Hydromorphological analysis of the Guadalquivir fluvial system (Southern Spain). [Technical report]

Hydromorphological analysis of the Guadalquivir fluvial system (Southern Spain) M. Rinaldi, R. Baena Escudero, L. Nardi, B. García Martínez, I. Guerrero Amador Technical Report Departamento de Geografía Física y Análisis Geográfico Regional Universidad de Sevilla May 2018 FACULTAD DE GEOGRAFÍA E HISTORIA DEPARTAMENTO DE GEOGRAFÍA FÍSICA Y ANÁLISIS GEOGRÁFICO REGIONAL ! 1! Table of contents 1. Introduction 2 2. Materials and methods 3 3. Delineation and characterization of spatial units 4 3.1 Region 4 3.2 Catchment 5 3.3 Landscape units 7 3.4 Segments 9 3.5 Reaches 10 4. Assessment of hydromorphological conditions 18 4.1 Further reach characterization 18 4.2 Assessment methodology: MQI and HMQI 21 4.3 Results 24 4.4 Comparative analysis with the IHG 31 5. Discussion and conclusions 33 References 37 Acknowledgements This research project has been conducted during a sabbatical leaving of Massimo Rinaldi (University of Florence) in Seville from September 2017 to May 2018, financially supported by University of Seville VI PPIT-US (Ayudas para la movilidad internacional del personal dedicado a la investigación, May 2017) and University of Florence. The staff of the Departamento de Geografía Física y Análisis Geográfico Regional, Universidad de Sevilla, is acknowledged for scientific and logistical support. Alfredo Ollero, University of Zaragozza, is acknovledged for providing information about the Índice Hidrogeomorfológico (IHG). Fernando Magdaleno Mas (CEDEX, Madrid) and Emilio Esteban Rodriguez Merino (Confederacion Hidrografica del Guadalquivir, Seville) are also acknowledged for providing information about the implementation of the Water Framework Directive in Spain and in the Guadalquivir catchment. ! 2! Hydromorphological analysis of the Guadalquivir fluvial system (Southern Spain) Massimo Rinaldi1, Rafael Baena Escudero2, Laura Nardi1, Belen García Martínez2, Inmaculada Guerrero Amador2 1 Dipartimento di Scienze della Terra, Università di Firenze 2 Departamento de Geografía Física y Análisis Geográfico Regional, Universidad de Sevilla 1. Introduction Following the introduction of the Water Framework Directive (WFD), numerous methodologies have been proposed to assess and monitor the “hydromorphology” (i.e. hydrology and geomorphology) of fluvial water bodies, which vary widely in terms of their concepts, aims, spatial scales, collected data and therefore their applicability (e.g., Fernandez et al., 2011; Belletti et al., 2015). Over the last few years, an evident trend has emerged from site-scale physical habitat surveys towards more sound geomorphologically based approaches, methods and frameworks of geomorphic condition assessment, in the attempt to understand river functioning and evolution as a basis for interpreting current conditions (e.g., Brierley and Fryirs, 2005; Ollero et al., 2007, 2011; Rinaldi et al., 2013; Fryirs, 2015). However, the implementation of these approaches is still quite limited, whereas methods not based on physical processes remain the most widely applied to assess hydromorphology (Belletti et al., 2015). Therefore, a need still exists to promote a more comprehensive, process-based hydromorphological assessment that considers the character and dynamics of river reaches and how these are affected by present and past natural and humaninduced changes within the catchment as well as within the reach. In response to this need, new methodologies have been developed and/or extended within the REFORM (REstoring rivers FOR effective catchment Management) project (2011-2015), funded by the European Union's FP7 Programme. Specifically, two complementary approaches have been developed for hydromorphological assessment: (i) an open-ended approach - the REFORM hydromorphological framework (Gurnell et al., 2016), and (ii) a set of more specific hydromorphological assessment procedures which incorporates a set of clearly defined stages and steps - the REFORM hydromorphological assessment methods (Rinaldi et al., 2017). However, a need to further test these approaches exists, specifically in some areas of Europe with peculiar morpho-climatic and human conditions, in order to promote their use by public agencies in charge of monitoring and management of rivers. In particular, many fluvial systems of Southern Europe (including Spain and Italy) present some similar natural (e.g. the relatively high energy, bedload, and channel dynamics) and human characteristics (e.g. intensive sediment mining, dams) so that they notably differ from rivers of North Europe where most of hydromorphological methods have been developed and are currently used for the WFD implementation. The overall objective of this research project was to carry out an assessment of hydromorphological conditions of the medium – lower Guadalquivir River by testing the hydromorphological methodologies and tools developed in the context of the REFORM project to a new case study of Southern Spain. Furthermore, the study also aims at performing a comparative analysis of the REFORM hydromorphological assessment methods (MQI, Rinaldi et al., 2017) with hydromorphological methods more specifically developed in Spain (IHG, Ollero et al., 2007, 2011) by using the results on the Guadalquivir with the addition of some rivers of Italy. ! 3! 2. Materials and methods The analysis of hydromorphological conditions of the Guadalquivir fluvial system carried out in this research comprises two distinct phases: (i) delineation and characterization of spatial units (according to the REFORM hydromorphological framework); (ii) assessment of morphological conditions of the lower Guadalquivir river (according to the REFORM hydromorphological assessment methods). The various phases of delineation, characterization, and assessment have strongly made use of the existing knowledge on the Guadalquivir River concerning its hydromorphological changes, problems and causes conducted at the University of Seville (García Martínez et al., 1999; Baena Escudero, 2006; García Martínez & Baena Escudero, 2006, 2008; Guerrero Amador & Baena Escudero, 2010; García Martínez, 2016; Guerrero Amador, 2016; Baena Escudero et al., 2016). For example, the detailed delineation and mapping of ancient and recent terraces (García Martínez, 2016; Baena Escudero et al., 2016) facilitated the classification of confinement and was therefore useful for the segmentation and characterization phases. The delineation and characterization process allows to define boundaries between spatial units from larger to smaller scales (Gurnell et al., 2016). The larger spatial units are the region (>105 km2) and the catchment (102-105 km2) within which main landscape units (102-103 km2) are settled on the basis of catchment scale controls as geology and physiographic characteristics. Then, segments (101-102 km) are defined as section of the river controlled by similar valley-scale influences and river energy conditions. Segments are at first delineated by accounting for the intersection of the main river with the landscape units, as well as by considering a number of additional factors (e.g. major changes in valley gradient, catchment area, confluences of major tributaries, and lateral confinement). Within each segment one to several different reaches (10-1-101 km) can be identified along which boundary conditions are uniform displaying a near consistent internal set of process-form interactions. Reaches are thus delineated on the basis of: (i) valley setting, in terms of confinement; (ii) channel morphology (e.g. sinuous, meandering, etc.); (iii) presence of other discontinuities, such as assemblages of existing geomorphic units, and presence of river tributaries. The delineation and characterisation of spatial units were performed on the basis of the following material and methods: (1) Landscape units were defined by using GIS data (layers), i.e. maps on catchment geology and DEM. (2) Segments were delineated on GIS and by using available Orthophotos of 2010, supported by Google Earth images of the same period. (3) Reaches were delineated, only for the segment where the assessment of morphological conditions was then carried out, by using images Orthophotos of 2010, supported by Google Earth images, from which parameters characterizing river morphology (e.g. channel sinuosity) were calculated. Images were then integrated by field surveys to better characterise some aspects (i.e. sediment size and bank dynamics) and classify channel morphology by the Extended River Typology (Rinaldi et al., 2016a). Furthermore, daily discharge records from some gauging stations, covering a period of about 50 years of data series, were used to estimate flow properties and classify the Flow Regime Type (Bussettini et al., 2014; Rinaldi et al., 2016a). Following the delineation and characterization phase, an assessment of morphological conditions was carried out for the Medium - Lower Guadalquivir river. This was achieved by the application of the REFORM assessment (MQI, Rinaldi et al., 2013; 2017), followed by a comparative analysis with the hydromorphological method developed in Spain (IHG; Ollero et al., 2007, 2011). ! 4! 3. Delineation and characterization of spatial units 3.1 Region The region is a large geographic area that contains characteristic assemblages of natural ecological communities that are the product of broad influences of climate patterns. The Guadalquivir catchment is located within the Mediterranean region (Figure 3.1), with most of it classified as Mediterranean pluriseasonal oceanic (Mpo), and some small areas classified as Mediterranean pluriseasonal continental (Mpc) and Mediterranean xeric continental (Mxo). The fluvial regime within most of the catchment is classified as “Atlantic Pluvial substropical” (Figure 3.2). Figure 3.1 - Spanish biogeographic regions (from the Bioclimatic Map of Europe, www.globalbioclimatics.org). Figure 3.2 – Spanish types of fluvial regimes (from Tipos de Régimen Fluvial: Atlas Nacional de España, from Ministerio de Agricultura y Pesca, Alimentación y Medio Ambiente, 2017). ! 5! 3.2 Catchment The Guadalquivir river catchment (Figure 3.3) extends for an area of 57.527 km2, with a mean catchment elevation of 566 m (Sabater et al., 2009). Figure 3.3 – The Guadalquivir river catchment. The physiography of the catchment is characterised by the Guadalquivir depression, delimited by the Macizo Hesperico and by the Cordilleras Beticas (Figure 3.4). The Macizo Hesperico (max elevation of 1.323 m a.s.l.) is located in the upper part of the catchment and is mainly characterised by hilly areas. The Cordilleras Beticas comprise hilly and mountainous areas reaching elevations of over 3.000 m a.s.l. The Guadalquivir depression is characterised by gentle hills and the alluvial valley floor of the Guadalquivir and its main tributaries. Figure 3.4 – DEM of the Guadalquivir catchment. From a geological point of view, the following main sectors are usually distinguished (Figure 3.5): the Cordilleras Beticas, divided in Externas and Internas; the Macizo Hesperico; and the Neogenic and Quaternary deposits. The Cordilleras Beticas are predominantly characterised by carbonate sedimentary (marls, dolomites) and metamorphic (mica-schist, marble, gneiss) little erodible rocks; the Macizo Hesperico is predominantly characterised by siliceous, little erodible rocks, mainly including sandstones, quartzite, and granites; whereas the Neogenic and Quaternary ± 0 60 120 18030 Km CHG.Elevaciones1.tif Value 3399 0 ± 0 50 100 15025 Km Elevation m a.s.l ! 6! depression is composed of more erodible rocks, mainly marls, silt, and calcarenite, covered by detrital and alluvial deposits. Figure 3.5 – Geological sketch of the Guadalquivir catchment (from Mapa Geologico-Minero de Andalucia – Junta de Andalucia). (1) Cordilleras Beticas Externas; (2) Cordilleras Beticas Internas; (3) Macizo Hesperico; (4) Neogene and Quaternary. Most of the Guadalquivir catchment is characterised by a warm temperature (Mediterranean) climate with mild temperatures (annual average of 16.8°C) and a relative scarcity of rainfall (annual average of 630 mm) (Sabater et al., 2009). Concerning the land use, agriculture largely dominates in the hilly and plain areas, whereas forests and semi natural areas dominate on the highest hilly and mountain areas (Figure 3.6). Figure 3.6 – Land use of the Guadalquivir catchment. 1: Artificial surfaces; 2: Agricultural areas; 3: Forest and semi natural areas; 4: Wetlands; 5: Water bodies. ± 0 50 100 15025 Km schema_geologico Unidad Cordilleras Beticas Ext Cordilleras Beticas Int Macizo Hesperico Neogenos y Cuaternario 1 2 3 4 Livello_I 1 2 3 4 5 ± 0 50 100 15025 Km ! 7! 3.3 Landscape units Landscape units have been delineated by combining GIS data (layers) of geology and physiographic characteristics (elevation and slope), such that the dominant characteristics were broadly consistent internally for each unit. The following five landscape units have been identified (Figure 3.7). Figure 3.7 – Landscape units of the Guadalquivir river catchment. 1. Mountain areas of the Cordilleras Beticas; 2. Hilly areas of the Cordilleras Beticas; 3. Hilly areas of the Macizo Hesperico; 4. Hilly and plain areas of Neogene – Quaternary; 5. Alluvial and coastal deposits. 1. Mountain areas of the Cordilleras Beticas This landscape unit comprises the portions of the catchment with notable characteristics of mountainous areas, i.e. high relief (elevation) and steepness. These areas are predominantly (but not exclusively) higher than 800 m a.s.l., in accordance with the guidance from the Water Framework Directive (high: > 800 m; mid-altitude: 200-800 m; lowland: < 200 m), and a higher slope compared to hilly areas. In relation to the geological characteristics of the catchment, this unit includes the totality of the Cordilleras Beticas Externas, and other areas of the Cordilleras Beticas Internas, characterised by very resistant (predominantly siliceous) rocks. Areas higher than 800 m a.s.l. within the Macizo Hesperico and Neogene-Quaternary were not included in this unit but considered as hilly areas because of less steep slopes. 2. Hilly areas of the Cordilleras Beticas This landscape unit coincides with the Cordilleras Beticas Internas, excluding the portions at an elevation predominantly higher than 800 m a.s.l. with steep slopes (which have been comprised within the mountain areas). Therefore it comprises predominantly mid-altitude areas (200-800 m), characterised by relatively resistant (predominantly carbonates and metamorphic) rocks, and less steep slopes compared to the mountain areas. 3. Hilly areas of the Macizo Hesperico This landscape unit coincides with the Macizo Hesperico, predominantly composed by resistant rocks (quartzite, sandstones, and granitic rocks) but with lower elevation (predominantly 200-800 m) and less steep slopes compared to the mountain areas. ± 0 50 100 15025 Km Landscape_UnitsCorretto2Copy Unidad Unità Montana Cordilleras Beticas Externas Macizo Hesperico Neogenos y Cuaternario Unità di Pianura costiera e alluvionale 1 2 3 4 5 ! 8! 4. Hilly and plain areas of Neogene - Quaternary This landscape unit coincides with the Neogene – Quaternary deposits, excluding the alluvial and coastal deposits which are separately considered in the following landscape unit. It includes hilly areas characterised by relatively soft, erodible rocks and gentle slopes, and some intermontane plain. 5. Alluvial and coastal deposits The unit comprises the coastal plain, and the whole alluvial Quaternary deposits (including Pleistocenic terraces) of the Guadalquivir river and of some of the main tributaries, where these deposits are sufficiently wide. Other portions of alluvial plain of other tributaries of Guadalquivir river were not considered, but included in the previous unit, because of their generally very limited lateral extension or consists of pockets of floodplain. A B C D E F Figure 3.8 – Typical features of the landscape units. A: Mountain areas of the Cordilleras Beticas; B: Hilly areas of the Cordilleras Beticas; C: Hilly areas of the Macizo Hesperico; D: Hilly and plain areas of Neogene – Quaternary; E: Alluvial plain of the Guadalquivir; F: Coastal plain at the mouth of the Guadalquivir. ! 15! Figure 3.15 – Reach delineation of Segment 3. Reach Confinement Channel morphology Segmentation criterion (upstream boundary) 4_1 Partly Confined Sinuous Change in landscape unit (segment) 4_2 Partly Confined Meandering Change in morphology – Dam 4_3 Partly Confined Sinuous Change in morphology - Tributary 4_4 Unconfined Meandering Change in confinement and morphology 4_5 Partly Confined Meandering Change in confinement – Weir – Urbanisation 4_6 Unconfined Sinuous Change in confinement and morphology – Weir 4_7 Unconfined Meandering Change in morphology – Tributary 4_8 Unconfined Sinuous Change in morphology 4_9 Unconfined Meandering Change in morphology – Tributary 4_10 Unconfined Sinuous Change in morphology – Tributary Table 3.6 – Reach delineation of Segment 3. ! 16! Figure 3.16 – Reach delineation of Segment 5. Reach Confinement Channel morphology Segmentation criterion (upstream boundary) 5_1 Unconfined Meandering Change in segment – Change in Morphology – Dam - Tributary 5_2 Unconfined Meandering Tributary 5_3 Unconfined Sinuous Change in morphology - Tributary 5_4 Partly Confined Sinuous Change in confinement - Tributary 5_5 Unconfined Sinuous Change in confinement - Tributary 5_6 Unconfined Sinuous Dam 5_7 Unconfined Sinuous Change - Tributary Table 3.7 – Reach delineation of Segment 5. ! 17! Figure 3.17 – Reach delineation of Segment 6. Reach Confinement Channel morphology Segmentation criterion (upstream boundary) 6_1 Unconfined Sinuous Change in segment – Dam 6_2 Unconfined Straight Change in morphology - Tributary 6_3 Unconfined Sinuous Change in morphology 6_4 Unconfined Sinuous Tributary 6_5 Unconfined Sinuous Tributary 6_6 Unconfined Sinuous Tributary Table 3.8 – Reach delineation of Segment 6. ! 18! 4. Assessment of hydromorphological conditions The assessment of hydromorphological conditions was carried out by the application of Morphological Quality Index (MQI, Rinaldi et al., 2013; 2017) and the Hydro-Morphological Quality Index (HMQI, Rinaldi et al., 2016b) to the Medium - Lower Guadalquivir river (from Segment 3 to Segment 6). The MQI was originally developed for application in Italy, i.e. to cover the full range of physical conditions, morphological types, degree of artificial alterations and amount of channel adjustments occurring in this country. During the REFORM project, the method was extended and tested on a number of European streams (Belletti et al., 2018), which were under-represented or indeed totally unrepresented in the Italian context. The version used in this study (Rinaldi et al., 2016b) is a further extension of the REFORM version, which is characterized by a more emphasis on low-energy streams, the use of an Extended River Typology (ERT), and the inclusion of the Hydro-Morphological Quality Index (HMQI). 4.1 Further reach characterization Following the initial delineation of the river reaches, a further characterization has been carried out during the assessment phase by collecting additional information during the field visits (in particular, the dominant bed sediment) (Table 4.1). Based on bed sediment, an extended river typology comprising 22 river types has been applied (Figure 4.1, Rinaldi et al., 2016a). ! 19! Figure 4.1 – Extended River Typology (Rinaldi et al., 2016a). ! 20! R L (km) Cc Si W (m) S Se ERT 3_1 4149.40 U 1.72 35.41 0.0007 ÷ 0.0009 G-Co 14 3_2 1745.08 U 1.18 60.26 G-Co 13 3_3 4487.94 PC 1.33 31.47 G-Co 13 3_4 4528.44 U 2.02 26.80 G-Co 14 3_5 2561.79 U 1.09 36.15 G-Co 13 3_6 3016.05 PC 1.16 25.06 G-Co 13 3_7 4570.51 PC 1.80 24.08 G-Co 14 3_8 19378.80 U 1.94 27.04 G-Co 14 3_9 18497.37 U 1.56 53.64 G-Co 14 3_10 2763.46 U 1.10 56.46 G-Co 13 3_11 4151.93 C - 94.45 G-Co 7 3_12 6171.11 C - 40.53 G-Co 7 3_13 6173.34 U 1.19 42.78 G-Co 13 3_14 3891.08 C - 42.50 G-Co 7 3_15 5492.92 C - 54.53 G-Co 7 3_16 2717.77 U 1.05 54.67 G-Co 13 3_17 6637.32 U 1.30 55.58 G-Co 13 3_18 14780.06 C - 62.69 G-Co 7 3_19 12194.66 PC 1.07 100.72 G-Co 13 3_20 10719.84 PC 1.40 62.04 G-Co 13 4_1 8929.47 PC 1.28 95.66 0.0006 ÷ 0.0008 G 13 4_2 8288.60 PC 2.13 72.25 G 14 4_3 8966.81 PC 1.29 71.42 G 13 4_4 7071.33 U 1.96 90.76 G 14 4_5 10216.05 PC 1.50 114.03 G 14 4_6 5501.66 U 1.21 75.73 G 13 4_7 16340.32 U 1.64 48.79 G 14 4_8 10537.17 U 1.24 81.96 G 13 4_9 24861.78 U 1.64 56.95 G 14 4_10 15002.28 U 1.18 80.19 G 13 5_1 14743.16 U 1.60 71.47 0.0003 ÷ 0.0005 G 14 5_2 9834.33 U 2.48 70.04 G 14 5_3 27346.98 U 1.42 70.37 G 13 5_4 9904.74 PC 1.40 110.06 Sa 17 5_5 7917.30 U 1.34 182.60 Sa 17 5_6 5054.70 U 1.16 124.78 Sa 17 5_7 17476.02 U 1.45 181.83 Sa 17 6_1 12127.74 U 1.45 136.27 <0.0001 Sa 20 6_2 14529.79 U 1.02 159.73 Sa 20 6_3 18415.12 U 1.06 306.42 Si 20 6_4 13003.54 U 1.12 392.50 Si 20 6_5 28141.89 U 1.16 518.50 Si 20 6_6 19741.69 U 1.21 700.18 Si 20 Table 4.1 – Summary of reach characteristics of the Medium and Lower Guadalquivir River. R: reach; L: reach length (m); Cc: confinement class (C: C; PC: PC; U: U); Si: sinuosity index; W: channel width (m); S: mean bed slope; Se: bed sediment (dominant) (C=Clay, Si=Silt, Sa=Sand, G=Gravel, Co=Cobbles, B=Boulders); ERT: extended river typology. Based on the additional knowledge (sediment size, bed slope, etc.) and the further characterization obtained through the ERT, at this stage it is useful to discriminate Low-Energy streams (LE) from other medium-high energy streams. This distinction is useful to better define the ! 21! range of application of some indicators during the application of the MQI. In fact, some indicators are exclusive of low-energy streams, whereas other indicators are not applied to these category. The definition of low-energy streams is based on the following features: (i) Typical low-energy conditions are generally identified with unit stream power < 10 Wm-2 (Nanson & Croke, 1992). However, when unit stream power cannot be reliably estimated, bed slope is used as an alternative parameter. A bed slope ≤ 0.001 is a necessary (but not sufficient) condition to define low energy reaches. (iii) The physiographic context of the reach should be also considered: low-energy streams are typically associated with lowland or coastal plains, with low gradients and relatively low bedload, mainly composed of sand and finer sediment. (iv) Bed sediment is predominantly fine (sand, silt, clay), although (fine) gravel can occasionally occur. Bank material is also mostly cohesive. (v) Channel planform is typically single-thread (from straight to meandering) or multi-thread, anastomosing (i.e., low energy anabranching). In the case of the Guadalquivir River, all reaches from segment 3 to segment 6 have a bed slope <0.001, so this parameter by alone was not sufficient to discriminate between low and medium energy reaches. According to the previous criteria, low-energy reaches have been defined those reaches with predominantly fine bed sediment (silt and sand), further than having a bed slope <0.001, i.e. from reach 5_4 to 6_6. 4.2 Assessment methodology: MQI and HMQI The MQI is the key tool of the REFORM assessment methods. It derives from an original version developed for application in Italy, described in detail in Rinaldi et al. (2013). It can be classified as a ‘process-based’ method because it presents the following features (Rinaldi et al., 2017): (1) It explicitly considers processes which go beyond basic channel forms, i.e. it includes a series of indicators directly linked to the functioning of a series of basic processes expected in natural rivers (e.g., continuity in sediment and wood fluxes, bank erosion, lateral channel mobility); (2) The temporal component is explicitly accounted for, i.e. channel form is not limited to being considered in a static way, but its adjustments through time are addressed by a series of specific indicators; (3) Reference conditions are defined in terms of dynamic processes and functions that are expected to normally occur in a given physical context. This differs substantially from most current hydromorphological methods which define reference conditions in terms of a precise channel configuration or a set of channel characteristics. In fact, reference conditions for the MQI are river reaches in dynamic equilibrium, where the river is performing those morphological functions that are expected for a specific morphological typology and valley setting (i.e. confined vs. unconfined), and where artificial elements and interventions are absent or do not significantly affect the river dynamics at the catchment and reach scale. The MQI includes a set of twenty-eight indicators (Table 4.2), divided into the following three categories: (1) Geomorphological functionality evaluates whether or not the processes and related forms responsible for the correct functioning of the river are prevented or altered by artificial elements or by channel adjustments. (2) Artificiality assesses the presence and frequency of artificiality (artificial elements, pressures, interventions, management activities) independently of the effects of these artificial elements on channel forms and processes. (3) Channel adjustments assess relatively recent morphological changes (i.e. over about the last 100 years) that are indicative of a systematic instability related to human factors. ! ! 22! Indicators Assessed parameters Geomorphological functionality F1 – Longitudinal continuity in sediment and wood flux Presence of crossing structures altering sediment and wood continuity F2 – Presence of a modern floodplain Width and longitudinal length of a modern floodplain F3 – Hillslope–river corridor connectivity Presence and length of elements of disconnection on river sides F4 – Processes of bank retreat Presence/absence of retreating banks F5 – Presence of a potentially erodible corridor Width and longitudinal length of an erodible corridor F6 – Bed configuration–valley slope Identification of bed configuration and comparison with expected bed configuration based on valley slope F7 – Planform pattern Percentage of the reach length with alteration of planform pattern F8 – Presence of typical fluvial forms in the floodplain Presence/absence of fluvial forms in the alluvial plain F9 – Variability of the cross section Percentage of the reach length with alteration of the natural heterogeneity of cross section F10 – Structure of the channel bed Presence/absence of alterations of bed sediment F11 – Presence of in-channel large wood Presence/absence of large wood F12 – Width of functional vegetation Mean width of functional vegetation in the fluvial corridor F13 – Linear extension of functional vegetation Longitudinal length of functional vegetation along the banks Artificiality A1M –Upstream alteration of flows with potentially relevant effects on channel morphology Amount of changes of flows with potentially relevant effects on channel morphology (channel-forming discharges or higher) caused by interventions upstream A1H –Upstream alteration of flows without potentially relevant effects on channel morphology Amount of changes in low-flows (with no potentially relevant effects on channel morphology) caused by interventions upstream A2 – Upstream alteration of sediment discharges Presence, type, and position (drainage area) of relevant structures responsible for bedload interception (dams, check-dams, weirs) A3 – Alteration of flows in the reach Amount of changes in discharge within the reach A4 – Alteration of sediment discharge in the reach Type and density of structures intercepting bedload along the reach A5 – Crossing structures Spatial density of crossing structures A6 – Bank protections Length of protected banks A7 – Artificial levees Length and distance from the channel of artificial levees A8 – Artificial changes of river course Percentage of the reach length with artificial modifications of the river course A9 – Other bed stabilization structures Presence, spatial density and typology of other bed-stabilizing structures and revetments A10 – Sediment removal Existence of past and/or recent sediment mining activity A11 – Wood removal Existence and relative intensity of in-channel wood removal A12 – Vegetation management Existence and relative intensity of riparian vegetation cuts or aquatic vegetation removal Channel adjustments CA1 – Adjustments in channel pattern Changes in channel pattern from 1930s-1960s based on changes in sinuosity, braiding, and anastomosing indices CA2 – Adjustments in channel width Changes in channel width from 1930s-1960s CA3 – Bed-level adjustments Bed-level changes over the last 100 years Table 4.2 – Morphological Quality Index (MQI) and Hydro-Morphological Quality Index (HMQI): synthesis of indicators and assessed parameters (modified from Rinaldi et al., 2017). The overall assessment is carried out by making a synergic use of two types of methods: GIS analysis (using available databases and remotely sensed data such as aerial photos and satellite images) and field survey. The spatial scale of application is the river reach, as identified during the initial phase of segmentation. However, alterations of flow and sediment discharge require information at the segment and at the catchment scale on the types of interventions affecting these variables (i.e., dams, check dams, weirs, etc.). GIS analysis is carried out at the reach scale, while the field survey is focussed on representative sub-reaches (or ‘sites’). Field observations are used to verify and integrate those aspects which cannot be determined by remote sensing, further to evaluating those indicators which can be exclusively assessed in the field. The selection of the field sites for the Guadalquivir River was mainly driven by logistic constraints (the accessibility to the river was in general difficult) and was focussed on all crossing and transversal structures (bridges, dams, check dams). The evaluation is based on a scoring system. Scores and classes were defined and subsequently improved based on the results of a testing phase (Rinaldi et al., 2013). Three classes are generally ! 23! defined for each indicator (except for a limited number with two classes or more than three classes): (A) undisturbed conditions or negligible alterations; (B) intermediate alterations; (C) very altered conditions. For each indicator, we started by defining reference conditions for that indicator, corresponding to the absence or negligible presence of alterations (class A), a value of 0 being assigned to this class. A total score is computed as the sum of scores across all components and aspects. The Morphological Quality Index is then defined as MQI = 1 – Stot/Smax (1) where Stot is the sum of the scores, and Smax is the maximum score that could be reached when all appropriate indicators are in class C. The index is therefore directly proportional to the quality of the reach and inversely to the alterations, varying from 0 (minimum quality) to 1 (maximum quality), allowing the investigation of the full range of morphological conditions. According to this structure, reference conditions (i.e., class A for each indicator, corresponding to MQI = 1) are identified with the following: (i) the full functionality of geomorphic processes along the reach; (ii) the absence or negligible presence of artificial elements along the reach and to some extent (in terms of flow and sediment fluxes) in the catchment; and (iii) the absence of significant channel adjustments (configuration, width, bed elevation) over a temporal frame of about 100 years. The following classes of morphological quality were defined: (i) high, 0.85≤MQI≤1; (ii) good, 0.7≤MQI<0.85; (iii) moderate, 0.5≤MQI<0.7; (iv) poor, 0.3≤MQI<0.5; (v) bad, 0≤MQI<0.3. During the assessment and the compilation of the evaluation forms, some indicators may be affected by a lack of data or information or may require an interpretation that involves a certain degree of subjectivity. To help in indicating how certain the user feels concerning the answer, a degree of confidence (low, medium, high) and a second (alternative) choice in the classes can be expressed. This is calculated by taking the scores for the second choice (with low or medium confidence in the answer), and obtaining a range of variability rather than a single final value of the MQI. In the last version of the methodology used in this study, an original tool, the Hydro- Morphological Quality Index (HMQI), has been developed and applied to the Guadalquivir River. The Hydro-Morphological Quality Index (HMQI) is an extension of the MQI designed to assess the overall hydromorphological conditions of a stream reach as envisaged by the WFD. It includes all the same indicators of the MQI, with the addition of a sub-indicator (A1H) concerning the upstream alteration of flows without potentially relevant effects on channel morphology. The additional score of the A1H indicator (depending on the class, from A to C) is then added to the sum of the MQI scores (Stot), providing the Hydro-Morphological Quality Index (HMQI) defined in the same way of the MQI, i.e.: HMQI = 1 – Stot/Smax (2) where Smax remains the same of the MQI (i.e., the A1H score is added only to the numerator of the equation). Similarly to the MQI, the following classes of the overall hydro-morphological quality are defined: (i) high, 0.85≤HMQI≤1; (ii) good, 0.7≤HMQI<0.85; (iii) moderate, 0.5≤HMQI<0.7; (iv) poor, 0.3≤HMQI<0.5; (v) bad, 0≤HMQI<0.3. ! 24! 4.3 Results A summary of the MQI and HMQI results for the Medium – Lower Guadalquivir (from segment 3 to 6) is reported in Table 4.3, including the minimum and maximum values of the two indices based on answers with no high confidence. R MQI MQImin MQImax MQI class HMQI HMQImin HMQImax HMQI class 3_1 0.70 0.67 0.73 Good 0.63 0.59 0.65 Moderate 3_2 0.57 0.54 0.60 Moderate 0.49 0.46 0.52 Poor 3_3 0.56 0.53 0.58 Moderate 0.48 0.45 0.51 Poor 3_4 0.68 0.65 0.70 Moderate 0.60 0.57 0.62 Moderate 3_5 0.63 0.58 0.68 Moderate 0.55 0.51 0.60 Moderate 3_6 0.61 0.57 0.68 Moderate 0.54 0.49 0.60 Moderate 3_7 0.70 0.63 0.73 Good 0.63 0.56 0.65 Moderate 3_8 0.61 0.54 0.64 Moderate 0.54 0.46 0.56 Moderate 3_9 0.56 0.53 0.63 Moderate 0.49 0.45 0.55 Poor 3_10 0.73 0.68 0.75 Good 0.65 0.61 0.67 Moderate 3_11 0.54 0.47 0.56 Moderate 0.46 0.40 0.49 Poor 3_12 0.56 0.54 0.63 Moderate 0.49 0.46 0.56 Poor 3_13 0.66 0.64 0.70 Moderate 0.58 0.56 0.62 Moderate 3_14 0.60 0.57 0.69 Moderate 0.53 0.50 0.62 Moderate 3_15 0.68 0.64 0.76 Moderate 0.61 0.57 0.69 Moderate 3_16 0.69 0.65 0.71 Moderate 0.61 0.58 0.63 Moderate 3_17 0.59 0.57 0.63 Moderate 0.51 0.49 0.55 Moderate 3_18 0.54 0.50 0.56 Moderate 0.46 0.43 0.49 Poor 3_19 0.56 0.51 0.57 Moderate 0.49 0.43 0.49 Poor 3_20 0.56 0.48 0.56 Moderate 0.48 0.40 0.49 Poor 4_1 0.55 0.51 0.57 Moderate 0.47 0.44 0.49 Poor 4_2 0.59 0.56 0.63 Moderate 0.51 0.48 0.55 Moderate 4_3 0.58 0.53 0.63 Moderate 0.50 0.45 0.55 Moderate 4_4 0.58 0.52 0.61 Moderate 0.51 0.44 0.53 Moderate 4_5 0.43 0.42 0.48 Poor 0.35 0.34 0.40 Poor 4_6 0.61 0.57 0.65 Moderate 0.53 0.49 0.58 Moderate 4_7 0.68 0.65 0.73 Moderate 0.60 0.58 0.65 Moderate 4_8 0.57 0.54 0.60 Moderate 0.49 0.46 0.52 Poor 4_9 0.58 0.56 0.64 Moderate 0.51 0.49 0.56 Moderate 4_10 0.55 0.53 0.61 Moderate 0.47 0.45 0.53 Poor 5_1 0.60 0.58 0.61 Moderate 0.52 0.50 0.53 Moderate 5_2 0.69 0.65 0.72 Moderate 0.61 0.57 0.64 Moderate 5_3 0.48 0.46 0.51 Poor 0.40 0.39 0.44 Poor 5_4 0.59 0.54 0.61 Moderate 0.51 0.46 0.53 Moderate 5_5 0.55 0.51 0.58 Moderate 0.47 0.44 0.51 Poor 5_6 0.48 0.48 0.51 Poor 0.40 0.40 0.44 Poor 5_7 0.57 0.53 0.58 Moderate 0.49 0.45 0.50 Poor 6_1 0.55 0.55 0.55 Moderate 0.47 0.47 0.47 Poor 6_2 0.50 0.47 0.50 Moderate 0.42 0.40 0.42 Poor 6_3 0.51 0.51 0.51 Moderate 0.43 0.43 0.44 Poor 6_4 0.51 0.51 0.51 Moderate 0.43 0.43 0.44 Poor 6_5 0.49 0.49 0.50 Poor 0.42 0.42 0.42 Poor 6_6 0.59 0.59 0.60 Moderate 0.51 0.51 0.52 Moderate Table 4.3 – Summary of MQI and HMQI results for the Medium and Lower Guadalquivir River. R: reach; MQI: Morphological Quality Index; MQImin and MQImax: minimum and maximum values of MQI respectively based on answers with no high confidence; HMQI: Hydro-Morphological Quality Index; HMQImin and HMQImax: minimum and maximum values of HMQI respectively based on answers with no high confidence. ! 31! 4.4 Comparative analysis with the Índice Hidrogeomorfológico (IHG) Besides the analysis carried out by the MQI, some applications of the Index for Hydro- Geomorphological quality assessment of fluvial systems (IHG, Ollero et al., 2007, 2011) was carried out to verify the coherence and make an inter-comparison of the results obtained by the two methods. The IHG has been developed for the WFD in Spain, and used in a wide range of assessments of the status of Spanish rivers, and in a number or restoration initiatives. It is comparable with the MQI being considered a process-based, morphological method (Rinaldi et al., 2015) with some common features. A summary of the indicators of the IHG is reported in Table 4.5, including some correspondences with the MQI indicators. Indicators Sub-indicators Corresponding MQI indicators System functional quality 1. Discharge regime naturalness A. Presence and use of interventions of alteration of flow regime upstream A1, A3 2. Sediments availability and movement A. Presence and use of interventions of alteration of sediment regime upstream A2 B. Sediment removal / mining A10 C. Symptoms of low sediment mobility (armouring, clogging, growth of some vegetational species) F10 D. Disconnection/interception of sediment supply (along tributaries or hillslopes) F3, A2 3. Floodplain functionality A. Presence of longitudinal structures reducing lateral functioning A7 B. Presence of obstacles (roads, infrastructures) generally transversal - C. Land use in the floodplain (impermeable areas, high surfaces because of dredging or channelization) - Channel quality 4. Planform and channel morphology naturalness A. Artificial changes of channel course A8 B. Retrospective and progressive planform changes CA1, CA2 5. Continuity and naturalness of river bed and of longitudinal and vertical processes A. Reservoirs by dams or weirs A4 B. Minor crossing structures (bridges, fords, sills) A5, A9 C. Degraded bed topography (including incision) CA3 6. Riverbanks naturalness and lateral mobility A. Channelization, bank protections, including infrastructures and buildings on the bank A6 B. Other not natural elements or interventions modifying the natural bank morphology (F9) C. Limited lateral dynamics or instability F4 River corridor quality 7. Longitudinal continuity A. Longitudinal continuity of river corridor F13 8. Width A. Width of river corridor F12 B. Interaction with longitudinal continuity of river corridor - 9. Structure, naturalness and transversal connectivity A. Pressures on the river corridor (vegetation management, pasture, recreational use, aquifer exploitation, etc.) A12 B. Alteration of vegetation caused by invasive species - C. Infrastructures altering the transversal connectivity of the fluvial corridor - D. Interaction with longitudinal continuity of river corridor - Table 4.5 – Indicators of the Index for Hydro-Geomorphological quality assessment of fluvial systems (IHG) (from Ollero et al., 2007, 2008, 2011), and corresponding indicators of the MQI. Each parameter has an initial score of 10, corresponding to the natural state and functionality of the system. After the impacts and pressures are assessed, points are deducted from this initial value according to different criteria. The full IHG hydrogeomorphological assessment of each river reach is performed by adding the nine values obtained, with a highest possible score of 90 points. In this study, the IHG has been applied to all reaches of segments 5 and 6, given that the information available for these reaches was more detailed and suitable for the application of this ! 32! index. Furthermore, given that the points related to the Guadalquivir reaches showed a relatively limited range of values, the application of the IHG was extended to ten additional reaches of Italian rivers where the MQI was already available, in order to cover a wider range of values and therefore to make a more effective comparison between the two methods. River Reach MQI MQI Class HMQI HMQI Class IHG IHG Class Guadalquivir 5_1 0.60 Moderate 0.52 Moderate 55 Moderate Guadalquivir 5_2 0.69 Moderate 0.61 Moderate 55 Moderate Guadalquivir 5_3 0.48 Poor 0.40 Poor 42 Moderate Guadalquivir 5_4 0.59 Moderate 0.51 Moderate 46 Moderate Guadalquivir 5_5 0.55 Moderate 0.47 Poor 42 Moderate Guadalquivir 5_6 0.48 Poor 0.40 Poor 42 Moderate Guadalquivir 5_7 0.57 Moderate 0.49 Poor 39 Poor Guadalquivir 6_1 0.55 Moderate 0.47 Poor 48 Moderate Guadalquivir 6_2 0.50 Moderate 0.42 Poor 37 Poor Guadalquivir 6_3 0.51 Moderate 0.43 Poor 35 Poor Guadalquivir 6_4 0.51 Moderate 0.43 Poor 36 Poor Guadalquivir 6_5 0.49 Poor 0.42 Poor 32 Poor Guadalquivir 6_6 0.59 Moderate 0.51 Moderate 41 Poor Sentino 4_1 0.92 High 0.92 High 87 High Tagliamento 3_2 0.85 High 0.85 High 82 High Magra 3_2 0.76 Good 0.68 Moderate 68 Good Magra 5_2 0.63 Moderate 0.55 Moderate 53 Moderate Arno 7_1 0.38 Poor 0.30 Poor 33 Poor Arno 7_2 0.11 Bad 0.03 Bad 10 Bad Serchio 5_1 0.45 Poor 0.37 Poor 38 Poor Ombrone 2_2 0.46 Poor 0.38 Poor 33 Poor Cecina 2_4 1.00 High 1.00 High 90 High Cecina 3_7 0.79 Good 0.79 Good 76 High Table 4.6 – Results of the comparative analysis between IHG, MQI and HMQI. A B Figure 4.6 – Results of comparison between IHG, MQI, and HMQI for the Guadalquivir River (segments 5 and 6) and for a series of reaches of Italian Rivers. A) IHG versus MQI; B) IHG versus HMQI. The analysis of the results (Table 4.6) shows a statistically robust correlation either between the IHG and MQI, and the IHG and HMQI (Figure 4.6), despite the use of some different indicators. In terms of classes, 16 on 23 reaches result to fall in the same class of IHG and HMQI, whereas the correspondence between classes is slightly lower (14 on 23) between IHG and MQI. The regression between IHG and HMQI is also slightly higher than between IHG and MQI. This is explained by the fact that the IHG include the indicator ‘Discharge regime naturalness’, which account for the alterations of the overall hydrological regime, similarly to the HMQI, whereas the MQI only accounts for the alteration of flows having potential morphological effects. y"="100,77x"-"10,275" R²"="0,9311" 0" 10" 20" 30" 40" 50" 60" 70" 80" 90" 0" 0,1" 0,2" 0,3" 0,4" 0,5" 0,6" 0,7" 0,8" 0,9" 1" IHG$ MQI$ Guadalquivir"R." Italian"rivers" y"="91,535x"+"1,4555" R²"="0,93949" 0" 10" 20" 30" 40" 50" 60" 70" 80" 90" 0" 0,1" 0,2" 0,3" 0,4" 0,5" 0,6" 0,7" 0,8" 0,9" 1" IHG$ HMQI$ Guadalquivir"R." Italian"rivers" ! 33! 5. Discussion and conclusions The overall results of the MQI show that the majority of the Medium – Lower Guadalquivir, is in “moderate” quality conditions, with some reaches in “poor” conditions, but with several reaches close to the boundary with “poor” conditions. “Good” conditions are quite limited to only some reaches of segment 3. The situation is worst when the HMQI is used to assess the overall hydromorphological conditions, i.e. including the alterations of low flows without potential effects on channel morphology. Additionally to the MQI and HMQI, another index with similar approach and structure but more widely used for Spanish rivers, the IHG, has been applied to segments 5 and 6, in order to investigate on the coherence of results of the two different methods. Based on the IHG results, the reaches of segments 5 and 6 also range from poor and moderate conditions, coherently with the MQI and, still more, with the HMQI. Although the IHG has not been applied to segments 3 and 4, based on the strong correlations with HMQI and MQI, it can be argued that the reaches of segments 3 and 4 would also result in the same range of classes, i.e. poor and moderate. Given its structure, the MQI attempts to make an overall and comprehensive evaluation of river conditions, therefore promoting an understanding of pressure – response (i.e. cause – effect) conditions that can support the identification of possible management actions. In fact, given that the same type of pressure may result in different responses for different rivers, the artificiality indicators identify the potential elements of alteration, whereas the functionality and channel adjustment indicators assess the geomorphic responses (effects) to these disturbances, including past off-site impacts and adjustments. This synergic use of the different components of the assessment and their mutual feedbacks promotes a sound understanding of the river conditions and causes of alteration, which can be eventually used to identify management actions. This type of pressure – response understanding applied to the Guadalquivir River, and some general implications in terms of management and improvement of geomorphic conditions, are discussed below. The impressive number of dams, either along the river course and also in the upstream catchment, definitively represents the main cause of alteration of the geomorphic functioning of the river. A consequence of the extensive presence of dams is a substantial alteration of the flow regime. In terms of geomorphological functioning, the most important alterations are those which affect the range of channel-forming discharges (or higher). Although most of the dams are used for hydroelectric power, hydropeaking does not represent a significant problem because oscillations of water level are not carried out. However, the main consequence of the dams in terms of channel dynamics is certainly the dramatic reduction, or total annulment, of the bedload, i.e. the transport of the coarser fractions (cobble, gravel, and sand) of sediments. This alteration of bedload may have drastic consequences on channel morphology and dynamics. First of all, the very low bedload causes a drastic simplification of channel morphology, with the disappearance of instream geomorphic units within the bankfull channel such as depositional bars and islands. This in turns may cause a homogenization of the cross section geometry and a reduced channel dynamics, given that bank erosion processes and wood recruitment occur much less then expected for a sinuousmeandering river (even if of low-energy). The absence of instream geomorphic units, the lack of variability of channel width, as well as the insufficient levels of bank erosion have important implications in terms of ecological conditions, as these processes determine the availability and variability of physical habitats. The deficit of relatively coarse bedload is also a main factor determining adjustments of channel width and bed elevation. Whereas changes in channel pattern and width have been quantitatively evaluated by a comparison of aerial photos of 1956 and 2010, a systematic investigation on bedlevel changes was not possible because longitudinal bed profiles were lacking. However, bed incision in some specific sites along the segment 5 has been documented by previous studies (e.g., García Martínez, 2016; Guerrero Amador, 2016), and field evidences were used to qualitatively evaluate the class of incision (CA3), such as differences in elevation between the historic floodplain ! 34! (now a low terrace) and the portions of new floodplain developed after incision. These types of evidences, together with the existing information, strongly suggest that bed incision is widespread along most of the investigated portion of the Guadalquivir River. Dams can be considered as one of the main cause of bed incision, together with other possible causes (reforestation, straightening along the downstream segments, reduction of bedload from the catchment related to several dams along tributaries) that presumably determined a first phase of bed-level lowering already before the construction of dams along the investigated reaches. Whereas bed incision is likely to occur downstream, aggradation generally occur upstream the dam (along the reservoir area), but we interpreted that this aggradation occurred after a previous phase of more intense degradation so that the net result is still bed incision, and also that this secondary aggradation did not always affect the entire reach upstream of the dam. A consequence of incision is a reduction or disappearance of depositional bars because of the deficit of sediment and consequent erosion of existing depositional units. However, the most relevant consequence, in alluvial, unconfined or partly confined reaches, is usually the disconnection of the river with the surrounding floodplain. The historical (pre-incision) floodplain is abandoned and becomes a terrace, whereas a new, usually narrow, modern floodplain can be created at a lower level, as a result of lateral dynamics (i.e., bank erosion and creation of a new floodplain on the opposite side) or vertical sedimentation and vegetational encroachment of previous depositional bars. In the case of the medium – lower Guadalquivir, given the scarce lateral dynamics because of quite low-energy conditions and limited bedload, the main mechanism of new floodplain formation appears to be the deposition of large amounts of fine sediments on previous coarser depositional bars, and the subsequent colonization by vegetation and stabilization of these narrow surfaces. As previously observed, this type of modern floodplain frequently occurs along the Guadalquivir, but in most cases is quite narrow and often discontinuous (i.e., F2 in class B or C), where as the abandoned, historical floodplain is now intensively used by agriculture. It is still periodically inundated, but with remarkably lower frequency (higher return period), determining a substantial disconnection with the most frequent floods (i.e., floods with 1.5 or 2 years of return period, or channel-forming discharges). As already mentioned previously, a second major pressure on the Guadalquivir system is represented by the changes in land use, either in the catchment and in the alluvial plain. In the catchment, a considerable progressive increase of agricultural use and the subsequent deforestation (in particular the widespread olive cultivation in the medium – upper Guadalquivir catchment) has determined a remarkable increase in soil erosion, determining a supply of huge quantities of fine sediments to the river network, similarly to what has occurred in other river systems of the same region (e.g., González del Tánago et al., 2015). Concerning the alluvial plain, the intensive agricultural activity during the last decades has determined an occupation of the river corridor and a management of the naturally functioning riparian vegetation by its partial cutting in the areas external to the banks. An additional major alteration in the Guadalquivir concerns its bed substrate. The structure and heterogeneity of the channel bed sediment have several implications for the functionality of bedload processes and flow resistance, and are extremely important for the characteristics of aquatic physical habitats. The bed of the Guadalquivir is extensively affected by clogging or burial, where clogging refers to an excess of fine sediments, whereas burial or siltation is a special type of aggradation, where finer sediments (e.g. silt and sand) are deposited in a sufficiently thick layer to bury a coarser (e.g. gravel) river bed. Burial has not only direct ecological effects, but it also has morphological effects, since it buries bed forms and so simplifies bed morphological complexity. Similarly to clogging, burial is generally associated with an excessive input of fine sediments to the river channel caused by extensive bank erosion or soil erosion related to agricultural activity, land use changes (e.g. deforestation), trapping of fine sediments upstream dams, or even downstream by periodic releases of fine sediments during floods. ! 35! Some additional pressures concern the final part (segment 6) of the Guadalquivir. Along the urban reach of Seville, and also in the downstream reaches, historical straightening has determined a partially artificial planform pattern, with a substantial decrease in sinuosity and, for the reaches from 6_3 to 6_5, a reduction of the anabranching index (in these reaches the river was historically anastomosed with three distinct branches). These historical changes have obviously determined a notable simplification of the planform pattern and a substantial homogeneity of the cross-sections, and possibly have triggered some bed incision because of the increase of bed slope. Furthermore, the periodic dredging activity that occur along all the reaches downstream Seville to maintain navigability conditions of the river determines some additional remarkable alterations, such as to the bed substrate that is fatherly disturbed and degraded by this activity. It is also evident the widespread occurrence of mass failures related to the cohesive nature of the banks combined with bed incision related to upstream sediment deficit, and the shoreline erosion due to boat wake and waves related to the navigational activity along these reaches. Despite the remarkable number of hydromorphological pressures, problems and alterations, it is to be recalled that the prevailing MQI quality class is “moderate”, although some reaches in “poor” conditions. In fact, the following relatively positive aspects can be noted: (1) Apart the extensive presence of dams, the artificiality of the river is not excessive. For example, bank protections and artificial levées are occasionally localized in correspondence of some urban area or infrastructure close to the river. As a result, the river is generally not constrained in planform, as it is relatively free to move and modify its morphology. (2) As a consequence of the previous point, the potentially erodible corridor is generally continuous and sufficiently wide, differently from other areas (e.g., in most of the alluvial plains in Italy) where the valley floor is densely populated. (3) Riparian vegetation is generally present, although the width of the functional vegetation is not very wide because of the intensive agriculture close to the river margins. We recall that the only the geomorphic functioning of the vegetation is considered by the MQI, because species identification is not required and so alterations of vegetation caused by invasive species are not considered. Conversely, the presence of invasive species along all the investigated reaches has been revealed by the application of the IHG. The MQI structure provides a rational framework that is useful for identifying and prioritizing management strategies and restoration actions see Rinaldi et al., 2017). A first obvious prioritization rule consists of preserving current conditions for those indicators which are in class A and considering some possible actions for improving those indicators lying in classes B and C. In the case of the medium – lower Guadalquivir, as discussed previously, the main hydromorphological problem is represented by the strong longitudinal discontinuity of sediment transport caused by the several dams and weirs for hydroelectric use. These structures cause an interruption of the flux of relatively coarser bed sediment (cobbles, gravel, and coarse sand) that moves as bedload and of wood, whereas the finer sediment fractions (fine sand, silt, and clay) are transported as suspended load and are capable to move downstream, although part of this fine material is temporarily stored in the reservoirs and accumulated along the river channel as new floodplain deposits or on the banks. A solution of this problem is obviously not easy, given the high number of dams and weirs along the investigated reaches and in the upstream catchment. Measures allowing for the flux of part of the bedload downstream of dams (by-pass) may be explored to mitigate this problem. However, the evaluation of the negative effects of the dams, as well as of the navigability along the final reaches, should be discussed in a broader context, analysing other aspects such as sensitivity, morphological potential, and the integration with other goals and constraints. Further than longitudinal disconnection of bedload, the dams also cause alteration of flows. As previously discussed, the most important alterations for geomorphic processes are those which affect the range of channel-forming discharges (or higher). In this respect, avoiding substantial regulation of channel-forming discharges without interfering with the hydroelectric use of the dams might be a feasible management action. Furthermore, the periodic release of natural ! 36! channel-forming flows, known as “flushing flows” (Kondolf & Wilcock, 1996) or, more recently, as “experimental floods” (Magdaleno, 2017) have already been shown to have relevant effects on the ecogeomorphic conditions of rivers altered by flow regulation and should be considered as a possible management action. Concerning other main geomorphic problems of the Guadalquivir, such as bed substrate burial and bed incision, they are at least in part related to the presence of dams and obviously would be difficult to mitigate without some actions on the dams. However bed substrate burial is basically related to the large supply of fine sediments from the catchment. Preserving current conditions for those aspects which are in relatively good conditions is another more feasible management strategy for the Guadalquivir River. First of all, adequate policy actions should be followed to avoid construction of bank protections, levées or other artificial structures, except in the cases that they are strongly necessary. As previously noted, naturally functioning vegetation in the river corridor directly connected with the channel is generally observed along the Guadalquivir, except that for some portions of reaches very urbanised or strongly managed, although the presence of invasive species represents a further problem. Concerning this aspect, preservation of a sufficient width of riparian vegetation is an obvious management priority, but the increase of the width of vegetation might be a feasible goal. This might be obtained by avoiding excessive pressure by agriculture activity up to the proximity of the river margins and impeding cutting of riparian forests to gain new terrains for farming activities. Avoiding agricultural practices on the margins of the river would have also beneficial effects in terms of reducing inputs of fine sediments by excessive soil erosion, and therefore promoting instream sediment burial. Whereas the previous are conservative measures, an improvement measure aimed at further improving the conditions in the river corridor and in the channel would be to promote the concept of “erodible corridor” (Piégay et al., 2005) or “functional mobility corridor” (Rinaldi et al., 2009), i.e. delineating a corridor within the alluvial plain where to allow the occurrence of natural river processes (channel mobility, riparian vegetation growth, flooding, etc.). This would have many beneficial effects on many aspects (indicators) such as the heterogeneity of channel pattern and of cross sections, the lateral extension of a functioning vegetation corridor, the recruitment of wood within the channel and therefore the recreation of instream geomorphic units, etc. Obviously, as noted previously for the dams, this very challenging management option should be considered in a broader context, analysing other aspects such as sensitivity, morphological potential, and the integration with other human goals and constraints. ! 37! References Baena Escudero, R. (2006). Los ríos como oportunidad para la recuperación del patrimonio natural de las ciudades desde la cooperación institucional. In: Baena Escudero R. et al. (Eds), Ríos y ciudades europeas: espacios naturales, culturales y productivos, Universidad de Sevilla, 37-43. Baena Escudero, R., Guerrero Amador, I., García Martinez, B., Posada Simeon, J.C. (2016). Ocupación antrópica y riesgos de inundación en la llanura del Río Guadalquivir (sector Palma del Rio – Sevilla, España). Boletín de la Asociación de Geógrafos Españoles, 72, 149-167. Belletti, B., Rinaldi, M., Gurnell, A.M., Buijse, A.D., Mosselman, E., (2015). A review of assessment methods for river hydromorphology. Environmental Earth Sciences 73, 2079-2100. Belletti, B., Nardi, L., Rinaldi, M., Poppe, M., Brabec, K., Bussettini, M., Comiti, F., Gielczewski, M., Golfieri, B., Hellsten, S., Kail, J., Marchese, E., Marcinkowski, P., Okruszko, T., Paillex, A., Schirmer, M., Stelmaszczyk, M., Surian, N. (2018). Assessing restoration effects on river hydromorphology using the process-based Morphological Quality Index in eight European river reaches. Environmental Management, 61, 69-84. Brice, J.C. (1964). Channel patterns and terraces of the Loup Rivers in Nebraska. US Geological Survey Professional Papers, 422D, 1-41. Brierley, G.J., Fryirs, K.A. (2005). Geomorphology and River Management: Applications of the River Style Framework. Blackwell, Oxford, UK, 398 pp. Bussettini, M., Percopo, C., Lastoria, B., Mariani, S. (2014). A method for characterizing the stream flow regime in Europe. In: Lollino, G., Arattano, M., Rinaldi, M., Giustolisi, O., Marechal, J.C., Grant, G.E. (Eds), Engineering Geology for Society and Territory, Volume 3, Proceedings IAEG XII Congress, Springer International Publishing Switzerland. Fernández, D., Barquin, J., Raven, P.J. (2011). A review of river habitat characterisation methods: indices vs. characterisation protocols. Limnetica 30, 217–234. Fryirs, K.A. (2015). Developing and using geomorphic condition assessments for river rehabilitation planning, implementation and monitoring. WIREs Water 2, 649-667. García Martínez, M.B. (2016). Cambios hidromorfológicos en el Guadalquivir y sus afluentes Guadalbarcar, Rivera de Hueznar y Rivera de Huelva (Sector, Palma del Rio – Sevilla). Tesis Doctoral, Universidad de Sevilla, 627 pp. García Martínez, M.B., Guerrero Amador, I., Baena Escudero, R. (1999). La dinámica de meandros durante el cuaternario reciente en la conformación de la llanura aluvial del bajo Guadalquivir aguas arriba de Sevilla. In: Pallí Buxó L. & Roqué Pau C. (Eds), Avances en el estudio del Cuaternario español, 119-124. García Martínez, M.B., Baena Escudero, R. (2006). El impacto de las infraestructuras de la ciudad de Sevilla sobre el paisaje fluvial del Río Guadalquivir. In: Baena Escudero R. et al. (Eds), Ríos y ciudades europeas: espacios naturales, culturales y productivos, Universidad de Sevilla, 119- 127. García Martínez, M.B., Baena Escudero, R. (2008). El doble meandro abandonado del Guadalquivir en Cantillana (Sevilla): cambios de trazado y evolución geomorfológica. Geographicalia, 53, 101-119. González del Tánago, M., Bejarano, M.D., García de Jalón, D., Schmidt, J.C. (2015). Biogeomorphic responses to flow regulation and fine sediment supply in Mediterranean streams (the Guadalete River, southern Spain). Journal of Hydrology, 528,751-762. Guerrero Amador, I.C. (2016). Evaluación de daños por acción hídrica en puentes de Andalucía Occidental. Tesis Doctoral, Universidad de Sevilla, 638 pp. Guerrero Amador, I., Baena Escudero, R. (2010). Inundaciones fluviales y entorno urbano: consecuencias y estrategias en las ciudades de Praga (republica Checa) y Sevilla (España). In: Abellan F.C. et al. (Eds), Las escalas de la Geografía: del mundo al lugar, 1169-1188. Gurnell, A.M., Belletti, B., Bizzi, S., Blamauer, B., Braca, G., Buijse, A.D., Bussettini, M., Camenen, B., Comiti, F., Demarchi, L., García De Jalón, D., González Del Tánago, M., Grabowski, R.C., Gunn, I.D.M., Habersack, H., Hendriks, D., Henshaw, A., Klösch, M., ! 38! Lastoria, B., Latapie, A., Marcinkowski, P., Martínez-Fernández, V., Mosselman, E ., Mountford, J.O., Nardi, L., Okruszko, T., O’Hare, M.T., Palma, M., Percopo, C., Rinaldi, M., Surian, N., van de Bund, W., Weissteiner, C., Ziliani, L. (2016). A multi-scale hierarchical framework for developing understanding of river behaviour. Aquat. Sci. 78, 1-16. Kondolf, G.M., Wilcock, P.R. (1996). The flushing flow problem: defining and evaluating objectives. Water Resources Research, 32 (8), 2589–2599. Magdaleno, F. (2017). Experimental floods: A new era for Spanish and Mediterranean rivers? Environmental Science and Policy, 75, 10–18. Malavoi, J.R., Bravard, J.P. (2010). Elements d’hydromorphologie fluviale. ONEMA, Baume-Les- Dames, France, 224 pp. Ministerio de Agricultura y Pesca, Alimentación y Medio Ambiente (2017) – Protocolo de caracterización hidromorfológica de masas de Agua de la categorìa Rìos. Código M-R-HMF- 2015, Version 2. Nanson, G.C., Croke, J.C. (1992). A genetic classification of floodplains. Geomorphology, 4(6): 459- 486. Ollero, O.A., Ballarín, F.D., Díaz, B.E., Mora, M.D., Sánchez, F.M., Acín, N.V., Echeverría, A.M.T., Granado, G.D., Ibisate, G.A., Sánchez, G.L., Sánchez, G.N. (2007). Un índice hydrogeomorfológico (IHG) para la evaluación del estado ecológico de sistemas fluviales. Geographicalia 52, 113-141. Ollero, A., Ibisate, A., Gonzalo, L.E., Acín, V., Ballarín, D., Díaz, E., Domenech, S., Gimeno, M., Granado, D., Horacio, J., Mora, D., Sánchez, M. (2011). The IHG index for hydromorphological quality assessment of rivers and streams: updated version. Limnetica 30, 255-262. Piégay, H., Darby, S.E., Mosselman, E., Surian, N. (2005). A review of techniques available for delimiting the erodible river corridor: a sustainable approach to managing bank erosion. River Research and Applications, 21, 773–789. Rinaldi, M., Simoncini, C., Piégay, H. (2009). Scientific design strategy for promoting sustainable sediment management: the case of the Magra River (Central-Northern Italy). River Research and Applications, 25, 607-625. Rinaldi, M., Surian, N., Comiti, F., Bussettini, M. (2013). A method for the assessment and analysis of the hydromorphological condition of Italian streams: the Morphological Quality Index (MQI). Geomorphology 180-181, 96-108. Rinaldi, M., Nardi, L., Belletti, B., Bizzi, S., Brabec, K., Comiti, F., Demarchi, L., Giełczewski, M., Golfieri, B., Habersack, H., Hellsten, S., Kaufman, S., Klösch, M., Marchese, E., Marcinkowski, P., Muhar, S., Okruszko, T., Paillex, A., Poppe, M., Rääpysjärvi, J., Seppo, H., Schirmer, M., Stelmaszczyk, M., Surian, N., Van de Bund, W. (2015). Final report on methods, models, tools to assess the hydromorphology of rivers, Deliverable 6.2, Part 5, of REFORM (REstoring rivers FOR effective catchment Management), a Collaborative project (large-scale integrating project) funded by the European Commission within the 7th Framework Programme under Grant Agreement 282656. Rinaldi, M., Gurnell, A.M., González Del Tánago, M., Bussettini, M., Hendriks, D. (2016a). Classification of river morphology and hydrology to support management and restoration. Aquat. Sci. 78, 17-33. Rinaldi, M., Bussettini, M., Surian, N., Comiti, F., Gurnell, A.M. (2016b). Guidebook for the evaluation of stream morphological conditions by the Morphological Quality Index (MQI). Version 2, Istituto Superiore per la Protezione e la Ricerca Ambientale, http://www.isprambiente.gov.it/pre_meteo/idro/idromorfologia/MQI_final_Aug16_revOct16.pdf Rinaldi, M., Belletti, B., Bussettini, M., Comiti, F., Golfieri, B., Lastoria, B., Marchese, E., Nardi, L., Surian, N. (2017). New tools for the hydromorphological assessment and monitoring of European streams. Journal of Environmental Management, 202, 363-378. ! 39! Sabater, S., Feio, M.J., Graça, M.A.S., Muñoz, I., Romaní, A.M. (2009). The Iberian rivers. In: Tockner K., Uehlinger U., Robinson C.T. (Eds), Rivers of Europe. Academic Press, London, 113–149.!