Introgression of Day-Neutral Genes in Primitive Cotton Accessions: II. Predicted Genetic Effects

cotton are an important source of useful genetic variability (Percival, 1987; Meredith, 1991; McCarty and JenPrimitive accessions of cotton, Gossypium hirsutum L., may prokins, 1992; McCarty et al., 1995). Percival and Kohel vide useful traits for cultivar development. Genetic effects for yield, yield components, and fiber traits were analyzed for five generations (1990) reviewed the collection, distribution, and evaluaof day-neutral progenies. The genetic material was derived from introtion of Gossypium germplasm. gressing day-neutral genes from ‘Deltapine 16’ into 16 primitive accesMost tropical primitive accessions of cotton are pesions with single and multiple backcrosses creating 80 populations rennial, photoperiod sensitive, and do not flower under representing one to four doses of the unadapted accession. Yield and the long days of the U.S. cotton belt. In their native fiber traits were determined from field plot studies conducted for 3 yr. habitat they flower during short days in winter months Significant accession effects were detected for all the traits studied. and remain vegetative during long summer days. The Significant generation main effects were found for three yield traits use of these primitive accessions has been limited beand one fiber trait. As expected, yield was predicted to decrease with cause of their flowering response. A backcross-breeding more cycles of backcrossing to the accession. Accessions 3 generation program has been in place for a number of years to interactions were detected for some traits which indicated that not all generations were having equal effects. This genetic analysis provides incorporate day-neutral genes into the primitive accesuseful information when utilizing these accessions. sions (McCarty et al., 1979; McCarty and Jenkins, 1992, 1993). McCarty et al. (1995) evaluated F5, BC1F5, BC2F5, BC3F5, and BC4F5 progenies of 16 day-neutral germA genetic base could result in crop cultivars being highly vulnerable to stresses and it could plasm accessions for several agronomic and fiber traits. also restrict genetic gain. Therefore, it is important that Results indicated there was useful variability for the we expand the genetic diversity of cotton with new and traits studied in the day-neutral lines. The objective of unrelated sources of germplasm. Primitive accessions of the present study was to determine genetic effects of 16 accessions in five generations (F5, BC1F5, BC2F5, BC3F5, and BC4F5) for agronomic and fiber traits, and J.C. McCarty, Jr., and J.N. Jenkins, USDA-ARS, Crop Sci. Res. Lab., P.O. Box 5367, Mississippi State, MS 39762; J. Zhu, Agronomy Dep., to aid cotton breeders in using these accessions in their Zhejiang Agriculture Univ., Hangzhou 310029, China. Work was combreeding programs. pleted while J. Zhu was a visiting scientist at Mississippi State Univ. Contribution of the USDA-ARS in cooperation with the Mississippi Agric. and Forestry Exp. Stn. Received 16 Sept. 1996. *Corresponding MATERIALS AND METHODS author ([email protected]). Day-neutral genes from Deltapine 16 were introgressed into 16 primitive accession with single and multiple backcrosses Published in Crop Sci. 38:1428–1431 (1998). MCCARTY ET AL.: PREDICTED GENETIC EFFECTS OF PRIMITIVE COTTON ACCESSIONS 1429 Table 1. Predicted main effects of germplasm accessions for yield and fiber traits. Lint Lint Boll Seed 50% Span 2.5% Span Accession yield percentage size index Elongation Micronaire length length Strength kg ha21 % g % mm kNm kg21 T53 (A1) 221.5 0.23 0.27** 0.01 20.30** 0.20** 20.18** 20.60** 0.07 T78 (A2) 246.6** 21.34** 20.27** 0.05 0.17** 20.30** 20.15** 20.40** 20.35 T87 (A3) 29.4 20.05 0.12* 0.01 0.16** 20.07 0.04 0.09 3.57* T88 (A4) 5.4 0.33** 0.32** 0.10* 20.03 0.05 20.04 0.31* 24.25** T91 (A5) 276.8** 21.41** 20.21** 0.26* 0.07 0.12** 20.05 20.36** 5.03** T106 (A6) 126.2** 0.85** 0.05 20.09 0.13* 20.05 0.21** 0.64** 0.41 T119 (A7) 214.3 20.25 20.05 20.21* 0.03 20.14** 20.02 0.12* 22.25 T158 (A8) 22.4 0.05 0.07** 20.07 20.03 0.06 0.03 0.09 20.42 T168 (A9) 20.1 20.29* 0.02 20.11 20.12** 0.10* 20.27** 20.91** 22.03 T174 (A10) 104.8** 0.78** 20.05 20.23* 20.01 20.00 0.07 0.43** 1.65 T175 (A11) 20.0 0.41** 0.11 0.15* 0.03 0.08 20.02 20.07 22.30** T228 (A12) 253.6** 21.18** 20.15** 20.19* 20.24** 20.01 0.01 20.20** 2.63** T257 (A13) 25.8 20.33 20.10* 0.10 20.20** 0.04 0.16** 0.22** 4.18** T326 (A14) 230.8* 0.92** 20.16** 20.05 0.04 20.05 0.30** 0.79** 3.47* T612 (A15) 241.8** 0.83** 0.12** 0.15* 20.23** 20.06* 20.06 0.11 26.81** T1149 (A16) 1.6 0.43** 20.10* 20.03 0.53** 0.03 20.03 20.25* 22.63* *, ** Significantly different from zero at the 0.05 and 0.01 levels of probability, respectively. creating 80 populations representing one to four doses of the effects (Miller 1974; Zhu, 1989). Since there were 12 blocks for 3 yr, the degrees of freedom were 11 for the jackknifing. A unadapted accession. The development of the 80 populations were described by McCarty et al. (1995). Experiments were two-tail t-test was employed for testing significance of genetic conducted in 1989, 1990, and 1991 in which yield and fiber parameters studied. All the data analyses were conducted on traits were measured (McCarty et al., 1998). a PC computer with programs written in C language. Data for yield and fiber traits were analyzed by mixed linear model approaches. A linear model was used by including RESULTS AND DISCUSSION random factors for years as environments (E), accessions (A), generations (G), and their interaction (AG, AE, GE, AGE) We detected significant accession effects for all the along with block factor (B) and residual errors (e). Since traits studied (Table 1). Accessions 6 (T-106) and 10 random effects can not be estimated, they were predicted by (T-174) had large effects for high lint yield, high lint using an adjusted unbiased prediction (AUP) method (Zhu, percentage, and long 2.5% span length. Accession 6 also 1993; Zhu and Weir, 1996). Jackknifing over blocks within years was used to calculate the standard errors of random had higher elongation and long 50% span length, while Fig. 1. Predicted interaction effects of accession by generation for yield traits. Wide blank columns is the main effect of accession. Five narrow columns from left to right are generations F5, BC1F5, BC2F5, BC3F5, and BC4F5, respectively. 1430 CROP SCIENCE, VOL. 38, NOVEMBER–DECEMBER 1998 Table 2. Predicted main effects of generations for yield and fiber traits.† Lint Lint Boll Seed 50% Span 2.5% Span Generation yield percentage size index Elongation Micronaire length length Strength kg ha21 % g % mm kNm kg21 F5 73.1** 0.48** 20.09 20.09 0.19 20.11 20.09 0.00 22.26 BC1F5 56.7* 0.55** 0.12** 0.07 0.06 0.07* 0.01 0.00 20.78 BC2F5 10.1 0.55** 0.02 20.13 20.02 20.01 0.01 0.00 20.85 BC3F5 271.6* 20.78** 20.12** 0.02 20.18 0.02 20.00 0.00 1.96 BC4F5 268.2* 20.79** 0.06* 0.14 20.05 0.04 0.08 0.00 1.94 *, ** Significantly different from zero at the 0.05 and 0.01 levels of probability, respectively. † Because of rounding (two decimal places) small contributions of some traits appear as 0.00 in the table. Accession 10 had small seeds. Accession effects of these three yield traits but only for one fiber trait (Table 2). The highest lint yield was expected in the F5 generation two lines were not significant for other yield and fiber traits. Therefore Accessions 6 and 10 may be suitable which had more genetic background from the high yield cultivar (Deltapine 16) than other generations. Lint for utilization in cotton improvement programs. Accession 5 (T-091) had the highest positive main effect for yield decreased with more cycles of backcrossing to the accessions. This tendency was not observed for other fiber strength but was poor for yield traits and 2.5% span length. Main effects of fiber strength and both span traits. Since accession 3 generation interaction (AG) was lengths were large for Accession 13 (T-257) which also showed acceptable performance for yield traits. Therethe major component of genetic variation for all the traits (McCarty et al., 1998), there could be some deviafore, Accession 13 may be useful as breeding material for improving both fiber strength and length. tion for genetic behavior of accessions in different generations. The predicted AG interaction effects along Significant generation main effects were detected for Fig. 2. Predicted interaction effects of accession by generation for fiber traits. Wide blank columns is the main effect of accession. Five small columns from left to right are generations F5, BC1F5, BC2F5, BC3F5, and BC4F5, respectively. MCCARTY ET AL.: PREDICTED GENETIC EFFECTS OF PRIMITIVE COTTON ACCESSIONS1431 with the accession main effects, which were significantly to the improved yield ofBC2F5 of Accession 6. Thegenetic entry with the strongest fiber was found fromdifferent from zero (P # 0.05), are presented in Fig. 1for yield traits and in Fig. 2 for fiber traits. Although F5 of Accession 10, which increase fiber strength byabout 17.9 kNm kg21. Accession 10 also increased 2.5%main accession effects of lint yield were large for Acces-sions 6 and 10, not all the generations in these two span length (1.09 mm), 50% span length (0.58 mm), andeven lint yield (108 kgha21) in the F5 generation.accessions were equally important for improving yield.BC1F5 in Accession 6 and BC2F5 in Accession 10 tended Genotype, environment, and their interaction af-fected the phenotypic behavior of genetic entries de-to decrease yield by about 100 kgha21, while BC2F5 inAccession 6 and BC3F5 in Accession 10 increased yield rived from primitive accessions of cotton. It was re-vealed that genetic analysis of the primitive accessionover 150 kg ha21. The yield gain of BC2F5 in Accession6 was due to the high lint percentage and large bolls, of cotton could result in a better understanding of thebreeding merit for specific mating generations of differ-while high yield of BC3F5 in Accession 10 might be theresult of high lint percentage and median size of boll ent accessions.(indicating large number of bolls). Although Accession3 (T-087) had positive main effects for boll size but notREFERENCESfor lint yield and lint percentage, F5 in Accession 3 had McCarty, J.C., Jr., J.N. Jenkins, W.L. Parrott, and R.G. Creech. 1979.high lint yield with very small boll (indicating largeThe conversion of photoperiodic primitive race stocks stocks of cotton to day-neutral stocks. Miss. Agric. and Forestry Exp. Stn.number of bolls) and high lint percentage. It was impliedRes. Rep. 4(19):4.by this analysis that F5 in Accession 3 was a differentMcCarty, J.C., Jr., and J.N. Jenkins. 1992. Characteristics of 79 day-type of high yield material as compared with BC2F5 inneutral primitive race accessions. Miss. Agric. and Forestry Exp.Accession 6. Genetic entries derived from these twoStn. Tech. Bull. 184.materials might be further used as two parents of a cross McCarty, J.C., Jr., and J.N. Jenkins. 1993. Registration of 79 dayneutral primitive cotton germplasm lines. Crop Sci. 33:351.which might result in higher yield.McCarty, J.C., Jr., J.N. Jenkins, and B. Tang. 1995. Primitive cottonThough Accession 13 tended to have strong and longgermplasm: Variability for yield and fiber traits. Miss. Agric. andfiber because of the positive main effects, it was notForestry Exp. Stn. Tech. Bull. 202.true for its F5 generation. Actually BC4F5 in Accession McCarty, J.C., Jr., J.N. Jenkins, and J. Zhu. 1998. Introgression of day-neutral genes in primitive cotton accessions: I. Genetic vari-13 was good for fiber strength and both span lengths.ances and correlations. Crop Sci. 38:1425–1428 (this issue).It was found by the AG interaction analysis that F5 inMeredith, W.R., Jr. 1991. Contributions of introductions to cottonAccession 10 had stronger and longer fiber also withimprovement. p. 127–146. In H.L. Shands and L.E. Weisner (ed.)large value of boll size and seed index. Therefore, F5 inUse of plant introductions in cultivar development. Part I. CSSA,Accession 10 and BC2F5 in Accession 6 could serve asMadison, WI. Miller, R.G. 1974. The jackknife: A review. Biometrika 61:1–15.a breeding material for further improving fiber qualityPercival, A.E. 1987. The national collection of Gossypium germplasm.of cotton.So. Coop. Series Bull. 321.When there are significant accession main effects and Percival, A.E., and R.J. Kohel. 1990. Distribution, collection, andaccession 3 generation interaction effects, the geneticevaluation of Gossypium. Adv. Agron. 44:225–256.merit of an accession in a specific generation is deterZhu, J. 1989. Estimation of genetic variance components in the general mixed model. Ph.D. Dissertation, NC State University, Raleighmined by the sum of main and interaction effects (A 1(Diss. Abstr. DA8924291).AG). In the present study, Accession 6, in BC2F5 provedZhu, J. 1993. Methods of predicting genotype value and heterosis forto make the highest contribution to yield with a 312offspring of hybrids. J. Biomathematics, 8(1):32–44.kg ha21 increase over the population mean. The yield Zhu, J., and B.S. Weir. 1996. Diallel analysis for sex-linked and mater-nal effects. Theor. Appl. Genet. 92:1–9.components lint percentage and boll size contributed