Adult mortality in a low-density tree population using high-resolution remote sensing

Area of Science:

• Ecology
• Forestry
• Remote Sensing

Background:

• Quantifying tree mortality is crucial for understanding forest dynamics.
• Traditional field methods can be labor-intensive and may not cover large areas.
• Remote sensing offers a potential solution for landscape-scale ecological monitoring.

Purpose of the Study:

• To develop and validate a statistical framework for estimating canopy tree mortality rates using time-series remote sensing data.
• To precisely quantify annual adult mortality for Handroanthus guayacan on Barro Colorado Island, Panama.
• To assess the accuracy of remote sensing-based mortality estimates in the presence of detection and data gaps.

Main Methods:

• Utilized high-resolution remote sensing data synchronized with annual flowering of Handroanthus guayacan over an 11-year period.
• Employed a Bayesian state-space model accounting for tree detection probabilities and missing observations.
• Validated remote sensing estimates with an independent field sample and field-checked mortality data.

Main Results:

• Detected 1,006 individual adult trees across 18,883 observation attempts.
• Estimated an adult population of 1,135 Handroanthus guayacan individuals on Barro Colorado Island.
• Calculated a precise annual adult mortality rate of 0.2%·yr⁻¹ (SE = 0.1), closely matching the field-derived estimate of 0.33%·yr⁻¹.

Conclusions:

• The developed statistical framework enables accurate remote estimation of canopy tree mortality rates.
• This approach is effective even with variable detection rates and missing remote sensing observations.
• Remote sensing provides a powerful tool for landscape-scale forest health and mortality assessments.