Structural characteristics of soil fauna community
A total of 49 orders were included in the final dataset, including 14 classes in 4 phyla. We classified all sites into three climate types based on their geographic locations, thermal characteristics, and water availability, as follows: temperate, subtropical, and tropical. The vegetation types include temperate coniferous and broad-leaved mixed forest, temperate deciduous broad-leaved forest, subtropical evergreen broad-leaved forest, and tropical rain forest (Figure 1). In general, the number of orders declined significantly with increasing latitude of overall orders, phytophage, and predacity faunas (Figure 1). The soil fauna diversity in the temperate zone was significantly different from that in the other two climatic zones (ꭓ2 = 7.582,p = 0.02, Figure 1). The difference of phytophage (ꭓ2 = 0.352, p = 0.839, Figure 1) and predacity (ꭓ2 = 1.040, p = 0.595, Figure 1) orders were not significant in the three climatic regions (Figure S1). Among the 14 sites being studied, the Jianfengling site belonged to the tropical rain forest, showed the highest among all plots (Overall: 33, phytophage: 9, predacity: 7). In contrast, the Aershan site, which belonged to the coniferous and broad-leaved mixed forests in cold temperate zones, showed the lowest species diversity among all plots (Overall: 14, phytophage: 3, predacity: 3; Figure 1).
DCA analysis was conducted to show plot ordinations (Figure 2). The highest similarity in order compositions was observed between Aershan and Changbaishan. Further, Sapporo and Tahe were observed with similar order compositions with Aershan and Changbaishan, as all of these four sites were in high-latitude temperate regions. Analogously, high similarities in order compositions were observed among the Tiantongshan, Badagongshan, Xishuangbanna, and Jianfengling sites, which were all located in low-latitudes subtropical and tropical regions.
The Monte Carlo permutation test showed that the total effects of Latitude, MAT, MTCM, EMT, and SOC reached a significant level on overall orders, phytophage, and predacity soil fauna composition (Table 4). The results of RDA ordination showed that seven variables of spatial and environmental factors can explain 31.7%, 36.0%, and 42.4% variation of total information of soil animal composition at overall orders, phytophage, and predacity, respectively (Figure 3).Specifically, the 1st ordination axis explains 18.7%, 20.9%, and 26.4% variation information of overall orders, phytophage, and predacity, respectively, and the 2nd ordination axis explains 13%, 15.1%, and 16% variation information, respectively. Soil fauna community was observed to be strongly affected by spatial and environmental factors (Table 4). Community composition differed significantly in the different latitudes (R2 = 0.519, p = 0.023). Variations in fauna orders showed a strong response to soil physicochemical properties (p = 0.004, Figure 3a) and temperature (p = 0.033, Figure 3a). The dominant Lumbricida belonging to Oligochaeta exhibited a positive relationship with SOC or latitude but a negative relationship with SBD, MAP, MAT, MTCM, or EMT values (Figure 3a). However, for Malacostraca, the dominant Isopoda showed a negative relationship with SOC or Latitude, but a positive relationship with MAT, MTCM, or EMT (Figure 3a). The phytophage Mesogastropoda belonging to Gastropoda was positively correlated with SOC and negatively correlated with SBD (Figure 3b), whereas the predacity Isoptera belonging to Insecta was observed as the opposite (Figure 3c).