Taiwan Forestry Research Institute

Global climate change has substantially altered forest fire regimes, with important implications for the dynamic stability of forest ecosystems. Understanding post-fire recovery patterns is critical for assessing ecosystem resilience and informing effective conservation management. However, post-fire recovery varies widely, reflecting complex interactions among multiple environmental factors and being further influenced by secondary disturbances and human restoration efforts. To address this, we analyzed historical burned area from 1987 to 2023 in Taiwan’s high-mountain national parks, using the differenced normalized burn ratio (dNBR) derived from Landsat spectral reflectance to characterize post-fire spectral responses, including the pattern of recovery status and recovery time. We also employed the random forest model and SHAP analysis to assess how environmental factors and secondary disturbances influenced both recovery status and recovery time, identifying key drivers of post-fire regeneration. Approximately 60 % of the burned grids recovered by 2023, although some have shown no spectral recovery after thirty-two years. The average recovery time was 13.9 ± 8.5 years for forest vegetation and 11.9 ± 9.6 years for non-forest vegetation. Random forest modeling supported the influence of fire-related and environmental factors on post-fire recovery. Our findings indicate that high-mountain forest ecosystems are particularly vulnerable, and severe wildfires can likely trigger secondary disturbances such as landslides and soil erosion, significantly prolonging the post-fire recovery time. The limited impact of reforestation efforts highlights the need to strengthen proactive prevention strategies and management mechanisms to effectively reduce ecological losses and resource costs, thereby protecting these vulnerable humid high-mountain ecosystems. • We analyzed post-fire forest recovery time and its key drivers in national parks. • While delayed mortality is common, secondary disturbances are the key recovery driver. • Mean recovery time is 14 years, and approximately 60 % of burned areas recovered. • Artificial reforestation has limited success in steep, inaccessible terrain. • Fire prevention is more effective than restoration in humid mountainous region.

Abstract. 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.

The growth and optimization of Al(In)GaN back barrier by MOCVD for InAlGaN-based heterostructures is successfully demonstrated. At zero growth temperature gap (which was defined by ΔT g = ∣T g , channel -T g , BB ∣), the coherent growth of the back barrier was confirmed by perfectly ordered atoms, sharp interface quality, and absence of threading dislocations at the buffer/back-barrier/channel interfaces, including the smoothest surface of heterostructures. The electron transport properties of InAlGaN/GaN heterostructures were consistently affected by the growth temperature of the back barrier. Significant increase of electron mobility from 1560 to 1740 (cm 2 Vs −1 ) and decrease of sheet carrier density from 1.57 × 10 13 to 1.31 × 10 13 (cm −2 ) were attributed to the improvement of electron confinement by adapting a thin ( 3.5 nm) optimized Al(In)GaN back barrier to the conventional InAlGaN/GaN heterostructure. XPS chemical shifts of the N1s core level and band alignment calculation have also confirmed the influence of the growth temperature of the back barrier on the electron confinement. Moreover, a large positive shift of the threshold voltage (2 V), a considerable increase in maximum transconductance from 180 to 216 (mS mm −1 ), and suppression of the Kink effect of the devices were realized, which paves the way for the employment of the Al(In)GaN back barrier for high-frequency applications.

ABSTRACT Question Alpine regions are particularly vulnerable to climate change, with plants adapted to harsh conditions proving sensitive to shifting environmental factors. While warming can enhance productivity, the combined effects of warming and drought represent a severe threat. We investigated how concurrent warming and drought influence alpine vegetation in Taiwan, testing the hypotheses that (1) warming may increase productivity when water is sufficient, but co‐occurring severe drought and extreme heat will cause disproportionate declines and (2) severe agricultural drought events would lead to declines in mesic and cold‐adapted species, favoring drought‐tolerant species and increasing compositional homogenization. Location Fifteen summit plots in five high‐mountain regions of Taiwan. Methods We applied a multi‐scale approach integrating satellite‐based vegetation productivity indices, long‐term climate anomalies, field‐based diversity measurements, and species‐level climatic optima. Climate events were classified by intensity and frequency from 1951 to 2022, and vegetation changes were assessed using field survey and remote sensing data from 2019 to 2022. Results During 1951–2022, the study regions experienced an average of 1.08 severe droughts and 0.56 extreme drought events per decade, with a marked increase in extreme drought events since 2016. From 2019 to 2022, negative water balance and elevated temperatures led to reduced productivity. Productivity declined moderately under single‐stressor scenarios but dropped significantly when drought and warming co‐occurred. This corresponded with a decline in species richness and evenness, disappearance of mesic and cold‐adapted species, and increased vegetation homogenization. Conclusions Our findings indicate that the co‐occurring drought and extreme heat conditions substantially impact alpine vegetation resilience, driving species loss and community homogenization. With future climate scenarios projecting further decline, even humid alpine climatic region may face ongoing drought‐related threats. Our results underscore the importance of sustained field monitoring and integrated multi‐scale strategies to conserve the alpine biodiversity under climate change.

Abstract— A new species of Primulina (Gesneriaceae) from Cuc Phuong National Park, northern Vietnam is described based on morphological and molecular evidence. Primulina cucphuongensis is distinguished from the other Primulina species by its unique combination of morphological characters, including dark green leaf blade with densely white puberulous hairs at margin, violet corolla, and peduncle and pedicel with deep red puberulous hairs and trichomes as well as its distinct phylogenetic position.

Secondary growth is a key characteristic evolved from seed plants and generates secondary xylem-the most abundant tissue on Earth. Recent studies have uncovered xylem developmental lineages in eudicots and magnoliids of angiosperms. However, xylem development in gymnosperms, the other representative clade of seed plants, remained elusive. We performed single-cell transcriptomics for xylem cells of conifers (Cunninghamia lanceolata), the major clade in gymnosperms. Using Seurat and scVI-based cross-species integration, we reconstructed the xylem differentiation trajectories and revealed that the radial system is conserved across seed plants, while the axial system in C. lanceolata exhibits a composite lineage architecture resembling both eudicots and magnoliids. To validate these trajectories, we established a multi-modal spatial framework incorporating spatial transcriptomics, spatial proteomics, and spatial metabolomics. These three spatial layers provided orthogonal evidence confirming cell-type identities and trajectory inference. Additionally, we identified a xylem cell population unique to gymnosperms, suggesting a lineage-specific specialization. Together, our findings uncover a more complex ancestral xylem architecture in gymnosperms and propose a progressive simplification of axial developmental programs from gymnosperms to angiosperms, highlighting a trajectory of reductive evolution in seed plant vascular development.

The field of autonomous driving has seen continuous advances, yet achieving higher levels of automation in real-world applications remains challenging. A critical requirement for autonomous navigation is accurate map construction, particularly in novel and unstructured environments. In recent years, Simultaneous Localization and Mapping (SLAM) has evolved to support diverse sensor modalities, with some implementations incorporating machine learning to improve performance. However, these approaches often demand substantial computational resources. The key challenge lies in achieving efficiency within resource-constrained environments while minimizing errors that could degrade downstream tasks. This paper presents an enhanced ORB-SLAM2 (Oriented FAST and Rotated BRIEF Simultaneous Localization and Mapping, version 2) algorithm implemented on a Raspberry Pi 3 (ARM A53 CPU) to improve mapping performance under limited computational resources. ORB-SLAM2 comprises four main stages: Tracking, Local Mapping, Loop Closing, and Full Bundle Adjustment (BA). The proposed improvements include employing a more efficient feature descriptor to increase stereo feature-matching rates and optimizing loop-closing parameters to reduce accumulated errors. Experimental results demonstrate that the proposed system achieves notable improvements on the Raspberry Pi 3 platform. For monocular SLAM, RMSE is reduced by 18.11%, mean error by 22.97%, median error by 29.41%, and maximum error by 17.18%. For stereo SLAM, RMSE decreases by 0.30% and mean error by 0.38%. Furthermore, the ROS topic frequency stabilizes at 10 Hz, with quad-core CPU utilization averaging approximately 90%. These results indicate that the system satisfies real-time requirements while maintaining a balanced trade-off between accuracy and computational efficiency under resource constraints.

This paper proposes a 10× zoom-ratio optical design with intermediate optics to maximize image size and minimize the front diameter. Image size and front diameter have generally been regarded as inevitable trade-offs since the development of zoom optics started in the 1950s. As a result of fast computer-aided optical design, optics with a fixed focal length have gained success in large image planes and in very compact lenses; however, it remains a critical topic in high-ratio zoom optics. In the proposed high-zoom-ratio optics, the diameter remains fixed, whereas the image size increases by up to four times, which dramatically enhances performance. The proposed optical design exhibited excellent performance in highly demanding specifications.

The genus Myobia (Acari: Myobiidae) includes obligate ectoparasites of rodents, with M. murismusculi being the most frequently reported species in laboratory mouse colonies worldwide. However, morphological overlap and nomenclatural confusion have long hindered accurate species identification. This study aims to describe a new fur mite species from Mus musculus and to distinguish it from M. murismusculi and related taxa using detailed morphological and morphometric analyses. Fur mites were collected from laboratory mice in two facilities and examined using light and scanning electron microscopy. Morphological characters were described, measured, and illustrated. Sstandardized morphometric characters were analyzed across multiple Myobia species using principal component analysis (PCA) and permutational multivariate analysis of variance (PERMANOVA) to assess interspecific variation and species boundaries. The new species, Myobia murina sp. nov., differs from M. murismusculi by longer dorsal setae ve and si. PCA and t-SNE analyses revealed a distinct cluster for M. murina, and PERMANOVA confirmed significant morphological divergence from all examined congeners (R² = 0.873, p < 0.001). An updated identification key for female mites of the subgenus Myobia is also provided. This study demonstrates the utility of morphometric approaches in mite taxonomy and confirms M. murina sp. nov. as a morphologically and statistically distinct species. Accurate species identification is essential for laboratory animal health management and fur mite diagnostics.