Amenazas actuales de los felinos neotropicales: la ganadería en el punto de mira

Amenazas actuales de los grandes felinos neotropicales! La ganadería en el punto de mira Pablo Villalva Aguilar! Tesis Doctoral " CITACIÓN RECOMENDADA: Villalva, P. 2022. Amenazas actuales de los grandes felinos neotropicales: La ganadería en el punto de mira. Tesis Doctoral. Universidad de Sevilla, España.! RECOMMENDED CITATION: Villalva, P. 2022. Current threats of neotropical big cats: Cattle ranching in the spotlight. Universidad de Sevilla, Spain. " AMENAZAS ACTUALES DE LOS GRANDES FELINOS NEOTROPICALES: La ganadería en el punto de mira Pablo Villalva Aguilar! Tesis Doctoral! Universidad de Sevilla, 2022 " AMENAZAS ACTUALES DE LOS GRANDES FELINOS NEOTROPICALES: La ganadería en el punto de mira Memoria presentada por el Ldo. Pablo Villalva Aguilar para optar al título de Doctor por la Universidad de Sevilla Universidad de Sevilla,! Sevilla, 2022! Tutor! Dr. Juan Francisco Beltrán Gala! Profesor titular! Departamento de zoología! Universidad de Sevilla! Sevilla - España! Director! Dr. Francisco Palomares Fernandez! Investigador científico! Departamento de Biología de la conservación! Estación Biológica de Doñana-CSIC! Sevilla - ! Universidad de Sevilla! Sevilla - España! Directora! Dra. Marina Zanin Gregorini! Investigadora científica! Departamento de Zoología! Instituto de Biología! Universidade de São Paulo ! São Paulo - SP, Brasil! CONTENTS Chapter 1 General introduction 11 Chapter 2 The differential genetic signatures related to climatic landscapes for jaguars and pumas on a continental scale 29 Chapter 3 A continental approach to jaguar extirpation: a tradeoff between anthropic and intrinsic causes 65 Chapter 4 Uneven tolerance to human disturbance interferes to dominance interactions of top predators 93 Chapter 5 Perceptions and livestock predation by felids in extensive cattle ranching areas of two Bolivian ecoregions 117 Chapter 6 Tiger trade threatens big cats worldwide / Selling jaguars as tigers: impacts of Asian medicine on American big cats 141 Chapter 7 General discussion 155 Conclusiones 174 Agradecimientos 177 Resumen: Jaguares (Panthera onca) y pumas (Puma concolor) son los mayores depredadores del Neotropico y comparten rasgos de su biología, historia evolutiva, modos de vida y conflictos con el ser humano. Sin embargo, el nivel de amenaza entre ambas especies difiere. La distribución actual del jaguar se encuentra profundamente reducida y se categoriza como especie cercana a la extinción (Near Threaten), mientras que el puma, ampliamente distribuido por el Neotropico, se categoriza como de preocupación menor (Least Concern). En esta tesis doctoral estudio los mecanismos que subyacen a estas diferencias, profundizando en la interacción entre la ecología de las especies y las relaciones con el ser humano con la motivación de comprender el efecto que la presión antrópica ejerce sobre estas especies. En el primer capítulo analizamos cómo los patrones climáticos pasados y presentes han afectado a las poblaciones de grandes felinos de manera dispar. Los pumas responden a paisajes climáticos antiguos mientras que los jaguares responden tanto a paisajes climáticos antiguos como modernos, mostrando una mayor vulnerabilidad al cambio climático. Mediante el uso de predicciones climáticas evaluamos distintos escenarios sobre cada especie, obteniendo una aproximación del efecto potencial del cambio climático sobre cada especie en el futuro próximo. En el segundo capítulo usamos el jaguar como modelo de estudio. A través de un modelado ecológico a escala continental comprobamos si las causas de extinción se relacionan con características intrínsecas (e.g. distancia al borde de la distribución) o extrínsecas (e.g. presiones humanas) a la especie. Observamos que el patrón de extinción de la especie es una combinación de ambas causas, destacando la ganadería como la principal de ellas. Una vez establecida la ganadería como impulsor de extinción de los jaguares, en el tercer capítulo estudiamos cómo la presión humana muestra un profundo efecto, no sólo sobre las especies de felino de forma aislada, sino sobre sus relaciones de dominancia. Encontramos que las relaciones intragremiales naturales se ven moduladas por la presión humana y se discute cómo la comunidad de carnívoros neotropical está siendo desestabilizada por la mayor tolerancia de los pumas a las presiones humanas. Se muestra el efecto en cascada de las distintas presiones sobre la comunidad de carnívoros y en última instancia sobre el ecosistema completo. En el capítulo cuarto nos aproximamos al conflicto ganadero a través de una perspectiva socio-ecológica. A través de entrevistas a ganaderos de la Chiquitanía y del Pantanal bolivianos estudiamos la magnitud del conflicto y la percepción local hacia los felinos en estas dos ecoregiones. Encontramos que el conflicto ganadero está ampliamente extendido, aunque las pérdidas ganaderas se encuentran dentro de los límites de depredación de ganado establecidos para otros grandes carnívoros. La percepción local de estas especies es generalmente negativa y su persecución está generalizada afectando directamente al declive de sus poblaciones. En el quinto capítulo nos aproximamos a la emergente amenaza que supone el uso de felinos como sustitutos del tigre (Panthera tigris). A través de una revisión bibliográfica explicamos cómo la demanda de productos derivados del tigre, usados en medicina tradicional asiática, supone una amenaza a las poblaciones de grandes felinos de todo el mundo, incluyendo a los felinos neotropicales que habitan a miles de kilómetros. Los resultados de esta tesis nos permiten comprender mejor las amenazas a las que se encuentran sometidas las especies de felinos neotropicales, en muchos casos aplicable a la generalidad de grandes carnívoros. Destacamos el profundo efecto que la ganadería extensiva ejerce sobre sus poblaciones y recalcamos la importancia de la coexistencia con el ser humano para la conservación de los grandes carnívoros, que son piezas clave en el mantenimiento del equilibrio del ecosistema. 9 Chapter 1 a gran escala ya que se distribuyen (aún) por la mayor parte del Neotrópico habitando hábitats diversos. Como ocurre con la mayor parte de los grandes carnívoros, existe una polarización en la percepción social hacia los grandes felinos. Es común encontrar percepciones negativas hacia estas especies en las áreas rurales donde la coexistencia diaria con el hombre genera tensiones continuas. Sin embargo, existe una clara tendencia de aceptación en las ciudades, donde sus habitantes son ajenos a las realidades rurales. Esta polarización en la opinión de distintos sectores poblacionales permite una interesante integración de las ciencias sociales y la ecología, necesaria para obtener una visión más completa a la hora de implementar medidas efectivas para la coexistencia entre la fauna y el hombre. Por este motivo, estas especies icónicas sirven de modelo para involucrar a diferentes gobiernos, organismos internacionales y ONGs en la elaboración de planes de conservación a gran escala, necesarios para el mantenimiento de las especies y los ecosistemas a largo plazo (Sanderson et al. 2002 b). Perspectiva prehistórica e histórica de las amenazas de los grandes felinos Neotropicales Tanto jaguares como pumas tienen amplias distribuciones y han sufrido cambios en sus rangos de distribución mediados por la acción conjunta de los cambios paleoclimáticos y las presiones antrópicas (Barnoski et al. 2004). La distribución actual del jaguar se extiende desde el norte de México hasta la provincia de Corrientes al norte de Argentina. El puma tiene una distribución más amplia comprendiendo prácticamente todo el continente americano, desde Canadá hasta el sur de Chile (IUCN 2017). Ambas especies, con historias filogenéticas similares, han coexistido desde el Pleistoceno temprano a lo largo del continente americano durante (probablemente) más de un millón de años hasta la llegada al nuevo mundo del Homo sapiens en el Pleistoceno tardío. Existe cierta controversia sobre la llegada del H. sapiens a América (Cooper et al. 2019, Manning 2020, Cooper et al. 2020) pero parece claro que su llegada ocurrió no antes de 30.000 años a través de Beringia, durante la glaciación de Wisconsin (o 16 General introduction para los europeos Würm) (Goebel et al. 2008). Hace tan solo 16.500 años se abrió el pasillo del Pacífico permitiendo la migración hacia el Sur y el acceso al resto del continente americano (Lesnek et al. 2018, Batchelor et al. 2019). A partir de este momento el H. sapiens, que había permanecido retenido en el refugio glacial norteamericano, pudo avanzar hacia el sur colonizando el continente, comenzando la (pre)histórica coexistencia con los felinos americanos. Coincidiendo con el proceso de expansión postglacial, ocurrió el evento de Extinción de Megafauna del Cuaternario, una extinción masiva que acabó con el 80 % de las especies de megafauna mundial (Grayson & Meltzler, 2003, Wroe et al. 2004, Johnson, 2009). El origen de esta extinción masiva ha sido foco de un histórico debate para elucidar si la extinción fue mediada por los cambios climáticos o por la acción directa del hombre. Las evidencias apuntan a que, más que un factor único, fue el efecto aditivo de ambos factores la causa de este suceso (Barnoski et al. 2004); sin embargo, el papel de cada uno de ellos parece haber sido diferente en cada continente. Parece claro que en el norte de Eurasia los eventos de extinción de megafauna coincidieron con los pulsos climáticos, mientras que en Norteamérica fue el avance de la colonización del Homo sapiens (las culturas Clovis de cazadores-recolectores) la que se ha relacionado robustamente con la extinción de megafauna (Becerra-Valdivia & Highman 2019). En lo que respecta a los felinos, tanto las evidencias fósiles (Seymour 1989, Rodriguez et al. 2018) como las genéticas (Eizirik et al. 2001, Culver & PecorSlattery 2000) sugieren que los pumas y jaguares también fueron eliminados de Centro y Norteamérica coincidiendo con el paso de esta nueva especie de depredador top, y que posteriormente ocurrió la recolonización por individuos procedentes de poblaciones suramericanas. Estos datos sugieren que, en las etapas primigenias de la coexistencia los grandes felinos americanos ya entraban en conflicto con los humanos y que la combinación hombre-clima supone un tándem sumamente penetrante para la extinción de las especies, como muestran las diversas especies de felinos extintas durante dicho periodo expansivo (i.e Smilodon o P. onca augusta) y de las cuales sólo persistieron los actuales jaguares y pumas. A pesar de este primer encuentro exterminador para las especies de felino, las culturas paleoindias - evolucionadas a partir de la cultura Clovis - se caracterizaron por 17 Chapter 1 tener una visión homofelina, entendida como la capacidad de proyección en la imagen y virtudes propias del felino (algo así como un alter ego). Esta visión idólatra se puede encontrar por todos los rincones y culturas indígenas, como muestran las artes rupestres de Chiribiquete (Colombia), el templo al dios jaguar en Tikal (Guatemala), en las cabezas deformadas con forma felina por los Olmecas (Perry 1970), en la creencia de los pueblos Guaraníes de que el jaguar era el causante de los eclipses (Cadogan 1973) o los innumerables casos de emulación y proyección cultural en el imaginario histórico y mítico que encontramos hasta nuestros días (Castaño-Uribe 2017). Probablemente esta visión integradora, culturalmente más evolucionada que la de los pueblos Clovis, pudo propiciar la coexistencia y recolonización de Centro y Norteamérica por los grandes felinos, durante el Pleistoceno tardío, a partir de las poblaciones suramericanas (Eizirik 2001, Culver & Pecor-Slattery 2000). La llegada del hombre blanco al nuevo continente supuso una transición en cómo los humanos percibían a estas especies. Gradualmente se diluyó la proyección de los felinos como alter ego a través de la satanización de su figura, simbolizando una idolatría propia de de los herejes (Reichel-Dolmatoff 1978). Esta visión felino-fóbica se fue consolidando a medida que avanzaba por el continente la conversión de los pueblos indígenas al cristianismo que, sin duda, marcó un importante hito para la actual percepción hacia los grandes felinos americanos. Estas dos percepciones conceptuales (homofelina vs felino-fóbica) son el origen, o al menos un reflejo, de las diferentes percepciones sociales que encontramos en la actualidad. A partir del siglo XVI nació una estrecha relación entre indígenas y colonos a través del comercio de pieles. En un principio los indios americanos comerciaban con pieles de pequeños animales para conseguir las herramientas de acero del hombre moderno. Este trueque primigenio fue sustituido paulatinamente por una verdadera industria internacional que abastecía el mercado de moda norteamericano y europeo en el que las pieles de felinos eran muy demandadas (Carlos & Lewis 2015). Esta industria tuvo su auge durante los años 60 del siglo XX, afectando gran cantidad de especies de mamíferos y reptiles incluyendo a los grandes felinos. Se estima que unos 15.000 jaguares y 80.000 ocelotes fueron cazados cada año para el comercio peletero durante la década de los años 60 (Smith, 1976) y se incrementó, durante tan solo una década, en 18 General introduction un orden de magnitud la exportación de grandes felinos (jaguares y pumas) desde Brasil hacia Estados Unidos y Europa (Dowghty & Myers 1971). Durante este corto período de tiempo la industria peletera fue la mayor amenaza para las poblaciones de grandes felinos, especialmente los gatos pintados, como jaguares, ocelotes y margays, que vieron reducidas sus poblaciones drásticamente (Swank & Teer 1987). Con la aparición del convenio CITES se prohibió el comercio internacional de muchas especies silvestres, entre ellas el jaguar, incluido en el Apéndice I en el año 1973. A pesar de que el comercio se mantuvo durante los años 70, la industria peletera prácticamente desapareció a principio de los años 80 y, a través de la reducción de la demanda, los precios de las pieles disminuyeron drásticamente, haciendo insostenible el negocio de peletería de felinos salvajes (Swank & Teer 1987). Principales amenazas actuales para la conservación de los felinos Desde entonces y hasta hoy las principales amenazas de los grandes felinos neotropicales se han relacionado con la pérdida de hábitat y la persecución directa (Quigley et al. 2017, Nielsen et al. 2015). Es abrumadora la reciente documentación sobre el declive poblacional - especialmente sobre el jaguarpor la pérdida de hábitat 19 Número de exportaciones 0 12500 25000 37500 50000 1957 1960 1963 1966 1969 1976 Exportaciones de Panthera onca y probablemente Puma concolor desde Brasil a US, UK, Alemania e Italia. Adaptación de Dowgly & Myers (1971) & Smith (1976) Chapter 1 (por ejemplo, Ceballos et al. 2011, Espinosa et al. 2016, Hoogesteijn et al. 2016, Di Bitetti et al. 2016, Paviolo et al. 2016, Jedrzejewski et al. 2016). La cacería actual de felinos ocurre por distintos motivos, y está generalizada a lo largo del Neotrópico (Medellín et al. 2017). Es común que el conflicto que supone la depredación de ganado promueva su cacería (Castaño-Uribe et al. 2017). Durante el pasado siglo la ganadería bovina ha crecido de manera muy acusada en Latinoamérica, dominando el paisaje neotropical (Gilbert et al. 2018, Robinson et al. 2014). Ésta ha modificado ya biomas completos y afecta a los ecosistemas en los que se establece (véase el caso del cerrado brasileño, modificado por la combinación de la ganadería y los sistemas agrícolas intensivos). Actualmente el 27% de la producción mundial de carne es producida en Latinoamérica (FAO, 2020), aunque su superficie abarque poco más del 12 % de la superficie terrestre, dando una idea de la intensidad de esta industria en el Neotrópico. La ganadería supone quizás la mayor amenaza actual a los grandes felinos neotropicales a través del enraizado conflicto que emerge por la depredación de ganado. Conocer en profundidad el conflicto ganadero es uno de los objetivos transversales de esta tesis. Por otro lado, además del conflicto ganadero existen otros motivos para la persecución y cacería de los grandes felinos neotropicales. Existe una creciente demanda de productos derivados del tigre por parte de países asiáticos cuyo efecto ya ha salpicado al resto de grandes felinos del mundo. Las poblaciones de tigre asiático Panthera tigris no han podido soportar la presión de cacería de las últimas décadas hasta tal punto que existen tan solo unos 3000 individuos salvajes en libertad (Goodrich et al. 2015). Sin embargo la gran demanda de productos derivados del tigre ha promovido el millonario negocio de la cría de tigres en cautividad (EIA 2017) y ya ha afectado a las poblaciones de leopardos (P. uncia, P. pardus, N.nebulosa) y leones (P. leo) a través de su uso como sustitutos (Maheshwari & Niraj 2018, Nowell & Pervushina 2015, Williams et al. 2017). Parece que este emergente foco de persecución puede afectar a las poblaciones de otros felinos, entre ellos los neotropicales, como ya lo ha hecho a las de tigre. El cambio climático es otra de las amenazas actuales. Afecta a buena parte de las especies animales del mundo (Thomas 2010) y sin duda puede tener un importante 20 General introduction efecto sobre las especies de grandes felinos. Como ya se puso de manifiesto durante las oscilaciones climáticas del Pleistoceno, los cambios climáticos pueden suponer grandes eventos de extinción especialmente si se presentan en combinación con otras presiones. Dada la tendencia humana a repetir el pasado (Freud 1914) no es descabellado imaginar que en la actualidad nos encontremos en un escenario paralelo a la Extinción de Megafauna del Cuaternario, donde - de nuevo - un cambio climático combinado con la presión de cacería amenazan a las especies de gran tamaño. De mantenerse en la línea actual, el efecto del rápido cambio climático (esta vez con una eminente componente antrópica) en combinación con la persecución y caza de los grandes felinos (motivados por la ganadería y el tráfico ilegal) con certeza desembocará en una nueva extinción. Esta vez, el nuevo evento de extinción que se ha denominado la Sexta Extinción Masiva (Wake & Vredenburg 2008) tiene un indiscutible origen antropogénico (Estes et al. 2011). El objetivo de esta tesis es profundizar en el conocimiento de los factores que amenazan la conservación de los felinos neotropicales para poder ofrecer información fidedigna que pueda ser utilizada como base de políticas ambientales asistiendo su conservación a largo plazo. He hecho un especial esfuerzo para comprender las relaciones de los grandes felinos con la ganadería tanto desde un punto de vista 21 5000 10000 15000 20000 1985 1993 2000 2008 2012 2015 2018 Tigres salvajes Tigres cautivos Número de tigres salvajes y tigres cautivos. Fuentes: CITES SC70 Doc. 51 Annex 2 (Rev. 1) & EIA 2017. Cultivating demand. The growing threat of tiger farms. Chapter 1 ecológico como social, ambos necesarios para ofrecer soluciones realistas y factibles ante el preocupante problema. Cada capítulo tiene una aproximación metodológica diferente -análisis genéticos, entrevistas estructuradas, modelado ecológico o revisión bibliográficapermitiendo aproximamos a las amenazas de los felinos a nivel local y continental, esperando contribuir con una visión lo más completa posible sobre cómo las actividades humanas afectan a estas especies. Las distintas aproximaciones metodológicas, escalas y contextos permiten profundizar en los efectos que las perturbaciones humanas ejercen en los felinos. Por un lado, investigando las principales amenazas a escala global para entender el efecto e implicaciones macroecológicas del ser humano sobre los grandes depredadores ofreciendo, al mismo tiempo, una imagen de su estado de conservación actual. Por otro lado, profundizando en el escenario de coexistencia entre hombre y felinos a través de una visión socio-ecológica del conflicto ganadero que, como veremos, resulta ser la mayor amenaza actual de estas especies. Por último, ofrezco una revisión sobre la amenaza emergente que supone el tráfico de felinos hacia Asia y sus potenciales efectos en las poblaciones de felinos. 22 General introduction Referencias Abrahms, B. (2021). Human-wildlife conflict under climate change. Science, 373(6554), 484–485. doi: 10.1126/science.abj4216 Arias, M., Hinsley, A., Nogales-Ascarrunz, P., Negroes, N., Glikman, J. A., & MilnerGulland, E. J. (2021). Prevalence and characteristics of illegal jaguar trade in northwestern Bolivia. Conservation Science and Practice, 3(7). doi: 10.1111/csp2.444 Barnosky, A. D., Koch, P. L., Feranec, R. S., Wing, S. L., & Shabel, A. B. (2004). Assessing the causes of late pleistocene extinctions on the continents. Science, 306(5693), 70–75. doi: 10.1126/science.1101476 Becerra-Valdivia, L., & Higham, T. (2020). The timing and effect of the earliest human arrivals in North America. Nature, 584(7819), 93–97. doi: 10.1038/s41586-020-2491-6 Caragiulo, A., Dias-Freedman, I., Clark, J. A., Rabinowitz, S., & Amato, G. (2014). Mitochondrial DNA sequence variation and phylogeography of Neotropic pumas (Puma concolor). Mitochondrial DNA, 25(4), 304– 312. doi: 10.3109/19401736.2013.800486 Cardillo, M., Purvis, A., Sechrest, W., Gittleman, J. L., Bielby, J., & Mace, G. M. (2004). Human population density and extinction risk in the world’s carnivores. PLoS Biology, 2(7), E197. doi: 10.1371/journal.pbio.0020197 Carlos, A. M., & Lewis, F. D. (2002). Marketing in the Land of Hudson Bay: Indian Consumers and the Hudson’s Bay Company, 1670–1770. Enterprise and Society, 3(2), 285–317. doi: DOI: 10.1093/es/3.2.285 Casas-Marce, M., Marmesat, E., Soriano, L., Martínez-Cruz, B., Lucena-Perez, M., Nocete, F., … Godoy, J. A. (2017). Spatiotemporal dynamics of genetic variation in the iberian lynx along its path to extinction reconstructed with ancient DNA. Molecular Biology and Evolution, 34(11), 2893–2907. doi: 10.1093/ molbev/msx222 Castaño-Uribe, C., C. A. Lasso, R. Hoogesteijn, A. D.-P. y E. P. (2016). Conflictos entre felinos y humanos en America Latina. Crooks, K. R., & Soulé, M. E. (1999). Mesopredator release and avifaunal extinctions in a fragmented system. Nature, 400(6744), 563–566. doi: 10.1038/23028 Culver, M., Johnson, W. E., Pecon-Slattery, J., & O’Brien, S. J. (2000). Genomic ancestry of the American puma (Puma concolor). Journal of Heredity, 91(3), 186–197. doi: 10.1093/jhered/ 91.3.186 Cunningham, C. X., Johnson, C. N., Barmuta, L. A., Hollings, T., Woehler, E. J., & Jones, M. E. (2018). Top carnivore decline has cascading effects on scavengers and carrion persistence. Proceedings of the Royal Society B: Biological Sciences, 285(1892). doi: 10.1098/ rspb.2018.1582 Davis, L. G., Madsen, D. B., Becerra-valdivia, L., Higham, T., Sisson, D. A., Skinner, S. M., … Buvit, I. (2019). R ES E A RC H Late Upper Paleolithic occupation. 897(December), 891– 897. Davis, L. G., Madsen, D. B., Becerra-Valdivia, L., Higham, T., Sisson, D. A., Skinner, S. M., … Buvit, I. (2019). Late Upper Paleolithic occupation at Cooper’s Ferry, Idaho, USA, ~16,000 years ago. Science, 365(6456), 891– 897. doi: 10.1126/science.aax9830 Doughty, R. W., & Myers, N. (1971). Notes on the Amazon Wildlife Trade 1969 Compared Caititfi ( Tayassu tajacu ) Queixada ( T . pecari ) Capybara ( Hydrochoerus hydrochaeris ) Onga ( Panthera onca ). Biological Conservation, 3(July), 293–297. EIA. (2017). Cultivating Demand - The Growing Threat of Tiger Farms. In EIA International. Retrieved from https://eia-international.org/ report/cultivating-demand-growing-threattiger-farms Eizirik, E., Kim, J. H., Menotti-Raymond, M., Crawshaw, P. G., O’Brien, S. J., & Johnson, W. E. (2001). Phylogeography, population history and conservation genetics of jaguars (Panthera onca, Mammalia, Felidae). Molecular Ecology, 10(1), 65–79. doi: 10.1046/j.1365-294X.2001.01144.x Eren, M. I., Patten, R. J., O’Brien, M. J., & Meltzer, D. J. (2013). Refuting the technological cornerstone of the Ice-Age Atlantic crossing hypothesis. Journal of Archaeological Science, 40(7), 2934–2941. doi: 10.1016/j.jas.2013.02.031 Estes, J. A., Terborgh, J., Brashares, J. S., Power, M. E., Berger, J., Bond, W. J., … Wardle, D. A. (2011). Trophic downgrading of planet 23 Chapter 1 earth. Science, 333(6040), 301–306. doi: 10.1126/science.1205106 FAO. (2009). La agricultura mundial en la perspectiva del año 2050. Fao, 4. Retrieved from http://www.fao.org/fileadmin/templates/ wsfs/docs/I Figel, J. J., Botero-Cañola, S., Sánchez-Londoño, J. D., & Racero-Casarrubia, J. (2021). Jaguars and pumas exhibit distinct spatiotemporal responses to human disturbances in Colombia’s most imperiled ecoregion. Journal of Mammalogy, 102(1), 333–345. doi: 10.1093/jmammal/gyaa146 Foster, R. J. (2008). The ecology of jaguars (Panthera onca) in a human-influenced landscape. 345 pp. Retrieved from http:// eprints.soton.ac.uk/ 66711/1.hasCoversheetVersion/ RFoster_2008_PhD_thesis.pdf%5Cnhttp:// eprints.soton.ac.uk/66711/1/ RFoster_2008_PhD_thesis.pdf Freud, S. (1914). Recordar, repetir y reelaborar en Obras completas, tomo XII (B. A. A. Ediciones., ed.). Geldmann, J., Barnes, M., Coad, L., Craigie, I. D., Hockings, M., & Burgess, N. D. (2013). Effectiveness of terrestrial protected areas in reducing habitat loss and population declines. Biological Conservation, 161, 230–238. doi: 10.1016/j.biocon.2013.02.018 Goebel, T., Waters, M. R., & O’Rourke, D. H. (2008). The Late Pleistocene dispersal of modern humans in the Americas. Science, 319(5869), 1497–1502. doi: 10.1126/ science.1153569 Gray, C. L., Hill, S. L. L., Newbold, T., Hudson, L. N., Börger, L., Contu, S., … Scharlemann, J. P. W. (2016). Local biodiversity is higher inside than outside terrestrial protected areas worldwide. Nature Communications, 7(1), 12306. doi: 10.1038/ncomms12306 Grayson, D. K., & Meltzer, D. J. (2003). A requiem for North American overkill. Journal of Archaeological Science, 30(5), 585–593. doi: https://doi.org/10.1016/ S0305-4403(02)00205-4 Hoeks, S., Huijbregts, M. A. J., Busana, M., Harfoot, M. B. J., Svenning, J. C., & Santini, L. (2020). Mechanistic insights into the role of large carnivores for ecosystem structure and functioning. Ecography, 43(12), 1752–1763. doi: 10.1111/ecog.05191 IPCC. (2021). El cambio climático es generalizado, rápido y se está intensificando. Grupo Intergubernamental de Expertos Sobre El Cambio Climático (IPCC). Retrieved from https://www.ipcc.ch/site/assets/uploads/ 2021/08/IPCC_WGI-AR6-Press-ReleaseFinal_es.pdf Johnson, C. N. (2009). Ecological consequences of Late Quaternary extinctions of megafauna. Proceedings of the Royal Society B: Biological Sciences, 276(1667), 2509–2519. doi: 10.1098/rspb.2008.1921 Jones, K. R., Venter, O., Fuller, R. A., Allan, J. R., Maxwell, S. L., Negret, P. J., & Watson, J. E. M. (2018). One-third of global protected land is under intense human pressure. Science, 360(6390), 788–791. doi: 10.1126/ science.aap9565 Jordán, F. (2009). Keystone species and food webs. Phil. Trans. R. Soc. B3641733–1741 Lesnek, A. J., Briner, J. P., Lindqvist, C., Baichtal, J. F., & Heaton, T. H. (2018). Deglaciation of the Pacific coastal corridor directly preceded the human colonization of the Americas. Science Advances, 4(5), eaar5040. doi: 10.1126/sciadv.aar5040 Maheshwari, A., & Niraj, S. K. (2018). Monitoring illegal trade in snow leopards: 2003–2014. Global Ecology and Conservation, 14, e00387. doi: 10.1016/j.gecco.2018.e00387 Manning, S. W. (2020). Comment on “late Upper Paleolithic occupation at Cooper’s Ferry, Idaho, USA, ∼16,000 years ago.” Science, 368(6487), 1–5. doi: 10.1126/science.aaz4695 Marshall, L. G. (1982). the Great Mammals American Interchange The emergence of the. 76(4), 380–388. Mazak, V. (1981). Panthera tigris. Mammalian Species, 8235(152), 1. doi: 10.2307/3504004 Morcatty, T. Q., Bausch Macedo, J. C., Nekaris, K. A. I., Ni, Q., Durigan, C. C., Svensson, M. S., & Nijman, V. (2020). Illegal trade in wild cats and its link to Chinese-led development in Central and South America. Conservation Biology, 34(6), 1525–1535. doi: 10.1111/ cobi.13498 Myers, N. (1973). The spotted cats and the fur trade. In The World’s Cats: Ecology and Conservation (Vol. 1.; ed. R. L. Eaton, ed.). World Wildlife Safari, Winston, OR, USA. Nowell, K., & Pervushina, N. (n.d.). Review of implementation of resolution conf. 12.5 (rev. cop16) on Conservation of and trade in tiger and other (Vol. 5). Palomares, F., Gaona, P., Ferreras, P., & Delibes, M. (1995). Positive Effects on Game Species 24 General introduction of Top Predators by Controlling Smaller Predator Populations: An Example with Lynx, Mongooses, and Rabbits. Conservation Biology, 9(2), 295–305. doi: 10.1046/ j.1523-1739.1995.9020295.x Palomares, F., Fernández, N., Roques, S., Chávez, C., Silveira, L., Keller, C., & Adrados, B. (2016). Fine-Scale Habitat Segregation between Two Ecologically Similar Top Predators. PLOS ONE, 11(5), e0155626. Retrieved from https://doi.org/10.1371/ journal.pone.0155626 Production, F., Cervantes-Godoy, D., Dewbre, J., PIN, Amegnaglo, C. J., Soglo, Y. Y., … Swanson, B. E. FAO. (2017). The future of food and agriculture: trends and challenges. In The future of food and agriculture: trends and challenges (Vol. 4). Prugh, L. R., Stoner, C. J., Epps, C. W., Bean, W. T., Ripple, W. J., Laliberte, A. S., & Brashares, J. S. (2009). The rise of the mesopredator. BioScience, 59(9), 779–791. doi: 10.1525/ bio.2009.59.9.9 Radwan, T. M., Blackburn, G. A., Whyatt, J. D., & Atkinson, P. M. (2021). Global land cover trajectories and transitions. Scientific Reports, 11(1), 1–16. doi: 10.1038/s41598-021-92256-2 Raff, J. A., & Bolnick, D. A. (2015). Does Mitochondrial Haplogroup X Indicate Ancient Trans-Atlantic Migration to the Americas? A Critical Re-Evaluation. PaleoAmerica, 1(4), 297–304. doi: 10.1179/2055556315Z.00000000040 Reichel-Dolmatoff, G. (1978). El chamán y el jaguar : estudio de las drogas narcóticas entre los indios de Colombia ([1a ed].). Madrid: Madrid : Siglo XXI. Ripple, W. J., & Beschta, R. L. (2006). Linking a cougar decline, trophic cascade, and catastrophic regime shift in Zion National Park. Biological Conservation, 133(4), 397– 408. doi: 10.1016/j.biocon.2006.07.002 Robinson, T. P., Wint, G. R. W., Conchedda, G., Van Boeckel, T. P., Ercoli, V., Palamara, E., … Gilbert, M. (2014). Mapping the Global Distribution of Livestock. PLOS ONE, 9(5), e96084. Retrieved from https://doi.org/ 10.1371/journal.pone.0096084 Rodriguez, S. G., Méndez, C., Soibelzon, E., Soibelzon, L. H., Contreras, S., Friedrichs, J., … Zurita, A. E. (2018). Panthera onca (Carnivora, Felidae) in the late Pleistoceneearly Holocene of northern Argentina. Neues Jahrbuch Fur Geologie Und Palaontologie - Abhandlungen, 289(2), 177–187. doi: 10.1127/ njgpa/2018/0758 Samojlik, T., Selva, N., Daszkiewicz, P., Fedotova, A., Wajrak, A., & Kuijper, D. P. J. (2018). Lessons from Białowieża Forest on the history of protection and the world’s first reintroduction of a large carnivore. Conservation Biology, 32(4), 808–816. doi: 10.1111/cobi.13088 Sanderson, E. W., Jaiteh, M., Levy, M. A., Redford, K. H., Wannebo, A. V., & Woolmer, G. (2002). The human footprint and the last of the wild. BioScience, 52(10), 891–904. doi: 10.1641/0006-3568(2002)052[0891:THFATL] 2.0.CO;2 Sarmento, P., Carrapato, C., Eira, C., & Silva, J. P. (2019). Spatial organization and social relations in a reintroduced population of Endangered Iberian lynx Lynx pardinus. Oryx, 53(2), 344–355. doi: 10.1017/ S0030605317000370 Seymour, K. L. (1989). Panthera onca. Mammalian Species, (340), 1–9. doi: 10.2307/3504096 Smith, N. J. H. (1978). Human exploitation of terra firme fauna in Amazonia Ciencia e Cultura 1978 Nigel Smith.pdf. Ciência e Cultura, Vol. 30, pp. 17–23. Smith, N. J. H. (1976). Spotted Cats and the Amazon Skin Trade. Oryx, 13(4), 362–371. doi: 10.1017/S0030605300014095 Swank, W. G., & Teer, J. G. (1987). Status of the Jaguar, 1987. Oryx, 23(January), 14–21. R e t r i e v e d f r o m Swank_&_Teer_1987_Status_of_the_Jaguar.p df Thomas, C. D. (2010). Climate, climate change and range boundaries. Diversity and Distributions, 16(3), 488–495. doi: 10.1111/ j.1472-4642.2010.00642.x Turner A, A. M. (1997). The Big Cats and Their Fossil Relatives. Columbia University Press, New York. Wake, D. B., & Vredenburg, V. T. (2009). Are we in the midst of the sixth mass extinction? A view from the world of amphibians. In the Light of Evolution, 2, 27–44. doi: 10.17226/12501 Williams, V. L., Loveridge, A. J., Newton, D. J., & Macdonald, D. W. (2017). A roaring trade? The legal trade in Panthera leo bones from Africa to East-Southeast Asia. PLoS ONE, 12(10), 1–22. doi: 10.1371/ journal.pone.0185996 25 Chapter 2 INTRODUCTION The distribution and spatial arrangement of suitable niches determine the locations, sizes, and connectivity of populations (Guisan & Thuiller 2005; Soberón & Nakamura 2009), which in turn regulate gene flow and drive genetic patterns from the local to the continental scales (Barrientos et al. 2014; Loveless et al. 2016; Carvalho et al. 2017; Inoue & Berg 2017; Mairal et al. 2017; Noguerales et al. 2017; Koch et al. 2018). Changes in the climate have led to range shifts and the alteration of niche suitability across species distributions, making population connectivity transient over space and time (Aguilée et al. 2009; Licona-Vera et al. 2018; Koch et al. 2018). Consequently, gene flow has occurred in different directions and intensity throughout population history, leading to climatic landscape signatures in the current genetic structure of species (Pauls et al. 2013; Habel et al. 2015). Jaguar [Panthera onca (Linnaeus, 1758)] and puma [Puma concolor (Linnaeus, 1771)] are large-bodied and widely distributed felids that likely underwent range shifts mediated by paleoclimatic changes. The current jaguar distribution extends from the southwestern United States to Rio Negro province in northern Argentinean Patagonia (IUCN 2017). However, its late Pleistocene distribution extended somewhat further, from the southern half of the United States to Chilean Patagonia (Rodriguez et al. 2018). The puma has a wider current distribution, ranging from Canada to southern Chile (IUCN 2017), though its late Pleistocene distribution encompassed all of North America (Kurtén & Anderson 1980) and reached southwards to the Mar del Plata region in central-eastern Argentina (Chimento & Dondas 2018). Despite advances in molecular techniques, including non-invasive methods that have been used effectively for jaguar and puma (Roques et al. 2014; Palomares et al. 2016, 2017; Zanin et al. 2016), investigations of the genetic diversity and structure of both species have largely been limited to the local or regional scales (Ernest et al. 2003; Haag et al. 2010; Miotto et al. 2011; Pflüger & Balkenhol 2014; Balkenhol et al. 2014; Salzano et al. 2015; Srbek-Araujo et al. 2018). Some of these studies have shown that both species are susceptible to genetic isolation at the landscape scale due to habitat 32 Climatic landscapes loss and fragmentation (Haag et al. 2010; Miotto et al. 2011; Balkenhol et al. 2014; Srbek-Araujo et al. 2018). However, the patterns of genetic isolation of both species at the continental scale and the possible effects of climate changes on their genetic structure are not well understood. Continental studies on jaguar and puma have focused on niche prediction and its relationship to ecological patterns, which shows the effectiveness of climatic variables as predictors of felid niches (Tôrres et al. 2012; MartínezMeyer et al. 2013; Caruso et al. 2015; Silva et al. 2017; Zanin & Neves 2019). This study aims to estimate the effect of climate changes on the jaguar and puma, assessing whether the modern and paleoclimatic landscape influenced the modern isolation patterns of individuals and the genetic population structure across the Neotropics. Our main hypothesis was that current genetic patterns are an amalgamation of spatial and temporal changes in the distribution of suitable areas and connection routes for the species, which in turn have been driven by climatic landscapes. We expected a stronger genetic signature from climatic landscapes for jaguars than for pumas, at both individual and population levels, since pumas have a higher tolerance to climatic variation (Zanin & Neves 2019) that enables a more stable distribution over space and time. MATERIAL AND METHODS Genetic data sampling, DNA extraction, and individual identification Surveys for fecal samples were conducted in 21 localities within the historical sympatric distribution of the species. Jaguar samples were found in 15 of these localities and puma samples in 20 (Fig. 1 and Table 1). Fecal samples were collected between 2000 and 2017 by active search, by experienced researchers and field technicians, along dirt roads and trails. Samples were stored in 100 mL plastic containers with silica gel, and the sites were georeferenced by GPS. DNA was extracted from the fecal samples using protocols based on the GuSCN/silica method (Boom et al. 1990; Frantz et al. 2003), and then purified and concentrated by ultra-filtration using Microcon-30 (Millipore). Each batch of extractions (n = 12–15) 33 Chapter 2 included 1 negative PCR control to monitor contamination by exogenous DNA. DNA extractions of fecal samples were performed in a UV-sterilized laminar flow hood in an isolated laboratory that was specially designated for the manipulation of non-invasively sourced biological material. Fecal samples were screened for species identity using species-specific primers (Roques et al. 2011). Individual genotyping was conducted using an optimized set of 11 microsatellite markers for jaguar and 12 microsatellite markers for puma, following the protocols described in Roques et al. (2014) and Zanin et al. (20 16 ), res pe ct iv el y. A consensus genotype was constructed for each sample after 4 repetitions. Homozygous loci were assigned when the same allele was observed in at least 3 replicates as long as a different allele was not observed in the fourth. Heterozygous loci were those with 2 different alleles in at least 2 replicates. Climatic landscapes Climatic landscapes were generated by ecological niche modeling, for which we adopted 4 algorithms that use a correlative approach between 34 Figure 1 Sampling localities of feces distributed across the Neotropics for jaguar (N = 15) and puma (N = 20). In dark gray, species distribution extends according to the IUCN red list (IUCN 2017). Sampling localities were aggregated in demes to population-level analysis. Table 1 provides the names of sampling sites and demes. Climatic landscapes species occurrence data and climatic variables. The selected algorithms were as follows: generalized linear models—GLM (McCullagh & Nelder 1989), multivariate adaptive regression splines—MARS (Friedman 1991), maximum entropy modeling— MaxEnt (Phillips et al. 2006), and random forest (Breiman 2001). The data on species occurrence is from a recently published database of Felidae occurrences (Zanin & Neves 2019), comprised of records sampled between 1990 and 2016. The database is highly effective for macroecological studies due to the random point-pattern of occurrences and climatic niche predictability at a spatial resolution of 0.5 decimal degrees. Therefore, our modeling approach used the 214 spatially unique occurrences for jaguar and the 538 for puma available in the database (Zanin & Neves 2019). This study thus used a spatial resolution of 0.5 decimal degrees for consistency with the occurrence data. The current literature warns about the reconstruction of species paleoclimatic distribution by using only the current occurrences of species, which could underestimate the niche breadth and compromise the niche transferability to other temporal scenarios (MartínezFreiría et al. 2016; Faurby & Araújo 2018). Here, we opted against including species occurrences from historical distributions for several reasons. Firstly, the adopted dataset covers a wide variety of ecosystems, so our modeling should capture a broad range of climatic niches. Moreover, most recent contractions in the range of jaguar and puma in the Neotropics have occurred locally across different ecosystems. Thus, our estimation should generate an effective approximation of the species’ climatic niches since sampled occurrences cover areas environmentally similar to those in which the species has become extinct (similar to the study on elephants by Martínez-Freiría et al. 2016). Lastly, the addition of other occurrences can compromise the verified reliability and effectiveness of the dataset by inserting spatial bias, since historical occurrence registers can be less precise than recent ones (Matthews & Heath 2008; Tingley & Beissinger 2009). All of the selected algorithms employ pseudo-absences in the modeling procedure (Fielding & Bell 1997). We used a random subset of occurrences of other felids of the 35 Chapter 2 database on Felidae occurrences as pseudo-absences, which were spatially restricted to the Neotropics. We assumed that the observer (or equipment) responsible for a record would be able to determine the presence of other felid species at the same location and time of observation based on the taxonomical similarity and sampling methods used. We, therefore, considered that these pseudo-absences are a closer representation of real absence. The climatic data comprised 19 temperature and precipitation variables that formed the following scenarios: last maximum glaciation (LMG), Holocene, and modern (1950– 1999). The climatic data were generated by the Community Climate System Model (CCSM) and obtained from the Ecoclimate Project (Lima-Ribeiro 2015; Lima-Ribeiro et al. 2018). 36 Table 1 DNA amplification success for samples of jaguars and pumas. NS, number of samples; NI, number of individuals identified; QI, quality index, calculated by amplification in 4 PCRs (mean ± SD); NL, number of loci (mean ± SD) Climatic landscapes In order to reduce the number of climatic variables, we began by performing niche modeling though 25 repetitions, in which the occurrences were randomly split at 75% and 25% of the amount of data to be used for model calibration and validation, respectively. We selected 7 variables that best described the climatic niche of the species and repeated the ecological niche modeling using 50 repetitions and the same split rules. The predictions were multiplied by true skill statistic (TSS) and averaged through an ensemble forecasting approach (Allouche et al. 2006) to generate the final climatic characterization of the species’ ecological niche in which individual predictions contributed accordingly with their robustness (Araújo & New 2007). After characterizing each species’ climatic niche, we projected it to modern and past scenarios to construct climatic suitability landscapes for the jaguar and puma. Final predictions were evaluated by TSS, area under ROC curve (AUC) analysis, and kappa (Allouche et al. 2006; Peterson et al. 2008). All analyses were conducted in R software (R Core Team 2019) with the biomod2 package (Thuiller et al. 2009). Individual-based causal modeling of climatic and genetic landscapes We used an individual pairwise comparison to test the influence of spatial and climatic distances (or dissimilarities) on genetic dissimilarities for the jaguar and puma. Genetic dissimilarities were calculated using the Kosman Index (Kosman & Leonard 2005). Geographical (Euclidean) distance among individuals was used to test the pure effect of spatial distance or the isolation by distance hypothesis. The pure effect of climate was estimated by differences in climatic suitability through pairwise sites. The functional distances (i.e. the connectivity among individuals) were estimated by the least-cost route (sensu Adriaensen et al. 2003) and resistance surface (sensu McRae 2006), both assuming the climatic niche suitability map as a surface of permeability. The least-cost analysis draws the most efficient path between 2 sites according to the landscape permeability (Adriaensen et al. 2003). The resistance surface, on the other hand, has a more complex approach since it estimates the expected commute time between 2 sites by random walk movement (McRae 2006). The transition rule among 37 Chapter 2 grid cells was the same for both methods, calculated by mean permeability and considering a potential move to any of the 8 grid cells surrounding the focal one. The pure climatic effect and functional distances were calculated for each climatic landscape (modern, Holocene, and LMG), for the average climatic landscape (mean suitability among climatic landscapes), and for climatic instability (coefficient of variation of suitability among climatic landscapes). We thus tested 16 hypotheses of how the environment and species genetics could be correlated, with each hypothesis represented by a different geographical and/or climatic distance matrix. The dissimilarity matrices were calculated using the mmod (Winter 2013), vegan (Oksanen et al. 2019), and gdistance (van Etten 2012) packages in the R software. Hypothesis tests were conducted by causal modeling on resemblance matrix, which uses Mantel and partial Mantel tests to measure the degree of correlation among dissimilarity matrices using a simple relationship or by comparing them with a third matrix, respectively (Legendre & Legendre 1998). The causal modeling framework is effective in identifying paths of gene flow in complex landscapes, allowing the simultaneous test of spatial and non-spatial hypotheses (Cushman & Landguth 2010; Cushman et al. 2013). We tested a single response variable (genetic distance) against individual predictor variables (either climatic or geographical distance) either by a simple relationship or by controlling for the effect of the other predictor variable. All pairwise comparisons among predictor variables were tested, totaling 256 correlations for each species (16 Mantel and 240 partial Mantel tests). The Pearson correlation and Monte Carlo P-values were calculated over 10 000 permutations of the Mantel and partial Mantel tests using the vegan package in R software. Causal modeling can result in reduced Type II errors (Cushman et al. 2013), suggesting a low probability of rejection of spatial or climatic effects when they do affect genetic structure. However, it can also show elevated rates of Type I errors, which lead to acceptance of spurious correlations as true patterns (Cushman & Landguth 2010; 38 Climatic landscapes Cushman et al. 2013). Spurious correlations can be identified by evaluating the tendencies of how each variable works across the entire causal modeling framework. We assumed that an influential predictive variable (true hypothesis) would have a good ability to explain genetic dissimilarities in all models where it appears as a causal variable. However, it should decrease the explanation power of models when inserted as a control variable. To reduce the misinterpretation pattern generated by Type I and II errors in the causal modeling framework, we used linear mixed models to evaluate the conditional means of the Pearson correlation and the Monte Carlo P-value (response variables) as a linear function of the scenarios used as the causal variable (fixed effect) and controlled variable (random effect) (Gotelli & Ellison 2004). The random effect works as a grouping factor, which, in our study, allowed us to calculate shrinkage estimates of regression coefficients (Efron & Morris 1977). The ability of the full model (comprising fixed and random variables) to describe the tendencies of Pearson correlations and Monte Carlo P-values was evaluated by comparing it with the null model (composed only by random variables) through an analysis of variance (ANOVA), in which we adopted a 95% confidence interval. Linear mixed models were calculated with the lme4 package (Bates et al. 2015) in R software. Population structure and diversity The sampling sites were unevenly distributed within the species’ range distribution (Fig. 1 and Table 1). Thus, we aggregated them based on the largest home-range of each species described in the literature to conduct the population-level analyses. These groups or demes were used as a maximum k value for the cluster analysis, minimizing the spatial dependence of sampling sites. The home-range area adopted was 690 km2 for jaguar and 755 km2 for puma (Gonzalez-Borrajo et al. 2017), resulting in 8 demes for each species (Table 1 and Fig. 1). 39 Chapter 2 Population clusters were analyzed with TESS software (Chen et al. 2007; Durand et al. 2009), which employs a spatial Bayesian cluster analysis using Markov Chain Monte Carlo algorithms to identify k populations without a priori group definition. TESS assumes geographical continuity of allele frequencies, which would make neighboring sites more similar than distant sites (François et al. 2006). This approach enables cline and/ or cluster detection, making TESS effective for scenarios with isolation by distance (François & Durand 2010). We ran admixture models using 50 000 iterations after a burn-in period of 500 000 iterations, for k = 2–8 (maximum number of demes), with 20 independent runs for each k. The number of clusters was determined using the deviance information criterion (DIC) (Spiegelhalter et al. 2002). Cluster analysis can be influenced by differences in the distribution of sampling units and gaps in the study area, so we performed a second-level analysis when a cluster was composed of more than one deme. In these cases, we performed TESS with 50 independent runs, for k varying from 2 up to the number of demes. When the first-order cluster was formed by 2 demes, the number of clusters was determined through the graphical distribution of ancestry coefficients. CLUMPP software was used to average the admixture proportions of individuals over the replicates of the 20 most likely k (Jakobsson & Rosenberg 2007). Geographic maps of ancestry coefficients were drawn in the R software by the tess3r package (Caye et al. 2016) to enable a spatial visualization of cluster distribution, following Caye et al. (2016). The strength of genetic differentiation was estimated with Jost’s D, which is more appropriate than FST to compare populations with contrasting levels of genetic diversity (Jost 2008; Whitlock 2011; Assis et al. 2018). We evaluated genetic diversity for populations through allele richness, rarefied allele richness, and observed (Ho) and expected (He) heterozygosity under Hardy–Weinberg assumptions. The significance of the Hardy–Weinberg equilibrium was evaluated through a Bonferroni correction of P-values (Rice 1989). We also estimated inbreeding to measure the degree of substructure among them. We calculated FIS over the populations and loci using 10 000 permutations. These genetic diversity estimators 40 Climatic landscapes were performed in the R software using the adegenet (Jombart 2008), hierfstat (Goudet & Jombart 2011), and genetics (Warnes et al. 2019) packages. RESULTS Genetic extraction and genotyping success We genotyped 270 samples from jaguars and 363 from pumas. Both amplification and extraction were efficient, as indicated by the quality index and the number of complete loci (Table 1). We identified 71 unique genotypes (individuals) for jaguar and 106 for puma (Table 1). The number of individuals per sampling locality ranged between 0 and 9 for jaguar and between 0 and 14 for puma (Table 1), while the number of individuals per deme ranged between 1 and 21 for jaguar and 1 and 51 for puma. Climatic landscape effects on genetic similarity between individuals The evaluating statistics of niche predictions were, on average, higher for jaguars (TSS = 0.85 ± 0.05; AUC = 0.96 ± 0.02; Kappa = 0.80 ± 0.08) than for pumas (TSS = 0.73 ± 0.11; AUC = 0.91 ± 0.025 Kappa = 0.73 ± 0.11). Despite these differences, climatic 41 Model logLik deviance Chisq df P Jaguar FM–Pearson correlation 34.992 –69.983 NM–Pearson correlation 79.940 –159.881 89.898 15 <0.001 FM–P-value –64.208 128.417 NM–P-value –36.813 73.625 54.792 15 <0.001 Puma FM–Pearson correlation 159.25 –318.50 NM–Pearson correlation –279.12 –558.25 239.740 15 <0.001 FM–P-value –94.232 188.464 NM–P-value –19.503 39.005 149.460 15 <0.001 Table 2 Chi-square (Chisq) test to evaluate the ability of the mixed model (full model, FM) to describe the causal modeling results (Pearson correlation and Monte Carlo P-value) when compared with the null model (NM). Chapter 2 DISCUSSION Jaguars and pumas are the largest American felids and generally have similar ecological and biogeographical patterns of distribution (Wilson & Reeder 2005). Our analysis showed similar patterns of variation in genetic similarity based on climate suitability in the LMG scenario. However, the jaguar seems more vulnerable to climate changes than the puma because the modern climatic landscape has also led to differences between individuals. The higher susceptibility of jaguars to climate change is a likely reason why the species has more genetically structured populations than pumas, especially in South America. Meanwhile, changes in genetic variation in pumas have been continuous and gradual, suggesting a low influence of recent landscape changes, which is in line with the isolation by distance pattern observed between individuals. For both species, individuals located in LGM climatically suitable areas exhibited higher similarity if compared with individuals from climatically unsuitable areas, despite the distance among them, suggesting an effect of climatic exposure on the genetic configuration. LMG is largely recognized as a period of range contraction and the emergence of environmental refuges for many species (Stewart & Lister 2001; Provan & Bennett 2008; Sandel et al. 2011). However, jaguars and pumas have a large climatic niche (Zanin & Neves 2019) with a high dispersion capacity (Morato et al. 2016; Gonzalez-Borrajo et al. 2017), which enables them to move through less-suitable areas. As a result, this hinders the applicability of climatic refuge hypotheses in the classical interpretation as sites of high genetic diversification and speciation (Provan & Bennett 2008). On the other hand, some species with wide ranges seem to maintain large population sizes in climatically suitable areas (Tôrres et al. 2012; MartínezMeyer et al. 2013; Martínez-Gutiérrez et al. 2018). Therefore, these areas may function as a source of individuals by maintaining a larger population and thereby keeping a larger pool of alleles. The broad climatic tolerances and dispersion abilities of jaguars and pumas also suggest that the effect of climatic exposure was not a sudden event that isolated or connected individuals and populations. The most probable hypothesis is that 48 Climatic landscapes paleoclimate changes may have generated gradual and dynamic cycles of expansion and retraction of populations. These smoothing changes reduce the probability of allele fixation, reducing the risk of alleles disappearing in the population, and thereby enhancing genetic similarities (Aguilée et al. 2009). As a result, climatically suitable areas may have housed a diminished genetic differentiation and functioned as an “allele refuge” for both the jaguar and puma. The suitability of the modern climate also interacts positively with genetic similarity for jaguars. These areas could, therefore, also be keeping an allelic pool in their populations. Moreover, the connection routes between climatically suitable areas in the modern day seem to result in genetic exchange between individuals. Therefore, the effects of the modern climatic landscape on the jaguar probably go beyond those from the LMG landscape, suggesting vulnerability to recent anthropogenic impacts coming from climate changes. This result represents an alarming prospect for jaguar conservation, since it suggests the existence of fragile populations and a relatively short-term response of jaguars to climate changes. The genetic diversity of jaguars found in our study corroborates this warning since our estimates were similar to previous studies in which authors drew attention to the low genetic diversity and population bottlenecks (Roques et al. 2016; Srbek-Araujo et al. 2018). These results show the emerging need for securing populations located in climatically suitable areas, as well as providing connections over climatically suitable routes to facilitate range shifts into optimum areas. The threats can be more intense for marginal populations, which were poorly sampled in our study because they usually have a small population size and lower genetic diversity than core populations, which can drive genetic losses in marginal populations (Lesica & Allendorf 1995; Blevins et al. 2011; Davis et al. 2015). This study provides a map of climatic connection routes for jaguars (Fig. 3); however, our map does not incorporate land cover changes, one of the major threats (IUCN 49 Chapter 2 2020) affecting the jaguar’s genetic structure and demography (Haag et al. 2010; Zanin et al. 2015). Although some studies and field experiences have designed and implemented dispersion corridors for the jaguar from distributional up to local scales (Rabinowitz & Zeller 2010; Petracca et al. 2013, 2014; Rodríguez-Soto et al. 2013; Silveira et al. 2014; de la Torre et al. 2017), climatic suitability is not a common parameter to delineate the connection routes. Therefore, the design of corridors must integrate both threats, and existing corridors should be re-evaluated to guarantee their maximum effectiveness and long-term performance. Unlike jaguars, pumas have not been affected by the modern climatic landscape, showing that the latter species has not experienced a substantial change in genetic configuration between the study sites in recent times. This pattern is probably due to the broad climatic niche of the puma, the largest among felids (Zanin & Neves 2019). This enables pumas to occupy areas ranging from cold climates such as the Rocky Mountains to semi-arid environments like the Brazilian Caatinga. Despite the climatic plasticity, the positive effect of LMG on the puma’s genetic isolation shows its responsiveness to climatic depletion. Therefore, the current scenario of the puma’s genetic configuration must be viewed with caution because future climate changes are predicted to reach higher rates and velocity than paleoclimatic changes (Loarie et al. 2009; Schloss et al. 2012; Zanin & Albernaz 2016). The current genetic scenario of the puma, therefore, represents an opportunity to design appropriate conservation plans to avoid further genetic erosion. Some limitations of our approach may be considered concerning the interpretation and application of results. The grid cell resolution (0.5 decimal degrees) can considerably overestimate ecological niche breadth when compared with a higher resolution (Trived et al. 2008). Another relevant consideration has to do with the distance measured by the least-cost routes, which is constrained to only one connection path. Thus, the method ignores alternative routes and solutions (Rayfield et al. 2010; Parks et al. 2020). Moreover, the least-cost routes can be influenced by the resolution scale and by the method chosen to weight geographic and environmental distances (Adriaensen et al. 50 Climatic landscapes 2003; Beier et al. 2009). The resolution we use is appropriate for the macroecological investigation since it provides a good balance between suitable environmental representation and logistical feasibility (Zanin & Neves 2019), and may also present a large amount of biological information (Soberón 2010). However, a higher resolution and more realistic representation of species’ movement are essential to investigations in which precise inferences are needed, such as conservation actions or studies at local or regional scales. Despite these limitations, our results provide valuable information on the effects of climatic conditions on jaguars and pumas. Our findings lead us to suppose that future climate changes will create genetic isolation and diversity losses for pumas and aggravate the already alarming situation of jaguars. The impacts of climate changes on genetic diversity of species have been increasingly documented in the literature (Aitken et al. 2008; Rubidge et al. 2012; Row et al. 2014; Loveless et al. 2016; Lima et al. 2017). This amount of evidence makes clear that climatic projection scenarios have to be included in species conservation agendas to ensure evolutionary and demographic processes mediated by gene flow. In this way, providing refuges in suitable areas of climatic stability and dispersion routes among climatically vulnerable populations is essential to the conservation of the species. It is important to highlight that our sampling localities were in areas with relatively homogeneous land cover, meaning that our inferences are only an approximation of the species’ natural response to past climate change. However, jaguars and pumas are subjected to strong range contraction due to habitat loss and fragmentation, which can exacerbate the effects of climate changes on genetic population patterns. ACKNOWLEDGMENTS This work was supported by project CGL2010-16902 of the Spanish Ministry of Research and Innovation, project CGL2013-46026-P of Ministerio de Economía, Industria y Competitividad, excellence project RNM2300 of Junta de Andalucía 51 Chapter 2 (Spain), the Formación de Profesorado Universitario fellowship #AP2010-5373 from the Spanish Ministry of Education, and by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior Brasil (CAPES) (Finance Code 001). L.P.C. has a fellowship from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). M.Z. is supported by CAPES (grant number 88887.478136/2020-00) through the Program of National Cooperation in the Amazon (Programa Nacional De Cooperação Acadêmica na Amazônia). 52 Climatic landscapes REFERENCES Adriaensen F, Chardon JP, De Blust G et al. (2003). The application of ‘least-cost’ modelling as a functional landscape model. Landscape and Urban Planning 64, 233–47. Aguilée R, Claessen D, Lambert A (2009). Allele fixation in a dynamic metapopulation: Founder effects vs refuge effects. Theoretical Population Biology 76, 105–17. Aitken SN, Yeaman S, Holliday JA et al. (2008). Adaptation, migration or extirpation: climate change outcomes for tree populations. Evolutionary Applications 1, 95–111. Allouche O, Tsoar A, Kadmon R (2006). Assessing the accuracy of species distribution models: Prevalence, kappa and the true skill statistic (TSS). Journal of Applied Ecology 43, 1223–32. Araújo MB, New M (2007). Ensemble forecasting of species distributions. Trends in Ecology & Evolution 22, 42–7. Assis J, Serrão EÁ, Coelho NC et al. (2018). Past climate changes and strong oceanographic barriers structured low-latitude genetic relics for the golden kelp Laminaria ochroleuca. Journal of Biogeography 45, 2326– 36. Balkenhol N, Holbrook JD, Onorato D et al. (2014). A multi-method approach for analyzing hierarchical genetic structures: A case study with cougars Puma concolor. Ecography 37, 552–63. Barrientos R, Kvist L, Barbosa A et al. (2014). Refugia, colonization and diversification of an arid-adapted bird: Coincident patterns between genetic data and ecological niche modelling. Molecular Ecology 23, 390–407. Bates D, Mächler M, Bolker B, Walker S (2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67, 1–48. Beier P, Majka DR, Newell SL (2009). Uncertainty analysis of least-cost modeling for designing wildlife linkages. Ecological Applications 19, 2067–77. Blevins E, Wisely SM, With KA (2011). Historical processes and landscape context influence genetic structure in peripheral populations of the collared lizard (Crotaphytus collaris). Landscape Ecology 26, 1125– 36. Boom R, Sol CJ, Salimans MM et al. (1990). Rapid and simple method for purification of nucleic acids. Journal of Clinical Microbiology 28, 495–503. Breiman L (2001). Random Forests. Machine Learning 45, 5–32. Caruso N, Guerisoli M, Luengos Vidal EM et al. (2015). Modelling the ecological niche of an endangered population of Puma concolor: first application of the GNESFA method to an elusive carnivore. Ecological Modelling 297, 11–9. Carvalho C da S, Ballesteros-Mejia L, Ribeiro MC et al. (2017). Climatic stability and contemporary human impacts affect the genetic diversity and conservation status of a tropical palm in the Atlantic Forest of Brazil. Conservation Genetics 18, 467–78. Caye K, Deist TM, Martins H et al. (2016). TESS3: Fast inference of spatial population structure and genome scans for selection. Molecular Ecology Resources 16, 540–8. Chen C, Durand E, Forbes F, François O (2007). Bayesian clustering algorithms ascertaining spatial population structure: A new computer program and a comparison study. Molecular Ecology Notes 7, 747–56. Chimento NR, Dondas A (2018). First record of Puma concolor (Mammalia, Felidae) in the Early-Middle Pleistocene of South America. Journal of Mammalian Evolution 25, 381–9. Cushman SA, Landguth EL (2010). Spurious correlations and inference in landscape genetics. Molecular Ecology 19, 3592–602. Cushman S, Wasserman T, Landguth E, Shirk A (2013). Re-evaluating causal modeling with Mantel tests in landscape genetics. Diversity 5, 51–72. Davis DM, Reese KP, Gardner SC, Bird KL (2015). Genetic structure of Greater SageGrouse (Centrocercus urophasianus) in a declining, peripheral population. The Condor 117, 530–44. de la Torre JA, Núñez JM, Medellín RA (2017). Habitat availability and connectivity for jaguars (Panthera onca) in the Southern Mayan Forest: Conservation priorities for a fr ag me nt ed l an d sc ap e. Biological Conservation 206, 270–82. Defries R, Bounoua L, Collatz G (2002). Human modification of the landscape and surface climate in the next fifty years. Global Change Biology 8, 438–58. Durand E, Jay F, Gaggiotti OE, François O (2009). Spatial inference of admixture proportions and secondary contact zones. Molecular Biology and Evolution 26, 1963– 73. 53 Chapter 2 Efron B, Morri C (1977). Stein’s paradox in statistics. Scientific American 236, 119–27. Ernest H, Boyce W, Bleich V et al. (2003). Genetic structure of mountain lion (Puma concolor) populations in California. Conservation Genetics 4, 353–66. Faurby S, Araújo MB (2018). Anthropogenic range contractions bias species climate change forecasts. Nature Climate Change 8, 252–6. Fielding AH, Bell JF (1997). A review of methods for the assessment of prediction errors in conservation presence/absence models. Environmental Conservation 24, 38–49. Foley JA, Defries R, Asner GP et al. (2005). Global consequences of land use. Science 309, 570–4. François O, Ancelet S, Guillot G (2006). Bayesian clustering using hidden Markov random fields in spatial population genetics. Genetics 174, 805–16. François O, Durand E (2010). Spatially explicit Bayesian clustering models in population genetics. Molecular Ecology Resources 10, 773–84. Frantz AC, Pope LC, Carpenter PJ et al. (2003). Reliable microsatellite genotyping of the Eurasian badger (Meles meles) using faecal DNA. Molecular Ecology 12, 1649–61. Friedman J (1991). Multivariate adaptive regression splines. The Annals of Statistics 19, 1–141. Gonzalez-Borrajo N, López-Bao JV, Palomares F (2017). Spatial ecology of jaguars, pumas, and ocelots: A review of the state of knowledge. Mammal Review 47, 62–75. Goudet J, Jombart T (2011). hierfstat: Estimation and tests of hierarchical F-Statistics. R package version 0.04-22. Available from URL: https://CRAN.Rproject.org/package=hierfstat Gotelli NJ, Ellison AM (2004). A Primer of ecological statistics. Sinauer Associates, Sunderland, Massachusets. Guisan A, Thuiller W (2005). Predicting species distribution: offering more than simple habitat models. Ecology Letters 8, 993–1009. Haag T, Santos AS, Sana DA et al. (2010). The effect of habitat fragmentation on the genetic structure of a top predator: Loss of diversity and high differentiation among remnant populations of Atlantic Forest jaguars (Panthera onca). Molecular Ecology 19, 4906–21. Habel JC, Zachos FE, Dapporto L et al. (2015). Population genetics revisited—towards a multidisciplinary research field. Biological Journal of the Linnean Society 115, 1–12. Inoue K, Berg DJ (2017). Predicting the effects of climate change on population connectivity and genetic diversity of an imperiled freshwater mussel, Cumberlandia monodonta (Bivalvia: Margaritiferidae), in riverine systems. Global Change Biology 23, 94–107. IUCN (2017). Metadata: Digital distribution maps on the IUCN Red List of threatened species. Version 6.2, January 2019. Available from URL: https://www. iucnredlist.org/resources/ spatial-data-download IUCN (2020). The IUCN Red List of Threatened Species. Version 2020–1. Available from URL: http://www. iucnredlist.org/ Jakobsson M, Rosenberg NA (2007). CLUMPP: a cluster matching and permutation program for deal ing wit h l abe l s wit chi ng and multimodality in analysis of population structure. Bioinformatics 23, 1801–6. Jombart T (2008). adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics 24, 1403–5. Jost L (2008). GST and its relatives do not measure differentiation. Molecular Ecology 17, 4015–26. Koch JB, Vandame R, Mérida-Rivas J et al. (2018). Quaternary climate instability is correlated with patterns of population genetic variability in Bombushuntii. Ecology and Evolution 8, 7849–64. Kosman E, Leonard KJ (2005). Similarity coefficients for molecular markers in studies of genetic relationships between individuals for haploid, diploid, and polyploid species. Molecular Ecology 14, 415–24. Kurtén B, Anderson E (1980). Pleistocene Mammals of North America, 1st edn. Columbia University Press, New York. Legendre P, Legendre L (1998). Numerical Ecology, 2nd edn. Elsevier Science, Amsterdam. Lesica P, Allendorf FW (1995). When are peripheral populations valuable for conservation? Conservation Biology 9, 753– 60. Licona-Vera Y, Ornelas JF, Wethington S, Bryan KB (2018). Pleistocene range expansions promote divergence with gene flow between migratory and sedentary populations of 54 Climatic landscapes Calothorax humming birds. Biological Journal of the Linnean Society 124, 645–67. Lima JS, Ballesteros-Mejia L, Lima-Ribeiro MS, Collevatti RG (2017). Climatic changes can drive the loss of genetic diversity in a Neotropical savanna tree species. Global Change Biology 23, 4639–50. Lima-Ribeiro MS (2015). EcoClimate: A database of climate data from multiple models for past, present, and future for macroecologists and biogeographers. Biodiversity Informatics 10, 1–21. Lima-Ribeiro MS, Varela S, González-Hernández J et al. (2018). The ecoClimate Database. Available from URL: http://ecoclimate.org Loarie SR, Duffy PB, Hamilton H et al. (2009). The velocity of climate change. Nature 462, 1052–5. Loveless AM, Reding DM, Kapfer PM, Papes # M (2016). Combining ecological niche modelling and morphology to assess the range-wide population genetic structure of bobcats (Lynx rufus). Biological Journal of the Linnean Society 117, 842–57. Mairal M, Sanmartín I, Herrero A et al. (2017). Geographic barriers and Pleistocene climate change shaped patterns of genetic variation in the Eastern Afromontane biodiversity hotspot. Scientific Reports 7, 1–13. Martínez-Freiría F, Tarroso P, Rebelo H, Brito JC (2016). Contemporary niche contraction affects climate change predictions for elephants and giraffes. Diversity and Distributions 22, 432–44. Martínez-Gutiérrez PG, Martínez-Meyer E, Palomares F, Fernández N (2018). Niche centrality and human influence predict rangewide variation in population abundance of a widespread mammal: The collared peccary (Pecari tajacu). Diversity and Distributions 24, 103– 15. Martínez-Meyer E, Díaz-Porras D, Peterson AT, YáñezArenas C (2013). Ecological niche structure and rangewide abundance patterns of species. Biology Letters 9, 20120637. Matthews PE, Heath AC (2008). Evaluating historical evidence for occurrence of mountain goats in Oregon. Northwest Science 82, 286– 98. McCullagh P, Nelder JA (1989). Generalized Linear Models, 2nd edn. Springer US, Boston, MA. McRae BH (2006). Isolation by resistance. Evolution 60, 1551–61. Miotto RA, Cervini M, Figueiredo MG et al. (2011). Genetic diversity and population structure of pumas (Puma concolor) in southeastern Brazil: Implications for conservation in a human-dominated landscape. Conservation Genetics 12, 1447–55. Morato RG, Stabach JA, Fleming CH et al. (2016). Space use and movement of a Neotropical top predator: The endangered jaguar. PLoS ONE 28, 1–17. Noguerales V, Cordero PJ, Ortego J (2017). Testing the role of ancient and contemporary landscapes on structuring genetic variation in a specialist grasshopper. Ecology and Evolution 7, 3110–22. Oksanen J, Blanchet FG, Friendly M et al. (2019). vegan: Community ecology package. R package version 2.55. Available from URL: https://CRAN.R-project.org/ package=vegan Palomares F, Adrados B, Zanin M et al. (2017). A noninvasive faecal survey for the study of spatial ecology and kinship of solitary felids in the Viruá National Park, Amazon Basin. Mammal Research 62, 241–9. Palomares F, Fernández N, Roques S et al. (2016). Finescale habitat segregation between two ecologically similar top predators. PLoS ONE 11, e0155626. Parks SA, Carroll C, Dobrowski SZ, Allred BW (2020). Human land uses reduce climate connectivity across North America. Global Change Biology 26, 2944–55. Pauls SU, Nowak C, Bálint M, Pfenninger M (2013). The impact of global climate change on genetic diversity within populations and species. Molecular Ecology 22, 925–46. Peterson AT, Papes # M, Soberón J (2008). Rethinking receiver operating characteristic analysis applications in ecological niche modeling. Ecological Modelling 213, 63–72. Petracca LS, Hernández-Potosme S, ObandoSampson L, Salom-Pérez R, Quigley H, Robinson HS (2014). Agricultural encroachment and lack of enforcement threaten connectivity of range-wide jaguar (Panthera onca) corridor. Journal for Nature Conservation 22, 436–44. Petracca LS, Ramirez-Bravo OE, HernandezSantin L (2013). Occupancy estimation of jaguar Panthera onca to assess the value of 55 Chapter 2 east-central Mexico as a jaguar corridor. Oryx 48, 1–8. Pflüger FJ, Balkenhol N (2014). A plea for simultaneously considering matrix quality and local environmental conditions when analysing landscape impacts on effective dispersal. Molecular Ecology 23, 2146–56. Phillips SJ, Anderson RP, Schapire RE (2006). Maximum entropy modeling of species geographic distributions. Ecological Modelling 190, 231–59. Provan J, Bennett KD (2008). Phylogeographic insights into cryptic glacial refugia. Trends in Ecology and Evolution 23, 564–71. R Core Team (2019). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available from URL: https:// www.Rproject.org/ Rabinowitz A, Zeller KA (2010). A range-wide model of landscape connectivity and conservation for the jaguar. Panthera onca Biological Conservation 143, 939–45. Rice WR (1989). Analysing tables of statistical tests. Evolution 43, 223–5. Rodriguez SG, Méndez C, Soibelzon E et al. (2018). Panthera onca (Carnivora, Felidae) in the late Pleistocene-early Holocene of northern Argentina. Neues Jahrbuchfür Geologie und Paläontologie – Abhandlungen 289, 177–87. Rodríguez-Soto C, Monroy-Vilchis O, ZarcoGonzález MM (2013). Corridors for jaguar (Panthera onca) in Mexico: Conservation strategies. Journal for Nature Conservation 21, 438–43. Roques S, Adrados B, Chavez C et al. (2011). Identification of neotropical felid faeces using RCP-PCR. Molecular Ecology Resources 11, 171–5. Roques S, Furtado M, Jácomo ATA et al. (2014). Monitoring jaguar populations (Panthera onca) with noninvasive genetics: A pilot study in Brazilian ecosystems. Oryx 48, 361–9. Roques S, Sollman R, Jácomo A et al. (2016). Effects of habitat deterioration on the population genetics and conservation of the jaguar. Conservation Genetics 17, 125–39. Row JR, Wilson PJ, Gomez C et al. (2014). The subtle role of climate change on population genetic structure in Canada lynx. Global Change Biology 20, 2076–86. Rubidge EM, Patton JL, Lim M et al. (2012). Climateinduced range contraction drives genetic erosion in an alpine mammal. Nature Climate Change 2, 285–8. Salzano FM, Valdez FP, Haag T et al. (2015). Population genetics of jaguars (Panthera onca) in the Brazilian Pantanal: Molecular evidence for demographic connectivity on a regional scale. Journal of Heredity 106, 503– 11. Sandel B, Arge L, Dalsgaard B et al. (2011). The influence of late quaternary climate-change velocity on species endemism. Science 334, 660–4. Schloss CA, Nunez TA, Lawler JJ (2012). Dispersal will limit ability of mammals to track climate change in the Western Hemisphere. PNAS 109, 8606–11. Silva LG, Kawanishi K, Henschel P et al. (2017). Mapping black panthers: macroecological modeling of melanism in leopards (Panthera pardus). PLoS ONE 12, 1–17. Silveira L, Sollmann R, Jácomo ATA, Diniz Filho JAF, Tôrres NM (2014). The potential for large-scale wildlife corridors between protected areas in Brazil using the jaguar as a model species. Landscape Ecology 29, 1213– 23. Soberón J, Nakamura M (2009). Niches and distributional areas: concepts, methods, and assumptions. PNAS 106, 19644–50. Soberón JM (2010). Niche and area of distribution modeling: A population ecology perspective. Ecography 33, 159–67. Spiegelhalter DJ, Best NG, Carlin BP, van der Linde A (2002). Bayesian measures of model complexity and fit. Journal of the Royal Statistical Society: Series B (Statistical Methodology) 64, 583–639. Srbek-Araujo AC, Haag T, Chiarello AG et al. (2018).Worrisome isolation: noninvasive genetic analyses shed light on the critical status of a remnant jaguar population. Journal of Mammalogy 99, 397–407. Stewart JR, Lister AM (2001). Cryptic northern refugia and the origins of the modern biota. Trends in Ecology and Evolution 16, 608–13. Thuiller W, Lafourcade B, Engler R, Araújo MB (2009). BIOMOD—A platform for ensemble forecasting of species distributions. Ecography 32, 369–73. Tingley MW, Beissinger SR (2009). Detecting range shifts from historical species 56 Climatic landscapes occurrences: New perspectives on old data. Trends in Ecology and Evolution 24, 625–33. Torres N, Junior P, Santos T, Silveira L, Jacomo A, DinizFilho J (2012). Can species distribution modelling provide estimates of population densitites? A case study with jaguars. Diversity and Distributions 18, 615–27. Tôrres NM, De Marco P, Santos T et al. (2012). Can species distribution modelling provide estimates of population densities? A case study with jaguars in the Neotropics. Diversity and Distributions 18, 615–27. Trivedi MR, Berry PM, Morecroft MD, Dawson TP (2008). Spatial scale affects bioclimate model projections of climate change impacts on mountain plants. Global Change Biology 14, 1089–103. van Etten J (2012). gdistance: distances and routes on geographical grids. R package version 1.2-2. Available from URL: https://CRAN.Rproject.org/package= gdistance Warnes G, Gorjanc G, Leisch F, Man M (2019). Genetics: population genetics. R package version 1.3.8.1.2. Available from URL: https:// CRAN.Rproject.org/ package=genetics. Whitlock MC (2011). G’ST and D do not replace FST. Molecular Ecology 20, 1083–91. Wilson DE, Reeder DM (2005). Mammal Species of the World. A Taxonomic and Geographic Reference, 3rd edn. Johns Hopkins University Press, Baltimore, MA. Winter AD (2013). Package ‘mmod’: Modern measures of population differentiation, pp. 1– 7. R package version 1.3.3. Available from URL: https://github.com/ dwinter/mmod Zanin M, Adrados B, González N et al. (2016). Gene flow and genetic structure of the puma and jaguar in Mexico. European Journal of Wildlife Research 62, 461–9. Zanin M, Albernaz ALM (2016). Impacts of climate change on native landcover: seeking future climatic refuges. PLoS ONE 11, e0162500. Zanin M, Palomares F, Brito D (2015). The jaguar’s patches: viability of jaguar populations in fragmented landscapes. Journal for Nature Conservation 23, 90–7. Zanin M, Neves BS (2019). Current felid (Carnivora: Felidae) distribution, spatial bias, and occurrence predictability: Testing the reliability of a global dataset for macroecological studies. Acta Oecologica 101, 103488. Zhang Z (2018). Artificial Neural Network. In: Multivariate Time Series Analysis in Climate and Environmental Research. Springer International Publishing, Cham, pp. 1–35. 57 Jaguar extinction #####Chapter 3! A continental approach to jaguar extirpation: a tradeoff between anthropic and intrinsic causes! Journal for Nature Conservation 2022 65 Chapter 3 66 Jaguar extinction Human impacts are blamed for range contraction in several animal species worldwide. Remarkably, carnivores and particularly top predators are threatened by humans despite their key role in maintaining ecosystem balance and functions. Conservation strategies to allow human-carnivore coexistence are urgently needed and must be built on evidence and driven by knowledge of population risk at a broad scale. However, knowledge on wide distributed species is often based on regional expert opinions in which uncertainty is not quantifiable, making data incomparable across regions. Here we develop a method to assess endangerment status using information on range contraction and main threats using the jaguar Panthera onca as model species. The use of GLM with the main intrinsic and extrinsic drivers of jaguar extinction allowed us to assess endangerment status at continental and population scale. We found this method to be a valuable tool to obtain a broad picture of human-induced endangerment in a species. Intrinsic traits (summarized in the demographic contraction theory) and anthropic traits (based on agriculture, cattle and human densities) explained jaguar extinction highlighting the particular importance of livestock activity. Our results suggest that livestock ranching has a pervasive effect on the species likely due to habitat loss combined with retaliatory hunting. We highlight the need to rethink policies, practice and law enforcement in relation to livestock and suggest the development of action plans based in local evidence in those countries where endangered populations have been detected. We also recommend involving and encouraging land owners and private companies in the conservation of private lands that comprise much of the endangered jaguar range. Key words: Human-carnivore coexistence, extinction risk modelling, jaguar, retaliatory hunting. 67 Abstract Chapter 3 INTRODUCTION: Species extinctions generally start with the vanishing of particular populations that continues until no populations remain (Yackulic et al. 2011). Understanding the general dynamics of species range contraction is key for effective conservation, and predicting population extinction risk is an important step toward achieve this goal (Safi & Pettorelli, 2010). There are two main factors determining carnivores’ extinction risk (Purvis, 2000): 1) intrinsic traits, such as body mass (Inskip & Zimmermann, 2009), life history (Pearson et al., 2014) or population genetics (Frankham, 2005), and 2) extrinsic causes based on modern exposure to external anthropogenic threats (Bruskotter et al., 2017), including human-driven disease expansion (Pedersen et al., 2007). Extrinsic causes have been repeatedly suggested as important traits that affect direct or indirectly carnivore populations. In fact, human density (Woodrooffe, 2000), prey depletion (Craige et al., 2010), habitat loss (Cardinale et al., 2012) or retaliation (Jedrezewski et al., 2017) have been highlighted as the most important drivers of many carnivore species. Intrinsic traits may be summarized by the demographic contraction theory (Lawton 1993, Brown, 1995), which is derived from basic population dynamics. It assumes that environmental conditions and resources for species at the center of their distribution are more suitable than at the border, resulting in higher population growth rates and thus, higher abundance in central areas. This theory predicts that populations would be first extirpated along the range border (where density is lower) and would continue toward the center. On the other hand, more humanized landscapes are associated with higher risk of extirpation (Laliberte and Ripple 2004, Schipper et al. 2008, Hoffmann et al. 2010, Fisher 2011, Pomara et al. 2014), predicting that populations would be first extirpated in areas where human activities such as agriculture, cattle raising and urbanization dominate. Risk models have been widely used to provide valuable information for conservation purposes. Most risk model assessments, particularly for larger mammalian carnivores, are based on habitat suitability often omitting anthropic traits that are usually indicated as responsible for the non-explicated variation (Cardillo, 2004; Brashares et al. 2001; Harcourt et al. 2001; McKinney 2001; Ceballos & Ehrlich 2002; Parks & Harcourt 2002). Studies taking into account anthropic variables 68 Jaguar extinction (Rodriguez-Soto et al. 2011, 2013, 2017, Zarco-Gonzalez et al. 2013) become more realistic models but use to be local scaled (but see Jedrezejewski et al. 2018). However, knowledge of both drivers of population declines and ultimately species extinction should be one of the main focusses to facilitate the allocation of the limited resources to specific conservation interventions (Mace et al., 1993; Davies et al., 2008; Travers et al., 2016). During the past decade, knowledge about the ecology and conservation of top predators has increased widely. In particular, jaguars (Panthera onca) have received great attention since they have suffered a reduction to at least 46% (Sanderson et al., 2002) of their historic range. Since Sanderson et al. (2002) defined the Jaguar Conservation Units (henceforth JCU), many updates and revisions of jaguar status have been made and this percentage has been currently set to 51% in the last IUCN expert assessment (Quigley et al., 2017). Global experts made contributions and reviewed the state of the art completing the actual conservation snapshot for many regions (i.e., Quigley et al. 2017 and Medellín et al. 2016). This effort generated a large amount of information that not only promotes jaguar conservation but also enables the use of the species as a model to test for specific hypotheses about species range contraction. Again, most of the continental scale research has been focused on niche prediction and its relation to ecological patterns, demonstrating the efficiency of climatic variables as predictors of the felid niche suitability (Torres et al., 2012; Martínez-Meyer et al., 2013; Caruso et al., 2015; Zanin et al., 2019). However most continental scale approaches ignore anthropic disturbances, vital to understand species decline, and have been only assessed at a regional scale (Cavalcanti et al., 2010; Zimmermann et al., 2005; Rodriguez-Soto et al. 2011, 2013; Marchini et al., 2012; Villalva and Palomares, 2019). Moreover, ecological research on jaguars has largely focused on populations inhabiting protected areas (Foster et al., 2020). Consequently, few data are available on jaguars occupying the mosaic of human settlements and agriculture. These areas comprise typical unprotected landscapes, including jaguar movement corridors and private properties (Hoogestijn et al, 2015) that have been identified as vital to maintaining the genetic connectivity of jaguars (Zeller et al., 2013). Furthermore, global estimates are based on “expert opinion” which is burdened with a high and unquantifiable level of uncertainty 69 Chapter 3 (Akçakaya et al., 2000; Rodrigues et al., 2006) that may result in a possible over or underreporting of threats to jaguars in certain areas (Zeller, 2006). The role of spatially explicit models appears to be a valuable tool as they can be a compliment to expertbased information (Bernal, 2015). The aim of this study is to identify the endangerment of jaguars across their current range. As mentioned, jaguars are mainly threatened by intrinsic and external factors. The intrinsic threats are related to their large size, high trophic level, slow population dynamics and fragmentation of its populations that increases their extinction risk (Purvis et al., 2000). Thus, we expect them to exhibit a periphery-to-center extinction pattern. The extrinsic threats are related to human activities such as habitat loss on behalf of agriculture, cattle ranching (including persecution by predation on livestock) and urbanization (Boron et al., 2016, Quigley et al., 2017). We expect the probability of jaguar extirpation to increase in areas with higher human pressures compared to more natural areas. More concisely, we expect to find jaguars especially affected by livestock due to the combined effect of habitat loss and the deeply rooted retaliatory killing used to minimize economic losses, while agriculture and urbanization may moderately affect the species due to the lower conflict level with the main economic interests. Accordingly, we build an extinction threat model based on aforementioned intrinsic and extrinsic traits that will contribute to generating a comprehensive snapshot of current jaguar endangerment that may help in the elaboration of conservation strategies. 70 Jaguar extinction METHODS: We examine the extinction probability of jaguars for their entire distribution range using General Linear Models (GLM). We used current and historic distribution to construct a binomial response variable, assigning value zero to areas where jaguars still persist and value one to areas where jaguars have been extirpated (Fig. 1. Supp. 1). Current range (Quigley et al., 2017) was subtracted from historic range (Seymour, 1989) to obtain the area where jaguars have been locally extinct. To generate the dataset, we randomly displayed a set of 1000 points on current and extirpated areas extracting the same amount of data in each to avoid overdispersion. We used a set of predictor variables based on intrinsic and extrinsic traits. The intrinsic variable was defined as the distance to the border from the historic distribution grounded on the demographic range contraction hypothesis. The extrinsic predictors contained a selection of the main reported human drivers of jaguar extinction such as fragmentation, habitat loss and felid persecution by humans (Quigley et al., 2017). Thus, we selected urban, agriculture and cattle ranching as the main human disturbances related to the aforementioned drivers. Human density was used as a proxy for urban disturbance based on the SEDAC-NASA model GRUMP 2000 (CIESIN, 2017). This is a population estimate at the municipal level, corrected for night-light. Agriculture disturbance was based on Global Land Cover (Bartholome et al., 2002; Eva et al., 2004) from The European Council. We clustered all types of agriculture (Cultivated and managed areas, Mosaic cropland with tree and Mosaic cropland with shrub) and resampled from 1 to 10 Km2 pixel size obtaining the percentage of agriculture cover. Livestock pressure was acquired from Global Livestock of the World (Robinson et al., 2014) based on livestock data at a municipal level representing cattle density (number of heads/km2). All layers were unified at a pixel value of 100 Km2 based on a conservative home range size for a jaguar male (Gonzalez-Borrajo et al., 2017). When redefining pixel size, human and livestock pressure data took the mean value of the underlying pixels, while agriculture pressure, as explained above, was obtained as a percentage of cover regarding the 100 underlying pixels of 1 Km2 contained in each sample grid of 100 km2. 71 Chapter 3 We tested the effect of intrinsic and external traits on jaguar extinction using a logit link with species extinction as the response variable. We first constructed univariate models to graphically evince the relationship of extinction to each independent variable. Then, using an information-theoretic approach, we analyzed the effect of intrinsic and extrinsic variables on the probability of jaguar extinction by comparing the generated presence-absence data, guided by three general hypotheses: (i) extrinsic traits (human disturbance) caused jaguar extinction; (ii) intrinsic traits (inherent species contraction) caused extinction; and (iii) the combination of the two influenced extinction. We also constructed a null model that included no explanatory variables. Prior to modelling we searched for confounding effects among predictor variables using Pearson´s rank correlation to avoid multicollinearity (Zar, 1999). We made the best model spatially explicit within the jaguar current range using the logit function. Note that the intrinsic variable used here was the distance to the border from the current distribution range rather than the distance to the historic distribution. All other variables remained unchanged. Finally, we focused on Jaguar Conservation Units (Sanderson et al., 2002) to compare the endangerment status among the most important and stable jaguar 72 Variables AIC Ɓ S.E. p value Natural contraction Intercept 0.820 0.108 <0.001 Distance to border 1255 -0.211 0.022 <0.001 Anthropic variables Intercept -0.621 0.085 <0.001 Cattle density 1212 0.033 0.003 <0.001 Intercept -0.532 0.082 <0.001 Agriculture 1258 0.023 0.002 <0.001 Intercept -0.166 0.071 0.01 Human density 1339 0.020 0.004 <0.001 Table 1. Univariate models for jaguar extinction. Comparison of models based on AIC and Beta, Standard error (S.E.) and significance level (p-value) are also shown. Jaguar extinction populations. Values for each Jaguar Conservation Unit were calculated as the mean value of pixels contained in the polygon. We transformed and processed the data with Q-Gis (2.8.9-Wien), and raster (Hijmans, 2020), spatial (Venables & Ripple, 2002), terra (Hijman, 2021) and sf (Pebesma, 2018) packages in R, and carried out the statistical analyses using the MuMIn (Barton, 2020) package in R version 4.0.3 (R Development Core Team, 2021). RESULTS: Univariant models showed the relationship of all predictors with jaguar extirpation (Table 1). Distance to the border relate negatively with extinction whereas anthropic variables showed a positive relationship. Cattle density showed the lowest AIC among univariant predictors. Each variable affected extinction differently (Fig. 1). 73 Figure 1. Univariate models for cattle density (heads/Km²), agriculture coverage (%), human density (people/Km²) and distance to the border (DD) related to probability of extinction. References Akçakaya, H. R. et al. (2000) ‘Making Consistent IUCN Classifications under Uncertainty’, Conservation Biology. John Wiley & Sons, Ltd, 14(4), pp. 1001–1013. doi: https://doi.org/ 10.1046/j.1523-1739.2000.99125.x. Angelo, C. et al. (2016) El jaguar en el siglo xxi La perspectiva continental. Athreya, V. et al. (2013) ‘Big Cats in Our Backyards: Persistence of Large Carnivores in a Human Dominated Landscape in India’, PLOS ONE. Public Library of Science, 8(3), p. e57872. Available at: https://doi.org/ 10.1371/journal.pone.0057872. Bartholome, E. et al. (2002) ‘GLC 2000 Global Land Cover mapping for the year 2000’, European Comission Joint Research Center, (November), p. 62. Barton, K. (2020) ‘Multi-Model Inference. R package version 1.43.17’. Bernal-Escobar, A., Payan, E. and Cordovez, J. M. (2015) ‘Sex dependent spatially explicit stochastic dispersal modeling as a framework for the study of jaguar conservation and management in South America’, Ecological Modelling, 299, pp. 40–50. doi: https:// doi.org/10.1016/j.ecolmodel.2014.12.002. Boron, V. et al. (2016) ‘Jaguar Densities across Human-Dominated Landscapes in Colombia: The Contribution of Unprotected Areas to Long Term Conservation’, PLOS ONE. Public Library of Science, 11(5), p. e0153973. Available at: https://doi.org/10.1371/ journal.pone.0153973. Brashares, J. S., Arcese, P. and Sam, M. K. (2001) ‘Human demography and reserve size predict wildlife extinction in West Africa’, Proceedings of the Royal Society of London. Series B: Biological Sciences. Royal Society, 268(1484), pp. 2473–2478. doi: 10.1098/ rspb.2001.1815. Brown, J. H., Mehlman, D. W. and Stevens, G. C. (1995) ‘Spatial Variation in Abundance’, Ecology. John Wiley & Sons, Ltd, 76(7), pp. 2028–2043. doi: https://doi.org/ 10.2307/1941678.Cardillo, M. et al. (2004) ‘Human Population Density and Extinction Risk in the World’s Carnivores’, PLOS Biology. Public Library of Science, 2(7), p. e197. Available at: https://doi.org/10.1371/ journal.pbio.0020197. Bruskotter, J.T., Vucetich, J.A., Manfredo, M.J., Karns, G.R., Wolf, C., Ard, K., Carter, N.H., López-Bao, J.V., Chapron, G., Gehrt, S.D., Ripple, W.J. ‘Modernization, Risk, and Conservation of the World's Largest Carnivores. ’%BioScience, Volume 67, Issue 7, July 2017, Pages 646-655. https://doi.org/ 10.1093/biosci/bix049 Cardinale, B. J. et al. (2012) ‘Biodiversity loss and its impact on humanity’, Nature, 486(7401), pp. 59–67. doi: 10.1038/nature11148. Caruso, N. et al. (2015) ‘Modelling the ecological niche of an endangered population of Puma concolor: First application of the GNESFA method to an elusive carnivore’, Ecological Modelling, 297, pp. 11–19. doi: https://doi.org/ 10.1016/j.ecolmodel.2014.11.004. Cavalcanti, S. and Gese, E. (2009) ‘Spatial Ecology and Social Interactions of Jaguars (Panthera Onca) in the Southern Pantanal, Brazil’, Journal of Mammalogy, 90. doi: 10.1644/08-MAMM-A-188.1. Ceballos, G. and Ehrlich, P. R. (2002) ‘Mammal population losses and the extinction crisis.’, Science (New York, N.Y.), 296(5569), pp. 904–7. doi: 10.1126/science.1069349. Chapron, G. et al. (2014) ‘Recovery of large carnivores in Europe’s modern humandominated landscapes’, Science, 346, pp. 1517–1519. doi: 10.1126/science.1257553. CIESIN. (2017) ‘Global Population Density Grid Time Series Estimates’. Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). Available at: https://doi.org/ 10.7927/H47M05W2. Craigie, I. D. et al. (2010) ‘Large mammal population declines in Africa’s protected areas’, Biological Conservation, 143(9), pp. 2221–2228. doi: https://doi.org/10.1016/ j.biocon.2010.06.007. Cushman, S. A. et al. (2013) ‘Biological corridors and connectivity’, Key Topics in Conservation Biology 2. (Wiley Online Books), pp. 384– 4 0 4 . d o i : h t t p s : / / d o i . o r g / 10.1002/9781118520178.ch21. Davies, T. J. et al. (2008) ‘Phylogenetic trees and the future of mammalian biodiversity’, Proceedings of the National Academy of Sciences of the United States of America, 105(SUPPL. 1), pp. 11556–11563. doi: 10.1073/pnas.0801917105. De Angelo, C. et al. (2013) ‘Understanding species persistence for defining conservation actions: A management landscape for jaguars in the Jaguar extinction Atlantic Forest’, Biological Conservation, 159, pp. 422–433. doi: https://doi.org/10.1016/ j.biocon.2012.12.021. De la Torre, J. A. et al. (2018) ‘The jaguar’s spots are darker than they appear: assessing the global conservation status of the jaguar Panthera onca’, Oryx. 2017/01/23. Cambridge University Press, 52(2), pp. 300–315. doi: DOI: 10.1017/S0030605316001046. Eva, H. D. et al. (2004) ‘A land cover map of South America’, Global Change Biology. John Wiley & Sons, Ltd, 10(5), pp. 731–744. doi: https://doi.org/10.1111/ j.1529-8817.2003.00774.x. FAO. (2014) The future of food and agriculture: trends and challenges, The future of food and agriculture: trends and challenges. Available online. FAO and UNEP. 2020. The State of the World’s Forests 2020. Forests, biodiversity and people. Rome. doi.org/10.4060/ca8642en pp. 83 Fisher, D. O. and Blomberg, S. P. (2011) ‘Correlates of rediscovery and the detectability of extinction in mammals’, Proceedings of the Royal Society B: Biological Sciences. Royal Society, 278(1708), pp. 1090–1097. doi: 10.1098/rspb.2010.1579. Foster, R. J. et al. (2020) ‘Jaguar (Panthera onca) density and tenure in a critical biological corridor’, Journal of Mammalogy, 101(6), pp. 1622–1637. doi: 10.1093/jmammal/gyaa134. Foster, R. J. et al. (2020) ‘Jaguar (Panthera onca) density and tenure in a critical biological corridor’, Journal of Mammalogy, 101(6), pp. 1622–1637. doi: 10.1093/jmammal/gyaa134. Frankham, R. (2005) ‘Genetics and extinction’, Biological Conservation, 126(2), pp. 131–140. doi: https://doi.org/10.1016/ j.biocon.2005.05.002. Gilroy, J. J., Ordiz, A. and Bischof, R. (2015) ‘Carnivore coexistence: Value the wilderness’, Science, 347(6220), pp. 382 LP – 382. doi: 10.1126/science.347.6220.382-a. Gonzalez-Borrajo, N., López-Bao, J. V. and Palomares, F. (2017) ‘Spatial ecology of jaguars, pumas, and ocelots: a review of the state of knowledge’, Mammal Review. John Wiley & Sons, Ltd, 47(1), pp. 62–75. doi: https://doi.org/10.1111/mam.12081. González-Maya, J. et al. (2016) ‘Ecology and conservation of jaguars in Mexico: state of knowledge and future challenges’, in, pp. 273– 289 Hansen, M. C. et al. (2013) ‘High-Resolution Global Maps of 21st-Century Forest Cover Change’, Science, 342(6160), pp. 850 LP – 853. doi: 10.1126/science.1244693. Harcourt, A. H., Parks, S. A. and Woodroffe, R. (2001) ‘Human density as an influence on species/area relationships: double jeopardy for small African reserves?’, Biodiversity & Conservation, 10(6), pp. 1011–1026. doi: 10.1023/A:1016680327755. Hijmans, R (2020) ‘raster: Geographic Data Analysis and Modeling. R package version 3.3-13.’ Hijmans, R (2021). terra: Spatial Data Analysis. R package version 1.4-22. https://CRAN.Rproject.org/package=terra Hoffmann, M. et al. (2010) ‘The Impact of Conservation on the Status of the World’s Vertebrates’, Science, 330(6010), pp. 1503 LP – 1509. doi: 10.1126/science.1194442. Hoogesteijn, R. et al. (2016) ‘Conservacion de jaguares (Panthera onca) fuera de áreas protegidas: turismo de observacion de jaguares en propiedades privadas del Pantanal, Brasil / Jaguar (Panthera onca) Observation tourism in private properties of the Brazilian Pantanal’, in, pp. 259–274. Inskip, C. and Zimmermann, A. (2009) ‘Humanfelid conflict: A review of patterns and priorities worldwide’, ORYX, 43(1), pp. 18– 34. doi: 10.1017/S003060530899030X. Jędrzejewski, W. et al. (2017) ‘Human-jaguar conflicts and the relative importance of retaliatory killing and hunting for jaguar (Panthera onca) populations in Venezuela’, Biological Conservation, 209, pp. 524–532. doi: https://doi.org/10.1016/ j.biocon.2017.03.025. Jędrzejewski, W. et al. (2018) ‘Estimating large carnivore populations at global scale based on spatial predictions of density and distribution – Application to the jaguar (Panthera onca)’, PLOS ONE. Public Library of Science, 13(3), p. e0194719. Available at: https://doi.org/ 10.1371/journal.pone.0194719. Laliberte, A. S. and Ripple, W. J. (2004) ‘Range Contractions of North American Carnivores and Ungulates’, BioScience, 54(2), pp. 123– 1 3 8 . d o i : 10.1641/0006-3568(2004)054[0123:RCONAC ]2.0.CO;2. Lawton, J. H. (1993) ‘Range, population abundance and conservation’, Trends in 81 Chapter 3 Ecology & Evolution, 8(11), pp. 409–413. doi: https://doi.org/10.1016/0169-5347(93)90043O. Lucas, P. M., González-Suárez, M. and Revilla, E. (2016) ‘Toward multifactorial null models of range contraction in terrestrial vertebrates’, Ecography. John Wiley & Sons, Ltd, 39(11), pp. 1100–1108. doi: https://doi.org/10.1111/ ecog.01819. Mace, G. M., Possingham, H. P. and LeaderWilliams, N. (2006) ‘Prioritizing choices in conservation’, Key Topics in Conservation Biology, pp. 17–34.Marchini, S. and Macdonald, D. W. (2012) ‘Predicting ranchers’ intention to kill jaguars: Case studies in Amazonia and Pantanal’, Biological Conservation, 147(1), pp. 213–221. doi: https://doi.org/10.1016/j.biocon.2012.01.002. MacKinney, M. L. (2001) ‘Role of human population size in raising bird and mammal threat among nations’, Animal Conservation. John Wiley & Sons, Ltd, 4(1), pp. 45–57. doi: https://doi.org/10.1017/S1367943001001056. Martínez-Meyer, E. et al. (2013) ‘Ecological niche structure and rangewide abundance patterns of species’, Biology Letters. Royal Society, 9(1), p. 20120637. doi: 10.1098/ rsbl.2012.0637.McKinney, Medellín, R.A., J. Antonio de la Torre, Heliot Zarza, Cuauhtémoc Chávez, G. C. (2016) El jaguar en el siglo XXI. La perspectiva continental.Rodrigues, A. S. L. et al. (2006) ‘The value of the IUCN Red List for conservation.’, Trends in ecology & evolution (Personal edition), 21(2), pp. 71–6. doi: 10.1016/j.tree.2005.10.010. Medina, L. K. (2010) ‘When government targets “the state”: Transnational NGO government and the state in belize’, Political and Legal Anthropology Review, 33(2), pp. 245–263. doi: 10.1111/j.1555-2934.2010.01113.x. Mora, J. et al. (2016) ‘V. Estado de Conservacion del jaguar (Panthera onca) en Honduras.’, in, pp. 137–167. Nassar, P. (2013) ‘Instituto Nacional De Pesquisas Da Amazônia-Inpa Programa De PósGraduação Em Gestão De Áreas Protegidas Da Amazônia (Ppg-Mpgap-Inpa) Viabilidade Do Ecoturismo Científico Com Onça-Pintada Na Reserva Mamirauá, Amazônia’. Nassar, P., Ramalho, E. and Silveira, R. (2013) ‘Economic and market viability of scientific ecotourism related to the jaguar in várzea area in central amazonia’, Scientific Magazine UAKARI, 9, pp. 21–32. doi: 10.31420/ uakari.v9i2.145. ONU - Organização das Nações Unidas (2019) World population prospects 2019, Department of Economic and Social Affairs. World Population Prospects 2019. Available at: http:// www.ncbi.nlm.nih.gov/pubmed/12283219. Pacifici, M. et al. (2020) ‘Global correlates of range contractions and expansions in terrestrial mammals’, Nature Communications, 11(1), p. 2840. doi: 10.1038/s41467-020-1684-w. Parks, S. A. and Harcourt, A. H. (2002) ‘Reserve Size, Local Human Density, and Mammalian Extinctions in U.S. Protected Areas’, Conservation Biology. John Wiley & Sons, Ltd, 16(3), pp. 800–808. doi: https://doi.org/ 10.1046/j.1523-1739.2002.00288.x. Paviolo, A. et al. (2008) ‘Jaguar Panthera onca population decline in the Upper Paraná Atlantic Forest of Argentina and Brazil’, Oryx. 2008/10/14. Cambridge University Press, 42(4), pp. 554–561. doi: DOI: 10.1017/ S0030605308000641. Paviolo, A. et al. (2016) ‘A biodiversity hotspot losing its top predator: The challenge of jaguar conservation in the Atlantic Forest of South America’, Scientific Reports, 6(1), p. 37147. doi: 10.1038/srep37147. Payan, E., Rabinowitz, A. and Zeller, K. (2010) ‘A range-wide model of landscape connectivity and conservation for the jaguar, Panthera onca’, Biological Conservation, 143, pp. 939– 945. doi: 10.1016/j.biocon.2010.01.002. Pearson, R. G. et al. (2014) ‘Life history and spatial traits predict extinction risk due to climate change’, Nature Climate Change, 4(3), pp. 217–221. doi: 10.1038/nclimate2113. Pebesma E (2018). ‘Simple Features for R: Standardized Support for Spatial Vector Data’%The R Journal,%10(1), 439–446. doi:%10.32614/RJ-2018-009,%https://doi.org/ 10.32614/RJ-2018-009. Pedersen, A.B., Jones, K.E., Nunn, C.L., Altizer, S. (2007) ‘Infectious Diseases and Extinction Risk in Wild Mammals. ’ Conservation Biology, 21: 1269-1279.%https://doi.org/ 10.1111/j.1523-1739.2007.00776.x Petracca, L. S. et al. (2014) ‘Agricultural encroachment and lack of enforcement threaten connectivity of range-wide jaguar (Panthera onca) corridor’, Journal for Nature 82 Jaguar extinction Conservation, 22(5), pp. 436–444. doi: https:// doi.org/10.1016/j.jnc.2014.04.002. Pomara, L. Y. et al. (2014) ‘Demographic consequences of climate change and land cover help explain a history of extirpations and range contraction in a declining snake species’, Global Change Biology. John Wiley & Sons, Ltd, 20(7), pp. 2087–2099. doi: https://doi.org/10.1111/gcb.12510. Purvis, A. et al. (2000) ‘Predicting extinction risk in declining species’, Proceedings of the Royal Society of London. Series B: Biological Sciences. Royal Society, 267(1456), pp. 1947– 1952. doi: 10.1098/rspb.2000.1234. Quigley, H., Foster, R., Petracca, L., Payan, E., Salom, R. & Harmsen, B. (2017) Panthera onca (errata version published in 2018). The IUCN Red List of Threatened Species 2017. Robinson, T. P. et al. (2014) ‘Mapping the Global Distribution of Livestock’, PLOS ONE. Public Library of Science, 9(5), p. e96084. Available at: https://doi.org/10.1371/ journal.pone.0096084. Safi, K. and Pettorelli, N. (2010) ‘Phylogenetic, spatial and environmental components of extinction risk in carnivores’, Global Ecology and Biogeography, 19(3). doi: 10.1111/ j.1466-8238.2010.00523.x. Sanderson, E. W. et al. (2002) ‘Planning to Save a Species: the Jaguar as a Model’, Conservation Biology. John Wiley & Sons, Ltd, 16(1), pp. 58–72. doi: https://doi.org/10.1046/ j.1523-1739.2002.00352.x. Schipper, J. et al. (2008) ‘The Status of the World's Land and Marine Mammals: Diversity, Threat, and Knowledge’, Science, 322(5899), pp. 225 LP – 230. doi: 10.1126/ science.1165115. Seymour, K. L. (1989) ‘Panthera onca’, Mammalian Species, (340), pp. 1–9. doi: 10.2307/3504096. Swank, W. G. and Teer, J. G. (1987) ‘Status of the Jaguar, 1987’, Oryx, 23(January), pp. 14–21. Available at: Swank_&_Teer_1987_Status_of_the_Jaguar.p df. Terborg, J. et al. (1999) ‘The Role of Top Carnivores in Regulating Terrestrial Ecosystems’, in Wild Earth. Torres, N. M. et al. (2012) ‘Can species distribution modelling provide estimates of population densities? A case study with jaguars in the Neotropics’, Diversity and Distributions. John Wiley & Sons, Ltd, 18(6), pp. 615–627. doi: https://doi.org/10.1111/ j.1472-4642.2012.00892.x. Tortato, F. R. and Izzo, T. J. (2017) ‘Advances and barriers to the development of jaguar-tourism in the Brazilian Pantanal’, Perspectives in Ecology and Conservation, 15(1), pp. 61–63. doi: https://doi.org/10.1016/ j.pecon.2017.02.003. Tortato, F. R. et al. (2017) ‘The numbers of the beast: Valuation of jaguar (Panthera onca) tourism and cattle depredation in the Brazilian Pantanal’, Global Ecology and Conservation, 11, pp. 106–114. doi: https://doi.org/10.1016/ j.gecco.2017.05.003. Travers, H., Clements, T. and Milner-Gulland, E. J. (2016) ‘Predicting responses to conservation interventions through scenarios: A Cambodian case study’, Biological Conservation, 204, pp. 403–410. doi: https://doi.org/10.1016/ j.biocon.2016.10.040. Venables, W. N. & R. (2002) ‘Modern Applied Statistics with S. Fourth Edition’. Springer, New York. ISBN 0-387-95457-0. Villalva, P. and Palomares, F. (2019) ‘Perceptions and livestock predation by felids in extensive cattle ranching areas of two Bolivian ecoregions’, European Journal of Wildlife Research, 65(3). doi: 10.1007/ s10344-019-1272-8. Woodroffe, R. and Ginsberg, J. (1998) ‘Edge Effects and the Extinction of Populations Inside Protected Areas’, Science, 280(5372), pp. 2126–2128. doi: 10.1126/ science.280.5372.2126. Woodroffe, R. (2000) ‘Predators and people: using human densities to interpret declines of large carnivores’, Animal Conservation. John Wiley & Sons, Ltd, 3(2), pp. 165–173. doi: https:// doi.org/10.1111/j.1469-1795.2000.tb00241.x. Yackulic, C. B., Sanderson, E. W. and Uriarte, M. (2011) ‘Anthropogenic and environmental drivers of modern range loss in large mammals.’, Proceedings of the National Academy of Sciences of the United States of America. National Academy of Sciences, 108(10), pp. 4024–9. doi: 10.1073/ pnas.1015097108. Zanin, M. and Neves, B. dos S. (2019) ‘Current felid (Carnivora: Felidae) distribution, spatial bias, and occurrence predictability: testing the reliability of a global dataset for 83 Chapter 3 macroecological studies’, Acta Oecologica, 101, p. 103488. doi: https://doi.org/10.1016/ j.actao.2019.103488. Zar, J. H. (1999) Biological statistics. Prentice Hall, New Jersey. Zeller, K. (2007) Jaguars in the new millennium data set update: The state of the jaguar in 2006, Wildlife Conservation Society, Bronx, N e w Yo r k . Av a i l a b l e at : ht t p s : / / www.panthera.org/sites/default/files/ JaguarsintheNewMillenniumDataSetUpdate_0 .pdf. Zeller, K. A. et al. (2013) ‘The Jaguar corridor initiative: A range-wide conservation strategy’, Molecular Population Genetics, Evolutionary Biology and Biological Conservation of Neotropical Carnivores, pp. 629–657. Zimmermann, A., Walpole, M. J. and LeaderWilliams, N. (2005) ‘Cattle ranchers’ attitudes to conflicts with jaguar Panthera onca in the Pantanal of Brazil’, Oryx. 2005/09/28. Cambridge University Press, 39(4), pp. 406– 412. doi: DOI: 10.1017/S0030605305000992. Zimmermann, A. and Macdonald, D. W. (2010) ‘Jaguars , Livestock , and People in Brazil : Realities and Perceptions Behind The Conflict’. USDA National Wildlife Research Center - Staff Publications. 918. 84 Jaguar extinction 85 Suplementary material 1 Figure1. Jaguar current and historic distribution range (1) used to generate the binomial response variable. Cattle density (2), agriculture cover (3), human density (4) and distance to the border (5) are the independent variables used for modelling. Chapter 3 Suplementary material 1 Distance to the border Agriculture Human density Cattle density Distance to the border 1.000 Agriculture -0.14 1.000 Human density -0.11 0.09 1.000 Cattle density 0.03 0.57 0.09 1.000 Table1. Correlation matrix of the variables in jaguar historic range. 86 Jaguar extinction Supplementary material 2 Table 1. Probability of extinction for Jaguar Conservation Units (JCUs) based on the mean values of the model contained in each JCU. Extinct area corresponds to the percentage of the area that is out of the current range. Name, country, size and estimated population size based on expert opinions (Zeller, 2007) are also shown. Id JCU Name Country Size (km2) Population size Probability of extinction Extinct area (%) 1 Sierra Madre Occidental Mexico 11724 <50 - 100 2 Sierra Madre Occidental Mexico 13606 50-100 0.61 0 3 Sierra Madre Oriental Mexico 21570 50-100 0.59 0 4 Sierra Madre Oriental Mexico 1384 50-100 0.63 0 5 Nayarit Mexico 29329 >500 0.77 0 6 Veracruz Mexico 9547 100-200 0.67 46 7 Yucatán Mexico 63349 >500 0.55 1 8 Belize Belize 8561 >500 0.57 0 9 Guatemala Guatemala 5334 50-100 0.75 4 10 Honduras North Honduras 2824 - 0.80 0 11 Honduras South Honduras 2206 50-100 - 100 12 Reservas Tawatika, Rio Plátano Honduras - Nicaragua 36435 200-500 0.55 6 13 Costa Rica NW Costa Rica 5302 50-100 0.78 5 14 Reserva Indo Maiz Nicaragua 7018 100-200 0.60 0 15 P.N La Amistad Costa Rica - Panama 14041 100-200 0.64 6 16 P.N Corcovado Costa Rica 1801 100-200 0.60 0 17 Panamá S. Panama - Colombia 69729 100-200 0.64 1 18 Paramillo Colombia 8203 200-500 0.85 3 19 Cienaga Mogua Colombia 6887 50-100 0.81 0 20 Rio Magdalena Colombia 2619 <50 0.90 0 87 Chapter 3 Id JCU Name Country Size (km2) Population size Probability of extinction Extinct area (%) 21 Rio Apure Venezuela 1422 <50 0.87 0 22 Arismendi Venezuela 973 50-100 0.90 12 23 El Baul Venezuela 7095 50-100 0.82 2 24 Laguna de Tacarigua Venezuela 1616 <50 0.64 0 25 Tinigua Colombia 20948 50-100 0.73 0 26 Tuparro Colombia 22614 50-100 0.37 0 27 Jaua-Sarisariñama Venezuela 18654 200-500 0.40 0 28 Canaima Venezuela - Brazil - Guyana 81458 - 0.38 0 29 Guyana Guyana 3716 - 0.39 0 30 Surinam Surinam 16183 - 0.39 0 31 Guayana Francesa French Guyana - Brazil 71437 >500 0.45 0 32 Parima Tapirapeco Venezuela 56414 200-500 0.22 0 33 Ilha de Maraca Brazil 2807 <50 0.27 0 34 Mache-Chindul Ecuador 1350 50-100 - 100 35 Cotachi Cayapas Ecuador 2450 50-100 0.59 5 36 Manglares Cayapas Ecuador 210 <50 - 100 37 Awa Territory Ecuador 896 <50 0.64 0 38 Sumaco Galeras Ecuador 5748 200-500 0.59 17 39 Yasuni Ecuador 47372 >500 0.49 0 40 Yaigoje Apaporis Colombia 4126 <50 0.20 0 41 Pico da Nebuna Brazil 11705 50-100 0.13 0 42 Condor Kutuku Ecuador 7505 200-500 0.60 0 43 Jaú Brazil 37991 50-100 0.15 0 44 Rio Marañón Peru 22830 200-500 0.35 0 88 Jaguar extinction Id JCU Name Country Size (km2) Population size Probability of extinction Extinct area (%) 45 Area de Conservación Comunal Peru 9215 50-100 0.24 0 46 Amacayacu Peru 7504 50-100 0.19 0 47 Amazonas w Brazil 67306 50-100 0.27 0 48 P.N Amazonia Brazil 38246 >500 0.26 0 49 Paragominas Brazil 20122 200-500 0.65 0 50 Parauapebas Brazil 32045 >500 0.65 6 51 Maranhao Brazil 6225 100-200 0.77 0 52 Manu Peru 43219 200-500 0.48 0 53 Tambopata-Madidi Peru - Bolivia 58658 >500 0.47 0 54 Rondonia Brazil 53918 - 0.59 21 55 P.N Campos Amazonicos Brazil 99175 100-200 0.58 1 56 P.N Rio Novo Brazil 36627 50-100 0.52 3 57 P.N Xingu Brazil 66555 200-500 0.60 46 58 Araguacema Brazil 13323 - 0.71 30 59 Araguaia Brazil 29728 50-100 0.60 0 60 Paranaiba Brazil 45214 200-500 0.61 4 61 Serra das Confusoes Brazil 7163 <50 0.52 2 62 Isiboro Securé Bolivia 20198 - 0.52 0 63 Noel Kempff Bolivia 68229 >500 0.51 0 64 Chiapada dos Viadeiros Brazil 10243 <50 0.61 0 65 Grande Sertao Brazil 4618 <50 0.61 0 66 Cavernas do Peruaçu Brazil 2427 <50 0.69 0 67 Carrasco-Amboró Bolivia 8653 100-200 0.53 0 68 Kaa Iya Bolivia - Paraguay 88852 >500 0.49 0 Id JCU Name Country Size (km2) Population size Probability of extinction Extinct area (%) 89 Chapter 4 INTRODUCTION Anthropogenic activities have led to decline of high-ecologically-demanding species such as top predator carnivores, in name of tolerant species (Olden et al., 2018). When top predators are removed, mesopredator release may unbalance the underneath trophic levels (Soulé et al, 1988, Crooks & Soulé, 1999). This top-down regulation usually depends on a complex network of intraguild relationships, surpassing the overall idea of one single regulator and revealing how carnivore co-existence works on ecosystem maintenance (Monterroso, 2020). Despite the evidences of carnivores declining (Ripple et al., 2014), little is known about how anthropogenic activities impact intraguild relationships, which could either force or reduce co-existence by impeding, facilitating or unbalancing niche partitioning among species (Seveque et al, 2020). Habitat loss and human-wildlife conflict are the main anthropogenic activities causing carnivores population declines and distribution constraining (Di Minin et al., 2016), besides a primary conservation concern in the Neotropics where agriculture and livestock farms are still expanding. There are over 400 million cattle heads in the Neotropics (FAO, 2021), occupying more than 60% of its surface (Gilbert et al., 2018). This widespread industry favours a global conflict scenario, firstly by reducing the species native habitats and after by making big cats to share territory with livestock (eg. Zimmermann et al., 2005, Amit et al., 2013, Michalski et al., 2006, Llanos et al., 2020, Villalva and Palomares, 2019). Protection of territory is the main instrument to minimise these threats (Coetzee et al., 2014; Gray et al., 2016) and has increased in recent decades through the promotion of protected areas (Watson et al., 2014). However, the effectiveness of this measure is controversial (Rodrigues et al., 2004; Craige et al., 2010; Laurence et al., 2012), especially for top carnivores whose populations require vast territories (Woodroffe and Ginsberg, 2008). In summary, jaguars are dominant over pumas (Haines, 2006) and ocelots are subordinate to both species due to its smaller size (Elbroch and Kusler, 2018, Wallach et al., 2015). Jaguars and pumas exhibit different tolerance to habitat disturbance, in which jaguars seems to be prevalent in pristine habitats (Di Bitetti et al., 2010), so more sensitive to habitat degradation (i.e. Torres et al., 2017, but see Foster et al., 2010), 96 Felid interactions while pumas can inhabit more degraded landscapes and be excluded from pristine habitat by jaguars (Sollman et al., 2002). In fact, in numerous regions where jaguars have been locally extinct, pumas still persist (Nielsen et al., 2015, Quigley et al., 2017). Ocelots also tolerates some level of habitat disturbance, being observed in degraded landscapes where they coexist with pumas (Lima Massara et al., 2018) and even where pumas are absent (Boron et al., 2020). Moreover, ocelots are not deeply threatened by human-livestock conflict that largely affect jaguars and pumas (Inskip & Zimmermann, 2009). Despite this knowledge, we have a poor understanding of the mechanisms behind global co-existence patterns of these species (Linnell & Strand, 2000, Elmhagen & Rushton, 2007, Periquet et al., 2015). The lack of global comprehensive research is certainly due to the demanding effort of methods used for investigating coexistence (usually camera trapping or radio tracking) that restrains their use at broader scales. Therefore, a compiled dataset from broad research programs such as the Neocarnivore database (Nagys-reis et al., 2020) enables the investigation of distribution-scale patterns and other global data of socioeconomics and environmental variables allow to put human disturbance into account. Our goal was to examine whether anthropogenic disturbance affects the coexistence patterns of Neotropical top predators, investigating how potential human persecution, habitat cover and quality, and livestock affect the intragremial relationships of jaguar, puma, and ocelot at a continental scale. Based on the coexistence theory, which assumes a competitive dominance hierarchy correlated with body size (Palomares & Caro, 1999, Donadio & Burskirk, 2006, Linnell & Strand, 2000), we predicted that pumas should avoid spatial overlap with jaguars while ocelot should avoid spatial overlap with jaguars and pumas. However, this dominance hierarchy could change accordingly to anthropogenic influence in human dominated areas, increasing occurrences of puma and ocelot due to their higher environmental tolerance. Therefore, jaguars’ occurrence should be conditioned to larger extension of habitat amount than pumas and ocelots. Livestock ranching, in its turn, should reduce jaguar and puma’s presence by promoting population depletion due to livestock predation and, consequently, retaliatory persecution. Protected areas could stabilise 97 Chapter 4 human-mediated pressure by ensuring the offer of a native habitat and a refugee to retaliatory hunting. METHODS 2.1. Data gathering Neocarnivores database is a compilation of georeferenced data on carnivore distribution, obtained from studies conducted from 1818 to 2018, representing the largest data set on neotropical wild carnivores to date (Nagy-reis et al., 2020). From this database, we selected occurrences of jaguars, pumas, and ocelots since 2000, in order to discard the unwanted effect from old entries (such as the presence of jaguars in the Atlantic forest in the past century where the species is currently almost extirpated). To avoid data from casual encounters, we only kept data obtained by systematic sampling where methodology was equally capable of detecting the three species, i.e. camera trapping and genetic fecal sampling. We completed this dataset with 586 entries corresponding to new genetic samples, mostly from pumas and ocelots, collected by the authors. Using the species occurrence locations, a presence-absence layer was generated for each species at the resolution scale of 0.0833º decimal degree, around 100 km2 in the equator, which is the average adult jaguar male home range (GonzalezBorrajo et al., 2017). This resolution scale was kept as spatial resolution for all variables. Forest cover was included as a proxy for environmental quality outstanding as a principal component for the habitat of jaguar (Hoogesteijn and Mondolfi, 1992, De Angelo et al., 2013), puma (Regolin et al., 2017), and ocelot (Regolin et al., 2017). For this purpose, we used the Global Land Cover 2015 (Buchhorn et al., 2020), a 100 mresolution layer classifying land uses and cover at 23 classes, from which forest cover areas were selected to obtain the coverage percentage in a resampling process at 100 km2. We used the Global Human Influence Index as a proxy for human presence that may affect felids through direct persecution (i.e hunting) or through avoidance behaviours to humans. This dataset is created from nine global variables related to human population pressure, infrastructure, and access (WCS, 2005). Protected areas were obtained from the World Database on Protected Areas (UNEP - WCMC, 2015) 98 Felid interactions corresponding to the IUCN categories. Any pixel containing a portion of a protected area took the value of its category while unprotected territories took value zero. Livestock density accounts for the combined effect of habitat loss and retaliatory hunting related to livestock practices. We defined livestock pressure using the Gridded Livestock of the World (GLW) model (Gilbert et al., 2018), which compiles and harmonises the distribution of livestock cadaster data at a subnational level. In the ultimate version of this model (GLW3 areal-weighted), the distribution of livestock is free from the influence of any other variables, contrary to previous models based on RF algorithm that used a multiple layer approach (Robinson, 2014). This enhancement provides greater utility for ecological modelling by eliminating the effect of confounding variables, which could interfere in the interpretation of our study. The layers used in our study were goat and sheep densities, which were combined in a single layer (hereinafter goat + sheep), and bovine cattle density (hereinafter cattle). Goat and sheep densities were combined to avoid overloading the model with too many variables, keeping the interpretation simple, based on the difference in livestock size and husbandry practices. 2.2. Data Analysis Structural equation modelling (SEM) is a collection of procedures whereby complex hypotheses, particularly those involving networks of path relations, are evaluated against multivariate data (Bollen 1989, Grace 2006). This multi-equational method of data analysis is capable of representing a wide array of complex hypotheses about how system components interrelate based on the analysis of covariance relations. SEM enables a broad set of scientific questions providing countless possibilities for quantitative research. In this sense, univariate models are suitable for studying individual processes or responses while SEMs are appropriate for studying multiple processes by controlling the behaviour of the system, allowing to quantify the relative importance of the interaction between species (or processes), as well as the cascade effects (Grace, 2010). However, the use of SEM is only suitable for testing causal models grounded in solid understanding of the system (Dunham & Niewiarowski, 1996, Hatcher, 1994), making essential the robust construction of a priori model at 99 Chapter 4 light of specific questions (Hoyle et al., 2012). Our SEM was built based on solid knowledge about felid species, their ecosystems, and the main human disturbance affecting them. A basic causal model has endogenous and exogenous variables. A variable is endogenous if its value is determined or influenced by one or more independent variables (excluding itself) while an exogenous variable is a factor whose value is independent from other variables in the system. Intragremial relations among species (endogenous variables) constitute the core of the proposed model, once the effect of one species is investigated on others, which can be under influence of environmental variables (i.e. forest cover, human index, livestock, and protected areas, i.e. exogenous variables). After testing correlation between exogenous variables, eliminating the risk of unpredicted causal relations among them (Supplementary Material S1), we adopted the follow equations as predictor models composing SEM: (1) jaguar occurrence, considering the effect of protected areas, human index, cattle, goat+sheep, and forest cover; (2) puma occurrence, considering the effect of jaguar occurrence, protected areas, human index, cattle, goat+sheep, and forest cover; (3) ocelot occurrence, considering jaguar and puma occurrences, protected areas, human index, and forest cover. Livestock rarely compose ocelot’s prey base, so cattle and goat+sheep variables were excluded from ocelot equation. We standardised the variable set to compare the effects among predictors. The magnitude of standardised coefficients indicates the degree to which the predictor directly affects the criterion variable if the rest of variables remain constant. This means that the variation of one standard deviation in the presence of jaguars would suppose the resulting coefficient value standard deviations in the presence of pumas if the rest of the model remain unaltered. Mediation (indirect effect) was determined by the coefficient product of the indirect structural path, which added to the direct effect corresponded to the total effect. We performed statistical analysis using lavaan package (Rosseel, 2012), and transform data with raster (Hijmans & Van Etten, 2012), sp (Bivand et al., 2013) and 100 Felid interactions spatial (Venables & Ripple, 2002) packages in R version 4.0.3 (R Development Core Team, 2021). RESULTS We obtained 1459 independent presence-absence raster units distributed over the Neotropical region (Figure 1), covering most of the countries. No data were available for Suriname, Guyana, Nicaragua, Honduras, and El Salvador indicating lack of recent research in these countries. Ocelots (n = 1075) and pumas (n = 1005) occurrences covered the larger extent of study area, 73% and 68% respectively; jaguars (n = 506) were rarer, appearing in 35 % of the studied area. We found the three species 101 Figure 1. Study area map. Green dots represent a buffer zone of 10 km on rasterised presence data of jaguars, pumas and ocelots extracted from Neocarnivores dataset (Nagys Reis et al., 2020). Zoom spanning Yucatan and Pantanal regions as symbolical examples showing pixels of 100 Km2. Colours indicate the number of species detected in each pixel. Chapter 4 co-existing in 21% of the studied area, two species coexisted in 35%, and only one species in 44 %. SEM analysis showed a good fit (CFI = 1, RMSEA = 0.00, SRMR = 0.004), so the predictions of endogenous variables were robust. Jaguar presence was negatively related to human index, cattle, and goat + sheep and positively related to forest cover; pumas were positively influenced by jaguar and goat + sheep density; and ocelot, in its turn, were positively affected by forest cover, but showed negative association with puma presence (Table 1, Fig. 2). 102 Species Variable Estimate ± SE z value p value Jaguar Cattle -0.061 ± 0.028 -2.163 0.031 Goat + Sheep -0.060 ± 0.025 -2.415 0.016 Forest 0.233 ± 0.029 7.985 < 0.001 Human index -0.220 ± 0.027 -8.117 < 0.001 Protected areas -0.024 ± 0.025 -0.970 0.332 Puma Jaguar 0.129 ± 0.029 4.476 < 0.001 Cattle -0.001 ± 0.031 -0.013 0.990 Goat + Sheep 0.080 ± 0.027 2.959 0.003 Forest -0.025 ± 0.033 -0.763 0.445 Human index 0.037 ± 0.030 1.224 0.221 Protected areas -0.004 ± 0.027 -0.160 0.873 Ocelot Jaguar 0.050 ± 0.028 1.805 0.071 Puma -0.263 ± 0.025 -10.387 < 0.001 Forest 0.118 ± 0.029 4.123 < 0.001 Human index -0.004 ± 0.029 -0.130 0.897 Protected areas -0.001 ± 0.026 -0.051 0.959 Table 1. Summary of SEM regression model for endogenous variables explaining the relations among variables. Estimate ± Standard error, z value and pvalue are shown. Felid interactions Contrary to our expectations, we observed a positive effect of jaguar on puma and no effect on ocelot, instead the negative effect predicted to both species. The prediction of the negative effect of puma on ocelot was corroborated and it is twice bigger than the effect of jaguar on puma. Total effect (direct + indirect effect) did not show any substantial change from direct effects in the relationships among species (Supplementary Material S2). The effect of exogenous variables also shows the complexity of felids distribution and co-occurrence pattern: while both livestock variables adversely affected jaguar presence, cattle had no effect on puma and goat + sheep favoured them; human index negatively affect jaguar, having no effect on puma and ocelot; forest cover do not affect puma, but affect twice more jaguar than ocelot; and protected areas show no association to any feline presence. 103 Figure 2. Structural Equation Model (SEM) showing the relations among variables. Solid lines denote significant relations, dotted lines denote non significant relations. Width of each line is proportional to the relative strength of relationship and the number indicates the path coeficiente values. Significance of relations are indicated by asterisks (pvalue < 0.001 ***, pvalue < 0.01 ** and pvalue < 0.05 *). Chapter 4 DISCUSSION Anthropogenic activity’s effects on interspecific relationships Anthropogenic activities negatively affected jaguar, unlike pumas and ocelots, suggesting a behaviour alteration of individuals in areas under human presence or livestock production. There are strong evidence of human activities influencing species behaviours, causing nocturnity (Gaynor et al., 2018), movement restrictions (Tucker et al., 2018), and anti-predatory behaviours (Tambiling et al., 2015; Laundré et al., 2001; Berger, 1999). On jaguar, anti-predatory behaviours may operate through direct avoidance human activities (Llaneza et al., 2012, Oriol-Cotterill et al., 2015); if not, by changes on prey-base behaviour (Taber et al., 2006, Scognamillo et al., 2006, Gutierrez et al., 2017, Azevedo et al., 2018). The unexpected positive effect of jaguars on pumas may relay with local segregation – in any spatial, temporal, or trophic niche dimensions – promoting species coexistence. For jaguars and pumas, spatial segregation seems common at fine scale (Palomares et al., 2016), sometimes directly promoted by habitat selection (Foster et al., 2010, Sollman et al., 2012, Foster et al., 2013, Porfirio et al., 2017) and sometimes by synchromism with different prey base (Taber et al., 2006, Scognamillo et al., 2006, Gutierrez et al., 2017, Azevedo et al., 2018). Temporal segregation, in its turn, may rule when habitat availability is restricted by anthropogenic activities (Di Bitetti al., 2010, De la Torre et al., 2017), supporting our overall idea of human disturbances changing the structural rules of felids coexistence. The lack of evidences of anthropogenic activities affecting pumas, in addition to the positive effect of sheep + goat, supports the established idea of pumas being more tolerant to habitat disturbance than jaguars (Sollmann et al., 2012, De Angelo et al., 2011, but see Foster et al., 2010). Under these circumstances, we can assume that human disturbance is inverting the hierarchical competition dominance between large felids, favouring puma and increasing its position in the assemblage hierarchy. The widespread facilitation of pumas would cause ocelots’ distribution contraction, despite being less affected by anthropogenic disturbance. Therefore, pumas seem the key species mediating the unbalancing on community structure through their higher environmental tolerance and competitive dominance over other less tolerant felids. 104 Felid interactions Protected areas and habitat loss Jaguars and pumas are among the large mammals with greatest distribution contraction (Morrison et al., 2007) and, at least jaguars, with few viable populations in the long term (Zanin et al., 2015). Despite being a widespread conservation strategy worldwide (CBD, 2011), protected areas are not sufficient to increase felids occurrence probability, so breaking the distribution contraction. This revels the inefficiency of protected areas, despite being by insufficient extension (Zanin et al., 2021) or incapacity of guaranteeing a hunting-free refugee for felids population. For jaguars and ocelots, the extension of forest cover more than the level of protection is interfering in their presence. Therefore, to assure felids conservation, we need integrative strategies that consider not only protected areas, but any fragment and the anthropogenic matrix in the landscape where protected areas are embedded. In summary, our conclusions remark the need to focus conservation efforts at a landscape level, where issues are more complex and the potential for conflict is also higher. Livestock The asymmetrical effect of livestock on big cats reveals that the link of livestock conflict and big cats is more complex than assumed (Inskip & Zimmermann, 2009). Pumas’ higher tolerance to habitat disturbance allow greater withstand in areas under livestock pressure; moreover, goat and sheep can provide an increment of prey availability. This “minor” livestock farming, which positively influence pumas, is mostly located in dry regions (such as the Brazilian Caatinga, the Guajira - Barranquilla xerophytic scrub region in Colombia, or the Pre-Andean slopes) where prey availability is scarce due to the scant primary production. These areas also correspond to suboptimal habitats for jaguars where, with the exception of small patches in the Caatinga (Morato et al., 2014), the species have been locally extirpated (Quigley et al., 2017). Persecution for livestock predation is potentially asymmetric as well, revealed by previous findings about the ranchers’ inability to discriminate between felid species (Cavalcanti et al., 2010; Villalva & Palomares, 2019), making any livestock predation overturning on jaguars by being the most familiar species. Therefore, puma presence 105 Chapter 4 Biotropica, 29(2), 204–213. doi: https:// doi.org/10.1111/j.1744-7429.1997.tb00025.x Tambling, C. J., Minnie, L., Meyer, J., Freeman, E. W., Santymire, R. M., Adendorff, J., & Kerley, G. I. H. (2015). Temporal shifts in activity of prey following large predator reintroductions. Behavioral Ecology and Sociobiology, 69(7), 1153–1161. Retrieved from http://www.jstor.org/stable/43599868 Tortato, F. R., Izzo, T. J., Hoogesteijn, R., & Peres, C. A. (2017). The numbers of the beast: Valuation of jaguar (Panthera onca) tourism and cattle depredation in the Brazilian Pantanal. Global Ecology and Conservation, 11, 106–114. doi: https://doi.org/10.1016/ j.gecco.2017.05.003 Tucker, M. A., Böhning-Gaese, K., Fagan, W. F., Fryxell, J. M., Van Moorter, B., Alberts, S. C., … Mueller, T. (2018). Moving in the Anthropocene: Global reductions in terrestrial mammalian movements. Science, 359(6374), 466 LP – 469. doi: 10.1126/science.aam9712 Van Eeden, L. M., Eklund, A., Miller, J. R. B., López-Bao, J. V., Chapron, G., Cejtin, M. R., … Treves, A. (2018). Carnivore conservation needs evidence-based livestock protection. PLOS Biology, 16(9), e2005577. Retrieved from https://doi.org/10.1371/ journal.pbio.2005577 Venables, W. N. & Ripley, B. D. (2002). Modern Applied Statistics with S. Fourth Edition. Springer, New York. doi: ISBN 0-387-95457-0 Villalva, P., & Palomares, F. (2019). Perceptions and livestock predation by felids in extensive cattle ranching areas of two Bolivian ecoregions. European Journal of Wildlife Research, 65(3). doi: 10.1007/ s10344-019-1272-8 Wallach, A. D., Izhaki, I., Toms, J. D., Ripple, W. J., & Shanas, U. (2015). What is an apex predator? Oikos, 124(11), 1453–1461. doi: https://doi.org/10.1111/oik.01977 Watson, J. E. M., Dudley, N., Segan, D. B., & Hockings, M. (2014). The performance and potential of protected areas. Nature, 515(7525), 67–73. doi: 10.1038/nature13947 WCS, W. C. S.-, & University, C. for I. E. S. I. N.- C.-C. (2005). Last of the Wild Project, Version 2, 2005 (LWP-2): Global Human Influence Index (HII) Dataset (Geographic). Palisades, NY: NASA Socioeconomic Data and Applications Center (SEDAC). Retrieved from https://doi.org/10.7927/H4BP00QC Woodroffe, R., & Ginsberg, J. (1998). Edge Effects and the Extinction of Populations Inside Protected Areas. Science, 280(5372), 2126–2128. doi: 10.1126/ science.280.5372.2126 Zanin, M., Palomares, F., & Brito, D. (2015). The jaguar’s patches: Viability of jaguar populations in fragmented landscapes. Journal for Nature Conservation, 23, 90–97. doi: https://doi.org/10.1016/j.jnc.2014.06.003 Zanin, M., Palomares, F., & Mangabeira Albernaz, A. L. (2021). Effects of climate change on the distribution of felids: mapping biogeographic patterns and establishing conservation priorities. Biodiversity and Conservation, 30(5), 1375–1394. doi: 10.1007/ s10531-021-02147-1 Zimmermann, a., Walpole, M. J., & LeaderWilliams, N. (2005). Cattle ranchers’ attitudes to conflicts with jaguar Panthera onca in the Pantanal of Brazil. (August 2015). doi: 10.1017/S0030605305000992 Zimmermann, A., & Macdonald, D. W. (2010). USDA National Wildlife Research Center - Jaguars , Livestock , and People in Brazil : Realities and Perceptions Behind The Conflict. (February). 112 Felid interactions Suplemmentary material 113 Jaguar Puma Ocelot Cattle Goat + Sheep Protected Areas Forest Human Index Jaguar 1.00 Puma 0.09 1.00 Ocelot 0.07 -0.26 1.00 Cattle -0.24 0.00 -0.09 1.00 Goat + Sheep -0.16 0.07 -0.07 0.05 1.00 Protected Areas 0.09 -0.01 0.03 -0.25 -0.08 1.00 Forest 0.36 -0.01 0.14 -0.49 -0.20 0.19 1.00 Human Index -0.34 0.02 -0.08 0.33 0.22 -0.23 -0.42 1.00 S1. Correlation matrix (Pearson) showing the relation among each variable on the Structured Equation Model. Chapter 4 Effects on Jaguars Effects on Pumas Effects on Ocelots Variables IE TE IE TE IE TE Jaguar - - - 0.130 -0.034 0.016 Cattle - -0.061 -0.009 -0.010 0.000 - Goat + Sheep - -0.060 -0.078 0.002 -0.021 - Protected Areas - -0.024 -0.003 -0.043 0.010 0.000 Forest - 0.233 0.029 0.005 0.007 0.125 Human Index - 0.220 -0.028 0.009 -0.009 -0.014 S2. Indirect (IE) and total effects (TE) for Structural Equation Model in Fig.1. Values of direct effect can be found on Table 2. 114 Perceptions and livestock predation #####Chapter 5! Perceptions and livestock predation by felids in extensive cattle ranching areas of two Bolivian ecoregion ! European Journal of Wildlife Research 2019 117 Chapter 5 118 Perceptions and livestock predation Human-carnivore conflicts arise as one of the most urgent carnivore conservation issues worldwide. Jaguars (Panthera onca) and pumas (Puma concolor) coexist with livestock in much of their range and have been historically blamed for livestock predation. At present, livestock landscapes are increasing enlarging the conflict scenario. Knowledge of perception of locals is vital to understand the context in which the conflict emerges and data of predation on livestock become an effective tool to plunge into it. We assess local perceptions about felids and identify the bases of the conflict in Bolivian pantanal and dry forest ecoregion using livestock predation data. We interviewed local ranchers and crossed information with the governmental livestock database achieving a comprehensive study on the conflict, based on descriptive statistics for local perceptions and generalized linear mixed models for analyzing cattle predation by felids. The conflict appeared to be widespread since most ranchers suffered predation on cattle, especially in the pantanal. Data suggested low percentage of annual loss (1.8 %) with the exception of some punctual cases that may magnify the generalized negative perception towards felids. Factors related to cattle husbandry explained better felid predation on livestock rather than the habitat quality of the ranch. We recommend local administrations to recover livestock predation data and highlight the importance that husbandry practices may have to reduce cattle losses by felids. Key words: Apex felids; Livestock predation; Extensive ranching; Average modeling 119 Abstract Chapter 5 Introduction Protected areas appear to be insufficient to protect large carnivores from extinction (Woodroffe, 1998), as healthy populations of these species require vast areas to persist due to their wide home ranges, territoriality and prey necessities (Ripple et al., 2014). Conservation policies usually rely on protected areas often creating a fragmented landscape with islands of protected territories. However conservation of big carnivores require more integrated conservation measures that ensure wide connected areas (Rabinowitz and Zeller, 2010; Sanderson et al., 2002). In this context, human dominated landscapes are considered a more realistic scenario for carnivore conservation (Athreya et al., 2013; Chapron et al., 2014; but see Gilroy et al., 2015). Extensive cattle ranching is globally practiced and is characterized by maintaining vast areas of moderately modified landscapes and very low human density where freegrazing domestic animals coexist with wildlife. However , such a situation recurrently, leads to a conflict (Treves and Karanth, 2003) as carnivores may damage livestock, compete for prey or induce risky situations for people (Kleiven et al., 2004; Woodroffe et al., 2005). Generally it resulted in hunting of the carnivore species causing the decline of natural populations and ultimately affecting the whole ecosystem by topdown cascade processes (Estes et al., 2011). Felids are largely affected by human-carnivore conflict becoming an urgent conservation issue for these species (Woodroffe et al., 2005). Severity of the conflict tends to increase with body mass (Inskip and Zimmermann, 2009) reason why jaguars (Panthera onca) and pumas (Puma concolor), the biggest cats in South America, are positioned in a conflict hotspot. Both species have been intensively reported to kill livestock throughout South and Central America where they co-occur with humans in close proximity (Rabinowitz, 1986; Quigley and Crawshaw, 1992; Hoogesteijn, 1993; Mazolli et al., 2002; Saenz and Carrilo, 2002; Conforti and Azevedo, 2003; Polisar et al., 2003; Crawshaw, 2004; Graham et al., 2004; Azevedo & Murray, 2007; RosasRosas et al., 2008; Cavalcanti and Gese, 2010; Kissling et al., 2009). In a systematic review of human-felid conflict worldwide, Inskip & Zimmermann (2009) reported that the proportion of livestock loss by jaguars may ranged 0.3-2.3% while pumas sometimes were attributed for a higher number of medium-size domestic animals (up to 120 Perceptions and livestock predation 26% of sheep in 15 ranches). On the other hand, even more than 50% of total prey killed by jaguar may correspond to cattle, whilst it may comprise 15-43% of puma diet. So that, estimates on cattle predation by jaguars and pumas are highly variable among study sites and also did not provide consistent data on which species is more likely to prey upon cattle probably as the conflict strongly depend on the ecological context. As an extreme case, there are areas where pumas scarcely consume domestic species but jaguar did supplement diet with cattle (Forester et al., 2010), and vice versa (Scognamillo et al., 2003). Numerous potential factors influence the cost-benefit balance that encourage jaguar and puma predation on livestock including innate and learned behavior, health and status of individuals (Mondolfi and Hoogesteijn, 1986; Quigley and Crawshaw, 1992), space and resources competition among jaguars and pumas (Scognamillo et al., 2003; Forested et al., 2010), abundance and distribution of natural preys (Polisar et al., 2003, Zanin et al., 2015) and cattle management practices (Michalski et al., 2006, Palmeira et al., 2008). Some possibilities to reduce frequency of felid attacks on livestock arise by a correct management of human-felid coexisting areas. Frequency of cattle predation was inversely related to availability and vulnerability of natural prey and directly related to availability and vulnerability of livestock (Polisar et al. 2003). Ranches that possessed abundant and diverse natural prey experienced fewer felid problems (Mondolfi and Hoogesteijn, 1986; Hoogestein et al., 1993). Meanwhile, increasing availability of domestic preys in areas with large cattle herd size positively affected predation rates (Michalski et al., 2006, Zarco-González, 2013). Aditionally, measures that ensure monitoring and carefully control of livestock with special attention to maternities which are more affected by predation is strongly recomended (Palmeira et al., 2008). Identifying main factors that affect cattle predation by jaguars and pumas may contribute to reduce the frequency of large cat depredation on livestock which ultimately may go a long way towards maintaining cat populations. However, an effective approach to felid conservation should also consider perception of people inhabiting conflicting territories. Studies on local perception of the conflict reveals that killing domestic animals by carnivores is not the only or most important reason why people kill them (Cavalcanti et al., 2010). People demonstrates a 121 Chapter 5 128 Pantanal Dry forest People in the ranch Mean (range) / % /categories Mean (range) / % /categories Age 40.6 (18-68) 41.5 (19-72) Education level 2/15/13/6 4/6/20/19 Owner/cowboy 10/12 27/18 Perceptions Felids threat humans 54%% 43%% Felids threat cattle 91%% 93%% Cite cattle as prey 84%% 87%% Puma population trend 11/14/11 8/12/25 Jaguar population trend 10/8/13 7/11/29 Puma footprint presence 94%% 51%% Jaguar footprint presence 88%% 73%% Identify footprints 46%% 33%% Identify predated carcass 20%% 15%% Cattle loss 16.9%% 8.5%% Attitude Response to cattle predation 8/6/21 9/4/33 Felid extirpation 65%% 33%% Ranch characteristics Ranch size (ha) 4,064 (100-14,000) 2,062 (120-7,000) Cattle holdings 1,429 (2-8,000) 1,337 (200 - 6,000) Cattle density* 0.37 (0.2 - 0.4) 1.19 (0.08 - 4) Human presence 0/11/16 1/9/20 Habitat features Forest cover (%) 0.39 (0.1 - 0.9) 0.6 (0.1 - 0.9) npp (g/m2) 14,744 (11,754 - 18,989) 16,211 (12,000 - 21,200) Table 1. Summary of local perception and attitude to felids, ranch characteristics and habitat features for each ecoregion surveyed. Mean, % of respondents or categories are shown for the different variables. Education level is reported as the number of respondents with no studies/primary/secondary/superior studies. Population trend of both felids is shown as increasing/stable/declining. Response to cattle predation correspond to ranchers that had no response/try to banish the felid/kill the felid. Only positive presence is shown regarding to percentage data (i.e. presence of footprints, identifying predated carcass and footprints). Felid extirpation is referred to the percentage of ranchers that consider felid extinction as positive.Human presence is reported as numeric categories corresponding to no buildings/manager house/manager and cowboy house. Perceptions and livestock predation Perception and attitude towards jaguars and pumas Most respondents considered felids a threat to cattle (93 %) and almost half as well as a threat to humans. Regarding to perception on felid preys, cattle was cited as the most common prey for both predators (91%) while rarely wild animals were referred as exclusive preys (9%). Perceptions varied between ecoregion (pantanal vs. Dry forest) in many aspects. Ranchers in the Dry forest considered puma population in decline, while in the pantanal they affirmed that population is stable (X2 = 5.8983, P = 0.05239). Decline of jaguars was a general perception and did not varied between ecoregion (X2 = 3.2228, P = 0.1996). Regarding to perception of footprint presence, in the pantanal pumas were significantly (θ=0.3998, P = 0.1677) more encountered than in the Dry forest, unlike jaguars prints that were common in both regions. However when we ask ranchers to identify felid footprints there were no difference among region (X2 = 0.975, P = 0.3234). Only 38 % of all ranchers identified correctly both species even though most of them (74 %) believed that they knew. Interestingly, only 17 % of all ranchers were able to distinguish between carcasses predated by either felid species. 129 No studies Primary Secondary Superior 20 15 10 5 Bad Good Figure 2. Relation between ranchers education level and opinion about local extinction of big felids. Dark bars represent ranchers against local extinction while light bars represent ranchers in favor of local extinction. Fisher’s exact test: P < 0.001. Chapter 5 Attending to attitude of ranchers towards felids, poaching appeared to be the first response to cattle predation (70%) regardless of ecoregion (X2 = 2.4552, P = 0.293). However, felid extirpation was more accepted by ranchers in the pantanal than those in the Dry forest (X2 = 5.9837, P = 0.01444). The opinion about local extinction also varied between education level (Fig. 2) and number of losses. Those ranchers that stayed in school for longer considered negative the effects of local extinction (Fisher’s test: P = 0.0003), while those with higher number of loss were in favor of it (X2 = 5.7799, P = 0.0162). Most respondents in favor of local extinction cited the reduction of cattle predation as main motivation. Livestock predation and factors affecting it Overall we recorded 913 cattle killed by big felids during the previous year in all the surveyed ranches. In general, most of them (77%) loss a moderate amount of 130 0.2 0.4 0.6 0.8 1 Ranch Size Cattle holdings Cattle density Human presence Forest cover Attitude Cumulative AIC weight Figure 3. Cumulative AIC weight of covariates used in cattle loss models. Ranch size was the most influential variable, followed by other ranch features related to the abundance of domestic prey. Perceptions and livestock predation cattle (12.2 ± 2.8 cattle heads loss per ranch). Losses were significantly greater in the pantanal than in Dry forest (average 16.9 and 8.5 cattle heads per ranch, respectively; X2 = 10.634, P = 0.004908). Felid predation represented 1.8 % of the total holdings (range= 0-26%) in surveyed ranches for the whole study area. In many ranches (47 %) felid predation represented less than 1 % of total holdings. Model averaging evidenced the importance of ranch characteristics over ranchers attitudes, human presence and habitat features (Fig. 3). As expected, ranch size showed the greater significance to explain cattle loss, while cattle holdings and cattle density (interaction between both previous variables) were also of great importance, followed by human presence, forest cover and attitude, that was the less explicative in the set of tested variables. Regarding to the model, parameters showed a positive relation between cattle loss and ranch size alike cattle holdings (Table 2), meaning that larger ranches and greater holdings resulted in greater cattle loss. Cattle density and human presence relation was negative, prompting that greater cattle density and more residents in the ranch resulted in fewer cattle loss. 131 Variables Estimate S.E. P (Intercept) 0.996 0.950 ns Ranch size 0.001 0.000 *** Cattle holdings 0.001 0.000 ** Cattle density 0.000 0.000 *** Human presence -0.988 0.432 * Response 0.229 0.206 ns Forest cover 0.908 0.770 ns Table 2. Parameters estimates for the generalized linear mixed-effects model explaining cattle loss by felids. Coefficient estimate, standard error (S.E.) and test z significance (P) are shown. ns P < 0.1; *P < 0.05; **P < 0.01; ***P < 0.001. Chapter 5 Meanwhile, probability of loss (Fig. 4) supported the previous result showing a marked slope of ranch size that quickly reached the upper limit. Ranches over 4000 ha had a 90 % probability of losing cattle.Livestock holdings curve showed a softer slope, while cattle density increase reduced the probability of loss. Discussion Large carnivores interfere with human business becoming a serious problem for long term viability of their populations due to hunting (Woodroffe and Redpath, 2015). Extensive ranching areas preserve part of the natural landscape allowing apex carnivores to coexist with humans (Hoogesteijn and Hoogesteijn, 2011), thus to find strategies to minimize the conflict between ranchers and carnivores becomes vital to achieve the goal of conservation of large predators in exploited landscapes. Our results suggest that felid-human conflict is widespread in the study region. Most ranchers assured that felids are a threat to cattle and many considered them a 132 Probability of loss 0 Ranch size / Livestock holdings / Cattle density Ranch size Livestock holdings Cattle density 0 0.2 0.4 0.6 0.8 1 3500/2000/1 7000/4000/2 10500/6000/3 14000/8000/4 Figure 4. Probability of cattle loss in relation to three main variables of the best model. Dots represent empiric data (0 = no loss, 1 = at least 1 loss). Black dots and continuous line represent ranch size, grey dots and long dashed line represent livestock holdings and empty dots and short dashed line represent cattle density. Perceptions and livestock predation threat to humans what evidence the basis of the conflict. Moreover, the perception that both apex felids diet is mostly cattle illustrates the generalized negative attitude towards them. Previous poaching data in the area supported our findings by reporting a total of 347 jaguars and 230 pumas killed in 85 ranches in four years (Arispe, 2008; Arispe et al., 2006). On average, annual predated cattle was 1.8 % in the total sampled ranches, which is consistent with the reported 0.1 - 3% range for predation rate of felids worldwide (Jackson and Nowell, 1996). Maximum and minimum values corresponded to studies on livestock predation by jaguars and pumas, with annual losses of 0.1% in Los Llanos, Venezuela (Polisar et al., 2003) and 3.0 % in Sao Paulo, Brazil (Palmeira et al., 2008). In a similar Pantanal area predation rate was 2.3 % (Zimmerman et al., 2005). We find marked differences among ecoregions not only in cattle loss but in many other aspects. Regarding to the characteristics of respondents, the level of studies was notably higher in the Dry forest than in the Pantanal. Also, education level appeared to be important for the perception of felids (Fig. 2), according to results of Cavalcanti et al. (2010) that found a negative correlation between attitude towards jaguars and the number of years attending school. We found that level of studies was notably higher in the Dry forest compared to the Pantanal, which explains the better attitude towards big cats in the Dry forest when they were asked about felid local extinction. The fact that ranchers in the Pantanal were more often pro-extinction of felids highlight the importance of this ecoregion for conservation efforts. On the other hand, perceived felids population trend revealed the decline of pumas in the Dry forest, while jaguar population remain stable. Supporting this result, presence of puma footprints were very low in the Dry forest (51%) compared to those cited in the Pantanal (94%), but we must be aware of the lack of credibility of this data due to the inability to differentiate tracks of both species. Also, predation impact on livestock markedly varied among ecoregion. Cattle loss was greater in the Pantanal than in the dry forest what can be explained by differences in cattle husbandry characteristics between regions. From this scope, ranch sizes were in average six times larger in the Pantanal than those in the Dry forest (see Table 1), factor that is pointed as the most important effect on cattle loss in this and previos studies (Zimmermann et al., 2005, 133 Chapter 5 Zarco-González, 2013). Despite type of cattle system had not been able to be tested due to scarce data, entire dataset from the livestock cadaster showed that the Pantanal is basically characterized by a cattle breeding system while the dry forest by a breedingfattening system where proportion of calves/cow was 0.37 and 0.47 respectively (DSA, 2011). This information may have important implications for the higher number of cattle losses found in the Pantanal besides ranch size. In general, the percentage of predated cattle might be low as compared with cattle deaths due to disease or drought (Azevedo and Murray, 2007, but see Tortato et al. 2015). Even though we did not measure other death causes, we found exceptional high predation rates in some of the surveyed ranches (up to 26 % of total holdings) which were personally visited by the authors to corroborate veracity of these data. These particular cases occurred in ranches located in the edge with agricultural areas, where poaching levels on wild preys may be very high and hence densities of wild prey extremely low. Usually ranchers of the area are aware of these dramatic cases which may skew their perception of the conflict and magnify the magnitud of the conflict (Chavez and Gese, 2006; Chavez et al., 2005; Conover, 2002; Sillero-Zubiri et al., 2007). Modeling of cattle loss evidenced ranch size as the variable that contributed the most to explain cattle predation by felids. Similar effect was found in studies of jaguar and puma predation (Zimmermann et al., 2005, Zarco-González, 2013), even in other carnivores as suggested wolf predation on cattle in northern US (Treves and Karanth, 2003). Large ranches often harbor more disperse cattle which becomes certainly more difficult to handle compared to small ranches where cattle is more controlled. Moreover, large ranches may shelter higher amount of felids that may facilitate the encounter of cattle with eater cats. The amount of livestock (stockholdings) also affected predation on cattle. Greater amounts of cattle increase availability of potential prey which may promote opportunistic predation on cattle . Moreover high amounts of cattle in extensive systems turn the herd more difficult to handle. Both features, wide ranches and numerous herds usually lead to unmanaged cattle which has been highlighted as a key piece for cattle predation (Bagchi and Mishra, 2006; Hoogesteijn and Hoogesteijn, 2011). Cattle density (interaction between ranch size and total 134 Perceptions and livestock predation stockholdings) also standed out as an important factor explaining cattle predation, but with an inverse relationship (Fig. 4). High herd densities, especially when high proportion of adults exist, may increase the probability of defensive behavior (Tortato et al., 2015). Also, it may be related again to cattle management practices with high densities being readily handled in comparison to low densities that usually maintain cattle more disperse (Jori et al., 2006, Zarco-González, 2013). Human presence in the ranch also contributes to explain cattle losses as showed the model averaging ranking (see Fig.3), although the predictor for cattle loss was only marginally significant (see Table 2). We expected a positive relation between human presence and predation, due to more people in the ranch will reduce natural prey thus increasing predation on cattle (Foster et al., 2014). Instead of this, we found a negative relationship. This unexpected result suggest that more people in the ranch tend to have a dissuasive effect on probability of felid attacks by direct persecution or avoidance behavior of cats. Besides, more people may assure more accurate cattle handling that may reduce the number of loss as already mentioned (Palmeira et al., 2008). Interestingly, the response of ranchers to predation events apparently did not affect cattle loss (Fig. 3 and Table 2). This is at odds with the popular belief that the extirpation of the individual predator could lead to the extirpation of the conflict. A more satisfactory way to minimize predation on cattle should be focused on improving management practices that might allow a real reduction of carnivore damage. However, the response of killing the felid before cattle predation is very common in both ecoregions, suggesting that the conflict is widespread and deeply rooted in the culture of the people inhabiting the whole study area. We believe that extensive ranching can become an ally for big cat conservation. In order to mitigate the conflict we suggest two main recommendations: 1) To focus mitigation efforts in those ranches that have actually show high predation rates. This will locally attenuate the number of losses while, simultaneously, may help to start changing the general negative perception about predators, and 2) to encourage further investigation in determining which cattle management practices may reduce cattle predation. For this purpose a greater dataset than used in this study would be desirable. Thus, we suggest a systematic data recovery that could be easily handled by the 135 Chapter 5 institutions in charge of cattle cadaster. The new dataset should contain the actual large amount of ranch characteristics plus data on livestock predation in each ranch. Analysis of these data will help to locate conflict hotspots where focus strategies to conflict mitigation and to determine those management practices that affect cattle predation. In conclusion, we believe that there is an urgent need in further investigation to implement those improved cattle husbandry practices that might mitigate predation by felids to contribute to the conservation and recovery of big felids in extensive systems. This approach should be along with the implementation of appropriate education programs of inhabitants in rural areas to build a positive baseline of conservation values within the ranching community, increasing their knowledge on those ranch features and measures that may reduce cattle loss consequently. We also encourage to rise awareness about the apex carnivores intrinsic value and role in ecosystems. Acknowledgements We would like to thank Norka Rocha and Ricardo Barberí for their help undertaking questionnaires and the Anmi San Matías office (SERNAP) for the permission to conduct this research in protected areas. We also thank Miguel Camacho and Miguel Clavero for helpful comments on the first drafts and Eva Moracho and two anonymous referees that have improve notably this paper. Financial support of this work was provided by the Spanish Cooperation Agency (Agencia Española de Cooperación Internacional para el Desarrollo) through a MAEC-AECID personal grant for the author. 136 References Anderson, D.R. (2008). Model Based Inference in the Life Sciences: A Primer on Evidence. Springer New York. Arispe, R. (2008). Looking for opportunities to protect the jaguar (Panthera onca) in Santa Cruz, Bolivia. Jaguar News. Arispe, R., Rumiz, D., Venegas, C., Noss, A. (2006). El conflicto de la depredación de ganado por jaguar (Panthera Onca) en Santa Cruz, Bolivia. In conference: Conflito Homem-Animal. Ilheus. Athreya, V., Odden, M., Linnell, J.D.C., Krishnaswamy, J., Karanth, U. (2013). Big Cats in Our Backyards: Persistence of Large Carnivores in a Human Dominated Landscape in India. PLoS One 8, 2–9. Azevedo, F., Murray, D.L. (2007). Evaluation of potential factors predisposing livestock to predation by jaguars. J. Wildl. Manage. 71, 2379. Bagchi, S., Mishra, C. (2006). Living with large carnivores: Predation on livestock by the snow leopard (Uncia uncia). J. Zool. 268, 217–224. Burnham, K.P., Anderson, D.R. (2004). Model Selection and Multimodel Inference. Ecol. Modell. 172, 96–97. Burnham, K.P., Anderson, D.R. (2002). Model selection and multimodel inference: a practical information-theoretic approach. 2nd ed. New York, Springer-Verlag. Carvalho, E.A.R., Zarco-González, M., MonroyVilchis, O., Morato, R. (2015) Modelling the risk of livestock depredation by jaguar in the Transamazon Highway, Brazil. Basic Appl. Ecol. 16, 413-419. Cavalcanti, S.M.C., Gese, E.M. (2010). Kill rates and predation patterns of jaguars (Panthera onca) in the southern Pantanal, Brazil. J. Mammal. 91, 722–736. Cavalcanti, S.M.C., Marchini, S., Zimmermann, A., Gese, E.M., Macdonald, D.W. (2010). Jaguars, livestock, and people in Brazil: realities and perceptions behind the conflict. Biol. Conserv. wild felids 383–402. Chapron, G., Kaczensky, P., Linnell, J.D.C., Arx, M. von, Huber, D., Andrén, H., López-Bao, J.V., Adamec, M., Álvares, F., Anders, O., Balčiauskas, L., Balys, V., Bedő, P., Bego, F., Blanco, J.C., Breitenmoser, U., Henrik, B., Bufka, L., Raimonda, B., Paolo, C., Dutsov, A., Engleder, T., Fuxjäger, C., Claudio, G., Holmala, K., Hoxha, B., Iliopoulos, Y., Ionescu, O., Jeremić, J., Jerina, K., Kluth, G., Knauer, F., Kojola, I., Ivan, K., Miha, K., Jakub, K., Kunovac, S., Josip, K., Kutal, M., Liberg, O., Majić, A., Männil, P., Manz, R., Marboutin, E., Francesca, M., Melovski, D., Mersini, K., Yorgos, M., W., M.R., Nowak, S., Odden, J., Ozolins, J., Palomero, G., Paunović, M., Persson, J., Potočnik, H., Quenette, P.-Y., Rauer, G., Reinhardt, I., Rigg, R., Ryser, A., Salvatori, V., Tomaž, S., Stojanov, A., Swenson, J.E., László, S., Aleksandër, T., Váňa, E.T.-S.M., Veeroja, R., Wabakken, P., Wölfl, M., Wölfl, S., Zimmermann, F., Zlatanova, D., Boitani, L. (2014). Recovery of large carnivores in Europe´s modern human-dominated landscapes. Science (80-. ). 346, 1514–1517. Chavez, a S., Gese, E.M. (2006). Landscape use and movements of wolves in relation to livestock in a wildland-agriculture matrix. J. Wildl. Manage. 70, 1079–1086. Chavez, A., Gese, E.M., Krannich, R.S. (2005). Attitudes of rural landowners toward wolves in northwestern Minnesota. Wildl. Soc. Bull. 33, 517–527. Conforti, V., Azevedo, F. (2003) Local perceptions of jaguars (Panthera onca) and pumas (Puma concolor) in the Iguaçu National Park area, south Brazil. Bio Cons, 111, 215–221. Conover, M. (2002). Resolving Human-Wildlife Conflicts: The Science of Wildlife Damage Management. Lewis Publishers. Crawshaw, P. (2004) Depredation of domestic animals by large cats in Brazil. Hum Dim Wild, 9, 329–330. Dirección de Sanidad Agroalimentaria, 2011. Catastro Ganadero 2011. Secretaría de Desarrollo Productivo, Gobierno Departamental Santa Cruz de la Sierra. Estes, J. a, Terborgh, J., Brashares, J.S., Power, M.E., Berger, J., Bond, W.J., Carpenter, S.R., Essington, T.E., Holt, R.D., Jackson, J.B.C., Marquis, R.J., Oksanen, L., Oksanen, T., Paine, R.T., Pikitch, E.K., Ripple, W.J., Sandin, S. a, Scheffer, M., Schoener, T.W., Shurin, J.B., Sinclair, A.R.E., Soulé, M.E., Virtanen, R., Wardle, Chapter 6 Illegal wildlife trade is a major threat to biodiversity worldwide (Maxwell et al. 2016; Milner-Gulland 2018) and heavily contributes to the Anthropocene extinction crisis (Nijman 2010). Intense hunting pressure can result in Anthropogenic Allee effect on natural populations (Stephens et al. 1999) where severe reduction of population density ends up collapsing the biological effectiveness of a species. Continuous harvesting creates feedback loops in which a species become rare, and therefore increase its value, then stimulating further harvesting (Courchamp et al. 2006). This overexploitation may lead a species into an extinction vortex that is accelerated because of economic interests (Clavero 2016). There is a high-value illegal wildlife market worth an estimated USD 8-10 billion per year that is recognized as one of the largest global illicit businesses of transnational criminal organizations (Haken 2011). Mammals stand out as the dominant exploited group (Rosen & Smith 2010; Mc. Clenachan et al. 2016) primarily for consumption as bushmeat (Riple et al. 2016), but also there is an international luxury market of non-perishable products for medicinal (i.e. the traditional Chinese medicine) and decorative purposes that represents a high extinction threat to wild mammal populations. The disproportionate impact on these species is due not only to the extraordinary value that some animal parts reach, but also because they can be stockpiled and distributed globally, so that poaching occurs continuously (Webb 2016). Large-sized mammals are affected the most by illegal trade due to both intensive international demand and their intrinsic sensibility related to great spatial requirements and low population growth rates (McClenachan et al. 2016). Some widely known examples of transcontinental trade include tiger products (Mills & Jackson 1994; Nowell & Ling 2007), rhinoceros horns (Milner-Gulland & Leader-Williams 1992), and elephant ivory (Milner-Gulland & Beddington 1993; Scriber 2014). During the past half century, the intensively trade of these species bounded for Asian market has driven them to the actual endangered (tiger and Asian elephant) and critically endangered status (black rhino) (IUCN 2018). Main causes of their ongoing extinction process are not just the excessive reduction of population density, but also particularities of each species trade such as poaching preference on males that result in a strong biased sex ratio seriously affecting reproduction effectiveness, as it has already occurred in Asian elephant populations. The ecological impact of illegal trade on these species warns of 144 Selling jaguars as tiger the potential threat that similar species may face in a near future. Despite of the huge efforts to control and mitigate illegal trade, the extremely profitable luxury market explains why it is still currently in force. Traditional Chinese medicine is a high-demand medical practice widely used in China and surrounding countries that has strong influence on illegal wildlife trade. Animal-based preparations may include parts of endangered species such as tiger bones. Tiger parts are highly quoted in traditional Chinese medicine in relation to disparate medical uses (Ellis 2005). For example, McClenachan et al. (2016) reported that the tiger penis worth up to USD 470,000 kg-1 and a complete individual could reach USD 350,000, placing the species among the most expensive animals (along with black rhino) traded. Demand for tiger-derived products has never slowed down; even it may increase as traditional Chinese medicine is becoming more popular and recognized worldwide (OMS 2013). Both are the causes -high value and great demandof tiger overexploitation that has rapidly accelerated its decline over the last decades. Protective regulation that Chinese government approved in 1993 banning any trade of tiger parts has not contributed to effectively safeguard the species (EIA 2013). Actually, sale of tiger products has notably increased in the following years, especially since 2010, being China and Vietnam the main countries of origin and destination (Nowell 2014). Furthermore, a rapidly growing of captive facilities for tiger breeding (with nearly 260 establishments) has emerged during the last decade in Asia (especially China and Thailand) and South Africa (EIA 2017), where the number of captive tigers (7000-8000 individuals) almost doubles that of natural populations (3000-4000 individuals). Far from protecting the species, tiger farms have been involved in illegal trade of their parts and its derivatives (Nowell 2014). Lack of regulation to release captive-bred tigers has recently motivated (October 2018) Chinese government to reverse the ban, so today the trade of captive tigers (and rhinos) for medicinal, scientific and educational purposes has been legalized. This measure is likely to reactivate tiger illegal trade in the short term (but see Conrad 2012 or Biggs et al. 2013) due to the greater ease of laundering in the legal market. It should be noted the difficulty to differentiate between captive and wild animals especially when they are sold as derivatives (Stoner 2014), complicating analysis and monitoring. 145 Chapter 6 Insatiable demand of tiger parts under the new context of legal tiger trade entails that poaching threatens not just the world’s few remaining wild tigers but also other big cat species worldwide which may be marketed to consumers as tiger (Williams et al. 2015). Diverse felid species have been used as tiger substitutes in traditional medicine preparations being sold as tiger fakes; otherwise the market value drops drastically compared to authentic tiger products (Nowell 2014) making it affordable to another economic sector of society. Other Asian cats such as snow leopards, leopards, cloud leopards and Asian lions are also threatened as a result of the illegal trade of their parts (EIA 2017). Out from Asia, the transcontinental trade of African lions represents a clear and well-documented example that sheds light on tiger substitution by other felids (Koshoo 1997). Lion bones have intensively been trafficked from South Africa to East-Southeast Asia under a government agreement that allowed its use for wine preparations that traditionally containing tiger bones (Bahuer 2016). Since then, legal trade of lion bones has increased progressively, with about 6000 skeletons being sent during the last decade (Williams et al. 2017). This fact gives support to Koshoo´s predictions (1997), which declared "It is also clear that when tiger is decimated, the next target will be lion, followed by leopard and all other felines from Asia and Africa". However, it is possible that twenty years ago, the belief that strong demand for tiger parts could also threaten American big cat species was still unrealistic. Exponential growth of Chinese population generates an unprecedented demand for natural resources (see Bai et al. 2018) whose exploitation could trigger cascading effects through entire ecosystems even at tens of thousands of kilometres (LuqueLarena et al. 2018). In addition, Chinese expansion throughout the world creates an opportunity to explore new resources while preserves its deeply rooted culture and traditions in new places. South America has become the destination for a wide expatriate Chinese population as a consequence of the great economic investment of this country in the continent during the last 10 years (Perez 2017). Their economic activities, usually operating under low environmental and social standards, have raised the concern of critical observers for nature conservation. In this context jaguars (Panthera onca) are placed at the spotlight due to the new business opportunity of selling jaguar as tiger substitute in the traditional Chinese medicine market. This new 146 Selling jaguars as tiger tiger substitute has still not been well documented, probably because it is a recent phenomenon. First evidences of the emerging transcontinental jaguar trade comes from Bolivia and Surinam (Bale 2017; Bale 2018) and have been related to establishment of Chinese companies in the region. Specifically, Bolivian authorities seized more than 380 jaguar teeth between 2013-2016 and 11 different shipments containing up to 185 big cat fangs were sent from Santa Cruz de la Sierra destined to China in just one year (Rumiz & Rivero 2018). We believe that these evidences represent just the tip of the iceberg of an emerging phenomenon that seriously threatens America's largest cats. Lack of detection in other countries can be explained due to the difficulty of uncovering it, especially when dealing with derivatives. High price that tiger derivatives reach for traditional medicine (exceeding indeed the actual price of gold, diamonds or cocaine) (Biggs et al. 2013; McClenachan et al. 2016) is a claim for drug trafficking mafias whose decades of expertise in illegal business make detection even more complicated. The emergent threat of illegal trade on jaguars for Chinese medical purposes is added to the historic hunting pressure that the species has suffered through the extensive conflict with cattle. These two factors may drive jaguars to an overexploitation situation that can lead it to an extinction vortex in a short time. Increasing knowledge on the impact of illegal trade and developing intelligent-aid management tools are especially important in early stages of the process and ultimately could allow that jaguar and other felid species follow a different path from that of the tiger. Great advances in molecular biology have been developed for big cats, providing new insights to diverse biological questions that would otherwise be unapproachable. New techniques that have been positively tested are actually readyfor-service for its application to illegal trade (Shina 2017) i.e. for the challenging tiger trade in China or the emergent increase of jaguar trade in Bolivia. A genetic approach has already been strongly recommended by many involved actors in illegal wildlife trade such as Interpol that proposed the development of intelligent platforms to allow the prosecution, investigation and forensic capacity of local authorities (Interpol 2014). Furthermore, the economic limitation has been exceeded allowing wide application opportunities at an affordable economic cost. As example, a very useful approach has already been proposed for field management of Rhinos illegal trade from South Africa 147 Chapter 6 to China (Biggs et al. 2013), based on horn harvesting and genetic profiling to control the legal market. Also genetic approaches have been recently implemented to aid management of Bengal tiger trade by geo-locating seizure samples in Nepal (Karmacharya et al. 2018). The new Chinese law should promote establishing a regulatory non-governmental institution to differentiate between captive and wild tigers based on DNA profiling, but we also must pay special attention on the impact that the new regulation have on long-distance populations of other big cats. We underscore the need for a close genetic approach also in those Latin American countries where the conflict has already been detected, allowing information and management tools to enforce intelligence-aid measures at the yet initial stage. 148 References Bai, Z., Lee, M.R.F., Ma, L., Ledgard, S., Oenema, O., Velthof, G.L. et al. (2018) Global environmental costs of China’s thirst for milk. Global Change Biology, 24, 2198–2211. Bale, R. (2017) Where jaguars are ‘killed to order’ for the illegal trade. National Geograhic.https:// www.nationalgeographic.co.uk/animals/ 2018/09/where-jaguars-are-killed-orderillegal-trade Bale, R. (2018) On the trail of jaguar poachers. National Geograhic. h t t p s : / / www.nationalgeographic.com/magazine/ 2017/12/on-the-trail-of-jaguar-poachers/ Bauer, H., Packer, C., Funston, P.F., Henschel, P. & Nowell, K. (2016). Panthera leo. The IUCN Red List of Threatened Species http://www. iucnredlist.org [accessed 6 December 2018]. Biggs, D., Courchamp, F., Martin, R. & Possingham, H.P. (2013) Legal trade of Africa’s rhino horns. Science 339, 1038–1039. Clavero, M. (2016) Species substitutions driven by anthropogenic positive feedbacks: Spanish crayfish species as a case study. Biological Conservation, 193, 80-85. Conrad, K. (2012) Trade bans: a perfect storm for poaching? Tropical Conservation Science, 5, 245-254. Courchamp, F., Angulo, E., Rivalan, P., Hall, R.J., Signoret, L., Bull, L. et al. (2006) Rarity value and species extinction: the anthropogenic allee effect. PLoS Biol, 4, e415. Ellis, R. (2005) Tiger bone and rhino horn: The destruction of wildlife for traditional chinese medicine.%Shearwater Books of Island Press,%Washington D.C. 294 pp. Enviromental Investigation Agency (2013) Hidden in plain sight: China´s clandestine tiger trade. EIA 30 pp. Enviromental Investigation Agency (2017) Cultivating demand: The growing threat of tiger farms. EIA 23 pp. Haken,%J.%(2011) Transnational Crime in the Developing World.%Report for Global Financial Integrity, February, www.gfintegrity.org/ storage/gfip/documents/reports/transcrime/ gfi_transnational_crime_web.pdf (accessed 1 October 2018). IUCN (International Union for Conservation of Nature). 2018. Red List of Threatened Species. IUCN, Gland, Switzerland. Karmacharya, D., Sherchan, A.M., Dulal, S., Manandhar, P., Manandhar, S., Joshi, J. et al. (2018) Species, sex and geo-location identification of seized tiger (Panthera tigris tigris) parts in Nepal —A molecular forensic approach. PLoS ONE, 13(8), e0201639. Khoshoo, T.N. (1997) Conservation of India’s endangered mega animals: tiger and lion. Current Science, 73, 830–842. Luque-Larena, J., Mougeot, F., Arroyo, B. & Lambin, X. (2018) “Got rats?” Global environmental costs of thirst for milk include acute biodiversity impacts linked to dairy feed production. Global Change Biology, 24, 2752-2754. Maxwell, S.L, Fuller, R.A, Brooks, T.M & Watson, J. (2016) Biodiversity: The ravages of guns, nets and bulldozers. Nature, 536, 143-145. McClenachan, L., Cooper, A. & Dulvy, N. (2016) Rethinking trade-driven extinction risk in marine and terrestrial megafauna. Current Biology, 26, 1640-1646. Milner-Gulland, E. (2018) Documenting and tackling the illegal wildlife trade: Change and continuity over 40 years.%Oryx,%52, 597-598. Milner-Gulland, E. & Beddington, J.R. (1993) The exploitation of elephants for the ivory trade: an historical perspective. Proceedings of the Royal Society of London B,!252,%29–73. Milner-Gulland, E. & Leader-Williams, N. (1992) A model of incentives for the illegal exploitation of black rhinos and elephants: Poaching pays in Luangwa Valley, Zambia.%Journal of Applied Ecology,%29, 388-401. Mills, J.A. & Jackson, P. (1994) Killed for a cure: review of the world trade in Tiger Bone. Traffic International, Cambridge, United Kingdom. Nijman, V. (2010) An overview of international wildlife trade from Southeast Asia. Biodiversity and Conservation, 19, 1101-1114. Nowell, K. (2014) Review of implementation of Resolution Conf. 12.5 (Rev. CoP16) on Conservation and trade in tigers and other Appendix-I Asian big cats. IUCN/SSC Cat Specialist Group. CITES SC65 Doc.38 Annex 1. Nowell, K. & Xu, L. (2007) Lifting China's tiger trade ban would be a catastrophe for conservation. Cat News, 46, 28–29. Perez, M. (2017) Chinese investments in Latin America Opportunities for growth and diversification. CELAC. http://repositorio.cepal.org//handle/ 11362/41134 [Accessed on October 10, 2018). Ripple, W.,%Abernethy, K., Betts, M., Chapron, G., Dirzo, R., Galetti, M. et al. (2016) Bushmeat hunting and extinction risk to the world mammals. Royal Society Open Science, 3, 160498. https://doi.org/10.1098/rsos.160498 Rosen, G.E. & Smith, K.F. (2010) Summarizing the evidence on the international trade in illegal wildlife. EcoHealth, 7, 24-32. Rumiz, D. & Rivero, K. (2018) Peritaje de partes de fauna silvestre. Tecnical report. Museo de Historia Natural Noel Kempff Mercado. 15 pp. Scriber, B. (2014) 100,000 Elephants killed by poachers in just three years, landmark analysis finds. National Geographic. http:// news.nationalgeographic.com/news/2014/ 08/140818-elephants-africa-poaching-citescensus/Shina, M. & Arjun, R. (2017) Wildlife protection and biodiversity conservation – A review on utility of forensic DNA analysis. Species, 18, 77-90. Stoner, S. (2014) Tigers: Exploring the threat from illegal online trade. Traffic Bulletin, 26, 26-30. Stephens, P. & Sutherland, W. (1999) Consequences of the Allee effect for behaviour, ecology and conservation. Trends in Ecology & Evolution, 14, 401-405. Webb, T.J (2016) Conservation: Threatened by luxury. Current Biology, 26, 498-500. Williams, V.L., Newton, D.J., Loveridge, A.J. & Macdonald, D.W. (2015). Bones of contention: An assessment of the South African trade in African lion Panthera leo bones and other body parts. Traffic, Cambridge, UK & WildCRU, Oxford, UK. Williams, V.L., Loveridge, A.J., Newton, D.J. & Macdonald, D.W. (2017) A roaring trade? The legal trade in Panthera leo bones from Africa to East-Southeast Asia. PLoS ONE, 12(10), e0185996. World Health Organization (WHO) (2013). Estrategia de la OMS sobre medicina tradicional 2014-2023. Genebra: World Health Organization. 150 Chapter 6 Wildlife trafficking has become a billion-euro criminal industry supporting an international luxury market of medicinal and decorative animal-based products (1). Endangered species are disproportionately affected; continuous harvesting of their populations creates feedback loops in which a species become rare, which increases its value and stimulates further harvesting (2). The overexploitation of Asian tigers over centuries is a clear example of this phenomenon, which has led the species to its current endangered status (3). Some of the insatiable demand for tigers is now met by tigers raised and bred in captivity (4). In 1993, the Chinese government banned any trade of tigers, but in October 2018, the commerce of captive tigers was legalized for scientific, educational, and medicinal purposes (5). This measure may reactivate the illegal tiger trade, because a legal market provides sellers an outlet to traffic illegal products. China’s decision threatens not only the world’s few remaining wild tigers, but also most big cat species worldwide, which may be marketed to consumers as tiger or as a substitute for tiger. For example, more than 6000 lion skeletons were legally moved from Africa to Asia during the past decade, marketed as a legal alternative to tiger bones (6). Recently, employees of construction companies with access to jaguars in Bolivia and Suriname attempted to send them illegally to China to serve as tiger substitutes in traditional medicine (7, 8). Since 2013, more than 380 jaguar teeth have been seized by Bolivian authorities (9), and in 2016 alone, up to 185 big cat fangs were delivered to China (10). These data suggest a growing illegal trade that seriously threatens the world’s largest cats. Current molecular techniques can help stem the tide of illegal trade (11). The Chinese government should promote the establishment of a regulatory institution to differentiate between captive and wild tigers based on DNA profiling. If a product appears to be illegal, it should be seized and the seller should be penalized. Meanwhile, governments worldwide should pay special attention to the impact that tiger trade regulations have on the populations of other big cats that may be used as substitutes for tiger products. 152 Tiger trade threatens big cats worldwide References 1.T. J. Webb, Curr. Biol. 26, 498 (2016). 2.F. Courchamp et al., PLOS ONE 4, 12 (2006). 3. J. Waltson et al. PLOS Biol. 8, e1000485 (2010). 4.Environmental Investigation Agency (EIA) “Cultivating demand: The growing threat of tiger farms” (Tech. Rep. EIA, 2017). 5.The State Council, The People's Republic of China, “China to control trade in rhino and tiger products” (2018); http://english.gov.cn/policies/ latest_releases/2018/10/29/ content_281476367121088.htm. 6.V. L Williams, A. J. Loveridge, D. J. Newton, D. W. Macdonald, PLOS ONE 12, e0185996 (2017). 7.R. Bale, “Where jaguars are ‘killed to order’ for the illegal trade,” National Geographic (2017); www.nationalgeographic.co.uk/animals/2018/09/ where-jaguars-are-killed-order-illegal-trade. 8.R. Bale, “On the trail of jaguar poachers,” National Geographic (2018) www.nationalgeographic.com/ magazine/2017/12/on-the-trail-of-jaguarpoachers/. 9.D. Navia, “Fang trafficking is putting Bolivia´s jaguars in jeopardy,” Mongabay (2018); www.news.mongabay.com/2018/01/fangtrafficking-to-china-is-putting-bolivias-jaguars-injeopardy/. 10. D. Rumiz, K. Rivero, “Peritaje de partes de fauna silvestre” (Museo de Historia Natural, Technical Report, 2018). 11. M. Shina, R. Arjun, Species 18, 77 (2017). 153