PATTERNS OF ASSOCIATION BETWEEN CRUCIFERS AND THEIR FLOWER-MIMIC PATHOGENS: HOST JUMPS ARE MORE COMMON THAN COEVOLUTION OR COSPECIATION

Morphological and molecular phylogenies of animal parasites have often shown parallel cladogenesis, supporting hypotheses of coevolution. Few studies of the phylogenetic history for plants and their pathogens exist. Gene-for-gene interactions suggest that plant pathogens ought to have similar phylogenetic histories as their hosts. However, high dispersability combined with an inability to choose to leave if an inappropriate host has been landed on could increase the likelihood of host jumps and thus decrease phylogenetic congruence between plant pathogens and their hosts. In this study, I examined the pattern of association between the flower-mimicking crucifer rusts and their hosts by comparing independent host phylogenies (based on both cpDNA trnL-F introns and nuclear internal transcribed spacer [ITS] sequences) with that of their rust pathogens (based on ITS sequences). The expectation was that if the pathogens coevolved or cospeciated with their hosts, then their phylogenies should be congruent. Host-tracking coevolution can be differentiated from cospeciation by examining the times of divergence: If the pathogens are younger than the hosts, then it is likely that host tracking has occurred. For the crucifer rusts and their hosts, there was little evidence of parallel cladogenesis, suggesting that both cospeciation and coevolutionary tracking are rare. Instead, the most common pattern was one of host jumps to geographically associated taxa. There are at least three factors that may have contributed to the geographic structuring of the data. First, along the east-west transect stretching from the Rocky Mountains to California, large differences in rainfall and the timing of rainfall may reduce long-distance gene flow. Second, although dispersal of infectious spores is by wind, sexual reproduction of these fungi depends on insects, which move short distances. Third, host shifts are most likely to occur to geographically available taxa. Any species that grows adjacent to infected plants will be exposed to millions of spores, and the probability of eventual infection by a new mutant increases with greater exposure. Thus, patterns of association between the crucifers and their flower-mimic pathogens reflect jumps to geographically available hosts, which are not necessarily those that are most closely related. Corresponding Editor: C. Boggs