Ecología y gestión de depredadores generalistas: el caso del zorro
Transcription
Ecología y gestión de depredadores generalistas: el caso del zorro
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 =−∑ logpi, 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 <0.001 1.40 0.239 0.48 0.491 0.26 0.614 7.98 0.006 0.85 0.358 0.88 0.351 Cereal Seeds F p 0.34 1.83 2.19 0.06 1.51 0.05 0.88 0.560 0.179 0.142 0.812 0.222 0.820 0.351 55 Vegetal F p 12.67 7.11 0.37 1.33 2.50 0.05 4.90 0.001 0.009 0.543 0.252 0.116 0.823 0.029 Gastropoda F p 5.87 0.05 0.05 1.80 0.28 0.01 2.74 0.017 0.830 0.826 0.183 0.598 0.927 0.101 Birds Eggs F p F p 6.26 0.91 0.08 1.08 0.07 0.60 2.34 0.014 0.342 0.773 0.301 0.793 0.438 0.129 0.93 0.03 0.03 0.39 0.77 1.49 2.81 0.336 0.860 0.867 0.535 0.382 0.224 0.097 Figure 2.1. Percentage of volume (VOL; mean±SE) of the principal food classes (FO > 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 <20, and only three foxes were euthanized; the others were kept in captivity for subsequent behavioural experiments. Therefore, we only were able to show a descriptive list of the injuries observed by the veterinary in situ. Injuries were recorded according to the four categories established in the international scale of traumas: mild trauma, moderate trauma, moderate-severe trauma and severe trauma (International Organization for Standardization 1999). The method chosen for euthanizing foxes was the intravenous injection of T61 ® (Intervet), which is the method recommended for euthanizing dogs and cats due to its high speed, efficiency, ease of use and safety (Close et al. 1996; 1997; Gómez-Villamandos 2000). All 71 procedures were performed following approval by the competent authority (Castilla-La Mancha Regional Government). Statistical Analyses Cage-traps, baits and attractants Generalized Linear Mixed Models (GLMMs) were employed to analyze the effect of cage-trap type, and different baits and scent attractants on the efficiency and selectivity of the cage-traps. The individual trap with each combination of baits and attractants was utilized as the sample unit. In this analysis, the number of foxes captured in 1000 trapsnights was used as a measure of efficiency (Muñoz-Igualada et al. 2008). Since direct selectivity could not be calculated in many individual traps that produced no capture at all, non-target capture rate was used as a measure inversely related to selectivity (Muñoz-Igualada et al. 2008). Due to violations of normality and variance homogeneity of standardized residuals, dependent variables (i.e. efficiency and non-target capture rate) were square-root (x+1) transformed (Muñoz-Igualada et al. 2008). FAS and COLL scent attractants were excluded from these analyses because they were not employed in all study sites. Fixed factors included as explanatory variables in these models were: cage-trap type (CT01, CT02 and CT03), bait type (dead or alive), scent attractant (control, FU and VAL), and the interaction between bait and scent attractants. Study site was included as a fixed factor because number of levels was not enough to be considered as a random variable (Zuur et al. 2007). Since a given trap received several different treatments (bait x attractant), trap location (a categorical variable identifying each trap position in the fieldi.e. trap id) was included as a random variable in the models. Collarum vs. cage-traps Differences in the efficiency and non-target capture rate between cage-traps and Collarum were tested in Sites 2 and 3. In order to simplify the analysis we grouped cage-traps in those with and without scent attractant. Generalized Linear Mixed Models (GLMMs) were developed to test for differences between cage-traps and Collarum in terms of efficiency and non-target capture rate. The type of trapping device (CT_control, CT_attractant or Collarum) and the study site (Site 2 and 3) were included in these models as fixed factors. Trap location was included as a random effect. 72 All statistical analyses were performed using the lme4 package of the R statistical software (Bates and Maechler 2010; R Core Development Team 2013). The models were obtained with the function dredge of MuMin package (Barton 2012) and compared through the AICc criterion (Burnham and Anderson 2002). The coefficients of predictor variables were calculated through model-averaging (Burnham and Anderson 2002). We present the coefficient of variables resulting from the model-averaging for all models with a total cumulative weight of at least 90% (Arnold 2010). Results Overall captures A total effort of 3359 trap-nights produced the capture of 9 red foxes and 41 non-target animals of 13 species. Cage-traps captured a total of 6 foxes and 40 non-target animals, including carnivores, raptors, corvids and other game species, with an overall values of efficiency and selectivity of 2.9 foxes/1000 trap-nights and 13 % respectively (Table 3.2.). Collarum traps captured 3 foxes and one non-target animal, with overall values of efficiency and selectivity of 2.3 foxes/1000 trap-nights and 75 % respectively (Table 3.2.). Baits and scent attractants Cage-traps captured foxes with both types of baits, and using all attractants except with COLL and FAS (Figure 3.1.). Model-averaging for efficiency included eight models with a total cumulative weight of 90% (Table 3.3.). The explicative variables in order of relative importance were study site (0.96), scent attractant (0.78), bait (0.66), the interaction bait*scent attractant (0.47) and trap type (0.10). Capture rate of red foxes was higher in Site 2 (mean±SE: 18.51±13.11 foxes/1000 trap-nights) than in the two other sites (Site 1: 0.83±0.83, and Site 3 without captures; Table 3.4.). Only the interaction live bait*FU increased the efficiency of cage-traps to capture red foxes (Table 3.4.; Figure 3.1.). Non-target species were captured using cage-traps with both types of bait and all the attractants except COLL (Table 3.3.). Model-averaging for non-target captures included eight models with a total cumulative weight of 91% (Table 3.3.). The explicative variables in order of relative importance were trap type (1), scent attractant (0.84), bait (0.52), study site (0.30) and the interaction bait*scent attractant (0.26). Cage-trap types 73 CT01 and CT02 had a lower non-target capture rate than CT03 type, and VAL attractant produced a slightly lower capture rate of non-target species than FU and control (Table 3.4.; Figure 3.1.). 74 Table 3.2. Number of animals captured in the three study sites using cage-traps and Collarum devices, and the selectivity and efficiency of both methods. The total sampling effort, estimated as the number of trap-nights, is shown. Efficiency is the number of foxes captured in 1000 trap-nights. Selectivity is the percentage of foxes captured related to the total number of animals captured (included red foxes). Non-target CR is capture rate refers to the number of nontarget animals captured per 1000 trap-nights. Vv: red fox, Fs: European wildcat, Gg: European genet, Mf: stone marten, Hi: Egyptian mongoose, Mm: European badger, Ag: goshawk, Fc: feral cat, Clf: dog, and Others: Black-billed magpie (n= 1), azure magpie (n= 1), wild boar (n= 2), red-legged partridge (n= 19) and European wild rabbit (n= 1). Non Target captures Trap type Study Site Site 1 Cage-traps Year 2003 Trap-Nights Vv Fs Gg Mf Hi Mm Ag Fc Clf Others Total Efficiency Non-target CR Selectivity 601 1 0 2 0 2 1 0 0 0 2 7 1.6 11.6 12.5 810 5 1 1 0 0 0 1 0 3 15 21 6.17 25.9 19.2 657 0 2 2 1 0 0 0 1 0 6 12 0 18.3 0 2068 6 3 5 1 2 1 1 1 3 23 40 2.9 19.3 13 362 1 0 0 0 0 0 0 0 0 1 1 2.8 2.8 50 929 2 0 0 0 0 0 0 0 0 0 0 2.2 0 100 Overall 1291 3 0 0 0 0 0 0 0 0 1 1 2.3 0.77 75 TOTAL 3359 9 3 5 1 2 1 1 1 3 24 41 Site 2 2006 Site 3 Overall Site 2 Collarum 2006 Site 3 75 Table 3.3. Best models explaining the efficiency of cage-traps 1) for capturing foxes, and 2) for capturing non-target animals using different baits and scent attractants. AIC values of the set of GLMMs included a total cumulative weight of at least 0.90. 1) Fox Model site+bait+attractant+bait*attractant site+attractant site site+bait+attractant site+bait site+bait+attractant+trap type+bait*attractant bait+attractant+trap type+bait*attractant site+bait+attractant+trap type d.f. 10 7 5 8 6 LogLike -160.6 -165.0 -167.7 -164.4 -166.9 AICc Delta AICc weight 344.1 0.00 0.33 345.5 1.45 0.16 346.2 2.12 0.11 346.6 2.54 0.09 346.8 2.75 0.08 12 -159.7 347.7 3.64 0.05 8 9 -165.2 -164.1 348.3 348.6 4.24 4.56 0.04 0.03 trap type+attractant bait+trap type+attractant+bait*attractant bait+trap type+attractant site+trap type+attractant site+bait+trap type+attractant+bait*attractant trap type site+bait+trap type+attractant site+trap type 7 10 8 9 -203.3 -199.8 -202.6 -201.6 422.1 422.6 423.0 423.5 0.00 0.54 0.94 1.48 0.24 0.18 0.15 0.11 12 -198.0 424.4 2.35 0.07 5 10 7 -206.9 -200.8 -205.0 424.6 424.6 425.4 2.49 2.56 3.35 0.07 0.07 0.04 2) Nontarget Table 3.4. Model-average coefficients and standard errors (SE) of the variables included in the models explaining the efficiency of cage-traps to capture red fox and non-target animals. The intercept includes Site 2, live bait, control attractant and CT03 cage-trap type. Red fox Non-target Parameter SE Parameter SE 2.19*** 0.58 7.45*** 0.98 Site 1 -1.22* 0.61 -0.53 1.13 Site 3 -1.58** 0.51 -0.36 0.85 Alive bait 0.15 0.61 0.47 0.89 FU 0.16 0.60 -0.91 0.83 VAL 0.18 0.65 -1.63# 0.90 Alive bait*FU 1.80* 0.91 -0.60 1.51 Alive bait*VAL 0.03 0.99 -0.11 1.64 CT01 0.15 0.78 -4.26*** 1.14 CT02 -0.12 0.53 -5.58*** 0.85 Variable Intercept *p<0.05; **p<0.01; ***p<0.001 # p= 0.07 76 Figure 3.1. Average capture rate (±SE) (indv./1000 traps-nights) observed of both foxes and non-target animals captured with cage-traps using each type of bait (dead and alive) and scent attractants (Control: without attractant; FU: fox urine; VAL: valerian-extract; FAS: fatty acids scent; COLL: Collarum canine bait). Data of captures using FAS and COLL as attractants are shown, although these were not included in GLMMs analyses. Cage-traps vs. Collarum Capture rate of red foxes using both types of devices was higher in Site 2 (14.51±10.01 foxes/1000 trap-nights) than in Site 3 (0.91±0.61 foxes/1000 trap-nights) (Table 3.5.). The efficiency for capturing foxes differed between cage-traps and Collarum (Table 3.5.; Figure 3.2.); the average red fox capture rate was 3.24±1.97 foxes/1000 trap-nights using the Collarum, 0.87±0.86 foxes/1000 trap-nights using cage-traps without scent attractants and 6.34±4.86 using cage-traps with scent attractants (Table 3.5.; Figure 3.2.). Additionally, the capture rate of non-target species was significantly lower with Collarum (0.45±0.44 captures/1000 trap-nights) than that obtained with cage-traps either with scent attractants (11.39±5.66 captures/1000 trap-nights) or without scent attractants (23.46±11.25 captures/1000 trap-nights) (Table 3.5.; Figure 3.2.). 77 Table 3.5. Model standardized coefficients and standard errors (SE) for cage-traps with and without scent attractants (CT_Attractant and CT_Control, respectively) vs. Collarum analysis of efficiency of red fox captures and non-target captures. Intercept includes Site 2 and CT_Control. Red fox Non-target Variable Parameter SE Parameter SE Intercept 1.94** 0.41 3.66** 0.73 Site 3 -1.40* 0.41 -1.12 0.76 CT_Attractant 0.72 0.43 -0.88 0.66 Collarum 0.74 0.50 -1.62* 0.83 *p<0.05; **p<0.01 Figure 3.2. Means and SE of the observed capture rates of foxes and non-target animals (indv./1000 trap-nights) using cage-traps with and without scent attractants (CT_Attractant and CT_Control, respectively) and the Collarum restraint device in sites 2 and 3. Injuries None of the tested traps caused any serious injury or the death (i.e. severe trauma in ISO scale) of the red foxes captured. However, 83% and 100% of foxes captured with cagetraps and Collarum, respectively, showed injuries, corresponding to nine mild or moderate traumas (International Organization for Standardization 1999; Table 3.6.). 78 Most of the non-target animals captured with the cage-traps showed no injuries (76 %; n= 31). In the rest of the individuals, seven categories of injuries were detected, and on most occasions these referred to mild or moderate traumas, with the exception of two cases of severe traumas (Table 3.6.). Only one azure magpie died as a consequence of severe trauma (i.e. neck fracture caused by the trap door; Table 3.6.). The Collarum device captured only one non-target species, a wild boar, causing it “major cutaneous laceration” and death (Table 3.6.). 79 Table 3.6. Observed injuries in animals captured using cage-traps and Collarum devices. Data were obtained through veterinarian “in situ” inspections. Injuries recorded in the list of the International Organization for Standardization (1999) are shown: *Severe trauma category; the other injuries showed are included into Mild or Moderate trauma categories. Vv: red fox, Fs: European wildcat, Gg: European genet, Mf: stone marten, Hi: Egyptian mongoose, Mm: European badger, Ag: goshawk, Fc: feral cat, Clf: domestic dog, Pp: black-billed magpie, Cc: azure magpie, Ss: wild boar, Ar: red legged partridge and Oc: European wild rabbit. Cage traps Collarum Vv Fs Gg Mf Hi Mm Fc Cld Pp Ag Ss Cc Ar Oc (n=6) (n=3) (n=5) (n=1) (n=2) (n=1) (n=1) (n=3) (n=1) (n=1) (n=1) (n=1) (n=19) (n=1) Vv Ss (n=3) (n=1) Injury Claw damaged Oedematous swelling or hemorrhage 0 1 1 0 0 1 1 0 0 0 0 0 0 0 0 0 2 1 1 0 0 0 1 0 0 0 0 0 0 0 3 0 Minor cutaneous laceration 1 2 2 0 1 0 0 0 0 0 0 0 0 0 0 0 Minor subcutaneous soft tissue maceration or erosion (contusion) 1 1 2 0 0 0 0 0 0 0 0 0 0 0 0 0 Major cutaneous laceration, except on foot pads or tongue Chipped or fracture of a permanent tooth without exposing pulp cavity Any other fracture (e.g. neck fracture)* 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 2 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 Death* 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 No injuries 1 1 1 1 1 0 0 3 1 1 1 0 19 1 0 0 80 Discussion Our findings confirm the low selectivity and efficiency of cage-traps to capture foxes in Spain, reported in previous studies (Herranz 2000; Duarte and Vargas 2001; MuñozIgualada et al. 2008), and for other canids in North America (Way et al. 2002; Shivik et al. 2005). The Collarum device has been reported as highly selective to catch canids; up to 100% of selectivity in some areas (Shivik et al. 2000, 2005; Muñoz-Igualada et al. 2008; Junta de Andalucía 2010). Our results are partly in agreement with this, since only foxes were captured in one of the study sites (100% selectivity), but in the other one a wild boar was caught (50% selectivity). Effects of cage-trap type, bait type and scent attractants on efficiency and selectivity Our results show that, although cage-traps captured foxes with both baits, live baits increased their efficiency. This is in agreement with the only study that had previously tested for differences in efficiency of cage traps to capture foxes using both baits (Herranz 2000). This increase in fox captures was even more noticeable when live bait was combined with fox urine as scent attractant. This is not surprising as urine is used by foxes for scent marking, and plays an important role in olfactory communication and territoriality (Macdonald 1979; Arnold et al. 2011). However, Monterroso et al. (2011) observed that captive foxes showed more interest in other attractants than in the urine of their conspecifics, and this together with the low number of foxes captured in this study suggest that further works are needed to confirm our finding. According to our results, the capture rate of non-target species differed between cagetrap types; it was lower using CT01 and CT02 models than using CT03 (see Appendix A), likely because the former were smaller in size, which could deter carnivores to enter inside the trap (Shivik et al. 2005). In addition, CT03 cage-trap type has the bait chamber in a central position, while the others present a chamber annexed to one side of the trap. This difference might also explain the increased capture rate observed for CT03 cage-trap. Non-target species were captured using both baits and all scent attractants, excepting COLL. Previous works conducted in Spain have shown that cage-traps baited with live animals capture higher numbers of non-target species, especially mammalian carnivores and raptors (Herranz 2000; Duarte and Vargas 2001; Muñoz-Igualada et al. 2008). 81 Similarly, in our study more non-target wild predators were captured using live bait (13.61 captures/1000 trap-nights) than dead bait (6.48 captures/1000 trap-nights). We observed a reduction in the capture rate of non-target species using VAL as attractant, which could be related to the avoidance behaviour of competitors (Harrington et al. 2009). Nevertheless, it could also increase the captures of some felids of conservation concern (Jerosch et al. 2010; Monterroso et al. 2011); in fact, we captured a European wildcat using this attractant. Cage-traps versus Collarum Cage-traps using baits have been reported as an effective method to capture foxes and medium-sized canids, such as coyotes (Baker 1998; Baker 2001; Way 2012), but no information of non-target captures is reported in these studies. In contrast, professional trappers of France preferred using cage-traps to catch medium-sized mustelids, considering other methods more efficient to capture foxes (Ruette et al. 2003). Our results showed how the efficiency and selectivity of cage-traps for foxes may be improved combining baits and attractants; however, the capture rate of non-target species remains high compared to Collarum, which showed an acceptable efficiency to catch foxes. The Collarum may constitute a good alternative to other predator control methods, such as cage-traps, passive neck-snares, leg-snares, leg-hold traps, or foothold-traps, which are less selective regardless of their efficiency (Travaini 1996; Fleming et al. 1998; Way et al. 2002; Shivik et al. 2005; Muñoz-Igualada et al., 2008, 2010; Duarte et al. 2012). Although we captured one wild boar using the Collarum device, the selectivity and efficiency of this trap to capture foxes were acceptabl e, which agrees with previous studies (Shivik et al. 2000, 2005; Muñoz-Igualada et al. 2008). To our knowledge, we reported the first case in the literature of a non-canid species captured using this device, which could be explained by the high density of wild boar in the study area. Injures to captured animals We recorded only non-severe injuries during cage-trapping, as those observed in other studies (Way et al. 2002; Shivik et al. 2005; Muñoz-Igualada et al. 2008). Therefore, our findings suggest that cage-trapping produces minimal impact on the welfare of target animals trapped. Way (2012) recently proposed cage-trapping as an alternative 82 method to other unpopular devices to capture canids, such as foot traps, which typically cause significant injuries to captured animals. Nevertheless, cage-traps must be revised daily to assure a low level of injuries (Duarte and Vargas 2001). Injuries in non-target species were even lower than those observed in foxes, and only an azure magpie died in a cage-trap strangled by the guillotine-door, although this was apparently a highly unlikely event. Previous studies reported that 80% of the canids captured with Collarum devices showed no injuries (Shivik et al. 2000, 2005; Muñoz-Igualada et al. 2008), which is a percentaje lower than that obtained using other devices, such as cage-traps or leg-traps (Muñoz-Igualada et al. 2008). Our results are in concordance with this, since the three foxes captured with this device showed minor injuries, which were caused when they bit the device. These injuries could be reduced simply by rubber coating the snare. In contrast, the wild boar captured died likely because the diameter of the loop, designed for the neck of foxes, is too small for a wild boar neck. Management Implications Several studies have shown that cage-traps present a low efficiency and selectivity for catching wild canids (Duarte and Vargas 2001; Shivik et al. 2005; Muñoz-Igualada et al. 2008). Our results show how the combined use of live baits and scent attractants may improve their efficiency to capture red foxes and slightly increase their selectivity in central Spain. Nevertheless, this cannot prevent the capture of non-target protected species and selectivity levels obtained are still low (max.21%; see Appendix 3.2.). The selectivity of this trap depends ultimately on the willingness of gamekeepers or hunters to release captured non-target animals. Accordingly, we do not recommended the use of cage-traps for fox management in hunting estates within central Spain. However, the small sample size obtained in our study requires that more tests are carried out at both levels, traps design (e.g. mesh size, triggered systems, doors modifications) and the study of other attractants for target species and/or aversive for non-target species. The Collarum device can be an alternative to cage-traps since it is effective to catch foxes and highly selective. Our results suggest also that some minor modification can decrease the level of harm caused by this device during the capture To our knowledge, the Collarum device has not been tested yet in areas of continuous presence and high abundance of large threatened carnivores, such as the Iberian lynx (Lynx pardinus), 83 Iberian wolf (Canis lupus signatus) or brown bear (Ursus arctos). So, it is unknown whether these species can be captured by this type of trap. Therefore, its use should be forbidden in these areas at least until more information is gathered. The use of this new device should be carried out only by skilled technical personnel to ensure proper handling of trapped animals and always under strict supervision of the competent authorities in wildlife. Moreover, it is essential that traps are checked at least daily to avoid injuries and unnecessary animal suffering. Acknowledgements This study was funded by Consejería de Medio Ambiente of Junta de Comunidades de Castilla-La Mancha (projects 02-227/RN-52 and PREG-05-23). Special thanks go to landowners who facilitated the access to the private game estates, and to the field staff of the regional government of Castilla-La Mancha. Thanks to people who assisted us during the fieldwork, especially to S. Luna and L.E. Mínguez. We are indebted to O. Rodríguez, M. Reglero and R. Sobrino who examined the captured animals. 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 appropriate permits, provided by the concerned institutions. 84 CAPÍTULO 4: Experimental evaluation of live cagetraps for Black-billed magpies Pica pica management in Spain Este capítulo ha sido publicado en revista SCI: Díaz-Ruiz F, García JT, Pérez-Rodríguez L, Ferreras P (2010) Experimental evaluation of live cage-traps for Black-billed magpies Pica pica management in Spain. European Journal of Wildlife Research 56: 239-248. 85 Abstract Black billed magpies (Pica pica) are considered as a nest predator of game and nongame birds in Europe. In rural areas of Spain magpie control is commonly used as a management tool in small game hunting estates. Cage-traps with a magpie as a decoy are the legal method most commonly used for controlling magpies in Spain although its performance has not yet been experimentally tested. We evaluated the selectivity, efficiency and the effect of different factors on capture rate of these traps for magpie control and determine the effect of magpie removal on magpie density. Only 4 out of 197 captures corresponded to non-target species, which were released unharmed. Since the release of non-target captures depends on the daily checking of the trap and the trapper commitment, in order to guarantee the efficiency and selectivity of this method traps should be revised daily by full time, qualified trappers. The efficiency of this method is high during the breeding season, reducing magpie density in the area where the control is performed. Highest capture rates were obtained in the first days after cagetraps setting. Neither the gender nor the origin (local or foreign) of the decoy significantly affected the capture rate. Among male decoys, experimentally increased testosterone levels did not increase capture rates. According to our results, the tested cage-traps with a living decoy could be employed as an efficient and selective method for magpie population management in Spain, when used by full time, qualified trappers. Keywords: cage-traps, capture rate, black-billed magpie, selectivity, predator management. 86 Introduction The effect of predators on species with high socio-economic value frequently causes conflicts among social stakeholders (Thirgood et al. 2000; Sillero-Zubiri and Laurenson 2001). Such conflicts have often caused the persecution of predators through illegal and non-selective methods (Delibes-Mateos 2006), causing negative impacts on wildlife conservation (Villafuerte et al. 1998). Hunting of red-legged partridges (Alectoris rufa) is an activity of economic interest in many rural areas of central and southern Spain (Bernabeu 2000). Predation is regarded in many of these areas as one of the main causes of the partridge populations decline (Vargas 2002). Among the predators of red-legged partridges, corvids are assumed to have high impact on partridge nests (Yanes et al. 1998) and, consequently, they have been traditionally controlled. In these areas, the black-billed magpie (Pica pica) is the most abundant corvid species, and magpie control is commonly employed in small game hunting estates in Spain (Otero 1995). The black-billed magpie is a generalist species, living in a wide range of habitats (Birkhead 1991). It feeds on a broad spectrum of food types: seeds, fruits, ground invertebrates, carrion and small vertebrates. Eggs form only a small proportion of the magpie diet (Birkhead 1991; Martínez 1992), and the impact of magpies on bird populations is still controversial (Gooch 1991; Thomson et al. 1998; Chiron and Julliard 2007). Some studies performed in Spain during the red-legged partridge breeding period suggested that most eggs consumed by magpies belong to red-legged partridges (77.8 %, Herranz 2000). According to artificial nest experiments, magpies may be locally the most important predator of red-legged partridge nests, (Blanco-Aguiar et al. 2001) and magpie abundance may be regionally the best indicator of nest predation probability (Ferreras et al. 2006). Population dynamics of partridges can be negatively affected by nest predation (Potts 1980), and hence in places where magpies reach high densities, their removal may increase the breeding success of red-legged partridges and other game bird species (Martínez de Castilla and Martínez 2004). Black-billed magpies are in expansion in Europe since 1960 (Birkhead 1991) and a positive trend of 25% has been reported in Spain between 1995 and 2001 (SEO/BirdLife 2002). Effective management tools for abundant populations of magpies 87 can be therefore necessary for alleviating their pressure on declining species such as red-legged partridges in circumstances where predation on nests is particularly high. Many methods have been traditionally used in Spain for capturing magpies, including currently forbidden methods such as eagle owls (Bubo bubo) as decoys combined with mist-nests (Wang and Trost 2000), glued-branches (Boza 2002) or poisoned baits, the latter frequently used in Spain over the last decades (Hernandez and Margalida 2009) and legally prohibited since 1989 (Law 4 / 1989 on Conservation of Natural Areas and Flora and Wildlife) to be massive and non-selective methods. Currently, the methods legally employed for capturing magpies include shooting in communal roosts or driven hunting, nest destruction and cage-traps. Among these, cagetraps with a magpie as a decoy is the most commonly used method, likely due to their efficiency and their ease to operate. Gamekeepers, hunters and manufacturers assure that cage-traps with a magpie as a decoy are an effective and selective method for reducing magpie density. Popular recommendations for increasing the capture rate based on non-systematic observations include using foreign magpies (i.e., magpies originating from an area different from the one in which cages are being used) as decoys, and using them throughout the magpie breeding season. Also, these popular recommendations suggest to use male birds as decoys, as it is expected that their more active territorial behaviour, which is highly determined by testosterone levels (Wingfield et al. 1987), will be more effective in attracting conspecifics to traps. However, no experimental studies have tested these recommendations. On the other hand, conservationists claim that cage-traps aimed to capture magpies are often not selective and may negatively affect other species, particularly raptors, which enter into the traps trying to capture the decoy. The objectives of the present study were: 1) to assess the selectivity of cage traps, 2) to evaluate the efficiency of cage traps with a living decoy to capture black-billed magpies, 3) to determine the effect on the capture rate of several factors such as the gender and testosterone levels of decoys, the origin of decoys, the trapping season and the permanence time of traps in the same place and 4) to determine the effect of magpie removal on magpie density. 88 Materials and Methods Study Area The study was carried out in two hunting estates located in Castilla-La Mancha (Central Spain) during spring and autumn 2006. Area 1 (960 ha) was placed in the province of Ciudad Real, within an agricultural-dominated landscape. Natural vegetation layers were primarily bushes and some trees associated to riparian areas. Hunting is an important activity in this area, where the main game species are the Iberian hare (Lepus granatensis), the wild rabbit (Oryctolagus cuniculus) and the red-legged partridge. Magpies were not systematically controlled before the study, although they were occasionally shot during the hunting season. Area 2 (547 ha), located in the province of Toledo, was also dominated by agricultural landscape, and bushes and some trees were associated with hedgerows. The main game species were the Iberian hare and the wild rabbit; the density of the red legged partridge was low and hence it was not among the main game species. Magpies were not controlled in this study area previous to this study. Magpie density before the breeding season was similar in the two study areas (see below). Magpie trapping We evaluated the efficiency of four different models of cage traps commonly used in Spain for capturing magpies, all of them using a live magpie as decoy. Cage-traps have one central chamber for the decoy and several capture chambers around the decoy chamber, employing a guillotine-type door as capture system. Models 1-3 had four capture chambers, octagonal prism structure and similar size (approximately: 85 x 85 x 35 cm long x wide x high; See Appendix). Model 4 had two capture chambers and rectangular prism structure (model 4, 90 x 30 x 30 cm long x wide x high; See Appendix 4.1.). All cage-traps were made of metallic mesh of variable gauge: thick (3 mm) in model 1, medium (1.9 mm) in model 3 and light (1 mm) in models 2 and 4. Cage-traps were located near magpie nests (<50 m). For this purpose, nests were searched previous to the spring trapping experiments (February-March). Magpie nests are easily found during this season because deciduous trees lack leaves, magpie nests are large, distinctive and conspicuous (Birkhead 1991), and pairs are very active building and defending the nest. Traps were separated at least 50 meters among them, and under tree or shrub shade to avoid sunstroke of the decoy and captured animals in 89 the central hours of the day. All cage-traps were checked daily in the morning, all captures removed, and the decoy was provided with food and water ad libitum. In order to compare the effect of different factors on traps performance we defined capture-rate as the average number of magpies captured per day that a trap is operative (International Organization for Standardization 1999). Testosterone manipulation Testosterone causes aggressive and territorial behaviour in male birds (Wingfield et al. 1987), which could affect the capture efficiency of decoys. For this reason, ten male decoys, sexed through molecular techniques from blood samples (Fridolfsson and Ellegren 1999), were experimentally provided with testosterone implants. Implants consisted in 10 mm long silastic tubes (inner diameter of 1.47 mm, outer diameter of 1.97 mm) filled with crystallized testosterone (T-males; ICN Biomedicals, Irvine, CA) or empty (C-males). These tubes were sub-cutaneously implanted in the dorsal zone between the wings (Blas et al. 2006).To assess the efficiency of testosterone implants in creating significant differences in testosterone levels between T-males and C-males, we collected 0.3ml of blood from the brachial vein of all birds before and 5 days after implantation. Blood samples were stored cold (4ºC) and centrifuged within 4 hours, and plasma was subsequently stored at -80ºC until testosterone quantification. Plasma testosterone concentration was measured using a commercially available testosterone enzyme immunoassay (Elisa Kit EIA-1559 from DRG Diagnostics, Marburg, Germany). Before implanting, C- and T- males did not differ in plasma testosterone levels (t-test, t=0.87, p=0.39). However, after implanting, T-males showed higher testosterone levels than C-males (2.3±0.11 (SE) and 1.04±0.13 (SE) ng/ml, respectively; t-test, t=8.53, p<0.05). Testosterone levels of T-males after implantation were within the range found in control birds or all birds before manipulation. Although testosterone levels where not measured again during the rest of the experiment, our previous experiences indicate that implants of this size are fully active during at least 50 days. In addition, the visual inspection of implants through the skin indicated that they were still active (i.e. they were partially filled with testosterone) during the whole extension of the experiment. 90 Experimental design Four experiments were designed to test the effect of different factors on capture-rate (Table 4.1.). Experiments 1, 2 and 4 consisted of n (2-4) blocks or groups of traps of the same model. Each block included one trap of each treatment. For instance, each block in experiment 1 consisted of one trap with female decoy, one trap with T-male decoy and one trap with C-male decoy. Experiment 3 consisted in a single block of 9 traps of models 1, 2 and 3 (all with 4 capture chambers). Traps belonging to one block were set in the same area separated at least 50 m. - Experiment 1 was performed during spring in Area 1, using three decoy types: females (F), control males (CM) and males implanted with testosterone (TM). All decoys were from foreign origin. Fifteen cage-traps were installed and remained active for 13 days. The following variables were evaluated with this experiment: trap model, gender and testosterone of decoy and days since trap placement. Moreover, the effect of magpie removals on magpie density was examined together with data from experiments 2 and 4. - Experiment 2 was performed during spring in Area 1, using F and CM decoys with different origins, local (L) and foreign (F). Sixteen cage-traps were installed and remained active 10 days. In this experiment we evaluated the following variables: trap model, gender and origin of decoys and days since trap placement. Moreover, the effect of magpie removal on magpie density was examined together with data from experiments 1 and 4. - Experiment 3 was performed during autumn in Area 1, using F and CM decoys, all from local origin. We installed 9 cage-traps that remain active 20 days. This experiment was used for evaluating trapping season, together with experiments 1 and 2. - Experiment 4 was performed during the spring season in Area 2, using three types of decoy: F, CM and TM, all decoys from foreign origin. We installed 12 cage-traps that remained active during14 days. In this experiment we evaluated trap model, the gender and testosterone of decoy. Moreover, the effect of magpie removal on magpie density was examined together with data from experiments 1 and 2. 91 Table 4.1. Summary of field experiments: variables evaluated, decoy gender and testosterone level (F: female, CM: control male, TM: testosterone implanted male), decoy origin, area where the experiment was carried out, season, total number of traps employed and duration (days) of the experiment. Experiment Variables Gender and Exp.1 Testosterone and trap model Exp.2 Exp.3 Gender, origin and trap model Season Gender and Exp.4 testosterone and trap model Decoy gender & testos. F, CM & TM F & CM F & CM F, CM & TM Decoy origin Study Area Season Nr Traps Duration (days) Foreign Area 1 Spring 15 13 Area 1 Spring 16 10 Local Area 1 Autumn 9 20 Foreign Area 2 Spring 12 14 Foreign & Local Handling of captures Captured animals were examined for possible trap-related injuries. Non-target species were checked for injuries and released in the capture site. Trap selectivity was defined as the proportion of captured magpies in relation to the total number (target and nontarget) of captured animals (International Organization for Standardization 1999). The captured magpies were euthanized through an intraperitoneal injection of sodium pentobarbitone (200 mg/ml Dolethal Vetoquinol), as recommended for birds (Close et al. 1997). Data from necropsy of captured magpies (age, gender, physical condition) were used for further studies (authors in prep.). Some captured magpies were kept alive and used as decoys in further experiments, once sexed through molecular techniques from blood samples (Fridolfsson and Ellegren 1999). Those magpies used as decoys in the same study area where they were captured were considered as “local decoys”, whereas those captured elsewhere were considered as “foreign decoys”. Magpie density estimation The density of magpies in both study areas was estimated with the distance-sampling method (Burnham et al. 1980), which has been successfully employed to estimate the density of a number of bird species, including magpies (Newson et al. 2008). We 92 employed the Fourier series estimator as detection function. Surveys were carried out once a week during the trapping period, following a fixed route (21 km in Area 1 and 12 km in Area 2) with high visibility, starting two hours after sunrise. We indirectly assessed the effect of magpie removal on population density (experiment 1 and 2, in study area 1 and experiment 4 in study area 2) by relating the changes in the density of magpies with the number of animals captured in our cage-traps. Raptor and corvid species observed during line transects were recorded in order to assess the abundance of potential magpie predators or competitors which could enter the traps attracted by the magpie decoy, and related to trap selectivity. Data analysis We first modelled daily capture rate (number of magpies trapped each day that a trap was active) using data from both study areas during spring (exps.1, 2 and 4). Fixed factors included in this model were: number of days since the trap was installed and decoy origin (local or foreign). Study area, experimental block and trap location (a categorical variable identifying each trap position) were included as random effects. Trap location was nested within experimental block, because these random effects are not independent. Generalized mixed models with a Poisson error term and a log link function were used for this and for the remaining analyses of factors affecting capture rate (see below). Since the amount of time each cage trap was active differed among experiments (see Table 4.1.) and capture rate was significantly affected by time since trap installation (see results), we only considered in further analyses of spring data the captures obtained during the first 5 days after trap installation, when highest capture rates were obtained (see results). By this way, we controlled for experiment duration and did not include this variable in further models of spring data. In order to test for variation in capture rate among experimental treatments, we modelled the number of magpie captures in the first 5 days after trap installation, using data from spring experiments (exps. 1, 2 and 4), and the following explanatory factors: trap model (a categorical factor with four levels), type of decoy (female, Control-male and Testosterone-male), decoy origin (local or foreign) and study area (1 or 2). Models were fitted to all the data from experiments 1, 2 and 4. We ranked the obtained models according to their Akaike Information Criterion value (AIC) with respect to the principle of parsimony (Akaike 1973; Burnham and 93 Anderson 1998). The statistical significance of the parameters estimated was assessed using the Wald test. Finally, the effect of season on capture rate was analysed including only data from study area 1, where trapping was performed both in spring (experiments 1 and 2) and in autumn (experiment 3). Since the decay of capture rate along time since trap installation could differ between seasons, we considered the daily capture rate as dependent variable, and fitted generalized linear models to these data, including day since trap installation, season and their interaction as fixed effects. Results Selectivity A total effort of 708 trap-days during spring and autumn 2006 produced the capture of 193 magpies and 4 individuals of non-target species, which indicate a high selectivity of the trapping method (97.9% captures of the target species). Non-target captures were: common buzzard (Buteo buteo), genet (Genetta genetta), Western hedgehog (Erinaceus europaeus) and red-legged partridge. None of these captured animals resulted injured and they were released in the capture site. Despite the low capture rate of non-target species, several medium raptors, potential predators of magpies, were frequently seen in the surroundings (<100 m) of the traps during the daily checks and during the weekly transects for magpie density estimation (Table 4.2.). Effect of trapping time on capture rate Average capture rate during spring was 0.32 magpies/trap-days (see Table 4.3.). We checked for overdispersion in the data, and extra-dispersion scale was close to 1 (0.9), therefore it was not necessary to correct for this factor and the use of Poisson errors was appropriate for modelling capture rates. Capture rate significantly decreased over time since trap installation (Table 4.4.; Figure 4.1.). Other significant term included in the models was the interaction between time and decoy origin (see Table 4.4.). According to this, local decoys provided higher capture rate than foreign decoys during the first days, but lower during latter days (Figure 4.2.). Study area, block and trap location, included as random terms, did not result significant. 94 Table 4.2. Non-target species susceptible of being captured in the traps that were observed in the traps vicinity during daily trap checking and along weekly linear transects in both study areas. Linear transects Trap vicinity Area 1 Area 2 Area 1 Area 2 Circus sp. 7 6 7 0 Buteo buteo 6 13 4 1 Asio otus 0 0 1 0 Milvus migrans 6 0 0 0 Accipiter nisus 2 0 0 0 Hieraaetus pennatus 1 0 0 0 Corvus corone 0 0 1 0 Corvus monedula 36 108 0 0 Total length 219 km 60 km Table 4.3. Number of magpies captured, trapping effort and average capture rate during spring and autumn in each study site. Effort Capture rate Nr magpies captured (trap-days) Site Area 1 105 355 0,26 Area 2 62 168 0,37 Autumn Area 1 26 185 Season Spring Site Average 0.32 95 0.14 Table 4.4. Summary of results of the mixed model of capture rate including time since installation and decoy origin (data from spring in study areas 1 and 2). Area, block and trap location are controlled as random variables. Effect DF F P Time since installation 1,536 46.56 0.0001 Decoy origin 1,305 3.54 0.061 Time x decoy origin 1,536 4.74 0.030 Figure 4.1. Daily capture rate changes along time since trap installation during spring 2006 (experiment 1 and 2) and autumn 2006 (experiment 3) in area 1. 96 Figure 4.2. Expected capture rate as a function of time since trap installation (days) and the origin of the magpie employed as decoy. Effects of type of decoy and trap model on capture rate during spring The following analysis focused on the number of magpies captured during the first five days after trap installation, when capture rate is highest in all the experiments (see Figures 4.1. and 4.2.). None of the factors considered (trap type, gender-testosterone and decoy origin) resulted significant in the models (Table 4.5.). However trap type was included in the five models with lowest AIC and had the highest sum of Akaike weight (Tables 4.5. and 4.6.). The trap types with four capture chambers tend to have higher capture rates than the model with two capture chambers (Figure 4.3.). Effect of season Average capture rate in study area 1 during autumn (experiment 3) was lower (0.14 magpies/trap-day) than during spring (0.26 magpie/trap-day; see Table 4.3.; Figure 4.1.). However, only days since trap installation, but not season, resulted significant in generalized models including data from spring and autumn in study area 1 (Table 4.7. ). 97 Table 4.5. Significance of variables used in the mixed models for magpies captured during the five days since trap installation. Last column indicates the relative importance of each predictor variable estimated as the sum of the Akaike weights over all the models including each variable. Degr. of Variable freedom Wald test p Intercept 1 0.531 0.466 Trap type 3 0.992 0.609 0.618 Gender & testosterone 2 3.433 0.180 0.275 Origin 1 2.407 0.121 0.209 Σwi Table 4.6. Summary of mixed models for magpies captured during the five days since trap installation. Variables: Tr: trap type; DS: decoy gender and testosterone level; DO: decoy origin. Model Variables Degr. of AIC freedom ∆AIC Chi-2 p wi 1 Tr 2 172.219 0.000 3.886 0.143 0.106 2 Tr + DO*Tr 4 172.648 0.430 7.456 0.114 0.086 3 DO 1 173.211 0.992 0.893 0.345 0.065 4 Tr + DS*DO 4 173.508 1.289 6.597 0.159 0.056 5 DS*DO 2 173.587 1.368 2.518 0.284 0.054 6 DO + Tr 3 173.704 1.485 4.401 0.221 0.051 7 DO*Tr 2 173.850 1.631 2.254 0.324 0.047 8 DS 2 173.926 1.707 2.178 0.336 0.045 9 Tr + DS*DO + DO*Tr 6 174.075 1.856 10.030 0.123 0.042 98 Figure 4.3. Average number (±SE) of captured magpies per trap during the first 5 days after trap installation according to trap model (1-3, with four capture chambers; 4 with two capture chambers). Table 4.7. Summary of results of the mixed model of capture rate including time since installation and season (only data from study area 1). Variable Degr. Wald test p Σwi of freedom Intercept 1 0.223 0.637 Season 1 2.194 0.139 0.378 Time since installation 1 13.017 0.0003 1.000 Season x Time since installation 1 2.054 0.152 1.175 Effect of captures on magpie density Magpie density before the breeding season was estimated as 0.23±0.06 magpies/ha in study area 1 and 0.39±0.09 magpies/ha in study area 2. The magpies removal during the breeding season (spring) was followed by a decline in magpie density in both study areas (see Figure 4.4.). In area 1 the initial density declined coinciding with the first 60 99 magpies removed. Despite magpie density increased in the fourth census, the density at the end of the trapping season was lower than the initial density (Figure 4.4a.). After the trapping ceased, magpie density tended to increase. In area 2, the effect of trapping is clearer than in area 1. There, the initial density of magpies was reduced following the trapping season (Figure 4.4b.), and did not increase after the end of the trapping season. Figure 4.4. Changes in magpie density (magpies/ha ± SE) along time (dashed line) and accumulated captures during the trapping season (solid line) in study area 1 (A) and study area 2 (B). 100 Discussion The tested traps are highly selective for the capture of magpies, according to our results (97% selectivity). This is not the result of the absence of species susceptible to be captured in the traps. Both systematic and non-systematic surveys indicate that species susceptible to enter the traps are abundant in the study areas (see Table 4.2.). This is the case of magpie predators, such as medium-size raptors, and magpie competitors, such as other corvids (e.g. jackdaw (Corvus monedula); Högstedt 1980). Only a common buzzard was captured in the traps among the medium-size raptors able to capture magpies that were observed Accipiter nisus, Circus sp., Buteo buteo. However, we did not detect in any of the study areas the presence of goshawk (Accipiter gentilis), a reputed magpie predator (Mañosa 1994), which likely would enter the traps. Some small carnivores Genetta genetta, Martes foina, Mustela nivalis, and Mustela putorius which could be attracted by the decoy and captured in these traps, are likely present in the study areas according to distribution atlas (Palomo et al. 2007), although we lack quantitative data on their abundance. However, only a common genet was captured during the study. The release of non-target animals captured in magpie traps, when used as a management tool, depends totally on the trapper commitment, as it happens with other traps used for predator control (Duarte et al. 2001). Because of that, the training and the awareness of the trappers are necessary to guarantee the release of non-target captures. Traps must be checked daily to prevent long restraint periods which can reduce the animal welfare and eventually cause the death of both target and non-target species. The assessed traps resulted highly effective for the capture of magpies during the breeding season, producing an average capture rate of 0.32 magpies/trap and day. The daily capture rate was highest during the first day after trap installation (0.73-0.87 magpies/trap; Figure 4.1.). These values are much higher than those obtained in one-day attempts with bal-chatri traps using an adult female as decoy (0.43; Wang and Trost 2000), although other factors such as magpie density could have affected capture rate. The trap type was included in the best models of capture rate during the first five days, although it was not a significant term. In fact, number of capture chambers seems to increase the capture rate (although not significantly), since trap models with four 101 chambers tended to provide more captures than the model with two capture chambers (Figure 4.3.). The popular recommendation of using foreign magpies as decoys to increase captures is not supported by our data, since the origin of the decoy seems not to affect the capture rate. In fact, we obtained a similar number of captures when using the first magpie captured as decoy in the trap where it was captured (authors unpublished, data not included). Although both magpie males and females defend territories, this behaviour is more conspicuous in males (Baeyens 1981; Birkhead 1991). However, neither the gender of the decoy nor the testosterone level affected significantly the capture rate (Table 4.5.). Therefore, our data do not support the popular assumption about increasing captures by using male decoys from distant populations. The lower capture rate in autumn compared to spring (Table 4.3.) could be explained, at least partially, by the lower density just before autumn trapping period (0.17 and 0.23 magpies/ha, respectively for autumn and spring in study area 1). This lower density in autumn is probably a result of the magpies removed during the experimental trapping in spring. Also during autumn and winter, magpies are more sociable and not so aggressive when defending their territory (Eden 1989; Birkhead 1991). This lower territoriality during the non-breeding season can also explain the lower tendency of magpies to enter the traps. Recent studies show that predator control often do not reduce local predator abundances (Baker and Harris 2006; Beja et al. 2009). However, in our study, there was a strong decline in one area, whereas in the other the pattern was unclear (Figure 4.4.). In the study area 2 we observed an increase in magpie density one week after trapping started, which was followed by a density decrease during the next week (Figure 4.4b.). In both areas, these density fluctuations are probably due to sampling variability. In any case, in both study areas trapping was able to reduce magpie density during the breeding season of game species such as red-legged partridge and therefore the potential predation impact upon nests reduced. Management Implications The use of non-selective, illegal methods for predator control in Spain is one of the main causes of mortality for many predator species, both mammals and birds, some of 102 them endangered (Villafuerte et al. 1998). Therefore, it is necessary to identify selective methods for predator control to be used as management and conservation tools in particular situations of high abundance of generalist, non-protected predators. Such is the case of the traps tested in the present study, which have resulted highly selective and efficient. Some recommendations for using this type of traps for managing magpie populations can be drawn from our results. The breeding season is the most appropriate for effectively control magpie populations with these traps, since capture rate is higher in this period, the magpie density of unmanaged populations is lowest just before breeding and easier to be controlled. On the other hand, this period coincides with nesting of most bird species, including red-legged partridges, reducing in this way nest losses due to magpie predation. Traps located in the proximity of magpie nests are highly effective but their efficiency would increase if they are moved to a new location after 4-5 days. Either local or foreign magpies of any gender can be used as decoys in the traps with similar results in capture rate. Traps should be checked daily in order to avoid the reduction of welfare of captures and the personnel in charge of setting and manipulating these traps must be encouraged to liberate individuals of non-target species. Other likely side effects of the traps assessed should be considered before their generalized use. For instance, the effect of the reduction of magpie populations on Great spotted cuckoo populations (Clamator glandarius), a nest parasite specialized on magpie nests (Soler et al. 1996), should be scientifically evaluated and taken into account when authorising the use of traps for magpie control. Acknowledgements This study was funded by Consejería de Medio Ambiente of Junta de Comunidades de Castilla-La Mancha (project PREG-05-23). Patrick Fasolo kindly provided the first magpie decoys to start the trapping experiments, and shared with us his long experience with the use of the traps. Land owners and game owners of both study areas facilitated the access to estates and facilitated the field tests. Salvador Luna performed most of the field work. Luis Enrique Mínguez kindly assisted in the field work and solved most bureaucracy during the project development. Beatriz Arroyo provided helpful support with the statistical analyses. 103 Ethical standards All the experiments comply with the current Spanish laws, and were performed with the corresponding legal authorizations and following current guidelines for animal welfare. 104 CAPÍTULO 5: Assessing the influence of predator control on target and non-target predator populations using occupancy models Este capítulo se encuentra en preparación para ser enviado a una revista SCI: Díaz-Ruiz F, Caro J, Delibes-Mateos M, Arroyo B, Ferreras P (en preparación) Assessing the influence of predator control on target and non-target predator populations using occupancy models. 105 Abstract Lethal control of predators may affect the structure and composition of predator communities, and this can have far-reaching ecological consequences, including the precipitation of trophic cascades and species declines. Understanding the effects of predator control on predator communities is therefore of great interest for the conservation and management of wildlife. In the present study, we used camera traps and occupancy models to assess the influence of red fox (Vulpes vulpes) control on foxes (target species) and stone marten (Martes foina, non-target species) across 12 localities of Mediterranean environments in central Iberian Peninsula. Our results show that the intensity of fox control was not associated with red fox occupancy, whereas it was negatively related to red fox detectability. This suggests that fox control could decrease the species’ abundance if we assume a relationship between abundance and detectability as suggested by some authors. On the other hand, the intensity of fox control was positively related to stone marten occupancy, but unrelated to its detectability. Nevertheless, habitat composition and prey availability were more closely associated with site occupancy of both species than red fox control. Our study suggests that predator control could affect target (red fox) populations both spatially and numerically, differently than non target (stone marten) populations. Furthermore, red fox extractions could benefit subordinate sympatric mesocarnivores, such as stone marten, through a competitor release process. This work provides valuable information on the ecological consequences of fox control to be considered in the management of red fox populations conducted in game estates. Key words: competitor release, hunting, mesocarnivores, predator management, Vulpes vulpes, Martes foina 106 Introduction Changes in the relative abundance of sympatric carnivores can have far-reaching ecological consequences, including the precipitation of trophic cascades and species declines (Prugh et al. 2009; Levi and Willmers 2012). Lethal control of predators may be one of the main factors affecting the structure and composition of carnivore communities in areas where this management practice is performed. The main goal of predator control is reducing the incidence of predation on prey, and it has become a widely management tool used to preserve both prey species of conservation concern and human interests (Woodroffe et al. 2005). In this sense, the red fox is the species of predator most often persecuted because it usually impacts negatively on livestock or game species (Sillero-Zubiri et al. 2004). In Spain predator control is mostly focused on the red fox and it is usually carried out by hunters and game managers (Díaz-Ruiz and Ferreras 2013), who consider it is indispensable to reduce fox impact on small-game prey (Rios-Saldaña 2010; DelibesMateos et al. 2013). The red fox is a game species in Spain; it is legally culled during the regular hunting season (autumn-winter, pre-reproduction fox season) mainly by shooting (Delibes-Mateos et al. 2013). In some small game estates, exceptional permits are also granted to cull foxes outside the regular hunting season (Delibes-Mateos et al. 2013) by means of traditional methods, such as cage traps and neck snares (MuñozIgualada et al. 2008, 2010). Although game managers often consider that fox numbers are reduced as a result of intensive fox culling (but see Delibes-Mateos et al. 2013), the effect of predator control on fox populations is still controversial, as several studies have shown different, often opposed, results (Heydon et al. 2000; Baker and Harris 2006; Saunders et al. 2010; Mateo-Moriones et al. 2012; Berry et al. 2013). Fox control could also have effects on other sympatric mesocarnivores. On the one hand, the Competitor Release Hypothesis states that the reduction of a dominant competitor species may benefit other subordinate competitor that increases its abundance (Caut et al. 2007). This has been shown experimentally in UK, where the culling of Eurasian badgers (Meles meles), dominant competitor, was associated with an increase in densities of a subordinate competitor, the red fox (Trewby et al. 2008). On the other hand, non-selective predator control methods (e.g. cage-traps, poisoning, etc) are sometimes used in some hunting estates, resulting in the culling of both target and non-target mesocarnivore species (Duarte and Vargas 107 2001; Barrull et al. 2011). In this sense, recent studies based on theoretical simulation models indicate that certain levels of non-selective control of red fox populations could alter the carnivore communities with an increase in the abundance of the target predator (i.e. the red fox), and a reduction in the numbers of other non-target species (red fox “competitors”), such as the Eurasian badger, the stone marten (Martes foina) and the pine marten (Martes martes) (Casanovas et al. 2012; Lozano et al. 2013). It is therefore of particular interest to all stakeholders involved in the conservation and management of wildlife to understand the multiple effects of predator control on terrestrial carnivore communities, including the interactions between carnivore species. Detecting these effects on the populations of both target and non-target carnivores is not an easy task because these species are often cryptic and occurs at low densities, and therefore specific sampling methods are required to a reliable monitoring of their populations (Boitiani and Fuller 2000; Long et al. 2008). Camera traps constitute a good alternative to monitor rare, elusive species, such as mammalian carnivores and the effect of management interventions on carnivore populations (Johnson et al. 2009; Sarmento et al. 2011; Schuette et al. 2013). Camera-trapping data combined with new methodologies of statistical analysis allows characterizing the status and changes in the populations of these species. For example, Occupancy Models are often applied to camera trapping data to estimate the probability of site occupancy of a species; i.e. the proportion of sites occupied by the species (MacKenzie et al. 2002; MacKenzie et al. 2006). This is estimated from the detection/non-detection data obtained from several sampling sites and during several sampling occasions. Occupancy has been used as a surrogate of abundance for many inferential purposes, including habitat selection, population dynamics and distribution, or changes in population size (Royle and Nichols 2003; MacKenzie and Nichols 2004; MacKenzie et al. 2006). To obtain unbiased estimates of occupancy it is fundamental to account for detection probability; i.e. the probability of detecting the species, given its presence, during the independent survey of sampling sites (MacKenzie et al. 2002; Mackenzy et al. 2006) In the present study, we combined the use of camera traps and occupancy models to assess the influence of red fox control on occupancy and detection probability of populations of foxes and stone martens, a non-target predator, in Mediterranean environments within central Iberian Peninsula. In addition, we discuss about the 108 potential relationship between predator control and the abundance and spatial distribution of both species, and it implications for the predator management. Material and Methods Study areas The study was conducted in 12 localities within central Spain (Figure 5.1.), characterized by hot and dry summers, cold winters and most rainfall occurring during autumn-spring months (Mediterranean bioclimatic region; Rivas-Martínez et al. 2004). The landscape was heterogeneous, and the main habitats present in all localities were Mediterranean scrubland (mainly (Cistus spp.) in combination with holm oak (Quercus ilex) forests), mixed with cereal croplands and permanent crops, such as olive groves (Olea europaea) and vineyards (Vitis vinifera), and natural pastures. Other less abundant habitats included riparian areas and ‘dehesas’ (pastureland with savannah-like open tree layer, mainly dominated by Mediterranean evergreen oaks). Surface and habitat composition varied among localities (see Table 5.1. and Appendix 5.1. for a detailed description). All localities were hunting estates, with the exception of two protected areas (numbers 5 and 11 in Figure 5.1.), where hunting was not allowed. The main small game species were European wild rabbit (Oryctolagus cuniculus; hereafter rabbit), red-legged partridge (Alectoris rufa) and Iberian hare (Lepus granatensis). Red deer (Cervus elaphus) and wild board (Sus scrofa) were the main big game species. Hunting estates were managed to boost small game numbers, mainly by the provision of supplementary food and water, and predator control. The intensity of fox control varied among hunting estates (Table 5.1., and see below). Camera trap surveys We carried out camera trap surveys between 2010 and 2013 in late spring and summer (May-September, Table 5.1.), after Iberian mesocarnivores breeding season (Blanco 1998). We used two models of infrared-triggered digital cameras: Leaf River IR5 (Leaf River OutDoor Products, Taylorsville, Mississippi, USA) and HCO ScoutGuard (HCO OutDoor Products, Norcross, Georgia, USA). Camera stations were regularly deployed with an average distance of ~1.2 km among neighboring cameras, ensuring independence between them (Monterroso et al. 2011, 2013). The number of camera 109 traps deployed in each study locality varied from 14 to 20, according to locality surface (Table 5.1.). Figure 5.1. Situation of the study localities (1-12) in the Iberian Peninsula. Cameras were mounted on trees approximately 0.5–1.0 m off the ground and set to record time and date when triggered. Cameras operated 24 h a day for an average period of 28.4±0.4 days (mean±SE). We programmed cameras with the minimum time delay between consecutive photos to maximize the number of photos taken per captured individual (Monterroso et al. 2011; 2013), so assuring the species identification of each event. In order to increase the detection probability of mesocarnivores, we set the sensitivity of the infrared sensor at the highest level, and used Valerian scent and Iberian lynx (Lynx pardinus) urine as lures. This combination has been described as an effective attractant for a wide range of Iberian carnivores (Monterroso et al. 2011). Between 3 and 4 ml of each lure were put in two independent perforated plastic vials secured to a metal rod. Lures were set at 2-3 meters from each camera trap, and were replenished every two weeks, when cameras were inspected to check the batteries and to replace memory cards. Consecutive images of the same species within 30 min interval were considered as the same event, unless animals were clearly recognized as different individuals, and 110 those separated by a longer interval as independent events (Kelly and Holub 2008; Davis et al. 2011; Delibes-Mateos et al. 2014). Table 5.1. Description of study localities. The predominant landscape (agriculture or scrubland) is indicated along with the habitat types present in each area: Oa: open areas, Scr: scrubland, Wc: woody crops, Rip: riparian, Fo: forest, Dh: dehesa. “Red fox control” is the number of foxes culled per square km and year. “Control Method” is the main method employed to remove foxes: shooting, snaring or no control (No). “Cameras” indicate the number of camera-traps used in each locality. “Effort” (survey effort) is expressed as cameradays, or the sum of days each camera was active in the field in each locality. Study Locality type site Area and (Map (ha) Uses ID) Social hunting 1 2000 estate Small game Commercial 2 1600 hunting estate Small game Social hunting 3 5000 estate Small game Social hunting 4 3580 estate Small game 5 6 7 8 9 10 11 12 Landscape (Habitats types) Agricultural (Oa, Scr, Rip, Wc) Scrubland (Oa, Scr, Rip) Agricultural (Oa, Scr, Rip, Wc) Agricultural (Oa, Scr, Rip, Wc) Scrubland Protected area 2140 (Oa, Scr, Conservation Rip, Dh, Fo) Social hunting Scrubland 1560 estate Small (Oa, Scr, game Rip, Wc) Social hunting Agricultural 2140 estate Big (Oa, Scr, game Rip, Dh) Social hunting Agricultural 2000 estate Small (Oa, Scr, game Rip, Wc) Commercial Scrubland 900 hunting estate (Oa, Scr, Big game Rip, Dh, Fo) Commercial Agricultural 900 hunting estate (Oa, Scr, Small game Rip) Scrubland Protected area 2600 (Oa, Scr, Conservation Rip) Commercial Scrubland 1600 hunting estate (Oa, Scr, Big game Rip, Dh, Fo) Red fox control (foxes/km2year) Control Method 0.08 Shoot 2010 20 620 1.98 Snares 2010 15 424 0.89 Shoot 2011 18 493 0.43 Shoot 2011 17 485 0 No 2011 19 682 1.30 Snares 2011 20 645 0 No 2012 20 495 4.00 Snares 2012 20 503 0.10 Shoot/Cage trap 2012 15 417 2.70 Snares 2012 14 372 0 No 2012 20 529 0.70 Shoot 2013 18 463 111 Sampling Cameras Effort Year Selection of covariates Several factors could influence the probability of site occupancy and detectability of carnivores. Among these habitat composition and prey availability are the main factors explaining the presence/absence of carnivores at a given site (Long et al. 2011; Sarmento et al. 2011; Silva et al. 2013), although human disturbance, including predator control, may also play an important role (Long et al. 2011). Fox control data (i.e. human disturbance covariate) were gathered through face-to-face interviews with game managers of each hunting estate, conduct before field sampling (at the end of the regular hunting season, in February). We asked managers about the number of foxes removed in the previous hunting season (Table 5.1.). The intensity of red fox control (IFC) was estimated as the number of foxes removed per km2 and year (fox·year-1·km-2), and it was recorder at locality level. We confirmed during the interviews that the same fox control effort was developed in each locality for at least two years before our field samplings. Methods employed to control foxes varied between study localities, including shooting, cage-trapping and neck snaring (Table 5.1.). We also recorded data of covariates associated with habitat type and prey availability at each camera site to account for potential heterogeneity in the probability of occupancy and detectability. We used a Geographic Information System (QGIS version 1.8.0) to calculate the percentage of each habitat type within a circular buffer of 200-m radius around each camera trap (Ordeñana et al. 2010; Sarmento et al. 2011; Silva et al. 2013). We classified habitats into 6 different types using the combination of CORINE landcover 2006 (European Environment Agency; http://www.eea.europa.eu), Updated satellite orthophotos (National Geographic Institute, http://www.ign.es/), and field data recorded on site during installation-revision of camera traps. Main habitat types were scrublands (SCR), open areas (OA), woody crops (WC), riparian (RIP) and “dehesa” (DEH) (Appendix 5.1.). The European rabbit is the main prey of most mesocarnivores in the Iberian Peninsula (Delibes-Mateos et al. 2008a). Therefore, rabbit availability (hereafter RA) was assessed for each camera station as the number of independent detections per 100 trap days and used as a measure of prey availability (Kelly and Holub 2008; Davis et al. 2011; Monterroso 2013). 112 To avoid multicollinearity, we eliminated any covariate highly correlated with other covariates (Spearman rank correlation≥0.70). Thus, among habitat covariates, we eliminated OA as it was highly correlated with SCR. Prior to analysis we standardized all continuous covariates using the z-transformation (MacKenzie et al. 2006). Occupancy models We constructed detection histories for each camera trap placed on the 12 study localities. We divided each survey period into four 1-week sampling occasions during which the detection/non detection data of each target species was recorded (Sarmento et al. 2011; Monterroso 2013). For each species we developed single season occupancy models (MacKenzie et al. 2002) using the software PRESENCE 5.8 (Hines and Mackenzie 2013); these are based on the assumption that all sampling sites always present the same level of occupancy (i.e., either occupied or not) during the sampling period (MacKenzie et al. 2002). As our goal was to estimate the influence of covariates on occupancy and detectability simultaneously, we followed a two steps process to fit global models, as previously described (Sarmento et al. 2011; Harihar and Pandav 2012; Monterroso 2013). In the first step we selected independently the covariates that best explained detection probability (Habitat and IFC covariates) and occupancy (all covariates). We used a sequential modeling approach to find the best model for each parameter, by discarding uninformative variables (Arnold 2010). We first held occupancy (Ψ) constant and proceeded to find the best detection (p) model. In other words, we started building a full effect model for detection probability, and performed a backward-stepwise model selection to sequentially eliminate the covariate with the weaker effect size (β/SE). This process was kept until the deletion of an additional covariate led to an increase in AICc, keeping the variables included in the top model (Monterroso 2013). The same process was developed to find the best occupancy model, holding detection probability constant. In the second step we built for each species a global model that included all informative covariates selected in the top models of detection and occupancy probability developed in the first step. Then we followed the same procedure of backward-stepwise model selection as described above to find the final model set for each species, we selected as informative covariates for inference those that were included in models with ∆AICc< 2 113 units of the top-supported model (Burnham and Anderson 2002). We estimated overall AIC weights for individual variables by summing the AIC weights of all the candidate models in which they were included (Mackenzie et al. 2006). If no single model accounted for >90 % of the total model weights, we model-averaged by extracting the top 95 % model confidence set and recalculating model weights (Burnham and Anderson 2002). Model averaged estimates were calculated using the spreadsheet developed by B. Mitchell (http://www.uvm.edu/%7Ebmitchel/software.html). Results Camera-trapping survey Red fox was detected in all study localities with naïve occupancy (i.e. the raw proportion of. camera traps where the species was detected) ranging 0.20-0.95, and a total of 254 positive sampling occasions. Stone marten was detected in more than half of study localities (n=7, 58%) with a total of 65 positive sampling occasions, and naïve occupancy ranging 0-0.60. Besides red fox and stone marten, other 5 species of wild mesocarnivores were detected during the sampling period at different localities: common genet (Genetta genetta), Egyptian mongoose (Herpestes ichneumon), Eurasian badger (Meles meles), least weasel (Mustela nivalis) and wildcat (Felis silvestris) (Appendix 5.2.). Explanatory covariates of detection and occupancy According to the sequential modeling approach to select the covariates, the intensity of fox control was selected as covariate for fox detection to be included in the second step (Table 5.2.). The scrublands, riparian areas, dehesa, the intensity of fox control and rabbit availability were selected as covariates for red fox occupancy to be included in the second step (Table 5.3.). For the stone marten the proportion of scrubland and the intensity of fox control were selected for the detection probability (Table 5.2.). The proportion of scrubland, rabbit availability and the intensity of fox control were selected as covariates for stone marten occupancy to be included in the second step (Table 5.3.). 114 Table 5.2. Models obtained in the process developed for selecting detection covariates, by fitting Ψ constant: Ψ(.), for red fox and stone marten. Selected covariates are those that are included in the top model, which is marked in bold. IFC: intensity of fox control; RIP: % riparian habitat; SCR: % scrubland; WC: % woody crops. Model Red fox Ψ(.), p(IFC) Ψ(.), p(IFC+RIP) Ψ(.), p(IFC+RIP+SCR) Ψ(.), p(IFC+RIP+SCR+WC) Ψ(.), p(full) Ψ(.), p(.) AIC ∆AIC AIC wt K 2log L 895.72 896.28 897.90 899.60 901.60 906.45 0.00 0.56 2.18 3.88 5.88 10.73 0.436 0.330 0.147 0.063 0.023 0.002 3 4 5 6 7 2 889.72 888.28 887.9 887.6 887.6 902.45 Stone marten Ψ(.), p(IFC+SCR) Ψ(.), p(SCR) Ψ(.), p(IFC+RIP+SCR) Ψ(.), p(IFC+RIP+SCR+WC) Ψ(.), p(full) 421.18 421.40 421.50 423.42 425.38 0 0.22 0.32 2.24 4.20 0.313 0.280 0.267 0.102 0.038 4 3 5 6 7 413.18 415.4 411.5 411.42 411.38 Table 5.3. Models obtained in the process developed for selecting occupancy covariates, by fitting p constant: p(.), for red fox and stone marten. Selected covariates are those that are included in the top model, which is marked in bold. IFC: intensity of fox control; RIP: % riparian habitat; SCR: % scrubland; WC: % woody crops; DEH: % dehesa; RA: rabbit availability. AIC Model Red fox Ψ(SCR+RIP+DEH+IFC+RA), p(.) 904.22 Ψ(SCR+RIP+DEH+IFC), p(.) 905.44 Ψ(full), p(.) 905.81 Stone marten Ψ(SCR+IFC+RA), p(.) Ψ(SCR+RA), p(.) Ψ(SCR+RIP+IFC+RA), p(.) Ψ(SCR+RIP+DEH+IFC+RA), p(.) Ψ(full), p(.) 414.06 415.16 415.33 416.66 418.66 115 ∆AIC AIC wt K 2log L 0.00 1.22 1.59 0.501 0.272 0.226 7 6 8 890.22 893.44 889.81 0.00 1.10 1.27 2.60 4.60 0.403 0.233 0.214 0.110 0.040 5 4 6 7 8 404.06 407.16 403.33 402.66 402.66 Selection of global models and model averaging The second step for the red fox produced three models with ∆AICc< 2 (Table 5.4.). According to the model averaging, the proportion of scrubland, riparian areas and dehesa as well as rabbit availability and fox control intensity influenced red fox occupancy probability (Table 5.4.). The three habitat types were the most explicative covariates (w= 0.99) and were positively associated with occupancy probability; a marginally significant effect was obtained for the proportion of riparian areas (Table 5.5.; Figure 5.2a.). Rabbit availability was also positively associated with occupancy probability (Table 5.5.). Although with lower weight of evidence (w= 0.17), red fox control intensity was negatively related to the probability of red fox occupancy (Table 5.5.; Figure 5.3.). Red fox detection probability was affected negatively by red fox control intensity (Table 5.5. and Figure 5.2b.), which was included in the three top models (Table 5.4.). Table 5.4. Single season occupancy models (top ranked models ∆AIC < 2) for the red fox and stone marten including covariates previously selected for p and Ψ. IFC: intensity of fox control; RIP: % riparian habitat; SCR: % scrubland; DEH: % dehesa; RA: rabbit availability. Model Red fox Ψ(SCR+RIP+DEH+RA), p(IFC) Ψ(SCR+RIP+DEH), p(IFC) Ψ(SCR+RIP+DEH+RA+IFC), p(IFC) AIC ∆AIC AIC wt K 2log L 893.73 894.61 895.63 0.00 0.88 1.90 0.492 0.317 0.190 7 6 8 879.73 882.61 879.63 Stone marten Ψ(SCR+IFC+RA), p(.) Ψ(SCR+RA), p(.) Ψ(SCR+IFC+RA), p(SCR) 414.06 415.16 415.48 0.00 1.10 1.42 0.444 0.256 0.218 5 4 6 404.06 407.16 403.48 116 Table 5.5. Model averaged coefficients (β) and confidence intervals (CI 95%) of the covariates included in the set of models explaining the red fox and stone marten detectability p and occupancy Ψ. “wt” refers to covariate AIC weights. Habitat covariates: RIP riparian; DEH dehesa and SCR scrubland. IFC: intensity of fox control. RA: rabbit availability. **Significant covariates. *Marginally significant covariate. Red fox Ψ p Stone marten Ψ p Covariate intercept RIP* DEH SCR RA IFC β 0.53 0.39 0.29 0.22 0.32 -0.01 CI 95% (0.18, 0.88) (-0.05, 0.83) (-0.13, 0.72) (-0.10, 0.54) (-0.50, 1.14) (-0.17, 0.15) wt 0.99 0.99 0.99 0.66 0.17 intercept IFC** 0.03 -0.36 (-0.18, 0.25) (-0.6, -0.2) 1 Covariate intercept SCR** RA* IFC β -1.28 0.769 -3.84 0.28 CI 95% (-2.02, -0.3) (0.14, 1.39) (-7.72, 0.06) (-0.25, 0.81) 1.00 1.00 0.71 intercept SCR -1.27 0.04 (-1.78, -0.75) (-0.24, 0.33) 0.22 wt The sequential modeling approach for the stone marten produced three models with ∆AICc< 2 (Table 5.4.). According to the model averaging the proportion of scrubland, rabbit availability and fox control intensity explained stone marten occupancy (Table 5.5.). The proportion of scrubland was the most informative covariate and was positively associated with stone marten occupancy (Table 5.5.; Figure 5.4.). Stone marten occupancy was negatively related to rabbit availability and positively to red fox control intensity (Table 5.5.; Figure 5.3.). Stone marten detection probability was positively related to the proportion of scrubland, although with low weight of evidence (w= 0.22) (Tables 5.4., 5.5.). 117 Figure 5.2. (a) Relationship between occupancy probability (Ψ) of red fox and riparian, dehesa and scrubland habitat proportions. (b) Relationship between red fox detection probability (p) (solid line) and red fox control intensity (fox·year-1·km-2). Dashed lines represent the 95% CI estimated for detection probability. 118 Figure 5.3. Relationship between occupancy probability (Ψ) of red fox (grey lines) and stone marten (black lines) and red fox control intensity (fox·year-1·km-2). Dashed lines represent the 95% CI estimated for occupancy probability of each species. Figure 5.4. Relationship between occupancy probability (Ψ) of stone marten (solid line) and scrubland habitat proportion. Dashed lines represent the 95% CI estimated for occupancy probability. 119 Discussion The intensity of red fox control was related differently to both site occupancy and detection probability of red fox and stone marten. On the one hand, the intensity of fox control was not related to site occupancy of red fox, but it was negatively associated with red fox detectability. On the other hand, the intensity of red fox control was positively related to the stone marten occupancy probability but it was not associated with its detectability. Overall, habitat composition and prey availability were more important than red fox control to define site occupancy probability of both species. According to our results, the intensity of fox control was not related to the site occupancy of red fox. Although differences in occupancy probability have been interpreted as reflecting differences in abundance, it has been also suggested that occupancy and abundance address distinctly different aspects of population dynamics (MacKenzie and Nichols 2004; Mackenzy et al. 2006). Thus, some changes in population abundance may not be identified using occupancy estimates, and some changes in site occupancy may not reflect changes in abundance (Towerton et al. 2011). For example, if foxes use larger areas, then they may spend less time in any given part of those areas, thus influencing differently occupancy (increasing) and abundance (decreasing) (Towerton et al. 2011). When territorial adult foxes are removed from an area, this area is afterwards reoccupied by subadult foxes, which move through the landscape in search of available territories (Towerton et al. 2011). In our study area, it is likely that, after intensive fox control, fox population may be dominated by transient individuals, which have larger home-rages than territorial (Henry et al. 2005). This may explain not only the lack of association between fox extraction and red fox site occupancy, but also the apparent low relationship between the latter and population abundance; this seems to be related in this case with the spatial distribution of red foxes populations. Several authors have suggested that there is a close relationship between species’ abundance and its detectability (e.g. Royle and Nichols 2003; McCarhty et al. 2013). In this regard, the negative relationship between the detection probability of red foxes and the intensity of fox control might suggest that this could decrease fox abundance. This is in disagreement with other studies performed in the Iberian Peninsula that reported similar (Virgós and Travaini 2005), or even higher abundances of foxes (Beja et al. 120 2009) in areas where these were controlled than in areas without control. Nevertheless, these studies did not take into account the intensity of fox control, which could explain their contrasting results with respect to our study. It is also noteworthy that changes in the probability of detection may be the response of foxes to other factors like habitat type or behavioural responses to human disturbances. Interestingly, in our study area the intensity of fox control does not seem to be related to the overall activity levels of red foxes (Chapter 6). The difficulty to reduce fox populations has been demonstrated in several studies; the effect of fox control on its abundance is variable and depends on various factors, such as the control method used or the duration of control (Saunders et al. 2010). In our study, the method employed in localities with the highest rates of fox extraction was necksnaring (Table 1), which is considered an efficient method to remove red foxes in Spain (see Muñoz-Igualada et al. 2008, 2010). Furthermore, fox control effort was sustained at least during two years before our study in all localities. On the other hand, cagetrapping and/or shooting were the methods employed in localities with a low level of fox control (Table 5.1.), and these are usually ineffective to reduce fox abundance (Baker and Harris 2006; Saunders et al 2010). This could be also the reason why Beja et al. (2009) did not observe a lower number of foxes where predator control was employed, but data on fox control intensity were not considered by these authors. Red fox can use very different habitats due to its high ecological plasticity (SilleroZubiri et al. 2004; Sarmento et al. 2011). In this study, we found that red fox site occupancy increased with the proportion of riparian habitat, dehesas and scrubland. According to previous studies, scrublands and dehesas are determinant habitats in Mediterranean areas for Iberian carnivores, including the red fox (Mangas et al. 2008). The high relevance of riparian habitat according to our findings is probably due to the fact that the field sampling was carried out in hot, dry summer. During this season riparian habitat can become a key habitat for Iberian carnivores in Mediterranean ecosystems since they find there food, water and protection against high temperatures (Virgós 2001; Matos et al. 2009). Red fox site occupancy was also positively associated with rabbit availability, although this relationship was weaker than those found for habitat types. This is not surprising as rabbits are likely the most profitable prey for foxes in the Iberian Peninsula, and their consumption by foxes increases when rabbit numbers increase (Díaz-Ruiz et al. 2013; Delibes-Mateos et al. 2008b). 121 Within communities of Iberian mesocarnivores, red foxes could play a dominant competitor role over some sympatric species, such as the stone marten (Pereira et al. 2012; Monterroso 2013). In addition, the red fox is one of the carnivore species most often cited as killers of other mesocarnivores like Martes sp. (Palomares and Caro 1999). In this scenario, red fox extraction may benefit the competitor release of subordinate sympatric mesocarnivores, such as the stone marten. Our results are in concordance with this, since red fox extraction increased the occupancy probability of the stone marten. Similarly, in UK the experimental, selective control of Eurasian badgers, the dominant predator species, was associated with an increase in the densities of the red fox, the subordinate one (Macdonald et al. 2004; Trewby et al. 2008). In our study area localities with more intensive fox extraction used fox snaring, which is a controversial method in the Iberian Peninsula because it allows the capture of nontarget species (Duarte and Vargas 2001; Barrull et al. 2011). Different levels of nonselective control of red fox populations could alter the carnivore communities with an increase in the populations of the target species (i.e. red fox), while the populations of other non-target species (red fox “competitors”), such as the Eurasian badger, the stone marten and the pine marten, decrease markedly or even disappear (Casanovas et al. 2012; Lozano et al. 2013; Barrull et al. 2014). Contrarily, our results suggest that fox control may decrease fox abundance and may benefit the occurrence of a competitor the stone marten. These differences could be probably due to a mainly selective fox control in the studied localities or that at least the stone marten was not affected by potential non-selective fox control. However, it is unclear that non-selective predator control does not occur in our study area, as it has been reported in other Iberian regions (Beja et al. 2009; López et al. 2014). Stone marten occupancy was significantly higher in areas with higher proportion of scrubland. This is in accordance with previous studies, which suggest that stone martens occupy areas with high vegetation cover and structure complexity, including Mediterranean scrubland. These habitats provide diverse key feeding resources for stone martens, such as fruits, small mammals or birds (Barrientos and Virgós 2006; Virgós et al. 2010; Sarmento et al. 2011; Monterroso 2013) as well as refuge areas (Mangas et al. 2008). Stone martens present a generalist feeding behavior mainly based on the consumption of fruits and small mammals, although rabbits also may represent key resources for them during spring and summer seasons (Barrientos and Virgós 2006), 122 when our study was conducted. Surprisingly rabbit availability was negatively related to stone marten site occupancy. This might be an indirect effect of habitat since stone martens usually avoid landscapes dominated by open areas (Prigioni et al. 2008; Dudús et al. 2014), where rabbits tend to be more abundant (Virgos et al. 2003; Calvete et al. 2004). In fact, we found lower rabbit availability in sites dominated by closed habitats like scrublands, riparian areas and forests (authors, unpublished results). In conclusion, our work shows how the combined use of camera-trapping and occupancy models provide a useful tool to evaluate the relationship between management actions and changes in populations of managed species, especially for species that are difficult to monitor like carnivores. Nevertheless, some relationships and assumptions of the outcome of occupancy models are not entirely clear, such as the link between detection probability and population abundance, discussed here. From this perspective, rigorous experimental studies based on the combination of this new methodology and good quality abundance data of predators is essential to improve the knowledge on the effect of predator control on the population dynamics of target and non-target predators. Acknowledgements We are very grateful to land owners, game managers, game keepers and hunters who allowed us to work in their hunting estates, and to the staff of Cabañeros National Park and Ruidera Natural Park. Special thanks to people who assisted us during the fieldwork. This study was funded by project ref: CGL2009-10741, by the Spanish Ministry of Science and Innovation and EU-FEDER funds, EU 7th framework HUNTing for Sustainability project (212160, FP7-ENV-2007-1), and the project OAPN 352/2011 from the Spanish Organismo Autónomo Parques Nacionales. J. Caro had a postdoctoral contract financed by the European Social Fund (ESF) and the Junta de Comunidades de Castilla-La Mancha (Operational Programme FSE 20072013), and M. Delibes-Mateos a JAE-doc contract funded by CSIC and the ESF. P. Monterroso provided helpful support with the statistical analyses. 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 appropriate permits 123 by the concerned institutions. CAPÍTULO 6: Drivers of red fox (Vulpes vulpes) daily activity: prey availability, human disturbance or habitat structure? Este capítulo ha sido enviado a una revista SCI: Díaz-Ruiz F, Caro J, Delibes-Mateos M, Arroyo B, Ferreras P (enviado) Drivers of red fox (Vulpes vulpes) daily activity: prey availability, human disturbance or habitat structure? 124 Abstract Daily activity patterns in mammals depend on food availability, reproductive stage, habitat selection, intraspecific interactions and predation risk, among other factors. Some mammals exhibit behavioral plasticity in activity patterns, which allows them to adapt to environmental changes. A good example of this can be found in the red fox (Vulpes vulpes). This species is adapted to living in highly humanized environments, where it is often culled because it may affect human interests (e.g. through the consumption of game species or livestock). We assessed the potential main drivers of the daily activity patterns of the red fox in 12 Iberian Mediterranean areas through the use of camera traps. Among these, we considered main prey availability, degree of human disturbance (e.g. distance to human settlements, and intensity of predator control) and habitat structure. Our results revealed a predominantly crepuscular and nocturnal activity of foxes with local variations. Although overall daily activity of fox increased with rabbit availability, the temporal overlap with prey activity was generally low. In addition, diurnal activity was lower with higher levels of human disturbance (i.e. closer to human settlements) and increased in dense habitats. Prey availability may determine red fox daily activity rhythms in areas with low human disturbance. In contrast, the activity of foxes seems to be determined by other factors like human presence where human disturbance is higher. Our study shows that in highly adaptable species daily activity patterns are determined by several interacting drivers, resulting in complex behavioral patterns. Key words: camera trap, circadian rhythms, human disturbance, fox control, Oryctolagus cuniculus. 125 Introduction Daily activity patterns have been defined as adaptive sequences of routines that meet the time structure of the environment, shaped by evolution and fine-tuned to the actual state of the environment (Halle 2000). According to this, animals may exhibit behavioral plasticity in daily activity patterns to decrease mortality risk, balance energy expenditure and gain, and enhance their fitness (Monterroso et al. 2013). Thus, in mammals, daily activity is internally regulated by species-specific endogenous clocks (Kronfeld-Schor and Dayan 2003), but also by ecological factors such as nutritional requirements (Masi et al. 2009), temporal habitat selection (Chavez and Gese 2006), intraguild interactions (Di Bitetti et al. 2010) or predation risk (Lima and Dill 1990). Additionally, mammals, as well as other animals, show behavioral responses to environmental changes induced by human activities (Tuomainen and Candolin 2011). Predators are strongly constrained by prey availability, which is defined as the combination of prey abundance and their accessibility; prey can be abundant but inaccessible to predators when they are not active or are in inaccessible habitats (Ontiveros et al. 2005). Daily activity patterns in mammalian predators are thus considered mainly the result of innate activity rhythms and a response to prey activity (Giller and Sangpradub 1993), showing in some cases a high level of synchrony with their prey (Foster et al. 2013; Monterroso et al. 2013). Other external factors explaining daily activity patterns of mammalian predators include habitat structure or human disturbance. Predators frequently decrease their activity at daytime in open habitats (Chavez and Gese 2006), where predator control is conducted (Brook et al. 2012) or where human activities such as hunting or outdoor recreational activities are common (Belotti et al. 2012; Ordiz et al. 2012). We chose the red fox (Vulpes vulpes) as a model to study flexibility of mammalian predator daily activity patterns due to its high ecological plasticity and capacity of adaptation to environmental changes. The red fox is the most widely distributed mammalian carnivore of the world and it is found in many different habitats, where it can be abundant and feeds on a large variety of foods (Sillero-Zubiri et al. 2004; DíazRuiz et al. 2013). Although the species is a generalist predator, European wild rabbits (Oryctolagus cunniculus) are the most profitable prey in the central-southern Iberian Peninsula (Delibes-Mateos et al. 2008a; Díaz-Ruiz et al. 2013) where foxes include rabbits in their diet according to their abundance (Delibes-Mateos et al. 2008b). Red 126 foxes have adapted to living in highly humanized environments, where they take advantage of human subsidiary resources (Bino et al. 2010). In addition, it is often persecuted by humans because it preys on game species and livestock (Sillero-Zubiri et al. 2004). Daily rhythms of activity are among the least studied aspects of the ecology and biology of red foxes. Different studies have shown that red foxes are mainly nocturnalcrepuscular, a pattern that can be explained by ecological factors such as season, habitat and prey (Blanco 1986; Cavallini and Lovari 1994; Monterroso et al. 2013). Additionally, red fox activity may also be influenced by human activities such as livestock husbandry (Villar et al. 2013) or road traffic (Baker et al. 2007). Notwithstanding, to our knowledge no studies have examined the simultaneous influence of ecological (e.g. habitat and prey availability) and human-related factors on red fox activity. In Spain the red fox is a game species that can also be legally culled outside the hunting season with a special permit. Direct shooting and live trapping with cage traps and neck snares are the methods most used for legal culling (Delibes-Mateos et al. 2013; DíazRuiz and Ferreras 2013). Fox control is a widespread game management tool in Spain, where between 70-94% of hunting estates perform predator control, mainly targeted to red fox (Díaz-Ruiz and Ferreras 2013). In areas where predator control is carried out more intensively foxes are exposed to a higher ‘risk of predation’ by humans resulting from their capture or death (Reynolds and Tapper 1996). Thus, fox control could be related to stronger fox behavioral responses to human presence in these areas. In this sense, it is known that when hunting constitutes an important source of mortality in a given species, human presence itself may create a ‘landscape of fear’ and thereby provoke strong behavioral responses, as it happens in brown bears (Ursus arctos) (Martin et al. 2010; Ordiz et al. 2012). We evaluated the plasticity of red fox daily activity in environments with varying levels of prey availability and human disturbance (e.g. fox control and distance to human settlements) in Mediterranean areas from central Spain. According to previous studies on mammal predator’s activity we expected that foxes would adapt their activity pattern to that of their main prey, when this was available, but that this behavioral pattern could be disrupted in function of factors, such as habitat composition or human disturbance. 127 To assess this, we first tested whether the activity patterns of the red fox were related to the daily activity of its preferred prey (European wild rabbit). Secondly, we tested the relationships between the daily activity of red foxes and prey availability, human disturbance and habitat structure simultaneously. Material and Methods Study area The study was conducted in 12 localities within central Spain (Figure 6.1.), with Mediterranean-continental climate characterized by hot and dry summers, cold winters and most rainfall occurring during autumn-spring months (Rivas-Martínez et al. 2004). The landscape was heterogeneous and dominated by Mediterranean scrubland (mainly Cistus spp. in combination with holm oak Quercus ilex forests), mixed with cereal croplands and permanent crops such as olive groves (Olea europaea) and vineyards (Vitis vinifera) and natural pastures. Other less abundant habitats included riparian habitats, ‘dehesas’ (pastureland with savannah-like open tree layer, mainly dominated by Mediterranean evergreen oaks) and plantations of pine (Pinus spp.), eucalyptus (Eucalyptus spp.) and poplar (Populus spp.). Villages and scattered dwellings were interspersed in the landscape. Surface and habitat composition varied among localities (see Table 6.1. for a detailed description). Figure 6.1. Situation of the study localities (1-12) in the Iberian Peninsula. 128 Agriculture and livestock were the main economic activities in all localities, which were also hunting estates, with the exception of two protected areas (numbers 5 and 11 in Figure 6.1.), where hunting was not allowed. The main small game species were European wild rabbit (hereafter rabbit), red-legged partridge (Alectoris rufa) and Iberian hare (Lepus granatensis). Hunting estates were managed to improve small game populations, mainly by the provision of supplementary food and water, and predator control. The intensity of fox control varied among hunting estates (Table 6.1., and see below). Table 6.1. Description of study localities. The predominant landscape (agriculture or scrubland) is indicated along with the habitat types present in each area: Oa: open areas, Scr: scrubland, Wc: woody crops, Rip: riparian, Fo: forest, Dh: dehesa. ‘Red fox control’ refers to the number of foxes culled per square km and year. ‘Cameras’ indicate the number of cameratraps used in each locality. ‘Effort’ (survey effort) is expressed as camera-days, or the sum of days each camera was active in the field in each locality. Study site Area (Map ID) (km2) Locality use 1 20 Hunting estate 2 16 Hunting estate 3 50 Hunting estate 4 35.8 Hunting estate 5 21.4 Protected area 6 15.6 Hunting estate 7 21.4 Hunting estate 8 20 Hunting estate 9 9 Hunting estate 10 9 Hunting estate 11 26 Protected area 12 16 Hunting estate Landscape (Habitat types) Agricultural (Oa, Scr, Rip, Wc) Scrubland (Oa, Scr, Rip) Agricultural (Oa, Scr, Rip, Wc) Agricultural (Oa, Scr, Rip, Wc) Scrubland (Oa, Scr, Rip, Dh, Fo) Scrubland (Oa, Scr, Rip, Wc) Agricultural (Oa, Scr, Rip, Dh) Agricultural (Oa, Scr, Rip, Wc) Scrubland (Oa, Scr, Rip, Dh, Fo) Agricultural (Oa, Scr, Rip) Scrubland (Oa, Scr, Rip) Scrubland (Oa, Scr, Rip, Dh, Fo) 129 Red fox control Sampling Cameras Effort (foxes km-2 Year -1 year ) 0.08 2010 20 620 1.98 2010 15 424 0.89 2011 18 493 0.43 2011 17 485 0 2011 19 682 1.30 2011 20 645 0 2012 20 495 4.00 2012 20 503 0.10 2012 15 417 2.70 2012 14 372 0 2012 20 529 0.70 2013 18 463 Camera trap surveys Camera-trap surveys were carried out between 2010 and 2013 in late spring and summer (May-September, Table 6.1.), after the red fox breeding season (Blanco 1998) and when rabbits reach their highest annual numbers in the Iberian Peninsula (Blanco and Villafuerte 1993). We used two models of infrared-triggered digital cameras: Leaf River IR5 (LeafRiver OutDoor Products, Taylorsville, Mississippi, USA) and HCO ScoutGuard (HCO OutDoor Products, Norcross, Georgia, USA). Camera stations were regularly deployed with an approximate distance of 1.2 km among neighboring cameras, ensuring independence between them (Monterroso 2013). The number of camera traps deployed in each study locality varied from 14 to 20, proportionally to locality surface (range: 9-35.8 Km2 ; Table 6.1.). Cameras were mounted on trees approximately 0.5–1.0 m off the ground and set to record time and date when triggered. Cameras operated 24 h a day for an average period of 28.4±0.4 days (mean±SE). We programmed cameras with the minimum time delay between consecutive photos to maximize the number of photos taken per captured individual, and so assure the species identification of each event. In order to increase the detection probability of red fox, we set the sensitivity of the infrared sensor at the highest level, and used Valerian scent and Iberian lynx (Lynx pardinus) urine as lures. This combination has been described as an effective attractant for the red fox (Monterroso et al. 2011). Between 3 and 4 ml of each lure were put in two independent perforated plastic vials secured to a metal rod. Lures were set at 2-3 m from each camera trap, and were replenished every two weeks, when cameras were inspected to check the batteries and to replace memory cards. Consecutive images of the same species within 30 min interval were considered as the same event (unless animals were clearly different individuals) and those separated by a longer interval as independent events (Kelly and Holub 2008; Davis et al. 2011; Monterroso et al. 2013; Delibes-Mateos et al. 2014). Relationship between fox and rabbit activity patterns We studied the activity patterns of both red foxes and rabbits (its main prey) to estimate the probability of both species concurring in a time period. Probability density functions of activity for both species were estimated non-parametrically for each locality from their detection records using kernel density estimates (Ridout and Linkie 2009). Density 130 functions were only estimated in species and localities with > 10 records. We also estimated for each locality the coefficient of overlap Δ1 as suggested by Ridout and Linkie (2009) and Linkie and Ridout (2011) for small sample sizes. The coefficient of overlap ranges from 0 (no overlap) to 1 (complete overlap). The precision of this estimator was obtained through confidence intervals, as percentile intervals from 500 bootstrap samples (Linkie and Ridout 2011). These analyses were performed in R 3.0.1 (R Core Development Team 2013), using an adaptation of the scripts developed by Linkie and Ridout (2011) available at <http://www.kent.ac.uk/ims/personal/msr/overlap.html>. Relationship between fox activity, rabbit availability, human disturbance and habitat structure Records of red fox activity were assigned to one of three time periods (Monterroso et al. 2013): i) twilight (between one hour prior to one hour after sunrise and sunset); ii) diurnal; and iii) nocturnal periods, taking into account the time of sunset and sunrise in each study site during the sampling period. A rabbit availability index was calculated as the number of independent detections of rabbits per 100 trap days in each camera station (Monterroso 2013). Distance to human settlement has been frequently used as a proxy of human disturbance (Ordeñana et al. 2010; Ohashi et al. 2013). We calculated the distance (in kilometers) to the nearest human settlement from each camera using a Geographic Information System (QGIS 1.8.0; QGIS Development Team 2013). Fox control data were gathered through face-to-face interviews with game managers of each hunting estate, conducted before field sampling (at the end of the regular hunting season, in February). We asked managers about the number of foxes removed in the previous hunting season (Table 6.1.). We estimated intensity of fox control as the number of foxes removed per km2 and year (fox·year-1·km-2), and used it as another index of human disturbance. Activity patterns could vary between areas dominated by habitats with high vegetation cover (i.e. shelter for foxes) and those occupied by open habitats (i.e. without shelter). Hence, we grouped habitat types in two main categories: dense habitats (including scrubland, forests and riparian habitats) and open habitats (including ‘dehesas’, pasture 131 and crops). Habitat types surrounding each camera trap were identified from CORINE land-cover 2006 and updated satellite orthophotos (Instituto Geográfico Nacional, <http://www.ign.es/>) and checked during field works. Using QGIS 1.8.0, we calculated the percentage of each habitat type (i.e. open versus dense) within a buffer of 200 m radius around each camera trap (Ordeñana et al. 2010; Monterroso 2013). Either open or dense habitat was assigned to each camera-trap according to the prevailing category (>50%) within the buffer. Generalized Linear Mixed Models (GLMM) were employed to assess red fox activity as a function of time period (day, twilight and night), rabbit availability, human disturbance (fox control intensity and distance to human settlement) and habitat type. The response variable was the number of independent red fox detections for each camera in a given time period, fitted to a Poisson distribution and a log link function was used. We calculated the trapping effort in each camera for each period and locality as follows: trapping effort = nº camera-days × period duration in hours. Trapping effort was included as an offset in models. Camera trap identity was included as a random effect nested within study locality, to account for the non-independence of observations according to these factors. Fixed explanatory effects included: time period and habitat as categorical variables; distance to human settlement, intensity of fox control and rabbit availability as continuous variables; and all two-way interactions between time period and other variables. Analyses were carried out with R 3.0.1 with lme4 package (Bates and Maechler 2010; R Core Development Team 2013). We performed all possible combinations of these independent effects, as all of those models were biologically plausible. For this purpose we used the function dredge (library MuMIn; Bartoń 2012), selected the models with delta ΔAICc<2, and if no single model accounted for >90 % of the total model weights we calculated model-averaged parameter estimates for the variables included in those models (Burnham and Anderson 2002). We assessed whether models were affected by overdispersion, accepting dispersion parameter levels between 0.5 and 1.5 (Zuur et al. 2009). We checked for potential collinearity and redundancy of the explanatory variables by analysing the Variable Inflation Factor (VIF), eliminating variables with VIF values greater than 10 (Belsley et al. 1980). 132 Results During a total effort of 6128 trap-days (mean±SE: 511±27 trapping days·locality-1 ; Table 1) (all means are presented±SE), we obtained 610 independent detections of red foxes (51±14 detections·locality-1) and 1190 of rabbits (99±37 detections·locality-1 ; Table 6.2.). Table 6.2. Number of independent detections of red fox and rabbit and coefficient of overlap (Δ1) of daily activity patterns of red fox and rabbit in each locality. CI95% is the 95% bootstrap confidence interval. Study site Nº Red fox (Map ID) detections 1 17 2 4 3 35 4 77 5 38 6 22 7 17 8 39 9 89 10 48 11 180 12 44 Nº Rabbit detections 48 101 343 176 108 18 0 12 0 357 16 11 ∆1 0.48 0.33 0.43 0.60 0.49 0.46 0.26 0.24 0.35 CI 95% (0.33-0.67) (0.31-0.52) (0.36-0.56) (0.39-0.66) (0.36-0.72) (0.29-0.63) (0.15-0.32) (0.25-0.49) (0.11-0.56) Red fox activity patterns and overlap with rabbit activity Red foxes were detected in all the studied localities (Table 6.2.). Fox activity density functions slightly varied among localities, but in general, two major activity peaks occurred, one after sunset and another before sunrise (Figure 6.2a.). A preliminary test showed significant differences among the three defined time periods (Kruskal-Wallis test, H=25.73, p<0.001): activity was most intense during twilight, followed by nighttime and day-time (mean±SE: 1.02 ± 0.22, 0.79±0.11 and 0.17±0.03 detections·100 trapping-hours-1, respectively). This is in agreement with the results obtained in the second more complex approach using GLMMs (see below). Rabbits were detected in most localities (Table 6.2.). Rabbit activity density functions were similar in all localities, revealing a strong bimodal pattern, with a major activity peak occurring after sunrise and throughout the morning and a second peak before 133 sunset (Figure 6.2b.). Rabbit activity significantly differed among periods (KruskalWallis test, H=34.88, p<0.001): activity was most intense at day-time, followed by twilight-time and night-time (mean±SE: 1.47±0.42, 0.98±0.33 and 0.22±0.06 detections·100 trapping-hours-1, respectively). Figure 6.2. Kernel densities of red fox (a) and rabbit (b) activity in study localities (mean: solid line; range: dashed lines). Vertical dashed lines represent approximate sunrise and sunset times. Coefficients of overlap were estimated in nine localities with enough detection of rabbits and foxes (Table 6.2.). Coefficient of overlap between red fox and rabbit activity patterns varied widely among localities, ranging from 0.24 to 0.60 (0.40±0.04; Table 6.2. and Figure 6.3.), and it was not correlated with rabbit availability (Pearson´s correlation= 0.42, p>0.05). 134 Figure 6.2. Kernel densities of red fox (a) and rabbit (b) activity in study localities (mean: solid line; range: dashed lines). Vertical dashed lines represent approximate sunrise and sunset times. Rabbit availability, human disturbance and habitat structure as factors explaining red fox activity patterns Predictor variables showed VIF values below 10 (VIF values <1.26), and therefore all variables were included in the analysis. Five of the evaluated models showed ΔAICc <2, involving a total weight of 0.70 (Table 6.3.). None of these models were affected by overdispersion (dispersion parameter levels: 0.67-0.69). All these models included all the fixed variables, except fox control, which was not included in two of the selected models (Table 6.3.). Interactions between time period and the remaining fixed variables 135 were also included in the selected models (Table 6.3.). The most important variables explaining fox activity were time period, rabbit availability, distance to human settlement and habitat type, and the interactions between time period and either rabbit availability or habitat type (Table 6.4.). Fox control and other interactions between variables were less important to explain the variability in daily activity of foxes (relative importance < 0.6; Table 6.4.). Table 6.4. Model averaged coefficients and standard errors of the variables included in the five best models explaining the red fox activity (number of independent red fox detections for each camera in a given period). ‘RI’ is the relative variable importance from model average, ‘Time’ is the time period (day, night or twilight), ‘Distance’ is the distance to human settlement, and ‘Rabbit’ is the availability of rabbits. Variable Intercept Time: Twilight Time: Day Fox Control Habitat: Dense Distance Rabbit Twilight*Fox Control Twilight*Dense habitat Twilight*Distance Twilight*Rabbit Day*Fox Control Day*Dense habitat Day*Distance Day*Rabbit Estimate -3.576 0.031 -1.469 -13.1 -0.046 0.159 0.004 3.923 0.192 -0.001 0.001 -22.9 0.910 -0.076 -0.002 SE 0.455 0.264 0.347 19.2 0.253 0.062 0.001 11.06 0.237 0.043 4·10-04 11.5 0.223 0.037 0.001 z 7.860 0.119 4.232 0.686 0.181 2.736 3.870 0.369 0.811 0.036 2.037 1.979 4.082 2.043 2.273 RI 1 1 0.46 1 1 1 0.20 1 0.59 1 0.20 1 0.59 1 P value <0.001 0.905 <0.001 0.492 0.856 0.009 <0.001 0.712 0.417 0.971 0.041 0.047 <0.001 0.041 0.023 Model-averaged parameter estimates revealed that red fox activity was in general lowest during day-time, and increased with rabbit availability except during daylight (Twilight*Rabbit interaction and Day*Rabbit interaction, Table 6.4.; Figure 6.4a.). Diurnal activity of red foxes increased in dense habitats (Day*Dense habitat interaction, Table 6.4., Figure 6.4.). Red fox activity increased with increasing distance to human settlements (Table 6.4.; Figure 6.4b.), although that trend was less marked during daytime (Day*Distance interaction, Table 6.4.), when overall activity was lower anyway (Figure 6.4b.). Overall fox activity did not change strongly with fox control, but diurnal 136 activity decreased where fox control was more intense (Day*Fox control interaction, Table 6.4.; Figure 6.4c.). Figure 6.4. Model-averaged relationships between red fox activity (expressed as detections·100 trapping-hour-1) and: a) Rabbit availability (rabbits·100 trapping-day-1), b) Distance to human settlements (km), and c) Fox control (fox·year-1·km-2) during the three periods of the daily cycle (day, twilight and night) at two different habitat types (dense or open). For plotting the results, data were back-transformed. 137 Discussion Our results indicate that the red fox is mainly crepuscular and nocturnal in our study areas (Figure 6.2a.). This is in agreement with previous studies (Blanco 1986; Servin et al. 1991; Cavallini and Lovari 1994) and supports that the red fox is ‘facultative nocturnal’ (Monterroso 2013). As most canids, the red fox has specific evolutionary adaptations to the night (sight, hearing, smell, Sillero-Zubiri 2009). However, our results support that it is not phylogenetically constrained to nocturnality, and we found differences in the activity patterns associated with the different factors. During our study period (May-September) rabbits showed a main peak of activity in the first hours of the day and a slight peak about sunset (Figure 6.2b.), i.e. they were mainly diurnal, unlike foxes, which were mainly crepuscular and nocturnal (Figure 6.2a.). This means that the overlap between red fox and rabbit activity was in general low (mean coefficient of overlap=0.40) compared with that described for other mammalian predator-prey examples (mean coefficients of overlap>0.60; Foster et al. 2013; Monterroso et al. 2013). Therefore, our results disagree with the opportunistic hunting theory, which states that predators adjust their activity in order to reduce the foraging energy expenditure (Sunquist and Sunquist 1989); i.e. adapting their activity to that of their main prey species (Foster et al. 2013). Nevertheless, this partial lack of synchrony between predator and its main prey has been previously reported by Arias-Del Razo et al. (2011) and Monterroso et al. (2013), who interpreted this as a response of prey to reduce predation risk. This low overlap between rabbit and fox activity patterns may suggest that prey is not the most important factor explaining variations in red fox activity patterns. However, other results show that it has an important influence on them. For example, our findings also showed that the overall activity of red foxes was higher where rabbits were more available, especially during twilight. In that time period, the overlap between rabbit and fox activity was overall highest, and rabbits might be thus more accessible for foxes. Accordingly, a greater temporal overlap between fox and rabbit would be expected in areas with higher availability of rabbits. However, this was not observed in our study, as there was no relationship between the overlap index and the availability of rabbits. These results could indicate that red foxes do not need a high synchrony with rabbits where the latter are abundant, and/or that prey-predator patterns may be altered by 138 human disturbance, as it has been also suggested for wolves (Canis lupus) and moose (Alces alces) in Scandinavia (Eriksen et al. 2009; 2011). In our study, red fox activity was lower in areas closer to human settlements, particularly during twilight and night, the time of highest fox activity. Several studies have shown that human disturbance caused by activities such as agriculture, stockbreeding and outdoor leisure activities, which frequently take place in our study areas, affect the activity of mammal predators. For example, Matthews et al. (2006) and Belloti et al. (2012) demonstrated that tourist activities altered the activity patterns of black bears (Ursus americanus) in Yosemite National Park (USA) and Eurasian lynxes (Lynx lynx) in the Czech Republic, respectively. The effect of human disturbance on predator behavior is especially evident when hunting is an important source of mortality in a given species. In such case, just human presence may create strong behavioral responses through fear (Martin et al. 2010), which is in accordance with our results. Red fox culling by humans has been globally identified as an important cause of mortality in the red fox (Sillero-Zubiri et al. 2004). From this point of view, an effect of predator control on the activity pattern of the target species could be expected. For example, in some areas with intense predator control, canids decrease their activity, especially during the daylight period (Kitchen et al. 2000; Rasmussen and Macdonald 2011; Brook et al. 2012, but see Monteverde and Piudo 2011). However, in our study red fox decreased its overall activity and particularly its activity in daytime in areas with intense fox control (thus with higher direct mortality risk), but the importance of fox control on fox activity was lower than that of other factors in our study. It is possible that the effect of human activity would be stronger during the hunting season (Ciuti et al. 2012; Ohashi et al. 2013) than during our field sampling period but this has not been tested yet. The lack of a strong behavioral response of foxes to predator control intensity, together with the high influence of human presence on fox activity, could indicate that “fear to humans” could be an intrinsic behavior in foxes, accentuated by the historical persecution of this canid by humans in our study area (Vargas 2002). The circadian variations in habitat use by hunted species in human-modified landscapes are possibly a response to human presence (Sunquist 1989; Chavez and Gese 2006; Martin et al. 2010). Therefore, anti-predator behavior in terms of avoidance of human disturbance may explain the observed increase in fox diurnal activity in dense habitats 139 (Figure 6.4), which would be safer for the canid. In agreement with this, several studies have reported that red foxes in rural areas select habitats dominated by dense vegetation during daytime even with human presence (Cavallini and Lovary 1994; Reynolds and Tapper 1995; Janko et al. 2012). Our results show that the red fox presents a high degree of behavioral plasticity adjusting its daily activity rhythms to different ecological scenarios. In this sense, rabbit availability seems to drive fox daily activity rhythms in a scenario of low human disturbance where foxes mainly track rabbits, increasing their diurnal activity. However, where foxes are close to urbanized areas or culled, human disturbance may determine the activity of red foxes, which is reduced during daytime. Our findings show how wildlife adapts to different environmental conditions, including human disturbance, contributing reliable information about an adaptive species such as the red fox. Acknowledgements We are very grateful to land owners, game managers, game keepers and hunters who allowed us to work in their hunting estates, and to the staff of Cabañeros National Park and Ruidera Natural Park. Special thanks to people who assisted us during the fieldwork. This study was funded by project ref: CGL2009-10741, by the Spanish Ministry of Science and Innovation and EU-FEDER funds, EU 7th framework HUNTing for Sustainability project (212160, FP7-ENV-2007-1), and the project OAPN 352/2011 from the Spanish Organismo Autónomo Parques Nacionales. J. Caro had a postdoctoral contract financed by the European Social Fund (ESF) and the Junta de Comunidades de Castilla-La Mancha (Operational Programme FSE 20072013), and M. Delibes-Mateos a JAE-doc contract funded by CSIC and the ESF. 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 appropriate permits, by the concerned institutions. 140 DISCUSIÓN GENERAL Existen numerosos trabajos sobre la temática tratada en esta Tesis Doctoral realizados en diferentes ecosistemas de todo el mundo. No obstante, un alto porcentaje han sido desarrollados en sistemas muy simplificados, lo cual hace difícil aplicar sus resultados y conclusiones a sistemas diversos y de mayor complejidad como los presentes en la Península Ibérica. En este sentido, los resultados obtenidos en esta Tesis Doctoral aportan nueva información sobre diferentes aspectos de la ecología y gestión de depredadores generalistas en ecosistemas complejos, como los ibéricos, en donde el conocimiento científico es más escaso. La mayor parte de los trabajos de esta Tesis han sido desarrollados en ambientes Mediterráneos de la Península Ibérica, considerados entre los de mayor biodiversidad a nivel mundial (Blondel y Aronson 1999), y que ocupan la mayor parte de la Península Ibérica (Rivas- Martínez 1987; Rivas-Martínez et al. 2004). Teniendo esto en cuenta, la información aportada en esta Tesis puede ser de utilidad para mejorar la gestión de dos especies generalistas ampliamente distribuidas y generalmente abundantes como son el zorro y la urraca. Ecología trófica del zorro y la urraca Es fundamental estudiar los hábitos alimentarios de los depredadores para comprender su ecología, así como para entender el papel que desempeñan en los procesos ecológicos que ocurren en las comunidades de las que forman parte. Además, los hábitos alimentarios de los depredadores son la principal causa por la cual gran parte de estos son perseguidos por el hombre (Woodroffe et al. 2005) y, por lo tanto, conocerlos bien es de vital importancia para la gestión de estas especies. Tanto el zorro común (Vulpes vulpes, zorro en adelante) como la urraca (Pica pica) son especies ampliamente distribuidas y abundantes en gran parte de España, por lo que el estudio de su alimentación es especialmente relevante para evaluar su posible impacto sobre algunas presas de interés para la conservación, así como sobre especies cinegéticas de interés socioeconómico (Herranz 2000; Ruiz-Olmo et al. 2003; Fernandez de Simón 2013). Aunque puede pensarse que la alimentación de estas dos especies es un aspecto de su ecología suficientemente conocido, algunas cuestiones no han sido tratadas hasta la fecha. Los capítulos 1 y 2 de esta Tesis Doctoral aportan nueva información sobre la ecología trófica de estas especies que puede ser útil para mejorar la gestión de sus poblaciones. 141 Para una mejor comprensión de las estrategias tróficas de los depredadores a nivel de especie es fundamental su estudio a escala biogeográfica, en la que se tienen en cuenta un amplio rango de condiciones ambientales derivadas de la distribución de la especie (Daan y Tinbergen 1997). En las últimas décadas se han llevado a cabo estudios de este tipo con varias especies de carnívoros de tamaño medio como por ejemplo la gineta (Genetta genetta) (Virgós et al. 1999), la nutria (Lutra lutra) (Clavero et al. 2003) o el gato montés (Felis silvestris) (Lozano et al. 2006). Este tipo de estudios han supuesto una mejora substancial en el conocimiento sobre la ecología trófica de estas especies. Por ejemplo, han puesto de manifiesto que el gato montés, considerado como un especialista en micromamíferos en latitudes septentrionales, se alimenta en gran medida de conejos en latitudes meridionales (Oryctolagus cuniculus). De este modo, gracias a estos estudios, el gato montés ha sido reconsiderado como un depredador especialista facultativo, que adapta su alimentación a presas energéticamente más rentables cuando éstas son abundantes (Lozano et al. 2006). El zorro ha sido definido como un depredador generalista de amplio espectro trófico, que utiliza los recursos alimentarios de forma oportunista en función de su disponibilidad o abundancia (Macdonald y Reynolds 2004). Esta consideración está apoyada por los resultados obtenidos en numerosos trabajos realizados mayormente a escala local o regional, mientras que no existe ningún trabajo a escala biogeográfica similar a los citados anteriormente sobre otros carnívoros. El capítulo 1 de esta Tesis Doctoral analiza la ecología trófica del zorro a escala biogeográfica en la Península Ibérica a través de una exhaustiva revisión bibliográfica. Los resultados de este capítulo demuestran el carácter generalista y oportunista del zorro a esta escala, ya que su alimentación está relacionada con variables geográficas (latitud, longitud y altitud) y ambientales (hábitat y estacionalidad), que determinan en gran medida la presencia y abundancia de sus principales alimentos. Concretamente la dieta del zorro presenta un marcado patrón latitudinal, alimentándose principalmente de conejos e invertebrados en el sur de la Península Ibérica, mientras que en el norte su dieta está dominada por micromamíferos, frutos y semillas. Patrones similares han sido descritos para otros carnívoros de mediano tamaño como el tejón (Meles meles), el gato montés y la gineta (Virgós et al 1999; Virgós et al. 2005; Lozano et al. 2006). El conejo es una especie clave en ecosistemas Mediterráneos de la Península Ibérica y principal recurso alimentario para un importante número de depredadores ibéricos 142 (Delibes-Mateos et al. 2007, 2008a). Los resultados de esta Tesis demuestran la influencia de los conejos en diferentes aspectos de la ecología del zorro. Por un lado, son una presa preferida, base de la dieta de los zorros en el centro-sur de la Península Ibérica (capítulo 1). Por otro lado, la disponibilidad o abundancia de conejos influye notablemente sobre la ecología espacial del zorro, siendo más probable su presencia en aquellas zonas donde el lagomorfo es abundante (capítulo 5). Finalmente, desde el punto de vista comportamental se ha demostrado que la actividad diaria del zorro es mayor en las zonas donde el lagomorfo presenta una mayor disponibilidad (capítulo 6). Al igual que otras especies generalistas, el zorro presenta respuestas funcionales alimentarias como mecanismo adaptativo ante las variaciones de sus recursos tróficos (Hanski et al. 1991; Panek et al. 2013). En el centro-sur de España se ha observado cómo el consumo de su principal presa, el conejo, es denso-dependiente, es decir aumenta con la abundancia del lagomorfo (Delibes-Mateos et al. 2008b; Fernandez de Simón 2013). Así, cuando el conejo no es abundante el zorro adapta su alimentación incrementando el consumo de otros recursos secundarios (Ferreras et al. 2011). A escala biogeográfica los micromamíferos parecen ser uno de los principales recursos alternativos preferidos por los zorros (capítulo 1). Esto es especialmente interesante desde un punto de vista de conservación de otras especies presa como las aves, ya que, cuando hay disponibilidad de micromamíferos, no parecen ser seleccionadas como principal fuente sustitutoria del conejo. Por el contrario, Ferreras et al. (2011) comprobaron a escala local, en la Reserva Biológica de Doñana, cómo los zorros incrementaron significativamente el consumo de aves y carroñas de ungulados como respuesta a la marcada disminución de las poblaciones de conejo tras la llegada de la enfermedad hemorrágico vírica (EHVc). A pesar de la acentuada disminución de conejos en gran parte de España durante las últimas décadas (Delibes-Mateos et al. 2009), las poblaciones de zorro no parecen mostrar descensos significativos en su abundancia (Sobrino et al 2008; Fernandez de Simón 2013; pero ver Ferreras et al. 2011). Ante este escenario de baja disponibilidad de su presa principal y una abundancia del cánido relativamente constante, posiblemente haya incrementado la presión de depredación del zorro sobre otras presas, como por ejemplo algunas aves terrestres que nidifican en el suelo, para las que el zorro es uno de los principales depredadores (Yanes y Suárez 1996; Herranz 2000; Ruiz-Olmo et al. 2003). Sin embargo, hasta la fecha este aspecto no ha sido estudiado de forma específica. 143 El patrón de alimentación de la urraca ha sido descrito como el de un generalista que utiliza de forma oportunista diferentes recursos en función de su disponibilidad (Birkhead 1991). Los resultados obtenidos en esta Tesis indican que la principal fuente de variabilidad en la dieta de la urraca es la localización geográfica de sus poblaciones, mientras que factores intrínsecos como el sexo y la edad no parecen tener tanto peso en su alimentación (capítulo 2). El patrón de consumo de los distintos grupos tróficos en cada localidad está probablemente relacionado con su disponibilidad. De esta forma el consumo de cereal no varió entre las dos localidades estudiadas, donde la disponibilidad de este tipo de alimento era similar. Sin embargo, las urracas pueden seleccionar ciertos tipos de alimentos independientemente de su disponibilidad, como por ejemplo algunos grupos de invertebrados (Martínez et al. 1992; Kryštofková et al. 2011). Estudios previos han documentado una mayor depredación de nidos artificiales por parte de las urracas durante la fase de incubación (Suvorov et al. 2012), mientras que durante la fase de crianza de los pollos el consumo de invertebrados se incrementa significativamente (Martínez et al. 1992; Ponz et al. 1999). De esta manera, las diferencias observadas entre localidades en el consumo de aves y artrópodos (capítulo 2) podrían estar determinadas por la fase del ciclo reproductor en la que se encuentran las urracas. Además, esta hipótesis podría explicar la diferencia encontrada en la diversidad de la dieta, que sería menor durante la crianza, debido probablemente al elevado consumo de artrópodos (capítulo 2). El papel de la urraca como depredador de pequeños vertebrados, concretamente de aves, y sus huevos, es un aspecto de su ecología muy discutido. Numerosos trabajos muestran cómo este córvido consume aves y huevos durante la época de reproducción, aunque estos representan una proporción muy baja en su dieta (Birkhead 1991). En ese sentido, los resultados obtenidos en el capítulo 2 muestran cómo las urracas se alimentan principalmente de semillas de cereal y artrópodos en ambientes agrícolas durante la reproducción. En esta época los artrópodos son la principal fuente de proteínas para las urracas (Martínez et al. 1992). Cuando la disponibilidad de éstos es baja, las urracas utilizan, probablemente de forma secundaria, otras fuentes de proteína animal como pueden ser las aves o sus huevos, incrementándose bajo estas circunstancias la probabilidad de depredación (Birkhead 1991; capítulo 2). El uso masivo de pesticidas en la agricultura actual podría haber incrementado la presión de depredación sobre aves 144 por parte de las urracas, al disminuir la disponibilidad de invertebrados. No obstante, hacen falta estudios adicionales para corroborar esta hipótesis. Se encontraron restos de cáscara de huevo y plumas atribuibles a perdiz roja en tan solo dos mollejas, perteneciendo la mayor parte de restos de aves y huevos a paseriformes (capítulo 2). A priori los resultados del capítulo 2 pueden ser indicativos de que las urracas no suponen un problema para la dinámica poblacional de otras aves, tal y como se ha descrito para las poblaciones de varias especies de paseriformes (Gooch 1991; Chiron y Julliard 2007). Sin embargo, otros estudios identifican a la urraca como uno de los principales depredadores de nidos de diferentes aves, incluidos los de perdiz roja (Groom 1993; Herranz 2000; Roos y Pärt 2004; Ferreras et al. 2010), por lo que es posible que se subestime el consumo de huevos en los estudios de dieta. Esto hace que no se pueda descartar que la depredación de nidos por la urraca pueda representar un riesgo para el éxito reproductor de la perdiz roja en un escenario de alta abundancia de urracas y bajas densidades de perdiz. En estas condiciones, incluso una pequeña cantidad de huevos depredados podría representar un gran impacto en el éxito reproductor de la población de perdiz. Evaluación y mejora de los métodos de captura para el control de zorros y urracas Los métodos empleados para el control de depredadores generalistas suscitan controversia entre diferentes sectores. Por un lado, los conservacionistas consideran que los métodos hasta ahora utilizados no son selectivos, y que eliminan de forma ilegal individuos de especies de interés para la conservación (Virgós et al. 2010). Igualmente algunos autores consideran que los criterios para determinar si los métodos se adecúan o no a los estándares de captura no cruel no son suficientes ni adecuados (Iossa 2007). Por otro lado, los cazadores consideran que los métodos permitidos para controlar depredadores generalistas abundantes son escasos y poco eficaces (Delibes-Mateos et al. 2013). La resolución exitosa de este tipo de conflictos se produce cuando el resultado es aceptable para las distintas partes y ninguna de ellas hace valer sus intereses en detrimento de los de los demás (Redpath et al. 2013). Bajo esta perspectiva, la prohibición total del control de depredadores no sería la mejor manera de minimizar los conflictos entre cazadores, gestores de fauna y conservacionistas en relación a la gestión 145 de los depredadores generalistas. Por lo tanto, parece fundamental trabajar en la mejora de los métodos de control existentes así como en el desarrollo de nuevos sistemas que permitan un control selectivo y eficiente de especies generalistas que bajo ciertas condiciones pueden llegar a ser abundantes, como es el caso del zorro y la urraca. De hecho, la mejora de los métodos legales para el control de depredadores ha sido incluida en varias Estrategias de Conservación de algunos depredadores amenazados como el lince ibérico (Lynx pardinus) o el águila imperial ibérica (Aquila adalberti), como medida de lucha contra el uso de métodos ilegales y masivos como los cebos envenenados (Fernández-Olalla 2011), que han repuntado en España en los últimos años (Martínez-Abraín et al. 2013). El estudio de diferentes sistemas para el control de zorros y urracas quizás sea el aspecto sobre el control de depredadores al que mayor esfuerzo se ha dedicado en los últimos años en España (Díaz-Ruiz y Ferreras 2013). Esto se debe principalmente a la obligatoriedad del cumplimiento de los estándares establecidos en diferentes tratados internacionales en cuanto a la eficiencia de captura de las especies objetivo, la selectividad y los daños relacionados con la captura, necesarios para la homologación de los métodos y su uso legal. Sin embargo, y hasta la fecha, el esfuerzo dedicado a los métodos para el control de zorro ha sido muy superior al dedicado a los métodos para urraca (Díaz-Ruiz y Ferreras 2013). Esto probablemente se deba a que por un lado el control de dicho carnívoro es más habitual que el control de córvidos y a que por otro lado los métodos habitualmente empleados para zorro son menos efectivos y selectivos que los empleados para la urraca, y por tanto suscitan una mayor polémica. En el caso del zorro, se han evaluado en España dos métodos tradicionales (jaulastrampa y lazos) y dos nuevos sistemas de captura desarrollados en Norteamérica (Belisle y Collarum) (Díaz-Ruiz y Ferreras 2013). Según estos estudios, las jaulastrampa son poco eficaces y poco selectivas (Tabla 1). Sin embargo, pocos trabajos han evaluado diferentes variantes en su uso o modificaciones con vistas a incrementar su eficiencia de captura y selectividad. Hasta la fecha se ha probado el uso de diferentes cebos (vivos o muertos) (Herranz 2000; Ferreras et al. 2003, 2007), la combinación de cebos y atrayentes olorosos (Ferreras et al. 2003, 2007) y la incorporación de una apertura circular en las puertas de la jaula-trampa para facilitar la salida de especies de menor tamaño que el zorro (Junta de Andalucía 2010) (Tabla 7.1.). 146 En el capítulo 3 de esta Tesis se han analizado de forma conjunta los datos obtenidos en trabajos realizados en Castilla-La Mancha entre 2003 y 2007. El análisis de estos datos muestra cómo el uso combinado de cebo vivo con orina de zorro como atrayente oloroso incrementa la eficiencia de captura de zorros de las jaulas-trampa para esta especie. También se pudo comprobar un ligero descenso en la tasa de captura de especies no objetivo al utilizar valeriana como atrayente oloroso. Por lo tanto, el uso combinado de atrayentes y cebos puede mejorar la eficiencia de captura y selectividad de las jaulas-trampa. Sin embargo, no se encontró una combinación de cebos y atrayentes que consiguiera de forma simultanea incrementar la eficiencia de captura y la selectividad de las jaulas-trampa, y en cualquier caso la selectividad sigue siendo muy baja (< 25%), aunque hay que ser cautos con estos resultados, debido el pequeño tamaño de muestra. Los resultados de este capítulo coinciden con los de anteriores trabajos en desaconsejar la homologación de las jaulas-trampa para su uso como método de control de zorros en los cotos de caza. En cualquier caso estos resultados podrían servir para mejorar otros métodos destinados a la captura de zorros. Estudios previos han demostrado cómo el sistema Collarum (lazo propulsado de cuello) es eficiente y selectivo para la captura de coyotes (Canis latrans) en Norteamérica (Shivik et al. 2000, 2005). Los distintos ensayos realizados en España con este sistema empleando una versión específica para zorros, incluidos los descritos en el capítulo 3, han obtenido resultados similares en cuanto a su selectividad para capturar cánidos y una eficiencia aceptable para capturar zorros, que puede mejorarse modificando el diámetro de cierre del lazo (Tabla 7.1.). Aunque se ha señalado la posibilidad de utilizar esta trampa en zonas con presencia de especies amenazadas como el lince ibérico (Muñoz-Igualada et al. 2008), en la actualidad se desconoce el riesgo de captura para esta especie, así como para otros carnívoros amenazados como el lobo (Canis lupus) o el oso pardo (Ursus arctos). Se necesitan por tanto pruebas específicas en condiciones controladas (sin el lazo, con registro en video) en zonas con presencia continuada de las citadas especies amenazadas, o pruebas previas en cautividad que ayuden a esclarecer si la trampa es inocua para estas especies. Se considera que otros métodos como los lazos de acero tradicionales (con y sin tope), así como sus versiones norteamericanas (lazo “americano” y Wisconsin; ver Herranz 2000 y Muñoz-Igualada 2010 respectivamente) son efectivos para capturar zorros y muestran una mayor selectividad que las jaulas-trampa, siempre y cuando sean 147 instalados de forma correcta (Tabla 7.1.). Sin embargo, es necesario matizar que los lazos sin tope, están totalmente prohibidos en la actualidad por producir graves lesiones y sufrimiento innecesario a los animales capturados (Herranz 2000; Duarte et al. 2012). A pesar de que los lazos con tope son un método muy eficaz para capturar zorros, su homologación en ciertas comunidades autónomas ha suscitado polémica al no considerarse lo suficientemente selectivo (Barrull et al. 2011). El sistema de captura Belisle (lazo de pie) ha sido evaluado en dos trabajos, que indican una buena eficiencia de captura de zorros y una selectividad mayor que las jaulas-trampa, aunque no se ha considerado suficiente para poder ser homologado (Tabla 7.1.). De forma general los lazos con tope, el sistema Belisle y el Collarum para zorro cumplen con los estándares internacionales de captura no cruel, como demuestra que más del 80% de los animales capturados en los distintos estudios no presentaron lesiones graves (Muñoz-Igualada et al. 2008, 2010; Junta de Andalucía 2010). En el caso de las jaulas-trampa, la mayoría de trabajos realizados en España no han podido evaluar de forma conveniente los daños relacionados con las capturas de zorros, debido al bajo número de capturas conseguidas (< 20; Tabla 1). Sin embargo, la mayor parte de estos trabajos encuentran una baja frecuencia de lesiones graves tanto en zorros como en especies no objeto de control (Herranz 2000; Muñoz-Igualada et al. 2008; Junta de Andalucía 2010). Estos resultados coinciden con los descritos para las jaulas trampa usadas para capturar coyotes en Norteamérica (Way et al. 2002; Way 2012). Los resultados del capítulo 3 son similares a los obtenidos en trabajos previos para las jaulas-trampa y el sistema Collarum, con una baja frecuencia de lesiones graves. Sin embargo, el bajo número de capturas conseguidas con ambos sistemas impidió un análisis pormenorizado similar al realizado en otros trabajos (Muñoz-Igualada et al. 2008; Junta de Andalucía 2010). En cualquier caso para minimizar los daños es fundamental la revisión de las trampas al menos una vez cada 24 horas, conclusión en la que todos los trabajos coinciden. 148 Tabla 7.1. Resumen de la información recopilada en los estudios que han evaluado los sistemas de captura de zorro (JT: jaula-trampa, entre paréntesis el nº de entradas; LT: lazo tradicional; LA: lazo americano; LW: lazo Wisconsin; Bel.: Belisle; Coll.: Collarum). Los parámetros utilizados en cada estudio han sido estandarizados en función de la información recopilada. Esfuerzo: expresado en trampas-noche. Zorros: número de zorros capturados. No buscadas: número de capturas de especies distintas al zorro. Eficiencia de captura: individuos capturados/1000 trampas-noche. Selectividad ISO: expresado como el % de zorros capturados con respecto al total de capturas conseguidas. En negrita aparecen los resultados totales (TOTAL) para cada sistema de captura además de la media±ES para los parámetros Eficiencia de captura y Selectividad ISO (*). En la localidad de estudio se indica el número de localidades (N) donde se han evaluado los diferentes métodos de captura y la Comunidad Autónoma (AN: Andalucía; CLM: Castilla-La Mancha; CL: Castilla y León) (modificado de Díaz-Ruiz y Ferreras 2013). A. JAULAS-TRAMPA Eficiencia de captura de No buscadas Zorro Capturas Referencia Modelo Localidad Esfuerzo Zorros Herranz (2000) 1 Duarte y Vargas (2001) 3 Ferreras et al. (2003) 2 Moleón et al.(2003) 2 Ferreras et al. (2007) 2 Muñoz-Igualada et al. (2008) 3 Junta de Andalucía (2010) 2 JT(1) JT(1) JT(2) JT JT(1-2) JT JT(1-2) JT1-2) JT(2) JT(2) JT(2) JT(2) JT(2) JT(2)a N=1 CLM N=8 CLM N=1 CLM N=1 AN N=1 CLM N=1 AN Muñoz-Igualada et al. (2010) Junta de Andalucía (2010) 2 Duarte et al. (2012) TOTAL 3 3 ISO % 2576 2596 363 2160 927 1558 1117 736 515 540 140 127 409 417 14180 B. LAZOS 1 3 1 5 1 6 5 0 0 3 0 0 0 1 26 19 86 42 61 7 25 22 12 13 2 1 1 16 7 314 0.39 1.16 2.75 2.31 1.08 3.85 4.48 0.00 0.00 5.56 0.00 0.00 0.00 2.40 1.71 ± 0.50* 5 3 2 8 13 19 19 0 0 60 0 0 0 13 9.58 ± 4* 14 506 27 21 1.86 56 13 610 9838 22 21 5 8 1.62 2.13 81 72 LW-al paso N= 2CLM 8550 21 9 2.46 70 LW-al paso 5363 8 1 1.49 89 22 292 12 10 0.54 55 8568 13 7 1.52 65 32 319 124 61 1.66 ± 0.63* 69.8 ± 26.38* N=2 CLM N=4 CL N= 2 AN TOTAL Herranz (2000) 1 Selectividad LT/LA(con y N=9 CLM sin tope) LT-alar N= 2CLM LW-alar N= 2CLM N= 1 AN LW-alar LT-al paso (sin N= 1 AN tope) 149 Muñoz-Igualada et al. (2008) 3 Bel. N=4 CL Junta de Andalucía (2010) 2 Bel. N= 1 AN C. BELISLE 538 574 406 317 13 10 3 1 5 2 4 2 24.16 17.42 7.39 3.15 72 83 43 33 1537 8 3 5.20 73 35 16 11.47 ± 4.01* 60.89 ± 9.64* 1 2 10 8 2 2 1 20 46 1 0 1 1 0 0 0 0 3 2.76 2.15 18.69 14.23 5.57 6.73 1.24 9.72 7.64 ± 2.20* 50 100 91 89 100 100 100 100 91.22 ± 6.11* TOTAL 7372 D. COLLARUM 363 Ferreras et al. (2007) 2 Coll. N=2 CLM 929 535 562 3 Muñoz-Igualada et al. (2008) Coll. N=4 CL 359 297 Coll. 809 2 Junta de Andalucía (2010) N= 2 AN Coll.a 2057 TOTAL 5911 1 : Tesis Doctoral; 2: Informe técnico; 3:Artículo científico. a: Modelos modificados Las jaulas-trampa con reclamo vivo de urraca es uno de los métodos más empleados en España para el control de las poblaciones de este córvido, y los cazadores españoles lo consideran como un método eficaz para reducir las poblaciones de estas aves (DelibesMateos et al. 2013). A pesar de ello, antes del trabajo que constituye el capítulo 4 de esta Tesis Doctoral este método no había sido evaluado teniendo en cuenta los estándares internacionales anteriormente citados. Dos breves ensayos previos habían mostrado resultados contradictorios sobre la eficiencia de este sistema para capturar urracas. Así, Herranz (2000) no consiguió capturar ninguna urraca, mientras que Martínez de Castilla y Martínez (2004) lo consideran un método muy eficiente; además, ninguno de los dos trabajos aporta información sobre su selectividad y sobre los daños derivados de las capturas. En el capítulo 4 se evalúa por primera vez de forma experimental este método de captura para urracas atendiendo a criterios de eficiencia, selectividad y captura no cruel establecidos en los tratados internacionales. Los resultados indican que es un método eficiente y muy selectivo (98%) para la captura de urracas durante su época de reproducción. Además, se trata de un método de captura no cruel puesto que ninguno de los animales capturados mostró ninguno de los indicadores de malestar establecidos en los estándares. Al igual que en el caso de los métodos para zorros, es imprescindible la revisión de las trampas al menos una vez cada 24 horas. Hasta la fecha este método ha sido evaluado principalmente en zonas agrícolas. Solamente existe una breve prueba realizada en una zona mixta de monte mediterráneo y dehesas agrícolas durante época post-reproductora y con baja densidad de urracas, en la que se capturaron dos ginetas y ninguna urraca (Ferreras et al. 2007). 150 Por lo tanto, se necesitan nuevos estudios en escenarios más heterogéneos, con mayor diversidad y abundancia de otras especies susceptibles de ser capturadas, que permitan contrastar los resultados obtenidos en este trabajo. Efectos del control de depredadores sobre las poblaciones de zorros y urracas El principal objetivo del control de depredadores es disminuir el impacto de éstos sobre algunas especies presa, asumiendo por lo general que la extracción de depredadores conlleva una disminución efectiva en la abundancia de sus poblaciones. Gran parte de los trabajos existentes sobre control de depredadores evalúan el efecto sobre las poblaciones de presas que se pretenden fomentar, prestando menor atención al efecto sobre las poblaciones de la especie controlada (Smith et al. 2010). La evaluación experimental de la efectividad de las extracciones de zorros para reducir sus poblaciones es complicada, debido en parte a la dificultad de realizar estimas fiables de la abundancia de las poblaciones de los carnívoros (incluido el zorro), que requieren metodologías costosas y sofisticadas (Heydon et al. 2000; Schauster et al. 2002). El foto-trampeo es un método de muestreo alternativo que, combinado con el uso de nuevas y potentes herramientas de análisis estadístico, permite caracterizar el estado de la poblaciones y detectar los cambios asociados a la gestión en especies poco abundantes y elusivas, como los mamíferos carnívoros (Sarmento et al. 2011; Towerton et al. 2011; Cove et al. 2012; Schuette et al. 2013). Recientemente se ha utilizado el foto-trampeo combinado con modelos de ocupación (del inglés Occupancy models, Mackenzy et al. 2006) para determinar el efecto de las campañas de control de zorros sobre sus poblaciones y las de sus potenciales presas en Australia (Towerton et al. 2011). En dicho trabajo no se encontraron diferencias en la probabilidad de ocupación espacial de los zorros ni en sus índices de actividad tras las campañas de control. Los resultados obtenidos en el capítulo 5 son similares a estos, ya que no se observó ninguna relación entre la intensidad de extracción (control) de zorros y la probabilidad de ocupación. Igualmente, los resultados obtenidos en esta Tesis Doctoral han puesto de manifiesto la ausencia de relación entre la intensidad del control de zorros y la actividad diaria de este carnívoro (capítulo 6), al igual que lo observado en Australia (Towerton et al. 2011). Sin embargo, en el capítulo 5 se encontró una clara reducción en la probabilidad de detección de zorros en zonas donde el control era 151 intenso, lo que contradice la asunción de Towerton et al. (2011) sobre la asunción de una probabilidad de detección de zorros similar en periodos pre y post extracción de zorros. Recientemente se ha sugerido que la probabilidad de detección de una especie está directamente relacionada con su abundancia (McCarthy et al. 2013). Si esta relación existiera realmente en el caso de los zorros, los resultados del capítulo 5 indicarían que el control de depredadores, a partir de una cierta intensidad de extracción y con cierta continuidad temporal, podría disminuir la abundancia local de las poblaciones de zorro. Sin embargo, tanto los resultados de Towerton et al. (2011) como los obtenidos en los capítulos 5 y 6 sugieren que las estimas de ocupación y los índices de actividad posiblemente no sean buenas alternativas para evaluar cambios en las abundancias de las poblaciones de zorro, ya que probablemente reflejan cambios espaciales y temporales en el uso de los territorios. A pesar de ello, no se conoce con exactitud la relación entre la abundancia real de determinadas especies, como en este caso el zorro, y estos índices (estimas de ocupación, índices de actividad y probabilidad de detección, estimados a partir de datos de foto-trampeo), lo que motivo frecuente de discusión entre científicos (Anderson 2003; Royle y Nichols 2003; Mackenzie y Nichols 2004; McCarthy et al. 2013; Sollman et al. 2013). Se necesitan, por lo tanto, nuevos estudios basados en estimas fiables de la abundancia real de las poblaciones que ayuden a probar la validez de estas nuevas metodologías de monitoreo para determinar, entre otros aspectos, las consecuencias de determinadas medidas de gestión de fauna, como control de zorros. Dichos estudios contribuirán a la mejora en la toma de decisiones para la gestión de las poblaciones de las especies hacia las que se dirige la gestión. A diferencia de lo descrito previamente para los zorros, la abundancia de ciertas aves como las urracas puede estimarse de forma más precisa mediante diferentes métodos; entre otros destacan el conteo de nidos en época reproductora (Stoate y Szuczur 2001, 2005), o el método de muestreo de distancias (del inglés distance sampling) que permite obtener estimas de densidad absoluta (Newson et al. 2008). A pesar de ello, no existen muchos trabajos que hayan evaluado el efecto de las extracciones de urracas sobre sus poblaciones. En general indican que el control durante la época de reproducción es efectivo para la reducción de la abundancia de sus poblaciones a escala local y regional (Stoate y Szuczur 2001, 2005; Chiron y Julliard 2007). Igualmente en España, Herranz (2000) observó que mediante la caza de urracas adultas y la destrucción de sus nidos en 152 un coto cinegético se consiguió una rápida y significativa reducción de la abundancia del córvido. Los resultados obtenidos en el capítulo 4 muestran que las extracciones experimentales realizadas con jaulas-trampa pueden reducir a corto plazo y a escala local la densidad de urracas en lugares donde éstas son abundantes. Sin embargo, la respuesta de las poblaciones de urracas tras el cese de las extracciones fue diferente en las dos localidades de estudio. Probablemente estas diferencias se deban a las fechas de inicio y fin del trampeo, desarrollado durante diferentes fases del ciclo reproductor de la especie en cada una de las localidades. Se ha descrito que el control a lo largo de todo el ciclo reproductor consigue una disminución en la población mantenida a lo largo del tiempo, ya que se extraen parte de los individuos flotantes que rápidamente colonizan los territorios vacantes (Chiron y Julliard 2007). De esta forma los resultados del capítulo 4 sugieren que el trampeo desarrollado tan solo desde la puesta (inicio del trampeo) hasta la incubación (final del trampeo) permite la incorporación de individuos flotantes que probablemente completen el ciclo reproductor, contribuyendo a la recuperación de la población. Sin embargo, el trampeo desarrollado en fases más avanzadas del ciclo reproductor (eclosión-crianza de los pollos) no permitiría completar de forma exitosa la reproducción a los individuos flotantes incorporados a la población, manteniéndose la abundancia baja. Recientemente se ha descrito que el control de urracas intensivo y continuado en el espacio y en el tiempo a escala regional puede disminuir drásticamente las poblaciones de este córvido, llegando en algunos casos a existir riesgo de extinción local (Chiron y Julliard 2013). En ningún caso esto debería ser el objetivo o el resultado de cualquier plan de gestión de un depredador generalista autóctono, como el zorro o la urraca. Efectos sobre otras especies no objeto de control Por lo general se suele asumir o esperar un efecto del control de depredadores sobre las poblaciones de las especies que son objeto directo de la gestión, es decir, el depredador que es controlado, así como de la(s) presa(s) que se pretenden recuperar o fomentar (Saunders et al. 2010; Smith et al. 2010). Sin embargo, el control de depredadores puede afectar a otras especies que no son contempladas en los planes de gestión, como por ejemplo otros depredadores (Virgós y Travaini 2005), presas secundarias (Henke y Bryant 1999) o especies que dependen de la especie controlada para completar su ciclo reproductor (Martínez 2011). 153 El control de depredadores desarrollado en gran parte de los cotos de caza españoles podría jugar un papel importante en la composición y estructura de las comunidades de carnívoros presentes en los mismos. Por ejemplo, se ha sugerido que el control no selectivo e ilegal puede provocar una disminución de las abundancias de carnívoros y de la riqueza específica de sus comunidades (Virgós y Travaini 2005; Beja et al. 2009). La extracción ilegal de especies como la garduña o el tejón podría producir reducciones significativas en sus poblaciones (Barrull et al. 2014) ya que estas especies presentan menores tasas reproductivas y menor capacidad de dispersión que el zorro (Casanovas et al 2012). Igualmente la extracción ilegal de competidores podría beneficiar al zorro que al ver reducida la competencia podría incrementar su abundancia y expandir sus poblaciones como se ha mostrado en Reino Unido con la extracción de tejones (Trewby et al. 2008). Se estaría produciendo, por lo tanto, un efecto contrario al buscado con el control de sus poblaciones (Lozano et al. 2013). Dentro de las comunidades de mesocarnívoros ibéricos el zorro podría desempeñar un papel de competidor dominante sobre algunas especies simpátricas de menor tamaño, como la garduña (Pereira et al. 2012; Monterroso 2013). De hecho, se ha citado que los zorros pueden incluso dar muerte a individuos del género Martes sp. (Palomares y Caro 1999). Bajo estas condiciones, la extracción intensiva de zorros podría beneficiar a otros mesocarnívoros simpátricos subordinados, como la garduña. Los resultados del capítulo 5 están en concordancia con esto, ya que se observó que la extracción de zorros aumentaba la probabilidad de ocupación por parte de la garduña. Sin embargo, el papel que desempeña el zorro dentro de las comunidades de mesocarnívoros ibéricos no está claro, desconociéndose en gran medida las interacciones ecológicas entre estas especies (Monterroso et al. 2013). Se necesitan, por lo tanto, nuevos estudios sobre las interacciones ecológicas intragremiales que ayuden a conocer los posibles efectos de las medidas de control de zorros sobre la estructura de las comunidades de mesocarnívoros. En el caso de la urraca se desconoce por completo las consecuencias que el control intensivo de sus poblaciones pueda tener sobre otras especies. Así, el control podría perjudicar indirectamente al críalo (Clamator glandarius), un ave parásita de los nidos de urraca, que depende en gran medida de este córvido para completar su ciclo reproductor (Martínez 2011), o incluso a otras aves que utilizan los nidos abandonados de urraca para criar. Igualmente se desconoce el efecto del control sobre especies potencialmente competidoras de la urraca como podría ser el cernícalo común (Falco 154 tinnunculus) u otros córvidos como el arrendajo (Garrulus glandarius) y el rabilargo (Cyanopica cyanea) con los que comparte ciertas preferencias ecológicas (Chirón y Julliard 2007; Alonso 2010; Palomino et al. 2011). En estos casos podría existir un efecto de “liberación” de competidores como lo anteriormente descrito para algunas especies de mesocarnívoros. La evaluación de estos aspectos es, por lo tanto, imprescindible en futuros trabajos que pretendan completar el conocimiento sobre el control poblacional de este córvido. Efectos sobre el comportamiento de los depredadores objeto de control Diferentes trabajos indican que especies habitualmente cazadas pueden modificar sus patrones de comportamiento espacio-temporal como respuesta al riesgo de muerte que pueden suponer los humanos, siendo estos cambios más acentuados durante la época hábil de caza (Ordiz et al. 2012; Ohashi et al. 2013). No obstante, en algunos casos la simple presencia humana puede provocar igualmente fuertes respuestas comportamentales en estas especies a través del miedo a ser matados (Martin et al. 2010). Por ejemplo, en algunas zonas con intenso control de depredadores se ha observado como diferentes cánidos disminuyen su actividad diurna haciéndose más nocturnos (Kitchen et al 2000; Rasmussen y Macdonald 2011; Brook et al 2012). A pesar de estos ejemplos, uno de los aspectos menos estudiados sobre el control de depredadores es como este puede influir sobre el comportamiento de la especie que es objeto de control. En el caso del zorro, los humanos (y sus actividades) constituyen una de las principales causas de mortalidad para la especie a escala mundial (Sillero-Zubiri et al. 2004). Desde este punto de vista, cabría esperar cambios en el patrón de actividad de esta especie asociados a la intensidad del control de depredadores. Los resultados obtenidos en el capítulo 6 indican que la actividad del zorro es principalmente nocturna y crepuscular, y que la actividad diurna se reduce en zonas donde el control de depredadores es intenso. Los resultados de este capítulo también muestran que la actividad de los zorros es mayor en zonas con poca presencia humana. Esto sugiere que el “miedo a los humanos" podría ser un comportamiento intrínseco en el zorro, relativamente independiente del grado de persecución de estos, derivado de la larga historia de persecución que este cánido ha experimentado en gran parte de España (Vargas 2002). 155 Respecto a las urracas, Birkhead (1991) observó un comportamiento más esquivo de las mismas ante la presencia humana en zonas donde éstas eran habitualmente cazadas. Probablemente los cambios en el comportamiento de estas dos especies generalistas derivados de las actividades humanas no suponen una amenaza directa para su supervivencia y conservación, a diferencia de lo señalado para otras especies como el oso pardo (Ordiz et al. 2012). Sin embargo, se desconocen las consecuencias directas o indirectas que estos cambios en la actividad puedan suponer sobre otras especies, así como en el funcionamiento de los ecosistemas de los que forman parte. Futuras líneas de investigación La comprensión de la ecología de depredadores generalistas abundantes es fundamental para poder establecer medidas de gestión adecuadas que permitan conjugar la conservación de los ecosistemas con un uso sostenible de los recursos naturales presentes en estos. Los resultados obtenidos en esta Tesis Doctoral aportan información de valor sobre esta temática, aunque la información existente sigue siendo escasa y a menudo poco concluyente. De los resultados de este trabajo destacan algunos aspectos que deberían estudiarse en mayor detalle en el futuro con vistas a mejorar la gestión de estas especies. A continuación se proponen algunas líneas futuras de investigación suscitadas a raíz de esta Tesis Doctoral: El estudio de la alimentación de los depredadores generalistas, sigue siendo un aspecto fundamental para mejorar la gestión de estas especies. En el caso del zorro se hacen necesarios estudios a largo plazo que desvelen, entre otras cuestiones, el papel de los recursos alimentarios secundarios en la dinámica poblacional del cánido, y en la de sus principales presas. En el caso de la urraca son necesarios trabajos que determinen su papel como depredador de huevos y aves; en este sentido es fundamental combinar estudios experimentales en cautividad y campo para determinar hasta qué punto el consumo de estos tipos de alimento es subestimado por las metodologías convencionales (análisis de contenidos de mollejas o de egagrópilas). Para ninguna de las dos especies generalistas objeto de estudio en esta Tesis, se ha podido comprobar si pueden llegar a especializarse en el consumo de algún tipo de alimento en concreto a nivel específico. En este sentido diferentes trabajos muestran la importancia de la especialización individual en la 156 explotación de ciertos recursos alimentarios por parte de algunos depredadores generalistas y cómo esto puede ser determinante en la estructuración y dinámica de las comunidades de presas (Oro et al. 2005; Prught et al. 2008; Araújo et al 2011; Elbroch y Wittmer 2013). Estudios sobre si existe especialización trófica en ciertos grupos de presas a nivel de individuo en especies como el zorro y la urraca pueden aportar información relevante para mejorar la gestión de sus poblaciones y de algunas de sus presas (especialmente para especies amenazadas), como se ha mostrado en algunas especies de láridos a través del control selectivo de individuos (ver Sanz-Aguilar et al. 2009). Es necesario potenciar la investigación sobre la eficacia de medidas de gestión alternativas al control letal de depredadores. Para ello se necesitan estudios sobre la eficacia de estas actuaciones en la reducción del impacto de depredación de algunos depredadores generalistas. A continuación se presentan ejemplos de algunas medidas sobre las que sería interesante investigar: - Papel del acceso a fuentes de alimentación subsidiarias de origen antrópico para depredadores generalistas (Stone y Trost 1991; Bino et al. 2010), cuya eliminación o control podría servir para reducir las abundancias de los mismos. - Recuperación y fomento de depredadores apicales (Prugh et al. 2009). Sería necesario investigar sobre el papel de depredadores como el lobo y el lince ibérico en la estructuración de las comunidades tanto de otros depredadores como en las poblaciones de presas. - Condicionamiento aversivo por sabor. Sería deseable investigar si este método provoca una reducción de la tasa de depredación sobre ciertas especies (Marguire et al. 2009), y si permite mejorar la selectividad de algunos métodos de captura (Phillips y Whinche 2011). - Control de la fertilidad mediante anticoncepción y/o esterilización. Recientes estudios muestran que estas medidas pueden ser eficaces para reducir el impacto de depredación sobre ciertas presas (coyotes en Norteamérica, Seidler et al. 2014) y pueden ser relativamente eficaces para el control del tamaño poblacional de las poblaciones de zorro (McLeod y Saunders 2014). 157 Aunque el control letal debe ser una medida excepcional, es esperable que su uso a corto o medio plazo siga siendo habitual tanto en programas de conservación como de gestión de especies cinegéticas (Redpath et al. 2013), al menos hasta que se encuentren mediadas alternativas efectivas. Aunque se ha comprobado que es posible mejorar los sistemas de captura para depredadores generalistas (Phillips y Whinchell 2011; Short et al. 2012; capítulos 3 y 4), aún es necesario aclarar un buen número de aspectos. Por ejemplo, en esta Tesis Doctoral se ha demostrado que se puede aumentar tanto la efectividad como la selectividad de las cajas-trampa usando combinaciones de cebos y atrayentes y oloroso. Sin embargo, no se ha podido determinar una combinación de estos que permita mejorar la efectividad y selectividad de estos dispositivos. Por ello, es importante seguir investigando en este y otros aspectos que permitan la mejora de los métodos existentes, así como en el desarrollo de nuevos métodos que cumplan con los diferentes criterios establecidos para la captura no cruel. Se ha comprobado como las extracciones de urraca durante la época de reproducción son eficaces para reducir la abundancia de sus poblaciones a corto plazo. Las diferentes respuestas de las dos poblaciones de urracas estudiadas tras el cese del trampeo plantean la hipótesis de que un control desarrollado durante el ciclo reproductor completo, o al menos durante la fase de crianza de los pollos, podría reducir la abundancia de forma duradera. Serían necesarios nuevos trabajos experimentales a largo plazo que evalúen esta hipótesis. Igualmente, sería necesario evaluar las consecuencias ecológicas del control sobre otras especies no objetivo vinculadas a la dinámica poblacional de este córvido, como por ejemplo el críalo europeo. Los resultados de esta Tesis sugieren que la intensidad del control de zorros podría disminuir la abundancia de los mismos y a un mismo tiempo desencadenar procesos ecológicos como la “liberación de competidores”. Son necesarios estudios experimentales para confirmar estos resultados, e igualmente desvelar el efecto de esta medida de gestión sobre la diversidad y composición de las comunidades de carnívoros. Se necesitan nuevos estudios que permitan esclarecer la relación entre los parámetros estimados mediante metodologías de occupancy a partir de datos de 158 foto-trampeo, como por ejemplo la probabilidad de ocupación y la detectabilidad, con la abundancia real de las poblaciones de las especies estudiadas, ya que este aspecto fundamental es todavía causa de debate en la comunidad científica. CONCLUSIONES 1. Las variaciones en los hábitos de alimentación de los zorros ibéricos están relacionadas con variables geográficas, tipos de hábitat y estacionalidad, que a su vez determinan la disponibilidad de sus principales alimentos. Por lo tanto, la flexibilidad trófica de este depredador refleja los patrones biogeográficos de la distribución y abundancia de sus principales fuentes de alimento. No se encontró ninguna relación significativa de la diversidad de la dieta con las variables estudiadas. Estos resultados confirman al zorro como un depredador generalista y oportunista a una escala biogeográfica mayor de lo que se había descrito hasta ahora. 2. Aunque no llegan a especializarse en ninguno de sus principales recursos alimentarios, los zorros en la Península Ibérica consumen conejos como alimento principal en aquellos lugares donde estos son abundantes, y como principales presas sustitutorias los micromamíferos, frutos y semillas cuando el lagomorfo no es abundante. 3. La alimentación de la urraca durante su periodo reproductor en ambientes agrícolas del centro-sur de la Península Ibérica, se basa principalmente en artrópodos (mayormente coleópteros) y cereales. Las urracas incluyen en su dieta huevos y aves con baja frecuencia y en baja proporción, por lo que su impacto de depredación sobre este tipo de presas no parece importante. No obstante, se desconocen los posibles sesgos asociados a la metodología de estudio para la estima del consumo de estos alimentos. 4. Tanto el consumo de los principales grupos alimentarios como la diversidad de la dieta de la urraca varió entre localidades, sin una influencia clara de factores intrínsecos como el sexo y la edad. El patrón de alimentación observado coincide con el de una especie generalista que utiliza los recursos en función de 159 su disponibilidad, aunque se podría explicar también en parte por la fase del ciclo reproductor. 5. El uso combinado de cebos y atrayentes puede mejorar la eficiencia de captura y selectividad de las jaulas-trampa para capturar zorros. La combinación cebo vivo-orina de zorro incrementó de forma significativa la eficiencia para capturar zorros, mientras que el uso de extracto de valeriana consiguió disminuir ligeramente la tasa de captura de especies no objeto de control. Sin embargo, ninguna combinación de cebos y atrayentes de las ensayadas permitió de forma simultanea incrementar la tasa de capturas de zorros y disminuir la de especies no buscadas, no alcanzándose en ningún caso los umbrales mínimos de selectividad establecidos como requisitos para su homologación. 6. El sistema de captura Collarum mostró una mayor selectividad que la obtenida para las jaulas-trampa y una aceptable eficiencia de captura de zorros, superior a la de las jaulas-trampa sin atrayentes. Los resultados obtenidos con este sistema indican que es una alternativa aceptable a métodos tradicionales como las jaulastrampa para el control poblacional de zorros en cotos de caza de características similares a los estudiados. 7. Las jaulas-trampa con reclamo de urraca viva, utilizadas durante la época de reproducción son un método eficaz y muy selectivo para el control de las poblaciones de urracas en ambientes agrícolas donde el córvido es abundante. Estos resultados cumplen los estándares de captura establecidos para poder utilizar estas trampas como método de control en medios agrícolas; sin embargo se desconoce su funcionamiento en ambientes más complejos, donde exista una mayor probabilidad de capturas de especies no buscadas. 8. Los sistemas de captura evaluados en esta Tesis no produjeron lesiones consideradas como indicadores de malestar a ninguna de las especies objetivo, tanto zorro como urraca. En cualquier caso es fundamental que en las campañas de control efectuadas con cualquier sistema de captura, todas las trampas instaladas sean revisadas al menos una vez cada 24 horas para evitar sufrimiento innecesario a los animales capturados. 160 9. Las extracciones de urracas consiguieron disminuir a corto plazo las densidades del córvido en las dos localidades de estudio. Sin embargo, la respuesta de las poblaciones tras el cese del control fue distinta entre ambas localidades. Se observó una recuperación de la población tras el cese del control cuando las extracciones se realizaron en las primeras fases del ciclo reproductor (puestaincubación), mientras que cuando se realizaron en fases más avanzadas de la reproducción (eclosión-crianza de los pollos) la población se estabilizó tras el cese del control a densidades menores. 10. La probabilidad de ocupación espacial por parte de los zorros no se vio afectada por la intensidad del control de sus poblaciones, estando determinada principalmente por el tipo de hábitat predominante. Por el contrario, la probabilidad de detección de zorros disminuye con el incremento en la intensidad de su control. Si se confirmase la relación positiva entre detectabilidad y abundancia sugerida por algunos autores, estos resultados sugerirían que la intensidad de control podría disminuir la abundancia de zorros. 11. La intensidad del control de zorros estuvo relacionada con el incremento en la probabilidad de ocupación espacial de la garduña. Estos resultados sugieren que el control intensivo de zorros puede desencadenar procesos de “liberación de competidores” debido a la disminución numérica del zorro. Al mismo tiempo, los resultados aportan nueva información sobre el papel desempeñado por cada especie en una relación competitiva intragremial. De esta forma el zorro desempeñaría un papel de competidor dominante y la garduña el de competidor subordinado. 12. El zorro mostró una actividad principalmente crepuscular y nocturna solapando parcialmente con la actividad de su principal presa, el conejo. Las variables más relacionadas con su actividad fueron la disponibilidad de conejos y la presencia humana, independientemente de la intensidad de control de zorros. La disponibilidad de conejo parece determinar los ritmos de actividad diaria en situaciones de baja perturbación humana, en las que los zorros incrementan su actividad diurna para aumentar el solapamiento con la actividad de los conejos. Sin embargo, la actividad de los zorros se reduce durante el día cerca de zonas urbanizadas o donde el zorro es sometido a intenso control, como consecuencia 161 de las perturbaciones humanas. Estos resultados demuestran que el zorro presenta una elevada plasticidad comportamental que le permite adaptarse de forma exitosa a diferentes condiciones ambientales, incluyendo la perturbación humana. 162 REFERENCIAS Akaike H (1973) Information theory and an extension of the maximum likelihood principle. En: Petrov BN, Csaki F (Eds.) 2nd international symposium on information theory. Akademiai Kiado, Budapest, Pp 267–281 Albone ES, Fox MW (1971). Anal gland secretion of the red fox. Nature 233: 569–570 Alonso CL (2010) Arrendajo – Garrulus glandarius. En: Enciclopedia Virtual de los Vertebrados Españoles. Salvador, A., Bautista, L. M. (Eds.). Museo Nacional de Ciencias Naturales, Madrid. http://www.vertebradosibericos.org/ Amores F (1975) Diet of the red fox (Vulpes vulpes) in the western Sierra Morena (South Spain). Doñana Acta Vertebrata 2: 221–239 Angelo IS (2000) Aspectos da ecologia da raposa (Vulpes vulpes) no Nordeste Algarvio. Relatorio de estagio para obtensçao da Licenciatura em Biologia Apliacada a los Recursos Animais Terrestres. Facultade de Ciências da Universidade de Lisboa, Departamento de Zoología e Antropología. Lisboa Angulo E (2003) Factores que afectan a la distribución y abundancia del conejo en Andalucía. Tesis Doctoral, Universidad Complutense de Madrid Angulo E, Villafuerte R (2003) Modelling hunting strategies for the conservation of wild rabbit populations. Biological Conservation 115: 291–301 Anderson DR (2003) Response to Engeman: index values rarely constitute reliable information. Wildlife Society Bulletin 31: 288–291 Aranda Y, Isern-Valberdú J, Pedrochi C (1995) Dieta estival del zorro Vulpes vulpes L. en pastos del Pirineo Aragonés: Relación con la abundancia de artrópodos. Lucas Mallada 7: 9–20 Araújo MS, Bolnick DI, Layman CA (2011) The ecological causes of individual specialization. Ecology Letters 14: 948–958 Arias-Del Razo I, Hernández L, Laundré JW, Myers O. (2011) Do predator and prey foraging activity patterns match? A study of coyotes (Canis latrans), and lagomorphs (Lepus californicus and Sylvilagus audobonii). Journal of Arid Environments 75: 112–118 Arnold TW (2010) Uninformative parameters and model selection using Akaike's information criterion. Journal of Wildlife Management 74: 1175–1178 Arnold J, Soulsbury CD, Harris S (2011) Spatial and behavioural changes by red foxes (Vulpes vulpes L., 1758) in response to artificial territory intrusion. Canadian Journal of Zoology 89: 808–815 163 Arroyo B, Delibes-Mateos M, Díaz-Fernández S, Viñuela J (2012) Hunting management in relation to profitability aims, red-legged partridge hunting in central Spain. European Journal of Wildlife Research 58: 847–856 Baeyens G (1981) The role of the sexes in territory defence in the Magpie (Pica pica). Ardea 69: 69–82 Baker PJ, Robertson CPJ, Funk SM, Harris S (1998) Potential fitness benefits of group living in the red fox, Vulpes vulpes. Animal Behavior 56: 1411–1424 Baker PJ, Harris S, Robertson CPJ, Saunders G, White PCL (2001) Differences in the capture rate of cage-trapped red foxes Vulpes vulpes and an evaluation of rabies control measures in Britain. Journal of Applied Ecology 38: 823–835 Baker PJ, Harris S (2003) A review of the diet of foxes in rural Britain and a preliminary assessment of their impact as a predator. En: Tattersall F, Manly WM (Eds.) Conservation and Conflict – Mammals and Farming in Britain, 120– 140. Linnean Society Occasional Publication No. 4,Westbury Publishing, Otley, UK Baker PJ, Harris S (2006) Does culling reduce fox (Vulpes vulpes) density in commercial forests in Wales, UK? European Journal of Wildlife Research 52: 99–108 Baker PJ, Dowding CV, Molony SE, White PCL, Harris S. (2007) Activity patterns of urban red foxes (Vulpes vulpes) reduce the risk of traffic-induced mortality. Behavioral Ecology 18: 716–724 Ballesteros T, Degollada A (2002) Dieta de la guineu (Vulpes vulpes) al Parc Natural de Sant Llorenç del Munt i l´Obac. V Trobada d´Estudiosos de Sant Llorenç del Munt i l´Obac. Monografies, 35, Barcelona Balme GA, Slotow R, Hunter LTB (2009) Impact of conservation interventions on the dynamics and persistence of a persecuted leopard (Panthera pardus) population. Biological Conservation 142: 2681–2690 Barea-Azcón J, Virgós E, Ballesteros-Duperón E, Moleón M, Chirosa M (2007) Surveying carnivores at large spatial scales: a comparison of four broad-applied methods. Biodiversity and Conservation 16: 1213–1230 Barea-Azcón JM, Ballesteros-Duperón E, Gil-Sánchez JM, Virgós E (2010) Badger Meles meles feeding ecology in dry Mediterranean environments of the southwest edge of its distribution range. Acta Theriologica 55: 45–52 Barrientos JA (1988) Bases para un curso práctico de entomología. Asociación española de Entomología. Departamento de Biología Animal de la Universidad de Salamanca. 754 Pp. Barrientos R, Virgós E (2006) Reduction of potential food interference in two sympatric carnivores by sequential use of shared resources. Acta Oecologica 30: 107–116 164 Barrull J, Mate I (2007) Alimentació dels mamífers carnívors en el Parc Natural de la Serra de Monsant. III Jornades del Parc Natural de la Serra de Monsant, Octubre 2007. Barcelona Barrull J, Mate I, Casanovas JG, Salicrú M, Gosálbez J (2011) Selectivity of mammalian predator control in managed hunting areas: an example in a Mediterranean environment. Mammalia 75: 363–369 Barrull J, Mate I, Salicrú M, Palet J, Casanovas JG, Gosàlbez J, Ruiz-Olmo J (2014) Differential response of a carnivore community to predator control: a spatiotemporal observational study. Italian Journal of Zoology 81:2, 271–279 Barton K (2012) MuMIn, Multi–model inference. R package version 1.7.2. http,//CRAN.R–project.org/package=MuMIn Bates D Maechler M (2010) lme4: Linear Mixed–effects Models Using S4 Classes. R package version 0.999375–37. http://lme4.r–forge.r-project.org/ Burnham KP, Anderson DR (2002) Model selection and multimodel inference : a practical information-theoric aproach. Springer 2nd edn. Verlag, New York, USA, 488pp Beasley JC, Olson ZH, Beatty WS, Dharmarajan G, Rhodes OE Jr (2013) Effects of culling on mesopredator population dynamics. PLoS ONE 8(3): e58982 Begon M, Harper JL, Townsend CR (1996) Ecology: individuals, populations and communities. Blackwell Science, Oxford Beja P, Gordinho L, Reino L, Loureiro F, Santos-Reis M, Borralho R (2009) Predator abundance in relation to small game management in southern Portugal: conservation implications. European Journal of Wildlife Research 55: 227–238 Belsley DA, Kuh E, Welsch RE (1980) Regression Diagnostics: Identifying Influential Data and Sources of Collinearity. Wiley, New York Belotti E, Heurich M, Kreisinger J, Šustr P, Bufka L (2012) Influence of tourism and traffic on the Eurasian lynx hunting activity and daily movements. Animal Biodiversity and Conservation 35: 235–246 Beltrán JF (1991) Temporal abundance pattern of the wild rabbit in Doñana, SW Spain. Mammalia 55: 591–599 Bermejo T, Guitián J (2000) Fruit consumption by foxes and martens in NW Spain in autumn: a comparison of natural and agricultural areas. Folia Zoologica 49: 89– 92 Bernabeu RL (2000) Evaluación económica de la caza en Castilla-La Mancha. Tesis Doctoral, Universidad de Castilla–La Mancha Bernard J (2008) Normativa española y europea sobre control de predadores. En: Garrido JL (Ed.) Especialista en control de predadores. FEDENCA-ESCUELA ESPAÑOLA DE CAZA, pp 47–64 165 Berry O, Tatler J, Hamilton N, Hilmer S, Hitchen Y, Algar D (2013) Slow recruitment in a red-fox population following poison baiting: a non-invasive mark–recapture analysis. Wildlife Research 40: 615–623 Bielby J, Donnellyb CA, Popec LC , Burkec T, Woodroffea R (2014) Badger responses to small-scale culling may compromise targeted control of bovine tuberculosis. Proceedings of the National Academy of Sciences of the United States of America 111: 9193–9198 Bino G, Dolev A, Yosha D, Guter A, King R, Saltz D, Kark S (2010) Abrupt spatial and numerical responses of overabundant foxes to reduction in anthropogenic resources. Journal of Applied Ecology 47: 1262–1271 Birkhead TR (1991) The Magpies. The ecology and behaviour of Black-billed and Yellow-billed Magpies. T and A D Poyser, Londres. 270 Pp Blanco JC (1986) On the diet, size and use of home range and activity patterns of a red fox in central Spain. Acta Theriologica 31: 547–556 Blanco JC (1988) Estudio ecológico del zorro en la Sierra de Guadarrama. Tesis Doctoral. Universidad de Oviedo Blanco JC, Reig S, Cuesta L (1992) Distribution, status and conservation problems of the wolf Canis lupus in Spain. Biological Conservation 60:73–80 Blanco JC, Villafuerte R (1993) Factores ecológicos que influyen sobre las poblaciones de conejos. Incidencia de la enfermedad hemorrágica. Informe Inédito. Empresa de Transformación Agraria, S.A., Madrid Blanco, J.C. (1998) Mamíferos de España, vol. I & II. Editorial Planeta Blanco-Aguiar JA, García JF, Ferreras P, Viñuela J, Villafuerte R (2001) Effect of game management on artificial nest predation in central Spain. En 25th International Union of Game Biologists (IUGB) and the 9th International Symposium Perdix, Limasol, Chipre Blanco-Aguiar JA, Virgos E, Villafuerte R (2003) Perdiz Roja (Alectoris rufa). En: M. R. y D.M.J. C. (Eds.) Atlas de las aves reproductoras de España (Pp. 212–213). Dirección General de Conservación de la Naturaleza y Sociedad Española de Ornitología. Madrid, España Blanco-Aguiar JA (2007) Variación espacial en la biología de la perdiz roja (Alectoris rufa): una aproximación multidisciplinar. Tesis Doctoral. Universidad Complutense de Madrid Blas J, Perez-Rodriguez L, Bortolotti GR, Vinuela J, Marchant TA (2006) Testosterone increases bioavailability of carotenoids: insights into the honesty of sexual signalling. Proceedings of the National Academy of Sciences of the United States of America 103:18633–18637 Blondel J, Aronson J (1999) Biology and Wildlife of the Mediterranean Region. Oxford University Press, Oxford, UK 166 Boitani L, Fuller T (2000) Research Techniques in Animal Ecology Methods: Controversies and Consequences. Columbia University Press, New York Boitiani L, Ciucci P, Raganella-Pelliccioni E (2010) Ex-post compensation payments for wolf predation on livestock in Italy: a tool for conservation? Wildlife Research 37: 722–730 Borralho R, Rego F, Vaz-Pinto P (1996) Is driven transect sampling suitable for estimating red-legged partridge Alectoris rufa densities? Wildlife Biology 2: 259–268 Boza MD (2002) El trampeo y demás artes de caza tradicionales en la Península Ibérica. Editorial Hispano Europea S.A, Barcelona Brand CJ, Keith LB, Fisher CA (1976) Lynx responses to changing snowshoe hare densities in central Alberta. Journal of Wildlife Management 40: 416–428 Braña F, Del Campo JC (1980) Estudio de la dieta del zorro (Vulpes vulpes L.), en la mitad occidental de la Cordillera Cantábrica. Boletín Ciencias Naturale I.D.E.A. 26: 135–146 Brook LA, Johnson CN, Ritchie EG (2012) Effect of predator control on behavior of an apex predator and indirect consequences for mesopredator suppression. Journal of Applied Ecology 49: 1278–1286 Buitron D (1988) Female and male specialization in parental care and its consequences in black-billed magpies. Condor 90:29–39 Bur MT, Stapanian MA, Bernhardt G, Turner MW (2008) Fall diets of Red-breasted Merganser (Mergus serrator) and Walleye (Sander vitreus) in Sandusky Bay and adjacent waters of western lake Erie. American Midland Naturalist 159: 147– 161 Burnham KP, Anderson RR (1998) Model selection and inference: a practical information-theoretic approach. Springer, New York Burnham KP, Anderson RR, Laake JL (1980) Estimation of density from line transect sampling of biological populations. Wildlife Monographs 72: 1–202 Burnham KP, Anderson DR (2002) Model selection and multimodel inference.2nd edn. New York: Springer Calvete C, Estrada R, Villafuerte R, Osacar JJ, Lucientes J (2002) Epidemiology of viral haemorrhagic disease (VHD) and myxomatosis in free-living population of wild rabbits. Veterinary Record 150: 776–782 Calvete C, Estrada R, Angulo E, Cabezas-Ruíz S (2004) Habitat factors related to wild rabbit conservation in an agricultural landscape. Landscape Ecology 19: 531– 542 167 Calviño F, Canals JLS, Bas S, De Castro A, Guitián J (1984) Régimen alimenticio del zorro (Vulpes vulpes L.) en Galicia, Noroeste de la Península Ibérica. Boletín Estación Central Ecología 13: 83–90 Calzada J (2000) Impacto de la Depredación y Selección de Presa del Lince y el Zorro Sobre el Conejo. Tesis Doctoral. Universidad de León Capinera JL (2010) Insects and Wildlife: Arthropods and their Relationships with Wild Vertebrate Animals.Wiley-Blackwell, Oxford, UK Carvalho JC (2001) Distribuçao e abundancia do Coelho-bravo: influência da qualidade do habitat e dos predadores. Disertação para obtenção do Grau de Mestre em Ciências do Ambiente. Universidade do Minho. Braga Carvalho JC, Gomes P (2004) Foods habits and trophic niche overlap of the red fox, European wild cat and common genet in the Peneda-Gerês National Park. Journal of Zoology London 263: 275–283 Carvalho JC, Cardoso P, Crespo LC, Henriques S, Carvalho R, Gomes P (2011) Biogeographic patterns of spiders in coastal dunes along a gradient of mediterraneity. Biodiversity and Conservation 4: 873–894 Casanovas JG, Barrull J, Mate I, Zorrilla JM, Ruiz-Olmo J, Gosálbez J, Salicru M (2012) Shaping carnivore communities by predator control:competitor release revisited. Ecological Research 27: 603–614 Casas F (2008) Gestión agraria y cinegética: efectos sobre la perdiz roja (Alectoris rufa) y aves esteparias protegidas. Tesis Doctoral. Universidad de Castilla La Mancha Castroviejo J, Palacios F, Garzón J, Cuesta L (1975) Sobre la alimentación de los cánidos ibéricos. XII Congress of the IUGB, Lisboa (Portugal) Caut S, Casanovas JG, Virgós E, Lozano J, Witmer GW, Courchamp F (2007) Rats dying for mice: Modelling the competitor release effect. Austral Ecology 32: 858–868 Cavallini P, Lovari S (1994) Home range, habitat selection and activity of the red fox in a Mediterranean coastal ecotone. Acta Theriologica 39: 279–287 Chacarov N, Krüger O (2010) Mesopredator release by an emergent superpredator: A natural experiment of predation in a three level guild. PLoS ONE 5(12): e15229 Chapman RF (1998) The Insects: Structure and Function. Cambridge University Press, Cambridge, UK Chavez AS, Gese EM (2006) Landscape Use and Movements of Wolves in Relation to Livestock in a Wildland–Agriculture Matrix. Journal of Wildlife Management 70: 1079–1086 Chinery, M (1997) Guía de campo de los insectos de España y de Europa. Ediciones Omega, S.A., Barcelona 168 Chiron F, Julliard R (2007) Responses of songbirds to magpie reduction in an urban habitat. Journal of Wildlife Management 71: 2624–2631 Chiron F, Julliard R (2013) Assessing the effects of trapping on pest bird species at the country level. Biological Conservation 158: 98–106 Ciuti S, Northrup JM, Muhly TB, Simi S, Musiani M, Pitt JA, Boyce MS (2012) Effects of humans on behaviour of wildlife exceed those of natural predators in a landscape of fear. Plos One 7: e50611 Clavero M, Prenda J, Delibes M (2003) Trophic diversity of the otter (Lutra lutra L.) in temperate and Mediterranean freshwater habitats. Journal of Biogeography 30: 761–769 Close B, Banister K, Baumans U, Bernoth EM, Bromage N, Bunyan J (1996) Recommendations for euthanasia of experimental animals, Part 1. Laboratory Animals 30: 293–316 Close B, Banister K, Baumans U, Bernoth EM, Bromage N, Bunyan J, Erhardt W, Flecknell P (1997) Recommendations for euthanasia of experimental animals: Part 2. Laboratory Animals 31:1–32 Conferencia Sectorial de Medio Ambiente (2011) Directrices técnicas para la captura de especies cinegéticas predadoras: homologación de métodos de captura y acreditación de usuarios. Documento aprobado por la Conferencia Sectorial de Medio Ambiente, 13 de julio de 2011 Conover M (2002) Resolving human-wildlife conflicts: the science of wildlife damage management. Lewis Publisher, New York. 418 pp Contesse P, Hegglin D, Gloor S, Bontadina F, Deplazes P (2004) The diet of urban foxes (Vulpes vulpes) and the availability of anthropogenic food in the city of Zurich, Switzerland. Mammalian Biology 69: 81–95 Corbelle-Rico EJ, Rico-Boquete E (2008) La actividad de las juntas de extinción de animales dañinos en España. En Actas de resumen del IX Congreso de la Asociación de Historia Contemporánea, Pag. 1944–1968, Universidad de Murcia, Murcia, España Cove MV, Jones BM, Bossert AJ, Clever DR, Dunwoody RK, White BC, Jackson VL (2012) Use of camera traps to examine the mesopredator release hypothesis in a fragmented Midwestern landscape. American Midland Naturalist 168: 456–465 Crooks KR, Soulé ME (1999) Mesopredator release and avifaunal extinctions in a fragmented system. Nature 400: 563–566 Daan S, Tinbergen JM (1997) Adaptation of life histories. En: Krebs JR, Davies NB (Eds.) Behavioural Ecology: an Evolutionary Approach, 311–333. WileyBlackwell, Oxford, UK Davis ML, Kelly MJ, Stauffer DF (2011) Carnivore co-existence and habitat use in the Mountain Pine Ridge Forest Reserve, Belize. Animal Conservation 14: 56–65 169 Day MG (1966) Identification of hair and feather remains in the gut a faeces of stoats and weasels. J Zool 148:201–217 De Cavalho Alves AS (1995) A Raposa (Vulpes vulpes silacea Miller, 1907) na Regiao de Onor no Parque Natural de Montesinho. Ecologia Trófica, Uso do Espaço e Uso do Tempo. Relatorio de estágio para obtençao da licenciatura em recursos faunísticos e ambiente. Facultade de Ciências da Universidade de Lisboa, (Portugal) Delibes M, Hiraldo F (1981) The rabbit as prey in the Iberian Mediterranean ecosystem. En: Myers K, MacInnes CD (Eds.) Proceedings of the I World Lagomorph Conference. University of Guelph, Ontario, Canada. Pp 614–622 Delibes-Mateos M (2006) Relaciones entre los cambios poblacionales del conejo, la gestión cinegética, el hábitat y los depredadores: implicaciones para la conservación. Tesis Doctoral, Universidad de Castilla-La Mancha Delibes-Mateos M, Redpath S, Angulo E, Ferreras P, Villafuerte R (2007) Rabbits as a keystone species in southern Europe. Biological Conservation 137: 149 –156 Delibes-Mateos M, Delibes M, Ferreras P, Villafuerte R (2008a) Key role of European rabbits in the conservation of the western Maditerranean basin hot spot. Conservation Biology 22: 1106–1117 Delibes-Mateos M, Fernández de Simón J, Ferreras P, Villafuerte R (2008b) Feeding responses of the red fox (Vulpes vulpes) to different wild rabbit (Oryctolagus cuniculus) densities: a regional approach. European Journal of Wildlife Research 54: 71–78 Delibes-Mateos M, Ferreras P, Villafuerte R (2008c) Rabbit populations and game management: the situation after 15 years of rabbit haemorrhagic disease in central-southern Spain. Biodiversity and Conservation 17: 559–574 Delibes-Mateos M, Ferreras P, Villafuerte R (2009) European rabbit populations trends and associated factors: a review of the situation in the Iberian Peninsula. Mammal Review 39: 124–140 Delibes-Mateos M, Díaz-Fernández S, Ferreras P, Viñuela J, Arroyo B (2013) The role of economic and social factors driving predator control in small game estates in central Spain. Ecology and Society 18(2): 28 Delibes-Mateos M, Díaz-Ruiz F, Caro J, Ferreras P (2014) Activity patterns of the vulnerable guiña (Leopardus guigna) and its main prey in the Valdivian rainforest of southern Chile. 79: 393–397 Dell’Arte GL, Laaksonen T, Norrdahl K, Korpimaki E (2007) Variation in the diet composition of a generalist predator, the red fox, in relation to season and density of main prey. Acta Oecologica 31: 276–281 Devesa JA (1991) Las gramineas de Extremadura. Grasses of Extremadura, Spain. University of Extremadura, Badajoz (Spain). 358 pp 170 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 Díaz-Fernández S, Viñuela J, Arroyo B. (2012) Harvest of red-legged partridge in central Spain. J.Wildlife Manage 76:1354–1363 Díaz-Ruiz F, García JT, Pérez-Rodríguez L, Ferreras P (2010) Experimental evaluation of live cage-traps for black-billed magpies Pica pica management in Spain. European Journal of Wildlife Research 56: 239–248 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 Díaz-Ruiz F, Ferreras P (2013) Conocimiento científico sobre la gestión de depredadores generalistas en España: el caso del zorro (Vulpes vulpes) y la urraca (Pica pica). Ecosistemas 22: 40–47 Di Bitetti MS, De Angelo CD, Di Blanco YE, Paviolo A (2010) Niche partitioning and species coexistence in a Neotropical felid assemblage. Acta Oecologica 36: 403– 412 Dickman CR (2010) The impacts and management of foxes Vulpes vulpes in Australia. Mammal Review 40: 181–211 Dos Santos Correia MP (1993) A comunidade de carnívoros da Reserva Natural da Serra da Malcata: uma partilha de recursos. Relatorio de estágio para obtençao da licenciatura em recursos faunísticos e ambiente. Facultade de Ciências da Universidade de Lisboa, Lisboa Dos Santos Marquez IJ (2003) Study of the interactions between the red fox (Vulpes vulpes Linnaeus) and the free-roaming cat (Felis catus, Linnaeus) in the wood of Valdelatas, Madrid. Estágio Professionalizante da Licenciatura em Biologia Aplicada a os Recursos Animais, variante Terrestres. Facultade de Ciências, Universidade do Lisboa, Lisboa Duarte J, Vargas JM (2001) Son selectivos los controles de predadores en los cotos de caza? Galemys 13: 1–9 Duarte J., Farfán MA, Fa JE, Vargas JM (2012) How effective and selective is traditional Red Fox snaring? Galemys 24: 1–11 Dudús L, Zalewski A, Koziol O, Jakubiec Z, Król N (2014) Habitat selection by two predators in an urban area: The stone martenand red fox in Wrocław (SWPoland). Mammalian Biology 79: 71–76 Eden SF (1989) The social organisation of non-breeding magpies Pica pica. Ibis 131: 141–153 171 Elbroch LM, Wittmer HU (2013) The effects of puma prey selection and specialization on less abundant prey in Patagonia. Journal of Mammalogy 94: 259–268 Eriksen A, Wabakken P, Zimmermann B, Andreassen HP, Arnemo JM, Gundersen H, Liberg O, Linnell J, Milner JM, Pedersen HC, Sand H, Solberg E, Storaas T (2009) Encounter frequencies of GPS-collared wolves and moose in a Scandinavian wolf territory. Ecological Research 24: 547–557 Eriksen A, Wabakken P, Zimmermann B, Andreassen HP, Arnemo JM, Gundersen H, Liberg O, Linnell J, Milner JM, Pedersen HC, Sand H, Solberg E, Storaas T (2011) Activity patterns of predator and prey: a simultaneous study of GPScollared wolves and moose. Animal Behavior 81: 423–431 Erpino MJ (1968) Age determination in the Black-billed Magpie. Condor 70:91–92 Esmeriz JP (2001) Ecologia alimentar da raposa (Vulpes vulpes Linnaeus 1758) no Parque Natural da Serra da Estrela. Disertaçao de Mestrado em Ecologia Aplicada apresentada à Facultade de Ciências da Universidade do Porto, Porto Estrada A, Caro J, Beja P, Borralho L, Casas F, Delibes-Mateos M, Díaz-Fernández S, Gordinho L, Reino L, Viñuela J, Arroyo B (2012) Does hunting management affect non-game bird species? En: Proceedings of the International Conference on Hunting for Sustainability: Ecology, Economics and Society. IREC, Ciudad Real, Pp. 97 European Union–Canada–Russian Federation (1998) Agreement on International Humane Trapping Standards between the European Community, Canada and the Russian Federation. Official Journal L 42 of 14 February 1998, 43–57 Fedriani JM (1996) Dieta anual del zorro Vulpes vulpes en dos hábitats del Parque Nacional de Doñana. Doñana Acta Vertebrata 23: 143–152 Fedriani JM, Ferreras P, Delibes M (1998) Dietary response of the Eurasian badger, Meles meles, to a decline of its main prey in the Doñana National Park. Journal of Zoology 245: 214–218 Fedriani JM, Palomares F, Delibes M (1999) Niche relations among three sympatric Mediterranean carnivores. Oecología 121: 138–148 Fedriani JM, Delibes M. (2009) Seed dispersal in the Iberian pear, Pyrus bourgaeana: a role for infrequent mutualists. Ecoscience, 16: 311–321 Fernández JM, Ruiz de Azua N (2005) Dieta y solapamiento trófico primaveral del zorro rojo Vulpes vulpes y de Martes sp. en simpatría en Álava (norte de España). Ecología 19: 167–182 Fernández de Simón J (2013) Predation and the recovery of European rabbit populations in central-southern Iberian Peninsula. An assessment of the predator pit hypothesis. Tesis Doctoral. Universidad de Castilla la Mancha Fernández-Juricic E, Jokimäki J, McDonald JC, Melado F, Toledano A, Mayo C, Martín B, Fresneda I, Martín V (2004) Effect of opportunistic predation on 172 antipredator bahavioural responses in a guild ground foragers. Oecologia 140:183–190 Fernández-Olalla M (2011) Seguimiento y gestión de sistemas depredador presa: aplicación a la conservación de fauna amenazada. Tesis Doctoral. Universidad Politécnica de Madrid, Madrid Ferreras P, Aldama JJ, Beltrán JF, Delibes M (1994) Immobilization of the endangered iberian lynx with Xylazine and Ketamine-hydroclhloride. Journal of Wildlife Diseases 30: 65–68 Ferreras P, Terriza J, López B, Reglero M, Castro F (2003) Homologación de métodos de control de predadores en Castilla la Mancha: bases científicas. Convenio de Colaboración entre la Consejería de Agricultura y Medio Ambiente de la Junta de Comunidades de Castilla la Mancha y el IREC-CSIC. Informe inédito Ferreras P, Villafuerte R, Mateo A (2006) Incidencia de la depredación sobre nidos de perdiz roja (Alectoris rufa) en Navarra. Departamento de Medio Ambiente, Ordenación del Territorio y Vivienda del Gobierno de Navarra- IREC-CSIC. Informe inédito Ferreras P, Luna S, Díaz-Ruiz F (2007) Evaluación de la selectividad y eficacia de métodos de control de depredadores para urraca y zorros en Castilla-La mancha. Convenio de Colaboración entre la Consejería de Agricultura y Medio Ambiente de la Junta de Comunidades de Castilla-La Mancha y el IREC-CSIC. Informe inédito Ferreras, P., Buenestado, P., Díaz-Ruiz, F. and Fernández de Simon (2009) Desarrollo de índices de abundancia de depredadores generalistas y niveles para su control en Castilla-La Mancha. Convenio de Colaboración entre la Consejería de Agricultura y Medio Ambiente de la Junta de Comunidades de Castilla la Mancha y el IREC-CSIC. Informe inédito Ferreras P, Mateo-Moriones A, Villafuerte R (2010) Influencia de la depredación sobre la perdiz roja en Navarra. Gobierno de Navarra, IREC-CSIC. Informe Final Inédito Ferreras P, Travaini A, Zapata SC, Delibes M (2011) Short-term responses of mammalian carnivores to a sudden collapse of rabbits in Mediterranean Spain. Basic and Applied Ecology 12: 116–124 Ferreira C, Touza J, Rouco C, Díaz-Ruíz F, Fernández de Simón J, Ríos-Saldaña CA, Ferreras P, Villafuerte R, Delibes-Mateos M (2013) Habitat management as a generalized tool to boost European rabbit (Oryctolagus cuniculus) populations in the Iberian Peninsula: A cost-effectiveness analysis. Mammal Review 44: 30–43 Fleming PJS, Allen LR, Berghout MJ, Meek PD, Pavlov PM, Stevens P, Strong K, Thompson JA, Thompson PC (1998) The performance of wild-canids traps in Australia, efficiency, selectivity and trap-related injuries. Wildlife Research 25: 327–338 173 Foster VC, Sarmento P, Sollmann R, Tôrres N, Jácomo ATA, Negrôes N, Fonseca C, Silveira L (2013) Jaguar and puma activity patterns and predator-prey interactions in four Brazilian biomes. Biotropica 45: 373–379 Fournier DA, Skaug HJ, Ancheta J, Ianelli J, Magnusson A, Maunder M, Nielsen A, Sibert J (2012) AD Model Builder, using automatic differentiation for statistical inference of highly parameterized complex nonlinear models. Optimization Methods and Software 27: 233–249 Fridolfsson AK, Ellegren H (1999) A simple and universal method for molecular sexing of non-ratite birds. Journal of Avian Biology 30: 116–121 Futuyma DJ, Moreno G (1988) The evolution of ecological specialization. Annual Review of Ecology and Systematics 19: 207–233 Gálvez MR (2004). Régimen jurídico de la actividad cinegética en España: análisis de las disposiciones autonómicas e intervención pública. Tesis Doctoral. Universidad de Málaga. García JT, García FJ, Alda F, González JL, Aramburu MJ, Cortés Y, Prieto B, Pliego B, Pérez M, Herrera J, García-Román L (2012) Recent invasion and status of the raccoon (Procyon lotor) in Spain. Biological Invasions 14: 305–1310 Garrote G, Perez de Ayala R, Pereira P, Robles F, Guzman N, García FJ, Iglesias MC, Hervás J, Fajardo I, Simón M, Barroso JL (2011) Estimation of the Iberian lynx (Lynx pardinus) population in the Doñana area, SW Spain, using capture– recapture analysis of camera-trapping data. European Journal of Wildlife Research 57: 355–362 Giller PS, Sangpradub N (1993) Predatory foraging behaviour and activity patterns of larvae of two species of limnephilid cased caddis. Oikos 67: 351–357 Gomes A, Pereira J, Bugoni L (2009) Age-specific diving and foraging behavior of the Great grebe (Podicephorus major). Waterbirds 32:149–156 Gómez-Villamandos RJ (2003) Anestesia, analgesia y eutanasia. En: Bienestar Animal, Pp. 37–42 González LM, Oria J, Sánchez R, Margalida A, Aranda A, Prada L, Caldera J, Molina, JI (2008) Status and habitat changes in the endangered Spanish Imperial Eagle Aquila adalberti population during 1974–2004: implications for its recovery. Bird Conservation International 18: 242–259 González-Prat F (1995a) Resultados preeliminares de la dieta del zorro (Vulpes vulpes L.) en Cataluña. II Jornadas de Conservación y Estudio de los Mamíferos (SECEM), Soria González-Prat F (1995b) Importancia del zorro (Vulpes vulpes L.) en la distribución del palmito (Chamaerops humilis) en la cordillera litoral catalana. II Jornadas de Conservación y Estudio de los Mamíferos (SECEM), Soria 174 Gooch S, Baillie SR, Birkhead TR (1991) Magpie Pica pica and songbirds populations. Restrospective investigation of trends in population density and breeding success. Journal of Applied Ecology 28: 1068–1086 Goodrich JM, Seryodkin I, Miquelle DG, Bereznuk SL (2011) Conflicts between Amur (Siberian) tigers and humans in the Russian Far East. Biological Conservation 144: 584–592 Gompper ME, Kays RW, Ray JC, Lapoint SD, Bogan DA, Cryan JR (2006) A comparison of noninvasive techniques to survey carnivore communities in Northeastern North America. Wildlife Society Bulletin 34: 1142–1151 Gortázar C (1999) Ecología y Patología del Zorro (Vulpes vulpes, L.) en el Valle Medio del Ebro. Consejo de Protección de la Naturaleza de Aragón, Zaragoza Goszczynski J, Jedrzejewska B, Jedrzejewski W (2000) Diet composition of badgers (Meles meles) in a pristine forest and rural habitats of Poland compared to other European populations. Journal of Zoology 250: 495–505 Graham K, Beckerman AP, Thirgood S (2005) Human-predator-prey conflicts: ecological correlates, prey losses and patterns of management. Biological Conservation 122: 159–171 Groom DW (1993) Magpie Pica pica predation on Blackbird Turdus merula nests in urban areas. Bird Study 40: 55–62 Guitián J, Callejo A (1983) Structure dúne communaute de carnivores dans la cordillere cantabrique occidentale. Terre Vie 37: 145–161 Guitián J, Munilla I (2010) Responses of mammal dispersers to fruit availability: rowan (Sorbus aucuparia) and carnivores in mountain habitats of northern Spain. Acta Oecologica 36: 242–247 Hadjisterkotis N (2003) The effect of corvid shooting on the populations of owls, kestrels and cuckoos in Cyprus, with notes on corvid diet. Zeitschrift für Jagdwissenschaft 49:50–60 Halle S (2000) Ecological relevance of daily activity patterns. En: Halle S, Stenseth NC (Eds.) Activity patterns in small mammals: An ecological approach. Springer, New York, Pp. 67–90 Hanski I, Hansson L, Henttonen H (1991) Specialist predators, generalist predators, and the microtine rodent cycle. Journal of Animal Ecology 60: 353–367 Harding EK, Doak DF, Albertson JD (2001) Evaluating the effectiveness of predator control: the non-native red fox as a case study. Conservation Biology 15: 1114– 1122 Harihar A, Pandav B (2012) Influence of connectivity, wild prey and disturbance on occupancy of tigers in the human-dominated western Terai Arc landscape. PLos ONE 7(7): e40105 175 Hartová-Nentvichová M, Šálek M, Červený J, Koubek P (2010) Variation in the diet of the red fox (Vulpes vulpes) in mountain habitats: effects of altitude and season. Mammalian Biology 75: 334–340 Harrington LA, Harrington AL, Macdonald DW (2009). The smell of new competitors, the response of American mink, Mustela vison, to the odours of otter, Lutra lutra and polecat, M. putorius. Ethology 115: 421–428 Harris S, Soulsbury C, Iossa G (2006) A scientific review on proposed humane trapping standards in Europe. The ISO Standards and the European proposal for a proposed Directive on humane trapping standards. University of Bristol, School of Biological Sciences, Bristol, United Kingdom Headland TN, Greene HW (2011) Hunter–gatherers and other primates as prey, predators, and competitors of snakes. Proceedings of the National Academy of Sciences of the United States of America 108: 1470–1474 Helldin JO, Liberg O, Gloersen G (2006) Lynx (Lynx lynx) killing red foxes (Vulpes vulpes) in boreal Sweden – frequency and population effects. Journal of. Zoology 270: 657–663 Henke SE, Bryant FC (1999) Effects of coyote removal on the faunal community in western Texas. Journal of Wildlife Management 63: 1066–1081 Henry C, Poulle ML, Roeder JJ (2005) Effect of sex and female reproductive status on seasonal home range size and stability inrural red foxes (Vulpes vulpes). Ecoescience 12: 202–209 Hernandez M, Margalida A (2009) Poison-related mortality effects in the endangered Egyptian vulture (Neophron percnopterus) population in Spain. European Journal of Wildlife Research 55: 415–423 Herranz J, Yanes M, Suárez F (1999) Efecto de los predadores sobre la caza menor y evaluación de sistemas selectivos para regular los niveles de predación. Informa final. CSIC, Madrid. Pp 318 Herranz J (2000) Efectos de la depredación y del control de depredadores sobre la caza menor en Castilla-La Mancha. Tesis Doctoral. Universidad Autónoma de Madrid Herrera CM (1976) A trophic diversity index for presence–absence food data. Oecologia 25: 187–191 Heydon MJ, Reynolds JC (2000) Demography of rural foxes (Vulpes vulpes) in relation to cull intensity in three contrasting regions of Britain. Journal of Zoology 251: 265–276 Heydon MJ, Reynolds JC, Short MJ (2000) Variation in abundance of foxes (Vulpes vulpes) between three regions of rural Britain, in relation to landscape and other variables. Journal of Zoology 251: 253–264 176 Hill RA, Dunbar RIM (2002) Climatic determinants of diet and foraging behaviour in baboons. Evolutionary Ecology 16: 579–593 Hines JE, Mackenzie D (2013) Presence Softwere 5.8 Hiraldo F, Delibes M, Calderón J (1979) El quebrantahuesos Gypaetus barbatus (L.) Sistemática, taxonomía, biología, distribución y protección. Monografías 22. Ministerio de Agricultura - Instituto Nacional para la Conservación de la Naturaleza, Sevilla Högstedt G (1980) Prediction and test of the effects of interspecific competition. Nature 283: 64–66 Holt AR, Davies ZG, Tyler C, Staddon S (2008) Meta-analysis of the effects of predation on animal prey abundance: evidence from UK vertebrates. PLoS ONE 3:e2400 Hounsome T, Delahay R (2005) Birds in the diet of the Eurasian badger Meles meles: a review and meta-analysis. Mammal Review 35: 199–209 Hudson P, Dobson AP, Newborn D (1992) Do parasites make prey vulnerable to predation? Red grouse and parasites. Ecology 61: 681–692 Inskip C, Zimmerman A (2009) Human-felid conflict: a review of patterns and priorities worldwide. Oryx 43: 18–34 InstitutoGeografico Nacional (2013) CORINE land-cover 2006 and satellite orthophotos. http://www.ign.es/ International Organization for Standardization (1999) TC191. Animal (mammal) traps. Part 5, methods for testing restraining traps. International Standard ISO/DIS 10990–5. International Organization for Standardization, Geneva, Switzerland Iossa G, Soulsbury CD, Harris S (2007) Mammal trapping: a review of animal welfare standards of killing and restraining traps. Animal Welfare 16: 335–352 Janko C, Schröder W, Linke S, König A (2012) Space use and resting site selection of red foxes (Vulpes vulpes) living near villages and small towns in Southern Germany. Acta Theriologica 57: 245–250 Jerosch S, Götz M, Klar N, Roth M (2010) Characteristics of diurnal resting sites of the endangered European wildcat (Felis silvestris silvestris), implications for its conservation. Journal for Nature Conservation 18: 45–54 Jiguet F (2002) Arthropods in diet of Little Bustards Tetrax tetrax during the breeding season in western France. Bird Study 49:105–109 Johnson A, Vongkhamheng C, Saithongdam T (2009) The diversity, status and conservation of small carnivores in a montane tropical forest in northern Laos. Oryx 43: 626–633 177 Jorgenson JW, Novonty M, Carmack M, Copland GB, Wilson S, Katona S, Whitten WK (1978) Chemical scent constituents in the urine of the red fox (Vulpes vulpes) during the winter season. Science 199: 796–798 Junta de Andalucía (2010) Estandarización de métodos de captura no crueles de acuerdo con la norma ISO10990-5. Informe Inédito Final. Consejería de Medio Ambiente. Junta de Andalucía Kareiva P, Marvier M (2012) What is Conservation Science? BioScience 62: 962–969 Kauhala K, Helle P, Helle E (2000) Predator control and the density and reproductive success of grouse populations in Finland. Ecography 23: 161–168 Kay B, Gifford E, Perry R, Van de Ven R (2000) Trapping efficiency for foxes (Vulpes vulpes) in central New South Wales: age and sex biases and the effects of reduced fox abundance. Wildlife Research 27: 547–552 Keedwell RJ, Maloney RF, Murray DP (2002) Predator control for protecting kaki (Himantopus novaezelandiae) lessons from 20 years of management. Biological Conservation 105: 369–374 Kelly MJ, Holub EL (2008) Camera trapping of carnivores: trap success among camera types and across species, and habitat selection by species, on Salt Pond Mountain, Giles County, Virginia. Northeast Naturalits 15: 249–262 Kidawa D, Kowalczyk R (2011) The effects of sex, age, season and habitat on diet of the red fox Vulpes vulpes in northeastern Poland. Acta Theriol. 56: 209–218 Kitchen AM, Gese EM, Schauster ER. (2000) Changes in coyote activity patterns due to reduced exposure to human persecution. Canadian Journal of Zoology 78: 853– 857 Kjellander P, Nordstrom J (2003) Cyclic voles, prey switching in red fox, and roe deer dynamics – a test of the alternative prey hypothesis. Oikos 101: 338–344 Klare U, Kamler JF, Macdonald DW (2011) A comparison and critique of different scat-analysis methods for determining carnivore diet. Mammal Review 41: 294– 312 Knauer F, Küchenhoff H, Pilz S (2010) A statistical analysis of the relationship between red fox Vulpes vulpes and its prey species (grey partridge Perdix perdix, brown hare Lepus europaeus and rabbit Oryctolagus cuniculus) in Western Germany from 1958 to 1998. Wildlife Biology 16: 56–65 Kopij G (2005) Diet of some insectivorous passerines in semi-arid regions of South Africa. Ostrich 76:85–90 Krebs CJ (2002) Beyond population regulation and limitation. Wildlife Research 29: 1– 10 Kronfeld-Schor N, Dayan T (2003) Partitioning of time as an ecological resource. Annual Review of Ecological Systems 34: 153–181 178 Kryštofková M, Fousová P, Exnerová A (2011) Nestling diet of the common magpie (Pica pica) in urban and agricultural habitats. Ornis Fennica 88: 138–146 Langley PJW, Yalden DW (1977) The decline of rarer carnivores in Great Britain during the nineteenth century. Mammal Review 7: 95–16 Le Vaillant M, Le Bohec C, Prud’Homme O, Wienecke B, Le Maho Y, Kato A, RopertCoudert Y (2013) How age and sex drive the foraging behaviour in the king penguin. Marine Biology 160:1147–1156 Levi T, Wilmers CC (2012) Wolves-coyotes-foxes: a cascade among carnivores. Ecology 93: 921–929 Lima SL, Dill LM (1990) Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology 68: 619–640 Linkie M, Ridout MS (2011) Assessing tiger-prey interactions in Sumatran rainforests. Journal of Zoology 284: 224–229 Liu Y, Wang SY, Shi L (2011) Effects of age and sex on diet composition of Phrynocephalus grumgrzimailoi. Chinese Journal of Zoology 46:111–117 Ljung PE, Widemo F, Ericsson G (2014) Trapping in predator management: catching the profile of trap users in Sweden. European Journal of Wildlife Research 60: 681–689 Lodé T (1997) Trophic status and feeding habits of the European Polecat (Mustela putorius L. 1758). Mammal Review 27: 177–184 Long RA, MacKay P, Zielinski WJ, Ray JC (2008) Noninvasive survey methods for carnivores. Island Press.Washington. Long RA, Donovan TM, MacKay P, Zielinsky WJ, Buzas JS (2011) Predicting carnivore occurrence with noninvasive survey and occupancy modeling. Landscape Ecology 26: 327–340 López G, López-Parra M, Garrote G, Fernández L, Rey-Wamba T, Arenas-Rojas R, García-Tardío M, Ruiz G, Zorrilla T, Moral M, Simón MA (2014) Evaluating mortality rates and casualties in a critically endangered felid across its whole distribution range. European Journal of Wildlife Research 60: 359–366 López-Bao JV, Sazatornil V, Llaneza L, Rodríguez A (2013). Indirect effects on heathland conservation and wolf persistence of contradictory policies that threaten traditional free-ranging horse husbandry. Conservation Letters 6:448– 455 López-Martín JM (1999) Inedited data from this author López-Martín JM (2010) Zorro – Vulpes vulpes. En: Enciclopedia Virtual de los Vertebrados Españoles. Salvador, A., Cassinello, J. (Eds.). Museo Nacional de Ciencias Naturales, Madrid. http://www.vertebradosibericos.org/ 179 Loureiro F, Bissonete JA, Macdonald DW, Santos-Reis M (2009) Temporal variation in the availability of Mediterranean food resources: do badgers Meles meles track them? Wildlife Biology 15: 197–206 Lozano J, Casanova JG, Cabezas-Díaz S, Virgós E, Mangas JG (2006a) El control de depredadores en España, más que discutible. Quercus 239: 80–82 Lozano J, Moleón M, Virgós E (2006b) Biogeographical patterns in the diet of the wildcat, Felis sylvestris Schreber, in Eurasia: factors affecting the trophic diversity. Journal of Biogeography 33: 1076–1085 Lozano J, Casanovas JG, Virgós E, Zorrilla JM (2013) The competitor release effect applied to carnivore species: how red foxes can increase in numbers when persecuted. Animal Biodiversity and Conservation36.1: 37–46 Lozano J, Casanova JG, Cabezas-Díaz S, Virgós E, Mangas JG (2006) El control de depredadores en España, más que discutible. Quercus 239: 80–82 Lozano J, Casanovas JG, Virgós E, Zorrilla JM (2013) The competitor release effect applied to carnivore species: how red foxes can increase in numbers when persecuted. Animal Biodiversity and Conservation 36.1: 37–46 Luque-Larena JJ, Mougeot F, Viñuela J, Jareño D, Arroyo L, Lambin X, Arroyo B. (2013) Recent large-scale range expansión and outbreaks of the common vole (Microtus arvalis) in NW Spain. Basic and Applied Ecology 14: 432–441 Macdonald DW (1979) Some observations and field experiments on the urine marking behaviour of the red fox, Vulpes vulpes L. Zeitschrift fur Tierpsychologie 51: 1– 22 Macdonald DW, Buesching CD, Stopka P, Henderson J, Ellwood SA, Baker SE (2004) Encounters between two sympatric carnivores: red foxes (Vulpes vulpes) and European badgers (Meles meles). Journal of Zoology 263: 385–392 Macdonald DW, Reynolds JC (2004). Red Fox. En: Sillero-Zubiri C, Hoffmann M, MacDonald DW (Eds.) (2004) Canids: Foxes, Wolves, Jackals and Dogs. IUCN/SSC Canid Specialist Group, Gland, Switzerland and Cambridge, UK, Pp.129–136 Mackenzie D, Nichols JD, Lachman GB, Droege S, Royle JA, Langtimm CA (2002) Estimating site occupancy rates when detection probabilitiesare less than one. Ecology 83: 2248–2255 Mackenzie D, Nichols JD (2004) Occupancy as a surrogate for abundance estimation. Animal Biodiversityand Conservation 27: 461–467 Mackenzie D, Nichols JD, Royle JA, Pollock KH, Bailey LL, Hines JE (2006) Occupancy estimation and modeling: inferring patterns and dynamics of species occurrence. Academic Press, Oxford Madden CF, Arroyo B, Amar A (en prensa) A review of the impacts of crows, ravens and Eurasian Magpies on bird productivity and abundance. Ibis 180 Maguire GS,D, Stojanovic D, Weston MA (2009) Conditioned taste aversion reduces fox depredation on model eggs on beaches. Wildlife Research 36: 702–708 Mangas JG, Lozano J, Cabezas-Díaz S, Virgós E (2008) The priority value of scrubland habitats for carnivoreconservation in Mediterranean ecosystems. Biodiversity and Conservation17: 43–51 Mañosa S (1994) Goshawk diet in a Mediterranean area of northeastern Spain. Journal of Raptor Research 28: 84–92 Marks CA (2010) Haematological and biochemical responses of red foxes (Vulpes vulpes) to different capture methods and shooting. Animal Welfare 19: 223–234 Márquez C, Vargas JM, Villafuerte R, Fa JE (2013) Risk mapping of illegal poisoning of avian and mammalian predators. The Journal of Wildlife Management 77: 75–83 Marti R, Del Moral JC (Eds.) (2003) Atlas de las aves reproductoras de España. SEO/BirdLife y Direccion General de Conservacion de la Naturaleza. Madrid Martín M (2008) Caracterització de la dieta de la guineu (Vulpes vulpes) al Parc del Garraf. En: V Trobada d’Estudiosos del Garraf. Diputació de Barcelona, Barcelona. Pp. 81–88 Martin J, Basille M, Van Moorter B, Kindberg J, Allaine D, Swenson JE (2010) Coping with human disturbance: Spatial and temporal tactics of the brown bears (Ursus arctos). Canadian Journal of Zoology 88: 875–883 Martínez JG (2011) Urraca – Pica pica. En: Enciclopedia Virtual de los Vertebrados Españoles. Salvador, A., Morales, M. B. (Eds.). Museo Nacional de Ciencias Naturales, Madrid. http://www.vertebradosibericos.org/ Martínez JG, Soler M, Soler JJ, Paracuellos M, Sánchez J (1992) Alimentación de los pollos de urraca (Pica pica) en relación con la edad y disponibilidad de presas. Ardeola 39: 35–48 Martínez-Abraín A, Crespo J, Jiménez J, Pullin A, Stewart G, Oro D (2008) Friend or foe: societal shifts from intense persecution to active conservation of top predators. Ardeola 55: 111–119 Martínez-Abraín A, Crespo J, Berdugo M, Gutiérrez L, Lafuente A, Mañas A, de Miguel JM (2013) Causes of human impact to protected vertebrate wildlife parallel long-term socio-economical changes in Spain. Animal Conservation 16: 286–294 Martínez de Castilla A, Martínez A (2004) ¿Es beneficioso controlar las urracas? Trofeo Diciembre 2004: 58–62 Masi S, Cipolletta C, Robbins MM (2009) Western lowland gorillas (Gorilla gorilla gorilla) change their activity patterns in response to frugivory. American Journal of Primatology 71: 91–100 181 Mateo–Moriones A, Villafuerte R, Ferreras P (2012) Does fox control improve red– legged partridge (Alectoris rufa) survival? An experimental study in Northern Spain. Animal Biodiversity and Conservation 35: 395–404 Matos HM, Santos MJ, Palomares F,.Santos-Reis M (2009) Does riparian habitat condition influence mammalian carnivore abundance in Mediterraneanecosystems? Biodiversity and Conservation18: 373–386 Matthews SM, Beecham JJ, Quigley H, Greenleaf SS, Leithead HM (2006) Activity patterns of American black bears in Yosemite National Park. Ursus 17: 30–40 McCade RE, McCade TR (1984) Of slings and arrows: a historical retrospection. En: Halls LK, (Ed.) White-Tailed Deer: Ecology and Management. Pp. 19–72. Stackpole Books, Harrisburg McCarthy MA, Moore JL, Morris WK, Parris KM, Garrard GE, Vesk PA, Rumpff L, Giljohann KM, Camac JS, Bau SS, Friend T, Harrison B, Yue B (2013) The influence of abundance on detectability. Oikos 122:717–726 McLeod SR, Saunders G (2014) Fertility control is much less effective than lethal baiting for controlling foxes. Ecological Modelling 273: 1– 10 Medina FM, Bonnaud E, Vidal E, Tershy BR, Zabaleta ES, Donlan CJ, Keitt BS, Corre ML, Horwath SV, Nogales M (2011) A global review of the impacts of invasive cats on island endangered vertebrates. Global Change Biology 17: 3503–3510 Millán J, Gortázar C, Tizzani P, Buenestado FJ (2002) Do helminths increase the vulnerability of released pheasants to fox predation? Journal of Helminthology 76: 225–229 Miller JR, Hobbs NT (2000) Recreational trails, human activity, and nest predation in lowland riparian areas. Landscape Urban Planing 50:227–236 Moberly RL,White PCL,Webbon CC, Baker PJ, Harris S (2004) Modelling the costs of fox predation and preventive measures on sheep farms in Britain. Journal of Environmental Management 70: 129–143 Monterroso P, Moreira P, Castro D, Nery F, Sarmento P (2006) Predação sobre coelhobravo (Oryctolagus cuniculus) em ecossistemas Mediterrâneos. En: Ferreira C, Alves PC (Eds.) Gestao de populações de coelho-bravo (Oryctolagus cuniculus algirus), Federaçao Alentejana de Caçadores (Eds.). Pp. 49–79 Monterroso P, Alves PC, Ferreras P (2011) Evaluation of attractants for non-invasive studies of Iberian carnivore communities. Wildlife Research 38: 446–454 Monterroso P (2013) Ecological interactions and species coexistence in Iberian mesocarnivores communities. Tesis Doctoral, Universidad de Oporto, Portugal Monterroso P, Alves PC, Ferreras P (2013) Catch me if you can: diel activity patterns of mammalian prey and predators. Ethology 119: 1044–1056 182 Monteverde MJ, Piudo L (2011) Activity patterns of the culpeo fox (Lycalopex culpaeus magellanica) in a non-hunting area of northwestern Patagonia, Argentina. Mammal Study 36: 119–125 Musiani M, Paquet PC (2004) The practices of wolf persecution, protection, and restoration in Canada and the United States. BioScience 54: 50–60 Muñoz-Igualada J, Shivik JA, Domínguez FG, Lara J, González LM (2008) Evaluation of cage-traps and cable restraint devices to capture red foxes in Spain. Journal of Wildlife Management 72: 830–836 Muñoz-Igualada J, Shivik JA, Domínguez FG, González LM, Moreno AA, Olalla MF, García CA (2010) Traditional and new cable restraint systems to capture fox in central Spain. Journal of Wildlife Management 74: 181–187 Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity hotspots for conservation priorities. Nature 403: 853–858 Narváez M, Blanco JC, Barrios L (2008) Estudio de los carnívoros predadores de la ZEPA de la Sierra Norte de Sevilla. En: Cuenca I, Menor A (Eds.) Investigación científica y conservación en el Parque Natural Sierra Norte de Sevilla, D.G. Sostenibilidad en la Red de Espacios Naturales, Consejería de Medio Ambiente, Junta de Andalucía. Pp. 241–254 Negrões NM (2000) Conflicto Homem-Raposa: o exemplo da região de Castro Laboreiro. Disertação para prestação de provas de Mestrado em Ciências do Ambiente-Ramo Qualidade Ambiental. Departamento de Biologia, Universidade do Minho, Braga Newsome AE, Catling PC, Cooke BD, Smyth R (2001) Two ecological universes separated by the Dingo Barrier fence in semi-arid Australia: interactions between landscapes, herbivory and carnivory, with and without dingoes. Rangeland Journal 23: 71–98 Newson SE, Evans KL, Noble DG, Greenwood JJD, Gaston KJ (2008) Use of distance sampling to improve estimates of national population sizes for common and widespread breeding birds in the UK. Journal of Applied Ecology 45: 1330– 1338 Newson SE, Rexstad EA, Baillie SR, Buckland ST, Aebischer NJ (2010) Population change of avian predators and grey squirrels in England: is there evidence for an impact on avian prey populations? Journal of Applied Ecology 47:244–252 Nicoll M, Norris K (2010) Detecting an impact of predation on bird populationsdepends on the methods used to assess the predators. Methods in Ecology and Evolution 1: 300–310 Njau JK, Blumenschine RJ (2012) Crocodylian and mammalian carnivore feeding traces on hominid fossils from FLK 22 and FLK NN 3, Plio-Pleistocene, Olduvai Gorge, Tanzania. Journal of Human Evolution 63: 408–417 183 Ohashi H, Saito M, Horie R, Tsunoda H, Noba H, Ishii H, Kuwabara H, Hiroshige Y, Koike S, Hoshino Y, Toda H, Kaji K. (2013) Differences in the activity pattern of the wild boar Sus scrofa related to human disturbance. European Journal of Wildlife Research 59: 167–177 O’Mahony D, Lambin X, MacKinnon JL, Coles CF (1999) Fox predation on cyclic field vole populations in Britain. Ecography 22: 575–581 Ontiveros D, Pleguezuelos JM, Caro J (2005) Prey density, prey detectability and food habits: the case of Bonelli’s eagle and the conservation measures. Biological Conservation 123: 19–25 Ordeñana MA, Crooks KR, Boydston EE, Fisher RN, Lyren LM., Siudyla S., Haas CD, Harris S, Hathaway SA, Turschak GM, Miles AK, Van Vuren DH (2010) Effects of urbanization on carnivore species distribution and richness. Journal of Mammalogy 91: 1322–1331 Ordiz A, Støen OG, Sæbø S, Kindberg J, Delibes M, Swenson JE. (2012) Do bears know they are being hunted? Biological Conservation 152: 21–28 Oro D, De León A, Mínguez E, Furness RW (2005) Estimating predation on breeding European storm-petrels (Hydrobates pelagicus) by yellow-legged gulls (Larus michahellis). Journal of Zoology 265: 421–429 Otero C (1995) Control de depredadores en la gestión integrada de un territorio. En: Predación, caza y vida silvestre. Fundación “La Caixa”/Aedos, Barcelona, Pp 151–180 Packer C, Ikanda D, Kissui B, Kushnir H (2005) Lion attacks on humans in Tanzania. Nature 436: 927–928 Padial B, Avila E, Gil-Sánchez JM (2002) Feeding habits and overlap among red fox (Vulpes vulpes) and stone marten (Martes foina) in two Mediterranean mountain habitats. Mammal Biology 67: 137–146 Palomares F, Ruiz-Martínez I (1994) Die Ditche des Rotfuchses und die Beute an Niederwild während der Periode der Jungenaufzucht im Südosten Spaniens. Zeitschrift für Jagdwissenschaft 40: 145–155 Palomares F, Gaona P, Ferreras P, Delibes M (1995) Positive effects on game species of top predators by controlling smaller predator populations: an example with lynx, mongooses, and rabbits. Conservation Biology 9: 295–305 Palomares F, Caro TM (1999) Interspecific killing among mammalian carnivores. The American Naturalist 153: 492–508 Palomares F (2001) Comparison of 3 methods to estimate rabbit abundance in Mediterranean environment. Wildlife Society Bulletin 29: 578–585 Palomares F, Rodríguez A, Godoy JA, Goritz F, Jewgenow K, Roldán E, López-Bao JV, Soto C (2010) Reproducción, genética, condición física, y tamaño de la población de los linces del Parque Nacional de Doñana en relación con un 184 programa de alimentación suplementaria: evaluación científica de varios métodos de estudio y seguimiento. Informe Inédito. EBD-CSIC, Organismo Autónomo de Parques Nacionales, MIMAM. España Palomino D, Carrascal LM, Potti J (2011) Distribution of Azure-winged Magpies Cyanopica cooki in Spain: both local and large-scale factors considered. Acta Ornithologica, 46: 71–82 Palomo LJ, Gisbert J, Blanco JC (2007) Atlas y Libro Rojo de los Mamíferos Terrestres de España. Dirección General de Conservación de la Naturaleza – SECEMSECEMU, Madrid, 586 Pp. Panek M (2013) Long-term changes in the feeding pattern of red foxes Vulpes vulpes and their predation on brown hares Lepus europaeus in western Poland. European Journal of Wildlife Research 59: 581–586 Pärt T (2001) The effects of territory quality on age-dependent reproductive performance in the northern wheatear, Oenanthe oenanthe. Animal Behaviour 62: 379–388 Penteriani V, Delgado MM (2010) Búho real – Bubo bubo. En: Enciclopedia Virtual de los Vertebrados Españoles. Salvador A, Bautista LM (Eds.). Museo Nacional de Ciencias Naturales, Madrid. http://www.vertebradosibericos.org/ Pereira P, Alves da Silva A, Alves J, Matos M, Fonseca C.(2012) Coexistence of carnivores in a heterogeneous landscape: habitatselection and ecological niches. Ecological Research27: 745–753 Phillips RB, Whinche CS (2011) Reducing nontarget recaptures of an endangered predator using conditioned aversion and reward removal. Journal of Applied Ecology 48: 1501–1507 Pianka ER (1966) Latitudinal gradients in species diversity: a review of concepts. American Naturalist 100: 33–46 Piorno V (2006) Gestión cinegética y conservación del conejo de monte. Tesis Doctoral, Universidad de Vigo Ponz A, Gil-Delgado JA, Barba E (1999) Factors affecting prey preparation by adult Magpies feeding nestlings. Condor 101: 818–823 Ponz A, Gil-Delgado JA (2004) Biología reproductiva de la urraca Pica pica en un área de montaña de Aragón. Ardeola 51: 411–423 Potts GR (1980) The effects of modern agriculture, nest predation and game management on the population ecology of partridges (Perdix perdix and Alectoris rufa). Advances in Ecological Research 11: 2–79 Progioni C, Balestrieri A, Remonti L, Cavada L (2008) Differential use of food and habitat by sympatric carnivores in theeastern Italian Alps. Italian Journal of Zoology 75: 173–184 185 Prugh LR, Arthur SM, Ritland CE (2008) Use of faecal genotyping to determine individual diet. Wildlife Biology 14: 318–330 Prugh LR, Stoner CJ, Epps WC, Bean WT, Ripple WJ, Laliberte AS, Brashares JS (2009) The rise of the mesopredator. BioScience 59: 779–791 Pullin AS, Stewart GB (2006) Guidelines for systematic review in conservation and environmental management. Conservation Biology 20: 1647–1656 Putman RJ (1984) Facts from faeces. Mammal Review 14: 79–97 QGIS Development Team (2013) QGIS Geographic Information System. Open Source Geospatial Foundation Project. http://qgis.osgeo.org R Development Core Team (2012). R, A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3900051-07-0, URL http,//www.R-project.org/ Rasmussen GSA, Macdonald DW. (2011) Masking of the zeitgeber: African wild dogs mitigate persecution by balancing time. Journal of Zoology 286: 232–242 Recio MR, Virgós E (2010) Predictive niche modelling to identify potential areas of conflicts between human activities and expanding predator populations: a case study of game management and the grey mongoose, Herpestes ichneumon, in Spain. Wildlife Research 37: 343–354 Redpath SM, Young J, Evely A, Adams WM, Sutherland WJ, Whitehouse A, Amar A, Lambert RA, Linnell JDC, Watt A, Gutiérrez RJ (2013) Understanding and managing conservation conflicts. Trends in Ecology and Evolution 28: 100–109 Reynolds JC, Aebischer N (1991) Comparison and quantification of carnivore diet by faecal analysis: a critique, with recommendations, based on a study of the fox Vulpes vulpes. Mammal Review 21: 97–122 Reynolds JC, Goddard HN, Brockless MH (1993) The impact of local fox (Vulpes vulpes) removal on fox populations at two sites in southern England. Gibier Faune Sauvage 10: 319–334 Reynolds JC, Tapper SC (1995) The ecology of the red fox Vulpes vulpes in relation to small game in rural southern England. Wildlife Biology 1: 105–119 Reynolds JC, Tapper SC (1996) Control of mammalian predators in game management and conservation. Mammal Review 26: 127–156 Reynolds JC, Stoate C, Brockless MH, Aebischer NJ, Tapper SC (2010) The consequences of predator control for brown hares (Lepus europaeus) on UK farmland. European Journal of Wildlife Research 56: 541–549 Ridout MS, Linkie M (2009) Estimating overlap of daily activity patterns from camera trap data. Journal of Agricultural, Biological, and Environmental Statistics 14: 322–337 186 Ríos-Saldaña CA (2010) Los Planes Técnicos de Caza de Castilla-La Mancha y su Aplicación en la Gestión y Conservación de las Especies Cinegéticas. Tesis Doctoral Universidad de Castilla-La Mancha Ripple WJ, Wirsing AJ, Wilmers CC, Letnic M (2013) Widespread mesopredator effects after wolf extirpation. Biological Conservation 160: 70–79 Ripple WJ, Estes JA, Beschta RL, Wilmers CC, Ritchie EG, Hebblewhite M, Berger J, Elmhagen B, Letnic M, Nelson M, Schmitz OJ, Smith DW, Wallach AD, Wirsing AJ (2014) Status and ecological effects of the world’s largest carnivores. Science 343: 1241484 Ritchie EG, Johnson CN (2009) Predator interactions, mesopredator release and biodiversity conservation. Ecology Letters 12: 982–998 Rivas-Martínez S (1987) Nociones sobre Fitosociología, Biogeografía y Bioclimatología. In: Peinado M, Rivas-Martínez S (eds) La Vegetación de España, 15–57. Universidad de Alcalá de Henares, Madrid Rivas-Martínez S, Penas A, Díaz TE (2004) Mapa Bioclimático de Europa, Bioclimas, http://www.globalbioclimatics.org/form/maps.htm Rodríguez A, Delibes M (2002) Internal structure and patterns of contraction in the geographic range of the Iberian lynx. Ecography 25: 314–328 Rodríguez A, Delibes M (2003) Population fragmentation and extinction in the Iberian lynx. Biological Conservation 109: 321–331 Rodríguez A, Delibes M (2004) Patterns and causes of non-natural mortality in the Iberian lynx during a 40-year period of range contraction. Biological Conservation 118: 151–161 Roos S, Pärt T (2004) Nest predators affect spatial dynamics of breeding red-backed shrikes (Lanius collurio). Journal of Animal Ecology 73:117–127 Roper TJ, Mickevicius E (1995) Badger Meles meles diet: a review of literature from the former Soviet Union. Mammal Review 25: 117–129 Rosalino LM, Santos-Reis M (2009) Fruit consumption by carnivores in Mediterranean Europe. Mammal Review 39: 67–78 Rosalino LM, Rosa S, Santos-Reis M (2010) The role of carnivores as Mediterranean seed dispersers. Annales Zoologici Fennici 47: 195–205 Rosenzweig ML (1995) Species Diversity in Space and Time. Cambridge University Press, Cambridge, UK Roughton RD (1982) A synthetic alternative to fermented egg as a canid attractant. The Journal of Wildlife Management 46: 230–234 Rowcliffe JM, Carbone C (2008).Surveys using camera traps: are we looking to a brighter future? Animal Conservation 11: 185–186 187 Royle JA, Nichols JD (2003) Estimating abundance from repeated presence-absence data or point counts. Ecology 84:777–790 Ruette S, Stahl P, Albaret M (2003) Factors affecting trapping success of red fox Vulpes vulpes, stone marten Martes foina and pine marten M. martes in France. Wildlife Biology 9: 11–19 Ruiz-Olmo J, Blanch F, Vidal F (2003) Relationships between the red fox and waterbirds in the Ebro delta natural park, N.E. Spain. Waterbirds 26: 217–225 Sala OE, Chapin III FS, Armesto JJ, Berlow E, Bloomfield J, Irzo R, Huber-Samwald E, Huenneke KF, Jackson RB, Kinzig A, Leemans R, Lodge DM, Mooney HA, Oesterheld M, Poff NL, Sykes MT, Walker BH, Walker M, Wall DH (2000) Global biodiversity scenarios for the year 2100 Science 87: 1770–1774 Šálek M, Kreisinger J, Sedláček F, Albrecht T (2010) Do prey densities determine preferences of mammalian predators for habitat edges in an agricultural landscape? Landscape and Urban Planning 98: 86–91 Salo P, Banks PB, Dickman CR, Korpimaki E (2010) Predator manipulation experiments: impacts on populations of terrestrial vertebrate prey. Ecological Monographs 80: 531–546 Sangay T, Vernes K (2008) Human–wildlife conflict in the Kingdom of Bhutan, patterns of livestock predation by large mammalian carnivores. Biological Conservation 141: 1272–1282 Sanderson EW, Jaiteh M, Levy MA, Redford KH, Wannebo AV, Woolmer G (2002) The Human footprint and the last of the wild. BioScience 52: 891–904 Santos MJ, Matos HM, Baltazar C, Grilo C, Santos-Reis M (2009) Is polecat (Mustela putorius) affected by ‘mediterraneity’? Mammalian Biology 74: 448–455 Sanz-Aguilara A, Martínez-Abraín A, Tavecchiaa G, Mínguez E, Oro D (2009) Evidence-based culling of a facultative predator: Efficacy and efficiency components. Biological Conservation 142: 424–431 Sarmento P (1996) Feeding habits of the weasel Mustela nivalis in relation to prey abundance. Oikos 26: 378–384 Sarmento P, Cruz J (1998) Ecologia e conservação do lince-ibérico e da comunidade de carnívoros da Serra da Malcata. Inedited Report, Reserva Natural da Serra da Malcata, ICN/RNSM Sarmento P, Cruz J, Eira C, Fonseca C (2009) Habitat selection and abundance of common genets Genetta genetta using camera capture-mark-recapture data. European Journal of Wildlife Research 56:59–66 Sarmento PB, Cruz J, Eira C, Fonseca C (2011) Modeling the occupancy of sympatric carnivorans in a Mediterranean ecosystem. European Journal of Wildlife Research 57: 119–131 188 Saunders G, Harris S (2000) Evaluation of attractants and bait preferences of captive red foxes (Vulpes vulpes). Wildlife Research 27: 237–243 Saunders GR, Gentle MN, Dickman CR (2010) The impacts and management of foxes Vulpes vulpes in Australia. Mammal Review 40: 181–211 SEO/BIRDLIFE (2002) Tendencias de las poblaciones de aves comunes en España (1996–2001). Programa SACRE. Informe 2001. SEO/BirdLife, Madrid Schauster ER, Gese EM, Kitchen AM (2002) An evaluation of survey methods for monitoring swift fox abundance. Wildlife Society Bulletin 30: 464–477 Schuette P, Wagner AP, Wagner ME, Creel S (2013) Occupancy patterns and niche partitioning within a diverse carnivore community exposed to anthropogenic pressures. Biological Conservation 158: 301–312 Seal US, Kreeger TJ (1987). Chemical immobilization of furbearers. En: Novak M, Baker JA, Obbard ME, Malloch B (Eds.) Wild Furbearer Management and Conservation in North America. Ontario Trappers Association, Ontario Seidler RG, Geseb EM, Conner MM (2014) Using sterilization to change predation rates of wild coyotes: A test case involving pronghorn fawns. Applied Animal Behaviour Science 154: 83–92 Sergio F, Hiraldo F (2008) Intraguild predation in raptor assemblages: a review. Ibis 150: 132–145 Sergio F, Caro T, Brown D, Clucas B, Hunter J, Ketchum J, McHugh J, Hiraldo F (2008) Top predators as conservation tools: Ecological rationale, assumptions, and efficacy. Annual Review of Ecology, Evolution and Systematics 39: 1–19 Servin J, Rau JR, Delibes M (1991) Activity pattern of the red fox Vulpes vulpes in Doñana, SW Spain. Acta Theriologica 36: 369–373 Shivik JA, Gruver KS, De Liberto TJ (2000) Preliminary evaluation of new cable restraints to capture coyotes. Wildlife Society Bulletin 28: 606–613 Shivik JA, Martin DJ, Pipas MJ, Turnan J, DeLiberto TJ (2005). Initial comparison: jaws, cables, and cage-traps to capture coyotes. Wildlife Society Bulletin 33: 1375–1383 Short MJ, Weldon AW, Richardson SM, Reynolds JC (2012) Selectivity and injury risk in an improved neck snare for live-capture of foxes. Wildlife Society Bulletin 36: 208–219 Sillero-Zubiri C, Laurenson MK (2001) Interactions between carnivores and local communities: conflict or coexistence? En: Gittleman J, Funk K, MacDonald D, Wayne R (Eds.) Carnivore conservation, conservation biology series 5. Cambridge University Press, Cambridge, Pp 282–312 Sillero-Zubiri C, Schwitzer D (2004) Management of wild canids in human-dominated landscapes. En: Sillero-Zubiri C, Hoffman M, Macdonald DW (Eds.) Canids: 189 foxes, wolves, jackals and dogs. Status survey and conservation action plan. IUCN, Gland, Pp. 257–266 Sillero-Zubiri C, Hoffmann M, Macdonald DW (2004) Canids: Foxes, Wolves, Jackals and Dogs. Status Survey and Conservation Action Plan. IUCN/SSC Canid Specialist Group. Gland, Switzerland and Cambridge Sillero-Zubiri C (2009) Family Canidae (Dogs). En: Wilson DE, Mittermeier RA (Eds.) Handbook of the Mammals of the World. Vol. 1. Carnivores. Lynx Editions, Barcelona, Pp 352–446 Silva AP, Rosalino LM, Johnson PJ, Macdonald DW, Anderson N, Kilshaw K (2013) Local-level determinants of wildcat occupancy in Northeast Scotland. European Journal of Wildlife Research 59: 449–453 Sinclair ARE, Pech RP (1996) Density dependence, stochasticity, compensation and predator regulation. Oikos 75: 164–173 Sinclair ARE, Mduma S, Brashares JS (2003) Patterns of predation in a diverse predator-prey system. Nature 425: 288–290 Smith GC, Cheeseman C, Clifton-Hadley RS, Wilkinson D (2001) A model of bovine tuberculosis in the badger Meles meles: an evaluation of control strategies. Journal of Applied Ecology 38: 509–519 Smith RK, Pullin AS, Stewart GB, Sutherland WJ (2010) Effectiveness of predator removal for enhancing bird populations. Conservation Biology 24: 820–829 Smedshaug CA, Selas V, Lund SE, Sonerud GA (1999) The effect of a natural reduction of red fox Vulpes vulpes on small game hunting bags in Norway. Wildlife Biology 5: 157–166 Smith GC, Cheeseman C, Clifton-Hadley RS, Wilkinson D (2001) A model of bovine tuberculosis in the badger Meles meles: an evaluation of control strategies. Journal of Applied Ecology 38: 509–519 Smith RK, Pullin AS, Stewart GB, Sutherland WJ (2010) Effectiveness of predator removal for enhancing bird populations. Conservation Biology 24: 820–829 Sobrino R, Acevedo P, Escudero MA, Marco J, Gortázar C (2008) Carnivore population trends in Spanish agrosystems after the reduction in food availability due to rabbit decline by rabbit haemorrhagic disease and improved waste management. European Journal of Wildlife Research 55: 161–165 Soler JJ, Soler M (1991) Análisis comparado del régimen alimenticio durante el periodo otoño-invierno de tres especies de córvidos en un área de simpatría. Ardeola 38: 69–89 Soler JJ, Soler M, Martínez JG (1993) Grit ingestion and cereal consumption in five corvid species. Ardea 81: 143–149 190 Soler JJ, Martínez JG, Soler M, Møller AP (1999) Host sexual selection and cuckoo parasitism: an analysis of nest size in sympatric and allopatric magpie Pica pica populations parasitized by the great spotted cuckoo Clamator glandarius. Proceedings of the Royal Society of London B 266: 1–7 Soler JJ (2012) Críalo europeo – Clamator glandarius. En: Enciclopedia Virtual de los Vertebrados Españoles. Salvador, A., Morales, M. B. (Eds.). Museo Nacional de Ciencias Naturales, Madrid. http://www.vertebradosibericos.org/ Soler M, Alcalá N, Soler JJ (1990) Alimentación de la grajilla Corvus monedula en tres zonas del sur de España. Doñana Acta Vertebrata 17: 17–48 Soler M, Martínez JG, Soler JJ (1996) Effects of brood parasitism by the Great Spotted Cuckoo on the breeding success of the Magpie host: an experimental study. Ardeola 43: 87–96 Sollman R, Mohamed A, Samejima H, Wilting A (2013) Risky business or simple solution – Relative abundance indices fromcamera-trapping. Biological Conservation 159: 405–412 Soriguer RC (1981) Biología y dinámica de una población de conejos (Oryctolagus cuniculus) en Andalucía occidental. Doñana Acta Vertebrata 8: 1–379 Soriguer RC, Carro C, Fandos P, Márquez FJ (2003) La diversidad y abundancia de los micromamíferos ibéricos. En: Universidad de Jaen (Ed.) In Memorian al Prof. Dr. Isidoro Ruiz Martínez, Pp. 439–478 Statsoft (2001) STATISTICA for Windows. Versión. 6.0. Statsoft Inc., Tulsa StatSoft (2011) Statistica© (Data Analysis Software System), Version 10, Tulsa Stefanescu C, Herrando S, Páramo F (2004) Butterfly species richness in the north-west Mediterranean Basin: the role of natural and human-induced factors. Journal of Biogeography 31: 905–915 Stoate C, Szczur J (2001) Could game management have a role in the conservation of farmland passerines? A case study from a Leicestershire farm. Bird Study 48: 279–292 Stoate C, Szczur J (2005) Predator control as part of a land management system: impacts on breeding success and abundance of passerines. Wildlife Biology in Practice 1: 53–59 Stone E, Trost CH (1991). The effects of supplemental food on nest dispersion in blackbilled magpies. Condor 93: 452–454 Suárez F, Yanes M, Herranz J, Manrique J (1993) Nature reserves and the conservation of Iberian shrubsteppe passerines: the paradox of nest predation. Biological Conservation 64: 77–81 191 Such-Sanz A (2003) Dieta estival del zorro rojo (Vulpes vulpes, Linnaeus, 1758) en naranjales del País Valencià. VI Jornadas de la Sociedad Española para la Conservación y estudio de los Mamíferos (SECEM). Ciudad Real Sunquist ME, Sunquist FC (1989) Ecological constraints on predation by large felids. En: Gittleman J (Ed.) Carnivore behavior, ecology, and evolution, Vol 1. Cornell University Press, Ithaca, New York, Pp 283–301 Sunquist ME (1989) Comparison of spatial and temporal activity of red foxes and gray foxes in north-central Florida. Florida Field Naturalist 17: 11–18 Suvorov P, Svobodová J, Koubová M, Dohnalová L (2012) Ground nest depredation by European Black-billed Magpies Pica pica: an experimental study with artificial nests. Acta Ornithologica 47: 55–61 Tapper SC, Potts GR, Brockless MH (1996) The effect of an experimental reduction in predation pressure on the breeding success and population density of grey partridges Perdix perdix. Journal of Applied Ecology 33: 965–978 Teerink BJ (1991) Hair of West -European mammals. Atlas and identification key. Cambridge University Press: Cambridge Thirgood SJ, Redpath SM, Haydon DT, Rothery P, Newton I, Hudson PJ (2000b) Habitat loss and raptor predation: disentangling long- and short-term causes of red grouse declines. Proceedings of the Royal Society of London Series 267: 651–656 Thirgood SJ, Redpath SM, Newton I, Hudson P (2000a) Raptors and red grouse: conservation conflicts and management solutions. Conservation Biology 14: 95– 104 Thomson DL, Green RE, Gregory RD, Baillie SR (1998) The widespread declines of songbirds in rural Britain do not correlate with the spread of their avian predators. Proceedings of the Royal Society of Lond B 265: 2057–2062 Torre I (2004) Distribution, Population Dynamics and Habitat Selection of Small Mammals in Mediterranean Environments: the Role of Climate, Vegetation Structure, and Predation Risk. Tesis Doctoral. Universidad de Barcelona Torre I, Arrizabalaga A, Díaz M (2002) Ratón de campo (Apodemus sylvaticus). Galemys 14: 1–26 Towerton AL, Penman TD, Kavanagh RP, Dickman CR (2011) Detecting pest and prey responses to fox control across the landscape using remote cameras. Wildlife Research 38: 208–220 Travaini A, Laffitte R, Delibes M (1996) Leg-hold trapping red-foxes (Vulpes vulpes) in Doñana National Park, efficiency, selectivity, and injuries. Journal of Wildlife Research 1: 52–56 Treves A, Karanth KU (2003) Human-carnivore conflict and perspectives on carnivore management worldwide. Conservation Biology 17: 1491–1499 192 Treves A, Naughton-Treves L, Harper EK, Mladenoff DJ, Rose RA, Sickley TA, Wydeven AP (2004) Predicting human-carnivore conflict, a spatial model derived from 25 years of data on wolf depredation on livestock. Conservation Biology 18: 114–125 Treves A, Naughton-Treves L (2005) Evaluating lethal control in the management of human-wildlife conflict. En: RW Woodroofe, Thirgood S, Rabinowitz A (Eds.) People and Wildlife, Conflict or Coexistence? Cambridge University Press, Cambridge. Pp. 86–106 Treves A, Bruskotter J (2014) Tolerance for predatory wildlife. Science 344: 476–477 Trewby LD, Wilson GJ, Delahay RJ, Walker N, Young R, Davison J, Cheeseman C, Robertson PA, Gorman ML, McDonald RA (2008) Experimental evidence of competitive release in sympatric carnivores. Biology Letters 4: 170–172 Tryjanowski P, Sparks TH, Kamieniarz R, Panek M (2009) The relationship between hunting methods and sex, ageand body weight in a non-trophy animal, the red fox. Wildlife Research 36: 106–109 Tuomainen U, Candolin U (2011) Behavioural responses to human-induced environmental change. Biological Reviews 86: 640–657 United States of America–European Community (1998) Standards for the humane trapping of specified terrestrial and semi-aquatic mammals between the United States of America and the European community. Official Journal L 219 of 7 August 1998: 0026–0037 Urios V, Plou J (1986) Estudio de la densidad y alimentación del zorro común (V. vulpes) en la provincia de Valencia. Informe inédito. Conselleria de Agricultura de Valencia. Valencia Valkama J, Korpimaki E, Arroyo B, Beja P, Bretagnolle V, Bro E, Kenward R, Mañosa S, Redpath SM, Thirgood SJ, Viñuela J (2005) Birds of prey as limiting factors of gamebird populations in Europe, a review. Biological Reviews 80: 171–203 Valverde JA (1971) El lobo español. Montes 159: 229–241 Vargas JM (2002) Alerta cinegética. Reflexiones sobre el futuro de la caza en España. Otero, Madrid, España. 398 Pp Vericad JR (1971) Estudio faunístico y biológico de los mamíferos del Pirineo. Informe inédito. CSIC, Jaca Vickery J, Carter N, Fuller RJ (2002) The potential value of managed cereal field margins as foraging habitats for farmland birds in the UK. Agriculture, Ecosystems & Environment 89: 41–52 Villafuerte R, Viñuela J, Blanco JC (1998) Extensive predator persecution caused by population crash in a game species: the case of red kites and rabbits in Spain. Biological Conservation 84: 181–188 193 Villar N, Lambin X, Evans D, Pakeman R, Redpath S (2013) Experimental evidence that livestock grazing intensity affects the activity of a generalist predator. Acta Oecologica 49: 12–16 Virgós E, Llorente M, Cortés Y (1999) Geographical variation in genet (Genetta genetta L.) diet: a literature review. Mammal Review 29: 119–128 Virgós E (2001) Relative value of riparian woodlands in landscapes with different forest cover for mediumsizedIberian carnivores. Biodivers Conservation 10: 1039– 1049 Virgós E (2002) Are habitat generalists affected by forest fragmentation? A test with Eurasian badgers (Meles meles) in coarse-grained fragmented landscapes of Central Spain. Journal of Zoology 258: 313–318 Virgós E, Cabezas-Díaz S, Malo A, Lozano J, López-Huertas D (2003) Factors shaping European rabbit abundance in continuous and fragmentedpopulations of central Spain. Acta Theriologica 48: 113–122 Virgós E, Revilla E, Mangas G, Barea-Azcón JM, Rosalino LM, De Marinis AM (2005) Revisión de la dieta del tejón (Meles meles) en la Península Ibérica: comparación con otras localidades de su área de distribución natural. En: Virgós E, Revilla E, Mangas JG, Domingo-Roura X (Eds.) Ecología y Conservación del Tejón en Ecosistemas Mediterráneos. Sociedad Española para la Conservación y Estudio de los Mamíferos, Málaga. Pp. 67–80 Virgós E, Travaini A (2005) Relationship between small-game hunting and carnivore diversity in central Spain. Biodiversity and Conservation 14: 3475–3486 Virgós E, Cabezas-Díaz S, Mangas JG, Lozano J (2010) Spatial distribution models in a frugivorous carnivore, the stone marten (Martes foina): is the fleshy-fruit availability a useful predictor? Animal Biology 60: 423–436 Virgós E, Lozano J, Mangas JG (2010) El debate sobre el control de los depredadores. Quercus 292: 80–81 Voigt DR, Macdonald DW (1984) Variation in the spatial and social behaviour of the red fox. Acta Zoologica Fennica 171: 261–265 Vulla E, Hobson KA, Korsten M, Leht M, Martin AJ, Lind A, Männil P, Valdmann H, Saarma U (2009) Carnivory is positively correlated with latitude among omnivorous mammals: evidence from brown bears, badgers and pine martens. Annales Zoologici Fennici 46: 395–415 Wang SW, Macdonald DW (2009) Feeding habits and niche partitioning in a predator guild composed of tigers, leopards and dholes in a temperate ecosystem in central Bhutan. Journal of Zoology 277: 275–283 Wang XH, Trost CH (2000) Trapping territorial black-billed Magpies. Journal of Field Ornithology 71: 730–735 194 Way J G, Ortega IM, Auger P J, Strauss EG (2002) Box-trapping eastern coyotes in southeastern Massachusetts. Wildlife Society Bulletin 30: 695–702 Way JG (2012) Seasonal and study site variation in box-trapping eastern coyotes. Canids News 15.2 (On-line). URL: http,//www. http,//canids.org/canidnews/15/Box-trapping_eastern_coyotes.pdf Webbon CC, Baker PJ, Cole NC, Harris S (2006) Macroscopic prey remains in the winter diet of foxes Vulpes vulpes in rural Britain. Mammal Review 36: 85–97 Whitfield DP, McLeod DRA, Watson J, Fielding AH, Haworth PF (2003) The association of grouse moor in Scotland with the illegal use of poisons to control predators. Biological Conservation 114: 157–163 Whitfield DP, Fielding AH, Mcleod DRA, Morton K, Stirling-Aird P, Eaton MA (2007) Factors constraining the distribution of Golden Eagles Aquila chrysaetos in Scotland: Capsule between 1992 and 2003 persecution appeared to be the main influential factor. Bird Study 54: 199–211 Whittaker RJ, Araújo MA, Jepson P, Ladle RJ,Watson JEM, Willis KJ (2005) Conservation biogeography: assessment and prospect. Diversity and Distributions 11: 3–23 Williams D, Acevedo P, Gortázar C, Escudero MA, Labarta JL, Marco J, Villafuerte R (2007) Hunting for answers: rabbit (Oryctolagus cuniculus) population trends in northeastern Spain. European Journal of Wildlife Research 53: 19–28 Wingfield JC, Ball GF, Dufty AM, Hegner RE, Ramenofsky M (1987) Testosterone and aggression in birds. American Scientist 75: 602–608 Woodroffe R, Frank, LG (2005) Lethal control of African lions (Panthera leo): local and regional population impacts. Animal Conservation 8: 91–98 Woodroffe R, Thirgood S, Rabinowitz A (2005) People and wildlife, conflict or coexistence? Cambridge University Press: Cambridge Yanes M, Suyárez F (1996) Incidental nest predation and lark conservation in an Iberian semiarid shrubsteppe. Conservation Biology 10: 881–887 Yanes M, Martín JM; Suárez F (1997) Aspectos ecológicos del zorro (Vulpes vulpes) en la R.O. de Las Almoladeras, Almería: su implicación en la gestión de un enclave estepario protegido. Investigación y gestión 1: 73–80 Yanes M, Herranz J, De la Puente J, Suárez F (1998) La perdiz roja. Identidad de los depredadores e intensidad de la depredación. En: I Curso. La perdiz roja. FEDENCA, Alcobendas, Madrid, Pp 135–147 Yom-Tov Y, Yom-Tov S, Barreiro J, Blanco JC (2007) Body size of the red fox Vulpes vulpes in Spain: the effect of agriculture. Biological Journal of the Linnean Society 90: 729–734 195 Zapata SC, Travaini A, Ferreras P, Delibes M (2007) Analysis of trophic structure of two carnivore assemblages by means of guild identification. European Journal of Wildlife Research 53: 276–286 Zar JH (1984) Biostatistical Analysis. Prentice-Hall, Englewood Cliffs, New Jersey, USA Zhou YB, Newman C, Xu WT, Buesching CD, Zalewski A, Kaneko Y, Macdonald DW, Xie ZQ (2011) Biogeographical variation in the diet of Holarctic martens (genus Martes, Mammalia: Carnivora: Mustelidae): adaptive foraging in generalists. Journal of Biogeography 38: 137–147 Zinkivskay A, Nazir F, Smulders TV (2009) What–Where–When memory in magpies (Pica pica). Animal Cognition 12: 119–125 Zuberogoitia I, González-Oreja JA, Zabala J, Rodríguez-Refojos C (2010) Assessing the control/eradication of an invasive species, the American mink, based on field data; how much would it cost? Biodiversity and Conservation 19: 1455–1469 Zuur AF, Ieno EN, Smith GM (2007) Analysing ecological data. Springer, New York Zuur AF, Ieno EN, Walker NJ, Saveliev AA, Smith GM (2009) Mixed effects models and extensions in ecology with R. New York: Springer Science+Business Media 196 APÉNDICES 197 Appendix 1.1. Studies of the diet of the red fox in Iberia used in this review, with an indication of the latitude (Lat.), longitude (Long.) and altitude (Alt.) where the study took place, year, sample size, duration of the study, predominant habitat, season and type of material. The Map ID (see Fig. 1) is also shown. Lat. Long. Amores (1975) 38.17 -5.25 Map ID 1 Study duration 22 Habitat Season Material 1973 Sample size 121 M. Scrub Annual Stomach Angelo (2000) 37.38 -7.63 2 1999 81 9 - Annual Scat Angelo (2000) 37.38 -7.63 2 1999 42 3 - Winter Scat Ballesteros & Degollada (2002) 41.68 2.02 3 1997 140 12 Forest Annual Scat Barrull and Mate (2007) 41.32 0.97 4 2006 985 21 - Annual Scat Barrull and Mate (2007) 41.32 0.97 4 2006 354 5,25 - Autumn Scat Barrull and Mate (2007) 41.32 0.97 4 2006 205 5,25 - Spring Scat Barrull and Mate (2007) 41.32 0.97 4 2006 241 5,25 - Summer Scat Barrull and Mate (2007) 41.32 0.97 4 2006 185 5,25 - Winter Scat Bermejo and Guitián (2000) 42.57 -6.63 5 1990 30 4 - Autumn Scat Bermejo and Guitián (2000) 42.85 -6.82 6 1990 44 3 - Autumn Scat Blanco (1986) 40.75 -4 7 1984 97 7 Forest Annual Scat Blanco (1986) 40.75 -4 7 1984 30 2 Forest Autumn Scat Blanco (1986) 40.75 -4 7 1984 18 2 Forest Spring Scat Blanco (1986) 40.75 -4 7 1984 49 3 Forest Summer Scat Blanco (1988) 40.75 -4 7 1985 414 27 Forest Annual Scat Blanco (1988) 40.75 -4 7 1985 90 6 Forest Autumn Scat Blanco (1988) 40.75 -4 7 1985 104 9 Forest Spring Scat Blanco (1988) 40.75 -4 7 1985 131 6 Forest Summer Scat Blanco (1988) 40.75 -4 7 1985 92 6 Forest Winter Scat Reference Year 198 Lat. Long. Narváez, et al. (2008) 37.92 -5.75 Map ID 8 Study duration 24 Habitat Season Material 2005 Sample size 138 M. Scrub Annual Scat Braña & Del Campo (1980) 43.1 -5.83 9 1978 68 36 Forest Annual Stomach Calviño et al. (1984) 42.83 -8.22 10 1978 429 84 - Annual Stomach Calviño et al. (1984) 42.83 -8.22 10 1978 230 21 - Autumn Stomach Calviño et al. (1984) 42.83 -8.22 10 1978 24 21 - Spring Stomach Calviño et al. (1984) 42.83 -8.22 10 1978 49 21 - Summer Stomach Calviño et al. (1984) 42.83 -8.22 10 1978 158 21 - Winter Stomach Calzada (2000) 37.15 -6.43 11 1996 1295 35 M. Scrub Annual Scat Carvalho and Gomes (2004) 41.82 -7.78 12 1999 193 12 M. Scrub Annual Scat Carvalho (2001) 41.82 -7.78 12 2000 193 12 M. Scrub Annual Scat Carvalho (2001) 41.82 -7.78 12 2000 38 3 M. Scrub Autumn Scat Carvalho (2001) 41.82 -7.78 12 2000 44 3 M. Scrub Spring Scat Carvalho (2001) 41.82 -7.78 12 2000 65 3 M. Scrub Summer Scat Carvalho (2001) 41.82 -7.78 12 2000 46 3 M. Scrub Winter Scat Castroviejo et al. (1984) 37.15 -6.43 13 1984 81 12 - Annual Stomach Castroviejo et al. (1984) 40.42 -3.67 14 1984 103 12 - Annual Stomach Castroviejo et al. (1984) 39.5 -6.33 15 1984 230 12 - Annual Stomach Castroviejo et al. (1984) 42.42 -8.63 16 1984 120 12 - Annual Stomach Castroviejo et al. (1984) 43.42 -7.25 17 1984 75 12 - Annual Stomach De Carvalho and Alves Alexandre (1994) 41.92 -6.6 18 1993 656 12 Forest Annual Scat Delibes Mateos et al. (2007) 38.48 -4.5 19 2002 35 3 - Summer Scat Delibes Mateos et al. (2007) 38.48 -4.5 19 2002 24 3 - Summer Scat Delibes Mateos et al. (2007) 38.48 -4.5 19 2002 25 3 - Summer Scat Reference Year 199 Lat. Long. Delibes Mateos et al. (2007) 38.48 -4.5 Map ID 19 2002 Sample size 114 Study duration 3 Delibes Mateos et al. (2007) 38.48 -4.5 19 2002 31 3 Dos Santos Correia (1993) 40.3 -7.07 20 1991 241 15 Esmeriz (2001) 40.58 -7.58 21 1998 207 12 Esmeriz (2001) 40.58 -7.57 21 1998 176 Esmeriz (2001) 40.58 -7.57 21 1998 Esmeriz (2001) 40.58 -7.57 21 Esmeriz (2001) 40.58 -7.57 Esmeriz (2001) 40.58 Esmeriz (2001) Reference Habitat Season Material - Summer Scat - Summer Scat Annual Scat Forest Annual Scat 12 Forest Annual Scat 26 3 Forest Autumn Scat 1998 48 3 Forest Autumn Scat 21 1998 92 3 Forest Spring Scat -7.57 21 1998 23 3 Forest Spring Scat 40.58 -7.57 21 1998 73 3 Forest Summer Scat Esmeriz (2001) 40.58 -7.57 21 1998 65 3 Forest Summer Scat Esmeriz (2001) 40.58 -7.57 21 1998 16 3 Forest Winter Scat Esmeriz (2001) 40.58 -7.57 21 1998 40 3 Forest Winter Scat Fedriani (1996) 37.15 -6.43 11 1996 129 24 M. Scrub Annual Scat Fedriani (1996) 37.15 -6.43 11 1996 164 24 Agr/Dehesa Annual Scat Fedriani et al. (1999) 37.15 -6.43 11 1993 293 26 M. Scrub Annual Scat Fedriani et al. (1999) 37.15 -6.43 11 1993 99 M. Scrub Spring Scat Fedriani et al. (1999) 37.15 -6.43 11 1993 123 M. Scrub Summer Scat 43 -2.83 22 1997 191 5 Forest Spring Scat González-Prat (1995)a 41.77 4.45 23 1995 - 3 - Autumn Scat González-Prat (1995)a 42.5 0.58 24 1995 - 3 - Winter Scat González-Prat (1995)b 41.3 1.8 25 1995 - 12 M. Scrub Annual Scat Guitian and Callejo (1983) 42.55 -7.27 26 1978 38 60 Agr/Dehesa Annual Scat Fernández and Ruiz de Azua (2005) Year 200 Lat. Long. Herranz et al. (1999) 39.5 -2.83 Map ID 27 1975 Sample size 105 Study duration 12 Dos Santos Marques (2003) 40.53 -3.67 28 2002 70 López (1999) 41.8 1.1 29 1999 López (1999) 41.8 1.1 29 López (1999) 41.8 1.1 López (1999) 41.8 López (1999) Reference Year Habitat Season Material - Annual Stomach 8 M. Scrub Annual Scat 661 22 - Annual Scat 1999 113 6 - Autumn Scat 29 1999 248 6 - Spring Scat 1.1 29 1999 143 6 - Summer Scat 41.8 1.1 29 1999 155 4 - Winter Scat Martín (2008) 41.3 1.8 25 1999 428 11 Annual Scat Monterroso et al. (2006) 37.65 -7.63 30 2005 45 12 M. Scrub Annual Scat Negrões (2000) 42.03 -8.13 31 1998 490 24 Agr/Dehesa Annual Scat Negrões (2000) 42.03 -8.13 31 1997 90 6 Agr/Dehesa Autumn Scat Negrões (2000) 42.03 -8.13 31 1998 144 6 Agr/Dehesa Spring Scat Negrões (2000) 42.03 -8.13 31 1998 151 6 Agr/Dehesa Summer Scat Negrões (2000) 42.03 -8.13 31 1997 105 6 Agr/Dehesa Winter Scat Padial et al. (2002) 37.07 -3.55 32 1997 132 12 M. Scrub Annual Scat Padial et al. (2002) 37.13 -3.38 33 1997 74 12 Forest Annual Scat Palomares and Ruiz-Martínez (1994) 37.37 -2.83 34 1991 38 1 M. Scrub Spring Scat Palomares and Ruiz-Martínez (1994) 37.37 -2.83 34 1991 90 1 M. Scrub Spring Scat Sarmento et al. (1999) 40.3 -7.07 20 1996 306 12 M. Scrub Annual Scat Sarmento et al. (1999) 40.3 -7.07 20 1996 63 3 M. Scrub Autumn Scat Sarmento et al. (1999) 40.3 -7.07 20 1996 59 3 M. Scrub Spring Scat Sarmento et al. (1999) 40.3 -7.07 20 1996 125 3 M. Scrub Summer Scat Sarmento et al. (1999) 40.3 -7.07 20 1996 59 3 M. Scrub Winter Scat 201 Lat. Long. Such-Sanz (2003) 38.95 -0.58 Map ID 35 Such-Sanz (2003) 39.07 -0.27 Urios and Plou (1986) 39.42 Vericard (1971) Yanes et al. (1996) Reference Study duration 4 Habitat Season Material 1998 Sample size 41 Agr/Dehesa Summer Scat 36 1998 40 4 Agr/Dehesa Summer Scat -0.83 37 1985 237 24 - Annual Stomach 42.52 -0.75 38 1971 66 12 Forest Annual Stomach 36.83 -2.42 39 1993 69 3 M. Scrub Spring Scat Year 202 Appendix 1.2. Fox diets as described in the reviewed studies (see Figure 1 and Appendix S1). The information is presented as the frequency of occurrence (FO) of each prey group.We also indicate the values of trophic diversity (Herrera diversityindex, D) recorded for each study Reference ID Lagomorph Micromammals Birds Reptiles Invertebrates Fruits/seed Carrion/garbage D Amores (1975) 1 71.1 38.8 29.7 0 60.3 0 6.6 12.5 Angelo (2000) 2 11.59 6.04 4.79 0.25 58.69 16.37 0 12.1 Angelo (2000) 2 29.29 16.16 9.09 0 12.12 30.3 3 8.3 Ballesteros and Degollada (2002) 3 1.3 44 12 2 60 60 0 10.3 Barrull and Mate (2007) 4 1.22 12.69 7.51 1.01 41.83 70.86 11.06 7.4 Barrull and Mate (2007) 4 1.41 12.15 4.24 0.56 29.38 80.51 11.86 7.8 Barrull and Mate (2007) 4 0.49 16.1 11.22 0 58.05 43.41 20 8.3 Barrull and Mate (2007) 4 1.24 8.3 8.3 2.9 57.26 70.54 3.73 7.4 Barrull and Mate (2007) 4 1.62 15.67 8.65 0.54 27.57 83.24 9.19 7.5 Bermejo and Guitián (2000) 5 0 16.6 3.3 0 26.6 100 20 9.5 Bermejo and Guitián (2000) 6 0 70.4 22.7 0 20.4 4.5 9.1 9.9 Blanco (1986) 7 9.3 38.1 0 0 47.4 70.1 26.8 12.5 Blanco (1986) 7 0 23.38 0 0 63.3 70 26.66 10.6 Blanco (1986) 7 17.85 67.85 0 0 60.7 7.15 3.55 9.7 Blanco (1986) 7 8.23 34.3 0 0 85.74 46.97 34.66 8.4 Blanco (1988) 7 22 40.3 11.1 2.2 47.3 33.4 20 5.2 Blanco (1988) 7 13.3 28.9 5.6 0 34.4 61.2 22.2 7.0 Blanco (1988) 7 16.3 51 12.5 6.7 53.8 18.3 18.3 4.9 Blanco (1988) 7 10.7 29.8 13.7 1.5 74.8 54.9 27.5 5.1 Blanco (1988) 7 50 52.2 7.6 0 6.5 4.3 8.7 8.3 Narváez et al. (2008) 8 0.7 44.2 6.5 2.9 82.6 8 5 7.7 203 Reference ID Lagomorph Micromammals Birds Reptiles Invertebrates Fruits/seed Carrion/garbage D Braña and Del Campo (1980) 9 0 64.7 23.52 5.8 45.6 40 19 8.5 Calviño et al. (1984) 10 5.6 53.4 16.5 0.23 49.9 54.1 71.6 5.7 Calviño et al. (1984) 10 2.6 44.7 18.4 2.6 34.2 78.9 2.6 5.4 Calviño et al. (1984) 10 20.5 59.1 6.8 2.3 31.8 2.3 11.4 6.9 Calviño et al. (1984) 10 27.7 24.6 7.7 13.8 55.4 4.6 3.1 6.1 Calviño et al. (1984) 10 10.9 87 13 0 8.7 15.2 0 6.4 Calzada (2000) 11 73 6 9 0 72 16 16 9.1 Carvalho and Gomes (2004) 12 19.6 65.7 10.8 7.8 27.5 17.7 2.9 5.8 Carvalho (2001) 12 17.1 51.3 10.9 5.7 34.7 21.2 4.1 5.8 Carvalho (2001) 12 2.6 44.7 18.4 2.6 34.2 78.9 2.6 6.4 Carvalho (2001) 12 20.5 59.1 6.8 2.3 31.8 2.3 11.4 6.8 Carvalho (2001) 12 27.7 24.6 7.7 13.8 55.4 4.6 3.1 6.2 Carvalho (2001) 12 10.9 87 13 0 8.7 15.2 0 9.8 Castroviejo et al. (1984) 13 0 29 2.5 1.2 3.7 61 37.3 11.1 Castroviejo et al. (1984) 14 13 47 0 0 21 61 25 12.7 Castroviejo et al. (1984) 15 14 34 12 0 35 42 26.5 8.7 Castroviejo et al. (1984) 16 6 33 0.8 0 53 57 10 10.3 Castroviejo et al. (1984) 17 0 34 0 0 25 54.6 26 16.9 De Carvalho and A. Alexandre (1994) 18 7.93 96.96 86.44 0 0 28.97 15.55 12.5 Delibes Mateos et al. (2007) 19 37.2 31.4 62.9 20 91.4 37.1 11.4 3.2 Delibes Mateos et al. (2007) 19 69.6 30.4 39.1 8.7 95.6 17.4 8.7 4.0 Delibes Mateos et al. (2007) 19 84 40 48 12 88 16 12 3.5 Delibes Mateos et al. (2007) 19 49.1 37.7 51.7 14.9 92.1 23.7 7.9 3.6 204 Reference ID Lagomorph Micromammals Birds Reptiles Invertebrates Fruits/seed Carrion/garbage D Delibes Mateos et al. (2007) 19 19.3 48.4 51.6 16.1 93.5 19.3 16.1 3.6 Dos Santos Correia (1993) 20 11.1 23.4 9.9 3.5 36.2 17.3 9.7 6.3 Esmeriz (2001) 21 29 42 27 0 46 51 0 12.1 Esmeriz (2001) 21 1 56 11 0 81 20 0 14.0 Esmeriz (2001) 21 4 51 37 0 51 67 0 8.6 Esmeriz (2001) 21 0 41 3 0 95 27 0 11.5 Esmeriz (2001) 21 26 48 22 0 45 25 0 8.5 Esmeriz (2001) 21 0 73 21 0 82 3 0 11.4 Esmeriz (2001) 21 28 29 26 0 47 76 0 8.1 Esmeriz (2001) 21 0 40 3 0 90 28 0 11.5 Esmeriz (2001) 21 31 43 38 0 38 50 0 8.0 Esmeriz (2001) 21 2 90 12 0 49 10 0 9.9 Fedriani (1996) 11 55.8 7 7 8.5 83.7 24.8 27.1 4.9 Fedriani (1996) 11 53 6.1 16.5 9.1 93.3 6.7 23.1 5.2 Fedriani et al. (1999) 11 53.6 0 0 0 89.1 14.7 22.5 16.8 Fedriani et al. (1999) 11 55.6 0 0 0 92.9 2 19.2 11.7 Fedriani et al. (1999) 11 65 0 0 0 90.2 13 22.8 10.8 Fernández & Ruiz de Azua (2005) 22 1.1 64.3 17.8 2.1 62.3 37.7 17.2 5.9 González-Prat (1995)a 23 33.33 25.93 55.56 0 14.81 66.67 25.93 5.9 González-Prat (1995)a 24 6.67 20 46.67 0 13.33 46.67 93.34 6.4 González-Prat (1995)b 25 24.8 0 18.55 0 8.25 100 15.3 13.2 Guitian and Callejo (1983) 26 0.64 18.82 1.94 0.43 38.96 38.09 0.21 10.5 Herranz et al. (1999) 27 32.4 34.3 37.1 0 20 80 10.5 8.2 205 Reference ID Lagomorph Micromammals Birds Reptiles Invertebrates Fruits/seed Carrion/garbage D Dos Santos Marques (2003) 28 66 24 17 0 11 30 11 9.0 López (1999) 29 33.13 35.1 3.63 0.45 8.17 21.33 1.51 8.3 López (1999) 29 27.43 32.74 2.65 2.65 12.39 36.28 1.76 7.2 López (1999) 29 37.5 41.53 4.44 0 6.85 9.27 1.2 9.2 López (1999) 29 34.97 36.36 4.9 0 12.59 20.28 0 9.8 López (1999) 29 29.03 25.81 1.94 0 2.58 32.9 1.93 9.6 Martín (2008) 25 8.64 8.4 3.7 0 14.7 35.05 5.14 11.1 Monterroso et al. (2006) 30 52.9 44.7 19.9 0 16.5 30.5 0 12.6 Negrões (2000) 31 17.1 44.1 16.3 0.8 54.1 43.1 16.9 5.4 Negrões (2000) 31 17.8 41.1 12.2 1.1 48.9 63.3 10 5.5 Negrões (2000) 31 22.2 47.9 11.8 2.1 47.2 15.3 25.7 5.3 Negrões (2000) 31 12.6 22.5 15.9 0 78.1 68.9 13.9 6.5 Negrões (2000) 31 16.2 72.4 26.7 0 33.3 26.7 15.2 6.4 Padial et al. (2002) 32 24.2 27.3 6.1 4.6 28 41 28.8 5.2 Padial et al. (2002) 33 0 52.7 4.1 1.4 32.4 42 34.2 9.9 Palomares and Ruiz-Martínez (1994) 34 0.7 58.5 8.1 22.2 54.8 14.1 13.3 6.0 Palomares and Ruiz-Martínez (1994) 34 0 72.2 10 11.1 17.8 0 2.2 10.5 Sarmento et al. (1999) 20 7.5 39.9 11.4 7.5 59.1 41.3 7.5 5.3 Sarmento et al. (1999) 20 9.5 27 7.9 3.2 66.7 57.1 11.1 5.5 Sarmento et al. (1999) 20 1.7 59.3 22 16.9 74.6 11.9 10.2 5.4 Sarmento et al. (1999) 20 6.4 28.8 11.2 4 48.8 56.8 3.2 6.1 Sarmento et al. (1999) 20 13.6 62.7 6.8 10.2 69.5 18.6 10.2 5.1 Such-Sanz (2003) 35 22 34 34 10 61 85 0 5.9 206 Reference ID Lagomorph Micromammals Birds Reptiles Invertebrates Fruits/seed Carrion/garbage D Such-Sanz (2003) 36 25 82 85 3 72 48 0 5.7 Urios and Plou (1986) 37 9.4 13.4 26.3 0.9 17 48.8 12.3 6.5 Vericard (1971) 38 7.6 39.93 15.1 1.5 15.5 39.69 15.47 6.2 Yanes et al. (1996) 39 97.1 2.9 11.6 1.4 49.2 0 8.7 8.7 207 Appendix 2.1. Detailed description of magpie diet composition. The No. of analyzed gizzards (Gizzard), and the minimum No. of items found (Items) for each food group are shown. For each food group, we also present the frequency of occurrence (FO), the relative frequency of occurrence (RF) and the average % volume (VOL). Data is independently presented in terms of overall magpie diet (Total) and in each study area (A1 and A2). Gizzards Items (n = 1016) FO RF VOL Total (n = 118) A1 (n = 61) A2 (n = 57) Total A1 A2 Total A1 A2 Total A1 A2 Total A1 A2 Food type Coleoptera Formicidae Isopoda Hymenoptera Dermaptera Araneida Diptera Arthropoda larva Hemiptera Arthropoda Gastropoda Hordeum sp. Avena sp. Triticum sp. Indet. Seeds Cereal seed Fruit Eggs Other vegetal Passeriforme Galliforme Birds Apodemus sylvaticus Felis sp. Indet. mammal Mammals Reptile Non-food remains Gastrolith Plastic 98 29 8 5 5 5 1 1 3 111 11 27 13 8 31 79 5 6 40 15 1 20 2 1 1 4 1 47 25 5 2 2 3 0 1 2 56 10 19 2 7 13 43 5 5 27 13 1 17 2 1 1 4 1 51 4 3 3 3 2 1 0 1 55 1 8 11 1 18 36 0 1 13 2 0 3 0 0 0 0 0 195 165 9 6 6 5 1 2 4 393 20 274 114 36 66 490 5 6 39 15 1 22 2 1 1 4 1 79 149 6 2 2 3 0 2 2 245 19 212 6 21 27 266 5 5 26 13 1 19 2 1 1 4 1 116 16 3 4 4 2 1 0 2 148 1 62 108 15 39 224 0 1 13 2 0 3 0 0 0 0 0 83.05 24.58 6.78 4.24 4.24 4.24 0.85 0.85 2.54 94.07 9.32 22.88 11.02 6.78 26.27 66.95 4.24 5.08 33.90 12.71 0.85 16.95 1.69 0.85 0.85 3.39 0.85 77.05 40.98 8.20 3.28 3.28 4.92 0.00 1.64 3.28 91.80 16.39 31.15 3.28 11.48 21.31 70.49 8.20 8.20 44.26 21.31 1.64 27.87 3.28 1.64 1.64 6.56 1.64 89.47 7.02 5.26 5.26 5.26 3.51 1.75 0.00 1.75 96.49 1.75 14.04 19.30 1.75 31.58 63.16 0.00 1.75 22.81 3.51 0.00 5.26 0.00 0.00 0.00 0.00 0.00 20.04 16.96 0.92 0.62 0.62 0.51 0.10 0.21 0.41 40.39 2.06 28.16 11.72 3.70 6.78 50.36 0.51 0.62 4.01 1.54 0.10 2.16 0.21 0.10 0.10 0.41 0.10 13.53 25.51 1.03 0.34 0.34 0.51 0.00 0.34 0.34 41.95 3.25 36.30 1.03 3.60 4.62 45.55 0.86 1.03 4.45 2.23 0.17 3.08 0.34 0.17 0.17 0.68 0.17 29.82 4.11 0.77 1.03 1.03 0.51 0.26 0.00 0.51 38.05 0.26 15.94 27.76 3.86 10.03 57.58 0.00 0.26 3.34 0.51 0.00 0.77 0.00 0.00 0.00 0.00 0.00 29.69 5.76 1.84 1.97 0.47 0.64 0.21 0.17 0.39 41.14 3.07 14.05 4.92 2.92 14.20 36.10 1.55 2.63 10.75 1.20 0.04 3.87 0.05 0.01 0.01 0.07 0.21 14.18 10.07 1.84 1.34 0.25 1.07 0.00 0.33 0.10 29.16 5.89 18.77 1.48 4.26 11.92 36.43 3.00 3.61 16.20 2.21 0.08 5.90 0.10 0.02 0.02 0.13 0.41 46.30 1.16 1.84 2.63 0.70 0.19 0.44 0.00 0.70 53.96 0.05 9.00 8.61 1.49 16.65 35.75 0.00 1.58 4.93 0.12 0.00 1.70 0.00 0.00 0.00 0.00 0.00 10 5 8 3 2 2 37 6 33 3 4 3 8.47 4.24 6.78 4.92 1.69 3.51 3.64 0.62 3.25 0.51 0.39 0.77 0.51 0.13 0.25 0.18 0.79 0.07 208 Appendix 2.2. 209 Appendix 2.3. 210 Appendix 3.1. Description and design of the cage-traps models evaluated. Units of size descriptions are in cm Model Size (width x length x height) Capture Entrances Live Bait Chamber Capture system A B C D E 1020 x 2000 x 1000 450 x 950 x 500 360 x 1450 x 550 450 x 1520 x 500 450 x 2300 x 500 1 1 2 2 2 no no lateral lateral central guillotine-type door+outrigger guillotine-type door+outrigger guillotine-type door+outrigger guillotine-type door+outrigger guillotine-type door+outrigger Model A (CT01 type) 211 Model B (CT01 type) Model C (CT02 type) 212 Model D (CT02 type) Model E (CT03 type) 213 Appendix 3.2. Overall results obtained for each bait-attractant combination (A). Overall results obtained for each trap-type-bait-attractant combination (B). Effort: number of trap-nights Efficiency: foxes/1000 trap-nights; NTcr: non-targets /1000 trap-nights. (A) Bait Attractant Effort Fox Control 654 1 COLL 40 0 Dead FAS 138 0 FU 199 0 VAL 72 0 Control 434 1 COLL 44 0 Alive FAS 39 0 FU 205 3 VAL 243 1 Captures Non-target 7 0 1 2 0 14 0 1 11 4 Efficiency 1.53 0.00 0.00 0.00 0.00 2.31 0.00 0.00 14.63 4.12 NTcr 10.71 0.00 7.25 10.05 0.00 32.30 0.00 25.64 53.66 16.46 Selectivity 13 0 0 7 0 21 20 (B) CT Type Bait Dead A Alive Dead B Alive Dead C Alive TOTAL Attractant Control COLL FAS FU VAL Control COLL FAS FU VAL Control COLL FAS FU VAL Control COLL FAS FU VAL Control COLL FAS FU VAL Control COLL FAS FU VAL Effort 8 nt nt 23 8 137 21 16 72 86 303 nt nt 120 16 nt nt nt nt nt 343 40 138 56 48 297 23 23 133 157 2068 Captures Fox Non-target Efficiency NTcr Selectivity 0 125.00 0 1 0 nt nt nt nt nt nt nt nt nt nt 0 43.48 0 1 0 0 0 0 0 1 12 7.30 87.59 8 0 0 0 0 0 1 0 62.5 0 27.78 152.78 2 11 15 0 46.51 0 4 0 1 4 3.30 13.20 20 nt nt nt nt nt nt nt nt nt nt 0 1 0 8.33 0 0 0 0 0 nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt nt 0 2 0 5.83 0 0 0 0 0 0 7.25 0 1 0 0 0 0 0 0 0 0 0 0 6.73 0 2 0 0 0 0 0 0 0 0 0 1 0 7.52 0 100 6.37 0 1 0 100 6 40 2.90 19.34 13 214 Appendix 4.1. Trap models employed in the experiments: models 1–3 have four capture chambers (a), whereas model 4 has two capture chambers (b). a b 215 Appendix 5.1. Habitat composition of studied sites (%) and rabbit availability (rabbit detections per 100 trap days) for each study locality (Map ID). Localities are ordered according to the increasing intensity of fox control. 216 Appendix 5.2. Carnivores detected during camera trap surveys in each locality (Study site). For each species and locality we show the naïve site occupancy (i.e. proportion of cameras that recorded the presence of the species). We show the overall mean naïve occupancy (mean (SE)). “Detection” is the proportion of localities where each species was present. “1-week positive” is the number of positive 1-week sampling occasions and respective proportion (in brackets) over all sampling occasions for mesocarnivores in each of the study localities. Study site (Map ID) Red fox Stone marten Common genet Egyptian mongoose Eurasian badger Least weasel Wildcat 1 2 3 4 5 6 7 8 9 10 11 12 0.35 0.20 0.56 0.59 0.63 0.50 0.40 0.50 0.87 0.57 0.95 0.67 nd nd nd nd 0.32 0.60 0.20 0.40 0.27 nd 0.35 0.33 nd nd nd nd 0.26 0.25 0.05 0.05 0.20 0.07 0.10 nd nd nd nd 0.12 nd nd 0.05 0.05 0.07 0.21 0.10 nd nd nd nd nd 0.05 0.05 0.05 0.10 0.07 nd 0.30 nd 0.10 nd 0.06 0.06 nd nd nd 0.05 nd nd nd 0.06 nd 0.07 nd nd nd nd 0.10 nd nd nd 0.10 nd Detection (%) 100 58 58 50 50 42 25 Mean Naïve Occupancy (SE) 0.57 (0.06) 0.21 (0.06) 0.08 (0.03) 0.05 (0.02) 0.05 (0.02) 0.03 (0.01) 0.02 (0.01) 1-week positive 254 (31.7 %) 65 (8.1 %) 22 (2.7 %) 17 (2.1 %) 12 (1.5 %) 5 (0.6 %) 5 (0.6 %) 217 218