We also developed studies on the general evolutionary mechanisms at the origin of new diseases on plants (in natural ecosystems or crops) caused by fungi, by using various approaches. We used mathematical modeling to show the possibility of an evolutionary mechanism unrecognized so far, allowing the rapid emergence of new pathogenic species by specialization onto new hosts: the model showed that the life cycle of fungi mating within their host largely facilitated adaptation, specialization and genetic divergence. Comparative analyses carried out by the team confirmed that fungi reproducing within their host plant diversified more easily by adaptation to new hosts, without requiring evolving other mechanisms of reproductive isolation. The monitoring across several years of populations of Venturia inaequalis, a pathogenic fungi responsible for apple scab, having recently circumvented a resistance introduced in cultivated apple trees, corroborated the existence and the importance of this mechanism of host shift and specialization (in collaboration with INRA Angers, C. Lemaire). This monitoring indeed showed that, as predicted by the model, mating within its host facilitated adaptation and differentiation, and allowed rapid emergence of new pathogenic lineages on new hosts.

Phylogenetic and comparative studies conducted by the team showed in addition that pathogenic fungi diversify mainly by host shifts and not by co-divergence (diversification following that of their hosts), which was long the admitted hypothesis. Another aspect of research on the emergence of new species specialized on different hosts consisted in carrying out experiments on the Microbotryum anther-smut pathogens, closely related but specialized on different host plants. Experimental crosses revealed cryptic species and showed that the genetic incompatibilities between species, responsible for reproductive isolation, evolved linearly with genetic distance and were mainly due to chromosomal rearrangements, and not to incompatibilities between alleles, as it is more often thought.