观测与模式研究生物固氮及其对大气氮沉降增加的响应—以青藏高原高寒草原为例

N2 are abundant in the atmosphere, but it cannot be utilized by organisms, until be converted to reactive forms. In pre-industrial time, biological nitrogen fixation (BNF) and NOx production by lightning were the two main processes(we called natural N fixation, or NNF, when combine those two together ) that converted atmospheric N2 to reactive forms that could be used by plants. There is still existing large uncertainty regarding quantifying the biological N fixation rate for the global terrestrial ecosystems. From the earlier estimation of190(100-290) Tg N a-1 to current 58 (40-100) Tg N a-1。In the previous study, based on the mass balance theory, we have developed a modelling approach to simulate the ecosystems’ total NNF rate in the DyN-LPJ model. Our estimation values for global terrestrial ecosystems were 350(230-470) Tg N a-1，almost 6 times larger than the latest estimation。Our estimated N fixation rate for Tibetan Plateau alpine steppe were about 0-2g N m-2 a-1，Whether our estimation are consistent with the local observation？As BNF was energy costly microbiological processes, the BNF might be decreasing with the increasing atmospheric N deposition, then resulting maximum N absorbing rate by terrestrial ecosystems. We then hypothesized that, this maximum N absorbing rate might be equals to the BNF rate. In this project, we try to measure the (1) symbiotic BNF rate under the soil moisture gradient along the north-east transect of alpine steppe, this measurement will be used to verify the new developed N fixation model, and (2) measure the BNF function of legume species to increasing N addition by using the 15N dilution technique; These observations will be used to further developing the existing NNF model for their reaction to increasing atmospheric N deposition. This reaserch would be helpful to correctly quantifying the BNF of Tibetan Plateau alpine steppe and provide a possible effective method for simulating the terrestrial ecosystems BNF rate in the current earth system models.

当前对全球陆地生态系统生物固氮量的估计存在很大不确定性，早期估计约为100-290TgN/a，而最新估计约40-100TgN/a。作为微生物耗能过程，固氮功能可能随大气氮沉降增加而减弱，但尚缺乏机理模型反映这个过程。申请人在原有LPJ模式基础上开发了氮循环机理过程DyN-LPJ，进而基于物质平衡原理构建了全球陆地生态系统自然固氮模型，其结果接近早期估算量的上限，是最新估计的6倍！对青藏高原生物固氮量的模拟（0-2gN/m2/a）与初步(单点)实测结果更接近。因此希望在更大空间和更长时间尺度下检验这个预测结果，并发展模型对外源氮增加的反馈方程。本项目拟1）利用15N稀释法沿青藏高原东西水分梯度测定豆科植物共生固氮量；2）观测固氮功能对外源氮增加的响应曲线，发展DyN模式中生物固氮过程对外源氮增加的反馈。为正确评估(高寒草原)生物固氮量提供科学依据，并为地球系统模式中生物固氮的模拟提供新方法。

为了正确评估青藏高原高寒草原的生物固氮速率及其对大气氮沉降增加的响应，并为地球系统模式发展有效的生物固氮模型，本研究 （1）依托青藏高原纳木错站，应用氮同位素稀释法，观测了豆科植物伴生群落和豆科植物建群群落的生物固氮速率，发现其变化范围在0.86-2.77 g N m-2 yr-1,固氮速率主要取决于豆科植物生物量。多梯度的模拟氮沉降增加实验表明，豆科植物固氮功能，%Ndfa值在0.8 g N m-2 施氮水平下开始减弱，而豆科植物生物量在2 g N m-2 yr-1的梯度下急剧下降；（2）延高寒草原从东到西观测了86个样地的生物固氮量，固氮速率范围在0-5 g N m-2 yr-1之间；（3）应用上述观测校正了基于物质平衡原理的潜在固氮速率模型，在0.1°分变率的气候资料驱动下模拟了青藏高原潜在固氮速率空间分布格局，从东南向西北递减，平均固氮速率为1.4±1.5 g N m-2 yr-1, 总固氮量约为 3.3 Tg N yr-1。其中高寒草原的平均固氮速率为 1.2±1.5 g N m-2 yr-1，总固氮量约为 0.8 TgN yr-1。根据目前观测到的当前青藏高原接受到的大气氮沉降速率范围，0.16-0.32 g N m-2 yr-1，青藏高原高寒草原生物固氮速率尚未达到受大气氮沉降影响的关键转折点。本研究为正确评估高寒草原生物固氮量提供科学依据, 并为地球系统模式中生物固氮的模拟提供新方法。