Monday, March 02, 2020

Genetic study shows the red panda is actually two separate species

Will Dunham

WASHINGTON (Reuters) - Red pandas, the bushy-tailed and russet-furred bamboo munchers that dwell in Asian high forests, are not a single species but rather two distinct ones, according to the most comprehensive genetic study to date on these endangered mammals.

Scientists said on Wednesday they found substantial divergences between the two species - Chinese red pandas and Himalayan red pandas - in three genetic markers in an analysis of DNA from 65 of the animals.

The recognition of the existence of two separate species could help guide conservation efforts for a mammal adored by many people even as its numbers dwindle in the wild, they added.

Chinese red pandas are found in northern Myanmar as well as southeastern Tibet, Sichuan and Yunnan provinces in China, while Himalayan red pandas are native to Nepal, India, Bhutan and southern Tibet in China, the researchers said.

International experts have estimated a total population of roughly 10,000 red pandas in the wild.

“To conserve the genetic uniqueness of the two species, we should avoid their interbreeding in captivity,” said Chinese Academy of Sciences conservation biologist Yibo Hu, who along with colleague Fuwen Wei led the study published in the journal Science Advances. “Interbreeding between species may harm the genetic adaptations already established for their local habitat environment.”

Scientists previously recognized red pandas as divided into two subspecies. While it had been proposed that these were separate species, the new study was the first to provide the genetic data necessary to allow such a judgment.

The Himalayan red panda is the scarcer of the two and needs urgent protection because of low genetic diversity and small population size, Hu said. The Yalu Zangbu River most likely marks the geographical boundary separating the two species, not the Nujiang River as previously believed, Hu added.

The two species also differ in coloration and skull shape.

“The Himalayan red panda has more white on the face, while the face coat color of the Chinese red panda is redder with less white on it. The tail rings of the Chinese red panda are more distinct than those of the Himalayan red panda, with the dark rings being more dark red and the pale rings being more whitish,” Hu said.

Slightly bigger than a domestic cat, red pandas have thick fur, a short snout and pointed ears, spending much of their life in trees and dining mostly on bamboo. Major threats to red pandas include deforestation and degradation of their habitat due to human development.

Despite similar names, red pandas and giant pandas are not closely related. Giant pandas are one of the world’s eight bear species.

Red pandas, with no close living relatives, are sometimes called living fossils as the only remaining member of the Ailuridae mammalian family. They are probably most closely related to a group that includes weasels, raccoons and skunks.

There are two distinct red panda species, according to DNA analysis

by Bob Yirka , Phys.org

Genetic analysis shows two red panda species — Distinguishing morphological Chinese red panda. (B and D) The Himalayan red panda. (A and B) The face coat color of the Chinese red panda is redder with less white on it than that of the Himalayan red panda. (C and D) The tail rings of the Chinese red panda are more distinct than those of the Himalayan red panda, with the dark rings being more dark red and the pale rings being more whitish. Photo credit: (A) Yunfang Xiu, Straits (Fuzhou) Giant Panda Research and Exchange Center, China; does not require permission. (B) Arjun Thapa, Institute of Zoology, Chinese Academy of Sciences. (C) Yibo Hu, Institute of Zoology, Chinese Academy of Sciences. (D) Chiranjibi Prasad Pokheral, Centr55555555al Zoo, Jawalkhel, Lalit Credit: *Science Advances* (2020). DOI: 10.1126/sciadv.aax5751

Red pandas are cat-sized mammals that look more like red raccoons than pandas. They have red coats, masked faces and ringed tails. They live at high altitudes, and like pandas, eat bamboo. They once lived across much of Eurasia, but they are now listed as endangered—their population has dwindled to just 10,000. They live only in the southeastern and southern parts of the Qinghai-Tibetan Plateau. Prior study has shown that there are two varieties of the red panda: the Chinese red panda and the Himalayan red panda. Chinese red pandas are known for having redder faces with less of the white wisps in their fur. Their tail rings are also darker with more white between them. Scientists have wondered for some time whether the two varieties were actually distinct species. In this new effort, the researchers have found evidence that shows that they are, indeed, two distinct species—a finding that could have an impact on efforts to save them both.

The work involved sequencing the genomes of 65 wild red pandas. The researchers collected muscle, skin and blood samples from seven population areas. The analysis was focused on finding differences in DNA base pairs on the Y chromosome. The data showed that not only are the two varieties of red panda different species, but that the Chinese red panda has three populations that are genetically different. It also showed that the line that divides the two species is not the Nujiang River, as previously thought, but the Yalu Zanbu River.

The researchers suggest their findings are important for the survival of both species, because it will help to concentrate specific conservation activities that are best suited for each of them. Also, it will help to prevent interbreeding that could lead to one species overtaking another.

Distinguishing morphological differences between two red panda species. (A and C) The Chinese red panda. (B and D) The Himalayan red panda. (A and B) The face coat color of the Chinese red panda is redder with less white on it than that of the Himalayan red panda. (C and D) The tail rings of the Chinese red panda are more distinct than those of the Himalayan red panda, with the dark rings being more dark red and the pale rings being more whitish. Photo credit: (A) Yunfang Xiu, Straits (Fuzhou) Giant Panda Research and Exchange Center, China; does not require permission. (B) Arjun Thapa, Institute of Zoology, Chinese Academy of Sciences. (C) Yibo Hu, Institute of Zoology, Chinese Academy of Sciences. (D) Chiranjibi Prasad Pokheral, Central Zoo, Jawalkhel, Lalitpur, Nepal; does not require permission. Credit: Science Advances (2020). DOI: 10.1126/sciadv.aax5751

Ancient pandas weren't exclusive bamboo eaters, bone evidence suggests

More information: Yibo Hu et al. Genomic evidence for two phylogenetic species and long-term population bottlenecks in red pandas, Science Advances (2020). DOI: 10.1126/sciadv.aax5751

Swamp wallabies conceive new embryo before birth—a unique reproductive strategy

by Forschungsverbund Berlin e.V. (FVB)

Marsupials such as kangaroos or wallabies are known for their very different reproductive strategies compared to other mammals. They give birth to their young at a very early stage and significant development occurs during a lengthy lactation period in which the offspring spends most of its time in a pouch. Although in some marsupials new ovulation happens only a few hours after giving birth, the regular consecutive stages of ovulation, fertilization, pregnancy and lactation are respected—with one exception: Reproduction specialists from the Leibniz Institute for Zoo and Wildlife Research (Leibniz-IZW), Germany, and the University of Melbourne, Australia, recently demonstrated that swamp wallabies ovulate, mate and form a new embryo before the birth of the previous offspring. They thereby continuously support embryos and young at different development stages before and after birth. These findings are published in the Proceedings of the National Academy of Sciences.

Using high-resolution ultrasound to monitor reproduction in swamp wallabies during pregnancy, Prof Thomas Hildebrandt (Leibniz-IZW and University of Melbourne), Dr. Brandon Menzies and Prof Marilyn Renfree (both from University of Melbourne) were able to confirm what has been suspected for a long time: swamp wallaby females ovulate, mate and form a new embryo whilst already carrying a full-term fetus that they will soon give birth to. The new embryo enters embryonic diapause until the new-born offspring leaves the pouch nine months later. Thus, when the embryonic diapause is included, females are continuously pregnant throughout their reproductive life, a unique reproductive strategy that completely blurs the normal staged system of reproduction in mammals.

This phenomenon is made possible by two anatomically completely separated uteri and cervices connected to ovaries by their oviducts. "This is true for all marsupials, but the unique overlapping reproductive cycles seem to be a special feature of the swamp wallabies," says Renfree. Normally, ovulation alternates between the two ovaries. "All female macropodid marsupials—essentially kangaroos, wallabies and a few other groups of species—except the swamp wallaby have an oestrous cycle longer than the duration of their pregnancy, so females come into oestrus, ovulate and mate within hours after birth." It has been suspected for some time that swamp wallabies might conceive during an active pregnancy, because the oestrous cycle of the swamp wallaby is shorter than the duration of their pregnancy and there have been reports about mating before the birth of the previous offspring. Such a "superfetation" has previously been only described (by Leibniz-IZW scientists) for the European brown hare where females copulate again three to four days before the birth of the incumbent young, forming new conceptuses during an active pregnancy.

In order to confirm superfetation in swamp wallabies, the scientists removed the pouch young of ten females to reactivate the dormant blastocysts (early stage embryo). They then monitored the development of the blastocyst in four of these ten females using high-resolution ultrasound. All females gave birth at around 30 days after the young had been removed. Parallel to the embryo development in one uterus, the scientists closely examined the opposite ovary. There, follicles started to appear and grow. At day 26 of the pregnancy the ultrasound examination showed that the conceptus had developed into a fetus with the head, limbs and heartbeat clearly visible—and at day 28 and 29 the largest follicle in the opposite (contralateral) ovary had ovulated and a new corpus luteum was evident. The other six females that were not scanned with ultrasound were regularly examined for sperm. Sperm was identified in the urogenital tract one to two days before birth but at no other time. "These results clearly demonstrate that swamp wallabies ovulate and mate one to two days before birth, during an existing pregnancy," says Hildebrandt.

Pregnancies of eutherian mammals (most mammals, i.e. the most taxonomically diverse of the three branches of mammals) greatly exceed the length of the oestrous cycle, so during mammalian evolution, there has been selection pressure to extend the duration of pregnancy. Among marsupials (who form a second taxonomic branch of mammals), gestation in most macropodids encompasses almost the entire duration of the oestrous cycle. The swamp wallaby takes this one step further with its pre-partum oestrus, allowing this marsupial's gestation length to exceed the oestrous cycle length.

Sadly, many of these unique animals have been lost in the current disastrous bushfires in Australia this summer.

The climb to the pouch begins in utero

More information: Brandon R. Menzies el al., "Unique reproductive strategy in the swamp wallaby," PNAS (2020). www.pnas.org/cgi/doi/10.1073/pnas.1922678117

Logging to start in bushfire-ravaged New Zealand forests this week

by David Lindenmayer and Doug Robinson, The Conversation

New South Wales' Forestry Corporation will this week start "selective timber harvesting" from two state forests ravaged by bushfire on the state's south coast.

The state-owned company says the operations will be "strictly managed" and produce timber for power poles, bridges, flooring and decking.

Similarly, the Victorian government's logging company VicForests recently celebrated the removal of sawlogs from burnt forests in East Gippsland.

VicForests says it did not cut down the trees—they were cut or pushed over by the army, firefighters or road crews because they blocked the rood or were dangerous. The company said it simply removed the logs to put them "to good use."

However the science on the impacts of post-fire logging is clear: it can significantly impair the recovery of burned ecosystems, badly affect wildlife and, for some animal species, prevent recovery.

We acknowledge that for safety reasons, some standing and fallen burnt trees must be removed after a fire. But wherever possible, they should remain in place.

Damaging effects

Hollows in fire-damaged trees and logs provide critical habitat for animal species trying to survive in, or recolonize, burned forests.

Detailed studies around the world over the past 20 years, including in Australia, have demonstrated the damage caused by post-fire logging.

Indeed, the research shows post-fire logging is the most damaging form of logging. Logging large old trees after a fire may make the forests unsuitable habitat for many wildlife species for up to 200 years.

Long-term monitoring data from extensive field surveys shows hollow-dependent mammals, such as the vulnerable greater glider, generally do not survive in areas burned and then logged. Research by the lead author, soon to be published, shows populations are declining rapidly in landscapes dominated by wood production.

Forests logged after a fire have the lowest bird biodiversity relative to other forests, including those that burned at high severity (but which remain unlogged). Critical plants such as tree ferns are all but eradicated from forests that have been burned and then logged.

Soils remain extensively altered for many decades after post-fire logging. This is a major concern because runoff into rivers and streams damages aquatic ecosystems and kills organisms such as fish.

A double disturbance

Fire badly disrupts forest ecosystems. Animals and plants then begin recovering, but most forests and the biota they support simply cannot deal with the second intense disturbance of logging so soon after a first one.

For example, young germinating plants are highly vulnerable to being flattened and destroyed by heavy logging machinery. And in an Australian context, post-fire logging makes no sense in the majority of eucalypt-dominated ecosystems where many tree species naturally resprout. This is an essential part of forest recovery.

Logs provide shade, moisture and shelter for plants, and rotting timber is food for insects—which in turn provide food for mammals and birds.

Living and dead trees are also important for fungi—a food source for many animals, including bandicoots and potoroos which have been heavily impacted by the fires.

Similarly on burnt private land, removing damaged and fallen trees will only hinder natural recovery by removing important animal habitat and disturbing the soil. If left, fallen trees will provide refuge for surviving wildlife and enable the natural recovery of forests.

While the sight of burnt timber can be disheartening, landholders should resist the urge to "clean up."

It doesn't add up

Research in North America suggests debris such as tree heads, branches and other vegetation left by post-fire logging not only hinders forest regeneration, but can make forests more prone to fire.

And the economics of logging, particular after a fire, is dubious at best. Many native forest logging operations, such as in Victoria's East Gippsland, are unprofitable, losing millions of taxpayer dollars annually.

Timber is predominantly sold cheaply for use as woodchips and paper pulp and fire-damaged timber is of particularly poor quality. Even before the fires, 87% of all native forest logged in Victoria was for woodchips and paper pulp.

Post-fire logging certainly has no place in national parks. But for the reasons we've outlined, it should be avoided even in state forests and on private land. Million hectares of vegetation in Australia was damaged or destroyed this fire season. The last thing our forests need is yet more disturbance.

VicForests response: VicForests told The Conversation that timber currently being removed by VicForests, at the direction of the Chief Fire Officer, is from hazardous trees that were cut or knocked over to enable the Princes Highway to be re-opened.

It said the timber would be used for fence restoration, firewood and to support local mills "protecting jobs, incomes and families. It would otherwise be left in piles on the side of the highway."

"Any further post-fire recovery harvesting will occur in consultation with government including biodiversity specialists and the conservation regulator, following careful assessment and protection of high conservation values," VicForests said.

The company said post-fire recovery harvesting, particularly of fire-killed trees, does not increase fire risk.

"Sensitive harvesting including the retention of habitat trees and active re-seeding is more likely to result in a successfully regenerated forest and a supportive environment for threatened species. This regenerating forest will have the same fire risk as natural regeneration following bushfire."

Forestry Corporation of NSW response: Forestry Corporation of NSW said in a statement that small-scale selective timber harvesting operation will begin on the south coast this week.

The company's senior planning manager Dean Kearney said the Environment Protection Authority, with the input of scientific experts "has provided Forestry Corporation with site-specific conditions for selective timber harvesting operations in designated parts of Mogo and South Brooman State Forests. These areas were previously set aside for timber production this year but have now been impacted by fire."

"Strictly-managed selective timber harvesting will help prevent the loss of some high-quality timber damaged by fire, including material that will be in high demand for rebuilding, while ensuring the right protections are in place for key environmental values, particularly wildlife habitat, as these forests begin regenerating," he said.

"The harvesting conditions augment the already strict rule set in place for forest operations and include requirements to leave all unburnt forest untouched and establish even more stringent conditions to protect water quality, hollow-bearing trees and wildlife habitat."

Wildlife needs fire-damaged and dead trees after fires

Provided by The Conversation

Transforming potato waste into a new industry for Australia

by Fight Food Waste CRC

Four of the largest potato producers in Australia want to convert 100% of their potato waste into commercial benefit through their partnership with the Fight Food Waste Cooperative Research Centre (CRC).

Over the next three years, The Mitolo Group, Zerella Fresh, Thomas Foods International Fresh Produce, The South Australian Potato Company, together with Industry Association; Potatoes South Australia Inc, and The University of Adelaide will invest nearly $1m in this research and development to save up to 100,000 tonnes of potatoes currently going to waste every year.

Chief Executive of Potatoes South Australia Robbie Davis says that this is a fantastic opportunity for Australia, particularly South Australia as it is the largest potato growing state.

"We are seeing up to 40% of potatoes rejected because they do not meet retail specifications. At the same time Australia is importing 20,000 tonnes of potato starch each year, and it just doesn't make sense that we're not using these huge volumes of potatoes for alternative purposes," she said.

A large focus of this project is the potential development of an Australian potato starch industry which would provide additional revenue for Australian potato companies; potentially $1000 a tonne for extracted starch instead of the current value of $0-10 a tonne for the waste.

"Potato starch is used broadly across the food industry, from bioplastics and packaging, to coatings and adhesives. We also want to use the waste from the waste, so after extracting the potato starch, there will be further opportunities using the residual waste from this first stage," says Ms Davis.

The four Australian potato companies that have partnered with the Fight Food Waste CRC are leaders in their industry and recognise the opportunity this represents to the industry.

Professor Vincent Bulone from the University of Adelaide is leading this research project from his world-class analytical centre for complex carbohydrate analysis, Adelaide Glycomics. The project is in line with the University's industry engagement priority on agrifood and wine.

"There are different forms of starch in potatoes that can be used in different products. For example, existing research suggests that the less digestible starches in potatoes, the so-called 'resistant starches," can be used to make superior pre-biotics that help prevent infections," says Professor Bulone.

"Another known starch component can be used to engineer low GI foods, and the skins of the potatoes themselves contain bioactives that can be used for a range of commercial products like nutraceuticals."

Fight Food Waste CRC CEO Dr. Steven Lapidge is thrilled to have such a transformational project underway so early in the Fight Food Waste CRC's journey and sees the partnership between all of the potato producers as a great example of what CRCs can achieve.

"We're looking to develop new products from current waste streams that will deliver additional profit to potato producers through domestic and export sales.

"Through investing in research and development we aim to deliver new high-value commercial opportunities for the participants of this project.

"This project is exactly what the CRC is all about; delivering real benefit for Australian businesses across the whole of the value chain."

Explore further

Summer drought may shrink supplies of French spuds

More information: Fight Food Waste: www.fightfoodwastecrc.com.au

Provided by Fight Food Waste CRC

New eggplant varieties resistant to extreme conditions

by Universitat Politècnica de València

The Universitat Politècnica de València (UPV), through the Institute of Conservation and Improvement of Valencian Agrodiversity (COMAV), leads EggPreBreed II, an international project that will help in developing new varieties of eggplants more resistant to extreme drought conditions, and to two of the most serious pathologies that affect this crop, such as the Fusarium fungus and nematodes. These phytopathological enemies of eggplant can cause significant crop losses and their incidence is expected to be even greater due to climate change.

Eggplant is one of the thirty-five crops considered as most important for world food security and, as such, is included in Annex 1 of the FAO International Treaty on Plant Genetic Resources for Food and Agriculture (ITPGRFA).

The EggPreBreed II project will help to develop new varieties of this crop with better properties. For this, the COMAV team works in new crosses of eggplants grown with related wild species such as, for example, Solanum incanum and Solanum elaeagnifolium, which have a high resistance to drought. With these crosses EggPreBreed II will allow to obtain genetic material of eggplant that in most of its characters—fruit size, color, composition, etc. - is indistinguishable from a presently cultivated eggplant, but they have fragments of the wild species genome that can confer properties of great value to the market.

"By introducing genes from these wild species we can improve drought tolerance, get higher eggplant production with the same amount of water. But, in addition, we can also take advantage of other characteristics of interest of wild species, for example, their high content in phenolic compounds, of great interest for human health. Therefore, this project opens the door to a new generation of commercial varieties of eggplant that can have greater efficiency in the use of water and with better functional properties," explains Jaime Prohens, director of COMAV-UPV and principal investigator of EggPreBreed II.

Seven years of research

The project is a continuation of research that the COMAV-UPV team has been developing since 2013. In these years, they have obtained different lines of introgression—genetic material of eggplant that includes a fragment of the genome of other donor species, in this case wild relatives.

"The objective now is to refine these materials, improve them to reduce the unfavorable characteristics of wild species, such as the presence of prickles or the intense bitterness of the fruit, obtaining pure lines resistant to stress caused by climate change, Fusarium and nematodes. Having this material will make it easier for seed companies to use and incorporate it into breeding pipelines in order to develop new commercial varieties in a crop as important worldwide as eggplant," says Prohens.

The project is part of the global initiative "Adapting Agriculture to Climate Change: Collecting, Protecting and Preparing Wild Relatives", led by the Global Crop Diversity Trust (Crop Trust) and funded by the Norwegian Government.

"In the first phase of this project, Jaime and his team managed to cross domestic varieties of eggplant with 15 different wild relatives, an impressive achievement," indicates Benjamin Kilian, representative of the Crop Trust, who says that "we are pleased to support the scientists of the UPV in this second phase; we are sure that the fruits of this effort will help to adapt our agriculture to climate change."

The EggpreBreed-II project also involves an Egyptian university (University of Kafrelsheikh), involved in the analysis of resistance to Fusarium and nematodes, as well as several seed companies from the Philippines, Egypt, France and Spain that will test the commercial utility of the plant materials developed in the project and will incorporate them in their breeding programs to develop the new varieties.

All agronomic and genetic data generated throughout this project will be shared in the future on the online platform Germinate, a database platform that provides a standard and common interface to genetic resources collections.

International project to create climate change-resistant eggplants

More information: Germinate: ics.hutton.ac.uk/get-germinate/

International Treaty for Plant Genetic Resources for Food and Agriculture
www.upv.es/visor/rtv/63121/c

Machine learning picks out hidden vibrations from earthquake data

by Jennifer Chu, Massachusetts Institute of Technology

Over the last century, scientists have developed methods to map the structures within the Earth's crust, in order to identify resources such as oil reserves, geothermal sources, and, more recently, reservoirs where excess carbon dioxide could potentially be sequestered. They do so by tracking seismic waves that are produced naturally by earthquakes or artificially via explosives or underwater air guns. The way these waves bounce and scatter through the Earth can give scientists an idea of the type of structures that lie beneath the surface.

There is a narrow range of seismic waves—those that occur at low frequencies of around 1 hertz—that could give scientists the clearest picture of underground structures spanning wide distances. But these waves are often drowned out by Earth's noisy seismic hum, and are therefore difficult to pick up with current detectors. Specifically generating low-frequency waves would require pumping in enormous amounts of energy. For these reasons, low-frequency seismic waves have largely gone missing in human-generated seismic data.

Now, MIT researchers have come up with a machine learning workaround to fill in this gap.

In a paper appearing in the journal Geophysics, they describe a method in which they trained a neural network on hundreds of different simulated earthquakes. When the researchers presented the trained network with only the high-frequency seismic waves produced from a new simulated earthquake, the neural network was able to imitate the physics of wave propagation and accurately estimate the quake's missing low-frequency waves.

The new method could allow researchers to artificially synthesize the low-frequency waves that are hidden in seismic data, which can then be used to more accurately map the Earth's internal structures.

"The ultimate dream is to be able to map the whole subsurface, and be able to say, for instance, 'this is exactly what it looks like underneath Iceland, so now you know where to explore for geothermal sources,'" says co-author Laurent Demanet, professor of applied mathematics at MIT. "Now we've shown that deep learning offers a solution to be able to fill in these missing frequencies."

Demanet's co-author is lead author Hongyu Sun, a graduate student in MIT's Department of Earth, Atmospheric and Planetary Sciences.

Speaking another frequency

A neural network is a set of algorithms modeled loosely after the neural workings of the human brain. The algorithms are designed to recognize patterns in data that are fed into the network, and to cluster these data into categories, or labels. A common example of a neural network involves visual processing; the model is trained to classify an image as either a cat or a dog, based on the patterns it recognizes between thousands of images that are specifically labeled as cats, dogs, and other objects.

Sun and Demanet adapted a neural network for signal processing, specifically, to recognize patterns in seismic data. They reasoned that if a neural network was fed enough examples of earthquakes, and the ways in which the resulting high- and low-frequency seismic waves travel through a particular composition of the Earth, the network should be able to, as they write in their paper, "mine the hidden correlations among different frequency components" and extrapolate any missing frequencies if the network were only given an earthquake's partial seismic profile.

The researchers looked to train a convolutional neural network, or CNN, a class of deep neural networks that is often used to analyze visual information. A CNN very generally consists of an input and output layer, and multiple hidden layers between, that process inputs to identify correlations between them.

Among their many applications, CNNs have been used as a means of generating visual or auditory "deepfakes"—content that has been extrapolated or manipulated through deep-learning and neural networks, to make it seem, for example, as if a woman were talking with a man's voice.

"If a network has seen enough examples of how to take a male voice and transform it into a female voice or vice versa, you can create a sophisticated box to do that," Demanet says. "Whereas here we make the Earth speak another frequency—one that didn't originally go through it."

Tracking waves

The researchers trained their neural network with inputs that they generated using the Marmousi model, a complex two-dimensional geophysical model that simulates the way seismic waves travel through geological structures of varying density and composition.

In their study, the team used the model to simulate nine "virtual Earths," each with a different subsurface composition. For each Earth model, they simulated 30 different earthquakes, all with the same strength, but different starting locations. In total, the researchers generated hundreds of different seismic scenarios. They fed the information from almost all of these simulations into their neural network and let the network find correlations between seismic signals.

After the training session, the team introduced to the neural network a new earthquake that they simulated in the Earth model but did not include in the original training data. They only included the high-frequency part of the earthquake's seismic activity, in hopes that the neural network learned enough from the training data to be able to infer the missing low-frequency signals from the new input.

They found that the neural network produced the same low-frequency values that the Marmousi model originally simulated.

"The results are fairly good," Demanet says. "It's impressive to see how far the network can extrapolate to the missing frequencies."

As with all neural networks, the method has its limitations. Specifically, the neural network is only as good as the data that are fed into it. If a new input is wildly different from the bulk of a network's training data, there's no guarantee that the output will be accurate. To contend with this limitation, the researchers say they plan to introduce a wider variety of data to the neural network, such as earthquakes of different strengths, as well as subsurfaces of more varied composition.

As they improve the neural network's predictions, the team hopes to be able to use the method to extrapolate low-frequency signals from actual seismic data, which can then be plugged into seismic models to more accurately map the geological structures below the Earth's surface. The low frequencies, in particular, are a key ingredient for solving the big puzzle of finding the correct physical model.

"Using this neural network will help us find the missing frequencies to ultimately improve the subsurface image and find the composition of the Earth," Demanet says.

Artificial intelligence improves seismic analyses

Meteorite observation network sets out to catch a falling star

by University of Glasgow

British scientists are turning their eyes to the skies to track meteorites before they land on UK soil—and they're looking for volunteers to help them recover the space rocks whenever and wherever they fall.

The UK Fireball Network, led by researchers from the University of Glasgow and Imperial College London is setting out to place 10 cameras across the country to catch glimpses of the spectacular natural firework displays caused when meteoroids enter the Earth's atmosphere. It is part of a wider network called the Global Fireball Observatory operated by researchers in Australia at Curtin University

On February 16, the network's cameras in Lincoln and Cambridge both saw their first fireball, which dropped meteorites into the North Sea. While those meteorites are impossible to recover, the network team expect that future sightings that fall on the UK landmass can be properly triangulated to determine a landing area and send out a search party to find the meteorites.

Dr. Luke Daly, of the University of Glasgow's School of Geographical and Earth Sciences, is one of the leaders of the UK Fireball Network. He said: "Meteorites provide scientists with invaluable insights into other planets and our solar system. A good deal of what we know about the surface of Mars, for example, comes from analysis of chunks of the planet which were blasted off its surface by asteroid impacts millions of years ago and drifted in space before falling to Earth.

"Meteorites enter our atmosphere all the time, but the UK hasn't had a great track record of finding them in recent years—in fact, it's been nearly 30 years since one was last seen dropping into a back garden in Glatton in Cambridgeshire, and more than 100 since one was observed in Scotland.

"Camera networks like ours, and those of partner organizations like UKMON and SCAMP, give us a great chance to capture fireballs on multiple cameras. Together we are building an integrated pipeline to use all the data from all the networks to track fireballs called the UK Fireball Alliance. Catching images on more than one camera allows us to estimate not just where they land, but calculate the trajectory of their arrival, which allows us to calculate where in space it came from. Pooling all our resources maximizes the chances of capturing these elusive events.

"It was very exciting to capture our first images of a fireball caught by two of our observatories, although it was obviously disappointing that the material it dropped fell into the sea. When we do catch sight of a fireball dropping meteorites on land, we'll need the help of volunteers to help comb the countryside to find them, so anyone interested in making a little bit of history by getting involved can follow us on Twitter at @FireballsUK."

The network's approach to meteorite observation has already been successful in Australia, where the Desert Fireball Network established by Curtin University has observed and recovered numerous meteorite falls in the outback. Initial funding for the UK Fireball Network was supplied by the Desert Fireball Network and the Australian Research Council, along with the University of Glasgow.

Currently the UK Fireball Network has set up six of their 10 cameras at sites in England, Scotland and Wales. Over the next couple of months, the remaining cameras will be placed in other locations, including Northern Ireland.

Meteorite search about to begin

Online STEM demonstrations can be as effective as classroom teaching

by Anna Fiorentino, Harvard University

YouTube has become the go-to for quick tutorials on almost any topic, from how to replace a zipper to how to install a water heater. But could some of the most memorable parts of a STEM course—live demonstrations—be brought to the screen effectively? In a new paper, Harvard researchers show for the first time that research-based online STEM demonstrations not only can teach students more, but can be just as enjoyable.

Researchers hope these findings will help spur the creation of a catalog of free online STEM video demonstrations to supplement lectures at institutions that cannot conduct their own. "We have an incredible group of scientists who present live demos for our students, but very few schools have these dedicated resources," said co-author Logan McCarty, director of science education in the Department of Physics, who oversees Harvard's Lecture Demonstration team. "With YouTube and other online channels, we can share Harvard's technical and pedagogical expertise with the world."

The research was based on previous literature by Kelly Miller, a lecturer in applied physics and co-author with McCarty. The previous article, published in 2013 by Miller and Eric Mazur, Balkanski Professor of Physics and Applied Physics, showed that students often misunderstand lecture demonstrations. They turned to science demos after hearing time and again that they are students' favorite part of the lecture. After all, who could forget a ball levitating on a sound wave or a laser bending into a tank of water?

"Our research suggests that when live demos are unavailable, videos can provide students with an equally effective—or possibly even more effective—learning experience," said co-author Louis Deslauriers, director of science teaching and learning in the Faculty of Arts and Sciences. "Even when live demonstrations are available, it may be helpful to supplement them with high-quality videos."

Their paper in the February issue of Physical Review Physics Education Research was spun into motion by first author Greg Kestin, a preceptor in physics who produces a series with NOVA called "What the Physics?!"

A sample of one of the video demonstrations shown to physics students. “Shoot the Monkey” illustrates gravitational acceleration by releasing a suspended stuffed animal while simultaneously shooting it with a cannon ball. The goal: to determine where to aim in order to hit the falling monkey. Credit: Greg Kestin

"There was already a rich array of findings about how to bring multimedia into education to increase student learning and motivation," said Deslauriers. "What Greg brings which no one else has is a unique combination of expertise in physics, video production, and cognitive psychology of multimedia presentations."

The investigators identified several ingredients in improving learning through online demonstrations. A number of these advantages relied on the filmmaker's ability to set the demonstrations in slow motion, the use of graphics to visualize abstract concepts, and the ability to direct attention to the most important features to prevent information overload.

For the study, they split an introductory physics class into two groups. The first watched a video of a short science demonstration in a classroom setting, while the other group saw the experiment live in the same classroom. They then flipped the groups and performed a second demonstration to reduce bias. Although the demos were virtually identical, the group who first saw the live presentation watched gestures, blackboard illustrations, and physical props, such as large wooden arrows, while the group who viewed video demos followed superimposed graphics, slow-motion video, and animations. In the end, students from both groups who watched the online video scored higher on tests on the material presented, without any reported decrease in enthusiasm.

"Students don't always know where to focus their attention in live demonstrations, even with a chalkboard and props, but in a video format we can highlight and zoom in on particular parts by setting narration at just the right moment to reinforce learning. A video can also present contrasting cases, which is known as a potent learning tool," said Deslauriers.

"Lecture demos have been a core component of STEM courses for over 100 years, but not every school has the resources for them," said Kestin. "We wanted to understand how much students were learning in video demos in the hopes that with the right support we can bring the same excitement to schools everywhere."Study shows students in 'active learning' classrooms learn more than they think

More information: Greg Kestin et al. Comparing the effectiveness of online versus live lecture demonstrations, Physical Review Physics Education Research (2020). DOI: 10.1103/PhysRevPhysEducRes.16.013101

LA REVUE GAUCHE - Left Comment