J Syst Evol ›› 2006, Vol. 44 ›› Issue (6): 627-640.DOI: 10.1360/aps06059

• Research Articles •

Morphological, karyological and molecular delimitation of two gentians: Gentiana crassicaulis versus G. tibetica

1ZHANG Xiao-Lan, 1GE Xue-Jun, 2LIU Jian-Quan*, 3YUAN Yong-Ming

1. 1(South China Botanical Garden, the Chinese Academy of Sciences, Guangzhou 510650, China)

2(Laboratory of Biological Evolution and Adaptation, Northwest Plateau Institute of Biology, the Chinese Academy of Sciences, Xining 810001, China)

3(College of Life Sciences, Beijing Normal University, Beijing 100875, China)
ljqdxy@public.xn.qh.cn
• Received:2006-03-29 Published:2006-11-18

Abstract: Gentiana crassicaulis Duthie ex Burk. and G. tibetica King ex Hook. f. are confusing gentians. The former has smaller corolla and blue-purple limbs whereas the latter has larger corolla and purple-brown limbs. Intermediate types are often found in southeast Xizang where both species are putatively considered to occur. A question has been raised whether these intermediate types were produced by introgression between the two species, which further contribute to the taxonomic ambiguity. In the present study, we comprehensively compared morphological variations, chromosome numbers, and sequences of two DNA fragments from populations of the intermediate types and those of the two typical species. We sequenced cpDNA trnS-G and nuclear ITS fragments of 69 individuals of 10 populations, counted chromosome number of 26 individuals of six populations and measured morphological variation of 54 specimens. We found that corolla size is highly correlated with ploidy level. The populations of typical G. crassicaulis from Yunnan were diploid (2n=26) with small corolla while those of G. tibetica from western Xizang were tetraploid (2n=52) with large corolla. The populations of the intermediate types from southeast Xizang were tetraploid with large corolla, and a few individuals with purple to brown limbs. These features suggest that the intermediate types show more similarities to G. tibetica. However, DNA sequence evidence unambiguously grouped those populations with G. crassicaulis. Both nuclear and chloroplast DNA showed consistent results: the intermediate populations showed identical cpDNA and very similar ITS sequences to typical G. crassicaulis. No sequence variation beyond what occurred in typical G. crassicaulis and G. tibetica was found from multiple individuals of the intermediate types. Thus the hypothesis of a recent introgression or hybridization between the two species could be rejected. Our results demonstrated that polyploidization had contributed to the intraspecific variation in G. crassicaulis and that the tetraploid individuals of this species showed morphological convergence with tetraploid G. tibetica, which blurred the morphological delimitations of the two species in southeast Xizang. Nevertheless, molecular markers can discriminate convergence and correctly attribute the intermediate populations or individuals to the correct species circumscription. Our results provided a case study to circumscribe the closely related species based on a comprehensive combination of morphological, karyological and molecular evidence. In addition, the geographical distributions of diploids and tetraploids within G. crassicaulis and the implications for the origin and dispersals of this alpine species were also discussed.