THE ROUTINE DETERMINATION OF IN VITRO DIGESTIBILITY OF ORGANIC MATTER IN FORAGES-AN INVESTIGATION OF THE PROBLEMS ASSOCIATED WITH CONTINUOUS LARGE-SCALE OPERATION

A procedure and equipment which make possible a continuous output of 250 to 300 determinations of in vitro digestibility of organic matter per week are described.
The effect on in vitro digestibility values of removing rumen liquor from sheep at various times before and after feeding, mixing of rumen liquors from different sheep, and nitrogen supplementation of the inoculum are investigated. Results are given of a comparison of the effect of soil contamination on in vitro data when expressed as digestibility of dry matter and digestibility of organic matter. The levels of within and between-batch errors are stated and the accuracy of prediction of feeding value is discussed.     

A TWO-STAGE TECHNIQUE FOR THE IN VITRO DIGESTION OF FORAGE CROPS

A simple technique for the determination in vitro of the dry- or organic-matter digestibility of small (0·5 g) samples of dried forages is described. It involves incubation first with rumen liquor and then with acid pepsin. Using 146 samples of grass, clover and lucerne of known in vivo digestibility ( Y ), the regression equation Y = 0·99 X– 1·01 (S.E. ± 2·31) has been calculated, where X = in vitro dry-matter digestibility. This technique has been used for the study of the digestibilities of plant breeder's material, of the leaf and stem fractions of herbage and of herbage consumed by animals.     

Sources of error with in vitro digestibility assay of pasture feeds

A series of experiments was undertaken to determine population statistics for in vitro organic matter digestibility ( in vitro OMD) data and to examine the effects of basal diet, donor animal and precollection fasting interval on the activity and specificity of rumen fluid inoculum. The experiments utilized wether sheep, a diverse set of pasture grass and legume feeds prominent in the Australian subtropics and the Tilley and Terry in vitro digestibility procedure running under the operating pressure of a practicing feeds evaluation laboratory.
The standard errors of in vitro OMD estimates for within and between batch runs were ±0·88 × 10⁻² and 0·62 × 10⁻², respectively. These error terms were used to develop protocols to accept, reject or scale raw in vitro OMD data. Differences between donor animals in the activity of rumen fluid were highly significant. Extending the precollection fasting interval beyond 16 h was associated with a substantial decline in inoculum activity.
An in vitro-in vivo calibration relationship based on fifteen test feeds and using lucerne ( Medicago sativa ) as basal diet was described by the linear model y = 1·3 x-0·195±4·9 × 10⁻² r = 0·79 (y = in vivo OMD, x = in vitro OMD). Despite large effects of basal diet on both the absolute values and relative ranking of test feeds, neither the RSD nor r values were improved using alternative diets to Lucerne chaff.
The results highlight the need to formally standardize the analytical and biological components of the in vitro digestibility procedure to safeguard the integrity of data.     

THE USE OF THE IN VITRO TECHNIQUE IN THE DETERMINATION OF THE DIGESTIBILITY OF GRASS/LEGUME MIXTURES

The in vitro dry-matter digestibility was determined of two samples of Digitaria decumbens , one sample of Medicago sativa , and one sample of Trifolium repens and nine mixtures prepared from these samples.
When the results were compared with in vivo digestibilities the standard errors of the regressions for the three combinations of legume, grass and legume/grass mixtures were ±0.6, ±0.6 and ±1.5 percentage units, respectively.
It is considered that in vivo digestibility of mixtures of grasses and legumes can be estimated by the in vitro technique, provided that suitable grass and legume standards are included.     

PREDICTING ORGANIC-MATTER DIGESTIBILITY FROM IN VIVO AND IN VITRO DETERMINATIONS OF DRY-MATTER DIGESTIBILITY

The DM digestibility (DMD) and organic matter digestibility (OMD) of 36 tropical grass samples were determined both in vivo and in vitro. Regressions relating OMD to DMD were calculated. In vivo OMD=1·02 in vivo DMD+ 0·1, r=0·99. In vitro OMD=1·05 in vitro DMD—4·7, r=0.99. At the same DMD, the in vitro method predicted OMD 2·8 to 3·4 percentage units lower than the in vivo method. This difference was caused by the difference in mean apparent digestibility of the ash, in vitro 71%, in vivo 47%. In vivo DMD was correlated (r=0·92, P<0·01) with in vitro DMD and the regression relating these two factors had a residual standard deviation of ±1·42 digestibility units. This deviation was reduced to ±1·9 digestibility units when OMD was determined. For both DMD and OMD significant differences were found between grass species in the relation between in vivo and in vitro digestibility. To overcome biased estimates of OMD it was suggested that conversion equations should be avoided, and that the in vitro method be standardized with samples of known in vivo OMD as similar as possible to those being tested.     

The effects of forage preservation method and proportion of concentrate on digestion of cell wall carbohydrates and rumen digesta pool size in cattle

Six bulls with ruminal and duodenal cannulae were used in a 6 × 6 Latin square design with a 2 × 3 factorial arrangement of treatments. Barn-dried grass (G) and direct-cut silage (S) from the same sward were fed together with 250 (L), 500 (M) and 750 (H) g kg^-1 total dry matter (DM) of a barley-based concentrate (barley 875 and rape-seed meal 125 g kg^-1) at the level of 80 g DM kg^-1 live weight^0·75. Rumen and total digestibility of cell wall constituents were measured by a double marker and total collection method. Rumen pool sizes of dietary constituents were estimated by emptying the rumen. Particle-associated enzyme activities were measured from rumen particulate material and feed particles incubated in nylon bags in the rumen. Neutral detergent fibre (NDF) and hemicellulose digestibility were higher (P<0·05) for G diets than for S diets while no differences were observed between the forages in acid detergent fibre (ADF) and cellulose digestibility. Both rumen and total digestibility of cell wall constituents decreased with increasing level of concentrate. The proportion of total cell wall digestion in the rumen was unaffected by the forage preservation method and the proportion of concentrate. NDF, and especially cellulose digestibility, declined quadratically with increasing level of concentrate. Dietary effects on particle-associated carboxymethylcellulase and xylanase activity were consistent with those observed in cell wall digestion. There were no differences between the forages in rumen pool size of total ingesta or any dietary constituent. Rumen pool size of total ingesta decreased with increasing level of concentrate. On the other hand, DM content of ingesta increased with the level of concentrate, while the pool size of DM, NDF and ADF declined quadratically reaching a minimum on M level of concentrate. The differences in rumen NDF pool size were mainly in the digestible fraction. Calculation of digestion kinetic parameters showed that both the rate of passage and especially the rate of digestion were markedly depressed in animals fed on the highest level of concentrate. The results suggest that a reduced fate of digestion of NDF with high concentrate diets can be partially compensated for by an increased NDF retention time in the rumen of animals fed at a restricted level of feeding.     

APPLICATIONS OF THE TWO-STAGE IN VITRO DIGESTIBILITY METHOD TO VARIETY TESTING

The in vitro digestibility method described by Tilley and Terry (5) has been applied to varieties of kale, rape, flat poll cabbage, clover, lucerne and silage maize in trials at the National Institute of Agricultural Botany. Several small modifications of the method are described. The method was developed originally for grass and had not previously been used for the brassicas and maize. Steps taken to check in vitro against in vivo results are described briefly in the appendix.
The digestible dry-matter content of the leaf in kale and rape was constant at about 80% but that of the stem varied over a wide range. Varietal differences were found, while a significant negative correlation was established between digestible dry matter and crude fibre. Some difficulties were encountered in applying the method to maize.     

SYSTEMATIC TESTING OF QUALITY IN GRASS VARIETIES

Methods of testing grass varieties are discussed and the efifects of differences in management and of nitrogen fertilizer on quality are shown. In a grazing-management system with 9 cuts per season, cocksfoot was higher than perennial ryegrass in protein and fibre and lower in soluble-carbohydrate content and in vitro dry-matter digestibility throughout the season. All species showed the lowest value for digestibility in August.
Curves for digestibility in primary growth were plotted for 4 varieties of ryegrass for leaf and stem fractions. The tetraploid variety Reveille was slightly more digestible than S24 in leaf and stem throughout the period, while at ear emergence S23 was lower in digestibility in both fractions than S24.
There were significant differenees in composition and digestibility between experimental centres, but relative differences between varieties were consistent. The level of N fertilizer did not materially alter the comparison between varieties in quality data.     

The prediction of the in vivodigestibility of the diet of sheep and cattle grazing indigenous hill plant communities by in vitrodigestion, faecal nitrogen concentration or indigestible' acid-detergent fibre

Herbage and faeces samples were retained from an intake and in vivo digestibility trial using material harvested at three different stages from five indigenous hill plant communities (those dominated by Agrostis-Festuca., Nardus stricta., Molinia caerulea., Eriophorum vaginatum and Tri-chophorum caespitosum) and from sown swards of ryegrass or white clover. Samples of the herbages and of their separate components, together with extrusa samples of the same herbages recovered from pen-fed sheep and cattle oesophageal fistulates, were digested in vitro using rumen liquor. Measurements were made also of faecal nitrogen concentration (FN) and of indigestible acid-detergent fibre (IADF) using the samples from the original trial.
The range of in vitro disappearance values of the herbages (IVOMD; 0809-0278) was slightly wider than those of in vivo digestibility (OMD; 0-796-0-37I). Differences between OMD and IVOMD were greatest at low quality and relationships were best described by two separate linear regressions of OMD on IVOMD for (a) ryegrass, white clover, Agrostis-Festuca and Nardus (RSD; 0 0185) and, (b) Molinia, Tricho-phorum and Eriophorum (RSD; 00246). In vitro disappearance values of extrusa were higher than those of herbages offered by 00503 and 00156 units for sheep and cattle respectively, partly because of greater levels of ensalivation, especially in the sheep. The relationships of OMD to IADF and to FN were poor, though inclusion of faeces output and fitting parallel lines for communities in the regression of OMD on FN reduced the RSD to 0020. We conclude that in vitro digestion of samples of extrusa, using appropriate standards, is the best method of diet digestibility prediction for sheep and cattle grazing these communities.     

Anatomical structure and digestibility of reed canarygrass cultivars and hybrid ryegrass

Anatomical structure and fibre chemistry influence the digestibility of forage grasses. The objective of this study was to evaluate the anatomical structure, digestibility, and fibre composition of leaf blades from a high alkaloid and two low alkaloid cultivars of reed canarygrass ( Phalaris arundinacea , L.) and one cultivar of tetraploid hybrid ryegrass ( Lolium hybridum , Hausskn.). Cultivars were grown in a greenhouse in a randomized complete block design with four replications. Leaf blades were either cut into 10-mm sections or dried and milled to disrupt anatomical structure.
Reed canarygrass had smaller cells and airspaces and slower tissue digestion than ryegrass. The in vitro disappearance of ryegrass sections in buffered rumen fluid was 60% greater ( P <0·01) than that of reed canarygrass after 6, 12 and 24 h of digestion. By contrast, the in vitro disappearance of milled ryegrass was similar to or less ( P <0·05) than that of reed canarygrass. In vitro dry matter digestibility and fibre composition of the grasses were similar except for acid detergent fibre, which was greater ( P <0·01) in ryegrass than in reed canarygrass. Reed canarygrass cultivars were similar in anatomical structure, digestibility, and fibre composition. It was concluded that the digestion of reed canarygrass, relative to ryegrass, was reduced by the compact arrangement of tissues and that alkaloid content did not influence the digestion of reed canarygrass cultivars.     

A NOTE ON THE SUITABILITY OF THE IN VITRO DIGESTIBILITY TECHNIQUE FOR VERY SMALL PLANT SAMPLES

The two-stage in vitro digestibility technique was scaled down to use 0.1 g samples. The accuracy of predicting in vivo DM digestibility by the use of micro-samples was studied, using 35 samples of four grass species of known in vivo digestibility. When the sample size was reduced from the normal 0.5 g to 0.1 g, the residual standard deviation of the regression relating in vitro to in vivo DM digestibility was increased from ±2.5 to ±3.4 digestibility units. Grinding the samples more finely than through a 1.0 mm screen did not improve the accuracy of predicting in vivo digestibility. It was concluded that the in vitro method may be used with micro-samples where necessary, but with less accuracy than the macro-technique. For maximum accuracy, the use of standards of known in vivo digestibility is necessary.     

THE USE OF RUMEN-FISTULATED STEERS IN THE DIRECT DETERMINATION OF NUTRITIVE VALUE OF INGESTED HERBAGE IN GRAZING EXPERIMENTS

A rumen-fistulated steer was used for the manual collection of samples of freshly swallowed herbage, in a grazing-management experiment on a perennial ryegrass sward. Diurnal and seasonal changes in the in vitro digestibility of the herbage selected by the grazing animal were studied under both strip- and continuous-grazing methods of management.
There was no appreciable change in digestibility as the sward was grazed down from upper to lower layers under strip-grazing management in April and May. In June to October a within-day fall in digestibility was found, much of which was attributable to an increase in the amount of old dead herbage grazed from the lower regions of the sward. Dead herbage taken in by the grazing steer was considerably lower in digestibility in August than in May. The in vitro digestibility of herbage samples, cut to ground level before and after grazing in a strip-grazed treatment, fell markedly as the proportion of dead herbage in the sample increased, giving a high negative correlation.
In a continuous-grazing management there was no pattern of diurnal variation, and the seasonal variation in digestibility of the ingested herbage was less than in strip grazing. The implications of these results are discussed in relation to indirect methods of digestibility determination (faecal-index technique), the measurement of herbage intake, and to some aspects of grazing management.     

A STUDY OF THE POSSIBLE ROLE OF CHELATION IN THE OCCURRENCE OF HYPOMAGNESAEMIA IN SHEEP

Wethers fitted with permanent rumen cannulae were used in a study of the effect of various doses of ammonium salts and E. D.T. A. (ethylene-diamine-tetra-acetic acid, disodium salt) on the pH of rumen digesta and the levels of serum calcium and magnesium. Doses of up to 200 g ammonium acetate had little effect on the pH of rumen contents or on serum Mg levels, but appeared to produce a fall in serum Ca. Ammonium carbonate, however, when given alone or with E. D.T. A., caused a marked increase in the pH of rumen fluid and a temporary fall in the serum Ca level. When given singly, neither ammonium carbonate nor E. D.T. A. significantly affected the levels of serum Mg which were, however, reduced when these compounds were given together, possibly because the increased pH of rumen contents produced by the ammonium carbonate was conducive to chelation of Mg by the E. D.T. A. High levels of ammonia may occur in rumen liquor after the ingestion of young nitrogen-rich herbage, and it is suggested that chelation of dietary Mg by such compounds as amino acids and peptides, which abound in young herbage, may be a factor concerned in the low 'availability' of this form of Mg.     

Comparison of two in vitro procedures using rumen liquor-pepsin or pepsin-cellulase for prediction of forage digestibility

Seventy-three samples of grass and legume herbage of known in vivo digestibility were used to compare an in vitro procedure comprising incubation with pepsin followed by cellulase (Jones and Hayward, 1975) with the Tilley and Terry (1963) procedure using rumen liquor followed by pepsin. The results obtained using the pepsin-cellulase method proved to be highly repeatable. On the basis of the residual standard deviation of the regression of in vivo dry matter digestibility (DMD) on dry matter digested in vitro (g/100 g DM), the pepsin-cellulase method was shown to be only slightly less accurate than the rumen liquor-pepsin method for combinations of both primary growth and regrowth of grasses (r.s.d. ±1.80 and ±1.46 respectively). However, the pepsin-cellulase method was noticeably less accurate than the rumen liquor-pepsin method when used on legume herbages (r.s.d. ±3.17 and ±1.91) or on grass and legume herbages combined (r.s.d. ±3.80 and ±1.60). For the rumen liquor-pepsin method one regression equation predicting DMD having the form, Y= 1.02 X? 0.41, was permissible. For the pepsin-cellulase method separate regression equations were necessary for grasses (species and growth stages combined) having the form, Y = 0.56 X+ 34.7, and for each legume species. It was concluded that the pepsin cellulase method was of value for grass samples, and, within species, for legumes because of its accuracy and precision and its speed and convenience relative to the rumen liquor-pepsin procedure; however, for legume species combined, and presumably for grass-legume mixtures, it was not sufficiently accurate.     

Bovine faeces as a source of micro-organisms for the in vitro digestibility assay of forages

The aim of this study was to investigate whether bovine faeces could replace rumen liquor collected from fistulated sheep in the determination of organic matter digestibility (OMD) of forages in vitro . The first experiment compared two inocula that had been prepared by mixing 83 and 333 g of cattle faeces with 1 litre of artificial saliva according to the first stage of the in vitro technique of Tilley and Terry ( Journal of the British Grassland Society, 18, 104–111, 1963). Inocula made from 333 g of faeces gave higher values of OMD than that prepared from 83 g of faeces and was used in subsequent studies. When a 48-h acid pepsin digestion, the second stage of the Tilley and Terry (1963 ) technique, was included, the OMD values and the ease of filtering of undigested residues were increased. The second experiment compared digestibility determined with inocula produced from faeces with that produced from rumen liquor. The OMD of a ryegrass hay determined with an inoculum prepared from faeces was 0·468 compared with 0·528 when using an inoculum prepared from rumen liquor. The third experiment investigated the accuracy of estimating the OMD of eight contrasting forages determined using rumen liquor from three sheep ( y ) and faeces from two cows ( x ). All regressions were significant ( P  < 0·001) and had residual standard deviations (r.s.d.) of between ±0·019 and ±0·022. The final study compared faeces collected from cattle fed with either hay or concentrates. The OMD of forage was higher from the hay-fed animals, but the r.s.d. values of the regressions were similar. It was concluded that cattle faeces shows potential as an alternative to rumen liquor collected from rumen-fistulated sheep for use in the in vitro digestibility assay of forages.     

Particle size reduction of the leaves of a tropical and a temperate grass by cattle

This study investigates forage particle breakdown, which is a factor of potential importance for the low intake of forages, particularly tropical grasses. A comparison was made between fresh leaf blades of a tropical grass, green panic ( Panicum maximum var. trichoglume ) and a temperate grass, Italian ryegrass ( Lolium multiflorum ). Leaf blades were chopped into 50 mm lengths and fed to cattle with oesophageal and rumen fistulae. Particle size was assessed on chewed samples immediately after eating and after digestion for 6, 12, 24, 48, 96 h and 3 weeks in nylon bags in the rumen.
Chewing during eating reduced particle size more in length than in width and more in green panic than in ryegrass. Mean length was decreased 9·2-fold and 4·7-fold for the two species respectively, and mean width 5·4-fold and 2·3-fold. Green panic leaf had a higher cell wall content and higher dry matter content than ryegrass leaf.
Digestion caused a substantial reduction in the width of the particles but not in the length. Width reduction occurred more rapidly in the ryegrass which was reduced to narrow fibres within 12–24 h of digestion while in green panic the same extent of degradation took < 48 h. After 96 h digestion, mean width of the chewed material had been reduced 40-fold in ryegrass compared with only 165-fold in green panic; all particles of both species were < 1 mm in diameter. The mean length of particles after 96 h digestion was similar to that of the chewed feed. Ryegrass was more digestible than green panic, a difference of 150 g kg DM⁻¹ which was maintained from within the first 6 h of digestion up to 96 h digestion. After three weeks the digestibility of ryegrass was c. 50 g kg DM⁻¹ higher than that of green panic but the proportion of cell wall that was ultimately indigestible was the same (0465) for both species.     

Age of regrowth as a factor affecting the nutritive value of hay of kikuyu grass (Pennisetum clandestinum) offered to lambs

The changes in chemical composition of hay of kikuyu grass ( Pennisetum clandestinum ) harvested at 50, 70 and 90 d of regrowth, and its effect on intake, digestibility, fermentation and microbial protein synthesis in the rumen (Experiment 1) and oxygen uptake by portal-drained viscera (PDV) (Experiment 2) were evaluated. The experiments were carried out with Polwarth × Texel crossbreed lambs with a mean live weight (s.e.m.) of 35(3·0) kg housed in metabolic cages. Organic matter (OM), neutral-detergent fibre (NDF) and nitrogen (N) intake, as well as rumen ammonia-N concentration, decreased linearly with age of regrowth ( P  <   0·05). Acid-detergent fibre (ADF) and indigestible NDF intakes were similar for all treatments. Apparent digestibility of organic matter (OM), NDF and N, as well as true digestibility of OM, microbial protein synthesis in the rumen, N retention, pH of rumen fluid and sugars, amino acids and peptide concentrations in rumen fluid were similar for all treatments. Age of regrowth also did not affect the kinetics of passage of the particulate phase through the digestive tract nor oxygen uptake by PDV. Heat production by PDV represented an average of 0·039 of digestible energy (DE) intake. Increasing the age of regrowth of kikuyu grass from 50 to 90 d did not affect digestibility nor the efficiency of DE use by PDV tissues of lambs but it reduced the nutritive value of the hay due to a lower intake of OM. Intake of hay appeared to be most limited by the ADF and indigestible NDF concentrations of the hay.     

THE ASSESSMENT OF THE NUTRITIVE VALUE OF SILAGE BY DETERMINATION OF IN VITRO DIGESTIBILITY ON HOMOGENATES PREPARED FROM FRESH UNDRIED SILAGE

A report is presented on studies on the relationship between in vitro digestibility of 18 silage samples and their in vivo digestibilities. Various analytical techniques are compared and the method adopted, on grounds of suitability for routine operation and accuracy of prediction of in vivo data, is based on analysis of homogenates of fresh silage. The homogenates, sufficiently uniform to permit of volumetric subsampling at acceptable error levels, are prepared in a new design of homogenizer. The impact of the % dry matter of the silage, as fed, on the accuracy of in vitrol in vivo relationships is considered and a bivariate regression plane is proposed.     

Intake and digestibility of components of forage rape (Brassica napus) by sheep

Herbage from forage rape ( Brassica napus cv. Lair), harvested in the autumn, was separated into different morphological components and offered ad libitum to weaned lambs and adult sheep in two indoor pen-feeding experiments. In vivo. digestibility and the voluntary intake of herbage components were measured and related to chemical composition and in vitro digestibility. The organic matter digestibility of all plant components, especially of petiole, was high; lamina, 0·847: petiole, 0·892; upper stem. 0·865; lower stem, 0·771. Lambs and adults digested 'leaf' (0·83) vs 0·846) or 'stem' (0·782 vs 0·789) equally well. Both lambs and adults ate much less of the components or of whole crop than would be predicted with grass crops of similar digestibility or fibre content. With the exception of lamina, intake was closely related positively to digestibility and negatively lo fibre content. Low intakes of lamina were associated with high concentrations of glucosinolates. In vitro digestibility values obtained on small samples of forage show that the process is adequate for the prediction of in vivo digestibility.     

THE USE OF CATTLE WITH OESOPHAGEAL FISTULAE IN GRASSLAND EXPERIMENTS

Oesophageal-fistulated calves and cattle were used in various experiments during 1963 to 1966; results for 1966 are reported.
The method of inserting the fistula, the routine care of the animals and the methods of sampling through the fistula, are described.
Individual stall-feeding experiments in 1966 gave recovery values of 99.7% for herbage extruded through the fistulae, in relation to herbage eaten. The in vitro organic-matter digestibility of the grass offered and of the extrusa samples were identical. The digestibility of the grass fed in these experiments was high, but was comparable with that of herbage used in the grazing experiments reported. Grazing studies on grass/white clover pastures are reported. The digestibility was determined on all samples and they were separated into various plant fractions. Results obtained were compared with herbage samples cut at the same time. In most cases the digestibility and % N of the extrusa sample was higher than that of the corresponding herbage sample. The quantity of dead material present increased, and more was eaten, as the season advanced; it was of low digestibility. Results of intake studies with grazing animals are presented and the use of oesophagea-istulated cattle as an aid in such studies is discussed.