Measuring dairy cow welfare with real-time sensor-based data and farm records: a concept study.

Welfare assessment of dairy cows by in-person farm visits provides only a snapshot of welfare and is time-consuming and costly. Possible solutions to reduce the need for in-person assessments would be to exploit sensor data and other routinely collected on-farm records. The aim of this study was to develop an algorithm to classify dairy cow welfare based on sensors (accelerometer and/or milk meter) and farm records (e.g. days in milk, lactation number). In total, 318 cows from six commercial farms located in Finland, Italy and Spain (two farms each) were enrolled for a pilot study lasting 135 days. During this time, cows were routinely scored using 14 animal-based measures of good feeding, health and housing based on the Welfare Quality® (WQ®) protocol. WQ® measures were evaluated daily or approximately every 45 days, using disease treatments from farm records and on-farm visits, respectively. WQ® measures were supplemented with daily temperature-humidity index to account for heat stress. The severity and duration of each welfare measure were evaluated, and the final welfare index was obtained by summing up the values for each cow on each pilot study day, and stratifying the result into three classes: good, moderate and poor welfare. For model building, a machine-learning (ML) algorithm based on gradient-boosted trees (XGBoost) was applied. Two model versions were tested: (1) a global model tested on unseen herd, and (2) a herd-specific model tested on unseen part of the data from the same herd. The version (1) served as an example on the model performance on a herd not previsited by the evaluator, while version (2) resembled a custom-made solution requiring in-person welfare evaluation for model training. Our results indicated that the global model had a low performance with average sensitivity and specificity of 0.44 and 0.68, respectively. For the herd-specific version, the model performance was higher reaching an average of 0.64 sensitivity and 0.80 specificity. The highest classification performance was obtained for cows in poor welfare, followed by cows in good and moderate welfare (balanced accuracy of 0.77, 0.71 and 0.68, respectively). Since the global model had low classification accuracy, the use of the developed model as a stand-alone system based solely on sensor data is infeasible, and a combination of in-person and sensor-based welfare evaluation would be preferable for a reliable welfare assessment. ML-based solutions, even with fair discriminative abilities, have the potential to enhance dairy welfare monitoring.

Animal board invited review: Quantification of resilience in farm animals.

Resilience, when defined as the capacity of an animal to respond to short-term environmental challenges and to return to the prechallenge status, is a dynamic and complex trait. Resilient animals can reinforce the capacity of the herd to cope with often fluctuating and unpredictable environmental conditions. The ability of modern technologies to simultaneously record multiple performance measures of individual animals over time is a huge step forward to evaluate the resilience of farm animals. However, resilience is not directly measurable and requires mathematical models with biologically meaningful parameters to obtain quantitative resilience indicators. Furthermore, interpretive models may also be needed to determine the periods of perturbation as perceived by the animal. These applications do not require explicit knowledge of the origin of the perturbations and are developed based on real-time information obtained in the data during and outside the perturbation period. The main objective of this paper was to review and illustrate with examples, different modelling approaches applied to this new generation of data (i.e., with high-frequency recording) to detect and quantify animal responses to perturbations. Case studies were developed to illustrate alternative approaches to real-time and post-treatment of data. In addition, perspectives on the use of hybrid models for better understanding and predicting animal resilience are presented. Quantification of resilience at the individual level makes possible the inclusion of this trait into future breeding programmes. This would allow improvement of the capacity of animals to adapt to a changing environment, and therefore potentially reduce the impact of disease and other environmental stressors on animal welfare. Moreover, such quantification allows the farmer to tailor the management strategy to help individual animals to cope with the perturbation, hence reducing the use of pharmaceuticals, and decreasing the level of pain of the animal.

A generalised additive model to characterise dairy cows' responses to heat stress.

Heat stress is one of the most critical issues jeopardising animal welfare and productivity during the warm season in dairy cattle farms. The global trend of increase in average and peak temperatures is making the problem more and more serious. Many devices have been introduced in livestock farms to monitor and control temperature-humidity index, as well as animal behaviour and production parameters. The consequent availability of collected databases has increasingly enhanced the research aimed to understand the consequences of heat stress in cattle, in relation to genetic, reproductive, productive and behavioural features. Moreover, these investigations laid the foundations for the development, calibration, validation and test of numerical models quantifying the individual responses to heat stress conditions. In this work, a generalised additive model with mixed effects has been developed to analyse the relationship between milk production, animal behaviour and environmental parameters based on data surveyed in 2016 in an Italian dairy farm. Each cow has been characterised in terms of her response to heat conditions, and the results led to define three classes of susceptibility to heat stress within the herd. These attributes have then been related to the various phenotypic parameters collected by the precision livestock farming devices used in the farm. The study provides a model to understand the effects of heat stress conditions on individual animals in relation to the main parameters describing their rearing conditions; moreover, the results contribute to improve the herd management by lending indications to define targeted treatments according to the cow's characteristics.

Perceptions and practices of Finnish dairy producers on disbudding pain in calves.

Disbudding causes pain-related distress and behavioral changes in calves. Local anesthesia and non-steroidal anti-inflammatory drugs are effective for treating disbudding-related pain. Dairy producers play a key role in whether or not calves to be disbudded are properly medicated. Pain and distress related to disbudding of calves often remains untreated. Thus, we conducted this study to characterize perceptions and practices of dairy producers on disbudding and disbudding-related pain management. A questionnaire was sent to 1,000 randomly selected Finnish dairy producers (response rate: 45%). Our aim was to investigate producer perceptions about disbudding-related pain, the perceived need for pain alleviation before disbudding, and how these perceptions affect the valuing and use of pain alleviation before disbudding. More than 70% of Finnish dairy farms disbud their calves. Producers who ranked disbudding-related pain and need for pain alleviation higher called a veterinarian to medicate calves before disbudding more often than producers who ranked disbudding pain and need for pain alleviation lower. Among respondents who disbudded calves on their farms, 69% stated that disbudding caused severe pain, 63% stated that pain alleviation during disbudding is important, and 45% always had a veterinarian medicate their calves before disbudding. Producers with a herd healthcare agreement with their veterinarian estimated disbudding-related pain to be higher and had a veterinarian medicate calves more often than producers without such an agreement. Producers with tiestall systems and producers who did not use disbudding valued pain alleviation prior to disbudding higher than producers with freestalls and producers who used disbudding.

Short communication: Lameness impairs feeding behavior of dairy cows.

The automated, reliable, and early detection of lameness is an important aim for the future development of modern dairy operations. One promising indicator of lameness is a change in the feeding behavior of a cow. In this study, the associations between feeding behavior and lameness were evaluated. A herd of 50 cows was investigated during the winter season in a freestall barn. Feeding behavior, feed intake, milk yield, and body weight were monitored using electronic feeding troughs and an automated milking system. Gait scoring every second week was used as a measure of lameness. To analyze the effect of lameness on feeding behavior and milk yield, linear mixed models were used. Cows with more severe lameness spent less time feeding per day (104 ± 4, 101 ± 4, and 91 ± 4 min/d for lameness scores 2, 3, and 4, respectively). An interaction between parity and lameness score was detected, with severely lame primiparous cows spending the least time feeding. Severely lame cows fed faster; however, their body weights were lower than for less-lame cows. Increase in lactation stage was associated with longer daily feeding time, longer duration of feeding bouts, and lower feeding rate. Worsening of gait was associated with lower silage intake and less time spent feeding even before severe lameness was scored. The results indicate that lameness is associated with changes in feeding behavior and that such changes could be considered in the future development of remote monitoring systems. It should also be noted that impaired feeding behavior along with lameness can put the welfare of especially early lactating primiparous cows at risk.

Dairy producer attitudes to pain in cattle in relation to disbudding calves.

Pain is an important indicator of poor welfare of livestock. Despite this, pain has largely gone unrecognized in farm animals due to attitudes of producers and veterinarians, although they play a key role in monitoring and managing the perception of animal pain. Producer attitudes toward animal welfare influence livestock management and production. The aim was to quantify dairy producer attitudes to the painfulness of various cattle diseases and disbudding, a painful routine procedure performed on farm to ensure safer handling of cattle. A questionnaire on disbudding-related opinions and practices was sent to 1,000 Finnish dairy producers (response rate: 45%). Attitudes toward disbudding were gauged using a 5-point Likert scale and attitudes to cattle pain scored on an 11-point numerical rating scale. Principal components analysis was used to assess the loadings, which were further tested for differences between producer gender and housing systems with Mann-Whitney U-tests, and between herd milk yield, herd size, and age and work experience of producers with a Kruskal-Wallis test. Four main factors were identified: factor I ("taking disbudding pain seriously"), factor II ("sensitivity to pain caused by cattle diseases"), factor III ("ready to medicate calves myself"), and factor IV ("pro horns"). Female producers took disbudding pain more seriously, were more sensitive to pain caused to cattle by diseases, and were more ready to medicate disbudded calves than male producers. Producers with tie-stalls favored horns over producers with freestalls. Male producers with tie-stalls were sensitive to cattle pain and preferred horns over male producers with freestalls. Female producers with freestalls were more ready to medicate calves, but did not prefer horns more than female producers with tie-stalls. Taking disbudding seriously correlated with sensitivity to pain caused by cattle diseases. Producers with low-milk-yielding herds were less willing to medicate calves and more willing to keep cattle with horns than producers with higher-yielding herds. Older producers were more sensitive to cattle pain than middle-aged and younger producers. No effect was established for taking disbudding pain seriously: the pro-horn factor was associated with work experience, age, and herd size. Women rated pain higher and were more positive toward pain medication for animals than men. Maintaining horns are more important for producers with tie-stalls than for those with freestalls.

Influence of crate height during slaughter transport on the welfare of male turkeys.

1. Poultry are usually transported in crates which provide the birds with very limited space. Slaughter transport of male turkeys is often carried out using crates that are 40 cm or less in height where it is not possible for them to stand up. There is little information on how this physical restriction over many hours affects the birds. 2. The aim of the study was to compare the welfare of male turkeys transported in crates 40 cm and 55 cm in height. Observations on the birds' behaviour during lairage, carcass damage and meat quality were carried out after four commercial slaughter transport journeys. 3. Birds in 40 cm crates panted more and lay down more than birds in 55 cm crates during lairage. A large percentage of the carcasses had some damage. Significantly more birds from the 55 cm crates had scratches on their backs than birds from the 40 cm crates. There was no significant difference in meat quality between birds transported in the two crate heights. 4. Both positive and negative effects of increased crate height were established and there is no evidence from this study that merely increasing crate height improves turkey welfare. Other solutions should therefore be sought in order to improve the welfare of birds during transport.

Measurement of acceleration while walking as an automated method for gait assessment in dairy cattle.

The aims were to determine whether measures of acceleration of the legs and back of dairy cows while they walk could help detect changes in gait or locomotion associated with lameness and differences in the walking surface. In 2 experiments, 12 or 24 multiparous dairy cows were fitted with five 3-dimensional accelerometers, 1 attached to each leg and 1 to the back, and acceleration data were collected while cows walked in a straight line on concrete (experiment 1) or on both concrete and rubber (experiment 2). Cows were video-recorded while walking to assess overall gait, asymmetry of the steps, and walking speed. In experiment 1, cows were selected to maximize the range of gait scores, whereas no clinically lame cows were enrolled in experiment 2. For each accelerometer location, overall acceleration was calculated as the magnitude of the 3-dimensional acceleration vector and the variance of overall acceleration, as well as the asymmetry of variance of acceleration within the front and rear pair of legs. In experiment 1, the asymmetry of variance of acceleration in the front and rear legs was positively correlated with overall gait and the visually assessed asymmetry of the steps (r ≥ 0.6). Walking speed was negatively correlated with the asymmetry of variance of the rear legs (r=-0.8) and positively correlated with the acceleration and the variance of acceleration of each leg and back (r ≥ 0.7). In experiment 2, cows had lower gait scores [2.3 vs. 2.6; standard error of the difference (SED)=0.1, measured on a 5-point scale] and lower scores for asymmetry of the steps (18.0 vs. 23.1; SED=2.2, measured on a continuous 100-unit scale) when they walked on rubber compared with concrete, and their walking speed increased (1.28 vs. 1.22 m/s; SED=0.02). The acceleration of the front (1.67 vs. 1.72 g; SED=0.02) and rear (1.62 vs. 1.67 g; SED=0.02) legs and the variance of acceleration of the rear legs (0.88 vs. 0.94 g; SED=0.03) were lower when cows walked on rubber compared with concrete. Despite the improvements in gait score that occurred when cows walked on rubber, the asymmetry of variance of acceleration of the front leg was higher (15.2 vs. 10.4%; SED=2.0). The difference in walking speed between concrete and rubber correlated with the difference in the mean acceleration and the difference in the variance of acceleration of the legs and back (r ≥ 0.6). Three-dimensional accelerometers seem to be a promising tool for lameness detection on farm and to study walking surfaces, especially when attached to a leg.

Measures of weight distribution of dairy cows to detect lameness and the presence of hoof lesions.

There is increasing interest in automated methods of detecting lame cows. Hoof lesion data and measures of weight distribution from 61 lactating cows were examined in this study. Lame cows were identified with different numerical rating scores (NRS) used as thresholds (NRS >3 and NRS >or=3.5) for lameness. The ratio of weight applied to a pair of legs (LWR) when the cow was standing was calculated using a special weigh scale, and the cows were gait scored using a 1 to 5 NRS. Hoof lesions were scored and the cows placed into 1 of 4 mutually exclusive categories of hoof lesion: a) no lesions, b) moderate or severe hemorrhages, c) digital dermatitis, and d) sole ulcers. Regression analysis and receiver operating characteristic (ROC) curves were used to analyze the relation between hoof lesions and LWR. A clear relationship was found between NRS and LWR for the cows with sole ulcers (R(2)=0.79). The LWR could differentiate cows with sole ulcers from sound cows with no hoof lesions [area under the curve (AUC)=0.87] and lame cows from nonlame cows with lameness thresholds NRS >3 (AUC=0.71) and NRS >or=3.5 (AUC=0.88). There was no relationship between LWR and NRS for cows with digital dermatitis. Measurement of how cows distribute their weight when standing holds promise as a method of automated detection of lameness.

Detection of clinical mastitis with the help of a thermal camera.

Increasing dairy farm size and increase in automation in livestock production require that new methods are used to monitor animal health. In this study, a thermal camera was tested for its capacity to detect clinical mastitis. Mastitis was experimentally induced in 6 cows with 10 microg of Escherichia coli lipopolysaccharide (LPS). The LPS was infused into the left forequarter of each cow, and the right forequarters served as controls. Clinical examination for systemic and local signs and sampling for indicators of inflammation in milk were carried out before morning and evening milking throughout the 5-d experimental period and more frequently on the challenge day. Thermal images of experimental and control quarters were taken at each sampling time from lateral and medial angles. The first signs of clinical mastitis were noted in all cows 2 h postchallenge and included changes in general appearance of the cows and local clinical signs in the affected udder quarter. Rectal temperature, milk somatic cell count, and electrical conductivity were increased 4 h postchallenge and milk N-acetyl-beta-D-glucosaminidase activity 8 h postchallenge. The thermal camera was successful in detecting the 1 to 1.5 degrees C temperature change on udder skin associated with clinical mastitis in all cows because temperature of the udder skin of the experimental and control quarters increased in line with the rectal temperature. Yet, local signs on the udder were seen before the rise in udder skin and body temperature. The udder represents a sensitive site for detection of any febrile disease using a noninvasive method. A thermal camera mounted in a milking or feeding parlor could detect temperature changes associated with clinical mastitis or other diseases in a dairy herd.

Use of force sensors to detect and analyse lameness in dairy cows.

Force sensors were used to detect lameness in dairy cows in two trials. In the first trial, leg weights were recorded during approximately 12,000 milkings with balances built into the floor of the milking robot. Cows that put less weight on one leg or kicked frequently during milking were checked first with a locomotion scoring system and then with a clinical inspection. A locomotion score of more than 2 was considered lame, and these cows' hooves were examined at hoof trimming to determine the cause and to identify any hoof lesions. In the second trial 315 locomotion scores were recorded and compared with force sensor data. The force sensors proved to be a good method for recognising lameness. Computer curves drawn from force sensor data helped to find differences between leg weights, thus indicating lameness and its duration. Sole ulcers and white line disease were identified more quickly by force sensors than by locomotion scoring, but joint problems were more easily detected by locomotion scoring.

A probabilistic neural network model for lameness detection.

A 4-balance system for measuring the leg-load distribution of dairy cows during milking to detect lameness was developed. Leg weights of 73 cows were successfully recorded during almost 10,000 robotic milkings over a period of 5 mo. Cows were scored weekly for locomotion, and lame cows were inspected clinically for hoof lesions. Unsuccessful measurements, caused by cows standing outside the balances, were removed from the data with a special algorithm, and the mean leg loads and number of kicks during milking were calculated. To develop an expert system to automatically detect lameness cases, a model was needed, and a classifying probabilistic neural network model was chosen for the task. The data were divided into 2 parts and 5,074 measurements from 37 cows were used to train a classifying probabilistic neural network model. The operation of the model was evaluated for its ability to detect lameness in the validating data set, which had 4,868 measurements from 36 cows. The model was able to classify 96.2% of the measurements correctly as sound or lame cows, and 100% of the lameness cases in the validation data were identified. The number of measurements (equal to the number of milkings) causing false alarms was 1.1%. The model developed has the potential to be used as an on-farm decision aid and can be used in a real-time lameness monitoring system.

Contactless measurement of cow behavior in a milking robot.

We have worked on automatically measuring the behavior of dairy cows during automatic milking. A milking robot offers a unique possibility for a dynamic measurement of physical data. Four strain gauge scales were installed into a milking robot in order to measure the weight of each leg separately, and a laser distance sensor was placed next to the robot in order to measure the radial movement of the cow's body surface. The data were collected into a PC. Three video cameras were installed to observe the system, and the data were recorded digitally. From the data, the dynamic weight or load of each leg and the respiration rate of a cow could be measured. Different stages of milking were observed, and the changes in behavior during milking were analyzed. The acquired information could be used to judge a cow's restlessness and welfare--for example, leg health and stress.