Data provided by Hake (1980, The genome of Zea mays: Its organization and homology to related grasses, Ph.D. dissertation, Washington University, St. Louis) have been used here to draw two modified Wagner tree diagrams. Hake used DNA/DNA hybridization and restriction analysis to measure relationships of two maize races, three races of Mexican annual teosinte and one population each of Z. diploperennis, Z. luxurians and Tripsacum laxum. Four "drivers" were used, here labeled Midwest Dent maize ("M"), Z. mexicana - Balsas race (central Guerrero; "B"), Z. diploperennis (Jalisco; "D") and Z. luxurians (SE Guatemala; "G"), and there were 1-10 "tracers" per driver. Some results are especially curious: Z. luxurians is as close to Tripsacum laxum ("T") as to Ladyfinger popcorn ("P") and Balsas teosinte, while Balsas teosinte and Midwest Dent maize are over twice as far from T. laxum as from each other (Table 1).

Classical Wagner tree diagrams are unable to adequately account for the variation, so a modification has been tried. Upward sloping, compound branches indicate divergence of taxa, and each segment accounts for a part of the total divergence. Horizontal lines are mathematical shortcuts between taxa necessitated seemingly by hybridization and do not count toward divergence (Diagrams A and B).

Once values have been assigned to each of the diverging branch segments, a table (matrix) of modeled distances for that diagram is calculated, and a table of differences between the two distance matrices is calculated (Tables 2-5). The lengths of branch segments are readjusted to minimize the values in the last table.

So far it has been much easier to account for the results using Diagram A, where G is an offshoot of a T-G branch long separated from maize and Mexican annual teosinte, but which later interacts with proto-maize once. Diagram B overly shortens the B-D distance, and calls for two hybridizations of G with proto-maize between two hybridizations of G with T, all quite early. The closeness of G and T here is reflected in their sharing terminal chromosome knobs, inflorescence form and stem structure, so it might be proposed that SE Guatemalan teosinte, G, be separated from Zea.

Robert McK. Bird and Sarah C. Hake

Return to the MNL 56 On-Line Index

Return to the Maize Newsletter Index

Return to the Maize Genome Database Page