Net primary productivity (NPP) is a core ecological indicator within terrestrial ecosystems, representing the potential of vegetation growth to offset anthropogenic carbon emissions. Thus, assessing NPP in a given region is crucial for promoting regional ecological restoration and sustainable development. This study utilized the CASA model and GEE to calculate the annual average NPP in Shandong Province (2001–2020). Through trend analysis, Moran’s Index, and PLS−SEM, the spatiotemporal evolution and driving factors of NPP were explored. The results show that: (1) From 2001 to 2020, NPP in Shandong showed an overall increasing trend, rising from 254.96 to 322.49 g C·m⁻²/year. This shift was accompanied by a gradual eastward movement of the NPP centroid, indicating significant spatial changes in vegetation productivity. (2) Regionally, 47.9% of Shandong experienced significant NPP improvement, 27.6% saw slight improvement, and 20.1% exhibited slight degradation, highlighting notable spatial heterogeneity. (3) Driver analysis showed that climatic factors positively influenced NPP across all four periods (2005, 2010, 2015, 2020), with the strongest impact in 2015 (coefficient = 0.643). Topographic factors such as elevation and slope also had positive effects, peaking at 0.304 in 2015. In contrast, human activities, especially GDP and nighttime light intensity, negatively impacted NPP, with the strongest negative effect in 2010 (coefficient = −0.567). These findings provide valuable scientific evidence for ecosystem management in Shandong Province and offer key insights for ecological restoration and sustainable development strategies at the national level.

Subtropical forests play an important role in the global carbon cycle and climate change mitigation. In order to understand the effects of climate factors on soil carbon in subtropical forest ecosystems, it is necessary to make full use of carbon sequestration potential. Soil organic carbon (SOC) and soil alkali-hydrolyzed nitrogen (SAN) were tested in 255 plots of subtropical forests in Zhejiang Province, and their forest reserves from 2020 in Zhejiang Province were compared with those from 2010. The results showed that SOC content significantly increased, but SAN content decreased over those ten years. Combined with random forest (RF) and correlation analysis, the contribution of different climate factors (temperature, precipitation, etc.) to soil carbon storage was analyzed, and the main driving factors were evaluated. The RF model explained that winter (December to February) and spring (March to May) were the most dominant drivers to the 0–10 cm and 10–30 cm increases in SOC. There was a significant positive correlation between precipitation and SOC accumulation (0–30 cm) during winter and spring. The minimum temperatures in summer (June to August) and autumn (September to November) were negatively correlated with SOC accumulation (0–30 cm). Increasing the precipitation or irrigation (cloud seeding) in winter could improve the carbon sequestration capacity of subtropical forest soils. This study provides a new perspective on the sensitivity and potential response of the carbon cycle to climate change in subtropical forest ecosystems.

As a crucial forest resource in southern China and a significant economic forest species for forestry production, moso bamboo has a notable influence on carbon stocks across the entire bamboo forest ecosystem. Studying the impact of different management measures on carbon stocks in moso bamboo forests and soil carbon stocks can assist bamboo forest operators in incorporating the carbon sequestration capacity of bamboo into forest production and management decisions, which can contribute to achieving carbon sequestration, emission reduction, and sustainable development in the decision-making processes of forest production and management. In this study, we utilized a randomized block design to investigate the changes in moso bamboo forests’ carbon stocks and soil carbon stocks under different management measures across three intensities: high-intensity intensive management (HT), moderate-intensity intensive management (MT), and regular management (CK). Additionally, we employed meta-analysis methods to enhance the accuracy of our conclusions. The experimental results showed that MT increased the carbon storage in moso bamboo forests by 19.86%, which was significantly different from CK (p < 0.05), while there was no significant difference between the HT group and the MT and CK groups. For soil carbon stocks, in the 10–30 m and 0–50 m soil layers, HT decreased soil carbon storage by 29.89% and 22.38%, while MT increased soil carbon storage by 64.15% and 31.02%, respectively. Both HT and MT were significantly different from CK (p < 0.05). However, for the soil layers of 0–10 m and 30–50 m, there was no significant difference between the treatments within the experimental group. The results of the meta-analysis indicate that, compared to traditional regular management, intensive management, especially high-intensity intensive management, can significantly increase the carbon storage in bamboo forests (p < 0.05). However, it will significantly reduce soil carbon storage (p < 0.05). Moreover, a significant difference in soil carbon storage is observed only within the 0–20 cm soil layer group. Therefore, from the perspective of the long-term ecological benefits of bamboo forest management, the selection of management measures should prioritize reasonable and moderate-intensity intensive management. Additionally, adopting appropriate and moderate-intensity fertilization, ploughing, and other management methods is recommended to enhance the productivity of moso bamboo forests while concurrently protecting the natural environment and improving the carbon sequestration capacity of moso bamboo forests.

As global climate change intensifies and human activities escalate, changes in vegetation cover, an important ecological indicator, hold significant implications for ecosystem protection and management. Shandong Province, a critical agricultural and economic zone in China, experiences vegetation changes that crucially affect regional climate regulation and biodiversity conservation. This study employed normalized difference vegetation index (NDVI) data, combined with climatic, topographic, and anthropogenic activity data, utilizing trend analysis methods, partial correlation analysis, and Geodetector to comprehensively analyze the spatiotemporal variations and primary driving factors of vegetation cover in Shandong Province from 2001 to 2020. The findings indicate an overall upward trend in vegetation cover, particularly in areas with concentrated human activities. Climatic factors, such as precipitation and temperature, exhibit a positive correlation with vegetation growth, while land use changes emerge as one of the key drivers influencing vegetation dynamics. Additionally, topography also impacts the spatial distribution of vegetation to a certain extent. This research provides a scientific basis for ecological protection and land management in Shandong Province and similar regions, supporting the formulation of effective vegetation restoration and ecological conservation strategies.

In recent years, the automatic recognition of tree species based on images taken by digital cameras has been widely applied. However, many problems still exist, such as insufficient tree species image acquisition, uneven distribution of image categories, and low recognition accuracy. Tree leaves can be used to differentiate and classify tree species due to their cognitive signatures in color, vein texture, shape contour, and edge serration. Moreover, the way the leaves are arranged on the twigs has strong characteristics. In this study, we first built an image dataset of 21 tree species based on the features of the twigs and leaves. The tree species feature dataset was divided into the training set and test set, with a ratio of 8:2. Feature extraction was performed after training the convolutional neural network (CNN) using the k-fold cross-validation (K-Fold–CV) method, and tree species classification was performed with classifiers. To improve the accuracy of tree species identification, we combined three improved CNN models with three classifiers. Evaluation indicators show that the overall accuracy of the designed composite model was 1.76% to 9.57% higher than other CNN models. Furthermore, in the MixNet XL CNN model, combined with the K-nearest neighbors (KNN) classifier, the highest overall accuracy rate was obtained at 99.86%. In the experiment, the Grad-CAM heatmap was used to analyze the distribution of feature regions that play a key role in classification decisions. Observation of the Grad-CAM heatmap illustrated that the main observation area of SE-ResNet50 was the most accurately positioned, and was mainly concentrated in the interior of small twigs and leaflets. Our research showed that modifying the training method and classification module of the CNN model and combining it with traditional classifiers to form a composite model can effectively improve the accuracy of tree species recognition.