Predicting lameness in dairy cattle using untargeted liquid chromatography-mass spectrometry-based metabolomics and machine learning.

Lameness in dairy cattle is a highly prevalent condition that impacts on the health and welfare of dairy cows. Prompt detection and implementation of effective treatment is important for managing lameness. However, major limitations are associated with visual assessment of lameness, which is the most commonly used method to detect lameness. The aims of this study were to investigate the use of metabolomics and machine learning to develop novel methods to detect lameness. Untargeted metabolomics using liquid chromatography-mass spectrometry (LC-MS) alongside machine learning models and a stability selection method were utilized to evaluate the predictive accuracy of differences in the metabolomics profile of first-lactation dairy cows before (during the transition period) and at the time of lameness (based on visual assessment using the 0-3 scale of the Agriculture and Horticulture Development Board). Urine samples were collected from 2 cohorts of dairy heifers and stored at -86°C before analysis using LC-MS. Cohort 1 (n = 90) cows were recruited as current first-lactation cows with weekly mobility scores recorded over a 4-mo timeframe, from which newly lame and nonlame cows were identified. Cohort 2 (n = 30) cows were recruited within 3 wk before calving, and lameness events (based on mobility score) were recorded through lactation until a minimum of 70 d in milk (DIM). All cows were matched paired by DIM ± 14 d. The median DIM at lameness identification was 187.5 and 28.5 for cohort 1 and 2, respectively. The best performing machine learning models predicted lameness at the time of lameness with an accuracy of between 81 and 82%. Using stability selection, the prediction accuracy at the time of lameness was 80 to 81%. For samples collected before and after calving, the best performing machine learning model predicted lameness with an accuracy of 71 and 75%, respectively. The findings from this study demonstrate that untargeted LC-MS profiling combined with machine learning methods can be used to predict lameness as early as before calving and before observable changes in gait in first-lactation dairy cows. The methods also provide accuracies for detecting lameness at the time of observable changes in gait of up to 82%. The findings demonstrate that these methods could provide substantial advancements in the early prediction and prevention of lameness risk. Further external validation work is required to confirm these findings are generalizable; however, this study provides the basis from which future work can be conducted.

A randomised controlled trial to evaluate the impact of indoor living space on dairy cow production, reproduction and behaviour.

As a global society, we have a duty to provide suitable care and conditions for farmed livestock to protect animal welfare and ensure the sustainability of our food supply. The suitability and biological impacts of housing conditions for intensively farmed animals is a complex and emotive subject, yet poorly researched, meaning quantitative evidence to inform policy and legislation is lacking. Most dairy cows globally are housed for some duration during the year, largely when climatic conditions are unfavourable. However, the impact on biology, productivity and welfare of even the most basic housing requirement, the quantity of living space, remains unknown. We conducted a long-term (1-year), randomised controlled trial (CONSORT 10 guidelines) to investigate the impact of increased living space (6.5 m2 vs 3 m2 per animal) on critical aspects of cow biology, behaviour and productivity. Adult Holstein dairy cows (n = 150) were continuously and randomly allocated to a high or control living space group with all other aspects of housing remaining identical between groups. Compared to cows in the control living space group, cows with increased space produced more milk per 305d lactation (primiparous: 12,235 L vs 11,592 L, P < 0.01; multiparous: 14,746 L vs 14,644 L, P < 0.01) but took longer to become pregnant after calving (primiparous: 155 d vs 83 d, P = 0.025; multiparous: 133 d vs 109 d). In terms of behaviour, cows with more living space spent significantly more time in lying areas (65 min/d difference; high space group: 12.43 h/day, 95% CI = 11.70-13.29; control space group: 11.42 h/day, 95% CI = 10.73-12.12) and significantly less time in passageways (64 min/d), suggesting enhanced welfare when more space was provided. A key physiological difference between groups was that cows with more space spent longer ruminating each day. This is the first long term study in dairy cows to demonstrate that increased living space results in meaningful benefits in terms of productivity and behaviour and suggests that the interplay between farmed animals and their housed environment plays an important role in the concepts of welfare and sustainability of dairy farming.

Clearance of Maedi-visna infection in a longitudinal study of naturally infected rams is associated with homozygosity for the TMEM154 resistance allele.

Maedi-visna (MV) is a lentiviral disease of sheep responsible for severe production losses in affected flocks. There are no vaccination or treatment options with control reliant on test and cull strategies. The most common diagnostic methods used at present are combination ELISAs for Gag and Env proteins with virus variability making PCR diagnostics still largely an experimental tool. To assess variability in viral loads and diagnostic tests results, serology, DNA and RNA viral loads were measured in the blood of 12 naturally infected rams repeatedly blood sampled over 16 months. Six animals tested negative in one or more tests at one or more time points and would have been missed on screening programmes reliant on one test method or a single time point. In addition the one animal homozygous for the 'K' allele of the TMEM154 E35K SNP maintained very low viral loads in all assays and apparently cleared infection to below detectable limits at the final time point it was sampled. This adds crucial data to the strong epidemiological evidence that this locus represents a genuine resistance marker for MV infection and is a strong candidate for selective breeding of sheep for resistance to disease.

Automated detection of lameness in sheep using machine learning approaches: novel insights into behavioural differences among lame and non-lame sheep.

Lameness in sheep is the biggest cause of concern regarding poor health and welfare among sheep-producing countries. Best practice for lameness relies on rapid treatment, yet there are no objective measures of lameness detection. Accelerometers and gyroscopes have been widely used in human activity studies and their use is becoming increasingly common in livestock. In this study, we used 23 datasets (10 non-lame and 13 lame sheep) from an accelerometer- and gyroscope-based ear sensor with a sampling frequency of 16 Hz to develop and compare algorithms that can differentiate lameness within three different activities (walking, standing and lying). We show for the first time that features extracted from accelerometer and gyroscope signals can differentiate between lame and non-lame sheep while standing, walking and lying. The random forest algorithm performed best for classifying lameness with an accuracy of 84.91% within lying, 81.15% within standing and 76.83% within walking and overall correctly classified over 80% sheep within activities. Both accelerometer- and gyroscope-based features ranked among the top 10 features for classification. Our results suggest that novel behavioural differences between lame and non-lame sheep across all three activities could be used to develop an automated system for lameness detection.

Schmallenberg virus neutralising antibody responses in sheep.

BACKGROUND: Schmallenberg virus (SBV) is a midge borne virus of cattle and sheep. Infection is typically asymptomatic in adult sheep but fetal infection during pregnancy can result in abortion, stillbirth, neurological disorders and malformations of variable severity in newborn animals. It was first identified in Germany and the Netherlands in 2011 and then circulated throughout Europe in 2012 and 2013. Circulation in subsequent years was low or non-existent until summer and autumn 2016, leading to an increased incidence of deformed newborn lambs and calves in 2016-17. This study reports SBV circulation in October 2016 within a group of 24 ewes and 13 rams. The ewes were monitored at 3 times points over an 11 week period (September to December 2016). RESULTS: Most ewes displayed an increase in SBV VNT with antibody titre increases greater in older, previously exposed ewes. Two ewes had SBV RNA detectable by RT-qPCR, one on 30/09/16 and one on 04/11/16. Of these ewes, one had detectable serum SBV RNA (indicating viraemia) despite pre-existing antibody. The rams had been previously vaccinated with a commercial inactivated SBV vaccine, they showed minimal neutralising antibody titres against SBV 8 months post-vaccination and all displayed increased titre in October 2016. CONCLUSION: This data suggests that SBV circulated for a minimum period of 5 weeks in September to October 2016 in central England. Ewes previously exposed to virus showed an enhanced antibody response compared to naïve animals. Pre-existing antibody titre did not prevent re-infection in at least one animal, implying immunity to SBV upon natural exposure may not be life-long. In addition, data suggests that immunity provided by killed adjuvanted SBV vaccines only provides short term protection (< 8 months) from virus.

Feature Selection and Comparison of Machine Learning Algorithms in Classification of Grazing and Rumination Behaviour in Sheep.

Grazing and ruminating are the most important behaviours for ruminants, as they spend most of their daily time budget performing these. Continuous surveillance of eating behaviour is an important means for monitoring ruminant health, productivity and welfare. However, surveillance performed by human operators is prone to human variance, time-consuming and costly, especially on animals kept at pasture or free-ranging. The use of sensors to automatically acquire data, and software to classify and identify behaviours, offers significant potential in addressing such issues. In this work, data collected from sheep by means of an accelerometer/gyroscope sensor attached to the ear and collar, sampled at 16 Hz, were used to develop classifiers for grazing and ruminating behaviour using various machine learning algorithms: random forest (RF), support vector machine (SVM), k nearest neighbour (kNN) and adaptive boosting (Adaboost). Multiple features extracted from the signals were ranked on their importance for classification. Several performance indicators were considered when comparing classifiers as a function of algorithm used, sensor localisation and number of used features. Random forest yielded the highest overall accuracies: 92% for collar and 91% for ear. Gyroscope-based features were shown to have the greatest relative importance for eating behaviours. The optimum number of feature characteristics to be incorporated into the model was 39, from both ear and collar data. The findings suggest that one can successfully classify eating behaviours in sheep with very high accuracy; this could be used to develop a device for automatic monitoring of feed intake in the sheep sector to monitor health and welfare.

The Applied Development of a Tiered Multilocus Sequence Typing (MLST) Scheme for Dichelobacter nodosus.

Dichelobacter nodosus (D. nodosus) is the causative pathogen of ovine footrot, a disease that has a significant welfare and financial impact on the global sheep industry. Previous studies into the phylogenetics of D. nodosus have focused on Australia and Scandinavia, meaning the current diversity in the United Kingdom (U.K.) population and its relationship globally, is poorly understood. Numerous epidemiological methods are available for bacterial typing; however, few account for whole genome diversity or provide the opportunity for future application of new computational techniques. Multilocus sequence typing (MLST) measures nucleotide variations within several loci with slow accumulation of variation to enable the designation of allele numbers to determine a sequence type. The usage of whole genome sequence data enables the application of MLST, but also core and whole genome MLST for higher levels of strain discrimination with a negligible increase in experimental cost. An MLST database was developed alongside a seven loci scheme using publically available whole genome data from the sequence read archive. Sequence type designation and strain discrimination was compared to previously published data to ensure reproducibility. Multiple D. nodosus isolates from U.K. farms were directly compared to populations from other countries. The U.K. isolates define new clades within the global population of D. nodosus and predominantly consist of serogroups A, B and H, however serogroups C, D, E, and I were also found. The scheme is publically available at https://pubmlst.org/dnodosus/.