| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dep. of Agronomy and Plant Genetics, Univ. of Minnesota, 411 Borlaug Hall, 1991 Buford Circle, St. Paul, MN 55108
* Corresponding author (berna022{at}umn.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
lµ) not found in LH 227 x LH 295. Cateto, AS5, ASA(FNI)C5, and ASA(FI)C5 had the highest llµ values for yield. Cluster analysis of SSR data grouped these four populations in one cluster and the Minnesota inbreds in a different cluster. We concluded that Cateto, a synthetic from South America adapted to the northern Corn Belt, is a suitable population for improving LH 227 x LH 295.
Abbreviations: BSSS, Iowa Stiff Stalk Synthetic • PCA, principal components analysis • SSR, simple sequence repeat
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Previous studies have considered the use of exotic maize germplasm to improve U.S. maize (Wellhausen, 1965; Hallauer and Sears, 1972; Michelini and Hallauer, 1993). Most of these studies have been conducted in the central and southern U.S. Corn Belt and have focused on later-maturing inbreds (approximately 120 d to relative maturity; Goodman, 1985; Goodman et al., 2000), because the exotic populations were of tropical origin and, consequently, were more adapted to warmer climates. Geadelmann (1986) found that if selection is done for adaptation to northern climates, exotic germplasm could be useful for improving northern Corn Belt maize. In addition to exotic populations, other materials adapted to northern climates, such as open-pollinated cultivars, historical landraces, and synthetic populations, may be useful in improving maize in the northern Corn Belt.
Because little of the research concerning exotic populations has been done with germplasm from the northern Corn Belt, this study would be valuable to private and public breeders alike for future improvement of northern maize inbreds. Our objectives were to (i) identify the best populations among a diverse group of 17 early populations for improving an elite single cross, LH 227 x LH 295, and (ii) determine the genetic diversity among the populations and Minnesota inbreds from BSSS and non-BSSS heterotic groups.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
The LH 227 x Py crosses, the LH 295 x Py crosses, and LH 227 x LH 295 were evaluated in 2002 in Lamberton, Rosemount, and Waseca, MN. The planting dates were 3 May 2002 at Lamberton and Rosemount, and 30 April 2002 at Waseca. A randomized complete block design with two replications was used at each location. LH 227 and LH 295 were evaluated in a separate but adjacent inbred trial with six replications at the same three locations. At Waseca and Lamberton, the entries were grown in two-row plots, each row 6.7 m long and spaced 0.76 m apart. The plant population density was 58687 plants ha–1. At Rosemount, the entries were grown in two-row plots, each row 6.1 m long and spaced 0.76 m apart. The plant population density was 64555 plants ha–1. We were concerned that high plant population densities would mask differences among the populations, particularly the landraces. We therefore chose reduced plant population densities at all three locations. We measured plant height (cm) on a plot basis as the distance from the soil surface to the tip of the tassel. Ear height (cm) was measured on a plot basis as the distance from the soil surface to the bottom of the main ear. We measured root lodging (%) by recording the number of plants leaning at an angle greater than 45° degrees from the vertical. Stalk lodging (%) was measured by recording the number of plants with stalks broken below the main ear. Data for root and stalk lodging were recorded at two locations, Waseca and Rosemount. Yield (Mg ha–1, adjusted to 155 g H2O kg–1) and grain moisture (g H2O kg–1) data were recorded during mechanical harvest.
Identifying the Best Populations for Inbred Improvement
Dudley (1987) developed a model, validated by Hogan and Dudley (1991), to identify populations with favorable dominant alleles absent in an elite single cross. There are four classes of loci, i, j, k, and l, for an I1 (LH 227) x I2 (LH 295) single cross (Table 2). These four classes represent possible combinations of favorable alleles from each parent. In the population, the average frequency of favorable alleles in each class was unknown, but were denoted by i, j, k, and l. We assumed the inbreds were homozygous, the populations were in random-mating equilibrium, dominance was complete, epistasis was absent, and the genotypic values, µ, were constant for all loci (Dudley, 1984). Populations with the highest l have the highest frequency of favorable alleles absent in I1 x I2. The l cannot be directly estimated, but Dudley (1987) showed that a related statistic, llµ, can be estimated from phenotypic data on I1 x Py and I2 x Py, I1 x I2, I1, and I2. The number of class l loci, which is denoted also by l, is constant for a given I1 x I2 cross. The population with the highest llµ will therefore have the highest frequency of favorable dominant alleles at class l loci, l.
|
lµ. The average frequency of recessive alleles in the population was denoted by 
j at class j loci and k at class k loci. To estimate llµ with a minimum amount of bias, the upper and lower limits of j and k were estimated from phenotypic data as outlined by Dudley (1987). The upper and lower limits of j and k then determined which of four equations, given by Dudley (1987), was used to calculate llµ. The llµ values were calculated only for yield, for which favorable alleles are dominant.
Analysis of variance, combined across locations, was performed in SAS (SAS Institute, 2001). Replications and locations were considered as random effects, whereas crosses were considered as fixed effects. The standard errors of the entry means for each trait were calculated. The variance of the llµ estimate was calculated as the variance of a linear function of entry means, with the mean squares due to location-entry interaction being used as the error variance. Significance (P = 0.05) of llµ was determined by a z test.
We then predicted the outcome of introgressing the germplasm from the population into the elite single cross. Improvement of LH 227 x LH 295 would be accomplished by either forming an F2 population from LH 227 x Py and using LH 295 as the tester [(LH 227 x Py)F2 x LH 295], or forming an F2 population from LH 295 x Py and using LH 227 as the tester [(LH 295 x Py)F2 x LH 227]. For a given population, if the yield of LH 227 x Py was greater than the yield of LH 295 x Py, the inbred to be improved (crossed to Py) was LH 295. If the yield of LH 227 x Py was less than the yield of LH 295 x Py, the inbred to be improved (crossed to Py) was LH 227. If epistasis is absent, the mean of an F2 x tester cross [e.g., (LH 227 x Py)F2 x LH 295] is equal to the mean of the corresponding three-way cross [(LH 227 x Py) x LH 295] (Fehr, 1991, p. 433–435). For convenience, both (LH 227 x Py)F2 x LH 295 and (LH 295 x Py)F2 x LH 227 are referred to as a three-way cross. The trait means of LH 227 x Py, LH 295 x Py, and LH 227 x LH 295 were used to calculate the predicted outcome of the three-way cross for each population. If the population was to be crossed to LH 227, the predicted mean of the three-way cross for each trait was calculated as 1/2[(LH 227 x LH 295) + (LH 295 x Py)]. If the population was to be crossed to LH 295, the predicted mean of the three-way cross for each trait was then calculated as 1/2[(LH 227 x LH 295) + (LH 227 x Py)].
Simple Sequence Repeat Analysis
Approximately 50 plants of each population and of A619, A632, A679, and A682 were grown in a growth chamber. The leaf material was harvested and bulked for each population and inbred. The bulk tissue samples were freeze-dried, ground into powder, and stored. The DNA was extracted from each bulk population sample by the CTAB method (Saghai-Maroof et al., 1984). Because of poor DNA recovery, Golden Glow was eliminated from marker analysis.
Primer sequences for 82 random SSR markers, chosen to span the genome, were synthesized by Integrated DNA Technologies (Coralville, IA) according to sequences published in MaizeDB (www.maizedb.org; verified 20 Feb. 2004). A MJ Tetrad thermocycler was used to perform the PCR reactions. The PCR reaction mixture consisted of 20 ng µL–1 working stock DNA, 1.5 µL Promega Taq, dilute SSR marker, and a PCR mix consisting of 100 mM dNTPs, 25 mM MgCl2 buffer, 10 x PCR buffer (Promega), and ddH2O. The total reaction volume was 20µL. The samples were heated to 94°C for 2 min to completely denature the DNA. The samples were maintained at 94°C for 30 s, 57°C for 1 min, and 72°C for 1 min. This cycle was repeated 35 times. The samples were then incubated at 72°C for 10 min.
The samples were loaded onto a 3% agarose gel that contained trace amounts of ethidium bromide for visualization of the bands. The samples were electrophoresed for approximately 4.5 h at 120 W. Visualization of the bands was done with Quantity One, Version 4.1.1 (BioRad, 1998). The resulting bands were counted and tallied for each population and inbred.
The Nei-Li distances (Nei and Li, 1979) among the populations and inbreds were calculated as Dij = 1 – (2Nij/Tij), where Nij was the number of bands that population or inbred i and j had in common, and Tij was the sum of the number of bands in i and j. The Nei-Li distances were used to perform cluster analysis and principal components analysis (PCA) in Multi-Variate Statistical Package, version 3.13d (Kovach Computing Services, 2002).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
|
|
lµ (0.82) for yield (Table 4). The three populations with the next highest llµ values were AS5 (0.73), ASA(FNI)C5 (0.69), and ASA(FI)C5 (0.61). The populations with the lowest llµ were KASS (–1.43), AS6 (–1.27), CG Syn B (–1.19), and Longfellow (–0.63). These negative llµ values for yield were most likely because of failure of the assumption of complete dominance (Dudley, 1987).
The llµ for yield was highly correlated (r = 0.94, P < 0.05) with the predicted yield of the three-way cross, [(LH 227 x Py)F2 x LH 295] or [(LH 292 x Py)F2 x LH 227]. The two populations with highest llµ, Cateto and ASA(FI)C5, also had the highest predicted yield in the three-way cross. Cartea et al. (1996) stated that, because of high correlation, predicted yield of the three-way cross and llµ for yield were equally effective for identifying populations for improvement of the elite single cross. But to accurately rank the populations in terms of the frequency of favorable alleles, Hogan and Dudley (1991) stated the llµ for each population should be used.
The elite single cross affects the ranking, based on llµ, of the populations. In a previous study by Dudley (1988), Mexican Dent and Caribbean Flint had higher llµ for yield than Cateto. But in our study, the yield llµ for Cateto was greater than llµ for Mexican Dent (0.03) and Caribbean Flint (–0.38). Dudley's study considered improving B73 x Mo17, which has a relative maturity of 117 d (MBS, 2002). In contrast, LH 227 x LH 295 has a relative maturity of 99 d (MBS, 2002). The differences in the genetic background and maturity of B73 x Mo17 and LH 227 x LH 295 led to the different rankings of populations based on llµ. Therefore, the effects of a population on one single cross will not predict the effects of the population on a different single cross.
Genetic Diversity among Populations and Inbreds
Of the 82 SSR sequences screened, 55 showed polymorphisms among the populations and inbreds and resulted in 232 bands. The average number of bands per marker was 4.3. The genetic distances among all of the populations and inbreds ranged from 0.15 to 0.70 with an average of 0.43. The genetic distances between the populations and inbreds ranged from 0.41 to 0.70, with an average of 0.52. When comparing the four populations with the highest llµ for yield [Cateto, AS5, ASA(FNI)C5, and ASA(FI)C5] to inbreds A619, A632, A679, and A682, the genetic distances ranged from 0.41 to 0.59, with an average of 0.51. This result indicated that, relative to the BSSS and non-BSSS inbreds, the diversity of the four selected populations (0.51) was similar to the diversity of all of the populations (0.52).
Two distinct groups, denoted by Group I and Group II, were formed by cluster analysis of Nei-Li distances (Fig. 1) . Group I included the inbreds A619, A632, A679, and A682, and the populations KASS, KALD, and V360. The detailed pedigrees of KASS and KALD are unknown. Because they were grouped with A619, A632, A679, and A682, we believe they share similar genetic background to the inbreds. When examining the specific groupings found within this cluster, the inbreds with similar backgrounds did not immediately cluster together. A619 (non-BSSS) clustered first with A632 (BSSS) rather than with A682 (non-BSSS). The clustering of V360, an early Spanish flint population, (Lee, 2002) with A679 (BSSS) cannot be explained from the pedigree data. The clustering of A619 with A632 and of A679 with V360 may be explained, however, by their similar maturities. Group II consisted solely of populations. Within Group II, ASA(FI)C5 and ASA(FNI)C5 clustered closely together because they were developed from the same population base (Johnson and Geadelmann, 1989). ASA(FNI)C5 and ASA(FI)C5 then clustered with AS5 in a subcluster of Group II. All three populations were developed at the University of Minnesota by crossing early maturing inbreds and populations (Geadelmann, 1986; Johnson and Geadelmann, 1989). Cateto, which comprised six proprietary Argentine inbreds, one with 12.5% US Corn Belt germplasm (Gerrish, 1983), was in a different subcluster in Group II. This result indicated that there was enough of a genetic distance between Cateto and the Minnesota populations to separate them within Group II. Cateto, AS5, ASA(FNI)C5 and ASA(FI)C5 had large genetic distances from the inbreds and were grouped separately from the inbreds.
|
lµ can therefore be separated into two groups.
|
lµ for yield. We expected populations that were elite and adapted to perform best. KASS and KALD, which were both developed from the cross between two commercial inbreds, had llµ for yield that were not significantly different from zero (P = 0.05). Earlier maturity, as evidenced by lower grain moisture at harvest, is generally associated with lower yield. KASS and KALD had the lowest means for grain moisture (Table 4), and this result may account for the low llµ for these two populations.
The genetic distances and the results for llµ suggested that suitable germplasm for elite inbred improvement exists within little-used publicly available populations that were adapted to the northern Corn Belt. In conclusion, we identified four populations [Cateto, AS5, ASA(FNI)C5, and ASA(FI)C5] as donors of favorable alleles for improving the yield of the elite single cross LH 227 x LH 295. These four populations had high llµ for yield and were genetically distant from northern maize inbreds similar to LH 227 and LH 295. These populations may also improve other inbred BSSS x non-BSSS crosses, related to LH 227 x LH 295, that are widely grown in the northern Corn Belt.
Received for publication May 27, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Smith Intellectual Property Protection for Plant Varieties in the 21st Century Crop Sci., July 1, 2008; 48(4): 1277 - 1290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. M. Rodriguez, R. A. Malvar, A. Butron, A. Ordas, and P. Revilla Maize Populations as Sources of Favorable Alleles to Improve Cold-Tolerant Hybrids Crop Sci., September 1, 2007; 47(5): 1779 - 1786. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. J. Kutka and M. E. Smith How Many Parents Give the Highest Yield in Predicted Synthetic and Composite Populations of Maize? Crop Sci., September 1, 2007; 47(5): 1905 - 1913. [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 | |||
lµ, testcross yield, and predicted trait means for different three-way crosses in maize.
