McDowell hypothesized that drought could kill trees either through thirst or starvation, and that owing to their different coping strategies, juniper would die of thirst while piñon would starve. Since the hypothesis is based on fundamental plant biology, and because juniper and piñon manage risk so differently, studying them could reveal basic mechanisms of death that can be tested and tweaked to model mortality elsewhere.

McDowell first tested his hypothesis in a drought experiment in central New Mexico. One set of trees was irrigated, another deprived of water, a third received whatever the sky provided, and all were poked and probed. The piñons in the "droughted" plots, nudged by beetles, perished first, but within a few years junipers, which beetles ignored, died too. Apparently, neither strategy was enough to protect the trees from long-term drought. Rather than perishing of thirst or hunger alone, both species died from some combination of both.

The Frijoles Mesa experiment adds another variable: heat. On the mesa in mid-August, McDowell pried open an acrylic cylinder enclosing a diminutive, maybe 6-foot-tall juniper, and invited me to wedge myself inside. The tree was alive, but had the scrappy look of Charlie Brown's Christmas tree. A fan roared on and off. The air was warm, the experience claustrophobic. After a minute or two, I showed myself out.

The chambers are kept at a consistent 9 degrees Fahrenheit above ambient temperatures, the sort of weather all these trees may have to cope with in the latter half of this century, especially during a drought. The study is young, but McDowell has found evidence of heat's disruptive effects. Elevated temperatures seem to cause both piñon and juniper to devour their carbohydrate reserves more quickly, for instance. "But it's not that clean," he says. During winter, extra warmth can boost photosynthesis.

Tree mortality is a complex and dynamic process. But despite all the remaining questions, a flurry of research over the last five years has helped crystallize an important message, says Breshears. "We have gained a huge amount of confidence that, under warmer conditions, we're going to get a lot more mortality."

In mid-August, I hiked into the Sangre de Cristo Mountains outside Santa Fe with Park Williams, a 32-year-old climatologist who, until recently, worked out of the Los Alamos National Laboratory with McDowell. When we strolled past Hyde Park Lodge, his eyes began to dance. In two weeks, the California native would marry his girlfriend at the lodge. He had proposed to her underneath a coast redwood, his favorite tree.

A half-mile or so up a steep trail, we gained a sweeping view of the fortresses of pine and fir on facing hills. Williams wore a navy trucker hat and aviator sunglasses with blue-blocking lenses. "When I first look at this mountainside, it looks totally green," he remarked, handing me his sunglasses. The lenses made it easier to see red-orange flecks in the blanket of green, like the first autumn leaves snagged in a lawn. "I think we're seeing the beginning of something that in another one or two years will be much more widespread."

Williams hasn't studied this forest, but his offhand prediction has some basis. Last fall, he authored a high-profile study concluding that if climate models' temperature projections are correct, and if carbon emissions remain at current levels, most mature conifers in the Southwest could die by 2050 or soon after. The tall ponderosa haunted by Mexican owls? Mostly gone. The old piñon that produce sweet nuts prized by New Mexicans? For the most part, toast. Douglas fir, the largest conifers native to Arizona and New Mexico? Them, too.

Williams made a convincing – and frightening – case that warmer temperatures alone could kill the trees, even without changes in rain and snowfall. Using tree rings from piñon, ponderosa and Doug fir – the species that occupy the Southwest's warm and dry, and cool and wet niches – Williams created something called a "forest drought stress index." It showed, surprisingly, that drought stress is driven as much by growing season temperatures as winter snowpack.

Drought is not always a problem of scarce rain or snow, though that's how we usually think of it. Hot weather can also impose drought conditions on plants. Minor temperature increases have an outsized effect on the amount of water the atmosphere can hold: When the temperature goes up, the atmosphere gets a lot spongier. The relationship is exponential. Stubbornly set on fulfilling its potential, warm air sucks water more greedily from both plants and soil. If the water supply it's drawing on becomes depleted, the tension begins to strain a tree's water columns. Picture an eager child sucking the last drops of a milkshake from a straw: The water columns, like the straw, collapse. That's bad news for trees, no matter their coping strategy.