Bark beetle outbreaks have devastated spruce forests across Central Europe in recent years, driven by hotter, drier conditions and stressed tree stands. Once visible symptoms appear, it is often too late for effective intervention. But what if we could detect infestations earlier, before discoloration, defoliation, or death?
A new peer-reviewed study from Europe-LAND researcher Dr. Lucie Kupkova and her team explores exactly this question. Published in Ecological Indicators, the article examines whether hyperspectral imagery from unmanned aerial vehicles (UAVs), often called drones, can detect signs of bark beetle (Ips typographus) infestation during the “green stage”, when trees still appear healthy to the naked eye.
The Study
Over the 2022 growing season, the research team conducted nine UAV flights over a spruce monoculture in Krkonoše National Park (Czechia), capturing hyperspectral data at 3 cm resolution. They closely monitored 23 infested and 23 healthy trees, using field inspections, thermal modeling, and vegetation indices to track the infestation timeline.
Using statistical methods including Jeffries-Matusita distance, Analysis of Similarities, Wilcoxon rank-sum tests, and Difference-in-Differences, the study showed that:
- Infestation was detectable roughly three weeks after the initial attack, but two weeks before visible crown discoloration.
- The most informative wavelengths were in the red-edge (692–706 nm) and red (626–644 nm) regions—well before changes were visible in standard RGB imagery.
- Dividing the green stage into three substages helped clarify the timeline and detection potential:
- G-HS (hidden symptoms, early bark boring),
- G-BS (visible bark symptoms, green crown),
- G-YS (green-yellow stage, early crown stress).
Implications for Forest Management
Detecting infestation during the early green stage is vital for effective management, allowing for sanitary felling and targeted response before outbreaks escalate. Current detection often occurs too late, once crowns yellow or dieback begins, beetles may have already dispersed.
By combining high-resolution drone imagery with hyperspectral analysis and time-series data, this method offers a more precise and timely tool for public authorities and forest managers tackling climate-related pest outbreaks.
Learn More
Read the full article:
https://www.sciencedirect.com/science/article/pii/S1470160X2500799X