Effects of feeding whole-cracked rapeseeds, nitrate, and 3-nitrooxypropanol on composition and functional properties of the milk fat fraction from Danish Holstein cows

The aim of this study was to determine the individual and combined effects of supplementing fat (FAT), nitrate (NITRATE) and 3-nitrooxypropanol (3-NOP) on compositional and functional properties of milk fat. An 8 × 8 incomplete Latin square design was conducted with 48 lactating Danish Holstein cows over 6 periods of 21 d each. Eight diets were 2 × 2 × 2 factorially arranged: FAT (30 or 63 g crude fat/kg DM), NITRATE (0 or 10 g nitrate/kg DM), and 3-NOP (0 or 80 mg 3-NOP/kg DM) and cows were fed ad libitum . Milk samples were analyzed for general composition, fatty acids (FA) and thermal properties of milk fat. Milk fat content was decreased by FAT and NITRATE and increased by 3-NOP. The changes in FA composition were mainly driven by the FAT × 3-NOP interaction. FAT shifted milk FA composition toward lower content of saturated FA (SFA) and greater contents of mono-and poly-unsaturated FA (MUFA and PUFA), whereas these effects of FAT were smaller in combination with 3-NOP. However, 3-NOP had no effects on SFA, MUFA and PUFA in low fat diets. FAT lowered solid fat content in milk fat because of decreased SFA content. The onset crystallization temperature of milk fat was decreased by 3-NOP when supplemented in low fat diets. According to the FAT × 3-NOP interaction, supplementation of fat without 3-NOP shifted peak temperature of low melting fraction of milk fat toward low temperature as a result of decreased proportion of C16:0, and increased proportions of C18:1 cis -9, C18:1 trans -11, C18:2 cis -9, and CLA cis -9, trans -11. In conclusion, no additive effects were observed among FAT, NITRATE and 3-NOP on chemical and thermal properties of milk fat and fat supplementation largely changed milk FA composition and in turn affected the thermal properties of milk fat.


INTRODUCTION
The global dairy sector faces a challenge of lowering enteric methane (CH 4 ) emissions from dairy cows.Research on feeding strategies has mainly been focused on the potential of inhibiting ruminal methanogenesis, and some of the most promising strategies include supplementation of fat, and addition of feed additives/rumen modifiers such as nitrate and 3-nitrooxypropanol (3-NOP) (Boadi et al., 2004;Honan et al., 2021).The CH 4 -suppressing effect of dietary fat is associated with (1) decreasing the intake of fermentable organic matter, (2) lowering the activity of ruminal methanogens and number of protozoa, and (3) providing an alternative hydrogen sink through biohydrogenation of unsaturated fatty acids (UFA) (Johnson and Johnson, 1995;Giger-Reverdin et al., 2003;Beauchemin et al., 2009).Being an electron acceptor, nitrate competes with methanogenesis and reduces excess hydrogen otherwise available for CH 4 synthesis.Furthermore, nitrate-derived nitrite exerts an inhibition effect on methanogenic archaea (Liu et al., 2017;Honan et al., 2021).The synthetic feed additive, 3-NOP, lowers enteric CH 4 emission through the inhibition of methyl-coenzyme M reductase, a nickel enzyme, which catalyzes the last step of CH 4 synthesis in rumen archaea (Duin et al., 2016).
Ruminal fermentation yields hydrogen and volatile fatty acids, the precursor molecules for the mammary de novo milk FA synthesis.Methanogenesis removes a considerable proportion of hydrogen while biohydrogenation of UFA also uses hydrogen to some extent.In milk fat, approximately 50% of the FA is synthesized de novo, which includes the production of short-and medium-chain FA (SCFA and MCFA).The remaining 50% stems from circulating blood, either feed-derived Effects of feeding whole-cracked rapeseeds, nitrate, and 3-nitrooxypropanol on composition and functional properties of the milk fat fraction from Danish Holstein cows Gayani M. S. Lokuge, 1 * C. Kaysen, 1 M. Maigaard, 2 P. Lund, 2 L. Wiking, 1 and N. A. Poulsen 1 or mobilized.Of the feed-derived FA, some undergo biohydrogenation: mainly: 18: 3 n-3, 18:2 n-6 and 18:1 cis 9. Therefore, changes in milk FA composition by these mitigation strategies are expected (Grummer, 1991;Rego et al., 2009;Bayat et al., 2018).Changes in milk FA profile affect the nutritional value of milk and dairy products e.g., by replacement of saturated FA (SFA) with mono-or poly-unsaturated fatty acids (MUFA or PUFA).Furthermore, FA composition affects physical properties of milk such as solid fat content (SFC), crystallization and melting of milk fat, which in turn influence the production and quality parameters of fat-based dairy products including texture, oxidative stability, spreadability, and mouthfeel (Bobe et al., 2003;Couvreur et al., 2006;Smet et al., 2010).
The supplementation of rapeseed as a fat source to dairy cows has been shown to decrease SFA and increase MUFA and PUFA (Givens et al., 2003;Samková and Kalač, 2021), whereas changes in FA composition by nitrate supplementation remains inconsistent.Almeida et al. (2022) reported a decrease in SFA and an increase in MUFA and PUFA in milk after nitrate supplementation of the diet.Combined use of nitrate and fat (docosahexaenoic acid) showed no interaction effect between fat and nitrate on milk FA composition (Klop et al., 2016).Similarly, inconsistent results of the effect of 3-NOP on milk FA have been reported.Hristov et al. (2015), and Melgar et al. (2020b) reported that 3-NOP increased SFA without affecting UFA content, whereas Melgar et al. (2021) reported a decrease in PUFA in milk from cows fed 3-NOP.However, no studies have investigated the impact of combined use of rapeseed, nitrate and 3-NOP on FA composition in milk.
In a recent study by Maigaard et al. (2023), it was shown that the individual CH 4 mitigating effect of whole-cracked rapeseed, nitrate and 3-NOP were between 6 and 23%, but without any additive effects when fed in combination, the combined use negatively affected production performances.Using the same experimental set-up, the objective of this study was to examine the impacts of adding whole-cracked rapeseed, nitrate and 3-NOP in different combinations on overall milk composition, FA composition, fat globule size, vitamin E content and thermal properties of milk fat.We hypothesized that (1) supplementation of whole-cracked rapeseed would largely decrease SFA and increase UFA in milk fat and possibly mask the minor changes in FA composition from adding nitrate and 3-NOP to the feed, and (2) changes in milk FA composition by wholecracked rapeseed would result in lower SFC and a shift in offset temperature of melting (T-offset) and onset temperature of crystallization (T-onset) toward lower temperature.

Experimental design
The feeding trial was conducted by the Department of Animal and Veterinary Sciences at Aarhus University, AU Viborg -Research Centre Foulum, Denmark as described by Maigaard et al. (2023).In short, a total of 48 lactating Danish Holstein cows were blocked in 6 blocks based on parity (8 cows/block) and DIM.Cows within a block were assigned to one of the 8 experimental diets in each of 6 periods of 21 d each, according to an 8 × 8 incomplete Latin square design.Experimental diets were arranged in a 2 × 2 × 2 factorial arrangement: 2 levels of fat (low fat (LF): 30 g crude fat/kg DM and high fat (HF): 63 g crude fat/kg DM), with or without dietary nitrate (5Ca(NO 3 ) 2 •NH 4 NO 3 •10H 2 O; Silvair®; Cargill Incs) (UREA; 0 g nitrate/kg DM or NIT; 10 g nitrate/kg DM), and with or without 3-NOP (Bovaer®; DSM Nutritional Products) (BLANK; 0 mg/kg DM or NOP; 80 mg/kg DM).Whole cracked rapeseeds were added as fat source.The diets without nitrate were supplemented with urea to maintain isonitrogenous diets and the diets without 3-NOP were mixed with the carrier compound of 3-NOP as a placebo.Cows were fed ad libitum with a partial mixed ration consisting of corn silage, grass-clover silage, soybean meal, rapeseed meal, rolled beet pulp, and NaOH treated wheat.The roughage: concentrate ratio was 50:50.Cows were offered additional concentrate of up to 1.2 kg DM/d as feed bait in the GreenFeed system used for measuring gas emissions.

Milk sampling and analysis of detailed milk composition
Milk samples were collected at d 19-20 of each experimental period.Cows were milked twice daily at 0545 and 1630 h and a representative milk sample from each cow per period was prepared by mixing 4 consecutive afternoon and morning milk samples.All the samples were stored at 4°C until the analyses at Department of Food Science, Aarhus University, Denmark.Samples from individual cows were used for compositional analysis, whereas samples within period and treatment were pooled for the analysis of SFC, crystallization and melting behavior.
Whole milk samples were used to measure fat, protein, casein, lactose, milk urea nitrogen (MUN) and citrate by infrared spectroscopy (Milkoscan FT2, FOSS Analytical).Whole milk samples were skimmed by centrifugation at 2643 × g for 30 min at 4°C.The cream fraction was stored at −20°C until the analysis of FA profile, vitamin E and thermal properties.

Milk fat globule size (MFGS) and fatty acid (FA) composition
MFGS in full milk was determined by laser diffraction using Mastersizer (Malvern Instruments, Malvern) (Wiking et al., 2003).The volume-based diameter, d(4,3) was reported and each sample was analyzed in technical triplicates.
FA composition was analyzed as previously described by Lokuge et al. (2023).Briefly, isolated milk fat from the samples were dissolved in heptane and then mixed with sodium methylate to convert the FA into methyl esters.FA methyl esters (FAME) were separated and identified by a gas chromatography system (7890A, Agilent Technologies) equipped with a flame ionization detector, a Restek RT-2560 capillary column (100 m, 0.25 µm ID, 0.20 µm df), and He as the carrier gas.The injector temperature was maintained at 250°C and used in splitless mode.The detector temperature was held at 300°C.FA were identified based on retention time comparisons with pure methyl ester standards (Supelco 37 component FAME mix, Supelco Inc. C12:1 was used as an internal standard.The contents of FA were calculated as the weight proportion of total identified FA and FA present in more than 0.5 g/100 g of total FA were reported.

Analysis of α-and γ-tocopherol
For α-and γ-tocopherols analysis, the procedure outlined by Havemose et al. (2004) was followed with a few modifications.Shortly, milk samples were mixed in 1% ethanolic ascorbic acid in a 1:1 ratio.Saturated potassium hydroxide solution was added and mixed.For the saponification, the samples were heated in the oven at 70°C for 30 min.Following the cooling, Milli-Q water and heptane to 2-propanol (100:2) were added, mixed and centrifuged at the speed of 1700 × g for 3 min at 4°C.The supernatant was transferred to brown vials and analyzed using a HPLC system (HP 1100, Agilent Technologies) equipped with a fluorescence detector, ZORBAX SB-C8 column (4.6 × 150 mm, 5 µm; Agilent Technologies) and heptane to 2-propanol (100:2) as mobile phase.The external standards of αand γ-tocopherols were used for the quantification.

Solid fat content (SFC) of milk fat
The SFC profiles of milk fat were determined by pNMR (Minispec mq20, Bruker) using the direct method .The liquified milk fat was added into pNMR tubes and held at 65°C for 30 min to obliterate thermal history, followed by fast cooling of the tubes.After 1 h incubation in a water bath at 5°C, SFC was determined with NMR.Then the temperature was increased to 10°C and the SFC was measured again after 1 h of incubation.This procedure was repeated for the following temperatures: 15, 20 and 30°C.

Crystallization of milk fat
The crystallization and melting behavior of milk fat was determined by differential scanning calorimetry (DSC) (Q2000, TA Instruments) in which nitrogen was used to purge the system.Samples of 5-15 mg of melted fat were sealed in hermetic aluminum pans.An empty pan was used as a reference.The applied time-temperature program was as previously explained by Smet et al. (2010) with some modifications.The sample was held at 65°C for 15 min, cooled at the rate of 10°C/min to −40°C and held for 5 min.This was followed by another heating of the sample at a rate of 20°C/min up to 65°C.The samples were run in duplicates, and the crystallization and melting properties were evaluated with Universal Analysis Software (TA Instruments).Onset crystallization temperature (Tonset), offset melting temperatures (T-offset), and peak temperatures of low and high melting fractions (LMF and HMF) were measured from DSC curves, as illustrated in Figure 1.

Statistical analyses
The data on overall milk composition, MFGS, FA composition and vitamin E content were statistically analyzed using a linear mixed model in R 3.6.3(R Core Team, 2019; https: / / www .r-project.org/),under the lmer function from the lme4 package (Bates et al., 2015).The model was: Data on SFC, crystallization and melting behavior was analyzed using a linear model since they were pooled samples, in the form of: where Y ijklmn and Y ijkm are the dependent variables, µ is the overall mean, F i is the fixed effect of FAT (i = LF, HF), N j is the fixed effect of NITRATE (j = UREA, NIT), O k is the fixed effect of 3-NOP (k = BLANK, NOP), M l is the fixed effect of parity (l = primiparous, multiparous), P m is the fixed effect of the period (1 to

Milk fat globule size (MFGS) and fatty acid (FA) composition
Table 2 presents the effects of the treatments on FA composition and MFGS.No interaction effects were found, and FAT or NITRATE had no effect on MFGS.Milk from cows fed NOP diets had a larger MFGS than BLANK diets (3.78 vs 3.71 µm, P < 0.05).A considerable number of 2-way interactions were observed for FA composition, but no 3-way interactions were observed.
No effect of NITRATE or any interactions with NI-TRATE was found on total SFA, MUFA and PUFA.However, FAT × 3-NOP interactions showed that FAT decreased the proportion of SFA and increased the proportions of MUFA and PUFA, however, these effects were smaller when FAT was combined with 3-NOP (P < 0.001) (Figure 2.A, 2.B and 2.C).

5
Standard error of Estimated Marginal Mean.P-values are reported for the selected effects and their interactions.Differences were declared significant if P < 0.05.

Solid fat content (SFC)
The SFC of milk fat from 5 to 30°C is shown in Figure 3. SFC was greatly affected by FAT and a few interaction effects were observed.The FAT × 3-NOP interaction showed that FAT decreased SFC at 5°C and this effect remained consistent when FAT was supplemented with 3-NOP (LF -BLANK: 54 .1,HF -BLANK: 39 .7,LF -NOP: 51 .9, and HF-NOP: 40.6%, P < 0.05).Furthermore, the FAT × 3-NOP interaction on SFC at 30°C revealed that FAT decreased SFC with or without presence of 3-NOP while 3-NOP decreased SFC at 30°C only in LF diets, not in HF diets (P < 0.05).According to the FAT × NITRATE interaction on SFC at 30°C, FAT decreased SFC irrespective of NITRATE whereas NITRATE decreased it only in LF diets (P < 0.05).Milk fat from cows fed HF diets had a lower SFC than LF diets at 10°C (34.6 vs 46.7%) at 15°C (26.5 vs 36.3%) and at 20°C (13.2 vs 19.9%) (P < 0.001 for all).
SFC at 5 and 10°C was lower in milk fat from cows fed NIT diets than UREA diets (45.8 vs 47.3% and 39.9 vs 41.4%, P < 0.05 for both).At 20°C, milk fat from cows fed NOP diets had lower SFC than BLANK diets (15.9 vs 17.2%, P < 0.05).

Crystallization and melting properties
Figure 4 shows the crystallization and melting curves of milk fat from cows fed HF diets and LF diets.According to the data presented in Table 4, T-onset varied from 11.8 to 14.5°C across the treatments.In the FAT × 3-NOP interaction, 3-NOP lowered T-onset in LF diets but not in HF diets (LF -BLANK: 13 .8,HF -BLANK: 12 .5,LF -NOP: 12 .0,and HF -NOP: 12 .4°C,P < 0.05).The melting curves consisted of 2 distinct peaks where melting peak 1 referred to LMF and melting peak 2 referred to HMF (Figure 4).T-offset after HMF and the mean peak temperature of LMF and HMF was reported.When fed separately, FAT, NITRATE or 3-NOP had no effect on T-offset.The FAT × 3-NOP interaction showed FAT decreased the peak temperature of LMF in BLANK diets but not in NOP diets (LF -BLANK: 7 .88,HF -BLANK: 6 .41,LF -NOP: 8 .10 and HF -NOP: 7 .71°C,P < 0.05).No interaction effects were found on peak temperature of HMF.Milk fat from cows fed LF diets had a higher peak temperature in HMP compared with HF diets (28.3 vs 26.4°C, P < 0.01).

DISCUSSION
This study was part of an experimental feeding trial and the results on dry matter intake (DMI), milk production and CH 4 emission are reported in Maigaard et al. (2023).We found reduced milk protein content by HF diets, which probably was due to the dilution effect caused by increased milk yield in cows fed HF diets (Maigaard et al., 2023).These results are in agreement with previous studies (Hellwing et al., 2014;Larsen et al., 2016).The lower milk protein content in cows fed NIT diets is likely a response to lower DMI reported in those cows (Maigaard et al., 2023).A decreased availability of ruminal or post-ruminal N and glucogenic precursors due to reduced DMI could impair milk protein synthesis (Rigout et al., 2003;Almeida et al., 2022).In line with our study, Klop et al. (2016) and van Zijderveld et al. (2011) also reported a lower milk protein content upon nitrate supplementation.Decreased casein content by FAT and NITRATE indicates the positive correlation between milk protein and casein contents.The decreasing effect of HF diets on  MUN may reflect increased milk N efficiency in cows fed HF diets (Maigaard et al., 2023).This effect of FAT on MUN content became smaller when fat was supplemented with 3-NOP.Therefore, it can be suggested that the effect of 3-NOP on MUN depends on the level of fat in the basal diet.In previous studies, 3-NOP has been reported to increase MUN in milk from cows fed doses of 3-NOP from 60 to 200 mg/kg of DM with varying fat concentration from 4.9 to 5.8% in the diet (Melgar et al., 2020a;2020b and2021).Although the statistical analysis revealed that lactose content was affected by the treatments, the numerical differences are negligible across the treatments.
The current study found that rapeseed supplementation in the form of cracked-seed lowered milk fat content, which is consistent with the findings from Givens et al. (2003).This is attributed to lowered mammary de novo fat synthesis as evidenced by decreased proportion of SCFA milk from cows fed HF diets.The increase in milk fat content by 3-NOP was in line with previous studies (Lopes et al., 2016;Melgar et al., 2020b and2021).This could be attributable to an increase in mammary de novo derived FA due to increased butyrate production by 3-NOP or increased body fat mobilization due to lower DMI or concentration effect by lowered milk yield (Maigaard et al., 2023).The citrate content in milk is negatively correlated to de novo FA synthesis in the mammary gland (Garnsworthy et al., 2006) which is agreed with the effect of FAT.However, 3-NOP as well as NITRATE increased de novo synthesized FA.Therefore, the increased citrate content in milk from cows fed 3-NOP in the FAT × 3-NOP interaction and milk from cows fed NIT diets is unlikely to be related   to de novo FA synthesis but might be attributed to a concentration effect resulting from lowered milk yield (Maigaard et al., 2023).The larger MFGS from cows fed NOP diets could be related to the accompanying increase of milk fat content.When more milk fat globules need to be formed, membrane materials can be limited which lead to the formation of larger MFGS (Wiking et al., 2004).HF diets decreased the proportion of SFA and increased the proportions of MUFA, and PUFA, and within the FAT × 3-NOP interaction.However, these effects became smaller when HF diets supplemented with 3-NOP, suggesting that the impact of 3-NOP on FA composition (SFA, MUFA and PUFA) depends on the level of fat in cow's diet.NITRATE had no effect on the proportions of SFA, MUFA and PUFA irrespective of the level of fat in the diet.The significant reductions in the proportions of C10, C12:0, C14:0, and C16:0 in milk by HF diets (within 3 interactions) led to a lower proportion of SFA in milk from cows fed HF diets.The decrease in the proportion of SCFA and some MCFA by HF diets reflected lowered mammary de novo FA synthesis.Increased uptake of LCFA from HF diets by the mammary gland can inhibit the de novo synthesis of SCFA and MCFA (Bauman et al., 1970).Also, the lower proportions of C8:0, and C10:0 in milk from cows fed HF-UREA and HF-NITRATE diets in comparison with milk from cows fed LF-BLANK diets within the FAT × NITRATE interaction can merely be explained by effect of FAT on mammary de novo FA synthesis.However, the lower proportion of C10:0 in cows fed HF-NIT than HF-UREA is most likely related to the difference in DMI, where cows fed HF-NIT diets showed lower DMI than cows fed HF-UREA diets (Maigaard et al., 2023).In contrast, Klop et al. (2016) found no interaction between nitrate (21 g/kg DM) and fat (supplemented by docosahexaenoic acid) on the level of SCFA in milk.The increased proportion of SCFA by NITRATE, 3-NOP and their combination indicated the increased production of butyrate by these treatments in the present study (Maigaard et al., 2023).This suggested that butyrate appears to be the primary substrate for the synthesis of SCFA in the mammary gland as opposed to acetate under the lowered methane emission by nitrate, 3-NOP or their combination.Previously, Melgar et al. (2020a) proposed that butyrate, rather than acetate, appears to become the main substrate for SCFA synthesis in the mammary gland under inhibited CH 4 emission by 3-NOP.HF diets lowered the proportions of C13:0, C16:0, C16:1 and C17:0 compared with LF diets and 3-NOP lowered them in LF diets but not in HF diets which caused a lower proportion of MCFA in milk fat from cows fed LF-NOP diets than cows fed LF-BLANK diets within the FAT × 3-NOP interaction.
The LCFA (≥C18:0 FA and half of C16:0) originates from feed and body reserves mobilization.Less than 10% of milk FA is typically accounted for by the mobilization of body fat, with this percentage could increase in cows with negative energy balance (Shingfield et al., 2008).NITRATE increased the proportion of LCFA in milk because of increased proportions of C18:0, C18:1 trans-11, and CLA cis-9, trans-11 by NIT diets.Lowered DMI by NITRATE could have increased mobilization of body fat reserves, hence increased proportions of C18:0 and CLA cis-9, trans-11 in milk from cows fed NIT diets.Due to the larger intake of preformed FA in cows fed HF diets, a pronounced effect of FAT was observed within the FAT × 3-NOP interaction as evidenced by increased proportion of LCFA in milk from cows fed both HF-BLANK and HF-NOP diets.C18:1 cis-9 was the most abundant FA in milk from cows fed HF diets while C16:0 was the most abundant FA in milk from cows fed LF diets.C18:1 cis-9 is the major FA present in rapeseed, averaging 27.1 g/100 g DM in whole rapeseed (Samková and Kalač, 2021).The higher intake of this FA was reflected in the increased proportion of C18:1 cis-9 in milk by HF diets within the FAT × 3-NOP interaction.The increased proportions of C18:0 by HF diets without interaction effect and increased proportion of C18:1 trans-11 and CLA cis-9,trans-11 by HF diets in the FAT × 3-NOP interaction were attributed to higher intake of C18:3 n-3 and C18:2 n-6.Givens et al. (2003) and Fearon et al. ( 2004) also showed similar changes in FA composition with increased supplementation of cracked/whole rapeseed.The increasing effect of FAT on C18:1 trans-11 and CLA cis-9,trans-11 became smaller in NOP diets but greater in NIT diets, while increasing effect of FAT on C18:0 remained unchanged in both NOP and NIT diets.This could be due to the differences in DMI or degree of biohydrogenation, the lower proportion of intermediate products indicates higher use of rumen hydrogen.However, a noticeable difference of hydrogen emission was not observed between HF-BLANK and HF-NOP treatments as well as between HF-UREA and HF-NIT treatments (Maigaard et al., 2023).An increased proportion of C18:0 by 3-NOP, NITRATE and their combination could be due to increased fat mobilization upon decreased DMI in those cows (Maigaard et al., 2023).Feeding HF diets increased the proportion of C18:3 n-3 in milk reflecting the higher intake of this FA.
α-tocopherol is the major form of vitamin E in milk and present in the milk fat globule membrane, where it acts as a chain breaking antioxidant and quencher of singlet oxygen in milk.The γ-tocopherol has high functional value as it can trap the nitrogen oxide species (Khan et al., 2019).In the present study, the α-and γ-tocopherol contents in milk were within the range of earlier findings in milk from Danish Holstein cows (Havemose et al., 2004).Increased α-tocopherol in milk by HF diets is attributed to an increased intake of vitamin E. Previous studies by use of whole-cracked rapeseed (Givens et al., (2003) and ground rapeseed (Larsen et al., 2016) also have reported to increase vitamin E content in milk.The increasing effect of 3-NOP on tocopherols may be due to increased fat content or changes in transfer efficiency of tocopherols from the diet.However, further studies are needed to investigate the underlying reasons.
The proportion of SFC and the effect on melting and crystallization behavior of milk fat at different temperatures are important for the physical and rheological properties of milk fat-based products such as spreadability of butter and mouth-feel.The T-onset is known to be decreased by increased proportion of UFA in milk (Couvreur et al., 2006;Smet et al., 2010).Although milk fat from cows fed HF diets had a higher proportion of UFA than LF diets in the present study, the T-onset was not affected by FAT within the FAT × 3-NOP interaction whereas 3-NOP decreased T-onset and this effect was absent when 3-NOP was supplemented in HF diets.Therefore, we assumed that the lowered T-onset by 3-NOP in LF diets might be related to the increased proportions of short-chain SFA and the decreased proportion of C16:0 by 3-NOP.
Melting points of FA in milk varies over a broad temperature range and thus milk fat melts ranging from −40 to 40°C (Figure 4).Not all the samples showed a typical curve with 3 melting peaks representing melting fractions: low, medium (MMF) and high.Therefore, we combined the first 2 peaks into one peak and considered this as LMF which included both LMF and MMF which melts from −40 to 20°C.T-offset is the temperature ending the melting and can also be considered as melting offset temperature of HMF.The lack of changes in T-offset by treatments is in agreement with Buldo et al. (2013), who reported no effect of FA composition on T-offset of HMF.The relationship between individual FA and melting point of milk fat was previously investigated and reported that melting point of the MMF of milk fat was positively correlated with C16:0 and negatively correlated with C18:1 cis-9, CLA cis-9,trans-11, C18:1 trans-11, C18:1 trans-9 and C14:1 (Buldo et al., 2013).These relationships were in agreement with the current study.Within the FAT × 3-NOP interaction, FAT lowered peak temperature of LMF (in which includes MMF), probably due to the decreased proportion of C16:0, and increased proportions of C18:1 cis-9, C18:1 trans-11, C18:2 cis-9, and CLA cis-9,trans-11 found in milk from cows fed HF diets.The decrease in C16:0 explains the lowered peak T of HMP by HF diets.
The shift of the FA profile toward a decreased proportion of SFA and increased proportion of UFA by HF diets could explain the reduction in the proportion of SFC in milk fat from cows fed HF diets.The results are in agreement with previous studies on the use of linseed (Smet et al., 2010) and rapeseed (Murphy et al., 1995;Fearon et al., 2004).The decreasing effect of FAT on SFC was pronounced even when FAT was combined with NITRATE or 3-NOP.Reduction of SFC by NOP diets at 20°C is thought to be due to increased proportion of short-chain SFA and decreased proportion of MCFA, particularly C14:0 and C16:0 by NOP diets.Decreases in SFC at 5 and 10°C by NIT diets may be accompanied by increased proportions of short-chain SFA and long-chain USFA upon NITRATE supplementation.The decreasing effects of 3-NOP and NITRATE on SFC remained in LF diets when the interactions occur with FAT.To our knowledge, this is the first study to report changes in physical and thermal properties of milk by supplementation of whole-cracked rapeseed, nitrate and 3-NOP, but further studies are needed to document how these effects manifests into dairy product quality if these mitigation strategies are implemented at a larger scale.

CONCLUSIONS
Combining fat, nitrate and 3-NOP did not exert additive effects and led to only minor changes in gross milk composition.Milk FA composition was shifted toward a lower proportion of SFA and a higher proportion of MUFA and PUFA by increasing fat level in the diet and these effects were smaller when HF diets were supplemented with 3-NOP showing that the effect of 3-NOP on FA composition depended on the fat level in the basal diet.The addition of nitrate alone or in combination with FAT or 3-NOP had no effect on the proportions of SFA, MUFA and PUFA in milk fat.The lower proportion of SFC in milk from cows fed HF diets remained unchanged even with the combination of HF diets with NIT or NOP.Treatments led to minor changes in thermal behavior of milk fat.However, these changes would be different dependent on the source and level of fat as well as the dose of nitrate and 3-NOP, which highlights the need of further studies.
Lokuge et al.: Methane-mitigation strategies and milk composition Lokuge et al.:  Methane-mitigation strategies and milk composition 6), FN ij, FO ik, FM il, NO jk, NM jl , OM kl, FNO ijk , FNM ijl , FOM ikl, NOM jkl and FNOM ijkl denote all the possible 2, 3 and 4-way interactions between FAT, NITRATE, 3-NOP and Parity whereas, C n is the random effect of the cow (n = 1 to 48) and e ijklmn and e ijkm are the random residual errors, which are assumed to be independent and normally distributed with constant variance.Estimated marginal means and standard error of mean were calculated using the emmeans package (Lenth, 2019) for both models.Pairwise comparisons were conducted by Tukey post hoc test.Differences were declared significant if P < 0.05.
Figure 1.The crystallization and melting curves derived from the analysis of a milk fat sample using differential scanning calorimetry (DSC).T-onset = temperature at the beginning of crystallization; LMF = low melting fraction; HMF = high melting fraction; T-offset = temperature at the ending of melting; Peak T = peak temperature.
of fat at the levels of LF and HF (30 or 63 g crude fat/kg of DM, respectively), NITRATE = addition of nitrate at the levels of UREA and NIT (0 or 10 g nitrate/kg of DM, respectively), 3-NOP = supplementation of 3-NOP at the levels of BLANK and NOP (0 or 80 mg/kg of DM, respectively). 1 Standard error of Estimated Marginal Mean.P-values are reported for the selected effects and their interactions.Differences were declared significant if P < 0.05.Table2.Effects of dietary addition of fat, nitrate and 3-nitrooxypropanol and their combinations on fatty acid (FA) composition and milk fat globule size (MFGS) (FA are given in g/ 100 g of total FA) of fat at the levels of LF and HF (30 or 63 g crude fat/kg of DM, respectively), NITRATE = addition of nitrate at the levels of UREA and NIT (0 or 10 g nitrate/kg of DM, respectively), 3-NOP = supplementation of 3-NOP at the levels of BLANK and NOP (0 or 80 mg/kg of DM, respectively).

Figure 2 .
Figure 2. Interaction between FAT and 3-NOP on the proportions of (A) saturated fatty acid (SFA) (B) monounsaturated fatty acid (MUFA) and (C) polyunsaturated fatty acid (PUFA) in milk fat (P < 0.001 for all).FAT = supplementation of fat at the levels of LF and HF (30 or 63 g crude fat/kg of DM, respectively), 3-NOP = supplementation of 3-NOP at the levels of BLANK and NOP (0 or 80 mg/kg of DM, respectively).Bars represent treatment means and Error bars represent the SEM.
Lokuge et al.: Methane-mitigation strategies and milk composition Table 3. Effects of dietary addition fat, nitrate and 3-nitrooxypropanol and their combinations on tocopherols in milk of fat at the levels of LF and HF (30 or 63 g crude fat/kg of DM, respectively), NITRATE = addition of nitrate at the levels of UREA and NIT (0 or 10 g nitrate/kg of DM, respectively), 3-NOP = supplementation of 3-NOP at the levels of BLANK and NOP (0 or 80 mg/kg of DM, respectively). 1 Standard error of Estimated Marginal Mean.P-values are reported for the selected effects and their interactions.Differences were declared significant if P < 0.05.

Figure 3 .
Figure 3.Effect of dietary addition of FAT, NITRATE and 3-NOP and their combinations on solid fat content (SFC) at different temperatures.FAT = supplementation of fat at the levels of LF and HF (30 or 63 g crude fat/kg of DM, respectively), NITRATE = addition of nitrate at the levels of UREA and NIT (0 or 10 g nitrate/kg of DM, respectively), 3-NOP = supplementation of 3-NOP at the levels of BLANK and NOP (0 or 80 mg/kg of DM, respectively).
of fat at the levels of LF and HF (30 or 63 g crude fat/kg of DM, respectively), NITRATE = addition of nitrate at the levels of UREA and NIT (0 or 10 g nitrate/kg of DM, respectively), 3-NOP = supplementation of 3-NOP at the levels of BLANK and NOP (0 or 80 mg/kg of DM, respectively).T-onset = The onset crystallization temperature, T-offset = The offset melting temperature, Peak T of LMF = Peak temperature of low melting fraction, Peak T of HMP = Peak temperature of high melting fraction.1 Standard error of Estimated Marginal Mean.P-values are reported for the selected effects and their interactions.Differences were declared significant if P < 0.05.
Lokuge et al.: Methane-mitigation strategies and milk composition Lokuge et al.: Methane-mitigation strategies and milk composition

Table 1 .
Lokuge et al.:Methane-mitigation strategies and milk composition Effect of dietary addition of fat, nitrate and 3-nitrooxypropanol and their combinations on milk composition