Study Of Soil Anti-erodibility Of Different Land Use Types In Karst Valley Region

Soil anti-erodibility indicator is an important parameter in simulating of soil erosion process and erosion intensity in soil erosion prediction model, which is also currently the hot issues and frontier scientific issues of soil erosion mechanism research. In this study, five typical land use types including slope farmland, abandoned farmland, grassland, shrub land and woodland in karst plateau valley region were studied as research objects, the study was conducted by combining field sampling and laboratory analysis method analysed the variation characteristics of soil texture and structure, the stability and fragmentation mechanism of aggregate, carbon sequestration mechanism of water stability of aggregate of different land use types; established evaluation index system for soil corrosion resistance and determined key factors used for denote soil erosion durability in Karst area through sensitivity analysis and dimensionality reduction; and explored the applicability of existing erodibility K value model for Karst area in China, to provide a scientific basis for studying soil erosion process in Karst area, establishing soil erosion prediction model and regional rocky desertification control and ecological restoration. The main results and conclusions of this research are as follows:(1) The characteristics of key anti-erodibility indicators of different land use types were studied. Woodland and shrub land have the better water-holding capacity among five kinds of land use types, while slope farmland has the worst water retention capacity. The total soil porosity relatively reduced and the capillary porosity as well as soil water retention enhanced during plant rehabilitation. Revegetation promoted the accumulation of soil organic matter and content of soil total nutrients(Total N, P and K) were the largest in woodland. Cation exchange capacity(CEC) of soil layers was the largest in woodland and which illustrated that woodland have a stronger function in storage and release of soil available nutrients. Soil particle size distribution was highest in coarse silt sizes(0.05~0.01 mm) and contents of soil coarse sand and fine sand relatively reduced along the vegetation restoration which indicating that soil texture has shifted from coarse and ossification to refine texture. Contents of 1~0.25 mm and 0.25~0.005 mm microaggregate were the largest in soil layers of different land use types, and restoration significantly increased soil aggregate reunion status and decreased dispersion rate. Woodland has the greatest aggregate reunion degree and reunion status, while slope farmland has the lowest among five kinds of land use types.(2) Characteristics of soil aggregate stability of different land use types were discussed by Yoder wet sieving method. Air-dried aggregates measured by Yoder wet sieving method were maximum in >5 mm aggregates and minimum in <0.25 mm aggregates, while water-stable aggregates were maximun in <0.25 mm aggregates. The content of water-stable aggregates which <0.25 mm significantly reduced and those which are more larger in grain size significantly increased due to revegetation and which promoted the formation of soil aggregate structure. Woodland has the greatest aggregate stability and soil anti-erodibility of different land use types, followed by grassland and shrub land, and slope farmland has the worst aggregate stability. Revegetation promoted the increase of SOC in each soil layer, and contribution rate of SOC associated in water-stable aggregate to total SOC was the largest in aggregate <0.25 mm of each soil layer along vegetation restoration.(3) Aggregate broken mechanism and break sensitivity to different crushing actions of different land use types were revealed by Le Bissonnais method. Three treatments of LB method showed that fast wetting(FW) treatment has a higher crushing degree of grain size aggregates and 0.5~0.2 mm aggregates has the maximum content. Aggregates size mainly concentrated in 2~1 mm and 1~0.5 mm under slow wetting(SW) treatment. However, wet stiring(WS) has small crushing effect on the aggregates considering that aggregate size mainly concentrated in 5~2 mm and 2~1 mm. Which illustrated that soil aggregates would have destructive effect when suffer rainstorm of flood irrigation, while aggregates have small destructive effect in continuous rain or mechanical tillage in karst region. The break of aggregate in karst region mainly due to internal dissipation of soil aggregates, followed by swelling and crushing effect of clay while mechanical vibration has less effect on aggregate break. Slope farmland was sensitive to aggregate dissipation and abandoned farmland is sensitive to mechanical crushing effect, however, woodland, shrub land and grassland could maintain high aggregate stability when suffer dissipation, clay swelling and mechanical crushing effect. Index of aggregate stability between Yoder wet sieving method and Le Bissonnais method have a good correlation, >0.25 mm aggregate of Yoder wet sieving has good aggregate information carrying capacity and was more sensitive to aggregate break. In addition, relative higher level of organic matter and clay content can effectively resist the impact of soil dissipation and clay swelling on destruction of soil aggregates(4) A preliminary anti-erodibility evaluation index system was established and the best key factors to evaluate anti-erodibility were determined and established comprehensive index model of anti-erodibility evaluation. 26 indicators belong to five categories including inorganic clay category, water-stable aggregate category, microaggregate category, organic colloids category and soil mechanics category were selected for correlation analysis, the results showed that anti-erodibility index of both same category and different category have certain correlation,and showed that a prominent problem that erodibility information was overlapped among different indicators. According to sensitivity analysis, anti-erodibility evaluation index was divided into four sensitivity levels including strong sensitivity level, high-level sensitive, medium sensitivity and weak level sensitivity. Principal component analysis to simplify anti-erodibility evaluate indicators, and the result showed that the cumulative variance contribution rate of seven main components reached 84.061%,soil aggregate category and SOC have the largest impact on soil anti-erodibility and followed by micro-aggregates category and clay category among the seven main components. Comprehensive score calculated by principal component analysis indicated that woodland occupied the highest comprehensive score,followed by grassland, shrubland and abandoned farmland, while slope farmland was the lowest. The best seven indicators(x1~x7) in karst gorge region were determined by comprehensive results of correlation analysis, sensitivity analysis and dimension reduction analysis, which include WSA>0.5mm, WSA>0.25 mm, Dwet, Clay <0.002 mm, soil organic matter, aggregation degree and MWDwet. Meanwhile, anti-erodibility evaluation equation was established with the selected factors.(5) Applicability of different erodibility estimate models to evaluate K Values in karst region were analyzed and discussed, also, the existed erodibility K value models were examinated and corrected by mathematical method. Erodibility K values estimated by USLE, RUSLE2, EPIC, Shirazi and Torri model were quite different, and the estimated values by USLE model is maximum and that by Shirazi model is minimum. USLE and RUSLE2 have the same law in estimating erodibility K values, USLE model "overestimated" the K values on the whole and EPIC model "undervalued" erodibility K values. The estimated value of Torri model has a significantly positive correlation with that of RUSLE2 model. Uncertainty analysis of different estimate modes performed as KTorri < KRUSLE < KUSLE < KEPIC < KShirazi, which indicated that Torri model owned the highest precision in evaluating erodibility K values among five different models. Considering the lake of long term soil erosion monitoring data in karst area, Torri model was the preferred method to estimate erodibility K values when there lack background information. The coefficient between observed soil erodibility K values and K values predicted by optimized Torri model increased, indicating that prediction accuracy improved. The estimation accuracy of corrected USLE, RUSLE2 and EPIC model referenced to Torri model have been improved. Evaluation results of AEI model proposed in this research were significantly negatively correlated to K values calculated by Torri model, which indicating that AEI model has strong applicability in karst region. Soil erodibility K values in each layer during different vegetation restoration stages expressed as slope farmland > abandoned farmland > grassland > shrub land > woodland, which means revegetation have significantly decreased soil erodibility K value in each layer and enhanced soil anti-erodibility..