3.1. SWC variability with landcover types
SWC data collected from 10 sampling plots across different landcover types are shown in Fig. 3 and Table 2. Data for each plot conform to normal distribution, with the absolute value of Skewness andKurtosis of < 1. There were some differences (from 9.01 to 16.55 cm3 cm-3) in surface SWC corresponding to various landcover types, but most were not significant. For example, brushland (15.81 cm3cm-3) exhibited significant differences compared with planted forest (12.28 cm3 cm-3) and grassland (13.30 cm3 cm-3) (plot #6 in Fig. 3), and grassland (13.35 cm3cm-3) also exhibited unnoteworthy compared with planted and mixed forests (10.53 cm3cm-3) (plot #9 in Fig. 3). The differences between natural (16.5 cm3 cm-3) and planted forests (11.82 cm3 cm-3) were notable (mean value between plot #10 and plot #2).
To further quantify SWC differences among the landcover types, SWC was summarized using SPSS by landcover type (Fig. 4). Results showed a decreasing trend from natural forest, brushland, grassland and planted forest to mixed forest, representing changes in SWC over several decades of plantation succession in this semi- and arid region. SWC in planted forest and mixed planted forest was lower than that in the brush and grass areas, with the only significant difference found between the natural forest and other landcover types.
Furthermore, standard deviations in natural forest (plot #10) and planted forest (plot #1) were 6.16 cm3cm-3and 3.5 cm3cm-3, respectively; while in brushland and grassland (plot #8 and #3), they were 1.18 cm3cm-3 and 2.62 cm3cm-3, respectively (Table 2). A larger standard deviation indicated that the forest had greater water storage capacity compared to other two landcover types.