The greater the genetic variation of a species, the stronger its capacity to adapt to intricate and ever-changing environments [24]. Consequently, examining the genetic diversity of species is instrumental in analyzing their biological characteristics. In this study, the average values of HE and HO of S. tzumu were recorded as 0.469 and 0.515, respectively. Notably, the value of HO matched that of Litsea auriculata, a nationally protected plant of the Lauraceae family (HE=0.515) [25], but fell short in comparison to other Lauraceae plants, such as Ocotea rotundata (HE=0.76 and HO=0.65) [26]. Furthermore, the values of HE and HO of 13 S. tzumu populations were inferior to those of other woody plants, including Xanthoceras sorbifolium (HE=0.53 and HO=0.72) and Olea europaea L. (HE=0.60 and HO=0.75) [27, 28].
The predominant distribution of genetic variation within the S. tzumu population was observed to be within populations (71%, p < .01), whereas the genetic variation among populations was minimal (10%, p < .01). These findings aligned with the outcomes of previous studies on perennial plants [29] and further supported the notion that the genetic differentiation of S. tzumu primarily occurred within populations (93.6%, p < .001) [13]. Similar to Zhang's investigation, which focused on genus Phoebe, the primary source of variation in this species was within population (about 75.5%) rather than between populations (about 24.5%) [30]. Furthermore, the overall genetic differentiation (FST) was calculated to be 0.103, suggesting a moderate level of differentiation. Genetic diversity played a crucial role in adapting to extreme environments, and insufficient diversity could impede survival advantages in the face of environmental changes. Despite the current wide distribution of S. tzumu, it may lack sufficient genetic diversity to adapt to climate change. The inbreeding coefficient (FIS) was calculated to be 0.215. It was worth noting that a high inbreeding coefficient may result in reduced genetic diversity and an increase in genetic structure within the 13 S. tzumu populations [31]. The number of migrants (Nm=2.172) indicated a moderate level of gene flow occurring among populations. The primary reproductive strategy of the Camphoraceae plant is out-crossing, a characteristic that is also observed in S. tzumu. S. tzumu possesses amphimerotic flowers, and its pollen is dispersed by insects, while birds aid in the dispersion of fruits. These two modes of transmission serve as the main sources of gene flow among populations. The transmission of this information occurs through the fruits, contributing to gene flow [15]. Based on the genetic structure analysis results of 13 populations of S. tzumu, S. tzumu could be divided into four most suitable clusters: cluster 1 was distributed in Hunan Province (HS and SS populations), cluster 2 was distributed in Jiangxi Province (LS and ML populations), cluster 3 was mostly located in the eastern part of the populations, and cluster 4 was partly from the eastern part of the populations (Fig. 1c). The PCA results also supported the division of cluster 1 and cluster 2 (Fig. 2). But a small number of these populations had already had some genetic exchange. The results of PCA also supported the situation. The composition of LS, ML, SS and HS populations was mostly similar, but there were differences, which may be related to the geographical boundary of Mount Luoxiaoshan, a mountain range bordering between Hunan Province and Jiangxi Province. In addition, most groups in the east were similar. This was related to the eastern hilly area and less high mountains. As a result, these groups exchanged genes to a large extent. Compared with the Hunan population, the Jiangxi population had partial gene exchange with most eastern populations, because the geographical distance between them was relatively close, and the eastern region was relatively flat, so gene flow did not become a major geographical barrier. However, the Hunan population was far away from the eastern population, and was blocked by mountains and the Yangtze River, so there was no similar gene exchange.
The Multiple Matrix Regression with Randomization (MMRR) analysis is a linear regression model that aims to quantify the impact of multiple explanatory variables, such as environmental factors and geographical distance, on genetic diversity. The results of MMRR analyses demonstrated that genetic distance was influenced by both geographical distance and environmental distance (r = 0.57, p < .01; Fig. 3b). Notably, IBE (r = 0.56, p < .01; Fig. 3c) exerted a significant influence on genetic distance. In the event of unfavorable environmental conditions, such as cold frosts or inclement weather during the pollination period, the pollination success of S. tzumu could be easily compromised [15], resulting in blooming without fruiting. Meanwhile, the distribution pattern of S. tzumu populations was found to be associated with elevation, slope direction, slope position, and humus thickness [32]. Research indicated that relative humidity (accounting for 26.2% of permutation importance), average temperature during the driest quarter (16.6%), annual precipitation (12.6%), and mean diurnal temperature range (10.3%) were the primary factors influencing the distribution of S. tzumu in China [14]. Moreover, a portion of the genetic variation could be accounted for geographical distance (r = 0.56, p < .01; Fig. 3a). The S. tzumu populations in HS and SS, as well as LS and ML, exhibited gene exchange and were distinct from other populations due to geographic distance. This could explain the results of the previous genetic structure analysis and principal component analysis. But environmental distance of S. tzumu in this study had a greater and profounder impact on its genetic diversity than geographical distance (βE = 0.46, p < .01; βD = 0.16, p < .01; Fig. 3b), which could be attributed to its preference for warm and humid habitats based on its biological characteristics [33].
To effectively elucidate the genetic structure of S. tzumu populations, we screened and developed 11,862 SNP loci. The findings revealed that S. tzumu populations exhibited a moderate level of genetic diversity. Simultaneously, it had come to our attention that the existing S. tzumu resources had suffered significant damage due to deforestation. In remote areas characterized by relatively underdeveloped economies, limited transportation, and insufficient protection and public awareness, it became imperative to implement in situ conservation measures and bolstered the dissemination of information regarding the protection of S. tzumu to deter indiscriminate deforestation. As for the development and utilization of S. tzumu resources, we propose placing emphasis on selective breeding of individual camphor trees, complemented by asexual reproduction. Additionally, it is crucial to formulate scientifically informed and comprehensive breeding strategies, manage S. tzumu genetic resources in a scientific manner, and enhance the sustainable utilization of these resources.