A Meta-Analytic Approach to Predict Methane Emissions from Dairy Goats Using Milk Fatty Acid Profile

sustainability Article A Meta-Analytic Approach to Predict Methane Emissions from Dairy Goats Using Milk Fatty Acid Profile Francisco Requena 1, Francisco Peña 2, Estrella Agüera 1and Andrés Martínez Marín2,* 1Department of Cell Biology, Physiology and Immunology, University of Córdoba, Ctra. Madrid-Cádiz, km. 396, 14071 Córdoba, Spain; v02r[email protected] (F.R.); [email protected] (E.A.) 2Department of Animal Production, University of Córdoba, Ctra. Madrid-Cádiz, km. 396, 14071 Córdoba, Spain; [email protected] *Correspondence: [email protected] Received: 20 April 2020; Accepted: 10 June 2020; Published: 13 June 2020   Abstract: The aim of this work was to develop an equation to predict methane yield (CH 4 , g/kg dry matter intake) from dairy goats using milk fatty acid (FA) profile. Data from 12 research papers (30 treatments and 223 individual observations) were used in a meta-regression. Since most of the selected studies did not extensively report milk fat composition, palmitic acid (C16:0) was selected as a potential predictor. The obtained equation was: CH4(g/kg dry matter intake) =0.525 ×C16:0 (% in milk fat). The coefficient of determination (R 2 =0.46), the root mean square error of prediction (RM SPE =3.16 g/kg dry matter intake), and the concordance correlation coefficient ( CCC =0.65 ) indicated that the precision, accuracy and reproducibility of the model were moderate. The relationship between CH 4 yield and C16:0 content in milk fat would be supported by the fact that diet characteristics that increase the amount of available hydrogen in the rumen for archaea to produce CH4, simultaneously favor the conditions for the synthesis of C16:0 in the mammary gland. The obtained equation might be useful, along with previous published equations based on diet characteristics, to evaluate the environmental impact of dairy goat farming. Keywords: methane; goats; milk; fatty acids 1. Introduction Goat farming is considered an important source of income and nutrition for poor and marginal farmers around the world. The resilience of this species to harsh conditions is particularly remarkable, but their enteric methane (CH 4 ) emission along with productive performance should be taken into account in the changing climate scenario [ 1 ]. The livestock sector is responsible for 14.5% of the global anthropogenic greenhouse gas emissions (GHG), 44% in the form of CH 4 and the remaining part almost equally shared between nitrous oxide (N2O) and carbon dioxide (CO2). Small ruminants (s heep and goats ) contribute 6.5% of total livestock sector emissions of GHG, with over 55% in the form of CH4from enteric fermentation [2]. Several equations to predict enteric CH 4 emission from dairy cows using animal and diet characteristics as well as milk fatty acid (FA) profiles have been developed through meta-analyses of published research [ 3 – 6 ]. The most common predictors in those equations are body weight, milk yield and composition, dry matter intake (DMI), diet composition, and the contents of iso FA and rumen biohydrogenation-derived FA in milk fat [ 3 – 6 ]. The basis of the proposed equations is the direct relationship between diet and rumen microbiota and the intrinsic connection between the lipid metabolism of ruminal bacteria and the milk FA profile [ 7 ]. Some equations for predicting CH 4 emission based on animal and diet characteristics have been also proposed for goats at any Sustainability 2020,12, 4834; doi:10.3390/su12124834 www.mdpi.com/journal/sustainability Sustainability 2020,12, 4834 2 of 9 productive stage [ 8 ]. However, to our knowledge no predictive equations based specifically on milk FA profile are available for dairy goats. Direct comparisons of milk FA responses to the same diets show dissimilarities between dairy cows and goats [ 9 , 10 ], which supports the assumption that the equations obtained in dairy cows cannot be used as proxy to predict CH 4 emission from dairy goats. A meta-analysis of published research data obtained in dairy goat studies could help to quantitatively relate CH 4 emission to milk FA profile in these animals, as an alternative to the scarcity of knowledge on the matter. Therefore, the objective of this study was to perform a meta-analysis and a meta-regression to quantify the relationship between CH4and milk FA profile in dairy goats. 2. Materials and Methods Research papers on dairy goats were searched in Google Scholar using the operator “allintext:” and introducing the keywords: “methane”, “milk fatty acids” and “goats”. The search was limited to the year 2000 onwards. Studies were selected if they simultaneously reported milk FA profile (g/100 g FA) and CH 4 yield (g/kg DMI) or CH 4 production (g/d or L/d) along with daily DMI. Any treatment where CH 4 emission was manipulated by administering chemical additives to the animals was excluded. Twelve peer reviewed research papers that involved a total of 30 treatments with 223 individual observations fulfilled the requirements (Table 1) [ 11 – 22 ]. The experimental diets were very varied and included hays, silages, agro-industrial by-products, cereals, protein concentrates and fat sources. In the selected studies, CH 4 emission was measured using the sulfur hexafluoride technique ( n=1 ), open-circuit respiration chamber (n=6) and open-circuit mask system (n=5). All the studies determined the FA contents in milk fat by gas chromatography. Table 1. Details of the 12 experiments included in the meta-analysis. Author n1Method 2Breed 3DIM 4Forage Type [11] 9 RC M-G 57 Alfalfa hay [12]6 RC M-G 60 Alfalfa +oat hay Alfalfa hay +olive by-product silage Alfalfa hay +tomato surplus silage [13] 10 OCM M-G LL Alfalfa hay [14]10 OCM M-G LL Alfalfa hay Maralfalfa hay [15]10 OCM M-G LL Alfalfa pellets Lemon leaves pellets [16]10 OCM M-G ML Alfalfa pellets Orange leaves pellets [17] 6 SF6AxB n.s. Berseem hay [18] 8 OCM M-G ML Alfalfa hay [19] 8 RC M-G 13 Alfalfa hay [20] 3 RC S 106 Italian ryegrass silage [21]3 RC S 106 Native pasture Pasture hay Non-forage diet [22] 8 RC M-G ML Alfalfa hay 1 Number of animals per treatment. 2 Method of CH 4 emission measurement (RC: Respiratory chamber; OCM: Open-circuit mask; SF 6 : Sulphur hexafluoride). 3 M-G: Murciano-Granadina; AxB: Alpine x Beetal; S: Saanen. 4Days in milk (LL: Late lactation; ML: Mid lactation; n.s.: Not specified). The data on diet composition, experimental design, number of animals per treatment, body weight, DMI, milk yield, milk fat, milk fat FA profile, molar proportions of volatile fatty acids (VFA) in rumen fluid and CH 4 emission reported in the studies were recorded and stored in a Microsoft Excel spreadsheet. The factors to convert units were 0.71 g/L CH 4 and 55.65 kJ/g CH 4 [ 8 ]. Since most of the selected studies did not extensively report milk fat composition, the milk fat contents of several FA that are highly related to CH 4 emission were not available [ 3 , 6 , 23 ]. Thus, we decided to select palmitic Sustainability 2020,12, 4834 3 of 9 acid (C16:0) as a potential predictor because of its positive relationship with CH 4 emission reported in the literature [ 3 , 23 ] and its negative relationship with dietary factors that are known to alter the rumen environment, like feeding high concentrate diets, elevated consumption of unsaturated FA or both [9,10,24,25]. All the statistical analyses were performed with the MIXED procedure of SAS University Edition 3.8 (SAS Institute, Cary, NC, USA). A regression analysis (meta-regression) was conducted to study the prediction of CH 4 yield (g/kg DMI) using C16:0 content (%) in milk fat [ 26 ]. Visual inspection of the data showed that the relationship between the variables was linear (Figure 1), thus a linear regression was performed. The statistical model included the fixed effect of C16:0 content in milk fat and the random effects of study and study × C16:0 content interaction. The data were weighed with the square root of the number of animals used in each treatment using the WEIGHT statement. Statistical significance was declared at p<0.05. Sustainability 2020, 12, x FOR PEER REVIEW 3 of 9 FA that are highly related to CH4 emission were not available [3,6,23]. Thus, we decided to select palmitic acid (C16:0) as a potential predictor because of its positive relationship with CH4 emission reported in the literature [3,23] and its negative relationship with dietary factors that are known to alter the rumen environment, like feeding high concentrate diets, elevated consumption of unsaturated FA or both [9,10,24,25]. All the statistical analyses were performed with the MIXED procedure of SAS University Edition 3.8 (SAS Institute, Cary, NC, USA). A regression analysis (meta-regression) was conducted to study the prediction of CH4 yield (g/kg DMI) using C16:0 content (%) in milk fat [26]. Visual inspection of the data showed that the relationship between the variables was linear (Figure 1), thus a linear regression was performed. The statistical model included the fixed effect of C16:0 content in milk fat and the random effects of study and study × C16:0 content interaction. The data were weighed with the square root of the number of animals used in each treatment using the WEIGHT statement. Statistical significance was declared at p < 0.05. Figure 1. Plot of observed methane (CH4) yield vs. palmitic acid (C16:0) content in milk fat. 3. Results The descriptive statistics of the data compiled from the studies included in the meta-analysis are shown in Table 2. Despite the uniformity in the breeds used in the studies that were included in the dataset (Table 1), body weight, dry matter intake, milk yield and composition, and CH4 emission showed considerable variation. The ranges of intakes of diet components varied widely, which reflected the substantial differences between diets in the studies. The correlation between CH4 yield and proportion of C16:0 in milk fat showed that both were positively linearly related not only in the pooled data (Figure 1) but also within nine out of the 11 experiments with two or more treatments. Table 2. Descriptive statistics of the data compiled from the studies included in the meta-analysis. Mean Minimum Maximum Standard Deviation Body weight (kg) 44.9 34.0 55.0 6.07 Dry matter intake (kg/d) 1.75 0.99 2.69 0.443 CP 1 intake (kg/d) 0.30 0.18 0.47 0.076 NDF 2 intake (kg/d) 0.57 0.32 1.15 0.205 EE 3 intake (kg/d) 0.06 0.01 0.19 0.041 NFC 4 intake (kg/d) 0.61 0.20 1.01 0.262 Milk yield (kg/d) 1.92 0.94 3.69 0.831 Milk fat (%) 4.86 2.96 6.90 1.103 Palmitic acid (% in milk fat) 29.8 20.5 44.5 6.20 Methane (g/d) 28.1 12.3 68.8 14.33 Methane (g/kg dry matter intake) 15.7 9.0 26.6 4.36 1 Crude protein. 2 Neutral detergent fiber. 3 Ether extract. 4 Non fibrous carbohydrates. Figure 1. Plot of observed methane (CH4) yield vs. palmitic acid (C16:0) content in milk fat. 3. Results The descriptive statistics of the data compiled from the studies included in the meta-analysis are shown in Table 2. Despite the uniformity in the breeds used in the studies that were included in the dataset (Table 1), body weight, dry matter intake, milk yield and composition, and CH 4 emission showed considerable variation. The ranges of intakes of diet components varied widely, which reflected the substantial differences between diets in the studies. The correlation between CH 4 yield and proportion of C16:0 in milk fat showed that both were positively linearly related not only in the pooled data (Figure 1) but also within nine out of the 11 experiments with two or more treatments. Table 2. Descriptive statistics of the data compiled from the studies included in the meta-analysis. Mean Minimum Maximum Standard Deviation Body weight (kg) 44.9 34.0 55.0 6.07 Dry matter intake (kg/d) 1.75 0.99 2.69 0.443 CP 1intake (kg/d) 0.30 0.18 0.47 0.076 NDF 2intake (kg/d) 0.57 0.32 1.15 0.205 EE 3intake (kg/d) 0.06 0.01 0.19 0.041 NFC 4intake (kg/d) 0.61 0.20 1.01 0.262 Milk yield (kg/d) 1.92 0.94 3.69 0.831 Milk fat (%) 4.86 2.96 6.90 1.103 Palmitic acid (% in milk fat) 29.8 20.5 44.5 6.20 Methane (g/d) 28.1 12.3 68.8 14.33 Methane (g/kg dry matter intake) 15.7 9.0 26.6 4.36 1Crude protein. 2Neutral detergent fiber. 3Ether extract. 4Non fibrous carbohydrates. Sustainability 2020,12, 4834 4 of 9 The intercept of the obtained regression equation was not significant (p=0.82) and was removed from the model. The slope of the final no-intercept regression Equation (1) was highly significant ( p<0.001 ) and the coefficient of determination (R 2 ) and the root mean square error of prediction (RMSEP) indicated that the goodness of fit and the accuracy of the model were moderate. Neither the mean bias (1.49 g/kg DMI; p=0.62) nor the linear bias ( − 0.089 g/kg DMI; p=0.64) were significant, thus almost all the model prediction error was due to random variation (Figure 2). Mor eover, the bia s correction factor, which reflects accuracy, was 0.96 and the concordance correlation coefficient or reproducibility index was 0.65. CH4(g/kg DMI) =0.525 (±0.0252) ×C16:0 (% in milk fat) R2=0.46; RMSEP =3.16 g/kg DMI (20.1%) (1) Sustainability 2020, 12, x FOR PEER REVIEW 4 of 9 The intercept of the obtained regression equation was not significant (p = 0.82) and was removed from the model. The slope of the final no-intercept regression Equation (1) was highly significant (p < 0.001) and the coefficient of determination (R2) and the root mean square error of prediction (RMSEP) indicated that the goodness of fit and the accuracy of the model were moderate. Neither the mean bias (1.49 g/kg DMI; p = 0.62) nor the linear bias (−0.089 g/kg DMI; p = 0.64) were significant, thus almost all the model prediction error was due to random variation (Figure 2). Moreover, the bias correction factor, which reflects accuracy, was 0.96 and the concordance correlation coefficient or reproducibility index was 0.65. Figure 2. Plot of observed minus predicted vs. predicted methane yield (CH4, g/kg dry matter intake). Those results aimed to further evaluate model adequacy. However, the number of studies in the dataset was not large enough to split them into development and evaluation subsets. Therefore, as a substitute for assessing the model performance, we decided to compare CH4 yield predicted using Equation (1) or using the relationship between CH4 yield and the molar proportions of VFA (acetate, propionate and butyrate) in rumen fluid as described in recent research [27]. The results from nine studies (21 treatments and 187 individual observations) in the dataset that reported molar proportions of VFA in rumen fluid were used for the comparison [11–16,18,19,22]. The descriptive statistics of the variables used in the calculations and of the predicted CH4 yields according to both methods are shown in Table 3. Table 3. Descriptive statistics of the data in the nine studies that were used to compare methane (CH4) yield predicted using either palmitic acid content in milk fat (CH4C16:0) or molar proportions of volatile fatty acids (VFA) in rumen fluid (CH4VFA). Mean Minimum Maximum Standard Deviation Palmitic acid (% in milk fat) 29.4 20.5 44.5 6.77 Acetate (mol/100 mol VFA) 62.1 53.5 67.6 4.23 Propionate (mol/100 mol VFA) 15.4 11.1 26.7 3.73 Butyrate (mol/100 mol VFA) 16.8 10.6 22.4 2.72 CH4 (g/kg dry matter intake) 14.8 9.0 21.1 3.37 CH4C16:0 (g/kg dry matter intake) 15.4 10.8 23.4 3.55 CH4VFA (g/kg dry matter intake) 11.6 6.7 14.0 2.07 As shown in Table 4, all the measures of precision (Pearson r), accuracy (root mean square error of prediction, RMSPE; and bias correction factor, BCF) and reproducibility (concordance correlation coefficient, CCC) of CH4 yield prediction were better when applying Equation (1) than when using the molar proportions of VFA as predictors. Moreover, the latter method underpredicted mean CH4 yield as ~22%, whereas Equation (1) overpredicted of mean CH4 yield as ~4%. It is noticeable that CH4 (g/kg DMI) = 0.525 (±0.0252) × C16:0 (% in milk fat) R2 = 0.46; RMSEP = 3.16 g/kg DMI (20.1%) (1) Figure 2. Plot of observed minus predicted vs. predicted methane yield (CH 4 , g/kg dry matter intake). Those results aimed to further evaluate model adequacy. However, the number of studies in the dataset was not large enough to split them into development and evaluation subsets. Therefore, as a substitute for assessing the model performance, we decided to compare CH 4 yield predicted using Equation (1) or using the relationship between CH 4 yield and the molar proportions of VFA (acetate, propionate and butyrate) in rumen fluid as described in recent research [ 27 ]. The results from nine studies (21 treatments and 187 individual observations) in the dataset that reported molar proportions of VFA in rumen fluid were used for the comparison [ 11 – 16 , 18 , 19 , 22 ]. The descriptive statistics of the variables used in the calculations and of the predicted CH 4 yields according to both methods are shown in Table 3. Table 3. Descriptive statistics of the data in the nine studies that were used to compare methane (CH 4 ) yield predicted using either palmitic acid content in milk fat (CH 4C16:0 ) or molar proportions of volatile fatty acids (VFA) in rumen fluid (CH4VFA). Mean Minimum Maximum Standard Deviation Palmitic acid (% in milk fat) 29.4 20.5 44.5 6.77 Acetate (mol/100 mol VFA) 62.1 53.5 67.6 4.23 Propionate (mol/100 mol VFA) 15.4 11.1 26.7 3.73 Butyrate (mol/100 mol VFA) 16.8 10.6 22.4 2.72 CH4(g/kg dry matter intake) 14.8 9.0 21.1 3.37 CH4C16:0 (g/kg dry matter intake) 15.4 10.8 23.4 3.55 CH4VFA (g/kg dry matter intake) 11.6 6.7 14.0 2.07 As shown in Table 4, all the measures of precision (Pearson r), accuracy (root mean square error of prediction, RMSPE; and bias correction factor, BCF) and reproducibility (concordance correlation coefficient, CCC) of CH 4 yield prediction were better when applying Equation (1) than when using the Sustainability 2020,12, 4834 5 of 9 molar proportions of VFA as predictors. Moreover, the latter method underpredicted mean CH 4 yield as ~22%, whereas Equation (1) overpredicted of mean CH 4 yield as ~4%. It is noticeable that none of the models showed mean or linear bias, which indicated that almost all the prediction error was due to random variation in both cases. Table 4. Prediction analysis of methane (CH 4 ) yield in nine studies using either palmitic acid content in milk fat (CH 4C16:0 ) or molar proportions of volatile fatty acids in rumen fluid (CH 4VFA ) as independent variables. RMSEP 1 g/kg DMI RMSEP % Mean Bias Linear Bias BCF 2Pearson rCCC 3 CH4C16:0 2.36 16 3.49 p=0.13 −0.27 p=0.07 0.98 0.77 0.76 CH4VFA 4.41 30 6.62 p=0.11 −0.29 p=0.39 0.54 0.43 0.23 1Root mean square error of prediction. 2Bias correction factor. 3Concordance correlation coefficient. 4. Discussion The number of studies, treatments and individual observations (12, 30 and 223, respectively) was lower and the ranges of body weight, DMI and CH 4 yield in our dataset were narrower than in previous research, where 42 studies that involved 211 treatments and 978 individual observations were used and the ranges of body weight, DMI and CH 4 yield in the dataset were 14.5 to 58.7 kg, 0.34 to 2.35 kg/d and 6.5 to 39.0 g/kg DMI, respectively [ 8 ]. Those differences are due to the fact that our study was limited exclusively to lactating dairy goats and only papers that reported simultaneously CH 4 emission and C16:0 content in milk fat were included. The ruminal contents of VFA have been related to CH 4 emission using stoichiometric equations because the proportions of acetate, propionate and butyrate determine the amount of hydrogen available for utilization by archaea [ 27 , 28 ]. However, our results showed that CH 4 yield predicted using Equation (1) would be more precise, accurate and reproducible than predictions using the molar proportions of VFA in rumen fluid (Table 4), at least under conditions similar to those used in the experiments included in the dataset (Table 1). The role of C16:0 as a predictor of enteric CH 4 emission would be supported by current scientific knowledge. First, acetate production from carbohydrate fermentation by rumen cellulolytic bacteria is inextricably linked to methanogenesis because it is the main source of hydrogen to be combined with carbon dioxide by archaea to release CH 4 [ 29 , 30 ]. Second, acetate and to a lesser extent β -hydroxybutyrate are the main substrates for de novo synthesis of C16:0 in the mammary gland [ 31 ]. Third, approximately 50% of C16:0 incorporated into milk fat comes from de novo synthesis [ 31 , 32 ]. Fourth, rumen cellulolytic bacteria are very sensitive to dietary changes, especially those which imply low ruminal pH, high intake of polyunsaturated FA (PUFA) or both [ 33 – 35 ]. Finally, in connection with the latter, the increased availability of long-chain unsaturated FA (either absorbed intact from the diet or derived from incomplete rumen biohydrogenation of dietary PUFA) to the mammary gland alters the activity ratio of the key enzymes acetyl-CoA carboxylase and fatty acid synthetase involved in de novo synthesis of FA [ 24 ]. Therefore, any diet modifications that negatively affect rumen cellulolytic bacteria would result in lowered availability of both hydrogen for archaea to produce CH 4 and acetate for the mammary gland to synthesize de novo C16:0, and, in most cases, would also impair the activities of the enzymes involved in the latter. As a result, CH 4 emission and C16:0 content in milk fat would decrease simultaneously [3,23]. Despite our results, other milk FA are expected to be better predictors of CH 4 emission in dairy ruminants than C16:0 [ 3 , 6 , 23 ]. Odd and branched-chain FA from microbial origin and trans FA arising from incomplete ruminal biohydrogenation of dietary unsaturated FA have aroused much attention in the last decade because they are directly related to rumen bacteria and their responses to diet characteristics [ 36 , 37 ]. Cellulolytic bacteria are rich in iso FA [ 38 ] and are the main responsible of Sustainability 2020,12, 4834 6 of 9 ruminal biohydrogenation of dietary unsaturated FA [ 39 ]. A decline in the abundance of cellulolytic bacteria promotes a shift in rumen biohydrogenation pathways with the increased formation of trans-10, trans-12 C18:2 and other biohydrogenation intermediate FA that have been associated with milk fat depression [ 40 ]. Unfavorable ruminal conditions to cellulolytic bacteria not only will diminish CH 4 emission due to the shortage of hydrogen for archaea but also will decrease the contents of iso FA and rise the contents of trans FA in milk fat due to a lower bacterial growth and a more incomplete rumen biohydrogenation process. These effects are well reflected in the predictive equations developed for predicting CH 4 emission from dairy cows (e.g., iso C16:0 adds up and trans-10 C18:1 and trans-11 C18:1 subtract in Van Lingen’s equation [ 6 ]). Principal component analyses of FA contents in caprine milk fat would suggest that, at least, iso C14:0, iso C16:0 and trans-11 C18:1 could be potential predictors of CH4emission from dairy goats [41,42]. Applying Equation (1) to published results obtained in dairy goats, it can be predicted that the consumption of 30 g/d of linseed oil would decrease CH 4 yield by ~14% and the reduction would rise up to ~40% at an intake of 125 g/d [ 41 , 42 ]. Those predicted changes are in line with the findings in dairy cows that were fed increasing levels of linseed fat [ 43 ]. Furthermore, from the original data of published research [ 44 ], Equation (1) predicted that CH 4 yield would diminish ~18% after the first 72 h from the inclusion of 48 g/d of linseed oil in the diet of dairy goats. 5. Conclusions As more research papers that simultaneously report enteric CH 4 emission and extensive milk FA profile in dairy goats become available to develop better predictive equations through meta-regression, a simple equation based on C16:0 content in milk fat might be useful, along with previous published equations based on diet characteristics, to evaluate the environmental impact of dairy goat farming. Author Contributions: Conceptualization, A.M.M., F.R. and F.P.; methodology, A.M.M. and F.R.; formal analysis, A.M.M. and F.R.; investigation, A.M.M.; data curation, A.M.M., F.P. and E.A.; writing—original draft preparation, A.M.M.; writing—review and editing, A.M.M., F.R., F.P. and E.A. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Conflicts of Interest: The authors declare no conflict of interest. References 1. Pragna, P.; Chauhan, S.S.; Sejian, V.; Leury, B.J.; Dunshea, F.R. Climate change and goat production: Enteric methane emission and its mitigation. Animals 2018,8, 235. [CrossRef] [PubMed] 2. Gerber, P.J.; Steinfeld, H.; Henderson, B.; Mottet, A.; Opio, C.; Dijkman, J.; Falcucci, A.; Tempio, G. Tackling Climate Change through Livestock—A Global Assessment of Emissions and Mitigation Opportunities; Food and Agriculture Organization: Rome, Italy, 2013. 3. Bougouin, A.; Appuhamy, J.A.D.R.N.; Ferlay, A.; Kebreab, E.; Martin, C.; Moate, P.J.; Benchaar, C.; Lund, P.; Eug è ne, M. Individual milk fatty acids are potential predictors of enteric methane emissions from dairy cows fed a wide range of diets: Approach by meta-analysis. J. Dairy Sci. 2019 ,102, 10616–10631. [CrossRef] [PubMed] 4. Ramin, M.; Huhtanen, P. Development of equations for predicting methane emissions from ruminants. J. Dairy Sci. 2013,96, 2476–2493. [CrossRef] [PubMed] 5. Santiago-Juarez, B.; Moraes, L.E.; Appuhamy, J.A.D.R.N.; Pellikaan, W.F.; Casper, D.P.; Tricarico, J.; Kebreab, E. Prediction and evaluation of enteric methane emissions from lactating dairy cows using different levels of covariate information. Anim. Prod. Sci. 2016,56, 557. [CrossRef] 6. Van Lingen, H.J.; Crompton, L.A.; Hendriks, W.H.; Reynolds, C.K.; Dijkstra, J. Meta-analysis of relationships between enteric methane yield and milk fatty acid profile in dairy cattle. J. Dairy Sci. 2014 ,97, 7115–7132. [CrossRef] [PubMed] Sustainability 2020,12, 4834 7 of 9 7. Negussie, E.; de Haas, Y.; Dehareng, F.; Dewhurst, R.J.; Dijkstra, J.; Gengler, N.; Morgavi, D.P.; Soyeurt, H.; van Gastelen, S.; Yan, T.; et al. Invited review: Large-scale indirect measurements for enteric methane emissions in dairy cattle: A review of proxies and their potential for use in management and breeding decisions. J. Dairy Sci. 2017,100, 2433–2453. [CrossRef] 8. Patra, A.K.; Lalhriatpuii, M. Development of statistical models for prediction of enteric methane emission from goats using nutrient composition and intake variables. Agric. Ecosyst. Environ. 2016 ,215, 89–99. [CrossRef] 9. Foug è re, H.; Delavaud, C.; Bernard, L. Diets supplemented with starch and corn oil, marine algae, or hydrogenated palm oil differentially modulate milk fat secretion and composition in cows and goats: A comparative study. J. Dairy Sci. 2018,101, 8429–8445. [CrossRef] 10. Toral, P.G.; Chilliard, Y.; Rouel, J.; Leskinen, H.; Shingfield, K.J.; Bernard, L. Comparison of the nutritional regulation of milk fat secretion and composition in cows and goats. J. Dairy Sci. 2015 ,98, 7277–7297. [CrossRef] 11. Abecia, L.; Toral, P.G.; Mart í n-Garc í a, A.I.; Mart í nez, G.; Tomkins, N.W.; Molina-Alcaide, E.; Newbold, C.J.; Y á ñez-Ruiz, D.R. Effect of bromochloromethane on methane emission, rumen fermentation pattern, milk yield, and fatty acid profile in lactating dairy goats. J. Dairy Sci. 2012,95, 2027–2036. [CrossRef] 12. Arco-P é rez, A.; Ramos-Morales, E.; Y á ñez-Ruiz, D.R.; Abecia, L.; Mart í n-Garc í a, A.I. Nutritive evaluation and milk quality of including of tomato or olive by-products silages with sunflower oil in the diet of dairy goats. Anim. Feed Sci. Technol. 2017,232, 57–70. [CrossRef] 13. Criscioni, P.; Fern á ndez, C. Effect of rice bran as a replacement for oat grain in energy and nitrogen balance, methane emissions, and milk performance of Murciano-Granadina goats. J. Dairy Sci. 2016 ,99, 280–290. [CrossRef] [PubMed] 14. Criscioni, P.; Marti, J.V.; P é rez-Baena, I.; Palomares, J.L.; Larsen, T.; Fern á ndez, C. Replacement of alfalfa hay (Medicago sativa) with maralfalfa hay (Pennisetum sp.) in diets of lactating dairy goats. Anim. Feed Sci. Technol. 2016,219, 1–12. [CrossRef] 15. Fern á ndez, C.; Mart í , J.V.; P é rez-Baena, I.; Palomares, J.L.; Ib á ñez, C.; Segarra, J.V. Effect of lemon leaves on energy and C–N balances, methane emission, and milk performance in Murciano-Granadina dairy goats. J. Anim. Sci. 2018,96, 1508–1518. [CrossRef] 16. Fern á ndez, C.; P é rez-Baena, I.; Marti, J.V.; Palomares, J.L.; Jorro-Ripoll, J.; Segarra, J.V. Use of orange leaves as a replacement for alfalfa in energy and nitrogen partitioning, methane emissions and milk performance of Murciano-Granadina goats. Anim. Feed Sci. Technol. 2019,247, 103–111. [CrossRef] 17. Heidarian Miri, V.; Tyagi, A.K.; Ebrahimi, S.H.; Mohini, M. Effect of cumin (Cuminum cyminum) seed extract on milk fatty acid profile and methane emission in lactating goat. Small Rumin. Res. 2013 ,113, 66–72. [CrossRef] 18. Ib á ñez, C.; Criscioni, P.; Arriaga, H.; Merino, P.; Espin ó s, F.J.; Fern á ndez, C. Murciano-granadina goat performance and methane emission after replacing barley grain with fibrous by-products. PLoS ONE 2016 ,11. [CrossRef] 19. Marcos, C.N.; Carro, M.D.; Fern á ndez Yepes, J.E.; Haro, A.; Romero-Huelva, M.; Molina-Alcaide, E. Effects of agroindustrial by-product supplementation on dairy goat milk characteristics, nutrient utilization, ruminal fermentation, and methane production. J. Dairy Sci. 2020,103, 1472–1483. [CrossRef] 20. Rapetti, L.; Bava, L.; Tamburini, A.; Crovetto, G.M. Feeding behaviour, digestibility, energy balance and productive performance of lactating goats fed forage-based and forage-free diets. Ital. J. Anim. Sci. 2005 ,4, 71–83. [CrossRef] 21. Rapetti, L.; Crovetto, G.M.; Galassi, G.; Sandrucci, A.; Sued, G.; Tamburini, A.; Battelli, G. Effect of maize, Rumen-protected fat and whey permeate on energy utilisation and milk fat composition in lactating goats. Ital. J. Anim. Sci. 2002,1, 43–53. [CrossRef] 22. Romero-Huelva, M.; Ramos-Morales, E.; Molina-Alcaide, E. Nutrient utilization, ruminal fermentation, microbial abundances, and milk yield and composition in dairy goats fed diets including tomato and cucumber waste fruits. J. Dairy Sci. 2012,95, 6015–6026. [CrossRef] [PubMed] 23. Chilliard, Y.; Martin, C.; Rouel, J.; Doreau, M. Milk fatty acids in dairy cows fed whole crude linseed, extruded linseed, or linseed oil, and their relationship with methane output. J. Dairy Sci. 2009 ,92, 5199–5211. [CrossRef] [PubMed] Sustainability 2020,12, 4834 8 of 9 24. Mart í nez Mar í n, A.L.; N ú ñez S á nchez, N.; Garz ó n Sigler, A.I.; Blanco, F.P.; De La Fuente, M.A. Relationships between the daily intake of unsaturated plant lipids and the contents of major milk fatty acids in dairy goats. Span. J. Agric. Res. 2015,13, 2171–9292. [CrossRef] 25. Serment, A.; Schmidely, P.; Giger-Reverdin, S.; Chapoutot, P.; Sauvant, D. Effects of the percentage of concentrate on rumen fermentation, nutrient digestibility, plasma metabolites, and milk composition in mid-lactation goats. J. Dairy Sci. 2011,94, 3960–3972. [CrossRef] [PubMed] 26. Sauvant, D.; Schmidely, P.; Daudin, J.J.; St-Pierre, N.R. Meta-analyses of experimental data in animal nutrition. Animal 2008,2, 1203–1214. [CrossRef] 27. Williams, S.R.O.; Hannah, M.C.; Jacobs, J.L.; Wales, W.J.; Moate, P.J. Volatile fatty acids in ruminal fluid can be used to predict methane yield of dairy cows. Animals 2019,9, 1006. [CrossRef] 28. Alemu, A.W.; Dijkstra, J.; Bannink, A.; France, J.; Kebreab, E. Rumen stoichiometric models and their contribution and challenges in predicting enteric methane production. Anim. Feed Sci. Technol. 2011 ,166, 761–778. [CrossRef] 29. Palarea-Albaladejo, J.; Rooke, J.A.; Nevison, I.M.; Dewhurst, R.J. Compositional mixed modeling of methane emissions and ruminal volatile fatty acids from individual cattle and multiple experiments1. J. Anim. Sci. 2017,95, 2467–2480. [CrossRef] 30. Tapio, I.; Snelling, T.J.; Strozzi, F.; Wallace, R.J. The ruminal microbiome associated with methane emissions from ruminant livestock. J. Anim. Sci. Biotechnol. 2017,8, 7. [CrossRef] 31. Urrutia, N.L.; Harvatine, K.J. Acetate dose-dependently stimulates milk fat synthesis in lactating dairy cows. J. Nutr. 2017,147, 763–769. [CrossRef] 32. Shorten, P.R.; Pleasants, T.B.; Upreti, G.C. A mathematical model for mammary fatty acid synthesis and triglyceride assembly: The role of stearoyl CoA desaturase (SCD). J. Dairy Res. 2004 ,71, 385–397. [CrossRef] [PubMed] 33. Mosoni, P.; Chaucheyras-Durand, F.; B é ra-Maillet, C.; Forano, E. Quantification by real-time PCR of cellulolytic bacteria in the rumen of sheep after supplementation of a forage diet with readily fermentable carbohydrates: Effect of a yeast additive. J. Appl. Microbiol. 2007,103, 2676–2685. [CrossRef] [PubMed] 34. Sun, Y.Z.; Mao, S.Y.; Zhu, W.Y. Rumen chemical and bacterial changes during stepwise adaptation to a high-concentrate diet in goats. Animal 2010,4, 210–217. [CrossRef] 35. Yang, S.L.; Bu, D.P.; Wang, J.Q.; Hu, Z.Y.; Li, D.; Wei, H.Y.; Zhou, L.Y.; Loor, J.J. Soybean oil and linseed oil supplementation affect profiles of ruminal microorganisms in dairy cows. Animal 2009 ,3, 1562–1569. [CrossRef] 36. Li, F.; Yang, X.J.; Cao, Y.C.; Li, S.X.; Yao, J.H.; Li, Z.J.; Sun, F.F. Effects of dietary effective fiber to rumen degradable starch ratios on the risk of sub-acute ruminal acidosis and rumen content fatty acids composition in dairy goat. Anim. Feed Sci. Technol. 2014,189, 54–62. [CrossRef] 37. Zhang, Y.; Liu, K.; Hao, X.; Xin, H. The relationships between oddand branched-chain fatty acids to ruminal fermentation parameters and bacterial populations with different dietary ratios of forage and concentrate. J. Anim. Physiol. Anim. Nutr. 2017,101, 1103–1114. [CrossRef] [PubMed] 38. Vlaeminck, B.; Fievez, V.; Cabrita, A.R.J.; Fonseca, A.J.M.; Dewhurst, R.J. Factors affecting oddand branched-chain fatty acids in milk: A review. Anim. Feed Sci. Technol. 2006,131, 389–417. [CrossRef] 39. Lourenço, M.; Ramos-Morales, E.; Wallace, R.J. The role of microbes in rumen lipolysis and biohydrogenation and their manipulation. Animal 2010,4, 1008–1023. [CrossRef] 40. Rico, D.E.; Preston, S.H.; Risser, J.M.; Harvatine, K.J. Rapid changes in key ruminal microbial populations during the induction of and recovery from diet-induced milk fat depression in dairy cows. Br. J. Nutr. 2015 , 114, 358–367. [CrossRef] 41. Bernard, L.; Shingfield, K.J.; Rouel, J.; Ferlay, A.; Chilliard, Y. Effect of plant oils in the diet on performance and milk fatty acid composition in goats fed diets based on grass hay or maize silage. Br. J. Nutr. 2009 ,101, 213–224. [CrossRef] 42. G ó mez-Cort é s, P.; C í vico, A.; De La Fuente, M.A.; N ú ñez S á nchez, N.; Peña Blanco, F.; Martinez Marin, A.L. Effects of dietary concentrate composition and linseed oil supplementation on the milk fatty acid profile of goats. Animal 2018,12, 2310–2317. [CrossRef] [PubMed] Sustainability 2020,12, 4834 9 of 9 43. Martin, C.; Ferlay, A.; Mosoni, P.; Rochette, Y.; Chilliard, Y.; Doreau, M. Increasing linseed supply in dairy cow diets based on hay or corn silage: Effect on enteric methane emission, rumen microbial fermentation, and digestion. J. Dairy Sci. 2016,99, 3445–3456. [CrossRef] [PubMed] 44. Mart í nez Mar í n, A.L.; G ó mez-Cort é s, P.; G ó mez Castro, G.; Ju á rez, M.; P é rez Alba, L.; P é rez Hern á ndez, M.; de la Fuente, M.A. Time-dependent variations in milk fatty acid content of goats fed 3 different plant oils. J. Dairy Sci. 2013,96, 3238–3246. [CrossRef] [PubMed] © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).