落叶松（Piñon pines）适应干旱环境，但长期干旱仍会导致其死亡。一组落叶松被剥夺水分，并在实验室环境中经历了数月。然而，若将温度仅升高几度，情况会迅速恶化。亨利·亚当斯（Henry Adams）及其同事在2009年发表于《美国国家科学院院刊》（Proceedings of the National Academy of Sciences）的一篇论文中量化了这一现象。
Piñon pines are adapted to dry conditions, but prolonged drought can still kill them. A group of piñons was deprived of water and left to endure several months in a laboratory setting. Raise the temperature by just a couple of degrees, however, and things go downhill rather quickly. Henry Adams and his colleagues quantified this phenomenon in a 2009 paper in the proceedings of the National Academy of Sciences.
二十株成熟的落叶松从新墨西哥州的一处牧场被转移至亚利桑那州的一个气候可控生物圈中。其中十株被置于接近其自然温度的条件下，另外十株则被置于比其自然条件高出约4.3摄氏度的环境中。每组中各有五株树木被停止供水。根据该研究，温暖环境中未浇水的树木在平均18周后死亡，而常温环境中的树木平均存活25.1周，存活时间缩短了28%。为何高温会让树木“挨饿”？
Twenty mature piñon pines were transferred from a ranch in New Mexico to a climate-controlled biosphere in Arizona. Ten trees were kept in conditions similar to their natural temperature, and the other 10 were placed in an environment that was approximately 4.3 degrees Celsius warmer than their natural condition. Water was withheld from five trees from each of these groups. According to the study, the unwatered trees in the warm room died after an average of 18 weeks, compared with 25.1 weeks for those in the normal room, a 28% shorter survival time. Why did heat leave the trees starving?
研究人员探究了高温缩短树木存活时间的原因。干旱主要通过两种方式杀死树木。第一种是树木关闭叶片气孔，停止制造食物，并耗尽所有储存能量。第二种是树木内部水流被气泡阻塞。由于高温下储存能量消耗更快，只有第一种机制取决于温度。因为两组树木的水分胁迫指标相同，没有理由认为树木发生了水力阻塞。两组未浇水树木的光合作用均迅速下降，到第三周已接近零。呼吸速率在第三、四周开始出现差异。温暖环境中的树木消耗储存碳储备的速度更快。
The researchers examined what caused higher temperatures to shorten the trees' survival. Drought can kill a tree in two main ways. The first is that the tree can close up its leaf openings, not make its food anymore, and use up all its stored energy. The second is that it can be blocked by air bubbles from the flow of water inside. Because stored energy is depleted more quickly at higher temperatures, only the first mechanism would depend on temperature. As the measures of water stress were the same in both groups, there was no reason to think the trees had become hydraulically blocked. Photosynthesis declined rapidly in both groups of unwatered trees, approaching zero by the third week. Respiration rates began to diverge during the third and fourth weeks. The warmer trees consumed their stored carbon reserves faster.
《西南气候展望》（Southwest Climate Outlook）2010年4月刊登的一篇对亨利·亚当斯的采访（由CLIMAS发布）提供了一个比喻来解释这些发现。
An April 2010 interview with Southwest Climate Outlook, published by CLIMAS, offered a comparison to explain the findings.
Adams 将两名马拉松运动员作比较，他们在各自的比赛前吃了同样的餐食，一名在芝加哥参赛，另一名则在气温达 100 华氏度的图森参赛。在高温下奔跑的马拉松运动员会更早耗尽能量储备。停止吸收二氧化碳的树木仍在消耗储存的能量。因此，较热的树木比较凉爽的树木更早耗尽能量储备。更热的干旱会给森林带来什么影响？
Adams compared two marathoners who had eaten the same meal before their respective races, one in Chicago and the other in Tucson in 100-degree Fahrenheit heat. The marathoner running in high temperatures exhausts his energy reserves earlier. Trees that stop taking in carbon dioxide continue consuming stored energy. As a result, warmer trees deplete their energy reserves sooner than cooler ones. What can hotter droughts mean for the forest?
Adams 及其同事还回顾了历史。研究人员分析了美国西南部四个州 103 年的区域气候记录，发现 2002 年的森林枯死事件与一次持续 26.1 周的干旱同时发生，这是他们数据集中唯一一次被识别出的为期六个月的区域性干旱。由于高温使存活时间缩短了 28%，持续 18.7 周的干旱也可能产生类似的毁灭性影响，而记录显示至少发生过五次此类干旱。
Adams and his colleagues also looked at history. The researchers analyzed 103 years of regional climate records from four southwestern states and found that the 2002 forest die-off coincided with a 26.1-week drought, the only six-month regional drought identified in their dataset. With survival time reduced by 28% in the heat, a drought lasting 18.7 weeks could have similarly devastating effects, and the records show at least five such droughts.
根据他们的模型，作者预测，仅温度升高就可能使足以导致森林枯死的严重干旱发生频率增加五倍。他们将此视为一个保守预测，因为该预测未考虑干旱频率增加或树皮甲虫活动加剧的因素。他们还提醒，实验仅测试了一个树种。
Based on their model, the authors projected that temperature increases alone could make droughts severe enough to cause forest dieback five times more frequent. They considered this a conservative projection because it did not account for increases in drought frequency or bark beetle activity. They also cautioned that the experiment tested only one species.
CLIMAS 访谈解释了实地面临的利害关系。访谈称，2002-2003 年的枯死事件波及四角地区约 4500 平方英里的油松林，留下干枯的树林，增加了火灾风险，并减少了吸引游客的秋季景色。访谈还讨论了不列颠哥伦比亚省另一次枯死事件，覆盖面积约 50200 平方英里，给木材业带来压力，并将储存的碳释放到大气中。Adams 补充说，人们依赖森林吸收二氧化碳，因此大规模枯死可能会削弱这一作用。Adams 对该研究的局限性持公开透明态度。
The CLIMAS interview explains what is at stake on the ground. It says the 2002-2003 die-off hit about 4,500 square miles of piñon forest in the Four Corners region, leaving dry groves that raise fire risk and reduce the fall color that draws tourists. The interview also discussed a separate die-off in British Columbia that covered about 50,200 square miles and had placed pressure on the timber industry while releasing stored carbon into the atmosphere. Adams added that people rely on forests to absorb carbon dioxide, so large die-offs could undermine that role. Adams was transparent about the limitations of the study.
由于实验中的树木是在容器中种植的，他警告不要将存活率直接推广到野生树木，后者可能存活更长时间。研究团队后来在亚利桑那州弗拉格斯塔夫复制了该实验，方法是将单叶松移至更温暖的地点，并使用防水布遮挡雨水。
Because the trees were grown in containers for the experiment, he cautioned against extending the survival rates directly to wild trees, which would likely survive longer. The research team later replicated the experiment in Flagstaff, Arizona, by moving the piñons to a warmer site and using tarps to block the rain.