Computational Models Ecosystems

Quantification of pasture ground cover and botanical composition for sustainable grazing management

A combination of high resolution multi-spectral imagery and supervised learning effectively distinguishes between pasture species and seasonal changes in botanical composition.

Seasonal pasture monitoring can increase the efficiency of pasture utilization for livestock grazing. Effective monitoring and management will help ensure long-term sustainability.  

Manual monitoring of pasture over large areas is often infeasible due to time and financial constraints. To date, remote sensing and Geographic Information System (GIS) approaches do not have the temporal (sub-weekly) or spatial (< 20 m) resolution to facilitate grazing management decision making at the paddock level.

A new study conducted by Dr. Iffat Ara, a Research Fellow in Spatial Science at the Tasmanian Institute of Agriculture, University of Tasmania, colleagues from University of Tasmania and AgroInsider used high-resolution multispectral data from ESA’s satellite, Sentinel-2, to monitor pasture cover and botanical composition over an extended duration (seasonal) at paddock/field scales as a case study in Triabunna, Tasmania, Australia.

“This satellite image is freely available with high spatial (10 m), temporal (nearly weekly) and multispectral resolution to monitor pasture at paddock scale” says Ara.

With data in hand, the authors set out to determine how it could be used to model seasonal changes in pasture growth and composition at the paddock scale (i.e. 50-100 ha). First, the authors measured ground cover and botanical composition over a twelve-month period from August 2019 to July 2020. Then, they processed the satellite data, converting reflectance data to vegetation indices based on greenness, namely normalized difference vegetation index (NDVI). Supervised classification was then used to map key pasture species classes (native, improved, and mixed) and observe how such classes change seasonally.

The authors demonstrated that 10 m multispectral imagery can be reliably used to distinguish between pasture species as well as seasonal changes in botanical composition. Across seasons and paddocks, the approach predicted pasture classes with 75-81% accuracy.

They were also able to identify changes in botanical composition seasonally in response to biophysical factors (primarily climate) and grazing behaviour, with seasonal highs in spring and troughs in autumn. Ara explained “interestingly, we found that within-paddock variability is more useful for management than paddock average cover. This is because the extent and duration of very low ground cover puts the paddock/field at risk of adverse grazing outcomes, such as soil erosion and loss of pasture biomass, soil carbon and biodiversity.”

These results indicate that satellite imagery can be used to support more timely grazing management decisions for the benefit of pasture production and the improvement of environmental sustainability.

Source of funding: This project is funded by the Smart Farm Grant, the Australian National Landcare Program, Department of Agriculture and Water Resources, Government of Australia.