| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dep. of Plant Sciences, North Dakota State Univ., Fargo, ND 58105-5051
* Corresponding author (marcelo.carena{at}ndsu.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIX 1REFERENCES |
|---|
Abbreviations: GCA, general combining ability • LSC, Lancaster Sure Crop • RCBD, randomized complete block design • SCA, specific combining ability • SOV, source of variation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIX 1REFERENCES |
|---|
The cyclical use of a limited number of inbred lines across the U.S. Corn Belt for more than 30 years has reduced the genetic diversity of hybrids (Hallauer and Miranda, 1988). The importance of genetic diversity has been emphasized since the shift from double-cross to single-cross hybrids, and especially after the southern corn leaf blight (caused by Bipolaris maydis [Nisik. & Miyake] Shoem.) outbreak on hybrids carrying the T-cytoplasm for male sterility in the USA in the 1970s (Duvick, 1981; Hallauer and Miranda, 1988). Improved open-pollinated populations are valuable for the development of inbred lines, especially when considering the importance of having genetic diversity in the germplasm pool. Differences in background, origin, and level of heterozygosity within and among populations are the basis of that diversity.
The identification of populations as sources of inbred lines is based on their agronomic performance, presence of useful genetic variance, high population mean, and the heterosis observed from using them in crosses. Midparent heterosis values provide the basis for the identification of heterotic patterns among a fixed set of populations, and average heterosis and specific heterosis are the components in the expression of midparent heterosis. Average heterosis is indicative of the superiority of population crosses over the midparent values, while specific heterosis indicates the heterosis observed in certain crosses (Hallauer and Miranda, 1988). Therefore, the utilization of midparent heterosis values is both a practical and effective method to identify heterotic responses among parents.
The diallel mating design has been largely utilized to identify heterotic patterns when the number of parents is small. Several methods have been proposed for the analysis of a set of n parents and their progeny. When data from parents and one set of F1s are available (reciprocal crosses not included), Griffing's Method 2 (Griffing, 1956), and Analyses II and III proposed by Gardner and Eberhart (1966) are suitable. The three methods are related, and similar information can be obtained from all of them. Griffing's Method 2 and Gardner and Eberhart's Analyses II provide estimates that are linear functions of the estimates of Gardner and Eberhart's Analysis III, and no major advantage or disadvantage can be found across methods (Baker, 1978). When the primary interest of the experimenter is the performance of the parents in crosses, the subdivision of sources of variation given by Gardner and Eberhart Analysis III has been widely used for the evaluation of open-pollinated populations (Gardner and Eberhart, 1966; Gerrish, 1983; Mungoma and Pollak, 1988). Information about the performance of the populations per se, as well as in crosses, is provided on the basis of the components as follows: Parents, Parents vs. Crosses, and Crosses (partitioned into GCA and SCA).
The objective of this research was to identify alternative heterotic patterns for hybrid development in the northern Corn Belt. We evaluated the potential of population crosses among 10 maize open-pollinated populations from diverse and common origins.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIX 1REFERENCES |
|---|
Crossing Procedure
Three-paired rows of about 20 plants each were grown to obtain the population cross seed. Crosses were made using each male parent to pollinate no more than two ear shoots. The ears obtained from the six rows were harvested, dried and shelled individually, and two balanced bulks were obtained and put in cold storage for future experimentation. This procedure was repeated for each of the 45 population crosses.
Six rows of about 120 plants each were also planted to increase the parental populations per se. Chain crosses within each population were made using each male parent to pollinate only one shoot. The ears obtained from each population were harvested, dried and shelled individually, and two balanced bulks were obtained and put in cold storage for future experimentation.
Experimental Design
The 10 parental populations, their respective 45 crosses, and 9 checks (3 public elite hybrids, 4 private single-cross hybrids, and 2 improved public open-pollinated varieties) were evaluated in experiments with two replicates, arranged in a partially balanced lattice design (8 by 8) at four locations in North Dakota (Fargo, Casselton, Barney, and Oakes), and in a rectangular lattice (10 by 7) at three locations in northern Iowa (Calumet, Kanawha, and Nashua) in 2002. Six additional entries were added in all Iowa experiments as fillers, but they were not included in the analyses. Experimental units consisted of single-row plots and two-row plots in North Dakota and Iowa, respectively.
Characters Studied
Grain yield (adjusted to 155 g kg–1 grain moisture and expressed as Mg ha–1), grain moisture at harvest (g kg–1), and stalk lodging (%) were measured at all seven locations. Root lodging (%) was measured at all locations except Nashua, IA. The experiments were overseeded and thinned to the desired stand at each location. Plots were machine harvested at all locations except Casselton, ND, where the experiment was hand-harvested to record ear measurements (data not shown). Six observations for both grain yield and grain moisture were missing at Oakes, and four and three observations for grain yield and grain moisture, respectively, were missing at the Fargo location. Missing data were estimated by means of the intrablock formula for incomplete block designs (Cochran and Cox, 1992).
Statistical Procedures
Analyses of variance were performed for all traits at each location, as well as across locations using the SAS Lattice procedure for 8 by 8 lattices and SAS GLM procedure for 10 by 7 lattices (SAS Institute, Inc., 1990). Efficiency of the lattices relative to a randomized complete block design (RCBD) was calculated following Fisher's procedure (Steel et al., 1997). Because the F-tests indicated significant differences among the genotypes for most of the traits studied, the genotype source of variation (SOV) was partitioned following Gardner and Eberhart's Analysis III model (Gardner and Eberhart, 1966).
General and specific combining ability effects were calculated using DIALLEL-SAS (Zhang and Kang, 1997) for Griffing's Method 4. Combined analyses of variance for all traits were performed to detect significant heterotic expressions. Error variances were tested for homogeneity using the Bartlett's test (Steel et al., 1997) before combining data. Combined error mean squares were calculated by pooling the corresponding individual error mean squares weighted by their respective degrees of freedom. Means adjusted by blocks were used when the relative efficiency of lattices was higher than 105% when compared with the RCBD (Cochran and Cox, 1992); otherwise, unadjusted means were used for the combined analysis. Intrablock error mean squares instead of RCBD error mean squares were used as a denominator in the F-test when relative efficiency of lattice designs was greater than 105% compared with the RCBD. For the combined analyses, locations were considered random effects and genotypes (entries and checks) fixed effects. Fisher's protected LSDs were used to compare means at P 
0.05. Because a nonsignificant interaction between genotypes and locations was found for the traits under study, a pooled error mean square was used as the error term to calculate LSD values for each trait. Expression of heterosis was measured for all traits as midparent heterosis (Falconer and Mackay, 1997). The Student's t test of significance was used to test the null hypothesis that midparent heterosis, GCA, and SCA effect values were equal to zero. Simple linear regression analyses were performed to study the degree of association of the performance of the populations per se and their corresponding GCA values for grain yield and grain moisture at harvest.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIX 1REFERENCES |
|---|
|
0.01). Because error mean squares were found to be homogeneous, means from the seven locations were combined. Combined F-tests indicated that significant differences (P 0.05) existed for Genotypes, as well as for all orthogonal components, except SCA (Table 1). The average grain yield for populations was 4.28 Mg ha–1, and ranged from 3.51 Mg ha–1 for BS5 to 5.02 Mg ha–1 for BS22(R)C7. The low performance of BS5 per se was expected because it was not improved for grain yield. In contrast, BS22(R)C7, NDSAB(MER)C12, NDSM(M)C7, and BS21(R)C7 have undergone several cycles of recurrent selection for grain yield.
General combining ability values ranged from –0.74 for BS5 to 1.07 for BS21(R)C7 (Table 2). The range of GCA effects found in this study was similar to that reported by Mungoma and Pollak (1988) and Hallauer and Sears (1968). The former study reported GCA values from –0.87 to 1.49 when studying 10 temperate (from early to late maturity) maize populations and their crosses across seven locations in Iowa, Missouri, Tennessee, Kentucky, and Kansas in 1985. Hallauer and Sears (1968) reported GCA values that ranged from –2.65 to 2.31 when evaluating nine temperate populations and their crosses across six environments in Iowa. Similarly, Mickelson et al. (2001) evaluated the heterotic relationships among nine temperate and subtropical populations in a broad set of environments in Mexico, Zimbabwe, and the USA. They found that the highest GCA value of 1.34 corresponded to BSSS(R)C11 across locations. A positive linear relationship between grain yield of the populations per se and their corresponding GCA values was found (P < 0.03), but the degree of association between the two variables, measured as the coefficient of determination, was 0.46, indicating a weak association.
|
The average grain yield for crosses was 5.12 Mg ha–1, and ranged from 3.87 Mg ha–1 for NDSG(M)C15 x NDSM(M)C7 to 6.75 Mg ha–1 for BS21(R)C7 x BS22(R)C7 (Table 3). Grain yield of the latter cross was higher than the mean of the checks (6.22 Mg ha–1) and not significantly different from the third highest yielding hybrid check. The high grain yield values observed for some of the crosses are a valuable measure of the potential of the populations to be used either as sources of inbred lines for hybrid development or directly as population hybrids.
|
Twenty-one of the population crosses had midparent heterosis values different from zero at P 0.05 (Table 3). Midparent heterosis ranged from –0.26 Mg ha–1 for NDSM(M)C7 x NDSG(M)C15 to 2.15 Mg ha–1 for BS21(R)C7 x CGL(S1–S2)C5. These values represent –9 and 54% of the corresponding midparent grain yield, respectively. The average midparent heterosis [0.84 Mg ha–1 (19.6%)] was similar to the magnitude of heterosis reported by Hallauer and Miranda (1988). They calculated a value of 19.5% midparent heterosis averaged across 40 experiments from 1893 to 1979, by analyzing a total of 611 populations and 1394 population crosses. Because heterosis has been described as one of the best measures of genetic diversity (Hallauer and Eberhart, 1966; Hallauer and Miranda, 1988, Mungoma and Pollak, 1988), parental populations of top performing crosses are expected to be genetically more diverse than the populations that exhibited little, absent, or negative heterosis.
The best heterotic response corresponded to the population cross BS21(R)C7 x CGL(S1–S2)C5. The population BS21(R)C7 has 54% Reid Yellow Dent (RYD) and no LSC background. On the other hand, CGL(S1–S2)C5 was developed by intercrossing 26 elite inbreds from the Lancaster heterotic group. Therefore, the presence of the two most exploited heterotic groups in the parents' background explains the high level of heterosis observed in the cross. However, the heterosis observed in other crosses between parents belonging to the RYD and LSC groups was not as expected. The population hybrid CGSS(S1–S2)C5 x CGL(S1–S2)C5 is the cross that best represents the heterotic pattern RYD x LSC in this study, based on the background of the parental lines. Crosses involving populations from RYD and LSC heterotic groups, which were selected for earliness and adaptation to cooler environments, have retained the same level of performance observed in their original environments (Moreno-Gonzalez et al., 1997). However, the CGSS(S1–S2)C5 x CGL(S1–S2)C5 cross had a midparent heterosis value [0.90 Mg ha–1 (21%)] that was close to the observed average heterosis. Its grain yield was 4.94 Mg ha–1, ranking below the crosses mean (5.12 Mg ha–1), and it was significantly different from the 10 highest yielding crosses. Midparent heterosis between RYD and LSC has already been reported to be either similar or smaller than the observed heterosis between other maize cultivars, such as Leaming and Midland, Leaming and LSC, and Midland and RYD (Kauffman et al., 1982).
The best heterotic combinations for CGSS(S1–S2)C5 were obtained when crossed with BS21(R)C7 and BS22(R)C7 (1.75 and 1.38 Mg ha–1 midparent heterosis, respectively). The three populations have a great proportion of RYD in their background (100, 54, and 45%, respectively), and therefore, they are in some way related (Hallauer et al., 2000; University of Guelph, 2002). However, BS21(R)C7 and BS22(R)C7 also have a high proportion of non-BSSS in their background (46 and 55%, respectively) (Hallauer et al., 2000), and in the case of BS21(R)C7, a high GCA value, which explains its excellent performance in combinations.
The highest midparent heterosis of crosses involving CGL(S1–S2)C5 were those with BS21(R)C7, LEAMING(S)C4, and BS22(R)C7. As mentioned before, the high heterosis observed in the cross with BS21(R)C7 is explained by its high GCA value. The combination of LSC [represented by CGL(S1–S2)C5] and Leaming [LEAMING(S)C4] has already been reported as being an alternative heterotic pattern (Kauffman et al., 1982).
The cross BS21(R)C7 x BS22(R)C7 also represents an alternative heterotic pattern. The population hybrid had both a high grain yield (6.75 Mg ha–1, ranking first among all crosses) and the third highest midparent heterosis value (1.95 Mg ha–1). Both parents have a few of the same RYD inbred lines in their backgrounds (A619 and W153R). The high midparent heterosis observed is the result of the reciprocal recurrent selection scheme utilized for their improvement.
The heterotic patterns Leaming x RYD and Leaming x LSC have been previously reported as alternatives to the most exploited RYD x LSC pattern (Kauffman et al., 1982). In this study, the Leaming heterotic group [represented by LEAMING(S)C4] combined well with BS21(R)C7 or BS22(R)C7 (midparent heterosis values of 1.95 and 1.96 Mg ha–1, respectively). The pattern with BS22(R)C7 is especially interesting because their population hybrid had the second highest grain yield among all crosses, and was not different from the best cross or the third highest yielding check.
The populations BS5, NDSG(M)C15, NDSM(M)C7, and NDSCD(M)C10 were overall the poorest combiners. The best hybrid combinations included BS21(R)C7. NDSG(M)C15 combined well with BS21(R)C7 despite both having Minnesota 13 in their backgrounds (100 and 18%, respectively). This population hybrid had the fourth highest midparent heterosis value (1.93 Mg ha–1).
Grain Moisture at Harvest
Error mean squares were found to be homogeneous among the four North Dakota locations as well as among the three Iowa locations, but the seven locations could not be grouped because of heterogeneity observed between the two groups. Therefore, data were combined in two groups, one having the four North Dakota locations and the second the three Iowa locations. Interactions among locations and genotypes were not found in either group (Table 1).
For the North Dakota locations group, differences existed for Genotypes (P 0.05), as well as for all its orthogonal components, except SCA and Checks. In the case of the Iowa group, differences were found for Genotypes, Entries, Crosses, GCA, and Entries vs. Checks (P 0.05).
Differences in grain moisture among the populations per se were large, especially when considering North Dakota locations. The populations developed in North Dakota and Canada had lower values of grain moisture at harvest when compared with the four populations from Iowa (Table 4). It was expected because the populations developed in Iowa mature later than the populations from either North Dakota or Canada (Appendix 1). Differences among means of crosses having one population in common were less important. The average grain moisture of populations ranked higher than the mean of population crosses. The mean of checks was lower than the mean of both populations and population crosses (P 0.05).
|
A correlation between grain moisture content of parents per se and their GCA values was found for both groups of locations (r = 0.97, and r = 0.95, for North Dakota and Iowa, respectively). Thus, for this group of genotypes, selection of the best populations as breeding material for earliness could be done using either their performance per se or their GCA values. Similar results were reported by Nevado-Burgos (1988) when studying the natural stratification of early, medium, and late flowering plants in open-pollinated populations using a diallel-mating design. He found a correlation (P < 0.01; r = 0.96) between the GCA effects and maturity of the populations.
The GCA sum of squares was greater than the SCA and Parents vs. Crosses (average heterosis) sum of squares (92949, 26228, and 6507, respectively), indicating that grain moisture variation among crosses was mainly due to additive genetic effects rather than nonadditive genetic effects.
The cross LEAMING(S)C4 x BS22(R)C7 had the lowest midparent heterosis value for grain moisture (Table 4). Also, the same cross was among the five highest crosses for grain yield midparent heterosis (P 0.01). Therefore, the two populations form different and complementary heterotic groups.
Midparent heterosis values were negative in most cases (Table 4). Average midparent heterosis was –14.2 g kg–1 (–6.1% of the mean of parents) in the North Dakota group of locations.
Root and Stalk Lodging
Root and stalk lodging are considered important agronomic characters in selection for improved cultivars because reduced root and stalk lodging should improve their agronomic performance (Fehr, 1991). Therefore, negative midparent heterosis, GCA, and SCA values are desirable for both traits.
Differences were observed among populations for root lodging, but not for stalk lodging (Table 1). For root lodging, populations ranged from 9.6% for BS21(R)C7 to 34.8% for NDSG(M)C15. Differences were found for both traits among crosses and for GCA (Table 2), but no differences were found for SCA effects or average heterosis. Hence, the relative contribution of additive genetic effects was greater than the corresponding contribution of nonadditive effects for both traits.
Midparent heterosis values for root lodging ranged from –8.2% for BS5 x NDSG(M)C15 to 5.9% for BS22(R)C7 x CGSS(S1–S2)C5 (Table 5). Only 15 of the 45 crosses had desirable (negative) midparent heterosis values. The population cross with the top midparent heterosis value, and the only one different from zero (P 0.05) was BS5 x NDSG(M)C15. Five population crosses and three populations had means lower than the mean of the checks. The crosses were BS21(R)C7 x CGSS(S1–S2)C5, BS21(R)C7 x NDSM(M)C7, BS21(R)C7 x LEAMING(S)C4, NDSAB(MER)C12 x BS21(R)C7, and BS21(R)C7 x BS22(R)C7; and the populations were BS21(R)C7, CGSS(S1–S2)C5, and BS22(R)C7. None of them differed from any of the checks.
|
0.05). Six population crosses, viz., BS21(R)C7 x BS22(R)C7, BS22(R)C7 x CGL(S1–S2)C5, LEAMING(S)C4 x CGSS(S1–S2)C5, NDSAB(MER)C12 x CGSS(S1–S2)C5, LEAMING(S)C4 x BS22(R)C7, and BS21(R)C7 x CGL(S1–S2)C5 had stalk lodging means below the mean of the checks. None of these genotypes differed from any of the checks (P 0.05), demonstrating that intrapopulation selection has been effective in improving this trait. It is important to note that two of the commercial single-cross hybrids were genetically modified for resistance to European corn borer [Ostrinia nubilalis (Hübner)].
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIX 1REFERENCES |
|---|
Because of its high grain yield and midparent heterosis values observed for both grain yield and grain moisture, the cross BS22(R)C7 x LEAMING(S)C4 deserves attention. To obtain inbred lines better adapted to the region, particular emphasis should be placed on the improvement of both populations for earliness.
A strong association was found among the performance of populations per se and their GCA for grain moisture at harvest. This indicates that selection of populations for this trait would be equally effective using either the performance of the populations per se or their GCA value. The population CGSS(S1–S2)C5 had the highest frequency of positive alleles for earliness. This improved population would be the first choice if selection for earliness were the primary objective. The performance of this population per se and in crosses was also excellent for root and stalk lodging.
Five population crosses and three populations [BS21(R)C7, CGSS(S1–S2)C5, and BS22(R)C7] did not differ from the checks for root lodging. In the case of stalk lodging, none of the population crosses had midparent heterosis values different from zero. However, several population crosses had lower stalk lodging means than the mean of the checks. Therefore, certain population crosses have shown evidence of outstanding agronomic performance in addition to grain yield performance.
Alternative heterotic patterns were found for breeding programs in the northern Corn Belt. We suggest BS21(R)C7 and CGSS(S1–S2)C5 as good choices for opposite heterotic groups based on the data evaluated.
|
APPENDIX 1 |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIX 1REFERENCES |
|---|
|
NDSG(M)C15 was developed from NDSG by 15 cycles of mass selection for yield and standability at low plant densities. NDSG, an unreleased experimental synthetic, was derived from the open-pollinated variety Minnesota 13 by two cycles of mass selection for kernel size and several cycles for improved agronomic appearance. The proportion of individuals selected was one percent on a sample size of 3000 individuals, and selection was performed at a plant density of 20000 plants ha–1 (Cross, 1984).
NDSM(M)C7 was obtained after seven cycles of mass selection on NDSM. NDSM originated by intercrossing 13 yellow dent inbred lines of approximately AES100 to AES300 maturity that were chosen for good general combining ability for resistance to stalk breakage and high grain yield. The proportion of individuals selected was one percent on a sample size of 3000 individuals, and selection was performed at a plant density of 20000 plants ha–1 (Cross and Wanner, 1991).
BS5 is an early synthetic maize population developed in central Iowa (Iowa State University, 2001) by intercrossing 23 dent and flint elite inbred lines. The approximate composition is as follows: 36% from Minnesota 13, 24% RYD, 26% European Flint, and 15% from other sources. BS5 provides an early population for recurrent selection and line development. BS5 has a maturity rating of AES 400-500.
BS21(R)C7 is a genetically broad-based population developed after seven cycles of reciprocal recurrent selection on BS21 using BS22 as tester, primarily for improved grain yield and root and stalk strength. BS21 is a synthetic developed by the cross of BS5 and BS20 maize populations. BS20 is a maize synthetic composed of 12 yellow dent inbred lines. The background of the lines belong 85% to the RYD group. Therefore, the composition of BS21 is 54% RYD, 18% Minnesota 13, 13% European Flint, and 15% from either other or unknown sources. The maturity classification of BS21 is AES500-600 (Hallauer et al., 2000). Two cycles of recombination and mild selection for earliness were completed on BS21(R)C7 at NDSU in 2000.
BS22(R)C7 is a genetically broad-based synthetic cultivar developed after seven cycles of reciprocal recurrent selection on BS22 using BS21 as tester, primarily for improved grain yield and root and stalk strength. Two cycles of recombination and mild selection for earliness were completed at NDSU before evaluation. BS22 is a synthetic population that was developed by intercrossing 16 inbred lines. The background of the inbred lines is 45% RYD, 13% LSC, 9% Minnesota 13, and 34% from either other or unknown sources. Its maturity classification is AES500-600 (Hallauer et al., 2000). Both BS21(R)C7 and BS22(R)C7 are improved sources of corn germplasm for use in areas of higher latitudes or in areas desiring earlier maturity (Iowa State University, 2001).
CGL(S1–S2)C5 is a maize population synthesized from 26 elite inbred lines and one single-cross hybrid from the ‘Lancaster’ heterotic group. Its maturity rating is AES 200. It was improved by five cycles of S1–S2 recurrent selection at the University of Guelph, Canada (University of Guelph, 2002).
CGSS(S1–S2)C5 has been synthesized using 18 inbred lines from the BSSS heterotic group at the University of Guelph, in Canada, and improved by five cycles of S1–S2 recurrent selection (University of Guelph, 2002). Its maturity rating is AES200.
LEAMING(S)C4 was developed by three cycles of S1–S2 recurrent selection and one cycle of half-sib selection with A632 as tester from the old open-pollinated landrace Leaming at Iowa State University (Carena and Hallauer, 2001). It has a yellow dent endosperm, good yield potential, and feeding value. Leaming was a very popular cultivar in the U.S. Corn Belt until RYD replaced it around 1920. Two cycles of recombination and mild selection for earliness were completed at NDSU before evaluation. LEAMING(S)C4 has a maturity rating of AES500.
|
ACKNOWLEDGMENTS |
|---|
Received for publication May 11, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSAPPENDIX 1REFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  X. M. Fan, H. M. Chen, J. Tan, C. X. Xu, Y. M. Zhang, Y. X. Huang, and M. S. Kang A New Maize Heterotic Pattern between Temperate and Tropical Germplasms Agron. J., June 16, 2008; 100(4): 917 - 923. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Carena, C. Eno, and D.D. Wanner Registration of NDBS11(FR-M)C3, NDBS1011, and NDBSK(HI-M)C3 Maize Germplasm Journal of Plant Registrations, May 1, 2008; 2(2): 132 - 136. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
