This experiment was performed at the Desertification Combating Experimental Site of the Qinghai Gonghe Desert Ecosystem Research Station (99°45′–100°30′E, 36°03′–36°40′N), which altitude is 2871 m. The climate belongs to the transition zone of alpine arid desert and semiarid grassland. The mean annual air temperature is 2.4°C. The mean annual precipitation is 246.3 mm, and the mean annual potential evaporation is 1716.7 mm. The mean annual frost-free period is 91 days, and the total solar radiation is 6631.69 MJ·m^-2·a^-1. The mean annual number of windy days is 50.6 days, and the maximum wind speed reaches 40 m·s^-1. The main vegetation type of the study area is sand-fixing plantations, dominated by the tree species Populus cathayana Rehd and Populus simonii Carr. and the shrub species C. intermedia, C. korshinskii Kom., Salix cheilophila Schneid., Salix psammophila C. Wang et Chang Y. Yang and Hippophae rhamnoides Linn. The zonal soil is chestnut soil, and the azonal soils are aeolian sandy soil, meadow soil and bog soil. The soil of C. intermedia plantations in this study is aeolian sandy soil (Li et al., 2023).

The field experiment was conducted from June 15, 2022 to August 17, 2024. From June 15 to 20, 2022, fine root samples of C. intermedia were dug in 4-, 9-, 11-, 16- and 22-years old plantations (The afforestation methods and initial densities of these plantations are the same). To minimize damage to the shrub, the sampling depth was 0–20 cm. Soil samples were collected simultaneously, with 3 replicates for each plantation age. In the laboratory, the fine root with a diameter greater than 2 mm and the inelastic dead root were removed, and then the fine root samples were washed and divided into 3 diameters, 1 mm< D ≤ 2 mm, 0.5 mm< D ≤ 1 mm and D ≤ 0.5 mm. All fine root samples were dried in the oven at 65°C for 48 h. A part of the fine root was used for the initial nutrient element determination, and the other fine root was cut into small root segments of 5 cm. Then, the 2.000 g fine root sample was weighed with a thousandth balance and putted into a 10×15 cm nylon decomposition bags (80-µm mesh, prevents root growth in the litter bags), marked and sealed. On June 28, 2022, fine root decomposition bags of different plantation ages were buried in the corresponding plantations at a depth of 0–20 cm. From August to April of the following year, the temperature was very low, which resulting the experiment could not be carried out during this period, therefore, fine root samples were collected during the growing season from May to October. In the first year of this experiment, after the fine roots were buried, the soil environment was disturbed, which requires some time to stabilize. Therefore, sampling was conducted only once at the end of the growing season in the first year, and it was defined as the first decomposition stage. During the growing season of the second year (2023, May to October), sampling was conducted approximately every two months, which are the second, third and fourth decomposition periods respectively. In the third year, sampling was conducted only once in the middle of the growing season, which was the fifth decomposition period. Samples were collected on September 17, 2022 (81 days), June 30, 2023 (367 days), August 26, 2023 (424 days), October 22, 2023 (481 days), and August 17, 2024 (780 days). On the sampling day, 15 decomposition bags were collected for each diameter of fine root, and soil samples were collected at the same time, with 3 replicates each plantation age. The samples were taken to the laboratory, the live root in the decomposition bag and the soil attached to the surface of the bag were removed, then the samples were washed under running water and dried in the oven at 65°C for 48h to obtain dry mass. In this study, a total of 1350 bags of fine root samples and 90 soil samples were obtained. The weighed fine root samples were ground and passed through 100-mesh sieve for the determination of C, N, P and K contents. Soil samples were also ground and passed through 100-mesh screens for the determination of soil organic carbon (SOC), soil total nitrogen (STN), soil total phosphorus (STP) and soil total potassium (STK) content.

The soil moisture and soil temperature in 0–20 cm depth of different plantation ages were monitored continuously by ECH₂O soil moisture monitoring system. As environmental factors, mean monthly soil moisture and soil temperature during the growing season from May to October were analyzed.

Fine root C and N content was quantified using a CHNOS Elemental Analyzer (Elemental Analyzer, Elementar Analysensysteme GmbH, Germany). Fine root P and K contents were quantified using an Inductive Coupled Plasma Emission Spectrometer (ICAP 6300 ICP-OES Spectrometer, Thermo Scientific, USA). SOC was determined by potassium dichromate and sulfuric acid method, STN was determined by Semimicro-Kjeldahl Method, STP and STK were obtained with the HF-HCLO₄-HNO₃ digestion method using a 6300 ICP-AES (Forestry industry standard of the People’s Republic of China, 1999).

Fine root dry mass remaining and nutrients remaining were expressed as follows (Fu et al., 2021):

Where M is mass remaining (%), M ₀ is the initial dry weight of fine root, M _t is the dry weight of fine root after decomposition during time t; N is nutrient element remaining (%), N ₀ is the nutrient’s initial concentration, and N _t is the nutrient’s concentration after time t.

The fine root decomposition rate constant k was calculated using the following exponential equation (Olson, 1963):

t is decomposition time (year), and a is a fitting parameter.

Time needed to reach 50% and 95% mass loss were estimated using the k value:

One-way ANOVA and Duncan’s multiple range test was used to analyze the difference of initial nutrients content and mass remaining of fine root at different plantation ages and diameters. Correlation analysis was used to analyze the relationship among fine root mass loss, nutrient release, fine root nutrient content and soil characteristics. The main influencing factors of fine root decomposition and nutrient release were analyzed by redundancy analysis (RDA). All statistical analyses were conducted using Excel 2019, SPSS 19.0, CANOCO 5.0 and Origin software.

The C content of 1 mm< D ≤ 2 mm and 0.5 mm< D ≤ 1 mm fine root was significantly higher than that of D ≤ 0.5 mm fine root (P< 0.05) (except 11-year-old). The K content of 1 mm< D ≤ 2 mm and 0.5 mm< D ≤ 1 mm in 11-, 16- and 22-year-old plantations was significantly higher than that of D ≤ 0.5 mm fine root (P< 0.05). There was no significant difference in N content of fine root among different diameters (P > 0.05) (except 16-year-old). With the plantation age increased, the N content of 3 diameters fine root was significantly increased (P< 0.05), and the K content was significantly decreased (P< 0.05), the C and P content did not change regularly with the increase of plantation age ( Table 1 ).

Fine root mass remaining decreased with the increase of decomposition time, and the decomposition rate in the first decomposition period (81 days) accounted for the highest percentage in the whole decomposition period, ranging from 56.55 to 71.28% ( Table 2 ; Figure 1 ). Fine root mass remaining in all plantations increased significantly with the decrease of fine root diameter (P< 0.05). After 780 days of decomposition, fine root mass remaining of 1 mm< D ≤ 2 mm, 0.5 mm< D ≤ 1 mm and D ≤ 0.5 mm were 27.31-36.17%, 28.12-42.36% and 43.24-54.54%, respectively. With the plantation age increased, fine root mass remaining of 1 mm< D ≤ 2 mm increased significantly (P< 0.05) (except 9-year-old), and the fine root mass remaining of 0.5 mm< D ≤ 1 mm and D ≤ 0.5 mm changed irregularly.

The fine root decomposition coefficients of all plantations decreased with the increase of fine root diameter. With the plantation age increased, fine root decomposition coefficients of 1 mm< D ≤ 2 mm decreased slowly (except 9-year-old), and the decomposition coefficient of 0.5 mm< D ≤ 1 mm and D ≤ 0.5 mm first increased and then decreased ( Table 3 ). The time of fine root decomposition to achieve 50% and 95% mass loss of 1 mm< D ≤ 2 mm increased with the plantation age increase. The time of fine root decomposition to achieve 50% and 95% mass loss of D ≤ 0.5 mm is much higher than that of the other two diameter fine roots.

Fine root nutrient remaining decreased with the increase of decomposition time. In all plantations, fine root nutrient remaining increased with the decrease of fine root diameter ( Figure 2 ). After 780 days of decomposition, fine root N remaining of 1 mm< D ≤ 2 mm, 0.5 mm< D ≤ 1 mm and D ≤ 0.5 mm were 25.25-33.87%, 24.17-34.41%, 34.39-53.52%, respectively; C remaining were 24.70-32.40%, 25.93-37.39%, 39.63-51.06%, respectively; P remaining were 17.96-25.39%, 19.2-34.93%, 30.58-38.78%, respectively; K remaining were 3.87-7.04%, 5.44-10.42%, 8.46-14.63%, respectively. The decomposition rate in the first decomposition period (81 days) accounted for the highest percentage in the whole decomposition period, ranging from 56.55 to 71.28% ( Table 2 ; Figure 1 ). In the first decomposition period, the fine root nutrient release rate accounted for the highest percentage in the whole decomposition period. The release rate of K element was the fastest, and the percentage of K release rate in the first decomposition period was 76.86-94.73%, followed by N (58.87-87.53%) and C (52.95-82.12%), P is relatively slowly (45.10-72.06%) ( Figure 3 ). The enrichment of C, N and P elements occurred in the late decomposition period, and the enrichment of D ≤ 0.5 mm fine roots in 4 and 22 year old plantations was more obvious. The enrichment of K element began in the middle stage of decomposition.

The mass loss and nutrient release of 3 diameter fine roots were positively correlated with their K, C, N and P contents (P< 0.01), and negatively correlated with the STK content (P< 0.05) ( Figure 4 ). The results of RDA analysis of mass loss, nutrient release and environmental factors of different diameter fine root showed that the total interpretation of axis 1 and axis 2 of 1 mm< D ≤ 2 mm, 0.5 mm< D ≤ 1 mm, and D ≤ 0.5 mm fine root were 98.15%, 98.23% and 94.28%, respectively ( Figure 5 ). The K content of different diameter fine roots had the highest explanation for the mass loss and nutrient release, reaching 84.60%, 91.50% and 68.50%, respectively, followed by the C content (12.00%, 5.10% and 20.20%), and all reached a very significant level (P< 0.01) ( Table 4 ).

In this study, we found that the C and K contents of fine roots with 1mm< D ≤ 2 mm and 0.5mm< D ≤ 1 mm were higher than those D ≤ 0.5 mm, which was consistent with previous studies. For example, Wang et al. (2022) reported that the lower-order fine root (commonly< 0.5 mm in diameter) with lower carbon concentration, while the higher-order roots contained higher concentrations of initial C content (Yang et al., 2019). Jani et al. (2015) reported that the initial carbon content of coarse roots (>1 mm diameter) was significantly higher than fine roots (<1 mm diameter) for Trifolium incarnatum and Vicia villosa Roth. With the increase of plantation age, the N content of all fine roots was significantly increased (P< 0.05), and the K content was significantly decreased (P< 0.05). C. intermedia is a leguminous shrub, the large number of rhizobia with high nitrogenase activity in its root, which can fix free nitrogen in the air (Li et al., 2014). Potassium involves in the adjustment of membrane penetration in host cell and a series of assimilation process, promotes plant growth and improves photosynthetic efficiency, therefore ensures the nodule formation and nitrogenase activity in legumes (Hu and Schmidhalter, 2005). Previous studies on the allocation and cycling characteristics of main nutrients for C. intermedia in this region showed that the N content of root system was significantly negatively correlated with fine root K content, indicating that the nitrogen fixation process of the root system consumes its own K, and mainly consumes the K of the fine root with D ≤ 2 mm (Li et al., 2023). It probably because that with the increase of plantation age, the fine root nitrogen fixation ability increased, and it needs to consume more K element, which leads to a significant increase in the N content and a significant decrease in the K content.

In the first decomposition period (81d), the fine root mass decomposition and nutrient release rate reached more than 50% (except P release), and the K element release was the fastest (76.86-94.73%). It is generally believed that in the early stage of decomposition, the decomposition process is greatly influenced by abiotic actions, and soluble substances such as carbohydrates in fine root leachate rapidly, while in the later stage of decomposition, the decomposition process is mainly influenced by biological action, most soluble compounds are consumed, and insoluble substances such as lignin and cellulose remain, which are slowly degraded by microorganisms (Lv et al., 2020). Zhou et al. (2022) also reported that annual percentage mass loss of root during the first stage contributed about 75%–85% of the total mass loss over the 2-year decomposition period. Liu et al. (2021) found that the fine root of A. halodendron decomposed rapidly in the initial period, more than 40% of the mass was lost within the first 33 days. The release rate of P is relatively slowly, evidence has been reported suggesting that P is tended to be immobilized during root decomposition in forests (Gargaglione et al., 2018). Güsewell and Freeman (2005) reported that litter with N/P > 22 tends to be immobilized during decomposition. In this study, the N/P was greater than 22, and N/P increased with the decomposition time, which might be one of the reasons for the slower release of P element.

The K released right from the first decomposition period without immobilization in 3 fine root diameters because K is not the structural component of plant litter and not bound into any known organic compounds (Bhattarai et al., 2022), it is mainly present in the solution of plant cells in the form of ions that are released directly mainly through leaching (Osono and Takeda, 2004). Xu et al. (2006) also observed that the K releases rapidly from the beginning of decomposition.

The decomposition coefficient and nutrient release rate of fine root increased with the increase of fine root diameter, and the decomposition rate of fine root with D ≤ 0.5 mm was the slowest. Previous studies have shown that lower-order roots (commonly< 0.5 mm in diameter) are not lignified and are rich in nutrients, especially for root tips, which have higher levels of acid-insoluble compounds, while higher-order roots are the opposite. These difficult to decompose acid-insoluble compounds are the main factor causing the slower decomposition rate of lower-order roots (Sun et al., 2016; Yang et al., 2019). In addition, the smaller the diameter of the lower-order roots, the greater the proportion of root bark tissue, the higher the content of secondary defense substances such as tannin in the root bark cell wall, and the more difficult it is to decompose (Chen et al., 2001). Other studies suggested that the lower-order fine roots are more easily colonized by ectomycorrhizal fungi and the chitin rich mycorrhizal sheath may reduce the fine roots decomposition (Guo et al., 2008). Studies on the root decomposition of various temperate tree species also show that the content of acid-insoluble substances in roots with a diameter of less than 0.5 mm is greater than that in roots with a diameter of 0.5-3.0 mm (Hendricks et al., 2000). Xiong et al. (2013) also indicated that the decomposition rate of low-order root was significantly negatively correlated with the content of acid-insoluble substances, very fine root (< 0.5 mm) decayed slower than 0.5-2.0 mm root in temperate and subtropical tree species. Fan and Guo (2010) reported that lower-order roots (commonly< 0.5 mm in diameter) decomposed at a slower rate than higher-order roots (> 0.5 mm in diameter). The very fine roots (< 0.5 mm) with rapid turnover rate but slow decomposition, may promote the accumulation of soil C and N in terrestrial ecosystems (Fan and Guo, 2010).

The mass loss and nutrient release of fine root were positively correlated with their K, C, N and P contents (P< 0.01). Among the indicators monitored in this study, the K content of fine root had the highest explanation for the mass loss and nutrient release (68.50-91.50%), followed by the C content (5.10-20.20%), and both reached a very significant level (P< 0.01). This result was consistent with previous studies, for example, Wang et al. (2017) reported that fine root decomposition rate of Pinus massoniana was significantly positively correlated with the initial concentrations of N, P, K and Ca. Wang et al. (2022) reported that the fine root decomposition constant was significantly positively correlated with the initial C concentrations. Yang et al. (2019) also found that the low decomposition rate of the lower-order fine root was due to their relatively low carbon concentration. K, as a non-structural element, is mainly dissolved in plant cells and is a nutrient element that is more prone to leaching in litter (Aponte et al., 2012). K is also crucial for maintaining osmotic potential and activating enzymes. Previous studies have also shown that the release rate of K element from fine roots is faster than that of other elements, because K has a high solubility and can exude rapidly from the decomposed fine roots (Cao et al., 2018). In this study, K and C had a significant impact on the decomposition and nutrient release of fine root, it’s probably because the highly soluble K element and the easily decomposing soluble carbohydrates had a stimulating effect on the microbial community, thereby promoting its decomposition (Fan and Guo, 2010). Soil microbial community is the main decomposer in the soil decomposition system and also affect the decomposition of fine roots (Chapin et al., 2002). However, this study did not measure microbial indicators, and their impact on fine root decomposition needs further investigation. In addition, soil animals feed on fungi and bacteria, break up organic matter, spread microbial propagules and change nutrient availability, which affecting the activities of microbial communities, and then influencing the decomposition process of fine roots (Smith et al., 2009; Ruan et al., 2005). In this study, decomposition bags pore size was 80-µm, and the influence of soil animals on fine root decomposition was ignored. There are still many deficiencies in comprehensively and systematically revealing the process and mechanism of fine root decomposition with different plantation ages.