High-ResolutionSpatiotemporal Dynamics of AirborneeDNA in a Forest Ecosystem

Abstract Global biodiversity decline necessitates scalable, long-term monitoring systems. Airborne environmental DNA (eDNA) offers a noninvasive approach for multitrophic assessment, but its high spatiotemporal dynamism and low concentrations demand advanced sampling technology. We developed an automated, high-frequency (4 h), multichannel airborne eDNA sampler based on negative-pressure cyclone centrifugation, enabling remote, continuous collection across vertical forest strata. Deploying three samplers across a tropical rainforest canopy gradient (floor, mid-, upper canopy) for 6 days, we obtained 108 samples with DNA yields ranging from 10 to 784 ng per sample. Metabarcoding detected 786 taxa spanning bacteria, fungi, insects, and chordates. Results revealed taxon-specific patterns: bacterial richness peaked during daytime; insects and chordates exhibited clear vertical stratification, with the highest richness in the upper canopy. Time dominated community variation, while insects showed the strongest relative vertical structure. Environmental drivers differed by taxon and scale, with local richness associated mainly with temperature, shortwave radiation, and precipitation, while stronger regional effects (50–100 km) were observed for bacteria, fungi, and insects than for chordates. By resolving fine-grained spatiotemporal dynamics of airborne eDNA, this automated, stratified platform provides an operational framework for long-term terrestrial biodiversity monitoring.