A critical need to evaluate methods to reduce production costs of Upland cotton (Gossypium hirsutum L.) to help maintain its. competitive market position, prompted experiments designed to evaluate the influence of row width and plant population on yield and fiber properties. Upland cotton was grown in 25.4‐, 50.8‐, 76.2‐, and 101.6‐cm‐spaced rows with population levels of 128,000 (low) and 256,000 (high) plants/ha at three locations in Georgia. Significant differences in yield due to test location and to row width were found, but differences in plant populations did not affect yield. Seeding with a modified grain drill produced 88% of the plants desired at the low population level and 82% at the high population level, but left much to be desired in uniformity of seed distribution, placement, and coverage. We found no difference in lint percentage and seed size except that due to location. Boll size was affected by plant population and location but not by row width. The only fiber characteristic affected by row width was fiber fineness. Earliness was influenced only by location. Significant interactions of row width × location were found for yield, number of plants, lint percentage, and fineness. The population × location interaction was significant for boll size and number of plants. The second order interaction of population × row width × location was significant for boll size. Results relative to yield were encouraging and indicated that the narrow‐row, high‐population system might provide partial solutions to some production problems.

A 3‐year study with Upland cotton (Gossypium hirsutum L.) was conducted at Experiment, Ga., to determine the effects of plants/hill, within‐row hill spacings, and plant populations on lint yield. Highest yields were obtained when the populations were within a range of 96,000 to 144,000 plants/ha. Below this population range yield decreased sharply; above this population range yield was reduced but the rate of decline was more gradual than the decline in the low population range. The highest average yield (919 kg lint/ha) was obtained from five plants/hill spaced 40 cm apart (120,412 plants/ha). When all hill spacings were combined, the mean yield from five plants/hill (879 kg/ha) was significantly (P = .05) higher than yield from two plants/hill (812 kg/ha). Yield differences among hill populations of three, four, or five plants were not significant. When the plants/hill treatments were combined, yields from plants spaced 20 and 40 cm apart were significantly higher (P = .01) than yield from plants spaced 60 cm apart. The plants/hill ✕ hill spacings interaction was not significant.

To determine the effects of variations in plants per hill, spacing arrangements, and plant populations on agronomic and fiber properties of cotton (Gossypium hirsutum L.) we conducted a 2‐year study at Experiment, Ga., using ‘Atlas’ cotton on Cecil scl. Each trial was a randomized block split‐plot with number of plants (2, 3, 4, or 5 per hill) as the main plots and within‐row spacings (20, 40, and 60 cm) as subplots. The variations in plants/ hill and spacings represent a range in population of 32,110 to 240,825 plants/hectare. Significant differences in yield due to years and to within‐row spacings were found, but number of plants per hill did not affect yield. The plants in hills spaced 20 and 40 cm apart produced higher yields than the plants spaced 60 cm. A population range of 96,330 to 144,495 plants/ha produced best yields. We found differences in lint percentage and boll size due to years, plants/hill, and within‐row spacings. Lint percentage was highest and bolls were largest from plants at the 60‐cm spacing and when there were 2 plants/hill. The only significant difference in seed size was that associated with years. Fiber length differences were due to years, plants/hill, and within‐row spacings; the longest fibers were produced by 3 plants/hill and by plants at 40‐cm spacing. Differences in fiber strength were due to years and within‐row spacings; the strongest fibers were produced by plants spaced 60 cm apart. Fiber fineness was influenced by years only. The following interactions were found: Years × plants/hill: boll size, fiber length, fiber fineness, yield, lint percentage, and seed size. Years × spacings: lint percentage, boll size, and fiber fineness. Plants/hill × spacings: lint percentage. There were no significant second order interactions. Although mean differences were statistically significant, they are so small that they may be of no practical value. However, results demonstrate the relative stability of the characters stodied over a wide range of planting patterns and population densities.

The purpose of this study was to obtain thrips injury data on Upland cotton varieties currently planted in the Southeast and to associate the effects of the injury with yield. Yield and thrips injury data were obtained on 16 varieties at Experiment, Georgia, in 1960, and thrips injury data on 12 varieties at Jackson, Georgia, in 1961. In 1960, a highly significant (P=.01) negative correlation (0.52) between yield and injury‐index was obtained. The linear regression showed a decrease of 392 pounds seed cotton per acre for each unit increase in thrips injury‐index. A highly significant correlation (0.73) between the 1960 and 1961 thrips injury‐index data was obtained.The order of rank of thrips resistance for varieties tested was ‘Empire,’ ‘DeKalb 108,’ ‘Dixie King,’ ‘Rex,’ ‘Auburn 56,’ ‘Plains,’ ‘Stoneville 7,’ ‘Coker 100A,’ and ‘Deltapine 15.’Empire and varieties having Empire in their genetic background exhibited more resistance to thrips injury than did other varieties.The value of Empire genes in a cotton breeding program for thrips resistance is pointed out.

Seed of ‘Empire WR 61’ and ‘Coker 100A WR’ Upland cotton (Gossypium hirsutum L.) from a common source were selfed or grown in isolated blocks for one generation at nine locations (sources). These seed were subsequently planted at Experiment, Georgia, in 1965 and 1966. Lint yield and seedling vigor data indicated that differences in progeny performance each year, and for both years combined, were associated with seed source. Seed source did not affect lint percent, seed index, boll size, fiber length, fiber strength, and fiber fineness either year. Correlations between progeny yield and vigor rating (r = 0.988 in 1965, r = 0.977 in 1966, and r = 0.988 for the combined data) were highly significant (P≤.Ol) indicating that vigor ratings were good indicators of progeny performance. Correlations between progeny yield and September minimum temperature at seed source (r = _0.519 in 1965, r = _0.553 in 1966, and r = _0.540 for the combined data), and between progeny yield and August rainfall at seed source (r = 0.524 in 1965, r = 0.535 in 1966, and r = 0.532 for the combined data) were significant (P≤.05) and each accounted for about 27% of the variability among seed sources. Higher temperatures and lower rainfall were associated with lower yield and poorer seedling vigor.