Characterising Forages for Ruminant Feeding*

R. A. Dynes*, D. A. Henry and D. G. Masters CSIRO Livestock Industries, Private Mail Bag 5, Wembley, WA 6913, Australia ABSTRACT : Forages are the most important feed resource for ruminants worldwide, whether fed as pastures, forage crops or conserved hay, silage or haylage. There is large variability in the quality of forages so measurement and prediction of feeding value and nutritive value are essential for high levels of production. Within a commercial animal production system, methods of prediction must be inexpensive and rapid. At least 50% of the variation in feeding value of forages is due to variation in voluntary feed intake. Identification of the factors that constrain voluntary feed intake allows these differences to be managed and exploited in forage selection. Constraints to intake have been predicted using combinations of metabolic and physical factors within the animal while simple measurements such as the energy required to shear the plant material are related to constraints to intake with some plant material. Animals respond to both pre- and post-ingestive feedback signals from forages. Pre-ingestive signals may play a role in intake with signals including taste, odour and texture together with learned aversions to nutrients or toxins (post-ingestive feedback signals). The challenge to forage evaluation is identification of the factors which are most important contributors to these feedback signals. Empirical models incorporating chemical composition are also widely used. The models tend to be useful within the ranges of the datasets used in their development but none can claim to have universal application. Mechanistic models are becoming increasingly complex and sophisticated and incorporate both feed characteristics and use of biochemical pathways within the animal. Improvement in utilisation through the deliberate selection of pasture plants for high feeding value appears to have potential and has been poorly exploited. Use of Near Infrared Reflectance Spectroscopy is a simple method that offers significant potential for the preliminary screening of plants with genetic differences in feeding value. Near Infrared Reflectance Spectroscopy will only be as reliable as the calibration sets from which the equations are generated. (Asian-Aust. J. Anim. Sci. 2003. Vol. 16, No. 1 : 116-123) Key Words : Forage Evaluation, Feeding Value, Nutritive Value, Sheep, Near Infrared Reflectance Spectroscopy ** This paper was presented at an 2002 International Symposium on “Recent Advances in Animal Nutrition” held in New Delhi, India (September 22, 2002). *Corresponding Author: R. A. Dynes. Tel: +61-8-93336000, Fax: +61-8- 93878991, E-mail: Robyn.Dynes@csiro.au CHARACTERISING FORAGES FOR RUMINANT FEEDING 117 FORAGE QUALITY Forage value is a function of its contribution to animal performance (reproduction, meat, wool and milk). For grazing animals, the challenge is to exploit forages to optimise forage utilisation and/or to maximise the genetic potential of the individual animal to grow products of economic value. The majority of Australian forage is utilised in this ‘extensive’ grazing system. For feedlot animals, the challenge is to optimise intake of the forage on offer to maximise the genetic potential of the animal. This latter use of forages is increasing in Australia in response to demands from feedlot beef, intensive dairy industries and export fodder markets. Forages fed in situ ideally supply all essential nutrients for the grazing animal while forages fed in intensive industries are usually in the ration to supply fibre for the rumen. It is critical this distinction between feeding systems and the objective of using the forage is considered when evaluating forage quality. Feeding value and nutritive value are two terms commonly used to describe the quality or value of a forage for animal production (Ulyatt, 1973). Feeding value refers to animal production responses when feed available does not limit voluntary feed intake and is a function of voluntary feed intake and nutritive value. Nutritive value refers to the responses in animal production per unit of intake and is a function of digestibility of nutrients and the efficiency with which the nutrients are used for maintenance or production. The distinction between nutritive value and feeding value is an important one and Ulyatt has estimated that variation in voluntary feed intake accounts for at least 50% of the variation that is observed in feeding value of forages. Frequently the availability of a forage will limit voluntary feed intake in both extensive and intensive systems. Constraints to intake Ruminants grazing forage or fed in pen experiments usually fail to consume sufficient nutrients to meet the needs for maximum production and hence achieve their full genetic potential. It follows that identifying constraints to intake of forages will allow us to manage or overcome the limitation. Weston (1982, 1985, 1996) proposed that voluntary feed intake of forages was regulated by an interplay between the rate of clearance of dry matter from the rumen and the amount of useful energy that is available to the animal, relative to the animals capacity to use the energy. In this conceptual model, forage diets generally fail to provide sufficient energy to meet the capacity of the animal to use energy due to a number of constraints, including the resistance of forage organic matter to removal from the rumen, low diet palatability, difficulty in forage harvesting and prehension, and environmental stress. Consequently, an energy deficit relative to the capacity to use energy exists, which Weston quantified as the forage consumption constraint (FCC). FCC can be calculated as the difference between the quantity of forage that the animal actually consumes and the amount of that forage that the animal needs to consume to meet its capacity to use energy when any constraints are absent. In general, for both weaner sheep (Weston, 1996) and adult sheep (Weston and Davis, 1991), the forage consumption constraint tends to decrease as the energy content of the diet increases. Similarly, Forbes and Provenza (2000) recently proposed a minimal total discomfort model. Briefly, signals from various families of visceral receptors interacting at organs sensitive to metabolic fluxes, together with adipose tissue, social stimuli and environmental factors are integrated at the central nervous system in an approximately additive manner to generate a total signal of discomfort. The key factors in both models are potential for clearance of organic matter from the rumen and availability of essential nutrients. Crude protein, organic matter digestibility and metabolisable energy represent important classes of nutrients while neutral detergent fibre (NDF) indicates an index of bulk. Both models can provide a framework to assess the extent to which different factors constrain intake. In focusing on factors which constrain voluntary feed intake we have utilised the relationship between energy required to shear plant material and FCC to identify factors limiting voluntary feed intake. For example, forage consumption constraint was predicted from energy required to shear plant material (Baker and Dynes, 1999) for penned sheep fed different genotypes of dry mature subterranean clover (from Taylor et al., 1989) (Table 1). Energy required to shear plant material increased as voluntary feed intake decreased. However, constraint to intake for sheep fed Mt Barker and Mt Helena subterranean clover was significantly under that predicted from shear suggesting, some factor(s) other than resistance of the plant material to shear constrained intake of these cultivars. The maturity of forages significantly affects the energy required to shear material. For example, the energy required to shear subterranean clover, annual grasses and broadleaf weeds was between 6 and 10 KJ/m2 and increased as the plants matured (Dynes and Henry, 2002). Pasture species and time of sampling significantly affected energy required to shear the plant material however surprisingly there was no effect of level of feed on offer in the range 600 to 2500 kg dry matter per hectare, despite visual differences in sward structure. Predicted FCC was low (9 and 20 g OM/kg W 0.75) and requires validation with measures of intake in grazing animals to permit the comparisons presented in Table 1. 118 DYNES ET AL. Despite the importance of selection to nutrient intake by grazing animals, the mechanisms by which animals show discrimination among forages in their feeding behaviour are not well understood. Diet composition of sheep grazing forages in situ, commonly matches the botanical composition of the sward (Dynes et al., 1999). However, sheep will demonstrate strong selection pressure. Animals grazing an annual sward similar to those above but including saltbush (Atriplex spp.) will largely avoid the saltbush while any alternative forages are available (Norman et al., 2002). Further, sheep show preference within a single species of saltbush for some individual plants over others and this cannot be explained simply by soluble ash content or other nutritive value traits (H. Norman, unpublished). Palatability is a term defined as any characteristics of the feed which inhibits intake of forage whether the forage is offered alone or as a choice. If an animal rejects a forage then clearly that forage will be of reduced feeding value even if its nutritive value is high. We suggest the current use of palatability is too broad and may in part reflect the need to better understand factors which drive selection in animals and how components of the forage stimulate feedback signals. If we consider palatability within the conceptual model framework then feedback signals to the central nervous system can be grouped into pre-ingestive and post-ingestive feedback signals. Pre-ingestive feedback signals include taste, odour, and texture of forages, and in the long term also reflects social learning and learning developed through an aversion to post-ingestive responses to a nutrient or toxin (Provenza and Pfister, 1991). Post-ingestive signals incorporate those proposed by Weston (1996) and Forbes and Provenza (2000) and including tissue capacity to utilise energy and signals relating to rumen load and organic matter clearance from the rumen. In the case of saltbush, post-ingestive feedback from salt load would be expected but other factors influencing avoidance/selection of saltbush require identification (Masters et al., 2001). Pre-ingestive feedback may constrain intake of some forages. The magnitude of constraint may vary between and within plant species depending on the physiological state of the animal. For example, weaner sheep fed spring harvested, dried subterranean clover (Dynes, 1996) appeared to be constrained by pre-ingestive feedback signals. Intake of the subterranean clover was lower (Table 2) than animals fed optimal or energy limiting diets, despite there being no apparent limitation to gut load on the subterranean clover diet. Over several weeks intake increased gradually and did not respond to additional protein or energy supplements, suggesting individuals were ‘used to’ the negative signals. Our ability to predict the role of both pre and postingestive feedback signals in feeding value of forages is critical for high producing dairy cows. Here there are very high demands for maximum productivity and high dry matter intake. Similarly the introduction of novel forage species into farming systems requires careful evaluation for negative pre- and post-ingestive feedback signals. Predicting performance Forage evaluation evolved through the 20th century in parallel with increasing understanding of the factors, which drive animal performance (Reid, 1994). Research on voluntary feed intake, digestion and utilisation in the 1950- 1970’s (Crampton, 1957; Blaxter, 1962) together with the development of routine in vitro digestibility (Tilley and Terry, 1963), and fibre analysis methods (Van Soest and Wine, 1967) were fundamental cornerstones of modern forage evaluation. Mathematical modeling together with laboratory Table 1. Voluntary feed intake, dry matter digestibility and constraint to forage intake (FCC) for dry, mature subterranean clovers, (T. subterraneum), fed to adult wether sheep in pens Subterranean clover Dry matter intake (g/d) Digestibility of dry matter (%) Energy required to shear (KJ/m2 ) Constraint to forage intake (FCC) (g OM/kg BW0.75.d) Predicted constraint to forage intake (FCC) (g OM/kg BW0.75.d) Spencers Brook 381 48.6 23.8 96 105 Collie A 499 49.9 22.8 88 100 Mt Helena A 605 52.3 14.5 77 54 69S30.5.4.1 600 55.6 15.9 70 62 Mt Barker 732 58.0 12.3 61 42 Standard error 46.6 1.73 0.54 2.4 - (Baker and Dynes, 1999) Table 2. Daily organic matter intake (g/kg metabolic liveweight W0.75) and gut load (g OM/kg reticulo-rumen fleece free weight) of weaner sheep offered an optimal or energy limiting diet or spring harvested subterranean clover Optimal diet Energy limiting Sub clover Voluntary feed intake 67.5±0.75a 40±1.20b 31±0.75c Digesta OM/kg RFW 25±1.5a 39±2.4b 26±1.3a a Within a row, means significantly different p<0.001. (Dynes, 1996) CHARACTERISING FORAGES FOR RUMINANT FEEDING 119 analysis of forages currently underpins ruminant feeding systems. Many systems are based on a general positive correlation between voluntary feed intake and digestibility (Minson et al., 1964; Blaxter et al., 1966). However, this relationship is empirical and digestibility accounts for only about 60% of the variation in voluntary feed intake. Forages of similar digestibility can differ in voluntary feed intake and nutritive value, and forages of similar nutritive value can differ in feeding value. Further, the digestibility and nutritive value of forage can be altered by changing particle size (see Minson, 1990 for examples). Globally, three main systems are in common use for predicting requirements of ruminants, all are based on relationships between intake and digestibility: Australia (SCA, 1990) and the associated computer package (GrazFeed, Freer et al., 1997), United Kingdom (ARC, 1980; AFRC, 1990), USA (NRC, 1984, 1987). The NRC equations are built on a very large number of data sets while both ARC and SCA are developed from smaller data sets. In general, animal performance is well predicted within the digestibility range these equations were developed from, but no one system can claim universal application because of the empirical way in which they were established. Consequently, if the characteristics of a forage fall outside the range for forages used in the equation development then predictions will not be accurate. For example, Poppi (1996) used equations from AFRC, NRC and SCA to predict dry matter intake where animal weight or dry matter digestibility were varied for growing cattle fed tropical diets. He found that the accuracy of the predictions depended upon the liveweight of the animal, it’s body condition, and the diet being fed. SCA based on roughages predicted higher intakes at increasingly high digestibility whereas the ARC and NRC equations yielded similar intakes once the digestibility reached approx. 0.8, again reflecting the digestibility range from which they were developed. The use of plant characteristics that predict ‘constraints to intake’ of forages requires consideration. Weston and Davis (1991) in a study of 14 forages showed the relationship of FCC with either the biomechanical characters or the fibre composition of the forages were better than the relationships between voluntary feed intake and biomechanical characters or fibre. Energy required to comminute was positively correlated with FCC (R=0.96, Weston and Davis, 1991) as was energy required to shear (R=0.94, S. Baker, unpublished). These biomechanical characters are good predictors of FCC when other factors such as limiting amino acids do not constrain intake. In an extensive study of oaten hays fed to sheep, energy required to shear together with crude protein and NDF content of the hay were the best predictors of animal performance (Baker et al., 1998). In the USA, a hay grading system based on laboratory analyses was proposed to get around the costly and lengthy requirement of in vivo digestion trials, and the standardization issues related to in vitro digestibility techniques (Rohweder et al., 1978). The relative feed value index (RFV) was developed to rank cool season legumes, grasses and mixtures by an estimate of potential digestible dry matter intake, calculated from digestible dry matter and dry matter intake. The acid-detergent fibre (ADF) and NDF analyses described by Goering and Van Soest (1970) were the chemical assays of choice to estimate in vivo dry matter digestibility and dry matter intake respectively. Subsequent analyses have shown the relationship between dry matter digestibility and ADF to be variable even within tropical and temperate grass species, and NDF concentration to be unsatisfactorily variable for predicting dry matter intake in grasses (Moore et al., 1996; Moore and Kunkle, 1999). These conclusions are likely to reflect the requirement for different regression equations for different plant species and possibly even locations (Rohweder et al., 1978, see Table 3). Surprisingly, when 24 alfalfa hays were grown under irrigation at one location and fed to lambs, only 1% of the variation in actual intake was accounted for by the prediction of intake from NDF. Further, only 20% of the variation in measured digestible dry matter (DDM) was accounted for by the equation predicting DDM from ADF. Hay grading systems are evolving in Australia and a standard grading system now operates. Forage exports are dominated by long chop oaten hay, with 500,000 tonnes exported in the 2000 season (RIRDC, 2000). Japan is the largest export market followed by Korea, Singapore, Taiwan, Malaysia and UAE accounting for the balance (RIRDC, 2000). Hays are commonly selected for export using subjective criteria of colour, texture and taste. Oaten hays differ significantly in nutritional parameters and vary due to a number of factors including season, growing location, drying time and weather damage. Oaten hays from a single processor in eastern Australia selected for export based on similar colour and texture were highly variable in all characteristics measured (Table 4). Expansion of Australia’s market share requires the development of objective standards for oaten hays that will predict the value of the forage in the mixed ration. These apparent weaknesses in the existing systems reflect gaps in our understanding of the interactions between the physiology, biochemistry and nutrition of the animal rather than a failure of forage evaluation. Table 3. Variation in linear correlations (R) for ADF with DDM and for NDF with DMI Sample origin ADF:DDM NDF:DMI Alfalfa-3 locations 0.83 - 0.91 0.44 - 0.75 Various grasses, various locations 0.73 - 0.92 0.43 - 0.94 (Rohweder et al., 1978) 120 DYNES ET AL. Mechanistic models Mechanistic models may play an increasing role in predicting performance of ruminants. Mechanistic models are based on theory and relate processes and responses at different hierarchical levels that include whole animal and tissue biochemistry (Beever et al., 2000). Mechanistic models analyse the whole system in terms of key components and their interactions with each other (Beever et al., 2000). The models commonly have a complex range of inputs for forage and digestive parameters such that they remain principally a research tool. At least 4 distinct areas must be considered in reviewing mechanistic modelling of nutrient intake, digestion, utilisation and output of meat, wool or milk. These include, intake regulation (Poppi et al., 1994), rumen function (for example Baldwin et al., 1987; Dijkstra et al., 1992), metabolism in the lactating cow (Baldwin et al., 1987b) and metabolism in the growing animal (Gerrits et al., 1997a,b). The success of mechanistic models in industry will depend both on the accuracy of prediction of animal performance and the inputs being measured (predicted) accurately, rapidly and inexpensively. The prediction of shear from Near Infrared Spectroscopy (NIRS) has the potential as a plant characteristic to predict animal performance directly (Baker et al., 1998) or as an input into mechanistic models if ongoing research supports its role as a predictor of ease of particle size reduction in the rumen. Plant selection and breeding The major criteria for selection of improved forage varieties have been optimisation of yield and digestibility along with the need to minimise disease susceptibility and maintain other agronomic traits (Beever, 1993). Relatively little regard has been given to the deliberate selection and breeding of plants for other components of feeding value for ruminants including potentially problematic secondary compounds. This is despite pasture and fodder plants ultimately being required to fit within a grazing or fodder conservation system where the final ‘consumer’ is the animal. Further, because of the risk of litigation, duty-of-care issues are now considered extremely important prior to the release of new varieties. This is particularly important where new varieties are developed from germplasm with little background information on feed quality and the presence of secondary compounds is possible. However selecting ‘safe’ varieties is not necessarily selecting highly productive varieties. Collaboration between plant breeders and animal scientists is essential to ensure the ‘best’ species are selected. There is an opportunity to develop new forage varieties with vastly improved (and known) feed quality characteristics, and to lessen the risk of litigation by developing rapid, accurate and inexpensive screening methodologies that can be utilised in the selection and breeding processes. Accessions from a novel breeding program for perennial legumes in Western Australia (Table 5) reflects the substantial variation in some nutritive traits that may be exploited in conventional plant breeding and selection programs. Obviously before such traits can be incorporated into a program, correlations with other agronomic and morphological characteristics must be established. The range in both in vitro digestibility of dry matter (IVDMD) and tannin content are important since some accessions have IVDMD and tannin contents which would significantly limit voluntary feed intake and/or rumen function and constrain animal performance. Our challenge in forage evaluation is to identify plant characteristics which can be used alone or as inputs to models to predict the value of a forage for animal feed. NIRS offers significant potential to be used in selection and breeding programs. It is inexpensive, rapid and requires only small amounts of plant material. However given the potential novelty and diversity of some ‘new’ plant genotypes, extreme care is necessary to ensure that samples are adequately represented in the calibration set. Not only is the validity of the calibration set important but also the suitability of laboratory methods to provide accurate measures. Extreme caution is required in use of in vitro derived predictions of biologically important components for example fermentable metabolisable energy, where little validation with in vivo measurements exists (Beever and Moulds, 2000). Further, it is unlikely that some specific secondary compounds such as betaines, coumerins, oxalates and nitrates will be able to be measured using NIRS and traditional wet chemistry laboratory techniques will be Table 4. Average, range and coefficient of variation (CV) moisture content, energy required to shear (Shear energy kJ/m2 ) plant material, in vitro digestibility (IVD), acid detergent fibre content (ADF), neutral detergent fibre content (NDF), soluble carbohydrate (SC) and crude protein (CP) of export oaten hay produced in New South Wales, Australia in 1999/2000 growing season Moisture content (%) Shear energy (kJ/m2 ) IVD (% DM) ADF (% DM) NDF (% DM) SC (% DM) CP (% DM) Mean 9 14 57 36 59 18 6 Range 3-15 10-19 46-63 28-46 52-72 1-30 3-9 CV 22.1 12.9 4.8 9.2 7.2 31.2 16.5 CHARACTERISING FORAGES FOR RUMINANT FEEDING 121 required, with final selection supported by animal grazing studies. THE FUTURE Forage evaluation will be increasingly a component of both intensive and extensive farming systems. Improving characterisation of forages must maintain focus on identifying the factors that are important for animal production and the levels of precision needed for practical application. Dewhurst and Webster’s (1989) summary remains highly relevant: robust measurements suitable for routine use, adequate description of processes of metabolism and digestion. Finally we must provide a better prediction of nutritive value and ultimately animal performance than is currently available. The challenge in intensive systems is to deliver nutrient based systems of feed characterisation. This requires comprehensive carbohydrate and protein characterisation with due recognition to all significant nutritional entities which are likely to have different metabolic fates (Beever et al., 2000). While sugar could be described adequately as one entity, descriptions of starch and fibre would need to represent potential ruminal degradability (both rate and maximal extent). Crude protein would need to be accounted for by content of true protein, peptides and amino acids and ammonia. Extensive farming systems will demand inputs for precision farming systems. Opportunities now exist to measure forage quality remotely, whether by vehicle mounted NIRS-type instruments or by airborne and satellite-borne sensors. Research is now underway to develop hyperspectral imagery for the measurement of protein content and digestibility in forages. Ultimately these will require integration into models of ruminant requirements for allocation of forage resources. REFERENCES AFRC. 1990. Nutritive requirements of ruminant animals: energy. Technical Committee on Responses to Nutrients. Report No 5. Nutr. Abstr. Rev. (Series B) 60:729-804. AFRC. 1991. Voluntary intake of cattle. Technical Committee on Responses to Nutrients, Report No 8. Nutr. Abstr. Rev. (Series B) 61:815-823. Agricultural Research Council. 1980. The Nutrient Requirements of Ruminant Livestock. CAB International. Wallingford, UK. Baker, S. K. and R. A. Dynes. 1999. Evaluation of the feeding value of pasture legumes. In: Genetic Resources of Mediterranean Pasture and Forage Legumes (Ed. S. J. Bennett and P. S. Cocks). Proceedings of an International Workshop on the Genetic Resources of Plant and Forage Legumes in Mediterranean Environments. Kluwer Academic Publishers. Netherlands. pp. 120-131. Baker, S. K., R. A. Dynes, D. A. Henry and D. B. Purser. 1998. Objective specification of the quality of hays and fodder. Rural Industries Research and Development Corporation Report. Canberra, Australia. Baldwin, R. L., J. France and M. Gill. 1987a. Metabolism of the lactating cow. I. Animal elements of a mechanistic model. J. Dairy Res. 54:77-105. Baldwin, R. L., J. H. M. Thornley and D. E. Beever. 1987b. Metabolism of the lactating cow. II. Digestive elements of a mechanistic model. J. Dairy Res. 54:107-131. Beever, D. E. and F. L. Mould. 2000. Forage evaluation for efficient ruminant livestock production. In: Forage Evaluation in Ruminant Nutrition (Ed D. J. Givens). CABI Publishing. Wallingford, UK. p 15-43. Beever, D. E., J. France and G. Alderman. 2000. Prediction of response to nutrients by ruminants through mathematical modelling and improved feed characterization. In: Feeding Systems and Feed Evaluation Models (Ed. M. K. Theodorou and J. France). CAB International Publishing. New York, Table 5. Average (and range) in in vitro dry matter digestibility (IVDMD), crude protein (CP) and total tannin content of some alternative perennial pasture legumes harvested in June and October 2000 June October Species name Common name IVDMD (%) CP (%) Tannin content (%) IVDMD (%) CP (%) Dorycnium hirsutum Hairy canary clover 52.3 (47 - 55) 12.9 (11 - 14) 3.87 (1.8 - 6.7) 55.4 (49 - 61) 13.6 (6 - 19) Hedysarum coronarium Sulla 69.8 (65 - 74) 17.3 (17 - 18) 3.02 (1.3 - 4.7) 68.7 (66 - 70) 15.3 (13 - 18) Lotus corniculatus Birdsfoot trefoil 67.1 (50 - 75) 15.7 (7 - 20) 1.80 (0a - 6.1) 66.7 (51 - 77) 16.5 (6 - 21) Lotus tenuis Narrow-leaf trefoil 72.8 (67 - 77) 17.5 (12 - 20) 0.26 (0a - 0.7) 69.4 (55 - 77) 18.1 (11 - 24) Melilotus officinalis Yellow sweet clover 72.9 (69 - 76) 14.5 (11 - 17) 0.07 (0.02 - 0.15) 75.8 (74 - 79) 19.6 (7 - 29) Trifolium fragiferum Strawberry clover 74.9 (72 - 79) 21.7 (19 - 25) 0.15 (0.05 - 0.25) 76.2 (75 - 78) 17.9 (17 - 19) Trifolium hybridum Alsike clover 77.8 (74 - 82) 18.8 (15 - 21) 0.08 (0.05 - 0.11) 75.5 (76 - 80) 19.4 (12 - 27) a Tannin content undetectable (less than approx 0.01 %). (Henry et al., 2002) 122 DYNES ET AL. Wallingford. pp. 275-297. Beever, D. E. 1993. Ruminant animal production from forages – present position and future opportunities. Proceedings of the XVII International Grassland Congress. Palmerston North, New Zealand. pp. 535-542. Blaxter, K. L. 1962. The Energy Metabolism of Ruminants. Hutchinson, London. Blaxter, K. L., F. W. Wainman and J. L. Davidson. 1966. The voluntary feed intake of food by sheep and cattle in relation to their energy requirements for maintenance. Anim. Prod. 8:75- 83. Crampton, E. W. 1957. Interrelations between digestible nutrient and energy content, voluntary dry matter intake and the overall feeding value of forages. J. Anim. Sci. 16:546-552. Dewhurst, R. J. and A. J. F. Webster, 1989. Development of a practical deterministic model for the prediction of true metabolizable energy in forages and compound feeds. In: Energy Metabolism of Farm Animals. European Association of Animal Production Publication No. 43. (Ed. Y. Van der Honing and W.H. Close) Pudoc, Wageningen, The Netherlands, pp. 223-230. Dijkstra, J., H. D. St. C., Neal, D. E. Beever and J. France. 1992. Simulation of nutrient digestion, absorption and outflow in the rumen: model description. J. Nutr. 122:2239-2256. Dynes, R. A. 1996. The animal; options for managing intake. Aust. J. Agric. Res. 47:277-87. Dynes, R. A. and D. A. Henry. 2002. Biomechanical properties of plant species in annual pastures grazed to different levels of feed on offer. Anim. Prod. Aust. (in press). Forage and Grazing Terminology Committee. 1991. Terminology for Grazing Lands and Grazing Animals. Pocahontas Press. Blacksburg, Virginia. Forbes, J. M. and F. D. Provenza. 2000. Integration of learning and metabolic signals into a theory of dietary choice and food intake. In: Ruminant Physiology Digestion, Metabolism, Growth and Reproduction. (Ed P. B. Cronje). CABI Publishing. Wallingford, UK. pp. 3-19. Freer, M., A. D. Moore and J. R. Donnelly. 1997. Grazplan: Decision support system for Australian grazing enterprises-2. The animal biology model for feed intake, production and reproduction and the GrazFeed DSS. Agric. Syst. 54:77-126. Gerrits, W. J. J., J. Dijkstra and J. France. 1997a. Description of a model integrating protein and energy metabolism in preruminant calves. J. Nutr. 127:1229-1242. Gerrits, W. J. J., J. France, J. Dijkstra, M. W. Bosh, G. H. Tolman and S. Tamminga. 1997b. Evaluation of a model integrating protein and energy metabolism in pre-ruminant calves. J. Nutr. 127:1243-1252. Goering, H. K. and P. J. Van Soest. 1970. Forage fiber analysis (apparatus, reagents, procedures, and some applications). Agric. Handbook No 379, US Department of Agriculture, Agricultural Research Service. Henry, D. A., G. Moore and R. A. Dynes. 2002. Variability in nutritive value allows better selection of perennial pasture legumes. Anim. Prod. Aust. (in press). Masters, D. G., H. C. Norman and R. A. Dynes. 2001. Opportunities and limitations for animal production from saline land. Asian-Aust. J. Anim. Sci. 14, Special Issue, 199- 211. Minson, D. J. 1990. Forage in Ruminant Nutrition. Academic Press. London, UK. Minson, D. J., C. E. Harris, W. F. Raymond and R. J. Milford. 1964. The digestibility and voluntary feed intake of S22 and H.I. ryegrass, S170 tall fescue, S48 timothy, S215 meadow fescue and Germinal cocksfoot. J. Br. Grassl. Soc. 19:298-305. Moore, J. E., J. C. Burns and D. S. Fisher. 1996. Multiple regression equations for predicting relative feed value of grass hays. Proc. American Forage and Grassland Council, June 13- 15, Vancouver, BC AFGC, Georgetown TX. Moore, J. E. and W. E. Kunkle. 1999. Evaluation of equations for estimating voluntary intake of forages and forage-based diets. J. Anim. Sci. Suppl. 77:204. Norman, H. C., D. G. Masters, R. A. Dynes, D. A. Henry and M. J. Lloyd. 2002. Liveweight change and wool growth in young sheep grazing a mixed saltbush and balansa clover pasture. Anim. Prod. Aust. (in press). National Research Council. 1984. Nutrient Requirements of Beef Cattle 6th Edition. National Academy of Sciences. Washington, DC. National Research Council. 1987. Predicting Feed Intake of FoodProducing Animals. National Academy of Sciences. Washington, DC. Poppi, D. P. 1996. Predictions of food intake in ruminants from analyses of food composition. Aust. J. Agric. Res. 47:489-504. Poppi, D. P., M. Gill and J. France. 1994. Integration of theories of intake regulation in growing ruminants. J. Theor. Biol. 167:129-45. Provenza, F. D. and J. A. Pfister. 1991. Influence of plant toxins on food ingestion by herbivores. In: Recent Advances on the Nutrition of Herbivores (Ed. Y. W. Ho, H. K. Wong, N. Abdullah and Z. A. Tajuddin). Malays. Soc. Anim. Prod. Kuala Lumpur. pp. 199-206. Reid, R. L. 1994. Milestones in forage research (1969-1994). In: Forage Quality, Evaluation and Utilization (Ed. G. C. Fahey Jr). Based on the National Conference on Forage Quality, Evaluation and Utilization. University of Nebraska, Lincoln. 13-15 April 1994. pp. 1-58. RIRDC. 2000. Fodder Industry Atlas. Rural Industry Development and Investment Corporation. Canberra, Australia. Rohweder, D. A., R. F. Barnes and N. Jorgensen. 1978. Proposed hay grading standards based on laboratory analyses for evaluating quality. J. Anim. Sci. 47:747-759. Standing Committee on Agriculture (SCA). 1990. Ruminants, Feeding Standards for Australian Livestock. CSIRO Publications, Australia. Taylor, G. B., R. C. Rossiter, L. Klein and W. J. Collins. 1989. The intake and digestibility of dry mature subterranean clovers. Proc. XVI International Grasslands Congress. Nice, France. pp. 809-810. Tilley, J. M. A. and R. A. Terry. 1963. A two-stage technique for in vitro digestion of forage crops. J. Br. Grassl. Soc. 18:104-111. Ulyatt, M. J. 1973. The feeding value of herbage. In: Chemistry and Biochemistry of Herbage (Ed. G. W. Butler and R. W. Bailey). Academic Press. London, UK. pp. 131-178. Van Soest, P. J. and R. H. Wine. 1967. Use of detergents in the analysis of fibrous feeds. IV. Determination of plant cell wall constituents. J. Assoc. Offic. Anal. Chem. 50:50-55. Weston, R. H. 1982. Animal factors affecting voluntary feed intake. CHARACTERISING FORAGES FOR RUMINANT FEEDING 123 In: Nutritional Limits to Animal Production from Pasture (Ed. J. B. Hacker). Commonwealth Agricultural Bureaux. Slough. pp. 183-198. Weston, R. H. 1985. The regulation of feed intake in herbage-fed ruminants. Proc. Nutr. Soc. Aust. 10:55-62. Weston, R. H. 1996. Some aspects of constraint to forage consumption by ruminants. Aust. J. Agric. Res. 47:175-198. Weston, R. H. and P. J. Davis. 1991. The significance of four forage characters are constraints to voluntary feed intake. Proc. of the III International Symposium on the Nutrition of Herbivores. Selangor Darul Ehsan, Malaysia. p. 31. Wilson, J. R. and P. M. Kennedy. 1996. Plant and animal constraints to voluntary feed intake associated with fibre characteristics and particle breakdown and passage in ruminants. Aust. J. Agric. Res. 47:199-226. Wilkins, R. J. 2000. Forages and their role in animal systems. Forage Evaluation in Ruminant Nutrition. (Ed. D. I. Givens, E. Owen, R. F. E. Axford and H. M. Omed). CAB Publishing.