Orchids distribution models
Like most terrestrial biodiversity hotspots worldwide, orchids hold a prominent place in the flora of the Hengduan Mountains (Crain & Fernandez, 2020; Crain & White, 2013; Parsons & Hopper, 2003; Perez-Escobar et al., 2017; Souza Rocha & Luiz Waechter, 2010; Vollering et al., 2016; Zhang et al., 2015b). Since orchid centers of diversity frequently correspond to hotspots of other species groups, prediction and analysis of their suitability provide an available approach to knowledge of plants' fundamental geographic distribution patterns in biodiversity hotspots, and it also would provide ancillary benefits for biodiversity conservation beyond orchids (Anderson et al., 2008; Seaton et al., 2010; Souza Rocha & Luiz Waechter, 2010; Xing & Ree, 2017). Considering the broad ecological fitness of orchids (widely distributed in terrestrial ecosystems other than polar and extremely arid deserts) (Souza Rocha & Luiz Waechter, 2010) and their critical conservation status (all wild orchids are CITES-listed) (Luo et al., 2003), the protection value among different orchids should be fully expressed when using the SDMs for habitat analysis. In addition, due to the specificity of orchids to their environment and their adaptation to microhabitats (Kaur et al., 2021), the omission of model-predicted results caused by the species bias after performing spatial autocorrelation to occurrences in hotspots with high environmental heterogeneity should be allowed to be considered. Consequently, our methodology provides a valuable and comprehensive framework for spatial planning of conservation patterns in orchid hotspots based on ecological suitability. Of course, this approach is also applicable to the protection value assessment and hotspot analysis of orchids in other hotspot areas and supports relevant studies on biodiversity conservation for species groups with similar attributes.
Orchids geographical patterns and their influencing factors
The southern section of the Hengduan Mountains is geographically diverse, experiencing long periods of complex and rapid geological movement, resulting in rugged topography and significant environmental heterogeneity (Marchese, 2015; Wang et al., 2012; Yu et al., 2020). These features favor higher levels of plant diversity and rich pollinator pools for species dispersal and diversification (Acharya et al., 2011; Crain et al., 2014; Rewicz et al., 2017; Zhang et al., 2015b). The broad geographic specialization and species formation opportunities provided for orchids can contribute to explaining their wide distribution in the study area. It would also be consistent with the assertion that habitat heterogeneity is often considered an important driver of diversity (Crain & White, 2013; Perez-Escobar et al., 2017). Examples of orchid diversity studies in Central and South America indicate that the rapid growth of mountain ranges, volcanic activity, and glaciation at high altitudes proved to be the main drivers of orchid evolution and speciation (Crain & Fernandez, 2020; Dodson, 2003; Kirby, 2011).
The significance of vegetation and elevation as prominent elements affecting orchid suitability has been confirmed in research worldwide (Acharya et al., 2011; Bernardos et al., 2007; Borrero et al., 2022; Djordjevic et al., 2020; Faruk et al., 2021; Hemrova et al., 2019; Jacquemyn et al., 2008; Perez-Escobar et al., 2017; Souza Rocha & Luiz Waechter, 2010; Timsina et al., 2016; Vollering et al., 2016). The existence of differentiated vegetation types composed of different population-building species, given the ecological pattern of regional habitats, implies that for a particular orchid, the same vegetation type may contain more potentially suitable, specialized microenvironments (such as soil with the mycorrhizal environment, stable pollinators, appropriate hydrothermal conditions) (Djordjevic et al., 2020; Kelly et al., 2013) for population development and dispersal. The difficulty of crossing natural geographical barriers created by different vegetation types also explains that only five vegetation types are the most suitable habitats in our model. Elevation change affects the distribution of orchids by influencing temperature conditions, which, as we know, affects seed germination, phenology, and population density of the orchid family (Zhang et al., 2018), and to some extent, determines the physiological upper and lower limits of geographical distribution. The possible explanation for the modeled unsuitability of orchids to higher mean temperatures of the hottest season is that the typical non-zonal dry and hot river valley climate in the study area is characterized by hot and dry summers with burning wind effects is at odds with the indication that the cool-moist environment is a possible explanation for the abundance of orchids (Crain & Fernandez, 2020). Meanwhile, studies suggest that the optimal water-energy dynamics obtained under cool-moist conditions can promote high biodiversity levels (O'Brien, 2006). By analyzing the geographic attributes intimately associated with orchid suitability, we would identify the most important environmental features that support the distribution of orchids in the Hengduan Mountains and ultimately benefit orchid conservation efforts in various respects.
Despite the existence within the study area of a wide range of suitable habitats for orchids, the unevenness of diversity was also reflected, enabling the identification of diverse hotspots. Mountainous areas with enormous vertical elevation variations, such as the Muli mountain plain, the Shaluri mountain system, and the Dasetsu mountain, were the diversity hotspots in the study area, showing peak levels of mountain diversity similar to the results of many mountain research projects (Acharya et al., 2011; Zizka & Antonelli, 2018). The distribution pattern of protection hotspots resembled the diversity pattern in this study, which conforms to the assertion that diverse and heterogeneous environments in mountainous areas can reduce the risk of climate-driven extinction by providing large quantities of alternative habitats within a short distance (Mosbrugger et al., 2018). Hence, these areas can serve as refuges for ancient species. Another interesting finding is that these diversity and protection hotspots were located in the northwestern part of the study area (Muli, Yangyuan, Dacheng, and Yajiang counties) in the Jinsha River basin (Yalong River DC). The geological evidence suggests that the climatic conditions in the river valley have been relatively stable since the Late Tertiary, which, together with the edge effect and channeling benefits, allowed new species to survive and reproduce, and species exchange was ensured (Zhu, 2014), giving rise to many endemic components, both paleo- and neo-endemic species. These results are also consistent with relevant flora studies in the Hengduan Mountains (Lang, 1990; Li & Li, 1993; Xu et al., 2014). These regions are priorities and core areas for orchid diversity conservation and potentially for new species diversification centers.
Differences between hotspot layers
It was clear that the differences between the layers were noticeable (Fig. 2), and there was a significant deviation between the habitat map and the diversity and protection hotspots. The highly suitable habitat in the southwest corner and west of the study area was not an essential diversity and protection core. The potential explanation is that these two parts belong to the Nujiang River basin, the Gaoligong Mountains, which is most strongly influenced by the habitat movement of the Himalayan orogeny, with younger geological age, shorter time of species formation and endemism (Li & Li, 1993), thus their species richness and protection value relatively low, even though the vertical variation is significant and enormous. Yet the high-quality habitat can provide sufficient space for current orchid diversification and development means that it would be an essential part of conservation planning. The situation may be dissimilar in the northeast corner, where there was no significant clustering about the richness or protection hotspot, despite the high predicted suitability. We speculate that these regions, a part of the Yunnan-Guizhou plateau, are faintly affected by river erosion, causing relatively less environmental heterogeneity (Sichuan Vegetation Cooperation Group, 1980), thus resulting in impacts on species diversity. At the same time, the relatively flat plateau surface has long been more susceptible to disturbance by human activities such as grazing, and it may be difficult for endangered species to remain. As a result, it is harder to generate diversity or protection hotspots. A particular situation arose in the northern part (Yajiang County), where protection hotspots appeared, but the model results did not show high-quality habitats. The possibility may be that for some endangered orchids, there are factors affecting survival, and targeted investigation and conservation efforts to address the causes of endangerment are formidable measures to avoid biodiversity loss.
Considering Fig. 3, all layer centers showed a gradually decreasing position change from east to west. One of the reasonable explanations is the retreating process of the ancient Mediterranean Sea from west to east (Sun, 2002). Other is that the changes in water and heat conditions brought about by the influence of the southeast monsoon are also possible causes. Another interesting finding is that the high-quality diversity and protection hotspots centers of mass were overall northward compared to other centers, and all located in Muli County (one of the counties with the highest orchid richness and endemism in China), which is closely related to the original forest vegetation preserved under stable geological conditions. According to our survey results and previous flora studies, Muli may be the central location for the evolution and diversification of one large flora of orchids. All of these imply that these hotspots maybe hold an equally important position in the whole of China, even Central Asia orchid floras.
Orchids multi-scenario conservation planning and applications
Multi-scenario conservation planning makes rational and effective utilization of resources, providing a reference basis for biodiversity conservation for different purposes at present and in the future. It was tough to cover all diversity and protection hotspots at any scale in our study if only habitat suitability predictions were used as the reference for conservation planning. This also proved our initial assumption, that the existence of species bias. Using the spatial overlay technology, we divided the biodiversity conservation plans corresponding to different scenarios. The strict conservation scenario (SS), based on the principle of protecting the most suitable habitat, the smallest buffer zones for diversity and protection hotspots, ensures the fundamental species pool of the region to a certain extent and is the bottom line to support the regional biodiversity (Catano et al., 2020). The economical conservation scenario (ES) is one of the most recommended methods under current conditions and is likely to be practiced. It protects the most suitable habitats and expands the scope of protection for the most endangered and richest areas to achieve the best protection effect. The positive conservation scenario (PS) is the most optimistic conservation plan, which can meet the maximum development of orchids in regional biodiversity hotspots. Undoubtedly, the managers will spend massive protection resources in choosing such a conservation scenario.
Our conservation plan formed four distinct clusters that contained divergent geographical attributes and conservation priorities. For Cluster Ⅰ and Cluster Ⅲ, although the diversity and protection value were less prominent than those of Cluster Ⅱ, there still exists a diverse and heterogeneous environment brought about by the enormous altitude difference, which, together with the warm and humid airflow brought about by the southwest monsoon, provides favorable conditions for the development and breeding of orchids, as well as numerous tropical orchid components have been developed here(Acharya et al., 2011; Perez-Escobar et al., 2017). Cluster II was the core of orchid biodiversity conservation for the south of the Hengduan Mountains. It has a long history of formation and stable climatic conditions that permit ample time for plants to enter, diversify, speciation, and spread. The complex mountainous environment with significant changes in vertical climate zones, frequent glacier movements, and upward and downward climatic shifts make the region include massive unique species, and many endangered plants can flourish here (Zhu, 2014). In the above three clusters, we propose to use the current protected areas as the basis and expand to hot spots outside the region. For Cluster IV, setting a string of protected sites to compose conservation corridors is a good choice. When facing global climate change threats, they provide vital migration corridors for the orchid population dispersal and are conducive to other species groups (Keeley et al., 2018; Pellerin et al., 2022). It is also a critical channel for species diversification centers to spread in all directions.
Considerations for future research
While our study affords a valuable framework for assessing the suitability of orchids in biodiversity hotspots and the results should provide significant information for investigating orchid distribution patterns in the Hengduan Mountains, there are several considerations when migrating to other scales or geographic regions for biodiversity studies. The orchid protection evaluation is currently based on the attributes of the species. We recommend that obtain enough survey data, such as the number of species, threat factors, and levels, to conduct more detailed conservation assessments, which would be more meaningful. Although these data are hard to obtain at large scales, we suggest that researchers try them at small scales. Of course, using separate or local lists for evaluation is valuable for local conservation efforts to advance in any case. Secondly, our quantification of species diversity did not address the issue of diversity weights within the cell grid, and being able to take β diversity into account in future studies would be a substantial expansion of this research. In addition, finer remote sensing data, acquisition of environmental variables (Krner, 2007), and fine gradient delineation of variables would facilitate the study of corresponding geographic attributes of target species within hotspots and will facilitate conservation efforts.