4 Discussion
Species richness, including that of rodents, is characterized by spatial heterogeneity (Gaston, 2000). The results of this study indicate that there is a wide variety of rodents, including endemic species, in the Hengduan Mountains and surrounding areas of Southwest China. The Hengduan Mountain area is considered an ecological corridor between the Palaearctic and Oriental realms, connecting the northern and southern fauna. It is an important center for the origin of species (Wu et al., 2016). This may explain why southwest China has the highest level of endemism. Southwest China is rich in plant resources and has a complex topography and a high degree of habitat heterogeneity. The area’s climate ranges from tropical to temperate and is dominated by high temperatures and abundant precipitation (Liu et al., 2015; Shrestha et al., 2017). These findings may contribute to the diversity of rodents in China. In comparison, the extreme climatic conditions of the Qinghai-Tibet Plateau, Tarim Basin, and northern China may limit the distribution of species intolerant to climatic factors (Hu et al., 2017; Ding et al., 2019). The results show that the border areas of China’s Yunnan, Sichuan, and Tibet autonomous regions, especially the Qionglai, Yunling, Minshan, and Gaoligong mountain systems in the Hengduan Mountains, have the highest richness values of all studied areas due to their diverse topography, hydrothermal conditions, and habitats. The distribution pattern of rodent species richness may be explained by a combination of habitat heterogeneity, climate seasonality, energy-water, and human factors (Figure 3).
The close relationship between species richness and habitat heterogeneity and the variance partitioning results showed that habitat heterogeneity is the most influential predictor variable for describing the species richness of all and non-endemic rodent species in China. It is also an important factor in the richness distribution pattern of endemic species. Habitat heterogeneity is considered the most critical factor in shaping biological distribution patterns. Habitat variation creates a microhabitat for species reproduction by creating dramatic changes in climate and habitat differentiation on a small scale, enabling species to coexist locally (Carmignotto et al., 2022). Southwest China has a variety of hydrothermal conditions and habitats that result from a combination of topographic changes caused by elevation differentiation (Huang et al., 2011; Liu et al., 2015). As a result, the site had the highest richness values for rodent species. In this study, we used the elevation range as one of the main predictors of habitat heterogeneity, arguably the best indicator of topographic variation. Meanwhile, southwest China is also the richest region in terms of plant species (Shrestha et al., 2017; Dakhil et al., 2019; Sun et al., 2020; Pandey et al., 2020). Abundant plant resources provide a large amount of food for rodents and a suitable shelter for unfavorable climatic conditions (Barreto, et al., 2019). In addition, the Qinghai-Tibet Plateau’s uplift has profoundly changed the geomorphology of mainland China, mountains, ravines, and canyons, resulting in huge mountain height differences. Quaternary ice sheet intrusion drives rodent migration from higher to lower elevations, and the cyclical effects of warming cause rodents to return to higher elevations (Barreto et al., 2019). Faunas that cannot migrate can only adapt to the differentiation of low- and medium-altitude environments, thus causing high endemic species richness in rodents. In addition, highly heterogeneous tropical and subtropical mountain ranges may be cradles of biodiversity and thus dominate in terms of species richness and concentrations of endemic species with narrow distributions.
The results showed that climatic seasonality is the dominant factor in the distribution patterns of endemic species of rodents in China. Conversely, it explained the least of all the species and non-endemic species richness patterns. This suggested that seasonal temperature changes have a profound effect on endemic species richness (Penjor et al., 2022). Previous research has also identified a significant role of climate seasonality in maintaining patterns of species richness in endemic species of small mammals (Amori et al., 2011; Wu et al., 2013; Hu et al., 2017). The temperature in China fluctuates widely, with more extreme temperatures in the north than in the south. This might be an important factor in northern China’s restricted distribution of endemic species. Our findings support the theory of tropical niche conservation and the inability of endemic rodent species to adapt to northern environments. Their ability to migrate from south to north is consistent with our findings (Romdal et al., 2013). This seasonality in southwest China creates stable climatic conditions for endemic species to survive under harsh conditions. Araújo et al. (2008) and Dakhil et al. (2019) reported the significant role of climate stability during the warmest season of the Quaternary glaciation. Climatic stability is an ecological indicator of the range stability of subtropical fauna in the high-altitude regions of southwest China, including the eastern Tibetan Plateau (Thuiller, 2004; Huang et al., 2011). Due to climate variability and the unsystematic variation in daily maximum and minimum temperatures, modifying the thermal environment experienced by rodents to increase the body’s tolerance level enables them to be widely distributed geographically.
Our results also revealed the importance of anthropogenic variables in the distribution patterns of total and non-endemic species. However, its effect on endemic species richness was minimal. Thus, it can be expected that anthropogenic activities will significantly impact the distribution of rodents. Anthropogenic activities can drive the contraction and expansion of species. Many species have lost significant areas of distribution owing to increased land use and other human activities (Ceballos & Ehrlich, 2002; Di Marco & Santini, 2015). Human activities have also contributed to the range expansion of many other species, and many regional native species have experienced increased range through human-mediated dispersal and the ability to thrive in anthropogenic landscapes (Li et al., 2015). However, endemic species have more specialized habitat requirements, are concentrated in specific geographic areas, and may deviate ecologically from a wide range of species (Tomašových & Jablonski, 2016). Therefore, the anthropogenic impact on rodent endemics is limited. However, endemic species are confined to limited areas, and their habitats are often fragile and highly susceptible to reduction or even disappearance by external factors. Therefore, we cannot ignore human-interference scenarios. In southwest China, anthropogenic disturbances, such as natural resource exploitation, tourism, and land use change, are gradually increasing, accelerating the impact of human activities on wildlife. Therefore, protecting this center of species richness and endemism is a big challenge.
On a large scale, energy-water did not have a significant direct relationship with the distribution pattern of rodent species richness in China. This finding differs slightly from Hu et al. (2017) and Wu et al. (2013) at the local scale. The spatial pattern of species richness and the importance of biotic and abiotic predictor variables in influencing species richness may depend to a large extent on the spatial scale of the sampling unit (Qian & Kissling, 2010). In some cases, species richness patterns and potential drivers can be reasonably explained by ecological requirements and the evolutionary history of species grouping (Wu et al., 2013). Rodents are phytophagous or omnivorous; therefore, the pattern of species richness may be more closely related to plant species richness (Hawkins & Pausas, 2004). At a large spatial scale, the direct impact of energy-water on plants is often more significant (Zhang et al., 2015). Although the effect of energy-water on the distribution pattern of rodent species richness is not very significant at large scales, on a smaller scale, rodent dispersal behavior is temperature oriented, and species may spread more rapidly when temperature changes (Wu et al., 2013). Therefore, temperature and precipitation may strongly influence rodent species richness at a regional scale.
In summary, we against using all species or non-endemic species richness as proxies for endemic species richness. This is consistent with the results of previous studies (Orme et al., 2005; Isik, 2011). The reasons for the formation of endemic and widespread species differ. Endemic species formation is often closely related to environmental and anthropogenic changes. Some endemic species may have been widely distributed; however, due to climate change during the geological and historical period, their habitats have shrunk or even disappeared, forcing them to confine their distribution to a narrow range; or newly evolved species have not yet spread to a larger geographic area; or destruction of the environment limits their ranges. In addition, we found that the reasons for determining the distribution of non-endemic and endemic species richness differed.
Although this study contributes to a more nuanced understanding of rodent species richness patterns and their drivers in China, it has several limitations. For example, there are 271 species of rodents in China (Wei et al., 2021), however, only the 237 species with relatively complete data were analyzed in this study, which does not provide a complete picture of the distribution pattern of rodents in China. The variables used in MaxEnt partially overlap with the factors involved in the regression model, which will inevitably affect the regression results. In addition, the species distribution model assumes that species distribution is influenced only by environmental variables, without interactions between organisms and biological dispersal constraints. In practice, species distribution is also influenced by biological factors such as competition, predation, and disease. A more sophisticated and integrated SDM for species range prediction is required to consider these factors, which is an important direction for future model development (Kearney & Porter, 2009).