Research on generative AI in education often asks whether these tools improve learning outcomes. Collaborative design raises a prior question: how should teaching organize AI-supported work so that disciplinary knowledge, creativity, collaboration, and human judgment come together? Integrative learning names that aim: connecting disciplinary knowledge with situated experience and synthesizing insights across cases. Integration matters in teacher education, where preservice teachers must apply principles to ambiguous professional situations. We examined an AI-assisted interactive-narrative (AI-IN) assignment in which preservice teachers designed branching professional-ethics games through a whole-part-whole sequence: course-wide orientation, collaborative development of a focal case with AI-supported possibility generation, and cross-case reintegration. We used self-reports from a 480-person primary cohort and an independent 257-person same-course cohort to ask how students appraised the design, whether the learning ratings shared residual covariance in a one-factor model, and which conditions accompanied appraisal. Allowing residual covariance between the two perceived-learning ratings improved the one-factor model in both cohorts (scaled Delta chi & sup2;(1) = 11.99 and 14.20, respectively; both p < .001); the localized association was stable across sensitivity analyses. This localized covariance may reflect perceived-learning content or shared-method influences, including wording and placement; it establishes neither learning gains nor a distinct latent domain. In the primary cohort, favorable appraisal accompanied clearer roles, efficient collaboration, greater contribution, better challenge-engagement balance, and positive AI attitudes. Students reported challenges in translating ethics into plots, selecting platforms, and managing workload. AI-IN offers a testable design proposition for integrative learning: AI supports possibility generation while students retain responsibility for selection, verification, ethical judgment, and synthesis.
While microalgae-based wastewater treatment has been widely studied, this work presents the first demonstration of using self-flocculating Spirulina platensis CMB-02 for treating rare earth element (REE) wastewater-a previously unexplored application where conventional microalgae systems face biomass recycling challenges. The airlift photobioreactor (AL-PBR) exhibited superior performance, attaining a biomass yield of 0.31 g/L and achieving the highest removal efficiencies for total ammonia nitrogen (TAN) and total inorganic nitrogen (TIN) at 75.12% and 72.06%, respectively. Based on these findings, a 50 L AL-PBR was developed for continuous cultivation of S. platensis CMB-02 to treat REEs wastewater. Over a 33-day operation, the TAN and TIN concentrations were reduced from approximately 60 mg/L in the influent to stable effluent levels of 3.29-12.30 mg/L and 5.46-15.20 mg/L, respectively, meeting discharge standards. Additionally, the biomass of S. platensis CMB-02 was effectively maintained within the range of 0.51-0.57 g/L. Furthermore, transcriptomic analysis revealed that NaHCO3 supplementation enhances carbon metabolism and ammonia nitrogen assimilation pathways in S. platensis CMB-02 (associated genes showed 1.41- to 3.35-fold upregulation), promoting rapid cell growth and improving ammonia nitrogen uptake from REEs wastewater. This study establishes a first-of-its-kind integrated solution combining strain selection, reactor engineering, and metabolic enhancement for REE wastewater remediation.
Plants allocate photosynthetically fixed carbon (C) between metabolic needs, biomass production (BP), and non-structural carbon (NSC), a buffer between carbon supply and demand. While conceptual representation of alternative regulatory mechanisms of NSC has advanced significantly in recent decades, quantitative assessments of NSC's contribution to the ecosystem C budget have not been possible. We estimated C flux to or from the NSC pool (FNSC) as the difference between gross primary productivity (GPP), plant respiration (Rp), and BP in a shortleaf pine forest. Annual GPP remained relatively stable, but FNSC varied during the 3-year study period: 77-350 g C m⁻² yr⁻¹ in dry years (2022-2023) and 41 g C m⁻² yr⁻¹ in the wet year (2024), representing 2-18% of annual GPP. In contrast, C allocation to BP was ∼43% higher in the wet year. Flux-based FNSC (37 ± 84 g C m⁻² yr⁻¹) closely matched FNSC estimates from monthly tissue NSC content (65 ± 77 g C m⁻² yr⁻¹) during 2024 (Feb-Dec). FNSC was inversely related to BP, Biomass Production Efficiency, and plant water status. These findings are consistent with the surplus carbon hypothesis, which posits sink strength-driven C allocation, and can serve as the conceptual foundation for next-generation carbon allocation algorithms.
Using high-resolution satellite observations and numerical simulations, this study investigates how orographically induced disturbances stimulate boundary-layer horizontal convective rolls (HCRs) and precondition the convective environment prior to afternoon convection over Hainan Island in southern China. As indicated by organized cloud streets, the island's HCR characteristics undergo a pronounced transition as the prevailing lowlevel flow shifts during the summer monsoon onset. When low-level monsoonal southwesterlies dominate, elongated ridge-dependent HCRs extending over 100 km become prevalent and are closely aligned with the major mountain ridges. This pattern contrasts sharply with the densely packed but randomly distributed typical HCRs observed when low-level southeasterlies prevail before monsoon onset. Numerical simulations show that elongated stripe-like convergence zones develop extending from the ridges as the lower troposphere destabilizes in response to late-morning surface heating. The associated updrafts drive the formation of broader and deeper roll circulations than the typical densely distributed HCRs. The upward branches of these ridge-dependent HCRs promote horizontal moisture pooling and vertical moistening, whereas the downward branches produce the opposite effect, extending from the mountain ridges toward the northeastern part of the island where afternoon convection frequently occurs. In addition to downstream surface heating, warm air advected from the ridge regions further strengthens low-level ascent and horizontal convergence, and they also help form a locally neutral layer, thereby reducing convective inhibition beneath the elongated cloud streets. These results suggest that, when the background low-level flow is favorably aligned with local orographic features, orographicallyinduced mechanical disturbances can organize ridge-dependent HCRs and substantially reshape the convective environment.
Magnetic mineral inclusions (MMI) are physically protected by host minerals and can preserve stable magnetic signals, making them important carriers for paleomagnetic reconstruction and provenance tracing. However, their occurrence and environmental significance within in situ weathering crust soils remain poorly constrained. Here, we investigated nine weathering crust soil profiles developed on basalt, granite, and sandstone along the southeastern coast of China. Acid leaching, targeted chemical treatments, rock-magnetic measurements, and microscopic observations were combined to characterize residual magnetic signals under contrasting lithological, climatic, and pedogenic settings. Following previous MMI studies, residual magnetic signals after acid treatment were used as an operational indicator of MMI-related magnetic components. Microscopic observations provided direct evidence for the presence of mineral-hosted magnetic inclusions. Residual magnetic signals after acid treatment generally followed the lithological order basalt > sandstone > granite, although substantial variability occurred within the same lithology, suggesting that parent lithology is one important factor influencing MMI-related magnetic signals. These signals also varied under prolonged chemical weathering and pedogenesis in warm and humid tropical–subtropical environments, although no simple monotonic relationships were observed with temperature, precipitation, or chemical weathering intensity. Several profiles showed distinct double-peak patterns of residual magnetic signals after acid treatment in the A (topsoil) and B (illuvial) horizons, suggesting that particle translocation, sorting, and selective preservation may influence their vertical distribution. Validation experiments further showed that clay aggregation and organic-matter cementation can shield non-intracrystalline magnetic particles and generate “pseudo-inclusion” residuals. Thus, residual magnetic signals after acid treatment provide provides a useful operational proxy for MMI-related magnetic components, but cannot be unconditionally equated with genuine intracrystalline MMI under all soil conditions. These findings improve our understanding of MMI preservation and modification during long-term weathering and help refine the interpretation of MMI-based provenance and environmental signals.
