The vegetation of deserts and arid environments

A valley of bare rock and sand, almost free of snow, stretches towards the horizon. — The Labyrinth, in Wright Valley in the McMurdo Dry Valleys, Antarctica, looking east (photograph from November 2013). Dturme, CC BY-SA 4.0, source

Antarctica is a desert. Antarctica is not a desert.

Both sentences are true: they simply do not refer to the same place — and this is the best illustration of how scientific categories are drawn.

What defines a desert

Not sand, not heat, not dunes.

It is the lack of water.

The threshold commonly used by climatologists is about 250 millimetres of precipitation per year. Below it, the term is a true desert; between 250 and 500 mm, a semi-arid or semi-desert environment.¹

Essential note: there is no single or universal definition of desert. This is not a cautious caveat: it is a fact accepted by all reference works.

Antarctica, from two angles

If the 250 mm threshold is applied:

The interior of the continent receives less than 50 millimetres per year (water equivalent), almost entirely in the form of snow. That is more than five times less than the threshold: it is a hyper-arid desert.
The coastal fringe receives up to about 400 millimetres in some places. That is well above the threshold: the edge is therefore not a desert.²

So is Antarctica a desert? The centre, yes. The edges, no.

What changes is not the nature of the continent, but where one decides to look and to set the boundary.

Why a single figure is not enough

Take two geographical areas that both receive less than 250 millimetres of water per year:

Phoenix, in Arizona: classed without hesitation among the deserts.
The North Slope of the Brooks Range, in Alaska: generally excluded from desert classifications.

Same rainfall, two distinct categories. Why?

Because the real water balance compares rainfall with evapotranspiration, that is, with the water demanded by plants and the air. In Alaska, the extreme cold prevents water from evaporating: the little moisture available remains stored in the soil. In Arizona, the heat immediately dissipates what little precipitation falls.³

A figure for the water that falls is not enough: one has to look at what the ecosystem does with it.

How a plant survives in a desert

A desert plant faces two difficulties at the same time: finding water, and not losing it. Yet the very mechanism by which a plant feeds makes it transpire. To capture carbon dioxide from the air, it must open tiny pores on the surface of its leaves, the stomata — and through these same pores, its water vapour escapes. Feeding and drying out are one and the same act.⁴

There are three strategies, very different from one another.

Storing. The cactus, the aloe and the agave thicken their stems or leaves into reservoirs — they are called succulents. The compact, often spherical, shape is not decorative at all: it is the one that offers the least surface for the most volume, and therefore the least evaporation. The spines are leaves reduced to almost nothing, which no longer transpire, protect against herbivores and slow the wind right at the surface of the plant.⁵

Many of these plants have also reversed their timetable. Often, they open their stomata only at night, when the air is cool and humid, store the carbon dioxide in the form of malic acid, and use it the next day with the shutters closed. This process has a name, CAM metabolism — and it changes not the place of the exchange, but its timing. It is slow — many of these plants grow slowly — but no land plant uses water as efficiently.⁶

Seeking it out. Others store nothing and rely on their roots. Some deploy a very extensive network at shallow depth: the roots of a saguaro spread out about as far as the cactus is tall, but rarely go deeper than ten centimetres. They absorb the slightest shower within a few hours, before it evaporates. Others, like certain acacias, send a single root down tens of metres to the water table. These two tactics are mutually exclusive: it is impossible to be everywhere at once.⁷

Waiting. The third strategy is simply not to be there. Many desert plants spend most of their time as a seed, in a dormancy that can last for years. Sufficient rain is needed — and some seeds germinate only above a precise rainfall threshold — for them to sprout, flower and produce seeds again within a few weeks. This explains the spectacular blooms that sometimes follow a storm in a desert that looks dead.⁸

These three responses have nothing in common, and yet two unrelated plants often end up with the same form. That is the subject of the next section.

Evolutionary convergence: two plants, one answer

A North American cactus and an African cactus-like euphorbia, placed side by side, show the same silhouette: a green, fleshy mass, full of water, leafless and covered in spines.

Yet they are not related. Their lineages separated a very long time ago, and an ocean has kept them isolated for a long time.

They did not inherit these traits from a common ancestor: they independently solved the same problem — surviving drought under extreme sunlight — by developing the same structures. This is the phenomenon of evolutionary convergence.

More remarkable still: a simultaneous evolution

Cacti are often assumed to be very ancient. They are not. Their lineage separated from its closest relatives about thirty-five million years ago, and the groups that have the most species today only really diversified between ten and five million years ago.⁹

Here is the surprising result. The major diversifications of the cacti of the Americas, the ice plants of South Africa and the agaves of North America are contemporaneous.¹⁰

Three continents. Unrelated lineages. The same moment.

This is not a coincidence, but the trace of a common cause, on a global scale — a cooling and drying of the world’s climate at that time. Plants did not invent succulence whenever they could. They invented it when the world demanded it, and everywhere at once.

How to tell them apart with certainty

There is an infallible anatomical criterion, visible to the naked eye: the areoles.

These are small felted or downy cushions on the surface of the stem, from which the spines emerge. They are exclusive to cacti (Cactaceae); no other plant family has them.¹¹

A downy cushion at the base of the spines → It is a cactus. Its spines are modified leaves, borne on a very short shoot.
Spines emerging directly from the stem (often in pairs) → It is a euphorbia. Its spines are modified stipules or shoots.

From a distance, the strategy looks the same. Up close, the organs used are entirely different.

⚠ Safety warning: the sap must never be tested
One sometimes reads that it is enough to cut the plant to see whether a white liquid flows out. This must be avoided at all costs. The latex of euphorbias is toxic and can be extremely irritating. Mere contact followed by rubbing the eye can cause severe keratitis or temporary loss of sight. The visual criterion of the areoles is more than enough.¹²

The geographical exception

Cacti are strictly native to the Americas, with one exception: Rhipsalis baccifera, found naturally in tropical Africa and Madagascar. Carried by migratory birds or transported by early ships, it illustrates the rule that biology always leaves a margin for exceptions.¹³

The safeguards: desertification and greening

Two contemporary subjects are frequently oversimplified when the dynamics of arid environments are discussed:

1. Desertification: a debated notion

Desertification does not refer to the natural advance of existing deserts, but to land degradation in arid, semi-arid and dry sub-humid areas, mainly as a result of human activities and climatic variations.¹⁴

The global extent and severity of this process are the subject of major methodological debates among researchers: the boundaries between multi-year rainfall fluctuation (temporary drought) and irreversible degradation of plant cover remain difficult to draw from satellite imagery alone.¹⁵

2. Greening projects and “great green walls”

Many initiatives seek to halt aridification through massive reforestation (such as the Great Green Wall in the Sahel). While the stated aim is to stabilise soils and support agriculture, ecologists point out several conditions for effectiveness:

Planting non-native monocultures often consumes more groundwater than it regenerates.
One of the finest successes comes from the natural regeneration of trees led by the farmers themselves (FMNR), as in Niger, with local agreements between farmers and herders on livestock routes.¹⁶

An inhabited and shaped environment

A desert is not an empty or dead space.

Arid zones are inhabited by specialised flora and fauna, as well as by human societies that have spent millennia developing sophisticated techniques for collecting and conserving water (qanats, foggaras, terraces, impluvia).

Describing these environments as virgin or sterile lands masks the technical history of local populations and the fine adaptation of plant species to scarcity.

Note 1

There are almost as many definitions of deserts and classification systems as there are deserts in the world. […] In this now widely accepted system, extremely arid lands have at least 12 consecutive months without rainfall, arid lands have less than 250 millimeters of annual rainfall, and semiarid lands have a mean annual precipitation of between 250 and 500 millimeters.

U.S. Geological Survey, “What Is a Desert?”, in Deserts: Geology and Resources, USGS, page last modified 18 December 2001