Tropical forests and their biodiversity

A winding river crosses a vast green forest, seen from an aeroplane. — The Solimões river, in the middle of the Amazon rainforest in Brazil, seen through an aeroplane window: it looks like a great snake, as Indigenous peoples say (photograph by lubasi, 2009). lubasi, CC BY-SA 2.0, source

No, the Amazon does not produce twenty per cent of the planet’s oxygen: three words were lost along the way.

Where the figure comes from

The original figure was correct. In its full form, it reads as follows:

the Amazon supplies around twenty per cent of the oxygen produced by photosynthesis on dry land.¹

On dry land: three words.

Without them, the sentence becomes: twenty per cent of the oxygen, full stop. The oxygen of the entire atmosphere.

It is not a lie, and nobody rigged anything. It is a detail that got lost along the way, as thousands get lost every day, and the figure went on travelling without it.

This site describes elsewhere the case of a figure on biodiversity that lost its source along the way and has been circulating since 2009. It is exactly the same mechanism.

The corrected figure

If the ocean is included, what do the calculations give?

Nine per cent, according to Yadvinder Malhi, an Oxford ecologist who specialises in the Amazon. Six per cent, according to the climate scientist Jonathan Foley.²

A third observes that the two values are equivalent within the uncertainty of the calculation — in other words, they do not really contradict each other.³

So this is where we stand: between six and nine per cent, instead of twenty.

But that is still not the main point.

The fact that changes everything

Trees produce oxygen. They also consume it.

They respire — especially at night, when photosynthesis stops. What they do not use is consumed by the countless micro-organisms that break down dead leaves and fallen wood.

The arithmetic is quick.

The Amazon’s net balance is close to zero.⁴

The details are known: Amazonian plants themselves consume a little over half of the oxygen they produce, and the micro-organisms that break down dead matter take care of the rest.⁵

The researcher who made this calculation adds a striking remark: this holds for the Amazon or for any other biome — and he specifies the timescale, which matters. Over the periods that concern us, that is, less than a few million years, the balance of any ecosystem is close to zero. A forest in equilibrium does not fill the atmosphere with oxygen: it produces it and consumes it, in roughly equal measure.⁶

This point should be kept in mind, because the next section depends entirely on it.

So where does the oxygen we breathe come from?

From photosynthesis, yes — but from a very ancient photosynthesis.

From organisms that died hundreds of millions of years ago, whose matter never rotted and ended up buried under sediments.

This is the key: oxygen builds up in the atmosphere when carbon escapes decomposition.⁷

If everything that grows eventually decomposes, all the oxygen produced is consumed again, and the account returns to zero. Part of it has to be removed from the cycle — buried, compressed, forgotten — for the balance to rise.

The air we breathe is therefore a geological legacy, not current production.

Why there are more species towards the equator

The number of species increases steadily from the poles towards the equator. This holds for trees, birds, insects, and almost every group.

Specialists call this the latitudinal diversity gradient, and a synthesis of nearly six hundred cases has confirmed that it holds very generally.⁸

Nobody knows why.

This is no exaggeration

The reference synthesis, published by some twenty researchers in 2007, opens by noting that how and why this gradient arose remains unresolved. A 2019 review says it again in the same terms: it is one of the most widely studied patterns in all of ecology, and no consensus has been reached about its causes. The disagreement concerns the relative importance of the processes involved, not the existence of the pattern.⁹

The what is solid. The why is not.

This pattern has been measured for a century and a half, and we still do not know what produces it.

Six explanations, no winner

They are all plausible, and none has been ruled out, even if some researchers are betting mainly on time and area.¹⁰

More time — the tropics are geologically older and have had more time to diversify.
More area — tropical environments covered vast expanses for fifty-five million years. A study of more than 12,500 species of woody plants supports this idea: transitions between tropical and temperate environments are rare, and most temperate lineages arose after the global cooling that began 34 million years ago.
More mutations — higher temperature, faster metabolism.
More niches — more ways of making a living.
More specialised enemies — herbivores and diseases would prevent any one species from dominating.
An ecological limit that would set a ceiling on the number of species a place can support.¹¹

Testing the fifth, and what it showed

Researchers set out to check the specialised-enemies hypothesis. The idea to be tested: the more numerous a tree species is in one place, the more its enemies settle there, and the harder it is for its young shoots to survive — which leaves room for the others.

The scale of the work is worth stressing: more than three thousand species, nearly two million four hundred thousand trees, in twenty-four plots spread across the globe.¹²

Result: this effect is stronger in the tropics than in temperate latitudes.¹³

But the most interesting result lies elsewhere. It is not only the intensity that changes with latitude — it is also the species it strikes. In the tropics, the effect strikes rare species more than in temperate forests.¹⁴

If this is confirmed, the consequence is elegant: rare species there would be protected by their own rarity, which would allow them to persist instead of disappearing. A mechanism that, rather than eliminating the weak, prevents the strong from taking everything.¹⁵

The challenge, a year later

Another team replied that this result may have arisen from a methodological artefact — that is, from an effect produced by the method of calculation, not by nature. It adds that several different processes can explain the same observed pattern.¹⁶

The original authors replied in the same issue.

What is at stake here deserves attention. This is not a hypothesis that is proved and then put away: it is a hypothesis that is tested, whose test is challenged, and whose discussion continues. This is the normal state of a live question, and that is why none of the six explanations has won yet.

And the figures are themselves ranges

Among insects, there are about 1 million named species. For all terrestrial arthropods — insects, spiders and their relatives — the total is estimated at 7 million; older estimates went as high as 30 million, a figure now considered very unlikely.¹⁷

Researchers have narrowed the range, but it remains wide: the number of species on Earth is known only approximately.

To end with a concrete example: on a study plot on Barro Colorado Island, in Panama, the first census, completed in 1982, counted approximately 240,000 stems belonging to 303 species of trees and shrubs more than 1 cm in diameter at breast height.¹⁸

The measurement trap: counting felled trees

Of all the questions concerning tropical forests, deforestation is probably the one that generates the most contradictory figures.

This is not because researchers lie. It is because they are not measuring the same thing.

The definition that changes everything

What is a “forest”? The answer varies according to the institution giving it.

The FAO (Food and Agriculture Organization of the United Nations) defines a forest as an area of more than 0.5 hectares with trees more than 5 metres tall and a tree canopy cover of more than 10%.
Some governments set the cover threshold at 30%, others at 70%.
Satellite data measure the loss of tree cover, which covers a clear-cut in primary forest just as much as the harvest of an oil palm or eucalyptus plantation.¹⁹

Result: an area clear-felled and then replanted as a monoculture of fast-growing trees can be counted as a restored forest under some criteria, even though it has lost most of its biodiversity.

Outdated historical data

For decades, assessment relied solely on reports submitted by states, which were often imprecise or politically slanted. The arrival of freely available satellite imagery (such as the Landsat programme) transformed measurement from the 2000s onwards.

But historical comparisons remain tricky: comparing a 1990 figure calculated from administrative reports with a 2026 figure established by high-resolution satellite observation amounts to comparing two different quantities.

As with the oxygen figure, the complexity lies in the definitions. Before debating whether a figure has gone down or up, one must ask: what exactly was counted?

An inversion to avoid

This page has just shown that the Amazon does not produce the oxygen attributed to it.

It should not be concluded that it is unimportant.

This is exactly the misreading this site tries to help avoid, and we come across it everywhere. Correcting an exaggerated figure does not mean the subject was of no interest. It means that that was not the right reason.

We make the same point in our page on the great centres of learning, where correcting a legend in no way diminishes what really happened.

Checking means replacing a bad reason with a good one. It does not mean taking the reason away.

Note 1

Malhi and Coe reckon it stems from the fact that the Amazon contributes around 20 percent of the oxygen produced by photosynthesis on land—which may have erroneously slipped into public knowledge as “20 percent of the oxygen in the atmosphere.”

Zimmer, Katarina, “Why the Amazon Doesn’t Really Produce 20 % of the World’s Oxygen”, National Geographic, 28 August 2019