Fiber Quality In Upland Cotton Research Articles

Fiber quality of upland cotton under different irrigation depths

ABSTRACTAiming to evaluate the effect of irrigation depths on fiber quality of upland cotton, an experiment was conducted from July to December 2010 in Barbalha-CE, Brazil. The treatments consisted of a factorial combination of two upland cotton cultivars (BRS Aroeira and BRS Araripe) and five irrigation depths (260.93, 418.93, 514.21, 711.81 and 894.68 mm), arranged in a split-plot design with four replicates. A line-source sprinkler irrigation system was used and irrigation depth in the control treatment was calculated according to the crop evapotranspiration. The analysed fiber quality variables were: fiber percentage, length, uniformity, short-fiber index, resistance, elongation at rupture, micronaire index, maturity, degree of yellowing, reflectance degree and count strength product (CSP) index. The irrigation depths influenced fiber length, short-fiber index, strength, micronaire index, maturity and reflectance degree. The cultivars influenced fiber percentage, length and color (degree of yellowing). The best results of fiber quality were found with irrigation depths of 514.21 and 418.93 mm for the upland cotton cultivars BRS Araripe and BRS Aroeira, respectively.

Identification of stable QTLs controlling fiber traits properties in multi-environment using recombinant inbred lines in Upland cotton (Gossypium hirsutum L.)

Cotton fiber is widely used as the raw materials for the textile industry. With the development in spinning technology, the improvement of cotton fiber quality is becoming more and more important. However, negative correlation between yield and fiber quality is an obstacle for cotton improvement. Molecular marker assisted selection provides a potential methodology to break the negative correlation. The main objective of this research was to construct a genetic linkage map and mapping QTL for fiber quality, further to facilitate marker assisted selection for fiber quality traits in Upland cotton (Gossypium hirsutum L.). A genetic linkage map, consisting of 581 loci and spanning 3714.4 cM of the cotton genome, was constructed using recombinant inbred lines population derived from the cross GX1135 × GX100-2, with an average interval of 6.39 cM between adjacent loci. Fiber quality traits were investigated in recombinant inbred lines population sampled by lines from three environments respectively, and each followed a randomized complete block design with two replications. Twenty quantitative trait loci were detected for four fiber quality traits: eight for fiber length, three for fiber strength, five for fiber elongation, and four for fiber micronaire. Among these QTLs, qFL-chr5-2 and qFL-chr10-1 for fiber length, qFS-chr1-1 for fiber strength, and qFM-chr19-1 for fiber micronaire were detected again, which verified the previous results in F2, F2:3 and F2:4 populations (four environments), therefore these major QTLs were stable and especially useful for marker assisted selection to improve fiber quality in Upland cotton.

Combining ability, heritability, and genotypic correlations for lint yield and fiber quality of Upland cotton in delayed planting

Delayed planting is a necessary practice in Upland cotton (Gossypium hirsutum L.) production when double cropping cotton with small grain crops. All current cultivars used for delayed planting were selected under environments of normal planting. The information about genetic values of parents in delayed planting will help determine if it is necessary for testing parents under environments of delayed planting before their use in development of cultivars for the delayed planting system. General combining ability (GCA), specific combining ability (SCA), broad sense heritability, and genetic correlation were compared between two treatments of planting date, the normal planting (mid-April to early May) and delayed planting (late May). Ten F2 hybrid populations derived from a half diallel crossing design and their five parents were planted in 2012 and 2013 with four replicates each year. GCA effects for lint yield and two yield components of boll no. and seeds per boll were different between delayed planting and normal planting. Both direction changes of SCA effects for yield components were observed between treatments of planting date. However, effects of SCA×year were not significant in delayed planting. Reductions in broad sense heritability of yield traits were observed in the delayed planting (0.01–0.25) compared with that of the normal planting (0.10–0.88). Combining ability effects and heritability for fiber properties were similar between treatments of planting date. Genotypic correlation between lint yield and fiber strength became more negative, −0.83, in delayed planting compared with −0.34 in normal planting. The change of genotypic correlations between lint yield and fiber quality in the delayed planting implies an increase of complexity for interrelationships among fiber traits in the double cropped cotton. Overall, the results should inform cotton breeders the need of testing parents for lint yield under environments of delayed planting.

Linkage Map Construction and Quantitative Trait Locus Analysis of Agronomic and Fiber Quality Traits in Cotton

The superior fiber properties of Gossypium barbadense L. serve as a source of novel variation for improving fiber quality in Upland cotton (G. hirsutum L.), but introgression from G. barbadense has been largely unsuccessful due to hybrid breakdown and a lack of genetic and genomic resources. In an effort to overcome these limitations, we constructed a linkage map and conducted a quantitative trait locus (QTL) analysis of 10 agronomic and fiber quality traits in a recombinant inbred mapping population derived from a cross between TM‐1, an Upland cotton line, and NM24016, an elite G. hirsutum line with stabilized introgression from G. barbadense. The linkage map consisted of 429 simple‐sequence repeat (SSR) and 412 genotyping‐by‐sequencing (GBS)‐based single‐nucleotide polymorphism (SNP) marker loci that covered half of the tetraploid cotton genome. Notably, the 841 marker loci were unevenly distributed among the 26 chromosomes of tetraploid cotton. The 10 traits evaluated on the TM‐1 × NM24016 population in a multienvironment trial were highly heritable, and most of the fiber traits showed considerable transgressive variation. Through the QTL analysis, we identified a total of 28 QTLs associated with the 10 traits. Our study provides a novel resource that can be used by breeders and geneticists for the genetic improvement of agronomic and fiber quality traits in Upland cotton.

Genetic Analysis for Quantitative and Quality Characters in Three Single Crosses of Upland Cotton

Gene effects were studied for 12 characters in three crosses of Gossypium hirsutum by six generation mean model to determine the potential for improvement of yield components and fiber quality in upland cotton. Additive genetic variance (gene action) was predominantly operating in the inheritance of boll weight, single plant yield, and lint index. In the presence of epistasis dominance and additive × additive gene effects were observed for boll weight, single plant yield, lint index, ginning outturn, 2.5 percent span length, bundle strength fineness and uniformity ratio. Boll weight, lint index, seed index, ginning outturn, fineness, bundle strength and uniformity ratio were predominantly under the control of dominance and dominance × dominance gene action. Duplicate type epistasis was noticed for all the studied characters and the possible breeding strategies for improvement of lint yield and quality traits were discussed.

Genetic dissection of the introgressive genomic components from Gossypium barbadense L. that contribute to improved fiber quality in Gossypium hirsutum L.

We crossed Luyuan343, containing the Gossypium barbadense (GB) genomic component, with Lumianyan22, a high-yielding Gossypium hirsutum (GH) variety, to evaluate the introgressive exogenous genomic components that contribute to fiber quality in upland cotton. A total of 158 of 334 mapping simple sequence repeats (SSR) loci screened from 18,467 SSR primer pairs were identified as putative GB introgression loci. Twenty-four quantitative trait loci (QTLs) for fiber quality and lint percentage were detected by WinQTLCart 2.5 based on three phenotypic datasets collected over 2 years in two different locations. Of these QTLs, 20 were mapped in the introgression chromosomal regions; the GB alleles contributed to improved fiber quality at eight loci, which were located in the introgression segments of chromosome (Chr.) 3, Chr.7 and Chr.25. A total of eight digenic epistasis and nine significant QTL × environment (QE) interactions were identified by QTLNetwork-2.2. Fiber elongation was found to be most easily influenced by environment since all QTLs for this trait showed significant QE interaction with large effects. Fiber length, fiber strength, and lint percentage were scarcely affected by environment as no or minor QE interaction effects for these traits were detected. Genome-wide identification of the introgressive GB components and mapping of the fiber-related QTLs indentified in this study will be beneficial for the simultaneous marker-assistant selection of improved fiber quality and lint yield in upland cotton breeding.

Construction of a linkage map and QTL mapping for fiber quality traits in upland cotton (Gossypium hirsutum L.)

With the development in spinning technology, the improvement of cotton fiber quality is becoming more and more important. The main objective of this research was to construct a high-density genetic linkage map to facilitate marker assisted selection for fiber quality traits in upland cotton (Gossypium hirsutum L.). A genetic linkage map comprising 421 loci and covering 3814.3 cM, accounting for approximately 73.35% of the cotton genome, was constructed using an F2 population derived from cross GX1135 (P1)×GX100-2 (P2). Forty-four of 49 linkage groups were assigned to the 26 chromosomes. Fiber quality traits were investigated in F2 population sampled from individuals, and in F2:3, and F2:4 generations sampled by lines from two sites and one respectively, and each followed a randomized complete block design with two replications. Thirty-nine quantitative trait loci were detected for five fiber quality traits with data from single environments (separate analysis each): 12 for fiber length, five for fiber uniformity, nine for fiber strength, seven for fiber elongation, and six for fiber micronaire, whereas 15 QTLs were found in combined analysis (data from means of different environments in F2:3 generation). Among these QTLs, qFL-chr5-2 and qFL-chr14-2 for fiber length were detected simultaneously in three generations (four environments) and verified further by combined analysis, and these QTLs should be useful for marker assisted selection to improve fiber quality in upland cotton.

Breeding for improved yarn quality: Modifying fiber length distribution

For the last two decades, cotton breeders have used HVI (High Volume Instrument) as their primary and often sole source of fiber quality data when making plant selections. Fiber data generated by Advanced Fiber Information System technology is also now available to plant breeders, and provides additional information on length characteristics and fiber maturity. Two methods of evaluating fiber quality of upland cotton (Gossypium hirsutum L.) in a breeding program were compared. Yarn quality of cottons selected using data from each of these methods was evaluated. One method used only HVI data for plant selections while the other method used only data from AFIS. One critical difference between the selection methods was the use of fiber length distributions as the only selection criteria in the AFIS method. Line development began with 15 intraspecific F2 populations in 2005. Selections were made in the F2 and F3 generations using fiber data generated from either HVI or AFIS instruments. Evaluation continued until F5 lines were selected in 2007. In the F5 generation, 10 lines from each selection method and 4 commercial varieties were planted at three locations. Whole plots were harvested for each location and carded 30Ne count ring spun yarns were produced from each of the F5 breeding lines and 4 commercial varieties. Yarn properties evaluated include elongation, single end tenacity, work of rupture, coefficient of variation, thin places, thick places, neps, and hairiness.Both selection methods resulted in F5 lines with better fiber quality and yarn quality than commercial varieties. F5 lines from both methods had lower short fiber content by number, SFCn, than the commercial varieties. Yarn tenacity, elongation, CV, thin places were all improved by both selection methods compared to commercial varieties. Results from different locations indicate fiber maturity has a significant impact on various fiber and yarn properties. Fiber and yarn data indicate it is possible to improve fiber length distribution using either selection method.It is important for breeders to understand the relationships that exist between overall fiber quality, specific fiber properties, and yarn quality. All of these factors interact and are critical to the development of cottons that can compete in a global market. Understanding these interactions will allow breeders to more effectively use fiber data, from HVI or AFIS, for selection purposes to improve yarn quality.

Inheritance of long staple fiber quality traits of Gossypium barbadense in G. hirsutum background using CSILs

Gossypium hirsutum is a high yield cotton species that exhibits only moderate performance in fiber qualities. A promising but challenging approach to improving its phenotypes is interspecific introgression, the transfer of valuable traits or genes from the germplasm of another species such as G. barbadense, an important cultivated extra long staple cotton species. One set of chromosome segment introgression lines (CSILs) was developed, where TM-1, the genetic standard in G. hirsutum, was used as the recipient parent and the long staple cotton G. barbadense Hai7124 was used as the donor parent by molecular marker-assisted selection (MAS) in BC(5)S(1–4) and BC(4)S(1–3) generations. After four rounds of MAS, the CSIL population was comprised of 174 lines containing 298 introgressed segments, of which 86 (49.4%) lines had single introgressed segments. The total introgressed segment length covered 2,948.7 cM with an average length of 16.7 cM and represented 83.3% of tetraploid cotton genome. The CSILs were highly varied in major fiber qualities. By integrated analysis of data collected in four environments, a total of 43 additive quantitative trait loci (QTL) and six epistatic QTL associated with fiber qualities were detected by QTL IciMapping 3.0 and multi-QTL joint analysis. Six stable QTL were detected in various environments. The CSILs developed and the analyses presented here will enhance the understanding of the genetics of fiber qualities in long staple G. barbadense and facilitate further molecular breeding to improve fiber quality in Upland cotton.

Genetic effects of introgression genomic components from Sea Island cotton (Gossypium barbadense L.) on fiber related traits in upland cotton (G. hirsutum L.)

The germplasm with exotic genomic components especially from Sea Island cotton (Gossypium barbadense L. Gb) is the dominant genetic resources to enhance fiber quality of upland cotton (G. hirsutum L., Gh). Due to low efficiency of phenotypic evaluation and selection on fiber quality, genetic dissection of favorable alleles using molecular markers is essential. Genetic dissection on putative Gb introgressions related to fiber traits were conducted by SSR markers with mapping populations derived from a cross between Luyuan343 (LY343), a superior fiber quality introgression line (IL) with genomic components from Gb, and an elite Upland cotton cv. Lumianyan#22 (LMY22). Among 82 polymorphic loci screened out from 4050 SSRs, 42 were identified as putative introgression alleles. A total of 29 fiber-related QTLs (23 for fiber quality and six for lint percentage) were detected and most of which clustered on the putative Gb introgression chromosomal segments of Chr.2, Chr.16, Chr.23 and Chr.25. As expected, a majority of favorable alleles of fiber quality QTLs (12/17, not considering the QTLs for fiber fineness) came from the IL parent and most of which (11/12) were conferred by the introgression genomic components while three of the six (3/6) favorable alleles for lint percentage came from the Gh parent. Validation of these QTLs using an F8 breeding population from the same cross made previously indicated that 13 out of 29 QTLs showed considerable stability. The results suggest that fiber quality improvement using the introgression components could be facilitated by marker-assisted selection in cotton breeding program.

Germplasm Potential for Continuing Improvement of Fiber Quality in Upland Cotton: Combining Ability for Lint Yield and Fiber Quality

Exotic germplasm, that is, germplasm without commercial applicability, can broaden the genetic base in upland cotton (Gossypium hirsutum L.). The objectives of this study were to determine combining ability and identify parents from exotic germplasm for breeding. Twelve regionally adapted exotic germplasm lines were crossed with one cultivar and three elite germplasm lines in a North Carolina Design II scheme. The 48 F2 hybrids were evaluated at two locations with 4 and 3 replicates each in 2008 and 2009, respectively. General combining ability (GCA) effects were significant (p ≤ 0.001) for lint yield and all fiber properties. General combining ability effects were more important than specific combining ability (SCA) effects for most traits. Significant (p ≤ 0.05) favorable GCA effects were detected in all exotic germplasm parents for different fiber properties. Positive GCA effects for lint yield and favorable GCA effects for at least one fiber property were detected in some exotic parents: SP156 and SP224 for lint yield and strength; SP192, SP205, and JC65 for lint yield, micronaire, and elongation; and SP192 and SP224 for lint yield and short fiber content and fineness, respectively. Significant high parent heterosis (HPH) for lint yield was detected in F2 hybrids of FM832 × SP205 (14%) and MD15 × SP205 (26%). These results provide evidence for the potential of these germplasm lines in breeding for continuing improvement of lint yield and fiber quality.

Construction of IF<SUB>2</SUB> Population and Its Application in Studies on Genetic Ef-fects and Heterosis for Fiber Quality in Upland Cotton (<I>G. hirsutum</I> L.)

Construction of IF<SUB>2</SUB> Population and Its Application in Studies on Genetic Ef-fects and Heterosis for Fiber Quality in Upland Cotton (<I>G. hirsutum</I> L.)

Associations among Lint Yield, Yield Components, and Fiber Properties in an Introgressed Population of Cotton

ABSTRACTBroadening the genetic base is essential for genetic improvement of lint yield and fiber quality in upland cotton (Gossypium hirsutum L.). An exotic germplasm population derived from multiple crosses between G. barbadense L. and G. hirsutum was obtained from USDA‐ARS at Las Cruces, NM. Experiments were designed to evaluate this germplasm for genotypic variation in lint yield and fiber quality. Two hundred germplasm lines and five cultivars were planted at two locations in 2006 and one location in 2007. Significant (P ≤ 0.001) genotypic variation for yield and fiber quality was identified. In path coefficient analysis, bolls per square meter had the largest direct effect on lint yield, while boll weight and lint percentage had secondary direct effects on yield. Lint yield was negatively associated with fiber quality. Lint per seed was favorably correlated with lint percentage, elongation, span lengths, and fineness. Based on stepwise analysis, 50% length, short fiber content, fineness, maturity ratio, and 2.5% length contributed to strength, in decreasing order of significance. Short fiber content was negatively correlated with fineness (r = −0.41). These results indicate that this germplasm is a useful genetic resource for genetic improvement of lint yield, fiber quality, and analysis of interrelationships between lint yield and fiber quality.

Molecular Tagging and Source Analysis of QTL for Elite Fiber Quality in Upland Cotton

With the development and application of molecular biology technology in recent years,several genetic maps developed from different combinations have been constructed and lots of QTL controlling fiber quality or yield components detected,which makes the breeding procedure more effective,immediate and time-saving.In this paper,both F2 and F2:3 populations were constructed using two elite upland cotton (Gossypium hirsutum L.) strains (variety),7235 and Yumian 1 as parents.Linkage analysis with SSR molecular markers revealed that 80 loci were mapped into 19 linkage groups that covered 1 147.8 cM genetic distance.A total of 36 QTLs were detected by composite interval mapping method,including 6 for fiber length,8 for fiber strength,7 for micronaire value,8 for fiber elongation and 7 for fiber uniformity;and explained 3.4%–14.4%,5.0%–42.4%,6.5%–11.7%,5.1%–19.4%,and 6.9%–14.0% of the corresponding phenotypic variations,respectively.Analysis of QTL distribution indicated that D7,D8,and D9 chromosomes were enriched with QTLs controlling fiber strength in both parents,among them those existed in D7 and D8 chromosomes were closely linked and two QTLs controlling fiber strength located respectively in D7 and D9 chromosomes were in the complete same position,which could be related with that both 7235 and Yumian 1 bear the pedigree of elite fiber accession PD4381.The utilization potential of the QTL of super quality in the pyramiding breeding was also discussed.

Yield and fiber quality of Upland cotton as influenced by nitrogen and potassium nutrition

Nutrient stress in Upland cotton ( Gossypium hirsutum L.) depresses lint yield, particularly of late-season fruit (bolls), and may disrupt fiber development. A 2-year (1999 and 2000) study was conducted outdoors in large pots to determine individual effects of nitrogen (N) and potassium (K) stress at flowering stage on lint yield and fiber quality. Treatments were half-strength nutrient solution from emergence to crop maturity (control), 20% and 0% of control N from first flower onward, and 20% and 0% of control K from first flower onward in 1999 and first square onward in 2000. Leaf N and K were determined every 2–3 days from an uppermost, fully expanded leaf on the main-stem of five plants selected at random. Mature bolls were harvested from sympodial (fruiting) branches only and grouped according to week of anthesis across a 35-day flowering period, providing five flowering groups, from which fiber length, strength, and micronaire were determined. Fiber length was not consistently altered by stress, suggesting early stages of fiber development were indirectly affected by plant N and K status. Nitrogen deficiency decreased yield through early termination of reproductive growth. In 1999, although flowering group four of N-deficient cotton had low length, strength, and micronaire, values for weighted-sum micronaire (whole-plant micronaire) increased under N stress by about 12% in 0% N treatment and about 18% in 20% N treatment. In general, N and K stress had opposite effects on weighted-sum micronaire. The year by N treatment interaction was significant for weighted-sum strength, due to weak fibers in N-deficient cotton in 1999, but no treatment difference in 2000. Apparently, crop response to N stress was influenced by environment, as flowering groups with low quality fiber also comprised a large fraction of total lint, and thus placed heavy demands on plant N and carbohydrate reserves. Severe K deficiency in 2000 decreased yield and lint weight boll −1, and micronaire values of 3.7 or less were evident in flowering groups two, three and four. Results support evidence that N stress indirectly affects cotton growth, as N deficiency decreased fiber length, strength and micronaire primarily in flowering groups with large percentage of bolls. Results from 2000 support evidence that K deficiency adversely affects reproductive growth, boll weight, and sugar translocation in cotton.

Inheritance of fiber quality and lint yield in a chemically mutated population of cotton

The narrow germplasm base of the upland cotton (Gossypium hirsutum L.), grown on the Texas high plains historically, has limited improvement of fiber quality. Chemical mutagenesis and subsequent selection have helped the development of lines with improved fiber quality in cultivars adapted to this region. This study was conducted to determine the inheritance of improvements in fiber quality. M3 lines with divergent fiber properties of micronaire, length, and strength were selected from a population of Paymaster HS 200 treated with 3% v/v ethyl methanesulfonate (EMS) for two hours. The 115 selected lines of M4 and M5 generation were evaluated for fiber quality and lint yield. Regression of the M4 and M5 on the M3 generation, as well as the M5 on the M4 was used to generate narrow sense heritability coefficients. Significant variations were observed between the mutant lines in all generations except for lint yield in the M5 (1997). The highest heritability estimates were found in fiber length (h 2= 0.29** to 0.46**). Micronaire and strength showed intermediate heritability estimates of h 2= 0.14 to 0.19, while lint yield had a very low heritability estimate of h 2= 0.03. Fiber length and strength were correlated (r= 0.58** to 0.46**) in all the three generations. The mutants identified in these studies have the potential to improve fiber quality of upland cotton without introducing alien genes that may reduce adaptation to short growing season production regions.

Association of Fiber Quality Parameters and Within‐Boll Yield Components in Upland Cotton

Lint yield and fiber quality in upland cotton, Gossypium hirsutum L., are interrelated through a series of individual components such as fiber length and the number of fibers produced on each seed. Numerous studies have reported the relationships among various components of yield and yield per unit land area, but none have reported on the relationships among the most basic within‐boll yield components and fiber quality parameters. The objectives of this study were (i) to determine the association of fiber quality parameters with basic within‐boll yield components for six diverse cotton genotypes and (ii) to determine if repulsion phase linkage explained the high negative correlations reported among fiber quality parameters and lint yield. The second objective was investigated by comparing the associations of fiber length, strength, and micronaire with within‐boll yield components among selected F1 populations developed by mating parents with similar direction of general combining ability (GCA) for fiber quality and within‐boll lint yield components with selected F1 populations derived by mating parents with dissimilar direction of GCA. Parents and F1s were grown at College Station, TX, in 1989 and 1992. Fiber quality parameters were determined by high volume instrumentation. Within‐boll lint yield components were determined by direct measurement or through calculations. Fiber strength and length were negatively associated with the most basic within‐boll lint yield components. Repulsion phase linkage appears to play a role in the negative association of fiber quality and within‐boll lint yield but pleiotropic effects could not be ruled out.

Combining Ability for Within‐Boll Yield Components in Cotton, Gossypium hirsutum L

Lint yield and fiber quality in upland cotton, Gossypium hirsutum L., are interrelated through a series of individual components that include bolls/unit land area, mean fiber length and weight/unit length, and a series of within‐boll components. Numerous studies have investigated the relationships among various components of lint yield and yield per unit land area, but none have reported on the relationships among the most basic within‐boll yield components and fiber quality. The objective of this study was to evaluate the combining ability for within‐boll lint yield components among a group of cotton genotypes that varied by date of release, type of release, originating program, and their fiber quality parameters, especially fiber bundle strength. Four cultivars, one each released in 1905, 1943, 1979, and 1981, and two modern germplasm lines were crossed in a half diallel. Parents and F1s were grown at College Station, TX, in 1989 and 1992. Twenty‐five and 100 normal bolls were hand harvested from each plot of a randomized complete block design in 1989 and 1992, respectively. Fiber quality parameters were determined by high volume instrumentation. Within‐boll yield components were determined by direct measurement or through calculations. Genotypes having good general combining ability (GCA) estimates for fiber quality exhibited negative GCA estimates for the most basic within‐boll yield components. Among these genotypes, three‐way crosses, modified backcross, or recurrent selection procedures would be required to select for improved fiber quality and simultaneously increase the number of harvestable fibers per unit of seed surface area.

Breeding Methods for Improving Yield and Fiber Quality of Upland Cotton (Gossypium hirsutum L.)1

In the introgression of Gossypium arboreum L. ✕ G. thurberi Tod. fiber strength to medium staple Upland cotton (G. hirsutum L.), we have been most successful with the pedigree method of hybridization and selection. Hybridization is required for the recombination of genes for high lint yield and fiber strength or length, and selection is necessary to locate the desirable recombinants in segregating populations.In 25 years of selection within breeding lines after the F5 generation, only two highly desirable recombinants, Earlistaple‐7 and Pee Dee 4381‐54, have been found. Yield improvement in Earlistaple‐7 is attributed to fusarium wilt resistance rather than to recombination of genes for lint yield and fiber quality. We cannot rule out the possibility of intercrossing or outcrossing in the isolated seed increase of Pee Dee 4381 from which Selection‐54 was obtained.Random intermating, backcrossing, and composite‐crossing also were tried as methods to improve yield and fiber quality of Upland cotton. A valuable breeding line (Q) was developed by three cycles of backcrossing to three G. hirsutum parents (with selection only for high lint percentage) and a composite‐cross of progenies from each of these backcross populations. No desirable recombinants were found with random intermating. Failure of this method was attributed to: 1) broad germplasm pools, 2) nonelimination of undesirable plants from the populations, and 3) study of small populations of 10,000 plants.We believe that any breeding method that gives maximum combinations of genes for yield and fiber quality is desirable, and selection of promising recombinants in early generations is necessary to maintain populations of practical size.