Diversity and distribution of parasitic angiosperms in China

Abstract Parasitic plants are an important component of vegetation worldwide, but their diversity and distribution in China have not been systematically reported. This study aimed to (1) explore floral characteristics of China's parasitic plants, (2) map spatial distribution of diversity of these species, and (3) explore factors influencing the distribution pattern. We compiled a nationwide species list of parasitic plants in China, and for each species, we recorded its phylogeny, endemism, and life form (e.g., herb vs. shrub; hemiparasite vs. holoparasite). Species richness and area‐corrected species richness were calculated for 28 provinces, covering 98.89% of China's terrestrial area. Regression analyses were performed to determine relationships between provincial area‐corrected species richness of parasitic plants and provincial total species richness (including nonparasitic plants) and physical settings (altitude, midlongitude, and midlatitude). A total of 678 species of parasitic angiosperms are recorded in China, 63.13% of which are endemic. Of the total, 59.73% (405 species) are perennials, followed by shrubs/subshrubs (14.75%) and vines (1.47%). About 76.11% (516 species) are of root hemiparasites, higher than that of stem parasites (100, 14.75%), root holoparasites (9.00%), and endophytic parasites (0.15%). A significant positive relationship is found between the area‐corrected species richness and the total species richness, which has been previously demonstrated to increase with decreasing longitude and latitude. Moreover, more parasitic species are found in the southwest high‐altitude areas than low areas. Consistently, the area‐corrected species richness increases with increasing altitude, decreasing latitude, and decreasing longitude, as indicated by regression analyses. China is rich in parasitic flora with a high proportion of endemic species. Perennials and root hemiparasites are the dominant types. The spatial distribution of parasitic plants is largely heterogeneous, with more species living in southwest China, similar to the distribution pattern of Chinese angiosperms. The positive relationship between parasitic and nonparasitic plant species richness should be addressed in the future.

There is a high species diversity of parasitic plants partly because of a higher mutation rate compared to their nonparasitic relatives (Bromham et al., 2013). It has been estimated that there are ~4,500 parasitic plant species, accounting for about 1% of the whole angiosperms in the world (Heide-Jørgensen, 2008). Moreover, parasitic species are not derived from a monophyletic group, and they have independently evolved at least 12 times (Bellot & Renner, 2013;Naumann et al., 2013;Westwood, Yoder, Timko, & Depamphilis, 2010), which indicate that parasitic plants could be diverse both evolutionally and ecologically. Consistently, parasitic plants are widely distributed in various natural and seminatural ecosystems ranging from tropical rain forest to Arctic tundra, and moreover, they differ in life forms, for example stem vs. leaf parasite and hemiparasite vs.
holoparasite (Poulin, 2011;Stewart & Press, 1990). Several reports have recorded the species richness at the national level (e.g., 151 parasitic angiosperms in Nepal, O'Neill & Rana, 2016;146 in Turkey, Sürmen, Kutbay, & Yilmaz, 2015). Nevertheless, the diversity of parasitic angiosperms, as well as the factors influencing the diversity, has seldom been well recorded for countries with large terrestrial area and high species richness.
Besides, there is little research which deals with the factors contributing to geographical pattern of parasitic angiosperms across different climates. Watson (2009) proposed "the host-quality hypothesis" to account for nonrandom distribution pattern of parasitic plants. Joel et al. (2013) pointed out that the majority of parasites had a wide host range, which was mainly influenced by host geographical distribution and ecological relationships. Luo, Sui, Gan, and Zhang (2015) contended that host compatibility interacting with seed dispersal determined small-scale distribution of the mistletoe Dendrophthoe pentandra (Loranthaceae) in Xishuangbanna, southwest China. In fact, the distribution of a parasitic plant flora in a certain area generally results from biological (i.e., dispersal vector and host availability) and environmental factors (i.e., altitude, area, longitude, and latitude).
China has a large terrestrial area of 9.60 million km 2 and is the third largest country in the world. China's territory stretches 5,200 km from east to west and 5,500 km from south to north (ECCPG, 1985a), ranging between tropical, subtropical, warm-temperate, temperate, and cold-temperate biome. Because of a wide range of climate, combined with highly complex topography and wide range of habitats, China has a tremendous diversity of plant and animal species (Wu, 1980;Zhang, 2007a,b), with a recent record of ~34,450 indigenous higher plant species (Zhao, Li, Liu, & Qin, 2016). Assuming that parasite species richness is proportional to host species diversity, we may speculate that the spatial pattern of parasite species richness is

| Data sources
Only the species that obtain nutrients from host plants by haustorium were included in this study, according to the definition of parasitic plants by Heide-Jørgensen (2008). We did not include epiphytes, stranglers, and mycoheterotrophic plant species (or saprophytes) because epiphytes and stranglers do not uptake water and nutrients from their hosts and mycoheterotrophic plants obtain nutrients by means of hypha rather than haustorium. We also excluded alien, cultivated, or naturalized plants. A typical example is Santalum album, distributed in Pacific islands, which has been widely cultivated in Guangdong and Taiwan of China for more than one thousand years (ECFRPS, 1959(ECFRPS, -2004.
Data on parasitic species were mainly collected from published literatures and floras. First, a database was initially created from two books, Flora Reipublicae Popularis Sinicae (ECFRPS, 1959(ECFRPS, -2004 and Flora of China (Wu, Raven, & Hong, 1994-2012. The former, consisting of 80 volumes, contains a comprehensive list of Chinese vascular plants. The latter, consisting of 25 volumes, is the English revision of the former. We searched for such words as "parasit*," "hemiparasit*," and "holoparasit*" in English or in Chinese from the books. If a species' description contains one of these words, the species was considered as parasitic. Then, its taxonomic status (family and genus, species, subspecies, varieties, or forms), functional group type (root hemiparasites, root holoparasites, stem parasites, and endophytic parasites; Těšitel, 2016), life form (herb/shrub), endemism (native/alien), and distribution location (Provinces within China) were recorded. In particular, we assigned Cuscuta species as holoparasites, because they have no roots and their leaves are too small to contribute significant photosynthetic carbohydrate to plants (McNeal, Arumugunathan, Kuehl, Boore, & dePamphilis, 2007;Těšitel, 2016). We also referred to published literature (e.g., research articles, local floras, monographs, collections, and reports; Ding, Li, Fu, & Yang, 2010;Joel et al., 2013;Liu, 2013Li & Ding, 2005;Zhang, 2007a,b), as well as websites (www. cvh.ac.cn; http://foc.eflora.cn/) to update the checklist. For example, Monochasma savatieri, a root hemiparasite indeed, was recorded by Zhang et al. (2015) and hence was amended to the checklist, although it was not mentioned elsewhere. After sorting out the checklist of parasitic angiosperms, we arranged all families and genera according to APG IV (2016). The final version of the checklist was shown in Table S1.
China is officially consisted of 32 provinces/autonomous regions (minority-dominated regions)/municipalities. According to the studies addressing biological diversity (Huang, Chen, Ying, & Ma, 2011;Weber, Sun, & Li, 2008), the municipalities Beijing and Tianjin were emerged into Hebei Province, and the municipalities Shanghai and Chongqing were merged into Jiangsu Province and Sichuan Province, respectively. Therefore, we created 28 units (provinces/autonomous regions/municipalities) in this study, covering 98.89% terrestrial area of China. For each unit, the number of native angiosperms, area, altitude, midlongitude, and midlatitude were derived from geographical data based on Diversity and geographic distribution of endemic species of seed plants in China (Huang, Ma, & Chen, 2014).

| Data analyses
We first calculated area-corrected species richness of parasitic plants from the raw species number (species richness) for each provincial unit as D = N/log (A), where N is the number of parasitic species and A is the unit area (Rejmánek & Randall, 1994;Xing, Zhang, Fan, & Zhao, 2016). Then, we conducted linear regression analyses to determine the relationships between log-scale parasitic area-corrected species richness and physical setting (altitude, midlongitude, and midlatitude) and also log-scale total species number within the unit (total species richness). Moreover, multiple regression analyses were carried out to determine the primary factors (among altitude, midlongitude, and midlatitude) on the area-corrected species richness across China (Xue, 2011). All statistical analyses were performed using SPSS 22.0 (SPSS Inc., Chicago, IL, USA) and ORIGIN 8.6 (Origin Laboratory Corporation, Northampton, MA, USA).

| Geographical distributions
The Results of multiple regression analysis showed that altitude accounted for the most of variation in area-corrected species richness among provinces, followed by latitude and longitude (Table 4).

| Floristic richness
To the best of our knowledge, this study is the first reporting the nationwide diversity and distribution of parasitic plant species in  (Guo et al., 2016;Li, 1998;Zhang et al., 2016). A complete floristic inventory of parasitic plants provides basic information for building control or conservation strategies to effectively managing these parasitic plants.
Our study can be also of international use. We recorded 678 parasitic plant species, accounting for 2.28% of China's total angiosperm species; this proportion is much higher than the value worldwide (1%, Těšitel, 2016), also higher than that of Turkey that has 146 parasitic angiosperms, corresponding to 1.29% of its flora (Sürmen et al., 2015), but slightly lower than that of Nepal that has 151 parasitic species, occupying 2.93% of its total angiosperm species (Joshi, Joshi, & Joshi, 2000;O'Neill & Rana, 2016). Importantly, the endemism is pronounced in China's parasitic flora, with almost two-thirds species endemic to China. This figure is much higher than that of Turkey (13.01%) and Nepal (8.61%). This high proportion of endemism may be ascribed to the extremely high richness of native plant species in China, which is often assumed to result from the diversity of climate and topography.
From south to north, Chinese vegetation covers tropical rainforest, subtropical evergreen forest, temperate deciduous forest, and boreal forest, spanning several climate zones (Wu, 1980). From east to west, Chinese topography is characterized by terrain, ranging from plains in the eastern provinces (>500 m at altitude) to basins in the middle ones (1,000-2,000 m at altitude), to the Qinghai-Tibet Plateau in the west provinces (above 4,000 m at altitude; ECCPG, 1985b). Such a climate and geographic diversity may result in a wide range of habitats and facilitate local species differentiation, giving rise to endemism (Huang, Ma, & Huang, 2017;Wen et al., 2016). For example, the high species richness of the genus Pedicularis (including 441 herbaceous species

| Life forms
Perennial herbs predominate the recorded parasitic plants in China.
This is presumably because these herb species may have a short annual growing period but can accumulate biomass for several years to produce seeds and complete life histories, which allows them to use abundance ephemeral resource of host plants (Wu, 1980). in Sichuan). This is partly because these parasitic plants are mostly living in mountain and subalpine areas, where the winter is often long and frigid, leading to a short annual growth period. For example, of the highly diversified Pedicularis, 151 species are found in the alpine areas of southwest China (Li et al., 2002).
Among the four categories of parasitic plants, more than 3/4 is root hemiparasites. This figure is close to the proportion in Nepal, in which 108 of the total 151 parasitic plant species are root hemiparasites; indeed, at the global scale, it is estimated that more than half of the total parasites is root hemiparasites. The predominance of root hemiparasites may be ascribed to the fact that stable underground soil temperature and moisture (relative to the aboveground microclimate) facilitate plant growth and that they can be avoid of aboveground herbivores. More importantly, root hemiparasites are able to obtain more sunlight and limited resources in poor-nutrition habitats than in rich-nutrition habitats, thus making them have higher competitive ability over other parasitic types (Dueholm et al., 2017;Press & Phoenix, 2005). Additionally, it is also possible that the dispersal of root hemiparasites could be more advantageous than stem ones because seeds transported with soil are more likely to locate their hosts than those transported by animals for stem parasites.

| Floristic distribution
We have shown that parasitic plants widely distribute throughout different regions of China, spanning across different biomes. This is consistent with the extensive distribution on a global scale, ranging from the tropics to the Arctic. Such a wide distribution is associated with the fact that parasitic plants are a diverse group of angiosperms with regard to their morphology, taxonomy, and phylogeny, which enable them to adapt various habitats (Press & Phoenix, 2005). However, the distribution of species richness is not even among provinces. For example, the richness is particularly high in the southwest China, but is extremely low in the northwest and northeast provinces. Such a distribution is consistent with the prediction of host-quality hypothesis (Watson, 2009), which claims that parasites should present a clumped distribution.
Indeed, the distribution pattern of parasitic plants is similar to that of nonparasitic higher plants, especially of endemic plants in China (Huang et al., 2011), as indicated by the close across-province relationship between the total species richness and parasitic plant richness revealed in our study. This positive relationship could be because high host diversity can foster parasites diversity (Hechinger & Lafferty, 2005 China. This could be because altitude influences both climate and geomorphology, and hence, it is more likely to be indicator than longitude and latitude for the species distribution of parasitic plants.
In summary, our study for the first time provides a comprehensive checklist of China's parasitic angiosperms, which cover 2.28% of Chinese angiosperms and are mostly endemic to China. Of the recorded parasites, perennial herbs and root hemiparasites are the predominating life forms. Moreover, the distribution of species richness is heterogeneous and the richness is the highest in the southwest of China, similar to the distribution pattern for Chinese angiosperms. The leading factors responsible for distribution include latitude, longitude, and altitude, of which altitude accounts for most of variation in specie richness among provinces. In conclusion, our results indicate that China is rich in parasitic plant species and their distribution is generally similar to overall distribution pattern of Chinese angiosperms. We suggest that the positive relationship between nonparasitic and parasitic plant species richness deserves future studies.

ACK N OWLED G EM ENTS
We thank Dr. Zheng Huang for his assistance in data analysis. This study was financially supported by Three New Forestry Engineering

CO N FLI C T O F I NTE R E S T
None declared.

DATA ACCE SS I B I LIT Y
Data available from the Dryad Digital Repository.

AUTH O R CO NTR I B UTI O N S
G. Z. and S. S. conceived the ideas; G. Z. and Q. L. collected the data; Q. L. and G. Z. analyzed the data; and G. Z. and S. S. led the writing.

TA B L E 4 Multiple regression analyses
showing the influence of the key environmental factors on area-corrected species richness of parasites for 28 provinces in China (n = 28)