PATRONES DE COLONIZACIÓN ... -...

PATRONES DE COLONIZACIÓN POSTINCENDIO

DE AVES DE HÁBITATS ABIERTOS

EN PAISAJES MEDITERRÁNEOS

Memoria presentada por

Elena López Zozaya

Para optar al grado de Doctora por la Universitat de Lleida

Director: Tutor:

Dr. Lluís Brotons Dr. Jose Antonio Bonet Grup d’Ecologia del Paisatge Departament de Producció Vegetal Àrea de Biodiversitat i Ciència Forestal Centre Tecnològic Forestal de Catalunya Universitat de Lleida

Diseño portada: Laia Espasa y Elena López Zozaya

Fotografías: Lluís Gustamante, Josep Rost, Josep Mª Puig (Jepi) y Marcel Gil

La batalla de la vida no siempre la gana el hombre más fuerte, o el más ligero,

porque tarde o temprano, el hombre que gana, es aquel que cree poder hacerlo

R. Kipling

AGRADECIMIENTOS/AGRAÏMENTS

La presente tesis se ha realizado en el marco del proyecto DINDIS (CGL2005-

00031) subvencionado por el Ministerio de Educación y Ciencia y mi trabajo durante estos cuatro años ha sido posible gracias a la beca predoctoral FPI. Además, gracias por concederme las ayudas del Programa Nacional de Formación de Recursos Humanos de Investigación para estancias fuera del centro receptor, con las que he podido conocer otros centros de investigación como el South African National Biodiversity Institute en Ciudad del Cabo y el Centro de Ecologia Aplicada Prof. Baeta Neves en Lisboa.

Gràcies al meu director de tesi, Lluís Brotons, per donar-me l’oportunitat de treballar en aquest projecte i sobretot, per la seva qualitat humana. Ha estat un plaer treballar amb ell. Gracias también a Santiago Saura por introducirme en el mundo de la teoría de grafos y los índices de conectividad. Muchas gracias a Jose Antonio Bonet por facilitarme los trámites con la Universitat de Lleida y a Francisco Moreira del Centro de Ecologia Aplicada Prof. Baeta Neves en Lisboa por su simpatía y por darme la posibilidad de trabajar en un sitio increíble.

També vull agrair al Centre Tecnòlogic Forestal de Catalunya el seu recolzament com a institució i a totes les persones que hi han treballat i hi treballen i amb els que he compartit molts moments inolvidables a Solsona. Gràcies especialment als companys de l’Àrea de Biodiversitat (David(s), Magda, Laura, Núria(s), Miguelito, Sara, Dani...). Gràcies Assu, Miquel i Adrián per la correcció de la tesi.

Gràcies a totes les persones que han participat als censos tots aquests anys, especialment gràcies a Lluís Gustamante per l’interès que ha mostrat en el projecte des del principi i el seu bon ”rollo” i a Marc Anton per l’ajuda prestada amb la creació de la base de dades. Gràcies també a totes les persones que van col·laborar voluntàriament en la realització del Atles d’ocells nidificants a Catalunya.

Vull agrair a la Meritxell, la Laia, la Montse, la Iria, la Franzoni i la Maria pels bons moments viscuts amb vosaltres; les escapades de cap de setmana, els sopars, els riures,... Gràcies al veí d’escala i company de feina, Miquel, per ajudar-me amb l’estadística i introduir-me al mon de R. Gràcies també al veí de plaça i una de les primeres persones que vaig conèixer a Solsona, Oscarils, per la seva alegria i energia.

Aquesta tesi no hauria estat possible sense en Toni. Per tot lo compartit durant aquests anys i la teva paciència sempre, però sobretot, durant els últims mesos de la tesi. Per la següent etapa que comencem a partir d’ara! Gracias a mis padres y hermanos/as por vuestro apoyo y vuestra ayuda constante durante tooooodos estos años. A Petri, Jose, Rakel y Javi por estar ahí siempre. A las amigas de Pamplona, Cris, Raquel, Aitana, Anne y Olatz y a Adriano porque sois incondicionales y siempre os tengo a mi lado. A todos vosotros y a los que me haya olvidado infinitas gracias!

ÍNDICE

1. RESUMENES .........................................................................................................................................1

1.1. RESUMEN...........................................................................................................................................1 1.2. RESUM...............................................................................................................................................2 1.3. SUMMARY .........................................................................................................................................3

2. INTRODUCCIÓN GENERAL .............................................................................................................5

2.1. LOS INCENDIOS EN LA CUENCA MEDITERRÁNEA ...............................................................................5 2.2. RESPUESTAS DE LA BIODIVERSIDAD A LOS INCENDIOS.......................................................................7 2.3. LAS AVES DE HÁBITATS ABIERTOS EN CATALUÑA .............................................................................8 2.4. CONECTIVIDAD DEL PAISAJE............................................................................................................11

3. OBJETIVOS DE LA TESIS................................................................................................................15

4. ARTÍCULOS ........................................................................................................................................16

4.1. ARTÍCULO I......................................................................................................................................17 4.2. ARTÍCULO II ....................................................................................................................................37 4.3. ARTÍCULO III...................................................................................................................................61 4.4. ARTÍCULO IV...................................................................................................................................83 4.5. ARTÍCULO V..................................................................................................................................107

5. DISCUSIÓN GENERAL ...................................................................................................................131

5.1. FACTORES QUE DETERMINAN LA COLONIZACIÓN POSTINCENDIO...................................................133 5.1.1. Patrón temporal ....................................................................................................................133 5.1.2. Patrón espacial .....................................................................................................................135

5.1.2.1. Factores que actúan a escala regional ............................................................................................ 136 5.1.2.2. Factores que actúan a escala local ................................................................................................. 137

5.2. PROCESOS HISTÓRICOS QUE AYUDAN A PREDECIR LAS NUEVAS COLONIZACIONES ........................139 5.3. APLICACIÓN DE LOS ÍNDICES DE CONECTIVIDAD A MODELOS DE DISTRIBUCIÓN DINÁMICOS .........141

6. CONCLUSIONES FINALES............................................................................................................143

7. REFERENCIAS BIBLIOGRÁFICAS..............................................................................................145

8. PUBLICACIONES.............................................................................................................................153

Resumen

1

1.1. Resumen

El fuego es una perturbación natural de los ecosistemas mediterráneos que durante miles de años ha contribuido a favorecer la dinámica paisajística. Algunos estudios sugieren que el régimen actual de incendios podría causar importantes cambios en la estructura de las comunidades. La capacidad de las especies para soportar o beneficiarse de perturbaciones depende de una serie de rasgos específicos que permiten la ocupación de las zonas afectadas. Un grupo de especies que parece beneficiarse de la presencia de zonas quemadas son las aves de hábitats abiertos, algunas de ellas con especial interés para la conservación. El objetivo de esta tesis ha sido identificar los factores determinantes de la colonización postincendio de las aves de hábitats abiertos. Para ello, se ha creado una base de datos con información avifaunística de todas las zonas quemadas en Cataluña (NE España) desde el año 2000 y se han utilizado datos del Atles d’ocells nidificants de Catalunya 1999-2002, imágenes satélite Landsat, parámetros topográficos y cartografía de los usos del suelo y del inventario forestal nacional. Además, para estudiar el patrón espacial a escala regional se han utilizado dos índices de conectividad: el Índice Integral de Conectividad (IIC) y la Probabilidad de Conectividad (PC). Por último, se ha utilizado información ornitológica recogida siete años después de un incendio que afectó un bosque de pino laricio (Pinus nigra) para analizar el efecto en la comunidad de aves de la variabilidad en el patrón de regeneración postincendio. Los resultados mostraron una gran heterogeneidad temporal y espacial en los patrones de colonización postincendio de las aves de hábitats abiertos. A escala temporal, nuestros resultados sugieren que procesos como la atracción de conespecíficos contribuyen a explicar el retraso que se produce entre que una zona quemada es potencialmente buena para la colonización de especies de hábitats abiertos y el momento en el que tiene lugar la colonización. A escala espacial, se ha confirmado la importancia en el proceso de colonización postincendio de factores que actúan a diferentes niveles (regional y local), como la conectividad y la calidad del hábitat de la zona quemada y, en menor medida, el tamaño del incendio. Además, los cambios en el paisaje inducidos por la escasa capacidad regenerativa del pino laricio después del fuego, conducen al mantenimiento a largo plazo del hábitat adecuado para las especies asociadas a hábitats abiertos. Por otra parte, se destaca la importancia del uso de índices de conectividad para predecir la respuesta en la distribución de estas especies a las perturbaciones. Por último, nuestros resultados sugieren que los cambios en los usos del suelo producidos en las últimas décadas del siglo XX, han provocado un cambio en los procesos ecológicos que actúan sobre los reservorios en las dinámicas de las especies de hábitats abiertos en la región mediterránea. Se ha pasado de un conjunto de hábitats abiertos relativamente estáticos (zonas agrícolas y prados) a un mosaico de hábitats donde los incendios juegan un doble papel; por una parte, los incendios crean el hábitat adecuado para la colonización de las especies de hábitats abiertos y por otra parte, nuestros resultados indican que las zonas quemadas actúan como hábitat fuente, proporcionando individuos a los nuevos hábitats que sucesivamente aparecen en el paisaje.

Resum

2

1.2. Resum

El foc és una pertorbació natural dels ecosistemes mediterranis que durant milers d’anys ha contribuït a afavorir la dinàmica paisatgística. Alguns estudis suggereixen que el règim actual d’incendis podria causar importants canvis en l’estructura de les comunitats. La capacitat de les espècies per suportar o beneficiar-se de pertorbacions depèn d’una sèrie de característiques específiques que permetin l’ocupació de les zones afectades. Un grup d’espècies que es beneficia de la presència de zones cremades és el de les aus d’hàbitats oberts, algunes amb especial interès per a la conservació. L’objectiu d’aquesta tesi ha estat la identificació dels factors determinants de la colonització postincendi de les aus d’hàbitats oberts. Per aconseguir-ho, s’ha creat una base de dades amb informació avifaunística de totes les zones cremades a Catalunya (NE Espanya) des de l’any 2000 i s’han utilitzat dades de l’Atles d’ocells nidificants de Catalunya 1999-2002, imatges satèl·lit Landsat, paràmetres topogràfics i cartografia dels usos del sòl i de l’inventari forestal nacional. A més a més, per estudiar el patró espacial a escala regional s’han utilitzat dos índexs de connectivitat: l’Índex Integral de Connectivitat (IIC) i la Probabilitat de Connectivitat (PC). Per acabar, s’ha utilitzat informació ornitològica recollida set anys després d’un incendi que va afectar un bosc de pinassa (Pinus nigra) per analitzar l’efecte en la comunitat d’aus de la variabilitat en el patró de regeneració postincendi. Els resultats han mostrat una gran heterogeneïtat temporal i espacial en els patrons de colonització postincendi de les aus d’hàbitats oberts. A escala temporal, els resultats suggereixen que processos com l’atracció d’individus conspecífics contribueixen a explicar el retard que es produeix des de que una zona cremada és potencialment bona per a la colonització d’espècies d’hàbitats oberts fins que té lloc realment la colonització. A escala espacial, s’ha confirmat la importància en el procés de colonització de factors que actuen en diferents nivells (regional i local), com la connectivitat i la qualitat de l’hàbitat de la zona cremada i, en menor mesura, la mida de l’incendi. A més a més, els canvis en el paisatge induïts per l’escassa capacitat regenerativa de la pinassa després del foc, condueixen al manteniment a llarg termini de l’hàbitat adequat per a les espècies d’aus associades a hàbitats oberts. D’altra banda, es destaca la importància de l’ús d’índexs de connectivitat per predir la resposta en la distribució d’aquestes espècies a les pertorbacions. Finalment, els resultats suggereixen que els canvis en els usos del sòl produïts en les últimes dècades del segle XX han provocat un canvi en els processos ecològics, els quals actuen sobre els reservoris en les dinàmiques de les espècies d’hàbitats oberts en la regió mediterrània. S’ha passat d’un conjunt d’hàbitats oberts relativament estàtics (zones agrícoles i pastures) a un mosaic d’hàbitats on els incendis tenen un doble paper; per una banda, els incendis creen l’hàbitat adequat per a la colonització de les espècies d’hàbitats oberts i, per l’altra, els nostres resultats indiquen que les zones cremades actuen com a hàbitat font, proporcionant individus als nous hàbitats que sucessivament apareixen en el paisatge.

Summary

3

1.3. Summary

Fire is a natural disturbance in Mediterranean ecosystems that has contributed to

favor landscape dynamics over millennia. Some studies suggest that the current fire regime may cause important shifts in community structure and composition. Species' capability to withstand or profit from disturbances, such as fires, depend on a series of specific traits allowing the effective occupation of burned areas. A group of species that seems to benefit from the occurrence of fires are bird species occupying open-habitats, some of which are threatened. The general objective of this thesis has been to identify factors determining colonisation process of open-habitat bird species after fire in Mediterranean landscapes. For this purpose, a database has been created with bird information on all burned areas in Catalonia (NE Spain) since 2000 and data from the Atles d’ocells nidificants de Catalonia 1999-2002, Landsat satellite images, topographic parameters and cartography of the land-uses and the Spanish National Forest Inventory has been used. In addition, in order to study the spatial pattern at a regional scale two connectivity indexes have been used: the Integral Index of Connectivity (IIC) and the Probability of Connectivity (PC). Lately, bird data collected seven years after a fire occurred in a Black Pine (Pinus nigra) forested area has been used to analyze the effect of variability in post-fire regeneration patterns on the bird community. Our results showed a large temporal and spatial heterogeneity in the post-fire colonization patterns of the open-habitat bird species. At a temporal scale, processes such as conspecifics attraction may explain the delay between when habitat is potentially adequate for species colonization and when colonization process really takes place. At a spatial scale, the results confirmed the importance in the post-fire colonisation of factors acting at different levels (regional and local), such as the landscape connectivity and the quality of the burnt area and, to a lesser extent, the fire size. In addition, landscape changes induced by the low regeneration capacity of the Black Pine after fire may lead to large temporal maintenance or increase of habitat suitable for species linked to open habitats. On the other hand, our results highlighted the importance of using connectivity indexes for predicting species distribution changes in response to disturbances. Finally, the results suggest land use changes in recent decades have produced a shift in the ecological processes acting in the reservoirs for open-habitat bird species dynamics in Mediterranean areas: from a more permanent habitat network constituted by relatively static open habitats (grassland and farmland) to a shifting mosaic of habitat patches where fires plays a double role; on one hand, fires create an adequate habitat for open-habitat bird colonisation and, on the other hand, our results indicate that the burned areas may act as source habitats, providing individuals to the new habitats that appear on the landscape.

Introducción general

5

2. INTRODUCCIÓN GENERAL

2.1. Los incendios en la cuenca Mediterránea

Los incendios forestales son una perturbación natural de los ecosistemas

mediterráneos que durante miles de años han contribuido a favorecer la dinámica

paisajística (Trabaud et al. 1993). Hay indicios de que el fuego fue utilizado ya en el

paleolítico para facilitar la caza y la recolección de vegetales, y hay claras evidencias

del uso del fuego en la cuenca mediterránea durante el neolítico (Naveh 1975). A partir

de estas fechas, en el mediterráneo han vivido diversas culturas, con diversas densidades

de población y variados usos del suelo (Pausas 2009). Estos cambios han modificado la

frecuencia y el grado de recurrencia de los incendios.

Algunos estudios sugieren que el régimen actual de incendios podría causar

importantes cambios en la estructura de las comunidades (Kazanis & Arianotsou 2004,

Rodrigo et al. 2004, Arnan et al. 2006). En las últimas décadas del siglo XX, la

frecuencia y la extensión de incendios han aumentado de manera exponencial debido

principalmente a los cambios de usos del suelo, el cambio climático y el aumento de la

población, siendo el primero probablemente el más importante (Moreno et al. 1998,

Pausas & Vallejo 1999, Pausas 2004a). Por ejemplo, en España, entre 1960 y 1973 la

superficie media anual quemada era de 50.000 ha (y el número de incendios < 2.000),

mientras que a partir de 1974, la superficie media anual quemada fue de 215.000 ha

(≈8.550 incendios) y en algunos años (1978, 1985, 1989, 1994) la superficie afectada

por incendios superó las 400.000 ha (Moreno et al. 1998, Pausas & Vallejo 1999). Este

incremento se ha observado a pesar del aumento paralelo que se ha realizado en el

esfuerzo de extinción. Sin embargo, la superficie afectada por incendios varía

notablemente entre años y se observa una cierta tendencia a la disminución en los

últimos años (Pausas 2004b). La gran variabilidad interanual de la superficie quemada

se relaciona principalmente con las características climáticas del año (Díaz-Delgado et

al. 2004), es decir, se observa que durante las últimas décadas hay más incendios cada

año, pero sólo en los años secos los incendios afectan a grandes superficies (Malamud

1998). Las consecuencias de los incendios que afectan grandes superficies son mucho

más importantes de lo que se esperaría por su número ya que son responsables de la

mayoría del área quemada (Moreno et al. 1998) (Foto 1). Entre 1968 y 1997, en España,


6

estos incendios sólo representaron un 0.8% del número total de incendios y sin embargo

supusieron más del 60% de la superficie total quemada.

Foto 1 Aspecto de una zona afectada por un incendio ocurrido en el año 2003 que quemó 1.800 ha de matorral en la Granja d’Escarp (Cataluña, NE España). En el mismo año, únicamente 3 de los 14 incendios forestales que afectaron más de 50 ha de bosque supusieron el 76% de la zona quemada en Cataluña.

Estudios recientes han observado que el aumento en el número de incendios ha

afectado la sensibilidad del bosque Mediterráneo al fuego (Peñuelas 1996, Piñol et al.

1998). Los incendios tienden a concentrarse espacialmente (Vázquez & Moreno 2001),

lo que implica una mayor recurrencia que puede acabar repercutiendo en la

regeneración, como se ha demostrado al analizar la recuperación de la cubierta vegetal

después de incendios repetidos (Díaz-Delgado et al. 2002). Por último, en los últimos

años se ha producido un cambio en las zonas impactadas por los incendios hacia áreas

no estrictamente mediterráneas y tradicionalmente menos afectadas por esta

perturbación (Espelta et al. 2002, Rodrigo et al. 2004, Pausas et al. 2009). A pesar de su

importancia, el efecto de estos cambios en la configuración del paisaje y sus

consecuencias en los diferentes componentes de la biodiversidad es todavía

desconocido.


7

2.2. Respuestas de la biodiversidad a los incendios

La contribución de las perturbaciones a la estructura y funcionalidad de los

sistemas biológicos es un paradigma de la teoría de la ecología en general y en

particular de la ecología del paisaje (Forman 1995). Las perturbaciones de origen

natural tienen un papel fundamental en el mantenimiento de la heterogeneidad de las

condiciones ambientales que los organismos experimentan a lo largo del tiempo y del

espacio (Pickett et al. 1989, Brawn et al. 2001). La capacidad de las especies para

soportar o beneficiarse de perturbaciones, como en el caso de los incendios, depende de

una serie de rasgos específicos que permiten una ocupación eficaz de la zona quemada.

En primer lugar, la capacidad de las plantas para rebrotar o la fidelidad al lugar en

organismos móviles podrían permitir que las especies permanecieran en el área después

del impacto del fuego. En segundo lugar, el potencial para reproducirse y la habilidad

para dispersarse permitirán a especies ausentes antes de la perturbación colonizar los

nuevos hábitats. Entender la capacidad de las especies para colonizar nuevos hábitats ha

sido un tema central en el desarrollo de la teoría de islas (McArthur & Wilson 1967) o

la teoría de las metapoblaciones (Hanski 1999). Estas teorías consideran la dinámica de

colonización y extinción como los determinantes de la distribución espacial de las

especies a grandes escalas (Purves et al. 2007).

En el caso de especies pioneras, asociadas a los primeros estadios de la sucesión

vegetal, como aquellos originados tras una perturbación, se asume que la dispersión les

permite la fácil colonización de estos hábitats, lo que se denomina “síndrome del

colonizador” (Platt & Connell 2003). Modelos teóricos sobre la evolución de la

capacidad de dispersión de organismos móviles, como las aves y los mamíferos, han

demostrado como un aumento en la variabilidad temporal ligada a la dinámica de las

perturbaciones favorece la dispersión (Platt & Connell 2003). Estos resultados apoyan

la idea de que las especies asociadas a estadios iniciales de la sucesión vegetal son

probablemente buenos colonizadores (Pickett et al. 1989). Sin embargo, estos modelos

también indican que una variabilidad amplia en el terreno espacial podría también estar

relacionada con la presencia de especies que tienen una capacidad de dispersión

reducida (Johst et al. 2002). Hasta la fecha, existen muy pocos estudios que analicen

hasta que punto el paisaje mosaico está limitando la colonización de nuevos hábitats

originados después de una perturbación, como es el caso de por ejemplo los incendios


8

que ocurren en la región mediterránea (Rodrigo et al. 2004, Kiss & Magnin 2006,

Brotons et al. 2008).

La mayoría de estudios que analizan el efecto del fuego se han limitado a

evaluar el proceso de recuperación de la comunidad a escala local (Brawn et al. 2001,

Herrando et al. 2002). Muchos han ignorado las limitaciones impuestas a este proceso

por el contexto paisajístico en el que ocurren (Foto 2). Dado que las perturbaciones son

a menudo uno de los factores más importantes que afectan a los cambios en la

disponibilidad de hábitat a nivel de paisaje, el enlace entre colonización, heterogeneidad

paisajística y dinámica de las perturbaciones es clave para entender los procesos que

lideran los patrones ecológicos que se observan en el sistema (Herrando & Brotons

2002, Johst et al. 2002). En el presente trabajo nos concentraremos en estadios iniciales

de la sucesión vegetal en paisajes mediterráneos espacio-temporalmente heterogéneos

originados por el impacto de incendios forestales. Se utilizarán las aves de hábitats

abiertos propias de sistemas mediterráneos como modelo de estudio.

Foto 2 Ejemplo del contexto paisajístico en el que se producen los incendios. A la izquierda, zona quemada próxima a una zona urbanizada. A la derecha, zona quemada que rodea una zona agrícola.

2.3. Las aves de hábitats abiertos en Cataluña

Cataluña, con una superficie aproximada de 32.091 km2, es una región

localizada en el noreste de España con una gran heterogeneidad de hábitats; desde áreas

montañosas en el Pirineo y numerosas cadenas interiores (con una altitud que alcanza

los 3.143 m) hasta una línea de costa a lo largo del mar Mediterráneo. El bosque y el

matorral suponen aproximadamente el 60% de su cobertura vegetal, mientras que las

zonas agrícolas conforman buena parte del resto del territorio. El plan general de


9

política forestal ha dividido el territorio en tres regiones bioclimáticas (DGMN 1994,

Burriel et al. 2000-2004) (Fig. 1). Cada región es ecológica y climáticamente

homogénea. Además, cada una de las regiones tiene una extensión parecida, pero un

porcentaje diferente de área forestal y ha sido afectada de manera diferente por el fuego

en los últimos años.

Fig. 1 Representación de Cataluña en el contexto europeo (en negro) y detalle de sus regiones bioclimáticas.

El clima dominante es el mediterráneo templado, con inviernos húmedos y

suaves y veranos secos y calurosos (Clavero et al. 1996), hecho que favorece la

ocurrencia de incendios. El régimen de incendios de las últimas dos décadas del siglo

XX ha tenido un gran impacto sobre el paisaje de Cataluña, afectando un total de

258.000 ha, de las cuales el 50% fueron bosques (principalmente coníferas, 86%) y 31%

matorrales (Díaz-Delgado et al. 2004). A pesar del gran impacto de los incendios sobre

los bosques de Cataluña su extensión no ha disminuido debido al avance de los bosques

y la sucesión secundaria tras el abandono rural. Por otra parte, durante el mismo periodo

el 12% de la superficie quemada había sufrido algún incendio en años anteriores, lo que

indica un alto grado de recurrencia. Además, los incendios que quemaron grandes

superficies supusieron la mayoría del área quemada. Durante el periodo 1988-1999 los

incendios más grandes fueron responsables de menos del 0,4% del total de incendios en

Cataluña pero contaron con más del 75% de la superficie quemadas (Peix 1999). El

cambio en el régimen de incendios en las últimas décadas es representativo del ocurrido

en el área mediterránea. Así, Cataluña ofrece un marco excelente para el estudio del

impacto del fuego sobre la biodiversidad en la región mediterránea.


10

Numerosos trabajos se han centrado en estudiar la respuesta de la comunidad de

aves a los incendios (Pons & Prodon 1996, Moreira et al. 2001, Herrando et al. 2002,

Jacquet & Prodon 2009). Estos estudios han revelado la presencia durante los primeros

años después del fuego de especies de hábitats abiertos con diferente grado de

especialización en hábitats rocosos, matorrales y zonas agrícolas (Fig. 2). Este punto es

especialmente importante porque muchas de estas especies forman parte de las especies

más amenazadas en Europa (Birdlife Internacional 2004). En los últimos años estas

especies han visto como sus poblaciones disminuían rápidamente, principalmente como

respuesta a la perdida y degradación del hábitat a causa de la intensificación agrícola,

del abandono rural y de la homogeneización del paisaje (Preiss et al. 1997, Sanderson et

al. 2005, Vepsalainen et al. 2005). Sin embargo, en la región mediterránea los incendios

parecen contrarrestar la perdida de estos hábitats proporcionando hábitat adecuado para

estas especies (Moreira & Russo 2007). Pons & Bas (2005) observaron que de las 22

especies de hábitats abiertos que utilizan zonas recientemente quemadas en la Península

Ibérica y Francia, 17 tenían un estatus de conservación desfavorable en Europa. Por lo

que respecta a las aves de hábitats abiertos en Cataluña, muchas de las especies han

experimentado una expansión a escala regional favorecida por los incendios (Estrada et

al. 2004, Brotons et al. 2008, Vallecillo et al. 2009).

Fig. 2 Ejemplo de algunas de las especies que ocupan zonas recientemente quemadas. De izquierda a derecha, la cogujada montesina (Galerida theklae), el escribano hortelano (Emberiza hortulana) y la curruca rabilarga (Sylvia undata). Dibujos cedidos por el Institut Català d’Ornitologia (ICO), http://www.ornitologia.org.

A pesar de que es conocida la ocurrencia de estas especies durante los primeros

años después del fuego (Pons & Prodon 1996, Herrando et al. 2002), los factores que

determinan la capacidad de las especies de hábitats abiertos a colonizar las zonas

quemadas son poco conocidos. Algunos estudios han demostrado que el tipo de

respuesta que presentan estas especies ante los incendios es variable según el contexto

paisajístico y las limitaciones impuestas por la dispersión (Brotons et al. 2005, Pons &


11

Bas 2005). En este contexto, la conectividad con poblaciones fuente será un factor clave

para entender el patrón espacial en el proceso de colonización, especialmente a escala

regional (Hanski et al. 1999).

2.4. Conectividad del paisaje

La clásica teoría de metapoblaciones proporciona un marco teórico adecuado

para el estudio del proceso de colonización (Hanski 1999). Según esta teoría, una

población está compuesta por subpoblaciones espacialmente separadas pero conectadas

por dispersión. Estrechamente relacionado con esta teoría, los sistemas fuente-sumidero

(source-sink, en inglés) son muy útiles para explicar los diferentes gradientes

poblacionales de una determinada especie (Pulliam 1988). De esta forma, un hábitat

fuente (source) es aquel que alberga una población abundante y productiva en la que la

natalidad excede la mortalidad y la emigración excede a la inmigración, mientras que un

hábitat sumidero (sink) alberga una población con un balance negativo entre la natalidad

y la mortalidad (la producción de nuevos individuos no es suficiente para compensar la

mortalidad adulta), estando avocada esta última a la extinción en ausencia de

inmigración (Pulliam 1988). Ambas teorías subrayan la importancia en el proceso de

colonización de la conectividad del paisaje para facilitar el intercambio de flujos

ecológicos a través del territorio (Taylor et al. 1993). En este sentido, un paisaje se

define como una porción heterogénea del territorio compuesta por un mosaico de

cubiertas y usos del suelo (teselas) que interaccionan entre sí (Forman & Godron 1986).

En este trabajo, una tesela corresponde a una porción de hábitat en la cual las aves de

hábitats abiertos encuentran todos los recursos necesarios, bióticos y abióticos, para su

supervivencia y reproducción. Estos hábitats se caracterizan por zonas rocosas con

escasa vegetación, matorrales, áreas de cultivo de secano y/o prados.

Asimismo, un paisaje puede ser representado a través de estructuras de grafos.

Un grafo es un par ordenado ),( ENG = donde N es un conjunto de nodos y E es un

conjunto de enlaces que relacionan estos nodos (Harary 1969). En este contexto, se

podría hacer una analogía (grafo-paisaje) en la que un paisaje (grafo) vendría

representado por un grupo de teselas de hábitats (nodos) conectadas o no (mediante

enlaces) entre sí (ver figura 3). Los nodos corresponden a teselas de hábitat rodeadas de

terreno inhóspito (matriz de no-hábitat), mientras que los enlaces representan la


12

conexión funcional entre dos nodos, es decir, la capacidad potencial de un organismo

para dispersarse directamente entre dos teselas de hábitat. La existencia o no de enlaces

entre cada par de teselas se determina en función de las distancias existentes entre las

teselas y la capacidad de dispersión de la especie.

Fig. 3 Representación de un paisaje (grafo) con 5 nodos y 4 enlaces. El nodo (i) representa la tesela de hábitat que se origina tras un incendio y los nodos j, k, l y m son teselas de hábitat presentes en el paisaje. Los individuos de una especie x que se encuentren en los nodos j, k i l podrán colonizar la nueva tesela de hábitat que se crea como consecuencia del fuego (i). Sin embargo, los individuos del nodo m no podrán colonizar la nueva tesela de hábitat (i) porque la distancia entre estos dos nodos es superior a la distancia máxima de dispersión de la especie x. Los movimientos de dispersión se producen cuando los individuos juveniles

dejan su lugar de nacimiento en busca de un territorio para reproducirse (natal

dispersal, en inglés) o cuando los individuos adultos abandonan su último lugar de

reproducción hasta encontrar uno nuevo donde volver a reproducirse (breeding

dispersal, en inglés) (Greenwood & Harvey 1982, Paradis et al. 1998). El movimiento

de los organismos en el territorio es un fenómeno de gran importancia en la dinámica

poblacional de las especies porque afecta los patrones de distribución, la abundancia, las

dinámicas de colonización y extinción, la estructura de la población y el flujo genético

(MacArthur & Wilson 1967, Hanski 1999, Clobert et al. 2001, Bullok et al. 2002). A

pesar de su importancia, la información disponible sobre la capacidad de dispersión de

las especies es muy escasa debido a la dificultad que supone el seguimiento de los

individuos a largas distancias (Wiens 2001). En el caso de las aves de hábitats abiertos,

únicamente existe información precisa sobre la distancia de dispersión de una especie,

el escribano hortelano (Dale et al. 2005). En su estudio, Dale et al. (2005) anillaron

todos los individuos machos (n=301) en todas las zonas de cría conocidas en Noruega y

durante 7 años siguieron los movimientos de dispersión de estos individuos, obteniendo

i

l k

j

m

enlace

nodo


13

una distancia mediana de dispersión de aproximadamente 12 km. En el presente trabajo,

se utiliza esta distancia como la referencia a la capacidad de dispersión de las aves de

hábitats abiertos, asumiendo que la distancia de dispersión de este grupo de especies es

relativamente similar (considerando el papel funcional, el comportamiento y la masa

corporal de las especies).

Por otra parte, la modelización de un paisaje en forma de grafo permite utilizar

operaciones y algoritmos propios de la teoría de grafos para obtener índices de

conectividad ecológicos. En el presente trabajo, se utilizan dos índices de conectividad:

el Índice Integral de Conectividad (IIC) (Pascual-Hortal & Saura 2006) y la

Probabilidad de Conectividad (PC) (Saura & Pascual-Hortal 2007) implementados en el

software Conefor Sensinode (CS) (Saura & Torné 2009, http://www.conefor.org). La

diferencia entre los dos índices es que el IIC se basa en un modelo de conexión binario

(los nodos están o no conectados), mientras que el PC se basa en un modelo

probabilístico (hay una cierta probabilidad de conexión entre dos nodos, pij).

Estos índices están basados en el concepto de disponibilidad de hábitat, que

integran tanto la topología del hábitat (conectividad estricta y relaciones espaciales)

como ciertos atributos de las teselas (área, calidad del hábitat, abundancia de las

poblaciones de las especies analizadas,…) (Pascual-Hortal & Saura 2006), es decir,

tienen en cuenta no sólo el número de teselas alcanzables desde cualquier otra (grado de

conectividad) sino la calidad de la tesela. Una de las ventajas de estos índices respecto a

otros índices de conectividad es la posibilidad de calcular la importancia relativa de

cada una de las teselas (dIIC o dPC, según se utilice el índice IIC o PC

respectivamente). Esta importancia individual de tesela se obtiene calculando el índice

de conectividad para todo el paisaje completo y observando la diferencia en el valor del

índice al recalcularlo nuevamente para un paisaje en el que hubiera desaparecido una de

las teselas (Saura & Pascual-Hortal 2007, Saura & Rubio 2010). Además, el valor de la

importancia relativa de cada una de las teselas se puede dividir en tres términos de

acuerdo con las diferentes maneras en que una tesela de hábitat puede afectar la

conectividad y disponibilidad de hábitat en el paisaje: 1) conectividad dentro de la tesela

(intra-patch connectivity, en inglés); 2) cantidad de flujo de dispersión recibido por las

otras teselas que forman parte del paisaje; 3) la contribución de la tesela a la

conectividad entre otras teselas de hábitat como un elemento conector o stepping stones

(Saura 2008, Saura & Rubio 2010).


14

En este trabajo, se utiliza este segundo término para calcular la cantidad de flujo

de dispersión que recibe cada una de las nuevas teselas de hábitat que aparecen en el

paisaje como consecuencia de los incendios. Este término se calcula de la siguiente

manera:

∑≠ +

=n

ji ij

ji

fluxi nl

aadIIC

1 (1)

*ij

n

jiji

fluxi paadPC ∑

≠= (2)

donde ai es el valor del atributo del nodo que aparece en el paisaje, es decir, del

incendio. En este estudio, este valor se considera igual a 1 para asegurarse la

independencia entre la estimación y el valor del atributo del incendio. Por lo tanto, la

cantidad de flujo estimada depende exclusivamente de las fuentes de colonizadores y los

límites impuestos por la dispersión. aj es el valor del atributo en cada una de las otras

teselas de hábitat. nlij es el número mínimo de enlaces necesarios para alcanzar el nodo i

desde el nodo j. *ijp es definido como el producto máximo de probabilidades entre las

teselas i y j. El producto máximo de probabilidades de un camino entre las teselas i y j

es aquel en que pasando sólo una vez por cada tesela del camino, se obtiene el máximo

producto de las probabilidades de dispersión directa entre teselas (pij) de todos los

caminos posibles entre aquellas dos teselas. *ijp es calculado mediante algoritmos de

grafos.

Para caracterizar el grado de conexión entre cada par de teselas se realiza una

modelización de las probabilidades de dispersión (pij) ajustando la probabilidad de

dispersión estimada por la distancia mediana de dispersión de la especie (fijada en

p=0.5) a una función exponencial negativa de la distancia entre teselas (e.g. Bunn et al.

2000, Urban & Keitt 2001). En este trabajo se considera que la probabilidad de

dispersión entre el nodo i y el nodo j es la misma que la probabilidad de dispersión entre

el nodo j y el nodo i. Además se utiliza la distancia euclídea (línea recta) entre los pares

de nodos.

Esta nueva aproximación nos permitirá estudiar el patrón espacial en el proceso

de colonización postincendio a una escala regional. Para ello se utilizarán diferentes

fuentes de información disponibles a nivel de Cataluña como el Atles dels Ocells

Nidificants de Catalunya 1999-2002 (Estrada et al. 2004) (ver artículos 2 y 3) y el mapa

de usos del suelo de Cataluña del 1997 (ver artículo 5).

Objetivos de la tesis

15

3. OBJETIVOS DE LA TESIS

El objetivo general de esta tesis ha sido identificar los factores determinantes de

la colonización postincendio de las aves de hábitats abiertos. Este objetivo general se

divide en los siguientes objetivos más específicos:

1. Desarrollar una base de datos adecuada para el análisis de la dinámica espacio-

temporal de las aves de hábitats abiertos en zonas afectadas por el fuego.

2. Estudiar el efecto del tamaño del incendio, la conectividad del paisaje y el

tiempo transcurrido desde el impacto del fuego en el proceso de colonización

postincendio del escribano hortelano (Emberiza hortulana).

3. Determinar la importancia relativa de la calidad del hábitat, el tamaño del

incendio y la conectividad del paisaje en la colonización postincendio de las

aves de hábitats abiertos.

4. Analizar el efecto en la comunidad de aves de la variabilidad en el patrón de

regeneración postincendio. Evaluar el impacto del cambio del paisaje en el valor

de conservación de las especies.

5. Profundizar sobre el impacto de las trayectorias históricas del paisaje y el papel

de las perturbaciones en la colonización postincendio de las aves de hábitats

abiertos.

4. ARTÍCULOS

17

4.1. Artículo I

I- MONITORING SPATIAL AND TEMPORAL DYNAMICS OF BIRD

COMMUNITIES IN MEDITERRANEAN LANDSCAPES AFFECTED

BY LARGE WILDFIRES

The desolate appearance of burnt areas provokes fear and panic in most people… However, fires are a natural component of the dynamics of most ecosystems,

and more so in the Mediterranean than in many other regions. (Blondel and Aronson 1999, p. 143.)

Artículo I

18

SUMMARY

Aims: To present a bird-monitoring database in Mediterranean landscapes affected by

wildfires and evaluate: 1) the spatial and temporal variability in the bird community

composition and 2) the influence of pre-fire habitat configuration in the composition of

bird communities.

Location: Catalonia (NE Spain)

Methods: The DINDIS database results from the monitoring of bird communities

occupying all areas affected by large wildfires in Catalonia since 2000. We used bird

surveys conducted from 2006 to 2009 and performed a principal components analysis to

describe the main gradients of variation in the composition of bird communities, which

were used as descriptors of bird communities in subsequent analyses. We then analysed

the relationships of these community descriptors with bioclimatic regions within

Catalonia, time since fire and pre-fire vegetation (forest or shrubland).

Results: We have conducted 1918 bird surveys in 567 transects distributed in 56 burnt

areas. Eight out of the twenty most common detected species have an unfavourable

conservation status, most of them being associated to open-habitats. Both bird

communities’ descriptors had a strong regional component and were related to pre-fire

vegetation, and to a lesser extent to the time since fire.

Conclusions: The responses of bird communities to wildfires are heterogeneous,

complex and context dependent. Large-scale monitoring datasets, such as DINDIS,

might allow identifying factors acting at different spatial and temporal scales that affect

the dynamics of species and communities, giving additional information on the causes

under general trends observed using other monitoring systems.

Keywords: dispersion, landscape context, open-habitat species, pre-fire vegetation,

recently burnt areas, species colonisation

Artículo I

19

INTRODUCTION

Humans have shaped the Mediterranean Basin during millennia through forest

exploitation, raising of livestock and agriculture (Blondel and Aronson, 1999), using

fire as a main management tool. The resulting agro-sylvo-pastoral systems created upon

a fine mosaic of interconnected habitats, resulting in a high global biodiversity (Covas

and Blondel, 1998; MacDonald et al., 2000; Fabbio et al., 2003). Nevertheless, this

biodiversity is highly threatened due to climatic changes, biological invasions and

especially due to widespread land-use changes (Sala et al., 2000).

Significant changes in the socioeconomic structure, during the second half of the

last century, have induced extensive modifications in the composition and nature of

prevailing landscapes in the European part of the Mediterranean Basin (Blondel and

Aronson, 1999). The abandonment of traditional activities has lead to a progressive loss

of open habitats and to an increase in forests and dense shrublands through secondary

succession (Preiss et al., 1997). At the same time, several economic incentives for

afforestation of old fields (e.g. Fry, 1989; Stoate et al., 2001) offered farmers the option

to obtain some economic profit from their lands, promoting an increased of forest cover.

Due to the large and continuous fuel accumulation, the number and intensity of

wildfires have exponentially increased (Moreno et al., 1998; Pausas and Vallejo, 1999;

Pausas, 2004). Additionally, climatic changes seem to contribute to the current fire

pattern, since the higher frequency of dry summers has been related to the higher

amount of burnt area (Piñol et al., 1998; Pausas, 2004). Thus, with more fires occurring

in the area, the importance of fire disturbance in determining landscape structure and

explaining species dynamics on Mediterranean ecosystems is increasing.

Wildfires affect habitats with different vegetation covers and structure, which

after the disturbance converge towards simpler and open habitat configuration. These

new open areas are pioneered by bird species preferring open and low vegetation

habitats (Prodon et al., 1984). Together with species showing site fidelity after fire,

species of open habitats form the bulk of post-disturbance avian communities associated

with fire dynamics (Pons and Prodon, 1996; Herrando et al., 2002; Ukmar et al., 2007).

Although the patterns of post-fire change in bird community composition are well

known, especially in Mediterranean systems (Lawrence, 1966; Prodon and Lebreton,

1981; Jacquet and Prodon, 2009), relatively little research has been directed at spatial

analysis of fire effects; such as the role of unburnt patches, the relationship between fire

extent and direct impacts, and the process of species colonisation (Stuart-Smith et al.,

Artículo I

20

2002). The mechanisms that determine the pool of pioneer species that colonise open

habitat originated by fire disturbances are specially unknown.

In this work, we present a bird-monitoring database (DINDIS-Bird distribution

dynamics in Mediterranean landscapes affected by fires) aimed at the large-scale

monitoring of bird communities in burnt habitats. This database started in 2006 and

includes the sampling of all areas affected by large wildfires after 2000 in Catalonia, a

region dominated by Mediterranean climate and located in the north-eastern corner of

Spain (Fig. 1). The monitoring arises as an extension and improvement of a previous

large-scale study conducted from 2002 to 2005 (Brotons et al., 2005; Pons and Bas,

2005). Here, we used data gathered by DINDIS from 2006 to 2009 in order to describe

spatial and temporal patterns in the composition and structure of bird communities

occupying burnt areas in relation to bioclimatic regions within Catalonia and time since

fire. Finally, we evaluated the impact of pre-fire vegetation (forest or shrubland) in the

bird community composition.

MATERIAL AND METHODS

The DINDIS database

The DINDIS database aims at monitoring bird communities in wildfires

occurring in Catalonia. Most fires in the region are severe, including crown fires

strongly affecting both, forest canopy and undergrowth, and causing widespread tree

mortality (Rodrigo et al., 2004). The project started in 2006 and monitors all wildfires

occurring from 2000 in which more than 50ha of forest and/or shrubland were affected

by fire.

Within each fire perimeter, we established a series of line transects in order to

estimate bird presence and abundance (Bibby et al., 2000). Each survey lasted 15 min

and was divided in 3 sections of 5 minutes each. Transects were characterised by four

points collected with a GPS at minute 0, 5, 10 and 15. Birds were counted, when heard

or seen, and were allocated into one of the four distance bands (0-25 m, 25-50 m, 50-

100 m, >100m). Surveys were conducted once every breeding season (10th May–15th

June) in good weather conditions (i.e. without rainfall or strong wind) during the first 3

hours after sunrise by experienced ornithologists at a speed of about 2 km/h (Bibby et

al., 2000).

The number of transects conducted at each wildfire increased non-linearly with

burnt area (logarithmically; NumberTran sec ts =14.8 * Log10BurntArea(ha) − 24.6;

Artículo I

21

R2=0.97). The establishment of transectes observed the following criteria; a) they were

conducted entirely across burnt wildland avoiding, when possible, unburnt patches

(forest or farmland) and fire edges, b) the minimum distance between two transects was

150 m, c) the minimum distance between transects and fire edge was 50 m , d) transects

were conducted preferably on existing trails in order to allow future repetition of the

transects after vegetation recovery and e) in the largest burnt areas, transects are

distributed in a number of representative locations covering habitat heterogeneity within

fire perimeter. The location of transects was maintained in consecutive years, although

in few case this was not possible (e.g. due to vegetation growth).

We assigned each fire to a bioclimatic region to account for potential effects of

spatial variation in bird community composition. With this aim, we defined three

different regions in Catalonia (“South”, “Northwest” and “Northeast”) differing broadly

in climatic patterns and dominant forest species (derived from DGMN 1994 and Burriel

et al., 2000-2004) (see Fig. 1, Appendix 1).

Data analysis

The present database includes data collected from 1918 bird surveys performed

at 567 transects established in 56 burnt areas. We used the bird surveys to search for

spatial and temporal patterns in the composition of bird communities occupying

recently burnt areas. We selected individuals detected within 100 m belts (the three

closest census bands) on both sides of the track and used the presence and absence of

those species detected in more than 5% of the censuses to construct the final matrix

(Table 1). First, we performed a principal components analysis (PCA) using each

individual bird census as the sampling unit in order to summarize the main gradients of

variation in the composition of bird communities. The first two principal components

(PCs) extracted by the PCA were interpreted as a function of the loadings and

ecological requirements of bird species. Values of censuses along these gradients were

kept as descriptors of bird communities.

For further analyses, the sampling unit was every combination of wildfire and

census year (wildfire/year), instead of individual censuses. A wildfire/year sample was

the result of calculating the mean value of the bird community descriptors (PCA1 and

PCA2) of all transects conducted within a fire in a particular census year. We performed

an analysis of variance (ANOVA) and a simple regression to analyze variations in bird

community descriptors in relation to the bioclimatic regions and the number of years

after fire respectively.

Artículo I

22

Finally, we analysed whether the characteristics of pre-fire vegetation influenced

the bird community descriptors. As shown in other studies, pre-fire vegetation

commonly constrains post-fire bird species composition (Herrando et al., 2002). With

this aim, we selected forest and shrubland fires, defining them as those in which more

than 70% of the burnt area was covered by any of the two habitat types before wildfire

occurrence. Then, we ran one-way ANOVAs testing the influence of pre-fire vegetation

(two levels: “forest” and “shrubland”) on bird community descriptors. All statistical

analyses were run with Statistica V 7, StatSoft, Inc. 1984-2004.

Table 1 List of the bird species detected in more than 5% of the bird surveys (N= 1918) and their frequency of occurrence (in %).

Scientific name Common name Acronym Occurrence

Sylvia melanocephala Sardinian Warbler SYLALA 61.4 Serinus serinus Serin SERSER 38.1 Sylvia undata Dartford Warbler SYLUND 34.1 Lullula arborea Woodlark LULARB 32.6 Sylvia cantillans Subalpine Warbler SYLCAN 30.9 Turdus merula Blackbird TURMER 29.9 Carduelis cannabina Linnet CARINA 29.4 Saxicola torquatus Stonechat SAXTOR 28.7 Emberiza calandra Corn Bunting MILCAL 24.8 Parus major Great Tit PARMAJ 24.6 Hippolais polyglotta Melodious Warbler HIPPOL 23.4 Oenanthe hispanica Black-eared Wheatear OENHIS 23.4 Carduelis carduelis Goldfinch CARLIS 22.9 Luscinia megarhynchos Nightingale LUSMEG 21.6 Galerida theklae Thekla Lark GALTHE 19.0 Carduelis chloris Greenfinch CARCHL 17.9 Emberiza cia Rock Bunting EMBCIA 16.9 Lanius senator Woodchat Shrike LANSEN 15.7 Troglodytes troglodytes Wren TROTRO 15.6 Columba palumbus Word Pigeon COLPAL 14.5 Passer domesticus House sparrow PASDOM 14.3 Fringilla coelebs Chaffinch FRICOE 12.8 Cyanistes caeruleus Blue Tit PARCAE 10.9 Lophophanes cristatus Crested Tit PARCRI 10.6 Emberiza cirlus Cirl Bunting EMBCIR 10.5 Alectoris rufa Red-legged Partridge ALERUF 10.4 Certhia brachydactyla Short-toed Tree-creeper CERBRA 8.7 Garrulus glandarius Jay GARGLA 8.7

Artículo I

23

Table 1 continued

Scientific name Common name Acronym Occurrence Anthus campestris Tawny Pipit ANTCAM 8.6 Emberiza hortulana Ortolan Bunting EMBHOR 8.2 Phylloscopus bonelli Bonelli’s Warbler PHYBON 7.7 Streptopelia turtur Turtle Dove STRTUR 7.4 Dendrocopos major Great Spotted DENMAJ 6.2 Petronia petronia Rock Sparrow PETPET 5.9 Sturnus vulgaris Starling STUVUL 5.9 Erithacus rubecula European Robin ERIRUB 5.8 Oriolus oriolus Golden Oriole ORIORI 5.4 Sylvia atricapilla Blackcap SYLATR 5.2 Sylvia hortensis Orphean Warbler SYLHOR 5.2

RESULTS

Summary of the data from DINDIS

A total of 56 wildfires, widely distributed across Catalonia, have occurred from

2000 to 2007 (Fig. 1, Appendix 2). None of the fires that occurred in 2008 in the study

area fulfilled the conditions being sample (i.e. a minimum of 50 ha burnt area). In 2006,

2007, 2008 and 2009 we conducted bird surveys in 44, 51, 45 and 47 sites respectively.

In the present year, 2010, we expect to survey 41 sites (Appendix 3). According to

Catalonia land use maps of 1997 and 2002, 35% of the studied burnt areas (n=20) were

mostly covered by shrubland before the fire, 21% (n=12) were covered by forest, with

the rest of burnt sites (n=24) being a mixture of these two habitat types (Appendix 2).

Most of the fires that occurred in the “south” region were shrubland fires (13 out of 25),

most of those that occurred in the “northwest” region were forest fires (8 out of 15) and

the majority of the fires that occurred in the ”northeast” region were mixture fires (10

out of 16).

Artículo I

24

Fig. 1 Geographical location of fires occurring in Catalonia from 2000 to 2007. Numbers correspond to wildfire codes (see Appendix 2). The three bioclimatic regions (delimited by the black thin line) with the total number of fires within each one are shown. The map on the lower right corner shows the location of the study region (Catalonia), shown in black colour.

Data on bird communities was collected at different stages of successional, from

one to nine years after fire, thus allowing the analysis of the short-temporal responses of

birds to fire. Most of the data were concentrated in the first six years after fire

occurrence, since there was little information on sites where fire disturbance occurred

between seven and nine years (Fig. 2).

Fig. 2 Number of burned areas monitored according to the years elapsed since the last wildfire.

Artículo I

25

Fire size varied from 50 to 6,389 ha. The number of transects conducted within

fire perimeter went from 2 to 31 (Appendix 2). A total of 1918 bird surveys have been

conducted describing the temporal variation of 567 transects; with a mean transect

length of 440 m (SD 95 m, range 381-526 m).

Ninety-seven species have been detected with a total of 36,806 individuals

observed. Raptors, aerial feeders (swallows, swifts and bee-eaters) and crepuscular

species were excluded from the analyses (Bibby et al., 2000). Eight out of the 20 most

common species have an unfavourable conservation status (SPEC 1, 2 or 3); most of

them associated to open-habitat species, highlighting the role of fires for threatened

species.

General patterns in the observed species gradient

The first two principal components extracted by the PCA accounted for 16.2%

of the original variability in bird community composition (Fig. 3). PCA1 represented a

gradient of tree density, ranging from open-habitat and shrubland birds (forest-avoiding

species) to forest species. The second factor PCA2 represented a biogeographical

gradient separating Mediterranean species (positive values) from inland species

(negative values).

Fig. 3 Factors loadings bird species for the first two principal components extracted by the PCA (eigenvalues and variance explained is in the parenthesis). Complete names of the species acronyms are given in table 1.

PCA1 and PCA2 were significantly different among the three bioclimatic

regions, although the explained variance was greater when analyzing PCA2 (Table 2).

Burnt sites within the northwest region were characterized by inland species whereas

Artículo I

26

burnt areas within the northeast region were distinguished by Mediterranean species. At

the same time, fires within the south region had more forest-avoiding species than the

others regions (Fig. 4).

Fig. 4 The arrangement of fire sites within the south region (open diamond), the northwest region (filled squares) and the northeast region (red triangle) monitored during fieldwork in the two axes of the principal component analysis.

On the other hand, PCA1 and PCA2 were also significantly related to time since

fire (Table 2). Time since fire was negative related to PCA1, indicating a temporal shift

towards a community dominated by forest-avoiding species, and positive related to

PCA2, thus suggesting a community change towards relatively more Mediterranean

species. Nevertheless, the explained variance was very low for both axes (Table 2). This

result might suggest that, above to a certain level, the bird community response to fire at

short-temporal scale depend on other factors rather than the number of years elapsed

after fire.

Table 2 Effects of bioclimatic region (three levels, see Figure 1), the time since fire (continuous predictor) and pre-fire vegetation (two levels; “forest” and “shrub”) on bird community descriptors (PCA1 and PCA2).

Dependent variable Effect Type of analysis d.f. F-value p-value R2

Region One-way ANOVA 2 9.62 <0.01 0.09

PCA1 Time Since Fire Simple regression 1 8.60 <0.01 0.04

Pre-fire vegetation One-way ANOVA 1 43.43 <0.001 0.30

Region One-way ANOVA 2 45.49 <0.001 0.33

PCA2 Time Since Fire Simple regression 1 6.03 <0.05 0.03

Pre-fire vegetation One-way ANOVA 1 13.72 <0.001 0.12

Artículo I

27

The structure of pre-fire vegetation (i.e. whether it was a forest or shrubland) had

a significant effect on both PCA factors, although its effect was much stronger in the

case of PCA1 (Table 2). Burnt forest areas showed higher proportion of forest species

comparing to burnt shrubland areas (Fig. 5).

Fig. 5 The arrangement of forest fires (filled squares) and shrubland fires (open circles) monitored during fieldwork in the two axes of the principal component analysis.

DISCUSSION

Our results indicate a large temporal and spatial variability in the species

response to fire. Post-fire bird community varied with time elapsed since fire and in the

different bioclimatic regions, indicating that multiple factors acting at different spatial

and temporal scales determine species composition after fire. Jacquet and Prodon (2009)

studied during 28 years the post-fire vegetation and bird succession in a Mediterranean

oak woodland and showed that the most important factor determining post-fire bird

community in the long-term scale is vegetation recovery; first starting with open-habitat

species, then shrubland species and finally forest species. In our work, we focussed on

the first stage of the successional process, when habitat is characterized by low habitat

structure (herbaceous and shrubland vegetation) and bird community is most often

dominated by open-habitat species.

We found bird community composition to be strongly constrained by pre-fire

vegetation. Fires that occurred in shrubland habitats, mostly concentrated in the

southern region of Catalonia, contained more forest-avoiding species than fires affecting

forests. The strong site tenacity, philopatry and habitat tolerance of some forest species

and the persistence of standing dead trees might explain the presence of these species in

Artículo I

28

the first years following fire (Prodon et al., 1987; Pons and Prodon, 1996; Hutto, 2006).

Nevertheless, these species usually disappear during the following years and do not

reappear until the vegetation attains a woody appearance (Jacquet and Prodon, 2009).

This would explain the results obtained in this study where bird community

composition in forest burnt areas tended to shift towards more forest-avoiding species

with time since fire, in what would seem the opposite pattern to that expected from bird

responses to vegetation succession.

Additionally, fires sites within the northwest region had more inland species,

whereas fires within the northeast region hosted more Mediterranean species. This result

suggests that other factors acting at the regional scale determine the bird community

response to fire. The habitat context and the species dispersal capacity might explain

difference in the bird community at regional scale. Hence, Brotons et al., (2005) suggest

that open-habitat species colonisation of recently burnt areas relies on relatively short-

dispersal distance since they showed that only those fires located nearby open-habitat

species population sources might have the chance of being colonised by these species.

Our data suggests that these factors and its interactions may account for the strong

spatial variability in bird community responses to fire.

Finally, some of the most common species detected in the studied wildfires

during the study period were, as expected, associated to open-habitats. These species are

among the most threatened European bird species due to land abandonment and

agricultural intensification (BirdLife International, 2004), and Mediterranean burnt

areas seem to counteract these negative processes by providing suitable habitat for them

(Moreira and Russo, 2007). In this sense, this database represent a powerful tool to

identify ecological processes affecting open-habitat bird dynamics, giving additional

information on the causes under general trends observed using other monitoring

systems. For instance, a recent study has shown that the expansion in Catalonia of the

ortolan bunting Emberiza hortulana, a species that has suffered a great population

decline across Europe, is due to the occupancy of recently burnt areas (Brotons et al.,

2008). On the other hand, it is also relevant the abundance of the Dartford warbler

Sylvia undata. This species has been newly updated as near threatened (BirdLife

International, 2009) due to its population declines in Spain. Nevertheless, it was one of

the most common species in the studied sites (34% of occurrence), indicating a high

occupancy of burnt areas.

In conclusion, we have shown that the bird community response to fire is

heterogeneous and complex. The type of vegetation prior to the fire, and the subsequent

Artículo I

29

post-fire management, and the landscape context are essential features in order to

understand bird community composition after fire. These factors and their interactions,

there might be a lower probability of finding a population source in a forest burnt area

than in a shrubland burnt areas, probably determine the capacity of pioneer species,

mostly open-habitat bird species, to arrive to the burnt areas (Brotons et al., 2008).

Further studies are needed to complete and document the progression towards a

shrubland and forest species composition. Here, especial attention must be taken on the

current fire regime. Recurrent fires may reduce the sprouting capacity of many

Mediterranean plants; large wildfire may induce homogeneity and species that were

previously less affected by fire might be more prone to fire (Pausas et al., 2008). As a

result, plant succession may be hampered and open vegetation may become permanent.

In this context, the use of a large and an extensive bird-monitoring database including

common bird species in burnt areas may offer a reliable tool to investigate such

complex impacts.

ACKNOWLEDGEMENTS

We thank the ornithologists who participated in the bird surveys. This work has

received financial support from the projects Consolider-Ingenio Montes (CSD2008-

00040), CGL2008-05506-CO2 ⁄BOS and CGL2005-2000031⁄BOS granted by the

Spanish Ministry of Education and Science. E.L.Z. (FPI fellowship) received financial

support from the Spanish Ministry of Education and Science.

REFERENCES

Bibby, C.J., Burgess, N.D., Hill, D.A. and Mustoe, S.H. 2000. Bird Census Techniques.

Second Edition Academic Press, London.

BirdLife International. 2004. Birds in Europe: population estimates, trends and

conservation status. BirdLife International, Cambridge, UK (BirdLife Conservation

Series No. 12).

BirdLife International. 2009. Species factsheet: Sylvia undata. Downloaded from

http://www.birdlife.org on 17/5/2010

Blondel, J. and Aronson, J. (Eds) 1999. Biology and wildlife of the Mediterranean

region. Oxford University Press, Oxford.

Brotons, L., Pons, P. and Herrando, S. 2005. Colonization of dynamic Mediterranean

landscapes: where do birds come from after fire? Journal of biogeography, 32: 789-

798.

Artículo I

30

Brotons, L., Herrando, S. and Pons, P. 2008. Wildfires and the expansion of threatened

farmland birds: the Ortolan Bunting, Emberiza hortulana, in Mediterranean

landscapes. Journal of Applied Ecology, 45: 1059-1066.

Burriel, J.A., Gracia, C., Ibàñez, J.J., Mata, T. and Vayreda, J. 2000-2004. Inventari

Ecològic i Forestal de Catalunya. CREAF, Bellaterra. ISBN de la sèrie: 84-931323-

0-6.

DGMN (Direcció General del Medi Natural). 1994. Plà General de Política Forestal.

DARP: Departament d’Agricultura, Ramaderia i Pesca, Barcelona, Spain.

Covas, R., and Blondel, J. 1998. Biogeography and history of the Mediterranean bird

fauna. Ibis, 140: 395-407.

Fabbio, G., Merlo, M. and Tosi, V. 2003. Silvicultural management in maintaining

biodiversity and resistance of forests in Europe-the Mediterranean region. Journal of

Environmental Management, 67: 67-76.

Fry, G.L.A. 1989. Conservation in agricultural ecosystems. In Spellerberg, I.F.,

Goldsmith, F.B. and Morris, M.G. (Eds.). The scientific management of temperate

communities for conservation, pp. 415-443. Blackwell Science, Southampton.

Herrando, S., Brotons, L., del Amo, R. and Llacuna, S. 2002. Bird community

succession after fire in a dry Mediterranean shrubland. Ardea, 90: 303-310.

Hutto, R.L. 2006. Toward meaningful snag-management guidelines for postfire salvage

logging in North American conifer forests. Conservation Biology, 20: 984-993.

Jacquet, K. and Prodon, R. 2009. Measuring the postfire resilience of a bird–vegetation

system: a 28-year study in a Mediterranean oak woodland. Oecologia, 161: 801–

811.

Lawrence, G.E. 1966. Ecology of vertebrate animals in relation to chaparral fire in the

Sierra Nevada foothills. Ecology, 47: 278-291.

MacDonald, D., Crabtree, J.R., Wiesinger, G., Dax, T., Stamou, N., Fleury, P.,

Gutiérrez-Lazpita, J. and Gibson, A. 2000. Agricultural abandonment in mountain

areas of Europe: environmental consequences and policy response. Journal of


Moreno, J.M., Vázquez, A. and Vélez, R. 1998. Recent history of forest fires in Spain.

In J.M. Moreno (Ed.). Large fires. Backhuys Publishers, Leiden, The Netherlands,

pp. 159-185.

Moreira, F. and Russo, D. 2007. Modelling the impact of agricultural abandonment and

wildfires on vertebrate diversity in Mediterranean Europe. Landscape ecology, 22:

1461-1476.

Artículo I

31

Pausas, J.G. and Vallejo, R. 1999. The role of fire in European Mediterranean

Ecosystems. In E. Chuvieo (Ed.), Remote sensing of large wildfires in the European

Mediterranean Basin, Springer, Berlin, pp. 3-16.

Pausas, J.G. 2004. Changes in fire and climate in the Eastern Iberian Peninsula

(Mediterranean Basin). Climatic change, 63: 337-350.

Pausas, J.G., Llovet, J., Rodrigo, A. and Vallejo, R. 2008. Are wildfires a disaster in the

Mediterranean basin?-A review. International Journal of Wildland Fire, 17: 713-

723.

Piñol, J., Terradas, J. and Lloret, F. 1998. Climate warming, wildfire hazard and

wildfire occurrence in coastal Eastern Spain. Climatic Change, 38: 345-357.

Pons, P. and Prodon, R. 1996. Short term temporal patterns in a Mediterranean

shrubland bird community after wildfire. Acta Oecologica, 17: 29-41.

Pons, P. and Bas, J.M. 2005. Open-habitat birds in recently burned areas: The role of

the fire extent and species’ habitat breadth. Ardeola, 52: 119-131.

Preiss, E., Martin, J.L. and Debussche, M. 1997. Rural depopulation and recent

landscape changes in a Mediterranean region: consequences to the breeding

avifauna. Landscape Ecology, 12: 51-61.

Prodon, R. and Lebreton, J.D. 1981. Breeding avifauna of a Mediterranean succession:

the holm oak and cork oak series in the eastern Pyrenees. Analysis and modelling of

the structure gradient. Oikos, 37: 21-38.

Prodon, R., Fons, R. and Peter, A.M. 1984. L’impact du feu sur la végétation, les

oiseaux et les micromammifères dans diverses formations des Pyrénées-Orientales :

premiers résultats. Revue d'Ecologie (la Terre et la Vie), 39: 129-158.

Prodon, R., Fons, R. and Athias-Binche, F. 1987. The impact of fire on animal

communities in Mediterranean area. In: Trabaud L. (ed.) The role of fire on

ecological systems: 121-157. SPB Academic Publ., The Hague.

Rodrigo, A., Retana, J. and Picó, X.F. 2004. Direct regeneration is not the only response

of Mediterranean forests to large fires. Ecology, 85: 716-729.

Sala, O.E., Chapin, F.S., Armesto, J.J. and 16 other authors. 2000. Global biodiversity

scenarios for the year 2100. Science, 287: 1770-1774.

Stoate, C., Boatman, N.D., Borralho, R.J., Rio Carvalho, C., de Snoo, G.R. and Eden, P.

2001. Ecological impacts of arable intensification in Europe. Journal of


Stuart-Smith, K., Adams, I.T., Karl., W. and Larsen, K.W. 2002. Songbird communities

in a pyrogenic habitat mosaic. International Journal of Wildland Fire, 11: 75–84.

Artículo I

32

Ukmar, E., Battisti, C., Luiselli, L. and Bologna, M.A. 2007. The effects of fire on

communities, guilds and species of breeding birds in burnt and control pinewoods in

central Italy. Biodiversity and Conservation, 16: 3287-3300.

Artículo I

33

Appendix 1 Summary description of the three bioclimatic regions of Catalonia, Spain, according to dominant forest species, mean annual temperature (T), and mean annual precipitation (P)

Note: Data are from DGMN (1994) and Burriel et al. (2000-2004).

Region

Mean

annual

T (ºC)

Mean annual

P (mm) Dominant forest species

1. South 14.41 53391 Pinus halepensis, Quercus ilex, Pinus nigra

2. Northwest 10.00 810.42 Pinus sylvestris, Pinus nigra, Quercus ilex,

Quercus humilis

3. Northeast 13.02 841.99 Quercus ilex, Quercus humilis, Pinus sylvestris,

Quercus suber

34

Artí

culo

I

App

endi

x 2

Des

crip

tion

of th

e ar

eas

affe

cted

by

fire,

in c

hron

olog

ical

ord

er. T

he d

escr

iptio

n in

clud

es th

e na

me,

cod

e, m

ain

habi

tats

bef

ore

the

fire,

yea

r whe

n th

e fir

e oc

curr

ed, f

ire a

rea,

num

ber o

f lin

e tra

nsec

ts c

ondu

cted

with

in fi

re p

erim

eter

in 2

006,

200

7, 2

008

and

2009

.

Fire

NA

ME

Fi

re C

OD

E

Yea

rM

ain

HA

BIT

AT

Fi

re a

rea

(ha)

20

0620

0720

0820

09

Alb

iol

29

2000

Shru

blan

d64

5.48

1717

1617

C

ampo

sine

s 34

20

00

Shru

blan

d 52

.11

2 2

2

Font

rubi

23

20

00

Fore

st (P

. hal

epen

sis)

87

.57

3 3

3

Gar

rigue

lla

2 20

00

Shru

blan

d 63

89.1

9 31

31

31

Oliv

ella

25

20

00

Shru

blan

d an

d P.

hal

epen

sis

382.

77

12

12

10

Pr

atde

Com

pte

33

2000

Fo

rest

(P. h

alep

ensi

s)

275.

85

13

14

13

B

adal

ona

17

2001

Sh

rubl

and

66.5

1 3

3

C

adaq

ues

1 20

01

Shru

blan

d 17

44.1

1 25

25

C

aste

lldef

els

26

2001

Sh

rubl

and

53.7

3 2

2

C

ubel

ls

42

2001

Sh

urbl

and

and

Que

rcus

coc

cife

ra

207.

18

11

11

Esca

la

3 20

01

P. h

alep

ensi

s and

shru

blan

d 37

1.88

11

11

11

12

La

Flor

esta

40

20

01

Shru

blan

d 10

9.89

5

5

V

allb

onaA

noia

21

20

01

P. h

alep

ensi

s, P.

pin

ea, s

hrub

land

and

Que

rcus

ilex

51

.75

2 2

Cas

tellb

isba

l02

19

2002

Sh

rubl

and

126.

72

5 5

5

Tivi

ssa

31

2002

Sh

rubl

and

and

P. h

alep

ensi

s 53

.91

2 2

2

Vila

verd

Lilla

27

20

02

Shru

blan

d 49

5.27

15

15

15

Alc

over

28

20

03

Shru

blan

d 21

5.37

10

10

10

10

C

alde

sMal

avel

la03

8

2003

P.

pin

ea a

nd sh

rubl

and

60.2

1 2

2 2

2 C

aste

llbel

liVila

r 16

20

03

Shru

blan

d an

d P.

hal

epen

sis

296.

91

14

14

13

12

Dal

tmar

Ole

rdol

a 24

20

03

Shru

blan

d 12

9.24

6

6 6

6 G

ranj

aEsc

arp

39

2003

Sh

rubl

and

1794

.33

22

22

22

22

Jorb

a 22

20

03

Shru

blan

d 64

.35

3 3

3 3

Mas

anet

Gra

n 9

2003

Q

. sub

er, P

. pin

ea, P

. pin

aste

r, Q

. ile

x an

d sh

rubl

and

920.

52

21

21

21

21

Mas

anet

Petit

10

20

03

Q. i

lex,

Q. s

uber

and

shru

blan

d 62

.64

2 2

2 2

Mas

quef

a 20

20

03

Fore

st (P

. hal

epen

sis)

58

.77

3 3

2 3

Plat

jaA

ro

5 20

03

Shru

blan

d, Q

. sub

er, P

. hal

epen

sis a

nd P

. pin

ea

318.

51

12

12

12

12

Sant

Feliu

Gui

xol

6 20

03

Shru

blan

d, Q

. sub

er, P

. hal

epen

sis a

nd P

. pin

ea

514.

44

18

18

18

18

Sant

Llor

ensS

aval

l 15

20

03

Fore

st (P

. hal

epen

sis)

44

83.2

6 30

30

30

30

Se

lvan

era

41

2003

Sh

rubl

and

121.

86

5 5

5 5

Tala

man

ca

14

2003

Fo

rest

(P. h

alep

ensi

s)

191.

07

7 7

7 7

35

Artí

culo

I

App

endi

x 2

con

tinue

d Fi

re N

AM

E

Fire

CO

DE

Y

ear

Mai

n H

AB

ITA

T

Fire

are

a (h

a)

2006

20

07

2008

20

09

Mon

tani

ssel

l43

20

04Sh

rubl

and

and

Pinu

s nig

ra su

bsp.

salz

man

ni80

.82

44

44

Mon

tgri

4 20

04

Shru

blan

d 50

0.85

16

16

16

16

B

alsa

reny

12

20

05

Fore

st (P

. hal

epen

sis)

85

8.87

19

20

19

19

B

orge

sdel

Cam

p 30

20

05

Shru

blan

d an

d P.

hal

epen

sis

105.

75

3 3

3 3

Cal

desM

alav

ella

05

7 20

05

P. p

inea

and

shru

blan

d 78

.57

3 3

3 2

Car

dona

11

20

05

Fore

st (P

. hal

epen

sis)

12

16.1

7 21

21

21

21

C

aste

llbis

bal0

5 18

20

05

Shru

blan

d an

d P.

hal

epen

sis

206.

73

6 6

6 6

Mar

gale

f 38

20

05

P. h

alep

ensi

s and

shru

blan

d 38

4.21

13

13

13

13

Pa

lmad

eEbr

e 37

20

05

Shru

blan

d 90

.18

4 4

4 4

Pere

llo

32

2005

Sh

rubl

and

96.5

7 5

5 5

5 Po

blad

eMas

aluc

a 35

20

05

Shru

blan

d an

d P.

hal

epen

sis

104.

04

4 4

4 4

Rib

aRoj

a05

36

2005

P.

hal

epen

sis a

nd sh

rubl

and

605.

25

18

18

18

18

Roc

afor

t 13

20

05

P. h

alep

ensi

s and

shru

blan

d 78

3.72

20

20

18

20

V

ilade

cans

44

20

05

Shru

blan

d 57

.78

3 3

3 3

Cap

man

y 47

20

06

Shru

blan

d an

d Q

. sub

er

239.

85

11

11

11

C

iste

lla

46

2006

P.

hal

epen

sis a

nd sh

rubl

and

200.

52

9

7 7

LaFe

bro

50

2006

P.

sylv

estri

s and

shru

blan

d 50

2 2

2 O

gern

48

20

06

Fore

st (P

inus

nig

ra su

bsp.

salz

man

ni)

86.0

4

4 4

4 V

ande

llos

51

2006

Sh

rubl

and

1129

.5

21

21

20

V

enta

llo

45

2006

Fo

rest

(P. h

alep

ensi

s)

768.

33

20

20

20

V

imbo

di

49

2006

P.

hal

epen

sis a

nd sh

rubl

and

115.

38

5

5 5

Mon

troig

delC

amp

53

2007

Fo

rest

(P. h

alep

ensi

s)

287.

1

12

12

N

avas

StSa

lvad

orTo

56

20

07

Fore

st (P

. hal

epen

sis)

22

1.76

9

9 R

ibaR

oja0

7 52

20

07

Shru

blan

d an

d P.

hal

epen

sis

91.8

4

4 Sa

llent

55

20

07

Fore

st (P

. hal

epen

sis)

64

.17

3 3

Torr

edeF

onta

ubel

la

54

2007

Sh

rubl

and

419.

58

12

12

TO

TA

L

453

527

449

488

Artículo I

36

Appendix 3 Proportion of burned areas surveyed from 2006 to 2009 and planned for 2010. Black, grey and white indicating 100%, 10-20% and 0% of wildfires monitored respectively. The total number of wildfires monitored per year is also shown. Note that in 2008 none of the wildfires occurring in Catalonia fulfilled the conditions to be included in DINDIS.

WILDFIRES

2000 2001 2002 2003 2004 2005 2006 2007 2009 Total

2006 44

2007 51

2008 45

2009 47

CEN

SUS

YEA

R

2010 41

37

4.2. Artículo II

II- FUNCTIONAL CONNECTIVITY DETERMINES THE POST-FIRE

COLONISATION OF AN OPEN-HABITAT BIRD SPECIES

Artículo II

38

SUMMARY

Wildfires are certainly a key important natural disturbance in Mediterranean

terrestrial ecosystems. After an intense fire, habitats with different vegetation cover and

structure converge towards structurally simpler open habitats. This early successional

stage has been shown to be used by many open-habitat bird species, which are able to

reach these new suitable habitats through colonisation. Adopting a regional scale

perspective, we assessed to what degree colonisation of recently burnt areas is

constrained by dispersal or by the amount of post-fire suitable habitat for an open

habitat bird species. We focused on the Catalan population of Ortolan Buntings

(Emberiza hortulana) and estimated the potential dispersal flux received by each

recently burnt area using available regional scale atlas data and connectivity metrics

based on graph theory. We evaluated our predictions using fine-grained bird data

gathered from a set of wildfires occurring in the region. Our results showed that species

occurrence on recently burnt areas was primarily driven by the amount of estimated

dispersal flux received by these areas and to a lesser extent to the amount of habitat

created by the fire itself. Species occurrence tended also to increase with time since fire,

suggesting that effective colonisation was partly driven by stochastic ecological and

behavioural processes. We suggest that to reproduce mechanistically the observed

patterns, the prediction of species’ responses to disturbances at large-spatial scales

should explicitly integrate species responses to habitat changes but also information on

dispersal constraints imposed by species ecology.

Keywords: Ecological networks, Emberiza hortulana, graph theory, landscape context,

population sources

Artículo II

39

INTRODUCTION

Predicting species dynamics is a key question in ecology and a challenge in a

changing world. Due to the complexity of community dynamics, the use of surrogates is

essential in order to develop operational tools that allow consistent predictions of the

consequences of such changes in biodiversity (With and Crist 1995). Species

persistence at the metapopulation level depends on a stochastic equilibrium between

local extinctions and colonisations of suitable habitat patches. Both processes are

considered as the key components of the metapopulation theory (Hanski 1999), which

has heavily relied on patch size and isolation in order to explain species spatial

dynamics (e.g. Hanski 1994, Moilanen and Hanski 1998). Smaller population sizes are

believed to be more prone to extinction, while the colonisation probability of an empty

patch is largely determined by its isolation from surrounding populations. Traditionally,

these studies have focused on these processes using models that simplified assumptions

concerning the long-term persistence of patches and the time-scale separation between

local and global dynamics. However, most habitats are affected, at least to some degree,

by dynamics related to disturbances and their subsequent successional processes

(Biedermann 2005, Vuilleumier et al. 2007).

The nature of patchiness induced by a disturbance affects the level of resource

availability in the disturbed patches, the species survival in the patch and the rate of

colonisation and success of establishment of new species (Sousa 1984, Pickett and

White 1985). Wildfires represent an important natural disturbance in Mediterranean

terrestrial ecosystems (Trabaud 1994, Whelan 1995), shaping the landscapes into their

present mosaic-like patterns (Piussi 1992, Naveh 1994). After an intense fire, habitats

with different vegetation cover and structure converge towards structurally simpler open

habitats that will progressively tend to more complex vegetation structure (Trabaud and

Lepart 1980, Lloret et al. 1999). This early successional stage has been shown to be

used by many open-habitat bird species (Herrando et al. 2002, Pons and Bas 2005),

which are able to reach these new suitable habitats through colonisation. Several studies

have focused on factors affecting bird community responses to fire (Pons and Prodon

1996, Moreira et al. 2001, Herrando et al. 2002, Jacquet and Prodon 2009) but there is

rather limited knowledge on the factors determining species capacity to colonise burnt

areas. Recent studies have suggested a strong role of dispersal on post-fire bird species

colonisation (Brotons et al. 2005, 2008).

Artículo II

40

This work focused on a Mediterranean population of Ortolan Buntings

(Emberiza hortulana), an open-habitat bird species linked to post-fire burnt areas (Dale

and Olsen 2002, Revaz et al. 2005, Pons and Bas 2005). This species has suffered large-

scale declines across Europe in the last decades (BirdLife International 2004). This

decrease appears to be related to general habitat loss and degradation as a consequence

of homogenization of agricultural landscapes and farming intensification (Dale 2001,

Vepsalainen et al. 2005). Nevertheless, in Catalonia (North-eastern Iberian Peninsula)

this species has almost doubled its breeding range in the last 20 years (Estrada et al.

2004). Whereas the main trends in agricultural practice appear to be similar to those

prevailing in other European regions (Sirami et al. 2007), in Catalonia wildfires have

been recently suggested as a potential cause of the expansion of the species by creating

suitable habitats for the colonisation and persistence of their populations (Brotons et al.

2008).

The objective of the present study was to assess to what degree the colonisation

of recently burnt areas by the Ortolan Bunting is constrained by dispersal or by fire size.

Using the metapopulation framework as a baseline, we hypothesised that, colonisation

dynamics of new appearing patches are likely to be more related to connectivity from

potential colonisers’ sources than to the local availability of habitat in the newly burnt

areas. We estimated the dispersal flux received from nearby population sources to

recently burnt areas using available regional scale atlas data and connectivity metrics

based on graph theory and evaluated our predictions using fine-grained bird data

specifically gathered from an exhaustive set of burnt areas for these purposes. We

hypothesised that if post-fire colonisation is highly constrained by dispersal,

connectivity metrics will better predict the colonisation of Ortolan Buntings in recently

burnt areas than the amount of habitat generated by fire disturbances. Finally, if

stochastic processes such as the probability of finding a new habitat patch or finding a

mate constrain effective species colonisation of recently burnt areas, we expect that time

since disturbance will play an additional role in explaining species occurrence patterns

in burnt areas.

MATERIAL AND METHODS

Study area

The study was carried out in Catalonia (32,091 km2), a region dominated by

Mediterranean climate and located in the north-eastern corner of the Iberian Peninsula

Artículo II

41

(between c. 0º15É and 3º20É longitude and 40º30Ń and 42º40Ń latitude). During the

last decades, landscape changes related to socio-economical dynamics in the region are

representative of those that occurred across Mediterranean Europe. On one hand, there

has been a marked land abandonment of less productive farmlands which has increased

shrubland and forest cover through vegetation succession (Debussche et al. 1999). On

the other hand, this process has simultaneously led to fuel accumulation and, therefore,

to an increase in wildfire impact with the consequent return to open habitats (Le

Houerou 1990).

In this context, open-vegetation habitats like shrublands in our study area mainly

arise through fire and secondary succession after abandonment of less productive

cropland. Fires have had a strong impact in Catalan landscapes and the size and

frequency of burnt areas have greatly increased in recent years (Pausas 2004, González

and Pukkala 2007). Between 1975 and 1998, about 234,000 ha of forest and shrubland

burned, and a few large fires accounted for most of the burnt area (Díaz-Delgado et al.

2004).

Post-fire bird sampling

We used fire perimeters provided by the Departament Medi Ambient i Habitatge

(DMAH) and selected all wildfires that occurred between 2000 and 2006 and had

affected a minimum of 50 ha of woodland or shrubland, resulting in a total of 49 burnt

areas ranging from 51 to 4,497 ha (Appendix). Forest and shrub land covers were

estimated using land use maps of Catalonia of 1997 and 2002 originated from remote

sensing imagery with a spatial resolution of 30 m (Viñas and Baulies 1995). All burnt

areas were located in mountain massifs, with Mediterranean climatic conditions, at low-

mid altitudes (100-1,300 m above sea level), and were formerly dominated by pine

and/or oak forests or dense shrubland (Appendix). Most fires in the region are severe,

including crown fires strongly affecting both, forest canopy and undergrowth, and

causing widespread tree mortality (Rodrigo et al. 2004).


estimate the presence or absence of the focal species (Bibby et al. 2000). The number of

transects conducted at each burnt site increased non-linearly with burnt area, ranging

from two to thirty (Appendix). Transects were located with the condition that they had

to be conducted entirely across burnt wildland avoiding, when possible, unburnt patches

(forest or farmland) and fire edges. In the largest burnt areas, transects are distributed in

Artículo II

42

a number of representative locations covering habitat heterogeneity within fire

perimeter.

Bird surveys were conducted during the breading season (10th May – 15th June)

of 2006-2009, by twelve experienced ornithologists at a speed of 2km/h approximately,

three hours after sunrise and under good weather conditions (i.e. without rainfall or

strong wind) (Bibby et al. 2000). Each bird survey lasted for 15 min and covered about

460m in length (SD 95m, range 381-526m). Within each transect, we noted the

occurrence (presence or absence) of the focal species either seen or heard within 100m

belts on both sides of the track. Overall, 479 transects were distributed in the selected

burnt sites.

Post-fire potential dispersal flux

We used a graph theory-based approach to estimate the potential dispersal flux

of Ortolan buntings into the studied burnt areas following fire. The use of graph

structures and algorithms has been shown to be a powerful and effective way of both

representing the landscape pattern and performing relatively complex connectivity

analysis (Urban and Keitt 2001, Pascual-Hortal and Saura 2006, Urban et al. 2009). A

graph represents a landscape as a set of habitat patches (‘nodes’ in graph theory

literature) and ‘links’ that represent the functional connections between particular pairs

of habitat patches (Urban and Keitt 2001).

In this work we differentiated two types of nodes: (1) the nodes corresponding to

the habitat patches explaining Ortolan Bunting distribution, before the impact of

selected wildfires (n=945) and (2) the nodes corresponding to the new suitable patches

originated after the studied fires (n=49). We used available regional scale data on

species distribution derived from the Catalan Breeding Bird Atlas (Estrada et al. 2004)

to identify the potential population sources. The Atlas provides the probability of

occurrence of the Ortolan Bunting in 1 x 1 km squares covering all Catalonia, because

of field sampling and niche-based modelling (Estrada et al. 2004). In this sense, atlas

data for all the burnt sites used in this study were collected before the fire event, which

allowed their used for further analyses. Based on this data source, the distribution of the

Ortolan Bunting is concentrated in central and Northwest Catalonia in a number of

scrubby massifs near the coast and locally in the Pyrenees up to 2,400 m (Pons 2004).

For this study, we calculated the probability of species occurrence in each of the 2 x 2

km squares (as the mean value of the four adjacent 1 x 1 km squares) and selected as

Artículo II

43

disperser’s sources those squares with a value greater than 0.1 (the smallest probability

of species occurrence where it was found during field work), resulting in a total of 945

nodes (2 x 2 km squares) characterized by the probability of species occurrence (Fig. 1).

We used a spatial resolution of 2 x 2 km squares in order to maximize modelling

process due to Conefor Sensinode software (see below) is not optimal for graph

structures with large number of nodes. In order to characterize the links we used

centroid to centroid Euclidean distances between nodes and an average breeding

dispersal distance of 12 km, close to the median breeding distance observed in a

population of this species in Norway (Dale et al. 2005, 2006). Links were considered

symmetric (undirected graphs).

We used the Conefor Sensinode software (CS) (Saura and Torné 2009,

http://www.conefor.org) to estimate the dispersal flux received on recently burnt areas

by using two connectivity metrics: the integral index of connectivity (IIC) (Pascual-

Hortal and Saura 2006) and the probability of connectivity (PC) (Saura and Pascual-

Hortal 2007). These metrics are suited to analyse the impact in connectivity of a

variation in the habitat pattern (Pascual-Hortal and Saura 2006, Saura and Pascual-

Hortal 2007), such as those created by wildfires at different spatial scales. The two

metrics differ from one another in that IIC is based on a binary connection model (nodes

are either connected or not connected) while PC relies on a probabilistic connection

model (there is a certain probability of direct dispersal between two nodes, pij). For IIC

we assigned a link between two nodes if the distance between them was smaller or

equal than 12 km. For PC, a probability of direct dispersal pij = 0.5 was set for that

median distance of 12 km, and a negative exponential function matching to those values

was used to obtain the pij between every two nodes as a function of the distance between

them (e.g. Urban and Keitt 2001, Saura and Pascual-Hortal 2007). The IIC and PC

metrics can be partitioned in three separate fractions considering the different ways in

which a habitat patch can contribute to the total habitat connectivity and availability in

the landscape (Saura 2008, Saura and Rubio 2009 in press). Among these, we were

interested in the fraction estimating the potential dispersal flux received in each new

node (burnt area) added to the landscape, which was calculated as follows for each

individual new node:

∑≠ +

⋅=n

ji ij

jii nl

aaIICflux

1 (1)

Artículo II

44

*ij

n

jijii paaPCflux ⋅⋅= ∑

≠

(2)

ai is the attribute of the new node appearing in the landscape (wildfire). In this study, the

attribute value of the nodes corresponding to the burnt sites was considered as 1 in order

to ensure independence between the flux estimation and the size of the burnt areas.

Therefore, the flux value depended exclusively on the colonisers’ sources and the

dispersal constrains of the species, but not on a characterising attribute of the fires

themselves. aj is the attribute of each of the nodes in the original species distribution

(before the wildfires, n = 945), corresponding to a probability of species occurrence. nlij

is the minimum number of links needed to reach node i from node j, and *ijp is defined

as the maximum product probability of all possible paths between nodes i and j (product

of the pij’s in each of the links in the path between i and j). Both nlij and *ijp consider the

contribution of nodes that function as stepping stones and facilitate the dispersal

between other habitat areas (Saura and Pascual-Hortal 2007, Saura and Rubio in press).

Data analysis

We performed two types of analyses that differ in the group of selected fires and

the fieldwork data used. First, we analyzed whether species occurrence patterns in the

burnt sites depended on the amount of dispersal flux rather than on the habitat

availability within fire perimeters. We selected wildfires occurring between 2000 and

2005 (n=42) and census information from two consecutive years (2006 and 2007). We

combined the bird information into one unique observation in order to focus on the

colonisation process rather than on the temporal variation of both survey-visits.

Furthermore, we used two descriptors of bird presence in the burnt area: 1) the

presence/absence of the species in each fire location. We considered that the species

was present in a fire when it was detected in any of the two survey-visits. 2) The

frequency of occurrence (i.e. total number of occurrences given a total number of

transects conducted within fire sites). We used the higher of the two counts for this

analysis. The presence/absence matrix represented a more conservative approach with

the respect to the effect of colonisation. Both descriptors (presence/absence and

frequency of species occurrence in the selected burnt sites) were used as response

variables in generalized linear models (GLM) with a binomial error and a logit link. The

potential dispersal flux received in each fire and the burnt area were used as fixed

factors. Additionally, we included the age of the selected burnt areas (whether it

Artículo II

45

occurred before or after 2004) and its interaction with the dispersal flux received in each

fire and with the burnt area as fixed factors. We performed one model for each of the

two connectivity metrics used (IICfluxi and PCfluxi).

In the second analysis, we explicitly analyzed the role of time since fire in the

post-fire colonisation of the Ortolan Bunting. We selected those fires from which the

species information was collected the first three years after the fire (fires occurring in

2005 and 2006, n=19). In this analysis, we used the same two descriptors of bird

presence used in the previous analysis, but we focussed in the temporal variation within

and between selected burnt sites. In this sense, we used the presence/absence and the

frequency of occurrence in the burnt locations in each of the three first years after the

fire event. We included here the observers and the burnt sites as random factors by

conducting generalized linear mixed models (GLMM). Since nine out of the twelve

observers that participated in the surveys did not conduct surveys in more than five of

the selected burnt areas, we decided to lump them all into one category, thus reducing

the number of categories of observers to four. The burnt sites and observers were

included as random factors in order to account for the dependency between data

obtained in different years in the same sites and differences between observers

observations. We performed one model for each of the two connectivity metrics used

(IICfluxi and PCfluxi). All statistics analyses were conducted using R software. In

addition, we used lme4 library when performing the GLMM (Pinheiro and Bates 2000).

Finally, we computed Moran’s I autocorrelation test in order to check for

residuals’ spatial autocorrelation to ensure that the occurrence patterns are spatially

independent (Legendre et al. 2002). We used wildfires occurring between 2000 and

2005 (n=42) and the frequency of the species occurrence in 2006 and 2007 in this

analysis. We used this information because it is the most consistent regarding the fire

sites and the year when census were conducted. In addition, it considers most of the

burnt sites used in the present paper. We used the UTM of each wildfire (centroid point)

as spatial coordinate and defined nine distance classes according to the median and

maximum breeding dispersal distance (12km and 45km respectively) observed in a

Norwegian population (Dale et al. 2005, 2006). These distance classes were: 0-6km, 6-

12km, 12-18km, 18-24km, 24-30km, 30-36km, 36-42 km, 42-48km, >48km. We

analyzed separately bird data recorded in 2006 and 2007. Significance was tested using

500 permutations. The results showed that correlograms were not significant (p>0.05)

Artículo II

46

neither globally nor considering each of the distance classes. These results allowed us to

fit the models without explicitly considering the spatial autocorrelation among fire sites.

RESULTS

Occurrence patterns of Ortolan Bunting in post-fire areas

We found a high spatial heterogeneity in patterns of species occurrence between

fires (Fig. 1). Using the bird censuses data collected in two consecutive years (2006 and

2007) within the sites burnt between 2000 and 2005 (42 distinct locations), the Ortolan

Bunting was recorded in 8 different wildfires (mostly located on the central plateau)

(Fig. 1). In seven of these sites, we detected the species both in 2006 and in 2007, with a

tendency for the species to be more common in 2007, including the detection in a new-

burned site.

Fig. 1 Ortolan Bunting probability of occurrence in Catalonia at 2 x 2 km squares. The map on the lower right corner shows Ortolan Bunting presence (in black) and absence (in white) in the studied wildfires recorded in the fieldwork. Numbers correspond to wildfire codes (see Appendix).

Using the species information obtained in fires monitored from the first three

year after fire (19 fires occurring in 2005 and 2006), we found a high temporal

Artículo II

47

variability in post-fire species occurrence patterns with a trend in increasing occurrence

with time since fire. One year after the fire, the species was recorded in only one out of

the nineteen burnt areas. Two years after the fire the species remained in that site,

showing an increase in the total number of transects where the species was detected (4

transects in 2006 versus 10 transects in 2007), and was additionally detected in two new

sites. Finally, three years after the fire, the species was present in a new site and

expanded to new transects within the same wildfires where it was recorded the previous

year.

Influence of connectivity, burnt size and time since fire on species occurrence in

the studied sites

Our results showed that, independently on whether we analyzed

presence/absence or frequency of species occurrence, the post-fire colonisation patterns

of Ortolan Buntings in the studied burnt sites was significantly related to the potential

dispersal flux calculated from nearby population sources and to a lesser extent to the

burnt area (Table 1 and Fig. 2).

Table 1: Results from the GLM, analyzing to what degree connectivity and post-fire suitable area constrained species colonisation patterns.

Presence/Absence Frequency of Occurrence

Estimate SE P R2 Estimate SE P R2 IIC 66.28 64.59IICflux 11.484 5.263 * 6.466 1.154 *** Area (A) 0.006 0.003 * 0.0004 0.0001 ** Age n.s. -0.794 0.432 0.07 A x Age n.s. n.s. IICflux x Age n.s n.s PC 62.22 76.24PCflux 7.117 2.873 * 6.550 0.981 *** Area (A) 0.003 0.001 * 0.0006 0.0001 *** Age n.s. -1.443 0.517 ** A x Age n.s. n.s. PCflux x Age n.s n.s Two different descriptors of the bird presence were used: Presence/absence and the frequency of occurrence. Models were run using the amount of potential dispersal flux (Xflux), burnt area, the age of the fire and its interaction with connectivity metrics and with the burnt area. (n.s.= p-value > 0.05; * =p-value < 0.05; ** = p-value < 0.01; *** = p-value < 0.001).

Artículo II

48

The stronger the dispersal flux a site was estimated to receive, the greater

probability of species occurring within that site and the higher the frequency recorded

(Fig. 2a). The amount of burnt area was also found to significantly increase species

occurrence (Fig. 2b) although this relation was much weaker than the one established

between occurrence and the estimated potential flux (Table 1). In addition, the

frequency of occurrence was also significantly related to whether the fire occurred

before or after 2004. The older the fire was the greater frequency of occurrence

recorded. Finally, both connectivity indices used in this paper (IIC and PC) showed

similar results (Table 1).

Fig. 2 Relationship between the highest frequency of occurrence of the Ortolan Bunting recorded in 2006 and 2007 on wildfires occurring between 2000 and 2005 (n=42) and the potential flux (a) and the burnt area (b). We used the results from the Probability of Connectivity (PC) as an example. Numbers correspond to wildfire codes where the species was found (see Appendix).

When analyzing the effect of time since fire on post-fire colonisation patterns,

we did find differences in the results concerning the response variable. When we

analyzed the presence/absence, we did not observe a significantly relation neither with

the time since fire nor with the interaction with the amount of potential flux received

from population sources (Table 2). Nevertheless, we showed that the frequency of

species occurrence within a burnt area was significantly related to the time since fire

and the amount of potential dispersal flux (Table 2 and Fig. 3). The older the burnt area

was, the higher the frequency of the species recorded. We did not find significant

differences when exploring the effect with the interaction. Both connectivity indices

(IIC and PC) showed similar results (Table 2).

a) b)

Artículo II

49

Table 2 Results from the GLMM, analyzing the effect of time since fire on species colonisation patterns.

Presence/Absence Frequency of Occurrence

Estimate SE P R2 Estimate SE P R2 IIC 64.23 87.62Fixed effects IICflux 10.148 4.874 * 8.275 2.226 *** Time since fire (T) 1.0 0.285 *** IICflux x T n.s. Random effects Fire Site (1.9e-13; 4.3e-7) (0.073; 0.27) Observer (0; 0) (0.092; 0.303) PC 64.44 86.96Fixed effects PCflux 7.914 2.346 *** 6.635 1.366 *** Time since fire (T) n.s. 1.024 0.289 *** PCflux x T n.s. n.s. Random effects Fire Site (0; 0) (0.373; 0.611) Observer (3e-11; 5.47e-6) (5.3e-10; 2.3e-5) Two different descriptors of the bird presence were used: Presence/absence and the frequency of occurrence. Models were run using time since fire (one, two or three years), the amount of potential dispersal flux (Xflux) and the interaction between both variables. Fire sites and observers were included as random factors. The variance and standard deviation of the random factors are shown. (n.s.= p-value > 0.05; * =p-value < 0.05; ** = p-value < 0.01; *** = p-value < 0.001).

1 2 3

Time since fire (yr)

0.0

0.1

0.2

0.3

0.4

0.5

Freq

uenc

y of

occ

urre

nce

Fig. 3 Relationship between Ortolan Bunting frequency of occurrence on wildfires occurring in 2005 and 2006 (n=19) and time elapsed since fire.

Artículo II

50

DISCUSSION

Our results showed strong spatial differences in the post-fire occurrence of the

Ortolan Bunting between geographically distinct sites in Catalonia. The pattern of

species occurrence within a site was primarily associated with the amount of potential

dispersal flux received from nearby population sources and to a lesser extent with the

amount of suitable habitat created by the fire itself, suggesting that the connectivity

between habitat patches plays a major role in the colonisation of recently burnt areas.

These results support the predictions from metapopulation theory that colonisation

process is greatly constrained by the connectivity of the habitat patch network

(Moilanen and Hanski 2001) and build on previous findings on birds in dynamic

landscapes suggesting that dispersal and landscape context are important factors in

determining post-disturbance coloniser structure in bird communities (Brotons et al.

2005).

Wildfires generate the habitats needed for open-habitat bird species like the

Ortolan Bunting by providing resources such as nest site, food availability or lower

predation (Pons and Prodon 1996, Dale and Olsen 2002). However, the geographical

location of the new burnt sites and the limited dispersal capacity of the Ortolan Bunting

appeared to determine the colonisation process. Thus, the absence of the species in

wildfires located c. 50 km from the nearest potential sources of colonisers (e.g. as in

Granja d’Escarp (Fire code=37) or Caldes de Malavella I and II (fire codes 6 and 7

respectively)) (see Fig. 1), contrasts with the high frequency of the species occurrence

in fire sites where this distance was the order of few km (e.g. Cardona, with fire

code=10). This pattern clearly shows that the species colonisation of burnt areas was

probably facilitated by short distances to potential colonizers’ sources. For a species

such as the Ortolan Bunting that relies on the availability of patchy and dynamic

habitats (such as the ones created after fire), dispersal capability is crucial in locating

areas in which to establish breeding territories and is a key factor in the overall

dynamics of Ortolan Bunting distribution. Nevertheless, many steps and decisions have

to be made by an Ortolan Bunting before settling in a new habitat patch: leaving the

previous patch, finding the new patch and finally deciding whether to stay in the patch

or not. Constraints at different stages may contribute to dispersal limitations. Leaving a

previously occupied patch is not probably a limiting factor since previous studies

suggests that movements by Ortolan Buntings do not depend on the detection of new

suitable patches at the time of leaving and that unsuitable habitat stretching beyond the

Artículo II

51

perceptual range does not necessarily present a barrier to this open habitat bird (Dale et

al. 2006). Second, the searching behaviour of the Ortolan Bunting seemed to follow

some kind of threshold giving up tactic suggesting some kind of search cost. However,

this search cost did not appear to be related to travelling time or energy consumed but

rather to the risk of not finding a new breeding area with a vacant territory or mate if

search is continued (Dale et al. 2006). Many species appear to use conspecifics’

presence or density to assess habitat quality in a heterogeneous landscape (Reed and

Dobson 1993, Stamps et al. 2005). In these cases, dispersal constraint is expected to be

derived from behavioural processes linked to habitat selection patterns in complex

landscapes. Dale et al. (2006) suggest that in declining populations with patches above

their carrying capacity and surplus of suitable habitat, dispersal males may have more

opportunities to settle at most sites inspected. In fact, they found that the male Ortolan

Buntings move surprisingly long distances (up to 45 km) during a short time (Dale et al.

2005). This matches the distance within we detected the species on recently burnt areas

suggesting that this value may be related to the limits of dispersal for this Bunting. In

this context, long distance dispersal may lead to colonisation of suitable habitat patches

created after fires far from the main breeding areas and it may be an effective way to

persist in a dynamic system. Our finding of strong dispersal constraints driving post-fire

colonisation of new habitats patches suggest that reports on this species (Dale et al.

2006) may be also extended to shifting habitat mosaics in which perturbations generate

a surplus of new adequate habitat.

On the other hand, our results indicated that the frequency of species occurrence

increased with the number of years elapsed since the fire event. This might occur due to

new immigrants arriving to the burnt locations or to local recruitments within burns.

Whether species frequency of occurrence increase with time since fire due to local

recruitment or new colonisations cannot be determined, probably due to the short-term

of the study (during the first three years after the disturbance) and/or the low species

occurrence in the selected sites. Nevertheless, the fact that several of the burnt sites

occurring in 2003 located close to the main species population on the central plateau

(e.g. Talamanca (Fire code=13) and Castellbell (Fire code=15)) (personal observations)

were colonised in 2008 suggests that the former is likely to occur in several cases. The

colonisation process might be constrained by stochastic factors not entirely ensured in

highly connected patches. For instance, scarcity of mates in the new appearing habitats

may prevent effective settlement of males immediately after fire. Time since fire may

Artículo II

52

contribute to effective individual settlement and boost effective colonisation by

conspecifics attraction mechanisms (Dale et al. 2006). On the other hand, colonisation

might be also limited by vegetation regrowth. Nevertheless, the Ortolan Bunting is a

typical pioneer colonizer of the early stages of vegetation succession and many studies

have already detected it the first year after fire (Pons and Prodon 1996, Herrando et al.

2002, Brotons et al. 2005, Pons and Bas 2005). Since all fires occurred in the early

months of the year, areas sampled one year after fire have considerable vegetation

regrowth ranging from 20 to 80 % of vegetation cover in at 0-0.25 m height (own

unpublished data). In this context, species colonisation with time elapsed since fire may

not be related to vegetation regrowth but rather to a matter of species capacity to arrive

to the recently created suitable habitat (dispersal constraints sensu, Brotons et al. 2005).

We highlight the importance of long-term studies in order to understand better this

process.

Pons and Bas (2005) showed that fire size was an important factor in order to

determine the species richness and composition within the burnt area. Nevertheless, our

results suggest that fire size appeared to have a secondary role in determining post-fire

colonisation. Actually, patch size is expected to have a role on post-fire colonisation

process since stochastic dispersal events may be more probable in large patches. This is

probably because they are more likely to be found by migrants. However, in the case of

the Ortolan Bunting this effect was largely overridden by the effect of connectivity and

dispersal pressure. In this context, although large fires may play a critical role in seeding

new populations by creating large favourable habitat patches, we have shown that these

areas should be located close to population sources connected by dispersion. In case of

being colonised, large fires may act as reservoirs of population playing an important

role in determining future species distribution changes.

Strong dispersal constraints on colonisation process of open habitat bird species

appear as a greatly relevant result in the current landscape dynamic affecting many

Mediterranean regions. Land abandonment in agricultural areas of marginal productivity

is driving a general decrease in suitable habitats eventually leading to local extinctions

for species such as the Ortolan Bunting (Sirami et al. 2007, Brotons et al. 2008). In this

context, wildfires are the main large scale disturbance creating new suitable open

habitats and thus offering the opportunity for these species to colonise new sites. Our

results suggest that open-habitat bird species dynamics seem to be greatly dependent on

Artículo II

53

the interaction between current species distribution and the spatial pattern of fire

occurrence (Vuilleumier et al. 2007).

Finally, in a context of global change it seems of utmost importance to extend

this study to other Mediterranean species in order to understand and to be able to predict

the patterns of biodiversity change and their interactions with disturbances and

ecosystem pressures at sufficiently large spatial and temporal scales. Further insights in

this respect may be as well obtained through direct monitoring of the actual dispersal

patterns (radiotracking, mark-release-recapture) of selected individuals of these

disturbance-sensitive species. This may allow closer and more detailed insights into the

factors limiting and enhancing colonization and the use of habitat in burnt areas.

Use of connectivity indices

This paper highlights the importance of using connectivity measures in order to

predict species distribution in a changing landscape. In this sense, models based

exclusively in the amount of habitat in the landscape or in the disturbed sites will not be

realistic in species such as the one studied here if they do not explicitly include dispersal

capacity. This is a key assumption incorporated in habitat or niche based static

distribution modelling applied to the dynamic prediction of changes in species

distribution under scenarios of environmental change (Guisan and Thuiller 2005). The

prediction of species distributional responses to disturbances should explicitly include

dispersal constraints and the context in which they occur in order to be able to mimic

observed patterns. One possibility is to integrate habitat models and information

generated by using graph theory analysis, as per example the amount of dispersal flux

received, as estimated through metrics like IIC and PC, which, among others, may be

particularly suited for these purposes (Saura and Pascual-Hortal 2007, Saura and Rubio

in press, Visconti and Elkin in press). In general, graph models offer a versatile and

flexible representation of habitat mosaics and can provide insights into a variety of

ecological questions at both the patch and landscape level (Urban et al. 2009). Our

results, together with other recent researches (McRae and Beier 2007, Neel 2008),

contribute to demonstrate the ability to explain relevant ecological processes through a

graph representation of the landscape.

This should not however preclude the value and usefulness of other available

connectivity metrics and approaches. For instance, spatially explicit metapopulation

models (e.g. Hanski 1999, South 1999) are able to consider the population dynamics

Artículo II

54

associated with individual habitat patches and with an explicit modeling of birth,

mortality, emigration and immigration processes. Although the graph metrics used here

were able and sufficient to capture the effects of functional isolation on colonization

patterns (absence/presence records), metapopulation models of different kinds would be

needed when the connectivity analysis requires a more detailed evaluation of the

demographic dynamics and temporal persistence of the target populations. This would

require however that the information needed to parameterize these models is available at

sufficiently wide scales to be usable for projecting biodiversity patterns under different

change scenarios, which may limit in practice the scope of application of some of the

available models and metrics related to landscape connectivity (Calabrese and Fagan

2004).

ACKNOWLEDGEMENTS

We thank all the volunteers who participated in the Catalan Breeding Bird Atlas

(CBBA). We also thank M. Pla for the laborious GIS work and C. Puerta-Piñero for her

help on spatial autocorrelation analyses. This work has received financial support from

the projects CGL2005-00031/BOS and CONSOLIDER-MONTES granted by the

Spanish Ministry of Education and Science and has been supported by a Marie Curie

International Reintegration grant within the 6th European Community Framework

programme. L.B. benefited from a Ramón y Cajal contract from the Spanish

government and E.L.Z. (FPI fellowship) received financial support from the Spanish

Ministry of Education and Science.

REFERENCES

Bibby, C.J., Burgess, N.D., Hill, D.A. and Mustoe, S.H. 2000. Bird Census Techniques.

Second Edition. - Academic Press, London.

Biedermann, R. 2005. Incidence and population dynamics of the leaf beetle Gonioctena

olivacea in dynamic habitats. - Ecography 28: 673-681.

BirdLife International. 2004. Birds in Europe: population estimates, trends and

conservation status. - BirdLife International, Cambridge, UK (BirdLife

Conservation Series No. 12).

Brotons, L., Pons, P. and Herrando, S. 2005. Colonization of dynamic Mediterranean

landscapes: where do birds come from after fire? - J. Biogeogr. 32: 789-798.

Artículo II

55

Brotons, L., Herrando, S. and Pons, P. 2008. Wildfires and the expansion of threatened


landscapes. - Journal of Applied Ecology 45: 1059-1066.

Calabrese, J.M. and Fagan, W.F. 2004. A comparison-shopper's guide to connectivity

metrics. - Frontiers in Ecology and the Environment 2: 529-536.

Dale, S. 2001. Causes of population decline of the Ortolan Bunting in Norway.

Proceedings of the 3rd International Ortolan Bunting Symposium (Bunting Studies

in Europe), pp. 33–41. Norwegian University of Life Sciences, Aas, Norway.

Dale, S. and Olsen, B.F.G. 2002. Use of farmland by Ortolan Buntings (Emberiza

hortulana) nesting on a burned forest area. - Journal für Ornithologie 143: 133–144.

Dale, S., Lunde, A. and Steifetten, Ø. 2005. Longer breeding dispersal than natal

dispersal in the Ortolan Bunting. - Behavioural Ecology 16: 20-24.

Dale, S., Steifetten, Ø. Osiejuk, T.O., Losak, K. and Cygan, J.P. 2006. How do birds

search for breeding areas at the landscape level? Interpatch movements of male

Ortolan Bunting. - Ecography 29: 886-898.

Díaz-Delgado, R., Lloret, F. and Pons, X. 2004. Spatial patterns of fire occurrence in

Catalonia, NE, Spain. - Landscape Ecology 13: 731-745.

Debussche, M., Lepart, J., Dervieux, A. 1999. Mediterranean landscape changes:

evidence from old postcards. - Global Ecology and biogeography 8: 3-15.

Estrada, J., Pedrocchi, V., Brotons, L. and Herrando, S. 2004. Atles dels ocells

nidificants de Catalunya (1999-2002). - Lynx, ICO (Institut Català d'Ornitologia),

Barcelona.

González, J.R. and Pukkala, T. 2007. Characterization of forest fires in Catalonia

(north-east Spain). - European Journal of forest research 126: 421-429.

Guisan, A. and Thuiller, W. 2005. Predicting species distribution: offering more than

simple habitat models. - Ecology Letters 8: 993-1009.

Hanski, I. 1994. A practical model of metapopulation dynamics. – J. Anim Ecol. 63:

151-162.

Hanski, I. 1999. Metapopulation ecology. - Oxford University Press, Oxford.

Herrando, S., Brotons, L., del Amo, R. and Llacuna, S. 2002. Bird community

succession after fire in a dry mediterranean shrubland. - Ardea 90: 303-310.

Jacquet, K. and Prodon, R. 2009. Measuring the postfire resilience of a bird–vegetation

system: a 28-year study in a Mediterranean oak woodland. - Oecologia 161: 801–

811.

Artículo II

56

Legendre, P., Dale, M.R.T., Fortin, M.J., Gurevitch, J., Hohn, M. and Myers, D. 2002.

The consequences of spatial structure for the design and analysis of ecological field

surveys. - Ecography 25: 601-615.

Le Houerou, H.N. 1990. Agroforestry and sylvopastoralism to combat land degradation

in the Mediterranean Basin: old approaches to new problems. - Agric Ecosyst

Environ 33: 99–110.

Lloret, F., Verdu, M., Noe, F-H. and Alfonso, V-B. 1999. Fire and resprouting in

Mediterranean ecosystems : insights from an external biogeographical region, the

Mexical shrublands. - American Journal of Botany 86: 1655-1661.

McRae, B.H. and Beier, P. 2007. Circuit theory predicts gene flow in plant and animal

populations. - Proceedings of the National Academy of Sciences of the USA 104:

19885-19890.

Moilanen, A. and Hanski, I. 1998. Metapopulation dynamics: Effects of habitat quality

and landscape structure. - Ecology 79: 2503-2515.

Moilanen, A. and Hanski, I. 2001. On the use of connectivity measures in spatial

ecology. - Oikos 95: 147-151.

Moreira, F., Ferreira, P.G., Rego, F.C. and Buting, S. 2001. Landscape changes and

breeding bird assemblages in northwestern Portugal: implications for fire

occurrence. - Landscape Ecology 16: 557-567.

Neel, M.C. 2008. Patch connectivity and genetic diversity conservation in the federally

endangered and narrowly endemic plant species Astragulus albens (Fabaceae). -

Biological Conservation 141: 938-955.

Naveh, Z. 1994. The role of fire and its management in the conservation of

Mediterranean ecosystems and landscapes. - In Moreno, J. M. and Oechel, W. C.

(ed) The role of fire in Mediterranean type ecosystems. Springer-Verlag, New York,

New York, USA, pp. 163-186.

Pascual-Hortal, L. and Saura, S. 2006. Comparison and development of new graph-

based landscape connectivity indices: towards the priorization of habitat patches and

corridors. - Landscape Ecology 21: 956-967.

Pausas, J.G. 2004. Changes in fire and climate in Eastern Iberian Peninsula

(Mediterranean basin). - Climatic change, 63: 337-350.

Pickett, S.T.A. and White, P.S. 1985. The ecology of natural disturbance and patch

dynamics. - Academic Press, London.

Artículo II

57

Pinheiro, J.C. and Bates, D.M. 2000. Mixed-effects Models in S and SPLUS. -

Springer-Verlag, New York, USA.

Piussi, P. 1992. Environmental changes in forests. Examples from the south of Europe. -

In Teller, A., Mathy, P. and Jeffers, J. N. R. (ed) Responses of forest ecosystems to

environmental changes. Elsevier Applied Science, London, UK, pp. 298–309.

Pons, P. and Prodon, R. 1996. Short term temporal patterns in a Mediterranean

shrubland bird community after wildfire. - Acta Oecologica 17: 29-41.

Pons, P. 2004. Hortolà Emberiza hortulana. In Catalan Breeding Bird Atlas (1999–

2002) (eds J. Estrada, V. Pedrocchi, L. Brotons and S. Herrando). Lynx, Barcelona.

Pons, P. and Bas, J.M. 2005. Open-habitat birds in recently burned areas: The role of

the fire extent and species’ habitat breadth. - Ardeola 52: 119-131.

Reed, J.M. and Dobson, A.P. 1993. Behavioural constraints and conservation biology:

conspecifics attraction and recruitment. - Trends Ecol Evol 8: 253-256.

Revaz, E., Posse, B., Gerber, A., Sierro, A. and Arlettaz, R. 2005. Quel avenir pour le

bruan Ortolan Emberiza hortulana in Suisse? - Nos Oiseaux 52: 67-82.

Rodrigo, A., Retana, J. and Picó, X.F. 2004. Direct regeneration is not the only response

of Mediterranean forests to large fires. - Ecology 85: 716-729.

Saura, S. and Pascual-Hortal, L. 2007. A new habitat availability index to integrate

connectivity in landscape conservation planning: Comparison with existing indices

and application to a case study. - Landscape Urban Plann. 83: 91-103.

Saura, S. 2008. Evaluating forest landscape connectivity through Conefor Sensinode

2.2. In: Lafortezza, R. et al. (eds), Patterns and Processes in Forest Landscapes:

Multiple Use and Sustainable Management. Springer, pp. 403-422.

Saura, S. and Torné, J. 2009. Conefor Sensinode 2.2: a software package for quantifying

the importance of habitat patches for landscape connectivity. - Environmental

Modelling and Software 24: 135-139.

Saura, S. and Rubio, L. in press. A common currency for the different ways in which

patches and links can contribute to habitat availability and connectivity in the

landscape. - Ecography. DOI 10.1111/j.1600-0587.2009.05760.x.

Sirami, C., Brotons, L., Martin, J.L. 2007. Vegetation and songbird response to land

abandonment: from landscape to census plot. - Diversity and Distributions 13: 42-

52.

Sousa, W.P. 1984. The role of disturbance in natural communities. - Ann. Rev. Ecol.

Sys. 15: 353-391.

Artículo II

58

South, A. 1999. Dispersal in spatially explicit population models. - Conservation

Biology 13: 1039-1046.

Stamps, J.A., Krishnan, V.V. and Reid, M.L. 2005. Search cost and habitat selection by

dispersers. - Ecology 86: 510-518.

Trabaud, L. and Lepart, J. 1980. Diversity and stability in garrigue ecosystems after fire.

- Vegetatio 43: 49-57.

Trabaud, L. 1994. Postfire plant community dynamics in the Mediterranean Basin. In

Moreno, J. M. and Oechel, W. C. (ed) The role of fire in Mediterranean type

ecosystems. Springer-Verlag, New York, New York, USA, pp. 1-15.

Urban, D. and Keitt, T. 2001. Landscape connectivity: a graph theoretic perspective. -

Ecology 82: 1205-1218.

Urban, D.L., Minor, E.S., Treml, E.A. and Schick, R.S. 2009. Graph models of habitat

mosaics. - Ecology Letters 12: 260-273.

Vepsalainen, V., Pakkala, T., Piha, M. and Tiainen, J. 2005. Population crash of the

Ortolan Bunting Emberiza hortulana in agricultural landscapes of southern Finland.

- Annales Zoologici Fennici 42: 91-107.

Visconti, P. and Elkin, C. 2009. Using connectivity metrics in conservation planning –

when does habitat quality matter? - Diversity and Distributions 15: 602-612.

Viñas, O. and Baulies, X. 1995. 1:250 000 Land-use map of Catalonia (32,000 km2)

using multitemporal Landsat-TM data. - Int J Remote Sens 16: 129-146.

Vuilleumier, S., Wilcox, C., Cairns, B.J. and Possingham, H.G. 2007. How patch

configuration affects the impact of disturbances on metapopulation persistence. -

Theoretical Population Biology 72: 77-85.

Whelan, R.J. 1995. The ecology of fire. Cambridge University Press, Cambridge, UK.

With, K.A. and Crist, T.O. 1995. Critical thresholds in species’ responses to landscape

structure. - Ecology 76: 2446-2459.

Artículo II

59

Appendix Description of the areas affected by fire used in this study, in chronological order. The description includes the name, code, year when the fire occurred, fire area, main habitat before the fire and number of line transects conducted within fire perimeter.

Fire NAME Fire code Year Fire

area (ha) Main habitats Number

of transects

Albiol 27 2000 615.24 Shrubland 17 Camposines 32 2000 52.38 Shrubland 2 Fontrubi 22 2000 88.11 Forest (P. halepensis) 3 PratdeCompte 31 2000 268.56 Forest (P. halepensis) 13 Badalona 16 2001 66.69 Shrubland 3 Cadaques 1 2001 1685.34 Shrubland 25 Castelldefels 24 2001 54 Shrubland 2 Cubells 40 2001 212.13 Shrubland and Quercus 11 Escala 2 2001 374.31 P. halepensis and shrubland 11 LaFloresta 38 2001 71.55 Shrubland 4 VallbonaAnoia 20 2001 59.4 P. halepensis, P. pinea, 2 Castellbisbal02 18 2002 119.25 Shrubland 5 Tivissa 29 2002 54.63 Shrubland and P. halepensis 2 VilaverdLilla 25 2002 498.6 Shrubland 15 Alcover 26 2003 218.7 Shrubland 10 CaldesMalavella03 7 2003 60.21 P. pinea and shrubland 2 CastellbelliVilar 15 2003 308.97 Shrubland and P. halepensis 14 DaltmarOlerdola 23 2003 129.33 Shrubland 6 GranjaEscarp 37 2003 1845.54 Shrubland 22 Jorba 21 2003 63.63 Shrubland 3 MasanetGran 8 2003 65.88 Q. suber, P. pinea, 2 MasanetPetit 9 2003 963.18 Q. ilex, Q. suber and shrubland 21 Masquefa 19 2003 60.48 Forest (P. halepensis) 3 PlatjaAro 4 2003 330.57 Shrubland, Q. suber, 12 SantFeliuGuixol 5 2003 513.72 Shrubland, Q. suber, 18 SantLlorensSavall 14 2003 4497.93 Forest (P. halepensis) 30 Selvanera 39 2003 131.31 Shrubland 5 Talamanca 13 2003 190.98 Forest (P. halepensis) 7 Montanissell 41 2004 81.18 Shrubland and Pinus 4 Montgri 3 2004 540.72 Shrubland 16 Balsareny 11 2005 863.64 Forest (P. halepensis) 19 BorgesdelCamp 28 2005 104.76 Shrubland and P. 3 CaldesMalavella05 6 2005 82.44 P. pinea and shrubland 3 Cardona 10 2005 1264.77 Forest (P. halepensis) 21 Castellbisbal05 17 2005 209.52 Shrubland and P. halepensis 6 Margalef 36 2005 393.03 P. halepensis and 13 PalmadeEbre 35 2005 95.85 Shrubland 4 Perello 30 2005 96.3 Shrubland 5 PobladeMasaluca 33 2005 105.3 Shrubland and 4 RibaRoja05 34 2005 613.35 P. halepensis and 18

Artículo II

60

Appendix continued

Fire NAME Fire code Year Fire

area (ha) Main habitats Number

of transects

Rocafort 12 2005 790.47 P. halepensis and 20 Viladecans 42 2005 59.22 Shrubland 3 PalmadeEbre 35 2005 95.85 Shrubland 4 Perello 30 2005 96.3 Shrubland 5 PobladeMasaluca 33 2005 105.3 Shrubland and 4 RibaRoja05 34 2005 613.35 P. halepensis and 18 Rocafort 12 2005 790.47 P. halepensis and 20 Viladecans 42 2005 59.22 Shrubland 3 Capmany 45 2006 264.69 Shrubland and Q. suber 11 Cistella 44 2006 208.8 P. halepensis and shrubland 7 LaFebro 48 2006 51 P. sylvestris and shrubland 2 Ogern 46 2006 90.81 Forest (Pinus nigra 4 Vandellos 49 2006 1142.55 Shrubland 21 Ventallo 43 2006 802.62 Forest (P. halepensis) 20 Vimbodi 47 2006 117.18 P. halepensis and shrubland 5

61

4.3. Artículo III

III-INFLUENCE OF HABITAT QUALITY, PATCH SIZE AND

CONNECTIVITY ON COLONISATION OF OPEN-HABITAT BIRD

SPECIES AFTER FIRE

Artículo III

62

ABSTRACT

Many declining species in the Mediterranean region are open-habitat birds that

often use recently burnt areas. However, not all burnt sites have the same probability of

being colonised. Using the metapopulation theory as a baseline, we analyzed the relative

importance of patch size, connectivity and habitat quality on the colonisation of five

open-habitat bird species following wildfires occurred in Catalonia (northeast of Spain)

between 2005 and 2006. Bird surveys were conducted once per year during the breeding

season one and two years after the fire. Habitat quality was measured in terms of pre-

fire habitat, fire severity, aspect and slope. We used a hierarchical modelling approach

to progressively consider variations in transect length, factors acting at regional scale

and factors acting at local scale. Our results showed that the occurrence of colonisers in

burnt areas was associated to connectivity and habitat quality but not to patch size. For

most of the species, both parameters seemed to be of equal importance, although in

some cases the surrogates of habitat quality appeared to be of greater importance in

determining species effective colonisation. In addition, species’ responses to local

descriptors depended on species habitat requirements. Our findings suggest that both

regional and local factors are crucial in determining the patterns of species colonisation

in recently burnt areas and bring doubts on the role of patch size on the colonisation

process.

Keywords: aspect, ecological networks, graph theory, microhabitat, pre-fire vegetation,

severity, slope

Artículo III

63

INTRODUCTION

Forest fires are one of the major disturbances of Mediterranean ecosystems all

over the world (Naveh 1975, Pickett and White 1985). In particular, fires burn large

surfaces of forest and shrubland each year in all countries around the Mediterranean

basin (Terradas and Piñol 1996). This creates a mosaic of new habitats that become

suitable for species occupying open and early-successional habitats. For birds, these

areas are especially important because they provide habitat for a number of species that

are among the most threatened species in Europe (Birdlife International 2004). In the

last decades, the study of bird succession in burnt areas has centered researchers’

attention (Prodon et al. 1984, Pons and Prodon 1996, Jacquet and Prodon 2009). These

studies showed that open-habitat bird species occupy recently burnt areas but they

progressively leave the sites when habitat is no longer adequate. Nevertheless, the

processes by which open-habitat-associated birds are able to colonise these areas are

still poorly known (but see Brotons et al. 2005).

Investigations on the mechanism of species colonisation following fire and other

disturbances often emphasize the importance of patch size and distance to sources (e.g.,

Bengtsson 2002, Kiss and Magnin 2003, Cristoni et al. 2004, Lindberg and Bengtsson

2005). These studies have also demonstrated that species response might vary with the

kind and severity of disturbance, as well as with the local heterogeneity derived from

both biotic and abiotic factors (see also Magoulick and Kobza 2003, Norkko et al.

2006). In this study, we analyze the effect of local and regional parameters on the

colonisation of five open-habitat bird species following wildfires occurred in Catalonia

(northeast of Spain) between 2005 and 2006. The spatial extent and heterogeneity of the

fires provided an ideal opportunity to analyze factors constraining post-fire open-habitat

bird colonisation. We present our initial hypotheses concerning species responses to fire

size, connectivity (the inverse of isolation) and local habitat heterogeneity (Table 1). We

then synthesize the full set of response variables in the Discussion to address the

relative importance of these parameters on the colonisation process.

According to the metapopulation theory (Hanski 1999), the probability of

colonisation of a habitat patch is related to its area and its effective isolation. Large

areas receive more immigrants and thus, the probability of colonisation is greater than

in smaller patches (Hill et al. 1996, Kuussaari et al. 1996). The degree of isolation

depends on both the distance to potential sources of colonisation and the dispersal

distance of the colonizer. In this sense, a recent study has shown that open-habitat bird

Artículo III

64

species rely on short distance dispersal to colonise recently burnt areas (Brotons et al.

2005); we therefore expected species colonisation to be greatest in large areas and in

areas receiving high amount of migrants (Table 1).

Furthermore, local habitat structure is likely to define the quantity and quality of

available resources for a given species (Thomas et al. 2001). In this paper, we present

hypotheses on four local parameters that may influence post-fire colonisation of open-

habitat bird species (Table 1). These factors were pre-fire habitat, fire severity, aspect

and slope. First, the high density of standing and laying dead burnt trunks after a forest

fire would lead to reduce foraging area and offer less potential area for species

colonisation than shrub burned areas (Llimona et al. 1993, Pons & Prodon 1996).

Second, we expected that greater fire intensities would lead to reduced survival of the

pre-fire plant community and offer more potential area for colonisation by open-habitat

bird species (Turner et al. 1997, Kotliar et al. 2007). Third, we examined relationships

between aspect and species post-fire colonisation. Several studies have found that open-

habitat birds select south-facing due to a greater availability of bare ground surfaces on

warmer south-facing areas (Dale 2000, Menz et al. 2009). We, therefore, tested the

prediction that species colonisation should be highest in south facing areas. Forth, flat

areas might be more optimal to nest location since these areas facilitate predator

detection through improving the field of view from the nest (Götmark et al. 1995,

Whittinghan et al. 2002). Finally, since the studied species have different habitat

requirements in the study region (Estrada et al. 2004), we expected that the effects of

local parameters on colonisation process may depend on the species. To our knowledge,

this is the first study to analyze the effect of local factors and to integrate factors acting

at both scales (local and regional) on post-fire bird colonisation.

Table 1 Summary of hypotheses about the effect of regional and local factors on open-habitat bird post-fire colonisation that were tested in this study.

FACTORS HYPOTHESIS

Fire Size Greater in larger patches REGIONAL

SCALE Dispersal flux Increasing with increasing the amount of dispersal flux received from neighbour sources of population

Pre-fire habitat Decreasing in pre-fire forest Burn severity Increasing with increasing burn severity Aspect Greatest in south-facing slopes

LOCAL SCALE

Slope Greatest in shallow

Artículo III

65

METHODS

Focal species

For this study, we selected five open-habitat bird species with preferences for

shrub-like habitats and dry, extensive farmlands in Mediterranean landscapes (Estrada

et al. 2004): common Linnet (Carduelis cannabina), Corn Bunting (Emberiza

calandra), Woodlark (Lullula arborea), Black-eared Wheatear (Oenanthe hispanica)

and Sardinian Warbler (Sylvia melanocephala). We selected these species because

previous works have shown that they benefit from fire impact on vegetation (Pons and

Prodon, 1996; Herrando et al. 2002). Additionally, they were detected in more than

25% of the samples, thus providing sufficient data to perform the analyses.

Study area and bird surveys

The study area was located in Catalonia (northeast of Spain), and included 15

sites burned between 2005 and 2006 separated by a minimum distance of 10 km (Fig.

1). One hundred and fifty-nine line transects were distributed among the 15 sites (3 to

21 stations/site) (Fig. 1), always among forest and/or shrublands burned area, thus

avoiding unburnt patches and farmland areas. In order to minimize bird double counts,

we separated neighbouring transects by a minimum distance of 150m. Similarly, the

line transect were always >50m from the fire perimeter to reduce edge effects. At each

transect the species’ occurrences were recorded within 100m belts on both sides of the

track. Although a broader band would allow more birds to be recorded, we chose a

100m external radius to make an adequate correspondence between habitat quality

variables and bird records.

Bird surveys were conducted once per year during the breading season (10th May

– 15th June) one and two years after the fires, in good weather conditions (i.e. without

rainfall or strong wind), during the first 3 hours after sunrise by experienced

ornithologists at a speed of about 2km/h (Bibby et al. 2000). Each survey lasted 15 min

and covered about 480m in length (SD 88m, range 280-700 m). Because of the lack or

the scarcity of trees, visibility of birds was consistently high over all sites.

Artículo III

66

Fig. 1 Geographical location of the fifteen wildfires occurred between 2005 and 2006 in Catalonia and fire size (ha). For each fire, we show in brackets, the total number of transects conducted within burnt sites.

Independent variables

Variables at the landscape or regional scales included the fire size and the

connectedness of the burnt areas with source populations (Wiens et al. 1993) and were

measured at fire level (i.e. all transects conducted within a burnt site had the same

estimates concerning both variables). On the other hand, variables at the local scale

were measured within a 100m radius around each transect.

Fire size varied from 86 to 1438 ha (Fig. 1). We used a measure of connectivity,

which takes into account the distance to potential source populations, the species

dispersal capacity and the quantity of immigrants provided by the source populations.

This connectivity measure is based on graph theory. A graph represents a landscape as a

set of habitat patches (‘nodes’ in graph theory literature) and ‘links’ that represent the

functional connections between particular pairs of habitat patches (Urban and Keitt

2001). In this work, we differentiated two types of nodes: (1) the nodes corresponding

to the habitat patches explaining species’ distribution, before the impact of the studied

wildfires and (2) the nodes corresponding to the new suitable patches originated after

the fires (n=15). We used available regional scale data on species distribution derived

from the Catalan Breeding Bird Atlas (Estrada et al. 2004) to identify the potential

source populations. The Atlas provides the probability of occurrence of the focal species

Artículo III

67

in 1 x 1 km squares covering all Catalonia, based on field sampling and subsequent

niche-based modelling (Estrada et al. 2004). For this study, we calculated the

probability of species occurrence in each of the 4 x 4 km squares (as the mean value of

the 16 adjacent 1 x 1 km squares). We used a spatial resolution of 4 x 4 km squares to

maximize modelling process due to Conefor Sensinode software (see below) is not

optimal for graph’ structures with large number of nodes. In addition, for each species

we calculated the mean value of those squares with a probability of occurrence greater

than 0. We then selected as potential sources those squares with a probability of

occurrence greater than that calculated value. In order to characterize the links we used

centroid to centroid Euclidean distances between nodes. Links were considered

symmetric (undirected graphs). Furthermore, we used a mean dispersal distance of 12

km for all the species, which may reflect constrains to effective colonisation imposed by

their limited movement abilities. Although dispersal is a key factor in order to

understand metapopulation dynamics, this parameter is frequently unknown with

acceptable accuracy due to practical difficulties involved in tracking movement of

individuals over large distances (Wiens 2001). We did not find accurate information of

any of the studied species but we did find information of another species occupying

recently burned areas, the Ortolan bunting (Emberiza hortulana). Dale et al. (2006)

found that the mean dispersal distance of this species was 12km. Assuming that the

dispersal distance of early-successional species might be relatively similar (considering

the similar functional role, behaviour and body mass they present), we used this

distance as a reference of the dispersal abilities of the studied species.

Based on the nodes and links matrices for each species and the mean dispersal

distance, the potential dispersal flux was calculated as follows:

*ij

n

jijii paaFluxDispersal ⋅⋅= ∑

≠

(1)

where ai is the attribute of the new node appearing in the landscape (wildfire). In this

study, the attribute value of the nodes corresponding to the burnt sites was considered as

1 in order to ensure independence between the flux estimation and the size of the burnt

areas. Therefore, the flux value depended exclusively on the colonisers’ sources and the

dispersal constrains of the species, but not on a characterising attribute of the fires

themselves. aj is the attribute of each of the nodes in the original species distribution

(before the wildfires), corresponding to a probability of species occurrence. *ijp is

defined as the maximum product probability of all possible paths between nodes i and j

Artículo III

68

(product of the pij’s in each of the links in the path between i and j). In this study, a

probability of direct dispersal pij = 0.5 was set for that median distance of 12 km, and a

negative exponential function matching to those values was used to obtain the pij

between every two nodes as a function of the distance between them (e.g. Urban and

Keitt 2001; Saura and Pascual-Hortal 2007). Additionally, *ijp considers the contribution

of nodes that function as stepping stones and facilitate the dispersal between other

habitat areas (Saura and Pascual-Hortal 2007; Saura and Rubio 2010). This parameter is

implemented in Conefor Sensinode software (CS) (Saura and Torné 2009, available at

http://www.conefor.org). The potential dispersal flux varied from species and ranged

between 0.2 and 151.6 (Table 2).

Table 2 Number of 4x4 UTM squares describing species potential sources, range of attribute values of selected squares and amount of potential dispersal flux received in the selected burnt areas.

Linnet Corn Bunting Woodlark Black-eared

Wheatear SardinianWarbler

Nº 4x4 UTM CBBA 922 821 918 448 1018

Attribute (HSA) 0.34-0.96 0.34-0.98 0.32-0.92 0.16-0.98 0.55-0.99

Potential dispersal flux 3.8-117.0 15.7-89.6 3.7-110.1 0.2-79.6 15-151.6

On the other hand, the following variables were measured at transect-level. Pre-

fire vegetation was determined from the information derived by the Spanish Forest Map

(scale 1:50,000), recently developed within the recent Third Spanish National Forest

Inventory (Ministerio de Medio Ambiente 2006). This map was obtained for Catalonia

from the interpretation of aerial photographs, combined with pre-existing maps and

field inventory data. The main forest species affected were Aleppo Pine Pinus

halepensis (73%) together with Black Pine Pinus nigra (14%), with Cork Oak Quercus

suber and Stone Pine Pinus pinea. We calculated the pre-fire percentages of shrubland

and forest within 100m radius around each transect. Since both variables were highly

correlated (r2=0.99), we decided to eliminate the percentage of shrubland for further

analysis. Thus, the percentage of forest area represented a gradient of pre-fire

vegetation, with decreasing forest cover while increasing shrub vegetation.

Fire severity was estimated through changes in the NDVI (Normalized

Difference Vegetation Index; Mather 1999). For this purpose, images describing the pre

and post-fire conditions from Landsat-5 Thematic Mapper and Landsat-7 Enhanced

Artículo III

69

Thematic Mapper Plus (ETM+) satellites (spatial resolution 30 x 30 m) were employed

after applying geometric (Pala and Pons 1995) and radiometric corrections (Pons and

Solé-Sugrañes 1994) and a cloud mask. Cloud mask was created with semiautomatic

methods from thermal band and photo-interpretation (Cea et al. 2005). Near-infrared

and red bands were used to calculate NDVI. The average NDVI values within the 100m

radius transect were obtained after extracting all possible non-wildland area affected by

fire (i.e. mainly small patches of agricultural areas). In order to minimize NDVI

variations due to seasonal and inter-annual differences in phenology and by the

particular climatic conditions for each year, we calculated for each date and in each

transect the quotient between the average NDVI measurements of each transect and the

average NDVI measurements of unburned, neighbouring areas. This quotient is referred

to as QNDVI and provides vegetation monitoring that is more independent of year and

period of year (Díaz-Delgado et al. 2002). Unburned areas were selected adjacent to

burned ones using the Spanish Forest Map (scale 1:50,000). These unburned areas were

selected to have surficial geology, topography, and vegetation similar to that of adjacent

burned areas. The size of the unburned area at each burnt site increased non-linearly

with the burned size, ranging from 50 to 525 ha. Note that we obtained one unburned

area for each burned site and not for individual transects.

Since fire events always display a sudden drop of NDVI values we measured

this decrease as a measure of fire severity. Therefore, we calculated fire severity in each

transect using equation 2 (Díaz-Delgado et al. 2003). This parameter ranges between 0

and 1, with greater values with increasing fire severity.

iNDVINDVI

iNDVINDVI

Q fire-Post Q fire-PreQ fire-Post Q fire-Pre

+−

=i

iiSeverityFire (2)

where Pre-fire QNDVI i and Post-fire QNDVI i is the QNDVI value of transect i from the

immediate pre-fire and post-fire images.

Lately, we also calculated the average slope (º) using a Digital Elevation Model

(DEM) generated from 1:50 000 topography maps and the proportion of south and

north-facing pixels.

Data analysis

We used generalized linear mixed models (GLMM) with a binomial error

distribution and a logit link function to asses the relative importance of patch size,

connectivity and habitat quality on post-fire colonisation. The dependent variable taken

Artículo III

70

for each transect was the presence/absence of the species. We considered that the

species was present in a transect when it was detected in any of the two survey-visits.

Explanatory variables were first tested for pairwise correlation, using spearman’s

coefficients (rs) before inclusion in the modelling. A value of rs=0.7 was used as the

acceptable lower limit to correlation. We did not find high correlations between

explanatory variables. In addition, since positive spatial auto-correlation was found at

short lag distances for all the species, probably due to correlation between transects

within burned sites rather than correlation between sites, we included fire site as random

factor. This approach takes account of dependencies within transects. A Spline

correlogram of the Pearson residuals of the final model for each species (i.e. that

containing significantly variables) showed that there were no longer any obvious

increase in spatial autocorrelation at short lag distances. This confirmed that the main

source of spatial auto-correlation at short lag distances is indeed the dependency

between transects within burned sites.

We used a hierarchical modelling approach with three steps to progressively

consider variations in sampling effort between transects, factors acting at regional scale

and factors acting at local scale. A backward stepwise selection procedure (P to

enter<0.05) was conducted to select only significant factors. Significant variables at

each step were included in subsequent steps. For each species, we developed a model

according to these steps:

Step 1. Sampling effort. Because changes in bird colonisation between transects could

be influenced by differences in transect length, we included it as the first

controlling variable in all models.

Step 2. Regional factors. Regional processes (connectivity and patch size) may produce

changes in bird occurrence after fire (Brotons et al. 2005, Pons and Bas 2005).

Therefore, we included the fire size (log-transformed) and the amount of

potential dispersal flux for each species.

Step 3. Local factors. We then tested whether local descriptors used as surrogates of

variability in post-fire habitat structure (pre-fire vegetation, fire severity, slope

and north and south-facing pixels) contributed to explain current post-fire

colonisations. This step allowed us explicitly to assess the contribution of the

“quality” of the transects in the colonisation by comparing differences in the

deviance (McCullagh and Nelder 1989).

Artículo III

71

All statistics analyses were conducted using R software and lme4 (Pinheiro and Bates

2000) and ncf (Bjørnstad 2005) packages for GLMM and spatial auto-correlation

analyses respectively.

RESULTS

Species assemblage after fire

The number of colonised burnt sites was relatively similar between the studied

species (between 8 and 12 sites). However, the sites were different, suggesting a large

heterogeneity in the species capacity to colonise burnt areas (Table 3). On the other

hand, the transects conducted within a colonised site were not equally used, suggesting

a species preference for transect occupancy (Table 3).

Table 3 Number of stations in which species were and were not observed during the fieldwork in the studied sites. For geographical location of the sites, see figure 1. The total number of colonised burned sites is also shown.

FIRE NAME Linnet Corn Bunting Woodlark Black-eared Wheatear

Sardinian Warbler

Caldes 0 0 1-2 0 2-1 Cardona 5-16 15-6 21-0 12-9 0 Balsareny 4-15 4-15 6-13 7-12 0 Rocafort 0 2-18 14-6 6-14 4-16 Castellbisbal 0 0 0 0 6-0 BorgesCamp 0 0 0 1-2 3-0 Perello 4-1 0 1-4 5-0 5-0 PalmaEbre 3-1 2-2 3-1 4-0 4-0 Margalef 10-3 8-5 8-5 9-4 11-2 Ventallo 8-12 3-17 12-8 0 17-3 Cistella 0 6-1 7-0 0 7-0 Capmany 3-5 6-2 6-2 0 6-2 Ogern 0 4-0 4-0 0 0 Vimbodi 1-4 1-4 5-0 0 0 Vandellos 3-18 0 0 9-12 16-5 TOTAL 9 10 12 8 11

Habitat quality, fire size and connectivity explaining species post-fire colonisation

Local descriptors and connectivity were important in predicting the probability

of post-fire occurrence of open-habitat bird species (Table 4). For most of the species

both parameters seemed to be of equal importance, although this is not the case for the

Black-eared Wheatear and especially for the Woodlark for which surrogates of habitat

quality appeared to be of greater importance in determining species effective

colonisation (Fig. 2).

Artículo III

72

Exp

lain

ed v

aria

nce

(R2 )

0

10

20

30

40

50Connectivity Habitat quality

Linnet CornBunting

Woodlark Black-earedWheatear

Sardinian Warbler

Fig. 2 Total variance explained by connectivity and habitat quality variables analyzing the species probability of colonization.

The post-fire colonisation of all the studied species except the Woodlark was

significantly related to the amount of dispersal flux received from source populations,

indicating an important role of connectivity on species colonisation (Table 4). Transects

that were highly connected were more likely to be colonized. Additionally, the “quality”

of the transects was important in predicting probabilities of occurrence (Table 4),

although this relation depended on individual species. Aspect had a strong effect with

transects facing northwards showing a higher probability of being colonised by the

Sardinian Warbler and a lower probability by the Linnet and transects facing south

showing a higher probability of being colonised by the Black-eared Wheatear. The

colonisation of a transect was also significantly affected by slope; transects located in

flat areas were more likely to be colonised by the Woodlark and marginally related to

the Corn Bunting occurrence. The proportion of forest before the fire also affected the

probability of colonisation, with patches with a high proportion of forest before the fire

having a lower probability of being colonised by the Linnet and the Black-eared

Wheatear. Fire severity was only marginally related to the occurrence of the Black-

eared Wheatear. In addition, the probability of colonisation largely varied between fire

sites, as indicated by the large standard deviation of the random factor (Table 4). Fire

size and transect length did not contribute significantly when added to the models and

were consequently excluded from them. Final models explained 31% of the deviance in

colonisation (SD=10.68, n=5) (Table 4).

Artículo III

73

Table 4 Results from the GLMM, analyzing the influence of fire size, connectivity and habitat quality on the colonisation of recently burnt areas of five open-habitat bird species. The explained deviance (R2) and standard deviation of the random factor of the final model (that containing significant variables) also shown.

Species Coef. P R2

Linnet Connectivity 0.03 0.021 18.25 Pre-fire Vegetation -1.81 0.048 % North -2.81 0.051 Random effects Fire Site 1.42 Corn Bunting Connectivity 0.06 0.008 27.43 Slope -0.01 0.06 Random effects Fire Site 1.47 Woodlark Slope -0.16 0.022 29.14 Random effects Fire Site 2.70 Black-eared Wheatear Connectivity 0.08 0.003 32.61 Pre-fire Vegetation -3.79 0.003 Fire Severity 10.49 0.06 % South 3.79 0.003 Random effects Fire Site 1.98 Sardinian Warbler Connectivity 0.08 0.019 47.59 % North 4.26 0.015 Random effects Fire Site 3.28

DISCUSSION

The occurrence of colonisers in burnt areas was associated to the amount of

dispersal flux received from source populations and to the local heterogeneity in habitat

characteristics within fire sites. These results support the hypothesis that local and

regional processes determine bird colonisation of post-fire habitats in Mediterranean

landscapes (Pons and Bas 2005, Brotons et al. 2005).

The probability of post-fire colonisation of all the studied species except for the

Woodlark decreased with increasing isolation from neighbour’ source populations

(Hanski 1999, Hanski & Ovaskainen 2003). The greater the distance to potential

sources the lower the probability of being colonised, suggesting that dispersal capability

is crucial in locating new suitable areas (Johst et al. 2002). This result is consistent with

those of other studies documenting differences in the post-fire occurrence of coloniser

species between geographically distinct sites. Brotons et al. (2005) studied the dispersal

Artículo III

74

ability of open-habitat bird species to reach recently burnt areas. They reported that only

those burnt areas located close to nearby species populations were colonised,

highlighting the role of short-dispersal distance for colonisation. In the case of the

Woodlark, an increase in the regional population after the Atlas fieldwork period (i.e. in

the period 2002-2008) may explain our result (www.sioc.cat, ICO 2008), since it is

likely that more immigrants were available for species colonisation.

The probability of colonisation also depended on the local factors measured as

surrogates of habitat quality (Thomas et al. 2001). Habitat characteristics associated

with aspect and slope influenced species probability of colonisation. The vegetation

recovery and nest site location may account for this result. Additionally, species

occurrence in recently burnt areas also depended on the pre-fire habitat. Many previous

studies have highlighted the importance of pre-fire vegetation in studies of disturbances

effects on avifauna (Pons and Prodon 1996, Herrando et al. 2002). Zozaya et al. (In

press a) showed that shrubland burned areas contain more open-habitat bird species

than fires affecting forests, probably because open-habitat species were already present

before the fire (Pons and Prodon 1996). In contrast, fires that occurred in forested areas

include a high density of snags, thus preventing species colonisation (Moreira et al.

2003). Here, post-fire management activities might play an important role by converting

a burned forest structure to open grassland and shrubland, thus facilitating open-habitat

bird species colonisation (Castro et al. 2010). In addition, forested areas tend to be

associated with more productive sites (Thuiller et al. 2003), which may enhance

vegetation recovery and reduce the time window in which open habitats are potentially

available for colonisation by open-habitat birds. Nonetheless, not all forested areas

respond equally to fire (Rodrigo et al. 2004). The low regeneration capacity of Pinus

nigra after fire may lead to large temporal maintenance or increase of habitat suitable

for species linked to open vegetation (Zozaya et al. In press b). This may be particularly

important in the current fire regime context since these areas have experienced an

incremental impact of fires in the last decades (Pausas et al. 2008).

On the other hand, our findings on the effect of fire severity were in contrast to

our initial hypothesis. All the studied species except the Black-eared Wheatear showed

non-significant changes across the entire burn severity gradient. This result is in contrast

to those of forest studies that mostly found positive associations between the abundance

of bird species and increasing burn severity (Kirkpatrick et al. 2006, Koivula and

Schmiegelow 2007, Kotliar et al. 2007). The difference might be the consequence of a)

Artículo III

75

the majority of fires in the Mediterranean areas are typically characterised as severe

stand-replacing disturbances and thus, the mosaic of fire severities generated within the

burned areas is relatively small (Broncano and Retana 2004) and/or b) the location of

the transects avoid non-burnt areas and therefore the gradient characterised was a

moderate-high gradient. Given the current fire regime, we might expect fire severity to

play an important role in structuring animal and plant communities (Sousa 1984).

Unfortunately, this study faced important limitations associated to the severity gradient

considered. Further studies are needed to obtain more results of fire effects on post-fire

colonization by open-habitat birds in Mediterranean landscapes.

It should be pointed out that the species-factor relationships largely depended on

species requirements for habitat selection, indicating interspecific differences in

response to fire (McCauley 2007). Linnets were more likely to colonise shrubland

burned areas and avoid patches orientated north, whereas the Sardinian Warbler

occupied transects that were north facing. The Corn Bunting and the Woodlark had a

higher probability of colonisation in transects located in shallow areas. Lately, the

Black-eared Wheatear was the most sensible species being more likely to colonise

shrubland burned areas and occupied south-facing and high-severity transects. This

result illustrates how diverse post-fire colonisation patterns may be exhibited when local

factors are quantified and included in the analysis.

Finally, it is surprising that fire size did not influence in any of the five species

studied. In theory, large patches received more immigrants since they are more likely to

be found by migrants and/or immigrants are more likely to remain in a large patch once

they have arrived (Hill et al. 1996, Kuussaari et al. 1996). A recent work on the Ortolan

Bunting have found that fire size partly explains species occurrence after fire, albeit this

relationship was found to be weaker than that with isolation (Zozaya et al. submitted).

The difference with this other work might be due to the number of studied sites is three

times smaller than in this other study (15 versus 44) and additionally the fire size range

was again smalles (86-1438 ha versus 51-4497ha). Nonetheless, the fact that all the

species studied showed the same strong pattern suggests that is a more general

phenomenon and bring doubts about the effect of patch size on species colonisation. A

better knowledge of whether this parameter is a key element in species colonisation

needed to be further investigated.

In conclusion, this study provides further insight into the roles of local and

regional processes in the colonisation process of open-habitat bird species across the

Artículo III

76

habitat gradient appeared after fire. Connectivity plays an important role in the

colonisation process; however, habitat quality must also be considered in models to

ensure species colonisation. This is especially relevant since most of the declining

species in Mediterranean areas are open-habitat species (Prodon 2000) and fires are

becoming the main disturbance creating new suitable habitats at regional scale (Lloret et

al. 2002, Moreira & Russo 2007), thus being a key component in the species

distribution (Vallecillo et al. 2009).

ACKNOWLEDGEMENTS

We thank all the volunteers who participated in the Catalan Breeding Bird Atlas

(CBBA) and the ornithologists who participated in the bird surveys conducted in the

burned areas. We also thank L. Juarez for the laborious GIS work and comments by M.

Clavero in a previous draft of the manuscript. This work has received financial support

from the project CGL2005-00031/BOS granted by the Spanish Ministry of Education

and Science. E.L.Z. (FPI fellowship) received financial support from the Spanish


REFERENCES

Bibby C.J., Burgess N.D., Hill D.A., Mustoe S.H. (2000) Bird Census Techniques.

Second Edition Academic Press, London.

Bengtsson J. (2002) Disturbance and resilience in soil animal communities. European

Journal of Soil Biology 38: 119-125.

BirdLife International (2004) Birds in Europe: population estimates, trends and


Series No. 12).

Bjørnstad O.N. (2005) ncf: spatial nonparametric covariancefunctions. R package

version 1.0-8.

Broncano M.J., Retana J. (2004) Topography and forest composition affecting the

variability in fire severity and post-fire regeneration occurring after a large fire in

the Mediterranean basin. International Journal of wildland fire 13: 209-216.

Brotons L., Pons P., Herrando S. (2005) Colonization of dynamic Mediterranean

landscapes: where do birds come from after fire? Journal of biogeography 32: 789-

798.

Artículo III

77

Castro J., Moreno-Rueda G., Hódar J.A. (2010) Experimental test of postfire

management in pine forests: impact of salvage logging versus partial cutting and

non-intervention on bird–species assemblages. Conservation biology 24: 810-819.

Cea C., Cristóbal J., Serra P., Pons X. (2005) Mejoras en la detección semiautomática

de nubes y sombras en imágenes LANDSAT. XI Congreso Nacional de

Teledetección, 21-23 septiembre 2005

Cristoni C., Colangelo M.A., Ceccherelli V.U. (2004) Spatial scale and meiobenthic

copepod recolonisation: testing the effect of disturbance size in a seagrass habitat.

Journal of Experimental Marine Biology and Ecology 298: 49– 70.

Dale S. (2000) The importance of farmland for Ortolan Buntings nesting on raised peat

bogs. Ornis fennica 77: 17-25.

Dale S., Steifetten Ø., Osiejuk T.O., Losak K., Cygan J.P. (2006) How do birds search

for breeding areas at the landscape level? Interpatch movements of male Ortolan

bunting. Ecography 29: 886-898.

Díaz-Delgado R., Lloret F., Pons X., Terradas J. (2002) Satellite evidence of decreasing

resilience in Mediterranean plant communities after recurrent wildfires. Ecology 83:

2293-2303.

Díaz-Delgado R., Lloret F., Pons X. (2003) Influence of fire severity on plant

regeneration by means of remote sensing imagery. Int. J. Remote Sensing 24: 1751-

1763.

Estrada J., Pedrocchi V., Brotons L., Herrando S. (2004) Atles dels ocells nidificants de

Catalunya (1999-2002) (Lynx, ICO (Institut Català d'Ornitologia), Barcelona.

Götmark F., Blomqvist D., Johansson O.C., Bergkvist J. (1995) Nest site selection: a

trade-off between concealment and view of the surroundings? J. Avian Biol. 26:

305–312.

Hanski I. (1999) Metapopulation ecology. Oxford University Press, Oxford.

Hanski I., Ovaskainen O. (2003) Metapopulation theory for fragmented landscapes.


Herrando S., Brotons L., del Amo R., Llacuna S. (2002) Bird community succession

after fire in a dry Mediterranean shrubland. Ardea 90: 303-310.

Hill J.K., Thomas C.D., Lewis O.T. (1996) Effects of habitat patch size and isolation on

dispersal by Hesperia comma butterflies: implications for metapopulation structure.

Journal of animal ecology 65: 725-735.

Artículo III

78

ICO (2008) Programa SOCC. Sisè informe del Segiment d’Ocells Comuns a Catalunya.

ICO, Barcelona. http://www.ornitologia.org/monitoratge/reportsocc6.pdf

Jacquet K., Prodon R. (2009) Measuring the postfire resilience of a bird–vegetation

system: a 28-year study in a Mediterranean oak woodland. Oecologia 161: 801-811.

Johst K., Brandl R., Eber S. (2002) Metapopulation persistence in dynamic landscapes:

the role of dispersal distance. Oikos 98: 263-270.

Kirkpatrick C., Conway C.J., Jones P.B. (2006). Distribution and relative abundance of

forest birds in relation to burn severity in southeastern Arizona. J. Wildl. Mgmt. 70,

1005–1012.

Kiss L., Magnin F. (2003) The impact of fire on some Mediterranean land snail

communities and patterns of post-fire recolonization. J. Moll. Stud. 69: 43-53.

Koivula M.J., Schmiegelow F.K.A. (2007) Boreal woodpecker assemblages in recently

burnt forested landscapes in Alberta, Canada: effects of post-fire harvesting and

burn severity. Forest Ecology Management 242: 606-618.

Kotliar N.B., Kennedy P.L., Ferree K. (2007) Avifauna responses to fire in

southwestern montane forests along a burn severity gradient. Ecological

Applications, 17: 491–507.

Kuussaari M., Nieminen M., Hanski I. (1996) An experimental study of migration in the

Glanville fritillary butterfly Melitaea cinxia. Journal of animal ecology 65: 791-

801.

Lindberg N., Bengtsson J. (2005) Population responses of oribatid mites and

collembolans after drought. Applied Soil Ecology 28: 163–174.

Llimona F., Matheu E., Prodon R. (1993) Role of snag persistence and of tree

regeneration in postfire bird successions: comparison of pine and oak forests in

Montserrat (Catalonia, NE Spain). In, L. Trabaud & R. Prodon (Eds.): Fire in

Mediterranean Ecosystems, pp. 315-331. Commission of the European

Communities. Bruxelles.

Lloret F., Calvo E., Pons X., Díaz-Delgado R. (2002) Wildfires and landscape patterns

in the Eastern Iberian Peninsula. Landscape Ecology 17: 745-759.

Magoulick D.D., Kobza R.M. (2003) The role of refugia for fishes during drought: a

review and synthesis. Freshwater Biology 48: 1186-1198.

Mather P.M. (1999) Computer processing of remotely sensed images. 2nd edn.

JohnWiley & Sons, New York, New York, USA.

Artículo III

79

McCauley S.J. (2007) The role of local and regional processes in structuring larval

dragonfly distributions across habitat gradients. Oikos 116: 121-133.

McCullagh P., Nelder J.A. (1989) Generalized Linear Models. Monographs on

Statistics and Applied Probability. Chapman & Hall, London, UK.

Menz M.H.M., Brotons L., Arlettaz R. (2009) Habitat selection by Ortolan bunting

Emberiza hortulana in post-fire succession in Catalonia: implications for the

conservation of farmland populations. Ibis 151: 752-761.

Ministerio de Medio Ambiente. 2006. Tercer Inventario Forestal Nacional. Barcelona.

Dirección General de Conservación de la Naturaleza, Madrid.

Moreira F., Delgado A., Ferreira S., Borralho R., Oliveira N., Inácio M., Silva J.S.,

Rego F. (2003) Effects of prescribed fire on vegetation structure and breeding birds

in young Pinus pinaster stands of northern Portugal. Forest Ecology and

Management 184: 225-237.

Moreira F., Russo D. (2007) Modelling the impact of agricultural abandonment and

wildfires on vertebrate diversity in Mediterranean Europe. Landscape ecology 22:

1461-1476.

Naveh Z. (1975) The evolutionary significance of fire in the Mediterranean region.

Vegetatio 29: 199-208.

Norkko A., Rosenberg R., Thrush S.F., Whitlatch R.B. (2006) Scale- and intensity-

dependent disturbance determines the magnitude of opportunistic response. Journal

of Experimental Marine Biology and Ecology 330: 195–207.

Pala V., Pons X. (1995) Incorporation of relief in polynomial-based geometric

corrections. Photogrammetric engineering and remote sensing 61: 935-944.

Pausas J.G., Llovet J., Rodrigo A., Vallejo R. (2008) Are wildfires a disaster in the

Mediterranean basin?-A review. International Journal of Wildland Fire 17: 713-

723.

Pickett S.T.A., White P.S. (1985) The ecology of natural disturbance and patch

dynamics. Academic Press, London.

Pinheiro J.C., Bates D.M. (2000) Mixed-effects Models in S and SPLUS. Springer-

Verlag, New York, USA.

Pons P., Prodon R. (1996) Short term temporal patterns in a Mediterranean shrubland

bird community after wildfire. Acta Oecologica 17: 29-41.

Pons P., Bas J.M. (2005) Open-habitat birds in recently burned areas: The role of the

fire extent and species’ habitat breadth. Ardeola 52: 119-131.

Artículo III

80

Pons X., Solé-Sugrañes L. (1994) A simple radiometric correction model to improve

automatic mapping of vegetation from multispectral satellite data. Remote Sensing

of Environment 48: 191-204.

Prodon R., Fons R., Peter A.M. (1984) L’impact du feu sur la végétation, les oiseaux et

les micromammifères dans diverses formations des Pyrénées-Orientales : premiers

résultats. Revue d'Ecologie (la Terre et la Vie) 39: 129-158.

Prodon R (2000) Landscape dynamics and bird diversity in the north Mediterranean

basin: conservation issues. In: Trabaud L (ed) Life and environment in the

Mediterranean. WIT Press, Southampton, pp 261–299.

Rodrigo A., Retana J., Picó X.F. (2004) Direct regeneration is not the only response of

Mediterranean forests to large fires. Ecology 85: 716-729.

Saura S., Pascual-Hortal L. (2007) A new habitat availability index to integrate

connectivity in landscape conservation planning: Comparison with existing indices

and application to a case study. Landscape Urban Plann. 83: 91-103.

Saura S., Torné J. (2009) Conefor Sensinode 2.2: a software package for quantifying the

importance of habitat patches for landscape connectivity. Environmental Modelling

and Software 24: 135-139.

Saura S., Rubio L. (2010) A common currency for the different ways in which patches

and links can contribute to habitat availability and connectivity in the landscape.

Ecography 33: 523-537.

Sousa W.P. (1984) The role of disturbance in natural communities. Annual Review of

Ecology and Systematics 15: 353-391.

Terradas J., Piñol J. (1996) Els grans incendis: condicions meteorològiques i de

vegetació per al seu desenvolupament. In: Terradas J. (ed.), Ecologia del Foc. Proa,

Barcelona, España, pp. 63-75.

Thomas J.A., Bourn N.A.D., Clarke R.T., Stewart K.E. Simcox D.J., Pearman G.S.,

Curtis R., Goodger B. (2001) The quality and isolation of habitat patches both

determine where butterflies persist in fragmented landscapes. Proc. R. Soc. Lond. B

268: 1791-1796.

Thuiller W., Vayreda J., Pino J., Sabate S., Lavorel S., Gracia, C. (2003) Large-scale

environmental correlates of forest tree distributions in Catalonia (NE Spain). Global

Ecology and Biogeography 12: 313-325.

Artículo III

81

Turner M.G., Romme W.H., Gardner R.H., Hargrove W.W. (1997) Effects of fire size

and pattern on early succession on Yellowstone national park. Ecological

Monographs 67: 411-433.

Urban D., Keitt T. (2001) Landscape connectivity: a graph theoretic perspective.

Ecology 82: 1205-1218.

Vallecillo S., Brotons L., Thuiller W. (2009) Dangers of predicting bird species

distributions in response to land-cover changes. Ecological Applications 19: 538-

549.

Whittingham M.J., Percival S.M., Brown A.F. (2002) Nest site selection by Golden

Plover: why do shorebirds avoid nesting on slopes? J. Avian Biol. 33: 184-190.

Wiens J.A., Stenseth N.C., Van Horne B., Ims R.A. (1993) Ecological mechanisms and

landscape ecology. Oikos 66: 369-380.

Wiens J.A. (2001) The landscape context of dispersal. Dispersal (ed. by J. Clobert, E.

Danchin, A.A. Dhondt & J.D. Nichols), pp. 96-109. Oxford Univ. Press.

Zozaya E.L., Brotons L., Herrando S., Pons P., Rost J., Clavero M. (in press a)

Monitoring spatial and temporal dynamics of bird communities in Mediterranean

landscapes affected by large wildfires. Ardeola

Zozaya E.L., Brotons L., Vallecillo V. (in press b) Bird community responses to

vegetation heterogeneity following non-direct regeneration of Mediterranean forests

after fire. Ardea

83

4.4. Artículo IV

IV- BIRD COMMUNITY RESPONSES TO VEGETATION

HETEROGENEITY FOLLOWING NON-DIRECT REGENERATION OF

MEDITERRANEAN FORESTS AFTER FIRE

Artículo IV

84

SUMMARY

Mediterranean forest has been shown to be highly resilient to fires, showing a rapid

recovery after disturbance. However, in some cases tree regeneration fails in direct

regeneration leading to radical changes in landscape composition. In this study, we analyzed

the effect of variability in post-fire regeneration patterns on a Mediterranean bird community

and evaluated the impact of landscape changes on the conservation value of the bird species

using the new diverse landscape mosaic arising from non-direct regeneration after a large

forest fire. We used data from a large fire that occurred in the central Catalonia (NE Spain) in

1998. The fire affected about 26,000 ha of a land mosaic mainly covered by Black Pine Pinus

nigra forests and farmland dominated by cereal crops. We used line transects to estimate bird

abundance and gathered information on dominant vegetation covers and landscape variables.

Redundancy analysis (RDA) and generalized linear models were used to explore how the

measured environmental variables explain bird species data and to analyze to which degree

post-fire heterogeneity in vegetation affected the conservation value of the bird community

that appears after the fire. Factors describing main patterns in post-fire landscape explained up

to 31.2% of the total variability in bird community composition and described three main

groups of bird species sharing similar ecological requirements. Additionally, 71% of the

studied species significantly responded to one of the three first vegetation gradients

distinguished in the study area. Finally, conservation value of the bird community

significantly decreased in areas dominated by Q. humilis resprouters and significantly

increased in shrubland areas. Overall, our results suggest that large fires affecting non-direct

regenerating forest types lead to a new and radically different mosaic landscape offering new

opportunities to species with unfavourable conservation status at the European level.

Keywords: Conservation value, large fires, open-habitat bird species, Pinus nigra,

regeneration patterns, shrubland

Artículo IV

85

INTRODUCTION

Fires are the most important natural disturbance in Mediterranean regions, and exert a

decisive role in the dynamics and structure of plant and animal communities (di Castri &

Mooney 1973, Picket & White 1985, Whelan 1995). The effect of fires on biological diversity

is highly variable and depends, among other factors, on fires’ characteristics such as extent

and intensity (Sousa 1984), the initial state of the ecosystem (Christensen 1993, Foster et al.

1998) and the biotic and abiotic environment (Foster et al. 1998).

Due to their ability to fly, birds can often avoid the direct effects of flames through

moving into adjacent habitats not affected by the passage of fire (Lawrence 1966).

Nevertheless, the drastic modification of habitat has important consequences on the re-

colonisation of the burnt area. Due to their strong site tenacity, philopatry, habitat tolerance,

and to the persistence of standing dead trees, some forest bird species can return to burnt

territories the first breeding season after fire, but they progressively leave their breeding

territories and do not reappear until the vegetation attains a woody appearance (Prodon et al.

1987, Pons & Prodon 1996). Apart from these cases, the post-fire succession of bird

communities is closely linked with vegetation recovery, first starting with open-space species,

then shrubland species and finally forest species (Prodon et al. 1984, Jacquet & Prodon 2009).

Overall, forest bird species are the most damaged by fire (Ukmar et al. 2007) while open-

habitat species seems to have greatly benefited from it (Pons & Bas 2005). In fact, recent

literature has highlighted the role of burnt areas in Mediterranean region in the maintenance

of open-habitat bird species populations (Brotons et al. 2008, Vallecillo et al. 2009). This is

especially relevant since these species are among the most threatened species in Europe

(Birdlife International 2004). Pons & Bas (2005) showed that 17 out of 22 open-habitat bird

species used recently burnt areas in Iberia and southern France had an unfavourable

conservation status in Europe. Thus, beyond the direct effect of fire, the structure of bird

communities after fire seems to be highly dependent on the effect of fire on habitat

composition, increasing species diversity with habitat heterogeneity and time since

disturbance (Herrando et al. 2002, Herrando et al. 2003, Ukmar et al. 2007, Vallecillo et al.

2008).

In this sense, it is widely accepted that Mediterranean vegetation is highly resilient to

fire effects; this means that the same pre-disturbance community is restored only a few

decades after the disturbance (Hanes 1971, Lloret 1998). Nevertheless, recent studies showed

that heterogeneous landscapes can be originated in relatively homogeneous forest areas after a

large forest fire, in cases in which dominant pre-fire tree forest species fail in direct

Artículo IV

86

regeneration (e.g. large forest fires of Black Pine Pinus nigra in Catalonia, North-Eastern

Spain, Retana et al. 2002; Rodrigo et al. 2004). Under such circumstances, the change in

forest cover results in post-fire environmental conditions completely different from those of

unburnt Pinus nigra forests. Hereafter, we refer to this process as a non-direct regeneration

event. In addition, the low colonisation ability of Pinus nigra (Ordoñez et al. 2004) allows

differences in forest structure to prevail for decades, favouring persistence of new appearing

species. Since bird communities respond to changes in vegetation composition and structure

caused by fire (Prodon & Lebreton 1981), these landscape changes, originated in a non-direct

regeneration scenario, are expected to enhance bird diversity and favour the persistence of

colonisers.

To date, most studies on the effects of fire on birds in the Mediterranean Basin have

been carried out in study areas highly resilient to fire where local community return to its

former state after fire disturbance (e.g. Prodon & Lebreton 1981, Prodon et al. 1984, Pons &

Prodon 1996, Herrando et al. 2002, Jacquet & Prodon 2009). However, to the best of our

knowledge, there is no study analyzing post-fire bird community under a non-direct

regeneration scenario. In this work, we examined to which degree spatial differences in bird

community structure appearing after fire track patterns of vegetation recovery. In the case of

the diverse landscape mosaic arising as response to the lack of direct regeneration of pre-fire

dominating pines, we expected to find a mosaic of bird communities matching the

heterogeneity in vegetation. Finally, we analyzed to what extent post-fire heterogeneity in

vegetation recovery affects the conservation value of bird community.

METHODS

Study area

The study area is located in the Solsonès county (between 41º59’ and 41º44’ North

and 1º21’ and 1º39’ East, Lleida, North-Eastern Spain), in an area characterized by a marked

altitudinal gradient, decreasing towards the south, with altitudes that range from 450 to 950m

above sea level. In July 1998, several fires burnt around 26,000 ha (Fig. 1), mostly affected by

crown fire (sensu Turner et al. 1994, 1997); that is, all trees were killed and canopy needles

were completely burnt.

According to the data collected in 1993 for the Ecological Forest Inventory of

Catalonia (IEFC) 67% of the total burnt area affected forested lands with the remaining land

dominated by cereal crop-fields (Gracia et al. 2000). Hence, burnt area comprised a

continuous forest mass on sloping areas with agricultural patches located in flat terrains. The

Artículo IV

87

main forest species affected were Black Pine Pinus nigra (74%) together with Aleppo Pine

Pinus halepensis (11%), with Holm Oak Quercus ilex and the deciduous species Lusitanian

Oak Quercus faginea. The understory was mainly covered by Downy Oak Quercus humilis.

Pinus nigra is a non-resprouter species with a regeneration strategy based on germination. As

its seeds are dispersed in spring (Skordilis & Thanos 1997, Alvarez et al. 2007), summer fires

prevent the regeneration and recovery of stands of this species, leading to regeneration of a

different type of forest dominated by resprouting species such as oaks (Habrouk et al. 1999,

Rodrigo et al. 2004). Concretely in the study area, the forest landscape changed to a mosaic of

different habitats dominated by different Quercus species, shrubland and open grasslands with

some remains of unburnt Pinus nigra (Retana et al. 2002); increasing habitat heterogeneity.

Field surveys

We used line transects to estimate bird presence and abundance (Bibby et al. 2000). In

this method, observer travels along a line recording bird species and abundance. Each census

lasted for 20 minutes and covered about 700m in length (range 602-820m). Birds were

counted, when heard or seen, within 100m belts on both sides of the track. Censuses were

conducted in 2005, 7 years after the fire. Each transect was surveyed twice, with one visit in

the early breeding season (19 April-24 May) and one in the middle of the breeding season (24

May- 24 June), allowing approximately 1 month between visits to the same transect. The

higher of the two counts per species was used as the dependent variable for further analysis.

Raptors, aerial feeders (swallows, swifts and bee-eaters) and crepuscular species were

excluded from the analysis because this method is not appropriate to assess their abundance

(Bibby et al. 2000).

Transects were distributed within the burnt area using a random stratified sampling.

First, we randomly located 25 points within the fire perimeter. At each of these points

(approximately 2km radius), 4 survey transects were defined (Fig. 1). The criteria used to

select transect location around each point were: 1) Transects were preferably and completely

conducted across burnt natural habitat and therefore avoiding large patches of non-burnt

forest or farmland along the track. 2) Transects had to be easily accessible from walking trails.

3) Transects had to represent the main burnt habitat types occurring near the random point. 4)

The minimum distance between transects was 200m. All bird surveys were performed by the

same observer, to avoid likely observer errors, and were always conducted in good weather

conditions (i.e. without rainfall or strong wind). All transects were conducted within 3 hours

from sunrise.

Artículo IV

88

Fig. 1 Map of the large forest fire occurred in the Solsonès county in 1998. Fire perimeter (grey line) and transects locations (black lines) are shown. The area shown represents burnt habitats, with grey representing burnt forests and yellow accounting for agricultural areas (mainly cereals). The map on the left shows the geographical location of the study area in the North-eastern Spain (in black).

Additionally, the observer also described landscape and habitat post-fire conditions.

Habitat characteristics were recorded along each transect using a modification of the cover

estimation method proposed by Prodon & Lebreton (1981), which involves a visual

estimation of the relative percentage cover of each variable within a defined area, in this case

the transect. The following vegetation layers were measured: bare ground, rock cover,

herbaceous vegetation (0-0.25m), shrubby vegetation (0.25-1m) and an overall assessment of

the cover of three regenerating tree species (P. halepensis, Q. ilex and Q. humilis). These

covers were taken to be representative of the whole length of transects, including a 100m belt

on both sides. They were recorded in both census visits and mean values were used as

explanatory variables for further analysis. A recent work carried out in the same study area

have shown that our field vegetation cover estimates reliably represented major components

of variability in vegetation cover along the transects by using Satellite Landsat data

(Normalized Difference Vegetation Index, NDVI, Pettorelli et al. 2005) (Menz et al. 2009).

Artículo IV

89

Finally, considering the importance of landscape variables on their contribution to

explain ecological processes such as regeneration pattern (Turner et al., 1994) and fauna

distribution (Izhaki & Adar 1997, Vallecillo et al. 2008, Menz et al. 2009), the following

variables were estimated within 250 m belts of the transects: Slope was calculated using a

Digital Elevation Model (DEM) generated from 1:50 000 topography maps. Aspect was

measured as the proportion of south and north-facing pixels. The relative abundance of

surrounding agricultural fields, unburned patches, standing and laying dead burnt trunks and

streams within the 100m belt at each side of the transect were estimated in the field

(categorical, 1-3 with increasing abundance). The stream character of the transects was

recorded as it has been previously observed that these areas often exhibit higher vegetation

recovery rates.

Data analyses

The relationship between the bird species and the environmental variables was

analyzed. Due to relatively large number of environmental variables (initially 14) we firstly

performed a Principal Components Analysis (PCA, Statistica V.6) (StatSoft, Inc. 2001) to

reduce number of variables and reduce colinearity between variables. We used a PCA with a

varimax normalized rotation. This procedure maximizes the correspondence between the

factors and the original variables. We retained the minimum number of components where all

original variables where represented and used them for further analyses.

We performed a Redundancy Analysis (RDA; ter Braak 1986) to relate bird

community with main factors described in the PCA. This method assumes that species are

responding linearly to the ordination axis. Species linear response was tested by using a

Detrended Correspondence Analysis (DCA). Species abundance was Log-transformed in

order to prevent skewed distribution and Monte-Carlo permutation test was performed to

determine the significance of the first ordination axis and that of all canonical axes together.

This RDA analysis was run within the CANOCO program version 3.2 (ter Braak 1988).

Additionally, species-specific analyses were performed using generalized linear models

(GLM) and generalized linear mixed models (GLMM) to describe ecological requirements for

each species (Dobson 1990), (R Development Core Team 2008). We used the abundance of

each species found in each transect as the response variable and the main vegetation gradients

described by the PCA as fixed factors for this analysis. We used a Poisson distribution and a

log-link function for the dependent variable. In the cases where there was a significant site

effect on species distribution we performed a GLMM considering site as a random factor in

Artículo IV

90

the analysis. However, we performed a GLM when site did not have an effect on the species

distribution. Either for the RDA analysis to the GLM and GLMM analysis explained above,

we only considered those species recorded in more than 5% of the transects (see Appendix).

Finally, we analyzed to which degree post-fire heterogeneity in vegetation recovery

affects the conservation value of the bird community that appeared 7 years after the fire. With

this purpose, we used an index (Equation 1) which takes into account the conservation status

and abundance of the species recorded in each transect (Pons et al. 2003). The status was

based on the classification of Birdlife International (2004) in categories of “Species of

European Conservation Concern”, hereafter SPEC. A SPEC value was assigned to each

species in geometric progression of increasing conservation concern (SPEC valuei: Non-

SPEC =1, SPEC-3 =2, SPEC-2 =4, SPEC-1 =8). In the present study, species belonging to all

categories except SPEC-1 were recorded (see Appendix). Abundance was logarithm-

transformed to balance its contribution to the global index.

[ ]∑=

×+=k

iii valueSPECAIndexonConservati

1)1(log (1)

Where k is the species richness and Ai the abundance of species i recorded in each transect and

relative to an area of 1 ha.

The influence of main vegetation gradients on the Conservation Index relative to each transect

was then approached by a GLM. We used the Conservation Index on each transect as the

response variable and the main vegetation gradients described by the PCA as fixed factors.

Normal distribution of the dependent variable was checked by using plot package with R

software.

RESULTS

Post-fire regeneration patterns

The principal component analysis summarised environmental variability in post-fire

regeneration patterns in seven factors explaining 80% of the original variability in the

regeneration patterns (Table 1). The first factor (Quercus humilis regeneration) explained

about 34% of the variability and represented a gradient of decreasing regeneration of Q.

humilis, dominating northern slopes, to areas with virtually no regeneration of this deciduous

oak species and increasing surface of bare soil in southern dominated slopes. The second

factor (Farmland) separated zones characterized by the presence of farmland, extensive cereal

fields and not burnt patches in flatter areas from those homogenously burnt in more abrupt

topography. The third factor (Shrubland) represented a gradient of decreasing shrub cover

Artículo IV

91

with increasing low, herbaceous vegetation. The fourth factor (Pine regeneration) was related

to the pine resprouters of the species P. halepensis. The fifth factor (Standing trunks) was

related mainly to the amount of standing and laying dead burnt trunks generated in post-fire

salvage logging and management activities undergone during the first two years after fire. The

sixth factor (Stream) was related to the presence of riparian areas in which vegetation had

recovered rapidly but only locally along the stream sides. Finally the last factor (Quercus ilex

regeneration) identified locations with strong re-sprouting of Q. ilex.

Table 1 Principal Components Analysis performed on habitat characteristics and landscape variables. Main contributing variables are given in bold letters.

F1

(Q.humilis reg.)

F2 (Farm-land)

F3 (Shrub-land)

F4 (Pine reg.)

F5 (Standing

trunks)

F6 (Stream)

F7 (Q. ilex Reg.)

HABITAT CHARACTERISTICS

Rock cover -0.64 -0.14 0.37 -0.11 0.13 0.14 0.43 Bare ground -0.65 -0.27 0.50 -0.04 -0.01 -0.15 0.06 Herbaceous vegetation -0.06 -0.07 -0.84 -0.05 -0.15 0.22 -0.04 Shrub vegetation -0.33 -0.08 0.75 0.01 0.07 0.16 0.15 Quercus ilex resprouts -0.24 0.14 0.13 0.05 0.01 0.01 0.87 Pinus halepensis resprouts -0.08 -0.06 0.03 0.94 -0.03 -0.11 0.04 Quercus humilis resprouts 0.82 0.17 -0.22 -0.15 0.09 -0.21 -0.11 LANDSCAPE VARIABLES

Slope -0.35 -0.57 0.44 0.13 0.02 0.27 -0.13 North orientation 0.84 -0.17 -0.12 0.04 -0.08 -0.02 0.03 South orientation -0.79 -0.06 0.23 -0.14 0.05 -0.07 -0.02 Agricultural fields -0.01 0.78 -0.14 -0.12 0.04 -0.14 0.35 Unburnt forest 0.07 0.87 0.13 0.05 -0.01 0.06 -0.13 Stream 0.20 -0.09 -0.08 -0.12 -0.02 0.90 0.02 Standing dead burnt trunks 0.16 0.07 0.09 -0.20 0.89 -0.04 0.13 Laying dead burnt trunks -0.32 -0.10 0.18 0.39 0.70 0.03 -0.20 Eigenvalue 6.27 2.13 1.85 1.38 1.05 0.91 0.84 % Total variance 34.86 11.81 10.30 7.69 5.83 5.07 4.64 Seven factors loading for each individual variable were obtained using a varimax normalized rotation. The percentage of accumulated variation is 80.2%

Bird community

Overall, we found that factors describing main patterns in post-fire landscape

explained up to 31.2% of the total variability in bird community composition. A plot of the

first two axes of the RDA is shown in Figure 2. The first axis reflected a gradient of

Artículo IV

92

vegetation structure from shrub lands to deciduous Q. humilis resprouters. The second axis

separated farmland from the remainder land use types. The plot described three main species

groups that share similar ecological requirements. The first group was characterised by

species using areas where resprouters of Quercus species dominated the post-fire

regeneration, especially in north-facing slopes, and included the Melodious Warbler

(Hippolais polyglotta), the Subalpine Warbler (Sylvia cantillans) and the Blackcap (Sylvia

atricapilla). The second group described the bird community using burnt mosaic areas

prevailing near farmland in which small patches of non-burnt pine forests were still present

after the fire and was characterised by species such as the Turtle Dove (Streptopelia turtur),

the Woodchat Shrike (Lanius senator) and the Golden Oriole (Oriolus oriolus). Finally, a

third group defined open habitat species using sparse vegetation with very poor or no post-fire

tree regeneration such as the Ortolan Bunting (Emberiza hortulana), the Tawny Pipit (Anthus

campestris) and the Black-eared Wheatear (Oenanthe hispanica).

Fig.2 Associations estimated between the bird species and the environmental variables by the Redundancy Analysis. Only the bird species which are well characterized by the first two axes are shown. Complete names of the species acronyms are given in Appendix.

In bird specific analysis, we found that 71% of the studied species significantly

responded to one of the three first vegetation gradients distinguished in the PCA (Table 2).

Whereas the Q. humilis regeneration gradient had contradicting effects on different species (9

species responding positively to the gradient and 8 negatively), the farmland gradient had a

Artículo IV

93

generally positive effect on many species with up to 18 species positively affected by this

gradient and only 5 negatively related to it (Table 2). Also the shrubland gradient tended to

positively influence more species (9) than exert negative influences (only 3). The regeneration

of other trees rather than deciduous oaks, namely pines and Holm Oaks tended to have strong

negative effects on species with 4 and 7 negatively and only 1 and 1 positively related to pine

and Holm Oak respectively. Finally, the presence of specific features such standing dead

trunks or streams on the transects tended to have a positive effect on bird abundance with 5

and 9 positively and only 1 and 1 species negatively related to these two gradients

respectively (Table 2).

On the other hand, the Conservation Index of the bird species was significantly related to

post-fire regeneration patterns (F = 7.06; df = 7, 92; P < 0.0001), indicating different bird

conservation value depending on post-fire vegetation patterns. In order to further analyse the

possible relation between vegetation patterns and conservation objectives, we considered only

those vegetation gradients estimated in the PCA that explained slightly more than 50% of

total variation (i.e. F1 (Q. humilis regeneration); F2 (Farmland); and F3 (Shrubland)). In this

sense, the Conservation Index of the bird species in the burnt area significantly decreased in

areas where resprouters of Q. humilis dominate (F = 25.02; d.f. = 1, 98; P < 0.0001) and

significantly increased in shrubland areas (F = 12.32; d.f. = 1, 98; P < 0.001) (Fig. 3). On the

other hand, the Conservation Index of the bird species was not significant related to the

farmland habitat type (P > 0.05) (Fig. 3).

94

Artí

culo

IV

Tab

le 2

Rel

atio

nshi

p be

twee

n bi

rd s

peci

es a

nd th

e se

ven

fact

ors

obta

ined

by

the

PCA

as

test

ed w

ith g

ener

aliz

ed li

near

mod

el a

nd g

ener

aliz

ed li

near

mix

ed

mod

elsa .

Spec

ies n

ame

Que

rcus

hum

ilis

rege

nera

tion

Farm

land

Shru

blan

d Pi

ne

rege

nera

tion

Stan

ding

trun

ks

Stre

amQ

uerc

us il

ex

rege

nera

tion

Ran

dom

Fa

ctor

b

Red

-legg

ed P

artri

dge

(Ale

ctor

is ru

fa)

n.

s W

ood

Pige

on (C

olum

ba p

alum

bus)

+++

n.

s Tu

rtle

Dov

e (S

trep

tope

lia tu

rtur

)

+++

*

* C

omm

on C

ucko

o (C

ucul

us c

anor

us)

n.

s H

oopo

e (U

pupa

epo

ps)

++

n.s

Wry

neck

(Jyn

x to

rqui

lla)

+ n.

s G

reen

Woo

dpec

ker (

Picu

s vir

idis

) +

++

+

n.

s G

reat

Spo

tted

Woo

dpec

ker

(Den

droc

opos

maj

or)

n.

s

Thek

la L

ark

(Gal

erid

a th

ekla

e)

*

* C

rest

ed L

ark

(Gal

erid

a cr

ista

ta)

+

- n.

s W

oodl

ark

(Lul

lula

arb

orea

)

n.s

Skyl

ark

(Ala

uda

arve

nsis

)

n.s

Taw

ny P

ipit

(Ant

hus c

ampe

stri

s)

-

n.

s W

ren

(Tro

glod

ytes

trog

lody

tes)

+++

n.

s N

ight

inga

le (L

usci

nia

meg

arhy

ncho

s)

+++

+++

-

++

+

* St

onec

hat (

Saxi

cola

torq

uata

) -

n.s

Bla

ck-e

ared

Whe

atea

r (O

enan

the

hisp

anic

a)

- - -

- ++

+

*

*

Roc

k Th

rush

(Mon

ticol

a sa

xatil

is)

- - -

- - -

++

-

* *

Bla

ckbi

rd (T

urdu

s mer

ula)

++

++

* M

elod

ious

War

bler

(Hip

pola

is p

olyg

lotta

) ++

+

- - -

-

+++

*

* D

artfo

rd W

arbl

er (S

ylvi

a un

data

) - -

- - -

- ++

+

- *

* Su

balp

ine

War

bler

(Syl

via

cant

illan

s)

+++

++

*

* Sa

rdin

ian

War

bler

(Syl

via

mel

anoc

epha

la)

+

- - -

*

95

Artí

culo

IV

Tab

le 2

C

ontin

ued

Spec

ies n

ame

Que

rcus

hum

ilis

rege

nera

tion

Farm

land

Shru

blan

d Pi

ne

rege

nera

tion

Stan

ding

trun

ks

Stre

amQ

uerc

us il

ex

rege

nera

tion

Ran

dom

Fa

ctor

b

Orp

hean

War

bler

(Syl

via

hort

ensi

s)

++

+

n.s

Bla

ckca

p (S

ylvi

a at

rica

pilla

) ++

+

- -

- +

+++

*

Bon

elli’

s War

bler

(Phy

llosc

opus

bon

elli)

+

+++

+

*

* Lo

ng-ta

iled

Tit (

Aegi

thal

os c

auda

tus)

+

++

n.

s C

rest

ed T

it (P

arus

cri

stat

us)

+ ++

+ ++

- -

-

* B

lue

Tit (

Paru

s cae

rule

us)

+++

+ +

n.s

Gre

at T

it (P

arus

maj

or)

++

++

*

Gol

den

Orio

le (O

riol

us o

riol

us)

++

+

* *

Gre

at G

rey

Shrik

e (L

aniu

s mer

idio

nalis

)

+

n.s

Woo

dcha

t Shr

ike

(Lan

ius s

enat

or)

++

+

-

* Ja

y (G

arru

lus g

land

ariu

s)

++

* *

Car

rion

Cro

w (C

orvu

s cor

one)

++

n.

s St

arlin

g (S

turn

us v

ulga

ris)

+

- -

- n.

s R

ock

Spar

row

(Pet

roni

a pe

tron

ia)

+

- -

n.s

Cha

ffin

ch (F

ring

illa

coel

ebs)

+++

n.

s Se

rin (S

erin

us se

rinu

s)

*

Gol

dfin

ch (C

ardu

elis

car

duel

is)

- -

n.s

Linn

et (C

ardu

elis

can

nabi

na)

- - -

+

* *

Cirl

Bun

ting

(Em

beri

za c

irlu

s)

++

* R

ock

Bun

ting

(Em

beri

za c

ia)

-

+

*

* O

rtola

n B

untin

g (E

mbe

riza

hor

tula

na)

- - -

- +

-

* *

Cor

n B

untin

g (E

mbe

riza

cal

andr

a)

++

+++

- - -

- -

- *

* a

Neg

ativ

e re

latio

nshi

ps: (

-) p

<0.5

, (- -

) p<0

.01,

(- -

-) p

<0.0

01; P

ositi

ve re

latio

nshi

ps: (

+) p

<0.5

, (++

) p<0

.01,

(+++

) p<0

.001

. b

Ana

lysi

s of t

he si

te e

ffec

t on

spec

ies d

istri

butio

n: n

.s (n

o si

gnifi

cant

), *

(p<0

.5),

** (p

<0.0

1)

Artículo IV

96

Fig. 3 Relationship between bird Conservation Index and main factors describing vegetation recovery obtained by the PCA. Significant differences are indicated in the top right with n.s. (no significant) or ** (p<0.01).

DISCUSSION

With this work, we have shown that bird community strongly responded to post-fire

heterogeneity in vegetation recovery after fire. The landscape heterogeneity arising from the

variable patterns of vegetation regeneration creates a mosaic of suitable habitats for bird

species with different habitat requirements. From the conservation point of view, especially

important are the open habitats originated in areas where non-tree regeneration takes place

leading to areas dominated by open shrub lands.

Bird species respond to changes in the habitat structure and composition such as the

ones generated after fire (e.g. Prodon & Lebreton 1981, Hobson & Schieck 1999, Brawn et al.

2001). In Mediterranean region, landscapes are assumed to be highly resilient to fire (Trabaud

& Lepart 1980, Lloret et al. 1999). Thus, depending on the vegetation type bird composition

and richness can recover after 1 year in dry grassland, some 15 years in a Cork Oak Quercus

suber forest and probably around one century in a mature Evergreen Oak Quercus ilex forest

(Prodon et al. 1984, Prodon 1988, Jacquet & Prodon 2009). However, in cases in which

ecosystems do not return to pre-fire conditions such as in Pinus nigra forests, post-fire

environmental conditions are expected to determine bird structure inducing important changes

on bird composition and richness. In this study we analyzed, for the first time, the bird

community found 7 years after a large fire that affected a Pinus nigra forested area. Our

results have shown that post-fire bird community is strongly associated to vegetation recovery

even when strong vegetation changes occurred due to non-direct regeneration of dominant

forest tree species after a large fire.

Artículo IV

97

The new mosaic of habitats created 7 years after fire facilitates the appearance of a

large variety of bird species matching the main regeneration patterns. This result is in

agreement with previous studies suggesting that habitat heterogeneity at local scale enhance

the occurrence of a rich bird community (Dunning et al. 1992, Brotons et al. 2005). This

association between spatial heterogeneity and bird richness has been repeatedly reported

(Wiens 1989). In our case, we found three main landscape gradients determining spatial

variability of bird community structure. The principal one showed the incipient formation of a

new forest dominated by young resprouters of Q. humilis prevailing mainly in the north-

facing slopes (Espelta et al. 2003). In this habitat we found a low number of specialist bird

species, mainly undergrowth species, such as the Blackcap and the Subalpine Warbler. These

species find availability of enough trophic resources, breeding sites and refuge within the

stands. A mature stage of this forest type would favour species preferring canopy cover such

as the Bonelli’s Warbler (Phylloscopus bonelli) enhancing bird richness and abundance

(Camprodon & Brotons 2006).

The second main landscape gradient was a mosaic of habitats, where patches of non-

burnt forests, burnt areas and agriculture lands coexist. This landscape prevailed in flat areas

as the result of a less aggressive behaviour of fire in gentle slopes together with a more

effective impact of fire fighting efforts in more developed and accessible areas. In this mosaic

a combination of farmland, shrubland and some forest bird species coexist. Many of the

farmland species are species that typically feed in the fields but nest in the edge or in forest

patches as the Woodlark (Lullula arborea) and the Golden Oriole (Oriolus oriolus) (Brotons

et al. 2004). These species benefit from the coexistence of open habitats, as the originated

after fire (i.e. farmland and shrublands) (Fuller et al. 2004; Vallecillo et al. 2008). Forest bird

species were present in this mosaic landscape as a consequence of the use of unburnt patches

of pine trees left relatively untouched (Herrando & Brotons 2002). However, common forest

bird species as Short-toed Tree-creeper (Certhia brachydactyla), Firecrest (Regulus

ignicapilla) and Chiffchaff (Phylloscopus collybita) were absent or present in very low

number after the fire. These species concentrate their activity in the canopy stratum and this

habitat was only present in small patches, surrounded by unsuitable farmland or burnt habitat

and this might not be enough for supporting large numbers of these species on the area.

Although strong site fidelity of forest and shrubland birds after fire has been reported (Pons &

Bas 2005), we do not think that this process is likely to explain the presence of these species

seven years after the fire. We rather suggest that unburnt forest patches are likely to act as

remaining islands of the original pine forest habitats (Brotons et al. 2004) and thus host an

Artículo IV

98

impoverished forest bird community from which more specialist forest species such as the

Nuthatch (Sitta europaea) have been lost and others have decreased their occurrence rate

(Estrada et al. 2004).

Others common forest species such as the Wren (Troglodytes troglodytes) and the

Nightingale (Luscinia megarhynchos) were positive associated to riparian areas, where

vegetation regeneration after the fire was rapid but only locally along the stream sides. These

areas are rich on fruits and invertebrates and many bird species find them as adequate

habitats. In contrast, the presence in certain areas of P. halepensis and Q. ilex resprouters

affected negatively the presence of certain species such as the Sardinian warbler (Sylvia

melanocephala) and the Blackcap. This could be due to the relatively high density of trees

found within these areas. Finally, although the main objective of this study was not to analyze

post-fire management activities effect on bird community after a large fire, our results shown

that leaving trunks within the fire perimeter had more positive effects than negative effect on

the bird community as shown for the Green Woodpecker (Picus viridis) and the Crested Tit

(Parus cristatus). This result is in agreement with previous results than highlight the

importance of standing dead trunk for bird diversity in terms of nest site availability, perching

sites and foraging substrates (Hutto 2006).

Finally, low shrubs combined with bare ground and grasslands located in south-facing

slopes dominated the third main landscape gradient. In these areas, pre-fire abundance of

resprouters was low most likely due to historical impact of agriculture and/or grazing

(Mosandl & Kleinert 1998, Gomez 2003), leading to poor tree vegetation recovery. This

habitat provides an adequate environment for a number of shrubland bird species, such as

Ortolan Bunting or Tawny Pipit (Menz et al. 2009). This habitat, arising from the lack of tree

regeneration, is especially relevant since most of the species inhabiting such open habitats are

among the most threatened species in Europe (BirdLife International 2004). This was

corroborated by our results. The significant positive relation between the species associated to

this habitat type and their conservation value highlight the important role of open habitats for

species conservation. Several studies have shown the use of recently burnt areas by these

species (Herrando et al. 2002, Brotons et al. 2005, Pons & Bas 2005), however as vegetation

succession takes place these species tend to disappear (Prodon et al. 1987). The finding of our

study have interesting implication for bird conservation in Mediterranean areas since the

persistence of open-habitat species seven years after the fire suggests that non-direct

regeneration process might create the appropriate habitats for their conservation.

Artículo IV

99

Management implications

Perhaps not coming as a surprise, the mosaic of habitats arising after fire has

originated a rich and diverse bird community (Blondel & Aronson 1999, Brotons et al. 2004).

Although the impact of fires is thought to disappear in later succession stages as vegetation

recovers, non-direct regeneration processes as the ones described in the present study or

repeated fire impact (Díaz-Delgado et al. 2002) may lead to large temporal maintenance or

increase of habitat suitable for species linked to open vegetation (Brotons et al. 2008). Here,

management should be undertaken in order to maintain the new heterogeneous landscape and

conserve an important number of species that are considerably declining elsewhere in Europe.

In addition, maintenance of heterogeneous landscapes might avoid the violent advance of

large and catastrophic wildfires (Lloret et al. 2002).

In order to preserve bird diversity in low shrubland landscape, prescribed burning

and/or grazing by large herbivores or livestock farming should be considered (Pons et al.

2003). Controlled burning is now a widely used management tool that can help at the same

time to prevent large-scale catastrophic wildfires (Hardy & Arno 1996, Miller & Urban 2000).

Besides, bird species associated to farmland areas might be favoured by traditional farming

(Casals et al. 2007). In view of land abandonment trend, compensatory payments to farmers

should be provided to keep traditional farming.

On the other hand, thinning would be recommended in areas where forest regenerates

(Q. humilis, Q. ilex and P. halepensis) to enhance the growth of remaining trees and favour

the presence of forest bird species on these areas while reducing fire risk associated to dense

stands (Gonzalez et al. 2006). The regenerating oak stands are especially important since they

may lead to forest habitats more resilient to fires (Moreira et al. 2001, Díaz-Delgado et al.

2002).

Overall, the results of this study suggest that landscape changes induced by non- direct

tree regeneration after fire might be viewed as offering promising management opportunities

for the conservation of many species. The low rate of vegetation recovery under non-direct

tree regeneration leads for a long term availability of suitable habitats for bird species under

conservation concern associated to low shrubland and farmland habitat. This may halt the

general decreasing trend of many of these species in large areas of the Mediterranean region

associated to land abandonment processes prevailing in the last 40 years (Sirami et al. 2007).

Artículo IV

100

ACKNOWLEDGEMENTS

This work is a contribution to the European Research Group GDRE “Mediterranean

and mountain systems in a changing world” and has received financial support from the

projects Consolider-Ingenio Montes (CSD2008-00040), CGL2008-05506-CO2 ⁄BOS and

CGL2005-2000031⁄BOS granted by the Spanish Ministry of Education and Science. S.V. (FI

fellowship) received financial support from the CUR of the DIUE from the Generalitat de

Catalunya and the European Social Fund. L.B. benefited from a Ramon y Cajal contract from

the Spanish government and E.L.Z. (FPI fellowship) received financial support from the

Spanish Ministry of Education and Science.

REFERENCES

Alvarez R., Valbuena L. & Calvo L. 2007. Effect of high temperatures on seed germination

and seedling survival in three pine species (Pinus pinaster, P. sylvestris and P. nigra).

International Journal of Wildland Fire 16: 63-70.

Bibby C.J., Burgess N.D., Hill D.A. & Mustoe S.H. (eds) 2000. Bird Census Techniques.

Second Edition. Academic Press, London.

Birdlife International (eds) 2004. Birds in Europe: population estimates, trends and

conservation status. Cambridge, UK: BirdLife International.

Blondel J. & Aronson J. (eds) 1999. Biology and wildlife of the Mediterranean region. Oxford

University Press, Oxford.

Brawn J.D., Robinson S.K. & Thompson F.R. 2001. The role of disturbance in the ecology

and conservation of birds. Annu. Rev. Ecol. Syst. 32: 251-276.

Brotons L., Herrando S. & Martin J.L. 2004. Bird assemblages in forest fragments within

Mediterranean mosaics created by wild fires. Landscape Ecology 19: 663-675.

Brotons L., Pons P. & Herrando S. 2005. Colonization of dynamic Mediterranean landscapes:

where do birds come from after fire? Journal of biogeography 32: 789-798.

Brotons L., Herrando S. & Pons P. 2008. Wildfires and the expansion of threatened farmland

birds: the Ortolan Bunting, Emberiza hortulana, in Mediterranean landscapes. Journal of

Applied Ecology 45: 1059-1066.

Camprodon J. & Brotons L. 2006. Effects of undergrowth clearing on the bird communities of

the northwestern Mediterranean coppice Holm oak forests. Forest Ecology and


Casals P., Baiges T., Bota G., Chocarro C., de Bello F., Fanlo R., Sebastià M.T. & Taull M.

2007. Silvopastoral systems in NE of the Iberian Peninsula. A Multifunctional

Artículo IV

101

perspective. In: Rigueiro-Rodríguez A, McAdam J, Mosquera-Losada MR (Eds.)

Agroforestry in Europe: Current status and future prospects (in press). Springer-Berlag,

Berlin. ISBN 978-1-4020-8271-9.

Christensen N.L. 1993. Fire regimes and ecosystems dynamics. Fire in the environment: the

ecological, atmospheric, and climatic importance of vegetation fires (ed. by P.J. Crutzen

and J.G. Goldammer), pp. 233–244. Dahlem Workshop Reports. Environmental Sciences

Research Report 13, John Wiley and Sons Ltd, Chichester.

Díaz-Delgado R., Lloret F., Pons X. & Terradas J. 2002. Satellite evidence of decreasing

resilience in Mediterranean plant communities after recurrent wildfires. Ecology 83: 2293-

2303.

di Castri F. & Mooney H.A. (eds) 1973. Mediterranean type ecosystems: origin and structure.

Springer-Verlag, Berlin.

Dobson A.J. (eds) 1990. An introduction to Generalized Linear Models. Chapman & Hall,

London.

Dunning J.B., Danielson B.J. & Pulliam H.R. 1992. Ecological processes that affect

populations in complex landscapes. Oikos 65: 169-175.

Espelta J.M., Retana J. & Habrouk A. 2003. Resprouting patterns after fire and response to

stool cleaning of two coexisting Mediterranean oaks with contrasting leaf habitats on two

different sites. Forest Ecology and Management 179: 401-414.

Estrada J., Pedrocchi V., Brotons L. & Herrando S. (eds) 2004. Atles dels ocells nidificants de

Catalunya (1999-2002) (Lynx, ICO (Institut Català d'Ornitologia), Barcelona.

Foster D.R., Knight D.H. & Franklin J.F. 1998. Landscape patterns and legacies resulting

from large, infrequent forest disturbances. Ecosystems 1: 497-510.

Fuller R.J., Hinsley S.A. & Swetnam R.D. 2004. The relevance of non-farmland habitats,

uncropped areas and habitat diversity to the conservation of farmland birds. Ibis 146: 22-

31.

Gomez J.M. 2003. Spatial patterns in long-distance dispersal of Quercus ilex acorns by jays in

a heterogeneous landscape. Ecography 26: 573-584.

Gonzalez J.R., Palahí M., Trasobares A. & Pukkala T. 2006. A fire probability model for

forest stands in Catalonia (north-east Spain). Ann. For. Sci. 63: 169–176.

Gracia C., Burriel J.A., Ibáñez J.J., Mata T. & Vayreda J. 2000. Inventaria ecològic i forestal

de Catalunya. Regió forestal IV. (Centre de Recerca Ecològica I Aplicacions Forestals,

Barcelona.)

Artículo IV

102

Habrouk A., Retana J. & Espelta J.M. 1999. Role of heat tolerance and cone protection of

seeds in the response of three pine species to wildfires. Plant. Ecol. 145: 91-99.

Hanes T.L. 1971. Succession after fire in the chaparral of southern California. Ecological

monographs 41: 27-52.

Hardy C.C. & Arno S.F. (eds) 1996. The Use of Fire in Forest Restoration. General Technical

Report. Seattle, WA: Department of Agriculture, US Forest Service.

Herrando S., Brotons L., del Amo R. & Llacuna S. 2002. Bird community succession after

fire in a dry Mediterranean shrubland. Ardea 90: 303-310.

Herrando S., Brotons L. & Llacuna S. 2003. Does fire increase the spatial heterogeneity of

bird communities in Mediterranean landscapes? Ibis 145: 307-317.

Hobson K.A. & Schieck J. 1999. Changes in bird communities in boreal mixedwood forest:

Harvest and wildfire effects over 30 years. Ecological applications 9: 849-863.

Hutto R.L. 2006. Toward meaningful snag-management guidelines for postfire salvage

logging in North American conifer forests. Conservation Biology 20: 984–993.

Izhaki I. & Adar M. 1997. The effects of Post-Fire Management on Bird Community

succession. International Journal of Wildland Fire 7: 335-342.

Jacquet K. & Prodon R. 2009. Measuring the postfire resilience of a bird–vegetation system: a

28-year study in a Mediterranean oak woodland. Oecologia 161: 801–811.

Lawrence G.E. 1966. Ecology of vertebrate animals in relation to chaparral fire in the Sierra

Nevada foothills. Ecology 47: 278-291.

Lloret F. 1998. Fire, canopy cover and seedling dynamics in Mediterranean shrubland of

northeastern Spain. Journal of Vegetation Science 9: 417-430.

Lloret F. Verdu M, Noe F-H & Alfonso V-B. 1999. Fire and resprouting in Mediterranean

ecosystems : insights from an external biogeographical region, the mexical shrubland.

American Journal of Botany 86: 1655-1661.

Lloret F., Calvo E., Pons X. & Diaz-Delgado R. 2002. Wildfires and landscape patterns in the

Eastern Iberian Peninsula. Landscape Ecology 17: 745-759.

Menz M.H.M., Brotons L. & Arlettaz R. 2009. Habitat selection by Ortolan bunting Emberiza

hortulana in post-fire succession in Catalonia: implications for the conservation of

farmland populations. Ibis 151: 752-761.

Miller C. & Urban D.L. 2000. Modelling the effects of fire management alternatives on Sierra

Nevada mixed-conifer forests. Ecol. Applications 10: 85–94.

Artículo IV

103

Moreira F., Regol F.C. & Ferreira P.G. 2001. Temporal (1958–1995) pattern of change in a

cultural landscape of northwestern Portugal: implications for fire occurrence. Landscape

Ecology 16: 557-567.

Mosandl R. & Kleinert A. 1998. Development of oaks (Quercus petraea (Matt.) Liebl.)

emerged from bird-dispersed seeds under old-growth pine (Pinus silvestris L.) stands.

Forest Ecology and Management 106: 35-44.

Ordoñez J.L., Franco S. & Retana J. 2004. Limitation of the recruitment of Pinus nigra in a

gradient of post-fire environmental conditions. Ecoscience 11:296-304.

Pickett S.T.A. & White P.S. (eds) 1985. The ecology of natural disturbance and patch

dynamics. Academic Press, London.

Pons P. & Prodon R. 1996. Short term temporal patterns in a Mediterranean shrubland bird

community after fire. Acta Oecologica 17: 29-41.

Pons P., Lambert B., Rigolot E. & Prodon R. 2003. The effects of grassland management

using fire on habitat occupancy and conservation of birds in a mosaic landscape.

Biodiversity and Conservation 12: 1843-1860.

Pons P. & Bas J.M. 2005. Open-habitat birds in recently burned areas: The role of the fire

extent and species’ habitat breadth. Ardeola 52: 119-131.

Prodon R. & Lebreton J.D. 1981. Breeding avifauna of a Mediterranean succession: the Holm

oak and Cork oak series in the eastern Pyrenees. Analysis and modelling of the structure

gradient. Oikos 37: 21-38.

Prodon R., Fons R. & Peter A.M. 1984. L’impact du feu sur la végétation, les oiseaux et les

micromammifères dans diverses formations des Pyrénées-Orientales: premiers résultats.

Rev. Eco. Terre Vie 39: 129-158.

Prodon R., Fons R. & Athias-Binche F. 1987. The impact of fire on animal communities in

Mediterranean area. In: Trabaud L. (ed.) The role of fire on ecological systems: 121-157.

SPB Academic Publ., The Hague.

R Development Core Team. 2008. R: A language and environment for statistical computing.

Vienna: R Foundation for statistical computing.

Retana J., Espelta J.M., Habrouk A., Ordoñez J.L. & Solà-Morales F. 2002. Regeneration

patterns of three Mediterranean pines and forest changes after a large wildfire in

northeastern Spain. Ecoscience 9: 89–97.

Rodrigo A., Retana J., & Picó X.F. 2004. Direct regeneration is not the only response of


Artículo IV

104

Sirami C., Brotons L. & Martin J.L. 2007. Vegetation and songbird response to land

abandonment: from landscape to census plot. Diversity and Distributions 13: 42–52.

Skordilis A., & Thanos C.A. 1997. Comparative ecophysiology of seed germination strategies

in the seven pine species naturally growing in Greece. In ‘Basic and applied aspects of

seed biology’ (Eds. R.H.Ellis, A.I. Murdoch, and T.D. Hong) pp. 623-632. (Kluwer

Academica, Dordrecht, The Netherlands).

Sousa W.P. 1984. The role of disturbance in natural communities. Annual Review of Ecology

and Systematics 15: 353-391.

Ter Braak C.J.F. 1986. Canonical Correspondence Analysis: A New Eigenvector Technique

for Multivariate Direct Gradient Analysis. Ecology 67: 1167-1179.

Ter Braak C.J.F. 1988. CANOCO: an extension of DECORANA to analyze species –

environmental relationships. Vegetation 75: 159-160.

Turner M.G., Hargrove W.W., Gardner R.H. & Romme W.H., 1994. Effect of fire on

landscape heterogeneity in Yellowstone National Park, Wyoming. Journal of Vegetation

Science 5: 731-742.

Turner M.G., Dale V.H. & Everham E.H. 1997. Fires, hurricanes and volcanoes: comparing

large disturbances. BioScience 47: 758-768.

Trabaud L. & Lepart J. 1980. Diversity and stability in garrigue ecosystems after fire.


Ukmar E., Battisti C., Luiselli L. & Bologna M.A. 2007. The effects of fire on communities,

guilds and species of breeding in burnt and control pinewoods in central Italy.

Biodiversity and Conservation 16: 3287-3300.

Vallecillo S., Brotons L. & Herrando S. 2008. Assessing the response of open-habitat bird

species to landscape changes in Mediterranean mosaics Biodiversity Conservation 17:

103-119.

Vallecillo S., Brotons L. & Thuiller W. 2009. Dangers of predicting bird species distributions

in response to land-cover changes. Ecological Applications 19: 538-549.

Whelan, R. J. (eds) 1995. The ecology of fire. Cambridge University Press, New York.

Wiens J.A. (eds) 1989. The Ecology of Bird Communities. (Vols I & II. Cambridge

University Press, Cambridge, UK).

WEBsites:

StatSoft, Inc. (2001) STATISTICA (data analysis software system), version 6.

http://www.statsoft.com

Artículo IV

105

Appendix List of the bird species found in the study area showing their conservation status according to the indices derived from the EU Birds Directive (70/409/CEE); the highest abundance on both visits done in the field survey (number of bird individuals seen or heard) and percentage of transects occupied by each species (occurrence).

Species name Acronym Conserva-tion status Abundance Occurrence

Red-legged Partridge (Alectoris rufa) Ale_ruf SPEC 2 85 50%Common Quail (Coturnix coturnix) Cot_cot SPEC 3 4 4%Word Pigeon (Columba palumbus) Col_pal Non-SPEC 48 34%Turtle Dove (Streptopelia turtur) Str_tur SPEC 3 54 35%Common Cuco (Cuculus canorus) Cuc_can Non-SPEC 13 13%Nightjar (Caprimulgus europaeus) Cap_eur SPEC 2 2 2%European Soller (Coracias garrulus) Cor_gar SPEC 2 1 1%Hoopoe (Upupa epops) Upu_epo SPEC 3 13 10%Wryneck (Jynx torquilla) Jyn_tor SPEC 3 24 22%Green Woodpecker (Picus viridis) Pic_vir SPEC 2 14 12%Great Spotted Woodpecker (Dendrocopos major) Den_maj Non-SPEC 32 27%Thekla Lark (Galerida theklae) Gal_the SPEC 3 19 14%Crested Lark (Galerida cristata) Gal_cri SPEC 3 6 6%Woodlark (Lullula arborea) Lul_arb SPEC 2 246 91%Skylark (Alauda arvensis) Ala_arv SPEC 3 10 10%Tawny Pipit (Anthus campestris) Ant_cam SPEC 3 17 17%White Wagtail (Motacilla alba) Mot_alb Non-SPEC 1 1%Wren (Troglodytes troglodytes) Tro_tro Non-SPEC 30 26%European Robin (Erithacus rubecula) Eri_rub Non-SPEC 3 3%Nightingale (Luscinia megarhynchos) Lus_meg Non-SPEC 166 75%Whinchat (Saxicola rubetra) Sax_rub Non-SPEC 2 2%Stonechat (Saxicola torquatus) Sax_tor Non-SPEC 172 80%Black-eared Wheatear (Oenanthe hispanica) Oen_his SPEC 2 76 39%Northern Wheatear (Oenanthe oenanthe) Oen-oen SPEC 3 1 1%Rock Thrush (Monticola saxatilis) Mon_sax SPEC 3 43 30%Blue-Rock Trush (Monticola solitarius) Mon_sol SPEC 3 4 4%Blackbird (Turdus merula) Tur_mer Non-SPEC 56 48%Mistle Thrush (Turdus viscivorus) Tur_vis Non-SPEC 1 1%Song Thrush (Turdus philomelos) Tur_phi Non-SPEC 1 1%Cetti´s Warbler (Cettia cetti) Cet_cet Non-SPEC 1 1%Great Reed-Warbler (Acrocephalus arundinaceus) Acr_aru Non-SPEC 1 1%Melodious Warbler (Hippolais polyglotta) Hip_pol Non-SPEC 134 58%Dartford Warbler (Sylvia undata) Syl_und SPEC 2 355 80%

Artículo IV

106

Appendix continued

Species name Acronym Conserva-tion status Abundance Occurrence

Subalpine Warbler (Sylvia cantillans) Syl_can Non-SPEC 363 90%Sardinian Warbler (Sylvia melanocephala) Syl_mel Non-SPEC 44 32%Orphean Warbler (Sylvia hortensis) Syl_hor SPEC 3 18 17%Blackcap (Sylvia atricapilla) Syl_atr Non-SPEC 30 23%Garden Warbler (Sylvia borin) Syl_bor Non-SPEC 1 1%Common Whitethroat (Sylvia communis) Syl_com Non-SPEC 3 3%Bonelli’s Warbler (Phylloscopus bonelli) Phy_bon SPEC 2 53 38%Spotted Flycatcher (Muscicapa striata) Mus_str SPEC 3 2 2%Long-tailed Tit (Aegithalos caudatus) Aeg_cau Non-SPEC 18 9%Crested Tit (Parus cristatus) Par_cri SPEC 2 12 9%Blue Tit (Parus caeruleus) Par_cae Non-SPEC 13 10%Great Tit (Parus major) Par_maj Non-SPEC 89 56%Short-toed Tree-creeper (Certhia brachydactyla) Cer_bra Non-SPEC 3 3%Golden Oriole (Oriolus oriolus) Ori_ori Non-SPEC 35 25%Great Grey Shrike (Lanius meridionalis) Lan_mer SPEC 3 62 49%Woodchat Shrike (Lanius senator) Lan_sen SPEC 2 35 26%Jay (Garrulus glandarius) Gar_gla Non-SPEC 44 33%Common Raven (Corvus corax) Cor_corax Non-SPEC 4 2%Carrion Crow (Corvus corone) Cor_cor Non-SPEC 15 12%Starling (Sturnus vulgaris) Stu_vul SPEC 3 18 11%Rock Sparrow (Petronia petronia) Pet_pet Non-SPEC 27 9%Chaffinch (Fringilla coelebs) Fri_coe Non-SPEC 8 7%Serin (Serinus serinus) Ser_ser Non-SPEC 69 47%Greenfinch (Carduelis Chloris) Car_chl Non-SPEC 1 1%Goldfinch (Carduelis carduelis) Car_car Non-SPEC 23 14%Linnet (Carduelis cannabina) Car_can SPEC 2 297 83%Cirl Bunting (Emberiza cirlus) Emb_cir Non-SPEC 85 51%Rock Bunting (Emberiza cia) Emb_cia SPEC 3 227 87%Ortolan Bunting (Emberiza hortulana) Emb_hor SPEC 2 155 61%Corn Bunting (Emberiza calandra) Emb_cal SPEC 2 239 76%

107

4.5. Artículo V

V- RECENT FIRE HISTORY DETERMINES SPECIES DISTRIBUTION

DYNAMICS IN LANDSCAPES DOMINATED BY LAND

ABANDONMENT

Artículo V

108

ABSTRACT

Aim: Mediterranean landscapes are suffering two opposing forces leading to large-scale

changes in species distribution: land abandonment of less productive areas and an

increase in wildfire impact. Here, we test the hypothesis that fires occurred in recent

decades drive the pattern of expansion of early-successional bird species by aiding in the

process of colonisation of newly burnt areas.

Location: The study was carried out in Catalonia (NE Iberian Peninsula). We selected

44 burnt sites occurring between 2000 and 2005 to model colonisation patterns under

different assumption of potential colonisers’ sources and evaluate the colonisation

estimates with empirical data on 6 bird species especially collected for this purpose.

Methods: We first defined three landscape scenarios serving as surrogates of potential

colonisers’ sources: open-habitats created by fire, shrublands and farmlands. Then, we

used a parameter derived from a functional connectivity metric to predict species

colonization dynamics on the selected sites by each particular scenario. Finally, we

evaluated our colonisation estimates for each scenario with the presence/absence of each

selected species in the studied locations by using generalized lineal models.

Results: The pattern of occurrence of the focal species on the studied burnt areas was

significantly related to the amount of dispersal flux received from previous fires and to a

lesser extent to the one coming from farmland areas, while it was poorly related to the

one originated from more stable, shrubland areas.

Main conclusions: We suggest that land use changes in recent decades have produced a

shift in the ecological processes acting as reservoirs for open-habitat bird species

dynamics in Mediterranean areas. From a more permanent habitat network constituted

by relatively static open habitats (grassland and farmland) before the middle of the 20th

century to a shifting mosaic of habitat patches where fires plays a key role for population

distribution and persistence.

Keywords: Ecological networks, fire history, colonisation dynamic, open-habitat bird

species, potential dispersal flux, habitat configuration, land use changes, land

abandonment, shrubland, connectivity

Artículo V

109

INTRODUCTION

Conservation of species diversity at local, regional, and continental scales has

received increasing attention as human pressures and modification of ecosystems

increase. Much of the research has focused on the effects of habitat loss and subsequent

landscape fragmentation on species dynamics (e.g. Villard et al., 1999; Alderman et al.,

2005). Nevertheless, in some cases, species responses appear to be more complex than

initially expected because the impacts of current landscape change interact with previous

changes affecting those landscapes (de Blois et al., 2001, Kuussaari et al., 2009). This is

for instance the case of the effect of past agricultural use on present plant communities

in temperate and tropical forests (Dupouey et al., 2002; Heckenberger et al., 2003;

Dambrine et al., 2007) or on current stream biodiversity (Harding et al., 1998), where

past land uses have long lasting impacts on present biodiversity patterns.

Furthermore, Lindborg & Eriksson (2004) went a step further and found that

present plant species diversity was related not only to the past use on a local scale but

also to the historical landscape structure (related to agricultural use). One of the

mechanisms by which we may expect that historical landscape trajectories influence

species diversity is that site connectivity influences colonisation and extinction rates

(Hanski, 1999). If landscape structure is altered, but local populations at the site are

maintained, either as remnant populations (Eriksson, 1996) or as stable albeit isolated

populations, sites with high historical connectivity will maintain a high diversity in

comparison of sites with lower historical connectivity. Additionally, if landscape

changes affect connectivity then we might expect to find a strong signal of previous

changes on present species distribution dynamics, especially in colonisation patterns of

new habitats.

The Mediterranean basin provides good examples of dynamic landscapes

suffering large-scale changes in the last decades. Since the middle of the 20th century,

important and rapid changes have occurred in this area mainly affecting the availability

and the spatial arrangement of open habitats. On one side, a growing demand for higher

economical productivity and population concentration in the cities has prompted two

contrasting processes: land use intensification in areas most favourable to agriculture

and land abandonment in agricultural areas of marginal productivity (Bouma et al.,

1998; Russo, 2007). On the other side, wildfires frequency and extent have increased

due to fuel accumulation after the abandonment of rural traditional activities (e.g.

firewood extraction), higher fire risk resulting from climate change and more numerous

Artículo V

110

ignition sources (Piñol et al., 1998; Pausas et al., 2004). Whereas the expected

consequence of agricultural abandonment is the reduction of the extent of open spaces

with a concurrent regrowth and expansion of woodlands through natural succession

(Preiss et al., 1997; Moreira et al., 2001), fires are expected to have the opposite effect

by creating new habitats, maintaining the availability of open patches (Lloret et al.,

2002). These counteracting events might have important consequences in the dynamics

of the species occupying open habitats. In the past, the distribution dynamics of these

species were probably dominated by the presence of those more permanent habitats

(farmland and grazed grasslands and shrub lands). However, with the marked increase in

fire footprint in the landscape and the vegetation closure resulting from rural land

abandonment, burnt habitats may have overtaken the role in sourcing population

dynamics. If the hypothesis of historic fires acting as population sources for open habitat

species holds true, species colonization of recently created habitats, such as the ones

originated after new fires, will be more likely in sites located close to previously burnt

areas than in sites located near to other more stable open habitat types.

The aim of this study was to test the hypothesis that fires occurred in recent

decades drive new colonisations in early successional bird species in the Mediterranean.

We focused in the region of Catalonia and selected six open-habitat bird species that

make extensive use of recently burnt areas in relatively high densities (Pons & Prodon,

1996; Herrando et al., 2002; Pons & Bas, 2005). We specifically address three

questions: 1) which is the role of wildfires history on the colonization of recent burnt

areas by open-habitat bird species? 2) For how long do the consequences of fire impacts

determine current colonization patterns? 3) Which is the role of more static habitat types

such as farmlands and shrublands on the current post-fire colonization patterns?

METHODS

Study area

Catalonia (32 091 km2) is a region located in northeast of the Iberian Peninsula

comprising a range of habitat types from mountainous areas in the Pyrenees and inland

chains (with an altitude up to 3143 m) to a long coastline along the Mediterranean Sea.

The climate is mainly Mediterranean temperate, with maritime influence in the coast and

a cold influence in the Pyrenees. Landscape changes related to socio-economical

dynamics in this region may be representative of those that occurred in the whole

Mediterranean Europe during the last decennia: abandonment of traditional activities,

Artículo V

111

agricultural intensification and afforestation of formerly cultivated lands (Preiss et al.,

1997; Stoate, 2001). Between 1989 and 2008, around 230 000 ha of non-irrigated cereal

crops and grasslands were abandoned and the forest cover increased by approximately

130 000 ha (DAR). This increase has occurred even when during this period about

170000 ha of forest and/or shrub has been burnt (DMAH and CREAF).

Focal species

For this study, we selected six open-habitat bird species with preferences for

shrub-like habitats and dry, extensive farmlands in Mediterranean landscapes (Estrada et

al., 2004): Tawny Pipit (Anthus campestris), common Linnet (Carduelis cannabina),

Corn Bunting (Emberiza calandra), Ortolan Bunting (Emberiza hortulana), Thekla Lark

(Galerida theklae) and Black-eared Wheatear (Oenanthe hispanica). We selected these

species because they have been associated with colonization processes in recently burnt

areas (Pons & Prodon, 1996; Herrando et al., 2002). In addition, most of them appear to

show significant species expansion at the regional level favoured by fire (Estrada et al.,

2004; Brotons et al., 2008; Vallecillo et al., 2009).

Site selection

We selected 44 wildfires affecting an area larger than 50 ha of forest and/or

shrubland between 2000 and 2005 (Fig. 1). We used these sites in order to model the

species colonisation patterns under different assumptions of potential colonisers` sources

and to evaluate the colonisation estimates using empirical data on the focal species

collected in the field (see following sections). Fire perimeters were provided by the

DMAH. Forest and/or shrubland area within selected fires was estimated combining fire

perimeters with the land use map of Catalonia of 2002 (available at

http://mediambient.gencat.cat). Selected wildfires occurred in late winter (late January-

early March) or in summer (June-August), and ranged between 52 and 6278 ha

(Appendix 1). They were severe fires, which affected the forest canopy and undergrowth

and caused widespread tree mortality. They were located in mountain massifs, with

Mediterranean climatic conditions and at low-mid altitudes (100-1300 m above sea

level). Pine and/or oak forests or dense shrublands formerly dominated the burnt areas

(Appendix 1).

Artículo V

112

Fig. 1 Geographical location of the forty-four studied wildfires in the map of Catalonia. Numbers correspond to wildfires’ code (see Appendix 1). The three defined regions (delimited within the map of Catalonia) with the total number of fires within each one are shown.

Field methods


estimate the presence/absence of the focal species (Bibby et al., 2000). The number of

transects conducted at each burnt site increased non-linearly with burnt area, ranging

from 2 to 30 (Appendix 1). The establishment of transects observed the following

criteria: a) they were conducted entirely across burnt wildland avoiding, when possible,

unburnt patches (forest or farmland) and fire edges, b) the minimum distance between

two consecutive transects was 150m, c) the minimum distance between transects and fire

edge was 50m, d) transects were conducted preferably on existing trails in order to allow

future repetition of the transects after vegetation recovery and e) in the largest burnt

areas, transects are distributed in a number of representative locations covering habitat

heterogeneity within fire perimeter. This made a total of 453 transects distributed in the

44 studied sites (Appendix 1).

Bird surveys were conducted during the breading season (10th May – 15th June)

of 2006 and 2007, by experienced ornithologists at a speed of 2km/h approximately,

three hours after sunrise and under good weather conditions (i.e. without rainfall or

strong wind) (Bibby et al., 2000). Each survey lasted 15 min and covered about 460m in

Artículo V

113

length (SD 95m). The occurrence of the species was registered within 100m belts on

both sides of the transect.

Landscape scenarios

We identified three landscape scenarios characterized by the cover type that may

act as the source habitat (shrubland or farmland), and in the case of the shrublands by the

type of disturbance that generates the shrub cover (fire or others) (Fig. 2). We used the

land use map of Catalonia of 1997 (just before the occurrence of the selected wildfires)

to identify these scenarios. In each scenario, we identified the 4 x 4km squares in which

the abundance of the corresponding open habitat type was higher than 50ha (the

minimum amount of suitable habitat where species population was considered sufficient

to provide immigrants). The following scenarios were analysed:

- (1) Shrubland associated to fire (fire scenario). This scenario corresponded to

the prevailing shrubland within wildfires that affected Catalonia in the last

twenty years of the 20th century. Fire perimeters were provided by DMAH for

the period 1986-1999 and by CREAF for fires that occurred between 1980 and

1985. Prevailing shrublands within fires were estimated combining fire

perimeters with the land use map of Catalonia of 1997. Additionally, we

considered the shrubland associated to fires from different periods to allow for

the identification of potential long-term responses in the population sources:

from 1980 to 1999 (193 000ha of shrubland, corresponding to 431 squares), from

1980 to 1989 (96 000ha, corresponding to 263 squares) and from 1990 to 1999

(97 000ha, corresponding to 251 squares).

- (2) Non fire-related shrubland scenario. This corresponded to shrublands not

affected by fire and thus more stable from a temporal perspective. They may

have been originated from other landscape processes such as vegetation

succession after the abandonment of agricultural lands or growth limitation due

to site conditions. These shrubland areas have not suffered any fire in the last 20

years of the 20th century although they could have been affected by this

perturbation earlier on. In total 663 000ha of shrublands corresponded to this

scenario (1709 squares).

- (3) Farmland scenario corresponded to the non-irrigated cereal crops present in

the study area. We used this habitat type and did not include other farmland

habitats, such as vineyard and fruit trees, because all the studied species used

Artículo V

114

cereal crops but not other type of irrigated or intensive farmland (Estrada et al.,

2004). In total the farmland scenario accounted for 472 000 ha (corresponding to

1096 squares).

If the hypothesis of fire history overtaking and leading species distribution dynamics

holds true, we expect that the fire scenario explain current species post-fire colonization

better than the other two scenarios (Fig. 2).

Fig. 2 A schematic representation of the landscape scenarios used in the present paper in order to analyze the colonisation patterns of open-habitat bird species after fire.

Modelling patterns of species colonisation under different landscape scenarios

As colonisation is largely dependent on the degree of isolation of habitat patches

(Hanski, 1999), we considered the amount of potential dispersal flux received on the

studied burnt areas from previously existing habitat areas as the parameter that may

explain species capacity to colonise these new sites. This parameter will vary, for a

given newly burnt area, with the considered landscape scenarios and the particular

spatial arrangement of habitat patches in each of them, as it will affect the distance to

colonisers’ sources and the feasibility of open habitat birds to reach the new habitat areas

created by disturbances.

Artículo V

115

We estimated the amount of dispersal flux using an approach based on graph

theory (Appendix 2). A graph represents a landscape as a set of habitat patches (‘nodes’

in graph theory literature) and ‘links’ that represent the functional connections between

pairs of habitat patches (Urban & Keitt, 2001). In particular, we used the Conefor

Sensinode software (http://www.conefor.org) to analyze the connectivity in the

landscape graph (Saura & Torné, 2009). In this work, the set of nodes comprised the

colonisers’ sources (4 x 4 km squares, different in each landscape scenario) plus the new

suitable unoccupied areas corresponding to the studied wildfires (the latter analysed

through the “nodes to add” option in the Conefor Sensinode) (Appendix 2). Each

previously existing node was characterized through an attribute corresponding to the

percentage of suitable area (shrub or farmland, depending on the scenario). On the other

hand, the links were characterized through a probability of direct dispersal between two

squares (pij) derived from the centroid to centroid Euclidean distances between every

pair of habitat nodes (Appendix 2). The pij values were obtained through a negative

exponential function of the abovementioned distance matching to the mean dispersal

distance value. We used a mean dispersal distance of 12 km for all the species, which

may reflect constrains to effective colonisation imposed by their limited movement

abilities. Although dispersal is a key factor in order to understand metapopulation

dynamics, this parameter is frequently unknown with acceptable accuracy due to

practical difficulties involved in tracking movement of individuals over large distances

(Wiens, 2001). For all the species studied in this work, accurate information on dispersal

was only available for the Ortolan Bunting in the study by Dale et al., (2006), where that

mean distance value of 12 km was found. Assuming that the dispersal distance of early-

successional species might be relatively similar (considering the similar functional role,

behaviour and body mass they present), we used this distance as a reference of the

dispersal abilities of all this group of species.

The dispersal flux was estimated from the probability of connectivity (PC) metric

(Saura & Pascual-Hortal, 2007) implemented in Conefor Sensinode. This metric is

suited to analyse the impact in connectivity of a variation in the habitat pattern (Pascual-

Hortal & Saura, 2006; Saura & Pascual-Hortal, 2007; Saura et al., in press), such as

those created by wildfires at different spatial scales (Zozaya et al., submitted). In

particular, we considered one of the three fractions quantified by the PC index that

specifically evaluates the amount of potential dispersal flux received by a certain patch

within the landscape (Saura & Rubio, in press) (Appendix 2). This fraction was

Artículo V

116

simplified in order to make the estimated flux depend exclusively on the sources of

colonisers (characteristics and distance to the source habitat) and the dispersal constrains

of the species, but not on a distinctive attribute of the fires themselves (burnt area),

resulting in the following formula:

*ij

n

jiji paPCflux ⋅= ∑

≠

(1)

Where PCfluxi is the estimated dispersal flux received in the burnt area i, aj is the

attribute of each of the nodes (already existing before the wildfires) in each of the

landscape scenarios described above and n is the total number of nodes existing in the

landscape before the new wildfires (different in each landscape scenario). *ijp is defined

as the maximum product probability of all possible paths between nodes i and j (product

of the pij’s in each of the links in the best path (in terms of probability) between i and j). *ijp considers the contribution of nodes that function as stepping stones and facilitate the

dispersal between other habitat areas (Saura & Pascual-Hortal, 2007; Saura & Rubio, in

press).

Data analysis

We used generalized linear modelling (GLM) to examine the potential role of the

different landscape scenarios explaining the colonisation dynamics of the focal species

in the selected burnt sites. We used the presence/absence of the focal species in the

selected sites as the response variable in a model with a binomial error and a logit link

function. We considered presence when a species was recorded in any of the two visits

and absence otherwise. The amount of potential dispersal flux was used as a fixed factor.

Additionally, we explicitly incorporated the spatial structure of colonisation patterns and

analyzed the interaction of the amount of dispersal flux and three bioclimatic regions

within the study area in order to analyze whether significant response on species

dynamic might be hidden by an effect of regional variability. In this sense, we divided

Catalonia in three different regions (South, Northwest, Northeast) based on differences

in climate and dominant forest species (derived from DGMN 1994 and Burriel et al.,

2000-2004) (see Fig. 1, Table 1). For each GLM model fitted, we estimated the adjusted

generalized coefficient of determination (R2) in order to allow comparability between

landscapes scenarios. All statistical analyses were run within R software.

Artículo V

117

Table 1 Characterization of the three regions considered in the study in Catalonia, NE Iberian Peninsula, according to mean annual temperature (T), mean annual precipitation (P), and dominant forest species.

Note: Data are from DGMN (1994) and Burriel et al. (2000-2004)

RESULTS

Species assemblage after fire

We found differences between species in the number of colonised burnt areas,

suggesting a large heterogeneity in the species capacity to colonise recently burnt areas.

The most common species (occurring in more than 25 locations) were the common

Linnet and the Black-eared Wheatear, followed by the Thekla Lark, the Corn Bunting

and the Tawny Pipit. In contrast, the Ortolan Bunting was the rarest species with less

than 10 sites accounting with the species (Fig. 3). Nevertheless, all the focal species

showed similar colonisation patterns between regions. Most of the colonised sites were

distributed between the south and northwest regions, whereas few sites were colonised

in the northeast region, indicating an effect of factors acting at regional scale that

constrain current colonisation (Fig. 3).

Region

Mean

Annual T

(ºC)

Mean

annual P

(mm)

Dominant forest species

South 14.41 533.91 Pinus halepensis, Quercus ilex, Pinus nigra

Northwest 10.00 810.42 Pinus sylvestris, Pinus nigra, Quercus ilex,

Quercus humilis

Northeast 13.02 841.99 Quercus ilex, Quercus humilis, Pinus sylvestris,

Quercus suber

Artículo V

118

Fig. 3 Frequency distribution for the six open-habitats bird species in the 44 burnt areas and pie chart of the percentage of species presence within the south region (in black), the northwest region (in white) and the northeast region (black dotted).

Landscape scenarios explaining species post-fire colonisation

The percentage of explained deviance (R2) by the landscape scenarios was

relatively low (Table 2). Nevertheless, we still believe it is significant given the

complexity of the process here analyzed and the simplicity of the scenarios described.

The occurrence of all the focal species except the common Linnet was

significantly related to the amount of dispersal flux received from any of the scenarios

described by old fires (Table 2), indicating an important role of fire history in current

colonisation dynamics. Species post-fire colonisation was more likely in sites located

nearby previously burnt areas. The colonisation patterns of the Tawny Pipit and the

Ortolan Bunting were significantly related to the dispersal flux received from fires

occurring in the 80s and 90s (Table 2). Besides, the colonisation of the Thekla Lark was

marginally related (P=0.05) to fires occurring in this period, but this relation depended

on the region where fire occurred (Table 2). The interaction between colonisers’ sources

and regions was also found to be significant in the new post-fire colonisation of the

Woodlark and the Black-eared Wheatear for the period from 1990-1999 and 1980-1989

respectively. In the case of the Tawny Pipit and the Thekla Lark, the colonisation pattern

was significantly related to different fire periods. In the case of the Tawny Pipit, the fires

occurring in the whole analyzed period (from 1980 to 1999) were the best predictor of

0

5

10

15

20

25

30

35

40

Common linnet Black-earedwheatear

Thekla lark Corn bunting Tawny pipit Ortolanbunting

Artículo V

119

colonisations, whereas in the Thekla Lark this corresponded the fires in the 90s (Table

2).

Moreover, the colonisation dynamics of three studied species (Corn Bunting,

Ortolan Bunting and Thekla Lark) were significantly related to the amount of dispersal

flux received from nearby farmlands, suggesting the importance of strong connections

with farmland for the persistence of open-habitat birds (Table 2). Nevertheless, the

colonisation patterns of the studied birds were poorly related to the landscape scenario

corresponding to shrubland areas not affected by fire. Only the occurrence pattern of the

common Linnet was found to be significantly related to this scenario (Table 2).

Table 2 Summary of the GLM results analyzing the role of three landscape scenarios (fire, shrubland and farmland) on the colonisation patterns of six early-successional species after fire.

Dispersal flux (D) Region x D

Species Landscape scenarios Coef. P R2 P R2

Tawny pipit Fire (80-99) 0.10 * 9.07 ns Fire (90-99) 0.15 0.05 7.91 ns

Common linnet Shrubland 0.07 * 9.61 ns

Corn bunting Fire (90-99) ns * 27.90 Farmland 0.05 * 10.47 ns

Ortolan bunting Fire (80-99) 0.10 * 9.45 ns Farmland 0.09 ** 28.08 ns

Thekla lark Fire (80-99) ns 0.05 20.72 Fire (90-99) ns * 27.18 Fire (80-89) ns * 23.80 Farmland 0.04 * 8.09

Black-eared wheatear Fire (80-89) ns * 32.53

The amount of dispersal flux received in each studied burnt area by each particular landscape scenario (D) was used as a fixed factor. Additionally, its interaction with three bioclimatic regions within the study region was also analyzed.

Notes: Only significant terms of any of the two models were included. Significance (P), coefficients (Coef.) and percentage of explained deviance (R2) from each model in the analyses are shown. (n.s.= p-value > 0.05; * = p-value < 0.05; ** = p-value < 0.01).

Artículo V

120

DISCUSSION

Our results support the hypothesis that current fire regime plays a critical role in

the colonisation of new suitable habitats by open habitat birds in Mediterranean areas.

Wildfires occurring in the last decades are acting as sources of immigrants for current

species post-fire colonisation. According to our results, which indicate a positive effect

of the oldest fires considered in this study (10-20 years old) on current colonisation

dynamics for some of the studied species, fire impact on species dynamics may be

extended over relatively long periods of time (Vallecillo et al., 2009). It is interesting to

note that, larger temporal lags may not extend over longer periods since Mediterranean

vegetation is highly resilient to fire effects (Hanes, 1971; Lloret, 1998) and open-habitat

birds tend to disappear from disturbance sites some years after the fire event (Prodon &

Lebreton, 1981; Jacquet & Prodon, 2009). In this case, individuals might be forced to

leave the areas and colonise new high quality ones, suggesting that wildfires might

function as a non-permanent colonisers’ sources.

Furthermore, the described farmland network contributed to explain the

colonisation patterns of three out of the six focal species. The variability in the

importance of the farmland scenario between species might be related to differences in

the use of the cropped areas. For instance, in Catalonia the common Linnet occurs in a

wide variety of habitats (Borràs et al., 2004) and therefore colonisers may not

necessarily come only from the cropped areas, whereas the corn bunting is strongly

dependent on this habitat type (Estrada & Orta, 2004). Although the importance of

cropped areas for feeding, foraging and breeding for the focal species is well known

(e.g. Sanderson et al., 2005), our study highlights the importance of maintaining a highly

connected farmland network for supporting the species spatial-temporal dynamics and

persistence. This result might suggest that in spite of the global decrease of cereal cover

in Catalonia during the last decades and the detrimental effect of agricultural

intensification on farmland species (Mañosa et al., 1996; Brotons et al., 2004), the

current network of cropped areas still plays an important role as a permanent source of

open-habitat birds’ immigrants. Nevertheless, if the extensive farmland cover continuous

to decrease, as it seems to be the current trend, processes such as habitat degradation or

fragmentation, already occurring in other European countries, might have a great

negative impact on these species (Donald et al., 2001). This is of utmost importance

since all the studied species have an unfavourable conservation status in Europe (SPEC

1 or 2) (BirdLife International, 2004).

Artículo V

121

On the other hand, only one out of the six studied species was significantly

related to the amount of potential flux received from more stable, non-burnt shrublands,

indicating that this scenario may be less suited for early successional species (Sirami et

al., 2007). Land abandonment favours the recovery of natural vegetation dynamics

(Debussche et al., 1987) leading to the encroachment of open habitats that were

maintained open by grazing, agriculture or by traditional harvesting for firewood

(Debussche et al., 1999; Romero-Calcerrada & Perry, 2004). Although Mediterranean

shrubland birds might first benefit from the initial shift from grassland and old fields to

open shrublands, their populations may ultimately decline as succession proceeds into

closer woodlands (Preiss et al., 1997). Our results suggest that habitat patches composed

by shrublands originated from former farmlands or grasslands might not be effective as

sources of long-term sources of colonizers that promote open-habitat species persistence.

In addition to the need for considering large-scale processes in order to

understand colonisation dynamics, regional variation seemed to increase the explanatory

power of current species post-fire colonisation. The large contributions of regional

effects explaining variability in colonisation patterns for some species (i.e. Thekla Lark,

the Corn Bunting and Black-eared Wheatear) suggest that important spatially dependent

factors were not explicitly included in this study. Eventually, older fires can only

considered as population’ sources if they were initially colonised, i.e. whether they were

close to other population sources (Brotons et al., 2005). Other factors favouring a

possible mismatch between potential and realized colonisation dynamics are the

potential time lags of the species response to environmental changes originated after fire

(Austin, 2002). Species colonisation increase with time since fire due to stochastic

processes (Zozaya et al., submitted). Finally, the spatial resolution used in the present

paper (4x4km2) and the minimum amount of suitable habitat in order to provide

immigrants (50ha) might have sub-estimated the importance of certain landscape

scenarios. Nevertheless, we still think that the results shown in the present paper are

valuable to highlight the main processes constraining colonisation process at large scales

and to provide novel insights into the current interactions between landscape and bird

species dynamics in Mediterranean areas.

In general, our study might suggest that large-scale land changes in recent

decades have produced a shift in the ecological processes acting as reservoirs for open-

habitat bird species dynamic in Mediterranean areas. From a more permanent habitat

network constituted by relatively static open-habitats (grassland and farmland) before

Artículo V

122

the middle of the 20th century to a shifting mosaic of habitat patches where fire

disturbance plays a double role. On one side, it creates suitable habitat for open-habitat

bird species colonisation and on the other side, it may act as a source of immigrants in

the colonisation of new appearing suitable habitats.

ACKNOWLEDGEMENTS

We thank the ornithologists who participated in the bird surveys. We also thank

Magda Pla for the laborious GIS work and Miquel de Cáceres for his support in statistics

and R programme software. This work has received financial support from the projects

CGL2005-00031/BOS, AGL2009-07140/FOR and CONSOLIDER-MONTES granted

by the Spanish Ministry of Education and Science and has been supported by a Marie

Curie International Reintegration grant within the 6th European Community Framework

programme. E.L.Z. (FPI fellowship) received financial support from the Spanish


REFERENCES

Alderman, J., McCollin, D., Hinsley, S.A., Bellamy, P.E., Picton, P. & Crockett, R.

(2005) Modelling the effects of dispersal and landscape configuration on population

distribution and viability in fragmented habitat. Landscape Ecology, 20, 857-870.

Austin, M.P. (2002) Spatial prediction of species distribution: an interface between

ecological theory and statistical modelling. Ecological Modelling, 157, 101-118.

Bibby, C.J., Burgess, N.D. & Hill, D.A. (2000) Bird Census Techniques. Second Edition

Academic Press, London.



Series No. 12).

Borràs, A., Senar, J.C., Cabrera, J. & Cabrera, A. (2004) Passerell comú Carduelis

cannabina. In Catalan Breeding Bird Atlas (1999–2002) (eds J. Estrada, V.

Pedrocchi, L. Brotons & S. Herrando). Lynx, Barcelona.

Bouma, J., Varallyay, G. & Batjes, N.H. (1998) Principal land use changes anticipated in

Europe. Agriculture Ecosystems and Environment, 67, 103-119.

Brotons, L., Mañosa, S. & Estrada, J. (2004) Modelling the effects of irrigation schemes

on the distribution of steppe birds in Mediterranean farmland. Biodiversity and

Conservation, 13, 1039–1058.

Artículo V

123

Brotons, L., Pons, P. & Herrando, S. (2005) Colonization of dynamic Mediterranean

landscapes: where do birds come from after fire? Journal of Biogeography, 32, 789-

798.

Brotons, L., Herrando, S. & Pons, P. (2008) Wildfires and the expansion of threatened

farmland birds: the ortolan bunting, Emberiza hortulana, in Mediterranean

landscapes. Journal of Applied Ecology, 45, 1059-1066.

Burriel, J.A., Gracia, C., Ibáñez, J.J., Mata, T. & Vayreda, J. (2000-2004) Inventari

Ecològic i Forestal de Catalunya (IEFC). CREAF, Bellaterra.

Dale, S., Steifetten, Ø., Osiejuk, T.O., Losak, K. & Cygan, J.P. (2006) How do birds

search for breeding areas at the landscape level? Interpatch movements of male

Ortolan bunting. Ecography, 29, 886-898.

Dambrine, E., Dupouey, J.-L., Laüt, L., Humbert, L., Thinon, M., Beaufils, T. &

Richard, H. (2007) Present forest biodiversity patterns in France related to former

roman agriculture. Ecology, 88, 1430-1439.

de Blois, S., Domon, G. & Bouchard, A. (2001) Environmental, historical, and

contextual determinants of vegetation cover: a landscape perspective. Landscape

Ecology, 16, 421-436.

Debussche, M., Rambal, S. & Lepart, J. (1987) Land-use changes in the humid

Mediterranean region: estimate of the hydrological impact. Acta Oecologica, 8, 317-

332.

Debussche, M., Lepart, J. & Dervieux, A. (1999) Mediterranean landscape changes:

evidence from old postcards. Global Ecology and Biogeography, 8, 3-15.

DGMN (Direcció General del Medi Natural). (1994). Plà General de Política Forestal.

Donald, P.F., Green, R.E. & Heath, M.F. (2001) Agricultural intensification and the

collapse of Europe’s farmland bird populations. Proceedings of the Royal Society of

London, Series B: Biological Sciences, 268, 25–29.

Dupouey, J.L., Dambrine, E., Laffite, J.D. & Moares, C. (2002) Irreversible impact of

past land use on forest soils and biodiversity. Ecology, 83, 2978-2984.

Eriksson, O. (1996) Remnant dynamics of plants: a review of evidence for remnant,

source-sink and metapopulations. Oikos, 77, 248-258.

Estrada, J. & Orta, J. (2004) Cruixidell Emberiza calandra. In Catalan Breeding Bird

Atlas (1999–2002) (eds J. Estrada, V. Pedrocchi, L. Brotons & S. Herrando). Lynx,

Barcelona.

Artículo V

124

Estrada, J., Pedrocchi, V., Brotons, L. & Herrando, S. (2004) Atles dels ocells nidificants

de Catalunya (1999-2002). Lynx, ICO (Institut Català d'Ornitologia), Barcelona.

Hanes, T.L. (1971) Succession after fire in the chaparral of southern California.

Ecological monographs, 41, 27-52.

Hanski, I. (1999) Metapopulation ecology. Oxford University Press, Oxford.

Harding, J.S., Benfield, E.F., Bolstad, P.V., Helfman, G.S., Jones, E.B.D. (1998) Stream

biodiversity: The ghost of land use past. Proceedings of Natural Academy of

Sciences, 95, 14843-14847.

Heckenberger, M.J., Kuikuro, A., Kuikuro, U.T., Russell, J.C., Schmidt, M., Fausto, C.

& Franchetto, B. (2003) Amazonia 1492: pristine forest or cultural parkland?

Science, 301, 1710-1714.

Herrando, S., Brotons, L., del Amo, R. & Llacuna, S. (2002) Bird community succession

after fire in a dry mediterranean shrubland. Ardea, 90, 303-310.

Jacquet, K. & Prodon, R. (2009) Measuring the postfire resilience of a bird–vegetation

system: a 28-year study in a Mediterranean oak woodland. Oecologia, 161, 801-811.

Lindborg, R., Eriksson, O. (2004) Historical landscape connectivity affects present plant

species diversity. Ecology, 85, 1840-1845.

Lloret, F. (1998) Fire, canopy cover and seedling dynamics in Mediterranean shrubland

of northeastern Spain. Journal of Vegetation Science, 9, 417-430.

Lloret, F., Calvo, E., Pons, X. & Diaz-Delgado, R. (2002) Wildfires and landscape

patterns in the Eastern Iberian Peninsula. Landscape Ecology, 17, 745-759.

Kuussaari, M., Bommarco, R., Heikkinen, R.K., et al. (2009) Extinction debt: a

challenge for biodiversity conservation. Trends in Ecology and Evolution, 24, 564-

571.

Mañosa, S., Estrada, J., Folch, A., Bonfil, J., Gonzàlez-Prat, F. & Orta J. (1996) Bird-

habitat relationships in the Catalan steppes. Conservación de las Aves Esteparias y

sus Hábitats (ed. by J. Fernández Gutiérrez J. & J. Sanz-Zuasti), pp. 153–161. Junta

de Castilla y León, Valladolid, Spain.

Moreira, F., Rego, F.C. & Ferreira, P.G. (2001) Temporal (1958–1995) pattern of

change in a cultural landscape of northwestern Portugal: implications for fire

occurrence. Landscape Ecology, 16, 557-567.

Pascual-Hortal, L. & Saura, S. (2006) Comparison and development of new graph-based

landscape connectivity indices: towards the priorization of habitat patches and

corridors. Landscape Ecology, 21, 956-967.

Artículo V

125

Pausas, J.G. (2004) Changes in fire and climate in the eastern Iberian Peninsula

(Mediterranean basin). Climatic Change, 63, 337-350.

Piñol, J., Terradas, J. & Lloret, F. (1998) Climate warming, wildfire hazard and wildfire

occurrence in coastal Eastern Spain. Climatic Change, 38, 345-357.

Pons, P. & Prodon, R. (1996) Short term temporal patterns in a Mediterranean shrubland

bird community after wildfire. Acta Oecologica, 17, 29-41.

Pons, P. & Bas, J.M. (2005) Open-habitat birds in recently burned areas: The role of the

fire extent and species’ habitat breadth. Ardeola, 52, 119-131.

Preiss, E., Martin, J.L. & Debussche, M. (1997) Rural depopulation and recent landscape

changes in a Mediterranean region: consequences to the breeding avifauna.

Landscape Ecology, 12, 51-61.

Prodon, R. & Lebreton, J.D. (1981) Breeding avifauna of a Mediterranean succession:

the holm oak and cork oak series in the eastern Pyrenees. Analysis and modelling of

the structure gradient. Oikos, 37, 21-38.

Prodon, R., Fons, R. & Athias-Binche, F. (1987) The impact of fire on animal

communities in Mediterranean area. The role of fire on ecological systems (ed. by L.

Trabaud), pp. 121-157. SPB Academic Publishing, The Hague.

Romero-Calcerrada, R. & Perry, G.L.W. (2004) The role of land abandonment in

landscape dynamics in the SPA “Encinares del rio Alberche y Cofio”, Central Spain,

1984-1999. Landscape and Urban Planning, 66, 217-232.

Russo, D. (2007) The effects of land abandonment on animal species in Europe:

conservation and management implications. Integrated assessment of vulnerable

ecosystems under global change in European Union. Project report. European

Commission, Community Research, Sustainable development, global change and

ecosystems.

Sanderson, F.J., Donald, P.F. & Burfield, I.J. (2005) Farmland birds in Europe: from

policy change to population decline and back again. Ecology and conservation of

steppe-land birds (ed. by G. Bota, M.B. Morales, S. Mañosa & J. Camprodon), pp.

211-236. Lynx Edicions & Centre Tecnològic Forestal de Catalunya, Barcelona.

Saura, S. & Pascual-Hortal, L. (2007) A new habitat availability index to integrate

connectivity in landscape conservation planning: comparison with existing indices

and application to a case study. Landscape and Urban Planning, 83, 91-103.

Artículo V

126

Saura, S. & Torné, J. (2009) Conefor Sensinode 2.2: a software package for quantifying

the importance of habitat patches for landscape connectivity. Environmental

Modelling and Software, 24, 135-139.

Saura, S. & Rubio, L. (in press) A common currency for the different ways in which

patches and links can contribute to habitat availability and connectivity in the

landscape. Ecography. DOI 10.1111/j.1600-0587.2009.05760.x.

Saura, S., Estreguil, C., Mouton C. & Rodríguez-Freire, M. (in press) Network analysis

to assess landscape connectivity trends: application to European forests (1990-2000).

Ecological Indicators.

Sirami, C., Brotons, L. & Martin, J.L. (2007) Vegetation and songbird response to land

abandonment: from landscape to census plot. Diversity and Distributions, 13, 42-52.

Stoate, C., Boatman, N.D., Borralho, R.J., Rio Carvalho, C., de Snoo, G.R. & Eden, P.

(2001) Ecological impacts of arable intensification in Europe. Journal of

Environmental Management, 63, 337-365.

Urban, D. & Keitt, T. (2001) Landscape connectivity: a graph theoretic perspective.

Ecology, 82, 1205-1218.

Vallecillo, S., Brotons, L. & Thuiller, W. (2009) Dangers of predicting bird species

distributions in response to land-cover changes. Ecological Applications, 19, 538-

549.

Villard, M-A., Trzcinski, M.K. & Merriam, G. (1999) Fragmentation effects on forest

birds: relative influence of woodland cover and configuration on landscape

occupancy. Conservation Biology, 13, 774-783.

Wiens, J.A. (2001) The landscape context of dispersal. Dispersal (ed. by J. Clobert, E.


WEBsites:

CREAF. Centre de Recerca Ecológica i Aplicacions Forestals

(http://www.creaf.uab.es/cat/index.htm)

DAR Departament d’Agricultura, Alimentació i Acció Rural

(http://www20.gencat.cat/portal/site/DAR)

DARP: Departament d’Agricultura, Ramaderia i Pesca, Barcelona, Spain.

(http://www.gencat.cat/)

DMAH Departament de Medi Ambient i Habitatge

(http://www20.gencat.cat/portal/site/dmah)

Artículo V

127

Appendix 1 Description of the areas affected by fire that were analysed in this study, in chronological order. The description includes the name, code (see Figure 1), year when the fire occurred, main habitats before the fire, fire area and number of line transects conducted within the fire perimeter in 2006 and 2007 to estimate the presence of the open-habitat bird species.

Fire Name

Fire code Year Main habitat Fire area

(ha) No. Transects

Albiol 29 2000 Shrubland 645.48 17 Camposines 34 2000 Shrubland 52.11 2 Fontrubi 23 2000 Forest (P. halepensis) 87.57 3 Garriguella 2 2000 Shrubland 6389.19 31 Olivella 25 2000 Shrubland and P. halepensis 382.77 12 PratdeCompte 33 2000 Forest (P. halepensis) 275.85 13 Badalona 17 2001 Shrubland 66.51 3 Cadaques 1 2001 Shrubland 1744.11 25 Castelldefels 26 2001 Shrubland 53.73 2

Cubells 42 2001 Shurbland and Quercus coccifera 207.18 11

Escala 3 2001 P. halepensis and shrubland 371.88 11 LaFloresta 40 2001 Shrubland 109.89 5

VallbonaAnoia 21 2001 P. halepensis, P. pinea, shrubland and Q. ilex 51.75 2

Castellbisbal02 19 2002 Shrubland 126.72 5 Tivissa 31 2002 Shrubland and P. halepensis 53.91 2 VilaverdLilla 27 2002 Shrubland 495.27 15 Alcover 28 2003 Shrubland 215.37 10 CaldesMalavella03 8 2003 P. pinea and shrubland 60.21 2 CastellbelliVilar 16 2003 Shrubland and P. halepensis 296.91 14 DaltmarOlerdola 24 2003 Shrubland 129.24 6 GranjaEscarp 39 2003 Shrubland 1794.33 22 Jorba 22 2003 Shrubland 64.35 3

MasanetGran 9 2003 Q. suber, P. pinea, P. pinaster, Q. ilex and shrubland 920.52 21

MasanetPetit 10 2003 Q. ilex, Q. suber and shrubland 62.64 2 Masquefa 20 2003 Forest (P. halepensis) 58.77 3

PlatjaAro 5 2003 Shrubland, Q. suber, P. halepensis and P. pinea 318.51 12

SantFeliuGuixol 6 2003 Shrubland, Q. suber, P. halepensis and P. pinea 514.44 18

SantLlorensSavall 15 2003 Forest (P. halepensis) 4483.26 30 Selvanera 41 2003 Shrubland 121.86 5 Talamanca 14 2003 Forest (P. halepensis) 191.07 7

Artículo V

128

Appendix 1 Continued

Fire Name

Fire code Year Main habitat Fire area

(ha) No. Transects

Montanissell 43 2004 Shrubland and Pinus nigra subsp. salzmanni 80.82 4

Montgri 4 2004 Shrubland 500.85 16 Balsareny 12 2005 Forest (P. halepensis) 858.87 19 BorgesdelCamp 30 2005 Shrubland and P. halepensis 105.75 3 CaldesMalavella05 7 2005 P. pinea and shrubland 78.57 3 Cardona 11 2005 Forest (P. halepensis) 1216.17 21 Castellbisbal05 18 2005 Shrubland and P. halepensis 206.73 6 Margalef 38 2005 P. halepensis and shrubland 384.21 13 PalmadeEbre 37 2005 Shrubland 90.18 4 Perello 32 2005 Shrubland 96.57 5 PobladeMasaluca 35 2005 Shrubland and P. halepensis 104.04 4 RibaRoja05 36 2005 P. halepensis and shrubland 605.25 18 Rocafort 13 2005 P. halepensis and shrubland 783.72 20 Viladecans 44 2005 Shrubland 57.78 3

129

Artí

culo

V

App

endi

x 2

A s

chem

atic

rep

rese

ntat

ion

of th

e st

eps

follo

wed

to a

sses

s th

e am

ount

of p

oten

tial d

ispe

rsal

flu

x re

ceiv

ed o

n th

e ne

wly

bur

nt a

reas

. In

step

1, s

tudi

ed b

urnt

si

tes

(in b

lack

) w

ithin

the

north

east

reg

ion

and

an e

xam

ple

of o

ne o

f th

e la

ndsc

ape

scen

ario

s (s

hrub

land

with

in f

ires

from

199

0 to

199

9) c

onsi

dere

d in

this

stud

y (s

quar

es s

how

n in

gre

y). I

n st

ep 2

, nod

es a

nd d

ista

nces

file

s ob

tain

ed to

repr

esen

t the

land

scap

e as

a g

raph

. In

step

3, e

stim

atio

n of

the

disp

ersa

l flu

x on

th

e se

lect

ed b

urnt

site

s by

usi

ng C

onef

or S

ensi

node

sof

twar

e. S

ee M

etho

ds:

Mod

ellin

g pa

ttern

s of

spe

cies

col

onis

atio

n af

ter

fire

unde

r di

ffere

nt la

ndsc

ape

scen

ario

s St

ep 1

) Lan

dsca

pe d

escr

iptio

n

Step

2) N

odes

and

Dis

tanc

es m

atric

es

Step

3) E

stim

atio

n of

the

disp

ersa

l flu

x

NO

DE

FIL

E

Nod

e ID

Attr

ib.

Nod

e Ty

pe1

6.87

0

2 5.

26

0 3

15.8

0

. .

. .

. .

215

1 1

216

1 1

D

IST

AN

CE

FIL

E

Nod

e ID

iN

ode

IDj

Dis

tanc

e 1

2 20

.80

1 3

800.

20

. .

. .

. .

215

216

100

Discusión general

131

5. DISCUSIÓN GENERAL

La mayoría de trabajos existentes sobre el impacto del fuego en las aves son

estudios locales donde se analiza las diferencias entre la comunidad antes y después de

la perturbación (Herrando et al. 2002, Kotliar et al. 2007, Ukmar et al. 2007) o cómo

cambia la comunidad a lo largo del tiempo transcurrido después del incendio (Lawrence

1966, Prodon & Lebreton 1981, Pons & Prodon 1996, Jacquet & Prodon 2009). Sin

embargo, a menudo nos encontramos con resultados contradictorios. Por ejemplo,

después del descenso inicial en la riqueza de especies justo pasado el incendio, algunos

estudios han mostrado que la riqueza de especies aumenta (Pons & Prodon 1996), otros

que no cambia (Wilson et al. 1995, Artman et al. 2001) y otros que disminuye (Pons

1999). Estas diferencias indican que la respuesta de las aves a los incendios es compleja

y heterogénea. Por ello, la creación de una base de datos que contenga información

avifaunística en una gran variabilidad de incendios parece una pieza clave para entender

los patrones en la respuesta de las aves a los incendios.

En el artículo 1 hemos presentado una base de datos (DINDIS, distribución de la

dinámica de aves en paisajes mediterráneos afectados por incendios) con información

de la comunidad de aves, durante el periodo reproductor, de todas las zonas quemadas

en Cataluña desde el año 2000 (Fig. 4). Los resultados muestran una fuerte variabilidad

temporal y espacial en la respuesta de las aves. La comunidad de aves varió en el

tiempo y en las diferentes regiones bioclimáticas, indicando que múltiples factores

actúan a diferentes escalas espaciales y temporales, determinando la composición de las

especies después del fuego.

Así, los incendios que ocurren dentro de la región noroeste de Cataluña

presentan más especies de aves de tierra adentro (inland species, en inglés) y los

incendios que ocurren dentro de la región noreste tienen más especies mediterráneas.

Este dato sugiere que otros factores que actúan a escala regional limitan la respuesta de

la comunidad de aves al fuego. El contexto paisajístico y la capacidad de dispersión de

las especies podrían explicar estas diferencias (Moreira et al. 2003, Pons & Bas 2005).

Discusión general

132

Fig. 4 Localización geográfica de los incendios ocurridos en Cataluña desde el año 2000 dentro de las tres regiones bioclimáticas.

A escala temporal y en el periodo para el cual hemos recopilado información (de

1 a 10 años después del incendio), hemos observado un cambio de la comunidad

avifaunística de las zonas analizadas: de una comunidad característica de especies

forestales a una comunidad con especies de hábitats abiertos. La vegetación antes del

incendio parece explicar en parte este resultado. Una zona donde el hábitat dominante

antes del incendio era el matorral, presenta después del incendio una comunidad de aves

constituida mayoritariamente por especies de hábitats abiertos. Sin embargo, la

comunidad de aves después del incendio en una zona forestal está caracterizada por

especies forestales, principalmente debido a la fuerte fidelidad al lugar y la presencia de

árboles muertos en pie (Prodon et al. 1987, Pons & Prodon 1996, Hutto 2006). Estas

especies suelen desaparecer en los primeros años y no reaparecen hasta que la estructura

del hábitat no tiene un porte arbustivo (Jacquet & Prodon 2009) y en su lugar se

observan especies de aves de hábitats abiertos (Pons & Bas 2005). Es interesante

destacar que a corto plazo, este cambio en la composición de especies es precisamente

el patrón opuesto que se espera de la respuesta de las aves con la sucesión vegetal

(Prodon & Lebreton 1981).

A pesar de la presencia en algunos de los incendios de especies forestales

durante los primeros años, las aves más abundantes durante los primeros estadíos de la

sucesión vegetal estuvieron asociadas a hábitats abiertos, contribuyendo este resultado a

recalcar la importancia de esta perturbación en la dinámica de estas especies (Moreira &

Russo 2007, Brotons et al. 2008, Vallecillo et al. 2009)

Discusión general

133

5.1. Factores que determinan la colonización postincendio

En general, observamos una gran heterogeneidad espacial y temporal en los

patrones de colonización postincendio de las aves de hábitats abiertos (Artículos 2, 3 y

4). La probabilidad de colonización postincendio varió con el tiempo transcurrido

después del fuego y dependió de factores que actúan a diferentes escalas espaciales,

como es la heterogeneidad a escala local producida por la pendiente, la orientación, la

vegetación antes del incendio y la severidad del incendio, así como factores que actúan

a escala de paisaje como la conectividad del paisaje y en menor medida del tamaño del

incendio (Fig. 5).

COLONIZACIÓNPOSTINCENDIO

Escala Local Escala regional

Pendiente

Orientación

Vegetaciónantes del incendio

Severidad del incendio

Conectividad

Tamaño del incendio

COLONIZACIÓNPOSTINCENDIO

Escala Local Escala regional

Pendiente

Orientación

Vegetaciónantes del incendio

Severidad del incendio

Conectividad

Tamaño del incendio

Fig. 5 Factores determinantes del patrón espacio-temporal de la colonización postincendio de las aves de hábitats abiertos.

5.1.1. Patrón temporal

En el artículo 2 de la presente tesis observamos que el número de escribanos

hortelanos aumentó con el número de años transcurridos tras el incendio. Este resultado

Discusión general

134

podría deberse principalmente a dos procesos; 1) nuevos individuos colonizan la zona

quemada y/o 2) individuos que nacen en los primeros años permanecen en la zona en

años posteriores (reclutamiento). Sin embargo, el hecho de que varias de las zonas

quemadas situadas cerca del núcleo poblacional más importante fuesen colonizadas por

la especie después de 4 o 5 años (observación personal), parece indicar que es más

probable la hipótesis de nuevos colonizadores que de reclutamiento. Estas nuevas

colonizaciones podrían estar asociadas a otros procesos estocásticos no considerados en

este trabajo, como por ejemplo la atracción de individuos de la misma especie

(conespecíficos) (Foto 3). La escasez de conespecíficos justo después del incendio

podría hacer que una zona potencialmente buena no fuera colonizada. Sin embargo, el

tiempo transcurrido después del incendio podría contribuir al establecimiento efectivo

de individuos e impulsar la colonización de nuevos individuos por mecanismos de

atracción (Dale et al. 2006).

Foto 3 Macho de escribano hortelano cantando sobre una piedra en el incendio de Cardona el 22 de mayo del 2010, 5 años después del incendio. Estos cantos tienen diferentes funciones: atraer a las hembras, defender el territorio de otros machos y, además, puede servir de reclamo para otros individuos de la misma o de otra especie.

Por otra parte, la colonización de nuevos individuos podría estar relacionada con

los mecanismos de regeneración de la vegetación. El escaso recubrimiento arbustivo y/o

la elevada presencia de árboles muertos en pie inmediatamente después del incendio

podría explicar que el hábitat fuera más adecuado en años posteriores al incendio y

como consecuencia que un mayor número de individuos colonizaran la zona durante los

años siguientes a la perturbación. Sin embargo, en el caso del escribano hortelano no

creemos que esta sea la causa de un mayor número de individuos a medida que va

pasando el tiempo ya que numerosos estudios han mostrado que es una de las primeras

especies que llega a la zona quemada pudiendo incluso observarse el primer año

Discusión general

135

después del incendio (Pons & Prodon 1996, Herrando et al. 2002, Brotons et al. 2005,

Pons & Bas 2005). Además, la mayoría de los incendios estudiados ocurrieron en los

primeros meses del año y las zonas muestreadas el primer año después del fuego

tuvieron una vegetación de entre 0-0,25 m de altura que varió entre el 20 y el 80% de la

cubierta vegetal (observación personal). Por lo tanto, parece que otros procesos

estocásticos contribuyen a explicar el retraso que se produce entre que una zona

quemada es potencialmente buena para la colonización de especies de espacios abiertos

y el momento en el que realmente tiene lugar la colonización. En este sentido,

destacamos la importancia de nuevos estudios que analicen explícitamente algunos de

los procesos aquí mencionados.

5.1.2. Patrón espacial

En los artículos 2 y 3 hemos utilizado la teoría de las metapoblaciones como

punto de partida para cuantificar el patrón de variabilidad espacial en la colonización

postincendio. Según esta teoría, la colonización de una nueva tesela de hábitat depende

del tamaño de la tesela, el grado de aislamiento y la calidad del hábitat (Hanski 1999,

Thomas et al. 2001).

Los resultados mostraron grandes diferencias espaciales entre las distintas zonas

estudiadas respecto a la ocurrencia de aves de hábitats abiertos. La probabilidad de

colonización de estas especies dependió del grado de aislamiento y de la calidad del

hábitat y en menor medida del tamaño del incendio. Para la mayoría de las especies, la

conectividad y las variables utilizadas como indicadores de la calidad del hábitat

tuvieron la misma importancia, aunque este no fue el caso de la collalba rubia

(Oenanthe hispanica) y particularmente de la totovía (Lullula arborea) donde los

parámetros utilizados como indicadores de la calidad del hábitat tuvieron una mayor

importancia en determinar la colonización efectiva de las especies (Fig. 6).

Discusión general

136

Var

ianz

a ex

plic

ada

(R2 )

0

10

20

30

40

50Conectividad Calidad de hábitat

Pardillocomún

Miliariacalandra

Totovía Collalbarubia

Curruca cabecinegra

Fig. 6 Varianza explicada por la conectividad y la calidad del hábitat en la colonización postincendio de 5 especies de hábitats abiertos.

A continuación se explicará la importancia en el proceso de colonización

postincendio de estos tres parámetros (tamaño del incendio, grado de aislamiento y

calidad de hábitat) en función de la escala espacial en la que participan. En este sentido,

asumimos que el área y la conectividad juegan un papel a escala de paisaje o regional

mientras que la calidad del hábitat de la zona quemada interviene a escala local.

5.1.2.1. Factores que actúan a escala regional

Los incendios crean un hábitat adecuado para las aves de hábitats abiertos

proporcionando recursos como disponibilidad de alimento, sitios de cría y una baja

depredación (Pons & Prodon 1996, Dale & Olsen 2002). Sin embargo, nuestros

resultados indican que la localización geográfica de las zonas quemadas y la capacidad

de dispersión de las especies son las piezas claves para que estas especies puedan

colonizar estas áreas (Artículo 2 y 3). La probabilidad de colonización postincendio

disminuyó cuando aumentó el grado de aislamiento (Hanski 1999, Hanski &

Ovaskainen 2003), es decir, cuanto mayor fue la distancia a fuentes potenciales de

colonizadores. Estos resultados subrayan la importancia de la capacidad de dispersión

de las especies en la localización de nuevos hábitats (Johst et al. 2002) y están en

consonancia con otros estudios que muestran diferencias en la ocurrencia de especies

colonizadoras entre diferentes áreas quemadas. Brotons et al. (2005) estudiaron la

capacidad de dispersión de las especies de hábitats abiertos para alcanzar las áreas

Discusión general

137

recientemente quemadas. En su estudio, demostraron que solo aquellas zonas situadas

cerca de poblaciones fuente fueron colonizadas, destacando la importancia de la

distancia de dispersión.

Sorprendentemente, nuestros resultados sugieren que el tamaño del incendio no

es un factor clave en el proceso de colonización de las zonas estudiadas (Artículo 2 y 3).

En teoría, una parcela de hábitat grande presenta una mayor probabilidad de ser

encontrada por nuevos individuos que una parcela de hábitat pequeña y además, es más

probable que estos individuos encuentren hábitat adecuado (Hill et al. 1996, Kuussaari

et al. 1996). Sin embargo, en el caso del escribano hortelano (Artículo 2) este efecto fue

contrarrestado por el efecto de la conectividad y la capacidad de dispersión y en el caso

del Artículo 3, el tamaño del incendio no estuvo relacionado con la probabilidad de

colonización de ninguna de las 5 especies estudiadas. Nuestros resultados indican que a

pesar de que los grandes incendios pueden jugar un papel crítico en el asentamiento de

nuevas poblaciones mediante la creación de grandes teselas de hábitat, estas áreas deben

estar localizadas cerca de poblaciones fuente conectadas por dispersión. Por otra parte,

las diferencias observadas entre el artículo 2 y 3 podrían deberse a que en este último, el

número de incendios es casi tres veces menor que en el estudio del escribano hortelano

(15 versus 44) y el rango de tamaño de los incendios es mucho más pequeño (86-1.438

ha versus 51-4.497ha). En este sentido, más información sobre el papel del tamaño del

incendio en el proceso de colonización parece ser necesaria.

5.1.2.2. Factores que actúan a escala local

La probabilidad de colonización también dependió de factores locales utilizados

como indicadores de la calidad del hábitat (Thomas et al. 2001). Los factores que

limitaron la colonización estuvieron principalmente relacionados con la topografía

(pendiente y orientación) y la vegetación antes del incendio aunque también se

analizaron otros parámetros como la severidad del fuego. Además, la relación de estos

factores dependió de los requerimientos de cada una de las especies, indicando

diferencias interespecíficas en la respuesta al fuego (McCauley 2007).

En cuanto a la vegetación antes del incendio, muchos estudios han subrayado la

importancia de este parámetro para entender la comunidad de aves después del incendio

(Prodon & Lebreton 1981, Pons & Prodon 1996, Pons & Bas 2005). En el artículo 1

Discusión general

138

observamos que los incendios ocurridos en zonas de matorral tiene un mayor porcentaje

de especies de hábitats abiertos que los incendios ocurridos en un bosque,

probablemente porque las especies ya estaban presentes antes del incendio (Pons &

Prodon 1996). Por el contrario, los incendios ocurridos en un bosque incluyen una alta

densidad de árboles muertos, y no ofrecen el hábitat adecuado para la colonización de

estas especies inmediatamente después de los mismos (Moreira et al. 2003) (Foto 4). En

este caso, la gestión postincendio juega un papel muy importante ya que convierte una

estructura de bosque quemado a hábitats abiertos o de matorral, facilitando de esta

manera la colonización por parte de las especies de hábitats abiertos (Castro et al. 2009)

(Foto 4). Además, las zonas boscosas tienden a estar asociadas con zonas más

productivas (Thuiller et al. 2003), lo que favorece la recuperación de la vegetación y

reduce el tiempo en el que los hábitats abiertos están potencialmente disponibles para la

colonización de las aves de hábitats abiertos. Sin embargo, no todas las zonas boscosas

son iguales ya que tal y como hemos visto en el artículo 4 los cambios en el paisaje

inducidos por la escasa capacidad regenerativa del pino laricio (Pinus nigra) después

del fuego conducen al mantenimiento a largo plazo del hábitat adecuado para las

especies de hábitats abiertos.

Foto 4 Elevada densidad de árboles muertos en pie justo después de un incendio en una zona forestal (izquierda) y comienzo de las tareas de tala y recogida de los árboles muertos quemados en otro de los incendios estudiados (derecha).

Por otra parte, la severidad del incendio no parece tener un impacto en el

proceso de colonización de las especies de hábitats abiertos. En su estudio sobre el

impacto de la severidad en las especies, Kotliar et al. (2007) encontraron que la mayoría

de las especies mostraba un impacto positivo o neutro. Sin embargo, en nuestro estudio

sólo la collalba rubia respondió positivamente. Las diferencias en el diseño entre las dos

zonas de estudio podría explicar estas divergencias en los resultados, ya que Kotliar et

Discusión general

139

al. (2007) estudiaron un único incendio a altitudes más altas. Más información sobre el

impacto directo del fuego en el proceso de colonización resulta imprescindible.

5.2. Procesos históricos que ayudan a predecir las nuevas colonizaciones

El paisaje mediterráneo está sufriendo dos fuerzas opuestas: el abandono de las

tierras no productivas y el incremento del impacto de los incendios (Moreira & Russo

2007). Ambos procesos han afectado principalmente la disponibilidad y la disposición

espacial de los hábitats abiertos. El fuego se ha convertido en el principal proceso a

escala de paisaje que crea nuevos hábitats favorables para las especies de hábitats

abiertos (Brotons et al. 2008, Vallecillo et al. 2009). En este contexto, en el artículo 5 se

testó la siguiente hipótesis: Si los incendios lideran la dinámica de distribución de estas

especies, se esperaría que la trayectoria histórica de los incendios explique la

colonización postincendio de las aves de hábitats abiertos mejor que los otros dos

escenarios (otros matorrales y zonas agrícolas) (Fig. 7).

Fig. 7 Patrones de paisaje que proporcionan hábitat potencialmente adecuado para las aves de hábitats abiertos (matorral originado como consecuencia de los incendios ocurridos en Cataluña en las dos últimas décadas del siglo XX, otros matorrales y zonas agrícolas). Estos paisajes podrían actuar como hábitat fuente en la colonización de nuevos hábitats que aparecen en el paisaje como consecuencia de nuevos incendios.

Nuestros resultados apoyaron la hipótesis inicial de que el régimen actual de

incendios juega un papel importante en la colonización por parte de especies de hábitats

abiertos de nuevos hábitats originados como consecuencia de nuevos incendios. La

probabilidad de colonización fue mayor en áreas situadas cerca de otras zonas

previamente afectadas por el fuego (Artículo 5). Además, el efecto positivo de los

Discusión general

140

incendios más antiguos (10-20 años) indica que el impacto del fuego en la dinámica de

las especies de hábitats abiertos se extiende a periodos de tiempo relativamente largos

(Vallecillo et al. 2009). Sin embargo, es poco probable que este impacto se extienda a

periodos temporales más largos ya que en general la vegetación mediterránea es muy

resistente (resilience, en inglés) al efecto del fuego (Hanes 1971, Lloret 1998) y las

especies de hábitats abiertos tienden a desaparecer algunos años después del fuego

(Prodon & Lebreton 1981, Jacques & Prodon 2009). En este caso, los individuos se

verían forzados a abandonar estas áreas y a colonizar nuevas teselas de hábitats, lo que

sugiere que los incendios funcionarían como fuentes no permanentes de colonizadores.

Además, la conectividad de las zonas agrícolas también jugó un papel relevante

en la colonización postincendio de especies de hábitats abiertos (Artículo 5). Aunque la

importancia de estas zonas para la alimentación y la cría de estas especies es bien

conocida (e.g. Sanderson et al. 2005), nuestros resultados destacan el papel de zonas

agrícolas bien conectadas en la dinámica espacio-temporal de estas especies. Este

resultado sugiere que, a pesar de la disminución global de la cubierta de cultivos

extensivos durante las últimas décadas y el efecto negativo de la intensificación en las

especies agrícolas (Mañosa et al. 1996, Brotons et al. 2004), la actual configuración

espacial de zonas agrícolas todavía juega un papel notable como fuente permanente de

colonizadores de especies de hábitats abiertos. Sin embargo, si la disminución de las

zonas agrícolas continúa, como parece ser la tendencia actual, procesos como las

degradación del hábitat o la fragmentación, actualmente presentes en otros países de

Europa, podrían traer consecuencias dañinas sobre dichas especies (Donald et al. 2001).

Finalmente, sólo una de las seis especies estudiadas estuvo relacionada con la

conectividad de las zonas de matorral no afectadas por el fuego (Artículo 5). Este

resultado nos indica que este escenario de paisaje es menos adecuado para la

persistencia de especies de hábitats abiertos (Sirami et al. 2007). Uno de los procesos

relacionados con este escenario es el abandono rural. En este sentido, el abandono rural

favorece la recuperación de la vegetación (Debussche et al. 1987) lo que conduce a la

desaparición progresiva y la invasión de los hábitats abiertos (Debussche et al. 1999,

Romero-Calcerrada & Perry 2004). Aunque las especies de aves que ocupan el matorral

mediterráneo se beneficiarían en un primer momento del cambio de pradera y viejos

campos a un matorral abierto, sus poblaciones por último disminuirían tan pronto la

sucesión vegetal siga su camino a un matorral denso (Preiss et al. 1997).

Discusión general

141

En resumen, los cambios en los usos del suelo en las últimas décadas han

originado un cambio de los procesos ecológicos que actúan sobre los reservorios en la

dinámica de especies de hábitats abiertos en la región mediterránea. De una

configuración espacial de hábitats abiertos constituida por hábitats relativamente

estáticos (zonas agrícolas y prados) antes de la mitad del siglo XX, a un conjunto

dinámico de teselas de hábitats donde el fuego juega un doble papel. Por una parte, los

incendios crean un hábitat adecuado para la colonización de estas especies y por otro,

las zonas quemadas actúan como hábitat fuente, proporcionando individuos a los

nuevos hábitats que aparecen en el paisaje (Artículo 5).

5.3. Aplicación de los índices de conectividad a modelos de distribución

dinámicos

Nuestros resultados destacan la importancia del uso de índices de conectividad

para predecir la distribución de especies en un paisaje dinámico (Artículo 2 y 3). De esta

manera, modelos basados exclusivamente en la cantidad de hábitat en el paisaje no

serán realistas para especies como las estudiadas en el presente trabajo si no se incluye

la capacidad de dispersión de las especies. Los modelos de predicción de las respuestas

en la distribución de las especies a las perturbaciones deberían incluir explícitamente los

límites impuestos por la dispersión y el contexto en el que ocurren para predecir los

cambios en el patrón espacial. Una posibilidad sería la integración en los modelos de

hábitat de la información generada tras utilizar análisis relacionados con la teoría de

grafos, como por ejemplo la cantidad potencial de flujo de dispersión estimada a partir

de los índices de conectividad estudiados en la presente tesis, IIC y PC (Saura &

Pascual-Hortal 2007, Saura & Rubio 2010, Visconti & Elkin 2010). En general, las

estructuras de grafos ofrecen una versátil y flexible representación de los mosaicos de

hábitat y puede proporcionar nuevos detalles sobre cuestiones ecológicas diversas a

escala local y de paisaje (Urban et al. 2009). Nuestros resultados junto con otros

recientemente publicados (McRae & Beier 2007, Neel 2008), contribuyen a demostrar

la habilidad de explicar procesos ecológicos relevantes a través de una representación

del paisaje de estructura de grafos.

Conclusiones finales

143

6. CONCLUSIONES FINALES

1. La base de datos creada durante la elaboración de esta tesis, DINDIS, proporciona

un marco de estudio adecuado en el analisis de los factores que actúan a diferentes

escalas (espaciales y temporales) en la colonización postincendio de las aves de

hábitats abiertos (Artículo 1).

2. La respuesta de las aves a los incendios es compleja y heterogénea. El tipo de

vegetación antes del incendio y la posterior gestión postincendio y el contexto

paisajístico en el cual tienen lugar los incendios son factores esenciales para

entender la comunidad de aves que aparece después de la perturbación (Artículo 1).

3. Los incendios crean hábitat adecuado para numerosas especies de aves de hábitats

abiertos con un gran interés para la conservación, algunas de las cuales son

relativamente abundantes en estas zonas (Artículo 1).

4. A escala temporal, determinados procesos estocásticos contribuyen a explicar el

retraso que se produce entre que una zona quemada es adecuada para la

colonización de especies de aves de hábitats abiertos y cuando tiene lugar realmente

la colonización (Artículo 2).

5. A escala espacial, la colonización postincendio de las aves de hábitats abiertos está

limitada por tres factores: la conectividad, la calidad del hábitat y el tamaño del

incendio, siendo los dos primeros los más importantes. En este sentido:

a. Los factores que determinan la colonización a escala regional son la

localización geográfica del incendio y la capacidad de dispersión de las

especies (Artículo 2 y 3).

b. Los factores que limitan la colonización a escala local están principalmente

relacionados con la topografía (pendiente y orientación) y la vegetación antes

del incendio (Artículo 3).

c. El efecto del tamaño del incendio sobre la colonización parece ser

contrarestado por el efecto de la conectividad. Por lo tanto, a pesar de que los

Conclusiones finales

144

grandes incendios pueden jugar un papel crítico en el asentamiento de nuevas

poblaciones mediante la creación de grandes teselas de hábitat favorable, estas

áreas deben estar localizadas cerca de poblaciones fuente conectadas por

dispersión (Artículo 2 y 3).

6. La predicción de las respuestas en la distribución de las especies a las

perturbaciones debería incluir explícitamente las limitaciones impuestas por la

dispersión y el contexto donde tienen lugar. Una posibilidad es integrar en los

modelos de hábitat información generada a partir de análisis utilizando la teoría de

grafos, como por ejemplo la cantidad potencial de flujo de dispersión, como la

estimada a través de los índices IIC y PC (Artículo 2 y 3).

7. Los cambios en el paisaje inducidos por la escasa capacidad regenerativa del pino

laricio (Pinus nigra) después del fuego conducen al mantenimiento a largo plazo del

hábitat adecuado para las aves asociadas a hábitats abiertos (Artículo 4).

8. Los cambios en los usos del suelo ocurridos durante las últimas décadas del siglo

XX en la región mediterránea han provocado un cambio en los procesos ecológicos

que actúan sobre los reservorios en las dinámicas de especies de hábitats abiertos.

De un conjunto de hábitats abiertos relativamente estáticos (zonas agrícolas y

prados) a un mosaico de hábitats donde los incendios juegan un doble papel. Por una

parte, los incendios crean el hábitat adecuado para la colonización de las especies de

hábitats abiertos y por otra parte, las zonas incendiadas actúan como hábitat fuente,

proporcionando individuos a los nuevos hábitats que aparecen en el paisaje

(Artículo 5).

Referencias bibliográficas

145

7. REFERENCIAS BIBLIOGRÁFICAS

Arnan X., Rodrigo A., Retana J. (2006) Post-fire recovery of Mediterranean ground ant

communities follows vegetation and dryness gradients. Journal of Biogeography 33:

1246-1258.

Artman V.L., Sutherland E.K., Downhower J.F. (2001) Prescribed burning to restore

mixed-oak communities in Southern Ohio: effects on breeding-bird populations.

Conservation biology 5: 1423-1434.



Series No. 12).

Brawn J.D., Robinson S.K., Thompson F.R. (2001) The role of disturbance in the

ecology and conservation of birds. Annual review of ecology and systematics 32:

251-276.

Brotons L., Mañosa S., Estrada J. (2004) Modelling the effects of irrigation schemes on

the distribution of steppe birds in Mediterranean farmland. Biodiversity and

Conservation 13: 1039-1058.

Brotons L., Pons P., Herrando S. (2005) Colonization of dynamic Mediterranean

landscapes: where do birds come from after fire? J. Biogeogr. 32: 789-798.

Brotons L., Herrando S., Pons P. (2008) Wildfires and the expansion of threatened


landscapes. Journal of Applied Ecology 45: 1059-1066.

Bullock J. M. et al. (2002) Dispersal ecology. Blackwell.

Bunn A.G., Urban D.L., Keitt T.H. (2000) Landscape connectivity: A conservation

application of graph theory. J. Environ. Manage. 59: 265-278.

Burriel J.A., Gracia C., Ibáñez J.J., Mata T., Vayreda J. (2000-2004) Inventari Ecològic

i Forestal de Catalunya (IEFC). CREAF, Bellaterra.

Castro J., Moreno-Rueda G., Hódar J.A. (2010) Experimental test of postfire

management in pine forests: impact of salvage logging versus partial cutting and

non-intervention on bird–species assemblages. Conservation biology 24: 810-819.

Clavero P., Martin Vide J., Raso Nadal J.M. (1996) Atles climatic de Catalunya.

Termopluviometria. Generalitat de Catalunya (Departament de Politica Territorrial i


146

Obres Públiques), Institut Cartogràfic de Catalunya i Departament de Medi

Ambient, Barcelona.

Clobert J., Danchin E., Dhondt A.A., Nichols J.D. (2001) Dispersal. Oxford University

Press.

Dale S., Olsen B.F.G. (2002) Use of farmland by Ortolan Buntings (Emberiza

hortulana) nesting on a burned forest area. Journal für Ornithologie 143: 133–144.

Dale S., Lunde A., Steifetten O. (2005) Longer breeding dispersal than natal dispersal in

the ortolan bunting. Behavioral ecology 16: 20-24.

Dale S., Steifetten O., Osiejuk T.O., Losak K., Cygan J.P. (2006) How do birds search

for breeding areas at the landscape level? Interpatch movements of male Ortolan

Bunting. Ecography 29: 886-898.

Debussche M., Rambal S., Lepart J. (1987) Land-use changes in the humid

Mediterranean region: estimate of the hydrological impact. Acta Oecologica 8: 317-

332.

Debussche M., Lepart J., Dervieux A. (1999) Mediterranean landscape changes:

evidence from old postcards. Global Ecology and Biogeography 8: 3-15.

Díaz-Delgado R., Lloret F., Pons X. (2002) Satellite evidence of decreasing resilience in

Mediterranean plant communities after recurrent wildfires. Ecology 83: 2293-2303.

Díaz-Delgado R., Lloret F., Pons X. (2004) Spatial patterns of fire occurrence in

Catalonia, NE, Spain. Landscape Ecology 19: 731-745.

DGMN (Direcció General del Medi Natural). (1994). Plà General de Política Forestal.

Donald P.F., Green R.E., Heath M.F. (2001) Agricultural intensification and the

collapse of Europe’s farmland bird populations. Proceedings of the Royal Society of

London, Series B: Biological Sciences 268: 25–29.

Espelta J.M., Rodrigo A., Habrouk A., Meghelli N., Ordoñez J.L., Retana J. (2002)

Land use changes, natural regeneration patterns, and restoration practices after a

large wildfire in NE Spain: challenges for fire Ecology and landscape restoration.

Pag. 315-324 en: L. Trabaud y R. Prodon (editores). Fire and Biological processes.

Backhuys, Leiden.

Estrada J., Pedrocchi V., Brotons L., Herrando S. (2004) Atles dels ocells nidificants de

Catalunya (1999-2002). Lynx, ICO (Institut Català d'Ornitologia), Barcelona.

Forman R.T.T., Godron M. (1986) Landscape Ecology. John Wiley, New York.

Forman R. (1995) Lands mosaics: the ecology of landscape and regions. Cambridge

University Press. Cambridge.


147

Greenwood P.J., Harvey P.H. (1982) The natal and breeding dispersal of birds. Ann Rev

Ecol Syst 13:1–21.

Hanes T.L. (1971) Succession after fire in the chaparral of southern California.

Ecological monographs 41: 27-52.

Hanski I. (1999) Metapopulation ecology. Oxford University Press, Oxford.

Hanski I., Ovaskainen O. (2003) Metapopulation theory for fragmented landscapes.


Harary F. (1969) Graph theory. Reading, MA.

Herrando S, Brotons L. (2002) Forest bird diversity in Mediterranean areas affected by

wildfires: a multi-scale approach. Ecography 25: 161-172.

Herrando S., Brotons L., del Amo R., Llacuna S. (2002) Bird community succession

after fire in a dry mediterranean shrubland. Ardea 90: 303-310.

Hill J.K., Thomas C.D., Lewis O.T. (1996) Effects of habitat patch size and isolation on

dispersal by Hesperia comma butterflies: implications for metapopulation structure.

Journal of animal ecology 65: 725-735.

Hutto R.L. (2006) Toward meaningful snag-management guidelines for postfire salvage

logging in North American conifer forests. Conservation Biology 20: 984-993.

Jacquet K., Prodon R. (2009) Measuring the postfire resilience of a bird–vegetation

system: a 28-year study in a Mediterranean oak woodland. Oecologia 161: 801-811.

Johst K., Brandl R., Eber S. (2002) Metapopulation persistence in dynamic landscapes:

the role of dispersal distance. Oikos 98: 263-270.

Kazanis D., Arianotsou M. (2004) Factors determining low Mediterranean ecosystems

resilience to fire: the case of Pinus halepensis forests. In ‘Proceedings of the 10th

MEDECOS Conference’, 25 April–1 May 2004, Rhodes, Greece.

Kiss L., Magnin F. (2003) The impact of fire on some Mediterranean land snail

communities and patterns of post-fire recolonization. J. Moll. Stud. 69: 43-53.

Kotliar N.B., Kennedy P.L., Ferree K. (2007) Avifauna responses to fire in

southwestern montane forests along a burn severity gradient. Ecological

Applications 17: 491-507.

Kuussaari M., Nieminen M., Hanski I. (1996) An experimental study of migration in the

Glanville fritillary butterfly Melitaea cinxia. Journal of animal ecology 65: 791-

801.


148

Lawrence G.E. (1966) Ecology of vertebrate animals in relation to chaparral fire in the

Sierra Nevada foothills. Ecology 47: 278-291.

Lloret F. (1998) Fire, canopy cover and seedling dynamics in Mediterranean shrubland

of northeastern Spain. Journal of Vegetation Science 9: 417-430.

Malamud B.D., Morein G., Turcotte D.L. (1998) Forest fires: an example of self-

organized critical behavior. Science 281: 1840-1842.

Mañosa S., Estrada J., Folch A., Bonfil J., Gonzàlez-Prat F., Orta J. (1996) Bird-habitat

relationships in the Catalan steppes. Conservación de las Aves Esteparias y sus

Hábitats (ed. by J. Fernández Gutiérrez J. & J. Sanz-Zuasti), pp. 153–161. Junta de

Castilla y León, Valladolid, Spain.

MacArthur R.H., Wilson E.O. (1967) The theory of island biogeography. Princeton

University Press, Princeton, NJ.

McCauley S.J. (2007) The role of local and regional processes in structuring larval

dragonfly distributions across habitat gradients. Oikos 116: 121-133.

McRae B.H., Beier P. (2007) Circuit theory predicts gene flow in plant and animal

populations. Proceedings of the National Academy of Sciences of the USA 104:

19885-19890.

Moreira F., Ferreira P.G., Rego F.C., Bunting S. (2001) Landscape changes and

breeding bird assemblages in northwestern Portugal: implications for fire

occurrence. Landscape Ecology 16: 175-187.

Moreira F., Delgado A., Ferreira S., Borralho R., Oliveira N., Inácio M., Silva J.S.,

Rego F. (2003) Effects of prescribed fire on vegetation structure and breeding birds

in young Pinus pinaster stands of northern Portugal. Forest Ecology and


Moreira F., Russo D. (2007) Modelling the impact of agricultural abandonment and

wildfires on vertebrate diversity in Mediterranean Europe. Landscape ecology 22:

1461-1476.

Moreno J.M., Vázquez A., Vélez R. (1998) Recent history of forest fires in Spain. In

Moreno, J. M. (ed.), Large Fires, Backhuys Publishers, Leiden, The Netherlands, pp.

159-185.

Naveh Z. (1975) The evolutionary significance of fire in the Mediterranean region.



149

Neel M.C. (2008) Patch connectivity and genetic diversity conservation in the federally

endangered and narrowly endemic plant species Astragulus albens (Fabaceae).

Biological Conservation 141: 938-955.

Paradis E., Baillie S.R., Sutherland W.J., Gregory R.D. (1998) Patterns of natal and

breeding dispersal in birds. J Anim Ecol 67:518-536.

Pascual-Hortal L., Saura S. (2006) Comparison and development of new graph-based

landscape connectivity indices: towards the priorization of habitat patches and

corridors. Landscape Ecology 21: 956-967.

Pausas J.G., Vallejo R. (1999) The role of fire in European Mediterranean ecosystems.

In ‘Remote Sensing of Large Wildfires in the European Mediterranean Basin’. (Ed.

E Chuvieco) pp. 3–16. (Springer: Berlin).

Pausas J.G. (2004a). Changes in fire and climate in the eastern Iberian Peninsula

(Mediterranean Basin). Climatic Change 63: 337-350.

Pausas J.G. (2004b). La recurrencia de incendios en el monte mediterráneo. Avances en

el estudio de la gestión del monte Mediterráneo. Págs. 47-64 Vallejo V.R., Alloza

J.A. eds. Fundación CEAM, 2004.

Pausas J. G., Keeley J.E. (2009) A burning story: the role of fire in the history of life.

Bioscience 59: 593-601.

Peñuelas J. (1996) Sobre el canvi climàtic i la manera com pot repercutir en els

incendis. En “Ecologia del Foc”. Terradas, J (edit.). Editorial Proa. Barcelona.

Pickett S.T.A., Kolasa J., Armesto J.J., Collins S.L. (1989). The ecological concept of

disturbance and its expression at various hierarchical levels. Oikos 54: 129-136.

Piñol J., Terradas J., Lloret F. (1998) Climate warming, wildfire hazard, and wildfire

occurrence in coastal eastern Spain. Climate Change 38: 345-357.

Platt W., Connell J. (2003) Natural disturbances and directional replacement of species.

Ecological Monographs 73: 507-522.

Peix J. (1999) Foc Verd II. Programa de gestió del risc d.incendi forestal. Barcelona:

Departament d.Agricultura, Ramaderia i Pesca, Generalitat de Catalunya, p. 26.

ISBN: 84-393-4789-8.

Pons P., Prodon R. (1996) Short term temporal patterns in a Mediterranean shrubland

bird community after wildfire. Acta Oecologica 17: 29-41.

Pons P. (1999): Brulage dirige et incendie sauvage: ont-t-ils l’un et l’autre le meˆme

impact sur l’avifaune? Foret Mediterraneene 20: 103-113.


150

Pons P., Bas J.M. (2005) Open-habitat birds in recently burned areas: The role of the

fire extent and species’ habitat breadth. Ardeola 52: 119-131.

Preiss E., Martin J.L., Debussche M. (1997) Rural depopulation and recent landscape

changes in a Mediterranean region: consequences to the breeding avifauna.

Landscape Ecology 12: 51-61.

Prodon R., Lebreton J.D. (1981) Breeding avifauna of a Mediterranean succession: the

holm oak and cork oak series in the eastern Pyrenees. Analysis and modelling of the

structure gradient. Oikos 37: 21-38.

Prodon R., Fons R., Athias-Binche F. (1987) The impact of fire on animal communities

in Mediterranean area. In: Trabaud L. (ed.) The role of fire on ecological systems:

121-157. SPB Academic Publ., The Hague.

Pulliam H.R. (1988) Sources, sinks and population regulation. American Naturalist

132: 652-661.

Purves D.W., Zavala M.A.. Ogle K., Prieto F., Benayas J.M.R. (2007) Environmental

heterogeneity, bird-mediated directed dispersal, and oak woodland dynamics in

Mediterranean Spain. Ecological monographs 77: 77-97.

Rodrigo A., Retana J., Picó X.F. (2004) Direct regeneration is not the only response of


Romero-Calcerrada R., Perry G.L.W. (2004) The role of land abandonment in landscape

dynamics in the SPA “Encinares del rio Alberche y Cofio”, Central Spain, 1984-

1999. Landscape and Urban Planning 66: 217-232.

Sanderson F.J., Donald P.F., Burfield I.J. (2005) Farmland birds in Europe: from policy

change to population decline and back again. Ecology and conservation of steppe-

land birds (ed. by G. Bota, M.B. Morales, S. Mañosa & J. Camprodon), pp. 211-

236. Lynx Edicions & Centre Tecnològic Forestal de Catalunya, Barcelona.

Saura S., Pascual-Hortal L. (2007) A new habitat availability index to integrate

connectivity in landscape conservation planning: comparison with existing indices

and application to a case study. Landscape and Urban Planning 83: 91-103.

Saura S. (2008) Evaluating forest landscape connectivity through Conefor Sensinode

2.2.–In: 812 Lafortezza, R. et al. (eds), Patterns and Processes in Forest Landscapes:

Multiple Use and 813 Sustainable Management. Springer, pp. 403–422.

Saura S., Torné J. (2009) Conefor Sensinode 2.2: a software package for quantifying the

importance of habitat patches for landscape connectivity. Environmental Modelling

and Software 24: 135-139.


151

Saura S., Rubio L. (2010) A common currency for the different ways in which patches

and links can contribute to habitat availability and connectivity in the landscape.

Ecography 33: 523-537.

Sirami C., Brotons L., Martin J.L. (2007) Vegetation and songbird response to land

abandonment: from landscape to census plot. Diversity and Distributions 13: 42-52.

Taylor P.D., Fahrig L., Henein K. and Merriam G. (1993) Connectivity is a vital

element of landscape structure. Oikos 3: 571–573.

Thomas J.A., Bourn N.A.D., Clarke R.T., Stewart K.E. Simcox D.J., Pearman G.S.,

Curtis R., Goodger B. (2001) The quality and isolation of habitat patches both

determine where butterflies persist in fragmented landscapes. Proc. R. Soc. Lond. B

268: 1791-1796.

Thuiller W., Vayreda J., Pino J., Sabate S., Lavorel S., Gracia C. (2003) Large-scale

environmental correlates of forest tree distributions in Catalonia (NE Spain). Global

Ecology and Biogeography 12: 313-325.

Trabaud L. (1993) Recovery following fire of woody plant communities in Alberes

(western Pyrenees, France). Vie et milieu 43: 43-51.

Ukmar E., Battisti C., Luiselli L., Bologna M.A. (2007) The effects of fire on

communities, guilds and species of breeding birds in burnt and control pinewoods in

central Italy. Biodiversity and Conservation 16: 3287-3300.

Urban D., Keitt T. (2001) Landscape connectivity: A graph-theoretic perspective.

Ecology 82: 1205-1218.

Urban D.L., Minor E.S., Treml E.A., Schick R.S. (2009) Graph models of habitat

mosaics. Ecology Letters 12: 260-273.

Vallecillo S., Brotons L., Thuiller W. (2009) Dangers of predicting bird species

distributions in response to land-cover changes. Ecological Applications 19: 538-

549.

Vázquez A., Moreno J.M. (2001) Spatial distribution of forest fires in Sierra de Gredos

(Central Spain). Forest ecology and management 147: 55-65.

Vepsalainen V., Pakkala T., Piha M., Tiainen J. (2005) Population crash of the ortolan

bunting Emberiza hortulana in agricultural landscapes of southern Finland. Annales

Zoologici Fennici 42: 91-107.

Visconti P., Elkin C. (2009) Using connectivity metrics in conservation planning –

when does habitat quality matter? Diversity and Distributions 15: 602-612.


152

Wiens J.A. (2001) The landscape context of dispersal. Dispersal (ed. by J. Clobert, E.


Wilson C.W., Masters R.E., Bukenhofer G.A. (1995) Breeding bird response to pine-

grassland community restoration for Red-cockaded woodpeckers. Journal of wildlife

management 59: 56-67.

.

Publicaciones

153

8. PUBLICACIONES

Zozaya E.L., Brotons L., Vallecillo S. (aceptado) Bird community responses to

vegetation heterogeneity following non-direct regeneration of Mediterranean forests

after fire. Ardea.

Zozaya E.L., Brotons L., Herrando S., Pons P., Rost J., Clavero M. (aceptado)

Monitoring spatial and temporal dynamics of bird communities in Mediterranean

landscapes affected by large wildfires. Ardeola.

Zozaya E.L., Brotons L., Saura S., Pons P., Herrando S. (enviado) Functional

connectivity determines the post-fire colonisation of an open-habitat bird species.

Journal of Avian Biology.

Zozaya E.L., Brotons L., Saura S. (manuscrito) Recent fire history determines species

distribution dynamics in landscapes dominated by land abandonment.

Zozaya E.L., Brotons L. (manuscrito). Influence of habitat quality, patch size and

connectivity on colonisation of open-habitat bird species after fire.

La ignorancia genera confianza más frecuentemente que el conocimiento, son aquellos que saben poco, y no esos que saben más,

quienes tan positivamente afirman que este o aquel problema nunca será resuelto por la ciencia

Charles Darwin

PATRONES DE COLONIZACIÓN ... -...

Documents

Transcript of PATRONES DE COLONIZACIÓN ... -...