The work of Kato, Longley, Blumenschein, McClintock and Brown has provided us with abundant data on chromosome knob frequencies for most regions and subdivisions of maize and teosinte. A reanalysis of these data has been part of a 15-year survey of variation patterns in the genus Zea, based on iterative analysis of the many data tables. It has been possible to describe groupings of taxa (collections and races, Table 1) and of variables (knob presence or frequency, Table 2).
Table 1. The population samples used to calculate the averages in Table
3. Generally prior classifications can be used, but Balsas is divided because
of significant differences between the northwest and southeast ends of
its range. There is remarkably little difference between the Guanajuato
and Michoacan populations of Central Plateau teosinte.
| Taxonomic unit | Samples used for averages |
| Zea luxurians | Cutler 202 A; Wilkes 51764, 51839, 51850 |
| Zea mexicana | |
| Huehuetenango | Wilkes San Antonio Huixta, Tzibaj, Monajil |
| Balsas, SE | Kato K-69-13; Beadle 1972 El Salado (3 collections); Wilkes 47335 |
| Balsas, NW | Kato K-67-13, -14; Wilkes 47890, 47942 |
| Central Plateau | from Guanajuato:
Kato K-89-1, -2, -7. -10, -11;
from NE Michoacan: Kato K-69-3, -4, -5, -6, -8, -9; Wilkes 45470 |
| Chalco | Kato K-65-2,
K-66-1, K-67-1, -2, -3, K-68-1, -2, -4,
-6, K-69-12; 1970 T-1, T-2, T-3, T-4 |
| Nobogame | Wilkes Nobogame (2 collections) |
| Zea mays | |
| Olotillo | Yucatan 37; Guatemala
108, 130, 131, 134, 314, 322,
600, 744, 769, 875 |
| Nal-tel | Campeche 18,
29, 37, 102; Yucatan 7, 102, 129, 146;
Guatemala 52, 115, 145, 260 |
| Pollo | Cundinamarca 401; Pollo Segregaciones |
| Conico | Guanajuato 2; Hidalgo 22; Mexico 3, 182, 207; Tlaxcala 1 |
| Coroico | Bolivia 992, 1035, 1063, 1071 |
| Chulpi | Chile 435, 438; Ecuador 387, 434, 598, 697 |
Table 2. Knobs arranged into sets according to their
intercorrelations and listed in the order of their average frequency. A
is the Andean set and is the only one present in dozens of high altitude
races. C1 and C2 contain the most frequent knobs in the Caribbean and Middle
American lowland races. One or two more sets of internal knobs could be
defined for the latter region, usually found in low frequency.
| Set | Knobs |
| Terminal knobs | |
| T1 | 2S3, 1S3, 6L5, 4L3, 10L3, 2L2, 7L3, 5L2, 9L3, 8L3 |
| T2 | 4S2, 3S2, 5S2 |
| T3 | 9S, 7S, 8S |
| Internal knobs | |
| C1 | 5L1, 8L1, 4L1, 3L1, 2L1, 6L2 |
| C2 | 1S2, 8L2, 9L2 |
| A | 7L1, 6L3 |
In maize there is great contrast between Andean-Amazonian races with high frequencies of knobs at the 7L1 and 6L3 positions (Knob set A) and Caribbean-Middle American races with four sets of knobs in varying frequencies (Sets A, C1, C2 and T3). Within each of these regional subdivisions, frequencies vary greatly.
In annual teosinte the major variation is between the Guatemalan teosintes with many knobs terminal on long and short chromosome arms and Mexican races with few terminal knobs and many internal knobs. However, tassels and spikelets of Zea luxurians of southeast Guatemala differ greatly from those of race Huehuetenango of northwest Guatemala which has the Z. mexicana form (Table 3). Mexican annual teosintes differ more subtly; all are fairly similar in knob patterns to Caribbean-Middle American maizes, with Nobogame the least so. Chalco seems to be the most like maize in knob pattern, as well as plant form. Central Plateau, including several geographic populations with somewhat different knob frequencies, is much like Chalco in several ways. There is considerable variation in the Balsas race, resolvable into a northwest group of populations, in southern Michoacan, with levels of knobs 7S, 8S, 4L1, 2L1, 1S2, 3L2, 7L2 and 3S1 higher than in a southeast group near Chilpancingo, Gro. (El Salado, Mazatlan, etc.). The latter group has shorter plants with more tillers and narrower leaves than the Michoacan group. Nobogame of Chihuahua in northwest Mexico has low knob frequencies and several traits indicative of introgression with local maize. Several teosinte populations are not on Table 3. A Central Balsas group (Kato's no. 8-11 and 16), a western state of Mexico group (no. 32-37), and collections from eastern Michoacan (no. 49-54) are intermediate between two or more of the other groups.
Table 3. Chromosome knob frequencies, by sets of
knobs, using data from Kato (1975, Mass. Agr. Exp. Sta. Bull. 635), Longley
and Kato (1965, CIMMYT Res. Bull. 1) and McClintock (in Ramirez E. at al.,
1960, Races of Maize in Bolivia; in Timothy et al., 1961, Races of Maize
in Chile; in Timothy et al., 1963, Races of Maize in Ecuador). Few knob
positions are occupied in all members of a race. The population samples
used and the knob set compositions are listed in Tables 1 and 2. N' is
the number of population samples used per race, and N" is the number of
plants used for the knob counts.
| Percent of knob positions filled | ||||||||
| Terminal sets | Internal sets | |||||||
| Taxonomic unit | N':N" | T1 | T2 | T3 | C1 | C2 | A | |
| Zea luxurians | 4:18 | 73 | 69 | 0 | 0 | 0 | 0 | |
| Zea mexicana | ||||||||
| Northwest Guatemala | ||||||||
| Huehuetenango | 3:26 | 53 | 76 | 75 | 0 | 0 | 0 | |
| Central Mexico | ||||||||
| Balsas, SE | 5:25 | 0 | 20 | 27 | 44 | 18 | 86 | |
| Balsas, NW | 4:25 | 0 | 20 | 63 | 67 | 40 | 60 | |
| Central Plateau | 12:59 | 0 | 2 | 13 | 60 | 42 | 42 | |
| Chalco | 14:33 | 0 | 8 | 36 | 75 | 27 | 52 | |
| Northwest Mexico | ||||||||
| Nobogame | 2:16 | 0 | 0 | 3 | 20 | 4 | 3 | |
| Zea mays | ||||||||
| Caribbean pattern | ||||||||
| Olotillo | 11:35 | 0 | 0 | 39 | 94 | 79 | 97 | |
| Nal-tel | 12:64 | 0 | 1 | 37 | 89 | 61 | 98 | |
| Pollo | 2:15 | 0 | 0 | 27 | 74 | 20 | 63 | |
| Conico | 6:49 | 0 | 0 | 16 | 39 | 10 | 41 | |
| Andean pattern | ||||||||
| Coroico | 4: 6 | 0 | 0 | 0 | 7 | 0 | 100 | |
| Chulpi | 6: 9 | 0 | 0 | 0 | 0 | 0 | 67 | |
There is considerable correlation between many of the knobs--7S correlates with 9S, 8L2 with 9L2, etc. The "sets" of Tables 2 and 3 are groups of knobs demonstrating similar trends for at least the major types of maize and teosinte. The correlations between knobs are important in two ways--a reduced set of factors seems to be responsible for their frequencies and many fewer measurements are needed to derive the same patterns. Knobs have varying degrees of independent control, so they have varying importance in taxonomic and evolutionary models. In low frequencies there are 19 other knobs, all with some geographical-racial relationships--two terminal knobs are limited to northwest Guatemala teosinte (average frequency of 13%), seven internal knobs are shared by Mexican teosinte and Caribbean-Middle American maize (averages, 26% and 7%, respectively), and ten internal knobs are restricted to Mexican teosinte (average, 4%).
Robert McK. Bird
Return to the MNL 54 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage