Unsaturated constructed wetlands for the remediation of cylindrospermopsin and microcystin-LR.

The increasing risk of cyanotoxin contamination of water resources can be mitigated using nature-based solutions. However, removal mechanisms remain insufficiently understood, particularly under unsaturated conditions. This study investigated the removal of cylindrospermopsin (CYN) and microcystin-LR (MC-LR) in mesocosm constructed wetlands (CW) operated in unsaturated mode. Vegetation (Juncus, Phragmites, unplanted), porous media (sand vs. gravel), season/loading conditions, and environmental drivers were assessed together with microbial community structure (16S rRNA gene metabarcoding) and cyanotoxin transformation product (TP) candidates. Planted systems, especially Juncus, achieved high cyanotoxin removal, MC-LR (up to 97%) and CYN (up to 90%), comparable to saturated systems despite shorter contact times. Associations between cyanotoxin attenuation, rhizosphere oxygenation, and prokaryotic biodiversity were found. Porous media had no statistically significant effect on cyanotoxin removal. The mlrA pathway was not consistently detected and most mlr-associated TPs were absent. Although these observations do not exclude involvement of the mlrA pathway, they suggest that alternative and/or complementary degradation mechanisms may occur. Suspect screening identified nine tentatively annotated TP candidates, including five not previously reported in the peer-reviewed literature, more effectively attenuated in planted mesocosms. CW dimensions were estimated using the P-k-C* model. Overall, this study highlights the potential importance of plant-microbe interactions for cyanotoxin removal.