**3. Results**

#### *3.1. Soil Macronutrient Stocks and Allocation*

There is no significant difference in C stocks of the 0−20 cm soil layer, but the N, P, and K stocks in the 0−20 and 20−40 cm soil layers increased first and then decreased with forest stand age (Figure 2). The 14- and 25-year-old plantations displayed the highest P stock and K stock, respectively, whereas the lowest values of N and P stocks were all observed in the 30-year-old plantation. The K stocks were higher in the 20−40 cm soil layer than in the 0−20 cm soil layer except in the 14-year-old plantation; the C, N, and P stocks declined with increasing soil depth across all sites. The sequence of soil macronutrient stocks was as follows: C stock > K stock > N stock > P stock, except in the 20−40 cm layer of the 25-year-old plantation.

**Figure 2.** Soil macronutrient stocks across a chronosequence of Masson pine plantations. Different lowercase letters indicate significant differences in plantations with different stand ages (*p* < 0.05).

#### *3.2. Variations of Soil Macronutrient Stoichiometry*

Under all plantation types, C:N, C:P, N:P, C:K, N:K, and P:K ratios were higher in the 0−20 cm soil layers than in the 20−40 cm, except for P:K in the 14-year-old plantation and N:P in the 25-year-old plantation (Figure 3). The ratios of C:N and C:P first decreased and then increased, with a maximum level in the 30-year-old plantation. The ratios of C:K in the 0−20 and 20−40 cm soil layers were highest in the 30- and 14-year-old plantations, respectively. The variation in the N:K ratio was similar to that in the C:K ratio, but the N:K ratio was much smaller, ranging from 0.10 to 0.67 for all depths. The lowest C:K, N:K, and P:K ratios were all observed in the 25-year-old plantation. The stoichiometric traits of the soil were ranked as C:P > C:N > C:K > N:P > N:K > P:K in the 0−20 cm soil layer of the 7-, 14-, and 30-year-old plantations, whereas in the 20−40 cm soil layer, the sequence was C:P > C:N > N:P > C:K > N:K > P:K except for the 14-year-old plantation (Figure 3).

**Figure 3.** The macronutrient stoichiometry of soil at different depths in Masson pine plantations. Different lowercase letters indicate significant differences in plantations with different stand ages (*p* < 0.05).

#### *3.3. Relationships between Macronutrient Stocks and Soil Physicochemical Properties*

#### 3.3.1. Controls of Macronutrient Stocks at Various Plantations

Stepwise multiple regression was performed to determine soil physicochemical properties affecting the variability of soil macronutrient stocks within each Masson pine plantation (Table 2). For the 7-year-old plantation, 97.5% of the variation in the C stock could be explained by the AP, MWD, and organic matter together. The proportion of >0.25 mm was the most dominant factor for the C stock in the 14-, 25-, and 30-year-old plantations. Soil bulk density showed a positive correlation with N and P stocks in the 14-year-old plantation. The available N contributed 42.8% of the N stock variation in the 7-year-old plantation, and the available K explained most of the P stock variation in the 7- and 30-year-old plantations. The available P and organic matter were the most dominant controls for the N stock in the 25- and 30-year-old plantations, and the available P showed a negative relationship with the N stock. Most of the variation in the K stock in different plantations could be explained by soil aggregate stability.


**Table 2.** Models of soil macronutrient stocks at different stand ages.

Independent variables considered include the available N, P, and K; soil aggregate stability (MWD, GMD,WSA > 0.25 mm and D); and organic matter (OM), pH, CEC, SBD, and Ks.

#### 3.3.2. Controls of Macronutrient Stocks at Various Soil Depths

Redundancy analysis was used to explore the correlations between macronutrient stocks and soil physicochemical properties at various soil depths (Figure 4). The first ordination axis explained 87.2% of the variation, while the second axis accounted for only 5.8% in the 0−20 cm soil layer. The available K and P:K ratio explained most of the variation in soil macronutrient stocks. Negative correlations were observed between the C, N, P, and K stocks and Ks and MWD. The C:N ratio, C:P ratio, and available K were the most important factors associated with the N, P, and K stocks, respectively. Axes 1 and 2 captured 88.8% and 7.7% of the total variance in the 20−40 cm soil layer, respectively, and the variance could be explained mainly by the geometric mean diameter and P:K ratio. The stability of aggregates was strongly associated with C, N, and K stocks, while the P stock showed a positive correlation with the C:K, N:K, and P:K ratios.

**Figure 4.** Ordination plots of redundancy analysis (RDA) of macronutrient stocks and soil physicochemical properties at various soil depths.

For the total soil system, axis 1 explained 72.5% of the variability in macronutrient stocks, primarily related to organic matter, the C:P ratio, the available N, and the available P. Axis 2 described 23.4% of the variation, mainly associated with the available K, CEC, and the P:K ratio. The stocks of C, N, and P were positively correlated with the available N, P, K, and organic matter and negatively related to pH. Moreover, there was a strong positive relationship between the C stock and the C, N, P, and K stoichiometry. The K stock was positively correlated with CEC, MWD, GMD, and WSA > 0.25 mm but negatively correlated with the C:K, N:K, and P:K ratios and D (Figure 4).
