Search

In the dryland and degraded regions, soil moisture is the primary factor determining ecological restoration. Proper land preparations and vegetation restoration can improve soil moisture and benefit land restoration. Identifying their effects on soil moisture is thus essential for developing suitable management strategies. In this study, four typical land preparation techniques (level ditches, fish-scale pits, zig terraces, and level benches) and non-native vegetation types (Prunus armeniaca, Platycladus orientalis, Pinus tabulaeformis, and Caragana microphylla) were planted on a semiarid loess hilly catchment, China. Soil moisture was monitored by a time-domain reflectometer, and its spatial-temporal variations were analyzed during the 2014-2015 growing seasons. The following results were captured. (a) Soil moisture was highest for the combination of P. orientalis and fish-scale pits (P. orientalis/fish-scale pits, 10.37%), followed by P. armeniaca/level ditches (10.23%), P. orientalis/zig terraces (9.67%), C. microphylla/level benches (8.62%), P. tabulaeformis/fish-scale pits (8.05%), and P. tabulaeformis/zig terraces (7.72%). (b) Fish-scale pits offered a better rainwater-harvesting capacity during the rainy months, whereas zig terraces had a higher water retention capacity under extremely dry conditions (as indicated by the temporal soil moisture variation and soil water retention curves). (c) P. tabulaeformis consumed more water than other vegetation types (26.7% lower than that of P. orientalis). (d) Soil moisture in shallower layers (0-80 cm) was more affected by land preparations and was more effected by vegetation types in deeper soil layers (80-180 cm). We thus suggest that fish-scale pits may be a better choice in wetter regions whereas zig terraces and P. orientalis are more suitable in the driest parts of the Loess Plateau to restore the fragile ecosystems. [ABSTRACT FROM AUTHOR]

A new method is proposed to establish the statistical fractal model for liver diseases classification. Firstly, the fractal theory is used to construct the high-order tensor, and then Generalized N-dimensional Principal Component Analysis (GND-PCA) is used to establish the statistical fractal model and select the feature from the region of liver; at the same time different features have different weights, and finally, Support Vector Machine Optimized Ant Colony (ACO-SVM) algorithm is used to establish the classifier for the recognition of liver disease. In order to verify the effectiveness of the proposed method, PCA eigenface method and normal SVM method are chosen as the contrast methods. The experimental results show that the proposed method can reconstruct liver volume better and improve the classification accuracy of liver diseases. [ABSTRACT FROM AUTHOR]