- Email: [email protected]
Nutritive Value of Poultry By-Product Meal
Nutritive Value of Poultry By-Product Meal.2. Comparisons of Methods of Determining Protein Quality R. R. ESCALONA P., G. M. PESTI, 1 and P. D. VAUGHTERS Department of Poultry Science, University of Georgia, Athens, Georgia 30602 (Received for publication July 24, 1985)
1986 Poultry Science 65:2268-2280 INTRODUCTION Phillips ( 1 9 8 1 , 1 9 8 2 ) developed t h e saturat i o n kinetics (SK) m o d e l for fitting n u t r i e n t response data (Morgan et al., 1 9 7 5 ) into a bioassay for estimating p r o t e i n quality. He suggested t h a t t h e SK m e t h o d was better t h a n previously used bioassays with rats. An objective of these studies was to c o m p a r e results from SK bioassays t o several o t h e r bioassays for t h e protein quality of p o u l t r y b y - p r o d u c t meal (PBPM) with t h e chick. After a suitable assay was identified, t h e protein quality of several PBPM samples was d e t e r m i n e d . Chemical assays for protein quality were also performed o n t h e same samples. T h e m e t h o d s of evaluating p r o t e i n quality used h e r e have application in ranking t h e relative values of different samples of t h e same p r o d u c t or different p r o d u c t s . As p o i n t e d o u t b y Harper ( 1 9 8 1 ) , t h e y are n o t used in practical feed f o r m u l a t i o n , n o r d o t h e y indicate w h a t
'To whom correspondence should be addressed.
c o m p l e m e n t a r y value t w o or m o r e proteins m a y have. T h e m e t h o d s t o be c o m p a r e d w e r e : Saturation Kinetics Model. (Morgan et al., 1 9 7 5 , as proposed b y Phillips, 1 9 8 1 , 1982). Phillips ( 1 9 8 1 ) suggested t h a t t h e r e was s o m e value in k n o w i n g t h e response of test animals over a wide range of n u t r i e n t intakes. T h e response data could be fitted to a single fourp a r a m e t e r sigmoid curve, with each p a r a m e t e r of curves of different protein sources being c o m p a r e d to indicate p r o t e i n quality. Data were derived from t h e present studies so t h a t SK curves could be fitted t o t h e same d a t a as t h e o t h e r m e t h o d s of evaluation. Slope-Ratio Assay (SR). (Hegsted and Chang, 1 9 6 5 a , b ) . These a u t h o r s felt t h a t there was s o m e t h i n g inherently valid a b o u t comparing t h e slopes of response lines depicting t h e relationship b e t w e e n b o d y weight gain and nitrogen intake. Because t h e y were making linear a p p r o x i m a t i o n s of sigmoid functions, t h e i n h e r e n t value of t h e t e c h n i q u e is n o t obvious. Proteins were ranked t h e same b y t h e SR assay and t h e m o r e traditional protein efficiency ratio m e t h o d in Hegsted and Chang's studies.
2268
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
ABSTRACT Studies were conducted to compare several methods of determining the protein quality of poultry by-product meal (PBPM). A sample of PBPM was compared to a soybean meal (SBM) sample fortified with .5% of L-methionine. Each was incorporated into corn starch and poultry oil-based diets at 2.5% increments between 0 and 30% protein. Diets were fed for 7 days starting when the chicks were 8 days old. Slope-ratio and saturation kinetics models were fitted for the two protein sources. Protein efficiency ratio (PER), net protein ratio, and net protein utilization were calculated for each diet. The PER was the most discriminating method of estimating protein quality, especially at lower levels (2.5% protein: PBPM = .65, SBM + methionine (Met) = 4.23). At higher levels, no differences could be detected (30% protein: PBPM = 2.45, SBM + Met = 2.31). The 6% protein level was chosen for further studies because there were large differences in PER, and the chicks were in positive nitrogen balance. For eight samples of PBPM (from three different processing facilities), Kjeldahl N ranged from 8.64 to 10.22% gross energy (GE) from 4.49 to 5.36 kcal/g; Ca from 3.6 to 8.84%, and P from 2.04 to 2.55%. A Monday morning sample (fresh broiler waste) was compared to a Friday afternoon sample from the same processing facility (broiler and hatchery waste, and DAF sludge) in semipurified diets at 6% protein. Significant differences in gain were detected in two experiments, but PER in only one. For the five samples tested by all methods, similar rankings were found by PER, pepsin digestibility, total lysine and Carpenter's available lysine. (Key words: Slope-ratio assay, protein efficiency ratio, net protein ratio, net protein utilization, pepsin digestibility, lysine availability)
POULTRY BY-PRODUCT MEAL
Available Lysine (ah YS). This was according to the measurement by Carpenter (1960). The binding of free e-amino groups of lysine with Sanger's reagent (fluorodinitrobenzene) was measured to indicate the lysine that has not
been rendered unusable to the bird by processing or overheating. Pepsin Digestibility (PD). (Adapted by Johnstone and Coon 1979). This is an in vitro measure of the ability of a purified enzyme to hydrolyze the protein present. MATERIALS AND METHODS
General. Day-old male broiler chicks from a commercial hatchery were used in five experiments. Chicks were housed in temperature-controlled battery brooders with raised wire floors and constant illumination. Diets in mash form and water were supplied on an ad libitum basis. Mortality was recorded daily and feed consumption data adjusted to account for any deaths, on a chick-day basis. Dietary protein (N X 6.25) was determined by the method of the Association of Official Analytical Chemists (AOAC, 1970). Percent moisture was determined by drying in a forced draft oven at 105C for 16 hr. Poultry by-product meals were obtained from three processing facilities in the Southeastern United States. They have previously been described (Pesti et al.). Samples # 1 , #2, and #3 were random samples taken from Plant A (continuous flow). Sample #4 was obtained through regular commercial channels and was reported to be from Plant B (continuous flow). Samples #5, #6, and #7 were from Plant B and were obtained on Monday morning (broiler waste only; #5), Monday afternoon (broiler and hatchery waste; #6), and Friday afternoon (broiler and hatchery wastes and DAF sludge, #7). Sample #8 was obtained from a plant with batch processing of broiler waste only, no hatchery waste or DAF sludge. Chemical Analysis. Pepsin digestibility was determined by the method of the AOAC (1970) with the exception that .002% pepsin was used (Johnstone and Coon, 1979). Available lysine was determined by the method of Carpenter (1960). Amino acid analyses were conducted by another laboratory using a modification of the method of Moore et al. (1958). Descriptive statistics, analysis of variance, and orthogonal comparisons were calculated using the general linear model procedures described in the SAS User's Guide (SAS, 1982). Duncan's new multiple range test was used to separate means (SAS, 1982). Experiment 1. Four hundred-eightly male chicks were randomly distributed into 80 pens
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
Protein Efficiency Ratio (PER). A method of expressing numerically the growth-promoting value of proteins was proposed by Osborne et al. in 1919. Protein efficiency ratio was defined as "the gain in body weight per gram of protein consumed." It consisted of measuring the growth (gain in weight) of growing rats fed different protein levels, and the highest obtained value was taken as the PER. The accuracy of this method has been frequently criticized on the basis that: 1) the results are influenced by the levels of protein consumed; 2) no allowance is made for the quantity of protein used for maintenance; and 3) gain in body weight does not necessarily correspond to gain in body protein (Mitchell, 1924; Hegsted and Worcester, 1947; Bender, 1956; Bender and Doel, 1957; Summers and Fisher, 1961; Harper, 1981). Net Protein Ratio (NPR). A modification of the original PER was proposed by Bender and Doel (1957). They included in the experiment a group of animals consuming a nonprotein diet for a similar period of time. The calculations involved the difference in weight between the groups, instead of only body weight gain. Net protein ratio was believed to be more precise and reproducible and to overcome the basic weakness of PER (Bender and Doel, 1957; Jansen, 1978), as it measures the protein utilized for maintenance as well as growth. Bender and Doel (1957) stated that NPR values are not influenced by food intake. As with any growth method, it assumes that the gain in body weight is indicative of the new protein tissue, which is not always valid. The procedure was defined by its originators as gain in weight of the test group plus loss of weight of proteinfree group, divided by the protein consumed. Net Protein Utilization (NPU). A derivation from the Thomas-Mitchell procedure (Mitchell, 1923) for the biological value (BV) determination is the NPU method (Bender and Miller, 1953; Miller and Bender, 1955). Miller and Bender (1955) defined it as the difference of body nitrogen content between a group fed a protein diet and a group fed a protein-free diet, expressed as a percentage of the nitrogen intake of the test group. The BV can be estimated by dividing NPU by a digestibility coefficient (Miller and Bender, 1955; Allison, 1964).
2269
2270
ESCALONA ET AL.
nitrogen. Carcasses were then ground by blender homogenization and refrozen until nitrogen and water determinations were conducted. The SK model was fitted to the data for each protein source: bKi + r m a x i n Kl + In where r = body weight gain in grams, I = nitrogen intake (g), b = intercept, r m a x = asymptotic value of r when I -» °°, and n = apparent kinetic order of the response when I -> O, and Ki is the "nutrition constant." The SR model was fitted to the data from protein levels of 15% or less by the equation: r = b0 + M S B M + MPBPM
TABLE 1. Composition of basal, summit, and control diets used in Experiment 1 (%)
Ingredients G r o u n d yellow corn Corn starch Poultry b y - p r o d u c t meal Soybeal meal, dehulled DL-Methionine L-Methionine Poultry oil Limestone, ground Phosphate, defluorinated Salt, NaCl Potassium chloride Selenium p r e m i x 3 Vitamin p r e m i x 4 Mineral p r e m i x 5 Composition b y calculation ME kcal/g Protein, % Calcium, % Phosphate, %, n o n p h y t a t e Methionine
Protein free
PBPM 1
Soybean2 meal
Practical control 57.50
89.43
44.02 49.59
30.18 60.91
6.0
6.0
3.89
5.0 31.0 .15
.31 6.0 .44 1.78 .08
6.0 .65 1.75 .40
.38
.10
.25 .05
.25 .05
.25 .05
.05 .25 .05
3.81 0 1.28 .72
3.47 30.0 2.95 .84 .50
3.12 30.0 .99 .47 .74
3.12 23.0 1.09 .51 .54
1
Poultry by-product meal Sample 4.
2
Mixture of 99.5 parts of soybean meal plus .5 part of L-Methionine.
3
Provides: .1 mg selenium/kg diet as sodium selenite.
* Provides (per kg diet): Vitamin A, 5500 IU; vitamin D 3 1100 ICU; vitamin E, 11 IU; riboflavin, 4.4; Ca panthotenate, 12 mg; nicotinic acid, 44 mg; choline CI, 220 mg; vitamin B 1 2 , 6.6 Mg; vitamin B6 2.2 mg; menadione, 1.1 mg (as MSBC); folic acid, .55 mg; d-biotin, .11 mg; thiamine, 2.2 mg (as thiamine mononitrate); ethoxyquin, 125 mg. 'Provides (ppm diet): Mn, 60; Zn, 50; Fe, 30; Cu, 5; I, 1.05; Ca, 75 (minimum), 90 (maximum).
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
of six birds each. Three replicates (pens) were used for each of the test rations, which also included a nonprotein diet. Chicks were fed a practical starter ration (Table 1) for the first 7 days, then fasted overnight. Experimental diets were provided from the 8th to the 14th day. A protein-free basal diet was mixed with the summit diets (Table 1) containing a poultry by-product meal (Sample #4, 30% protein) or a mixture of soybean meal (SBM) plus methionine (Met) (30% protein). Each was incorporated in appropriate quantities into a series of 12 diets (Table 2) such that the nitrogen content varied from .4 to 4.8% (2.5 to 30% protein, N x 6.25). On the 14th day, feed was removed for a 6 hr period to allow the intestinal tract to empty. After this, two birds/pen were killed by cervical dislocation (without loss of blood) for carcass analysis. The carcasses were frozen, cut into small pieces, and cooled further in liquid
P O U L T R Y BY-PRODUCT M E A L
where b 0 = intercept, I = nitrogen intake, and bi and b 2 are the coefficients for SBM and PBPM, respectively. The three other models for assessing protein nutritional quality were computed by the following formulas:
a pretreatment period of 8 days was used. The experimental diets were fed on Days 9 to 14. All diets were prepared to contain a constant protein level of 6%, based on determined protein content (Table 3). Five pens of 10 male chicks each were fed each diet in Experiment 2, eight pens of 10 male chicks each in Experiment 3 (Table 4).
Gain in body weight PER;
2271
Protein consumed RESULTS
Gain test group + loss protein-free group NPR
Body nitrogen test group — body nitrogen protein-free group Nitrogen consumed
Experiments 2 and 3. The procedures followed in these two trials were similar to that in Experiment 1, except that in Experiment 3,
T A B L E 2. Summary of intake-response data of chicks fed diets containing various levels of poultry meal or 199 parts soybean meal plus 1 part L-methionine or a protein-free diet (Experiment Dietary protein
Dietary nitrogen
Feed consumed
Nitrogen consumed
(n>)
by-product 1)
Nitrogen retained 1
Body 'weight gain
-.37 ±
.04
-11 ±
-.17 .17 .47 .79 1.24 1.92 2.27 2.61 2.88 3.20 3.58 3.63
.02 .06 .07 .02 .11 .14 .26 .05 .14 .16 .13 .13
-2 13 26 37 57 83 96 107 111 123 133 131
(g) Protein-free diet
.0
.0
105 ±
5
0
±0
2
Poultry b y - p r o d u c t meal-based diets 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0
.4 .8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8
2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0
.4 .8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8
107 ± 124 ± 130± 135 ± 163 ± 171 ± 187 ± 186 ± 182 ± 199 ± 175 + 179 ±
4 2 5 3 6 10 11 7 4 35 2 13
.43 .99 1.56 2.17 3.26 4.09 5.24 5.97 5.09 7.94 7.69 8.57
± .02 ± .01 ± .05 + .11 ± .11 + .24 ± .30 ± .21 ± .25 ± :1.38 ± .11 ± .64
+ ± + ± + + ± ± ± ± + ±
± 1 ± 2 ± 3 ± 3 + 5 ± 6 ± 11 ± 2 ± 2 ± 6 ± 5 ± 5
Diets based on 199 parts soybean meal plus 1 part L-methionine 118± 7 158 ± 8 203 ± 16 201 ± 2 237 ± 7 245 ± 6 245 + 5 228± 5 244+ 8 225 ± 9 226 + 10 2 3 4 ± 19
.47 1.26 2.44 3.20 4.73 5.89 6.86 7.31 8.80 9.00 9.94 11.23
+ ± ± ± ± ± ± ± ± ± ± ±
.03 .06 .19 .04 .15 .15 .15 .17 .29 .35 .44 .91
.18 .85 1.83 2.32 3.38 3.66 4.21 4.31 4.81 4.42 4.35 4.50
± .02 ± .11 ± .25 ± 6.96 ± .16 ± .37 ± .09 ± .03 ± .22 ± .25 ± .36 ± .35
12± 41 ± 82 ± 102 + 145 + 155 ± 175 + 175 ± 191 ± 171 ± 163 ± 162 ±
1 5 11 9 7 15 4 1 9 10 14 13
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
NPU =
Experiment 1. At lower levels of dietary protein, carcass nitrogen deposition appeared to be fairly constant. After a 10% level of dietary protein was reached, the rate of carcass nitrogen increased (Fig. 1). The quadratic regression equation describing the effect of dietary protein content ( X t ) on the percent carcass nitrogen (Y) was determined:
Protein consumed
2272
ESCALONA ET AL. TABLE 3. Composition of experimental diets used in Experiment 2 (%)
Poultry by-product meal 1 Ingredients
1
Soybean meal + methionine
9.64
12.37
12.31
80.67 6.00 .073
76.93 6.00 1.78
76.93 6.00 1.78
1.53
2.46
2.46
.25 .05 .31
.25 .05
.25 .05
.12
.12
11.11 .11
80.40 6.00 1.59 1.7 .25 .05 .32
4.1 6 1.5
4.05 6 1.5
3.9 6 1.5
.52
.52
3.9 6 1.5
.52
.52
.089
.199
Described by Pesti et al. (1986).
2
Provides (per kg diet): Vitamin A, 5500 1U; vitamin D 3 , 1100 1CU; vitamin E, 11 IU; riboflavin, 4.4; Ca panthothenate, 12 mg; nicotinic acid, 44 mg; choline Cl, 220 mg; vitamin B 1 2 , 6.6 Mg; vitamin B 6 , 2.2 mg; menadione, 1.1 mg (as MSBC); folic acid, .55 mg; d-biotin, .11 mg; thiamine, 2.2 mg (as thiamine mononitrate); ethoxyquin, 125 mg. 3
Provides (ppm diet): Mn, 60; Zn, 50; Fe, 30; Cu, 5; I, 1.05; Ca, 75 (minimum), 90 (maximum).
TABLE 4. Composition of experimental diets used in Experiment 3 (%) Poultry by-product meal 1 Ingredients Poultry by-product meal Corn starch Poultry oil Limestone, ground Phosphate, defluorinated Vitamin premix 2 Mineral premix 3 Potassium chloride Composition by calculation ME, kcal/g Protein, % Calcium, % Phosphate, %, nonphytate 1
1
4
5
7
8
9.5
9.9
82.1
81.0
9.64 80.55
11.11 80.78
9.65 80.65
6.0
6.0
6.0
1.18 1.44
1.53 1.78
6.0 .05
6.0
1.52 1.56
1.56
.25 .05 .2
.25 .05 .2
.25 .05 .2
1.5 .25 .05 .2
4.1 6 1.5 .52
4.1 6 1.5 .52
4.1 6 1.5 .52
.25 .05 .2 4.1 6 1.5 .52
1.39
4.1 6 1.5 .52
Described by Pesti et al. (1986).
2
Provides (per kg diet): Vitamin A, 5500 IU; vitamin D 3 1100 ICU; vitamin E, 11 IU; riboflavin, 4.4; Ca panthothenate, 12 mg; nicotinic acid, 44 mg; choline Cl, 220 mg; vitamin B 1 2 , 6.6 Mg; vitamin B 6 2.2 mg; menadione, 1.1 mg (as MSBC); folic acid, .55 mg; d-biotin, .11 mg; thiamine, 2.2 mg (as thiamine mononitrate); ethoxyquin, 125 mg. 3
Provides (ppm diet): Mn, 60; Zn, 50; Fe, 30; Cu, 5; I, 1.05; Ca, 75 (minimum), 90 (maximum).
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
Poultry by-product meal Soybean meal DL-Methionine Corn starch Poultry oil Limestone, ground Phosphate, defluorinated Vitamin premix 2 Mineral premix 3 Potassium chloride Salt (NaCl) Composition by calculation ME energy, kcal/g Protein, % Calcium, % Phosphate, % nonphytate Methionine
Soybean meal
7
5
POULTRY BY-PRODUCT MEAL
3.0
•
2.9 2.8 O O
• •
2.7 i
2.2
•
en <
<
2.4
•
•
•
2.1 2.0 0
i
i
i
5
10
15
i
20
i
i
25
30
% DIETARY PROTEIN FIG. 1. Relationship between the carcass nitrogen of chicks and dietary protein level (Experiment 1).
Y = 2.43 3 - .0072XJ + .00052Xi (52.79)* (-1.03X2.34)* The values in parenthesis are Student t-values and the asterisk indicates significance at a 5% level. The coefficient of determination (R 2 ) was .34. Mean values were used in calculating NPU, NPR, SK, and SR models. Summaries of intake-response data and PER, NPR, and NPU values as determined for both protein sources are shown in Tables 2 and 5, respectively. Weight gains of chicks increased with an increase in protein level for each protein source (Fig. 2). Gains were consistently higher for SBM + Met. Maximum gain values were reached at the 27.5% level for PBPM and the 22.5% level for SBM + Met, but at higher protein levels the gains decreased (Fig. 2). Feed consumption increased with an increase in protein level for each protein, with greater intakes for SBM + Met (Table 2). Maximum intakes were seen at approximately 17.5%
protein for both protein sources. Nitrogen intake and nitrogen retention (Table 2) increased linearly for both proteins, but greater values were observed for SBM + Met. For the saturation kinetics model (Table 6), there were no obvious differences in intercepts (b), the kinetic order of the response as intake approaches zero (n), or the maximum response ( r m a x ) ; there was considerable overlapping of the confidence limits. There was, however, a considerable difference in the nutrition constant (Kj) with the better quality protein having the lower Kj value, as would be expected from the formula. In the slope-ratio analysis, SBM + Met was clearly the better quality protein with a slope ratio of 1.95. The three conventional indexes (Table 5; Figs. 3 to 5) indicated that SBM + Met was a better protein source at the lower levels. The PER showed a greater difference between protein sources than did either of the other measures. Variable PER values were observed; they increased up to a maximum (15% PBPM and 7.5% SBM + Met) and then began to decline (Fig. 3). On the other hand, NPR and NPU values appeared to be relatively constant, especially for PBPM. Similarly, significant curvilinear (quadratic) relationships (P<.01) between PER or NPU and protein level were found; this relationship was not significant (P> .29) for NPR. Significant correlations between PER and NPU (r = .79) and PER and NPR (r = .72) were found (Table 7). The correlation between NPR and NPU, although significant, was lower (r = .63). The NPU increased from the 2.5% level up to approximately 10% for both proteins, being more marked for SBM + Met (Fig. 4). While NPU values for PBPM remained relatively constant above that level, SBM + Met showed a progressive decline in NPU as the dietary protein level increased. The PER and NPR gave different responses. Protein efficiency ratio rises to a maximum for both proteins. Beyond this maximum, PER came to plateau and then declined slowly at higher intakes. The NPR, although relatively constant for PBPM, declined for SBM + Met as dietary protein increased (Fig. 5). Experiments 2 and 3. As expected, PER values were higher for the reference protein (soybean meal) than PBPM; and a significant difference was observed due to L-methionine supplementation (Table 7). No significant differences in PER were observed between
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
2.3'
• • • •• t • • • • • • • • • • • • • • • • • • • !
2.5
o a:
x
'v'
2.6
2273
2
1
± .62 ± .26 ± .39 ± .07 ± .31 ± .01 ± .08 ± .05 ± .12 + .11 ± .02
Mixture of 199 parts of SBM plus 1 part Met.
5.19 5.35 5.09 4.91 4.19 4.08 3.83 3.47 3.03 2.61 2.31
± .38 ± .26 ± .15 ± .12 ± .14 ± .29 ± .07 + .04 ± .32 ± .06 ± .11
2.09 2.62 2.74 2.78 3.26 2.94 2.88 2.72 2.59 2.77 2.45
(X± SEM) 4.28 ± .46
SBM + Met
-.65 ±.39
PBPM
Ratio of (PBPM/SBM + Met) X 100.
10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0
2.5 5.0 7.5
(%)
Protein
1
Protein efficiency ratio
-15 40 49 54 57 78 72 75 78 86 106 106
(%)
Index
2
3.30 3.80 3.71 3.51 3.30 3.67 3.26 3.16 2.98 2.81 2.99 2.65
( + .25 + .39 + .25 ± .14 ± .11 ± .14 ± .29 ± .08 ± .04 ± .35 ± .06 ± .13
PBPM
42 58 61 63 62 82 75 78 81 87 108 108
(%) ± .67 ± .64 ± .21 ± .39 + .06 + .31 ± .02 ± .08 ± .04 ± .12 ± .10 ± .13
A)
7.89 6.54 6.05 5.62 5.27 4.48 4.33 4.06 3.66 3.22 2.78 2.46
Index
SBM + Met
Net protein ratio
TABLE 5. Estimates of the protein quality of poultry by-product meal (PBPM) and a so plus methionine (Met) mixture by various methods (Experiment 1)
ed from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
POULTRY BY-PRODUCT MEAL
210 r
0 O
0 /O
V
-= 150 -
<
0
\ °
180 -
0
o 0
X 120-
o
°4
LJ 90
f
7
Q O 60 CD
/• *
i
A
?
-
J
Protein Source » PBPM SBM
i° r y
\
0 " 5
i
i
i
i
10
15
20
25
i
30
% DIETARY PROTEIN FIG. 2. Relationship between body weight gain and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
chicks fed PBPM Samples #5 or #7. However, responses from these samples were significantly different than those from the SBM and SBM + Met reference proteins.
Chicks fed the soybean meal + Met diet had significantly better weight gains than the others. Chicks fed the PBPM diets had lower gains, especially sample #7. Feed intake was consistently higher with the SBM + Met diet. However, chicks fed PBPM #5 and the SBMbased diet were not significantly different in feed intake (P>.05). Orthogonal comparisons (Table 8) indicated a significant difference (P = .045) in gain, but not in feed efficiency or PER (P = .222), between samples #5 and #7. [P values for feed efficiency and PER are identical since efficiency (gain/consumption) and PER (gain/consumption X .06) are linearly related]. Chemical Assays. Protein efficiency ratio (Table 7), aLYS (Table 8), total lysine (LYS, Table 9), and pepsin digestibility all ranked the samples in similar orders: 5, 8, 1, 4, 7; 1, 8, 5, 4, 7; 1, 8, 5, 4, 7, and 1, 8, 7, 5, 4, respectively. The amino acid pattern of these samples was fairly uniform (Table 9). The patterns were also similar to that in the NRC (1984) nutrient composition tables. Essential amino acids in the PBPM were slightly higher than NRC values with the exceptions of histidine (lower) and tyrosine (much higher). DISCUSSION In Experiment 1, there was a high correlation between PER, NPU, and NPR, which is in agreement with the observations of Block
TABLE 6. Coefficients for the saturation kinetics model and slope-ratio assay fitted to the response data from Table 2 Soybean meal
Poultry by-product meal Confide nee limits
Confidence limits
Saturation kinetics Analysis Intercept Maximum response Nutrition constant Kinetic order Intake At maintenance At inflection point Response At inflection point Slope ratio analysis Intercept 1 Slopes 1
Mean
Lower
Upper
Mean
Lower
Upper
-4.07 202 2.84 1.70
-10.03 190 2.58 1.44
1.88 213 3.1 1.95
-8.36 180 4.17 1.53
-14.16 151 3.26 1.22
-2.56 209 5.08 1.84
-10.7 5.72
-19.3 6.62
.285 1.27
.559 1.49
38.3 -15.5 12.01
24.2 -10.7 11.56
A common intercept was fitted for the two lines.
-19.3 12.46
-15.5 6.17
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
*
>-
30
/ ^
2275
2276
ESCALONA ET AL.
FIG. 3. Relationship between protein efficiency ratio and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
Protein Source PBPM SBM
\
Protein Source •—»PBPM
o
o
0^
o SBM
\
<
or
-°\\ o\
° o
o
>v N
l±J
IO
o o
or OL
t • - •
0
5
10
15
20
25
30
% DIETARY PROTEIN FIG 4. Relationship between net protein utilization and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
0
X
V 8
1k
i 15
i 20
• 1
1
5
10
T 25
i 30
% DIETARY PROTEIN FIG. 5. Relationship between net protein ratio and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
% DIETARY PROTEIN
and Mitchell (1946), Bender (1956), and Henry (1965). The good curvilinear relationship between PER and protein level observed in this study confirms that the finding of Hegsted and Worcester (1947) and Sammonds and Hegsted (1977), with rats, is similar for chicks. Osborne and Mendel (1916), Osborne et al. (1919), and Mitchell (1924) had also reported the influence of protein intake on protein efficiency, although their data covered smaller ranges. Bender (1956) showed the PER variability with feed consumption and the relative constancy of NPU and NPR. Similar observations were made in this experiment. Although subject to the criticisms described earlier, the PER index showed a greater spread than the other techniques. This allowed a better discrimination of protein quality between samples, especially at lower levels. At higher levels, differences could not be detected. Woodham (1967) had reported greater differences at low dietary protein for PER values. Scott et al. (1982) stated the greater sensitivity of chickens to differences in protein quality at
2277
POULTRY BY-PRODUCT MEAL TABLE 7. Protein efficiency ratios (PER), body weight gains, and feed consumption of broilers fed protein sources as 6% protein in semipurified diets (Experiments 2 and 3) PER
Consumption
Gain
Ranking
Mean + SE
source
UP Experiment 2 SBM 1 SBM + Met PBPM # 5 PBPM # 7
4.6 5.3 3.6 3.2
± ± ± ±
.3b .3a .lc .2C
4 8
+
3b,2
68±4a 37 ± 2 C 28 ± l d
176 + 212 ± 168+ 148 ±
4b'5 la 5b 4C
165 154 161 157 164
5a 3a 3a 4a 5a
Experiment 3 #1 #4 #5 #7 #8
3.2 3.1 3.6 2.9 3.5
± .2 ± ,2ab ± .2a ± .lb ±.la
3rd 4th 1st 5th 2nd
31 29 35 27 35
± ± ± ± ±
2ab b 2
3a lb la
± ± ± ± ±
O r t h o g o n a l c o m p a r i s o n s (significan<:e probabilities) Experiment 2
PBPM 5 vs. 7
a,b,c
Gain
PER
.045
.222
Means with the same letter are not significantly different (P>.05; Duncan's new multiple range test).
SBM = Soybean meal, PBPM = Poultry by-product meal. Mean + SE.
lower levels. These authors recommend using 10% protein in the diet, but our results (Table 5 and Fig. 3) indicate that differences may be obscured at this level. Protein efficiency ratio is well correlated to the other methods, NPU, and NPR. Protein efficiency ratio is simpler to determine. No time is consumed in preparing carcasses for analysis (as in NPU), and the inclusion of a group fed a protein-free diet is not neces-
sary (as for NPU and NPR). The wide use and better protein quality discrimination of PER in this study indicate that it offers a simpler and more reproducible procedure for evaluating protein quality. All these considerations led us to choose the PER method as the most useful one to be used in further experiments. Any overriding advantage of either the SK or SR assays is not apparent from our studies. Many more data points are needed for these
TABLE 8. Chemical estimations of the quality of protein from several poultry by-product meals. Sample no.
Available lysine
Plant (g/100 g) A A A B? B B B C
3.03 2.85 3.12 2.19 2.50 2.52 2.05 2.52
± ± ± ± ± ± ± ±
.15' . 0 8 'ab .11' . 0 2 'cd .12,'be .14 1be .15' .10 1be
Ranking
Pepsin digestibility
2nd 3rd 1st 7th 6th 4th 8th 5th
87.0 89.9 86.4 77.3 78.0 75.4 79.0 85.8
( g / 1 0 0 g protein 4.82 ± . 2 4 a b 4.46±.13abc 4.92 ± . 1 7 a 3.63 ± . 0 3 d 4.00 ± 1 8 ^ d 4.17 ± . 2 3 b ™ 3.80 ± . 1 8 c d 4.05 ± . 1 5 d
Ranking
(g/100 g) ± ± ± ± ± ± ± ±
.5a .4a ^ab .0e
•°r • 2 di
.5 .lb
Means ± SE within columns with different superscripts are significantly different (P<.05).
1st 2nd 3rd 7th 6th 8th 5th 4th
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
PBPM PBPM PBPM PBPM PBPM
ab
5 S 5 S o nJ, 2 . x a g- o SJ S" 5 & a « ^ 2 .x 3 p - n > ° x " S 2 . 3 = • ^ - I !•%
rt
5 =• 3 5 S 3 " =
to w w 4i M sj I-* ujojN))---vjNJO-f».4i.4^ WOvUiwixw^lv)w'w>^Ot-1lw^Oyiv0^vo'tO \ O O \ v ) O v i - 1 ^ U i U i \ O H W O ^ w O 0 0 N t O W N ) l - '
in
to to -t* i-» to
NJ
NJ
to
-F* i- 1 t o
>-*
i-* ^
t o t o •£• i- 1 t o i-*
W ^ t O M ^ h J O v ^ - ^ - t '
W
W
W
H
^
W
O
V
-
^
^
^
w i w N j H v ) w y \ y j ^ w
wsawONoowMvoNNOvobHobbe^i^bo
H^SO^OOWHOOWMWHOO^Ul^OHHOJVOW
t-*
tO
O
>-»
h - t O U J H-i W
W W | g M M t O U l w 4 » W
W
W
M
M
a
t
O
U
l
W
^
W
w ^l u i ^ b b o b v w b o ^ w ' ^ v i ^ b ^ u ' u i y i i - ' ys i ! -
M
W C O ' s l 0 0 H U i ^ O v ^ t O ^ l H s l \ O 0 0 t O U l W O
t o t o 4* >-» t o
t o tO 4^ <-- t o
•-*
i- 1
J
w W N ) H < I W O \ ^ ^ ^
W
W
W
M
^
w
a
^
^
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
to
to
y j i > y i t o b Q w \ o w b \ I - ' t j j ^ \ o b to N© In bo U» bo to,
ON
O
w t o t o 4^ t o t o •-* >-* ui4itoi-''<]UJON4»-t^^w O J O C O Wt W ' \ I O V4 l w b N J M N ) t O i - 1 b \ W U t s W vv l \ 0 C O O s i - a ^ W ' - N M ^ U i O ^ O v 4 ' U i O ' - ' J O W
N) tO ^
t- 1 tO
tO
•£»
© O
to
OvitoyjwNiOMODyii-'O-f'CoovOvooMOvOow
H-
^
to o j ^ b \ S i l n s ) b o u i b o h J N i N j ^ o t o l i ^ ' v 4 4*>--'bNb O 1+
- J ^ i - ' t O t O P M \ O O \ u i t o 0 0 i - » W ' s j t o \ O O i - ' 0 \ a v O 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ ,
i
t o O i w - ^ w ^ o ^ w O - f ' M O ^ V i w O O ' O ' - ' O ^ O
©
b b b b H b b b
to
00 ©
to to ON U\ © O w O
H b b ' * ' b b b w H
to w Lfi h-» O N © -£• O VO O
to O J to >— ^J -v] O Wl
' H H i - b
(M Ul -^ ON ^O © 00 O ©
O © © O
I V X3 VNCHVDS3
POULTRY BY-PRODUCT MEAL
Sample #5 as a better protein source when five samples of PBPM were compared. Sample #7 showed the lowest PER values, while the other three samples were intermediate between Samples #5 and #7. The reproducibility of PER was demonstrated when the mean PER value for Sample #5 was exactly the same in Experiments 2 and 3. Chapman et al. (1959) and Henry (1965) have also reported good reproducibility for the PER of casein with rats. Although in Experiment 2 the higher PER values were obtained at higher feed intakes, in Experiment 3, no relation between PER values and feed intake was observed, because consumption was not significantly different between the experimental diets. Mitchell (1924) observed the highest protein efficiency in rats with larger intakes, similar to the results of Experiment 2. It is not clear why there were differences in the pattern of consumption between Experiments 2 and 3. Perhaps a change in the chicks or environment caused the more uniform consumption in Experiment 3. Ingredient sources were identical. Random effects may have been the cause. The rankings of these samples were very similar for PER, LYS, aLYS, and pepsin digestibility if Sample #5 is omitted from the analyses. Sample 5 is an enigma, being ranked first by PER, third by LYS and aLYS, and fourth by pepsin digestibility, among samples ranked by all four methods. The amount of the limiting amino acid will normally determine the protein quality of a feedstuff. It may be that lysine is not the first limiting amino acid in these samples, or that some other amino acid may interefere with its utilization. An advantage of one of these methods over the others for routine analysis is not apparent from these studies. The PER determinations are time and resource intensive, and differences in PER do not necessarily indicate differences in performance when the feedstuffs are incorporated into practical diets at low levels. Of the chemical methods, Carpenter's available lysine procedure requires the least-specialized equipment (a fume hood, reflux apparatus, and spectrophotometer), and should indicate if nutritional value has been lowered by overheating. Pepsin digestibility is simpler if the specialized apparatus is available. It indicates if poor quality proteins were used in making the product. Total amino acid analysis is the most difficult, requiring specialized equipment, but it yields the most usable data for feed formulation.
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
assays. A large advantage would be necessary to justify them. As Phillips (1982) has noted, there are theoretical reasons not to compare r max o r b values, yet extra data has to be generated to fit them in order to estimate the other parameters. The effect of better quality proteins having lower Kj values also complicates interpreting the results. The PER and SR assays are very similar, with both relating body weight gains to nitrogen (or protein) intake. The only real difference is that the SR values are pooled over a range of intakes, and PER is computed for one intake level. The arbitrary value generated by the SR assay was very similar to an arbitrary value made by a ratio of the PER values at 7.5% protein (1.94 vs. 2.04) in this experiment. The PER values have just as much inherent value, in our opinion, and are derived with considerably less work. The almost constant nitrogen deposition (on a percent of carcass basis) at lower protein levels (Fig. 1) may be ascribed to the predominant function of dietary nitrogen in maintaining the nitrogenous integrity of the tissues. Once nitrogen equilibrium was reached, nitrogen was used in increasing proportions for growth. The significant quadratic effect of protein level determines its contribution to the model, indicating that this curvilinear effect of protein level should be considered in predicting nitrogen deposition in growing chicks. The choice of a 6% protein level used in the semipurified diets fed in Experiments 2 and 3 was made somewhat arbitrarily. The better discrimination between different proteins by PER at the lower levels was the basis for this decision. Another important consideration in choosing this level was the fact that the chicks were in positive nitrogen balance, an important consideration to some researchers. The significant difference in PER and gain between SBM- and PBPM-fed chicks indicated that SBM is a superior protein source (Table 7). As expected, better results were obtained when SBM was supplemented with L-methionine. This established that the assay was sensitive enough to detect this known difference in protein quality. Although PER values of chicks fed the two PBPM were not significantly different, gains and feed intakes were. This suggests a better balance in amino acid composition or a greater availability in Sample #5 than in #7, or perhaps a toxic factor in the DAF sludge in Sample #7. Again, PER ranked
2279
2280
ESCALONA ET AL.
However, availability and digestibility are n o t considered. Each assay m a y be a p p r o p r i a t e u n d e r specific circumstances. ACKNOWLEDGMENTS
REFERENCES Allison, J. B., 1964. The nutritive value of dietary protein. In Mammalian Protein Metabolism II: 41. H. N. Monro, and J. B. Allison, ed. Academic Press, New York, NY and London, England. Association of Official Agricultural Chemists, 1970. Official Method of Analysis, 11th ed., Washington, DC. Bender, A. E., 1956. Relation between protein efficiency and net protein utilization. Br. J. Nutr. 10:135-143. Bender, A. E., and B. H. Doel, 1957. Biological evaluation of protein: a new aspect. Br. J. Nutr. 11:140-147. Bender, A. E., and D. S. Miller, 1953. Constancy of the N / H 2 0 ratio of the rat and its use in the determination of the net protein value. Proc. Biochem. Soc. Biochem. J. 53:vii-viii. (Abstr) Block, R. J., and H. H. Mitchell, 1946. The correlation of amino acid composition of proteins with their nutritive value. Nutr. Abstr. Rev. 16:249-278. Carpenter, K. J., 1960. The estimation of the available lysine in animal protein foods. Biochem. J. 77:604-610. Chapman, D. G., R. Castillo, and J. A. Campbell, 1959. Evaluation of protein in foods. 1. A method for the determination of protein efficiency ratios. Can. J. Biochem. Physiol. 37: 679-68'6. Harper, A. E., 1981. McCollum and directions in the evaluation of protein quality. J. Agric. Food Chem. 29:429-435. Hegsted, D. M., and Y. Chang, 1965a. Protein utilization in growing rats. I. Relative growth index as a bioassay procedure. J. Nutr. 85:159-168. Hegsted, D. M., and Y. Chang, 1965b. Protein utilization in growing rats at different levels of intake. J. Nutr. 87:19-25. Hegsted, D. M., and J. Worcester, 1947. A study of the relation between protein efficiency and gain in weight on diets of constant protein content. J. Nutr. 3 3:685-693. Henry, K. M., 1965. A comparison of biological methods with rats for determining the nutritive
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
S u p p o r t e d b y State and Hatch funds allocated t o t h e Georgia Agricultural E x p e r i m e n t Stations of t h e University of Georgia. T h e a u t h o r s gratefully acknowledge L. O. Faust and M. A. S t a t o n , for their able technical assistance, and F. J. Ivey and Monsanto C o m p a n y , Nutrition Chemicals Division, Chesterfield, MO, for providing t h e a m i n o acid analyses.
value of proteins. Br. J. Nutr. 19:125-135. Jansen, G. R., 1978. Biological evaluation of protein quality. Food Technol. 32:52-56. Johnstone, J., and C. N. Coon, 1979. The use of varying levels of pepsin for pepsin digestion studies with animal proteins. Poultry Sci. 58: 1271-1273. Miller, D. S., and A. E. Bender, 1955. The determination of the net protein utilization of proteins by a shortened method. Br. J. Nutr. 9:382-388. Mitchell, H. H., 1923. A method of determining the biological value of protein. J. Biol. Chem. 58: 873-903. Mitchell, H. H., 1924. The nutritive value of proteins. Physiol. Rev. 4:424-478. Moore, S., D. H. Spackman, and W. H. Stein, 1958. Chromatography of amino acids on sulfonated polystyrene resins. Anal. Chem. 30:1185-1190. Morgan, P. H., L. P. Mercer, and N. W. Flodin, 1975. General model for nutritional responses of higher organisms. Proc. Natl. Acad. Sci. USA. 72:4327. National Research Council, 1984. Nutrient Requirements of Poultry. Nutrient Requirements of Domestic Animals. Natl. Acad. Sci., Washington, DC. Osborne, T. B., and L. B. Mendel, 1916. A quantitative comparison of casein, lactalbumin, and edestin for growth or maintenance. J. Biol. Chem. 26:1-23. Osborne, T. B., L. B. Mendel, and E. L. Ferry, 1919. A method of expressing numerically the growth promoting value of protein. J. Biol. Chem. 37:223-229. Pesti, G. M., L. O. Faust, H. L. Fuller, N. M. Dale, and F. H. Benoff, 1986. Nutritive value of poultry by-product meal. 1. Metabolizable energy. Poultry Sci. 65: (in press). Phillips, R. D., 1981. Linear and nonlinear models for measuring protein nutritional quality. J. Nutr. 111:1058-1066. Phillips, R. D., 1982. Modification of the saturation kinetics model to produce a more versatile protein quality assay. J. Nutr. 112:468-473. Sammonds, K. W., and D. M. Hegsted, 1977. Animal bioassay: A critical evaluation with special reference to assessing nutritive value for die human. Pages 68-79 in Evaluation of Protein for Humans. C. E. Bodwell, ed. The AVI Pub. Co., Westport, CT. SAS Institute, Inc., 1982. SAS User's Guide: Basics. SAS Inst. Cary, NC. Scott, M. L., M. C. Nesheim, and R. J. Young, 1982. Nutrition of the Chicken. M. L. Scott Associates, Publishers, W. F. Humphrey Press, Inc., Geneva, NY. Summers, J. D., and H. Fisher, 1961. Net protein values for the growing chickens as determined by carcass analysis: Exploration of the method. J. Nutr. 75:435-442. Woodham, A. A., 1967. A chick growth test for the evaluation of protein quality in cereal-based diets. 1. Development of the method. Br. Poult. Sci. 9:53-63.
1986 Poultry Science 65:2268-2280 INTRODUCTION Phillips ( 1 9 8 1 , 1 9 8 2 ) developed t h e saturat i o n kinetics (SK) m o d e l for fitting n u t r i e n t response data (Morgan et al., 1 9 7 5 ) into a bioassay for estimating p r o t e i n quality. He suggested t h a t t h e SK m e t h o d was better t h a n previously used bioassays with rats. An objective of these studies was to c o m p a r e results from SK bioassays t o several o t h e r bioassays for t h e protein quality of p o u l t r y b y - p r o d u c t meal (PBPM) with t h e chick. After a suitable assay was identified, t h e protein quality of several PBPM samples was d e t e r m i n e d . Chemical assays for protein quality were also performed o n t h e same samples. T h e m e t h o d s of evaluating p r o t e i n quality used h e r e have application in ranking t h e relative values of different samples of t h e same p r o d u c t or different p r o d u c t s . As p o i n t e d o u t b y Harper ( 1 9 8 1 ) , t h e y are n o t used in practical feed f o r m u l a t i o n , n o r d o t h e y indicate w h a t
'To whom correspondence should be addressed.
c o m p l e m e n t a r y value t w o or m o r e proteins m a y have. T h e m e t h o d s t o be c o m p a r e d w e r e : Saturation Kinetics Model. (Morgan et al., 1 9 7 5 , as proposed b y Phillips, 1 9 8 1 , 1982). Phillips ( 1 9 8 1 ) suggested t h a t t h e r e was s o m e value in k n o w i n g t h e response of test animals over a wide range of n u t r i e n t intakes. T h e response data could be fitted to a single fourp a r a m e t e r sigmoid curve, with each p a r a m e t e r of curves of different protein sources being c o m p a r e d to indicate p r o t e i n quality. Data were derived from t h e present studies so t h a t SK curves could be fitted t o t h e same d a t a as t h e o t h e r m e t h o d s of evaluation. Slope-Ratio Assay (SR). (Hegsted and Chang, 1 9 6 5 a , b ) . These a u t h o r s felt t h a t there was s o m e t h i n g inherently valid a b o u t comparing t h e slopes of response lines depicting t h e relationship b e t w e e n b o d y weight gain and nitrogen intake. Because t h e y were making linear a p p r o x i m a t i o n s of sigmoid functions, t h e i n h e r e n t value of t h e t e c h n i q u e is n o t obvious. Proteins were ranked t h e same b y t h e SR assay and t h e m o r e traditional protein efficiency ratio m e t h o d in Hegsted and Chang's studies.
2268
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
ABSTRACT Studies were conducted to compare several methods of determining the protein quality of poultry by-product meal (PBPM). A sample of PBPM was compared to a soybean meal (SBM) sample fortified with .5% of L-methionine. Each was incorporated into corn starch and poultry oil-based diets at 2.5% increments between 0 and 30% protein. Diets were fed for 7 days starting when the chicks were 8 days old. Slope-ratio and saturation kinetics models were fitted for the two protein sources. Protein efficiency ratio (PER), net protein ratio, and net protein utilization were calculated for each diet. The PER was the most discriminating method of estimating protein quality, especially at lower levels (2.5% protein: PBPM = .65, SBM + methionine (Met) = 4.23). At higher levels, no differences could be detected (30% protein: PBPM = 2.45, SBM + Met = 2.31). The 6% protein level was chosen for further studies because there were large differences in PER, and the chicks were in positive nitrogen balance. For eight samples of PBPM (from three different processing facilities), Kjeldahl N ranged from 8.64 to 10.22% gross energy (GE) from 4.49 to 5.36 kcal/g; Ca from 3.6 to 8.84%, and P from 2.04 to 2.55%. A Monday morning sample (fresh broiler waste) was compared to a Friday afternoon sample from the same processing facility (broiler and hatchery waste, and DAF sludge) in semipurified diets at 6% protein. Significant differences in gain were detected in two experiments, but PER in only one. For the five samples tested by all methods, similar rankings were found by PER, pepsin digestibility, total lysine and Carpenter's available lysine. (Key words: Slope-ratio assay, protein efficiency ratio, net protein ratio, net protein utilization, pepsin digestibility, lysine availability)
POULTRY BY-PRODUCT MEAL
Available Lysine (ah YS). This was according to the measurement by Carpenter (1960). The binding of free e-amino groups of lysine with Sanger's reagent (fluorodinitrobenzene) was measured to indicate the lysine that has not
been rendered unusable to the bird by processing or overheating. Pepsin Digestibility (PD). (Adapted by Johnstone and Coon 1979). This is an in vitro measure of the ability of a purified enzyme to hydrolyze the protein present. MATERIALS AND METHODS
General. Day-old male broiler chicks from a commercial hatchery were used in five experiments. Chicks were housed in temperature-controlled battery brooders with raised wire floors and constant illumination. Diets in mash form and water were supplied on an ad libitum basis. Mortality was recorded daily and feed consumption data adjusted to account for any deaths, on a chick-day basis. Dietary protein (N X 6.25) was determined by the method of the Association of Official Analytical Chemists (AOAC, 1970). Percent moisture was determined by drying in a forced draft oven at 105C for 16 hr. Poultry by-product meals were obtained from three processing facilities in the Southeastern United States. They have previously been described (Pesti et al.). Samples # 1 , #2, and #3 were random samples taken from Plant A (continuous flow). Sample #4 was obtained through regular commercial channels and was reported to be from Plant B (continuous flow). Samples #5, #6, and #7 were from Plant B and were obtained on Monday morning (broiler waste only; #5), Monday afternoon (broiler and hatchery waste; #6), and Friday afternoon (broiler and hatchery wastes and DAF sludge, #7). Sample #8 was obtained from a plant with batch processing of broiler waste only, no hatchery waste or DAF sludge. Chemical Analysis. Pepsin digestibility was determined by the method of the AOAC (1970) with the exception that .002% pepsin was used (Johnstone and Coon, 1979). Available lysine was determined by the method of Carpenter (1960). Amino acid analyses were conducted by another laboratory using a modification of the method of Moore et al. (1958). Descriptive statistics, analysis of variance, and orthogonal comparisons were calculated using the general linear model procedures described in the SAS User's Guide (SAS, 1982). Duncan's new multiple range test was used to separate means (SAS, 1982). Experiment 1. Four hundred-eightly male chicks were randomly distributed into 80 pens
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
Protein Efficiency Ratio (PER). A method of expressing numerically the growth-promoting value of proteins was proposed by Osborne et al. in 1919. Protein efficiency ratio was defined as "the gain in body weight per gram of protein consumed." It consisted of measuring the growth (gain in weight) of growing rats fed different protein levels, and the highest obtained value was taken as the PER. The accuracy of this method has been frequently criticized on the basis that: 1) the results are influenced by the levels of protein consumed; 2) no allowance is made for the quantity of protein used for maintenance; and 3) gain in body weight does not necessarily correspond to gain in body protein (Mitchell, 1924; Hegsted and Worcester, 1947; Bender, 1956; Bender and Doel, 1957; Summers and Fisher, 1961; Harper, 1981). Net Protein Ratio (NPR). A modification of the original PER was proposed by Bender and Doel (1957). They included in the experiment a group of animals consuming a nonprotein diet for a similar period of time. The calculations involved the difference in weight between the groups, instead of only body weight gain. Net protein ratio was believed to be more precise and reproducible and to overcome the basic weakness of PER (Bender and Doel, 1957; Jansen, 1978), as it measures the protein utilized for maintenance as well as growth. Bender and Doel (1957) stated that NPR values are not influenced by food intake. As with any growth method, it assumes that the gain in body weight is indicative of the new protein tissue, which is not always valid. The procedure was defined by its originators as gain in weight of the test group plus loss of weight of proteinfree group, divided by the protein consumed. Net Protein Utilization (NPU). A derivation from the Thomas-Mitchell procedure (Mitchell, 1923) for the biological value (BV) determination is the NPU method (Bender and Miller, 1953; Miller and Bender, 1955). Miller and Bender (1955) defined it as the difference of body nitrogen content between a group fed a protein diet and a group fed a protein-free diet, expressed as a percentage of the nitrogen intake of the test group. The BV can be estimated by dividing NPU by a digestibility coefficient (Miller and Bender, 1955; Allison, 1964).
2269
2270
ESCALONA ET AL.
nitrogen. Carcasses were then ground by blender homogenization and refrozen until nitrogen and water determinations were conducted. The SK model was fitted to the data for each protein source: bKi + r m a x i n Kl + In where r = body weight gain in grams, I = nitrogen intake (g), b = intercept, r m a x = asymptotic value of r when I -» °°, and n = apparent kinetic order of the response when I -> O, and Ki is the "nutrition constant." The SR model was fitted to the data from protein levels of 15% or less by the equation: r = b0 + M S B M + MPBPM
TABLE 1. Composition of basal, summit, and control diets used in Experiment 1 (%)
Ingredients G r o u n d yellow corn Corn starch Poultry b y - p r o d u c t meal Soybeal meal, dehulled DL-Methionine L-Methionine Poultry oil Limestone, ground Phosphate, defluorinated Salt, NaCl Potassium chloride Selenium p r e m i x 3 Vitamin p r e m i x 4 Mineral p r e m i x 5 Composition b y calculation ME kcal/g Protein, % Calcium, % Phosphate, %, n o n p h y t a t e Methionine
Protein free
PBPM 1
Soybean2 meal
Practical control 57.50
89.43
44.02 49.59
30.18 60.91
6.0
6.0
3.89
5.0 31.0 .15
.31 6.0 .44 1.78 .08
6.0 .65 1.75 .40
.38
.10
.25 .05
.25 .05
.25 .05
.05 .25 .05
3.81 0 1.28 .72
3.47 30.0 2.95 .84 .50
3.12 30.0 .99 .47 .74
3.12 23.0 1.09 .51 .54
1
Poultry by-product meal Sample 4.
2
Mixture of 99.5 parts of soybean meal plus .5 part of L-Methionine.
3
Provides: .1 mg selenium/kg diet as sodium selenite.
* Provides (per kg diet): Vitamin A, 5500 IU; vitamin D 3 1100 ICU; vitamin E, 11 IU; riboflavin, 4.4; Ca panthotenate, 12 mg; nicotinic acid, 44 mg; choline CI, 220 mg; vitamin B 1 2 , 6.6 Mg; vitamin B6 2.2 mg; menadione, 1.1 mg (as MSBC); folic acid, .55 mg; d-biotin, .11 mg; thiamine, 2.2 mg (as thiamine mononitrate); ethoxyquin, 125 mg. 'Provides (ppm diet): Mn, 60; Zn, 50; Fe, 30; Cu, 5; I, 1.05; Ca, 75 (minimum), 90 (maximum).
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
of six birds each. Three replicates (pens) were used for each of the test rations, which also included a nonprotein diet. Chicks were fed a practical starter ration (Table 1) for the first 7 days, then fasted overnight. Experimental diets were provided from the 8th to the 14th day. A protein-free basal diet was mixed with the summit diets (Table 1) containing a poultry by-product meal (Sample #4, 30% protein) or a mixture of soybean meal (SBM) plus methionine (Met) (30% protein). Each was incorporated in appropriate quantities into a series of 12 diets (Table 2) such that the nitrogen content varied from .4 to 4.8% (2.5 to 30% protein, N x 6.25). On the 14th day, feed was removed for a 6 hr period to allow the intestinal tract to empty. After this, two birds/pen were killed by cervical dislocation (without loss of blood) for carcass analysis. The carcasses were frozen, cut into small pieces, and cooled further in liquid
P O U L T R Y BY-PRODUCT M E A L
where b 0 = intercept, I = nitrogen intake, and bi and b 2 are the coefficients for SBM and PBPM, respectively. The three other models for assessing protein nutritional quality were computed by the following formulas:
a pretreatment period of 8 days was used. The experimental diets were fed on Days 9 to 14. All diets were prepared to contain a constant protein level of 6%, based on determined protein content (Table 3). Five pens of 10 male chicks each were fed each diet in Experiment 2, eight pens of 10 male chicks each in Experiment 3 (Table 4).
Gain in body weight PER;
2271
Protein consumed RESULTS
Gain test group + loss protein-free group NPR
Body nitrogen test group — body nitrogen protein-free group Nitrogen consumed
Experiments 2 and 3. The procedures followed in these two trials were similar to that in Experiment 1, except that in Experiment 3,
T A B L E 2. Summary of intake-response data of chicks fed diets containing various levels of poultry meal or 199 parts soybean meal plus 1 part L-methionine or a protein-free diet (Experiment Dietary protein
Dietary nitrogen
Feed consumed
Nitrogen consumed
(n>)
by-product 1)
Nitrogen retained 1
Body 'weight gain
-.37 ±
.04
-11 ±
-.17 .17 .47 .79 1.24 1.92 2.27 2.61 2.88 3.20 3.58 3.63
.02 .06 .07 .02 .11 .14 .26 .05 .14 .16 .13 .13
-2 13 26 37 57 83 96 107 111 123 133 131
(g) Protein-free diet
.0
.0
105 ±
5
0
±0
2
Poultry b y - p r o d u c t meal-based diets 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0
.4 .8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8
2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0
.4 .8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4 4.8
107 ± 124 ± 130± 135 ± 163 ± 171 ± 187 ± 186 ± 182 ± 199 ± 175 + 179 ±
4 2 5 3 6 10 11 7 4 35 2 13
.43 .99 1.56 2.17 3.26 4.09 5.24 5.97 5.09 7.94 7.69 8.57
± .02 ± .01 ± .05 + .11 ± .11 + .24 ± .30 ± .21 ± .25 ± :1.38 ± .11 ± .64
+ ± + ± + + ± ± ± ± + ±
± 1 ± 2 ± 3 ± 3 + 5 ± 6 ± 11 ± 2 ± 2 ± 6 ± 5 ± 5
Diets based on 199 parts soybean meal plus 1 part L-methionine 118± 7 158 ± 8 203 ± 16 201 ± 2 237 ± 7 245 ± 6 245 + 5 228± 5 244+ 8 225 ± 9 226 + 10 2 3 4 ± 19
.47 1.26 2.44 3.20 4.73 5.89 6.86 7.31 8.80 9.00 9.94 11.23
+ ± ± ± ± ± ± ± ± ± ± ±
.03 .06 .19 .04 .15 .15 .15 .17 .29 .35 .44 .91
.18 .85 1.83 2.32 3.38 3.66 4.21 4.31 4.81 4.42 4.35 4.50
± .02 ± .11 ± .25 ± 6.96 ± .16 ± .37 ± .09 ± .03 ± .22 ± .25 ± .36 ± .35
12± 41 ± 82 ± 102 + 145 + 155 ± 175 + 175 ± 191 ± 171 ± 163 ± 162 ±
1 5 11 9 7 15 4 1 9 10 14 13
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
NPU =
Experiment 1. At lower levels of dietary protein, carcass nitrogen deposition appeared to be fairly constant. After a 10% level of dietary protein was reached, the rate of carcass nitrogen increased (Fig. 1). The quadratic regression equation describing the effect of dietary protein content ( X t ) on the percent carcass nitrogen (Y) was determined:
Protein consumed
2272
ESCALONA ET AL. TABLE 3. Composition of experimental diets used in Experiment 2 (%)
Poultry by-product meal 1 Ingredients
1
Soybean meal + methionine
9.64
12.37
12.31
80.67 6.00 .073
76.93 6.00 1.78
76.93 6.00 1.78
1.53
2.46
2.46
.25 .05 .31
.25 .05
.25 .05
.12
.12
11.11 .11
80.40 6.00 1.59 1.7 .25 .05 .32
4.1 6 1.5
4.05 6 1.5
3.9 6 1.5
.52
.52
3.9 6 1.5
.52
.52
.089
.199
Described by Pesti et al. (1986).
2
Provides (per kg diet): Vitamin A, 5500 1U; vitamin D 3 , 1100 1CU; vitamin E, 11 IU; riboflavin, 4.4; Ca panthothenate, 12 mg; nicotinic acid, 44 mg; choline Cl, 220 mg; vitamin B 1 2 , 6.6 Mg; vitamin B 6 , 2.2 mg; menadione, 1.1 mg (as MSBC); folic acid, .55 mg; d-biotin, .11 mg; thiamine, 2.2 mg (as thiamine mononitrate); ethoxyquin, 125 mg. 3
Provides (ppm diet): Mn, 60; Zn, 50; Fe, 30; Cu, 5; I, 1.05; Ca, 75 (minimum), 90 (maximum).
TABLE 4. Composition of experimental diets used in Experiment 3 (%) Poultry by-product meal 1 Ingredients Poultry by-product meal Corn starch Poultry oil Limestone, ground Phosphate, defluorinated Vitamin premix 2 Mineral premix 3 Potassium chloride Composition by calculation ME, kcal/g Protein, % Calcium, % Phosphate, %, nonphytate 1
1
4
5
7
8
9.5
9.9
82.1
81.0
9.64 80.55
11.11 80.78
9.65 80.65
6.0
6.0
6.0
1.18 1.44
1.53 1.78
6.0 .05
6.0
1.52 1.56
1.56
.25 .05 .2
.25 .05 .2
.25 .05 .2
1.5 .25 .05 .2
4.1 6 1.5 .52
4.1 6 1.5 .52
4.1 6 1.5 .52
.25 .05 .2 4.1 6 1.5 .52
1.39
4.1 6 1.5 .52
Described by Pesti et al. (1986).
2
Provides (per kg diet): Vitamin A, 5500 IU; vitamin D 3 1100 ICU; vitamin E, 11 IU; riboflavin, 4.4; Ca panthothenate, 12 mg; nicotinic acid, 44 mg; choline Cl, 220 mg; vitamin B 1 2 , 6.6 Mg; vitamin B 6 2.2 mg; menadione, 1.1 mg (as MSBC); folic acid, .55 mg; d-biotin, .11 mg; thiamine, 2.2 mg (as thiamine mononitrate); ethoxyquin, 125 mg. 3
Provides (ppm diet): Mn, 60; Zn, 50; Fe, 30; Cu, 5; I, 1.05; Ca, 75 (minimum), 90 (maximum).
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
Poultry by-product meal Soybean meal DL-Methionine Corn starch Poultry oil Limestone, ground Phosphate, defluorinated Vitamin premix 2 Mineral premix 3 Potassium chloride Salt (NaCl) Composition by calculation ME energy, kcal/g Protein, % Calcium, % Phosphate, % nonphytate Methionine
Soybean meal
7
5
POULTRY BY-PRODUCT MEAL
3.0
•
2.9 2.8 O O
• •
2.7 i
2.2
•
en <
<
2.4
•
•
•
2.1 2.0 0
i
i
i
5
10
15
i
20
i
i
25
30
% DIETARY PROTEIN FIG. 1. Relationship between the carcass nitrogen of chicks and dietary protein level (Experiment 1).
Y = 2.43 3 - .0072XJ + .00052Xi (52.79)* (-1.03X2.34)* The values in parenthesis are Student t-values and the asterisk indicates significance at a 5% level. The coefficient of determination (R 2 ) was .34. Mean values were used in calculating NPU, NPR, SK, and SR models. Summaries of intake-response data and PER, NPR, and NPU values as determined for both protein sources are shown in Tables 2 and 5, respectively. Weight gains of chicks increased with an increase in protein level for each protein source (Fig. 2). Gains were consistently higher for SBM + Met. Maximum gain values were reached at the 27.5% level for PBPM and the 22.5% level for SBM + Met, but at higher protein levels the gains decreased (Fig. 2). Feed consumption increased with an increase in protein level for each protein, with greater intakes for SBM + Met (Table 2). Maximum intakes were seen at approximately 17.5%
protein for both protein sources. Nitrogen intake and nitrogen retention (Table 2) increased linearly for both proteins, but greater values were observed for SBM + Met. For the saturation kinetics model (Table 6), there were no obvious differences in intercepts (b), the kinetic order of the response as intake approaches zero (n), or the maximum response ( r m a x ) ; there was considerable overlapping of the confidence limits. There was, however, a considerable difference in the nutrition constant (Kj) with the better quality protein having the lower Kj value, as would be expected from the formula. In the slope-ratio analysis, SBM + Met was clearly the better quality protein with a slope ratio of 1.95. The three conventional indexes (Table 5; Figs. 3 to 5) indicated that SBM + Met was a better protein source at the lower levels. The PER showed a greater difference between protein sources than did either of the other measures. Variable PER values were observed; they increased up to a maximum (15% PBPM and 7.5% SBM + Met) and then began to decline (Fig. 3). On the other hand, NPR and NPU values appeared to be relatively constant, especially for PBPM. Similarly, significant curvilinear (quadratic) relationships (P<.01) between PER or NPU and protein level were found; this relationship was not significant (P> .29) for NPR. Significant correlations between PER and NPU (r = .79) and PER and NPR (r = .72) were found (Table 7). The correlation between NPR and NPU, although significant, was lower (r = .63). The NPU increased from the 2.5% level up to approximately 10% for both proteins, being more marked for SBM + Met (Fig. 4). While NPU values for PBPM remained relatively constant above that level, SBM + Met showed a progressive decline in NPU as the dietary protein level increased. The PER and NPR gave different responses. Protein efficiency ratio rises to a maximum for both proteins. Beyond this maximum, PER came to plateau and then declined slowly at higher intakes. The NPR, although relatively constant for PBPM, declined for SBM + Met as dietary protein increased (Fig. 5). Experiments 2 and 3. As expected, PER values were higher for the reference protein (soybean meal) than PBPM; and a significant difference was observed due to L-methionine supplementation (Table 7). No significant differences in PER were observed between
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
2.3'
• • • •• t • • • • • • • • • • • • • • • • • • • !
2.5
o a:
x
'v'
2.6
2273
2
1
± .62 ± .26 ± .39 ± .07 ± .31 ± .01 ± .08 ± .05 ± .12 + .11 ± .02
Mixture of 199 parts of SBM plus 1 part Met.
5.19 5.35 5.09 4.91 4.19 4.08 3.83 3.47 3.03 2.61 2.31
± .38 ± .26 ± .15 ± .12 ± .14 ± .29 ± .07 + .04 ± .32 ± .06 ± .11
2.09 2.62 2.74 2.78 3.26 2.94 2.88 2.72 2.59 2.77 2.45
(X± SEM) 4.28 ± .46
SBM + Met
-.65 ±.39
PBPM
Ratio of (PBPM/SBM + Met) X 100.
10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0
2.5 5.0 7.5
(%)
Protein
1
Protein efficiency ratio
-15 40 49 54 57 78 72 75 78 86 106 106
(%)
Index
2
3.30 3.80 3.71 3.51 3.30 3.67 3.26 3.16 2.98 2.81 2.99 2.65
( + .25 + .39 + .25 ± .14 ± .11 ± .14 ± .29 ± .08 ± .04 ± .35 ± .06 ± .13
PBPM
42 58 61 63 62 82 75 78 81 87 108 108
(%) ± .67 ± .64 ± .21 ± .39 + .06 + .31 ± .02 ± .08 ± .04 ± .12 ± .10 ± .13
A)
7.89 6.54 6.05 5.62 5.27 4.48 4.33 4.06 3.66 3.22 2.78 2.46
Index
SBM + Met
Net protein ratio
TABLE 5. Estimates of the protein quality of poultry by-product meal (PBPM) and a so plus methionine (Met) mixture by various methods (Experiment 1)
ed from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
POULTRY BY-PRODUCT MEAL
210 r
0 O
0 /O
V
-= 150 -
<
0
\ °
180 -
0
o 0
X 120-
o
°4
LJ 90
f
7
Q O 60 CD
/• *
i
A
?
-
J
Protein Source » PBPM SBM
i° r y
\
0 " 5
i
i
i
i
10
15
20
25
i
30
% DIETARY PROTEIN FIG. 2. Relationship between body weight gain and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
chicks fed PBPM Samples #5 or #7. However, responses from these samples were significantly different than those from the SBM and SBM + Met reference proteins.
Chicks fed the soybean meal + Met diet had significantly better weight gains than the others. Chicks fed the PBPM diets had lower gains, especially sample #7. Feed intake was consistently higher with the SBM + Met diet. However, chicks fed PBPM #5 and the SBMbased diet were not significantly different in feed intake (P>.05). Orthogonal comparisons (Table 8) indicated a significant difference (P = .045) in gain, but not in feed efficiency or PER (P = .222), between samples #5 and #7. [P values for feed efficiency and PER are identical since efficiency (gain/consumption) and PER (gain/consumption X .06) are linearly related]. Chemical Assays. Protein efficiency ratio (Table 7), aLYS (Table 8), total lysine (LYS, Table 9), and pepsin digestibility all ranked the samples in similar orders: 5, 8, 1, 4, 7; 1, 8, 5, 4, 7; 1, 8, 5, 4, 7, and 1, 8, 7, 5, 4, respectively. The amino acid pattern of these samples was fairly uniform (Table 9). The patterns were also similar to that in the NRC (1984) nutrient composition tables. Essential amino acids in the PBPM were slightly higher than NRC values with the exceptions of histidine (lower) and tyrosine (much higher). DISCUSSION In Experiment 1, there was a high correlation between PER, NPU, and NPR, which is in agreement with the observations of Block
TABLE 6. Coefficients for the saturation kinetics model and slope-ratio assay fitted to the response data from Table 2 Soybean meal
Poultry by-product meal Confide nee limits
Confidence limits
Saturation kinetics Analysis Intercept Maximum response Nutrition constant Kinetic order Intake At maintenance At inflection point Response At inflection point Slope ratio analysis Intercept 1 Slopes 1
Mean
Lower
Upper
Mean
Lower
Upper
-4.07 202 2.84 1.70
-10.03 190 2.58 1.44
1.88 213 3.1 1.95
-8.36 180 4.17 1.53
-14.16 151 3.26 1.22
-2.56 209 5.08 1.84
-10.7 5.72
-19.3 6.62
.285 1.27
.559 1.49
38.3 -15.5 12.01
24.2 -10.7 11.56
A common intercept was fitted for the two lines.
-19.3 12.46
-15.5 6.17
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
*
>-
30
/ ^
2275
2276
ESCALONA ET AL.
FIG. 3. Relationship between protein efficiency ratio and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
Protein Source PBPM SBM
\
Protein Source •—»PBPM
o
o
0^
o SBM
\
<
or
-°\\ o\
° o
o
>v N
l±J
IO
o o
or OL
t • - •
0
5
10
15
20
25
30
% DIETARY PROTEIN FIG 4. Relationship between net protein utilization and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
0
X
V 8
1k
i 15
i 20
• 1
1
5
10
T 25
i 30
% DIETARY PROTEIN FIG. 5. Relationship between net protein ratio and dietary protein (Experiment 1). PBPM = Poultry by-product meal, SBM = Soybean meal.
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
% DIETARY PROTEIN
and Mitchell (1946), Bender (1956), and Henry (1965). The good curvilinear relationship between PER and protein level observed in this study confirms that the finding of Hegsted and Worcester (1947) and Sammonds and Hegsted (1977), with rats, is similar for chicks. Osborne and Mendel (1916), Osborne et al. (1919), and Mitchell (1924) had also reported the influence of protein intake on protein efficiency, although their data covered smaller ranges. Bender (1956) showed the PER variability with feed consumption and the relative constancy of NPU and NPR. Similar observations were made in this experiment. Although subject to the criticisms described earlier, the PER index showed a greater spread than the other techniques. This allowed a better discrimination of protein quality between samples, especially at lower levels. At higher levels, differences could not be detected. Woodham (1967) had reported greater differences at low dietary protein for PER values. Scott et al. (1982) stated the greater sensitivity of chickens to differences in protein quality at
2277
POULTRY BY-PRODUCT MEAL TABLE 7. Protein efficiency ratios (PER), body weight gains, and feed consumption of broilers fed protein sources as 6% protein in semipurified diets (Experiments 2 and 3) PER
Consumption
Gain
Ranking
Mean + SE
source
UP Experiment 2 SBM 1 SBM + Met PBPM # 5 PBPM # 7
4.6 5.3 3.6 3.2
± ± ± ±
.3b .3a .lc .2C
4 8
+
3b,2
68±4a 37 ± 2 C 28 ± l d
176 + 212 ± 168+ 148 ±
4b'5 la 5b 4C
165 154 161 157 164
5a 3a 3a 4a 5a
Experiment 3 #1 #4 #5 #7 #8
3.2 3.1 3.6 2.9 3.5
± .2 ± ,2ab ± .2a ± .lb ±.la
3rd 4th 1st 5th 2nd
31 29 35 27 35
± ± ± ± ±
2ab b 2
3a lb la
± ± ± ± ±
O r t h o g o n a l c o m p a r i s o n s (significan<:e probabilities) Experiment 2
PBPM 5 vs. 7
a,b,c
Gain
PER
.045
.222
Means with the same letter are not significantly different (P>.05; Duncan's new multiple range test).
SBM = Soybean meal, PBPM = Poultry by-product meal. Mean + SE.
lower levels. These authors recommend using 10% protein in the diet, but our results (Table 5 and Fig. 3) indicate that differences may be obscured at this level. Protein efficiency ratio is well correlated to the other methods, NPU, and NPR. Protein efficiency ratio is simpler to determine. No time is consumed in preparing carcasses for analysis (as in NPU), and the inclusion of a group fed a protein-free diet is not neces-
sary (as for NPU and NPR). The wide use and better protein quality discrimination of PER in this study indicate that it offers a simpler and more reproducible procedure for evaluating protein quality. All these considerations led us to choose the PER method as the most useful one to be used in further experiments. Any overriding advantage of either the SK or SR assays is not apparent from our studies. Many more data points are needed for these
TABLE 8. Chemical estimations of the quality of protein from several poultry by-product meals. Sample no.
Available lysine
Plant (g/100 g) A A A B? B B B C
3.03 2.85 3.12 2.19 2.50 2.52 2.05 2.52
± ± ± ± ± ± ± ±
.15' . 0 8 'ab .11' . 0 2 'cd .12,'be .14 1be .15' .10 1be
Ranking
Pepsin digestibility
2nd 3rd 1st 7th 6th 4th 8th 5th
87.0 89.9 86.4 77.3 78.0 75.4 79.0 85.8
( g / 1 0 0 g protein 4.82 ± . 2 4 a b 4.46±.13abc 4.92 ± . 1 7 a 3.63 ± . 0 3 d 4.00 ± 1 8 ^ d 4.17 ± . 2 3 b ™ 3.80 ± . 1 8 c d 4.05 ± . 1 5 d
Ranking
(g/100 g) ± ± ± ± ± ± ± ±
.5a .4a ^ab .0e
•°r • 2 di
.5 .lb
Means ± SE within columns with different superscripts are significantly different (P<.05).
1st 2nd 3rd 7th 6th 8th 5th 4th
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
PBPM PBPM PBPM PBPM PBPM
ab
5 S 5 S o nJ, 2 . x a g- o SJ S" 5 & a « ^ 2 .x 3 p - n > ° x " S 2 . 3 = • ^ - I !•%
rt
5 =• 3 5 S 3 " =
to w w 4i M sj I-* ujojN))---vjNJO-f».4i.4^ WOvUiwixw^
in
to to -t* i-» to
NJ
NJ
to
-F* i- 1 t o
>-*
i-* ^
t o t o •£• i- 1 t o i-*
W ^ t O M ^ h J O v ^ - ^ - t '
W
W
W
H
^
W
O
V
-
^
^
^
w i w N j H v ) w y \ y j ^ w
wsawONoowMvoNNOvobHobbe^i^bo
H^SO^OOWHOOWMWHOO^Ul^OHHOJVOW
t-*
tO
O
>-»
h - t O U J H-i W
W W | g M M t O U l w 4 » W
W
W
M
M
a
t
O
U
l
W
^
W
w ^l u i ^ b b o b v w b o ^ w ' ^ v i ^ b ^ u ' u i y i i - ' ys i ! -
M
W C O ' s l 0 0 H U i ^ O v ^ t O ^ l H s l \ O 0 0 t O U l W O
t o t o 4* >-» t o
t o tO 4^ <-- t o
•-*
i- 1
J
w W N ) H < I W O \ ^ ^ ^
W
W
W
M
^
w
a
^
^
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
to
to
y j i > y i t o b Q w \ o w b \ I - ' t j j ^ \ o b to N© In bo U» bo to,
ON
O
w t o t o 4^ t o t o •-* >-* ui4itoi-''<]UJON4»-t^^w O J O C O Wt W ' \ I O V4 l w b N J M N ) t O i - 1 b \ W U t s W vv l \ 0 C O O s i - a ^ W ' - N M ^ U i O ^ O v 4 ' U i O ' - ' J O W
N) tO ^
t- 1 tO
tO
•£»
© O
to
OvitoyjwNiOMODyii-'O-f'CoovOvooMOvOow
H-
^
to o j ^ b \ S i l n s ) b o u i b o h J N i N j ^ o t o l i ^ ' v 4 4*>--'bNb O 1+
- J ^ i - ' t O t O P M \ O O \ u i t o 0 0 i - » W ' s j t o \ O O i - ' 0 \ a v O 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ 1+ ,
i
t o O i w - ^ w ^ o ^ w O - f ' M O ^ V i w O O ' O ' - ' O ^ O
©
b b b b H b b b
to
00 ©
to to ON U\ © O w O
H b b ' * ' b b b w H
to w Lfi h-» O N © -£• O VO O
to O J to >— ^J -v] O Wl
' H H i - b
(M Ul -^ ON ^O © 00 O ©
O © © O
I V X3 VNCHVDS3
POULTRY BY-PRODUCT MEAL
Sample #5 as a better protein source when five samples of PBPM were compared. Sample #7 showed the lowest PER values, while the other three samples were intermediate between Samples #5 and #7. The reproducibility of PER was demonstrated when the mean PER value for Sample #5 was exactly the same in Experiments 2 and 3. Chapman et al. (1959) and Henry (1965) have also reported good reproducibility for the PER of casein with rats. Although in Experiment 2 the higher PER values were obtained at higher feed intakes, in Experiment 3, no relation between PER values and feed intake was observed, because consumption was not significantly different between the experimental diets. Mitchell (1924) observed the highest protein efficiency in rats with larger intakes, similar to the results of Experiment 2. It is not clear why there were differences in the pattern of consumption between Experiments 2 and 3. Perhaps a change in the chicks or environment caused the more uniform consumption in Experiment 3. Ingredient sources were identical. Random effects may have been the cause. The rankings of these samples were very similar for PER, LYS, aLYS, and pepsin digestibility if Sample #5 is omitted from the analyses. Sample 5 is an enigma, being ranked first by PER, third by LYS and aLYS, and fourth by pepsin digestibility, among samples ranked by all four methods. The amount of the limiting amino acid will normally determine the protein quality of a feedstuff. It may be that lysine is not the first limiting amino acid in these samples, or that some other amino acid may interefere with its utilization. An advantage of one of these methods over the others for routine analysis is not apparent from these studies. The PER determinations are time and resource intensive, and differences in PER do not necessarily indicate differences in performance when the feedstuffs are incorporated into practical diets at low levels. Of the chemical methods, Carpenter's available lysine procedure requires the least-specialized equipment (a fume hood, reflux apparatus, and spectrophotometer), and should indicate if nutritional value has been lowered by overheating. Pepsin digestibility is simpler if the specialized apparatus is available. It indicates if poor quality proteins were used in making the product. Total amino acid analysis is the most difficult, requiring specialized equipment, but it yields the most usable data for feed formulation.
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
assays. A large advantage would be necessary to justify them. As Phillips (1982) has noted, there are theoretical reasons not to compare r max o r b values, yet extra data has to be generated to fit them in order to estimate the other parameters. The effect of better quality proteins having lower Kj values also complicates interpreting the results. The PER and SR assays are very similar, with both relating body weight gains to nitrogen (or protein) intake. The only real difference is that the SR values are pooled over a range of intakes, and PER is computed for one intake level. The arbitrary value generated by the SR assay was very similar to an arbitrary value made by a ratio of the PER values at 7.5% protein (1.94 vs. 2.04) in this experiment. The PER values have just as much inherent value, in our opinion, and are derived with considerably less work. The almost constant nitrogen deposition (on a percent of carcass basis) at lower protein levels (Fig. 1) may be ascribed to the predominant function of dietary nitrogen in maintaining the nitrogenous integrity of the tissues. Once nitrogen equilibrium was reached, nitrogen was used in increasing proportions for growth. The significant quadratic effect of protein level determines its contribution to the model, indicating that this curvilinear effect of protein level should be considered in predicting nitrogen deposition in growing chicks. The choice of a 6% protein level used in the semipurified diets fed in Experiments 2 and 3 was made somewhat arbitrarily. The better discrimination between different proteins by PER at the lower levels was the basis for this decision. Another important consideration in choosing this level was the fact that the chicks were in positive nitrogen balance, an important consideration to some researchers. The significant difference in PER and gain between SBM- and PBPM-fed chicks indicated that SBM is a superior protein source (Table 7). As expected, better results were obtained when SBM was supplemented with L-methionine. This established that the assay was sensitive enough to detect this known difference in protein quality. Although PER values of chicks fed the two PBPM were not significantly different, gains and feed intakes were. This suggests a better balance in amino acid composition or a greater availability in Sample #5 than in #7, or perhaps a toxic factor in the DAF sludge in Sample #7. Again, PER ranked
2279
2280
ESCALONA ET AL.
However, availability and digestibility are n o t considered. Each assay m a y be a p p r o p r i a t e u n d e r specific circumstances. ACKNOWLEDGMENTS
REFERENCES Allison, J. B., 1964. The nutritive value of dietary protein. In Mammalian Protein Metabolism II: 41. H. N. Monro, and J. B. Allison, ed. Academic Press, New York, NY and London, England. Association of Official Agricultural Chemists, 1970. Official Method of Analysis, 11th ed., Washington, DC. Bender, A. E., 1956. Relation between protein efficiency and net protein utilization. Br. J. Nutr. 10:135-143. Bender, A. E., and B. H. Doel, 1957. Biological evaluation of protein: a new aspect. Br. J. Nutr. 11:140-147. Bender, A. E., and D. S. Miller, 1953. Constancy of the N / H 2 0 ratio of the rat and its use in the determination of the net protein value. Proc. Biochem. Soc. Biochem. J. 53:vii-viii. (Abstr) Block, R. J., and H. H. Mitchell, 1946. The correlation of amino acid composition of proteins with their nutritive value. Nutr. Abstr. Rev. 16:249-278. Carpenter, K. J., 1960. The estimation of the available lysine in animal protein foods. Biochem. J. 77:604-610. Chapman, D. G., R. Castillo, and J. A. Campbell, 1959. Evaluation of protein in foods. 1. A method for the determination of protein efficiency ratios. Can. J. Biochem. Physiol. 37: 679-68'6. Harper, A. E., 1981. McCollum and directions in the evaluation of protein quality. J. Agric. Food Chem. 29:429-435. Hegsted, D. M., and Y. Chang, 1965a. Protein utilization in growing rats. I. Relative growth index as a bioassay procedure. J. Nutr. 85:159-168. Hegsted, D. M., and Y. Chang, 1965b. Protein utilization in growing rats at different levels of intake. J. Nutr. 87:19-25. Hegsted, D. M., and J. Worcester, 1947. A study of the relation between protein efficiency and gain in weight on diets of constant protein content. J. Nutr. 3 3:685-693. Henry, K. M., 1965. A comparison of biological methods with rats for determining the nutritive
Downloaded from http://ps.oxfordjournals.org/ at New York University on September 3, 2014
S u p p o r t e d b y State and Hatch funds allocated t o t h e Georgia Agricultural E x p e r i m e n t Stations of t h e University of Georgia. T h e a u t h o r s gratefully acknowledge L. O. Faust and M. A. S t a t o n , for their able technical assistance, and F. J. Ivey and Monsanto C o m p a n y , Nutrition Chemicals Division, Chesterfield, MO, for providing t h e a m i n o acid analyses.
value of proteins. Br. J. Nutr. 19:125-135. Jansen, G. R., 1978. Biological evaluation of protein quality. Food Technol. 32:52-56. Johnstone, J., and C. N. Coon, 1979. The use of varying levels of pepsin for pepsin digestion studies with animal proteins. Poultry Sci. 58: 1271-1273. Miller, D. S., and A. E. Bender, 1955. The determination of the net protein utilization of proteins by a shortened method. Br. J. Nutr. 9:382-388. Mitchell, H. H., 1923. A method of determining the biological value of protein. J. Biol. Chem. 58: 873-903. Mitchell, H. H., 1924. The nutritive value of proteins. Physiol. Rev. 4:424-478. Moore, S., D. H. Spackman, and W. H. Stein, 1958. Chromatography of amino acids on sulfonated polystyrene resins. Anal. Chem. 30:1185-1190. Morgan, P. H., L. P. Mercer, and N. W. Flodin, 1975. General model for nutritional responses of higher organisms. Proc. Natl. Acad. Sci. USA. 72:4327. National Research Council, 1984. Nutrient Requirements of Poultry. Nutrient Requirements of Domestic Animals. Natl. Acad. Sci., Washington, DC. Osborne, T. B., and L. B. Mendel, 1916. A quantitative comparison of casein, lactalbumin, and edestin for growth or maintenance. J. Biol. Chem. 26:1-23. Osborne, T. B., L. B. Mendel, and E. L. Ferry, 1919. A method of expressing numerically the growth promoting value of protein. J. Biol. Chem. 37:223-229. Pesti, G. M., L. O. Faust, H. L. Fuller, N. M. Dale, and F. H. Benoff, 1986. Nutritive value of poultry by-product meal. 1. Metabolizable energy. Poultry Sci. 65: (in press). Phillips, R. D., 1981. Linear and nonlinear models for measuring protein nutritional quality. J. Nutr. 111:1058-1066. Phillips, R. D., 1982. Modification of the saturation kinetics model to produce a more versatile protein quality assay. J. Nutr. 112:468-473. Sammonds, K. W., and D. M. Hegsted, 1977. Animal bioassay: A critical evaluation with special reference to assessing nutritive value for die human. Pages 68-79 in Evaluation of Protein for Humans. C. E. Bodwell, ed. The AVI Pub. Co., Westport, CT. SAS Institute, Inc., 1982. SAS User's Guide: Basics. SAS Inst. Cary, NC. Scott, M. L., M. C. Nesheim, and R. J. Young, 1982. Nutrition of the Chicken. M. L. Scott Associates, Publishers, W. F. Humphrey Press, Inc., Geneva, NY. Summers, J. D., and H. Fisher, 1961. Net protein values for the growing chickens as determined by carcass analysis: Exploration of the method. J. Nutr. 75:435-442. Woodham, A. A., 1967. A chick growth test for the evaluation of protein quality in cereal-based diets. 1. Development of the method. Br. Poult. Sci. 9:53-63.