The authors applied semi‑supervised deep‑learning to super‑resolution images of modern and fossil grass pollen, training convolutional neural networks to extract abstract morphological features. These features were used to quantify past grass community diversity and C3:C4 ratios in a 25,000‑year lake‑sediment record, revealing a marked diversity loss during the last glacial and a gradual decline of C4 grasses in the Holocene.

The authors introduce AdaPoinTr, a geometry-aware transformer that predicts the alpha‑shape of coniferous tree crowns from incomplete terrestrial or mobile laser‑scanning point clouds, focusing on crown reconstruction rather than full tree completion. Trained on synthetically generated partial crowns, the model consistently improves crown shape similarity and reduces height estimation bias across three diverse forest datasets, providing a cost‑effective solution for enhanced 3D forest structural monitoring.

The study introduces a hybrid modeling framework that integrates a logistic ordinary differential equation with a Long Short-Term Memory neural network to form a Physics-Informed Neural Network (PINN) for predicting wheat plant height. Using only time and temperature as inputs, the PINN outperformed other longitudinal growth models, achieving the lowest average RMSE and reduced variability across multiple random initializations. The results suggest that embedding biological growth constraints within data‑driven models can substantially improve prediction accuracy for plant traits.

The study employed ultra large‑scale 2D clinostats to grow tomato (Solanum lycopersicum) plants beyond the seedling stage under simulated microgravity and upright control conditions across five sequential trials. Simulated microgravity consistently affected plant growth, but the magnitude and direction of the response varied among trials, with temperature identified as a significant co‑variant; moderate heat stress surprisingly enhanced growth under simulated microgravity. These results highlight the utility of large‑scale clinostats for dissecting interactions between environmental factors and simulated microgravity in plant development.

The study cloned the resistance genes Rmo2 and Rwt7 from barley and wheat, revealing them as orthologous tandem kinase proteins (TKPs) with an N‑terminal heavy metal‑associated (HMA) domain. Domain‑swapping experiments indicated that the HMA domain dictates effector specificity, supporting a model of TKP diversification into paralogs and orthologs that recognize distinct pathogen effectors.

The authors used a bottom‑up thermodynamic modelling framework to investigate how plants decode calcium signals, starting from Ca2+ binding to EF‑hand proteins and extending to higher‑order decoding modules. They identified six universal Ca2+-decoding modules that can explain variations in calcium sensitivity among kinases and provide a theoretical basis for interpreting calcium signal amplitude and frequency in plant cells.

Mutations in the plastid division gene PARC6 and the granule initiation gene BGC1 were combined to generate wheat plants with dramatically enlarged A-type starch granules, some exceeding 50 µm, without affecting plant growth, grain size, or overall starch content. The parc6 bgc1 double mutant was evaluated in both glasshouse and field trials, and the giant granules displayed altered viscosity and pasting temperature, offering novel functional properties for food and industrial applications.

The researchers performed a large‑scale field trial with 105 wheat (Triticum aestivum) genotypes inoculated by Fusarium culmorum, combining quantitative deoxynivalenol (DON) profiling and untargeted metabolomics to uncover molecular signatures of infection. Sesquiterpene‑derived metabolites tracked toxin accumulation, whereas glycosylated diterpene conjugates were enriched in low‑DON samples, indicating a potential defensive metabolic pathway.

The study assessed how well common deep learning models (ResNet, EfficientNet, Inception, MobileNet) generalize across different tomato pest and disease image datasets. While models performed well on the dataset they were trained on, they suffered substantial accuracy drops when applied to other datasets, indicating that architectural changes alone cannot overcome dataset variability. The results highlight the necessity for more diverse, representative training data to improve real-world deployment of PPD diagnostic tools.

The study investigates the wheat Pm3 NLR allelic series, revealing that near-identical Pm3d and Pm3e alleles confer broad-spectrum resistance by recognizing multiple, structurally diverse powdery mildew effectors. Using chimeric NLR constructs, the authors pinpoint specificity-determining polymorphisms and demonstrate that engineered combinations of Pm3d and Pm3e further expand effector recognition, showcasing the potential for durable wheat protection through NLR engineering.