רמי פוזיס

Floral scents play a key role in plant–pollinator interactions, shaping visitation by beneficial pollinators and ultimately influencing plant reproductive success. These scents are not static and often follow daily rhythms that create temporal patterns requiring detailed chemical profiling. To capture these dynamics, Gas Chromatography–Mass Spectrometry (GC–MS) is the standard method for analyzing floral volatiles, producing high-resolution and complex datasets. Interpreting such datasets increasingly relies on computational tools. However, most existing approaches do not account for circadian dynamics or background signals when integrating volatile emission patterns with pollinator activity. To address this gap, we developed ScentChrono, a computational pipeline that integrates clustering of volatile features with temporal emission patterns. We tested ScentChrono on Erucaria microcarpa (Brassicaceae), where it outperformed RAMClust, a widely used spectral clustering tool, in sensitivity, precision, and consistency. Targeted analysis identified 18 volatile compounds, while ScentChrono detected 14 compounds that consistently aligned with temporal emission patterns. Of these, 9 compounds including benzaldehyde, methyl benzoate, and benzyl acetate were identified by both approaches and displayed diurnal rhythms, with midday emission peaks matching pollinator activity. Our results demonstrate that ScentChrono captures both floral volatile profiles and their diurnal rhythms, providing a robust framework for understanding how floral chemistry shapes pollinator behavior.