Research Blog

An Unconventional Career Map

On December 27, 2015

In Climate/Global Change, Computers/Technology, Environment/Sustainability, Field Research, Guest Post

Landscape ecologist Jennifer Swenson has a very special set of skills that come from having an “unconventional academic trajectory.”

Jennifer Swenson is an associate professor of the practice in geospatial analysis at the Nicholas School of the Environment

Her diverse and interesting career began at UC Santa Barbara, where she chose to learn about International Relations and Geography. “I wanted to be versatile and globally aware,” she said. Undergraduate school is where she learned the geography techniques she now uses in her research, and received stacks of reading for international relations.

After this, Swenson spent three years giving bicycle tours, working at a ski resort, and other jobs, until she went on a conservation trip to Ecuador to work for an NGO (non-governmental organization). She was able to use her geospatial techniques (GIS) to map trails and land cover change in Ecuadorian national parks, and also to evaluate forest corridors for an endangered species of monkey.

Connectivity between habitat remnants for critically endangered primate, Callicebus oenanthe, in San Martin, Peru. Presented at the 2nd Simposio de Primatologia en el Peru (Iquitos, November 2013) & at the Remote Sensing forConservation Symposium (London, May 2014) Schaffer-Smith, Swenson, Bóveda-Penalba, Murrieta-Villalobos

She learned many things, including how to manage a lab, and also became fluent in Spanish thanks to the total immersion. “It’s just another barrier,” she says, to have to use English. Plus, it is useful for reading papers that haven’t been translated.

“Everyone should learn a second or third language and have the opportunity to be immersed in that country.”

After this she went back to graduate school and got a Ph.D. in forest science at Oregon State.

Swenson’s research at Duke is often about conservation or biodiversity, and occasionally ecosystem studies. She is still using her special skills to try to do the greatest good.

“Its great to work towards that, but sometimes its hard to detect that you are doing change,” she says. “I still keep trying to forge ahead and do whatever I can for the environment. In the end, all those students that we train and send out will do great things, and that’s how we have the greatest impact for the environment.”

Guest Post by Caleb Caton, a senior at the North Carolina School of Science and Math

Checking in on Air Pollution — With an Expert!

By Karl Bates

On December 23, 2015

In Climate/Global Change, Global Health, Guest Post, Medicine

I had the pleasure of meeting and interviewing Duke Professor Junfeng “Jim” Zhang, an avid environmental health researcher who has done amazing work on air pollution research. Because of my own interest in air pollution and its adverse health effects, I began by trying to grasp Dr. Zhang’s work through his detailed scientific explanations of his projects, such as looking at the human health effects of nano-technology.

Junfeng (Jim) Zhang is a professor of global and environmental health in the Nicholas School and the Duke Global Health Institute

From my reading on the projects I was interested in, I did not expect Zhang to be able to step back and capture the importance of his research in simple terms because his projects were quite complicated (such as testing human health effects due to chemically altered diesel fuel). However, it turned out that Zhang is well-versed in communicating both crucial details of his research and the overall meaning for human health.

The most captivating aspect of my interview with Zhang was our discussion of his contribution to the 2007 Nobel Peace Prize, which was awarded to the Intergovernmental Panel on Climate Change. Dr. Zhang essentially concluded that burning and replanting trees is not necessarily carbon-neutral, disproving the common view that replanting burned trees is always carbon-neutral. His arguments really sold the importance of his research to me and I very quickly agreed with his views on the consequences of his research.

My interview with Zhang also revealed just how important scientific research is in adding evidence and findings to support a side on the growing global issues of environmental pollution and protection. It really amazed me how researchers play such a crucial supportive role in not only protecting the world’s environment, but advancing the quality of human life.

UPDATE – Professor Zhang spoke with BBC about air pollution in China. Read the story here.

Guest Post by Peter Cheng, a senior at the North Carolina School of Science and Math.

Improving Machine Learning With an Old Approach

By Karl Bates

On December 21, 2015

In Computers/Technology, Guest Post, Mathematics

Computer scientist Rong Ge has an interesting approach to machine learning. While most machine learning specialists will build an algorithm which molds to a specific dataset, Ge builds an algorithm which he can guarantee will perform well across many datasets.

Rong Ge is an assistant professor of computer science.

A paper he wrote as a postdoc at Microsoft Research, Escaping From Saddle Points — Online Stochastic Gradient for Tensor Decomposition , describes how a programmer can use the imprecision of a common machine learning algorithm, known as stochastic gradient descent, to his advantage.

Normally this algorithm is used to speed up a slow learning process by only approximating the correct solution rather than working harder to get precision; however, Ge and his colleagues found that the small amount of “noise” created by the algorithm can be the saving grace of an algorithm which would otherwise be trapped by its own perfectionism.

“This algorithm is not normally used for this purpose,” Ge says, “It is normally used as a heuristic to approximate the solution to a problem.”

Noise allows the algorithm to escape from something called a “saddle point” on the function which the stochastic gradient is trying to optimize, which looks sort of like a sine wave. Ge describes gradient descent as being like a ball rolling down a hill. When on the slope of the hill it always seeks the lowest point, but if it is at a saddle point, a high point on a “ridge” between two different slopes, it will not start rolling.

Stochastic gradient descent remedies this problem by jostling the ball enough to start it rolling down one of the slopes. But one cannot be sure which way it will roll.

The results he has obtained relate to a certain branch of machine learning known as unsupervised learning. One common problem from unsupervised learning is “clustering,” in which the algorithm attempts to find clusters of points which are similar to each other while different from the other points in the set. The algorithm then labels each of the clusters which it has found, and returns its solution to the programmer.

A key requirement for the final result of the algorithm to be correct is that the two slopes end at low points of equal depth, which represent optimal solutions. Often two solutions will appear different at first glance, but will actually represent the same set of clusters, different only because the labels on clusters were switched. Ge said one of the hardest parts of his process was designing functions that have this property.

These results are guaranteed to hold so long as the dataset is not provided by someone who has specifically engineered it to break the algorithm. If someone has designed such a dataset the problem becomes “NP hard,” meaning that there is little hope for even the best algorithms to solve it.

Hopefully we will see more growth in this field, especially interesting results such as this which find that the weaknesses associated with a certain algorithm can actually be strengths under different circumstances.

GraysonYork

Guest post by Grayson York, a junior at the North Carolina School of Science and Math

Federal Budget Raises NIH, Other R & D

By Karl Bates

On December 17, 2015

In Data, Faculty, Statistics

Reversing more than a decade of flat growth in research funding, the federal budget proposal announced Wednesday includes a $2 billion increase for the National Institutes of Health, the major source of Duke’s federal research funding. The increase would bring NIH’s budget to $32.1 billion.

Budget trends by agency. Graph courtesy of AAAS.

The budget must pass both houses of Congress and be signed by the president to take effect, but the figures being announced are the result of negotiations between House and Senate budget committees.

According to Chris Simmons, Duke’s associate vice president for government relations, other research funds Duke relies on are also slated to increase.

The National Science Foundation’s budget would increase to $7.46 billion, up $119.3 million over 2015.

The Department of Energy’s Office of Science would grow 5.6 percent to $5.35 billion.

NASA science programs would increase 6.6 percent to $5.6 billion.

Simmons added that most of the major research accounts at the Department of Defense (Basic, Applied and Advanced Technology Development) will receive increased funding. “Unfortunately Air Force Research and DARPA will see a cut in their funding by $20 and $25 million, respectively.”

“This expansion of R&D funding is very encouraging news,” said Duke Vice Provost for Research Larry Carin. “Federal investments in university research have long been a powerful engine of the American economy and we’re heartened to see policymakers returning to that commitment.”

Long-term trends by R&D purpose. Note the little spike for the economic stimulus package. Graph courtesy of AAAS.

Within the increased NIH funding, the budget includes $200 million for the Precision Medicine Initiative, a $350 million increase for Alzheimer’s disease research and an $85 million increase for the BRAIN Initiative.

“Duke has strong research positions in all three of these areas, so we’re pleased by those particular expansions,” Carin said. “Now, of course, our faculty will have to get out there and write the grants that bring that money to North Carolina.”

Post by Karl Leif Bates

Middle Schoolers Ask: What's it Like to be a Scientist?

By Robin Smith

On December 16, 2015

In Animals, Behavior/Psychology, Biology, Chemistry, Environment/Sustainability, Physics, Science Communication & Education, Students

When a group of local middle schoolers asked four Duke postdocs what it’s like to be a scientist, the answers they got surprised them.

For toxicologist Laura Maurer, it means finding out if the tiny silver particles used to keep socks and running shirts from getting smelly might be harmful to your health.

For physics researcher Andres Aragoneses, it means using lasers to stop hackers and make telecommunications more secure.

And for evolutionary anthropologist Noah Snyder-Mackler, it means handling a lot of monkey poop.

The end result is a series of short video interviews filmed and edited by 5th-8th graders in Durham, North Carolina. Read more about the project and the people behind it at http://sites.duke.edu/pdocs/, or watch the videos below:

Five Duke Papers Crack the Altmetric 100

By Karl Bates

On December 14, 2015

In Animals, Behavior/Psychology, Biology, Medicine

The numbers are in, and five papers with Duke authors cracked the Top 100 Altmetric scores for 2015.

Example of an Altmetric analysis.

Yeah, it all seems a little gimmicky and meta, but the scores can be useful. Altmetric (to which Duke has an institutional membership) combines multiple counts of news stories, social media chatter and professional citations on an academic paper to give it a single score. Obviously, the system’s greatest strength is comparing this to other Altmetric scores, but it’s actually a lot of fun.

Duke’s biggest score – a very impressive Altmetric 2294 – came in at #5 on the list. “Estimating the reproducibility of psychological science” attracted a lot of attention in Science, spawning 74 news stories and nearly 2,000 tweets. Postdoctoral researcher Nina Strohminger of the Kenan Institute for Ethics is one of the authors from 125 institutions on the paper that suggests psychology has some housekeeping to do.

At number 28 with an Altmetric of 1,279, came “Global, regional and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013 (here comes the colon!): a systematic analysis for the Global Burden of Disease Study 2013.” This Lancet paper, backed by the Bill and Melinda Gates Foundation, is every bit as massive and important as its title. Among its thousands of authors is our own Terrie Moffitt. It garnered 39 news stories and 1400 tweets and has already been incorporated into nine Wikipedia entries.

A companion paper with another big title for another big study, “Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013,” came in at #36 on the list with an Altmetric of 1180. Its authors might be jealous of #28, but it’s mostly the same folks! Eighty people saw fit to post this one on Facebook and 60 on Google+.

Two papers out of the now-defunct NSF think-tank the National Evolutionary Synthesis Center (NESCent) rounded out our top 100 at #72 and #87.

What are the largest ocean giants?

Craig McClain of NESCent and Duke Biology led “Sizing Ocean Giants,” an analysis that tries to get the right dimensions on a bunch of intimidating ocean creatures including the giant clam and the colossal squid (which turns out to be only a third the size of the less impressively named giant squid). The paper’s Altmetric of 954 was led by 24 news stories, 24 blog posts and almost 900 tweets. McClain also leads a very popular marine science blog “Deep Sea News” which probably aided the story’s social presence.

Number 87 was “Synthesis of phylogeny and taxonomy into a comprehensive tree of life,” which included Karen Cranston of NESCent and Duke Biology. This hugely ambitious effort to draw a tree of life for the whole planet at once earned an Altmetric of 895 by garnering 21 news stories, 12 blogs and nearly 900 tweets. And it too has been incorporated into Wikimedia – once so far.

It’s a brave new world out there in academic publishing.

Post by Karl Leif Bates

Dam Good Research on Invasive Beavers in Patagonia

By Devin Nieusma

On December 11, 2015

In Animals, Behavior/Psychology, Biology, Environment/Sustainability, Field Research

Alejandro Pietrek and a subject of his research. Photo by Duke Forward.

For three years, Duke student Alejandro Pietrek has bravely grappled with some unusual marauders of the forests and steppes of Patagonia: invasive beavers. A biology graduate student, Pietrek recently presented his dissertation on the “Demography of invasive beavers in the heterogeneous landscapes of Patagonia.” Pietrek has studied over two dozen colonies of beavers in order to answer three questions:

“How do differences between habitats affect the demography of invaders?”
“How does density dependence affect the abundance and distribution of invaders post-establishment?”
“How can we manage biological invasions?”

Pietrek began by explaining how the furry rodents started ravaging the natural habitat of Patagonia, beginning with 20 beavers intentionally introduced into the forests of Tierra del Fuego in 1946. By the late 1960s, the growing colonies spread to the continent.

Today there are an estimated 100,000 individuals in Patagonia, disrupting the regional habitats and destroying biodiversity.

The biologist tackled his first question via quantitative science. “Very simply mathematical models have shown that speed of invasion is determined by two main things: population growth rate and movement,” he said. How the different habitats of Patagonia affected this invasion was what he worked to find out, by measuring colony size and the number of kits. Pietrek directly observed 25 colonies in each habitat over three years, using binoculars and seemingly endless patience. “It was very fun,” he said. He found that steppe habitats tended to have higher numbers of beavers and kits compared to forest.

Beaver Dam – Tierra del Fuego National Park, Argentina – Photograph by Anne Dirkse via Wikimedia Commons

Pietrek believed that the answer to the first question was counterintuitive, and he explored the possible reason in the second, where he figured the cause to be density dependence, as beavers in the steppe were more likely to survive in higher populations, and thus were dependent on living in large colonies for survival. In the forest, colony size wasn’t as important to the survival of the beavers. Pietrek found that as population density increased, the animals’ choice of landform changed: with denser numbers, the beavers were more likely to choose u-shaped valleys and plains than canyons. He noted the importance of identifying the preferred habitat of beavers, as it may allow easier detection of the presence of the invasive species.

Finally, Pietrek applied his findings toward the management of biological invasions. “One thing we can do is to build a model to predict the spread of beavers,” he said. He observed that beavers spread on average 7.8 km per year, though he also used individual-based models as well in order to track juvenile beavers. He found that young beavers tended to disperse and form new colonies, and formed another model in order to track this dispersal pattern. Juvenile beavers will first search for mates within their original colonies, only moving along if none can be found. These findings make for easier tracking of beavers across the landscape, allowing for easier management of their population growth.

Post by Devin Nieusma, Duke 2019

Seeing the Research for the Trees

By Karl Bates

On December 9, 2015

In Biology, Climate/Global Change, Environment/Sustainability, Field Research, Students

The Duke Forest is more than just a place to run the trails or harvest timber. It’s also an important living laboratory for Duke’s research community.

On Dec. 4, we joined the annual tour of research sites in the 7,000-acre forest, led by forest Director Sara Childs and Operations Manager Jenna Schreiber. Nearly two dozen of us learned about water and bugs, climate change and nanoparticle pollution.

Maggie Zimmer opened up the equipment box for her show and tell of the hydrology experiment.

At the first stop, in the Edeburn Division south of Hillsborough, Nicholas School graduate student Maggie Zimmer showed us a densely instrumented watershed for studying how a raindrop reaches a stream. A little valley of 130 hectares is studded with wells and dammed by a weir that measures every drop flowing out of the watershed. Zimmer and her thesis advisor Brian McGlynn are trying to get a handle on how a drop of water falling on a leaf or the ground eventually makes its way through several feet of soil and clay, in and around chunks of old rock, to the stream.

It’s not as simple as you think, says Zimmer, who has hand-augured 35 test wells in the study area and spent many dark, wet nights tending to her delicate equipment. For example, the rain gauge measures .01 millimeters at a time!

Across the road from the hydrology lab, we visited a global warming forest built by Jim Clark’s research team and overseen by lab manager Jordan Siminitz.

Jordan Siminitz showed us inside one of the warming forest test chambers.

There are 24 plastic enclosures for studying how temperature increases in the soil might affect the growth of young trees. The warming scenarios were produced by a network of propane-heated pipes under the soil in each enclosure. The funding that built the site and operated it for four years has stopped, but the trees are still there and the team is hopeful they can restart the experiment.

Here and in Harvard Forest, the team was looking at soil temperature increases of 3 degrees or 5 degrees Celsius. The surprising finding out of four years of data was that southern tree species seemed to be more adversely affected by the temperature increase than northern species.

“Long term research like this is really hard to get funding for,” Childs said. But without long term studies, we won’t know much about what to expect from climate change. Incidentally, NC State was conducting a parallel study of ants and warmer soils in the same experimental booths, but they’ll be shutting down this year as well.

Duke Forest Director Sara Childs checked out a pickled Southern Pine Beetle.

At the next stop, we found nattily uniformed NC Forest Service ranger Philip Ramsey standing next to an elaborate plastic contraption like 10 black funnels in a series leading down to a reservoir of antifreeze at the bottom. It’s a pheromone trap for the Southern Pine Beetle and its predator, the Clerid beetle. All is well with those bugs for now, but the devastating enemy of ash trees, the Emerald Ash Borer, is on the march and due to arrive any month now, Ramsey said, passing around pickled specimens of the bugs for our inspection.

Our last stop was an update on the nanotechnology test site called the mesocosm facility – 30 boxes filled with water, silt and plant life. They’re meant to mimic a tiny slice of a shoreline ecosystem to see how various nanoparticle materials are taken up by plant and animal life.

Steve Anderson (at right) explained the mesocosm test chambers to the tour group.

Research analyst Steve Anderson from Emily Bernhardt’s lab explained the latest experiments on what happens to all the poisonous stuff infused into anti-bacterial socks and pressure-treated lumber. The good news so far is that nanoparticies don’t seem to get taken up by ecosystems as readily as some had feared.

This isn’t really forest research per se, but where else are you going to put 30 big bunkers of mud, surrounded by an electrified raccoon fence, a super-fine frog fence and a Quonset hut enclosure for the cooler months?

Duke Forest houses 71 research projects at the moment, 16 of them started in just the last year. We’ll look forward to more fun discoveries on next year’s tour!

Follow Duke Forest on Facebook or subscribe to their updates to catch this and other tours.

Post by Karl Leif Bates

Science a waste of money? “Wastebook” misses big picture

By Robin Smith

On December 7, 2015

In Animals, Behavior/Psychology, Biology, Engineering, Faculty, Science Communication & Education

Duke biologist Sheila Patek explains the big picture behind a recent study on sparring mantis shrimp. Photograph by Roy Caldwell.

Sheep in microgravity. An experiment involving a monkey in a hamster ball on treadmill. These are among more than 100 descriptions of what Senator Jeff Flake, Republican of Arizona, deems wasteful federal spending in “Wastebook: The Farce Awakens,” released on Tuesday, Dec. 8. The latest in a series originally launched by retired Senator Tom Coburn, each “Wastebook” targets a range of federally-funded projects, many of them science-related, which the authors declare a waste of taxpayers’ money.

But what do the researchers behind these projects have to say? We asked Duke biologist Sheila Patek, whose work on fighting mantis shrimp was singled out in Flake’s latest report, to tell us her side of the story:

“What do we stand to learn from basic research on mantis shrimp? It turns out, a lot,” Patek said.

“First, mantis shrimp strike with weapons operating at the same acceleration as a bullet in the muzzle of a gun, yet they achieve high performance without explosive materials. They use a system based on muscles, springs and latches and neutralize their opponents with impact-resistant armor. This research helps us understand how animals survive when they have lethal weapons at their disposal but do not actually kill the opponent — something that could change the way we look at future defense systems,” Patek said.

“Second, these crustaceans have properties of extreme acceleration that are of great interest to military and manufacturing engineers. Mantis shrimp use a toothpick-sized hammer that can break snail shells in water that humans can only break with a larger hammer in air. Their small, lightweight hammer resists fracture over thousands of uses. Our research has already led to the development of novel engineered materials that resist impact fracture, based directly on mantis shrimp hammers,” Patek said.

“Third, mantis shrimp do something else that humans cannot: strike in water at the speed of cars on a major highway without causing cavitation, a phenomenon that occurs in systems with rapid motion (like propellers) where implosive bubbles emit heat, light and sound with energy sufficient to wear away steel. Naval engineers have been trying to solve this problem since the invention of the submarine. When we understand how mantis shrimp avoid cavitation during the rotation phase of their strikes while effectively using cavitation during their impact phase, the knowledge will undoubtedly improve the capabilities of ships, submarines, torpedoes and other machines,” Patek said.

“Research that helps us understand and apply the mechanics and evolutionary diversity of natural systems to create a better and safer society for all of us is a wise investment for this country.”

Post by Robin A. Smith, Senior Science Writer

Geography and the Web: A new frontier for data vizualization

By Anika Radiya-Dixit

On December 3, 2015

In Animals, Biology, Climate/Global Change, Computers/Technology, Data, Environment/Sustainability, Field Research, Visualization

A GIS Day earth cake made by the Collegiate Baker

You might be forgiven if you missed GIS Day at The Levine Science Research Center Nov. 18, but it was your loss. Students and faculty enjoyed a delightful geography-themed afternoon of professional panels, lightning talks, and even a geospatial research-themed cake contest.

What is GIS and why is it important?

Geographic information systems (GIS) give us the power to visualize, question, analyze, and interpret data to understand relationships, patterns, and trends in the world around us. Those who work with data and analytics have a responsibility to contribute to this change by helping us make the right decisions for our future. As noted during ESRI’s 2015 User Conference in the video below, “We have a unique ability to impact and shape the world around us. [Yet] for all of our wisdom, our vast intellectual marvels, we still choose a path of unsustainability and continue to make decisions that negatively impact the Earth and ourselves. […]We must accept our responsibility as stewards of the Earth. […] We must apply our best technology, our best thinking, our best values. Now is the time to act. Now is the time for change.”

How does GIS help?

Doreen Whitley Rogers, Geospatial Information Officer for the National Audubon Society, led a lively discussion about GIS and the World Wide Web at Duke’s GIS Day. She said GIS is essential to understand what is happening in the geographic space around us. As GIS becomes increasingly web-based, efficiently distributing the system to other people is crucial in a time when new data about the environment is being created every second.

3D map displaying the height of buildings that birds hit windows

3D map displaying the height of buildings at which birds fly into windows in Charlotte, NC

Rogers and her team are aiming to move authoritative GIS data to web for visualizations and create a centralized system with the potential to change our culture and transform the world. As the technology manager, she is working on bringing the information to people with proper security and integrity.

In order to get people to use GIS data in a generalized way, Rogers needed to implement several core capabilities to assist those integrating GIS into their workflow. These include socializing GIS as a technology to everybody, creating mobile apps to work with data in real time, and 3D maps such as this one of bird-strikes in downtown Charlotte.

Case Studies

ClimateWatch helps us predict the seasonal behaviour plants and animals.

Mobile apps connecting to the GIS platform promise a strong “return on mission” due to the vast number of people using maps on phones. By mobilizing everyone to use GIS and input data about birds and geography in their area, the platform quickly scales over millions of acres. In the Bahamas, an app allows users to take pictures to support bird protection programs.

ClimateWatch is an app that gives us a better understanding of how bird habitats are affected during temperature and rainfall variations – motivating people to speak up and act towards minimizing anthropogenic climate change. Developed by Earthwatch with the Bureau of Meteorology and The University of Melbourne, the app enables every Australian to be involved in collecting and recording data to help shape the country’s scientific response to climate change.

Virtual simulation of scenic flights as an endangered bird.

Virtual simulation of scenic flights from the perspective of an endangered bird.

Apps such as the 3-D flight map give users the vicarious thrill of cruising through nature landscapes from the view of endangered birds.

With the movements toward cleaning air and water in our communities, our planet’s birds will once again live in healthier habitats. As the Audubon Society likes to say: “Where birds thrive, people prosper.”

For more information about bird-friendly community programs, you can visit Audubon‘s site or send them a message.

Doreen Rogers after her presentation on National GIS day.

To learn more about data visualization in GIS, you can contact Doreen Whitley Rogers via email here.

Post by Anika Radiya-Dixit