animal Animal (2014), 8:7, pp 1191–1200 © The Animal Consortium 2014 doi:10.1017/S1751731114000998 Grazing increases the concentration of CLA in dairy cow milka M. N. Lahlou1, R. Kanneganti1, L. J. Massingill1, G. A. Broderick1, Y. Park2, M. W. Pariza3, J. D. Ferguson4 and Z. Wu4† 1 US Dairy Forage Research Center, USDA-ARS, Madison, WI 53706, USA; 2Department of Food Science, University of Massachusetts, Amherst, MA 01003, USA; Food Research Institute, University of Wisconsin, Madison, WI 53706, USA; 4New Bolton Center, University of Pennsylvania, Kennett Square, PA 19348, USA 3 (Received 18 June 2013; Accepted 17 March 2014; First published online 30 April 2014) An experiment was conducted to examine whether increased CLA in milk of dairy cows fed fresh pasture compared with alfalfa and corn silages was because of ruminal or endogenous synthesis. Eight Holsteins were fed a total mixed ration using alfalfa and corn silages as the forage source in confinement or grazed in a replicated crossover design. The proportion of total fatty acids as CLA (primarily c9, t11-18:2) in g/100 g was 0.44 v. 0.28 in ruminal digesta, 0.89 v. 0.53 in omasal digesta and 0.71 v. 1.06 in milk during confinement feeding and grazing, respectively. Blood plasma CLA was 0.54 v. 1.05 mg/l for the two treatments, respectively. The increased concentration of CLA in milk with grazing likely resulted from increased synthesis through desaturation of t11-18:1 in the mammary gland. Keywords: CLA, pasture, milk fatty acids, dairy cow Implications CLA c9, t11-18:2 has been shown to be beneficial to human health, and its concentration in milk of dairy cows was increased when cows were grazed on fresh pasture compared with receiving conserved feeds. The increase was not explained by more production in the rumen, but likely because of increased synthesis in the mammary gland. Introduction C9, t11-18:2 is the most abundant isomer of CLA in milk of dairy cows, and because of its health benefits to humans, various feeding strategies have been studied for their effect on the content of the fatty acid in milk (Palmquist et al., 2005). Grazing (Kelly et al., 1998; Dhiman et al., 1999; Schroeder et al., 2003) is one of the most effective approaches identified for increasing CLA, compared with feeding stored feeds, the most common practice in the modern dairy industry. Understanding the origin of milk CLA is important, and sources include both ruminal synthesis and endogenous contributions (Palmquist et al., 2005). Ruminal synthesis is a result of partial biohydrogenation of 18:2n-6 and 18:3n-3. a Trade names and the names of commercial companies are used in this report to provide specific information. Mention of a trade name or manufacturer does not constitute a guarantee or warranty of the product by the USDA or an endorsement over products not mentioned. † E-mail: zwu@vet.upenn.edu The biohydrogenation of 18:2n-6 initiates with isomerization of c9, c12-18:2 into c9, t11-18:2, the primary isomer of CLA, as the first intermediate (Kepler et al., 1966). The subsequent step is reduction of the cis-9 double bond, resulting in t11-18:1 as the second intermediate. The final step is another reduction, resulting in 18:0. The rumen bacterium Butyrivibrio fibrisolvens is the primary species that carries out the isomerization and the first reduction step (Kepler et al., 1966). Bacteria that carry out the last reduction are separate species, including some Fusocillus and Ruminococcus strains (Kemp et al., 1975). Formation of intermediates during biohydrogenation of 18:3n-3 is not well understood. The biohydrogenation pathway begins with isomerization of the double bond at carbon 12, forming a conjugated cis-9, trans-11, cis-15 triene (Kim et al., 2009). Hydrogen is subsequently added to the double bonds, likely in the order of carbon positions 9, 15 and 11 (Dawson and Kemp, 1970). Loor et al. (2002) reported distinct increases in t11, c15-18:2 in blood plasma and milk when cows were grazed compared with being fed a total mixed ration. However, no typical CLA would be formed with this 9 > 15 > 11 preference order (Jenkins et al., 2008). Other preference scenarios are 9 > 11 > 15 and 15 > 9 > 11, with the latter being the one that would yield c9, t11-18:2 as an intermediate. However, such a preference order has not been documented (Dawson and Kemp, 1970; Destaillats et al., 2005). Observed increases in milk CLA when cows are grazed have not been well explained by ruminal biohydrogenation because the main fatty acid in grasses and clovers is 18:3n-3 (Palmquist, 1988). 1191 Lahlou, Kanneganti, Massingill, Broderick, Park, Pariza, Ferguson and Wu Table 1 Ingredients and composition of the total mixed ration, supplement and pasture forage Item Ingredient Alfalfa silage Corn silage Shelled corn Soybean meal Expeller soybean meal2 Blood meal CaCO3 CaHPO4 MgO NaHCO3 Salt Trace mineral and vitamin mix3 Chemical composition CP NDF ADF Total mixed ration (g/kg of dry matter) Supplement (g/kg of dry matter)1 Pasture forage (g/kg of dry matter) 190 270 350 40 110 10 5 7 2 6 8 2 – – 825 – 100 20 15 10 5 10 10 5 – – – – – – – – – – – – 175 236 156 149 93 33 180 398 288 1 The supplement was offered to grazing cows at 10 kg/day (dry matter). SoyPlus (West Central Coop; Ralston, IA, USA). Each kilogram contained 0.32 g of Se, 0.43 g of Co, 1.03 g of I, 13.35 g of Cu, 23.99 g of Fe, 51.00 g of Mn, 62.01 g of Zn, 7,000,000 IU of vitamin A, 2,222,000 IU of vitamin D and 17,630 IU of vitamin E. 2 3 Endogenous synthesis is a result of Δ9 desaturation of t11-18:1 (Griinari et al., 2000; Morales et al., 2000). T11-18:1, derived from the rumen during biohydrogenation of 18:2n-6, as discussed above, can be desaturated as a result of the action of Δ9-desaturase (Palmquist et al., 2005; Mosley et al., 2006), forming c9, t11-18:2. The enzyme acts on the bond at carbon 9 in 16:0, 18:0 as well as t11-18:1 in the presence of NADPH, flavin and ferrous ion (Moore and Christie, 1979). The most recent studies have shown that endogenous synthesis contributes a large proportion of total CLA in milk, with estimates including 64 (Griinari et al., 2000), 78 (Corl et al., 2001), 94 (Kay et al., 2004) and 93% to 97% (Piperova et al., 2002; Mohammed et al., 2009), depending on diet composition, oil supplementation and grazing. However, reasons for increased synthesis with regimens that are known to be effective in increasing milk CLA remain unclear. The objective of this study was to determine the concentration of CLA in ruminal and omasal contents and in milk of cows fed conserved forages or pasture to gain insights into the mechanism of observed increases in milk CLA. Material and methods The experimental protocol was approved by the Institutional Animal Care and Use Committee of the College of Agricultural and Life Sciences, University of Wisconsin, Madison, United States of America. Animals and dietary treatments Eight ruminally cannulated multiparous Holsteins were paired based on milk yield and days in milk (ranging from 16 to 290) and utilized in a replicated crossover design in June 1192 to July. Cows within a pair were fed a total mixed ration consisting of alfalfa silage, corn silage and a concentrate mix (Table 1) or grazed for 21 days, and then switched to the other regimen for 21 more days. In each period, the first 14 days served as adaptation and the last 7 days for sample collection. Grazed cows were also offered a supplement mix at 10 kg/day (dry matter basis) using ingredients similar to those included in the concentrate portion of the total mixed ration for the other group of cows. The allowance was used to provide 40% of the total feed intake, assuming that animals would obtain 60% of their intake from the pasture based on previous season estimates (Wu et al., 2001). The use of these values before the commencement of the experiment was necessary, but as reported below, the actual amount of the supplement consumed was less than offered. Cows in the total mixed ration group were housed in tie stalls and offered the diet at 0830 h daily for ad libitum consumption (5% to 10% refusal). Actual amounts of feed offered and refused by individual animals were recorded daily to obtain net intake. Cows in the pasture group grazed continuously except for ∼ 6 h/day when taken to the milking parlor and then tie stalls to receive the supplement. Refusals of the supplement were recorded. The pasture was about 20 ha in size, divided into 0.4 to 0.8 ha paddocks by electric fences. It was routinely fertilized with 56 kg/ha of nitrogen in two applications annually, one in June and one in August. The June application for the year the experiment was conducted fell 2 weeks before the beginning of the trial and the August application occurred after the trial. Determined as described below, the pasture used for this experiment contained grasses (bluegrass 35%, quackgrass 28%, timothy grass 16%, and bromegrass, CLA in milk orchardgrass and unidentified species 21%) and clovers (red clover 70%, white clover 30%) at ∼55 : 45, as well as some dead matter (∼10%, dry matter basis). A new paddock was allocated every 3 days, and the stocking rate was estimated to be ∼6 cows/ha. Rotation length was 28 days, and grazed forages were a result of re-growth, thus similar in phenological stage between the two periods. Water was available all times in paddocks. Sampling and analysis Cows were milked at 0600 and 1700 h, with milk weights recorded each time. Milk samples were collected at milking on days 16 and 20 of each period and analyzed for fat, protein, lactose and solid-not-fat (AgSource Milk Analysis Laboratory, Menomonie, WI, USA). An aliquot of the milk samples was pooled, extracted for fat with hexane and analyzed for fatty acids upon methylation with 1.097 M methanolic HCl (Chin et al., 1992) using a gas chromatograph (Hewlett-Packard model 5890 Series II; HewlettPackard, Brookfield, WI, USA) fitted with a flame ionization detector and a 3396 A auto-integrator. The analysis used a fused silica capillary column (60 m × 0.32 mm i.d., 0.25 μm film thickness; Supelco, Inc., Bellefonte, PA, USA). A mixture of known fatty acids was used as a standard (Nu-Chek Prep, Inc., Elysian, MN, USA). The peaks identified for CLA included a major one for c9, t11-18:2 and a minor one for t10, c1218:2. Because the minor peak was identified for only some samples, the two were combined when identified in calculating the proportion of total CLA. Likewise, other isomers such as t7, c9-18:2 and t8, c10-18:2 were not separated from these peaks, and were included in total CLA. Blood was sampled (10 ml) 3 h after feeding on day 19 in each period from the coccygeal vessels into heparinized vacutainers, and centrifuged at 2200 × g for 15 min. Plasma was kept at − 20°C before analysis for CLA (c9, t11-18:2) (Chin et al., 1992) using HPLC (Beckman 125/166; Beckman Instruments, Inc., Palo Alto, CA, USA). Alfalfa silage, corn silage and refusals were sampled daily and kept frozen to generate weekly composites. Corn, soybean meal and blood meal were sampled weekly. All weekly samples were determined for dry matter by oven-drying at 60°C for 48 h. Refusals were used only for net intake calculations. Pasture forage was sampled before grazing from alternate paddocks for botanical and nutrient (CP, NDF, ADF) composition analysis; a total of four paddocks were sampled in each period. For botanical analysis, herbage from 10 randomly selected strips (0.01 × 1.0 m each) in each of the sampling paddocks was cut at ground level and separated into plant species by hand; the species profile was determined on the dry weight of samples (60°C oven-dried). For nutrient analysis, herbage from a second randomly selected area was cut at 9 cm above the ground to obtain 0.5 kg of wet material. Ten such areas were cut in each of the sampling paddocks. The collected herbage was subsampled, oven-dried (60oC), and pooled within paddock to generate four composites for each period. Ruminal samples were collected every 2 h on day 19 of each period, beginning at 0830 h, until six samples were collected, representing 0, 2, 4, 6, 8 and 10 h after feeding. This schedule did not include night times, and thus could result in biases if changes in rumen contents over time were different between the dietary treatments. Samples were strained through two layers of cheesecloth. The liquid was measured for pH (Corning® pH Meter 360i; Corning, Inc., Corning, NY, USA) and kept at − 20°C until analyzed for NH3-N (Broderick and Kang, 1980) using an auto analyzer (Technicon Auto Analyzer II; Technicon Instruments Corp., Tarrytown, NY, USA) and for volatile fatty acids (Brotz and Schaefer, 1987) using GLC (Varian Vista 6000; Varian Instrument Group, Walnut Creek, CA, USA). The solids were pooled within animal and period and lyophilized. Omasal samples were obtained at 0830 and 1530 h during days 16 and 19 in each period. A one-hand opening and closing flip-top vial (35 ml; www.capitol-dairy-solutions. com) was held in the left hand with the cap closed and inserted through the rumen cannula into the omasum. The cap then was opened using the thumb to allow the vial to be filled with digesta. The vial was then closed and withdrawn. Two vials of samples were obtained at each sampling time. The samples (16 vials) were pooled within cow and period and lyophilized. Dried samples of feeds (including pasture forage), ruminal contents and omasal contents were ground through a Wiley mill using a 1-mm screen (Arthur H. Thomas, Philadelphia, PA, USA). Ground feed samples were analyzed for dry matter (105°C), CP (LECO FP-2000 Nitrogen Analyzer; Leco Instruments, Inc., St. Joseph, MI, USA) and NDF (using heat stable α-amylase and Na2SO3) plus ADF (Robertson and Van Soest, 1981) using the ANKOM200 Fiber Analyzer (ANKOM Technology, Fairport, NY, USA). Chemical analyses were expressed on a dry matter basis (105°C). Nutrient content of the total mixed ration and the supplement was computed from average analyses of the ingredients over the experiment. Ground ruminal and omasal samples were analyzed for fatty-acid concentration and profile (Chin et al., 1992) using the same gas chromatograph used for milk fatty-acid analysis described above. Heptadecanoic acid was used as an internal standard, assuming sample content of this acid was negligible. Peaks for c9, t11 and t10, c12 isomers were detected for most of the digesta samples. Statistical analysis Data were processed for ANOVA using two models by the GLMs procedure of SAS (2000). Model 1 was based on a crossover design and used for all data except for ruminal pH, volatile fatty acids and NH3-N, all of which were analyzed with Model 2 according to a crossover split-plot design. The two models were as follows: Y ¼ μ + R + C ðRÞ + P + T + E (Model 1) Y ¼ μ + R + C ðRÞ + P + T + RT + H + RH + HC ðRÞ + HP + TH + RTH + E ðModel 2Þ where Y is the observation, μ the overall mean, R the replicate, C(R) the cow within replicate, P the period, T the treatment, 1193 Lahlou, Kanneganti, Massingill, Broderick, Park, Pariza, Ferguson and Wu E the error term, RT the interaction between replicate and treatment, H the hour, RH the interaction between replicate and hour, HC(R) the interaction between hour and cow within replicate, HP the interaction between hour and period, TH the interaction between treatment and hour and RTH the interaction among replicate, treatment and hour. In Model 2, replicate, cow within replicate, period and treatment were in the main plot and the remaining terms were in the subplot. Treatment effects were tested by replicate and treatment interactions. Results and discussion One cow in the total mixed ration group during period 1 was off feed before sample collection began and was removed from the trial, leaving seven cows contributing to the data analyzed. In addition, as mentioned above, some fatty acids in some of the milk and digesta samples were not identified, including those of 10 carbons or shorter, t10, c12-18:2 and t11-18:1. We did not think it appropriate or meaningful to assign a ‘0’ to these samples and then conduct a statistical analysis to compare treatment means. Instead, the values for t10, c12-18:2, minor if detected, were combined with c9, t11-18:2 and reported as ‘CLA.’ T11-18:1 was detected in most of the milk and digesta samples and thus reported separately. Lactation performance and ruminal fermentation Mean consumption of the supplement by grazing cows was 8.4 kg/day (dry matter; Table 2). Using the 40 : 60 split for intake from the supplement and pasture and the nutrient content of the feed sources (Table 1), the diet consumed by grazing cows would contain 168, 276 and 186 g/kg CP, NDF and NDF, respectively, compared with 175, 236 and 156 g/kg for the total mixed ration (Table 1). Feed starch was not analyzed, but the ration formulation showed 350 g/kg for the total mixed ration and 620 g/kg for the supplement. These values were relatively high because of the use of high amounts of corn. Likewise, feed ingredients were not analyzed for fatty acids for this experiment, but forages from the same pasture were analyzed in the previous season, and the analysis showed 174 and 393 g/kg of the total fatty acids as 18:2n-6 and 18:3n-3, respectively (unpublished). Using these values and values listed in the NRC guidelines (2001) for Table 2 Feed intake and milk production of cows fed a total mixed ration or grazed Item Total mixed ration Grazing s.e.m. P 20.3 8.41 – – 35.2 32.2 1.138 28.2 0.979 48.5 1.691 85.1 146 28.4 32.0 0.907 27.6 0.776 47.2 1.335 83.1 136 1.4 0.9 0.047 0.4 0.045 0.4 0.052 0.5 28 0.02 0.89 0.02 0.34 0.02 0.05 0.01 0.03 0.79 Feed intake (kg/day) Milk Yield (kg/day) Fat (g/kg) Fat (kg/day) Protein (g/kg) Protein (kg/day) Lactose (g/kg) Lactose (g/day) Solids-not-fat (g/kg) Somatic cell count (1000/ml) 1 Supplement intake only; forage intake from pasture is not included. Table 3 Ruminal pH, NH3-N concentration and volatile fatty-acid profile measured during a 10-h period after feeding in cows fed a total mixed ration or grazed P Item pH NH3-N (mg/dl) Volatile fatty acids (mM) Proportion (mol/100 mol) Acetate Propionate Isobutyrate Butyrate Isovalerate Valerate Acetate : propionate 1194 Total mixed ration Grazing s.e.m. Diet Diet × hour 5.94 7.6 56.9 5.84 12.1 64.4 0.03 0.3 1.6 0.09 0.07 0.01 0.002 0.11 0.22 62.0 23.4 0.84 10.8 1.56 1.41 2.75 63.1 20.7 1.06 11.9 1.68 1.60 3.25 0.9 0.6 0.02 0.2 0.04 0.09 0.11 0.52 0.17 0.05 0.18 0.05 0.50 0.04 1.00 0.99 0.01 0.91 0.02 0.42 0.94 CLA in milk Table 4 Fatty acids in ruminal contents, omasal contents, blood plasma and milk fat of cows fed a total mixed ration or grazed Fatty acid Total mixed ration Grazing s.e.m. P 29.9 33.9 0.7 0.1 0.13 0.53 14.91 0.09 49.13 8.40 4.75 17.07 0.44 3.42 0.11 0.45 14.51 0.10 45.64 9.63 7.07 17.18 0.28 4.94 0.01 0.04 0.21 0.01 1.82 0.93 0.86 1.44 0.04 0.46 0.05 0.22 0.22 0.21 0.23 0.39 0.11 0.67 0.02 0.06 31.4 34.0 1.3 1.9 0.14 0.62 15.73 0.09 48.77 8.96 6.39 16.83 0.89 1.50 0.54 0.09 0.48 14.52 0.09 46.51 9.67 9.32 17.28 0.53 1.47 1.05 0.01 0.03 0.09 0.01 1.67 0.46 0.30 1.18 0.08 0.05 0.16 0.01 0.02 0.01 0.78 0.38 0.32 0.01 0.80 0.03 0.66 0.07 0.80 0.55 0.15 2.05 3.38 12.49 36.29 1.43 14.09 2.06 21.66 4.37 0.71 0.92 0.24 0.84 0.61 0.14 1.74 2.54 9.50 29.18 1.76 14.96 3.57 29.51 4.69 1.06 1.12 0.10 0.07 0.03 0.04 0.18 0.18 0.55 0.90 0.06 0.65 0.27 1.02 0.15 0.07 0.03 0.06 0.68 0.17 0.80 0.28 0.02 0.01 0.01 0.01 0.38 0.01 0.01 0.17 0.02 0.01 0.16 Ruminal contents Total fatty acids (g/kg of dry matter) Fatty acid composition (g/100 g of total fatty acids) 12:0 14:0 16:0 c9-16:1 18:0 t11-18:1 c9-18:1 18:2n-6 CLA1 18:3n-3, n-6 Omasal contents Total fatty acids (g/kg of dry matter) Fatty acid composition (g/100 g of total fatty acids) 12:0 14:0 16:0 c9-16:1 18:0 t11-18:1 c9-18:1 18:2n-6 CLA1 18:3n-3, n-6 Blood plasma CLA (mg/l)1 Milk fatty acid composition (g/100 g of total fatty acids) 4:0 6:0 8:0 10:0 12:0 14:0 16:0 c9-16:1 18:0 t11-18:1 c9-18:1 18:2n-6 CLA1 18:3n-3, n-6 20:4n-6 1 Primarily c9, t11-18:2, including t10, c12-18:2, if any. the other feed ingredients, as well as the 40 : 60 split for supplement and pasture intake, the total mixed ration and the diet consumed during grazing were calculated to contain 33 and 35 g/kg long-chain fatty acids, of which 48.1% and 32.4% were 18:2n-6 and 9.4% and 24.7% were 18:3n-3, respectively. The estimated intake was 322 and 235 g/day for 18:2n-6 and 63 and 179 g/day for 18:3n-3, totaling 385 and 404 g/day, for the two regimens, respectively. These estimates were consistent with observations that the predominant fatty acid is 18:2n-6 in cereal grains and 18:3n-3 in grasses and clovers. The estimates also suggest that, because both fatty acids are extensively biohydrogenated, combined outflow of these fatty acids from the rumen would be similar between the diets. Milk yield was 6.8 kg/day lower (P < 0.05) for cows grazed than fed the total mixed ration (Table 2). The concentrations of milk fat were similar between the feeding systems, and both were low, reflecting the estimated high dietary starch content. The solids-not-fat proportion of milk of grazing cows was reduced compared with those on the total mixed ration (P < 0.05), largely because of a lower (P < 0.05) lactose concentration. Yield of milk protein also was lower 1195 Lahlou, Kanneganti, Massingill, Broderick, Park, Pariza, Ferguson and Wu CLA Total fatty-acid concentration in ruminal digesta was higher (P < 0.01) for cows grazed than fed the total mixed ration (Table 4). The concentration for c9-18:1 tended to be higher (P < 0.01) in ruminal digesta of grazed cows than for those fed the total mixed ration. Also, CLA concentration in ruminal digesta was lower (P < 0.05) for grazing cows. Some small decreases in 12:0, 14:0 and 16:0 were observed (P < 0.10) in omasal digesta during grazing relative to total mixed ration feeding. However, the concentration of c9-18:1 increased (P < 0.01). The concentration of CLA in blood plasma was higher (P < 0.10) for grazing than for the total mixed ration. The CLA proportion of milk fatty acids was ∼50% higher (P < 0.05) during grazing than during total mixed ration feeding (1.06 v. 0.71 g/100 g). The proportions for 4:0, 6:0 and 8:0 in milk fat are listed in Table 4, but because not all of them were detected in some of the samples, as indicated above, the values may have been underestimated and could be biased. Low values could be a result of loss of methyl esters during analysis or extremely low concentrations that were below the sensitivity of the flame ionization detector. The proportions for 12:0, 14:0 and 16:0 were lower (P < 0.05) for grazing cows than for those fed the total mixed ration. In contrast, c9-16:1, t11-18:1, c1118:1 and 18:3n-3 and n-6 were present in higher (P < 0.05) concentrations, along with CLA, in milk fat of grazing cows. The increase in the concentration of CLA in milk during grazing compared with feeding a total mixed ration observed 1196 in this experiment was in accordance with previous reports (Kelly et al., 1998; Dhiman et al., 1999; Loor et al., 2002). These studies showed that grazing can effectively increase milk CLA, and the CLA concentration is somewhat proportional to the contribution of pasture to the total feed intake. The origin of milk CLA has been ascribed to both ruminal and endogenous syntheses. Rumen-based theories concerning the mechanism of increased milk CLA synthesis when cows are grazed include: (1) reduced digesta retention in the rumen, which would increase the passage of biohydrogenation intermediates from the rumen to the small intestine (Kay et al., 2004); (2) rapid release of fatty acids from grazed herbage in the rumen, which would result in inhibited biohydrogenation at the latter step, thus causing accumulation of intermediates formed in the rumen (Noble et al., 1974; Kim et al., 2009); (3) altered microbial populations and associated biohydrogenation pathways (Polan et al., 1964) and (4) decreased ruminal pH (as shown in Table 3), which would result in shifts in bacterial populations (Kim et al., 2002; Piperova et al., 2002; Palmquist et al., 2005). None of these hypotheses is relevant for the present study because grazing cows had ruminal and omasal CLA concentrations that were only about 60% of the cows fed the total mixed ration (Table 4), despite their higher milk CLA concentrations. In addition, a plot (Figure 1) using cows from both treatments reveals a trend for a negative relationship between the proportion of CLA in milk fat and the proportion in omasal fat; the CLA in milk would include isomers that are usually present in higher amounts in milk than in the rumen (such as t7, c9-18:2). These data suggest that increased milk CLA during grazing observed in this experiment was not a result of more production in the rumen. The observation is consistent with the analysis of Mohammed et al. (2009) and Halmemies-Beauchet-Filleau et al. (2013a and 2013b) for the effect of grazing on milk CLA compared with feeding chopped grass or hay. Discussion should be made on several other factors that could be related to increased milk CLA in this study. First, while cows usually lose BW after taken onto a pasture, the data on milk fatty acids for this short-term experiment might not have been confounded by adipose mobilization, as suggested by the similar low proportions of 18:0 in milk, a pre-formed, major fatty acid in the adipose tissue, for both 1.4 Conjugated linoleic acid in milk (g/100 g of fatty acid) (P < 0.05) when cows were grazed than fed the total mixed ration. A balance calculation showed that cows would consume 7 kg/day forage from the pasture, in addition to the 8.4 kg/day from the supplement, to produce 28.4 kg/day milk when their energy balance was zero and protein balance was positive (180 g/day). The intake from the pasture would be even lower if cows were in fact losing BW, which was likely the case based on visual observations (cows were not weighed). In contrast, cows fed the total mixed ration had an energy balance of 21 MJ/day and a protein balance of 400 g/day with 35.2 kg/day milk. These estimates suggest that the intake from the pasture was not sufficient to support a higher milk yield above the energy maintenance requirement that usually substantially increases because of grazing activities (NRC, 2001). Milk decreases are a common observation when cows are changed from a total mixed ration to grazing, often accompanied with BW losses, and the change is considered to result from insufficient intake from pasture because of small bite sizes (Wu et al., 2001). Mean ruminal pH, obtained from measurements taken during a 10-h period after feeding, was low for both treatments (5.84 and 5.94 for grazing and total mixed ration, respectively, P < 0.10; Table 3), reflecting the high starch content of the diets. The NH3-N and total volatile fatty-acid concentrations were higher (P < 0.10) for grazing, reflecting more rapid fermentation and degradation of protein for fresh herbage than conserved forages. Isobutyrate and isovalerate were slightly higher (P < 0.05) for grazing cows, as was the acetate to propionate ratio. 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Conjugated linoleic acid in omasal digesta (g/100 g of fatty acid) Figure 1 Concentration of CLA (primarily c9, t11-18:2, including t10, c12-18:2, if any) in milk and omasal digesta when cows were fed conserved forages (light markers) or grazed (dark markers). 15.0 g/100 g of fatty acid g/100 g of fatty acid CLA in milk 10.0 5.0 0.0 15.0 10.0 5.0 0.0 14:0 40.0 g/100 g of fatty acid g/100 g of fatty acid 12:0 30.0 20.0 10.0 0.0 60.0 40.0 20.0 0.0 18:0 15.0 g/100 g of fatty acid g/100 g of fatty acid 16:0 10.0 5.0 0.0 12.0 9.0 6.0 3.0 0.0 Trans 18:1n-7 20.0 g/100 g of fatty acid g/100 g of fatty acid Cis 18:1n-9 16.0 12.0 8.0 4.0 0.0 1.5 1.0 0.5 0.0 18:2+18:3 Omasal CLA Milk Omasal Milk Figure 2 Omasal and milk fatty-acid proportions when cows were fed conserved forages (light bars) or grazed (dark bars). feeding systems (Figure 2). These proportions were also lower than those in the omasal contents, contrasting to those for 12:0, 14:0 and 16:0 that can be synthesized in the mammary gland. On the other hand, milk c9-18:1 was higher for grazing, which would suggest increased adipose contribution, but its omasal concentration was also higher. Second, milk CLA was reported to increase when dietary starch was increased from 150 to 250 g/kg (Cabrita et al., 2007), but decreased when corn, a starchy feed source, was increased from 365 to 588 g/kg of the diet (Griinari et al., 1998), or remained unchanged when corn was replaced by citrus pulp, a non-starch feed source (Solomon et al., 2000). Regardless of the inconsistency of reports in the literature, changes in milk CLA observed in these studies were extremely small, especially compared with those caused by oil supplementation. Dietary starch in the present study would be within the range of the amounts used in the above-cited studies. Furthermore, the amounts would be similar between the treatments based on formulation and milk fat concentrations. It was unlikely that dietary starch caused a bias for the effect of pasture v. conserved feeds on milk CLA. Lastly, red clover contains polyphenol oxidase, which may have a protective effect on lipids from ruminal metabolism, thus increasing the transfer of polyunsaturated fatty acids from the forge to milk (Lee et al., 2009). However, substituting red clover for grasses in the diet had little effect on milk CLA (Lee et al., 2009) or the formation of c9, t11-18:2 or t11-18:1 in the rumen (Halmemies-Beauchet-Filleau et al., 2013c), although flows of 18:2n-6 and 18:3n-3 at the omasum showed some increases (Halmemies-Beauchet-Filleau et al., 2013c). From results of the present study, endogenous synthesis was more likely responsible for increased milk CLA when cows were fed pasture compared with conserved feeds. The proportion of fatty acids as CLA increased in milk compared with omasal contents for both regimens, as did the proportions for 12:0, 14:0 and 16:0 (Figure 2), all of which are known to be synthesized entirely or partially in the mammary gland. In addition, CLA was the only fatty-acid group that had a lower concentration in omasal contents but a higher concentration in milk during grazing than during total mixed ration feeding. Lastly, the proportion of t11-18:1 was lower in milk than in the digesta (Figure 2). While this was the case for both treatments, the observation is consistent with the understanding that CLA can be formed from t11-18:1 in 1197 1198 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 Omasal trans 18:1(n-7) (g/100 g of fatty acid) Conjugated linoleic acid in milk (g/100 g of fatty acid) the mammary gland. More directly, calculated productto-substrate ratios for Δ9-desaturase (Perfield et al., 2002) for grazed cows were 0.06, 1.97 and 0.30 for c9-16:1/16:0, c9-18:1/18:0 and c9, t11-18:2/t11-18:1 in milk, respectively, compared with 0.01, 0.20 and 0.05 in omasal contents; the ratio for the c9, t11-18:2/t11-18:1 pair was sixfold higher in milk than in omasal contents, and the difference was 1.5 times that when cows were fed the total mixed ration. Data from all these lines suggest that CLA in milk was synthesized to a large extent at a post-absorptive stage and the synthesis increased when cows were grazed. The results are consistent with the report of Yang et al. (1999) that the activity of Δ9-desaturase was higher in grazing beef cattle than in feedlot cattle, and of Mohammed et al. (2009) that ruminal concentration of CLA has little to do with its final concentration in milk because of the involvement of de novo synthesis. There are several possible reasons for endogenous synthesis of CLA to increase when cows are grazed. First, greater formation of t11-18:1 in the rumen could lead to more c9, t11-18:2 synthesized in the mammary gland through increased substrate availability (Halmemies-Beauchet-Filleau et al., 2013a and 2013b). This mechanism may also offer some explanations for increased production of CLA in milk associated with 18:3n-3, whose classical biohydrogenation pathway does not involve formation of c9, t11-18:2, as discussed above. Mohammed et al. (2009) also reported that increased CLA cannot be fully explained by the outflow of 18:2n-6 and 18:3n-3 from the rumen and suggested that the effect of these fatty acids on milk CLA is primarily through formation of t11-18:1 during their biohydrogenation. However, no relationship was shown between omasal t11-18:1 proportion and milk CLA proportion in this experiment (Figure 3). Piperova et al. (2002) also reported that duodenal flow of t11-18:1 was larger than milk CLA output. Included in Figure 3 also is a plot of milk CLA against omasal 18:2n-6 and 18:3n-3 concentrations, a measure that reflects the intake, and more importantly, the metabolism of these fatty acids in the rumen through biohydrogenation. The plot shows no relationships between omasal and milk concentrations of these fatty acids, consistent with the report of Mohammed et al. (2009). Second, the availability of the essential cofactors involved in the desaturase activity could increase when the forage source was grazed pasture. The third possibility is related to milk fat liquidity. According to Moore and Christie (1979), a liquid form is necessary for milk fat droplets to move to the surface of the secretory cells to be pinched off during milk fat secretion and for fat globules to be suspended in milk. Normally, fat liquidity is assured by acylation of short-chain fatty acids (low melting point) to the sn-3 position of the glycerol moiety. In the absence of short-chain fatty acids, c9-18:1, also low in melting point (14°C), could be substituted. In the present study, fatty acids of 10:0 to 16:0 carbons decreased from 52.2 to 43.0 g/100 g in milk fat during grazing compared with total mixed ration feeding (Table 4), consistent with the literature (Christie, 1979; Timmen and Patton, 1988; Kelly et al., 1998) that Conjugated linoleic acid in milk (g/100 g of fatty acid) Lahlou, Kanneganti, Massingill, Broderick, Park, Pariza, Ferguson and Wu 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 Omasal 18:2 and 18:3 (g/100 g of fatty acid) Figure 3 Relationship between the concentration of t11-18:1 and polyunsaturated octadecenoic acids (18:2n-6 and 18:3n-3) in omasal fatty acids and the concentration of CLA (primarily c9, t11-18:2, including t10, c12-18:2, if any) in milk fatty acids when cows were fed conserved forages (light markers) or grazed (dark markers). short-and medium-chain fatty acids (C4–16) decrease when cows receive their forage from pasture. In that case, CLA synthesis may increase as a compensatory mechanism to maintain milk fat liquidity, because t11-18:1 (melting point 44oC) does not exist in a liquid form at body temperature and physicochemically resembles 18:0 (melting point 69oC) rather than c9-18:1 (melting point 14oC). While plausible, the liquidity theory is difficult to study. Toral et al. (2013) found in a meta-analysis that the variability of the melting point of milk fatty acids was narrower than the variability of milk fatty-acid concentrations, suggesting that melting point may be involved in the control of milk fatty-acid composition, although the effect might be small. Conclusions It has been amply demonstrated that grazing dairy cows increases the CLA concentration in milk compared with feeding stored feeds, and this was again shown in the current study. In addition, this study showed that CLA accounted for a smaller proportion of long-chain fatty acids in ruminal and omasal digesta in pasture-fed cows than in cows fed stored feeds, thus discounting increased ruminal formation as a reason for increased CLA concentrations in milk. Rather, increased endogenous synthesis via desaturation of t11-18:1 might be the primary mechanism for the increase. Current hypotheses for increased endogenous synthesis include increased substrate availability and compensation for deceased short- and medium-chain fatty acids that are important for milk fat liquidity. CLA in milk Acknowledgments This study was supported in part by a grant from the American Cancer Society, grant number IRG-58-41-10. The authors wish to acknowledge the leadership Dr L. D. Satter played in the project. In addition, J. E. Delahoy helped verify the identification of fatty acids on the gas chromatograph. References Broderick GA and Kang JH 1980. Automated simultaneous determination of ammonia and total amino acids in ruminal fluid and in vitro media. Journal of Dairy Science 63, 64–75. Brotz PG and Schaefer DM 1987. Simultaneous determination of lactic acid and volatile fatty acids in microbial fermentation extracts by gas-liquid chromatography. Journal of Microbiological Methods 6, 139–144. Cabrita ARJ, Bessa RJB, Alves SP, Dewhurst RJ and Fonseca AJM 2007. Effects of dietary protein and starch on intake, milk production, and milk fatty acid profiles of dairy cows fed corn silage-based diets. Journal of Dairy Science 90, 1429–1439. Chin SF, Liu W, Storkson JM, Ha YL and Pariza MW 1992. Dietary sources of conjugated linoleic acid, a newly recognized class of anticarcinogens. Journal of Food Composition and Analysis 5, 185–197. Christie WW 1979. The effect of diet and other factors on the lipid composition of ruminant tissues and milk. Progress in Lipid Research 17, 245–277. Corl BA, Baumgard LH, Dwyer DA, Griinari JM, Philipis BS and Bauman DE 2001. The role of Δ9-desaturase in the production of cis-9, trans-11 CLA. The Journal of Nutritional Biochemistry 12, 661–666. Dawson RMC and Kemp P 1970. Biohydrogenation of dietary fats in ruminants. In Physiology of digestion and metabolism in the ruminant (ed. AT Phillipson), pp. 504–518. Oriel Press, Newcastle-upon-Tyne, England. Destaillats F, Trottier JP, Galvez JMG and Angers P 2005. Analysis of α-linolenic acid biohydrogenation intermediates in milk fat with emphasis on conjugated linolenic acids. Journal of Dairy Science 88, 3231–3239. Dhiman TR, Anada GR, Satter LD and Pariza MW 1999. Conjugated linoleic acid content from cows fed different diets. Journal of Dairy Science 82, 2146–2156. Griinari JM, Dwyer DA, McGuire MA, Bauman DE, Palmquist DL and Nurmela KVV 1998. Trans-octadecenoic acids and milk fat depression in lactating dairy cows. Journal of Dairy Science 81, 1251–1261. Griinari JM, Corl BA, Lacy SH, Chouinard PY, Nurmela KVV and Bauman DE 2000. Conjugated linoleic acid is synthesized endogenously in lactating dairy cows by Δ9-desaturase. Journal of Nutrition 130, 2285–2291. Halmemies-Beauchet-Filleau A, Kairenius P, Ahvenjärvi S, Crosley LK, Muetzel S, Huhtanen P, Vanhatalo A, Toivonen V, Wallace RJ and Shingfield KJ 2013a. Effect of forage conservation method on ruminal lipid metabolism and microbial ecology in lactating cows fed diets containing a 60:40 forage-to-concentrate ratio. Journal of Dairy Science 96, 2428–2447. Halmemies-Beauchet-Filleau A, Kairenius P, Ahvenjärvi S, Toivonen V, Huhtanen P, Vanhatalo A, Givens DI and Shingfield KJ 2013b. Effect of forage conservation method on plasma lipids, mammary lipogenesis, and milk fatty acid composition in lactating cows fed diets containing a 60:40 forage-toconcentrate ratio. Journal of Dairy Science 96, 5267–5289. Halmemies-Beauchet-Filleau A, Vanhatalo A, Toivonen V, Heikkilä T, Lee MRF and Shingfield KJ 2013c. Effect of replacing grass silage with red clover silage on ruminal lipid metabolism in lactating cows fed diets containing a 60:40 forage-to-concentrate ratio. Journal of Dairy Science 96, 5882–5900. Jenkins TC, Wallace RJ, Moate PJ and Mosley EE 2008. Board-invited review: recent advances in biohydrogenation of unsaturated fatty acids within the rumen microbial ecosystem. Journal of Animal Science 86, 397–412. Kay JK, Mackle TR, Auldist MJ, Thompson NA and Bauman DE 2004. Endogenous synthesis of cis-9, trans-11 conjugated linoleic acid in dairy cows fed fresh pasture. Journal of Dairy Science 87, 369–378. Kelly ML, Kolver ES, Bauman DE, VanAmburgh ME and Muller LD 1998. Effect of intake of pasture on concentration of conjugated linoleic acid in milk of lactating cows. Journal of Dairy Science 81, 1630–1636. Kemp P, White RW and Lander DJ 1975. Hydrogenation of unsaturated fatty acids by five bacterial isolates from sheep rumen, including a new species. Journal of General Microbiology 90, 100–114. Kepler CR, Hirons KP, McNeill JJ and Tove SB 1966. Intermediates and products of the biohydrogenation of linoleic acid by Butyrivibrio fibrisolvens. The Journal of Biological Chemistry 241, 1350–1354. Kim EJ, Huws SA, Lee MRF and Scollan ND 2009. Dietary transformation of lipid in the rumen microbial ecosystem. Asia-Australian Journal of Animal Science 22, 1341–1350. Kim YJ, Liu RH, Rychlik JL and Russell JB 2002. The enrichment of a ruminal bacterium (Megasphaera elsdenii YJ-4) that produces the trans-10, cis-12 isomer of conjugated linoleic acid. Journal of Applied Microbiology 92, 976–982. Lee MRF, Theobald VJ, Tweed JKS, Winters AL and Scollan ND 2009. Effect of feeding fresh or conditioned red clover on milk fatty acids and nitrogen utilization in lactating dairy cows. Journal of Dairy Science 92, 1136–1147. Loor JJ, Herbein JH and Polan CE 2002. Trans 18:1 and 18:2 isomers in blood plasma and milk fat of grazing cows fed a grain supplement containing solventextracted or mechanically extracted soybean meal. Journal of Dairy Science 85, 1197–1207. Mohammed R, Stanton CS, Kennelly JJ, Kramer JKG, Mee JF, Glimm DR, O’Donovan M and Murphy JJ 2009. Grazing cows are more efficient than zerograzed and grass silage-fed cows in milk rumenic acid production. Journal of Dairy Science 92, 3874–3893. Moore JH and Christie WW 1979. Lipid metabolism in the mammary gland of ruminant animals. Progress in Lipid Research 17, 347–395. Morales MS, Palmquist DL and Weiss WP 2000. Effects of fat source and copper on unsaturation of blood and milk triacylglycerol fatty acids in Holstein and Jersey cows. Journal of Dairy Science 83, 2105–2111. Mosley EE, Shafii B, Moate PJ and McGuire MA 2006. Cis-9, trans-11 conjugated linoleic acid is synthesized directly from vaccenic acid in lactating dairy cattle. Journal of Nutrition 136, 570–575. Noble RC, Moore JH and Harfoot CG 1974. Observations on the pattern on biohydrogenation of esterified and unesterified linoleic acid in the rumen. British Journal of Nutrition 31, 99–108. NRC 2001. Nutrient requirements of dairy cattle, 7th edition. National Academy Press, Washington, DC, USA. Palmquist DL 1988. The feeding value of fats. In Feed science (ed. ER Orskov), World Animal Science B4, pp. 293–311. Elsevier Science Publishers, Amsterdam, The Netherlands. Palmquist DL, Lock AL, Shingfield KJ and Bauman DE 2005. Biosynthesis of conjugated linoleic acid in ruminants and humans. Advances in Food Nutrition Research 50, 179–217. Perfield JW II, Bernal-Santos G, Overton TR and Bauman DE 2002. Effects of dietary supplementation of rumen-protected conjugated linoleic acid (CLA) in dairy cows during established lactation. Journal of Dairy Science 85, 2609–2617. Piperova LS, Sampugna J, Teter BB, Kalscheur KF, Yurawecz MP, Ku Y, Morehouse KM and Erdman RA 2002. Duodenal and milk trans octadecenoic acid and conjugated linoleic acid (CLA) isomers indicate that postabsorptive synthesis is the predominant source of cis-9-containing CLA in lactating dairy cows. Journal of Nutrition 132, 1235–1241. Polan CE, McNell JJ and Tove SB 1964. Biohydrogenation of unsaturated fatty acids by rumen bacteria. Journal of Bacteriology 88, 1056–1064. Robertson JB and Van Soest PJ 1981. The detergent system of analysis. In The analysis of dietary fibre in food (ed. WPT Kames and O Theander), Chapter 9, pp. 123–158. Marcel Dekker Inc., New York, NY, USA. SAS 2000. SAS® user’s guide, statistics, version 8. SAS Institute, Inc., Cary, NC, USA. Schroeder GF, Delahoy JE, Vidaurreta I, Bargo F, Gagliostro GA and Muller LD 2003. Milk fatty acid composition of cows fed a total mixed ration or pasture plus concentrates replacing corn with fat. Journal of Dairy Science 86, 3237–3248. Solomon R, Chase LE, Ben-Ghedalia D and Bauman DE 2000. The effect of nonstructural carbohydrate and addition of full fat extruded soybeans on the concentration of conjugated linoleic acid in the milk fat of dairy cows. Journal of Dairy Science 83, 1322–1329. 1199 Lahlou, Kanneganti, Massingill, Broderick, Park, Pariza, Ferguson and Wu Timmen H and Patton S 1988. Milk fat globules: fatty acid composition, size and in vivo regulation of fat liquidity. Lipids 23, 685–689. Toral PG, Bernard L, Chilliard Y and Glasser F 2013. Short communication: dietinduced variations in milk fatty acid composition have minor effects on the estimated melting point of milk fat in cows, goats, and ewes: insights from a meta-analysis. Journal of Dairy Science 96, 1232–1236. 1200 Wu Z, Kanneganti VR, Massingill LJ, Wiltbank MC, Walgenbach RP and Satter LD 2001. Milk production of fall-calving dairy cows during summer grazing of grass pasture or grass-clover pastures. Journal of Dairy Science 84, 1166–1173. Yang A, Larsen TW, Smith SB and Tume RK 1999. Δ-9 desaturase activity in bovine subcutaneous adipose tissue of different fatty acid composition. Lipids 34, 971–997.