One of the most significant, long-term public U.S. Upland cotton {Gossypium hirsutum L.) germplasm enhancement programs is known as the Pee Dee germplasm program. The unique, genetic foundation of the Pee Dee germplasm was created using germplasm from Upland, Sea Island {Gossypium barbadense L.), and primitive diploid cottons. Since the program's inception in 1935, the Pee Dee germplasm program has released >80 improved germplasm lines and cultivars. In this study, the agronomic and fiber quality performance of Pee Dee germplasm was evaluated across southeastern U.S. environments to estimate genetic improvement within the Pee Dee germplasm program. Results suggest that the Pee Dee germplasm enhancement program has (i) maintained usable genetic variation and (ii) maintained high fiber quality potential while concomitantly improving agronomic performance. Although the results highlight the need to continue improving lint percent, lint yield, and bolls m'^, there is also evidence to suggest that Pee Dee germplasm can continue being utilized to develop the next generation of high-fiberquality and high-yielding cotton cultivars.

Genotype x environment (G x E) interactions and trait correlations significantiy impact efforts to deveiop high-yield, high-quality, and environmentally stable Upland cotton {Gossypium hirsutum L.) cultivars. Knowledge of both can and should be used to design optimal breeding programs and effective selection criteria. In this study, we examined the GxE interactions and trait correlations present in the 70-yr Pee Dee cotton germplasm enhancement program. Since beginning in 1935, the Pee Dee program has employed a variety of unique germplasm and breeding methods to release >80 improved germplasm lines and cuitivars. Results suggest that significant G X E interactions exist for several agronomic and fiber quality performance traits that are mostly due to changes in magnitude. Negative genotypic correlations still persist between lint percent/lint yield and fiber length/fiber strength. However, apparently the breeding methods and selection criteria used over 70 yr have lessened the negative relationship between agronomic performance and fiber quality over time to some degree. The results provide cotton breeders a resource to select specific Pee Dee germplasm lines for increased environmental stability. Cotton breeders can also use the information herein to select specific Pee Dee germplasm lines that represent rare recombination events that combine high yield and fiber quality potential.

Eighty-five isolates of Phytophthora infestans from 33 tomato and 8 potato fields in North Carolina, South Carolina, and Tennessee were collected from 1993 to 1995 and tested for mating type, sensitivity to metalaxyl, and allozyme genotype at glucose-6-phosphate isomerase (Gpi) and peptidase (Pep) loci. One isolate collected from potato in eastern North Carolina in 1994 was the A1 mating type, whereas all other isolates from potato and tomato were the A2 mating type. Six isolates were sensitive to metalaxyl (<40% growth at 1.0 μg of metalaxyl per ml), nineteen isolates were intermediate in sensitivity to metalaxyl (>40% growth at 1.0 μg of metalaxyl per ml and <40% growth at 100 μg of metalaxyl per ml), and sixty isolates were resistant to metalaxyl (<40% growth at 1.0 and 100 μg of metalaxyl per ml). Four different allozyme genotypes at the Gpi and Pep loci were identified. The single A1 isolate found on potato in eastern North Carolina had the dilocus allozyme genotype Gpi 86/100, Pep 92/100 and was identified as the US-1 genotype. Fifty-five isolates had the dilocus allozyme genotype Gpi 100/111, Pep 100/100 and were classified as the US-7 genotype, whereas twenty-four isolates were Gpi 100/111/122, Pep 100/100 and were classified as the US-8 genotype. Two isolates that were sensitive to meta-laxyl and seventeen isolates that were intermediate in sensitivity to metalaxyl were found among the US-7 and US-8 genotypes. In addition, five isolates had the allozyme genotype Gpi 100/100, Pep 92/100 (similar to the previously reported US-6 genotype), but they were the A2 mating type and either sensitive or intermediate in response to metalaxyl. These isolates composed a new genotype not previously reported in the United States and were designated as US-18. The US-7 genotype was more frequent on tomato in western North Carolina and the US-8 genotype was present on potato in eastern North Carolina, indicating that different inoculum sources are responsible for epidemics on the two crops in different regions of the state.

In the last 8 yr, cotton (Gossypium hirsutum L.) growers in North Carolina have experienced variations in the year-to-year lint yield averages that range from a 56% increase to a 49% decrease. This variability results in wild fl uctuations in income and a desire for more stable yields. Genetic structure may contribute to stability. This study was conducted to determine the impact of heterozygosity and heterogeneity on lint yield stability. Lint yield was observed in 18 environments over 3 yr among four population types that included homozygous lines grown in pure stands, homozygous lines grown in blended stands, hybrids grown in pure stands, and hybrids grown in blended stands. Comparisons were made using trait means, standard deviations, and the coeffi cients of variation (CVs) calculated over environments. There was no signifi cant difference between lines grown in pure stands and blended lines with respect to yield or stability. Hybrids had a lower CV (were more stable) than homozygous lines. This stability was attributed to the hybrids and blends of hybrids out-yielding the homozygous lines and blends of homozygous lines in the low-yielding environments, but having similar yields in the high-yielding environments. These results do not support growing blends to increase stability or yield; however, growing hybrid cultivars could result in increased yields while reducing variability compared with current production practices.

Historically, the primary objective of cotton (Gossypium hirsutum L.) breeding programs was to improve the quantity and quality of cotton fiber. Because of the added value of cottonseed and its many uses, including a feed and human food source, there is interest in developing cotton breeding programs that focus improvement efforts simultaneously on cotton fiber and seed. Genetic analysis of cottonseed traits, such as protein and oil, is a prerequisite to building new joint fiber and seed cotton breeding programs. In this study, we conducted a genetic analysis of a diverse set of elite upland cotton germplasm for cottonseed protein and oil. Environment was responsible for a sizeable portion of the total variation for protein and oil, and genetics accounted for a larger portion of variation for oil than protein. Genotype × environment (G × E) interactions significantly impacted oil but not protein. Genotypic correlation analysis found a strong, negative relationship between protein and oil. Positive genotypic correlations were found for protein and several agronomic traits including lint yield; whereas, negative correlations were found between oil and lint yield along with other agronomic traits. Overall, results showed very little association between protein, oil, and fiber quality traits. These findings indicate that altering protein and oil seed composition will impact yield and yield component traits. However, alterations in seed composition should not impact fiber quality.