发表在新一期英国《Nature Climate Change》杂志上的一项研究认为，尽管地球整体性变暖会使部分农作物减产，但在全球某些地区，大气中二氧化碳浓度升高会减轻这一影响，此外可能有更复杂的因素同时发挥作用。

在新研究中，一个国际研究小组分析从另一项大范围现场实验中获得的农作物模型和数据认为，虽然全球升温和水资源短缺使农作物减产，但考虑到二氧化碳浓度增加会减少农作物对水分的需求，到2080年，玉米、大豆、小麦和大米4种农作物用水效率都将提高。研究预测，如果将所有因素考虑在内，到2080年，在依靠降雨灌溉区域，小麦平均产量将提高10％；在人工灌溉区域，小麦将平均减产4％。而玉米几乎在所有种植区域的产量都将下降，平均下降约8％。研究没有对大豆和大米的产量变化作出结论。

论文主要作者、美国哥伦比亚大学气候系统研究中心环境科学家德尔芬·德里格强调，这项结论并不意味着二氧化碳是“人类的朋友”，只是提醒相关研究人员，在讨论气候变化长期影响时应将二氧化碳浓度对农作物的直接影响计算在内。（来源：新华社 张莹）

Regional disparities in the beneficial effects of rising CO₂ concentrations on crop water productivity

Abstract  Rising atmospheric CO₂ concentrations ([CO₂]) are expected to enhance photosynthesis and reduce crop water use1. However, there is high uncertainty about the global implications of these effects for future crop production and agricultural water requirements under climate change. Here we combine results from networks of field experiments1, 2 and global crop models3 to present a spatially explicit global perspective on crop water productivity (CWP, the ratio of crop yield to evapotranspiration) for wheat, maize, rice and soybean under elevated [CO₂] and associated climate change projected for a high-end greenhouse gas emissions scenario. We find CO₂ effects increase global CWP by 10[0;47]%–27[7;37]% (median[interquartile range] across the model ensemble) by the 2080s depending on crop types, with particularly large increases in arid regions (by up to 48[25;56]% for rainfed wheat). If realized in the fields, the effects of elevated [CO₂] could considerably mitigate global yield losses whilst reducing agricultural consumptive water use (4–17%). We identify regional disparities driven by differences in growing conditions across agro-ecosystems that could have implications for increasing food production without compromising water security. Finally, our results demonstrate the need to expand field experiments and encourage greater consistency in modelling the effects of rising [CO₂] across crop and hydrological modelling communities.

原文链接：http://www.nature.com/nclimate/journal/vaop/ncurrent/full/nclimate2995.html