Evidence for the tri-hybrid origin of Tripsacum andersonii Gray
-- Marc Barré, Julien Berthaud, Diego González-de-León and Yves Savidan

Tripsacum andersonii has 64 chromosomes (Levings et al., Crop Sci. 16:63-66, 1976). Since this counting, this species was postulated to be the result of a hybridization event between a Zea (10 chr) and a Tripsacum (3x=54) species. De Wet et al. (Amer. J. Bot. 70:706-711, 1983) proposed T. latifolium (2n=2x=36) as the putative Tripsacum parent based on its highly unique morphological features. Studies by Talbert et al. (Amer. J. Bot. 77:722-726, 1990) have suggested that the Zea genome is from Zea luxurians and the Tripsacum genome is from T. maizar or T. laxum rather than from T. latifolium (2x). This conclusion was based on analysis of restriction fragments revealed by an rDNA probe (pzmr1) of BamHI-digested DNA: clearly, a 1.6 kb band is present in T. andersonii, T. maizar and T. laxum but not in T. latifolium (2x), T. dactyloides or T. peruvianum.

Since then we have surveyed the diversity of wild Tripsacum populations in Mexico and have assembled a large collection. Among the T. latifolium accessions, we found two types that have the same gross morphology except that one has paired sessile spikelets and is diploid (as described for the botanical type of the species), while the other is triploid and has paired spikelets, one sessile, one shortly pedicellate. From these unpublished observations we had derived the hypothesis that triploid T. latifolium would be a hybrid between a diploid T. latifolium and another Tripsacum species belonging to the Fasciculata section (to explain pedicellate spikelets) and, as a corollary, that this hybrid is the putative Tripsacum progenitor of T. andersonii.

Using the same probe/enzyme combination as Talbert et al., we analyzed DNA samples from different species of Tripsacum to determine the occurrence of the 1.6 kb band and test the robustness of a conclusion based on the presence/absence of such a band. Some results are shown in Table 1 and Figure 1.

Three bands of interest were detected in our collections: "A", a high intensity 1.6 kb band similar to that described by Talbert et al.; "B" a low intensity 1.6 kb band; and "C" a low intensity 1.9 kb band that is always found when B is present (Fig. 1).

We believe that bands B and C are here reported for the first time and that they are essentially different from band A but have, so far, no bearing on the evidence used for supporting our hypothesis on the origin of triploid T. latifolium.

As was recorded by Talbert et al., band A is absent from diploid T. latifolium. It is also absent from most other species with the exception of T. manisuroides, T. maizar, T. laxum, triploid T. latifolium and T. andersonii (Table 1). This narrow distribution suggests that one of the latter species could be the putative Tripsacum progenitor of T. andersonii. On the basis of their morphological differences with T. andersonii, it is improbable that T. manisuroides, T. maizar or. laxum would be good candidates. By contrast, the morphologically unique resemblance between T. latifolium and T. andersonii (supporting De Wet et al.'s observations on the diploid), as well as precisely the adequate number of chromosomes (54) and the presence of the diagnostic A band, make triploid T. latifolium the best putative progenitor of T. andersonii. Under this hypothesis, we would also propose that the two collected triploid T. latifolium accessions, both having pedicellate spikelets, may well be derived from hybridization events between diploid T. latifolium (no 1.6 kb band) and T. laxum (1.6 kb band), which has pedicellate spikelets and is the only species that we have found to be sympatric with T. latifolium in our surveys.

Table 1. Distribution of bands A (1.6 kb, intense), B (1.6 kb, faint) and C (1.9 kb, faint) in samples from our Tripsacum collection (S.A. = South America; MEX = Mexico).
Pop # Species Origin Ploidy Pattern # in Fig. 1
507 andersonii S.A. 64 chr. A a
528 australe australe S.A. 2x none
544 australe hirsutum S.A. 2x none
606 cundinamarce S.A. 2x none
612 S.A. 2x none
57 bravum MEX 2x BC
15 MEX 4x BC
38 MEX 4x BC
121 MEX 4x BC
127 MEX 4x BC
132 MEX 4x BC
148 MEX 4x BC
111 dactyloides hispidum MEX 2x BC
151 MEX 2x BC
67 MEX 4x BC
37 dactyloides mexicanum MEX 4x BC
38 MEX 4x BC
39 MEX 4x BC
40 MEX 4x BC
60 MEX 4x BC
83 MEX 4x BC
156 MEX 4x none
14 intermedium MEX 4x BC
96 jalapense MEX 4x BC
126 lanceolatum MEX 4x BC
142 MEX 4x BC
77 latifolium MEX 2x none c
106 MEX 2x none b
73 MEX 3x A d
109 MEX 3x A e
76 laxum MEX 2x A f
95 manisuroides MEX 2x A
3 maizar MEX 2x A h
21 MEX 2x A i
99 MEX 2x A g
39 pilosum MEX 2x none
46 MEX 2x none
47 MEX 2x none
139 MEX 2x none
68 MEX 4x BC
49 zopilotenseMEX2xBC


In conclusion, we propose that the data discussed above support the hypothesis that the genetic constitution of T. andersonii was derived from two hybridization events :
    1) T. latifolium (2x) x T. laxum (2x) => T. latifolium (3x=54)
    2) T. latifolium (3x) x Zea luxurians(2n=20) => T. andersonii (54+10 chromosomes)

The first event may have occurred at least twice given that the two T. latifolium populations have different isozyme profiles (data not shown). That the second event has probably been unique is strongly suggested by at least two lines of evidence: more than 20 different accessions of T. andersonii from several South American countries show exactly the same morphology and the same isozyme pattern (data not shown).

T. andersonii may therefore be an example of how an apparently very improbable set of events can give rise to a new species.

Figure 1. Luminograph of Southern blots probed with rDNA probe pzmr1 after digestion with BamHI. Lanes: a: T. andersonii; b&c: T. latifolium (2x), d&e:T. latifolium (3x), f: T. laxum; g h&i: T. maizar.

Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors

Return to the MNL 68 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page