Towards the exploitation of the American hazelnut (Corylus americana)

In the Midwest U.S. landscape dominated by the corn-soybean rotation, agroforestry systems can be particularly valuable for increasing the provisioning and regulatory capacity of the agricultural landscape. However, these systems have not yet been broadly integrated into the landscape of this region since they are mostly relegated to marginal lands. A growing body of literature suggests a path to increase the adoption of agroforestry in the Midwest U.S. lies in the incorporation of low-input food-producing tree species that provide economic incentives for farmers. While existing varieties and breeding selections of tree fruits and nuts provide the opportunity for initial system development and integration, their broad adaptability to the Midwest U.S. requires genetic improvement with respective to target environments. This dissertation begins by summarizing literature on the genetic improvement of underutilized temperate U.S. tree crops and their wild relatives, with emphasis on their strategic integration into the Midwest U.S. agricultural landscape. Subsequently, hazelnut is the focus of the three experimental chapters, with the theme of characterizing american hazelnut (Corylus american) germplasm for eastern filbert blight (EFB) resistance. EFB, caused by the fungus Anisogramma anomala, is a primary limitation to european hazelnut (Corylus avellana) cultivation in eastern North America. C. americana is the endemic host of A. anomala and, despite its tiny, thick-shelled nuts, is a potentially valuable source of EFB resistance and climatic adaptation. Interspecific hybrids (C. americana × C. avellana) have been explored for nearly a century as a means to combine EFB resistance with wider adaptability and larger nuts. While significant progress was made in the past, the genetic diversity of the starting material was limited, and additional improvements are needed for expansion of hazelnut production outside of Oregon, where 99% of the U.S. crop is currently produced. Towards this end, this Ph.D. research sought to expand the availability of characterized C. americana germplasm through: i) evaluating american and interspecific hybrid hazelnut (C. americana × C. avellana) germplasm for EFB resistance, ii) evaluating wild american hazelnut germplasm for genetic diversity and structure, and iii) mapping EFB resistance quantitative trait loci (QTL) in C. americana OSU 403.040. In the first study, to improve our understanding of C. americana as a donor of EFB resistance, 29 diverse EFB-resistant C. americana accessions were crossed with EFB-susceptible C. avellana selections (31 total progenies) to produce 2031 F1 plants. Additionally, new C. americana germplasm was procured from across the native range of the species – 1335 plants from 122 seed lots representing 72 counties and 22 states. The interspecific hybrid progenies and a subset of the american collection (616 trees from 62 seed lots) were field planted and evaluated for EFB response following inoculations and natural disease spread over seven growing seasons. EFB was rated on a scale of 0 (no EFB) to 5 (all stems containing cankers). Results showed that progeny means of the interspecific hybrids ranged from 0.96 to 4.72. Fourteen of the 31 progenies were comprised of at least one-third EFB-free or highly tolerant offspring (i.e., ratings 0 to 2), transmitting a significant level of resistance/tolerance. Several corresponding C. americana accessions that imparted a greater degree of resistance to their hybrid offsprings were also identified. In addition, results showed that 587 of the 616 (95.3%) C. americana plants evaluated remained completely free of EFB, confirming reports that the species rarely expresses signs or symptoms of the disease and should be further studied and used in breeding. In the second study, the genetic diversity and structure of new C. americana (272 individuals) collected from 33 seedlots across the species’ native range are reported. Two-thousand fifty-three SNPs were discovered using a genome-by-sequencing approach and support a heterozygous collection (HE = 0.276, HO = 0.280) with moderate differentiation (FST = 0.108) and low inbreeding (FIS = -0.136). Bayesian model-based and neighbor-joining (NJ) clustering corroborate an uppermost clustering level of K = 3. The NJ dendrogram depicts many small subgroups equally distant from common ancestry. Discriminant analysis of principal components reveals between-sub-group variation (K = 15) within the NJ dendrogram and allows the identification of 19 consensus subgroups. Fifty-one accessions were selected for inclusion within a core set based upon 95% representation of the observed allelic variation. Breeders can now exploit the breadth of genetic diversity held within this collection during development of interspecific hybrids. In the final study, a genetic linkage map was developed using a genome-by-sequencing approach and used to identify QTL associated with EFB resistance from the C. americana selection OSU 403.040 from Nebraska U.S. A bi-parental mapping population comprised of 121 seedling trees was evaluated for EFB under high disease pressure in New Jersey, where A. anomala is endemic and highly genetically diverse. With EFB response represented by the percent of diseased wood, a total of three QTLs were discovered on linkage groups (LG) 3, 6, and 11 that respectively represent 62.6%, 23.3%, and 11.1% of the phenotypic variation. EFB resistance from OSU 403.040 appears new based upon it being only the second mapped source from C. americana and due to it mapping to three loci – all other sources of EFB resistance in Corylus spp. are monogenic and map to a single locus. Additionally, OSU 403.040 likely exhibits resistance to a broad range of A. anomala given the genetically diverse A. anomala environment under which it was selected. Such durability is requisite for the development of a feasible commercial variety for the eastern U.S. and highlights a priority for its inclusion in gene pyramiding schemes with resistant C. avellana