4. DISCUSSION

4.1 The C:N:P characteristics of soil and alfalfa

In our study, the change of SOC content was not obvious in 1-5 year stands, and even has a downward trend, but significantly increased in 7-9 year stands (Fig. 3I). Soil C in traditional farmland was found to have a similar trend (Jones et al., 2012, Li et al., 2011, Fiorini et al., 2020), while it was different with the increasing trend of organic C in forest or other species stand (Paul et al., 2002, Chang et al., 2014, Zhiyanski et al., 2016, Guo et al., 2020). A possible explanation is that alfalfa as a perennial leguminous forage was managed with the 3-4 times mowing per year, thereby, the decomposition rate of soil C has exceeded the rate of sequestration from litter and rhizome deposition (Vesterdal et al., 2002). As for the accumulation of SOC in 7-10 year stands, it can be attributed to the decrease in alfalfa biomass at the old-age period (Fig. 4I), so the OC taken away from the soil by mowing would decrease too. In addition, as the planting age of alfalfa increases, the increase of recalcitrant litter and organic matter in the soil would also enhance the C accumulation (Paul et al., 2003, Kubar et al., 2019, Li et al., 2019). The change of soil N content with planting age was similar to SOC. Due to the decrease of alfalfa’s absorption of soil C and N and nitrogen-fixation by root system (Yang et al., 2011a, Griesmann et al., 2018), the TN and AN content of the soil increased significantly in 7-10 year stands (Fig. 3II, Tables S1), which led the downward trend of C:N in 7-10 year stands (Fig. 3IV). The soil TP content did not fluctuate greatly in alfalfa planting stands, only increased in 9 and 10 year stands but was far less than the increase of C and N, so soil C: P and N: P ratios significantly increased in the 7-10 year stands (Fig. 3). This phenomenon is similar to previous studies (Wang et al., 2015), it probably due to the deep root system of alfalfa that brings mineral elements such as phosphorus to the top soil or the consequently weakness of P absorption (Li et al., 2011).
Alfalfa C, N and C:N were affected significantly by stand age in our study (Fig. 2I, II and IV). But the C:N ratio change of legumes was not like forest trees (Zhang et al., 2019a). The C:N ratio of woody plants would increase with growth age due to the increased proportion of xylem (Yang & Luo, 2011b, Zhang et al., 2019b). In our study, alfalfa C:N peaked at 7-year and then declined (Fig. 2IV). As a perennial legume, we harvest the branches and leaves of alfalfa, so that it was often in a status of regeneration, unable accumulate C as much as the wooden species. The C content does not have large fluctuation in different age groups, but the N content has increased significantly in 7-10 year stands, which resulting in a decrease in C:N content (Fig. 2). The P content decreased first and then increased with the alfalfa planting age (Fig. 2III), so it was understandable that the C:P ratio had the opposite trend, and the maximum value in 7-year stand (Fig. 2V), same with the C:N ratio. Thus, it suggested that at after 7 years, the growth of alfalfa was restricted, and the ability of plants to absorb soil P decreased. The content of TN and TP changes very similar in alfalfa, and both are significantly affected by stand age. N and P of plants were very active in metabolism, sensitively affected by plantation years (Wang et al., 2014) and other non-biological factors (Reich & Oleksyn, 2004b, Sardans et al., 2011b, Song et al., 2020). Similarly to previous research, the P content in plant leaves would decrease with the growth age (Fan et al., 2015a, Elser et al., 2010, Wang & Zheng, 2020) (Fig.2 III). The N:P ratio of alfalfa was negatively correlated with P content (Tables S2), which was consistent with studies in Molinia caerulea ,Carex flava (Gusewell, 2004) and Tibetan Plateau meadow (Hong et al., 2014). Plant N:P ratios are often used to reveal element limitations in ecosystems, generally believed that 14<N:P<16 is suitable for plant growth (Reich & Oleksyn, 2004a). However, different species may have different N:P thresholds. Braakhekke et al. (1999) measured the data of 74 grassland plots and found that when grassland plant N:P<10, their growth was mainly restricted by N, and when N:P>14, their growth was mainly restricted by P. Considering that alfalfa is a grass plant, so according to this report, we can clearly see that the growth of alfalfa in the 1-year stand and 2-year stand was mainly restricted by N, and from the 5th year and after, it was restricted by P (Fig. 2VI).

4.2 The interaction between alfalfa growth and soil stoichiometry.

Both N and P in alfalfa come from soil, so the changes of N and P in the soil will definitely affect the stoichiometric ratio in the alfalfa (Fig. 6), which verify previous research (McGroddy et al., 2004a, Yang & Luo, 2011b, Liao et al., 2014, Xie et al., 2020). The C:N ratio of alfalfa had increased significantly in the 7-year stand, because the N content in the alfalfa has decreased (Fig. 2). But the N content in the soil has increased significantly at the meantime (Fig. 3II). The reason for this may be that the plant’s biological nitrogen-fixing ability was weakened when it grew to a certain age (Jacobs et al., 2007, Lynd & Ansman, 1993) (McKenzie & McLean, 1984, Hwangl & Flores, 1987), so that a large amount of N cannot be transported into plant for its growth and development, but stayed in the soil. Then, the accumulation of N would accelerate root growth (Kou et al., 2019) (Fig. 4II). The accompanying decrease in the aboveground biomass of alfalfa in the 7-year stand (Fig. 4I) also illustrates its problem of growth limitation.
The C:P ratio of soil to alfalfa was closely related in this study (Fig. 6), and the C:P ratio in plants can indicate the efficiency of P use, symbolizing the absorption balance of C and P from soil into plants (Schindler, 2003), so the change of plant C:P ratio was caused by at least one of soil OC or TP. In our study, the plant C:P ratio was positively correlated with SOC (Tables S2). So, the C:P ratio of alfalfa was affected heavily by OC than TP, this was inconsistent with previous research results (Wang et al., 2015), the possible reason is that in the Mu Us Desert area, the soil P content is too scarce. In fact, that alfalfa cultivation had little effect on the TP content in the soil, only a slight increase in 9-year and 10-year stands (Fig. 3III).
The N:P ratio in soil and alfalfa was closely related (Fig. 6), however, the increase of N content in the soil did not directly affect the N content of alfalfa (Fig. 6). Cui et al. (2010) indicated that the N supplement in the soil would increase the N content in the plant leaves, which was inconsistent with us, probably reason is that the increase of N content in the our alfalfa stands mainly comes from the biological nitrogen fixation of alfalfa instead of the artificial N fertilizer in Cui’s research. But the increase of N content in the soil was significantly positively related to the N:P ratio of alfalfa (Tables S2). Plant N:P ratios can reflect their growth rate, Agren’s (2004) research shows that plants with higher growth rates had lower N:P ratios because higher growth rates require more rRNA, DNA and ATP etc. input to produce the phosphorus-containing compounds needed for growth (Fan et al., 2015b, LeBauer & Treseder, 2008, Ren et al., 2016), while they was a major P-bank of plants, therefore, an increase in phosphorus-containing compounds content would result in a disproportionate increase in P concentration in plant cells, thereby reducing the N:P ratio. The same phenomenon was found in our research. And as the alfalfa stands age increases, the N:P ratio had been increasing (Fig. 2VI), consistent with the report by Wang et al (Wang et al., 2014, Wang et al., 2015). The growth rate of alfalfa affects its yields, thus we find that its yields was significantly negatively correlated with the N:P ratio (Fig. 5), and there would be the maximum yield in the third year (Fig. 4I).
Our research also shows that the N:P ratio of soil and plants was positively correlated with root-shoot ratio (Fig. 5). In the early stage of alfalfa growth, due to the rapid growth of underground root system, the P content in alfalfa increased rapidly and resulting in a lower N:P ratio (Fig.2III and VI). With the growth of alfalfa, it needs more nutrients (OC, TN, TP) supplied by the soil (Fig. 3I, II and III). So, plant would product more and deeper root systems to cope with this situation (Fig. 4II). Kou’s (2019) research also said that the addition of N (similarly with the increase of N in 7-10 years, Fig. 3II) in P-limited soil would increase the root production. The formation of equal weight roots consumes more photosynthetic products, therefore, the prosperity of the root system would lead to a decline in aboveground production in nutrient limited ecosystems (Gersani et al., 2001, Laird & Aarssen, 2005, Ma et al., 2010, Maina et al., 2002), nutrient resorption from senescing leaves and stems can reduce the loss of nutrients inside the plant and to less depend on soil nutrients (Sohrt et al., 2018). But the available P content was mainly concentrated in the surface layer of the soil (Li et al., 2020, Barbosa et al., 2015, Cassagne et al., 2000). Therefore, the content of P in alfalfa did not increase with the growth of age, and the content of P and AP in the upper soil did not decrease significantly or even increased in 9-year and 10-year stands (Tables S1) due to the return of plant litter (Chen et al., 2016a, Liu et al., 2020), but it was not obvious compared with the increase of soil N content. Consequently, if we want to improve this situation through artificial fertilization, we should increase the proportion of P fertilizer used, and fertilize in deep soil.
We found that the N:P ratio of alfalfa was significantly related to soil microbial biomass C, N (Tables S2). In agricultural management, microbial biomass C and N are used as early indicators of nutrient cycling and organic matter dynamics in the soil (Xiao et al., 2018, Joergensen et al., 1995). In 7-year stand, the microbial biomass C and N had decreased significantly, and the growth of alfalfa was limited, which indicating that this area was no longer suitable for alfalfa cultivation at this time. Alfalfa aboveground biomass is negatively correlated with soil nutrients (Fig S1I,II and III), after 7 years of planting, although the growth of alfalfa continued to deteriorate, soil nutrient content (Tables S1), microbial biomass C, N and soil enzyme activity began to rose (Tables S3), all of those characteristics were significantly higher than FL. These phenomena mainly come from the growth of legumes, the return of litter, the increase of soil microbial communities, the decomposition of soil minerals and organic matter are all its functions (Iqbal et al., 2019, Kohmann et al., 2018). In summary, with the age of cultivation, alfalfa responds to more and more severe P restrictions by changing the allocation of aboveground and underground biomass, meanwhile increase the soil nutrient content, enzyme activity and microbial biomass, which make the soil more suitable for vegetation growth or agricultural production (Boerner et al., 2005).