Natural'spreadable butters, i. e. those produced by cows'milk without any artificial processing, are known. However, such butters having the ability to spread straight from the fridge'often lack taste, texture and may include unnatural ingredients'.

Meanwhile, standard'butter which has superior flavour, generally lacks spreadability. The spreadability of butter is generally dependent on milk fat content.

The physical and chemical properties of milk fat are largely determined by its chemical composition.

Milk fat composition can be modified through manipulation of the diet of the dairy cow.

Manipulation of the fatty acids present in the milk can lead to a more spreadable'butter by increasing the levels of C181 and C182.

Meanwhile, it is recently been suggested that increase in the human diet of conjugated fatty acids may have anti-carcinogenic properties, and may be usable in the treatment of heart diseases. Some conjugated fatty acids, e. g. conjugated linoleic acid (CLA), are known to be contained in cows'milk.

One object of the present invention is to provide the benefit of spreadable butter and increased CLA to humans through cows'milk.

Thus, according to one aspect of the present invention, there is provided a feed product for cows having final oils content in the range 13-18gm per 10Ogm.

Preferably, the feed has a final oils content in the range 14-16gm per 100gm of feed.

In a second aspect of the present invention, there is provided a feed product for cows wherein the oil content of the feed is over 55% mono-unsaturated fats.

Preferably, the oil content comprises approximately 60% (5*-.) mono-unsaturated fats.

Such an oils content or mono-unsaturated content is achievable using certain high-oil constituents such as rapeseed, naked oats, canola and high oleic sunflower oil. Such constituents typically have the following oil types (g per 100g fat). Saturated Mono-Oleic Acid Poly-unsat unsaturated Rapeseed 7. 0 62. 9 61. 7 30. 2 Naked Oats 41. 0 38. 0 Canola 6. 0 58. 4 56. 1 30. 2 High Oleic 9. 0 80. 0 75. 0 9. 0 Sunflower In one preferred embodiment of the present invention, rapeseed is used. More preferably, the rapeseed is cracked in the feed, but nevertheless whole oilseed is used.

One example of a feed according to the present invention comprises 180g of protein, 160g of oil and 80g crude fibre, per kg of fresh weight.

According to a third embodiment of the present invention, there is provided a method of increasing the C181, C182 and/or conjugated linoleic acid (CLA) content in a cow's milk, comprising feeding to a lactating cow a feed product as described above, based on the feed level of 3-5kg/cow/day.

According to a fourth aspect of the present invention, there is provided a milk product obtained

from a lactating cow fed with a feed as hereinbefore defined, and/or obtainable from the method hereinbefore described.

According to a fifth aspect of the present invention, there is provided a spreadable butter product whenever prepared from a milk as hereinbefore defined.

The present invention extends to a method of producing a spreadable butter product comprising churning a milk product as hereinbefore defined.

More generally, the present invention provides a spreadable butter product having an iodine value in the range 41-49 gI2/100g milk fat (IV). Iodine values can be tested using the (WIJS) British Standard 684 Part 2,2.13 1996. This provides an iodometric determination of the action of iodine on fat.

Preferably, the spreadable product has an iodine value in the range 43-47 IV.

The spreadable butter product of the present invention can be separately defined as having a solid fat content in the range 42-48% of total fat at 5°C.

The measurement can be determined using NMR and the method of British Standard 684 Part 2,2.22 (Bruker/QUB). In this test, oils are conditioned at

60°C, followed by 0°C for 30 minutes and 5°C for 30 minutes prior to the test in an NMR spectrometer, e. g. a minispec 120 apparatus.

More preferably, the spreadable butter product has a solid fat content in the range 43-46% of total fat.

The present invention provides a spreadable butter which is naturally'produced from cow's milk, rather than requiring the addition of additives or significant re-working to get better spreadability'. Nevertheless, the butter product of the present invention may still include additives well known in the art, and also be re-worked, if desired or necessary.

Embodiments of the present invention will now be described by way of example only and with reference to the accompanying Tables 1 to 8 and Figures 1 to 7 as labelled.

Experimental A total of 64 Holstein-Friesian dairy cows were used in this study. Thirty two cows (16 first lactation and 16 multiparous cows) were in mid lactation (average calving date 12 January 16.2 days) with thirty two (12 first lactation and 20 multiparous cows) in late lactation (average calving date 3 October 58.9 days) i. e. mid lactation and late lactation cows were in milk for 130 16.2 and 231 + 58.9 days respectively prior to commencement of the trial. Within stage of lactation groups the 32 cows

were blocked in groups of 4 on the basis of calving date, parity and milk yield, liveweight and condition score in the week prior to commencement of the study, and allocated to 4 concentrate treatments at random.

The experiment commenced on 22 May and continued for 20 weeks to 9 October. Prior to commencement of the trial all cows had previously been turned out to pasture in late March and from 10 April they rotationally grazed perennial ryegrass swards in a 21 day rotation, with cows moved to a new paddock daily following afternoon milking. All cows received 2 kg/day of a standard dairy concentrate containing 180g crude protein/kb fresh weight from turnout to the start of the trial.

Treatments Two concentrates were manufactured to contain either low (LO) or 200g oil/kg fresh concentrate (high oil, HO) through inclusion of full fat rapeseed. Both concentrates were formulated to contain 180 g crude protein/kg fresh weight and contained barley, wheat, maize and sugar beet pulp as the principal energy sources with soya bean meal as the primary protein source.

The four treatments involved Added Oil level (g/cow/day) Treatment 1 3kg day of LO (0) Treatment 2 2kg/day of LO + (200) lkg/day of HO Treatment 3 lkg/day of LO + (400) 2kg/day of HO Treatment 4 3kg day of HO (600) All 64 cows rotationally grazed together as a single group with cows moved to a new paddock daily following afternoon milking and concentrates offered in two equal feeds per day during milking. As a result of grass shortage and poor weather conditions a buffer feed of grass silage was offered once daily from 11 September (week 16) to the end of the trial.

Pre-trial Assessment Milk yields were recorded daily for all cows for two weeks prior to commencement of the trial, when all animals were on a standard diet. Furthermore, in weeks-1 and-2, a 3 day composite milk sample was retained for each cow for analysis of fat, protein and lactose concentrations. In week-1, a 2 day composite sample was retained for each cow for fatty acid analysis. For each treatment group if cows, during week-1 a 1 litre milk sample was retained for assessment of iodine value and a 20 litre bulk sample retained for butter processing and subsequent assessment (see chemical assessment later). All

cows were weighed and condition scored in weeks-1 and-2.

Trial Assessments Milk yields recorded daily for all cows throughout the 20 weeks of the trial. Each week a 3-day composite milk sample was retained from each animal and analysed for fat, protein and lactose concentration. A 2-day composite milk sample was also retained from each animal every 2 weeks and analysed for fatty acid composition. Each week, a 1-day bulked composite sample was retained for each treatment group with samples analysed for iodine value. Similarly, every 2 weeks, a 1-day bulked composite sample (20 litres) for each treatment group was retained for butter processing and subsequent assessment.

Cows were weighed once weekly and condition scored fortnightly throughout the trial. All cows were also blood sampled at 4-week intervals (via tail vein) with samples analysed for urea, hydroxybutyrate, glucose, total protein, albumin and globulin. In addition all health events were recorded during the trial including incidence of lameness, mastitis, digestive and reproductive problems.

Feed Assessments Twice weekly, samples of fresh herbage, representative of that being consumed by the cows, were obtained with fresh samples analysed for oil content and fatty acid composition. A sub sample of fresh herbage was also oven-dried and used to determine oven DM, crude protein, water soluble carbohydrate acid detergent fibre and ash concentration.

Samples of each concentrate type were obtained daily, bulked over each week and analysed for oven DM, acid detergent fibre, ash, crude protein, starch, oil and fatty acid composition.

Chemical Assessments Individual milk samples (200 cm3) were collected from each cow every fortnight during feeding trial for fatty acid analysis. Bulked milk samples (20 1) were collected during the same week from each treatment group for separation and churning into butter. Cone penetrometry measurements were made on the butters and, after clarification, fatty acid analysis, NMR and oxidative stability tests were carried out using the clarified milk fats.

Fatty acid profiles were obtained for milk samples from individual animals and bulk butter samples every second week during the trial period. Milk fat was extracted from liquid milk using chloroform/methanol (2: 1 v/v), whilst bulk butter samples were clarified (BS 769: 1961). Fatty acid

methyl esters (FAMEs) were prepared by the cold methylation method to increase volatity of the acids.

Separation and quantification of the milk fat fatty acids was carried out using a capillary column (0.32 mm i. d., 100m length), WCOT fused silica coated CP Sil 88 (Chrompack, London) in a Varian Star 3400 gas chromatograph (Varian Associates Ltd., Walton-on- Thames, Surrey) with a Star Software Integration system. The chromatograph was equipped with the Varian 8200 autosampler, septa programmed injector and FID. A sample aliquot of 0. 1y1 was injected onto the column at 50°C and the oven temperature ramped during heating to 225°C to improve resolution and separation. Internal and external standards were used for identification and recovery efficiency purposes. Results were expressed as g per kg of milk fat.

Percent solid fat contents of the milk fats from each treatment group were measured using pulsed nuclear magnetic resonance (NMR) in a Bruker Minispec PC 120 series NMR Analyser (Bruker Spectrospin Ltd., Coventry).

For each measurement, approximately 4g of clarified milk fat was transferred onto a glass NMR tube, melted at 60°C for 30 minutes to destroy previous thermal history and crystalised at 0°C for 9G minutes. Parallel measurements of solid fat content were then made in duplicate at 5, 10,15,20 and

25°C, following tempering of the milk fat for 30 minutes at the measuring temperature. Results were expressed as % solid fat content.

Oxidative stability of the milk fats was assessed using the Schaal Oven Test to accelerate oxidation and peroxide value determinations.

Clarified milk fats (1000cm') were placed in clean glass beakers in laboratory ovens set at 63°C (+/- 1°C). Samples for analysis were removed weekly initially and more frequently as required drying the storage period (20-40 days). Peroxide value determinations were carried out by the AOCS official Method Cd 8-53 (AOCS, 1990), with results expressed as milliequivalents of peroxide oxygen per kg of sample (mEq kg-1).

Statistical Analysis Animal performance data were analysed as a completely randomised 8 treatment study with 8 replicates per treatment, and compared using the t- test. Data for milk yield, milk composition, liveweight and condition score were adjusted using data recorded in the two week pre-experimental period as covariates. The statistical significant of the results of chemical analysis of milk components was assessed by analysis of variance.

Gompertz curves were fitted to the NMR and oxidative stability data. The Gompertz curve is an a asymmetrical sigmoidal curve which has the form;

y=a+c*exp (-exp [-b (t-m)]) Equation 1 For the NMR data, y is the solid fat content at temperature t, and a, c, b and m are parameters to be estimated. For the oxidative stability data, y is the peroxide value at time, t. The conclusion of the induction phase of oxidation was identified by drawing a line tangential to the point of inflexion to cut the lower asymptote.

Results Animal health was generally very satisfactory for all cows through the trial.

The low and high-oil concentrates had similar acid- detergent fibre, ash and crude protein concentrations but differed markedly in oil content.

There were small variations in oil content through the experiment (Figure 2) particularly with the high oil concentrate. Starch concentration was also lower in the high-oil concentration reflecting the high inclusion rate of whole rape-seed. Using the actual oil and dry matter contents of the two concentrates, oil intakes for the four concentrate treatments were; 151; 305; 459 and 613 g/day for treatments 1, 2,3 and 4 respectively, indicating levels of added oil from whole rapeseed oil of 154, 308 and 462 g/day for treatments 2,3 and 4 respectively.

Treatment effects on animal performance are presented in Tables 5 to 8. Small increases in milk

yield with increases level of oil inclusion in the concentrate were observed with both mid and late lactation cows but these effects were not statistically significant (P>0.05). Changes in milk yield with time across the four oil levels are presented in Figures 3 and 4. A proportion of the late lactation cows were dried off from week 16 onwards and hence milk yield data are only presented for the first 16 weeks of the trial for this group.

Milk yields declined with both groups throughout the summer period, but the results do suggest a trend towards higher levels of performance at the higher oil levels as the season progressed, particularly with the mid lactation cows.

Increasing the oil concentrate significantly (P<0.01) depressed milk fat content with both mid and late lactation cows, with a mean depression of 0.40% at the highest level of inclusion. Milk protein content tended to decrease with increased level of oil in the concentrate with both mid and late lactation cows, although this effect was not statistically significant and much less evident that the decrease in fat content. There were no significant treatment effects on milk lactose content.

The iodine value of milk (Tables 7 and 8), based on group samples, increased with increasing level of oil inclusion. The changes in iodine value with time are presented in Figures 5 and 6. Iodine values increased rapidly with the high oil

concentrates and treatment effects were largely established within two weeks of the start of the trial. The data presented in Figures 5 and 6 indicate considerable fluctuation in iodine value through the 20 weeks of the trial. However, seasonal effects were consistent across treatments, suggesting that these effects were related to a common factor e. g. changes in grass composition or differences in herbage intake which may be mediated through changes in prevailing weather conditions.

For example the low iodine values on all treatments in week 15 may be related to low herbage intakes as a result of heavy rainfall at this time. Silage feeding was introduced in week 16 and iodine values recovered to previous levels. Peak iodine values were recorded in week 7 (week commencing 3 July) of the trial and tended to decline throughout the rest of the summer period.

There were no significant treatment effects on cow liveweight or liveweight change with either mid of late lactation cows, with all cows gaining liveweight throughout the trial.

Discussion The chemical composition of grazed herbage during the trial was in accord with normal values for high quality grazed swards, with the exception of water soluble carbohydrate concentrations, which were below normal as a result of dull, wet conditions during the latter weeks of the trial (see Figure lc). The low-and high-oil concentrates contained

almost identical crude protein concentrations, but differed markedly in oil content as planned. The lower starch and higher acid detergent fibre content of the high oil concentrates reflects the high inclusion rate of whole rapeseed (40% inclusion rate). These changes in concentrate composition would be expected to produce milk of marginally lower milk protein content, given the strong positive correlation between starch content of the concentrate and milk protein content.

Increasing the level of oil in the concentrate consistently increased the iodine value of milk fat, indicative of increases in the levels of polyunsaturated fatty acids. Treatment effects were established very rapidly i. e. approximately 64% of the full treatment effects obtained at week 6 were achieved within two weeks of commencing the trial.

Whilst there were considerable seasonal fluctuations in iodine value through the trials with all 4 treatments, there was no evidence of a diminishing response in iodine value to additional oil with prolonged feeding of high oil diets. The seasonal fluctuations appear to be related to changes in grass intake or herbage composition. Overall treatment effects on iodine value are summarised in Figure 7. These data indicate a consistent linear response in iodine value to added oil, best described by the relationships; Mid lactation cows IV = 36.35 + 0.023 X Ruz=0. 96 Equation 7

Late lactation cows IV = 37.74 + 01021 X Ruz=0. 98 Equation 8 Where IV represents iodine value of milk fat (g I2/100g milk fat) and X represents level of added oil (g/cow/day).

Iodine values of milk were consistently higher for late lactation cows, but the response in iodine value to added oil was similar, with mean responses of 2.3 and 2.1 g I2/100g milk fat per 100g added oil for mid and late lactation cows respectively.

In conclusion 1. Animal health was highly satisfactory throughout the trial, even at the highest level of added oil (462 g/cow/day). Indeed, the small increases in milk yield with increases oil level with mid lactation cows suggests an improvement in the energy status of animals on these treatments.

2. Effects of added oil in the diet on milk composition and iodine value were consistent throughout the grazing season. Full treatment responses were obtained within 6 weeks of introducing the high-oil concentrates and responses were maintained throughout the remaining 14 weeks of trial.

3. With the exception of milk yield, responses in animal performance to added oil were similar with both mid-and late-lactation cows.

4. Increasing the level of oil in the diet up to 462g added oil/cow/day, through inclusion of full fat rape seed in the concentrate, resulted in linear responses in milk fat and protein concentrations and iodine value, as summarised in Table 9.

Table 9 Summary of responses in milk yield, milk composition and iodine value for 100g added oil to the diet. Mid-lactation Late-lactation Milkyield +0. 32 0 (kg/cow/day) Milk fat (%)-0. 088-0.079 Milk protein (%)-0. 034-0.036 Iodine value +2. 3 +2.1 (gI2/100g milk fat) This data can be used to determine the level of added oil required to achieve a specific change in iodine value, and the consequences on milk production and composition.

For example, assuming the objective is to increase iodine value in milk by 10 g I2/100g milk fat, then for mid-lactation cows the level of added oil required is 435g and this will result in an increase

in milk yield of 1.39 kg/cow/day, with decreases in milk fat and protein concentrations of 0.38 and 0.15t respectively. This calculation assumes a concentrate feed level of 3kg/dow/day.