Summary
The limitation to the breeding for improved quality in tropical species has always been the numbers of individuals needed to be screened, the need to grow plants to mature individuals before testing for quality and the cost of chemical analyses required for screening and the need to concurrently screen for both quality and agronomic characters. In general, this has proved a major hurdle which has not been successfully overcome.
Molecular biology has provided a method of rapid screening of individuals for the presence of specific genes or gene sequences that control specific biological processes. In relation to quality, genes controlling lignin biosynthesis appear to be one of the most important in controlling digestibility. Molecular techniques are capable of rapidly screening large numbers of individuals in a very early stage of their development (eg. seedling stage), and, combined with traditional selection procedures, show promise in tropical pasture improvement. Kikuyu (Pennisetum clandestinum) is an important summer-growing grass for the animal industries of Australia. It is high yielding and responsive to nitrogen and water. However, as with other tropical grasses, its quality is limited by the levels of lignin which reduce digestibility relative to temperate grasses.
There have been attempts to improve the quality of tropical grasses but the difficulty of selecting for high yields, vigour and high quality concurrently has seen these attempts fail to achieve the goals either by improving quality at the expense of vigour or failing to deliver sufficient improvement in quality to substantially lift animal production. There have been concerns that any selection for improved quality in tropical species may be accompanied by a reduction in agronomic characters such as plant vigour or yield. Kikuyu grows in many areas of Australia but little is known about the genetic variation existing in the natural populations. There are only a restricted number of cultivars and most were derived from an introduction which was used to develop the cultivar Whittet. Unlike most of the other tropical grass species, there is no collection of accessions in the Australian Tropical Genetic Resource Centre at Biloela.
This series of experiments was designed to investigate Australian kikuyu resources to determine whether sufficient variation exists in it to sustain a breeding program for improved quality. All experiments except one used single plants on a 1.5 m square grid as the experimental unit and a randomised block design with replicates varying from 3 to 10. The final experiment used a strip plot design, where single plants were sown on either a 0.5 or 1.5 m grid. All plants except the cultivar Crofts in Experiment 1 were genetically different individuals, either coming from seed or being selected at random from within swards of kikuyu. Experiment 1 investigated the variation in material generated from treating Whittet kikuyu seed with two chemicals known to cause mutagenesis in grasses and in a selection of commercial cultivars. Experiment 2 investigated variation in pre-release populations of four commercial cultivars and a breeding population from NSW. Experiments 4 and 6 investigated the variation in performance of ecotypes of kikuyu collected from regions within Australia.
There were two sites, one at Wollongbar on a plateau area in northern NSW and a second at Mutdapilly in coastal south eastern Queensland. DNA was extracted from each plant and subjected to a modified DAF (DNA amplification fingerprinting) analysis to determine the genetic relatedness of the genotypes to each other. Experiments 5 and 7 investigated the variation within a separate population of mutagenised Whittet seed at the same two sites. The final experiment (Experiment 3) investigated the ability of kikuyu selections with varying ability to produce runners and with different growth habits, to colonise bare areas after establishment. Selections, accessions and cultivars showed considerable variation in both physical and chemical attributes.
In Experiment 1, Cultivars A and B, WK 9, WK 12, WK 39, WK 46 and WK 85 were higher yielding, Cultivar A and WK 42 produced more runners and Crofts, WK 9, Cultivars A and B and WK 42 were leafier than Whittet. In Experiment 2, Whittet was the highest yielding cultivar. The breeding line CND8 was consistently the lowest yielding, both in forage and runner yield and was a less erect type compared with Noonan, Breakwell, Whittet and a common, seeding selection. Noonan and Breakwell were only moderate yielding, a result consistent with their commercial performance. In experiments 4 and 6, significant differences in plant height and yield, runner development, and the crude protein, invitro digestibility and acid and neutral detergent fibre content of leaf was demonstrated between the 10 ecotypes and 6 cultivars evaluated. There was a 4-fold range in plant yield and a 10-fold range in runner production between the ecotypes. Crude protein ranged from 21 to 26%, in vitro digestibility from 67 to 80%, neutral detergent fibre from 53 to 45% and acid detergent fibre from 26 to 18%. Experiments 5 and 7 suggest that the two mutagenic chemicals (either used singly or in combination) randomly resulted in gene mutations but had no effect on specific plant parameters. There were no significant differences in any agronomic or quality parameter between the four mutagenic seed lots at Mutdapilly and only stem yield was significantly affected in one sampling at Wollongbar.
Analysis of the genetic fingerprints of the ecotypes evaluated in Experiments 4 and 5 indicated that they formed two broad groupings. Most of the regional ecotypes grouped with common kikuyu as represented by the material collected from Wollongbar, whilst the Beechmont ecotype grouped with the cultivars Whittet, Noonan and Crofts. The only exception was the Atherton Tablelands ecotype that was not aligned with either group. The variation between individual plants was also substantial. There was also greater variation in the populations generated by mutagenesis than from naturally-occurring populations or within regional ecotypes. In Experiment1, individual plants from Cultivar A, Cultivar B, WK 9, WK 12 and WK 42 demonstrated combinations of high yield and high quality during autumn, while WK 85 showed superior runner vigour but no superior quality, over Whittet. Cultivar A and B did not show the same superiority in the winter or late summer but the other entries continued to perform well. In Experiment 2, most of the elite material (ie. high yielding combined with high quality attributes) came from Whittet, and to a lesser extent from Noonan and an accession of common kikuyu.
Generally the elite individuals from Experiments 4 and 6 came from the best performing cultivars (Whittet and Noonan) and ecotypes (Wollongbar, Numinbah Valley, Gympie and Victoria) but even within lower yielding ecotypes, outstanding plants emerged. The maximum improvement in quality attributes of individual plants was 14% in crude protein, 5% in in vitro digestibility and 6% in ME and a reduction in ADF and NDF of 20 and 16% respectively. However, the levels of improvement declined considerably from these values to around 3-5% when the added selection pressure for high forage yields and runner production was also taken into account. In Experiment 3, sward spread in the first two months reflected the potential runner production predicted from previously conducted experiments. Experiment 1 predicted that WK 42 and WK 64 would produce the most runner mass and this initially was the case. However in the longer term, runner production did not reflect a sward's ability to produce foliage over the period from sowing to maximum sward coverage of 7 months. Selections performed similarly at the two densities; the lower density planting just took longer to cover the bare areas. Sward height was a better predictor but even it only accounted for 22.5% of the variation in forage yield.
To select a successful kikuyu cultivar the following physical attributes would need to be assessed:- runner spread, runner yield and runner vigour, together with forage characteristics such as total and component yield and sward height. It was concluded that, even from this limited evaluation of the material available, there was potential to produce elite lines of kikuyu which would show superior yield, runner production (measures of sward vigour) combined with higher quality parameters (lower fibre, higher digestibility and higher crude protein).
The results also demonstrate that improvements in quality and agronomic attributes were not mutually exclusive and that a breeding program could be expected to achieve improved quality in kikuyu while retaining the vigour of current cultivars. Whilst no immediate use can be made of the results of these experiments, it does strengthen a case for a breeding program to improve the quality of kikuyu. Such improvements can be expected to have a substantial effect on improved returns from intensive animal industries which utilise kikuyu. In the long term, success in this area is likely to have far wider implications on the whole grazing industry because the principles derived here can be equally applied to other tropical grasses.
