Catena 70 (2007) 366 – 374 www.elsevier.com/locate/catena
Spatial distribution of rock fragments in dolines: A case study in a semiarid Mediterranean mountain-range (Sierra de Gádor, SE Spain) Xiao-Yan Li a,b,⁎, Sergio Contreras b , Albert Solé-Benet b a
b
Institute of Land Resources, College of Resources Science and Technology, Beijing Normal University, No. 19, Xinjiekouwai Street, Beijing 100875, P. R. China Estación Experimental de Zonas Áridas, Consejo Superior de Investigaciones Científicas, General Segura 1, 04001 Almería, Spain Received 3 July 2005; received in revised form 28 August 2006; accepted 13 November 2006
Abstract Rock fragments (RFs) at the soil surface have great effects on the intensity of various hydrologic and geomorphic processes. However, little information is available on the spatial distribution of rock fragments (RF) in the dolines, which may be of importance in understanding overland flow and subsequent recharge in limestone karst landscapes. This study analysed spatial variability of RF cover and size in different topographic positions (top, upper, middle and lower position) in three dolines in Sierra de Gádor (Almería province, south-east Spain). The results indicated that cover percentage of small RFs (5–20 mm) increased but large RF cover (250–600 mm and N 600 mm) decreased from the upper position to the lower position of the dolines. Small RFs were usually resting on the soil surface while most large RFs tended to be partly embedded in the soil surface. Total RF cover and D50 (median diameter) of the surface RFs greater than 5 mm tended to increase with slope gradient. © 2006 Elsevier B.V. All rights reserved. Keywords: Rock fragments; Doline; Topographic positions; Mediterranean; SE Spain
1. Introduction Dolines, typical features of limestone terrains, are relatively shallow, bowl-shaped depressions ranging in diameter from a few to more than 1000 m (White, 1988; Ford and Williams, 1989). The formation of dolines mainly results from limestone solution and collapse (Cramer, 1941; Ford and Williams, 1989). In many areas, dolines can act as funnels concentrating near surface waters and forming important point-sources of recharge to limestone aquifers (Gunn, 1983).
⁎ Corresponding author. Institute of Land Resources, College of Resources Science and Technology, Beijing Normal University, No. 19, Xinjiekouwai Street, Beijing 100875, P. R. China. Tel./fax: +86 10 58802716. E-mail address:
[email protected] (X.-Y. Li). 0341-8162/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.catena.2006.11.003
For most dolines, soil surfaces are characterized by the presence of rock outcrops and RFs, which are products of geochemical, hydrologic and geomorphic process. Surface RFs, in turn, can affect the intensity of various soil hydrological processes such as surface sealing, infiltration, evaporation, runoff generation, runoff energy dissipation and erosion by water (Abrahams and Parsons, 1994; Brakensiek and Rawls, 1994; Poesen and Lavee, 1994; Poesen et al., 1994; Valentin, 1994; Van Wesemael et al., 1995, 1996). Yair and Lavee (1974) reported that large boulders could increase runoff while small stones decrease it. In fact, RF cover at the soil surface has an ambivalent effect on infiltration rate and on overland flow generation, which depends on various factors such as position, size and cover of RFs as well as structure of the fine earth (Poesen et al., 1990; Poesen and Lavee, 1994). When free on the soil surface, RFs generally prevent the soil from sealing and the infiltration increases, but embedded in the surface, they participate in the
X.-Y. Li et al. / Catena 70 (2007) 366–374
establishment of a continuous crust inhibiting infiltration and reinforcing runoff (Poesen and Lavee, 1994). Generally, infiltration in stony soil increases for small RFs but there exists a threshold of stone size (Valentin and Casenave, 1992). Above this threshold, infiltration decreases because of the less accessible surface for water flow (Valentin, 1994). Until now, most studies have dealt with the effects of RF cover on the intensity of earth surface processes pertaining to agricultural lands, whereas rangelands have been less well studied (Poesen et al., 1998; Cerdà, 2001). This is particularly true for quantitative information on spatial distribution of RFs on rangelands. Simanton et al. (1994) and Simanton and Toy (1994) noted that hillslope morphology influenced RF cover and consequently runoff and sediment yields; they found a logarithmic increase of RF cover with slope angle for hillslopes formed in weakly consolidated coarse Quaternary alluvium in semiarid rangelands of Arizona USA. Poesen et al. (1998) reported that the spatial variation of RF cover and RF size along semiarid hillslopes and transects in the Mediterranean was largely controlled by hillslope gradient. However, very little attention has been paid to spatial distribution of RF in dolines and its effects on soil hydrological process, which may be important in understanding overland flow and recharge processes in limestone karst landscapes. Therefore the objective of this study was to investigate spatial variability of RF cover and size with respect to topographic position, slope and aspect in three dolines in Sierra de Gádor (Almería, SE Spain).
367
and 1800 m a.s.l. (Vallejos et al., 1997). Most dolines and sinkholes are formed in this area. Soils are very thin and rocky, and vegetation is generally sparse. Vegetation cover is 50–60% and consists mainly of patchy dwarf perennial shrubs (30–35%) and grasses (20–25%). The woody shrubs include Genista pumila (7–13%), Thymus serpylloides (4– 17%) and Hormathophylla spinosa (2–6%). The herbaceous stratum is dominated by Festuca scariosa (14–20%) and Brachypodium retusum (2–4%). The landscape is a mosaic of rock outcrops or bare soil with RF patches interspersed with vegetation patches. Vegetation usually grows on
2. Study area Sierra de Gádor is located in southeastern Spain just west of the town of Almería (Fig. 1). It is a mountain range reaching 2246 m a.s.l. and consisting of a thick series of Triassic limestone and dolomites, i.e. highly permeable, fractured rocks with intercalated calcschists of low permeability underlain by impermeable metapelites of Permian age (Vandenschrick et al., 2002; González et al., 2003). A system of faults delimits the footslopes of Sierra de Gádor on the fringe of the coastal plain (Campo de Dalías). Sierra de Gádor has a Mediterranean climate characterized by dry-hot summers and wet-warm spring, autumn and winter. The climate has strong altitudinal gradients in annual precipitation (P) and temperature (T). Mean annual precipitation is 260 mm in Alhama de Almería (520 m a.s.l.) and approximately 650 mm on the summit (2246 m a.s.l.). The pluviometric gradient estimated is 23 mm/100 m. Mean annual temperature is 9 °C on the summit and 18 °C at the foot of the mountain. The thermometric gradient is about − 0.4 °C/100 m (Contreras et al., 2005). Llano de los Juanes, with an elevation of about 1660 m a. s.l., is a relatively flat area (Fig. 1) corresponding to a well developed karstic plateau. Isotopic studies have shown that the main recharge area to the deep aquifers of Campo de Dalías lies in this karstic plain with altitude between 1400
Fig. 1. Location of the study area.
368
X.-Y. Li et al. / Catena 70 (2007) 366–374
fractured rocks or ground littered with RFs. Soils are composed dominantly of silt (40–55%) and clay (26–52%). Organic matter content is relatively high ranging from 1.6– 11.9%. The land use pattern of the study area had been considered as a sylvopastoral system before 18th century (Gómez Cruz, 1992); however, large-scale deforestation occurred during the 18th and 19th century (Pulido-Bosch et al., 1993; Vandenschrick et al., 2002). At present, the land use is grazing by sheep and goats.
3. Methods Three representative dolines were selected in Llano de los Juanes to evaluate spatial cover of RF and RF size distribution (Fig. 1). Doline 1 and doline 3 are circular whereas doline 2 is a little oval-shaped (Fig. 2). The configurations of the three dolines characterize convexstraight-concave from top to bottom and therefore they can be divided into four topographic positions: top, upper,
Fig. 2. Views of the three dolines in the study. Kermes oaks (Quercus coccifera) and hawthorns (Crataegus monogyna) are growing at the bottom of doline 2 and 3, respectively.
X.-Y. Li et al. / Catena 70 (2007) 366–374
369
The roots of Crataegus can reach a depth up to 6.5 m although thick structural roots dominate over fine roots at the surface level (Silva et al., 2003). Eight transects in doline 1 and doline 3 were established from center to the ridge top along the directions of N, NW, W, SW, S, SE, E and NE (Fig. 3). A total of fifteen transects in doline 2 were measured from two centers at the bottom along the above same directions (Fig. 3). The length of each transect varied from approximate 40 to 60 m depending on length from the top to the bottom. For the top position of the dolines, five 10 m transects were established randomly for each doline. RF cover and surface conditions were recorded by the line-intercept method of Mueller-Dombois and Ellenberg (1974) along transects. RF size (the longest diameter) and position (embedded or not embedded), vegetation type and bare soil were recorded at 2 m intervals along a tape measure run from the bottom of the dolines. RFs were classified as partly embedded if a scar was left at the surface after removal. Conversely fragments that were not embedded but simply resting on the soil surface (no scars when removed) were classified as free fragments (Valentin and Casenave, 1992). Five RF size classes were used in the study according to Poesen et al. (1998): 5–20 mm (medium gravel), 20–75 mm (coarse gravel), 75–250 mm (cobble), 250–600 mm (stone) and N 600 mm (boulder). Soil surface with RFs smaller than 5 mm was classified as “uncovered soil surface”. Such particles were considered to be part of the fine fraction of the soil since they are easily transported by interrill flow, while RFs N 5 mm were considered to behave as relatively immobile particles on the interrill areas where they form erosion pavements (Poesen, 1987; Poesen et al., 1998). A total number of 970 points were sampled in the three dolines, of which 614 RFs were measured for rock size analysis. The median diameter (D50) of the RFs N5 mm was Table 1 Characteristics of the soil surfaces for the different topographic positions in the three dolines Length (m)
Fig. 3. Configuration of the three dolines compiled from a digital elevation model at 10 m resolution. Sampling–measuring transects are represented by solid lines. Dash lines show the limits between the different topographic positions of the dolines: T (top), U (upper), M (middle) and B (bottom).
middle and lower (Fig. 3). Length, vegetation cover and physical properties of soils for the different topographic positions in the three dolines are presented in Table 1. Two kermes oaks (Quercus coccifera) and hawthorns (Crataegus monogyna) are growing at the bottom of the dolines 2 and 3 (Fig. 2), respectively. Both Q. coccifera and C. monogyna are resprouter species with deep root system architecture.
Plant cover (%)
Organic matter (%)
Particle size composition USDA (%) Sand
Silt
Clay
4.95 6.76 8.06 4.37
14.21 16.33 12.37 13.14
50.46 49.08 51.26 52.54
35.33 34.59 36.37 34.32
77 70 64 49
3.35 7.15 6.88 7.63
4.43 21.91 14.68 6.67
44.38 49.57 55.19 41.15
51.19 28.52 30.13 52.18
28 71 63 67
1.66 11.97 11.13 6.45
16.39 15.45 14.26 16.69
40.53 57.93 53.70 45.83
43.08 26.62 32.04 37.48
Doline 1 Bottom Middle Upper Top
14 20 12 10
62 55 53 47
Doline 2 Bottom Middle Upper Top
12 10 12 10
Doline 3 Bottom Middle Upper Top
16 12 14 10
370
X.-Y. Li et al. / Catena 70 (2007) 366–374
it was highest in the lower position, followed by the middle and upper position respectively. For doline 3, the total RF cover and uncovered soil surface percentage distribution was opposite to that of doline 1 and doline 2. The total RF percentage decreased and the uncovered soil surface percentage increased from the lower position to the upper position of the doline. However, the general trend of RF size distribution was similar: small RFs tended to be concentrated in the lower position and large RFs in the upper position. There was no large difference in distribution of the percentages of various RF sizes for the top position between the three dolines. As far as the RF position was concerned, small RFs were usually resting on the soil surface while most large RFs tended to be partly embedded in the soil surface (Fig. 5). For
Fig. 4. Frequency of rock fragments with different sizes in relation to topographic positions for the three dolines.
calculated as well. Slope gradients were measured at 20 m intervals for each transect. The distribution of the slope gradients for the three dolines is shown in Fig. 3, which is generated from a 10 m resolution digital elevation model (Junta de Andalucía, 2005). 4. Results Fig. 4 shows percentages of various RF sizes relative to topographic position for each doline. For doline 1 and doline 2, total RF percentage (N 5 mm) was highest in the top or upper position of the doline, followed in decreasing order by the middle and lower position respectively. Small RF (5– 20 mm) percentage invariably decreased and large RF percentage (250–600 mm and N 600 mm) increased from the lower position to the upper position of the doline. In the case of the uncovered soil surface percentage (b5 mm), however,
Fig. 5. Percentage of embedded rock fragments for the different rock sizes at the different topographic positions of the three dolines.
X.-Y. Li et al. / Catena 70 (2007) 366–374
371
north-east. Doline 2, with an oval shape and average slope gradient of 15% slightly tilted to the north, showed a similar distribution pattern of total RF cover to that of doline 1 though the largest size RFs lay in the direction of the southwest. In the case of doline 3, which was quite circular in shape and had an average slope gradient of 8%, the RFs with highest cover and largest size were found on the south-west slope with the highest gradients. 5. Discussion
Fig. 6. Embedded rock fragments in a well-structured topsoil on side slope (A) and in a sealed topsoil at bottom (B) of doline 3.
Spatial distribution of RFs in dolines is a reflection of coupled geochemical, hydrological and geomorphic processes. The fractured structure of limestone and dolomite leads to unusual water and air circulation (Atalay, 1997). When water penetrates into rock fissures and holes and diffuses into limestones rocks, in situ weathering processes take place and erosion occurs quickly by solution. Dissolution of the rock results in red clayey residues. These red soils (terra rossa) favor water retention and plant root penetration. Root residues, producing organic acids, may be effective in chemical weathering (Atalay, 1997). These processes enlarge cracks and leave more fine soil particles transported to the lower part of the joints. A large fracture allows more water to flow through it and more carbonates to come into solution, leading to more soil at the bottom. Finally the cracks are growing into each other. If there is direct contact with the surface then a depression is
doline 1 and doline 2, embedded RF percentage was lower in the lower position than in the upper position; but this was reversed for doline 3, i.e. embedded RF percentage was high in the lower position and low in the upper position. From a visual field assessment, it was clear that RFs were generally embedded in a surface seal (i.e. a topsoil layer with essentially textural pore spaces) at the bottom of the doline, but in a topsoil layer with structural porosity in the middle and upper part of the dolines (Fig. 6). The side slopes of the three dolines were not steep. Both total RF cover percentage and D50 of the surface RFs N 5 mm tended to increase with slope gradient (Fig. 7), although a considerable scatter of points was evident for the total RF cover percentage, and the correlation is poor (correlation coefficient R2 was 0.07 and the significance level p was 0.14). However, the general trend is consistent with results obtained by Poesen et al. (1998) on semiarid hillslopes. The influence of slope aspect on spatial distribution of RFs in the three dolines is indicated in Fig. 8. In general, the total RF cover tended to be a little higher on southern and western side slopes than on the northern and eastern slopes, respectively. For doline 1, which was circular-shaped and slightly tilted to the south-west, with an average slope gradient of 6%, the total RF cover was highest in the NW direction which corresponded to the highest flanks of the doline, whereas the largest size RFs were clearly found on the slope with the highest gradients in the direction of the
Fig. 7. Relationship between slope gradient and (A) median diameter (D50) (cm) as well as (B) total cover percentage of rock fragments N5 mm. Analysis was based on 63 data pairs obtained from the three dolines.
372
X.-Y. Li et al. / Catena 70 (2007) 366–374
Fig. 8. Circular diagrams showing the cumulative percentage of rock fragments with different sizes in relation to orientation for the three dolines; rock sizes were classified as five levels: C1 = 5–20 mm, C2 = 20–75 mm, C3 = 75–250 mm, C4 = 250–600 mm, C5 = larger than 600 mm; which were grouped into three main classes in the diagram: ‘C1 + C2’ denotes rock sizes up to 75 mm, ‘C1 + C2 + C3’ denotes rock sizes up to 250 mm, and ‘C1 + C2 + C3 + C4 + C5’ denotes all rock fragments with different sizes.
directly formed (solution doline). Otherwise the hole is growing under the surface until the construction becomes too weak and collapses (collapse doline) (Jennings, 1971). The results of this study in dolines 1 and 2 indicated that rock outcrops and large RFs were distributed in the upper positions of the dolines and small RFs in the lower ones. This spatial pattern might be closely related to limestone solution in the upper positions and transport-sedimentation processes in the lower positions. However, the formation of dolines by dissolution process and spatial RF distribution by hydrolog-
ical and geomorphic processes operate at different time scales, which may differ by two or three orders of magnitude. For the transport-sedimentation processes, once a doline is developed, the centripetal focusing of flow and hence corrosion enhance the spatial pattern of RFs. Fine soil particles coming from the dissolution of the calcareous parent rock are washed away by the overland flow from the upper position of the doline to the bottom, resulting in RF concentration at the surface in the upper positions and fine soil sediment at the bottom. Small RFs may be moved by overland flow from the upper to the lower position of the doline. Poesen (1987) reported that RF transport distance was controlled more by fragment size than by fragment shape. He found that RFs up to 9 cm in diameter could be moved by overland flow during a moderate rainfall event and that rill or other types of concentrated flow were responsible for downslope movement of larger diameter RFs. Moreover, there are other factors responsible for the spatial variability of RF occurrence in hillslopes, such as: bioturbation, trampling (sheep and goats) and digging and wallowing (wild boar) (Poesen et al., 1998). Contrary to the general RF pattern found in doline 1 and 2, in the bottom sector of doline 3 RFs were partly embedded in a mantle of clay, which might reveal a long-term clay sediment concentration. The clayey soil with a low infiltration rate at the bottom of the doline would promote the ponding of runoff. The deep clayey soil has also a high water holding capacity which may provide enough water for plant species with high water requirements like hawthorns in the doline 3 (Table 1, Fig. 2). Doline 3 is within a completely flat area and very close to a reasonably large rock outcrop. Weathering by freeze–thawing generates large amounts of RFs which would be transported to the bottom by overland flow; this may account for a higher percentage of RFs in the lower position than in the upper position in doline 3 (Fig. 4), and also for the embedded RF percentage distribution (Fig. 5). Small RFs rest on the soil surface while most large RFs tend to be partly embedded in the soil surface. This is closely related to the fact that rock outcrops in the upper position of dolines are often embedded into the soil; dissolution and ice crystal pressure in cracks can cause larger RFs to become smaller, while frost heaving and overland flow can change RF position. The selective removal of fines by overland flow may be the reason for RFs being generally embedded in a surface seal at the bottom of the doline but resting on a topsoil layer with structural porosity in the middle and upper part of the doline. From the analysis of RF distribution in relation to orientation and slope gradient, we found that RFs with the highest cover percentage and the largest size were generally distributed on the slope with highest slope gradients in doline 2 and doline 3, e.g., sectors SW–S–SE–E in doline 2 and sectors W–SW–S in doline 3 (Fig. 8), which are correspondingly related to the position of large rock outcrops. However, in doline 1, the spatial distribution of RFs follows
X.-Y. Li et al. / Catena 70 (2007) 366–374
a particular pattern which might be related to the overall tilting of the doline towards south-west. Total RF cover and D50 of the surface RFs N 5 mm tended to increase with slope gradient. This is because runoff flow velocities, and thus transport capacities, tend to increase with increasing slope gradient. A higher transport capacity would increase the amount and size of soil material that could be transported by runoff (Simanton et al., 1994). This would lead to a relation between increased surface RF cover and increasing slope gradient. Parson and Abrahams (1987) reported that the mean diameter of soil surface particles on Mojave Desert debris slope was positively correlated with slope gradient. 6. Conclusions Dolines have an important role to play in the infiltrationrecharge events that take place in karstified semiarid landscapes. Because of this, studying spatial distribution of RF cover and RF size in their side slopes is an important topic to understand the spatial dynamic of soil moisture and derived hydrological processes. In dolines with typical convex-straight-concave profiles, such as those described in Llano de los Juanes, the spatial distribution of RF follows some patterns: small RFs tend to be concentrated in the lower position and rock outcrops and large RF in the upper position of dolines; small RFs rest on the soil surface while most large RFs tend to be partly embedded in the soil surface; total RF cover and D50 of the surface RF N 5 mm tend to increase with slope gradient. The total RF cover percentage is a little higher on southern and western side slopes than on the northern and eastern slopes, respectively. These common spatial patterns are the result of distinctive processes that take place along the topographic profiles. In the upper, convex sectors of the profiles with lower gradient slopes, limestone dissolution processes dominate over the transport and accumulation processes which are more important in the straight and concave sectors respectively. However, at the human temporal scale, dissolution processes are slower than transport-sedimentation processes. For that reason, in dolines which are close to each other, differences between spatial patterns of RF and vegetation cover are interlinked to other land use factors which might have an important role in promoting redistribution processes of water and sediments along these topographic profiles. Acknowledgements The two first authors acknowledge the Spanish Ministerio de Educación y Ciencia for the fellowships SB2000-0476 and AP2001-135. The research was also carried out as a part of the projects RECLISE (REN2002-04517-CO2-02) and CANOA (GLC2004-04919-C02-01) funded by the Spanish National R+D Programme of the MCYT, IRASEM research contract funded by Instituto del Agua (Junta de Andalucía) and the Foundation for the Author of National Excellent
373
Doctoral Dissertation of PR China for Dr. Xiao-Yan Li (Grant No. 200426). J.M. Calaforra, F.L. Perez Sanchez, H. Lavee, and an anonymous referee are specially thanked for their constructive review of the manuscript and Alfredo Duran and Montse Guerrero for field and laboratory support respectively. References Abrahams, A.D., Parsons, A.J., 1994. Hydraulics of interrill flow on stonecovered desert surfaces. Catena 23, 111–140. Atalay, I., 1997. Red Mediterranean soils in some karstic regions of Taurus mountains, Turkey. Catena 28, 247–260. Brakensiek, D.L., Rawls, W.J., 1994. Soil containing rock fragments effects on infiltration. Catena 23, 99–110. Cerdà, A., 2001. Effects of rock fragment cover on soil infiltration, interrill runoff and erosion. European Journal of Soil Science 52 (1), 59–68. Contreras, S., Boer, M.M., Domingo, F., Puigdefábregas, J., Lázaro, R., 2005. Desarrollo y aplicación de una aproximación descendente para la estimación local del drenaje anual medio en Sierra de Gádor (Almería). In: López-Geta, J.A., Rubio, J.C., Martín, M., Machuca, M. (Eds.), VI Simposio del Agua en Andalucía. GME, Sevilla, Spain, pp. 377–388. Cramer, H., 1941. Die systematic der Karstdolinen. Neues Jahrbuch fur Mineralogie, Geologie und Palaontologie 85, 293–382. Ford, D.C., Williams, P.W., 1989. Karst Geomorphology and Hydrology. Unwin Hyman, London. Gómez Cruz, M., 1992. Atlas histórico-forestal de Andalucía: siglo XVIII. Universidad de Granada, Granada, Spain. González, A., Domínguez, P., Franqueza, P.A., 2003. Sistema costero de Sierra de Gádor. Observaciones sobre su funcionamiento y relaciones con los ríos Adra y Andarax, y con el mar. TIAC, Alicante, Spain, pp. 423–432. Gunn, J., 1983. Point-recharge of limestone aquifers — a model from New Zealand karst. Journal of Hydrology 61, 19–29. Jennings, J.N., 1971. An Introduction to Systematic Geomorphology: Karst. The M.I.T. Press, Cambridge Massachusetts. Junta de Andalucía, 2005. Modelo Digital del Terreno de Andalucía. DVD. Mueller-Dombois, D., Ellenberg, H., 1974. Aims and Methods of Vegetation Ecology. John Wiley and Sons, New York. Parson, A.J., Abrahams, A.D., 1987. Gradient-particle size relation on quartz monzonite debris slopes in the Mojave desert. Journal of Geology 95, 423–452. Poesen, J., 1987. Transport of rock fragments by rill flow — a field study. In: Bryan, R.B. (Ed.), Rill Erosion. Catena Suppl., vol. 8, pp. 35–54. Poesen, J., Lavee, H., 1994. Rock fragments in top soils: significance and processes. Catena 23, 1–28. Poesen, J., Ingelmo-Sanchez, F., Mucher, H., 1990. The hydrological response of soil surfaces to rainfall as affected by cover and position of rock fragments in the toplayer. Earth Surface Processes and Landforms 15, 653–671. Poesen, J., Torri, D., Bunte, K., 1994. Effects of rock fragments on soil erosion by water at different spatial scales: a review. Catena 23, 141–166. Poesen, J.W., van Wesemael, B., Bunte, K., Solé-Benet, A., 1998. Variation of rock fragment cover and size along semiarid hillslopes: a case study from southeast Spain. Geomorphology 23, 323–335. Pulido-Bosch, A., Martin Rosales, W., Vallejos, A., Molina, L., Navarrete, F., de Simon, E., 1993. The Southern Catchment area of the Sierra de Gador and its impact in the Campo de Dalias. In: Pulido-Bosch, A. (Ed.), Some Spanish Karstic Aquifers. Universidad de Granada, Granada, Spain, pp. 159–181. Silva, J.S., Rego, F.C., Martins-Loução, M.A., 2003. Root distribution of Mediterranean woody plants — introducing a new empirical model. Plant Biosystems 137 (1), 63–72. Simanton, J.R., Toy, T.J., 1994. The relation between surface rock fragment cover and semiarid hillslope profile morphology. Catena 23, 213–225.
374
X.-Y. Li et al. / Catena 70 (2007) 366–374
Simanton, J.R., Renard, K.G., Christiaensen, C.M., Lane, L.J., 1994. Spatial distribution of surface rock fragments along Catenas in semi-arid Arizona and Nevada, USA. Catena 23, 29–42. Valentin, C., 1994. Surface sealing as affected by various rock fragment covers in West Africa. Catena 23, 87–97. Valentin, C., Casenave, A., 1992. Infiltration into sealed soils as influenced by gravel cover. Soil Science Society of America Journal 56, 1667–1673. Vallejos, A., Pulido-Bosch, A., Martín-Rosales, W., Calvache, M.L., 1997. Contribution of environmental isotopes to the knowledge of complex hydrologic systems. A case study: Sierra de Gador (SE Spain). Earth Surface Processes and Landforms 22, 1157–1168. van Wesemael, B., Poesen, J., de Figueiredo, T., 1995. Effects of rock fragments on physical degradation of cultivated soils by rainfall. Soil and Tillage Research 33, 229–250.
van Wesemael, B., Poesen, J., Kosmas, C.S., Danalatos, N.G., Nachtergaele, J., 1996. Evaporation from cultivated soils containing rock fragments. Journal of Hydrology 182, 65–82. Vandenschrick, G., van Wesemael, B., Frota, E., Pulido-Bosch, A., Molina, L., Stievenard, M., Souchez, R., 2002. Using stable isotope analysis (δD–δ18O) to characterise the regional hydrology of the Sierra de Gador, south east Spain. Journal of Hydrology 265, 43–55. White, W.B., 1988. Geomorphology and Hydrology of Karst Terrains. Oxford University Press, New York. Yair, A., Lavee, H., 1974. Areal contribution to runoff on scree slopes in an extreme arid environment. A simulated rainstorm experiment. Zeitschrift fur Geomorphologie. Supplementband 21, 106–121.
