Ecología y gestión de depredadores generalistas: el caso del zorro (Vulpes vulpes) y la urraca (Pica pica)

Memoria presentada por Francisco Díaz Ruiz para optar al grado de Doctor

VºBº Directores Dr. Pablo Ferreras de Andrés

Dr. Miguel Delibes Mateos

Instituto de Investigación en Recursos Cinegéticos (IREC-CSIC-UCLM-JCCM) Departamento de Ciencia y Tecnología Agroforestal Universidad de Castilla-La Mancha

Índice INTRODUCCIÓN GENERAL ................................................................................................. 4 Relación histórica entre hombres y depredadores ................................................................... 4 Impacto de la depredación en las presas: depredadores especialistas y generalistas ................ 6 Factores que favorecen a los depredadores generalistas ......................................................... 8 Ecología del zorro y la urraca: paradigma de especies generalistas....................................... 11 El control de depredadores como herramienta de gestión y conservación ............................. 12 Efectos derivados del control de depredadores ..................................................................... 14 Efecto sobre las presas .................................................................................................... 14 Efecto sobre los depredadores generalistas objeto de control ............................................ 15 Efecto sobre especies que no son objeto de control .......................................................... 16 El control de depredadores en España .................................................................................. 18 Regulación legal del control de depredadores .................................................................. 20 Métodos de control de depredadores generalistas ............................................................. 21 Efectos del control de depredadores en España ................................................................ 23 OBJETIVOS Y ESTRUCTURA DE LA TESIS ...................................................................... 25 CAPÍTULO 1: Biogeographical patterns in the diet of an opportunistic predator: the red fox Vulpes vulpes in the Iberian Peninsula ..................................................................................... 27 Abstract .............................................................................................................................. 28 Introduction ........................................................................................................................ 29 Material and Methods ......................................................................................................... 32 Results ................................................................................................................................ 35 Discussion .......................................................................................................................... 40 Acknowledgements ............................................................................................................. 45 CAPÍTULO 2: Factors affecting the feeding habits of black-billed magpies Pica pica during the breeding season in Mediterranean Iberia.................................................................................. 46 Abstract .............................................................................................................................. 47 Introduction ........................................................................................................................ 48 Material and Methods ......................................................................................................... 49 Results ................................................................................................................................ 52 Discussion .......................................................................................................................... 58 Acknowledgements ............................................................................................................. 60 Ethical standards ................................................................................................................. 61

1

CAPÍTULO 3: An evaluation of cage-traps and the Collarum device to capture red foxes (Vulpes vulpes). Can the performance of cage-traps be improved by baits and scent attractants? ............................................................................................................................................... 62 Abstract .............................................................................................................................. 63 Introduction ........................................................................................................................ 65 Material and Methods ......................................................................................................... 67 Results ................................................................................................................................ 73 Discussion .......................................................................................................................... 81 Acknowledgements ............................................................................................................. 84 Ethical standards ................................................................................................................. 84 CAPÍTULO 4: Experimental evaluation of live cage-traps for Black-billed magpies Pica pica management in Spain .............................................................................................................. 85 Abstract .............................................................................................................................. 86 Introduction ........................................................................................................................ 87 Materials and Methods ........................................................................................................ 89 Results ................................................................................................................................ 94 Discussion ........................................................................................................................ 101 Acknowledgements ........................................................................................................... 103 Ethical standards ............................................................................................................... 104 CAPÍTULO 5: Assessing the influence of predator control on target and non-target predator populations using occupancy models ..................................................................................... 105 Abstract ............................................................................................................................ 106 Introduction ...................................................................................................................... 107 Material and Methods ....................................................................................................... 109 Results .............................................................................................................................. 114 Discussion ........................................................................................................................ 120 Acknowledgements ........................................................................................................... 123 Ethical standards ............................................................................................................... 123 CAPÍTULO 6: Drivers of red fox (Vulpes vulpes) daily activity: prey availability, human disturbance or habitat structure? ............................................................................................ 124 Abstract ............................................................................................................................ 125 Introduction ...................................................................................................................... 126 Material and Methods ....................................................................................................... 128 Results .............................................................................................................................. 133 Discussion ........................................................................................................................ 138 Acknowledgements ........................................................................................................... 140 2

Ethical standards ............................................................................................................... 140 DISCUSIÓN GENERAL ...................................................................................................... 141 Ecología trófica del zorro y la urraca ................................................................................. 141 Evaluación y mejora de los métodos de captura para el control de zorros y urracas ............ 145 Efectos del control de depredadores sobre las poblaciones de zorros y urracas ................... 151 Efectos sobre otras especies no objeto de control ............................................................... 153 Efectos sobre el comportamiento de los depredadores objeto de control ............................. 155 Futuras líneas de investigación .......................................................................................... 156 CONCLUSIONES ................................................................................................................ 159 REFERENCIAS ................................................................................................................... 163 APÉNDICES ........................................................................................................................ 197 Appendix 1.1. ................................................................................................................... 198 Appendix 1.2 .................................................................................................................... 203 Appendix 2.1. ................................................................................................................... 208 Appendix 2.2. ................................................................................................................... 209 Appendix 2.3. ................................................................................................................... 210 Appendix 3.1. ................................................................................................................... 211 Appendix 3.2. ................................................................................................................... 214 Appendix 4.1. ................................................................................................................... 215 Appendix 5.1. ................................................................................................................... 216 Appendix 5.2. ................................................................................................................... 217

3

INTRODUCCIÓN GENERAL Relación histórica entre hombres y depredadores El hombre tiene una larga historia de coexistencia con los depredadores que probablemente comenzó como una relación depredador-presa, en la que los primeros homínidos habrían sido presas de los grandes depredadores (Headland y Greene 2011; Njau y Blumenschine 2012). El hombre, como presa, desarrolló en primer lugar un sentimiento de temor ante los depredadores por riesgo a ser depredado. Con el paso del tiempo, el hombre se convirtió en un eficiente depredador al aprender a utilizar diversas herramientas que le confirieron la capacidad de defenderse de los depredadores y la posibilidad de cazar grandes presas. (McCade y McCade 1984; Vargas 2002). Desde ese momento, el hombre percibe a otros depredadores como competidores por alimentarse de presas de interés humano (Conover 2002; Vargas 2002). La persecución de los depredadores por parte del hombre pudo comenzar, por lo tanto, hace muchísimos años por lo que se trataría de una actividad muy antigua, y extendida por todo el mundo. Quizás los casos más conocidos sean los de los grandes carnívoros como el lobo (Canis lupus) en Europa, Asia y América (Musiani y Paquet 2004; SilleroZubiri y Schwitzer 2004) o los grandes felinos en África, Asia y América (Woodroffe y Frank 2005; Balme et al. 2009; Inskip y Zimmerman 2009), los cuales consumen diferentes especies de ganado o incluso atacan a los propios humanos (Treves y Karanth 2003). No obstante, existen también numerosos ejemplos de otros depredadores de menor tamaño que han sido perseguidos por ser potenciales depredadores de especies de caza menor, piscícolas, ganado e incluso por ser considerados como perjudiciales para la agricultura. Entre estos destacan carnívoros de pequeña y mediana talla (Reynolds y Tapper 1996; Virgós y Travaini 2005), rapaces en general (Villafuerte et al.1998; Thirgood et al. 2000a; Whitfield et al. 2003, Whitfield et al. 2007) e incluso algunos córvidos (Hadjisterkotis 2003; Madden et al. en prensa). En este sentido España no ha sido una excepción, y la persecución de depredadores ha sido una actividad muy extendida y arraigada desde hace mucho tiempo como así acreditan diferentes documentos históricos. Archivos históricos constatan una persecución organizada e impuesta de osos (Ursus arctos), lobos y zorros comunes (Vulpes vulpes, zorro en adelante) ya desde la Edad Media (Vargas 2002). Pero quizás

4

el mejor ejemplo de la sistematización de esta persecución sea la creación a mediados del siglo XX de las conocidas “Juntas provinciales de extinción de animales dañinos y protección a la caza” promovidas y financiadas por la administración pública. La finalidad de estas Juntas fue la erradicación de aquellas especies consideradas entonces como dañinas, entre las que se incluían carnívoros, rapaces y córvidos, para la que no existía ningún tipo de restricción en cuanto a los métodos utilizados (Vargas 2002; Corbelle-Rico y Rico-Boquete 2008). Esta persecución ejercida por el hombre ha contribuido al declive de algunas especies a lo largo del tiempo (Langley y Yalden 1977; Villafuerte et al. 1998; Ripple et al. 2014). En España la larga historia de persecución de depredadores contribuyó probablemente a la regresión y rarefacción de las poblaciones de muchas especies de depredadores, como el lobo (Valverde 1971; Blanco et al. 1992) o el lince ibérico (Lynx pardinus) (Rodríguez y Delibes 2002; 2004) e incluso de grandes rapaces necrófagas como el quebrantahuesos (Gypaetus barbatus), que desapareció por completo del sur de la Península Ibérica (Hiraldo et al. 1979). La percepción sobre parte de los depredadores comienza a cambiar entre mediados y finales del siglo XX, al menos en aquellas regiones del planeta más desarrolladas. Esto se debe en gran parte a al éxodo de personas del medio rural a las grandes urbes industrializadas y a el inicio de una conciencia social sobre la conservación de la biodiversidad (Conover et al. 2002), concepto que no será definido como tal hasta los años 80 (Kareiva y Marvier 2012). En relación a esta nueva conciencia social de conservación se crean nuevas medidas de protección para la fauna silvestre mediante diferentes leyes y normativas que incluyen la protección de un número importante de depredadores. Por ejemplo, en 1954 se promulgó en Reino Unido la Ley de Protección de las Aves, según la cual un gran número de rapaces pasaron a ser especies protegidas (Whitfield et al. 2003). Igualmente, en España este cambio de tendencia se ve reflejado a finales de los años 60 con la aprobación de la Orden General de Vedas de 1966, que prohíbe la caza de algunas especies consideradas nocivas hasta entonces como por ejemplo el lince ibérico. Pocos años después, la de la Ley de caza de 1970 regula y limita las especies que se pueden cazar así como las épocas y zonas para hacerlo (Vargas 2002).

5

Posteriormente, en la Convención de Washington de 1973 sobre Comercio Internacional de Especies Amenazadas de Flora y Fauna Silvestres (CITES) se redacta el primer catálogo internacional de especies protegidas frente a la explotación comercial, en el que se recogen un gran número de especies de depredadores. A pesar de la protección legal de muchos de estos depredadores, la persecución ilegal de gran parte de ellos ha continuado hasta nuestros días, como atestigua el reciente repunte del uso de cebos envenenados para controlar estas especies (Márquez et al. 2013; Martínez-Abraín et al. 2013). Aunque la percepción de los depredadores por la sociedad actual ha variado considerablemente en el último siglo (Martínez-Abraín et al. 2008), ésta sigue dependiendo de los intereses de diferentes grupos sociales o sectores. Así, la percepción y actitudes hacia los depredadores es diferente entre los grupos interesados en la conservación (p. ej. conservacionistas) y otros sectores con intereses productivos y de explotación de especies que son potenciales presas para los depredadores, como los ganaderos o cazadores (Treves y Bruskotter 2014). Algunos miembros de estos sectores siguen considerando hoy en día que los depredadores son perjudiciales porque consumen especies de cierto valor económico (Reynolds y Tapper, 1996; Graham et al. 2005). En ocasiones esto también ocurre porque se considera que los depredadores pueden ser peligrosos para el propio hombre (Packer et al. 2005; Goodrich et al. 2011). Hoy en día la problemática derivada de la actividad de los depredadores (es decir, los daños ocasionados por la depredación) es a menudo gestionada mediante el control letal de estos depredadores (en adelante, control de depredadores). Este se basa en la eliminación de individuos de la especie “problemática” con la intención de reducir la abundancia de sus poblaciones y disminuir de esta forma la presión de depredación sobre las presas. Esta medida de gestión es fuente de conflicto entre los diferentes sectores citados anteriormente, ya que su aplicación solo beneficia o satisface las pretensiones de una de las partes implicadas en el conflicto, lo cual dificulta la resolución de los mismos (Redpath et al. 2013).

Impacto de la depredación en las presas: depredadores especialistas y generalistas Como se ha mencionado en la sección anterior, existe la creencia relativamente extendida entre diferentes sectores de que los depredadores impactan negativamente 6

sobre las poblaciones de sus presas. Desde este punto de vista es importante conocer el impacto real de la depredación sobre las presas. Para ello, en primer lugar se deben de diferenciar los efectos de la depredación sobre el ganado, cuyas poblaciones están controladas por el hombre, de los efectos sobre las poblaciones de presas silvestres. En estas últimas, la dinámica poblacional está modulada por diferentes factores, tanto intrínsecos (p. ej. estado fisiológico, genética, comportamiento social, competencia intraespecífica) como extrínsecos (p. ej. hábitat, disponibilidad de alimento, climatología, parásitos, enfermedades y depredación), que a menudo interactúan entre sí (Sinclair y Pech 1996; Krebs 2002). Los depredadores son, por lo tanto, un factor más en la dinámica de las poblaciones de presas silvestres y sus efectos pueden ir desde la regulación (proceso por el que el depredador devuelve a la población de la presa a su densidad de equilibrio) hasta la limitación (proceso por el que el depredador establece la densidad de equilibrio de la presa) de las poblaciones de presas (Krebs 2002). El balance positivo o negativo de estos efectos sobre las presas depende en gran medida de la biología y abundancia de las presas, la abundancia del propio depredador/es, así como de la biología y la ecológica trófica de éste (Sinclair y Pech 1996; Sinclair et al. 2003). La teoría ecológica clasifica a las especies en dos grandes grupos, especialistas y generalistas, en función de la amplitud de nicho ecológico que presentan, definido este según varios ejes tanto bióticos como abióticos (p. ej. alimentación, hábitat, climatología, altitud, etc.) (Futuyma y Moreno 1988). Según esta teoría las especies especialistas presentarían una reducida amplitud de nicho ecológico en la cual sus poblaciones pueden conseguir un rendimiento ecológico óptimo, mientras que el nicho ecológico de las especies generalistas presenta una mayor amplitud. En el caso concreto de los depredadores, se distinguen depredadores generalistas, que tienen un amplio nicho trófico (alimentación variada), y depredadores especialistas, con un nicho trófico reducido (poca variedad de presas). No obstante, existen grupos ecológicos intermedios, como los denominados depredadores especialistas facultativos, que pueden adaptar su estrategia a las condiciones dominantes, cambiando su presa principal cuando otras presas más rentables están disponibles (Glasser 1982). Debido a su reducida amplitud trófica, los depredadores especialistas presentan cambios en el tamaño poblacional asociados a la densidad de su principal presa (i.e. respuesta numérica). Por ello no suelen representar un riesgo para las poblaciones de sus presas 7

(Begon et al. 1996), aunque existen algunas excepciones (ver p. ej. Hanski et al 1991). Estas características les permiten un desarrollo óptimo en condiciones ambientales estables y homogéneas, pero sin embargo, les limita considerablemente su capacidad de respuesta ante cambios ambientales. Por el contrario los depredadores generalistas presentan una serie de características biológicas que les confieren una gran flexibilidad ecológica (Begon et al. 1996). Se alimentan de varios tipos de presas en función de su abundancia, cambiando la tasa de depredación sobre su presa principal ante la variación de la densidad de la misma (i.e. respuesta funcional). Dicho de otro modo, pueden adaptarse a alimentarse de presas secundarias cuando su principal presa disminuye de abundancia. Los depredadores generalistas suelen presentar altas tasas de reproducción por lo que sus poblaciones pueden llegar a ser abundantes. El incremento en la abundancia de los depredadores generalistas puede provocar un notable impacto negativo para algunas poblaciones de presas simplemente por el aumento en el riesgo de depredación, es decir aumento de la depredación incidental (Thirgood et al. 2000b; Valkama et al. 2005; Prugh et al. 2009; Eagan et al. 2011; Ripple et al. 2013). Altas densidades de este tipo de depredadores pueden reducir e incluso extinguir las poblaciones de ciertas presas, provocando importantes desajustes en la estructura y estabilidad de las comunidades en las que se encuentran (Prugh et al. 2009).

Factores que favorecen a los depredadores generalistas Actualmente gran parte de los sistemas naturales han sido fuertemente modificados por la mano del hombre (Sanderson et al. 2002), lo que parece haber beneficiado a muchos depredadores generalistas. Esto se debe principalmente al efecto combinado de la rarefacción de depredadores apicales (del inglés top predators, depredadores claves en la regulación de los procesos ecológicos de las comunidades de los que forman parte; Sergio et al. 2008), a la modificación y fragmentación de hábitats y al incremento de recursos alimentarios derivados de la actividad humana (Prugh et al. 2009). Durante el pasado siglo las poblaciones de muchas especies de depredadores apicales se han visto reducidas a escala mundial, debido principalmente a la persecución humana y a la modificación y pérdida de sus hábitats o el de sus principales presas. Tal ha sido el 8

caso de grandes carnívoros como osos, lobos y grandes felinos en todo el mundo (Ripple et al. 2014) y grandes rapaces como el águila real (Aquila chrysaetos) (Whitfield et al. 2007) o el búho real (Bubo bubo) en algunas zonas de Europa (Penteriani y Delgado 2010). En la Península Ibérica también existen dos casos muy reconocidos, el lince ibérico y el águila imperial ibérica (Aquila adalberti) (Rodríguez y Delibes 2002; 2004; González et al. 2008). Los depredadores apicales, muchos de ellos considerados como especialistas, actúan como especies clave en los ecosistemas limitando las poblaciones de otros depredadores menores, ya sea por depredación directa o por exclusión competitiva (Palomares y Caro 1999; Sergio e Hiraldo 2008). De esta manera, la presencia de depredadores apicales puede resultar beneficiosa para sus presas al disminuir la tasa de depredación por depredadores de tamaño medio (los llamados “mesodepredadores”) (Palomares et al. 1995; Sergio e Hiraldo 2008). Ante este escenario de ausencia de depredadores apicales, los mesodepredadores a menudo generalistas, pueden beneficiarse aumentando su abundancia y rango de distribución según la denominada Hipótesis de “liberación de mesodepredadores” (del inglés Mesopredator Release Hypothesis; Crooks y Soulé 1999). Numerosos estudios han encontrado evidencias por todo el mundo que confirman esta hipótesis (Prugh et al. 2009; Ritchie y Johnson 2009). Algunos ejemplos son el aumento de coyotes en Norteamérica tras la regresión de la poblaciones de lobos (Ripple et al. 2013), la limitación de las poblaciones de zorro por el lince boreal (Lynx lynx) en Suecia (Helldin et al. 2006), o la limitación de meloncillos (Herpestes ichneumon) por el lince ibérico en España (Palomares et al 1995). De esta forma la regresión de las poblaciones de lince ibérico en el siglo pasado (Rodríguez y Delibes 2003) probablemente haya contribuido al aumento de la abundancia y distribución de algunos carnívoros generalistas de tamaño medio como el observado recientemente para el meloncillo (Recio y Virgós 2010). Existen también ejemplos en aves como el descrito en Alemania para el búho real y dos rapaces de tamaño medio como el azor (Accipiter gentilis) y el busardo ratonero (Buteo buteo) (Chacarov y Krüger 2010). La fragmentación y degradación de hábitats debido al creciente desarrollo de diferentes actividades humanas (p. ej. agricultura, explotación maderera, infraestructuras, etc.) han sido reconocidas entre los factores con mayor impacto sobre la biodiversidad (Sala et al. 2000). Aparte de la anteriormente citada disminución de depredadores apicales, la transformación de algunos hábitats ha facilitado el incremento de recursos alimentarios 9

para muchos depredadores generalistas, como por ejemplo diferentes especies de roedores (Thirgood et al. 2000b; Šálek et al. 2010; Luque-Larena et al. 2013). Thirgood y colaboradores (2000b), por ejemplo, mostraron cómo en Escocia los aguiluchos pálidos (Circus cyaneus) se beneficiaron del incremento en la abundancia de pequeños roedores debido al aclarado de los brezales por el pastoreo, lo que supuso el incremento de depredación incidental sobre el lagópodo escocés (Lagopus lagopus scoticus). La actividad agrícola también puede incrementar la abundancia de ciertas presas consumidas habitualmente por numerosos depredadores generalistas, como son algunos micromamíferos (Luque-Larena et al. 2013) o el caso de algunos invertebrados y pequeñas aves, asociados a los linderos entre cultivos (Vickery et al. 2002). Igualmente los productos derivados de los cultivos agrícolas también pueden beneficiar a ciertos depredadores generalistas omnívoros que incluyen de forma frecuente alimentos como frutos y semillas en su alimentación. Este es el caso de algunos carnívoros de tamaño medio y algunos córvidos (Soler et al. 1993; Rosalino y Santos-Reis 2009). Por otro lado en ambientes fuertemente antropizados la actividad humana genera un importante volumen de desperdicios (p. ej. basureros, merenderos, restos de granjas, etc) que son fuente de alimentación suplementaria para muchos de estos depredadores generalistas. De esta forma se ha observado cómo los zorros que habitan las periferias de pueblos en entornos rurales, o incluso en las grandes ciudades, incluyen en su dieta una importante proporción de alimentos de origen antrópico como basura o carroña de ganado (Contesse et al. 2004; Webbon et al. 2006). Esta fuente de alimentación puede suponer un aumento de la supervivencia y, por tanto, de la abundancia de las poblaciones de zorros en estos ambientes (Bino et al. 2010). De forma similar algunos córvidos también pueden verse beneficiados por estas fuentes de alimentación antrópicas. Por ejemplo, se ha observado cómo la reducción de alimento subsidiario tras el cierre de varias piscifactorías provocó una reducción de la densidad de nidos de urraca (Pica pica) en una región de Norteamérica (Stone y Trost 1991). Más recientemente se ha señalado que una alta disponibilidad de alimento de origen antrópico puede favorecer la reproducción, supervivencia de adultos y abundancia local de diferentes especies de córvidos (Marzluff y Neatherlin 2006).

10

Ecología del zorro y la urraca: paradigma de especies generalistas El zorro y la urraca representan el paradigma de especies generalistas debido a su gran flexibilidad ecológica en cuanto a requerimientos de hábitat, alimentación, parámetros reproductivos y capacidad de adaptación a los cambios en el medio, como los producidos por la actividad humana. Por todo ello, pueden llegar a alcanzar elevadas abundancias (Birkhead 1991; Sillero-Zubiri et al. 2004). El zorro es el carnívoro de tamaño medio más abundante y ampliamente distribuido en todo el mundo. Especie de distribución holártica, se encuentra en grandes áreas del Paleártico, incluida la Península Ibérica (Blanco 1998; Sillero-Zubiri et al. 2004). No presenta requerimientos específicos de hábitat, estando presente tanto en ambientes naturales como en ambientes fuertemente antropizados e incluso en el centro de grandes ciudades (Contesse et al. 2004; Sillero-Zubiri et al. 2004; Webbon et al. 2006). Se considera un depredador oportunista y omnívoro, que incluye en su dieta alimentos vegetales, animales y desperdicios de origen antrópico (Díaz-Ruiz et al. 2013). Presenta respuestas funcionales ante la disminución en la disponibilidad de su principal fuente de alimento en cada situación, adaptándose al consumo de otros alimentos secundarios (Ferreras et al. 2011). Se trata de una especie monoestra, es decir, que solo tiene un ciclo reproductor al año (Voigt y Macdonald 1984), presentando una alta tasa de reproducción, con tamaños de camada variables en función de los recursos disponibles, que oscilan entre 1 y 12 cachorros (López-Martín 2010). El zorro dispone de mecanismos de reproducción compensatoria, aumentando su productividad en situaciones de alta mortalidad (Heydon y Reynolds 2000). Por lo general, una parte importante de su población está compuesta por individuos no reproductores sin territorios definidos, por lo que el proceso de recolonización de territorios vacíos puede ser rápido cuando hay una mortalidad alta de adultos territoriales (Reynolds et al. 1993; Cavallini 1996). La urraca también es una especie ampliamente distribuida y abundante en muchas zonas de Asia, el oeste de Norteamérica y Europa, incluida la Península Ibérica (Birkhead 1991; Martínez 2011). Aunque se encuentra en diferentes tipos de hábitats, que van desde áreas naturales a zonas urbanas, suele alcanzar las mayores densidades en ambientes agrícolas humanizados (Martínez 2011). La urraca es un generalista omnívoro en cuanto a sus hábitos alimentarios que consume un amplio espectro de

11

alimentos de origen vegetal y animal, pudiendo beneficiarse a su vez de recursos alimenticos de origen antrópico (Birkhead 1991). A diferencia del zorro, su papel como depredador de aves, tanto protegidas como cinegéticas, no está claro, aunque algunos trabajos indican que puede consumir huevos, pollos e incluso adultos de algunas especies de estas aves (Groom 1993; Herranz 2000; Fernández-Juricic et al. 2004; Roos y Pärt 2004). Presentan un solo ciclo reproductor al año, y una alta tasa de reproducción, con puestas de entre 4 y 10 huevos (Birkhead 1991; Martínez 2011). En caso de pérdida de la puesta pueden efectuar una puesta de sustitución como mecanismo de compensación de pérdidas en la población (Pónz y Gil-Delgado 2004). Una parte importante de sus poblaciones está formada por individuos no reproductores que pueden reemplazar rápidamente cualquier pérdida de algún miembro de las parejas reproductoras, y completar de forma exitosa la reproducción (Birkhead 1991). Tanto zorro como urraca están considerados por algunos sectores de la sociedad como especies perjudiciales para diferentes intereses humanos, como la agricultura, la actividad cinegética o la ganadería, en prácticamente todo su rango de distribución; lo cual hace que a menudo sean objeto de control (Birkhead 1991; Sillero-Zubiri et al. 2004).

El control de depredadores como herramienta de gestión y conservación Actualmente los impactos de los depredadores sobre algunos intereses humanos se gestionan de forma diferente según el grado de protección de los mismos. De esta forma, los impactos o daños generados por especies amenazadas suelen gestionarse a través de compensaciones y subvenciones a los afectados o mediante translocaciones de individuos, evitando la eliminación legal de los depredadores amenazados. Este tipo de gestión está normalmente asociado a los daños producidos al ganado por grandes depredadores como los lobos (pagos de indemnizaciones; aunque en algunas zonas también se autoriza su caza), o que pueden afectar a la integridad física de las personas como es el caso de los grandes felinos en algunas zonas (translocaciones de individuos conflictivos) (Boitiani et al. 2010; Goodrich et al. 2011; Treves y Bruskotter, 2014). En cambio, el control letal de depredadores es una medida habitual de gestión de la depredación causada por depredadores generalistas abundantes (Treves y NaughtonTreves 2005). Se utiliza como herramienta de gestión en la conservación de ecosistemas y especies amenazadas, como medida sanitaria para el control de zoonosis, como 12

protección del ganado o en la gestión cinegética (Prught et al. 2009; Beja et al. 2009; Saunders et al. 2010; Baesley et al. 2013). El control de depredadores introducidos, por ejemplo, es una herramienta utilizada a menudo en acciones de conservación en zonas donde estas especies han causado un gran impacto ecológico o pueden llegar a hacerlo. Un claro ejemplo es el control de las poblaciones de zorro en Australia, donde el cánido ha contribuido a la extinción de varias especies de vertebrados autóctonos, representando un grave problema para la conservación de la fauna nativa (Saunders et al. 2010). Igualmente el control de gatos domésticos asilvestrados (Felis catus) es una acción de gestión habitual para la recuperación de fauna en numerosas islas de todo el mundo, ya que su depredación ha contribuido al declive poblacional e incluso extinción de numerosas especies (Medina et al. 2011). La eliminación de algunos depredadores generalistas que actúan como reservorios de enfermedades ha sido una herramienta empleada para el control y erradicación de algunas zoonosis. Algunos ejemplos son el control poblacional de tejones (Meles meles) empleado en Reino Unido para minimizar el riesgo de trasmisión de la tuberculosis bovina (Smith et al. 2001; Bielby et al. 2014), el control de las poblaciones de zorros para limitar el avance de la rabia en gran parte de Europa (Holmala y Kauhala 2006) o el control de mapaches (Procyon lotor) en Norteamérica por ser reservorio de estas y otras enfermedades infecciosas (Baesley et al. 2013). Sin embargo, el control de depredadores generalistas por motivos de conservación y sanidad, solo se realiza en casos excepcionales y bajo un estricto seguimiento por parte de la administración. Por el contrario, el control de depredadores generalistas con fines cinegéticos es una medida ampliamente extendida en diferentes zonas de todo el mundo (Reynolds y Tapper 1996) debido a que los cazadores lo consideran con frecuencia fundamental para aumentar la abundancia de las especies cinegéticas (Delibes-Mateos et al. 2013; Ljung et al. 2014). Aunque en algunas zonas se controlan grandes depredadores para fomentar especies de caza mayor (Musiani y Paquet 2004), el control orientado a depredadores de pequeña o mediana talla, para el fomento de especies de caza menor, es probablemente mucho más común y extendido. En Reino Unido, por ejemplo, es muy común el control de zorros, tejones, pequeños mustélidos y córvidos como la urraca y la corneja negra (Corvus corone) para fomentar las poblaciones de aves cinegéticas como la perdiz gris (Perdix perdix) o los lagópodos (Tapper et al. 1996, Thirgood et al. 2000a). En Francia el trampeo de pequeños y medianos carnívoros como zorros, garduñas (Martes foina) y 13

martas (Martes martes) es una práctica habitual (Ruette et al. 2003). También se controla la urraca de forma sistemática en gran parte del país al considerarse una especie dañina para la caza (Chiron et al. 2013). Igualmente en Suecia el control de zorros, tejones y urracas es una medida muy extendida para fomentar las poblaciones de varias especies de caza menor como los lagópodos y las liebres (Lepus sp.) (Ljung et al. 2014). En Portugal el control legal de zorros, meloncillos y urracas es una medida muy empleada para fomentar las poblaciones de perdiz roja (Alectoris rufa), conejo de monte (Oryctolagus cuniculus) y liebre ibérica (Lepus granatensis) (Beja et al. 2009).

Efectos derivados del control de depredadores Los diferentes efectos derivados del control de depredadores son uno de los principales puntos de controversia que genera esta actividad. Esto es debido en parte a la falta de conocimiento científico, pero también a que los resultados obtenidos en los trabajos que han estudiado estos efectos son a menudo contrapuestos o poco concluyentes. Como se ha indicado anteriormente, con el control de depredadores se pretende un efecto beneficioso sobre las presas que se quieren fomentar. Sin embargo, por lo general no se consideran los efectos sobre especies no relacionadas directamente con el control. En este sentido podríamos agrupar los efectos derivados del control de depredadores en tres categorías: 1) Efecto sobre las presas que se pretenden fomentar, 2) Efecto sobre los depredadores objeto del control y 3) Efecto sobre otras especies que no son objeto de control. Efecto sobre las presas Existe gran controversia en cuanto a la efectividad del control de depredadores para fomentar las poblaciones de ciertas presas. Por un lado, diversos trabajos no encuentran un efecto significativo del control de depredadores sobre el incremento de las presas (Kauhala et al. 2000; Keedwell et al. 2002). Por ejemplo, el control de múltiples depredadores durante 20 años en una zona de Nueva Zelanda provocó cierto efecto positivo a corto plazo en las poblaciones de kaki (Himantopus novaezelandiae), un ave amenazada, pero dicho efecto desapareció posteriormente pese a mantener el control (Keedwell et al. 2002). Por el contrario, varias revisiones indican que el control de depredadores puede producir mejoras en las poblaciones de presas bajo ciertas condiciones (Holt et al. 2008; Salo et al. 2010; Smith et al. 2010). Estas revisiones coinciden en señalar que la eficacia del control depende de varios factores como la

14

duración e intensidad de las extracciones, el número de especies de depredadores controlados, el tipo de depredador (autóctono o exótico), el tipo de presa que se intenta recuperar, etc. Estos trabajos también señalan la importancia de los métodos de seguimiento de las poblaciones de presas como algo fundamental para poder determinar los efectos del control de sus depredadores. Efecto sobre los depredadores generalistas objeto de control La mayor parte de trabajos científicos existentes sobre control de depredadores evalúan el efecto que éste tiene sobre las poblaciones de presas que se pretende fomentar (ver apartado anterior), mientras que pocos evalúan el efecto sobre las poblaciones de la especie objeto del control. Normalmente se asume que la extracción de un número de animales conlleva una reducción del tamaño de la población. Sin embargo, no siempre es así debido a que algunas especies que se pretenden controlar, como los depredadores generalistas, presentan mecanismos para compensar reducciones en sus poblaciones. El zorro y la urraca son un claro ejemplo en ese sentido, como se ha señalado anteriormente. Una parte importante de las poblaciones de zorro y urraca está constituida por individuos no reproductores que contribuyen a la rápida respuesta demográfica frente a actuaciones de control (Birkhead 1991; Cavallini 1996). Se ha descrito que una eliminación de individuos adultos territoriales, sin reducir la disponibilidad de alimento, va seguida de la ocupación de los territorios vacíos por individuos flotantes (Reynolds et al. 1993; Chiron y Juliard 2013). Además de la rápida ocupación de territorios, las poblaciones de estos depredadores pueden responder a la extracción con mecanismos de reproducción compensatoria, aumentando la productividad (Heydon y Reynolds 2000) o haciendo puestas de reposición (Pónz y Gil-Delgado 2004). Varios trabajos han puesto de manifiesto la dificultad de reducir las poblaciones de zorro, incluso empleando métodos de control masivos como cebos envenenados específicos (Saunders et al. 2010). A menudo el control sólo es eficaz a corto plazo (Harding et al. 2001), y en algunos casos ineficaz para reducir las densidades (Baker y Harris 2006). Por el contrario, en un estudio observacional realizado a gran escala en Inglaterra se comprobó que el control de zorros mediante distintos métodos puede reducir sustancialmente la abundancia de este carnívoro en un amplio rango de circunstancias (Heydon y Reynolds 2000). En cualquier caso, la evaluación 15

experimental de la efectividad de los métodos de captura de zorros para reducir sus poblaciones es complicada, debido en parte a la dificultad de realizar estimas fiables de su abundancia. Estas suelen requerir en el caso de los carnívoros metodologías costosas y sofisticadas (Heydon et al. 2000; Schauster et al. 2002). La efectividad del control de depredadores para reducir la densidad de urracas ha sido menos estudiada que en el caso del zorro. No obstante, diferentes trabajos encuentran como el control de urracas puede ser efectivo en la reducción de sus poblaciones a escala local y regional (Stoate y Szuczur 2001, 2005; Chiron y Julliard 2007). Recientemente se ha descrito cómo el control intensivo de urracas continuado en el espacio y en el tiempo propiciaba el descenso de las poblaciones así como la desestructuración de la población reproductora, que estaba dominada por individuos jóvenes en zonas donde el control era más intensivo (Chiron y Julliard 2013). Aparte de los efectos sobre la abundancia y la dinámica poblacional de la especie controlada, las extracciones realizadas mediante el control de depredadores también puede tener efectos a nivel comportamental cuando este es una importante causa de mortalidad para la especie. En Australia, por ejemplo, se ha observado como los dingos modifican sus ritmos de actividad diarios de acuerdo a si sus poblaciones son o no controladas; son más nocturnos en zonas con que en zonas sin control (Brook et al. 2012). Efecto sobre especies que no son objeto de control El control de depredadores puede tener efectos negativos sobre otras especies que no son objeto del control, tanto cuando el control es selectivo, es decir, solo se extrae la especie objeto de control, como cuando no lo es, extrayéndose también otras especies. La hipótesis de la liberación de competidores (del inglés “Competitor Release Hypothesis”) propone como la eliminación de una especie dominante dentro de una comunidad puede ser aprovechada por otra especie subordinada que, ante la falta de su competidor, incrementa su abundancia (Caut et al. 2007). Aunque esta hipótesis se basa en una aproximación teórica realizada para una comunidad de roedores sometida a control, este efecto puede darse también en las comunidades de depredadores como por ejemplo en los mesocarnívoros (Barrull et al. 2014). Cuando el control es selectivo se puede producir un aumento de otros depredadores subordinados. En Reino Unido, por ejemplo, se observó cómo, tras el control selectivo de un depredador dominante como el 16

tejón, realizado para frenar la expansión de la tuberculosis, la abundancia de zorros (competidor subordinado) incrementó (Trewby et al. 2008). Sin embargo, el control de depredadores desarrollado en algunas fincas de caza no es selectivo y se eliminan ilegalmente especies de mesocarnívoros, que a priori no son objeto de control (Duarte y Vargas 2001; Barrull et al. 2011). Estudios recientes basados en modelos teóricos de simulación han puesto de manifiesto que diferentes niveles de control no selectivo de las poblaciones de zorros podrían alterar las comunidades de carnívoros con un aumento en la abundancia de la especie objetivo, es decir el zorro. Por el contrario, las poblaciones de otras especies no objetivo (competidores del zorro) como el tejón, la garduña y la marta (Martes martes) podrían reducirse notablemente o incluso desaparecer debido a las menores tasas reproductivas de estas especies (Casanovas et al. 2012; Lozano et al. 2013). Pero el control de depredadores no solo puede tener efectos sobre otros depredadores que a priori no son objeto de control sino que puede afectar de forma indirecta a otras especies no relacionadas directamente con el control. El control intensivo de depredadores puede perjudicar a la diversidad y estructuración de algunos grupos de presas secundarias, como se ha comprobado en Norteamérica para el control de coyotes y las comunidades de micromamíferos (Henke y Bryant 1999). En dicho estudio se observó que en zonas de baja abundancia de coyote debido a su intenso control, las comunidades de roedores eran menos diversas y estaban dominadas por pocas especies que se libraron de la depredación de los coyotes, y desplazaron por competición a otras especies de la comunidad. Otro ejemplo del posible efecto indirecto del control de depredadores sobre otras especies sería el del control de urracas y el críalo (Clamator glandarius). El críalo es un ave parásita de los nidos de urraca que en gran medida depende de ésta para completar su ciclo reproductor (Martínez 2011; Soler 2012). Por lo tanto, tanto el críalo como otras aves que utilizan para criar los nidos abandonados de urraca, podrían verse perjudicados cuando éstos son destruidos como medida de control (Birkhead 1991). Además, la extracción intensa de estos depredadores generalistas puede tener efectos sobre diferentes procesos ecológicos en los que estas especies desempeñan diferentes funciones. Por ejemplo el zorro es un importante dispersor de semillas de ciertas plantas y también puede regular las poblaciones de ciertas presas consideradas como plaga por 17

el hombre (Hanski et al. 1991; Fedriani y Delibes 2009). Igualmente la urraca juega un papel de control biológico sobre ciertos grupos de invertebrados potencialmente perjudiciales para los cultivos (Birkhead 1991).

El control de depredadores en España En España el control de depredadores es una práctica bastante extendida que se usa tanto como parte de la gestión cinegética como para la conservación de ecosistemas y especies amenazadas. En relación al segundo de los casos, existen varios ejemplos de control de depredadores introducidos, como el visón americano (Neovison vison) por su impacto sobre diferentes presas así como por ser competidor del autóctono y amenazado visón europeo (Mustela lutreola) (Zuberogoitia et al. 2010). Más reciente es el control de mapaches, el cual ha colonizado varias zonas de España a partir de las liberaciones de particulares, y tiene un gran potencial como depredador, como competidor de otros depredadores autóctonos y como reservorio de enfermedades (García et al. 2012). Aparte del control de depredadores exóticos, en España también se han controlado depredadores autóctonos como medida para la conservación de especies amenazadas. Por ejemplo, en los Pirineos se han controlado zorros y translocado otros mesocarnívoros generalistas para la protección del urogallo (Tetrao urogallus) (Fernández-Olalla 2011). Sin embargo, el control de depredadores generalistas por motivos de conservación se realiza en España de forma puntual, en casos excepcionales y bajo un estricto seguimiento por parte de la administración, siendo el control ligado a la gestión cinegética mucho más común y extendido a lo largo de gran parte del país. La caza menor es un recurso económico importante en muchas áreas rurales de España (Bernabeu, 2000). Las principales especies de caza menor son la perdiz roja, el conejo de monte y la liebre. En las últimas décadas la abundancia de las poblaciones silvestres de estas especies ha sufrido una importante disminución en gran parte de la Península Ibérica, siendo más acusada en la perdiz roja y el conejo (Blanco-Aguiar et al. 2003; Blanco-Aguiar 2007; Delibes-Mateos et al. 2009). Esto parece haber provocado un incremento en el uso de métodos para el control de depredadores (tanto legales como el ilegales) con la intención de recuperar estas especies (Villafuerte et al. 1998; Márquez et al. 2012). Al igual que lo descrito anteriormente, en España el control de depredadores es una medida muy extendida en gran parte de los cotos de caza, principalmente de caza

18

menor, como así lo confirman varios estudios (Tabla 1). El zorro y la urraca son las especies en las que se suele centrar este control (Tabla 1) (Díaz-Ruiz y Ferreras 2013). Tabla 1. Trabajos que han estudiado la extensión del uso del control de depredadores en España como herramienta de gestión cinegética. N es el tamaño muestral de cada trabajo.

a

porcentaje de provincias en las que el control se realiza con una intensidad media-alta; el resto de los encuestados reconoció un baja intensidad en el control de depredadores. b Información no disponible. c porcentaje de los cotos que realizan control en los que se realiza sobre cada especie o grupo de especies

Referencia

Zona de estudio

Tipo de

Datos

Áreas

N

Control Zorro Córvidos

95%Cotos de caza Angulo 2003

Andalucía

Entrevistas

menor-

personales con los

mayor

307

48%

-b

-b

47

66%a

-b

-b

60

70%

95%c

5% c

5365

94.4%

82% c

56% c

59

90%

85% c

80% c

gestores de los cotos 5% áreas protegidas

Piorno 2006

Delibes-Mateos 2008

Encuestas a técnicos

Cotos de

España

de caza de las

caza

Peninsular

Administraciones

menor-

Provinciales

mayor Cotos de

Entrevistas Centro-Sur

caza

personales con Cazadores-Gestores

menormayor Cotos de

Rios-Saldaña 2010

Castilla-La

Planes técnicos de

caza

Mancha

caza

menormayor

Delibes-Mateos et al. 2013

Entrevistas Centro

personales con los gestores de los cotos

19

Cotos de caza menor

A pesar de tratarse de una actividad legal y regulada, el control de depredadores, y especialmente el desarrollado en la gestión cinegética, es una actividad que genera gran controversia en la sociedad española con posicionamientos opuestos entre diferentes grupos sociales: ecologistas, conservacionistas, científicos, administración, cazadores y ganaderos (Herranz 2000; Lozano et al. 2006; Virgós et al. 2010). Esto es debido, al menos en parte, a la poca información disponible sobre diferentes aspectos relacionados con esta actividad, como son la idoneidad de los métodos de control empleados así como los efectos derivados del control de depredadores (Díaz-Ruiz y Ferreras 2013). Regulación legal del control de depredadores Actualmente el control de depredadores en España está regulado por cuatro ordenamientos: el internacional, el comunitario, el estatal y el autonómico, a través de diferentes normativas (Tabla 2). La mayor parte de estas normativas se refieren a los métodos de control, prohibiendo de forma general aquellos masivos y/o no selectivos, e incluyen anexos donde se enumeran los diferentes métodos que quedan completamente prohibidos, como por ejemplo el uso de cebos envenenados (p. ej. Convenio de Berna 1979) o el de cepos (Reglamento (CEE) nº 3254/91 de 1991). Estas normativas coinciden en dejar una vía de excepción a la norma general, merced a la cual se pueden autorizar determinados métodos bajo unos supuestos que justifiquen su uso (entre ellos daños a la fauna). Las diferentes normativas autonómicas vigentes en España son las que establecen las especies que pueden ser objeto de control (Gálvez 2004). Por lo general solamente se permite el control de ciertos depredadores generalistas, que en su mayoría están catalogados como especies cinegéticas. En concreto, y salvo algunas excepciones según cada región, se permite controlar cuatro especies silvestres: el zorro, la urraca, la grajilla (Corvus monedula) y la corneja negra. También se suele permitir de forma excepcional el control de otras dos especies de depredadores domésticos asilvestrados: el gato y el perro (Canis lupus familiaris). Generalmente los depredadores cinegéticos pueden ser cazados con armas de fuego durante la época hábil de caza. Además, se permite el uso excepcional de otros métodos de captura fuera de la temporada cinegética para controlar tanto estas dos especies domésticas como las cinegéticas. Los permisos de control excepcional son concedidos por la administración regional según diferentes criterios, que no siempre son los establecidos en estas normativas (Bernard 2008).

20

Tabla 2. Normativas vigentes en España sobre control de depredadores. Nivel Legislativo

Normativas vigentes - Convención sobre la conservación de la vida silvestre y el medio natural de Europa (“Convenio de Berna”. Berna, 19-IX-1979) - Acuerdo entre la Unión Europea, Canadá y la Federación Rusa sobre

Internacional

métodos de captura no cruel (Decisión 98/142/CE del Consejo de 26 de Enero de 1998) - Acuerdo ente la Unión Europea y los Estados Unidos de América sobre métodos de captura no cruel (Decisión 98/487/CE de 13 de Julio de 1998) - Directiva 79/409/CEE, relativa a la conservación de las aves silvestres (“Directiva de Aves”). - Directiva 92/43/CEE relativa a la conservación de los Hábitats naturales y de la fauna y flora silvestres (“Directiva Hábitats”). - Reglamento (CEE) nº 3254/91 del Consejo, de 4 de noviembre de 1991,

Unión Europea por el que se prohíbe el uso de cepos en la Comunidad - Reglamento (CE) nº 1771/94 de la Comisión, de 19 de julio de 1994, sobre comercialización de pieles de animales salvajes - Reglamento (CE) nº 35/97 de la Comisión de 10 de enero de 1997, sobre la certificación de pieles - 97/602/CE: Decisión del Consejo de 22 de julio de 1997 - Ley 42/2007 de Conservación del Patrimonio Natural y de la Biodiversidad. Título III. Capítulo IV – De la protección de las especies en relación con la caza y la pesca continental Estatal

- Directrices técnicas para la captura de especies cinegéticas predadoras: homologación de métodos de captura y acreditación de usuarios. Aprobadas por la Conferencia Sectorial de Medio Ambiente. 13 de julio de 2011

Autonómica

- Leyes y Reglamentos Autonómicos de Ordenación de la Caza

Métodos de control de depredadores generalistas Uno de los principales motivos de controversia en relación al control de depredadores es la efectividad y selectividad de los métodos utilizados. Por lo general los cazadores consideran que los métodos permitidos por la legislación vigente son pocos eficaces 21

para controlar a los depredadores (Delibes-Mateos et al. 2013). En los últimos años algunas comunidades autónomas han iniciado el proceso legal de homologación de determinados métodos de control de depredadores generalistas basándose en ensayos de campo, en sendos acuerdos internacionales sobre métodos de captura no cruel (Ver Tabla 2), y en una Norma ISO (International Organization for Standardization 1999) sobre evaluación de métodos de captura y retención de mamíferos (Díaz-Ruiz y Ferreras 2013). Como se ha señalado anteriormente, el zorro y la urraca son las dos principales especies en las que se centra el control de depredadores en España. En este sentido, actualmente los principales métodos de captura utilizados con carácter excepcional para el control poblacional de estas especies son los lazos y jaulas-trampa para la captura de zorros y jaulas-trampa para la captura de urracas (Delibes-Mateos et al. 2013). Diferentes trabajos han evaluado de forma empírica la eficiencia de captura de las especies objetivo, la selectividad y los daños relacionados con la captura de varios de estos métodos utilizados habitualmente para controlar zorros y urracas en España (Díaz-Ruiz et al. 2013). Recientemente se ha aprobado un documento, consensuado entre las administraciones central y autonómicas, que recoge las directrices para establecer qué métodos pueden homologarse para realizar control de depredadores (Conferencia Sectorial de Medio Ambiente 2011). Sin embargo, la citada Norma ISO y su interpretación han suscitado controversia y críticas entre científicos que la consideran insuficiente e incluso errónea en algunos de sus planteamientos, tanto en lo relativo a bienestar animal como en algunos conceptos aplicados a los dispositivos de captura (Iossa et al. 2007; Virgós et al. 2010). Métodos para el control de zorros

Las jaulas-trampa para zorros consisten en un compartimento de captura con una o dos puertas de entrada, que se cierran mediante un balancín al ser pisado por el animal, y un compartimento opcional para el cebo (Fig. 1). Pueden utilizarse con cebo vivo o muerto (Ferreras et al. 2003; 2007; Muñoz-Igualada et al. 2008). Tanto los lazos tradicionales actuales como dos versiones norteamericanas más complejas (“Lazo Americano” y “Lazo Wisconsin”) consisten en un cable de acero en el que en uno de sus extremos presenta un lazo corredizo con un tope (salvo en el modelo “sin tope”) para que este no 22

se cierre totalmente sobre el cuello del animal, fijándose el otro extremo al terreno para retener al animal capturado (Muñoz-Igualada et al. 2010). En España se han evaluado también dos nuevos sistemas diseñados en Estados Unidos para la captura de cánidos, las trampas Belisle y Collarum (Shivik et al. 2000). Las trampas Belisle (Edouard Belisle, Saint Veronique, PQ, Canadá) consisten en un lazo de acero propulsado que retiene al animal por la extremidad al accionar una pletina central de disparo (Shivik et al. 2000; Muñoz-Igualada et al. 2008). La trampa Collarum (Wildlife Control Supplies, East Granby, CT, USA) es también un lazo de acero propulsado que retiene al animal por el cuello (Shivik et al. 2000; Ferreras et al. 2007; Muñoz-Igualada et al. 2008). En este último caso, el sistema de disparo precisa de una respuesta activa del animal ante un atrayente oloroso. Ambos lazos propulsados se instalan enterrados, quedando tan sólo visible en la superficie, en el caso del Collarum, el disparador con el atrayente (Ferreras et al. 2007; Muñoz-Igualada et al. 2008). Métodos para el control de urracas

Las jaulas-trampa para capturar urracas son el método más empleado para controlar urracas en España ya que los cazadores las consideran eficaces para reducir las abundancias del córvido (Delibes-Mateos et al. 2013). Por lo general estas trampas tienen un compartimento central donde se coloca una urraca viva que actúa como reclamo y una serie de compartimentos de captura (2 o 4) alrededor que se accionan de forma independiente (Ferreras et al. 2007). Efectos del control de depredadores en España En España existen pocos trabajos que hayan estudiado los diferentes efectos del control de depredadores. De esta forma la efectividad del control de depredadores para fomentar las presas en España está poco clara. El único trabajo experimental de este tipo realizado en España evaluó la efectividad del control selectivo de depredadores (zorro y urraca) para mejorar la supervivencia de la perdiz roja (Mateo-Moriones et al. 2012). El control de depredadores mejoró la supervivencia de los pollos, especialmente de aquéllos de más de un mes de edad, pero no mejoró la supervivencia de los adultos ni de los nidos, ni el tamaño de las poblaciones de perdiz. Herranz (2000) describe resultados similares referidos al control de urracas en un coto de caza de Castilla-La Mancha, donde tras el control se incrementó el tamaño de bando de las perdices pero no se consiguió incrementar sus poblaciones ni las de paloma torcaz (Columba palumbus). Del mismo modo, en un trabajo reciente realizado en el centro de España no se encontró 23

ninguna relación entre la intensidad de control de zorros y las densidades de perdiz roja (Díaz-Fernández et al. 2013). Por el contrario, Delibes-Mateos et al. (2008c) hallaron que el control de depredadores y el manejo de hábitat fueron las dos únicas medidas de gestión relacionadas con la tasa de cambio en la abundancia de conejo en cotos de caza del centro-sur de España entre 1993 y 2002. Igualmente, Virgós y Travaini (2005) observaron mayores abundancias de conejo en cotos de caza con gestión cinegética intensiva que en zonas donde no se realizaba este tipo de gestión. Varios estudios han evaluado también la efectividad del control de depredadores para incrementar especies de interés para la conservación en España. Por ejemplo, en un experimento realizado en el Pirineo el control de zorro y las translocaciones de marta, garduña y gato montés (Felis silvestris) no produjeron mejoras en el éxito reproductor del urogallo en Pirineos (Fernández-Olalla 2011). Por el contrario, la declaración de un área protegida en Almería, y la consiguiente prohibición de utilizar control de depredadores, repercutió negativamente en las poblaciones de paseriformes esteparios (Suárez et al. 1993). Estos resultados concuerdan con los obtenidos más recientemente por Estrada et al. (2012), quienes observaron mayores densidades de ciertas aves esteparias en cotos de caza donde se realizaba control de zorros. Por lo tanto, la efectividad del control de depredadores para fomentar las presas en España está poco clara. El efecto de las extracciones sobre las poblaciones de los depredadores controlados igualmente ha sido poco estudiado en España, encontrando resultados dispares. Así en Doñana no se encontró ninguna respuesta poblacional clara a las extracciones de zorros realizadas por personal del Parque Nacional durante cuatro años, probablemente debido a una baja intensidad y gran variabilidad interanual de extracción (Palomares et al. 2010). Igualmente Virgós y Travaini (2005) no encontraron diferencias en la presencia de zorros entre zonas cinegéticas (donde se asumía el uso de métodos de control de depredadores) y zonas sin caza del centro de la Península Ibérica. En un experimento realizado en Pirineos se consiguió reducir la densidad de zorros en una de las zonas de estudio durante uno de los años de estudio. No obstante, esto no se consiguió en otras dos zonas de trabajo ni en la misma zona durante los otros dos años que duró el estudio (Fernández-Olalla 2011). De forma similar las extracciones experimentales realizadas en dos localidades en Navarra redujeron la abundancia en una de las localidades de

24

estudio, mientras que este efecto no fue tan evidente en la otra localidad (MateoMoriones et al. 2012). Prácticamente no existen estudios que hayan evaluado experimentalmente el efecto de las extracciones de urraca sobre sus poblaciones. Herranz (2000) observó una reducción significativa de la población de urracas en un coto de caza tras una campaña de control mediante destrucción de nidos y caza de adultos; sin embargo, no aportó información sobre la evolución tras cesar el control. Un experimento realizado en Navarra, no pudo evaluar el efecto de las extracciones sobre las poblaciones de urracas por ser éstas muy poco abundantes (Mateo-Moriones et al. 2012). Por último, en España no se ha estudiado de forma experimental el efecto del control de depredadores sobre otras especies no relacionadas directamente con el control. Hasta la fecha solamente un estudio observacional ha evaluado el efecto del control no selectivo de zorros en otras especies de mesocarnívoros como el tejón y la garduña (Barrull et al. 2014). No existe ningún trabajo similar en el caso de las urracas ni estudios sobre el efecto potencial del control de depredadores sobre el comportamiento de la especie objetivo del control.

OBJETIVOS Y ESTRUCTURA DE LA TESIS Como queda patente en lo anteriormente dicho, el conocimiento científico en materia de control de depredadores es escaso, especialmente en España (Díaz-Ruiz y Ferreras 2013). El objetivo principal de esta tesis es, por tanto, contribuir al conocimiento científico sobre la gestión del zorro y la urraca, mediante el estudio de diferentes aspectos relacionados como la ecología trófica de estas especies, la adecuación y mejora de los métodos empleados para su control y las implicaciones ecológicas derivadas del control de sus poblaciones. Esta tesis pretende aportar avances en el conocimiento científico para mejorar la gestión de los depredadores generalistas y, por lo tanto, de las especies que pueden verse afectadas por el control de dichos depredadores. Para la consecución del objetivo principal, en esta tesis se plantean los siguientes objetivos parciales: 1) Analizar la ecología trófica de las dos especies seleccionadas como modelo de estudio, el zorro y la urraca, por ser la alimentación el principal motivo en el que se basa el control de sus poblaciones. En el capítulo 1 se plantea un estudio de la alimentación 25

del zorro a escala biogeográfica de la Península Ibérica, una perspectiva espacial más amplia a la descrita hasta ahora, para definir patrones de su alimentación que ayuden a una mejor compresión de la flexibilidad trófica del cánido. El objetivo del capítulo 2 es caracterizar la dieta de las urracas durante su época de reproducción en zonas agrícolas del centro de España para determinar la frecuencia de consumo de ciertos alimentos como huevos y aves, y estudiar la influencia de diferentes factores intrínsecos (sexoedad) y extrínsecos (localidad) en la composición de su alimentación. 2) Evaluar la efectividad y selectividad de los métodos de captura usados con mayor frecuencia en España para controlar zorros y urracas. Además, se pretende analizar diferentes formas de mejorar la efectividad y selectividad de estos métodos de captura. En concreto se evalúa el uso combinado de diferentes cebos y atrayentes para mejorar la eficiencia de captura y selectividad de las jaulas-trampa para zorros, así como la evaluación de nuevos sistemas de captura alternativos como el sistema Collarum (capítulo 3). Igualmente se evalúan las jaulas-trampas habitualmente empleadas para el control de las poblaciones de urraca, ensayando diferentes variantes de uso con la intención de mejorar este método de control (capítulo 4). 3) Analizar posibles efectos del control de depredadores sobre las especies objeto de control así como sobre otras especies. Por un lado estudiar los efectos de las extracciones de estos depredadores sobre la abundancia de sus poblaciones. En concreto, estudiar el efecto a corto plazo de las extracciones experimentales de urracas sobre sus poblaciones (capítulo 4). Por otro lado estudiar si el gradiente de intensidad de control de depredadores está relacionado con la probabilidad de ocupación y detección de depredadores objeto de control, como el zorro, y de otros que a priori no lo son, como la garduña (capítulo 5). Por otro lado, se pretende evaluar si el control de zorros tiene algún efecto sobre el comportamiento de esta especie (capítulo 6). La Tesis está estructurada en 6 capítulos en formato de artículos científicos. Alguno de ellos está publicado en revistas incluidas en el “Science Citation Index”, otros están actualmente en revisión o en preparación para su publicación. Se incluye una discusión general en la que se destacan los resultados más significativos obtenidos en los diferentes capítulos de esta tesis. Finalmente se proponen futuras líneas de investigación surgidas de este trabajo y las principales conclusiones obtenidas en cada capítulo.

26

CAPÍTULO 1: Biogeographical patterns in the diet of an opportunistic predator: the red fox Vulpes vulpes in the Iberian Peninsula

Díaz–Ruiz F, Delibes–Mateos M, García–Moreno JL, López–Martín JM, Ferreira C, Ferreras P (2013) Biogeographical patterns in the diet of an opportunistic predator, the red fox Vulpes vulpes in the Iberian Peninsula. Mammal Review 43: 59-70

27

Abstract Biogeographical diversity is central to the trophic ecology of predators. Understanding the biogeographical trophic patterns of generalist predators, such as the red fox Vulpes vulpes, is particularly challenging because of their wide distributions, broad trophic spectra and high ecological plasticity, which often generate conflicts with humans. We reviewed 55 studies from the Iberian Peninsula concerning the diet of the red fox to describe its trophic patterns from a biogeographical perspective. We considered the frequency of occurrence of seven food groups and characterized each study site according to environmental variables. We tested relationships between geographical variables and each food group independently, and assessed the consumption of lagomorphs in relation to the other food groups. We also tested the relationships between trophic diversity, the main food groups, latitude and altitude, and finally investigated changes in the consumption of all food groups in relation to habitat type and seasonality. We found a latitudinal pattern in the diet of the red fox, which was characterized by a greater consumption of lagomorphs and invertebrates in southern areas, and a higher intake of small mammals and fruits/seeds in northern regions. Additionally, the consumption of invertebrates increased from east to west, while fruit/seed consumption increased from west to east. Consumption of lagomorphs decreased, and of small mammals increased, with altitude. Trophic diversity was not associated with geographical variables. The intake of lagomorphs and small mammals was greatest in Mediterranean scrub and forest, respectively. Reptiles and invertebrates were consumed mostly during summer; fruits/seeds in autumn. Iberian red foxes show variation in their feeding habits associated with environmental variables, which are in turn associated with the availability of their main prey. Foxes select rabbits where they are abundant, and feed on small mammals and fruits/seeds where lagomorphs are scarce.

Keywords: carnivore, feeding patterns, generalist predator, Portugal, Spain

28

Introduction Feeding habits have been one of the most studied features of carnivore ecology. The traditional approach to studies ofcarnivore diets is to investigate the feeding habits of species (mainly in terms of diet composition) at local or regional scales (e.g. Brand et al. 1976; Zapata et al. 2007; Wang and Macdonald 2009). Comprehensive studies of carnivore trophic ecology at broader geographical scales have only recently been undertaken (e.g. Clavero et al. 2003; Lozano et al. 2006; Zhou et al. 2011). The study of trophic biogeographical patterns of predators is fundamental to understanding their ecology and life history strategies (Daan and Tinbergen 1997). For instance, defining a species as a trophic generalist or specialist is only relevant in the context of extensive ecological studies in which variation in feeding behaviour among populations over a broad range of environmental conditions is considered (Lozano et al. 2006). Investigations of the diet of medium-sized carnivores at large biogeographical scales have included studies of the Eurasian badger (Meles meles) (Roper and Mickevicius 1995; Goszczynski et al. 2000; Hounsome and Delahay 2005); the polecat (Mustela putorius) (Lodé 1997); the common genet (Genetta genetta) (Virgós et al. 1999); the Eurasian otter (Lutra lutra) (Clavero et al. 2003); the European wildcat (Felis silvestris) (Lozano et al. 2006); and the Holarctic martens, (Martes sp.) (Zhou et al. 2011). Surprisingly, this type of study is lacking for the red fox (Vulpes vulpes), which is the world’s most widespread member of the order Carnivora (Sillero- Zubiri et al. 2004) and one of the most abundant carnivore species in the Iberian Peninsula (Blanco 1998; Palomo et al. 2007) and elsewhere. Environmental and climatic conditions affect food availability, and can have an impact on dietary composition and diversity (Hill and Dunbar 2002). Thus, variations in the distribution of potential prey species across biogeographical regions have been postulated to affect the feeding habits of medium-sized carnivores. For instance, dietary diversity in wildcats increases at lower latitudes (i.e. Mediterranean areas; Lozano et al. 2006), where potential prey richness is greater (Rosenzweig 1995). Latitudinal gradients have also been observed in relation to dietary diversity and in the consumption of particular prey. For example, the Eurasian otter’s diet is more diverse in southern localities, while further north the species is more piscivorous, predating upon a large diversity of fish families (Clavero et al. 2003). Similarly, food availability can vary along altitudinal gradients, and this can affect the dietary composition of carnivores. For 29

instance, small mammals (mice, voles and shrews) are the primary food of martens, but are less frequently consumed at lower altitudes, where other food resources are more abundant and are available throughout the year (Zhou et al. 2011). Diet is one of the most studied aspects of the ecology of the red fox. Most studies indicate that the red fox is a generalist predator that uses resources according to their availability and hence is opportunistic in its behaviour (e.g. Webbon et al. 2006; Dell’Arte et al. 2007). However, most studies were undertaken at local or regional scales, and specific studies describing biogeographical patterns in the red fox diet are lacking. Although some studies have shown variations in the feeding habits of foxes based on environmental variables including habitat type (Fedriani 1996; Gortázar 1999), the effects of latitude, longitude and altitude on the composition of fox diets at a larger scale remain unknown. Similarly, there is a lack of information about how the consumption by foxes of some preferred prey, such as lagomorphs or small mammals, varies spatially at biogeographical scales. The ecological features of red foxes can bring them into conflict with human activities where their prey is of economic or conservation concern (Baker and Harris 2003). For example, predation by foxes is often regarded as one of the factors preventing the recovery of small game (Reynolds and Tapper 1995; Smedshaug et al. 1999; Beja et al. 2009; Knauer et al. 2010), and farmers consider predation of livestock by foxes to cause economic losses (Moberly et al. 2004). Furthermore, several researchers have reported negative impacts of fox predation on species of conservation concern (Yanes and Suárez 1996; Ruiz-Olmo et al. 2003; Dickman 2010). However, predators, including generalists such as red foxes, play major roles in ecological processes by limiting populations of pest species (O’Mahony et al. 1999; Newsome et al. 2001), reducing the transmission of disease (Hudson et al. 1992; Millán et al. 2002) and acting as seed dispersers (Guitián and Munilla 2010; Rosalino et al. 2010). Our ability to understand biogeographical patterns is crucial for developing efficient management programs in the context of human usage (Whittaker et al. 2005). From this perspective, a large-scale study of the trophic ecology of the red fox could provide valuable knowledge concerning its ecosystem functions and improve management of this predator. The Iberian Peninsula is included in the Mediterranean Basin hotspot (Myers et al. 2000) and is thereby an interesting site for the study of biogeographical patterns (e.g. 30

Carvalho et al. 2011). It includes distinct Atlantic (Northern Iberia), Mediterranean (Central and Southern Iberia) and Alpine (Pyrenees mountains) biogeographical regions (Rivas-Martínez 1987; Figure 1.1.), and is characterized by high environmental heterogeneity because of its climatic and physiographical complexity (the altitude ranges from 0m at sea level to 3479m above sea level at Sierra Nevada, Granada, Spain). The variability in environmental conditions underpins the diversity in community composition and structure in this region (Blondel and Aronson 1999, Stefanescu et al. 2004). Several patterns in the distribution and abundance of the main prey species of Iberian predators have been described. For instance, wild rabbits Oryctolagus cuniculus, which are a key prey for red foxes and other Iberian predators (Delibes and Hiraldo 1981; Calzada 2000; Ferreras et al. 2011), are most abundant at central–southern latitudes (Villafuerte et al. 1998), and small mammals show a gradient in abundance and species richness from south to north (Soriguer et al. 2003). The theory of feeding specialization predicts an increase in dietary diversity when the preferred prey becomes scarce (Futuyma and Moreno 1988). In this study, we tested this prediction in relation to the red fox and rabbits as its preferred prey. Although the Iberian Peninsula is a relatively small biogeographical area, its high environmental variability and biodiversity justifies a biogeographical analysis of the diet of resident generalist carnivores such as the red fox. Our main objective was to describe the trophic biogeographical patterns of the red fox in the Iberian Peninsula, based on a comprehensive literature review. Specifically, we: (i) evaluated changes in consumption by red foxes of main food groups in relation to geographical variables (latitude, longitude and altitude); (ii) analysed the relationships between red fox dietary diversity, consumption of its main prey and geographical variables; (iii) assessed the relationships between the consumption of different food groups and habitat type and season; and (iv) interpreted patterns in the diet of this generalist predator from a biogeographical perspective.

31

Figure 1.1. Geographical distribution in the Iberian Peninsula of studies of the diet of the red fox (Vulpes vulpes) included in this review. Biogeographical regions are shown, and the numbers represent study site identifiers (ID; see Appendix 1.1.).

Material and Methods Literature compilation and standardization of dietary data Various sources of information were used to review the available literature comprehensively, as recommended by Pullin and Stewart (2006). Search engines (ISI Web of Science and Google Scholar) were used to identify relevant scientific studies containing information about the trophic ecology of the red fox in the Iberian Peninsula.We searched for terms that were identified using the following combinations of keywords: ‘red fox’ or ‘Vulpes vulpes’ and ‘diet’ or ‘feeding’ and ‘Iberian Peninsula’, ‘Spain’ or ‘Portugal’. We consulted several zoological bibliographical data bases including the Zoological Record (http://scientific.thomson.com/products/zr/) and the bibliographical data set of the Spanish Society for the Conservation and Study of Mammals (http://www.secem.es/Secem_la_biblioteca.htm). We also sought information on the topic from informal contacts with expert researchers (colleagues working in 32

different institutions – universities and environmental public administration – in Spain and Portugal). This provided us with less readily accessible sources of information, including unpublished or unedited studies (e.g. PhD theses, MSc and BSc dissertations, and public administration data bases). We compiled a total of 55 published and unpublished studies concerning the diet of the red fox in Portugal and Spain, spanning the period 1971–2008. Some authors reported data pooled annually, others reported data pooled seasonally, and several provided both annual and seasonal data. To simplify the statistical procedures, two independent data bases were created for analysis: one comprising annual data and the other seasonal data. These data bases were analysed independently (see Statistical analyses). To standardize data from different geographical areas (for later comparison and analysis), we excluded studies: (i) with small sample sizes (scat or stomachs; n < 30 for anual studies and n < 15 for seasonal studies); (ii) reporting data for only one prey group; (iii) containing duplicated information, e.g. academic dissertations later published as scientific articles; and (iv) reporting only relative frequency of occurrence (RF, expressed as the percentage of times one food ítem occurs in relation to the total times all food items occur) or percentage biomass. This last exclusion meant that we only considered studies reporting the frequency of occurrence (FO, expressed as the percentage of scats/stomachs containing a particular food item) for the various food groups. RF values are considered to be highly suitable for interpopulation comparisons in diet studies (Clavero et al. 2003), and biomass is considered a direct measure of the energetic value of prey items consumed (Reynolds and Aebischer 1991), and therefore the best approximation to the true diet (Klare et al. 2011). However, only a small proportion of the reviewed studies presented RF or biomass information, while FO is widely used in carnivore diet studies and was used in most of the red fox studies considered in this review. Moreover, FO can be used to assess whether a predator behaves as an opportunist or as a specialist forager (Klare et al. 2011), and it is considered a valid parameter for comparative purposes (Reynolds and Aebischer 1991; Klare et al. 2011). The application of the four exclusion criteria above resulted in a final set of 37 studies that were further analysed to describe red fox feeding patterns in the Iberian Peninsula. These studies were carried out in 39 locations distributed throughout the region (Figure. 33

1; for more detailed information, see Appendices 1.1. and 1.2.). The data were highly heterogeneous among the variables, which reflected the diversity of environmental conditions in the Iberian Peninsula. For example, a broad altitudinal range (20– 1425m) was included, and various habitat types were represented, including several types of Mediterranean scrub, agricultural lands, dehesas (savannah-like formations that combine pastures with intermittent cereal cultivation in park-like oak woodlands; Blondel and Aronson 1999) and forests containing various tree species (e.g. Pinus sp. and Quercus pyrenaica). Variable selection From each study we derived the following parameters: respective geographical variables (latitude and longitude, in degrees; and altitude, in metres) either from the study itself or, if they were not provided in the study, from Google Earth (http://earth.google.com); the source of food materials analysed (scats or stomach contents); and the simple size, study duration, season, habitat, and FO of each food group (see Appendices 1.1. and 1.2.). We categorized dietary items into the following main groups: lagomorphs (mainly European wild rabbits; see Results), small mammals (rodents and insectivores), birds, reptiles, invertebrates, fruits/seeds, and carrion/garbage (mainly large mammals and leftover food of anthropogenic origin). Four seasons were considered: spring (March– May), summer (June–August), autumn (September–November) and Winter (December– February). The habitat type at each location was categorized as Mediterranean scrub, forest or agricultural–dehesa (agricultural land and dehesas), according to the descriptions given in each study. We calculated Herrera’s trophic diversity index (D; Herrera 1976) from the FO data as an index of the trophic diversity for each diet. The index is computed according to the formula

= −∑

log pi, where p is the

frequency of occurrence of the various prey categories (i). This index is recommended for presence–absence food data, because other diversity indices such as the Shannon index cannot be calculated from this type of data (Herrera 1976). To test for bias caused by the study duration, sample size or source of analysed food material (scats or stomach contents; Putman 1984), we followed the approach of earlier authors (Lozano et al. 2006; Zhou et al. 2011) and used multivariate analysis of covariance with the study duration and simple size as covariates, food material as a fixed factor and the FO of each of the seven food groups as response variables. 34

To avoid temporal pseudo-replication, we considered only those studies in which annual information on the Iberian fox diet was provided: 30 studies and localities, including a total of 9459 samples (stomachs and scats; see Appendix 1.1. and 1.2.). Therefore, analyses of the relationship of the consumption of various food groups to geographical variables and habitat type were performed using the anual data base. The testing of seasonal variation was based only on those studies in which seasonal data were reported: 18 studies and 20 localities, including a total of 5027 samples (stomachs and scats; see Appendices 1.1. and 1.2.). The relationships between geographical variables (latitude, longitude and altitude) and the FO of each food group were tested using simple regression analyses. In view of the potential importance of wild rabbits in the diet of red foxes, we used a simple regression analysis to investigate the relationships between the lagomorph FO (mainly wild rabbits; see Results) and the FO of other food groups. To evaluate whether trophic specialization occurred in Iberian red foxes, we tested the relationships between diet diversity (Herrera D index) and the FO of each of the four main food groups (lagomorphs, small mammals, invertebrates and fruits/ seeds) using data from annual studies. We applied general linear models (GLMs) using a normal distribution for errors of the response variable (Herrera D index) and an identity link function. One-way analysis of variance was used to test the effect of habitat type on the FO of each food group. We assessed seasonal variations in the diet by performing separate one-way analyses of variance with the FO of each food group as a dependent variable. We conducted Tukey’s post-hoc tests to assess differences between pairs of habitat types and seasons. Prior to statistical analyses, the FO for each food group and the Herrera D index values (dependent variables) were arc sine and log transformed, respectively, to achieve normality (Zar 1984), which was assessed visually from normal probability plots. All statistical analyses were performed using Statistica 6.0 software (StatSoft 2001).

Results We found no significant effect of study duration (F7,26 = 0.86, P = 0.55), sample size (F7,26 = 0.73, P = 0.64), source of analysed food material (scats or stomach contents; F7,26 = 0.43, P = 0.11) or the interaction between sample size and food material (F7,26 = 1.04, P = 0.42) on the FO of food groups in the diet. Thus, for further analyses we 35

pooled data from studies with differing durations, sample sizes and sources of analysed food material. Overal diet Iberian red foxes consume a wide range of food items. Invertebrates were the most frequent food group in their diet (mean FO±SD, 40.1±25.5%), followed by fruits/ seeds (38.9±22.0%), small mammals (34±20.9%), lagomorphs (20.6±22.0%), carrion/garbage (15.3±14.2%), birds (13.4±15.3%) and reptiles (1.8±2.8%). Coleoptera and Orthoptera species were the most common among the invertebrates, and both wild and cultivated fruits were included among the fruits/sedes consumed. The most common small mammal prey was Apodemus sylvaticus, followed by Microtus spp., Crocidura spp. and Eliomys quercinus. Wild rabbit was the dominant species among the lagomorphs, while hares Lepus spp. Were rare in the red fox diet (only identified in 6 of the 27 studies that recorded lagomorphs; FO = 1.2±0.43%). For this reason, we will use indistinctly ‘rabbits’ and ‘lagomorphs’ from now on in the text. The large mammals reported as fox food items included Cervus elaphus, Dama dama, Sus scrofa, Bos taurus, Ovis aries and Capra hircus, and were presumably consumed as carrion. Among birds in the fox diet, the most common species consumed were Columba spp., Alectoris rufa, Galerida spp. and Anas spp. Several reptile species were consumed, including Psammodromus spp., Malpolon monspessulanus and Elaphe scalaris. Geographical patterns (latitude, longitude and altitude) We found a negative and statistically significant relationship between latitude and the FO of lagomorphs (R2 = 0.19, F1,35 = 8.47, P = 0.006; Figure 1.2a.) and invertebrates (R2 = 0.11, F1,35 = 4.37, P = 0.04; Figure 1.2b.), and a positive and significant relationship between latitude and the FO of small mammals (R2 = 0.16, F1,35 = 6.78, P = 0.01; Figure 1.2c.) and fruits/sedes (R2 = 0.12, F1,35 = 5.04, P = 0.03; Figure 1.2d.). Therefore, at lower latitudes, lagomorphs and invertebrates were more frequently eaten, while at higher latitudes small mammals and fruits/seeds were more commonly consumed. Only the FO of invertebrates and fruits/seeds were significantly related to longitude. The consumption of invertebrates increased towards the east (R2 = 0.12, F1,35 = 4.95, P = 0.03), whereas that of fruits/seeds increased towards the west (R2 = 0.16, F1,35 = 6.99, P = 0.01).

36

1.4

1.2

1.2

Small mammals

1.4

1.0 0.8 0.6

0.8 0.6 0.4

0.2

0.2

0.0

0.0

1.4

1.6

R2 = 0.11

b

1.4

1.2

R2 = 0.12

d

1.2

1.0 0.8 0.6

1.0 0.8 0.6

0.4

0.4

0.2

0.2

0.0 36

R2 = 0.16

c

1.0

0.4

1.6

Invertebrates

1.6

R2 = 0.19

a

Fruit/Seed

Lagomorph

1.6

37

38

39

40

41

42

43

44

0.0 36

37

38

39

Latitude

40

41

42

43

44

Latitude

Figure 1.2. Relationships between latitude and the frequency of occurrence (FO; arc sine transformed) of (a) lagomorphs (b) invertebrates (c) small mammals and (d) fruits/seeds in the diet of the red fox. Each point represents one study site (see Figure 1.1.).

Altitude was significantly and negatively associated with the FO of lagomorphs (R2 = 0.29, F1,30 = 12.67, P = 0.001; Figure 3a), and positively associated with that of small mammals (R2 = 0.27, F1,30 = 11.31, P = 0.002, Figure 1.3b.). Thus, the consumption of lagomorphs decreased with altitude, and that of small mammals increased.

Is the red fox specialized on rabbits in the Iberian Peninsula? The consumption of wild rabbits (represented by lagomorphs) was significantly and negatively related to the consumption of both small mammals (R2 = 0.15, F1,35 = 6.23, P = 0.02) and fruits/seeds (R2 = 0.17, F1,35 = 8.41; P = 0.006). The GLM results suggest that diet diversity was not significantly associated with latitude (F1,25 = 0.33, P > 0.5), altitude (F1,25 = 0.552, P > 0.4) or the FO of the four main food groups (lagomorphs: F1,25 = 0.126, P > 0.7; small mammals: F1,25 = 0.004, P > 0.9; invertebrates: F1,25 = 0.253, P > 0.6; and fruits/seeds: F1,25 = 0.196, P > 0.6). 37

1.4 R2 = 0.29

a 1.2

Lagomrph

1.0 0.8 0.6 0.4 0.2 0.0 1.4 R2 = 0.27

b Small mammals

1.2 1.0 0.8 0.6 0.4 0.2 0.0 0

200

400

600

800

1000

1200

1400

1600

Altitude

Figure 1.3. Relationships between altitude (in metres) and the frequency of occurrence (FO; arc sine transformed) of (a) lagomorphs and (b) small mammals in the diet of the red fox. Each point represents one study site (see Figure 1.1.).

Habitat type and seasonality We found a significant relationship between habitat type and the FO of lagomorphs (F2,21 = 8.10, P = 0.002) and small mammals (F2,20 = 4.05, P = 0.03) in red fox diet. The FO of lagomorphs was higher in Mediterranean scrub than in forest (Figure 1.4a.), but the opposite was observed for small mammals (Figure 1.4b.). A significant seasonal relationship in the red fox diet was found for reptiles (F3,53 = 3.34, P = 0.02), invertebrates (F3,53 = 9.45, P < 0.0001) and fruits/seeds (F3,53 = 11.49, P < 0.0001). The FO of reptiles increased from winter to summer (Figure 1.5a.); invertebrates were mostly consumed in summer, and their occurrence in the diet was lowest in winter (Figure 1.5b.); and fruits/seeds were consumed most in autumn and 38

least in spring (Figure 1.5c.). Marginally significant differences were found for lagomorphs (F3,53 = 2.40, P = 0.07), which were consumed most in summer (Figure 1.5d.). 1.0 0.9

a

Lagomorphs

0.8

A

0.7

A, B

0.6 0.5 0.4

B

0.3 0.2 0.1 0.0

Small mammals

1.0 0.9 0.8 0.7

B

b

A, B

A

0.6 0.5 0.4 0.3 0.2 0.1 0.0

M. Scrub

Forest

Agri./Dehesa

(N=12)

(N=9)

(N=3)

Figure 1.4. Frequency of occurrence (FO; arc sine transformed; means±SE) of (a) lagomorphs and (b) small mammals in the diet of the red fox as a function of habitat type. Means marked with the same letter are not significantly different from one another (P < 0.05; Tukey’s post-hoc test). M. scrub, Mediterranean scrub; Agri., agricultural lands.

39

0.25

a

B A, B

0.15 A, B

0.10

A

0.05

Invertebrates

Reptiles

0.20

0.00

1.0 0.8 0.7

C

b B, C

A, B A

1.0

c A, C C

A

0.6 0.5 0.4

B

0.3

Lagomorphs

Fruits/Seeds

0.9

1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.9 0.8 0.7 0.6 0.5

d A A, B

A, B

0.4 0.3

B

0.2 0.1

0.2 0.1 0.0

0.0

Winter

Spring

(N=11)

(N=15)

Summer (N=18)

Autumn

Winter

Spring

(N=13)

(N=11)

(N=15)

Summer Autumn (N=18)

(N=13)

Figure 1.5. Frequency of occurrence (FO; arc sine transformed; means±SE) of (a) reptiles (b) invertebrates (c) fruits/seeds and (d) lagomorphs in the diet of the red fox, as a function of season (marginally non-significant for lagomorphs, P = 0.07). Means marked with the same letter are not significantly different from one another (P < 0.05; Tukey’s post-hoc test).

Discussion Biogeographical variations in the diet of the red fox in Iberia Generalist predators feed on different food resources according to their abundance and availability (Futuyma and Moreno 1988). This study confirms that the red fox is a generalist predator; its trophic patterns can be explained by geographical variables, habitat type and seasonality. These factors determine directly the abundance and availability of its main foods [e.g. wild rabbits are more abundant at southern latitudes (Villafuerte et al. 1998) and in Mediterranean scrubland habitats (Calvete et al. 2004); small mammals are more abundant at northern latitudes (Soriguer et al. 2003) and in forest habitats (Torre et al. 2002)]. Latitude influences the feeding patterns of many medium-sized carnivores (Clavero et al. 2003; Hounsome and Delahay 2005; Lozano et al. 2006; Zhou et al. 2011). Some researchers relate dietary patterns in the abundance and diversity of prey species with the latitudinal gradient described in Eurasia, which

40

increases towards the south (Pianka 1966; Blondel and Aronson 1999). Our results are consistent with these findings as we observed a latitudinal gradient in the consumption of lagomorphs, invertebrates, small mammals and fruits/seeds by red foxes. The increase in the consumption of lagomorphs, mainly wild rabbits, towards southern Iberia is a consequence of the greater abundance of this prey at these latitudes (Villafuerte et al. 1998). The same pattern in rabbit intake has been shown for other medium-sized Iberian carnivores including the wildcat (Lozano et al. 2006), the badger (Virgós et al. 2005; Barea-Azcón et al. 2010) and the polecat (Santos et al. 2009). This feeding pattern could explain the negative latitudinal gradient found in the body size of Iberian red foxes, which contradicts Bergmann’s Rule (Yom-Tov et al. 2007). The high occurrence of invertebrates in the red fox diet in southern regionsmay be explained by the greater availability of this food type at low latitudes (Chapman 1998; Blondel and Aronson 1999) and is in agreement with studies of the diet of other medium-sized Iberian generalist carnivores including the genet (Virgós et al. 1999). The positive relationship between latitude and small mammal consumption by Iberian red foxes corresponds to a south–north gradient in the abundance and species richness of this prey group (Blanco 1998; Soriguer et al. 2003). The decrease in rabbit abundance in northern regions of the Iberian Peninsula also promotes the switch to small mammals as the main prey in these areas. This pattern was also observed by Zhou et al. (2011) in Holarctic marten species at a larger biogeographical scale. The consumption of fruits/seeds by the red fox is greater in northern regions than in southern regions. However, this pattern is opposite to that described for other Eurasian generalist carnivores, which decrease their consumption of plant matter and increase carnivory with increasing latitude (Virgós et al. 1999; Goszczynski et al. 2000; Vulla et al. 2009; Zhou et al. 2011). In some of these studies, this pattern is explained by a reduction in primary production with increasing latitude, but the narrow latitudinal range covered in the present study leads us to believe that the higher consumption of fruits/seeds is likely to be due to the greater availability of this resource in the north of the Iberian Peninsula. The FO of invertebrates in the fox diet increases from east to west, while that of fruits/seeds increases from west to east. Rosalino and Santos-Reis (2009) were not able to explain a similar longitudinal gradient found in fruit/seed consumption by medium41

sized carnivores in Iberia because of the absence of data on the availability of plant species producing fruits and seeds. Invertebrates are an alternative food source for some omnivorous species, especially larger carnivorous mammals, where larger prey items are not available (Capinera 2010). However, as there is currently no information on the availability of invertebrates over a longitudinal gradient in Iberia, we have no data to enable us to interpret our results. The decrease in consumption of lagomorphs by foxes with increasing altitude could be because of the reduced presence and abundance of rabbits above 1000m (Blanco 1998; Palomo et al. 2007), but the consumption of small mammals by foxes increased in high altitude areas. This is in contrast with previous findings that the species richness and abundance of small mammals decreases at higher altitudes (Torre 2004). However, the altitudinal range considered in this study (only three localities were higher than 1400m; see Appendix S1) did not include altitudes that may limit the presence of most small mammals consumed by the red fox (Palomo et al. 2007), which prevents us from confirming this trend in small mammal consumption. Thus, the increased intake of small mammals seems to be a functional response to the reduced availability of lagomorphs at higher altitudes, as Hartová-Nentvichová et al. (2010) found for red foxes in the mountains of the Czech Republic. Is the red fox specialized on rabbits in the Iberian Peninsula? A negative relationship between a given food group and dietary diversity is usually interpreted as indicating trophic specialization (Futuyma and Moreno 1988; Fedriani et al. 1998; Lozano et al. 2006). A negative relationship at a regional scale between lagomorph consumption and dietary diversity has been described for red foxes (DelibesMateos et al. 2008) and for other small and medium-sized Mediterranean carnivores (Sarmento 1996; Lozano et al. 2006; Santos et al. 2009). However, we did not find any significant relationship between dietary diversity and the consumption of lagomorphs or other prey, or geographical variables, perhaps because of the high trophic flexibility of the fox in the Iberian Peninsula. These results suggest that, at the scale of the peninsula, only small mammals and fruits/seeds are eaten by foxes as alternatives to lagomorphs. This confirms the opportunistic and generalist feeding behaviour of the red fox, as has consistently been reported for different geographical areas and at various scales (e.g. Kjellander and Nordstrom 2003, Dell’Arte et al. 2007). 42

Habitat type and seasonality We observed a high intake of lagomorphs by red foxes in the Mediterranean scrubland, where wild rabbits reach higher densities (Fedriani 1996; Palomares 2001; Calvete et al. 2004). In contrast, Fedriani (1996) found no difference in consumption of wild rabbits by red foxes in adjacent áreas of scrubland and dehesa habitat in Doñana (southwest Iberian Peninsula), despite higher rabbit density in the scrubland patches. This is probably a consequence of the larger scale considered in our review, where habitats were clearly differentiated between studies. The preference for forests shown by the small mammal species most frequently consumed by foxes (e.g. Apodemus sylvaticus; Torre et al. 2002), together with the low abundance of rabbits in this type of habitat, explains why foxes include in their diet a greater proportion of small mammals in forests than in others habitats. Several researchers have reported marked seasonality in the diet of the red fox (Dell’Arte et al. 2007; Hartová-Nentvichová et al. 2010). Mediterranean ecosystems have marked climatic seasonality, with hot dry summers and cold wet winters (Blondel & Aronson 1999); thus, some trophic resources for carnivores are only seasonally available (Virgós 2002). We also observed a marked seasonality in the diet of the red fox, which is a result of the seasonal availability of some food groups at the Iberian scale. Populations of Orthoptera and Coleoptera, the invertebrates most consumed in summer, increase dramatically during this season (Aranda et al. 1995; Loureiro et al. 2009). The availability of cultivated and wild fruits is greatest in summer and autumn (Loureiro et al. 2009), when they are most consumed by foxes. The annual abundance of wild rabbits in the Iberian Peninsula peaks in the spring–summer period (Soriguer 1981; Beltrán 1991). At this time the greater availability of juvenile rabbits and the susceptibility of the rabbit population to myxomatosis (Calvete et al. 2002) may make this prey more vulnerable to predation and consumption as carrion by foxes, so that rabbits may provide a valuable energy source for foxes during the highly critical breeding period. This explains the observed seasonal increase in the FO of lagomorphs from spring to summer (Figure 1.5d.). However, in areas where rabbits are very abundant, their availability is high throughout the year (Angulo and Villafuerte 2003), which could explain the lack of statistically significant differences between seasons in the FO of lagomorphs in the red fox diet.

43

Conclusions Biogeographical variation in the feeding habits of Iberian red foxes are associated with geographical variables, hábitat type and season, which affect the availability of alternative potential foods (Figure 1.6.). Our results confirm that the feeding habits of the red fox, a generalist predator, vary widely both spatially and temporally, even within a relatively small biogeographical area such as the Iberian Peninsula. Therefore, we demonstrate that the flexibility of this generalist predator really reflects the biogeographical patterns of distribution and abundance of its main food sources. Understanding these patterns in the feeding ecology of the red fox, the most abundant carnivore in the Iberian Peninsula, will facilitate the understanding of the geographical variations in its abundance and behaviour, and improve the management and conservation of this species.

Figure 1.6. Conceptual model illustrating the biogeographical patterns found in the consumption of the main food groups by the Iberian red fox, in relation to geographical variables (LAG, lagomorphs; SM, small mammals; F/S, fruits/seeds; INV, invertebrates). The white arrows represent latitudinal (LATITUDE) and longitudinal (LONG) gradients, and the grey arrow shows the altitudinal gradient (ALTITUDE).

44

Acknowledgements We are especially grateful to Drs P. C. Alves and C. Gortázar for providing unpublished data to be included in this review. We thank also Drs. Jennings and Hackländer, and two anonymous referees whose comments greatly improved the manuscript. M. Delibes-Mateos currently holds a Juan de la Cierva research contract awarded by the Ministerio de Ciencia e Innovación and the European Social Fund. C. Ferreira was supported by a PhD grant (Ref. SFRH/BD/22084/2005) funded by the Fundação para a Ciência e Tecnologia of the Ministério da Ciência, Tecnologia e Ensino Superior, Portuguese government. Financial support for the study was provided by the Spanish MICINN Project CGL2009-10741 from Spanish Plan Nacional de I+D and FEDER funds.

45

CAPÍTULO 2: Factors affecting the feeding habits of black-billed magpies Pica pica during the breeding season in Mediterranean Iberia

Este capítulo ha sido enviado a una revista SCI: Díaz-Ruiz F, Zarca JC, Delibes-Mateos M., Ferreras (enviado) Factors affecting the feeding habits of black-billed magpies Pica pica during the breeding season in Mediterranean Iberia.

46

Abstract Feeding habits of the black-billed magpie are of conservation and management interest for researchers, conservationists and hunters since magpies are considered as predators of eggs and chicks of both songbirds and gamebirds. The aim of this study was to characterize the feeding habits of magpies during the breeding season of birds (i.e. magpies and sympatric birds) in agricultural environments of central Spain, and to assess the occurrence and incidence of birds and eggs in the magpie’s diet. Diet was determined by the analysis of gizzards contents from 118 magpies. We tested the effect of locality, age and sex on diet composition and diet diversity through multivariate analysis of variance (MANOVA) and general lineal models (GLM). Magpies presented a generalist diet, which included a wide range of foods. Arthropods and cereal seeds were the most frequent food groups (frequency of occurrence, FO >60 %). Eggs and birds were consumed only occasionally (FO < 6% and 17% respectively; percentage of volume, VOL, < 4%), and more frequently during magpie incubation stage. We did not find overall significant differences in diet related with age and sex. Significant effects were only found for the interaction between sex and age and between them and locality. Our findings suggest that magpies do not seem to pose an important threat for the conservation of birds in Mediterranean agricultural environments, under the conditions found during this study. Nevertheless, more complex studies in different scenarios (i.e. different population sizes of magpies and prey) and at longer temporal scales are necessary to clarify this controversial issue.

Key words: bird conservation, egg predation, feeding habits, generalist diet, predator control

47

Introduction Feeding habits is an important and widely studied aspect of animal ecology and a fundamental component for understanding the biology and ecology of species. Some species are perceived as harmful for human interests, frequently because of their feeding habits. For instance, some predators can consume species of human interest such as game species or livestock (Woodroffe et al. 2005). From this point of view, the information provided by studies on predator feeding habits may be relevant to guide appropriate policy and management decisions (López-Bao et al. 2013) that facilitate human-wildlife coexistence. Feeding habits of the black-billed magpie (Pica pica, hereafter the magpie) give rise to controversial interpretations between researchers, conservationists and hunters. In Europe, magpies are considered as a harmful bird species by some conservationists and hunters because of their predation on eggs and chicks of songbirds and gamebirds (Birkhead 1991; Herranz 2000). As a consequence, control of magpie populations is widespread in Europe (Hadjisterkotis 2003), particularly in southern regions (Chiron and Julliard 2013; Díaz-Ruiz and Ferreras 2013). In Spain, magpie control is mostly performed by hunters and game managers, who consider magpies as high efficient predators of nests of red-legged partridges (Alectoris rufa) (Delibes-Mateos et al. 2013; Díaz-Ruiz and Ferreras 2013), a small game species of socioeconomic relevance (DíazFernandez et al. 2012). Magpies feeding habits have been object of several studies focusing on different issues, e.g. seasonal diet composition, food selection, diet of nestlings or differences between feeding patterns of rural and urban magpies (Birkhead 1991; Soler and Soler 1991; Martínez et al. 1992; Ponz et al. 1999; Kryštofková et al. 2011). These studies describe magpies as generalist predators that feed on a broad spectrum of food types. In general, eggs form only a small proportion of magpie diet (Birkhead 1991; Martínez et al. 1992), although some studies have shown that magpies are one of the main predators of artificial and natural nests (Groom 1993; Herranz 2000; Miller and Hobbs 2000; Roos and Pärt 2004). Nevertheless, the impact of magpies on bird populations remains still unclear, due to contrasting results (Gooch et al. 1991; Stoate and Szczur 2001; Thomson et al. 1998; Chiron and Julliard 2007; Newson et al. 2010), particularly in the Iberian Peninsula, where the number of studies on this issue is low. In addition, other basic

48

aspects of the feeding habits of magpies, such as how these are affected by intrinsic factors (e.g. age or sex) remain largely unknown. Differences in feeding behaviour related to age and sex have been shown for several vertebrate species, e.g. reptiles (Liu et al. 2011), mammals (Kidawa and Kowalczyk 2011) and birds (Le Vaillant et al. 2013). In bird species, foraging behaviour may differ between males and females, in order to avoid intraspecific competition for food resources (Le Vaillant et al. 2013). Moreover, individuals improve with age their knowledge of the environment and their ability to prospect for food, which means that older individuals can expand the range of available dietary items, or focus on more profitable foods, increasing their foraging efficiency (Pärt 2001; Gomes et al. 2009). Biometrical differences occur between sexes and age classes in magpies; males are larger than females and adults are larger than yearlings (Birkhead 1991; Martínez 2011). Furthermore, during the breeding period males and females take on different roles, e.g. only females incubate (Buitron 1988). In addition, magpies can remember the type of food they hoarded, in which location, and when this hoarding took place (Zinkivskay et al. 2009), and this ability may be more accentuated in more experienced adult birds than in yearlings. Therefore, these biological and behavioural differences linked to age and sex may be a source of variation in the magpie’s diet as observed in the case of other birds (Le Vaillant et al. 2013; Pärt 2001; Gomes et al. 2009). On the one hand, larger individuals may capture larger prey, such as birds, and more experienced individuals may have learned to exploit resources not used by less experienced individuals, such as nests. Although these aspects may be very relevant for magpies’ management, they have not been tested or described so far for this species. In the present study or main goal was to characterize the diet of magpies during their breeding season in agricultural rural areas of central Iberia. Our specific aims were to examine: (1) the occurrence and importance of birds and eggs in the diet of magpies and (2) whether age, sex and area may be sources of variation in the feeding habits of magpies.

Material and Methods Study Area Magpie feeding habits were studied in two hunting estates located in central Spain (Area 1: 960 ha, 39º 4.5´ N, 3º 54´ W; Area 2: 547 ha, 39º 33´ N, 3º 12´ W), during 49

spring 2006. Both study areas were within the Mediterranean bioclimatic region (RivasMartínez et al. 2004), and were similar in habitat composition: an agricultural dominated landscape with some interspersed patches of natural vegetation (mainly Mediterranean bushes and some trees in riparian areas and hedgerows). Main crops were cereals (~50 and 70% of total surface) and, to a lesser extent, vineyards and olive groves. Hunting was an important activity in both estates, and the main game species were Iberian hare (Lepus granatensis), European wild rabbit (Oryctolagus cuniculus) and red-legged partridge. Partridge density was low in both estates (less than 0.36 partridges/ha, authors, unpublished data), within the range of other agricultural regions of the Iberian Peninsula (Borralho et al. 1996; Duarte and Vargas 2001). Both hunting estates harbor an important community of small breeding birds, including species of families such as Alaudidae or Fringillidae (Martí and Del Moral 2003). Magpie density in both study areas (Area 1: 0.23 magpies/ha, Area 2: 0.39 magpies/ha, before breeding season; see Díaz-Ruiz et al. 2010) was above average values reported in other areas of Europe (Birkhead 1991). Sample collection

Magpies were captured during an experimental evaluation of cage-traps as live capture methods for magpie population management (see for more details Díaz-Ruiz et al. 2010). Magpies were captured during their breeding season of 2006 (Birkhead 1991; Soler et al., 1999; Ponz and Gil-Delgado 2004): during May in Area 1 and during late May-early June in Area 2. Birds were euthanized using standard procedures and following current guidelines of animal welfare (Close et al. 1997). Age was determined from the shape and appearance of the first outermost primaries; this method allows to differentiate between first-year (hereafter young) and older magpies (hereafter adult) (Erpino 1968; Birkhead 1991). Sex was determined for each individual by the assessment of gonadal development during laboratory necropsies. Gizzard contents were extracted and placed in 70% alcohol in labeled plastic tubes for subsequent analyses. Gizzard contents analysis Magpie diet was determined through the analysis of gizzard contents, a frequent methodology used in diet study of several bird species (Jiguet 2002; Kopij 2005; Bur et al. 2008). Gizzard contents were analysed in the laboratory following the methodology 50

described in corvid diet studies (Soler et al. 1990; Soler and Soler 1991; Herranz 2000). Food items were identified to the lowest possible taxonomic level using published literature (Day 1966; Barrientos 1988; Devesa 1991; Teerink 1991; Chinery 1997), as well as a dedicated reference collection of seeds, invertebrates, bird eggs and mammal hairs. The thickness of eggshells was measured with a digital calliper (precision 0.01 mm) to assign the eggs at least to the family level (Herranz 2000). All identified items were pooled in nine food classes: arthropod, gastropod, cereal seed, fruit, other vegetal, bird, egg, reptile and mammal, and two non-food items: gastrolith and plastic (Table 2.1. and Appendix 2.1.). We estimated the minimum number of individuals per food class present in each gizzard by: the presence of whole individuals or diagnostic hard structures (e.g. thorax, elytrum, chelicerae or heads) for invertebrates; cereal grain husk and fruit seeds; for vertebrates we assumed a minimum number of one since usually only feathers, hair or small fragments of eggshell appeared. We calculated three dietary indices frequently used in diet studies (Soler et al. 1993; Herranz 2000; Hadjisterkotis 2003; Kryštofková et al. 2011): the frequency of occurrence (FO) expressed as the percentage of gizzards in which a food item was found, the relative frequency of occurrence (RF) expressed as the percentage of times a food item occurs in relation to the total times all food items occur, and the percentage of volume (VOL) estimated as the percentage of total volume corresponding to a certain food item upon the total content of each gizzard. Data analysis We used VOL of each food class in the statistical analyses because this index considers the amount of each food class in each magpie gizzard. The individual gizzard was considered as the sampling unit in the statistical analyses. In order to test the effect of study area, age (adult or young) and sex on diet composition and diversity we conducted two statistical approaches. First, we pooled all food classes in four main categories to avoid groups with very low FO (< 5 %; e.g. fruits, reptiles and mammals). The four categories were: invertebrates (arthropods and gastropods), cereal seeds, vegetal (encompassing fruits and other vegetal material, see below) and vertebrates (eggs, birds, reptiles and mammals). We used multivariate analysis of variance (MANOVA) with the VOL of each main food category as response variables and the study area, age and sex and all interactions

51

between them as fixed factors. Using these main categories, we calculated diet diversity of each gizzard using the Shannon diversity index (

= ∑

lg

). Differences in

′ were tested using General Linear Models (GLM), which included the same factors as in MANOVA. Second, we assessed the factors explaining the consumption of the principal food classes (FO ≥ 5%) present in both study areas (arthropod, cereal seed, other vegetal, gastropod, bird and egg). For this, we performed independent GLMs with the VOL of each food class as dependent variable and study area, age, sex and all interactions as fixed factors. A negative relationship between a given food and dietary diversity is usually interpreted as indicative of trophic specialization (Futuyma and Moreno 1988). We tested whether magpies specialize on any food class through Pearson´s correlation analysis between the principal food classes (FO ≥ 5%) and H’. Prior to statistical analyses, the VOL for each food class and H’ values (dependent variables) were log (x+1) transformed to achieve normality (Zar 1984), which was assessed visually from normal probability plots of residuals. All statistical analyses were performed using Statistica 10.0 software (Statsoft INC 2011) and the significance level was set at α = 0.05.

Results A total of 118 gizzards were collected and analyzed in the laboratory, achieving a similar sample size for each study area (61 from Area 1, 57 from Area 2), age (51 adult, 67 young), and sex (48 females, 70 males). Overall, we identified 1016 food items in the gizzard contents belonging to 26 taxonomic groups (Table 2.11 and Appendix 2.1.). Diet composition Magpies consumed a wide range of food items among which arthropods and cereal seeds were the most frequent classes (total FO of 94.07% and 66.95% respectively), followed by other vegetal (FO of 33.90%) and birds (FO of 16.95%). Other food classes (gastropods, mainly small snails, bird eggs, fruits, mammals and reptiles) were present in lower FO (< 10%, Table 2.1. and Appendix 2.1.). Coleoptera and formicidae species represented 90% of the items consumed among the artrhropoda (Appendix 2.1.). We were able to identify 84% of the seeds found in the gizzards, and most of them 52

corresponded to Hordeum sp. (64%), Avena sp. (27%) and Triticum sp. (9%) (Appendix 2.1.). The “other vegetal” class was composed mainly by grass stalk and leaves of unidentified herbaceous plants, likely from cereal crops. We only could differentiate bird remains to the taxonomic order level by the microscopic structure of feathers (Day 1966). Most bird remains belonged to passeriformes (n = 15), and only one of them corresponded to galliformes (Appendix A). Bird egg remains always appeared highly fragmented, making very difficult the identification of the species that had produced them. Nevertheless, according to the thickness of eggshells, four (< 0.09 mm) were compatible with eggs produced by small birds (likely passeriformes), one (0.14 mm) with those of doves and one with those of partridges (0.23 mm, Herranz 2000). The rest of vertebrate prey items were remains of two Apodemus sylvaticus, hairs of one Felis sp., and one undetermined mammal and reptile species, respectively (Appendix 2.1.).

Table 2.1. Magpie diet composition in central Spain. For each food class, we present the number of gizzards containing remains (Gizzards), the frequency of occurrence (FO) and the average % volume (VOL). Data is independently presented in terms of overall magpie diet (Total) and for each study area (A1 and A2). More detailed data on diet composition are shown in the Appendix 2.1.

Arthropoda

Total (n = 118) 111

Gizzards A1 (n = 61) 56

A2 (n = 57) 55

Gastropoda

11

10

1

Cereal seeds

79

43

36

Fruits

5

5

0

Other vegetal

40

27

13

Eggs

6

5

1

Birds

20

17

3

Mammals

4

4

0

3.39

Reptiles

1

1

0

0.85

Food class

FO Total

A1

VOL A2

94.07 91.80 96.49 9.32 16.39 8.20

5.08

3.07

A2

5.89

0.05

36.10 36.43 35.75

0.00

33.90 44.26 22.81

A1

41.14 29.16 53.96

1.75

66.95 70.49 63.16 4.24

Total

1.55

3.00

0.00

10.75 16.20

4.93

8.20

1.75

2.63

3.61

1.58

16.95 27.87

5.26

3.87

5.90

1.70

6.56

0.00

0.07

0.13

0.00

1.64

0.00

0.21

0.41

0.00

Influence of locality, age and sex on diet composition and diversity Our first approximation showed that overall diet varied significantly between study areas and that there was a statistically significant effect of the sex-area interaction, and a marginal statistical effect of the interaction sex-age on diet variation (Table 2.2.). Only 53

VOL of seeds did not differ between localities (Tukey post-hoc, Appendix 2.2.). Males fed similarly in both areas but females from Area 1 fed more on vegetal and less on invertebrates than females from Area 2 (Tukey post-hoc; Appendix 2.3.).

Table 2.2. Results of MANOVA using the four main food categories as response variables: invertebrates (arthropod and gastropod), cereal seeds, vegetal (encompassing fruit and other vegetal material, see below) and vertebrates (egg, bird, reptile and mammal) and three fixed factors (Study Area, Age and Sex) and all possible interactions. Statistically significant variables are highlighted in bold and marginally significant ones in italic. Variables Study Area Sex Age Study Area*Age*Sex Study Area*Sex Study Area*Age Sex*Age

Value 0.75 0.94 0.98 0.97 0.88 0.98 0.92

F4, 107 9.15 1.71 0.68 0.82 3.48 0.63 2.38

P-value < 0.001 0.154 0.609 0.513 0.010 0.645 0.056

Significant differences in the consumption of principal food classes were mainly related to the study areas (Table 2.3.). Magpies consumed more arthropods, less other vegetal, less gasthropods and less birds in Area 2 than in Area 1 (Figure 2.1.). The only significant difference due to sex was a larger consumption of other vegetal by females (mean ± se: 15.11±2.93) than males (7.89±2.39). The interactions between sex and area significantly affected the consumption of arthropods (Figure 2). The effect of the interaction sex-age on the consumption of other vegetal group VOL was statistically significant (Table 2.3.; Figure 2.3.).The consumption of bird eggs was not significantly affected by any of the factors considered. Magpie diet was significantly more diverse in Area 1 than in Area 2 (Table 2.3.; Figure 2.4.), while sex, age and interactions did not represent significant differences in diet diversity (Table 2.3.). H’ was significantly and positive correlated with the VOL of cereal seeds and other vegetal material (Pearson´s correlation: 0.36 and 0.39 respectively; p < 0.05). VOL of arthropods was significantly and negatively correlated with VOL of cereal seeds, eggs, birds and vegetal groups (Pearson´s correlation: -0.40, 0.33, -0.24 and -0.20 respectively; p < 0.05). 54

Table 2.3. Results of the General Linear Models (GLMs) performed to assess the effect of different factors on the consumption of the principal food classes (FO ≥ 5%) by magpies and on diet diversity (H´). Degrees of freedom were 1,110 in all F tests. Statistically significant variables are highlighted in bold.

Variables Study Area Sex Age Study Area*Sex*Age Study Area*Sex Study Area*Age Sex*Age

Diet Diversity (H´) F p 16.04 0.17 1.17 0.01 0.02 1.98 2.13

0.014 0.677 0.280 0.888 0.874 0.874 0.147

Arthropoda F p 25.35 5%) consumed by magpies in both study areas. *: Statistically significant differences; NS: nonsignificant differences.

Figure 2.2. Variation in the percentage of volume of arthropods (VOL) consumed by magpies (mean±SE) in function of the study area and sex. *: Statistically significant differences between pair of means (Tukey’s post-hoc test); NS: non-significant differences (Tukey’s post-hoc test).

56

Figure 2.3. Variation in the percentage of volume (VOL) of vegetal food consumed by magpies (mean±SE) in function of age and sex. *: Statistically significant differences between pair of means (Tukey’s post-hoc test); NS: non-significant differences (Tukey’s post-hoc test).

Figure 2.4. Differences in magpie diet diversity (H´; means±SE) between study areas.

57

Discussion Our findings show that, during the breeding season, magpies fed on different food types, with varying importance between localities, and that the most frequently consumed food classes were cereal seeds and arthropods. This is in agreement with previous studies conducted in Spain, which indicated that, although both food classes are consumed throughout the year, the consumption of invertebrates increases during the breeding season, when their availability is higher (Soler and Soler 1991; Martínez et al. 1992; Herranz 2000). Magpie predation on eggs and birds Eggs were detected in a low proportion and volume in magpie gizzards (< 6%), in accordance with most previous studies (Birkhead 1991). A higher occurrence of eggs in magpie diet has been recorded in a previous study conducted in central Spain (FO = 1320 %; Herranz 2000); a large proportion of these were attributed to red-legged partridges (77-80 %). In contrast, only one of the egg remains found in our study (17%) coincided with the partridge egg thickness. This suggests that partridge eggs do not represent an important food for magpies during the breeding season in the study areas. However, several studies conducted in the Iberian Peninsula have shown that magpies are one of the main predators of dummy partridge nests (Herranz 2000; Blanco-Aguiar et al. 2001; Ferreras et al. 2010). From this perspective, we cannot discard that magpie nest predation could represent a risk for partridge breeding success in a scenario of high magpie abundances and low partridge densities, where even a small number of partridge eggs predated by each magpie could represent a large impact on the breeding success of the partridge population. In addition, partridge nest predation by magpies may be underestimated in diet studies, which hardly detect remains of predated eggs, i.e., eggshells (Chiron and Julliard 2007). This is probably because magpie behaviour of egg predation and ingestion varies with egg size. Thus, while smaller eggs are entirely swallowed, including the eggshell, larger ones are broken and only the egg content and small eggshell pieces are swallowed (Suvorov et al. 2012), decreasing the likelihood of eggshells ingestion. We found a relatively high consumption of passerines (12.7 % FO) compared with data reported in other studies performed during the breeding season (FO < 8 %; Birkhead 1991; Herranz 2000; Kryštofková et al. 2011). It has been suggested that magpie 58

predation on breeding birds may be related to high bird densities (Birkhead 1991). However, Fernández-Juricic et al. (2004) found that magpie predation on birds was opportunistic and was mainly observed during the breeding season, regardless of bird abundance. Magpies might increase their predatory pressure on birds when invertebrates, the principal animal component of their diet, are less available. Sources of magpie diet variation and consumption of other food groups The consumption of the other main food groups, except cereal seeds and eggs, varied between localities. This pattern was potentially related to food availability, as suggested by the similar consumption of cereal seeds between areas, which had similar cereal crop surfaces. Nevertheless, we must be cautious with this interpretation for two reasons. First, we did not have data about the availability of the other food groups, and second magpies can select food items independently to their availability; e.g. some invertebrate groups (Martínez et al. 1992; Kryštofková et al. 2011). Alternatively, differences in the consumption of arthropods, birds and eggs between areas may be explained by the different breeding stages when samples (i.e. gizzards) were collected: during the incubation stage in Area 1 and during the stage of nestling feeding in Area 2. In this sense, Suvorov et al. (2012) showed that magpies predated dummy nests more frequently at the incubation stage than during the stage of nestling feeding because during this stage magpies select invertebrates to feed nestlings (Martínez et al. 1992). This may also explain the lower diet diversity found in Area 2. During our study an important proportion of young magpies were also reproductive (all captured young females showed brood patch, indicating they were breeders), and therefore this may explain that young magpies presented a similar feeding behavior to adults. Globally, we did not find differences in diet composition associated with age and sex, and only the interaction between the locality and these intrinsic factors significantly affected magpie diet. During the breeding season males regularly feed females (Buitron 1988), so it would be expected that the diet was similar between sexes. However, we observed that adult females included in their diet significantly more vegetal food than adult males. Breeding females spend most of the time in the nest during incubation and hatching (Buitron 1988), where vegetal food, which they can easily consume, is probably more available, supplementing food provided by males. Also, female magpies consumed more arthropods than males in Area 2. During the nestling feeding stage

59

males increase the supply of food to the female and chicks (Buitron 1988), being invertebrates the main food brought to chicks (Martínez et al. 1992; Ponz et al. 1999). In this sense, the male probably reduces the consumption of invertebrates in order to provide most of their catch to the nest. Magpie diet diversity Our results indicate that magpies do not specialize in any food during our study since diet diversity was not related negatively to the occurrence of any of the main food classes (Futuyma & Moreno 1988). In contrast, diet diversity was positively related to the amount of cereal seeds and other vegetal in the diet. This suggests that magpies need to supplement their diet including many different animal food types, although it is predominantly vegetarian. Invertebrates are the principal contribution of protein in a large number of birds (Capinera 2010), including magpies in agricultural landscapes within central Spain. Arthropods consumption was negatively associated with the consumption of other animal sources of proteins, such as birds or eggs, suggesting that these may be a secondary and occasional source of protein for magpies during the breeding season (Birkhead 1991). Conclusions Overall we found no evidence that magpies pose a threat to the conservation of birds since magpies include in their diet eggs and birds in a low proportion, regardless of the age and sex of magpies. However, the possible sources of bias associated with our study methodology, such as the quantification of these bird remains and eggs, as well as the fact that even a low rate of predation may affect a prey when the predator is abundant, make us to be cautious with this conclusion. Thus, more complex and experimental studies at larger time-spatial scales are necessary, including localities with different densities of magpies and potential bird prey. Diet data should be complemented with the monitoring of the abundances of potential bird prey species and magpies, prey breeding success and predation rate of magpies on nests, chicks and adults birds.

Acknowledgements We are very grateful to land owners and game managers who allowed us to work in their hunting estates. We thank people who assisted us during the fieldwork, especially S. Luna and L.E. Minguez. We acknowledge Dr. J.T. García and Dr. E. Pérez-Ramírez for necropsy and sexing of magpies. This study was funded by Consejería de Medio 60

Ambiente of Junta de Comunidades de Castilla-La Mancha (Project PREG-05-23). M. Delibes-Mateos is currently supported by a JAE-DOC contract funded by CSIC and the European Social Fund.

Ethical standards This work was performed in compliance with current Spanish legislation, and follows the European Union’s recommendations regarding animal welfare. All procedures were carried out with all legal permits required by the concerned administrations.

61

CAPÍTULO 3: An evaluation of cage-traps and the Collarum device to capture red foxes (Vulpes vulpes). Can the performance of cage-traps be improved by baits and scent attractants?

Este capítulo se encuentra en preparación para ser enviado a una revista SCI:

Díaz-Ruiz F, Delibes-Mateos M., Ferreras P (en preparación) An evaluation of cagetraps and the Collarum device to capture red foxes (Vulpes vulpes). Can the performance of cage-traps be improved by baits and scent attractants?

62

Abstract Carnivore predation on prey of human interest, such as game species or livestock, leads frequently to the lethal control of predators. This constitutes a serious conservation problem in many places across the world, since non-target species of conservation concern are frequently removed. In Spain, cage-trapping is one of the most widespread methods used by hunters to control red foxes (Vulpes vulpes), although its low efficiency and selectivity have been frequently reported. From this perspective, these control methods need urgently to be improved, and its performance compared to that of new alternative devices, such as the Collarum restraint device. The aim of this study were to test whether the use of different baits and scent attractants may improve the selectivity and efficiency of cage-traps, to compare the performance of different cagetraps designs with that of the Collarum restraint device, and to analyse the injuries caused by both methods to captured animals. Fieldwork was conducted in three study sites in central Spain during 2003 and 2006/07. We tested the effect of two types of baits (dead or alive), four scent attractants, and their combinations on the efficiency and selectivity of three cage-trap types commonly used to control foxes in Spain. During 2006/07, we also compared the Collarum restraint device with cage-traps in terms of efficiency and selectivity. Injuries caused to animals by both capture methods were also described. Cage-traps captured a total of six foxes and 40 individuals of 13 non-target species, including protected carnivores and raptors, with an overall effort of 2068 trapnights. The use of live baits and fox urine increased the efficiency of cage-traps independently of the cage-trap type. In addition, the capture rate of non-target animals was lower with cage-traps with chamber for bait adjacent to the capture chamber and with traps of one capture chamber. It was also slightly lower using valerian scent as attractant. The Collarum restraint device was more selective (50-100%) than cage-traps (12-29%) and more efficient than cage-traps without attractant, but as efficient as cagetraps with attractants. Animals captured with both types of traps showed no indicator of poor welfare. Our results suggest that live baits and scent attractants may improve the efficiency and selectivity of cage-traps for capturing red foxes. Even so, non-target species, including some protected ones, can be still captured, and selectivity levels are still very low (0-21%) and therefore the use of this method is not recommended for managing foxes in Spain. The Collarum restraint device may be an acceptable selective alternative to traditional methods in areas with similar carnivore composition than that

63

existing in our study sites. Further studies are necessary to test the selectivity in other areas with different composition of carnivore communities. Although our results show that the selectivity of trapping methods can be improved, the decision of releasing captured non-target animals depends ultimately on the trapper. For this reason, it is of key importance that fox management is carried out by skilled technical personnel and always supervised by wildlife competent authorities. Keywords: red fox, cage-traps, Collarum restraint device, capture efficiency, selectivity, predator control, game management.

64

Introduction Lethal control of predators is widespread all over the world (Treves and Karanth 2003; Woodroffe et al. 2005), because humans usually see these species as competitors for shared, limited resources, such as game species (e.g. Valkama et al. 2005) or livestock (e.g. Treves et al. 2004; Sangay and Vernes 2008). Intensive predator removal has caused the local extinction of several species of conservation concern, and massive contractions of the geographic ranges of many others (e.g. Whitfield et al. 2003). Methods of predator control may result in the death of protected species. On the one hand, some legal methods are not selective, and therefore non-target protected species are captured (e.g. Duarte and Vargas 2001; Way et al. 2002). On the other hand, come managers employ illegal, unselective methods, such as poisoning (e.g. Whitfield et al. 2003; Márquez et al. 2012), based on their belief that legally permitted methods are not efficient to reduce predator numbers (Delibes-Mateos et al. 2013). The removal of predator species of conservation concern causes frequent clashes between conservationists and hunters and game managers (Virgós et al. 2010). In biodiversity conflict management, success occurs when the outcome is acceptable to both sides and when neither party is asserting its interests to the detriment of others (Redpath et al. 2013). Under this perspective, banning totally predator control would not be the best way to minimise conflicts between hunters and conservationists in relation to predator management. In this regard, finding efficient and selective control methods to legally reduce the numbers of some generalist/opportunistic predators could help to reduce these tensions between hunters and conservationists. In Spain, hunting is a very important socioeconomic activity and one of the most important leisure rural activities; thus, >77 % of the territory is covered by hunting estates (Rios-Saldaña 2010; Arroyo et al. 2012). Hunters and game managers employ several different management tools, including predator control, to boost game species numbers (see Angulo 2003; Arroyo et al. 2012). The use of predator control is widespread in some Spanish regions (Ríos-Saldaña 2010; Díaz-Ruiz and Ferreras 2013). For example, in central Spain most small-game estates (~ 90%) use some type of predator control (Delibes-Mateos et al. 2013). The main predators legally controlled in Spain are red foxes (Vulpes vulpes), feral cats (Felis catus) and feral dogs (Canis lupus familiaris), among carnivores, and magpies (Pica pica), among birds. Nevertheless, the detrimental effect of illegal predator control on some protected species of conservation 65

concern, including raptors and carnivores, has been frequently reported (e.g. Villafuerte et al. 1998; Márquez et al. 2013). Spanish hunters argue frequently that the current legal predator control methods are inefficient, and therefore they request more effective methods to cull predators, and especially red foxes (Delibes-Mateos et al. 2013). For example, cage-traps, which are one of the most frequently employed methods to legally control foxes are usually considered as inefficient by Spanish hunters (Delibes-Mateos et al. 2013). In fact, the efficiency of cage-traps to capture foxes in Spain is extremely low; capture rate ranges between 1.2 and 5 foxes per 1000 trap-nights, and levels of selectivity are far from acceptable (Herranz 2000; Duarte and Vargas 2001; Muñoz-Igualada et al. 2008). Given that this is neither acceptable for conservationists (selectivity) nor for hunters (efficiency), it is urgent to explore possibilities of improving both the efficiency and selectivity of cage-traps in Spain. For example, some scent attractants could be used to achieve this goal, since not all the species respond equally to different scent attractants (Monterroso et al. 2011). In addition, Iberian predators show different feeding strategies; some species feed exclusively on live prey (e.g. the European wildcat (Felis silvestris); Lozano et al. 2006), while others can also scavenge (e.g. red fox; Díaz-Ruiz et al. 2013). This suggests that the probability of capturing different species could change in function of the type of bait (alive or dead) used. To our knowledge, only Herranz (2000) previously tested for differential attraction effects using both dead and alive baits in Spain, but this author did not evaluate any scent attractant. Methods alternative to cage-traps have been used to capture other canids with success in terms of efficiency, selectivity and injuries to both target and non-target. For example, the Collarum restraint device (hereafter Collarum), a powered neck snare designed to the live capture of canids (see Shivik et al. 2000), has shown up to 87% efficiency for coyotes (Canis latrans; Shivik et al. 2005). In Spain, the Collarum has been tested for capturing foxes only in two studies developed in northern and southern Spain respectively (Muñoz-Igualada et al. 2008; Andalucía 2010). These studies consider the Collarum as highly selective and its efficiency as higher than that of traditional cagetraps (Muñoz-Igualada et al. 2008), but still far from the efficiency obtained for coyotes (Shivik et al. 2005).

66

In this paper, our goals were: 1) to evaluate the efficiency and selectivity of different types of cage-traps traditionally used for the capture of red foxes in Spain; 2) to test whether the use of different baits and scent attractants improve the selectivity and efficiency of different cage-traps types; 3) to compare the performance of cage-traps and the Collarum restraint device in terms of efficiency and selectivity; and 4) to describe the injures caused to foxes and non-target species by both capture methods.

Material and Methods Study areas Fieldwork was performed in 3 sites of central Spain (Ciudad Real province), one private and two public estates, during 2003 and 2006/2007 (Table 3.1.). The climate was typical Mediterranean characterised by wet, mild winters and warm, dry summers with a marked drought period (Rivas-Martínez et al. 2004). The landscape was similar between study sites i.e. Mediterranean scrubland (mainly Cistus spp. in combination with holm oak (Quercus ilex) forests), mixed with cereal croplands, riparian habitats, ‘dehesas’ (pastureland with savannah-like open tree layer, mainly dominated by Mediterranean evergreen oaks) and pine (Pinus spp.) plantations (Table 3.1.). Study sites selection was based on three criteria: 1) a high habitat heterogeneity that favoured the presence of a diverse wildlife community, including both prey and predators, 2) a medium-high red fox abundance, which allowed us to test trap efficiency for capturing foxes, and 3) a high diversity of other potentially capturable predators, including protected ones, which allowed us to asses trap selectivity. The three study sites were situated in the distribution area of several Iberian terrestrial carnivores such as European wildcat, stone marten (Martes foina), small-spotted genet (Genetta genetta) and Eurasian badger (Meles meles); the Egyptian mongoose (Herpestes ichneumon) was also present in Site 1 (Palomo et al. 2007). Our study sites also held raptors, such as common buzzards (Buteo buteo), goshhawks and sparrowhawks (Accipiter sp.), Bonelli's eagles (Aquila fasciata), Spanish Imperial eagles (Aquila adalberti), Golden eagles (Aquila chrysaetos) or Eagle owls (Bubo bubo) (Martí and Del Moral 2003).

67

Table3.1. Description of study sites. The geographical location, the year and season when trapping was performed are shown.

Study site

Location

Year (Season)

Site 1

38˚27´40´´N 3˚34´5´´ W

2003 (SeptemberNovember)

Site 2

38˚ 58´ 2´´ N 4˚ 8´47´´ W

Site 3

39˚ 0´ 2´´ N 4˚ 23´55´´ W

Area (ha)

Main habitat types

Main land uses

3700

Pine plantations (Pinus pinaster), Managed publicly for forestry Mediterranean scrub (Cistus sp.), holm oak production and big game forest (Quercus ilex), and cereal crops

2006 (July-December)

1000

“Dehesa” (a typical Mediterranean formation of sparse oaks and underlying cereal crops), Managed privately for livestock, holm oak forest with Mediterranean scrub, and cereal agriculture, and big game riparian vegetation

2006/2007 (November-March)

1500

“Mixed” forests of pine (Pinus sp.) and holm Managed publicly for forestry, and oak with Mediterranean scrub. small and big game

68

The presence of these species and foxes was previously confirmed by the technical staff of the public estates (i.e. Sites 1 and 3; Junta de Comunidades de Castilla-La Mancha, unpublished data), by colleagues from our Institute (Site 2, P. Acevedo, unpublished data), and during the revision of the traps in this study. Since no direct measure of the abundance of potentially capturable species was available, we performed nocturnal spotlight-counts chiefly to estimate fox relative abundance (kilometric abundance index, KAI) at the beginning of the study and during the trapping season (Ruette et al. 2003). KAIs estimated were apparently higher in site 2 (0.26 foxes km-1, 42.4 km surveyed) than in site 1 (0.016 foxes km-1, 60 km surveyed), and in site 3 (0.02 foxes km-1, 66.4 km surveyed). The European wildcat (Felis silvestris) was only observed in Site 1, (0.016 wildcats km-1), and feral cats in Site 3 (0.03 km-1). Efficiency and selectivity definitions of control methods We used the parameters described previously by the International Organization for Standardization (1999) for testing restraining traps for mammals. The number of foxes captured per 1000 trap-nights was used to assess trapping system efficiency. We evaluated two parameters related to the selectivity: direct selectivity, or the percentage of foxes captured related to the total number of animals captured (including red foxes), and the non-target capture rate, or number of non-target captures per 1000 trap-nights (inversely related to selectivity). Trap types evaluated We used three types of cage-traps used in Spain for capturing foxes. These types had one or two capture entrances that employed a guillotine-type door and a tread trigger system, differing in design details. CT01 type had one entrance and one capture chamber; used exclusively with dead baits placed in the capture chamber, CT02 type had two entrances, one capture chamber and a lateral bait chamber and CT03 type had two entrances, two capture chamber and a central bait chamber (see Appendix 3.1.). Some of these types included different commercially available models that slightly differed in their characteristics, such as measures or mesh size, as described in Appendix 3.1. The Collarum neck restraint device is a specific trap to selectively capture canids, such as coyotes, foxes, dogs and dingoes (Canis lupus dingo) (Shivik et al. 2000, 2005). It

69

uses a baited pull-tab that triggers a pair of coil-spring powered throw-arms that propels a cable loop over the head onto the neck of a fox (Muñoz-Igualada et al. 2008). We tested the commercially available red fox version (Wildlife Control Supplies, East Granby, Connecticut, USA). The Collarum traps were tested in Sites 2 and 3. Baits and scent attractants We tested two types of bait (dead or alive) for possible effects on the efficiency and selectivity of cage-traps. Chicken carcasses and lamb meat were used as dead baits. Common quails (Coturnix coturnix), red legged partridges (Alectoris rufa), and helmeted guinea fowl (Numida meleagris) were used as live baits. Dead baits were placed inside the traps secured with wire to avoid that animals took them away, and they were weekly replaced. Live baits were placed in an independent chamber that was adjacent to, or inside the trap, depending on the trap model (see Appendix 3.1.). We also tested the effect of scent attractants on cage-trap efficiency and selectivity. We tested four types of scent attractants previously used to attract red foxes (Saunders and Harris 2000; Monterroso et al. 2011): red fox urine (hereafter FU), valerian-extract solution (hereafter VAL), containing valeric acid found in urine and anal-sac secretions of fox (Albone and Fox 1971; Jorgenson et al. 1978), fatty acids scent (hereafter FAS), a mixture of seven volatile fatty acids found in fermented eggs (Roughton 1982), and Collarum canine bait (Wildlife Control Supplies, East Granby, Connecticut, USA; hereafter COLL), a commercial canids-specific attractant. Scent attractants were impregnated on a piece of chalk tied to an iron stick with elastic bands, driven to the ground inside the cage-trap. The chalk was moistened with the attractant (1-5 cc) with the help of a syringe and was replenished every 3-4 days. Dead and live baits were tested in all study sites, but only FU and VAL scent attractants were used in all study sites. FAS was tested in Sites 2 and 3, and COLL only in Site 3. Moreover, traps without any scent attractant were used as control in all the localities. We followed a block design in each study site, with the treatment randomly assigned to each trap within a block, regardless of the trap type. Three treatments were simultaneously tested in Site 1: control, FU and VAL. Four treatments were tested in Site 2, control and three scents (FU, VAL and FAS) being simultaneously deployed after an initial period with the control treatment in all the traps. Five treatments were simultaneously tested in Site 3: control, FU, VAL, FAS and COLL. The minimum 70

distance between neighbouring traps was 100 m. Traps were placed near shrubs or other resources that increase the probability of animal presence (e.g. ponds, water courses, edges of dense vegetation, etc.). Handling of animals and injuries All captured animals were examined in situ by a wildlife veterinarian for possible traprelated injuries. For veterinarian inspection, both foxes and non-target carnivores captured were immobilized with a combination of Ketamine hydrochloride (50 mg/ml, Imalgene ® 50, Merial) and Xylazine hydrochloride (20 mg/ml Rompun®, Bayer); this was injected intramuscularly in the animal's hindquarters, using recommended doses for small and medium size carnivores (15 mg Ketamine + 1-1.5 mg Xylacine per Kg; Seal and Kreeger 1987). To do so, animals were transferred from the cage-trap to a squeeze cage that allows their physical immobilization, and prevent damage to both them and the veterinary (Ferreras et al. 1994). Non-target carnivores were marked with a subcutaneous transponder (ID-100, Trovan®), which allowed their identification in case of recapture. The drug effect was reversed using Yohimbine (0.15 mg per Kg; Seal and Kreeger 1987). Once fully recovered from anesthesia and after the veterinary checked that no serious injuries compromised their survival, animals were released in the capture site (Harris et al. 2006). A correct evaluation of injuries caused by trapping methods to target species requires the examination through a post-mortem necropsy of at less 20 captured animals (European Union-Canada-Russian Federation 1998; United States of America-European Community 1998; International Organization for Standardization 1999). In our study, the number of captured foxes was