A large-scale study with data from Brazilian Cerrado streams found that about half of the field sampling effort preserved the ability of most habitat and biological indices to distinguish stream disturbance levels, while greatly reducing time and costs. Macroinvertebrate indices remained robust under reduced sampling, and only fish richness was more sensitive to lower effort. The findings suggest that faster, more cost-effective stream assessments could support biomonitoring programs in tropical regions with limited funding.

Researchers developed an automated cold plasma and bubble system for dual modality function of treating food-industry wastewater and transforming into nutrient-rich liquid fertilizer for hydroponic farming. This automated electrified flow system runs on a low energy, reduced wastewater organic load while simultaneously improving plant growth, offering a scalable and promising route toward sustainable agriculture and water reuse.

Researchers have developed "MetaRL," a meta reinforcement learning model that provides near-optimal, individualized investment strategies for goals-based wealth management, i.e., the portfolio composition and the decision to take a goal or not. Pre-trained on thousands of scenarios, MetaRL achieves 97.8% of optimal utility while operating over 100 times faster than traditional methods. The model delivers immediate "zero-shot" solutions in milliseconds and remains robust against changing market regimes and stochastic inflation, enabling high-throughput, sophisticated financial planning for a broad range of investors.

A widely used method for measuring how well streams absorb excess nutrients has a hidden flaw: it systematically overestimates uptake length under high-nutrient conditions. Researchers at Duke Kunshan University have derived a corrected zero-order analytical approach that better captures stream nutrient processing when nutrients are abundant, improving the accuracy of tools used to assess river health and guide restoration decisions.

The researchers coupled the vehicle-track dynamics model with the UIUC model to provide a comprehensive description of vehicle-induced ballast settlement — encompassing both its full evolution and cumulative dynamic effects. Special emphasis was placed on the influence of sleeper voids.

Researchers develop an interpretable hybrid feature selection model to overcome the "curse of dimensionality"in genomic selection, pinpointing key genetic markers for global food security.

Human cognition relies on the seamless integration of multiple senses, allowing the brain to associate, infer, and even imagine across modalities. Replicating this capability in artificial systems has long remained a challenge, particularly under strict energy constraints. This study presents a bioinspired multisensory framework that integrates vision, touch, hearing, smell, and taste within a self-powered architecture. By enabling cross-modal association and adaptive reconfiguration, the system allows one sensory input—such as touch or sound—to trigger corresponding representations in other sensory domains. Beyond conventional recognition, the framework demonstrates higher-level cognitive functions, including inference and generative pattern creation. These advances point toward a new generation of intelligent machines capable of human-like perception and cognition.

Achieving both high efficiency and high selectivity remains a central challenge in catalytic hydrogenation reactions, as rapid reactant activation often leads to overly strong binding of intermediates and unwanted over-reduction. A new photocatalytic strategy now demonstrates how this long-standing trade-off can be overcome. By harnessing nonequilibrium charge carriers generated under visible light, the system promotes efficient hydrogen dissociation while steering the reaction toward selective semihydrogenation. Active hydrogen species migrate away from their formation sites to neighboring catalytic surfaces, where product formation and desorption are energetically favored. This spatial separation of reaction steps enables near-complete conversion under ambient conditions while maintaining excellent selectivity, offering a powerful new design principle for advanced catalytic systems.