Què és l' · PDF filesedimentos, morfología del cauce, hábitat)...

Comunicació ICRA: Ramon Balasch, cap de premsa de l'ICRA, 667 55 05 90. Clipmèdia Comunicació: [email protected]

ICRA Roda de premsa

Presentació del llibre Procesos hidrosedimentarios en medios fluviales

de Ramon J. Batalla i Álvaro Tena amb pròleg de Damià Barceló

Barcelona, 16 de novembre de 2016

L'Edifici H20, seu de l'ICRA, està cofinançat en un 50% pel Ministerio de Economia y Competitividad (MINECO) i pel Fons Europeu de Desenvolupament Regional (FEDER) en el marc del Programa Operatiu FEDER de

Catalunya 2007-2013).

mailto:[email protected]


ÍNDEX

1. Fitxa tècnica

2. Índex del llibre.

3. Pròleg de Damià Barceló

4. Introducció

5. Capítol 1. Morfodinàmica fluvial (primeres pàgines)

ANNEXOS: Materials i articles relacionats


Estudio geográfico

C/ San Salvador, 8 25005 Lleida (Spain) Tel. +34 973 23 66 11 Fax. +34 973 24 07 95www.edmilenio.com [email protected]

Pedidos: 902 885 [email protected]

www.nusdellibres.com

Procesos hidrosedimentarios en medios fluvialesRamon J. Batalla y ÁlvaRo tena

El funcionamiento natural de los ríos depende de la interacción entre los distintos elementos y procesos físicos (flujo de agua y de sedimentos, morfología del cauce, hábitat) y biológicos (comuni-dades bentónicas, vegetación acuática y de ribera, metabolismo) que forman el cauce, así como de las características físico-quí-micas del agua (temperatura, turbidez, composición). Este libro presenta una serie de técnicas y métodos para la observación, medición y modelización de diversos de estos procesos, desar-rollados y aplicados en el marco del Proyecto de investigación SCARCE-Consolider, financiado por el Ministerio de Economía y Competitividad entre 2009-2014.

Colección: Alfa, 60

ISBN: 978-84-9743-732-5

17 x 24 cm

276 páginas

Rústica con solapas

PVP: 20,00 €

Geografía Física

1. El libro contiene suficientes elementos bási-cos para que su lectura sea de lo más amena para todo este conjunto de profesionales del sector así como para científicos y técnicos aje-nos a esta materia.

2. Siete capítulos que cubren los diferentes aspectos de funcionamiento de los ríos, cen-trados en elementos físicos, biológicos, co-munidades bentónicas, vegetación y la cali-dad fisicoquímica del río.

3. De interés no solo para estudiantes univer-sitarios en las especialidades de hidráulica, hidrología, ecología, química ambiental, mo-delización ambiental y geografía y también para los responsables de la gestión y técnicos de los organismos de cuenca y las agencias del agua.

.

Interés comercial y editorial

9 788497 437325

Ramon J. Batalla es geógrafo, especializado en Geomorfología Fluvial. Obtuvo el doctorado en Geografía por la UB en 1993 y es profesor de Geografía Física en la Universitat de Lleida desde 1996. Lidera el Grupo de Investigación sobre Dinámica Fluvial —RIUS— y es investigador adscrito al Instituto Catalán de Investigación del Agua —ICRA. Ha dedicado su carrera al estudio de los procesos fl uviales, particularmente el transporte de sedimentos y la dinámica de los cauces, y los efectos de las actividades antrópicas en los ríos y las cuencas de drenaje.

Álvaro Tena es geógrafo, especializado en Geomorfología Fluvial. Realizó su tesis doctoral en la Universitat de Lleida (2012), donde luego continuó como investigador postdoctoral dentro del proyecto Consolider-SCARCE. Actualmente trabaja como investigador postdoctoral del Centre National de la Recherche Scientifique, en Lyon. Su actividad investigadora está relacionada con el transporte de sedimentos y la dinámica morfosedimentaria, así como el análisis de los efectos de las actividades antrópicas en los procesos fluviales.

Otros títulos de la colección:

9 788497 432894 9 788497 433921

‹ 5

ÍNDICE

Prólogo ...................................................................................................... 7

Introducción ............................................................................................. 11

Capítulo 1. Morfodinámica fluvial ........................................................ 19

Capítulo 2. Medición de procesos hidrosedimentarios ...................... 75

Capítulo 3. Modelación hidráulica de cauces ..................................... 109

Capítulo 4. Modelización hidrológica de cuencas .............................. 157

Capítulo 5. Transporte de microcontaminantes orgánicos asociados a materia y sedimentos en suspensión .............................. 187

Capítulo 6. Sedimentos y macroinvertebrados ................................... 219

Capítulo 7. La madera muerta y la morfología fluvial: ecología y gestión ................................................................................... 245

‹ 7

PRÓLOGO

Damià Barceló

En los alrededores de Navidad del 2015 el amigo Ramon Batalla, profesor de la Universitat de Lleida (UdL) y científico del Institut Català de Recerca de l’Aigua (ICRA), sito en Girona, me pidió si podía redac-tar el prólogo de este libro del que es alma mater y editor, juntamente con su discípulo Álvaro Tena, y que ahora tiene el lector en sus manos.

No fue una petición hecha al azar sino que se debe a que Ramon fue el investigador principal de la UdL en el proyecto SCARCE Evalu-ación y predicción de los efectos del cambio global en la cantidad y calidad del agua en ríos ibéricos financiado por el Ministerio de Ciencia e Innovación, posteriormente denominado Ministerio de Economía y Competitividad, que coordiné entre los años 2009 y 2014.

El proyecto SCARCE tiene dos objetivos fundamentales, uno de ellos más científico, de investigación básica cuya meta fue analizar las pautas a largo plazo y los mecanismos que operan en la hidrología, la calidad de agua, etc., y el segundo más relacionado con la gestión de cuencas atendiendo al impacto del cambio climático sobre ellas y la huella huma-na, ambos elementos del cambio global. La integración es un elemento fundamental de SCARCE donde la investigación básica se dirige a ayudar a las políticas normativas de la Unión Europea, en especial la Directiva Marco del Agua. El objetivo último y fundamental del proyecto es, por tanto, el de analizar de manera conjuntas los diferentes factores de es-trés que afectan al ecosistema fluvial y poder influir positivamente en los planes hidrológicos de cuenca exigidos por la Directiva. El proyecto, además, aborda cuestiones importantes tales como: de qué manera hay que gestionar la cuenca cuando hay escasez de agua, cómo afectan a los diversos elementos del ecosistema fluvial tales como macro invertebrados y biofilm los factores de estrés más relevantes como la intermitencia del caudal, los nutrientes y los contaminantes emergentes.

He de señalar que ha sido un gran placer coordinar el proyecto SCARCE y me voy a permitir detallar algunos de sus aspectos más

8 ›

notorios. A lo largo de los cinco años de duración se realizaron mues-treos extensivos de aguas, sedimentos y peces en cuatro cuencas hidro-gráficas: Ebro, Llobregat, Guadalquivir y Júcar a fin de determinar la calidad química y ecológica tanto en condiciones de escasez como de abundancia de agua. Además se programaron diferentes experimentos tanto a escala de laboratorio como en subcuencas para determinados estudios de transporte de sedimentos, modelización e impactos en los ecosistemas entre otros.

El proyecto SCARCE generó una gran actividad científica en forma de tesis doctorales, congresos y artículos, todos ellos en inglés, pu- blicándose los resultados en revistas de gran prestigio internacional. Este libro es la primera gran obra del proyecto SCARCE en castellano con la finalidad de que tenga una mayor difusión y esté a disposición no solo de los científicos que trabajan en materias vinculadas a la calidad de las aguas, y la hidrología y la morfología de los ríos, ya sean profesores universitarios o estudiantes de doctorado y máster, sino también entre los gestores y técnicos al servicio en los distintos organismos de cuenca y agencias del agua quienes precisan de textos científico-técnicos que les sean de utilidad en el ejercicio de sus labores y funciones para la correcta gestión medioambiental de las cuencas.

El libro contiene suficientes elementos básicos y aplicados para que su lectura sea de lo más amena para todo este conjunto de pro-fesionales del sector así como para científicos y técnicos ajenos a esta materia. Debido a la formación y especialidad de los editores, el libro está muy centrado en la morfología del río, integrando, junto con los cambios en los usos del suelo y las características hidroclimáticas de las cuencas, los aspectos de dinámica morfosedimentaria y de transporte de sedimentos así como su relevancia como hábitat tanto para peces como macroinvertebrados. El libro plantea diferentes escenarios dentro de lo que denominamos cambio global, como por ejemplo el cambio de uso del suelo, para poder predecir diferentes estrategias de adaptación a corto y largo plazo.

El libro consta de siete capítulos que cubren los diferentes aspectos de funcionamiento de los ríos, centrados en elementos físicos, como agua y sedimentos, biológicos, comunidades bentónicas, y vegetación así como la calidad fisicoquímica del río. Desde la introducción general sobre el tema, hasta la medición de los procesos hidrosedimentarios, la modelización hidráulica e hidrológica de cuencas, el transporte de con-taminantes orgánicos asociados a sedimentos y materiales en suspensión, los impactos de la regulación en la hidrología, geomorfología y macroin-vertebrados, hasta la materia muerta de los ríos, como los troncos, las

‹ 9

ramas gruesas, los árboles, que son un elemento importante en la gestión del río proporcionando abundante material orgánico y refugio para organismos potenciando así el funcionamiento ecológico del río.

No puedo dejar de dar las gracias a Ramon por su iniciativa para hacer este libro y a Álvaro por su trabajo como coeditor, así como a todos los autores que han redactado los diferentes capítulos que son reconocidos especialistas en cada uno de los temas tratados. A todos ellos vaya mi reconocimiento por el esfuerzo realizado para dar una difusión tan amplia del proyecto SCARCE en lengua castellana.

Confío en que este libro tenga un gran interés no solo para los diferentes estudiantes universitarios en las especialidades de hidráulica, hidrología, ecología, química ambiental, modelización ambiental y geografía sino también para los responsables de la gestión y técnicos de los organismos de cuenca y las agencias del agua. La información básica y aplicada, que se ha compilado en este libro, es relevante y puede ayudar a tomar decisiones para una gestión más eficaz de nuestras cuencas, especialmente en situaciones difíciles, o incluso críticas, como por ejemplo cuando se producen avenidas, inundaciones y sequías.

Damià Barceló

Director del ICRA- Institut Català de Recerca de l’Aigua, Girona,

y Profesor de Investigación del IDAEA-CSIC, Barcelona

‹ 11

El funcionamiento de los sistemas fluviales depende de la interacción entre sus elementos físicos (agua, sedimentos, morfología), biológicos (comunidades bentónicas, vegetación) y de la calidad fisicoquímica del agua (Figura 1). Las relaciones entre ellos son variables en el espacio y en el tiempo, dependen del comportamiento de la cuenca de drenaje aguas arriba, y determinan el estado ecológico actual y futuro de los ríos.

Las cuencas y los ríos experimentan di-rectamente los efectos del cambio global en el planeta. Por ejemplo, los ríos en España han perdido una media del 10% del caudal en las últimas décadas, ha disminuido notablemente su dinamismo físico y han visto reducida su biodiversidad. La reducción en la disponibilidad de los recursos hídricos y sus efectos sobre los ecosistemas acuáticos debe mitigarse con una correcta gestión de dichos recursos teniendo en cuenta criterios de funcionamiento ecológico. El estudio de la dinámica fluvial de un río permite avanzar en el conocimiento de las condiciones biofísicas de referencia de dichas masas de agua y la comparación con tramos alterados, así como el diseño y la aplicación de programas de rege-neración ambiental, recuperación de caudales y mejora del estado ecológico de los ríos.

Un caso de alteración hidrosedimentaria y biogeoquímica ampliamente conocido y docu-mentado lo tenemos, por ejemplo, en los ríos regulados. En regiones de clima muy variable, como es el caso de las cuencas mediterráneas es

Ramon J. Batalla1,2

es geógrafo especializado en Geomorfología Flu-vial. Doctor en Geografía por la Universidad de Barcelona (1993) y Profesor Titular de Geografía Física de la Universidad de Lleida desde 1999. Lidera el Grupo de Investigación sobre Dinámica Fluvial —RIUS— y es investigador adscrito al Ins-tituto Catalán de Investigación del Agua —ICRA— desde 2011 como responsable de la Línea de Procesos Hidrológicos. Ha dedicado su carrera al estudio de los procesos fluviales, sobre todo el transporte de sedimentos y la dinámica de cauces, y el impacto de las actividades antrópicas en los ríos y las cuencas de drenaje.

Álvaro Tena1,3

es geógrafo especializado en Geomorfología Flu-vial. Realizó su tesis doctoral en la Universidad de Lleida (2007-2012), donde luego continuó como investigador postdoctoral dentro del proyecto Consolider-SCARCE (2012-2014). Actualmente trabaja como investigador Postdoctoral del Cen-tre National de la Recherche Scientifique, en l’École Normale Supérieure de Lyon. Su actividad investigadora está relacionada con el transporte de sedimentos, la dinámica sedimentaria y los procesos geomorfológicos asociados.

1. Grupo de Investigación de Dinámica Fluvial. RIUS. Universidad de Lleida, Lleida. [email protected]

2. Instituto Catalán de Investigación del Agua, Girona. [email protected]

3. ENS de Lyon, UMR 5600 - EVS CNRS, Plateforme ISIG, 15, Lyon Cedex 0, [email protected]

INTRODUCCIÓN

Ramon J. Batalla, Álvaro Tena

12 ›

Relaciones entre los diferentes componentes biofísicos en el medio fluvial. (Figura: Damià Vericat)

Figura 1

donde históricamente se ha producido una mayor regulación del caudal y también donde las alteraciones ecológicas provocadas por la presencia de una presa son mayores. El impacto es importante, principalmente, por la alteración del régimen natural de caudales de los ríos generando, en consecuencia, impactos en la biota (Ward y Stanford, 1979) y en el transporte de sedimentos, ya que altera el régimen hidrosedimentario (Williams y Wolman, 1984). Las crecidas ven disminuida su magnitud y frecuencia, la aportación hídrica anual se ve reducida, y tienen lugar cambios hidrológicos estacionales (Petts, 1984). Para su estudio, es nece-saria una caracterización del estado ecológico de las masas de agua en relación a su régimen hídrico y sedimentario. La restauración ecológi-ca de los ríos ha ganado importancia a nivel internacional y nacional, aunque todavía son pocos los trabajos relacionados con el diseño de métodos de renaturalización total o parcial y el estudio de sus efectos directos sobre el funcionamiento físico de los ríos. A nivel nacional, el diseño y puesta en práctica de medidas de restauración del régimen hídrico y sedimentario de ríos regulados mediante la implementación de crecidas de mantenimiento ya ha sido utilizada, por ejemplo, en el

‹ 13

Ebro. Las crecidas controladas existentes en programas de gestión hi-drológica generan i) una eliminación de los elementos finos y vegetales del cauce, ii) un control sobre la entrada de vegetación madura en el cauce activo, y iii) una removilización del lecho de gravas generando lugares de freza de mejor calidad. Para una gestión sostenible de los ríos mediterráneos y para poder recuperar un funcionamiento normal y un equilibrio de los ecosistemas son necesarios este tipo de crecidas que, además, derivan en funciones de mantenimiento y renovación del ecosistema fluvial. Se trata de conocimientos y acciones fundamentales para el diseño e implementación de los Planes de Gestión de Cuenca en el marco de la Directiva Marco del Agua (2000/60/EC).

Este libro presenta una serie de métodos y técnicas para la mo-nitorización y modelización de procesos fluviales y de cuenca relacio-nados con el flujo de agua, sedimentos y contaminantes, la dinámica morfosedimentaria de los lechos, y las comunidades de invertebrados bentónicos que en ellos habitan. Los diferentes capítulos son el fruto de los distintos trabajos de investigación desarrollados en el marco del Proyecto SCARCE-Consolider, financiado por el Ministerio de Economía y Competitividad, y desarrollado entre los años 2009-2014 por una se-rie de grupos de investigación dentro de los cuales se encuentran los autores de este trabajo. Como ejemplo, la Figura 2 presenta el marco conceptual que guió las interacciones entre los diferentes paquetes de trabajo relacionados con el análisis hidrosedimentario en el proyecto. Las ilustraciones que el libro contiene son el resultado de los distintos estudios llevados a cabo en el contexto del proyecto.

SCARCE ha sido un proyecto multidisciplinar cuyo principal objetivo era describir y predecir la relevancia de los impactos del cambio global sobre la disponibilidad de agua, su calidad y los servicios ecosistémicos en las cuencas del Mediterráneo de la Península Ibérica, así como sus impactos en la sociedad humana y la economía. Para ello, el proyecto reunió la experiencia de científicos líderes en muy diversos campos, como la hidrología, la geomorfología, la química, la ecología, la ecotoxicolo-gía, la economía, la ingeniería y la modelización, sin precedentes en el marco del Programa Consolider. Además en el proyecto se involucraron activamente las autoridades del agua y sectores afines. Tal y como explica la propia página web del proyecto (www.idaea.csic.es/scarceconsolider/publica/P000Main.php) el agua es un factor fundamental para el desa-rrollo socioeconómico en toda la cuenca mediterránea, a pesar de su variabilidad temporal. La manipulación antropogénica cada vez mayor de la hidrología y la exacerbación del cambio climático han dado lugar a una variabilidad temporal aún mayor. En particular, los cursos de agua

14 ›

de la cuenca mediterránea están sometidos a severas alteraciones en su régimen de flujo, debido a una disminución en el número de días de precipitación y un aumento de días con lluvias intensas. El desequilibrio entre los recursos de agua disponibles durante las sequías prolongadas y el aumento de la demanda de agua son la causa de los principales problemas ecológicos y económicos. En consecuencia, la disponibilidad de agua se ha convertido en una cuestión importante para todos los gobiernos en las regiones mediterráneas. Sin embargo, las consecuencias del cambio global no solo afectarán a la disponibilidad de agua, sino también a su calidad y a los servicios ecosistémicos.

Ejemplo de interacción entre trabajos y paquetes de trabajo en el marco de SCARCE. (Figura: Damià Vericat)

Figura 2

‹ 15

En este contexto D. Vericat y R. J. Batalla presentan un primer capítulo en el que se explica la morfo-dinámica como el estudio de los procesos erosivos y sedimentarios que afectan la forma y la evo-lución de los cauces de los ríos. El análisis de las relaciones entre las formas y los procesos permite entender los mecanismos de formación y modificación de las unidades geomorfológicas. Se trata de relaciones causa-efecto interrelacionadas y altamente variables en el espacio y en el tiempo. En el capítulo se describen los aspectos fundamentales para el estudio de la dinámica fluvial, y se presentan una serie de variables de interés y se analizan diversos métodos de adquisición de datos de campo y de postproceso, poniendo el énfasis en la importancia de las escalas espaciales y temporales, y al análisis crítico de la incertidumbre asociada a los resultados. Finalmente, se identifican una serie de aspec-tos a tener en cuenta para planificar, diseñar y evaluar aproximaciones metodológicas para el estudio de la morfo-dinámica fluvial.

En esta misma línea J. A. López-Tarazón y A. Tena enfatizan que el control de variables hidrosedimentarias y morfológicas es fundamental para entender la dinámica fluvial, para caracterizar y modelizar los pro-cesos fluviales asociados, e incluso para proporcionar el soporte técnico a proyectos de rehabilitación de tramos fluviales alterados. En su capítulo recogen una serie de metodologías ampliamente utilizadas en los campos de la hidrología y la geomorfología fluvial, pero que también pueden ser utilizados en otras disciplinas científicas relacionadas (p. ej. ecología fluvial, eco-geomorfología, etc.). El capítulo presenta además una serie de directrices y protocolos (internacionalmente aceptados y homologados) para el muestreo y la medida del caudal y del transporte de sedimento, así como de la estructura y características físicas del cauce fluvial, que son las variables consideradas por los autores como fundamentales para la ejecución de cualquier tipo de proyecto (tanto técnico como científico) en los campos de investigación anteriormente mencionados.

En su capítulo F. J. Vallés e I. Andrés explican como el flujo de agua y sedimentos en un cauce natural es un proceso dependiente del tiempo y en el que intervienen las tres dimensiones espaciales. Estos flujos además interaccionan con los contornos, lecho y márgenes, a través de procesos de erosión y deposición, cambiando de esta forma la morfología fluvial. Obviamente, el material erosionado es transportado por el flujo hasta su deposición. El flujo bifase resultante, de agua y sedimentos, también modifica la forma de interacción con los contornos fluviales. La modelación matemática de todos estos procesos, tanto para predicciones hidráulico-morfológicas como con fines ingenieriles, requie-re de un alto grado de esquematización. En el capítulo se presentan,

16 ›

ilustrados con diferentes casos de estudio, los conceptos fundamentales de la modelación hidráulica considerando los principales procesos en la interacción cauce-flujo.

G. Bussi y F. Francés explican la importancia de los modelos ma-temáticos que se implementan para predecir la variable de estado de interés en escenarios de no observación de la misma, sea en el tiempo o en el espacio. En su capítulo las variables de interés son los inputs de agua y sedimentos de un tramo de río, provenientes de su cuenca vertiente. La propuesta es la utilización de modelos distribuidos, ya que permiten obtener resultados en cualquier punto del territorio, aunque no sea aforado, y tener mejor en cuenta las no linealidades asociadas a los eventos de crecida de agua y sedimentos, fundamentales en la construcción de la morfodinámica fluvial.

Por su parte A. Ginebreda, A. Tena y Damià Barceló, examinan de forma cualitativa y cuantitativa el transporte fluvial de microcontaminan-tes orgánicos asociado a la materia en suspensión. En primer lugar se indican las propiedades físico-químicas y demás condiciones ambientales que determinan el equilibrio de partición de un compuesto orgánico entre las fases sólida y disuelta, mencionándose las familias de contaminan-tes asociados a materia en suspensión más relevantes y que, por ello, han sido objeto de atención por parte de la legislación ambiental. En segundo lugar se describen los métodos de cálculo de los flujos másicos de un contaminante asociados a la materia en suspensión durante un intervalo de tiempo a través de una sección de río, así como el de los balances másicos entre dos puntos de medida consecutivos. Así mismo, se describen de forma resumida los métodos experimentales asociados a la toma de muestra, medida de parámetros hidrológicos, técnicas de extracción de contaminantes y métodos analíticos para su cuantificación comúnmente empleados. Todo ello se ilustra con un ejemplo del Bajo Ebro (Flix) tomado de la experiencia de los autores, en el que se cuantifica el balance másico de diversas familias de contaminantes (hidrocarburos aromáticos policíclicos, DDT y sus productos de transformación, PCB y otros compuestos organoclorados) durante una crecida de río provocada artificialmente.

G. Lobera e I. Muñoz describen los efectos de la regulación sobre las comunidades de invertebrados de los ríos mediterráneos. Las autoras indican que los embalses modifican el régimen de caudal y la dinámica morfosedimentaria del cauce. Estas alteraciones suelen provocar cambios en la composición y abundancia de los macroinvertebrados bentónicos, que son indicativos de la degradación del medio, y muchas veces no pueden ser detectados mediante la utilización de índices bióticos. Por

‹ 17

esta razón, proponen un protocolo de muestreo y cálculo de indicadores para evaluar el impacto hidromorfológico en ríos regulados y su efecto en la comunidad de macroinvertebrados. Como ejemplo, se presenta un estudio realizado en el río Siurana para evaluar el impacto de la regulación en la hidrología, la geomorfología y la comunidad de ma-croinvertebrados, siguiendo una aproximación de contraste entre tramo control y tramo impactado.

Finalmente, J. R. Diez, M. M. Sarriegui y A. Elósegi explican la im-portancia de la madera en la morfología de los cauces fluviales. Aunque antiguamente la madera muerta se consideraba como un elemento ajeno y fuente de problemas en los cauces, desde hace décadas se sabe que es una pieza fundamental de los mismos, dado que modela la morfología del cauce, retiene sedimentos y materia orgánica, proporciona comida y refugio a innumerables organismos y potencia el funcionamiento de los ecosistemas fluviales. Sin embargo, las autoridades hidráulicas, junto a los ribereños, dedican importantes recursos a retirar los árboles y troncos del cauce en prevención de riesgos de erosión o de inundación, para facilitar la migración de peces, o por simples razones estéticas. De ese modo, se contribuye a la simplificación de ríos y arroyos, de su biodi-versidad y a la pérdida de su funcionamiento y, en último término, de los servicios ecosistémicos asociados. En el capítulo los autores describen el origen y dinámica de la madera muerta en los cauces y el papel que desempeña sobre la morfología y sobre los procesos ecológicos de ríos y arroyos. También proponen una guía metodológica para su gestión, de forma que se maximicen sus funciones ecológicas mientras se minimiza el riesgo asociado.

Agradecimientos

Este libro se ha elaborado en el marco del Proyecto Consolider SCARCE 2010 CSD2009-00065 financiado por el Ministerio de Economía y Competitividad.

Referencias bibliográficas

Petts, g. e. 1984. Impounded Rivers. Perspectives for Ecological Management. Wiley: Nueva York; 326 p.

Ward, J. a.; stanFord, J. a. 1979. The ecology of Regulated Streams. Plenum Press: Nueva York.

Williams, g. P.; Wolman, m. g. 1984. Downstream Effects of Dams on Alluvial Rivers. US Geological Survey Professional Paper 1986.

‹ 19

CAPÍTULO 1

MORFODINÁMICA FLUVIAL

Damià Vericat, Ramon J. Batalla

Damià Vericat1,2

es geógrafo especializado en Dinámica Fluvial. Se doctoró en Geografía en la Universidad de Lleida en 2005, donde actualmente es investigador Ra-món y Cajal. Es investigador adscrito del Centro Tecnológico Forestal de Cataluña y líder del área de Hidrología. Su investigación se centra en el estudio de la dinámica morfológica y sedimentaria de sistemas fluviales. El objetivo de su investigación recae en la interacción entre hidráulica, transporte de sedimentos y morfología fluvial en múltiples escalas espaciales y temporales.

Ramon J. Batalla1,3

es geógrafo especializado en Geomorfología Flu-vial. Doctor en Geografía por la Universidad de Barcelona (1993) y Profesor Titular de Geografía Física de la Universidad de Lleida desde 1999. Li-dera el Grupo de Investigación sobre Dinámica Flu-vial —RIUS— y es investigador adscrito al Instituto Catalán de Investigación del Agua —ICRA— desde 2011 como responsable de la Línea de Procesos Hidrológicos. Ha dedicado su carrera al estudio de los procesos fluviales, particularmente el trans-porte de sedimentos y la dinámica de cauces, y el impacto de actividades antrópicas en los ríos y las cuencas de drenaje.

1. Grupo de Investigación de Dinámica Fluvial. RIUS. Universidad de Lleida, Lleida. [email protected], [email protected]

2. Centro Tecnológico Forestal de Catalunya, Solsona. [email protected]

3. Instituto Catalán de Investigación del Agua. Girona. [email protected]

Resumen

Se entiende como morfodinámica el estudio de los procesos erosivos y sedimentarios que afectan la forma y la evolución de los cauces de los ríos. El análisis de las relaciones entre las formas y los procesos permite entender los mecanismos de formación y modificación de las unidades geomorfológicas. Se trata de relaciones causa-efecto interrelacionadas y altamente variables en el espacio y en el tiempo. En este capítulo se des-criben diferentes aspectos fundamentales para el estudio de la morfodinámica fluvial. Se presentan una serie de variables de interés y se analizan diversos métodos de adquisición de datos de campo y de postproceso, poniendo el énfasis en la importancia de las escalas espaciales y temporales, y al análisis crítico de la incertidumbre asociada a los resultados. Finalmente, se iden-tifican una serie de aspectos a tener en cuenta para planificar, diseñar y evaluar aproximaciones metodológicas para el estudio de la morfodinámica fluvial.

Abstract

We understand as morpho-dynamics as the study of erosion and sedimentation processes affecting the form and evolution of fluvial systems. The relations between form and processes allow understanding the mechanisms of formation of geomorpholog-ical units. These are interrelated cause-effect relations highly variable in space and time. This chapter describes main funda-mental aspects for the study of morpho-dynamics in rivers. First, we present and describe a list of variables of interest. Secondly, various methods of field data acquisition and post-processing are discussed; giving emphasis to the spatial and temporal scales, and the analysis of data uncertainty. Finally, we identify a number of key points to consider when field data acquisition and post-processing designs to study morpho-dynamics in fluvial systems are prepared.

20 ›

Introducción

La forma de un río es el resultado de la interacción de numero-sos procesos controlados principalmente por el régimen de crecidas, el transporte de sedimentos y la vegetación (y la madera) en el cauce y en la llanura de inundación. A nivel geomorfológico, un río se puede dividir en distintas unidades o componentes caracterizadas por su forma, geometría y estructura sedimentaria. La evolución temporal de estas unidades está regida por los procesos erosivos y sedimentarios de los materiales que las conforman, procesos que a su vez están determinados por la competencia del flujo circulante y el subministro y disponibilidad de sedimentos. Al mismo tiempo, la efectividad hidráulica del flujo en un punto determinado (competencia para movilizar los materiales) está influenciada por las características morfológicas y sedimentarias de dichas unidades. Así pues, se trata de relaciones causa-efecto interrelacionadas y altamente variables en el espacio y en el tiempo.

Las unidades geomorfológicas que se observan en los sistemas flu-viales se pueden agrupar en: (a) unidades de cauce y (b) unidades de llanura de inundación. Dentro de estas dos categorías existen numerosas formas caracterizadas por su geometría, los materiales que las conforman y los procesos dominantes que controlan su formación y evolución. Por ejemplo, dentro de las unidades de cauce se encuentran los rápidos y pozas (Figura 1). Estas unidades difieren substancialmente a nivel topo-gráfico y granulométrico, forman secuencias que se repiten con cierta frecuencia (espacio), y disponen de características hidráulicas distintas y variables en función del caudal. Consecuentemente, los procesos erosivos y de sedimentación en estas unidades serán variables y controlaran su formación y presencia a lo largo del tiempo.

En este capítulo se presentan diversos métodos para el estudio de la morfodinámica fluvial, poniendo el énfasis en la importancia de las escalas espaciales y temporales, y al análisis crítico de la incertidumbre asociada a los resultados mediante casos de estudio con un contrasta-do dinamismo. Cabe destacar que en este capítulo no se presenta una descripción detallada de las distintas unidades de cauce o de llanura de inundación, aunque se aconseja su revisión antes de acometer cualquier trabajo. Por ejemplo, Fryirs y Brierley (2013) presentan una descripción pormenorizada de las principales unidades geomorfológicas fluviales (de cauce en el capítulo 8; y de llanura de inundación en el capítulo 9), acompañada de un análisis de su forma, y una explicación de los prin-cipales procesos responsables y su interacción.

Este capítulo se divide en cuatro secciones. En primer lugar se des-cribe en qué consiste el estudio de la morfodinámica fluvial, detallando

‹ 21

cuáles son los aspectos fundamentales a tener en cuenta y los principales atributos o variables que se pueden obtener y analizar en relación con las unidades de cauce. Seguidamente se presenta una revisión de métodos para el estudio de la morfodinámica fluvial. A continuación se identifi-can algunas de las oportunidades y retos que actualmente nos ofrecen los avances tecnológicos en este campo. Finalmente, se presentan una serie de aspectos a tener en cuenta para planificar, diseñar y evaluar aproximaciones metodológicas para el estudio de la morfodinámica en sistemas fluviales.

Figura 1

Ejemplos de unidades geomorfológicas en el tramo alto del río Ésera aguas abajo del municipio de Campo (cuenca del río Cinca) y en el tramo alto del río Cinca

aguas arriba de Escalona (cuenca del río Ebro). En el caso del río Ésera se observan secuencias de barras laterales con poca vegetación acompañadas de secuencias de rápidos y pozas. En el caso del río Cinca se observa una amplia llanura de

inundación con diferentes niveles de movilidad. Se trata de un río que históricamente había tenido un patrón trenzado con un patrón actual sinuoso o wandering debido

a la reducción del aporte sedimentario por motivos naturales y antrópicos. La vegetación de ribera ha estabilizado parte de los depósitos. Se observa vegetación

arbustiva y arbórea en lo que históricamente era la llanura activa del patrón trenzado. Se pueden identificar fácilmente barras laterales o diagonales alternadas con rápidos y pozas, y algunas zonas que se podrían caracterizar como de glides o runs. Se aconseja revisar los capítulos 8 y 9 de Fryirs y Brierley (2013) para obtener una correcta distinción de varias unidades de cauce o de llanura de inundación que se

pueden observar en medios fluviales.


ANNEXOS: Materials i articles relacionats

“Dynamics of suspended sediment borne persistent organic pollutants in a large regulated Mediterranean river (Ebro, NE Spain)” de S. Quesada, A. Tena, D. Guillén, A. Ginebreda, D. Vericat, E. Martínez, A. Navarro-Ortega, R.J. Batalla, D. Barceló (Elsevier, Agosto 2013)

“Geomorphic status of regulated rivers in the Iberian Peninsula” de G. Lobera, P. Besné, D. Vericat, J.A. López-Tarazón, A. Tena, I. Aristi, J.R. Díez, A. Ibisate, A. Larrañaga, A. Elosegi, R.J. Batalla (Elsevier, July 2014)

“Coupling channel morphology and ecological diversity in a gravel bed river: morphsed conceptual approach and experimental design” de D. Vericat, R.J Batalla, C.N. Gibbins, J. Brasington, J., A. Tena, M. Béjar, E. Muñoz-Narciso, E. Ramos, G. Lobera, C. Buendia, J.A. López-Tarazón M. Smith, J. Wheaton, R. López, J. Verdú i A. Palau


Science of the Total Environment 473–474 (2014) 381–390

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv

Dynamics of suspended sediment borne persistent organic pollutants in alarge regulated Mediterranean river (Ebro, NE Spain)☆

S. Quesada a, A. Tena b, D. Guillén a, A. Ginebreda a,⁎, D. Vericat b,d,e, E. Martínez a, A. Navarro-Ortega a,R.J. Batalla b,c,d, D. Barceló a,c

a Department of Environmental Chemistry, IDAEA-CSIC, c/Jordi Girona 18-26, E-08034 Barcelona, Spainb Department of Environment and Soil Sciences, University of Lleida, E-25198 Lleida, Spainc Catalan Institute for Water Research (ICRA), Emili Grahit 101, E-17003 Girona, Spaind Forest Technology Centre of Catalonia, E-25280 Solsona, Spaine Institute of Geography and Earth Sciences, Aberystwyth University, Wales, Ceredigion SY23 3DB, UK

G R A P H I C A L A B S T R A C T

☆ This is an open-access article distributed under the tedistribution, and reproduction in any medium, provided t⁎ Corresponding author. Tel.: +34 934006100.

E-mail address: [email protected] (A. Gin

0048-9697/$ – see front matter © 2013 The Authors. Pubhttp://dx.doi.org/10.1016/j.scitotenv.2013.11.040

a b s t r a c t
a r t i c l e i n f o
Article history:Received 7 August 2013Received in revised form 6 November 2013Accepted 7 November 2013Available online 28 December 2013

Keywords:Suspended sedimentsPersistent organic pollutantsFlushing flowsPollutants' transportMass loadsRiver Ebro

Mediterranean rivers are characterized by highly variable hydrological regimes that are strongly dependent onthe seasonal rainfall. Sediment transport is closely related to the occurrence of flash-floods capable to deliverenough kinetic energy to mobilize the bed and channel sediments. Contaminants accumulated in the sedimentsare likely to be mobilized as well during such events. However, whereas there are many studies characterizingcontaminants in steady sediments, those devoted to the transport dynamics of suspended-sediment borne pol-lution are lacking. Here we examined the occurrence and transport of persistent organic microcontaminantspresent in the circulating suspended sediments during a controlled flushing flow in the low part of the RiverEbro (NE Spain) 12 km downstream of a well-known contaminated hot-spot associated to a nearby chloro-alkali industry. Polycyclic aromatic hydrocarbons (PAHs) and semi-volatile organochlorine pollutants (DDTand related compounds, DDX; polychlorinated byphenils, PCBs; and other organochlorine compound, OCs)were measured in the particulate material by GC–MS and GC–MS/MS, using previously developed analytical

rms of the Creative Commons Attribution-NonCommercial-No Derivative Works License, which permits non-commercial use,he original author and source are credited.

ebreda).

lished by Elsevier B.V. All rights reserved.

http://crossmark.crossref.org/dialog/?doi=10.1016/j.scitotenv.2013.11.040&domain=pdf

http://dx.doi.org/10.1016/j.scitotenv.2013.11.040



http://www.sciencedirect.com/science/journal/00489697

382 S. Quesada et al. / Science of the Total Environment 473–474 (2014) 381–390

methods. The concentration levels observedwere compared to previously reported values in steady sediments inthe same river and discussed on a regulatory perspective. Hydrographs and sedigraphs recorded showed a peak-flowof 1300 m3 s−1 and a corresponding peak of suspended sediments of 315 mg L−1. Combination of flowdis-charge, suspended sediments and pollutants' concentrations data allowed for quantifying the mass flows (massper unit of time) and setting the load budgets (weight amount) of the different pollutants transported by theriver during the monitored event. Mean mass-flows and total load values found were 20.2 mg s−1 (400 g) forPAHs, 38 mg s−1 (940 g) for DDX, 44 mg s−1 (1038 g) for PCBs and 8 mg s−1 (200 g) for OCs. The dynamic pat-tern behavior of PAHs differs substantially to that of organochlorine pollutants, thus reflecting different pollutionorigins.

© 2013 The Authors. Published by Elsevier B.V. All rights reserved.

1. Introduction

Water scarcity is present in natural conditions in the Mediterraneanarea due to the characteristic highly variable river flows and the occur-rence of low flow characteristic of the Mediterranean climate. Alterna-tion of extreme events such as droughts and flash-floods is commonin the region, a fact that will likely accentuate according to the previ-sions of the IPCC (Lehner et al., 2006). Sediment transport is closely re-lated to the occurrence of flood events (Batalla et al., 1995; Garcia et al.,2000; Vericat and Batalla, 2010) capable to deliver enough kinetic ener-gy to mobilize the riverbed and entrain the sediments. Likewise, it isknown that contaminants accumulated in the sediments are mobilizedduring such energetic events. However, whereas there aremany studiescharacterizing contaminants in steady sediments (Fernández et al.,1999; Carrizo and Grimalt, 2006; Lacorte et al., 2006; Navarro-Ortegaet al., 2010), only few are devoted specifically to the transport dynamicsof pollutants associated to the suspended sediment load (Gómez-Gutiérrez et al., 2006). In addition, quantitative information concerningthe load budgets of persistent organic pollutants associated to particu-late material transport is scarce (Schwientek et al., 2013; Rügner et al.,2013). Within this context, flushing flows (FF), which are controlledfloodflows performed from a dam for a given environmental and/or en-gineering purpose (Kondolf and Wilcock, 1996; Batalla and Vericat,2009), offer some advantageous characteristics (in comparison to natu-ral floods) to examine the role of the suspended loads in entraining andtransporting associated contaminants. It has been shown that duringFFs suspended sediment concentration doubles that of natural floods,although discharges are typically much lower (Tena et al., 2012b). Inturn, flashiness, measured as the rate of discharge increment per unittime, may attain an order of magnitude higher during FFs than duringnatural events. Consequently, FFs exhibit higher transport capacitythan their natural counterparts despite their considerably lowermagni-tude and duration. Furthermore, and owing to their intrinsic character-istics (i.e. hydrograph design, programmability), they offer obviouspractical advantages, particularly in what monitoring logistics isconcerned.

FFs have been used typically to mitigate dam-induced impacts,mobilizing accumulated sediment and scouring the channel (Milhous,1982), and maintaining large morphological features (Reiser et al.,1989). Under appropriate conditions, they are increasingly used as atool for themaintenance and enhancement of aquatic and riparian hab-itat (Brookes, 1995; Kondolf and Wilcock, 1996; Downs et al., 2002). Inparticular, FFs have been implemented in the lower Ebro River since2002 and have been extensively monitored and subsequently analyzed,even modeled (Batalla and Vericat, 2009; Tena et al., 2011, 2012a,2012b). In the present work, we examine the dynamics of persistentorganicmicrocontaminants associated to the suspended sediments dur-ing a particular FF that was performed in the lower River Ebro (NESpain) in June 2012. Measurements were taken 12 km downstream ofa well-known contaminated hot-spot (Amaral et al., 1996; Olivareset al., 2010; Soto et al., 2011) associated to a nearby chloro-alkali indus-try (Flix) (Fig. 1).

A total of 42 semi-volatile persistent organic compounds corre-sponding to the families of polycyclic aromatic hydrocarbons (PAHs),

DDT and its metabolites (DDX), polychlorinated biphenyls (PCBs) andother organochlorine compounds (OCs)weremonitored. The objectivesof the paper are (a) to analyze the presence of persistent organic pollut-ants associated to the suspended sediment load; (b) to characterizetheir variation during one of the artificial releases (flushing-flow) thatare regularly performed in this river; (c) to provide a quantitativeassessment of the mass-flows and load budgets of the organicmicrocontaminants transported during the FF; and (d) to assess therisk associated to such transport, also by putting the results in the cur-rent regulatory context.

2. Site description: The lower Ebro

The Ebro is the largest river in the Iberian Peninsula flowing into theMediterranean Sea, with a basin draining a total of 85,534 km2 (Fig. 1).It is characterized by an interannual variability associated with its in-trinsic Mediterranean character. Mean discharge recorded in Tortosa(i.e. the lowermost downstream gauging station) for the period1912–2012 is 436 m3 s−1, but flows vary from less than 50 m3 s−1 inthe very dry seasons to more than 12,000 m3 s−1 (i.e. the major floodrecorded ever that occurred in October 1907 Novoa, 1984).

Runoff in the Ebro is regulated by a series of dams. In particular, inthe lower Ebro (where we focus our study) the Mequinenza–Ribarroja–Flix dam complex is located; it owns a total storage capacityof ~1.7 km3 (i.e. 1 km3 = 1 × 109 m3), the largest in the catchment.Regulation has led to a decrease of an average of 25% the magnitudeof natural frequent floods (i.e. Q2–Q25, where Qi is the discharge associ-ated with i years recurrence interval) in the reach downstream thedams (Batalla et al., 2004). Sediment transfer has also been altered;for instance, Vericat and Batalla (2006) reported a mean trapping effi-ciency for suspended sediment at around 90% in the dam complex,whereas bedload is fully captured. Together with hydrological and sed-imentological alterations because of regulation (Batalla et al., 2004;Vericat and Batalla, 2006; Tena et al., 2012a) ecological consequenceshave also been observed in the lowermost part of the catchment(Prats et al., 2011; Sabater et al., 2011); for instance, themassive growthof macrophytes (Batalla and Vericat, 2009; Tena et al., 2012a). Withinthis context, FFs have been designed and implemented in the lowerEbro since 2002 with a twofold objective: (i) controlling macrophytepopulations and (ii) maintaining certain degree of sedimentary activityin the channel. However, river regulation is not the only impact in thelower Ebro. The Flix Reservoir is heavily affected by wastes from achloro-alkali plant. The result of the accumulation of waste disposalfrom the factory through decades is a deposit of hazardous industrialsolid waste (200,000–360,000 T) which contains large amounts ofPCBs, DDT, hexachlorocyclohexane (HCB) and heavy metals (Hg)(Fernández et al., 1999). Today, this deposit is a major concern and de-contamination works are being carried out. This decontamination in-volves the removal of the contaminated sludge of the river bydredging, a subsequent processing at a nearby treatment plant, trans-portation and disposal in landfill of contaminatedwastes. Consequently,the present study can be used as a tool for the evaluation of the effects ofthis dredging process on the Ebro River downstream from Flix.

Fig. 1. The Ebro River basin with the location of the study area.

383S. Quesada et al. / Science of the Total Environment 473–474 (2014) 381–390

In order to examine the dynamics of the microcontaminants of con-cern during the June 2012 FF downstream Flix dam, two sampling sec-tions were established: Ascó Monitoring Section (AMS), located 12 kmdownstream the damand Pas de l'AseMonitoring Section (PAMS), locat-ed 4 km downstream AMS and 16 km downstream the Flix dam (Fig. 1).

3. Methods

3.1. Flow monitoring

Flow and sediment transport were monitored at the two sectionspreviously selected (AMS and PAMS) during the FF performed in June2012. Data on water release from the dam complex was provided bythe dam operator (Endesa Generación SA); moreover, discharge is rou-tinelymeasured by EbroWater Authorities (ConfederaciónHidrográficadel Ebro, CHE) at AMS (Fig. 1) (CHE, 2012; SAIH, 2012). There, waterstage (h) is recorded every 15 min by an OTT Water Level and trans-formed into discharge (Q) using an ad-hoc calibrated h-Q rating curve.In order to corroborate the official CHE data calibration, discharge mea-surements were done by means of acoustic methods (i.e. ADCP RiverSurveyor M9®, Sontek) during the flood. The hydrograph at PAMS (lo-cated 4 km downstream from AMS) has been directly routed fromAMS by means of the Muskingum method (Tena et al., 2011, 2012a).

3.2. Sampling of sediment and chemical loads

Suspended sediment concentrations (SSC¸ inmg L−1) have been ob-tained from (i) direct sampling in AMS and (ii) indirect data (i.e. turbid-ity measurements) in PAMS: (i) Suspended sediments were sampled

during the event by means of a cable-suspended bucket sampler at30 min interval from the Ascó Bridge (200 meter upstreamAMS). Sam-pling was carried out in a single vertical, in the midpoint of the trans-verse section of the river, according to (Vericat and Batalla, 2006). Theresulting water samples were lately vacuum filtered at the laboratoryby means of glass microfiber filters (Filter-Lab, 1.2 μm pore size),dried and weighed to determine SSC. The method reported by Tenaet al. (2011) was followed to determine the organic matter content;then, the percentage of organic matter was subtracted from the filterweight to obtain the net particulate suspended sediment load.

(ii) Suspended sediment concentrations were also obtained bymeans of an indirect method: water turbidity, a proxy of suspendedsediment, and was recorded every 15 min at the water quality stationof PAMS by means of a Hach SS6 turbidity probe. Subsequently, turbid-ity series were calibrated with a total of 150 water samples already ob-tained during floods, and weekly or fortnightly during low flows (Tenaand Batalla, 2013). A correlation between pair values of turbidity (inNTU; i.e. Nephelometric Turbidity Units) and SSC (mg L−1) wasestablished for the PAMS turbidity probe. The relation fitted to a linearregression (SSC = 0.83 × NTU − 1.3) was then used to obtain the cor-responding SSC values (Tena et al., 2011; Tena and Batalla, 2013) (seeFig. S1, Supporting material). The suspended sediment mass-flowtransported during the FF was calculated by multiplying the hourlySSC (in mg L−1) by Q (in m3 s−1).

In turn, water samples, specifically taken for the chemical analysis ofthe suspended sediments were also sampled during the flash-floodevent from AMS using the methodology explained in (i). Each collectedwater samplewas transferred to a 10 L polypropylene flask, labeled andstored in mobile refrigerators to be properly transported to the


laboratory. At the endof the sampling campaign a total of 13water sam-ples were collected intensively from 9:45 to 15:30 with frequencies be-tween 15 and 30 min.

3.3. Chemical analysis

3.3.1. Chemicals4,4-DDE d8, 4,4′-DDT d8, d4, β-HCH, δ-HCH, dicofol, dicofol d8, γ-

HCH d6, hexaclorobenzene 13C, trifluralin d14, Pesticide-Mix 208,(100 μg mL−1), PAHs-Mix 63, PAH-Mix 9 (deuterated) and PCB-Mix19 were purchased from Dr Ehrenstorfer (Augsburg, Germany). La-beled PCBs (#31, #52, #118, #153 and #194) were purchased fromWellington Laboratories (Ontario, Canada).

3.3.2. Sample extractionThe first step in the laboratory was the filtration of the water sam-

ples with 0.7 μm pore size glass fiber filters on a vacuum filter systemfor the extraction of the Suspended Sediments (SS). After filtration,each water sample resulted in 8 filters that were dried over silica gelin a desiccator until constant weight. The filters that constitute eachsample were spiked with 15 μL of the surrogate solution at 1 ng g−1

and extracted by pressurized liquid extraction (PLE) using a DionexASE 350 accelerated solvent extractor (ASE) (Sunnyvale, CA, USA). The22 mL ASE stainless steel cells were packed as follows: 2 g of Florisilpowder was placed at the outflow side of the cell and 4 g more wasadded to the filter package. The remaining space was filled withHydromatrix (Merck, Darmstadt, Germany) and before closing the cella cellulose filter was placed at the bottom. A mixture of hexane:dichlo-romethane (1:1) was used as extraction solvent.

The simultaneous extraction and clean-up of the 42 analyzed com-pounds were briefly optimized according to previous studies that ana-lyzed the same compounds in sediments and soils samples(Hildebrandt et al., 2009; Navarro-Ortega et al., 2010). Extracts wereevaporated at room temperature to nearly dryness using a TurboVapLV from Caliper LifeSciences (Hopkinton, MA, USA) and reconstitutedwith hexane to a final volume of 100 μL into glass amber vials for gaschromatography.

3.3.3. Instrumental analysisFor PAHs, GC–MS analyses were carried out using a Trace 2000 gas

chromatograph (Thermo Electron, USA) coupled to a mass spectrome-ter from Thermo Electron. In order to increase the sensibility and selec-tivity of organochlorine pollutants (DDX, PCBs and OCs), the analysiswas performed using GC–MS/MS adapting the method described inSanchís et al. (2013). A Trace GC-Ultra coupled to a triple quadrupole(QqQ) mass spectrometer TSQ Quantum (ThermoFischer Scientific)was employed. In both cases chromatographic separation was per-formed on a fused silica capillary column of 30 m × 0.25 mm I.D., and0.25 μm film thickness, using helium as carrier gas (HP-5MS (J&W Sci-entific, Folsom, CA, USA) for PAHs and DB-5 MS (Agilent Technologies,CA, USA) for organochlorine pollutants). The mass spectrometer wasoperated in the electron ionization mode with an ionizing energy of70 eV and the extracts were injected in the splitless mode. Oven tem-peratureswere programmed followingNavarro-Ortega et al. (2010). Al-though the presented multiresidue method is performed with only oneextraction, the use of isotopic-labeled surrogates (16 labeled PAHs, 2 la-beled DDX, 6 labeled PCBs and 4 labeled OCs) for the quantification ofcompounds and the control of method blanks guarantees a good qualityof the analytical results. Data acquisition was carried out in the selectedion monitoring (SIM) mode. Each compound was separately quantifiedusing a five-point calibration of mixed standard solutions in the rangefrom 50 to 1000 mg L−1. Method detection limits ranged respectivelyfrom 0.18 to 0.54 ng g−1 for PAHs, 0.002 to 0.025 ng g−1 for DDX,0.001 to 0.0261 ng g−1 for PCBs and 0.001 to 0.005 ng g−1 for OCs(see Table S1 in Supporting Material).

3.4. Computation of pollutants' mass-flows and loads

Mass-flowmij (mg s−1) for every single pollutant j at time ti is calcu-lated according to Eq. (1):

mij ¼ Qi � SSCi � cij � 10−6 ð1Þ

whereQi is the river discharge inm3s−1, SSCi is the suspended sedimentinmg L−1, cij is the concentration of pollutant j in ng g−1 and 10−6 is theappropriate unit conversion factor.

Total (cumulated) load of a given pollutant along the FF correspondsto the area under the mass-flow vs time curve (i.e., the integral curve).Thus, load of pollutant j corresponding to time interval comprised be-tween ti and ti + 1 can be approximated by Eq. (2):

Lij ¼tiþ1−ti� � � mij þm iþ1ð Þ j

� �

2� 10−3 ð2Þ

where the load Lij is expressed in g, the mass-flows mij and m(i + 1)j inmg s−1, the time interval in s and 10−3 is the corresponding conversionunit factor.

The total cumulated load of pollutant j along the flushing-flow eventLj (in g) is obtained by summation over all the time intervals:

L j ¼Xn

i¼0

Lij ð3Þ

4. Results and discussion

4.1. Flood hydrology and sediment load

Here, we summarize the general patterns of the hydrology and thesuspended sediment load of the FFs in the lower Ebro, emphasizingthe characteristics of the one performed in June 2012. The event moni-tored did not exactly resembled the habitual hydrograph implementedduring the last years, but followed a new design elaborated to increasethe effectiveness of the artificial releases onmacrophyte removal, whileminimizing the geomorphic effects of the clearwater releases. Peakflows registered at AMS during the event (i.e. 1284 m3s−1 and1323 m3s−1; Fig. 2) were remarkably higher that those performed fol-lowing the ancient FF design (i.e. 1185 m3s−1 and 1135 m3s−1). Thepeak discharge of the FF had a recurrence period of 1.6 year, estimatedfor the post-dam flow regime (1965–2012). Mean discharge during theevent was 753 m3 s−1 and ca. 40 hm3 of water were released duringthe FF, a value that is in the same range than other studied events. How-ever, the energy expenditure in the channel, expressed as the rate ofdischarge increment per unit time (Flashiness Index, FI, as per Batallaand Vericat, 2009) was higher in this event. Mean flood FIs obtainedduring FFs in the period 2002–2011 was 149 m3s−1 h−1; while in thepresent release the FI was 230 m3s−1 h−1, a fact that highlights thehigher energy expenditure in the channel and thus the accordinglyhigher sediment transport. Suspended sediment concentration (SSC)and river flow discharge (Q) show a direct non-linear dependence(see Fig. S2, Supporting material) fitting a power type equationSSC = 0.012·Q1.35 (eq. 4), which is consistent with previous observa-tions (Batalla and Vericat, 2009).

Mean SSC at AMS during the FF was 155 mg L−1 ranging from aminimum of 40 mg L−1 to a maximum of 315 mg L−1 (Fig. 2). SSCswere higher than those observed in the same section during previousevents that typically did not exceed 220 mg L−1. Further downstream,SSC in PAMS was lower, ranging from a minimum of 4 mg L−1 to amaximum of 151 mg L−1 observed during the peak discharge. Valuesobserved at PAMSwere in the same order ofmagnitude than inpreviousFFs.

Fig. 2. Graphs showing river discharge (Q) and Suspended Sediment Concentration (SSC) profiles along time during the FF event at the locations AMS and PAMS respectively.


4.2. Occurrence of persistent organic pollutants

All the persistent organic pollutants analyzed in the present studyshow relatively high octanol–water partition constants (i.e., logKow ≥ 4), the only exceptions being lightest PAHs and some HCH iso-mers (log Kow 3.4 to 4). Therefore they are likely to be found in bothsteady sediments and particulate suspended material (it may be seenas a mobilized fraction of sediments) rather than in the aqueous phasein which they show a very low solubility. Only HCB andHCH are detect-able in water, though the dissolved fraction was excluded from thepresent study, which is exclusively focused on transport of pollutantsby particulate suspended material. Occurrence data of the various pol-lutants (mean, median, standard deviation, maximum, minimum, coef-ficient of variation and detection frequency) measured in suspendedsediments during the FF event studied are reported in Table 1.Concerning PAHs, the overall concentration is around 100 ng g−1

which falls within the range described in the Ebro river for steady sedi-ments (ca. 200 ng g−1) (Hildebrandt et al., 2009). These PAHs concen-trations are typical of rural environments rather than urban andindustrial areas (Liu et al., 2012). This also applies to single constituentsthat are also consistent with sediment levels within a factor up to 2.Main differences are observed only for fluoranthene, chrysene andbenzo(a)pyrene, the former two showing lower concentrations in partic-ulate material (respectively 10 ng g−1 and 7.5 ng g−1 vs. 120 ng g−1

and 31.4 ng g−1 in sediments) while the latter shows the oppositetrend (24 in particulate material and 6 ng g−1 in sediments). In fact,benzo(a)pyrene is the major constituent of the PAHs mixture (24% onaverage), followed by phenanthrene (19%) andfluoranthene (11%).Mix-ture profile is consistent with a pyrolytic (combustion) origin (indexesIP/276 and BaA/228 ca. 0.35 and 0.34 respectively) (Arias et al., 2010).Furthermore, predominance of less volatile compounds in the mixturewould be indicative of local pollution influence rather than long range at-mospheric transport (Fernández and Grimalt, 2003). Generally, PAHcompounds exhibited fairly constant concentrations during the FF,which are reflected in individual low to moderate standard devia-tions and coefficients of variation (CV) (20% to 64%). Total PAHsranged from 76 to 146 ng g−1 with a CV of 21.5%. These results indi-cate that there is a low connection between these compounds andthe studied FF. The concentrations of PAHs found in the particulateare likely due to diffuse pollution rather than to the mobilization ofthe contaminated sediments.

Even thoughDDT and PBCs are banned since long time and their for-mer production in Flix discontinued, the presence of contaminated sed-iments constitutes a known source for these compounds downstream inthe Ebro. As mentioned before, one of the specific objectives of thepresent study is to characterize their mobilization and transport pat-terns. DDT and its transformation compounds DDE and DDD are fully

detectable in particulate material, in good concordance with previouslydata reported in the same location for steady sediments. The averagevalues are comprised between 0.92 and 187 ng g−1, the latter corre-sponding to 4,4′-DDT. These values are slightly higher than those re-ported for steady sediments of the same river section in the last10 years: 11 to 63.5 ng g−1(Navarro et al., 2006), 4 to 30 ng g−1

(Lacorte et al., 2006) and 3 to 48 ng g−1(Navarro-Ortega et al., 2010)indicating as a probable source the polluted sediments located in Flix in-dustrial complex. Peak values of 4,4′-DDT up to 1400 ng g−1 are not-withstanding, being the main contributor to this family of compounds(75% on average). Such a high value of the parent compound point to re-cent illegal releases in concordance with the existing literature (Lacorteet al., 2006; Gómez-Gutiérrez et al., 2007; Bosch et al., 2009).

Measured concentration ranges for PCBs in particulate suspendedmaterial (13 to 1445 ng g−1) are fairly coincident with those alreadydescribed for steady sediments in this reach of the Ebro river(Fernández et al., 1999), being again their most likely source the pollut-ed pack of sediments located in Flix industrial complex, in which theywere formerly manufactured. Major PCB constituents are the less vola-tile congeners 170, 180 and 194with average levels and contributions of88 ng g−1 (26%), 157 ng g−1 (46%) and 61 ng g−1 (18%) respectively.

Other OCs are largely dominated by HCB (95% contribution on aver-age), a known by-product generated in the chloro-alkali industry locatednearby upstream (Flix) and in less extent to HCH, being among them iso-mersβ andγ (lindane) themost relevant. δ-HCH isomer aswell as aldrinare detected in trace amounts while α HCH, and dicofol are below thelimit of detection. Range levels for HCB are in agreement with those de-scribed in the literature for sediments (Navarro-Ortega et al., 2010),being the maximum somewhat higher, which is consistent with theproximity of the Flix hot-spot.

In sharp contrast with PAHs, range levels of all organochlorine pol-lutants (OCs, DDX and PCBs) detected in suspended particulatematerialshow a high variation along the period observed (CV% usually higherthan 100%). That fact can be interpreted in terms of the characteristicsof the pollution sources involved, as deeply explained in Section 4.3.While PAHs show a diffuse source not influenced by the FF, the rest ofcompounds present in the contaminated sediments of the Flix damhave been mobilized downstream during the FF together with thesediments.

4.3. Mass-flows and load budgets

Floods, including both natural and FFs, are the main responsible ofsediment transport in rivers. In the case of rivers regulated by damsthe total load transported during natural floods is usually much higherthan during artificial floods, mainly due to their larger peaks and dura-tion (Batalla and Vericat, 2009). However, in dry years, and in the

image of Fig.�2

Table 1Mean, median, standard deviation, maximum, minimum concentrations (ng g−1), coefficient of variation and detection frequency of pollutants measured in suspended solids during theFF and reference regulatory values (see text).

Mean Median Std. Dev. Max Min C.V. % D.F. % ISQC TEC PEC RCRA

Compounds (ng g−1) (ng g−1) (ng g−1) (ng g−1) (ng g−1) (ng g−1) (ng g−1) (ng g−1) (ng g−1)

PAHsNaphthalene bld bld bld bld bld – 0 34.6 176 561 –

Acenaphthylene bld bld bld bld bld – 0 5.87 5.9 128 –

Acenaphthene 2.774 2.684 1.287 5.627 1.247 46.4 100 6.71 6.71 89 –

Fluorene 4.951 4.161 2.490 10.835 2.431 50.3 92.3 21.2 77.4 536 –

Phenanthrene 18.878 19.024 5.148 26.845 12.153 27.3 100.0 41.9 204 1170 –

Anthracene 0.526 0.471 0.336 1.129 0.042 63.9 92.3 46.9 57.2 845 –

Fluoranthene 10.784 10.040 3.061 15.986 6.175 28.4 100 111 423 2230 –

Pyrene 9.610 9.081 2.525 14.546 6.178 26.3 100 53 195 1520 –

Benzo(a)anthracene 3.731 3.691 0.933 5.449 1.981 25.0 100 31.7 108 1050 –

Crysene 7.478 7.254 1.686 10.600 4.691 22.5 92.3 57.1 166 1290 –

Benzo(b)Fluoranthene 6.862 7.003 1.528 9.231 4.612 22.3 100 – 240 13,400 –

Benzo(k)fluoranthene 4.946 5.001 0.977 6.215 3.817 19.8 69.2 – 240 13,400 –

Benzo(a)pyrene 24.159 21.986 8.714 43.844 13.713 36.1 100 31.9 150 1450 –

Indenoe(1.2.3-c.d)pyrene 3.672 3.920 0.949 4.472 2.624 25.8 23.1 – 200 3200 –

Dibenzo(a.h)anthracene bld bld bld bld bld – 0 6.22 33 135 –

Benzo(g.h.i)perylene 7.373 7.118 1.496 9.531 5.217 20.3 92.3 – 170 3200 –

∑ PAHs 99.8 97.1 21.5 146.4 75.6 21.5 100

OrganochlorinesHCB 64.661 3.363 129.788 435.958 0.817 200.7 100 – – – 20α-HCH bld bld bld bld bld – 0 – 6 100 –

β-HCH 1.114 1.327 0.643 2.182 0.178 57.7 100 – 5 210 –

γ-HCH (Lindane) 1.794 1.489 1.141 5.203 0.937 63.6 100 0.94 2.37 4.99 –

δ-HCH 0.058 0.083 0.063 0.218 0.024 109.0 58.3 – – – –

Aldrin 0.059 0.063 0.065 0.222 bld 110.0 66.7 – 2 80 –

Dicofol bld bld bld bld bld – – – – – –

∑ OC 67.7 6.5 130.0 440.1 3.7 192.0 100

DDX2.4-DDE 12.292 6.826 16.035 60.975 3.179 141.3 100 Σ = 1.42a Σ = 3.2a Σ = 31.3a –

4.4′-DDD 12.032 1.617 26.307 95.674 0.533 236.9 100 Σ = 3.54b Σ = 4.9b Σ = 28.0b –

2.4′-DDT 0.918 0.347 1.344 4.885 bld 158.6 91.7 Σ = 1.19c Σ = 4.2c Σ = 62.9c –

2.4′-DDD 8.039 4.158 9.661 33.169 1.611 120.2 100 –b –b –b –

4.4′-DDE 17.830 7.487 22.691 68.890 2.405 127.3 100 –a –a –a –

4.4′-DDT 187.219 30.142 418.910 1467.047 5.810 223.8 100 –c –c –c –

∑ DDX 238.3 48.8 472.5 1651.7 19.3 198.3 100 – 5 572 –

PCBPCB 18 0.513 0.246 0.744 2.703 0.087 145.0 100 – – – –

PCB28(3) + PCB31 0.203 0.172 0.115 0.466 0.091 56.7 100 – – – –

PCB44 0.717 0.401 0.682 2.484 0.102 95.2 100 – – – –

PCB52 1.102 0.668 1.227 3.729 0.035 111.3 100 – – – –

PCB 101 16.671 8.393 18.289 56.503 4.734 109.7 100 – – – –

PCB 118 2.012 1.988 1.698 5.140 0.600 84.4 77.8 – – – –

PCB 138 2.611 1.679 2.981 11.320 0.483 114.2 100 – – – –

PCB 153 6.884 4.413 5.664 19.065 1.926 82.3 100 – – – –

PCB 170 87.858 2.407 151.006 415.638 0.733 171.9 66.7 – – – –

PCB 180 157.079 8.310 262.031 687.324 1.273 166.8 100 – – – –

PCB 194 60.872 7.149 93.898 233.049 1.231 154.3 100 – – – –

PCB 149 5.571 3.569 4.569 15.815 1.721 82.0 100 – – – –

∑ PCB 342.1 38.9 538.5 1445.6 13.1 157.4 100 34.1 59.8 676.0 –

bld: below limit of detection.a Sum of 2.4′-DDE and 4.4′-DDE.b Sum of 2.4′-DDD and 4.4′-DDD.c Sum of 2.4′-DDT and 4.4′-DDT.


absence of natural floods, the contribution of FFs to the annual loadmaybecome remarkable. The sediment load estimated during the June 2012FF in the River Ebro varies between the two monitoring sections. Thesuspended sediment load estimated at AMS by means of direct sam-pling attained 5766 T (i.e. from 9:00 am to 6:00 pm) while at PAMSwas at around 3010 T for the same period of time. This last value fitswith the mean value observed for all the other previous FFs (mean of3040 T, with a standard deviation of 1512 T); in contrast, the estimatedload at AMS was the largest ever registered.

Sediment dynamics in regulated rivers, thus sediment supply-limitedsystems, are rather complex and not only function of the hydrology. Inthe lower Ebro several parameters such as antecedent conditions, sedi-ment availability exhaustion, flashiness, loss of energy due to flow

routing and overflooding have been discussed to play an importantrole in sediment dynamics (Tena et al., 2012a). Within this context, thedifference observed in the sediment load of the studied sections couldbe attributed to any of the previous parameters or a combination ofthem. In this case, even there is a short distance between the twosections (e.g. 4 km), sharp changes in channel width or depth, watervelocity, as well as the distance from the Flix Dam influences directlythe sediment dynamics. For instance, sediment resuspension due toenergy expenditure in the channel is higher in the upstream section(i.e. AMS); while such resuspension could turn into deposition inthe downstream section due to the energy loss, thus increasing the sed-iment storage in this reach and reducing the sediment transported inPAMS.


Concerning the contaminants, mass-flows (mg s−1), as well as totalloads (i.e., cumulated loads expressed in g) transported by suspendedsediments along the FF time period for the different specific compoundsand families analyzed were calculated as described in Section 3.4. Re-sults are summarized in Table S2 (Supportingmaterial). Observed aver-age mass-flow of PAHs along the flushing-flow event was about20 mg s−1, with a maximum of 34 mg s−1 coincident with that ofSSC. Such a low variation in PAHsmass flows (CV = 42%) indicates ho-mogeneity of the pollution source, which is consistent with a local dif-fuse pollution related to combustion processes. Owing to this, totalPAHs mass-flow (mΣPAH) shows a more linear correlation withsuspended solids (expressed either as SSC or turbidity) than with dis-charge Q. While both fit a power equation Y = A · Xα, the exponentsα characterizing the dependence of PAH mass-flow (Y = mΣPAH) ondischarge (X = Q) and suspended solids (X = NTU, SSC) are 2.6 and1.65 respectively, the latter clearly closer to unity. Similar to what is ob-served in the concentration pattern, the minimum and the maximummass-flow during the FF for the rest of compounds differ significantly,especially for HCB, 4,4′-DDT and the heavier PCBs indicating high de-pendence with the FF.

Fig. 3a to d shows the evolution of mass-flows (mg s−1) along time,as well as the cumulated load (i.e., their integral curve, g) for the fourfamilies of compounds. The observed profile is consistent with the re-sults previously shown. PAHs show a variable mass-flow over the peri-od of time evaluatedwith no clear tendency and no correlation betweenthe maximum mass-flow and the maximum Q (Fig. 3a). Consequently,total load grows steadily along time, reaching at the end of themonitor-ing period a value of ca. 403 g.

Remarkably, organochlorine pollutants (DDX, PCBs and OCs) show atotally different pattern behavior, characterized by sharp peaks relatively

Fig. 3. Graphs showing mass-flows, actual and basal cumulated loads transported by the rive(c) PCBs and (d) OCs.

early on time (at ca. 1 h 15 min) followed by a rapid subsequent decay.Thus, for instance, OCs have amaximummass-flow of 66.1 mg s−1 withan average of 8.2 mg s−1; DDX247.9 and 38.2 mg s−1 and PCBs 217 and44.8 mg s−1, respectively. Such a high ranges of variability are wellreflected on CVs of 150–200%. Consequently, cumulative load curvesgrow up very quickly (Fig. 3b–d).

For comparative purposes, we also provide an estimation of whatcan be called basal loads (Fig. 3a to d), calculated using the averageriver discharge recorded in the period 1912–2012 (436 m3 s−1), a cor-responding basal SSC value of 36.6 mg L−1 obtained using Eq. (4)(Section 4.1.), the same time interval of the period surveyed(19,800 s) and the average concentration level of each familymeasuredin this study (Table 1). Basal values found for the four families are rep-resented in Fig. 3a to d. Cumulated basal loads for the various familiesare respectively PAHs, 31 g; OCs, 21.5 g; DDX, 75.3 g and PCBs,108.1 g. Comparing these load values with the actual ones measuredduring the event, i.e., PAHs, 403 g; OCs, 207.5 g; DDX, 940.5 g andPCBs, 1038.9 g; this indicates that even though the peak dischargereached during the FF merely exceeds the basal flow by a factor of 3(436 vs. 1300 m3 s−1) the transport of contaminants is increased by afactor up to 10 to 13.

Afinalwarning remark about load budgets as presented in this studyis necessary: it must be kept in mind that the given figures are conser-vative since they only take into account the fraction of pollutantsadsorbed on suspended sediment without accounting for the dissolvedfraction.Whereas for the majority of pollutants this fraction is expectedto be negligible due to the very low water solubility of the concernedcompounds, this is not the case of few of them, such as HCH isomers,HCB and the lightest PAHs (naphthalene, acenaphthene, acenaphthyl-ene, fluorene, phenanthrene) whose solubility in water is perceptible.

r along time during the FF event for the different pollutants' families (a) PAHs; (b) DDX;

image of Fig.�3


4.4. Regulatory assessment

The concentrations obtained were also compared with some legis-lated limits in sediments assuming that the majority of the SS obtainedcome from the resuspension of sediments due to increasing energy dur-ing the flash-flow event and that pollutants in SS coming from air pollu-tion is in comparison insignificant in this case. In this way, threedifferent guidelines that specially address sediment managementwere considered: The Interim freshwater Sediment Quality Guidelines(ISQC) (Canadian Council of Ministers of the Environment, 2002); theThreshold Effect Concentration (TEC; below which adverse effect arenot expected to occur) and the Probable Effect Concentration (PEC,above which adverse effects are expected to occur more often thannot), both from MacDonald et al. (2000); and the guideline defined bythe Resource Conservation and Recovery Act (RCRA, USEPA, 1976).

As far as EU legislation is concerned, Directive 2013/39/UE(European Council, 2013) (an extension of theWater Framework Direc-tive) explicitly sets annual average andmaximum allowable concentra-tion Environmental Quality Standards (respectively referred to as AA-EQS and MAC-EQS) in surface waters, while EQS for sediments andbiota are generally left to the provisions of member states. Thus,recalculating particulate concentration (ng g−1) values to concentra-tions in water (ng L−1) obtained in the present study we found thatHCB exceeds the MAC-EQS (59.9 vs 50 ng L−1) while DDX (sum ofDDT, DDE and DDD) and 4,4′DDT exceed the AA-EQSs set by the

Fig. 4.Measuredmean andmaximum concentrations of different compounds analyzed on partitified with an inverted black triangle (▼) indicate noncompliance with any reference regulator

above referred Directive (respectively 33.6 and 26.1 ng L−1 mean con-centrations in water compared to AA-EQSs of 25 and 10 ng L−1).

Among the PAH compounds analyzed, benzo(a)pyrene presented amaximum concentration of 43.8 ng g−1 (Table 1 and Fig. 4a), which isslightly above the most restrictive limit (ISQC, 31.7 ng g−1). Special at-tention has to be paid to this compound considered as one of the mostcarcinogenic among PAHs. In the case of organochlorine pollutants animportant contribution of the DDX (DDT, DDE and DDD) is observed(Table 1 and Fig. 4b), especially 4,4′-DDT, although this compound isforbidden in the EU (and thus in Spain) for agricultural purposes since1977. Samples showed DDX maximum concentration (1651 ng g−1)above the PEC limits (572 ng g−1) and mean concentrations(238.3 ng g−1) between the TEC and the PEC limit. The minimumDDX concentration found during the FF, 19.3 ng g−1, also exceeds theTEC, resulting in a high potential risk for the living organisms in theriver ecosystem in steady conditions which is increased dramaticallyduring FF. Similarly, the sum of all PCBs analyzed, which maximumsum concentration (1445.6 ng g−1) exceeds by far the PEC limit(676 ng g−1) and with a mean concentration between the TEC(59.8 ng g−1) and the PEC limits. Both the maximum and the meanconcentration of PCBs also exceed the ISQC established at 34.1 ng g−1.In this case the minimum PCB concentration found is underneath thethree thresholds considered and consequently the potential risk forthe living organisms is a direct consequence of the FF and its remobiliza-tion of the contaminated sediments accumulated in the Flix dam.

culatematerial compared to some reference and regulatory values (see text). Values iden-y levels. (a) PAHs; (b) DDX and PCBs; (c) OCs.

image of Fig.�4


Other compounds detected at concentrations exceeding regulatoryreferences are (Table 1 and Fig. 4c): HCB, with maximum concentration(435,96 ng g−1) and mean concentration (64.7 ng g−1) far above theRCRA limits (20 ng g−1) and γ-HCH, with maximum concentration(5.2 ng g−1) above the PEC limit (4.99 ng g−1) and mean concentra-tion (1.79 ng g−1) between the TEC (2.37 ng g−1) and the PEC limit.In both cases the minimum concentration found is underneath thethresholds indicating the same behavior than the previously explainedfor PCBs.

5. Conclusions

Programmed FFs allow studying both quantitatively and qualitative-ly how polluted sediments are transported under controlled conditions.The main conclusions of this research can be drawn as follows:

(i) The suspended sediment load estimated at the monitoring sec-tions was, on average, 4400 tones (for a total runoff of 40 hm3).This load is in the range of those observed in previous FFs (i.e.performed since 2002 in this river). Overall, suspended sedimentconcentrations were slightly higher than those measured inprevious releases, a fact that may be a response to the newhydrograph design that enhances energy expenditure in thechannel,

(ii) Concentration of some relevant pollutants found in thesuspended sediments, such as HCB, DDT and PCBs, largelyexceeded existing reference regulatory values established forsediments.

(iii) Evolution of concentrations, mass-flows and total cumulatedload budgets along time was estimated and compared to abasal or average scenario. Our estimations show that transportof pollutants mostly occurs under high flow conditions.

The remarkablemobilization of pollutants during flood flows shouldbe taken into consideration for the futuremanagement of contaminatedrivers, especially those located downstream from dams, fromwhich pe-riodical water releases are performed. This is particularly important inMediterranean Rivers, for which this hydrological behavior is consub-stantial to the precipitation regime.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.scitotenv.2013.11.040.

Acknowledgments

This work was supported by the Spanish Ministry of Economy andCompetitiveness through the Consolider-Ingenio 2010 project SCARCE(CSD2009-00065) and by the Generalitat de Catalunya (ConsolidatedResearch Group: Water and Soil Quality Unit 2009-SGR-965). The re-search leading to these results has received funding from theEuropean Comunities 7th Framework Programme under Grant Agree-ment No. 603629-ENV-2013-6.2.1-Globaqua.

References

Amaral OC, Otero R, Grimalt JO, Albaigès J. Volatile and semi-volatile organochlorinecompounds in tap and riverine waters in the area of influence of a chlorinated organ-ic solvent factory. Water Res 1996;30:1876–84.

Arias AH, Vazquez-Botello A, Tombesi N, Ponce-Vélez G, Freije H, Marcovecchio J.Presence, distribution, and origins of polycyclic aromatic hydrocarbons (PAHs)in sediments from Bahía Blanca estuary, Argentina. Environ Monit Assess2010;160:301–14.

Batalla RJ, Vericat D. Hydrological and sediment transport dynamics of flushing flows: im-plications for management in large Mediterranean rivers. River Res Appl 2009;25:297–314.

Batalla RJ, Sala M, Werritty A. Sediment budget focused in solid material transport in asubhumid Mediterranean drainage-basin. Z Geomorphol 1995;39:249–64.

Batalla RJ, Gómez CM, Kondolf GM. Reservoir-induced hydrological changes in the EbroRiver basin (NE Spain). J Hydrol 2004;290:117–36.

Bosch C, Olivares A, Faria M, Navas JM, del Olmo I, Grimalt JO, Piña B, Barata C. Identifica-tion of water soluble and particle bound compounds causing sublethal toxic effects. A

field study on sediments affected by a chlor-alkali industry. Aquat Toxicol 2009;94:16–27.

Brookes A. River channel restoration: theory and practice. Changing river channels.Chichester: Wiley & Sons; 1995.

Canadian Council of Ministers of the Environment. Canadian sediment quality guidelinesfor the protection of aquatic life. Winnipeg: Canadian Council of Ministers of theEnvironment; 2002.

Carrizo D, Grimalt JO. Rapid and simplified method for the analysis of polychlo-ronaphthalene congener distributions in environmental and human samples by gaschromatography coupled to negative ion chemical ionization mass spectrometry.J Chromatogr A 2006;1118:271–7.

CHE. Confederación Hidrográfica del Ebro. Retrieved July 2012, fromhttp://www.chebro.es, 2012.

Downs PW, Kondolf GM, Skinner KS. Rivers and streams. In: Davy A, Perrow M, editors.Handbook of ecological restoration. New York: Cambridge University Press; 2002.[Restoration in practice: 267–296].

European Council. Directive 2013/39/EC of the European Parliament and of the council of12 August 2013 amending directives 200/30/EC and 2008/105/EC as regards prioritysubstances in the field of water policy; 2013.

Fernández P, Grimalt JO. On the global distribution of persistent organic pollutants.Chimia 2003;57:514–21.

Fernández MA, Alonso C, González MJ, Hernández LM. Occurrence of organochlorineinsecticides, PCBs and PCB congeners in waters and sediments of the Ebro River(Spain). Chemosphere 1999;38:33–43.

Garcia C, Laronne JB, Sala M. Continuous monitoring of bedload flux in a mountaingravel-bed river. Geomorphology 2000;34:23–31.

Gómez-Gutiérrez AI, Jover E, Bodineau L, Albaigés J, Bayona JM. Organic contaminantloads into the Western Mediterranean Sea: estimate of Ebro River inputs.Chemosphere 2006;65:224–36.

Gómez-Gutiérrez A, Garnacho E, Bayona JM, Albaigés J. Assessment of the Mediterraneansediments contamination by persistent organic pollutants. Environ Pollut 2007;148:396–408.

Hildebrandt A, Lacorte S, Barceló D. Occurrence and fate of organochlorinated pesticidesand PAH in agricultural soils from the Ebro River Basin. Arch Environ Contam Toxicol2009;57:247–55.

Kondolf GM, Wilcock PR. The flushing flow problem: defining and evaluating objectives.Water Resour Res 1996;32:2589–99.

Lacorte S, Raldúa D, Martínez E, Navarro A, Diez S, Bayona JM, et al. Pilot survey of a broadrange of priority pollutants in sediment and fish from the Ebro river basin (NE Spain).Environ Pollut 2006;140:471–82.

Lehner B, Doll P, Alcamo J, Henrichs T, Kaspar F. Estimating the impact of global change onflood and drought risks in Europe: a continental, integrated analysis. Clim Change2006;75:273–99.

Liu Y, Beckingham B, Ruegner H, Li Z, Ma L, Schwientek M, et al. Comparison of sedimen-tary PAHs in the rivers of Ammer (Germany) and Liangtan (China): differencesbetween early- and newly-industrialized countries. Environ Sci Technol 2012;47:701–9.

MacDonald DD, Ingersoll CG, Berger TA. Development and evaluation of consensus-basedsediment quality guidelines for freshwater ecosystems. Arch Environ Contam Toxicol2000;39:20–31.

Milhous RT. Effect of sediment transport and flow regulation on the ecology of gravel-bedrivers. Gravel-bed rivers. Chichester: Wiley & Sons; 1982.

Navarro A, Tauler R, Lacorte S, Barceló D. Chemometrical investigation of the presenceand distribution of organochlorine and polyaromatic compounds in sediments ofthe Ebro River Basin. Anal Bioanal Chem 2006;385:1020–30.

Navarro-Ortega A, Tauler R, Lacorte S, Barceló D. Occurrence and transport of PAHs, pes-ticides and alkylphenols in sediment samples along the Ebro River Basin. J Hydrol2010;383:5–17.

Novoa M. Precipitaciones y avenidas extraordinarias en Catalunya. Proceedings ofthe Jornadas de Trabajo sobre Inestabilidad de laderas en el Pirineo, Barcelona;1984.

Olivares A, Quirós L, Pelayo S, Navarro A, Bosch C, Grimalt JO, et al. Integrated biologicaland chemical analysis of organochlorine compound pollution and of its biologicaleffects in a riverine system downstream the discharge point. Sci Total Environ2010;408:5592–9.

Prats J, Armengol J, Marcé R, Sánchez-Juny M, Dolz J. Dams and reservoirs in the lowerEbro River and its effects on the river thermal cycle. In: Barceló D, Petrovic M, editors.The Ebro River Basin. Springer Berlin Heidelberg; 2011. p. 77–95.

Reiser DW, Wesche TA, Estes C. Status of instream flow legislation and practices in NorthAmerica. Fisheries 1989;14:22–9.

Rügner H, Schwientek M, Beckingham B, Kuch B, Grathwohl P. Turbidity as a proxy fortotal suspended solids (TSS) and particle facilitated pollutant transport in catch-ments. Environ Earth Sci 2013;69:373–80.

Sabater S, Artigas J, Gaudes A, Muñoz I, Urrea G, Romaní AM. Long-term moderate nutri-ent inputs enhance autotrophy in a forested Mediterranean stream. Freshw Biol2011;56:1266–80.

SAIH. Ebro (automatic hydrological information system of the Ebro Basin). http://195.55.247.237/saihebro/index.php. [Visited 28th June 2012. 2012].

Sanchís J, Martínez E, Ginebreda A, Farré M, Barceló D. Occurrence of linear and cyclicvolatile methylsiloxanes in wastewater, surface water and sediments from Catalonia.Sci Total Environ 2013;443:530–8.

Schwientek M, Rügner H, Beckingham B, Kuch B, Grathwohl P. Integrated monitoring ofparticle associated transport of PAHs in contrasting catchments. Environl Pollution2013;172:155–62.

Soto DX, Roig R, Gacia E, Catalan J. Differential accumulation of mercury and other tracemetals in the food web components of a reservoir impacted by a chlor-alkali plant



http://refhub.elsevier.com/S0048-9697(13)01321-1/rf0005


























http://www.chebro.es

http://www.chebro.es
































































http://195.55.247.237/saihebro/index.php

http://195.55.247.237/saihebro/index.php










(Flix, Ebro River, Spain): Implications for biomonitoring. Environ Pollut 2011;159:1481–9.

Tena A, Batalla RJ. The sediment budget of a large river regulated by dams (the lowerRiver Ebro, NE Spain). J Soils Sediments 2013;13:966–80.

Tena A, Batalla RJ, Vericat D, López-Tarazón JA. Suspended sediment dynamics in a largeregulated river over a 10-year period (the lower Ebro, NE Iberian Peninsula).Geomorphology 2011;125:73–84.

Tena A, Batalla RJ, Vericat D. Reach-scale suspended sediment balance downstream fromdams in a large Mediterranean river. Hydrol Sci J 2012a;57:831–49.

Tena A, Ksiazek L, Vericat D, Batalla RJ. Assessing the geomorphic effects of aflushing flow in a large regulated river. River research and applications;2012b.

USEPA. Region 5, resource conservation and recovery act, ecological screening levels,2003; 1976.

Vericat D, Batalla RJ. Sediment transport in a large impounded river: the lower Ebro, NEIberian Peninsula. Geomorphology 2006;79:72–92.

Vericat D, Batalla RJ. Sediment transport from continuous monitoring in a perennialMediterranean stream. Catena 2010;82:77–86.



















Science of the Total Environment 508 (2015) 101–114

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv

Geomorphic status of regulated rivers in the Iberian Peninsula

G. Lobera a,⁎, P. Besné b, D. Vericat a,c,d, J.A. López-Tarazón a,e, A. Tena a, I. Aristi f, J.R. Díez g,A. Ibisate b, A. Larrañaga f, A. Elosegi f, R.J. Batalla a,c,h

a Fluvial Dynamics Research Group — RIUS, University of Lleida, Lleida, E-25198 Catalonia, Spainb Department of Geography, Prehistory and Archaeology, University of the Basque Country, 01006 Vitoria-Gasteiz, Spainc Forest Sciences Centre of Catalonia, Solsona, E-25280 Catalonia, Spaind Institute of Geography and Earth Sciences, Aberystwyth University, Ceredigion SY23 3DB, Wales, UKe School of Natural Sciences and Psychology, Liverpool John Moores University, L3 3AF Liverpool, UKf Faculty of Science and Technology, University of the Basque Country, 48080 Bilbao, Spaing Faculty of Education, University of the Basque Country, 01006 Vitoria-Gasteiz, Spainh Catalan Institute for Water Research, Girona, E-17003 Catalonia, Spain

H I G H L I G H T S

• Hydrogeomorphic alterations of 74 river sites in four large basins are analysed.• An index to assess the channel’s Geomorphic Status (GS) has been developed.• Flow regulation causes a reduction of the flow competence and loss of river’s dynamism.• Formerly active bars disappear as they are encroached by vegetation.• Channel stabilization limits river’s activity and degrades the fluvial ecosystem.

⁎ Corresponding author.E-mail address: [email protected] (G. Lobera).

http://dx.doi.org/10.1016/j.scitotenv.2014.10.0580048-9697/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t
a r t i c l e i n f o
Article history:Received 23 July 2014Received in revised form 17 October 2014Accepted 17 October 2014Available online xxxx

Editor: D. Barcelo

Keywords:Geomorphic StatusFlow regulationDamsIberian rivers

River regulation by dams modifies flow regimes, interrupts the transfer of sediment through channel networks,and alters downstream bed dynamics, altogether affecting channel form and processes. So far, most studies onthe geomorphic impacts of dams are restricted to single rivers, or even single river stretches. In this paper weanalyse the geomorphic status of 74 river sites distributed across four large basins in the Iberian Peninsula (i.e.47 sites located downstream of dams). For this purpose, we combine field data with hydrological data availablefromwater agencies, and analyse historical (1970) and current aerial photographs. In particular, we have developeda Geomorphic Status (GS) index that allows us to assess the physical structure of a given channel reach and itschange through time. The GS encompasses a determination of changes in sedimentary units, sediment availability,bar stability and channel flow capacity. Sites are statistically grouped in four clusters based on contrasted physicaland climate characteristics. Results emphasise that regulation changes river's flow regime with a generalizedreduction of the magnitude and frequency of floods (thus flow competence). This, in addition to the decreasedownstream sediment supply, results in the loss of active bars as they are encroached by vegetation, to the pointthat only reaches with little or no regulation maintain exposed sedimentary deposits. The GS of regulated riverreaches is negatively correlated with magnitude of the impoundment (regulation). Heavily impacted reachespresent channel stabilization and, in contrast to the hydrological response, the distance and number of tributariesdo not reverse the geomorphic impact of the dams. Stabilization limits river dynamics and may contribute to theenvironmental degradation of the fluvial ecosystem. Overall, results describe the degree of geomorphologicalalteration experienced by representative Iberian rivers mostly because of regulation, challenging the successfullong-term implementation of river basin management programmes.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

The form of an alluvial channel results from the interaction betweenwater flow, sediment flux and basin landscape features, and it may

change over time as a result of the continuous interplay betweennatural and human factors (e.g. Church, 2002; Rinaldi et al., 2013).Although channel forms such as gravel bars may appear stable, thegrains of which it is composed are replaced regularly (i.e. annualflood) by new sediment from upstream. These fluvial processes aremaintained in quasi-equilibrium through a spatial and temporaldynamic balance between water and sediment transport that

http://crossmark.crossref.org/dialog/?doi=10.1016/j.scitotenv.2014.10.058&domain=pdf




http://www.sciencedirect.com/science/journal/00489697

www.elsevier.com/locate/scitotenv

102 G. Lobera et al. / Science of the Total Environment 508 (2015) 101–114

promotes the creation andmaintenance of habitats, and ensures eco-systems' integrity (Petts, 2000).

Rivers in Mediterranean regions are particularly affected by humanimpacts which modify hydrology, sediment fluxes, and channel formsat different scales. Dams stand out among these impacts, as they alterflow regimes, interrupt the sediment transfer, and subsequently changedownstream erosion and deposition patterns (e.g. Brandt, 2000; Vericatand Batalla, 2006). Overall, the impacts of dams in dryland rivers tend tobe more pronounced than those in more humid regions, since channelform and river ecology are adapted to highly variable flows (Gasithand Resh, 1999; Batalla et al., 2004). Following impoundment, theriver undergoes a complex adjustment in channel shape that involveschanges in width, depth, bed level, slope, bed material (i.e. grain size),bedforms and planform configuration (Brandt, 2000). As geomorphologyis fundamental for river's ecosystem functioning, adjustments in physicalhabitat typically bring important effects for both the biota and ecosystemprocesses (Elosegi et al., 2010). Responses vary according to channelsubstrate and shape, flow regime, valley gradient, and the geophysicalhistory (Merritt and Cooper, 2000). In particular, dams reduce highflows (hence flow competence and the associated physical disturbance),and often increase base flows (Batalla et al., 2004); altogether makingthese fragile environments more suitable for exotic species not adaptedto marked hydrological cycles (Kondolf, 1997).

Owing to the highly variable and complex responses of river systems,most studies have been undertaken to simply examine the dam impacton a specific fluvial site (e.g. Surian, 1999; Gilvear, 2004; Vericat et al.,2006; Wyzga et al., 2012), whereas few studies cover broader spatialand temporal scales (e.g. Graf, 2006; Fitzhugh and Vogel, 2011; Batallaand Vericat, 2011). In parallel, indices have been developed to evaluatethe physical river quality, mostly for restoration purposes (e.g. Ravenet al., 1997; LAWA, 2000; Ollero et al., 2008; Rinaldi et al., 2013), all of

Fig. 1. The Iberian Peninsulawith location of the study sites and reservoirs in eachbasin. The inseare represented by symbols that indicate the physiographic and climatic cluster to which they

them involving field work, and nearly all ignoring the temporalevolution of river channels. Nowadays, the high availability of data(e.g. hydrological data and aerial images) and the technologies for theiranalysis (i.e. GIS) provide new opportunities to make a more detailedanalysis of the geomorphological changes over time, and also allowexamining the current channel state in relation to pre-dam conditions(i.e. taken as reference conditions for the comparative analysis in thispaper).

Themain objective of this paper is to examine the effects of the damson the channel geomorphology of 74 selected river reaches in the IberianPeninsula. To accomplish this objective we present a methodology thatintegrates flow series and climate data, catchment scale information(i.e. GIS layers) and digital elevation models, ancient and contemporaryaerial imagery, and field data. The study sites are first statisticallyclassified (clustered) based on their climatic and hydrological characteris-tics (i.e.multivariate analysis). Subsequently, changes on the hydrologicaland flood regimes in each cluster are analysed with special emphasis tothose reaches downstream from dams. The geomorphic status of studyreaches is assessed by means of a novel Geomorphic Status (hereafterGS) index that is presented to evaluate channel's geomorphic activity;results are finally coupled and interpreted at the light of the hydrologicalchanges, together with a discussion on the factors that most influenceriver channel activity.

2. Study sites

We analyse a selection of river sites in four contrasting basins in theIberian Peninsula (Fig. 1): Ebro (85,530 km2), Llobregat (4923 km2),Júcar (21,632 km2) and Guadalquivir (57,527 km2). Catchments aremostly located in the Mediterranean region, altogether encompassingan extensive latitudinal, thermal, rainfall and hydrological gradient.

t represents the number of reservoirs and their storage capacity for each of the basins. Sitesbelong (see Methods section for further information on clustering analysis).

103G. Lobera et al. / Science of the Total Environment 508 (2015) 101–114

The basins are characterised by frequent periods of low dischargeor even long dry periods during low-rainfall seasons, and flashyevents during wet seasons (i.e. Mediterraneity), but with avery different recurrence and intensity (i.e. the degree ofMediterraneity varies between sites and basins). Mean annual pre-cipitation at the basin-scale ranges from 520 to 700 mm, but itgreatly varies regionally, from i.e. N2000 mm in the Ebro headwa-ters to i.e. b300 mm in the Ebro Depression. In general, precipita-tion decreases on a north–south and west–east gradient. Anextensive network of dams for hydropower and irrigation purposeshave been built in the region, starting in the late 19th century, ac-counting for a considerable impoundment capacity (to express it weuse the Impoundment Ratio — IR, defined as the ratio between the reser-voir storage volume and the mean annual runoff expressed in percentage,asper Batalla et al. (2004)). There are187 largedams in theEbrobasin (res-ervoir storage volume V=8022 hm3, i.e. 1 hm3=1 × 106m3, mean IR=0.6 i.e. 60%), 118 in the Guadalquivir (V = 8192 hm3, IR = 1), 19 in theJúcar (V = 2741 hm3, IR = 2.2), and 3 in the Llobregat (V = 223 hm3,IR = 0.46). Dams typically alter flood frequency and magnitude,and subsequently channels react by reducing their morphologicalcomplexity (the presence of river forms) and activity (sedimentmobility that, in turn, affects river forms). For this study, a totalof 74 study sites were selected from 36 rivers representing small,medium and large basins (i.e. 24 sites in the Ebro, 14 in theLlobregat, 15 in the Júcar and 21 in the Guadalquivir). A total of47 sites are located downstream from dams (i.e. thus consideredregulated), with the rest not regulated (Fig. 1). Regulated sites are4 km to more than 100 km downstream from dams. The capacityof the reservoirs included in this study span several orders of mag-nitude, from 1 to more than 1000 hm3. Data base has beenfragmented in order to address each of the objectives and the cor-responding analysis.

Fig. 2. (A) Example ofmapping of sediment units (active and vegetated bars) and active channeof the River Guadiana Menor (i.e. GUAM, Guadalquivir basin). (B) Maps of the GUAM drainageprecipitation) variables that have been used for the cluster classification.

3. Methods

Data base includes information of 74 study sites (see location inFig. 1). Data have been fragmented according to the different analy-ses (i.e. Fig. 2 exemplifies how data is fragmented): a) classification(or clustering) based on their physiographic and climatic character-istics (i.e. multivariate analysis); b) analysis of the hydrological dif-ferences between sites within each group; c) assessment of thegeomorphic situation of the sites by means of the Geomorphic Status(GS) index; and, finally, d) examination of the relation between thecurrent river's geomorphic status and the hydrological alterationscaused by regulation. Following, we present a brief description of eachof the analysis (see Fig. 2 for a complete view of data fragmentationand analysis):

a) Classification (clustering): For this analysis all sites were used (n =74). The multivariate analysis provides the number of statisticallysignificant clusters based on physiographic and climatic characteris-tics.

b) Hydrological regime: A total of 46 sites provide the required data toperform the analysis (see details of the indices used to measure hy-drological alteration in the next sections). Of the total, 30 sites are lo-cated downstream from dams and 16 are located in rivers notregulated.

c) Channelmorphology: A total of 42 sites provide the required data toperform the geomorphic analysis (see details on the GS in the nextsections); of them, 31 sites are located downstream from damsand 11 are located in not regulated rivers.

d) Effects of regulation on the geomorphic activity: This analysis isperformed only for the sites located downstream from damswhere hydrological information and GS could be coupled. A totalof 19 river stretches were considered for this part of the analysis.

l width transects along a reach, illustrated by two images (historical and current) of a reachbasin that show some of the physical (stream order and slope) and climate (mean annual

Fig. 3.Workflow summary for the hydrologic and the geomorphic analysis. Note that themain sets of analyses are indicated togetherwith the objectives, the number of sites used in eachanalysis and the main outcomes (see text in Section 3.3 for more details).


In the following sections, we first describe in detail the differentcomponents of the data base and introduce the main variables thatwere calculated. Secondly, we provide details on how data analyseswere performed and, finally, we present the Geomorphic Status (GS)index, an indicator specifically developed here to assess the geomorphicstatus of regulated river reaches.

3.1. Data compilation and fragmentation and extraction of variables

Table 1 summarizes the main variables calculated from the sixcomponents of the data base i.e. five components are based on archivalinformation (e.g. flow series, aerial photos), in addition to the directfield observations on the current river morphology specificallyundertaken for this study.

3.1.1. Physiographic variables (catchment)Catchments upstream from each site were characterised using

ESRI® ArcMap™ 9.3 (Table 1). The spatial analysis was based on adigital elevation model (DEM) with a cell size ranging from 20 × 20to 30 × 30 m. ESRI Spatial Analyst® was used to process each DEMand delineate the drainage area (A), the stream network (L) andthe stream order (SO) following Strahler's classification in eachsite. ESRI 3D Analyst Tools® were used to define the mean basinslope (Sb). The Gravelius index (KG) was also calculated to representthe basin shape, as this may influence the hydrological response (seeTable 1 for definition). Geology of the catchments was describedwith the average Rock Resistance Class value (RRC; as per Claytonand Shamoon, 1998), re-drawn and estimated from the 1:1,000,000Geological Map of the Iberian Peninsula (Spanish Geological andMining Institute— IGME). This index classifies lithologies in 6 classesaccording to their resistance to erosion (i.e. 1 very weak to 6 veryresistant).

3.1.2. Climatic variables (catchment)Monthly precipitation and evapotranspiration data from the period

1950–2009 were obtained from the SIA (Integrated Water informationSystem, Spanish Ministry of Agriculture, Food and Environment; formore information see www.magrama.gob.es). Data are provided inraster format of 1 × 1 km pixel resolution. From the raw data severalclimatic variables were elaborated (see Table 1 for details): (1) meanannual precipitation (Pm), (2) Aridity Index (AI), here simplytaken as the ratio between mean annual precipitation and meanannual evapotranspiration and, (3) the coefficient of variation betweenmonthly precipitation means (MC); the last expresses the variabilitybetween months and reflects the degree of Mediterraneity of a givensite. All variables were derived using ESRI ArcMap® 9.3 from the digitalrasters.

3.1.3. Hydrological variables (at-a-site)Hydrological data were obtained from daily streamflow series

compiled from various water agencies in Spain, namely the CatalanWater Agency — ACA, the Hydrographic Studies Center — CEDEX, andthe Automatic Hydrological Information System — SAIH. Hydrologicalresponses through time were analysed for both the regulated and thenon-regulated sites.

The analyses of the effects of dams on river's hydrologywere limitedto the sites where pre-dam data was available (n = 30) that presentsseries varying from 6 years to 83,with an average of 33 years per series.We are aware that there are some limitations associated to length of theflow series, thus this may have a certain influence on the final results;however, given the intrinsic nature of the study (large catchments ona broad scale) we believe that the general patterns described in thepaper will remain very similar and that the final effect of the time serieslength shall be considered asminor. Dailyflowswere analysedusing theIndicators of Hydrologic Alteration (IHA) software (Richter et al., 1996;Mathews and Richter, 2007) that producesmore than 30 variables, fromwhich we selected four (1-day maximum flow, number of high flow

http://www.magrama.gob.es

Table 1Basin and stream reach characteristics used to define study sites and analyse the impact ofregulation.

Component/variables Definition

PhysiographicDrainage area (A, km2) Area drained by the river (at the study site) and

its tributariesChannel length (L, km) Length of the main river and their tributariesMean basin slope (Sb, %) Mean slope of the drainage areaGravelius index (KG) Basin area divided by the corresponding

catchment perimeterStream Order (SO) Stream Order by Strahler (1957)Rock Resistance Class (RRC) Classification according to resistance to the

erosion

ClimaticMean annual precipitation(Pm in mm)

Mean annual precipitation of the drainage area

Monthly variation coefficient(MC in %)

Variation between mean month precipitation

Aridity Index (AI) Quotient between ETP and precipitation

HydrologicalQa Day mean annual flowQ2, Q10 and Q25 Return period of 2, 10 and 25 years1 day maximum flow (DMF) Mean of the highest single daily value each yearBase flow index (BFI) 7-day minimum flow divided by the average

annual flowNumber of low flow pulses(NLF)b

Mean of the number of low flow pulses each year

Number of high flow pulses(NHF)a

Mean of the number of high flow pulses each year

Number of reversals (NR) Mean of the number of flow variation betweenconsecutive days

Torrentiality Mean maximum discharge between meandischarge

GeomorphicActive bars (NA) Number of active barsVegetated bars (NV) Number of vegetated barsTotal bars (NB) Number of total barsChannel width (W) Mean of active channel width of the stream reach

Dam impactDistance (Dr) Distance to the dam from the study siteReach Impoundment Ratio(IRr)

Impoundment ratio according to the upstreamreservoir

Total Impoundment Ratio(IRt)

Impoundment ratio according to the sum ofreservoirs

Tributaries Number of tributaries to the dam from the study site

Field-basedShear stress Calculated following the DuBoys approachArmouring Ratio between median surface and subsurface

bed-material

a High pulse defined as the 75th flow percentile of the flow frequency distribution(as per Richter et al. 1996).

b Low pulse defined as the 25th flow percentile of the flow frequency distribution(as per Richter et al. 1996).


pulses, number of low flow pulses and number of reversals; see Table 1for notations and definitions) because of their relevance for channelgeomorphology (Graf, 2006). In addition, three more variables (i.e. dis-charges corresponding to a return period of 2, 10 and 25 years, calculat-ed by means of the Extreme Gumbel Value distribution) and added tocomplete the analyses (Table 1). All variables were calculated for thepre- and post-dam records. The ratio between post- and pre-dam re-cords indicates the effect of regulation on each hydrological variable.

Additionally, hydrological data of not regulated siteswas analysed inorder to study the existence of changes that suggest a response to otherenvironmental pressures (either natural or human-induced), such aschanges in land use, increase in water demand, or climate change. Ofthe 27 not regulated sites, only 16 presented suitable enough data forthe analysis (i.e.flow series had a length between22 and 96 years). Series

were also analysed bymeans of IHA, although in this caseweworkeddatarecords as a whole, looking for temporal trends from linear regression foreach variable over time. Finally, basin torrentiality, taken as a proxy of theMediterraneity character of the basin, was also calculated for bothregulated and not regulated river reaches; torrentiality is representedby the ratio between the mean of the maximum discharges in the series(annual instantaneous maximum — Qci or annual daily maximum — Qc

depending on data availability) and the mean discharge (Qn).

3.1.4. Geomorphic variablesA series of geomorphic variables were calculated and compared

from two sets of aerial photographs i.e. 1970s and the current (2011).It shouldbenoted that geomorphic variables provide a general evaluationof the at-a-moment channel activity, as we do not aim for a detailedcharacterisation of channel evolution. Images from the late 1970s, freelyavailable at National Geographical Institute — IGN web site (www.ign.es), were taken as the reference image i.e. most dams in Spain werebuilt between 1960 and 1970s and we considered the geomorphologyof the river reaches downstream from the dams not yet severely affected.The current imageswere also obtained through the IGNweb site by usingthe ‘Iberpix Viewer’.

In order to extract the variables across a representative reach, wedefined the optimal channel length as the one that equals 20 timesthe width of the active channel (following Walter and Tullos, 2010criteria). Four variables were extracted in both set of images (Fig. 3;Table 1): (1) number of active lateral and point bars (NA), taken aspotential unstable features that are frequently inundated during highflows; (2) number of vegetated bars (NV) (i.e. N50% of the bar area iscovered by vegetation), that are relatively stable forms only inundatedduring larger floods; (3) total number of bars (NB); and (4) mean activechannel width (W) calculated from 10–20 transects (depending on theriver size) along the main channel. It is worth to mention that thesevariables have a different character than the ones that can be obtainedby spatial disaggregation and aggregation procedures as, for instant,those described by Alber and Piégay (2011). Such approaches are usedto support and automate large-scale characterisation of fluvial systems,while the objective of this paper is to obtain local variables that provideinformation about the geomorphic status of different river reaches locatedin different catchments at a given moment.

3.1.5. Impact variablesTo estimate the degree of hydrological impact affecting each site and

to assess the influence of regulation on river morphology, we usedthe Impoundment Ratio (IR). In our case, two IRs were calculated(Table 1): 1) reach IR (IRr) that relates to the closest reservoir locatedupstream of the study site and, 2) total IR (IRt) that relates to theimpoundment capacity of all reservoirs located upstream from thestudy site. Reservoir capacity and mean annual runoff were obtainedfrom water agencies. Mean annual runoff of sites where official flowrecords are not available was obtained from the CEDEX-SIMPA Rainfall–Runoff Integrated Modelling System. Finally, the distance from the damto the study site and the number of tributaries between the dam andthe site were also calculated (Table 1). Only the tributaries with a streamorder equal or higher to half of that of the study site were taken intoconsideration; the role of smaller tributaries on the mainstream riverflow and channel forms was considered negligible.

3.1.6. Field-based variablesData base was completed with ad-hoc field data obtained in 2010

and 2011. All the 74 study sites were surveyed, although not all ofthem were suitable to obtain the complete set of variables (i.e. sed-iments not exposed, too deep for wading, etc.); 26 sites were finallyanalysed (Table 7). Channel slope were measured by means of aGeodimeter® 422 Total Station and a Leica® TCRP1201 RoboticTotal Station. To characterise surface sediments on the river bed,the b axis of a minimum of 100 particles was measured using a

http://www.ign.es

http://www.ign.es


gravel template with squared holes at 1/2 ϕ unit classes (particlesfiner than 8 mm were not included, as per Wolman (1954)). Parti-cles were sampled in open bars at sedimentologically equivalentpositions, typically bar head to avoid downstream fining (Rice andChurch, 1998). Subsurface material was sampled in the same ex-posed bars using the volumetric method (Church et al., 1987), i.e.a representative patch of the bar (typically of 1 m2) was spray-painted to differentiate the surface from the subsurface material(Lane and Carlson, 1953), the volume of the subsurface sample de-termined based on the weight of the largest particle (the weight ofthe largest particle never represented more than 1% of the totalsample, as per Church et al. (1987) criteria in gravel-bed rivers).Subsurface materials were extracted and transported to the labora-tory where they were dried and sieved for a later size-class classifi-cation according to the Wentworth scale.

Finally, current channel activity was characterised by combiningtopographical surveys with surface sediment data. This way, chan-nel potential activity is characterised from shear stress (calculatedfollowing the DuBoys approach) at bankfull stage, and riverbedarmouring (ratio between median surface and median subsurfacebed-materials from grain-size distributions) (see Results andDiscussion section and Table 7 for complete data).

3.2. Assessment of the Geomorphic Status of regulated rivers

An index to assess the Geomorphic Status (GS) and the degree ofchange of each regulated site was developed based on the variablesextracted in the geomorphic component of the data base (Table 1).The GS includes four dimensionless indices:

1. Changes in Sedimentary Units (SU). It represents the difference inthe geomorphic complexity of the river reach. SU is calculated as:

SU ¼ 1þ NB=Lð Þpost� �h i

= 1þ NB=Lð Þpre� �h i

where NB is the number of bars, L is channel length (km) and preand post refer to the reference and current images, respectively. Aresult of 1 implies no change, whereas values N1 or b1 indicate anincrease or decrease, respectively, of the number of sedimentaryunits.

2. Changes in Sediment Availability (SA). It indicates the river dynamismand the availability of sediment coming from upstream, and it isrepresented by the degree of bar activity. If NBpost is 0 then SA mustbe 0 (if no bars are exposed we consider that sediment availability isnegligible in the reach); in contrast, if NBpost is N0 then SA can becalculated as follows:

SA ¼ 1þ NA=NBð Þpost� �h i

= 1þ NA=NBð Þð Þpreh i

where NA is the number of active bars. If results equal 1 therehave not been major changes in sediment availability, whereasvalues N1 imply an increase on sediment deposition and b1 areduction of sediment availability.

3. Changes in Bar Stability (BS). The presence of vegetation in bars is anindicator of stability. BS evaluates the difference in vegetation coverthrough time, thus complementing SA. As in the previous case,if NBpost is 0 then BS must be 0; but, if NBpost is N0 then BS can becalculated as:

BS ¼ 1þ NV=NBð Þpre� �h i

= 1þ NV=NBð Þpost� �h i

whereNV is the number of vegetated bars and pre and post refer to thereference and current images, respectively. A bar has been consideredas vegetatedwhen N50% of its area is covered by vegetation. Values N1

imply a reduction of the vegetation cover and b1 an increase of thenumber of vegetated bars over time.

4. Changes in Channel Flow (CF) capacity. This index evaluates thevariation on the active channel width, which relates mainly withchanges in the frequency and magnitude of flood events. CF iscalculated by means of:

CF ¼ Wpost=Wpre

where W is the mean width of the active channel (in m) in thereach, and pre and post refer to the reference and current images,respectively. Values N1 imply an increase on the active width, sug-gesting an increase of the frequency and magnitude of competentevents.

GS is calculated as the sum of the previous indices (i.e. GS =SU+ SA+ BS+ CF). A GS equal or close to 4 implies an overall mainte-nance on the geomorphic characteristics of the reach,whereas values N4would indicate an increment of the geomorphological activity in thereach, and values b4 would suggest a tendency towards channel stabili-zation or degradation (i.e. loss of geomorphic diversity, simplification ofthe channel pattern, disappearance of sedimentary active areas, togetherwith bed armouring and incision). It is worth to mention than the appli-cability of this index is limited to the characteristics of the river reach(e.g. channel width, presence of expose bars) and to the availability ofappropriate aerials (e.g. scale, flow conditions) that allow calculatingthe necessary variables.

3.3. Data analysis

(1) Multivariate analysis (clustering)Study sites were first classified into homogeneous subregions

according to their basin physical (drainage basin area, mean basinslope, Gravelius index, rock resistance, channel length and streamorder) and climatic (Pm, AI and MC) variables (Table 1). To this end, aSpearman correlation matrix was used to detect redundancy amongthese variables. Next, Principal Components Analysis (PCA) was appliedto assess the weight of each of remaining variable (i.e. for instance, rockresistance that was lately eliminated due to its low significance).Finally, the five remaining variables were used to create homogeneoussubregions after applying the k-means clustering method by means ofStatSoft STATISTICA® 7.0. In order to define the number of clusters, twotechniques were used: (1) V-fold cross-validation for which repeatedrandom samples are drawn from the data for the analysis and therespective model is then applied to compute predictive classification;and (2) trial and error approach to select themost appropriatednumberof clusters which adjust to an ANOVA criterion for a significant levelp b 0.001 (Fang et al., 2012). Once sites were classified into statisticallysignificant clusters, one-way ANOVA was conducted to determinewhich variables are the main responsible to differentiate the groups.Additionally, Tukey post-hoc test was also applied to define wheredifferences are related to the existence of three or more groups.

(2) Hydrological analysisThe second set of analyses is performed to examine the influence

of physical and climatic conditions on river hydrology. As stated,hydrological data were available for 46 sites (30 regulated and 16 notregulated). Natural conditions (before dam construction) were assessedby calculating the median and the first and third quartiles (i.e. quantiles25th and75th) of eachof thehydrological variables for each cluster. Sub-sequently, the degree of change for both regulated and non-regulatedreaches was examined. In the first case, the change due to the regulationwas calculated by means of the ratio between post-dam and pre-damvalues of variables. In this way, a value N1 represents an increase of themagnitude of the variable analysed. In relation to the non-regulated


sites, the hydrological response was evaluated by means of linearregressions.

(3) Geomorphological analysis

The hydrological analyses were followed by geomorphologicalanalyses. We assessed the differences between the reference (i.e., endof 1970s) and the current (i.e., 2011) channel geomorphology at eachstudy site. Morphological variables (Table 1) were examined at eachof the 74 study sites; although 20 of them did not show any visiblegeomorphological structure (i.e. bars) already in the reference image.In addition, 12 sites were also excluded because either they were toosmall or the riparian vegetation precluded any correct measurementof the river channel. Thereby, analyses were finally carried out for 42reaches. The remaining sites were subsequently grouped after the clus-ters previously established to examine the degree of influence of phys-iographic and climatic conditions on channel morphology. After thisclassification, median and selected quartiles of each geomorphic vari-able both for the reference and the current conditions were calculated.A multiple regression model was applied to derive statistical relationsbetween dam impact parameters (as independent variables) andboth hydrologic and geomorphic parameters (as dependent variables).

Fig. 4. Box plots of the climatic and physiographical characteristics for each cluster: (A) drainag(mm), (E) monthly variation coefficient of precipitation (%).

Finally, the relation between hydrological and geomorphological re-sponses is established in rivers affected by regulation.

4. Results and discussion

4.1. Site classification in relation to physiographic and climatic variables

The Spearman correlation test indicates that stream length (L),stream order (SO) and aridity (AI) had to be excluded because theyare redundant (r N 0.9, p b 0.05) with drainage basin area (i.e. L andSO) and mean annual precipitation (i.e. AI), respectively. The first twofactorial axes in the PCA explained 71.5% of the variance; owing to thisanalysis we also eliminated the Rock Resistance Class (RRC) because itexplains b0.01 of the variance for the first two axes. The rest of thevariables (drainage basin area, mean basin slope, Gravelius index (seeTable 1 for definition),mean annual precipitation andmonthly variationcoefficient) were taken for the next step. The subsequent V-fold crossvalidation and trial and error approach grouped the study sites intofour climatic and physiographic subregions (Fig. 4). The one-wayANOVA reveals a significant difference (p b 0.05) between the descrip-tive variables of the clusters. However, the Tukey post-hoc test indicatesthat not all variables present significant differences (p b 0.05) between

e area (km2), (B) mean basin slope (%), (C) Gravelius index, (D)mean annual precipitation

Fig. 5. Box plots of the quotient between post-dam and pre-dam for each hydrological variable (as per Richter et al. (1996); see Table 1 for details) (Qa: daily mean annual flow, DMF: 1-daymaximum flow, NLF: number of low flow pulses, NHF: number of high flow pulses, NR: number of reversals, flood with 2, 10 and 25 years of return period, respectively). Horizontal dottedline indicates no change, N1 imply a reduction and b1 imply an increase, between pre-dam and post-dam periods. The medians (central bar), 25th–75th quantiles (box) and non-outlierrange (whiskers) are shown in the plot (for clarity outliers not shown).

Table 3Temporal trend (expressed and the slope of the linear function) of each hydrologicalvariable in the not regulated sites with more than 20 years of data available (valueshighlighted with *means that correlations are significant at 0.05 confidence level).


clusters. The final cluster and the variables that allowdistinguishing oneto the other are defined as follows:

– Cluster #1 (13 sites) grouped river sites with high mean basin slopeand high annual precipitation, characteristics typical of headwaterrivers;

– Cluster #2 (22 sites) grouped reaches with large drainage areas andhighGravelius index (i.e. elongated basin shape), whichwere locatedin lower-most parts of the Ebro, Júcar and Guadalquivir basins;

– Cluster #3 (16 sites) grouped reaches in the Guadalquivir basincharacterised by a high degree of Mediterraneity and low annualprecipitation; and

– Cluster #4 (23 sites) encompasses a mixture of reaches with nocharacteristically different features, and includes the majority of thestudy sites in the Llobregat and the Júcar basins.

4.2. Hydrological responses in regulated and non-regulated rivers

In regulated rivers (30 sites), the annual streamflow and themagnitude and frequency of floods showed a generalized decrease(Fig. 5). This decrease is especially remarkable for the high-frequencyand low-magnitude events (i.e. DMF andQ2). Conversely, the frequencyof lowdischarges and the number of reversals increased (see Table 1 fordefinitions). This increment may be due to particular dam operations,such as particular releases for water supply (Graf, 2006). Hydrologicalvariables related to flow magnitude (DMF, Q2, Q10 and Q25) werenegatively correlated with the degree of regulation and positively

Table 2Multiple regression between hydrologic variables (§ indicates the ratio between post andpre-dam values) and degree of regulation (IRr), number of tributaries and distance to thedam (*statistically significant at 0.05 and +significant at 0.1 confidence level).

IRr No. of tributaries Distance (km)

§Qaa −0.28 0.09 0.16

§DMFb −0.50* 0.12 −0.03§NLFc −0.40* −0.34 0.35§NHFd −0.31 −0.40 0.55*§NRe −0.44* −0.43+ 0.63*§Q2

f −0.44* 0.47* −0.26§Q10

f −0.30 0.35 −0.26§Q25

f −0.29 0.34 −0.25

a Daily mean annual flow (as per Richter et al., 1996; see Table 1 for definition).b 1-day maximum flow.c Number of low flow pulses.d Number of high flow pulses.e Number of reversals.f Flood with 2, 10 and 25 years of return period, respectively.

with the number of tributaries (Table 2). This fact indicates that theimpacts of dams are attenuated downstream in relation to tributaryinflows. In addition, the hydrological variables related to frequency ofhydrological events (NLF, NHF and RDF) tend to increase below damsin sites with low regulation, in some tributaries and in sites located farfrom the reservoir.

In turn, not regulated rivers (16 sites) present more heterogeneousresponses i.e. not all sites have significant changes in the frequencyand magnitude of annual floods (as per Richter et al. (1996), seeTable 1 for notations and definitions) during the periodwith historic re-cords (Table 3). Frequency of low pulses (NLF) increased at 7 sites anddid not change at 9, the frequency of high pulses (NHF) increased at 2sites, decreased at 4 and did not change at 10, whereas the number ofreversals (NR) increased at 6 sites, decreased at 3 and did not changeat 7. Otherwise, the magnitude of annual floods (DMF) increased in 1site, decreased in 4, and did not change in 11. Hydrological changes inrivers not affected by dams may reflect the influence of other factorssuch as climate variability, land uses changes and water abstractions(Arora and Boer, 2001; Lorenzo-Lacruz et al., 2012; Rinaldi et al.,2013). For instance, a recent study of 187 river basins in the IberianPeninsula (Lorenzo-Lacruz et al., 2012) showed a generalized decreaseof annual flow, corroborating previous findings by Gallart and Llorens

Site DMFa NLFb NHFc NRd

Ebro OCA −0.427 −0.036 −0.038 −0.378MAT 1.491 0.190* −0.235* −0.608MAR −0.006 0.085* −0.021 0.623*EBR1 −0.336* 0.109* 0.011 0.320ALG 0.769 0.045 0.009 0.911*

Llobregat LLO1 −0.031 0.229* −0.069 1.683*ANO1 −0.148* 0.043* −0.056* 0.032ANO3 −0.719* 0.301* 0.187* 2.838*

Júcar CAB2 −0.333 −0.011 −0.071* −0.624*CAB3 −0.453 −0.309 −0.404* −1.443*JUC1 −0.204 −0.104 0.077 −0.515

Guadalquivir GEN1 −0.466 0.024 −0.030 0.614*GUAN −0.759* 0.060* −0.025 −0.531*YEG −3.413 0.025 −0.033 2.204*GUAA 0.659 −0.055 −0.068 1.225GUAR 3.312* 0.060 0.326* 2.137*

a 1-day maximum flow (as per Richter et al., 1996; see Table 1 for definition).b Number of low flow pulses.c Number of high flow pulses.d Number of reversals.

Fig. 6.Box plots of analysed hydrological variables (as per Richter et al. (1996); see Table 1 for details). (A)Qa: dailymean annualflow, (B)DMF/Qa: 1-daymaximumflowdivided bydailymeanannual flow, (C) BFI: base flow index, (D) NR: number of reversals, (E) NLF: number of low flow pulses, (F) NHF: number of high flow pulses. The medians (central bar), 25th–75th quantiles(box) and non-outlier range (whiskers) are shown in the plot (outliers not shown).


(2004). This tendency is generally related to the pattern of decreasingprecipitation found in the Mediterranean basin (Xoplaki et al., 2004),but also with an increase of the forested land and an expansion of theirrigated surface. The increase of the evapotranspiration caused by theabandonment of cultivated fields and its replacement by shrubs andforests (Hill et al., 2008) togetherwith the increment of the temperature(Chaouche et al., 2010), result in a reduction of the runoff generation inthe headwaters (most of the not regulated rivers in this paper arelocated in headwaters). Thus, land use seems to have lesser influenceon them than on average runoff. Three sites showing a negative pattern(CAB2, CAB3 and ANO1) have suffered a significant increase in forestand natural vegetation cover (Gallart et al., 2011; Herráiz-Hernansanzand Serrano-Gil, 2013). These results are consistent with those obtainedby López-Moreno et al. (2006), who showed a general decrease in floodintensity despite no change in the frequency and distribution of the pre-cipitation in the central Spanish Pyrenees. Altogether, results fit with thepattern described in the IPCC, 2007 for theMediterranean region, whichreported an increase in heavy precipitation events that causes an in-crease of flooding, but a decrease in total runoff that reduces meanand minimum flows.

Results show that hydrological responses differed between riverclusters (Fig. 6). Headwater sites (Cluster #1) were characterisedby frequent pulses of high and low flow. In contrast, lower reaches

(Cluster #2) exhibit higher mean daily flow and higher number ofreversals. Mediterranean sites (Cluster #3) showed minimum baseflow (i.e. ephemeral rivers), the largest 1-day maximum flow andminimum number of low and high flow pulses. Finally, sites in Cluster#4 showed intermediate behaviour but higher recurrence of high andlow pulses.

4.3. Geomorphological response to river regulation

Fig. 7 shows the number of active (NA) and vegetated (NV) bars(expressed in percentage) in relation to the local impounded ration(IRr) for the reference and contemporary periods. In the late 1970smost sites presented an elevated percentage of active bars (Fig. 7A)with a median of 59% of active bars over the total number of barsfor all the sites. This pattern suggests that although impoundmentwas present in some of the analysed river reaches, river channelsare still active because the time since dam construction was stilltoo short and probably below the required reaction time neededfor the effects being observable. The pattern reverseswhen the currentsituation is analysed (Fig. 7B). Currently, most bars have beenencroached by vegetation, and only sites with little or no regulationmaintain exposed bars i.e. the median percentage of vegetated barsreaches nowadays 97%. During periods of relatively small floods,

0

20

40

60

80

100

Nu

mb

er o

f B

ars

(NB

, %)

Active bars Vegetated bars

0

20

40

60

80

100

0 20 40 60 80 100 120 140 160

Nu

mb

er o

f b

ars

(NB

, %)

IRr (%)

Active bars Vegetated bars

A

B

Fig. 7. Percentage of active and vegetated bars for each site in relation to the degree of regulation (local Impoundment Ratio— IRr; see Section 3.1 in the text for reference) in the referenceimage (A) and the current situation (B).


riparian vegetation is able to colonize fresh gravel deposits, leading tothe conversion of active channel bars to relict deposits and floodplain(Nelson et al., 2013). The stabilization limits river dynamics and con-tributes to the environmental degradation of the fluvial ecosystem(Ollero, 2010; Magdaleno et al., 2012), thus reducing the habitat foraquatic and riparian wildlife (Graf, 2006).

The GS of the 42 study sites (Fig. 3) was negatively correlated to IRr,but not to IRt (Table 4, Fig. 8), thus showing that the effect of the localregulation is more important than that of the total basin; a dam locatedimmediately upstream from a study reach controls the flow regime andthe sediment trapping (bedload) directly affects the reach; whereas theeffects of other dams located further upstream are presumably morediffuse. The hydrological effects of dams propagate further downstreamtoo, and their effects on sediment movement and the sediment budgets(thus channel form) tend to recover (although never completely) as theriver merges with sedimentary active tributaries. Distance for partialrecovery of sediment load and channel activity (i.e. looseness of surfacegravel) has been reported from few tens of kilometres (River Ebro, Petts

Table 4Correlation matrix between Geomorphic Status — GS and degree of regulation (as perBatalla et al., 2004), where IRr means the impoundment ratio in relation to the singleupstream reservoir and IRtmeans the impoundment ratio related to all upstream reservoirs(*statistically significant at 0.05 confidence level).

GS IRr IRt

GS 1 −0.35* 0.02IRr −0.35* 1 0.24IRt 0.02 0.24 1

and Gurnell, 2005) to hundreds of km (River Mississippi, Williams andWolman, 1984). It is also worth pointing out the fact that some of thenot regulated sites with a very low GS (i.e. ANO2, ANO3, CAB2, CAB3)can be related to the reduction of the floods due to the hugereforestation of the basin. On the other hand, MAG1 also shows a verylow GS due to land use changes but, in this case, to the fact that riparianforests have been substituted by agriculture, thus limiting channel mo-bility, a phenomenonwhich is typically enhanced by engineeringworksthat are usually associated (e.g. embankments and rip-raps).

Finally, themultiple regression analysis between geomorphic indices(i.e. SU, SA, BS, CF and GS) and impact variables (i.e. IRr, number oftributaries, and distance to the dam) yielded only a few statisticallysignificant relations (Table 5). Overall, regulation was negativelycorrelated with the abundance of active areas and GS. Furthermore, thenumber of tributaries and the distance from dams were significantlycorrelated only with change in flow capacity, with β coefficients of−0.47and +0.51, respectively. Distance and number of tributaries could not re-verse the local impact of the dam, since it is known that, for many rivers,headwaters provide typically more than 3/4 of the sediment load (e.g.Petts and Gurnell, 2005) which, in this situation, is captured by dams.

At-a-site complexity was related to upstream basin characteristics.Headwater sites (Cluster #1) presented a complex channel structurein the reference image (Fig. 9), but most (90%) were subsequentlyaffected by regulation. Sites in this cluster showed the largest reductionin the number of total and active bars, and a substantial increase in thevegetated ones. The lowermost reaches (Cluster #2) showed littlechanges in morphological structures between periods of analysis (here100% of the sites are regulated). Results suggest that the rivers were

Fig. 8. Box plot of the Geomorphic Status— GS in relation to different degrees of impoundment. Horizontal dotted line indicates no change, N4 implies an improvement and b4 impliesa deterioration, between reference and the current periods. The medians (central bar), 25th–75th quantiles (box) and non-outlier range (whiskers) are shown in the plot (outliers notshown).


already stabilized (i.e. most of the bars were vegetated in the referenceimage) before dam construction. This finding is found to be consistentwith a study in the lower Rhône River (Provansal et al., 2014) wherethe changes in the sediment balance and its consequences on channelmorphology were already important following reforestation andengineering works at the beginning of the 20th century; consequently,dams built up over the last 50 years would have been of lesser impact.The Mediterranean sites (Cluster #3) showed the highest variabilityregarding channel features, mainly in relation to active bars. It mustbe noted that more than half of the sites of this group were excludedfrom the analysis because they did not show any geomorphologicalstructure in the reference aerial photographs. The main reason for thisrelies on the fact that in the 1970s (reference image) the margins andriver banks were already occupied by agriculture. Sites in Cluster #4(mid-size basins) showed a notable geomorphic complexity in thereference image, but less active bars andmore vegetation in the presentsituation.

Overall, rivers in Cluster #1 showed the greater decline in thesedimentary activity (Fig. 4). In general terms, the degree of activityof a given channel depends directly on the flow competence and thesediment supply. Cluster #1 are headwater sites that generally havea great potential for bedload transport, on which river forms are builtup, and exhibit frequent competent episodes (flows). Therefore,when such a dynamic system is altered by a dam, the system cuts-offand channel forms adapt to it, typically losing dynamism.

Table 5Multiple regression between geomorphic variables and impoundment ratio (IRr), numberof tributaries and distance to the dam (+statistically significant at 0.1 confidence level).

IRr Number of tributaries Distance (km)

SUa −0.25 −0.18 0.26SAb −0.34+ 0.01 −0.03BSc −0.32 0.01 −0.05CFd −0.12 −0.47+ 0.51+

GSe −0.35+ −0.16 0.17

a Changes in sedimentary unities.b Changes in sediment availability.c Changes in bar stability.d Changes in channel flow capacity.e Index of Geomorphic Status — GS.

4.4. Hydrological and geomorphological responses

Hydrological alteration implies a change in the flow energy budgetbelow the dam, and this generally brings a geomorphic adjustment ofthe river downstream. In our case, a correlation analysis betweenhydrological (Qa, NLF, NHF, NR, Q2, Q10 and Q25; Table 1) and geomor-phological (i.e. SU, SA, BS, CF and GS; Table 1) variables for the 19 riverreaches (Fig. 3) was carried out (Table 6). Results showed that thereduction in reach complexity is positively correlated with a decreasein the magnitude and frequency of high flows (Q 2, Q 10 and Q25).The decline in active bars and the increment in vegetated bars wereparticularly related to Q2, and consequently, the decrease in GSfollowed mostly with the decline of relatively frequent floods. Thedischarge that transports most of the sediment over the long termis commonly termed ‘effective’ discharge, whereas the dischargeperforming most of the geomorphic work in the channel is termeddominant, both generally related to bankfull level, whose recurrenceintervals have been accepted between 1 and 2 years for many rivers(Leopold et al, 1964). Other studies (e.g. Surian, 1999) suggest thatchannel morphology has adjusted to the post-dam relatively frequentfloods (i.e. up to around bankfull, Q2), rather than to the large events. Itis worth noting that changes in hydrology and the reduction in the activewidth were not significantly correlated in our work (Table 6), suggestingthat other factors may influence such reduction as vegetation encroach-ment and human activities (e.g. agricultural intrusion). Frequently, regu-lation increases summer flows for irrigation purposes (Batalla et al.,2004), and may cause the growth and stabilization of vegetation in thechannel (Magdaleno and Fernández, 2011). Although such vegetationmay decrease channel mobility, riparian areas provide important otherfunctions to the fluvial system such as nutrient cycling, flood attenuation,carbon dioxide sequestration, sediment deposition and wildlife habitat(Sharitz and Mitsch, 1993). Moreover, in the absence of recurrent floods,fluvial territory is progressively invaded by human activities yielding areduction in the active area. The riparian vegetation communities arevery sensitive to changes in flood frequency, and species not adapted toinundation or saturation are restricted to the higher riparian elevations;therefore, the alteration of flow in regulated rivers has also effects onthe riparian communities, and may result in changes in the communityassemblage by reducing the biodiversity of wetland-adapted plants andanimals (Alldredge and Moore, 2014).

Finally, in addition to the historical analysis, an examination of thecurrent status of the study sites was also carried out based on field

Table 6Correlation matrix between geomorphic and hydrologic variables (§ indicates the ratiobetweenpost and pre-damvalues and *indicates significant correlation at 0.05 confidencelevel).

§DMFa §NLFb §NHFc §NRd §Q 2e §Q10

e §Q 25f

SUf 0.15 0.00 −0.31 0.08 0.66* 0.65* 0.65*SAg 0.19 −0.11 0.02 0.04 0.51* 0.41 0.40BSh 0.18 −0.10 0.03 0.06 0.51* 0.40 0.39CFCi −0.02 0.01 −0.15 0.10 −0.13 −0.09 −0.07GSj 0.18 −0.08 −0.12 0.09 0.57* 0.49 0.49

a 1-day maximum flow (as per Richter et al., 1996; see Table 1 for definition).b Number of low flow pulses.c Number of high flow pulses.d Number of reversals.e Flood with 2, 10 and 25 years of return period, respectively.f Changes in sedimentary unities.g Changes in sediment availability.h Changes in bar stability.i Changes in channel flow capacity.j Index of Geomorphic Status.

Fig. 9.Boxplots of the geomorphic parameters for reference and current streamconditions ineach cluster: (A) number of total bars (NB) divided by channel length. (B) number of activebars (NA) divided by channel length. (C) number of vegetated (NV) bars divided by channellength. The medians (central bar), 25th–75th quantiles (box) and non-outlier range(whiskers) are shown in the plot (outliers not shown).


data and hydrological series available at the selected sites (Table 7). Theaimof this analysiswas to determinedifferences between regulated andnot regulated rivers regarding their present geomorphological activity.Overall sampled sites in not regulated rivers showed a higher degreeof fluvial activity than their regulated counterparts, except for twocases; these two sites (GUAN and JUC1, Table 7) belong to groups #3and #4 (i.e. areas showing marked Mediterranean trends). Not-regulated sites are less armoured and have high instantaneoustorrentiality (mean of Qci/Qn = 55), altogether suggesting that channelsare still active (Table 7). In contrast, sites in rivers experiencing regulationshow a lower degree of fluvial activity; except for three sites (CAR1, GAL2and ZAD, Table 7), the rest belongmostly to groups #1 and #2 (i.e. head-water basins where most of dams in the basins are located, and lowlandbasins where fluvial landscapes use to be naturally gentle andgeomorphic processes of low intensity) and show armoured riverbeds.At these sites instantaneous torrentiality is much lower than in theMediterranean sites (mean of Qci/Qn = 13.6), altogether suggestingless intensemorphosedimentary processes, and reflecting the influenceof upstream regulation in their water and sediment budgets, thus inchannel form and dynamics. Rivers downstream from a dam responsemainly to the lack of sediment (i.e. sediment eroded from the bed isnot replace with solid load coming from upstream); but also to thereduction in the competence of flood discharge that, in turn, favoursthe encroachment of vegetation in formerly active areas, that,subsequently, reduces the local in-channel availability of sediment.This feedback phenomena is well-know and our results point to thesame direction, corroborating previous observations reported elsewhere(e.g. Williams and Wolman, 1984; Surian, 1999, Vericat et al., 2006).Moreover, our analysis put these findings into a broader perspective byexamining channel activity in regulated rivers along a latitudinal gradientthat covers different climatic subregions (here represented by fourstatistically significant clusters). The geographical amplitude of thiswork also points out the internal differences in each of the regions,a fact that demonstrates the complexity of the adjustments thattake place in a river affected by upstream reservoirs, precluding the useof simple generalizations at assessing implications for river conservationand implementing classic restoration measures.

5. Final remarks

This work examines the effects of flow regulation on the geomorphol-ogic activity in selected river sites of four large basins in the IberianPeninsula. A Geomorphic Status (GS) index that allows assessing thepotential geomorphic activity (i.e. dynamism) at a given river reach andits change through time has been used to assess channel response toregulation; in addition, ad-hoc field data is analysed to describe thecurrent activity of the study reaches. Hydrological trends geomorphicactivity is also examined in not regulated rivers that allow illustratingchanges beyond regulation. Conclusions from the work can be drawn asfollows:

1) Regulation of these Iberian rivers produces a decrease in the annualstream flow and in the magnitude and frequency of floods, togetherwith an increase in the frequency of lowdischarges and in the numberof reversals. These hydrological changes result mainly in the loss ofactive bars as they are encroached by vegetation, to the point thatonly sites with little or no regulation display active bars.

2) Accordingly, the geomorphic activity (as represented by GS) declineswith regulation (best represented by the local impoundment ratio,IRr). In contrast to hydrology, distance and number of tributariesdoes not reverse the geomorphic impact of the dam. Field datacorroborates the reduction of the morphosedimentary activityin river reaches located downstream from dams, with less energyexpenditure and higher riverbed armouring.

3) Headwaters sites are especially affected by regulation, reducing theiractivity and forcing them to experience low-energetic processes

Table 7Hydrological and geomorphological indicators derived from field data at selected sites in not regulated and regulated rivers. Mean values are discussed in text (see details in Sections 4.2and 4.3).

Locationa Torrentialityb Dynamismc

Site Basin Cluster # Instant.d Dailye Shear stressf Armouringg

Not regulated RS Ebro 1 ndh nd 4.4 ndBOR Guadalquivir 1 nd nd 56.8 ndCAC Guadalquivir 3 27.2 9.8 4.6 1.39GUAL Guadalquivir 3 nd nd 18.2 3.08GUAN Guadalquivir 3 17.4 12.5 12.1 3.28HER Guadalquivir 3 nd nd 151.9 ndPIC Guadalquivir 3 nd nd 20.7 1.63ALG Ebro 4 87.8 34.8 34.5 1.20MAR Ebro 4 27.3 9.2 61.7 1.33MAT Ebro 4 146.3 32.2 10.5 1.35ANO3 Llobregat 4 64.9 16.8 34.9 ndCAB1 Júcar 4 nd nd 34.4 6.01JUC1 Júcar 4 14.0 6.9 91.6 ndMean 55.0 17.5 41.3 2.4

Regulated ARG Ebro 1 15.4 12.7 1.3 8.54CIN1 Ebro 1 9.4 5.8 29.0 1.55CIN2 Ebro 1 12.0 7.4 nd 1.09ESE Ebro 1 10.6 5.7 7.8 5.85CAR1 Llobregat 1 24.7 9.7 18.9 1.18CAR2 Llobregat 1 10.1 9.5 16.2 ndJUC4 Júcar 2 2.8 2.3 31.2 ndEBR6 Ebro 2 nd nd nd 1.54GAL2 Ebro 4 20.7 12.8 nd 1.63HUE Ebro 4 9.4 5.6 8.2 ndZAD Ebro 4 20.4 15.7 5.9 6.59Mean 13.6 8.7 14.8 3.6

a See Section 3.3.1 in the text for details.b Ratio between the maximum discharge and mean discharge; used here as a proxy of Mediterraneity.c Force and resistance factors indicating degree of geomorphic activity in the channel.d Ratio between the mean of the annual maximum instantaneous discharge — Qci and the mean discharge.e Ratio between the mean of the annual maximum daily discharges — Qc and the mean discharge.f Shear stress at bankfull (N/m2).g Armouring ratio (ratio between median surface diameter and median subsurface diameter; see Section 3.1.6. for details).h No data available; due to technical difficulties the site could not be (fully) surveyed.


such as those situations typical from lowland rivers with highly sta-ble beds.

4) Finally, the reduction in reach complexity relates positively with thedecrease in the magnitude and frequency of high flows; particularlythe decline in active bars and the increment in vegetated bars arewell correlated with Q2 (assumed here as a proxy of the bankfulldischarge).

Overall, our results indicate that dams have progressively altered thegeomorphic status of Iberian rivers and raises concerns about thepossibilities to attain a good status according to the various EUdirectives related to rivers and river habitats. The novel GeomorphicStatus index complements the available Hydrological Alteration indicesand may aid further developments of the current tools to characterisethe hydromorphology of altered systems, as it is required by RiverBasin Management Plans. Information shall inform as well on a varietyof restoration actions, such as the design and implementation offlushingflows (i.e. artificialflow releases fromdams aiming at enhancingmorphosedimentary processes in the channel), together with the simul-taneous injection of sediments, to recreate channel forms and the associ-ate habitat for targeted species.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.scitotenv.2014.10.058.

Acknowledgements

This research has been carried out within the framework of theConsolider Ingenio 2010 CSD2009-00065 Project funded by the SpanishMinistry of Economy andCompetitiveness. Damià Vericat is in receipt ofa Ramon y Cajal Fellowship (RYC-2010-06264) funded by the SpanishMinistry of Science and Innovation. The authors acknowledge the

support from the Economy and Knowledge Department of the CatalanGovernment through the Consolidated Research Group 2014 SGR 645(RIUS — Fluvial Dynamics Research Group). Special thanks are due tothe Ebro, Júcar and Guadalquivir Water Authorities and the CatalanWater Agency for their collaborative support during the investigation,providing assistance and useful data. We are grateful to Lorea Floresfor her assistance during fieldwork and to Patrick Byrnewho undertooka helpful revision of the first version of the manuscript. Finally we fullythank Hervé Piégay and Fernando A.L. Pacheco for their positive andhelpful reviews that have greatly improved the final version of themanuscript.

References

Alber A, Piégay H. Spatial disaggregation and aggregation procedures for characterizingfluvial features at the network-scale: application to the Rhône basin (France).Geomorphology 2011;125:343–60.

Alldredge B, Moore G. Assessment of riparian vegetation sensitivity to river hydrologydownstream of a Major Texas Dam. River Res Appl 2014;30:230–44.

Arora VK, Boer GJ. Effects of simulated climate change on the hydrology of major riverbasins. J Geophys Res 2001;106:3335–48.

Batalla RJ, Vericat D. An appraisal of the contemporary sediment yield in the Ebro Basin. JSoils Sediments 2011;11:1070–81.

Batalla RJ, Gómez CM, Kondolf GM. Reservoir-induced hydrological changes in the EbroRiver basin (NE Spain). J Hydrol 2004;290:117–36.

Brandt SA. Classification of geomorphological effects downstream of dams. Catena 2000;40:375–401.

Chaouche K, Neppel L, Dieulin C, Pujol N, Ladouche B, Martin E, et al. Analyses ofprecipitation, temperature and evapotranspiration in a French Mediterranean region inthe context of climate change. CR Geosci 2010;342:234–43.

Church M. Geomorphic thresholds in riverine landscapes. Freshw Biol 2002;47:541–57.Church MA, McLean DG, Wolcott JF. River bed gravels: sampling and analysis. In: Thorne

CR, Bathurst JC, Hey RD, editors. Sediment transport in gravel-bed rivers. John Wileyand Sons; 1987. p. 43–88.

Clayton K, Shamoon N. A new approach to the relief of Great Britain II. A classification ofrocks based on relative resistance to denudation. Geomorphology 1998;25:155–71.


























Elosegi A, Díez J, Mutz M. Effects of hydromorphological integrity on biodiversity andfunctioning of river ecosystems. Hydrobiologia 2010;657:199–215.

Fang NF, Shi ZH, Li L, Guo ZL, Liu QJ, Ai L. The effects of rainfall regimes and land usechanges on runoff and soil loss in a small mountainous watershed. Catena 2012;99:1–8.

Fitzhugh T, Vogel RM. The impact of dams on flood flows in the United States. River ResAppl 2011;27:1192–215.

Gallart F, Llorens P. Observations on land cover changes and water resources in theheadwaters of the Ebro catchment, Iberian Peninsula. Phys Chem Earth 2004;29:769–73.

Gallart F, Delgado J, Beatson SJV, Posner H, Llorens P, Marcé R. Analysing the effect of globalchange on the historical trends of water resources in the headwaters of the Llobregatand Ter river basins (Catalonia, Spain). Phys Chem Earth 2011;36:655–61.

Gasith A, Resh VH. Streams in Mediterranean climate regions: abiotic influences and bioticresponses to predictable seasonal events. Annu Rev Ecol Syst 1999;30:51–81.

Gilvear DJ. Patterns of channel adjustment to impoundment of the upper river Spey,Scotland (1942–2000). River Res Appl 2004;20:151–65.

Graf WL. Downstream hydrologic and geomorphic effects of large dams on American rivers.Geomorphology 2006;79:336–60.

Herráiz-Hernansanz J, Serrano-Gil O. Despoblación y abandono de las actividadesagropecuarias tradicionales en la Serranía Alta de Cuenca: medidas de desarrollosostenible en un territorio rural desfavorecido. In: Camacho-Ballesta JA, Jiménez-Olivencia Y, editors. Desarrollo Regional Sostenible en tiempos de crisis. Universidadde Granada; 2013. p. 141–66.

Hill J, Stellmes M, Udelhoven Th, Röder A, Sommer S. Mediterranean desertification andland degradation. Mapping related land use change syndromes based on satelliteobservations. Global Planet Change 2008;64:146–57.

IPCC. Climate change 2007, synthesis report. Core Writing Team. In: Pachauri RK, ReisingerA, editors. Contribution of working groups I, II and III to the fourth assessment reportof the Intergovernmental Panel on Climate Change. Geneva, Switzerland: IPCC; 2007.[104 pp.].

Kondolf GM. Hungry water: effects of dams and gravel mining on river channels. EnvironManage 1997;21(4):533–51.

Lane EW, Carlson EJ. Some factors affecting the stability of canals constructed in coarsegranular material. Proceedings of the 5th IAHR Congress. Madrid: IAHR; 1953. p.37–48.

LAWA. Gewässerstrukturgütekartierung in der Bundesrepublik Deutschland -Verfahrenfür kleine und mittelgroße Fließgewässer. Berlin: Kulturbuch-Verlag; 2000. p. 1–145.

Leopold LB, Wolman MG, Miller JP. Fluvial processes in geomorphology. San Francisco:Freeman; 1964 [526 pp.].

López-Moreno JI, Beguería S, García-Ruiz JM. Trends in high flows in the central SpanishPyrenees: response to climatic factors or to land-use change? Hydrol Sci J 2006;51(6):1039–50.

Lorenzo-Lacruz J, Vicente-Serrano SM, López-Moreno JI, Morán-Tejedo E, Zabalza J.Recent trends in Iberian streamflows (1945–2005). J Hydrol 2012;414–415:463–75.

Magdaleno F, Fernández JA. Hydromorphological alteration of a large Mediterraneanriver: relative role of high and low flows on the evolution of riparian forests andchannel morphology. River Res Appl 2011;27:374–87.

Magdaleno F, Fernández JA, Merino S. The Ebro river in the 20th century or theecomorphological transformation of a large and dynamic Mediterranean channel.Earth Surf Process Landf 2012;37:486–98.

Mathews R, Richter BD. Application of the indicators of hydrologic alteration software inenvironmental flow setting. J Am Water Resour Assoc 2007;43:1400–13.

Merritt DM, Cooper DJ. Riparian vegetation and channel change is response to riverregulation: a comparative study of regulated and not regulated streams in the GreenRiver basin, USA. Regul Rivers Res Manag 2000;16:543–64.

Nelson NC, Erwin SO, Schmidt JC. Spatial and temporal patterns in channel change on theSnake River downstream from Jackson Lake dam, Wyoming. Geomorphology 2013;200:132–42.

Ollero A. Channel changes and floodplain management in the meandering middle EbroRiver, Spain. Geomorphology 2010;117:247–60.

Ollero A, Ballarín D, Díaz E, Mora D, Sánchez M, Acín V, et al. IHG: Un índice para lavaloración hidrogeomorfológica de sistemas fluviales. Limnética 2008;27:171–87.

Petts GE. A perspective on the abiotic processes sustaining the ecological integrity of runningwaters. Hydrobiologia 2000;422–423:15–27.

Petts GE, Gurnell AM. Dams and geomorphology: research progress and future directions.Geomorphology 2005;71:27–47.

Provansal M, Dufour S, Sabatier F, Anthony EJ, Raccasi G, Robresco S. The geomorphicevolution and sediment balance of the lower Rhône River (southern France) overthe last 130 years: hydropower dams versus other control factors. Geomorphology2014;219:27–41.

Raven PJ, Fox PJA, Everard M, Holmes NTH, Dawson FH. River habitat survey: a new systemfor classifying rivers according to their habitat quality. In: Boon PJ, Howell DL, editors.Freshwater quality: defining the indefinable? Edinburgh: The Stationery Office; 1997.p. 215–34.

Rice S, Church M. Grain size along two gravel-bed rivers: statistical variation, spatial patternand sedimentary links. Earth Surf Process Landf 1998;23:345–63.

Richter BD, Baumgartner JV, Powell J, Braun DP. Amethod for assessing hydrologic alterationwithin ecosystems. Conserv Biol 1996;10:1163–74.

Rinaldi M, Surian N, Comiti F, Bussettini M. A method for the assessment and analysis ofthe hydromorphological condition of Italian streams: the Morphological QualityIndex (MQI). Geomorphology 2013;180–181:96–108.

Sharitz RR,MitschWJ. Southern floodplain forests. In:MartinWH, Boyce SG, Echternacht AC,editors. Biodiversity of the southeastern United States: lowland terrestrial communities.New York, NY: John Wiley & Sons; Inc.; 1993. p. 311–72.

Surian N. Channel changes due to river regulation: the case of the Piave River, Italy. EarthSurf Process Landf 1999;24:1135–51.

Vericat D, Batalla RJ. Sediment transport in a large impounded river: the lower Ebro, NEIberian Peninsula. Geomorphology 2006;79(1–2):72–92.

Vericat D, Batalla RJ, Garcia C. Breakup and reestablishment of the armour layer in a highlyregulated large gravel-bed river: the lower Ebro. Geomorphology 2006;76:122–36.

Walter C, Tullos DD. Downstream channel changes after a small dam removal: using aerialphotos and measurement error for context; Calapooia River, Oregon. River Res Appl2010;26:1220–45.

Williams GP, Wolman MG. Downstream effects of dams on Alluvial Rivers. US geologicalsurvey professional paper 1286. Washington, DC: USGS; 1984.

Wolman MG. A method of sampling coarse river-bed material. Am Geophys Union Trans1954;35:951–6.

Wyzga B, Zawiejska J, Radecki-Pawlik A, Hajdukiewicz H. Environmental change,hydromorphological reference conditions and the restoration of Polish Carpathianrivers. Earth Surf Landf 2012;37:1213–26.

Xoplaki E, González-Rouco JF, Luterbacher J, Wanner H. Wet season Mediterraneanprecipitation variability, influence of large-scale dynamics and trends. ClimateDynam 2004;23:63–78.









































































































0

50

100

150

200

250

6:00 12:00 18:00

Turb

idit

y (N

TU)

Date

Geomorphic characteristics and dynamics are key factors controlling the microhabitat conditions which are key to maintaining the ecological diversity of a particular fluvial system.

The maximum ecological diversity of a system is reached when the heterogeneity of physical

habitat conditions (including flow velocity, substrate, etc) is maximised.

Channel geometry and bed composition adjust toward a new configuration when water and sediment supply are changed (Wilcock, 1998: Science 280: 410-412)

Physical stressors (e.g. floods) change

over temporal and spatial scales, and determine the disturbance regime (frequency and magnitude) experienced by a given reach and, consequently, the associated ecological diversity.

Disturbance is considered the main factor affecting the organization of riverine communities and contributes to key ecological processes (Yount and Niemi, 1990: Env. Man. 14(5): 547-569).

Human-induced alterations, such as gravel mining (to which all subsequent discussion relates), represent an additional, non-natural, stressor which modifies physical and ecological processes and dynamics.

MorphSed (www.morphsed.es) is analysing and comparing physical (i.e. geomorphic) and ecological processes and their spatio-temporal dynamics in unaltered sections and mined reaches. The project is developed in a 13-km long reach of the Upper River Cinca (Central Pyrenees) with a strong fieldwork component to link: (a) flow htdraulics, (b) channel morphology, (c) bed sedimentology, (d) Sediment transport, (e) ecology, and (f) sediment management (Figure 1).

A total of 4 specific objectives will provide the basis to answer main research questions (Figure 2 and 3).

COUPLING CHANNEL MORPHOLOGY AND ECOLOGICAL DIVERSITY IN A GRAVEL BED RIVER: MORPHSED CONCEPTUAL APPROACH AND

EXPERIMENTAL DESIGN

1. MORPHSED: PROJECT RATIONALE 3. DATA STRUCTURE

D. Vericat 1,2,12, R.J Batalla 1,2,3, C.N. Gibbins 4, J. Brasington, J.5, A. Tena1, M. Béjar1, E. Muñoz-Narciso1, E. Ramos1, G. Lobera1, C.

Buendia3,1, J.A. López-Tarazón 6,1, M. Smith7, J. Wheaton8, R. López1, J. Verdú1 and A. Palau9, 1

©2014 (All rights reserved)

2. AIM, OBJECTIVES, AND RESEARCH QUESTIONS

For further information on the project contact: Dr. Damia Vericat, [email protected]

Data structure is divided in 4 main blocs (Figure 4). Each bloc has a series of key variables and specific methods and techniques to obtain and post-process field data, and elaborate products.

www.morphsed.es

Acknowledgements

4. GRAPHICAL PRELIMINARY OUTCOMES

MO

RP

HS

ED

This research is carrying out within the framework of a research project funded by the Spanish Ministry of Economy and Competiveness (CGL2012-36394). The sixth author has a PhD grant funded by the University of Lleida while the seventh has a contract supported by MophSed and the eight a PhD grant from the Spanish Ministry of Economy and Competiveness (BES-2013-067223). The first author is founded by a Ramon y Cajal Fellowship (RYC-2010-06264). Hydrological data is being supplied by the Ebro Water Authorities. We thank all scientific collaboration and support provided by Celso Garcia and Colin Rennie. We thank all support by the Ebro Water Authorities, the interest of ENDESA in terms of the methodology, and all logistics provided by Acciona. We also thank Girolibre for providing the platform to obtain the aerials.

1

2 3 4

5 6 7

8 9 10

Hydraulics

Morphology

Sedimentology

Sediment Transport

Ecology

Management

Figure 1. Linking field observations and modelling approaches for a better understanding of eco-geomorphological processes:

insights into sediment management strategies.

Objective 1

Study the spatial variability of morphological and sedimentological changes associated to flood magnitude and frequency

Objective 2

Assess the spatial range of gravel mining impacts on the morphosedimentary characteristics and on sediment transport dynamics

Objective 3

Assess the impacts of gravel extractions on benthic invertebrate communities

Objective 4

Determine the most suitable metrics for detecting and monitoring the impacts of gravel extraction on the ecological status of highly dynamic gravel-bed rivers and their recovery Figure 2. MorphSed specific objectives.

Eco

log

ica

l sta

tus*

Regime

State 1

Transient

State

Time

Variability

Average

Changed physical

state

Relaxation time

Gra

vel M

inin

g

Regime

State 2

HYPOTHESIS AND RESEARCH QUESTIONS

Figure 3. Hypothetical trajectory of ecological status* following gravel mining (modified from the original idea form Petts and Gurnell, 2005: Geom. 71(1-2): 27-47 about fluvial metamorphosis after dam

closure), and associated key research questions. *Note ecological status is a broad term here.

RQ1. What is the relaxation time associated to physical human-based pressures?

RQ2. What are the key controlling variables for

physically and ecologically-based predictors?

SURFACE MODELLING

• Aerials from a manned autogiro (at an altitude that guarantees a maximum of 5 cm pixel resolution photos) are taken after competent events (see workflow).

• Ground Control Points at a density that guarantees an accurate post-processing are set up in the field (in evaluation after a first complete flight using 220 GCPs)

• Point clouds from the aerials are obtained by means of the Structure from Motion approach (using Agisoft software).

• Wet channel bathymetry is obtained by means of optical bathymetric mapping.

SEDIMENT TRANSPORT

• Suspended sediment transport is monitored by means of two high range turbidity meters (Endress+Hauser Turbimax W CUS41) and one short range turbidity meter (maximum 3000NTU, Analite NEP9000). Probes will be installed in the three monitoring stations (see location in the aerials; planned in November 2013).

• Bedload transport will be inversely estimated by means of the morphological approach and based on the step lengths provided by paired erosion-sedimentation areas after competent events. Bedload transport will be partitioned across the study reach to assess the variability of bedload flux across the study reach.

• Bed disturbance during mining will be assessed by means of aDcp-GPS spatially distributed surveys. A filed-based calibration will be used to transform bed velocity measurements (aDcp bottom track position vs GPS position) to bedload transport rates.

FLOW AND HYDRAULICS

• Discharge is continuously recorded in the Escalona Gauging Station (supervised by the Ebro Water Authorities).

• Water stage will be recorded at the three monitored stations by means of Druk PDCR1730 pressure transducers (to be installed in Nov.).

• Flow gauging will be performed regularly in order to establish and maintain a d-Q relationship for the monitored stations.

• A time-lapse camera will be installed in the Ainsa and in the Escalona monitoring stations (LtR Arcon). Cameras will take pictures every 15 minutes in order to examine the areas flooded in each competent event (to be installed in Nov.).

ECOLOGICAL STATUS

• Macroinvertebrates surber samples are obtained in the 5 sampling sites after competent events. 5 replicates are obtained in each site. The replicates are weighted in relation to the extension of the morphological unites in each site. Each site has a length between 200 and 300 meters (a first campaign is performed, 14-16 October 2013). Animals are identified in the lab. Different indices (taxon and trait-based) will be calculated (in evaluation and discussion).

• An electrofishing campaign was performed at the 5 sampling sites in October 2013. The objective is to obtain fish data in terms of communities, structure of the population, densities and biomass.

Figure 4. MoprhSed data structure. Main data variables in each bloc are presented together with some details of the methods.

FLOODS

Benthic Samples Drift Samples Electrofishing

See Béjar et al. talk

Aerials Point Clouds Roughness

Topographic Change

See Múñoz-Narciso poster

Turbidity Water samples Bed mobility

Bedload

aDcp Gaugings Modelling

GRAVEL MINING

Flow Hydraulics

Sediment Transport

Morphology & Sedimentology

Ecological Status

CGL2012-36394

http://www.morphsed.es/






Què és l' · PDF filesedimentos, morfología del cauce, hábitat)...

Documents

Transcript of Què és l' · PDF filesedimentos, morfología del cauce, hábitat)...