J. Dairy Sci.

100:7088–7105
https://doi.org/10.3168/jds.2016-12383
© American Dairy Science Association®, 2017.

Feeding a concentrate rich in rapeseed oil improves fatty

acid composition and flavor in Norwegian goat milk
R. A. Inglingstad,*1 S. Skeie,* G. E. Vegarud,* T. G. Devold,* Y. Chilliard,† and M. Eknæs‡
*Faculty of Chemistry, Biotechnology and Food Science (KBM), Norwegian University of Life Sciences (NMBU), 1430 Ås, Nroway
†UMR1213 Herbivores, Institut National de la Recherche Agronomique (INRA)-Clermont-Theix-Lyon, F-63122 Saint-Genès-Champanelle, France
‡Faculty of Biosciences (BIOVIT), Norwegian University of Life Sciences (NMBU), 1430 Ås, Norway

ABSTRACT INTRODUCTION

Impaired quality due to a high content of free fatty Norwegian goat milk has been of variable quality in
acids (FFA) and off-flavors has caused challenges in respect to both rennet coagulation properties and off-
the development of Norwegian goat milk products. The flavors, which may be a problem in the production of
present study aimed to examine the effect of lipid-sup- cheese. The problem is most prominent in mid lactation,
plemented concentrates on milk fat content, fatty acid which sometimes coincides with the time when the goats
composition, FFA, lipoprotein lipase activity, sensory are let out on mountain or rangeland pasture (Eknæs
properties, and size of milk fat globules of goat milk. et al., 2006). The off-flavors of goat milk are correlated
Thirty goats assigned to 3 experimental groups were with the content of free fatty acids (FFA; Dønnem et
fed different concentrates from 60 d in milk (DIM) until al., 2011), and goats deficient in αS1-CN [goats homo-
late lactation (230 DIM). The diets were (1) control zygous for a deletion in exon 12 of the gene encoding
concentrate (no added fat); (2) control concentrate αS1-CN, which causes low or no synthesis of this protein
with 8% (added on air-dry basis) hydrogenated palm in their milk (E12–00)] have higher frequencies of FFA
oil enriched with palmitic acid (POFA); and (3) control and off-flavors compared with goats not deficient in
concentrate with 8% (added on air-dry basis) rapeseed αS1-CN [goats with no or only one deletion in exon 12
oil (RSO). The POFA group produced milk with the of the gene encoding αS1-CN; these goats express this
highest fat content, and fat content was positively cor- protein in their milk (E12–11/E12–01); Dagnachew et
related with the mean size of milk fat globules. Goats al., 2011]. Free fatty acids result from the action of
in the RSO group had a higher content of long-chain lipoprotein lipase (LPL), a highly potent enzyme that
and unsaturated fatty acids, whereas milk from goats in hydrolyses mainly triglycerides (TG) into glycerol and
the POFA group had a higher content of palmitic and FFA. However, LPL does not reach its full lipolytic
palmitoleic acids (C16:0 and C16:1 cis). The control potential in milk because its substrate (TG) is localized
group produced milk with a higher content of short-, in milk fat globules (MFG) surrounded by a protective
medium-, odd-, and branched-chain fatty acids com- membrane—the milk fat globule membrane (MFGM;
pared with the 2 other groups. The content of FFA in Deeth, 2006). In cow milk, LPL is associated with the
milk was low in early and late lactation and peaked in casein micelles. However, in goat milk it is probably
mid lactation (90 DIM). A high content of FFA was associated with the MFGM (Chilliard et al., 2003). If
correlated with poor sensory properties (tart/rancid the MFGM is damaged or broken, LPL will have im-
flavor). The RSO group produced milk with lower mediate access to TG, resulting in excessive lipolysis.
content of FFA and off-flavors in mid lactation and a Previous studies (Eknæs et al., 2009) have shown that
higher proportion of unsaturated fatty acids. Therefore, extra supplements of saturated fat (C16:0 and C18:0)
replacement of palm oil with rapeseed oil as a lipid in goat feed improves the milk flavor. Calcium salts of
source in dairy goat feed would be favorable. palm oil-derived fatty acids are largely used as dairy
Key words: rapeseed oil, palm oil, goat milk, fatty energy supplements (Onetti and Grummer, 2004;
acid composition, free fatty acids Rabiee et al., 2012). However, the use of palm oil is
severely criticized, both from an environmental sustain-
ability point of view (Wilcove et al., 2013) and from a
consumer health perspective, because it increases milk
palmitic acid, which is not recommended (Hayes et al.,
1997; Shingfield et al., 2008). Hence, the search for a
Received December 1, 2016.
Accepted May 15, 2017. good substitute to palm oil in animal feed becomes in-
1
Corresponding author: [email protected] creasingly important.
7088
 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7089

Rapeseeds could be a more sustainable substitute for (Akofeed Gigant 60, AarhusKarlshamn Sweden AB;
palm oil in the dairy feed. Oilseeds were introduced in POFA), and (3) control concentrate with 8% (air-dry
Norway in the late 1950s; however, the land used for basis) rapeseed oil (AarhusKarlshamn Sweden AB;
cultivation of oilseeds today is less than 1% of total RSO). The treatment designations CON, POFA, and
cultivated farmland (Granlund et al., 2010). Several RSO are also used to describe the milk from the goats
experiments feeding rapeseeds or rapeseed oil to dairy on the respective treatments. The experimental groups
goats have shown positive effects on milk quality in were balanced according to lactation number, kidding
terms of milk flavor and fatty acid composition. Activ- date, milk yield, and BW. Each group consisted of 7
ity of LPL and content of FFA were reduced (Ollier goats heterozygous (E12–01) for the deletion in exon
et al., 2009), which improve the sensory quality of the 12 at the αS1-CN locus (CSN1S1), whereas 3 were ho-
milk. Contents of SFA, especially C16:0, decreased, and mozygous (E12–00) for this deletion. From 0 to 130
contents of trans-C18:1 isomers, linoleic, linolenic, and DIM and from 200 to 230 DIM, the goats were stabled
conjugated linoleic acids increased (Gulati et al., 1997; indoor in individual pens and received silage according
Mir et al., 1999; Andrade and Schmidely, 2006; Ollier et to appetite (10% refusals). From 130 to 200 DIM, the
al., 2009). The trans isomers C18:1 trans-6–9 are found goats were grazing free-range in a mountain pasture
in hydrogenated vegetable oils and believed to have a about 1,000 m above sea level. The mountain pasture
negative effect on human health, whereas C18:1 trans-11 consisted of boggy areas with sedges (mainly Carex
and the CLA isomer C18:2 cis-9,trans-11 found in ru- nigra and Carex rostrata) and drier areas with grass,
minant milk have neutral or health-promoting effects herbs, and shrubs (mainly Deschampsia cespitosa, Salix
(Shingfield et al., 2008). We therefore hypothesize that spp., and Betula spp.; Eknæs et al., 2006). The goats
rapeseed oil in the goats’ diet will reduce the frequency received 0.9 kg of the experimental concentrate per day
of off-flavors and the content of FFA, and increase the until the start of the mountain grazing season, and 0.7
proportion of UFA in the milk fat. kg/d thereafter. The concentrate was fed individually
The aim of this study was thus to evaluate the ef- with no refusals to the goats 3 times a day during the
fect of lipid-supplemented concentrate fed to Norwe- indoor period and twice a day during the mountain
gian dairy goats on the following milk parameters: fat grazing period. Further details on animal feeding man-
content, fatty acid composition, lipolysis, LPL activity, agement and production performances are described in
FFA, and MFG size. Two lipid sources were compared Eknæs et al. (2017).
(saturated palm oil and unsaturated rapeseed oil) with
a concentrate with no added fat (control feed). Milk Feed Production and Analysis
samples from individual goats were collected through-
out the lactation cycle. We also investigated and report The experimental concentrate mixtures were pro-
separately milk protein composition and cheesemaking duced by the Centre for Feed Technology at the Nor-
properties (Inglingstad et al., 2016) and goats’ energy wegian University of Life Sciences, and the fat contents
balance and milk production parameters (Eknæs et al., and fatty acid composition of the concentrates and the
2017) from the same experiment. silage are shown in Table 1.
Silage was harvested at the experimental farm at
MATERIALS AND METHODS Ås, Norway, and consisted of grass of timothy (Phleum
pratense), meadow fescue (Festuca pratensis), and red
Experimental Design clover (Trifolium pratense). After cutting, the grass was
wilted to achieve a DM content of 250 g/kg of DM.
Thirty Norwegian dairy goats in second to fourth The grass was baled using an Orkel HiQ (Orkel AS,
lactation, kidding in February 2011 (average kidding Fannrem, Norway) roundbaler, preserved with Ensil1
date: February 17 ± 8 d) and with average BW 2 (Felleskjøpet Agri SA, Lillestrøm, Norway; 4 L/t), and
d after kidding of 54.4 ± 6.7 kg were fed a control wrapped in 6 layers of plastic film (Trio Wrap, Tri-
diet until 60 DIM (preparatory period). The control oplast AB, Sweden).
concentrate consisted of barley, rapeseed meal (Expro Fatty acid composition in silage and concentrate was
00SF, AarhusKarlshamn Sweden AB, Karlshamn, Swe- analyzed by Vitas (Oslo, Norway). Dried and milled
den), soybean meal, beet pulp, molasses, and mineral/ samples were accurately weighed (200 mg) and trans-
vitamin premix. At 60 DIM, the goats were assigned ferred to soda lime tubes. Internal standard, methyl
to 3 experimental groups each of 10 goats receiving 3 tricosanoate (Nu-Chek Prep, Waterville, MN), and 900
types of feed. The diets were (1) control concentrate µL of 3 N methanolic HCl were added, and samples
(CON), (2) control concentrate with 8% (air-dry ba- were methylated for 2 h at 80°C. Aliquots of 100 µL of
sis) hydrogenated palm oil enriched with palmitic acid FAME were transferred to GC vials, diluted with hex-
Journal of Dairy Science Vol. 100 No. 9, 2017
7090 INGLINGSTAD ET AL.

ane and injected (1 µL, splitless) on an Agilent 7890A Total Content of Fat and FFA
Gas Chromatograph System with flame-ionization
detector (FID; Agilent Technologies, Palo Alto, CA). The total content of fat and FFA in milk were ana-
The FAME were separated on a Supelcowax 30 m × lyzed by FTIR/milkoscan (MilkoScan Combifoss 6500;
250 µm × 0.2 µm column (Supelco Inc., Bellefonte, Foss, Hillerød, Denmark). To avoid microbial growth,
PA). Helium (99.9999%, AGA, Oslo, Norway) was used 1 tablet of Bronopol (2-bromo-2-nitropropane-1, 3-diol;
as carrier gas at a constant flow of 3.25 mL/min. The D&F Control Systems Inc., Dublin, CA) was added to
GC oven temperature was 70°C for 0.5 min, raised by the samples, which were kept at 4°C until analysis. The
30°C/min to 170°C, thereafter 1.5°C/min to 221, and fat content was measured in milk stored for 36 h. To
finally 50°C/min to 255 for 5 min. study lipolysis in milk during storage, the content of
The chemical composition of the feed was analyzed FFA was measured in raw milk samples after 36 and 84
as described by Eknæs et al. (2017). h. To obtain a 0-sample, the content of FFA was mea-
sured in milk pasteurized (63°C/30 min) immediately
Milk Sampling
after milking (pasteurization inactivates LPL activity).

Milk was sampled 6 times throughout the lactation Extraction, Separation, and Analysis of Milk Lipids
period at 30, 60, 90, 120, 190, and 230 DIM. The goats
were milked at 0600 and 1530 h. Milk yield was mea- The total lipids were extracted from 1 mL of milk
sured on 3 subsequent days every second week during as described by Steinshamn et al. (2014) using a modi-
indoor feeding and on 2 subsequent days every fourth fied method of Folch et al. (1957). Internal standards
week during mountain grazing. Because milk yield was (trinonadecanoin, 2.5 mg, and trinonadecanoic acid,
lower in the evening, morning and evening milks were 0.2 mg; Larodan Fine Chemicals, Malmö, Sweden) dis-
mixed in the ratio 1.5:1. Preparation and analysis were solved in chloroform were added to the samples before
performed on milk stored at 4°C for 84 h (age of milk lipid extraction.
when processed by the dairies), without preservatives The extracted lipids were re-dissolved in
unless otherwise specified. At 190 DIM, the goats were hexane:chloroform:methanol (95:3:2, vol/vol) and
on summer mountain pasture and because of limited loaded on 500-mg aminopropyl cartridges (Hypersep
laboratory facilities, the milk samples were stored in spe, Thermo Scientific, Bellefonte, PA) for solid-phase
Styrofoam boxes with crushed ice and cooling elements extraction. The cartridges were conditioned with 7 mL
during the 5-h transfer by car to the University (Ås, of hexane before loading. Neutral lipids (mainly TG)
Norway) for preparation and analyses. were eluted with 5 mL of chloroform, and then FFA

Table 1. Chemical and fatty acid composition of control concentrate, concentrates with added 8% fatty acids
from palm oil (POFA) or rapeseed oil (RSO), and silage

Concentrate

Item Control POFA RSO Silage

DM (g/kg) 912 889 897 253
Chemical composition (g/kg of DM)        
 CP 196 191 195 138
  Crude fat 22 107 110 34
 Starch 343 280 299  
 aNDF1 187 172 168 531
  Total N 32.5 34.5 34.9 23.4
 Ash 73 72.5 69.5 77.2
Fatty acid composition (g/100 g of total FAME)        
 C14:0 0.3 0.7 0.1 0.5
 C16:0 17.9 52.6 8.1 13.3
 C16:1 cis-9 0.3 0.1 0.3 1.5
 C18:0 1.8 25.9 2.4 1.6
 C18:1 cis-9 15.1 4.5 48.9 3.1
 C18:1 cis-11 2.3 0.7 3.1 0.5
 C18:2 cis-9,12 37.9 8 22.9 14.8
 C18:3 cis-9,12,15 4.5 1 6.9 41
 C20:0 0.2 0.3 0.6 0.8
  Unidentified FAME 19.7 6.2 6.7 22.9
1
NDF assayed with α amylase and expressed exclusive of residual ash.

Journal of Dairy Science Vol. 100 No. 9, 2017

 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7091

were eluted with 5 mL of 2% acetic acid in diethyl Instruments, Malvern, UK). The method is previously
ether. The solvents were evaporated to dryness under a described by Michalski et al. (2001) with some modi-
gentle stream of N2. fications: 0.1% SDS solution was replaced by MilliQ
Transesterification and esterification of fatty acids water.
from TG and FFA, respectively, were performed ac- The composition of phospholipids and cholesterol in
cording to Devle et al. (2014), and FAME in hexane the MFGM were analyzed by Vitas AS (Oslo, Norway).
were transferred to GC vials and stored at −20°C until Six milk samples (3 with high and 3 with low content of
analysis. FFA) from the sampling at 90 DIM were chosen for this
The FAME were analyzed using a Carlo Erba GC analysis, and only milk from goats on the POFA feed
8000 GC equipped with a Carlo Erba EL 980 autosam- was used to omit confounding factors such as lactation
pler and a Carlo Erba AS V570 FID (Carlo Erba In- stage and feed. The milk samples were kept at −20°C
struments, Milano, Italy). The column used was a 50-m before analysis. For identification and quantification
CP-Sil 88 capillary column with an inner diameter of of polar lipids and sphingomyelin, samples were accu-
0.25 mm and 0.20-µm film thickness (Varian, Agilent rately weighed, dissolved in isopropanol, shaken, and
Technologies, Matriks, Norway). Helium (99.9999%, centrifuged (4,000 × g at 10°C for 10 min) before the
AGA, Oslo, Norway) was used as carrier gas at a con- supernatants were transferred to new vials. The pellets
stant pressure of 75 kPa. The injected sample (1 µL) were washed with isopropanol and centrifuged (4,000
was split in a ratio of 1:25. The GC oven temperature × g at 10°C for 10 min), and the resulting superna-
was 60°C for 3 min, raised by 10°C/min to 140°C and tants were pooled with the first ones. The superna-
held for 1 min, raised by 10°C/min to 160°C and held tants were evaporated to dryness and dissolved in a
for 1 min, and finally raised by 2.5°C/min to 210°C and mix of chloroform:methanol:water and analyzed with
held for 20 min. The FID signals were transferred to a an Agilent 1100 normal phase liquid chromatography
Total Chrom workstation for interpretation and cali- system using an evaporative light scattering detector.
bration. The FAME were identified by comparing the Separation was performed on a PVC-SIL-NP 250 × 4.6
retention time of standards (Larodan Fine Chemicals; mm HPLC column (YMC, Dinslaken, Germany) us-
Sigma-Aldrich, St. Louis, MO) in addition to fatty acid ing hexane, isopropanol, acetonitrile, chloroform, tert-
profiles of goat milk obtained by GC-MS in a previous butyl methyl ether, water, and acetic acid as mobile
study (Steinshamn et al., 2014). phase. Analytes were calibrated against known stan-
It is important to note that use of polypropylene dards (Lipoid GmbH, Köln, Germany). Samples for
tubes is a common practice in most laboratories today; quantification of cholesterol were accurately weighed
however, in contact with organic solvents, fatty acids and dissolved in methanol and a sodium hydroxide
are released from the material. We detected consider- solution, and then incubated for 60 min at 50°C to
able amounts of C16:0 and C18:0 fatty acids in blank hydrolyze cholesteryl esters to free cholesterol and then
samples originating from both the tubes and the SPE centrifuged. Cholesterol was determined by an Agilent
cartridges; these fatty acids were therefore omitted from reversed-phase liquid chromatography system using a
the results. This problem was only observed in the frac- diode array detector. Separation was performed on an
tion containing the FFA; the blanks from the fraction Eclipse XDB-C8 150 × 4.6 particle size 5-µm column
of neutral lipids (TG) were free from contamination. from Agilent using methanol with ammonium acetate
We strongly recommend use of acid-washed glassware as mobile phase. Analytes were calibrated against stan-
for sample preparation of FFA and a frequent control dards from Sigma-Aldrich.
of blank samples.
Sensory Evaluation of Milk
Measurement of LPL Activity
Three panelists trained in sensory evaluation of goat
The LPL activity was measured in milk sampled at
milk performed the analysis. The analysis was done by
30, 60, 120, 190, and 230 DIM and kept at −20°C before
scoring using a scale from 1 to 5, where 1 indicates
analysis. Activity of LPL (EC 3.1.1.34) was measured
pronounced flavor deviation and 5 indicates good milk
using an artificial emulsion containing [3H]-triolein
without off-flavors. The panel’s average score for each
(Bernard et al., 2005).
sample was used in calculations.
Measurement of MFG Size and Composition
of the MFGM Statistical Analysis

The size of the milk fat globules was measured by Analysis of variance was performed using the MIXED
laser light scattering using a Mastersizer 2000 (Malvern procedure (Littell et al., 1998) of SAS (ver. 9.4, SAS In-
Journal of Dairy Science Vol. 100 No. 9, 2017
7092 INGLINGSTAD ET AL.

stitute Inc., Cary, NC). Samples were collected 4 times 0.001). Often, a lower fat content can be explained by
(3 times for LPL activity) from each goat during the higher milk yield. However, the milk yield was not sig-
experimental period; therefore, a repeated-measures nificantly different between any of the feeding groups,
procedure was used. The covariance structure of the but decreased gradually during the course of lactation
repeated measurements was chosen by comparing po- (Table 2). The fat content in milk was highest shortly
tential structures using Akaike’s and Schwarz’s Bayes- after kidding (30 DIM) and in POFA milk (POFA >
ian information criteria (Wolfinger, 1996). Variance RSO > CON; Table 2). A high fat content in early
components covariance structure proved useful for all lactation is in accordance with other studies (Chilliard
the milk data. The values at 60 DIM were used as a et al., 2003; Eknæs et al., 2006). Extra dietary lipids
covariate. The ANOVA with repeated measures of the fed to goats are known to increase the fat content in
milk data was performed according to the following the milk (Chilliard et al., 2007; Sanz Sampelayo et al.,
model: 2007; Tudisco et al., 2015). However, differences in fat
content (P = 0.02) between the POFA and RSO groups
Yijkl = µ + Ai + Bj + A × B(ij) + Ck + A were not expected. One of the limitations of feeding
× C(ik) + d(Ai,Ck) + εijkl, fat supplements to dairy cows is the inhibition of their
rumen microbial activity (Palmquist, 1984), which af-
where Yijkl is the dependent variable; µ is the intercept; fects fiber digestion in the rumen (Palmquist, 1984)
Ai is the fixed effect of concentrate type, i = 1, 2, 3 and probably depresses fiber intake (Khorasani et al.,
(CON, POFA, RSO); Bj is the fixed effect of DIM, j = 1996). Furthermore, feeding a surplus of unsaturated
1, 2,…,4 (DIM 90, 120, 190, 230); A × B(ij) is the inter- fats to cows may cause milk fat depression (Bauman
action between concentrate type i and DIM j; Ck is the and Griinari, 2003; Chilliard et al., 2007). However,
fixed effect of genotype at the CSN1S1 locus, k = 1, 2 goats seem to tolerate addition of unsaturated fat well,
(E12–00, E12–01); A × C(ik) is the interaction between and hence it is possible to alter the milk fatty acid
concentrate type i and genotype k; d(Ai,Ck) is random composition by feeding (Chilliard et al., 2007). In total,
effect of goat within concentrate type and genotype and 32 fatty acids were identified and quantified from the
εijkl represents the experimental error. TG fraction (Tables 3, 4, 5, and 6). The most abundant
The difference between means was estimated by the fatty acids were palmitic acid (C16:0; Table 3), stearic
calculated least squares means (LSM). Differences were acid (C18:0), and oleic acid (C18:1 cis-9; Table 5). The
considered statistically significant when P < 0.05. composition of all fatty acids except arachidonic acid
Principal component analysis (PCA) of the TG fatty (C20:4 cis-5,8,11,14) were influenced by lactation stage
acids composition and partial least squares regression . Moreover, the interaction between feeding group and
using the FFA content and composition as X variables lactation stage was significant for most fatty acids (Ta-
and sensory scores as Y variables were conducted using bles 3 to 6). The contents of short- and medium-chain
The Unscrambler X 10.3 (CAMO Process AS, Oslo, SFA (C4:0–C14:0) were highest in early and late stages
Norway). The fatty acid data (for TG and FFA) were of lactation, whereas UFA were highest in mid lacta-
standardized by dividing each response variable by its tion. The PCA separated goats into 3 distinct clusters
standard deviation to ensure equal contribution in the according to feed group (Figure 1). Comparison of the
model. The sensory data were not weighted as the same score plot (goats; Figure 1A) with the loading plot
scale was used during analysis. Full cross-validation (milk fatty acids; Figure 1B) shows that goats in the
was used to validate the data set. CON group produced milk fat with a higher content
The data from the analysis of the MFGM composition of de novo synthesized SFA with <16 carbon atoms
(content of phospholipids and cholesterol) of samples (except for C4:0). Moreover, the CON milk fat had
with high versus low content of FFA were analyzed by a a higher proportion of odd- and branched-chain fatty
2 sample t-test using R Statistics (version 3.2.5; R Core acids (Tables 3 and 4) compared with the POFA and
Team, 2016). Differences were considered statistically RSO groups. The odd- and branched-chain fatty acids
significant when P < 0.05. originate from ruminal metabolism of branched-chain
AA, propionate, and butyrate (Chilliard et al., 2003).
The higher content of de novo synthesized fatty acids
RESULTS AND DISCUSSION
and odd- and branched-chain fatty acids in CON milk
Fat Content and Fatty Acid Composition fat than in POFA and RSO milk fat is probably because
of Triglycerides supplementation of dietary fat inhibits de novo synthe-
sis and alters the ruminal microflora. A higher content
The fat content in milk was influenced both by DIM of C16:0 and C16:1 cis throughout lactation character-
(P < 0.001) and type of lipid supplementation (P < ized the POFA milk fat compared with the CON and
Journal of Dairy Science Vol. 100 No. 9, 2017
 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7093

RSO milks (Table 3, Figure 1). These results are in putative effects of lactation stage and of the change
accordance with previous results using the same source in basal forage from silage to mountain pasture at 190
of saturated lipids (Eknæs et al., 2009) and reflects the DIM were confounded. However, the different milk
high C16:0 in POFA, and the Δ9-desaturation of a part fatty acid composition observed at 190 DIM is more
of C16:0. The RSO milk fat had higher proportions of likely an effect of grazing rather than lactation stage
fatty acids with 18 or more carbons, especially C18:0 because lactation was in its slowly declining phase and
and C18:1 cis-9. However, linoleic (C18:2 cis-9,12) and similar effects of changing the basal diet from hay or
linolenic acid (C18:3 cis-9,12,15) were higher in CON silage to grass were reported previously (e.g., Chilliard
milk, but only when the goats were grazing mountain et al., 2007; Steinshamn et al., 2014).
pastures, and the content of C20:4 cis-5,8,11,14 was not Pasture and supplementation of rapeseed oil increased
influenced by feed (Table 5). This is most likely due to the content of long-chain and unsaturated fatty acids in
the almost complete biohydrogenation in the rumen of the milk fat in present study, and intake of such fatty
linoleic and linolenic acids from dietary rapeseed oil, acids is recommended for human consumption by health
which are transformed into C18:0 and several cis or authorities and nongovernmental organizations (FAO,
trans intermediates (Chilliard et al., 2007). 2010). Moreover, some fatty acids, such as branched-
The shift in basal diet from silage to mountain chain fatty acids, vaccenic acid (C18:1 trans-11), and
pasture from 130 to 200 DIM caused an increase in CLA, are thought to have health-promoting properties
long-chain SFA (C17, C18, C20, and C22), vaccenic, (Shingfield et al., 2008).
linoleic, and linolenic acids at the expense of medium- If milk is to be processed into products such as cheese
chain (C10–C14) and odd- and branched-chain fatty and butter, a high proportion of UFA may result in
acids in milk fat from all feed groups (Tables 3 to 6) products with a softer texture. Protein composition and
and C16:0 in CON and POFA groups (Table 3). The cheesemaking properties of the POFA and RSO milks

Table 2. Milk yield, fat content, size of milk fat globules (MFG), lipoprotein lipase activity (LPL), concentration of free fatty acids after 84 h
of storage (FFA84h), and flavor scores in milk from goats fed a control (CON) feed with no extra fat, or concentrates supplemented with either
palm (POFA) or rapeseed oil (RSO)

Preparatory period
(DIM) Experimental period (DIM) P-value

Feed ×
Item   Feed 301 601   90 120 1902 230 SEM Feed DIM DIM
Milk yield (kg/d) CON 3.09 2.99 3.21w 2.99w 2.45x 1.91y 0.114 0.685 <0.001 0.333
POFA 3.27 3.16 3.08w 2.73x 2.59x 1.79y
RSO 3.15 3.11 3.05w 2.86w 2.59x 1.96y
SEM3 0.230 0.211
Fat (g/kg) CON 52.0 44.4 35.8b,x 31.8b,y 34.3b,xy 33.5b,xy 0.166 <0.001 <0.001 0.490
POFA 52.8 43.0 45.3a,x 39.8a,y 40.1a,y 39.9a,y
RSO 51.6 48.3 40.0b,x 34.1b,y 37.2ab,xz 34.6b,yz
SEM3 0.227 0.173
MFG CON 3.39 3.24 3.03b,x 3.02b,xy 3.00b,xy 2.87y 0.085 0.049 <0.001 <0.001
 (d43, µm) POFA 3.51 3.32 3.41a,x 3.30a,y 3.22ab,y 2.93z
RSO 3.47 3.48 3.30b,x 3.16ab,y 3.34a,x 2.84z
SEM3 0.103 0.090
LPL (nmol/min CON 489 496 403x 386a,xy 381a,y 8.677 0.002 <0.001 0.003
  per milliliter) POFA 425 421 402x 360b,y 344b,y
RSO 418 421 395x 331c,y 334b,y
SEM3 6.756 9.319
FFA84h (mM) CON 0.23 0.42 1.74a,x 1.25a,xy 1.03y 0.32b,z 0.221 0.008 <0.001 0.105
POFA 0.17 0.38 2.01a,x 1.35a,y 1.27y 0.73a,z
RSO 0.32 0.41 0.81b,x 0.41b,xy 0.86x 0.23b,y
SEM3 0.178 0.199
Flavor score CON 4.67 4.20 2.77b,yz 2.47b,z 3.47b,y 4.27x 0.364 0.009 <0.001 0.016
POFA 4.70 4.00 2.91b,y 3.41b,xy 3.71ab,xy 3.81x
RSO 4.20 3.90 4.11a 4.76a 4.56a 4.56
SEM3 0.389 0.389  
a–c
LSM within a column with different letters are significantly different (P < 0.05) from each other.
w–z
LSM within a row with different superscript letters are significantly different (P < 0.05) from each other.
1
All goats received the control concentrate in the preparatory period.
2
Mountain pasture.
3
Standard error of mean for the preparatory period.

Journal of Dairy Science Vol. 100 No. 9, 2017

7094 INGLINGSTAD ET AL.

Table 3. Short- and medium-chain fatty acids (% of total FA in triglycerides) in milk from goats fed either a concentrate with palm (POFA)
or rapeseed oil (RSO) or a control (CON) feed with no extra fat

Preparatory period
(DIM) Experimental period (DIM) P-value

Feed ×
Item   Feed 301 601   90 120 1902 230 SEM Feed DIM DIM
C4:0 CON 3.47 2.09 2.11b,x 1.63b,y 1.48y 1.58b,y 0.06 0.014 <0.001 0.559
POFA 3.47 2.19 2.39a,x 1.81a,y 1.54z 1.78a,y
RSO 3.20 1.95 2.32b,x 1.71ab,yz 1.56y 1.74ab,z
SEM3 0.173 0.249
C6:0 CON 2.31 2.19 2.20a,x 1.69a,z 1.64a,z 1.99y 0.06 0.006 <0.001 0.254
POFA 2.35 2.10 1.95b,x 1.42b,z 1.45b,z 1.90x
RSO 2.30 2.14 2.20a,x 1.59a,z 1.56ab,z 2.02y
SEM3 0.101 0.249
C8:0 CON 2.33 2.16 2.06a,x 1.71a,y 1.49a,z 1.99a,x 0.07 <0.001 <0.001 0.014
POFA 2.44 2.11 1.55b,y 1.23b,z 1.21b,z 1.72b,x
RSO 2.42 2.29 1.97a,x 1.51a,y 1.40ab,y 1.99a,x
SEM3 0.119 0.136
C10:0 CON 9.05 8.85 8.54a,x 7.08a,y 5.64a,z 8.57a,x 0.29 <0.001 <0.001 <0.001
POFA 9.33 9.17 5.52c,y 4.42c,z 4.31b,z 6.86b,x
RSO 9.45 9.64 6.86b,y 5.32b,z 4.99ab,z 7.83a,x
SEM3 0.557 0.566
C12:0 CON 4.62 4.13 3.72a,x 3.21a,z 2.45a,y 4.62a,w 0.15 <0.001 <0.001 <0.001
POFA 4.53 4.64 2.24c,y 1.99b,z 1.87b,z 3.46b,x
RSO 4.73 4.93 2.67b,y 2.29b,z 2.10ab,z 3.75b,x
SEM3 0.314 0.425
C14:0 CON 11.51 12.24 12.89a,x 11.42a,y 9.16a,z 14.34a,w 0.29 <0.001 <0.001 <0.001
POFA 11.50 12.53 8.69c,y 7.77b,z 7.43b,z 11.37c,x
RSO 11.90 12.67 9.54b,x 8.47b,y 7.84b,z 12.32b,w
SEM3 0.489 0.462
C14:1 cis-9 CON 0.14 0.17 0.15a,y 0.17a,y 0.09z 0.37a,x 0.02 <0.001 <0.001 0.002
POFA 0.14 0.19 0.09b,z 0.11b,z 0.08z 0.24b,y
RSO 0.14 0.17 0.09b,z 0.11b,z 0.08z 0.23b,y
SEM3 0.012 0.018
C16:0 CON 24.87 28.71 34.99b,y 33.66b,y 28.76b,z 37.92a,x 0.77 <0.001 <0.001 <0.001
POFA 24.66 27.57 42.29a,w 39.15a,xy 36.28a,z 40.01a,x
RSO 24.79 27.70 23.91c,y 21.78c,z 23.37c,yz 28.70b,x
SEM3 0.855 0.754
C16:1 cis CON 0.63 0.70 0.68b,y 0.80a,x 0.54a,x 1.04a,w 0.04 <0.001 <0.001 <0.001
POFA 0.67 0.70 0.97a,y 0.89a,y 0.62a,z 1.08a,x
RSO 0.63 0.65 0.41c,z 0.48b,z 0.41b,z 0.66b,y
SEM3 0.065 0.048
C16:1 trans CON 0.16 0.43 0.29b,y 0.31b,y 0.41x 0.20b,z 0.02 <0.001 <0.001 <0.001
POFA 0.15 0.44 0.26b,z 0.32b,y 0.38x 0.22b,z
RSO 0.12 0.41 0.40a,y 0.44a,y 0.41y 0.29a,z
SEM3 0.34 0.020  
a–c
LSM within a column with different letters are significantly different (P < 0.05) from each other.
w–z
LSM within a row with different superscript letters are significantly different (P < 0.05) from each other.
1
All goats received the control concentrate in the preparatory period.
2
Mountain pasture.
3
Standard error of mean for the preparatory period.

were analyzed in a parallel study. A doughy cheese and C18:1 cis-9 (Coppa et al., 2011), which are the
with a high moisture content was obtained from milk most abundant SFA and UFA in POFA and RSO milks,
of goats fed rapeseed oil, whereas cheese with a lower respectively. However, the influence of fatty acid com-
moisture content, better texture, and a higher content position on content of free AA and cheese ripening has
of free amino acids, indicating a higher proteolysis rate, not previously been reported in the literature, to our
was obtained from milk of goats fed palm oil (Inglings- knowledge, and why different fatty acid composition
tad et al., 2016). Because no differences in the protein results in different content of free AA is yet unknown.
composition were detected, these effects in the final Goat milk contains a higher proportion of the “goat
cheeses are likely to be explained by the different fatty fatty acids”—caproic (C6:0), caprylic (C8:0), and cap-
acid composition in RSO and POFA milks. The fatty ric (C10:0) acids—compared with other ruminant milk
acids that influence cheese texture are mainly C16:0 (Chilliard et al., 2003). These fatty acids were influ-
Journal of Dairy Science Vol. 100 No. 9, 2017
 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7095
Table 4. Odd- and branched-chain fatty acids (% of total FA) in milk from goats fed either palm (POFA) or rapeseed oil (RSO) or a control
(CON) feed with no extra fat

Preparatory period
(DIM) Experimental period (DIM) P-value

Feed ×
Item   Feed 301 601   90 120 1902 230 SEM Feed DIM DIM
C11:0 CON 0.70 0.44 0.42a,y 0.37a,y 0.27z 0.69a,x 0.03 0.003 <0.001 0.701
POFA 0.60 0.50 0.28b,y 0.24b,y 0.22y 0.56b,x
RSO 0.57 0.40 0.30b,y 0.26b,yz 0.21z 0.58b,x
SEM3 0.061 0.054
C14:0 iso CON 0.18 0.14 0.12a,y 0.13xy 0.08z 0.15a,x 0.01 0.046 <0.001 0.037
POFA 0.15 0.15 0.06b,z 0.12x 0.09yz 0.11b,xy
RSO 0.13 0.13 0.10a,y 0.12x 0.09y 0.13ab,x
SEM3 0.026 0.015
C15:0 iso CON 0.31 0.34 0.32a,y 0.38a,x 0.30y 0.28z 0.02 <0.001 <0.001 0.028
POFA 0.25 0.40 0.23b,z 0.29b,y 0.25y 0.27y
RSO 0.28 0.35 0.25b,z 0.34ab,y 0.24z 0.28z
SEM3 0.027 0.027
C15:0 anteiso CON 0.26 0.34 0.25a,z 0.37a,y 0.27a,z 0.34a,y 0.02 <0.001 <0.001 0.215
POFA 0.28 0.34 0.17b,z 0.25b,xy 0.21b,y 0.28b,x
RSO 0.23 0.34 0.19b,z 0.29b,y 0.23ab,z 0.31ab,y
SEM3 0.040 0.018
C15:0 CON 0.65 0.73 0.77a,y 0.77a,y 0.63a,z 0.85a,x 0.02 <0.001 <0.001 0.001
POFA 0.62 0.75 0.55b,y 0.54c,y 0.52b,y 0.71b,x
RSO 0.66 0.78 0.60b,y 0.61b,y 0.55b,z 0.76b,x
SEM3 0.036 0.044
C16:0 iso CON 0.20 0.24 0.28a,x 0.26a,x 0.15a,z 0.18a,y 0.01 <0.001 <0.001 0.013
POFA 0.20 0.23 0.18b,x 0.19b,x 0.12b,y 0.14b,y
RSO 0.22 0.23 0.21b,x 0.20b,x 0.12b,z 0.16ab,y
SEM3 0.016 0.017
C17:0 iso CON 0.29 0.43 0.41a,y 0.50a,x 0.42a,y 0.24z 0.02 <0.001 <0.001 0.005
POFA 0.33 0.38 0.30b,y 0.41b,x 0.35b,y 0.22z
RSO 0.32 0.39 0.43a,x 0.55a,w 0.38a,y 0.25z
SEM3 0.038 0.029
C17:0 anteiso CON 0.57 0.32 0.36a,y 0.34a,y 0.24b,z 0.35y 0.02 0.021 <0.001 <0.001
POFA 0.52 0.32 0.23b,z 0.30ab,y 0.23b,z 0.35y
RSO 0.55 0.35 0.27b,yz 0.26b,z 0.31a,yx 0.32x
SEM3 0.031 0.026
C17:0 CON 0.97 0.94 0.86a,z 0.92a,y 0.96a,x 1.01a,w 0.02 <0.001 <0.001 0.003
POFA 1.00 0.95 0.53c,z 0.58c,y 0.74c,x 0.71c,x
RSO 1.00 0.97 0.64b,z 0.70b,y 0.84b,w 0.79b,x
SEM3 0.018 0.024
C17:1 cis-9 CON 0.54 0.54 0.29a,z 0.39a,y 0.28a,z 0.32a,z 0.02 <0.001 <0.001 0.424
POFA 0.55 0.54 0.23b,z 0.32b,y 0.25ab,z 0.28ab,y
RSO 0.55 0.47 0.20b,z 0.27b,y 0.21b,z 0.26b,y
SEM3 0.049 0.044  
a–c
LSM within a column with different letters are significantly different (P < 0.05) from each other.
w–z
LSM within a row with different superscript letters are significantly different (P < 0.05) from each other.
1
All goats received the control concentrate in the preparatory period.
2
Mountain pasture.
3
Standard error of mean for the preparatory period.

enced by feed, DIM, and genotype. The proportion of been reported in bovine milk (Wiking et al., 2004).
goat fatty acids was higher in CON milk fat (Table 6) A decline in MFG size with increasing intake of fresh
and in milk fat from goats of E12–01 genotype, whereas grass (Couvreur et al., 2006) and unsaturated lipids
C16:0 was higher in milk fat from E12–00 goats (Table (Lopez et al., 2008) has been reported in cow milk.
7). However, we observed no decrease in MFG size at the
end of the grazing period (at 190 DIM, Table 2). The
Mean Diameter of the MFG average size of MFG has been reported to vary between
goat breeds; smaller MFG were reported in milk from
The mean size of the MFG decreased during lacta- French Alpine (2.76 µm; Attaie and Richter, 2000) and
tion from 3.46 µm at 30 DIM to 2.88 µm at 230 DIM, Sarda (2.73 µm) than in Saanen (3.63 µm; Pisanu et al.,
and a similar effect of lactation on the MFG size has 2013). The MFG size of Norwegian dairy goats seems
Journal of Dairy Science Vol. 100 No. 9, 2017
7096 INGLINGSTAD ET AL.

to be in between those breeds, with an average of 3.2 positive relationship between fat content and the mean
µm in mid lactation (120 DIM). The above-mentioned diameter of the MFG of goats (Cebo et al., 2012), cows
effects of breed, diet, and lactation stage are probably (Wiking et al., 2004), and buffaloes (Ménard et al.,
confounded with fat content. Several authors report a 2010), a trend also confirmed in our study (r = 0.72,

Table 5. Long-chain fatty acids (% of total FA in triglycerides) in milk from goats fed either palm (POFA) or rapeseed oil (RSO), or a control
(CON) feed with no extra fat

Preparatory period
(DIM) Experimental period (DIM) P-value

Feed ×
Item   Feed 301 601   90 120 1902 230 SEM Feed DIM DIM  
c,y c,x b,w b,z
C18:0 CON 9.89 9.14 7.82 8.71 15.02 4.71 0.46 <0.001 <0.001 <0.001
POFA 9.20 9.61 9.45b,y 10.65b,x 14.18b,w 6.13b,z
RSO 9.13 9.34 15.76a,y 17.02a,x 19.54a,w 9.48a,z
SEM3 1.096 0.640
C18:1 cis-9 CON 21.10 19.65 16.25c,y 19.54c,x 21.48b,w 14.55c,z 0.55 <0.001 <0.001 <0.001
POFA 21.98 18.96 18.65b,z 22.64b,x 21.11b,y 18.39b,z
RSO 21.43 18.50 24.39a,y 27.93a,x 24.88a,y 22.28a,z
SEM3 1.381 1.227
C18:1 trans-9 CON 0.33 0.33 0.26c,y 0.31c,y 0.40c,x 0.16c,z 0.02 <0.001 <0.001 <0.001
POFA 0.33 0.31 0.52b,y 0.65b,x 0.57b,y 0.40b,z
RSO 0.30 0.33 0.90a,x 1.01a,w 0.80a,y 0.49a,z
SEM3 0.042 0.031
C18:1 trans-11 CON 1.16 1.17 0.90b,y 1.21b,x 1.54b,w 0.64b,z 0.09 <0.001 <0.001 <0.001
POFA 1.17 1.31 0.56c,z 0.72c,z 1.25c,y 0.58b,z
RSO 1.26 1.25 1.70a,y 1.89a,y 2.24a,x 1.09a,z
SEM3 0.063 0.060
C18:2 cis-9,12 CON 1.50 1.46 1.30a,y 1.49a,x 2.56a,w 1.05a,z 0.06 <0.001 <0.001 <0.001
POFA 1.48 1.34 1.00b,y 1.19b,x 1.96b,w 0.86b,z
RSO 1.54 1.42 1.31a,y 1.54a,x 2.07b,w 1.06a,z
SEM3 0.066 0.082        
C18:2 trans-9,12 CON 0.17 0.17 0.13b,z 0.18b,y 0.18b,y 0.12b,z 0.01 <0.001 <0.001 0.004
POFA 0.15 0.18 0.12b,z 0.16b,y 0.15b,y 0.12b,z
RSO 0.17 0.17 0.26a,z 0.34a,y 0.27a,z 0.28a,z
SEM3 0.012 0.014        
C18:2 cis-9,trans-11 CON 0.58 0.60 0.54b,z 0.77a,x 0.64a,y 0.59a,yz 0.04 <0.001 <0.001 0.019
POFA 0.58 0.61 0.31c,z 0.45b,y 0.49b,y 0.43b,y
RSO 0.62 0.59 0.66a,z 0.86a,y 0.71a,z 0.63a,z
SEM3 0.034 0.034        
C18:3 cis-9,12,15 CON 0.41 0.42 0.32a,z 0.42a,y 0.94a,x 0.32a,z 0.02 <0.001 <0.001 <0.001
POFA 0.47 0.44 0.21b,z 0.30b,y 0.67c,x 0.24b,z
RSO 0.46 0.43 0.34a,z 0.46a,y 0.77b,x 0.35a,z
SEM3 0.049 0.023        
C20:0 CON 0.20 0.22 0.18b,y 0.26b,y 0.87a,x 0.18c,y 0.03 <0.001 <0.001 <0.001
POFA 0.17 0.21 0.18b,y 0.26b,y 0.66b,x 0.22bc,y
RSO 0.19 0.22 0.37a,z 0.50a,y 0.82a,x 0.32a,z
SEM3 0.025 0.015        
C20:1 cis-11 CON 0.17 0.10 0.05b,z 0.12b,y 0.10y 0.04b,z 0.01 <0.001 <0.001 0.002
POFA 0.15 0.15 0.07b,z 0.11b,y 0.12y 0.06ab,z
RSO 0.13 0.11 0.12a,z 0.22a,y 0.11z 0.09a,z
SEM3 0.044 0.025        
C20:4 cis-5,8,11,14 CON 0.19 0.11 0.10ab 0.13 0.13 0.09 0.02 0.119 0.067 0.361
POFA 0.13 0.11 0.05b,z 0.07 0.13y 0.04z
RSO 0.15 0.12 0.13a 0.11 0.10 0.07
SEM3 0.023 0.016        
C22:0 CON 0.18 0.15 0.16a,y 0.18b,y 0.29a,x 0.12b,z 0.01 <0.001 <0.001 0.456
POFA 0.14 0.15 0.12b,yz 0.1c,y 0.24b,x 0.08c,z
RSO 0.15 0.12 0.18a,y 0.23a,x 0.29a,w 0.14a,z
SEM3 0.038 0.017          
a–c
LSM within a column with different letters are significantly different (P < 0.05) from each other.
w–z
LSM within a row with different superscript letters are significantly different (P < 0.05) from each other.
1
All goats received the control concentrate in the preparatory period.
2
Mountain pasture.
3
Standard error of mean for the preparatory period.

Journal of Dairy Science Vol. 100 No. 9, 2017

 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7097

n = 186; P < 0.0001). This also explains why MFG ported to have larger MFG than those of the “weak”
size did not decline during the grazing period, because (0/0) genotype (Cebo et al., 2012); however, the effect
the fat content in the milk increased at this stage. The of genotype in the present study was not significant
total surface area of the MFG increases with decreasing (Table 7).
size, and MFGM material may be a limiting factor. In
this way, an increase in fat content would explain the LPL Activity and Lipolysis
increase in MFG size.
Polymorphism in CSN1S1 may also influence MFG Lipoprotein lipase activity decreased between 120
size as goats of the “strong” genotype (A/A) are re- and 230 DIM (Table 2), which is in agreement with a

Table 6. Fatty acid (FA) groups in milk fat from goats fed either palm (POFA) or rapeseed oil (RSO) or a control (CON) feed with no extra fat

Preparatory period Experimental period P-value

Feed ×
Item   Feed1 303 603   90 120 1904 230 SEM2 Feed DIM DIM
Goat FA5 CON 13.69 13.20 12.79a,x 10.47a,y 8.76z 12.54a,x 0.40 <0.001 <0.001 <0.001
POFA 14.12 13.38 9.02c,y 7.07b,z 6.97z 10.48b,x
RSO 14.12 14.07 11.04b,y 8.43ab,z 7.95z 11.85a,x        
SEM6 2.35 2.52                
Short- and medium- CON 35.52 33.79 33.46a,x 28.85a,y 23.41a,z 35.69a,w 0.737 <0.001 <0.001 <0.001
chain7 POFA 35.65 35.08 23.75c,x 20.24c,y 19.22b,y 29.29c,w
RSO 35.97 35.83 27.14b,x 22.66b,y 20.89b,z 31.98b,w        
SEM6 1.669 1.702                
8
C16 CON 25.87 30.08 36.22b,x 35.01b,x 29.84b,y 39.32a,w 0.771 <0.001 <0.001 <0.001
POFA 25.68 28.95 43.69a,w 40.54a,x 37.40a,y 41.44a,x
RSO 25.76 28.99 24.95c,x 22.91c,y 24.32c,xy 29.82b,w        
SEM6 0.864 0.781                
Odd- and branched- CON 4.67 4.42 4.09a,x 4.44a,w 3.61a,y 4.42a,w 0.080 <0.001 <0.001 <0.001
chain9 POFA 4.50 4.56 2.70c,z 3.19c,x 2.93c,y 3.57c,w
RSO 4.50 4.45 3.20b,y 3.61b,x 3.20b,y 3.85b,w        
SEM6 0.123 0.117                
10
Long SFA CON 11.23 10.45 9.02b,y 10.06c,x 17.15b,w 6.03b,z 0.478 <0.001 <0.001 <0.001
POFA 10.51 10.92 10.28b,y 11.62b,x 15.83b,w 7.14b,z
RSO 10.47 10.65 16.95a,y 18.45a,x 21.49a,w 10.73a,z        
SEM6 1.089 0.657                
11
Long UFA CON 26.13 24.54 20.13b,y 24.56c,x 28.25b,w 17.88c,z 0.665 <0.001 <0.001 <0.001
POFA 26.99 23.96 21.74b,x 26.63b,w 26.73b,w 21.44b,x
RSO 26.60 23.38 30.02a,y 34.63a,w 32.17a,x 26.61a,z        
SEM6 1.550 1.377                
Σ MUFA CON 24.55 23.45 18.81c,y 22.68c,x 24.78b,w 17.18c,z 0.59 <0.001 <0.001 <0.001
POFA 25.40 22.96 21.29b,z 25.69b,y 24.30b,y 21.13b,z
RSO 24.87 22.22 28.25a,y 32.29a,x 29.11a,y 25.24a,z        
SEM6 4.79 4.26                
Σ PUFA CON 2.84 2.75 2.28b,z 2.86b,y 4.32a,x 2.08a,z 0.10 <0.001 <0.001 <0.001
POFA 2.81 2.68 1.66c,z 2.10c,y 3.29c,x 1.67b,z
RSO 2.94 2.73 2.58a,y 3.20a,x 3.83b,w 2.32a,z        
SEM6 0.39 0.32                
a–c
Means within a column with different superscripts differ (P < 0.05).
w–z
Means within a row with different superscripts differ (P < 0.05).
1
CON = basal concentrate containing no additional fat; POFA = basal concentrate supplemented with hydrogenated palm oil enriched with
palmitic acid; RSO = basal concentrate supplemented with rapeseed oil.
2
Standard error of means for the experimental period.
3
All goats received the control concentrate in the preparatory period.
4
Mountain pasture.
5
Sum of C6:0, C8:0, and C10:0.
6
Standard error of mean for the preparatory period.
7
Sum of C4:0, C6:0, C8:0, C10:0, C11:0, C12:0, C14:0 iso, C14:0, C15:0 iso, C15:0 anteiso, C14:1 cis-9, and C15:0.
8
Sum of C16:0 iso, C16:0, C16:1 trans-9, and C16:1 cis-9.
9
Sum of C11:0, C14:0 iso, C15:0 iso, C15:0 anteiso, C15:0, C16:0 iso, C17:0 iso, C17:0 anteiso, C17:0, and C17:1 cis-9.
10
Sum of C17:0, C18:0, C20:0, and C22:0.
11
Sum of C17:1 cis-9, C18:1 trans-9, C18:1 trans-11, C18:1 cis-9, C20:1 cis-11, C18:2 trans-9,12, C18:2 cis-9,12, C18:3 cis-9,12,15, C18:2 cis-
9,trans-11, and C20:4 cis-5,8,11,14.

Journal of Dairy Science Vol. 100 No. 9, 2017

7098 INGLINGSTAD ET AL.

Figure 1. Principal component analysis (PCA) of the triglyceride (TG) fatty acid profile in milk at 90 DIM. The score plot (top panel)
shows the distribution of the samples indicated by feeding groups: control feed with no extra fat (■); basal concentrate supplemented with
hydrogenated palm oil enriched with palmitic acid (♦); or basal concentrate supplemented with rapeseed oil (●). The corresponding loading
plot (bottom panel) shows the distribution of variables (fatty acids; c = cis, t = trans, ant = anteiso). More than 70% of the variation was
explained by principal components (PC) 1 and 2. The PCA revealed a similar distribution of the loadings and scores at 120 to 230 DIM. Color
version available online.

Journal of Dairy Science Vol. 100 No. 9, 2017

 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7099
Table 7. Main effects on selected parameters in milk from goats either homozygous (E12–00) or heterozygous (E12–01) for the deletion in exon
12 in CSN1S1 and fed either palm (POFA) or rapeseed oil (RSO) or a control (CON) feed with no extra fat

Genotype P-value

Feed ×
Item1   Feed E12–00 E12–01 SEM Feed Genotype Genotype
Fat (g/kg) CON 34.2 33.5b        
POFA 39.5 43.0a 1.98 <0.001 0.287 0.448
RSO 35.7 37.3b        
MFG (d43; μm) CON 2.96 3.00b        
POFA 3.15 3.28a 0.115 0.049 0.304 0.893
RSO 3.12 3.21ab        
LPL activity (mmol/min) CON 393a 387a        
POFA 367ab 368a 8.37 0.002 0.505 0.929
RSO 355b 352b        
FFA84h (mM) CON 1.27ab 0.94        
POFA 1.78a,w 0.91x 0.271 0.008 0.021 0.294
RSO 0.67b 0.49        
Flavor score CON 2.87b 3.62        
POFA 2.69b,x 4.22w 0.469 0.009 0.029 0.161
RSO 4.51a 4.49        
C6:0 (%) CON 1.85a 1.90a        
POFA 1.61b 1.75b 0.072 0.006 0.029 0.676
RSO 1.77ab 1.92a        
C8:0 (%) CON 1.72a 1.91a        
POFA 1.35b 1.51b 0.106 <0.001 0.008 0.771
RSO 1.58ab 1.86a        
C10:0 (%) CON 7.06a 7.85a        
POFA 5.04b 5.52c 0.409 <0.001 0.012 0.704
RSO 5.72b 6.78b        
C12:0 (%) CON 3.25a 3.75a        
POFA 2.32b 2.46c 0.214 <0.001 0.037 0.601
RSO 2.53b 2.88b        
C16:0 (%) CON 36.24b,w 31.42b,x        
POFA 39.96a 38.91a 0.962 <0.001 <0.001 0.090
RSO 25.98c,w 22.90c,x        
C18:0 (%) CON 8.60b 9.53c        
POFA 9.41b 10.80b 0.639 <0.001 0.005 0.723
RSO 14.55a 16.35a        
Goat FA (%) CON 10.62a 11.65a        
POFA 8.00b 8.78b 0.570 <0.001 0.010 0.741
RSO 9.06b,x 10.58a,w        
a–c
Means within a column with different superscripts differ (P < 0.05).
w,x
Means within a row with different superscripts differ (P < 0.05).
1
MFG = milk fat globule, LPL = lipoprotein lipase, FFA84h = free fatty acids after 84 h of storage; Goat FA = sum of C6:0, C8:0, and C10:0.

study on cow milk (Chazal and Chilliard, 1986) but is content of FFA in milk samples that were pasteurized
contrary to results obtained in cow milk cream (Ah- shortly after milking (at time 0, before the milk was
rné and Björck, 1985). Previous studies on goat milk cooled) was low; however, it varied among individual
showed low activity in early and late lactation and high goat milk samples. The increase in the content of
LPL activity between 60 and 210 DIM (Chilliard et FFA from 0 to 84 h after milking is shown in Figure
al., 2003). The RSO milk had lower LPL activity than 2. Post-milking lipolysis was lower in RSO milk at 90
CON (P = 0.001) and POFA (P = 0.02), which is sup- to 120 DIM, compared with POFA and CON milk, as
ported by other studies also showing a lower LPL activ- previously observed (Chilliard et al., 2003; Ollier et al.,
ity when goats were fed unsaturated lipids (Bernard 2009). This is likely to be linked to the decrease in
et al., 2005; Ollier et al., 2009). This decrease is likely milk LPL activity that was observed in this feeding
to be due to mammary LPL being more orientated to- group. The positive effect of inclusion of rapeseed oil
ward blood capillary (for TG fatty acid uptake) than in the feed seems to decline after 120 DIM as the dif-
toward milk secretion when goats were fed unsaturated ferences in post-milking lipolysis were smaller between
lipids (Chilliard et al., 2003). The LPL activity did not the feeding groups at 190 and 230 DIM (Figure 2). This
vary between E12–00 and E12–01 goats (Table 7). The decrease in the effect of RSO on milk lipolysis in late

Journal of Dairy Science Vol. 100 No. 9, 2017

7100 INGLINGSTAD ET AL.

(vs. peak) lactation is likely to be due to the simultane- with milk from the CON (1.7 mM, P = 0.05) and POFA
ous decrease in lipolysis of the CON group at 190 and (1.9 mM, P = 0.02) groups at 90 DIM, when the prob-
230 DIM. lem with high level of FFA was most prominent (Figure
3A). The content of FFA did not increase while the
Free Fatty Acids and Milk Flavor goats were grazing mountain pasture (Figure 3) which
is in line with results from another study (Steinshamn
Off-flavors caused by high concentrations of FFA is et al., 2014). However, the concentration of FFA was
challenging for the development of goat milk products. higher in E12–00 goats compared with E12–01 goats, as
Concentrations of FFA above 1.2 mM make the milk expected (Table 6; Dagnachew et al., 2011).
unsuitable for human consumption. Lactation stage, feed group, and genotype all influ-
In early lactation, the content of FFA was low in gen- enced milk flavor. The best flavor scores were obtained
eral, although large variations were seen among goats in early and late lactation and from E12–01 goats,
(from 0.1 to 1.4 mM at 30 DIM). The FFA content whereas milk at mid lactation and from E12–00 goats
peaked in mid lactation (90 DIM, Figure 3A), which is received lower flavor scores (Figure 3B, Table 6). Milk
in agreement with Chilliard et al. (2003). At this stage, samples with low flavor scores (1 and 2) were described
some individual goats displayed an extremely high con- as tart or rancid. The average sensory score for RSO
centration (up to 4.3 mM) of FFA in milk. However, milk was 4.5, which was higher than for CON (3.3, P
other goats produced milk with a low concentration = 0.004) and POFA (3.5, P = 0.01) milks (Figure 3B).
(0.1 mM) of FFA throughout the lactation, even in mid This conflicts with results from another study, where
lactation. Interestingly, milk from the RSO group had a surplus of saturated fat from palm oil in the diet
a remarkably lower content of FFA (0.8 mM) compared decreased off-flavors in the milk compared with a diet

Figure 2. Milk lipolysis as shown by free fatty acid (FFA) content measured after 0, 36, and 84 h of storage in milk from goats receiving
different lipid-supplemented concentrates [control feed with no extra fat (—■—); basal concentrate supplemented with hydrogenated palm oil
enriched with palmitic acid (···♦···); or basal concentrate supplemented with rapeseed oil (– –●– –)] from 90 to 230 DIM. Different letters indi-
cate significant differences (P < 0.05) between the feeding groups.

Journal of Dairy Science Vol. 100 No. 9, 2017

 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7101

enriched with sunflower oil (rich in UFA; Eknæs et al., thresholds (Brennand et al., 1989), and are believed
2009). However, other studies have shown that supple- to be strong contributors to the flavors of goat milk
mentation with unsaturated fat may reduce LPL activ- (Chilliard et al., 2003). More than 20 of the FFA were
ity, content of FFA, and goaty flavor in milk (Skjevdal, analyzed to identify specific fatty acids responsible for
1979; Chilliard et al., 2003). Indeed, previous studies the tart and rancid flavor; however, no branched-chain
have shown a correlation between the content of FFA in variants of caprylic acid were identified in our samples.
milk and off-flavors (Deeth and Fitz-Gerald, 2006), and The fatty acid profiles of the FFA fraction appeared
this is supported by the current experiment (Figures 3 similar to the profiles of TG; however, the difference
and 4, r = −0.84). The content of FFA obtained from between samples in total content was much greater.
the routine milk analysis (Milkoscan/FTIR) therefore Samples with a high content of FFA received low flavor
gives a good indication of the sensory quality of the scores and were characterized as tart or rancid (Figure
milk concerning off-flavors. 4). Those samples had a high content of all the identi-
Liberation of short- and medium-chain fatty acids fied fatty acids. Thus, we were not able to link any
(C6:0–C10:0) is believed to be responsible for the spe- specific fatty acid(s) to either goaty flavor or to specific
cific goaty flavor (Brandsaeter and Abate, 1959), and off-flavors (tart or rancid). The total number of FFA
perhaps also tart or rancid flavors (Deeth and Fitz- measured by FTIR from the routine analysis was a bet-
Gerald, 2006). Branched-chain variants (methyl and ter indicator for off-flavors than the concentration of
ethyl) of caprylic acid are volatiles with very low flavor the individual FFA (Figure 4).
The FAME profiles of the FFA fraction in 2 samples
with extremely high (4.3 mM) and low (0.1 mM) con-
tents of FFA are shown in Figure 5. The sample with
the extremely high content of FFA (A) contained peaks
of both short- and long-chain fatty acids, whereas the
profile of the sample with a low content of FFA con-
tained only fatty acids originating from polypropylene
contamination (details in Materials and Methods).
Some goats produced milk with a high (>0.8 mM)
concentration of FFA throughout the whole lactation
period, whereas others had a high concentration only
in mid lactation but acceptable levels at the start and
end of lactation. Moreover, some goats consistently
produced milk with a good flavor and low content of
FFA. We know from previous studies that a genetic fac-
tor (of the CSN1S1 locus) is linked to high content of
FFA in the milk of Norwegian dairy goats (Dagnachew
et al., 2011; Dagnachew and Ådnøy, 2014); however, in
our study, we observed that goats of genotypes other
than E12–00 may also produce milk with a high con-
tent of FFA and off-flavors. Milk LPL activity and FFA
content are reported to be higher in French goats of
“weak” genotypes of CSN1S1 (Chilliard et al., 2003);
however, no correlation was found in the present study
between the LPL activity and content of FFA in milk
(r = −0.187, n = 155) nor to the genotype of CSN1S1
(Table 6).

Composition of Phospholipids and Cholesterol

Figure 3. (A) Free fatty acid (FFA) content, and (B) milk fla- of the MFGM in Selected Samples
vor scores measured in milk stored for 84 h. Samples were measured
throughout a whole lactation from goats receiving different lipid-sup-
plemented concentrates [control feed with no extra fat (—■—); basal As differences in LPL activity cannot explain the in-
concentrate supplemented with hydrogenated palm oil enriched with creased FFA content in some samples, we hypothesize
palmitic acid (···♦···); or basal concentrate supplemented with rape- that substrate availability may explain the differences
seed oil (– –●– –)] from 61 to 230 DIM. The vertical line indicates start
of the experimental period. Values at 30 and 60 DIM are means; values in FFA content among the milk samples. The MFGM
from 90 to 230 DIM are LSM. Error bars represent SEM. protects TG from lipolytic degradation, and the stabil-
Journal of Dairy Science Vol. 100 No. 9, 2017
7102 INGLINGSTAD ET AL.

Figure 4. Partial least squares regression of free fatty acid (FFA) profile, total FFA, and flavor scores of all samples from 30 to 230 DIM.
The score plot (top panel) shows the distribution of samples indicated by their flavor score [1–5, where 5 is best (no off-flavors) and 1 indicates
high degree of off-flavors] and the corresponding loading plot (bottom panel) shows the distribution of variables (fatty acids, c = cis, t = trans;
flavors). The first principal component (PC 1) explains 64 and 50% of the variation of the x and y variables, respectively, and PC 2 explains 11
and 15% of the variations in the x and y variables, respectively. Color version available online.

Journal of Dairy Science Vol. 100 No. 9, 2017

 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7103

Figure 5. Comparison of 2 gas chromatograms of the fraction of free fatty acids (FFA) in goat milk: (A) extremely high (4.3 mM), and (B)
low (0.1 mM) total content of FFA; c = cis, t = trans, aiso = anteiso. Color version available online.

ity of the MFGM may depend on its composition. We The goat milk samples with a high degree of lipolysis
therefore examined possible differences in phospholipid also displayed a low content of phosphatidylethanol-
and cholesterol composition of the MFGM to explain amine. Phosphatidylinositol and phosphatidylserine
the different degree of lipolysis in the milk samples. were not detected in any of the samples. The relation-
Examination of the phospholipid composition and cho- ship between lipolysis and MFGM composition should
lesterol content in 6 goat milk samples revealed differ- be further investigated to unravel the mechanism of
ences between milk with a high content of FFA (2.9 ± lipolysis in some milk samples.
1.25 mM) versus those with a low content of FFA (0.5
± 0.17 mM; Figure 6). Samples with a high content of
FFA had a higher content of cholesterol (3.41 ± 0.21
mg/g of fat) compared with those with low content of
FFA (2.72 ± 1.79 mg/g of fat; P = 0.04). Cholesterol
and sphingomyelin are the major constituents of lipid
rafts in MFGM, which are the structures involved in
different cellular processes, and cholesterol is reported
to affect the MFGM organization (Murthy et al., 2015).
A hydrolyzed variant of phosphatidylcholine, lysophos-
phatidylcholine (also called lysolecithin), was found in
higher concentrations in samples with a high content
of FFA (5.07 vs. 1.78 mg/g of fat); however, the dif-
ference was not statistically significant (P = 0.15).
Lysophospholipids are known to have a strong affinity
for both LPL and lipoproteins and may aid the LPL
by disruption of the MFGM (Deeth and Fitz-Gerald, Figure 6. Content of polar lipids and cholesterol in milk from
1983). Sundheim et al. (1983) showed that exogenous goats with high degree of lipolysis (white) and low degree of lipolysis
lysophosphatidylcholine enhances cow milk lipolysis (black). FFA = free fatty acids, PE = phosphatidyletanolamine, LPE
= lysophosphatidyletanolamine, PC = phosphatidylcholine, SPM =
when activated by blood serum, but not in milk with- sphingomyelin, LPC = lysophosphatidylcholine, CHOL = cholesterol.
out serum addition. Values are means ± SD; ns = P > 0.05, *P < 0.05.

Journal of Dairy Science Vol. 100 No. 9, 2017

7104 INGLINGSTAD ET AL.

CONCLUSIONS Attaie, R., and R. L. Richter. 2000. Size distribution of fat globules in
goat milk. J. Dairy Sci. 83:940–944. https://doi.org/10.3168/jds.
S0022-0302(00)74957-5.
This study showed that it is possible to alter the milk Bauman, D. E., and J. M. Griinari. 2003. Nutritional regulation of
fat content and composition of goat milk by altering milk fat synthesis. Annu. Rev. Nutr. 23:203–227.
the lipid composition of the concentrate fed. Feeding Bernard, L., J. Rouel, C. Leroux, A. Ferlay, Y. Faulconnier, P. Le-
grand, and Y. Chilliard. 2005. Mammary lipid metabolism and
goats with a concentrate supplemented with rapeseed milk fatty acid secretion in alpine goats fed vegetable lipids. J.
oil increased the content of most UFA and decreased Dairy Sci. 88:1478–1489. https://doi.org/10.3168/jds.S0022-
the content of SFA in goat milk. In addition, rape- 0302(05)72816-2.
Brandsaeter, E., and V. Abate. 1959. Flavour compounds in goat’s
seed oil supplements resulted in lower LPL activity in milk and goat whey cheese. 2. Occurrence of free fatty acids in goat
the milk, lower content of FFA, and a higher flavor whey cheese. Meldinger fra Norges Landbrukshogskole 38:1–12.
score compared with milk from goats receiving palm Brennand, C. P., J. K. Ha, and R. C. Lindsay. 1989. Aroma properties
and thresholds of some branched-chain and other minor volatile
oil supplements or goats receiving a diet without fat fatty acids occurring in milkfat and meat lipids. J. Sens. Stud.
supplements. Both types of oil increased the fat content 4:105–120. https://doi.org/10.1111/j.1745-459X.1989.tb00461.x.
in the milk, which correlates positively with the milk Cebo, C., C. Lopez, C. Henry, C. Beauvallet, O. Ménard, C. Bevilac-
qua, F. Bouvier, H. Caillat, and P. Martin. 2012. Goat αs1-casein
fat globule size. The presence of off-flavors and the total genotype affects milk fat globule physicochemical properties and
content of FFA in milk detected by the routine analysis the composition of the milk fat globule membrane. J. Dairy Sci.
(Milkoscan) were highly correlated. The positive effect 95:6215–6229. https://doi.org/10.3168/jds.2011-5233.
Chazal, M. P., and Y. Chilliard. 1986. Effect of stage of lactation,
of inclusion of rapeseed oil in the diet of goats has po- stage of pregnancy, milk yield and herd management on seasonal
tential for development of new feeding strategies with variation in spontaneous lipolysis in bovine milk. J. Dairy Res.
feed based on sustainably produced lipid sources. 53:529–538.
Chilliard, Y., A. Ferlay, J. Rouel, and G. Lamberet. 2003. A review
of nutritional and physiological factors affecting goat milk lipid
ACKNOWLEDGMENTS synthesis and lipolysis. J. Dairy Sci. 86:1751–1770.
Chilliard, Y., F. Glasser, A. Ferlay, L. Bernard, J. Rouel, and M.
Doreau. 2007. Diet, rumen biohydrogenation and nutritional qual-
The Norwegian Research Council (NFR, Lysaker, ity of cow and goat milk fat. Eur. J. Lipid Sci. Technol. 109:828–
Norway ) and TINE BA (Oslo, Norway) financed this 855. https://doi.org/10.1002/ejlt.200700080.
study. We thank the staff at the Animal Production Coppa, M., A. Ferlay, F. Monsallier, I. Verdier-Metz, P. Pradel, R.
Didienne, A. Farruggia, M. C. Montel, and B. Martin. 2011. Milk
Experimental Centre of NMBU for animal care and fatty acid composition and cheese texture and appearance from
assistance with sampling procedures, Irene Comi (De- cows fed hay or different grazing systems on upland pastures. J.
partment of Chemistry, Biotechnology and Food Sci- Dairy Sci. 94:1132–1145. https://doi.org/10.3168/jds.2010-3510.
Couvreur, S., C. Hurtaud, C. Lopez, L. Delaby, and J. L. Peyraud.
ences, NMBU, Norway) for technical assistance during 2006. The linear relationship between the proportion of fresh grass
milk sampling, Claes-Gøran Fristedt (Department of in the cow diet, milk fatty acid composition, and butter prop-
Animal and Aquacultural Sciences, NMBU, Norway) erties. J. Dairy Sci. 89:1956–1969. https://doi.org/10.3168/jds.
S0022-0302(06)72263-9.
for measurements of the milk fat globule size, Kari Dagnachew, B. S., and T. Ådnøy. 2014. Additive and dominance ef-
Olsen (Department of Chemistry, Biotechnology and fects of casein haplotypes on milk composition and quality in Nor-
Food Sciences, NMBU, Norway) for assistance in the wegian dairy goats. Small Rumin. Res. 122:59–69. https://doi.
org/10.1016/j.smallrumres.2014.07.020.
GC-FID analysis of the fatty acids, and Cyril Labonne Dagnachew, B. S., G. Thaller, S. Lien, and T. Adnoy. 2011. Casein
(INRA, France) for assistance in LPL analysis. Harald SNP in Norwegian goats: additive and dominance effects on milk
Volden and Tormod Ådnøy (Department of Animal and composition and quality. Genet. Sel. Evol. 43:31.
Deeth, H., and C. Fitz-Gerald. 1983. Lipolytic enzymes and hydrolytic
Aquacultural Sciences, NMBU, Norway) are acknowl- rancidity in milk and milk products. Pages 195–239 in Develop-
edged for their advice regarding the experimental de- ments in Dairy Chemistry—2. Springer, New York, NY.
sign and statistical analysis, respectively. We are very Deeth, H. C. 2006. Lipoprotein lipase and lipolysis in milk. Int. Dairy
J. 16:555–562.
grateful for the sensory evaluation of the milk samples Deeth, H. C., and C. H. Fitz-Gerald. 2006. Lipolytic enzymes and
performed by Knut Erik Grindaker, Helga Kvamsås, hydrolytic rancidity. Pages 481–556 in Advanced Dairy Chemistry.
and Kåre Johan Vassbotn (TINE). TINE provided Vol. 2: Lipids. P. F. Fox and P. L. H. McSweeney, ed. Springer,
New York, NY.
Milkoscan analysis free of charge. Devle, H., E. K. Ulleberg, C. F. Naess-Andresen, E.-O. Rukke, G.
Vegarud, and D. Ekeberg. 2014. Reciprocal interacting effects of
proteins and lipids during ex vivo digestion of bovine milk. Int.
REFERENCES Dairy J. 36:6–13. https://doi.org/10.1016/j.idairyj.2013.11.008.
Dønnem, I., Å. T. Randby, and M. Eknæs. 2011. Effect of grass silage
Ahrné, L., and L. Björck. 1985. Lipolysis and the distribution of li-
harvesting time and level of concentrate supplementation on goat
pase activity in bovine milk in relation to stage of lactation and
milk quality. Anim. Feed Sci. Technol. 163:118–129. https://doi.
time of milking. J. Dairy Res. 52:55–64. https://doi.org/10.1017/
org/10.1016/j.anifeedsci.2010.10.013.
S002202990002389X.
Eknæs, M., Y. Chilliard, K. Hove, R. A. Inglingstad, L. Bernard, and
Andrade, P., and P. Schmidely. 2006. Influence of percentage of con-
H. Volden. 2017. Effects of feeding palm oil fatty acids or rape-
centrate in combination with rolled canola seeds on performance,
seed oil throughout lactation on energy status, body composi-
rumen fermentation, and milk fatty acid composition in dairy
goats. Livest. Sci. 104:77–90.

Journal of Dairy Science Vol. 100 No. 9, 2017

 GOAT MILK FLAVOR AND FATTY ACID COMPOSITION 7105
tion and milk production in Norwegian dairy goats. J. Dairy Sci. chim. Biophys. Acta 1848:2308–2316. https://doi.org/10.1016/j.
100:7588–7601. https://doi.org/10.3168/jds.2017-12768. bbamem.2015.06.014.
Eknæs, M., K. Kolstad, H. Volden, and K. Hove. 2006. Changes in Ollier, S., C. Leroux, A. de la Foye, L. Bernard, J. Rouel, and Y.
body reserves and milk quality throughout lactation in dairy Chilliard. 2009. Whole intact rapeseeds or sunflower oil in high-
goats. Small Rumin. Res. 63:1–11. forage or high-concentrate diets affects milk yield, milk composi-
Eknæs, M., Ø. Havrevoll, H. Volden, and K. Hove. 2009. Fat content, tion, and mammary gene expression profile in goats. J. Dairy Sci.
fatty acid profile and off-flavours in goats milk—Effects of feed 92:5544–5560. https://doi.org/10.3168/jds.2009-2022.
concentrates with different fat sources during the grazing season. Onetti, S. G., and R. R. Grummer. 2004. Response of lactating
Anim. Feed Sci. Technol. 152:112–122. https://doi.org/10.1016/j. cows to three supplemental fat sources as affected by forage in
anifeedsci.2009.04.006. the diet and stage of lactation: A meta-analysis of literature.
FAO. 2010. Fats and fatty acids in human nutrition. Report of an Anim. Feed Sci. Technol. 115:65–82. https://doi.org/10.1016/j.
expert consultation. Food and Agriculture Organization of the anifeedsci.2004.02.009.
United Nations (FAO), Rome, Italy. Palmquist, D. L. 1984. Calcium soaps of fatty acids with varying un-
Folch, J., M. Lees, and G. H. Sloane Stanley. 1957. A simple method saturation as fat supplements for lactation cows. Can. J. Anim.
for the isolation and purification of total lipids from animal tissue. Sci. 64:240–241. https://doi.org/10.4141/cjas84-242.
J. Biol. Chem. 226:497–509. Pisanu, S., G. Marogna, D. Pagnozzi, M. Piccinini, G. Leo, A. Tan-
Granlund, L. L., R. Eltun, E. E. Hohle, L. Nesheim, W. Waalen, and ca, A. M. Roggio, T. Roggio, S. Uzzau, and M. F. Addis. 2013.
M. Åssveen. 2010. Biodiesel fra norske jordbruksvekster, Utred- Characterization of size and composition of milk fat globules from
ning for prosjekt [“Opportunities for Norwegian production of bio- Sarda and Saanen dairy goats. Small Rumin. Res. 109:141–151.
diesel from agricultural crops”] (Bioforsk, Trans.) Vol. 5. www. https://doi.org/10.1016/j.smallrumres.2012.07.024.
bioforsk.no. Rabiee, A. R., K. Breinhild, W. Scott, H. M. Golder, E. Block, and
Gulati, S. K., E. B. Byers, Y. G. Byers, J. R. Ashes, and T. W. Scott. I. J. Lean. 2012. Effect of fat additions to diets of dairy cattle
1997. Effect of feeding different fat supplements on the fatty acid on milk production and components: A meta-analysis and meta-
composition of goat milk. Anim. Feed Sci. Technol. 66:159–164. regression. J. Dairy Sci. 95:3225–3247. https://doi.org/10.3168/
https://doi.org/10.1016/S0377-8401(96)01117-0. jds.2011-4895.
Hayes, K. C., P. Khosla, T. Hajri, and A. Pronczuk. 1997. Saturated Sanz Sampelayo, M. R., Y. Chilliard, P. Schmidely, and J. Boza. 2007.
fatty acids and LDL receptor modulation in humans and monkeys. Influence of type of diet on the fat constituents of goat and sheep
Prostaglandins Leukot. Essent. Fatty Acids 57:411–418. milk. Small Rumin. Res. 68:42–63. https://doi.org/10.1016/j.
Inglingstad, R. A., M. Eknæs, L. Brunborg, T. Mestawet, T. G. De- smallrumres.2006.09.017.
vold, G. E. Vegarud, and S. B. Skeie. 2016. Norwegian goat milk Shingfield, K. J., Y. Chilliard, V. Toivonen, P. Kairenius, and D. I.
composition and cheese quality: The influence of lipid supplement- Givens. 2008. Trans fatty acids and bioactive lipids in ruminant
ed concentrate and lactation stage. Int. Dairy J. 56:13–21. https:// milk. Pages 3–65 in Bioactive Components of Milk. Vol. 606. Z.
doi.org/10.1016/j.idairyj.2015.12.010. Bösze, ed. Springer, New York, NY.
Khorasani, G. R., E. K. Okine, and J. J. Kennelly. 1996. Forage source Skjevdal, T. 1979. Flavour of goat’s milk: A review of studies on the
alters nutrient supply to the intestine without influencing milk sources of its variations. Livest. Prod. Sci. 6:397–405.
yield. J. Dairy Sci. 79:862–872. https://doi.org/10.3168/jds.S0022- Steinshamn, H., R. A. Inglingstad, D. Ekeberg, J. Mølmann, and M.
0302(96)76435-4. Jørgensen. 2014. Effect of forage type and season on Norwegian
Littell, R. C., P. R. Henry, and C. B. Ammerman. 1998. Statisti- dairy goat milk production and quality. Small Rumin. Res. 122:18–
cal analysis of repeated measures data using SAS procedures. J. 30. https://doi.org/10.1016/j.smallrumres.2014.07.013.
Anim. Sci. 76:1216–1231. Sundheim, G., T.-L. Zimmer, and H. Astrup. 1983. Induction of milk
Lopez, C., V. Briard-Bion, O. Menard, F. Rousseau, P. Pradel, and lipolysis by lipoprotein components of bovine blood serum. J.
J.-M. Besle. 2008. Phospholipid, sphingolipid, and fatty acid com- Dairy Sci. 66:400–406.
positions of the milk fat globule membrane are modified by diet. R Core Team. 2016. R: A language and environment for statistical
J. Agric. Food Chem. 56:5226–5236. https://doi.org/10.1021/ computing. R Foundation for Statistical Computing, Vienna, Aus-
jf7036104. tria. www.R-project.org.
Ménard, O., S. Ahmad, F. Rousseau, V. Briard-Bion, F. Gaucheron, Tudisco, R., B. Chiofalo, L. Addi, V. L. Presti, R. Rao, S. Calabro’,
and C. Lopez. 2010. Buffalo vs. cow milk fat globules: Size dis- N. Musco, M. Grossi, M. I. Cutrignelli, V. Mastellone, P.
tribution, zeta-potential, compositions in total fatty acids and Lombardi, and F. Infascelli. 2015. Effect of hydrogenated palm
in polar lipids from the milk fat globule membrane. Food Chem. oil dietary supplementation on milk yield and composition, fatty
120:544–551. https://doi.org/10.1016/j.foodchem.2009.10.053. acids profile and stearoyl-CoA desaturase expression in goat
Michalski, M.-C., V. Briard, and F. Michel. 2001. Optical parameters milk. Small Rumin. Res. 132:72–78. https://doi.org/10.1016/j.
of milk fat globules for laser light scattering measurements. Lait smallrumres.2015.10.006.
81:787–796. Wiking, L., J. Stagsted, L. Bjorck, and J. H. Nielsen. 2004. Milk fat
Mir, Z., L. A. Goonewardene, E. Okine, S. Jaegar, and H. D. Scheer. globule size is affected by fat production in dairy cows. Int. Dairy
1999. Effect of feeding canola oil on constituents, conjugated J. 14:909–913.
linoleic acid (CLA) and long chain fatty acids in goats milk. Wilcove, D. S., X. Giam, D. P. Edwards, B. Fisher, and L. P. Koh.
Small Rumin. Res. 33:137–143. https://doi.org/10.1016/S0921- 2013. Navjot’s nightmare revisited: Logging, agriculture, and
4488(99)00016-4. biodiversity in Southeast Asia. Trends Ecol. Evol. 28:531–540.
Murthy, A. V. R., F. Guyomarc’h, G. Paboeuf, V. Vié, and C. Lopez. https://doi.org/10.1016/j.tree.2013.04.005.
2015. Cholesterol strongly affects the organization of lipid mono- Wolfinger, R. D. 1996. Heterogeneous variance: Covariance structures
layers studied as models of the milk fat globule membrane: Con- for repeated measures. J. Agric. Biol. Environ. Stat. 1:205–230.
densing effect and change in the lipid domain morphology. Bio-

Journal of Dairy Science Vol. 100 No. 9, 2017