Using Lightweight Unmanned Aerial Vehicles to Monitor Tropical Forest Recovery

Background

This study examines the ability of unmanned aerial vehicles (UAVs) to accurately monitor and measure secondary forest growth in southern Costa Rica. These UAVs may help to fill an important gap in our ability to measure and monitor the success of restoration projects in terms of measuring structural complexity and community composition in remote or mountainous areas where field-based monitoring is otherwise difficult. Using the 'Ecosynth' method, the UAVs simply combine aerial photographs from digital cameras with a computer software to create 3D models of vegetation. This UAV technology can be priced at $300-a few thousand dollars and is a more cost-effective alternative to LiDAR, which can be $20,000 for a single flight.

Research Goals & Methods

In particular, this study compared the accuracy of UAV measurements/predictions to field-based measurements of canopy height, above ground biomass, canopy roughness, canopy openness, and frugivorous birds. Field-based measurements were conducted on 13 sites, each with three plots of a different restoration treatment (passive, active plantation, and intermediate "island" strategies).

Conclusions & Takeaways

Overall the Ecosynth canopy height was a strong predictor of field canopy height, although accuracy decreased with low canopy height associated with the passive restoration plots. Above-ground biomass was able to be estimated with Ecosynth, suggesting the potential importance of this methodology in being incorporated into REDD+ programs without adding large expenses. Parameters such as canopy openness and canopy height were successful in predicting frugivorous bird presence at an accuracy similar to those of the field-based measurements. While UAV and Ecosynth methodology is not a replacement for field-based monitoring, it can help researchers and landowners to monitor the progress of restoration projects at the local level, especially where field-based monitoring is difficult.