RESUMEN
The generally held opinion that seeds of Lunaria remain at the replum after detachment of the two valves and then wind causes a shaking or rattling of the replum with its diaphragm, thus launching the seeds, is challenged. In a sparse forest in the Swabian Alb, the first author noticed flying valves of Lunaria rediviva to which the narrow-winged flat seeds are attached. Investigations with SEM and histology have shown that the valves secrete a glue only at those sites where the seeds rest on the valves before valve tissues die. Further analysis has shown (using the periodic acid-Schiff reaction) that the glue consists of polysaccharides. After detachment and dispersal of the valves, the adhesive strength continuously decreases. This is the first report for a sticky valve exudate in the Brassicaceae. Because of the adhesion of Lunaria seeds to their valves for some time, the 1st order diaspore is a mericarp, in a broad sense, and can be interpreted as an adaptation to long-distance dispersal by stronger winds. In this context, the 'flying carpets' of Lunaria are more effective and transport more than one seed. Molecular studies assigned Lunaria to the tribe Biscutelleae, which now contains the angustiseptate genera Biscutella and Megadenia as well as the latiseptate genera Lunaria and Ricotia. The valves in Ricotia can easily be detached (studied in herbarium material and a living plant), but, in contrast to Lunaria, the ripe seeds remain at the replum and its diaphragm, respectively.