Ecology

1.FAO—Food and Agriculture Organization of the United Nations. World Crops Production. http://www.wptc.to/releases-wptc.php (2016).2.WPTC—World Processing Tomato Council. World production estimate. http://www.wptc.to/releases-wptc.php. (2016).3.Silva, Y. P. A. et al. Characterization of tomato processing by-product for use as a potential functional food ingredient: Nutritional composition, antioxidant activity and bioactive compounds. Int. J. Food Sci. Nutr. 70, 150–160 (2019).Article 

Google Scholar 
4.Pereira, M. A. B. et al. Postharvest conservation of structural long shelf life tomato fruits and with the mutant rin produced, in edaphoclimatic conditions of the southern state of Tocantins. Ciênc. Agrotec. 39, 225–231 (2015).Article 

Google Scholar 
5.Brummell, D. A. & Harpster, M. H. Cell wall metabolism in fruit softening and quality and its manipulation in transgenic plants. PCW. 47, 311–340 (2001).CAS 

Google Scholar 
6.Meli, V. S. et al. Enhancement of fruit shelf life by suppressing N-glycan processing enzymes. Proc. Natl. Acad. Sci. USA 107, 2413–2418 (2010).ADS 
CAS 
Article 

Google Scholar 
7.Samimi-Akhijahani, H. & Arabhosseini, A. Accelerating drying process of tomato slices in a PV-assisted solar dryer using a sun tracking system. Renew. Energy 123, 428–438 (2018).Article 

Google Scholar 
8.Tripathy, P. P. Investigação da secagem solar da batata: efeito da geometria da amostra na cinética de secagem e na mitigação das emissões de CO2. J. Ciênc. e Tecnol. Alim. 52, 1383–1393 (2015).CAS 

Google Scholar 
9.Badaoui, O., Hanini, S., Djebli, A., Haddad, B. & Benhamou, A. Experimental and modelling study of tomato pomace waste drying in a new solar greenhouse: Evaluation of new drying models. Renew. Energy 133, 144–155 (2019).Article 

Google Scholar 
10.Mohsen, H. A., El-Rahmam, A. A. & Hassan, H. E. Drying of tomato fruits using solar energy. Int. J. Agric. Eng. 21, 204–215 (2019).
Google Scholar 
11.César, L. V. E., Lilia, C. M. A., Octavio, G. V., Isaac, P. F. & Rogelio, B. O. Thermal performance of a passive, mixed-type solar dryer for tomato slices (Solanum lycopersicum). Renew. Energy 147, 845–855 (2020).Article 

Google Scholar 
12.Kingsly, A. R. P., Singh, R., Goyal, R. K. & Singh, D. B. Thin-layer drying behavior of organically produced tomato. Am. J. Food Tech. 2, 71–78 (2007).Article 

Google Scholar 
13.Miller, G. L. Use of dinitrosalicylic acid reagent for determination of reducing sugar. Anal. Chem. 31, 426–428 (1959).CAS 
Article 

Google Scholar 
14.Silva, F. A. S. A. & Azevedo, C. A. V. Versão do programa computacional Assistat para o sistema operacional Windows. Rev. Bras. Prod. Agroindustriais 4, 71–78 (2002).Article 

Google Scholar 
15.Klunklin, W. & Savage, G. Effect on quality characteristics of tomatoes grown under well-watered and drought stress conditions. Foods 6, e56 (2017).Article 

Google Scholar 
16.Azeez, L., Adebisi, S. A., Oyedeji, A. O., Adetoro, R. O. & Tijani, K. O. Bioactive compounds’ contents, drying kinetics and mathematical modelling of tomato slices influenced by drying temperatures and time. J. Saudi Soc. 10, 120–126 (2019).
Google Scholar 
17.Correia, A. F., Loro, K. A. C., Zanatta, S., Spoto, M. H. F. & Vieira, T. M. F. S. Effect of temperature, time, and material thickness on the dehydration process of tomato. Int. J. Food Sci. 1, e970724 (2015).
Google Scholar 
18.Eswara, A. R. & Ramakrishnarao, M. Solar energy in food processing—A critical appraisal. J. Food Sci. Technol. 50, 209–227 (2013).CAS 
Article 

Google Scholar 
19.Castillo, C. P., Silva, F. B. & Lavalle, C. An assessment of the regional potential for solar power generation in EU-28. Energy Policy 88, 86–99 (2016).Article 

Google Scholar 
20.Tampakis, G., Tsantopoulos, G. & Arabatzis, I. R. Citizens’ views on various forms of energy and their contribution to the environment. Renew. Sust. Energ. Rev. 20, 473–482 (2013).Article 

Google Scholar 
21.Tsantopoulos, G. & Arabatzis, T. G. Stilianos Public attitudes towards photovoltaic developments: Case study from Greece. Energy Policy 71, 94–106 (2014).Article 

Google Scholar 
22.Tiwari, R. B. Application of osmo-air dehydration for processing of tropical frepical fruits in rural areas. Indian Food Ind. 24, 62–69 (2005).
Google Scholar 
23.Goula, A. M. & Adamopoulos, K. G. Retention of ascorbic acid during drying of tomato halves and tomato pulp. Drying Technol. 24, 57–64 (2006).CAS 
Article 

Google Scholar 
24.McAlpine, R. D., Cocivera, M. & Chen, H. Photooxidation and reduction of ascorbic acid atudied by E.S.R. Can. J. Chem. 51, 1682–1686 (1973).CAS 
Article 

Google Scholar 
25.Santos, P. H. S. & Silva, M. A. Retention of vitamin C in drying processes of fruits and vegetables—A review. Drying Technol. 26, 1421–1437 (2008).CAS 
Article 

Google Scholar 
26.Santos-Sánchez, N. F., Valadez-Blanco, R., Gómez-Gómez, M. S., Pérez-Herrera, A. & Salas-Coronado, R. Effect of rotating tray drying on antioxidant components, color and rehydration ratio of tomato saladette slices. LWT Food Sci. Technol. 46, 298–304 (2012).Article 

Google Scholar 
27.Yadav, A. K. & Singh, S. V. Y. Osmotic dehydration of fruits and vegetables: A review. J. Food Sci. Technol. 51, 1654–1673 (2014).Article 

Google Scholar 
28.Gunhan, T., Demir, V., Hancioglu, E. & Hepbasli, A. Mathematical modeling of drying of bay leaves. Energy Convers. Manag. 46, 1667–1679 (2005).Article 

Google Scholar 
29.Sacilik, K. & Unal, G. Dehydration characteristics of kastomonu garlic slices. Biosyst. Eng. 92, 207–215 (2005).Article 

Google Scholar 
30.Instituto Adolfo Lutz. Métodos Físico-Químicos Para Análise de Alimentos 1020 (Instituto Adolfo Lutz, 2008).
Google Scholar  More

NEWS AND VIEWS
25 August 2021

African tropical montane forests store more carbon than was thought

The inaccessibility of African montane forests has hindered efforts to quantify the carbon stored by these ecosystems. A remarkable survey fills this knowledge gap, and highlights the need to preserve such forests.

Nicolas Barbier

0

Nicolas Barbier

Nicolas Barbier is at AMAP, Université de Montpellier, IRD, CNRS, INRAE, CIRAD, Montpellier 34980, France.

View author publications

You can also search for this author in PubMed
 Google Scholar

Share on Twitter
Share on Twitter

Share on Facebook
Share on Facebook

Share via E-Mail
Share via E-Mail

Download PDF

In a paper in Nature, Cuni-Sanchez et al.1 report the assembly of a large database of tree inventories for 226 mature montane-forest plots in 12 African countries. The authors analyse the data to determine the amount of aboveground biomass and carbon stored in these highly diverse and threatened ecosystems. Their results suggest that African montane forests store more carbon than was previously thought, and the findings should help to guide efforts to conserve these ecosystems.Cuni-Sanchez and colleagues measured trunk diameters and heights of the trees in plots, and identified the botanical species to deduce wood density — an approach that constitutes the gold standard for estimating the biomass, and thus the amount of carbon, contained per unit of forest area. This method involves the use of general statistical equations for describing tree form, called allometric models, and considers only the aboveground parts of trees. It therefore disregards several other pools of carbon, notably in the roots and soil. The overall approach might seem crude, but recognizing and measuring the many hundreds of tree species found on steep, cloud-shrouded slopes (Fig. 1), let alone the underground carbon, without visiting the sites, will long remain difficult, even with the best drones and satellite systems.

Figure 1 | Montane forest in Boginda, Ethiopia. Cuni-Sanchez et al.1 use data from a survey of montane tropical forests in Africa to quantify the amount of carbon stored above ground in these ecosystems.Credit: Bruno D’Amicis/Nature Picture Library

Anyone who has conducted field inventories in tropical mountains knows that measuring and identifying 72,336 trees, often just a few steps away from the void, is an amazing feat. For comparison, a previously reported study2 based its estimates of the carbon stored in montane African forests on as few as seven plots. The study also brings together contributions from numerous researchers and institutions, including many in Africa, to greatly increase the size of the data set, which is also a remarkable achievement. Even so, the total area of forest studied is less than 150 hectares, whereas African montane forest covers about 100,000 times that area, inevitably raising questions about how representative the inventory is.Statisticians might raise their eyebrows at the sampling design. As is usually the case in meta-analyses, the data set was neither homogeneous (for example, there is a roughly tenfold variation in the plot sizes), nor were the sites selected at random. However, the authors did their best to rule out possible biases induced by sampling artefacts.
Read the paper: High aboveground carbon stock of African tropical montane forests
Cuni-Sanchez et al. chose not to discuss one tricky aspect of surveys of this sort (extensively discussed elsewhere2): how should the land area of a steep slope be measured? The authors followed standard practice, which is to measure the extent of forest plots and of land-cover types in reference to horizontal, planimetric areas (that is, the areas that would be represented on a 2D map, as if seen from the air). This tends to overestimate aboveground carbon because the sloped surface area is greater than that of the planimetric area — which means that the tree density of the planimetric area is higher than it is on the slope. By contrast, the use of planimetric areas underestimates total montane-forest area (by about 40%; see ref. 2). These two biases should roughly cancel each other out when estimating carbon stocks, or changes to stocks, for a region or country. But care should be taken not to combine data acquired using planimetric and non-planimetric areas in future meta-analyses, because the resulting estimates could end up well off the mark.One might expect that trees in mature African montane forests would be, on average, shorter — and therefore store less carbon — than their lowland counterparts, because of their lower environmental temperatures and shallow soils, frequent landslides and strong winds. However, this is not what Cuni-Sanchez et al. report. Instead, they find that average aboveground carbon stocks are not significantly different from those of mature lowland forests. This contrasts with the situation in the neotropics and southeast Asia, where montane forests store, on average, less carbon than do lowland forests.However, the new results fit with the 2016 discovery that the tallest African trees (81.5 metres) grow on Mount Kilimanjaro3, the highest mountain in Africa. African forests, in general, tend to contain fewer but larger-statured tree stands than does, for example, Amazonia4. The current study confirms that this peculiarity applies even at high altitudes.The authors investigate several possible drivers for the variations in biomass observed at different sites in their study, including topography, climate, landslide hazard, and even the presence of elephants or certain conifers (Podocarpaceae), but were unable to identify any clear pattern. Many environmental, historical and biological effects probably interact, with each of these effects varying greatly in ways that are poorly captured by available data sets. These effects must therefore be disentangled before a predictive model of African montane carbon distribution can be developed.
Tropical carbon sinks are saturating at different times on different continents
Nevertheless, Cuni-Sanchez and colleagues’ study underlines a crucial message: African montane forests are immensely valuable, and not only because they host the source of the River Nile, mountain gorillas and ecosystems such as mysterious lichen-covered forests. They also store vast amounts of carbon, and thereby have a key role in tackling climate change. Of course, this immense intrinsic value does not preclude intense human exploitation of these ecosystems, which can lead to rapid degradation and deforestation. For instance, on the basis of satellite monitoring, Cuni-Sanchez and colleagues report that Mozambique lost nearly one-third of its montane forests between 2000 and 2018.There is, however, the faint hope that putting a financial value on carbon, and the establishment of economic incentives to avoid deforestation in tropical countries, might help to check the flood of damage5. The aim is to reward African countries — for which montane forest sometimes constitutes the last remaining forests — for their conservation endeavours, and for renouncing efforts to access the timber and ore in these ecosystems, even when such resources are otherwise desperately lacking. By gathering the best-available data to provide precise, country-level estimates of average aboveground carbon content in African montane forests, Cuni-Sanchez and colleagues’ study will add weight to such efforts — not least because the new estimates are, on average, two-thirds higher than the values reported by the Intergovernmental Panel on Climate Change6.The next step should be to extend measurements in these forests, particularly by continuing to support national forest-inventory efforts. These inventories target all vegetation types, rather than just the most intact forests, and all carbon pools, using standardized protocols and systematic sampling methods. Remote sensors, both in the sky and in space, should also be used to fully map the detailed spatial variation of forest diversity, structure and dynamics. But there is no excuse for delaying policymaking — we already know enough to justify immediate decisive action to preserve yet another of Earth’s threatened treasures.

Nature 596, 488-490 (2021)
doi: https://doi.org/10.1038/d41586-021-02266-3

References1.Cuni-Sanchez, A. et al. Nature 596, 536–542 (2021).Article 

Google Scholar 
2.Spracklen, D. V. & Righelato, R. Biogeosciences 11, 2741–2754 (2014).Article 

Google Scholar 
3.Hemp, A. et al. Biodivers. Conserv. 26, 103–113 (2017).Article 

Google Scholar 
4.Lewis, S. L. et al. Phil. Trans. R. Soc. B 368, 20120295 (2013).PubMed 
Article 

Google Scholar 
5.Venter, O. et al. Science 326, 1368 (2009).PubMed 
Article 

Google Scholar 
6.Domke, G. et al. in 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories Vol. 4 (eds Calvo Buendia, E. et al.) Ch. 4, 4.48 (IPCC, 2019).
Google Scholar 
Download references

Competing Interests
The author declares no competing interests.

Related Articles

Read the paper: High aboveground carbon stock of African tropical montane forests

Southeast Amazonia is no longer a carbon sink

Tropical carbon sinks are saturating at different times on different continents

See all News & Views

Subjects

Biogeochemistry

Environmental sciences

Ecology

Latest on:

Biogeochemistry

High aboveground carbon stock of African tropical montane forests
Article 25 AUG 21

The Montreal Protocol protects the terrestrial carbon sink
Article 18 AUG 21

Amazonia as a carbon source linked to deforestation and climate change
Article 14 JUL 21

Environmental sciences

Brazilian road proposal threatens famed biodiversity hotspot
News 17 AUG 21

‘Polluter pays’ policy could speed up emission reductions and removal of atmospheric CO2
News & Views 16 AUG 21

The world’s species are playing musical chairs: how will it end?
News Feature 04 AUG 21

Ecology

Can artificially altered clouds save the Great Barrier Reef?
News Feature 25 AUG 21

Great Barrier Reef: accept ‘in danger’ status, there’s more to gain than lose
World View 18 AUG 21

Brazilian road proposal threatens famed biodiversity hotspot
News 17 AUG 21

Jobs

Postdoctoral Research Scientist

CRUK Beatson Institute for Cancer Research
Glasgow, United Kingdom

Bioinformatics Software Engineer

CRUK Beatson Institute for Cancer Research
Glasgow, United Kingdom

Computational Biologist

CRUK Beatson Institute for Cancer Research
Glasgow, United Kingdom

Research Fellow – Berbeco Lab

Dana-Farber Cancer Institute (DFCI)
Boston, MA, United States

Nature Briefing
An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.

Email address

Yes! Sign me up to receive the daily Nature Briefing email. I agree my information will be processed in accordance with the Nature and Springer Nature Limited Privacy Policy.

Sign up More

Download PDF

In place of its normal load of cars and vans, the repurposed ferry boat sported a mobile science laboratory and a large fan on its deck as it left Townsville, Australia, in March. Researchers dropped anchor in a coral lagoon some 100 kilometres offshore and then fired up the cone-shaped turbine, which blew a mist of seawater off the back of the boat. What happened next came as a welcome surprise: after briefly drifting along the ocean surface, the plume ascended into the sky.Looking a bit like a jet engine, this mist machine is at the centre of an experiment that, if successful, could help to determine the future of the Great Barrier Reef. Three-hundred and twenty nozzles spewed a cloud of nano-sized droplets engineered to brighten clouds and block sunlight — providing a bit of cooling shade for the coral colonies below. Scientists used sensors aboard the ferry, drones and a second boat to monitor the plume as it migrated skyward.The experiment wasn’t big enough to significantly alter the clouds. But preliminary results from the field tests — which were shared exclusively with Nature — suggest that the technology might perform even better than computer models suggested it would, says Daniel Harrison, an oceanographer and engineer at Southern Cross University in Coffs Harbour, Australia, who is heading up the research. “We are now very confident that we can get the particles up into the clouds,” Harrison says. “But we still need to figure out how the clouds will respond.”Harrison’s project is the world’s first field trial of marine cloud brightening, one of several controversial geoengineering technologies that scientists have studied in the laboratory for decades. The research has been driven by fear that humans might one day be forced to deliberately manipulate the Earth’s climate and weather systems to blunt the most severe impacts of global warming.For many Australians, that day arrived in 2017, when a marine heat wave spurred massive coral bleaching and death across much of the 2,300-kilometre Great Barrier Reef. That crisis hit just a year after another bleaching event along the reef, which supports more than 600 species of coral and an estimated 64,000 jobs in industries such as tourism and fishing. Research suggests that the reef lost more than half of its coral between 1995 and 2017, as a result of warming waters, tropical storms and predatory starfish (A. Dietzel et al. Proc. R. Soc. B. 287, 20201432; 2020).
These corals could survive climate change — and help save the world’s reefs
The project has raised concerns among some scientists abroad, in part because the Australian group has published little about its work. Environmentalists outside Australia objected to the project last year after news of the first trial broke, and there could be similar criticism when details of the 2021 trial emerge.Harrison stresses that the cloud-brightening project is about local adaptation to climate change, not global geoengineering, because its application would be limited in both space and time. It’s also just one part of a larger Aus$300 million (US$220 million) Reef Restoration and Adaptation Program (RRAP) launched last year by Australia to investigate and develop techniques and technologies to save the country’s reefs. Many of the proposals, from cloud brightening to breeding heat-tolerant corals, would represent unprecedented human interventions in the natural reef system.Ecological modelling suggests that a large-scale intervention involving multiple strategies — including a fleet of mist machines — could prolong the life of the reef while governments work to eliminate greenhouse-gas emissions. The goal now is to work out what’s achievable in the real world, says Cedric Robillot, executive director of the RRAP.“You need to consider every angle, from the fundamental science to the very pointy end of engineering, if you want to succeed,” Robillot says. “It’s not enough to just prove you could do it. You need to explain how you would do it.”Into the cloudsHarrison conducted his first field test in March 2020: a three-day proof-of-concept expedition on a small car ferry with four scientists, one representative from a local Indigenous group, and two shipping containers for equipment and sleeping quarters. The team had a minimal Aus$400,000 budget and limited scientific instrumentation to monitor the mist, but it was enough to document that the plume flowing out of their mist machine rode a draught of warm air high into the sky.It was the first time they had witnessed this phenomenon. Their models had suggested that evaporation of the brine droplets would cool the plume, which would then float across the surface of the ocean, only slowly mixing upwards into the low-lying marine clouds. The models also indicated a risk that the tiny droplets might merge and drop out of the air. Instead, brine droplets floated along the surface of the ocean for half a kilometre without coalescing, gradually losing water and weight to evaporation along the way. And then they shot upwards.

A marine heat wave in 2017 caused coral bleaching along much of Australia’s Great Barrier Reef.Credit: Juergen Freund/Nature Picture Library

“We didn’t expect that at all,” Harrison says, “but it turned out we were doing this experiment in the middle of a rising air mass.”The scientists feared it was a fluke. Although years of research and development have gone into the nozzles, initially led by a separate American team, this was the first time anybody had ever deployed them in the field with fresh seawater. The team also didn’t know what to expect from clouds and aerosols in that region, because research on the reef has focused almost exclusively on what happens below the water, not the conditions above.For Harrison, the 2020 experiment was more than enough to justify moving forward with another, larger trial in March 2021. But it did raise eyebrows among some scientists and observers abroad, where geoengineering research has met strong opposition and struggled to attract funding.
IPCC climate report: Earth is warmer than it’s been in 125,000 years
Most of the concern has centred on a form of solar geoengineering that involves injecting reflective material into the stratosphere to block sunlight at a global scale. But cloud brightening has also been studied as a potential global intervention, and it has attracted criticism from some environmental groups who argue that it carries inevitable ecological risks and detracts from efforts to limit greenhouse gases.Some scientists, as well as environmental advocates who follow geoengineering research, told Nature that they were surprised to see the experiment move forward without more scrutiny — or without published research to justify such an investment.Critics also worry that Australia is setting the wrong kind of precedent by rebranding a solar-geoengineering experiment that could have regional impacts as a local adaptation project. “One could say that there should have been some level of consultation with the outside world,” says Janos Pasztor, who heads the Carnegie Climate Governance Initiative, an advocacy group in New York City that has been pushing for a global debate over geoengineering governance in the United Nations.Harrison says scientists in the programme have consulted with regulatory authorities, as well as with the general public and Indigenous groups that have historic claims on the reef. He also readily acknowledges trying to avoid getting embroiled in a debate about solar geoengineering, arguing that the project would be more akin to cloud-seeding operations that are designed to promote rain and that are not considered to be geoengineering. One of the next modelling efforts, however, will be to explore any potential regional and global implications, he says.

A plume of seawater droplets rises up into the sky during a field trial in March 2021.Credit: Brendan Kelaher/SCU

Others question the Australian government’s motivations in funding such work. Under the conservative prime minister Scott Morrison, the government has yet to strengthen its climate pledge under the 2015 Paris agreement, as many nations have done in the past year. Morrison has personally ruled out committing to net-zero emissions. Pushing for a technological fix to global warming without moving to aggressively curb greenhouse gases is “sheer lunacy”, says Peter Frumhoff, chief climate scientist for the Union of Concerned Scientists, an advocacy group in Cambridge, Massachusetts.Some researchers, however, are pleased to see marine cloud brightening move from theory to the field, including US scientists working on a similar project that has been struggling to get into the field for nearly a decade. “This is an early example of how climate disruption can drive interest in these things,” says Sarah Doherty, an atmospheric physicist who manages the Marine Cloud Brightening Project at the University of Washington in Seattle. Members of the team provided the initial nozzle design and have been tracking the Australian group’s progress.Coral crisisThe first time that scientists observed a major bleaching event along the Great Barrier Reef was in 1998, and the second event followed four years later. In both cases, corals expelled the algae that live within them and that provide colour and energy through photosynthesis. Most of the corals eventually recovered. But in 2016 and 2017, many corals bleached and then died across two-thirds of the reef.
First sun-dimming experiment will test a way to cool Earth
“It was absolutely horrifying,” says David Wachenfeld, chief scientist at the Great Barrier Reef Marine Park Authority, which manages the reef. The clear message from those events was that the traditional approach to managing corals and coral reefs would not be enough, he adds. “Our hand was forced.”In 2018, the Australian government allocated Aus$6 million to a consortium of universities and government research institutes for a feasibility study focused on potentially radical strategies that could be applied across the reef. Researchers reviewed some 160 ideas, including putting live corals on ice for long-term preservation and synthetically engineeering new varieties that can tolerate the warmer waters. Many approaches proved too costly and energy intensive, but 43 interventions were singled out for further study. Marine cloud brightening drew support in part because it theoretically provides direct relief precisely when and where corals need it most.Much of the emphasis of the programme is on helping corals to adapt and repopulate the reef, including efforts to improve coral aquaculture operations so that they can produce millions of corals per year rather than thousands. For Madeleine van Oppen, a coral geneticist at the Australian Institute of Marine Science near Townsville, the RRAP programme helps to integrate her team’s work on assisting coral evolution to make them more heat tolerant.Thanks to the RRAP, she says, data from those projects are now being fed directly into models that enable researchers to assess the potential benefits — as well as the risks — of releasing new strains of coral and microalga into the wild. The programme is also raising ecological questions, such as whether the introduction of new coral species can propagate disease, or whether a new variety of more heat-tolerant corals might displace corals struggling to survive.

Researchers are testing specialized nozzles that create jets of seawater mist.Credit: Alejandro Tagliafico/SCU

“It sort of speeds up the whole path from research to implementation in the field,” says van Oppen.In the long run, the models indicate that without interventions, the extent of coral on the reef could shrink by well over 60% by 2070 compared with 2020 levels (S. A. Condie et al. R. Soc. Open Sci. 8, 201296; 2021). But simulations suggest that Australia could cut those losses in half with a three-pronged approach focused on propagating heat-tolerant corals, controlling outbreaks of the predatory crown-of-thorns starfish and brightening clouds to take the edge off of heat waves. Crucially, the latest modelling also suggests that without the cooling provided by Harrison’s cloud brightening project, the other interventions might not amount to much.Testing the windWhen Harrison’s group returned to the field this year, they had more-powerful drones as well as other aerosol sensors on a second boat. As in the previous year’s experiment, each time they created a plume, it rose into the sky after the droplets lost around 90% of their water to evaporation. The likely explanation, Harrison says, is that the reef is creating its own weather as warm water along the shallow corals heats the air above.Many more droplets are making it into the clouds than the scientists had initially calculated, but Harrison says their mist machine might need to be scaled up by a factor of 10 — from 320 to around 3,000 nozzles — to produce enough particles to brighten nearby clouds by around 30%. His team’s modelling suggests that this could in turn reduce the incoming solar radiation on the reef locally by around 6.5%. Even then, the operation would require 800–1,000 stations to cover the length of the Great Barrier Reef.
Fevers are plaguing the oceans — and climate change is making them worse
But it’s unclear whether that spray of salty droplets will have the desired effect, says Lynn Russell, an atmospheric chemist at the Scripps Institution of Oceanography in La Jolla, California, who has studied cloud brightening. Russell has not seen the latest — and as-yet unpublished — results, but questions whether there are enough of the low layered clouds considered suitable for cloud brightening.Harrison acknowledges such concerns and says that his team sees more of these clouds on the southern part of the reef. His team’s modelling suggests the technology will also work on the clouds that are common across the rest of the reef in summer. Even then, he says, it remains unclear how much coverage a full-scale cloud-brightening operation could provide across the entirety of the reef. More measurements, and detailed modelling, are needed to provide answers.For now, Harrison has secured funding for another two years, and he needs to demonstrate progress. The RRAP is testing all 43 approaches and will redistribute resources to projects that show potential, Robillot says. But he stresses that no amount of science and engineering will preserve the reef in its current form. “Even if we do all of this, the system that you’ll end up with is not going to be the Great Barrier Reef that we know today,” Robillot says. “You might, however, retain a very functional ecosystem.”That’s enough to keep Harrison going, and his team is already preparing for a trip into the field in 2022. The scientists plan to run the mist machine at higher pressure, which should produce a sixfold increase in the number of particles, and they will use new instrumentation to determine how particles alter clouds. They are also investigating an entirely different nozzle technology that could reduce the number of nozzles needed by a factor of 1,000.Harrison is more confident today than he was even a year ago that cloud brightening might work over the reef, but he is also realistic about the future if governments fail to limit carbon emissions. “There are only so many clouds available, and there is only so much you can brighten them,” he says. “Eventually, climate change just overwhelms things.”

Nature 596, 476-478 (2021)
doi: https://doi.org/10.1038/d41586-021-02290-3

Related Articles

First sun-dimming experiment will test a way to cool Earth

Fevers are plaguing the oceans — and climate change is making them worse

These corals could survive climate change — and help save the world’s reefs

IPCC climate report: Earth is warmer than it’s been in 125,000 years

Subjects

Ocean sciences

Conservation biology

Climate change

Climate sciences

Latest on:

Ocean sciences

Possible poriferan body fossils in early Neoproterozoic microbial reefs
Article 28 JUL 21

Cruise ships could sail now-icy Arctic seas by century’s end
Research Highlight 09 JUL 21

Shark mortality cannot be assessed by fishery overlap alone
Matters Arising 07 JUL 21

Climate change

Control methane to slow global warming — fast
Editorial 25 AUG 21

Five principles for climate-resilient cities
Correspondence 24 AUG 21

Great Barrier Reef: accept ‘in danger’ status, there’s more to gain than lose
World View 18 AUG 21

Jobs

Postdoctoral Research Scientist

CRUK Beatson Institute for Cancer Research
Glasgow, United Kingdom

Bioinformatics Software Engineer

CRUK Beatson Institute for Cancer Research
Glasgow, United Kingdom

Computational Biologist

CRUK Beatson Institute for Cancer Research
Glasgow, United Kingdom

Research Fellow – Berbeco Lab

Dana-Farber Cancer Institute (DFCI)
Boston, MA, United States

Nature Briefing
An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.

Email address

Yes! Sign me up to receive the daily Nature Briefing email. I agree my information will be processed in accordance with the Nature and Springer Nature Limited Privacy Policy.

Sign up More

Conceptualization of organismal to tumor biogeography through clone phylogeniesWe first map organismal biogeographic concepts and models to the process of migration and colonization of cancer cells during metastasis. Tumors are populations consisting of a diversity of cancer cells with different genetic profiles that may represent different lineages in the clone phylogeny. We use the example in Fig. 1, which contains a phylogeny of 17 clones found in one primary tumor (P) and four metastases (M1–M4). Events occurring along a branch in a phylogeny are anagenetic events, which include diversification, extinction, and expansion12,14. In organismal evolutionary biology, anagenetic events are not directly observed except through the fossil record. However, one can map the collection of genetic variants that likely arose on individual lineage in a phylogeny. In many cancers, sequencing of temporally sampled biopsies’ can directly reveal anagenetic events similar to the sequencing of ancient DNA in paleogenomics.The other types of evolutionary events in the phylogeny are cladogenetic, including genetic divergences and dispersals (Fig. 1). Genetic differences observed among species and populations are the key to detect cladogenetic events reconstructed in molecular phylogenies of living descendants. In cancer, temporal sampling of biopsies can reveal cladogenetic events that produced extinct descendants.In biogeography, genetic divergence results in the diversification of lineages within an area. Sometimes, the term duplication is used, but we avoid its further use because of the confusion it may cause in evolutionary genomics. Divergence events are also observed in a clone phylogeny, particularly when clone lineages diverge from each other within a tumor or across tumors. The exact opposite of genetic diversification can also be observed when lineages partially or fully disappear from the phylogeny. Extinction can occur due to random chance, selection, or environmental pressures. Even though extinction is rarely discussed in tumor clone phylogenetics, it happens frequently.Phylogenies also reveal movements of lineages between locations (geographic areas or body parts) when the locations of individual cells, species, or populations are known5,6,7,8,9,15. When lineages accumulate genetic differences along a branch in the phylogeny, and the evolved lineages migrate to a new area, we observe an expansion event. Expansions differ from dispersals in such that the growth of a population occurs in the same place. This movement of cells of a clone from one location to another, where they would potentially form a metastasis, results in the dispersal of  these cells of that clone to additional areas, which is modeled by a dispersal rate (d) in organismal biogeography. When a clone genetically diverges following its migration, then a distant dispersal event is said to have occurred. Similarly, when a clone diverges from the rest of the clones within a tumor and disperses to another tumor, we have observed an expansion event. Thus, clone phylogenies can give insights into the origin and trajectory of cancer cells between tumors.When a clone is no longer present at a location, it is extinct at that location. Extinctions are modeled by an extinction rate (e) in biogeographic models. As a result of extinction, the range of descendent clones on a phylogeny can be smaller than the ancestors. Biogeography models also have a parameter (J) to consider founder events that establish new populations from a few individuals. In phylogenies, founder events can be detected if only one or a few cells are found to have moved from one location to another to start diversifying in a new area. Both distant dispersal and founder events may result in forming a new colony of cells, i.e., a new metastasis in the case of cancer cell migrations. The primary distinction between dispersal and founder events is the relative number of migrating cells. Founder events are due to one or a few cells, whereas dispersal events involve a larger number of migrating cells. Founder events are expected to be more common in tumor evolution because metastases are thought to be formed by the spread of only one or a few cancer cells. These biogeographic events have been mathematically modeled and implemented in various approaches to infer species migration events12, which are directly applicable in the inference of cancer cell migrations between tumors.Model fitsWe began by analyzing the statistical fits of six biogeographic models (Table 1) to 80 computer-simulated tumor evolutionary datasets. Simulations enable us to assess the performance of computational approaches and reveal potential caveats associated with their use because the ground truth is known. These datasets were simulated using four main clone migration schemes defined by the different number of migrating clones (1–3), the small and large number of tumor areas (5–7 tumors, m5 datasets; 8–11 tumors, m8 datasets), and the different types of source areas of migration (primary or metastasis). The following seeding scenarios reflect this complexity of the clone migration schemes: monoclonal single-source seeding (mS), polyclonal single-source seeding (pS), polyclonal multisource seeding (pM), and polyclonal reseeding (pR) (see “Methods” section).Table 1 Phylogenetic and biogeographic events considered in seven biogeographic models used for analysis.Full size tableWe considered biogeographic models that weigh genetic divergence, dispersal/expansion, and extinction events differently (Table 1). We also explored the provision of including founder events in our models on the accuracy of detecting clone migrations. The parameterization of the aforementioned events results in models with two free parameters, i.e., dispersal rate (d) and extinction rate (e), and models with three free parameters by adding the founder-event speciation (J); see “Methods” section for more details.Overall, we tested six biogeographic models for their fit to the tumor data, three models with two free parameters and three others with three free parameters. BAYAREALIKE, DEC, and DIVALIKE models have two parameters each. They are nested within their respective models that add the founder effect, resulting in a model with three free parameters (hereinafter +J models). We used the BioGeoBEARS software for all model fit analyses. In data analysis, we first inferred phylogeny of cancer cell populations (clone phylogeny) using an existing method16, followed by the use of BioGeoBEARS to infer the clone migration history in which the clone phylogeny is used along with the location of tumor sites in which each clone is observed (Fig. 2). BioGeoBEARS estimates the probabilities of annotating internal nodes with tumor locations. These annotations are then used to derive cancer cell migration paths when two adjacent nodes are annotated with different tumor locations. In these analyses, we assumed the correct clone phylogeny because our focus was not assessing the impact of errors in a phylogeny on the accuracy of clone migration inferences. We also compared the accuracy of migration histories reconstructed using biogeographic models in BioGeoBEARS with those obtained from the approaches that do not model biogeographic processes (BBM9, MACHINA5, and PathFinder7).Figure 2Data analysis pipeline using BioGeoBEARS in R14 to infer clonal migration histories.Full size imageWe first conducted Likelihood Ratio Tests (LRTs) to examine the improvement offered by considering founder events in modeling tumor migrations. In this case, the fit of the BAYAREALIKE, DEC, and DIVALIKE models was compared to their +J counterparts, respectively. The null hypothesis was rejected for more than 50% of the datasets (BAYAREALIKE: 71.25%, DEC: 60%, and DIVALIKE: 53.75%; P  More

The Fur-regulated T6SS1 plays an important role in iron acquisition in C. necator
To explore the function of T6SS1 (Reut_A1713 to Reut_A1733) in C. necator (Fig. S1A), we analyzed the T6SS1 promoter and identified a Fur binding site (AGAAATA) upstream of gene reut_A1733. This Fur binding site was highly similar to the Fur-box reported in E. coli [38], with a probability score of 2.25 (out of a maximum score = 2.45) (Fig. S1B), which was calculated by applying the position weight matrix to a sequence [39]. Incubation of the T6SS1 promoter probe with purified Fur protein led to decreased mobility of the probe in the electrophoretic mobility shift assay, suggesting a direct interaction between Fur and the T6SS1 promoter (Fig. 1A). To further determine the function of Fur on the expression of T6SS1, a single-copy PT6SS1::lacZ fusion reporter was introduced into the chromosomes of C. necator wild-type (WT), Δfur deletion mutant, and the Δfur(fur) complementary strain. Compared to WT, the PT6SS1::lacZ promoter activity was significantly increased in the Δfur mutant (about 2.2-fold), and this increase could be restored by introducing the complementary plasmid pBBR1MCS-5-fur (Fig. 1B). Similar results were obtained by analyzing the expression of T6SS1 core component genes (hcp1, clpV1, vgrG1, and tssM1) with qRT-PCR (Fig. S1C). These results demonstrate that the expression of T6SS1 in C. necator is directly repressed by Fur, the master regulator of genes involved in iron homeostasis in many prokaryotes [40, 41].Fig. 1: Regulation of T6SS1 expression by Fur.A The interactions between His6-Fur and the T6SS1 promoter examined by EMSA. Increasing amounts of Fur (0, 0.03, 0.06, 0.13, 0.25, and 1.0 μM) and 10 nM DNA fragments were used in the assay. A 500 bp unrelated DNA fragment (Control A) and 1 µM BSA (Control B) were included in the assay as negative controls. B Fur represses the expression of T6SS1. β-galactosidase activities of T6SS1 promoter from chromosomal lacZ fusions in relevant C. necator strains were measured. C Iron uptake requires T6SS1. Stationary-phase C. necator strains were washed twice with M9 medium. Iron associated with indicated bacterial cells were measured with ICP-MS. The vector corresponds to the plasmid pBBR1MCS-5 (B) and pBBR1MCS-2 (C), respectively. Data are represented as mean values ± SD of three biological replicates, each with three technical replicates. **p  More

AffiliationsAnimal Production Systems group, Wageningen University & Research, Wageningen, The NetherlandsAbigail Muscat, Evelien M. de Olde, Raimon Ripoll-Bosch & Imke J. M. de BoerFarming Systems Ecology group, Wageningen University & Research, Wageningen, The NetherlandsHannah H. E. Van ZantenPublic Administration and Policy group, Wageningen University & Research, Wageningen, The NetherlandsTamara A. P. Metze & Catrien J. A. M. TermeerPlant Production Systems group, Wageningen University & Research, Wageningen, The NetherlandsMartin K. van IttersumAuthorsAbigail MuscatEvelien M. de OldeRaimon Ripoll-BoschHannah H. E. Van ZantenTamara A. P. MetzeCatrien J. A. M. TermeerMartin K. van IttersumImke J. M. de BoerCorresponding authorCorrespondence to
Imke J. M. de Boer. More

Abundances of N2 fixing symbioses in the WTNATo date, the various marine symbiotic diatoms are notoriously understudied, and hence our understanding of their abundances and distribution patterns is limited [7]. In general, these symbiotic populations are capable of forming expansive blooms, but largely co-occur at low densities in tropical and subtropical waters with a few rare reports in temperate waters [26,27,28,29, 39,40,41,42]. The Rhizosolenia-Richelia symbioses have been more commonly reported in the North Pacific gyre [26, 27, 31], and the western tropical North Atlantic (WTNA) near the Amazon and Orinoco River plumes is an area where widespread blooms of the H. hauckii-Richelia symbioses are consistently recorded [28, 29, 42,43,44,45,46,47].In the summer of 2010, bloom densities (105−106 cells L−1) of the H. hauckii-Richelia symbioses were encountered at multiple stations with mesohaline (30–35 PSU) surface salinities (Supplementary Table 1). The R. clevei-Richelia symbioses were less abundant (2–30 cells L−1). Similar densities of H. hauckii-Richelia have been reported in the WTNA during spring (April–May) and summer seasons (June–July) (28–29; 46). In fall 2011, less dense symbiotic populations (0–50 cells L−1) were observed, and the dominant symbioses was the larger cell diameter (30–50 µm) H. membranaceus associated with Richelia. Previous observations of H. membranaeus-Richelia in this region are limited and reported as total cells (i.e., 12-218 cells) and highest numbers recorded in Aug–Sept in waters near the Bahama Islands [43]. On the other hand, Rhizosolenia-Richelia are even less reported in the WTNA, and most studies by quantitative PCR assays based on the nifH gene (for nitrogenase enzyme for N2 fixation) of the symbiont (44; 46–7). Unlike qPCR which cannot resolve if the populations are symbiotic or active for N2 fixation, the densities and activity reported here represent quantitative counts and measures of activity for symbiotic Rhizosolenia.The WTNA is largely influenced by both riverine and atmospheric dust deposition (e.g., Saharan dust) [48], including the silica necessary for the host diatom frustules, and trace metals (e.g., iron) necessary for photosynthesis by both partners and the nitrogenase enzyme (for N2 fixation) of the symbiont. We observed similar hydrographic conditions (i.e., low to immeasurable concentrations of dissolved N, sufficient concentrations of dissolved inorganic P and silicates, and variable surface salinities; 22; 28–29; 40–47) as reported earlier that favor high densities of H. hauckii-Richelia blooms. Unfortunately our data is too sparse to determine if these conditions are in fact priming and favoring the observed blooms of the H.hauckii-Richelia symbioses in summer 2010, and to a lesser extent in the Fall 2011.A biometric relationship between C and N activity and host biovolumeThe diatom-Richelia symbioses are considered highly host specific [10, 11], however, the driver of the specificity between partners remains unknown. We initially hypothesized that host selectivity could be related to the N2 fixation capacity of the symbiont. Moreover, it would be expected that the larger H. membranaceus and R. clevei hosts which are ~2–2.5 and 3.5–5 times, respectively, larger in cell dimensions than the H. hauckii cells would have higher N requirements (Supplementary Table 2). In fact, recently it was reported that the filament length of Richelia is positively correlated with the diameter of their respective hosts [22]. Thus, to determine if there is also a size dependent relationship between activity and cell biovolume, the enrichment of both 15N and 13C measured by SIMS was plotted as a function of symbiotic cell biovolume.Given the long incubation times (12 h) and previous work [32] that show fixation and transfer of reduced N to the host is rapid (i.e., within 30 min), we expected most if not all of the reduced N, or enrichment of 15N, to be transferred to the host diatom during the experiment (Fig. 1). Therefore, we measured and report the enrichment for the whole symbiotic cell, rather than the enrichment in the individual partners (Supplementary Table 2; Fig. 2). The enrichment of both 13C/12C and 15N/14N was significantly higher in the larger H. membranaceus-Richelia cells (atom % 13C: 1.5628–2.0500; atom % 15N: 0.8645–1.0200) than the enrichment measured in the smaller H. hauckii-Richelia cells (atom % 13C: 1.0700–1.3078; atom % 15N: 0.3642–0.7925) (Fig. 2) (13C, Mann–Whitney p = 0.009; 15N, Mann–Whitney p 50 symbiotic cells in a chain) were reported at station 2 with fully intact symbiotic Richelia filaments (2–3 vegetative cells and terminal heterocyst), and at station 25 chains were short (1–2 symbiotic cells) and associated with short Richelia filaments (only terminal heterocyst). Moreover, the symbiotic H. hauckii hosts possessed poor chloroplast auto-fluorescence at station 25 [46]. Given that the cells selected for NanoSIMS were largely single cells, rather than chains, we suspect that these cells were in a less than optimal cell state, which was also reflected in the low 13C/12C enrichment ratios and low estimated C-based growth rates (0.30–57 div d−1). These are particularly reduced compared to the growth rates recently reported for enrichment cultures of H. hauckii-Richelia (0.74–93 div d−1§) (Supplementary Table 2) [33].In 2011, higher cellular N2 fixation rates (15.4–27.2 fmols N cell−1 h−1) were measured for the large cell diameter H. membranaceus-Richelia, symbioses. Despite high rates of fixation, cell abundances were low (4–19 cells L−1), and resulted in a low overall contribution of the symbiotic diatoms to the whole water N2 ( >1%) and C-fixation ( >0.01%). The estimated C-based growth rates for H. membranaceus were high (1.9–3.5 div d−1), whereas estimated N-based growth rates (0.3–4 div d−1) were lower than previously published (33; 52–53). Hence the populations in 2011 were likely in a pre-bloom condition given the low cell densities.Estimating symbiotically derived reduced N to surface oceanTo date, determining the fate of the newly fixed N from these highly active but fragile symbiotic populations has been difficult. Thus, we attempted to estimate the excess N fixed and potentially available for release to the surround by using the numerous single cell-specific rates of N2 fixation determined by SIMS on the Hemiaulus spp.-Richelia symbioses (Supplementary Materials). Because the populations form chains during blooms and additionally sink, we calculated the size-dependent sinking rates for both single cells and chains ( >50 cells). Initially we hypothesized that sinking rates of the symbiotic associations would be more rapid than the N excretion rates, such that most newly fixed N would contribute less to the upper water column (sunlit).The sinking velocities were plotted (Fig. 5) as a function of cell radius at a range (min, max) of densities and included two different form resistances (∅ = 0.3 and 1.5). As expected, the combination of form resistance and density has a large impact on the sinking velocity. For example, a H. hauckii cell of similar radius (10 μm) and density (3300 kg m−3) but higher form resistance (0.3 vs. 1.5) sinks twice as fast at the lower form resistance (Fig. 5). This points to chain formation (e.g., increased form resistance) as a potential ecological adaptation to reduce sinking rates. Recently, colony formation was identified as an important phenotypic trait that could be traced back ancestrally amongst both free-living and symbiotic diatoms that presumably functions for maintaining buoyancy and enhancing light capture [22].Fig. 5: The influence of cell characteristics on estimated sinking velocity for symbiotic Hemiaulus spp.The range of diatom sinking speed predicted using the modified Stokes approximation for diatoms [74] and accounting for the symbioses (cylinders) having varying cell size characteristics (form resistance by altering chain length, density; Supplementary Table 4). Note that form resistance increases with chain length and that the longest chains would have sinking speeds less than 10 m d−1.Full size imageThe concentration of fixed N surrounding a H. hauckii and H. membranaceus cell were modeled (Supplementary Materials; Supplementary Table 4; Fig. 6). First, the cellular N requirement (QN, mol N cell−1) for a cell of known volume, V, as per the allometric formulation of Menden-Deuer and Lessard [71] is calculated by the following.$${{{{{{{mathrm{Q}}}}}}}}_{{{{{{{mathrm{N}}}}}}}} = (10^{ – 12}/12) times 0.76 ;times, {{{{{{{mathrm{V}}}}}}}}^{^{0.189}}$$
(1)
Fig. 6: The simplified case of diffusive nitrogen (N) exudate plumes for non-motile symbioses.The concentration of dissolved N (nmol L−1) is presented at of varying cell sizes (3 µm and 30 µm) for H. hauckii-Richelia (A and B, respectively) and H. membranaceus-Richelia (C and D, respectively) growing at specific growth rates of 0.4 d−1 (dashed red lines) or 0.68 d−1 (solid black lines). Exudation follows the same principle as diffusive uptake as per Kiorboe [72] in the absence of turbulence.Full size imageVolume calculations assume a cylindrical shape; whereas exudation assumes an equivalent spherical volume. Then, using published growth rates of 0.4 d−1 and 0.68 d−1 for the symbioses [52, 53], N uptake rate (VN) necessary to sustain the QN was determined. N loss was assumed to be a constant fraction (f) of the VN; this fraction was assumed to be 7.5% and 11% for H. hauckii and H. membranaceus, respectively, or the estimated excess N which was fixed given the assumed growth rate [31]. The excretion rate (EN) of the individual cells was then calculated as$${{{{{{{mathrm{E}}}}}}}}_{{{{{{{mathrm{N}}}}}}}} = {{{{{{{mathrm{fQ}}}}}}}}_{{{{{{{mathrm{N}}}}}}}}$$
(2)
The concentration of fixed N surrounding the cell (Cr) was iteratively calculated by the following:$${{{{{{{mathrm{C}}}}}}}}_{{{{{{{mathrm{r}}}}}}}} = {{{{{{{mathrm{E}}}}}}}}_{{{{{{{mathrm{N}}}}}}}}/(4pi * {{{{{mathrm{D}}}}}}* {{{{{mathrm{r}}}}}}_{{{{{mathrm{{x}}}}}}}) + {{{{{{{mathrm{C}}}}}}}}_{{{{{{{mathrm{i}}}}}}}}$$
(3)
The concentric radius (rx) as per Kiørboe [72] uses a diffusivity of N assumed to be 1.860 × 10−5 cm2 sec−1 and the background concentration of N (Ci) is assumed to be negligible. Figure 5 presents the results for the two symbioses: H. membranaceus and H. hauckii at the two growth rates and as chains or singlets. Mean sinking rates for cells with a high form resistance (e.g., chains) are More

Late Quaternary climate reconstructionsPrecipitationOur reconstructions of Late Quaternary precipitation are based on outputs from a statistical emulator of the HadCM3 general circulation model62. The emulator was developed using 72 3.75° × 2.5° resolution snapshot climate simulations of HadCM3, covering the last 120k years and in 2k year time steps from 120k to 22k years ago and 1k-year time steps from 21k years ago to the present, where each time slice represents climatic conditions averaged across a 30-year post-spin-up period28,63. The emulator is based on grid-cell-specific linear regressions between the local time series of HadCM3 climate data and four time-dependent forcings, given by the mean global atmospheric CO2 concentration and three orbital parameters: eccentricity, obliquity, and precession. The values of these four predictors are known well beyond the last 120k years; thus, applying them to the calibrated grid-cell-specific linear regressions allows for the statistical extrapolation of global climate up to 800k years into the past62. The emulated climate data have been shown to correspond closely to the original HadCM3 simulations for the last 120k years, and to match long-term empirical climate reconstructions well62.Here, we used precipitation data from the emulator, denoted ({bar{{{{bf{P}}}}}}_{{{{{rm{HadCM3}}}}}_{{{{rm{em}}}}}}(t)), of the last 300k years at 1k-year time steps, (tin {{{{bf{T}}}}}_{300k}). The data were spatially downscaled from their native 3.75° × 2.5° grid resolution, and subsequently bias-corrected, in two steps, similar to the approach described in ref. 64, whose description we follow here. Both steps use variations of the delta method65, under which a high-resolution, bias-corrected reconstruction of precipitation at sometime t is obtained by applying the difference between lower-resolution present-day simulated and high-resolution present-day observed climate—the correction term—to the simulated climate at time t. The delta method has been used to downscale and bias-correct palaeoclimate simulations before (e.g. for the WorldClim database66), and, despite its conceptual simplicity, has been shown to outperform alternative methods commonly used for downscaling and bias-correction67.A key limitation of the delta method is that it assumes the present-day correction term to be representative of past correction terms. This assumption is substantially relaxed in the dynamic delta method used in the first step of our approach to downscale ({bar{{{{bf{P}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}(t)) to a ~1° resolution. This method involves the use of a set of high-resolution climate simulations that were run for a smaller but climatically diverse subset of T300k. Simulations at this resolution are computationally very expensive, and therefore running substantially larger sets of simulations is not feasible; however, these selected data can be very effectively used to generate a suitable time-dependent correction term for each (tin {{{{bf{T}}}}}_{300k}). In this way, we can increase the resolution of the original climate simulations by a factor of ∼9, while simultaneously allowing for the temporal variability of the correction term. In the following, we describe the approach in detail.We used high-resolution precipitation simulations from the HadAM3H model63, generated for the last 21,000 years in 9 snapshots (2k year time intervals from 12k to 6k years ago, and 3k year time intervals otherwise) at a 1.25° × .83° grid resolution, denoted ({bar{{{{bf{P}}}}}}_{{{{rm{HadAM}}}}3{{{rm{H}}}}}(t)), where (tin {{{{bf{T}}}}}_{21k},)represents the nine time slices for which simulations are available. These data were used to downscale ({bar{{{{bf{P}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}(t)) to a 1.25° × 0.83° resolution by means of the multiplicative dynamic delta method, yielding$${bar{{{{bf{P}}}}}}_{ sim 1^circ }(t)mathop{=}limits^{{{{rm{def}}}}}{bar{{{{bf{P}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}^{ boxplus }(t)cdot frac{{bar{{{{bf{P}}}}}}_{{{{rm{HadAM}}}}3{{{rm{H}}}}}(hat{t})}{{bar{{{{bf{P}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}^{ boxplus }(hat{t})}.$$
(1)
The ⊞-notation indicates that the coarser-resolution data were interpolated to the grid of the higher-resolution data, for which we used an Akima cubic Hermite interpolant68, which, unlike the bilinear interpolation, is continuously differentiable but, unlike the bicubic interpolation, avoids overshoots. The time (hat{t}in {{{{bf{T}}}}}_{21k}) is chosen as the time at which climate was, in a sense specified below, close to that at time (tin {{{{bf{T}}}}}_{300k}). In contrast to the classical delta method (for which (hat{t}=0) for all (t)), this approach does not assume that the resolution correction term, ((frac{{bar{{{{bf{P}}}}}}_{{{{rm{HadAM}}}}3{{{rm{H}}}}}(hat{t})}{{bar{{{{bf{P}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}^{ boxplus }(hat{t})})), is constant over time. Instead, the finescale heterogeneities that are applied to the coarser-resolution ({bar{{{{bf{P}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}(t)) are chosen from the wide range of patterns simulated for the last 21k years. The strength of the approach lies in the fact that the last 21k years account for a substantial portion of the glacial-interglacial range of climatic conditions present during the whole Late Quaternary. Following ref. 64, we used global CO2, a key indicator of the global climatic state, as the metric according to which (hat{t}) is chosen; i.e. among the times for which HadAM3H simulations are available, (hat{t}) is the time at which global CO2 was closest to the respective value at the time of interest, t.In the second step of our approach, we used the classical multiplicative delta method to bias-correct and further downscale ({{{{bf{P}}}}}_{ sim 1^circ }(t)) to a hexagonal grid69 with an internode spacing of ~55 km (~0.5°),$${bar{{{{bf{P}}}}}}_{ sim 0.5^circ }(t)mathop{=}limits^{{{{rm{def}}}}}{bar{{{{bf{P}}}}}}_{ sim 1^circ }^{ boxplus }(t)cdot frac{{bar{{{{bf{P}}}}}}_{{{{rm{obs}}}}}(0)}{{bar{{{{bf{P}}}}}}_{ sim 1^circ }^{ boxplus }(0)},$$
(2)
where ({{{{bf{P}}}}}_{{{{rm{obs}}}}}(0)) denotes present-era (1960–1990) observed precipitation70.We reconstructed land configurations for the last 300k years using present-day elevation71 and a time series of Red Sea sea level72. For locations that are currently below sea level, the delta method does not work. For these locations, precipitation was extrapolated using a inverse distance weighting approach. With the exception of a brief window from 124–126k years ago, sea level in the past was lower than it is today; thus, present-day coastal patterns are spatially extended as coastlines move, but not removed. For all (tin {{{{bf{T}}}}}_{300k}), maps of annual precipitation ({bar{{{{bf{P}}}}}}_{ sim 0.5^circ }(t)) with the appropriate land configuration are available as Supplementary Movie 1.Based on these data representing 30-year climatological normals at 1k-year time steps between 300k years ago and the present, we generated, for each millennium, 100 maps representing 10-year average climatologies as follows. We used 3.75° × 2.5° climate simulations from the HadCM3B-M2.1 model, providing a 1000-years-long annual time series of annual precipitation for each millennium between 21k years ago and the present73. Millennia were simulated in parallel; thus, the 1000-years-long time series representing each millennium is in itself continuous, but the beginnings and ends of the time series of successive millennia generally do not coincide. For (tin {{{{bf{T}}}}}_{21k}), we denote the available 1000 successive maps of annual precipitation by ({{{{bf{P}}}}}_{{{{rm{HM}}}}}^{(1)}(t),ldots ,{{{{bf{P}}}}}_{{{{rm{HM}}}}}^{(1000)}(t)). We used these data to compute the relative deviation of the climatic average of each decade within a given millennium, and the climatic average of the 30-year period containing the specific decade as$${{{{boldsymbol{epsilon }}}}}_{HM}^{(d)}(t)mathop{=}limits^{{{{rm{def}}}}}frac{{sum }_{i=1+(d-1)cdot 10}^{dcdot 10}{{{{bf{P}}}}}_{{{{rm{HM}}}}}^{(i)}(t)}{{sum }_{n=1+(d-2)cdot 10}^{(d+1)cdot 10}{{{{bf{P}}}}}_{{{{rm{HM}}}}}^{(n)}(t)},d=1,ldots ,100$$
(3)
Finally, we applied these ratios of 10-year to 30-year climatic averages to the previously derived 1k-year time step climatologies to obtain, for each (tin ,{{{{bf{T}}}}}_{300k}), 100 sets of 10-year average annual precipitation,$${{{{bf{P}}}}}_{ sim 0.5^circ }^{(d)}(t)mathop{=}limits^{{{{rm{def}}}}}{bar{{{{bf{P}}}}}}_{ sim 0.5^circ }(t)cdot {{{{boldsymbol{epsilon }}}}}_{HM}^{(d), boxplus }(hat{t}),,d=1,ldots ,100$$
(4)
where, analogous to our approach in Eq. (1),(, boxplus ) denotes the interpolation to the ~55 km hexagonal grid, and where (hat{t}) is chosen as the time at which global CO2 was closest to the respective value at time t.AridityThe Köppen aridity index used here is defined as the ratio of annual precipitation (in mm) to the sum of mean annual temperature (in °C) and a constant of 33 °C (cf. Eq. (8)). This measure of aridity was found to be the most reliable one of a set of alternative indices in palaeoclimate contexts30.Decadal-scale mean annual temperature data between 300k years ago and the present were created using analogous methods to those previously applied to reconstruct precipitation. 3.75° × 2.5° resolution emulator-derived simulations of mean annual temperature of the past 300k years at 1k time steps62, denoted ({bar{{{{bf{T}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}(t)), were first downscaled by means of the additive dynamic delta method, using 1.25° × 0.83° HadAM3H simulations of mean annual temperature of the past 21k years, denoted ({bar{{{{bf{T}}}}}}_{{{{rm{HadAM}}}}3{{{rm{H}}}}}(t)), yielding, analogous to Eq. (1),$${bar{{{{bf{T}}}}}}_{ sim 1^circ }(t)mathop{=}limits^{{{{rm{def}}}}}{bar{{{{bf{T}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}^{ boxplus }(t)+left({bar{{{{bf{T}}}}}}_{{{{rm{HadAM}}}}3{{{rm{H}}}}}(hat{t})-{bar{{{{bf{T}}}}}}_{{{{rm{HadCM}}}}{3}_{{{{rm{em}}}}}}^{ boxplus }(hat{t})right).$$
(5)
Analogous to Eq. (2), Next, present-day observed mean annual temperature, ({bar{{{{bf{T}}}}}}_{{{{rm{obs}}}}}(0)), was used to further downscale and bias-correct the data by means of the additive delta method to obtain$${bar{{{{bf{T}}}}}}_{ sim 0.5^circ }(t)mathop{=}limits^{{{{rm{def}}}}}{bar{{{{bf{T}}}}}}_{ sim 1^circ }^{ boxplus }(t)+left({bar{{{{bf{T}}}}}}_{{{{rm{obs}}}}}(0)-{bar{{{{bf{T}}}}}}_{ sim 1^circ }^{ boxplus }(hat{t})right).$$
(6)
For all (tin {{{{bf{T}}}}}_{300k}), maps of mean annual temperature ({bar{{{{bf{T}}}}}}_{ sim 0.5^circ }(t)) with the appropriate land configuration are available as Supplementary Movie 1.Finally, we incorporated HadCM3B-M2 simulations of mean annual temperature of the past 21k years, ({{{{bf{T}}}}}_{{{{rm{HM}}}}}^{(1)}(t),ldots ,{{{{bf{T}}}}}_{{{{rm{HM}}}}}^{(1000)}(t)) for (tin {{{{bf{T}}}}}_{21k}), to obtain 10-year average mean annual temperature,$$begin{array}{c}{{{{bf{T}}}}}_{ sim 0.5^circ }^{(d)}(t)mathop{=}limits^{{{{rm{def}}}}}{bar{{{{bf{T}}}}}}_{ sim 0.5^circ }(t)+{left(mathop{sum }limits_{i=1+(d-1)cdot 10}^{dcdot 10}{bar{{{{bf{T}}}}}}_{{{{rm{HM}}}}}^{(i)}(t)-mathop{sum }limits_{n=1+(d-2)cdot 10}^{(d+1)cdot 10}{bar{{{{bf{T}}}}}}_{{{{rm{HM}}}}}^{(n)}(t)right)}^{ boxplus },\ d=1,ldots ,100end{array}$$
(7)
Based on these data, the Köppen aridity index at the same spatial and temporal resolution is calculated as$${{{{bf{A}}}}}_{ sim 0.5^circ }^{(d)}(t)mathop{=}limits^{{{{rm{def}}}}}frac{{{{{bf{P}}}}}_{ sim 0.5^circ }^{(d)}(t)}{{{{{bf{T}}}}}_{ sim 0.5^circ }^{(d)}(t)+33}.$$
(8)
Comparison with empirical proxiesLong-term proxy records
Long-term proxy records allow us to assess whether simulations capture key qualitative dynamics observed in the empirical data. The lack of direct long-term time series reconstructions of annual precipitation and mean annual temperature makes it necessary to use proxies related to these two climate variables. Proxies providing temporal coverage beyond the last glacial maximum are not only extremely sparse in North Africa and Southwest Asia, but even the few records that exist are affected by environmental factors other than the specific climate variables considered here. For example, reconstructions of past wetness and aridity use proxies that reflect not only rainfall conditions but also the interaction of precipitation with other local and non-local hydro-climatic variables, e.g. river discharge or hydrological catchment across a larger area. Here, we have not attempted to correct for such processes, but assumed that the simulated climate at the site where the empirical record was taken provide a suitable approximation of the potentially broader climatic conditions relevant for the proxy data. Realistic climate simulations would therefore be expected to match major qualitative trends of the empirical records, rather than exhibit a perfect correlation with the data. We compared our precipitation simulations against three long-term humidity-related empirical proxies (Fig. 4a). Proxy 174 provides a time series of Dead Sea lake levels, for which wet and dry periods are associated with high-stand and low-stand conditions, respectively. Proxy 219 from the southern tip of the Arabian Peninsula was obtained from a marine sediment core that allows for reconstructing past changes in aridity over land from the stable hydrogen isotopic composition of leaf waxes (δDwax). Proxy 318 is an XRF-derived humidity index from a core near the Northwest African coast. Temperature simulations were compared against two long-term records of δ18O, which varies over time as a result of temperature fluctuations (in addition to other factors), from the Peqiin and Soreq caves in Northern Israel75 (Fig. 4e). Overall, the simulated data capture key phases observed in the empirical records well for both precipitation (Fig. 4b–d) and temperature proxies (Fig. 4f–h).Fig. 4: Comparison of our data to long-term proxy records.a Geographical locations of empirical proxies on a map of present-day annual precipitation. b–d Comparisons of simulated annual precipitation against the three wetness proxies. e Geographical locations of empirical proxies on a map of present-day mean annual temperature. f, g Comparisons of simulated mean annual temperature against the two δ18O records. Black lines represent the simulated climatological normals at 1k-year intervals (Eqs. (2) and (6)), grey shades represent the 10th and 90th percentile of the decadal simulations (n = 100; Eqs. (4) and (7)).Full size image

Pollen-based reconstructions
Pollen records used to empirically reconstruct past climate do not reach as far back in time as the above-described proxy records and are not available at the same temporal resolution; however, in contrast to those proxies, they can be used to quantitatively estimate local annual precipitation and mean annual temperature directly. Here, we used the dataset of pollen-based reconstructions of precipitation and temperature for the mid-Holocene (6k years ago) and the last glacial maximum (21k years ago)76 (Fig. 5a). Our precipitation and temperature data are overall in good agreement with the empirical reconstructions (Fig. 5b–e). During the mid-Holocene, our simulations suggest slightly less precipitation at low levels than most of the empirical records (Fig. 5d), while our data match the empirical reconstruction available from a very arid location during the last glacial maximum very well (Fig. 5e).Fig. 5: Comparison of our data to pollen-based climate reconstructions from the mid-Holocene and the last interglacial period.a Geographical locations and timings of pollen records. b–e Comparisons of our data against empirical reconstructions. Vertical centre measures and error bars represent the empirical reconstructed values and their uncertainties, respectively; horizontal centre measures and error bars represent simulated climatological normals at 1k-year intervals (Eqs. (2) and (6)) and the 10th and 90th percentile of the simulated decadal data (n = 100; Eqs. (4) and (7)), respectively.Full size image

Interglacial palaeolakes on the Arabian Peninsula
Finally, we plotted time series of our precipitation simulations in three locations in which palaeolakes have been dated to the last interglacial period, following the approach in ref. 24, in which the authors tested whether their climate simulations predicted higher rainfall during the last interglacial period than at present at palaeolake sites on the Arabian Peninsula. Figure 6 shows the locations of three palaeolakes in the northeast (western Nefud near Taymal; proxy 1), the centre (at Khujaymah; proxy 2), and the southwest (at Saiwan; proxy 3) of the peninsula24 (described in detail in refs. 23,77), and our precipitation data in these locations. In two out of the three locations, our data predict that more rainfall occurred at the estimated timings of the palaeolakes than at any point in time since; in the third location, slightly more rainfall than during the dated time interval is simulated only for a period around 8k years ago.Fig. 6: Comparison of our data against the dates of three palaeolakes on the Arabian peninusla.a Geographical locations of the lakes. b–d Time series of our precipitation data. Black lines represent the simulated climatological normals at 1k-year intervals (Eqs. (2) and (6)), grey shades represent the 10th and 90th percentile of the decadal simulations (n = 100; Eq. (4)). Horizontal error bars represent the estimated dates of the lakes24.Full size image
Determining the minimum precipitation and aridity tolerance required for out-of-Africa exitsWe denote by ({{{bf{X}}}}={({lambda }_{1},{phi }_{1}),({lambda }_{2},{phi }_{2}),ldots },)the set of longitude and latitude coordinates of the hexagonal grid with an internode spacing of ~55 km (~0.5°)69 that are contained in the longitude window [15°E, 70°E] and the latitude window [5°N, 43°N] (shown in Fig. 3). We denote by ({{{bf{E}}}}) the set of the present-day elevation values of the coordinates in ({{{bf{X}}}}) (in meters)78, i.e. ({{{bf{E}}}}({x}_{i})) is a positive number in a point ({x}_{i}=({lambda }_{i},{phi }_{i})) if ({x}_{i}) is currently above sea level, and negative if ({x}_{i}) is currently below sea level. We denote by (s(t)) the sea level (in meters) at the time (tin {{{{bf{T}}}}}_{300k}) (where ({{{{bf{T}}}}}_{300k}) represents the last 300k years in 1k time steps), for which we used a long-term reconstruction of Red Sea sea level72. In particular, we have (s(0)=0) at present day. For each millennium (tin {{{{bf{T}}}}}_{300k}), we denote by (bar{{{{bf{X}}}}}(t)) the subset of points in (X) that are above sea level:$$bar{{{{bf{X}}}}}(t)mathop{=}limits^{{{{rm{def}}}}}{xin {{{bf{X}}}}:{{{bf{E}}}}(x), > , s(t)}$$
(9)
Based on the precipitation map ({{{{bf{P}}}}}_{ sim 0.5^circ }^{(d)}(t)) for a decade (d=1,ldots ,100) in millennium (t) (Eq. (4)), and a given precipitation threshold value (p) (in mm year−1), we denote by ({mathop{{{{bf{X}}}}}limits^{=}}_{p}^{(d)}(t)) the subset of (bar{{{{bf{X}}}}}(t)) that would be suitable grid cells for humans assuming that they cannot survive in areas where precipitation levels are below (p):$${mathop{{{{bf{X}}}}}limits^{=}}_{p}^{(d)}(t)mathop{=}limits^{{{{rm{def}}}}}left{xin bar{{{{bf{X}}}}}(t):{{{{bf{P}}}}}_{ sim 0.5^circ }^{(d)}(t)ge pright}$$
(10)
We then determined whether there was a connected path in ({mathop{{{{bf{X}}}}}limits^{=}}_{p}^{(d)}(t)) between an initial point, for which we used ({x}_{{{{rm{start}}}}}=(32.6^circ {{{rm{E}}}},10.2^circ {{{rm{N}}}})), and any point in a set of coordinates outside of Africa, defined as ({{{{bf{X}}}}}_{{{{rm{end}}}}}mathop{=}limits^{{{{rm{def}}}}}{(lambda ,phi )in {{{bf{X}}}}:lambda, > , 65^circ {{{rm{E}}}},{{{rm{or}}}},phi , > , 37^circ N}). This was defined to be the case if there was a finite sequence$${x}_{{{{rm{start}}}}}to {x}_{1}to {x}_{2}to ldots to {x}_{n}in {{{{bf{X}}}}}_{{{{rm{end}}}}}$$
(11)
of points ({x}_{i}in {mathop{{{{bf{X}}}}}limits^{=}}_{p}^{(d)}(t)) such that the distance between any two successive points ({x}_{i}) and ({x}_{i+1}) was less or equal to the maximum internode spacing of the grid (X). Based on this approach, the critical precipitation threshold below which no connected path exists for the precipitation map ({{{{bf{P}}}}}_{ sim 0.5^circ }^{(d)}(t)) was determined using the following bisection method. Beginning with ({hat{p}}_{0}=1000) mm y−1 and ({check{p}}_{0}=0) mm y−1, for which a connected path between ({x}_{{{{rm{start}}}}}) and ({{{{bf{X}}}}}_{{{{rm{end}}}}}) exists, respectively, for all and for no (t) and (d), the values ({hat{p}}_{k}) and ({check{p}}_{k}) were iteratively defined as$$ , left.begin{array}{c}{check{p}}_{k+1}mathop{=}limits^{{{{rm{def}}}}}frac{{{hat{p}}_{k}+{check{p}}_{k}}}{2}\ {hat{p}}_{k+1}mathop{=}limits^{{{{rm{def}}}}}{hat{p}}_{k}hfillend{array}right},{{{rm{if}}}},{{{rm{a}}}},{{{rm{connected}}}},{{{rm{path}}}},{{{rm{exists}}}},{{{rm{for}}}},p=frac{{{hat{p}}_{k}+{check{p}}_{k}}}{2}\ , left.begin{array}{c}{check{p}}_{k+1}mathop{=}limits^{{{{rm{def}}}}}{check{p}}_{k}hfill\ {hat{p}}_{k+1}mathop{=}limits^{{{{rm{def}}}}}frac{{{hat{p}}_{k}+{check{p}}_{k}}}{2}end{array}right}, {{{rm{else}}}}$$
(12)
For all (k), the sought critical precipitation threshold, denoted ({p}_{{{{rm{crit}}}}}^{(d)}(t)), is bounded above by ({hat{p}}_{k}) and bounded below by ({check{p}}_{k}). For (kto infty ), both values converge to ({p}_{{{{rm{crit}}}}}^{(d)}(t)). Here, we defined$${p}_{{{{rm{crit}}}}}^{(d)}(t)mathop{=}limits^{{{{rm{def}}}}}frac{,{hat{p}}_{10}+{check{p}}_{10}}{2},$$
(13)
which lies within 1 mm y−1 of the true limit value.To specifically determine the precipitation tolerance required for a northern (Fig. 1a) or southern (Fig. 1b) exit, we rendered the passage of the respective other route impassable by removing appropriate cells from the grid. When investigating the southern route, we additionally assumed that no sea level and precipitation constraints applied within a ~40 km radius around the centre of the Bab al-Mandab strait.For aridity, the procedure is identical, with the exception that ({mathop{{{{bf{X}}}}}limits^{=}}_{p}^{(d)}(t)) is defined based on the relevant aridity map, ({{{{bf{A}}}}}_{ sim 0.5^circ }^{(d)}(t)), and the value 4.0 is used for the initial upper threshold (denoted ({hat{p}}_{0}) above).Width of the Strait of Bab al-MandabSimilar to ref. 52, we reconstructed the minimum distance required to cover on water in order to reach the Arabian peninsula (present-day west coast of Yemen) from Africa (present-day Djibouti and southeast Eritrea). We used a 0.0083° (~1 km at the equator) map of elevation and bathymetry78 and a time series of Red Sea sea level72 to reconstruct very-high-resolution land masks for the last 300k years. For each point in time, we determined the set of connected land masses, and the distances between the closest points of any two land masses. The result can be graph-theoretically represented by a complete graph whose nodes represent connected land masses and whose edge weights correspond to the minimum distances between land masses. The path involving the minimum continuous distance on water was then determined by solving the minmax path problem whose solution is the path between the two nodes representing Africa and the Arabian Peninsula that minimises the maximum weight of any of its edges (Fig. 1b grey shades).Analyses were conducted using Matlab R2019a79.Reporting summaryFurther information on research design is available in the Nature Research Reporting Summary linked to this article. More