Hydrodynamic cooperation allows groups of spores to disperse more effectively than single spores. When many spores are ejected simultaneously, they create a wind and travel many times further than singly ejected spores; this wind also enables spores to travel round obstacles (e.g. leaves and branches). This synchronized ejection of hundred of thousands of spores is called "puffing" and is observed in apothecial fungi which represent more than 9000 of the 35 000 described species of Ascomycetes. For example, the pathogen Sclerotinia sclerotiorum, must travel from fruit bodies found at ground level to the flowers of target crops, sometimes more than a meter off of the ground. It has long been known that any single spore decelerates very rapidly after ejection, because of its small size: in still air the typical range of a singly ejected spore is ~4mm. Moreover, singly ejected spores are easily checked by unfavorable airflows or blocked by obstacles.

By combining experiments using cultured S. sclerotiorum with modelling and mathematical analysis we demonstrate that synchronized ejection of spores allows apothecial fungi like S. sclerotiorum to overcome these range constraints. Rather like the cyclists in a peloton, many spores moving together sculpt a flow of air that almost totally cancels the drag upon each individual spore. The jet of spores travels 20 or more times further than a singly ejected spore, enabling S. sclerotiorum spores to infect the flowers of e.g. soybeans.

High speed imaging of spore ejection reveals that synchronicity is dynamically self-organized, and mathematical analysis shows how this spore coordination is stable even despite a potential for cheating. Specifically, we prove how cooperation among spores to create the wind can be stabilized against the conflicting interests of individual spores by a previously unreported kind of policing, the physical (or hydrodynamic) targeting of cooperative benefits to spores that cooperate maximally to produce the favorable airflow. Synchronous spore ejection may therefore provide a model for the evolution of stable, self-organized behaviors. This work was published in collaboration with Marcus Roper, Mahesh Bandi, Ann Cobb, Helene Dillard and Anne Pringle in the Proceedings of the National Academy of Science in 2010.