Marine Isotope Stage 11 (MIS 11; ca. 423–362 ka) is generally considered to be the best analogue for the present interglacial (Holocene), and investigation of it will improve our understanding of current climate variability and assist in predictions of future climate change. However, many recent studies primarily focus on the structure and duration of MIS 11. Little research has focused on climate warmth and stability recorded in the Chinese loess-paleosol sequences (LPS) during the S4 paleosol formation (equivalent to MIS 11). On the basis of previous work, this study presents a high-resolution record (ca. 75 a/cm) that spans from MIS 1 to MIS 15, as preserved in the thickest known Jingyuan loess section on the western Chinese Loess Plateau (CLP). This LPS is almost 165 m thick and was sampled from the upper part of L6 to the modern soil at 2-cm depth intervals. Measurements of magnetic susceptibility, mean grain size and >63 μm particle content, carbonate content, total organic carbon, and soil color of samples were made to reconstruct the paleoclimate variation, and a grain-size age model was used to constrain the chronological framework. The primary results show that a generally warm-humid climate dominated the S4 paleosol development, but the climate condition was extremely unstable during the whole of MIS 11. Two obviously different climate regimes controlled the MIS 11 climate variation: the early part of MIS 11 was extremely warm and stable, but the latter part was relatively cool (non-glacial) and unstable. This climate pattern was consistent with records on the central CLP and wavelet analysis suggested that it was forced by the 65°N insolation variability modulated by a quasi-100-ka cycle. In addition, a multi-proxy comparative study on the climate conditions during S0 to S5 paleosol development indicates that the period of S4 development might be the warmest interglacial of the past 650 ka. However, the climate condition during S4 development was not the most humid episode as recorded in Xifeng and Luochuan loess sections on the central CLP. On the contrary, it was drier than both the MIS 15 and the present interglacial on the western CLP, which is somewhat similar to the present climate pattern on the central CLP.

The infiltration of water into soil is one of the most important soil physical properties that affect soil erosion and the eco-environment, especially in the Pisha sandstone area on the Chinese Loess Plateau. We studied the one-dimensional vertical infiltration of water in three experimental soils, created by mixing Pisha sandstone with sandy soil, irrigation-silted soil, and loessial soil, at mass ratios of 1:1, 1:2, 1:3, 1:4, and 1:5. Our objective was to compare water infiltration in the experimental soils and to evaluate the effect of Pisha sandstone on water infiltration. We assessed the effect by measuring soil bulk density (BD), porosity, cumulative infiltration, infiltration rate and saturated hydraulic conductivity (Ks). The results showed that Pisha sandstone decreased the infiltration rate and saturated hydraulic conductivity in the three experimental soils. Cumulative infiltration over time was well described by the Philip equation. Sandy soil mixed with the Pisha sandstone at a ratio of 1:3 had the best water-holding capacity. The results provided experimental evidence for the movement of soil water and a technical support for the reconstruction and reclamation of mining soils in the Pisha sandstone area.

Soil respiration (SR) in crop field is affected by environmental factors, agronomic practices and crop types. To clarify how planting density affects the SR dynamics in switchgrass (Panicum virgatum L.) field on the semi-arid Loess Plateau, this research investigated diurnal and seasonal changes of soil respiration rate (R_S) under three different row-spacing treatments (20, 40 and 60 cm) in the fourth growing year of switchgrass. Results showed that R_S presented a pronounced seasonality under all row-spacing treatments. The highest daily average R_S values appeared in August, while the lowest (P<0.05) were in September for each row-spacing. Diurnal variations of R_S exhibited single-peak curves in each month. Daily average R_S increased significantly as row-spacing enlarged during May and August but there was no significant difference among row-spacing treatments in September. Soil water storage in the depth of 0–100 cm had no significant difference (P>0.05) among the row-spacing treatments, and similar results were found for soil temperature in 0–15 cm soil depth. Soil respiration temperature sensitivity (Q10) values were 1.0–3.7 during the growing months, which were strongly correlated with air temperature in May and June and the soil temperature at 15 cm depth in August. Higher aboveground biomass production and lower RS in most growth months indicated that 20 cm row-spacing treatment was beneficial for increasing the carbon fixation in switchgrass field. Results also implied that it is necessary to take into account the in?uence of phenology and root growth of switchgrass on soil respiration for accurately evaluating the carbon cycle in the region.

Soil organic matter content in water-stable aggregates (WSA) in the arid ecosystems (abandoned agricultural lands especially) of China is poorly understood. In this study, we examined the WSA sizes and stability, and soil organic carbon (OC) and nitrogen (N) contents in agricultural lands with abandonment ages of 0, 3, 12, 20, 30 and 40 years, respectively, in the Minqin Oasis of Northwest China. The total soil OC and N contents at depths of 0–20, 20–40 and 40–60 cm in abandoned agricultural lands were compared to those in cultivated land (the control). Agricultural land abandonment significantly (P<0.05) influenced the distribution of MWD (mean weight diameter), and OC and N contents. There were significant increases in MWD and the proportion of macroaggregates (sizes >0.25 mm) as the age of agricultural land abandonment increased. The effect of abandonment ages of agricultural lands on MWD was determined by the changes of OC and N accumulation in WSA sizes >2 mm. The total OC and N contents presented a stratification phenomenon across soil depths in this arid ecosystem. That is, both of them decreased significantly at depths of 0–20 and 40–60 cm while increased at the depth of 20–40 cm. The WSA sizes <0.053 mm had the highest soil OC and N contents (accounting for 51.41%–55.59% and 42.61%–48.94% of their total, respectively). Soil OC and N contents in microaggregates (sizes 0.053–0.25 mm) were the dominant factors that influenced the variations of total OC and N contents in abandoned agricultural lands. The results of this study suggested that agricultural land abandonment may result in the recovery of WSA stability and the shifting of soil organic matter from the silt+clay (<0.053 mm) and microaggregate fractions to the macroaggregate fractions. However, agricultural land abandonment did not increase total soil OC and N contents in the short-term.

Grassland ecosystem is an important component of the terrestrial carbon cycle system. Clear comprehension of soil organic carbon (SOC) storage and potential of grasslands is very important for the effective management of grassland ecosystems. Grasslands in Inner Mongolia have undergone evident impacts from human activities and natural factors in recent decades. To explore the changes of carbon sequestration capacity of grasslands from 2000 to 2012, we carried out studies on the estimation of SOC storage and potential of grasslands in central and eastern Inner Mongolia, China based on field investigations and MODIS image data. First, we calculated vegetation cover using the dimidiate pixel model based on MODIS-EVI images. Following field investigations of aboveground biomass and plant height, we used a grassland quality evaluation model to get the grassland evaluation index, which is typically used to represent grassland quality. Second, a correlation regression model was established between grassland evaluation index and SOC density. Finally, by this regression model, we calculated the SOC storage and potential of the studied grasslands. Results indicated that SOC storage increased with fluctuations in the study area, and the annual changes varied among different sub-regions. The SOC storage of grasslands in 2012 increased by 0.51×10¹² kg C compared to that in 2000. The average carbon sequestration rate was 0.04×10¹² kg C/a. The slope of the values of SOC storage showed that SOC storage exhibited an overall increase since 2000, particularly for the grasslands of Hulun Buir city and Xilin Gol League, where the typical grassland type was mainly distributed. Taking the SOC storage under the best grassland quality between 2000 and 2012 as a reference, this study predicted that the SOC potential of grasslands in central and eastern Inner Mongolia in 2012 is 1.38×10¹² kg C. This study will contribute to researches on related methods and fundamental database, as well as provide a reference for the protection of grassland ecosystems and the formulation of local policies on sustainable grassland development.

The plant ecosystems are particularly sensitive to climate change in arid and semi-arid regions. However, the responses of vegetation dynamics to climate change in Central Asia are still unclear. In this study, we used the normalized difference vegetation index (NDVI) data to analyze the spatial-temporal changes of vegetation and the correlation of vegetation and climatic variables over the period of 1982–2012 in Central Asia by using the empirical orthogonal function and least square methods. The results showed that the annual NDVI in Central Asia experienced a weak increasing trend overall during the study period. Specifically, the annual NDVI showed a significant increasing trend between1982 and 1994, and exhibited a decreasing trend since 1994. The regions where the annual NDVI decreased were mainly distributed in western Central Asia, which may be caused by the decreased precipitation. The NDVI exhibited a larger increasing trend in spring than in the other three seasons. In mountainous areas, the NDVI had a significant increasing trend at the annual and seasonal scales; further, the largest increasing trend of NDVI mainly appeared in the middle mountain belt (1,700–2,650 m asl). The annual NDVI was positively correlated with annual precipitation in Central Asia, and there was a weak negative correlation between annual NDVI and temperature. Moreover, a one-month time lag was found in the response of NDVI to temperature from June to September in Central Asia during 1982–2012.

Growth of annual plants in arid environments depends largely on rainfall pulses. An increased understanding of the effects of different rainfall patterns on plant growth is critical to predicting the potential responses of plants to the changes in rainfall regimes, such as rainfall intensity and duration, and length of dry intervals. In this study, we investigated the effects of different rainfall patterns (e.g. small rainfall event with high frequency and large rainfall event with low frequency) on biomass, growth characteristics and vertical distribution of root biomass of annual plants in Horqin Sandy Land, Inner Mongolia of China during the growing season (from May to August) of 2014. Our results showed that the rainfall patterns, independent of total rainfall amount, exerted strong effects on biomass, characteristics of plant growth and vertical distribution of root biomass. Under a constant amount of total rainfall, the aboveground biomass (AGB), belowground biomass (BGB), plant cover, plant height, and plant individual and species number increased with an increase in rainfall intensity. Changes in rainfall patterns also altered the percentage contribution of species biomass to the total AGB, and the percentage of BGB at different soil layers to the total BGB. Consequently, our results indicated that increased rainfall intensity in future may increase biomass significantly, and also affect the growth characteristics of annual plants.

Dew is an important source of water which significantly influences the physiological status of vegetation and the microclimate environment. For quantifying the characteristics of dew events and analyzing the underlying mechanism of dew formation in different ecosystems, we measured, based on the flux-profile method, the amount, frequency and duration of dew events in two croplands, an arid artificial oasis cropland in Zhangye, Gansu province and a sub-humid cropland in Luancheng, Hebei province in China. The results showed that dew events were observed in a total of 69 days in Zhangye, which accounted for 59% of the growing season (from 28 May to 21 September, 2012), while 128 days in Luancheng, which accounted for 79% of the growing season (from 5 April to 13 September, 2008). The frequencies of dew events were 2.8 and 2.4 times of those of precipitation in Zhangye and Luancheng, respectively. In addition, the dew amount reached up to 9.9 and 20.2 mm in Zhangye and Luancheng, which accounted for 9.5% and 4.1% of precipitation, respectively. The average amount of dew was 0.14 and 0.16 mm/night in Zhangye and Luancheng, respectively and the duration of dew events ranged from 0.5 to 12.0 h in the two study sites. Dew amounts were associated with the gradient of atmospheric water vapor concentration and dew duration (P<0.001) in both the two sites. The result implies that dew events play a more important role in crop growth in arid areas in comparison to sub-humid areas considering the dew occurrence frequency and the amount per night.

The transpiration rate of plant is physically controlled by the magnitude of the vapor pressure deficit (VPD) and stomatal conductance. A limited-transpiration trait has been reported for many crop species in different environments, including Maize (Zea mays L.). This trait results in restricted transpiration rate under high VPD, and can potentially conserve soil water and thus decrease soil water deficit. However, such a restriction on transpiration rate has never been explored in maize under arid climatic conditions in northwestern China. The objective of this study was to examine the transpiration rate of field-grown maize under well-watered conditions in an arid area at both leaf and whole plant levels, and therefore to investigate how transpiration rate responding to the ambient VPD at different spatial and temporal scales. The transpiration rates of maize at leaf and plant scales were measured independently using a gas exchange system and sapflow instrument, respectively. Results showed significant variations in transpiration responses of maize to VPD among different spatio-temporal scales. A two-phase transpiration response was observed at leaf level with a threshold of 3.5 kPa while at the whole plant level, the daytime transpiration rate was positively associated with VPD across all measurement data, as was nighttime transpiration response to VPD at both leaf and whole plant level, which showed no definable threshold vapor pressure deficit, above which transpiration rate was restricted. With regard to temporal scale, transpiration was most responsive to VPD at a daily scale, moderately responsive at a half-hourly scale, and least responsive at an instantaneous scale. A similar breakpoint (about 3.0 kPa) in response of the instantaneous leaf stomatal conductance and hourly canopy bulk conductance to VPD were also observed. At a daily scale, the maximum canopy bulk conductance occurred at a VPD about 1.7 kPa. Generally, the responsiveness of stomatal conductance to VPD at the canopy scale was lower than that at leaf scale. These results indicate a temporal and spatial heterogeneity in how maize transpiration responses to VPD under arid climatic conditions. This could allow a better assessment of the possible benefits of using the maximum transpiration trait to improve maize drought tolerance in arid environment, and allow a better prediction of plant transpiration which underpin empirical models for stomatal conductance at different spatio-temporal scales in the arid climatic conditions.

 community structure responds strongly to anthropogenic disturbances, which greatly influence community stability. The changes in community structure, aboveground biomass (AGB), biodiversity and community stability associated with different management practices were studied with a three-year field investigation in a temperate steppe of Inner Mongolia, China. The species richness, Shannon-Wiener index, evenness, plant functional type abundance, AGB, temporal community stability, summed covariance, scaling coefficient and dominant species stability were compared among areas subjected to long-term reservation (R), long-term grazing (G), mowing since enclosure in 2008 (M) and grazing enclosure since 2008 (E). Site R had higher perennial grass abundance and lower species richness than sites G, M and E, although the AGB was not significantly different among the four sites. The species structure varied from a single dominant species at site R to multiple dominant species at sites G, M and E. The long-term reservation grassland had lower biodiversity but higher stability, whereas the enclosed grassland with/without mowing had higher biodiversity but lower stability. Different stability mechanisms, such as the compensatory dynamics, mean-variance scaling and dominant species stability were examined. Results showed that community stability was most closely related to the relative stability of the dominant species, which supports the biomass ratio hypothesis proposed by Grime.

Species richness is an important indicator of species diversity. Different sampling intensities will very likely produce different species richness values. Substantial efforts have already been made to explicitly quantify the spatial variability of soil properties in different ecosystems. However, concerns still remain on how to characterize the effect of different sampling intensities on plant species richness within a given region. This study characterized the spatial variability of plant species richness and the species distribution pattern in a 25-hm² sand dune plot in northeastern Inner Mongolia, China by using an intense sampling method (n=10,000). We also evaluated the overall effect of information loss associated with the spatial variability and distribution patterns of species richness under various scenarios of sampling intensities (n=10,000 to 289). Our results showed that semi-variograms of species richness were best described by the spherical and exponential models. As indicated by the nugget/sill ratio, species richness was different in terms of the strength of the spatial relationship. The different spatial metrics of species richness with increasing sampling intensities can represent different responses of the spatial patterns when compared with the reference set (n=10,000). This study indicated that an appropriate sampling intensity should be taken into account in field samplings for evaluating species biodiversity properly. A sampling intensity of n>2,500 for species richness yielded satisfactory results to resemble the spatial pattern of the above-quantified reference set (n=10,000) in this sand dune region of China.

The acclimatization of plant xylem to altered environmental conditions has attracted considerable attention from researchers over several decades. Plants growing in natural environments must seek a balance between water uptake and the water loss of leaves from evaporation. Thus, the adaptation of xylem to different soil textures is important in maintaining plant water balance. In this study, we investigated the xylem changes of cotton (Gossypium herbaceum L.) xylem in sandy, clay and mixed soils. Results showed that soil texture had a significant effect on xylem vessel diameter and length of stems and roots. Compared with G. herbaceum growing in the clay soil, those plants growing in the sandy soil developed narrower and shorter xylem vessels in their roots, and had a higher percentage of narrow vessels in their stems. These changes resulted in a safer (i.e. less vulnerable to cavitation), but less-efficient water transport system when soil water availability was low, supporting the hydraulic safety versus efficiency trade-off hypothesis. Furthermore, in sandy and mixed soils, the root:shoot ratio of G. herbaceum increased twofold, which ensures the same efficiency of leaves. In summary, our finding indicates that the morphological plasticity of xylem structure in G. herbaceum has a major role in the acclimatization of this plant species to different soil textures.

Weedy plants affect the biodiversity and ecosystem function as well as the crop and fodder plant production. However, adequate management requires detailed knowledge of the taxonomic identity of these plants. Here, we focused on a hemiparasitic Pedicularis species (Orobanchaceae), which occurs at high densities and results in significant biomass reductions in forage grasses in Bayanbulak Grassland of Xinjiang. The identity of this target species is not clear, with conflicting reports in publications and in herbarium collections. Hence, clear and management-relevant information on demography and reproductive ecology is difficult to be obtained from the literature. Therefore, we analyzed field and archival materials collected from Xinjiang in order to clarify the identity and distribution of the target species. Morphological analyses suggested that the populations at Bayanbulak Grassland should be Pedicularis kansuensis Maxim. rather than P. verticillata L. which has been accepted in the available literature. Phylogenetic analysis with a combination of three barcodes (matK, rbcL and trnH-psbA) uniting a clade of P. kansuensis and individuals from Bayanbulak Grassland populations with 100% bootstrap support, confirmed the target species to be P. kansuensis. Anatomical investigations and field observations showed that the target species is an annual or biennial herb, which also fits with the life cycle as P. kansuensis. Based on archive material and field observations, we verified that the distribution of P. kansuensis is mainly concentrated in the Tianshan Mountains of Xinjiang.

Spatio-temporal variations of vegetation phenology, e.g. start of green-up season (SOS) and end of vegetation season (EOS), serve as important indicators of ecosystems. Routinely processed products from remotely sensed imagery, such as the normalized difference vegetation index (NDVI), can be used to map such variations. A remote sensing approach to tracing vegetation phenology was demonstrated here in application to the Inner Mongolia grassland, China. SOS and EOS mapping at regional and vegetation type (meadow steppe, typical steppe, desert steppe and steppe desert) levels using SPOT-VGT NDVI series allows new insights into the grassland ecosystem. The spatial and temporal variability of SOS and EOS during 1998–2012 was highlighted and presented, as were SOS and EOS responses to the monthly climatic fluctuations. Results indicated that SOS and EOS did not exhibit consistent shifts at either regional or vegetation type level; the one exception was the steppe desert, the least productive vegetation cover, which exhibited a progressive earlier SOS and later EOS. Monthly average temperature and precipitation in preseason (February, March and April) imposed most remarkable and negative effects on SOS (except for the non-significant impact of precipitation on that of the meadow steppe), while the climate impact on EOS was found to vary considerably between the vegetation types. Results showed that the spatio-temporal variability of the vegetation phenology of the meadow steppe, typical steppe and desert steppe could be reflected by the monthly thermal and hydrological factors but the progressive earlier SOS and later EOS of the highly degraded steppe desert might be accounted for by non-climate factors only, suggesting that the vegetation growing period in the highly degraded areas of the grassland could be extended possibly by human interventions.