Ecosystem primary productivity is a key ecological process influencing many ecosystem services, including carbon storage. Thus, clarifying how primary productivity in terrestrial ecosystems responds to climatic variability can reveal key mechanisms that will drive future changes in the global carbon budget. Satellite products of canopy greenness are widely used as proxies for vegetation productivity to evaluate how ecosystems respond to climate variability. However, to what degree inter-annual variations in productivity are consistent with greenness and how this relationship varies spatially remains unclear. Here we investigated the strength of the coupling between inter-annual variations in leaf area index (LAI, a measure of greenness) and ecosystem gross primary productivity (GPP) derived from eddy covariance towers, i.e., the r² of the LAI-GPP relationship. Overall, inter-annual GPP and LAI were highly coupled (i.e., high r²) in arid grasslands, but were fully decoupled in mesic evergreen broadleaf forests, indicating that this relationship varies strongly along aridity gradients. A possible mechanism of the spatial variation in the LAI-GPP relationship is that the tradeoff between ecosystem structure (LAI) and physiology (photosynthesis per unit leaf area) becomes stronger in more humid climates. Land models overestimated the r² of LAI-GPP correlation for most ecosystem types and failed to capture the spatial pattern along aridity gradients. We conclude that relying on greenness products for evaluating inter-annual changes in vegetation productivity may bias assessments, especially in tropical rainforest ecosystems. Our findings may also reconcile observed disparities between responses in greenness and GPP during drought in Amazon forests.

生态系统初级生产力是影响包括碳储存在内的许多生态系统服务的关键生态过程。因此，阐明陆地生态系统的初级生产力如何响应气候变化可以揭示推动未来全球碳预算变化的关键机制。冠层绿度的卫星产品被广泛用作植被生产力的代表，以评估生态系统如何应对气候变化。然而，生产力的年际变化在多大程度上与绿色一致，以及这种关系在空间上如何变化仍不清楚。在这里，我们研究了叶面积指数（LAI，一种绿色度的度量）的年际变化与来自涡流协方差塔的生态系统总初级生产力（GPP）之间的耦合强度，即r ² LAI-GPP 关系。总体而言，年际 GPP 和 LAI 在干旱草原中高度耦合（即高r ²），但在中干常绿阔叶林中完全解耦，表明这种关系随干旱梯度变化很大。LAI-GPP 关系空间变化的一个可能机制是生态系统结构 (LAI) 和生理学（每单位叶面积的光合作用）之间的权衡在更潮湿的气候中变得更强。土地模型高估了r ²大多数生态系统类型的 LAI-GPP 相关性，未能捕捉沿干旱梯度的空间格局。我们得出结论，依靠绿色产品来评估植被生产力的年际变化可能会使评估产生偏差，尤其是在热带雨林生态系统中。我们的研究结果也可能调和在亚马逊森林干旱期间观察到的绿色和 GPP 响应之间的差异。