How much clearcutting can nature sustain? A study determines the ecological limit for boreal forest birds

University of Jyväskylä - Jyväskylän yliopisto

LinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

The researchers identified the ecological limits below which populations of bird species associated with old-growth forests can remain stable or recover from past declines. Pictured is a goldcrest.

view more 

Credit: Aleksi Lehikoinen

A research led by the University of Jyväskylä (Finland) evaluated, for the first time, an ecological threshold for the amount of clearcutting in boreal forests. According to the results, maintaining populations of bird species specialized in old-growth forests requires either a forest rotation period of at least 90 years on average or approximately one-fifth of the forest area be excluded from clearcut management. The study provides a concrete target for the ecologically sustainable use of forests.

Clearcutting is a common forest management method worldwide, especially in boreal forests, but it is also a significant factor in the decline of forest biodiversity. Although the effects of individual harvests are relatively well understood, there has been little information to date on the long-term and cumulative effects of clearcutting at the landscape level. 

Researchers analyzed national bird monitoring program data from South and Central Finland covering the years 2006 to 2020 and the European maps of forest disturbances, depicting the accumulation of clearcut areas. 

- We noticed that the abundance of common forest species and species specialized in old-growth forests was lower in landscape where large areas of clearcutting had accumulated over 20 years, summarises Senior Researcher Rémi Duflot from the University of Jyväskylä. 

Defining a "safe operating space" for forests 

A key finding of the study was the definition of a so-called “safe operating space” for forest ecosystems. The researchers identified the ecological limits below which populations of bird species associated with old-growth forests can remain stable or recover from past declines. 

- Based on our results, to maintain forest bird populations, the proportion of clearcutting should not exceed c. 10 percent of the forest landscape over a ten-year period. This corresponds to a rotation period of approximately 90 years in forestry or to the exclusion of approximately one-fifth of forest areas from clearcutting management, explains Duflot. 

The researchers also emphasize that ecological thresholds may vary by region and among different groups of organisms. For example, lichens and mosses may be more demanding than forest birds. Therefore, the results should be supplemented by studies on other species groups before general thresholds are adopted. 

A tool for more sustainable forest management 

The results provide decision-makers, forest owners and forestry sector stakeholders with a concrete metric for planning ecologically sustainable forest management. 

- Our study offers a clear and easily communicable goal that can be used to reconcile forest use with the protection of biodiversity. The method is also suitable for use around the world, says Professor in Applied Ecology Mikko Mönkkönen from the University of Jyväskylä.  

The results were published in the Journal of Applied Ecology. 

---

Journal

Journal of Applied Ecology

DOI

10.1111/1365-2664.70552 

Method of Research

Data/statistical analysis

Subject of Research

Not applicable

Article Title

Impacts of clearcut accumulation on boreal forest bird communities: Defining a safe operating space at the landscape level

 

Using math to follow the money: a machine-learning tool to catch stablecoin laundering

ELSP

Facebook

XLinkedInWeChat
WhatsAppEmail

 

image: 

A behaviour-based AI screens USDT and USDC transfers on Ethereum, sorting wallets into three groups: sanctioned/frozen, cybercrime, and normal. Simpler tree models outperform graph-based AI when the network fragments.

view more 

Credit: Luciano Juvinski, Haochen Li, Alessio Brini / Duke University Pratt School of Engineering

Researchers at Duke University built one of the first large-scale datasets of stablecoin transfers and used it to train machine-learning models that flag suspicious cryptocurrency wallets. Published in the Blockchain Journal, the study detects illicit activity with high accuracy and, importantly, tells apart different types of criminal behavior, helping compliance teams act on real threats while sparing innocent users from frozen assets.

Almost every crime leaves a financial trail. Whether the offense is fraud, trafficking, or a cyberattack, criminals eventually need to move, hide, and cash out their proceeds. Increasingly they do it with stablecoins, digital currencies pegged to the U.S. dollar that move billions of dollars a day. In 2024, an estimated $51 billion was laundered through cryptocurrencies.

A new Duke University study addresses this problem. As privacy tools make parts of the crypto ecosystem harder to observe, centralized stablecoins such as Tether’s USDT and Circle’s USD Coin (USDC) stay visible, because their issuers must keep an auditable record to remain convertible into regular money. The team used that visibility to build a large-scale dataset of Ethereum wallet transfers and to set a baseline for spotting money laundering from on-chain behavior alone.

“Stablecoins have become the connective tissue of illicit finance, but that same visibility is what makes them auditable. We wanted to show you can catch laundering by studying behavior, not identities,” says Luciano Juvinski, lead author.

The team compared several families of artificial intelligence, from simple linear methods to deep neural networks and graph-based models. The clearest result: carefully engineered, behavior-based models known as tree ensembles beat the more complex graph approaches, which lost accuracy as the transaction network fragmented and grew harder to trace.

The models do more than raise a single alarm. They separate distinct types of illicit behavior. Wallets tied to cybercrime move funds fast and scatter them across many addresses, while sanctioned or frozen wallets leave a constrained, static footprint. That difference matters in practice. Under new rules such as the European Union’s MiCA and the U.S. GENIUS Act, compliance teams must decide which wallets to act on, and a wrong call can mean an innocent person’s money is locked.

By blocking laundering with mathematics, the researchers argue, it becomes possible to disrupt the flow of illicit funds and raise the cost of crime itself. The work offers a foundation that future systems can build on to move this kind of detection from research into everyday compliance.

The study was conducted by Luciano Juvinski, Haochen Li, and Alessio Brini at the Duke University Pratt School of Engineering, and grew out of the Duke Master in FinTech program. The dataset and code are publicly available to support further research.

The paper, “StableAML: machine learning for behavioral wallet detection in stablecoin anti-money laundering on Ethereum,” was published in the Blockchain Journal.

Juvinski L, Li H, Brini A. StableAML: machine learning for behavioral wallet detection in stablecoin anti-money laundering on Ethereum. Blockchain. 2026, https://doi.org/10.55092/blockchain20260007

---

Journal

Blockchain

DOI

10.55092/blockchain20260007 

Method of Research

Data/statistical analysis

Subject of Research

Not applicable

Article Title

StableAML: machine learning for behavioral wallet detection in stablecoin anti-money laundering on Ethereum

Article Publication Date

1-Sep-2026

 

Green and efficient maize varieties could raise global yields while reducing nitrogen losses

Researchers map genetic hotspots, real-world adoption gaps and global benefits for maize breeding aligned with Sustainable Development Goals (SDGs)

Science China Press

Facebook

XLinkedInWeChatBlueskyMessageWhatsAppEmail

Maize is central to global food, feed and industrial systems. It is also exposed to mounting pressure from climate change, resource constraints and environmental pollution linked to intensive fertilizer use. Breeding higher-yielding maize is therefore no longer sufficient on its own. The crop must also become more efficient in its use of nutrients and more resilient to environmental stress.

In a Short Communication published by Science Bulletin, researchers led by Xiangyuan Wan and Xun Wei at the University of Science and Technology Beijing, with collaborators from China Agricultural University, Zhejiang University, Wageningen University & Research and the International Maize and Wheat Improvement Center, provide a global assessment of how G&E maize varieties could contribute to SDGs.

The researchers classified 48 G&E maize traits into four groups: biotic stress resistance, abiotic stress tolerance, ideal plant morphology and architecture, and efficient nutrient use. These traits include insect resistance, drought and heat tolerance, nitrogen use efficiency and compact plant architecture suitable for high-density planting.

To locate genetic entry points for multi-trait improvement, the researchers integrated 27,516 QTNs and 3,272 QTLs into 691 QTN clusters and 386 QTL clusters. Across the four trait categories, they detected 293 common genomic regions. Among 524 previously reported genes associated with these traits, 227 were located within 98 common clusters. These 98 regions can be interpreted as priority genomic hotspots, giving breeders and geneticists a tractable set of regions for fine mapping, gene editing, multi-omics profiling and molecular design breeding.

The authors also compiled 539 maize varieties worldwide that carry one or more G&E traits. Most were developed through hybrid breeding or genetic modification. The current portfolio is dominated by technically accessible traits such as insect resistance and herbicide tolerance, while more complex traits such as nitrogen use efficiency, cold tolerance and salt tolerance remain underrepresented.

A meta-analysis of 1709 field observations from 96 studies showed that G&E maize varieties increased yield by 10.1% overall, and 12.7% after trim-and-fill adjustment. Yield effects differed by continent, breeding technology and trait type. Varieties combining insect resistance and drought tolerance showed particularly large gains in the compiled studies. The nitrogen results showed both promise and a warning. G&E varieties improved nitrogen utilization efficiency, the conversion of absorbed nitrogen into grain yield, by 16.7%. However, nitrogen uptake efficiency declined by 13% in the available dataset. It nevertheless highlights a key breeding challenge: improving yield and aboveground nitrogen use without weakening root-based nitrogen acquisition.

To estimate future global potential, the team applied random forest models to 561,359 gridded soil and climate observations. Under a full-adoption scenario for ideal G&E maize varieties, the models projected an 18.1% increase in global maize yield, equal to 145.78 Tg additional grain per year, and a 26.6% reduction in reactive nitrogen losses, equal to 1.49 Tg less reactive nitrogen per year. These projections represent upper-bound biological potential. When the modelled gains are scaled to current adoption levels in low-efficiency target regions, the near-term benchmark is about a 9% yield gain and a 13% reduction in reactive nitrogen losses.

The analysis revealed a three-stage translation gap. First, research has not yet produced commercial varieties that reliably combine three or more G&E traits. Second, many reported varieties have not reached commercial production, especially in regions where expected benefits would be high. Third, deployed varieties only achieve their full value when paired with suitable fertilization, planting density, pest control and market conditions. The study concludes that realizing the potential of G&E maize will require coordinated action across genetics, breeding, regulation, seed systems and crop management. AI-based genomic selection, gene editing, synthetic biology and multi-environment field trials could help assemble beneficial allele combinations. Policy and market interventions will also be needed so that improved varieties reach farmers in high-need regions.

---

Journal

Science Bulletin

DOI

10.1016/j.scib.2026.08.082 

Method of Research

Meta-analysis

Article Title

Green and efficient maize varieties synergize global yield and nitrogen sustainability

 

The YouRban project tests the first itinerant plant for treating composite waste

The truck developed by the Department of Mechanical Engineering of Politecnico di Milano is a mobile factory that brings on-site treatment of composite materials to major events, companies and remote areas

Politecnico di Milano

WeChatBluesky
Email

 

image: 

One of the parts of the Yourban truck

view more 

Credit: Politecnico di Milano

Bringing advanced recycling of composite materials and the ecological transition directly to where waste is generated, shortening the distances between scientific research, the industrial fabric and citizens: this is the objective of YouRban, a European project coordinated by the Department of Mechanical Engineering of Politecnico di Milano, which has just ended. The technological core of the project is a mobile and itinerant truck-based plant designed and implemented for the on-site treatment and reprocessing of end-of-life composite materials, and in particular fibreglass. It is a highly versatile material: it is used for skis, tennis rackets, some automotive components, wind turbine blades and construction materials.

The project involved seven partners from Italy, Spain, Austria and Lithuania, and was funded by the European Union within the New European Bauhaus (NEB) programme. Specifically, the truck was developed in collaboration with the start-up Fibereuse Tech and the Origoni&Steiner studio, according to a modular and scalable model: this makes it replicable and customisable based on the specific needs of industrial districts or the different types of materials to be treated.

“The real strength of YouRban was to demonstrate that technological innovation can and must leave laboratories to meet people and territories. Seeing the transformation of waste into new products live allowed citizens to experience first-hand the value of the circular economy,” explains Marcello Colledani, faculty member in the Department of Mechanical Engineering and head of the YouRban project.

With the end of the testing phase, in fact, the most promising future practical scenarios for this new technology are now opening up. Among the main applications, in addition to raising citizens’ awareness through participation and the development of local events, there is the provision of on-demand services for companies: the truck can in fact travel to different production districts to treat production waste directly on site, eliminating waste transport costs. “Another highly significant scenario is the management of major events, such as the Fiera dell’Artigianato or Design Week, true institutions of the Milan area where the high concentration of temporary installations requires rapid and virtuous solutions for treating and transforming residues,” adds Colledani. Finally, the flexibility of the vehicle makes it possible to manage logistics in remote and hard-to-reach areas, bringing recycling directly on site: it can be used in wind farms, for example, facilitating the treatment of turbine blades that have reached the end of their life.

YouRban was also created with a strong mission of participation and public awareness: the plant is in fact transparent, and visually shows citizens the concrete value of circularity thanks to windows that make it possible to see the plant in operation even though it is inside a closed container.

The YouRban project validated its technologies in the field with two important public stops in Milan and Barcelona and a third, in reduced form, in Vienna. On these occasions, the Department of Design of Politecnico di Milano was also involved: “Our role was to support the creativity of the project development of the designers and artists selected through open calls,” explains Stefano Maffei, faculty member in the Department of Design. The international designers transformed the composite materials recycled by the mobile plant into functional prototypes of urban furniture and objects, concretely demonstrating the technical and aesthetic feasibility of the circular economy.

The truck will be presented at the Engineering Festival, organised as every year by Politecnico di Milano at its Leonardo and Bovisa campuses on 19 and 20 September. It will be possible to see the truck in operation, subject to press accreditation by replying to this email, on Saturday 19 September, at the Bovisa campus, in via Raffaele Lambruschini.

---

DOI

10.3030/101135997 

The entire YouRban truck from a distance

Materials produced through the recycling process

Two girls watch the truck at work during the launch event in Milan

Credit

Politecnico di Milano

 

New study uses neural networks to reveal how experience shapes learning

University of Utah Health

Facebook

WeChat
WhatsAppEmail

 

image: 

Jim Heys, PhD.

view more 

Credit: Charlie Ehlert / University of Utah Health

A new study has used a type of machine learning called a neural network to reveal how different kinds of training can change how learning happens—both in machines, and in living brains.

“We can use these complex models to make specific predictions about the functions of the brain regions we’re interested in,” says Jack Bowler, PhD, postdoctoral fellow in neurobiology at University of Utah Health and first author of the study. “If you get at the most abstract level, it’s a fairly good analogy for how we think the brain has to work.”

As neural networks learn, they adopt patterns of activity that are strikingly similar to real-life neuron firing patterns in a part of the brain that’s important for learning, the researchers found.

What’s more, the type of training affects both the activity pattern, and the ability to solve a task, in similar ways between neural networks and living brains. For both computational and living models, starting by learning a more simple task prepares the learner to excel at complex tasks. On the other hand, poorly structured learning experiences can bias learners to make specific, predictable errors when things get more complicated.

The results are published in Nature Neuroscience.

Neural networks learn—and fail—like living creatures

Jim Heys, PhD, associate professor of neurobiology at U of U Health and senior author on the study, compares the process of scaling from simple to complex tasks to learning math.

“Everyone knows that when you’re trying to learn calculus, you first learn order of operations, and then you learn algebra, and you learn trigonometry, and you build up these concepts systematically,” Heys says. “But why that is the case?”

Researchers trained the neural networks on a relatively complex task—responding in a specific way only after presented with two stimuli of different durations (a “go trial”), and not responding when the two stimuli are the same length (“no go”). It’s the virtual equivalent of a complex learning task the team had previously developed to study learning in a mouse model, in which mice are rewarded for responding to timed patterns of smells.

For a mouse, learning to respond only to the correct patterns is a complex task that requires multiple training steps, building from simple to complex. The same is true for a neural network, the researchers found. Training a simpler task first by only giving the network “go trials” made the network respond more accurately once it progressed to the full, complex task with both “go” and “no go” trials.

In contrast, when researchers trained the networks on the complex task without providing a simpler one first—essentially, jumping straight to calculus before learning algebra—the networks tended to make repeated, predictable errors.

They often responded too early to “go” trials, jumping the gun after the first long stimulus. Importantly, this is the exact same error that living animals make when trained improperly,  which shows that the networks can be used to predict how real brains learn.

Neural networks predict patterns of real-life brain activity

The researchers then measured neuron-level brain activity in mice as they responded to trials, focusing on a specific brain region known to be involved in task learning. The patterns of neuron activity they observed looked very similar, on a large scale, to the patterns of activity the neural networks adopted during trials.

Appropriately trained neural networks, like appropriately trained mice, moved through a cyclic pattern of activity over the course of a single trial, ending the trial in about the same state they began. During “go” trials, both neural networks and mice adopted a different pattern of activity: while they still moved through a cycle, the activity swerved off its regular path during the time period that the mouse—or the network—needed to respond to the trial.

The similarities in activity and behavior between neural networks and real-life brains suggest that the networks can be especially useful as a first step to understanding human learning, Bowler says. Discoveries with computational models help generate useful, targeted questions about how we think, which can then be tested with animal models—helping scientists learn vital information while using the smallest number of research animals.

“By modeling up front, we were able to move much more quickly to useful hypotheses and reduce the number of animals we needed to test,” Bowler says. “It helps us work with animals more respectfully.”

The findings could ultimately guide development of better trainings that help people scale up their learning, the researchers say. By revealing how the brain learns when it’s healthy, the team hopes that their work could also be useful for understanding what goes wrong in diseases that affect complex thinking.

“We’re understanding fundamental principles of how the brain works,” Heys says. “So when it breaks, like in diseases like Alzheimer’s disease, which affect really high level complex cognitive functions first, we can understand how to fix it.”

 

###

This research is published in Nature Neuroscience as “Structured experience shapes strategy learning and neural dynamics in the medial entorhinal cortex.”

The work was supported by the National Institutes of Health, including the National Institute of Mental Health (DP2 MH129958-01), the National Science Foundation (IOS-2145814), the Life Sciences Research Foundation with support from the Simons Foundation – Collaboration on the Global Brain, and the University of Utah. Content is solely the responsibility of the authors and does not necessarily represent the official views of the funders.

---

Journal

Nature Neuroscience

DOI

10.1038/s41593-026-02409-7 

Method of Research

Experimental study

Subject of Research

Animals

Article Title

Structured experience shapes strategy learning and neural dynamics in the medial entorhinal cortex

Article Publication Date

3-Sep-2026