Technical note: Validation of an automatic recording system to assess behavioural activity level in sheep (Ovis aries)

Behavioral Fingerprinting: Acceleration Sensors for Identifying Changes in Livestock Health

During disease or toxin challenges, the behavioral activities of grazing animals alter in response to adverse situations, potentially providing an indicator of their welfare status. Behavioral changes such as feeding behavior, rumination and physical behavior as well as expressive behavior, can serve as indicators of animal health and welfare. Sometimes behavioral changes are subtle and occur gradually, often missed by infrequent visual monitoring until the condition becomes acute. There is growing popularity in the use of sensors for monitoring animal health. Acceleration sensors have been designed to attach to ears, jaws, noses, collars and legs to detect the behavioral changes of cattle and sheep. So far, some automated acceleration sensors with high accuracies have been found to have the capacity to remotely monitor the behavioral patterns of cattle and sheep. These acceleration sensors have the potential to identify behavioral patterns of farm animals for monitoring changes in behavior which can indicate a deterioration in health. Here, we review the current automated accelerometer systems and the evidence they can detect behavioral patterns of animals for the application of potential directions and future solutions for automatically monitoring and the early detection of health concerns in grazing animals.

Interplay Between Stress and Reproduction: Novel Epigenetic Markers in Response to Shearing Patterns in Australian Merino Sheep (Ovis aries)

In this study, we determined the effect(s) of early shearing on Australian Merino ewes (Ovis aries) and their lambs. To test this research question, we used a suite of field and laboratory methods including GPS collars, wool cortisol, and epigenetic change between ewes and lambs identified using Illumina NovaSeq RRBS. Once shorn ewes (n = 24) were kept on their full fleece throughout the entire gestation period, whereas twice (early) shorn ewes (n = 24) had their wool shorn pre-joining. Top-knot wool sample was taken from ewes during pre-joining, day 50 (mid-gestation), and day 90 (late gestation) for laboratory analysis. Ewes were pregnancy scanned at mid-gestation to determine whether they were early or late parturition (this confirmation is provided by the pregnancy scanner based on fetus size). Top-knot wool sample was also taken from the lambs at weaning for hormone and wool quality testing. Ear tissue was taken from ewes at day 50 (mid-gestation) and from lambs at lamb marking for DNA analysis. Results showed that twice or early shorn ewes grazed 10% higher and maintained stronger body condition than once shorn ewes. Wool cortisol levels were also significantly lower in the early shorn ewes between mid- and late gestation. Lambs bred from twice shorn ewes had on average better visual wool quality parameters in terms of micron, spin finesses, and curvature. For the DNA methylation results, when comparing a group of once sheared with twice sheared ewes, we have discovered one locus (Chr20:50404014) that was significantly differentially methylated [False Discovery Rate (FDR) = 0.005]. This locus is upstream of a protein-coding gene (ENSOARG00000002778.1), which shows similarities to the forkhead box C1 (FOXC1) mRNA using BLAST searches. To further our understanding of the potential interaction between pregnancy status and shearing frequency of the ewes, we performed further differential methylation analysis using a combination of shearing treatment and pregnancy scanning status. The comparisons (1) late pregnancy vs. early pregnancy for ewes with one shearing treatment and (2) late pregnancy vs. early pregnancy for sheep with two shearing treatments were carried out to identify associations between loci and pregnancy duration for sheep with either one or two shearing events. We discovered that 36 gene loci were significantly modulated either between different shearing treatments or late vs. early pregnancy status of ewes. This result suggests that maternal pregnancy and nutritional status during gestation influence DNA methylation. We further investigated DNA methylation in lambs and identified 16 annotated gene loci that showed epigenetic modulation as a result of being born from an early or late stage pregnancy. From the genomics data, we pointed out that ewes go through epigenetic modifications during gestation, and there is a degree of intra-individual variation in the reproductive performance of ewes, which could be due to combination of intrinsic (genetic and physiological) and extrinsic (management and climatic) factors. Collectively, this research provides novel dataset combining physiological, molecular epigenetics, and digital tracking indices that advances our understanding of how Merino ewes respond to shearing frequency, and this information could guide further research on Merino sheep breeding and welfare.

An open-source research tool to study triaxial inertial sensors for monitoring selected behaviors in sheep

The use of automated systems for monitoring animal behavior provides information on individual animal behavior and can be used to enhance animal productivity. However, the advancement of this industry is hampered by technology costs, challenges with power supplies, limited data accessibility, and inconsistent testing approaches for confirming the detection of livestock behaviors. Development of open-source research tools similar to commercially available wearable technologies may contribute to the development of more-efficient and affordable technologies. The objective of this study was to demonstrate an open-source, microprocessor-based sensor designed to monitor and enable differentiation among selected behaviors of adult wethers. The sensor was comprised of an inexpensive espressif ESP-32-WROOM-32 microprocessor with Bluetooth communication, a generic MPU92/50 motion sensor that contains a three-axis accelerometer, three-axis magnetometer, a three-axis gyroscope, and a 5-V rechargeable lithium-ion battery. The open-source Arduino IDE software was used to program the microprocessor and to adjust the frequency of sampling, the data packet to send, and the operating conditions. For demonstration purposes, sensors were placed on six housed sheep for three 1-h increments with concurrent visual behavioral observation. Sensor readings (x-, y-, and z-axis) were summarized (mean and SD) within a minute and compared to animal behavior observations (also on a by-minute basis) using a linear mixed-effect model with animal as a random effect and behavioral classifier as a fixed effect. This analysis demonstrated the basic utility of the sensor to differentiate among animal behaviors based on sensed data (P < 0.001). Although substantial additional work is needed for algorithm development, power source testing, and network optimization, this open-source platform appears to be a promising strategy to research wearable sensors in a generalizable manner.

Validation of the Unesp-Botucatu composite scale to assess acute postoperative abdominal pain in sheep (USAPS)

A scale with robust statistical validation is essential to diagnose pain and improve decision making for analgesia. This blind, randomised, prospective and opportunist study aimed to develop an ethogram to evaluate behaviour and validate a scale to assess acute ovine postoperative pain. Elective laparoscopy was performed in 48 healthy sheep, filmed at one preoperative and three postoperative moments, before and after rescue analgesia and 24 hours after. The videos were randomised and assessed twice by four evaluators, with a one-month interval between evaluations. Statistical analysis was performed using R software and differences were considered significant when p <0.05. Based on the multiple association, a unidimensional scale was adopted. The intra- and inter-observer reliability ranged from moderate to very good (intraclass correlation coefficient ≥ 0.53). The scale presented Spearman correlations > 0.80 with the numerical, simple descriptive, and visual analogue scales, and a correlation of 0.48 with the facial expression scale. According to the mixed linear model, the scale was responsive, due to the increase and decrease in pain scores of all items after surgery and analgesic intervention, respectively. All items on the scale demonstrated an acceptable Spearman item-total correlation (0.56-0.76), except for appetite (0.25). The internal consistency was excellent (Cronbach's α = 0.81) and all items presented specificity > 0.72 and sensitivity between 0.61-0.90, except for appetite. According to the Youden index, the cut-off point was ≥ 4 out of 12, with a diagnostic uncertainty zone of 4 to 5. The area under the curve > 0.95 demonstrated the excellent discriminatory capacity of the instrument. In conclusion, the Unesp-Botucatu pain scale in sheep submitted to laparoscopy is valid, reliable, specific, sensitive, with excellent internal consistency, accuracy, discriminatory capacity, and a defined cut-off point.

Can accelerometer ear tags identify behavioural changes in sheep associated with parturition?

On-animal sensor systems provide an opportunity to monitor ewes during parturition, potentially reducing ewe and lamb mortality risk. This study investigated the capacity of machine learning (ML) behaviour classification to monitor changes in sheep behaviour around the time of lambing using ear-borne accelerometers. Accelerometers were attached to 27 ewes grazing a 4.4 ha paddock. Data were then classified based on three different ethograms: (i) detection of grazing, lying, standing, walking; (ii) detection of active behaviour; and (iii) detection of body posture. Proportion of time devoted to performing each behaviour and activity was then calculated at a daily and hourly scale. Frequency of posture change was also calculated on an hourly scale. Assessment of each metric using a linear mixed-effects model was conducted for the 7 days (day scale) or 12 h (hour scale) before and after lambing. For all physical movements, regardless of the ethogram, there was a change in the days surrounding lambing. This involved either a decrease (grazing, lying, active behaviour) or peak (standing, walking) on the day of parturition, with most values returning to either pre-partum or near-pre-partum levels (all P < 0.001). Hourly changes also occurred for all behaviours (all P < 0.001), the most marked being increased walking behaviour and frequency of posture change. These findings indicate ewes were more restless around the time of parturition. Further application of this research should focus on development of algorithms that can be used to identify onset of lambing and/or time of parturition in pasture-based ewes.

Welfare Assessment on Pasture: A Review on Animal-Based Measures for Ruminants

Simple Summary

Welfare assessment in outdoor and extensive systems has rarely been investigated, and little is known about the most appropriate indicators. This study aimed at compiling a list of animal-based measures of welfare for domestic ruminants raised on outdoor/extensive systems by means of a systematic review. Out of 810 papers retrieved, 52 matched the inclusion criteria and went through an in-depth analysis. According to available literature, 45 indicators have been used to assess welfare on pasture, often following different methodologies. Most indicators were measured by observers even if the use of sensor technologies increased in recent years. Considering the growing interest in pasture-based or grass-fed products, it is suggested that welfare assessment in outdoor/extensive farming systems is carried out by following shared methodologies in order to provide evidence of the higher animal welfare claims that these products often imply compared to indoor systems.

Abstract

Outdoor and extensive farming systems allow animals to behave in a natural way and are often perceived as welfare friendly. Nonetheless, the natural environment poses multiple challenges to the welfare of animals, sometimes hampering their capacity to cope. Welfare assessment in outdoor and extensive systems has been rarely investigated, and little is known about the most appropriate indicators. The aim of this review was to identify animal-based measures of welfare to apply in extensive and pasture-based systems in domestic ruminants. Through the use of a dedicated software for systematic reviews, 810 papers were screened and a total of 52 papers were retained for in-depth analysis. ABM resulting from these papers were initially divided according to the species (cattle and small ruminants, including sheep and goats) and then to four principles: comfort, behavior, feeding and health. The results showed that welfare data were collected applying different methodologies, with an increasing use of sensors in recent years. The need to herd and restrain animals for individual data collection is one of the major constraints to data collection in extensive farming systems. It is suggested that welfare assessment in outdoor/extensive farming systems is carried out by following shared procedures in order to provide evidence of the higher animal welfare claims that these products often imply compared to indoor systems.

Identifying Sheep Activity from Tri-Axial Acceleration Signals Using a Moving Window Classification Model

Behaviour is a useful indicator of an individual animal’s overall wellbeing. There is widespread agreement that measuring and monitoring individual behaviour autonomously can provide valuable opportunities to trigger and refine on-farm management decisions. Conventionally, this has required visual observation of animals across a set time period. Technological advancements, such as animal-borne accelerometers, are offering 24/7 monitoring capability. Accelerometers have been used in research to quantify animal behaviours for a number of years. Now, technology and software developers, and more recently decision support platform providers, are integrating to offer commercial solutions for the extensive livestock industries. For these systems to function commercially, data must be captured, processed and analysed in sync with data acquisition. Practically, this requires a continuous stream of data or a duty cycled data segment and, from an analytics perspective, the application of moving window algorithms to derive the required classification. The aim of this study was to evaluate the application of a ‘clean state’ moving window behaviour state classification algorithm applied to 3, 5 and 10 second duration segments of data (including behaviour transitions), to categorise data emanating from collar, leg and ear mounted accelerometers on five Merino ewes. The model was successful at categorising grazing, standing, walking and lying behaviour classes with varying sensitivity, and no significant difference in model accuracy was observed between the three moving window lengths. The accuracy in identifying behaviour classes was highest for the ear-mounted sensor (86%–95%), followed by the collar-mounted sensor (67%–88%) and leg-mounted sensor (48%–94%). Between-sheep variations in classification accuracy confirm the sensor orientation is an important source of variation in all deployment modes. This research suggests a moving window classifier is capable of segregating continuous accelerometer signals into exclusive behaviour classes and may provide an appropriate data processing framework for commercial deployments.

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.

Predicting Lameness in Sheep Activity Using Tri-Axial Acceleration Signals

Simple Summary

Monitoring livestock farmed under extensive conditions is challenging and this is particularly difficult when observing animal behaviour at an individual level. Lameness is a disease symptom that has traditionally relied on visual inspection to detect those animals with an abnormal walking pattern. More recently, accelerometer sensors have been used in other livestock industries to detect lame animals. These devices are able to record changes in activity intensity, allowing us to differentiate between a grazing, walking, and resting animal. Using these on-animal sensors, grazing, standing, walking, and lame walking were accurately detected from an ear attached sensor. With further development, this classification algorithm could be linked with an automatic livestock monitoring system to provide real time information on individual health status, something that is practically not possible under current extensive livestock production systems.

Abstract

Lameness is a clinical symptom associated with a number of sheep diseases around the world, having adverse effects on weight gain, fertility, and lamb birth weight, and increasing the risk of secondary diseases. Current methods to identify lame animals rely on labour intensive visual inspection. The aim of this current study was to determine the ability of a collar, leg, and ear attached tri-axial accelerometer to discriminate between sound and lame gait movement in sheep. Data were separated into 10 s mutually exclusive behaviour epochs and subjected to Quadratic Discriminant Analysis (QDA). Initial analysis showed the high misclassification of lame grazing events with sound grazing and standing from all deployment modes. The final classification model, which included lame walking and all sound activity classes, yielded a prediction accuracy for lame locomotion of 82%, 35%, and 87% for the ear, collar, and leg deployments, respectively. Misclassification of sound walking with lame walking within the leg accelerometer dataset highlights the superiority of an ear mode of attachment for the classification of lame gait characteristics based on time series accelerometer data.

Large-Brained Animal Models of Huntington's Disease: Sheep

The limitations of using small-brained rodents to model diseases that affect large-brain humans are becoming increasingly obvious as novel therapies emerge. Huntington's disease (HD) is one such disease. In recent years, the desirability of a large-brained, long-lived animal model of HD for preclinical testing has changed into a necessity. Treatment involving gene therapy in particular presents delivery challenges that are currently unsolved. Models using long-lived, large-brained animals would be useful, not only for refining methods of delivery (particularly for gene and other therapies that do not involve small molecules) but also for measuring long-term "off-target" effects, and assessing the efficacy of therapies. With their large brains and convoluted cortices, sheep are emerging as feasible experimental subjects that can be used to bridge the gap between rodents and humans in preclinical drug development. Sheep are readily available, economical to use, and easy to care for in naturalistic settings. With brains of a similar size to a large rhesus macaque, they have much to offer. The only thing that was missing until recently was the means of testing their neurological function and behavior using approaches and methods that are relevant to HD. In this chapter, I will outline the present and future possibilities of using sheep and testing as large animal models of HD.