Taxonomical Treatment Of Ficus Sect. Ficus (Moraceae)

Acorrding to Berg’s system, Ficus sect. Ficus comprises about 37 species all over the word, with the majorrity distributed in subtropical of China. Many of these species are difficult to be delimitated morphologically because of the continuous variations. This study based on two ways to re-evaluate the species status of this section:1) Molecular analyses.29 species and 151 samples of sect. Ficus were included, based on the nuclear ribosomal internal transcribed spacer (ITS) and external transcribed spacer (ETS), and the plastid psbA-trnH region; 2) Morphological analyses. Most of the related primary literature and types,479 specimens from HSNU, IBSC and IBK were inspected. The results show that:1. The phylogenetic trees revealed that F. abelii Miq., F. erecta Thunb., F. formosana Maxim., F. gasparriniana Miq., F. pandurata Hance, F. pyriformis Hook, et Arn, F. stenophylla Hemsl., F. tannoensis Hayata and F. vaccinioides Hemsl. ex King of subsect. Frutescentiae have close relationship and compose a core complex. The morphological characters support that the "ghost species" F. fusuiensis S. S. Chang, F. undulata S. S. Chang and F. sinociliata Z. K. Zhou & M. G. Gilbert should be considered as transitional forms of the species in this core complex. Hence, we currently combined these twelve species as one (F. erecta Thunb. sensu lato), established nine varieties in this species (i.e., F. erecta var. erecta, var. stenophylla, var. pyriformis, var. abelii, var. formosana, var. pandurata, var. gasparriniana, var. tannoensis and var. vacciniodes), and confirmed fourteen related new synonyms. Redescription of the species and check-list of its varieties are provided.2. Based on both molecular and morphological results, F. potingensis Merr.& Chun should not included in F. tuphapensis Drake but belongs to sect. Eriosycea. F. filicauda Hand.-Mazz. and F. neriifolia Sm. maybe one species. Morphologically, F. heteromorpha Hemsl. is distinct from F. variolosa Lindl. ex Benth., however, they show a close relationship on the phylogenetic trees. It probably means that introgression or hybridization exists between the two species, and/or there exists hysteresis effect and plasticity in the evolved morphological characters.3. Due to the existence of interspecific gene flow, both "F. gasparriniana" and "F. pandurata" have two or more different genetic background respectively. Some samples morphologically labeled "?F. gasparriniana" or "?F. pandurata" are out of the core complex on the phylogenetic trees, but clustered with F. chapaensis Gagnep., F. daimingshanensis S.S. Chang or F. tuphapensis, which means that they should be members of F. chapaensis, F. daimingshanensis or F. tuphapensis.4. The type of F. pyriformis var. brevifolia Gagnep. has been corrected as a member of F. variolosa; the types of F. silhetensis Miq. and F. silhetensis var. annamica Gagnep. should belong to F. tuphapensis Drake.5. Subsect. Ficus is native to the Middle East and Mediterranean, and it’s far from subsect. Frutescentiae on the phylogenetic trees. It infers that these two groups should not belong to the same section or subgenus.