Coats of many colours

The unvaccinated are at risk as evolution accelerates the covid-19 pandemic

Research is unravelling the virus’s deep secrets


For much of 2020 the covid-19 virus was, in genetic terms, a little dull. 

Early in the pandemic a version of sars-cov-2 that was slightly different from the one originally sequenced in Wuhan, and spread a bit better, came to dominate the picture outside China. 

But after that it was just a case of a letter or two of genetic code changing here and there. 

Sometimes such mutations proved useful for working out where infections were coming from. 

But none of them seemed biologically relevant. 

By September Salim Abdool Karim, a South African epidemiologist, was beginning to find his monthly updates on new mutations “quite boring”. 

He considered dispensing with them altogether.

He was soon glad that he hadn’t. In the last months of 2020 researchers around the world began to see variants of the virus with not just one or two mutations but ten or 20. 

What was more, some of these new variants turned out to have new properties—to spread faster, to shrug off antibodies, or to do both.

The first of them, now called Alpha, appeared in Britain in September. 

By November scientists sequencing virus samples were becoming alarmed at the rate of its spread. 

Each infection with the original virus, as sequenced in Wuhan in January 2020, had been estimated to lead to roughly 2.5 subsequent infections in the absence of countermeasures like masks, social distancing and lockdowns. 

Under the same conditions the “reproductive number” for Alpha was reckoned to be almost twice as large: four or five.


By November Dr Karim was sitting in his office gobsmacked by evidence of a variant similarly studded with mutations, now called Beta, in South Africa. 

The Gamma variant, formally identified only in 2021, was beginning to make itself felt in Brazil and would go on to ravage South America. 

Delta, a key factor in the catastrophic Indian epidemic a few months later, raised the transmissibility bar yet further. 

British scientists estimate that in unvaccinated populations not taking precautions its reproductive number may be as high as eight. In mid-June, only two months after it first appeared there, Delta had almost fully displaced Alpha in England (see chart 1). 

It now threatens the rest of the world (see map).


All the variants are more transmissible to some extent. 

Laboratory tests on human airway cells in Petri dishes have shown that Delta replicates more avidly in them than do earlier variants. 

That would seem to suggest that a smaller initial dose is needed for an infection to take hold. 

It also means that the amount of virus lurking in people’s airways is probably higher.

Swabs taken from people’s nostrils and throats during testing back this notion up. 

The amount of virus found in samples from people infected with Delta is higher than for other variants. 

That probably means that people are exhaling more virus than those infected by an older variant and thus that every encounter between an infected and uninfected person poses a greater risk of transmission.

Vaccination slows this spread down, but it does not stop it. 

The current vaccines do not stop all infections by any version of the virus. 

Nor do they stop infected people from passing the virus on, though they do make it significantly more difficult. 

People vaccinated with Pfizer or AstraZeneca jabs who are subsequently infected with Alpha are about half as likely to pass it on as the unvaccinated are.

British studies have found Delta to be around 60% more transmissible than Alpha. 

They put roughly three-quarters of that effect down to the fact that it is easier to catch if you are not vaccinated and about a quarter to the increased ease with which Delta infects people who have been vaccinated. 

Around half of the adults infected in a recent Delta outbreak in Israel were fully vaccinated with the Pfizer vaccine.

Happily, studies of vaccines made by Western companies show them to reduce deaths and severe cases of the disease in people infected with every sort of sars-cov-2. 

This protection means none of the new variants is anything like as potent a public-health threat to a largely vaccinated population as the original version was to an unvaccinated one. 

Delta’s increased transmissibility, along with relaxed restrictions on travel and socialising, has seen the number of infections and cases in Britain beginning to climb again. 

But thanks to widespread vaccination, deaths have barely moved. 

Deaths are, by their nature, a lagging indicator of infection; but widespread vaccination of the most vulnerable is working as hoped.


The dangers posed to the unvaccinated and partially vaccinated mean that there is still a public-health case for keeping infections from spreading. 

Here, unfortunately, the degree to which variants can evade vaccine-produced immunity makes things a lot harder than once they seemed. 

“If there is a certain degree of immune escape, even if you were to vaccinate 100% of the population, it’s going to keep coming at you for some period of time,” says Adam Kucharski of the London School of Hygiene and Tropical Medicine.

In a population where 60% are immune, either through vaccination or from a past bout of covid-19, the introduction of a variant with a reproductive number of eight would cause a sharp surge in infections unless lockdowns and similar interventions were established right away (see chart 2). 

For unvaccinated populations the situation is much worse. 

If no precautions are taken, a reproductive number of eight produces a far more dramatic crisis in an unvaccinated population than one of two or three does. 

And last year provided ample evidence of how bad things get even with a lower R. 

Other things being equal, a highly transmissible virus means more deaths and a more acute stress on the health-care system.

Spikes for speedy spread

Other things may not be equal; the danger posed to the unvaccinated by a new variant may not be exactly the same as that posed by older versions. 

In Britain those infected with the Alpha variant saw a higher level of severe disease than those infected with the original version, but no corresponding increase in deaths.

Whether Delta does the same is unclear. 

Comparisons with other variants in countries that can measure such things well are made hard to assess by the large numbers of vaccinated people in those populations. 

The picture emerging from a British symptom-tracking app called Zoe suggests that Delta is presenting with symptoms closer to those for the common cold than those seen with other variants. 

They rarely have shortness of breath, the hallmark symptom of covid-19 with the variants that dominated the first year of the pandemic. 

Oddly, vaccinated people who then get infected tend to sneeze more—which is good for the virus not just because sneezes spread diseases but also because it allows covid-19 to be mistaken for hayfever.

So far, though, differences in the severity of disease caused by the different variants have been eclipsed by the simple, deadly fact of their high-speed spread. There is ample room for that to continue. 

Less than 1% of people in low-income countries have had even one dose of vaccine. 

In sub-Saharan Africa Delta is fuelling outbreaks that are crushing hospitals and killing health-care workers.

Rich countries, including Australia, Japan and South Korea, where the first wave was largely avoided and vaccination has not been a high priority now look highly vulnerable. 

By the end of June the risk of Delta had seen almost half of Australia put under lockdown orders. 

Delta is the dominant strain in Russia, where a vaccination rate of 12% and misinformation-driven vaccine scepticism seem set to make its spread easy.

The variants make vaccination programmes more urgent than ever. 

But though they may march on through the alphabet for some time to come, there is some reason to hope that they will not get all that much worse as they do so. 

They may be running out of evolutionary room to manoeuvre.

For a clearer understanding of what is going on, focus on the spike protein that adorns the outer envelope of sars-cov-2 particles. 

You can think of it, as you can of any protein, as being like a paper chain in which every link can have one of 20 colours. 

The gene for spike specifies the sequence in which those colours appear in the protein’s 1,273-link long chain. 

Mutations in the gene can change the colour of one specific link, add a few new links, or cut some links out. 

In the Alpha variant six of those links have different colours from those in the Wuhan sequence, and in a couple of places a link or two are missing altogether. 

The Delta spike has five distinctive mutations.

In reality the links in the chain are 20 different types of amino acid. 

Each type has subtly different chemical and physical properties. 

At the time that the chain is created the laws of physics require it to fold up into something more compact. 

The specific shape into which it folds is determined by its unique sequence of amino acids, as laid out in the gene. 

And that shape underlies all the protein’s future capabilities. 

Shape is almost everything in the world of proteins. 

It is through their shapes that proteins recognise each other. 

It is through changes of shape that they act.


Each of the now-familiar protuberances on the surface of sars-cov-2 particles is composed of three copies of the spike protein slotted together into a “trimer” shaped a bit like a golf tee (see chart 3). 

In the cup of these tees are the virus’s receptor-binding domains (rbds). 

Each of the trimer’s constituent proteins can be open or closed at any given time. 

When they are open ace2, a protein found on the surface of some human cells, fits quite nicely into the rbd’s carefully contrived nobbliness.

Acey deucey

The ace2 receptor is the virus’s main target; it normally attacks only those cells that display it. 

The act of glomming on to an ace2 molecule changes the spike protein’s shape, revealing a “cleavage site” which is suited to attack by another protein on the cell’s surface. 

As a result the spike gets cut in two—which sounds bad for the virus, but is in fact the necessary next step in infection. 

It is only after the spike is sliced asunder that the membranes of the virus and the cell can merge.

Tyler Starr, a researcher at the Fred Hutchinson Cancer Research Centre in Seattle, describes the rbd as a “big, squishy interface” that mutations can reshape quite easily. 

In 2020 he, Jesse Bloom and their colleagues sought to examine this mutability by making versions of the sars-cov-2 rbd in which individual amino acids in the protein paper-chain were replaced by alternatives with different properties. 

These mutant proteins were then tested to see how well they stuck to ace2; those that did best, the researchers reasoned, might be mutations that evolution would favour. 

They were right.

In the original Wuhan genome the 501st position in the spike chain is occupied by an amino acid called asparagine. 

When the scientists in Seattle put an amino acid called tyrosine there instead, the rbd bound to ace2 more tightly; it turns out that the change twists a key part of the rbd round by about 20 degrees, making the fit a bit more snug. 

Mutations which cause just that substitution, known as n501y (or sometimes “Nelly”) subsequently turned up in the Alpha, Beta and Gamma variants. 

Another change they spotted, now called e484k (or “Eek”), was found in both Beta and Gamma.

Changes to the rbd can also reduce its susceptibility to antibodies. 

Antibodies also work by recognising shapes, and though they recognise various other bits of the spike protein, notably another region in the trimer’s head called the n-terminal domain (ntd), the most effective of them are specific to particular aspects of the rbd. 

Some changes to the rbd, such as n501y, do not make it less recognisable to antibodies. 

Others, such as e484k, do. 

Being a lot less susceptible to some antibodies seems to help e484k’s possessors to infect people who have been vaccinated.

The rbd is not the only part of the spike protein where mutations matter. 

In a preprint published on June 22nd Ravindra Gupta, a molecular virologist at Cambridge University, and his colleagues put forward an argument as to why Delta is both more infectious and better at evading immunity than other variants. 

It is based on a substitution at site 681, which is at the point where, after the rbd meets ace2, the protein is cleft in two.

Not ai, therefore em

Dr Gupta says p681r, helped by two shape-modifying mutations elsewhere, makes it easier for the protein to be cut up and thus get into cells. 

Its presence also means that, once a cell starts producing particles, their spike proteins can get on to the cell’s surface pre-cut. 

That can lead to virus particles which are shorn of the rbds which antibodies recognise and ready to fuse with any nearby cell. 

It can also encourage infected cells to clump together with others. 

Dr Gupta’s lab has found evidence of these cell clumps in a living model of the human respiratory system.

A full validation of this work will require a detailed picture of the Delta variant’s structure—something which is not yet available. 

In theory, it should be possible to predict the shape of a protein using nothing but the sequence of amino acids described by its gene and the laws of physics. 

Doing so from first principles, though, is impossible. 

DeepMind, an ai company which is part of Google, has shown that machine learning can help a lot. 

But as yet its capabilities are best demonstrated on small single proteins. 

This approach is not much good if the protein is large, anchored in a membrane, and naturally found in a dimer or trimer, as spike is. 

DeepMind has not made any predictions of spike’s structure public.

The best tool for seeing spike’s structure in detail is cryo-electron microscopy. 

Copies of the protein in question are flash frozen using liquid nitrogen (hence cryo); once they are immobilised beams of electrons are bounced off them and used to build up pictures (hence microscopy). 

Bing Chen, who has run a series of cryo-em experiments on the spike protein at Harvard, is at pains to stress the time, effort and computer power required to turn thousands of pictures of the protein taken from every conceivable angle into a three-dimensional image which comes close to resolving the positions of every single atom. 

But there is no better way to appreciate the changes in the fine details of the protein’s structure brought about by the variants’ different mutations.

On June 24th Dr Chen’s group published long-awaited structures for the Alpha and Beta spike variants. 

They show the way in which the protein’s complex folding allows mutations that are at some distance from each other in paper-chain terms to have effects on the overall shape that it would be near impossible to predict from the sequence alone. 

A pair of mutations found called a570d and s982a, for example, act to slightly loosen up the protein’s structure in Alpha. 

That makes the rbd open up more. 

The group is now working on a structure for Delta which might confirm Dr Gupta’s insights.

Studies of this sort help reveal how the mutations in the variant spikes work together. 

But how did these variants come to have so many mutations in the first place?

Mutations are normally expected to crop up one at a time; but the named variants each emerged with a whole set of them. 

That is what has given them sudden and surprising effects.

One way in which they could have emerged fully formed is by evolving in people with compromised immune systems who had very long drawn out sars-cov-2 infections. 

In such cases the virus would be able to continue replicating itself in their bodies again and again, accumulating a number of mutations as it did so. 

The time required for such a process would help explain why the variants only started to appear towards the end of last year. 

Studies of five such people have shown that they developed a number of the mutations now seen in variants.

Not all the mutations in the variants are in the spike gene, and some of those affecting other proteins will doubtless also prove to have importance. 

One of Alpha’s mutations appears to give it an advantage when dealing with a non-antibody-using arm of the immune system. 

Non-spike mutations probably explain why Delta’s symptoms appear different. 

But spike still dominates the discussion. 

Its structure is crucial to the vaccines. 

And it also seems unusually mutable.

Dr Starr thinks this mutability may be a consequence of the virus’s origin in bats. 

He points out that most viruses have binding domains that cannot tolerate much mutation, and so they evolve ways of hiding them away from pesky antibodies. 

The sars-cov-2 rbds are too large for such protection. 

That would seem like a problem for the virus. 

But it may be a price worth paying if a larger, more open rbd is easier for evolution to reshape.

The reason that Dr Starr thinks evolvability might be a benefit worth paying for is that, in bats, ace2 is much more diverse than it is in humans. 

That means viruses which use the receptors as a target need to be able to adapt the mechanisms by which they do so. 

The tolerance for mutations that has made new variants of rbd possible in humans may be the “by-product of this arms race...between virus and bats”.

Avoiding Omega

If mutation is comparatively easy, though, it also has its limits. 

In their experiments last year Dr Starr and his colleagues identified changes to the rbd that seemed advantageous but which do not turn up in the real world—presumably because real spike proteins cannot contort themselves enough to accommodate them.

Seeing similar mutations crop up in different variants also suggests that evolution is sampling a somewhat limited number of possibilities. 

“The fact is that you’re starting to see recurring mutations,” says Dr Chen. 

“That would be an indication that there are probably not that many places that the virus can mutate.” 

Strains with radically different ways of becoming more transmissible or evasive may be beyond evolution’s reach.


Another cause for optimism is that spike is not the only part of the process that is complex and mutable. 

The immune system is, too. 

The initial infection is the first stage of a protracted struggle in which the immune system has various strategies at its disposal. 

A study by Jackson Turner of the Washington University School of Medicine and his colleagues which was published in Nature on June 28th showed that the immune response produced by infection with sars-cov-2 is long lasting, robust and multifaceted. 

Among other things, some of the b-cells which produce antibodies produce more effective ones later in the course of infection than earlier on. 

This may be part of the reason why they provide better protection against severe disease than they do against infection.

It is quite possible, though, that not all vaccines will do so equally well. 

Hundreds of millions of doses of two vaccines made by Chinese companies, Sinopharm and Sinovac, have been sold to low and middle-income countries; they look like being a large part of the world’s vaccine supply for the rest of the year. 

But there are some doubts about their efficacy, especially against new variants. 

The original clinical trial of the Sinovac vaccine found a lower efficacy than in any other covid-19 vaccine trial, just 51%. 

Studies of the vaccine’s use in Uruguay and Indonesia have been a great deal more encouraging. 

But there is rising concern in Bahrain, Chile, the Seychelles, Turkey and the uae, all of which have relied on Chinese jabs. 

The uae and Bahrain are worried enough to have started offering a third shot of Pfizer’s vaccine to people who have already been given two shots of Sinopharm’s.

Third shots are being looked at by some other governments, too, including Britain’s. 

The fact that current vaccines protect people against severe disease and death even when infected by the new variants makes the idea that variant-specific vaccines analogous to seasonal flu jabs will be necessary look less likely. 

The easier alternative of offering people who have been vaccinated twice a third shot, though, perhaps using one of the other vaccines, has advocates.

But there is as yet no evidence that it is necessary. 

And third shots pale as a priority compared with first and second shots for those who have had neither, and now need them more than ever.  

Missing ingredients

The bottlenecks which could constrain emission cuts

The green revolution risks running short of minerals, money and places to build



At 107 metres, the three carbon-fibre blades of a Haliade-X marine wind turbine are longer than the wingspan of any airliner ever made. 

The generator which transforms their rotation—over 300km an hour at the tip—into power requires over 100 powerful magnets made of exotic metals and untold lengths of coiled-up copper. 

The blades, generator and associated gubbins, weighing around 900 tonnes all-in, have to be installed on a pylon so tall that the blade-tips reach almost as high above the waves as the pinnacle of the Transamerica Pyramid rises over the 600 block of San Francisco’s Montgomery Street.

In May, President Joe Biden’s administration announced the approval of Vineyard Wind, a wind farm off the coast of Massachusetts which will require ge, an American industrial giant, to supply 60 of these airliner-skyscraper-stick-insect hybrids. 

With a planned capacity of 800 megawatts (mw) Vineyard Wind would on its own increase America’s offshore-wind capacity by a factor of roughly 25. 

But it will not be on its own. 

Mr Biden has set a target of 30,000mw (30 gigawatts, gw) of offshore wind by 2030, the equivalent of 37 such projects. 

Britain, China and Germany have ambitions on a similar scale. 

Bernstein, a research firm, reckons that the world’s offshore-wind capacity may reach 254gw by 2030, more than seven times today’s level.

A decade ago this would have seemed pure fantasy. 

Today companies are rushing to meet the demand. 

A battalion of European energy companies led by Equinor, Orsted and Royal Dutch Shell are competing to build in American waters. 

Equinor is backing a new wind-tower factory in Albany. 

Dominion Energy, a utility, is teaming up with a Texan shipbuilder to construct a vessel that can install turbines along America’s east coast. 

In Britain a rush to secure offshore-wind leases in February led to companies bidding so much for the privilege that the returns risk being nugatory.

Similar booms are under way across the world of renewable energy and electric-vehicle making. 

The reason is simple. 

Governments have said they want to cut greenhouse-gas emissions dramatically. 

Decades of subsidy and support, along with some inspired entrepreneurialism, have made available a range of technologies ready to do so. 

The time is ripe to push those technologies as hard as possible—both to battle rising temperatures and, governments hope, advance their countries’ role in a new green economy.

However, the fact that wind farms, solar farms and battery-powered vehicles are now cost-competitive does not mean they can be built at whatever pace politicians choose. 

They require raw materials—sometimes, as with the Haliade-X turbines, in prodigious amounts—siting permits, infrastructure for transmission, recharging and the like. 

They also need lots of capital. 

And the necessary materials, sites and capital are all, to various extents in different places, in short supply. 

The price of lithium has more than doubled in the past year. 

Copper prices are up by about 70%. 

Fights are breaking out over permits for new mines, wind and solar farms. 

Capital remains poorly allocated; while big companies rush for offshore-wind projects around Britain, poorer countries with rising emissions remain starved for investment. 

If efforts to ease those constraints fail, the world’s decarbonisation plans will stall instead of soar.




The Paris agreement of 2015 calls for a world in which average temperatures never climb more than 2°C above those of the preindustrial age, and ideally rise no further than 1.5°C. 

To that end most large economies have now committed themselves to “net zero” emissions—a notional state where the amount of greenhouse gas emitted is matched by the amount absorbed by natural and artificial “sinks”—by the middle of the century. 

In the long term, meeting or even approaching those goals is going to require new technologies, even new industries. 

But any serious attempt also requires the prompt use of the tools already to hand. 

Over the coming decade urgent research and development aimed at creating future tools must take place in tandem with a massive deployment of technologies which already exist.

Greenie in a bottle

In the past, such energy transitions have been slow affairs, and also cumulative ones; new technologies such as those of steam and oil added to the total energy budget rather than simply replacing what was there before them. 

Climate action requires the current transition to be both fast and total. 

A lack of precedent does not make the challenge impossible. 

But it reinforces the need for foresight and imagination in trying to overcome impediments.

The deployment of renewable technologies is already, by the standards of the past, a remarkable success. 

In 2019 installed solar capacity was almost 15 times higher than it was in 2010; for wind power, which got started earlier, the figure was a more modest but still impressive 3.4 times. 

Building capacity has driven down prices, thus making more capacity affordable and driving prices down further. 

Over the past decade the “levelised costs” of solar, offshore wind and onshore wind—figures that take into account initial investment in equipment and construction, financing and maintenance—dropped by 83%, 62% and 58% respectively, according to Bloombergnef, a research group. 

Two-thirds of humankind now lives in countries where wind and solar power offer the cheapest new electrical-generating capacity.

By the standards of the future, though, this is, if not paltry, certainly unsatisfactory. 

Without further intervention, says Seb Henbest, Bloombergnef’s chief economist, “The natural rate of change is far, far too slow to achieve climate targets.” 

In May the International Energy Agency (iea), an intergovernmental group founded in the 1970s to protect access to fossil fuels, published a report on how to abandon them that underscored Mr Henbest’s message.

Looking at pathways by which the world could reach net zero by 2050, the iea confirmed that a lot of what was needed in the near term could be done with existing technologies. 

With a rapid expansion of renewable generation and electric cars through the 2020s electricity and transport could account for more than 70% of the envisaged drop in energy-related emissions. 

But following this path sees the world of 2030 building wind and solar farms at about four times the pace of 2020. 60% of new-car purchases would have to be electric, compared with about 5% today. 

Annual clean-energy investment, already at an all-time high, would have to exceed $4trn by 2030, three times its average over the past five years. 

And the market for key minerals needed to build clean-energy kit would expand nearly seven-fold.



There is an ethereal charm to replacing fuels won from the depths of the Earth with the barely corporeal powers of sun and wind. 

But doing so at scale still requires millions of tonnes of raw materials to be mined. 

Batteries depend on cobalt, lithium and nickel; neodymium and other rare-earth elements (which despite their name are not necessarily rare, though some are) make the magnets for electric generators and motors; the veins and arteries of the green economy run with copper.

The supply chains on which this all depends pose at least two big problems. 

The first is one of concentration. 

The mining and processing of minerals needed for renewables is far more geographically concentrated than the drilling of oil and gas; that should be troubling to anyone with a sense of how the distribution of fossil fuels has influenced history and geopolitics. 

Chinese firms control a large share of many crucial mineral supply chains and of the wherewithal for making batteries, a point anxiously underlined in a review of critical supply chains published by the White House on June 8th.



The second problem concerns underinvestment, particularly in metals. 

Revenues from coal, the dirtiest fossil fuel, continue to exceed those from the minerals that today’s technologies for providing a cleaner future require (see chart 2).

Investment in new projects for lithium, nickel and copper were rising before the pandemic, but at less than $25bn the figure in 2019 was only about 5% of the amount invested on upstream oil and gas. 

And mines require sustained effort; it can take well over a decade to get one up and running.

If the prospect of huge booms in renewables and electric vehicles has not encouraged investment, price signals produced by shortages as those booms get booming may do better. 

But there are issues that go beyond price. 

Some investors find a lot of the mining sector off-putting, either because of genuine ethical concern or because they fear tarnishing their environmental and social credentials. 

They have a point. 

Lithium mining in Chile has triggered legal fights over water in the Atacama. 

More than 70% of cobalt is mined in the Democratic Republic of Congo, with a history of corruption and what the sector euphemistically dubs “artisanal” mining by poor men, women and children.

American, European and Asian politicians are eager to boost mining within their countries’ borders. 

Their citizens may prove less keen. 

Environmental opposition to a rare-earths mine in Greenland helped topple the ruling party in an election there in April. 

In Minnesota conservation groups are worried about a proposed copper-and-nickel mine’s effect on creeks and rivers; in May Mr Biden’s government agreed to reconsider the mine’s permits. 

Its supply-chain review recommends both easing permitting for new mines and limiting their environmental impact; that looks likely to be a hard balancing act.

Rubbed up the wrong way

Benchmark Mineral Intelligence, a research group, recently concluded that in the second half of this decade the world’s lithium demand might be more than twice the level of supply. 

Truly severe shortages could conceivably reverse the long-term trend towards cheaper batteries. Battery costs have declined by 83% since 2012. 

But those savings have come more from design and process improvements and economies of scale than from frugality with inputs. 

Raw materials now represent 50-70% of battery costs, up from 40-50% five years ago, making prices more vulnerable to expensive commodities.

Process changes will still reduce some of the supply gaps. 

Innovations which spare raw materials can spread very quickly—diamond wire saws, which reduce the amount of silicon wasted in the making of solar cells, went from novelty to industry standard almost overnight. 

There will be ever more scope for recycling. 

And there will be substitutions. 

Driven more by concerns over sustainability than price per se, turbine-makers are moving away from the balsa wood often used in their big blades. 

Andreas Nauen, the boss of Siemens Gamesa, a turbine manufacturer, says his company will be using foam instead of balsa by the middle of the decade. 

In February Elon Musk, the boss of Tesla, an electric-car maker, called the availability of nickel the “biggest concern” as the business scales up; he would like to swap nickel-based cathodes for ones made with iron.

Ingenuity can be a powerful force. 

But it cannot be expected always to offset all the effects of the price signals which drive it. 

And it cannot do everything all at once. 

Tesla is still interested enough in nickel to have become an adviser to a nickel mine in New Caledonia.

Another potential shortage is land. Researchers at Princeton University have modelled transition pathways which take America to net zero by 2050. 

They found that the area occupied by solar and wind farms by 2030 might be about 160,000 square kilometres (62,000 square miles). 

That is less than 2% of the surface area of continental America. 

But it is around six times the area currently covered by the water in all the country’s reservoirs—or a little more than the area of Illinois.

Land used for wind farms can be used for other farming, too, and turbines have spread across swathes of America’s Great Plains without too much opposition. 

But for some technologies and places new projects may still depend less on resource abundance than on concern about local impacts and the political heft or legal budgets of those who live nearby. 

American offshore wind is still in its infancy in part because rich people who enjoy their views of the open ocean have fought hard to smother it in its crib—a cause which, in Massachusetts, has brought together Kennedys and Kochs. 

The problem is not restricted to well-off countries. 

In Indonesia, disputes over land rights have seriously slowed the deployment of renewables.

Building infrastructure to deliver green power from panels and pylons in plains and deserts to the places where it is needed faces some of the same challenges. 

Grids that are both bigger and smarter than today’s are needed to make use of intermittent renewable sources at the scales being envisaged later this decade. 

“There will be no renewables without networks,” says Armando Martínez, who leads the grid business of Iberdrola, a big utility. 

The iea estimates that annual spending on electricity grids should more than triple by 2030.

But hurdles to grid investments remain stubbornly high. 

Disagreement over the siting of transmission lines from wind farms in Germany’s north to factories and cities in its south has helped sustain southern coal- and gas-fired power stations. 

In America a transmission line must receive approval from each state it crosses and, in some states, approval from each county. 

The result is that such projects can take more than a decade to build, if they are built at all. 

In Vietnam the growth of solar power in recent years has overwhelmed the country’s ability to transmit it to consumers. 

Forced curtailments of power from solar farms depress their profitability. 

Upgrades to the grid are sorely needed, but to date there has been little way for the private sector to provide it—Vietnam Electricity, or evn, has a monopoly over the country’s transmission and distribution.

Lighting the lantern

Such disincentives point to the biggest supply constraint, especially in developing countries: that of capital. 

Despite rising interest in green investment, serious attempts to meet the Paris goals will require a further surge in finance for green energy and electrification.



The biggest shortfall is in emerging economies other than China, which are expected to account for most of the rise in emissions in the coming decades. 

Those markets saw just $150bn in clean-energy investment in 2020, down 8% from a year earlier, according to new analysis from the iea, World Bank and World Economic Forum. 

In 2019 India attracted just $8bn in clean-energy finance, less than a tenth of China’s total and a sixth of America’s, according to Bloombergnef. 

Other middle-income and poor countries saw even less investment (see chart 3).

Enel, an Italian utility, is the largest foreign investor in green energy in emerging markets. 

To warrant the company’s investment, according to Francesco Starace, its boss, a country must have natural resources, such as ample sun or wind, be in need of infrastructure and, most important, “It has to have a legal and regulatory framework we can trust.” 

A survey by Bloombergnef found that, on average, countries without policies to support clean energy, such as auctions for supply and liberalised electricity sectors, attract one-seventeenth as much clean-energy investment as emerging markets with clearer policies. 

Government support of entrenched domestic coal production and use, as in India and Indonesia, muddies prospects further.




Regulatory and political uncertainties push a country’s levelised costs up. 

Renewable projects have low operating costs (the sun and wind are free) but require a lot of capital upfront. And in many emerging markets capital is expensive. 

The average cost of capital for a wind project in Indonesia is about four times that of one in Germany. 

Investors and politicians in rich countries claim to want to help, but they are not yet doing enough. 

Signatories to the Principles for Responsible Investment, convened by the un, aim to promote sustainable finance, but some 90% of signatories are not active in emerging markets. 

Rich countries have failed to provide the $100bn a year in climate finance that they promised developing countries in Paris.

Politicians and investors are just starting to face these constraints. 

The g7 meeting on June 11th-13th may see well-targeted green aid announced. 

Vietnam is contemplating reforms to encourage private investment in its grid. 

Investors are working to harmonise disclosure of climate risks; governments may do the job for them. 

The most important catalyst to broader green investment, argues Ed Morse of Citigroup, a bank, would be pricing to account for the environmental and social costs of carbon.

Such measures point to a new phase in the green revolution. 

The engineering which allows the spinning blades of a single wind turbine to power a thousand homes, or uses lithium from desiccated lake beds to store power from sunlight in the floor of a sedan, is remarkable. 

But it has to be fed the materials it needs, found places to stand, integrated into the rest of the world’s infrastructure and paid for. 

Innovation and investment in mining, pressure on the politics of land use and new catalysts for private investment, especially in emerging markets, are less iconic. 

But they are no less necessary.

The Case Against Zoos

By Emma Marris

                Photographs by Peter Fisher for The New York Times


After being captives of the pandemic for more than a year, we have begun experiencing the pleasures of simple outings: dining al fresco, shopping with a friend, taking a stroll through the zoo. 

As we snap a selfie by the sea lions for the first time in a year, it seems worth asking, after our collective ordeal, whether our pleasure in seeing wild animals up close is worth the price of their captivity.

Throughout history, men have accumulated large and fierce animals to advertise their might and prestige. 

Power-mad men from Henry III to Saddam Hussein’s son Uday to the drug kingpin Pablo Escobar to Emperor Charlemagne all tried to underscore their strength by keeping terrifying beasts captive. 

William Randolph Hearst created his own private zoo with lions, tigers, leopards and more at Hearst Castle. 

It is these boastful collections of animals, these autocratic menageries, from which the modern zoo, with its didactic plaques and $15 hot dogs, springs.

The forerunners of the modern zoo, open to the public and grounded in science, took shape in the 19th century. 

Public zoos sprang up across Europe, many modeled on the London Zoo in Regent’s Park. 

Ostensibly places for genteel amusement and edification, zoos expanded beyond big and fearsome animals to include reptile houses, aviaries and insectariums. 

Living collections were often presented in taxonomic order, with various species of the same family grouped together, for comparative study.


The first zoos housed animals behind metal bars in spartan cages. 

But relatively early in their evolution, a German exotic animal importer named Carl Hagenbeck changed the way wild animals were exhibited. 

In his Animal Park, which opened in 1907 in Hamburg, he designed cages that didn’t look like cages, using moats and artfully arranged rock walls to invisibly pen animals. 

By designing these enclosures so that many animals could be seen at once, without any bars or walls in the visitors’ lines of sight, he created an immersive panorama, in which the fact of captivity was supplanted by the illusion of being in nature.

Mr. Hagenbeck’s model was widely influential. 

Increasingly, animals were presented with the distasteful fact of their imprisonment visually elided. 

Zoos shifted just slightly from overt demonstrations of mastery over beasts to a narrative of benevolent protection of individual animals. 

From there, it was an easy leap to protecting animal species.

The “educational day out” model of zoos endured until the late 20th century, when zoos began actively rebranding themselves as serious contributors to conservation. 

Zoo animals, this new narrative went, function as backup populations for wild animals under threat, as well as “ambassadors” for their species, teaching humans and motivating them to care about wildlife. 

This conservation focus “must be a key component” for institutions that want to be accredited by the Association of Zoos and Aquariums, a nonprofit organization that sets standards and policies for facilities in the United States and 12 other countries.

This is the image of the zoo I grew up with: the unambiguously good civic institution that lovingly cared for animals both on its grounds and, somehow, vaguely, in their wild habitats. 

A few zoos are famous for their conservation work. 

Four of the zoos and the aquarium in New York City, for instance, are managed by the Wildlife Conservation Society, which is involved in conservation efforts around the world. 

But this is not the norm.



While researching my book on the ethics of human interactions with wild species, “Wild Souls,” I examined how, exactly, zoos contribute to conservation of wild animals.

A.Z.A. facilities report spending approximately $231 million annually on conservation projects. 

For comparison, in 2018, they spent $4.9 billion on operations and construction. 

I find one statistic particularly telling about their priorities: A 2018 analysis of the scientific papers produced by association members between 1993 and 2013 showed that just about 7 percent of them annually were classified as being about “biodiversity conservation.”

Zoos accredited by the A.Z.A. or the European Association of Zoos and Aquaria have studbooks and genetic pedigrees and carefully breed their animals as if they might be called upon at any moment to release them, like Noah throwing open the doors to the ark, into a waiting wild habitat. 

But that day of release never quite seems to come.

There are a few exceptions. 

The Arabian oryx, an antelope native to the Arabian Peninsula, went extinct in the wild in the 1970s and then was reintroduced into the wild from zoo populations. 

The California condor breeding program, which almost certainly saved the species from extinction, includes five zoos as active partners. 

Black-footed ferrets and red wolves in the United States and golden lion tamarins in Brazil — all endangered, as well — have been bred at zoos for reintroduction into the wild. 

An estimated 20 red wolves are all that remain in the wild.

The A.Z.A. says that its members host “more than 50 reintroduction programs for species listed as threatened or endangered under the Endangered Species Act.”

Nevertheless, a vast majority of zoo animals (there are 800,000 animals of 6,000 species in the A.Z.A.’s zoos alone) will spend their whole lives in captivity, either dying of old age after a lifetime of display or by being culled as “surplus.”



The practice of killing “surplus” animals is kept quiet by zoos, but it happens, especially in Europe. 

In 2014, the director of the E.A.Z.A. at the time estimated that between 3,000 and 5,000 animals are euthanized in European zoos each year. 

Early in the pandemic, the Neumünster Zoo in northern Germany coolly announced an emergency plan to cope with lost revenue by feeding some animals to other animals, compressing the food chain at the zoo like an accordion, until in the worst-case scenario, only Vitus, a polar bear, would be left standing. 

The A.Z.A.’s policies allow for the euthanasia of animals, but the president of the association, Dan Ashe, told me, “it’s very rarely employed” by his member institutions.

Mr. Ashe, a former director of the U.S. Fish and Wildlife Service, suggested that learning how to breed animals contributes to conservation in the long term, even if very few animals are being released now. 

A day may come, he said, when we need to breed elephants or tigers or polar bears in captivity to save them from extinction. 

“If you don’t have people that know how to care for them, know how to breed them successfully, know how to keep them in environments where their social and psychological needs can be met, then you won’t be able to do that,” he said.

The other argument zoos commonly make is that they educate the public about animals and develop in people a conservation ethic. 

Having seen a majestic leopard in the zoo, the visitor becomes more willing to pay for its conservation or vote for policies that will preserve it in the wild. 

What Mr. Ashe wants visitors to experience when they look at the animals is a “sense of empathy for the individual animal, as well as the wild populations of that animal.”

I do not doubt that some people had their passion for a particular species, or wildlife in general, sparked by zoo experiences. 

I’ve heard and read some of their stories. I once overheard two schoolchildren at the Smithsonian’s National Zoo in Washington confess to each other that they had assumed that elephants were mythical animals like unicorns before seeing them in the flesh. 

I remember well the awe and joy on their faces, 15 years later. 

I’d like to think these kids, now in their early 20s, are working for a conservation organization somewhere. 

But there’s no unambiguous evidence that zoos are making visitors care more about conservation or take any action to support it. 

After all, more than 700 million people visit zoos and aquariums worldwide every year and biodiversity is still in decline.


In a 2011 study, researchers quizzed visitors at the Cleveland, Bronx, Prospect Park and Central Park zoos about their level of environmental concern and what they thought about the animals. 

Those who reported “a sense of connection to the animals at the zoo” also correlated positively with general environmental concern. 

On the other hand, the researchers reported, “there were no significant differences in survey responses before entering an exhibit compared with those obtained as visitors were exiting.”

A 2008 study of 206 zoo visitors by some members of the same team showed that while 42 percent said that the “main purpose” of the zoo was “to teach visitors about animals and conservation,” 66 percent said that their primary reason for going was “to have an outing with friends or family,” and just 12 percent said their intention was “to learn about animals.”

The researchers also spied on hundreds of visitors’ conversations at the Bronx Zoo, the Brookfield Zoo outside Chicago and the Cleveland Metroparks Zoo. 

They found that only 27 percent of people bothered to read the signs at exhibits. 

More than 6,000 comments made by the visitors were recorded, nearly half of which were “purely descriptive statements that asserted a fact about the exhibit or the animal.” 

The researchers wrote, “In all the statements collected, no one volunteered information that would lead us to believe that they had an intention to advocate for protection of the animal or an intention to change their own behavior.”

People don’t go to zoos to learn about the biodiversity crisis or how they can help. 

They go to get out of the house, to get their children some fresh air, to see interesting animals. 

They go for the same reason people went to zoos in the 19th century: to be entertained.

*

A fine day out with the family might itself be justification enough for the existence of zoos if the zoo animals are all happy to be there. 

Alas, there’s plenty of heartbreaking evidence that many are not.

In many modern zoos, animals are well cared for, healthy and probably, for many species, content. Zookeepers are not mustache-twirling villains. 

They are kind people, bonded to their charges and immersed in the culture of the zoo, in which they are the good guys.

But many animals clearly show us that they do not enjoy captivity. 

When confined they rock, pull their hair and engage in other tics. 

Captive tigers pace back and forth, and in a 2014 study, researchers found that “the time devoted to pacing by a species in captivity is best predicted by the daily distances traveled in nature by the wild specimens.” 

It is almost as if they feel driven to patrol their territory, to hunt, to move, to walk a certain number of steps, as if they have a Fitbit in their brains.

The researchers divided the odd behaviors of captive animals into two categories: “impulsive/compulsive behaviors,” including coprophagy (eating feces), regurgitation, self-biting and mutilation, exaggerated aggressiveness and infanticide, and “stereotypies,” which are endlessly repeated movements. 

Elephants bob their heads over and over. 

Chimps pull out their own hair. 

Giraffes endlessly flick their tongues. 

Bears and cats pace. 

Some studies have shown that as many as 80 percent of zoo carnivores, 64 percent of zoo chimps and 85 percent of zoo elephants have displayed compulsive behaviors or stereotypes.


Elephants are particularly unhappy in zoos, given their great size, social nature and cognitive complexity. 

Many suffer from arthritis and other joint problems from standing on hard surfaces; elephants kept alone become desperately lonely; and all zoo elephants suffer mentally from being cooped up in tiny yards while their free-ranging cousins walk up to 50 miles a day. 

Zoo elephants tend to die young. 

At least 20 zoos in the United States have already ended their elephant exhibits in part because of ethical concerns about keeping the species captive.

Many zoos use Prozac and other psychoactive drugs on at least some of their animals to deal with the mental effects of captivity. 

The Los Angeles Zoo has used Celexa, an antidepressant, to control aggression in one of its chimps. 

Gus, a polar bear at the Central Park Zoo, was given Prozac as part of an attempt to stop him from swimming endless figure-eight laps in his tiny pool. 

The Toledo Zoo has dosed zebras and wildebeest with the antipsychotic haloperidol to keep them calm and has put an orangutan on Prozac. 

When a female gorilla named Johari kept fighting off the male she was placed with, the zoo dosed her with Prozac until she allowed him to mate with her. 

A 2000 survey of U.S. and Canadian zoos found that nearly half of respondents were giving their gorillas Haldol, Valium or another psychopharmaceutical drug.

Some zoo animals try to escape. Jason Hribal’s 2010 book, “Fear of the Animal Planet,” chronicles dozens of attempts. 

Elephants figure prominently in his book, in part because they are so big that when they escape it generally makes the news.

Mr. Hribal documented many stories of elephants making a run for it — in one case repairing to a nearby woods with a pond for a mud bath. 

He also found many examples of zoo elephants hurting or killing their keepers and evidence that zoos routinely downplayed or even lied about those incidents.

Elephants aren’t the only species that try to flee a zoo life. 

Tatiana the tiger, kept in the San Francisco Zoo, snapped one day in 2007 after three teenage boys had been taunting her. 

She somehow got over the 12-foot wall surrounding her 1,000-square-foot enclosure and attacked one of the teenagers, killing him. 

The others ran, and she pursued them, ignoring all other humans in her path. 

When she caught up with the boys at the cafe, she mauled them before she was shot to death by the police. 

Investigators found sticks and pine cones inside the exhibit, most likely thrown by the boys.

Apes are excellent at escaping. 

Little Joe, a gorilla, escaped from the Franklin Park Zoo in Boston twice in 2003. 

At the Los Angeles Zoo, a gorilla named Evelyn escaped seven times in 20 years. 

Apes are known for picking locks and keeping a beady eye on their captors, waiting for the day someone forgets to lock the door. 

An orangutan at the Omaha Zoo kept wire for lock-picking hidden in his mouth. 

A gorilla named Togo at the Toledo Zoo used his incredible strength to bend the bars of his cage. 

When the zoo replaced the bars with thick glass, he started methodically removing the putty holding it in. 

In the 1980s, a group of orangutans escaped several times at the San Diego Zoo. 

In one escape, they worked together: One held a mop handle steady while her sister climbed it to freedom. 

Another time, one of the orangutans, Kumang, learned how to use sticks to ground the current in the electrical wire around her enclosure. 

She could then climb the wire without being shocked. 

It is impossible to read these stories without concluding that these animals wanted out.


“I don’t see any problem with holding animals for display,” Mr. Ashe told me. 

“People assume that because an animal can move great distances that they would choose to do that.” 

If they have everything they need nearby, he argued, they would be happy with smaller territories. 

And it is true that the territory size of an animal like a wolf depends greatly on the density of resources and other wolves. 

But then there’s the pacing, the rocking. 

I pointed out that we can’t ask animals whether they are happy with their enclosure size. 

“That’s true,” he said. 

“There is always that element of choice that gets removed from them in a captive environment. 

That’s undeniable.” 

His justification was philosophical. In the end, he said, “we live with our own constraints.” 

He added, “We are all captive in some regards to social and ethical and religious and other constraints on our life and our activities.”

What if zoos stopped breeding all their animals, with the possible exception of any endangered species with a real chance of being released back into the wild? 

What if they sent all the animals that need really large areas or lots of freedom and socialization to refuges? 

With their apes, elephants, big cats, and other large and smart species gone, they could expand enclosures for the rest of the animals, concentrating on keeping them lavishly happy until their natural deaths. 

Eventually, the only animals on display would be a few ancient holdovers from the old menageries, animals in active conservation breeding programs and perhaps a few rescues.


Such zoos might even be merged with sanctuaries, places that take wild animals that because of injury or a lifetime of captivity cannot live in the wild. 

Existing refuges often do allow visitors, but their facilities are really arranged for the animals, not for the people. 

These refuge-zoos could become places where animals live. 

Display would be incidental.

Such a transformation might free up some space. 

What could these zoos do with it, besides enlarging enclosures? 

As an avid fan of botanical gardens, I humbly suggest that as the captive animals retire and die off without being replaced, these biodiversity-worshiping institutions devote more and more space to the wonderful world of plants. 

Properly curated and interpreted, a well-run garden can be a site for a rewarding “outing with friends or family,” a source of education for the 27 percent of people who read signs and a point of civic pride.

I’ve spent many memorable days in botanical gardens, completely swept away by the beauty of the design as well as the unending wonder of evolution — and there’s no uneasiness or guilt. 

When there’s a surplus, you can just have a plant sale.


Emma Marris is an environmental writer and the author of the forthcoming book “Wild Souls: Freedom and Flourishing in the Non-Human World.”

Photographs by Peter Fisher. Mr. Fisher is a photographer based in New York.