Net Energy for Swine: Application to Low Protein Diets
by Jean Noblet - Institut National de la Recherche Agronomique, Station de Recherches Porcines, 35590 St Gilles (France)
The cost of feed is at least 50% of the total cost of pig meat production with the energy component representing the greatest proportion. Therefore, from a practical point of view, it is important to estimate precisely the energy value of feeds, either for least-cost formulation purposes or for adapting feed supply to energy requirements of animals. Evaluation of energy content of pig feeds is usually based on their digestible (DE) or metabolizable (ME) energy contents. However, the closest estimate of the " true " energy value of a feed is given by its net energy (NE) content which takes into account the metabolic utilization of ME. Therefore, NE systems have been proposed. In addition, NE is the only system in which energy requirements and diet energy values are expressed on a same basis which should be independent of the feed.
The most recent NE systems were established from measurements carried out in fattening (Schiemann et al., 1972) or in growing (Just, 1982a) pigs fed mainly cereal-based diets; they used the Weende fractionation method of organic matter. These systems have been used in many countries, especially in Northern Europe. However, according to changes in the pig industry (more by-products in feed formulation, modern lean-type pigs, progress in analytical procedures, ...) and some limitations in the above systems, it appeared as necessary to reconsider these proposals (Noblet et al., 1993a, 1994a and 1994b ; CVB, 1994 ; Noblet, 1996).
The increased concentration of pig herds and the environment constraints (i.e. reduction of soil and water pollution by nitrogen from manure) have forced to propose new feeding strategies for growing pigs. The most efficient technique consists in reducing the amount of nitrogen intake and keeping the amino acids supplies adequate for meeting the animals requirements. Phase-feeding systems combined with the use of low protein diets supplemented with industrial amino acids have therefore been proposed. Under an ideal situation, nitrogen output can be reduced by about 50%, as compared to conventional feeding strategies (Dourmad et al., 1992 ; Bourdon et al., 1995).
The first purpose of this paper is to present available energy systems (DE, ME and NE) for evaluation of pig feeds, with more emphasis on NE systems. The second purpose is to consider the consequences of lowering the level of protein in pig diets on their energy value, according to energy system. Some biochemical aspects of protein utilization and results of growth trials will also be considered. Information on methodological aspects has been given by Noblet (1996).
Energy Utilization of Pig Feeds
For most pig diets, the digestibility coefficient of energy (DCe or DE:gross energy ratio) varies between 70 and 90% for most pig diets. The variations are associated with differences in fecal digestibility of the nutrients constituting organic matter. However, most of the variation of DCe is associated with the presence of fiber which is less digestible than other nutrients (below 50%) and reduces the apparent fecal digestibility of crude protein and fat (Noblet and Pérez, 1993). In addition, the digestive utilization of fiber is variable with its botanical origin (Chabeauti et al., 1991). An example of the effect of fiber level on DCe of compound feeds in growing pigs is given in Table 1. Such an equation should not be applied to raw materials where specific relationships are to be used (Noblet and Henry, 1993).
Literature studies indicate that DCe is also affected by factors not related to the diet itself. In growing pigs, DCe increases with body weight (BW) (Noblet and Shi, 1994), with larger differences for high fiber feeds. The largest effect of BW is observed when adult sows and growing pigs are compared: digestibility coefficients are superior in all cases for the sows, the difference being greater with fibrous diets or ingredients (Fernandez et al., 1986 ; Noblet and Shi, 1993 ). In addition, the difference also depends on the origin of fiber. For instance, the values measured for wheat bran and corn gluten feed in growing pigs represented 90 and 70% of the value recorded in sows, respectively (Noblet and Bourdon, 1997).
The ME content of a feed is the difference between DE and energy losses in urine and gases (mainly as methane). In growing pigs, average energy lost in methane is equivalent to 0.4% of DE intake (Noblet et al., 1994a). In sows fed at maintenance level, methane production represents a much higher proportion of DE intake (1.5 vs 0.4% in growing pigs ; Noblet and Shi, 1993).
Energy lost in urine represents a variable percentage of DE since urinary energy is highly dependent on the amount of nitrogen in urine; the urinary nitrogen will mainly depend on the amount of digestible protein and, therefore, on the crude protein (CP) content of the diet. Consequently, the ME:DE ratio is linearly related to dietary protein content (Tables 1 and 2). In most situations, the ME:DE ratio of complete feeds is relatively constant and equivalent to about 0.96. However, that mean value cannot be applied to single feed ingredients (Shi and Noblet, 1993).
Metabolic Utilization of ME
NE is defined as ME minus heat increment associated with metabolic utilization of ME and also to the energy cost of ingestion and digestion. The ratio between NE and ME (or k ; kg for NE in growing pigs) corresponds to the efficiency of utilization of ME for NE. Apart from variations due to the final utilization of ME (protein gain vs fat gain vs milk production vs ...), k varies according to the chemical characteristics of the feed since nutrients (oses, amino acids, long-chain fatty acids or volatile fatty acids) are not used with similar efficiencies. In studies conducted with growing pigs, kg was increased when fat and starch contents were higher and reduced when protein or fiber contents were enhanced (Table 1). Similar trends were observed in adult sows fed at maintenance energy level (Noblet et al., 1993a). The variations of kg are due to differences in efficiencies of ME utilization of nutrients: 90, 82, 73, 58 and 58% when ME was provided by digestible ether extract, starch, sugars, digestible crude protein and digestible fiber, respectively. Corresponding efficiencies of DE of nutrients for NE were 90, 82, 72, 50 and 54% (Noblet et al., 1994a).
Metabolic Utilization of Energy from Protein
In Table 2, the different steps of energy utilization, expressed as a function of digestible nutrient contents, are presented. The equations indicate that most of the variation in energy losses occurring in the DE to ME step are related to the utilization of digestible crude protein (DCP) in connection with the excretion of nitrogen as urea . From the coefficients of DCP or catabolized DCP (DCPc) in the DE and ME prediction equations, the energy loss can be estimated as about .7 to .8 kcal per g of DCPc. Comparable estimates can be obtained from the studies of Schiemann et al. (1972) in near-maturity pigs and Noblet et al. (1993) in adult sows at maintenance energy level (.8 and .7 kcal/g of DCPc, respectively). Even the N in urine is not only as urea, it can be noticed that these values are close to the energy content of urea (.87 kcal/g N x 6.25). The ME value of DCPc is then equivalent to about 4.7 kcal/g.
In the ME to NE step, energy losses as heat increment in growing pigs concern all nutrients : about 2.0, 1.0, .75 and 1.2 kcal per g of DCP, digestible ether extract, starch and digestible fiber, respectively (Table 2). According to the same approach, it was 1.6 kcal per g of DCP in maintenance adult sows (Noblet et al., 1993a) or growing pigs (Just, 1982b) and 2.3 kcal per g of DCP in near-maturity pigs (Schiemann et al., 1972). In all studies, the heat increment due to the metabolic utilization of DCP is significantly higher than for starch or ether extract.
The heat increment of DCP as measured in the above studies corresponds to the combination of heat increments of DCP for protein gain and DCP for maintenance (ATP production) and, in some instances, for lipid synthesis. Biochemical approaches indicate that, as compared with glucose or fat, about 20% more ME is required for ATP formation when it is supplied by protein (Krebs, 1964 ; Armstrong, 1969 ; Schulz, 1975). It must also be mentioned that increased protein supply is associated with a higher protein turnover (Reeds et al., 1981) and an increased mass of visceral organs (Noblet et al., 1987) with subsequently enhanced heat production. As mentioned before, the catabolism of protein is associated with urea synthesis whose energy content is considered in the DE to ME step ; but urea synthesis also requires energy which is dissipated as heat. According to the estimates of Buttery and Boorman (1976), it would be equivalent to about .6 kcal per g of DCP (on the basis of about 5 ATP per mole of urea). Finally, when DCP or amino acids are used for protein deposition, additional energy as ATP for peptide bond and associated increased body protein turnover is required. The estimates of additional energy for protein deposition are quite variable ; .6 to .7 kcal per kcal retained which is equivalent to an energetic efficiency of 60% (so-called kp) can be suggested. This value is remarkably comparable to the efficiency of utilization of ME of DCP for NE.
Overall, there are many factors which explain the high rate of heat production when energy is supplied by DCP or when the amount of DCP in the diet is increased. But, further studies are required to evaluate their respective contributions. In addition, the coefficient affected to DCP in the NE equations has been obtained with diets whose protein levels increased from conventional to excessive levels. The situation of low protein diets supplemented with synthetic amino acids has not been extensively considered. The metabolic studies of Fuller et al. (1987), Noblet et al. (1987) and Moehn and Susenbeth (1995) would confirm the expected reduced heat loss with low protein and amino acids supplemented diets in growing pigs. Growth trials (see Tables 6 to 9) also confirm the same hypothesis.
From the available information (Table 2), it is proposed that the NE value of 1 g of DCP is about 2.7 kcal (or 2.4 kcal per g of CP if the DC of CP is 90%) while the estimated NE content of 1 g of starch is 3.4 kcal; starch is supposed to be 100% digestible. Consequently, the substitution of 1 g of protein by one g of starch should increase the NE value of the diet by about 1 kcal; on a ME basis, the energy value would hardly be changed and it would be decreased on a DE basis (minus .7 kcal/g). In practical terms, for a diet containing 2200 kcal NE per kg, the DE content would be reduced by about .30% and the NE value would be increased by about .45% for each 1% reduction of the CP level and its replacement by starch. Ewan (1991) suggested that NE value of a diet was increased by about .6% for each 1 percent reduction of the CP level. Addition of both estimates (.45% plus .30%, i.e. .75%) gives an indication of the additional DE to supply in order to maintain the same NE supply when starch is replaced by protein in the diet. In a similar approach, Noblet et al. (1987) suggested a slightly higher figure (1%) from results of an experiment conducted in young growing pigs.
Prediction of Net Energy Value
All published NE systems combine the utilization of ME for maintenance and for growth (Just, 1982; Noblet et al., 1994a and 1994b) or for fattening (Schiemann et al., 1972) by assuming similar efficiencies for maintenance and energy retention. Equations proposed by Noblet et al. (1993c) can be used to predict NE value of feeds in a maintenance situation. One important aspect is that NE value is directly dependent on the estimate of fasting heat production used in the calculation; the practical consequence is that absolute NE values obtained under different measurement conditions or with different hypotheses cannot be compared.
In studies conducted at INRA, each diet was measured for its chemical composition, DE, ME, and digestible nutrient contents with the objective of using these data as predictors of NE content. Different linear regression models were tested. But, for practical application, such regression equations should use easily available criteria, either from feeding tables or at the laboratory level (i.e. from accurate and not expensive analyses). The most important NE prediction equations we obtained on 61 diets fed to 45-50 kg Large White boars are presented in Table 3. They can be applied to both single feedstuffs and mixed diets. The first one (NEg2) is based on digestible nutrient contents which can be calculated from chemical characteristics and digestibility coefficients of the different fractions. Digestibility coefficients in most feedstuffs are available in Dutch (CVB, 1993) or German or Danish feeding tables; it can be assumed that starch is 100% digestible. The two other equations (NEg4 and NEg7) take into account the DE or ME contents and some chemical characteristics. They can also be applied from information available in feeding tables which indicate DE or ME contents of most feedstuffs used in pig diets formulation. Reliable information on digestibility of energy or of nutrients is then necessary for prediction of NE content of pig feeds.
The studies we conducted in heavier pigs or in adult sows indicate that the equations obtained in young growing pigs (Table 3) are applicable at all stages of pig production (Noblet, 1996). In the case of adult sows fed at maintenance, the measured NEm value is higher than the calculated NEg; but the difference between NEm and NEg is not explained by any chemical characteristic of feeds. This means that the equations proposed in Table 3 and based on DE, ME or digestible nutrient contents are able to determine an acceptable hierarchy between feeds when they are used in a maintenance situation. The superiority of NEm over NEg will be taken into account in the calculation of requirements. But, as shown above, the digestibility coefficients of energy or chemical constituents vary according to body weight of animals with major differences between adult sows and growing pigs (especially for high fibre feeds). Net energy values should then be different for both stages.
Comparison of DE, ME and NE Systems
From the equations reported in Tables 1, 2 and 3, it is obvious that the hierarchy between feeds obtained in the DE or ME systems will vary in the NE system according to their specific chemical composition. Since NE represents the best estimate of the "true" energy value of a feed, the energy value of protein or fibrous feeds is overestimated when expressed on a DE (or ME) basis. On the other hand, fat or starch sources are underestimated in a DE system. These conclusions are more clearly demonstrated in Tables 4 and 5.
These tables clearly demonstrate that the hierarchy between feeds is dependent on the energy system, the biggest differences being observed for ingredients whose chemical composition is quite different from that of standard diets (fat sources and protein- and/or fiber-rich ingredients). Results in least-cost formulation will therefore depend on the energy system. Unpublished results show that diets have lower protein contents when formulated on a NE concept than on a DE basis and a subsequent higher supplementation of synthetic amino acids.
Performance and Utilization of Energy in Growing Pigs Fed Low Protein Diets
The effect of reducing the protein level in diets for growing pig has been studied in many experiments but under very variable conditions. If we limit our purpose to experiments in which essential amino acids supply across diets was kept constant, correctly balanced and above requirements for optimal growth, the number of experiments becomes considerably lower. The main difficulties in interpreting data are then differences in feeding strategy (ad libitum vs restricted feeding) or the absence of information on protein and energy value of diets or on body composition of pigs. Most important and synthetic results of some recent experiments are presented in Tables 6 to 9.
The effect of lowering the dietary CP level (at constant supply of essential amino acids) on voluntary feed intake (VFI) is not quite clear and results are contradictory. The differences in response are probably related to factors such as genotype or sex, nature of ingredients, sub-limiting or excessive levels of some amino acids and environmental conditions. However, the general trend would be a slight increase of VFI with the reduction of CP level. In most experiments conducted under ad libitum feeding or even restricted feeding (on a DE or ME scale), the growth was not affected and the carcasses at slaughter contained more fat and less lean when the CP level is reduced (Tables 6 and 7). This situation is quite logical under ad libitum feeding since the trend is an over-consumption of energy. If carcasses are fatter when low CP diets are fed at similar DE or ME intakes, this means that more energy is retained in the body and also more energy is available from the diet (Tables 6 and 7). In comparable experiments but at controlled NE intakes, CP level affected neither growth of the animals nor body composition at slaughter (Tables 8 and 9) ; in addition, the feed conversion ratio was affected by CP level when expressed on a DE basis and independent on CP level when expressed on a NE basis (Table 8). In the previous experiments (Tables 6 and 7), the feed conversion ratio, expressed on a DE or ME basis, was not affected by CP level while an increased fat content in the carcass suggests an increased energy requirement per unit of BW gain.
The combination of these results indicates that the energy value of low protein diets is underestimated and/or the energy value of conventional diets is overestimated when it is expressed on a DE basis. These conclusions are quite consistent with the differences in the relative energy values of different feeds in DE and NE systems (Tables 4 and 5) and the high heat production and energy loss as urea associated with the utilization of excessive DCP (Table 2).
Theoretically, energy requirements of animals and energy values of feeds can be expressed on the same basis when the NE concept is used for both of them. Furthermore, NE value of a feed should be a better predictor of the growth response of pigs than DE value, since it is closer to the " true " energy value of feeds. These assumptions are illustrated in Tables 8 and 9 in the case of diets with variable CP contents. The same conclusion was given for diets with variable fat contents (Noblet, 1996). In addition, difficulties in the interpretation of increased fatness of carcasses of pigs fed low protein diets (or also high fat diets) are quite reduced when differences in NE intakes are considered. Net energy systems are therefore highly preferable in such situations.
However, when changing from DE (or ME) to NE systems and using low protein diets, attention should be paid to the reduction of amino-acids levels and the subsequent risk of sub-limiting available amino-acids supplies to the pigs. Consequently, it is highly suggested to adopt an accurate protein evaluation system (available or digestible amino-acids) when a NE system is used and/or CP level is reduced, in order to adapt feed composition more precisely to requirements and growth potential of the pig. An other consequence is that the utilization of low CP diets may result in fatter carcasses in connection with the slight over-consumption of feed and its higher NE value. Reduction of CP level with sufficient supplies of essential amino acids should then be combined with strategies for limiting energy intakes, especially with animals depositing rather large amounts of fat (castrates).
It must also be stated that prediction of net energy value of pig feeds depend directly on their digestible (or metabolizable) energy or digestible nutrients contents. These latter quantities depend firstly on dietary characteristics of the feed, but they are also affected by (bio)technological treatments, animal factors and interactions between these factors and feed composition. Improvements in the prediction of energy value of pig feeds will therefore come mainly from a better knowledge of factors of variation of energy and nutrients digestibility. In the above paragraphs, we have particularly insisted on the differences between adult sows and growing pigs.
One major objective in the reduction of CP level in growing pigs or sows diets is to reduce the nitrogen output from pig production. This can be significantly achieved with good knowledge of amino acids and energy requirements of animals and accurate estimation of the nutritional value of feeds (available amino acids, net energy). An other objective in reducing the CP level (and/or increasing the fat content) might be to decrease the amount of heat which is dissipated by the animals. A potentially high heat production can represent a limiting factor in pigs kept under hot climatic conditions or with a limited heat tolerance ; they will then react by a lower VFI and subsequent reduced performance. The use of low CP and/or high fat diets might then be an attractive solution to reduce heat stress (Lopez et al. 1994). Such strategies deserve further experimental studies and confirmation.
Armstrong, D.G. 1969. Cell bioenergetics and energy metabolism. In : W. Lenkeit, K. Breirem, and E. Craseman (Ed.) Handbuch der Tierernährung. pp 385-414. Verlag P. Parey, Hamburg, Germany.
Buttery, P.J., and K.N. Boorman. 1976. The energetic efficiency of amino acid metabolism. In : D.J.A. Cole, K.N. Boorman, P.J. Buttery, D. Lewis, R.J. Neale, and H. Swan (Ed.) Protein metabolism and nutrition. pp 197-206. Butterworths, London , UK.
CVB (Cental Veevoederbureau). 1994. Veevoedertabel. Gegevens over chemische samenstelling, verteerbaarheid en voederwaarde van voedermiddelen. CVB: Lelystad (The Netherlands).
Bourdon D., J.Y. Dourmad, and Y. Henry. 1995. Réduction des rejets azotés chez le porc enc roissance par la mise en uvre de l'alimentation multiphase, associée à l'abaissement du taux azoté. Journées Rech. Porcine Fr. 27 :269-278
Chabeauti, E. J. Noblet, and B. Carré. 1991. Digestion of plant cell walls from four different sources in growing pigs. Anim. Feed Sci. Techn. 32 : 207-213.
Cromwell, G.L., M.D. Lindermann, G.R. Parker, K.M. Laurent, R.D. Coffey, H.J. Monegue, and J.R. Randolph. 1996. Low protein, amino acid supplemented diets for growing-finishing pigs. J. Anim. Sci. 74(Suppl. 1): 174 (Abstr.).
Dourmad, J.Y., D. Guillou, and J. Noblet. 1992. Development of a calculation model for predicting the amount of N excreted by the pig: effect of feeding, physiological stage and performance. Livest. Prod. Sci. 31: 95-107.
Dourmad, J.Y., Y. Henry, D. Bourdon, N. Quiniou, and D. Guillou. 1993. Effect of growth potential and dietary protein input on growth performance, carcass characteristics and nitrogen output in growing-finishing pigs. In: M.W.A. Verstegen, L.A. den Hartog, G.J.M. van Kempen, and J.H.M. Metz (Ed.) Nitrogen flow in pig production and environmental consequences. pp 206-211. Pudoc. Wageningen, The Netherlands.
Ewan, R.C., 1991. Energy utilization in swine nutrition. In : E.R. Miller, D.E. Ullrey, and A.J. Lewis (Ed.) Swine Nutrition. pp 121- . Butterworth-Heinemann, Boston, USA.
Fernandez, J.A., H. Jorgensen, and A. Just. 1986. Comparative digestibility experiments with growing pigs and adult sows. Anim. Prod. 43 :127-132.
Fuller, M.F., A. Cadenhead, G. Mollison, and B. Sève. 1987. Effects of the amount and the quality of dietary protein on nitrogen metabolism and heat production in growing pigs. Br. J. Nutr. 58:277-285.
Just, A. 1982a. The net energy value of balanced diets for growing pigs. Livest. Prod. Sci., 8:541-555.
Just, A. 1982. The net energy value of crude (catabolized) protein for growth in pigs. Livest. Prod. Sci. 9 : 349-360.
Krebs, H.A., 1964. The metabolic fate of amino acids. In : H.N. Munro and J.B. Allison (Ed.) Mammalian protein metabolism. pp 125-170. Academic Press, London, UK.
Lopez, J., R.D. Goodband, G.L. Allee, G.W. Jesse, J.L. Nelssen, M.D. Tokach, D. Spiers, and B.A. Becker. 1994. The effects of diets formulated on an ideal protein basis on growth performance, carcass characteristics, and thermla balance of finishing gilts housed in a hot, diurnal environment. J. Anim. Sci. 72 :367-379.
Moehn, S., and A. Susenbeth. 1995. Influence of dietary protein content on efficiency of energy utilization in growing pigs. Arch. Anim. Nutr. 47: 361-372.
Noblet, J., Y. Henry, and S. Dubois. 1987. Effect of protein and lysine levels in the diet on body gain composition and energy utilization in growing pigs. J. Anim. Sci. 65 :717-726.
Noblet, J., and Y. Henry. 1993. Energy evaluation systems for pig diets: a review. Livest. Prod. Sci. 36: 121-141.
Noblet, J., and J.M. Perez. 1993. Prediction of digestibility of nutrients and energy values of pig diets from chemical analysis. J. Anim. Sci. 71: 3389-3398.
Noblet, J., and X.S. Shi. 1993. Comparative digestibility of energy and nutrients in growing pigs fed ad libitum and adult sows fed at maintenance. Livest. Prod. Sci. 34: 137-152.
Noblet, J., X.S. Shi, and S. Dubois. 1993a. Metabolic utilization of dietary energy and nutrients for maintenance energy requirements in sows: basis for a net energy system. Br. J. Nutr. 70 : 407-419.
Noblet, J., H. Fortune, C. Dupire, and S. Dubois. 1993b. Digestible, metabolizable and net energy values of 13 feedstuffs for growing pigs: effect of energy system. Anim. Feed Sci. Technol. 42 : 131-149.
Noblet, J., and X.S. Shi. 1994. Effect of body weight on digestive utilization of energy and nutrients of ingredients and diets in pigs. Livest. Prod. Sci. 37 : 323-338.
Noblet, J., H. Fortune, X.S. Shi, and S. Dubois. 1994a. Prediction of net energy value of feeds for growing pigs. J. Anim. Sci. 72 : 344-354.
Noblet, J., X.S. Shi, and S. Dubois. 1994b. Effect of body weight on net energy value of feeds for growing pigs. J. Anim. Sci. 72 : 648-657.
Noblet, J., 1996. Digestive and metabolic utilization of energy in pig feeds: comparison of energy systems. In : P.C. Garnsworthy, J. Wiseman and W. Haresign (Ed.) Recent Advances in Animal Nutrition. pp 207-231. Nottingham University Press, Nottingham, UK.
Noblet, J., and D. Bourdon. 1997. Valeur énergétique comparée de onze matières premières chez le porc en croissance et la truie adulte. Journées Rech. Porcine en France 29 (in press)
Reeds, P.J., M.F. Fuller, A. Cadenhead, G.E. Lobley, and J.D. McDonald. 1981. Effects of changes in the intakes of protein and non-protein energy on whole-body protein turnover in growing pigs. Br. J. Nutr. 45 : 539-546.
Schiemann, R., Nehring, K., Hoffmann, L., Jentsch, W. and Chudy, A., 1972. Energetische Futterbevertung und Energienormen. VEB Deutscher Landwirtschatsverlag, Berlin, 344 pp.
Schulz, A.R. 1975. Computer-based method for calculation of the available energy of proteins. J. Nutr. 105 :200-207.
Shi, X.S., and J. Noblet. 1993. Digestible and metabolizable energy values of ten ingredients in growing pigs fed ad libitum and sows fed at maintenance level; comparative contribution of the hindgut. Anim. Feed Sci. Technol. 42:223-236.
Valaja, J., T. Alaviuhkola, and K. Suomi. 1993. Reducing crude protein content with supplementation of lysine and threonine in barley-rapeseed meal- pea diets for growing pigs. Agr. Sci. Finl. 2 : 117-123.
Tables and Figures available in PDF version.