Mediterranean river systems of Andalusia, southern Spain, and associated deltas: A source to sink approach

Marine Geology 222–223 (2005) 471 – 495 www.elsevier.com/locate/margeo

Mediterranean river systems of Andalusia, southern Spain, and associated deltas: A source to sink approach C. Liquete, P. Arnau, M. Canals *, S. Colas GRC Geociencies Marines, Dep. d’Estratigrafia, Paleontologia i Geociencies Marines, Facultat de Geologia, Universitat de Barcelona, E-08028 Barcelona, Spain Received 8 June 2004; received in revised form 28 December 2004; accepted 15 June 2005

Abstract The northern shores of the semi-isolated Mediterranean Sea with its low tidal range and a relatively voluminous fluvial supply of sediments constitute an ideal delta forming environment. In this paper, we examine the present-day and multi-decadal behaviour of 26 river systems from Andalusia, southern Spain, forming deltas in the Alboran Sea, the westernmost basin in the Mediterranean Sea. Eastern, Central and Western Andalusian watersheds have been defined based on geomorphological, climatological and hydrological characteristics. A comprehensive data set has been compiled, including satellite images, aerial photographs, a digital elevation model, thematic maps, time series of precipitation, temperature and water discharge, and the damming history of individual river basins. This data set has been used to analyse basin morphology and hydrology, and anthropogenic impact. Several modelling approaches have been applied to obtain the water budget and mean annual sediment yield of 12 of the 26 studied river systems. In addition, the periodicities of water discharge events and their possible link with North Atlantic Oscillation (NAO) fluctuations have been also studied. A decreasing trend has been observed in most water discharge time series during the last decades, which has been attributed to natural factors. Although it could have been expected that the diminution of water discharge may have caused a reduction in sediment load, calculated sediment discharge time series do not show any significant tendency. In general, sediment yield shows an opposite relationship with basin area. A comparative analysis of Spanish Mediterranean deltas indicates that in terms of sediment transport Andalusian river systems are quite efficient despite the small size of their catchments. Repetitive flood events and the consequent suspension plumes off river mouths play a major role in the development of deltaic and prodeltaic bodies. Nowadays, 42% of the study area is regulated, although to date the effect of dam building is hardly noticeable on river mouths. D 2005 Elsevier B.V. All rights reserved. Keywords: river/delta systems; southern Spain; water discharge; sediment load

* Corresponding author. Tel.: +34 93 402 13 60; fax: +34 93 402 13 40. E-mail address: [email protected] (M. Canals). 0025-3227/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.margeo.2005.06.033

472

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

1. Introduction Deltas form where the supply of sediment by rivers exceeds the dispersal capacity of the receptor basin (McManus, 2002). Thus, the semi-enclosed Mediterranean Sea with its micro-tidal range and its relatively important supply of sediments constitutes an ideal delta-forming place. Anthropogenic influence in the Mediterranean region dates from millennia, though it has increased during the last centuries and, specially, during the 20th century (Poulos and Collins, 2002; Vorosmarty et al., 2003). Mediterranean river basins and associated coastal environments are a perfect place to study the processes governing the transfer of water and sediment from continent to ocean and the historically recent human impacts on such transfer. The Mediterranean Basin (3800
800 km approximately) is located at mid-latitude, between 338N and 458N. Dominant winds and consequent atmospheric climate are strongly conditioned by mountain ranges such as the Alps and the Central Massif in Central Europe, the Pyrenees, the Iberian Massif and the Betic Cordillera in Spain, the Taurus Mountain range in the Anatolia Peninsula, and the Lebanese mountains (UNEP, 2002). These mountainous barriers together with a considerable land area surrounding a relatively small sea make Mediterranean weather and climate more continental in origin than marine (Hopkins, 1985). There are also important differences within the Mediterranean watershed. Yearly rainfall varies from more than 1500 mm over the European mountain ranges to less than 100 mm inland of northern Africa and western Asia (UNEP, 2002). This explains why runoff is at least one order of magnitude greater in the northern watershed than in the southern one (Cruzado, 1979; Liquete et al., 2004). During the Holocene, the main Mediterranean river systems (Ebro, Rhone, Po, Axios, Seyhan and Nile) formed 10- to 40-m-thick deltas, while during the 20th century the Ebro, Po and Nile deltas underwent coastal retreats of more than 10 m (Poulos and Collins, 2002). Nevertheless, excluding these few large rivers, most of the fluvial systems draining into the Mediterranean Sea are relatively small and flow into relatively narrow littoral zones. Farnsworth and Milliman (2003) and Milliman and Syvitski (1992) pointed out the importance of small mountainous rivers, usually intermittent and event-driven, to the global

sediment budget. However, the prime problem when studying these systems is most often the lack of continuous hydrological data. Their transport patterns are just starting to receive scientific attention (e.g., Kettner and Syvitski, 2003; Kineke et al., 2003; Pavanelli and Pagliarani, 2002). In addition, environmental changes may have a greater impact on these small streams than on larger rivers (Syvitski, 2003). Fluvial sediment load is sensitive to many hydrologic, geomorphic, climatologic and anthopogenic factors. Parameters such as relief, precipitation, air temperature, runoff, vegetal cover and lithology must be taken into account when evaluating erosion in a river basin (Dendy and Bolton, 1976; Jansen and Painter, 1974; Ludwig, 1997; Pinet and Souriau, 1988; Probst, 1992). These parameters have been considered in this paper to examine the present-day and multi-decadal behaviour of river/delta systems from Andalusia flowing into the Alboran Sea, the westernmost basin in the Mediterranean Sea. It is the first attempt to systematically study a set of neighbouring river systems feeding deltas along 400 km of shoreline in southern Spain and to quantify hydrological and sedimentary budgets.

2. Regional setting Dozens of river systems feed deltaic and prodeltaic bodies along the Mediterranean shoreline of Spain. These range from the large Ebro system (982 km long and 85 708 km2 of river basin) to small rivers that are dry most of the year. All Andalusian fluvial systems flowing into the Alboran Sea and forming deltas have been considered in this work (Fig. 1). These include 26 river basins draining the Betic Cordillera Internal Zone of Andalusia (with peaks up to 3500 m in height), one of the hottest, driest and least vegetated regions in Spain. The eastern part of the study area can be regarded as one of the most tectonically active zones in the Iberian Peninsula (Gimenez et al., 2000). The Betic Cordillera formed during the Alpine orogeny between 20 and 5 Myr ago (Fontbote and Vera, 1983). The Cordillera resulted from the convergence between the African and the Eurasian plates at a rate of about 0.5 cm/yr (Garcia et al., 2003; Sanz de Galdeano et al., 1995). The collision between the two plates caused continuous compression leading to the

L C A

F O

C Europe D Africa E

G

M

H

E Spain P

B

J

D 5∞W

4∞

F G H J

Catalan rivers system Ebro river system Levantine rivers system Andalusian rivers system

K L M O P

Pyrenees Cantabric Cordillera Iberian Massif Central System Betic Cordillera

3∞W

Eastern watershed

Western watershed

37°N

1 14 15 16 17

A

18 - 23 24

13

12 11 10

0

25 - 26

15

30

Sea n a r lbo

45 km

5°W

7

3-4

2

Figure 3

Gauging stations Dams Meteorological stations

36°N

8-9

65

4°W

River systems 1 Andarax 2 Adra 3 Huarea 4 Albuñol del Tranco 5 Haza de Trigo 6 Gualchos 7 Guadalfeo 8 Verde Almuñecar

9 10 11 12 13 14 15 16 17

Seco Chillar Patalamara Algarrobo Guaro Guadalmedina Guadalhorce Fuengirola Verde Marbella

18 19 20 21 22 23 24 25 26

Guadaiza Guadalmina Dos Hermanas Guadalmansa Castor Padrón Guadiaro Guadarranque Palmones

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

Central watershed

Figure 2

37°N

K

A Atlantic Ocean B Mediterranean Sea

36°N

3°W

473

Fig. 1. Location map of the studied river systems including gauging stations, meteorological stations and dams. The names of the river basins are given in the inset, numerically ordered from east (1) to west (26). Only those river systems supporting deltas/prodeltas at their mouths have been considered. White lines mark the boundaries between adjacent river basins. Continuous black lines correspond to main rivers (thicker) and tributaries (thinner). A Digital Elevation Model (DEM) constructed from the last Shuttle Radar Topography Mission is included as background. Dashed black lines correspond to the boundaries between Eastern, Central and Western Andalusian watersheds. Blue boxes point to the areas represented in Figs. 2 and 3. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

474

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

the study area consist of limestone, marl, dolomite and slate (Fontbote and Vera, 1983; IGME, 1969). The Betic Cordillera behaves as a natural barrier that isolates a narrow coastal zone from the rest of the Iberian Peninsula. It explains the presence of two principal climatic zones within the study area. Firstly, the Alboran coastal area, a strip of land 3–35 km wide located between the shoreline and the Betic foothills. This zone shows a typical Mediterranean climate with a mean annual temperature of 18–19 8C, ranging from subtropical to the west (400–900 mm yr 1 of preci-

formation of NW–SE and NE–SW trending fault systems (Estrada et al., 1997; Galindo-Zaldivar et al., 1997; Sanz de Galdeano and Vera, 1992). The modern physiography was already established by the end of the Pliocene, when the Strait of Gibraltar formed (Fontbote and Vera, 1983). Although faulting has been a very active process, recent local studies (Galindo-Zaldivar et al., 1997; Garcia et al., 2003) indicate that landscape evolution has been primarily driven by the incision of the drainage network and by regional uplift. The main rock types outcropping in 5°30'W 0

10

20

5°0'W

4°30'W

30 km

37°0'N

37°0'N

14 15

36°30'N

16 17

36°30'N

18 - 23 River systems 14 Guadalmedina 15 Guadalhorce 16 Fuengirola 17 Verde Marbella 18 Guadaiza 19 Guadalmina 20 Dos Hermanas 21 Guadalmansa 22 Castor 23 Padrón 24 Guadiaro 25 Guadarranque 26 Palmones

24

25 - 26 36°0'N

5°30'W

5°0'W

36°0'N

4°30'W

Fig. 2. Large sediment plume off the Guadalhorce River mouth (15) and smaller ones off the Guadiaro (24), Guadarranque (25) and Palmones (26) rivers as seen on a georeferenced MODIS image from the 10th of December 2003 combined with an 82-m resolution DEM onland. Note the Atlantic Jet entering into the Alboran Sea across the Strait of Gibraltar. Thick white line corresponds to the outer boundary of the Mediterranean watershed in Andalusia while thin white lines represent limits between the different basins. Rivers are indicated by black lines. Red-shadowed areas are those presently regulated behind dams. See location in Fig. 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

pitation and no frost) to subdesertic to the east (less than 300 mm yr 1, usually as torrential rain, and more extreme temperatures denoted by a mean annual temperature of 17–21 8C) (IEA, 2003). The second climatic zone corresponds to the Betic mountains, whose climate is colder (mean annual temperature of 13 8C and frequent frost) and rainier (up to 1000 mm yr 1) than in the coastal stretch (IEA, 2003). The mean climatic conditions are disrupted at least once a year by violent storms leading to flood events, often of a catastrophic nature (e.g., Romero Cordon et al., 2003). These floods carry large amounts of sediment in single episodes and trigger the formation of large suspension plumes off the river mouths (Fig. 2) that often feed deltas (Fig. 3) and prodeltas. Alongshore sediment redistribution by waves and coastal currents contributes to delta reshaping and prevents the preservation of distributary channel-fed deltaic fingers. Surface oceanographic currents on the Alboran Sea are relatively strong due to the influence of the incoming Atlantic Jet (that enters through the Strait of Gibraltar with a speed of around 1 m s 1), the gyres and counter-gyres it induces, and the habitual windy conditions (e.g. Heburn and La Violette, 1990; Perkins et al., 1990; Sarhan et al., 2000; VargasYan˜ez et al., 2002). Only a few Alboran Sea prodeltas have been studied in detail. Internal structure of the Guadalhorce Prodelta, for instance, shows at least two progradation/aggradation cycles overlying the last 3°10'0"W

475

post-glacial transgressive system track. Guadalhorce Prodelta deposits average 22 m thick and extend as far as 10 km offshore (Fernandez-Salas et al., 2003).

3. Materials and methods 3.1. Data compilation and analysis A comprehensive data set from the Iberian Peninsula river systems flowing into the Alboran Sea has been compiled. The data set includes satellite images, aerial photographs, a digital elevation model (DEM), thematic maps, time series of precipitation, temperature and water discharge, and the damming history of individual river basins. This data set has been used to analyse basin morphology, hydrology, and anthropogenic impact. Several modelling approaches have been applied to obtain water budget and sediment discharge data. 3.1.1. Basin morphology Morphometric data include basin area, basin length, basin maximum elevation, river maximum elevation, river length, river slope, delta area, delta perimeter and delta coastal length (see Section 4.1 and Table 2) (Canals et al., 2004). An 82-m resolution DEM based on the NASA’s Shuttle Radar Topography Mission (SRTM) has been 3°9'0"W

3°8'0"W

0 36°45'30"N

36°45'0"N

0.2 0.4 0.6 km

36°45'30"N

36°45'0"N

4 3°10'0"W

3 3°9'0"W

3°8'0"W

Fig. 3. Small triangular deltas formed at the mouths of Huarea (3) and Albun˜ol del Tranco (4) streams as seen on an aerial photograph taken in 1997. Delta areas are 0.59 km2 and 0.63 km2, respectively.

476

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

implemented into a GIS. The DEM allowed an accurate morphometric analysis that, jointly with georeferenced aerial photographs (MAPA, 1997) and thematic maps of lithology (IGME, 1969), vegetation coverage and land-use (CMA, 1999), led to a quantitative characterisation of the 26 river basins in the study area which have developed deltaic systems at their mouths (Figs. 1 and 3). River basin lithology and vegetation indexes have been obtained from the data compilation (CMA, 1999; IGME, 1969, 1972) following Jansen and Painter (1974) and Probst (1992). Jansen and Painter (1974) related rock hardness to their geological history and defined a lithology index that ranges from 2 to 6. Probst (1992) linked soil erosion to rock nature and created an index ranging from 1 (= metamorphic rocks) to 40 (= alluvion). Vegetation index ranges from 1 (= desert) to 4 (= forest) after Jansen and Painter (1974), and from 0 (= desert) to 6 (= forest) after Probst (1992). Hypsometry, longitudinal profile graphs and cross-correlations between various parameters have been produced from the available data.

Monthly terrestrial air temperature time series from 1950 to 1999 were extracted from the Willmott and Matsuura (2001) database. This database results from both a DEM-assisted interpolation and a Climatologically Aided Interpolation of Global Historical Climatology Network v.2 and Legates and Willmott (1990a,b) data. The spatial resolution is 0.58
0.58 longitude/latitude. The observation of these data shows that there has been a warming trend in the study area of at least 1 8C during the last 50 yrs. Statistically significant periodic events were identified within the discharge and pluviometric time series by means of spectral analysis, namely the Fourier spectrum for unevenly sampled data. The analysis was carried out with 12 monthly water discharge and rainfall data series and 3 daily discharge series. Only spectrum levels exceeding a 99.9% peak-based critical limit have been assumed as significant frequencies. This means that there is less than a 1 in 1000 probability for the identified peak to arise strictly from chance.

3.1.2. Basin hydrology The hydrological data set consists of meteorological time series, daily and monthly discharge data, and hydrological runoff. Annual sediment discharge data were estimated from simple empirical models, whose validity has been carefully assessed. Precipitation and water discharge time series for 12 of the 26 river systems under study have been gathered from bConfederacion Hidrografica del SurQ basin authority and from the Spanish bInstituto Nacional de MeteorologiaQ. The length of the time series ranges from 10 to 89 yrs including temporary gaps of variable duration (Table 1). Hydrological runoff (l km 2 s 1) was calculated when data were available. Between the non-monitored rivers, there are several ephemeral streams or bramblasQ that usually carry a relatively large volume of water during the stormy season while they are dry during the rest of the year. When it was possible, gauging stations were chosen following two main criteria: (i) proximity to river mouth, and (ii) duration of the time series (Table 1). Selected meteorological stations within individual river basins are as close as possible to the mean altitude of the basin in order to extract mean precipitation time series (Table 1).

3.1.3. Anthropogenic impact Information related to 37 mid-size to small dams located in the studied river systems was integrated in the GIS (Fig. 1 and Table 3). Data were mostly provided by the Spanish bMinisterio de Medio AmbienteQ and were subsequently improved by means of satellite images and the DEM. Present land-use in the studied river basins has been also analysed from thematic maps (CMA, 1999) and aerial photographs (MAPA, 1997). 3.2. Modelling Water budgets were obtained for the 12 river systems in Table 1, from where time series of potential and actual evapotranspiration, total water supply, superficial runoff and possible infiltration were derived (Table 4). Generally, hydrological models approximate soil humidity change by the expression P = AET + Q + I, where P is precipitation, AET is actual evapotranspiration, Q is fluvial surface runoff, and I is ground infiltration. Annual actual evapotranspiration was calculated after Turc (1954) and Pike (1964), whose proposed formulae are, respectively, AET = P/(0.9 + ( P/L)2)1/2, and AET/PET = (Pt/PET)/ (1 + (Pt/PET)2)1/2, where P is the mean annual precipitation, L is a function of the mean annual tempera-

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

477

Cumulated gaps (yr)

Distance to river mouth (km)

Altitude (m)

Rainfall data period

Canjayar

466

1942-2000

20.5

42.2

Abrucena

975

1942-2000

Adra

Fte. de Marbella

292

1983-2000

1

19.8

Nechite

975

1946-2000

7

Guadalfeo

Narila

1110

1942-2000

17

58.7

Portugos

1300

1941-2000

River

Meteorological station

Hydrological data period

Andarax

2

Gauging station

1

No. in Fig.1

Altitude (m)

Table 1 Descriptors of the gauging stations from where water discharge time series have been obtained and of the meteorological stations from where rainfall time series have been extracted

8

˜ Verde Almunecar

Cazulas

321

1967-2000

2.5

13.3

Lentegi

631

1943-2000

11

Patalamara

Torrox

217

1982-2000

6.5

5.7

Competa

636

1942-2000

12

Algarrobo

La Umbria

139

1942-2000

4

5.6

Competa

636

1942-2000

13

Guaro

Salto del Negro

143

1950-2000

15.5

21.4

Periana

547

1942-1999

14

Guadalmedina

Casabermeja

512

1942-2000

23.5

37.0

Casabermeja

547

1945-2000

15

Guadalhorce

El Chorro

310

1911-2000

40

69.4

Antequera

460

1917-2000

24

Guadiaro

Corchado

204

1911-2000

13.5

45.4

Ronda (central)

600

1935-1998

25

Guadarranque

La Almoraima

16

1952-1961

0

12.2

Castellar de la Ftra.

240

1945-1981

36.7

˜ El Castano

180

1951-2000

26

Palmones

Charco Redondo

136

1981-2000

0.5

Numbers to the left of river names correspond to those from Fig. 1. Grey shadings separate the Eastern (upper), Central (middle) and Western (lower) Andalusian watersheds.

ture, PET is the annual potential evapotranspiration, and Pt is the total annual precipitation. After AyalaCarcedo (1996), Turc’s formulae are particularly appropriate for Spanish watersheds. However, Pike’s equations seem to be more reliable at least when precipitation does not reach 300 mm yr 1. PET is the amount of water available to evapotranspirate in a certain moment and place, while AET is what is actually evapotranspirated. We calculated monthly PET after Thornthwaite (1948): PETm = 16 (10T m/I)a , where I is an annual heat index, T m is the mean monthly temperature, and a is a parameter that depends on I. Mean annual sediment yield (SY in t km 2 yr 1) time series were calculated for the same 12 river systems following four basic models of mechanical erosion. (a) Dendy and Bolton (1976) formulae: SY ¼ 1280Q0:46 ð1:43  0:26logAÞ; when Qb2 in: SY ¼1958e0:055Q ð1:43  0:26logAÞ; when Qz2 in:

(b) Jansen and Painter’s (1974) general equation: SY ¼  2:032 þ 0:100logQ  0:314logA þ 0:750logH þ 1:104logP þ 0:368logT  2:324logV þ 0:786logL (c) 4-parameters Probst’s (1992) expression: lnSY ¼ 1:5610 þ 0:9655lnS þ 0:0023Q þ 0:5692lnP  0:8660V (d) 5-parameters Probst’s (1992) model: lnSY ¼  0:0723 þ 1:0280lnS þ 0:0365L þ 0:6932lnP þ 0:0016Q  0:7516V where Q is measured runoff, A basin area, H mean basin height, P precipitation, T temperature, V vegetation index, L lithology index, and S slope (units and scales are not constant between the different formulae). Obviously, some inaccuracies could arise by

478

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

using models that were not derived from the study area. For example, Jansen and Painter (1974) and Probst (1992) take into account river basins larger than those in the study area, while Dendy and Bolton equations consider all-size basins. It must be noticed that sediment retention behind dams has not been taken into account. Three of the modelled river systems have no dams, while only five have dams older than 1983 (Table 3 and Fig. 1). Note also that water discharge data were taken from gauging stations situated at variable relative locations within each river basin (Table 1 and Fig. 1). Finally, an approximation of the suspended sediment load carried under natural conditions by the studied river systems was obtained, as well as its evolution through the last decades. Bed-load was not calculated, it is generally assumed to represent less than 10% of the suspended load (Milliman and Meade, 1983; Pinet and Souriau, 1988).

4. Results 4.1. Hydromorphology Main morphometric results of the 26 studied river systems are shown in Table 2. River slope ranges from 0.007 to 0.144, while river length varies between 4.7 and 154.2 km. Based on these two parameters (Fig. 4), the authors have classified 17 streams (3–6, 8–12, 14, 16–23 in Fig. 1) as torrents, characterised by elongate basin areas, no or only intermittent tributaries, axial gradients greater than 0.03, and stream lengths shorter than 34 km. Thirteen of these torrents do not have any reservoir along their course. The remaining nine river systems are larger, have a more complex drainage network and are all partially regulated. The torrential character has certainly had a major impact on sediment transport and discharge as evidenced by the fact that Andalusian deltas are collectively the largest in the Iberian Peninsula Mediterranean watershed with respect to river basin areas (Fig. 5). In addition, deltas have a typical triangular shape (Fig. 3) interpreted as indicative of seaward development. Data from the Spanish Harbour Authority (Puertos del Estado, 1983– 2004) show that wave activity cannot be the cause of this difference. Mean wave activity on the north-

ern Alboran inner shelf ranges usually from 0 to 1 m height and from 3 to 6 s period, being similar to the mean Spanish Mediterranean coastal conditions. Mean river length is 36 km. Longitudinal profiles are usually concave up and the axial gradient becomes steeper upstream despite some local pronounced steps at mid and lower course in specific rivers. The thirdorder polynomial has been established as a best fit for all longitudinal profiles of Andalusian Mediterranean rivers with R 2 values between 0.966 and 0.999. The whole Andalusian Mediterranean watershed can be subdivided into Eastern Andalusian (from Andarax to Guadalfeo, 1–7 in Fig. 1), Central Andalusian (from Verde Almun˜ecar to Fuengirola, 8–16 in Fig. 1) and Western Andalusian (from Verde Marbella to Palmones, 17–26 in Fig. 1) watersheds, each of them displaying specific features and following different hydrological patterns. 4.1.1. Eastern Andalusian watershed Climate in Eastern Andalusia is semiarid, colder and drier than over most of the studied river basins (IEA, 2003). Therefore, Eastern Andalusia has the most extreme winters and the least vegetated area. Data from Moreira and Ojeda (1992) shows that erosion risk in the Eastern zone is the highest of the study area. Estimated soil loss reaches values of more than 300 t ha 1 yr 1 in most of the watershed (CMA, 1997). The Four seasonal index gives an idea of the variability of precipitation along the year (Ludwig, 1997, from Fournier, 1960). It is defined P modified as ( P2i /P), where P i is the total precipitation of each month and P is the total mean annual precipitation (both in mm). This parameter shows that Eastern Andalusia is more variable than the Central and Western watersheds. Basin area and main river length within this subset range from 24 to 2160 km2 and from 9 to 74 km, respectively (Fig. 1 and Table 2). Since river sources are in Sierra Nevada, where the highest mountain peaks in the Iberian Peninsula are found, Eastern Andalusian river basins reach the maximum elevations of the whole study area. In particular, the Guadalfeo reaches 2793 m river elevation and 3243 m basin elevation. Accordingly, Eastern systems show the largest mean axial gradient (Table 2), reflecting a strong torrential character. Hypsometry plots illustrate that the largest rivers within this subset, i.e. Andarax,

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

479

Haza de Trigo

6

Gualchos

7

Guadalfeo

8

˜ Verde-Almunecar

Delta area (km2)

River maximum

Basin maximum elevation (m) 2481

988

2047

74.0

0.028

3

16.8

1.8

9.4

14.0

7.6

750.7

2682

1075

2277

51.4

0.044

70

10.1

1.8

2.2

6.8

3.3

40.8

1172

492

1120

12.3

0.091

0

5.0

2.6

0.6

4.1

2.1

118.1

1335

821

1267

20.0

0.063

0

5.0

2.6

0.6

4.3

1.8

24.2

1378

565

1267

8.7

0.144

0

5.0

2.6

0.2

2.6

1.2

% Regulated area

2160.5

Mean basin elevation (m)

Delta coastal length (km)

Albunol ˜ del Tranco

5

Delta perimeter (km)

4

Vegetation index

Huarea

Lithology index

3

River slope

Adra

River length (km)

Andarax

2

elevation (m)

1

Basin area (km 2 )

No. in Fig.1

River system

Table 2 Morphometric descriptors of the 26 river/delta systems represented in Fig. 1

74.7

1755

645

1080

13.4

0.080

0

5.0

2.6

0.2

2.4

1.1

1312.2

3243

1253

2793

72.5

0.038

79

7.3

3.2

8.6

15.7

6.0

100.6

1624

701

1502

23.4

0.064

0

5.0

3.6

0.2

2.5

1.0

9

Seco

21.4

1069

376

533

10.5

0.051

0

5.0

3.6

0.3

2.6

1.0

10

Chillar

55.7

1588

727

1428

16.2

0.088

0

4.0

5.4

0.6

3.5

1.1

11

Patalamara

50.0

1534

571

1457

17.5

0.083

0

4.6

4.0

0.8

3.9

1.9

12

Algarrobo

64.8

1667

705

1509

21.0

0.072

0

5.0

3.0

0.2

2.4

1.0

13

Guaro

611.4

1990

555

1184

48.6

0.024

64

7.8

2.9

2.5

7.4

3.7

14

Gudalmedina

184.6

1342

543

1222

49.3

0.025

90

4.0

2.9

0.7

4.2

1.7

15

Guadalhorce

3180.9

1703

522

1022

154.2

0.007

59

11.2

3.2

5.1

9.8

4.1

16

Fuengirola

129.5

993

264

749

23.4

0.032

0

15.0

4.6

1.9

6.8

2.5

17

Verde-Marbella

155.3

1746

664

1410

33.9

0.042

92

1.0

5.2

0.3

2.8

1.2

18

Guadaiza

49.2

1287

579

1270

21.2

0.060

78

1.0

5.2

0.5

2.8

1.2

19

Guadalmina

66.8

1242

536

939

29.0

0.032

80

1.0

5.2

1.6

5.4

2.6

20

Dos Hermanas

3.8

140

105

129

4.7

0.027

0

1.0

5.2

0.2

2.4

1.1

21

Guadalmansa

66.1

1091

513

904

23.4

0.039

71

1.0

5.2

0.8

3.9

1.9

22

Castor

20.4

1052

443

917

13.3

0.069

0

1.0

5.2

0.2

1.6

0.6

23

Padron

22.8

1360

397

978

12.7

0.077

0

1.0

5.2

0.6

3.4

1.4

24

Guadiaro

1489.5

1609

548

1398

101.0

0.014

3

4.2

4.0

5.8

11.7

4.2

25

Guadarranque

268.6

552

202

452

40.6

0.011

54

4.0

5.0

3.5

8.5

2.5

26

Palmones

316.5

759

235

401

45.9

0.009

50

4.0

5.0

5.2

10.9

3.6

River slope was calculated as the (main) river maximum height/river length ratio. Lithology and vegetation indexes were calculated after Probst (1992) and range from 1 (hardly erodible) to 40 (easily erodible), and from 0 (desertic or nil soil protection by vegetation) to 6 (forested or soil highly protected by vegetation), respectively. Grey shadings separate the Eastern (upper), Central (middle) and Western (lower) Andalusian watersheds.

Adra and Guadalfeo (1, 2 and 7 in Fig. 1), have most of their areal development at mid altitudes, while lower altitudes represent a relatively small part of their total basin area (Fig. 6A). Eastern Andalusian river systems feed well-defined delta systems with areas ranging from 0.2 km2 to 9.4 km2 (Table 2).

4.1.2. Central Andalusian watershed Both temperature and precipitation show halfway values between the Eastern and the Western areas. This subset includes primarily zones with very high erosion risk (mainly between river systems 10 and 14 in Fig. 1), but also sectors with zero or very low risk

480

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

Spanish Harbour Authority (Puertos del Estado, 1983–2004) show that Central Andalusian wave activity is slightly lower than in the surrounding areas.

Fig. 4. River slope vs. river length plot of Andalusian river systems (black triangles numbered after Fig. 1) compared with other river systems in the Iberian Peninsula flowing into the Mediterranean Sea. High river slopes and limited lengths highlight the torrential character of Andalusian river systems. Catalan river systems extend northwards of the Ebro River to the Pyrenean divide, while Levantine river systems comprise the region between the Ebro River and the study area.

(Moreira and Ojeda, 1992). Estimated sediment loss is predominantly very high, often larger than 300 t ha 1 yr 1 (CMA, 1997). Central Andalusian rivers are quite steep and most of them (all but the Guadalhorce Basin, the largest of the study area) follow rather straight paths. The reason is that the Betic mountains that constitute the source area are only between 3 and 30 km away from the shoreline. Only the three westernmost rivers (Guadalmedina, Guadalhorce and Fuengirola; 14–16 in Fig. 1) show a small step in their middle course. The longest rivers of this cluster (13 to 15 in Fig. 1) have most of their basin area between 300 and 700 m of altitude, while the shortest streams (8 to 12 and 16 in Fig. 1) have a better distributed areal coverage (Fig. 6B). A noticeable exception is the Fuengirola River basin, whose maximum hypsometric level is located between sea level and 100 m elevation. The ratio between delta area and basin area is lower than in Western Andalusian rivers (Fig. 5), possibly reflecting lower precipitation rates and consequent weaker transport of sediment to river mouths. Sediment dispersion by waves cannot explain this decrease as data from the

4.1.3. Western Andalusian watershed Precipitation is more abundant and uniform (i.e. less torrential) than in the Eastern and Central Andalusian watersheds. Besides, this area has a more widespread vegetal cover and its lithology consists of generally softer rocks. Erosion risk is also lower in this subset than in the Central and Eastern ones (Moreira and Ojeda, 1992). The estimated soil loss is of less than 50 t ha 1 yr 1 in most of the watershed (CMA, 1997). In Western Andalusia, basin area varies between 3.8 and 1489.5 km2, while river length ranges from 5 to 101 km. Altitudes are substantially lower than in the rest of the study area (maximum height is 1746 m and mean one 474 m). In general, rivers have a smoother axial gradient and a more variable longitudinal profile than the ones further east. Hypsometry plots show that river systems 23, 25 and 26 in Fig. 1 have most of their basin area between 0 and 100 m of altitude (Fig. 6C). River systems 17 to 22 in Fig. 1 have a strong torrential character and their hypsography is more evenly distributed. The Guadiaro River

Fig. 5. Delta area/basin area ratio of Andalusian river systems expressed as percentage, compared with Catalan and Levantine river sets of the Iberian Peninsula. The largest relative delta magnitude correspond to some of the Western Andalusian river systems.

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

(A)

(B)

481

(C)

Fig. 6. Hypsometry of (A) the Guadalfeo River (7 in Fig. 1) considered as representative of the most noticeable Eastern Andalusian rivers; (B) the Chillar Stream (10 in Fig. 1) showing the typical torrential morphology abundant in the Central Andalusian watershed; and (C) the Guadarranque River (25 in Fig. 1) from the westernmost part of the study area, with a smoother axial gradient that concentrates most of the basinal area on lowlands. Hypsometry is shown by the grey bar histogram referred to the lower horizontal axis. River longitudinal profiles are represented by the continuous black lines referred to the upper horizontal axis.

(24 in Fig. 1), in spite of being the largest of this subset, also has a very well distributed areal coverage. 4.2. Damming history Damming is a significant issue in the whole Mediterranean watershed, and river systems in the Iberian Peninsula are intensively regulated. The most dramatic case is the Ebro Basin, with 95.9% of its area now behind dams. However, river systems in the Andalusian Mediterranean watershed are comparatively much less affected. Only half of the 26 river systems considered are regulated by a total of 37 midsize to small dams (Table 3) mainly for irrigation and flood control. Energy generation by these dams is less than 2% of the Spanish hydroelectrical power production (MMA, 2000). The damming history of the Andalusian Mediterranean watershed can be divided into four stages, I– IV (Table 3 and Fig. 7). The initial stage (stage I) started in 1905 with the building of the small Llano de la Leche dam, followed by three new reservoirs constructed before 1925 (Fig. 7B). In stage II, extending till the 1970s, only two additional dams were finished (Fig. 7A), but just one of them, the Gaitanejo dam within the Guadalhorce basin (15 in Figs. 1 and 7), brought the regulated total area up to 2000 km2. The third stage was characterised by river damming

reactivation. Two sub-stages can be distinguished within that period, from 1971 to 1980 (stage IIIa) and from 1981 to 1990 (stage IIIb) (Fig. 7A). During the former, total reservoir capacity reached almost 600 hm3 and dams were only present in the Central and Western Andalusian watersheds. In stage IIIb, the number and capacity of reservoirs were doubled (Fig. 7A). Regulated area reached nearly 3400 km2 affecting for the first time the Eastern watershed (Fig. 7D). Stage IV extends from 1991 to the present. During the first half of the 1990s, 12 new small dams and reservoirs were built (Table 3). Nowadays, total reservoir capacity is about 1100 hm3, and almost 4800 km2 (representing 42% of the study area) are regulated. The Central Andalusian watershed is the most affected one, comprising more than 50% of the total regulated area. It is essentially due to the Guadalhorce River system. In contrast, the Western Andalusian subset is the least regulated one. With the exception of El Limonero and La Concepcion dams (Table 3), the remaining 35 dams are located in the middle and upper course of river basins. The Verde Marbella River (17 in Figs. 1 and 7) has the highest percentage of regulated area (92%), while the Guaro Basin contains more dams (9) than any other system. Due to the recent construction (within the last 35 yrs) of all but six of the dams, it is to be expected that

482

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

2

Adra

7

Guadalfeo

13

14

15

Guaro

Guadalmedina

Guadalhorce

17

Verde Marbella

18

Guadaiza

Distance to river mouth (km)

% Basin area regulated

Area regulated by dam (km2)

Dam height (m)

Capacity (Hm3)

Date of construction

Andarax

Altitude (m)

1

Isfalada

Lobera

1246.0

1986

0.38

23.0

33.03

1.5

82.92

Finana ˜

Bco. del Castanar ˜

1114.5

1995

0.21

28.0

36.10

3.2

75.81

Beninar

Grande de Adra

363.0

1983

68.20

87.0

528.97

70.5

20.80

Beznar

Izbor

485.0

1986

53.60

134.0

302.31

23.0

30.04

Rules

Guadalfeo

243.0

2003

112.00

130.0

1035.76

78.9

20.20

Stream

Reservoir

No. in Fig.1

River system

Table 3 Descriptors of the reservoirs in the study area

La Vinuela ˜

Zapaton

230.0

1986

170.00

96.0

116.27

19.0

21.75

Alcaucin

Alcaucin

264.0

1995

0.18

15.5

58.51

28.6

23.72

Almanchares

Almanchares

265.0

1995

0.07

21.1

12.16

30.6

18.74

13.92

32.8

20.34

13.0

48.00

32.8

18.35

Bermuza

Bermuza

265.0

1995

0.27

Granados

Rubite

265.0

1995

0.08

La Cueva

La Cueva

265.0

1995

1.00

32.1

81.41

46.2

34.08

Rubite

Rubite

267.0

1995

0.09

15.0

50.98

54.5

17.07

Seco

Alcaucin

248.0

1995

0.27

2.50

54.9

23.31

Solano

Solano

283.0

1995

0.80

30.0

58.28

64.4

33.18

Agujero

Guadalmedina

104.0

1924

5.00

165.75

89.8

5.37

El Limonero

Guadalmedina

104.0

1983

24.80

95.0

165.75

89.8

5.37

Cde. de Guadalhorce

Ardales

341.0

1921

84.00

74.1

269.48

8.5

71.02 66.97

Gaitanejo

Guadalhorce

302.0

1927

0.20

34.0

1731.26

54.4

Guadalteba

Guadalteba

362.0

1972

153.30

84.0

967.62

54.4

71.65

Guadalhorce

Guadalhorce

362.0

1973

125.80

74.4

480.45

54.4

71.72

Tajo de la Encantada

Guadalhorce

302.0

1978

4.30

36.0

1736.67

54.6

64.42

T. Encantada (sup.)

-

586.5

1978

3.00

38.2

-

54.6

65.74

El Tomillar

Arroyo Pelones

101.7

1995

2.92

47.0

5.04

54.8

17.36

Casasola

Campanillas

153.5

1999

23.64

76.0

190.94

60.8

24.02

La Concepcion

Verde

102.5

1971

57.00

89.5

142.34

91.7

5.69

Guadaiza

Guadaiza

131.5

1995

0.24

23.5

38.30

77.9

7.35

Llano de la Leche

La Leche

59.5

1905

0.20

20.0

4.86

7.3

4.64

Guadalmina

167.5

1995

0.17

21.0

48.85

80.4

10.49 11.52

19

Guadalmina

Guadalmina La Zagaleta

Senegal

239.0

0.25

14.0

2.71

80.4

21

Guadalmansa

Guadalmansa

Guadalmansa

156.0

1998

0.10

26.0

46.69

70.6

9.29

24

Guadiaro

Montejaque

Gaduares

693.0

1924

36.00

84.0

48.69

3.3

86.54

25

Guadarranque

Guadarranque

Guadarranque

78.0

1965

87.74

72.0

145.59

54.2

16.82

Deposito DD.1

Prior

30.0

1972

0.55

23.0

21.24

6.7

8.57

26

Palmones

Charco Redondo

Palmones

81.4

1983

81.60

71.7

96.13

37.1

23.63

Charco Redondo D.R.

Majadillas

80.0

1984

0.50

22.4

95.15

37.1

23.99

Valdeinfierno

Valdeinfierno

112.0

1985

0.15

13.7

19.98

43.4

27.21

La Hoya

Arroyo La Hoya

132.6

2000

22.36

50.5

26.30

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

most of the downstream impact on the corresponding deltas, i.e. the induced reduction in sediment volumes reaching river mouths, is still to come. Reservoir siltation and capacity reduction is a potential problem in the Andalusian Mediterranean watershed, as exemplified by neighbouring areas. For example, several dams of the nearest two basins westwards from the Palmones River (26 in Fig. 1), show a trapping efficiency of more than 96% (CHG, 1994). However, the impact of dams on deltas may likely be less severe than could be expected, because of the lack of dams in the lower and, partly, mid-river courses where erosion rates are still high (Moreira and Ojeda, 1992). 4.3. Water discharge Discharge and runoff peak December to February in most of the Andalusian Mediterranean river systems, as related to late fall and winter precipitation maxima. The only exceptions are the Guadalfeo and Guadalhorce rivers (7 and 15 in Fig. 1). This seasonal pattern is illustrated in Fig. 8. The delayed response of the Guadalfeo and Guadalhorce rivers is attributable to their relatively large basinal area penetrating deeply inland towards the high cordilleras where the influence of snowmelt is greater. However, the location of the gauging stations from where the data series have been obtained for the various river systems may influence the variability of the observed hydrological patterns. For instance, water discharge trends measured in upstream gauging stations such as Narila in the Guadalfeo course probably show a relatively instantaneous response to the mountainous climate regime. In general, water discharge trends show a zonal behaviour matching with the Eastern, Central and Western Andalusian watersheds classification. Nevertheless, some extreme events affecting the whole study area have been identified. Widespread discharge maxima occurred in 1945–1947, 1969–1970, 1989– 1990 and 1995–1998. Historical water discharge and runoff maxima exceed mean values 3 to 27 times,

483

with rivers from the Central Andalusian watershed being in general the most extreme ones. During the 1980s, a generalised decrease in water discharge occurred, what we link primarily to reservoir effects as there were no signs of decline in rainfall or rise in evapotranspiration. Although the spectral analysis of mean monthly discharge data reveals only the typical seasonal periodicity, the analysis of daily discharge data from selected rivers illustrates at least four significative supra-annual periodicities of 8–11, 5–6.5, 3–4 and 1.1–1.4 yrs. The 3–4 yrs periodicity is the most common one. In addition, four relatively long and continuous monthly and annual discharge time series were compared to the NAO index. These series came from the Andarax, Guadalfeo, Algarrobo and Guadiaro river systems (1, 7, 12 and 24 in Fig. 1) and lasted 38, 47, 54 and 76 yrs, respectively. Annual discharge data showed a better correlation with NAO than monthly discharge data. The best correlation index between NAO and discharge curves (R = +0.66) was that of the Andarax, followed by the Guadiaro (R =  0.42), Algarrobo (R =  0.35), and Guadalfeo (R =  0.11). Although these coefficients of correlation are low, a negative relationship between fluvial discharge and the NAO index can be observed in the last three cases (Fig. 9). On the other hand, it was graphically observed that maximum monthly discharge events often occur when the NAO index undergoes the most extreme variations. This has been the case for the 1947 and 1989–1990 peak discharges of the Guadiaro, the 1943, 1947 and 1997 of the Guadalfeo, the 1947 and 1997 of the Algarrobo, and the 1947, 1979–1980 and 1990 of the Andarax. Water budget calculations allow us to discriminate between years in which infiltration caused a net water loss to fluvial water discharge (i.e. an increase of the groundwater reservoir reflected by negative infiltration values) and years with a net water gain for rivers (represented by positive infiltration values) (Table 4). It must be noticed that riverine discharge series were

Notes to Table 3: Agujero reservoir on the Guadalmedina River system was incorporated into the subsequent El Limonero reservoir in 1983. Area regulated by dam corresponds to the upstream basin area from each single dam, whilst the percentage of basin area regulated refers to the overall evolution of each river system. Numbers in bold indicate the accumulated area regulated in each basin percentage. See also Figs. 1 and 7. Grey shadings separate the Eastern (upper), Central (middle) and Western (lower) Andalusian watersheds.

484

2

B

Regulated area: 489 km (4 %)

Stage I (until 1925)

C

Stage IIIa (until 1980)

1

24

25-26

30

1200

A

IIIb I

20

II

1

2

IIIa

IV

800 400

10 0

0

D

Stage IIIb (until 1990)

14 15

13 12 11 10

65

8-9 7

3-4

2

3

3-4

Total capacity (hm )

18-23

40

65

8-9 7

1900-05 1906-10 1911-15 1916-20 1921-25 1926-30 1931-35 1936-40 1941-45 1946-50 1951-55 1956-60 1961-65 1966-70 1971-75 1976-80 1981-85 1986-90 1991-95 1996-00 2001-03

16

Number of reservoirs

17

13 12 11 10

16

17 18-23 24

2

25-26

Regulated area: 2265 km (20 %)

Stage IV (present)

Central watershed

E

Eastern watershed

Western watershed 1 14 15 17

13 12 11 10

65

8-9 7

3-4

1

2

14 15

16

17

18-23

6 5 3-4

8-9 7

2

16

18-23

24

25-26

13 12 11 10

24

2

Regulated area: 3362 km (30 %)

25-26

2

Regulated area: 4761 km (42 %)

0

15

30

45 km

Fig. 7. Damming history (A) of the study area in four stages: (B) 1905–1925, (C) until 1980, (D) until 1990, and (E) 1990–nowadays. Graph A represents reservoir quantity as an histogram and capacity growth as a continuous black line. The numbers of the river basins and the coordinates’ ticks correspond to those of Fig. 1. Basin area of the studied river systems is light grey-shadowed and regulated area behind dams is dark grey-shadowed. White lines mark the boundaries between adjacent river basins. Continuous black lines correspond to main rivers (thicker) and tributaries (thinner). Stages’ names correspond to those discussed in the text.

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

14 15

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

485

Fig. 8. Mean monthly discharge histograms and calculated mean monthly runoff curves of Guadalfeo and Palmones rivers (7 and 26 in Fig. 1) based on Narila and Charco Redondo gauging stations, respectively. The former time series is 41 years long while the latter is 19 years long.

obtained only from specific gauging stations and not as a basin-wide runoff average. Some local trends in infiltration time series have been observed. The 1970s and 1980s were decades when there was a widespread loss of superficial water throughout infiltration, while from 1995 to 1998 river systems tended to gain groundwater. Commonly, river systems receive more soilwater than they lose, as denoted by the mean infiltration values in Table 4. However, some episodes of groundwater loss have occurred. For example, the Algarrobo River system (12 in Fig. 1) showed an infiltration rate of  600.46 mm yr 1 in 1963–1964,

what we link to the excessive mean rainfall (1106 mm yr 1) of 1962–1963. Mean PET and AET values in Spain, 862 mm yr 1 and 464 mm yr 1, respectively (MMA, 2000), can be compared with the maximum, minimum and mean values obtained for each Andalusian watershed (Table 4). Since PET depends only on temperature and solar radiation it is spatially homogeneous, while AET, taking rainfall data into account, shows differences between the three Andalusian watersheds. AET peaks are synchronous across the entire study area, although rivers from the Eastern watershed show

Fig. 9. Time series of the Guadiaro River (24 in Fig. 1) mean annual discharge (continuous line) and the mean annual NAO index (dashed line). A negative relationship between the two curves can be observed, although the correlation index is rather low (R =  0.42). Guadiaro discharge data are from the Corchado gauging station (Table 1).

486

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

Table 4 Hydric budget (average results in mm yr 1) of the three Andalusian watersheds including potential evapotranspiration (PET), actual evapotranspiration (AET) and ground infiltration (I) maximum, minimum and mean values

PET

AET

I

Maximum Minimum Mean Maximum Minimum Mean Maximum Minimum Mean

Eastern watershed

Central watershed

Western watershed

980.33 681.06 818.69 759.62 109.17 395.21 575.25 35.14 63.75

1121.77 665.50 922.57 827.46 184.97 494.92 395.10 600.46 35.86

1052.94 762.04 906.38 912.28 237.72 607.35 972.61 186.30 82.29

Negative infiltration values in the Central watershed indicate a net surface water loss towards the groundwater reservoir.

lower values due to their scarcer rainfall input. Historical trends of individual river systems show that mean AET either stayed constant or slightly decreased. 4.3.1. Eastern Andalusian watershed Hydrological time series were available for the three largest rivers of this watershed (Andarax, Adra and Guadalfeo; 1, 2 and 7 in Fig. 1) (Fig. 10A). These rivers have the least and most constant runoff values of the analysed fluvial systems, which denotes that they carry a low volume of water related to their basin areas. Time series from Fuentes de Marbella station show that the Adra has the greatest discharge (Fig. 10A), although its basin is neither the largest nor the rainiest. The possible reasons are that the Adra drains a significant area of Sierra Nevada and that its gauging station is located at the mid-course, while in the other Eastern basins data have been gathered from the upper course. The only dam in the Adra basin is the Beninar dam, which was constructed in 1983, just when our hydrological data start. Between the three rivers analysed in this subset, the Adra is the only one that does not dry up during the summer months. The Guadalfeo River water discharge has markedly decreased since the first half of the 20th century (Fig. 10A). This cannot be due to damming, since the Narila gauging station is upstream of the main dam interrupting the river course, which in addition was constructed only in 2003. Possible causes are the natural increasing aridness trend of the study area

reinforced by anthropogenic influence such as the growing agriculture production, which has been quadruplicate since the 1970s. Time series from the Canjayar gauging station show that the Andarax River has also diminished its overall water discharge but to a much lesser extent than the Guadalfeo (Fig. 10A). Regulation might have promoted this tendency but it is not likely to be the only reason since only two small dams was built in 1986 and 1995 in two headwall tributaries. Water discharge of the Eastern Andalusian rivers has a relatively slow response to rainfall. We suggest that its main control comes throughout groundwater reserves. Maximum peaks usually occur one to several months after the strongest rainstorms, when a smaller downpour refill the water reserve. In addition, fluvial response is not proportional to rain and, therefore, precipitation is not its only controlling factor. Secondary controls are likely to be (a) the variable soil water reserve accumulated by different rock types within the same river basin (from sand and arenite to micacite and slate), and (b) the extreme temperatures and consequent changing evaporation rates in such a semiarid climate. 4.3.2. Central Andalusian watershed Hydrological information of six river systems of the Central Andalusian watershed was available (Fig. 10B). The Guadalhorce River is the largest one and also carries the greatest water volume, with maximum monthly discharge values up to 130 m3 s 1. Its mean discharge was considerably greater during the first half of the 20th century (9.0 m3 s 1 approximately) than in the last two decades (about 5.5 m3 s 1). Data from El Chorro-Gaitanejo gauging station, located upstream the existing dams, do not allow to assess the impact of regulation on such flow reduction. The remaining five rivers follow a common seasonal trend with a typical double-peaked winter. Some of them, such as the Verde Almun˜ecar and Patalamara (8 and 11 in Fig. 1), in spite of their small size, do not dry up completely during summer. They all respond directly, though not proportionally, to rainfall with a delay of less than a month. Therefore, underground water accumulation in these river basins is not as important as in the Eastern Andalusian cluster. The Patalamara and Algarrobo (11 and 12 in Fig. 1), which are the two smallest fluvial systems in this group,

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

487

A

B

Fig. 10. Monthly water discharge time series from (A) the three main rivers in the Eastern Andalusian watershed, Andarax (black), Adra (red) and Guadalfeo (green) (1, 2 and 7 in Fig. 1), and (B) six rivers in the Central Andalusian watershed, Verde Almun˜ecar (purple), Patalamara (yellow), Algarrobo (green), Guaro (red), Guadalmedina (blue) and Guadalhorce (black) (8 and 11–15 in Fig. 1). Data from the latter river are represented using a different vertical scale (right vertical axis) to ease comparison. Horizontal axes and the overimposed grid help to identify hydrological years and corresponding discharge values. The gauging stations from where these data were taken are described in Table 1. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

show the highest runoff values (3.9 and 4.6 l km 2 s 1 on average) and tend to lose water by infiltration. 4.3.3. Western Andalusian watershed Hydrological time series are available for the three westernmost rivers, the Guadiaro, Guadarranque and Palmones (24, 25 and 26 in Fig. 1). As in the previous group, the river system with the largest basin area (the Guadiaro) shows the greatest discharge. It varies from 100 m3 s 1 in winter months (with peak values in February) to nearly drought conditions during the summer (Fig. 11). Western Andalusian rivers show the greatest runoff of the study area. The maximum mean monthly value (30 l km 2 s 1) is attained in December by the Guadarranque. Western Andalusian discharge time series are rather similar and no significant long-term trends are

discernible. The reaction of these rivers to rainfall is more rapid and proportional than in the other Andalusian subsets (Fig. 11) and, as a result, their water discharge series show a large intra-annual variability including frequent summer droughts. It seems that these river systems do not gain as much water by infiltration as the preceding ones. 4.4. Sediment discharge Suspended sediment load has been calculated for the 12 river systems included in Tables 1 and 5 using the empirical erosion models of Dendy and Bolton (1976), Jansen and Painter (1974), and the 4- and 5variables Probst’s (1992) models (see Section 3.2). Except for the first, the results obtained from the other three models show similar and synchronized

488

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

Fig. 11. Mean monthly Guadiaro River discharge and mean monthly precipitation on its basin since hydrological year 1942–1943. The dashed horizontal line represents the averaged mean discharge of this river. Rains are relatively abundant and directly control fluvial discharge. Fluvial droughts are usual during the summer while winter discharge peaks can reach 100 m3 s 1. Data come from Corchado gauging station and Ronda (central) meteorological station (Table 1).

fluctuations. We conclude that Dendy and Bolton’s (1976) forecast is inappropriate to simulate flash floods and the immediate sediment discharge peaks characteristic of the study area, likely because this model was based on deposition processes inside reservoirs. In order to verify the other three models, calculated results were checked against (i) soil erosion rates from an area close to the Andarax River basin (1 in Fig. 1) (Canton et al., 2001), (ii) sediment yield data from Murcian catchments, a few kilometers northeastwards of the study area (Verstraeten et al., 2003), and (iii) selected global sediment load records (Walling and Fang, 2003). After comparison with those data, the 4-variable multiple regression model of Probst (1992) proved to be the most trustworthy for our input data. Sediment yield time series usually show no significant trends, although sometimes a slight decrease or increase can be inferred (Fig. 12). In general,

sediment discharge series are similar within the same Andalusian cluster. Sediment yield averages are of the same order of magnitude for the Central and Eastern Andalusian river systems, but they are smaller in the Western Andalusian watershed. According to the average values in Table 5, the Adra River transports more sediments per area than any other system, while the Andarax would be the most competent river (the one transporting more sediment per water discharge, i.e. kg m 3). However, the highest punctual sediment yield value (710 t km 2 yr 1) was attained by the Algarrobo Stream in 1963–1964. Overall maximum loads for Andalusian river systems correspond to hydrological years 1962–1963 and 1996–1997, while minimum loads occurred in 1994–1995. Comparison between river systems is very useful to relate sediment discharge to basin parameters, for instance to investigate the relative importance of basin

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

489

0.4 1.0 0.6 0.3 0.2 0.3 0.8 0.2 7.4 10.8 2.6 0.3

0.2 1.3 0.4 2.8 3.9 4.6 1.3 1.0 2.3 7.2 9.5 1.1

4.1 16.3 6.9 5.3 0.6 3.8 11.6 5.0 130.3 104.5 26.6 6.3

1.9 21.7 5.3 52.2 12.2 58.1 18.9 27.1 41.0 70.2 99.0 19.8

5.7 4.8 2.7 0.4 0.1 0.4 1.6 0.4 2.8 1.1 0.1 0.1

Mean sediment yield (t km–2 yr–1)

Mean sediment load (kg s–1)

Runoff historical maximum (l km–2 s–1)

273.4 375.4 536.9 578.5 531.9 523.7 482.0 459.8 393.5 477.4 666.4 678.2

Discharge historical maximum (m3 s–1)

13.5 16.7 17.2 19.4 19.4 18.9 18.3 18.3 14.8 16.5 19.3 19.3

Mean runoff (l km–2 s–1)

318.9 445.1 717.3 747.5 657.3 657.3 593.9 562.0 497.5 613.7 1035.9 1049.6

Mean discharge (m 3 s–1)

Mean actual evapotranspiration (mm yr–1)

Andarax Adra Guadalfeo Verde Patalamara Algarrobo Guaro Guadalmedina Guadalhorce Guadiaro Guadarranque Palmones

Mean annual temperature (ºC)

1 2 7 8 11 12 13 14 15 24 25 26

Mean precipitation (mm yr–1)

No. in Fig.1

River system

Table 5 Meteorological and hydrological characteristics of the 12 river systems for which suspended sediment discharge has been calculated

81.5 201.4 65.1 120.9 66.5 176.0 84.1 62.5 28.3 23.9 11.7 10.2

Gauging and meteorological stations are described in Table 1 and located in Fig. 1. Grey shadings separate the Eastern (upper), Central (middle) and Western (lower) Andalusian watersheds.

area. This is well illustrated by the 65 km2 Algarrobo System, which shows the second highest mean annual sediment yield, and the 1500 km2 Guadiaro Basin, with one of the lowest sediment yield values (Table 5). The Algarrobo is not regulated, and the Guadiaro has only one reservoir (Table 3) that can be disregarded since it only affects about 3% of the basin area. In the period 1942–2000, the Algarrobo would have transported 11 400 t yr 1 (6.5
105 t total) of suspended sediment towards the ocean, while the Guadiaro would have carried 35 563 t yr 1 (20
105 t total). In summary, a river system like the Guadiaro, 23 times larger than the Algarrobo, would have supplied a total sediment load to the ocean only three times greater.

5. Discussion 5.1. Hydrological behaviour Modelled water budget based on precipitation and fluvial discharge time series shows that contributions

from underground water reserves play a principal role in Eastern Andalusian river systems and that such importance decreases westwards. This can be explained by the porous nature of the soil types and the aridness of the Eastern zone. Actual evapotranspiration also rises from east to west, where the climate is wetter and a denser vegetation cover is present. Synchronous peaks in evapotranspiration along the whole study area were also found. Infiltration is, however, highly variable amongst the three Andalusian watersheds. Although a long-term regional trend shows up, each river basin presents an individual behavior with gains or loses of interstitial water that do not follow any supra-basinal pattern. Domingo et al. (2001) interpreted that the estimated deficit of precipitation compared with actual evapotranspiration in arid zones of the easternmost part of the study area is compensated by infiltration of channel flow during flash floods originating in the upper part of the catchments. Peak runoff values are reached by the rivers in the Western Andalusian watershed, where discharge is more directly dependant on rainfall and where seaso-

490

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

Fig. 12. Sediment yield time series of the Guadalfeo (A), Verde Almun˜ecar (B), Algarrobo (C) and Guadiaro (D) river systems (7, 8, 12 and 24 in Fig. 1, respectively). Data were calculated after the 4-variables multiple regression model of Probst (1992). Algarrobo shows the highest values, followed by Verde Almun˜ecar, Guadalfeo and Guadiaro. Continuous black lines show the long term mean sediment yield trends. Guadalfeo, Algarrobo and Guadiaro river systems underwent a minor decrease during the last decades while Verde Almun˜ecar experienced a slight increase. Overall sediment load variations seem to be of low significance for the time period considered in our analysis.

nal variability is greater. This makes summer droughts more frequent and severe than in the watersheds to the east in spite of the denser vegetation cover and the overall wetter weather. Similar water discharge trends between nearby rivers (e.g., the Guadiaro and Guadarranque or the Guadiaro and Palmones) demonstrate a zonal behaviour coincident with the three Andalusian watersheds classification (see Section 4.3). However, regional (i.e. more widespread) tendencies and events have also been found. Most of the water discharge time series indicate a long-term decrease. Natural reasons have been suggested to explain this trend, which is in agreement with the general tendency towards aridity recorded in the Western Mediterranean since 5.4 ky BP (Goy et al., 2002). Nevertheless, neither rainfall nor evapotranspiration data show sufficiently significant trends throughout the last decades. During specific time intervals, evapotranspiration seems to play a stronger role than precipitation in governing water discharge. For instance, low discharge values during the 1980s were determined more by higher temperatures and related higher evapotranspiration than by low precipitations. In addition, the Mediterranean flu-

vial systems of the Iberian Peninsula could be markedly impacted by global warming (Ayala-Carcedo, 1996). However, separating the impact of recent climate change from the natural interannual variability and longer-term tendencies is far from obvious, specially when direct anthropogenic influences on river systems become more and more significant. Southern Spain and the Alboran Sea region are influenced by the Atlantic oceanographic and atmospheric circulation. Goy et al. (2002) deduced a decadal periodicity in coastal deposits from SE Spain and suggested a direct relation with NAO and solar activity variations. Such an intricate relationship between natural processes in the study area is illustrated by the variable correlation between supra-annual fluvial discharge periodicities and the NAO index (see Section 4.3). Although spectral analyses are not statistically too consistent, the fact that the strongest discharge events occur hand in hand with the largest NAO fluctuations is rather significant. We interpret that low correlation indexes reflect the relative larger influence of local (small scale) environmental controls compared with global (large scale, such as the NAO) ones. Some individual analyses show that

mean annual discharge time series and the NAO index evolution are often opposite (Fig. 9), meaning that the greatest Andalusian water discharge events could be favoured by negative NAO situations. In fact, negative NAO situations are known to promote wet climate over the Iberian Peninsula. Some water discharge time series allow us to compare pre- and post-damming conditions. We have observed that (i) reservoirs do not avoid peak discharge events, (ii) the delay between rainfall and fluvial discharge has been confirmed as essentially independent from river regulation, and (iii) when a decreasing water volume trend through time is appreciable, it could be related to natural controls (e.g. temperature or evapotranspiration trends, rainfall, etc.). Several factors could help explaining this behaviour, such as the relatively low height of dams and small size of reservoirs, their preferential location in upper river courses, and the dominance of short-lived, naturally governed processes over man-made influences. Therefore, the effect of dam building is barely noticeable at river mouths and in northern Alboran deltas/prodeltas. 5.2. Sedimentological behaviour Although a reduction in sediment load could be in agreement with the described general decreasing tendency in water discharge through time, calculated sediment discharge time series for each river system do not change significantly, showing only in some cases very subtle trends (Fig. 12). Globally, most of the world’s rivers show either declining loads or no relevant trends with only very few rivers showing increasing tendencies both in runoff and sediment load (Walling and Fang, 2003). In our study area, 5 of the 12 analysed rivers (numbers 1, 12, 13, 14 and 24 in Fig. 1) show a minor decrease in their annual sediment yield, 4 of them (7, 8, 15 and 26 in Fig. 1) undergo a minor increase, and the remaining 3 (2, 11 and 25 in Fig. 1) have too short time series (less than 12 annual values) to identify long-term trends. A linear relationship between measured runoff (mm yr 1) and estimated sediment yield (t km 2 yr 1) can be observed in most of the studied river systems (Fig. 13). When the water discharge increases, for instance due to storm events, the relative fluvial sediment transport is enhanced.

491

Sediment yield (t km-2 yr-1)

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

-1

Runoff (mm yr )

Fig. 13. Sediment yield and runoff plot of the Palmones River (26 in Fig. 1), from the Western Andalusian watershed. Similar linear correlations are observed in most of the studied river systems.

The Eastern and Central Andalusian watersheds show similar sediment yield values, while these are lower in the Western area. The highest values have been found in the Adra River (2 in Fig. 1 and Table 5), directly draining part of Sierra Nevada in the Eastern watershed, and in two small streams from the Central watershed called the Algarrobo and Verde Almun˜ecar (12 and 17 in Fig. 1) (Table 5 and Fig. 12). Rivers of the Eastern watershed carry the highest mean sediment load (Table 5), likely because the semi-desertic character of their catchments promotes the availability of sediment particles to be transported. Dams and reservoirs are traditionally thought to be the most important anthropogenic influence on fluvial sediment fluxes. According to Vorosmarty et al. (2003), European regulated river basins display the highest mean sediment retention (50%) compared to the 25–30% of global sediment flux trapped in large artificial impoundments. Nevertheless, in the Andalusian systems, dams do not seem to have major influence on river discharge. For example, the strongly regulated Guadalhorce River basin still delivers massive suspended sediment discharges to the sea following strong storm events (Fig. 2). However, we must notice that most of the dams and reservoirs in the Andalusian rivers are relatively small and young and, therefore, the continuation of dam building in the near future could change appreciably the present situation. Data mostly from public agencies (Moreira and Ojeda, 1992; IEA, 2003) state that part of the study

492

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

area undergoes important erosion problems caused by the combination of soil composition, steep slopes, insufficient vegetation cover, climate and inappropriate agricultural practices. Soil erosion risk grows from west to east. Actions to prevent hydric erosion and consequent damages by means of reforestation and dike construction into streams are on-going (e.g., SGMA, 2001). Although at present human influence over Andalusian riverine sediment load does not seem to be as important as natural factors, we suggest that this could soon change. The impact of progressing desertification, already a key problem in Andalusia, should be assessed. It is expected that desertification will enhance the release of sedimentary particles able to be transported by rivers to delta and prodelta areas even with decreasing water discharge trends. 5.3. Formation of deltas and prodeltas A comparative analysis of Spanish Mediterranean delta systems (namely the delta area/basin area ratio, Fig. 5) indicates that, in terms of sediment transport, Andalusian river systems are quite efficient independently of the small size of their catchments. This confirms that the importance of small basins as sediment feeders to the ocean is relatively high. The importance of small mountain rivers to global sediment discharge to the ocean compared with that from vast watersheds has been widely recognised (Milliman and Syvitski, 1992; Farnsworth and Milliman, 2003). In spite that all the studied river systems fall into the standard category of small rivers, the smallest amongst them usually carry the largest sediment loads. This is well illustrated by the Algarrobo and Guadiaro river systems (12 and 24 in Fig. 1). The latter is around 23 times larger than the former but its sediment load is only 3 times greater. A similar negative linear relationship between sediment yield and basin area was found in rivers from southeastern Asia and the islands of Oceania draining young mountain chains (Farnsworth and Milliman, 2003). Such relation was interpreted as a result of (i) the ability of these rivers to respond to flash floods and (ii) the lack of areas within river basins to store flood-driven sediment. These two characteristics are observed in the Andalusian watersheds, whose hypsometry

demonstrates that only a few sites could act as intermediate sediment sinks, and whose discharge is often governed by flood events. In other words, most of the sediment load the Andalusian rivers carry is funneled directly to deltas and prodeltas. For instance, following stormy episodes and flash floods in 1973, quasiinstantaneous sediment accumulations of the Albun˜ol del Tranco River (4 in Fig. 1) extended 200 m offshore the river mouth and were rapidly reshaped (Romero Cordon et al., 2003). Sediment discharge during specific flood events can be volumetrically more important than total sediment discharge during the rest of the year. But not all the continental material remains near the river mouth. On the one hand, the narrowness of the northern Alboran continental shelf (2–10 km) likely favors a quick transfer to deeper sedimentary environments. Fabres et al. (2002) have convincingly demonstrated the close relationship between particle flux peaks on the deep continental slope south of Malaga and Guadalhorce and Guadiaro river floods (15 and 24 in Fig. 1). This is well illustrated by certain Central Andalusian streams (9–12 in Fig. 1) that have the smallest deltas related to their basin area (Fig. 5), despite having undergone numerous flood events. Sediment transfer towards deeper marine areas seems to be the best explanation, as in the Central watershed the continental shelf is the narrowest of the study area (3.5 km mean width) and the coastline is relatively protected against wave activity. On the other hand, after flood events large suspended plumes spread over broad areas off river mouths for days (Fig. 2), being affected by mesoscale structures and coastal currents. The dispersion of such plumes must play a major role in the development of muddy delta-attached or detached prodeltas and other mud-covered areas beyond the continental shelf. We have demonstrated that water and sediment discharge regimes in the Mediterranean watershed of Andalusia, as well as subsequent delta evolution, are controlled not only by the regional weather patterns, including a marked seasonal variability, but also by the orography and hypsometry of river basins dominated by pronounced slopes. We also interpret as significant factors the generally scarce vegetation cover with an eastward tendency towards desertification, and the increasing damming pressure.

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

6. Conclusions All Andalusian river systems flowing into the Alboran Sea and feeding well-defined deltas have been systematically studied to understand the main processes governing the transfer of water and sediment from continent to ocean, and the historically recent human impacts on such transfer. They are 26 relatively small, event-driven rivers flowing through one of the hottest, driest and less vegetated regions in the Iberian Peninsula. Water and sediment discharge analyses were carried out for the 12 river systems with available data series. Based on the river slope, river length and basin shape parameters the authors have classified 17 streams as torrents, most of them unregulated. The torrential character is reflected on water discharge and sediment transport patterns. Usually, the smallest streams show the greatest sediment yield values and the best distributed areal coverage (which implies a lack of sink areas for fluvial sedimentation). Actually, the studied deltas are the largest of the Spanish Mediterranean coast with respect to their basin areas, fact that could not be explained by the wave activity. The role of the studied rivers as sediment feeders to the shoreline and ultimately to the ocean is therefore relatively important, being even enhanced by the narrowness of the continental shelf and by the broad dispersion of suspended sediment plumes. Only half of the 26 river systems considered in this paper are nowadays regulated. Damming has been especially active since the 1970s, and has led to a present regulation of 42% of the study area. To date the effect of dam building has been barely noticeable on river mouths. Existing impoundments are not efficient in flattening peak discharge events and the time lag between rainfall and fluvial discharge has been confirmed as independent from river regulation. Whenever a decreasing trend in river discharge occurred, it was longer-term and probably naturally forced. Several factors could explain the minor impact of dams, including their location (mostly in mid and upper courses), their relatively small size, and the dominance of fast-occurring natural processes over man-made impacts. Eastern, Central and Western Andalusian watersheds have been distinguished based on geomorphological, climatological and hydrological characteristics.

493

Eastern Andalusia is climatically more changeable and dry than the Central and Western watersheds. Soil erosion risk and torrential character grows from west to east. This is likely the reason why Eastern watershed carry the highest mean sediment load and why Western rivers show the lowest sediment yield values. Water budget calculations demonstrate that (a) underground water reserves play a principal role in fluvial discharge, specially in the arid Eastern Andalusian watershed, (b) actual evapotranspiration increases from east to west although some regionally synchronic peaks have also been observed, and (c) infiltration is highly variable from one watershed to another. A decreasing trend has been observed in most water discharge time series during the last decades, which has been attributed to natural factors. In contrast, other available data series such as rainfall, calculated evapotranspiration and calculated sediment load do not show consistent trends throughout the same time period. However, a linear relationship between runoff and estimated sediment yield is commonly observed. The spectral analysis of daily discharge data points to a quite consistent periodicity of 3–4 yrs and other supra-annual cycles. This suggests a possible link between water discharge and NAO fluctuations. The strongest fluvial discharge events occur simultaneously with the largest NAO fluctuations, and some mean annual discharge time series seem to be opposite to the evolution of NAO. However, correlation indexes are low, meaning that local (small scale) factors are much more important than hemispheric ones (NAO) governing discharge. Only extreme NAO values may have an influence on the longterm development of deltas and prodeltas fed by Andalusian Mediterranean rivers.

Acknowledgements The authors acknowledge the information provided by the bConfederacion Hidrografica del SurQ, bMinisterio de Medio AmbienteQ and CEDEX, as well as the MODIS images supplied by GSFC/ DAAC (NASA). We also thank Victor Centella for his invaluable technical support. C. Liquete benefited from a fellowship attributed by the Spanish

494

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495

bMinisterio de Ciencia y TecnologiaQ. This research work has been supported by the Concerted Action bEuropean Coordination on Mediterranean and Black Sea ProdeltasQ (EURODELTA; EVK3-CT2002-20001), the Spanish RTD project PRODELTA (REN2002-02323) and the European EUROSTRATAFORM project (EVK3-2002-00079). GRC Geociencies Marines is funded by Generalitat de Catalunya excellency research grants program. References Ayala-Carcedo, F.J., 1996. Reduccion de los recursos hidricos en Espan˜a por el posible cambio climatico. Tecno Ambiente 64, 43 – 49. Canals, M., Arnau, P., Liquete, C., Colas, S., Casamor, J.L., 2004. Catalogue and Data Set on River Systems from Mediterranean Watersheds of the Iberian Peninsula. Univ. Barcelona. 220 pp. Canton, Y., Domingo, F., Sole-Benet, A., Puigdefabregas, J., 2001. Hydrological and erosion response of a badlands system in semiarid SE Spain. J. Hydrol. 252, 65 – 84. CHG (Confederacion Hidrografica del Guadalquivir), 1994. Redaccion de la Propuesta del Plan Hidrologico del Guadalete-Barbate, 49 pp. CMA (Consejeria de Medio Ambiente, Junta de Andalucia), 1997. La problematica asociada al suelo y al paisaje. Plan Andaluz de Medio Ambiente 1997–2002, pp. 81 – 90. CMA (Consejeria de Medio Ambiente, Junta de Andalucia), 1999. Mapas de Usos y Coberturas Vegetales de Andalucia 1:50.000. Cruzado, A., 1979. Climatology and hydrology of the Mediterranean Region. Report on the state of pollution in the Mediterranean Sea, UNEP, Chap. 2. Dendy, F.E., Bolton, G.C., 1976. Sediment yield–runoff–drainage area relationships in the United States. J. Soil Water Conserv. 31, 264 – 266. Domingo, F., Villagarcia, L., Boer, M.M., Alados-Arboledas, L., Puigdefabregas, J., 2001. Evaluating the long-term water balance of arid zone stream bed vegetation using evapotranspiration modelling and hillslope runoff measurements. J. Hydrol. 243, 17 – 30. Estrada, E., Ercilla, G., Alonso, B., 1997. Pliocene–Quaternary tectonic-sedimentary evolution of the NE Alboran Sea (SW Mediterranean Sea). Tectonophysics 282, 423 – 442. Fabres, J., Calafat, A., Sanchez-Vidal, A., Canals, M., Heussner, S., 2002. Composition and spatio-temporal variability of particle fluxes in the Western Alboran Gyre, Mediterranean Sea. J. Mar. Syst. 33–34, 431 – 456. Farnsworth, K.L., Milliman, J.D., 2003. Effects of climatic and anthropogenic change on small mountainous rivers: the Salinas River example. Glob. Planet. Change 39, 53 – 64. Fernandez-Salas, L.M., Lobo, F.J., Hernandez-Molina, F.J., Somoza, L., Rodero, J., Diaz del Rio, V., Maldonado, A., 2003. High-resolution architecture of late Holocene highstand

prodeltaic deposits from southern Spain: the imprint of highfrequency climatic and relative sea-level changes. Cont. Shelf Res. 23, 1037 – 1054. Fontbote, J.M., Vera, J.A., 1983. La Cordillera Betica. In: Rios, J.M. (Ed.), Geologia de Espan˜a, Tomo II. IGME, Madrid, pp. 205 – 342. Fournier, F., 1960. Climat et Erosion. Presses Universitaires de France, Paris. 201 pp. Galindo-Zaldivar, J., Jabaloy, A., Gonzalez-Lodeiro, F., Aldaya, F., 1997. Crustal structure of the central Betic Cordillera (SE Spain). Tectonics 16, 18 – 37. Garcia, A.F., Zhu, Z., Ku, T.L., Sanz de Galdeano, C., Chadwick, O.A., Chacon Montero, J., 2003. Tectonically driven landscape development within the eastern Alpujarran Corridor, Betic Cordillera, SE Spain (Almeria). Geomorphology 50, 83 – 110. Gimenez, J., Surin˜ach, E., Goula, X., 2000. Quantification of vertical movements in the eastern Betics (Spain) by comparing levelling data. Tectonophysics 317, 237 – 258. Goy, J.L., Zazo, C., Dabrio, C.J., 2002. A beach-ridge progradation complex reflecting periodical sea-level and climate variability during the Holocene (Gulf of Almeria, Western Mediterranean). Geomorphology 50, 251 – 268. Heburn, G.W., La Violette, P.E., 1990. Variations in the structure of anticyclonic gyres found in the Alboran Sea. J. Geophys. Res. 95 (C2), 1599 – 1613. Hopkins, T.S., 1985. Physics of the sea. In: Margalef, R. (Ed.), Western Mediterranean. Pergamon Press, Oxford, pp. 100 – 125. IEA (Instituto de Estadı´stica de Andalucia), 2003. Andalucia: Datos Basicos, Sevilla, 87 pp. IGME (Instituto Geologico y Minero de Espan˜a), 1969. Mapa Litologico de Espan˜a 1:500.000 sheet 3, Madrid. IGME (Instituto Geologico y Minero de Espan˜a), 1972. Mapa Geologico de Espan˜a 1:200.000 sheets 82, 83, 84 and 87, Madrid. Jansen, J.M.L., Painter, R.B., 1974. Predicting sedimentary yield from climate and topography. J. Hydrol. 21, 371 – 380. Kettner, A.J., Syvitski, J.P.M., 2003. Sediment flux from Apennine rivers: influence of reservoirs. COMDELTA, Aix-En-Provence, 26–28 Oct 2003, Abstr. Book, pp. 56 – 57. Kineke, G.C., Mullenbach, B., Geyer, W.R., 2003. Sediment dispersal on the shallow Apennine margin. COMDELTA, Aix-EnProvence, 26–28 Oct 2003, Abstr. Book, p. 58. Legates, D.R., Willmott, C.J., 1990a. Mean seasonal and spatial variability in gauge-corrected, global precipitation. Int. J. Climatol. 10, 111 – 127. Legates, D.R., Willmott, C.J., 1990b. Mean seasonal and spatial variability in global surface air temperature. Theor. Appl. Climatol. 41, 11 – 21. Liquete, C., Canals, M., Arnau, P., Urgeles, R., Durrieu de Madron, X., 2004. The impact of humans on strata formation along Mediterranean margins. Oceanography 17 (4), 42 – 51. Ludwig, W., 1997. Continental erosion and river transport of organic carbon to the world’s oceans. Sci. Geol., Mem. 98 (Strasbourg, 196 pp.). McManus, J., 2002. Deltaic responses to changes in river regimes. Mar. Chem. 79, 155 – 170.

C. Liquete et al. / Marine Geology 222–223 (2005) 471–495 MAPA (Ministerio de Agricultura, Pesca y Alimentacion), 1997. Sistema de Informacion Geografica Oleicola, URL: http://w3. mapya.es/dinatierra_v3. Milliman, J.D., Meade, R.H., 1983. World-wide delivery of river sediment to the oceans. J. Geol. 91, 1 – 21. Milliman, J.D., Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sedimentary discharge to the ocean: the importance of small mountainous rivers. J. Geol. 100, 525 – 544. MMA (Ministerio de Medio Ambiente), 2000. Libro Blanco del Agua en Espan˜a, Jacaryan S.A., Madrid, 637 pp. Moreira, J.M., Ojeda, J., 1992. Evaluacion de riesgos de erosion en Andalucia. Andalucia, Una Vision Inedita desde el Espacio, Publicaciones Electronicas de la Consejeria de Medio Ambiente de la Junta de Andalucia, Serie tecnica, pp. 265 – 270. Pavanelli, D., Pagliarani, A., 2002. Soil and water monitoring water flow, turbidity and suspended sediment load, from an Apennine catchment basin, Italy. Biosyst. Eng. 83 (4), 463 – 468. Perkins, H., Kinder, T.H., La Violette, P.E., 1990. The Atlantic inflow in the northwestern Alboran Sea. J. Phys. Oceanogr. 20, 242 – 263. Pike, J.G., 1964. The estimation of annual runoff from meteorological data in a tropical climate. J. Hydrol. 2, 116 – 123. Pinet, P., Souriau, M., 1988. Continental erosion and large scale relief. Tectonics 7, 563 – 582. Poulos, S.E., Collins, M.B., 2002. Fluviatile sediment fluxes to the Mediterranean Sea: a quantitative approach and the influence of dams. In: Jones, S., Frostick, L.E. (Eds.), Sediment Flux to Basins: Causes, Controls and Consequences, Spec. Publ.-Geol. Soc. Lond., vol. 191, pp. 227 – 245. Probst, J.L., 1992. Geochimie et hydrologie de l’erosion continental. Mecanismes, bilan global actuel et fluctuations au cours des 500 derniers millions d’annees. Sci. Geol., Mem. 94 (Strasbourg, 161 pp.). Puertos del Estado, 1983–2004. Banco de Datos, Oceanografia y Meteorologia, Parametros Oceanograficos, Oleaje. URL: http:// www.puertos.es. Romero Cordon, E., Alvarez Vigil, L., Garrido Perez, L., De Las Heras Martin, F., Duran Valsero, J.J., Madrid Garcia, R., 2003. The catastrophic flood event of the 18th of October, 1973, in the Rambla of Albun˜ol (La Rabita, Granada, Spain). In: Thorndycraft, V.R., Benito, G., Barriendos, M., Llasat, M.C. (Eds.), Palaeofloods, Historical Floods and Climatic Variability: Applications in Flood Risk Assessment, Proc. PHEPHRA Workshop, Barcelona, 16–19 Oct 2002, pp. 327 – 332.

495

Sanz de Galdeano, C., Vera, J.A., 1992. Stratigraphic record and paleogeographical context of Neogene basins in the Betic Cordillera, Spain. Basin Res. 4, 21 – 36. Sanz de Galdeano, C., Lopez Casado, C., Delgado, J., Peinado, M.A., 1995. Shallow seismicity and active faults in the Betic Cordillera. A preliminary approach to seismic sources associated with specific faults. Tectonophysics 248, 293 – 302. Sarhan, T., Lafuente, J.G., Vargas, M., Vargas, J.M., Plaza, F., 2000. Upwelling mechanisms in the northwestern Alboran Sea. J. Mar. Syst. 23 (4), 317 – 331. SGMA (Secretaria General de Medio Ambiente), 2001. Resolucion de 19 de septiembre de 2001 sobre la evaluacion de impacto ambiental del proyecto de restauracion hidrologico forestal en la cuenca del Guadalmedina de la Confederacion Hidrografica del Sur. BOE 248 (16-10-2001), 38101 – 38102. Syvitski, J.P.M., 2003. Supply and flux of sediment along hydrological pathways: research for the 21st century. Glob. Planet. Change 39, 1 – 11. Thornthwaite, C.W., 1948. An approach toward a rational classification of climate. Geogr. Rev. 38, 85 – 94. Turc, L., 1954. Le bilan d’eau des sols. Relation entre les precipitations, l’evaporation et l’ecoulement. Ann. Agron. 5, 491 – 569. UNEP (United Nations Environment Programme), 2002. Regionally based assessment of persistent toxic substances. Mediterranean Regional Report, 11 pp. Vargas-Yan˜ez, M., Plaza, F., Garcia-Lafuente, J., Sarhan, T., Vargas, J.M., Velez-Belchi, P., 2002. About the seasonal variability of the Alboran Sea circulation. J. Mar. Syst. 35, 229 – 248. Verstraeten, G., Poesen, J., de Vente, J., Koninckx, X., 2003. Sediment yield variability in Spain: a quantitative and semiquantitative analysis using reservoir sedimentation rates. Geomorphology 50, 327 – 348. Vorosmarty, C.J., Meybeck, M., Fekete, B., Sharma, K., Green, P., Syvitski, J., 2003. Anthropogenic sediment retention: major global impact from registered river impoundments. Glob. Planet. Change 39, 169 – 190. Walling, D.E., Fang, D., 2003. Recent trends in the suspended sediment loads of the world’s rivers. Glob. Planet. Change 39, 111 – 126. Willmott, C.J., Matsuura, K., 2001. Terrestrial air temperature and precipitation: Monthly and annual time series (1950–1999), Version 1.02. URL: http://climate.geog.udel.edu/~climate/ html_pages/download.html.