Introduction

Camelina (Camelina sativa L.) is an oilseed crop that has been cultivated for centuries in Europe for a variety of purposes (Zubr 2003), but it is a relatively new crop in western Canada that is being cultivated for, inter alia, biofuel production. Camelina seed typically contains 39%–47% oil (Zubr 2003), and its mechanical (expeller) processing to extract oil for industrial use produces camelina expeller meal (CEM) as a byproduct. On average, CEM contains 38% crude protein (CP) and 10%–20% residual oil (Sarramone et al. 2020), thus potentially making it a good protein and energy source that could replace more expensive supplements such as canola meal (CM) in dairy cow diets. Also, camelina oil is a rich source of unsaturated fatty acids (UFA), of which C18:3n3 is the most abundant at 25%–42% of total fatty acid (FA) (Sarramone et al. 2020). In addition to increasing the milk content of desirable omega-3 FA such as C18:3n3, feeding CEM could also increase the milk content of cis-10, trans-12 CLA and cis-9, trans-11 CLA. These CLA are produced from the incomplete ruminal biohydrogenation of UFA such as C18:3n3 and have been identified as having numerous health benefits in humans, such as the prevention of cancer and inflammation (Viladomiu et al. 2016).

A potential problem of feeding CEM is the presence of antinutritional factors (e.g., glucosinolates), which might pose a health risk to animals as they have negative effects on the thyroid gland and cardiovascular system (Tripathi and Mishra 2007); so it is important to determine a safe inclusion level for CEM in dairy diets that could at least maintain production without negatively affecting cow health. In studies conducted in Europe, Hurtaud and Peyraud (2007) observed that feeding camelina seed or meal (up to 10% of dietary DM) tended to decrease feed intake with no adverse effects on milk production, whereas Halmemies-Beauchet-Filleau et al. (2011) reported that feeding up to 20% extruded CEM versus feeding camelina oil tended to reduce feed intake without negatively affecting milk production. CEM is not registered for use in dairy diets in Canada and we are not aware of any published local studies that have investigated its effects on production in dairy cows. Therefore, our objective was to determine the effects of feeding graded levels of CEM (up to 10% of dietary DM) on production and milk FA profiles in dairy cows. Our hypothesis was that CEM could partially replace CM in dairy diets without negatively affecting milk production.

Materials and methods

Eight multiparous Holstein cows (718 ± 64 kg BW; 127 ± 14 DIM at the start of the experiment) housed in individual tie-stalls at the University of Saskatchewan Rayner Dairy Research and Teaching Facility were used in this study. To calculate appropriate sample size, a sample size calculation tool ( https://www.bu.edu/researchsupport/compliance/animal-care/working-with-animals/research/sample-size-calculations-iacuc/) was used. Briefly, data from previous studies conducted in the Rayner facility indicated that the standard deviation for milk yield was ∼3. With P = 0.05 and 90% power, ∼12 cows per treatment were required to detect a 10% difference in milk yield. However, due to inadequate amounts of CEM being available to complete a feeding study with 12 cows per treatment, only 8 cows were used. The experimental design was a replicated 4 × 4 Latin square with experimental periods of 28 days (consisting of 18 days of dietary adaptation and 10 days of data and sample collection). All experimental procedures were approved by the University of Saskatchewan Animal Care Committee (UCACS Protocol No. 20040048) and followed the guidelines of the Canadian Council on Animal Care on the care and use of experimental animals (CCAC 2000). Because CEM is not registered for use in dairy diets in Canada, a research permit for its use as a feed ingredient was obtained from the Canadian Food Inspection Agency prior to the initiation of this study.

The four dietary treatments consisted of a standard barley silage-based diet containing CM as the principal source of protein or diets formulated to contain 5%, 7.5%, and 10% CEM (as percentage of diet DM) as a partial replacement for CM. The chemical compositions of CEM and CM, and the ingredient and chemical compositions of the four experimental diets are presented in Table 1. The maximum inclusion level of 10% was used as it is the limit set by the US Food and Drug Administration, and it is the maximum safe inclusion level for cattle (Lawrence et al. 2016). The batch of CEM that was used was supplied by Linnaeus Plant Sciences Inc. (Vancouver, BC). Briefly, the manufacturing process involved the camelina seed being mechanically sieved to clean it, followed by oil extraction by pressing the seed through a single-screw press where the temperatures reached were 25–40 °C. The remaining CEM was allowed to air-cool to room temperature and packaged into tote bags until use. Cows were fed experimental diets as total mixed rations (TMR) at 0900 and 1600 h for ad libitum intake and had free access to water.

Table 1.

Chemical composition of CEM and CM, and ingredient and chemical composition of TMR containing graded levels of CEM fed to dairy cows.

During the last 10 days of each experimental period, individual cow feed intake was recorded daily. Samples of TMR were collected on days 25, 26, and 27 and stored at −20 °C. Experimental cows were milked at 0430, 1230, and 1900 h and milk weights were recorded. At each milking on days 25, 26, and 27, milk samples were collected in plastic vials containing 2-bromo-2-nitropropane-1-3-propanediol as a preservative and were stored at 4 °C before being sent to the CanWest DHI Milk Testing Laboratory (Edmonton, AB) for compositional analysis. Milk samples were also collected on days 25, 26, and 27 in plastic vials without preservative and were pooled proportionally based on milk yield into a composite sample that was stored at −20 °C.

After the experiment, frozen TMR samples were thawed at room temperature, pooled per cow per collection period and subsequently dried in an oven at 60 °C for 48 h. Dried TMR samples were ground through a 1 mm screen using a Christy–Norris mill (Christy and Norris Ltd., Chelmsford). Dried TMR samples were sent to Cumberland Valley Analytical Services Inc. (Maugansville, MD) to be analyzed for ash, CP, crude fat, ADF, NDF, starch, and minerals. Pooled milk samples that were sent to the CanWest Laboratory were analyzed for fat, CP, lactose, and milk-urea nitrogen (MUN) using mid-infrared spectroscopy (Foss MilkoScan 7, Foss Analytical, Hillerød, Denmark). Pooled milk samples without preservatives were sent to the Lipid Analytical Services Ltd. (Guelph, ON) for FA analysis as described by AlZahal et al. (2009), with identification of FA peaks by comparison of retention times with known FA methyl-ester standards (Nu-Chek Prep Inc., Elysian, MN).

All data on feed intake, milk production, and milk composition were analyzed as a replicated 4 × 4 Latin square using the MIXED procedure of SAS according to the following model: Y_ijkl = µ + S_i + P_j + C_k(i) + T_l + ST_il + E_ijkl, where Y_ijkl is the dependent variable, µ is the overall mean, S_i is the random effect of square i, P_j is the random effect of period j, C_k(i) is the random effect of cow k (within square i), T_l is the fixed effect of dietary treatment l, ST_il is the interaction between square i and treatment l, and E_ijkl l is the residual error. Orthogonal polynomial contrasts were used to test for linear, quadratic, and cubic effects of the dietary level of CEM. Significance was declared at P ≤ 0.05 and trends at 0.05 < P < 0.10.

Results and discussion

Feeding graded levels of CEM did not influence dry matter intake (DMI) (P > 0.05; Table 2). In agreement with our observations, feeding 10% CEM to growing dairy heifers had no negative effects on DMI (Lawrence et al. 2016). However, Hurtaud and Peyraud (2007) observed a tendency (P = 0.089) for a reduced DMI when ∼10% CEM was fed to dairy cows as a substitute for soybean meal. Based on an extensive review of the literature, Tripathi and Mishra (2007) concluded that dietary supplementation with glucosinolate-rich feedstuffs like CEM will generally result in decreased DMI partly due to reduced diet palatability as a result of the bitter taste of the degradation products of glucosinolates. In the current study, calculated glucosinolate intakes for cows fed CEM based on mean DMI, and dietary inclusion levels and glucosinolate content of CEM (21.4 µmol/g) were 1.08–2.14 µmol/g of diet. These glucosinolate intakes were well below the threshold levels of 11.7–24.3 µmol/g that can depress DMI (Tripathi and Mishra 2007); so it was not surprising that DMI was unaffected in the present study.

Table 2.

Effects of feeding graded levels of CEM in experimental diets on production and milk FA profiles in dairy cows in Experiment 1¹.

Feeding graded levels of CEM did not influence milk yield (P > 0.05; Table 2). It is well established that milk yield is positively correlated with DMI; so the lack of response in milk yield with dietary addition of CEM was not surprising as DMI was unaffected by diet. However, it should be noted that the current study was conducted with a rather small sample size (n = 8), and this could have limited our ability to detect statistical differences for production parameters. Because CEM is a relatively new feedstuff, studies that have compared feeding graded levels of CEM with corn silage- or barley silage-based diets on milk yield in dairy cows are scarce. Although DMI tended to decrease with 10% supplemental CEM (41.1% CP and 13.2% crude fat) in a study by Hurtaud and Peyraud (2007), this was not accompanied by a decrease in milk yield likely because the decrease in DMI was quite small (∼1.1 kg/day). Recently, Lawrence et al. (2016) evaluated the effects of supplemental CEM as a replacement for distiller grains with solubles on growth performance of dairy heifers and reported that average daily gain (ADG) and bodyweight (BW) were not influenced by CEM, which was congruent with the lack of effects of supplemental CEM on DMI.

Milk fat and CP contents decreased linearly, whereas MUN concentration decreased quadratically, as dietary content of CEM increased. When feeding concentrate-rich diets that are high in polyunsaturated fatty acids (PUFA), ruminal biohydrogenation of PUFA results in the formation of intermediates such as C18:1 trans-10 and trans-10, cis-12 CLA, which are potent inhibitors of de novo milk fat synthesis when they reach the mammary gland (Bauman et al. 2011). It is likely that these intermediates were responsible for the observed decrease in milk fat content as dietary CEM increased in the present study as supported by the observed increase in milk content of trans-10, cis-12 CLA. Also, we observed linear decreases in milk fat contents of most of the short- to medium-chain saturated FA (i.e., C4:0–C16:0) that are predominantly derived from de novo synthesis in the mammary gland, thus suggesting that de novo fat synthesis was suppressed. Hurtaud and Peyraud (2007) also reported that feeding CEM decreased milk fat content. The decrease in MUN concentration is consistent with results from a previous study when CEM was fed to dairy cows (Halmemies-Beauchet-Filleau et al. 2011). Camelina meal has a greater rumen undegradable protein (RUP) content compared to CM (316 versus 275 g/kg CP; Colombini et al. 2014), and providing camelina meal as an N source in a dual-flow continuous culture system resulted in a lower NH₃-N concentration compared to CM (Brandao et al. 2018). Taken together, results from these two studies provide evidence that camelina meal has a lower rumen degradable protein (RDP) content than CM. As MUN is derived from hepatic ureagenesis using ruminally derived NH₃-N as the primary source of N, a lower ruminal NH₃-N concentration from the lower RDP content when CEM was fed would be expected to decrease MUN concentration as observed in the present study. The observed response in milk CP content with greater levels of CEM can be attributed to the decrease in MUN concentration.

Feeding increasing amounts of CEM resulted in linear decreases (P < 0.01) in milk content of most of the short- to medium-chain FA (C6:0–C16:0). Of the major long-chain FA, feeding increasing amounts of CEM resulted in linear increases (P < 0.01) in C18:1, C18:2n6, C18:2 cis-9, trans-11, C18:2 trans-10, cis-12, total CLA, and C18:3n3, and a linear decrease in C20:4n6 (P < 0.01) (Table 2). Also, milk content of total saturated fatty acids (SFA) linearly decreased (P < 0.01), whereas that of total monounsaturated fatty acids (MUFA), PUFA, and omega-3 FA linearly increased (P < 0.01) as dietary content of CEM increased. The major PUFA that are contained in camelina oil are C18:2n6 and C18:3n3 (Sarramone et al. 2020). Although these PUFA can undergo extensive biohydrogenation in the rumen, which would potentially reduce their transfer into milk, substantial amounts of these FA can escape ruminal biohydrogenation when PUFA-rich diets are fed and can be available for incorporation into milk, thus elevating their content in milk as observed in the present study. Contrary to our results, other studies have reported that feeding CEM (Hurtaud and Peyraud 2007) or camelina oil (Halmemies-Beauchet-Filleau et al. 2011) had no influence on the secretion of C18:2n6 and C18:3n3 in milk.

We observed linear increases in milk fat contents of cis-9, trans-11 CLA and trans-10, cis-12 CLA, which are important intermediates in the incomplete ruminal biohydrogenation of dietary C18:2n6 and C18:3n3 (Halmemies-Beauchet-Filleau et al. 2011). These results suggest that the greater dietary content of PUFA-rich CEM induced a shift in ruminal biohydrogenation pathways, leading to the ruminal accumulation and, subsequently, greater intestinal absorption of these intermediates. These responses in milk enrichment with cis-9, trans-11 CLA, trans-10, cis-12 CLA, and total CLA when CEM was fed are in agreement with other studies (Hurtaud and Peyraud 2007; Halmemies-Beauchet-Filleau et al. 2011). Biologically active CLA isomers such as cis-9, trans-11 CLA and trans-10, cis-12 CLA have positive health effects in humans, such as being anti-carcinogenic and anti-inflammatory (Viladomiu et al. 2016); so the enrichment of milk with these FA is desirable. Typically, milk fat content of total CLA ranges from 0.3 to 0.4 g/100 g FA (Lindmark Månsson 2008), a range which encompasses the total CLA content that was observed in cows fed the control diet in the present study. Overall, milk fat content of total CLA increased from 0.21 to 0.50 g/100 g FA when diets contained 5%–10% CEM when compared to the control diet, which is a relatively substantial increase that could potentially improve the CLA status of human consumers. A possible shortcoming of this study was that the FA compositions of CEM and CM were not measured, thus potentially limiting our ability to attribute the observed responses in milk FA profiles to the dietary substitution of CM with CEM. Of the major UFA that would result in alterations in milk FA composition due to the production of biohydrogenation intermediates in the rumen, a previous batch of CEM obtained from the same crushing plant as CEM used in the present study contained (percent of total FA) 22.4% C18:2n6 and 27.3% C18:3n3 (unpublished data), whereas that of CM contained 20.1% C18:2n6 and 9.3% C18:3n3 (Canola Council of Canada 2019); thus, CEM is a richer source of these UFA in addition to having a four-fold greater crude fat content (14.8% versus 3.5%). Taken together, we can surmise that the greater UFA and crude fat contents of CEM would result in CEM having a far greater influence on milk FA composition when graded levels are fed as a replacement for CM.

In conclusion, the inclusion of CEM up to 10% in barley silage-based diets as a partial replacement for CM had no detrimental effects on DMI and milk yield; however, milk fat and CP contents decreased. Feeding CEM resulted in important alterations in milk FA composition, which were mainly characterized by increases in the content of desirable omega-3 FA (such as C18:3n3) and CLA isomers (particularly cis-9, trans-11 CLA) that have been demonstrated to have human health benefits. This is an important finding as it shows that feeding CEM could be a viable strategy to favorably manipulate the FA composition of bovine milk, thus making it more appealing to today's health-conscious consumers.