The desert that creates the rainforest

This is probably the most amazing thing I learned all weekend. The Amazon rainforest—with all its plant and animal life, and all its astounding biodiversity—could not exist as we know it without the patch of African desert pictured above.

The rainforest is amazing, but the soil it produces isn't very nutrient rich. All the minerals and nutrients that fertilize the rainforest have to come from someplace else. Specifically: Africa. Scientists have known for a while that this natural fertilizer is crossing the Atlantic in the form of dust storms, but science writer Colin Schultz ran across a 2006 paper in the journal Environmental Research Letters that not only produces evidence for a much larger trans-oceanic transfer of dust than was previously assumed ... it also pinpoints the exact (and astoundingly small) location where all the fertilizer in the Amazon is coming from.

The research paper, itself, is pleasantly readable, as far as these things go, so I'm going to quote directly from it. One quick note before I launch into this quote. The authors are measuring the mass of the dust in teragrams (or Tg). As you're trying to wrap your head around this, it might be helpful to know that 1 Tg = 1 million tons.

A total of 140 (± 40) Tg is deposited in the Atlantic ocean and 50 (± 15) Tg reach and fertilize the Amazon basin. This is four times an older estimate, explaining a paradox regarding the source of nutrients to the Amazon forest. Swap et al suggested that while the source for minerals and nutrients in the Amazon is the dust from Africa, it was estimated that only 13 Tg of dust per year actually arrive in the Amazon. However, they pointed out that 50 Tg are needed to balance the Amazon nutrient budget.

Here we show a remarkable arrangement in nature in which the mineral dust arriving at the Amazon basin from the Sahara actually originates from a single source of only ~ 0.5% of the size of the Amazon: the Bodélé depression. Located northeast of Lake Chad (17°N, 18°E) near the northern border of the Sahel, it is known to be the most vigorous source for dust over the entire globe.

Basically, these 2006 calculations account for all the fertilization needs of the Amazon, while previous calculations left a weird gap in between the amount of dust the rainforest needed and the amount the scientists thought was getting there.

Also: The place the dust is coming from is a single, highly specific region. As Alexis Madrigal pointed out at The Atlantic, we're talking about a patch of desert only 1/3 the size of Florida supplying the nutrient needs of a jungle that is roughly the same size as all 48 contiguous United States. Mind, blown.

Read the full research paper at Environmental Research Letters

Check out The Atlantic's write up on this, including a satellite photo of the dust storms in question.

Follow the guy who started it all—the very smart, very entertaining, and very tall Colin Schultz

Via Bart King



  1. Forgive me for asking  a potentially silly question… but how come the Sahara provides all those nutrients for the Amazon, and (almost) none for itself?

    1. Who says it doesn’t provide any nutrients for itself? Isn’t the whole point that the Sahara is nutrient rich? Heck, isn’t that why Egypt is/was able to support to support so much agriculture? Simply add water and you’re good to go!

      1.  Duh. Of course. So glad i prefaced my question with a warning. Sometimes my brain just goes on strike.

        1. It wasn’t a silly question, don’t worry about it. I’m sure a lot of people were thinking the same thing.

  2. Not having read the paper and not having the time to do so (i.e, tracking down references, looking for counter arguments) in the near future, I will make the off-the-wall comment that this hypothesis raises red flags.  Taking this from the summary (granted, not the actual paper), “…
     these 2006 calculations account for all the fertilization needs of the Amazon, while previous calculations left a weird gap in between the amount of dust the rainforest needed and the amount the scientists thought was getting there. …”  

    In other words an explanation not based on empirical real data but on a thought experiment.   While thought experiments — “I see this, I wonder how it occurs” — are useful and can be valid they can also be dead wrong.  Until I learn more I am taking this conclusion with a grain of salt

    1. You don’t have time to read the paper, but you do have time to write a criticism about the paper?

      1. Well, yes.   It takes about 5 minutes to write a comment about the red flags that the paper sets up in my mind and to caution the non-scientific readers on BB not to just jump on the bandwagon because the conclusion is so intriguing.    It would take at least 30 minutes — and a cup of coffee — to actually read, understand, and followup references and rebuttals to the paper.  So, yes, I have the time to critique/caution but not to read.

        A tip of the hat to Maggie who, evidently, has read the paper and who says that I should read it.  I’ll try to do so later today.

        1. Or you could read the first two paragraphs of the introduction and see that it is based on satellite observations. You aren’t being asked to referee the paper.

          Satellite observations show continuous dust transport across 5000 km from the Saharan sources to the Caribbean Sea and North America in the Northern summer and to the Amazon basin during the Northern winter [1, 2]. Due to the annual cycle in winds over the Sahara, the winter Saharan dust sources are different from the summer sources [3]. In the summer, dust fluxes reaching the Tropical Atlantic shore originate mainly from the northwest and central-west parts of the Sahara. During winter, strong surface winds (the Harmattan winds) occur along the southern border of the Sahara, activating sources on the border of the Sahel, notably the Bodélé depression in Northern Chad.
          Analysis of satellite data [4] shows that out of the 240 ± 80 Tg (1 Tg = 1012 g = one million tons) of dust transported annually from Africa to the Atlantic ocean, 140 Tg are transported in the summer and 100 Tg in the winter. A total of 140 ± 40 Tg is deposited in the Atlantic ocean and 50 ± 15 Tg reach and fertilize the Amazon basin. This is four times an older estimate, explaining a paradox regarding the source of nutrients to the Amazon forest. Swap et al suggested [5] that while the source for minerals and nutrients in the Amazon is the dust from Africa, it was estimated that only 13 Tg of dust per year actually arrive in the Amazon. However, they pointed out that 50 Tg are needed to balance the Amazon nutrient budget.

          1. No referee, certainly. And I can skim the first couple of paragraphs as well as anyone — they make a claim but do not support it with anything more than, to paraphrase, “we found 50 Tg of dust; ergo this is the stuff needed to make the Amazon nutrients complete.”  

            Now this entire topic is out of my field of expertise so I can not do an immediate evaluation and may be making a snap judgement. But I am being asked to believe that a single source point that is, what, 7000 years old, fuels the entire Amazon basin — something that has been existing for millions of years?  Perhaps I am reacting poorly to the initial line in the BB post … ”
            could not exist as we know it [Amazon] without the patch of African desert pictured above …” and perhaps I am focusing too much on the nutrient claim but I am skeptical. Thus I need to read the followup responses to the 2006 paper.  Has there been more proof?  A rebuttal?   I am not saying that I disbelieve the amount of dust blowing over — I’ll take the satellite interpretation as a given.  I am also not suggesting that the Earth is a bunch of unrelated ecosystems.  Certainly one part of the Earth can affect another. What I am skeptical about is the implication that the Bolele depression is that important to the Amazon. Maybe you can throw me a line and point out supporting research.  I am spending my lunch hour doing some looking into supporting research but have not found any.  Respectable journal citations would be best.  

            Oh, I should say that there are the ‘cited by’ references which I am skimming through but which specific ones I should look at would be helpful.

          1. Antinous.

            In your role as a moderator I would expect more of a learning experience from you.  My post was in no way offensive to anyone — no personal slurs, no swearing, etc.  It brought up to, at least in my mind, a valid question.  It give a link for followup reading.  So why is it “junk”?   

            If you want to see less ‘junk’ on BB then you need to inform us on what is ‘junk’.

          2. Announcing that you haven’t read the paper and then proceeding to critique it is junk. It serves no purpose. It helps no one. It amuses no one. It just wastes bandwidth.

          3. Fair enough suggestion that my comment was junk because I made comments without reading the paper.

            Although I thought that I was raising “red flags” instead of critiquing the paper (it was the second poster – Colin – who initially used the word ‘criticism’ and then I carried the word onward).  As I pointed out it easily takes a half-hour to read a paper.  Much more to track down references to the point where a critique can be made.  However it only 5 minutes to look at the first couple of lines or the summary the paper and write a cautionary note on the forum.  As I know, as we should all know,  not all scientific papers are true in all particulars. I wonder how many of the people who are saying “yes, yes” to the conclusions in the paper — especially the part about the Bodele depression providing most of the nutrients of the Amazon (to paraphrase the part of the argument that I got stuck on) — actually read the paper in detail.   Agreeing to the conclusion without reading the paper is just the flip side of disagreeing with the conclusion without reading the paper.   Both could be considered “junk” comments.

            Note 1:  Comments from knowledgeable environmentalists who did not read the paper in detail should not be considered junk.  Of course it has to tell which of the commenters fall into this category. 

            Note 2:  Most of the paper was not about nutrients in the Amazon but rather the amount of dust.  I had no disagreement with that.  

            Note 3: As with much scientific research we have to accept the experts in that particular field.  It still should not stop people from questioning conclusions especially when done in a respectful manner.  

    2. OK.  Having read the paper and previous (e.g., Swap and post-papers (‘cited-by’ papers; e.g., Bristow, ) to the limit that I can read them — they are out of my direct scientific field (which is genomics) — and not having the time/energy to track down papers that are not on-line (although, being at a university, I can see many journal articles on-line) all I can say is what I have to say to many things outside of my field — “ok, I will accept your word on this.”  

      There is no doubt that there is dust from Africa (and the Bodele depression) being carried to the Amazon.  There is no doubt that the dust contains many nutrients/minerals.  There is no doubt that the Amazon is a nutrient-poor environment.   However I still did not find a direct link that Bodele dust equals Amazon survival.  Perhaps this simply can not be measured or quantified.  Or like many environmental effects the link is hard to prove in a direct manner (I should point out that this also applies to other fields — I am not trying to discount indirect reasoning as a valid method).  As the Swap paper points out, “…the discrepancy [in calculations] may also be due to an overestimation of the necessary nutrient influxes and a simultaneous underestimation of the efficiency in recycling nutrients in this [Amazon] ecosystem.”

  3. It supposedly accounts for half the necessary minerals, not “nutrients”.

    But it really drives home how small, seeming unrelated changes can affect things on a global scale in non-obvious ways. Change the weather pattern in a very specific, relatively tiny desert in Africa, and the Brazilian Rain forest dies.  Scary.

  4. This reminds me of the coolest experience in an ecology class I took in Puerto Rico many years ago:  we were hiking atop a mountain peak on the eastern portion of the isle and examining the stunted (wind shorn) elfin cloud forest there. I climbed into a tree and found clay in the crotches of the branches. How did it get there?!? The only way to the ground was down, and the peak is shrouded in cloud all the time so the ground was wet (i.e. no dust). It was explained to me that I was holding a bit of the Sahara in my fingers.

    1. When I was in Puerto Rico, I was warned about “allergies” to Sahara dust, that the smoggy atmosphere I was seeing on a hot day was due to the dust and that you would get hay fever type symptoms. Didn’t believe it until I experienced the symptoms myself –  bleary watery eyes, dry sinuses, a need to sneeze without sneezing. I knew something was up because I was in a jungle but having a reaction to something as though I were in a field of flowers.

    1. They were both rather nice savannas. The Sahara even supported a largish human population. Broadly and globally (and as I remember it from grad school nigh-on 20 years ago), the climatic history of the last few million years is: deserts and jungles grow together, as do glaciers and savannas. Glacial climate regimes are “equable” and interglacial climate regimes are…not.

  5.  IANA farmer, but whats up with the nutrient profile in the amazonian soil?  More diversity = more nutrients, right?  Are the nutrients blown in from Africa each year getting removed somehow, after uptake & plant death?

    1.  All that rain has the effect of washing away many of the soluble nutrients.  That’s why the soil fertility is so low.

      1. This is half true. All that growth also draws nutrients from the soil; most of the nutrients in the Amazon are above ground. This is why you have to burn down a rainforest before you can farm on it.

  6. I wonder whether this favors epiphytes, plants that grow on other plants rather than in the soil.  If the nutrients fall from above, perhaps there’s that much less benefit to being close to the ground.

    1. I think I’m right in saying that rain forest soils are typically relatively thing and nutrient poor because all the rain leaches them out.

  7. These types of findings highlight how fragile some of the complex ecosystems on our planet are. Point to point relationships like this are easily broken.

    We can talk about cutting carbon emissions and other ways to reduce our impact, but at some point it feels like any impact at all, no matter how small, has the potential to be catastrophic to connections like this.

    I just totally depressed myself. Happy Monday.

    1.  Not to mention the effect “cleaner” energy sources would have.  Consider: what happens to the rainforest if you put a bunch of solar cells in this part of the desert?

  8. “Follow the guy who started it all… Colin Shultz”

    To be totally fair, I think the guy that started it all was Koren et al.

    1. Yes, I thought that very strange.  The article celebrates someone finding the article, rather than the people who wrote the article, as if science is primarily the art of reading through dusty tomes for forgotten knowledge.

  9. I thought the opening sentence, “This is probably the most amazing thing I learned all weekend,” was hilarious, but I can’t figure out if it was intended to be funny or not. My guess is that it’s just a reflection of how much information we winding up taking in while meandering around the internet.

    The “probably” is what really puts it over the top for me.

    1. Maggie has been at an environmental science conference over the weekend (and is/was still there today). Funny that she left out that context – I thought nothing of it because she’s been talking about the conference on Twitter.

  10. Fascinating.  Makes me wonder if LA’s draining of Owens Lake has had similar effects on downwind farms and forests.  The dry Owens Lake bed has become one of the largest sources of particulates in the US, and it’s a desert lake with no outlet, and thus has large accumulations of salts and minerals and dead halophytic algae.

    Since those particulates are viewed as pollution, measures are being taken to reduce the dust.  I wonder if the alteration in fertility downwind will be noticed.

  11. I knew about the dust exchange- and that their climates balance one another as well- but I did NOT know that all of the nutrient rich dust came from one tiny spot.
    That is absolutely fascinating.
    I’ve read that as the Atlantic currents slow and mutate due to warming oceans (as well as changes in land temps and climates), this exchange will likely be compromised, and even potentially reversed to a certain degree. As if the Amazon didn’t have enough problems already.
    Anyway, thanks Maggie- this was a great piece of info I’ve stored away in my mental filing cabinet.

  12. “It comes from this patch of desert, specifically Africa”

    Chad is one country on a 11.7million sq mile continent of 58 countries 900 million people 4000 ethnic groups and… at least 9 deserts depending on how you define them. I guess “specifically, Africa” is the best American geographic knowledge can manage.  

    1. I said “specifically Africa” not because I think Africa is a country (the quote I chose talks about Chad) but because I thought it would be particularly surprising that the nutrients for the Amazon came from a whole separate continent across the ocean. Sorry that wasn’t clear. 

      1. I think this is a perspective thing. Saying something is in “Africa” isn’t terribly helpful in narrowing down a location for people in Africa because its so enormous and diverse. Writing about an “astoundingly small” location, “Specifically, NE Chad” would have been much more helpful and well, actually specific. The way you put it would be like me saying, “Fisherman’s Wharf is a tourist attraction in North America”, or “the Western Hemisphere”.   

        #africaisacountry is a hashtag on twitter devoted to this kinds of thing.    

        Thanks for your reply. My last ritalin should be wearing off in an hour or so. 

  13. So, when the central plains of America turn into a dust bowl, which part of the world will it feed it’s nutrient minerals to?

  14. So, wouldn’t the rainforest become increasingly rich in nutrients as vast amounts of plant matter fall to the floor and decay? Isn’t that how every forested landscape replenishes its nutrients to some extent? 

Comments are closed.