Heterosis On Eucalyptus Hybrids And Molecular Marker-assisted Selection

Genetic parameters were estimated for growth and cold hardiness traits using factorial-mating hybrids of E. urophylla×E. tereticornis. Fast growth and cold hardiness hybrids were selected. Two hundreds and six EST-SSR markers were developed by utilizing the EST resources of Eucalyptus in GenBank. And these EST-SSR markers plus genomic SSR markers published previously were used to evaluate genetic variation among 20 different genotypes of eucalyptus. Correlation analysis between genetic similarity based on SSR and phenotypic data of hybrids was carried out to predict heterosis. Main conclusion as followed:1. Genetic parameters were estimated for CH over ages 0.5, 1.5 and 2.5 using factorial mating hybrids of E. urophylla×E. tereticornis and maternal open-pollinated progenies of E. urophylla. Variance component due to female effect was statistically non-significant in CH over all ages in both hybrids and progenies, whereas male effect was in the hybrids consistently significant (P<0.01 or 0.05) and female×male effect was significant at age 2.5 (P<0.05) but not significant at ages 0.5 and 1.5. Both additive and dominance variance coefficients remained relatively stable over ages. Within the hybrids, the dominance/additive variance ratio (σD2 /σA2) was lower than 1.0 and fluctuated markedly with age, ranging from 0.10 at age 0.5 to 0.86 at age 1.5. Estimates of narrow-sense heritability (h2) for CH were low to moderate, ranging from 0.06 to 0.29. Age-age additive genetic correlation varied greatly with population and age combination. Cold hardiness is under a low to moderate level of genetic control and could be improved by hybrid breeding in Eucalyptus.2. We employed 57 Eucalyptus urophylla×E. tereticornis hybrids for selection in both growth and cold hardiness traits. The results indicated that for 4-year-old growth traits, including height (H4), diameter at breast height (D4) and volume (V4), maternal and paternal effects was significant upon 0.01 or 0.05 level, while the female×male interaction effects was not reached to the significant level. However, the female, male and their interaction effects were all significant on cold hardiness for 2.5-year-old testing. Additive genetic correlations between volume and cold hardiness were 0.28 and 0.69 at the ages of 1.5 and 2.5, respectively, which may indicate that the selection on growth would increase cold hardiness in a certain way. A total of 11 superior hybrids and 12 individual trees were selected out as their combined performance in growth and cold hardiness were concerned.3. Ten female Eucalyptus urophylla and ten male E. tereticornis were selected as parents to analyze the combining ability of growth traits and cold hardiness. The results indicated that for 4-year-old growth traits, including height (H4), diameter at breast height (D4) and volume (V4), maternal general combining ability (GCA) was significant at 0.01 level but no significant ??in male parents, while special combing ability (SCA) was significant in height trait but no significant in DBH and volume traits. However, maternal and paternal GCA and SCA were all significant in 2.5-year-old cold hardiness. No. 3,7,9,10 of E. urophylla and No.11,12,18 of E. tereticornis had high GCA value, while No. 9 had the highest GCA value in cold hardiness. Among different combinations, E.u05×E.t12 had the highest SCA value in growth traits and E.u06×E.t14 was the best in cold hardiness.4. A set of 206 simple sequence repeat (SSR) markers were developed from 8,266 unigenes assembled with 23,977 Eucalyptus expressed sequence tags (ESTs) deposited in GenBank. A total of 891 primer pairs were designed, of which 636 and 216 yielded polymerase chain reaction (PCR) products and scorable sequencing traces, respectively, in maternal E. urophylla (P1) and/or paternal E. tereticornis (P2) parents of an F1 mapping population. Sequence alignment revealed that 206 amplicons were identical to the respective ESTs, whereas the other 10 were probably resulted from PCR errors. All the 206 EST-SSRs were tested on 10 E. urophylla and 10 E. tereticornis genotypes, and 186 were scorable in allele detection, including 171 being polymorphic and 15 monomorphic. For the 186 scorable markers, a total of 1201 alleles were obtained and the number of alleles per locus ranged from 1 to 15 (mean 6.46). For the 171 polymorphic markers, the observed heterozygosity (Ho), the expected heterozygosity (He) and the polymorphic information content (PIC) varied with locus from zero to 1.0 (mean 0.4545), 0.0987 to 0.9359 (mean 0.6942) and 0.0940 to 0.9059 (mean 0.6409), respectively. A relatively high level of genetic variation was observed among the 20 eucalypt genotypes.5. Three groups with different definition containing 167 genomic linear DNA sequences in GenBank were downloaded and the flanking regions of 47 SSR containing sequences were designed as primer for PCR. After optimizing, 7 of them were successfully re-sequenced to detect reliability of SSR sequences. To evaluating the polymorphism within two species, E. urophylla and E. tereticornis, all of the 7 markers were genotyped via an automated sequencer and of which 5 were available to produce an average of 7 alleles per locus, an average observed heterozygosity of 0.5399 and an average of expected heterozygosity of 0.7818. Polymorphism information content (PIC) value varied from 0.6711 to 0.7891, with a mean of 0.7262.6. Ninety genomic SSR were selected to evaluate the genetic variation within two species. A total of 998 alleles were obtained and the number of alleles per locus ranged from 3 to 18 (mean 11). For the 90 polymorphic markers, the observed heterozygosity (Ho), the expected heterozygosity (He) and the polymorphic information content (PIC) varied with locus from zero to 1.0 (mean 0.5805), 0.1693 to 0.9559 (mean 0.8197) and 0.1572 to 0.9265 (mean 0.7798), respectively. Correlation analysises were carried out between Jaccard similarity based on 171 EST-SSR and 90 genomic SSR markers and hybrid performance, heterosis, SCA and GCA respectively. The results showed that: significant positively correlation was found between EST-SSR derived genetic similarity and growth traits on 1.5 years age. Both EST-SSR and genomic SSR markers genetic similarities were significant negatively correlated with hybrid performance and GCA on CH traits at 2.5 year age. Moreover, hybrid performances on all traits over ages were significant positively correlated with heterosis, GCA and SCA. Heterosis was significant positively correlated with SCA but only at special age with GCA, and the correlation coefficient was low.In conclusion, heterosis on Eucalyptus may be a dynamic process variated following environment on different traits and ages. Besides genetic control, hetersis may affected by gene expression in different ages or directely affected by environment. It was unilateral and inaccurate by using one way to predict heterosis on Eucalyptus.. Further study should be concerned on how to select molecular markers for their relationship to the alleles and linked to QTLs involved in the heterotic traits considered.