According to the main areas of distribution of C. funebris plantations in the Sichuan Basin and based on a literature review and field investigation, four representative sites in northeast Sichuan Province, China, were established. The four research sites were GA, DY, MY and SN City, which respectively represented low mountain, middle hill, shallow hill and high hill areas ( Figure 1 ). All sites were in a subtropical monsoon climate zone, and all C. funebris trees were aged 25 to 30 years. Undergrowth vegetation included primarily Myrsine africana, Vitex negundo, Coriaria nepalensis, Smilax china, Oplismenus compositus, Carex brunnea, Cyperus rotundus, Ficus tikoua, Adiantum capillus-veneris, Parathelypteris glanduligera, and Pteris cretica. Additional information on the research sites was provided in Table 1 .

In May 2019, forest subconditions (density and age, among other characteristics) were recorded at each site, which were consistent with little disturbance and well-developed natural vegetation, and standard lands were identified for forest subsites at each research site. The soil depth at the sample sites was approximately 30 cm. The standard plot size was 20 m × 30 m, and three plots were set up at each of the four sites for a total of 12 plots. In July 2019, three average C. funebris trees (trees of approximately the same average diameter at breast height (DBH), height, and mean diameter) were selected in each plot, and surface vegetation was removed 1 to 1.5 m from each trunk base. In four directions in the southeast and northwest, 20 cm × 20 cm × 20 cm blocks of soil were excavated, and surface soil and other impurities were carefully removed. We first determine the position of the target tree taproots according to its growth direction, and collect root samples. Then we combine the color (purple red), skin texture, smell (C. funebris essential oil), elasticity and the ease of root separation from the central column to put the live C. funebris roots into the fine sieve and remove other plant roots. Finally, we put the complete root segment of C. funebris with five orders of fine roots in the collection bags. Simultaneously with root collection, soil samples for determination of chemical composition were collected near each standard root sampling point to a depth consistent with that of root collection (0–20 cm). The diameter of the soil drill was 5 cm. To measure SBD and soil porosity (SP), soil samples at 0–20 cm were also collected with a 100 cm³ ring knife. We put the roots and soil in a self-sealing bag, stored them in the refrigerator, and then took them back to the lab.

Soil attached to fine roots was removed with low-temperature deionized moisture in the laboratory, and then roots were put in a 15-cm-diameter Petri dish containing low-temperature deionized moisture (2-4°C). Fine roots were graded according to the method of Pregitzer et al. (2002). First-order fine roots were derived from second-order fine roots, second-order fine roots were derived from third-order fine roots, and so on up to fifth-order fine roots. In addition, single roots attached to higher-order fine roots were also classified as first-order fine roots ( Figure S1 ; Pregitzer et al., 2002). Tweezers were used to remove each order of roots, which were put into a Petri dish containing a mixture of ice and moisture, and the root number was recorded. An Epson digital scanner (Expression 10000XL 1.0; Epson, Suwa City, Japan) was used to scan roots, and root image analysis system software (Win RHIZO Pro2009c; Canadian Ward Precision (Beijing) Technology Trading Co., Ltd., Beijing, China). It was used to analyze the morphological characteristics of fine roots of different root orders, including root D, length, surface area, and volume. After roots were scanned, they were oven-dried at 65°C to a constant weight, and dry mass was determined. Data on fine root D, length, surface area, and volume measured according to root image analysis system software were combined with fine root dry weights to calculate SRL, SRA, RTD and D (direct measurement). The calculation formula is as follows:

Where L is the root length, M is the root dry weight, S is the root surface area, V is the root volume, CR is the change rate, Max is the maximum value, and Min is the minimum value.

At each standard point, soil moisture (SW) was measured by drying (Schmugge et al., 1980). Organic carbon (OC) was measured by a potassium dichromate oxidation–external heating method (Sato et al., 2014). Alkaline nitrogen (AN) was determined by an alkali-diffusion method (Keeney and Bremner, 1966). Available phosphorus (AP) was determined by sodium bicarbonate extraction (Li et al., 2010). Total nitrogen (TN) was determined by a semi-micro Kjeldahl method (Trikilidou et al., 2020) and total phosphorus (TP) was determined by an alkali fusion-molybdenum antimony colorimetric method (Chapin, 1983). ST was measured with a button thermometer (DS1921G; Shanghai Bobamban Electronic Technology Co., Ltd., Shanghai, China). In May 2019, a thermometer was buried at a 10-cm depth in the soil at each site and set to record every 2 hours, and data were read when root systems were sampled (Lee and Wang, 2017). SBD and SP were measured using the tangent loop method (Chen et al., 2021).

Data were analyzed by SPSS 20.0 software (SPSS Inc., Chicago, IL, United States). Effects of root order and location on the morphological characteristics of fine roots were analyzed by two-way ANOVA, whereas the morphological characteristics of fine roots and differences in soil physicochemical properties were analyzed by one-way ANOVA. The least significant difference (LSD) method was used for multiple comparisons, with significant differences at P< 0.05. Pearson correlations were used to examine relations between different soil indices and root order morphology. Redundancy analysis (RDA) of fine root morphological characteristics and soil environmental factors was conducted with Canoco software (version 5.0) and get contribution rate. For all statistical tests, the significance level was 0.05. The figure was produced using RStudio (version 4.1.0).

The SRL, SRA, D and RTD of fine roots were significantly affected by root order and site (P< 0.05). However, the interaction effect between root order and site on fine root indices was not significant (P > 0.05; Table S1 ). In the four research sites, differences in SRL, SRA, D and RTD of fine roots were observed primarily in first- and second -order fine roots, and differences in third- to fifth -order fine roots were not significant ( Table 2 ). The values of the fine root morphology differed at the four research sites. We calculated its change rate using Formula 4. Under the influence of different soil environments, the SRL change rate of first -order fine roots was 53%, the SRA change rate was 20%, the D change rate was 23%, and the RTD change rate was 29%, while the SRL change rate of second -order fine roots was 36%, the SRA change rate was 21%, the D change rate was 26%, and the RTD change rate was 27% ( Table 2 ). Among the four sites, there were no significant differences in soil TP content, but there were significant differences in AN, AP, OC, and TN contents, and SW, ST, SBD, and SP ( Table S2 ). After difference significance analysis (LSD), the four sites showed significant differences in first- and second -order fine root morphology, while there was no difference in third- to fifth -order fine roots ( Table 2 ). This indicated that third- to fifth -order fine root morphology was not affected by the physical and chemical properties of the soil. Therefore, we further analyzed the correlation of soil AN, TN, AP, OC, SW, SBD, ST, SP, and first- and second -order fine root morphology.

The results showed that SRA and SRL of first- and second -order fine roots were directly proportional to TN content ( Figures 2C, A , respectively). The TN content was negatively correlated with the RTD of first -order fine roots and D of first- and second -order fine roots ( Figures 2D, B , respectively). The content of AN was positively correlated with SRL and SRA of the first- and second -order fine roots ( Figures 3A, C , respectively), and negatively correlated with RTD and D ( Figures 3D, B , respectively). OC content did not show significant changes with SRL, SRA, D and RTD of first- and second-order fine roots ( Figure 4 ). AP increased, SRA increased in first- and second -order fine roots, and SRL increased in first -order roots. AP content was positively correlated with SRA and SRL of first- and second -order fine roots ( Figures 5A, C , respectively). AP content was negatively correlated with the RTD of first-order fine roots and D of second-order fine roots ( Figures 5B, D , respectively). SW was positively correlated with D and RTD of first- and second-order fine roots ( Figures 6B, D , respectively); it was negatively correlated with SRL of first -order fine roots and SRA of first- and second -order fine roots ( Figures 6A, C , respectively). ST, SBD, and SP were not correlated with SRL, SRA, D, and RTD for first- and second -order fine roots ( Figures 7 – 9 , respectively).

In different soils, SW, AN, TN, and AP were related to fine root morphology. Among those soil properties, AN, TN, and SW content were strong. Soil OC, ST, SBD, and SP were not significantly related to fine root morphology.

Because of the relations between soil environmental factors and fine root morphology, RDA was used to analyze the importance of relations between fine root morphology (response variables) and soil environmental factors (explanatory variables). The first-sort axis explained 82.14% of total spatial variation, and the second-sort axis explained 6.42%, resulting in 88.56% of the total variation explained ( Figure 10 ). Thus, the first two axes well reflected the relations between fine root morphology and soil environmental factors. In Figure 10 , the angles between the SRL, SRA, AN, and TN were sharp, and the angle between the SW was obtuse for the first- and second -order fine roots. The angle between the D, RTD, and AP, AN, and TN was obtuse, and the angle between the SW was sharp for the first- and second -order fine roots. The angle between SRL, SRA, D, and RTD and SBD, ST, OC, and SP of first- and second -order fine roots was close to the right angle. This indicated that the SRL, SRA of first- and second -order fine roots were positively correlated with AP, AN, and TN and were negatively correlated with SW. However, the D and RTD of first- and second -order fine roots were positively correlated with SW and negatively correlated with AN, AP, and TN. The SRL, SRA, D, RTD, and while SBD, OC, SP, and ST of first- and second-order fine roots did not show a correlation. According to the RDA, AN was the main factor influencing fine root morphology, contributing 76% ( Figure 11 ). The results indicated that the fine root morphology was also influenced by SBD, which contributed 8%. But the contribution rate was only 8%. Therefore, among the environmental indicators, soil AN had the greatest effect on fine root morphology and was the leading factor driving changes in the characteristics of fine root morphology.

With the increase in root order, fine roots were mainly divided into absorbing roots and transporting roots (Qin et al., 2021). It was generally believed that the fourth- and fifth-roots were transportation roots (Guo et al., 2008; McCormack et al., 2015). However, the classification of absorbing roots seemed unclear. Some literature found that the morphology of first- and second -order roots changed with the soil environment (Li et al., 2017), while others found that first- to third -order fine roots changed (Gu et al., 2014; Wang et al., 2019). Our research results showed that only the first- and second -order fine roots changed with the soil environment, while the third -order fine root did not respond. According to the hypothesis of the ephemeral root model, the main physiological function of low roots (first- and second -order fine roots) was to absorb nutrients and moisture (Pregitzer et al., 2002), and they responded to the availability of soil resources through morphological plasticity (Trumbore and Gaudinski, 2003; Wang et al., 2013). And the thicker roots (third- to fifth -order fine roots) were mainly responsible for storing and transporting nutrients and supporting the aboveground parts (Bassirirad et al., 1993; Ouimet et al., 2008). Our research results supported this hypothesis. In addition, we had previously found that the action of third -order fine roots would slip (when the environment was different, the third root -order fine root function was unstable). It tended to absorb function in some environments and transport function in other environments (Li et al., 2022). It was generally accepted that fine roots differ in structure depending on the root order (Guo et al., 2008). However, studies have found that fine roots with different root orders might have similar structural functions. Such as cortical thickness (Gu et al., 2014), third -order fine roots would change from absorbing roots to transporting roots as soil depths varied (Wang et al., 2019). Consequently, we speculated that the third -order roots functions were varied due to the different structure and function among different root orders and different soil environment. And the specific reasons were needed to go a step further study.

Nitrogen was an important plant nutrient element (Shen et al., 2022), and the nitrogen cycle directly affected ecosystem productivity (Gruber and Galloway, 2008; LeBauer and Treseder, 2008; Xu and He, 2020). Nitrogen in the soil would affect fine root morphology (Brassard et al., 2009; Razaq et al., 2017), thus affecting the chemical composition of plants (Li et al., 2010; Mukta and Sreevalli, 2010). AN was a form of nitrogen in forest soil that was more critical than TN (Wallis et al., 2010). AN was an important indicator for evaluating soil nutrients and could affect nutrient cycling at the ecosystem level (Reich et al., 1997; Gong et al., 2020). This study found that the most significant environmental factor affecting the morphological changes of first- and second -order fine roots was soil AN (76%), and similar literature has been reported (Yu et al., 2017; Zhang et al., 2020). In many studies, AN significantly changed the morphological characteristics of roots (Drew, 1975; Wang et al., 2013; Chen et al., 2016). As the increase in AN content led to increase in photosynthesis and transpiration in trees (Carbone and Trumbore, 2007; Wang and Liu, 2014), the activity and capacity of plant-absorbing roots increased, leading to changes of fine root morphology. While the morphology and physiological activity of coarse roots were not affected (Adamtey et al., 2010; Bekku et al., 2011; Wang et al., 2017). AN was not easily leached from soil (Chen et al., 2016), and it could be directly absorbed by plants to affect growth (Schimel and Bennett, 2004; Li et al., 2013; Yu et al., 2020). This might mean that AN could change the morphology of fine roots by directly absorbing roots from plants. With the increase in soil nitrogen content, the respiration rate and activity of fine roots were enhanced (Wang et al., 2017; Bergmann et al., 2020), the nutrient absorption efficiency of fine roots was improved (Geng and Jin, 2022), and the transport capacity of fine roots was improved, thus driving the morphology of fine roots (Wang et al., 2018). Nitrogen in the soil changed the respiration rate of fine roots by affecting the concentration of root nitrogen (Jia et al., 2011), thus increasing the absorption and transport of ions in fine roots to maintain root growth (Lucas et al., 2011). The increase in soil nitrogen content increased the SRL and SRA of fine roots, decreased RTD (Makita et al., 2009; Noguchi et al., 2013; Geng and Jin, 2022), and changed the morphology of first- and second- order fine roots (Wang et al., 2013). AN in the soil also affected soil microbial function (Levy-Booth et al., 2014; Cong et al., 2015). For this reason, we inferred that AN might also promote root respiration by affecting the function of soil microorganisms, and the ions absorption by fine roots was affected by the increase of fine root activity. Finally, fine root morphology changed.

In many forest soil environments, phosphorus was the main limiting factor (Lang et al., 2017). Studies have found that the content of soil phosphorus increased, with the SRL and SRA of fine roots increasing and the RTD decreasing (Wurzburger and Wright, 2015; Jia et al., 2021). And this study also confirmed this conclusion. Root morphology were related to phosphatase activity (Ushio et al., 2015). Phosphatase activity in roots was positively correlated with SRL and SRA and negatively correlated with D (Ushio et al., 2015; Lugli et al., 2019). Hence, we inferred that the change in phosphorus content might stimulate the phosphatase activity of roots and promote the response of fine root morphology. However, phosphorus did not flow in the soil, and it was not readily available to plants (Aerts, 1999; Ma et al., 2020). Only a small amount of AP dissolved in soil solutions could be directly absorbed and utilized by plants (Fujita et al., 2010; Lambers, 2022). In the subtropical region of China, the content of soil phosphorus was relatively low (He et al., 2003), while the diffusion of AP in soil was relatively slow (Lewis and Quirk, 1967). Conifers might have conservative life strategies (Coomes et al., 2005), thus showing low morphological plasticity (Grassein et al., 2010). This may be the reason why AP contributes less to fine root morphology than AN. The soil was the largest repository of OC in the terrestrial ecosystem (Post et al., 1982). Plants released OC into the soil through their roots through photosynthesis, which eventually became soil organic matter (Rasse et al., 2005; Clemmensen et al., 2013; Jackson et al., 2017), and the process was usually called root deposition (Jones et al., 2004). However, root deposition occurred unidirectionally, so plants rarely used carbon in reverse (Trolldenier, 1987). Although the roots might also obtain a small amount of previously lost carbon element from the soil under the action of root exudates when OC existed in dissolved form, the roots could hardly directly control this process (Jones and Darrah, 1996). This meant that the fine root morphology was difficult to respond to the OC in the soil. This may be why OC in this study did not have a significant correlation with fine root morphology.

We found that with the decrease in SW content, D and RTD decreased, while SRL and SRA increased. And the same findings have been reported in previous studies. Such as, when SW content decreased, root SRL increased and D decreased (Zhou et al., 2018). According to the theory of the optimal allocation of resources, when the moisture supply was insufficient, plants would increase the root absorption area (Hertel et al., 2013), and the root moisture absorption efficiency would increase (Dhiman et al., 2017). The decrease of fine root D and RTD and the increase of SRL and SRA indicated that fine roots adapted to a thinner and longer morphology (White and Kirkegaard, 2010; McCormack et al., 2015). It could increase the contact area between plants and SW and improve the efficiency of soil volume use (Wasson et al., 2012). Less D and more SRL could reduce the ectoplasmic barrier of root xylem moisture to increase moisture transmission (Wasson et al., 2012; Comas et al., 2013), which was conducive to fine roots using soil nutrients more effectively (Preditzer et al., 2000; Lobet et al., 2014). This might be the reason why D and RTD decreased and SRL and SRA increased when SW content decreased in our study. At the same time, some studies found that when moisture content decreased, transpiration and respiration decreased, and the biomass of fine roots decreased (Zang et al., 2014). For example, in order to reduce moisture loss caused by transpiration, plants would allocated biomass to more durable roots under drought conditions (Brunner et al., 2015). As a result, we extrapolated that when the moisture content was low, plants might also be able to adjust fine root morphology according to biomass distribution so as to optimize moisture absorption and make the plant still grow well when moisture was insufficient. When the soil is well hydrated, the fine roots increase D and RTD, and the cell walls of phloem cells were strongly lignified, which improves the efficiency of moisture transport (Blokhina et al., 2003). This experiment found that D and RTD decreased and SRL and SRA increased when SW content increased. This may be the result of fine root morphology change caused by SW distribution. However, in the RDA analysis of this study, the contribution of SW to fine root morphology was small (contribution rate of 1.8%, Figure 9 ). It could be seen that SW might contribute to fine root morphology, but it was not the main environmental factor that causes fine root morphology changes.

ST was related to fine root growth. It had been found that the temperature of Flakaliden coniferous forest increased the SRL of fine roots (Leppalammi-Kujansuu et al., 2013). Norwegian spruce roots were also affected by ST (Kilpeläinen et al., 2019). In warm soil, SRL and SRA of spruce absorbing roots was increased, while RTD was decreased (Parts et al., 2018). At the same time, a high-temperature environment might reduce the growth of fine root D (Lyr and Garbe, 1995; Qin et al., 2007). The root length and root surface area were not affected at 24°C but decreased seriously at 28°C. This indicated that root morphology might have some adaptability to temperature (Sefloo et al., 2021). Nevertheless, fine root morphology did not show a response to ST in our study. From this point of view, it might be that the ST did not reach the appropriate level of fine root morphology change. Furthermore, some studies believed that the response of plants to temperature would vary from species to species. The law of factor complementarity held that the temperature difference would also be compensated by nitrogen, phosphorus, potassium, and other nutrient elements (Maurel and Nacry, 2020). Accordingly, we speculated that different species or nutrient complementation might also affect our research results.

The growth of fine roots would change with SBD (He et al., 2022). Some studies had shown that when roots grew in soil with large SBD and small pores, the axial growth of cells would be limited while the tangential growth would increase (Bengough et al., 1997). The length of the root system would also increase significantly with the increase in SBD (Ola et al., 2018). In hard soils, root growth might also differ between young and mature growth stages (Burr-Hersey et al., 2017). In this study, fine root morphology was not affected by SBD but showed a certain contribution rate. Synergistic or antagonistic effects of soil physical and chemical properties might weaken the impact of soil compaction. Some studies also showed that the growth of fine roots was the result of the combined effects of SW, ST, and nutrients (Preditzer et al., 2000). As a consequence, we deduced that the results of this study might be due to the combined effect of moisture and nutrients. This might also be caused by other reasons, and we look forward to further research.