Growth, nutrient utilization, and body composition of dairy calves fed milk replacers containing different amounts of protein

Abstract

Male Holstein calves <1 wk of age were allowed a 2-wk adaptation period after purchase, and then were blocked by BW and assigned randomly within block to either a baseline slaughter group or one of four experimental groups (n = 8 to 9 per group). Treatments were isocaloric milk replacers (12.5% solids) fed at 12% of BW that contained 16.1, 18.5, 22.9, or 25.8% CP (DM basis) from whey protein sources. After a 6-wk feeding period, all calves were slaughtered and the weights and chemical composition of the viscera-free carcasses (VFC; including head, hide, feet, and tail) were determined. Gain of BW (0.38, 0.45, 0.56, and 0.62 kg/d) and gain:feed ratio (0.51, 0.59, 0.71, and 0.78) increased linearly (P < 0.001) as dietary CP increased; rate of change in body length, wither height, and heart girth also increased linearly (P ≤ 0.05). Balance measurements conducted during wk 3 and 4 of the experimental period showed that both absorbed N (16.9, 20.0, 25.8, and 30.6 g/d) and retained N (7.6, 9.0, 13.2, and 15.6 g/d) increased linearly (P < 0.001) as dietary CP increased. Retained N as a percentage of absorbed N increased linearly (P < 0.01) as dietary CP increased (44.3, 44.7, 50.7, and 50.9%), whereas biological value was unaffected (71.1, 68.7, 69.5, and 67.3%; P = 0.26). Digestible energy and ME represented 94.5 and 89.7% of intake energy, respectively, and were not affected by dietary CP content. Plasma urea N concentration increased linearly (2.9, 3.3, 4.6, and 6.0 mg/dL) as dietary CP increased. Contents of water (68.2, 69.1, 70.2, and 70.5%; P < 0.001) and protein (19.6, 20.0, 20.0, and 20.2%; P < 0.10) in VFC increased linearly, whereas contents of fat (7.2, 6.2, 5.5, and 5.2%; P < 0.001) and ash (5.1, 5.2, 4.8, and 4.7%; P < 0.02) decreased linearly as dietary CP increased. Trends in visceral tissue composition were similar to those for VFC. The content of water in VFC tissue gain increased, whereas contents of fat and energy decreased, as dietary CP increased. Final VFC energy and gain of energy in VFC were not affected by dietary CP. At similar initial ME intakes, increasing dietary CP (i.e., increasing protein:energy) linearly increased ADG, gain:feed, N retention, and deposition of lean tissue in VFC, demonstrating that diet composition can markedly affect components of body growth in preruminant dairy calves.

Introduction

Accurate definition of protein requirements for dairy calves is important to ensure an appropriate supply of AA for rapid structural growth and lean tissue deposition, while minimizing cost and excess N excretion when protein is overfed. Several authors have noted problems with the ability of previous NRC (1989) standards to predict growth of dairy heifers (Waldo et al., 1997; Davis and Drackley, 1998; Diaz et al., 2001). In particular, protein requirements stated for milk-fed calves (NRC, 1989) did not agree with experimental evidence or with specifications derived from other nutrient-requirement systems (Davis and Drackley, 1998).

Except for the comprehensive study by Donnelly and Hutton (1976a,b), effects of dietary protein supply in preruminant dairy calves have largely been evaluated by changes in BW gain, which is an imprecise, inaccurate, and often biased predictor of lean body mass (Owens et al., 1993). Increases in BW gain do not necessarily reflect increases in lean tissue deposition because of potential differences in fat deposition and in weights of gut fill and gut tissue mass (Stobo et al., 1966). Increased protein intake in milk-based diets may increase lean tissue deposition and decrease fat deposition (Donnelly and Hutton, 1976b). Diaz et al. (2001) reported that calves fed amounts of protein similar to those fed by Donnelly and Hutton (1976b) gained more protein and less fat than calves in the earlier study. Some evidence indicates that continued selection for milk production has resulted in dairy cattle that are leaner than those of 20 to 30 yr ago (Murphy et al., 1991; Waldo et al., 1997), perhaps due to larger mature size (Waldo et al., 1997) or increases in somatotropin secretion (Brown et al., 1989). Our objective, therefore, was to quantify the relationships between protein content of milk replacers and growth, body composition, and nutrient utilization in preruminant Holstein calves.

Materials and Methods

Management of Calves and Experimental Design

All procedures were conducted under protocols approved by the University of Illinois Laboratory Animal Care Advisory Committee. The experiment was designed to include eight calves in an initial (baseline) slaughter group and eight calves on each of four experimental treatments (i.e., 40 calves total) distributed over two replicates of 20 calves. In anticipation of a higher than normal death loss for shipped-in calves, 50 male Holstein calves, < 1 wk old, were purchased from a sale barn in two separate groups (i.e., replicates). The second group of 26 calves was purchased and used after completion of the first group of 24 calves. The calves were on experiment for an 8-wk period, which consisted of a 2-wk standardization period and a 6-wk experimental period. The experiment was conducted from July through November of 1997.

Calves were fed whole milk at a rate of 8% of BW divided into two equal feedings at 0800 and 1700 for 2 d after arrival. Calves were supplemented with an electrolyte solution (Advance Arrest; Milk Specialties Co., Dundee, IL) twice during d 1 at 1200 and 2200. Calves were fed with nipple buckets. The oral electrolyte solution was offered via nipple bucket; if calves refused to drink, the solution was intubated with an esophageal feeder. Calves were vaccinated upon arrival with bovine rotavirus-coronavirus vaccine, bovine rhinotracheitis parainfluenza vaccine, Escherichia coli antiserum, and Actinomyces pyogenes-Escherichia coli-Pasteurella multocida-Salmonella typhimurium antiserum (Bo-Bac-2X; Boehringer Ingelheim Animal Health, Inc., St. Joseph, MO) according to standard operating procedures at the University of Illinois Dairy Research Unit.

From d 3 to 14, the first group of calves was fed a standard milk replacer (Milk Specialties Co., Dundee, IL) that contained 20% protein and 20% fat. The milk replacer was reconstituted to 12.5% solids and fed at 10% of BW daily in two equal feedings at 0800 and 1700. The second group of calves began on this same schedule, but because of excessive diarrhea, on d 8, calves were switched to whole milk (10% of BW divided into two equal feedings) for the rest of the standardization period. Diarrhea seemed to lessen after this switch was made; reasons for improvement are unknown but could include different protein sources (casein vs. whey proteins), higher digestibility, or bioactive factors present in whole milk but not in milk replacer. Although the different standardization diets between the two groups of calves might be a confounding factor for experimental treatments, body composition of the calves slaughtered at the start of the experimental period was similar between the two groups of calves and the interaction of replicate and diet composition was significant for very few variables (see Results and Discussion). Calves had free access to fresh water at all times, but no access to hay or starter at any time during the experiment. All calves remained in individual calf hutches throughout the standardization period.

Seven calves (four in replicate 1, three in replicate 2) died during the adaptation period. On d 14, surviving calves were blocked by weight into four blocks of five calves each. Calves from each block were randomly assigned to either an initial slaughter (baseline) group (n = 8; 4 per replicate) or one of four dietary treatments. The dietary treatments were isocaloric, all-milk-protein milk replacers that contained 16.1, 18.5, 22.9, or 25.8% CP. Remaining calves in excess of the eight assigned per treatment were allocated randomly to experimental groups. On d 15, prior to the morning feeding, calves in the baseline group were transported to the University of Illinois Meat Science Laboratory and slaughtered for determination of body composition. Calves that were assigned to the four dietary treatments were fed milk replacer (reconstituted to 12.5% solids) at 12% BW daily in two equal feedings at 0800 and 1700. The amount of milk replacer fed was adjusted weekly as calves grew.

On d 31, calves were placed into metabolism stalls where they remained from d 31 to 42. Urine and feces were collected during d 36 to 42. Two calves were deemed unsuitable to be moved to the metabolism stalls because of leg problems; these calves (one on 18.5% CP and one on 22.9% CP) were maintained in their hutches from d 31 to 42 and no balance data were collected. On d 42, calves were returned to the individual calf hutches until d 56. On d 56, all calves were transported to the Meat Science Laboratory for slaughter and determination of body composition.

Housing Facilities

Individual calf hutches were placed on the north and east sides of a one-story building. The hutches were aligned in two rows north to south, and openings of the hutches faced to the middle of the rows. Each hutch was bedded with 10 to 15 cm of corncobs that were topped with a layer of straw.

During the 11-d collection period, calves were moved from the hutches to metabolism stalls that measured 122 × 43 cm. This size allowed the calf enough room to stand up and lie down, but restricted any movement back and forth or side to side. The calves were secured with a head stanchion and were tied with a rope halter. Calves stood on grates that allowed for the collection of feces and urine. Urine was directed into a plastic bucket via an aluminum funnel, which was located under the front half of the grate. The metabolism stalls were placed inside a building that was cooled and dehumidified by window air conditioners.

Sampling and Analysis of Milk Replacer

Each milk replacer was sampled (50 g) weekly, and samples were combined by treatment. An additional 10-g sample was taken daily at 0830 during the collection period and combined by treatment. The samples from the collection period were kept separate from the weekly samples. All samples were stored at −20°C until they were analyzed for N content by Kjeldahl (AOAC, 1984), energy by bomb calorimetry (1261 Isoperibol Calorimeter; Parr Instrument Co., Moline, IL), and fatty acids (FA). The FA were determined by gas chromatography (Shimadzu GC-17A; Shimadzu Scientific Instruments, Inc., Columbia, MD) of methyl esters formed by acid-catalyzed transesterification (Sukhija and Palmquist, 1988).

Utilization of Energy and Nitrogen

Total collections of feces and urine were made from d 36 to 42. Feces were collected in aluminum pans measuring 41.9 × 30.5 × 6.35 cm. Pans were changed daily at 0900. Collection pans were checked at 0730, 1200, 1630, and 2200 to ensure that feces were not contaminated with urine. Feces were dried at 55°C for 3 d, composited by calf, ground in a Wiley mill, and then stored in glass containers until laboratory analysis. Fecal samples were analyzed for contents of energy (1261 Isoperibol Calorimeter; Parr Instrument Co.), N (AOAC, 1984), and FA (Suhkija and Palmquist, 1988).

Urine was collected in 11.4-L plastic buckets. Urine was acidified by addition of 100 mL of 50% HCl to the buckets daily to minimize the loss of ammonia. Urine buckets were checked for fecal contamination at 0730, 1200, 1630, and 2200 and were changed at 0930 daily. The volume of urine was measured, and 1% of the volume was saved and combined by calf. At the end of the collection period, the composited urine samples were frozen and stored at −20°C until analysis. Urine was analyzed for gross energy (1261 Isoperibol Calorimeter; Parr Instrument Co.) by adsorbing urine onto cellulose pellets and N by Kjeldahl (AOAC, 1984) to determine energy and N utilization.

Apparent ME intake was estimated by subtracting gross energy lost in feces and urine from gross energy in feed consumed. Energy loss from combustible gas is not significant for calves fed a liquid diet (Gonzalez-Jimenez and Blaxter, 1962; Holmes and Davey, 1976) and was not measured.

Sampling and Analysis of Blood

Blood was sampled from each calf by puncture of the jugular vein every Monday at 0700, which was before feeding. Samples were collected into separate evacuated tubes (Vacutainer; Becton Dickinson and Co., Rutherford, NJ) containing sodium heparin or EDTA and were immediately placed on ice. Tubes then were centrifuged at 14,000 × g for 15 min to obtain plasma. Plasma was collected and stored at −20°C until analyzed for NEFA (kit number 990-75409; Wako Pure Chemical Industries, Ltd., Osaka 541, Japan), total protein (kit number 541-2; Sigma-Aldrich Chemical Co., St. Louis, MO), glucose (kit number 315-500; Sigma-Aldrich Chemical Co.), and urea N (kit number 535-A; Sigma-Aldrich Chemical Co.).

Body Growth Measurements and Determination of Body Composition

Calves were weighed every Monday at 0900, after feeding was complete. Stature measurements, including heart girth, wither height, and body length, were made at this time. Calculations of ADG of BW and stature measurements were made from these measurements. Rectal temperatures were measured to monitor calf health.

Calves were slaughtered for body composition analysis at the University of Illinois Meat Science Laboratory. Calves were weighed before slaughter, which was 14 to 16 h after the previous afternoon feeding; calves were not fed on the morning of slaughter. This weight was assumed to represent shrunk weight. The heart, kidneys, liver, and the empty digestive tract were weighed individually, and then were composited and ground as the visceral fraction. The entire viscera-free carcass, including head, hide, hooves, and tail, was ground and is referred to as the viscera-free carcass (VFC) fraction. Viscera and VFC were ground and mixed three times through a whole carcass grinder (model 801 GP15; Autio Co., Inc., Astoria, OR) fitted with a 1.3-cm plate. Representative samples were collected and frozen at −30°C.

Before analysis, subsamples of frozen viscera and VFC were ground and mixed through a grinder (model 52HF; Butcher Boy Ltd., Beith, U.K.) fitted with a 3-mm plate. Samples were refrozen until analysis. Frozen samples of ground viscera and VFC were analyzed for contents of water (AOAC, 1995), total lipid by chloroform-methanol extraction (Novakofski et al., 1989), N (AOAC, 1984), and ash (AOAC, 1995). Subsamples of ground frozen tissue were lyophilized at −30°C until weight was constant. Lyophilized samples were analyzed for content of energy by bomb calorimetry (1261 Isoperibol Calorimeter; Parr Instrument Co.).

Because weights and samples of viscera for some calves were lost, changes in viscera components and whole-body components could not be calculated for all calves. Consequently, only data for changes in composition of the VFC over the 6-wk experimental period are presented. Initial composition of the VFC in calves slaughtered at the end of the experimental period was assumed to be the same as that of the eight baseline calves slaughtered at the start of the experimental periods. The composition of VFC gain for each calf was calculated as the amounts of water, protein, fat, ash, and energy at slaughter minus the calculated amounts of those components present in each calf at the start of the experiment, as estimated from the average VFC composition of baseline calves.

Statistical Analysis

Data were subjected to regression analysis for a randomized complete block design using the GLM procedures of SAS (version 8; SAS Inst., Inc., Cary, NC). The full model tested for each variable contained effects of replicate, the percentage of CP in milk replacer as a continuous variable, the percentage of CP², and the interactions of CP and CP² with replicate. The interaction of replicate and CP² was not significant for any variable. Interactions and effects were removed sequentially from the model if P > 0.10. Type-III sums of squares were used in the analysis. Because of the dose-response nature of our experiment, designed to determine linear or curvilinear responses of calves to increasing protein contents (and decreasing non-protein energy sources) in milk replacers, means separation tests were not justified (Morris, 1999) and were not performed.

Blood variables were analyzed using the MIXED procedure of SAS. In addition to the factors described above, the model contained the effects of week as a repeated factor and the interactions of week with CP and CP². Values for samples obtained from individual calves at the end of the 2-wk adaptation period were included as covariates. Effects of replicate and calf were specified as random terms. The first-order autoregressive approach was used to fit the covariance structure of the repeated factor. Terms were removed sequentially from the model when P > 0.10. The arithmetic means are presented throughout along with the parameter estimates and standard errors from the final regression model for each variable.

Results and Discussion

Based on principles summarized elsewhere (Davis and Drackley, 1998), our original hypothesis was that dietary CP content could be limited to about 18 to 20% in dairy calves fed restricted amounts of milk replacer, without compromising growth. Such formulations would minimize cost to producers and minimize the environmental impacts of calf-rearing enterprises. Testing that hypothesis required detailed investigations of effects of diet composition on N and energy balance as well as body composition, because BW gain alone may not be an accurate predictor of lean mass deposition (Stobo et al., 1966; Owens et al., 1993).

Formulation and Nutrient Composition of Diets

The formulated chemical compositions of the four treatments and the standardization milk replacer are found in Table 1. The experimental diets were formulated to contain 14, 18, 22, and 26% CP, and the standardization diet was 20% CP. The desired CP contents were produced by varying the relative amounts of dried whey and whey protein concentrate. To manufacture isocaloric milk replacers, the amounts of fat and lactose decreased as the content of CP increased. The ratio of lactose:fat was kept similar among diets and averaged 2.35:1. All diets were formulated to provide adequate vitamins and minerals as specified by the NRC (1989). Whole milk fed to the second group of calves during the standardization period averaged 3.4% fat and 3.2% CP on an as-fed basis.

Table 1.

Formulated composition of milk replacers^a

		Formulated crude protein content, %
Nutrient	Standardization dietb	14	18	22	26
Dry matter, %	96.9	97.0	96.9	96.9	96.8
Crude protein, %	20.0	14.0	18.0	22.0	26.0
Crude fat, %	20.0	22.0	20.7	19.4	18.1
Ash, %	6.9	7.3	7.1	6.8	6.5
Metabolizable energy, kcal/kg	4,263	4,267	4,266	4,265	4,264
Lactose, %	46.9	49.9	47.9	45.9	43.8
Calcium, %	1.1	1.0	1.0	1.1	1.1
Phosphorus, %	0.7	0.7	0.7	0.7	0.7
Copper, mg/kg	11.5	11.3	11.4	11.6	11.7
Zinc, mg/kg	104	104	104	105	105
Selenium, mg/kg	0.3	0.3	0.3	0.3	0.3
Sodium, %	0.4	0.4	0.4	0.5	0.5
Chloride, %	1.4	1.5	1.4	1.4	1.3
Cobalt, mg/kg	1.5	1.5	1.5	1.5	1.5
Iodine, mg/kg	7.2	5.6	6.7	7.8	8.8
Iron, mg/kg	106	106	106	106	106
Magnesium, %	0.1	0.1	0.1	0.1	0.1
Manganese, mg/kg	44.2	45.1	44.5	44.0	43.4
Potassium, %	1.1	1.3	1.2	1.1	1.0
Vitamin A, IU/kg	66,056	66,056	66,056	66,056	66,056
Vitamin D, IU/kg	23,065	23,065	23,065	23,065	23,065
Vitamin E, IU/kg	220	220	220	220	220
Thiamin, mg/kg	15.7	15.8	15.8	15.7	15.6
Riboflavin, mg/kg	22.3	22.2	22.2	22.3	22.3
Vitamin B6, mg/kg	6.4	6.9	6.5	6.2	5.9
Vitamin B12 μg/kg	60.2	42.8	54.4	65.9	77.5
Pantothenic acid, mg/kg	44.1	45.8	44.7	43.5	42.4
Niacin, mg/kg	46.0	47.4	46.5	45.5	44.6
Folic acid, mg/kg	3.4	3.4	3.4	3.4	3.4
C, mg/kg	100	100	100	100	100
Biotin, mg/kg	0.3	0.4	0.4	0.3	0.3
Choline, mg/kg	1,445	1,480	1,456	1,432	1,408

^aAs reported by the manufacturer (as-fed basis).

^bFed during d 3 to 14 for Replicate 1. Calves in Replicate 2 were fed whole milk (contained 3.4% fat and 3.2% CP on an as-fed basis) during d 8 to 14.

Our data were analyzed and are discussed as a function of increasing CP content in the milk replacers. Because composition of isocaloric diets by necessity is confounded by the decreasing amounts of fat and lactose as CP increased, a more accurate description of the diets is that the ratio of protein:nonprotein energy sources increased linearly at a constant energy density. Previous studies using lambs (Jagusch et al., 1970; Norton et al., 1970) and calves (Gerrits et al., 1996) have established that increasing CP content, and the corresponding increases of the protein:energy ratio, have marked effects on growth and body composition of milk-fed preruminants. This principle held true for calves fed isocaloric diets with increasing protein:energy ratios (Donnelly and Hutton, 1976a,b) as well as for older calves fed increasing amounts of protein with fixed amounts of non-protein energy sources (Gerrits et al., 1996); the latter situation resulted in confounding because of greater ME intake by calves as protein supply was increased. Statistical interpretation of data from our study would not change if regression analyses were performed using the increasing CP:energy ratio rather than the CP content; consequently, our results are presented as a function of increasing dietary CP content for simplicity and ease of interpretation in the field.

The measured chemical composition of the experimental milk replacers varied slightly from the formulated composition as shown in Table 2. The negative control was designed to be 14% CP, but deviations during manufacturing resulted in an actual content of 16.1% CP. Therefore, the contents of CP in milk replacers fed varied from 16.1 to 25.8%. Milk replacers were formulated to be isocaloric; measured gross energy content increased slightly as the content of CP increased. Total FA concentration decreased as the amount of fat decreased (i.e., as CP increased), and averaged 19.0, 18.1, 17.1, and 16.8 g/100 g of milk replacer.

Table 2.

Measured chemical composition of the experimental milk replacers

	Formulated crude protein content, %
Component	14	18	22	26
DM, %	97.0	97.2	97.1	97.0
CP, % of DM	16.1	18.5	22.9	25.8
Gross energy, Mcal/kg of DM	5.04	5.06	5.13	5.11
Fatty acids, g/100 of DM
16:0	4.69	4.49	4.08	4.03
18:0	2.60	2.47	2.20	2.16
18:1cis-9	7.56	7.15	6.54	6.37
18:1cis-11	0.54	0.51	0.43	0.42
18:2	1.92	1.85	2.06	1.98
Total	19.0	18.1	17.1	16.8
Total fat, %a	21.1	20.1	19.0	18.6

^aTotal fatty acids/0.9 (Davis and Drackley, 1998).

Most previous studies on protein requirements for dairy calves used skim milk based milk replacers (Jacobsen, 1969; Donnelly and Hutton, 1976a,b). Because of its increased cost, dried skim milk has been replaced in milk replacer formulations by whey proteins, principally whey protein concentrate and dried whey (Davis and Drackley, 1998). Although whey proteins are highly acceptable sources of protein in milk replacers for young calves (Terosky et al., 1997; Davis and Drackley, 1998; Lammers et al., 1998), no studies were available that have systematically quantified the responses to increasing contents of whey protein in milk replacer.

Consequently, diets used in our experiment are more representative of milk replacers used at the present time.

General Calf Health

Calves in both replicates scoured during the 2-wk adaptation period. In the first replicate, 83.3% of calves scoured and the average duration was 3.5 d, whereas 88% of calves scoured for an average of 4 d in the second replicate. During the experimental period, some scouring was noted during the first 10 d on the experimental diets; 67.7% of calves in replicate 1 and 8% of calves in replicate 2 scoured, with an average duration of 4 d. Two calves from the first replicate died. Both calves were being fed the 22% CP milk replacer, but treatment was not a suspected cause of death. One calf died of heat stroke on d 6 and the other of unknown causes on d 5 of the experimental period. Because scouring occurred mainly during the adaptation period and the first week of the experimental period, we are confident that comparisons among treatments were not compromised.

Body Weight, Total Dry Matter Intake, Efficiency of Gain, and Stature Changes

Calves on all treatments had similar initial BW as designed (Table 3). Final BW increased linearly as the amount of CP in the diet increased. The ADG of BW increased linearly as dietary CP increased. Our results agree with previous findings on effects of dietary CP content in calves of similar age and BW fed skim milk-based milk replacers (Donnelly and Hutton, 1976a). The mean DMI for the 6-wk experimental period increased slightly as dietary CP increased (Table 3), although the regression coefficient for the linear effect of increasing CP did not achieve statistical significance (P < 0.11). Calves were fed at a constant percentage of BW, adjusted weekly as calves grew; consequently, the greater DMI for calves fed higher CP in milk replacers is a result of the greater growth rates stimulated by those diets.

Table 3.

Means and parameter estimates for dry matter intake, body weight, efficiency of gain, and stature measurements for calves fed milk replacers that differed in crude protein content

	CP in milk replacer, %				Parameter estimatesa
Variable	16.1	18.5	22.9	25.8	Intercept	CP linear	P b	R2
n	8	9	8	8
Body weight, kg
Initial	44.4	45.4	45.4	44.9	44.9 ± 3.92	0.03 ± 0.19	0.85	0.001
Final	60.1	64.2	68.4	70.7	43.7 ± 6.63	1.06 ± 0.32	0.002	0.27
ADGc	0.38	0.45	0.56	0.62	0.022 ± 0.083	0.024 ± 0.004	<0.001	0.61
DMI, kg/d	0.739	0.771	0.785	0.798	0.660 ± 0.070	0.0055 ± 0.0033	0.11	0.08
Gain:feedc	0.507	0.587	0.708	0.778	0.120 ± 0.066	0.027 ± 0.003	<0.001	0.76
Wither height, cm
Finald	85.2	87.5	86.7	87.4	81.15 ± 3.54	0.33 ± 0.17	0.32	0.35
Rate of change, cm/d	0.113	0.124	0.168	0.156	0.029 ± 0.055	0.0054 ± 0.0026	0.05	0.12
Heart girth, cm
Final	93.0	94.7	96.7	97.2	86.53 ± 3.81	0.427 ± 0.180	0.03	0.35
Rate of change, cm/de	0.208	0.228	0.283	0.302	0.203 ± 0.056	0.010 ± 0.002	<0.001	0.39
Body length, cm
Finalcd	79.9	81.6	82.2	82.1	66.43 ± 4.17	0.52 ± 0.20	0.07	0.72
Rate of change, cm/dc	0.231	0.234	0.287	0.277	0.015 ± 0.068	0.0073 ± 0.0032	0.03	0.68

^aValues are estimates ± SE. The R² represents the coefficient of determination for the final regression model for each variable.

^bProbability of greater F for the linear effect of increasing CP.

^cReplicate effect (P < 0.05).

^dInteraction of replicate by CP linear (P < 0.10).

^eReplicate effect (P < 0.10).

Efficiency of gain (gain:feed) increased linearly as dietary CP increased. The gain:feed ratios for calves fed milk replacer containing 25.8% CP were similar to the highest values reported by Diaz et al. (2001) for calves fed a 30% CP milk replacer in amounts to allow ADG of about 1 kg/d. Increases in gain:feed have been clearly demonstrated in response to increased rates of milk replacer feeding that allow greater ADG (Khouri and Pickering, 1968; Diaz et al., 2001). The higher values obtained in our study also are comparable to gain:feed ratios reported for the young of other species that consume milk ad libitum, including lambs (Hodge, 1974; Greenwood et al., 1998) and pigs (Hodge, 1974; Kim et al., 2001). Improvement of gain:feed in animals fed for ad libitum intakes relative to animals fed at restricted intakes is attributed to a dilution of maintenance expenditures. It is noteworthy, however, that gain:feed varied from 0.51 to 0.78 in our calves fed at the same restricted percentage of BW, with the only difference being diet composition (i.e., increasing CP content and CP:energy, and decreasing lactose and fat contents at similar ME concentrations).

Final wither height, heart girth, and body length, as well as rates of change of these measurements, are shown in Table 3. Measurements at the beginning of the experiment did not differ among dietary treatments (data not shown). Final wither height did not differ significantly among dietary groups, but rate of change in wither height increased linearly as dietary CP increased. Final heart girth, rate of change in heart girth, and rate of change in body length increased linearly (P < 0.05) as dietary CP increased; final body length tended (P < 0.07) to increase linearly. Increases in stature measurements demonstrate that increases in BW and ADG of BW as dietary CP increased were increases in frame size, not just gain of gut fill or body fat.

A plot of BW vs. week of treatment is shown in Figure 1. Calves were fed the standardization diet during wk -2 and -1. At wk 0, calves were placed on the experimental diets. During the experimental period, increases in body weight of calves were greater as dietary CP increased. During wk 3 and 4, calves were placed into metabolism stalls and rates of gain slowed. Because DMI did not change during this time, lower rates of gain indicate decreases in efficiency of gain. Decreased gain:feed evidently resulted from the changes in housing and environment when calves were confined in the metabolism stalls. Rates of gain increased again during wk 5 and 6 when calves were returned to hutches. Whether these changes in ADG due to housing affected composition of gain during those times cannot be determined from our data. Because calves on all dietary treatments went through the changes of housing together, we assume that impacts on body composition and nutrient utilization were uniform across dietary treatments and are discussed accordingly.

Figure 1.

Mean BW of calves fed milk replacers differing in content of CP.

Nitrogen and Energy Balances

Nitrogen utilization data from the balance study during wk 3 and 4 are in Table 4. As the amount of CP in the diet increased, N intake increased linearly as designed. Fecal N showed a significant quadratic relationship, increasing as dietary CP increased from 16.1 to 22.9%, and then decreasing for calves fed 25.8% CP. Quantities of N absorbed, excreted in urine, and retained increased linearly as dietary CP increased. Absorbed N expressed as a percentage of intake N (i.e., N apparent digestibility) was not affected significantly but tended (P < 0.08) to demonstrate a quadratic response. Calculated true digestibility of N was affected in a quadratic manner, with highest values for calves fed the lowest and highest CP diets. No biological explanation is offered for this pattern of response, although differences were relatively small. Retained N expressed either as a percentage of intake N or as a percentage of absorbed N increased linearly as dietary CP increased.

Table 4.

Means and parameter estimates for utilization of nitrogen and energy by calves fed milk replacers that differed in crude protein content

	CP in milk replacer, %				Parameter estimatesa
Variable	16.1	18.5	22.9	25.8	Intercept	CP linear	P b	CP quadratic	P c	R2
n	8	8	7	8
BW, kg	51.8	54.3	56.6	57.6
DMI, kg	0.74	0.78	0.80	0.82
Intake N, g/d	19.2	23.3	29.4	33.9	−4.6 ± 2.6	1.49 ± 0.12	<0.001	NSh	—	0.83
Fecal N, g/dd	2.3	3.3	3.6	3.3	−11.7 ± 6.7	1.41 ± 0.65	0.04	−0.03 ± 0.02	0.05	0.40
Urine N, g/d	9.3	11.0	12.6	15.0	0.5 ± 1.5	0.55 ± 0.07	<0.001	NS	—	0.68
Absorbed N, g/d	16.9	20.0	25.8	30.6	−5.9 ± 2.7	1.40 ± 0.13	<0.001	NS	—	0.80
Absorbed N, % of intake N	87.7	85.5	87.8	90.2	126.3 ± 26.2	−4.3 ± 2.6	0.106	0.11 ± 0.06	0.083	0.39
True N digestibility, %e	96.3	92.9	93.8	95.5	153.8 ± 26.3	−5.3 ± 2.6	0.05	0.13 – 0.06	0.05	0.33
Retained N, g/dd	7.6	9.0	13.2	15.6	−7.6 ± 2.1	0.86 ± 0.10	<0.001	NS	—	0.75
Retained N, % of intake Nd	39.0	38.4	44.6	45.9	20.7 ± 5.7	0.85 ± 0.26	0.003	NS	—	0.45
Retained N, % of absorbed Nd	44.3	44.7	50.7	50.9	27.8 ± 6.0	0.80 ± 0.28	0.008	NS	—	0.38
Biological value, %de	71.1	68.7	69.5	67.3	72.9 ± 5.4	−0.28 ± 0.25	0.26	NS	—	0.19
Intake energy, Mcal/d	3.73	3.99	4.10	4.24	3.02 ± 0.39	0.048 ± 0.018	0.016	NS	—	0.18
Fecal energy, Mcal/dg	0.18	0.25	0.23	0.22	0.19 ± 0.07	0.0027 ± 0.0033	0.43	NS	—	0.14
Urinary energy, Mcal/d	0.19	0.19	0.19	0.19	0.17 ± 0.02	0.001 ± 0.0007	0.25	NS	—	0.04
DE, Mcal/d	3.55	3.74	3.88	4.02	2.86 ± 0.38	0.045 ± 0.018	0.019	NS	—	0.18
DE, % of intake energyd	95.1	93.7	94.4	94.8	93.7 ± 1.6	0.010 ± 0.076	0.89	NS	—	0.13
ME, Mcal/d	3.36	3.55	3.68	3.82	2.69 ± 0.37	0.044 ± 0.018	0.019	NS	—	0.18
ME, % of intake energyg	90.1	88.9	89.7	90.2	88.2 ± 1.6	0.045 ± 0.077	0.56	NS	—	0.14

^aValues are estimates ± SE. The R² represents the coefficient of determination for the final regression model for each variable.

^bProbability of greater F for the linear quadratic effect of increasing CP.

^cProbability of greter F for the quadratic effect of increasing CP.

^dReplicate effect (P < 0.05).

^eCalculated assuming metabolic fecal N of 2.2 g of N/kg of DMI (Donnelly and Hutton, 1976a).

^fCalculated according to Thomas-Mitchell equation (Mitchell, 1924).

^gReplicate effect (P < 0.07).

^hNonsignificant (P > 0.10).

Davis and Drackley (1998) concluded from a summary of experiments that an average of 30 g of N was retained per kilogram of BW gained by milk-fed calves. This value was adopted by the NRC in its most recent revision of Nutrient Requirements of Dairy Cattle (NRC, 2001). If this value is applied to our measured N retention values, calves would be predicted to gain 0.23, 0.27, 0.40, and 0.47 kg/d when fed milk replacers containing 16.1, 18.5, 22.9, and 25.8% CP, respectively. Actual ADG during wk 3 to 4 when N balance data were collected were 0.15, 0.23, 0.35, and 0.38 kg/d, respectively. Therefore, ADG predicted by N retention overestimated actual ADG in agreement with others (Gerrits et al., 1996; Diaz et al., 2001). Possible sources of error in N balance studies include losses of N in scurf that is not measured, incomplete collection of feces or urine due to splashing or evaporation, and losses of volatile N compounds in feces during oven drying. The amount of N retained per kilogram of actual ADG, based on our measured N retention and BW gains, would be 19.7, 25.3, 26.8, and 24.4 g of N/kg ADG.

Biological value (BV) was calculated to determine the efficiency of deposition of digested N in the bodies of the calves. The BV was calculated using the Thomas-Mitchell equation (Mitchell, 1924) as follows:

where N intake, fecal N, and urinary N were measured during the metabolism trial. Endogenous urinary N was calculated as 0.2 g/kg^0.75 (ARC, 1965; Davis and Drackley, 1998; NRC, 2001), and metabolic fecal N was calculated using a value of 2.2 g of N/kg of DMI (Donnelly and Hutton, 1976a; Davis and Drackley, 1998). The BV was not affected significantly (P = 0.26) by treatment, although means trended downward as dietary CP increased. By definition, BV refers to the efficiency of use of absorbed protein for growth above maintenance; therefore, there were no significant differences among diets in how efficiently the calves utilized N for growth. However, gross efficiency of N use by calves (i.e., N retained as a proportion of N intake) increased as dietary CP increased, probably because of the stimulation of greater growth rates as dietary CP increased. The latter values may be more relevant for use in determining effects of nutrition on N budgets for dairy operations.

The NRC (2001) assumed a constant value for BV of 80% for protein in whole milk or milk replacers based on milk proteins. We are aware of no studies other than that of Donnelly and Hutton (1976a) in which BV has been measured over a range of dietary CP contents in preruminant calves. In that experiment, milk replacers based on dried skim milk were fed; CP content ranged from 15.7 to 31.5%. The BV were calculated by the same methodology used in our study and decreased from 83% for the 15.7% CP diet to 56% for the 31.5% CP diet. Babella et al. (1988) calculated the true BV for a milk replacer (28% CP) based entirely on whey proteins to be 66.2%, similar to the BV for our highest CP diet (67.3%). Terosky et al. (1997) reported a value for “apparent biological value” of 73.7% for a milk replacer (20.7% CP) based entirely on whey proteins (whey and whey protein concentrate). Diaz et al. (2001) reported that the “biological value of absorbed protein” ranged from 57.0 to 70.0% for a milk replacer (30% CP) fed at different DMI. However, in both of these latter studies (Terosky et al., 1997; Diaz et al., 2001), values were calculated as retained N divided by N apparently absorbed and were not corrected for estimated endogenous N and metabolic fecal N. Consequently, their values are comparable to our values for N retained as a percentage of N absorbed. Correction of those values for endogenous urinary N and metabolic fecal N would increase means to greater than 80% in both studies. Slightly lower values in our study may coincide with the decreased gain:feed observed during the time calves were housed in metabolism stalls (Figure 1). We conclude, therefore, that the NRC (2001) value of 80% for BV is a reasonable estimate, but that additional research should investigate factors that impact BV in calves fed whey protein-containing milk replacers.

Energy balance data are shown in Table 4. Intake energy increased linearly as dietary CP increased and averaged 4.02 Mcal/d during the week of the metabolism study. Greater energy intake is explained by the slight increase in energy content of the milk replacers as the amount of CP increased (Table 2) and by the tendency for greater consumption of milk replacer as calves grew faster (Table 3). Daily fecal and urinary energy losses were not different among treatments and averaged 0.22 and 0.19 Mcal/d, respectively. Both DE and ME increased linearly as dietary CP increased. Expressed as percentages of total intake energy, DE and ME were unaffected by dietary CP content and averaged 94.5 and 89.7%, respectively. Therefore, changes in the sources of energy (i.e., increased CP and decreased fat and lactose) in the diets did not affect the digestibility or metabolizability of energy.

The NRC (2001) assigned a value for DE of 0.97 × gross energy for whole milk and milk-derived products, and ME was assumed to be 0.96 × DE; thus, ME would be 93% of gross energy. Those values were obtained from studies in which whole milk or skim milk-based milk replacers were fed to calves (Davis and Drackley, 1998). Calculations from data reported by Diaz et al. (2001) indicate that DE was 95.0% of gross energy and ME was 91.9% of gross energy in milk replacers based on whey proteins. Taken together with our data, the values for DE and ME used by NRC (2001) appear to be slightly high for young calves consuming whey-protein-based milk replacers.

Although the content of total FA in the milk replacers decreased as CP increased, total FA intake(x̅ = 140 g/d) did not differ significantly among treatments (data not shown). The lack of differences among treatments can be explained by the fact that as calves fed milk replacers containing higher CP content and lower total FA grew faster, they were fed larger amounts of milk replacer. Total FA absorbed per day was not affected by treatments (data not shown) and averaged 136.7 g/d. The apparent total tract digestibility of total FA increased linearly as dietary CP increased (97.3, 96.0, 97.8, and 98.4% for calves fed 16.1, 18.5, 22.9, and 25.8% CP, respectively), perhaps as a result of the decreasing fat content of the diet as CP increased. Although statistically significant, these differences were small and likely of little biological significance.

Blood Metabolites

Concentrations of metabolites in plasma sampled before feeding are summarized in Table 5. Plasma urea N increased linearly as dietary CP content increased. The interaction of diet × time was significant (Figure 2). Urea N decreased from wk 0 to 3 in calves fed 16.1, 18.5, and 22.9% CP, but increased at wk 1 for calves fed 25.8% CP. These data may indicate that calves fed the milk replacer containing 25.8% CP did not utilize dietary N as efficiently as calves fed milk replacers with lower CP content. Alternately, the distinct decrease in plasma urea N with time in calves fed 16.1, 18.5, or 22.9% CP may indicate that those calves did not receive adequate amounts of protein in their diets compared with calves fed 25.8% CP. Urea N concentrations in our study were lower than the values (7.0 to 8.0 mg/dL) reported by Terosky et al. (1997) and those (9.3 to 13.1 mg/dL) reported by Diaz et al. (2001); calves in the latter study were fed a milk replacer that contained 30% CP.

Table 5.

Means and parameter estimates for concentrations of compounds in plasma from calves fed milk replacers that differed in crude protein content

	CP in milk replacer, %				Parameter estimatesa
Variable	16.1	18.5	22.9	25.8	Intercept	CP linear	P b	CP quadratic	P c
n	8	9	8	8
Urea N, mg/dLd	2.9	3.3	4.6	6.0	−3.82 ± 1.49	0.37 ± 0.070	<0.001	NSf	—
Total protein, g/dLe	5.4	5.2	5.6	5.8	1.96 ± 0.68	0.027 ± 0.018	0.15	NS	—
Glucose, mg/dLe	77.8	80.5	83.0	78.2	−10.14 ± 39.74	7.02 ± 3.95	0.09	−0.166 ± 0.094	0.09
NEFA, μeq/Le	311	294	368	337	122  ± 86	5.2 ± 3.9	0.20	NS	—

^aValues are estimates ± SE.

^bProbability of greater F for the linear effect of increasing CP.

^cProbability of greater F for the quadratic effect of increasing CP.

^dInteraction of week and linear effect of increasing CP (P < 0.05).

^eWeek effect (P < 0.001).

^fNonsignificant (P > 0.10).

Figure 2.

Concentration of urea in plasma from calves fed milk replacers differing in content of CP. The interaction of the linear effect of increasing CP content and week was significant (P < 0.05). Standard errors for individual CP × week means ranged from 0.3 to 1.6 mg/dL.

The regression of total protein concentration in plasma on dietary CP content was not significant (P = 0.15; Table 5). Concentrations of glucose and NEFA were not affected significantly by milk replacer CP content, although a tendency (P < 0.09) for a quadratic effect was detected for plasma glucose. Means tended to be greater for the intermediate CP contents. Treatment × time interactions were not significant for total protein, glucose, or NEFA in plasma, but time effects were significant for each (Figure 3). Concentrations increased during wk 3 and 4 when calves were placed into metabolism stalls. Increases of glucose and NEFA might reflect greater sympathetic nervous system activity or adrenal catecholamine release in response to the change in housing environment, and might also be related to the lower gain:feed observed when calves were in metabolism stalls.

Figure 3.

Concentrations of metabolites in plasma by week of study, averaged across all dietary treatments. Bars represent standard errors. The effect of week was significant for each variable (P < 0.001).

Composition of Viscera-Free Carcass and Visceral Tissue

Means and parameter estimates for final shrunk BW and components of VFC from baseline calves and calves fed experimental milk replacers are presented in Table 6. Final shrunk BW averaged 94.2% of final live weight, which is similar to data of Diaz et al. (2001) and assumptions made by NRC (2001). Quadratic effects were detected for VFC weight expressed as percentages of final live weight or shrunk weight, but differences were small and unexplained. Water content in final VFC increased linearly, and content of protein tended (P < 0.10) to increase linearly as dietary CP increased. In contrast, contents of fat, ash, and energy decreased linearly as dietary CP increased. Although means for baseline calves were not compared statistically with calves fed experimental treatments, VFC for calves fed the lowest CP milk replacer contained substantially more fat and energy, and less water, than baseline calves. Increasing dietary CP resulted in VFC composition becoming closer to that of baseline calves. Chemical composition of VFC for baseline calves was very similar to that reported for digesta-free whole body composition (including viscera) of baseline calves in the study of Donnelly and Hutton (1976b) and empty body (including viscera) of baseline calves used by Diaz et al. (2001).

Table 6.

Means and parameter estimates for weight and components of visceral-free carcass (VFC) of calves fed milk replacers that differed in crude protein content

		CP in milk replacer, %				Parameter estimatesa
Variable	Baselineb	16.1	18.5	22.9	25.8	Intercept	CP linear	P c	CP quadratic	P d	R2
n	8	8	9	8	8
Shrunk BW, kge	43.7 ± 3.5	56.5	60.4	64.6	66.8	43.2 ± 6.4	1.01 ± 0.30	0.002	NSg	—	0.35
% of final live weightf	95.2 ± 0.2	94.0	94.1	94.4	94.5	94.3 ± 1.0	0.032 ± 0.046	0.49	NS	—	0.38
VFC, kg	34.4 ± 2.9	44.5	46.6	49.5	52.2	32.1 ± 5.05	0.77 ± 0.24	0.003	NS	—	0.25
% of live weight	74.9 ± 3.6	74.0	72.5	72.2	74.0	104.7 ± 15.2	−3.28 ± 1.49	0.04	0.079 ± 0.035	0.04	0.49
% of shrunk BW	78.7 ± 3.4	78.7	77.1	76.5	78.3	110.4 ± 14.8	−3.43 ± 1.44	0.03	0.081 ± 0.034	0.03	0.66
VFC composition
Water, %f	72.36 ± 0.57	68.25	69.10	70.16	70.53	64.16 ± 0.90	0.240 ± 0.042	<0.001	NS	—	0.55
Protein, %	19.45 ± 0.26	19.61	20.05	20.02	20.24	18.91 ± 0.64	0.052 ± 0.030	0.10	NS	—	0.08
Fat, %	3.96 ± 0.23	7.15	6.25	5.50	5.17	10.08 ± 0.89	−0.195 ± 0.042	<0.001	NS		0.41
Ash, %	4.70 ± 0.26	5.09	5.17	4.82	4.70	5.94 ± 0.42	−0.048 ± 0.020	0.02	NS		0.15
Energy, Mcal/kg of DM	5.04 ± 0.07	5.44	5.29	5.29	5.14	5.82 ± 0.14	−0.025 ± 0.006	<0.001	NS	—	0.34
VFC components
Water, kg	24.87 ± 2.17	30.34	32.19	34.72	36.78	20.02 ± 3.45	0.65 ± 0.16	<0.001	NS	—	0.34
Protein, kg	6.68 ± 0.61	8.71	9.34	9.92	10.56	5.89 ± 1.04	0.18 ± 0.05	<0.001	NS	—	0.30
Fat, kg	1.36 ± 0.22	3.19	2.93	2.73	2.74	3.84 ± 0.65	−0.045 ± 0.031	0.15	NS	—	0.07
Ash, kg	1.62 ± 0.32	2.26	2.40	2.38	2.43	2.09 ± 0.26	0.013 ± 0.012	0.29	NS	—	0.04
Energy, Mcal	110.4 ± 14.8	76.86	76.38	78.09	79.39	71.64 ± 9.63	0.29 ± 0.46	0.53	NS	—	0.01

^aValues are estimates ± SE. The R² represents the coefficient of determination for the final regression model for each variable.

^bValues are means ± SD. Calves were fed a milk replacer containing 20% CP (Replicate 1) or whole milk (Replicate 2).

^cProbability of greater F for the linear effect of increasing CP.

^dProbability of greater F for the quadratic effect of increasing CP.

^eReplicate effect (P < 0.06).

^fReplicate effect (P < 0.05).

^gNonsignificant (P > 0.10).

Amounts of water and protein in the final VFC increased linearly as dietary CP increased (Table 6). The VFC for calves on all treatments contained a ratio of water:protein of approximately 3.48:1. The trend for decreased amounts of fat in the VFC as dietary CP increased did not achieve statistical significance (P = 0.15). As dietary CP increased, the protein:fat ratio in the final VFC increased from 2.73:1 to 3.85:1. Amounts of ash and energy in final VFC were not affected by dietary CP content. We also calculated energy content in VFC from amounts of fat and protein using energy values obtained for fat and protein from calf tissue of 9.290 kcal/g fat and 5.545 kcal/g protein (Donnelly and Hutton, 1976b). Values calculated with this approach (77.9, 79.0, 80.4, and 84.0 Mcal) were slightly higher than those calculated from results of bomb calorimetry of total VFC tissue (76.9, 76.4, 78.1, and 79.4 Mcal, respectively).

The amounts of components in VFC gained over the 42-d experiment and the composition of VFC gain on a percentage basis can be found in Table 7. Gains in VFC weight and amounts of water and protein gained increased linearly as dietary CP increased. The amount of fat gained in VFC tended (P = 0.10) to decrease linearly; amounts of ash and energy gained were not affected by diet. The sum of ADG of water, protein, fat, and ash in VFC (data not shown) accounted for an average of 69% of ADG of BW across diets. Given that VFC represented 73% of final live BW, this result suggests that changes in visceral mass, gut fill, and blood accounted proportionately for the remainder of ADG of BW. When components of VFC gain were expressed as percentages of the total VFC gain, water content of gain increased, whereas content of fat decreased as dietary CP increased. The content of protein in VFC gain was not affected by diets; ash content tended (P = 0.06) to decrease linearly.

Table 7.

Means and parameter estimates for the amounts of components gained and the composition of gain in viscera-free carcass (VFC) of calves fed milk replacers that differed in crude protein content

	CP in milk replacer, %				Parameter estimatesa
Component	16.1	18.5	22.9	25.8	Intercept	CP linear	P b	R2
n	8	9	8	8
Gain of VFC weight, kg	11.25	12.67	15.56	18.64	−1.02 ± 2.87	0.75 ± 0.14	<0.001	0.49
Gain in VFC
Water, kg	6.31	7.64	10.17	12.51	−3.95 ± 2.01	0.63 ± 0.095	<0.001	0.58
Protein, kg	2.25	2.74	3.32	4.03	−0.55 ± 0.65	0.17 ± 0.031	<0.001	0.51
Fat, kg	1.88	1.59	1.39	1.41	2.53 ± 0.57	−0.046 ± 0.027	0.10	0.09
Ash, kgcd	0.70	0.81	0.78	0.86	0.10 ± 0.28	0.031 ± 0.013	0.20	0.18
Energy, Mcal	30.56	29.09	30.78	32.61	25.45 ± 6.40	0.25 ± 0.30	0.41	0.02
Composition of gain
Water, %	58.38	59.43	65.74	66.12	43.47 ± 4.13	0.91 ± 0.20	<0.001	0.41
Protein, %	20.25	21.59	20.72	22.01	18.81 ± 1.85	0.11 ± 0.09	0.21	0.05
Fat, %	15.36	12.66	8.70	6.74	29.32 – 3.24	−0.89  – 0.15	<0.001	0.52
Ash, %c	6.01	6.32	4.84	5.12	7.67 ± 1.40	−0.13 ± 0.065	0.06	0.25
Energy, Mcal/kgc	2.75	2.31	2.00	1.74	4.37 ± 0.23	−0.099 ± 0.011	<0.001	0.76

^aValues are estimates ± SE. The R² represents the coefficient of determination for the final regression model for each variable.

^bProbability of greater F for the linear effect of increasing CP.

^cReplicate effect (P < 0.05).

^dEffect of replicate × CP linear (P < 0.07).

Because of the increased water, increased protein, and decreased fat, the energy content per kilogram of VFC gain decreased linearly as dietary CP increased (Table 7). Energy gain in VFC calculated from gains of fat and protein, using energetic values for fat and protein from calf tissue of 9.290 kcal/g of fat and 5.545 kcal/g of protein (Donnelly and Hutton, 1976b), were 29.9, 30.0, 31.3, and 35.4 Mcal for 16.1, 18.5, 22.9, and 25.8% CP, respectively, which agreed well with values calculated from results of bomb calorimetry of baseline and final VFC tissue (30.6, 29.1, 30.8, and 32.6; Table 7).

Although data were not available for all calves, weights and composition of the visceral fraction in baseline and treatment calves are shown in Table 8. Diet composition did not affect weights of liver or heart. Kidney weight increased quadratically, with highest values for calves fed the diet containing 22.9% CP. Total viscera weight (including digestive tract, heart, lungs, liver, kidneys, spleen, and pancreas) increased linearly as dietary CP increased. In contrast, total visceral weight expressed as a percentage of live BW tended (P < 0.06) to decrease linearly. Weights of VFC and total viscera accounted for approximately 89.5 and 88.7% of final live BW for baseline and treatment calves, respectively; the remainder of live BW would be accounted for by blood (approximately 5% of BW), digesta fill (approximately 2% of BW), and loss of hydration from live BW to shrunk BW (Bartlett, 2001). Similar proportions of body components have been reported by others (Bartlett, 2001; Diaz et al., 2001; Tikofsky et al., 2001).

Table 8.

Means and parameter estimates for weights and composition of visceral fraction for calves fed milk replacers that differed in crude protein content

		CP in milk replacer, %				Parameter estimatesa
Variable	Baselineb	16.1	18.5	22.9	25.8	Intercept	CP linear	P c	CP quadratic	P d	R2
Visceral weights
n	4	4	4	4	4
Liver, kg	0.87 ± 0.18	1.09	1.13	1.25	1.19	0.89 ± 0.23	0.014 ± 0.011	0.22	NSh	—	0.10
Heart, kg	0.27 ± 0.02	0.39	0.42	0.44	0.43	0.33 ± 0.08	0.004 ± 0.004	0.26	NS	—	0.08
Kidneys, kg	0.21 ± 0.06	0.25	0.31	0.33	0.29	−1.08 ± 0.51	0.13   – 0.05	0.02	−0.003 ± 0.001	0.03	0.38
Total, kg	6.38 ± 0.69	9.34	9.65	10.22	10.59	7.26 ± 1.23	0.13 ± 0.06	0.04	NS	—	0.25
Total, % BW	14.57 ± 1.20	16.25	15.38	15.78	14.78	17.79 ± 1.13	−0.11 ± 0.05	0.06	NS	—	0.21
Visceral composition
n	8	5	7	4	6
Water, %e	81.55 ± 0.50	76.82	77.32	78.66	80.07	72.89 ± 1.51	0.31 ± 0.067	<0.001	NS	—	0.65
Protein, %f	12.94 ± 0.20	11.80	11.91	11.94	12.46	10.04 ± 0.96	0.073 ± 0.34	0.10	NS	—	0.25
Fat, %f	5.32 ± 0.07	11.94	11.61	10.70	8.16	19.69 ± 1.86	−0.40 ± 0.083	<0.001	NS	—	0.56
Ash, %e	0.86 ± 0.08	0.62	0.57	0.56	0.61	0.41 ± 0.12	0.0032 ± 0.0053	0.55	NS	—	0.45
Energy, Mcal/kg DMg	6.21 ± 0.09	7.02	7.14	6.94	6.73	9.36 ± 1.09	−0.104 ± 0.059	0.10	NS	—	0.22

^aValues are estimates ± SE. The R² represents the coefficient of determination for the final regression model for each variable.

^bValues are means ± SD. Calves were fed a milk replacer containing 20% CP (Replicate 1) or whole milk (Replicate 2).

^cProbability of greater F for the linear effect of increasing CP.

^dProbability of greater F for the quadratic effect of increasing CP.

^eReplicate effect (P < 0.05).

^fReplicate effect (P < 0.10).

^gEffects of replicate (P < 0.10) and replicate × CP linear (P < 0.10).

^hNonsignificant (P > 0.10).

Water content of final visceral tissue increased linearly, and protein content tended (P < 0.10) to increase linearly, as dietary CP content increased (Table 8). Fat content of visceral tissue decreased linearly, whereas ash content was unaffected. As a result of these changes, energy content of visceral tissue tended (P < 0.10) to decrease linearly as dietary CP increased. Similar to VFC, the viscera of calves at the end of the experiment contained a greater proportion of fat and lower proportions of water and protein than viscera of baseline calves, but increasing dietary CP percentage resulted in changes toward the composition of baseline calves.

Together, therefore, our data indicate that calves contained greater amounts of lean tissue and less fat in the final VFC (and, by inference from limited data in Table 8, the whole body) as dietary CP content increased. Likewise, the content of fat in VFC gain decreased and lean components increased. Our results indicate that diet composition, specifically the CP content or perhaps the ratio of protein:nonprotein energy sources (lactose and fat), can markedly impact body composition and composition of gain in young milk-fed calves, even when total energy deposition does not differ. Increasing protein intake seems to cause a repartitioning of energy deposition from fat to protein. Similar results have been shown previously for calves (Donnelly and Hutton, 1976b; Gerrits et al., 1996), lambs (Jagusch et al., 1970; Norton et al., 1970), and pigs (Caperna et al., 1991; Cameron et al., 1999). Equations derived in our experiment to describe responses of body composition and gain components to increasing dietary CP percentage were, in general, similar to those reported by Donnelly and Hutton (1976b).

Energy and Protein Utilization

Crude estimates of energy utilization can be made from our data. At the mean BW of calves across treatments during the balance study (55 kg; Table 4), ME requirements for maintenance (0.100 Mcal/kg of BW^0.75; NRC, 2001) would be 2.02 Mcal. Total ME intake during the balance study averaged 3.60 Mcal/d. Because intake and BW data obtained during the balance study approximate means for the entire 42-d experiment, they can be used to estimate the efficiency of ME use for body energy retention. Using means across dietary treatments for energy gains in VFC (Table 7) and estimates for visceral tissues derived from the incomplete data in Table 8, whole body energy gain was approximately 0.92 Mcal/d. The ME available for tissue deposition was 1.58 Mcal/d (i.e., 3.60 minus 2.02); consequently, efficiency of ME use for tissue energy retention was 0.58 (i.e., 0.92/1.58). This estimate is lower than the value of 0.69 assigned by NRC (2001) but is similar to estimates for calves of this size made from more recent data sets (Bartlett, 2001; Diaz et al., 2001; Tikofsky et al., 2001) and for milk-fed calves from 80 to 240 kg of BW (Gerrits et al., 1996). Maintenance requirements specified by NRC (2001) are lower than earlier estimates (e.g., Roy, 1980); if maintenance is underestimated the apparent efficiency of ME use for growth also would be underestimated.

Rates of protein deposition have the greatest impact on ADG of BW because of the associated water deposited with protein as lean tissue (Roy, 1980). Relationships between digestible CP intake and either protein gain in VFC or protein deposition determined from N retention were studied by regression analysis (Figure 4). A linear relationship (R² = 0.88) existed between whole body protein gain estimated from N retention and total digestible CP intake (Figure 4A). This equation predicted that additional digestible CP intake was used for retained CP with an efficiency of 66%. Also of interest is the intercept, which predicts that at zero digestible CP intake, endogenous loss of CP would be 23.4 g/d (or 3.7 g/d of N). At the average BW of all calves during the study (55 kg), endogenous urinary N loss predicted by the equation (endogenous urinary N, g/d = 0.2 × kg BW^0.75) of the ARC (1965), as adopted by NRC (2001), would be 4.0 g/d.

Figure 4.

Relationships between digestible CP intake and whole-body CP retention (Panel A) and visceral-free carcass (VFC) protein gain (Panel B) plotted for individual calves. Panel A. Whole-body CP retention was calculated as N retention × 6.25. The line shows the regression: y = 0.659 (± 0.044) x − 23.36 (± 6.46), R² = 0.885. Panel B. The line shows the regression: y = 0.515 (± 0.064) x + 0.13 (± 9.28), R² = 0.680.

The linear relationship between VFC protein gain and digestible CP intake (Figure 4B) did not account for as much variance (R² = 0.68) but was still highly significant. This equation predicted that additional digestible CP was used for VFC deposition with an efficiency of 52%. The ratio of the coefficients of these two equations is 0.79 (i.e., 52/66). Although, as discussed earlier, balance trials overestimate N deposition as protein (e.g., Gerrits et al., 1996; Diaz et al., 2001), this ratio is close to the percentage of live BW represented by VFC (73.2%; Table 6). Assuming that the visceral tissues accrete protein proportionally to VFC, as suggested by data in Table 8 and as shown in our more recent studies (Bartlett, 2001), the rate of whole body protein deposition (i.e., VFC plus viscera) would be closer to, but still lower than, that predicted by the N balance determination. This difference again reflects the overestimation of body N retention by balance studies.

We detected no evidence of curvilinearity in response of protein deposition to increasing digestible CP intake, likely because our highest dietary CP content was below the breakpoint where protein deposition would be maximized. Donnelly and Hutton (1976a) found that BV of milk protein in milk replacers decreased markedly between CP contents of 29.6 and 31.5%. In calves of similar age and BW as those used in our study, rates of protein accretion were maximal and body fat content minimal when milk replacer contained 29.6% CP (Donnelly and Hutton, 1976a,b). Gerrits et al. (1996) found that protein deposition responded linearly as digestible CP intakes increased to 600 g/d in calves between 80 and 160 kg of BW; however, a plateau of protein deposition was reached with greater digestible CP intakes in older calves (160 to 240 kg).

Marginal efficiencies of protein deposition, or the rate at which additional digestible CP is deposited in tissue, are determined from the slopes of the lines in Figure 4. Rates of either of these determinations (66 and 52%) are higher than those found by Gerrits et al. (1996) for heavier calves (30%) and those measured by Donnelly and Hutton (1976a) for Friesian calves of similar age and BW (45%). In both of these studies, skim milk protein and casein were the primary protein sources. Our values obtained using whey protein-based milk replacers are similar to the previous values for N retained as a percentage of N absorbed of 74% reported by Terosky et al. (1997) and 63% reported by Diaz et al. (2001) for milk replacers based entirely on whey proteins. In the latter experiment, however, only one CP content (30%) was evaluated at three intakes. Our data support the findings of the previous studies indicating that whey protein is used with high efficiency for protein deposition in young dairy calves.

Young milk-fed ruminants deposit more of the dietary energy as fat when digestible protein supply limits rates of tissue protein deposition (Jagusch et al., 1970; Norton et al., 1970; Donnelly and Hutton, 1976a,b). Using equations in NRC (2001), calves at the start of the study (45 kg of BW) fed reconstituted milk replacer (4.56 Mcal of ME/kg of DM) at 12% of BW would have an energy-allowable growth rate of about 0.48 kg/d. The NRC (2001) predicts a requirement for apparently digestible protein (ADP) of 141 g to meet that rate of gain, which would be equivalent to 152 g/d of CP from milk proteins (NRC, 2001). This amount of CP would necessitate a milk replacer containing 22.5% CP on a DM basis to match ADP and ME requirements. Consequently, calves fed milk replacers containing 16.1 or 18.5% CP consumed insufficient protein and deposited greater amounts of dietary energy as fat.

Using the NRC (2001) assumptions, most of which were derived earlier by Davis and Drackley (1998) from several experimental data sets, the calculated ADP requirements based on measured growth rates and DMI in our experiment were 121, 138, 168, and 176 g/d of ADP when calves were fed 16.1, 18.5, 22.9, and 25.8% CP, respectively. Calves actually consumed 105, 123, 159, and 186 g of digestible CP on average; therefore, the NRC (2001) equation seemed to overpredict ADP requirements at lower dietary CP contents and underpredict them at the highest CP content. Despite the fact that predicted supply of ADP at the start of the study should have been adequate for maximal growth in calves fed milk replacers containing either 22.9 or 25.8% CP according to NRC (2001) assumptions, ADG, gain:feed, N retention, and protein deposition in VFC increased linearly, and fat content of VFC and visceral tissues decreased linearly over this range of dietary CP contents. Our results suggest either 1) that assumptions inherent in the NRC (2001) equations are not correct for calves under these circumstances, or 2) that additional CP stimulates protein deposition in some fashion beyond just provision of AA for protein synthesis.

In support of the latter possibility, Gerrits et al. (1998) showed that increasing digestible CP intake, but not increased protein-free energy intake, stimulated increased concentrations of IGF-1 in plasma of preruminant calves between 80 and 240 kg of BW. Likewise, our own recent evidence (Bartlett, 2001) showed that IGF-1 increased, as did lean tissue growth, in response to increased dietary CP content at the same ME intake in calves similar to those in the present study. Direct evidence for increased IGF-1 synthesis and increased skeletal muscle protein synthesis in response to increased AA supply (via abomasal casein infusion) at constant energy intake also has been reported in older steers (Maloney et al., 1998). Increases in circulating IGF-I in response to increased protein or AA intake are believed to result from increased transcription of IGF-I messenger RNA in liver, as demonstrated in sheep by Pell et al. (1993).

Implications

Our research showed that increasing crude protein in milk replacers from 16 to 26% and the corresponding increase in protein:energy ratio linearly increased growth rates of calves, even though total energy deposition remained unchanged. Increased average daily gains were shown to reflect increases in structural tissues and lean tissue deposition, not additional fat deposition. As growth rate increased, calves also grew larger more efficiently and retained a greater proportion of ingested protein at the same metabolizable energy intake. Our results indicate that manipulating milk replacer composition can markedly alter the characteristics of body growth in young dairy calves. Linear increases of lean tissue deposition with increasing dietary crude protein are not well predicted by current (NRC, 2001) guidelines. Our data on digestion and utilization of milk-based diets will allow further refinements in equations to predict growth and nutrient requirements of preruminant calves.