Long-term changes in the abundance and distribution of tree species in the temperate forests of South Korea remain poorly understood. We investigated the changes in tree species composition in temperate mountainous forests using survey data from 130 permanent plots (0.1 ha) from the past 15 years (1998–2012) distributed across Mts. Jiri and Baegun, South Korea. The tree communities showed positive net changes in terms of stand density, richness, diversity, and evenness. At the species level, the change in relative species composition has been mainly driven by species such as Quercus mongolica, Carpinus laxiflora, Quercus serrata, Quercus variabilis, Styrax japonicus, Lindera erythrocarpa, and Pinus densiflora. These changes were categorized into five groups representing gradual increase or decrease, establishment, extinction, or fluctuation in species populations. At the community level, the changes in species composition showed consistent and directional increases in the annual rate of change for the mean species traits, including stand prevalence, pole growth rate, adult growth rate, and adult stature. Based on additive models, topographic variables (elevation, latitude, longitude, slope, topographic wetness index, and curvature) were more strongly associated with the distribution of species diversity than climate variables (annual mean minimum and maximum temperatures, temperature seasonality, annual rainfall, rainfall seasonality). Elevation was the most significant driver, followed by latitude and longitude. This study reveals the dynamics of change in tree species composition and distribution along topographical and climate gradients in South Korea and contributes to a broader understanding of temperate forest ecosystems for the purpose of better forest management.

Waterlogging induces oxidative damage by accumulation of reactive oxygen species due to stomatal closure. Plants alter their physiological and molecular mechanisms to reduce and adapt to oxidative stress. This mechanism of adaptation to stress, known as hardening, can support future stress tolerance. Pinus densiflora seedlings were grown under waterlogging treatment for three years and then transplanted to another site to identify the waterlogging hardening effect on transplanting. Transcriptome analysis was conducted before and after transplanting, and physiological factors were measured after transplanting. After transplanting, wounding stress is the main cause of transplant stress, and 13 genes related to phenylpropanoid were upregulated for the recovery of wounded roots in waterlogged hardened seedlings. The leaf starch and soluble sugar content of the waterlogged hardened seedlings were 50.3% and 40.5% lower due to the formation of cell walls. However, auxin-related genes were downregulated in waterlogging hardened seedlings, resulting in a lower tendency for height growth in hardened waterlogged seedlings. Waterlogging hardening mitigated transplant stress by wounding more than non-hardening, whereas waterlogging hardening may negatively affect seedling height. Our study provides evidence for the hardening effect of long-term waterlogging on transplanted P. densiflora seedlings.

Forests are essential for maintaining the ecological balance of the planet and providing critical ecosystem services. Amidst an increasing rate of global forest loss due to various natural and anthropogenic factors, many countries are committed to battling forest loss by planting new forests. Despite the reported national statistics on the land area in plantations, accurately delineating boundaries of planted forests with remotely sensed data remains a great challenge. In this study, we explored several deep learning approaches based on Convolutional Neural Networks (CNNs) for mapping the extent of planted forests in the Korean Peninsula. Our methodology involved data preprocessing, the application of data augmentation techniques, and rigorous model training, with performance assessed using various evaluation metrics. To ensure robust performance and accuracy, we validated the model’s predictions across the Korean Peninsula. Our analysis showed that the integration of the Near Infrared band from 10 m Sentinel-2 remote sensing images with the UNet deep learning model, incorporated with unfrozen ResNet-34 backbone architecture, produced the best model performance. With a recall of 64% and precision of 76.8%, the UNet model surpassed the other pixel-based deep learning models, including DeepLab and Pyramid Sense Parsing, in terms of classification accuracy. When compared to the ensemble-based Random Forest (RF) machine learning model, the RF approach demonstrates a significantly lower recall rate of 55.2% and greater precision of 92%. These findings highlight the unique strength of deep learning and machine learning approaches for mapping planted forests in diverse geographical regions on Earth.