An important problem currently faced by karst rocky desertification management is the degradation of vegetation fruits, which encourages the search for a solution for the sustainable growth of vegetation. Water and fertilizer are key factors affecting crop quality (the formation of soluble solids, protein, amino acids, soluble sugar content, etc.). A comprehensive review of research related to water–fertilizer coupling and crop quality improvement may be beneficial for rocky desertification control. This study analyzed 427 related papers, and the results showed that: (1) the number of published papers showed an “S” curve growth trend over time, indicating that research in the field of water–fertilizer coupling has entered a mature stage after rapid growth, yet it is still in the initial stage in karst areas; (2) the research was mainly theoretical, but also included technology research and development, model construction, demonstration and extension, and monitoring and evaluation; (3) except for research related to greenhouse facilities, the research areas are mainly located in arid and semi-arid regions. The study further revealed (4) the frontier theory of water–fertilizer coupling and crop quality research. Finally, it was found that the theoretical and technical studies related to water–fertilizer coupling and crop quality improvement in karst areas are insufficient. Whether the existing theories and techniques are applicable to karst areas needs further verification. In addition to the areas of selecting germplasm with high water/fertilizer utilization efficiency and tolerance to rocky desertification, further improvement in the dynamic model of water–fertilizer coupling and the construction of a more applicable water–fertilizer coupling model could also be future research hotspots.

Under extreme drought and climate change, golden pear trees have experienced problems such as yield reduction, dryness and death. This suggests that we know very little about the mechanisms regulating pear tree growth, assuming that meteorological factors positively influence plant sap flow. Based on this, we used the heat ratio method to monitor the sap flow of pear trees from June to December 2020, and recorded the changes in various environmental factors. The results showed that: (1) Sap flow velocity has obvious radial variability in tree sections; the sap flow velocity during the day was significantly higher than that at night (p < 0.05) and was higher in the growing season than in the non-growing season. (2) All environmental factors, except relative humidity and precipitation, were positively correlated with sap flow, vapor pressure deficit and photosynthetically active radiation, which are the key factors affecting daytime flow, and vapor pressure deficit and plant water potential are the key factors affecting nighttime flow. The linear regression results also showed that the daytime sap flow had a significant positive effect on the nighttime sap flow (p < 0.05). (3) The contribution of night flow to total daily flow varied from 17.3% to 50.7%, and most of the non-growing season values were above 40%. The results show that nighttime sap flow accounts for a significant portion of the pear tree’s water budget. Continuous irrigation during fruit enlargement and non-growing seasons will increase fruit yield and maintain plant sap flow activity to avoid death due to drought.

In the context of global climate change, drought events significantly impact plants, and studying the response of plant sap flow to environmental factors is of great significance for plant protection and sustainable development. This article uses the thermal pulse method to obtain trunk sap flow data, combined with real-time monitoring of environmental factor data, to compare the differences in sap flow characteristics between the subsequently deceased and surviving Zanthoxylum planispinum var. dingtanensis during a natural drought event. Additionally, the xylem vulnerability curve of Z. planispinum was fitted through laboratory measurements to estimate its hydraulic safety margin ($HSM$). The results showed that: (1) Compared with the surviving individuals, the subsequently deceased Z. planispinum exhibited a higher and more fluctuating sap flow rate, maintaining a consistently higher transpiration rate even after the onset of drought. (2) As soil moisture decreased, the sap flow of surviving Z. planispinum was promptly constrained by soil water availability and remained at a low, conservative level. In contrast, the sap flow of the deceased individuals continued to be driven by meteorological factors, failing to downregulate transpiration. (3) The xylem vulnerability curve revealed a $P_{50}$ of −2.2 MPa. During the severe drought, the minimum field branch water potential (${\psi }_{min}$) dropped to −3.0 MPa, resulting in a negative $HSM$ ($HSM$ = −0.8 MPa). This suggests that the xylem tension severely breached the embolism threshold, likely triggering catastrophic hydraulic failure, which emerged as the primary driver of the observed mortality in Z. planispinum. Therefore, for plants with low $HSM$s, when they are subjected to drought stress accompanied by highly fluctuating sap flow—a condition indicating that water supply cannot match the transpirational demand—timely manual intervention (e.g., supplemental irrigation) may be considered to mitigate hydraulic risks and ensure survival.

Accurate prediction of soil organic carbon (SOC) in cropland is important for food production, sustainable soil management, and carbon sequestration. Although digital soil mapping (DSM) has been widely used in the prediction of SOC, most of the current DSM studies use only a single remote sensing image and a single machine learning (ML) approach, and few studies apply multi-temporal remote sensing images and ensemble methods. This study explores the accuracy of the prediction of surface SOC in cropland by comparing multi-temporal Sentinel-2A remote sensing with random forest (RF), support vector machine (SVM), gradient boosted decision trees (GBDT), extreme gradient boosted decision trees (XGBoost), and a stacking ensemble method consisting of these four ML approaches. The potential of multi-temporal remote sensing data and the stacking ensemble method for SOC prediction is discussed. To this end, 76 sampling points were selected in the study area, soil samples were collected at depths of 0–10 cm and 10–20 cm for each soil profile, and a total of 152 soil samples were obtained. Remote sensing variables extracted from topography, climate, and Sentinel-2A images on 13 January and 31 August 2023 were used as predictor variables. The results showed that the stacking ensemble method with multi-temporal predictor variables outperformed all single models and variable combinations. However, the overall predictive accuracy remained moderate, with the best performance for 0–10 cm (R² = 0.386, RMSE = 4.782, MAE = 3.36) and 10–20 cm (R² = 0.425, RMSE = 4.484, MAE = 4.031). The relatively low R² values, despite the use of advanced methods, highlight the inherent challenges of SOC prediction in highly fragmented karst croplands. This study demonstrates the incremental benefit, rather than a universal high accuracy, of combining multi-temporal Sentinel-2 imagery with a stacking ensemble to improve SOC mapping in such complex environments.