Convolutional neural networks (CNNs) have demonstrated their efficacy in remote sensing applications for mountain forest classification. However, two-dimensional convolutional neural networks (2D CNNs) require a significant manual involvement in the visual interpretation to obtain continuous polygon label data. To reduce the errors associated with manual visual interpretation and enhance classification efficiency, it is imperative to explore alternative approaches. In this research, we introduce a novel one-dimensional convolutional neural network (1D CNN) methodology that directly leverages field investigation data as labels for classifying mountain forest types based on multiple remote sensing data sources. The hyperparameters were optimised using an orthogonal table, and the model’s performance was evaluated on Mount Emei of Sichuan Province. Comparative assessments with traditional classification methods, namely, a random forest (RF) and a support vector machine (SVM), revealed superior results obtained by the proposed 1D CNN. Forest type classification using the 1D CNN achieved an impressive overall accuracy (OA) of 97.41% and a kappa coefficient (Kappa) of 0.9673, outperforming the U-Net (OA: 94.45%, Kappa: 0.9239), RF (OA: 88.99%, Kappa: 0.8488), and SVM (OA: 88.79%, Kappa: 0.8476). Moreover, the 1D CNN model was retrained using limited field investigation data from Mount Wawu in Sichuan Province and successfully classified forest types in that region, thereby demonstrating its spatial-scale transferability with an OA of 90.86% and a Kappa of 0.8879. These findings underscore the effectiveness of the proposed 1D CNN in utilising multiple remote sensing data sources for accurate mountain forest type classification. In summary, the introduced 1D CNN presents a novel, efficient, and reliable method for mountain forest type classification, offering substantial contributions to the field.

Grasslands play a vital role in the global ecosystem. Efficient and reproducible methods for estimating the grassland aboveground biomass (AGB) are crucial for understanding grassland growth, promoting sustainable development, and assessing the carbon cycle. Currently, the available methods are limited by their computational inefficiency, model transfer, and sampling scale. Therefore, in this study, the estimation of grassland AGB over a large area was achieved by coupling the PROSAIL model with the support vector machine regression (SVR) method. The ill-posed inverse problem of the PROSAIL model was mitigated through kernel-based regularization using the SVR model. The Zoigê Plateau was used as the case study area, and the results demonstrated that the estimated biomass accurately reproduced the reference AGB map generated by zooming in on on-site measurements (R² = 0.64, RMSE = 43.52 g/m², RRMSE = 15.13%). The estimated AGB map also maintained a high fitting accuracy with field sampling data (R² = 0.69, RMSE = 44.07 g/m², RRMSE = 14.21%). Further, the generated time-series profiles of grass AGB for 2022 were consistent with the trends in local grass growth dynamics. The proposed method combines the advantages of the PROSAIL model and the regression algorithm, reduces the dependence on field sampling data, improves the universality and repeatability of grassland AGB estimation, and provides an efficient approach for grassland ecosystem construction and planning.

A new orchid species, Hemipilia avisoides, is described from Songpan County and Maoxian County, Sichuan Province, China. Morphologically, H. avisoides is most similar to H. hui, but can be distinguished by the combination of its involute middle lip lobe that is smaller than the lateral lobes, floral bracts that are 5 mm long and are always shorter than the ovary, a leaf that is appressed to the substrate and is adaxially green with white lines along 7–9 principal veins and the subterranean stem with a solitary sheath at its base. The floral morphology of H. avisoides is presented by utilising in vivo micro-CT scanning and 3D visualisation.

Developing an effective vegetation classification method for mountain–plain transition zones is critical for understanding ecological patterns, evaluating ecosystem services, and guiding conservation efforts. Existing methods perform well in mountainous and plain areas but lack verification in mountain–plain transition zones. This study utilized terrain data and Sentinel-1 and Sentinel-2 imagery to extract topographic, spectral, texture, and SAR features as well as the vegetation index. By combining feature sets and applying feature elimination algorithms, the classification performance of one-dimensional convolutional neural networks (1D-CNNs), Random Forest (RF), and Multilayer Perceptron (MLP) was evaluated to determine the optimal feature combinations and methods. The results show the following: (1) multi-feature combinations, especially spectral and topographic features, significantly improved classification accuracy; (2) Recursive Feature Elimination based on Random Forest (RF-RFE) outperformed ReliefF in feature selection, identifying more representative features; (3) all three algorithms performed well, with consistent spatial results. The MLP algorithm achieved the best overall accuracy (OA: 81.65%, Kappa: 77.75%), demonstrating robustness and lower dependence on feature quantity. This study presents an efficient and robust vegetation classification workflow, verifies its applicability in mountain–plain transition zones, and provides valuable insights for small-region vegetation classification under similar topographic conditions globally.

The biogeography research of orchids through species distribution models (SDMs), a vital tool in the biogeography field, is critical to understanding the fundamental geographic distribution patterns and identifying conservation priorities. The correspondence between species occurrence and environmental information is crucial to the model's performance. However, ecological preferences unique to different orchid species, such as their life forms, are often overlooked during the modeling process. This oversight can introduce bias and increase model uncertainty. Additionally, human activities, as an important potential predictor, have not been quantified in any orchid SDMs. Taking the Hengduan Mountains as an example, we preprocessed all orchid species' occurrences based on physiological characteristics. Choosing five spatial factors related to human activities to quantify the interference and enter into models as HI factor. Using different modeling methods (GLM, MaxEnt, and RF) and evaluation indices (AUC, TSS, and Kappa), diverse modeling strategies have been constructed in the study. A double‐ranking method has been adopted to select the critical orchid distribution regions. The results showed that classification models based on physiological characteristics significantly improved the model's accuracy while adding the HI factor had the same effect but the absence of enough significance. Suitability maps indicated that highly heterogeneous mountainous areas were vital for the distribution of orchids in the Hengduan Mountains. Different distribution patterns and critical regions existed between various orchid life forms geographically – terrestrial orchids were dominant in the mountain, and mycoherterophical orchids were primarily located in the north, more influenced by vegetation and temperature. Critical regions of epiphytic orchids were in the south due to a greater dependence on precipitation and temperature. These studies are informative for understanding the orchids' geographic distribution patterns in the Hengduan Mountains, promoting conservation and providing references for similar research beyond orchids.

Forests are widely distributed in terrestrial ecosystems, covering about one-third of the global land area. They play a key role in sequestering carbon, releasing oxygen, mitigating climate change, and maintaining ecosystem balance. The ecology of the Tibetan Plateau is very fragile, but the impact of environmental change on regional forest ecosystems is not yet clear. Located in the Eastern Tibetan Plateau, the Shaluli Mountain has the richest biodiversity and the widest distribution of forests on the Tibetan Plateau. Assessing the dynamics of forest change is the basis for correctly formulating forest management measures, and is important for regional biodiversity conservation. However, traditional field surveys have the shortcomings of high cost, being time-consuming, and having poor regional coverage in forest dynamics monitoring. Remote sensing methods can make up for these shortcomings. Therefore, in this study, satellite remote sensing images were used to extract forest information from 2000 to 2020 in Shaluli Mountain, and the main drivers of forest change were analyzed with full consideration of the Spatially Stratified Heterogeneity (SSH) of environmental factors. The results found that the forest area increased from 23,144.20 km² in 2000 to 28,429.53 km² in 2020, and the average Percentage of Forest Cover (PFC) increased from 19.76% to 21.67%, with significant improvement in forest growth. The annual minimum temperature (TMN), altitude, annual maximum temperature (TMX), and annual precipitation (PRE) were the main driving factors of forest change, with an average driving power (q-value) of 0.4877, 0.2706, 0.2342, and 0.2244, and TMN was the primary limiting factor for forest growth. In addition, the driving power of all environmental factors on forest change increased from 2000 to 2020. The results of this study can provide a basis for the development of forest management strategies, and provide reference materials for regional biodiversity conservation.