Search

Direct measurements of methane (CH4) from individual animals are difficult and expensive. Predictions based on proxies for CH4 are a viable alternative. Most prediction models are based on multiple linear regressions (MLR) and predictor variables that are not routinely available in commercial farms, such as dry matter intake (DMI) and diet composition. The use of machine learning (ML) algorithms to predict CH4 emissions from across-country heterogeneous data sets has not been reported. The objectives were to compare performances of ML ensemble algorithm random forest (RF) and MLR models in predicting CH4 emissions from proxies in dairy cows, and assess effects of imputing missing data points on prediction accuracy. Data on CH4 emissions and proxies for CH4 from 20 herds were provided by 10 countries. The integrated data set contained 43,519 records from 3,483 cows, with 18.7% missing data points imputed using k-nearest neighbor imputation. Three data sets were created, 3k (no missing records), 21k (missing DMI imputed from milk, fat, protein, body weight), and 41k (missing DMI, milk fat, and protein records imputed). These data sets were used to test scenarios (with or without DMI, imputed vs. nonimputed DMI, milk fat, and protein), and prediction models (RF vs. MLR). Model predictive ability was evaluated within and between herds through 10-fold cross-validation. Prediction accuracy was measured as correlation between observed and predicted CH4, root mean squared error (RMSE) and mean normalized discounted cumulative gain (NDCG). Inclusion of DMI in the model improved within and between-herd prediction accuracy to 0.77 (RMSE = 23.3%) and 0.58 (RMSE = 31.9%) in RF and to 0.50 (RMSE = 0.327) and 0.13 (RMSE = 42.71) in MLR, respectively than when DMI was not included in the predictive model. When missing DMI records were imputed, within and between-herd accuracy increased to 0.84 (RMSE = 18.5%) and 0.63 (RMSE = 29.9%), respectively. In all scenarios, RF models out-per

Direct measurements of methane (CH4) from individual animals are difficult and expensive. Predictions based on proxies for CH4 are a viable alternative. Most prediction models are based on multiple linear regressions (MLR) and predictor variables that are not routinely available in commercial farms, such as dry matter intake (DMI) and diet composition. The use of machine learning (ML) algorithms to predict CH4 emissions from across-country heterogeneous data sets has not been reported. The objectives were to compare performances of ML ensemble algorithm random forest (RF) and MLR models in predicting CH4 emissions from proxies in dairy cows, and assess effects of imputing missing data points on prediction accuracy. Data on CH4 emissions and proxies for CH4 from 20 herds were provided by 10 countries. The integrated data set contained 43,519 records from 3,483 cows, with 18.7% missing data points imputed using k-nearest neighbor imputation. Three data sets were created, 3k (no missing records), 21k (missing DMI imputed from milk, fat, protein, body weight), and 41k (missing DMI, milk fat, and protein records imputed). These data sets were used to test scenarios (with or without DMI, imputed vs. nonimputed DMI, milk fat, and protein), and prediction models (RF vs. MLR). Model predictive ability was evaluated within and between herds through 10-fold cross-validation. Prediction accuracy was measured as correlation between observed and predicted CH4, root mean squared error (RMSE) and mean normalized discounted cumulative gain (NDCG). Inclusion of DMI in the model improved within and between-herd prediction accuracy to 0.77 (RMSE = 23.3%) and 0.58 (RMSE = 31.9%) in RF and to 0.50 (RMSE = 0.327) and 0.13 (RMSE = 42.71) in MLR, respectively than when DMI was not included in the predictive model. When missing DMI records were imputed, within and between-herd accuracy increased to 0.84 (RMSE = 18.5%) and 0.63 (RMSE = 29.9%), respectively. In all scenarios, RF models out-per

The objective of this study was to evaluate short-term variations of trans fatty acids (TFA) in plasma lipoproteins and ruminal fermentation parameters of non-lactating cows subjected to ruminal pulses of vegetable oils. Three non-lactating, non-pregnant Holstein cows, each with a ruminal cannula, were arranged in a 3 × 3 Latin square design with three-day pulsing periods and four-day washout intervals between treatments. Cows were treated with single ruminal pulses of: (1) control (skimmed milk (SM); 500 mL); (2) soybean oil (SO; 250 g/d in 500 mL of SM) and (3) partially-hydrogenated vegetable oil (PHVO; 250 g/d in 500 mL of SM). Time changes after infusion in TFA contents were only observed for plasma C18:1 trans-4, trans-5 and trans-12, and high-density lipoprotein fraction C18:1 trans-9. After ruminal pulses, concentration of acetate decreased linearly; molar concentrations of propionate and valerate increased linearly; molar concentrations of butyrate and isovalerate changed quadratically and were greater at 1 h than at other times. There was an accumulation of several C18:1 TFA in plasma and lipoproteins, especially on the third day of pulsing. Overall, naturally occurring C18:1 TFA isomers (produced during ruminal biohydrogenation of SO) and preformed TFA (supplied by PHVO) elicited differential TFA partitioning and transport in plasma and lipoproteins.

The objective of this study was to determine the effect of long-term supplementation of unsaturated oil on lipid metabolism and transcription of genes involved in lipid metabolism in subcutaneous adipose tissue (SAT) of mid-lactating dairy cows. The objective was achieved by supplementing dairy cows with soybean oil (SO; high in linoleic acid) or fish oil (FO; high in EPA and DHA) for 63 days (nine weeks). Cows were fed a control diet with no added lipid, or diets containing SO or FO (n = 5 cows/group). At the onset of the experiment (day 0) and on days 21, 42, and 63 of supplementation, blood and SAT samples were collected from each animal. Oil supplementation increased cholesterol and NEFA in plasma, with a greater effect of SO compared to FO. Concentration of BUN was lower in SO compared to control and FO at the end of the trial. Transcription of few genes was affected by dietary lipids: FABP4 had lowest expression in FO followed by SO and control. ACACA and FASN had higher expression in FO. Transcription of SCAP was higher but expression of INSIG1 was lower in SO. Overall, results revealed that compared to control, SO and FO had lipogenic effect in SAT.

This study analyzed effects of vegetable oils fed to dairy cows on abundance of genes related to lipid metabolism in milk somatic cells (MSC). During 63 days, 15 cows were allocated to 3 treatments: a control diet with no added lipid the same diet supplemented with olive oil (OO, 30 g/kg DM) or hydrogenated vegetable oil (HVO, 30 g/kg DM). On days 21, 42 and 63, MSC were obtained from all cows. Relative abundance of genes involved in lipid metabolism in MSC from cows fed control on days 42 and 63 was compared with relative abundance at day 21 to evaluate fold-changes. Those genes without changes over the time were selected to analyze effects of OO and HVO. Compared with control, on day 42, PLIN2 and THRSP were upregulated by OO. Compared with control, on day 21, HVO up regulated ACACA, down regulated FABP3, and on day 63 THRSP and FABP4 were down regulated. Dietary oil supplementation (3% DM) had a modest nutrigenomic effect on different biological functions such as acetate and FA activation and intra-cellular transport, lipid droplet formation, and transcription regulation in MSC.

This study evaluated which lipoprotein fractions are involved in C18:1 trans fatty acid (tFA) transport and whether there is a difference in transport between naturally occurring C18:1 tFA isomers [produced during rumen biohydrogenation of soybean oil (SO) or fish oil (FO)] and preformed tFA [supplied by partially hydrogenated vegetable oil (PHVO)]. Two non-lactating non-pregnant rumen fistulated cows (body weight ± SD = 778 ± 88 kg), were grouped in a 2 × 3 crossover design with 3 d rumen pulsing periods. Treatments were: 1) SO [control; 250 mL/d in 500 mL/d of skimmed milk (SM)]; 2) FO (250 g/d in 500 mL/d of SM) and 3) FO + PHVO (125 g/d FO + 125 g/d PHVO in 500 mL/d of SM). Treatments were supplied via rumen fistula. Fish oil increased C18:1 t9 concentration in high density lipoprotein triglyceride (HDL-TG) and low density lipoprotein triglyceride (LDL-TG). FO+PHVO increased C18:1 t9 concentration in very low density lipoprotein triglyceride (VLDL-TG). FO increased C18:1 t11 concentration in low density lipoprotein cholesterol esters (LDL-CE) and VLDL-TG. FO+PHVO increased C18:1 t11 concentration in VLDL-TG. Results showed that in bovine plasma there may be a differential transport mechanism in which the VLDL fraction is the main lipoprotein transporting tFA (primarily associated with TG) and is likely to be more responsive than other fractions to variation in supply of dietary tFA.

The objective of this study was to evaluate the effects of supplementation of dairy cows with different fatty acid sources (soybean oil (SO) and fish oil (FO)) on milk production, milk composition, milk fatty acid profile, and physicochemical and sensory characteristics of ice cream. During 63 days, fifteen Holstein cows averaging 198 ± 35 days in milk were assigned to three groups: control diet with no added lipid (n = 5 cows); and supplemented diets with SO (n = 5 cows; unrefined SO; 30 g/kg DM) or FO (n = 5 cows; FO from unrefined salmon oil; 30 g/kg DM). Milk production, milk fat, and milk protein were not affected by treatments. Saturated fatty acids in milk fat were decreased with SO and FO compared with control. C18:2 cis-9, cis-12 was increased with SO whereas C18:2 cis-9, trans-11, C20:3n-3, C20:3n-6, C20:5n-3, and C22:6n-3 were the highest with FO. Draw temperature and firmness were higher in SO compared to control and FO ice creams. Melting resistance was higher in FO compared with control and SO ice creams. Supplementation of cow diets with SO and FO did not have detrimental effects on milk production, or ice cream physicochemical and sensory characteristics.

E. Vargas-Bello-Pérez, and P.C. Garnsworthy. 2013. Trans fatty acids and their role in the milk of dairy cows. Cien. Inv. Agr. 40(3): 449-473. Lipids obtained from dairy products are an important part of the human diet in many countries. Approximately 75% of the total consumption of fat from ruminant animals comes from bovine milk fat. Trans fatty acids (tFA) are produced during biohydrogenation of mono- and poly-unsaturated FA in the rumen. They are mixtures of positional and geometrical isomers that are incorporated into the milk fat of lactating cows. The most important sources of tFA in the human diet are partially hydrogenated vegetable oils and ruminant milk and meat products. Ruminant-derived lipids often contain 1-8% of total fatty acids as tFA, which are predominantly 18:1 isomers. The most common FA in ruminant fat is vaccenic acid (18:1 trans-11) (VA), accounting for 60-80% of total tFA. Unlike other tFA, VA can be converted to rumenic acid (RA) through the action of stearoyl coenzyme-A desaturase. Today, consumers are becoming aware of the relationship between dietary fat, health maintenance, and disease prevention. These concerns have increased the need to investigate the metabolic fate and bioactivity of dietary FA. By altering the nutrition of cows, farmers can markedly and rapidly modulate the FA composition of milk FA. The largest changes can be obtained either by feeding animals high-quality forage, particularly fresh pasture, or by adding plant or marine oils to the diet. Given that economic factors define future profits for farmers, diet manipulation may be the most practical and appropriate approach to change milk's FA composition., E. Vargas-Bello-Pérez y P.C. Garnsworthy. 2013. Ácidos grasos trans y su rol en la leche de vacas lecheras. Cien. Inv. Agr. 40(3): 449-473. La grasa de productos lácteos es muy importante en la dieta humana en muchos países. Aproximadamente, el 75% del consumo total de grasa de productos de rumiantes proviene de la grasa láctea. Los ácidos grasos trans (AGT) son producidos durante la hidrogenación ruminal de AG poliinsaturados. Son una mezcla de isómeros que son incorporados en la grasa láctea de vacas lactantes. La principal fuente de AGT de la dieta humana proviene de los aceites vegetales parcialmente hidrogenados y de la leche y carne de rumiantes. Usualmente los lípidos de la leche y carne de rumiantes contienen entre 1 a 8% de AGT del total de AG presentes, y en su mayoría compuestos por isómeros 18:1. El AGT más común en la grasa de rumiantes es el vaccénico (18:1 trans-11), el cual representa entre un 60 a 80% del total de AGT. A diferencia de los AGT de los aceites vegetales parcialmente hidrogenados, los AGT de rumiantes (ácido vaccénico) pueden ser convertidos en ácido ruménico a través de la enzima estearoil coenzima-A desaturasa. Actualmente, los consumidores están conscientes de la relación entre la grasa de la dieta, la salud y la prevención de enfermedades. Estas preocupaciones han motivado la necesidad de investigar el destino metabólico y la bioactividad de los AG de la dieta. Mediante la manipulación de la nutrición de la vaca lechera, los productores pueden modular y alterar la composición de la grasa láctea. Los cambios más grandes se pueden observar mediante la calidad del forraje de la dieta (pastos) o mediante la utilización de suplementos vegetales o marinos en la dieta. Dado que la factibilidad económica está definida por las ganancias de los productores, la manipulación de la dieta de la vaca lechera es la alternativa más factible y práctica para modificar la composición de la grasa láctea.