Grasses and legumes comprising poor to good quality temperate and tropical species were fed to either cattle or sheep in 36 digestibility experiments. Cell wall in these forages was the ash-free and protein-free residue after sequential extraction with acidpepsin, organic solvents and either water for grasses or ammonium oxalate for legumes. The average amount of cell wall digested per 100 g forage OM was 40-0 + 0-59 g in grasses and 19-8 ± 1-85 g in legumes. It was considered that within grasses and within legumes the physiology of ruminant digestion, rather than forage quality, was the main determinant of the average amount of cell wall digested and the difference between grasses and legumes was due to interaction of the ruminant digestion process with the physical structure of the cell wall. Of forage factors governing variation about the physiological average, the total cell wall had some effect on the amount of cell wall digested, but the lignin concentration in the cell wall had no effect.Among grasses and legumes the average, apparently undigested, protein-free non-cell-wall component was 6-2 ±0-13 g per 100 g forage OM. This component and digestible protein relative to total protein varied among different sets of data. It was concluded that only the component of digestible organic matter which was governed by the relative proportions of cell walls and cellular contents was predictable from chemical composition. It was considered that selection in plant breeding should be based on both digestible cell wall and cell-wall content instead of digestible organic matter.

SUMMARYGrowing beef cattle grazed six pasture types (all-grass or grass-legume mixtures) for 4 months on each. Live weight (mean of ten animals) and dietary energy concentration were measured at 2- or 3-weekly intervals. Values for feed intake estimated by back-calculating from feeding standards were on average 2·5% higher (range, 10% less to 12 % higher) than intake values estimated directly from dietary energy concentration and live weight. In contrast to this level of agreement of means, differences between methods within pastures were so large as to suggest the presence of overriding effects on predictors of voluntary intake. It was concluded that until these overriding effects are explained, there are no means of validating measured intakes in grazing cattle, or of accurately predicting the amount of beef produced from pasture.

Faeces and the residues from the in vitro digestion of 13 forages were sequentially extracted with acidpepsin, ethyl alcohol, hot water and diethyl ether and the quantity of the extracts was adjusted to a protein-free basis and expressed per 100 g of original forage organic matter. Undigested material in in vivo and in vitro experiments were respectively, 3.37 and I .88 g (&O. 12) for ethyl alcohol extracts, 1.13 and 1.10 g (*O.OS) for hot water extracts, 0.15 and 0.12 g (kO.02) for diethyl ether extracts and 1.44 ( 5 0 . 2 2 ) and Og for 0.1 N-HC~ extracts. The diflerence between values for in vivo and in vitro experiments was considered to represent endogenous material present in faeces.The quantity of cell walls digested in vivo per 100 g forage organic matter was 41 ' 5 g i I 'OI for grasses and 21.5 & 0.74 for legumes.

The performance of steers grazing Townsville stylo (Stylosanthes humilis)-native grass pasture, with and without superphosphate fertilizer, was studied at two stocking rate (0.41 and 0.82 beasts ha-1) on the sub-coastal spear grass region of the dry tropics of north Queensland over a period of three and a half years. Fertilizer increased yield and quality of pasture, but invasion of the fertilized areas by annual grass species was occurring by the end of the study. Liveweight gains on fertilized Townsville stylo pastures were greater than on unfertilized Townsville stylo, where performance was similar to that recorded on untreated native pasture at stocking rates of 0.25 and 0.41 beasts ha-1. Liveweight gains at the lighter stocking rate were greater than at the higher level on the Townsville stylo-native grass pasture, but at the stocking rates applied to native pasture there was no significant effect of stocking rate. A stocking rate of 0.41 beast ha-1 on fertilized Townsville stylo was safe in all years. Wastage of material was high owing to mould development on the Townsville stylo which remained as standing hay during the winter.

Total cell wall, in vitro digested cell wall and fermentation-gas production were determined in the separated tops and stubble of five tropical and two temperate grass species grown under controlled temperatures and humidities. As the day/night temperatures increased from 18/10 to 25/17 C C the total cell wall and in vitro digested cell wall increased. With a further increase to 32/24 °C the total cell wall increased, but not the in vitro digested cell wall. In vitro digested cell-wall values were also calculated from a previously derived relationship between in vitro digested cell wall and total cell wall. The differences between observed and calculated values increased (negatively) with increasing growth temperature suggesting that the in vitro digested cell wall was depressed with increasing growth temperatures, but the extent of this depression was small.In vitro gas production from the fermentation of plant tops or stubble in buffered rumen fluid for 24 h was significantly affected by growth temperature and humidity, apparently because of changes in chemical composition induced by the treatments. The volume of gas produced between 24 and 48 h fermentation times was appreciably lower from stubble than from plant tops and this was thought to be due to a higher resistance of part of the cell wall of stubble to digestion by rumen bacteria. Gas production in this period was slightly lower in the tops of grasses grown at the lower temperatures, but this could have been an indirect effect from an associated decrease in the total cell wall.TMTROTyTTPTTOTtf digestible dry matter, and they were led to conclude that the digestibility of cell-wall material was The digested cell wall of both tropical and tern-reduced at the higher temperatures. As the in vitro perate grass species when expressed as an amount digestible cell wall was not directly determined in digested per 100 g forage organic matter tends to this experiment, the conclusion is not unequivocal, be constant (Moir, 1972a(Moir, , 1974. Variation about For the direct determination of in vitro digestible the 'constant' digested cell wall could be due to cell wall, protein and ash corrections are required genetic variation within grass species, or to envir-on the cell-wall residue from extraction of the on mental effects on the cell wall during growth of plant, and on the similarly extracted residue from the plant. A knowledge of factors influencing the in vitro digestion (Moir, 19726). This is a formidable digested cell wall could help plant breeders in the task when there are a large number of plant samples selection of grasses with a high content of digestible for analysis and may be impracticable in pot cell wall.experiments that yield only small amounts of plant In a factorial experiment in which grasses were tissues. Moir (1976) has described a simple method grown at different controlled temperatures and for demonstrating differences between plants in the humidities, Wilson, Taylor & Dolby (1976) found properties of their cell walls by measuring the that t...