Research

Background

Cattle exhibit complex behaviour when grazing rangelands that includes preferential selection across the landscape (e.g., hills vs. valleys, near water sources, etc.) and plant species. Several factors influence where and what cattle graze, including nutritional needs, their sense of smell, learned behaviours, and perhaps even genetics. While previous work has examined grazing selection and preference in certain environments, few studies have attempted to correlate grazing behaviours and selectivity with both genetics and performance data.

In addition, there is growing interest in the use of technology such as GPS eartags to support traditional grazing management practices such as cross fencing, water and mineral placement, along with virtual fencing, to ensure that grazing utilization and animal performance is optimized in diverse pasture environments.

Objectives

The objectives of this study are to:

• Evaluate virtual fencing for controlling the spatial distribution of cattle and associated forage use while grazing
• Quantify cattle performance metrics during grazing
• Test different GPS-based technologies for use in pasture systems
• Conduct high resolution mapping of grazing lands using drones
• Evaluate cattle grazing use patterns and associations with animal performance
• Quantify differences in animal activity on pasture
• Assess changes in cattle activity and grazing patterns in response to weather extremes
• Construct and evaluate DNA barcoding for plant species identification
• Characterize cattle diets using fecal DNA barcoding
• Evaluate enteric methane production of cattle while grazing

What They Did

The team deployed Nofence virtual fence (VF) collars on livestock across five complementary field trials from 2022 to 2024. Virtual fencing controls animal distribution by setting GPS-based boundaries via an app. When an animal approaches a boundary, the collar emits an escalating audio cue (siren). If the animal does not turn back, it receives a brief electrical pulse. Cattle were systematically trained to associate audio warnings with the physical pulse to learn the boundaries before trial initiation.

In addition to virtual fencing, researchers monitored three distinct cattle populations over three years (2021–2023): 1) a youthful Kinsella Composite (KC) herd (20 heifers and 40, 3-yr-old cows with calves) in each of three years; 2) a multi-aged purebred Angus herd (ages 3–11; 182 hd); and 3) a multi-aged KC crossbred herd (133 hd). They measured average daily gain (ADG), calf growth, and 205-day adjusted weaning weights. Activity budgets (lying time, step counts, posture transitions) were tracked 24/7 on a subset of animals using IceQube+ pedometers fit to the cows’ legs, and habitat use was tracked with neck-mounted LOTEK GPS collars. Diet selection was analyzed by building a DNA reference library of 167 plant species and using fecal DNA barcoding to identify the plant species consumed. The DNA barcoding method was calibrated in drylot trials using heifers fed known rations. Lastly, enteric greenhouse gas emissions (methane and carbon dioxide) were measured continuously on pasture using GreenFeed stations positioned near water sources, and rumen fluid samples were collected to analyze volatile fatty acids (VFAs) and rumen microbial communities.

What They Learned

Across multiple years and seasons, cattle successfully learned to respect the virtual boundaries while rotational grazing, both as heifers and cows with uncollared calves, with containment exceeding 99.5% of the time. About half the animals never escaped. When brief escapes occurred (averaging 137 minutes), cattle voluntarily returned to the herd, allowing the collars to resume normal function. Most importantly, virtual fence interactions had no measurable impact on cattle weight gains or conception rates.

Cattle fell into three clear behavioral groups: low stimuli (rapidly trained, required few cues from the collar), moderate stimuli (willingly grazed near and interacted with the boundary, responsive to audio warnings), and high stimuli (13–14% of the herd – these frequently tested the boundary and received a disproportionate 26–33% of the shocks). These learning patterns persisted in the same animals as heifers and cows. Interactions with the virtual boundary increased with forage depletion, but did not lead to increased escapes. In addition to being effective for rotational grazing, VF collars could be used to exclude grazing from targeted riparian/wetland areas, with cattle rapidly occupying these areas once access was provided.

Out of over 80 units deployed, only one collar completely failed, and the NoFence collars maintained a 90% battery charge even in a winter that had several days of -30°C temperatures due to robust solar charging capability.

High-resolution drone LiDAR and multispectral mapping demonstrated a 92.9% habitat classification accuracy, and was used for assessing cattle habitat use on native Parkland range. Under summer conditions cattle distinctly preferred upland grasslands and lowlands due to abundant and accessible forage and gradually switched to open shrublands as grazing progressed. Forests and closed shrublands were typically avoided, except during heatwaves (Temperature-Humidity Index ≥ 68, temperatures >30°C) and periods of increased cold (Wind Chill Index < -7°C), when cattle prioritized shelter seeking behavior by shifting use from open habitats to aspen forest and closed shrublands. Heat stress also increased overall animal movement rates, including shifting daily activity patterns from normal foraging to active night-time grazing and brief midday travel to access water, with extended rest periods in-between.

Cow-calf aggregate weights (sum of cow weight and calf weight) peaked in cows aged 4 to 9. However, purebred Angus cows aged 10 and 11 tended to trade-off their own body condition and maintenance over milk production, leading to a significant drop in calf weaning weights. Crossbred cows with greater heterosis (hybrid vigour) had calves with greater ADG in the fall.

Young cows with high drylot feed efficiency (low residual feed intake (RFI)) showed better weight gains on limited fall pasture, but there was no correlation between drylot RFI and enteric methane emissions on pasture. Furthermore, methane emissions on pasture were highly variable and had low repeatability, meaning individual animals dramatically re-ranked their emission rates between drylot and pasture environments.

Fecal DNA barcoding was able to track dietary additions and removals within two days. However, fecal read counts were biased with woody/shrub species overrepresented, and legumes and sedges underestimated due to high digestibility. Free-ranging diets of cows tended to contain more plant species than that of heifers, regardless of season of grazing, and cows with greater British breed composition had greater overall dietary diversity during late fall when foraging conditions rapidly degraded.

What it Means

Virtual fencing is a promising way to reduce labour and infrastructure costs while optimizing pasture utilization. However, the collars used in this project are not commercially available in Canada yet. While other systems do exist, producers will need to weigh the network connection used, capital investments such as the collars themselves or associated base stations, contract agreements, and overall cost:benefit for their operation and use cases.

This project demonstrated several impacts of cow age and breed composition on overall productivity. Production of the cow/calf unit on pasture did not peak until cows were 4 years of age, suggesting a potential benefit to retaining cattle until after this age; while not surprisingly, cows over 9 years of age either started to wean lighter calves or sacrificed their own weight gain. Cows with more hybrid vigour also weaned heavier calves, with the greatest benefit arising due to an ability of crossbred cows to better support fall calf growth.

Landscape matters. Cattle preferred to graze upland grasslands and wetlands, especially earlier in the season. Brush or forest encroachment will limit grazing opportunities for cattle, but some wooded landscape features may be necessary to allow cattle to cope with periods of weather stress.