SYMPOSIUM: LYSINE. Ideal Protein in Turkeys 1

SYMPOSIUM: LYSINE Ideal Protein in Turkeys 1 J. D. FIRMAN2 and S. D. BOLING3 Department of Animal Science, University of Missouri-Columbia, Columbia, Missouri 65211 ABSTRACT This paper will review current and previous research related to our work on ideal proteins for turkeys. The concept of ideal protein is not new, but has not been researched in turkeys to any extent. Ideal proteins may be defined as the exact balance of amino acids needed for maximal growth. Ideal proteins are based on digestible amino acids. Several pieces of information need to be in place before one can actually run an experiment on the digestible amino acid requirements and ideal protein. The first of these is data on digestible amino acid content of feedstuffs for turkeys. The second necessity is a low protein diet that can be used to titrate the amino acids on a digestible basis. These requirements led us to formulate a very low protein starter diet (approximately 18% intact crude protein + amino acids) with corn and soybean meal that would support maximal growth. To formulate a diet at this crude protein level, it was necessary to use the ideal protein research from other species and previous research with turkeys. We now have an estimate of the ideal protein for turkeys. The future of this work and amino acid nutrition of turkeys will be reviewed. (Key words: turkey, amino acids, ideal protein, digestible amino acids) 1998 Poultry Science 77:105 110 INTRODUCTION Protein (as amino acids; AA) is one of the major cost components of the diet of turkeys. Protein also has major effects on performance as well as the overall cost of the finished product. Considering the major importance of feed to the turkey industry, there is, relatively speaking, little research on the feeding of turkeys. This dearth of research has led to wide variation in feeding programs for turkeys relative to protein or AA. As our scientific and nutritional knowledge has advanced, the field of poultry nutrition has taken advantage of this knowledge. Initially, poultry were fed on a total crude protein basis (Figure 1). This basis did not take into account any of the AA needs of the animal. With more research, we then moved to formulation on an AA requirement basis. Although the data set for this is not very complete (there is a good deal of extrapolation) for the turkey, it is still a major step towards the true requirement of the bird. The industry then moved to routine laboratory analyses for AA and formulas to predict AA content based on the crude protein content Received for publication April 14, 1997. Accepted for publication August 5, 1997. 1Contribution from the Missouri Agricultural Experiment Station. Journal Series Number 12,630. 2To whom correspondence should be addressed. 3University of Illinois, Department of Animal Sciences, 286 ASL, 1207 W. Gregory Drive, Urbana, IL 61801. of the feedstuffs. Most recently, we have begun to use AA digestibility values to formulate diets with the same AA content using different feedstuffs. We are now moving into ideal proteins and the use of tissue accretion methods to predict retention efficiency on our way to truly defining the exact requirements for AA in the turkey. This paper will cover some of the concepts and data that relate to ideal proteins and digestible AA requirements and will suggest the direction that AA nutrition of turkeys may head to in the future. BACKGROUND The data on protein and AA requirements of turkeys have been recently reviewed (Firman, 1994a). Briefly, there are few data on the AA requirements for turkeys. The general AA data found in the NRC go back to the early 1970s with research from Warnick and Anderson (1973). These data were collected for Bronze turkeys with what we would consider to be nonstandard diets during the starter period only. Much of the later numbers have been extrapolated from this starter period. Based on these early data, it appeared that lysine and sulfur AA (SAA) are most limiting in commercial diets. Dr. Potter s group at VPI (Stas and Potter, 1982; Jackson et al., 1983; Jackson and Potter, 1984) performed a number of studies using deletion methods to determine that threonine, valine, and isoleucine were next limiting in corn-soybean meal diets. We have also used a low protein diet and found the same AA are limiting based on an addition method (Firman, 1994a). 105

106 FIGURE 1. Development of nutritional knowledge based on formulation methodology and inputs for amino acid nutrition. Before research on ideal protein determination could begin, a good deal of information needed to be in place. Procedures and data for determining digestible AA in feedstuffs for turkeys had to be developed. We have done a good deal of work with cecectomized turkeys to calculate digestibilities of a variety of feedstuffs (Firman, 1992; Firman and Remus, 1993). The next step was formulation of a diet with reduced protein levels in which AA could be added back and maximal growth achieved. The diet needed to be low enough in crude protein from intact sources to allow for titration of the individual AA. A workable diet with these levels of protein and similar growth to a 28% crude protein diet was achieved several years ago (Boling and Firman, 1997a) and was used for all subsequent work. Once these prerequisites were in place, our attention turned to digestible AA requirements and ideal protein. IDEAL PROTEIN The concept of an ideal protein is not new. As early as 1946, Mitchell and Block discussed the concept of a perfect balance of AA. A variety of attempts have been made to determine an ideal AA pattern through carcass composition (Price et al., 1953; Williams et al., 1954; Summers and Fisher, 1961; Robel and Menge, 1973). Although carcass composition may be used as a starting point for an ideal ratio, it does not take into account the dynamics of the live animal such as maintenance costs. Dean and Scott (1965) were the first to use dietary testing to begin determination of ideal proteins. Since that time, a number of studies have sought to determine FIRMAN AND BOLING the ideal ratio for both the chick and the pig (Fuller et al., 1989, Wang and Fuller, 1989; Chung and Baker, 1992, Baker and Chung, 1992; Baker and Han, 1994; Brown, 1994). Let s take a moment and look at what ideal protein is, what it can be used for, and what it probably shouldn t be used for. Although there are several ways to define ideal protein, it is ideally the theoretically exact balance of AA that meet the animal s needs. There should be no excess, no deficiency, and as little of the AA should be used for energy as possible. Nitrogen excretion would be minimized in this situation and all dispensable AA should be provided in another form (i.e., not indispensable AA). The concept is that all of the AA can be related to lysine (by choice) and that if the lysine needs of the animal increase due to genetics, etc. then the AA pattern remains the same relative to the lysine requirement. The ratio probably changes throughout the growth cycle of the animal: as it moves to periods of feather growth vs breast meat accretion, for example. The ideal ratio is useful from several standpoints. It allows for determination of digestible AA requirements that can then be easily modified through the changes in the requirement for lysine. This methodology allows for rapid research responses to improving genetics and the new AA levels that such genetic makeup might require. The methodology has allowed us to formulate very low protein diets for determination of AA requirements. It forces us to move to digestible formulation, which I believe is useful in terms of more accurately meeting the turkeys needs, reducing overfeeding, and correctly placing a monetary value of feedstuffs based on their utilization capability for the animal. Pricing of ingredients, based on the actual usable nutrient content, may be the most useful portion of the switch to digestible requirements and formulation. In theory this concept sounds good, but there are also some inherent difficulties when we try to use the ideal values for practical formulation. Use of the ideal protein has taken some criticism based on the apparent lack of practicality in formulations. This criticism has probably occurred for several reasons. First, one must expend the effort to set up the formulation software for using digestible AA or the ratio has little meaning. Second, the numbers should just be used as minimum requirement values for each AA, as it will not work to try to meet the exact requirements. Meeting the exact requirements means going very low on protein and adding back substantial amounts of crystalline AA (extremely expensive). Thus, the ideal ratio is probably most useful in determination of digestible requirements and the rapid determination of new requirements as new genetics or conditions arise. In the future, we can probably use feed intake data, environmental conditions, etc., to model the requirements for each operation to obtain ideal inputs to achieve the maximum return. The ideal ratio for the chick, pig, and our estimate for the turkey are noted below (Table 1).

LYSINE SYMPOSIUM 107 TABLE 1. Estimated ideal protein ratio for starting hen turkeys (chick and pig ratio for comparison) 1 Amino acid 2 Turkey Chick Pig Lysine 100 100 100 SAA 59 72 60 Threonine 55 67 65 Valine 76 77 68 Arginine 105 105... Histidine 36 31 32 Isoleucine 69 67 60 Leucine 124 100 111 Phe+Tyr 105 105 95 Tryptophan 16 16 18 1Expressed as a percentage of the lysine requirement. 2Estimates based on previous research with low protein turkey diets and the estimates for other species. EXPERIMENTAL METHODS The general methods used in the experiments reported were similar. All poults were housed in environmentally controlled rooms with temperature and humidity control. Stainless steel batteries were used with four hen poults per pen. Replications ranged from six to eight per treatment. All experiments were designed as randomized blocks with each treatment randomized within each block to account for environmental differences. Poults were allowed a 1-wk acclimation period and then weight-sorted on an individual basis. They were then removed from feed for 4 h, wing-- banded, and weighed into groups of four per pen with similar starting weights. The poults were then fed the experimental diets for 10 to 14 d, removed from feed for 4 h, and weighed. Feed intake was monitored for disappearance and all dead birds were weighed for feed conversion adjustment. Diets were similar, based on our low (approximately 18% intact crude protein) protein diet (Boling and Firman, 1997a). All AA were added to this ration in excess of the estimated ideal ratio except the test AA. This AA was then titrated, first in a broad range that would be considered from very deficient to above the National Research Council (NRC, 1994) requirement, which is expressed on a nondigestible basis. These initial experiments generated an initial curve. Follow-up experiments more closely defined the digestible requirement. A positive control diet consisting of corn and soybean meal that met or exceeded all NRC (1994) requirements was used to ascertain that maximal growth had been achieved by diets in the plateau region of the growth curve. All dietary components were analyzed for AA content at each source change and digestibility assays were run routinely. All data were initially analyzed via analysis of variance and where appropriate for breakpoints based on the method of Robbins (1986). LYSINE REQUIREMENTS The first series of studies focused on the lysine requirement for the obvious reason that this is the base AA that will be used for all future relationships. Initial studies focused on the 1.42 to 1.72% digestible lysine range. However, after several attempts in this range, with no differences in any treatment observed, it was determined that the plateau began at a lower point than this. The next study extended the range to 1.18% digestible lysine and a breakpoint was seen between 1.30 and 1.36%. This point also showed a break on feed efficiency. A repeat of this experiment showed similar results with the change in feed:gain from 1.30 to 1.36% digestible lysine causing a change of 1.97 to 1.77 in feed: FIGURE 2. Break point analysis of weight gain for poults fed graded levels of lysine. FIGURE 3. Break point analysis of feed:gain for poults fed graded levels of lysine.

108 FIRMAN AND BOLING TABLE 2. Growth performance of female Hybrid turkey poults fed graded levels of digestible lysine. (All measurements are average per bird per treatment) Digestible Weight Feed Feed: lysine gain intake gain (%) (g) (g:g) 1.26 268 472.8 1.77 1.29 1 269 457.7 1.71 1.32 294 503.0 1.72 1.35 288 483.8 1.68 1.38 1 280 472.0 1.69 1.41 300 511.3 1.70 Pooled SEM 1 7.4 (7.9) 11.0 (11.8) 0.03 (0.04) Significance 2 P < 0.05 P < 0.05 P < 0.05 1SEM differed in treatments 1.29 and 1.38%; n = 7; values shown in parentheses. 2Probability of whole model test. Please see Figure 2 and 3 for broken line analysis. gain. The next experiment then used closer increments focused around the 1.30 to 1.36% range. Similar results were attained (Table 2, Figures 2 and 3) with the breakpoint for growth at 1.32% digestible lysine and 1.34% digestible lysine for feed efficiency (Boling and Firman, unpublished date). This level is substantially lower than the NRC (1994) requirement. Based on an estimate of a typical diet, the 1.60% lysine noted in the NRC (1994) would be approximately 1.48% on a digestible basis. The reason for this discrepancy might relate to the relative lack of recent data on lysine requirements. The strains of birds fed today have different intake patterns than those used in the past, and the modern bird may be more efficient at utilization of feed. METHIONINE REQUIREMENTS These studies were similar to those noted above with the exception of titration of the SAA. Lysine was maintained slightly above the previously determined requirement to ascertain that no deficiency occurred and SAA were titrated initially at 0.77 to 1.19% digestible level. This range resulted in no growth response. This study was repeated again with similar results. The next study broadened the levels to 0.50 to 1.33% digestible SAA. A plateau occurred at 0.77% digestible SAA. The final experiment (Table 3, Figures 4 and 5) used a range of 0.59 to 0.83% with a breakpoint found at 0.76% for body weight gain and 0.75% for feed:gain (Boling and Firman, 1997b). CURRENT STUDIES Currently, we are working on threonine and valine requirements and the energy:lysine ratio. Although the ideal ratio noted previously is an estimate, it is relatively speaking, close to the correct levels. We have probably overestimated several of the AA in the ratio. Several FIGURE 4. Broken-line analysis on weight gain of poults fed graded levels of digestible SAA. FIGURE 5. Broken-line analysis on feed:gain of poults fed graded levels of digestible SAA.

LYSINE SYMPOSIUM 109 TABLE 3. Growth performance of female turkey poults fed graded levels of digestible SAA by DL-methionine addition. (All measurements are average per bird per treatment) Calculated Analyzed Digestible Total Weight Feed Feed: SAA SAA gain intake gain (%) (g) (g:g) 0.59 0.71 222 527.3 2.36 0.62 0.74 261 537.6 2.07 0.65 0.77 305 583.4 1.92 0.68 0.80 318 610.4 1.92 0.71 0.83 303 582.3 1.92 0.74 0.86 333 629.4 1.88 0.77 0.89 347 623.6 1.80 0.80 0.92 355 636.9 1.79 0.83 0.95 350 617.9 1.77 PC 1 354 569.5 1.61 Significance 2 P < 0.05 P < 0.05 P < 0.05 Pooled SEM 17.7 31.3 0.07 1Feed conversion was slightly improved in this trial for birds consuming the PC diet. This improvement is thought to be due to a different soybean meal source being used that had a higher ME content, thereby improving feed:gain. 2Probability of whole model test. Please see Figure 4 and 5 for broken-line analysis. studies have used different strains of turkey with no differences found. There may be some sex differences in AA requirements, although we have not yet studied this. Such research will be much more successful if a basic ratio is in existence. It is anticipated that the ratio will change with age as the relative proportion of tissues deposited changes. As we perform these trials, we are also looking at tissue deposition curves to determine the efficiency of AA deposition. FUTURE USE OF IDEAL PROTEIN The future of the ideal protein concept depends on a number of factors. The first of these is the continuation of ongoing research. New data will further define the ratio and provide for any changes that may occur over age or for different sexes. The ideal ratio may be the next step in industry formulation. Utilization of the ideal ratio as minimum digestible AA requirements will aid in equalization of dietary requirements across changing ingredient compositions. It will also allow for proper pricing of ingredients based on their utilizable AA content, much as we currently do for metabolizable energy content of feedstuffs. What may be the biggest use for ideal proteins is the potential for rapid determination of requirements for new strains of turkeys. The cost of turkey research during the later periods of growth makes many experiments prohibitively expensive. Thus, the use of an ideal ratio would allow for one experiment to determine the lysine requirement and all of the other AA requirements would be determined based on the relationship to lysine. This would solve one of the major problems that breeders have as they introduce new strains with more rapid growth rates. Another area of interest that has really seen little work is that of pollution. Although the focus of most efforts to deal with nitrogenous waste production have revolved around dilution of the final product, use of ideal proteins has the potential to reduce initial input of nitrogen through a reduction in the overfeeding of protein currently practiced. This idea has tremendous potential for the reduction of nitrogenous waste. FUTURE RESEARCH IN AA NUTRITION The ideal protein ratio is not the final step towards the exact needs of the animal (Figure 1). The next step appears to require a model in which we actually calculate the AA needed for deposition into muscle tissue. Although a number of attempts have been made to model nutritional needs of turkeys (Firman, 1994b), there really are not sufficient data available. When the ultimate product being produced is lean muscle, the focus should be on maximizing muscle tissue accretion with minimum inputs. This maximization of muscle will require that several inputs be determined. The protein deposition rates is one of the major inputs needed. This methodology, combined with AA composition of the whole body, will allow for quantitation of the AA needed. The digestibility of the feedstuff times the efficiency of utilization (units of digestible AA input relative to AA deposited) will disclose the relative loss of inputs. These data can be combined into the following equation: grams of protein deposited for period percentage AA composition efficiency of deposition aa digestibility grams of AA/period. These numbers would then be adjusted for intake for each period and a requirement could be determined.

110 Although there are a number of other approaches that may include items such as maintenance costs, it is the authors opinion that the above equation is the simplest approach. CONCLUSIONS The ideal protein concept has tremendous potential for improvements in formulation of turkey diets. Although the research data will continue to accumulate in the future, this conceptual foundation aids in providing a roadmap for this evolution. REFERENCES Baker, D. H., and Y. Han, 1994. Ideal amino acid profile for chicks during the first three weeks posthatching. Poultry Sci. 73:1441 1447. Baker, D. H., and T. K. Chung, 1992. Ideal protein for swine and poultry. BioKiowa Tech. Rev. 4:2 14. NutriQuest, Inc., Chesterfield, MO. Boling, S. D., and J. D. Firman, 1997a. A low-protein diet for turkey poults. Poultry Sci. 76:1298 1301. Boling, S. D., and J. D. Firman, 1997b. Digestible sulfur amino acid requirement of starting turkeys. Poultry Sci. 76: 873 877. Brown, R. H., 1994. Worldwide movement aimed at lowering protein in rations. Feedstuffs, June 6:24. Chung, T. K., and D. H. Baker, 1992. Ideal amino acid pattern for 10-kilogram pigs. J. Anim. Sci. 70:3102 3111. Dean, W. F., and H. M. Scott, 1965. The development of an amino acid reference diet for the early growth of chicks. Poultry Sci. 44:803 808. Firman, J. D., 1992. Amino acid digestibilities of soybean meal and meat meal in male and female turkeys of different ages. J. Appl. Poult. Res. 1:350 354. Firman, J. D., 1994a. Utilization of low protein diets for turkeys (review). BioKyowa Tech. Rev. 7:2 10. NutriQuest, Inc., Chesterfield, MO. Firman, J. D., 1994b. Turkey growth modeling: A metabolic approach. J. Appl. Poult. Res. 4:373 378. FIRMAN AND BOLING Firman, J. D., and J. C. Remus, 1993. Amino acid digestibilities of feedstuffs in female turkeys. J. Appl. Poult. Res. 2: 171 176. Fuller, M. F., R. McWilliam, T. C. Wang, and L. R. Giles, 1989. The optimum dietary amino acid pattern for growing pigs. Br. J. Nutr. 62:255 267. Jackson, S., and L. M. Potter, 1984. Influence of basic and branched chain amino acid interactions on the lysine and valine requirements of young turkeys. Poultry Sci. 63: 2391 2398. Jackson, S., R. J. Stas, and L. M. Potter, 1983. Relative deficiencies of amino acids and nitrogen per se in low protein diets for young turkeys. Poultry Sci. 62:1117 1119. Mitchell, H. H., and R. J. Block, 1946. Some relationships between the amino acid contents of proteins and their nutritive values for the rat. J. Biol. Chem. 163:599 620. National Research Council, 1994. Nutrient Requirements of Poultry. 8th rev. ed. National Academic Press, Washington, DC. Price, W. H., W. M. Taylor, and W. C. Russell, 1953. The retention of essential amino acids by the growing chick. J. Nutr. 51:413 422. Robbins, K. R., 1986. A method, SAS program, and example for fitting the broken-line to growth data. The University of Tennessee Agricultural Experiment Station Research Report 86-09, Knoxville, TN. Robel, E., and H. Menge, 1973. Performance of chicks fed an amino acid profile based on carcass composition. Poultry Sci. 52:1219 1221. Stas, R. J., and L. M. Potter, 1982. Deficient amino acids in a 22% protein corn-soybean meal diet for young turkeys. Poultry Sci. 61:933 938. Summers, J. D., and H. Fisher, 1961. Net protein values for the growing chicken as determined by carcass analysis: exploration of the method. J. Nutr. 75:435 442. Wang, T. C., and M. F. Fuller, 1989. The optimum dietary amino acid pattern for growing pigs. I. Experiments by amino acid deletion. Br. J. Nutr. 62:77 89. Warnick, R. E., and J. O. Anderson, 1973. Essential amino acid levels for starting turkey. Poultry Sci. 52:445 452. Williams, H. H., L. V. Curtin, J. Abraham, J. K. Loosli, and L. A. Maynard, 1954. Estimation of growth requirements for amino acid by assay of the carcass. J. Biol. Chem. 208: 277 286.