Accurate prediction of calving in dairy cows by applying feature engineering and machine learning.

Prediction of calving is key to dairy cow management. Current trends of increasing herd sizes globally can directly impact the time that farmers spend monitoring individual animals. Automated monitoring on behavioural and physiological changes prior to parturition can be used to develop machine learning solutions for calving prediction. In this study, we developed a machine learning algorithm for the prediction of calving in dairy cows. We demonstrated that temperature and activity index information retrieved from a commercial reticuloruminal bolus sensor can accurately predict calving from 1-day to 5-days in advance. The best prediction solution using data from 82 dairy cows, achieved up to 87.81 % in accuracy, 92.99 % in specificity, 75.84 % in sensitivity, 82.99 % in positive predictive value (PPV), 78.85 % in F-score, and 90.02 % in negative predictive value (NPV) on the test dataset when using information from 2-days in advance and all the subsets of feature characteristics (temperature + drinking + activity). The performance only decreased by 2.45 % points in accuracy, 0.74 % points in specificity, 6.41 % points in sensitivity, 2.45 % points in positive predictive value, 4.91 % points in F-score, and 2.44 % points in negative predictive value on the test dataset when using all feature characteristics and 5-days in advance information compared to using all features and information from 2-days in advance. Full evaluation of the performance of the prediction showed an improvement when using all the different subsets of feature characteristics together (temperature, activity, and drinking) compared to using temperature features only. When adding activity and drinking to the subset of temperature features, an average increase of 2.70, 1.52, 5.40, 4.39, 5.02, 2.13 % points in accuracy, specificity, sensitivity, PPV, F-score, and NPV, respectively, was obtained. Notably, evaluation of feature importance (i.e., relative weight of any given feature in relation to model prediction) showed that 3-5 (depending on the selected days in advance model) of the top ten features were derived from drinking behaviour, showing the relevance that this behaviour can have in the prediction of calving. This algorithm can provide a useful tool for automated calving prediction in dairy cows which has potential for improvement of health, welfare, and productivity in the dairy industry.

Developing and evaluating threshold-based algorithms to detect drinking behavior in dairy cows using reticulorumen temperature.

In this study, we assessed for the first time the use of a reticuloruminal temperature bolus and a thresholding method to detect drinking events and investigated different factors that can affect drinking behavior. First, we validated the detection of drinking events using 16 cows that received a reticuloruminal bolus. For this, we collected continuous drinking behavior data for 4 d using video recordings and ambient and water temperature for the same 4 d. After all the data were synchronized, we performed 2 threshold algorithms: a general-fixed threshold and a cow-day specific threshold algorithm. In the general-fixed threshold, a positive test was considered if the temperature of any cow fell below a fixed threshold; in the cow-day specific threshold, a positive test was considered when the temperature of specific cows fell below the threshold value deviations around the mean temperature of the cow for that day. The former was evaluated using a threshold varying between 35.7 and 39.5°C, and the latter using the formula µ-n10σ, where µ = mean of the temperature of each cow for one day, n = 1, 2, …, 20, and σ = standard deviation of the temperature of each cow on that day. The performance of the validation of detection using each of the threshold types was computed using different metrics, including overall accuracy, precision, recall (also known as sensitivity), F-score, positive predictive value, negative predictive value, false discovery rate, false omission rate, and Cohen's kappa statistic. The findings of the first study showed that the cow-day specific threshold of n = 10 performed better (true positives = 466; false positives = 167; false negatives = 165; true negatives = 8,416) than using a general-fixed threshold of 38.1°C (true positives = 449; false positives = 181; false negatives = 182; true negatives = 8,402). With the information gained in this first study, we investigated the different factors associated with temperature drop characteristics per cow: number of drops, mean amplitude of the drop, and mean recovery time. For this, we used data from 54 cows collected for almost 1 yr to build a mixed-effect multilevel model that included days in milk, parity, average monthly milk production, and ambient temperature as explanatory variables. Cow characteristics and ambient temperature had significant effects on drinking events. Our results provide a platform for automated monitoring of drinking behavior, which has potential value in prediction of health and welfare in dairy cattle.

Increased dietary protein in the second trimester of gestation increases live weight gain and carcass composition in weaner calves to 6 months of age.

Genetically similar nulliparous Polled Hereford heifers from a closed pedigree herd were used to evaluate the effects of dietary protein during the first and second trimester of gestation upon foetal, placental and postnatal growth. Heifers were randomly allocated into two groups at 35 days after artificial insemination (35 days post conception (dpc)) to a single bull and fed high (15.7% CP) or low (5.9% CP) protein in the first trimester (T1). At 90 dpc, half of each nutritional treatment group changed to a high- or low-protein diet for the second trimester until 180 dpc (T2). High protein intake in the second trimester increased birth weight in females (P=0.05), but there was no effect of treatment upon birth weight when taken over both sexes. Biparietal diameter was significantly increased by high protein in the second trimester with the effect being greater in the female (P=0.02), but also significant overall (P=0.05). Placental weight was positively correlated with birth weight, fibroblast volume and relative blood vessel volume (P<0.05). Placental fibroblast density was increased and trophoblast volume decreased in the high-protein first trimester treatment group (P<0.05). There was a trend for placental weight to be increased by high protein in the second trimester (P=0.06). Calves from heifers fed the high-protein treatment in the second trimester weighed significantly more on all occasions preweaning (at 1 month (P=0.0004), 2 months (P=0.006), 3 months (P=0.002), 4 months (P=0.01), 5 months (P=0.03), 6 months (P=0.001)), and grew at a faster rate over the 6-month period. By 6 months of age, the calves from heifers fed high nutrition in the second trimester weighed 33 kg heavier than those fed the low diet in the second trimester. These results suggest that dietary protein in early pregnancy alters the development of the bovine placenta and calf growth to weaning.

Evaluation of treatments for claw horn lesions in dairy cows in a randomized controlled trial.

Lameness is one of the most significant endemic disease problems facing the dairy industry. Claw horn lesions (principally sole hemorrhage, sole ulcer, and white line disease) are some of the most prevalent conditions. Despite the fact that thousands of animals are treated for these conditions every year, experimental evidence is limited on the most effective treatment protocols. A randomized, positively controlled clinical trial was conducted to test the recovery of newly lame cows with claw horn lesions. Animals on 5 farms were locomotion scored every 2wk. Cows were eligible for recruitment if they had 2 nonlame scores followed by a lame score and had a claw horn lesion on a single claw of a single foot. Following a therapeutic trim, enrolled cows were randomly allocated to 1 of 4 treatments: treatment 1-no further treatment (positive control; TRM), treatment 2-trim plus a block on the sound claw (TB), treatment 3-trim plus a 3-d course of the nonsteroidal anti-inflammatory drug (NSAID) ketoprofen (TN), treatment 4-trim plus a block plus ketoprofen (TBN). The primary outcome measure was locomotion score 35d after treatment, by an observer blind to treatment group. Descriptive statistics suggested that treatment groups were balanced at the time of enrollment, that is, randomization was successful. Based on a sound locomotion score (score 0) 35d after treatment, the number of cures was 11 of 45 (24.4%) for TRM, 14 of 39 (35.9%) for TB, 12 of 42 (28.6%) for TN, and 23 of 41 (56.1%) for TBN. The difference between TBN and TRM was significant. To test for confounding imbalances between treatment groups, logistic regression models were built with 2 outcomes, either sound (score 0) or nonlame (score 0 or 1) 35d after treatment. Compared with TRM, animals that received TBN were significantly more likely to cure to a sound outcome. Farm, treatment season, lesion diagnosis, limb affected, treatment operator, and stage of lactation were included in the final models. Our work suggests that lameness cure is maximized with NSAID treatment in addition to the common practices of therapeutic trimming and elevation of the diseased claw using a block when cows are newly and predominantly mildly lame.

Relationship between locomotor play of dairy calves and their weight gains and energy intakes around weaning.

Play behaviour has been proposed as a measure of good welfare in growing animals and locomotor play by calves is often reduced after weaning off milk. Adjusting weaning age according to individual calves' abilities to eat solid feed maintains energy intake and weight gain during weaning. We investigated the effects of this method of weaning on locomotor play of calves and the relationship between locomotor play and energy intake and weight gains. We measured the running behaviour of 56 Holstein heifer calves before and after weaning. Calves were housed in groups of eight, fed milk, grain starter and hay from automated feeders. Weaning began when their voluntary intake of grain starter reached either 200 or 400 g/day, and weaning was completed when starter intake reached either 800 or 1600 g/day. Before weaning, older calves ran less than young ones; and the duration of running correlated with weight gains and digestible energy intake. Immediately after weaning, digestible energy intake and locomotor play decreased but no correlation was observed between these variables. One week after weaning, the duration of running was correlated with both energy intake and weight gain. Digestible energy intake increased but locomotor play continued to decrease. The amount of running a calf does after weaning partly reflects energy intake and weight gain, supporting suggestions that locomotor play is good indicator of welfare and fitness of growing animals. However, the decline in locomotor play following weaning is not solely due to decreased energy intake.