Supplementing different ratios of short- and medium-chain fatty acids to long-chain fatty acids in dairy cows: Changes of milk fat production and milk fatty acids composition

The present study was performed at Beijing Cangdafu Dairy Farm (Beijing, China). Animals were cared for in accordance with the guidelines established by the Institute of Animal Science, Chinese Academy of Agriculture Science, Beijing, China.

Thirty-six lactating Holstein cows were blocked by average daily milk yield, DIM, and parity (183 ± 46, 21 ± 3.37 kg/d, and 1.83 ± 1.25, respectively) and then randomly assigned to 1 of 3 treatments, each treatment consisting of 12 cows. Treatment supplements were added to diets as follows: (1) 80 g/d of SMCFA mixture and 320 g/d of LCFA mixture (20SM80L; ratio of SMCFA to LCFA was 20:80); (2) 400 g/d of butterfat (40SM60L: ratio of SMCFA to LCFA was 40:60); or (3) 240 g/d of SMCFA mixture and 160 g/d of LCFA mixture (60SM40L; ratio of SMCFA to LCFA was 60:40). The FA compositions of the 3 treatment supplements are presented in Table 1. Except for butyrate in the SMCFA and trans FA in the LCFA, the FA compositions of the SMCFA mixture and LCFA mixture were similar to the FA derived from de novo synthesis and preformed FA in butterfat (Bright Dairy Co., Shanghai, China), which contain all FA with composition identical to milk. Because the amount of trans-10,cis-12 conjugated linoleic acid (CLA) in butterfat is less than the effective dose for milk fat depression (), the inhibitory effect of trans-10,cis-12 CLA in the butterfat was neglected. The SMCFA mixture included 6% caproic acid (C6:0), 4% caprylic acid (C8:0), 9% capric acid (C10:0), 10% lauric acid (C12:0), 32% myristic acid (C14:0) and 39% palmitic acid (C16:0; Guanhua Co. Ltd., Nanjing, China). The LCFA mixture included 59% cocoa butter (major FA by weight percentage were C12:0, 1.25%; C16:0, 24.64%; C18:0, 35.55%; cis-9 C18:1, 32.91%; C18:2, 3.45%; Linzhishanyang Co. Ltd., Jiangsu, China), 16% olive oil (major FA by weight percentage were: C16:0, 9.43%; C18:0, 2.86%; cis-9 C18:1, 72.81%; C18:2, 12.20%; Shijikangxin Ltd., Beijing, China), and 25% palm oil (major FA by weight percentage were: C16:0, 47.12%; C18:0, 4.69%; cis-9 C18:1, 37.39%; C18:2, 9.67%; Yihai Ltd., Shanghai, China).

Because about half of the C16:0 is formed de novo (the other half originates from feed and body adipose tissues; ), only 50% of the C16:0 found in the butterfat was included in the both SMCFA mixture and LCFA mixture. The SMCFA mixture provided 50% of the C16:0 and equivalent proportion (by weight) of C6:0, C8:0, C10:0, C12:0, and C14:0 FA as those found in the butterfat. The LCFA mixture provided 50% of the C16:0 and equivalent proportion (by weight) of C18 FA as found in the butterfat. The SMCFA and LCFA provided by the butterfat were approximately 40 and 60% (by weight) of the total FA, respectively. The butterfat supplemented in 40SM60L treatment was prepared from commercial butter by melting at 37°C and separating the oil layer when it became solid.

The total duration of the experiment was a single period of 9 wk, including a pretrial period of 1 wk. The FA supplements and butterfat were mixed thoroughly with 800 g Molasses Soyhulls (contains 40% soy molasses and 60% soy hulls; Jinhai Foods Inc., Qinhuangdao, China), and then mixed with 1 kg (wet weight) diet, and consumed by cows of each treatment every morning before feeding during the 8-wk experimental period. The diet was formulated to meet the nutrient requirements () of cows weighing 600 kg, averaging 150 DIM, and producing 30 kg of milk (3.5% FCM/d). The ingredient and chemical composition of the basal diet are shown in Table 2. Cows were housed in a tie-stall barn and fed individually. Diets were fed as a TMR 3 times daily.

The feed offered and feed refused were recorded for individual cows daily. Orts were restricted to 5 to 10% of intake on an as-fed basis. The offered TMR were sampled daily and frozen at −20°C. Samples of orts were collected from each cow, composited for each treatment, and frozen at −20°C until analysis. Weekly representative samples of TMR and orts were analyzed for DM content by oven-drying at 60°C for 48 h. Dietary formulations were adjusted weekly during the experiment period to account for small changes in ingredient DM content. Dried feed samples from each week were ground through a Cyclotec 1093 mill (Tecator, Hoganas, Sweden) using a 1-mm screen and analyzed for composition. Samples were further dried at 105°C for 2 h to determine the absolute DM, and chemical analyses were expressed on the basis of the final absolute DM. Content of CP in feed samples was determined using the macro-Kjeldahl nitrogen test with a Kjeltec digester 20 and Kjeltec System 1026 distilling unit (Tecator). Contents of NDF and ADF were determined using the basic procedure of . In the procedure for NDF determination, sodium sulfite was not used, but heat-stable amylase (type XI-A from Bacillus subtilis; Sigma-Aldrich Corp., St. Louis, MO) was added to the substrate before the procedure.

The daily DMI of individual cows was calculated by subtracting the weekly mean orts from the weekly mean feed offered. Cows were milked 3 times daily, and the milk yield of individual cows was recorded consecutively 3 d of each week. On the last day of each week, milk samples from individual cows were collected from 3 consecutive milkings and composited based on the average milk production at each milking (morning, afternoon, and night; volume ratio: 4:3:3). One aliquot of each milk sample was stored with preservative (bronopol tablet; D&F Control System Inc., Dublin, ON, Canada) at 4°C for analysis of fat, protein, lactose, TS, nonfat solids, CN, citric acid, and urea content using mid-infrared analysis (Foss MilkoScan, Foss Food Technology Corp., Eden Prairie, MN). Another aliquot was stored at −20°C until analyzed for FA composition.

Fatty acid analysis was done according to the method described by , with minor changes as indicated. The frozen milk samples from individual cows were thawed in a refrigerator at 4°C, and centrifuged at 5,000 × g for 20 min at 8°C to separate fat from milk. The fat cake was transferred to a 5-mL tube, and 20 mg of the fat was esterified using the method containing with NaOCH₃:methanol followed by HCl:methanol, described by . Fatty acid methyl esters were separated by using an Agilent 6890 GC (Agilent Technologies, Santa Clara, CA) fit ted with a flame-ionization detector. The samples containing methyl esters in hexane (2 μL) were injected through the split injection port (50:1) onto an HP-88 fused silica 100 m × 0.25 mm column with a 0.20-μm film (Agilent Technologies). The oven temperature was initially 120°C for 10 min and was increased to 230°C at 1.5°C/min and held at that temperature for 30 min. The injector and detector temperatures were maintained at 250°C and 280°C, respectively, and the total run time was 113.33 min. The qualitative external standard method was used in this study. Each peak was identified using known standards of FA and FA methyl esters (Nu-Chek Prep, Elysian, MN; Matreya, Pleasant Gap, PA; Supelco 37 Component FA methyl esters mix, Supelco Inc., Bellefonte, PA). The percentage of each FA was calculated by dividing the area under the FA peak by the sum of the areas for all reported FA peaks. The yield of FA was calculated by multiplying the individual milk FA by milk fat yield and by a coefficient of 93.3% (). Fatty acids were reported as grams per hundred grams of total FA.

The milk yield, milk composition, and milk FA composition were analyzed as a randomized block design with repeated measures using the MIXED procedures of SAS (version 9.0, SAS Institute Inc., Cary, NC). The fixed effects consist of treatment, week, and treatment × week interaction. The variance for cow nested within treatment was used as random error term to test the main effect of treatment. Data from the pretrial period were used as covariates for milk yield and milk composition analysis, and DIM throughout the whole experiment was used as covariate for milk FA composition. The covariance structure was auto-regressive order (1), and downward degrees of freedom using the Kenward-Roger method. Data are presented as covariate-adjusted least squares means with their standard error of the mean. All data were tested for linear and quadratic effects of the 3 supplements. The significance level was declared at P < 0.05 and the trend was declared at P = 0.05 to 0.10.

The model used for milk composition and milk FA analysis was

where Y_ijk = dependent variable measured at week k on the jth cow assigned to the ith treatment; μ = population mean; T_i = treatment effect; D_j(i) = random effect of the jth cow within the ith treatment; W_k = week effect; (TW)_ik = fixed interaction effect between treatment and week; b = the common regression coefficient of pre-treatment data (or DIM) of X_ij; φ_i = the slope deviation of the ith treatment from the common slope b; X_ij = the pretreatment data (or DIM) of cow j on treatment i; and E_ijk = random error associated with the jth cow assigned to the ith treatment at week k.