Advertisement

Euphytica

, 215:190 | Cite as

Identification of novel sources of genetic variation for the improvement of cold germination ability in upland cotton (Gossypium hirsutum)

  • Junghyun Shim
  • Ritchel B. Gannaban
  • Benildo G. de los Reyes
  • Rosalyn B. Angeles-ShimEmail author
Open Access
Article

Abstract

Upland cotton (Gossypium hirsutum) is inherently susceptible to low temperature stress especially during the early seedling growth and boll maturation stages. The goal of the study is to identify novel sources of genetic variation that can be used to improve cold tolerance of cotton during seed germination. Genetic diversity analysis of thirty accessions from the core Gossypium Diversity Reference Set (GDRS) and twenty recombinant inbred lines derived from intercrossing cotton mutants with altered fatty acid content profiles established genetic variation in the test germplasm based on simple sequence repeat (SSR) genotyping. The mutants clustered in a single clade, whereas the GDRS accessions were separated into four different clades. Screening for germination ability at 12 °C and 15 °C showed that the fatty acid mutants had a significantly better overall germination compared to the GDRS accessions. Hydropriming improved the germination rate and uniformity of the GDRS accessions at 12 °C and 15 °C but not those of the fatty acid mutants, which recorded a better overall germination at 15 °C even without hydropriming. The tolerance of the FA mutants to cold stress during germination is proposed to be conferred by the higher proportion of unsaturated to saturated fatty acids in the mutants compared to the GDRS accessions. Principal component analysis established phenotypic patterns of variation that is consistent with the observed genotypic variation in the test germplasm. Results of the study indicate the potential of the mutants and select GDRS accessions as donors in breeding for cold germination ability.

Keywords

Gossypium hirsutum Cotton Cold tolerance Germination Genetic diversity Fatty acids 

Introduction

Upland cotton (Gossypium hirsutum) is native to tropical and subtropical environments and grows best under long-season cultivation in warm climates. While maximum cotton production can be achieved when plants at the juvenile and adult vegetative phases are grown at a temperature range of 21–30 °C, optimum germination requires a higher temperature of 28–30 °C (Lehman 1925; Ludwig 1932; Stanway 1960). Decrease in temperature below 15 °C concomitantly reduces germination percentage and induces chilling injuries that often leads to seedling malformations, taproot loss, reduced vigor, increased occurrence of seedling diseases, poor stand establishment and ultimately, reduced yields. Similarly, low temperature stress at the later stages of growth negatively impacts the crop by delaying fiber elongation and reducing cell wall thickening, resulting in low cellulose production and overall decline in crop productivity (Kittock et al. 1986; Speed et al. 1996; Krzyzanowski and Delouche 2011).

The innate susceptibility of cotton to low temperature stress severely limits the planting window for the crop particularly in temperate regions. While late season planting provides an optimum range of temperature for seed germination, it runs the danger of having the crop mature late in the fall when unexpected cold snaps below 15 °C are very likely to occur and negatively impact yield potential, as well as fiber and seed quality (Gipson et al. 1969). Conversely, early season planting to ensure maturation of the crop under warmer temperatures risks poor germination, emergence and stand establishment due to cold stress in the early spring (Christiansen and Thomas 1969; Pettigrew 2002; Buxton et al. 1977).

Chemical priming of seeds before planting has been proven to mitigate the negative impacts of cold stress on seed germination during early season planting. Exogenous applications of abscisic acid and mefluidide before planting early in the season effectively provided cold tolerance to germinating seeds (Rikin et al. 1979, 1984; Li 1994). Once the seedlings emerge, foliar spraying with the plant growth regulating chemicals, ethephon, mefluidide, and/or diethanolamine efficiently reduced seedling damage. Such strategies to mitigate the adverse effects of cold stress have increased the yield of cotton planted in the first week of April in the mid-Southern US by 10% (Pettigrew 2002).

Despite the benefits afforded by chemical mitigation strategies to the overall growth and agronomic performance of cotton planted in the early spring, the additional costs required for these inputs limit the economic productivity of the crop. In the long term, breeding and cultivation of cotton varieties with improved tolerance to low temperature stress during the germination and early seedling stages would be the most economical and efficient approach to establish production stability of early planted cotton. Identification of novel genetic variation that can be exploited to improve cold germination ability is a key step towards the development of cotton varieties with high seedling vigor characterized by high potential to germinate, emerge and establish quickly in the field at suboptimal temperatures.

Natural populations and mutant libraries of cotton hold a tremendous amount of genetic variation that can be tapped to create novel gene combinations that can increase crop productivity under a wide range of agricultural ecosystems. The US National Cotton Germplasm Collection holds more than 10,000 cotton accessions representing 45 Gossypium species (Campbell et al. 2010; Hinze et al. 2015). A subset of this collection makes up the Gossypium Diversity Reference Set (GDRS) which represents 70% of the genetic diversity present in the total cotton germplasm collection. Molecular genetic characterization of 1982 tetraploid accessions of the GDRS using 105 SSR markers indicate the presence of intra-species genetic variation that can be used for trait improvement (Hinze et al. 2015). A significant portion of this variation is also captured in the core GDRS accessions that represent a good spectrum of allelic haplotypes present across the entire diversity panel (Hinze et al. 2017). Based on the results of extensive phenotypic, physiological and biochemical profiling, the core panel also represents a large portion of phenotypic variation for salt and dehydration stress tolerance potential (B.G. de los Reyes, unpublished data). Thus, the core GDRS provides a good baseline for the initial assessment of both genetic and phenotypic variation relevant to seedling cold tolerance potential.

Alongside the natural variation provided by the core GDRS panel, induced genetic variations such as those generated by chemical mutagenesis provide an additional and valuable genetic resource (Auld et al. 1998, 2000; Bechere et al. 2009, 2012; Aslam et al. 2016) that can be utilized to broaden the genetic base of crops which has been severely limited by the bottleneck effects of domestication and artificial selection such as cotton (Shim et al. 2018). In this study, genetic and phenotypic variation for cold germination ability was evaluated across a panel comprised of a subset of the core GDRS and chemically-induced mutant lines generated from the breeding program at Texas Tech University (Auld et al. 1998, 2000; Bechere et al. 2009, 2012). The goal was to establish a baseline for future comparison of genetic and phenotypic variation across a larger germplasm collection towards the identification of potential genetic donors for cold germination improvement in cotton.

Materials and methods

Plant materials

A subset of the core GDRS (Hinze et al. 2015, 2017) composed of 30 G. hirsutum accessions from the National Cotton Germplasm Center, as well as 20 recombinant inbred lines (F5 generation) that were derived from intercrossing the fatty acid mutants AFIS 1-1422-A5 and SCM3-7-3-A3 in the genetic background of Acala 1517-99 and SC 9023, respectively, (Thompson et al. 2019) were used in the study (Table 1). The mutant lines were generated at Texas Tech University from 1997 to 2008 by mutation induction of the Texas High Plains cotton cultivars SC 9023 and Acala 1517-99 using ethylmethanesulfonate (EMS). The EMS-induced mutants in the background of SC 9023 were selected primarily for tolerance to the herbicide, Imazamox, whereas those in the background of Acala 1517-99 were selected for fiber quality up to the M5 generation before they were analyzed for fatty acid composition. The M5 mutants AFIS 1-1422-A5 and SCM3-7-3-A3, which were identified to have lower palmitic acid (18.10–18.70%) and higher linoleic acid content (56.80–58.70%) compared to commercial cultivars, were crossed and advanced by single seed descent up to the F5 generation. The selected F5 lines have 17.30–19.60% palmitic acid, 1.90–3.00% stearic acid, 17.70–24.00% oleic acid and 58.20–61.80% linoleic acid. The inclusion of the fatty acid (FA) mutants in the study was based on reports of better and faster germination at cooler temperatures of seeds with higher proportion of unsaturated versus saturated fatty acids (Thompson et al. 2019; Linder 2000). On the other hand, the selected GDRS accessions have been proposed to represent a broad spectrum of variation for tolerance to abiotic stresses particularly salinity and drought. Both GDRS accessions and FA mutants were grown at the Horticultural Gardens of the Department of Plant and Soil Science in the winter of 2018 for seed multiplication and leaf sampling for DNA extraction.
Table 1

GDRS accessions and fatty acid mutants used in the genetic diversity and cold germination ability analysis

No.

Inventory no.

Species

Accession name

Country of origin

1

SA-0002

G. hirsutum

Algerian Brown

Algeria

2

SA-0033

G. hirsutum

Hopi

USA

3

SA-0165a

G. hirsutum

M.U. 3 UA 7-41

Trinidad and Tobago

4

SA-0298

G. hirsutum

Wonder Wilt Wannamaker’s

USA

5

SA-0300a

G. hirsutum

Rowden #2

USA

6

SA-0369a

G. hirsutum

D and PL 10-1

USA

7

SA-0582

G. hirsutum

 

USA

8

SA-0718

G. hirsutum

Arkansas 11 Nucala X Rowden 20-4

USA

9

SA-0857a

G. hirsutum

Acala Original

USA

10

SA-0881b

G. hirsutum

Missdel

USA

11

SA-1055ab

G. hirsutum

M 100

Uzbekistan

12

SA 1106ab

G. hirsutum

  

13

SA 1156a

G. hirsutum

  

14

SA-1232

G. hirsutum

AC 134 CB 4029

Pakistan

15

SA-1330a

G. hirsutum

Reba P 279 (Reba B-50 X Dpl Smo.)

Chad

16

SA-1406

G. hirsutum

S4727

Russia

17

SA-1412a

G. hirsutum

Chung Mein-Jue #7

China

18

SA-1512

G. hirsutum

Deltapine 50

USA

19

SA-1759

G. hirsutum

Chaco 510 Inta

Argentina

20

SA-1766

G. hirsutum

Ceix

El Salvador

21

SA-2580

G. hirsutum

Acala 1517-99

USA

22

SA-2895

G. hirsutum

Lambright 2020A

USA

23

SA-3284a

G. hirsutum

VIR-6654 SAC-24-4

Russia (Mexico)

24

SA-3403a

G. hirsutum

VIR-7137 Coker 201

Russia (USA)

25

SA-3781

G. hirsutum

Acala Royale

USA

26

TX-0307ac

G. hirsutum

 

Mexico

27

TEX 112ac

G. hirsutum

 

Guatemala

28

TEX 1556ac

G. hirsutum

 

Dominican Republic

29

TEX 1801ac

G. hirsutum

 

France

30

TEX 2150abc

G. hirsutum

 

Trinidad and Tobago

31

FA101-1

G. hirsutum

EMS mutant intercross (F5 generation)

USA

32

FA103-1

G. hirsutum

EMS mutant intercross (F5 generation)

USA

33

FA110-5

G. hirsutum

EMS mutant intercross (F5 generation)

USA

34

FA110-6

G. hirsutum

EMS mutant intercross (F5 generation)

USA

35

FA110-8

G. hirsutum

EMS mutant intercross (F5 generation)

USA

36

FA110-9

G. hirsutum

EMS mutant intercross (F5 generation)

USA

37

FA210-2

G. hirsutum

EMS mutant intercross (F5 generation)

USA

38

FA210-4

G. hirsutum

EMS mutant intercross (F5 generation)

USA

39

FA210-7

G. hirsutum

EMS mutant intercross (F5 generation)

USA

40

FA210-10A

G. hirsutum

EMS mutant intercross (F5 generation)

USA

41

FA301-1

G. hirsutum

EMS mutant intercross (F5 generation)

USA

42

FA301-3

G. hirsutum

EMS mutant intercross (F5 generation)

USA

43

FA302-3

G. hirsutum

EMS mutant intercross (F5 generation)

USA

44

FA303-1

G. hirsutum

EMS mutant intercross (F5 generation)

USA

45

FA303-3

G. hirsutum

EMS mutant intercross (F5 generation)

USA

46

FA304-1

G. hirsutum

EMS mutant intercross (F5 generation)

USA

47

FA304-2

G. hirsutum

EMS mutant intercross (F5 generation)

USA

48

FA306-8

G. hirsutum

EMS mutant intercross (F5 generation)

USA

49

FA307-3

G. hirsutum

EMS mutant intercross (F5 generation)

USA

50

FA309-3

G. hirsutum

EMS mutant intercross (F5 generation)

USA

aGDRS accessions exhibiting 34–75% germination in replicated viability testing

bGDRS accessions that were excluded from genetic diversity analysis due to several missing data

cGDRS accessions identified as landraces. The rest of the GDRS accessions are cultivars

Genetic diversity analysis using SSR markers

Genomic DNA was extracted from young leaf tissues of the experimental materials following the TPS method (Miura et al. 2009). A core panel of 105 SSR markers that are distributed across the 26 chromosomes of upland cotton and are known to discriminate among accessions within and among different Gossypium species (Supplementary Table 1) (Yu et al. 2012; Hinze et al. 2015) were used to genotype the experimental materials following a standard PCR protocol (Shim et al. 2015). The amplified PCR fragments were resolved in 3% agarose gel in 1X TBE and scored based on their molecular weight. SSRs are co-dominant markers and are not able to discriminate heterozygous alleles in polyploids. For this reason, each of the alleles identified by the SSRs were treated as independent loci and were scored as dominant markers. Allelic phenotypes based on the absence or presence of bands were scored as ‘1’ and ‘0’, respectively (López-Vinyallonga et al. 2015). Descriptive statistics for the SSR markers including the number of different alleles (Na), number of effective alleles (Ne) and expected heterozygosity (He) were generated using GenAlEx v 6.5b3 (Peakall and Smouse 2012). Polymorphism information content (PIC) of each individual SSR allele was calculated using the formula PIC = ∑ Pi2, where Pi is the frequency of the ith allele in the genotypes tested (Weir 1990). For dominant markers, this formula was simplified to PIC = 2PiQi, where Pi is the frequency of the present band and Qi is the frequency of the absent band (Tehrani et al. 2008). A genetic distance matrix based on the SSR scores was generated using GenAlEx v 6.5b3 and used to calculate similarity indices based on Jaccard’s coefficient. Genetic divergence among the experimental materials was determined by clustering analysis using the Unweighted Pair Group Method with Arithmetic Mean (UPGMA) subroutine in MEGA with 1000 bootstraps (Kumar et al. 2018).

Viability seed testing

Replicated warm germination tests were conducted to ensure that the observable differences in the germination ability were due to the genetic variation for the trait and not due to poor viability of the seed lot or other post-harvest storage effects. Thirty seeds of each genotype were surface-sterilized, rolled in moist blotting paper, placed inside a sealed container and kept inside a growth chamber set at a constant temperature of 30 °C. Seed germination based on a 2 mm radicle protrusion was recorded after 4, 8 and 12 days in the growth chamber. Genotypes with ≥ 80% germination were considered viable and used for cold germination screening.

Cold germination screening

All experimental materials were screened for germination ability at the minimum cardinal temperature of 15 °C and at the critically low temperature of 12 °C following a modified method by Raphael et al. (2017). Briefly, 30 seeds of each of the GDRS accessions and FA mutants were surface-sterilized with 70% ethanol (v/v) for 1 min and with 20% bleach (v/v) for 10 min before rinsing them three times with sterile, distilled water. Surface-sterilized seeds were placed in petri dishes lined with two sheets of sterilized, qualitative filter paper. A third sheet of qualitative filter paper was laid over the seeds before spraying the plates with approximately 6 ml sterile distilled water. Three replicates for each of the cotton accessions and mutant lines were maintained for 14 days at 12 °C, 15 °C or 30 °C (control).

Additionally, the effects of hydropriming in enhancing cold germination ability in cotton were also determined in triplicate set-ups. A total of 30 seeds of each GDRS accession and FA mutant were surface-sterilized and plated as described above. The seeds were hydrated by spraying approximately 6 ml of sterile distilled water on to the plates before sealing them with parafilm. The plates were then incubated at 30 °C for 8 h to allow imbibition. Instead of drying the seeds after 8 h of hydration, the plates were directly transferred in growth chambers set at 12 °C, 15 °C or 30 °C. All plate set-ups, with or without hydropriming, were transferred at 30 °C after 14 days of low temperature treatment to determine seed recovery.

Germination parameters

The number of germinated seeds in all set-ups was recorded daily for 14 days. Seeds with 2 mm radicle protrusion were considered germinated. Data on germination percentage (GP), mean germination time (MGT), mean daily germination (MDG), peak value (PV) and germination index (GI) were calculated for all experimental lines. GP was derived from the total of number of seeds that germinated/total number of seeds × 100. MDG was calculated as the germination percentage at the end of the germination test/the number of days to the end of the germination test. PV was derived from the cumulative full-seed germination percentages on any day/the number of days to reach these percentages (Djavansher and Pourbeik 1976; Orchard 1977). MGT was calculated using the formula \({\text{MGT}} = \sum {f \cdot x} /\sum f {}\) where f = seeds that germinated on day x. GI was established using the formula GI = (10 × n1) + (9 × n2) + ··· + (1 × n14), where n1, n2…n14 refer to the number of germinated seeds on the first, second, and subsequent days until the 14th day. The multipliers 10, 9 and 1 are weights given to the number of germinated seeds on the first, second and subsequent days, respectively (Scott et al. 1984; Bench et al. 1991).

Statistical analysis

A two-way analysis of variance (ANOVA) and a post hoc Tukey’s test with significance level set at p < 0.05 was carried out to examine the effects of temperature, genotype and interaction of both factors on the cold germination ability of the experimental materials. Patterns of variations in the ability of the test germplasm to germinate at low temperatures, with or without hydropriming, was also established by principal component analysis (PCA) using RStudio v1.1.463 (RStudio Inc 2015). The mean values for GP, MGT, MDG, PV and GI of all the test materials were used as basis for the PCA.

Results and discussion

Descriptive statistics of SSR markers used in genetic diversity assessment

Of the 105 SSR markers used for genotyping, 51 did not amplify or did not give consistently clear bands in 50% of the materials tested and therefore were eliminated from the analysis. Several missing data also excluded the GDRS accessions SA-0881, SA-1106, SA-1055 and TX-2150 from the genetic diversity assessment.

Fifty-four SSR markers consistently amplified targets in the experimental materials, generating a total of 88 alleles. Of the 54 SSRs, 22 were polymorphic. The observed number of alleles per SSR ranged from 1 to 4, with an average of 1.62 alleles per marker. The calculated number of effective alleles (Ne) ranged from 1.02 to 2.00, with a mean Ne of 1.50 ± 0.05 (Table 2).
Table 2

Summary statistics of the SSR markers used to genotype the subset of GDRS accessions and FA mutants

Descriptive statistics

Value

Total no. of markers used

105

Total no. of markers that amplified in > 50% of the experimental materials

51

Average no. of observed alleles per SSR

1.6

Average no. of effective alleles (Ne)

1.5 ± 0.05

Average expected heterozygosity (He)

0.30 ± 0.02

No. of polymorphic SSR markers

22

No. of independent alleles amplified by polymorphic SSRs

88

No. of polymorphic alleles amplified by polymorphic SSRs

42

PIC range of alleles that were considered as independent markers

0.04–0.50

Average PIC of alleles that were considered as independent markers

0.26

SSRs are co-dominant markers and can clearly distinguish between the homozygote and heterozygote forms of alleles when genotyping a diploid species. In genotyping an allopolyploid species such as cotton however, the ability of SSRs to discriminate heterozygous alleles is confounded by the amplification of duplicate loci in the homeologous genome or the presence of multiple alleles due to heterozygosity. To render our data suitable for analysis, each of the alleles generated by the 22 polymorphic SSRs were treated as an individual locus and scored as dominant markers. Following this convention, 42 independent loci were identified to be polymorphic across the experimental materials. Six loci that were generated by the SSR markers BNL2960, BNL1047, BNL1531, JESPR153, CIR218 and BNL1673 recorded a band frequency of ≤ 5%, indicating the presence of rare, informative bands that are unique to an accession or mutant line. The expected heterozygosity (He) of markers used across all accessions ranged from 0.02 to 0.49, with a mean He of 0.30 ± 0.02.

PIC values for each individual locus ranged from 0.04 to 0.50, with an average of 0.26 (Table 3). Twenty-one loci had PIC values of > 0.25, whereas five loci recorded lower PIC values between 0.12 and 0.25. The lowest PIC of 0.04 was shared by the six rare alleles that occurred at ≤ 5% frequency. Despite the low PIC values, these alleles were not excluded from the genetic diversity assessment to ensure that the full extent of allelic variation available in the experimental materials are represented. Overall, 62% of the individual locus recorded high PIC values, indicating the suitability of the markers in assessing the genetic diversity present in the experimental materials.
Table 3

Calculated polymorphism information content (PIC) values for independent SSR alleles that amplified polymorphic targets in the test germplasm

Locus No.

SSR Markers

No. of alleles amplified

Total no. of alleles detected from the test germplasm

Total no. cotton lines genotyped

Frequency of the ‘present’ allele

PIC (2PiQi)

1

BNL3545

1

37

45

0.82

0.292

2

 

2

20

45

0.44

0.494

3

BNL4071

1

4

46

0.09

0.159

4

BNL2960

1

40

46

0.87

0.227

5

 

2

35

46

0.76

0.364

6

 

3

9

46

0.20

0.315

7

 

4

1

46

0.02

0.043

8

BNL0530

1

41

46

0.89

0.194

9

 

2

6

46

0.13

0.227

10

NAU2140

1

30

46

0.65

0.454

11

 

2

16

46

0.35

0.454

12

BNL3474

1

36

46

0.78

0.340

13

 

2

10

46

0.22

0.340

14

MUSB1015

1

43

46

0.93

0.122

15

DPL0541

1

41

46

0.89

0.194

16

 

2

44

46

0.96

0.083

17

BNL1673

1

2

46

0.04

0.083

18

 

2

2

46

0.04

0.083

19

BNL3090

1

44

46

0.96

0.083

20

BNL1495

1

28

46

0.61

0.476

21

 

2

18

46

0.39

0.476

22

BNL4061

2

26

46

0.57

0.491

23

DPL0135

1

10

46

0.22

0.340

24

 

2

36

46

0.78

0.340

25

JESPR220

1

21

46

0.46

0.496

26

BNL1521

1

43

46

0.93

0.122

27

 

2

3

46

0.07

0.122

28

BNL1047

1

45

46

0.98

0.043

29

 

2

1

46

0.02

0.043

30

JESPR065

1

25

46

0.54

0.496

31

 

2

21

46

0.46

0.496

32

 

3

36

46

0.78

0.340

33

BNL1531

1

45

46

0.98

0.043

34

 

2

1

46

0.02

0.043

35

BNL4030

1

26

46

0.57

0.491

36

 

2

3

46

0.07

0.122

37

JESPR119

1

10

46

0.22

0.340

38

CIR218

2

2

46

0.04

0.083

39

JESPR153

1

10

46

0.22

0.340

40

 

2

38

46

0.83

0.287

41

 

3

44

46

0.96

0.083

42

 

4

1

46

0.02

0.043

Aside from establishing the informativeness of the SSR markers in differentiating among genotypes, PIC is also a function of the number of known alleles and their frequency distribution in a set of experimental materials (Botstein et al. 1980; Chesnokov and Artemyeva 2015) and is therefore reflective of the genetic diversity of the genotyped germplasm. The mean PIC value for the SSRs, which coincides with the mean He value, suggests high genetic variation among the genotypes tested.

The markers used in the study belong to a core set of SSRs that has been used to establish genetic diversity in a wide range of cotton germplasm (Yu et al. 2012; Hinze et al. 2015, 2016). Initial validation of the robustness of this core SSR set in establishing polymorphisms between and within species produced a PIC of 0.65/1 when differentiating between species and 0.29/1 when differentiating accessions within species (Yu et al. 2012). The use of the same core set of SSRs to define genetic diversity among 1541 G. hirsutum accessions produced a lower PIC value of 0.07 (Hinze et al. 2015). Discrepancies between the current and the previously reported PICs for the core SSRs are attributed to the differences in the number and composition of the test germplasm, as well as to the threshold used to establish the PIC values. Yu et al. (2012) scored the same set of SSRs as dominant markers with a maximum PIC of 1 to differentiate six G. hirsutum accessions. In the present study, the SSRs were used to genotype 20 FA mutants derived from intercrossing EMS-induced mutants in two different genetic backgrounds and 26 GDRS accessions that represent allelic haplotypes across the diversity panel of cotton. As in the previous report (Yu et al. 2012), the SSRs were scored as dominant markers capable of distinguishing only two alleles at a time. The maximum PIC value however, was pegged at 0.5 based on the formula Pi = Qi = 0.5 (Chesnokov and Artemyeva 2015; Hinze et al. 2015). The different PIC threshold, combined with the genetic nature and number of the GDRS accessions genotyped, contributed to the higher PIC value obtained in the current study. Nonetheless, the results support the suitability of the core SSR set in distinguishing G. hirsutum accessions as previously reported.

Genetic diversity across the core GDRS accessions and EMS mutants

Based on SSR analysis, the calculated genetic distance across the test germplasm ranged from 0.01 to 0.26, with an average of 0.093. UPGMA clustering using Jaccard’s coefficient grouped the test germplasm into five clusters i.e. clusters I, II, III, IV and V based on a minimum similarity threshold of approximately 38% (Fig. 1). The twenty FA mutants, along with the GDRS accessions SA-2580 and SA-1406 formed the biggest clade (cluster V), followed by the sixteen GDRS accessions that grouped into cluster IV with the TX-0307 landrace. The remaining GDRS accessions grouped into minor clades (clusters I, II and II) composed of 2–3 genotypes each.
Fig. 1

UPGMA clustering of GDRS accessions and FA mutants based on Jaccard’s similarity coefficient. Red broken line indicates genetic similarity threshold for the 5 major groupings as indicated by the colored bars labeled I, II, III, IV and V. (Color figure online)

Clustering of the FA mutants was based on a 50–100% genetic similarity. The FA mutants used in the study were recombinant inbred lines derived from crosses between EMS-induced mutants in the genetic background of SC 9023 and Acala 1517-99. Clustering of the FA mutants in a single group reflects the shared pedigree of the lines. Conversely, variation within the clade reflects the intrinsic genetic differences between the cultivars used to generate the mutants, in addition to the variations introduced by natural cross pollination of the FA mutants in the field. Outcrossing rates in upland cotton range from 5 to 50% in the US depending on the climatic conditions imposed by the geographic location, as well as on the cultivar-dependent, physio-anatomical structure of cotton flowers (Turner 1950; Moffet et al. 1980). The FA mutants were advanced to the M5 generation, and the intercrosses up to the F5 generation in the field. Possible natural outcrossing among the mutant lines and even with the different cotton cultivars planted in the field during the same time may have contributed to the variations observed within the FA mutant cluster.

The FA mutants 103-1 and 110-5 grouped together with SA-1406 and SA-2580, respectively at 70–80% similarity, suggesting a shared genetic lineage of the FA mutants with either GDRS accessions. SA-2580 is the cultivar Acala 1517-99 that was selected from progenies of a breeding line having an earlier version of the Acala cultivar in its pedigree (Cantrell et al. 2000). Acala was first cultivated in Texas in the early 1900s and since then has been widely utilized in breeding programs as a donor line for field selections and for the development of new cultivars with higher yield, better fiber quality and adaptability to the Texas High Plains conditions (Turner 1914). SA-2580 was one of the accessions used to generate the mutant populations and hence would naturally group with the FA mutants.

Relative to the FA cluster, accessions in cluster IV grouped together based on a lower genetic similarity ranging from approximately 42 to 85%. Cotton accessions comprising cluster IV originated from at least eight countries including the US, China, Mexico, Russia, Algeria, Chad, El Salvador and Argentina (Hinze et al. 2015). The tetraploid G. hirsutum is entirely Mesoamerican in origin despite having the ancestral genomes of cotton from Asia and Africa (A subgenome), as well as the Americas (D subgenome) (Wendel and Grover 2015). From the New World, G. hirsutum was introduced to over 50 countries across the globe (Wendel and Cronn 2003), including the eight countries where the GDRS accessions from cluster IV were collected. Clustering in a single clade underlines the shared genetic lineage of these GDRS accessions that possibly traces back to the ancestral G. hirsutum from Mesoamerica. Genetic divergence in members of this clade may be attributed to generations of adaptation to artificial selection for region-specific breeding targets, as well as to natural selection pressures that are unique to the environment where each germplasm was collected.

Narrower genetic similarities ranging from approximately 38 to 60% produced three clusters composed of 2–3 genotypes each. Clusters I and II included landraces (TEX 1801, TEX 1556 and TEX 112) from different geographic origins, whereas cluster III was composed of the genetically related Acala Original and Acala Royale from the US that grouped with cultivar AC 134 CB 4029 from Pakistan. Landraces are population complexes that are not necessarily high-yielding but are productive across different seasonal environments of a given locality (Mercer and Perales 2010; Frankel et al. 1998). Despite having adaptations to the edaphic and climatic conditions, as well as to traditional farming systems of a certain eco-geographical area, landraces lack formal crop improvement and are therefore genetically heterogeneous (Casañas et al. 2017). The higher degree of genetic variation in the landraces was reflected in the clustering of these accessions in smaller clades that were far removed from cultivars bred for productivity and quality traits.

Germination ability of the test germplasm under low temperatures stress

Viability testing resulted in a > 80% germination for all FA mutants and thirteen core-GDRS accessions, eliminating possible interference effects of seed lot quality and other post-harvest factors on the germination parameters observed for these lines. Seventeen core-GDRS accessions showed germination rates of 34 to 75% (Table 1). Re-testing the viability of these accessions using a different seed lot resulted in germination rates that were not significantly different from the initial results, suggesting an inherently lower germination ability of the 17 GDRS accessions under the specific conditions used in the experiment.

Germination is a quantitative developmental response of an individual seed that occurs in a specific time point under favorable physiological and environmental conditions. While two different seed genotypes may attain the same final germination percentage, one may germinate faster than the other (Orchard 1977; Scott et al. 1984; Kader et al. 1998). From a crop production standpoint, the ability of seeds to attain uniform germination in the shortest time possible is equally important to the ability of seeds to reach a maximum final germination percentage. For this reason, we used a time-to-event approach in evaluating cold germination ability of the test germplasm and considered the spread (MGT), duration (MDG) and high/low event (PV), alone or in combination with final germination percentage (GI) (Kader 2005).

Consistent with the empirically established effects of low temperature on the germination of cotton seeds (Krzyzanowski and Delouche 2011; Khetran et al. 2015; Cole and Wheeler 1974), the experimental materials recorded a general reduction in the overall germination, expressed in terms of mean GP, MGT, MDG, PV and GI, concomitant to decreasing temperatures from 30 to 12 °C (Table 4). The FA mutants recorded significantly higher mean values for GP, MDG, and GI at all temperature treatments, resulting in a higher, faster and a more uniform germination compared to both GDRS cultivars and landraces. At 15 °C and 30 °C, the GDRS landraces recorded a slightly but not significantly higher germination than the GDRS cultivars (Table 4; Supplementary Table 2). Imbibition at 30 °C prior to cold exposure facilitated a more uniform and faster radicle emergence in both GDRS cultivars and landraces at both 12 °C and 15 °C, although germination of the landraces were more spread throughout the 14-day test period compared to the cultivars. In contrast, hydropriming did not improve the germination of the FA mutants at 12 °C. At 15 °C, the FA mutants already registered high mean values for germination and hence were no longer screened for the effects of hydropriming on germination at this temperature. Despite the benefits of hydropriming in the overall germination of the GDRS at 15 °C, the observed improvements were not at par with the germination ability of non-hydroprimed seeds of the FA mutants germinated at 15 °C (Table 4; Supplementary Table 2). Most of the seeds that did not germinate within 14 days at low temperatures were able to recover and germinate within 5–7 days at 30 °C. The FA mutants recorded the highest recovery at both low temperatures, with or without hydropriming, followed by the GDRS cultivars and the landraces. Significant effects of temperature, genotype and the interaction of both factors on all the germination parameters measured were established using a two-way ANOVA (Table 5).
Table 4

Mean values of the germination parameters used to evaluate cold germination ability in the FA mutants, and GDRS cultivars and landraces

Germination parameter/group

Temperature treatment

12 °C

12 °C+

15 °C

15 °C+

30 °C

Germination percentage (GP; %)

Fatty acid mutants

33.32a

30.18b

94.73a

c

100.00a

GDRS cultivars

13.74b

53.05a

36.20b

92.40a

70.03b

GDRS landraces

11.13b

30.18b

42.87b

74.81b

76.65ab

Mean germination time (MGT)

     

Fatty acid mutants

12.22a

11.33a

5.88a

1.90b

GDRS cultivars

12.42a

10.78a

5.09a

5.41a

5.16a

GDRS landraces

12.14a

11.33a

7.28a

6.82a

4.49a

Mean daily germination (MDG)

Fatty acid mutants

0.07a

0.065a

0.81a

4.91a

GDRS cultivars

0.03b

0.082a

0.10b

0.31a

0.82b

GDRS landraces

0.03b

0.065a

0.14b

0.21a

0.51b

Peak value (PV)

     

Fatty acid mutants

0.49a

0.46a

2.96a

8.14a

GDRS cultivars

0.20b

0.59a

0.68b

1.68a

3.03b

GDRS landraces

0.19b

0.46a

0.73b

1.27a

1.99b

Germination index (GI)

Fatty acid mutants

0.57a

0.48b

4.81a

10.96a

GDRS cultivars

0.21b

0.78a

1.12b

4.39a

4.76b

GDRS landraces

0.21b

1.07a

0.98b

3.15a

4.84b

+With hydropriming treatment

Values within a column followed by different letters indicate significant differences at P < 0.05

CNo data. Effects of hydropriming not tested because of high GP obtained at 15 °C even without hydropriming

Table 5

Two-way ANOVA of the germination parameters used to evaluate cold germination ability of the FA mutants, and GDRS cultivars and landraces

Source

DF

Sum of squares

Mean squares

F

Germination percentage (GP; %)

Genotype

2

6604.71

3302.35

8.937*

Temperature

4

103688.45

25922.11

70.16*

Genotype × temperature

7

47160.231

6737.18

18.23*

Mean germination time (MGT)

Genotype

2

96.98

48.49

72.25*

Temperature

4

2001.48

500.37

745.52*

Genotype × temperature

7

462.61

66.09

98.46*

Mean daily germination (MDG)

Genotype

2

71.56

35.78

98.65*

Temperature

4

185.53

46.38

127.88*

Genotype × temperature

7

135.60

19.37

53.41*

Peak value (PV)

Genotype

2

162.28

81.14

136.29*

Temperature

4

616.57

154.14

258.92*

Genotype × temperature

7

238.60

34.09

57.26*

Germination index (GI)

Genotype

2

197.03

98.51

312.62*

Temperature

4

1376.12

344.03

1091.72*

Genotype × temperature

7

449.69

56.21

178.38*

*Significant at P < 0.001

PCA analysis based on the observed mean GP, MGT, MDG, PV and GI values for each genotype established phenotypic patterns of variation that is consistent with the observed genotypic variation in the test germplasm based on SSR marker profiles. More than 85% of the observed phenotypic variability in response to different temperature treatments was explained by principal component (PC) 1 (63.59–85.42%) and PC 2 (11.36–23.11%) (Fig. 2a–e). The FA mutants, and GDRS cultivars and landraces clustered separately from each other when germinated at 12 °C, 15 °C and 30 °C without hydropriming, although overlaps in the performance of the FA mutants and GDRS cultivars were observed at 12 °C (Fig. 2f, h, j). The more scattered distribution of both GDRS cultivars and landraces in the PCA biplot reflects the wider variability in the germination of the genotypes at these three temperatures. With hydropriming, the variability in the germination of the test germplasm at 12 °C was significantly reduced as indicated by the overlapping distribution of the genotypes in the PCA biplot (Fig. 2g). Similarly, hydropriming narrowed the phenotypic variation between the GDRS cultivars and non-hydroprimed FA mutants germinated at 15 °C (Fig. 2i).
Fig. 2

Patterns of variations in the cold germination ability of the test germplasm established using PCA based on the mean values of GP, MGT, MDG, PV and GI. Histogram representation of the proportion of variance contributed by each principal component (PC), with PC1 and PC2 accounting for most of the phenotypic variation observed in the test germplasm (a–e). Clustering of the FA mutants (red circle), GDRS cultivars (green triangle) and landraces (blue square) at 12 °C without imbibition (f) and with imbibition (g), at 15 °C without imbibition (h) and with imbibition (i) and at 30 °C (j). Clustering of the FA mutants in i is based on the mean values obtained for the all parameters measured from non-hydroprimed seeds. (Color figure online)

Overall, the observed differentiation of the FA mutants from the GDRS cultivars and landraces based on phenotypic measurements for cold germination ability coincides with the clustering of the test germplasm into separate clades based on genetic similarities that was established by SSR marker profiles. Particularly, the observed degree of germination ability of the FA mutants and GDRS accessions in response to low temperature stress correlates with the degree of genetic variability established for the test germplasm.

Potential significance of mutations that affect the germination of FA mutants

The proportion of stored unsaturated to the saturated fatty acid content in oilseeds have been proposed to significantly influence seed germination. Studies on Helianthus species from Texas and Canada showed that accessions with lower proportions of saturated fatty acids had higher and/or faster germination at low temperatures (4 °C/10 °C) compared to accessions with higher proportions of saturated fatty acids (Linder 2000; Meadows 2012). Similarly, cotton seeds that have been genetically engineered to produce higher palmitic acid content recorded a high germination rate at 28 °C that significantly decreased at18 °C (Liu et al. 2017).

The effects of fatty acid composition on the germination ability of oilseeds are based on the established principles of energetics of seed oil synthesis and oxidation. Saturated fatty acids store more energy per carbon and have higher melting points compared to unsaturated fatty acids. For example, the saturated, 16-carbon palmitic acid has a melting point of 62.9 °C, whereas the di-unsaturated, 18-carbon linoleic acid has a melting point of − 5 °C (Linder 2000). At optimum temperatures (28–30 °C), β-oxidation of the single-bonded carbons of saturated fatty acids generates a larger amount of energy than the cis double bonds of unsaturated fatty acids, providing the germinating seeds with maximum chemical energy for growth. At lower temperatures however, the intermolecular bonds that closely pack the polar molecules in saturated fatty acids, reduces the rate at which energy is produced (Lehninger 1993; Linder 2000), resulting in slower seed germination.

Conventional cotton seeds contain approximately 50% linoleic acid, 22–26% palmitic acid, 16–20% oleic and 2–3% stearic acid (Dowd et al. 2010; Dowd 2015; Liu et al. 2009). The FA mutants used in the study have been identified to have lower proportions of palmitic acid (17.30–19.60%) and higher proportions of linoleic acid (58.20–61.80%). Given the significant difference in the melting points of the two fatty acid chains, the reduced proportion of palmitic acid chain alone is expected to effectively lower the cumulative melting temperature of the fatty acids in the mutant seeds, facilitating faster catabolism of the lipid reserves at 12 °C and 15 °C, and providing the seeds the required energy to germinate faster and at a higher rate compared to the GDRS accessions.

Additionally, the higher proportion of unsaturated to saturated fatty acids in the FA mutants may account for the non-responsiveness of FA mutants to the effects of hydropriming when germinated at 12 °C. Germination is a highly regulated process that starts with the imbibition of water by the mature, dry seed. Water influx during imbibition triggers reorganization of the cell membrane from the hexagonal to the lamellar phase. During reorganization, the cell membrane becomes highly permeable causing solutes and low molecular weight metabolites to leak out of the cell. Low temperature during water imbibition exacerbates this cytoplasmic leakage, resulting in cold-induced, structural damages in the seeds, and therefore poor germination (Liu et al. 2017; Noblet et al. 2017; Plazek et al. 2018; Yu et al. 2015).

Membrane unsaturation has been established to reduce the permeability of the cell membrane by increasing its flexibility during reorganization in water-imbibing seeds (Plazek et al. 2018; Upchurch 2008). Recent phospholipidomic studies in corn show that incorporation of unsaturated linoleic acid in the membrane lipids increased cell membrane fluidity, facilitating the faster reorganization of the membranes from a gel to a crystalline liquid state even at a low temperature of 10 °C. The faster membrane remodeling significantly reduces electrolyte leakage during imbibition and rapidly restores cellular function, thereby facilitating faster and higher seed germination (Noblet et al. 2017). Conversely, earlier studies showing a significant decline in the germination of canola seeds engineered to have higher proportions (30–40%) of stearic acid compared to conventional cultivars have proposed the accumulation of saturated stearic acid in the membrane as the cause of poor seed germination at even optimum temperature (Thompson and Li, 1997).

In the current study, the higher proportion of linoleic acid in the FA mutants may have contributed to the unsaturation of the cell membrane, giving it more flexibility during reorganization even at the critically low temperature of 12 °C. The enhanced fluidity of the cell membrane due to membrane unsaturation allowed the FA seeds to rapidly restore normal cellular function, attaining their maximum germination potential at 12 °C without the need for hydropriming. Differences in the maximum germination potential of the FA mutants at 12 °C and 15 °C may be attributed as inherent genetic response to lower temperatures.

While the presumed uniqueness in the fatty acid profiles of these mutants seem to explain why they were able to sustain germination at low temperature in the context of cellular energetics (i.e. FA catabolism), the precise role of membrane properties as conferred by the FA profiles remains purely speculative, and needs to be validated with further studies.

Conclusion

Identification of sources of genetic variation that can be utilized to improve traits of agronomic importance is requisite for the development of an effective breeding strategy for any crop. An important component of this exercise is the purposeful characterization of genetic resources.

In the present study, we sought to evaluate the genotypic and phenotypic variation in the cold germination ability of a minimal set of cotton germplasm to identify sources of variation that can be used to improve germination of cotton under low temperature stress. Combined genetic and phenotypic screening established a wide range of genetic variation in the experimental materials that can provide the foundation for cold germination breeding in upland cotton.

At the minimum cardinal temperature of 15 °C, the FA mutants were identified as robust germinators, recording a higher overall germination of un-imbibed seeds compared to the imbibed seeds of the GDRS accessions. Even at the critically low temperature of 12 °C, the FA mutants outperformed the GDRS cultivars and landraces in terms of all the germination parameters measured, without requiring prior water imbibition treatment.

Generally, the GDRS cultivars were also identified as robust germinators at 15 °C but only when allowed to imbibe water at higher temperatures (28–30 °C) prior to cold germination. At the critically low temperature of 12 °C, the cultivars required prior imbibition at higher temperatures to achieve a minimum of 50% germination.

Lastly, the GDRS landraces were identified as moderate germinators that strictly require prior imbibition for an optimum mean germination of ≥ 74% at the minimum cardinal temperature of 15 °C. Like the GDRS cultivars, the landraces positively and significantly responded to hydropriming even at the critically low temperature of 12 °C. Improvement in the germination ability of the landraces however, was not at par with that of imbibed GDRS cultivars.

Based on controlled environment screening, the FA mutants constitute better donors of cold germination ability in cotton compared to the GDRS cultivars and landraces, although the unique and non-overlapping physiological properties of all the test germplasm make them ideal for potential use in genetic studies, particularly in understanding the molecular mechanisms underlying cold germination ability in cotton as affected by the fatty acid profiles in the seed. Screening for cold germination ability under actual field conditions is currently underway to validate the performance of the test germplasm under controlled conditions and establish their practical use in breeding programs.

Notes

Acknowledgements

This research was supported by Cotton Incorporated as part of the project entitled “Novel sources of seedling cold tolerance and vigor traits in cotton: Identification, characterization and use in marker-assisted breeding”. The authors would like to thank Dr. Dick Auld of Texas Tech University for sharing the seeds of the FA mutants.

Funding

Funding was provided by Cotton Incorporated (Grant No. 18-282).

Supplementary material

10681_2019_2510_MOESM1_ESM.xlsx (16 kb)
Supplementary material 1 (XLSX 15 kb)
10681_2019_2510_MOESM2_ESM.xlsx (21 kb)
Supplementary material 2 (XLSX 20 kb)

References

  1. Aslam U, Cheema HMN, Ahmad S, Khan IA, Malik W, Khan AA (2016) COTIP: cotton tilling platform, a resource for plant improvement and reverse genetic studies. Front Plant Sci 7:1863PubMedPubMedCentralCrossRefGoogle Scholar
  2. Auld DL, Ethridge MD, Dever JK, Dotray PA (1998) Chemical mutagenesis as a tool in cotton improvement. In: Proceedings of Beltwide cotton conference San Diego. National Cotton Council, pp 550–551Google Scholar
  3. Auld DL, Bechere E, Ethridge MD, Becker WD, Hequet E, Cantrell RG (2000) Registration of TTU 202-1107B and TTU 271-2155C mutant germplasm lines of upland cotton with improved fiber quality. Crop Sci 40:1835–1836Google Scholar
  4. Bechere E, Auld DL, Hequet E (2009) Development of naked-tufted seed coat mutants for potential use in cotton production. Euphytica 167:333–339CrossRefGoogle Scholar
  5. Bechere E, Turley RB, Auld DL, Zheng L (2012) A new fuzless seed locus in an upland cotton (Gossypium hirsutum L.) mutant. Am J Plant Sci 3:799–804CrossRefGoogle Scholar
  6. Bench AR, Fenner M, Edwards P (1991) Changes in germinability, ABA content and ABA embryonic sensitivity in developing seeds of Sorghum bicolor induced by water stress during grain filling. New Phytol 118:339–347CrossRefGoogle Scholar
  7. Botstein D, White RL, Skolnick M, Davis RW (1980) Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am J Hum Genet 32(3):314–331PubMedPubMedCentralGoogle Scholar
  8. Buxton DR, Melick PJ, Patterson LL, Pegelow EJ (1977) Relationships among cottonseed vigor and emergence. Agron J 69:677–681CrossRefGoogle Scholar
  9. Campbell BT, Saha S, Percy R, Frelichowski J, Jenkins JN et al (2010) Status of the global germplasm resources. Crop Sci 50:1161–1179CrossRefGoogle Scholar
  10. Cantrell RG, Roberts CL, Waddell C (2000) Registration of `Acala 1517-99’ cotton. Crop Sci 40(4):1200CrossRefGoogle Scholar
  11. Casañas F, Simo J, Casala J, Prohens J (2017) Toward an evolved concept of landrace. Front Plant Sci 8:145PubMedPubMedCentralCrossRefGoogle Scholar
  12. Chesnokov YV, Artemyeva AM (2015) Evaluation of the measure of polymorphism information of genetic diversity. Agric Biol 50(5):571–578Google Scholar
  13. Christiansen MN, Thomas RO (1969) Season-long effects of chilling treatments applied to germinating cottonseed. Crop Sci 9:672–673CrossRefGoogle Scholar
  14. Cole DF, Wheeler JE (1974) Effects of pre-germination treatments on germination and growth of cottonseed at suboptimal temperatures. Crop Sci 14(3):451–454CrossRefGoogle Scholar
  15. Djavansher K, Pourbeik H (1976) Germination value: a new formula. Silvae Genet 25(2):79–83Google Scholar
  16. Dowd MK (2015) Seed. In: Fang DD, Percy RG (eds) Cotton, 2nd edn. Agron Monogr, Wisconsin, pp 745–781Google Scholar
  17. Dowd MK, Boykin DL, Meredith WR, Campbell BT, Bourland FM, Gannaway JR, Glass KM, Zhang J (2010) Fatty acid profiles of cottonseed genotypes from the national cotton variety trials. J Cotton Sci 14:64–73Google Scholar
  18. Frankel OH, Brown AHD, Burdon JJ (1998) The conservation of plant biodiversity, 2nd edn. Cambridge University Press, CambridgeGoogle Scholar
  19. Gipson JR, Ray LL, Flowers CL (1969) Influence of night temperature on seed development of five varieties of cotton. In: Proceedings of the Beltwide cotton production research conferences, pp. 117–118Google Scholar
  20. Hinze LL, Fang DD, Gore MA, Scheffler BE, Yu JZ et al (2015) Molecular characterization of the Gossypium diversity reference Set of the US National Cotton Germplasm Collection. Theor Appl Genet 128:313–327PubMedCrossRefPubMedCentralGoogle Scholar
  21. Hinze LL, Gazave E, Gore MA, Fang DD, Scheffler BE et al (2016) Genetic diversity of the two commercial tetraploid cotton species in the Gossypium diversity reference set. J Hered 107(3):274–286.  https://doi.org/10.1093/jhered/esw004 CrossRefPubMedPubMedCentralGoogle Scholar
  22. Hinze LL, Hulse-Kemp AM, Wilson IW, Zhu QH, Llewellyn DJ et al (2017) Diversity analysis of cotton (Gossypium hirsutum L.) germplasm using CottonSNP63K array. BMC Plant Biol 17:37PubMedPubMedCentralCrossRefGoogle Scholar
  23. Kader MA (2005) A comparison of seed germination calculation formulae and the associated interpretation of resulting data. J Proc R Soc N S W 138:65–75Google Scholar
  24. Kader MA, Omari M, Hattar B (1998) Maximizing germination percentage and speed of four Australian indigenous tree species. Dirasat Agric Sci 25:157–169Google Scholar
  25. Khetran AS, Bashi W, Baloch S, Saleh A, Shah AN et al (2015) Influence of temperature regimes on germination of cotton (Gossypium hirsutum L.) varieties. J Biol Agric Healthc 5(11):173–177Google Scholar
  26. Kittock DL, Selley RA, Cain CM, Taylor BB (1986) Plant population and plant height effects on pima cotton lint yield. Agron J 78:534–538CrossRefGoogle Scholar
  27. Krzyzanowski FC, Delouche JC (2011) Germination of cotton seed in relation to temperature. Rev Bras Sementes 33:543–548CrossRefGoogle Scholar
  28. Kumar S, Stecher G, Li M, Knyaz C, Tamura K (2018) MEGA X: molecular evolutionary genetics analysis across computing platforms. Mol Biol Evol 35:1547–1549PubMedPubMedCentralCrossRefGoogle Scholar
  29. Lehman SG (1925) Studies on treatment of cotton seed. N C Agric Exp Stn Tech Bull 26:71Google Scholar
  30. Lehninger AI (1993) Biochemistry. Worth, New YorkGoogle Scholar
  31. Li PH (1994) Crop plant cold hardiness. In: Boote KJ et al (eds) Physiology and determination of crop yield. ASA, CSSA and SSSA, Madison, WIGoogle Scholar
  32. Linder CR (2000) Adaptive evolution of seed oils in plants: accounting for the biogeographic distribution of saturated and unsaturated fatty acids in seed oils. Am Nat 156(4):442–458PubMedCrossRefGoogle Scholar
  33. Liu Q, Singh S, Chapman K, Green A (2009) Bridging traditional and molecular genetics in modifying cottonseed oil. In: Paterson AH (ed) Genetics and genomics of cotton, vol 3. Crops and models. Plant genetics and genomics. Springer, New YorkGoogle Scholar
  34. Liu Q, Wu M, Zhang B, Shrestha P, Petrie J, Green AG, Singh SP (2017) Genetic enhancement of palmitic acid accumulation in cotton seed oil through RNAi down-regulation of ghKAS2 encoding β-ketoacyl-ACP synthase II (KASII). Plant Biotech J 15:132–143CrossRefGoogle Scholar
  35. López-Vinyallonga S, Soriano I, Susanna A, Montserra JM, Roquet C, Garcia-Jacas N (2015) The polyploid series of the Achillea millefolium aggregate in the Iberian Peninsula investigated using microsatellites. PLoS ONE 10(6):e012861CrossRefGoogle Scholar
  36. Ludwig CA (1932) The germination of cottonseed at low temperature. J Agric Res 44(4):367–380Google Scholar
  37. Meadows SA (2012) Evaluation of seed oil melting points of multiple species at a common latitude. MS Thesis, University of Texas, AustinGoogle Scholar
  38. Mercer KL, Perales H (2010) Evolutionary response of landraces to climatic change in centres of diversity. Evol Appl 3:480–493PubMedPubMedCentralCrossRefGoogle Scholar
  39. Miura K, Agetsuma M, Kitano H, Yoshimura A, Matsuoka M, Jacobsen SE, Ashikari M (2009) A metastable DWARF epigenetic mutant affecting plant stature in rice. Proc Natl Acad Sci USA 106:11218–11223PubMedCrossRefPubMedCentralGoogle Scholar
  40. Moffet JO, Cobb HB, Rummen DR (1980) Bees of potential value as pollinators in the production of hybrid cotton seed on the higher plains of Texas. In: Proceedings of Beltwide cotton production research conference (BCPRC`80), National Cotton Council of America, Memphis, TN, pp 268–270Google Scholar
  41. Noblet A, Leymarie J, Christophe Bailly (2017) Chilling temperature remodels phospholipidome of Zea mays seeds during imbibition. Sci Rep 7:8886PubMedPubMedCentralCrossRefGoogle Scholar
  42. Orchard T (1977) Estimating the parameters of plant seedling emergence. Seed Sci Tech 5:61–69Google Scholar
  43. Peakall R, Smouse PE (2012) GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research-an update. Bioinformatics 28:2537–2539PubMedPubMedCentralCrossRefGoogle Scholar
  44. Pettigrew WT (2002) Improved yield potential with early planting cotton production system. Agron J 94:997–1003CrossRefGoogle Scholar
  45. Płażek A, Dubert F, Kopeć P et al (2018) Seed hydropriming and smoke water significantly improve low-temperature germination of Lupinus angustifolius L. Int J Mol Sci 19(4):992.  https://doi.org/10.3390/ijms19040992 CrossRefPubMedCentralGoogle Scholar
  46. Raphael JPA, Gazola B, Nunes JGS, Maedo GC, Rosolem CA (2017) Cotton germination and emergence under high diurnal temperatures. Crop Sci 57:2761–2769CrossRefGoogle Scholar
  47. Rikin A, Atsmon D, Gitler C (1979) Chilling injury in cotton (Gossypium hirsutum L.): prevention by abscisic acid. Plant Cell Physiol 20:1537–1546CrossRefGoogle Scholar
  48. Rikin A, Atsmon D, Gitler C (1984) Quantitation of chill-induced release of tubulin-like factor and its prevention by abscisic acid in Gossypium hirsutum L. Plant Physiol 71:747–748CrossRefGoogle Scholar
  49. RStudio Team (2015) RStudio: Integrated Development for R. RStudio Inc., Boston, MAGoogle Scholar
  50. Scott S, Jones R, Williams W (1984) Review of data analysis methods for seed germination. Crop Sci 24:1192–1199CrossRefGoogle Scholar
  51. Shim J, Torollo G, Angeles-Shim RB, Cabunagan RC, Choi IR et al (2015) Rice tungro spherical virus resistance into photoperiod-insensitive japonica rice by marker-assisted selection. Breed Sci 65:345–351PubMedPubMedCentralCrossRefGoogle Scholar
  52. Shim J, Mangat PK, Angeles-Shim RB (2018) Natural variation in wild Gossypium species as a tool to broaden the genetic base of cultivated cotton. J Plant Sci Curr Res 2:005Google Scholar
  53. Speed TR, Kreig DR, Jividen G (1996) Relationship between cotton seedling cold tolerance and physical and chemical properties In: Dugger P and Richter DA (eds.), Proceedings of the Beltwide Cotton Conference, National Council, Memphis TN, pp 1170–1171Google Scholar
  54. Stanway V (1960) Laboratory germination of cottonseed at three different temperatures. Proc Assoc Off Seed Anal 50:97–100Google Scholar
  55. Tehrani MS, Mardi M, Saeidi H, Gharehyazi B, Assadi M (2008) Transferability of genomic and EST-microsatellites from Festuca arundinacea Schreb. to Lolium persicum Boiss. and Hohen. ex Boiss. Int J Bot 4:476–480CrossRefGoogle Scholar
  56. Thompson GA, Li C (1997) Altered fatty acid composition of membrane lipids in seeds and seedling tissues of high-saturate canolas. In: Williams JP, Khan MU, Lem NW (eds) Physiology, biochemistry and molecular biology of plant lipids. Springer, DordrechtGoogle Scholar
  57. Thompson CN, Hendon BR, Mishra D, Rieff JM, Lowery CC et al (2019) Cotton (Gossypium hirsutum L.) mutants with reduced levels of palmitic acid (C16:0) in seed lipids. Euphytica 215:112CrossRefGoogle Scholar
  58. Turner JH (1914) History of Acala cotton varieties bred for San Joaquin Valley, California. US Agricultural Research Service. Western RegionGoogle Scholar
  59. Turner JH (1950) Natural crossing in cotton. In: Cotton improvement conference proceedings II, p5Google Scholar
  60. Upchurch RG (2008) Fatty acid unsaturation, mobilization and regulation in the response of plants to stress. Biotechnol Lett 30(6):967–977PubMedCrossRefPubMedCentralGoogle Scholar
  61. Weir BS (1990) Genetic data analysis: methods for discrete population genetic data. Sinauer Associates Inc, SunderlandGoogle Scholar
  62. Wendel JF, Cronn RC (2003) Polyploidy and the evolutionary history of cotton. Botany Publication and Papers, p 23Google Scholar
  63. Wendel JF, Grover CE (2015) Taxonomy and evolution of the cotton genus, Gossypium In: Fang DD, Percy RG (eds.), Cotton. Agronomy Monographs, 2nd edn. American Society of Agronomy, Inc., Crop Science Society of America, Inc., and Soil Science Society of America, Inc, Wisconsin, pp 745–781Google Scholar
  64. Yu JZ, Fang DD, Kohel RJ, Ulloa M, Hinze LL et al (2012) Development of a core set of SSR markers for the characterization of Gossypium germplasm. Euphytica 187:203–213CrossRefGoogle Scholar
  65. Yu X, Li A, Li W (2015) How membranes organize during seed germination: three patterns of dynamic lipid remodeling define chilling resistance and affect plastid biogenesis. Plant Cell Environ 37(8):1391–1403CrossRefGoogle Scholar

Copyright information

© The Author(s) 2019

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors and Affiliations

  1. 1.Department of Plant and Soil Science, College of Agricultural Sciences and Natural ResourcesTexas Tech UniversityLubbockUSA

Personalised recommendations