut populations in New York State have a relatively high degree of genetic diversity despite the loss of millions of trees. In the caseof Shannon's diversity index the higher the value, the greater the genetic diversity within the population. Shannon's index is scaled on a range of 0 to 1, with 1 representing the maximum possible diversity for the species. The estimates are considered relatively high, but within the normal range for forest tree species. The genetic diversity estimates were similar to those found for congener species including white spruce (Rajora 1999), trembling aspen, and bigtooth aspen (Liu and Furnier 1993) (TABLE 4).
I also found no significant difference between restoration orchard and natural populations in New York State. All statistical tests reveal little genetic difference between the two populations (TABLE 1). Estimates of Shannon's Diversity Index (Hsp) for orchard and natural populations were not significantly different from one another. The AMOVA and Shannon's results show that most of the diversity is contained within the populations and not between them (TALBE 3). A larger between-population diversity component value would indicate larger genetic distance between the two populations. Nei's genetic distance is a direct measure of similarity where the larger the value, the greater the genetic difference between the two populations. Orchard and natural populations were only 1.9% genetically different from one another. I found, therefore, that the restoration orchards of the ACF are maintaining the natural diversity levels of American chestnut. This is extremely important for breeding blight resistant chestnut trees for reforestation.
I found a small amount of genetic difference amongst the three geographic regions of New York State. The statistical results delineate the regions into two distinct groupings: west/central and east (TABLE 2). My calculated estimates of Shannon's Diversity and genetic distance easily demonstrate this difference. However, the AMOVA statistics I obtained were not significant, indicating that the genetic distance may not be as great as first indicated. This makes sense considering that the environmental conditions across New York State are similar and no natural barriers to gene flow existed before the blight.

Acknowledgements:
I would like to thank Dr. William Powell and Charles Maynard for all their help and suggestions in this research. I would also like to thank Dr. Tom Kubisiak for supplying RAPD primer sequences and reaction conditions, and Bernadette Connors and Dr. Haiying Liang for laboratory assistance and guidance. My greatest appreciation also goes out to the members of the New York State ACF who assisted in the research for this study. Funding for this study was provided by The American Chestnut Foundation, The National Wild Turkey Federation, and a fellowship from the Roosevelt Wild Life Station at ESF.

About The Author: Ryan MacFee

B.A. in Biology from Hartwick College, Oneonta, NY, 1998
M.S. in Environmental and Forest Biology, SUNY-CESF, Syracuse, NY, 2001

My research interests involve using molecular biology techniques to answer forest ecology questions. Most of my past research has involved genetic engineering of plants, especially using Agrobacterium. I hope to continue studying population structure and distribution of plant species using genetic information.

For more information about the project:
American Chestnut Restoration Project (www.esf.edu/faculty/efb/powell/default.htm)

References:
Excoffier L, Smouse PE, Quattro JM (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data Genetics 131 , 479-491.

Huang H, Dane F, Kubisiak TL (1998) Allozyme and RAPD analysis of the genetic diversity and geographic variation in wild populations of the American chestnut (Fagaceae) American Journal of Botany 85 , 1013-1021.

Liu Z, Furnier GR (1993) Comparison of allozyme, RFLP, and RAPD markers for revealing genetic variation within and between trembling aspen and bigtooth aspen Theoretical and Applied Genetics 87 , 97-105.

Nei M (1978) Estimation of average heterozygosity and genetic distance from a small number of individuals Genetics 89 , 583-590.

Rajora OP (1999) Genetic biodiversity impacts of silvicultural practices and phenotypic selection in white spruce Theoretical and Applied Genetics 99 , 954-961.

Williams JGK, Kubelik AR, Livak KJ, Rafalski JA, Tingey SV (1990) DNA polymorphisms amplified by arbitrary primers are useful as genetic markers Nucleic Acids Research 18 , 6531-6535.

Yeh, FC, Chong DKX, Yang R-C (1995) RAPD variation within and among natural populations of trembling aspen (Populus tremuloides Michx.) from Alberta Journal of Heredity 86 , 454-460.