Multisensory Monitoring Using UAVs to Estimate Forage Height and Biomass in Andean Silvopastoral Systems Associated with Alnus acuminata.

The digital transformation of high-Andean livestock farming requires precise, efficient, and scalable tools to quantify structural vegetation variables and monitor forage productivity in silvopastoral systems. In this context, the present study evaluated the potential of the combined use of unmanned aerial vehicles (UAVs) equipped with RGB, multispectral, and LiDAR sensors to estimate vegetation height and forage biomass in silvopastoral systems associated with Alnus acuminata in the livestock basin of Molinopampa, Amazonas, Peru.

The research integrated remote sensing, photogrammetry, geospatial analysis, and artificial intelligence technologies to characterize the spatial and temporal dynamics of forage under different silvopastoral conditions, cutting frequencies, and distances from the tree component. For this purpose, georeferenced plots were established, where productive and structural variables were recorded using conventional field methodologies, complemented by UAV flights for the generation of RGB orthomosaics, digital surface models, LiDAR point clouds, and vegetation indices, including NDVI, NDRE, SAVI, among others.

The results showed significant spatial variation in forage biomass, with estimated values ranging from 1.1 to 1.6 t DM ha⁻¹ per cutting cycle, and pasture heights ranging from 12 to 25 cm. The greatest biomass accumulation was observed at intermediate distances from Alnus acuminata, where the moderate influence of the tree component favored pasture performance. Predictive models integrating LiDAR-derived structural metrics and multispectral vegetation indices achieved an R² above 0.75 and an RMSE below 0.15, with machine learning algorithms and linear regression models standing out for their higher predictive accuracy.

In conclusion, the integration of UAVs, remote sensing, geospatial analysis, and machine learning constitutes a robust and replicable tool for optimizing the monitoring of agronomic parameters, reducing evaluation costs and time, and strengthening the sustainable management of high-Andean silvopastoral systems under climate change scenarios.