Abstract
More than a decade ago, the Driver (arbuscular mycorrhizal fungal partners drive plant communities) and Passenger (AMF community dynamics follows changes in the host plant community) hypotheses were suggested to explain the mutual relationships of plant and AMF communities. We propose one more hypothesis: the Habitat hypothesis, which postulates that both plant and AMF communities follow changes in abiotic conditions. The null hypothesis for all three working hypotheses can be called the Independence hypothesis, which proposes that plant and AMF communities are unrelated. We investigate the assumptions of these hypotheses and the available evidence in support of them. We suggest that community dynamics during secondary succession, including those related to land-use changes, may be explained by the Driver hypothesis, while the dynamics of plant and AMF communities during primary succession may be explained by the Passenger hypothesis. Within-region co-variation of successionally stable plant and AMF communities may be explained by the Habitat hypothesis, while the Independence hypothesis may explain global patterns of plant and AMF communities. These suggestions are tentative, and more evidence from both descriptive and experimental studies is required to assess them. In particular, comparative information is needed about dispersal limitation of plant and AMF communities in dynamic landscapes.