American chestnut (Castanea dentata) was once a staple of eastern United States forests; however, chestnut blight (Cryphonectria parasitica) has severely threatened its significance. Since 1983, The American Chestnut Foundation (TACF) has engaged in blight-resistance breeding as a step toward restoring C. dentata to the wild. The Georgia chapter of TACF has been producing backcross lines in Georgia since 2005. The goal has been to introduce blight resistance from Castanea mollissima while recovering the C. dentata phenotype via successive backcrossing. This study focuses on the degree to which this combination of traits can be found among seeds and seedlings derived from the first BC3 and BC4 (third and fourth backcross) trees selected for blight resistance in Georgia. Seeds were collected from selected BC3 and BC4, C. mollissima, C. dentata, BC3F3, and C. henryi lines. Seeds were harvested in September 2015 and held at 4oC for 1 month. Seeds were weighed, their volumes measured, and dried material was assayed for sugar content. Duplicate seed batches were held at 4oC until planting in pots in early February 2016. Seedling traits (basal diameter, height, leaf number, leaf length and width) were measured for seedlings of each line on two dates in mid-April 2016.
Results showed that C. mollissima lines tended to differ significantly from C. dentata and backcross lines in seed mass, seed volume, and leaf number, basal area, height, and volume of stems, whereas C. dentata and backcross lines overlapped broadly in these traits. Seed sucrose and total sugar content, and leaf area and leaf length/width ratio showed non-significant trends for differences between C. mollissima and C. dentata/backcross lines. No consistent patterns were found among various lines in seed dry matter, glucose, stem height/basal area ratio, or various estimates of relative growth. These results suggest that seed and seedling characteristics of backcross hybrid chestnuts derived from the TACF breeding program do not differ in significant ways from those of C. dentata. C. henryi is a possible new source of blight resistance for this breeding program. Seeds and seedlings of C. henryi were similar in most respects to those of C. dentata, suggesting that hybrids of these species may produce seeds and seedlings of similar traits. Future plans are to assess lipid, protein, complex carbohydrate, and phenolic content of seed material to determine if these parameters are likewise similar among C. dentata, C. henryi and backcross lines.