Using Real-Time GNSS Tracking Tags to Monitor Alpaca Activity in an Australian Extensive Production System

Abstract

Australian alpacas contribute to a developing alternative fibre industry with an increasing number of larger-scale enterprises requiring real-time management options. This study aimed to investigate the ability of GNSS real-time tracking tags to monitor alpaca herd behaviour in an extensive production system and assess their suitability as a future management tool. A total of 32 alpacas were fitted with collar-mounted GNSS tracking livestock tags, and an additional 32 alpacas were used as a control group without tags. Both Huacaya (n = 32) and Suri (n = 32) breeds were included. There was no effect of treatment on body condition score change (p = 0. 3648). Breed had a significant effect on distance travelled (p < 0.0184), with Suri alpacas travelling 1.03 (±0.058) km and Huacayas 0.9 (±0.058) km per day. Season significantly impacted the distance travelled each day (p< 0.0001), with alpacas moving a greater distance in winter and spring compared to summer and autumn. The alpacas displayed an increase in activity between 0600 and 1600, with the majority (60%) of their activity occurring during daylight hours. This study outlines normal paddock behaviour for extensively raised alpacas in Australia and showcases the potential for GNSS remote monitoring technology to be utilised as a management tool.

1. Introduction

Technology is becoming an increasingly common tool for understanding and monitoring livestock behaviour worldwide. Global Navigation Satellite Systems (GNSS) estimate the positional location of the device [1]. GPS (Global Position System) is a common type of GNSS device used in agriculture. GNSS and accelerometers have a high accuracy of >98% (calculated as the percentage of well-classified instances [2]) when monitoring grazing and resting behaviours as well as movement tracking [2,3]. GNSS technology, when used with GIS or motion sensors, has a slightly reduced accuracy (still above 90%) [4], suggesting GNSS-based technology can provide accurate animal tracking for individual livestock in an extensive herd system. As this technology tracks movement, the technology and associated digital platforms have the ability to track and identify a variety of aspects to improve health and management, an example being their use in the dairy industry for real-time monitoring of health traits for animals. This includes rumination and movement, to identify abnormal behaviour (including heat stress) or important events such as calving [5,6] to advocate early intervention where needed for improved welfare outcomes. Outside of animal health and behaviour monitoring, GNSS tracking technology can assist producers in identifying grazing patterns and making informed grazing manangement decisions, saving both time and money [7,8]. GNSS tracking devices attached to sheep collars have also been successfully used to detect changes in individual and flock movements, determine activity patterns, and monitor animal welfare in extensively managed flocks [8,9,10,11]. The use of GNSS technology in extensive production systems extends beyond monitoring activity, as it can enable producers to see the location of their animals. In herds grazing large areas, using GNSS technology and digital platforms, producers can see where in the paddock individual animals from the herd are located, providing real-time information on pasture or vegetation being grazed, distance to water access, and commonly used areas (for example, for resting or birthing) [12].

The Australian alpaca industry is a growing industry focused on fibre production from small to large-scale enterprises [13]. It is largely unknown how alpacas raised in Australia behave in a grazing environment, particularly in an extensive production system, and as the primary livestock species in the herd (i.e., not as herd guardians). Despite alpacas commonly being compared to and co-grazed with sheep, the daytime grazing behaviour between the two species varies in the published literature [14,15,16]. Additionally, it is known that alpacas in higher altitude environments in their native habitat travel more distance then those at slightly lower altitudes to meet their nutritional requirements [17]. However, information on the distances travelled by alpacas raised out of their native habitat is limited, with minimal published literature for Australian production systems, particular across 24 h periods [16]. As the Australian alpaca industry continues to develop, understanding paddock behaviour is crucial to developing maagement and handling recommendations for optimal animal welfare outcomes [18,19]. Previously, GNSS technology has been mounted on alpaca collars to monitor the guardian behaviour when cohabitating with sheep [11]. In the Australian alpaca industry, alpacas are commonly kept with collars or neckbands as an additional method of identification, suggesting GNSS device attachment is a suitable option for real-time animal monitoring if reliable information can be collected. It is important to gain an understanding of alpaca grazing behaviour as this will assist producers in considering ways to improve animal health, welfare and production efficiency [7]. This study aimed to use GNSS real-time tracking tags to investigate baseline alpaca herd activity in an extensive production system and assess the suitability of the technology for future use in real-time monitoring and improved management through early identification of illness or injury.

2. Materials and Methods

2.1. Animal Usage and Location

A total of 64 adult female alpacas (32 Suri and 32 Huacaya) were inducted into the trial at a commercial alpaca property in the Southern Highlands, NSW, Australia. At the time of induction, animals were over 2 years old and had a body condition score (BCS) greater than 2 (scale 1–5). The study was conducted across 10 months from January to October 2024 to collect data across all seasons. A total of 5 animals were removed from the study at different time points due to ill-thrift (2) or death due to causes unrelated to the study (3). Body condition score measurements were collected by the same trained person in the middle of summer, autumn, winter and spring in the study year. The animals used in this study experienced regular interactions with humans. The University of Sydney Animal Ethics Committee (ethics number 2023/2333) granted animal ethics approval.

2.2. Herd Management

The alpacas were run together as one herd for the duration of the study, moving between four paddocks. Paddock size ranged from 0.89 ha to 2.38 ha, with all paddocks being used for this study located in the same area of the property. Paddock rotation was conducted by the producer based on pasture availability. This study was conducted concurrently with Boughey et al. [16]. The alpacas were housed in the same paddock for a period of 3 days in the middle of each season for visual behaviour monitoring [16]. The methodology and results of the visual behaviour monitoring component are reported in detail in Boughey et al. [16].

2.3. Real Time Tracking Tags

Smart Paddock Bluebell Cattle Tags (Smart Paddock, Melbourne, Australia) were used to monitor the behaviour of alpacas in a treatment group (n = 16 Huacaya, n = 16 Suri). The tags require a standard male button pin for application (to the ear or collar) and contain a small solar panel that provides power for the tags. The tags are also encased in a water-tight moulding. The tags were applied to a buckle collar, with the collar fitted to the alpaca (Figure 1). Collar fit was determined by ensuring two fingers could move between the collar and the alpaca’s neck. A control group (n = 16 Huacaya, n = 16 Suri) was included to assess any negative impacts on body condition. The control group had a Velcro neckband without a tag. The neckbands on the control group were lighter as they did not contain a tag to assess if the presence of the tags had an impact on the alpacas’ body condition score. Collars and neckbands were commonly used on the property as part of normal management practice, with the alpacas used in the study wearing long-term collars. All tags were replaced in late June due to a product fault with the initial batch of tags, due to the moulding on the tags not being water-tight. This resulted in some missing data due to the tags malfunctioning after significant rain events that the property experienced in late autumn. Replacement with new tags resolved this issue.

The tags were programmed to update the location (GPS ping) every 15 min to 1 h, depending on the tag’s battery voltage. Higher battery voltage, usually due to direct sunlight on the tag, increased the update frequency. With every update, the tags recorded the location coordinates (degrees), date, time, battery voltage, tag temperature, average and standard deviation acceleration data for the X-, Y-, and Z-axes. A LoRaWAN (Long Range Wide Area Network) gateway from Smart Paddock was used to receive data pings from the tags (Smart Paddock, Melbourne, Australia). Tag location and movement were visualised on the Smart Paddock online producer platform.

2.4. Spatio-Temporal Patterns

Density heat maps were generated using R (Version 4.4.2 [20]) from coordinate data points during a 3-day monitoring period in the middle of each season when the alpacas were in the same paddock (in line with the methodological approach published by [16], conducted concurrently with this study). The ggmap [21], ggplot2 [21] and sf [22] packages were utilised to conduct the base mapping and plotting the paddock outlines and coordinate points.

2.5. Statistics

Statistical analyses were conducted in R (Version 4.4.2 [20]) and Excel (Version 16.98 [23]). Data were pulled from the Smart Paddock Platform using AzureTableStor in R [20]. Data were cleaned in Excel [23], removing incomplete tag readings (latitude and longitude values equalling zero) and assigning each tag to a unique animal ID number to account for the change in tags during the trial. Data were categorised as day or night based on a day being between 0700 and 1700 and night between 1700 and 0500 inclusive. Distance travelled was calculated using the Haversine equation [24] via the distHaversine function in R, for every recorded latitude and longitude for each animal throughout the trial. Grouped average distance calculations were completed for the hour of the day and period (day and night) in R [20] for qualitative analysis. Linear models (LM) were run to investigate the effect of the tags on body condition score change between the treatment and control groups. LMs were also run on the distance data generated from the tags to investigate the effect of season, period (day and night) and breed on average distance travelled. A p-value of <0.05 was considered significant.

3. Results

3.1. Alpaca Health

There was no significant difference in body condition score (BCS) between the treatment and control groups (p = 0.3648). However, season had a significant impact on BCS change (p = 0.0066). The change in BCS between autumn and winter significantly differed compared to between summer and autumn (p = 0.0016), with a small increase between autumn and winter. Summer to autumn and winter to spring both indicated marginal increases. No other seasonal comparison displayed a difference.

3.2. Alpaca Activity Levels

The alpacas moved 1.08 (±0.72) km per day on average for the duration of the study. The average daily distance travelled (km) significantly varied between seasons (p < 0.0001) with higher activity levels in winter and spring (

Table 1. Average daily distance travelled per season.

Season	Average Daily Distance (km)	SE
Summer	1.07	0.007
Autumn	0.43	0.015
Winter	1.57	0.029
Spring	1.56	0.013

). There was a significant effect of breed on the distance travelled (p = 0.0108). On average, Huacaya alpacas travelled a shorter distance, moving 0.9 km (±0.058) compared to the Suri alpacas, who travelled 1.03 km (±0.058) per day.

The distance travelled per hour varied significantly (p <0.001). The activity record increased between 0400 and 0600, with high activity levels between 0600 and 1600 when levels decreased (Figure 2).

The difference in distance travelled between day and night periods (kilometres) was significant (p < 0.0001). The majority (60%) of the movement occurred during the day, with 40% during the night (Figure 3).

3.3. Spatio-Temporal Patterns

There was a change in the spatio-temporal pattern trends between seasons over the 3-day periods where the alpacas were monitored in the same paddock (Figure 4). There was a higher density in summer located on the right-hand side of the paddock, which is the location of a treeline providing shade. The locations were more widely spread in the cooler seasons of autumn and winter. Spring showed more variation in location compared to summer; however, animal location was more concentrated at the top of the paddock compared to the cooler seasons.

4. Discussion

With an increase in precision farming technologies available for use in livestock [8,10,25], it is relevant to explore opportunities to improve livestock health and welfare, as well as to reduce labour costs for producers [6,7,8]. Although precision farming and livestock monitoring technology includes satellite imagery for pasture management, water sensors, fence sensors, GNSS and accelerometer tags [4,10], this study focused on the use of GNSS tags to monitor alpaca behaviour in an extensive grazing system.

4.1. Alpaca Activity in an Extensive Production System

The use of GNSS tags across four seasons enabled changes in movement distances to be compared, a first in Australian alpacas. A significant effect of season was observed through the average daily distance travelled, with higher activity in winter and spring compared with summer. It is plausible to conclude that the reduced activity in summer could be due to higher temperatures and coincide with the higher amounts of resting, which was also reported in [16]. In the summer season, the alpacas in this study were also more concentrated in one area of the paddock, near the fence and treeline compared to being more uniformly distributed in the other seasons. This highlights that in warmer periods, alpacas display increases levels of resting as seen through reduced walking activity and grazing [16]. Interestingly, there was a significant drop in activity in autumn, indicated by almost half the daily distance travelled compared to summer and one-third of winter and spring. This could be attributed to adverse weather conditions experienced by the region, with large amounts of rain that resulted in localised flooding on the property. However, a study by Lachica et al. [26] also reported that grazing goats displayed a significant reduction between summer and autumn in the daily distance travelled while grazing, most likely due to seasonal variation in feed availability. As the grazing conditions and species are different, more research is required to understand if the reduction in movement by the alpacas seen in this study is a common occurrence or due to unexpected external conditions.

Matthews et al. [11] used a collar-mounted GNSS device to monitor the behaviour of two herd guardian alpacas cohabitating with sheep, reporting similar behaviour trends and diurnal activity between the species. It has been previously reported that alpacas feed regularly, interspersed with resting periods during the day [27,28]. Although in both of these studies, the alpacas were housed indoors overnight [27,28], a similar trend was observed in this study where alpacas remained in the paddock overnight. Activity levels, measured by the average distance travelled, began to increase between 0400 and 0500 before peaking at 0600 and remaining high for the duration of the daylight hours. Alpacas displayed a steep reduction in activity reaching minimal movement by 1800, which is reflected in the study by Scheibe et al. [27] despite the differences in nighttime housing. This study also found that even with the same space to move during the night, the alpacas conducted 60% of their movement between 0600 and 1600, which is supported by the trends reported for alpacas in other countries and production systems [28,29]. Understanding peak activity periods may assist in planning on-farm management practices, such as aligning supplement feeding with peak grazing times.

Understanding the spatio-temporal pattern of grazing livestock enables the visualisation of land use and herd movement [30], which is important for both animal and sustainable land management [31,32]. In this study, there were clear visual seasonal differences for paddock usages, which could be attributed to the location of shade and shelter as well as pasture availability. The spatio-temporal movement of grazing cattle has been shown to vary with the availability of feed or feed type preferences [32]. Improving the knowledge of paddock and environment use by grazing livestock has the potential to guide paddock design, pasture management practices, and improve environmental management through understanding the factors that drive livestock movement [31,32,33]. Currently, there are no comparable data for the spatio-temporal movement of alpacas, highlighting an area for further research to improve production efficiency and environmental management of alpacas in extensively raised systems.

This study examined both Suri and Huacaya behaviour, research which has been limited previously in Australian extensive production systems. In this study, there was a significant difference between the average daily distance travelled between the breeds, with Suri alpacas travelling further than Huacayas. Differentiated behaviour between different sheep breeds grazed in extensive production systems has been reported previously [34,35]. In sheep, it has been found that factors including selection pressure of predators, exposure to humans and body size can influence behaviour with regards to social spacing in herds, resulting in differences in grazing behaviour [34]. The social behaviour of alpacas in extensive systems, as the sole species being grazed, is not well researched, highlighting an area for future development to understand alpaca herd social structure in a commercial production setting. Variations in behaviour between breeds have been reported in other livestock, with cattle grazing in a mountainous grassland environment displaying differences in movement and travel distances between three cattle breeds [36]. However, in the study by Pauler et al. [36], the behavioural differences were attributed to variation in anatomical structures such as hoof size. In alpacas, it is unlikely that this explains differences in activity, as the breeds are similar in adult size and anatomical attributes aside from fleece structure. Further research is required to understand if this difference is a recurring trend and to identify the external and animal-related factors influencing the distance travelled per day.

4.2. Effect of Real-Time Tracking Technology on Animal Health

GNSS and accelerometer technology in ear tags or collars are becoming increasingly common in livestock production to improve animal health and optimise productivity [3,6,8,10]. A vital aspect to consider when trialling new technology for livestock management is ensuring that the technology does not result in negative impacts on health and welfare. The use of the collar mounted tags in this study displayed no significant reduction in body condition scores between the treatment and control groups, suggesting that the use of the tags has minimal negative impact on alpaca health when applied to the animal via a collar, supporting the continued use of this technology based on the information available from this study.

Real-time and continuous tracking technology can identify unwell or injured animals by using changes from normal behaviour [10,37]. This type of information can also be used to identify animals that are showing early signs of illness or ill-thrift [10]. The benefit of early detection is improved recovery rates, reduced disease spread and it can also reduce the economic burden of treating late-stage or widely spread issues. As the normal behaviour of cattle and sheep is more widely documented compared to alpacas, the real-time monitoring systems have been able to be used to create production associations with live movement data, such as comparing levels of rumination and with milk yield and quality in dairy cattle [38]. Furthering this, collar-mounted accelerometers were able to measure how long dairy cattle were experiencing heat stress through increased breathing motions, providing a practical application of early heat stress detection [6]. The benefit of remote animal monitoring technology in extensive livestock systems extends into utilising this tool to make informed production decisions, including in reproductive management [9]. Through understanding and identifying changes in normal behaviour, not only can producers identify health issues such as heat stress [6], but the information can also be used to detect oestrus as both cows and ewes are found to show increased activity in the mornings during the oestrus period [9,37]. Accelerometer technology, which is often paired with GNSS in monitoring tags and collars has also been successfully used to detect calving times [38], which enables producers to be aware of animal activity and provide early intervention if needed. The implementation of real-time monitoring collars in other livestock industries provides a starting point to base industry adoption and future development for this technology in other species including alpacas. However, it is crucial to have an understanding of normal behaviours in order to use real-time monitoring to detect behavioural changes for welfare or reproductive management.

4.3. Research Learnings and Future Opportunities

For GNSS and similar technology for real-time alpaca behaviour tracking to be implemented on-farm, key information around deviations from normal behaviour is required to alert a producer of the need for potential intervention [10]. The moulding issue impacting the water tightness of the tags resulted in false alerts due to the tags’ failure to send data to the gateway. The false alerts, alongside poor weather conditions in autumn, resulting in localised flooding, prevented visual observations of the animals, limiting this study’s ability to validate the accuracy of the low activity alerts. This was resolved with replacement tags. Future research and technology development in alpacas would benefit from focusing on the accuracy of detecting specific behaviours, such as low activity and inactivity, to alert producers to an issue. The results of this study have showcased that real-time tracking tags are a valuable tool for assessing the natural behaviour of alpacas in a free grazing environment as the tags supported longitudinal monitoring across 24-h periods, which has not previously been reported for alpacas in Australia. The potential benefits for day-to-day alpaca care and management include the capacity to remotely monitor herd and individual animal health and welfare, facilitating immediate remediation when required.

5. Conclusions

GNSS technology has previously been limited in monitoring alpaca behaviour. The Smart Paddock Bluebell tag highlighted the capacity for GNSS remote monitoring technology to be used as a research tool to investigate alpaca behaviour under commercial grazing conditions. The remote-monitoring technology resulted in no negative impacts on body condition score and possesses the ability to enable producers to monitor alpacas for illness or injury through absence of normal movement when visual observation is not possible. The results of this study showcase the potential for GNSS remote monitoring technology to be utilised in a practical, extensive production system for alpacas raised in Australia.