Molecular phylogeography of East Asian Boea clarkeana (Gesneriaceae) in relation to habitat restriction

Abstract

Subfamily Cyrtandroideae (Gesneriaceae) comprises a broadly distributed group of rocky-slope herbs, with China being the center of its distributional range. The normal growth of many species within the family is particularly dependent on special habitats. Due to the paucity of molecular studies, very little is known regarding East Asian herb phylogeographic pattern. Here, we investigate the molecular phylogeography of Boea clarkeana Hemsl., a unique resurrection herb endemic to China, focusing on geographically restrictive effects of habitat distribution on evolutionary history. Variation in three chloroplast DNA (cpDNA) intergenic spacers (psbA-trnH, rps12-rpl20, and trnL-trnF), the ribosomal internal transcribed spacer (ITS) and simple sequence repeats in expressed sequence tags (EST-SSRs) was investigated across 18 populations to assess genetic diversity, genetic structure and historical dynamics. Genetic diversity was low within populations (cpDNA, hS = 0.03, πS×10(3) = 0.17; ITS, hS = 0.16, πS×10(3) = 0.43) but high for species (cpDNA, hT = 0.82, πT×10(3) = 3.12; ITS, hT = 0.88, πT×10(3) = 6.39); 76 alleles were detected in this highly inbred species (FIS = 0.22), with a significantly low average of 1.34 alleles per locus. No cpDNA or ITS haplotypes were shared between regions. Based on cpDNA results, the Mt. Huangshan-Tianmu and Mt. Qinling-Daba haplotypes are ancestral; these two regions represent potential refugia. Although no evidence of significant retreat-migration phenomena during glacial cycles was detected, interglacial range expansion from northern Mt. Qinling-Daba was identified (121,457 yr BP). Rapid agricultural growth caused bottlenecks in many populations, especially on Mt. Huang-Tianmu. Habitat restriction and fragmentation, weak seed and pollen dispersal abilities, and long-term isolation caused by human-induced or environmental changes are considered the main causes of extinction of several populations and low genetic diversity within populations and regions. These analyses clarify the effects of habitat restriction on B. clarkeana, representing an evolutionary reference for similar gesneriads, and enrich our understanding of the molecular phylogeography of East Asian rocky-slope herbs.

Fig 1

(a) Distribution of 8 Chlorotypes of B. clarkeana. H-T, Mt. Huang and Tianmu; G, Guan Mts.; S, Shennongjia Mts.; Z, Zhangjiajie Mts.; W, Wu Mts.; WL, Wuling Mts.; N, Nan Mts.; Q, Qinling Mts.; D, Daba Mts.; DB, Mt. Dabie. (b) The TCS-derived Network of 8 Chlorotypes of B. clarkeana. Small open circles indicate mutational steps. Circle sizes represent sample sizes (n). (c) 50% Majority-Rule Consensus Neighbor-Joining Tree Obtained via Analysis of 8 Chlorotypes of B. clarkeana. Based on 1000 permutations, bootstrap values higher than 50% are indicated above branches.

Fig 2

(a) Distribution of 14 Ribotypes in B. clarkeana. (b) TCS-derived Network of 14 Ribotypes of B. clarkeana. Small open circles indicate mutational steps. Circle sizes represent sample sizes (n). (c) 50% Majority-Rule Consensus Neighbor-Joining Tree Obtained via the Analysis of 14 Ribotypes of B. clarkeana. Based on 1000 permutations, bootstrap values higher than 50% are indicated above branches.

Fig 3. PCoA of 18 populations of B. clarkeana.

See Table 1 for mountain codes.

Fig 4. UPGMA dendrogram depicting Nei’s (1978) genetic distances (D_A) between 18 populations of B. clarkeana, based on 15 EST-SSR Loci [45].

Numbers above branches indicate bootstrap values. See Table 1 for population and mountain codes.

Fig 5. Effective population size and years since a change for 18 populations of B. clarkeana, based on the MSVAR analysis.

N₁, mean current population size (over loci); N₀, mean ancestral population size; T, mean time (in years) since the population started to decline.