Distribution Of Soil Water And Its Effect On Carbon Process In Grassland Ecosystems On The Typical Loess Plateau

Soil water content is the key factor for vegetation restoration and ecological constructionon the Loess Plateau and has important influence on the critical ecological processes. Thedistribution of soil water is combined result of climatic, soil, vegetation, and topographicfactors, which can reflect the evolutional process. Understanding the distributioncharacteristics of soil water and the relationships with environmental factors at multiple scalescan therefore reveal the spatial-temporal evolutional patterns of soil water and the drivingmechanisms, thus this would provide scientific evidence for the rational management andsustainable utilization of soil water resource. The grassland on the Loess Plateau is animportant component of China’s grassland; therefore, understanding the effects of soil wateron carbon processe in grassland ecosystems will provide a scientific evidence to reveal theinteractive relationships between water and vegetation. The thesis clusters full-fledged fieldinvestigations and laboratory measurements. Along with the field works, the thesis presentsthe study of spatiotemporal distribution of soil water and its relationships to environmentalfactors at reginal, transect, and slope scales in the typical Loess Plateau. Following, the thesisanalyzes the effects of soil water on the net primary productivity, soil respiration, distributionof soil carbon and nitrogen, and herbaceous species diversity in the grassland ecosystems. Themain conclusions are congregated as the following:1. Soil water content within profiles across the typical Loess Plateau exhibited obviousspatial distribution patterns. There were decreasing trend in soil water content from thesoutheast to northwest in the horizontal direction. Mean soil water content within the0~5mprofile generally decreased, and then increased with increasing soil depth. Land use had asignificant impact on both the amount and profile distribution of soil water, with the mean soilwater content followed the sequence of cropland> forestland> grassland. The mean soilwater storage within the0~5m profile across the typical Loess Plateau is719mm. The rangefor soil water generally decreased with increasing soil depth, indicating that the spatialdependence of soil water increased with increases in the soil depth. Dominant factorsaffecting the spatial patterns in soil water across the typical Loess Plateau were precipitation,soil porosity, and slope gradient. Soil water distributions along the regional transect demonstrated an obvious decreasing trend from the south to the north. The mean soil waterstorage within the0~5m profile for the regional transect is735mm. Temporal variation inthe soil water content within the soil profile decreased with increasing soil depth, while thespatial variation increased along the south-north transect. Dominant factors affecting thespatial patterns in soil water for the regional transect were aridity, clay content, elevation, andslope gradient.2. The dried soil layers (DSL), occurred on the typical Loess Plateau demonstrated anobvious spatial distribution pattern. The thickness of DSL (DSLT) were considered to bemuch thicker in west and northwest areas, while much thinner in southeast areas. The meansoil water content for DSL (DSL-SWC) distributions demonstrated an obvious decreasingtrend from the southeast to the northwest. Land use types had a significant impact on soildesiccation, with the extent of which followed the sequence of grassland> forestland>cropland. The distribution patterns of DSL across the typical Loess Plateau were thecombined results of regional large-scale and local small-scale factors. Temporal stability ofDSL for the south-north transect was ascertained. The temporal stability of DSL was strongerwith decreases in the DSL-SWC and the formation depth of DSL (DSLFD). The drier theclimate and the lower the soil water retention, the stronger the temporal stability ofDSL-SWC and the easier the formation of permanent DSL.3. The temporal changes at the hill-slope scale mean soil water storage (SWS) decreasedwith increasing soil depth, while the spatial variation increased, indicating that the degree oftemporal-spatial variation of SWS were strongly dependent on sampling depth. The temporalstability of SWS within the soil profiles were strong on the typical watershed hill-slopes, andthe number of time-stable locations increased with increasing soil depth on both thehill-slopes, indicating that the SWS intended to be more temporally stable in deeper soil. Theselected representative time-stable locations could directly estimate the mean SWSs well onthe hillslopes (R2≥95%). Soil texture, organic carbon content, elevation, and properties ofthe vegetation affected the temporal stability of SWS for various layers; however these effectsdiffered between the two hill-slopes, implying that the dependence of SWS temporal stabilityon soil, plants, topography and on the scale of the hill-slope may differ at different hill-slopes.4. Precipitation is the dominant factor regulating the distribution of aboveground netprimary productivity (ANPP) of grassland on the Loess Plateau at the regional scale. Theresponse sensitivity of ANPP to the changes in precipitation differed among differentgrassland types, with the sensitivity following the sequence of meadow grassland> typicalgrassland> scrub grassland> desert grassland. A significant negative relationship betweenANPP and precipitation seasonal distribution (PSD) during growing seasons (May-September) were obtained, thus implying that more homogeneous precipitation distribution pattern favorshigher ANPP on the Loess Plateau; especially in areas with precipitation between340and440mm. Soil water storage also had a significant positive effect on ANPP and the degree ofpositive effect decreased with an increasing soil depth.5. Both surface soil water content (SWC) and NPP in grassland on the typical watershedhill-slope had strong spatial autocorrelation structures. Besides, the spatial cross-correlationstructures between SWC and NPP and other environmental variables were significant.Saturated soil hydraulic conductivity (Ks) based state-space model showed the bestperformance for simulating SWC, therefore they can be used to support SWC distributionestimations. State-space models that included clay content+Olsen phosphorus or Olsenphosphorus+NO3-N content could explain all the variations of NPP, hence they can be usedto simulate the NPP distribution on the hill-slope scale.6. Changes in soil water have significant effect on both carbon sequestration andemission in grassland ecosystems. Adding water significantly stimulated ANPP and fine rootbiomass for bunge needlegrass (BNE), alfalfa (ALF), fallow land for the recovery of naturalvegetation (DALF), and degraded alfalfa (DALF) communities, indicating that by addingwater is important for grassland ecosystems during the early growing stage. Water additionconsistently increased soil respiration rates across the four communities; which however, didnot change their seasonal variations. The addition of water significantly increased the Q10ofsoil respiration for ALF, NAF and DALF, but it had no significant effect on the Q10for BNE.The levels of δ13C and δ15N increased gradually with the succession of the ALF communityand decreased with the succession to the BNE community, while indicating that the dynamicsof the δ13C and δ15N isotopic signatures differed within the different re-vegetation stages ofsuccession. A pronounced inverse relationship developed between the soil water storage andthe natural abundance of δ13C/δ15N, implying that the primary processes in the arid grasslandecosystems are mainly controlled by water.7. Botanical diversity responded differently to the environmental variables in differentplant communities. Dominant factors affecting the distributions in herbaceous speciesdiversity were soil moisture and Ks in restored grasslands. Shrub density, slope gradient, andelevation accounted for the largest variations in species diversity in restored shrub-lands. Themaintenance of a moderate density of shrubs, taking conditions of soil and topographic intoaccount, soil moisture in particular, may therefore help to conserve biodiversity in thenorthern Loess Plateau.The above conclusions indicate that climate, soil, and topography are the major factorscontrolling soil water distribution in the typical Loess Plateau on the regional and transect scales. On and at the hill-slope scale, however, soil, topography, and vegetation are the majorfactors influencing the temporal-spatial changes in soil water content. By studying thedistribution patterns within different environmental factors, the thesis reveals that soilmoisture spatiotemporal processes and the driving mechanisms provide scientific evidencesfor the simulation; prediction in soil moisture dynamics and changes in the soil and waterenvironmental quality at multiple scales. Changes in soil moisture have significant effect ongrassnalnd ecosystem productivity, carbon and nitrogen cycles and species diversitydistribution. By studying the responses of fundamental ecological processes as well asvegetation biodiversity to alterations in soil moisture can help to revealing the interactionsbetween water and vegetation. Therefore, the thesis provides scientific support for thevegetation restoration and ecological construction in fragile ecosystems.