The natural abundance of stable nitrogen isotopes (δ¹⁵N) provides insights into the N dynamics of terrestrial ecosystems, the determination of which is considered an effective approach for gaining a better understanding ecosystem N cycling. However, there is currently little information available regarding the patterns and mechanisms underlying the variation in foliar-soil δ¹⁵N among mountain ecosystems. In this study, we examined the determinants of foliar-soil δ¹⁵N in association with N transportation rates along an elevational gradient in the Hengduan Mountains. Despite the relatively high levels of available N produced from high N fixation and mineralization, we detected the lowest levels of foliar δ¹⁵N at 3500 m a.s.l., reflecting the stronger vegetation N limitation at medium high elevations. The enhanced vegetation N limitation was driven by the combined effects of higher microbial immobilization and inherent plant dynamic (the shifts of δ¹⁵N in vegetation preference, including vegetation community) with changing climate along the elevational gradient. Unexpectedly, we established that soil δ¹⁵N was characterized by an undulating rise and uncoupled correlation with foliar δ¹⁵N with increasing elevation, thereby indicating that litter input might not be a prominent driver of soil δ¹⁵N. Conversely, soil nitrification and denitrification were found to make a more pronounced contribution to the pattern of soil δ¹⁵N along the elevational gradient. Collectively, our results serve to highlight the importance of microbial immobilization in soil N dynamics and provide novel insights that will contribute to enhancing our understanding of N cycling as indicated by foliar-soil δ¹⁵N along elevational gradients.

Plant leaf and fine root samples were washed with deionized water to remove dust particles and then dried at 60℃ for more than 48 h prior to further analysis. Soil samples were sieved through a 2-mm sieve, and having removed the residual vegetation, then dried at 60℃ to a constant weight. Both vegetation (i.e., leaves, litter, and roots) and soil samples were finely ground using a GT300 ball mill (PowteQ Inc., China) prior to isotope measurements. All N isotope and concentration analyses were carried out using an isotope ratio mass spectrometer. 

5-g samples of fresh soil were placed into 20-mL glass tubes, the headspace of which was replaced with a synthetic gas comprising O₂ and ¹⁵N₂ , Control samples were received unlabeled N₂ gas. The tubes were incubated at 20℃ in the dark for 9 days, and thereafter, the air and soil samples were collected and measured to calculate biological nitrogen fixation rate.

Four sub-samples of fresh soil were labeled with ¹⁵N, to which 2 mL of either (¹⁵NH₄)₂SO₄ or K¹⁵NO₃ solution was evenly added at an amount equivalent to 20 mg N kg^-1. Thereafter, one sub-sample of each ¹⁵N labeled portion was extracted with 2 M KCl (T0), and the other was incubated in the dark at 20℃ for 1 day (T1). The soil extract was used for determing the content of NH4⁺ and NO₃^- concentration and ¹⁵N abundance to calculate the gross mineralization, nitrification and microbial assimulation rate. 

The copy numbers of denitrifying genes were determined based on quantitative the real-time PCR analysis.

The detail information was in method.