Chromosome substitution lines with improved essential mineral nutrients and fiber quality traits in Upland cotton

Introgression of donor germplasm from Gossypium tetraploid species into Upland cotton (G. hirsutum L.) would help alleviate the constricted genetic base of Upland cotton and increase opportunities for genetic improvement. The objective of this research was to determine whether the mineral nutrition and fiber quality of cotton are affected by and might be improved using 11 quasi-isogenic BC5Sn interspecific chromosome substitution (CS) lines of Upland cotton. The CS lines were bred previously for disomic replacement of Upland cotton chromosomes 01, 04, 07, 15 (part), and 18 by homologs from G. barbadense L. and G. tomentosum Nutt. ex Seem., and another for part of Upland chromosome 08 by G. tomentosum. The CS lines were grown in two fields along with a related Upland inbred, TM-1. Leaf mineral nutrition of the CS lines was measured based on samples taken two weeks after flowering. Fiber quality was based on bolls hand-picked at harvest. CS lines involving 1 and 18 were associated with leaf nutrient improvement. CS-B18 improved leaf N, P, K, and Zn relative to TM-1. All fiber quality traits were affected by CS lines, but none led to more than two fiber quality improvements. CS-B15Lo and CS-B18 improved fiber strength and elongation, whereas CS-T15Lo improved fiber elongation and yellow index. Some leaf mineral and fiber quality traits exhibited positive correlations. Overall results suggest these chromosome substitutions broaden the genetic base and allow for potential enhancement of leaf mineral nutrition and fiber quality of Upland cotton.


Introduction
Use of chemical fertilizers in crop production in the past few decades played a significant role in maintaining soil fertility and in nearly doubling food production in developed and developing countries (Loneragan 1997).Improving essential element concentrations in crop species is critical for human and animal health and the improvement of crop species.Plants require N (nitrogen), P (phosphorus), K (potassium) in large quantities; Ca (calcium), Mg (magnesium), and S (Sulphur) in moderate quantities; and Fe (iron), Cu (copper), B (boron), Mn (manganese), Mo (molybdenum), and Zn (zinc) in very small quantities for growth and development (Alloway 1990;Brady and Weil 2002;Imtiaz et al. 2010).
Plants absorb mineral elements from the soil to meet their need for growth and development when available in sufficient amounts.When deficient, the elements have to be supplied externally as chemical fertilizers.But plants have developed different capacities of utilizing mineral nutrients from the soil when not present in sufficient amounts.This capacity is primarily controlled by the genetics of the plant and environmental factors.In a review article, Maillard et al. (2015) reported that plants have different strategies for acquisition and use of most macro-and some micro-nutrients when facing nutrient deficiencies.The first strategy is to increase the expression of genes encoding more or less nutrient-specific root transporters (Amtmann and Armengaud 2009;Gojon et al. 2009).The second, longer-term, strategy is to increase root growth and branching, which increase exploration of the soil (Gruber et al. 2013;Giehl and von Wirén 2014;Giehl et al. 2014).Maillard et al. (2015) observed that among several species, wheat (Triticum aestivum L.), barley (Hordeum vulgare L.), and oak (Quercus robur L.) were the most effective at remobilization of mineral elements from leaves for efficient use while poplar (Populus nigra L.) and maize (Zea mays L.) were the least efficient.It has been reported that remobilization of Zn from leaves to the grain is substantial in wheat and barley during post-anthesis (Maillard et al. 2015).They also mentioned that more Fe (80%) than Zn (50%) in wheat grains was derived from leaf remobilization.Normally a genotype can be considered nutrient-efficient when it provides an above-average yield at suboptimal soil nutrient levels (Graham 1984).However, a genotype can also be considered as nutrient-efficient if it converts high nutrient supply into yield comparatively better than other genotypes (Sattelmacher et al. 1994).For example, Moll et al. (1982) described nitrogen use efficiency as maize grain yield per unit N supply and Muchow (1998) reported that the efficiency of utilization of N could be defined as grain yield per unit N uptake in a study of maize and sorghum (Sorghum bicolor L.).
Cotton, a perennial shrub cultivated as a domesticated annual crop, is the most important natural textile fiber source as well as a major oilseed crop for human use and as a livestock feed.The worldwide economic impact of the cotton industry is estimated to be about $500 billion/year with an annual use of 115 million bales or 27 million metric tons of cotton fiber (National Cotton Council 2006, http:// natio nalco ttonc ouncil. com 2007).
Global market demands that cotton breeders need to shift selection parameters not only from yield to improving fiber qualities but also to making plants more resilient against emerging threats from new pests, diseases, and environmental changes.To meet these new challenges, breeders will have to use strategies with some innovative methods of genetic solutions to develop cotton lines resilient against biotic and abiotic stresses with improved yield and fiber quality.The solutions require adequate genetic variation in the cotton gene pool.However, Upland cotton cultivar development has relied heavily on crosses among cultivars and similarly elite lines with high yield and moderate fiber quality traits, and resulted in a discernibly narrow genetic base among cultivars (Van Esbroeck and Bowman 1998;Hinze et al. 2017).
Interspecific germplasm introgression could be a useful strategy for broadening the genetic base of Upland cotton because its genome is genetically buffered by polyploidy, and it shares the same chromosome number (2n = 52) and same gross cytogenetic architecture with the other tetraploid species of the primary gene pool, including G. barbadense (L.) and G. tomentosum (Nutt.ex Seem.).Nonetheless, previous attempts to use conventional breeding methods for interspecific breeding have been seriously impeded by natural barriers to transmission and recombination; without sufficient genetic resolution around them, potentially useful donor loci are co-inherited with undesirable factors that render them commercially unsuitable in terms of agronomic Vol.: (0123456789) performance, fiber yield and/or quality.Given the need for speed, scale and near perfection in varietal breeding, most cotton breeders are understandably averse to dedicating major long-term efforts to widecross breeding.We have demonstrated over the past several years that interspecific chromosome substitution (CS) lines can be a useful breeding and genetic tool for targeted introgression of donor species germplasm into Upland cotton (Jenkins et al. 2017a, b;Saha et al. 2013Saha et al. , 2017)).CS lines are quasi-isogenic to each other, so replicated experimentation enables facile statistical comparisons that identify CS lines with potentially valuable donor-derived traits and indicates their likely chromosomal locations.
Limited genetic information is available on the genetic control of mineral element status and the regulation and transport of the elements for sustaining the physiological and biochemical functions in crop plants including cotton (Baxter 2010;Salt et al. 2008;Singh et al. 2013).In this research, the genetic potential for nutrient uptake efficiency has been considered by measuring the variation in leaf nutrient content of different CS lines.Genetically improving cotton for its ability to uptake and efficiently utilize mineral nutrients could reduce the amount of applied fertilizers and thus reduce cost of production.The primary goal of this research was to identify CS lines with improved essential nutrient contents measured in the leaves of different CS lines grown under the same field environment.The specific objectives of this research were to: (1) identify the level of essential mineral elements N, P, K, Ca, Mg, Cu, Fe, Mn, Zn, B and fiber quality traits in selected CS lines and (2) determine the relationships of individual mineral element contents with fiber quality traits.

Field experiment
Our previous studies demonstrated that CS lines provided a tool for effective introgression of germplasm and detection of beneficial alleles from the alien tetraploid species into Upland cotton (Saha et al. 2017(Saha et al. , 2020;;Jenkins et al. 2017a, b;Ulloa et al. 2016).Each of the CS lines used in this study was bred previously by [1] first developing an isogenic hypoaneuploid derivative of the Upland cotton genetic standard TM-1, e.g., monosomic for Chromosome-01, or monotelodisomic for part of Chromosome 15; [2] hybridizing the hypoaneuploid TM-1 as seed parent with a donor species as pollen parent; [3] recovering the corresponding hypoaneuploid (BC0F1) hybrid progeny (hemizygous for donor chromosome, e.g., Chromosome-1, but heterozygous for all other chromosomes); [4] backcrossing the hypoaneuploid (BC0F1) hybrid progeny as pollen parent onto the previously used TM-1 hypoaneuploid female parent.Then [5] repeat Steps 3 and 4 iteratively, advancing the backcross generations while hemizygous, until attaining the BC5F1 hypoaneuploid; [6] selfpollinating and [7] recovering BC5F1S1 euploid (disomic substitution line, or "chromosome substitution").Thus, CS lines are expectedly quasi-isogenic with each other and Upland inbred TM-1, where each line differs by the disomic replacement of a specific chromosome pair of chromosomes or chromosome segments from the alien donor species (Fig. 1; Stelly et al. 2005;Saha et al. 2013Saha et al. , 2017;;Ulloa et al. 2016;Jenkins et al. 2017a, b).More detailed illustrations of the development of CS lines can be found in Saha et al. (2012).Twelve euploid (2n = 52) cotton lines were used in this experiment, including 11 CS lines and Upland inbred TM-1, the recurrent parent used in CS line development.The CS lines were generated by substituting Chromosomes 1, 4, 7, 15 long arm, and 18 of Upland cotton with corresponding chromosomes from G. barbadense L. (CS-B) and G. tomentosum Nuttall ex Seeman (CS-T).Also, one additional CS line, substitution of the long arm Chromosome eight (8Lo) from G. tomentosum (CS-T), was included.The 11 CS lines were designated as CS-B01, CS-B04, CS-B07, CS-B15Lo, CS-B18, CS-T01, CS-T04, CS-T07, CS-T08Lo, CS-T15Lo, and CS-T18.
The CS lines and TM1 were grown in two field locations at the R. R. Foil Plant Science Research Center of Mississippi State University, MS (33.4°N, 88.8° W) in 2014.The field design was a randomized complete block with four replications.The soil type at one of the fields was a Marietta loam (fine-loamy, siliceous, active, thermic Fluvaquentic Eutrudepts) and the soil type at the other field was a Leeper silty clay loam (fine, smectitic, nonacid, thermic Vertic Epiaquepts).Individual lines were grown in 12-m long single-row plots with row-to-row spacing of Vol:.(1234567890) 97 cm.The plant-to-plant spacing within each row was 10 cm (a total of 120 plants per plot).

Fiber trait analysis
A 25-boll sample was hand-harvested from first-position bolls near the middle nodes of plants in each plot prior to machine harvest.Boll samples were weighed and ginned on a 10-saw laboratory gin.Fiber samples were analyzed by Cotton Incorporated, Cary, NC using a high-volume instrument (HVI Uster Technologies).The following fiber properties were measured: strength, force required to break a bundle of fibers one tex unit in size; upper half mean fiber length (UHM), the average length of the longest one-half of the fibers in a sample; elongation, percent elongation of fibers before breaking; micronaire (Mic), a measure of fiber fineness or maturity by an airflow instrument that measures air permeability of a constant mass of cotton fibers compressed to a fixed volume; fiber uniformity ratio (UI), the ratio of the mean fiber length to the upper half mean fiber length determined from a fibrograph beard and expressed as a percentage; Rd, a measure of light reflectance with higher values denoting more white fibers; +b, a measure of yellowness with greater +b values representing greater yellow fiber color.

Leaf mineral nutrient analyses
The elements N, P, K, Ca, Mg, Cu, Fe, Mn, Zn, and B were measured in leaf samples collected 2 to 3 weeks after the first flower (4 Aug. 2014 and 6 Aug. 2014).In each plot, the youngest and fully expanded leaf, usually on the third or fourth node from the top, were collected from 20 plants.The samples were dried at 75 °C in a forced-air oven to constant weight, and ground to pass a 1.0-mm screen before analysis.
Prepared leaf samples were analyzed for total N concentration by an automated dry combustion method using a ThermoQuest C/N analyzer (CE Elantec, Inc., Lakewood, NJ).Concentrations of P, K, Ca, Mg, Cu, Fe, Mn, and Zn in the samples were determined using an inductively coupled dual axial argon plasma spectrophotometer (ICP, Thermo Jarrell-Ash Model 1000, Franklin, MA) following ashing and acid-digesting ≈ 0.2 g of the dried and ground samples.Briefly, the samples were ashed in a muffle furnace at 500 °C for 4 h, the ash was then digested by adding 1.0 mL of 6 M HCl for 1 h and 40 mL of a double-acid solution of 0.0125 M H 2 SO 4 and 0.05 M HCl for an additional 1 h.The digested solution was filtered using a 2 V Whatman (Maidstone, UK) filter paper before analysis with the ICP.

Statistical analysis
The data were analyzed as a randomized complete block design using the MIXED model analysis of SAS (Littell et al. 2002).The data from both fields were analyzed with field and genotype treatment effects as the fixed-effect factors and replication and its interaction with field and genotype as the randomeffect factor.In addition to comparison of the genotype means using LSD, group comparisons between genotypes with G. barbadense vs. G.tomentosum chromosome substitution were made for robust data interpretation.Pearson correlation coefficients were computed to test the relationships among the different mineral element concentrations and fiber traits.All differences mentioned in the discussion are significant at P ≤ 0.05 unless stated otherwise.
Given that CS lines and TM-1 are regarded as quasi-isogenic for all chromosome pairs except the substituted chromosome or chromosome segment pair, significant differences for any trait were inferred as due to either genes on the specific substituted chromosome or chromosome segment and/or epistasis between TM-1 genes and genes on the substituted chromosome or chromosome segment pairs of the alien species (Saha et al. 2012).We note, however, that in addition to the targeted substitution, each BC5S1-derived CS line is also expected to cryptically harbor one or more non-targeted donor segments due to serendipitous retention during backcrossing and inbreeding, so the chromosomal associations inferred for specific donor traits are regarded as tentative, and warrant follow-up breeding and SNP-aided segregation analysis.

Results
ANOVA indicated that genotypes very significantly affected 7 of the 10 essential mineral elements, including N, Ca, Mg, P, K, Zn, and B, but not Cu, Fe or Mn (Table 1).All fiber traits were also significantly affected by genotype.Overall results indicated that the 11 CS lines caused changes in cotton essential mineral nutrition and fiber quality.If all CS lines are highly isogenic, except for the substituted chromosomes, significant differences among the lines for all traits are likely due the substituted chromosomes and segments.Many such differences are likely due to the targeted substitution (Fig. 1), but a guarded interpretation is recommended, since cryptic substitutions may also exist (see "Discussion" section).

Mineral elements
Pairwise comparisons between CS lines targeting the same Upland chromosome or chromosome arm with substitution from G. barbadense and G. tomentosum, e.g., CS-B01 versus CS-T01, showed that significant differences exist between the groups for leaf Ca, P, Mn, B (data not shown).For example, CS-B01 versus CS-T01 differed significantly in leaf calcium; relative to TM-1, CS-B01 elevated the leaf Ca level by 10.0% whereas CS-T01 decreased the leaf Ca level by 14.4%.Such differences may be due to genetic differences between the respective G. barbadense and G. tomentosum homologs, but additional data are needed.Among all substitutions, CS-B01 and CS-B18 had the highest leaf nutrient levels for several of the elements (Fig. 2A).Considering all 11 CS lines, 7 of the 10 analyzed mineral elements were increased by one or more CS lines over TM-1.These increases included P 10.5% by CS-T18, K 33.3% by CS-B18, Ca 10.0% by CS-B01, Mg 13.0% by CS-B01, Mn 37.4% by CS-B18, Zn 20.7% by CS-B18, and B 19.0% by CS-B01 (Table 2).The results suggest that Upland cotton leaf mineral nutrition might be improved using these or other CS lines from the two respective alien species.None of the CS lines affected Fe and Cu, compared with TM-1, suggesting the genes for these elements might not be associated with any of the substituted chromosomes used in this investigation.CS-B01 significantly affected leaf N, Ca, Mg, Zn, and B suggesting the potential association of multiple elements with substituting Chromosome 1 from G. barbadense.CS-B01 had the highest Ca, Mg, and B values among all lines (Fig. 2A).The CS lines involving substitution of Chromosome 1 (A genome) or its segmental homeolog 15Lo (D genome) from both G. barbadense and G. tomentosum significantly affected leaf Ca levels compared to TM-1.
CS-B18 had the highest levels of several elements including K, Mn, and Zn.CS-T18 had the highest level among all lines for P, one of the major essential mineral elements for growth and development in cotton.
None of the substitutions from either species significantly increased leaf N, the most commonly applied and important plant nutrient in cotton production.But there were notable reductions in leaf N due to chromosome substitutions.CS-T15Lo reduced leaf N by 7.7% relative to TM-1 and CS-B04 reduced leaf N by 7.0% (Table 2).These results suggest improving cotton N nutrition may not be accomplished using the chromosome substitutions tested in this study.

Fiber traits
All fiber quality traits were highly significantly affected by genotype, and one or more CS line differed from TM1 for each trait.All CS lines differed from TM-1 in two or more fiber traits (Table 3).Comparison of G. barbadense versus G. tomentosum for the same substituted chromosome or chromosome arm (Chromosome 1, 4, 7, 15Lo, and 18) showed that G. barbadense differed from G. tomentosum for UHM, UI, Mic, and Rd fiber traits indicating the potential association of different alleles on the same chromosome from two different alien species (data not shown).

Uniformity index (UI)
UI ranged from 80.0 to 84.7% (Table 3), with 5 of the 11 CS lines differing significantly from TM-1 (83.5%).Only the UI of CS-B15Lo was higher, while four were lower, namely CS-B18, CS-T07, CS-T15Lo and CS-T18, two of which involve homologous substitutions of Chromosome 18, i.e., CS-B18 and CS-T18.The average UI of the 5 G. barbadense CS-B lines (83.5) was greater than the average UI of G. tomentosum CS-T lines (82.7).The highest UI among all lines occurred in CS-B07 and CS-B15Lo.

Fiber strength
Four CS lines (CS-B04, CS-B15L0, CS-B18, and CS-T07) significantly differed in fiber strength than TM-1, indicating their likely association with fiber strength (Table 3).CS-B15Lo produced the highest fiber strength among all lines suggesting the potential use of this line to improve fiber strength.Strength was also significantly improved by CS-B18 and CS-T07 but reduced by CS-B04.

Fiber elongation percentage
Eight of the 11 CS lines (exceptions CS-T04, CS-T07, and CS-T08Lo) differed significantly from TM-1 for fiber elongation percentage (Table 3).CS-B15Lo and CS-B18 showed the highest fiber elongation percentage among all lines.

Reflectance (Rd)
All CS lines with two exceptions (CS-T08Lo and CS-T18Lo) increased or had the same Rd values as TM-1 (Table 3).CS-T07 had the highest Rd of 74.7, and CS-T15Lo had the lowest with 67.5 among all lines compared with 72.4 for TM-1.

Yellowness (+b)
None of the five G. barbadense CS-B lines significantly affected fiber yellowness (Table 3).In contrast, one of the six G. tomentosum CS-T lines, namely CS-T15Lo, significantly affected yellowness.

Pearson correlations
Pearson correlation coefficients that indicate the strength and direction of pairwise relationships among mineral elements and/or fiber traits (Table 4) were quantified to detect positive relationships indicative of possible opportunities for simultaneous improvement, as well as negative correlation coefficients that would suggest antagonistic relationships.Significant to highly significant correlations were common between certain pairs of leaf mineral levels and between certain pairs of fiber quality traits.
A strong positive correlation (0.51) between leaf N and K content suggests similar potential mechanisms associated with the CS lines that might influence the absorption of these two important elements.Interestingly, some leaf mineral levels and fiber quality traits were significantly correlated too, e.g., leaf Mg levels and fiber Uniformity Index (0.31).Leaf Ca content had a highly significant negative correlation coefficient (− 0.46) with micronaire.Leaf K levels exhibited a remarkably large number of strong correlations with other mineral traits and fiber quality traits.Among fiber traits, UHM was positively correlated with UI but negatively correlated with micronaire (− 0.46), and fiber strength had a high positive correlation with elongation percentage (0.63).The highest positive correlation coefficient (0.65) was detected between UHM and UI.

Discussion
All extant AD-genome tetraploid Gossypium species, including G. hirsutum, G. barbadense and G. tomentosum are thought to have descended from a common ancestral tetraploid hybrid between two diploid species, one with an A-like genome, i.e., similar  to extant species G. herbaceum (L.) and one with a D-like genome, i.e., similar to G. raimondii (Ulbrich).Cytogenetic analyses have previously revealed that the allotetraploid cotton species have 52 chromosomes (2n), undergo bivalent pairing during meiosis, and have a grossly similar AADD disomic tetraploid genome architecture (Endrizzi et al. 1985).The A-subgenome and D-subgenome chromosomes of the tetraploid species have been designated as chromosomes 1-13 and 14-26, respectively (Brown 1980).
The diploid-like behavior of cotton is attained during meiosis by the preferential pairing of homologous chromosomes.Sequencing revealed the genomes of these three AD-genome species to be closely related (Chen et al. 2020), which seems to suggest gene transfer and utilization for genetic improvement should be feasible if not facile.However, they also discovered significant coverage of the Upland cotton genome by low-diversity haplotypic blocks, which could greatly complicate recombination-based dissection of donor germplasm to map and selectively recover introgressed beneficial donor alleles in affected parts of the genome.
Overall results from this study are congruent with the hypothesis that alleles of value may exist cryptically in the wild species, and not be detected until they are bred into a suitable genetic background or population, including interspecific chromosome addition or substitution lines (O'Mara 1940, Gerstel 1943), chromosome segment substitution lines (Eshed and Zamir 1995), backcross inbred lines (Wehrhahn and Allard 1965), and their derivatives.Novel alleles in the CS lines significantly affected and in some cases improved leaf mineral nutrition and fiber quality traits in Upland cotton.Several of the same CS-T lines with improved fiber qualities confirmed our previous results (Saha et al. 2017;Jenkins et al. 2017a, b).
Genetic variability provides a foundation for selection in a plant breeding program.Several previous studies reported that the level of genetic diversity is very low in G. hirsutum, especially among elite types (Bowman and Gutierrez 2003;Wendel and Cronn 2003;Saha et al. 2006;Udall and Wendel 2006).The rapid change in textile technology to high throughput rotor spinning and subsequently to air-jet spinning demands higher fiber qualities for efficient spinning.Upland cotton fiber with UHM of 26.7 mm, fiber strength of 250 kN m kg −1 , and micronaire between 3.5 and 4.9 is not discounted in the U.S. cotton fiber market, whereas 28.2 mm UHM, 265 kN m kg −1 fiber strength, and 3.8-4.6micronaire are the minimum requirements for world cotton fiber markets (Hequet et al. 2006;Kolbyn et al. 2010).
Conventional plant breeding strategies to improve fiber quality and other economically important traits using interspecific crosses have been hindered by complex antagonistic genetic relationships between the species at the whole-genome level.Attempts to incorporate genes from G. barbadense into Upland cotton have generally not achieved stable introgression of the G. barbadense fiber properties (Stephens 1949;McKenzie 1970).Such attempts using conventional plant breeding methods have often resulted in poor agronomic qualities of the progeny, distorted segregation, sterility, mote formation, and limited recombination due to incompatibility between the genomes (Reinisch et al. 1994).Our previous studies documented that CS lines can be used for targeted introgression of donor germplasm; once developed, panels of CS lines can be used to discover valuable genetic effects (Saha et al. 2006(Saha et al. , 2013(Saha et al. , 2017;;Jenkins et al. 2017a, b).Ideally, CS lines provide an opportunity to utilize specific traits from other species into Upland cotton by the selective introduction of just one individual donor chromosome into each Upland CS line, largely eliminating complications imparted by the genetic and epistatic effects of the other 25 alien chromosomes.In reality, however, serendipitous retention of some unselected donor segments during backcrossing and inbreeding is to be expected, so CS lines can also contain cryptic non-targeted segmental substitutions elsewhere in their genome.These added substitutions enhance the coverage of the donor genome provided by a CS line, which is an advantage, but the additional coverage also demands guarded interpretations about the locations of genetic effects.Thus, it is reasonable to infer from the association of a heritable trait effect with a given CS line that the responsible donor gene(s) is likely to be in the knowingly substituted chromosome or chromosome segment, but the conclusion is not assured because it is also possible that the donor gene(s) is located in an unrelated stowaway donor segment.Cautious interpretation is also warranted when differential effects are detected in pairwise comparisons between G. barbadense versus G. tomentosum substitutions for the same chromosome, e.g., CS-B01 versus CS-T01, since they merely suggest the possibility or likelihood but not the certainty that observed differences result in the respective donor chromosomes; direct or epistatic effects by unrelated but serendipitously substituted segments in either CS line could explain all or part of a real pairwise difference.Confirmatory evidence is desirable.In general, CS-line specific traits can entail genetic effects of the donor genes, as well as their epistatic interactions with the background Upland cotton genotype (Saha et al. 2004(Saha et al. , 2012)).
Results of this study showed that certain CS lines (genotypes) significantly affected leaf N, Ca, Mg, P, K, Zn, and/or B (Table 1), demonstrating that chromosome substitution generated genetic variability for mineral nutrition.Effects on Fe, Cu and Mn were nonsignificant.Leaf mineral element contents are dependent upon multiple factors in a complex manner involving morphological, physiological, genetic background, and growth stage of the plant.Climate, dwindling water availability, variable rain conditions, enduring soil degradation, and the rising cost of fertilizer enhances the need for the development of crop germplasm with the genetic potential to improve nutrient uptake efficiency as plants are developed that can grow well under low input agriculture conditions with resilience against abiotic stresses (Sandhu et al. 2016;Lynch 2019;Schneider and Lynch 2020).Results from our research suggest that genetic variation within the CS lines can be helpful for developing nutrient-efficient cotton germplasm with improved capacity using conventional breeding methods.The results indicate that both G. barbadense and G. tomentosum can serve as donors of useful foreign genetic material for the mineral nutrition of Upland cotton.
Results of fiber traits with some of these specific CS-B and CS-T lines were largely concordant with our previous results (Jenkins et al. 2017a, b;Saha et al. 2017).The average cotton chromosome is thought to contain roughly 3,000 genes (Chen et al. 2020), so it will be interesting to investigate if the multiple genetic effects associated with the substituted chromosome or chromosome segment in a CS line were due to many genes or one or more genes with pleiotropic effects on mineral elements or fiber quality traits.
Higher yields of modern cotton cultivars with improved fiber qualities will require germplasm with genetic potential for more efficient management of soil and fertilizer nutrients.A greater understanding of the genetic mechanism of the relationship between mineral elements, yield and fiber quality is also vitally important.CS lines like CS-B01 or CS-B18 could be instrumental and novel genetic resources to simultaneously improve several traits, including mineral nutrition and fiber quality in cotton.Our previous research reported CS-B18 had high tolerance against root-knot nematode, Meloidogyne incognita Kofoid and White [Chitwood] and Fusarium wilt, Fusarium oxysporom f. sp.vasinfectum Atk.Syn and Hans (races 1 and 4) disease in cotton (Ulloa et al. 2016).This same CS line in our current study showed that it had 33% higher K, 37% higher Mn, 21% higher Zn, 5% higher fiber strength and 9% higher elongation percentage than TM-1 (Tables 2, 3).
Most breeding programs aim to improve crops through hybridization, recombination and selection of superior lines.Genetic improvement of Upland cotton is likely impeded by its narrow genetic base and the localized recombinational inertia imposed by the many variously sized haplotypic blocks.Although the Gossypium genus comprises over 50 species, it has been challenging to discover, introgress and use valuable genes of wild species to create agronomically elite upland cotton cultivars.Alleles of wild species that might be used to enhance upland cotton performance for quantitative multigenic traits are challenging to discover in their natural genetic background, as their effects are often masked by an overwhelming preponderance of agronomically undesirable characteristics.Thus, there may be significant advantages to introgression and genetic dissection of donor germplasm prior to searching for beneficial genetic effects.The two donor allopolyploid species used in this study, G. barbadense and G. tomentosum, are closely related with Upland cotton with a common shared ancestral genome (Wendel and Cronn 2003).Although both species are amenable to hybridization and introgression into G. hirsutum, G. tomentosum, for which there are no cultivated forms, has very rarely been used in Upland cotton improvement programs.The significant effects of G. barbadense and G. tomentosum CS lines on Upland cotton leaf mineral nutrients and fiber quality underscore the value of CS lines as a tool for interspecific introgression, partial dissection and discovery of potentially useful genetic traits from the donor species.

Conclusion
We explored a panel of 11 interspecific chromosome substitution (CS) lines of upland cotton from two related AD-genome tetraploid species to discover effects of donor germplasm on leaf mineral nutrient levels and fiber quality traits.The quasi-isogenic background of the CS lines facilitated detection of potentially useful genetic effects.We found certain CS lines significantly affected specific leaf essential mineral elements N, Ca, Mg, P, K, Zn, and/or B, but not Fe, Cu or Mn.Given that the CS lines have a background genotype largely equal to the crop species, the observed effects of donor germplasm would include direct effects of donor genes and their epistatic interactions with the crop species genome.The findings suggest probable chromosomal locations of the donor factors affecting leaf essential mineral elements, but follow-up analyses are required.Correlations were significant between some leaf mineral nutrient traits and some fiber quality traits.Lastly, the findings suggest that CS-B18 can be used in Upland cotton breeding programs to potentially improve simultaneously several traits associated with essential mineral elements and fiber qualities.
Funding DMS received research support from Cotton Incorporated projects 13-466TX and 18-201.Other authors declare that no funds, grants, or other support were received during the preparation of this manuscript.

Declarations
Competing interests The authors have no relevant financial or non-financial interests to disclose.

Fig. 1
Fig. 1 Diagrammatic representation to illustrate the idealized genome composition of Upland chromosome substitution line CS-B18, relative to genomes of the recurrent parent background genotype (Upland cotton inbred TM-1) and the G. barbadense donor line (doubled haploid line 3-79)

Fig. 2 A
Fig. 2 A Leaf nutrient content (g/kg for N, P, and K; mg/kg for Mn and Zn) and fiber elongation (%) of TM-1 and the chromosome substitution line CS-B18 with error bars showing the standard error of the mean.B Leaf nutrient content (g/kg for Ca and Mg, mg/ kg for Cu and B) and micronaire (Mic) of TM-1 and the chromosome substitution line CS-B01 with error bars showing standard error of the mean

Table 1
Analysis of variance (ANOVA) regarding effects of field and chromosome substitutional lines (CS) on leaf nutrient concentration and fiber properties UHM upper half mean fiber length, UI uniformity index, Mic micronaire, Rd fiber reflectance, +b fiber yellowness

Table 2
Concentration of plant nutrients in the youngest fully expanded leaf (usually the fourth or third leaf from the top-most node) of chromosome substitution cotton lines (CS).Each value is an average of three replications and two fields Values followed by the same lower (among CS lines) or upper (group contrast) case letter within a column are not significantly different at P Vol:. (1234567890)

Table 4
Pearson correlation coefficients between mineral elements and fiber traits UHM upper half mean fiber length, UI