The objectives of this study were to determine the feasibility of measuring feed intake in commercial tie-stall dairies and infer genetic parameters of feed intake, yield, somatic cell score, milk urea nitrogen, body weight (BW), body condition score (BCS), and linear type traits of Holstein cows. Feed intake, BW, and BCS were measured on 970 cows in 11 Pennsylvania tie-stall herds. Historical test-day data from these cows and 739 herdmates who were contemporaries during earlier lactations were also included. Feed intake was measured by researchers once per month over a 24-h period within 7 d of 6 consecutive Dairy Herd Information test days. Feed samples from each farm were collected monthly on the same day that feed intake was measured and were used to calculate intakes of dry matter, crude protein, and net energy of lactation. Test-day records were analyzed with multiple-trait animal models, and 305-d fat-corrected milk yield, dry matter intake, crude protein intake, net energy of lactation intake, average BW, and average BCS were derived from the test-day models. The 305-d traits were also analyzed with multiple-trait animal models that included a prediction of 40-wk dry matter intake derived from National Research Council equations. Heritability estimates for 305-d intake of dry matter, crude protein, and net energy of lactation ranged from 0.15 to 0.18. Genetic correlations of predicted dry matter intake with 305-d dry matter, crude protein, and net energy of lactation intake were 0.84, 0.90, and 0.94, respectively. Genetic correlations among the 3 intake traits and fat-corrected milk yield, BW, and stature were moderate to high (0.52 to 0.63). Results indicate that feed intake measured in commercial tie-stalls once per month has sufficient accuracy to enable genetic research. High-producing and larger cows were genetically inclined to have higher feed intake. The genetic correlation between observed and predicted intakes was less than unity, indicating potential variation in feed efficiency.

The objectives of this study were to calculate the heritability of feed efficiency and residual feed intake, and examine the relationships between feed efficiency and other traits of productive and economic importance. Intake and body measurement data were collected monthly on 970 cows in 11 tie-stall herds for 6 consecutive mo. Measures of efficiency for this study were: dry matter intake efficiency (DMIE), defined as 305-d fat-corrected milk (FCM)/305-d DMI, net energy for lactation efficiency (NELE), defined as 305-d FCM/05-d NEL intake, and crude protein efficiency (CPE), defined as 305-d true protein yield/305-d CP intake. Residual feed intake (RFI) was calculated by regressing daily DMI on daily milk, fat, and protein yields, body weight (BW), daily body condition score (BCS) gain or loss, the interaction between BW and BCS gain or loss, and days in milk (DIM). Data were analyzed with 3- and 4-trait animal models and included 305-d FCM or protein yield, DM, NEL, or CP intake, BW, BCS, BCS change between DIM 1 and 60, milk urea nitrogen, somatic cell score, RFI, or an alternative efficiency measure. Data were analyzed with and without significant covariates for BCS and BCS change between DIM 1 and 60. The average DMIE, NELE, and CPE were 1.61, 0.98, and 0.32, respectively. Heritability of gross feed efficiency was 0.14 for DMIE, 0.18 for NELE, and 0.21 for CPE, and heritability of RFI was 0.01. Body weight and BCS had high and negative correlations with the efficiency traits (-0.64 to -0.70), indicating that larger and fatter cows were less feed efficient than smaller and thinner cows. When BCS covariates were included in the model, cows identified as being highly efficient produced 2.3 kg/d less FCM in early lactation due to less early lactation loss of BCS. Results from this study suggest that selection for higher yield and lower BW will increase feed efficiency, and that body tissue mobilization should be considered.

Traditionally, high-forage, low-concentrate diets fed ad libitum have been the primary system of feeding dairy heifers. However, high-concentrate diets can be fed at restricted intakes to reach desired rates of gain and increase nutrient efficiency. A total mixed ration containing high corn silage (CS; HCS: 77% CS, 23% concentrate) or low CS (LCS: 67% concentrate, 33% CS) was fed at restricted intakes in 2 trials to evaluate nutrient utilization by growing heifers. In the first trial, 4 ruminally cannulated heifers (298 +/- 16 kg of body weight) were fed to study differences in rumen pH, volatile fatty acid and ammonia concentrations, and mass of rumen contents. In situ determinations were made on the total mixed ration and CS. Low CS rations were digested more rapidly in situ when compared with HCS (4.5 vs. 2.3 +/- 0.3%/h), and no differences were observed in CS digestibility when incubated in the rumen of heifers fed either ration. Mean rumen pH tended to be lower for LCS than for HCS (5.9 vs. 6.2 +/- 0.1). Individual and total rumen volatile fatty acid concentrations and rumen ammonia concentration were not different between treatments. Total mass of rumen contents was lower for LCS. In the second trial, four 6-mo-old heifers (172 +/- 14 kg of body weight) and four 12-mo-old heifers (337 +/- 10 kg of body weight) were used. Digestibility of dry matter was greater for the LCS than the HCS diet in both age groups (76.3 vs. 71.1% for 12-mo-old heifers; 71.4 vs. 68.9% for 6-mo-old heifers). Apparent digestibility of N was not different between treatments; however, retained N was higher for the LCS diets for both age groups. Fecal output was significantly reduced in the LCS diets for both age groups. Feeding low-forage, high-concentrate diets to growing dairy heifers at restricted intakes, although more highly digestible, resulted in few significant differences in rumen fermentation patterns and lower fecal output.

The objectives of this study were to quantify the relationships of various definitions of feed utilization with both fertility and productive life. Intake and body measurement data were collected monthly on 970 cows in 11 tie-stall herds for 6 consecutive months. Measures of feed utilization for this study were dry matter intake (DMI), dry matter intake efficiency (DME, defined as 305-d fat-corrected milk/305-d DMI), DME with intake adjusted for maintenance requirements (DMEM), crude protein efficiency (defined as 305-d protein yield/305-d crude protein intake), and 2 definitions of residual feed intake (RFI). The first, RFI(reg), was calculated by regressing daily DMI on daily milk, fat, and protein yields, body weight (BW), daily body condition score (BCS) gain or loss, the interaction between BW and BCS gain or loss, and days in milk. The second, RFI(NRC), was estimated by subtracting 305-d DMI predicted according to their fat-corrected milk and BW from actual 305-d DMI. Data were analyzed with 8-trait animal models and included one measure of feed utilization and milk, fat, and protein yields, BW, BCS, days open (DO), and productive life (PL). The genetic correlation between DME and DO was 0.53 (± 0.19) and that between DME and PL was 0.66 (± 0.10). These results show that cows who had higher feed efficiency had greater DO (undesirable) and greater PL (desirable). Results were similar for the genetic correlation between DO and crude protein efficiency (0.42). Productive life had genetic correlations of -0.22 with BW and -0.48 with BCS, suggesting that larger, fatter cows in this study had shorter PL. Correlations between estimated breeding values for feed utilization and official sire genetic evaluations for fertility were in agreement with the results from the multiple-trait models. Selection programs intended to enhance feed efficiency should factor relationships with functional traits to avoid unfavorable effects on cow fertility.

Data from 879 Holstein cows from 11 tie-stall herds in Pennsylvania were analyzed to determine the effects of nutritional management practices on the level of genetic expression for milk, fat, and protein yields. Environments were defined according to the amount of dry matter refusals at the end of 24h for the average cow (DMR), diet crude protein percentage (CP), and diet NE(L) concentration. Sire predicted transmitting ability (PTA) was available for all cows, whereas 775 cows were genotyped and received a molecular breeding value (MBV) for milk, fat, and protein yields. Milk, fat, and protein yields were regressed on sire PTA and cow MBV independently in addition to combined breeding values (CBV) of sire PTA and cow MBV. Four-trait animal models with fat-corrected milk yield in high and low environments plus either body weight or body condition score in high and low environments treated as separate traits were also evaluated. Regressions on sire PTA (0.31 for fat yield to 0.54 for milk yield) were significantly lower in the 5 herds that had the lowest average DMR than in the 6 herds with highest average DMR (0.82 for fat yield to 1.11 for protein yield). The regressions of milk and protein yield on CBV were also significantly lower in the 5 herds with low NE(L) concentration in the ration than in herds that had high NE(L) concentration. Genetic correlations from animal models showed that large cows were more affected by low DMR, CP, and NE(L) concentration than smaller cows. Efforts to minimize feed wastage must ensure that cows receive adequate nutrient intake to avoid suppression of genetic potential for yield, particularly for larger cows.