The breeding objective at CIMMYT was to develop lines with a high level of resistance to Southwestern corn borer (SWCB) and improved agronomic traits. Earlier work at CIMMYT indicated that borer resistance may be under the control of several genes. Since there were no particular reasons for recovering the genotype of the chosen parental lines, we decided to develop a series of recombinant inbred lines, while selecting for resistance and then for agronomic performance at each successive cycle (S1 to S4).

The six S4 lines for which RFLP data are now available are part of the insect resistance breeding program at CIMMYT. These lines were derived from a single cross between Mp78:518 and an S3 line from CIMMYT's Population 47. Mp78:518 is a yellow endosperm line with very high multiple borer resistance (MBR) that was developed by the USDA-ARS in cooperation with the Mississippi Agricultural and Forestry Experiment Station; however, it is susceptible to ear rot when grown at CIMMYT stations in Mexico, develops iron chlorosis on calcareous soils and stalk lodges. The other parent , a white endosperm S3 line from CIMMYT�s Pop.47, is intermediate to susceptible to corn borers. In contrast to the first parent, this line is resistant to ear rots, tolerates the high pH of calcareous soils, and has a strong stalk. Each cycle was infested with Southwestern corn borer and only the most resistant fraction was carried forward to the next cycle. Additional selection pressure for improvement of other agronomic traits (particularly those contributed by the Pop.47 parent), was applied within the resistant fraction when appropriate.

Out of 101 loci analyzed for RFLPs, 91 were polymorphic between the two parents for at least one of the three enzymes used. All S4 lines were scored at each polymorphic locus and their genotypes determined (homozygous for one of the parents or heterozygous).

We are currently developing algorithms based on inbreeding theory (e.g., Haldane and Waddington, 1931, Genetics 16:357-374), for predicting the expected genomic composition of a given arm after n generations of self-fertilization based on the frequency of crossing over at each meiotic cycle. We shall then be able to calculate the probability of a given genomic composition, and thus estimate whether certain gene combinations significantly reflect the effects of selection for a specific trait. These calculations will take into account the multi-point maps now available for RFLP markers, as well as known cytogenetic features of maize such as chiasma frequencies along specific arms.

The overall expected heterozygosity of any particular S4 line (after five generations of selfing) is (1/2) = 3.125%. The observed values for heterozygosity were, on average, 4.5 times greater than expected.

The effects of selection on any particular area of the genome were estimated from the frequency of parental morphs along each chromosome. If no selection had taken place at a locus, one would expect a 1:1 contribution by each parent. For the six lines in the analysis, a total of 12 morphs (2 per line) is possible at any one locus. Ratios of 9:3, 10:2, 11:1 and 12:0 would be significant at the 90% confidence level (Fisher�s Exact Test).

Since Mp78:518 is expected to contribute the majority of resistance factors, significant deviations toward that particular parent may indicate areas involved in borer resistance. Six regions showed such deviations (in 1L, 2, 3L, 5L, 9S and 10L). Chromosome 2 has several potential regions of importance as almost all of the loci were skewed towards the resistant parent.

Deviations toward the other parent may indicate regions of importance for agronomic traits found in Pop. 47; however, one cannot rule out the possibility that this parent might have contributed resistance factors in these regions. Eleven areas showed such deviations (in 1S, 1L, 3S, 3L, 4, 5S, 5L, 6, 7S, 8L and 10L). Chromosomes 4 and 6 may contain several regions of importance to the Pop. 47 plant type.

The results to date indicate possible areas of resistance and/or desirable agronomic traits in the S4 lines so far studied. We now plan to use this information in future studies of the inheritance of borer resistance. These studies will involve further analysis of additional S4 lines (both selected and unselected) and segregation analysis of the QTL loci responsible for borer resistance.

Return to the MNL 64 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage