High Genetic Diversity and Insignificant Interspecific Differentiation in Opisthopappus Shih, an Endangered Cliff Genus Endemic to the Taihang Mountains of China

Opisthopappus Shih is endemic to the Taihang Mountains, China. It grows in the crevice of cliffs and is in fragmented distribution. This genus consists of two species, namely, O. taihangensis (Ling) Shih and O. longilobus Shih, which are both endangered plants in China. This study adopted intersimple sequence repeat markers (ISSR) to analyze the genetic diversity and genetic structure from different levels (genus, species, and population) in this genus. A total of 253 loci were obtained from 27 primers, 230 of which were polymorphic loci with a proportion of polymorphic bands (PPB) of up to 90.91% at genus level. At species level, both O. taihangensis (PPB = 90.12%, H = 0.1842, and I = 0.289) and O. longilobus (PPB = 95.21%, H = 0.2226, and I = 0.3542) have high genetic diversity. Their respective genetic variation mostly existed within the population. And genetic variation in O. longilobus (84.95%) was higher than that in O. taihangensis (80.45%). A certain genetic differentiation among populations in O. taihangensis was found (G st = 0.2740, Φst = 0.196) and genetic differentiation in O. longilobus was very small (G st = 0.1034, Φst = 0.151). Gene flow in different degrees (N m = 1.325 and 4.336, resp.) and mating system can form the existing genetic structures of these two species. Furthermore, genetic differentiation coefficient (G st = 0.0453) between species and the clustering result based on the genetic distance showed that interspecific differentiation between O. taihangensis and O. longilobus was not significant and could occur lately.


Introduction
Opisthopappus Shih (Asteraceae) is a genus of perennial herbaceous plants with two species, namely, O. taihangensis (Ling) Shih and O. longilobus Shih [1]. These two species are endemic to the Taihang Mountains (China) on the borders between Shanxi, Hebei, and Henan provinces. They grow mostly on cliff cracks, rock gaps in open forests below cliffs, and infertile soil at an elevation of 1000 m [1]. The flowers are white and large, with remarkably high ornamental and medicinal values [2] and potential genetic resource for chrysanthemum improvement [3][4][5]. Given the changes in their habitats and manmade damage, their distribution range is decreasing continuously. At present, these species are in an endangered status, listed as national rare and endangered plants in China [6]. O. taihangensis and O. longilobus can be distinguished according to their leaf morphology and the presence or absence of bracts below involucres [1]. The former is distributed in the south of the Taihang Mountain, mainly in Henan and Shanxi; meanwhile, the latter is distributed in the north of the Taihang Mountains, mainly in Hebei. The distribution of these two species basically regards the border between Hebei and Henan (36.07 ∘ N latitude) as the boundary. Based on relative specimen records, an overlapping distribution of these two species exists in the Linzhou district of the north of Henan.
Daxinganling Mountains-Taihang Mountains-Wu Mountains-Xuefeng Mountains is the borderline of the second and third classes of China Relief. The Taihang Mountains   2 The Scientific World Journal is northeast-southwest trend mainly but some parts nearly south-north trend, with the terrain of high north and low south. The altitude of most parts is above 1,200 m. The mountain is precipitous in the east and gradual in the west. Shanxi plateau is connected to the west, and the east part is from middle mountains, low mountains, hills to plains. This terrain allows the growth of rich seed plants in this area. The geographic distribution type of genera in seed plants is diversiform with a great amount of ancient, relict, original and endemic taxa of China, in which Opisthopappus and Taihangia Yu et Li are endemic taxa in this area [1,[7][8][9].
Generally, species with small geographic ranges tend to maintain less genetic diversity than those that are geographically widespread [10]. Several plant species distributed in cliffs show a relatively low genetic diversity [11][12][13][14], whereas a few of which have high genetic diversity [12,15,16]. High genetic diversity maintained in rare plants is attributable to many factors [17], such as the recent reduction of population size plus insufficient time for isolation, extensive and recurrent gene flow, outcrossing, and fruit dispersal via wind [18][19][20]. As an ancient habitat, cliffs provide a refuge for plants distributed therein to escape animal feed, interspecies competition, and human disturbance. However, a great heterogeneity also exists on a small scale; for example, differences are found in the proportion of plants to cliff cracks, soil quantity and quality, rock type, weathering degree, and number of surrounding plants [15]. Meanwhile, the differences in the orientation of vertical cliffs also cause differences in illumination, temperature, relative humidity, soil moisture, plant density, and wind speed [21]. This type of local environment diversity causes plants to exhibit polymorphism, which eventually results in high genetic diversity in populations [15,16]. In Lloydia serotina, species in cliffs have similar or higher genetic diversity than those in other areas [12]. Taihangia rupestris Yu et Li [22,23] also shows high genetic diversity. The results of cluster analysis show that T. rupestris var. rupestris and T. rupestris var. ciliate have significant genetic differentiation.
Currently, studies on Opisthopappus mainly focus on tissue culture [24,25], karyokinesis [26], meiosis [27,28], intergeneric crossing [3,5], pollen morphology [29], and phylogenetic evolution [4]. Wang and Yan [30] reported population genetic structure of two species in Opisthopappus by sequence related amplified polymorphism markers (SRAP). The results showed high genetic diversity existed in populations of the two species. The main aims of this study is to evaluate the genetic diversity and population genetic structure from different levels (genera, species, and populations) in Opisthopappus by adopting the inter-simple sequence repeat (ISSR) molecular marker. Furthermore, according genetic variation and differentiation, interspecific relationship and differentiation were analyzed and realized. The results can provide scientific data and theoretical foundation for the effective protection, sustainable development, and utilization of Opisthopappus species.  (LZ, LFS, and XT) populations were collected from the Taihang Mountains in Henan, Shanxi, and Hebei province to serve as materials ( Figure 1). A total of 200 individuals were collected ( Table 1). The spacing distance between individuals was more than 3 m. Fresh leaves in robust growth were selected for rapid drying using anhydrous silica gel. After drying completely, the leaves were placed in an ultra-cold storage freezer for cold storage.

Materials and Methods
Extraction of Genome DNA and ISSR-PCR Amplification. An ample amount of dried leaves was taken to extract genomic DNA using the improved CTAB method [31]. The mass and content of DNA were measured using a spectrophotometer to dilute it to 20 ng/ L. Twenty-seven primers with high polymorphism and good repeatability were screened from 54 ISSR primers for PCR amplification. The total volume of the ISSR-PCR reaction system was 20 L, including 20 ng to 50 ng template DNA, 1.5 mmol/L Mg 2+ , 1 × Taq DNA polymerase buffer solution (containing 10 mmol/L Tris-Hcl, 50 mmol/L Kcl, pH 9.0), 1.0 mol/L primer, 0.25 mmol/L dNTPS, and 1.0 U Taq DNA polymerase. The amplification program used was as follows: predenaturation for 5 min at 94 ∘ C; denaturation for 45 s at 94 ∘ C; annealing for 40 s at 55 ∘ C; extension for 70 s at 72 ∘ C, 37 cycles; and a final extension for 7 min at 72 ∘ C.
Genetic Diversity Analysis. According to the bands in the electrophoretogram, the positions with the same migration rate on the gel and with DNA bands were recorded as "1", and those without DNA bands were recorded as "0. " POPGEN 1.32 software [32] was used to calculate the genetic parameters: (1) percentage of polymorphic loci (PPB); (2) number of alleles ( ) and number of effective alleles ( ); (3) expected heterozygosity (He); (4) Nei's genetic diversity index ( ) and Shannon's information index ( ); (5) total genetic diversity index ( ) and genetic diversity index in populations ( ); (6) genetic differentiation coefficient ( The Scientific World Journal 3   (7) Nei's genetic distance ( ) and genetic identity ( ). Analysis of molecular variance (AMOVA) in GENALEX 6.4 software package [33] was used to calculate the genetic variation and genetic distance (Φ ). Based on Nei's genetic distance, NTSYS-pc 2.10e software [34] was used to conduct the unweighted pair-group method with arithmetic mean (UPGMA) cluster analysis. In addition, TFPGA software [35] was adopted to perform the Mantel test of correlation between the genetic and geographical distance between species groups.

Results
Genetic Diversity. A total of 253 ISSR bands were amplified, 230 of which were polymorphism bands. The PPB at genus level was 90.91%, 90.12% for O. taihangensis, and 95.21% for O. longilobus. Generally, at genus level, Opisthopappus showed higher genetic diversity. Nei's total genetic diversity index of Opisthopappus ( ) was 0.2139 (Table 2) (Table 3). At species level, all genetic parameters of O. longilobus were higher than those of O. taihangensis. Moreover, among the three populations of O. taihangensis, all genetic parameters of the GS population were the highest, whereas those of WML were the lowest. Meanwhile, among three populations of O. longilobus, all genetic parameters of the LZ population were the highest, whereas those of LFS were the lowest.
At population level, all the genetic parameters of LZ were the highest, whereas those of WML were the lowest (Table 3).
Genetic Differentiation and Gene Flow. Genetic differentiation coefficient ( ) between two species of Opisthopappus was 0.0453, which was lower than that among populations at species level (Table 2). Furthermore, genetic differentiation among populations in O. taihangensis was slightly higher than O. longilobus (Table 2).
At population level, genetic differentiation coefficient (Φ ) arranged from 0.122 to 0.427 (i.e., an average of 0.240) ( Table 4). The coefficient between LFS and WML was the highest (0.427), whereas that between XT and LFS was the lowest (0.122). For O. taihangensis, genetic differentiation between GS and YTS, GS and WML, and YTS and WML increased gradually; similarly, for O. longilobus, genetic differentiation between XT and LFS, LZ and XT, and LZ and LFS also increased gradually (Table 4).
Genetic Variation and Structure. The AMOVA results showed that most of genetic variation in Opisthopappus existed within population (77.87%), following among populations (16.15%), and between species (5.98%) ( Table 5). Most of the genetic variation of the two species, respectively, also existed within populations. Moreover, genetic variation of O. longilobus within populations (84.95%) was slightly higher than that of O. taihangensis (80.45%).
Cluster Analysis and Correlation between Geographical Distance and Genetic Distance. At population level, genetic distance between XT and GS population was lowest (0.0277) and that between LZ and WML was highest (0.1552) ( Table 6). According to UPGMA cluster analysis based on Nei's genetic distance (Figure 2), the respective populations of two species were not clustered into one clade at first, while XT and GS populations were clustered firstly and were then composed

Discussion
Genetic Diversity. In Opisthopappus, not only at genus level but also at species level, there was high genetic diversity.  [36,37] but were slightly lower than those of Taihangia rupestris, another cliff species endemic to the same area [22,23]. These indicated that two species of Opisthopappus both had higher genetic diversity. The degree of genetic diversity in O. longilobus was slightly higher than that of O. taihangensis. The AMOVA analysis showed that most of the genetic variation existed within populations and LZ had the highest genetic diversity among six populations. T. rupestris also had similar genetic diversity and genetic structure [22,23]. These two taxa (i.e., Opisthopappus and Taihangia) endemic to the Taihang Mountains showed similar genetic structures. For T. rupestris, ancient origin and comprehensive reproductive mechanisms might be the key contributing factors for its higher genetic diversity [23]. T. rupestris produced rich genetic variation through sexual reproduction during the long-term evolutionary process. The coexistence of asexual and sexual reproduction caused these variations to accumulate and spread [23]. Meanwhile, higher genetic diversity in Opisthopappus was closely related to its special habitats, population regeneration, and dispersal. Individuals of Opisthopappus all grow in cliff cracks with a slope of nearly 90 ∘ (Figures 3(a)-3(d)). No conditions for vegetative propagation are present, given that its seeds are small and light and easily spread by gravity and wind (Figure 3(e)). It can be    [39]. Wright [40] asserted that when gene flow > 1, genetic differentiation among populations due to genetic drift can be prevented. For O. taihangensis or O. longilobus, AMOVA showed that genetic variation mostly existed within population (80.45% and 84.95%, resp.). A similar phenomenon was found in T. rupestris [22,23], implying that habitat fragmentation and particularity do not cause significant genetic differentiation among populations in these endangered cliff plants endemic to the Taihang Mountains, and instead, promote them to adapt to the special habitats to sustain higher genetic diversity.
A significantly positive correlation exists between genetic and geographical distance for T. rupestris [22,23] and Centaurea corymbosa Pourret [41], another rare, cliff-dwelling species. By contrast, in the studies on Opisthopappus, the Mantel test revealed that no significant correlation exists between genetic and geographical distance ( = −0.2419, = 0.1845). Geographical isolation cannot explain the population differentiation that occurred in Antirrhinum charidemi and A. valentinum (endemic species distributed narrowly) [42] and Arctomecon humilis (an endangered plant) [43]. The absence of correlation between geographical and genetic distance in these species may be related to the rude interference of human beings at the least [43]. Human activity may break the balance between population migration and adaption, deepen habitat fragmentation [42], reduce the size of proper migration sites, and increase the isolated distance among populations [22,44,45]. O. taihangensis and O. longilobus both have a certain medical value, which led to serious human picking and mining. These activities cause great impact on the effective size of populations and their dispersal, affecting the natural gene flow among populations. They can explain the absence of correlation between geographical and genetic distance.
In T. rupestris, the clustering of the different populations into the two major clades agreed with the geographical distribution and its classification into the two varieties [22,23]. The moderate degree of gene flow shaped the genetic profiles of T. rupestris, homogenizing the genetic diversity of the two varieties, namely, T. rupestris var. rupestris and T. rupestris var. ciliate [23]. The two varieties were classified based on the difference in their leaf morphology [7]. They are distributed in the south and north of the Taihang Mountains, respectively. O. taihangensis and O. longilobus are similar to T. rupestris in terms of morphological difference and geographic distribution; however, the clustering result of this study shows that the respective populations of these two species do not gather but cross-cluster. Hence, interspecific genetic differentiation between two species is not significant.
Their interspecific differentiation is significantly later than the two varieties of T. rupestris [1]. O. taihangensis and O. longilobus are classified according to the split series of the upper leaves and the presence or absence of bracts below involucres. In O. taihangensis, the leaves are bipinnatifid and have pubescences attached to both sides; it does not have bracts below involucres. By contrast, most of the leaves of O. longilobus are pinnatifid except for basal leaves (bipinnatifid or near the bipinnatifid) and all leaves are smooth without pubescences; in addition, it has a pair of bracts below involucres [1]. Almost no genetic differentiation was found between the two subspecies of Dryas octopetala L. (Rosaceae) despite their significant morphological differences [46]. For Opisthopappus, the speed of interspecific genetic differentiation can be less than the speed of morphological differentiation, which may be closely related to mating system. Sexual reproduction will promote gene flow among populations, even between species, reducing the interspecific genetic differentiation. Meanwhile, in T. rupestris, the main mode of population regeneration is vegetative propagation [22], which is favorable to fix genetic characteristics in the level of population, subspecies, or species and thus enhance genetic differentiation.
Genetic Diversity and Resource Protection. The degree of genetic diversity to a certain extent reflects the capacity of species to adapt to the external environment. It not only restricts the capacity of individual and species to adapt to the environmental changes but also provides very important information for the realization and protection of species [14,47]. The results in this study have shown that the endangered status in Opisthopappus is not caused by the degree of genetic diversity but by population regeneration block and the difficulty to recovery due to habitat destruction and fragmentation. Therefore, an effective protection procedure should be developed to avoid a serious endangered situation; for example, in situ conservation through establishing a nature reserve, ex situ conservation in areas with serious habitat destruction, and manual promotion to natural regeneration can be carried out.