When the nations of the world ratified the Paris Agreement, they adopted two different goals for how much they hoped they could limit climate change. While pledging to keep global warming “well below” 2 degrees Celsius, they also said they would try to “limit the temperature increase to 1.5 degrees.”

It was a hopeful and ambitious moment in climate diplomacy—and one not yet matched by policy. The only way to keep warming that minimal is to cut emissions much faster than almost any country is currently planning. Many scientists and policymakers consider hitting the 2-degree goal extremely difficult, and the 1.5-degree goal all but impossible given today’s technology.

A study published this week in Science emphasizes the importance of both goals. Unless global temperatures are kept below 2 degrees, the biomes of the Mediterranean Basin will change more by the year 2100 than they have in the past 10,000 years.

Biomes are ecological regions that share similar flora, fauna, and weather. Desert, deciduous forest, and Alpine meadow are all types of biome. The study finds that many regions of the Mediterranean will undergo such stress that—under all but the kindest warming scenarios—biomes are likely to significantly change in the next century, as forests become scrubland and scrubland becomes desert.

“Under the [business-as-usual] scenario, all of southern Spain turns into desert, deciduous forests invade most of the mountains, and Mediterranean vegetation replaces most of the deciduous forests in a large part of the Mediterranean basin,” write the study’s authors.

The paper also begins to fill in a hole in available climate science: how the newer, leaner 1.5-degree goal improves on the 2-degree goal in different regions of the world. (Most biome models were built around the older 2-degree warming aim.) It finds that 1.5-degree averts most of the disruptive biome change.

“The difference between global warming of 1.5 degrees Celsius and 2 degrees Celsius—just a small 0.5 degrees—is significant and not harmless,” said Joel Guiot, a paleoclimatologist at Aix-Marseille University and one of the authors of the study, in an email. “It is really important to limit the warming to much less [than] 2 degrees to avoid changes not experienced during the last 10,000 years.”

The study, by Guiot and his colleague Wolfgang Cramer, is the first to run the Paris Agreement’s temperature goals through BIOME4, a global model of biome geography. BIOME4 is not the most state-of-the-art biome model, but it is the fastest—it doesn’t require months of data-center or supercomputer time. First, Guiot and Cramer used the model to estimate the next century of biome change in Europe, given various warming pathways for the planet. Then they compared it to model runs for the past 10,000 years. Those runs were restricted by knowledge gained from pollen sediment core data—evidence of old biomes trapped in sedimentary rock.

Guiot and Cramer’s maps of the Mediterranean Basin under various warming scenarios. The map labeled D shows the approximate 1.5-degree scenario, figure E the 2-degree scenario, figure F the 3-degree scenario, and figure G the business-as-usual scenario, where the atmosphere heats to more than 5 degrees Celsius above the pre-industrial average. Figure B shows one of most anomalous regional moments of the past 10,000 years, when Europe was colder than normal.  

Even under the “moderate” warming scenarios, the model estimates that Mediterranean scrubland will expand and Alpine forests will recede.

Simon Goring, a scientist at the University of Wisconsin, said that the paper convincingly argued that climate change will trigger biome shifts, and not just change how biomes work within their current regions.

“We know there have been short-term changes in climate, but this [21st-century warming] is outside the envelope of any of the current forest biomes in the region,” he told me. This especially applies in the Balkans, where cool mixed forest is forecast to get much warmer. “We’re pushing everything out of that boundary, because the forest can’t tolerate them.”

Part of that change is that biome changes will happen for different reasons in the century to come than they have in the Holocene. During the past epoch, biomes tended to shift because of changing precipitation patterns due to the changing location of the intertropical convergence zone. Temperature was relatively stable. In the years to come, biome change—and precipitation—will be driven by rising temperatures.

Jack Williams, a paleoclimatologist at the University of Wisconsin, said the study’s findings were important though preliminary. “Their takeaway message is that 1.5 degrees of warming, and the ecosystem changes that result, is at the upper end of what you see during the Holocene. And as you go beyond, 2 degrees above, 3 degrees above, that’s above what we’ve seen at the Holocene,” he told me. “That’s a message I would agree with, and it makes sense given the data.”

But BIOME4, the model used, did have some limits. Specifically, it doesn’t examine changes to the environment within biomes—such as whether deciduous forests, while remaining deciduous, could become drier or host a different mix of species. Forests don't have to become desert to fundamentally change—new blights, rain patterns, and invasive species can devastate them, leaving the same forest with many new trees.

These more difficult modeling problems are some of the challenges researchers will face as they try to understand the benefits of the 1.5-degree goal—before it becomes impossible.