پژوهش های علوم دامی (Mar 2022)
Study of feed efficiency based on residual feed intake and 4% fat-corrected milk in Iranian lactating Holstein dairy cows
Abstract
Introduction: To increase the revenue, understanding feed efficiency (FE) in dairy cows and its improvement is essential. Dry matter intake (DMI) is fundamentally important in nutrition because it establishes the amount of nutrients available to an animal for health and production. Residual feed intake (RFI) is calculated as the residual in the linear model to predict feed intake of individual animal (Connor et al, 2015). RFI is essentially the difference between an individual’s observed feed consumption and its predicted feed consumption (Bauman et al, 2012). An animal with a negative RFI consumes less feed than expected for its level of production thus is more efficient. RFI is independent of production level hence recent attention has been given to using RFI as a tool to assess the feed efficiency in dairy cattle for purposes of genetic selection. On the other hand, feed conversion efficiency (FCE) based on 4% fat corrected milk (FCM 4%) is also considered as a factor for calculating feed efficiency. Therefore, the objective of present experiment was to investigate the dynamics of RFI and FCM 4%/DMI in Iranian Holstein dairy cows. Material and methods: Thirty lactating Iranian Holstein cows (10 primiparous and 20 multiparous), averaging 594 ± 62.6 kg of body weight, 38.81 ± 6.22 kg of milk/d, and 94.5 ± 21.5 day postpartum, were fed a diet balanced with CPM Dairy V3 ration software. Diet consisted of 40% forage and 60% concentrate and was fed as total mixed ration (TMR). Cows were housed in individual tie stalls and milked three times daily (0800, 1600 and 2400 h). Cows were fed once per day a fresh diet after morning milking ad libitum and orts were removed and weighed daily before the next morning feeding. Water was available ad libitum. Milk yield was recorded electronically at each milking, and milk samples were obtained from 3 consecutive milkings per week. Milk samples were analyzed for fat, true protein, and lactose with infrared spectroscopy. Body weight (BW) for each cow was recorded 2 consecutive days per month immediately after the morning milking. Daily body weight change (∆BW) was calculated based on body weight for each cow at the beginning and end of each period (30- and 60-days period). Body condition score (BCS) was determined on a 5-point scale in 0.25 increments by a trained investigator and recorded for each cow at the beginning and end of each period (30- and 60-days period). Also, milk energy output (MilkE; Mcal/d), metabolic body weight for a cow (MBW), and energy expended for body tissue gain (ΔBodyE; Mcal/d) were estimated based on NRC 2001 equations. Dry matter intake for an individual cow during each 30- and 60-days period was regressed as a function of major energy sinks through four different models (model 1, model 2, model 3, and model 4) using Minitab software (version 19). To define RFI, DMI was modeled as follows: 〖RFI model 1: DMI〗_i= β_0+ β_1 MilkE_i+ β_2 lMBW_i+ β_3 WkBodyE_i+〖β_4 od〖∆BW〗_i +W〗_5 +WParity_i+ β_6 r〖Week of lactation〗_i+ ε_i 〖RFI model 2: DMI〗_i= β_0+ β_1 ×〖FCM 4%〗_i+ β_2 × MBW_i+ β_3 × 〖∆BW〗_i+ ε_i 〖RFI model 3: DMI〗_i= β_0+ β_1 ×MilkE_i+ β_2 × MBW_i+ β_3 × ∆BodyE_i+β_4 × Parity_i+ ε_i 〖RFI model 4: DMI〗_i= β_0+ β_1 ×MilkE_i+ ε_i Where DMIi was the observed DMI, MilkEi was the observed milk energy output, MBWi was the average BW0.75, ΔBodyEi was the estimated change in body energy, based on measured BW and BCS, ∆BWi was the daily weight body change, Parityi was the parity, Week of lactationi was week of lactation, and FCM 4%i was 4% fat-corrected milk for ith cow. RFI was defined as the error term (ε_i) in the model. Also, we reported Pearson correlation coefficient between DMI, RFI and FCM 4%/DMI with measured and estimated traits for a 30-days and 60-days period. Results and discussion: The obtained results from RFI in present study indicated that population of dairy cattle can be classified based on RFI. Adjusted R2 for models 1, 2, 3 and 4 were 88.51%, 78.82%, 80.05%, and 64.41%, respectively, in the period of 60-days. The mean ± SD for models 1, 2, 3 and 4 were 0 ± 0.86, 0 ± 1.25, 0 ± 1.18, and 0 ± 1.68 kg DM per a day, respectively. For model 1 (in 60-days period, full model), milk energy output (MilkE), metabolic body weight for a cow (MBW), energy expended for body tissue gain (ΔBodyE), daily body weight change (∆BW), and week of lactation were significant (P<0.05) except for parity which showed a significant trend (P<0.1). Partial regression coefficients for MilkE, MBW, ΔBodyE, ∆BW, parity, and week of lactation for the model 1 used to predict DMI were 0.566, 0.1, 3.46, -9.51, 4.13, and 0.1771, respectively. For model 3 (in 60-days period), milk energy output (MilkE), metabolic body weight for a cow (MBW), and energy expended for body tissue gain (ΔBodyE) were significant (P<0.05) except for parity. Partial regression coefficients for MilkE, MBW, ΔBodyE, and parity in the model 3, were 0.429, 0.1381, 0.791, and 0.229, respectively. Pearson correlation coefficient for RFI between model 1 with 2, 3 and 4 was 0.817, 0.728, and 0.515, respectively (P<0.01). The same trend was observed for a 30-days period. Pearson correlation coefficient for DMI between 60-days and 30-days period was 0.994 and for RFI between 60-days and 30-days period was 0.882. Also, based on Pearson correlation coefficient for DMI, RFI model 1 and FCM 4%/DMI with other biological parameters, we observed that there were the reasonable correlations, significant at the P=0.01. Surprisingly, there was a negative correlation between FCM 4%/DMI and milk protein percentage (P<0.0001). Also, there was a positive significant correlation between RFI model 1 with milk fat and protein percentage (P<0.1). Conclusion: Measuring feed efficiency through RFI in a 30-days period is predictable. Based on the adjusted model R2, the model 1 and its parameter describe the DMI in accurate way. However, it seem that scientific exploration necessary for finding other parameters to improve and predication accuracy of model 2. Therefore, finding the effective models would result in an accurate estimation of RFI for individual dairy cows, classifying efficient and inefficient dairy cows correctly and clarifying the reasons for these differences through a holistic approach. Further research is required to provide an explanation for the significant negative correlation between FCE with milk protein and positive significant trend between RFI model 1 with milk fat and protein percentage.
Keywords
