The objectives of this study were to describe on-farm mortality and to investigate cow- and herd-level risk factors associated with on-farm mortality in Midwest US dairy herds using lactation survival analysis. We analyzed a total of approximately 5.9 million DHIA lactation records from 10 Midwest US states from January 2006 to December 2010. The cow-level independent variables used in the models were first test-day milk yield, milk fat percent, milk protein percent, fat-to-protein ratio, milk urea nitrogen, somatic cell score, previous dry period, previous calving interval, stillbirth, calf sex, twinning, calving difficulty, season of calving, parity, and breed. The herd-level variables included herd size, calving interval, somatic cell score, 305-d mature-equivalent milk yield, and herd stillbirth percentage. Descriptive analysis showed that overall cow-level mortality rate was 6.4 per 100 cow-years and it increased from 5.9 in 2006 to 6.8 in 2010. Mortality was the primary reason of leaving the herd (19.4% of total culls) followed by reproduction (14.6%), injuries and other (14.0%), low production (12.3%), and mastitis (10.5%). Risk factor analysis showed that increased hazard for mortality was associated with higher fat-to-protein ratio (>1.6 vs. 1 to 1.6), higher milk fat percent, lower milk protein percent, cows with male calves, cows carrying multiple calves, higher milk urea nitrogen, increasing parity, longer previous calving interval, higher first test-day somatic cell score, increased calving difficulty score, and breed (Holstein vs. others). Decreased hazard for mortality was associated with higher first test-day milk yield, higher milk protein, and shorter dry period. For herd-level factors, increased hazard for mortality was associated with increased herd size, increased percentage of stillbirths, higher somatic cell score, and increased herd calving interval. Cows in herds with higher milk yield had lower mortality hazard. Results of the study indicated that first test-day records, especially those indicative of negative energy balance in cows, could be helpful to identify animals at high risk for mortality. Higher milk yield per cow did not have a negative association with mortality. In addition, the association between herd-level factors and mortality indicated that management quality could be an important factor in lowering on-farm mortality, thereby improving cow welfare.

The objectives of this study were to determine the effects of prepartum stocking density on social, lying, and feeding behavior of dairy animals and to investigate the relationship between social rank and stocking density. In total, 756 Jersey animals were enrolled in the study approximately 4 wk before expected calving date. This study used 8 experimental units (4 replicates × 2 pens/treatment per replicate), and at each replicate, one pen each of nulliparous and parous (primiparous and multiparous) animals per treatment was enrolled. The 2 treatments were 80% stocking density (80D, 38 animals per pen; each pen with 48 headlocks and 44 stalls) and 100% stocking density (100D, 48 animals per pen). Parous animals were housed separately from nulliparous animals. Animals at 254±3d of gestation were balanced for parity (parous vs. nulliparous) and projected 305-d mature-equivalent milk yield (only parous animals) and randomly assigned to either 80D or 100D. Displacements from the feed bunk were measured for 3h after fresh feed delivery on d 2, 5, and 7 of each week. Feeding behavior was measured for 24-h periods (using 10-min video scan sampling) on d 2, 5, and 7 on wk 1 of every replicate and d 2 and 5 for the following 4 wk. A displacement index (proportion of successful displacements from the feed bunk relative to all displacements the animal was involved in) was calculated for each animal and used to categorize animals into ranking categories of high, middle, and low. Seventy nulliparous and 64 parous focal animals in the 80D treatment and 89 nulliparous and 74 parous focal animals in the 100D were used to describe lying behavior (measured with data loggers). Animals housed at 80D had fewer daily displacements from the feed bunk than those housed at 100D (15.2±1.0 vs. 21.3±1.0 per day). Daily feeding times differed between nulliparous and parous animals at the 2 stocking densities. Nulliparous 80D animals spent 12.4±5.0 fewer minutes per day feeding than nulliparous 100D animals, whereas 100D parous animals tended to spend 7.6±4.5 fewer minutes per day feeding than 80D parous animals. The 2 treatments were not different in the number of lying bouts or lying-bout duration; lying time was longer for 100D on d -33, -29, and -26 and shorter on d -7, -5, and 0 than 80D. The interaction between treatment, parity, and social rank was associated with lying and feeding times. In summary, animals in the 80D treatment had a lower number of displacements from the feed bunk and spent more time lying down near parturition than 100D animals, and 80D nulliparous animals had reduced daily feeding time compared with 100D nulliparous animals. Although these results showed some potential behavior benefits of a prepartum stocking density of 80% compared with 100%, observed changes were small. However, greater stocking density cannot be recommended; more research is needed to evaluate the effects of stocking densities greater than 100% and with other breeds of cattle besides Jersey.

Summary The objective of this study was to compare accuracies of different Bayesian regression models in predicting molecular breeding values for health traits in Holstein cattle. The dataset was composed of 2505 records reporting the occurrence of retained fetal membranes (RFM), metritis (MET), mastitis (MAST), displaced abomasum (DA), lameness (LS), clinical endometritis (CE), respiratory disease (RD), dystocia (DYST) and subclinical ketosis (SCK) in Holstein cows, collected between 2012 and 2014 in 16 dairies located across the US. Cows were genotyped with the Illumina BovineHD (HD, 777K). The quality controls for SNP genotypes were HWE P‐value of at least 1 × 10−10; MAF greater than 0.01 and call rate greater than 0.95. The FImpute program was used for imputation of missing SNP markers. The effect of each SNP was estimated using the Bayesian Ridge Regression (BRR), Bayes A, Bayes B and Bayes Cπ methods. The prediction quality was assessed by the area under the curve, the prediction mean square error and the correlation between genomic breeding value and the observed phenotype, using a leave‐one‐out cross‐validation technique that avoids iterative cross‐validation. The highest accuracies of predictions achieved were: RFM [Bayes B (0.34)], MET [BRR (0.36)], MAST [Bayes B (0.55), DA [Bayes Cπ (0.26)], LS [Bayes A (0.12)], CE [Bayes A (0.32)], RD [Bayes Cπ (0.23)], DYST [Bayes A (0.35)] and SCK [Bayes Cπ (0.38)] models. Except for DA, LS and RD, the predictive abilities were similar between the methods. A strong relationship between the predictive ability and the heritability of the trait was observed, where traits with higher heritability achieved higher accuracy and lower bias when compared with those with low heritability. Overall, it has been shown that a high‐density SNP panel can be used successfully to predict genomic breeding values of health traits in Holstein cattle and that the model of choice will depend mostly on the genetic architecture of the trait.

Objectives of the current experiment were to evaluate plasma concentrations of metabolites and haptoglobin peripartum, uterine health and involution, and follicle growth and resumption of cyclicity of Holstein (HO) and Montbéliarde-sired crossbred cows. Cows (52 HO and 52 crossbred) were enrolled in the study 45 d before expected calving date. Cows had body weight and body condition score recorded on d -45, -14, 0, 1, 28, and 56 relative to calving. Dry matter intake was calculated for a subgroup of cows (25 HO and 38 crossbred) from 6 wk before to 6 wk after calving. Blood was sampled weekly from d -14 to 56 relative to calving for determination of glucose, nonesterified fatty acid, and β-hydroxybutyrate concentrations; from d -7 to 21 relative to calving for determination of haptoglobin concentration; and from d 14 to 56 postpartum for determination of progesterone concentration. Cows were examined at calving and on d 4, 7, 10, and 14 postpartum for diagnosis of postparturient diseases, on d 24 postpartum for diagnosis of purulent vaginal discharge, and on d 42 postpartum for diagnosis of subclinical endometritis. Uteri and ovaries were examined by ultrasonography every 3 d from d 14 to 41 postpartum. Milk yield and composition were measured monthly and yield of milk, fat, protein, and energy-corrected milk were recorded for the first 90 d postpartum. Body weight was not different between Holstein and crossbred cows, but HO cows had reduced body condition score compared with crossbred cows. Even though DMI from 6 wk before to 6 wk after calving tended to be greater for HO cows (16.8 ± 0.7 vs. 15.3 ± 0.5 kg/d), HO cows tended to have more pronounced decline in dry matter intake, expressed in percentage of body weight from d -15 to 0 relative to calving. Energy-corrected milk and nonesterified fatty acid and β-hydroxybutyrate concentrations were not different between breeds. No differences were observed in incidence of retained fetal membranes, metritis, and subclinical endometritis, but HO cows tended to be more likely to have pyrexia from d 0 to 15 postpartum (50.0 vs. 31.4%) and to have greater incidence of purulent vaginal discharge (44.2 vs. 26.5%) than crossbred cows. Holstein cows were more likely to have at least 1 uterine disorder postpartum than crossbred cows (63.5 vs. 36.7%). No differences between breeds were observed in uterine involution. Holstein cows had larger subordinate follicles (10.1 ± 0.4 vs. 8.9 ± 0.5) and a greater number of class III follicles (1.6 ± 0.1 vs. 1.2 ± 0.1) than crossbred cows. Furthermore, the first corpus luteum postpartum of HO cows was diagnosed at a slower rate compared with crossbred cows. Crossbred cows had improved uterine health compared with HO cows and this may have been a consequence of heterosis and (or) breed complementarity and less pronounced decrease in DMI during the last days of gestation.

The objective of this study was to estimate the cost of metritis in dairy herds. Data from 11,733 dairy cows from 16 different farms located in 4 different regions of the United States were compiled for up to 305 d in milk, and 11,581 cows (2,907 with and 8,674 without metritis) were used for this study. Metritis was defined as fetid, watery, red-brownish vaginal discharge that occurs ≤21 d in milk. Continuous outcomes such as 305-d milk production, milk sales