Nutrient Digestion and Lactation Performance of Dairy Cows Fed Combinations of Prilled Fat and Canola Oil' 1.C. JENKINS and 8. F. JENNY* Department of Animal, Dairy and Veterinary Sciences Clemson University C l e m , sc 29634 ABSTRACT
Four combinations of prilled fat and canola oil were fed to 10 lactating Holstein cows in a replicated 5 x 5 Latin square to determine whether mixing plant oil with a rumen inert fat had additive effects on digestive and lactation responses. Five diets of conmtrate and corn silage (l:l, DM basis) contained either no added fat (control) or 5% fat comprising 100,67,33, or 0% prilled fat and the remainder canola oil. The fat supplement containing 100% prilled fat appeared to be rum&-inert because it caused no changes in xuminal VFA concentration, acetate to propionate ratio, or total tract fiber digestion. F'rilled fat increased milk production, FCM, and milk fat percentage but decreased milk protein percentage, including casein content. Increasing canola oil in the fat supplement caused linear declines in ruminal VFA, acetate to propionate ratio, and milk production. Milk production efficiency (weight FWweight DMI) exceeded the control diet when fat supplements contained 100 or 67% prilled fat but dropped below control for 33 and 0% prilled fat. This study demonstrates additive effects of combining canola oil with hydrogenated, prilled fat on ruminal fermentation but nonadditive effects on milk production efficiency and milk composition. At low levels of supple-
mentation, plant oils, such as the canola oil used in this study, can inhibit ruminal fermentation but still maintain milk production efficiency. (Key words: fat combinations, lactation performance, digestion, dairy)
Abbreviation key: A:P = acetate to propionate ratio. INTRODUCTION
Lipids of plant and animal origin often interfere with microbial fermentation in the Nmen, thus limiting their level of inclusion in ruminant diets. Large reductions in fiber and energy digestibility often result from inhibition of fermentation. To counter this problem, as well as to improve handling, lipids commonly are modified by a number of processes (such as hydrogenation, conversion to calcium salts, prilling. and encapsulation) that minimize or eliminate these changes in fermentation. Modified fats generally are referred to as rumeninert or rumen-bypass fats to signify the relative absence of antimicrobial effects in the rumen. Lipid supplements combining rumen-inert fats with plant oils may reduce the cost of feeding fat without greatly inhibiting ruminal fermentation. Some change in ruminal fermentation by plant oils may even be beneficial. Fats and oils added to ruminant diets were reported to increase dietary N escape to the hindgut, as well as to increase efficiency of microbial protein synthesis (3, 11, 14). Starch digestion also could be shifted to the hindgut, causing increased glucose absorption and less Received August 5, 1991. fermentation losses (19). Accepted November 1, 1991. There is some evidence that lipid sources lTechoical Contribution Number 3205 of the South Carolina Agricultnral Experiment Statim Partidy sap may act synergistically when combined, giving ported by a grant from CBP Reso-. Inc., GnensboTo, nonadditive metabolic and production NC. responses. Rapeseed oil combined with tallow 2 M e n t address: Deparimea of airy science, WSin poultry diets had a synergistic effect on diet iaaa State University. Baton Range 70803. 1992 J Dairy Sci 75:79&803
796
797
FAT COMBINATIONS ON LACTATION AND DIGESTION TABLE 1. Fatty acid composition of lipid snpplancnts.
Fatty acid'
canola oil
Rilled fat
140 160 161 18:O 18:1 18:2
.03 5.48 20 3.00 62.16 28.46 .67
3.31 28.70 .79 47.88 1326 5.14 .92
2o:O
'Number of c a r b o n s : ~of double bonds.
metabolizable energy of nearly 20% compared with calculated values (16). Blended animalvegetable fat had fewer negative effects on fermentation in vitro than equal quantities of corn oil or tallow (12). In feeding trials, blended fats had little effect on ruminal fermentation and resembled rumen-inert fats more than their constituent animal fats or vegetable oils (21, 27). This raises doubts that metabolic and production effects of a lipid source can be predicted simply from fatty acid composition. This study was designed to determine whether digestive and milk production responses changed linearly as plant oil (canola)
replad rumen-inert fat (hydrogenated and prilled) in dairy diets. MATERIALS AND METHODS
Fatty acid compositions of the canola oil and prilled fat supplements are shown in Table 1. ?he prilled fat consisted of partially hydrogenated triglycerides containing 8 1% saturated fatty acids. Both fat supplements were mixed with other c o n c e n m ingredients as they were received from the supplier (CBP Resources, Inc., Greensboro,NC). Three concentrates (Table 2) were mixed for the study that included a control with no added fat and two concentrates with 10.9% added fat (DM basis) as either prilled fat or canola oil. The prilled pdt and canola oil concentrates were blended daily (2:l and 1:2), giving two additional levels of each fat suurce. Therefore, prilled fat comprised 100, 67,33, and 0% of the added fat in the four concentrates; the remainder was canola oil, Concentrates were combined with corn silage (1:l. DM basis) and fed as a TMR once daily at 0800 h. Cows were fed individually in a tiestall barn; adequate feed was given to ensure at least 10% weighback. Cows had ad
TABLE 2. Composition of diets Containing combinations of prilled fat and canola oil. DiCts' Item
Concentrate ingredients, % of DM prilled fat Canola oil COm Barley soybean meal Cottonseed hulls Deflorinated phosphate LimeStOtE
C
1009b
67%
33%
0%
0 0 312 23.7 31.8 7.9 2.4 1.2 1.1 .6 .04
10.9 0 18.9 23.3 33.8 7.8 2.4 12 1.1 .6 .04
7.3 3.6 18.9 23.3 33.8 7.8 2.4 1.2 1.1 .6 .04
3.6 7.3 18.9 23.3 33.8 7.8 2.4 1.2 1.1 .6 .04
0 10.9 18.9 23.3 33.8 7.8 2.4 1.2 1.1 .6 .04
salt Didcium phosphate Vitamins A and D2 Composition of total diet, % of DM 15.4 15.9 16.0 16.1 152 CP ADF 252 24.7 24.3 24.7 25.7 Fatty acids 22 55 6.0 6.1 5.7 1.67 1.88 1.88 1.88 1.88 m ~ Mcal/Lg ? kisthecontroldietwithnoaddedfat, N o r m b a s r e f e r t o t h e ~ ~ e o f p r i l l e d f a t i n t h c l i p i d ~ l r m e n t w i t h t h e remainder as canola oil. 2Supplied 220 IU of vitamin A and 44 IU of vitamin D&. htimatcd from NRC (17).
Jomnal of Dairy Science Vol. 75, No. 3, 1992
798
JE"SANDJE"Y
libitum access to water and were moved twice daily (at approximately 0200 and 1400 h) to a parlor for milking. Ten lactating Holstein cows (5 primiparous) averaging 83 DIM were fed the five experimental diets in a replicated 5 x 5 Latin square design. A multiparous and a primiparous cow were paired for each of the 2 1 d periods. Each cow was orally given a gelatin capsule (by bolus gun) at 0800 and 1700 h during the last 10 d of each period. Capsules contained 10 g of chromic oxide as an indigestible reference marker. Grab fecal samples were taken from cows the last 3 d of each period in a time sequence so that every 3 h over 24 h was represented (total of eight samples per period). Samples of ruminal contents (via stomach tube) were taken on the last day of each period at 1100 h. Milk production and composition data were taken from twice daily milk samples collected during the last 4 d of each period. The eight milk samples were composited and frozen. Feed intakes are reported as averages of the last 5 d of each period. Feed and fecal samples were dried at 55'C and ground in a Wiley mill (2-mm sieve; Arthur H. Thomas, Philadelphia, PA). They were analyzed for DM at l W C , for ADF (8). Kjeldahl N (1). and fatty acids (26). Chromic oxide content in fecal samples was determined by the pnxedure of Fenton and Fenton (7). Ruminal samples were transported immediately to the laboratory and strained through cheesecloth, and then pH was determined. Five milliliters of strained fluid were mixed with 1 ml of 1.2N H2SO4 and centrifuged at 30,000 x g for 15 min at 4'C. Internal standard (2-ethyl butyric acid) was added to the Supernatant. Volatile fatty acids were determined by GLC using a 183-cm x .32-cm stainless steel column packed with 20% NPGSD% H3PO4 on 60/80 Chromosorb W AW (Supelco, Inc., Bellefonte, PA). Oven temperatures were 125°C for the column and 2 W C for the injector and detector. Nitrogen was the Carrier gas. Thawed milk samples were analyzed for fat by the Babcock method and for total N by Kjeldahl determination (1). Milk N fractions were determined according to Rowland (23). Fat was separated from thawed milk by Centrifugation at 21,000 x g at 4'C for 30 min. Fatty acids were determined by GLC according to Sukhija and Palmquist (26) using a temperaJournal of Dairy Science Vol. 75, No. 3, 1991
ture program for the capillary column from W C (3 min) to 220°C (10 min) at 6"c/min. Helium was the carrier gas. Identity of fatty acids was determined by comparison of retention times with pure reference standards and verified by gas chromatography-mass spectroscopy (Hewlett-Packard 5890A GC with a 5971A mass selective detector, Avondale, PA). Statistical analysis was by least squares ANOVA using the general linear models procedwe of SAS (24). Sources of variation included in the model for the replicated 5 x 5 Latin square included replicate, animal within replicate, period within replicate, and diet. Least significant difference was used to compare the control diet with the 100% prilled fat diet. Response curves for the four levels of prilled fat (control diet not included) were determined by subdividing the diet sum of squares into orthogonal polynomials with linear, quadratic, and cubic components. Significance was declared at P < .05 unless otherwise noted. RESULTS AND DISCUSSION Fermentatton and Dlgestlon
No changes in nuninal VFA, except valerate, were seen when prilled fat was added to the diet, indicating that it was rumen-inert (Table 3). Grummer (10) also reported no effect of prilled fatty acids on ruminal VFA or on nutrient digestion. Inertness of these prilled fats can be largely attributed to high contents of saturated fatty acids that inhibit ruminal fermentation less than unsaturated fatty acids (22). Replacing @ed fat with canola oil in the diet linearly reduced total VFA concentration and acetate to propionate ratio (A:P). Acetate decreased, while all other VFA, except propionate, increased in ruminal samples as prilled fat content declined. As expected, canola oil, similar to other unsaturated plant oils (22), depressed fermentation, providing the basis in this study to determine the effects of replacing rumen-inert fats with antimicrobial oils. Combining prilled fat and canola oil had additive effects on ruminal fermentation, unlike the synergistic response reported previously for blended fats (12). Blended fats typically consist of triglycerides mixed with FFA,
799
FAT COMBINATIONS ON LAnATION AND DIGESTION
TABLE 3. Effect of replacing prilled fat with canoh oil in dairy diets on pH and VFA in ruminal contents. DiCtsl Item
Total WA," mM % of Total VFA Acetate (A)" Ropionate (p) AP 1SObutyratea Butyratea Isovakratea valerateqb
C 92.0 63.6 20.3 3.19 1 .o 11.5 2.1 1.5
10046
67%
33%
0%
SEM
103.6
982
94.5
79.2
5.6
61.4 232 2.70 1.1 10.5 2.3 1.5
58.8 23.3 254 1.3 122 2.7 1.8
58.8 22.6 2.61 1.5 12.5 2.7 1.9
.7 .8 .12
Four combinations of prilled fat and canola oil were fed to 10 lactating Holstein cows in a replicated 5 x 5 Latin square to determine whether mixing plant oil with a rumen inert fat had additive effects on digestive and lactation responses. Five diets of conmtrate and corn silage (l:l, DM basis) contained either no added fat (control) or 5% fat comprising 100,67,33, or 0% prilled fat and the remainder canola oil. The fat supplement containing 100% prilled fat appeared to be rum&-inert because it caused no changes in xuminal VFA concentration, acetate to propionate ratio, or total tract fiber digestion. F'rilled fat increased milk production, FCM, and milk fat percentage but decreased milk protein percentage, including casein content. Increasing canola oil in the fat supplement caused linear declines in ruminal VFA, acetate to propionate ratio, and milk production. Milk production efficiency (weight FWweight DMI) exceeded the control diet when fat supplements contained 100 or 67% prilled fat but dropped below control for 33 and 0% prilled fat. This study demonstrates additive effects of combining canola oil with hydrogenated, prilled fat on ruminal fermentation but nonadditive effects on milk production efficiency and milk composition. At low levels of supple-
mentation, plant oils, such as the canola oil used in this study, can inhibit ruminal fermentation but still maintain milk production efficiency. (Key words: fat combinations, lactation performance, digestion, dairy)
Abbreviation key: A:P = acetate to propionate ratio. INTRODUCTION
Lipids of plant and animal origin often interfere with microbial fermentation in the Nmen, thus limiting their level of inclusion in ruminant diets. Large reductions in fiber and energy digestibility often result from inhibition of fermentation. To counter this problem, as well as to improve handling, lipids commonly are modified by a number of processes (such as hydrogenation, conversion to calcium salts, prilling. and encapsulation) that minimize or eliminate these changes in fermentation. Modified fats generally are referred to as rumeninert or rumen-bypass fats to signify the relative absence of antimicrobial effects in the rumen. Lipid supplements combining rumen-inert fats with plant oils may reduce the cost of feeding fat without greatly inhibiting ruminal fermentation. Some change in ruminal fermentation by plant oils may even be beneficial. Fats and oils added to ruminant diets were reported to increase dietary N escape to the hindgut, as well as to increase efficiency of microbial protein synthesis (3, 11, 14). Starch digestion also could be shifted to the hindgut, causing increased glucose absorption and less Received August 5, 1991. fermentation losses (19). Accepted November 1, 1991. There is some evidence that lipid sources lTechoical Contribution Number 3205 of the South Carolina Agricultnral Experiment Statim Partidy sap may act synergistically when combined, giving ported by a grant from CBP Reso-. Inc., GnensboTo, nonadditive metabolic and production NC. responses. Rapeseed oil combined with tallow 2 M e n t address: Deparimea of airy science, WSin poultry diets had a synergistic effect on diet iaaa State University. Baton Range 70803. 1992 J Dairy Sci 75:79&803
796
797
FAT COMBINATIONS ON LACTATION AND DIGESTION TABLE 1. Fatty acid composition of lipid snpplancnts.
Fatty acid'
canola oil
Rilled fat
140 160 161 18:O 18:1 18:2
.03 5.48 20 3.00 62.16 28.46 .67
3.31 28.70 .79 47.88 1326 5.14 .92
2o:O
'Number of c a r b o n s : ~of double bonds.
metabolizable energy of nearly 20% compared with calculated values (16). Blended animalvegetable fat had fewer negative effects on fermentation in vitro than equal quantities of corn oil or tallow (12). In feeding trials, blended fats had little effect on ruminal fermentation and resembled rumen-inert fats more than their constituent animal fats or vegetable oils (21, 27). This raises doubts that metabolic and production effects of a lipid source can be predicted simply from fatty acid composition. This study was designed to determine whether digestive and milk production responses changed linearly as plant oil (canola)
replad rumen-inert fat (hydrogenated and prilled) in dairy diets. MATERIALS AND METHODS
Fatty acid compositions of the canola oil and prilled fat supplements are shown in Table 1. ?he prilled fat consisted of partially hydrogenated triglycerides containing 8 1% saturated fatty acids. Both fat supplements were mixed with other c o n c e n m ingredients as they were received from the supplier (CBP Resources, Inc., Greensboro,NC). Three concentrates (Table 2) were mixed for the study that included a control with no added fat and two concentrates with 10.9% added fat (DM basis) as either prilled fat or canola oil. The prilled pdt and canola oil concentrates were blended daily (2:l and 1:2), giving two additional levels of each fat suurce. Therefore, prilled fat comprised 100, 67,33, and 0% of the added fat in the four concentrates; the remainder was canola oil, Concentrates were combined with corn silage (1:l. DM basis) and fed as a TMR once daily at 0800 h. Cows were fed individually in a tiestall barn; adequate feed was given to ensure at least 10% weighback. Cows had ad
TABLE 2. Composition of diets Containing combinations of prilled fat and canola oil. DiCts' Item
Concentrate ingredients, % of DM prilled fat Canola oil COm Barley soybean meal Cottonseed hulls Deflorinated phosphate LimeStOtE
C
1009b
67%
33%
0%
0 0 312 23.7 31.8 7.9 2.4 1.2 1.1 .6 .04
10.9 0 18.9 23.3 33.8 7.8 2.4 12 1.1 .6 .04
7.3 3.6 18.9 23.3 33.8 7.8 2.4 1.2 1.1 .6 .04
3.6 7.3 18.9 23.3 33.8 7.8 2.4 1.2 1.1 .6 .04
0 10.9 18.9 23.3 33.8 7.8 2.4 1.2 1.1 .6 .04
salt Didcium phosphate Vitamins A and D2 Composition of total diet, % of DM 15.4 15.9 16.0 16.1 152 CP ADF 252 24.7 24.3 24.7 25.7 Fatty acids 22 55 6.0 6.1 5.7 1.67 1.88 1.88 1.88 1.88 m ~ Mcal/Lg ? kisthecontroldietwithnoaddedfat, N o r m b a s r e f e r t o t h e ~ ~ e o f p r i l l e d f a t i n t h c l i p i d ~ l r m e n t w i t h t h e remainder as canola oil. 2Supplied 220 IU of vitamin A and 44 IU of vitamin D&. htimatcd from NRC (17).
Jomnal of Dairy Science Vol. 75, No. 3, 1992
798
JE"SANDJE"Y
libitum access to water and were moved twice daily (at approximately 0200 and 1400 h) to a parlor for milking. Ten lactating Holstein cows (5 primiparous) averaging 83 DIM were fed the five experimental diets in a replicated 5 x 5 Latin square design. A multiparous and a primiparous cow were paired for each of the 2 1 d periods. Each cow was orally given a gelatin capsule (by bolus gun) at 0800 and 1700 h during the last 10 d of each period. Capsules contained 10 g of chromic oxide as an indigestible reference marker. Grab fecal samples were taken from cows the last 3 d of each period in a time sequence so that every 3 h over 24 h was represented (total of eight samples per period). Samples of ruminal contents (via stomach tube) were taken on the last day of each period at 1100 h. Milk production and composition data were taken from twice daily milk samples collected during the last 4 d of each period. The eight milk samples were composited and frozen. Feed intakes are reported as averages of the last 5 d of each period. Feed and fecal samples were dried at 55'C and ground in a Wiley mill (2-mm sieve; Arthur H. Thomas, Philadelphia, PA). They were analyzed for DM at l W C , for ADF (8). Kjeldahl N (1). and fatty acids (26). Chromic oxide content in fecal samples was determined by the pnxedure of Fenton and Fenton (7). Ruminal samples were transported immediately to the laboratory and strained through cheesecloth, and then pH was determined. Five milliliters of strained fluid were mixed with 1 ml of 1.2N H2SO4 and centrifuged at 30,000 x g for 15 min at 4'C. Internal standard (2-ethyl butyric acid) was added to the Supernatant. Volatile fatty acids were determined by GLC using a 183-cm x .32-cm stainless steel column packed with 20% NPGSD% H3PO4 on 60/80 Chromosorb W AW (Supelco, Inc., Bellefonte, PA). Oven temperatures were 125°C for the column and 2 W C for the injector and detector. Nitrogen was the Carrier gas. Thawed milk samples were analyzed for fat by the Babcock method and for total N by Kjeldahl determination (1). Milk N fractions were determined according to Rowland (23). Fat was separated from thawed milk by Centrifugation at 21,000 x g at 4'C for 30 min. Fatty acids were determined by GLC according to Sukhija and Palmquist (26) using a temperaJournal of Dairy Science Vol. 75, No. 3, 1991
ture program for the capillary column from W C (3 min) to 220°C (10 min) at 6"c/min. Helium was the carrier gas. Identity of fatty acids was determined by comparison of retention times with pure reference standards and verified by gas chromatography-mass spectroscopy (Hewlett-Packard 5890A GC with a 5971A mass selective detector, Avondale, PA). Statistical analysis was by least squares ANOVA using the general linear models procedwe of SAS (24). Sources of variation included in the model for the replicated 5 x 5 Latin square included replicate, animal within replicate, period within replicate, and diet. Least significant difference was used to compare the control diet with the 100% prilled fat diet. Response curves for the four levels of prilled fat (control diet not included) were determined by subdividing the diet sum of squares into orthogonal polynomials with linear, quadratic, and cubic components. Significance was declared at P < .05 unless otherwise noted. RESULTS AND DISCUSSION Fermentatton and Dlgestlon
No changes in nuninal VFA, except valerate, were seen when prilled fat was added to the diet, indicating that it was rumen-inert (Table 3). Grummer (10) also reported no effect of prilled fatty acids on ruminal VFA or on nutrient digestion. Inertness of these prilled fats can be largely attributed to high contents of saturated fatty acids that inhibit ruminal fermentation less than unsaturated fatty acids (22). Replacing @ed fat with canola oil in the diet linearly reduced total VFA concentration and acetate to propionate ratio (A:P). Acetate decreased, while all other VFA, except propionate, increased in ruminal samples as prilled fat content declined. As expected, canola oil, similar to other unsaturated plant oils (22), depressed fermentation, providing the basis in this study to determine the effects of replacing rumen-inert fats with antimicrobial oils. Combining prilled fat and canola oil had additive effects on ruminal fermentation, unlike the synergistic response reported previously for blended fats (12). Blended fats typically consist of triglycerides mixed with FFA,
799
FAT COMBINATIONS ON LAnATION AND DIGESTION
TABLE 3. Effect of replacing prilled fat with canoh oil in dairy diets on pH and VFA in ruminal contents. DiCtsl Item
Total WA," mM % of Total VFA Acetate (A)" Ropionate (p) AP 1SObutyratea Butyratea Isovakratea valerateqb
C 92.0 63.6 20.3 3.19 1 .o 11.5 2.1 1.5
10046
67%
33%
0%
SEM
103.6
982
94.5
79.2
5.6
61.4 232 2.70 1.1 10.5 2.3 1.5
58.8 23.3 254 1.3 122 2.7 1.8
58.8 22.6 2.61 1.5 12.5 2.7 1.9
.7 .8 .12
