A weakening nitrogen sensitivity of plant productivity over four decades (1976-2016) in a Tibetan alpine grassland

We conducted a four-level N addition experiment over 2011-2016 in a Tibetan alpine grassland, and collected data of aboveground net primary productivity (ANPP) from other five N addition experiments at the same site but different time periods over 1976-2012. Nitrogen response efficiency (ANPP response per unit N addition) generated from these six experiments was quantified to test the N sensitivity of plant productivity. Syntheses of these six experiments revealed that ANPP displayed a linear-plateau response to N addition and saturated at 9.3 g N m^-2yr^-1. N response efficiency was significantly affected by mean annual temperature and monthly precipitation at the key growing stage (i.e., leaf-out phase in April and fast-growing phase in June), and decreased over the past four decades by a rate of -0.68 g g^-1 N yr^-1. Such a decrease in N response efficiency was likely due to the consequences of improving soil nutrient availability, aggravating water limitation and shifting plant species composition along with global change. Our study highlighted that temporal shifting in ecosystem processes and driving factors on plant productivity should be considered in modelling to project carbon and N cycling under global change scenarios.Fig. 1 Responses of aboveground net primary productivity (ANPP, a) and N response efficiency (NRE, b) to N addition in our own manipulative experiment over 2011-2016.Fig. 2 Responses of aboveground net primary productivity (ANPP) to N addition (a) and changes in N response efficiency (NRE, b) across the six N addition experiments spanning from 1976 to 2016 at the Haibei station.<br>Fig. 3 A closing ANPP gap (the grey area) between the ambient (ANPP_ambient) and upper boundary of varied N additions (ANPP_upper) over 1976-2016 in a Tibetan alpine grassland. The solid line indicates a significant exponential increase in ANPP_ambient and the dash line represents a stable level of ANPP_upper. ANPP_upper, the top 10 percentile of ANPP_N observed under varied N additions (Fig. S2); and ANPP_ambient, ANPP from the control of the six N addition experiments and the long-term monitoring study.<br>Fig. 4 Changes in climate, soil nutrients and plant functional group composition over 1976-2016 in a Tibetan alpine grassland. (a-c) MAT (mean annual temperature) and MAP (mean annual precipitation), and humidity index; (d-f) soil total N, Olsen-P and pH in the 0-10 cm; (g-i) relative abundance of grass, sedge and forb in plant community biomass. <br>

本研究于2011-2016年在青藏高原高寒草地开展了4个梯度的氮添加实验，并收集了1976-2012年间同一站点、不同时期开展的另外5项氮添加实验的地上净初级生产力（aboveground net primary productivity, ANPP）数据。基于上述6项实验，本研究量化了氮响应效率（nitrogen response efficiency, NRE，即单位氮添加量对应的ANPP响应值），以探究植物生产力的氮敏感性。
对6项实验的综合分析表明，ANPP对氮添加呈现线性-平台型响应，且在9.3 g N m⁻² yr⁻¹时达到饱和。氮响应效率显著受关键生育期（即4月展叶期与6月快速生长期）的年平均气温与月降水量影响，并在过去40年间以-0.68 g g⁻¹ N yr⁻¹的速率下降。
氮响应效率的这种下降，可能与全球变化背景下土壤养分有效性提升、水分限制加剧以及植物物种组成改变有关。本研究强调，在全球变化情景下预测碳氮循环的模型中，应纳入生态系统过程与植物生产力驱动因子的时间动态变化。
图1 2011-2016年本研究开展的控制性氮添加实验中，地上净初级生产力（ANPP，图a）与氮响应效率（NRE，图b）对氮添加的响应。
图2 海北站1976-2016年间6项氮添加实验中，地上净初级生产力（ANPP）对氮添加的响应（图a）及氮响应效率（NRE）的变化（图b）。
图3 1976-2016年青藏高原高寒草地中，环境条件下地上净初级生产力（ANPP_ambient）与不同氮添加梯度下的地上净初级生产力上限（ANPP_upper）之间的收窄ANPP差值（灰色区域）。实线代表ANPP_ambient的显著指数增长趋势，虚线代表ANPP_upper的稳定水平。其中，ANPP_upper为不同氮添加处理下观测到的ANPP_N的前10%分位值（图S2）；ANPP_ambient为6项氮添加实验的对照组与长期监测研究中的ANPP值。
图4 1976-2016年青藏高原高寒草地的气候、土壤养分与植物功能群组成变化。(a-c) 年平均气温（mean annual temperature, MAT）、年平均降水量（mean annual precipitation, MAP）与湿度指数；(d-f) 0-10 cm土层的土壤全氮、Olsen磷与pH值；(g-i) 植物群落生物量中禾本科、莎草科与杂类草的相对丰度。