When employing remote sensing images, it is challenging to classify vegetation species and ground objects due to the abundance of wetland vegetation species and the high fragmentation of ground objects. Remote sensing images are classified primarily according to their spatial resolution, which significantly impacts the classification accuracy of vegetation species and ground objects. However, there are still some areas for improvement in the study of the effects of spatial resolution and resampling on the classification results. The study area in this paper was the core zone of the Huixian Karst National Wetland Park in Guilin, Guangxi, China. The aerial images (Am) with different spatial resolutions were obtained by utilizing the UAV platform, and resampled images (An) with different spatial resolutions were obtained by utilizing the pixel aggregation method. In order to evaluate the impact of spatial resolutions and resampling on the classification accuracy, the Am and the An were utilized for the classification of vegetation species and ground objects based on the geographic object-based image analysis (GEOBIA) method in addition to various machine learning classifiers. The results showed that: (1) In multi-scale images, both the optimal scale parameter (SP) and the processing time decreased as the spatial resolution diminished in the multi-resolution segmentation process. At the same spatial resolution, the SP of the An was greater than that of the Am. (2) In the case of the Am and the An, the appropriate feature variables were different, and the spectral and texture features in the An were more significant than those in the Am. (3) The classification results of various classifiers in the case of the Am and the An exhibited similar trends for spatial resolutions ranging from 1.2 to 5.9 cm, where the overall classification accuracy increased and then decreased in accordance with the decrease in spatial resolution. Moreover, the classification accuracy of the Am was higher than that of the An. (4) When vegetation species and ground objects were classified at different spatial scales, the classification accuracy differed between the Am and the An.

Soil conservation (SC) plays a vital role in preventing soil erosion and ensuring ecological security. While current research on SC primarily focuses on historical spatiotemporal variations, there remains a dearth of sufficient simulation research exploring future development scenarios. In this study, simulations were applied to the source of Yellow River (SYR), a representative ecologically fragile area. Satellite remote sensing and product data, including precipitation, soil, land use/cover, DEM, and SPOT/VEGETATION NDVI, were utilized. The historical and future evolutionary trends of SC in the SYR were quantitatively assessed using the Revised Universal Soil Loss Equation (RUSLE) and trend analysis method, and the geographical detector was employed to explore the forces driving spatial differentiations in SC. The results demonstrated that: (1) 2000–2020, the spatial heterogeneity of SC in the SYR was characterized by the distribution of “gradually decreasing from Southeast to Northwest”, demonstrated a trend of “increasing, decreasing, and then increasing”. (2) Under the diverse development scenarios, the trend of SC change in the SYR was predominantly rising, and the natural change scenario (NCS) > ecological conservation scenario (ECS) > economic expansion scenario (EES). (3) Slope was the most important single driver affecting the spatiotemporal differentiation of SC, and the interaction of slope with average annual precipitation, and NDVI on the spatiotemporal heterogeneity of SC had the strongest explanatory ability. The results can serve as a scientific basis for regional SC and ecological protection and construction of the SYR.

The preservation of biodiversity is crucial for maintaining ecological balance and promoting the harmonious development of mankind and nature. To formulate a biodiversity conservation plan for Guilin in China and accurately evaluate the impact of conservation measures on regional biodiversity, this study combined the InVEST model (v 3.7.0), the GeoSOS-FLUS model (v 2.3), and the landscape pattern index, analyzing the spatiotemporal changes in biodiversity from 2000 to 2020 in Guilin and simulating biodiversity under different development scenarios in 2040. The results showed the following: (1) The biodiversity index in Guilin displayed a declining trend, with the average annual value decreasing from 0.875 in 2000 to 0.870 in 2020. The area experiencing a reduction in biodiversity was primarily concentrated around the city. (2) The biodiversity level of each district and county had a large spatiotemporal difference, with the overall distribution characteristics of “higher in the northwest, southwest, and east, and lower in the northeast, southeast and central”. (3) The biodiversity hotspots were identified as priority areas for conservation; woodland and wetland were effectively protected, and the expansion of construction land and arable land was limited under the ecological protection scenario compared with the natural development scenario. (4) The annual average value of the biodiversity index of Guilin could reach 0.872 in 2040 after the adoption of ecological conservation measures, which is 0.013 higher than the natural development scenario. The results of this study can provide guidance for the construction of a national sustainable development model city and land use planning in Guilin, as well as a scientific reference for the creation of biodiversity conservation policies.

Overgrazing typically leads to grassland vegetation degradation and reduction, which in turn triggers a series of ecological problems. Therefore, it is crucial to understand the effects of different Grazing Intensities (GIs) on the Vegetation Ecosystem (VE) to achieve sustainable grazing development. This study proposes a new quantitative index, the Grazing Intensity and Vegetation Cover Harmonization Index (GVCI), based on multiple indicators such as fractional vegetation cover (FVC), net primary productivity (NPP), and GI. The GVCI was used to quantify the “Harmonization and Conflict” status between GI and the VE in 39 Prefecture-Level Cities (PLCs) of the Qinghai-Tibet Plateau (QTP) and to evaluate the sustainable development level of grazing in different regions. In addition, the Random Forest (RF) model was used to simulate the GVCI development trend of various PLCs from 2015 to 2040. The results showed the following: (1) The GVCI can effectively quantify the response relationship between GI and the VE. The overall GVCI of the QTP was in the “Harmonization” state, with the proportion of areas in the “Harmonization” state fluctuating upwards. (2) The level of economic development intuitively affects the harmonization between grazing and the VE. Gross Domestic Product (GDP) is one of the important indicators of economic development level. PLCs with higher GDP levels exhibited a strong positive correlation between the GVCI and regional GDP. (3) The simulation results indicate that an increasing number of PLCs on the QTP will shift toward a “Harmonization” state. However, some PLCs in the western regions were still in an “Overload” state, and there is a need for close monitoring of their grazing activities and VE dynamics. The GVCI proposed in this study provides a novel methodology for quantifying the complex relationship between GI and the VE. It offers important scientific support for the sustainable development of grazing in ecologically fragile areas such as the QTP. The research results can be a robust scientific basis for the government to formulate reasonable grazing plans.

Introduction: Chemoresistance and disease relapse remain the main obstacles responsible for treatment failure in leukemia. MicroRNAs (miRNAs) play essential roles in various physiological and pathological processes, including cell proliferation, differentiation, metabolism, and cancer development. The miR-106b-25 cluster consists of three miRNAs: miR-106b, miR-93 and miR-25. We have previously reported that miR-106b-25 was associated with chemoresistance by negatively regulated EP300 in breast cancer, but its role in hematological malignancies has not yet been elucidated. Here, we aim to clarify the biological role and underlying mechanisms of miR-106b-25 on drug resistance in leukemia. Methods: To see whether the miR-106b-25 was associated with the poor prognosis of AML patients, enriched LSCs (CD34 + cells) were isolated from the bone marrow of 18 newly diagnosed AML patients, the expression of miR-106b, miR-93, and miR-25 were examined, respectively. The expression levels of miR-106b, miR-93 and miR-25 were further determined in the doxorubicin-resistant leukemia cell line K562/A02 and HL60/ADR, compared with their parental cell lines. In addition, K562 cells were transduced with lentiviral vectors carrying miR-106b-25, and cell proliferation, drug resistance, colony-forming assay, apoptosis assays were performed to explore the function of miR-106b-25 overexpression in leukemia cells in vitro. To investigate the role of miR-106b-25 on tumor growth and overall survival after drug treatment, we performed xenotransplantation in nude mice using miR-106b-25 overexpressed K562 cells. To further clarify the function of each microRNA function in this cluster, K562 cells were also transduced with lentiviral vectors carrying individual miR-25, miR-93, or miR-106b separately. Cell proliferation, colony forming assay and cell apoptosis assay were also carried out subsequently. Simultaneously, RNA-sequencing was performed to reveal the underlying mechanisms of miR-106b-25 in the chemoresistance of myeloid leukemia. To experimentally confirm the direct target of the miR-106b-25 cluster in AMLs, we further performed a dual-luciferase reporter assay. Results: Upregulated miR-106b, miR-93 and miR-25 expression in enriched LSCs were significantly associated with shortened overall survival of AML patients. We also found miR-106b, miR-93 and miR-25 were significantly upregulated in drug-resistant leukemia cell lines compared with its parental cell lines. Overexpression of miR-106b-25 cluster promoted cell proliferation, led to resistance of K562 cells to doxorubicin, imatinib and ABT-737 (BCL-2 inhibitor) in liquid culture and drug-resistant colony-forming assays. Overexpression of miR-93 or miR-106b accelerated cell growth, and all the three miRNAs can promote drug-resistant colony-forming and inhibit cell apoptosis. RNA-sequencing (RNA-Seq) data revealed that multiple critical genes related to apoptotic pathways were downregulated after overexpressing miR-25, miR-93, miR-106b as well as the whole cluster, such as TP73, BAX, BAK1, Caspase-7, CDKN1A and BTG2. RT-qPCR confirmed that these genes are reduced with or without ABT-737 treatment. Luciferase assay further identified TP73 was a direct target of miR-93 and miR106b, BAK1 was a direct target of miR-25, and CASPASE-7 was a direct target of all these three miRNAs. Conclusions: In summary, we made the novel observation that miR-106-25 is associated with AML drug-resisitance and disease prognosis and identified TP73, BAK1 caspase-7 as a novel direct target of this cluster. Further studies revealed that the biological effects of miR-106b-25 cluster on leukemic cell proliferation, chemoresistance and apoptosis were mediated through regulation of apoptotic pathway. These findings indicate a promising diagnostic biomarker and a potential target therapeutic strategy for AML patients. Disclosures: No relevant conflicts of interest to declare.

Accurate tree species classification using high-resolution unmanned aerial vehicle (UAV) images is crucial for forest carbon cycle research, biodiversity conservation, and sustainable management. However, challenges persist due to high interspecies feature similarity, complex canopy boundaries, and computational demands. To address these, we propose an enhanced DeepLabV3+ model integrating Object-Based Image Analysis (OBIA) and a lightweight attention mechanism. First, an OBIA-based multiscale segmentation algorithm optimizes object boundaries. Key discriminative features, including spectral, positional, and vegetation indices, are then identified using Recursive Feature Elimination with Cross-Validation (RFECV). High-precision training labels are efficiently constructed by combining Random Forest classification with visual interpretation (RFVI). The DeepLabV3+ model is augmented with a lightweight attention module to focus on critical regions while significantly reducing model parameters. Evaluations demonstrate that the improved DeepLabV3+ model achieved overall accuracy (OA) of 94.91% and Kappa coefficient (Kappa) of 92.89%, representing improvements of 2.91% and 4.11% over the original DeepLabV3+ model, while reducing parameters to 5.91 M (78.35% reduction). It significantly outperformed U-Net, PSPNet, and the original DeepLabV3+. This study provides a high-accuracy yet lightweight solution for automated tree species mapping, offering vital technical support for forest carbon sink monitoring and ecological management.