Abstract:
EFFICIENCY OF FEED UTILISATION BY RUMINANTS R.A. LENG* I INTRODUCTION l Metabolisable energy content of a feed or its components has been used to assess its feeding value for ruminants. The demonstration that VFA energy could, at times, be used highly inefficiently when included in a roughage diet, led to the suggestion that feeds for ruminants should be evaluated in terms of their net energy. The establishment of calorimetry technology in nutritional research was a direct result of the apparent need to measure heat production in animals to develop systems of feed evaluation based on net or metabolisable energy. The formation in 1970 of 'The Feeding Stuffs Evaluation Unit' at the Rowett Research Institute was the forerunner of the attempts to develop such a system of feeding standards (see Blaxter 1982). These systems have been developed with a narrow group of *common' feeds with high digestibilities and usually high in protein. However, even on such feeds the relationship between measured metabolisable energy values and those calculated from chemical compositions, are poor. This led Blaxter (1982) to end a lecture given to the annual Nutrition Conference for Feed Xanufacturers at Nottingham in the following way. *'The two *best' equations (relating observed and computed XEs) based on an entirely chemical approach do not seem as satisfactory as one based on physical me.asure of heat of combustion and the biolo@cal one of in vitro digestibility'. Since this time there has been a growing questioning of the accuracy and usefulness of prediction of feed requirements based on analysis and a calculated metabolisable energy values of a feed. This is emphasised nuch illore where, low digestibility forages are being evaluated or where the husbander has little control over selection of the diet or, as in the case of grazing animals, has little knowledge of what an animal eats. More recently it has been suggested that the balanced nature of the nutrients absorbed are more important in determining intake and therefore nutritive value of a forage than its metabolisable energy content (Preston & Leng 198% Since the early 1970% there has been substantial research which showed that the utilisation of metabolisable energy of a diet is variable depending on its ingredients, The early observations of high heat increment when additional acetate was given to ruminants on top of a basal diet, have generally not been substantiated. The major conclusion is that the efficiency of utilisation of acetate is diet specific. grskov and Allan (1966) and Bull et al. (1970) showed the efficiency with which acetate was used for growth in sheep and cattle was not below expectation when added to a 'high quality*! diet. Over the last few years there has been considerable research effort to explain the differences between the early calorimetric work (see Blaxter 1962) and aore recent studies (Hovel1 et al. 1976; P(rskov et al. 1979) on the efficiency of utilisation of metabolisable energy of VFA for body gain. This paper reviews the literature on the efficiency with which ruminants utilise VFA energy and strongly supports the concept that the level of glucose availability is critical for the efficient utilisation of the products of digestion for growth, milk production and reproduction. :* Department of Biochemistry, Microbiology and Nutrition, University of New England, Armidale, New South Wales 2351, Australia That the level of glucose availability is a major factor influencing the efficiency of utilisation of nutrients by animals is a contentious issue. For this reason a wide cross-section of research is taken to illustrate that the efficiency of utilisation of metabolisable energy by ruminants is dependent on both the quantities of glucose and amino acids and perhaps long chain fatty acid available in relation to the physiological state of the anialal. The overall concept is not new and there have been notable supporters in the past including Professor Max Klieber, who stated in 1960 that 'metabolisable energy is not a homogeneous entity, instead it represents an assembly of nutrients or metabolites each of which is used with a specific efficiency for a particular productive purpose' (see Kronfeld 1982). In addition Linzell (1957) suggested that the rate of milk production was primarily effected by the glucose availability and Kronfeld (1976) suggested that milk yield is a function of glucose uptake by the mammary gland whereas the efficiency of milk production is a function of the uptake of long chain fatty acids. Leng and Preston (1976) postulated that the major limitation to growth and milk production in cattle fed sugar cane based diets was the availability of glucose. The evidence that glucose, long chain fatty acids and amino acids are essential for efficient utilisation of feed for gain and milk production is discussed below. II l RUMINANT PRODUCTIVITY - BACKGROUND Productivity of ruminants is a function of digestible feed intake and the efficiency with which the absorbed nutrients are used for productive functions. These are determined by: 23 4e The capacity of the feed to support an efficient microbial inilieu in the rumen. The capacity of the diet to supply the quantities and balances of nutrients required by the animal in different productive states. d A primary consideration for maximising intake is that the rate of fermentative digestion and the efficiency of microbial growth in the rumen is optimised. A deficiency of an essential nutrient for microorganisms will be reflected in (1) a lowered rate of digestion of feed; (2) a lower than optimum microbial growth efficiency in the rumen expressed as Y-ATP (g dry cells produced(mole ATP available) and (3) a lowered feed intake. The capacity of a feed to supply the quantities of nutrients required in proportions that are balanced to meet a particular productive function depends on: 8 d % Its potential digestibility and potential rate of digestion Its ability to support an efficient rumen, leading to a high microbial cell production relative to VF'A and a high propionate production rate relative to acetate and butyrate. Its ability to provide additional dietary materials which bypass the rumen fermentation to balance the nutrients arising from fermentative. digestion to meet the requirements of production. The requirement for nutrients are highest in lactating and pregnant anirilals in the last Z-3 weeks before term, they are lower in growing anitilals and lowest with animals that are mature. Draught animals represent a special case but where they are fed on 'low quality diets *' they may also be limited at tivces by the balance of absorbed nutrients (see Leng 198% The availability of nutrients from a diet The nutrients available for metabolism are: ?3 The products of fermentative digestion: 3 VFA which include acetate, propionate, butyrate and higher fatty acids the digestible components of microbes synthesised in the rumen (which are about 60% protein), 3 ?# The dietary nutrients escaping the rumen and available for digestion in the small intestine, including: 3 bypass protein and starch providing amino acids and glucose 2 long chain fatty acids from dietary lipids. Essentiality of nutrients for production It is perhaps not suprising that nutrients that appear to be critical for efficient production in ruminants (essential amino acids, glucose and long chain fatty acids) are the same as the principal nutrients present in milk. For the young animal (considering the extremely high growth rates that can be achieved), milk probably provides the best balance of nutrients for growth (ie. they are highly efficiently used). Milk is also an effective supplement for young ruminants on solid food since suckling stimulates the oesophageal groove reflex and directs milk into the intestines. Reauirement for dietarv bypass protein Although Virtanen (1966) was able to obtain considerable milk from cows fed semi-purified diets in which the only nitrogen was in the form of urea, production was not high and could be stimulated by feeding protein. In their studies, Virtanen and his colleagues also found it necessary to feed high levels of grain-starches and apparently some oil (fat) was given; both these components may have lowered the need for gluconeogenesis from amino acids (see later) sparing them for milk production. It has been hypothesised that the effects of feeding a protein meal per se in dairy cow diets may be two-fold: 8 On high quality diets there is an apparent increase in diet or fibre digestibility in the rumen (Oldham et al. 1985); this has been attributed to the continuous availability of amino acids and peptides which act as microbial growth stimulants. However, there is a possibility that urea simply increased rumen amnonia concentrations above a critical level which was limitiw microbial growth. Urea supplementation9 if it increased rumen ammonia levels could have stimulated the rate of starch digestion (Mehrez et al. 1977) which decreased rumen pH, and which in turn depressed cellulolytic activity where urea was the sole supplemental N source (see Terry et al. 1969; Mould et al. 198% it Supplementary protein provides post-ruminal amino acids for absorption. Simple calculations (Burroughs et al. 1971) indicate that body growth requires more amino acids, relative to energy3 than are produced in the rumen and a supply of dietary protein to the intestines is required to optimise milk production. j&skov (1970) summarised the situation in terms of availability of amino acids and energy and the ability of the microbial system in the rumen to provide these. This is shown in Fig. 1. The microbial ecosystem can supply a suitably balanced protein to energy ratio in the products of fermentative digestion to support maintenance and the first two thirds of pregnancy but not noderate to high growth nor growth of the pregnant uterus close to term or moderate to high levels of milk production. A deficiency of absorbed dietary amino acids for production will often limit feed intake and the efficiency of feed utilisation particularly in highly productive animals such as cattle in the last 60 days of pregnancy (Lindsay et al. 1982). Roy et al. (W7) have attempted to describe requirements of the ruminants for amino acids in terms of the amounts of microbial and dietary protein that became available on a particular diet. This is based on a standardised and calculated microbial growth efficiency and a constant value for protein escape from the rumen for various feed components. This appears to be a major oversimplification, where these predictions of protein availability are applied to a variety of diets based on crop residues through to grain-based concentrates and highly digestibile forages such as lucerne. This is discussed further below. Microbial arowth efficiencv and the reouirements for dietarv bvDass r>rotein Undoubtedly microbial growth efficiencies are variable depending on numerous factors, The factors that have been identified as being of major importance include: * Deficiencies of microbial nutrients such as ammonia, sulphur, amino acids and peptides. Microbial species present in the rumen. In particular a large population of protozoa decreases the protein yield to the intestines (see Bird and Leng 1985, Veira et al. 1984). Frequency of feeding. Hespell (1979) showed that rumen bacteria when starved, quickly lysed (some within 2 hours). In a mixed diet, soluble sugars are rapidly fermented and the organisms responsible may therefore grow and a proportion lyse according to availability of these soluble sugars. Feeding a concentrate diet six times daily as against once daily doubled the microbial growth yield from the rumen (Tamminga 1981). Rumen fluid outflow rate appears also outflow from the rumen. Thus in cold lactation, digesta outflow rates from and microbial growth yield appears to and Milligan 1978). to %ontrolTT microbial cell stress, late pregnancy or the rumen appear to increase increase (Weston 1979; Kennedy 3 3 * % Physiological state, as well as effecting rumen microbial growth yields through the effects on digesta flow from the rumen, may also effect the overall digestibility of microbial and dietary bypass protein through increases in the capacity of the small intestine and increased enzyme secretions (see Oldham 1984). Thus the reported values for digestibility of microbial protein in the intestines have varied from 70-952 (see Leng and Nolan 1984). The above factors indicate the uncertainty about the efficiency of microbial growth and therefore the availability from fermentative digestion of protein relative to energy (P/E). The theoretical variability of P/E with microbial efficiency is shown in Fig. 2. The point stressed here is that the theoretical values for the P/E ratio from fermentative digestion can change from about 10:l to over 4O:l (g protein/MJ available energy) (Preston and Leng 1985). Calculation of the ratio of protein/energy available for utilisation by ruminants, taking into account all the variables, clearly is not simply modeled. III. REQUIREMENTS FOR GLUCOSE Glucose is required for a number of essential change with physiological state in a similar pattern of amino acid requirements (see Fig. 1 c.f. Fig. 3) the present time there are no analysis of feed that availability of glucose precursors from a feed given functions. Requirements to the changing pattern (Leng et al. 1977). At indicate the likely to ruminants. It is the difficulty of predicting glucose requirements, the physiological and nutritional factors that effect this requirement and the esttilation of glucose availability that appears to jeopardise systems of feed evaluation based on metabolisable energy. This is an area in which there is considerable controversy and therefore the evidence for glucose being essential is discussed in more detail. Background: Most dietary carbohydrates are fermented to VFA and microbial cell components in the rumen with the result that glucose absorption is uinimal in rudnants. Glucose is largely synthesised from propionate (Leng et al. 1967), but amino acids may also contribute to gluconeogenesis (Lindsay 1980). In cattle on certain high grain diets appreciable, but variable, quantities of starch escape fermentative digestion (Waldo 1973). Ha&e, sorzhu,ril and rice appear t o have a capacity to escape microbial activity in the ruGlen, and therefore to be digested in the small intestines and absorbed as glucose when included in a diet for ruminants. Small amounts of starch abso rbed in this way could be detrimental whe re glucose availability is prec ariously bal anced since it could actu ally suppress glucoeogenesis (see Leng 1970). Although some of the amino acids arising from digestion of dietary and microbial proteins can be converted to glucose they probably represent a minor source of glucose on diets that support a high propionate fermentation (e.g. grains) in the rumen. Under certain conditions, however, where glucose requireuents are high (as in fattening animals on diets low in fat) and/or where propionate production in the rumen is low, amino acids could be used extensively for glucose synthesis. Propionate is the major precursor of glucose and probably accounts for 80-90X of the glucose synthesised in maintenance fed animals (Cridland 1984). The requirement for minimum quantities of glucose for efficient use of absorbed nutrients is inferred from a number of different research approaches. The rilajor ones are summarised below. Evidence from growth studies Forage diets: Corbett et al. (1966), in a classical calorimetric study, demonstrated that in Scotland the spring pasture was more efficiently used for fat synthesis by sheep than that grown in the autumn. In the context of this discussion important differences were the higher propionate concentration in tine WAS in the rumen of animals on spring pasture. The energy retention of sheep . on diets of spring or autumn pasture is shown in Fig. 4. Recently MacRae et al. (1985) have demonstrated that an infusion of casein (30 g/d) per abomasum into sheep given autumn harvested dried grass at a feeding level equivalent to maintenance or 1.5 times maintenaxe increased the efficiency of utilization of the herbage from Kg 0.45 to Kg ;57 and those authors suggested that this response was a result of the increased availability of glucose synthesised from the amino acids. Thomson (1978) found that the efficiency of utilization of ME for bodyweight gain in cattle was higher for a concentrate/forage diet based on lnaize and forage (clover or grass) than for barley and forage even though the metabolizability of the DM (ME/DM) was the same on all diets (Table 1). One explanation is the proportionately greater post-ruminal digestion of maize compared with barley. Clover, as compared to grass based diets, was also more efficiently used by the animal. There is higher protein (more bypass) and generally there appears to be a higher fat content in legumes than in grasses indicatin g the likely relative nutrient availabilities on the two diets. In addition, maize contains up to twice the content of oil found in barley. 0, spring herbage 0, autumn herbqge CP 4, fasting heat production CF 30 29 The composition of the pastures is shown below: spring pasture autumn pasture 13 16 EE 2.5 3.3 6.7 9.5 Ash NFE 48 43 VFA properties in the rumen were as follows: AC 0 Prop. But. spring pasture autumn pasture 66 72 24 17 7 8 Higher Acids 3 3 Fig. 4. Relationship between metabolisable energy intake and energy retention (kca1/24 hr) (after Corbett et al. 1966). Efficiency of utilization of metabolizable energy for fattening cattle (kf) according to the nature of the ingredients in the diet. The ration combinations containing maize grain and clover forage were used more efficiently than those containirq barley . grain or ryegrass forage (from Thomson 1978) Tabie 1. Further evidence for an inefficient utilization of feed being related to propionate concentrations in the rumen arises from the studies of Tudor and Xinson (1982) which showed that pangola grass was used more efficiently than setaria grass for tissue synthesis in sheep even though both were fed at the sam rates and had the same apparent digestibility (Table 2). The authors zlention higher concentrations of propionate in rumen fluid of sheep on payola as one possible explanation for the difference. It is also feasible that microbial growth efficiencies were higher on the pangola grass diets or ti?at the grasses differed in the quantities of fat they contained. ' Table 2. The metabolizable energy in pangola grass is efficiently for fattening sheep than is the grass, even though both grasses had the same metabolizable energy in the dry matter (from 1982) used more energy in setaria concentration of Tudor and i3inson Undoubtedly acetate given in quantities representing a significant proportion of the ruminant's intake is used with a variable efficiency. Acetate given to a fasting animal or when the basal diet is roughage is **burnt off' to some extent since heat increment is high (Amstrong and Blaxter 1957; Arl!lstrong et al, 1957; Tyrell et al. 19'79). However when given in a concentrate diet acetate is much more efficiently used by the animal (Rooke et al. 1963; @rskov and Allen 1966a; Hovel1 et al. 1976; Tyrell et al. 1979). These conclusions are illustrated by the research of Tyrell et al. (1979) which is shown_-.---_-.-_- 5. __ in Fig. Fig. 5. Effect of the basal diet on the efficiency of utilisation of acetic acid infused into the rumen of cattle. The highest retention of energy was on the diet rich in glucose precursors (Preston and Leng 1984 adapted from Tyrell et al. 1979). Evidence froia infusion of metabolites Zconoaides et al. (1978) exmined the interaction between dietary bypass protein (fish neal) and abomsally infused glucose in lambs fed a basal Jie'c or' zq;ar/oaten chaff suppleaented with urea. Feed intaKe and growtti rates xere increased by supplements of bypass protein. Glucose infusion had no er'~cct on feed intake but increased liveweight gain and feed conversion el'r'iciency (Table 3). Table 3. The effects of bypass protein (fish meal) and &,~cogenic enzr;;y (glucose infused into the abomasum) on feed intake, g;rokLh i-ate ailJ feed conversion in lanbs fed a basal diet of sugar/oat chaff (rui;len fermentable carbohydrate) (Economides et al. 1973) Acetate clearance Sarly studies ii? Australia, aLaed at exmining the role of $ucose as a yi,,lin~ substrate for the TCA cycle (ie. providing oxaloacetate), used acetate clearance rate as a Ineasure of glucose sufficiency (Table 4). The data show that) al;;lOst alHays, acetate clearance was r;lost rapid on those diets likeljr to ilZ;Ve a hi&l glucoqenic capacity (ie. diets high in protein or t:Jaize; OP where it coulil be expected that propionate would be a high proportion of th,o rumr? VFA). Tile rate of csearance of injected acetate on the lucerne hay diet was ,'li$ly correlated (Ii = 0.98) with feed intake. The increasiq; level of intaxe coullzl be expected to lead to greater escape of potentially Glucodmic nutii-ients. T11e sudsestion from this work is that there is a close correlatim ~txce;~ tile aoility of an aniinal to clear acetate and the availability of _I;iticose. Rminants mst control blood acetate within physiological likts, a11d Glersfore fe'ed intake and fermentation rate mst aatch the anii;laPs atiiiity to utilise acetate which i s'dependent on the availability of ~lucosc This hypothesis provided that the acetate is being used for fat synthesis. silould be tested since acetate clearance rate is relatively easy to Lleasure. The otiler point arising from these considerations is that the illtake of a gmrticular feed will be maxinised when nutrient availability is ''oalancecP witrl requiremnts. Therefore, the ability of the animal to clear acetate couici be used as an index of the 'balance' of the absorbed mArien'cs. This relationship could be especially useful in grazing studies to identify the effects of supplements. . Table 4 Acetate clearance in sheep on diets of high or low glucogenic potential. Acetate clearance is defined as the time for half an injected dose of acetate to be cleared from blood (after Reid 1958; Jarrett and Filsell 1960; Egan 1965; Weston 1966) The Quperior' nutritive value of propionic acid compared with acetic acid, observed in the original work of Armstrong and Blaxter (1957 and Armstrong et al. (1957) with starved mature sheep, and Armstrong et al. (1958) with feeding sheep, and the absence of differences in efficiency of use of these two VFA acids in the experiments of $rskov and Allen (1966) can also be explained in terms of the glucogenic potential of the basal diet. The diet used by Armstrong et al. (1958) was dried grass (low glucogenic potential) whereas @rskov and .Allen (1966) gave the different VF'A mixture to animals fed mainly on barley grain (high glucogenic potential). Evidence for the essentialitv of alucose on low-fat diets The need for glucose for high efficiency of feed utilization appears to be aore obvious where the diet is low in lipid and/or when the animal is growing at a fast rate and is close to maturity (ie. is fattening). The evidence from this is discussed below. Diets based on molasses: Fattening systems for cattle on molasses based diets supplemented with bypass protein, supported levels of growth comparable with those on grain based diets (Preston et al. 1967). Molasses-fed animals, however, had a lower feed conversion efficiency and a lower carcass fat content (Redferne and Creek 1972). In contrast, only low to moderate ievels of milk production were attained when maize was replaced by molasses in a diet fed to dairy cows (Fig. 6). The level of milk yield was closely related with ruaen propionate proportions and in addition the diet was extremely low in fat (Clark et al. 1972). Fig. 6. Effects of replacing maize with molasses on the pattern of rumen frementation and milk yield of Holstein cows (Clarke et al. 1975). Diets based on sugar cane: Rice polishings with a large proportion of broken rice grains and oil (1248%) was a better supplement than cassava root meal (low in fat) for growth of cattle on sugar cane based diets (see Preston et al. 1976). The starch in rice polishings escaped rumen fermentation almost totally (Elliott et al. 19781s whereas the starch of cassava root meal was ferlllented rapidly in the rumen (Santana and Hovel1 1979). .Glucose entry rates were higher in cattle fed sugar cane based diets when supplemented with rice polishings, rather than cassava root meal (Ravelo et al. 1978). On these diets the rice polishings provided considerable fat in addition to bypass protein and starch. Supplementary maize grain (with good rumen escape characteristics and with a relatively high content of oil) improved feed conversion efficiency In cattle fed sugar cane pith whereas the same amount of molasses energy (completely fermented in the rumen containing no fat) depressed feed conversion efficiency (Donefer cited by Pigden 1972). Evidence from animalsfed bv intraruminal and intragastric infusions Strongest evidence for the thesis comes from the studies reported by rskov et al. (1979) where growing lambs nourished by infusion of VF'A into the pl rumen and infusion of casein into the abomasum (no lipids were given) increased their nitrogen retention as the proportion of propionic acid in the infused VFA was increased (see Fig. 7). The effects were the same in sheep fed at maintenance or twice maintenance. Similar experiments reported in preliminary form also support the concept of glucose increasing the efficiency of body protein gain (Girdler et _ _ _-_ - _--_.-_-.- - . . . -. . __- .._-.-_- ____ al. 198% In the studies of Blaxter and his colleagues (see Blaxter 19621, there was apparently a positive linear relationship between the molar proportions of propionate in rumen VFA and the efficiency of utilization of metabolizable energy above maintenance for fattening of sheep (see Fig. 7). As these anizlals were of mature body size they were presumably fattening, as compared to the lambs in $rskov et al. (1979) experiments which would have been depositing mainly proteinaceous tissues and water. A comparison of the data summarised by Blaxter (1962) with the results obtained by arskov et al. (1979) with young lambs nourished by infusion of VFA into the rumen and casein into the abomasum, is given in Fig. 8. The data clearly show that N-balance (Fig. 7) (proteinaceous tissue deposition) \qas stinulated in young animals as the proportion of glucogenic energy increased. ---. -_ -_- ..---. There is obviously less need for glucogenic energy in the growiw animal when the body tissue gain is high in protein rather than lipid. The highly efficient use of both dietary energy and protein at the highest glucogenic energy to total energy ratio in the nutrient (ie. where the two sets of data coincide) is clearly the optimum balance of nutrients for maxirilum efficiency of growth (ie, a G/E ratio of 53% and a G/P ratio of 54 kJ/g). The optimum balance of these nutrients, however, would be dependent on the P/E ratio which is 10 for the infusions into young lambs (Orskov et al. 1979). A P/E ratio of 10 represents an efficiency of microbial growth in the rumen equivalent to approx. Y-ATP of 7 which is relatively low (Leng 1982). Furttier evidence for the essentiality of ;;iucose from studies of rdilk production Zroilfeld (1982) in his discussion of the metabolic deteminants 02~ rKk ~mhcLion in tile relatively high yielding dairy cow9 has hi&-lighted tiiIai;Z .l. k i7ate of r;iilk synthesis is a function of wmnary upta:<e of ~~uws~?. Efficiency of iililic synthesis is a function of the uptake of' lon,i; chain fatty acids by the inamary gland. .. J The evidence presented for these stateiaents are that: 15 Snzyrue activities in the im-mary gland are not rate li&.tinl; (3arttiann 1969). l-U% secretion in the perfused udder of seats varied directly with tile perfusate concentration of glucose but not with tile concenLrations of mine acids or acetate (Hardwick zt al. 1961; Liflzell 1967). xetotic COWS, irlilk yield responded linearly to steyiqise increxent s of infused glucose (Wonfeld 1970). Xi? leb -: .. `2: 7, 2: Tile a;aouilts of glucose passing through tile blood pool ii? lacta'cind cow (Xronfeld 1982) was linearly related to milk productiou. The uptake of glucose by the mamary gland, measured using Mood flow rates and A-V differences (Hart;ilann and Xronfeld 1973; Xponfeld et al. 1968), was 72 g/kg railk produced. A sir;rilar value was reported in Goats by Annison and Linzell (1964). , Trle partial efficiency of ailk production was increased wilen &rain replaced concentrates in a diet and the level of pi?opionate in ruwm fluid increased relative to acetate aild butyrate. IV INTERRELATIONSHIPS BETWEEN REQUIREMENTS FOR LCFA AND GLUCOSE :s 9b .. 52 l %Qucose is undoubtedly thai; needed Part of a ruuinant's requirements for ti to i>e oxidised to allow fat synthesis to occur, therefore the need for ,Jiucose is a function of the level of fat in a diet. This is an i,llportant area sime it cqhasises the role of dietary fat and its analysis in feed evaluatioa pJC&&Js and therefore considerable background infornation is diven. Lon:r chain fattv acid svnthesis and catabolim 3&<,xound: Long chain fatty acids are rapidly released froid fzed ,~lymrides in the rmen by bacterial lipase. In the ruen, long chain fatty acids xay be hydrogenated to give saturated acids but are otherwise uxhanged and are efficiently absorbed from the mall intestine following their -;loveriletrt down tne tract in digesta. Ligid say contribute approximately 140% of the Cry tiatter of a forage consulned depending on the species, its iilaturity a& t,'le cox>onmCs consmed. IMai>olis;J of long chain fatty acids in the an&al results in endogenws ace-late production (Annison and White 1962). The loi chain fatty acids of tile diet are efficiently incorporated into body tissues and into rilik fat. Thus the lipids of adipose tissues becone unsaturated if unsaturated long chain fattjr acia's are added directly i;o the abomasum (Ogilvie et al. 1961) or fed as Ttp~otected ** fats (Scott et al. 1970). Lindsay (1970) demonstrated a negligible synthesis of long chain fatty acids from acetate in sheep on a maintenance diet. He provides evidence in subsequent publications to support the concept that lipids of adipose tissue arise mainly from the uptake of circulating long chain fatty acids which are of dietary origin (Lindsay 1983) a concept that is supported by Thornton and Tume (1984). Long chain fatty acid synthesis is only fron acetate where animals have either a low low protein to energy ratio in the nutrients tract where the fat deposition may act as an likely to occur quantitatively availability of dietary fat or a absorbed from the alimentary energy sink. Lipogenesis from acetate requires considerable ATP and therefore fat synthesis is accompanied by considerable heat generation. Whilst fat synthesis from acetate almost certainly requires glucose to be oxidised most of adipose tissue long chain fatty acids probabl'y arise from dietary fat. On the other hand only half the fat in milk fat (about half the C 16 and all the C18 fatty acids) arise from dietary fat, the rest is synthesised from acetate and butyrate (ie. half the C 6 and all the CQaC18 fatty acids) (see Linzell 1968). The mammary gland has developed mechan