Progress 10/01/02 to 09/30/08
Outputs
OUTPUTS: A study of two crosses between Chinese and US soybean varieties adapted to similar latitudes was completed. The varieties, Tiefeng #27 and Shennong #6 from China, and Ohio FG1 and HS97-4534 from the US, were adapted to soyfood production. The study was begun in 1999 and carried out in parallel in Ohio and in Liaoning, China. The objective was to determine whether epistasis occurred for quantitative traits. Random lines from the biparental crosses and the first backcross to each parent were evaluated for 2 or 3 years. Statistical tests involving means and variances were used to detect epistasis for grain yield, maturity, lodging resistance, seed size, plant morphological traits, and chemical composition. A retrospective analysis of 20 years of data from early-generation yield tests was conducted to determine whether incorporating information on relationships between entries would have made a difference in the selections. Information on degree of relationship is relatively easy to obtain and incorporate into models for data analysis. Data from tests of F2:4 soybean lines conducted (in most cases) at three locations with two replications per location were analyzed in three ways: (A) with no information on relationships, (B) with information as to which entries were full- or half-sibs, and (C) with coefficient of parentage calculated for each pair of entries. The selection criterion was yield, adjusted for maturity using quadratic regression. Selection of the top 10% of lines was carried out with each method. PARTICIPANTS: Dr. Xie Futi, visiting scholar, The Ohio State University and professor, Shenyang Agricultural University, Liaoning, China TARGET AUDIENCES: Soybean breeders PROJECT MODIFICATIONS: Not relevant to this project.
Impacts
Analysis of biparental and backcross lines from two China x US crosses revealed significant additive x additive epistasis for plant height, maturity, seed size, yield, internode length, number of branches, harvest index, and content of protein and oil. There was little consistency in the presence of epistasis across crosses or locations, indicating interactions of gene action with background genotype and test environment. Further, in some cases simple digenic epistasis did not account for the results, implying higher-order interactions among loci. Some of the results can be explained by interactions affecting the timing of growth stages and interactions between such loci and the environment. The results indicate that breeders who make crosses between parents adapted to Northeastern China and the Midwestern United States should be prepared for a high frequency of poorly adapted progeny. There is a need to work out the complex genetics responsible for the timing of growth stages. Without such information, it will be difficult to predict the value of any particular Chinese parent as breeding material for US breeding programs (and vice versa) without actually making the cross. Field plot testing is costly, and information that would allow breeders to use test data more efficiently would lead to greater genetic gain. In our retrospective study of F2:4 lines, the effect on selection of including information on genetic relationships is small. Of the lines selected by a one method, 66% were also selected by the other two methods. Review of subsequent performance of F4-derived lines selected from the superior F2:4 lines suggests little difference among methods A, B, and C. We conclude that with 6 replications (3 locations x 2 replications/location) the observed mean performance of each line outweighs any additional information from its relatives. Attention will now be directed to the F2:3 generation, where the amount of replication of each line is much less, and information from relatives may carry more weight.
Publications
- S. K. St. Martin, Fu-ti Xie, Hui-jun Zhang, Wei Zhang, and Xian-jun Song. 2009. Epistasis for Quantitative Traits in Crosses between Soybean Lines from China and the United States. Crop Sci. 49:(in press, Jan.-Feb. issue).
|
Progress 01/01/07 to 12/31/07
Outputs
OUTPUTS: Field experiments were continued in 2007. A third and final year of data was obtained for a comparison of lines derived from a two-parent cross with lines derived from reciprocal backcrossing (BC1). Differences between these two populations reflect the extent to which linkage of favorable genes is responsible for a portion of the agronomic superiority of soybean genotypes. In 2007 we conducted generation advance of the final backcross populations for transfer of purple flower color from many sources to the common parent HS93-4118. By 2009 these populations should be ready for testing. Purple flower color is used as a simple, presumably neutral marker. Phenotypic differences between purple isolines and HS93-4118 will be due to putative linkages between the flower-color locus and genes for quantitative traits. We also advanced the population resulting from the final backcross of to Clark. In this population we hope to identify lines isogenic to Clark that have significantly
higher yield than the recurrent parent. Such lines, if found, will provide insight into genetic and physiological mechanisms of yield improvement.
PARTICIPANTS: Staff Steve St. Martin, PI Xie Futi, postdoctoral researcher Scott McIntyre, research associate Marcia Feller, reserch associate Andy Spring, research assistant The Ohio BioProducts Innovation Center provided funds for staff and new equipment.
TARGET AUDIENCES: The target audience is soybean breeders in the private and public sectors.
Impacts
All three experiments are aimed at enhancing our understanding of the genetics of quantitative traits in soybean. New knowledge will be disseminated through publications and will help breeders more efficiently improve soybean for yield and other economic traits. We have completed a preliminary analysis of the data from the comparison of biparental progenies with reciprocal backcrosses. There are significant differences in lodging resistance and maturity between groups of progeny. Specifically, the reciprocal backcross progeny are later and lodged more than the biparental progeny. There was no difference in yield between the progeny groups. The differences in lodging and maturity perhaps reflect interactions between sets of maturity genes that perhaps differ greatly between the parents. The implication for soybean breeders is that breaking up parental linkages does not have a negative impact on seed yield. This means that breeding procedures that employ extensive
recombination among parents should be no less effective than procedures that better conserve parental genotypes.
Publications
- No publications reported this period
|
Progress 01/01/06 to 12/31/06
Outputs
Soybean has been produced in China for more than 3000 years and in the USA for more than 200 years. Current soybean production in the Northern USA has its genetic base primarily in introductions from China . Beginning from similar genetic bases, breeders in China and North America have selected for adaptation to their respective environments, resulting in increased yield and improvements in other traits. Future yield gains may depend on an understanding of the changes made to soybean cultivars by breeding and selection. It would be valuable to know whether selection in one environment (China or USA) has led to similar improvements in the other environment. Knowledge of how morphological and physiological traits have responded to selection may also provide insight as to the nature of the genetic changes. Our objective was to compare, in both Liaoning province (China) and Ohio (USA) environments, a set of older cultivars from Liaoning with their modern counterparts
derived from breeding programs in Liaoning and Ohio. The experiment included four older cultivars grown in Liaoning in the early 20th century, and eight modern cultivars, four from Liaoning and four from Ohio. In Ohio, modern Liaoning cultivars yielded 22% and modern Ohio cultivars 78% greater than the old Liaoning cultivars. In Liaoning, the modern cultivars from both locations yielded about 50% greater than the old Liaoning cultivars. Modern Chinese and Ohio cultivars were shorter and more lodging resistant than old Liaoning cultivars in Ohio. In Liaoning, the Ohio cultivars were more lodging resistant than either set of Chinese cultivars. In Ohio, modern Ohio cultivars were lower in protein than old or modern Liaoning cultivars. In Liaoning, there were no significant differences in protein content among groups of cultivars. Assessment of photosynthetic rate of leaves during grain fill, performed only in Liaoning, showed that modern Ohio and Liaoning cultivars exceeded old Liaoning
cultivars. The results show that a significant portion of the yield gains made by breeders in each locations translates to the other location. Breeders have, however, applied selection pressure for different traits, resulting in phenotypic, and presumably also genotypic differences between modern Chinese and USA varieties. Exchange of cultivars should benefit breeders in both countries.
Impacts
In this experiment we assessed the genetic gain for yield and other transits that have been achieved by soybean breeders operating independently, but starting from similar genetic bases, in Liaoning province, China, and Ohio, USA. Breeders in both places have made progress in improving yield, and it appears that a significant portion of yield gains made in one location translate to the other. Changes in protein content and lodging resistance were equivalent between the two countries, reflecting different goals of the breeding programs. The results suggest that exchange of breeding material between China and USA is mutually beneficial.
Publications
- No publications reported this period
|
Progress 01/01/05 to 12/31/05
Outputs
The importance of epistasis, defined as the interaction between non-allelic genes, has been an open question in soybean breeding since the 1950s. If selection of superior homozygous cultivars operates on epistatic as well as additive gene action, then we can expect the progeny of the cross of two superior cultivars to be inferior to that of the parental mean, to the degree that favorable gene combinations present in the parents are broken up through Mendelian recombination. If epistasis is important then breeding methods that conserve gene combinations, such as partial backcrossing, might be preferable. Two crosses (Tiefeng#27 x HS97-4534 and Ohio FG1 x SN91-44) were made between food-type breeding lines from Liaoning Province, China and from Ohio. The Chinese lines were provided by Dr. Xie Futi, Shenyang Agricultural University. Random inbred lines were developed from these populations by single-seed descent. Also, random inbred lines were developed from the
reciprocal backcross generations, i.e. equal numbers of Ax(AxB) and Bx(AxB). In expectation, lines from the biparental (2X) and reciprocal backcross (RB) procedures have the same allele frequencies, but the RB lines have more of the parental gene combinations intact. At Plain City, Ohio, in 2005, 48 2X lines and 48 RB lines from each parental combination were tested, along with parents and checks, in a reps-within-sets design with two replications. Maturity, height, lodging, 100-seed weight, and yield were determined. Results are shown below. No mean separation is presented due to lack of a proper (genotype x environment-based) error term. There are no clear trends in the data at this point. Tiefeng#27 x HS97-4534 date mature lodging height (cm)g/100 seeds yield (kg/ha) [Tiefeng#27 24 1.2 74 22.0 3244] [HS97-4534 20 1.9 90 25.6 3234] [parental mean 22 1.6 82 23.8 3239] [2X lines 26 2.2 96 22.9 2955] [RB lines 34 2.5 101 23.6 2958] Ohio FG1 x SN91-44 date mature lodging height (cm)
g/100 seeds yield (kg/ha) [Ohio FG1 22 1.3 78 24.3 3420] [SN91-44 24 1.4 79 21.7 2775] [parental mean 23 1.4 78 23.0 3097] [2X lines 22 1.4 80 24.1 3227] [RB lines 23 1.5 79 23.1 3032]. The study will be continued in Ohio and is also being conducted in parallel in Liaoning Province by Dr. Xie.
Impacts
Does selection of superior cultivars lead to the accumulation of superior genes or superior gene combinations? If individual genes, operating independently, are responsible for superior performance, the breeders can use methods that freely break up gene combinations in existing cultivars. But if high performance requires synergistic or other types of interactions between genes, then breeders might be better advised to use methods that conserve existing favorable gene combinations. The current study, when complete, will contribute to an answer to this broad question, which breeders have asked for more than 50 years. The study compares breeding lines derived by two procedures from crosses of Ohio and Northeast Chinese food-type parental lines. In one procedure, lines were derived from the simple cross of the two parents, permitting recombination of genes. The second procedure involved reciprocal backcrosses to each parent, thus preserving more of the parental gene
combinations. Differences between the mean performance of the two groups of breeding lines will suggest whether gene combinations are important. The study will also help to determine whether food-type soybean cultivars from Northeast China have the potential to contribute to the development of food-type cultivars for Ohio.
Publications
- No publications reported this period
|
Progress 01/01/04 to 12/31/04
Outputs
An effort to backcross genes for improved yield into the old cultivar Clark has been underway since 1994. A sample of 30 BC2 lines from Clark(3) x HS89-5467 was tested at two locations in 2003, with two replications per location. Modern check varieties and double entries of the parents were included. The cultivar Clark (2381 kg/ha) outyielded HS89-5467 (2313 kg/ha), probably due to better partial resistance to Phytophthora, which was a yield-limiting factor at one of the locations. The mean yield of the 30 BC2 lines was 2569 kg/ha. The parents and BC2 mean were similar in days to maturity, and the BC2 lines had an average lodging score similar to that of HS89-5467 and better than that of Clark. None of the BC2 lines yielded better than the high-yielding, disease-resistant check, Kottman (3363 kg/ha). Four BC2 lines were selected and used to make the BC3 to Clark in the summer of 2004. The mean yield of the four selections was 2972 kg/ha, which provided a selection
differential of 403 kg/ha.
Impacts
The BC3 lines resulting from the crosses made in 2004 will be useful as a tool to study the genetic and physiological basis of soybean yield. Use of near-isogenic (backcross) lines eliminates much of the background genetic variation and facilitates molecular mapping of the genes responsible for the yield increase. The developing discipline of genomics can also make use of these lines to identify the genes that differ between Clark and the BC3 lines, some of which are responsible for the differences in yield.
Publications
- No publications reported this period
|
Progress 01/01/03 to 12/31/03
Outputs
Near-isogenic lines have proven to be a valuable tool for the study of soybean genetics and can be used in conjunction with modern tools such as marker mapping and proteomics to increase our understanding of the genetic mechanisms underlying economic traits. In the late 1980s a backcross program was initiated in this project to incorporate increased yield into the older soybean variety Clark, released in 1953. Clark was chosen because many other isogenic lines have been developed in a Clark background. The donor parent was HS89-5467, a high-yielding experimental line similar to Clark in maturity and possessing no major genes for disease resistance. The objective was to transfer genes conferring higher yield through improved physiological processes, not improved pest resistance. The procedure consists of two years of yield tests of BCF2-derived lines, followed by a backcross to Clark and a generation in which BC hybrid plants were intermated. Each cycle requires five
years. The second-year test of BC2 lines was conducted in 2003 at two locations, Lakeview and South Charleston, Ohio. Plots were 460 cm long and 120 cm wide, and the design was a simple 6 x 6 lattice. At South Charleston, the mean yield of the 30 BC2 lines was 2910 kg/ha, not significantly different from that of HS89-5467 (2930 kg/ha) and greater than that of Clark (2480 kg/ha). At Lakeview, flooding and phytophthora rot affected yields, and HS89-5467 was more severely damaged than Clark. The mean yield of BC2 lines was 2220 kg/ha, compared with 1670 kg/ha for HS89-5467 and 2280 kg/ha for Clark. There was significant variation for yield among BC2 lines. BC2 lines were similar to both parents in maturity and showed improvement over Clark in lodging resistance. Four superior BC2 lines will be selected for further backcrossing to Clark. The eventual product of this research will be lines similar to Clark but with alleles for improved yield from HS89-5467.
Impacts
Use of molecular markers and proteomics on the lines developed in this study will provide information on the genetic mechanism and genomic location of genes that increase soybean yield. This will increase our understanding of the genetic basis of productivity and facilitate further advances in soybean breeding..
Publications
- No publications reported this period
|
Progress 01/01/02 to 12/31/02
Outputs
North American soybean breeders have long been concerned that the narrowness of the genetic base might limit genetic progress. Published reports that measured genetic gain in North American soybean breeding were surveyed to assess the possibility that breeders will soon reach a plateau, or selection limit, beyond which further genetic gain will be difficult to obtain. Literature on selection limits in other species, including model systems, has yielded an empirical relationship between effective population size and long-term cumulative genetic gain. This relationship was applied to North American soybean breeding, using estimates of effective population size calculated by St. Martin in 1982 as well as more recent information. The results showed that the current breeding effort, taken as a whole, is not approaching a selection limit. Nevertheless, efforts to broaden the soybean germplasm will have long-term benefits in promoting genetic gain. Restrictions on germplasm
exchange may also restrict the genetic gain of individual breeding programs.
Impacts
The narrowness of the North American soybean germplasm base implies a restriction on the amount of genetic diversity available to plant breeders. In some long-term selection studies with model species, a limit (or plateau) has been reached, beyond which no further gain is possible. Calculations using data from the published literature showed that there is no immediate danger of reaching such a selection limit in North American soybean breeding, indicating that increases in yield are possible for the near future.
Publications
- St. Martin, S.K., and Geraldi, I.O. 2002. Comparison of three procedures for early generation testing of soybean. Crop Sci. 42:705-709.
- St. Martin, S. K. 2002. Selection limits: How close are we? Soyb. Genet. Newsl. (published on-line at ).
- Serunjogi, L. K.,St. Martin, S. K., and Bauer,W. D. 2002. Effects of genotype X bradyrhizobium inoculation or X fertilizer N interactions on genetic gains while selecting in soybean. Uganda J. Agric. Sci. 7:1-6.
|
Progress 01/01/01 to 12/31/01
Outputs
A long-term effort to transfer genes for increased yield from Korean soybean germplasm resulted in the development of four lines that will eventually be used in mapping studies. After earlier studies suggested that Korean plant introductions had the potential to enrich the soybean germplasm base, we initiated a program to transfer genes for increased yield into the cultivar `Zane' by backcrossing. Each cycle of backcrossing required five years, including two years of yield-testing in each cycle. Intermating of BC F1 plants was included in order to produce greater diversity of gene combinations. Four BC2 lines were selected as a result of 1998 tests and were tested at five locations in each of two years, 1999 and 2000. Mean yields of the lines were not significantly different from that of Zane, ranging from 1% less to 3% greater. The lines matured within 1 day of Zane. One line was taller than Zane and more susceptible to lodging. Marker data are being collected on the
lines to determine how they differ genetically from Zane. Yield tests were continued in 2001, and potential mapping populations are being developed by making paired crosses of the type A x Zane and A x Zane BC2 line, where A represents an unrelated parent. A cross involving a Zane BC2 line should display greater transgressive segregation that the corresponding cross with Zane if the BC2 line carries favorable genes for yield from the Korean introduction that are not present in the adapted parents. When the populations are complete, marker mapping will permit us to characterize any favorable chromosomal segments that derive from the Korean parent.
Impacts
The genetic base of current U.S. soybean varieties is generally regarded as narrow. This experiment has potential to expand the genetic base through a deliberate search for favorable genes. Characterization of any such genes will also be possible, adding to our basic knowledge of the genetics of soybean yield.
Publications
- ST. MARTIN, S. K., and FUTI, XIE. 2000. Genetic gain in early stages of a soybean breeding program. Crop Sci. 40:1559-1564.
|
Progress 01/01/00 to 12/31/00
Outputs
A field experiment to compare three different breeding procedures was completed. In the experiment, which was begun in 1986, breeding procedures based on early generation testing of F1-, F2-, and F3-derived lines were applied to two different populations. Selection was for grain yield, maturity, and lodging resistance. Application of the procedures resulted in groups of selected lines, stratified into maturity classes. Each group of selected lines was compared to a corresponding unselected control group to assess the progress produced by each method. Averaged across populations, the three methods produced similar gains for yield, approximately 4 percent of the unselected population. Selection based on F1-derived lines produced taller, more lodging-susceptible plants, particularly in the one of the populations, which had a relatively high proportion of unadapted germplasm in its background. For maximum genetic gain per year, selection based on F2-derived lines is
recommended over the other two.
Impacts
Although breeding methods based on early generation testing have been used for more than 50 years and have often been compared with other methods, this study is the only one to systematically compare different early generation testing procedures. The results will guide breeders toward more efficient methods.
Publications
- No publications reported this period
|
Progress 01/01/99 to 12/31/99
Outputs
A long-term effort to transfer genes for increased yield from Korean plant introductions was continued in 1998 and 1999. After earlier studies suggested that Korean plant introductions had the potential to enrich the soybean germplasm base, we initiated a program to transfer genes into the cultivar 'Zane' by backcrossing. Each cycle of backcrossing required five years, including two years of yield testing in each cycle. Intermating of BC F1 plants to produce greater diversity of gene combinations was a feature of the process. Ten BC2 lines were tested in 1998 at five Ohio locations. Four of these lines exceeded the mean yield of Zane by at least one standard error, while similar to Zane in maturity. Dr. Anne Dorrance tested these four lines for phytophthora resistance. Interestingly, although direct selection had never been practiced for disease response, three of the four selections proved to be resistant (Zane is susceptible). Thus, it appears that phytophthora
resistance is an important contributor to yield improvement under Ohio conditions. The four superior BC2 lines were re-tested in 1999. If the lines continue to yield better than Zane, they (particularly the one line that is not resistant to phytophthora) will be used in further experiments to elucidate the physiological nature of the difference between from Zane and to determine if the superior gene(s) captured in the line have potential to enhance yield of other adapted varieties.
Impacts
(N/A)
Publications
- No publications reported this period
|
Progress 01/01/98 to 12/31/98
Outputs
We used a procedure developed by St. Martin and McBlain (1991; Crop Sci. 31:1367-1369) to examine historical data from the OARDC soybean breeding program and obtain estimates of genetic gain for each stage. We examined the F2:3 and F2:4 testing stages of an early generation testing program that began in 1985. For the F2:3 stage, at which testing consisted of one or two replicates per entry in a single-row plot, genetic gain for yield was negative, averaging -1.37%. The negative gain was due primarily to strong simultaneous selection pressure for maturity and yield; average genetic gain for yield was positive in tests of later maturing material and negative in tests of early material. For the F2:4 stage of testing, conducted at two or three locations with two replicates per location, genetic gain averaged 3.66%. Examination of the performance of F4-derived lines from F2-derived lines suggested that more intense selection at the F2:4 stage was warranted in future tests.
There was no support for altering the selection intensity currently applied at the F2:3 stage.
Impacts
(N/A)
Publications
- WILCOX, J. R., and ST.MARTIN, S. K. 1998. Soybean genotypes resistant to Phytophthora sojae and compensation for yield losses of susceptible isolines. Plant Dis. 82:303-306.
- LEWERS, K.S., ST.MARTIN, S. K., HEDGES, B. R., and PALMER, R. G.. 1998. Effects of the Dt2 and S alleles on agronomic traits of F1 hybrid soybean. Crop Sci. 38:1137-1142. LEWERS,, K.S., ST.MARTIN, S. K., HEDGES, B. R., and PALMER, R. G. 1998. Testcross evaluation of soybean germplasm. Crop Sci. 38:1143-1149.
|
Progress 01/01/97 to 12/31/97
Outputs
Estimates of proportions of parental contribution to germplasm populations are useful to breeders. Amplified fragment-length polymorphism (AFLP) markers were applied to the parents and later cycles of a broad-based population that had undergone 2 cycles (8 years) of recurrent selection. Use of four primer pairs yielded 47 polymorphic bands. Estimates of parental contribution were consistent between primer pairs and conformed to those expected based on pedigrees. Standard errors of estimates were 3 to 5%, indicating satisfactory precision. Thus, the use of AFLP markers has promise as a relatively easy method for tracking parental contribution to populations. In this population, two cycles of selection did not eliminate the contribution of unadapted, exotic parents.
Impacts
(N/A)
Publications
- VANTOAI, T.T., PENG, J., and ST.MARTIN, S.K. 1997. Using AFLP markers to determine the contribution of parental genomes to populations. Crop Sci. 37:1370-1373.
- JENKINS, E.B., HAMMOND, R.B., ST.MARTIN, S.K., and COOPER, R.L. 1997. Effect of soil moisture and soybean growth stage on resistance to Mexican bean beetle (Coleoptera: Coccinellidae). J. Econ. Entomol.
|
Progress 01/01/96 to 12/30/96
Outputs
A method for identifying germplasm carrying valuable dominant genes that are notalready present in adapted lines was developed. The method is based on a testcross of germplasm lines to a set of adapted parent lines. A testcross statistic, which is a function of the performance of the hybrid and its adapted and germplasm-line parents, is used to identify lines with unexploited dominant genes. This method can be applied to soybean and other crops through the use of genetic male-sterility to produce hybrid seed. Three methods for production of hybrid seed were compared for efficiency in a two-year experiment conducted in Iowa, Ohio, and Ontario. A method based on close genetic linkage between male-sterility and color of the hypocotyl and flower proved to be space- and labor- efficient. Termed the "cosegregation method", this procedure can be used in testcross evaluation of germplasm.
Impacts
(N/A)
Publications
- LEWERS, K.S., ST.MARTIN, S.K., HEDGES, B.R., WIDRLECHNER, M. P., and PALMER, R.G.1996. Comparison of three methods of hybrid soybean seed production. Crop Sci. 36: (in press).
- ST.MARTIN, S.K., LEWERS, K.S., PALMER, R. G., and HEDGES, B.R. 1996. A testcross procedure for selecting exotic strains to improve pure-line cultivars in predominantly self-fertilizing species. Theor. Appl. Genet. 92:78-82.
|
Progress 01/01/95 to 12/30/95
Outputs
Early generation testing is a breeding method that, while not widely used, in theory offers some efficiencies not available with other methods. We initiated an experiment in 1986 to compare three versions of early generation testing applied to each of two soybean populations. The versions involve testing (A) S0- or F1-derived lines, (B) S1- or F2-derived lines, and (C) S2- or F3-derived lines, followed in each version by selection of relatively homozygous candidate varieties two generations later than the tested lines. In 1994, selections resulting from the versions A, B, and C were compared along with an unselected control. In population HS3, version A produced 76 kg/ha improvement in seed yield, B produced 45 kg/ha, and C produced 178 kg/ha. In the second population, derived by intermating 11 elite lines, version A produced 171 kg/ha improvement, B 53 kg/ha, and C 176 kg/ha. Selection in version A resulted in maturity 2.3 days later than the control; in version C the
response was 1.3 days; version B did not affect maturity. Mean lodging resistance was unaffected by selection. The comparison of selections will be continued in 1996.
Impacts
(N/A)
Publications
|
Progress 01/01/94 to 12/30/94
Outputs
An experiment was conducted to determine whether soybean genotypes respond differentially to inoculation with Bradyrhizobium japonicum. An experiment of this type is needed in connection with breeding soybeans for environments, such as Uganda, where B. japonicum is not native and where its introduction has been difficult. A set of 52 random lines from the broad-based population AP10 was planted with and without inoculation in B. japonicum-free fields at Jackson, OH, in 1991 and 1992. Although inoculation increased mean protein content from 36.6 percent to 40.3 percent, there was no genotype X inoculation interaction. Heritability and expected genetic gain for protein content was similar under both inoculation treatments. For yield, there were no genotype X inoculation interactions, and main effects of inoculation were not significant. Heritability for yield was 0.36 in the presence of B. japonicum, but only 0.02 in its absence. The results suggests that soybean
breeders could select effectively for high protein either with out without B. japonicum and that selections made under one inoculation treatment would retain their relative performance under the other treatment. Selection for yield should be carried out in the presence of B. japonicum.
Impacts
(N/A)
Publications
|
Progress 01/01/93 to 12/30/93
Outputs
The evaluation of experimental hybrids is being considered as a basis for identifying lines in the soybean germplasm collection that carry favorable genes not present in existing cultivars. This research is being done in collaboration with Iowa State University and Agriculture Canada (Harrow, Ontario, Research Station). There are two parts to the work: (1) Three methods for producing hybrid seed, each based on genetic male-sterility, were compared in 1993 with respect to yield and quality of seed and land and labor requirements. This comparison will be repeated in 1994. (2) The theoretical background for the design and analysis of testcross data has been developed. We propose mating a group of germplasm lines with a group of adapted lines in factorial fashion. From the results of such an experiment, statistics have been derived that indicate the potential of each germplasm line to contribute favorable genes not present in the cultivars.
Impacts
(N/A)
Publications
|
Progress 01/01/92 to 12/30/92
Outputs
Three statistical procedures were applied to regional soybean yield test data inan effort to identify a procedure that would reliably compare the yields of experimental strains from different stages of testing. The first of these methods, based on empirically-derived regression coefficients was published earlier by this project (St. Martin and McBlain, 1991). The second was selection based on yield expressed as a percentage of check cultivars common to the two stages. The third was based on best linear unbiased predictor (BLUP) procedures. The three methods achieved almost exactly the same rate (66%) of success in predicting the results of pairwise comparisons. This rate of success is too low to warrant much confidence in any of the methods. In future work, other methods will be examined. These methods may include more sophisticated BLUP procedures, Bayesian methods, and the Kalman filter.
Impacts
(N/A)
Publications
|
Progress 01/01/91 to 12/30/91
Outputs
The first five-year cycle of an experiment to transfer genes for yield by backcrossing was completed in 1991. On the basis of earlier work to examine genetic variability of germplasm accessions of different geographic origins, four accessions (PI's) from Korea were selected as donors. The variety `Zane' was used as the recurrent parent. Crosses between PI's and Zane were made in 1983, and F(subscript 2)-derived lines were tested in 1985 and 1986. Four selected F(subscript 2)-derived lines (from crosses between Zane and four different PI's) were crossed to Zane in 1987 to produce BC1 populations. In 1988, a portion of the BCI seed was used to intermate the four BC1 populations. In 1990, 42 F(subscript (superscript 2)-derived lines from each backcross population and 126 from the intermated population were evaluated and increased in unreplicated short-row plots. Selection for yield and maturity was imposed on each population, resulting in 12 lines from each backcross
population and 48 from the intermated population, which were evaluated along with Zane as a check in 1991. The 1991 test was conducted at two locations with two replications per location...The BC1 lines from the intermated population were very similar in mean yield, lodging and maturity to the lines from the non-intermated population. In comparison to Zane, the BC1 lines averaged three days later, 16% lower in yield, and similar in lodging resistance. Only one BC line had a numerically greater yield than Zane.
Impacts
(N/A)
Publications
|
Progress 01/01/90 to 12/30/90
Outputs
Development of cultivars requires testing in multiple stages, an expensive and lengthy process. Little research has been conducted on testing strategies. Our objectives were to develop a procedure to estimate genetic gains for each stage of testing and to illustrate the procedure using regional soybean data. In the procedure, a test in which a set of strains is evaluated is paired with a test in the next stage in which selections from the set are re-evaluated. Realized gains are estimated by comparison with checks common to the two tests. Regression of genetic gain on selection differential gives an estimate of the proportion of phenotypic superiority that is realized as genetic gain...Application of the procedure to regional test data revealed that genetic gains were small for all traits except grain yield. Averaged across maturity groups 00 to IV, genetic gains for yield averaged 3.0% (SE 0.4%) in the Uniform Preliminary Test and 1.7% (SE 0.4%) in the first year of
the Uniform Test. Genetic gains and regression coefficients of gain on selection differential for yield were near zero for maturity group 00, indicating that selection based on regional means was ineffective for this maturity group. The procedure may be valuable in identifying inefficient stages and in comparing genotypes in different stages of testing.
Impacts
(N/A)
Publications
|
Progress 01/01/89 to 12/30/89
Outputs
A study was designed to detect and measure residual effects of major resistance genes in soybean (Glycine max (L.)) on phytophthora root rot caused by Phytophthora megasperma f.sp. glycinea Kuan & Erwin (Pmg). Residual effects of major resistance genes are defined as rate-reducing or horizontal resistance attributable to major resistance genes when they are defeated by virulent isolate of a pathogen. The resistance genes Rps(1), Rps(3), Rps(4), Rps(5), and Rps(6) were studied in a factorial arrangement by constructing 28 ofthe 32 possible combinations of these genes in 'Harosoy' near-isogenic lines. Residual effects were measured in terms of broad sense tolerance, narrow sense tolerance and horizontal resistance by root inoculation of the near-isogenic lines with Pmg race 7 which is virulent against all five resistance genes. Of the five genes studies, only the Rps(5) gene demonstrated residual effects when defeated by Pmg race 7. Broad sense tolerance was lower when
the (defeated) dominant allele of this gene was present in the Harosoy background than when only recessive alleles were present. The decrease in broad sense tolerance was due primarily to a decrease in horizontal resistance, but also was due in part to a decrease in narrow sense tolerance (tolerance per se).
Impacts
(N/A)
Publications
|
Progress 01/01/88 to 12/30/88
Outputs
An experiment was conducted to compare two types of unbordered microplots, (1) hill plots, and (2) single, short row plots, for selection of single-plant progenies in soybeans. The genetic material consisted of 33 F(2)-derived lines, selected only for similar maturity. Microplot tests were conducted in eight environments, using a split-plot design with microplot type as the whole-plot factor. The lines were also tested in four-row, bordered plots at three locations for three years. Maturity, height, lodging, and seed yield were determined in each test. Genotypic correlations were obtained between each trait in each plot type and each trait in four-row plots. Variance and covariance components were used to calculate expected gains in four-row plots upon selection in each type of plot. Response to direct selection for yield was approximately 30% greater in short row plots than in hills. Selection for maturity and height wereequally effective in short rows and
hills, but short rows were more effective for selection for lodging resistance. Increased height and lodging was an indirect reponse associated with selection for yield in hill plots, but to a lesser extent in short rows. We concluded that short rows are an attractive alternative to hill plots for testing single-plant progenies. Based on expected gains for different resource allocations and on practical considerations, we recommend that short row plot tests be conducted at two or three locations with a single replication per location.
Impacts
(N/A)
Publications
|
Progress 01/01/87 to 12/30/87
Outputs
Computer programs simulating two methods of generation advance, early generationtesting (EGT) and single seed descent (SSD), were used to obtain estimates of genetic gain for the two methods. The EGT scheme involved testing and selection in the F(3) and F(4) of F(2)-derived lines, followed by testing of F(4)-derived lines from selected F(2) progenitors. In the SSD procedure there were two years of testing F(4)-derived lines. Identical F(1) genotypes and equal testing resources were assumed. Three genetic models (additive, additive plus complementary epistasis, and additive plus duplicate epistasis) and three levels of environmental variance were used in a factorial arrangement with 15 replications. The EGT procedure was superior to SSD in all combinations, except for one (additive plus duplicate epistasis, high environmental variation), in which EGT and SSD had equal gain...A male-sterile soybean strain was interplanted with eight pollen parents for three years in an
experiment to deterine whether parents mated using genetic male-sterility would contribute equally to the resulting population. The eight pollen parents were chosen so that combinations of simply-inherited marker traits would allow estimation of the frequency of their contribution by examination of the F(2) progeny. Deviations from equal frequency were statistically significant in all three years but probably not large enough to discourage the use of male-sterility-promoted intermating by soybean breeders.
Impacts
(N/A)
Publications
|
Progress 01/01/86 to 12/30/86
Outputs
The first performance data on a new group of breeding lines of a self-fertilizedcrop are usually obtained in microplots with a set of random lines from a series of crosses. Although replication of individual lines is limited, cross means are obtained relatively precisely. According to theory, a selection index incorporating line and cross means would be the most effective selection criterion. Five selection methods were applied to 8 sets of soybean lines, each set consisting of 20 F(2)-derived lines from each of 4 crosses. The 650 lines were evaluated in single 1.5 m row plots at 3 locations in 1985. Selection intensity was 10%. In 4 of the methods (selection ignoring cross structure, selection within crosses, selection confined to the best cross, and index selection), selection was based on a linear index of yield and maturity. In the fifth method, yield alone was used and cross structure was ignored. Composite populations from each method, together with an
unselected composite, were evaluated for each set at 3 locations in standard 4-row plots with two replications in 1986. Maturity of composites was determined on a random sample of individual plants at two locations. Mean yields and maturities did not differ across methods at a significance level below 40%. The composites will be tested again in 1987, but it appears that any advantage of index selection to incorporate cross and line means is small.
Impacts
(N/A)
Publications
|
Progress 01/01/85 to 12/30/85
Outputs
F(3) or F(4) bulks derived by crossing 12 soybean lines as females to each of four male lines were evaluated in five environments. Parental lines were chosen for maturity (late group II-early group III) and seed yield and included 6 adapted (A) and 6 exotic (E) females and 2 A and 2 E males. The objectives were to compare male and female effects on bulk performance both within and between the A and E groups and to determine whether female x male interactions occurred more frequently in association with bulks derived from A or E parents. Bulks derived from E males averaged 10% less in seed yield than those from A males; the corresponding difference for females was 4%. There were differences within both the A and E female groups for all traits except seed yield. Significant components of female x male interaction occurred for plant height, seed weight, and, to a lesser extent, protein concentration and lodging score. Female x male interactions, whenever they
occurred, were associated with bulks having at least one E parent. There were no female x male interactions for seed yield or soil concentration. It is likely that selection for yield per se would be the most effective method to identify parental strains that would produce superior progeny. If a testcross procedure were employed, the results of this experiment indicate that the choice of tester would not be critical.
Impacts
(N/A)
Publications
|
Progress 01/01/84 to 12/30/84
Outputs
An experiment to measure the inheritance of seed quality factors and their relationship to agronomic traits was carried out with three populations of random lines. Agronomic traits were measured in hill plots in two replications per environment, and the harvested seeds were subjected to seed quality tests. Standard germination percentage had heritabilities of 0.70, 0.86, and 0.46 (live-mean basis), respectfully, in the three populations with nonsignificant genotype x environment interaction. The heritability estimates for accelerated aging test score were 0.42, -0.26, and -0.04, with genotype x environment interaction highly significant in the last two populations. No consistent correlations between seed quality and agronomic traits were observed. A test to measure seedling emergence by estimating the force exerted by germinating seedlings was unsuccessful because of excessive variation from seed to seed within a seedlot. Development of a germplasm population, H
53, was completed in 1984. The population contains genetic variation for phytophthora root rot tolerance and insect resistance as well as yield and agronomic traits. It will be proposed for release in 1985 and used in future years in studies of breeding methodology.
Impacts
(N/A)
Publications
|
Progress 01/01/83 to 12/30/83
Outputs
A testcross study was conducted involving 12 soybean crosses (4 liens crossed toeach of 3 testers). Twenty-four F(2)-derived lines from each cross were evaluated in hill plots at 2 locations. Significant differences in general combining ability were observed for yield, maturity and yield adjusted for maturity. For yield and adjusted yield, specific combining ability effects were both statistically significant and large enough to be of practical importance (up to 12.1 and 8.7% of the grand mean for yield and adjusted yield, respectively). These results are suggestive of a substantial additive x additive epistatic component in the inheritance of yield. In this study, Pella had superior general combining ability for yield and also transmitted to its progeny populations a negative correlation between yield an lodging score, in contrast to the positive correlation usually observed...In the same 12 crosses, no evidence was found for linkage or pleiotropic effects of
common marker genes (I, T, W(1), R, L(2)) on yield...Sixteen F(3) maturity group bulks were yield-tested in 1981 at 3 locations. Lines from each bulk were extracted and tested in 1982 and 1983. The correlation between bulk yield and line mean yield for the 16 crosses was 0.69.
Impacts
(N/A)
Publications
|
Progress 01/01/82 to 12/30/82
Outputs
Two intermated populations developed at Iowa State University were evaluated in hill-plot tests at two locations in 1981. These were AP10 and AP14, developed by intermating 40 plant introductions and 40 varieties and elite lines, respectively. A total of 120 S(3) lines from each population were evaluated along with 7 adapted check varieties. Phytophthora root rot reduced stands, so that only 83 lines from AP10 and 100 from AP14 were harvested. Compared to the mean of the checks, AP10 averaged 3 days earlier and 33% lower in yield, while AP14 was two days later and 13% lower yielding. Forty-three lines from AP10 were selected and evaluated in row plots at two locations in 1982. In addition, twenty of these lines were intermated to initiate recurrent selection. Comparison with check varieties common to the 1981 and 1982 evaluations permitted estimates of the gain achieved from selection. The group of 43 lines gave an increase of 4.9% in yield over the original
population, but averaged 3 days later in maturity. For the group of 20 lines, estimated gain was 7.85, but maturity was only two days later. Realized heritability for yield was approximately 0.17. Studies in progress in 1982 include an evaluation of selection among crosses by testing maturity group bulks, a search for linkages between genes for pigments and yield genes, and a study of the correlation between seed weight and yield and seedling emergence. Development of intermated populations for recurrent selection studies continues.
Impacts
(N/A)
Publications
|
Progress 01/01/81 to 12/30/81
Outputs
A study of intergenotypic competition within hill plots of soybeans was conducted at two locations. Two-line blends were used to simulate introplot segregation. The maturity difference between blend components was not consistently related to the type of competition observed between components or to the relative contribution of components to total plot yield. It was concluded that maturity differences of up to ten days within a segregating entry could be tolerated without substantially reducing the effectiveness of selection for yield, provided an average maturity could be obtained for each plot. Approximately 200 S(3) lines from each of two synthetic populations, AP10 and AP14, were evaluated in hill-plot yield tests at two locations. Superior lines from these tests will be evaluated further in 1982, and one or both populations may serve as a base population for recurrent selection. An additional 800 lines, most from crosses involving exotic germplasm, were
evaluated in hill-plot yield tests. Artificial and natural crossing were used to initiate or continue development of ten synthetic populations to be used in future studies of gene action, linkage, and the effects of selection.
Impacts
(N/A)
Publications
|
|