Milk fat is the main component determining energy expenditure for milk production in dairy cows. It consists almost completely of triglycerides (Moate et al. Reference Moate, Chalupa, Boston and Lean2007). Due to proceeding analytical improvements, up to now more than 400 individual fatty acids (FA) in milk fat are documented (Jensen et al. Reference Jensen, Ferris and Lammi-Keefe1991; Moate et al. Reference Moate, Chalupa, Boston and Lean2007). During the last decades, milk FA composition has gained the interest of manufacturers and consumers as it influences nutritional, physical and flavour properties of dairy products (Bobe et al. Reference Bobe, Lindberg, Freeman and Beitz2007). Changes in the degree of FA unsaturation have an impact on oxidative stability during milk processing (Kay et al. Reference Kay, Weber, Moore, Bauman, Hansen, Chester-Jones, Crooker and Baumgard2005; Glantz et al. Reference Glantz, Lindmark Mansson, Stalhammar, Barström, Fröjelin, Knutsson, Teluk and Paulsson2009). The focus of research regarding the impact of milk FA on human health was especially set on conjugated linoleic acids (CLA) and other polyunsaturated FA (18:2, 18:3) (e.g. Kelsey et al. Reference Kelsey, Corl, Collier and Bauman2003; Palladino et al. Reference Palladino, O`Donovan, Murphy, McEvoy, Callan, Boland and Kenny2009). Dietary manipulation to directly influence the content of these FA is documented in over 100 publications during the last decade (Moate et al. Reference Moate, Chalupa, Boston and Lean2007). Thus nutrition is one of the main factors influencing milk FA composition.
Besides diet, lactational stage along with energy balance of dairy cows have an impact on FA profile in cows' milk (Kay et al. Reference Kay, Weber, Moore, Bauman, Hansen, Chester-Jones, Crooker and Baumgard2005; Stoop et al. Reference Stoop, Bovenhuis, Heck and van Arendonk2009). During the first couple of weeks after parturition, the occurrence of a negative energy balance (NEB) is common in dairy cows. The deficiency of nutrients and energy is compensated by mobilization of body reserves, predominantly adipose tissue associated with the release of FA. During insufficient supply and quality of feed, a NEB may also occur later in lactation as reported for pastured dairy cows averaging 94 days in milk (DIM) (Leiber et al. Reference Leiber, Kreuzer, Nigg, Wettstein and Scheeder2005). The specific effect of a NEB on milk FA composition was shown in a few earlier studies such as Luick & Smith (Reference Luick and Smith1963), who fasted high-yielding dairy cows for 5 d to induce a clinical ketosis or Dann et al. (Reference Dann, Morin, Bollero, Murphy and Drackley2005), who enhanced the post partum (pp) NEB by feed restriction.
The focus of the present study is set on the interactions between energy balance and milk FA profile in dairy cows. Contrary to earlier studies, the present study investigated effects of both the NEB in early lactation and a deliberately induced NEB by feed restriction at around 100 DIM.
Materials and Methods
Animal trial
Fig. 1 shows the experimental schedule of the present study. Detailed information on the animal trial and feeding regimen is given in Gross et al. (Reference Gross, van Dorland, Bruckmaier and Schwarz2011). In brief, the study was conducted with multiparous Holstein dairy cows (3·1±0·2 parities, mean±sem) kept in a free stall barn and covered three experimental periods beginning with parturition. Experimental period 1 (from parturition up to week 12 pp) included 40 cows that were treated similarly. However, only milk samples from 30 out of the 40 cows were obtained for FA analysis at all time points in period 1. In period 2 at around 100 DIM (week 14 pp), cows were divided equally into either a control (CON; n=20) or a restriction group (RES; n=20) with feed restriction for 3 weeks (weeks 15 to 17 pp, Fig. 1). After 3 weeks of the deliberately induced NEB, period 3 (weeks 18 to 21 pp, Fig. 1) started, where RES cows were (re)fed similarly to CON cows. Throughout the study, all animals, except RES cows in feed restriction, had free access to a partial mixed ration 1 (PMR 1; 33·7 % grass silage, 44·9 % corn silage, 6·5 % hay and 14·9 % concentrate on a DM basis). Intake of PMR was recorded individually by troughs connected to electronic balances. When milk yield was above 21 kg/d, additional concentrate (CONC, based on barley, wheat, corn kernels, soybean meal, dried sugar beet pulp with molasses) including a vitamin-mineral premix was fed individually according to milk yield in transponder feeding stations (0·4 kg CONC/kg milk; maximum of 8·9 kg CONC/d). In order to induce an energy deficiency of at least 30 % of cows' requirements at the start of period 2, RES cows received a limited amount of a similar diet as the CON cows, but mixed with an additional 25 % of hay (PMR 2) and a limited amount of CONC (0·4 kg/d). Nutrient values of PMR 1, PMR 2 and CONC are described in Gross et al. (Reference Gross, van Dorland, Bruckmaier and Schwarz2011).
Experimental design.
Determination of energy balance
Individual feed intake (PMR and CONC) and milk yield were recorded daily, body weight weekly. Components of the diets were analysed for crude ash, crude fibre, crude protein and crude fat according to the Weende analysis (Naumann et al. Reference Naumann, Bassler, Seibold and Barth2000) and from that their energy content was calculated according to the German Society of Nutrition Physiology (GfE, 2001). The energy content of PMR 1, PMR 2 and CONC was calculated by multiplying the energy density of the single components with their relative proportion in the diets. Energy content of milk as well as the energetic requirement for maintenance were determined according to GfE (2001). The difference between energy intake by feed and energy expenditure for maintenance and milk production (net energy lactation) results in the energy balance of the individual cow.
Analysis of milk composition, and FA composition of milk and diets
After parturition, cows were milked twice daily (5·00 and 15·00) and milk yield was recorded. Milk samples were pooled once per week from two consecutive milkings, Monday evening and Tuesday morning. Obtained milk aliquots were analysed for milk fat, protein and lactose by an infrared-spectrophotometer (MilkoScan-FT-6000, Foss Analytical A/S, Hillerød, Denmark) in the laboratory of the Milchprüfring Bayern e. V., Wolnzach, Germany. A second aliquot was stored at −20°C until analysis for FA composition. Milk FA composition was determined in weeks 1, 4, 6 and 12 pp of period 1, weekly during period 2 and in weeks 1, 2 and 4 of period 3.
Milk fat was extracted according to Bligh & Dyer (Reference Bligh and Dyer1959), modified by Hallermayer (Reference Hallermayer1976). Total fat from feed samples was extracted according to Naumann et al. (Reference Naumann, Bassler, Seibold and Barth2000). FA composition of milk and feed samples was determined using FA methyl esters (FAME) prepared by transesterification with trimethylsulphonium hydroxide (TMSH). FAME were injected into a gas chromatograph (GC 6890, Agilent Technologies, Waldbronn, Germany) with a flame-ionization detector and a DB23 column (Agilent Technologies, Waldbronn, Germany; 60 m×0·25 mm inner diameter, 0·25 μm film). The carrier gas was hydrogen (constant flow at 1·0 ml/min). Samples were injected by split injection (split ratio 1:100). The oven temperature was increased from 50°C to 175°C at 5 deg C per min, from 175°C to 220°C at 3·3 deg C per min, held for 7·25 min, increased from 220°C to 250°C at 10 deg C per min and held for 5 min. The detector temperature was 260°C. FA were quantified with Chromeleon 6·8 Chromatography Software (Dionex, Sunnyvale CA, USA) using a FAME mix (Sigma Aldrich, St. Louis MO, USA) as standard. FA composition of PMR 1, PMR 2 and CONC is shown in Table 1.
Fatty acid (FA) composition (g/100 g of fatty acid methyl esters, FAME) of experimental diets and concentrate (CONC)
† Partial mixed ration
‡ Concentrate
Statistical analysis
Data presented in text and tables are means±sem of individual measurements in each cow. Relations between energy balance and FA were expressed by the Pearson correlation coefficient. Changes in energy balance, feed intake, milk yield, milk composition and milk FA profile over time during lactation and feed restriction with subsequent realimentation were evaluated by a mixed model in SAS, version 9.2 (SAS Institute, Cary NC, USA) with group, week and the group×week interaction as fixed effect. Differences over time and within groups during feed restriction (period 2) and realimentation (period 3) were detected by Bonferroni's t test. P values <0·05 were considered to be significant.
Results and Discussion
Changes in milk composition and milk FA profile with altering energy balance post partum
Data on energy balance, feed intake, milk yield, milk composition and FA profile are shown in Tables 2 and 3. Lactational stage clearly affected daily milk yield and milk composition. Daily milk yield increased up to week 6 pp and declined thereafter (Table 2). Milk fat and protein contents decreased from week 1 pp up to week 6 pp and slightly increased thereafter, whereas lactose content was relatively constant from week 4 pp onwards (Table 2). With progressing lactation and improvement of energy balance through increasing feed intake, especially milk FA profile markedly changed (Table 3). Energy balance in dairy cows was most negative in week 1 pp and improved with increasing feed intake, but was still negative in week 6 pp (Table 2). In order to compensate the insufficient energy intake pp, considerable amounts of body fat are mobilized resulting in elevated plasma concentrations of non-esterified FA (NEFA) and beta-hydroxybutyrate (BHBA). During the NEB in early lactation, plasma glucose concentration in the present study showed a nadir of 3·30±0·04 mmol/l in week 2 pp, whereas plasma concentrations were highest for NEFA in week 2 pp (0·90±0·06 mmol/l) and for BHBA in week 3 pp (0·98±0·14 mmol/l; Gross et al. Reference Gross, van Dorland, Bruckmaier and Schwarz2011). After the observed peak of lactation, energy requirements could be met by consumed feed resulting in a positive energy balance.
Daily milk yield and milk composition in weeks 1, 4, 6, 12 (all animals), 17 and 21 (control group) post partum. Values without a common superscript letter are significantly different (P<0·05)
† Partial mixed ration
‡ Concentrate
Basically, milk FA can be derived from four major pathways: the diet, the mammary gland (de-novo synthesis), the rumen (biohydrogenation, bacterial degradation and synthesis), and body fat mobilization (Stoop et al. Reference Stoop, Bovenhuis, Heck and van Arendonk2009). Changes in milk FA composition during lactation originate from altered activities in these pathways (Van Knegsel et al. Reference Van Knegsel, van den Brand, Dijkstra, Tamminga and Kemp2005; Stoop et al. Reference Stoop, Bovenhuis, Heck and van Arendonk2009).
For the evaluation of changes in milk composition and milk FA profile in the present study with altering energy balance pp depending on the same feeding regimen, data from all animals in period 1 and CON group in periods 2 and 3 were evaluated in order to investigate the effect of continuous lactation. In the present study, most changes in milk FA profile took place during the observed NEB from weeks 1 to 6 pp, while FA composition was relatively constant between weeks 12 and 21 pp. FA up to 16:0 showed lowest proportions in week 1 pp that increased to relatively constant proportions from week 12 onwards (Table 3). These findings agree with earlier studies (Stull et al. Reference Stull, Brown, Valdez and Tucker1966; Palmquist et al. Reference Palmquist, Beaulieu and Barbano1993; Kay et al. Reference Kay, Weber, Moore, Bauman, Hansen, Chester-Jones, Crooker and Baumgard2005; Garnsworthy et al. Reference Garnsworthy, Masson, Lock and Mottram2006). Confirming results of Stoop et al. (Reference Stoop, Bovenhuis, Heck and van Arendonk2009), saturated FA (SFA), especially 16:0 increased from weeks 1 to 12 pp in the present study (Fig. 2), while monounsaturated FA (MUFA), mainly represented by 18:1,9c decreased until week 12 pp with improving energy balance (Table 2). The proportion of polyunsaturated FA (PUFA) was relatively constant from week 1 up to week 21 pp (Table 3). Due to the increased adipose tissue mobilization during the NEB in early lactation, preformed FA concentrations (sum of FA >C16) were greatest in week 1 pp and decreased in a similar pattern to that reported by Kay et al. (Reference Kay, Weber, Moore, Bauman, Hansen, Chester-Jones, Crooker and Baumgard2005). Oleic acid (18:1,9c) is the predominant FA in adipocytes and primarily released through lipolysis during NEB (Rukkwamsuk et al. Reference Rukkwamsuk, Geelen, Kruip and Wensing2000). Plasma NEFA and triglycerides are utilized by the mammary gland for milk FA synthesis (Moore & Christie, Reference Moore and Christie1979). The high transfer rate of 18:1,9c from plasma into milk fat (Tyburczy et al. Reference Tyburczy, Lock, Dwyer, Destaillats, Mouloungui, Candy and Bauman2008) confirms the elevated proportion of 18:1,9c in milk fat during the NEB pp in the present study. Especially long-chain FA are derived from plasma and incorporated into milk fat (Palmquist et al. Reference Palmquist, Beaulieu and Barbano1993) and inhibit the de-novo synthesis of short-chain FA by the mammary gland (Bauman & Davis, Reference Bauman, Davis, Larson and Smith1974). The observed increase in short-chain FA with progressing lactation in the present study is consistent with the decreasing adipose tissue mobilization at around weeks 4–6 pp (Garnsworthy & Huggett, Reference Garnsworthy and Huggett1992; Palmquist et al. Reference Palmquist, Beaulieu and Barbano1993).
Proportions (g/100 g of fatty acid methyl esters, FAME) of de-novo synthesized FA (<C16; circles), C16 FA (triangles) and C18:1,9c (squares) in milk fat during the first 21 weeks of lactation in dairy cows.
Milk fatty acid (FA) composition (g/100 g of fatty acid methyl esters, FAME) in weeks 1, 4, 6, 12 (all animals), 17 and 21 (control group) post partum. Values without a common superscript letter are significantly different (P<0·05)
Contrary to the findings of Stoop et al. (Reference Stoop, Bovenhuis, Heck and van Arendonk2009), trans FA slightly increased with improved energy balance of dairy cows in the present study (Table 3). The proportion of CLA in milk fat remained constant from week 1 to week 12 pp and increased slightly thereafter up to week 21 pp. Milk FA from de-novo synthesis (⩽C14 and part of C16) in the present study increased from week 1 pp up to week 12 pp (Fig. 2) in agreement with Palmquist et al. (Reference Palmquist, Beaulieu and Barbano1993) and Kay et al. (Reference Kay, Weber, Moore, Bauman, Hansen, Chester-Jones, Crooker and Baumgard2005). Although the FA synthesized de novo comprise approximately 40 % by weight over the entire lactation, preformed FA contribute a larger portion of the total FA in early lactation (Kay et al. Reference Kay, Weber, Moore, Bauman, Hansen, Chester-Jones, Crooker and Baumgard2005). Percentage of preformed FA decreased up to week 12 pp in the present study. Thereafter milk composition and milk FA profile were stable until the end of the study. Garnsworthy et al. (Reference Garnsworthy, Masson, Lock and Mottram2006) concluded that stage of lactation does not affect the relative incorporation of de-novo synthesized and preformed FA when the composition of diets remains constant. Because of the same feeding regimen in the present study, changes in milk FA profile regarding de-novo synthesized and preformed FA therefore reflect changes in energy balance of dairy cows.
Changes in milk and milk FA composition during feed restriction and subsequent realimentation
Due to feed restriction in period 2, RES cows experienced a significant NEB in weeks 15 to 17 pp being even more intense compared with the NEB occurring in the first 6 weeks of lactation (Tables 2 and 4). Besides energy supply the dietary composition also was changed in the group of RES cows during period 2 as the amount of the PMR and the concentrate was limited. Additionally, PMR 2 fed to RES cows included more hay than the PMR 1 fed to CON cows. Despite the higher induced NEB by feed restriction compared with the NEB in early lactation, plasma NEFA and BHBA concentrations were elevated only moderately in RES cows compared with CON cows during period 2 (0·23±0·02 v. 0·14±0·01 mmol/l and 0·62±0·03 v. 0·52±0·02 mmol/l, respectively).
Milk yield and milk composition for feed-restricted (RES) and control cows (CON) during feed restriction and realimentation. For differences over time within group, values without a common superscript are significantly different (P>0·05), significant differences between RES- and CON-group within week are marked with * (P<0·05)
† Partial mixed ration
‡ Concentrate
§ Energy balance
¶ Fat-corrected milk
Milk fat content was elevated during feed restriction for RES cows (Table 4). Changes in milk yield and composition have been discussed in Gross et al. (Reference Gross, van Dorland, Bruckmaier and Schwarz2011). Data on the effect of energy balance on milk fat composition are scarce (Stoop et al. Reference Stoop, Bovenhuis, Heck and van Arendonk2009). In the present study, milk FA profile of CON cows was stable during the whole time of periods 2 and 3. For RES cows, the proportion of most FA ⩽16:0 (e.g. 6:0, 10:0, 10:1, 12:0, 14:0, 16:0) was decreased during the NEB induced by feed restriction compared with the respective initial values, whereas preformed FA, especially 17:1,9c, 18:0 and 18:1,9c arising from body fat mobilization increased markedly during feed restriction (Table 5). These changes occurred rapidly within the first week of feed restriction (on average 3 d distance between the start of feed restriction and the next milking sample) and disappeared completely within one week of realimentation (4 d on average). The proportion of CLA in milk fat of RES cows was elevated at the start of feed restriction and adjusted immediately to initial values (Table 5). The rise of CLA in RES cows may be attributed to the changed diet towards less concentrate and more roughage. Despite the maintenance of the deliberately induced NEB by feed restriction at a relatively constant level, FA showed a tendency during the NEB to adjust towards the initial levels with progressing feed restriction. The pattern of decreasing short-chain FA and increase of long-chain FA during an induced NEB was reported in an earlier study of Luick & Smith (Reference Luick and Smith1963). During feed restriction (period 2), SFA decreased, while MUFA (especially 18:1,9c) increased for RES cows compared with CON cows (Fig. 3). PUFA were stable during period 2 and the following realimentation period in RES cows (Table 5).
Proportions (g/100 g of fatty acid methyl esters, FAME) of de-novo synthesized FA (<C16; circles), C16 FA (triangles) and C18:1,9c (squares) in milk fat of feed-restricted cows (filled symbols) and control cows (empty symbols) during feed restriction (weeks 15–17 post partum) and subsequent realimentation (weeks 18–19 post partum). Stars indicate a significant difference between the groups (P<0·05).
Milk fatty acid (FA) composition (g/100 g fatty acid methyl esters, FAME) for feed-restricted (RES) and control cows (CON) during feed restriction and realimentation. For differences over time within group, values without a common superscript are significantly different (P>0·05), significant differences between RES- and CON-group within week are marked with * (P<0·05)
Luick & Smith (Reference Luick and Smith1963) examined whether changes in milk FA composition during feed restriction are caused by a decreased synthesis of short-chain FA or by an increased incorporation of long-chain FA absorbed from plasma. According to Luick & Smith (Reference Luick and Smith1963), the failure to utilize beta-hydroxybutyrate occurs when its concentration in plasma is elevated, i.e. during fasting and ketosis. This in turn, accounts for the relatively high levels of oleic acid found in milk fat of fasting and ketotic cows (Luick & Smith, Reference Luick and Smith1963) as observed in the present study. Although less in their extent, milk FA in the present study clearly showed a similar pattern during a deliberately induced NEB by feed restriction at around 100 DIM compared with the NEB in early lactation. Compared with milk of dairy cows in positive energy balance (week 14 pp), proportions for single FA (e.g. 17:1,9c; 18:1,9c) were changed up to 80 % during the NEB in early lactation and during the deliberately induced NEB by feed restriction.
Van Haelst et al. (Reference Van Haelst, Beeckman, van Knegsel and Fievez2008) determined whether concentrations of specific FA in milk fat are suitable for the early detection of subclinical ketosis as mobilization of adipose tissue precedes development of ketosis (Reist et al. Reference Reist, Erdin, von Euw, Tschuemperlin, Leuenberger, Delavaud, Chilliard, Hammon, Kuenzi and Blum2003). Van Haelst et al. (Reference Van Haelst, Beeckman, van Knegsel and Fievez2008) suggested the elevated proportion of 18:1,9c as an interesting trait for prediction of subclinical ketosis, particularly since this FA was elevated in milk fat before diagnosis of ketosis. As milk FA changed with altering energy balance, it is obvious to identify milk FA indicating a NEB in dairy cows independent of their lactational stage. Therefore, correlations between milk FA and the energy balance of all cows were calculated. Correlation over all cows between energy balance and the proportion of 18:1,9c was 0·77. We wanted to investigate whether the correlation is higher in cows with a higher NEB. Thus the correlation between energy balance and the proportion of 18:1,9c in milk fat of 5 cows with the most negative EB in week 1 pp (−80·2 MJ NEL/d) was calculated, but did not show a closer relationship (r=0·62). The correlation between EB and 18:1,9c for RES cows with the highest NEB in the first week of feed restriction (−83·2 MJ NEL/d) was 0·92. From these results, an elevated proportion of 18:1,9c in milk fat can be confirmed to be a suitable marker for a NEB. The correlation between energy balance and a single FA during the NEB in early lactation and during the deliberately induced NEB ranged from 0·71 to 0·96. However, the low proportion and relatively high variation in changed FA (e.g. 11:0; 12:1; Tables 3 and 5) restrict the predictive value of these FA to indicate a NEB although changes were significant between RES and CON cows. On the contrary, the correlation between energy balance and groups of FA was higher in cows with a more intense NEB during feed restriction that ranged from 0·92 up to 0·98 for SFA, MUFA (predominantly 18:1,9c), de-novo synthesized (sum of FA<C16, r=0·94) and preformed FA (sum of FA >C16, r=0·92). The higher proportion of these summarized FA in milk fat makes their changes a more appropriate tool reflecting energy balance in dairy cows compared with a single FA represented at a low concentration.
Conclusions
Post-partum changes in milk FA profile in the present study followed previous findings. Short- and medium-chain FA up to C16 increased with the decreasing NEB pp, while long-chain FA, especially 18:1,9c decreased as mobilization of body fat reserves declined. The responses of FA profiles of cows' milk due to a NEB at two lactational stages in the present study – the NEB in early lactation and the deliberately induced NEB by feed restriction – was similarly directed. Despite the maintenance of a high NEB during the feed restriction period, changes in milk FA profile were less pronounced compared with changes during the NEB in early lactation and tended to adjust to the initial composition. However, milk FA profile changed within a few days after initiation of the deliberately induced NEB and showed no more differences within the first week of realimentation compared with control cows. For the dietary composition and feeding regimen in the present study, the close relationship with energy balance makes changes in 18:1,9c as well as in groups of FA (SFA, MUFA, de-novo synthesized and preformed FA) suitable indicators of the energy balance in dairy cows.
