Ecohydrology最新文献

The intensification of global change and the escalation of anthropogenic activities have increasingly exacerbated drought conditions in Central Asia (CA), threatening regional ecosystem sustainability and socioeconomic stability. Although previous studies have characterized drought spatiotemporal patterns, a systematic assessment of vegetation responses, particularly regarding lag effects, cumulative impacts and adaptive strategies, remains lacking. Therefore, this study utilized multisource remote sensing and hydrological datasets (e.g., Standardized Precipitation Evapotranspiration Index [SPEI], Normalized Difference Vegetation Index [NDVI], Solar-Induced Chlorophyll Fluorescence [SIF] and Gravity Recovery and Climate Experiment-Drought Severity Index [GRACE-DSI]) from 1982 to 2022, to quantify the lagged and cumulative effects of drought on vegetation and assess ecosystem responses through drought resistance and recovery capacity in CA. The results indicated that (1) vegetation responses to drought across CA exhibited significant lag and cumulative effects, with maximum SPEI-NDVI correlation coefficients (MCC) exceeding 0.4 at 4–8-month lags in 64.5% of CA, and particularly prominent cumulative impacts at 10–16-month scales in southern CA; (2) grasslands and croplands demonstrated the highest drought sensitivity (DS), with the most pronounced fluctuations in SIF, while approximately 67.2% of CA exhibited strong resistance though there with low resistance expanded markedly during extreme drought events (e.g., in 2001 and 2021); and (3) short- and medium-lived vegetation exhibited strong post-drought recovery capacity (i.e., recovery index [RI] > 0.75), contrasting with slower and more variable recovery in long-lived types, while mean vegetation recovery in CA had declined notably since 2020. This study advances the understanding of drought-vegetation coupling in CA and provides data-driven insights for identifying drought-prone regions, evaluating ecosystem functionality and informing adaptive environmental management strategies.

As environmental indicators, invertebrate communities are closely related to the changes in the water environment. Efficient and accurate monitoring of invertebrates is of great significance for providing references for water environment conservation. Environmental DNA (eDNA) metabarcoding permits a useful tool for biodiversity monitoring in aquatic ecosystems. Panlong River is a typical urban river suffering from various anthropogenic activities in Kunming, Yunnan, China, and the current characteristics of invertebrate communities are unclear. Therefore, this study investigated the spatial and seasonal patterns of invertebrates and the environmental stressors of Panlong River. Based on eDNA metabarcoding, 968 amplicon sequence variants (ASVs) of invertebrates belonging to Annelida, Arthropoda, Cnidaria, Gastrotricha, Mollusca, Platyhelminthes, Protozoa, Rotifera and Tardigrada were identified, with Arthropoda being the absolute dominant phylum. Distinct spatial and seasonal variations in the invertebrate communities (e.g., ASV numbers, dominant genera, relative abundances and Shannon–Wiener index) were observed. Macrothrix was the first dominant genus both during dry and wet periods, while the second abundant genus was different during dry (Acanthamoeba) and wet (Cricotopus) periods. A spatial–seasonal heterogeneity of the relation between the invertebrate communities and environmental factors was observed in Panlong River. C, COD andNH₄⁺ were deemed to be the crucial environmental stressors influencing the distributions of the invertebrate communities in Panlong River, which was consistent with the spatial and seasonal differences in pollution characteristics around Panlong River. This study provides insights on conserving the diversity of invertebrate communities and the management of Panlong River and similar urban rivers.

Dissolved oxygen (DO) levels in river basins are critical indicators of water quality and the health of aquatic ecosystems. Machine learning has been widely used for DO monitoring, but its application is often constrained by the necessity to acquire comprehensive datasets, including hydrological, water quality and meteorological data, to accurately capture system dynamics. An integrated modelling approach was developed that employed environmental fluid dynamics code (EFDC) to simulate watershed-wide hydrology and generate high-frequency data. The maximal information coefficient (MIC) was then applied for data dimensionality reduction. Following this, six machine learning algorithms—backpropagation (BP) neural network, k-nearest neighbours (KNN), kernel ridge regression (KRR), fuzzy neural network (FNN), support vector regression (SVR) and long short-term memory (LSTM)—were employed to create hybrid MIC predictive models for various lag times ranging from 1 h to 15 days. For each specific lag time, a random forest (RF) ensemble model was developed to enhance prediction accuracy. The main findings are as follows: (1) High-frequency hydrological data from EFDC improved machine learning model accuracy by effectively capturing hydrodynamic characteristics. (2) MIC-based dimensionality reduction enhanced model performance, with R² and NSE reaching 0.87 and 0.89, and MAE and RMSE reduced to 0.5 and 0.77. MIC also aided in selecting prediction lag times by identifying autocorrelation trends. (3) The RF ensemble model outperformed individual models, achieving R² and NSE of 0.93, with MAE and RMSE of 0.37 and 0.57. It also showed high early warning accuracy during low oxygen periods.

Peatlands recognized as areas of water surplus, are highly sensitive to climate change, which induces new ecohydrological realities, particularly under drought conditions. In raised bogs, where precipitation serves as the sole water source, it plays a vital role in maintaining water balance and sustaining the health of peatland vegetation. Integrating ground-based observations with remotely sensed data offers an effective approach for detecting water stress patterns over large territories, as reflected in the phenological dynamics of these ecosystems. Therefore, the present study assessed drought in the Čepkeliai raised bog (Lithuania), using the correlations between Vegetation Condition Index (VCI), Standardized Precipitation Index (SPI) and in situ water table (WT) measurements. Terra MODIS satellite data of the growing season were used for 12 years (2012–2023) to assess the long-term vegetation conditions and identify spatio-temporal drought patterns. Study results showed that WT, as a primary driver of vegetation health is closely linked with 30-day precipitation (especially in June). Multiple timescale SPI was also strongly correlated with VCI. Significant positive correlations between VCI and WT were found in June and September, confirming the importance of these critical periods for monitoring vegetation water stress in the raised bog. Analysis of the different habitats confirmed the sensitivity of active raised bog to WT fluctuations. The spatial distribution of VCI values during dry years (2015, 2016, 2018 and 2019) revealed areas most susceptible to water stress within the raised bog.

Forest floor evapotranspiration (Ef) is a critical hydrological process in forested landscapes, accounting for up to 50% of total evapotranspiration. However, existing Ef estimation models primarily rely on meteorological inputs and overlook forest structural influences, limiting their applicability across diverse forest types. This study developed a modelling framework that explicitly incorporates the relative yield index (Ry) as a three-dimensional structural indicator to estimate Ef using a modified Penman–Monteith equation. Ry represents the ratio of current to maximum stand volume for given tree height and stand density, providing a comprehensive measure of forest structure. The framework estimates under-canopy meteorological conditions including net radiation, temperature, vapour pressure deficit and wind speed based on Ry and above-canopy meteorological data. We tested the framework using lysimeter measurements in Japanese cypress plantations with contrasting structural characteristics (Ry = 0.57 and 0.82). Significant relationships were established between Ry, leaf area index (R² = 0.34) and canopy openness (R² = 0.71) using data from field inventory and literature review. The model successfully reproduced observed daily Ef with totals of 112.2–165.9 mm, representing 6.9%–13.2% of annual precipitation. Mean daily Ef ranged from 0.1 to 2.5 mm/day, comparable to previous studies. Residuals were smallest during low vapour pressure deficit conditions (< 0.5 kPa), with larger deviations during high atmospheric demand periods when stomatal regulation and soil moisture limitations become important. This Ry-driven framework provides a transferable, inventory-based tool for estimating forest floor evapotranspiration across managed plantations.