Androgenic haploids have been produced for at least 50 genera, most notably in the cereals and vegetable crops (Foroughi-Wehr & Wenzel 1989), and use has been made of this technology for the rapid production of homozygous lines in these plants. Woody plants generally have been recalcitrant, and there have been few successes, particularly for forest tree species, among a large number of reported attempts over the last 40 years. Plantlets have been obtained from pollen culture of species of Hevea, Aesculus, Citrus, Vitis, Malus, Litchi, Euphorbia, Poncirus, Lycium and Camellia (Chen 1987). Hardwood forest tree species for which pollen cultures have produced plants are Betula pendula (Radojevic & Kovoor 1986) and 13 species or interspecific hybrids of Populus (Chen 1987). Some of the resultant trees are now several years old in the field, and have flowered (Chen 1987). Regeneration of plants has been achieved with haploid megagametophyte cultures of a number of gymnosperms (Rohr 1987). Mostly these are cycads (Cycas, Zamia, Ceratozamia) or Ephedra, but included are two conifers - Picea abies and Sequoia sempervirens.
This record suggests that work aimed at deriving methods for the regeneration of plants from haploid tissues of forest tree species is likely to be long-term with a relatively high risk of an unsuccessful result.
In Brassica, for example, production of haploid plantlets from anther culture has been used as a method of fixing gamete arrays of homozygous plants. An F1 hybrid population is generated, and anther culture, with either spontaneous or colchicine induced chromosome doubling, is then used to fix true breeding homozygous diploid lines (Beversdorf & Kott 1987). In the absence of the anther culture technology, derivation of the homozygous lines would take several generations of inbreeding. Another culture therefore has had a direct and important application in the breeding of self-pollinated crops.
With the exception of Leucaena leucocephala, self-pollinating approaches are not a feature of breeding programmes for forest tree species, and the availability of haploids is not as immediately useful in these breeding programmes. The question to be asked then is - to what extent should different breeding approaches be contemplated for forest tree species, e.g. an inbreeding and hybridization approach such as in maize? Current impediments to a maize breeding approach for forest tree species are inbreeding depression (severe in many cases), and long generation intervals - several cycles of selfing could not realistically be contemplated. Even if haploids can be obtained e.g. in poplar, the generation interval would still be an impediment in relation to consecutive testing of homozygotes and heterozygotes. This approach is not likely to be advantageous in a species which is easily propagated vegetatively. Some non-industrial species, e.g. Gliricidia, flower early and can be selected at an early age. Sophisticated breeding programmes are unlikely to be warranted, however, for these species.
Induction of haploid plants thus does not have immediate application in forest tree improvement programmes. Such plants may be of some use in basic genetic studies, e.g. for studies of heterosis in forest tree species. As a long-term strategic research objective for industrial species, induction of haploid plants should be of low priority until such time as methods for early selection and the promotion of early flowering are available.