| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Plant and Soil Sciences, Oklahoma State Univ., Stillwater, OK 74078
b College of Veterinary Medicine, Oklahoma State Univ., Stillwater, OK 74078
c Texas Coop. Ext., Galveston Office, 5115 Highway 3, Dickinson, TX 77539
* Corresponding author (melaniebayles{at}starband.net)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: CE, cultivar x environment • CL, cultivar x location • CY, cultivar x year • CYL, cultivar x year x location • OAES, Oklahoma Agricultural Experiment Station
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Bridge et al. (1971) compared 13 obsolete cotton cultivars with three then current releases in Mississippi. The two recent cultivars adapted to the Delta Region yielded 111 to 636 kg ha–1 yr–1 more lint than the older ones. The rate of that increase in lint yield was later calculated to be 10.2 kg ha–1 yr–1 over the 1922–1962 time span studied (Bridge and Meredith, 1983). The more recent cultivars generally had higher lint percentages, smaller bolls, smaller seed, and higher micronaire values. Less progress over time was evident in the other fiber properties. In fact, some of the older cultivars had better fiber properties than did the more recent ones (Bridge et al., 1971). In a later study of 17 cultivars, Bridge and Meredith (1983) estimated that the rate of lint yield increase from 1910 through 1978 because of cultivar improvement was 9.5 kg ha–1 yr–1 compared with the 8.6 kg increase in yields actually observed in Mississippi. Cultivars released after 1944 generally had higher yields, smaller bolls, and longer fiber.
Turner et al. (1976) studied genetic advances from 1960 through 1974 in seven unspecified breeding programs using "Regional Cotton Variety Test" data. They calculated lint yield gains of 5 to 17% among the programs. Three of the seven programs increased yield by decreasing the number of seed per boll and increasing lint per seed. The other four did so by increasing boll number per unit area. Six of the programs produced longer fiber, and most had improved fiber strength and micronaire. All seven showed a trend toward lower fiber length uniformity.
Hoskinson and Stewart (1977) compared two obsolete cultivars with four then grown in Tennessee. They reported that breeders had increased lint yield primarily by increasing lint percentage and number of bolls per unit area. Lint percentage increases were attributed to a reduction in seed index and to an increase in lint per seed. The current cultivars had smaller bolls, but more seed per boll. The newer cultivars had increased fiber length and yarn tenacity by 10 to 15% while maintaining acceptable length uniformity and micronaire.
Miller (1977) utilized data from the Regional Cotton Variety Tests, the Alabama Agricultural Experiment Station cultivar tests, and from individual breeding programs to calculate changes in yield potential between 1965 and 1975. In each case, the lint yield of a current cultivar was compared with the cultivar it replaced. From the Regional Test data, yield of the newer cultivars in six coded breeding programs ranged from 7 to 17% higher than the older ones. One of the breeding programs was Paymaster, a company known for its stripper cultivars; but because the programs were coded, we do not know which results apply to it. In the Alabama tests, three coded programs showed yield increases from 2 to 17% over the same time period. At Lubbock, TX, under favorable conditions, the newer cultivars outyielded the older by 16%; but under stress conditions, their advantage was 26%. The cultivars at Lubbock were probably stripper types of cotton.
Wells and Meredith (1984a)(1984b, 1984c) reported several physiological differences between obsolete vs. then modern cotton cultivars, six from the Stoneville and six from the Deltapine breeding programs. The study was conducted in Mississippi with two planting dates. The modern cultivars made earlier, more complete transitions from vegetative to reproductive dry matter partitioning; partitioned more dry matter into reproductive structures without increasing total dry matter; had a greater proportion of their reproductive development at an earlier stage, i.e., when leaf area and mass were at a maximum; and generally produced a greater number of smaller bolls with higher lint percentages. The number of bolls contributed more to lint yield than boll size or lint percentage. Except for micronaire, changes in fiber properties since 1905 were small and showed little association with time of cultivar development. Micronaire was generally higher in cultivars released after 1950.
Meredith and Bridge (1984) reported that average lint yield in the USA remained essentially stable (and low) from 1866 to 1935, increased at a rate of 10.4 kg ha–1 yr–1 from 1936 through 1960, and then declined by 0.9 kg ha–1 yr–1 through 1980. The average rate of genetic improvement from 1910 through 1980 in three studies was continuous and ranged from the 9.5 to 10.2 kg ha–1 yr–1 in Mississippi already discussed (Bridge and Meredith, 1983; Bridge et al., 1971) up to 11.5 kg ha–1 yr–1 in South Carolina. The authors concluded that breeders could continue to increase the lint yield of cotton and that the 1960 through 1980 yield decline was not due to genetic causes.
Bridge and McDonald (1987) analyzed data from 23 yr (1964–1986) of cotton cultivar testing at Stoneville Pedigreed Seed Company in Mississippi. Over that time span, lint yield increased an average of 9.6 kg ha–1 yr–1. The average percentage of cotton harvested at first pick increased 0.5% yr–1, and the time from planting to final harvest decreased at the rate of 1.2 d yr–1. They then partitioned and analyzed the yield data from those tests into 1964–1968, 1969–1974, 1975–1980, and 1981–1986 time segments. Yield trends for those subsets were –66, –19, 45, and 11 kg ha–1 yr–1, respectively. They stated that the yield decline previously documented by Meredith and Bridge (1984) was no longer evident and that earlier maturing cultivars could be partially responsible for that trend reversal.
Among Acala seed stocks in California, Bassett and Hyer (1985) reported genetic advances in lint yield of 9 kg ha–1 yr–1 for the period 1939–1979. However, commercial production after 1960 did not reflect those gains. Fiber length increased significantly after the late 1950s; fiber strength increased steadily throughout the period; and micronaire remained in a relatively narrow, but desirable, range after 1949.
In South Carolina, Culp and Green (1992) studied 29 cultivars and Pee Dee germplasm lines released from 1945 through 1978 in the Eastern Region. Lint yield increases over that time of 10.5 and 15.1 kg ha–1 yr–1 were observed for the then modern vs. obsolete cultivars and among the germplasm lines, respectively. Within genetically related Pee Dee breeding material, the rate of increase was 20.6 kg ha–1 yr–1. Their findings indicate that simultaneous improvement for lint yield and fiber strength can be made in a breeding program if both traits are emphasized.
Meredith et al. (1997) compared eight obsolete cultivars (released 1938–1965) with eight more modern entries (released 1984–1993) on two Mississippi soils at two levels of N. They reported an average lint yield increase of 6.1 and 4.8 kg ha–1 yr–1 over the 16 entries for the high and low N rates, respectively. Yield increases at high N levels were much greater for the obsolete cultivars (9.0 kg ha–1 yr–1) than for the modern entries (1.5 kg ha–1 yr–1). At low N levels, those increases were 7.5 vs. 1.4 kg ha–1 yr–1, respectively.
In Arizona, Moser and Percy (1999) estimated the genetic gain for lint yield in Pima cotton (Gossypium barbadense L.) to be 16.9 kg ha–1 yr–1 from 1949 through 1991. Boll number per unit area and lint per seed increased over time. The more modern cultivars produced longer, more uniform, stronger, and coarser fibers. Tendencies toward greater lint percentage, smaller bolls, and smaller seed were also noted.
The Oklahoma Agricultural Experiment Station (OAES) began plant selection and progeny-row testing in cotton in 1914. Its first cultivar, Oklahoma Triumph 44 (Ware, 1936), was released in about 1918. At some point before 1955, the breeding program was redirected more toward cultivars adapted to stripper, rather than picker, harvest. As near as we can determine, 18 cultivars were released from this program through 1982. Seed of 12 of those 18 were still available at the time these experiments were conducted. The 12 cultivars were evaluated in multiple environments to estimate selection progress over the 1918–1982 time period for lint yield, six fiber properties, eight agronomic characters, and three diseases.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
Yield determinations were based on the weight of hand pulled cotton (equivalent to machine stripped cotton) harvested from each plot, converted into kilograms of lint per hectare. Before harvest, 15 mature bolls were randomly sampled from each plot in each test primarily to measure fiber properties. The samples were ginned on an eight-saw laboratory-type gin, and the lint was forwarded to the Cotton Quality Research Laboratory then at Oklahoma State University. In the laboratory, 2.5 and 50% span lengths were measured on the digital fibrograph and converted into millimeters. Uniformity index was calculated as the ratio of 50 to 2.5% span length, expressed as a percentage. Fiber fineness was measured on the micronaire in standard curvilinear micronaire units. Fiber strength was estimated on the stelometer using 0.0-mm (0-inch) gauge (T0) and 3.175-mm (1/8-inch) gauge (T1) measurements, converted into kilonewton meters per kilogram.
From various weights, measures, and counts taken while ginning the 15-boll samples, several agronomic characters could be calculated. Picked lint percentage was estimated as lint weight converted into a percentage of seedcotton weight; pulled lint percentage as lint weight converted into a percentage of the combined weights of seedcotton plus bur; boll size as weight of seedcotton in grams per boll; bur size as weight of the empty bur in grams per boll; lint weight per boll in grams; lint index as weight of lint in grams per 100 seed; and seed index as weight of 100 seed in grams. In 1982, an additional 15-boll sample was taken from each plot in the irrigated test near Chickasha to measure the lock tenacity (i.e., storm resistance) of the bolls for the cultivars. Samples were taken and measurements were made using the procedures outlined by McCall et al. (1982).
Measurement of Disease Reactions
Three replications of each cultivar in 1981 and four replications in 1982 were planted in a randomized complete-block experimental design on the Plant Pathology Research Farm, Oklahoma State University, Stillwater to determine their reactions to bacterial blight and Verticillium wilt. The soil type was a Norge loam (fine-silty, mixed, active, thermic Udic Paleustoll). Plots consisted of single rows, 6.7 m long, and 1.0 m apart. After emergence, seedlings within rows were thinned to approximately 15-cm intervals. All plots received frequent irrigations to enhance disease development. Other cultural practices were performed as judged necessary.
Reactions to bacterial blight were determined in 1981 by artificially inoculating plants at the six to eight true-leaf stage with an aqueous suspension of the causal organism at a concentration of approximately 5.0 x 105 viable bacterial cells per milliliter. Half of each row was treated with Race 1 of the causal organism, half with Race 2. Inoculum was applied to the abaxial side of leaves with a single-nozzle gun using a power sprayer at a pressure of 1.4 to 2.1 x 106 Pa. Fourteen days after inoculation, individual plants were scored for their disease reactions by the 0.0 (immune) to 4.0 (fully susceptible) grading system described by Brinkerhoff (1963). Before analyses, those grades were converted into a whole-number scale of 0 (for his 0.0 grade), 1 (for 0.1), 2 (for 0.2), 3 (for 1.0), 4 (for 1.2), 5 (for 2.3), and 6 (for 4.0). Reactions were not determined in 1982.
The soil in the area where these experiments were grown was highly infested with the Verticillium wilt causal organism. The first 20 plants in each plot were scored for their Verticillium wilt responses in 1981 and 1982. Plants were evaluated in mid-October on the basis of gross external symptoms and on vascular discoloration in cut stems of those plants without external symptoms. Grades were assigned by the 1 (no visible leaf symptoms; no vascular discoloration in stems) to 10 (defoliated; stems dead down to ground level) scale utilized by Verhalen et al. (1971).
Cultivars were evaluated for their response to the Fusarium wilt–root-knot nematode complex under field conditions as part of the then Regional (now National) Cotton Fusarium Wilt Testing Program at Tallassee, AL. The soil type at that location is a Cahaba loamy fine sand (fine-loamy, siliceous, semiactive, thermic Typic Hapludult). The experimental design for each cooperator in these tests is a randomized complete-block with four replications and with the center two rows in each replication occupied by a susceptible and a resistant check. Because a limit is placed on the number of entries that can be submitted each year, only five of the 12 cultivars in this study were included in the 1982 test (Kappelman, 1982b). Eleven of the 12 entries were included in the Alabama tests in 1 to 10 yr (Johnson and Williams, 1985a, 1985b; Kappelman, 1971, 1972, 1974, 1975, 1976, 1977, 1978, 1982a, 1982b, 1983; Kappelman et al., 1965; Kappelman and Moore, 1981; Lipscomb and Minton, 1964). One of the 12 (i.e., Stoneville 62) was not included in these tests. An entry labeled "Stoneville 62" was included in the 1982 test (Kappelman, 1982b), but later evidence as to its pedigree raised doubt that it was the cultivar developed by the OAES. Although environment affects Fusarium–nematode expression, the same susceptible check (‘Rowden’) was used in all these tests. Therefore, it was possible to compare cultivars tested in different years, with Rowden as the common reference point, with modifications in the technique described by Kappelman (1980). Those modifications entailed the calculation of individual entries (rather than over all entries) and with Rowden rows only within the immediate vicinity of the Oklahoma materials (rather than over the entire test area). Adjustments were accomplished as follows: Y = (AB)C–1 where Y = mean wilting percentage of a given cultivar adjusted for year effects and for location in the field; A = mean wilting percentage of that cultivar within a particular year; B = mean wilting percentage of the Rowden rows within the Oklahoma test material over all years; and C = mean wilting percentage of the Rowden rows within the Oklahoma test material in that particular year. Adjusted mean values for cultivars tested in multiple years were averaged over those years.
Statistical Analyses
Where possible, analyses of variance were combined over years and/or locations for all traits. Because the dryland experiment near Tipton in 1982 was lost, the overall array of experiments was unbalanced. Therefore, analyses were initially conducted for lint yield, each fiber property, and each agronomic character over the seven remaining environments. If the cultivar x environment (CE) interaction was not significant, means over the seven environments were utilized; if significant, analyses were conducted separately for the dryland and irrigated experiments, as suggested by the results of Murray and Verhalen (1970) in Oklahoma. If the CE interaction in the dryland analysis was not significant, means over those three environments were utilized; if significant, analyses were conducted separately in those experiments. The irrigated experiments were balanced over 2 yr and two locations. Thus, environmental effects were partitioned into those components. If the cultivar x year x location (CYL) interaction in the irrigated analysis was significant, analyses were conducted separately for those four tests; if not, the first-order interactions were considered. If the cultivar x year (CY) interaction was significant, years were analyzed separately over locations. If the cultivar x location (CL) interaction was significant, locations were analyzed separately over years. If neither of those interactions were significant, means over the four experiments were utilized. F tests for each source of variation were performed at the 0.05 probability level with the appropriate error term, assuming a mixed model with "cultivars" considered as a fixed variable and with "environments" (or "years" and "locations") as random variables. Linear and quadratic regression analyses were performed to estimate selection progress over time, if any, for the largest data sets possible without significant CE interactions, but with significant differences among cultivars. The major economic traits displaying significant trends were shown in figures; the others are in tables. Linear vs. quadratic presentation of trends was determined independently for each data set on the basis of their respective estimates of the coefficient of determination (R2).
Analyses of variance for reactions to Races 1 and 2 of the bacterial blight organism were conducted separately for the environment used to characterize them. Reaction to Verticillium wilt was analyzed over the two experiments in which it was measured. The significance of its CE interaction was used to decide whether to combine or separate the results of those tests for data presentation. Linear vs. quadratic regressions were calculated to describe the cultivars' reactions over time to the Fusarium wilt–nematode complex.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Significant differences among cultivars were detected for lint yield in six of the seven experiments, all except the irrigated test near Chickasha in 1982. Regression analyses for the six tests displaying significant differences are illustrated in Fig. 1A (dryland) and 1B (irrigated). All trends were linear except for the Tipton irrigated test in 1981 which was more completely described by a quadratic equation. The latter increased the linear test's R2 in that experiment from 0.78 to 0.86. The "fit" of the other five linear equations was not improved by the use of quadratic equations. Thus, five of the six equations indicated that a selection plateau had not yet been reached for lint yield in the OAES breeding program. The sixth suggested otherwise.
|
Fiber Length
A significant CE interaction for 2.5% span length in the combined analysis over all experiments necessitated separate dryland and irrigated analyses. No interaction was detected in the dryland analysis, so those results were averaged over tests. A significant CYL interaction in the irrigated tests required those four experiments to be analyzed separately. Significant differences among cultivars were obtained in all five data sets. Linear regression coefficients exhibited significant increases of 0.04 mm yr–1 on dryland (Fig. 2A)
and of 0.04 to 0.06 mm yr–1 under irrigation for this trait (Fig. 2B). The irrigated values were equal to or greater than those on dryland.
|
|
Uniformity index displayed no significant CE interaction in the combined analysis. Differences among cultivars were detected, but no significant trends were evident over time (Table 2). Similarly, Hoskinson and Stewart (1977) detected no changes over time for uniformity index. However, Moser and Percy (1999) found a 1.5% increase in Pima cotton over a 40-yr period. Conversely, Turner et al. (1976) reported that uniformity index had decreased from 1960–1975 in all seven breeding programs they investigated.
Fiber Fineness
Analyses for micronaire over the seven environments detected a significant CE response. This was also true within the dryland experiments but not the irrigated. Significant differences among cultivars were demonstrated in all four data sets (Table 2); but none of the regressions were significant over time, indicating stabilizing selection. Researchers in several other areas have shown gains in micronaire over time, particularly in cultivars released since 1950–1960 (Bridge et al., 1971; Turner et al., 1976; Wells and Meredith, 1984c; Moser and Percy, 1999).
Fiber Strength
The CE interactions were not significant in the combined analyses for T0 or T1 fiber strength; therefore, only one mean per cultivar was required for each trait over the seven experiments. Differences among cultivars were significant for both. Regression analyses over time for T0 (Table 2) and T1 (Fig. 3)
were significant as well. T0 fiber strength exhibited a linear increase of 0.5 kN m kg–1 yr–1. The overall trend for T1 was better described by a quadratic equation. Since the mid-1940s, the trend for T1 was upward in a generally linear fashion (y = 168 + 0.6x; R2 = 0.64). Increases in fiber strength over time were reported in three studies (Turner et al., 1976; Bassett and Hyer, 1985; Moser and Percy, 1999). Little or no progress over time was observed in several others (Bridge et al., 1971; Bridge and Meredith, 1983; Wells and Meredith, 1984c).
|
|
|
Boll and Bur Size
For boll size, CE interactions were significant in the combined and dryland analyses but not in the irrigated experiments. Differences among cultivars were significant in all four data sets (Table 3). A significant regression was not obtained in the 1981 Chickasha dryland test. Linear increases of 0.02 g seedcotton boll–1 yr–1 were calculated in the 1982 Chickasha dryland experiment as well as in the irrigated experiments. The Tipton dryland results were quadratic suggesting a plateau had been achieved, but that regression was significant only at the 0.10 probability level. The larger bolls contributed at least in part to the yield increases previously noted (Fig. 1A and 1B). The trend toward larger bolls among these cultivars is opposite in direction from that reported in picker-harvested cotton by Bridge et al. (1971) and by Wells and Meredith (1984c).
|
|
Lint index exhibited a significant CE interaction for the combined analysis, but not for the dryland experiments. The CYL interaction for the irrigated tests was not significant, but the CY and CL interactions were. Thus, data were presented over the dryland locations, for separate irrigated locations over years, and for separate years over irrigated locations. Significant differences were obtained in all five data sets (Table 5). Quadratic equations better fit all five data sets than did linear equations. Relatively large improvements were made in lint index through 1964. Since that time, cultivars released by the OAES tended to exhibit similar values. Three of the seven breeding programs studied by Turner et al. (1976) increased lint yield from 1960 through 1975 at least partially by increasing the lint produced per seed.
|
Lock Tenacity
Adverse weather at harvesttime often occurs in Oklahoma which can lead to substantial preharvest loss of seedcotton. In addition, much of the cotton in the state is harvested with once-over, stripper-type machines. As a consequence, the OAES has placed considerable emphasis over the years on developing cultivars with storm resistant or stormproof bolls. Lock tenacity was determined on bolls produced under irrigation in the 1982 experiment near Chickasha. An irrigated test was chosen because McCall et al. (1982) showed that the range of values observed under irrigation was greater and that the separation of boll types was more distinct than on dryland. One location was sufficient because they also showed that CE interaction components for this trait were minimal relative to the cultivar components. Differences among cultivars were significant. A linear regression analysis indicated that lock tenacity had increased over time in this program by an average of 1.7 g force yr–1 (Fig. 6) . None of the other cotton studies comparing breeding trends over time measured lock tenacity.
|
Bacterial blight response trends were similar for Races 1 and 2 of the organism (Fig. 7A and 7B) . When selection was practiced for blight resistance (Periods 2 and 4), acceptable levels were obtained; however, when selection was absent or relaxed (Periods 1 and 3, respectively), the cultivars developed were completely susceptible. The mean resistance grade for Race 1 in Period 4 (Fig. 7A), when selection was practiced for two or more diseases, was significantly higher (i.e., more susceptible) than in Period 2 when selection for only bacterial blight resistance was practiced. However, the mean grade for Period 4 was still significantly lower (more resistant) than the mean grades for Periods 1 or 3 when little or no selection was practiced. The mean grade for Race 2 in Period 2 (Fig. 7B) was not significantly different from that in Period 4, but it was considerably better than those in Periods 1 and 3. One environment was sufficient to measure this disease reaction because artificial inoculation was employed (thus, no escapes) and irrigation was used to optimize disease symptoms.
|
Susceptibility to the Fusarium wilt–root-knot nematode complex decreased (or resistance increased) at an average rate of about 0.5% yr–1 (Fig. 8) . The increase in resistance among OAES cultivars was undoubtedly due to direct selection for that trait in Periods 3 and 4 and was possibly partly due to selection for bacterial blight resistance in Periods 2 and 4 (Bird, 1982).
|
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Several trends described herein were the same as reported in a number of other programs (e.g., fiber strength). At least one was not (i.e., boll size). Some trends were derived in the OAES program (e.g., fiber length) that were not observed in others, specifically the Mississippi Delta, though, admittedly, the Delta already produced much longer fiber than did Oklahoma. At least one trait (i.e., lock tenacity) was studied in this research that was not reported in other such experiments. The plant requirements for stripper harvest differ to some extent from those for picker harvest, necessitating study of that trait.
Bridge et al. (1971) pointed out that cotton genotypes have historically changed to meet changing production conditions and requirements. It is therefore possible that obsolete cultivars may still contain "latent or unnoticed" genes that could be used to improve current and future cultivars. Studies of previous breeding efforts, such as this one, describe what has been accomplished over time and can provide current breeders with some ideas about future possibilities.
Postscripts
After the field portion of this study was conducted, an additional cultivar was released by the OAES. As described in its release notice (OAES, 1986), ‘Cencot’ was higher yielding than ‘Westburn M’ under dryland conditions and had higher picked and pulled lint percentages. It was more resistant to bacterial blight than Westburn M and had a higher micronaire. Westburn M was more resistant to Fusarium wilt–nematodes and had a stronger fiber. Both cultivars displayed similar degrees of earliness, fiber length, uniformity index, and stormproofness.
In 1986, the focus of this cotton breeding project was shifted by the OAES from the development of cultivars to the development of germplasms. The project was terminated by the OAES in early 1993.
One could ask (and rightfully so) why it took so long after the data were collected for this paper to be submitted for publication. Most of the data were collected in 1981 and 1982. Some of the disease grades for Fusarium wilt–root-knot nematodes were taken in 1985. As much as we would have preferred to publish this information earlier, the fact is that we did not. Because of family considerations, Bayles did not complete her Ph.D. dissertation until 1991. Family and other work responsibilities monopolized her time until 2002 when she joined Verhalen in his current job assignment. Work on this paper began shortly thereafter.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Received for publication July 26, 2004.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Related articles in Crop Science:
This article has been cited by other articles:
|
|
 |
|
  B. M. Schwartz and C. W. Smith Genetic Gain in Fiber Properties of Upland Cotton under Varying Plant Densities Crop Sci., July 1, 2008; 48(4): 1321 - 1327. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. W. Smith, P. S. Thaxton, S. Hague, E. Hequet, and D. Jones Registration of TAM 01E-22 Upland Cotton Germplasm Line with Improved Fiber Bundle Strength Journal of Plant Registrations, May 1, 2008; 2(2): 129 - 131. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. M. Schwartz and C.W. Smith Genetic Gain in Yield Potential of Upland Cotton under Varying Plant Densities Crop Sci., March 19, 2008; 48(2): 601 - 605. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
