I should have known there was a sticking point, diploperennis isn't winter hardy, so while those of us living in cold winter areas could make the initial crosses we would need people in warmer climates to do the grow outs and selection work.
There is some interesting reading on the subject in the extract below. It's from
Breeding perennial grain crops by T. S. Cox, M. Bender, C. Picone, D. L. Van Tassel, J. B. Holland, E. C. Brummer, B. E. Zoeller, A. H. Paterson, and W. Jackson (
www.landinstitute.org/vnews/display.v/ART/2002/06/01/3dcbf8a7874a8)There is still scope for hardy perennial sweetcorn but it may be some time coming. Still, the greater the challenge, the grater the success, so what do you say we give this a try anyway.
Extract C. Maize
1. Hybridization with tetraploid perennial teosinte
Efforts to develop perennial maize (Zea mays ssp. mays, 2n=20) have been sporadic at best; as in other crops, hybridization between maize and perennial relatives has led primarily to improvement of the annual crop (Wagoner, 1990a). Shaver (1964) first attempted development of maize-like perennials from crosses between colchicine-induced tetraploids of maize and a wild, perennial, tetraploid relative, Z. mays ssp. perennis (2n=40). Selection within the resulting tetraploid populations and backcrosses to tetraploid maize effectively increased the frequency of perennial progeny. Crosses to diploid maize produced perennial triploids, but all diploid selections were annual.
Shaver (1967) combined a postulated gene (pe) for perenniality with recessive genes for indeterminacy (id) and grassy tillers (gt) in a diploid background, to produce perennial plants; however, the idid genotype prevented production of ears. Because Shaver (1967) had developed a separate idid population in a different genetic background that did produce ears, he suggested that perennial diploids could also be made fertile if the genetic background were manipulated.
2. Hybridization with diploid perennial teosinte
Little further attention was paid to perennial maize until the dramatic discovery of a diploid species of perennial teosinte, Z. mays ssp. diploperennis (Iltis et al., 1979). Initial studies showed that inheritance of perenniality was relatively simple in maize/diploperennis crosses, but perenniality was inferred from tillering habit, a potentially misleading technique (Shaver,1964). In subsequent, larger-scale experiments, inheritance of tillering in progeny of similar inter-subspecific crosses was more complex, and perennial maize types were not recovered even in large segregating populations (Srinivasan and Brewbaker, 1999).
Genetic mapping in maize/annual teosinte crosses show that most traits of domestication separating the species are oligogenic, and the loci tend to be clustered on the map, through either linkage or pleiotropy (Doebley and Stec, 1993). A similar study of these traits, plus perenniality, in crosses between maize and diploid perennial teosinte would be of great value to any breeding program attempting to combine perenniality with the agronomic phenotype of maize. This would require substantial effort to evaluate large segregating populations for tillering, rhizome production, and capacity to produce seed over multiple seasons. Once the genomic regions of interest are identified, marker-assisted selection can be used to incorporate them into a maize background and eliminate unwanted alleles such as those conditioning hard glumes and shattering.
One serious obstacle to adoption of any teosinte-derived perennial grains is the lack of winterhardiness of these tropical species. There are no winterhardy species of Zea. Because the bulk of maize production and breeding occurs in temperate areas, there has been little incentive to develop perennials from crosses with Z. mays ssp. diploperennis. One possible approach has not been suggested to date: selection for rhizome depth. As we have seen, johnsongrass rhizomes also are not winterhardy if near the soil surface, but dispersal of the species into higher latitudes has been made possible by selection for deeper rhizomes. Superimposing selection for this undoubtedly complex trait on selection for perenniality and traits of domestication, not to mention yield, may entail a much larger effort than any breeding program is willing to undertake.
3. Hybridization with eastern gamagrass
The closest winterhardy relatives of maize are in the genus Tripsacum. Eastern gamagrass (T. dactyloides), for example, is currently grown as a perennial forage grass as far north in the western hemisphere as Kansas and Massachusetts, and can be grown in the Corn Belt (Voigt et al., 1995). T. dactyloides has been hybridized many times with maize, beginning with the work of Manglesdorf and Reeves (1931). Plants of the diploid (2n=36) or tetraploid (2n=72) races may be crossed with maize. If Tripsacum is used to pollinate maize, embryo rescue is necessary (deWet et al., 1973; James, 1979), but if maize is used as the male, some hybrid seed may be obtained without rescue (deWet et al, 1973). In addition, several strains of popcorn, when pollinated with tetraploid T. dactyloides, produce large amounts of hybrid seed that does not require embryo rescue (Kindiger and Beckett, 1992). Some have good crossability with diploid T. dactyloides as well. Contrary to typical results, Eubanks (1995, 1997) reported that a putative 20-chromosome hybrid between T. dactyloides and Z. diploperennis showed 93 to 98% pollen fertility.
Natural introgression between Tripsacum and maize has not been observed, but morphological (deWet et al., 1983) and molecular (Talbert et al., 1990; Dewald and Kindiger, 1998) evidence supports the hypothesis that the species T. andersonni is an intergeneric hybrid containing three genomes (54 chromosomes) from Tripsacum and 10 chromosomes from Zea in Tripsacum cytoplasm. The uniformity of this ancient natural hybrid indicates that T. andersonii arose from a single hybridization. It has been able to spread across tropical Latin America because of its vigorous perenniality (deWet et al., 1983; Dewald and Kindiger, 1998).
In addition to being perennial, tetraploid T. dactyloides is a facultative apomict. Perennial hybrids result from artificial crosses between tetraploid Tripsacum and maize (Farquharson 1957), and some seed-set can result from apomixis. The hybrids, derived from parents with different basic chromosome numbers and chromosomes of different sizes (those of maize being larger), are male sterile, with cytological behavior that is anything but regular. Harlan and deWet (1977) summarized methods for utilizing such hybrids in maize improvement. Either 28-chromosome or 46-chromosome hybrids — derived from diploid and tetraploid T. dactyloides parents, respectively — can be backcrossed to maize. In either case, Tripsacum chromosomes are eliminated with backcrossing. Elimination occurs more gradually in progeny of 46-chromosome hybrids, and the 20 chromosomes of the resulting backcross plants can contain significant genetic material from Tripsacum (Harlan and deWet, 1977; Stalker et al., 1977a and b). All 20-chromosome backcross plants derived to date have been annual and non-apomictic. Kindiger et al. (1996) derived an annual, 39-chromosome line that carried 9 Tripsacum chromosomes and displayed an intermediate level of apomixis.
Most hybridization with Tripsacum has been for the purpose of either elucidating the evolution of maize or transferring resistance or other genes to annual maize. The latter purpose implies backcrossing to maize. But development of perennial populations may require interpollinating plants in early backcross generations that still carry many Tripsacum chromosomes (Harlan and deWet, 1977), or even backcrossing to Tripsacum. As Harlan and deWet (1977) commented, "Apparently, if one wishes to contaminate maize with Tripsacum one should first contaminate Tripsacum with maize."
New approaches to perennial maize are being explored. An anomalous fertile hybrid between diploid T. dactyloides and maize was discovered by one of the authors (BEZ) in 1997 near the mouth of the Big Nemaha river in Richardson County, Nebraska, USA. This derivative of natural introgression between T. dactyloides and a putative commercial hybrid is being hybridized with gynomonoecious Tripsacum — both diploid (Dewald and Dayton, 1985; see below) and tetraploid (Salon and Earle, 1998) — and with tassel-seed popcorn to develop a 56-chromosome perennial cultivar for production of grain, forage, fiber, and fuel.
The difficulties encountered in introgressing apomixis from Tripsacum into maize (Kindiger et al., 1996) should temper hopes for a rapid synthesis of perenniality with high grain yield. Whatever the initial population, and even with marker-assisted selection, the process of recovering perennial, winterhardy segregants with maize-like ears, and then breeding for yield and other agronomic traits will be long and arduous.
