Wei Wang

Taiwan is characterized by high mountains density, with over 200 peaks exceeding 3000 m in elevation. The alpine treeline ecotone (ATE) is a transitional zone between different vegetation types. The species distribution, range variations, and movement patterns of vegetation within the ATE are crucial indicators for assessing the impact of climate change and warming on alpine ecosystems. Therefore, this study focuses on the Xue Mountain glacial cirques in Taiwan (approximately 400 ha) and utilizes WorldView-2 satellite images from 2012 and 2021 to compute various vegetation indices and texture features (GLCM). By integrating these features with Random Forest (RF) and U-Net models, we developed a classification map of the ATE in Xue Mountain. We analyzed changes in bare land, forest, krummholz, and shadows within the ATE from 2012 to 2021. The results indicate that the classification accuracy reached an overall accuracy (OA) of 0.838 when incorporating raw spectral bands along with vegetation indices and texture features (GLCM) (77 features in total). Feature importance ranking and selection reduced training time by 14.3 % while ensuring alignment between field survey treeline positions and classification results. From 2012 to 2021, tree cover density increased, with the total forest area expanding by approximately 10.09 ha. The upper limit of forest distribution shifted upslope by 32.00 ± 4.00 m, with the most significant area changes occurring between 3500 and 3600 m, while the 3700 to 3800 m range remained relatively stable. This study integrates remote sensing imagery with deep learning classification methods to establish a large-scale ATE classification map. The findings provide a valuable reference for the sustainable management of alpine ecosystems in the Xue Mountain glacial cirques in Taiwan.

Taiwan has the highest density of high mountains globally, with over 200 peaks exceeding 3,000 meters in elevation. The Alpine Treeline Ecotone (ATE) is a transitional zone between different vegetation types. The species distribution, range variations, and movement patterns of vegetation within the ATE are crucial indicators for assessing the impact of climate change and warming on alpine ecosystems. Therefore, this study focuses on the Xue Mountain glacial cirques in Taiwan (approximately 4 km²) and utilizes WorldView-2 satellite images from 2012 and 2021 to compute various vegetation indices and texture features (GLCM). By integrating these features with the Random Forest (RF) and U-Net models, we developed a classification map of the alpine treeline ecotone (ATE) in Xue Mountain. We analyzed changes in bare land, forest, krummholz, and shadows within the ATE from 2012 to 2021. The results indicate that the classification accuracy reached an overall accuracy (OA) of 0.838 when incorporating raw spectral bands along with vegetation indices and texture features (GLCM) (77 features in total). Feature importance ranking and selection reduced training time by 14.3 % while ensuring alignment between field survey treeline positions and classification results. From 2012 to 2021, tree cover density increased, with the total forest area expanding by approximately 0.101 km². The elevation of tree distribution rose by 14 m, with the most significant area changes occurring between 3,500 and 3,600 m, while the 3,700 to 3,800 m range remained relatively stable. This study integrates remote sensing imagery with deep learning classification methods to establish a large-scale alpine treeline ecotone (ATE) classification map. The findings provide a valuable reference for the sustainable management of alpine ecosystems in the Xue Mountain glacial cirques in Taiwan.

Taiwan has the highest density of high mountains globally, with over 200 peaks exceeding 3,000 meters in elevation. The Alpine Treeline Ecotone (ATE) is a transitional zone between different vegetation types. The species distribution, range variations, and movement patterns of vegetation within the ATE are crucial indicators for assessing the impact of climate change and warming on alpine ecosystems. Therefore, this study focuses on the Xue Mountain glacial cirques in Taiwan (approximately 4 km²) and utilizes WorldView-2 satellite images from 2012 and 2021 to compute various vegetation indices and texture features (GLCM). By integrating these features with the Random Forest (RF) and U-Net models, we developed a classification map of the alpine treeline ecotone (ATE) in Xue Mountain. We analyzed changes in bare land, forest, krummholz, and shadows within the ATE from 2012 to 2021. The results indicate that the classification accuracy reached an overall accuracy (OA) of 0.838 when incorporating raw spectral bands along with vegetation indices and texture features (GLCM) (77 features in total). Feature importance ranking and selection reduced training time by 14.3 % while ensuring alignment between field survey treeline positions and classification results. From 2012 to 2021, tree cover density increased, with the total forest area expanding by approximately 0.101 km². The elevation of tree distribution rose by 14 m, with the most significant area changes occurring between 3,500 and 3,600 m, while the 3,700 to 3,800 m range remained relatively stable. This study integrates remote sensing imagery with deep learning classification methods to establish a large-scale alpine treeline ecotone (ATE) classification map. The findings provide a valuable reference for the sustainable management of alpine ecosystems in the Xue Mountain glacial cirques in Taiwan.

Taiwan has the highest density of high mountains globally, with over 200 peaks exceeding 3,000 meters in elevation. The Alpine Treeline Ecotone (ATE) is a transitional zone between different vegetation types. The species distribution, range variations, and movement patterns of vegetation within the ATE are crucial indicators for assessing the impact of climate change and warming on alpine ecosystems. Therefore, this study focuses on the Xue Mountain glacial cirques in Taiwan (approximately 4 km²) and utilizes WorldView-2 satellite images from 2012 and 2021 to compute various vegetation indices and texture features (GLCM). By integrating these features with the Random Forest (RF) and U-Net models, we developed a classification map of the alpine treeline ecotone (ATE) in Xue Mountain. We analyzed changes in bare land, forest, krummholz, and shadows within the ATE from 2012 to 2021. The results indicate that the classification accuracy reached an overall accuracy (OA) of 0.838 when incorporating raw spectral bands along with vegetation indices and texture features (GLCM) (77 features in total). Feature importance ranking and selection reduced training time by 14.3 % while ensuring alignment between field survey treeline positions and classification results. From 2012 to 2021, tree cover density increased, with the total forest area expanding by approximately 0.101 km². The elevation of tree distribution rose by 14 m, with the most significant area changes occurring between 3,500 and 3,600 m, while the 3,700 to 3,800 m range remained relatively stable. This study integrates remote sensing imagery with deep learning classification methods to establish a large-scale alpine treeline ecotone (ATE) classification map. The findings provide a valuable reference for the sustainable management of alpine ecosystems in the Xue Mountain glacial cirques in Taiwan.

Taiwan has the highest density of high mountains globally, with over 200 peaks exceeding 3,000 meters in elevation. The Alpine Treeline Ecotone (ATE) is a transitional zone between different vegetation types. The species distribution, range variations, and movement patterns of vegetation within the ATE are crucial indicators for assessing the impact of climate change and warming on alpine ecosystems. Therefore, this study focuses on the Xue Mountain glacial cirques in Taiwan (approximately 4 km²) and utilizes WorldView-2 satellite images from 2012 and 2021 to compute various vegetation indices and texture features (GLCM). By integrating these features with the Random Forest (RF) and U-Net models, we developed a classification map of the alpine treeline ecotone (ATE) in Xue Mountain. We analyzed changes in bare land, forest, krummholz, and shadows within the ATE from 2012 to 2021. The results indicate that the classification accuracy reached an overall accuracy (OA) of 0.838 when incorporating raw spectral bands along with vegetation indices and texture features (GLCM) (77 features in total). Feature importance ranking and selection reduced training time by 14.3 % while ensuring alignment between field survey treeline positions and classification results. From 2012 to 2021, tree cover density increased, with the total forest area expanding by approximately 0.101 km². The elevation of tree distribution rose by 14 m, with the most significant area changes occurring between 3,500 and 3,600 m, while the 3,700 to 3,800 m range remained relatively stable. This study integrates remote sensing imagery with deep learning classification methods to establish a large-scale alpine treeline ecotone (ATE) classification map. The findings provide a valuable reference for the sustainable management of alpine ecosystems in the Xue Mountain glacial cirques in Taiwan.

The Matsu Islands Tern Refuge comprises eight reefs located at a relay station on the East Asian bird migration route, and it attracts many transiting, wintering, or breeding birds to inhabit and live on the reefs every year. In order to understand the compositions of plant communities as a breeding habitat for seabirds, we investigated the plant communities of the eight reefs. A total of 130 plots of 10 × 10 square meters were established, from which we found 107 species of plants in 102 genera and 51 families. Among this, we found one critically endangered (CR) species, four vulnerable (VU) species, and three near-threatened (NT) species. The result of two-way indicator species analysis (TWINSPAN) and indicator value (IndVal) showed 130 samples were divided into 11 vegetation types; most of the vegetation types had significant indicator species. We also use the two-way to present the plot of detrended correspondence analysis (DCA) by vegetation types and reefs. Moreover, this result reveals that these samples were more clearly cluster divided by islands. Our results reveal that the compositions and characteristics of plant communities were related clearly to the environmental factors for each reef in the Matsu Islands Tern Refuge. Canonical correspondence analysis (CCA) indicated that species composition of vegetation yielded high correlation with soil property, especially with soil pH. In addition, we found that the traces of bird activity is relevant to the characteristics and structures of plant communities. We found that the plant communities comprising low-grass shrubs would provide relatively soft nesting materials and sheltering effects for eggs or hatchlings for terns. Compared to low-grass shrubs, the traits of high-grass shrubs would not be beneficial to nest for breeding of terns on the ground, and no nested trace was found in these plant communities.

Fire is one of the principal factors influencing subalpine ecosystem succession. Species numbers and plant compositions are used to determine postfire disturbance, vegetation, structural change, and succession. Ecologists also integrate species diversity and mathematical models to enable researchers to obtain increasingly detailed insights into habitats during post-disturbance restoration processes. This study employed five species-abundance models, namely the niche preemption model, the broken-stick model, the log-normal model, the Zipf model, and the Zipf–Mandelbrot model, to perform fitting analysis on the abundance data of postfire species coverage in shrub grasslands near 369 Hut at Xue Mountain in Shei-Pa National Park, Taiwan. We performed the logarithmic transformation on plant-coverage areas for each period of postfire shrub-grassland succession, and then, based on histograms drawn for species–coverage distribution modes, the test results consistently showed normal distributions (p < 0.05). Species-coverage histograms measuring various periods showed that there were comparatively higher numbers of common species during postfire succession and that the numbers of rare species progressively increased. The fitting results of the five species-abundance models showed that although the most suitable abundance models for each period of postfire succession varied, the majority of these periods demonstrated decent fitting with respect to the Zipf–Mandelbrot model. These findings showed that fuel consumption provided nutrients in a manner that facilitated postfire regeneration. Moreover, dominant species, such as Yushania niitakayamensis, and Miscanthus transmorrisonensis, did not fully occupy growing spaces and resource availabilities; consequently, seeded species were able to grow.

Abies kawakamii (Hay.) Ito is an endemic species mainly distributed in Taiwan’s subalpine region with an elevation of 3,000–3,600 m above sea level. It is a significant ecological indicator species in high subtropical mountains regarding plant geography and climate change. Taiwan is an island in the Asian-Pacific region. Typhoons cause forest gaps belonging to small-scale disturbances in the forest ecosystem, which is critical in maintaining species diversity and forest regeneration. We set up seven long-term monitoring plots of Ab. kawakamii forest to investigate the stand structure, species composition, and environmental factors of Mount (Mt.) Xue in Shei-Pa National Park. Comparing tree numbers between two surveys in a decade. Most of the population structure of Ab. kawakamii showed a similar positive skewness distribution between the two surveys, indicating that the population of Ab. kawakamii forest in Mt. Xue would be stable. According to principal component analysis (PCA), population structure of Ab. kawakamii correlated with dominant understory species and saplings of Ab. kawakamii but may be related to their habitat condition, such as elevation. Death trees released space and resources as gaps in the subalpine forest, which initiated micro-succession in the Ab. kawakamii forest.

The contributions of urban forests and green spaces to sustainable development have been confirmed. Meanwhile, cities worldwide have shown that investments in urban forestry can greatly contribute to citizens’ quality of life. This study was conducted in urban forests in the Dadu Terrace of Taichung City, central Taiwan, which were frequently disturbed by fires and had grassland severely invaded by Panicum maximum after the forest degraded. We sampled 46 plots in Dadu Terrace to understand the relationship between the soil seed bank and vegetation as well as to evaluate the feasibility of applying soil seed bank transfers for ecological restoration in Dadu Terrace. The grassland was dominated by Panicum maximum. Forest vegetation was distinguished by cluster analysis into five types, i.e., Ficus microcarpa type, Acacia confusa type, Litsea glutinosa type, Cinnamomum camphora type, and Trema orientalis type. In the aboveground survey, we recorded 141 vascular plants, including 129 seed plants and 12 ferns. There were 40 identified species of naturalized plants. A total of 29,914 seedlings were recorded in the soil seed bank, with an average seed density of 9634 seeds/m² and a total of 91 species. There were 40 species of naturalized plants, accounting for 90.9% of the total seed reserves. This showed that Dadu Terrace was severely affected by the invasion of naturalized species. The species number and seed reserves of woody plants of the Panicum maximum type were significantly lower than those of forest vegetation. The composition of the soil seed bank was dominated by naturalized plants, indicating that the high frequency of fire reduced the proportion of native species and woody plants in the soil seed bank. Acacia confusa type was the main forest type in Dadu Terrace. Although several woody species and seed reserves were in its soil seed bank, the naturalized proportions were even higher. Trema orientalis type was the secondary forest type in Dadu Terrace; it had the smallest forest area. However, it was the only vegetation type with a greater tree seed abundance than herbs and the lowest proportion of naturalized seed abundance. Trema orientalis type vegetation has a relatively high soil transfer value for ecological restoration but lacks diversity. Our results revealed that the characteristics of the soil seed bank of Dadu Terrace make it challenging to restore the grassland to the forest by natural succession. Therefore, we suggest that artificial restoration is necessary for Dadu Terrace.