Abstract
Reef-building corals form symbiotic relationships with dinoflagellates of the genus Symbiodinium. Symbiodinium are genetically and physiologically diverse, and corals may be able to adapt to different environments by altering their dominant Symbiodinium phylotype. Notably, each coral species associates only with specific Symbiodinium phylotypes, and consequently the diversity of symbionts available to the host is limited by the species specificity. Currently, it is widely presumed that species specificity is determined by the combination of cell-surface molecules on the host and symbiont. Here we show experimental evidence supporting a new model to explain at least part of the specificity in coral-Symbiodinium symbiosis. Using the laboratory model Aiptasia-Symbiodinium system, we found that symbiont infectivity is related to cell size; larger Symbiodinium phylotypes are less likely to establish a symbiotic relationship with the host Aiptasia. This size dependency is further supported by experiments where symbionts were replaced by artificial fluorescent microspheres. Finally, experiments using two different coral species demonstrate that our size-dependent-infection model can be expanded to coral-Symbiodinium symbiosis, with the acceptability of large-sized Symbiodinium phylotypes differing between two coral species. Thus the selectivity of the host for symbiont cell size can affect the diversity of symbionts in corals.