Changes in the status of freshwater resources are a topic of major global, regional and local concern. This is especially so in the arid and semi-arid regions of China, where shortage of water resources plays a crucial role in limiting sustainable socioeconomic development, as well as in sustaining natural ecosystems. Recent climate change, as well as the effects of localized human activity, such as the use of water for irrigation agriculture, may have significant effects on the status of the water resources in the region. Here, we report the results of a study of changes in the areas of lakes in Gonghe Basin, northeastern Tibetan Plateau of China, over the last 60 years. The data were acquired from optical satellite images and demonstrate that the total water area of lakes in Gonghe Basin decreased significantly from the 1950s to 1980s. The cause is ascribed mainly to human activity including exploitation of farmland, against a background of increasing population; in addition, climatic data for the region demonstrate a minor drying trend during this period as the temperature increased slightly. After the construction of several reservoirs, significant amounts of water were redistributed to promote irrigation agriculture and we conclude that this caused a significant shrinkage of the natural lakes. However, both the area of farmland and the population size remained approximately constant after 1990. We conclude that the variation of the total area of lakes during the second period was mainly controlled by climatic factors (precipitation and temperature). As the regional temperature reached a new high, the area of some of the lakes decreased sharply before finally maintaining a relatively steady state. We emphasize that anthropogenic climate change and human activity have both significantly influenced the status of water resources in the arid and semi-arid regions of China.

The Lenglongling Mountains (LLM) located in northeastern part of the Tibet Plateau, belong to a marginal area of the East Asian summer monsoon (EASM) and are sensitive to monsoon dynamics. Two tree-ring width chronologies developed from six sites of Picea crassifolia in the LLM were employed to study the regional drought variability. Correlation and temporal correlation analyses showed that relationships between the two chronologies and self-calibrated Palmer Drought Severity Index (sc_PDSI) were significant and stable across time, demonstrating the strength of sc_PDSI in modeling drought conditions in this region. Based on the relationships, the mean sc_PDSI was reconstructed for the period from 1786 to 2013. Dry conditions prevailed during 1817–1819, 1829–1831, 1928–1931 and 1999–2001. Relatively wet periods were identified for 1792–1795 and 1954–1956. Spatial correlations with other fourteen precipitation/drought reconstructed series in previous studies revealed that in arid regions of Northwest China, long-term variability of moisture conditions was synchronous before the 1950s at a decadal scale (1791–1954). In northwestern margin of the EASM, most of all selected reconstructions had better consistency in low-frequency variation, especially during dry periods, indicating similar regional moisture variations and analogous modes of climate forcing on tree growth in the region.

Runoff is a major component of the water cycle, and its multi-scale fluctuations are important to water resources management across arid and semi-arid regions. This paper coupled the Distributed Time Variant Gain Model (DTVGM) into the Community Land Model (CLM 3.5), replacing the TOPMODEL-based method to simulate runoff in the arid and semi-arid regions of China. The coupled model was calibrated at five gauging stations for the period 1980–2005 and validated for the period 2006–2010. Then, future runoff (2010–2100) was simulated for different Representative Concentration Pathways (RCP) emission scenarios. After that, the spatial distributions of the future runoff for these scenarios were discussed, and the multi-scale fluctuation characteristics of the future annual runoff for the RCP scenarios were explored using the Ensemble Empirical Mode Decomposition (EEMD) analysis method. Finally, the decadal variabilities of the future annual runoff for the entire study area and the five catchments in it were investigated. The results showed that the future annual runoff had slowly decreasing trends for scenarios RCP 2.6 and RCP 8.5 during the period 2010–2100, whereas it had a non-monotonic trend for the RCP 4.5 scenario, with a slow increase after the 2050s. Additionally, the future annual runoff clearly varied over a decadal time scale, indicating that it had clear divisions between dry and wet periods. The longest dry period was approximately 15 years (2040–2055) for the RCP 2.6 scenario and 25 years (2045–2070) for the RCP 4.5 scenario. However, the RCP 8.5 scenario was predicted to have a long dry period starting from 2045. Under these scenarios, the water resources situation of the study area will be extremely severe. Therefore, adaptive water management measures addressing climate change should be adopted to proactively confront the risks of water resources.

 Since the Irtysh River flows through the important economic, ecological and social territories of China, Kazakhstan and Russia, the water quality issues growingly draw the attention of the water authorities from these countries. Therefore, a detailed study of the hydrochemical regime and toxicological indicators in Kazakhstan was carried out for understanding the regime dynamics of water quality and its affect factors. The combined assessment of maximum permissible concentration (MPC) of chemical components and biotesting method were proposed and performed for the study area. The results clearly showed that the concentrations of single chemical component at different locations are mostly under MPC standard in a basin scale. However, the watershed surface runoff and tributary stream flow from mining industry areas had high concentration of heavy metals and had significant impact on the water quality near Ust-Kamenogorsk. Furthermore, even the stream water generally meet MPC standard, the results of biotesting method show the toxicity level of water sample is lethal for the test objects of phytoplankton and Daphnia. The survival rates of most water samples are lower than 46.7%. Hereby, this study strongly suggests using combined water assessment methods to evaluate the water quality.

Arid and semi-arid areas are the most vulnerable regions to climate change. Clear understanding of the effects of climate change on ecosystems in arid and semi-arid regions and the ecosystem vulnerability is important for ecosystem management under the background of climate change. In this study, we conducted a vulnerability assessment on various ecosystems from 1982 to 2013 in western China with large areas of arid and semi-arid lands based on the Time-Integrated Normalized Difference Vegetation Index (TINDVI) data and climate data. The results indicated that grasslands were the most vulnerable ecosystem to climate change in western China, especially for those in Tibetan Plateau. Croplands in oases were not vulnerable to climate change compared to rain-fed croplands in semi-arid regions (e.g. Gansu and Inner Mongolia), which was attributed to the well-developed drip irrigation technology in oases. Desert and Gobi ecosystems were slightly vulnerable to climate change during the past several decades. The assessment results, as revealed in this study, can provide a reference for taking appropriate actions to protect the ecosystems in western China.

 The Loess Plateau of China has experienced extensive vegetation restoration in the past several decades, which leads to great changes in soil properties such as soil bulk, porosity, and organic matter with the vegetation restoration age. And these soil properties have great effect on the soil infiltration and soil hydraulic conductivity. However, the potential changes in soil hydraulic conductivity caused by vegetation restoration age have not been well understood. This study was conducted to investigate the changes in soil hydraulic conductivity under five grasslands with different vegetation restoration ages (3, 10, 18, 28 and 37 years) compared to a slope farmland, and further to identify the factors responsible for these changes on the Loess Plateau of China. At each site, accumulative infiltration amount and soil hydraulic conductivity were determined using a disc permeameter with a water supply pressure of –20 mm. Soil properties were measured for analyzing their potential factors influencing soil hydraulic conductivity. The results showed that the soil bulk had no significant changes over the initial 20 years of restoration (P>0.05); the total porosity, capillary porosity and field capacity decreased significantly in the grass land with 28 and 37 restoration ages compared to the slope farmland; accumulative infiltration amount and soil hydraulic conductivity were significantly enhanced after 18 years of vegetation restoration. However, accumulative infiltration amount and soil hydraulic conductivity fluctuated over the initial 10 years of restoration. The increase in soil hydraulic conductivity with vegetation restoration was closely related to the changes in soil texture and structure. Soil sand and clay contents were the most influential factors on soil hydraulic conductivity, followed by bulk density, soil porosity, root density and crust thickness. The Pearson correlation coefficients indicated that the soil hydraulic conductivity was affected by multiply factors. These results are helpful to understand the changes in hydrological and erosion processes response to vegetation succession on the Loess Plateau.

Analysis of spatial-temporal variations of desert vegetation under the background of climate changes can provide references for ecological restoration in arid and semi-arid areas. In this study, we used the Global Inventory Modeling and Mapping Studies (GIMMS) NDVI data from 1982 to 2006 and Moderate Resolution Imaging Spectroradiometer (MODIS) NDVI data from 2000 to 2013 to reveal the dynamics of desert vegetation in Hexi region of Northwest China over the past three decades. We also used the annual temperature and precipitation data acquired from the Chinese meteorological stations to analyze the response of desert vegetation to climatic variations. The average value of NDVImax (the maximum NDVI during the growing season) for desert vegetation in Hexi region increased at the rate of 0.65×10^–3/a (P<0.05) from 1982 to 2013, and the significant increases of NDVImax mainly appeared in the typical desert vegetation areas. Vegetation was significantly improved in the lower reaches of Shule and Shiyang river basins, and the weighted mean center of desert vegetation mainly shifted toward the lower reaches of the two basins. Almost 95.32% of the total desert vegetation area showed positive correlation between NDVImax and annual precipitation, indicating that precipitation is the key factor for desert vegetation growth in the entire study area. Moreover, the areas with non-significant positive correlation between NDVImax and annual precipitation mainly located in the lower reaches of Shiyang and Shule river basins, this may be due to human activities. Only 7.64% of the desert vegetation showed significant positive correlation between NDVImax and annual precipitation in the Shule River Basin (an extremely arid area), indicating that precipitation is not the most important factor for vegetation growth in this basin, and further studies are needed to investigate the mechanism for this phenomenon.

Croplands are often suffering from sand burial in dry regions of northern China. For studying this phenomenon, we carried out a case study of field experiment including four sand burial levels, i.e. shallow (1–3 cm), moderate (8–12 cm) and deep (15–20 cm) sand burials, and no sand burial (control, CK), in a typical agro-pastoral transitional zone in Naiman Banner of eastern Inner Mongolia. The aim of this study was to assess the impacts of sand burial on maize (Zea mays L.) productivity and the soil quality along a gradient of burial depths. Results showed that there was a strong negative effect of sand burial on maize productivity and soil quality, which significantly declined (P<0.05) under moderate and deep sand burial treatments. In comparison with the CK, the maize yield and above-ground biomass reduced by 47.41% and 39.47%, respectively. The soil silt and clay, soil water, soil organic carbon and total nitrogen contents under deep sand burial decreased by 67.85%, 40.32%, 86.52% and 82.11%, respectively, while microbial biomass carbon, microbial abundance and enzyme activity decreased by 89.78%, 42.28%–79.66% and 69.51%–97.71%, respectively. There was no significant effect on crop productivity and soil quality with shallow sand burial treatment. The correlations analysis showed that there was significant positive correlations of both maize yield and above-ground biomass with soil silt and clay, soil organic carbon and total nitrogen contents, pH, electrical conductivity, soil water content, microbial abundance and biomass and all tested soil enzyme activities. Stepwise regression analysis indicated that soil water and total nitrogen contents, urease, cellobiohydrolase and peroxidase activities were key determining factors for maize productivity. This combination of factors explains reason of the decreased maize productivity with deep sand burial. We found that degradation of cropland as a result of sand burial changed soil physical-chemical properties and soil enzyme activities in the plow layer, and decreased overall maize productivity. Furthermore, decreased soil enzyme activity was a better indicator to predict sandy cropland degradation.

The effects of long-term fertilization on pools of soil organic carbon (SOC) have been well studied, but limited information is available on the oxidizable organic carbon (OOC) fractions, especially for the Loess Plateau in China. We evaluated the effects of a 15-year fertilization on the OOC fractions (F1, F2, F3 and F4) in the 0–20 and 20–40 cm soil layers in flat farmland under nine treatments (N (nitrogen, urea), P (phosphorus, monocalcium phosphate), M (organic fertilizer, composted sheep manure), N+P (NP), M+N (MN), M+P (MP), M+N+P (MNP), CK (control, no fertilizer) and bare land (BL, no crops or fertilizer)). SOC content increased more markedly in the treatment containing manure than in those with inorganic fertilizers alone. F1, F2, F4 and F3 accounted for 47%, 27%, 18% and 8% of total organic carbon, respectively. F1 was a more sensitive index than the other C fractions in the sensitivity index (SI) analysis. F1 and F2 were highly correlated with total nitrogen (TN) and available nitrogen (AN), F3 was negatively correlated with pH and F4 was correlated with TN. A cluster analysis showed that the treatments containing manure formed one group, and the other treatments formed another group, which indicated the different effects of fertilization on soil properties. Long-term fertilization with inorganic fertilizer increased the F4 fraction while manure fertilizer not only increased labile fractions (F1) in a short time, but also increased passive fraction (F4) over a longer term. The mixed fertilizer mainly affected F3 fraction. The study demonstrated that manure fertilizer was recommended to use in the farmland on the Loess Plateau for the long-term sustainability of agriculture.

Being an important desert riparian forest in the lower reaches of the Heihe River Basin, Populus euphratica Oliv. forest functions as a natural barrier in maintaining and preserving the stability of local oases. Accordingly, accurately estimating the water use of P. euphratica is important for vegetation protection and water resource allocation. To date, little data are available for evaluating the hysteretic effects between sap flow and environmental variables, and for estimating the water use of desert riparian forest. In this study, we measured the sap flow velocity (V_s) of P. euphratica using the heat ratio method during the growing season of 2012. Based on the response of Vs to solar radiation (R_s) and vapor pressure deficit (VPD), we estimated the hourly V_s and daily V_s using the multivariable linear regression and a modified Jarvis-Stewart (JS) model, respectively. Hysteretic response of Vs to environmental variables was then evaluated using a sap flow model. We found the thresholds of Vs responses to R_s and VPD at both hourly and daily scales during the growing season, and successfully estimated the seasonal variations of hourly V_s and daily V_s using the JS model. At an hourly scale, the maximum Vs occurred earlier than the maximum VPD by approximately 0.5 h but later than the maximum R_s by approximate 1.0 h. At a daily scale, the maximum V_s lagged behind the maximum VPD by approximately 2.5 h while occurred earlier than the maximum R_s by approximately 2 h. However, hysteretic response of V_s was weakened when R_s and VPD were measured together using the JS model at both hourly and daily scales. Consequently, short-term and intensive field campaigns, where V_s and and environmental variables can be measured, may be used to estimate short-run sap flow and stand transpiration using only two environmental variables.

Black locust (Robinia pseudoacacia) is widely planted throughout the semi-arid Loess Plateau of China. The spatial distribution of this species at different ages is highly heterogeneous due to restoration and management practices. In this study, we aimed to compare the transpiration levels between different aged black locusts at the tree and stand scales, clarifying the physiological status of this species with different ages. Black locust trees with two representative age classes (12 and 28 years) were selected in the Yangjuangou catchment on the semi-arid Loess Plateau. Sap flux density (F_d) and environmental variables (solar radiation, air temperature, relative humidity and soil water content) were simultaneously monitored throughout the growing season of 2014. Tree transpiration (E_t) was the product of Fd and sapwood area (A_S), and stand transpiration (E_c) was calculated basing on the stand sap flux density (J_s) and stand total sapwood area (A_ST). Stomatal conductance (g_s) was measured in a controlled environment and hydraulic conductance was estimated using the relationship between transpiration rate and vapor pressure deficit (VPD). Our results showed that E_t and E_c were higher in the 28-year-old stand than in the 12-year-old stand. The g_s and hydraulic conductance of 28-year-old trees were also higher than those of 12-year-old trees, and the two parameters were thus the causes of variations in transpiration between different age classes. After rainfall, mean F_d increased by 9% in 28-year-old trees and by 5% in 12-year-old trees. This study thus suggests that stand age should be considered for estimating transpiration at the catchment and region scales in this area. These results provide ecophysiological evidences that the older black locust trees had more active physiological status than the younger ones in this area. These findings also provide basic information for the management of water resources and forests on the semi-arid Loess Plateau.

Provision of woodfuel is an important ecosystem service of dry forests and woodlands. However, charcoal production through selective logging of preferred hardwood species has the potential to alter the physiognomic composition of the residual or re-growth woodlands and may lead to their deterioration and degradation. This study, conducted through forest inventory in Mutomo District in Kenya, assessed the impact of charcoal production on unprotected dry woodlands in terms of tree density, targeted species basal area, species richness, evenness and Shannon diversity. The parameters of the disturbed woodlands were evaluated for significant differences with those of the neighbouring protected Tsavo East National Park, which served as a reference for an ecologically undisturbed ecosystem. By evaluating a consequence of tree harvesting for charcoal production, this study confirmed the overall significant differences between the protected and unprotected woodlands in all the tested parameters. To confirm if the differences in the land-covers of the woodlands had any influence on their degradation, all mentioned parameters were compared between the four differentiated classes and their respective control plots in the protected areas. At the “land-cover level”, the statistically significant difference in the basal area of tree species preferred for charcoal production between the protected and unprotected open trees confirms that the class with a high density of large mature trees is the prime target of charcoal producers. In addition, there seems to be a general trend of lower values of tree species richness, evenness and Shannon diversity for the unprotected woodlands subjected to charcoal production. On the other hand, the disturbed woodlands display the potential to recover through their comparably high saplings density. The findings make an important contribution to the discourse on the impact of charcoal production in dry woodlands, a topic that is highly controversial among researchers.

Different microhabitats along dune slope were found to affect the distribution of plant performances and soil properties in desertified ecosystems. However, the ground-active arthropod responses to rainfall-induced dune microhabitats were largely unknown in desertified regions. At dune top, midslope and bottom, ground-active arthropods were sampled by the method of pitfall traps in addition to the herbaceous and soil measurements during spring, summer and autumn from 2012 to 2013. Ground-active arthropod had a strong dynamics in time, seasonally and yearly in responses to rainfall-induced dune microhabitats and the variations were significant higher than those of soil and herbaceous properties. The abundance distribution of dominant taxa (i.e. Melolonthidae, Carabidae, Glaphyridae, Tenebrionidae and Formicidae families) among dune microhabitats was similar between seasons within the same year, whereas they differed markedly between two sampling years with varying rainfall patterns. A significant (P<0.05) difference in total abundance, taxa richness and Shannon index among dune microhabitats was found only in certain season time, particularly in 2013; however, no significant (P>0.05) differences were found among dune microhabitats when averaged on three sampling seasons within the year of either 2012 or 2013. In all, the taxonomical structure differed considerably from the community structure in ground-active arthropod response to rainfall-induced dune microhabitats. The spatial distribution of taxonomical groups among dune microhabitats was significantly affected by inter-annual rainfall changes, whereas that of community structure was affected by both intra- and inter-annual rainfall changes in desertified ecosystems.