Current trends in grass cultivar development are reviewed, with respect to the range of species involved, and the objectives and methodology within each species. Extrapolations and predictions are made about future directions and methodologies. It is assumed that selection will necessarily cater for the following environmental changes: (1) higher year-round temperatures, higher variability of rainfall incidence, and lower total winter and spring rainfall along the south of the continent; (2) higher nutrient and lime inputs as land utilisation intensifies; and (3) the grazing management requirements of the important pasture components will be increasingly defined and met in practice.The 'big four' species, perennial ryegrass, phalaris, cocksfoot and tall fescue, will continue to be the most widely sown species in temperate regions for many decades, with the latter 3 increasing most in area and genetic differentiation. However, species diversification will continue, especially with native grasses, legumes, and shrubs from fertile regions of Australia and exotics from little-explored parts of the world, such as South Africa, western North and South America, coastal Caucasus, and Iraq–Iran. By contrast, the recent high rate of species diversification in the tropics and subtropics will probably give way to a much lower rate of cultivar development by refinement and diversification within the established species. Domestication of native grasses will continue for amenity, recreational, land protection, and grazing purposes. As seed harvesting technologies and ecological knowledge improve, natural stands will become increasingly important as local sources of seed. It is suggested that many native grasses have been greatly changed by natural selection so as to withstand strong competition from introduced species under conditions of higher soil fertility and grazing pressure. Conversely, some introduced species are being selected consciously and naturally to persist in regions with irregular rainfall and less fertile soils. Therefore, the distinction between native and introduced grasses may be disappearing, and many populations of native species could now be as foreign to the habitats of pre-European settlement as are populations of introduced species that have been evolving here for 50–200 years. Methods used for genetic improvement will continue to be selection among both overseas accessions and the many native and introduced populations that have responded to natural selection in Australia. As well, there will be deliberate recurrent crossing and selection programs in both native and introduced species for specific purposes and environments. Increasingly, molecular biology methods will complement traditional ones, at first by the provision of DNA markers to assist the selection of complex traits, and for proving distinctness to obtain Plant Breeders' Rights for new cultivars. Later, genetic engineering will be used to manipulate nutritive value, resistance to fungal and viral diseases, and breeding systems, especially cytoplasmic male sterility and apomixis, to utilise heterosis in hybrid cultivars of grasses, particularly for dairying and intensive meat production.Areas where the practice and management of grass breeding and selection programs could be improved are highlighted throughout the review, and reiterated in a concluding statement. Most problems appear to stem from inadequate training in population ecology, population genetics, evolution, and quantitative inheritance.
Epistasis and natural selection shape the mutational architecture of complex traits