Breakthrough Research Integrates UAV Photography and Trajectory Tracking for Precise Grazing Assessment

A new study demonstrates how unmanned aerial vehicle aerial photography combined with trajectory tracking technology can provide detailed evaluations of relative grazing intensity in desert steppe household pastures. The research focuses on a specific case in Inner Mongolia, China, highlighting non-uniform livestock utilization patterns that traditional methods often overlook. This integrated framework offers herders and land managers a less labor-intensive way to monitor grassland use and support sustainable practices.

The full paper, titled "Evaluation of relative grazing intensity based on UAV aerial photography and trajectory tracking technology—a case study of a desert steppe household pasture," appears in the journal Computers and Electronics in Agriculture. It credits authors Baoping Meng, Yu Wang, Lei Dong, Jinrong Li, Shuhua Yi, Yanyan Lv, Shuixia Zhao, and Jian Wang. Readers can access the abstract and details through the ScienceDirect platform at https://www.sciencedirect.com/science/article/abs/pii/S016816992600685X.

Context of Desert Steppe Pastures in Inner Mongolia

Desert steppes represent a critical ecosystem in arid and semi-arid regions, characterized by low precipitation, sparse vegetation, and vulnerability to degradation from overgrazing. In Inner Mongolia, these landscapes support household-based pastoralism where families manage livestock on fenced or semi-fenced pastures. The study area covers approximately 3.87 square kilometers in the north of Darhan Mumingan Union Banner, falling within a mid-temperature semi-arid monsoon climate zone with an annual average temperature of 3.4 degrees Celsius. Precipitation in recent years has been below long-term averages, adding pressure on forage resources.

Household pastures in such regions often experience uneven livestock distribution, leading to localized overutilization near water sources, enclosures, or preferred grazing spots. Understanding these patterns helps maintain grassland productivity and prevents desertification. The research emphasizes how cattle herds show activity densities up to 16 times higher in some zones compared to others within the same pasture.

Defining Relative Grazing Intensity and Its Importance

Relative grazing intensity, often abbreviated as RGI, measures the quantity of grazing livestock per unit area over a defined time period. It serves as a key indicator of how intensively grassland resources are used. Moderate levels can promote plant diversity and soil health through natural nutrient cycling from manure, while excessive intensity causes vegetation decline, soil compaction, and reduced biodiversity.

In pastoral communities, balancing livestock numbers with available forage is essential for long-term economic viability and ecological stability. The case study illustrates how RGI varies spatially and temporally, with cumulative values from May to July showing a gradient decreasing from northwest and northeast edges toward the central pasture areas.

Limitations of Traditional Monitoring Approaches

Conventional techniques for assessing RGI rely on field observations, household surveys, recording of grazing trails, dung distribution mapping, or direct visual counts of animals. These methods demand significant labor and time, limiting the scale and frequency of data collection. Livestock populations fluctuate due to births, sales, and growth, requiring repeated surveys for accuracy. Additionally, pinpointing exact locations of herds across large areas proves challenging without advanced tools.

Satellite remote sensing has been explored in some grassland studies to infer intensity from vegetation changes, yet it often lacks the fine-scale resolution needed for household pasture heterogeneity. The new research addresses these gaps by combining high-resolution aerial data with movement tracking.

Photo by Charles MingZ on Unsplash

UAV Aerial Photography for Livestock Population and Distribution

Unmanned aerial vehicles, commonly known as UAVs or drones, capture images at centimeter-level spatial resolution, enabling clear identification and counting of cattle. In the study, UAV surveys achieved strong correlation with ground observations, yielding an R-squared value of 0.83 and a root mean square error of 22.56 sheep units per hectare. This accuracy supports reliable population estimates while revealing spatial activity regularities.

The distribution of cattle herds tends to increase initially with distance from the livestock enclosure before decreasing farther out, a pattern confirmed statistically with a p-value less than 0.001. Such insights help map preferred grazing zones and identify underutilized areas within the pasture.

Trajectory Tracking Technology for Time-Series Herd Data

Trajectory tracking, typically using GPS collars on livestock, records geographic positions at regular intervals to build detailed movement histories. This technology captures diurnal rhythms, habitat preferences, and long-term activity patterns that single aerial snapshots cannot provide. In the Inner Mongolia pasture, collar-based tracking revealed concentrated trajectory points in western, southern, and southwestern edges, with densities exceeding 320 sheep units per hectare in high-activity zones covering about 7.5 percent of the total area.

While effective for temporal data, trajectory methods alone may struggle with precise herd density mapping across entire pastures. Integration with UAV imagery compensates for this by adding spatial context and validation.

Integrated Framework and Study Methodology

The exploratory case study merged UAV photography for spatial distribution and population counts with trajectory tracking for temporal movement data. Researchers conducted surveys over the growing season, processing aerial images to count animals and overlay trajectory points on pasture maps divided into grids. This hybrid approach quantified non-uniform utilization and generated high-resolution RGI maps.

Validation against traditional ground-based methods showed high consistency, with R-squared values ranging from 0.927 to 0.984. The framework proved more efficient, requiring fewer personnel hours while delivering finer detail on heterogeneity caused by livestock behavior.

Key Results on Spatiotemporal RGI Variation

Findings confirmed pronounced spatial heterogeneity in livestock activity. Trajectory point densities varied dramatically across the pasture, underscoring that uniform division of total livestock by area underestimates localized pressures. UAV data further illustrated how herd ranges expand and contract relative to enclosure proximity.

Temporally, RGI accumulation from May through July displayed a clear directional trend, with higher values near boundaries tapering centrally. These patterns inform targeted management, such as rotational grazing or supplemental feeding in low-use zones to promote even utilization.

Photo by Charles MingZ on Unsplash

Benefits for Sustainable Grassland Management

The UAV and trajectory method reduces labor demands compared to exhaustive field surveys, making regular monitoring feasible for household operations. It captures fine-scale variations that support precision interventions, potentially improving forage regrowth, soil conditions, and overall pasture resilience in water-limited desert steppes.

Broader adoption could aid policy efforts addressing grassland degradation in Inner Mongolia and similar regions worldwide. By revealing actual utilization patterns rather than averages, managers can adjust stocking rates dynamically and preserve ecosystem services like carbon sequestration and biodiversity support.

Additional perspectives from related grassland research, such as studies on overgrazing causes available at https://ecologyandsociety.org/vol27/iss1/art8/, reinforce the value of technology-enhanced monitoring for herder decision-making.

Future Outlook and Potential Applications

This case study serves as a foundation for scaling the integrated technology to larger pastures or different grassland types. Advances in UAV battery life, image processing algorithms, and affordable tracking devices could further enhance accessibility. Future work might incorporate hyperspectral sensors or machine learning for automated species identification and biomass estimation alongside RGI mapping.

Collaboration between researchers, local herding communities, and agricultural extension services will be key to translating these tools into practical guidelines. The approach holds promise for global arid land management, contributing to food security and climate adaptation in pastoral systems.