Influence of Fix Schedule on the Location Accuracy of a Low-Cost GPS Data Logger on Cattle

Global positioning system (GPS) data loggers are commonly used to track the movements and distribution of both wild and domestic animals. However, the expense often poses a challenge for researchers. Recently, there has been a rise in the utilization of affordable and user-friendly GPS data loggers for tracking animal movements, albeit with compromised accuracy. We aimed to identify factors influencing the accuracy of a low-cost GPS data logger (I-gotU GT-600) and to enhance its location accuracy. Initial investigations revealed that recording intervals impacted the location error of the GPS data logger. To elucidate the relationship between recording intervals and location accuracy, we conducted stationary and motion tests. Our findings indicated that recording intervals of less than 15 sec substantially enhances the location accuracy of the low-cost GPS data logger. Our results highlight the relationship between the fix schedule and location accuracy for these GPS data loggers. Our study provides information that enhances the quality of data for researchers using low-cost GPS data loggers for short-term studies in various settings, such as zoos and livestock facilities.

Machine Learning Methods and Visual Observations to Categorize Behavior of Grazing Cattle Using Accelerometer Signals

Accelerometers worn by animals produce distinct behavioral signatures, which can be classified accurately using machine learning methods such as random forest decision trees. The objective of this study was to identify accelerometer signal separation among parsimonious behaviors. We achieved this objective by (1) describing functional differences in accelerometer signals among discrete behaviors, (2) identifying the optimal window size for signal pre-processing, and (3) demonstrating the number of observations required to achieve the desired level of model accuracy,. Crossbred steers (Bos taurus indicus; n = 10) were fitted with GPS collars containing a video camera and tri-axial accelerometers (read-rate = 40 Hz). Distinct behaviors from accelerometer signals, particularly for grazing, were apparent because of the head-down posture. Increasing the smoothing window size to 10 s improved classification accuracy (p < 0.05), but reducing the number of observations below 50% resulted in a decrease in accuracy for all behaviors (p < 0.05). In-pasture observation increased accuracy and precision (0.05 and 0.08 percent, respectively) compared with animal-borne collar video observations.

Exploring 'Wether' Grazing Patterns Differed in Native or Introduced Pastures in the Monaro Region of Australia

Monitoring livestock allows insights to graziers on valuable information such as spatial distribution, foraging patterns, and animal behavior, which can significantly improve the management of livestock for optimal production. This study aimed to understand what potential variables are significant for predicting where sheep spent the most time in native (NP) and improved (IP) paddocks. Wethers (castrated male sheep) were tracked using Global Positioning System (GPS) collars on 15 sheep in the IP and 15 in the NP, respectively, on a property located in the Monaro region of Southern New South Wales, Australia. Trials were performed over four six-day periods in April, July, and November of 2014 and March in 2015. Data were analyzed to understand various trends that may have occurred during different seasons, using random forest models (RFMs). Of the factors investigated, Normalized Difference Vegetation Index (NDVI) was significant (p < 0.01) and highly important for wethers in the IP, but not the NP, suggesting that quality of pasture was key for wethers in the IP. Elevation, temperature, and near distance to trees were important and significant for predicting residency of wethers in the IP, as well as the NP. The result of this study highlights the ability of predictive models to provide insights on behavior-based modelling of GPS data and further enhance current knowledge about location-based choices of sheep on paddocks.

Mapping Welfare: Location Determining Techniques and Their Potential for Managing Cattle Welfare—A Review

Several studies have suggested that precision livestock farming (PLF) is a useful tool for animal welfare management and assessment. Location, posture and movement of an individual are key elements in identifying the animal and recording its behaviour. Currently, multiple technologies are available for automated monitoring of the location of individual animals, ranging from Global Navigation Satellite Systems (GNSS) to ultra-wideband (UWB), RFID, wireless sensor networks (WSN) and even computer vision. These techniques and developments all yield potential to manage and assess animal welfare, but also have their constraints, such as range and accuracy. Combining sensors such as accelerometers with any location determining technique into a sensor fusion system can give more detailed information on the individual cow, achieving an even more reliable and accurate indication of animal welfare. We conclude that location systems are a promising approach to determining animal welfare, especially when applied in conjunction with additional sensors, but additional research focused on the use of technology in animal welfare monitoring is needed.

Spatial Distribution and Habitat Selection of Sarda Cattle in a Silvopastoral Mediterranean Area

Simple Summary

Silvopastoral systems support multiple uses, such as cattle grazing, timber harvesting and the provision of many ecosystem services. The management of livestock movement patterns plays a pivotal role in the sustainable use of silvopastoral systems. Uneven livestock distribution can lead to over- and under-grazed areas, negatively affecting plant and animal diversity, as well as ecosystem services. This study was conducted in a Mediterranean silvopastoral area grazed by autochthonous Sarda cattle to determine the spatial distribution and habitat selection of the cows, who were fitted with GPS tracking collars for this purpose. The total time spent by the animals in different areas was mapped to show the spatial distribution of the cattle in the landscape. Moreover, a preference index was computed for different areas and across different seasons. Overall, the areas where the animals drank and received supplementation were strongly preferred, whereas areas with predominantly rocks were strongly avoided. Grasslands were normally used in proportion to their presence in the area. Forest area was frequented by the cows more in the spring and the summer. These results, representing the first findings concerning Sarda cow grazing in silvopastoral areas, could help farmers to implement actions that help exploit the area more evenly by cows, limiting over- and under-grazing.

Abstract

The beef livestock system in Sardinia is based on suckler cows, often belonging to autochthonous breeds, such as the Sarda breed, and they often graze silvopastoral areas. Besides beef meat, silvopastoral systems (SPSs) provide several Ecosystem Services (ESs), such as timber provision, harvested as wood, and watershed protection. Livestock distribution is a critical factor for the sustainable use of SPSs (e.g., to avoid uneven grazing patterns) and information on patterns of spatial use are required. A study was conducted to determine: (i) the spatial distribution and (ii) the habitat selection of Sarda cattle grazing in a Mediterranean silvopastoral area. Over different seasons, 12 free-roaming adult Sarda cows were fitted with Global Positioning System (GPS) Knight tracking collars to calculate an index mapping of the incidence of livestock in the landscape (LRI) and a preference index (PI) for different areas. Since the PI data were not normally distributed, the Aligned Rank Transform (ART) procedure was used for the analysis. LRI was able to represent the spatial variability in resource utilization by livestock as a LRI map. Overall, the areas where the animals drank and received supplementation were strongly preferred by the cows, reaching PI values in the summer of 19.3 ± 4.9 (median ± interquartile range), whereas areas with predominantly rocks were strongly avoided (the worst PI value in the spring was 0.2 ± 0.6). Grasslands were, in general, used in proportion to their presence in the area, with slightly increased use in the spring (PI 1.1 ± 0.5). Forest area was avoided by cows, except in the spring when it was used in proportion to their presence in the area.

Review: Precision Livestock Farming technologies in pasture-based livestock systems

Precision Livestock Farming (PLF) encompasses the combined application of single technologies or multiple tools in integrated systems for real-time and individual monitoring of livestock. In grazing systems, some PLF applications could substantially improve farmers' control of livestock by overcoming issues related to pasture utilisation and management, and animal monitoring and control. A focused literature review was carried out to identify technologies already applied or at an advanced stage of development for livestock management in pastures, specifically cattle, sheep, goats, pigs, poultry. Applications of PLF in pasture-based systems were examined for cattle, sheep, goats, pigs, and poultry. The earliest technology applied to livestock was the radio frequency identification tag, allowing the identification of individuals, but also for retrieving important information such as maternal pedigree. Walk-over-weigh platforms were used to record individual and flock weights. Coupled with automatic drafting systems, they were tested to divide the animals according to their needs. Few studies have dealt with remote body temperature assessment, although the use of thermography is spreading to monitor both intensively reared and wild animals. Global positioning system and accelerometers are among the most applied technologies, with several solutions available on the market. These tools are used for several purposes, such as animal location, theft prevention, assessment of activity budget, behaviour, and feed intake of grazing animals, as well as for reproduction monitoring (i.e., oestrus, calving, or lambing). Remote sensing by satellite images or unmanned aerial vehicles (UAVs) seems promising for biomass assessment and herd management based on pasture availability, and some attempts to use UAVs to monitor, track, or even muster animals have been reported recently. Virtual fencing is among the upcoming technologies aimed at grazing management. This system allows the management of animals at pasture without physical fences but relies on associative learning between audio cues and an electric shock delivered if the animal does not change direction after the acoustic warning. Regardless of the different technologies applied, some common constraints have been reported on the application of PLF in grazing systems, especially when compared with indoor or confined livestock systems. Battery lifespan, transmission range, service coverage, storage capacity, and economic affordability were the main factors. However, even if the awareness of the existence and the potential of these upcoming tools are still limited, farmers' and researchers' demands are increasing, and positive outcomes in terms of rangeland conservation, animal welfare, and labour optimisation are expected from the spread of PLF in grazing systems.

Factors Affecting Site Use Preference of Grazing Cattle Studied from 2000 to 2020 through GPS Tracking: A Review

Understanding the behaviour of grazing animals at pasture is crucial in order to develop management strategies that will increase the potential productivity of grazing systems and simultaneously decrease the negative impact on the environment. The objective of this review was to summarize and analyse the scientific literature that has addressed the site use preference of grazing cattle using global positioning systems (GPS) collars in the past 21 years (2000–2020) to aid the development of more sustainable grazing livestock systems. The 84 studies identified were undertaken in several regions of the world, in diverse production systems, under different climate conditions and with varied methodologies and animal types. This work presents the information in categories according to the main findings reviewed, covering management, external and animal factors driving animal movement patterns. The results showed that some variables, such as stocking rate, water and shade location, weather conditions and pasture (terrain and vegetation) characteristics, have a significant impact on the behaviour of grazing cattle. Other types of bio-loggers can be deployed in grazing ruminants to gain insights into their metabolism and its relationship with the landscape they utilise. Changing management practices based on these findings could improve the use of grasslands towards more sustainable and productive livestock systems.

Seasonal dynamics of cattle grazing behaviors on contrasting landforms of a fenced ranch in northern China

The number of livestock per unit area is commonly used as a proxy of grazing pressure in both experimental studies and grassland management. However, this practice ignores the impact of landform heterogeneity on the spatial distribution of grazing pressure, leading to localized patches of degraded grassland. The spatial distribution of actual grazing density thus needs to be examined. Owing to the corresponding changes in resource availability and energy consumption as livestock move across an elevation gradient, we predict that livestock will preferentially use low-land and that different temporal patterns of grazing pressure will occur in the contrasting landforms. GPS location data and a machine learning technique were used to identify the seasonal pattern and the factors driving grazing pressure on a fenced ranch. Over both low-land and sand-dune landforms, the proportion of time that livestock spent on foraging increased from 63% in July to 67% in August and 69% in September, and non-foraging behavior decreased correspondingly. In low-land, the log-transformed average foraging density significantly increased from 0.61 (i.e., total foraging behaviors in 5 days measured at 50-s intervals per 10 × 10 m grid) in July to 0.66 in August and 0.88 in September, whereas there was no significant change on sand-dunes. From July to September, the relative area of low-land foraged by cattle accounted for 31%, 35%, and 36%, respectively, and in sand-dunes the proportions increased from 45% to 47% to 51%. In low-land, the foraging density was negatively correlated with biomass (P = .07), total digestible nutrients (P < .05), and crude protein (P = .06) and positively correlated with acid detergent fiber (P < .05), whereas no such relationships were observed in sand-dunes. Our results indicate that topographic features should be considered when managing livestock, especially during periods with adverse conditions of herbage quality and quantity.

Method for Classifying Behavior of Livestock on Fenced Temperate Rangeland in Northern China

Different livestock behaviors have distinct effects on grassland degradation. However, because direct observation of livestock behavior is time- and labor-intensive, an automated methodology to classify livestock behavior according to animal position and posture is necessary. We applied the Random Forest algorithm to predict livestock behaviors in the Horqin Sand Land by using Global Positioning System (GPS) and tri-axis accelerometer data and then confirmed the results through field observations. The overall accuracy of GPS models was 85% to 90% when the time interval was greater than 300–800 s, which was approximated to the tri-axis model (96%) and GPS-tri models (96%). In the GPS model, the linear backward or forward distance were the most important determinants of behavior classification, and nongrazing was less than 30% when livestock travelled more than 30–50 m over a 5-min interval. For the tri-axis accelerometer model, the anteroposterior acceleration (–3 m/s²) of neck movement was the most accurate determinant of livestock behavior classification. Using instantaneous acceleration of livestock body movement more precisely classified livestock behaviors than did GPS location-based distance metrics. When a tri-axis model is unavailable, GPS models will yield sufficiently reliable classification accuracy when an appropriate time interval is defined.

Effects of number of animals monitored on representations of cattle group movement characteristics and spatial occupancy

The number of animals required to represent the collective characteristics of a group remains a concern in animal movement monitoring with GPS. Monitoring a subset of animals from a group instead of all animals can reduce costs and labor; however, incomplete data may cause information losses and inaccuracy in subsequent data analyses. In cattle studies, little work has been conducted to determine the number of cattle within a group needed to be instrumented considering subsequent analyses. Two different groups of cattle (a mixed group of 24 beef cows and heifers, and another group of 8 beef cows) were monitored with GPS collars at 4 min intervals on intensively managed pastures and corn residue fields in 2011. The effects of subset group size on cattle movement characterization and spatial occupancy analysis were evaluated by comparing the results between subset groups and the entire group for a variety of summarization parameters. As expected, more animals yield better results for all parameters. Results show the average group travel speed and daily travel distances are overestimated as subset group size decreases, while the average group radius is underestimated. Accuracy of group centroid locations and group radii are improved linearly as subset group size increases. A kernel density estimation was performed to quantify the spatial occupancy by cattle via GPS location data. Results show animals among the group had high similarity of spatial occupancy. Decisions regarding choosing an appropriate subset group size for monitoring depend on the specific use of data for subsequent analysis: a small subset group may be adequate for identifying areas visited by cattle; larger subset group size (e.g. subset group containing more than 75% of animals) is recommended to achieve better accuracy of group movement characteristics and spatial occupancy for the use of correlating cattle locations with other environmental factors.

GNSS technology and its application for improved reproductive management in extensive sheep systems

The behaviour of Merino ewes during non-oestrus and oestrus were quantified using Global Navigation Satellite System (GNSS) tracking devices and direct visual observation. GNSS devices were attached to neck collars and deployed on mixed-age ewes (38 maiden and 40 experienced ewes) following hormonal oestrus synchronisation. The positional accuracy of the GNSS data was validated through a comparative study of GNSS estimates of each animal’s location compared with direct visual observations. Positional accuracy was estimated at 90–94%, for a 4-m and 6-m-buffer radius, respectively. Ewe speed of movement was calculated from the GNSS data and plotted against hour of the day to determine diurnal activity patterns during non-oestrus and oestrus days. Ewes showed increased speed of movement during the early morning of the anticipated day of oestrus compared with the non-oestrus day (P < 0.001). In addition, ewes that increased their speed of movement by 0.05 m/s received 1.4–28.4 times more mounts depending on the hour of the day (P = 0.02). Ewes also displayed an increased speed of movement in the period leading up to maximum sexual activity, defined as the hour in which ewes received the maximum number of mounts. Thereafter, ewe activity decreased. No difference in sexual activity was detected between maiden and experienced ewes. The present study has demonstrated a change in ewe diurnal activity at oestrus, suggesting the onset of sexual activity can be identified as a period of increased speed of movement followed by a return to ‘normal’ activity. The development of commercial remote autonomous monitoring technologies such as GNSS tracking to detect this change in behaviour could facilitate improved reproductive management of sheep in extensive systems.

Inferring behavioral states of grazing livestock from high-frequency position data alone

Studies of animal behavior are crucial to understanding animal-ecosystem interactions, but require substantial efforts in visual observation or sensor measurement. We investigated how classifying behavioral states of grazing livestock using global positioning data alone depends on the classification approach, the preselection of training data, and the number and type of movement metrics. Positions of grazing cows were collected at intervals of 20 seconds in six upland areas in Switzerland along with visual observations of animal behavior for comparison. A total of 87 linear and cumulative distance metrics and 15 turning angle metrics across multiple time steps were used to classify position data into the behavioral states of walking, grazing, and resting. Five random forest classification models, a linear discriminant analysis, a support vector machine, and a state-space model were evaluated. The most accurate classification of the observed behavioral states in an independent validation dataset was 83%, obtained using random forest with all available movement metrics. However, the state-specific accuracy was highly unequal (walking: 36%, grazing: 95%, resting: 58%). Random undersampling led to a prediction accuracy of 77%, with more balanced state-specific accuracies (walking: 68%, grazing: 82%, resting: 68%). The other evaluated machine-learning approaches had lower classification accuracies. The state-space model, based on distance to the preceding position and turning angle, produced a relatively low accuracy of 64%, slightly lower than a random forest model with the same predictor variables. Given the successful classification of behavioral states, our study promotes the more frequent use of global positioning data alone for animal behavior studies under the condition that data is collected at high frequency and complemented by context-specific behavioral observations. Machine-learning algorithms, notably random forest, were found very useful for classification and easy to implement. Moreover, the use of measures across multiple time steps is clearly necessary for a satisfactory classification.

Virtual herding for flexible livestock management – a review

Free-ranging livestock play a pivotal role globally in the conversion of plant tissue into products and services that support man’s many and changing lifestyles. With domestication came the task of providing livestock with an adequate plane of nutrition while simultaneously managing vegetation for sustainable production. Attempting to meld these two seemingly opposing management goals continues to be a major focus of rangeland research. Demand for multiple goods and services from rangelands today requires that livestock production make the smallest possible ‘negative hoof-print’. Advancements in global navigation satellite system, geographic information systems, and electronic/computing technologies, coupled with improved understanding of animal behaviour, positions virtual fencing (VF) as an increasingly attractive option for managing free-ranging livestock. VF offers an alternative to conventional fencing by replacing physical barriers with sensory cues to control an animal’s forward movement. Currently, audio and electrical stimulation are the cues employed. When VF becomes a commercial reality, manual labour will be replaced in large part with cognitive labour for real-time prescription-based livestock distribution management that is robust, accurate, precise and flexible. The goal is to manage rangeland ecosystems optimally for soils, plants, herbivores in addition to the plant and animal’s microflora. However, maximising the benefits of VF will require a paradigm shift in management by using VF as a ‘virtual herder’ rather than simply as a tool to manage livestock within static physical barriers.

Assessing herbivore foraging behavior with GPS collars in a semiarid grassland

Advances in global positioning system (GPS) technology have dramatically enhanced the ability to track and study distributions of free-ranging livestock. Understanding factors controlling the distribution of free-ranging livestock requires the ability to assess when and where they are foraging. For four years (2008–2011), we periodically collected GPS and activity sensor data together with direct observations of collared cattle grazing semiarid rangeland in eastern Colorado. From these data, we developed classification tree models that allowed us to discriminate between grazing and non-grazing activities. We evaluated: (1) which activity sensor measurements from the GPS collars were most valuable in predicting cattle foraging behavior, (2) the accuracy of binary (grazing, non-grazing) activity models vs. models with multiple activity categories (grazing, resting, traveling, mixed), and (3) the accuracy of models that are robust across years vs. models specific to a given year. A binary classification tree correctly removed 86.5% of the non-grazing locations, while correctly retaining 87.8% of the locations where the animal was grazing, for an overall misclassification rate of 12.9%. A classification tree that separated activity into four different categories yielded a greater misclassification rate of 16.0%. Distance travelled in a 5 minute interval and the proportion of the interval with the sensor indicating a head down position were the two most important variables predicting grazing activity. Fitting annual models of cattle foraging activity did not improve model accuracy compared to a single model based on all four years combined. This suggests that increased sample size was more valuable than accounting for interannual variation in foraging behavior associated with variation in forage production. Our models differ from previous assessments in semiarid rangeland of Israel and mesic pastures in the United States in terms of the value of different activity sensor measurements for identifying grazing activity, suggesting that the use of GPS collars to classify cattle grazing behavior will require calibrations specific to the environment and vegetation being studied.