Bletilla striata, a perennial orchid of both medicinal and ecological value, exhibits remarkable adaptability to bicarbonate-rich karst environments. To elucidate its physiological and electrophysiological responses to bicarbonate stress, seedlings were cultivated for 45 days under NaHCO₃ concentrations of 0, 5, 10, and 15 mM (n = 4), with the nutrient solution renewed daily. At 5 mM, biomass, chlorophyll content, electrophysiological traits, nutrient transport, metabolic indices, and conductance–resistance parameters did not differ significantly from controls, while intracellular water-use efficiency exhibited only a minor, non-significant increase—indicating stable physiological performance under low bicarbonate conditions. By contrast, higher concentrations (≥10 mM), particularly 15 mM, markedly reduced intracellular water-holding capacity (−35.90%), nutrient translocation capacity (−22.26%), and metabolic activity (−50.00%), alongside electrophysiological signatures of diminished capacitance (−48.69%) and elevated resistance (+147.61%), consistent with membrane injury and impaired ion transport. Although xylem pathways dominated HCO₃⁻ transport, the phloem—despite greater sensitivity—showed an increased relative contribution under stress, supporting partial compensatory allocation. Metabolically, severe stress induced a shift toward a “low-metabolism, high-efficiency” strategy, prioritizing water conservation over carbon assimilation. Collectively, Bletilla striata adopts a dual strategy: maintaining functional stability (and modest enhancement) under environmentally relevant bicarbonate concentrations, while shifting to conservative resource-use under excessive stress. These adaptive mechanisms highlight B. striata’s potential for ecological restoration and sustainable cultivation in bicarbonate-rich karst environments.

Soil moisture sensors deployed for long-term monitoring often suffer from prolonged data gaps caused by battery depletion, communication dropouts, or hardware failures. When such gaps overlap with irrigation events, key transient phases are obscured and become difficult for conventional imputers to recover. This study proposes HyDSoil, a hybrid diffusion-based imputation model tailored for event-centered block missingness in multichannel soil moisture time series. HyDSoil is first pretrained on a physically interpretable synthetic generator that mimics the baseline-rise-decay response to irrigation and then fine-tuned on field observations from the Baltimore Ecosystem Study dataset. During reverse diffusion, a mask-guided correction keeps observed values fixed while iteratively denoising missing regions. The denoising backbone integrates one-dimensional convolutions, gated recurrent units, and Transformer components to capture high-frequency event spikes, mid-range temporal dynamics, and long-range cross-depth dependencies, respectively. Experiments on both synthetic and real datasets show that HyDSoil reconstructs irrigation-driven peaks with higher fidelity and achieves consistent improvements over strong baselines in global metrics (MAE and DTW) as well as event-focused metrics (PTE and PAE). Ablation studies further verify the complementary contributions of the convolutional, recurrent, and attention branches, and confirm the benefit of synthetic pretraining for long-duration gaps. Overall, HyDSoil enables more reliable continuous soil moisture monitoring and supports precision irrigation analytics.

Plant electrophysiology is undergoing a profound paradigm shift from traditional phenomenological observation to systemic signal decoding, with mature methodologies from computational neuroscience and brain-computer interface technologies providing critical theoretical and engineering support for this interdisciplinary evolution. This review first systematically summarizes the evolution of flexible wearable electrodes and ultra-high impedance amplification hardware systems tailored to the ultra-slow signal dynamics and continuous morphological growth characteristics of plants. Second, we discuss the application pathways of introducing standardized sequential evoked paradigms from neuroscience—such as steady-state visual evoked potentials and event-related potentials—into the plant domain. This aims to replace traditional destructive stimuli with non-invasive, reproducible rhythmic stimulation to acquire data with high signal-to-noise ratios. In the dimension of data analysis, we explore modeling strategies that incorporate physics-informed neural networks and multi-modal heterogeneous sensor fusion technologies under the constraint of sample scarcity, aiming to resolve the equifinality problem inherent in single-modality electrical signal decoding. Building upon this decoding foundation, this paper proposes the construction of a bidirectional Plant-Computer Interface architecture, exploring the engineering feasibility of utilizing the plant itself as an active sensory node to directly drive closed-loop regulation within agricultural environments. Establishing cross-species standardized open-source datasets and unified hardware/software testing benchmarks will be the core driving force in overcoming current data fragmentation. Ultimately, the deep integration of multidisciplinary approaches will lay a rigorous scientific foundation for precision agricultural resource management and the development of next-generation bio-inspired intelligent hardware.

With the development of data mining technology, the analysis of event-related potential (ERP) data has evolved from statistical analysis of time-domain features to data-driven techniques based on supervised and unsupervised learning. However, there are still many challenges in understanding the relationship between ERP components and the representation of familiar and unfamiliar faces. To address this, this paper proposes a model based on Dynamic Multi-Scale Convolution for group recognition of familiar and unfamiliar faces. This approach uses generated weight masks for cross-subject familiar/unfamiliar face recognition using a multi-scale model. The model employs a variable-length filter generator to dynamically determine the optimal filter length for time-series samples, thereby capturing features at different time scales. Comparative experiments are conducted to evaluate the model’s performance against SOTA models. The results demonstrate that our model achieves impressive outcomes, with a balanced accuracy rate of 93.20% and an F1 score of 88.54%, outperforming the methods used for comparison. The ERP data extracted from different time regions in the model can also provide data-driven technical support for research based on the representation of different ERP components.