3. Result

3.1. Alfalfa C, N, P concentrations and stoichiometric ratio in differently age groups

The C, N and P concentrations of alfalfa were significantly affected by stand age (Fig. 2I, II, and III). The variation of the C content in the alfalfa fluctuated slightly from different age groups, with an average value of 453.0 g·kg-1, and the fluctuations did not exceed 4.7%. The change trend of C content was to increased first and then decreased, with significant maximum and minimum values in 7-year stand and 1-year stand, respectively. The N content in the plant varies from 28.6-45.7 g·kg-1, but there was no obvious trend in the change of alfalfa N concentration among in various years, and the minimum value was in 7-year stand, the maximum was in 10-year. P content decreased from 4.0 g·kg-1 to 1.9 g·kg-1 between 1-year to 7-year stand, but increased to 2.5 g·kg-1 in 9-year and 10-year stands.
The changes of alfalfa C:N, C:P and N:P were significantly related to stand age (Fig. 2IV, V, and VI). Alfalfa C:N and C:P had similar trend, and they all increased and then decreased with stand age. C:N had a significant highest value in 7-year stand and the lowest in 10-year stand. The 7-year stand also had the highest value of C:P, but the lowest C:P in the 1-year stand. The N:P ratios of alfalfa was positively correlated with the stand age (P <0.01), and there was over double increase in 10-year stand.

3.2. Soil OC, TN and TP concentrations and stoichiometric ratio in differently age groups

Soil OC was affected significantly by alfalfa stand age (Fig. 3I). The OC in the first two years was lower than 2.4 g·kg-1 of the FL, but it was increased with stand age. The significant maximum value of OC was 9.62 g·kg-1, appeared in 9-year stand. Soil TN was also significantly affected by the stand age (Fig. 3II), but the TN content were 0.12-0.23 g·kg-1 in 1-3 year stands, there was no obvious difference compared with the FL. There was a significant increase in 7-10 years, an increase of 4.5-7 times relative to the FL. The maximum TN appeared in 10-year stand and was significantly higher than other age groups. The fluctuation of soil TP was not as obvious as the SOC and soil TN (Fig. 3III), only in the 1-year, 9-year and 10-year stands had a significant increase, other age groups had no significant increase compared with the FL of 0.5 g·kg-1, and the maximum TP content appeared in 9-year stand, higher by 50.0% then FL.
Soil C:N was significantly reduced by the cultivation of alfalfa compared to the FL (Fig. 3IV). The first year of alfalfa planting fell by 45.1%, which was to the greatest extent. Then C:N started to rise, but it was always 16.7-34.0% lower than FL, and the maximum C:N appeared in 3-year stand in all alfalfa plantation. Soil C:P was slightly lower than the FL in the first two age groups, and then increased with stand age (Fig. 3V), The maximum C:P appeared in 10-year stand, which was 2.8 times of FL. Soil N:P ratio was significantly correlated with alfalfa stand age (Fig. 3VI). The soil N:P of all year stands were higher than FL, but it did not increase significantly except 5-10 year stands. The change trend of N:P was similar to the C:P, increased with stand age. The maximum value of C:P appeared in 10-year stand, was 4.9 times higher compared to the FL.

3.3. Relationships between alfalfa growth and soil stoichiometry

Alfalfa biomass was significant affected by growth age (Fig. 4I). It increased first and then decreases, with the largest biomass in the 2-year stand, the smallest in the 10-year stand, and the maximum value is 287.3% higher than the minimum value. Underground biomass generally increased with growth age (Fig. 4II), but the relatively higher in 1-year stand may be due to more weed roots in the newly established plantation of alfalfa. Therefore, it led to a root-shoot ratio of similar trends (Fig. 4III).
The biomass of alfalfa was negatively correlated to alfalfa N (R2=0.19, P<0.05) and positively related to alfalfa P (R2=0.26, P=0.02) (Fig. S1V and VI), so there were negatively related between alfalfa biomass and alfalfa C:P and N:P (Fig. 5). Underground biomass was negatively related to alfalfa P (R2=0.43, P<0.01) (Fig. S2VI) and positively related to alfalfa C:P and N:P (Fig. 5). The root-shoot ratio was positively related to alfalfa N (R2=0.27, P=0.02) (Fig. S3V), while negatively related to alfalfa P (R2=0.25, P=0.02) (Fig. S3VI), thus there was a positively correlation between root-shoot ratio and alfalfa N:P (Fig. 5).
The biomass of alfalfa was negatively related to soil OC (R2=0.70, P<0.01), TN (R2=0.81, P<0.01), TP (R2=0.42, P<0.01) (Fig. S1I, II and III), soil C:P ratio and soil N:P (Fig. 5). While underground biomass was positively related to soil OC (R2=0.65, P<0.01), TN (R2=0.69, P<0.01), TP (R2=0.34, P<0.01) (Fig. S2I, II and III), soil C:P ratio and soil N:P (Fig. 5). So it can be easily inferred that the root-shoot ratio was also significantly positively correlated with soil OC (R2=0.75 , P<0.01), TN (R2=0.82, P<0.01), TP (R2=0.43, P<0.01) (Fig. S3I, II and III), soil C:P ratio and soil N:P (Fig. 5).
The redundancy analysis in our article revealed the C, N, P, stoichiometric relationship between alfalfa and soil. The results showed that during the growth of alfalfa, the content of nutrient elements in the soil would be significantly affected by the C, N, P and stoichiometric ratio of alfalfa (the together explain value of total variation was 50.86%), alfalfa TP and N:P have the most significant influence (Fig. 6). Soil OC and alfalfa C content have a significant positive correlation, but N and P had no significant correlation between soil and alfalfa. The C:P and N:P ratios have significant positive correlation between soil and alfalfa (Fig. 6).