Spatial behavior of livestock in extensive systems lies at the nexus of space use, foraging decisions, animal growth performance, and eventual enteric CH₄ emissions (1). Management of the spatial distribution of cattle through physical or electric fencing is one of the most widespread practices in extensive rangelands, and the influence such management has on stock density and forage allocation is well known to influence animal growth performance (2–4). Over the last two decades, there has been a rapid growth in the development of precision livestock farming technologies, including on-animal or wearable sensors for grazing livestock. Most wearable sensor systems are passive, providing data to livestock managers that includes diagnostics for health, nutritional, or reproductive states or events, such as estrus and calving (5, 6). Animal location is often a crucial element of grazing systems, with data from Geographical Navigation Satellite Systems (GNSS, frequently referred to as GPS [Global Positioning System]) allowing the creation of map-based visualizations (7).

Virtual fencing (VF) systems represent a significant advancement over passive sensors, offering active livestock management and precision livestock farming technology. VF systems utilize two-way digital communication between livestock managers and each animal via its wearable device, referred to as a VF collar. Managers can then actively, but remotely, manage desired grazing areas for the herd, opening or restricting movement in near real-time.

Experimentally-informed knowledge on the behavior and welfare of beef cattle under VF-based management in extensive rangeland systems is needed, as this precision livestock management technology is experiencing rapid commercial uptake (8, 9). Yet, most evaluations of VF efficacy and effects on animal behavior or performance pertain to short durations in small, intensively managed systems over ≤60-d periods [e.g., 0.2–10 ha in size; (10–12)] or to more extensive grazing systems over ≤40-d periods [e.g., 2.1–414 ha; (9, 13)].

Using VF-based management in extensive grazing systems, such as semiarid rangelands, potentially enables managers to target higher forage quality, thereby enhancing the likelihood of achieving production goals (14–16). With conventional boundary systems (i.e., physically fenced pastures) on large rangeland parcels, it is often impractical to focus grazing, even if it is deemed useful for both farming and nature conservation objectives (17). Recent literature suggests that VF can overcome many of these issues in extensive systems, where opportunities exist to easily and flexibly manage grazing at a finer spatial scale, keeping livestock well away from risk areas (which may be seasonal or temporary) using virtual exclosures or focusing herds onto specified grazing areas (17). For instance, an initial study that deployed VF collars on beef cattle to curb utilization of burned sagebrush steppe demonstrated the efficacy of this technology in modulating post-wildfire herd distribution for land restoration purposes (13). Yet, little is known about how the use of this emerging technology on extensive rangeland can affect spatial behavior, growth performance, and enteric CH₄ emissions, the latter of which is one of the most challenging agricultural products to manage but also with the most potential to reduce the overall carbon footprint of the beef sector (18).

At the same time, prior studies of rotational grazing management in extensive systems via physical fences show that increasing stock density (number of animals per unit area at any given point in time) alters foraging behavior and reduces diet quality, which results in reduced weight gain (4, 19). Use of virtual fencing to subdivide pastures and rotate cattle sequentially through the subdivisions will increase stock density, with the magnitude of the increase depending on the size of subdivision areas. It remains unclear whether this effect of stock density on weight gain can be offset by implementing VF rotations in a manner that targets different plant communities when they are most palatable. In standard extensive grazing systems on western US rangelands, the current high costs of VF also necessitate clear economic benefits from combinations of hardware (e.g., wire fences) and labor savings or production benefits (20, 21). Yet, evidence for production benefits on extensive rangeland remains limited.

Researchers have also theorized that incorporating VF management could reduce enteric CH₄ emissions by enhancing animal growth performance without detrimental effects on welfare (15). As reviewed by Vargas, Ungerfeld (22), research has shown that ruminants on diets with higher fiber concentrations result in greater enteric CH₄ production (g/d) due to more favorable conditions for rumen methanogenesis. Thus, by coupling knowledge of how different forage types affect enteric emissions with grazing plans (23, 24), managers could potentially reduce enteric CH₄ emissions of beef cattle grazing extensive rangelands through adopting VF technology.

Assessing enteric methane (CH₄) production in extensive beef cattle production systems is one of the most challenging aspects of developing innovative, emissions-reducing production practices across the beef sector (18, 25). The cow-calf and stocker stages, where growing cattle graze forage before finishing in confinement, have been identified as the portion of the beef cattle life cycle when 89% of enteric CH₄ emissions are produced (26). Compared to confined settings, this production stage is the most difficult for measuring and managing CH₄ emissions (14, 27). Because enteric CH₄ is produced during the anaerobic fermentation of organic matter in the rumen and represents an energy loss for the animal that varies between 2 and 12% of the gross energy intake (28, 29), spatial management practices that influence animal distribution and access to forage of varying quality have the potential to impact production efficiency in extensive grazing systems.

In confined settings, the ability to manage intake and monitor growth performance in near real-time has allowed managers and researchers to develop alternative diets and interventions that lower CH₄ synthesis of fed-animal production systems (18, 27, 30) and the overall footprint of the beef cattle supply chain (29, 31, 32). Yet, the capacity to manage foraging decisions, spatial behavior, and concomitant growth performance in grazing systems is limited mainly to small, pasture-based, intensively managed grazing systems, where producers primarily manage consumption of the most digestible, nutrient-dense plant parts through physical or electric fencing (33, 34). For example, tannin-containing birdsfoot trefoil (Lotus corniculatus L., variety Langille) and small burnet (Sanguisorba minor Scop., variety Delar) have been shown to reduce enteric CH₄ emissions and urinary nitrogen (N) excretion in intensive grazing systems (34). Additionally, targeted grazing of timothy (Phleum L.) or Italian ryegrass (Lolium multiflorum L.) at an early vegetative state has been shown to be a successful method for reducing enteric CH₄ emissions in intensive grazing systems (35, 36). In contrast, practices for reducing the emissions of free-ranging livestock in extensive production environments remain limited (22, 27, 37–39), which is unfortunate as 78% (166 M ha) of the 214 M ha of US grazing land are extensive production systems (40). Furthermore, a recent review of feeding strategies to mediate enteric emissions in grassland systems by Vargas, Ungerfeld (22) found targeted-grazing research evaluating enteric CH₄ emissions in extensive grazing systems was absent from the literature. Thus, there is a growing need to identify where and which precision livestock management technologies can be used to increase production and reduce emissions, partly because enteric CH₄ emissions vary spatially and by production context (18, 26, 32).

Unlike controlled production settings, such as intensive-pasture management and confined animal feeding operations, measurements of enteric CH₄ emissions on extensive grazing systems have been relatively difficult to obtain. Before the availability of automated head-chamber systems (AHCS, i.e., GreenFeed, C-Lock, Inc., Rapid City, SD), techniques including the sulfur hexafluoride (SF₆) marker dilution method had been the primary means of collecting enteric emissions data on pasture (36, 41). At the same time, whole animal respiration chambers and head chambers served as the gold standard across research settings (42, 43). These tools enabled enteric CH₄ emissions measurements, yet their ability to collect gas flux data that reflects working production systems was limited in comparison to AHCS units (41), which are deployed in both confined and open grazing environments for long-duration measurement periods [i.e., weeks to months; (44, 45)]. As opposed to previous methods, which collect animal gas flux for single-day to three-day increments without the need for handling animals, AHCS measurement periods span as long as electrical power persists and bait remains available to attract individual animals to the automated head-chamber.

The same less-invasive element that allows AHCS technology to reflect working production operations, voluntary intake of bait by unconstrained livestock, also represents a limitation, especially in grazing systems. By having free choice to graze forage and consume AHCS bait on pasture, cattle visit AHCS units at a much lower daily rate in grazing systems [e.g., 3–4.5 visits/d; (45, 46)] than observed in confined settings [(44, 45), e.g., 3–4.5 visits/d; (46), e.g., 1–2 visits/d; (47, 48)]. Moreover, the percentage of individual cattle in a herd that will routinely use the AHCS varies widely, irrespective of production setting. For example, Alemu, Shreck (45) reported 48–84% of cattle acclimated to AHCS in confinement, and acclimation rates were equally variable for intensive-managed [e.g., 34–70%; (44, 47–49)] and extensive-managed grazing systems [e.g., 46–62%; (50, 51)]. To date, the variability in routine use of AHCS units across production systems has remained a methodological issue for researchers without a firm understanding of pertinent AHCS use drivers (41, 44). However, recent efforts have identified overall behavior, including feeding behavior and between-animal variation, as key animal-oriented factors driving this variability in grazing systems (25). Nonetheless, AHCS technology remains a flexible method to collect enteric gas flux data across ruminant livestock production settings.

Here, we investigate the influence of VF on spatial behavior, growth performance, and enteric CH₄ emissions of growing steers in an extensive livestock production system. We first evaluate the efficacy of the VF system in manipulating cattle distribution into desired sub-pastures. We then examine consequences for cattle movements on a daily basis within each rotation interval, for CH₄ emissions at the level of rotation intervals and the full grazing season, and finally for steer weight gain and CH₄ emissions intensity (g CH₄ emitted per kg beef produced) over the full grazing season.

We evaluated the efficacy of implementing rotational grazing management via VF technology in terms of its influence on steer spatial behavior and the consequences for both animal growth and CH₄ emissions. Overall, the VF system was successful in manipulating steer distribution as planned in the within-pasture rotational treatments. In one block where five of 30 steers often moved through the virtual fence boundaries on a daily basis, the herd still collectively spent 94–96% of their time within the desired grazing areas. In the second block, excursions through the VF boundary were rare, and collectively the herd spent more than 98% of their time within the desired grazing areas. The successful implementation of the grazing rotation via VF was reflected in patterns of forage biomass accumulation, with rested portions of the VF-managed pastures increasing in forage biomass much more rapidly than the full paddock being grazed following rains in the second half of the growing season (Figures 1A,B). One key result was that VF-management consistently reduced steer weight gain across both pasture types (by 12% for ACHS-acclimated steers and 9% for all steers in the experiment). For the grazing season, this translated into a 10.2 kg (22.4 lb) lower weight than control steers at the end of the grazing season. These findings are consistent with prior rotational grazing experiments at the study site which employed larger herds rather than subdividing pastures (4, 19). The increase in stock density of the VF-managed steers may have induced less selective grazing, as the sub-pastures were more homogenous in plant community composition than the controls.

In terms of the consequences for enteric CH₄ emissions, VF-managed steers in the diverse mosaic pasture-pair produced more CH₄ when grazing C₄ grass-dominated communities during the latter two-thirds of the grazing season and produced less CH₄ than controls while grazing a cool-season dominated sub-pasture early in the season. Averaged over the entire season, no differences were observed in CH₄ emissions, but VF-managed steers emitted 33% more CH₄ per kg of weight gain than controls. In the Sandy Plains pair of pastures, VF-managed steers produced less CH₄ than control steers in all four rotations, all involving grazing of cool-season dominated sub-pastures. However, this outcome occurred at the cost of a 12% decrease in kg gained per day, resulting in no differences in CH₄ emission intensity between VF versus control steers over the whole season. The variability in the direction and magnitude of differences in CH₄ emission rates across pastures and plant communities suggests that concentrating grazing on cool-season forage may be more likely to reduce enteric emissions, while the opposite is true for targeted grazing of warm-season forage. At the same time, a reduction in weight gain associated with targeted grazing can result in an overall increase in CH₄ emission intensity, which is ultimately the key metric for sustainable intensification of food production.

Our findings are viewed from the subset of animals that routinely used the AHCS across all pastures; therefore, conclusions about study treatments and enteric emissions from this unique empirical study must be drawn with caution. Nonetheless, our coupling of animal GPS tracking and AHCS data indicated that differences in spatial behavior were consistent between VF-managed and control steers irrespective of their AHCS-acclimation status. This finding supports the view that the subset of steers that were acclimated to the AHCS in each treatment were representative of the wider herd population. Such fusion of animal performance and enteric emissions information with movement behavior enhances our confidence in estimates of enteric CH₄ emission in extensive rangeland settings.

Additionally, we note that all analyses were conducted at the individual steer level, such that inferences are restricted to these specific pastures and weather conditions, and additional replication of VF treatments at the pasture level and seasons of study are needed to test the broader implications for ADG and CH₄ emissions in semiarid rangelands. However, issues related to spatial behavior under distinct types of rangeland management have not been considered before in enteric CH₄ studies. Thus, our experiment illustrates that multiple factors have potential impact on the management of the forage resource-growth performance-enteric emissions relationship and can contribute to reducing uncertainty around the development of sustainable grazing practices for extensive livestock farming.

Steers in this study were obtained from three different previous production environments (PPE). Interestingly, steers from these different sources differed substantially in their growth performance on shortgrass rangeland, with steers that spent the previous winter grazing shortgrass achieving significantly greater weight gains than steers backgrounded in drylots. Further, steers of different origins differed to only a limited degree in CH₄ emissions, such that the steers backgrounded on rangeland had significantly lower CH₄ emissions per kg of beef produced compared to steers from drylot origin. These three different groups of steers also differ in terms of their genetics; thus, we cannot determine whether background, genetics, or other factors are driving the differences in ADG. However, we can conclude that factors other than grazing management during the growing season can have large effects on ADG and CH₄ emission intensity, thereby muting the effects of grazing management.

Due to the greater quantity and quality of forage resources, shortgrass steppe pastures associated with the Sandy Plains ecological site (76), were expected to support enhanced steer growth performance (77) and concomitantly reduced enteric CH₄ emissions when herd distribution focused on available cool-season grass swards. This expectation was largely met, as pooled weight gains were 24% greater on the Sandy Plains pastures compared to the diverse mosaic pastures. However, differences in CH₄ emissions between the pasture types depended on spatial management, with both pasture types having the same CH₄ emission rates in the absence of VF management but 25% less emissions on the VF-managed Sandy Plains compared to the VF-managed diverse mosaic. Greater biomass in the diverse mosaic pair was primarily due to residual, standing dead forage carried over from the previous year (Figure 1), which may in tandem with increased stock density (4), have prevented selective foraging when steers were concentrated into the smaller VF pastures, particularly in the second rotation into the Salt Flat ecological site, where standing dead vegetation and stock density were greatest. In contrast, lower amounts of residual, standing dead biomass in the Sandy Plains pair of pastures may have facilitated selective foraging, even when the steers were concentrated into the VF sub-pastures. We note that the VF-managed sub-pastures in the Sandy Plains supported lower stock density than their diverse mosaic counterpart, which likely also played a role in mediating behavior-performance-emissions relationships that cannot be examined with our current experimental design.

Another important consideration in spatial distribution management is ensuring cattle have access to water sources (78). In the diverse mosaic block, targeted grazing of the eastern two-thirds of the pasture during the first two rotation intervals required including a long lane along the northern edge of the pasture for cattle to reach the water source. This grazing area polygon placement appears to have affected movement patterns, as these two rotation intervals had a larger difference in distance traveled per day and in area explored per day compared to the controls than the third interval (with no lane necessary) and all intervals in the Sandy Plains pasture (no lane necessary). Visual inspection of animal distribution based on GPS location data showed intensive use of the lane during the first two rotation intervals. This may have resulted in more uneven use of available forage. However, the first interval yielded a decrease and the second an increase in CH₄ emissions with VF management relative to controls. Further research is needed on how the spatial configuration of virtual sub-pastures, in relation to water and forage conditions, influences cattle distribution, performance, and welfare.

Periodic drought has been shown to decrease species richness, aboveground plant biomass, and concomitant yearling steer performance in grasslands (53, 79–81). This was also true in the current study, where precipitation and ADG were 78 and 82% of the historic average (53, 82). Under conditions of more rapid plant growth and an increase in forb plant components (77, 83), differences in animal performance across VF treatments might be less pronounced, and differences in CH₄ emissions more pronounced. Furthermore, the data presented here align with outcomes from a recent study conducted in a drought year (i.e., 33% of historical site average; 50), where AHCS routine use was 46% vs. our 48% in the current study and slightly lower CH₄ emissions, ~200 g CH₄/hd/d vs. ~214 g/hd/d reported here. Nonetheless, this study offers insights into the spatial behavior and performance of grazing animals, as well as the resulting enteric CH₄ emissions, during dry conditions typical of the western US shortgrass steppe region. These conditions are becoming increasingly common in a consistent, low productivity megadrought scenario in southwestern US rangelands (84). Future research on grazing animal phenotypic expression in extensive production systems merits further investigation, especially under more conducive non-drought grazing conditions that allow for adequate forage production.

Access to cool-season grass may not have been as critical to maintaining steer growth in the diverse mosaic pastures, but it may remain key for reducing enteric CH₄ emissions under VF management, as cool-season grasses have less complex cell structures and sugars. Greater CH₄ production per kg gained by VF-managed steers compared to control steers indicates that increased stocking density in the sub-pastures likely reduced the ability of steers to forage selectively (19). Therefore, a crucial aspect of target-grazing with VF technology is that it requires high capacity in adaptive decision-making to ensure that increased stocking density does not negatively impact selectivity and foraging patterns (4). This element of adjusting VF grazing area for further access to forage also limited capacity for selective foraging in the dry grazing season. Knowing which plant communities produce more enteric emissions at certain times of the year and avoiding grazing them during those periods can potentially reduce the operational footprint of the enterprise. The growing array of predictive grazing tools (85), including Rangeland Analysis Platform (RAP), GrassCast, and the U. S. Drought Monitor, combined with ground-based assessments of animal utilization and knowledge of plant community-specific enteric emissions characteristics, such as methanogenic potential of C₃- vs. C₄-photosynthetic pathway forage (86), can assist ranchers in making adaptive decisions that balance animal weight gains with CH₄ production.

Experimentation using precision livestock management technology, as in the current study, where VF technology, enteric emissions measurement infrastructure, and remote sensing allowed an improved understanding of forage-animal growth performance-emissions relationships, underlies the development of innovative grazing management practices. For instance, a pasture-level evaluation of animal growth performance and AHCS-measured enteric CH₄ emissions by Thompson, Maciel (87) in two standard Midwest grazing forage mixtures: simple alfalfa:orchardgrass (Dactylis glomerata L.) mixture and a complex multi-forage species mixture showed that adopting the complex mixture would not result in improved animal performance or reduced CH₄ emission intensity compared with a simple mixture for yearling steers and heifers. With this knowledge, a manager could then expect little utility of targeted-grazing management in these plant communities to impact performance and emissions.

Clear trade-offs are associated with employing VF for targeted grazing in extensive semiarid rangeland systems. Manipulating animal distribution into smaller sub-pastures increases stock density, which impacts foraging behavior, diet quality, and animal gains (4, 19, 88). Ranchers will need to employ experiential knowledge to provide sufficient amounts of high-quality forage in each sub-pasture to ensure growth performance and potentially reduce CH₄ emissions. This management acumen for ranchers is contingent upon knowledge of ecological sites, available plant diversity and phenological growth patterns, and forage quality patterns throughout the grazing season (23, 89). Other benefits of this approach in extensive rangelands include faster-growing livestock and flexible timing of the grazing season terminus, which can increase economic returns (90, 91) and potentially reduce herd-level CH₄ emissions intensity during their time on rangeland.

The costs of VF implementation and its advantage relative to conventional fencing will vary depending on both the specific application or technology adopted, as well as factors such as topography, cellular coverage, and pasture size (20). Furthermore, as noted in this study, weather impacts on forage quantity and quality, which affect livestock performance related to spatial management of the herd through VF, can have an impact on overall management costs, especially in terms of foregone revenue associated with reduced animal gains. Programs designed to incentivize VF adoption for the purpose of reducing enteric CH₄ emissions in grazing livestock should address these practical tradeoffs.

We executed an experiment evaluating the efficacy of VF in implementing rotational grazing management during the growing season in semiarid shortgrass rangeland and quantified the consequences for both cattle weight gain and CH₄ production. Despite the finding that 16% of the steers in one of the experimental pastures made daily excursions outside the desired grazing areas defined by the virtual fence boundary, the system was still effective in focusing 94–99% of the herd’s locations into desired grazing areas. Notably, VF management did not improve animal performance. Rather, we found consistently lower steer weight gains with VF management across both pairs of experimental pastures. It remains unclear whether this effect arose as a result of VF management on the degree of selective foraging by the cattle, increased energy expenditure, the specific types of plant communities onto which we chose to focus grazing, or some combination of all three factors. Across pastures, overall enteric CH₄ emissions (mean ± SE) was 214 g CH4/hd/d ± 9 and averaged among VF treatment levels VF-managed steers emitted 216 g CH₄/hd/d ± 5 vs. 221 g CH₄/hd/d ± 6 for non-VF-managed steers. Additionally, we found that AHCS subsets of individuals are fair representations of their broader herd in relation to growth performance and spatial behavior. Finally, our results provide no consistent evidence that VF management will simplistically reduce CH₄ emissions. In the diverse mosaic pair, VF management reduced emissions during the first rotation interval, when VF steers were concentrated on the C₃-dominated plant community, but increased emissions in the second and third interval when VF steers were concentrated on C₄-dominated areas. Additionally, due to the spatial arrangement of VF polygons and access to water, we could not fully compare responses to our control animals, which further demonstrates the difficulty of examining forage resource-growth performance-enteric emissions relationship in working livestock production landscapes. In the Sandy Plains pasture pair, VF management did reduce CH₄ emissions. Yet, the effect of VF management on steer weight gain was larger than the effect on enteric CH₄ emissions; therefore, VF management did not reduce CH₄ emissions per kg of beef produced. Although it has been postulated that VF is a potential tool to reduce CH₄ emissions, our findings show this is not a straightforward outcome, and that greater attention needs to be given to the effect of VF on cattle weight gain and the mechanisms underlying this outcome.

When VF management focused on grazing cool-season grass-dominated areas in the Sandy plains ecological site-associated pasture, enteric CH₄ emissions were reduced, but animal performance declined by 12% compared to non-VF-managed steers. Conversely, in the diverse mosaic-associated pastures, VF management did not impact growth performance but did increase CH₄ emission intensity. When VF-managed steers grazed in the warm-season grass-dominant sub-pastures, they likely did not have sufficient access to cool-season grasses to reduce CH₄ synthesis. Drought-induced reductions in forage quantity and quality impacted the effectiveness of targeted grazing sub-pastures to enhance animal performance while concurrently decreasing enteric CH₄ emissions. To improve precision grazing strategies, further research should incorporate measures of diet quality to understand the mechanisms driving relationships between animal growth performance and enteric CH₄ emissions across varying weather and climatic conditions.