Scientists Redraw the World’s Biomes Using Climate Alone, From Tundra to Tropical Rainforest

Every map of the world’s vegetation that you have ever seen rests on a quiet act of translation: converting the ceaseless churn of temperature, rainfall and seasonality into a finite set of recognizable living landscapes. A new study published in Vegetation Classification and Survey takes that translation further than most, proposing a complete hierarchical classification of the planet’s terrestrial biotic units that is anchored, from its largest category down to its smallest, in explicitly defined climatic envelopes. Drawing on data from 616 localities distributed across every continent, the authors distinguish four great domains, seven ecozones, nine biomes and twenty subbiomes, and then match them to a parallel bioclimatic scheme of four macrobioclimates, ten bioclimates, seven ombrotypes and seven thermotypes. The result is an ambitious attempt to do for the whole planet what regional phytosociologists have long done for single mountain ranges: to say precisely which climate produces which kind of nature, and where the boundaries between them lie.

The intellectual core of the paper is a definitional argument. The term biome, coined in the early twentieth century, has drifted badly, the authors contend, and they identify two specific ways in which modern usage has corrupted it. The first is the regional drift, in which units tied to a single continent or even a single genus are promoted to global rank, such as forests defined by the southern-hemisphere tree genus Nothofagus or grasslands named after the Argentine Pampa. The second is the dynamism drift, in which transient, human-created states, including an “intensive land-use biome,” are treated as natural units of the same standing as the boreal taiga. Both drifts, the authors argue, break the promise that makes the biome concept useful: that a biome is a zonal unit, determined primarily by climate, that can recur on several continents with entirely different floras. A single biome may be dominated by conifers in one region and by broadleaved evergreens in another, but the evergreen, oceanic character that defines it remains constant.

To keep the concept clean, the authors strip human influence out of the definition entirely, and their reasoning is more subtle than a simple preference for pristine nature. Human impact is modern, having become globally significant only since the Neolithic; it is selective, concentrating on fertile, young, often-disturbed landscapes while sparing old, infertile ones; and it is constantly changing as technology advances, making any snapshot of land use obsolete within decades. A biome defined by current land use would therefore be a moving target. Instead, the authors anchor each biome in the potential natural vegetation, the theoretical natural state that would prevail without anthropic disturbance. This choice has a practical payoff: by comparing any real landscape with its corresponding biome, one gains a yardstick for measuring the degree and type of human alteration, in the same spirit that potential natural vegetation has long been used in European vegetation science.

The classification itself is built on a four-level hierarchy. At the top sit the domains, defined by the broadest climatic governance: the cryocratic domain ruled by cold around the poles and high mountains, the mesocratic domain ruled by alternating warm and cold seasons in the intermediate belt, the xerocratic domain ruled by aridity in the subtropics, and the thermocratic domain ruled by perpetual warmth between the tropics. Each domain splits into ecozones characterized by seasonality of rainfall or temperature, ecozones into biomes defined by the physiognomy of the zonal potential natural vegetation, and biomes into twenty subbiomes resolved at higher climatic resolution. Most of these units are intercontinental by design. The tundra, the boreal forest, the temperate deciduous forests, the pluvial evergreen forests, the Mediterranean-type sclerophyll scrublands, the steppes, the deserts, the tropical pluviseasonal forests and the tropical rainforests each appear wherever their climatic envelope is met, whether the local flora is Holarctic, Gondwanan or something in between.

The empirical backbone is a principal component analysis of climatic parameters extracted from the CHELSA database for the 616 selected localities, supplemented with derived indices such as the number of frost days, the drought of the warm season, positive temperature, positive precipitation and the ombrothermic index. The first ordination axis, accounting for roughly forty percent of the variance, tracks pluviometric parameters, while the second, near thirty percent, tracks thermometric ones. On these axes the four domains separate cleanly, and the biomes of the cold, arid and tropical domains occupy distinct peripheral positions, while the mesocratic biomes overlap in the center, reflecting their shared intermediate position. The mountain variants of the cold biomes, such as the alpine tundras of temperate and tropical ranges, overlap with the mesocratic steppic biome, a statistical confirmation of the old observation that altitude compresses latitude.

Beneath the biome level, the paper revives and formalizes the paired concepts of thermotypes and ombrotypes to capture variation that a single climate type cannot. Ombrotypes, seven of them from hyper-arid to hyper-humid, are calculated as the quotient of positive precipitation to positive temperature, quantifying water availability across regions shaped by orographic rainfall and rain shadow. Thermotypes, from infra through thermo, meso, supra and oro to cryoro, are defined by positive temperature thresholds that shift with latitude, and they describe the thermal belts that stack vertically on every mountain. The data reveal a striking latitudinal rule: the warmer thermotypes vanish poleward in sequence, with the infra type disappearing above roughly thirty degrees of latitude, the thermo above forty, the meso above fifty, the supra above sixty and the oro above seventy, until only ice remains. The altitudinal limits of these belts also rise with continentality, so that the alpine zone begins far higher in the continental Rocky Mountains than at comparable latitudes in the oceanic Sierra Nevada of Spain.

Perhaps the most consequential finding concerns the profound asymmetry between the hemispheres. The southern hemisphere, the analysis shows, is substantially more oceanic than the northern, a direct consequence of how little land lies in its extratropical latitudes and how strongly land masses modify the flow of air and marine currents. The consequences ripple through the entire classification. One biome, the steppe, and one subbiome, the continental scrub and woodlands, are exclusive to the northern hemisphere, because both require the extreme continentality that only the vast Eurasian and North American interiors provide. Two others, the boreal forest and the temperate deciduous forests, are almost confined to it, surviving in the south only as narrow strips of deciduous southern beech in the Andes and a fragment in western Tasmania. Conversely, the pluvial evergreen forest biome, which thrives under oceanic conditions, spreads far more widely across the southern lands, clothing southern Chile, New Zealand, Tasmania and the South African coast.

The second asymmetry is longitudinal. The western sides of the continents, exposed to the descending arms of the subtropical high-pressure cells, host the deserts and the summer-dry, Mediterranean-type biome, which are largely absent from eastern seaboards. These drought-driven units then expand deep into continental interiors, most spectacularly across Eurasia, where the sheer size of the land mass generates rain shadows and continentality effects that push aridity thousands of kilometers from any ocean. Only Patagonia breaks the southern-hemisphere pattern, its shrublands owing their summer dryness to the rain shadow of the southern Andes combined with the westerlies, producing a cool, dry, oceanically moderated unit that the authors recognize as a subbiome of its own.

Methodologically, the paper makes a case for how such classifications should be built in the era of high-resolution global climate data. Rather than defining a small number of climatic types and then asking what vegetation happens to fall inside them, the authors work in the opposite direction, using large multidimensional parameter sets so that each biotic unit is characterized by its own climatic envelope in vector space. They are candid about the limits of the exercise: the boundaries between biotic units are transitional rather than sharp, because the units themselves grade into one another along climatic gradients. What the framework offers instead is a transparent, reproducible correspondence between climate and life at planetary scale, a modern heir to Humboldt’s isothermal lines and Troll’s mean continent, in which every mountain belt, every continental profile and every hemisphere’s peculiarities can be read as expressions of the same underlying climatic logic.

Subject of Research: Climatic classification of the world's terrestrial biomes and bioclimatic units

Article Title: Climatic definitions of the world’s terrestrial biomes

Article References: Loidi, J., Navarro-Sánchez, G., & Vynokurov, D. (2022). Climatic definitions of the world’s terrestrial biomes. Vegetation Classification and Survey, 3, 231-271. https://doi.org/10.3897/vcs.86102

Image Credits: AI Generated

DOI: 10.3897/VCS.86102

Keywords: biomes, bioclimatology, vegetation classification, macrobioclimates, thermotypes, ombrotypes, CHELSA climate data, principal component analysis, continentality, altitudinal zonation, potential natural vegetation, hemispheric asymmetry