UK United Kingdom

‘CO₂ is food for plants!’ What will higher emissions mean for crop productivity?

Life is woven out of air by light – Jacob Moleschott Before starting to write this article, I asked my eight-year-old boy what he knew about how plants grow. He answered: “Plants take food from the air…

Could more CO₂ in the atmosphere mean faster-growing crops? Jose Cabezas/AFP

Life is woven out of air by light – Jacob Moleschott

Before starting to write this article, I asked my eight-year-old boy what he knew about how plants grow.

He answered: “Plants take food from the air and use the sun to grow”. He was pretty right: CO₂ is food for plants, though CO₂ is also a pollutant.

Having some CO₂ in the atmosphere can be a good thing, as it provides us with warmth through the greenhouse effect. At the same time it is the source of carbon for plants on Earth.

Laboratory studies have shown that having more CO₂ in the air will increase the rate of photosynthesis and growth for most temperate plants, if they have enough water and nutrients. This is called CO₂ fertilisation.

This is because photosynthesis in plants such as wheat, rice, and soybean is source limited. Quite simply, “source limited” means that at present concentrations of atmospheric CO₂ (approximately 380 parts-per-million), photosynthesis rates are running at about ¾ of full capacity.

The leaves of summer crops (such as maize and sorghum) work differently, and they are less responsive to increases in atmospheric CO₂.

But in real life – “out in the field” – things can be dramatically different. There is much more to climate change effects than just CO₂ fertilisation. More on that in a moment.

Can one desire too much of a good thing?

Sometimes too much of a good thing can be a bad thing. This is particularly the case with greenhouse gases such as CO₂.

As levels of CO₂ increases in the atmosphere, air temperatures increase and global and regional climate patterns change. This has serious implications for the functioning of most biological systems.

Recent CO₂ atmospheric measurements indicate we are already overshooting safe climate targets, with important implications for our climate systems.

In Australia, evidence indicates feedback processes are likely to increase the frequency of El Niño events.

In Australia, El Niño events are linked with widespread droughts, heat waves, and bush fires. From 1902 to 2002 we had 24 El Niño events, of which 62% occurred after 1950. Between 1990 and 2002 there were six such events.

At this point I think we can all agree that CO₂ is both a food for plants, and a dangerous pollutant.

So what does this mean in terms of the potential for our planet to feed a growing population by 2050?

Does more CO₂ mean more to eat?

When we think about the growth of plants and how much they yield, CO₂ fertilisation and climate change will have different effects. There is also a difference between climate change impacts on crop yields and on the profitability of farm businesses.

One third of the world’s food is produced by determinate crops grown on smallholder farms in the developing world. Determinate crops are ones that bear their full yield at once. These are crops such as wheat, rice, maize and beans that die after a few months in the ground, after which their full yield is harvested.

The yield is a fraction of the total growth of the crop – from 0-50% – and depends on whether the crop faces serious water or heat stresses.

CO₂ fertilisation has a positive effect on crop growth. But climate change will affect both the fraction of the crop that is harvestable, and the total growth of the crop.

Heat, water and CO₂: it’s a complex package

Increased air temperatures are expected to expose crops to more very hot days. This will damage developing grains and reduce the harvestable yield.

Crops will need more water and will face water shortages more often. Increased temperatures mean the time from sowing to harvest will be shorter, so crops will grow less and have smaller yields.

Water is key in crop growth and production: 80% of the total agricultural land depends on rainfall. Changes in precipitation are likely to have fundamental impacts on food production.

Analyses of recent trends in global food production could be the best indicator of how climate change will affect global food production in future. One analysis shows that since 1980, higher CO₂ and climate change have had a slight positive effect on rice and soybean, and a negative effect on wheat and maize.

In Africa, maize is the main food staple. Researchers calculated that 1°C of global warming would reduce yields across the continent by 20% over 75% of the area if there is drought. Under optimal management, 1°C of warming would reduce maize yields by around 10% on 65% of the area.

It’s not all bad news

The challenges are significant, but we should be able to feed 9 billion people by 2050. Over the last 50 years the increase in agricultural production fed an additional 4 billion people with only an 11% increase in land area.

In Brazil there has been sustainable intensification of broad acre agriculture. Vietnam has changed its regulations on land ownership. Malawi has introduced smart subsidies on agricultural inputs.

Nowadays Brazil is leading the world as a global net food exporter; smallholder farmers from Vietnam are increasingly accessing international markets; and Malawi, one of the poorest countries in the world, exports maize to neighboring countries.

These are significant, if small, examples of how the right technologies and policies can generate incentives, opportunities and economic growth from agriculture, whatever our climate future holds.

So, is CO₂ just plant food? Or is it just a pollutant? As with all of these things, the answer is: “it depends”.

Articles also by This Author

Sign in to Favourite

Join the conversation

22 Comments sorted by

Comments on this article are now closed.

  1. Timothy Curtin

    Economic adviser

    I am bound to say I found Rodriguez’ thoughts on CO2, climate change and food production wholly inadequate and serially misleading – and I think his 8-year old son has more clue as there is no way in which CO2 is a pollutant.

    Seriously, I do not know why he feels able to write this: “One analysis shows that since 1980, higher CO₂ and climate change have had a slight positive effect on rice and soybean, and a negative effect on wheat and maize”. Nonsense – that paper (Lobell and Burke) he links…

    Read more
    1. Ian Enting

      Honorary Senior Associate, Faculty of Science at University of Melbourne

      In reply to Timothy Curtin

      so you claim that if going from 390 to 650 ppm gives 34% more yield,
      reducing from 390 to 350 would reduce yield by "quite possiblly as much".
      The first 200 years of growing wheat in Australia was with CO2 levels
      below 350ppm. In the incredibly likely event of reversing the last 20 years
      or CO2 increase, "wiping out Australia's farmland" is not going to be the result.

    2. Eric Ireland

      logged in via Facebook

      In reply to Timothy Curtin

      I looked up that Crimp 2008 paper you mentioned that was referenced in the Garaut report and tried entering different values into that formula. ( if anyone else is interested!)

      The thing you seem to have forgotten is that elevated atmospheric CO2 is projected to cause increased temperatures and decreased rainfall. Enter values for reduced rainfall and increased temperature into the formula, and you find that…

      Read more
  2. Timothy Curtin

    Economic adviser

    Ian - ever heard of the law of Diminishing Marginal Returns? The gain in yield from extra CO2 from 315 ppm to 390 ppm will be more cet.par. than from 390 to 465. Anyway, I gave you the CSIRO equation, get some data on rainfall and temperature for say Dalby (BoM could help) + CO2 from CDIAC, and tell us what you find when you reduce the CO2 back to 315 as wished by this Green Government.

    1. Eric Ireland

      logged in via Facebook

      In reply to Timothy Curtin

      I used the equation and got the coefficients for Dalby from Crimp 2008, Table 2 (assuming no change in current crop management). You are right, assuming no climate change, yield increase from 315ppm to 390ppm is 10%, from 390ppm to 465ppm is 8%.

      If CO2 is 550, temperature is 2 (i.e. a 2 degree increase) and rainfall 0.9 (i.e a 10% decrease), yield change is -2%. (I got these rough values from maps assuming 'medium emissions' from The increase in temperature and decrease in rainfall negate the effect of CO2 fertilisation. If there is no climate change, yield increases 26%.

      So our debate hinges on whether or not increasing atmospheric CO2 will lead to climate change, and if so, how that will effect temperature and rainfall in Dalby.

  3. Nicholas Aberle

    Sure, Brazil is growing more food, but they're doing this largely at the expense of cutting down the Amazon. Which is understandable from a country-level quest for economic development, but is a massive problem from a sequestering of atmospheric carbon point of view. (NB. I'm not putting the blame on Brazil).

    I think its worth acknowledging the trade-offs in growing more food.

  4. James Szabadics


    Overall this is a good article but it has one central point that I would argue with. Is there a study that shows EL-Nino frequency is linked to CO2 concentration?

    CO2 is higher today than at any time in the recent past. But look at El-Nino Frequency.

    This century shows more frequent La-Ninas than El-ninos as time has progressed. Since 2005 we have 4 La-Ninas and 3 El-ninos. I dont think you can show a reasonable link between CO2 concentration and EL-Nino frequency. I think there definitely is a link between El-nino and global warming and also a link between La-nina and global cooling. What causes them exactly is still something we are learning about but it doesnt seem to be CO2 related.

  5. Timothy Curtin

    Economic adviser

    As a follow up to my reply to Ian Enting, I have now done basic multivariate regression of wheat yields at Moree NSW from 1965 to 1999 (dates determined by data source) with respect to rainfall, mean temperature, and atmospheric CO2. As I said, the role of [CO2] is crucial, as it alone has a statistically significant coefficient, with t-stat at 3.18 and its p-value showing better than 95% confidence, whereas the other 2 variables are not stat. sig. The adj. R2 is 0.367, suggesting omitted variable(s…

    Read more
  6. Ross James


    There's a bit of a problem here. The Australian Government doesn't list CO2 as a pollutant. Of course, it's no more a pollutant than oxygen is. It's only the politicians that have no scientific background (most of them) that refer to CO2 as "carbon pollution".

    Back in 2007, polar ice was predicted to disappear by 2013 - alas, it failed to decrease any further, and Antarctic ice continued to grow. Global temperatures haven't increased over the past 13 years. Sea level experienced a record drop over the past 2 years, and has been dropping for the past 10 years. Plants, if anything, need more CO2.

    We need both scientists and politicians to stop relying on failed computer models, and lreport real data.

    1. Shane Perryman

      logged in via email

      In reply to Ross James

      Engineers who provide references to otherwise obvious misleading distortions would also help.

      You: "Global temperatures haven't increased over the past 13 years."

      In 2007 the IPCC reported this: "2005 and 1998 were the warmest two years in the instrumental global surface air temperature record since 1850."
      Compare your claim to the instrumental record.

      The trend is up.

      And wikipedia kindly provides a list of…

      Read more
    2. Ross James


      In reply to Shane Perryman

      Shane - you shouldn't use wikipedia as your authority in climate science. They based their comments on NASA GISS data. NASA has admitted that they are not the best on global temperature, and refer to Hadcrut as the authority (I have this in writing from them). GISS data is in conflict with Hadcrut, RSS, and UAH. You can check Hadcrut at

      Given the corruption and poor scientific process that's come out of the IPCC, I'm surprised that you dare mention them. They also use GISS, as it's data suits their cause.

      As I said, Australian National Standards doesn't list CO2 as a pollutant - only politicians try to do this. Something that's part of the natural carbon cycle can't be a pollutant.

      Re sea level rise, it's been generally rising for the past 20,000 years, though slower in recent times. It's nothing new.

    3. Shane Perryman

      logged in via email

      In reply to Ross James

      And here I was thinking that James was just a little misguided.

      Your reply shows you have an entrenched position (the same claim you level at others). You do not counter my demonstration of your misleading and wrong assertions, but choose instead to target the messengers (IPCC, wikipedia).

      In particular when you say "corruption and poor scientific process that's come out of the IPCC, I'm surprised that you dare". Really? I do dare.

      The IPCC compiles the results from other sources. That the poor…

      Read more
    4. Shane Perryman

      logged in via email

      In reply to Shane Perryman

      "it's data suits their cause."
      Forgot to ask... what exactly is their cause?
      Just so we can all be clear on the level of paranoia motivating your concerns.

    5. Shane Perryman

      logged in via email

      In reply to Shane Perryman

      Absolutely my final comment on this thread.

      Notice how James never really counters the argument.

      Language like "I'm surprised that..." and "You are surely aware..." contains within it an implicit assertion that his statement is widely known to be true and that somehow I am foolish to disagree.

      I have provided links James that counter your silliness.

      I find it amusing that James trumpets NOAA data as that suited his argument but the great Tim Curtin, otoh says they are all wrong. Apparently, he…

      Read more
  7. Timothy Curtin

    Economic adviser

    Many thanks, Eric, for your second comment. But you overlook that Crimp et al assume in their projections to 2030 various changes in temperature and rainfall.

    So your ballpark figure is "if there is no climate change, yield increases 26%" is unduly pessimistic if there are also increases in [CO2].

    1. Eric Ireland

      logged in via Facebook

      In reply to Timothy Curtin

      If Crimp et al assume various changes in temperature and rainfall, why include variables for those in the formula? What changes have they assumed?

  8. Timothy Curtin

    Economic adviser

    Shane Perryman: all very fine if your temperature data are to be trusted. But the HadleyCRU and GISS and all the others showing GMT from 1850 or 1880 to 2010 did not have any of the tropics before 1910, and far from complete before 1950, while since 1990 the northern stations in Canada and former USSR have mostly slipped off the rada screen or drifted south. So you series scrubbed the hot tropics for the early period and the cold north for the later. That's science for you! To be very specific, the…

    Read more
  9. Ross James


    Shane - I'm surprised that you have to ask about the IPCC. You are surely aware of the scientists that have pulled out because of the bad processes used. Just things like saying a report has been reviewed by 2,500 scientists. Reviewed doesn't mean they agreed with the report - just that they commented on it. Even the Vice-Chairman Prof Yuri Izrael, stated "there is no proven link between human activity and global warming". "I think the panic over global warming is totally unjustified. There is no…

    Read more
  10. Ross James


    Actually, the talk about the IPCC prompted me to re-visit previous Expert Comments. As you would expect, it's fairly long. However, are few points are of particular interest.

    Michael Ghil UCLA pointed out that he received Eamil request for review on April 24 2000, and given until May 8 to reply. He stated that "this was a much shorter than the time normally allowed for a single paper, let alone specialist journals.

    Anthony Lupo University of Missouri considered that the statement in the second…

    Read more
    1. Michael J. I. Brown

      ARC Future Fellow and Senior Lecturer at Monash University

      In reply to Ross James

      Many (perhaps all) of the above comments refer to IPCC AR3. Thus, many of these comments do not apply to more recent IPCC AR4 report.

  11. Timothy Curtin

    Economic adviser

    Shane, I fear you will not be much missed, as you avoid answering specific questions put to you, and then misrepresent others' comments. Contrary to what you just said, I specifically excluded ESRL-NOAA from my criticisms of NASA-GISS with it shameful data manipulation like its Met station selections for Scotland: "To be very specific, the Gistemp et al records for Scotland are based on just a handful of stations most of which are islands in the Gulf Stream, 2 on the mainland are airports, and the…

    Read more
  12. Trevor Ellice

    logged in via Facebook

    why is it assumed that higher temperature is associated with drought?

    higher ocean temperatures in the ENSO system whether off the coast of South America (El Nino) or over the Maritime Continent (la Nina) result in more rains. I don't get it.

    More heat, more rain, more CO2 = better plant growth - I would have thought?