Enunciating the carbon storage across various types of forests is a precondition for comprehending forest ecosystem carbon cycling. However, previous studies regarding forest carbon storage were primarily conducted in the general temperature zones, with a limited understanding of carbon storage in different forest types within climate transition zones. In this study, we employed biomass models to explore the carbon storage in three types of natural secondary forests (Pinus armandii forest, Quercus aliena forest, and Q. aliena–P. armandii mixed forest) in the transition zone between the northern subtropical and warm temperate regions of China. Furthermore, we used the variance decomposition analysis and random forest model to determine the key factors influencing carbon storage in three types of natural secondary forests. Our results indicated that the carbon storage of wood and soil layers in the Q. aliena–P. armandii mixed forest was significantly higher than that in the P. armandii and Q. aliena forests. Total carbon storage was ranked as follows: Q. aliena–P. armandii mixed forest (266.09 t/ha) > P. armandii forest (222.89 t/ha) > Q. aliena forest (212.46 t/ha). Our results also revealed that carbon storage of the wood layer was jointly regulated by environmental factors, plant physiological characteristics, and soil properties, while soil carbon storage was mainly affected by soil properties. These results highlight the significant advantages of mixed conifer–broadleaf forests in carbon storage, emphasizing the importance of mixed natural secondary forests in carbon cycling and ecosystem services. This study provides scientific evidence for enhancing forest carbon sink functions and developing forest conservation and management policies to combat climate change.

Plantation cultivation plays an important role in improving terrestrial ecosystem functions and services. Understanding the water-use patterns of major afforestation species is vital for formulating ecological restoration strategies and predicting the response of plantation to climate change. However, the impacts and drivers of forest types on water-use patterns of key tree species are poorly understood. Here, the combined methods of dual stable isotope of δD and δ¹⁸O and Bayesian mixed framework (MixSIAR) were employed to investigate the water-use patterns of Cunninghamia lanceolata (Chinese fir) in a monoculture, mixed forest with Cinnamomum camphora, and mixed forest with Alnus cremastogyne under different rainfall events in subtropical China. Furthermore, the relative contribution of different soil and plant factors to the water-use patterns of Chinese fir was quantified using a random forest model. Our results showed that Chinese fir in the mixed forests (with C. camphora or with A. cremastogyne) utilized less water from shallow soil compared to that in a monoculture but significantly improved the proportion of water absorbed from deep soil with the increase of 55.57%–64.90% and 68.99%–108.83% following moderate and heavy rainfall events, respectively. The most important factors contributing to the differences in water-use patterns of Chinese fir among monoculture and mixed forests were tree attributes (i.e., leaf biomass, eco-physiological regulation, and fine root biomass). These findings reveal that Chinese fir in mixed forests could optimize water-use patterns by adjusting plant properties for interspecific niche complementarity, improving the utilization of deep soil water. Overall, this study suggests that mixed-species plantations could improve water-use efficiency and reduce the sensitivity of tree species to precipitation change, indicating they are better able to cope with expected climate variability.

The climate-induced changes in soil water patterns pose a serious threat to subtropical plantations. Mixed species stands have been advocated as an efficient way to enhance ecosystem stability. However, little is known about their possible impact on the soil water-holding capacity in the subtropics. In this study, we employed a stable hydrogen isotope to assess the contribution of rainfall to soil water (CRSW) in a pure Chinese fir (Cunninghamia lanceolata) plantation and in two mixtures of Chinese fir with Cinnamomum camphora or with Alnus cremastogyne after three different magnitudes of rainfall events in subtropical China. Furthermore, we used structure equation modeling (SEM) to quantify the relative importance of vegetation and soil properties on the CRSW. The results indicated that the CRSW did not differ among these three Chinese fir plantations after light rainfall, whereas the CRSW of moderate and heavy rainfall to soil water were 15.95% and 26.06% higher in Chinese fir plantation with Cinnamomum camphora, and 22.67% and 22.93% higher in Chinese fir plantation with Alnus cremastogyne than that in the pure Chinese fir plantation, respectively. SEM analysis showed that the vegetation biomass and soil properties significantly affected the CRSW following light rainfall, but the soil properties were the most important factors influencing the CRSW under moderate and heavy rainfall. Our findings demonstrate that the mixed conifer–broad-leaved plantation is a more effective strategy for improving the soil water-holding capacity than the pure conifer plantation in subtropical regions, which is conducive to coping with the frequent seasonal droughts and extreme precipitation events.

Accurate registration of multi-platform light detection and ranging (LiDAR) point clouds is essential for detailed forest structure analysis and ecological monitoring. In this study, we developed a novel two-stage method for aligning terrestrial and unmanned aerial vehicle LiDAR point clouds in forest environments. The method first performs coarse alignment using canopy-level digital surface models and Fast Point Feature Histograms, followed by fine registration with Iterative Closest Point. Experiments conducted in six forest plots achieved an average registration accuracy of 0.24 m within 5.14 s, comparable to manual registration but with substantially reduced processing time and human intervention. In contrast to existing tree-based methods, the proposed approach eliminates the need for individual tree segmentation and ground filtering, streamlining preprocessing and improving scalability for large-scale forest monitoring. The proposed method facilitates a range of forest applications, including structure modeling, ecological parameter retrieval, and long-term change detection across diverse forest types and platforms.