Molecular systematics and evolution of the non-photosynthetic parasitic Cistanche (Orobanchaceae)

Abstract

The parasitc Orobanchaceae have been the focus of few phylogenetic and systematic studies. While previous studies targeted on the genera Pedicularis, Orobanche and Euphrasia, here the phylogenetic relationships, systematics and taxonomy of the genus Cistanche are in the focus. Published studies so far involved only a couple of Cistanche species along with other genera, e.g. to infer the backbone phylogeny of the family. In order to address the enigmatic species, the research was intended to investigate the molecular phylogenetic relationships among the species, inferring a thorough taxonomic revision, defining basic chromosome number in the genus, estimating genome sizes and biogeography of the genus.<br /> A large scale molecular data set based on the rapidly evolving plastid regions (trnL- F, trnS-fM, psbA-trnH) and nuclear ITS was undertaken using many populations of almost all species distributed worldwide. The phylogenetic trees obtained from MP, ML and BI are relatively consistent in the major clades and confirmed the monophyly of the species. Cistanche was found to contain four distinct lineages. Two subgenera Subcistanche and Cistanche can be recognized, with the latter splitting into three clades that represent the three different sections. Subgenus Subcistanche is placed as first branching lineage, followed by the species of section Heterocalyx. Thus the remaining sections Macrocalycinum and Cistanche share are sister group relationship. The combined data set of plastid markers and nuclear ITS, however, could not fully resolve the widespread complex sect. Cistanche and failed to phylogenetically group the distinct morphological taxa.<br /> In chapter three, genome size were measured for some representative species of which some chromosome counts were analyzed and presented in a phylogenetic context. The difference between the highest (31.88 pg / 2C in C. tubulosa subsp. nov.) and lowest (8.56 pg / 2C in C. ambigua) genome content was estimated at about nearly 4-fold. Chromosome counts showed that all analyzed accessions of Cistanche were diploid (n = 20), with only one exception among C. violacea vouchers that was triploid (n = 30). Following this study, no correlation was found between the genome size of Cistanche and environmental including climatic factors. Similarly the no phylogenetic trend was indicated. The results further indicated the variation in genome size did not arise from polyploidization. Therefore, a possible mechanism to explain the high variation in genome size argues for differences in the amounts of repetitive DNA, e.g. the accumulation of transposable elements.<br /> The fourth aspect of the study concerned the re-appraisal taxonomic level of the genus. The considerable purposes of the study were the determination of intraspecific relationships in the genus. The species of Cistanche have been hardly identified due to reduction of morphological characteristics commonly seen in holoparasitic lineages in the family. According to our latest taxonomical survey Cistanche falls into two subgenera based on morphological features of the calyx of which subgen. Subcistanche distributed in China and Mongolia and subgen. Cistanche contains all other species widely dispersed in the Old World. The monotypic species C. sinensis from the first subgenus includes quadricalyx and the second group with 5-equal to subequal lobes. The genus has not treated comprehensively since c. 80 years ago (Beck-Mannagetta, 1930 and regional floras); therefore, this is the first attempt for a through monographic treatment. As a consequence, nine new species/subspecies (four new species in sect. Cistanche, four in sect. Heterocalyx and one in sect. Macrocalycinum) and several new records are introduced to the 16 previously recognized species by Beck-Mannagetta (1930). A diagnostic key, species description, nomenclatural corrections and sectional distributions are provided.