RESUMO
In 1895 and 2001, rust fungi affecting Licania trees (Chrysobalanchaceae) in Brazil were described as Uredo licaniae by Hennings in the state of Goiás and as Phakopsora tomentosae by Ferreira et al. in the state of Amazonas, respectively. Recently, a Licania rust fungus collected close to the Amazonian type location sharing symptoms with the former two species was subjected to morphological examinations and molecular phylogenetic analyses using 28S nuc rDNA (ITS2-28S) and cytochrome c oxidase subunit III (CO3) gene sequences. Since the original type specimen of Ph. tomentosae is considered lost, we carefully reviewed the type description and questioned the identity of the telium, which justified the description of the fungus as a Phakopsora species. Furthermore, the additional revision of the type material described by Hennings revealed that Ph. tomentosae is a synonym of U. licaniae. Based on the morphological examinations, disease symptoms, and shared hosts, we concluded that the newly collected material is conspecific with U. licaniae. However, the phylogenetic analyses rejected allocation in Phakopsora and instead assigned the Licania rust fungus in a sister relationship with Austropuccinia psidii (Sphaerophragmiaceae), the causal agent of the globally invasive myrtle rust pathogen. We therefore favored a recombination of U. licaniae (syn. Ph. tomentosae) into Austropuccinia and proposed the new name Austropuccina licaniae for the second species now identified for this genus. The fungus shares conspicuous symptoms with A. psidii, causing often severe infections of growing leaves and shoots that lead to leaf necrosis, leaf shedding, and eventually to the dieback of entire shoots. In view of the very similar symptoms of its aggressively invasive sister species, we briefly discuss the current state of knowledge about A. licaniae and the potential risks, and the opportunity of its identification.
