Research Highlights

Our pick of the latest scientific literature

Breaching Cuvier's beaked whale (Ziphius cavirostris) Atlantic, USA.

The Cuvier's beaked whale, whose deep dives can last an hour or more, typically spends only minutes at the sea surface before plunging back underwater. Credit: Todd Pusser/NPL


A smiling whale makes a record deep dive

The elusive Cuvier’s beaked whale, already known for its prowess as a diver, turns out to have even more staying power than scientists thought.

The champion diver among whales can stay underwater for more than 3½ hours — an unparalleled diving record for any mammal.

Cuvier's beaked whales (Ziphius cavirostris), which live in deep waters around the world, are capable of reaching depths of almost 3,000 metres when hunting for squid and other food. Scientists had previously estimated that these whales could hold their breath for around 30 minutes before exhausting their oxygen store. But observations among Cuvier’s whales foraging off the coast of North Carolina suggest they can dive for far longer.

Nicola Quick at Duke University in Beaufort, North Carolina, and her team fastened satellite-linked tags on 23 whales and obtained records from almost 3,700 deep dives over a 5-year period. The data reveal that the median duration of foraging dives was around a full hour, before the animal returned to the surface to recover.

One individual made two extreme dives of 2 hours 57 minutes, and 3 hours 42 minutes, respectively, which might represent the true limit of the species’ extraordinary diving capacity, the researchers say.

The Curly is throwing to deliver the stone to the target location chosen by the curling AI.

Guided by an adaptive algorithm, Curly the robot makes highly accurate ‘throws’ of 20-kilogramme curling stones at a target. Credit: Korea University


A robot triumphs in a curling match against elite humans

Artificial intelligence helps to propel a machine to victory in an Olympic sport described as ‘chess on ice’.

A robot equipped with artificial intelligence (AI) can excel at the Olympic sport of curling — and even beat top-level human teams.

Players in a curling match slide stones across an ice rink towards targets. Success requires precision and strategy, but the game is less complex than other real-world applications of robotics. That makes curling a useful test case for AI technologies, which often perform well in simulations but falter in real-world scenarios with changing conditions.

Using a method called adaptive deep reinforcement learning, Seong-Whan Lee and his colleagues at Korea University in Seoul created an algorithm that learns through trial and error to adjust a robot’s throws to account for changing conditions, such as the ice surface and the positions of stones.

The team’s robot, nicknamed Curly, needed a few test throws to calibrate itself to the curling rink where it was to compete. But once acclimated, the robot won three out of four matches against elite human competitors. The algorithm reduced Curly’s throwing error — the mean distance from the target’s centre — to about one-third of that of a non-adapting robot.

A hand holding a mobile phone showing the profile picture for Twitter's own Twitter account.

Mining the biographies of a Twitter user’s followers can reveal the reach and audience of a paper the user shares. Credit: Getty


Twitter followers’ details reveal the power of a paper

Tweets about biology preprints generally come from members of the academic community — but politics also plays a part.

“UK biologist and writer.” That is the only information that outspoken evolutionary biologist and atheist Richard Dawkins uses to describe himself in his Twitter biography.

Laconic or vague profiles can pose a challenge for researchers interested in how information spreads on social networks and who is responsible for spreading it. But Jedidiah Carlson and Kelley Harris at the University of Washington in Seattle have found that studying the biographies of a Twitter user’s followers can paint a clearer picture of those users and the communities they reach.

The pair analysed tweets about 1,800 papers posted on the biology preprint server bioRxiv. Users associated with academia generated the majority of tweets for 96% of the papers. But political inclinations affected who tweeted about some papers: users with left-leaning lay audiences were the most likely to share ecology preprints, and those with more right-leaning lay followers favoured genetics. The authors noted that 10% of the papers they studied had a prominent right-wing white-nationalist audience.

The authors hope the work will motivate researchers to think about how their findings might be shared and misappropriated.

A stretch of tundra covered in grasses, with mountains in the distance.

Autumn on the Arctic tundra, which is becoming a deeper shade of green in some places as temperatures climb. Credit: Getty

Climate change

A frozen land goes green as Earth warms

Climate change drives vegetation gains in patches of the high Arctic tundra.

Parts of the treeless Arctic tundra have become greener as rising temperatures stimulate plant growth.

Low-resolution satellite imagery and some observations made on the ground have suggested that the Arctic tundra, an often-frozen landscape dotted with hardy small plants, has become greener since the 1980s. Now, Logan Berner at Northern Arizona University in Flagstaff and his colleagues have analysed high-resolution images from the Landsat Earth-observing satellites.

The team found that between 1985 and 2016, 37% of the Arctic tundra, including parts of western Eurasia and North America, grew substantially greener. And since the turn of the century, the highest latitudes have experienced the most intense greening.

Temperature records show that the Arctic air and soil grew warmer in summer, on average, over the study period. However, most areas did not become greener. And about 5% of the area studied became browner rather than greener.

Changes in Arctic vegetation affect how carbon cycles through the soil and atmosphere, how wildlife and people make use of the landscape, and how vulnerable the tundra is to wildfires.

Close-up of a green plant stem covered in fine, translucent stings, on a black background.

Beware the fuzz: the hair-like structures on the leaves and stems of the giant stinging tree (Dendrocnide excelsa) can inject a potent venom into the skin of the unwary. Credit: Institute for Molecular Bioscience/University of Queensland


How the giant stinging tree of Australia can inflict months of agony

A new type of peptide produces pain so intense that sometimes even morphine cannot quell it.

The unrelenting pain from an Australian tree’s sting is caused by newly identified peptides resembling those in spider venoms.

The leaves and stems of the giant stinging tree (Dendrocnide excelsa) carry what looks like an inviting fuzz, but is actually numerous tiny needles. When a passer-by brushes against the tree, the needles can inject their skin with a venom causing intense pain that sometimes lasts months and resists even morphine. Various molecules have been proposed as the venom’s active component, but none adequately accounted for the severity and duration of the pain.

Irina Vetter and Thomas Durek at the University of Queensland in Brisbane and their colleagues ventured out to harvest leaves of wild D. excelsa and the closely related tree Dendrocnide moroides. Chemical analysis of the venom-injecting needles revealed a previously unknown family of peptides. The researchers named them gympietides after gympie-gympie, the trees’ name in the Indigenous Gubbi Gubbi language.

Further investigation showed that gympietides not only generate pain signals, but also suppress a mechanism that stops those signals. The authors, all of them Dendrocnide victims, hope their findings spur research to develop a gympietide antidote.

An aerial view of the East Bay Municipal Utility District Wastewater Treatment Plant on April 29, 2020 in Oakland, California.

Some microfibres captured by waste-water treatment plants (pictured) end up in sewage sludge, which can be spread on land. Credit: Justin Sullivan/Getty

Environmental sciences

Fleece garments are cosy — but pollute land as well as water

Many plastic microfibres filtered out of waste water end up in landfills and at other terrestrial sites.

Synthetic clothing has released about 5.6 million tonnes of microfibres since 1950, polluting land and water alike.

Synthetic fabrics, such as polar fleece and nylon, shed microscopic plastic fibres when washed. But although previous research has identified microfibres in many diverse ecosystems, this is the first comprehensive estimate of the fibres’ final resting places.

Pulling information from global data sets on clothing lifetime, washing machine usage and waste-water treatment, Jenna Gavigan and her colleagues at the University of California, Santa Barbara, estimated the total mass of microfibres introduced to the environment between 1950 and 2016 to be 4.3 million to 7 million tonnes. They found that nearly half of that mass ended up in lakes, rivers and oceans, mostly by way of untreated waste water.

But almost as much made its way into landfills and land ecosystems, where the tiny fibres might affect microbial activity. And as waste-water treatments become more widespread in low-income countries, a growing proportion of the fibres are ending up in these terrestrial environments, the authors say.

A gate made of stone, with two giant lion statues.

The Lion’s Gate was the southwestern entrance to Hattusha, which is located in modern-day Turkey and was once the capital of the Hittite empire. Credit: Francesco Bandarin/UNESCO (CC BY-SA 3.0 IGO)


Ancient tax collectors amassed a fortune — until it went up in smoke

Grain stored in a burnt silo came from multiple farms in an empire in the sixteenth century BC.

Heaps of Bronze-Age grain retrieved from an enormous subterranean silo show the long reach of the tax collector, even 3,000 years ago.

Archaeologists discovered the silo in 1999 at Hattusha, in what is now Turkey. Founded in roughly 1650 BC, Hattusha was the capital of the Hittite empire, which grew into a superpower that rivalled the kingdom of Egypt. The silo covered an area that was roughly the size of a football pitch and, when discovered, held hundreds of tonnes of intact grain (below) in layers more than one metre thick.

Carbonised crop and weed material from the silo complex.

Carbonized remnants found in a burnt silo from the Bronze Age include, clockwise from upper left, wheat kernels mixed with weed seeds; barley; and two types of weed seed.Credit: C. Diffey et al./Antiquity

Amy Bogaard at the University of Oxford, UK, and her colleagues examined the wheat and barley in 5 of the silo’s 32 chambers. The intermixed weed seeds and the chemical profiles of the grains suggest that each chamber held cereals from a separate farming community, or perhaps multiple communities.

The authors say the silo contained grain collected as tax from people living across Hittite lands, and was a symbol of the Hittite king’s wealth — until fire devastated the structure shortly after its construction, leading the regime to abandon the wreckage.

Portrait of a little girl sitting with her head on her arms, staring at an old clock.

If the day seems to be moving slowly, fatigue in a group of neurons involved in time perception might be to blame. Credit: Getty


Exhausted neurons help make time seem ... to ... drag

A brain region that tires after repeated use is involved in distorted perceptions of time’s passage.

Time flies when you are busy, but when you’re bored, it seems as if the day will never end. Researchers have now found that specific neurons can grow weary if repeatedly exposed to the same stimulus — altering the brain’s perception of time.

Masamichi Hayashi at the National Institute of Information and Communications Technology in Suita, Japan, and Richard Ivry at the University of California, Berkeley, scanned volunteers’ brains while showing them a grey spot on a screen for a defined period of time, 30 times in a row. After this ‘adaptation’ phase, participants saw the grey spot again, but for different lengths of time. Then, they estimated how long the object had stayed on screen.

In some tests, the length of time that the visual stimulus was on screen was similar to that in the adaptation phase. Participants tended to respond by misjudging its duration, and activity decreased in a group of brain cells involved in time perception, indicating neuron fatigue. The activity of other neurons was unchanged, which can skew an individuals’ experience of time, the researchers say.

Rocky surface in foreground with grey-blue lake behind it and snow-covered mountains in the distance.

When swollen with glacial meltwater, Nepal’s Imja Lake posed such a hazard that the authorities mounted a massive campaign to lower its level. Credit: Getty

Climate change

Shrinking glaciers lead to growing lakes — and growing risks

Meltwater from thawing glaciers is expanding lakes, which could catastrophically burst their banks.

As glaciers melt in response to rising global temperatures, their waters are engorging nearby lakes. Now scientists estimate that the volume of Earth’s glacial lakes has grown by nearly 50% in the past three decades.

Dan Shugar at the University of Calgary in Canada and his colleagues analysed more than 250,000 images taken by the Landsat satellites. The scientists used a model that maps the outlines of all of the world’s lakes that are next to glaciers, but excluded those in Antarctica. The researchers then calculated how much water was stored in these lakes.

Between 1990 and 2018, the number of glacial lakes rose from 9,414 to 14,394. The lakes’ total surface area increased by 51%, and their volume by 48%. The extra water would fill more than 20 million Olympic swimming pools. Lake volumes increased most at high latitudes.

Bigger lakes pose greater hazards to mountain communities, because the water is often loosely dammed by glacier debris and can suddenly burst through, causing catastrophic flooding. If all of the water contained in these glacial lakes were to reach the ocean, it would increase global sea level by 0.43 millimetres.

Black Metaltail Hummingbird (Metallura phoebe) feeding on flower nectar, Colca Canyon, Southern Andes, Peru.

The black metaltail hummingbird, which thrives in the high Andes, has set a new low-temperature record for birds. Credit: Tui De Roy/NPL


Why some of the world’s zippiest birds go stiff and cold every night

Torpor saves energy for hummingbirds that live high in the Andes Mountains.

Some hummingbirds drink more than their own weight in nectar every day to power their energetically costly hovering flight and keep their tiny bodies warm. Astonishingly, however, hummingbirds can be found at up to 5,000 metres above sea level in the Andes Mountains of South America. How do they make it through the frosty nights?

The answer, researchers say, is that the minuscule birds go into torpor — a state of reduced metabolic activity and temperature that is not unlike hibernation, but only one night long. In an effort to find out more, Blair Wolf at the University of New Mexico in Albuquerque and his colleagues captured 26 hummingbirds, encompassing 6 species, from the Andean forest, and observed them overnight. Almost all of the birds turned into still, silent, ice cubes; the body of one black metaltail (Metallura phoebe) reached a low of 3.26°C — a new record for birds.

Torpor lasted from less than 3 hours to more than 12 hours, and longer bouts of torpor correlated with less loss of body mass overnight. The researchers speculate that in cold weather, torpor might extend over days in some species, becoming true hibernation.

Yellow, white and blue galaxies against a black background.

Among the galactic clusters examined by researchers seeking to understand the distribution of dark matter is MACSJ1206 (spread across centre of image), which distorts other galaxies’ light into smears and arcs. Credit: ESA/Hubble & NASA (CC BY 4.0)


Gigantic clusters of galaxies pose a new dark-matter puzzle

The Hubble Space Telescope shows that galactic ‘lenses’ of the invisible material are more common than computer models of galaxy clumping predict.

Examination of huge clusters of galaxies has revealed that dark matter — the invisible stuff that constitutes most of the Universe’s mass — forms more galactic-scale lumps than expected.

Massimo Meneghetti at the National Institute for Astrophysics in Bologna, Italy, and his colleagues surveyed 11 clusters that each contain around 1,000 galaxies and are massive enough to curve the light of more distant galaxies as it travels past them, making the distant galaxies look like long arcs when viewed from Earth. These ‘strong gravitational lenses’ had been imaged by the Hubble Space Telescope.

As expected, most of the lensing power was centred around the clusters’ middle regions, where dark matter forms a dense ‘halo’. But Hubble’s images also enabled a finer mapping, revealing dozens of smaller lenses — like “additional bubbles of little lenses”, says co-author Priyamvada Natarajan at Yale University in New Haven, Connecticut. These ‘subhaloes’ were more numerous, and their lensing was on average ten times stronger, than predicted by computer simulations of clusters formed according to existing theories of dark matter, the authors write.

A blue machine travels between rows of plants, spraying them from four hoses.

A worker sprays grapevines in France. Sensors affixed to grape leaves can detect the effects of the air pollutant ozone. Credit: Getty

Plant sciences

Etched for success: the ‘tattoos’ that could help crops to survive polluted air

Polymer electrodes deposited onto the leaves of crop plants provide an early warning of ozone damage.

Fingernail-sized electrodes ‘tattooed’ onto a leaf can reveal the invisible damage done by the ubiquitous air pollutant ozone, which threatens farmers’ harvests.

Exposure to just 20 parts per million of ozone in air for one hour can permanently cut yields of fruit crops such as grapes by 10%. Diagnosing ozone damage early is both key for protecting crops and difficult, because farmers must act before damage becomes visible on leaves.

To develop an early indicator of ozone damage, Trisha Andrew and her colleagues at the University of Massachusetts Amherst capitalized on one of the pollutant’s effects: when ozone kills a leaf’s cells, the leaf tissue becomes less electrically conductive. The researchers printed electrodes one micrometre thick onto leaves cut from apple trees and vines of Merlot, Chardonnay and Concord grapes. Then, they exposed the leaves to a variety of ozone levels. After measuring the leaves’ conductivity through the electrodes, the researchers observed that it was indeed lower at higher ozone levels.

The researchers envision tattooing leaves on living plants to monitor ozone exposure in vineyards and orchards.

Illustration of gold film punctuated with holes, illuminated by a laser beam whose movement causes a trapped particle to shift.

Moving a laser beam (red; artist’s impression) from one site on a gold film to another causes a trapped particle (blue) to move in concert. Credit: C. Hong et al./Nature Nanotechnol.

Nanoscience and technology

These ‘tweezers’ made of light gently grasp and move a single protein

A fresh approach allows a laser beam to manipulate molecules without exposing them to destructive levels of heat and light.

A laser beam focused by a microscope can trap and manipulate a range of minute objects, including viruses and cells. But such ‘optical tweezers’ do not work well for particles smaller than about 10 nanometres, because the laser light needed to hold such minuscule objects can easily damage them.

To circumvent this problem, Justus Ndukaife and his colleagues at Vanderbilt University in Nashville, Tennessee, fashioned a gold film patterned with tiny holes and placed it inside a sample chamber that they then filled with fluid. Next, the team shone a laser on the film and applied an alternating electric field to it, creating two opposing flows in the fluid.

The researchers captured a protein molecule roughly 7 nanometres wide in an island of motionless fluid between these flows. The team could move the trapped molecule by shifting the position of the laser.

Crucially, objects trapped in this way are held a few micrometres away from the focus of the laser beam, limiting their exposure to heat and light. The approach should allow researchers to grasp and study individual biological objects, and even sort them by size.

Regenerating tail of a male killifish 6 days after amputation.

A killifish regrew its tail only six days after the appendage was clipped at the brightly dotted horizontal line. Credit: Dr Wei Wang


Why some animals have the power of regeneration

Newly identified genetic elements help to replace missing tails and other body parts.

Scientists have pinpointed gene-activating stretches of DNA that help animals to regenerate limbs, fins and other tissues.

Some vertebrates can regenerate tissues after injury, but many cannot — suggesting that this trait has been lost and gained across the vertebrate family tree. Alejandro Sánchez Alvarado at the Stowers Institute for Medical Research in Kansas City, Missouri, and his colleagues studied zebrafish (Danio rerio) and African killifish (Nothobranchius furzeri), which can regenerate their tail fins, and identified genes that were active in the regenerating tissue.

There was little overlap in the active genes between the species. Those that did overlap, including one called inhibin beta A, were also active during regeneration of ear tissue in Cairo spiny mice (Acomys cahirinus), but not after ear injury in non-regenerative house mice (Mus musculus).

A type of non-coding DNA sequence called an enhancer influenced the activity of inhibin beta A. The team found that deleting the enhancer or replacing it with a human version hindered fin and heart regeneration in killifish. The researchers propose that changes to such enhancer sequences are responsible for differences in animals’ ability to rebuild their tissues.

Scenic Yumthang valley with view of river Teesta and majestic Himalayan range.

A 2015 rock avalanche in the Yumthang Valley, located in the eastern Himalayan range, triggered a blast of air that knocked down more than 1,000 trees. Credit: Shutterstock


The violent blasts that can add to an avalanche’s devastation

Scientists zero in on the factors that heighten the chance of ‘airblasts’ after a slope collapses.

Rock falls and rock avalanches, which are dangerous enough in their own right, sometimes unleash an additional peril: powerful blasts of air that can flatten trees more than a kilometre away. Now scientists have documented the conditions that make these ‘airblasts’ more likely.

Although some airblasts have proved fatal, little research has been done to document their destructive potential, and landslide risk assessments do not account for them. To fill this gap, Ivanna Penna at the Geological Survey of Norway in Trondheim and her colleagues analysed airblasts that have occurred around the world, including a previously unreported 2015 event in the Yumthang Valley in the Indian Himalayas.

Using data from both ground and aerial drone surveys, they mapped the destruction of the airblast that followed the Yumthang rock fall. This allowed them to estimate the event’s maximum wind speed, which was 385 kilometres per hour.

The authors determined that airblasts are most likely to follow rock avalanches on steep mountainsides. They also found that the most destructive airblasts occur in narrow valleys, which confine the airflow more strongly than wide valleys do.

Inhumation Sk 6 from Windmill Fields, Ingleby Barwick, North Yorkshire, disarticulated remains.

A grave in northern England held the remains of a woman who lived thousands of years ago — and the skulls of two people who had died some years earlier. Credit: T.J. Booth & J. Brück/Antiquity Pub.


An ancient whistle was crafted from a human thigh bone

Prehistoric people kept the bones of relatives and friends for generations as relicts.

Ancient Britons modified the bones of people in their communities and kept these relicts close at hand, often saving pieces of skeleton for decades after a person’s death, according to an analysis of artefacts from dozens of prehistoric sites.

Thomas Booth and Joanna Brück, then at the University of Bristol, UK, generated new radiocarbon dates for 54 human bones and associated animal bones, charcoal and one hazelnut shell collected at British archaeological sites dating from 2500 to 600 BC. All of the items are thought to have been deliberately placed on prepared surfaces.

A human femur turned into a musical instrument, found in Wiltshire

A human femur found near Stonehenge was shaped into a whistle and then buried with a man who probably knew the source of the bone.Credit: Wiltshire Museum

At 26 of 60 study sites included in the analysis, the team discovered that the human bones were older than surrounding materials, suggesting that the bones had been buried long after death. A human femur that had been converted into a whistle, for example, was found in another person’s grave near Stonehenge, and the Yorkshire grave of an adult woman contained two substantially older skulls.

This treatment suggests that human remains were often revered, rather than viewed with horror or disgust, the authors say.

Coloured scanning electron micrograph (SEM) of Geobacter sulfurreducens bacteria amid metallic waste.

Geobacter sulfurreducens bacteria (brown in this false-colour image, pictured digesting grey metallic waste) can fashion coppery networks that conduct electricity. Credit: EYE OF SCIENCE/SPL


Microbes with mettle build their own electrical ‘wires’

Hardy bacteria thrive on an antimicrobial metal — and turn it to their advantage.

Some bacteria that can transmit electricity turn out to have a hidden talent: when grown on a copper electrode, they construct a network of copper sulfide compounds that boosts their conductivity.

Copper is hostile to microorganisms, and as a result has long been used in antibacterial cladding for ships’ hulls, and in pots and pipes for drinking water. But while studying electrically conductive Geobacter sulfurreducens bacteria for use in fuel cells, Uwe Schröder and his colleagues at the Technical University of Braunschweig in Germany found that the microbes flourished on copper, forming tough layers known as biofilms.

What’s more, G. sulfurreducens biofilms on a copper electrode produced double the electrical current of those on a graphite electrode. Chemical analysis revealed copper sulfide solids deposited throughout the biofilms. The team concluded that the bacteria promote chemical reactions between the copper electrode and sulfate ions in their food source to form copper sulfide ‘wires’, which enhanced the flow of electricity within the biofilm.

The researchers hope their findings can help to improve the design and performance of fuel cells that take advantage of such electrically conductive bacteria.

Overhead shot of lots of full plates on a table, with hands reaching to take food or pour wine.

Scientists have identified a brain response that encourages eating even after appetite has been satisfied. Credit: Getty


The brain circuit that encourages eating for pleasure

Neurons that tamp down eating and drinking become sluggish after consumption of indulgent treats.

Sometimes it can feel impossible to say ‘no’ to a second slice of cake, even after a hearty meal. Now, scientists have found a brain circuit that could help to explain why it is so easy to overindulge.

Working in mice, Scott Sternson at Janelia Research Campus in Ashburn, Virginia, and his colleagues identified a region of the brainstem that is home to a set of neurons whose activity is influenced by food or water intake. The scientists found that activating the neurons inhibited eating, whereas blocking their activity caused mice to eat or drink more and for longer than usual, even if the animals were well fed and hydrated.

The researchers served the mice a variety of liquids while monitoring the neurons’ activity. The team found that neuronal activity was reduced when mice drank water or bitter-tasting compounds — but the reduction was even more marked when mice consumed tasty drinks such as a vanilla-flavoured beverage.

This suggests that a neural feedback loop allows the consumption of palatable food or drinks to trick the brain into wanting more.

Portrait of the Hongwu Emperor (1328-1398), the founder of Ming dynasty

The Chinese dynasty founded by the Hongwu emperor crumbled in part because of a severe drought, which was prolonged by a volcanic eruption. Credit: Fine Art Images/Heritage Images/Getty

Climate sciences

The eruption that helped to destroy one of China’s great dynasties

Cooling particles in a volcanic plume intensified the drought that toppled the Ming dynasty.

The collapse of China’s prosperous Ming dynasty, one of the most stable in Chinese history, has been attributed, in part, to the 1641 eruption of a volcano thousands of kilometres from the imperial capital in Beijing.

Geoscientists have long known that a mega-drought that parched eastern China between 1637 and 1643 was the most severe to affect the area during the last millennium, but they did not know precisely what made it so bad. Liang Ning at Nanjing Normal University in China, Zhengyu Liu at Ohio State University in Columbus and their colleagues looked at records of past temperatures, as well as ice-core records and climate models, to unravel the mystery.

The drought kicked off as a standard dry spell. Four years later, Mount Parker in the Philippines erupted. Volcanic particles blanketed the region, cooling the air more than the ocean’s surface and setting up weather patterns that weakened the East Asian monsoon. The monsoon rains were much lighter than usual and the drought lasted for another three years.

The Ming dynasty’s fall ushered in the Qing dynasty, which imposed conservative policies and ruled for nearly three centuries.

A Somali Sengi at the Assamo locality in Djibouti.

The long-nosed mammal called the Somali sengi thrives in the deserts of Djibouti. Credit: Steven Heritage


An elephant-nosed creature ‘lost to science’ was living just next door

The Somali elephant shrew, unseen by scientists for decades, is well-known to people in Djibouti’s rocky deserts.

Do not mourn the Somali sengi, also known as the Somali elephant shrew. It was considered ‘lost to science’ after decades without any sightings being recorded in the scientific literature, and was known to biologists only from museum specimens. But the tiny insect-eating mammal — with its long tufted tail, trunk-like nose, and adorable large, dark, liquid eyes — is apparently doing fine.

The Somali sengi (Galegeeska revoilii), one of 20 sengi species, was previously thought to be endemic to its namesake nation. But now the species has been found in the neighbouring Republic of Djibouti, according to Steven Heritage at Duke University’s lemur centre in Durham, North Carolina, and his colleagues.

After co-author Houssein Rayaleh at Association Djibouti Nature in Djibouti City saw sengis in the area, the team put out 1,259 live traps in Djibouti, and successfully trapped 8 Somali sengis. Local people aided the expedition with information about the creature’s abundance and preferred habitats — a reminder that vast troves of biological and ecological knowledge reside outside the scientific enterprise.