Milk fatty acid profile from grazing buffaloes fed a blend of soybean and linseed oils

The aim of the study was to examine the changes in milk fatty acid (FA) profile of grazing buffaloes fed either low (L, 276g/d) or high (H, 572g/d) doses of a blend (70:30, wt/wt) of soybean and linseed oils. Fourteen multiparous Mediterranean buffaloes grazing on a native pasture were fed 4 kg/day of a commercial concentrate containing no supplemental oil over a pre-experimental period of ten days. The baseline milk production and composition and m i l k FA profile were measured over the last three days. After this pre-experimental period the animals received the same concentrate added with either the L or H oil doses for 26 additional days. Milk yield (g/animal/day) did not differ at the start (1776 ± 522 and 1662 ± 291 for L and H, respectively, P<0.622) or at the end of the t r i a l (4590 ± 991 and 4847 ± 447 in L and H, respectively, P<0.543). Baseline milk fat content (g/kg) averaged 77.1 (±20.5) in L and 74.3 (±9.9) in H (P<0.10) and was reduced (P<0.031) to 60.7 (±23.6) and 49.4 (±11.2) (P<0.0031) respectively after L and H with no differences between treatments (P<0.277). Baseline milk protein content (L=43.2 ± 3.4 and H= 44.3 ± 6.9g/kg) increased after oil supplementation (P<0.0001) in both L (73.2 ± 6.0g/kg) and H (68.4 ± 4.9g/kg) without differences between oil doses (P<0.123). Milk fat content of 14:0 decreased after oil supplementation only in the H treatment (5.29 to 4.03, P<0.007) whereas that of 16:0 was reduced (P<0.001) at both L (24.49 to 19.75g/100g FA) and H (25.92 to 19.17g/100g FA) doses. The reduction of total content o f 12:0 to 16:0 was higher (P<0.052) in H (32.02 to 23.93g/100g FA) than L (30.17 to 25.45g/100g FA). Vaccenic acid content increased (P<0.001) from 5.70 to 13.24g/100g FA in L and from 5.25 to 16.77 in H, with higher results in the in H treatment (P<0.001). Baseline rumenic acid was sharply increased (P<0.001) in L (1.80 to 4.09g/100g FA, +127%) and H (1.60 to 4.61g/100g FA, +187%) with no differences between L and H (P<0.19). Overall, these results indicate a pronounced improvement in the nutritional value of milk fat from grazing buffaloes fed little amounts (0.276g/day) of a blend of soybean and linseed oils.


INTRODUCTION
Buffalo milk is a rich source of nutrients with higher levels of total protein, medium chain FA, conjugated linoleic acid (CLA) and contents of retinol and tocopherols than those of cow milk (Ahmad et al., 2013).Higher amounts of saturated fatty acids (SFA) and lower amounts of unsaturated fatty acids (UFA) were also reported in buffalo m i l k c o m p a r e d t o cow milk together with higher contents of myristic (14:0) and palmitic (16:0) FA (Ménard et al., 2010).These FA (14:0 and 16:0) are classed as atherogenic (Ulbritch and Southgate, 1991) if consumed in excess and associated to increased risk of heart disease (Stanton el al., 2003;Chilliard and Ferlay, 2004).The supplementation with polyunsaturated rich oils may decrease the concentration of the potentially atherogenic FA in cows (Rego et al., 2005) and in lactating buffaloes (Oliveira et al., 2009).Concerning the potential functional benefits of milk, a current special interest exists on conjugated linoleic acid (CLA) concentration, because it plays an important role in regulating plasma lipid concentrations and cardiovascular functions, reducing cancer incidence, as well as blocking tumor growth and metastasis from cancer breasts (Parodi, 1999).Between the natural CLA isomers, the cis-9 trans-11 18:2 or rumenic acid (RA) is the most present (80-90%) in milk fat.Vaccenic acid (VA, trans-11 18:1) is the main natural trans FA and precursor of RA in the mammary gland and other tissues.It may also have anticarcinogenic properties and can be metabolized by humans to RA (Stanton et al., 2003).Buffalo milk contains higher amounts of RA and VA than cow milk (Ahmad et al., 2013;Ménard et al., 2010).).Milk fat is considered the main natural source of RA and its concentration in milk is highly dependent on the type of diet and lipid supplementation in dairy cows (Chilliard et al., 2007) and buffaloes (Fernandes et al., 2007;Patiño et al., 2010, Oliveira et al. 2009).To our knowledge the information about the effect of providing a combination of soy and linseed oils on the full milk FA profile of grazing buffaloes is non-existent.The aim of this experiment is to evaluate the effect of supplementation with a blend (70:30 wt/wt) of soybean and linseed oils on the chemical composition of milk and the FA profile in lactating buffaloes in natural grassland.

MATERIALS AND METHODS
The experiment was conducted during the months of October -November 2013 at the commercial farm of Nuestra Señora de Itatí (Route 12, Km 1098, Itatí Department, site Yacareí Province of Corrientes, Argentina) with an 80 hectares field and natural grassland.Fourteen multiparous Mediterranean buffaloes (milked after two years of inactivity) between the second and the third part of their lactation were used (7 with L and 7 with H doses). Experimental procedures and buffalo management were performed according to the recommendations of good management practices (GMP) of the National Sanitarian Service (SENASA, Argentina).As the experiment progressed animals were gradually adapted to the milking routine and oxytocin was used to  ), rice bran (10%), pelleted sunflower (10%), soybean (5%) and wheat meals, bovine fat (11%) and a mineral premix (4%).Ten days prior to the start of the experiment the animals consumed pasture (63%) plus concentrate (27%) without oil to determine the baseline (B) milk composition and FA profile on samples taken over the last three days.After this period the animals were supplemented with a blend (70:30, wt/wt) of soybean and linseed oils at a low (L, 276g/d) and a high (H, 572g/d) doses mixed to the concentrate.Oil represented about 2.21 and 4.42% of total DM intake if a maximum DM intake of 12.5 kg/animal/day is assumed (Kathirvelan and Tyagi, 2009).Oils were manually mixed to the concentrate and thoroughly consumed during milking time in the morning.After milking animals were left in the grazing paddock until the next day.
Animals were milked mechanically once a day at 8am and individual samples (100 ml) were collected.An aliquot of 10mL was used for chemical analysis while the remaining 90mL were immediately frozen (-20) until FA determination by GLC.Before the start of oil supply milk yield was individually recorded and milk samples were obtained during three consecutive days to measure the B milk composition and FA profile.The procedure was repeated over the last 3 days of the experimental period.Milk samples were analyzed for fat, protein and nonfat solids by infrared spectrophotometry (Milk analyzing device Model: Master Eco).The calibration routine is done with reference to a standard pattern using buffalo milk quantified by other methods used values (Fat by the Gerber method; Acidity by the Dornic method; Protein by the Kjeldahl method).Milk FA composition was measured in individual samples from the three days of each period according to guidelines of Bulletin of International Dairy Federation N° 265/1991 ("Determination of free fatty acids in milk and milk products" ISO 15884-IDF 182: Milk fat, preparation of fatty acid methyl esters).

STATISTICAL ANALYSES
Data of milk production and composition and milk FA profile were analyzed using the Student T-test for paired observations.Differences between L and H oil doses were stated using the Student t-test for independent observations (n=7).Differences were considered significant with P<0.05 unless otherwise stated.Values are presented as means followed by standard deviation.

RESULTS
Oil content for linoleic (53.61 and 15.90%) and linolenic (0.36 and 50.39%)FA presented normal values for soybean and linseed oils respectively.At the start of the experiment, milk yield (g/animal/day) averaged 1776 (±522) and 1662 (±291) for L and H respectively (Table 1) without differences between treatment groups (P<0.622).At the end of the trial milk yield averaged 4590 and 4847g/animal/day in L and H (Table 1) without differences between oil doses (P<0.543).
Compared to the basal diet, milk fat content decreased in L and H whereas milk protein content increased after oil supplementation.Concentrations of de novo (4:0 a 15:1) and short chain FA (4:0 to 10:0) were reduced only in the H treatment (Table 2).Milk concentration of 12:0 was not affected, but the concentration of 14:0 decreased (-24%, P<0.007)only in H (Table 2).Concentrations of 16:0 were reduced (P<0.001) in both the L and H treatments (Table 2) without differences between the two doses (Table 2).The hypercholesterolemic fraction of milk fat (12:0 to 16:0) was reduced (P<0.001) in both treatments but the resulting decrease was higher (P<0.052) in H (-8.09g/100g FA) compared to L (-4.72g/100g FA).The atherogenic index of milk was reduced and was lower in H compared to L (P<0.056).The concentration of VA increased (P<0.001) in L and H and had a higher result (P<0.001) in H (16.77g/100g FA) than in L (13.24g/100g FA).Basal RA sharply increased in L (+127%) and H Arq. Bras.Med.Vet.Zootec., v.67, n.3, p.927-934, 2015 (+187%) without differences between treatments (P<0.195).The basal RA/VA ratio did not change after oil supplementation.Basal concentrations of trans-10 18:1 were low and increased by the oil-blend without differences between treatments.Basal concentrations of elaidic (trans-9 18:1) acid in milk increased after oil supplementation without differences between L and H.The basal n-6/n-3 ratio was slightly increased after oil supplementation with lower results in L compared to H doses.

DISCUSSION
The increase in basal milk production after oil supplementation was probably explained by a gradual adaptation of animals to milking and also by the use of oxytocin to promote the release of milk.No effect of lipid dose on milk production was observed.Supplementation with PUFA-rich oils did not affect milk yield in non-grazing (Gagliostro and Chilliard, 1992) or in grazing experiments (Schroeder et al., 2004) with dairy cows.Basal content of total solids in milk (17.6%) was in the normal range (17% to 18.1%) reported by Oliveira et al. (2009) using confined buffaloes and Ahmad et al. (2013) for their review on general composition of buffalo milk.
After oil feeding this parameter increased up to 21.89g/100g probably as the consequence of the higher milk protein content after the intake of oil.
The increase in total solids makes this milk ideal for processing into dairy products and may also contribute to significant energy savings in manufacturing the milk.Before oil feeding milk protein content was in the normal range (4.0 to 5.0%) reported by Oliveira et al. (2009) and Ahmad et al. (2013).After oil intake, this parameter showed a noticeable increase reaching values as high as 6.84 and 7.32%.Milk protein content was not affected when confined buffaloes were supplemented with different lipid sources (Oliveira et al., 2009).As synthesis of milk protein can be limited by energy availability some additional energy absorbed after oil intake could improve energy status of the buffaloes and enhance milk protein synthesis explaining in part the increase in milk protein content.In fact, a possible dilution effect of milk protein was not observed as milk production increased.Average basal values of fat (7.43-7.71%,Table 1) were below the normal concentration (8.3%) reported by Varrichio et al. (2007).The lower milk fat content after oil supplementation (L= -1.64g/100g and H= -2.49g/100g, Table 1) could be explained in part by a dilution effect as milk yield increased (Table 1) and was not reported for confined buffaloes supplemented with soybean oil at a rate of 2.21% of total DM intake (Oliveira et al., 2009).Fat content in milk (8.66 to 9.40g/100g) and in mozzarella cheese (21.05 to 25.27g/100g) were increased (P<0.05) by supplementation with soybean oil (Oliveira et al., 2009).In our trial the lower milk fat concentration could also be explained by the increase (and ulterior transfer to the udder) of some trans-FA formed by ruminal biohydrogenation which have antilipogenic properties (Shingfield et al., 2010).The uptake of trans-10, cis-12 CLA, trans-9, cis-11 CLA,
In Oliveira et al. (2009) feeding soybean oil resulted in an approximate 10% decrease in SFA relative to the control treatment or when oil was supplied as soybean grain.In Oliveira et al. (2009), baseline for the hypercholesterolemic FA (45.8g/100g) was higher than the average observed in our grazing experiment (30.15 to 32.02g/100g, Table 2).Beyond type of diet, milk concentration of 12:0 and 14:0 was also associated with higher body condition of buffaloes (Ahmad et al., 2013) and the difference with Oliveira's data may also be explained by the moderate body condition of our grazing animals.In the study by Fernandes et al. (2007) the hypercholesterolemic FA varied from 32.48 to 42.9%.When considering only the SFA able to raise LDL blood levels (lauric, mirystic and palmitic) feeding the oil-blend has reduced their concentrations with emphasis on H (23.93g/100g) compared to L (25.45g/100g) doses (Table 2).After supplementation with the oil-blend, the concentration of human atherogenic FA was lower than the values of 43g/100g reported by Oliveira et al. (2009).Supplementation with the high dose of oil-blend was the most effective way to reduce the total concentration of hypercholesterolemic FA and the atherogenic index of milk fat (Table 2).
Average baseline records for 18:0 (19.57g/100g) were well above the range (7.86-13.44)reported by Fernandes et al. (2007) for Brazilian buffaloes.Supplementary 18:2 did not enhance the concentration of 18:0 in milk fat (Table 2) suggesting that increased biohydrogenation of 18:2 (and also VA) in the rumen probably did not occur.Feeding soybean oil but not soybean grain reduced the concentration of 18:0 in milk fat of lactating buffaloes (Oliveira et al., 2009).
In our experiment, average basal content of oleic acid (28.04g/100g) was above the range (20.6 to 25.1g/100g) reported by Fernandes et al. (2007) for Brazilian buffaloes resulting also well above than observed in confined buffaloes (20.35g/100g) fed TMR diets (Oliveira et al., 2009) but near the values of 29.47g/100g reported by Qurshi et al. (2010).In confined diets (Oliveira et al., 2009), feeding soybean oil (but not soybean grain) increased oleic acid in milk up to 25g/100g, whereas in our experiment this FA decreased in both L and H oil doses (Table 2).As oils contained about 19% of oleic acid (Table 1) a high rumen hydrogenation of this FA seems probable.Some inhibition by trans isomers of delta-9 desaturase enzyme in the mammary gland (Shingfield et al., 2010) may also explain the observed reduction in oleic acid.In our trial, average basal concentration of RA (1.70g/100g FA) was in the upper range limit (1.02-1.77g/100g)reported by Fernandes et al. (2007) for buffaloes on commercial farms in Brazil although the average values observed were about 1.1g/100g in four of the five tested farms.
Baseline RA (Table 2) was 1.74 times higher than values of 0.98g/100g FA reported by Oliveira et al. (2009).The relative increase in milk RA after oil supplementation (127% and 188%) was higher than the obtained by Oliveira et al. (2009) in buffaloes supplemented with soybean grain (+13.3%) or soybean oil (+102%).The average RA concentration obtained after oil supplementation (4.35g/100g,Table 2) was higher than the total CLA concentration of 1.95g/100g when 2% of mustard oil was added to the diet of Murrah buffaloes (Kathirvelan and Tyagi, 2009).The relative increase in H oil treatment (188%, Table 2) was near to that informed (185%) by Kathirvelan and Tyagi (2009).In their experiment the average total CLA contents (g/100g milk fat) was 0.684, 1.212 and 1.95 when animals were fed groundnut cakebased concentrate, mustard cake-based concentrate and 2% of mustard oil respectively.
Baseline of total trans FA in milk (6.90g/100g,Table 2) was sharply increased (17.72g/100g) after oil supplementation.Intake of trans fatty acids has been associated with an increased risk of human cardiovascular disease but results from epidemiological studies have indicated that the intake of rumenic trans FA is innocuous or even protects against cardiovascular disease (Jakobsen et al., 2008;Bendsen et al., 2011).In the
Vaccenic acid represented 79.3% and 84.5% of total trans in basal or after oil supplementation respectively (Table 2).At about 2.21% of estimated total DMI the oil mix supplied apparently exceeded the desaturation capacity of the mammary gland to convert VA to RA because in doubling the dose an increase in the concentration of VA was obtained without further increases in the content of RA.The presence of VA in dairy products may also have beneficial health properties through its direct anticarcinogenic properties (Awad et al., 1995) or mediated through its endogenous conversion to RA by delta-9 desaturases in human tissues (Parodi, 2003).The conversion of VA to RA has been shown to prevent chemically induced carcinogenesis in rodents (Banni et al., 2001).In spite of the increased availability of linoleic and linolenic acids after oil supplementation its incorporation to milk fat was negligible suggesting a moderate efficiency in the transfer of these FA to milk and a high rate of rumen biohydrogenation in buffaloes.

CONCLUSIONS
In the extensive production conditions of the present work based on natural grassland diets, buffalo milk represented a good source of RA.Supplementation with a mixture (70:30) of soy and linseed oils at 2.21% of estimated DM intake reduced milk fat concentration, increased milk protein and total solid contents and induced profound changes in the milk FA profile.The concentration of the potentially human atherogenic FA and the atherogenic index of milk were attenuated while the RA concentration was sharply increased.This fact should contribute to alleviate the dieticians' criticism against dairy products for their ability to raise LDL cholesterol and to improve the health quality of buffalo milk and its image as perceived by the consumer.

Table 1 .
Milk yield and composition in lactation buffaloes before (Basal) and 23 days after (Final) supplementation with a blend (70:30, wt:wt) of soybean and linseed oils at a low (L, 276g/d) and a high (H, 572g/d) dose.
(1) a, b Student T-test for paired observations.Differences between L and H oil doses were not significant (Student Ttest for independent observations, n=7).