Genetic diversity among maize hybrids

Genetic diversity among and within cultivars Of OUT major crop species is d esirable to reduce potential impact of economic losses due to environmental and biological stresses. The Southern corn leaf blight (Bipolaris maydis N isik.) epidemic of maize Zea mays L.) in 1970 emphasized the importance of genetic diversity among maize hybrids to provide protection against this di sease causing organism. Genetic diversity values (GDVs) calculated from fie ld data and genetic distances calculated on the basis of restriction fragme nt length polymorphisms (RFLPs) have been suggested as measures of genetic diversity. Both methods have a theoretical range of zero (no genetic divers ity) to one (greater genetic diversity). Thirty-six hybrids produced from p arental lines having different levels of Reid Yellow Dent and Lancaster Sur e Crop germplasm were compared by both methods. GDVs were calculated for ni ne traits evaluated in seven environments, but only grain yield provided me aningful estimates of genetic diversisty. Genetic diversity values from fie ld data had large errors, but no genotype by environment interaction. Genet ic diversity values were able to identify hybrid comparisons with either ze ro or two parents in common. Genetic diversity values ranged from -0.05 to 1.12, based on comparison set means (GDVI), and from -0.01 to 1.10, based o n entry means (GDVII). The correlation between GDVs based on comparison set means (GDVI) and based on entry means (GDVII) was 0.94**. For comparisons of hybrids having more than 50% Reid Yellow Dent germplasm and having a com mon parent, the GDVs deviated significantly from the expected, based on ped igree information. Genetic distances ranged from 0.00 to 0.43 and were able to discriminate among hybrids that included zero, one, or both common pare nts. Correlation of genetic distances and average GDVs was 0.80**. Deviatio ns from the correlation were more common for comparisons of hybrids that ha d greater than 50% Reid Yellow Dent germplasm and sharing a common parent. Reasons for differences between estimates of GDV and genetic diversity were not obvious. Estimates based on GDVs and RFLP genetic distances could comp lement one another and partially overcome their individual limitations. It seems an initial screening with a highly polymorphic set of RFLP probes wou ld be desirable. Genetic distances based on these data would be able to dis criminate among hybrids with 0, 1, or both parents in common and would be q uicker and require less resources than GDV field data. If additional divers ity information is desired for the hybrids being compared, genetic diversit y estimates based on GDV would be appropriate. Although Nei's genetic dista nce and GDV provide essentially the same information, the methods would req uire caution to interpret them relative to each other to determine genetic diversity among maize hybrids.