首页 > > > 草被覆盖的影响在沙质土壤水文物理参数和水流的异质性.pdf

草被覆盖的影响在沙质土壤水文物理参数和水流的异质性.pdf

草被覆盖的影响在沙质土壤水文物理参数和水流的异质性.pdf

上传者: 紫幽 2013-02-23 评分1 评论0 下载0 收藏10 阅读量979 暂无简介 简介 举报

简介:本文档为《草被覆盖的影响在沙质土壤水文物理参数和水流的异质性pdf》,可适用于个人文书领域,主题内容包含Pedosphere():–,ISSNCNPcSoilScienceSocietyofChinaPublishedbyElsevierBVandSc符等。

Pedosphere 21(6): 719–729, 2011 ISSN 1002-0160/CN 32-1315/P c 2011 Soil Science Society of China Published by Elsevier B.V. and Science Press Grass Cover Influences Hydrophysical Parameters and Heterogeneity of Water Flow in a Sandy Soil1 L. LICHNER1,2, D. J. ELDRIDGE2, K. SCHACHT3, N. ZHUKOVA4, L. HOLKO1, M. SˇIR5 and J. PECHO6 1Institute of Hydrology, Slovak Academy of Sciences, Racianska 75, 83102 Bratislava (Slovakia) 2Department of Environment Climate Change and Water, School of Biological, Earth and Environmental Sciences, Unive- rsity of New South Wales, Sydney, NSW 2052 (Australia) 3Geography Department, Ruhr-University Bochum, Universitaetsstrasse 150, 44801 Bochum (Germany) 4M. Nodia Institute of Geophysics, 1 Alexidze str., 0193 Tbilisi (Georgia) 5Institute of Hydrodynamics, Academy of Sciences of the Czech Republic, Pod Patankou 30/5, 16612 Prague (Czech Repu- blic) 6Institute of Atmospheric Physics, Academy of Sciences of the Czech Republic, 1401 Bocni II, 14131 Prague (Czech Republic) (Received April 26, 2011; revised September 23, 2011) ABSTRACT Vegetation cover has a major effect on water flow in soils. Two sites, separated by distance of about 50 m, were selected to quantify the influence of grass cover on hydrophysical parameters and heterogeneity of water flow in a sandy soil emerging during a heavy rain following a long hot, dry period. A control soil (pure sand) with limited impact of vegetation or organic matter was obtained by sampling at 50 cm depth beneath a glade area, and a grassland soil was covered in a 10 cm thick humic layer and colonised by grasses. The persistence of water repellency was measured using the water drop penetration time test, sorptivity and unsaturated hydraulic conductivity using a mini disk infiltrometer, and saturated hydraulic conductivity using a double-ring infiltrometer. Dye tracer experiments were used to assess the heterogeneity of water flow, and both the modified method for estimating effective cross section and an original method for assessing the degree of preferential flow were used to quantify this heterogeneity from the images of dyed soil profiles. Most hydrophysical parameters were substantially different between the two surfaces. The grassland soil had an index of water repellency about 10 times that of pure sand and the persistence of water repellency almost 350 times that of pure sand. Water and ethanol sorptivities in the grassland soil were 7% and 43%, respectively, of those of the pure sand. Hydraulic conductivity and saturated hydraulic conductivities in the grassland soil were 5% and 16% of those of the pure sand, respectively. Dye tracer experiments revealed a stable flow with “air-draining” condition in pure sand and well-developed preferential flow in grassland soil, corresponding to individual grass tussocks and small micro-depressions. The grassland soil was substantially more water repellent and had 3 times the degree of preferential flow compared to pure sand. The results of this study reinforce our view that the consequences of any change in climate, which will ultimately influence hydrology, will be markedly different between grasslands and bare soils. Key Words: dye tracing, grassland soil, hydrophobicity, infiltration, preferential flow Citation: Lichner, L., Eldridge, D. J., Schacht, K., Zhukova, N., Holko. L., Sˇır, M. and Pecho, J. 2011. Grass cover influences hydrophysical parameters and heterogeneity of water flow in a sandy soil. Pedosphere. 21(6): 719–729. INTRODUCTION Global changes in the Earth’s climate are known to have substantial and irreversible effects on soil and ecological processes (Solomon et al., 2007). Cli- mate change projections for continental Europe, for example, suggest a general increase in air tempera- tures and large changes in rainfall patterns (Hardy, 2003). Lenderink and van Meijgaard (2010) predicted that the hourly precipitation extremes could increase 1Supported by the Slovak Scientific Grant Agency VEGA (Nos. 2/0042/11 and 2/0073/11) and by the Ministry of the Environment of the Czech Republic (No.VaV SP/lab/151/07). 2Corresponding author. E-mail: lichner@uh.savba.sk. 720 L. LICHNER et al. by 70% or even more by the end of this century. Changes to the summer precipitation regime have important implications for hydrology. Many areas of southern Slovakia are experiencing increases in the fre- quency and intensity of heavy rains following long hot, dry periods (Fasˇko et al., 2008), leading to increased surface runoff, soil erosion and deposition of sediment within rivers and streams, as well as a nutrient washout (Pekarova et al., 1999). The relationship between rain- fall intensity, runoff and infiltration capacity of the soil is clearly an important consideration for determining the amount of total catchment runoff and its compo- nents (overland flow versus subsurface flow) following high-intensity storms (e.g., Sˇanda and Cıslerova, 2009; Holko et al., 2011). Runoff is likely to be exacerbated by water repel- lency, as it decreases infiltration rates, enhances over- land flow and increases the risk of soil erosion (Cerda` et al., 1998; Doerr et al., 2000). Water repellency is a transient soil property, which tends to be both spa- tially and temporally highly variable. It often disap- pears after periods of prolonged soil wetting, but will usually re-emerge during drier periods when soil mois- ture falls below a critical threshold (Dekker et al., 2001). Water flow paths, once created, persist over time during summer, but over annual cycles their spa- tial arrangements can change completely (Wessolek et al., 2009). Slight reductions in soil water content are known to cause substantial reductions in soil wetta- bility (Czachor et al., 2010). Soil water repellency is caused mainly by long-chained amphiphilic molecules, which have both hydrophilic (with polar functional group) and hydrophobic (with non-polar hydrocarbon chain) ends. Amphiphilic molecules may be released from a wide range of plants, organisms, and decaying organic matter. On drying, the hydrophilic ends bond more strongly with each other and soil particles, lea- ving an exposed hydrophobic surface (Hallett, 2007), leading to the fluctuations described previously. Both the persistence and severity of water repellency are in- fluenced by land use and plant cover, as well as soil temperature, texture, pH, water content, and by soil organic carbon (SOC) and clay (mainly kaolinite) con- tent (Doerr et al., 2000; Dekker et al., 2001; Lichner et al., 2002; Arcenegui et al., 2008; Lachacz et al., 2009; Novak et al., 2009; Diehl et al., 2010). Wang et al. (2010) revealed that the wettability in water repellent soils was affected more by SOC than by soil texture and pH, whereas in wettable soils, soil texture and pH had a greater impact. In grassland soils, wettability and infiltration can be affected considerably by vegetation cover (Alaoui et al., 2011; Kodesˇova` et al., 2011). Infiltration is a cri- tical process in grasslands where resources such as wa- ter, litter, nutrients and biological activity are typically scaled at the level of individual plants (Schlesinger et al., 1996). In general, the enhancement of infiltration due to plants and soil animal activity can be expected in fine-textured soils, whereas plants and biological crusts are known to decrease infiltration in coarse- textured soils (Cerda`, 1999; Ravi et al., 2007; Caldwell et al., 2008; Capuliak et al., 2010; Cerda` and Doerr, 2010; Eldridge et al., 2010). Grass cover can induce water repellency in all soil types ranging from sands (Dekker et al., 2001) to clays (Dekker and Ritsema, 1996) by both root exudates and thatch (the layer of organic matter between the mineral soil and the green grass). In well-structured soils, roots are often found in the inter-aggregate pore network, and therefore, water repellent substances are found upon the faces of ag- gregates and/or larger structural elements (Gerke and Kohne, 2002). Repellency is generally absent within aggregates and larger structural elements where roots are absent (Dekker and Ritsema, 1996). Aamlid et al. (2009) found sand layers to be more strongly water re- pellent than the overlying organic thatch layer. This could be caused by leaching of amphiphilic compounds from litter and the organic layer, by grass root exu- dates, and/or by the hyphae and exudates of soil fungi (Bond, 1964). The objective of the research was to quantify the influence of grass cover on hydrophysical parameters and heterogeneity of water flow in a sandy soil emerg- ing during a heavy rain following a long hot, dry pe- riod. We hypothesised that soil water repellency, in- duced by the thatch and mucilages of grasses, as well as infilling of voids by organic matter, would reduce sub- stantially sorptivity, and unsaturated and saturated hydraulic conductivity. Soil water repellency should also result in an increase in the heterogeneity of water flow (fingered flow). The practical relevance of our re- search is in quantifying the influence of grass cover on both the soil properties and the heterogeneity of water flow in a sandy soil. MATERIALS AND METHODS Study area The field sites were located in the Borska nızˇina Lowland of southwest Slovakia. This area covers about GRASS COVER INFLUENCES SOIL WATER FLOW 721 410 km2 and consists of aeolian sand dunes. The region is in transition zone between temperate oceanic and continental climates. Precipitation is seasonal, avera- ges 550 mm per year and is mainly summer-dominant. The climate in summer tends to consist of long hot and dry spells interspersed with intense rainfall. Air temperature averages about 9 C. Temperatures of 50 C on the surface of bare soil at a glade lasted about 5 hours a day during a hot summer, as measured on July 28, 2005. The daily precipitation amounts and temperature measured in the Meteorological Station of the Slovak Hydrometeorological Institute in Moravsky Svaty Jan, at a distance of about 5 km from the study sites, in July 2008 and 2010 are shown in Fig. 1. The study area was located at Mlaky II near Sekule (48 37′ 10′′ N, 16 59′ 50′′ E) which is about 150 m a.s.l. Two sites, separated by a distance of about 50 m, formed the basis of our study. A control soil (pure sand) with limited impact of vegetation or or- ganic matter was obtained by sampling at 50 cm depth beneath a glade area. This was compared to a grass- land soil that was covered by a 10 cm thick humic layer and colonised by grasses (Calamagrostis epigejos, and Agrostis tenuis). The sand soil at the glade site sup- ported a sparse cover of mosses (mainly Polytrichum piliferum) and lichens (mainly Cladonia sp.), and oc- casionally, grasses (mainly Corynephorus canescens) (Sˇomsˇak et al., 2004). Some areas in the glade had ex- posed bare soil. Both the macroscopic and microscopic soil fungi have been recorded at this site (Lichner et al., 2007). The soil was a Regosol formed from windblown sand (IUSS Working Group WRB, 2006) and had a sandy texture (Soil Survey Division Staff, 1993). Physi- cal and chemical properties of soil samples are pre- sented in Table I. With the exception of organic car- bon content (responsible for the differences in soil wet- tability), soil properties were almost identical. The mineralogy of the aeolian sand was primarily siliceous sand (silica content up to 90%), with a low content of primary minerals (feldspars and micas) (Kalivodova et al., 2002). Compared with the 10-cm thick humic layer at the grassland soil, a thin humic layer a few milime- ters thick occurred below the moss cover in the glade soil. The soil beneath this thin humic layer is wettable all year round. Thin textural transitions as well as an accumulation of Fe-oxides and Fe-hydroxides were de- tected in a 1–2 cm thick layer in the soil profiles during excavation (Lichner et al., 2010). Field methods Soil water content was estimated by the gravime- Fig. 1 Mean daily air temperatures and daily precipitations measured at the Meteorological Station of the Slovak Hy- drometeorological Institute in Moravsky Svaty Jan (southwest Slovakia) in July–August 2008 and 2010. 722 L. LICHNER et al. TABLE I Physical and chemical properties of the soil from the studied site Site Depth Sand Silt Clay CaCO3 C pH(H2O) pH(KCl) cm % g kg1 Grassland 0–5 91.26 2.81 5.93 < 0.5 9.9 5.14 3.91 Pure sand 50–55 94.86 1.74 3.40 < 0.5 0.3 5.54 4.20 tric method after drying at 105 C (Kutilek and Nielsen, 1994). The persistence of water repellency was measured using the water drop penetration time (WDPT) test (Doerr, 1998). In this study, 58 5 µL drops of distilled water were placed onto the soil surface from a standard height of 10 mm above the surface, and the time required for infiltration was recorded. Field measurements of infiltration were performed using a mini disk infiltrometer (Decagon Devices, USA) under a negative tension h0 = 196 Pa. The mini disk infiltrometer had a diameter of 4.5 cm with liquid transport through sintered stainless steel disc at the base that was in contact with the soil. Measure- ments were made with water and ethanol, with the latter providing an infiltration measurement that was not influenced by water repellency. The cumulative infiltration I was calculated based on the Philip infiltration equation (Philip, 1957): I = C1t1/2 + C2t+ C3t3/2 + C4t2 + . . . (1) where C1, C2, C3, C4, . . . are coefficients, and t is time. Zhang (1997) proposed to estimate the sorptivity S(h0) and unsaturated hydraulic conductivity k(h0) at suction h0 0 from: I = C1(h0)t1/2 + C2(h0)t (2) where C1(h0) and C2(h0) are functions of suction h0. The sorptivity S(h0) and unsaturated hydraulic con- ductivity k(h0) at suction h0 0 can be calculated from: S(h0) = C1(h0)/A1 (3) and k(h0) = C2(h0)/A2 (4) where A1 and A2 are dimensionless coefficients. Eq. 3 was used to calculate the sorptivities of both water, Sw(196 Pa), and ethanol, Se(196 Pa), from the cu- mulative infiltration vs. time relationships taken with the mini disk infiltrometer during early-time (< 180 s) infiltration of water and ethanol, respectively. The in- dex of water repellency R was calculated from (Hallett et al., 2001): R = 1.95Se(196 Pa)/Sw(196 Pa) (5) Eq. 4 was used to estimate the hydraulic conduc- tivity k(196 Pa) in this study, using A2 = 1.8 for sandy soil and suction h0 = 196 Pa from Table II in the Mini Disk Infiltrometer User’s Manual, Ver- sion 6 (Decagon Devices, 2007). In our former calcu- lations we used A2 = 2.4 for sandy soil and suction h0 = 196 Pa from the Mini Disk Infiltrometer User’s Manual, Version 2, published in 2005, and this coef- ficient was changed to A2 = 1.73 in Version 8, pub- lished in 2010. Dohnal et al. (2010) modified the origi- nal Zhang (1997) formula to lower the relative error of the hydraulic conductivity estimates. Grass and crust cover was removed (scalped) before the measurements of WDPT, k(196 Pa), Sw(196 Pa), and Se(196 Pa). Scalping could result in an increase in some of these properties, as observed by Eldridge et al. (2000) for sorptivity and steady-state infiltration rate under both ponding and tension. Field measurements of infiltration under a small positive pressure head h0 = 196 Pa were also per- formed repeatedly at all sites using a double-ring infil- trometer with an inner-ring diameter of 24.5 cm, buffer ring diameter of 34.5 cm, and length of 23.5 cm. The first two and three terms of the Philip infiltration equa- tion (Eq. 1) can be used to estimate the saturated hy- draulic conductivity Ks. The first two terms are appli- cable to relatively short times as follows: I Swt1/2 +mKst (6) where Sw is the sorptivity of water, with coefficient m = 0.667 being the most frequently used value (Ku- tilek and Nielsen, 1994). Kutılek and Krejcˇa (1987) proposed to use three terms of the Philip infiltration equation to estimate the saturated hydraulic conductivity Ks: GRASS COVER INFLUENCES SOIL WATER FLOW 723 Ks (3C1C3)1/2 + C2 (7) Eqs. 6 and 7 were used to estimate the saturated hy- draulic conductivity Ks in this study. The tracer experiment at the grassland site was carried out at the 100 cm 100 cm plot using the method similar to Bachmair et al. (2009) and Ho- molak et al. (2009). First, the grass was mowed, and the plot was partitioned into two smaller subplots (50 cm 100 cm each) to apply different rainfall amounts (20 and 70 mm). Rainfall of 70 mm was chosen to rep- resent increased rainfalls expected as a consequence of the climate change. Brilliant blue dye was added at a concentration of 10 g L1 to the water used to simu- late rainfall and the dyed water was applied at a rate of about 1 mm min1. The tracer application was con- ducted manually with a watering can, and it took 20 and 60 minutes for application amounts 20 mm (on the 100 cm 100 cm plot) and next 50 mm (on the 50 cm 100 cm subplot), respectively. Thirty minutes after the first sprinkling, vertical sections were excavated at distances of 15, 20, 30, and 40 cm from the edge of the plot, and clean soil profiles were photographed with a digital camera. Similarly, 30 minutes after the sec- ond sprinkling, vertical sections were excavated at dis- tances of 60, 70, 80, 90, and 100 cm from the edge of the plot, and clean soil profiles were photographed. The tracer experiment in the pure sand was carried out at the 50 cm 100 cm plot, where 50 mm of dyed water was applied. The soil surface on this plot was flattened to prevent uneven infiltration of water due to ponding. The water dyed with brilliant blue was ap- plied manually with a watering can at a rate of about 1 mm min1. Seventy minutes after sprinkling, verti- cal sections were excavated at distances of 10, 20, 30, 40, and 50 cm from the edge of the plot, and clean soil profiles were photographed with a digital camera. All the two-dimensional pictures were then digitally cor- rected and georeferenced using standard GIS software according to the scales provided. To make dye patterns comparable among sites and rainfall amounts, two parameters were chosen for a characterization of vertical flow patterns: the effective cross section for water flow, ECS, and the degree of preferential flow, DPF. ECS was used to quantify the heterogeneity of water flow in soil. The approach pre- sented in Taumer et al. (2006) was modified so that the fraction of total water content change was deter- mined from the stained area. The picture of each verti- cal section was divided into i = 10 vertical bands with the width of 10 cm, and the numbers ni of stained 5 cm 5 cm pixels were calculated in each band. The number of stained pixels is not an integer if all the area of pixels was not stained. The fraction of total water content change fi, which is the ratio between the water content change in band i and the total water content change in vertical profile, was calculated for each band using fi = ni / i=10 i=1 ni with i=10 i=1 fi = 1 (8) The fractions fi were ranked in descending order and presented against the fraction of cross-sectional area, FCA (black dots in Figs. 2b and 3b). A beta dis- tribution p(x, α, β) = Γ(α+ β) Γ(α)Γ(β) x(α1)(1 x)β1 (α > 0, β > 0, 0 x 1) (9) was fitted to the data and the Levenberg-Marquardt al- gorithm (the non-linear least-square fitting) was used to optimize the parameters α and β. The Levenberg- Marquardt method is a popular alternative to the Gauss-Newton method of finding the minimum of a function that is a sum of squares of nonlinear func- tions. The fitted curve expresses the share of the water content changes as a function of the area share. Ac- cording to the definition in Taumer et al. (2006), ECS was then estimated as the fraction of the total area that corresponds to the 90% of water content change in vertical section (Figs.

该用户的其他资料

  • 名称/格式
  • 评分
  • 下载次数
  • 资料大小
  • 上传时间

用户评论

0/200
    暂无评论
上传我的资料

相关资料

资料评价:

/ 11
所需积分:1 立即下载
返回
顶部
举报
资料
关闭

温馨提示

感谢您对爱问共享资料的支持,精彩活动将尽快为您呈现,敬请期待!