in milk

1256

Received: 22 October 2009 Revised: 10 February 2010 Accepted: 17 February 2010 Published online in Wiley Interscience:

(www.interscience.wiley.com) DOI 10.1002/jsfa.3968

Influence of low-level supplementation

of grazing dairy cows with cereals or sugar beet pulp on the concentrations of CLA isomers

in milk

Manuela Renna, ^† Marius Collomb, Andreas M ¨unger and Ueli Wyss ^∗

Abstract

BACKGROUND: A wide range of isomer-specific health benefits have been attributed to conjugated linoleic acids (CLAs).

Little information is currently available on the influence of different feed components on the concentration of CLA isomers in ruminant-derived products. The aims of this study were to compare the effect of cereal mix or dried sugar beet pulp supplementations fed to grazing dairy cows on milk CLA isomeric distribution and to examine the monthly variation of CLA isomers during the grazing season.

RESULTS: The concentrations of the isomerst7c9,t10c12 andt10t12 were higher in milk from cows supplemented with cereals (P≤0.001). The milk of beet pulp-supplemented cows showed higher levels of the isomerst11c13 (P≤0.05),t9t11 (P≤0.001) andt7t9 (P≤0.01). Total CLA as well as seven other minor isomers were not significantly affected by the concentrate type.

Monthly variations occurred extensively for almost all detected isomers. Total CLAs showed the highest values at the beginning of the grazing season and in autumn in coincidence with plant regrowth.

CONCLUSION: These results show that even at low levels, supplement types can significantly influence the concentration of some CLA isomers in milk fat; an indication is given of the possibility to manipulate the animals’ diet to change the CLA isomeric profile in ruminant-derived products.

c 2010 Society of Chemical Industry

Keywords:bovine milk; conjugated linoleic acids; pasture; cereal mix; beet pulp

INTRODUCTION

Conjugated linoleic acids (CLAs) are a group of positional and geometric isomers of octadecadienoic acid (C18 : 2), with double bonds located in adjacent carbon atoms. Ruminant- derived food products (dairy and meat) are the major sources of CLAs in the human diet.¹A total of 16 different CLA isomers have been detected in dairy products by using high-resolution chromatography.²In lactating dairy cows, CLA isomers containing acis-9 double bond (the predominant onesc9t11 andt7c9) mainly originate from endogenous synthesis in the mammary gland by the activity of9-desaturase.^{3 – 5}The latter, also named stearoyl-CoA desaturase (SCD), is a key enzyme in lipid metabolism, allowing the endogenous synthesis of many unsaturated fatty acids (UFAs) by adding acis-9 double bond to some saturated and unsaturated FAs with a chain length of 10 to 18 carbon atoms.⁶Other minor CLA isomers, ranging from6,8 to13,15, originate from UFAs of feed origin that escape complete microbial biohydrogenation in the rumen.⁵Some of these isomers are present only in traces in milk fat.⁷

Increasing interest in CLAs is mainly due to their wide range of isomer-specific biological activities (protection against carcino- genesis, obesity, diabetes, arteriosclerosis and other inflammatory diseases), which have been highlighted in several biomedical

studies with a variety of cell cultures and mammalian models of human diseases.⁸ Some, but not all, of the biological activities observed in animals have been also reported in humans, but with less-marked effects.^9,10

Dietary factors are known to strongly influence milk fatty acid profiles. Concerning CLA isomers, while the effect of different diet components on C18 : 2c9t11 (rumenic acid, RA) concentrations has been widely investigated, there is a lack of knowledge concerning the same influence on the concentrations of minor isoforms in ruminant-derived products. Among dietary regimens, pasture feeding was found to provide high contents of RA in several studies.¹¹ A strong positive relationship between fresh forage intake and the content of other minor CLA isomers in bovine milk fat has also been reported recently.^{12 – 14}

∗ Correspondence to: Ueli Wyss, Agroscope Liebefeld-Posieux Research Station ALP, CH-1725 Posieux, Switzerland. E-mail: ueli.wyss@alp.admin.ch

† On secondment from Dipartimento di Scienze Zootecniche, Universit`a degli Studi di Torino, via L. da Vinci 44, 10095 Grugliasco (TO), Italy.

Agroscope Liebefeld-Posieux Research Station ALP, CH-1725 Posieux, Switzerland

1257

Pasture feeding alone is often not sufficient to meet total nutritional requirements in high producing lactating dairy cows.¹⁵ Consequently, in common practice different types of concentrates are often supplemented in order to cover energy, protein, mineral and vitamin deficiencies as well as to enhance individual milk yields. Cereals (characterised by high starch levels) and beet pulp (containing notable amounts of highly digestible fiber) are common diet components for dairy cows in the lowland regions of Europe. Notwithstanding, no studies are currently available on their influence on bovine milk CLA isomer concentrations when used as supplements to the pasture feeding. It is well known that diet constituents high in starch and low in fibre can result in a notable postprandial decrease in ruminal pH. As a consequence, alterations in the biohydrogenation pathways in the rumen can occur leading to increased formation of specific fatty acid (FA) intermediates (e.g. CLAt10c12) characterised by anti-lipogenic effects.^16,17 This phenomenon is known as milk fat depression (MFD) and has sometimes been observed in pasture-fed ruminants,¹⁸although whether it could be enhanced by starch-rich supplements to fresh grass-based diets has not yet been documented. A milk CLA isomeric profile could also be influenced by differentiation in the dietary PUFA supply in concentrate supplements.

The aims of this study were therefore (1) to compare the effect of a low level supplementation of corn–barley or beet pulp concentrates to grazing dairy cows on the concentrations of CLA isomers in milk fat; (2) to examine the variations of all detected CLA isomers throughout the grazing season; and (3) to assess if interactions between concentrate type and sampling month occur. The effects of concentrate type, month and their interaction on estimated9-desaturase activity were also studied.

EXPERIMENTAL

Animals, design and treatments

The experiment lasted 8 months (from April to November 2005) and was conducted at the Agroscope Liebefeld-Posieux Research Station, Switzerland (latitude: 46^◦4609N; longitude: 07^◦0619 E; altitude: 650 m a.s.l.; average annual rainfall: 1029 mm). Initially, 16 dairy cows [12 Red Holstein (RH), two Holstein Friesian (HF), and two Brown Swiss (BS)] were chosen for the experiment, although seven of these were excluded after 5 months as they neared the end of their lactation. The cows were attributed to two balanced groups. The first group (supplemented with cereals) was composed of seven RH and one HF cows from April to August and by remaining five RH cows from September to November.

The second group (supplemented with beet pulp) was composed of five RH, one HF, and two BS cows from April to August and by remaining three RH and one BS in the last 3 months of the trial. Group balancing was performed on the basis of the cows’

stage of lactation (means and standard deviations of days in milk (DIM) postpartum at the beginning of the experimental period:

first group 72.1±23.5; second group 75.6±24.5), lactation number (3.1±1.3 and 2.9±1.3, respectively) and milk yield (37.7±9.5 and 36.9±7.1 kg head⁻¹day⁻¹, respectively). The total CLA content did not statistically differ between the two groups at the beginning of the experiment (4.40±0.08 and 4.23±0.09 mg g⁻¹fat, respectively).

From the 28 March to the 11 April (pre-experimental period) all cows grazed during daytime (0800 to 1500 hours) and were offered a mixed ration (MR) consisting of 40% lucerne hay,

35% sugar beet pulp silage and 25% maize silage during the remainder of the day. In addition they were offered hay (third cut) and a concentrate supplement consisting of a combination of corn–barley mix, dried sugar beet pulp, protein feed and mineral mix. The experimental period ran between the 12 April and 13 November when the cows were allowed to pasture during both day and night although, until the end of April, were still offered diminishing quantities of MR and hay after milking. Throughout the experiment each group received concentrates formulated to diverge in their carbohydrate composition; either a starch based supplement (corn–barley at 1 : 1, CB) or a digestible fibre supplement (sugar beet pulp, BP). They were offered at variable levels dependent on milk yield but balanced for net energy supply (estimated according to Arrigo et al.¹⁹). In August, due to the drought conditions occurring in the region in summer 2005, fresh grass and low concentrate levels alone were not adequate to cover the nutritional requirements of the cows. For this reason varying amounts (average about 5 kg head⁻¹day⁻¹) of maize silage were added to the diet in both groups of cows for several weeks. A total (pre-experimental and experimental periods) of 122 milk samples have been analysed. Diet formulations during the pre-experimental and the experimental periods are presented in Table 1.

Grazing management and botanical composition of the paddocks

The grazing management system was a rotational one with 2–3 days grazing per paddock in each rotation.

The grazing paddocks consisted of about 78% Poaceae (with Lolium perenne L. as the most abundant species) and 11%

Fabaceae (almost entirely represented by Trifolium repens L.).

The remaining forage plants belonged to different botanical families (Asteraceae, Plantaginaceae, Polygonaceae, Apiaceae, Ranunculaceae), with Taraxacum officinale as the most repre- sentative one.

Sampling procedure and chemical analysis Forages

Representative pasture samples were hand-plucked once a month such as to approximate the height at which the cows grazed.

Samples were dried in an oven at 60^◦C for 24 h and stored at room temperature until analysis for chemical and FA compositions. The dried samples were ground (Brabender^mill; Brabender GmbH &

Co., Duisburg, Germany) to pass a 1 mm screen. Dry matter and total ash contents were determined after heating for 3 h at 105^◦C and 550^◦C, respectively, according to VDLUFA.²⁰Crude protein (CP) was determined using the Dumas method (AOAC,²¹ no.

968.06; Leco Model FP-2000; Leco Corp., St Joseph, MI, USA). Cell wall constituents were analysed according to standard protocols [acid detergent fibre (ADF): AOAC,²¹no. 973.18, expressed without residual ash after incineration at 500^◦C for 1 h; neutral detergent fibre (NDF): Mertens,²² assayed with heat stable amylase and expressed without residual ash]. The sugar content was analysed as described by Shannon²³with a colorimeter autoanalyser (Type II; Bran+Luebbe, Norderstedt, Germany). Crude fat was assayed by the Berntrop method according to VDLUFA.²⁴ Starch was analysed with a 241 polarimeter (Perkin Elmer, Wellesley, MA, USA) according to the standard protocol of VDLUFA (Naumann and Bassler²⁵).

For FA analysis, fat was extracted using a mixture of dichloromethane and methanol according to Winter.²⁶ After

1258

Table 1. Diet formulations (daily intake, in kg day⁻¹) in corn –barley (CB) and beet pulp (BP) supplemented cow groups along the pre-experimental (March^†) and the experimental (April^‡– November) periods

Pre-experiemental period Experiemental period

March^† April^‡ May June July Aug Sept Oct Nov

Dietary component All cows CB BP CB BP CB BP CB BP CB BP CB BP CB BP CB BP

Fresh forage (pasture) Restricted a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l. a. l.

CB concentrate (1 : 1) 3.0 4.1 – 4.1 – 3.6 – 3.0 – 2.7 – 1.3 – 1.3 – 1.3 –

BP concentrate 2.2 – 2.8 – 4.2 – 3.6 – 3.1 – 3.0 – 1.1 – 0.8 – 0.8

Protein concentrate 1.6 0.4 0.7 – – – – – – – – – – – – – –

Mineral mix concentrate 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.3 0.3 0.4 0.4 0.4 0.4

Hay (third cut) 3.2 1.0 1.0 – – – – – – – – – – – – – –

Maize silage – – – – – – – – – 5.4 4.8 – – – – – –

MR 29.1 7.9 8.7 – – – – – – – – – – – – 16.1 14.7

Bold font indicates the prevalent concentrate type supplementation.

†We referred to March as the period from 28 March to 11 April (pre-experimental period).

‡We referred to April as the period from 12–30 April.

a. l.,ad libitum: the cows were allowed to pasture only during the day until 11 April and during both day and night from 12 April to November.

MR, mixed ration: lucerne hay (40% of DM), sugar beet pulp silage (35%) and maize silage (25%).

Table 2. Monthly chemical composition (g kg⁻¹dry matter, unless otherwise stated) of the fresh grass (pasture) Month

Component April May June July Aug Sept Oct Nov

Main nutrients

DM (g kg⁻¹) 154 144 202 205 166 119 150 159

Ash 107 97 100 101 118 141 235 128

CP 252 226 186 226 205 241 183 217

NDF 349 419 443 432 433 407 339 213

ADF 198 225 252 238 233 259 198 383

Sugar 123 117 108 94 94 70 91 130

Fatty acids

C16 2.81 2.56 2.13 2.69 2.19 2.83 2.25 2.78

C18 0.25 0.23 0.25 0.46 0.27 0.22 0.21 0.28

C18 : 1 0.35 0.53 0.45 0.52 0.47 0.44 0.50 0.54

C18 : 2 3.51 2.98 2.40 2.67 1.75 3.39 2.19 2.99

C18 : 3 19.68 14.70 10.17 13.47 7.14 14.98 11.32 16.24

Other FAs^† 0.36 0.23 0.16 0.16 0.31 0.34 0.17 0.47

Total FAs 26.96 21.23 15.56 19.97 12.13 22.20 16.64 23.30

†Sum of C20+C22+C22 : 2+C22 : 6.

ADF, acid detergent fibre; CP, crude protein; DM, dry matter; FAs, fatty acids; NDF, neutral detergent fibre.

saponification with a solution of NaOH, the resulting FA salts were converted into fatty acids methyl esters (FAMEs) in the pres- ence of boron trifluoride according to ISO standard 5509.²⁷FAMEs were then separated and quantified by gas chromatography (HP 5890 series II GC) using a Supelcowax-10 capillary column (30 mL

×0.32 mm ID, 0.25µm film thickness; Supelco 2–4080; Supelco, Bellefonte, PA, USA). The column temperature was held at 170^◦C for 1 min, then raised at 2.5^◦C min⁻¹to 210^◦C and subsequently 0.5^◦C min⁻¹ to 220^◦C. The isotherm was maintained for 5 min before rising again at 15^◦C min⁻¹to a final temperature of 250^◦C, where it remained for 6 min. The temperature of both the injector and the flame-ionisation detector was maintained at 250^◦C; the injection volume was 1µL; the nitrogen constant linear flow rate was set at 25 mL min⁻¹. All analyses were done in duplicate. The

chemical composition of the fresh grass during the experimental period is presented in Table 2.

Milk

Cows were milked twice a day in a milking parlor, model 2×3, produced by Lemmer-Fullwood GmbH (Lohmar, Germany).

Individual milk yields (both morning and afternoon milkings) were recorded daily. Milk samples for gross composition and FA analyses were taken once a month and frozen at−20^◦C until analysed. Milk fat, protein and lactose percentages were determined by infrared spectroscopy (Combi-Foss^ 6000 FC; Foss, Hillerød, Denmark) according to FIL-IDF.²⁸

For FA analysis, milk fat extraction was obtained by centrifuga- tion at 5000×gfor 30 min. The resulting creams were churned

1259

at approximately 5^◦C. After the resulting molten butter had been filtered through a hydrophobic filter (1 PS folded filter; Whatman, Bottmingen, Switzerland), the pure milk fat was collected and stored at−20^◦C until analysis. Pure milk fat was dissolved in hex- ane and FAMEs were obtained bytrans-esterification of glycerides by using a solution of potassium hydroxide in methanol.²⁹ FAs were determined by a high-resolution gas chromatography(Agi- lent 6890; Agilent Technologies, Santa Clara, CA, USA) equipped with a CP-Sil 88 capillary column (100 m×0.25 mm ID, 0.20µm film thickness; Varian BV, Middleburg, Netherlands) as previously reported by Collomb and B ¨uhler.³⁰Quantification was assessed by using nonanoic acid as internal standard.

CLA isomers were analysed by silver-ion (Ag⁺) -HPLC (Agilent LC 1100) equipped with a photodiode array detector (234 nm) using three ChromSpher Lipid columns in series (stainless steel, 250×4.6 mm, 5µm particle size; Chrompack, Middleburg, Netherlands) according to Collombet al.³¹The solvent consisted of UV-grade hexane with 0.1% acetonitrile and 0.5% ethyl ether (flow rate 1 mL min⁻¹), prepared fresh daily. The injection volume was 10µL, corresponding to<250µg lipid. The HPLC areas for t7c9+t8c10+c9t11 (t = trans,c= cis) were added and used for comparison of the peak area of the three isomers from the GC chromatogram. The results were expressed as absolute values as mg g⁻¹fat.

Statistical analysis

All statistical analyses were performed using the SPSS software for Windows.³² Weekly averages of individual milk yields corre- sponding to the weeks when milk samples were also taken for chemical and FA composition analyses were computed and used for statistical analysis.

Normal Q–Q plots and the Kolmogorov–Smirnov test were used to check variables for normality. Variables which were not normally distributed were log-transformed prior to further statistical analysis, but the presented results were calculated from non-transformed data. Estimated marginal means of milk yield, gross components, FAs and desaturase index (DI) were computed using the General Linear Model (GLM) procedure according to the following model:

Y_ijkl=µ+CT_i+M_j+(CT×M_ij)+DIM_k+e_ijkl, whereY_ijkis the dependent variable;µis the overall mean; CT_i is the fixed effect of concentrate type (ican be CB or BP);Mjis the fixed effect of month (jcan be April, May, June, July, August, September, October or November); CT×M_ij is the interaction between concentrate type and month; DIM is a covariate; ande_ijk is an unexplained residual element.

Correlations were performed by computing Pearson’s or Spearman’s coefficients (if variables were normally or not normally distributed, respectively). Significance was declared atP≤0.05.

RESULTS AND DISCUSSION

Characteristics of the two experimental concentrates

The composition of the two experimental concentrates used for the main supplementation of the pasture-based diet is presented in Table 3. Main differences between the concentrates were related to their carbohydrate and FA compositions.

The BP concentrate was characterised by notably higher fibre contents while the CB concentrate contained a high starch

Table 3. Chemical composition (g kg⁻¹dry matter, unless otherwise stated) of the concentrates used for main supplementation of the pasture-based diet

Experimental concentrate Other Component Corn –barley Beet pulp Mineral mix^† Main nutrients

DM (g kg⁻¹) 873 905 917

Ash 43 61 435

CP 107 93 64

NDF 137 394 192

ADF 42 234 83

EE 35 10 65

Starch 656 0 NA

Sugar 34 100 31

NSCs^‡ 678 442 244

Fatty acids

C16 2.01 1.12 7.70

C18 0.30 ND 16.51

C18 : 1 3.50 0.52 8.25

C18 : 2 8.67 1.41 3.69

C18 : 3 0.38 0.18 0.16

C20 ND ND 0.24

C22 ND ND 0.18

Total FAs 14.86 3.23 36.73

†Supplemented to both groups of cows throughout the whole experimental period according to the schedule presented in Table 1.

‡Calculated as 1000−(NDF+CP+EE+ash).

ADF, acid detergent fibre; CP, crude protein; DM, dry matter; EE, ether extract; FAs, fatty acids; NA, not analysed; ND, not detected; NDF, neutral detergent fibre; NSCs, non-structural carbohydrates.

concentration. Concerning the FA composition, the cereal-based concentrate showed considerable higher total FA amounts, particularly more than six-fold higher total octadecenoic ( C18 : 1) and total octadecadienoic (

C18 : 2) acids relative to beet pulp. Also the total octadecatrienoic (

C18 : 3) acid content was higher in the CB concentrate (approximately double the values observed in BP), but the relative concentrations in both cases were very low if compared to those of the fresh grass.

Milk yield and gross composition

The type of concentrate significantly affected neither milk production nor milk fat, protein and lactose contents (Table 4).

The explanation can be related to the low amount of concentrates supplemented to the pasture feeding (F : C ratio, estimated according to Arrigoet al.,¹⁹approximately equal to 80 : 20 during the whole experimental period). Our results confirm the findings by Coulonet al.³³who reported no appreciable effect of different types of concentrates (starchy or containing highly digestible fibre) on milk production and its main constituents when the amount of concentrate in the diet is lower than 50% of the total feeding.

CLA isomer content in milk fat

Descriptive statistics of detected CLA isomers during the exper- imental period are presented in Table 5. Rumenic acid was the predominant isomer in milk fat. Its average concentration was equal to 13.06 mg g⁻¹ fat, accounting for 85.7% of total CLA.

1260

Table 4. Effect of the type of concentrate (corn –barley or beet pulp), month and their interaction on milk yield and gross composition Significance^† Corn –barley, Beet pulp,

Parameter n=55 n=51 SEM CT M CT×M

Milk yield (kg head⁻¹ day⁻¹)

27.3 27.5 0.9 NS ^∗∗ NS

ECM^‡(kg head⁻¹ day⁻¹)

26.8 26.6 0.8 NS ^∗∗∗ NS

Fat (g kg⁻¹) 38.6 39.1 1.3 NS 0.06 NS

Milk fat yield (kg head⁻¹ day⁻¹)

1.0 1.0 0.01 NS ^∗∗∗ NS

Protein (g kg⁻¹)

35.1 34.3 0.6 NS ^∗∗∗ NS

Lactose (g kg⁻¹)

47.1 47.2 0.4 NS NS NS

†Probability:^∗P≤0.05;^∗∗P≤0.01;^∗∗∗P≤0.001; thePvalue is shown if, not being significant, it shows a tendency 0.05<P<0.10; NS, not significant (P≥0.10).

‡Calculated according to Arrigoet al.¹⁹

CT, effect of concentrate type;M, effect of month; CT×M, interaction between concentrate type and month;n, number of samples; ECM, energy-corrected milk.

A high correlation between RA and trans-vaccenic acid (C18 : 1 t11, TVA) was found (Fig. 1a), reflecting the direct precursor and product relationship between these two FAs in the mam- mary gland. Pearson’s correlation coefficients were similar in the two treatment groups (0.90 and 0.89 for CB and BP cows, re- spectively;P ≤ 0.001) and also analogous to values previously reported in the literature.^34,35 An increase in both TVA and RA contents in milk fat after the transition from a prevailing MR diet (pre-experimental period) to a prevailing fresh grass-based diet (experimental period) was also observed (Figs 1a, 2a and m) probably as a consequence of the higherα-linolenic acid (C18 : 3 c9c12c15, ALA) content normally found in fresh forages relative to conserved ones.³⁶

C18 : 2t7c9 has been found to be the second most abundant CLA isomer in many commercial milk and dairy products^37,38as well as in controlled trials under different dietary conditions.^5,12,31 However, in this study CLA t7c9 accounted, in both groups of cows, for less than 3% of total CLA. Its concentrations were equal to 0.39±0.08 and 0.34±0.06 mg g⁻¹fat in the CB and BP groups, respectively.

In this study, thet11c13 isomer ranked second in concentration after RA (3.2 and 4.3% of total CLA in the CB and BP groups, respectively) rather thant7c9. The former has been observed as the second most represented isomer in milk fat from grazing ruminants^7,14,39and has been suggested as an appropriate marker of grass feeding. Such observations led to the hypothesis that ALA, the most representative FA in forage plants, could be the indirect precursor oft11c13 in the rumen³⁹by following a specific metabolic pathway,³¹ but such an assumption has still to be confirmed experimentally.

Thet11t13 isomer showed similar values to thet7c9 isomer, accounting for 2.8% of total CLAs. For all other individual isomers, the relative percentage did not exceed 1.3% of total CLAs.

The concentration of total CLAs and that of all individual isomers were comparable to values previously obtained in other trials with dairy cows managed according to low-input production systems (high percentage of grass-based diets, e.g. pasture

feeding conditions).^7,12,14,39A single exception was represented by the concentration of CLA t11c13 in CB-supplemented cows (0.50 mg g⁻¹milk fat), which was actually lower relative to values reported in the above-mentioned trials under high forage diets conditions (0.61–0.84 mg g⁻¹milk fat).

9-Desaturase activity

In order to estimate the9-desaturase activity in the mammary gland, the computation of different desaturase indexes (DIs) has been suggested.⁶Among them, the ratio between myristoleic and myristic acids (DI₁₄) is normally considered the best indicator because all myristoleic acid synthesis from myristic acid is regulated in the mammary gland by the activity of this enzyme.³In dairy cows,9-desaturase activity has been found to be influenced by genetics,^6,40,41 parity,⁴⁰stage of lactation^6,40 and nutrition.⁴² Individual variability seems to be of great extent.^42,43 However, factors affecting this enzyme in the mammary gland are still largely unexplored.

Differences in estimated9-desaturase activity among the cows involved in the current trial were conspicuous: at the beginning of the trial (pre-experimental period) the variation of DI₁₄ was 2.6-fold (from 0.034 to 0.088) among individuals consuming the same diet, confirming a key role of genetics.

No statistically significant variation was observed for DI₁₄be- tween the two groups of cows during the pre-experimental period.

Values of DI₁₄obtained in the current work were comparable to those obtained in previous studies.^6,43 This index was signifi- cantly affected by the type of concentrate (Table 6), showing higher values in the CB group relative to the BP one (0.100 vs.

0.089;P≤0.01). These results reflect a general higher activity of the enzyme in pasture-fed cows receiving starchy relative to di- gestible fibre supplementation. It is known that the metabolism of both insulin and non-structural carbohydrates (NSCs) is involved in the regulation of stearoyl-CoA desaturation, both parame- ters being able to increase SCD activity in rat liver.⁴⁴ Cows supplemented with the CB concentrate ingested higher levels

1261

Table 5. Descriptive statistics for CLA isomers (mg g⁻¹milk fat) in milk fat from cows fed pasture and supplemented with corn –barley (CB,n=55) or beet pulp-based (BP,n=51) concentrates^†

Isomers Treatment group Mean Median SD Min Max

%^‡of total FAs

% of total CLAs

% of totalc/t ort/tgroup cis/transisomers

C18 : 2c9t11 All cows 13.06 12.87 3.88 5.62 23.71 1.49 85.69 91.37

C18 : 2t11c13 CB 0.50 0.48 0.19 0.20 1.02 0.06 3.24 3.46

BP 0.64 0.55 0.28 0.31 1.75 0.07 4.34 4.64

C18 : 2t7c9 CB 0.39 0.40 0.08 0.14 0.73 0.05 2.67 2.85

BP 0.34 0.34 0.06 0.25 0.48 0.04 2.39 2.55

C18 : 2t8c10 All cows 0.17 0.16 0.04 0.09 0.31 0.02 1.12 1.19

C18 : 2c/t12,14 All cows 0.05 0.05 0.02 0.02 0.11 0.01 0.35 0.38

C18 : 2c11t13 All cows 0.03 0.03 0.01 0.01 0.07 <0.01 0.20 0.21

C18 : 2t10c12 CB 0.02 0.02 0.01 0.01 0.05 <0.01 0.14 0.15

BP 0.01 0.01 0.01 0.01 0.03 <0.01 0.08 0.08

cis/trans All cows 14.25 13.89 4.10 6.54 25.15 1.63 93.76 100

trans/transisomers

C18 : 2t11t13 All cows 0.40 0.39 0.12 0.21 0.75 0.05 2.77 44.30

C18 : 2t12t14 All cows 0.18 0.17 0.04 0.11 0.30 0.02 1.24 19.85

C18 : 2t9t11 CB 0.16 0.15 0.03 0.10 0.23 0.02 1.07 17.65

BP 0.18 0.18 0.03 0.11 0.26 0.02 1.27 20.16

C18 : 2t7t9 CB 0.08 0.08 0.02 0.04 0.14 0.01 0.55 8.95

BP 0.10 0.10 0.02 0.04 0.15 0.01 0.68 10.73

C18 : 2t10t12 CB 0.04 0.04 0.01 0.02 0.06 <0.01 0.27 4.45

BP 0.02 0.02 <0.01 0.02 0.03 <0.01 0.16 2.61

C18 : 2t6t8 All cows 0.02 0.02 0.01 <0.01 0.04 <0.01 0.13 1.99

C18 : 2t8t10 All cows 0.01 0.01 0.01 <0.01 0.08 <0.01 0.10 1.64

trans/trans All cows 0.90 0.87 0.18 0.57 1.37 0.10 6.24 100

CLA All cows 15.15 14.78 4.18 7.22 26.19 1.73 100 –

†For variables that were not significantly affected by the concentrate type, descriptive statistics refer to the entire dataset (n=106).

‡All percentages refer to the average values of the variable.

CLA, conjugated linoleic acid;c,cis;t,trans.

of NSC which could have directly and indirectly (by increas- ing circulating insulin levels) determined an enhancement in 9-desaturase activity relative to cows supplemented with BP.

Substantial seasonal variation of the C14:1c9/C14 : 0 ratio has been reported in dairy cows and mainly attributed to changes in the animals’ diet. Higher values of this ratio were observed in summer months with cows fed fresh grass-based diets relative to winter months with cows fed MR diets.⁴² This finding is also confirmed in the current study; the absolute lowest values of DI₁₄ were detected in the pre-experimental period and at the end of the grazing season (November), when MR substantially contributed to the diet (Fig. 2n).

The statistical analysis showed that DI14did not vary significantly from April to October. Significant differences were only detected between May, August and September (when the absolute highest values for the estimated9-desaturase activity were detected) and March and November (which showed, as previously mentioned, the absolute lowest values of DI14). The lack of statistically significant variations from April to October has probably to be attributed to the fact that the cows were fed the same fresh grass-based diet during this period. Also the inclusion of modest amounts of MR (April) and maize silage (August) in the diet did not substantially modify the 9-desaturase activity.

Effect of the experimental concentrates on the concentration of CLA isomers

CLA c9t11, t7c9and t11c13

In spite of the higher LA content found in cereals relative to beet pulp, RA as well as its direct precursor (TVA) were not significantly affected by the concentrate type (Table 6). When starchy and digestible fibre concentrates are eaten by lactating cows in similar quantities (as occurs in this study), the intake of forages by the cows is expected to be higher with the latter type, due to the fact that highly digestible fibre concentrates normally lead to a lesser substitution rate than starchy ones.³³Although daily intake of grass was not measured in the current study, it is possible to hypothesise a higher intake of forages (and consequently of ALA) in cows supplemented with BP. This could have compensated the lower LA content in BP relative to CB (as well as the lower 9-desaturase activity found in the BP-fed cows), determining the lack of significant differences in the TVA and RA concentrations between the two groups. Also, Cabidduet al.⁴⁵obtained similar results, showing only a weak effect of the concentrate source on milk and cheese RA content from sheep stall-fed with fresh forages and supplemented with beet pulp or corn concentrates at a comparable F : C ratio as that of the current study.

CLA t7c9 showed significantly higher levels in milk fat from cows supplemented with CB relative to cows supplemented with BP (P ≤ 0.001). This isomer is almost exclusively synthesised

1262

0 5 10 15 20 25

0 20 40 60

C18:1 t10-11 (mg g⁻¹ fat) C18:2 c9t11 (mg g−1 fat)

pre−experimental period

(a)

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35

0 5 10 15 20 25

C18:2 c9t11 (mg g⁻¹ fat) C18:2 t8c10 (mg g−1 fat)

pre−experimental period

(b)

Figure 1.(a) Regression between the concentrations of C18 : 2c9t11 and C18 : 1t10–11^†‡; and (b) correlation between the concentrations of C18 : 2c9t11 and C18 : 2t8c10^§in milk fat from cows fed pasture and supplemented with corn –barley (CB,

•

nis the number of samples.^†Peaks of C18 : 1t11 (TVA) and C18 : 1t10 were not separated in the chromatogram. As the concentration of C18 : 1t10 is normally negligible with respect to that of C18 : 1t11 in milk fat, their sum is referred in the text as TVA.^‡Regression equations:

CB group (dotted line), C18 : 2c9t11=0.4057×C18:1t10–11−0.0081;R²=0.81⁺⁺⁺

BP group (continuum line), C18 : 2c9t11=0.4084×C18:1t10–11−0.0487;R²=0.79⁺⁺⁺.

§Correlation coefficient:r=0.84;P≤0.001.

endogenously from C18 : 1 t7 by the9-desaturase enzyme in the mammary gland.⁵A higher9-desaturase activity has been estimated in the CB-supplemented cows, thus partly explaining the higher CLAt7c9 concentration found in milk fat in this group.

In addition, the FA composition of the two concentrates also played a key role in determining the observed highly significant influence of the concentrate type on CLAt7c9. Indeed, on the basis of the current knowledge it is well known that C18 : 1t7 is synthesised in the rumen, deriving from isomerisation of oleic acid (OA)^46,47as well as of othercis- ortrans-octadecenoic acids formed during the biohydrogenation of dietary unsaturated FAs.^47,48Even if the chromatographic conditions did not allow the complete separation of the t6–t7 and t8 octadecenoic acids, their sum was found significantly higher in the CB group relative to the BP one (P≤ 0.001). Compared to beet pulp, cereals contained about six-fold higher concentrations of both OA and LA (Table 3) that had probably determined a higher formation of C18 : 1 t7 in the rumen, thus determining also a higher absorption and availability of thistrans-octadecenoic acid in the mammary gland as a substrate for9-desaturase activity. Such considerations are also consistent with the statistically significant correlations found by Collombet al.³¹between the content of CLAt7c9 in milk fat and the content of both oleic and linoleic (C18 : 2c9c12) acids in the diet.

Milk fat from cows that had been allocated to the BP group showed significantly higher levels of the isomert11c13 (P≤0.05).

Previous findings showed a statistically significant correlation between dietary ALA and CLAt11c13.³¹The metabolic pathway of the biohydrogenation of ALA starts with an isomerisation to a conjugated triene (C18 : 3 c9t11c15) followed by a reduction that leads to the formation of C18 : 2t11c15.⁴⁹ Finally, another isomerisation could be responsible for the formation of CLA t11c13.³⁹ The higher concentration of this isomer found in the BP-supplemented cows also reflected the higher content of the combined C18 : 2t11c15 and C18 : 2t9c12 FA found in the same group (P≤0.05). C18 : 2t11c15 is the major C18 : 2transisomer produced in the rumen during the biohydrogenation of ALA⁵⁰ and we attributed the latter difference between dietary groups to this FA. The differences between the two groups with respect

to ALA supply from the supplements are likely to be negligible relative to total ALA intake (due to high intakes from forages). The obtained results seem to corroborate the hypothesis of a higher intake of forages in the BP group. Furthermore, thet11c13/t7c9 ratio, which essentially reflects the proportion of fresh forages in the diet,¹⁴ was also found to be significantly higher in cows supplemented with beet pulp relative to cows supplemented with cereals (1.73 vs. 1.29, respectively;P≤0.001).

Minor CLA isomers

CLA isoforms other than c9t11,t7c9 and t11c13 were present only in low (t8c10, t11t13,t12t14 and t9t11) or very low (c/t 12,14,c11t13,t10c12,t7t9,t10t12,t6t8 andt8t10) concentrations in both groups of cows. The isomerst10c12 andt10t12 showed significantly higher levels in the CB group (P≤0.001). The former has received noteworthy attention by researchers due both to its anti-obesity⁵¹and insulin-sensitising⁵²properties and its role in inhibiting mammary synthesis of milk fat (MFD) under typical dietary regimens.^18,53 This isomer is synthesised entirely in the rumen during the biohydrogenation of LA.⁵⁴Other CLA isomers with double bonds located in positions 10,12 have been found to follow a similar biosynthetic pathway.⁵⁵ Highly statistically significant correlations between the content of 10,12 isomers in bovine milk fat and dietary LA have been previously reported.³¹ The higher concentrations oft10c12 andt10t12 CLA isomers found in the CB group are, consequently, not surprising, if considering the higher LA levels found in cereals relative to beet pulp.

It is well known that diets high in starchy concentrates and low in fibre can determine MFD. The appearance of CLAt10c12 in milk fat is considered a marker for MFD and increased concentrations of this isomer are generally associated with marked changes in milk fat yields.⁵³ In our study, the CB-supplemented cows showed approximately doubled levels of t10c12 relative to the cows supplemented with BP, but the concentrations of this isomer were very low (0.02 vs. 0.01 mg g⁻¹ fat, respectively). No statistically significant differences were observed in the milk fat yield between the two groups (Table 4). When severe MFD occurs, thet10c12 levels are normally higher than those observed in this study.⁵³Low

1263

Table 6. Effect of the type of concentrate (corn –barley or beet pulp), month and their interaction on the concentrations of CLA isomers and other FAs related to CLA metabolism (mg g⁻¹fat) in milk fat

Significance^†

Isomers Corn –barley,n=55 Beet pulp,n=51 SEM CT M CT×M

cis/transCLA isomers

C18 : 2c9t11 13.38 12.57 0.65 NS ^∗∗∗ NS

C18 : 2t11c13 0.50 0.59 0.04 ^{∗ ∗∗∗} NS

C18 : 2t7c9 0.40 0.35 0.01 ^{∗∗∗ ∗∗∗} NS

C18 : 2t8c10 0.17 0.17 0.01 NS ^∗∗∗ 0.092

C18 : 2c/t12,14 0.05 0.05 0.00 NS ^∗∗ NS

C18 : 2c11t13 0.03 0.03 0.00 NS ^∗∗∗ NS

C18 : 2t10c12 0.02 0.01 0.00 ^{∗∗∗ ∗ ∗}

cis/trans 14.54 13.77 0.68 NS ^∗∗∗ NS

trans/transCLA isomers

C18 : 2t11t13 0.42 0.40 0.02 NS ^∗∗∗ NS

C18 : 2t12t14 0.18 0.18 0.01 NS ^∗∗∗ NS

C18 : 2t9t11 0.16 0.18 0.01 ^∗∗∗ 0.067 NS

C18 : 2t7t9 0.08 0.09 0.00 ^{∗∗ ∗} NS

C18 : 2t10t12 0.04 0.03 0.00 ^{∗∗∗ ∗∗∗ ∗∗∗}

C18 : 2t6t8 0.02 0.02 0.00 NS ^∗∗∗ NS

C18 : 2t8t10 0.01 0.01 0.00 NS NS NS

trans/trans 0.90 0.90 0.03 NS ^∗∗∗ NS

CLA 14.86 14.00 0.07 NS ^∗∗∗ NS

C18 : 2t11c13/C18 : 2t7c9 1.29 1.73 0.11 ^{∗∗∗ ∗∗∗} NS

Other FAs related to CLA metabolism

C18 : 1t6-8 1.51 1.27 0.06 ^{∗∗∗ ∗∗} NS

C18 : 1t10-11 32.60 31.75 1.41 NS ^∗∗∗ NS

C18 : 1c9 164.47 158.50 4.19 NS ^∗∗ NS

C18 : 2c9c12 11.23 10.50 0.30 ^{∗ ∗∗∗} NS

C18 : 2t11c15+t9c12 4.00 4.52 0.23 ^{∗ ∗} NS

C18 : 3c6c9c12 0.17 0.16 0.01 NS ^∗∗∗ NS

C18 : 3c9c12c15 7.81 7.76 0.24 NS ^∗∗∗ NS

9-desaturase activity (DI14)^‡ 0.100 0.089 0.004 ^{∗∗ ∗} NS

†Probability:^∗P≤0.05;^∗∗P≤0.01;^∗∗∗P≤0.001; thePvalue is shown if, not being significant, it shows a tendency 0.05<P<0.10; NS, not significant (P≥0.10).

‡Calculated as C14 : 1c9/C14 : 0.

CT, concentrate type;M, month; CT×M, interaction between concentrate type and month;n, number of samples; CLA, conjugated linoleic acid; DI, desaturase index;t,trans;c,cisFAs, fatty acids.

supplementations of starchy concentrates do not seem to enhance the phenomenon of MFD in pasture-fed ruminants. However, the increased CLAt10c12 concentration found in the CB group lead to the hypothesis that decreased F : C ratios (with increased levels of starchy concentrates) could cause the reduction of milk fat secretion in lactating cows at pasture.

The slightly but significantly higher concentrations of t9t11 (P ≤ 0.001) and t7t9 (P ≤ 0.01) isomers found in milk from the BP group are, indeed, unexpected and difficult to explain as both isomers had been previously reported to be positively correlated with dietary LA,³¹ which was actually greater in the CB supplement. Other minor isomers that had earlier been found to be positively correlated with LA (t8t10 and t8c10) or ALA (t11t13,t12t14,c/t 12,14 andc11t13) did not show significant differences between the two groups. Biohydrogenation pathways that lead to the synthesis of these minor CLA isomers in the rumen are not completely understood yet. It is possible to hypothesise that some modifications of the ruminal environment caused by concentrate type could have affected specific ruminal bacteria

that are involved, through their isomerases and reductases, in the biosynthesis of minor CLA isomers.

Monthly variations of the concentrations of CLA isomers RA showed up to 2.4-fold higher values during the experimental period relative to the pre-experimental period, confirming earlier findings that higher intakes of fresh grass increase the RA content in ruminant milk.^12,13,56The same trend was also observed for the concentrations of other minor isomers (t8c10,t12t14,c/t12,14, t11t13,t11c13,c11t13 andt7c9) (Fig. 2). Such results are consistent with the previous findings by Butleret al.,¹²Collombet al.¹⁴and Kraftet al.³⁹who found a strong association with high proportions of fresh forage in the diet and the concentrations of many CLA isomers in bovine milk fat.

Almost all CLA isomers were significantly affected by the sampling month and extensively fluctuated throughout the grazing season, with the exception oft9t11 andt8t10, the latter showing the lowest total variability (Table 6).

1264

0 2 4 6 8 10 12 14 16 18 20 22

Mar Apr May Jun Jul Aug Sep Oct Nov C18:2 c9t11 (mg g−1 fat)

d

a

bc bc cd a

ab

cd a

(a)

0.00 0.20 0.40 0.60 0.80 1.00 (c) 1.20

bc

a a

b ab

c

b ab

c C18:2 t11c13 (mg g−1 fat)

Mar Apr May Jun Jul Aug Sep Oct Nov

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 (e)

d bc

a

bc bc c

a ab

ab

C18:2 t11t13 (mg g−1 fat)

Mar Apr May Jun Jul Aug Sep Oct Nov C18:2 t7c9 (mg g−1 fat)

0.00 0.10 0.20 0.30 0.40 0.50 0.60

d ab

ab c

bc a

ab ab (b)

Mar Apr May Jun Jul Aug Sep Oct Nov c

0.00 0.05 0.10 0.15 0.20 0.25 0.30 (f)

d

ab a

c bc

a a

bcd

C18:2 t8c10 (mg g−1 fat)

Mar Apr May Jun Jul Aug Sep Oct Nov bc

0.00 0.01 0.02 0.03 0.04 0.05

d

ab a bc

cd

ab

bc bcd

C18:2 c11t13 (mg g−1 fat)

Mar Apr May Jun Jul Aug Sep Oct Nov cd

(d)

C18:2 t12t14 (mg g−1 fat) (g)

d

bc bc bc c

a bc

ab c

0.00 0.05 0.10 0.15 0.20 0.25 0.30

Mar Apr May Jun Jul Aug Sep Oct Nov

0.00 0.02 0.04 0.06 0.08 0.10

C18:2 c/t 12,14 (mg g−1 fat) (h)

c

b b b

b b

a b

a

Mar Apr May Jun Jul Aug Sep Oct Nov

Figure 2. Monthly variations of the concentrations of individual CLA isomers, TVA^†and estimated9-desaturase activity (DI14) in milk fat from cows fed pasture and supplemented with corn –barley (CB,

•