On the road - RIO Principal

Transcription

La portada es una versión editada de una fotografía tomada por Dr. Nuno Negrões en el
Camino del Control, Reserva Biológica de Doñana (España).
Ad Angela e Melo, per la terza volta
On the road:
los distintos impactos del tráfico motorizado
sobre poblaciones animales
Marcello D’Amico
Tesis Doctoral
Sevilla, 2015
Estación Biológica de Doñana, CSIC
Departamento de Biología de la Conservación
Universidad Pablo de Olavide
Facultad de Ciencias Experimentales
Departamento de Biología Molecular e Ingeniería Bioquímica
Doctorado en Estudios Medioambientales
On the road:
los distintos impactos del tráfico motorizado sobre poblaciones animales
Memoria presentada por el Licenciado en Ciencias Naturales Marcello D’Amico
para optar al título de Doctor por la Universidad de Pablo de Olavide
Fdo. Marcello D’Amico
Dr. Eloy Revilla y Dr. Jacinto Román,
Estación Biológica de Doñana (CSIC)
CERTIFICAN
Que los trabajos de investigación desarrollados en la Memoria de Tesis Doctoral “On
the road: los distintos impactos del tráfico motorizado sobre poblaciones animales” son
aptos para ser presentados por el Ldo. Marcello D’Amico ante el Tribunal que en su día
se designe, para aspirar al Grado de Doctor por la Universidad Pablo de Olavide.
Y para que así conste, y en cumplimiento de las disposiciones legales vigentes,
extendemos el presente certificado a 18 de Diciembre de 2014
Directores:
Fdo. Eloy Revilla
Tutor:
Fdo. Francisco Eduardo Narbona Fernández
Universidad Pablo de Olavide
Fdo. Jacinto Román
“And this was really the way that my whole road experience began,
and the things that were to come are too fantastic not to tell.”
Jack Kerouac, On the Road
Índice
Resumen..........................................................................................................................12
Introducción general........................................................................................................13
Capítulo 1, Atropellos:
Patrones de atropello en hábitats mediterráneos: quién, cuándo y dónde.......................24
Capítulo 2, Efecto barrera:
¿Cuánto tráfico es demasiado? El impacto de redes viarias heterogéneas en áreas
protegidas.........................................................................................................................86
Capítulo 3, Medidas de mitigación:
Comparación de muestreos genéricos y específicos: estimando la eficiencia de
diferentes pasos de fauna para micromamíferos............................................................116
Capítulo 4, Reproducción y comportamiento:
El tráfico motorizado como elemento de estrés que estimula los cuidados
parentales en aves..........................................................................................................140
Capítulo 5, Invasiones Biológicas:
Invasores en ruta: aves sinantrópicas alimentándose en las carreteras..........................168
Discusión general y Conclusiones.................................................................................188
Agradecimientos............................................................................................................197
11
Resumen
Resumen: Esta tesis doctoral describe patrones y mecanismos de impacto viario
sobre poblaciones animales, acerca de los temas principales de la Ecología de
carreteras: atropello, efecto barrera, medidas de mitigación, impacto sobre
comportamiento y reproducción, invasiones biológicas. En el primer capítulo vimos
que los ectotermos comunes y abundantes son los vertebrados mediterráneos con más
probabilidad de ser atropellados en Doñana (España), los anfibios durante las lluvias y
los reptiles (serpientes y lagartijas) coincidiendo con altas temperaturas. Las lagartijas
(y también los micromamíferos) son atropelladas sobre todo en grandes vías muy
transitadas. En el segundo capítulo observamos que la red viaria heterogénea (i.e. con
variación de tráfico y superficie viaria) de Doñana afecta la probabilidad de presencia
de ciervos y jabalíes, y especialmente la distancia al camino más cercano, aunque no
tenga tráfico ni pavimentación. La consecuencia a nivel de paisaje es una reducción de
probabilidad de presencia del 32% y del 45%, con una grande reducción de calidad de
hábitat en un enclave protegido. En el tercer capítulo demostramos que los muestreos
genéricos de eficacia de pasos de fauna para micromamíferos están sesgados hacía
especies generalistas de hábitat en detrimento de las especialistas, pero ambas parecen
preferir los ecoductos a los pasos subterráneos. En el cuarto capítulo vimos que los
abejarucos que crían en las áreas laterales de los caminos perciben el tráfico como un
elemento de estrés que estimula alarmas y estampidas. Sin embargo, las tasas de ceba
son mayores en nidos situados en carreteras de tránsito elevado, aumentando durante los
días laborales y con picos en las horas punta. En el quinto capítulo observamos que las
aves sinantrópicas invasoras en Nueva Zelanda usan las carreteras intermunicipales para
la búsqueda de comida, los que podría favorecer que estas actuaran como corredores de
invasión. Todos los resultados de esta tesis nos ayudan a comprender mejor el impacto
de carreteras y tráfico sobre la biodiversidad y como mitigarlo.
12
Introducción general
Road ecology (en español Ecología de carreteras) es una disciplina de la
Biología de la Conservación que estudia las interacciones entre organismos y ambientes
relacionados con redes viarias y vehículos (Forman et al., 2003). Este concepto fue
desarrollado por primera vez por Ellenberg et al. (1981) usando el término alemán
Straßenökologie; traducido posteriormente al inglés por Richard T.T. Forman, uno de
los padres de esta disciplina (Forman, 1998). En esta publicación, Forman irónicamente
se preguntaba: “¿Qué es enorme, conspicuo, y eludido por los ecólogos?”. La respuesta
era el sistema viario, que definía como “el gigante que nos abraza” (the giant embracing
us) y que de alguna forma nos une y amenaza al mismo tiempo. Como veremos más
adelante, se puede considerar un ámbito de investigación relativamente reciente, que
hunde sus raíces mayoritariamente en la Ecología, pero también en la Geografía y en la
Ingeniería civil (Forman et al., 2003; Coffin, 2007).
Importancia de la Ecología de carreteras, aspectos ecológicos:
La metáfora del gigante de Forman resulta especialmente acertada en los países
industrializados. En los Estados Unidos de América, por ejemplo, el 1% de la superficie
total del país está ocupado por carreteras (lo que equivale a un área del tamaño de
Austria; Forman, 1998; Forman and Alexander, 1998). Es probable que en otros países
industrializados más pequeños y densamente poblados (como los europeos, por
ejemplo) este porcentaje sea incluso mayor. Pero las redes viarias se están rápidamente
convirtiendo en gigantes también en el resto del mundo, y especialmente en Europa
oriental, India, China y América latina (van der Ree et al., 2011).
13
Los efectos de las redes viarias sobre el entorno no se limitan a la mera
destrucción de hábitat por la presencia de caminos y carreteras o a los atropellos de
fauna que ocurren a lo largo de estas infra-estructuras, sino que se extienden
considerablemente en los ambientes colindantes, principalmente a causa de las
emisiones asociadas al tráfico motorizado (e.g. contaminación química, acústica y
lumínica; Forman and Alexander, 1998; Forman et al., 2003). Los mismos autores que
estimaban que el 1% de la superficie de los Estados Unidos de América estaba ocupado
por carreteras, también consideraban que hasta el 20% del país podría sufrir los efectos
de estas emisiones (Forman, 1998; Forman and Alexander, 1998), en las que se han
denominado road-effect zones (Forman, 2000; Forman and Deblinger, 2000). Con el
tiempo, nuevos estudios resaltaron que los efectos de las infra-estructuras lineares
actúan de forma sinérgica, aumentando la extensión de estas zonas de impacto. Este
concepto es la base de la Teoría ecológica de las redes viarias (Ecological road-network
theory) que postula que los efectos acumulativos de distintas infra-estructuras lineares
pueden llegar incluso a saturar un paisaje con baja densidad de carreteras, no quedando
áreas libres de su efecto (Forman et al., 2003; Coffin, 2007). Finalmente, hay que
considerar que los impactos asociados a la red viaria pueden afectar a distintos hábitats
de forma diferente, componiendo zonas de impacto heterogéneas que, por otro lado,
siguen actuando de forma sinérgica, causando la que se ha denominado
hiperfragmentación (Trombulak and Frissell, 2000).
Importancia de la Ecología de carreteras, aspectos sociales
Las redes viarias conectan gente con gente, sociedades con sociedades, ciudades
con ciudades, haciendo posible el movimiento de personas y bienes. Sin ellas, la calidad
14
de la vida bajaría dramáticamente (Forman et al., 2003). Los humanos las utilizamos
intensamente; solo hace falta pensar que un habitante de un país industrializado pasa, de
media, una hora diaria viajando a lo largo de infra-estructuras lineares (Forman and
Alexander, 1998). Esta intensidad de uso aumenta inexorablemente la probabilidad de
accidentes, con lo que esto conlleva en términos económicos y de vidas humanas. Una
parte de estos accidentes hoy en día es debida a las colisiones entre vehículos y fauna,
cuya principal causa radica en el desarrollo extraurbano de las redes viarias (Forman et
al., 2003). Un ejemplo de la importancia de las colisiones entre vehículos y fauna en el
cómputo total de accidentes nos lo proporciona el caso de dos especies de cérvidos de
los Estados Unidos de América (Odocoileus hemionus y Odocoileus virginianus). En
este país, solo estas dos especies causan cada año aproximadamente 720000 accidentes,
que suelen comportar el fallecimiento de los animales atropellados, la implicación
traumática de unas 29000 personas (incluidas más de 200 muertes) e ingentes daños
económicos (Conover et al., 1995). Si consideramos que a lo largo de las redes viarias
estadounidenses se registran en total unas 33000 víctimas humanas al año, veremos que
solo estas dos especies causan un 0.6% de estas muertes (International Traffic Safety
Data and Analysis Group, 2014). El gasto económico que supondría acondicionar
adecuadamente las infra-estructuras lineares sería considerablemente inferior a los
perjuicios monetarios resultantes de las colisiones entre vehículos y fauna (Conover et
al., 1995; Forman et al., 2003). Resulta evidente, entonces, que la resolución de los
conflictos viarios entre humanos y fauna no aporta únicamente beneficios para las
poblaciones animales y, precisamente por esta razón, debería ser un objetivo explícito
de las políticas viarias (Forman, 1998). Todos los gobiernos implicados deberían
considerar, por lo tanto, la Ecología de carreteras como la principal herramienta para el
cumplimiento de este propósito (Forman et al., 2003).
15
Breve historia de la Ecología de carreteras
Los primeros trabajos de esta disciplina trataban problemas de erosión y drenaje
en pistas de tierra de Norteamérica a principio del siglo XX (Lay, 1992). La
relativamente temprana implementación de superficies pavimentadas significó la
resolución de estos problemas y la momentánea perdida de interés en este campo de
investigación (Forman et al., 2003).
Por otro lado, la pavimentación de pistas permitió el aumento de la velocidad de
circulación de los vehículos, incrementándose las colisiones entre vehículos y fauna,
dando impulso a un tema que ha sido desde siempre central en Ecología de carreteras
(Forman et al., 2003). Los atropellos de fauna representaban un impacto evidente y
llamativo sobre las poblaciones animales y, como hemos visto, tenían y siguen teniendo
una relevancia económica (sobre todo en Canadá, USA y Escandinavia, donde
vehículos motorizados y grandes mamíferos abundaban; Forman et al., 2003). Por lo
tanto, no es de extrañar que pronto algunos ecólogos y gestores ambientales empezaran
a interesarse en este tema. En realidad hasta los años ‘70 solo se llevaron a cabo una
decena de estudios (e.g. Barnes, 1936; Dickerson, 1939; Davis, 1940; Pickles, 1942;
Finnis, 1960; Hodson, 1962; Dunthorn and Errington, 1964) motivados por un interés
puntual de científicos y gestores (Forman et al., 2003). Incluso entonces, a pesar de que
todavía la Ecología de carreteras no se había desarrollado de una forma orgánica, la
importancia de esta disciplina resultaba evidente para algunos investigadores (“Few
forces have been more influential in modifying the earth than transportation”; Ullman,
1956).
Entre los años ‘70 y los ‘80 los estudios sobre atropellos de fauna se
intensificaron (e.g. McCaffery, 1973; Case, 1978; Coulson, 1982). La construcción en
16
Francia de los primeros pasos de fauna inspiró estudios preliminares sobre la eficacia de
esta medida de mitigación (Camby and Maizeret, 1985; Ballon, 1985), y empezaron a
publicarse trabajos sobre contaminación (Sherwood and Bowers, 1970; Hickman and
Colwill, 1982) y efecto barrera debido a infra-estructuras lineares (Oxley et al., 1974;
Mader, 1984). A partir de 1990 y gracias al desarrollo de la Ecología del paisaje
(Coffin, 2007), el interés de científicos y gestores hacía la Ecología de carreteras
aumentó considerablemente. Fue entonces cuando realmente se convirtió en una
disciplina, como demuestran las abundantes revisiones bibliográficas que se empezaron
a publicar entre el final de esa década y el principio de la siguiente (Forman, 1998,
2000, 2004; Forman and Alexander, 1998; Underhill and Angold, 1999; Trombulak and
Frissell, 2000; Seiler, 2001; Zong et al., 2002; Li et al., 2003). Aun así, en 1998 Richard
T.T. Forman y Lauren E. Alexander todavía consideraban que la Ecología de carreteras
era una disciplina que tenía las potencialidades para aportar soluciones a los problemas
viarios pero que aún no había cumplido adecuadamente con su cometido (Forman and
Alexander, 1998). Precisamente por esta razón los mayores expertos de Ecología de
carreteras se reunieron para llevar a cabo la publicación del libro “Road ecology:
science and solutions”. Por primera vez se revisaban de forma orgánica todos los
aspectos de esta disciplina, sugiriendo un cambio conceptual y apuntando a un enfoque
más científico e interdisciplinar (Forman et al., 2003).
17
Objetivos generales y particulares de esta tesis doctoral
El hilo conductor a lo largo de la bibliografía existente en Ecología de carreteras
es que la mayoría de trabajos realizados hasta el momento expone casos de estudio con
aproximaciones más o menos descriptivas, y por lo tanto con un rango amplio de
posibilidades de inferencia. En este contexto, probablemente, la necesidad más urgente
en Ecología de carreteras es fomentar la realización de estudios que no se limiten solo a
una descripción de patrones sino que investiguen los mecanismos que los generan. El
objetivo general de esta tesis doctoral es proporcionar unos ejemplos de esta clase de
estudios, aportando unas aproximaciones lo suficientemente generales como para
contribuir al desarrollo de la Ecología de carreteras pero que al mismo tiempo puedan
sugerir también como resolver problemáticas reales y mitigar amenazas concretas
relacionadas con el impacto de la presencia de infra-estructuras lineares y tráfico
motorizado. Más concretamente, esta tesis doctoral se centra en el impacto de carreteras
y trafico sobre poblaciones animales, desarrollándose a lo largo de cinco capítulos, cada
uno relativo a uno de los principales temas de Ecología de carreteras, que hemos
mencionado anteriormente: atropellos, efecto barrera, medidas de mitigación, impacto
sobre reproducción y comportamiento, e invasiones biológicas.
La mortalidad por atropello es uno de los impactos más llamativos sobre
poblaciones animales y conlleva un no desdeñable coste económico. Probablemente por
esta razón se lleva estudiando desde hace varias décadas (Conover et al., 1995; Forman
et al., 2003). Hasta la fecha se han recolectado a lo largo del planeta una cantidad
considerable de datos (Forman et al., 2003; Coffin, 2007), y un considerable número de
publicaciones están contribuyendo a salvar muchas poblaciones animales de la extinción
18
local o generalizada (Jones, 2000; Downs et al., 2014). A pesar de esto, muchos
estudios de mortalidad por colisión con vehículos se limitan a meros listados de las
especies atropelladas en lugares concretos, investigando raramente los mecanismos
(tanto temporales como espaciales o de características específicas intrínsecas) que se
asocian directamente al fenómeno. Entender estos mecanismos debería ser el principal
objetivo futuro de los estudios sobre atropellos, permitiendo el diseño e implementación
de medidas de mitigación más eficaces.
Lo mismo se puede decir de los trabajos relativos al efecto barrera, que en la
última década se han incrementado sobre todo en los países industrializados (Forman et
al., 2003; van der Ree et al., 2011). A pesar de que en buena parte sean estudios
descriptivos, han aportado indicaciones sobre los mecanismos que pueden causar la
evitación de infra-estructuras lineares en varias especies (e.g. McGregor et al., 2008;
Brehme et al., 2013). Esta clase de estudios necesitará en el futuro ubicar los fenómenos
descritos en el marco teórico establecido por Jaeger et al. (2005), que diferencia entre
evitación de superficie viaria, emisiones asociadas y el efecto de la presencia de
vehículos, y que resulta esencial para elegir la medida de mitigación más oportuna.
Con respecto a la eficacia de las medidas de mitigación, este resulta ser uno de
los temas principales en Ecología de carreteras, y en las últimas décadas ha aportado
grandes mejoras a la conectividad del paisaje (Forman et al., 2003). El siguiente paso
debería ser averiguar los mecanismos que determinan la eficacia de estas medidas de
mitigación, con el fin último de aumentarla para el rango más amplio posible de
especies. Muchos trabajos futuros probablemente serán aproximaciones más generales a
este tema, con el propósito de poder extrapolar las conclusiones de estos estudios
concretos a estructuras, especies y entornos distintos.
19
También hemos mencionado la existencia de estudios que tratan la relación entre
infra-estructuras lineares y tráfico a motor con modificaciones del comportamiento o
variaciones en el éxito reproductivo. Las cuestiones comportamentales asociadas al
tráfico que se han investigado hasta ahora en Ecología de carreteras son aquellas
relativas a colisiones y efecto barrera, y en menor medida a tolerancia y habituación
(Dean and Milton, 2003; Laurian et al., 2008; Malo et al., 2011). Apenas existen
trabajos acerca de capacidad cognitiva, personalidad, aprendizaje y transmisión cultural,
comunicación intra e inter-especifica, conducta territorial, social o sexual; cuando todas
estas aproximaciones en realidad serían seguramente muy valiosas a la hora de diseñar
medidas de mitigación. Análogamente, podemos observar que los estudios de biología
reproductiva realizados hasta ahora suelen limitarse a la descripción de un impacto
viario sobre el éxito reproductivo de una especie, normalmente a corto plazo (e.g.
Reijnen et al., 1997; Flaspohler et al., 2001). Todavía no se están llevando a cabo
trabajos sobre el impacto del tráfico en la calidad de los reproductores y la prole, con lo
que eso conlleva en términos de éxito reproductivo a largo plazo.
Finalmente, el estudio de invasiones biológicas se ha integrado con la Ecología
de carreteras especialmente en el caso de las plantas invasoras, mientras que apenas
existen estudios relativos a especies animales y mucho menos con extrapolaciones a
nivel de población y paisaje, que serían seguramente importantes para la contención de
este impacto. De todas formas, con respecto a la relación entre redes viarias e invasiones
biológicas, comportamiento animal o éxito reproductivo todavía quedan muchas
preguntas básicas por contestar.
En concreto, los objetivos particulares de cada capítulo son:
20
1) Atropellos: investigar en un mismo estudio que factores específicos
intrínsecos, temporales y espaciales afectan la probabilidad de atropello en
vertebrados.
2) Efecto barrera: investigar en una red viaria heterogénea (con variación en
intensidad de tráfico y superficie viaria), que factores determinan el efecto
barrera en ungulados (evitación de infra-estructuras lineares, de tráfico y
emisiones asociadas o de vehículos concretos), y su impacto en la calidad del
hábitat.
3) Medidas de mitigación: investigar en un mismo estudio y en
micromamíferos la eficacia de unas medidas de mitigación y de los métodos
que normalmente se usan para evaluarlas.
4) Reproducción y comportamiento: investigar algunos de los efectos
potencialmente negativos del tráfico sobre reproducción de aves que
aparentemente se benefician de la presencia de carreteras.
5) Invasiones biológicas: investigar la asunción previa a la existencia de un
patrón de dispersión de aves sinantrópicas invasoras a lo largo de carreteras.
21
Bibliografía
Balkenhol, N. and Waits, L.P. (2009). Molecular road ecology: exploring the potential of genetics for
investigating transportation impacts on wildlife. Molecular Ecology. 18: 4151-4164.
Ballon, P. (1985). Premières observations sur l’efficacité des passages à gibier sur l’autoroute A36
[Preliminary observations on the effectiveness of game species road-crossing structures on the
highway A36]. In: Routes et faune sauvage. Actes du colloque [Roads and wildlife. Symposium
proceedings]. SETRA (Ed.). Service d’Etudes Techniques de Routes et Autoroutes, Bagneaux
(France), pp. 311-316.
Barnes, M.D. (1936). The death-roll of birds on our roads. Naturalist. 85-86.
Brehme, C.S., Tracey, J.A., McClenaghan, L.R. and Fisher, R.N. (2013). Permeability of roads to
movement of scrubland lizards and small mammals. Conservation Biology. 27: 710-720.
Camby, A. and Maizeret, C. (1985). Perméabilité des routes et autoroutes vis-a-vis des mammifères
carnivores: exemple des études menées dans les Landes de Gascogne par radio-poursuite
[Permeability of roads and highways for carnivorous mammals: a radio-tracking case study from
Landes de Gascogne]. In: Routes et faune sauvage. Actes du colloque [Roads and wildlife.
Symposium proceedings]. SETRA (Ed.). Service d’Etudes Techniques de Routes et Autoroutes,
Bagneaux (France), pp. 183-196.
Case, R.N. (1978). Interstate highway road-killed animals: a data source for biologists. Wildlife Society
Bulletin. 6: 8-13.
Coffin, A.W. (2007). From roadkill to road ecology: a review of the ecological effects of roads. Journal of
Transport Geography. 15: 396-406.
Conover, M.R., Pitt, W.C., Kessler, K.K., DuBow, T.J. and Sanborn, W.A. (1995). Review of human
injuries, illnesses, and economic losses caused by wildlife in the United States. Wildlife Society
Bulletin. 23: 407-414.
Coulson, G.M. (1982). Road-Kills of macropds on a section of highway in central Victoria. Wildlife
Research. 9: 21-26.
Davis, W.B. (1940). Mortality of wildlife on a Texas highway. The Journal of Wildlife Management. 4:
90-91.
Dean, W.R.J. and Milton, S.J. (2003). The importance of roads and road verges for raptors and crows in
the Succulent and Nama-Karoo, South Africa. Ostrich: Journal of African Ornithology. 74: 181-186.
Dickerson, L.M. (1939). The problem of wildlife destruction by automobile traffic. The Journal of
Wildlife Management. 3: 104-116.
Downs, J., Horner, M., Loraamm, R., Anderson, J., Kim, H. and Onorato, D. (2014). Strategically
locating wildlife crossing structures for Florida panthers using maximal covering approaches.
Transactions in GIS. 18: 46-65.
Dunthorn, A.A. and Errington, F.P. (1964). Casualties among birds along a selected road in Wiltshire.
Bird Study. 11: 168-182.
Ellenberg, H., Müller, K. and Stottele, T. (1981). Straßen-Ökologie: Auswirkungen von Autobahnen und
Straßen auf Ökosysteme deutscher Landschaften [Road ecology: effects of highways and roads on
German ecosystems and landscapes]. In: Ökologie und Straße [Ecology and roads]. Walper, K.H.,
Ellenberg, H., Müller, K. and Stottele, T. (Eds.). Deutschen Straßenliga, Bonn (Germany), pp. 19-122.
Finnis, R.G. (1960). Road casualties among birds. Bird Study. 7: 21-32.
Flaspohler, D.J., Temple, S.A. and Rosenfield, R.N. (2001). Species-specific edge effects on nest success
and breeding bird density in a forested landscape. Ecological Applications. 11: 32-46.
Forman, R.T.T. (1998). Road ecology: a solution for the giant embracing us. Landscape Ecology. 13: 3-5.
Forman, R.T.T. (2000). Estimate of the area affected ecologically by the road system in the United States.
Conservation Biology. 14: 31-35.
Forman, R.T.T. (2004). Road ecology's promise: what's around the bend? Environment: Science and
Policy for Sustainable Development. 46: 8-21.
Forman, R.T.T. and Alexander, L.E. (1998). Roads and their major ecological effects. Annual Review of
Ecology and Systematics. 29: 207-231.
Forman, R.T.T. and Deblinger, R.D. (2000). The ecological road-effect zone of a Massachusetts (U.S.A.)
suburban highway. Conservation Biology. 14: 36-46.
Forman, R.T.T., Sperling, D., Bissonette, J., Clevenger, A.P., Cutshall, C.D., Dale, V.H., Fahrig, L.,
France, R., Goldman, C.R., Heanue, K., Jones, J.A., Swanson, F.J., Turrentine, T. and Winter, T.C.
(2003). Road ecology: science and solutions. Island Press, Washington DC (USA).
22
Hickman, A.J. and Colwill, D.M. (1982). The estimation of air pollution concentrations from road traffic.
Transport and Road Research Laboratory, Crowthorne (UK).
Hodson, N.L. (1962). Some notes on the causes of bird road casualties. Bird Study. 9: 168-173.
International Traffic Safety Data and Analysis Group (2014). Road Safety Annual Report. Organisation
for Economic Co-operation and Development (OECD)/International Transport Forum (ITF), Paris
(France).
Jaeger, J.A.G., Bowman, J., Brennan, J., Fahrig, L., Bert, D., Bouchard, J., Charbonneau, N., Frank, K.,
Gruber, B. and von Toschanowitz, K.T. (2005). Predicting when animal populations are at risk from
roads: an interactive model of road avoidance behavior. Ecological Modelling. 185: 329-348.
Jones, M.E. (2000). Road upgrade, road mortality and remedial measures: impacts on a population of
eastern quolls and Tasmanian devils. Wildlife Research. 27: 289-296.
Laurian, C., Dussault, C., Ouellet, J.-P., Courtois, R., Poulin, M. and Breton, L. (2008). Behavior of
moose relative to a road network. The Journal of Wildlife Management. 72: 1550-1557.
Lay M.G. (1992). Ways of the World: a history of the World's roads and of the vehicles that used them.
Rutgers University Press, New Brunswick (USA).
Li, Y., Hu, Y., Li, X. and Xiao, D. (2003). A review on road ecology. Chinese Journal of Applied
Ecology. 14: 447-452.
Mader, H.J. (1984). Animal habitat isolation by roads and agricultural fields. Biological Conservation. 29:
81-96.
Malo, J.E., Acebes, P. and Traba, J. (2011). Measuring ungulate tolerance to human with flight distance: a
reliable visitor management tool? Biodiversity and Conservation. 20: 3477-3488.
McCaffery, K.R. (1973). Road-kills show trends in Wisconsin deer populations. The Journal of Wildlife
Management. 37: 212-216.
McGregor, R.L., Bender, D.J. and Fahrig, L. (2008). Do small mammals avoid roads because of the
traffic? Journal of Applied Ecology. 45: 117-123.
Oxley, D.J., Fenton, M.B. and Carmody, G.R. (1974). The effects of roads on small mammals. Journal of
Applied Ecology. 11: 51-59.
Pickles, W. (1942). Animal mortality on three miles of Yorkshire roads. Journal of Animal Ecology. 11:
37-43.
Reijnen, R., Foppen, R. and Veenbaas, G. (1997). Disturbance by traffic of breeding birds: evaluation of
the effect and considerations in planning and managing road corridors. Biodiversity and Conservation.
6: 567-581.
Seiler, A. (2001). Ecological effects of roads: a review. Introductory research essay. Swedish University
of Agricultural Sciences Uppsala (Sweden), pp. 1-40.
Sherwood, P.T. and Bowers, P.H. (1970). Air pollution from road traffic-a review of the present position.
Road Research Laboratory, Earthworks and Foundations, and Climate and Environment Section,
Springfield (USA).
Spellerberg, I.A.N. (1998). Ecological effects of roads and traffic: a literature review. Global Ecology and
Biogeography. 7: 317-333.
Trombulak, S.C. and Frissell, C.A. (2000). Review of ecological effects of roads on terrestrial and aquatic
communities. Conservation Biology. 14: 18-30.
Ullman, E.L. (1956). The role of transportation and the bases for interaction. In: Man's role in changing
the face of the Earth. Thomas, J.W.Jr. (Ed.). Wenner-Gren Foundation for Anthropological Research
and the National Science Foundation University of Chicago Press (USA), pp. 862-880.
Underhill, J.E. and Angold, P.G. (1999). Effects of roads on wildlife in an intensively modified
landscape. Environmental Reviews. 8: 21-39.
van der Ree, R., Jaeger, J.A.G., van der Grift, E.A. and Clevenger, A.P. (2011). Effects of roads and
traffic on wildlife populations and landscape function: road ecology is moving toward larger scales.
Ecology and Society. 16: 48 [online].
Zong, Y., Zhou, S.Y., Peng, P., Liu, C., Guo, R. and Cheng, H. (2002). Perspective of road ecology
development. Acta Ecologica Sinica. 23: 2396-2405.
23
Capítulo 1:
Atropellos
Patrones de atropello en hábitats mediterráneos:
quién, cuándo y dónde.
_____________________________________________________________________________________
D'Amico, M. a, Román, J. a, de los Reyes, L.
b
and Revilla, E.
a
(Manuscript under
second review). Road-kill patterns in Mediterranean habitats: who, when and where.
a
Estación Biológica de Doñana CSIC, Seville (Spain).
b
Área de Conservación, Espacio Natural de Doñana, Almonte (Spain).
Road-kill patterns in Mediterranean habitats: who, when and where
Resumen: La mortalidad por atropello es el impacto más estudiado del tráfico y
una amenaza para la biodiversidad. Sin embargo, gran parte de la investigación se
enfoca en determinadas especies o medidas de mitigación, describiendo las causas
próximas de atropello y raramente investigando los mecanismos. Este estudio es una
aproximación más general, que analiza factores de historia natural, temporales y
espaciales que afecten el atropello de vertebrados en paisajes mediterráneos, que son
unos puntos calientes de biodiversidad con relevantes y poco estudiados impactos de
carreteras. Durante un año recolectamos atropellos en carreteras pavimentadas del
Parque Natural de Doñana. Encontramos 2368 atropellos de 66 especies (32% de la lista
de referencia), teniendo los ectotermos abundantes más probabilidad de ser atropellados.
Además, investigamos los factores temporales y espaciales que afectaran los patrones de
atropello de diferentes grupos de especies. Las lluvias coincidieron con los picos de
atropello de anfibios, mientras las altas temperaturas se asociaban a un aumento de
atropellos en serpientes, lagartijas y aves nidificantes. Lagartijas y micromamíferos
fueron atropellados sobre todo a lo largo de grandes vías con mucho tráfico y altos
límites de velocidad. Exhortamos priorizar medidas de mitigación capaces de reducir
permanentemente la mortalidad de ectotermos (anfibios y reptiles), como pasos de fauna
específicos y cercados impermeables en correspondencia de sus puntos calientes de
atropello. Al mismo tiempo, las medidas temporales de mitigación pueden aplicarse
durante los picos estacionales de atropellos. El presente trabajo proporciona algunas
recomendaciones para reducir el impacto de los atropellos en ambientes mediterráneos,
pero paralelamente intenta contribuir a un desarrollo más general de la ecología de
carreteras, sugiriendo algunas directrices generales para realizar estudios sobre la
mortalidad debida al tráfico motorizado.
25
Capítulo 1: Atropellos
Abstract: Road mortality is the most recognized traffic impact and a threat for
biodiversity. Nevertheless, most research on this topic deals with particular species or
with mitigation measures, describing road-kill proximate correlates but rarely
investigating the mechanisms. Here we provide an example of a more general
approximation, describing life history, temporal and spatial factors affecting vertebrate
road-kills in Mediterranean landscapes, which are a biodiversity hotspot with relevant
and little studied road impacts. During one year we recorded road casualties found in
paved roads within Doñana Natural Park. We found 2368 road-kills belonging to 66
species (32% of the reference list), with more abundant ectotherm species more likely to
be found as road-kills. We also investigated temporal and spatial factors affecting the
road-kill patterns of different taxonomic and functional groups. Rainfall events
coincided with road-kill peaks for amphibians, while high temperatures were associated
with increases in the number of casualties for snakes, lizards and nesting birds. Lizards
and small mammals were mainly road-killed along wider roads, with more traffic
volume and higher allowed speed. We suggest prioritizing the mitigation measures
which can permanently decrease the road mortality of ectotherm species (mostly
amphibians and reptiles), such as specific road-crossing structures and drift fences on
road-kill hotspots. Concurrently, group-specific temporal mitigation measures can be
applied
during
the
road-kill
seasonal
peaks.
The
present
work
provides
recommendations to decrease road-kill impact in Mediterranean environments, but
simultaneously tries to contribute to a more general development of road ecology
research, suggesting several useful guidelines to perform road mortality studies.
26
Introduction
Road-kill mortality is the most well-known impact of traffic (Forman and
Alexander, 1998; Trombulak and Frissell, 2000; Forman et al., 2003), but, in spite of
mitigation efforts, a very large number of vertebrates die worldwide on paved and
unpaved roads (Coffin, 2007; van der Grift et al., 2013). Road-kills are a considerable
threat for populations of several species (Fahrig et al., 1995; Mumme et al., 2000;
Taylor et al., 2002) and a relevant issue for human road safety, involving also high
economic costs (Conover et al., 1995; Forman et al., 2003; Huijser et al., 2009). As a
consequence, in the last decades the number of road-related studies has increased
considerably, improving our knowledge on what has been defined as road ecology
(Forman et al., 2003; Coffin, 2007). Most research has focused on a few emblematic
species or on particular mitigation measures (Forman et al., 2003; van der Grift et al.,
2013), and therefore at present there is still a need for more general approximations. For
example, most studies describing road-kill patterns investigate only the role of local
variables or proximate causes, thus their recommendations to limit road mortality are
difficult to translate to other localities or to other species. Road-kill patterns could
probably be better understood by additionally considering the role of more general
predictors.
An example of those more general descriptors is the study of species-specific
life history traits affecting the susceptibility to collision with vehicles. This is an
overlooked issue in road ecology with few published studies (Ford and Fahrig, 2007;
Barthelmess and Brooks, 2010; Cook and Blumstein, 2013), although the available
27
bibliography demonstrates the existence of strong patterns and relevant differences
among species (e.g. Ashley and Robinson, 1996; Erritzoe et al., 2003; Orlowski and
Nowak, 2006; Glista et al., 2008). An additional step should be testing the relevance of
other predictors which may be potentially correlating with both road-kill probability and
life-history traits, as is the case for example of species abundance, which has been
proposed in the past to be the underlying factor affecting road-kill probability (Ford and
Fahrig, 2007; Barthelmess and Brooks, 2010; Møller et al., 2011). Non-road-killed
species should be also included in the analysis, because they can provide valuable
information on key species-specific traits. Life history research can improve road
mortality knowledge, but at the same time it can also provide direct recommendations
that can be applied to the conservation of vulnerable groups of species for which data is
not yet available.
Road mortality research focused on life history traits can also improve our
understanding of temporal and spatial road-kill patterns, especially if the target species
are aggregated at relevant taxonomic or functional groups (e.g. amphibians or migrant
birds) that might be differentially affected by specific climatic or environmental
conditions.
Previous studies have found that temporal variations in road-kill patterns are
associated with seasonal behaviors such as dispersal or migration, thus helping to
predict outbreaks in the number of casualties (Bonnet et al., 1999; Gryz and Krauze,
2008; Smith-Patten and Patten, 2008). The drivers inducing these behaviors (e.g.
temperature or rainfall) have been suggested to be directly linked to the road-kill peaks
(Puky, 2005; Andrews et al., 2006; Glista et al., 2008), but usually without statistical
28
analyses. Conversely, understanding which additional factors affect seasonal behaviors
and the consequent temporal variability in road-kill patterns would potentially improve
mitigation measures, especially those that can be differentially managed in time (e.g.
traffic intensity).
Finally, a central issue in road mortality research is the spatial distribution of
casualties, which is probably the most investigated topic in road ecology. Most studies
have investigated the correlations between the presence of road-kills and some spatial
variables describing road features or traffic volume (Forman and Alexander, 1998;
Trombulak and Frissell, 2000; Forman et al., 2003). Nevertheless, the spatial
distribution of road-kills can be affected by a large amount of landscape and road
features, and some of them can act simultaneously (Clevenger et al., 2001a; 2003;
Forman et al., 2003). It is probably for this reason that several spatial variables (e.g.
traffic volume or surrounding habitat) have shown contradictory effects in different
studies (Forman et al., 2003). A simultaneous evaluation of the potentially relevant
predictors, obviously controlling for correlated factors, is the way to proceed.
In this study we aimed to comprehensively describe the life history, temporal
and spatial factors potentially affecting road-kill patterns, performing a detailed
selection of candidate predictors previously found or suggested as affecting road
mortality. A priori, we expect that some species are more prone to be found as road-kills
than others, species with particular life history traits directly or indirectly related to their
abundance or movement behavior. We also expect that temporal road-kill patterns
should be determined by predictors affecting temporal variations in species abundance
29
or activity, and/or by traffic variations along time. Finally, we expect that road-kill
spatial patterns should be related with predictors describing local variations in
abundance, road properties and traffic.
We tested these hypotheses focusing on terrestrial vertebrates in a typical
Mediterranean
landscape
(Doñana
Natural
Park,
south-western
Spain).
The
Mediterranean basin is a biodiversity hotspot (Mittermeier et al., 1998; Myers et al.,
2000) with an ancient and widespread road-network development, affecting even
protected areas and involving potential threats for wildlife (Ferreras et al., 1992; Gomes
et al., 2009; Grilo et al., 2009). Nevertheless, road-kill studies have been mainly
performed in temperate landscapes (Trombulak and Frissell, 2000; Forman et al., 2003),
with few contributions from Mediterranean areas (Malo et al., 2004; Carvalho and Mira,
2011; Garriga et al., 2012).
This study should provide some general management and conservation
recommendations for an emblematic protected area and for other Mediterranean
environments, while providing scientists and conservation biologists with some
guidelines to better understand the factors behind road-kill patterns.
30
Materials and Methods
Study area
Doñana Natural Area (36º59’ N, 6º26’ W; Figure 1) is a Biosphere Reserve with
a Mediterranean climate, characterized by a mosaic of urban, rural and natural
environments in which, depending on the level of protection (National and Natural
parks), some human activities are allowed. The local road-network is widespread, with
different types of linear infra-structures and traffic intensities. We surveyed all paved
roads within the Natural Park: one regional road (A494 Matalascañas-Mazagón, 23 km)
and three paved forestry/agricultural roads (Cabezudos, 5 km; A483-Hinojos, 11 km;
A483-Villamanrique de la Condesa, 16 km; Figure 1). All the surveyed roads were
located in the same ecosystem, Mediterranean forest and scrubland. Local terrestrial
vertebrate communities are composed by typically Mediterranean species.
31
Figure 1: Study area. Black lines represent the paved road-network. Bold lines are the surveyed road
sections: regional road A494 Matalascañas-Mazagón, forestry/agricultural road of Cabezudos,
forestry/agricultural road A483-Hinojos, and forestry/agricultural road A483-Villamanrique de la
Condesa.
Data collection
From March 2006 to February 2007 (July 2007 excluded) we surveyed driving
at 15 km/h the four roads twice per week (one day in the morning and one day in the
afternoon, randomly determined), on both directions. The observer (always the same
person, LdlR) searched for road-killed animals on road surface and along the adjacent
verges and ditches. We georeferenced, identified (when possible) and removed all the
road-killed animals.
32
We used bibliography and data from the Natural Processes Monitoring Team
(ESPN) of EBD-CSIC to compile a checklist of the species present in the study area in
association with the vegetation and land-uses adjacent to the roads (Appendix 1). A
priori we considered that all the species included in the checklist could be found as
road-kills.
For those species we compiled a set of life history variables previously found or
suggested as affecting road-kill probability (Ford and Fahrig, 2007; Barthelmess and
Brooks, 2010; Cook and Blumstein, 2013; Appendix 1 for more details). The candidate
predictors directly or indirectly described species as a function of their abundance or
movement. In the first case the candidate predictors were abundance (rare, common, or
abundant), size (body mass), breeding (average number of offspring per female-year),
activity (nocturnal or diurnal), habitat preferences (urban, farmlands, grasslands,
scrubland/woodland and freshwater), food habits (carnivorous, insectivorous or
herbivorous) and territoriality (territorial or not). In the second case the candidate
predictors were movement (terrestrial or flying), speed (fast or slow movers) and their
metabolic strategy (ectotherms or endotherms; see Appendix 1 for more details).
In the analysis of the temporal patterns of road-kill mortality we considered
species as belonging to taxonomic or functional groups (temporal groups hereafter) with
similar seasonal responses in abundance, behavior and activity: amphibians, snakes,
lizards, resident birds, migrant nesting birds, migrant wintering birds, transient migrant
birds and small mammals (Appendix 2 for more details).
For this set of analyses we selected three candidate predictors suggested by
previous studies as affecting seasonal variations in species abundance and consequently
33
in road-kill mortality (Puky, 2005; Andrews et al., 2006; Glista et al., 2008; Appendix 2
for more details): photoperiod variation (average of the daily variation for each month),
rainfall and temperature (both measured as the daily average measured in the week
prior to each survey). We also selected another candidate predictor related with
seasonal traffic variation (average of daily traffic per season). Finally, considering that
carcasses suffer temporal degradation (Antworth et al., 2005; Santos et al., 2011), we
also selected as candidate predictor the time elapsed since the previous survey.
We obtained photoperiod variation data from the U.S. Naval Oceanography
Portal
(http://www.usno.navy.mil/USNO/astronomical-applications).
Rainfall
and
temperature were calculated using data from the meteorological station of Palacio de
Doñana (ESPN). Traffic in the forestry/agricultural roads (Cabezudos, A483-Hinojos,
and A483-Villamanrique de la Condesa) was obtained using a magnetometer (TRAFx
Vehicle Counter Generation III; sampling during one week per season), while the traffic
of the regional road (A494 Matalascañas-Mazagón) was obtained from the Andalusian
Regional Government (sampling during one weekend and two weekdays per season at
an official counter station).
In the analyses of spatial patterns we classified the road-killed species with
enough observations into five taxonomic or functional groups with a priori similar
spatial behaviors (spatial groups hereafter): amphibians, snakes, lizards, small
passerines and small mammals (Appendix 3 for more details).
We divided the surveyed roads (55 km in total) into 1100 sections. These
sections were 50-m long with the aim to avoid spatial auto-correlation and
simultaneously describe changes in spatial predictors. Each section was characterized
34
by a set of candidate predictors previously found or suggested as affecting spatial
variations in the distribution of road-kills: species presence/abundance, the probability
of road-crossing at the specific site and traffic (Clevenger et al., 2001a; 2003; Forman et
al., 2003). The predictors potentially affecting the spatial variation in species abundance
were related to the adjacent landscape: macrohabitat (grassland, scrubland or
woodland) and distance to the nearest water body. The predictors potentially affecting
spatial variation in wildlife road-crossing were related to road verges: road verge habitat
(grass, grass with trees, grass with shrubs and trees, shrubs), presence/absence of roadfences, drainage culverts and wildlife road-crossing structures. The predictors
potentially affecting spatial variations in traffic and its speed were related to the road
itself, they were road width, annual average of daily traffic, maximum allowed speed,
and the presence/absence of road-shoulders, curves, slope changes in the road, and
vertical or horizontal road signs. We characterized all road sections in the field and
using a geographic information system (landscape variables).
Data analysis
First, we calculated the correlation between predictors in order to consider the
potential effect of collinearity in the life history analysis (Appendix 1). Abundance
strongly correlated with size, food habits, territoriality and movement. Therefore, we
fitted an initial generalized linear model (GLM) testing the relationship between a
dependent variable describing all the species initially considered in the checklist with a
value of 1 for those found road killed at least once in any of the roads and 0 for those
that were not found (i.e. species road-kill probability), and, as a predictor, the
35
abundance (model with a binomial error distribution and logit link function; GLIMMIX
procedure of SAS version 9.3; SAS Institute Inc., 2012).
As the probability of rare species to be detected as road-killed was much lower,
all subsequent life-history analyses were done on a reduced species checklist in which
we removed rare species (both the road-killed and not road-killed ones). For every lifehistory predictor we fitted a separate GLM (again with species road-kill probability as
response variable and with binomial error distribution and logit link function). We
considered as plausible those models at ΔAIC < 2 from the best model (Burnham et al.
2011). Finally, we calculated model support using Akaike weights (wAIC, ranging from
0 to 1, with larger numbers indicating greater support; Burnham and Anderson 2002).
In the temporal analyses we followed a similar approach to evaluate the
collinearity between predictors (Appendix 2). In this case, as the correlation was low,
we run a single model for each temporal group of species to evaluate the impact of the
selected predictors on the number of road-killed individuals per survey, using
generalized linear mixed models (GLMM) in which the surveyed road was included as
a random factor (negative binomial error distribution and log link function using
procedure GLIMMIX in SAS version 9.3).
In the spatial analyses, we found that several candidate predictors were strongly
correlated, especially road width, traffic and speed (Appendix 3). As a consequence, we
replaced these with a new explanatory variable named road magnitude, obtained from a
principal component analysis of those variables (PCA; FACTOR procedure in SAS
version 9.3). We also excluded from further analyses all the variables that correlated
36
with road width, traffic or speed. The implemented spatial predictors then were road
magnitude, presence/absence of wildlife road-crossing structures and distance to the
nearest water body. Finally, for every spatial group we evaluated the impact of the
selected predictors on the number of road-killed individuals accumulated during the
study in each of the 50-m road sections using GLMMs in which the surveyed road was
included as a random factor (negative binomial error distribution and log link function
using procedure GLIMMIX in SAS version 9.3).
Results
We detected 2368 road-killed vertebrates of which the majority (67%) were
amphibians (Table 1). They included both anurans and urodeles (1568 and 20
casualties, respectively). We could identify only a small fraction of the road-killed
amphibians, mostly corresponding to natterjack toads Epidalea calamita (381
casualties) and western spadefoot toads Pelobates cultripes (142 casualties). We found
299 reptile casualties (12.6% of the total), including snakes (mainly Malpolon
monspessulanus and Rhinechis scalaris) and lizards (mostly Psammodromus
manuelae). Birds accounted for 12.2% of the casualties (290), covering a large range of
orders (Erithacus rubecula was the most common). Finally, we found 191 (8.1%) roadkilled mammals (133 small mammals, mainly rodents).
37
Table 1: Road-killed species and groups. Number of road-killed individuals found and their assignation
to taxonomic or functional groups used in the different analyses.
Species
Number of road-kills
Amphibians
Discoglossus galganoi
1
Epidalea calamita
381
Pelobates cultripes
142
Pelophylax perezi
7
Pleurodeles waltl
20
Unidentified Amphibians
1037
Total Amphibians
1588
Reptiles
Coronella girondica
6
Hemorrhois hippocrepis
1
Macroprotodon brevis
4
Malpolon monspessulanus
39
Natrix maura
10
Rhinechis scalaris
39
Vipera latasti
8
Unidentified Snakes
22
Total Snakes
129
Acanthodactylus erythrurus
20
Blanus cinereus
1
Chalcides bedriagai
2
Chalcides striatus
1
Chamaeleo chamaeleon
2
Podarcis carbonelli
19
Podarcis vaucheri
1
Psammodromus manuelae
79
Tarentola mauritanica
15
Timon lepidus
3
Unidentified Lizards
33
Total Lizards
170
Birds
Alectoris rufa
2
Apus pallidus
1
Bubo bubo
1
Buteo buteo
2
Caprimulgus ruficollis
13
Columba livia domestica
6
Cyanistes caeruleus
1
Cyanopica cooki
4
Emberiza calandra
4
Erithacus rubecula
27
Ficedula hypoleuca
7
Fringilla coelebs
2
Hirundo rustica
1
38
Temporal group
Spatial group
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Amphibians
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Snakes
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Lizards
Resident
Nesting
Resident
Resident
Nesting
Resident
Resident
Resident
Resident
Wintering
Migrant
Resident
Nesting
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
-
Lanius senator
Lophophanes cristatus
Lullula arborea
Muscicapa striata
Oceanodroma leucorhoa
Parus major
Passer domesticus
Passer montanus
Phylloscopus collybita/ibericus
Saxicola rubetra
Saxicola rubicola
Serinus serinus
Strix aluco
Sylvia atricapilla
Sylvia cantillans
Sylvia melanocephala
Sylvia undata
Troglodytes troglodytes
Turdus philomelos
Unidentified Small passerines
Total Birds
Mammals
Apodemus sylvaticus
Canis lupus familiaris
Crocidura russula
Eliomys quercinus
Erinaceus europaeus
Felis silvestris catus
Herpestes ichneumon
Lepus granatensis
Mus spretus
Oryctolagus cuniculus
Pipistrellus pipistrellus
Rattus rattus
Vulpes vulpes
Unidentified Mammals
Unidentified Small mammals
Total Mammals
5
1
1
1
1
1
4
1
10
2
1
4
2
3
2
6
8
1
2
163
290
Nesting
Resident
Resident
Migrant
Resident
Resident
Resident
Migrant
Migrant
Resident
Resident
Resident
Migrant
Migrant
Resident
Resident
Resident
Wintering
-
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
Small passerines
5
2
1
2
17
2
1
4
4
17
1
2
5
9
119
191
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
Small mammals
39
We were able to identify 991 casualties as belonging to 66 different species
(Table 1), representing 32% of our species checklist (Appendix 1). The results of the
first GLM showed that abundance affected the species road-kill probability, with lower
probability for rare species (F = 9.13; p = 0.0002; Figure 2). The reduced dataset
included only the 148 abundant and common species, of which 60 were found as roadkills. The rest of the life history models were run with this reduced subset. The most
supported model was the one including metabolic strategy (Table 2). Overall, ectotherm
species had a higher probability to be found as road-kills than endotherms (Figure 2).
Table 2: Factors affecting species road-kill probability, model ranks by AIC weights (wAIC). ΔAIC
is the difference of a given AIC value compared to the smallest AIC value. The only supported model
(ΔAIC < 2) is marked with an asterisk. AIC weights indicate the relative support of every model.
Evidence ratio (ER) is the ratio of wAIC, comparing the best supported model with every competing one.
Model
Body weight
Offsprings
Habitat preference
Circadian activity
Feeding strategy
Territoriality
Movement type
Speed
Thermal strategy *
40
AIC
ΔAIC
wAIC
Rank
199.8
198.3
196.1
200.3
202.7
202.5
192.8
203.4
189.8
10.0
8.6
6.3
10.5
12.9
12.7
3.1
13.7
0.0
0.00
0.01
0.03
0.00
0.00
0.00
0.17
0.00
0.78
5
4
3
6
8
7
2
9
1
ER
154.4
72.2
23.7
194.4
648.7
587.0
4.7
939.2
1.0
Figure 2: Life history factors and species road-kill probability. Probability (± standard error) of species
present to be found as road kills as predicted by the initial model and the best supported model,
respectively. Left: species categorized by three coarse abundance categories (parameter estimates: β = 0.8 ± 0.4; rare = -1.8 ± 0.6, common = 0.5 ± 0.4, abundant = 0). Right: species included within the
abundant and common categories by their metabolic strategy (β = 1.1 ± 0.5; endotherms = -1.8 ± 0.5,
ectotherms = 0).
In the analysis of temporal patterns we considered 2156 road-kills (Table 1;
Appendix 2 for more information on the temporal distribution of each group). High
road-kill frequencies for lizards and resident birds occurred when the duration of the
photoperiod was increasing, while in the case of snakes, small mammals and migrant
birds they occurred when the photoperiod decreased (Table 3). Peaks in amphibian
road-kill mortality were associated with rainfall events, while in the case of lizards they
were associated with the lack of rain (Figure 3). Months with high temperatures
associated with increases in the number of casualties of snakes, lizards and nesting
birds, while months with low temperatures did so for amphibians and wintering birds
(Figure 3). Seasonal traffic intensity was associated with the number of road-killed
41
amphibians and snakes, with more casualties corresponding to periods with less traffic
(Table 3).
Table 3: Factors affecting temporal distribution of road-kills. In the table header we show the name of
the explanatory variable and its average value (range). Each row represents the results of the model for a
temporal group. For every explanatory variable with we give the parameter estimate (with standard error)
and, below, the p-value when p < 0.05.
Temporal variables
Temporal groups
Intercept
Photoperiod
variation
(minutes)
1.3 (-70, 71)
Temperature
(ºC)
2.0 (0, 15.3)
0.21 + 0.04
p < 0.0001
-0.14 + 0.06
p = 0.01
17.0 (6.2, 27.8)
-0.09 + 0.03
p = 0.003
0.17 + 0.03
p < 0.0001
0.26 + 0.03
p < 0.0001
-0.04 + 0.06
0.02 + 0.03
Amphibians
3.5 + 1.2
Snakes
-2.7 + 1.0
Lizards
-5.3 + 1.0
Resident birds
-2.3 + 0.7
Nesting birds
-7.1 + 2.0
0.006 + 0.006
-0.21 + 0.21
Wintering birds
-0.5 + 0.6
-0.018 + 0.005
0.06 + 0.05
Migrant birds
-2.7 + 0.9
Small mammals
-0.9 + 0.8
42
-0.005 + 0.003
Rainfall
(mm/day)
-0.005 + 0.002
p = 0.005
0.011 + 0.002
p < 0.0001
0.008 + 0.003
p = 0.01
-0.022 + 0.008
p = 0.007
-0.005 + 0.002
p = 0.003
0.004 + 0.027
0.22 + 0.09
p = 0.01
-0.15 + 0.04
p = 0.0005
Seasonal
traffic
(cars/day)
712 (98.3, 3217)
-0.002 + 0.001
p = 0.03
-0.0011 + 0.0003
p = 0.0005
Time since
previous survey
(days)
8 (0, 47)
-0.0004 + 0.0003
0.004 + 0.007
0.0004 + 0.0004
0.01 + 0.02
0.0002 + 0.0004
0.006 + 0.030
0.0002 + 0.0003
0.01 + 0.02
-0.008 + 0.017
-0.005 + 0.010
-0.01 + 0.06
-0.03 + 0.05
-0.0001 + 0.0004
0.02 + 0.02
-0.01 + 0.03
-0.01 + 0.02
0.0001 + 0.0001
0.016 + 0.009
Figure 3: Temporal road-kill patterns. Relationship between the number of road-kills found per survey
and temporal predictors: in the upper two graphs we show the effect of rainfall; in the lower five graphs
the effect of temperature. The solid line represents the estimated effect of the predictor, dotted lines are
the upper and lower 95% CI. For correspondent plots with raw data see Appendix 2.
43
Finally, for the spatial analyses we considered 1617 road-kills which could be
assigned to the spatial groups (Table 1; Appendix 3 for more information on the spatial
distribution of each group). Road magnitude was the most relevant factor affecting
road-kill patterns, with more lizard and small mammal casualties found in wider road
sections with more traffic and higher maximum speeds (Table 4; Figure 4). Road
sections near water had higher numbers of amphibian road-kills while those away from
water had more lizard casualties (Table 4). Overall, the presence of road-crossing
structures was not associated with the number of road-kills found for any group.
Table 4: Factors affecting the spatial distribution of road-kills. In the table header we show the name
of the explanatory variables, with its average value (range) in the case of continuous factors or its levels
in the case of categorical factors. Each row represents the results of the model for a spatial group. For
every explanatory variable with we give the parameter estimate (with standard error) and, below, the pvalue when p < 0.05. Road magnitude range (*) corresponds with 5-10 m of Road width (average = 8.79
m), 171-2499 cars/day of Traffic (average = 1547 cars/day) and 30-100 Km/h of Speed (average = 78.69
Km/h).
Spatial variables
Spatial groups
Intercept
44
Road
magnitude
0 (-2, 2.3) *
Crossing
structures
Absent
Present
Distance
from water (m)
1873 (0, 4632)
-0.0001 + 0.0001
p = 0.02
Amphibians
0.5 + 0.3
0.02 + 0.09
-0.06 + 0.14
0
Snakes
-2.1 + 0.5
-0.009 + 0.010
-0.06 + 0.46
0
-0.0001 + 0.0001
Lizards
-2.6 + 0.6
0.4 + 0.1
p = 0.001
0.1 + 0.5
0
-0.0001 + 0.0001
p = 0.002
Small passerines
-1.8 + 0.4
0.1 + 0.1
-0.1 + 0.3
0
-0.0001 + 0.0001
Small mammals
-2.0 + 0.8
0.7 + 0.3
p = 0.02
-0.2 + 0.4
0
-0.0001 + 0.0001
Figure 4: Spatial road-kill patterns and road magnitude. Relationship between the number of roadkills accumulated in all surveys and road magnitude. Road magnitude range corresponds with 5-10 m of
Road width (average = 8.79 m), 171-2499 car/day of Traffic (average = 1547 car/day) and 30-100 Km/h
of Speed (average = 78.69 Km/h). The solid line represents the estimated effect of the road magnitude
dotted lines are the upper and lower 95% CI. For correspondent plots with raw data see Appendix 3.
Discussion
In the present study we found that several life history, temporal and spatial
factors had a clear association with road-kill patterns. Life history analysis indicated
that abundant and ectotherm species had more probability to be road-killed, temporal
analysis highlighted the general factors influencing group-specific road-kill seasonality
(i.e. photoperiod variation, rainfall and temperature), and spatial analysis showed the
relevance of road features related with road magnitude (i.e. road width, traffic and
speed). These results improve our understanding on some of the underlying mechanisms
causing group-specific collisions in Mediterranean habitats.
45
The first interesting results on life history traits and road mortality were the
correlations among the candidate predictors. Previous studies already suggested that
several of these factors might be correlated, such as abundance with body size and food
habits (Ford and Fahrig, 2007; Barthelmess and Brooks, 2010).
In our case, we found that a coarse descriptor of species abundance correlated
with other predictors, making difficult its inclusion in further analysis. An initial
analysis confirmed that rare species were less likely to be found road-killed, as
previously suggested both directly (Møller et al., 2011) and indirectly (Ford and Fahrig,
2007; Barthelmess and Brooks, 2010).
The remaining analyses focusing only on abundant and common species showed
that ectotherm species (amphibians and reptiles) had a higher probability to be roadkilled. This is most probably because their metabolism causes slowness in amphibians
(Hels and Buchwald, 2001; Puky, 2005) and basking behavior in reptiles (Ashley and
Robinson, 1996; Tanner and Perry, 2007), but other reasons can be their relatively
lower environmental awareness and behavioral freezing responses to threats (Andrews
et al., 2006; Lima et al., 2014). Numbers and results make clear that mitigation
measures should focus on amphibians and reptiles. The establishment of drift fences
should be extensively implemented to permanently reduce their road-kill numbers
(Aresco, 2005; Puky, 2005), including roads with low traffic where basking behavior is
more frequent (Aresco, 2005; Jones et al., 2008).
The number of casualties showed strong variations along the year for the
different taxonomic and functional groups, as previously found (Ashley and Robinson,
1996; Erritzoe et al., 2003; Orlowski and Nowak, 2006), in association with periods of
46
the year in which there are increases in activity or abundance, such as during seasonal
mate search, juvenile dispersal or migration (Bonnet et al., 1999; Clevenger et al.,
2001b; Gryz and Krauze, 2008; Smith-Patten and Patten, 2008). Additionally, we found
a clear association among road-kill numbers and weather predictors, which usually
contribute by triggering the above mentioned seasonal behaviors (Fryxell and Sinclair,
1988; Lloyd, 1999; Gwinner and Helm, 2003). It is important to note that in
Mediterranean environments those variables can be quite unpredictable, in spite of
which they can be used by managers to determine when to enforce temporal mitigation
measures.
We found that photoperiod variation was the only temporal factor having a
significant association with seasonal peaks in road-kill mortality of small mammals and
both resident and migrant birds, most probably in correspondence with variations in
their presence and abundance (García et al., 2000; Mumme et al., 2000; Todd, 2001;
Moreno and Rouco, 2013). In these cases it would be therefore relatively easy to reduce
the group-specific casualties, through the implementation of temporary traffic-calmed
zones in the relevant spatial road-kill hotspots (Jaarsma and Willems, 2002; Jaarsma et
al., 2013). The temporary limitation of maximum speed can be alternatively applied
when traffic reduction is not possible (Sullivan et al., 2004; Hardy et al., 2006). The
positive association between temperature and both snake and lizard road-kills depends
on their increase in activity during the warm season (Koenig et al., 2002; Meek, 2009)
and the basking behavior often carried out on the road surface (Ashley and Robinson,
1996; Tanner and Perry, 2007). On the other hand, amphibian road-kill peaks occur
during events of high rainfall events, which in Mediterranean landscapes generate
breeding habitats and the seasonal migrations of toads and newts (Díaz-Paniagua et al.,
47
2005; Semlitsch, 2008). The same pattern has been previously suggested in boreal and
temperate landscapes (Clevenger et al., 2003; Glista et al., 2008), in which water is a
less limiting factor. The unpredictability of Mediterranean precipitations would make
difficult the planning of traffic-calmed zones, and therefore with the aim to limit
amphibian road-kill peaks we recommend the implementation of both temporary traffic
signs and drift fences in amphibian road-kill hotspots (Aresco, 2005; Puky, 2005), and,
if needed, crossing rescue campaigns (Puky, 2005; Beebee, 2013).
The implementation of temporary mitigation measures can be applied with more
accuracy for groups with more than one temporal factor affecting road-kill seasonality.
For example, snakes were more road-killed during the decrease of photoperiod length,
but especially in association with a quantifiable threshold of high temperatures, which
usually coincides with the early autumn, just after the hatching of most Mediterranean
snakes (Bonnet et al., 1999; Pleguezuelos and Feriche, 2006; Feriche et al., 2008).
Reaching the temperature threshold during other time of the year will not mark a similar
temporal road-kill peak, and could be therefore ignored by managers.
In the spatial analysis we found that several potentially relevant descriptors of
road characteristics were strongly correlated, as previously suggested for road width,
traffic volume and maximum allowed speed (Forman et al., 2002; Jaeger et al., 2005).
The composed road magnitude was the most relevant factor affecting the spatial
distribution of road-kills, with more lizard and small mammal casualties found in wider
road sections, which hold more and faster traffic. This pattern has been previously
described for a large amount of species (Inbar and Mayer, 1999; Coelho et al., 2008),
48
and implies a tangible threat for populations of several species (Rosen and Lowe, 1994;
Jones, 2000).
We did not find an effect of the presence of wildlife road-crossing structures on
road-kill spatial distribution, probably because they are actually correcting the threat
and because in our study area these structures are mainly focused on medium-sized
mammals just as carnivores of which we detected a relatively low number of casualties.
In our study area they are built for rare species with proved road-kill susceptibility, such
as the critically endangered Iberian lynx Lynx pardinus (Ferreras et al., 1992).
Nevertheless, it is clear that managers should also consider the implementation of
appropriate road-crossing structures for smaller vertebrates (e.g. with effective drift
fences; Glista et al., 2009; Pagnucco et al., 2012) in correspondence to the groupspecific road-kill risk areas, and the improvement of the structures already present. Our
results also show that mitigation measures should be prioritized along major roads.
Unfortunately, we could not discern the single effects of road width, traffic volume and
maximum allowed speed; and we could not even determine the potential relevance of
the correlated spatial variables that we needed to remove from the analyses (for example
the presence of shoulders, curves, traffic signals, etc.). The effects of all of these factors
on the road-kill spatial distribution of the different taxonomic and functional groups
should be explored in the future.
We believe that our results have relevant implications for the establishment of
measures to mitigate road-kill patterns regarding life history, temporal and spatial
causes in our study area and also in other Mediterranean environments. Our spatial
analysis suggests how to permanently decrease wildlife-vehicle collisions through the
49
implementation of drift fences and road-crossing structures in group-specific road-kill
risk areas. When these mitigation measures cannot be permanently applied, we also
suggest temporal actions which can reduce road-related mortality during the temporal
road-kill peaks. It is also very clear that managers should focus on abundant and
ectotherm species, many of them with conservation problems, as is the case of
amphibians.
Road ecology is an expanding discipline, and road mortality is its most
investigated issue. Currently many researchers, managers and also nonprofessional
environmental organizations are collecting data that should be used to improve our
general understanding of the problem. We hope that this study contributes to the
development of this discipline and to the conservation of threatened ecosystems.
Acknowledgements
The road-kill survey was realized by Doñana Conservation Area, while the
explanatory variable surveys were funded by the project 2008X0963 (Andalusian
Ministry of Environment; Andalusian Regional Government). The ESPN (EBD-CSIC)
provided unpublished data for life history variable survey, and Carmen Díaz-Paniagua,
Margarita Florencio and Carlos Ibañez suggested bibliographic references. The ESPN
also provided data for temporal variable survey. The Huelva Provincial District of
Public Works and Transportation (Andalusian Regional Government) provided traffic
data for the A494 Matalascañas-Mazagón regional road. Land-Rover España S.A.
kindly lent two vehicles for spatial variable survey, while Gemma Calvo, Clara Grilo,
Pablo Lucas and especially Andrea Barón helped with fieldwork. MD was supported by
50
JAE-PRE fellowship from CSIC; ER was supported by projects funded by the Spanish
Ministry of Science and Innovation co-funded by FEDER (CGL2009-07301/BOS,
CGL2012-35931/BOS). Sofía Conradi-Fernández and Angela Genovese provided
logistic support. Giulia Bastianelli, Manuela González-Suárez, Simone Santoro and
Laura Rios-Pena helped with statistical analyses; Anthony P. Clevenger, Nestor
Fernández, Zulima Tablado and Carlos Rodriguez provided helpful comments. Sol
Heber and three anonymous referees reviewed a draft version of the manuscript.
References
Aresco, M.J. (2005). Mitigation measures to reduce highway mortality of turtles and other herpetofauna
at a north Florida lake. Journal of Wildlife Management. 69: 549-560.
Ashley, E.P. and Robinson, J.T. (1996). Road mortality of amphibians, reptiles and other wildlife on the
Long Point Causeway, Lake Erie, Ontario. Canadian Field-Naturalist. 110: 404-412.
Andrews, K.M., Gibbons, J.W. and Jochimsen, D.M. (2006). Literature synthesis of the effects of roads
and vehicles on amphibians and reptiles. Federal Highway Administration (FHWA), U.S. Department
of Transportation, 151 pp., Washington DC (USA).
Antworth, R.L., Pike, D.A. and Stevens, E.E. (2005). Hit and run: effects of scavenging on estimates of
roadkilled vertebrates. Southeastern Naturalist. 4: 647-656.
Barthelmess, E.L. and Brooks, M.S. (2010). The influence of body-size and diet on road-kill trends in
mammals. Biodiversity and Conservation. 19: 1611-1629.
Beebee, T.J.C. (2013). Effects of road mortality and mitigation measures on amphibian populations.
Bonnet, X., Naulleau, G. and Shine, R. (1999). The dangers of leaving home: dispersal and mortality in
snakes. Biological Conservation. 89: 39-50.
Burnham, K.P. and Anderson, D.R. (2002). Model selection and multimodel inference. A practical
information-theoretic approach. Springer-Verlag, 488 pp., New York (USA).
Burnham, K., Anderson, D. and Huyvaert, K. (2011). AIC model selection and multimodel inference in
behavioral ecology: some background, observations, and comparisons. Behavioral Ecology and
Sociobiology. 65: 23-35.
Carvalho, F. and Mira, A. (2011). Comparing annual vertebrate road kills over two time periods, 9 years
apart: a case study in Mediterranean farmland. European Journal of Wildlife Research. 57: 157-174.
Clevenger, A.P., Chruszcz, B. and Gunson, K. (2001a). Highway mitigation fencing reduces wildlifevehicle collisions. Wildlife Society Bulletin. 29: 646-653.
Clevenger, A.P., Chruszcz, B. and Gunson, K.E. (2003). Spatial patterns and factors influencing small
Vertebrate fauna road-kill aggregations. Biological Conservation. 109: 15-26.
Clevenger, A.P., McIvor, M., McIvor, D., Chruszcz, B. and Gunson K. (2001b). Tiger salamander,
Ambystoma tigrinum, movements and mortality on the Trans-Canada highway in southwestern
Alberta. Canadian Field-Naturalist. 115: 199-204.
Coelho, I.P., Kindel, A. and Coelho, A.V.P. (2008). Roadkills of vertebrate species on two highways
through the Atlantic Forest Biosphere Reserve, southern Brazil. European Journal of Wildlife
Research. 54: 689-699.
Coffin, A.W. (2007). From roadkill to road ecology: a review of the ecological effects of roads. Journal of
Transport Geography. 15: 396-406.
51
Conover, M.R., Pitt, W.C., Kessler, K.K., DuBow, T.J. and Sanborn, W.A. (1995). Review of Human
injuries, illnesses, and economic losses caused by Wildlife in the United States. Wildlife Society
Bulletin. 23: 407-414.
Cook, T.C. and Blumstein, D.T. (2013). The omnivore’s dilemma: diet explains variation in vulnerability
to vehicle collision mortality. Biological Conservation. 167: 310-315.
Díaz-Paniagua, C., Gómez-Rodríguez, C., Portheault, A. and de Vries, W. (2005). Los Anfibios de
Doñana [The Amphibians of Doñana]. Organismo Autónomo de Parques Nacionales (Ministerio de
Medio Ambiente), 181 pp., Madrid (Spain).
Erritzoe, J., Mazgajski, T.D. and Rejt, L. (2003). Bird casualties on European roads: a review. Acta
Ornithologica. 38: 77-93.
Fahrig, L., Pedlar, J.H., Pope, S.E., Taylor, P.D. and Wegner, J.F. (1995). Effect of road traffic on
amphibian density. Biological Conservation. 73: 177-182.
Feriche M., Pleguezuelos J.M. & Santos X. (2008). Reproductive ecology of the Montpellier snake,
Malpolon monspessulanus (Colubridae), and comparison with other sympatric colubrids in the Iberian
Peninsula. Copeia. 2008: 279-285.
Ferreras, P., Aldama, J.J., Beltrán, J.F. and Delibes, M. (1992). Rates and causes of mortality in a
fragmented population of Iberian lynx Felis pardina Temminck, 1824. Biological Conservation. 61:
197-202.
Ford, A.T. and Fahrig, L. (2007). Diet and body size of North American mammal road mortalities.
Transportation Research Part D: Transport and Environment. 12: 498-505.
(2003). Road ecology: science and solutions. Island Press, 481 pp., Washington DC (USA).
Forman, R.T.T., Reineking, B. and Hersperger, A.M. (2002). Road traffic and nearby grassland bird
patterns in a suburbanizing landscape. Environmental Management. 29: 782-800.
Fryxell, J.M. and Sinclair, A.R.E. (1988). Causes and consequences of migration by large herbivores.
Trends in Ecology & Evolution. 3: 237-241.
García, L., Ibáñez, F., Garrido, H., Arroyo, J.L., Máñez, M. and Calderón, J. (2000). Prontuario de las
aves de Doñana. Anuario ornitológico de Doñana Nº 0 [Short guide of Doñana Birds. Ornithological
Doñana Yearbook Nº 0]. Estación Biológica de Doñana (EBD - CSIC). Ayuntamiento de Almonte,
113 pp., Almonte (Spain).
Garriga, N., Santos, X., Montori, A., Richter-Boix, A., Franch, M. and Llorente, G.A. (2012). Are
protected areas truly protected? The impact of road traffic on vertebrate fauna. Biodiversity and
Conservation. 21: 2761-2774.
Glista, D.J., DeVault, T.L. and DeWoody, J.A. (2008). Vertebrate road mortality predominantly impacts
amphibians. Herpetological Conservation and Biology. 3: 77-87.
Glista, D.J., DeVault, T.L. and DeWoody, J.A. (2009). A review of mitigation measures for reducing
wildlife mortality on roadways. Landscape and Urban Planning. 91: 1-7.
Gomes, L., C. Grilo, C. Silva, and A. Mira. 2009. Identification methods and deterministic factors of owl
roadkill hotspot locations in Mediterranean landscapes. Ecological Research. 24: 355-370.
Goosem, M. (2002). Effects of tropical rainforest roads on small mammals: fragmentation, edge effects
and traffic disturbance. Wildlife Research. 29: 277-289.
Grilo, C., Bissonette, J. and Santos-Reis, M. (2009). Spatial-temporal patterns in Mediterranean carnivore
road casualties: consequences for mitigation. Biological Conservation. 142: 301-313.
Gryz, J. and Krauze, D. (2008). Mortality of vertebrates on a road crossing the Biebrza Valley (NE
Poland). European Journal of Wildlife Research. 54: 709-714.
Gwinner, E. and Helm, B. (2003). Circannual and circadian contributions to the timing of avian
migration. In: Avian Migration. Berthold, P., Gwinner, E. and Sonnenschein, E. (Eds.). Springer
Berlin Heidelberg (Germany), pp. 81-95.
Hardy, A., Lee, S. and Al-Kaisy, A. (2006). Effectiveness of animal advisory messages on dynamic
message signs as a speed reduction tool: case study in rural Montana. Transportation Research
Record: Journal of the Transportation Research Board. 1973: 64-72.
Hels, T. and Buchwald, E. (2001). The effect of road kills on amphibian populations. Biological
52
Huijser, M.P., Duffield, J.W., Clevenger, A.P., Ament, R.J. and McGowen, P. (2009). Cost-benefit
analyses of mitigation measures aimed at reducing collisions with large ungulates in the United States
and Canada: a decision support tool. Ecology and Society. 14: 15 [online].
Inbar, M. and Mayer, R.T. (1999). Spatio-temporal trends in armadillo diurnal activity and road-kills in
central Florida. Wildlife Society Bulletin. 27: 865-872.
Jaarsma, C.F., van Langevelde, F. and Beunen, R. (2013). Landscape ecology and rural roads: traffic
calming for improving both landscape and wildlife? Ekologia. 32: 352-360.
Jaarsma, C.F. and Willems, G.P.A. (2002). Reducing habitat fragmentation by minor rural roads through
traffic calming. Landscape and Urban Planning. 58: 125-135.
Jones, R., Cable, J. and Bruford, M.W. (2008). An evaluation of non-invasive sampling for genetic
analysis in northern European reptiles. The Herpetological Journal. 18: 32-39.
Koenig, J., Shine, R. and Shea, G. (2002). The dangers of life in the city: patterns of activity, injury and
mortality in suburban lizards (Tiliqua scincoides). Journal of Herpetology. 36: 62-68.
Laurance, W.F., Goosem, M. and Laurance, S.G.W. (2009). Impacts of roads and linear clearings on
tropical forests. Trends in Ecology & Evolution. 24: 659-669.
Lima, S.L., Blackwell, B.F., DeVault, T.L. and Fernández-Juricic, E. (2014). Animal reactions to
oncoming vehicles: a conceptual review. Biological Reviews. [online]
Lloyd, P. (1999). Rainfall as a breeding stimulus and clutch size determinant in South African arid-zone
birds. Ibis. 141: 637-643.
Malo, J.E., Suárez, F. and Díez, A. (2004). Can we mitigate animal-vehicle accidents using predictive
models? Journal of Applied Ecology. 41: 701-710.
Meek, R. (2009). Patterns of reptile road-kills in the Vendée region of western France. The
Herpetological Journal. 19: 135-142.
Mittermeier, R.A., Myers, N., Thomsen, J.B., Da Fonseca, G.A.B. and Olivieri, S. (1998). Biodiversity
hotspots and major tropical wilderness areas: approaches to setting conservation priorities.
Møller, A.P., Erritzøe, H. and Erritzøe, J. (2011). A behavioral ecology approach to traffic accidents:
interspecific variation in causes of traffic casualties among birds. Zoological Research. 32: 115-127.
Moreno, S. and Rouco, C. (2013). Responses of a small-mammal community to habitat management
through controlled burning in a protected Mediterranean area. Acta Oecologica. 49: 1-4.
Mumme, R.L., Schoech, S.J., Woolfenden, G.E. and Fitzpatrick, J.W. (2000). Life and death in the fast
lane: demographic consequences of road mortality in the Florida scrub-jay. Conservation Biology. 14:
501-512.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B. and Kent, J. (2000). Biodiversity
hotspots for conservation priorities. Nature. 403: 853-858.
Orlowski, G. and Nowak, L. (2006). Factors influencing mammal roadkills in the agricultural landscape
of south-western Poland. Polish Journal of Ecology. 54: 283-294.
Pagnucco, K.S., Paszkowski, C.A. and Scrimgeour, G.J. (2012). Characterizing movement patterns and
spatio-temporal use of under-road tunnels by long-toed salamanders in Waterton Lakes National Park,
Canada. Copeia. 2012: 331-340.
Pleguezuelos, J.M. and Feriche, M. (2006). Reproductive ecology of a Mediterranean ratsnake, the ladder
snake Rhinechis scalaris (Schinz, 1822). The Herpetological Journal. 16: 177-182.
Puky, M. (2005). Amphibian road kills: a global perspective. In: Proceedings of the 2005 International
Conference on Ecology and Transportation (ICOET). Irwin, C.L., Garrett, P. and McDermott, K.P.
(Eds.). Center for Transportation and the Environment, North Carolina State University (USA), pp.
325-338.
Rosen, P.C. and Lowe, C.H. (1994). Highway mortality of snakes in the Sonoran desert of southern
Arizona. Biological Conservation. 68: 143-148.
Santos, S.M., Carvalho, F. and Mira, A. (2011). How long do the dead survive on the road? Carcass
persistence probability and implications for road-kill monitoring surveys. PLoS ONE. 6: e25383
[online].
53
SAS Institute Inc. (2012). What’s New in SAS® 9.3. SAS Institute Inc., 272 pp., Cary (USA).
Semlitsch, R.D. (2008). Differentiating migration and dispersal processes for pond-breeding amphibians.
Journal of Wildlife Management. 72: 260-267.
Smith-Patten, B.D. and Patten, M.A. (2008). Diversity, seasonality, and context of mammalian roadkills
in the southern Great Plains. Environmental Management. 41: 844-852.
Sullivan, T.L., Williams, A.F., Messmer, T.A., Hellinga, L.A. and Kyrychenko, S.Y. (2004).
Effectiveness of temporary warning signs in reducing deer-vehicle collisions during mule deer
migrations. Wildlife Society Bulletin. 32: 907-915.
Tanner, D. and Perry, J. (2007). Road effects on abundance and fitness of Galápagos lava lizards
(Microlophus albemarlensis). Journal of Environmental Management. 85: 270-278.
Taylor, S.K., Buergelt, C.D., Roelke-Parker, M.E., Homer, B.L. and Rotstein, D.S. (2002). Causes of
mortality of free-ranging Florida panthers. Journal of Wildlife Diseases. 38: 107-114.
Todd, L.D. (2001). Dispersal patterns and post-fledging mortality of juvenile burrowing owls in
Saskatchewan. Journal of Raptor Research. 35: 282-287.
van der Grift, E.A., van der Ree, R., Fahrig, L., Findlay, S., Houlahan, J., Jaeger, J.A.G., Klar, N.,
Madriñan, L.F. and Olson, L. (2013). Evaluating the effectiveness of road mitigation measures.
Biodiversity and Conservation. 22: 425-448.
54
Appendix 1 – Life history patterns
Table 1: Species checklist and life history analysis data set. Species column indicates the check-list
classified by taxonomic categories (Amphibians, Reptiles, Birds and Mammals). In the second column
the meaning of the acronym Rk is road-killed species, indicating if for the given species was found at
least once as a casualty (1) during the study period. The other columns are the ten predictors of the Life
Body mass (g)
Breeding
(offsprings/female/year)
Activity
Habitat preference
Food habits
Territoriality
Movement
Speed
Metabolic strategy
Amphibians
Alytes cisternasii
Bufo bufo
Discoglossus galganoi
Epidalea calamita
Hyla meridionalis
Lissotriton boscai
Pelobates cultripes
Pelodytes ibericus
Pelophylax perezi
Pleurodeles waltl
Triturus pygmaeus
Reptiles
Acanthodactylus erythrurus
Blanus mariae
Chalcides bedriagai
Chalcides striatus
Chamaeleo chamaeleon
Coronella girondica
Emys orbicularis
Hemorrhois hippocrepis
Macroprotodon brevis
Malpolon monspessulanus
Mauremys leprosa
Natrix maura
Natrix natrix
Podarcis carbonelli
Podarcis vaucheri
Abundance
Species
Road-killed
history analysis. Below the table there is a legend for the acronyms used.
0
0
1
1
0
0
1
0
1
1
0
1
1
1
3
3
2
3
1
2
2
3
6.4
149.3
9.3
12.5
2.3
0.5
17.8
2.5
15.4
25.2
2.5
73.0
6920.0
671.0
2247.0
316.0
175.0
2500.0
600.0
9000.0
750.0
250.0
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
S/W
S/W
Fw
S/W
S/W
Fw
S/W
Fw
Fw
Fw
Fw
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
No
No
No
No
No
No
No
No
No
No
No
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Slow
Slow
Slow
Slow
Slow
Slow
Slow
Slow
Slow
Slow
Slow
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
1
1
1
1
1
1
0
1
1
1
0
1
0
0
1
2
1
1
2
2
2
2
2
2
2
2
2
1
2
2
6.2
4.1
3.1
36.6
53.6
30.0
596.0
125.8
26.1
165.0
658.2
50.0
92.2
1.8
3.1
9.0
1.0
3.5
8.0
20.0
7.0
10.5
7.5
2.0
9.0
7.0
7.0
30.0
6.7
3.0
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
S/W
S/W
S/W
S/W
S/W
S/W
Fw
S/W
S/W
S/W
Fw
Fw
Fw
S/W
Ua
Ins
Ins
Ins
Ins
Ins
Carn
Ins
Carn
Carn
Carn
Ins
Carn
Carn
Ins
Ins
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
Yes
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Fast
Slow
Fast
Fast
Slow
Fast
Slow
Fast
Fast
Fast
Slow
Fast
Fast
Fast
Fast
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
55
Psammodromus hispanicus
Psammodromus manuelae
Rhinechis scalaris
Tarentola mauritanica
Timon lepidus
Vipera latasti
Birds
Accipiter gentilis
Accipiter nisus
Acrocephalus arundinaceus
Acrocephalus scirpaceus
Aegithalos caudatus
Alauda arvensis
Alcedo atthis
Alectoris rufa
Anas platyrhynchos
Anthus campestris
Anthus pratensis
Anthus spinoletta
Anthus trivialis
Apus apus
Apus pallidus
Aquila adalberti
Asio flammeus
Asio otus
Athene noctua
Bubo bubo
Bubulcus ibis
Burhinus oedicnemus
Buteo buteo
Calandrella brachydactyla
Caprimulgus europaeus
Caprimulgus ruficollis
Carduelis cannabina
Carduelis carduelis
Carduelis chloris
Carduelis spinus
Cecropis daurica
Certhia brachydactyla
Cettia cetti
Ciconia ciconia
Circaetus gallicus
Circus aeruginosus
Circus cyaneus
Circus pygargus
Cisticola juncidis
Clamator glandarius
Coccothraustes coccothraustes
Columba livia domestica
Columba oenas
Columba palumbus
Coracias garrulus
Corvus corax
Corvus monedula
Coturnix coturnix
Cuculus canorus
56
0
1
1
1
1
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
1
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
2
2
2
2
2
1
1
2
3
1
3
2
2
3
2
3
2
2
3
3
1
1
1
2
2
3
2
2
2
1
2
1
2
3
2
2
2
2
3
1
2
1
1
2
1
1
3
1
2
1
2
2
2
2
1.7
3.8
128.4
8.0
100.0
40.0
8.0
16.2
9.5
3.7
13.5
3.2
Diu
Diu
Diu
Noct
Diu
Noct
S/W
S/W
S/W
Ua
S/W
S/W
Ins
Ins
Carn
Ins
Ins
Carn
No
No
No
Yes
Yes
Yes
Terr
Terr
Terr
Terr
Terr
Terr
Fast
Fast
Fast
Fast
Fast
Fast
Ecto
Ecto
Ecto
Ecto
Ecto
Ecto
792.5
204.0
28.3
12.7
7.8
33.0
31.6
497.5
1077.1
23.7
18.2
21.8
24.0
44.9
41.3
3000.0
320.0
265.0
157.0
1980.0
345.0
443.5
848.3
22.2
65.0
75.0
17.9
15.7
25.9
12.6
22.2
8.2
13.1
3471.7
1699.5
584.5
436.5
310.2
8.4
146.2
54.0
267.5
299.8
499.2
130.8
1131.2
237.5
101.5
109.0
3.5
5.0
9.0
10.0
10.0
16.0
9.7
11.2
11.0
0.0
0.0
0.0
0.0
2.5
2.4
2.5
0.0
4.1
3.0
3.0
4.5
2.0
3.0
3.0
3.0
4.0
10.0
10.0
10.0
8.0
11.2
13.0
10.0
4.0
1.0
5.5
0.0
4.5
12.5
18.0
3.0
10.0
0.0
1.5
4.0
5.0
5.2
10.5
9.2
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Noct
Noct
Noct
Diu
Noct
Diu
Diu
Noct
Noct
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
S/W
S/W
Fw
Fw
S/W
S/W
Fw
S/W
Fw
Gr
S/W
Fw
S/W
Ua
S/W
S/W
Gr
S/W
S/W
S/W
S/W
Gr
S/W
Gr
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Fw
S/W
S/W
Fw
Fa
Fa
Fw
S/W
S/W
Ua
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Carn
Carn
Ins
Ins
Ins
Herb
Carn
Herb
Herb
Ins
Ins
Ins
Ins
Ins
Ins
Carn
Carn
Carn
Ins
Carn
Ins
Ins
Carn
Ins
Ins
Ins
Herb
Herb
Herb
Herb
Ins
Ins
Ins
Carn
Carn
Carn
Carn
Carn
Ins
Ins
Herb
Herb
Herb
Herb
Ins
Carn
Ins
Herb
Ins
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
No
No
Yes
Yes
Yes
No
Yes
Yes
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Terr
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Terr
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Terr
Fly
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Cyanistes caeruleus
Cyanopica cooki
Delichon urbicum
Dendrocopos major
Elanus caeruleus
Emberiza calandra
Emberiza cia
Emberiza cirlus
Emberiza hortulana
Emberiza schoeniclus
Erithacus rubecula
Falco columbarius
Falco naumanni
Falco peregrinus
Falco subbuteo
Falco tinnunculus
Ficedula hypoleuca
Fringilla coelebs
Galerida cristata
Galerida theklae
Gallinago gallinago
Gallinula chloropus
Hieraaetus pennatus
Himantopus himantopus
Hippolais polyglotta
Hirundo rustica
Iduna opaca
Jynx torquilla
Lanius meridionalis
Lanius senator
Lophophanes cristatus
Lullula arborea
Luscinia megarhynchos
Melanocorypha calandra
Merops apiaster
Milvus migrans
Milvus milvus
Monticola saxatilis
Motacilla alba
Motacilla cinerea
Motacilla flava
Muscicapa striata
Oenanthe hispanica
Oenanthe oenanthe
Oriolus oriolus
Otus scops
Parus major
Passer domesticus
Passer hispaniolensis
Passer montanus
Petronia petronia
Phoenicurus ochruros
Phoenicurus phoenicurus
Phylloscopus bonelli
Phylloscopus collybita/ibericus
Phylloscopus trochilus
1
1
0
0
0
1
0
0
0
0
1
0
0
0
0
0
1
1
0
0
0
0
0
0
0
1
0
0
0
1
1
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
1
1
0
1
0
0
0
0
1
0
2
2
3
2
1
2
1
1
1
2
3
1
1
1
2
2
3
2
3
3
2
3
2
2
2
3
1
1
2
2
1
2
2
2
3
1
3
1
3
1
3
2
2
2
1
1
2
3
2
2
1
2
2
2
3
2
10.9
74.5
19.5
77.9
232.5
48.8
23.2
25.6
19.2
24.0
16.0
187.0
152.0
889.2
211.0
232.5
14.5
22.0
41.4
36.8
108.0
326.7
842.0
205.0
11.2
19.3
9.9
33.9
51.4
32.5
11.1
26.1
19.7
64.5
54.9
828.5
1039.7
58.0
21.8
17.8
17.4
17.5
17.8
24.2
70.3
85.7
17.5
28.5
28.2
20.8
30.5
16.2
14.5
7.0
7.5
8.0
9.3
6.0
8.0
5.5
3.5
10.0
0.0
7.0
0.0
0.0
0.0
4.0
4.0
3.5
3.0
4.5
0.0
4.5
10.0
7.0
0.0
14.0
2.0
4.0
4.5
11.2
5.2
12.7
6.0
5.5
6.5
8.0
9.0
9.0
6.5
2.5
2.0
0.0
0.0
0.0
10.0
5.0
9.0
0.0
3.5
4.5
9.7
10.0
10.0
15.0
11.0
0.0
0.0
0.0
0.0
0.0
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Noct
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Gr
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Gr
Fw
S/W
Fw
S/W
Ua
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Fa
Fa
Fa
S/W
S/W
S/W
S/W
S/W
S/W
Ua
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Carn
Herb
Ins
Ins
Carn
Herb
Herb
Herb
Herb
Herb
Ins
Carn
Ins
Carn
Carn
Carn
Ins
Herb
Ins
Ins
Ins
Herb
Carn
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Carn
Carn
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Herb
Herb
Herb
Herb
Ins
Ins
Ins
Ins
Ins
Yes
No
No
Yes
Yes
Yes
No
Yes
No
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Terr
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
57
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Pica pica
Prunella modularis
Pyrrhula pyrrhula
Regulus ignicapilla
Remiz pendulinus
Riparia riparia
Saxicola rubetra
Saxicola rubicola
Serinus serinus
Streptopelia decaocto
Streptopelia turtur
Strix aluco
Sturnus unicolor
Sturnus vulgaris
Sylvia atricapilla
Sylvia borin
Sylvia cantillans
Sylvia communis
Sylvia conspicillata
Sylvia hortensis
Sylvia melanocephala
Sylvia undata
Tachybaptus ruficollis
Tetrax tetrax
Troglodytes troglodytes
Turdus iliacus
Turdus merula
Turdus philomelos
Turdus pilaris
Turdus torquatus
Turdus viscivorus
Tyto alba
Upupa epops
Vanellus vanellus
Mammals
Apodemus sylvaticus
Arvicola sapidus
Bos taurus
Canis lupus familiaris
Cervus elaphus
Crocidura russula
Dama dama
Eliomys quercinus
Eptesicus isabellinus
Equus caballus
Erinaceus europaeus
Felis silvestris catus
Felis silvestris silvestris
Genetta genetta
Herpestes ichneumon
Lepus granatensis
Lutra lutra
Meles meles
Microtus duodecimcostatus
Miniopterus schreibersii
Mus musculus
58
0
0
0
0
0
0
1
1
1
0
0
1
0
0
1
0
1
0
0
0
1
1
0
0
1
0
0
1
0
0
0
0
0
0
1
0
0
1
0
1
0
1
0
0
1
1
0
0
1
1
0
0
0
0
0
2
2
1
2
1
2
2
2
3
2
2
2
3
3
2
2
2
2
1
1
2
2
2
1
2
1
2
2
1
1
1
2
2
2
2
2
3
2
3
2
2
2
1
3
2
2
1
2
2
2
1
2
2
1
2
198.7
19.2
22.4
5.4
9.2
12.9
17.7
15.0
11.9
198.5
132.0
459.8
86.6
77.5
18.0
20.0
9.3
16.5
8.7
21.1
10.9
9.4
162.5
775.0
8.8
60.0
86.1
72.2
91.0
107.5
117.8
271.2
68.6
217.5
6.1
0.0
0.0
0.0
14.0
10.0
0.0
12.5
5.2
4.5
2.0
3.5
9.0
0.0
5.0
0.0
0.0
0.0
8.0
0.0
8.0
8.0
10.0
3.5
13.0
0.0
10.0
0.0
0.0
0.0
10.0
8.2
7.5
4.0
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Noct
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Diu
Noct
Diu
Diu
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Fw
Fa
S/W
S/W
S/W
S/W
S/W
S/W
Gr
S/W
S/W
Gr
Ins
Ins
Herb
Ins
Ins
Ins
Ins
Ins
Herb
Herb
Herb
Carn
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Herb
Ins
Ins
Ins
Ins
Ins
Ins
Ins
Carn
Ins
Ins
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
No
Yes
Yes
Yes
Yes
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Terr
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Fly
Terr
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
32.5
205.0
450000.0
13500.0
72000.0
9.3
51500.0
85.0
22.5
400000.0
850.0
3750.0
4250.0
1875.0
2982.5
2300.0
11350.0
11950.0
25.5
12.5
18.5
13.5
9.9
1.0
5.3
1.0
15.7
1.0
11.1
1.0
1.0
5.5
4.0
3.0
2.5
2.5
1.5
2.5
2.5
18.0
1.0
34.0
Noct
Noct
Diu
Diu
Noct
Noct
Noct
Noct
Noct
Diu
Noct
Noct
Noct
Noct
Diu
Noct
Noct
Noct
Noct
Noct
Noct
S/W
Fw
Gr
Ua
S/W
S/W
Gr
S/W
S/W
Gr
Fa
Ua
S/W
S/W
S/W
Gr
Fw
S/W
S/W
S/W
Ua
Herb
Herb
Herb
Carn
Herb
Ins
Herb
Ins
Ins
Herb
Ins
Carn
Carn
Carn
Carn
Herb
Carn
Carn
Herb
Ins
Herb
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
Yes
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Fly
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Terr
Fly
Terr
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Mus spretus
Mustela putorius
Nyctalus lasiopterus
Nyctalus leisleri
Oryctolagus cuniculus
Pipistrellus kuhlii
Pipistrellus pipistrellus
Rattus norvegicus
Rattus rattus
Suncus etruscus
Sus scrofa
Tadarida teniotis
Vulpes vulpes
1
0
0
0
1
0
1
0
1
0
0
0
1
2
1
1
1
3
1
1
2
2
2
3
1
2
13.0
940.0
58.0
16.5
1200.0
7.5
14.0
300.0
175.0
1.9
49000.0
38.0
6000.0
34.8
7.5
1.0
1.0
7.1
2.0
1.5
81.2
13.4
21.0
4.2
1.0
4.5
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
Noct
S/W
S/W
S/W
S/W
S/W
S/W
S/W
Ua
S/W
S/W
S/W
S/W
S/W
Herb
Carn
Carn
Ins
Herb
Ins
Ins
Carn
Herb
Ins
Herb
Ins
Carn
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Terr
Terr
Fly
Fly
Terr
Fly
Fly
Terr
Terr
Terr
Terr
Fly
Terr
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Fast
Legend: Abundance: 1 = Rare species; 2 = Common species; 3 = Abundant species (at least during a
period of the year). Activity: Diu = diurnal species; Noct = nocturnal species. Habitat preference: Ua =
urban areas; Fa = farmlands; Gr = grasslands; S/W = Mediterranean scrubland/woodland; Fw =
freshwaters. Food habits: Carn = carnivorous; Ins = insectivorous; Herb = herbivorous. Movement: Terr
= terrestrial species; Fly = flying species. Speed: Fast = fast species; Slow = slow species. Metabolic
strategy: Ecto = ectotherm species; Endo = endotherm species.
The checklist was compiled from the following references: Castroviejo, 1993;
Ibáñez et al., 1995; Blanco, 1998a, 1998b; García et al., 2000; Arnold and Ovenden,
2002; Mullarney et al., 2003; Díaz-Paniagua et al., 2005; Ministerio de Medio
Ambiente y Medio Rural y Marino, 2007a, 2007b; Salvador and Cassinello, 2010;
Salvador and Marco, 2010.
The values of predictors for life history analysis were obtained from the
bibliography: Valverde, 1958, 1967; Jaksic et al., 1982; Castroviejo, 1993; Ibáñez et al.,
1995; Blanco, 1998a, 1998b; Cramp, 1998; Salvador, 1998; García et al., 2000; Arnold
and Ovenden, 2002; Díaz-Paniagua et al., 2005; Pérez-Santigosa, 2007; Salvador and
Marco, 2010; Salvador and Martínez-Solano, 2010; and also using unpublished data of
ESPN.
59
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Endo
Bibliographic review for the selection of candidate predictors for Life history
analyses
Abundance: Abundant species can be more road-killed than rare species in birds
and mammals (Ford and Fahrig, 2007; Barthelmess and Brooks, 2010; Møller et al.,
2011). Size: Small mammals can be more road-killed than large mammals (Caceres,
2011), but medium-sized mammals can be more road-killed than these categories (Ford
and Fahrig, 2007; Barthelmess and Brooks, 2010). Small and large passerines can be
road-killed with different frequency in different habitats (Ramos et al., 2011). Breding:
Juveniles can be more road-killed than adults in reptiles, birds and mammals (Mumme
et al., 2000; Ciesiołkiewicz et al., 2006; Conard and Gipson, 2006). Activity: Nocturnal
species can be more road-killed than diurnal species (Caceres, 2011; Cook and
Blumstein, 2013). Habitat preference: Road-kill frequency can be different in different
habitats (Clevenger et al., 2003; Alves da Rosa and Bager, 2012). Food habits: Food
habits can affect road-kill probability of birds and mammals (Ford and Fahrig, 2007;
Barthelmess and Brooks, 2010; Cook and Blumstein, 2013). Territoriality: Juvenile
dispersal has been often associated with road-kill risk in reptiles, mammals and birds,
because juvenile individuals have to leave the parental territory, frequently moving
through highly anthropized matrix habitats (Bonnet et al., 1999; Massemin et al., 1998;
Grilo et al., 2009). For this reason we hypothesized that territorial species can be more
road-killed than non-territorial species. Movement: Terrestrial and flying species can be
differently affected by road mortality (Teixeira et al., 2013). Speed: Slowness has been
suggested to increase road-kill probability of amphibians, reptiles and mammals (Puky,
60
2005; Laporte et al., 2013). Metabolic strategy: Ectotherms can be more road-killed
than endotherms (Puky, 2005; Andrews et al., 2006).
Correlations between Life history candidate predictors
Table 2: Correlations between continuous candidate predictors. The performed analyses were
Pearson correlations between every possible pair of continuous candidate predictors (CORR procedure in
SAS). Strong correlations were considered when r > 0.75.
Size
Breeding
Size
1
-
Breeding
0.79
1
Table 3: Correlations between categorical candidate predictors. The performed analyses were
contingency tables between every possible pair of categorical candidate predictors (FREQ procedure in
SAS). When the 50% of dataset cells had expected counts less than 5, the contingency tables could not be
performed. In these cases the FREQ procedure provided a Fisher’s exact test (with F in the table). Fisher’s
exact test can be performed only for binary predictors, thus in the case in which one of the predictors was
not binary, we carried out a GLM (GLIMMIX procedure in SAS, binomial error distribution and logit
link function; glm in the table; in the cases of absence of binary predictors we applied poisson error
distribution and log link function). Significant variables (p < 0.05) are marked with an asterisk.
Abundance
Abundance
Food habits
Territoriality
Activity
Habitat preference
Movement
Speed
Metabolic strategy
-
Food
habits
* 0.04
-
Territoriality
Activity
* 0.02
* 0.048
-
0.60
0.47
0.34
-
Habitat
preference
0.77 glm
0.88
0.31glm
0.58 glm
-
Movement
*
0.04
0.11
0.07
* < 0.0001
* 0.007 glm
-
Speed
0.44 glm
0.99 glm
* < 0.0001 F
* < 0.0001 F
* 0.002 glm
* < 0.0001 F
-
61
Metabolic
strategy
0.61
* 0.007
* < 0.0001
* 0.003
* 0.02 glm
* < 0.0001
* < 0.0001 F
-
Table 4: Correlations between categorical and continuous candidate predictors. P-values of the
performed analyses were GLMs between every possible pair of categorical and continuous candidate
predictors (GLIMMIX procedure in SAS, binomial error distribution and logit link function; in the cases
of absence of binary predictors we applied poisson or negative binomial error distribution and log link
function). Significant variables (p < 0.05) are marked with an asterisk.
Abundance
Food habits
Territoriality
Activity
Habitat preference
Movement
Speed
Metabolic strategy
Size
* < 0.0001
* < 0.0001
0.11
0.82
* < 0.0001
* 0.01
0.36
0.09
Breeding
0.38
* < 0.0001
0.06
* 0.03
* < 0.0001
* 0.01
* 0.007
* 0.02
References
Alves da Rosa, C. and Bager, A. (2012). Seasonality and habitat types affect roadkill of neotropical birds.
Journal of Environmental Management. 97: 1-5.
Barthelmess, E.L. and Brooks, M.S. (2010). The influence of body-size and diet on road-kill trends in
mammals. Biodiversity and Conservation. 19: 1611-1629.
Arnold, N. and Ovenden, D. (2002). Reptiles y Anfibios. Guía de campo. Todas las especies de España y
de Europa [Reptiles and Amphibians. Field guide. All species of Spain and Europe]. Ediciones Omega
S.A., 319 pp., Barcelona (Spain).
Blanco, J.C. (1998a). Mamíferos de España (Volume I): Insectívoros, Quirópteros, Primates y Carnívoros
de la Península ibérica, Baleares y Canarias [Mammals of Spain (Volumen I): Insectivores,
Chiropterans, Primates and Carnivores of Iberian Peninsula, Balearic and Canary Islands].
GeoPlaneta, 457 pp., Barcelona (Spain).
Blanco, J.C. (1998b). Mamíferos de España (Volume II): Cetáceos, Artiodáctilos, Roedores y
Lagomorfos de la Península ibérica, Baleares y Canarias [Mammals of Spain (Volumen II): Cetaceans,
Artiodactylans, Rodents and Lagomorphs of Iberian Peninsula, Balearic and Canary Islands].
GeoPlaneta, 383 pp., Barcelona (Spain).
Caceres, N.C. (2011). Biological characteristics influence mammal road kill in an Atlantic Forest-Cerrado
interface in south-western Brazil. Italian Journal of Zoology. 78: 379-389.
Castroviejo, J. (1993). Memoria del Mapa del Parque Nacional de Doñana [Essay on the Map of Doñana
National Park]. Consejo Superior de Investigaciones Cientifícas (CSIC) y Agencia de Medio
Ambiente (AMA) de la Junta de Andalucia, 133 pp., Seville (Spain).
Ciesiołkiewicz, J., Orłowski, G. and Elżanowski, A. (2006). High juvenile mortality of grass snakes
Natrix natrix (L.) on a suburban road. Polish Journal of Ecology. 54: 465-472.
62
Conard, J.M. and Gipson, P.S. (2006). Spatial and seasonal variation in wildlife-vehicle collisions. The
Prairie Naturalist. 38: 251-260.
Cook, T.C. and Blumstein, D.T. (2013). The omnivore’s dilemma: diet explains variation in vulnerability
to vehicle collision mortality. Biological Conservation. 167: 310-315.
Cramp, S. (1998). The Complete Birds of the Western Palearctic. Oxford University Press. CD-ROM.
Oxford (UK).
Díaz-Paniagua, C., Gómez-Rodríguez, C., Portheault, A. and de Vries, W. (2005). Los Anfibios de
Doñana [The Amphibians of Doñana]. Organismo Autónomo de Parques Nacionales (Ministerio de
Medio Ambiente), 181 pp., Madrid (Spain).
Ford, A.T. and Fahrig, L. (2007). Diet and body size of North American mammal road mortalities.
Transportation Research Part D: Transport and Environment. 12: 498-505.
García, L., Ibáñez, F., Garrido, H., Arroyo, J.L., Máñez, M. and Calderón, J. (2000). Prontuario de las
aves de Doñana. Anuario ornitológico de Doñana Nº 0 [Short guide of Doñana Birds. Ornithological
Doñana Yearbook Nº 0]. Estación Biológica de Doñana (EBD - CSIC). Ayuntamiento de Almonte,
113 pp., Almonte (Spain).
Grilo, C., Bissonette, J. and Santos-Reis, M. (2009). Spatial-temporal patterns in Mediterranean carnivore
road casualties: consequences for mitigation. Biological Conservation. 142: 301-313.
Ibáñez, C., Guillén, A., Juste, J., Migens, E., Pérez-Jordá, J.L. and Ruiz, C. (1995). Quirópteros del
Parque Nacional de Doñana: especies, y tamaño y métodos de estima de sus poblaciones
[Chiropterans of Doñana National Park: species, and their population sizes and estimate methods].
Grupo de Investigación de Sistemática y Ecología de Quirópteros (EBD - CSIC), 114 pp., Seville
(Spain).
Jaksic, F.M., Greene, H.W., Schwenk, K. and Seib, R.L. (1982). Predation upon reptiles in Mediterranean
habitats of Chile, Spain and California: a comparative analysis. Oecologia. 53: 152-159.
Laporte, M., Silva Beaudry, C.-O. and Angers, B. (2013). Effects of road proximity on genetic diversity
and reproductive success of the painted turtle (Chrysemys picta). Conservation Genetics. 14: 21-30.
Massemin, S., Le Maho, Y. and Handrich, Y. (1998). Seasonal pattern in age, sex and body condition of
barn owls Tyto alba killed on motorways. Ibis. 140: 70-75.
Ministerio de Medio Ambiente y Medio Rural y Marino (2007a). Mamíferos [Mammals]:
http://www.mma.es/portal/secciones/biodiversidad/inventarios/inb/inventario_vertebrados/mamiferos/
index.htm [online]. In: Inventario Nacional de Biodiversidad, Vertebrados. Madrid (Spain).
Ministerio de Medio Ambiente y Medio Rural y Marino (2007). Reptiles [Reptiles]:
http://www.mma.es/portal/secciones/biodiversidad/inventarios/inb/inventario_vertebrados/index.htm
[online]. In: Inventario Nacional de Biodiversidad, Vertebrados. Madrid (Spain).
Møller, A.P., Erritzøe, H. and Erritzøe, J. (2011). A behavioral ecology approach to traffic accidents:
interspecific variation in causes of traffic casualties among birds. Zoological Research. 32: 115-127.
Mullarney, K., Svensson, L., Zetterström, D. and Grant, P.J. (2003). Guía de Aves. La guía de campo de
Aves de España y de Europa más completa [Bird Guide. The most complete Field Guide to the Birds
of Spain and Europe]. Ediciones Omega S.A., 400 pp., Barcelona (Spain).
501-512.
Pérez-Santigosa, N. (2007). Ecología del Galápago exótico, Trachemys scripta elegans, en la Península
Ibérica. Efectos sobre las poblaciones autóctonas de Mauremys leprosa y Emys orbicularis [Ecology
of the invasive pond turtle, Trachemys scripta elegans, in the Iberian Peninsula. Effects on the
autochthonous populations of Mauremys leprosa and Emys orbicularis]. University of Seville (Spain).
325-338.
Ramos, C., Pereira de Lima Júnior, D., Zawadzki, C.H. and Benedito, E. (2011). The biology and ecology
of birds is an important factor for explain the road kill frequencies? Neotropical Biology and
Salvador, A. and Marco, A. (2010). Reptiles [Reptiles]: http://www.vertebradosibericos.org/reptiles.html
[online]. In: Enciclopedia Virtual de los Vertebrados Españoles. Salvador, A. (Ed.). Sociedad de
63
Amigos del MNCN (SAM) y Museo Nacional de Ciencias Naturales (MNCN - CSIC), Madrid
(Spain).
Salvador,
A.
and
Martínez-Solano,
I.
(2010).
Anfibios
[Amphibians]:
http://www.vertebradosibericos.org/anfibios.html [online]. In: Enciclopedia Virtual de los
Vertebrados Españoles. Salvador, A. (Ed.). Sociedad de Amigos del MNCN (SAM) y Museo
Nacional de Ciencias Naturales (MNCN - CSIC), Madrid (Spain).
Teixeira, F.Z., Coelho, I.P., Esperandio, I.B., da Rosa Oliveira, N., Porto Peter, F., Dornelles, S.S.,
Delazeri, N.R., Tavares, M., Borges Martins, M. and Kindel, A. (2013). Are road-kill hotspots
coincident among different vertebrate groups? Oecologia Australis. 17: 36-47.
Valverde, J.A. (1958). An ecological sketch of the Coto Doñana. British Birds. 51: 1-23.
Valverde, J.A. (1967). Estructura de una comunidad Mediterránea de vertebrados terrestres [Structure of
a Mediterranean community of terrestrial vertebrates]. Consejo Superior de Investigaciones
Científicas (CSIC), 192 pp., Madrid (Spain).
64
Appendix 2 – Temporal patterns
Road-killed species included in Temporal groups
Amphibians: Discoglossus galganoi, Epidalea calamita, Pelobates cultripes,
Pelophylax perezi, Pleurodeles waltl. Snakes: Coronella girondica, Hemorrhois
hippocrepis, Macroprotodon brevis, Malpolon monspessulanus, Natrix maura,
Rhinechis scalaris, Vipera latasti. Lizards: Acanthodactylus erythrurus, Podarcis
carbonelli, Podarcis vaucheri, Psammodromus manuelae, Tarentola mauritanica,
Timon lepidus. Resident birds: Alectoris rufa, Bubo bubo, Buteo buteo, Columba livia
domestica, Cyanistes caeruleus, Cyanopica cooki, Emberiza calandra, Fringilla
coelebs, Lophophanes cristatus, Lullula arborea, Parus major, Passer domesticus,
Passer
montanus,
Saxicola
rubicola,
Serinus
serinus,
Strix
aluco,
Sylvia
melanocephala, Sylvia undata, Troglodytes troglodytes. Migrant nesting birds: Apus
pallidus, Caprimulgus ruficollis, Hirundo rustica, Lanius senator. Migrant wintering
birds: Erithacus rubecula, Turdus philomelos. Transient migrant birds: Ficedula
hypoleuca, Muscicapa striata, Phylloscopus collybita/ibericus, Saxicola rubetra, Sylvia
atricapilla, Sylvia cantillans. Small mammals: Apodemus sylvaticus, Crocidura
russula, Eliomys quercinus, Mus spretus, Rattus rattus.
65
Bibliographic review for the selection of candidate predictors for the temporal
analysis
Photoperiod variation: Temporal distribution of road-kills can vary according to
seasonal behaviors (Bonnet et al., 1999; Clevenger et al., 2001; Gryz and Krauze, 2008;
Smith-Patten and Patten, 2008) which can be related with photoperiod variation (Fryxell
and Sinclair, 1988; Lloyd, 1999; Gwinner and Helm, 2003). Rainfall: Amphibian roadkill peaks can coincide with rainfall (Puky, 2005; Glista et al., 2008). Temperature:
Temporal distribution of reptile road-kills can vary according to temperature (Andrews
et al., 2006). Seasonal traffic: Temporal distribution of road-kills can vary according to
traffic variation (Forman and Alexander, 1998; Trombulak and Frissell, 2000; Forman
et al., 2003). Time since previous survey: Temporal distribution of road-kills can be
affected by carcass persistence (Antworth et al., 2005; Santos et al., 2011).
Correlations between Temporal candidate predictors
Table 1: Correlations between continuous candidate predictors. The performed analyses were
Pearson correlations between every possible pair of continuous candidate predictors (CORR procedure in
SAS).
Photoperiod variation
Rainfall
Temperature
Seasonal traffic
Time since previous survey
66
Photoperiod
variation
1
-
Rainfall
Temperature
-0.29
1
-
-0.33
-0.10
1
-
Seasonal
traffic
0.001
-0.06
0.12
1
-
Time since
previous survey
-0.17
-0.02
0.18
0.06
1
Monthly road-kill patterns for every temporal group
Figure 1: Monthly road-kill pattern for amphibians. Casualties per month during the whole
study period.
Figure 2: Monthly road-kill pattern for snakes. Casualties per month during the whole study
period.
67
Figure 3: Monthly road-kill pattern for lizards. Casualties per month during the whole study
period.
Figure 4: Monthly road-kill pattern for resident birds. Casualties per month during the whole
study period.
68
Figure 5: Monthly road-kill pattern for migrant nesting birds. Casualties per month during
the whole study period.
Figure 6: Monthly road-kill pattern for migrant wintering birds. Casualties per month
during the whole study period.
69
Figure 7: Monthly road-kill pattern for transient migrant birds. Casualties per month during
the whole study period.
Figure 8: Monthly road-kill pattern for small mammals. Casualties per month during the
whole study period.
70
Factors affecting road-kill temporal patterns
Figure 9: Temporal road-kill patterns. Representations with raw data of significant
correlations between road-kills and temporal predictors. In the upper two graphs the significant
correlations between road-kills and rainfall; in the lower five graphs the significant correlations between
road-kills and temperature.
71
References
Antworth, R.L., Pike, D.A. and Stevens, E.E. (2005). Hit and run: effects of scavenging on estimates of
roadkilled vertebrates. Southeastern Naturalist. 4: 647-656.
Clevenger, A.P., McIvor, M., McIvor, D., Chruszcz, B. and Gunson K. (2001). Tiger salamander,
Ambystoma tigrinum, movements and mortality on the Trans-Canada highway in southwestern
Alberta. Canadian Field-Naturalist. 115: 199-204.
Fryxell, J.M. and Sinclair, A.R.E. (1988). Causes and consequences of migration by large herbivores.
Gryz, J. and Krauze, D. (2008). Mortality of vertebrates on a road crossing the Biebrza Valley (NE
Poland). European Journal of Wildlife Research. 54: 709-714.
Gwinner, E. and Helm, B. (2003). Circannual and circadian contributions to the timing of avian
migration. In: Avian Migration. Berthold, P., Gwinner, E. and Sonnenschein, E. (Eds.). Springer
Berlin Heidelberg (Germany), pp. 81-95.
Lloyd, P. (1999). Rainfall as a breeding stimulus and clutch size determinant in South African arid-zone
birds. Ibis. 141: 637-643.
325-338.
[online].
72
Appendix 3 – Spatial patterns
Road-killed species included in spatial groups
Amphibians: Discoglossus galganoi, Epidalea calamita, Pelobates cultripes,
Pelophylax perezi, Pleurodeles waltl. Snakes: Coronella girondica, Hemorrhois
hippocrepis, Macroprotodon brevis, Malpolon monspessulanus, Natrix maura,
Rhinechis scalaris, Vipera latasti. Lizards: Acanthodactylus erythrurus, Podarcis
carbonelli, Podarcis vaucheri, Psammodromus manuelae, Tarentola mauritanica,
Timon lepidus. Small passerines: Cyanistes caeruleus, Emberiza calandra, Erithacus
rubecula, Ficedula hypoleuca, Fringilla coelebs, Lophophanes cristatus, Lullula
arborea, Muscicapa striata, Parus major, Passer domesticus, Passer montanus,
Phylloscopus collybita/ibericus, Saxicola rubetra, Saxicola rubicola, Serinus serinus,
Strix aluco, Sylvia atricapilla, Sylvia cantillans, Sylvia melanocephala, Sylvia undata,
Troglodytes troglodytes. Small mammals: Apodemus sylvaticus, Crocidura russula,
Eliomys quercinus, Mus spretus, Rattus rattus.
Bibliographic review for the selection of candidate predictors for the spatial
analyses
Macrohabitat: Road-kill frequency can be different in different habitats
(Andrews et al., 2006; Alves da Rosa and Bager, 2012; Clevenger et al., 2003).
Distance to water: Amphibians can be mainly road-killed close to water bodies
(Trombulak and Frissell, 2000; Puky, 2005). Verge habitat: Road-kill frequency can be
73
different in correspondence to different road verge habitats (Saeki and Macdonald,
2004; Orlowski, 2008; Meek, 2009). Fences: Road-kill frequency can decrease in
correspondence to road-fences (Aresco, 2005; Bager and Fontoura, 2013). Culverts:
Road-kill frequency can decrease in correspondence to drainage culverts (Clevenger et
al., 2001; Aresco, 2005; Dodd et al., 2004). Crossing structures: Road-kill frequency
can decrease in correspondence to wildlife road-crossing structures (Clevenger and
Waltho, 2000; Grilo et al., 2008; Mata et al., 2008). Road width: Road-kill frequency
can decrease or increase according to road width increase (Forman et al., 2003; SmithPatten and Patten, 2008; van Langevelde et al., 2009). Traffic: Road-kill frequency can
decrease or increase according to traffic increase (Forman et al., 2003; Coelho et al.,
2008; van Langevelde et al., 2009). Speed: Road-kill frequency can decrease or increase
according to the increase of maximum allowed speed (Forman et al., 2003; Hobday and
Minstrell, 2008; van Langevelde et al., 2009). Shoulders: Road-kill frequency can
increase in correspondence to road-shoulders (Ashley and Robinson, 1996; Tanner and
Perry, 2007). Curves: Road-kill frequency can decrease or increase in correspondence
to curves (Jones, 2000; Forman et al., 2003; Snow et al., 2011). Slope changes: Roadkill frequency can increase in correspondence to slope changes in road (Forman et al.,
2003; Snow et al., 2011). Vertical signs: Road-kill frequency can decrease in
correspondence to road signs (Hobday and Minstrell, 2008; Glista et al., 2009).
Horizontal signs: Road-kill frequency can decrease in correspondence to road signs
(Hobday and Minstrell, 2008; Glista et al., 2009).
74
Correlations between Spatial candidate predictors
Table 1: Correlations between continuous candidate predictors. The analyses were Pearson
correlations between every possible pair of continuous candidate predictors regarding road and landscape
(CORR procedure in SAS). Correlations above r > 0.75 are marked with an asterisk.
Road width
Traffic
Speed
Distance from water
Road
width
1
-
Traffic
Speed
* 0.99
1
-
* 0.89
* 0.89
1
-
Distance
to water
0.19
0.14
0.01
1
Table 2: Correlations between categorical candidate predictors regarding road. The performed
analyses were contingency tables between every possible pair of categorical candidate predictors
regarding road (FREQ procedure in SAS). When 50% of dataset cells had expected counts below 5, the
contingency tables could not be performed. In these cases the FREQ procedure provided a Fisher’s exact
test (with F in the table). Figures are p-values, marked with an asterisk when p < 0.05.
Shoulder
Shoulder
Slope changes
Curves
Vertical signs
Horizontal signs
-
Slope
changes
* 0.03
-
Curves
0.66
0.25 F
-
Vertical
signs
* 0.01
0.25 F
0.47
-
Horizontal
signs
* < 0.0001
0.26 F
* 0.003
* < 0.0001
-
75
Table 3: Correlations between categorical candidate predictors regarding road verges. The analyses
were contingency tables between every possible pair of categorical candidate predictors regarding road
verges (FREQ procedure in SAS). When 50% of dataset cells had expected counts below 5, the
contingency tables could not be performed. In these cases the FREQ procedure provided a Fisher’s exact
test (with F in the table). Figures are p-values, marked with an asterisk when p < 0.05.
Verge habitat
Fences
Culverts
Crossing structures
Verge
habitat
-
Fences
Culverts
* < 0.0001
-
* 0.03
* 0.009
-
Crossing
structures
* 0.0007
* 0.0004
0.35 F
-
Table 4: Correlations between categorical candidate predictors regarding road and road verges.
The performed analyses were contingency tables between every possible pair of categorical candidate
predictors regarding road and road verges (FREQ procedure in SAS). When 50% of dataset cells had
expected counts less than 5, the contingency tables could not be performed. In these cases the FREQ
procedure provided a Fisher’s exact test (with F in the table). Fisher’s exact test can be performed only for
binary predictors, thus in the case in which one of the predictors was not binary, we carried out a GLM
(GLIMMIX procedure in SAS, binomial error distribution and logit link function; glm in the table).
Figures are p-values, marked with an asterisk when p < 0.05.
Shoulder
Verge habitat
Fences
Culverts
Crossing structures
76
* < 0.0001
* < 0.0001
* 0.04
* 0.0024
Slope
changes
* 0.01 glm
0.23
0.33 F
* 0.0001
Curves
* 0.02
0.19
0.13 F
* 0.03 F
Vertical
signs
0.34
* < 0.0001
0.23
* 0.02
Horizontal
signs
* 0.0007
* < 0.0001
0.18 F
0.21 F
Table 5: Correlations between categorical candidate predictors regarding road, road verges and
landscape. The performed analyses were contingency tables between every possible pair of categorical
candidate predictors regarding road, roadside and landscape (FREQ procedure in SAS). When 50% of
dataset cells had expected counts less than 5, the contingency tables could not be performed. In these
cases the FREQ procedure provided a Fisher’s exact test. Fisher’s exact test can be performed only for
binary predictors, thus in the case in which one of the predictors was not binary, we carried out a GLM
(GLIMMIX procedure in SAS, binomial error distribution and logit link function; glm in the table).
Figures are p-values, marked with an asterisk when p < 0.05.
Shoulder
Slope changes
Curves
Vertical signs
Horizontal signs
Verge habitat
Fences
Culverts
Crossing structures
Macrohabitat
* < 0.0001
* 0.06 glm
* < 0.0001
* < 0.0001
* < 0.0001
* < 0.0001
* < 0.0001
* 0.004
* 0.39 glm
Table 6: Correlations between categorical and continuous candidate predictors regarding road,
road verges and landscape. The performed analyses were GLMs between every possible pair of
categorical and continuous candidate predictors regarding road, road verges and landscape (GLIMMIX
procedure in SAS, binomial error distribution and logit link function; in the cases of absence of binary
predictors we applied poisson error distribution and log link function). Figures are p-values, marked with
an asterisk when p < 0.05.
Shoulder
Slope changes
Curves
Vertical signs
Horizontal signs
Verge habitat
Fences
Culverts
Crossing structures
Macrohabitat
Road
width
* < 0.0001
* 0.04
0.15
* 0.0003
* < 0.0001
* < 0.0001
* < 0.0001
* < 0.0001
0.47
* < 0.0001
Traffic
Speed
* < 0.0001
* 0.04
0.08
* 0.001
* < 0.0001
* < 0.0001
* < 0.0001
* < 0.0001
0.53
* < 0.0001
* < 0.0001
* 0.02
0.51
* < 0.0001
* < 0.0001
* < 0.0001
0.22
* < 0.0001
0.23
* < 0.0001
Distance
to water
* < 0.0001
0.43
* < 0.0001
0.90
* < 0.0001
* < 0.0001
* < 0.0001
0.10
0.87
* < 0.0001
77
Road-kill distribution for every spatial group
Figure 1: Road-kill distribution for amphibians. Black lines represent the paved road-network. The
surveyed roads (regional road A494 Matalascañas-Mazagón, forestry/agricultural road of Cabezudos,
Condesa) were divided into 50-m long sections in which we sampled the road-kills. These sections are
showed in this figure with different width, according with the number of road-killed amphibians
accumulated during the study (range 0-20).
78
Figure 2: Road-kill distribution for snakes. Black lines represent the paved road-network. The
showed in this figure with different width, according with the number of road-killed snakes accumulated
during the study (range 0-4).
79
Figure 3: Road-kill distribution for lizards. Black lines represent the paved road-network. The
showed in this figure with different width, according with the number of road-killed lizards accumulated
during the study (range 0-4).
80
Figure 4: Road-kill distribution for small passerines. Black lines represent the paved road-network.
The surveyed roads (regional road A494 Matalascañas-Mazagón, forestry/agricultural road of Cabezudos,
showed in this figure with different width, according with the number of road-killed small passerines
81
Figure 5: Road-kill distribution for small mammals. Black lines represent the paved road-network.
The surveyed roads (regional road A494 Matalascañas-Mazagón, forestry/agricultural road of Cabezudos,
showed in this figure with different width, according with the number of road-killed small mammals
82
Factors affecting road-kill spatial patterns.
Figure 6: Road-kill spatial distribution and road magnitude. Representations with raw data of
significant correlations between road-kills and road magnitude. Road magnitude range corresponds with
5-10 m of Road width (average = 8.79 m), 171-2499 car/day of Traffic (average = 1547 car/day) and 30100 Km/h of Speed (average = 78.69 Km/h).
References
Alves da Rosa, C. and Bager, A. (2012). Seasonality and habitat types affect roadkill of neotropical birds.
Journal of Environmental Management. 97: 1-5.
Ashley, E.P. and Robinson, J.T. (1996). Road mortality of amphibians, reptiles and other wildlife on the
Long Point Causeway, Lake Erie, Ontario. Canadian Field-Naturalist. 110: 404-412.
Aresco, M.J. (2005). Mitigation measures to reduce highway mortality of turtles and other herpetofauna
at a north Florida lake. Journal of Wildlife Management. 69: 549-560.
Bager, A. and Fontoura, V. (2013). Evaluation of the effectiveness of a wildlife roadkill mitigation
system in wetland habitat. Ecological Engineering. 53: 31-38.
Clevenger, A.P., Chruszcz, B. and Gunson, K. (2001). Drainage culverts as habitat linkages and factors
affecting passage by mammals. Journal of Applied Ecology. 38: 1340-1349.
Clevenger, A.P. and Waltho, N. (2000). Factors influencing the effectiveness of wildlife underpasses in
Banff National Park, Alberta, Canada. Conservation Biology. 14: 47-56.
Coelho, I.P., Kindel, A. and Coelho, A.V.P. (2008). Roadkills of vertebrate species on two highways
through the Atlantic Forest Biosphere Reserve, southern Brazil. European Journal of Wildlife
Research. 54: 689-699.
83
Dodd, C.K.J., Barichivich, W.J. and Smith, L.L. (2004). Effectiveness of a barrier wall and culverts in
reducing Wildlife mortality on a heavily traveled highway in Florida. Biological Conservation. 118:
619-631.
Grilo, C., Bissonette, J. and Santos-Reis, M. (2008). Response of carnivores to existing highway culverts
and underpasses: implications for road planning and mitigation. Biodiversity and Conservation. 17:
1685-1699.
Hobday, A.J. and Minstrell, M.L. (2008). Distribution and abundance of roadkill on Tasmanian
highways: human management options. Wildlife Research. 35: 712-726.
Mata, C., Hervás, I., Herranz, J., Suárez, F. and Malo, J.E. (2008). Are motorway wildlife passages worth
building? Vertebrate use of road-crossing structures on a Spanish motorway. Journal of
Environmental Management. 88: 407-415.
Meek, R. (2009). Patterns of reptile road-kills in the Vendée region of western France. The
Herpetological Journal. 19: 135-142.
Orlowski, G. (2008). Roadside hedgerows and trees as factors increasing road mortality of birds:
implications for management of roadside vegetation in rural landscapes. Landscape and Urban
Planning. 86: 153-161.
325-338.
Saeki, M. and Macdonald, D.W. (2004). The effects of traffic on the raccoon dog (Nyctereutes
procyonoides viverrinus) and other mammals in Japan. Biological Conservation. 118: 559-571.
Snow, N.P., Andelt, W.F. and Gould, N.P. (2011). Characteristics of road-kill locations of San Clemente
Island foxes. Wildlife Society Bulletin. 35: 32-39.
Tanner, D. and Perry, J. (2007). Road effects on abundance and fitness of Galápagos lava lizards
(Microlophus albemarlensis). Journal of Environmental Management. 85: 270-278.
van Langevelde ,F., van Dooremalen, C. and Jaarsma, C.F. (2009). Traffic mortality and the role of minor
roads. Journal of Environmental Management. 90: 660-667.
84
85
Capítulo 2:
Efecto Barrera
¿Cuánto tráfico es demasiado?
El impacto de redes viarias heterogéneas en áreas protegidas.
______________________________________________________________________
D'Amico, M. a, Périquet, S. b, Román, J.
a
and Revilla, E.
a
(Manuscript submitted).
How much traffic is too much? The impact of heterogeneous road-networks on
protected areas.
a
b
Université Claude-Bernard Lyon 1, Villeurbanne (France).
How much traffic is too much? The impact of heterogeneous road-networks on protected areas
Resumen: El efecto barrera es un impacto que afecta muchas poblaciones
animales y puede ser causado por una respuesta comportamental a carreteras (evitación
de superficies viarias o espacios vacíos), a las molestias asociadas (evitación de tráfico)
o a vehículos circulantes (evitación de vehículos). Muchos estudios han descrito efecto
barrera a lo largo de carreteras principales, pero pocos de ellos han investigado sus
causas, como por ejemplo el tráfico o la superficie viaria. Con este propósito medimos
el efecto barrera a lo largo de una red viaria heterogénea (i.e. con variación de tráfico y
superficie viaria) y además estimamos el consecuente deterioro de hábitat dentro de una
emblemática área protegida, la Reserva de la Biosfera de Doñana (España). Nuestras
especies de estudio fueron el ciervo Cervus elaphus y el jabalí Sus scrofa, y
muestreamos su presencia a lo largo de transectos de 200 m que empezaban en
carreteras de referencia (con más de 10 coches diarios) y que frecuentemente
intersecaban caminos sin pavimentar. La probabilidad de presencia de jabalíes fue
afectada por la mera distancia a la carretera más cercana (en la mayoría de casos
caminos sin pavimentación ni tráfico), y los ciervos sufrieron también la cercanía a
carreteras de referencias. El efecto barrera debido a la carretera más cercana no puede
estar relacionado ni con la superficie viaria ni con el tráfico asociado, y sugiere por
primera vez una evitación de espacios vacíos que podría estar basada en la asociación
entre estructuras lineares y la posibilidad de vehículos transitando. La consecuencia a
nivel de paisaje es una reducción de la probabilidad de presencia del 32% por ciervos y
del 45% por jabalíes. Este deterioro de hábitat reduce el valor de un área protegida para
nuestras especies de estudio. Por esta razón sugerimos el establecimiento de grandes
87
Capítulo 2: Efecto Barrera
áreas sin estructuras lineares, a través de la optimización de la red viaria, especialmente
dentro de áreas protegidas, con potenciales efectos positivos a nivel de ecosistema.
Abstract: The barrier effect is a road impact which affects several animal
populations and can be caused by a behavioral response to roads (surface or gap
avoidance), associated disturbances (traffic avoidance) or a circulating vehicle (vehicle
avoidance). Most past studies described barrier effect along major roads, but only few
of them investigated the causes, such as for example traffic volume or road surface.
With this purpose we measured the barrier effect along a heterogeneous road-network
(i.e. with variation in traffic volume or road surface), and we also estimated the
consequent loss of habitat quality within an emblematic protected area, Doñana
Biosphere Reserve (Spain). Our study species were the red deer Cervus elaphus and the
wild boar Sus scrofa, and we surveyed their presence of along 200-m transects starting
from reference roads (i.e. with more than 10 cars/day) and frequently intersecting
unpaved minor roads. The presence probability of wild boars was affected by the mere
distance to the nearest road (in most of cases unpaved road without traffic), and red deer
also suffered the nearness of reference roads. The barrier effect due to the nearest road
cannot be related to road surface or associated traffic, but suggests for the first time a
gap avoidance which may be based on the association between linear infrastructures and
the possibility of vehicles transiting along them. The consequence at landscape level is a
reduction in presence probability of 32% for deer and 45% for wild boar. This decrease
of habitat quality affects the conservation value of an emblematic protected area for the
selected model species. For this reason we suggest the establishment of large roadless
areas through the road-network optimization, especially within protected areas, with
potential positive effects for the whole ecosystems.
88
Introduction
Cross a road or avoid it. This individual choice can entail two kinds of impact
for animal populations: road-kill or barrier effect, respectively (Forman et al., 2003;
Grilo et al., 2012). The barrier effect has been defined as the consequence of a
behavioral response towards the road itself (road avoidance), the associated
disturbances (traffic avoidance) or a circulating vehicle (vehicle avoidance; Jaeger et al.,
2005; McGregor et al., 2008). Road avoidance may depend on the surface material
(surface avoidance; McGregor et al., 2008; Brehme et al., 2013) or the artificial
clearance in the vegetation (gap avoidance; Oxley et al., 1974; Ford and Fahrig, 2008),
while the traffic avoidance may be associated with road-related emissions and other
long-ranging disturbances (e.g.: noise, vibrations, lights, dust; Mader, 1984; Reijnen et
al., 1995; Forman and Alexander, 1998; Kaseloo, 2005). Traffic avoidance can also
results from changes in plant or animal communities due to clearance or treatment of
road verges (Oxley et al., 1974; Underhill and Angold, 1999). Finally, vehicle
avoidance is a short-range behavioral response of species having enough cognitive
ability and movement capacity to avoid being road-killed (Jaeger et al., 2005; Fahrig
and Rytwinski, 2009).
All these barrier effects typically arise within the road-effect zone, with an
extent determined by the features of both road and surrounding environment (Forman
and Alexander, 1998; Forman and Deblinger, 2000). Nevertheless, in the last century
89
road-networks have undergone a synergic development, which has caused the increase
and overlap of road-effect zones (Forman, 2000; Jaarsma and Willems, 2002). As a
result, the individual behavioral responses towards roads producing the barrier effect
can affect animal populations and even species, especially if their distribution is limited
to industrialized areas (i.e. with widespread and dense road-networks). The current
road-networks indeed fragment and deteriorate natural habitats (Andrews, 1990; Findlay
and Houlahan, 1997), imposing a filter on animal movement (Kramer-Schadt et al.,
2004; McDonald and St. Clair, 2004b). Large and connected populations become then
fragmented into small isolated ones (Vos and Chardon, 1998; Nellemann et al., 2001),
with a potential impact on their viability (Mumme et al., 2000; Lesbarrères et al., 2003).
Gene flow between isolated populations may also decrease (Keller and Largiadèr, 2003;
Riley et al., 2006), entailing a long-term threat to their genetic diversity (Lesbarrères et
al., 2006; Holderegger and Di Giulio, 2010). Finally, those isolated populations will
also have a lower colonization probability from other areas (Jaeger et al., 2005;
Eigenbrod et al., 2008).
Barrier effect is one of the best studied road-related impacts (Trombulak and
Frissell, 2000; Forman et al., 2003). Nevertheless, there are few studies determining the
factors behind the negative response of a given animal species towards certain road
features, such as road type or traffic volume. Typical barrier effect studies are usually
conducted along a heavily transited major road or highway, making difficult to separate
road avoidance from traffic avoidance (McGregor et al., 2008). This distinction could
be highlighted by measuring the barrier effect along a heterogeneous road-network, with
enough variation in road types (e.g. paved and unpaved roads) and traffic volumes.
90
Unfortunately, unpaved roads are comparatively less studied than paved ones, although
the former are most widespread within natural and protected areas. This entails a
potential threat for wildlife, because most species are more abundant in natural areas,
and they may be then mostly affected by unpaved roads.
The main objective of the present study is to determine for the first time the
barrier effect impact of a whole road-network, estimating the related loss of potential
habitat quality for a given species. We selected a road-network within a natural
protected area, Doñana National and Natural Parks (Spain). This emblematic Biosphere
Reserve has a heterogeneous road-network which might facilitate the distinction
between road avoidance and traffic avoidances, a distinction of paramount importance
to plan mitigation of road impacts (Jaeger et al., 2005).
With the aim to optimize the possibilities to detect barrier effects, we selected
two ungulates as model species, because ungulates are susceptible to human
disturbances (Frid and Dill, 2002; Stankowich, 2008), including road-related ones
(Alexander et al., 2005; Ito et al., 2013). These species were the red deer Cervus
elaphus Linnaeus, 1758 (which is frequently described as negatively affected by major
roads and associated traffic; Gagnon et al., 2007b; St. Clair and Forrest, 2009) and the
wild boar Sus scrofa Linnaeus, 1758 (usually less affected by human disturbance;
Frantz et al., 2012).
The consideration for the first time of a heterogeneous road-network should
provide a more complete overview of road-related barrier effects after controlling for
91
the environmental factors potentially affecting the probability of presence (such as
pasture or water availability). The main benefit will be a realistic approach to estimate
the loss in habitat quality imposed by the presence of roads, with the consequent
reduction in the conservation value of the selected protected area for the two model
species.
Study area
The fieldwork was carried out within Doñana National and Natural Parks (southwestern Spain: 36º59’ N, 6º26’ W), having the National Park a higher protection degree
(Figure 1). Doñana is an assemblage of protected areas characterized by Mediterranean
climate, with a mosaic of urban, rural and natural environments, including scrubland,
dunes and wetlands. Several regulated human activities are allowed within these
protected areas (e.g. cattle breeding, beekeeping, pine-cone harvesting, tourism,
research, etc.), aided by the widespread road-network present within the protected area
(Figure 1). This is characterized by different road types and traffic volumes, ranging
from highly travelled highways to isolated unpaved roads (Román et al., 2010a).
92
Figure 1: Study area. Into the left section the National Park is in dark grey and the Natural Park is in
light grey. In the main section the solid thin lines represent the road-network, and the dashed lines are the
reference roads. Black circles indicate the locations of transects. On the right we represent a transect; the
arrows A and B represent two examples of the explanatory variables denominated distance to the nearest
road and distance to the nearest reference road, respectively.
Data collection
Red deer and wild boar are present over the whole study area. In the autumn of
2009 (September and October), we randomly established 40 transects 200-m-long and
perpendicular to and starting from a reference road (Figure 1), which we defined as
93
roads with a traffic volume above 10 vehicles per day (Román et al., 2010b). Each
transect was subdivided into 20 sections, each 10-m-long (Figure 1). By definition, and
given the dense road-network present, transects could intersect with minor roads (i.e.
roads with low traffic intensity; Figure 1).
We determined the presence of both study species within each transect section
by walking along the transect line and georeferring the pellets found within a 1-m-wide
buffer. In the case of deer, we only considered those aggregations with more than 10
pellets. We characterized the sections with the distance to the nearest road and distance
to the nearest reference road measured from the center of every section; and the traffic
volume of the nearest reference road (Table 1; Figure 1). The available road-map was
updated using 2007 aerial orthophotos and ground validated in the spring of 2009.
Table 1: Factors potentially affecting deer and wild boar presence probabilities. Averages and range
values of the potential explanatory variables considered in the analyses. Traffic volume is expressed as
average daily cars (ADC). The level of protection of Natural Park is lower than National Park.
Variable
Environmental
Road-related
Average (+ SE)
Range
9438 (+ 7745)
289-23844
distance to nearest water-body (m)
379 (+ 553)
0-2954
distance to nearest shrub-patch (m)
protection level
556 (+ 770)
0-2509
National Park / Natural Park
distance to wetland (m)
distance to the nearest road (m)
63 (+ 51)
0-197
distance to the nearest reference road (m)
99 (+ 58)
5-197
1018 (+ 2091)
11-6800
traffic volume of the nearest reference road (ADC)
94
The traffic volume of the nearest reference road was calculated using
underground magnetometers (TRAFx Vehicle Counter Generation III; www.trafx.net)
deployed during the survey (autumn 2009). Each reference road was sampled during
two complete non-consecutive weeks. Traffic in minor roads was always below one car
per day.
The presence of both species is mediated by habitat quality. We controlled for
environmental variables potentially affecting this quality by characterizing transect
sections with variables describing pasture availability, which in Doñana is mediated by
the ground water level (distance to wetland and distance to nearest water-body; Table
1) and by the presence and proximity to the protection offered by thick vegetation
(distance to the nearest shrub-patch; Table 1). Within every section we determined the
corresponding values for these environmental variables using 2007 aerial orthophotos.
The protection level potentially affects presence and abundance because hunting is
allowed in the Natural Park (Table 1).
Data analysis
We used generalized linear mixed models (GLMM) and evaluated their
performance using Akaike information criterion (AIC). For each model the response
variable was the presence/absence of the target species within each transect section,
95
considering the identification code of each transect as a random factor. We used a
binomial error distribution and a logit link function.
We considered different hypotheses, depending on the assumed impact of roads
(Table 2). First, we considered the model with only environmental variables as the null
model explaining the presence of both species. The rest of the models contained always
the environmental variables. We considered the nearest road hypothesis (model adding
the distance to the nearest road), the nearest road reference hypothesis (two models,
one including distance to the nearest reference road and another including distance to
the nearest reference road plus traffic volume in that reference road), and a combination
of previously explained hypothesis (two models, one including both the distance to the
nearest road and the distance to the nearest reference road, and the full model; Table
2). We selected the most supported model using AIC, and calculated Akaike weights
(wAIC) to estimate the relative support for every given model (ranging from 0 to 1,
with larger numbers indicating greater support; Burnham and Anderson 2002). We
selected as plausible those models with ΔAIC < 2 (Burnham et al. 2011).
Finally, we used the best supported model to calculate the road-network impact
on the study area conservation value for each species. We previously measured within
every pixel (with a 10 m resolution) of the study area map the values corresponding to
each explanatory variable. For each pixel we then calculated deer and wild boar
presence probabilities considering both environmental and traffic variables, applying the
equation of the best supported model for every study species, respectively. As a result,
we obtained two presence probability maps (both of them considering traffic effects):
the first for deer and the second for wild boar. We then repeated the presence
96
probability calculation but replacing within every pixel the values corresponding to each
traffic variable with the value assuming there was not road/traffic effect (i.e. a pixel
with a distance from the nearest road or reference road corresponding to 200 m). Also
in this case we obtained two presence probability maps (both of them without traffic
effects) one species. For every map we calculated the overall presence probabilities as a
proxy of habitat quality for deer and wild boar, respectively over the entire study area
(sum of the presence probability for all pixels over the study area). Finally, for each
species we calculated the reduction in habitat quality imposed by the road-network on
both species, by obtaining the percentage reduction in habitat quality.
Results
During the whole survey we detected deer and wild boar excrements in 194 and
30 transect sections, respectively (i.e. 24.25% and 3.75% of the 800 surveyed transect
sections). On a wider scale, 32 and 12 transects had at least one transect section with
evidences of deer or wild boar presence, respectively (i.e. 80% and 30% of the 40
surveyed transects).
Regarding to model selections, for each species there were two supported
models according to the ΔAIC < 2 provision (Table 2). In the case of the deer, both
supported models belonged to the hypothesis combining the distance to the nearest
road and the distance to the nearest reference road (88% of wAIC), highlighting the
relevance of these explanatory variables when considered together. The distance to the
97
nearest road appears to be more relevant than the distance to the nearest reference
road, because the only model belonging to the nearest road hypothesis was at least ten
times more supported than any model of the nearest reference road hypothesis. The
traffic volume of the nearest reference road should be a less relevant explanatory
variable, since the addition of this factor to a model decreased the ΔAIC in both of
cases.
Concerning the wild boar, the first ranked model belonged to the nearest road
hypothesis and it was almost three times more supported than the second ranked model,
which included both distance to the nearest road and distance to the nearest reference
road (Table 2). On the other hand, the distance to the nearest reference road appears to
be a marginally relevant explanatory variable, since the models of the nearest reference
road hypothesis had not support.
98
Table 2: Factors affecting red deer and wild boar presence probabilities within each transect
section, model ranks by AIC weights (wAIC). In the Model column, the abbreviation env refers to the
four environmental variables designated as distance to wetland, distance to nearest water-body, distance
to the nearest shrub-patch and protection level; while the abbreviations road, reference and traffic
concern the three road-related variables designated as distance to the nearest road, distance to the nearest
reference road and traffic volume of the nearest reference road, respectively. See Table 1 for a
description of explanatory variables. ΔAIC is the relative difference of a given AIC value compared to the
smallest AIC value. The supported models (ΔAIC < 2) are highlighted in grey. AIC weights indicate the
relative support for every model (the weights of all the models in the candidate set have the sum of 1).
Deer
Evidence ratio (ER) is the ratio of wAIC, comparing the best supported model with every competing one.
ΔAIC
wAIC
Rank
env
764.8
40.0
0.00
6
env + road
env + reference road
env + reference road + traffic
env + road + reference road
env + road + reference road + traffic
env
env + road
env + reference road
env + reference road + traffic
env + road + reference road
env + road + reference road + traffic
727.1
731.9
733.0
723.6
724.8
251.2
248.8
252.7
252.7
250.7
252.0
3.5
8.3
9.4
0.0
1.2
2.4
0.0
3.9
3.9
1.9
3.2
0.09
0.02
0.01
0.57
0.31
0.14
0.46
0.07
0.07
0.18
0.08
3
4
5
1
2
3
1
4
4
2
6
Model
Null
Nearest road
Reference road
Combination
Wild boar
AIC
Hypothesis
Null
Nearest road
Reference road
Combination
ER
888459835
.8
5.8
62.2
110.0
1.0
1.8
3.4
1.0
7.1
7.1
2.6
5.0
Both deer and wild boar presence probabilities increased according to the
distance to the nearest road and the distance to the nearest reference road, considering
all the best supported models (Table 3; Figure 2). In the case of deer, the presence
probability decreased according to the traffic volume of the nearest reference road
(Table 3).
99
Figure 2: Distance to the nearest road. The relation between presence probability and distance to the
nearest road, for both deer (on the left) and wild boar (on the right). Dotted lines represent upper and
lower standard errors.
Table 3: Factors affecting red deer and wild boar presence probabilities within each transect
section, parameter estimates for supported models. In the Parameter estimates columns, values in
parentheses are standard errors (+ SE).
Wild boar
Deer
Explanatory variables
100
Intercept
distance to wetland
distance to nearest water-body
distance to nearest shrub-patch
protection level
distance to the nearest road
distance to the nearest reference road
traffic volume of the nearest reference road
Intercept
distance to wetland
distance to nearest water-body
distance to nearest shrub-patch
protection level
distance to the nearest road
distance to the nearest reference road
traffic volume of the nearest reference road
Variable
categories
National Park
Natural Park
National Park
Natural Park
First ranked
model
Second ranked
model
Parameter estimates (+ SE)
- 1.511 (+ 0.394)
- 0.009 (+ 0.003)
- 0.064 (+ 0.032)
- 0.043 (+ 0.018)
0.619 (+ 0.324)
0
0.771 (+ 0.248)
0.516 (+ 0.217)
- 3.151 (+ 0.736)
- 0.001 (+ 0.004)
- 0.156 (+ 0.037)
- 0.078 (+ 0.045)
0.155 (+ 0.673)
0
0.739 (+ 0.319)
-
- 1.462 (+ 0.397)
- 0.009 (+ 0.003)
- 0.045 (+ 0.037)
- 0.044 (+ 0.018)
0.723 (+ 0.344)
0
0.768 (+ 0.242)
0.521 (+ 0.215)
- 0.048 (+ 0.054)
- 3.065 (+ 0.781)
- 0.001 (+ 0.005)
- 0.152 (+ 0.082)
- 0.075 (+ 0.045)
0.128 (+ 0.673)
0
0.854 (+ 0.502)
0.145 (+ 0.483)
-
According to model selection, the presence probability maps were obtained
through the best supported models (in the case of the deer including both to the distance
to the nearest road and the distance to the nearest reference road, while for the wild
boar only the first variable). For each species, we used the environmental presence
probability map as a reference of habitat quality, in comparison to the map including
road-related variables (Figures 3 and 4). The percentage reduction in habitat quality was
32% for deer and 45% for wild boar, respectively.
101
Figure 3: Deer presence probability maps. On the left the deer presence probability map including only
environmental variables, on the right the map regarding the best supported model including road-related
variables (see Tables 2 and 3).
102
Figure 4: Wild boar presence probability maps. On the left the wild boar presence probability map
including only environmental variables, on the right the map regarding the best supported model
including road-related variables (see Tables 2 and 3).
Discussion
In the present study we showed the effects of a heterogeneous road-network on
the presence probabilities of two model species within a protected area. We observed
that, even when considering the environmental factors potentially influencing the spatial
behavior of both red deer and wild boar, the presence of a heterogeneous road-network
can determine a relevant loss of habitat quality at landscape level, with the consequent
reduction in the conservation value of the selected area.
Our results showed that deer and wild boar presence probabilities increased
according to the mere distance from the nearest road, which in the most of cases refers
to unpaved road virtually without traffic. Concerning this variable we could therefore
exclude that the study species were suffering traffic avoidance or road avoidance
depending on road surface, focusing then on road avoidance due to the artificial
clearance in matrix habitat (i.e. gap avoidance). This kind of barrier effect has been
previously observed for the red deer (Rost and Bailey, 1979; Meisingset et al., 2013;
also in our study area: Suárez-Esteban et al., 2013a), although the authors did not
discuss the subjacent mechanisms potentially responsible for it; while for wild boar this
is the first reference. The most well-known cases of gap avoidance are in birds and
gliding mammals eluding artificial clearances in the forest canopy (Laurance, 2004; van
der Ree et al., 2010). Artificial gap avoidance is usually interpreted as an anti-predator
103
behavior (Shepard et al., 2008; Tremblay and St. Clair, 2009), similarly to the barrier
effect produced by a natural clearance in the matrix habitat (Bélisle et al., 2001; Lees
and Peres, 2009). In densely vegetated habitats, this anti-predator behavior should be
limited to narrow road-effect zones virtually coincident with the linear infra-structures,
contrarily to gap avoidance we detected in red deer and wild boar which affected wide
areas. This difference probably occurs because Doñana road-network mainly spreads
across open habitats (Román et al., 2010c), which vegetation cannot provide the suitable
anti-predator cover for large ungulates (Stankowich, 2008). As a consequence, our study
species preferentially used the core areas of the patches, as far as possible from the
artificial gaps in the matrix habitat. Both red deer and wild boar instinctively avoids
humans and vehicles (Gagnon et al., 2007b; Marini et al., 2009; Rogala et al., 2011),
and predation attempts consistently amplify the perception of risk in prey species
(Stankowich and Blumstein, 2005). These ungulates are a long-lived species with
relatively high cognitive capacities (Kloppers et al., 2005; Morelle et al., 2014), and for
this reason they could be able to associate humans and vehicles with linear infrastructures. This mechanism might have been also promoted by hunters in the past and
poachers in the present, considering that hunting and poaching pressure in natural areas
usually spread through and from the road-network (Robinson et al., 1999; Papaioannou
and Kati, 2007). Therefore, the gap avoidance we observed in red deer and wild boars
might be for the first time classified as prolonged traffic avoidance, based on the
association between linear infra-structures and the memory (and related possibility) of
vehicle transit along them.
The distance from the nearest reference road also affected presence probabilities
just as expected, but mainly for the red deer, which has been widely described avoiding
104
vehicular traffic (Gagnon et al., 2007a, 2007b; St. Clair and Forrest, 2009). Both the
model species were then simultaneously affected by the usual traffic avoidance and the
prolonged one we suggested above.
Both kinds of barrier effect detected in the present study were synergically
determining the habitat fragmentation and loss we depicted in the presence probability
maps. The environmental maps showed that almost the whole considered protected area
could provide good presence probabilities for both deer and wild boars, with a relative
importance of the area protection level for deer. Nevertheless since we included road
effects the best areas suffered extreme fragmentation, while the previous low quality
areas became essentially unavailable. As a consequence, the habitat quality at landscape
level decreased almost one-third and one-half for deer and wild boar, respectively.
Similarly to main part of large ungulates, red deer and wild boars require sizeable areas
to establish viable home ranges (Carranza et al., 1991; Massei et al., 1997). In some
cases these areas can include linear infra-structures (Frantz et al., 2012; Meisingset et
al., 2013), but it has been widely shown that road fragmentation involves a decrease of
habitat quality, with negative consequences on behavior (Guangshun et al., 2006; Ciuti
et al., 2012), physiology (Millspaugh et al., 2001; Gasparik et al., 2012), reproduction
(Phillips and Alldredge, 2000) and survival (Malo et al., 2004; McCorquodale et al.,
2011). Although both red deer and wild boar are globally considered as Least Concern
species, the conservation of several of their subspecies and populations are explicitly
proclaimed by international legislations and also recommended by scientific literature
(Lovari et al., 2008; Oliver and Leus, 2008). On the other hand, the prolonged traffic
105
avoidance suggested by the present study can probably affect most of ungulate species,
which typically show relevant sensitivity to road-related disturbances (Stankowich,
2008; Ito et al., 2013). Mitigation measures should be implemented not only along
highways, but also along unpaved road-networks, especially within national and natural
parks which ideally should be the source areas for these species. The most suitable
mitigation measure for heterogeneous road-networks is their optimization through the
decommissioning of high-risk and redundant linear infra-structures (see for example:
Switalski et al., 2004; Forman, 2005; Dolan and Whelan, 2007), just as previously
suggested for our study area (Román et al., 2010d). This kind of actions or even the
establishment of large roadless areas have been already shown to benefit a wide range
of species from mammals to plants (Chen and Roberts, 2008; Grant et al., 2011;
Switalski and Nelson, 2011).
Unfortunately, in most of cases a road-network optimization cannot
decommission paved roads, although their negative impacts on ungulates are proven
(Stankowich, 2008; Huijser et al., 2009). Despite this, past studies showed that some
ungulate species can be able to cross major roads, for example by avoiding traffic rush
hours or in presence of extreme foraging or mating motivation (Baofa et al., 2006;
Meisingset et al., 2013). Several of these species may also present individual variation
in traffic tolerance (Clevenger and Waltho, 2000; Papouchis et al., 2001) and linear
infra-structures would represent more a filter than an impenetrable barrier for them, just
as previously suggested for other mammals and also birds (Alexander et al., 2005; Jones
and Pickvance, 2013). The purpose of possible mitigation measures should be to
enhance the permeability of this filter, attempting to not increase at the same time the
road-kill probability, which can threaten both ungulate populations (Groot Bruinderink
106
and Hazebroek, 1996) and human safety, and also imply high economic costs (Conover
et al., 1995; Bissonette et al., 2008). The most suitable mitigation measure in this case is
the simultaneous provision of a network of wildlife crossing structures supported by
road-exclusion fences (Clevenger et al., 2001; Olsson and Widen 2008). Concerning
ungulates, crossing structures showed in past studies a mutable effectiveness, from high
(Clevenger and Waltho, 2000; 2005; Mata et al., 2008) to relatively low (Rodríguez et
al., 1996; Mata et al., 2005). Most of efficient crossing structures were wildlife
overpasses (also called eco-ducts or land bridges: Van Wieren and Worm, 2001;
Clevenger and Waltho, 2005), which also represent the best solution to improve the
landscape connectivity for a large range of other species (McDonald and St. Clair,
2004a; Bond and Jones, 2008). On the other hand, even these efficient crossing
structures should be appropriately designed and managed with the purpose to assure
their effectiveness. For example, traffic noise and other emissions in the overpass
proximities should be properly eliminated or at least reduced (Reijnen et al., 1997;
Glista et al., 2009), and the presence of natural vegetation connecting the crossing
structure and the surrounding natural habitats should be favored depending on the target
species (Clevenger and Waltho, 2005; Jones et al., 2011).
Several ungulate species have been defined as keystone species and ecosystem
engineer in most part of their distribution range, because of their potential impact on
plant species assemblage (Côté et al., 2004; Sandom et al., 2013) and consequences on
ecological communities (McShea and Rappole, 2000; Ripple and Larsen, 2000).
Therefore, we should consider that the road-network is not only limiting ungulate
107
presence probabilities and related conservation threats, but also has consequences on the
whole ecosystem. For example, in our case study the extreme road-network
development through the scrubland areas can probably exacerbates the ungulate habitat
selection towards the highly productive ecotone with the wetlands, favoring the
overgrazing of these areas (Rogers and Myers, 1980; Soriguer, 1983) and the possible
consequences on other species. Another example recently described in Doñana showed
that the lack of ungulate herbivory along more travelled unpaved roads is one of the
factors facilitating the establishment of fleshy-fruit shrubs (Suárez-Esteban et al.,
2013b).
This is the first study in which the barrier effect was analyzed along a
heterogeneous road-network, and this innovative approach revealed an undescribed
category of road avoidance: the prolonged traffic avoidance due to the recognition of
linear infra-structures as possibility of motorized transit. We suggest that this type of
road avoidance is occurring for red deer and wild boar, but it could affect a large
number of species, at least the long-lived ones having high cognitive capacities. For this
reason we recommended to mitigate the impact of unpaved road-networks through their
optimization, especially within protected areas, with potential positive effects for the
whole ecosystem.
Acknowledgements
This study was funded by the project 2008X0963 (Andalusian Ministry of
Environment; Regional Government of Andalusia) and Land-Rover España S.A. kindly
108
lent two vehicles. MD was supported by JAE-PRE fellowship from CSIC, SP by ICTSRBD project (ICTS-2009-39). Andrea Barón helped with fieldwork and traffic model
development; Sofía Conradi-Fernández provided logistic support. Nestor Fernández,
Simone Santoro, Giulia Bastianelli, Zulima Tablado and Carlos Rodríguez provided
helpful comments; Manuela González-Suárez and Margarita Mulero-Pázmány reviewed
a draft version of the manuscript.
References
Alexander, S.M., Waters, N.M. and Paquet, P.C. (2005). Traffic volume and highway permeability for a
mammalian community in the Canadian Rocky Mountains. Canadian Geographer / Le Géographe
canadien. 49: 321-331.
Andrews, A. (1990). Fragmentation of habitat by roads and utility corridors: a review. Australian
Zoologist. 26: 130-141.
Baofa, Y., Huyin, H., Yili, Z., Le, Z. and Wanhong, W. (2006). Influence of the Qinghai-Tibetan railway
and highway on the activities of wild animals. Acta Ecologica Sinica. 26: 3917-3923.
Bélisle, M., Desrochers, A. and Fortin, M.-J. (2001). Influence of forest cover on the movements of forest
birds: a homing experiment. Ecology. 82: 1893-1904.
Bissonette, J., Kassar, C.A. and Cook, L.J. (2008). Assessment of costs associated with deer-vehicle
collisions: human death and injury, vehicle damage, and deer loss. Human-Wildlife Conflicts. 2: 1727.
Bond, A.R. and Jones, D.N. (2008). Temporal trends in use of fauna-friendly underpasses and overpasses.
Wildlife Research. 35: 103-112.
Burnham, K.P. and Anderson, D.R. (2002). Model selection and multimodel inference. A practical
information-theoretic approach. Springer-Verlag, 488 pp., New York (USA).
Burnham, K., Anderson, D. and Huyvaert, K. (2011). AIC model selection and multimodel inference in
behavioral ecology: some background, observations, and comparisons. Behavioral Ecology and
Sociobiology. 65: 23-35.
Carranza, J., Hidalgo de Trucios, S.J., Medina, R., Valencia, J. and Delgado, J. (1991). Space use by red
deer in a Mediterranean ecosystem as determined by radio-tracking. Applied Animal Behaviour
Science. 30: 363-371.
Chen, X. and Roberts, K.A. (2008). Roadless areas and biodiversity: a case study in Alabama, USA.
Biodiversity and Conservation. 17: 2013-2022.
Ciuti, S., Northrup, J.M., Muhly, T.B., Simi, S., Musiani, M., Pitt, J.A. and Boyce, M.S. (2012). Effects
of humans on behaviour of wildlife exceed those of natural predators in a landscape of fear. PLoS
ONE. 7: e50611 [online].
Clevenger, A.P., Chruszcz, B. and Gunson, K. (2001). Highway mitigation fencing reduces wildlifevehicle collisions. Wildlife Society Bulletin. 29. 646-653.
Clevenger, A.P. and Waltho, N. (2000). Factors influencing the effectiveness of wildlife underpasses in
Banff National Park, Alberta, Canada. Conservation Biology. 14: 47-56.
Clevenger, A.P. and Waltho, N. (2005). Performance indices to identify attributes of highway crossing
structures facilitating movement of large mammals. Biological Conservation. 121: 453-464.
109
Conover, M.R., Pitt, W.C., Kessler, K.K., DuBow, T.J. and Sanborn, W.A. (1995). Review of human
injuries, illnesses, and economic losses caused by wildlife in the United States. Wildlife Society
Bulletin. 23: 407-414.
Côté, S.D., Rooney, T.P., Tremblay, J.-P., Dussault, C. and Waller, D.M. (2004). Ecological impacts of
deer overabundance. Annual Review of Ecology, Evolution, and Systematics. 35: 113-147.
Dolan, L. and Whelan, P.M. (2007). Road decommissioning: minimising the adverse ecological effects of
roads in European agriculture landscapes. In: Proceedings of the 2007 International Conference on
Ecology and Transportation ICOET (Irwin, C.L., Nelson, D. and McDermott, K.P., Eds.). Center for
Transportation and the Environment and North Carolina State University, Raleigh (USA), p. 592.
Eigenbrod, F., Hecnar, S.J. and Fahrig, L. (2008). Accessible habitat: an improved measure of the effects
of habitat loss and roads on wildlife populations. Landscape Ecology. 23: 159-168.
Fahrig, L. (2007). Non-optimal animal movement in human-altered landscapes. Functional Ecology. 21:
1003-1015.
Fahrig, L. and Rytwinski, T. (2009). Effects of roads on animal abundance: an empirical review and
synthesis. Ecology and Society. 14: 21 [online].
Findlay, C.S. and Houlahan, J. (1997). Anthropogenic correlates of species richness in southeastern
Ontario wetlands. Conservation Biology. 11: 1000-1009.
Ford, A.T. and Fahrig, L. (2008). Movement patterns of eastern chipmunks (Tamias striatus) near roads.
Journal of Mammalogy. 89: 895-903.
Forman, R.T.T. (2000). Estimate of the area affected ecologically by the road system in the United States.
Forman, R.T.T. (2005). Good and bad places for roads: effects of varying road and natural pattern on
habitat loss, degradation, and fragmentation. In: Proceedings of the 2005 International Conference on
Ecology and Transportation ICOET (Irwin, C.L., Nelson, D. and McDermott, K.P., Eds.). Center for
Transportation and the Environment and North Carolina State University, Raleigh (USA), pp. 164174.
Forman, R.T.T. and Deblinger, R.D. (2000). The ecological road-effect zone of a Massachusetts (U.S.A.)
suburban highway. Conservation Biology. 14: 36-46.
(2003). Road ecology: science and solutions. Island Press, 481 pp., Washington (USA).
Frantz, A.C., Bertouille, S., Eloy, M.C., Licoppe, A., Chaumont, F. and Flamand, M.C. (2012).
Comparative landscape genetic analyses show a Belgian motorway to be a gene flow barrier for red
deer (Cervus elaphus), but not wild boars (Sus scrofa). Molecular Ecology. 21: 3445-3457.
Frid, A. and Dill, L. (2002). Human-caused disturbance stimuli as a form of predation risk. Conservation
Ecology. 6: 11 [online].
Gagnon, J.W., Schweinsburg, R.E. and Dodd, N.L. (2007a). Effects of roadway traffic on wild ungulates:
a review of the literature and case study of elk in Arizona. In: Proceedings of the 2007 International
Conference on Ecology and Transportation ICOET (Irwin, C.L., Nelson, D. and McDermott, K.P.,
Eds.). Center for Transportation and the Environment and North Carolina State University, Raleigh
(USA), pp. 449-458.
Gagnon, J.W., Theimer, T.C., Boe, S., Dodd, N.L. and Schweinsburg, R.E. (2007b). Traffic volume alters
elk distribution and highway crossings in Arizona. Journal of Wildlife Management. 71: 2318-2323.
Gasparik, J., Dobias, M., Capcarova, M., Smehyl, P., Slamecka, J., Bujko, J. and Gasparik, J.Jr. (2012).
Concentration of cadmium, mercury, zinc, copper and cobalt in the tissues of wild boar (Sus scrofa)
hunted in the western Slovakia. Journal of Environmental Science and Health, Part A. 47: 1212-1216.
Grant, A.S., Nelson, C.R., Switalski, T.A. and Rinehart, S.M. (2011). Restoration of native plant
communities after road decommissioning in the Rocky Mountains: effect of seed-mix composition on
vegetative establishment. Restoration Ecology. 19: 160-169.
Grilo, C., Sousa, J., Ascensão, F., Matos, H., Leitão, I., Pinheiro, P., Costa, M., Bernardo, J., Reto, D.,
Lourenço, R., Santos-Reis, M. and Revilla, E. (2012). Individual spatial responses towards roads:
implications for mortality risk. PLoS ONE. 7: e43811 [online].
110
Groot Bruinderink, G.W.T.A. and Hazebroek, E. (1996). Ungulate traffic collisions in Europe.
Guangshun, J., Jianzhang, M.A. and Minghai, Z. (2006). Spatial distribution of ungulate responses to
habitat factors in Wandashan Forest Region, northeastern China. Journal of Wildlife Management. 70:
1470-1476.
Holderegger, R. and Di Giulio, M. (2010). The genetic effects of roads: a review of empirical evidence.
Basic and Applied Ecology. 11: 522-531.
Huijser, M.P., Duffield, J.W., Clevenger, A.P., Ament, R.J. and McGowen, P. (2009). Cost-benefit
analyses of mitigation measures aimed at reducing collisions with large ungulates in the United States
and Canada: a decision support tool. Ecology and Society. 14: 15 [online].
Ito, T.Y., Lhagvasuren, B., Tsunekawa, A., Shinoda, M., Takatsuki, S., Buuveibaatar, B. and Chimeddorj,
B. (2013). Fragmentation of the habitat of wild ungulates by anthropogenic barriers in Mongolia.
PLoS ONE. 8: e56995 [online].
Jaarsma, C.F. and Willems, G.P.A. (2002). Reducing habitat fragmentation by minor rural roads through
traffic calming. Landscape and Urban Planning. 58: 125-135.
Jones, D.N., Bakker, M., Bichet, O., Coutts, R. and Wearing, T. (2011). Restoring habitat connectivity
over the road: vegetation on a fauna land-bridge in south-east Queensland. Ecological Management
and Restoration. 12: 76-79.
Jones, D.N. and Pickvance, J. (2013). Forest birds use vegetated fauna overpass to cross multi-lane road.
Oecologia Australis. 17: 147-156.
Kaseloo, P.A. (2005). Synthesis of noise effects on wildlife populations. In: Proceedings of the 2005
International Conference on Ecology and Transportation ICOET (Irwin, C.L., Nelson, D. and
McDermott, K.P., Eds.). Center for Transportation and the Environment and North Carolina State
University, Raleigh (USA), pp. 33-35.
Keller, I. and Largiadèr, C.R. (2003). Recent habitat fragmentation caused by major roads leads to
reduction of gene flow and loss of genetic variability in ground beetles. Proceedings of the Royal
Society B: Biological Sciences. 270: 417-423.
Kloppers, E.L., St. Clair, C.C. and Hurd, T.E. (2005). Predator-resembling aversive conditioning for
managing habituated wildlife. Ecology and Society. 10: 31 [online].
Kramer-Schadt, S., Revilla, E., Wiegand, T. and Breitenmoser, U.R.S. (2004). Fragmented landscapes,
road mortality and patch connectivity: modelling influences on the dispersal of Eurasian lynx. Journal
of Applied Ecology. 41: 711-723.
Laurance, S.G.W. (2004). Responses of understory rain forest birds to road edges in central Amazonia.
Ecological Applications. 14: 1344-1357.
Lees, A.C. and Peres, C.A. (2009). Gap-crossing movements predict species occupancy in Amazonian
forest fragments. Oikos. 118: 280-290.
Lesbarrères, D., Pagano, A. and Lodé, T. (2003). Inbreeding and road effect zone in a Ranidae: the case
of agile frog, Rana dalmatina Bonaparte, 1840. Comptes Rendus Biologies. 326: 68-72.
Lesbarrères, D., Primmer, C.R., Lodé, T. and Merilä, J. (2006). The effects of 20 years of highway
presence on the genetic structure of Rana dalmatina populations. EcoScience. 13: 531-538.
Lovari, S., Herrero, J., Conroy, J., Maran, T., Giannatos, G., Stubbe, M., Aulagnier, S., Jdeidi, T.,
Masseti, M., Nader, I., de Smet, K. and Cuzin, F. (2008). Cervus elaphus. IUCN Red List of
Threatened Species, www.iucnredlist.org, Version 2014 [online].
Mader, H.J. (1984). Animal habitat isolation by roads and agricultural fields. Biological Conservation. 29:
81-96.
Malo, J.E., Suárez, F. and Díez, A. (2004). Can we mitigate animal-vehicle accidents using predictive
models? Journal of Applied Ecology. 41: 701-710.
Marini, F., Franzetti, B., Calabrese, A., Cappellini, S. and Focardi, S. (2009). Response to human
presence during nocturnal line transect surveys in fallow deer (Dama dama) and wild boar (Sus
scrofa). European Journal of Wildlife Research. 55: 107-115.
Massei, G., Genov, P.V., Staines, B.W. and Gorman, M.L. (1997). Factors influencing home range and
activity of wild boar (Sus scrofa) in a Mediterranean coastal area. Journal of Zoology. 242: 411-423.
111
Mata, C., Hervás, I., Herranz, J., Suárez, F. and Malo, J.E. (2005). Complementary use by vertebrates of
crossing structures along a fenced Spanish motorway. Biological Conservation. 124: 397-405.
McCorquodale, S.M., Wik, P.A. and Fowler, P.E. (2011). Elk survival and mortality causes in the Blue
Mountains of Washington. Journal of Wildlife Management. 75: 897-904.
McDonald, W. and St. Clair, C.C. (2004a). Elements that promote highway crossing structure use by
small mammals in Banff National Park. Journal of Applied Ecology. 41: 82-93.
McDonald, W. and St. Clair, C.C. (2004b). The effects of artificial and natural barriers on the movement
of Small mammals in Banff National Park, Canada. Oikos. 105: 397-407.
McShea, W.J. and Rappole, J.H. (2000). Managing the abundance and diversity of breeding bird
populations through manipulation of deer populations. Conservation Biology. 14: 1161-1170.
Meisingset, E.L., Loe, L.E., Brekkum, Ø., Van Moorter, B. and Mysterud, A. (2013). Red deer habitat
selection and movements in relation to roads. Journal of Wildlife Management. 77: 181-191.
Millspaugh, J.J., Woods, R.J., Hunt, K.E., Raedeke, K.J., Brundige, G.C., Washburn, B.E. and Wasser,
S.K. (2001). Fecal glucocorticoid assays and the physiological stress response in elk. Wildlife Society
Bulletin. 29: 899-907.
Morelle, K., Lehaire, F. and Lejeune, P. (2014). Is wild boar heading towards movement ecology? A
review of trends and gaps. Wildlife Biology. 20: 196-205.
501-512.
Nellemann, C., Vistnes, I., Jordhøy, P. and Strand, O. (2001). Winter distribution of wild reindeer in
relation to power lines, roads and resorts. Biological Conservation. 101: 351-360.
Oliver, W. and Leus, K. (2008). Sus scrofa. IUCN Red List of Threatened Species, www.iucnredlist.org,
Version 2014 [online].
Olsson, M.P.O. and Widen, P. (2008). Effects of highway fencing and wildlife crossings on moose Alces
alces movements and space use in southwestern Sweden. Wildlife Biology. 14: 111-117.
Oxley, D.J., Fenton, M.B. and Carmody, G.R. (1974). The effects of roads on small mammals. Journal of
Applied Ecology. 11: 51-59.
Papaioannou, H.I. and Kati, V.I. (2007). Current status of the Balkan chamois (Rupicapra rupicapra
balcanica) in Greece: implications for conservation. Belgian Journal of Zoology. 137: 33-39.
Papouchis, C.M., Singer, F.J. and Sloan, W.B. (2001). Responses of desert bighorn sheep to increased
human recreation. Journal of Wildlife Management. 65: 573-582.
Phillips, G.E. and Alldredge, A.W. (2000). Reproductive success of elk following disturbance by humans
during calving season. Journal of Wildlife Management. 64: 521-530.
Reijnen, R., Foppen, R., Braak, C.T. and Thissen, J. (1995). The effects of car traffic on breeding bird
populations in woodland. III. Reduction of density in relation to the proximity of main roads. Journal
of Applied Ecology. 32: 187-202.
Reijnen, R., Foppen, R. and Veenbaas, G. (1997). Disturbance by traffic of breeding birds: evaluation of
the effect and considerations in planning and managing road corridors. Biodiversity and Conservation.
6: 567-581.
Riley, S.P.D., Pollinger, J.P., Sauvajot, R.M., York, E.C., Bromley, C., Fuller, T.K. and Wayne, R.K.
(2006). A southern California freeway is a physical and social barrier to gene flow in carnivores.
Molecular Ecology. 15: 1733-1741.
Ripple, W.J. and Larsen, E.J. (2000). Historic aspen recruitment, elk, and wolves in northern Yellowstone
National Park, USA. Biological Conservation. 95: 361-370.
Robinson, J.G., Redford, K.H. and Bennett, E.L. (1999). Wildlife harvest in logged tropical forests.
Science. 284: 595-596.
Rodríguez, A., Crema, G. and Delibes, M. (1996). Use of non-wildlife passages across a high speed
railway by terrestrial vertebrates. Journal of Applied Ecology. 33: 1527-1540.
112
Rogala, J.K., Hebblewhite, M., Whittington, J., White, C.A., Coleshill, J. and Musiani, M. (2011). Human
activity differentially redistributes large mammals in the Canadian Rockies National Parks. Ecology
and Society. 16: 24 [online].
Rogers, P.M. and Myers, K. (1980). Animal distributions, landscape classification and wildlife
management, Coto Doñana, Spain. Journal of Applied Ecology. 17: 545-565.
Román, J., Barón, A. and Revilla, E. (2010a). Descripción de la red de caminos del Espacio Natural
Doñana [Road-network characterization in Doñana Natural Area]. In: Evaluación de los efectos del
tránsito a motor sobre especies y comunidades de interés en el Espacio Natural de Doñana
[Evaluation of the effects of motorized traffic on species and communities of interest within Doñana
Natural Area] (Román, J., Barón, A. and Revilla, E., Eds.). Doñana Biological Station (CSIC) Seville
(Spain), pp. 7-40.
Román, J., Barón, A. and Revilla, E. (2010b). Número de vehículos que circulan por Doñana: intensidad
de tráfico [Number of circulating vehicles in Doñana: traffic intensity]. In: Evaluación de los efectos
del tránsito a motor sobre especies y comunidades de interés en el Espacio Natural de Doñana
[Evaluation of the effects of motorized traffic on species and communities of interest within Doñana
Natural Area] (Román, J., Barón, A. and Revilla, E., Eds.). Doñana Biological Station (CSIC) Seville
(Spain), pp. 41-68.
Román, J., Barón, A. and Revilla, E. (2010c). Evaluación del impacto directo de la presencia de caminos
sobre hábitats catalogados [Evaluation of the direct impact of the presence of unpaved roads on
protected habitats]. In: Evaluación de los efectos del tránsito a motor sobre especies y comunidades
de interés en el Espacio Natural de Doñana [Evaluation of the effects of motorized traffic on species
and communities of interest within Doñana Natural Area] (Román, J., Barón, A. and Revilla, E.,
Eds.). Doñana Biological Station (CSIC) Seville (Spain), pp. 69-84.
Román, J., Barón, A. and Revilla, E. (2010d). Propuesta de optimización de la red de caminos del
Espacio Natural Doñana [Road-network optimization proposal for Doñana Natural Area]. In:
Evaluación de los efectos del tránsito a motor sobre especies y comunidades de interés en el Espacio
Natural de Doñana [Evaluation of the effects of motorized traffic on species and communities of
interest within Doñana Natural Area] (Román, J., Barón, A. and Revilla, E., Eds.). Doñana Biological
Station (CSIC) Seville (Spain), pp. 191-217.
Rost, G.R. and Bailey, J.A. (1979). Distribution of mule deer and elk in relation to roads. Journal of
Wildlife Management. 43: 634-641.
Sandom, C.J., Hughes, J. and Macdonald, D.W. (2013). Rewilding the Scottish Highlands: do wild boar,
Sus scrofa, use a suitable foraging strategy to be effective ecosystem engineers? Restoration Ecology.
21: 336-343.
Shepard, D.B., Kuhns, A.R., Dreslik, M.J. and Phillips, C.A. (2008). Roads as barriers to animal
movement in fragmented landscapes. Animal Conservation. 11: 288-296.
Soriguer, R. (1983). Consideraciones sobre el efecto de los conejos y grandes herbívoros en los pastizales
de la Vera de Doñana. Doñana Acta Vertebrata. 10: 155-168.
Stankowich, T. (2008). Ungulate flight responses to human disturbance: a review and meta-analysis.
Biological Conservation. 141: 2159-2173.
Stankowich, T. and Blumstein, D.T. (2005). Fear in animals: a meta-analysis and review of risk
assessment. Proceedings of the Royal Society B: Biological Sciences. 272: 2627-2634.
St. Clair, C. and Forrest, A. (2009). Impacts of vehicle traffic on the distribution and behaviour of rutting
elk, Cervus elaphus. Behaviour. 146: 393-413.
Suárez-Esteban, A., Delibes, M. and Fedriani, J.M. (2013a). Barriers or corridors? The overlooked role of
unpaved roads in endozoochorous seed dispersal. Journal of Applied Ecology. 50: 767-774.
Suárez-Esteban, A., Delibes, M. and Fedriani, J.M. (2013b). Unpaved road verges as hotspots of fleshyfruited shrub recruitment and establishment. Biological Conservation. 167: 50-56.
Switalski, T.A., Bissonette, J.A., DeLuca, T.H., Luce, C.H. and Madej, M.A. (2004). Benefits and
impacts of road removal. Frontiers in Ecology and the Environment. 2: 21-28.
Switalski, T.A. and Nelson, C.R. (2011). Efficacy of road removal for restoring wildlife habitat: black
bear in the northern Rocky Mountains, USA. Biological Conservation. 144: 2666-2673.
Tremblay, M.A. and St. Clair, C.C. (2009). Factors affecting the permeability of transportation and
riparian corridors to the movements of songbirds in an urban landscape. Journal of Applied Ecology.
46: 1314-1322.
113
Underhill, J.E. and Angold, P.G. (1999). Effects of roads on wildlife in an intensively modified
landscape. Environmental Reviews. 8: 21-39.
van der Ree, R., Cesarini, S., Sunnucks, P., Moore, J.L. and Taylor, A. (2010). Large gaps in canopy
reduce road crossing by a gliding mammal. Ecology and Society. 15: 16 [online].
Van Wieren, S.E. and Worm, P.B. (2001). The use of a motorway wildlife overpass by large mammals.
Netherlands Journal of Zoology. 51: 97-105.
Vos, C.C. and Chardon, J.P. (1998). Effects of habitat fragmentation and road density on the distribution
pattern of the moor frog Rana arvalis. Journal of Applied Ecology. 35: 44-56.
114
115
Capítulo 3:
Medidas de mitigación
Comparación de muestreos genéricos y específicos:
la eficiencia de diferentes pasos de fauna para micromamíferos.
______________________________________________________________________
D'Amico, M. a, Clevenger, A.P. b, Román, J.
a
and Revilla, E.
a
(Manuscript under
second review). General versus specific surveys: estimating the effectiveness of
different road-crossing structures for small mammals.
a
b
Western Transportation Institute, Bozeman (USA).
General vs specific surveys: the effectiveness of different road-crossing structures for small mammals
Resumen: La eficiencia de pasos de fauna se ha controlado menos para micromamíferos que para especies emblemáticas. Además, a causa de las dificultades en la
identificación de rastros de micromamíferos, normalmente se realizan muestreos
genéricos sin determinación específica. Considerando que las proximidades de las
carreteras a menudo están ocupadas por ambientes degradados, nuestra hipótesis es que
los
tradicionales
muestreos
genéricos
puedan
estar
sesgados,
representando
principalmente las especies generalistas de hábitat en detrimento de las especialistas.
Por esta razón, comparamos un muestreo genérico con uno específico, seleccionando
tres especies modelo: una generalista (Peromyscus maniculatus), una especialista de
bosque (Myodes gapperi) y una de pradera (Microtus pennsylvanicus). Muestreamos en
cuatro tipos de pasos de fauna (ecoductos, puentes, a sección elíptica y a sección
cuadrangular) en el Parque Nacional Banff, Canadá. Posicionamos trampas de huellas a
lo largo de los pasos, y como controles también delante de sus entradas y en los bosques
colindantes. Como hipotetizamos, el muestreo genérico no detectó diferencias en la
ocurrencia de micromamíferos entre pasos de fauna y controles. En cambio, el
específico evidenció que la única especie que usaba consistentemente los pasos era la
generalista, seleccionando mayoritariamente los ecoductos (que además fueron los
únicos pasos usados por las especialistas). Entonces la eficiencia de los pasos de fauna
para micromamíferos especialistas podría ser sobrestimada, con implicaciones para su
conservación. Como posible solución, sugerimos mejorar el hábitat (por lo menos la
disponibilidad de cobertura) a lo largo de los pasos y en el espacio entre estos y los
ambientes colindantes. Estudios que como este investiguen la eficacia de los métodos de
muestreo tradicionales, son útiles por la correcta interpretación de la eficiencia de
nuestros esfuerzos para la conservación, con el propósito último de optimizarlos.
117
Capítulo 3: Medidas de mitigación
Abstract: The use of wildlife road-crossing structures (WCS hereafter) is less
monitored for small mammals than for emblematic species. Furthermore, due to the
objective difficulties in small mammal track identification, most the authors usually
carry out general surveys without species recognition. Considering that road proximities
are often characterized by degraded environments, our hypothesis was that traditional
general surveys may be biased, mainly representing habitat generalist species in
detriment to specialist ones. For that, we compared a general small mammal survey with
a species-specific one, selecting three study species: one habitat generalist (North
American deer mouse Peromyscus maniculatus), one forest specialist (southern redbacked vole Myodes gapperi) and one prairie specialist (meadow vole Microtus
pennsylvanicus). Our study was carried out along four types of WCS (overpasses, openspan underpasses, and both elliptical and box culverts), within Banff National Park,
Canada. We positioned footprint track-tubes along the WCS, and also as a control in
front of their entrances and within the surrounding woodlands. As we suggested, using
the traditional general survey we did not detect significant differences in small mammal
presence among WCS and controls. Instead, species-specific surveys showed that the
only species consistently using the WCS was the generalist one, mainly preferring
overpasses (which also were the only WCS used by specialist species). In most cases
the inadequacy of WCS for specialist small mammals could therefore be
underestimated, with serious conservation implications. As a possible solution, we
suggest to improve the habitat (or at least the cover availability) along the WCS and in
the space between them and surrounding environments. Studies like the present,
verifying traditional survey methods, are useful for the correct interpretation of the
effectiveness of our conservation efforts, with the final aim to optimize them.
118
Introduction
Several studies in the last decades have shown that motorized traffic can
negatively affect animal populations; road-kill mortality and barrier effect being the
most documented impacts (Forman and Alexander, 1998; Trombulak and Frissell, 2000;
Forman et al., 2003). Barrier effect may have especially relevant consequences at
population level, limiting animal movements and in some cases gene flow (Merriam et
al., 1989; Gerlach and Musolf, 2000; Keller and Largiadèr, 2003; Riley et al., 2006).
Wildlife road-crossing structures (WCS) are commonly used to reduce roadrelated mortality and increase landscape connectivity for animal populations (van der
Ree et al., 2007; Glista et al., 2009); however, their effectiveness may vary among taxa
(Rodríguez et al., 1996; Clevenger and Waltho, 2005). Nevertheless, in most cases the
efficacy of WCS has been explored for large mammals (Foster and Humphrey, 1995;
Clevenger and Waltho, 2005; Grilo et al., 2008), where small mammals have received
less attention (Porto Peter et al., 2013). Furthermore, the objective difficulties in
surveying small mammal WCS use cause many authors to consider them as a category,
without species identification (Rodríguez et al., 1996; Mata et al., 2003, 2005, 2007;
Ascensão and Mira, 2007). Conversely, different species often show distinct habitat
selection, territory structure and movement patterns, and all of them can influence
species-specific WCS effectiveness.
In our opinion, this ambiguity can arise where habitat generalist species are
sympatric with specialist ones. Indeed, habitat generalists should be more likely to use
119
WCS, which in most cases are characterized by degraded environments typical of road
proximities (Umetsu and Pardini, 2007; Freitas et al., 2012). In this context, a traditional
general survey of small mammal presence (i.e., without species identification) might
only represent the habitat of generalist species, to the detriment of specialist ones, in
some cases with relevant conservation implications.
With the aim to test this hypothesis, we chose a study system in which are
coexisting in sympatry a forest specialist species (southern red-backed vole Myodes
gapperi, hereafter referred to as red-backed vole; see Merritt, 1981) with a prairie
specialist species (meadow vole Microtus pennsylvanicus; Reich, 1981) and a generalist
species (the North American deer mouse Peromyscus maniculatus, hereafter referred to
as deer mouse; Baker, 1968; Wywialowski, 1987). In a general survey of WCS
effectiveness without species identification, the presence of the ubiquitous deer mouse
could hide the absence of the two habitat specialist voles. Indeed, previous studies
actually showed that when translocating these three species to the other side of a
highway, only few individuals returned to their own territory, and almost all were deer
mice (McDonald and St Clair, 2004b). These translocations generated to territorial
adults an extreme motivation to cross back the road, but the homing ratios and the WCS
efficacy were relatively low anyway, particularly for the two specialist species
(McDonald and St Clair, 2004a, 2004b). Therefore, we suggest that WCS use in a
natural context (without homing behavior due to translocation) could be even lower
than recorded up until now, mainly involving deer mice and almost non-existent for the
two vole species.
120
The purpose of the present study is therefore to verify whether traditional
surveys of WCS efficacy without species identification are actually an appropriate
method to sample the whole category of small mammals, or if they are only estimating
the use by more generalist species. In addition, we tested whether the suitability of this
type of survey depends on the type of WCS, with the aim to better understand which
WCS feature can improve across-road connectivity for the habitat specialist species.
Study area
Banff National Park (51º15’ N, 115º30º W) is situated in Alberta (Canada),
along the Bow River Valley (within Canadian Rocky Mountains, elevation range: 13003400 m). The valley floor (always below 2000 m of elevation) is characterized by the
presence of the Trans-Canada Highway (Figure 1). This is the major transportation
corridor through Banff and Yoho National Parks, with a traffic volume relatively high
for the region (an average of 17970 vehicles per day in 2008 and increasing 2.5% per
year: Highway Service Centre, Parks Canada, unpublished data). WCS were built along
the Trans-Canada Highway, the first 27 km (Phases 1 and 2) were completed in 1988,
the next 18 km (Phase 3A) in 1997, the final 30 km (Phase 3B) are nearly finished.
The climate of the park is continental with relatively long winters and short
summers (Janz and Storr, 1977), and the vegetation encompasses montane, subalpine
and alpine ecoregions (Achuff & Corns 1983; Holland and Coen, 1983). The
transportation corridor traverses the montane ecoregion, mostly characterized by
121
coniferous forests (Douglas fir Pseudotsuga menziesii, Canadian white spruce Picea
glauca, lodgepole pine Pinus contorta, and American quaking aspen Populus
tremuloides) and natural grasslands (Achuff & Corns 1983; Holland and Coen, 1983).
Deer mice and red-backed voles are the most abundant rodents in Banff National Park
(Holland and Coen, 1983), meadow voles are also present (McDonald and St Clair,
2004a, 2004b).
Figure 1: Study area map, with locations of the surveyed WCS (overpasses, open-span bridge passes,
elliptical culverts and box culverts).
122
Data collection
During September and October 2010, we sampled the presence/absence of the
three study species within and adjacent to four different types of WCS. We surveyed
two 50-m-wide wildlife overpasses, two open-span bridge underpasses (ca. 3 m high, 11
m wide), three elliptical metal culverts (ca. 4 m high, 7 m wide) and five concrete box
culverts (2.6 m high, 3.2 m wide; Figure 2) along the Phases 1, 2 and 3A of the TransCanada Highway (Figure 1).
Overpass habitat consists of sparse young trees, shrubs and open grassland,
whereas the wildlife underpasses are void of vegetation. There was no abrupt habitat
edge between overpass habitat and surrounding woodland, in contrast with the wildlife
underpasses which have grassland verge habitat (Figure 2).
Surveys were carried out using track-plates (Mayer, 1957; Zielinski and Truex,
1995; Clevenger et al., 2001). Every track-plate consisted of a tube of 30 cm in length
and 10 cm in diameter, with a sooted metal sheet as floor (Figure 2). We disposed these
track-tubes as sample lines along every surveyed WCS and also in their respective
controls (Figure 2). Along each sample line, we placed one track-tube every 10 m. The
WCS sample lines were composed by five track-tubes (50 m length), while the control
sample lines had four track-tubes (40 m length). Within every surveyed WCS we placed
two sample lines, one for each WCS side. The first control sample lines coincided to the
WCS entrances, where they were disposed in parallel to the highway and
perpendicularly to the WCS. In correspondence to overpasses, the entrance controls
123
were characterized by the same habitat of the WCS, whereas in underpasses and
culverts these controls were located along grassland verge habitat. The second control
sample lines were placed inside the contiguous woodland, one of them in each side of
the highway, 50 m away from the WCS. Also the woodland controls were positioned
parallel to the highway and perpendicularly to the WCS (Figure 2).
The survey period consisted of two continuous weeks for every WCS
(concentrated during September-October 2010), that is enough time for our study
species to get used to track-tubes (Nams and Gillis, 2003). The track-tubes were
checked only at the end of the survey period.
Species-specific track identification was carried out according to bibliography
(Murie, 1974; Elbroch, 2003; Kays and Wilson, 2009). Furthermore, we also trapped
some individuals for every study species, with the aim to analyze in visu their footprints
on our track-tubes.
124
Figure 2: Surveyed WCS types: 1) overpass; 2) open-span bridge pass; 3) elliptical culvert; 4) box
culvert. Tree symbols represents the woodland, while grass symbols correspond to roadside grasslands.
Circles represent WCS track-tubes (black circles along the overpass, grey circles within underpasses).
Rhombus represent both control track-tubes (WCS entrances and surrounding woodland).
125
Data analysis
We performed two analyses based on the presence/absence of footprints within
the track-tubes. In the first analysis we verified whether small mammals used the WCS,
while in the second one we analyzed the small mammal WCS preferences. In both
cases, we firstly considered the small mammals as a whole category without species
identification, and afterwards we separated them by species.
The first analysis (i.e. WCS use) was constituted by generalized linear mixed
models (GLMMs; McCullagh and Nelder, 1989) with binomial error distribution and
logit link function. In the first GLMM, the response variable was small mammal
presence/absence in every given track-tube (without species identification). The only
explanatory variable was the location of the track-tube, constituted by three categories:
WCS, entrance control or woodland control. The random variable was the identification
code of every WCS.
In the following species-specific three models of WCS use, the response
variables were deer mouse, red-backed vole and meadow vole presence/absence in
every given track-tube, respectively. The explanatory and random variables were the
same of the previous general model.
The second analysis (i.e. WCS preference) was also constituted by GLMMs with
binomial error distribution and logit link function. In the first GLMM, the response
variable was small mammal presence/absence in every given WCS track-tube (without
species identification). The explanatory variables were WCS type (four categories:
126
overpass, open-span, elliptical culvert or box culvert) and track-tube continuous
distance from the centre of the WCS. We used the identification code of every WCS as a
random variable.
In the following species-specific models of WCS preference, we only considered
the species that effectively used the WCS. The response variable was the species
presence/absence in every given WCS track-tube. The explanatory and random variables
were the same of the previous general model.
Results
In the 85% of track-tubes (within WCS and both controls; n = 310) we detected
the presence of at least one of our three study species. In 28 track-tubes we found the
simultaneous presence of two species. Deer mice were the most commonly detected
species (60% of total captures, present in 56% of track-tubes), followed by red-backed
vole (25% and 23%) and meadow vole (16% and 15%).
In 86% of WCS track-tubes (n = 119) we detected small mammal tracks
(without species identification). Both controls showed similar small mammal
occurrence, since we detected footprints in 83% of WCS entrance track-tubes (n = 95)
and 84% in the woodland track-tubes (n = 96). Indeed, we did not find significant
differences in small mammal presence among WCS and both controls (parameter
estimates and standard errors for WCS, WCS entrances and woodland: 1.78 + 0.38; 1.69
+ 0.39; 1.61 + 0.29; Figure 3).
127
Focusing on single species WCS use, we found that 98% of small mammal
occurrences within WCS were deer mice. This species occurred in 84% of WCS tracktubes, but it was also widely recorded in both controls, being detected in 76% of WCS
entrance track-tubes and 33% of woodland track-tubes. The specific GLMM showed
higher presence probability along WCS track-tubes than in control track-tubes
(parameter estimates and standard errors for WCS, WCS entrances and woodland: 1.65
+ 0.33; -0.25 + 0.30; -0.69 + 0.22; F = 27.33; p < 0.001; Figure 3).
Red-backed vole scarcely used the WCS, representing only 2% of WCS small
mammal records. It was present only in 6% of WCS entrance track-tubes. Nevertheless,
red-backed vole occurrence along woodland controls was high, amounting to 66% of
track-tubes. Indeed, the specific GLMM showed that red-backed voles used more
frequently woodland track-tubes than other ones (parameter estimates and standard
errors for WCS, WCS entrances and woodland: -3.86 + 0.65; -2.89 + 0.50; 0.71 + 0.32;
F = 42.02; p < 0.001; Figure 3).
Finally, for meadow vole we only detected one record in WCS track-tubes. On
the other hand, this species occurred in 46% of WCS entrance track-tubes and there
were not records on woodlands. Its specific GLMM confirmed than WCS entrance
track-tubes were significantly more used compared with other ones (parameter estimates
and standard errors for WCS, WCS entrances and woodland: -4.91 + 1.42; -0.19 + 1.03;
-4.70 + 1.04; F = 19.19; p < 0.001; Figure 3).
128
Figure 3: General and species-specific small mammal presence probability among WCS and
controls. Results from the WCS use analysis.
We detected small mammals at all WCS (without species identification). All
WCS types were highly used, amounting the small mammal presence to 85% along
overpass track-tubes, 75% in open-span underpasses and respectively 90% and 87%
within elliptical and box culverts (Figure 4). In fact, into the second analysis, the
general GLMM confirmed the lack of differences in small mammal occurrence among
different WCS types (parameter estimates and standard errors for overpasses, bridges,
elliptical culverts and box culverts: 1.27 + 0.73; 0.57 + 0.82; 0.84 + 0.82; 1.13 + 0.83;
Figure 4). The distance from the centre of the WCS did not affect small mammal
occurrence.
129
Figure 4: General and species-specific small mammal presence probability among different WCS
types. Columns represent the percentage of used track-tubes: white columns correspond to the whole
category of small mammals without species identification, lighter grey columns to deer mouse, darker
grey columns to red-backed vole, and black columns to meadow vole. Least mean squares with error bars
represent presence probabilities resulting from the WCS preference analysis, only for the categories with
enough data (i.e. small mammals and deer mice).
Focusing on single species, the deer mouse commonly used WCS, with presence
proportions of 75% in track-tubes of both overpasses and bridges, 90% in elliptical
culverts and 87% in box culverts (Figure 4). Indeed, the specific GLMM did not show
significant differences among WCS types (parameter estimates and standard errors for
overpasses, bridges, elliptical culverts and box culverts: 0.61 + 0.64; 0.56 + 0.74; -0.47
130
+ 0.71; -0.18 + 0.75; Figure 4). Additionally, the distance from the centre of the WCS
did not affect the deer mouse presence.
Concerning both red-backed and meadow voles, all their scarce WCS records (3
and 1 for 119 WCS track-tubes, respectively) regarded overpasses (Figure 4), but they
were insufficient to statistically estimate their specific WCS preferences.
Discussion
Our results highlighted strong discrepancies between general and specific
surveys of WCS effectiveness for small mammals, questioning the suitability of the
former method. Indeed, our first analysis series showed that surveying without species
identification we could infer that small mammals widely used WCS, without significant
differences compared with controls. Nevertheless, as a result of the specific models, we
found that the only species actually using the WCS was the habitat generalist deer
mouse. This species commonly uses both wildlife overpasses and underpasses
(Clevenger et al., 2001; McDonald and St Clair, 2004a; Meaney et al., 2007), and past
studies also recorded its relative indifference towards road disturbances (Yale-Conrey
and Mills, 2001; McDonald and St Clair, 2004b; and also see for the similar whitefooted mouse Peromyscus leucopus: Rytwinski and Fahrig, 2007; McGregor et al.,
2008). Furthermore, the deer mouse is a nocturnal species, and therefore its activity
usually coincides with the hours of lower traffic volume and consequent disturbance
(Goosem, 2002; McDonald and St Clair, 2004a; McGregor et al., 2008).
131
Our analyses also showed significantly more deer mouse presence within WCS
than along controls, but also past studies on this species found an apparent preference
towards degraded habitats compared with more conserved ones (Yale-Conrey and Mills,
2001; Bissonette and Rosa, 2009; Brehme et al., 2013). In our case this finding could be
consequence of interspecific displacement by dominant red-backed and meadow voles
(Grant, 1971; Bowker and Pearson, 1975; Crowell and Pimm, 1976), which were
widely present along surrounding woodlands and roadside grasslands. Other factors
could also explain this result, such as for example the relative low densities of predators
along roads, compared with more natural habitats (Lovallo and Anderson, 1996; Riley
et al., 2006; van Langevelde et al., 2009). The generalist deer mouse could then actively
select these road-related degraded areas, in a trade-off context between habitat quality
and predation risk (Rytwinski and Fahrig, 2007, 2013; Planillo and Malo, 2013).
The effectiveness of WCS for red-backed and meadow voles was then extremely
low, because both of these species were virtually absent along them, despite being
frequently found along contiguous habitats (woodlands and roadside grasslands,
respectively). As we hypothesized, their WCS use in this natural context was even
lower than recorded in previous studies which entailed translocations of individuals
(McDonald and St Clair, 2004a). The main cause of such recorded barrier effect should
be the well-known road-associated disturbances (Oxley et al., 1974; Kozel and Fleharty
1979; Mader, 1984), also amplified in the case of the meadow vole because its circadian
activity includes daily hours, which coincide with higher traffic intensity (McDonald
and St Clair, 2004a, 2004b). Both vole species usually tend to avoid the road surface,
including translocated territorial adults with a consequent extreme motivation to homing
132
behavior (Yale-Conrey and Mills, 2001; McDonald and St Clair, 2004a, 2004b). On the
other hand, this road avoidance could also depend to a more general rejection towards
unattractive habitats (Nams and Gillis, 1998; Witt and Huntly, 2001; Russell et al.,
2007). Our findings also confirm the already described potential importance of roadside
grasslands as habitat for meadow vole (Getz et al., 1978; Kirsch, 1997), especially in
heavily forested landscapes just as our study area (Clevenger et al., 2001). More in
general, this result is an additional evidence of the potential relevance of verge habitats
for small mammals (Bellamy et al., 2000; Ascensão et al., 2012; Ruiz-Capillas et al.,
2013).
Effectiveness surveys without species identification would errantly suggest
WCS suitability for all the small mammals, but we found significant differences
between species, being the more generalist one (i.e. the deer mouse) the only actually
using the WCS. Such misinterpretation produces an overestimation of WCS use by
habitat specialist species, as observed in our study concerning red-backed and meadow
voles. The consequences can entail obvious conservation implications, in the case of
endangered or keystone species. For example, some authors consider the red-backed
vole as a forest keystone species for its role dispersing woodland ectomycorrhizal
spores (Terwilliger and Pastor, 1999; Cook and MacDonald, 2001). In this context, the
long term viability of its fragmented populations could be threatened by the barrier
effect caused by roads and associated traffic (Browne and Ferree, 2007), with possible
repercussions on the ecosystem health (Cook and MacDonald, 2001). All that
considered, we want to highlight the necessity to improve WCS effectiveness for habitat
133
specialist species, with special regard to endangered and keystone ones, especially
across heavily fragmented landscapes.
The improvement of WCS efficacy principally depends on the understanding of
specific preferences regarding these infra-structures. In our study, the only WCS type
used by red-backed and meadow voles was the wildlife overpass. Previous published
experiments showed that our study species preferred underpasses in comparison to
overpasses, possibly because the latter ones at that time, recently constructed, were
scarcely vegetated (McDonald and St Clair, 2004a). These species selected then smaller
underpasses in comparison with larger ones, probably because the former can be
perceived as safer from predators (Rodriguez et al., 1996; McDonald and St Clair,
2004a). In our study the wildlife overpasses were already characterized by a mosaic of
young woodland and grasslands, which for both habitat specialist voles potentially
represents a matrix environment with the suitable vegetative cover to minimize the
perception of predation risk (Merritt, 1981; Reich, 1981). We suggest that translocated
voles probably would return to their territories through the shorter way or the safer one
(in the study of McDonald and St Clair, 2004a it was the wildlife underpasses); while in
a more natural context without translocations, they would choose to cross the highway
only in correspondence to suitable conditions (in our study provided by vegetated
wildlife overpasses).
Our findings suggest the importance of species-specific habitat quality and
principally of vegetative cover in determine the occurrence of habitat specialist rodents
along WCS. The first inference should be encouraging the establishment of overpasses
134
as privileged WCS building choice, when local topography make it possible; especially
considering the efficacy of such infra-structures also for other road-sensitive species
(Clevenger and Waltho, 2005; Bond and Jones, 2008). On the other hand, wildlife
underpasses are comparatively much more widespread than overpasses, and a strategy
focused to improve their effectiveness for small mammals would considerably optimize
the landscape connectivity for these species. A possible suggestion in this sense can be
the improvement of underpass habitat quality, providing dead woods and brushes as
artificial cover (Clevenger et al. 2001, McDonald and St Clair, 2004a). Finally, another
complementary way to improve both overpass and underpass attractiveness for small
mammals can be to avoid the vegetation cutting at the entrance of the WCS, decreasing
then predation risk perception while approaching the WCS (Rodriguez et al., 1996;
Clevenger et al., 2001; McDonald and St Clair, 2004a). All these possible solutions
have been often suggested in the past; nevertheless their organic application is rarely
carried out. For this reason we suggest to associate them to experimental studies
focused on validate their effectiveness and the actual improvement of the WCS.
The present study highlights the unsuitability of WCS effectiveness surveys for
small mammals without species identification, showing that they may be strongly
biased towards more generalist species. Therefore, we resolutely recommend the
application of species-specific surveys. Concerning our case of study, we also discussed
the reasons of WCS inefficiency for two habitat specialist species (i.e. red-backed and
meadow vole), and consequently we suggested several possible solutions. This kind of
studies, entailing the verification of traditional survey methods, can help in the correct
135
interpretation of the effectiveness of our conservation efforts, with the final aim to
optimize them.
Acknowledgements
This research was partially supported by the JAE-PRE PhD stay fellowship
granted to MD by the CSIC. MD was supported by the JAE-PRE fellowship from the
CSIC. Nikki Heim helped with fieldwork, Banff National Park (Parks Canada), Sofia
Conradi-Fernández and Marilù Bastianelli provided logistic support. Richard T.T.
Forman and the CarNe (EBD Carnivore Network, especially Nestor Fernández)
provided useful comments.
References
Ascensão, F., Clevenger, A.P., Grilo, C., Filipe, J. and Santos-Reis, M. (2012). Highway verges as habitat
providers for small mammals in agrosilvopastoral environments. Biodiversity and Conservation. 21:
3681-3697.
Ascensão, F. and Mira, A. (2007). Factors affecting culvert use by vertebrates along two stretches of road
in southern Portugal. Ecological Research. 22: 57-66.
Baker, R.H. (1968). Habitats and distribution. In: Biology of Peromyscus (Rodentia). King, J.A. (Eds.).
American Society of Mammalogists, Special Pubblication Number 2, Stillwater (USA), pp. 98-126.
Bellamy, P.E., Shore, R.F., Ardeshir, D., Treweek, J.R. and Sparks, T.H. (2000). Road verges as habitat
for small mammals in Britain. Mammal Review. 30: 131-139.
Bissonette, J.A. and Rosa, S.A. (2009). Road zone effects in small-mammal communities. Ecology and
Society. 14: 15 [online].
Bond, A.R. and Jones, D.N. (2008). Temporal trends in use of fauna-friendly underpasses and overpasses.
Wildlife Research. 35: 103-112.
Bowker, L.S. and Pearson, P.G. (1975). Habitat orientation and interspecific interaction of Microtus
pennsylvanicus and Peromyscus leucopus. American Midland Naturalist. 94: 491-496.
Browne, R.A. and Ferree, P.M. (2007). Genetic structure of southern Appalachian “sky island”
populations of the southern red-backed vole (Myodes gapperi). Journal of Mammalogy. 88: 759-768.
Clevenger, A.P., Chruszcz, B. and Gunson, K. (2001). Drainage culverts as habitat linkages and factors
affecting passage by mammals. Journal of Applied Ecology. 38: 1340-1349.
Clevenger, A.P., Ford, A.T. and Sawaya, M.A. (2009). Banff wildlife crossings project: integrating
science and education in restoring population connectivity across transportation corridors. Parks
Canada Agency, 165 pp., Radium Hot Springs (Canada).
136
Clevenger, A.P. and Waltho, N. (2005). Performance indices to identify attributes of highway crossing
structures facilitating movement of large mammals. Biological Conservation. 121: 453-464.
Coffman, C.J., Nichols, J.D. and Pollock, K.H. (2001). Population dynamics of Microtus pennsylvanicus
in corridor-linked patches. Oikos. 93: 3-21.
Collins, R.J. and Barrett, G.W. (1997). Effects of habitat fragmentation on meadow vole (Microtus
pennsylvanicus) population dynamics in experiment landscape patches. Landscape Ecology. 12: 6376.
Cook, J.A. and MacDonald, S.O. (2001). Should endemism be a focus of conservation efforts along the
North Pacific coast of North America? Conservation Biology. 97: 207-213.
Crowell, K.L. and Pimm, S.L. (1976). Competition and niche shifts of mice introduced onto small
islands. Oikos. 27: 251-258.
Delibes-Mateos, M., Redpath, S.M., Angulo, E., Ferreras, P. and Villafuerte, R. (2007). Rabbits as a
keystone species in southern Europe. Biological Conservation. 37: 149-156.
Elbroch, M. (2003). Mammal tracks and sign. A guide to North American species. Stackpole Books, 779
pp., Mechanicsburg (USA).
Ford, A.T., Clevenger, A.P. and Bennett, A. (2009). Comparison of methods of monitoring wildlife
crossing-structures on highways. Journal of Wildlife Management. 73: 1213-1222.
Foster, M.L. and Humphrey, S.R. (1995). Use of highway underpasses by Florida panthers and other
wildlife. Wildlife Society Bulletin. 23: 95-100.
Freitas, S.R., Alexandrino, M.M., Pardini, R. and Metzger, J.P. (2012). A model of road effect using line
integrals and a test of the performance of two new road indices using the distribution of small
mammals in an Atlantic forest landscape. Ecological Modelling. 247: 64-70.
Gerlach, G. and Musolf, K. (2000). Fragmentation of landscape as a cause for genetic subdivision in bank
voles. Conservation Biology. 14: 1066-1074.
Getz, L.L., Cole, F.R. and Gates, D.L. (1978). Interstate roadsides as dispersal routes for Microtus
pennsylvanicus. Journal of Mammalogy. 59: 208-212.
Goosem, M. (2002). Effects of tropical rainforest roads on small mammals: fragmentation, edge effects
and traffic disturbance. Wildlife Research. 29: 277-289.
Grilo, C., Bissonette, J. and Santos-Reis, M. (2008). Response of carnivores to existing highway culverts
and underpasses: implications for road planning and mitigation. Biodiversity and Conservation. 17:
1685-1699.
Grant, P.R. (1971). Experimental studies of competitive interaction in a two-species system. III. Microtus
and Peromyscus species in enclosures. Journal of Animal Ecology. 40: 323-350.
Hardy, A., Clevenger, A.P., Huijser, M. and Neale, G. (2004). An overview of methods and approaches
for evaluating the effectiveness of wildlife crossing structures: emphasizing the science in applied
science. In: Proceedings of the 2003 International Conference on Ecology and Transportation
(ICOET). Irwin, C.L., Garrett, P., McDermott, K.P. (Eds.). Center for Transportation and the
Environment and North Carolina State University, Raleigh (USA), pp. 319-330.
Howe, H.F., Brown, J.S. and Zorn-Arnold, B. (2002). A rodent plague on prairie diversity. Ecology
Letters. 5: 30-36.
Howe, H.F., Zorn-Arnold, B., Sullivan, A. and Brown, J.S. (2006). Massive and distinctive effects of
meadow voles on grassland vegetation. Ecology. 87: 3007-3013.
Holland, W.D. and Coen, G.M. (1983). Ecological (biophysical) land classification of Banff and Jasper
National Parks. Alberta Institute of Pedology (Publication M-83-2), 193 pp., Edmonton (Canada).
Janz, B. and Storr, D. (1977). The climate of the contiguous National Parks: Banff, Jasper, Kootenay,
Yoho. Application and Consultation Division (Meteorological Applications Branch, Atmospheric
Environment Service, Environment Canada), 324 pp., Toronto (Canada).
Kays, R.W. and Wilson, D.E. (2009). Mammals of North America. Princeton Field Guides. Princeton
University Press, 248 pp., Princeton (USA).
137
Keinath, D.A. and Hayward, G.D. (1999). Red-backed vole (Clethrionomys gapperi) response to
disturbance in subalpine forests: use of regenerating patches. Journal of Mammalogy. 84: 956-966.
Keller, I. and Largiadèr, C.R. (2003). Recent habitat fragmentation caused by major roads leads to
reduction of gene flow and loss of genetic variability in ground beetles. Proceedings of the Royal
Society B: Biological Sciences. 270: 417-423.
Kirsch, E.M. (1997). Small mammal community composition in cornfields, roadside ditches, and prairies
in eastern Nebraska. Natural Areas Journal. 17: 204-211.
Lin, Y.K. and Batzli, G.O. (1995). Predation on voles: an experimental approach. Journal of Mammalogy.
76: 1003-1012.
Lovallo, M.J. and Anderson, E.M. (1996). Bobcat (Lynx rufus) home range size and habitat use in
northwest Wisconsin. American Midland Naturalist. 135: 241-252.
Mata, C., Hervàs, I., Herranz, J., Suàrez, F. and Malo, J.E. (2003). Effectiveness of wildlife crossing
structures and adapted culverts in a highway in northwest Spain. In: Proceedings of the 2003
International Conference on Ecology and Transportation (ICOET). Irwin, C.L., Garrett, P.,
McDermott, K.P. (Eds.). Center for Transportation and the Environment and North Carolina State
Mata, C., Hervás, I., Herranz, J., Suárez, F. and Malo, J.E. (2005). Complementary use by vertebrates of
crossing structures along a fenced Spanish motorway. Biological Conservation. 124: 397-405.
Mayer, W.V. (1957). A method for determining the activity of burrowing mammals. Journal of
Mammalogy. 38: 531.
McCullagh, P. and Nelder, J.A. (1989). Generalized linear models. Monographs on statistics and applied
probabibilty. Chapman and Hall, 511 pp., London (UK).
McDonald, W. and St Clair, C.C. (2004a). Elements that promote highway crossing structure use by
small mammals in Banff National Park. Journal of Applied Ecology. 41: 82-93.
McDonald, W. and St Clair, C.C. (2004b). The effects of artificial and natural barriers on the movement
of small mammals in Banff National Park, Canada. Oikos. 105: 397-407.
McGuire, T.M. and Morrall, J.F. (2000). Strategic highway improvements to minimize environmental
impacts within the Canadian Rocky Mountain National Parks. Canadian Journal of Civil Engineering.
27: 523-532.
Meaney, C., Bakeman, M., Reed-Eckert, M. and Wostl, E. (2007). Effectiveness of ledges in culverts for
small mammal passage. Colorado Department of Transportation - Research Branch, 46 pp., Denver
(USA).
Merriam, G., Kozakiewicz, M., Tsuchiya, E. and Hawley, K. (1989). Barriers as boundaries for
metapopulations and demes of Peromyscus leucopus in farm landscapes. Landscape Ecology. 2: 227235.
Miller, B., Reading, R., Hoogland, J., Clark, T., Ceballos, G., List, R., Forrest, S., Hanebury, L.,
Manzano, P., Pacheco, J. and Uresk, D. (2000). The role of prairie dogs as a keystone species:
response to Stapp. Conservation Biology. 14: 318-321.
Nams, V.O. and Gillis, E.A. (2003). Changes in tracking tube use by small mammals over time. Journal
of Mammalogy. 84: 1374-1380.
Planillo, A. and Malo, J.E. (2013). Motorway verges: paradise for prey species? A case study with the
European rabbit. Mammalian Biology. 78: 187-192.
Porto Peter, F., Molina-Vacas, G., Rodriguez, J. and Grilo, C. (2013). Effects of roads on spatial
behaviour and abundance of small mammals: gaps in knowledge. Oecologia Australis. 17: 63-76.
Riley, S.P.D., Pollinger, J.P., Sauvajot, R.M., York, E.C., Bromley, C., Fuller, T.K. and Wayne, R.K.
(2006). A southern California freeway is a physical and social barrier to gene flow in carnivores.
Molecular Ecology. 15: 1733-1741.
Rodríguez, A., Crema, G. and Delibes, M. (1996). Use of non-wildlife passages across a high speed
railway by terrestrial vertebrates. Journal of Applied Ecology. 33: 1527-1540.
Ruiz-Capillas, P., Mata, C. and Malo, J.E. (2013). Road verges are refuges for small mammal populations
in extensively managed Mediterranean landscapes. Biological Conservation. 158: 223-229.
138
Rytwinski, T. and Fahrig, L. (2007). Effect of road density on abundance of white-footed mice.
Landscape Ecology. 22: 1501-1512.
Rytwinski, T. and Fahrig, L. (2013). Why are some animal populations unaffected or positively affected
by roads? Oecologia. 173: 1143-1156.
Smith, A.T. and Foggin, M. (1998). The plateau pika (Ochotona curzoniae) is a keystone species for
biodiversity on the Tibetan plateau. Animal Conservation. 2: 235-240.
Terwilliger, J. and Pastor, J. (1999). Small mammals, Ectomycorrhizae, and conifer succession in beaver
meadows. Oikos. 85: 83-94.
Umetsu, F. and Pardini, R. (2007). Small mammals in a mosaic of forest remnants and anthropogenic
habitats - evaluating matrix quality in an Atlantic forest landscape. Landscape Ecology. 22: 517-530.
van der Ree, R., Gulle, N., Holland, K., van der Grift, E., Mata, C. and Suarez, F. (2007). Overcoming the
barrier effect of roads - How effective are mitigation strategies? In: Proceedings of the 2007
International Conference on Ecology and Transportation (ICOET). Irwin, C.L., Garrett, P.,
McDermott, K.P. (Eds.). Center for Transportation and the Environment and North Carolina State
van Langevelde, F., van Dooremalen, C. and Jaarsma, C.F. (2009). Traffic mortality and the role of minor
roads. Journal of Environmental Management. 90: 660-667.
Witt, W.C. and Huntly, N. (2001). Effects of isolation on red-backed voles (Clethrionomys gapperi) and
deer mice (Peromyscus maniculatus) in a sage steppe matrix. Canadian Journal of Zoology. 79: 15971603.
Wywialowski, A.P. (1987). Habitat structure and predators: choices and consequences for rodent habitat
specialists and generalists. Oecologia. 72: 39-45.
Yale-Conrey, R.C. and Mills, L.S. (2001). Do highways fragment small mammal populations? In:
Proceedings of the 2001 International Conference on Ecology and Transportation (ICOET). Irwin,
C.L., Garrett, P., McDermott, K.P. (Eds.). Center for Transportation and the Environment and North
Carolina State University, Raleigh (USA), pp. 448-457.
Zielinski, W.J. and Truex, R.L. (1995). Distinguishing tracks of marten and fisher at track-plate stations.
Journal of Wildlife Management. 59: 571-579.
139
Capítulo 4:
Reproducción y comportamiento
El tráfico motorizado como elemento de estrés
que estimula los cuidados parentales en aves.
_____________________________________________________________________________________
Blas, J. a, Abaurrea, T. a, D’Amico, M. a, Barcellona, F. a, Revilla, E. a, Román, J. a and
Carrete, M.
ab
(Manuscript under second review). Motor traffic as a stressor eliciting
parental care in birds.
a
Doñana Biological Station (EBD-CSIC), Seville (Spain).
b
University Pablo de Olavide, Seville (Spain).
Motor traffic as a stressor eliciting parental care in birds
Resumen: Algunas aves acuden a carreteras para la búsqueda de alimento y
lugares de nidificación, pero el estrés asociado al tráfico puede amenazar su éxito
individual. Entender estos procesos puede ser relevante dentro de áreas protegidas,
donde el poco tráfico debido a actividades de conservación puede paradójicamente
transformar hábitats adecuados en trampas ecológicas. En este estudio averiguamos si
los abejarucos Merops apiaster que crían en las áreas laterales de las vías perciben el
tráfico como un elemento de estrés que estimule alarmas, estampidas y un decremento
de cuidados parentales, o si alternativamente demuestran habituación. Comparamos
individuos sometidos a diferentes niveles de tráfico dentro de una Reserva de la
Biosfera en España. Los marcados cambios diarios y semanales en niveles de tránsito
proporcionaban un diseño cuasi-experimental para discernir los efectos del tráfico per
se. Alarmas y estampidas fueron mayores en zonas de tráfico intenso, donde también se
detectaron atropellos. Los abejarucos perciben el tránsito como un elemento de estrés y
son incapaces de modular respuestas de evitación de riesgo. Contrariamente a nuestras
predicciones las tasas de ceba parentales fueron mayores en carreteras de tránsito
elevado, aumentando durante días laborales y con picos en horas punta, y aumentando
potencialmente el éxito reproductivo de los abejarucos. Por otro lado, hasta que no se
determinen las consecuencias netas sobre salud individual, niveles de estrés y eficacia
biológica, se deberían mantener los actuales niveles de tráfico y reducir los límites de
velocidad (en la temporada de cría) para disminuir la mortalidad. Debido a que los
resultados esperados pueden parecer opuestos si monitoreados durante el apareamiento
o la cría, aconsejamos cautela a la hora de inferir eficacia biológica a partir de
comportamientos aislados o de determinados estadios del ciclo vital de una especie.
141
Capítulo 4: Reproducción y comportamiento
Abstract: Several avian species are attracted to roads for foraging or nesting,
but individual performance can be jeopardized by exposure to traffic. Understanding the
effects of traffic can be relevant within protected areas, where a relative low level of
transit allowing conservation-oriented activities could paradoxically transform suitable
habitats into ecological traps. Here we questioned whether roadside breeding bee-eaters
Merops apiaster perceive traffic as a stressor eliciting predation-avoidance behaviors
and decreasing parental care, or they alternatively show habituation. Comparisons were
established between birds exposed to contrasting traffic levels within a Biosphere
Reserve in Spain. The marked daily and weekly changes in road use provided a quasiexperimental design to disentangle the effects of traffic per se from other sources of
variability. Alarm calls and risk-avoidance behaviors were related to traffic intensity,
being considerably higher in breeding colonies beside high-traffic roads compared to
controls. Road-kills were only detected in colonies exposed to high traffic. Bee-eaters
perceive traffic as a stressor and were unable to down-modulate risk-avoidance
responses. Contrary to our predictions parental feeding rates were highest along hightraffic roads, increasing during weekdays and peaking during traffic rush hours. Traffic
positively affects nesting feeding rates, potentially improving the breeding performance.
Until the net consequences of traffic on individual health, stress levels and fitness are
determined, managers are encouraged to maintain the present transit levels and reduce
speed limits (at least during breeding season) to minimize mortality. Because the
expected outcome of traffic on performance can be opposed when responses are
monitored during mating or brooding, caution should be taken before inferring fitness
consequences from isolated behaviors or life history sub-stages.
142
Introduction
The development of motor vehicles and the associated expansion of road
networks throughout the last century have largely contributed to the dramatic increase
of anthropogenic impacts currently affecting most habitats worldwide (Trombulak and
Frissell, 2000; Forman et al., 2003). Among other effects, roads and motor traffic often
result in increased wildlife casualties (Clevenger et al., 2003; Coelho and Kindel, 2008)
to the extent of changing demographic parameters (McLellan and Shackleton, 1988;
Ferreras et al., 1992), facilitate the introduction of exotic species (Mortensen et al.,
2009; D'Amico et al., 2013) and reduce the availability of suitable habitats through
barrier effects and fragmentation (Vos and Chardon, 1998; McGregor et al., 2008),
thereby reducing biodiversity (Findlay and Bourdages, 2000; Blasco et al., 2008).
However, wild birds and also other vertebrates can show contrasting behavioral
responses to roads and motor traffic. While some species display behavioral plasticity in
response to this kind of human disturbance (e.g. progressive reduction of flushing
distances; Stankowich and Blumstein, 2005; Rees et al., 2005) and may even benefit
from traffic (Peris and Pescador, 2004; Fahrig and Rytwinski, 2009), others show
disrupted activity patterns and use of space (Bautista et al., 2004; Bennett et al., 2013),
sexual and territorial behavior (Bee and Swanson, 2007; Dalal et al., 2011), and parental
care (McGowan and Simons, 2006; Schroeder et al., 2012), as well as elevated stress
levels (Crino et al., 2011; Strasser and Heath, 2013), potentially reducing individual
fitness and long-term population viability (Frid and Dill, 2002; Gibbs and Shriver,
2002).
143
A major goal of the global Biosphere Reserve network is the preservation of
biodiversity, but many reserves are affected by the impacts of roads and associated
traffic (e.g. Ramp Wilson and Croft, 2006; Coelho and Kindel, 2008; Velando and
Munilla, 2011). Even if public access is regulated or restricted, management and
research activities often involve the transit of motor vehicles within the reserves. This
provides grounds for a paradox that has been rarely tested: that conservation-related
road traffic can actually become a perturbation potentially jeopardizing the preservation
of biodiversity. Management of human disturbance thus requires a prior understanding
of the mechanisms by which anthropogenic stressors promote changes in individual
behavior, and how they may affect population distribution and abundance (Boyle and
Samson, 1985; Sorice et al., 2006).
Here, we test whether moderate levels of road traffic associated to management
and research activities within a Biosphere Reserve modify the behavior of roadside
breeding birds. As study model we used a population of European bee-eaters Merops
apiaster Linnaeus, 1758, a burrow-nesting insectivorous bird, present in Doñana
National Park and MAB Biosphere Reserve (Spain). The particularities of this bird
population offer a unique model to study the effects of traffic on wildlife behavior for
several reasons.
First, because the flat landscape lacks natural soil cliffs (typically selected by
bee-eaters for nesting elsewhere; Casas-Crivillé and Valera, 2005; Smalley et al., 2013),
birds readily dig their burrows in the slight slopes of the edges and ditches of unpaved
roads (thus literally breeding beside vehicle transit areas). Second, because access to
144
most areas within the National park is restricted to authorized park staff and researchers,
our model is suitable for assessing the specific impact of scientific and management
activities within a Biosphere Reserve. Finally, because recent studies have revealed high
variability in the spatial and temporal dynamics of road transit within the study area
(Revilla et al., 2010) we can (i) perform an a priori selection of sampling locations
exposed to contrasting vehicle transit levels, and also (ii) perform an a priori, quasiexperimental sampling design aimed at comparing behaviors within and between days
differing in traffic intensity (e.g. weekdays vs weekends; rush vs post-rush traffic hours;
see details below).
We addressed the following hypotheses: (1) road traffic is perceived by breeding
bee-eaters as a predatory threat (risk-disturbance hypothesis; Frid and Dill, 2002;
Fernández-Juricic et al., 2003), and (2) increased levels of traffic interfere with normal
breeding activities reducing parental care. According to the risk-disturbance hypothesis,
we predicted a higher frequency of risk-avoidance responses (such as those elicited by
the presence of potential predators, e.g. vocalization of alarm calls, massive flushing
and temporary abandonment of nesting activities; Kossenko and Fry, 1998; Petrescu
and Adam, 2001) among birds breeding beside high-traffic roads compared to birds
breeding beside low-traffic roads. We also predicted a direct positive association
between traffic intensity and risk-avoidance behaviors. Our second hypothesis is an
offset of the risk-disturbance hypothesis based on ample evidence that wildlife exposed
to stressful stimuli require shifting energy and time away from behaviors such as
foraging (Gill et al., 1996; Williams et al., 2006) and parental care (McGowan and
145
Simons, 2006; Baudains and Lloyd, 2007), allowing us predicting a negative association
between traffic intensity and nestling feeding rates.
Study area
The study was carried out in Doñana National Park (southwestern Spain; Figure
1), a Biosphere Reserve and UNESCO World Heritage Site, Ramsar Site and European
Community Special Protection Area considered one of the most important wetlands in
the world (García-Novo and Marín, 2006). Here some 427 species of vertebrates
(including 6 million migratory birds) coexist with traditional human activities such as
open-range cattle-raising, agriculture, shell-fishing and agro-environmental tourism.
The main habitats are marshlands, Mediterranean scrublands and coastal sand dunes
(Valverde, 1958).
The European bee-eater is a medium-sized (27-29 cm), migratory, monogamous
and highly gregarious bird (Cramp, 1985; Valera, 2009). Bee-eaters are colonial
breeders nesting inside ground cavities excavated at the end of 70-200 cm deep
burrows, generally in vertical or very sloped cliffs (Casas-Criville and Valera, 2005)
although in our study area they nest in nearly flat sandy soils (Alvarez and Hiraldo,
1974). Birds are present in Doñana between March and August, starting socio-sexual
activity and nest-digging in April-May, laying one asynchronous clutch (4-7 eggs) in
May-June, and brooding the nestlings in June-July. The species shows little sexual
146
dimorphism, and both sexes incubate, brood and care for the nestlings (Cramp, 1985;
Jones et al., 1991). Bee-eaters feed almost exclusively on flying insects caught on the
wing (Krebs and Avery, 1984; 1985; Kossenko and Fry, 1998).
Figure 1: Map of the study area, with the Iberian Peninsula and Doñana Biosphere Reserve outlined in
grey, and enlarged study area showing the four main bee-eater colonies located beside high- or low-traffic
roads (dashed and dotted lines, respectively). Note that for ease of presentation not all the roads are
represented.
147
Data collection
Access to the study area is restricted to park staff and authorized researchers,
implying that traffic levels are exclusively related to management and scientific
activities. Recent studies have revealed a high variability in the spatial and temporal
dynamics of road transit therein, allowing us selecting a priori bee-eaters colonies with
similar habitat but exposed to contrasting vehicle transit levels.
The highest traffic intensity of the unpaved road-network occurs through the
gravel road leading to the Palacio de Doñana (i.e. Research Station in Figure 1). Here
vehicle transit during spring averages 80.9 cars/day (Revilla et al., 2010) and shows the
following particularities: (i) traffic intensity dramatically decreases during weekends
compared to weekdays, and (ii) weekday traffic shows a marked daily peak between
14:00-15:00 (traffic rush time onwards). Bee-eaters nests are frequent along the ditches
of this road and show several colonial aggregations that will be further referred to as
high-traffic colonies. On the other extreme, the lowest intensity of traffic in the park
occurs in secondary, dirt roads, where average transit is less than one order of
magnitude below the gravel road described above (less than two cars/day; Revilla et al.,
2010). For comparative purposes we chose two low-traffic bee-eaters colonies located
in the ranch of Hato Barrera (Figure 1).
In order to have a more specific knowledge of the traffic affecting the selected
study colonies, we also placed automatic magnetometers (TRAFx Vehicle Counter
Generation III) beside the roads crossing our high-traffic and low-traffic bee-eaters
colonies during two weeks of May 2012.
148
During the early stages of the reproductive season (May 2012), when bee-eaters
were engaged in nest-digging and high socio-sexual activity we performed 26
behavioral observation sessions in four focal breeding sites (two high- and two lowtraffic colonies). The observations were performed from camouflaged blinds during 60
minutes, recording (i) total number of birds present in the colony (maximum number of
birds counted every other minute), (ii) number of passing predators (both raptors and
carnivores), (iii) number of passing vehicles, (iv) number of alarm calls, (v) number of
flock flushing events (sudden take-off of all the birds present in the colony or at least 10
birds simultaneously, followed by rapid spiral upwards described as panic-flight in
Cramp, 1985). No other source of disturbance was detected during the observations.
Walking transects were performed upon finalization of the observation sessions with the
aim of recording road kills and injured birds along the road segments adjacent to the
colonies.
During the late stages of the reproductive season (July 2012), when bee-eaters
were brooding the nestlings, we monitored parental feeding rates in nests located beside
the high (N = 75) and low (N = 29) traffic roads described above. Feeding events were
recorded using a small size video camera (GoPro HD Hero ®) attached to 15-20 cm
long sticks and placed on the ground, 50-70 cm away from the entrance of each focal
nest. During the 24 hours prior to data collection, dummy cameras were left in
recording position to allow habituation before actual data recording. The resulting 30minutes long video clips (60 fps; 848x480 pixels) were visualized on color computer
149
screens to count the number of feeding events. These occurred at the entrance of the
tunnel or inside the burrow, but in all cases we verified adults carrying a prey upon
arrival and leaving the nest without it. All the recordings were performed in the
afternoons (14:00- 19:30 h), allowing us to establish comparisons between a priori rush
(14:00- 15:00 h) and post-rush traffic hours (15:00-19:30 h) as well as between days
differing in traffic intensity (i.e. weekdays vs weekends) in nests exposed to high traffic.
Data analysis
Differences in traffic intensity (i.e. average daily or hourly vehicle passages)
between and within colonies (data from the TRAFx Vehicle Counter, see above) were
analyzed by means of t-tests aimed at verifying consistency with the patterns described
by Revilla et al., (2010), which were considered the a priori working hypotheses (i.e.
higher traffic in high-traffic roads, during weekdays and rush hour).
The frequency of potential risks (i.e. either predator and vehicle passages per
hour) and risk-avoidance behaviors (i.e. either alarm calls and flock flushing events per
hour) recorded through direct observations as well as the nestling feeding frequencies
extracted from video recordings were analyzed through generalized linear mixed models
(GLMMs; SAS; Littel et al., 1996). Models used normal or poisson error distribution
and identity or logarithmic link function, respectively; depending on the fit of response
variable to normal function (Kolmogorov-Smirnov test; McCullagh and Nelder, 1989;
150
Zuur et al., 2007). The models included colony identity as a random factor, and
considered sampling date and hour as covariates.
All the GLMMs regarding the frequencies of potential risks (i.e. either predator
and vehicle passages per hour) and risk-avoidance behaviors (i.e. either alarm calls and
flock flushing events per hour) tested the differences between high- and low-traffic
colonies, while the models concerning nestling feeding frequency tested the differences
between: high- and low-traffic colonies, weekdays and weekend (only considering hightraffic colonies), rush hours and post-rush hours (only considering weekdays of hightraffic colonies).
Results
The average daily vehicle passages through high-traffic bee-eater colonies were
more than twenty times higher than through low-traffic colonies (mean and SE = 22.43
± 5.86 vs 0.86 ± 1.29 vehicles/day; t = 3.63; p < 0.01; Figure 2).
Restricting the analyses to the high-traffic colonies, the average daily vehicle
passages during weekdays were more than two times higher than during weekends
(27.30 ± 7.71 vs 10.25 ± 2.52 vehicles/day; t = 2.102; one tailed p = 0.03). In the lowtraffic colonies, weekday effects were absent.
In the high-traffic colonies, the average hourly vehicle passages were almost
three times higher at rush traffic hours compared to post-rush hours during weekdays
(8.40 ± 1.17 vs 2.83 ± 0.24 vehicles/hour; t = 4.67; p < 0.001), but not during weekends.
In the low-traffic colonies there were no rush hours.
151
Figure 2: Frequency of vehicle passage (mean ± SE vehicles per hour) beside the bee-eater nesting
grounds characterized as high-traffic areas (A) and low-traffic areas (B) within Doñana National Park.
Light grey and dark grey bars indicate weekdays and weekends respectively.
The estimated number of adults in the four focal breeding sites was 35 and 40 in
the high-traffic colonies, and 20 and 45 in the low-traffic ones. There were not
differences comparing predator passage per hour (black kites Milvus migrans in all the
observations) between high- and low-traffic areas, while high-traffic areas experienced
significantly more vehicle passage per hour than low-traffic areas during the
observations sessions (F1,21 = 13.83; p < 0.005).
Bee-eaters responded to predators with alarm calls (in 88% of the occasions),
flock flushing (88%) or both risk-avoidance behaviors simultaneously (81%). Vehicle
passage elicited the vocalization of alarm calls (62%), flock flushing (100%) or both
risk-avoidance behaviors simultaneously (in 62% of the occasions).
152
Alarm calls per hour were unrelated to predator passage per hour (and also to
sampling hour), but positively related to vehicle passage per hour (F1,20 = 11.52; p <
0.01; Figure 4A) and negatively related to sampling date (F1,20 = 5.22; p = 0.03).
Flock flushing events per hour were positively associated to both predator and
vehicle passage per hour (F1,18 = 18.79; p < 0.001 and F1,18 = 53.05; p < 0.001
respectively; Figure 4B) and negatively related to sampling date and hour (F1,18 = 12.60
; p < 0.01 and F1,18 = 4.96; p = 0.04).
Although no road fatalities were recorded in the low-traffic colonies, four birds
were found road-killed in the high-traffic ones.
Figure 3: Frequency (mean ± SE events per hour) of predators and vehicle passages, alarm calls
and flock flushing events, recorded in bee-eater colonies located in high-traffic areas (dark circles) and
low-traffic areas (light circles).
153
Figure 4: Frequency of alarm calls (A) and flock flushing events (B) occurring in bee-eater colonies in
relation to the number of vehicles registered during one-hour observation sessions (N = 26 sessions).
Black and white dots indicate observations performed in high-traffic areas and low-traffic areas
respectively.
Nestling feeding frequency was statistically higher in nests located in high-traffic
colonies compared to nests located in low-traffic colonies (F1,63 = 51.06; p < 0.01;
Figure 5A), decreased with sampling date (F1,63 = 34.57; p < 0.01) and was independent
of the sampling hour.
Restricting the analyses to high-traffic colonies, nestling feeding frequency was
higher during weekdays compared to weekends (F1,65 = 35.73; p < 0.01 ; Figure 5B),
even accounting for the effects of both sampling date and hour (F1,65 = 10.84; p < 0. 01
and F1,65 = 8.57; P < 0.01).
During the weekdays the effect of traffic rush time was statistically significant,
whit nestling feeding frequency increasing at rush traffic hours compared to post-rush
hours (F1,51 = 16.78; p < 0.01; Figure 5C).
154
Figure 5: Feeding rate (mean ± SE feeds per 30 minutes) in bee-eater nests as a function of (A) study
area (high- and low-traffic areas respectively), (B) day of the week (only nests in high-traffic areas), and
(C) rush traffic time (only nests in high-traffic areas, triangles and circles indicate weekdays and
weekends respectively).
Discussion
Our results concerning alarm calls and flock flushing events as a response to
both predators and vehicles indicate that motor traffic is perceived by European beeeaters as a threat comparable to a predator, supporting the risk-disturbance hypothesis
(Frid and Dill, 2002; Fernández-Juricic et al., 2003). Thus, even when our study sites
did not differ in the abundance of natural predators, individuals living in colonies
subjected to different degree of vehicle traffic experienced variable levels of risk which
can have consequences for their fitness.
Alarm calls are typically used by bee-eaters when detecting predators and in
response to conspecific distress calls (Jilka and Ursprung, 1980; Kossenko and Fry,
1998). As a response, the typical reactions are risk-avoidance flights and the rapid
155
retreat of nestlings inside the burrows (Reid, 1974; Jilka and Ursprung, 1980) likely
imposing energetic costs to signalers and receptors (Bergstrom and Lachmann, 2001).
Although alarm vocalizations were not systematically elicited by vehicle passage, they
were generally associated to this perturbation and their frequency was positively and
statistically related to traffic intensity. Moreover, flock flushing consistently occurred
when we observed passing vehicles (100%, N = 92). This result is expected according
to the risk-disturbance hypothesis which posits that fleeing probability increases when
the threats approach more directly, more quickly and are larger in size (Frid and Dill,
2002; Fernández-Juricic et al., 2003).
The road-kills recorded in high-traffic colonies may reinforce a lack of
habituation to this stressor (Lima et al., 2014), but even in absence of fatal outcomes
bee-eaters likely maximize fitness by overestimating rather than underestimating risks,
because the costs of overestimation (e.g. resource utilization in traffic-avoidance
behaviors) have milder fitness consequences than the cost of underestimating (which
might be immediate death; Bouskila and Blumstein, 1992; Byrnes et al., 2012). The
recorded responses to traffic may impose other less immediate costs, such as larger
energy expenditure.
The effects of traffic on parental provisioning rates were completely opposed to
our second hypothesis. Breeding bee-eaters provided more, rather than less food to
nestlings in colonies located beside high-traffic roads, and here feeding rates increased
during the days of the week experiencing more traffic and during rush traffic hours. The
quasi-experimental nature of our sampling design allowed us to rule out the possibility
156
that these results were an artifact of daily rhythms in adult hunting activities (Inglisa
and Galeotti, 1993) or parental adjustments in food provisioning as nestlings’ growth
(Lessells and Avery, 1989). Although both factors could affect feeding rates (explaining
why sampling date and/or hour were statistically significant in some of our models)
they would be unable to account for the decreased parental provisioning recorded during
weekends (when local traffic decreases), and the recorded increase at rush traffic times
(which does not occur on weekends, when rush traffic peaks are negligible).
Previous studies on other avian species have reported increased foraging
efficiency along roads or roadside habitats, but to our knowledge no study has
previously shown a positive effect of traffic intensity per se. For example, Florida
scrub-jays Aphelocoma coerulescens show higher rates of energy intake on roadside
habitats (Morgan et al., 2010) and male white-crowned sparrows Zonitrichia leucophrys
oriantha increase nestling feeding rates with decreased distance to roads (Crino et al.,
2011).
At least three alternative, non-mutually exclusive hypotheses may explain the positive
association we found between traffic levels and parental feeding in bee-eaters, namely
(1) facilitation, (2) unmatched cue and (3) interference-displacement hypothesis.
First, road transit could elevate the availability or detectability of food resources,
facilitating (improving) the foraging efficiency of adult birds and hence nest
provisioning rates. This is the most frequently invoked mechanism explaining the
foraging benefits associated to roadside habitats: vehicle-killed animals including
arthropods facilitate foraging to a number of avian species, mostly corvids (e.g. Morgan
157
et al., 2010; Yamac and Kirazli, 2012) and the abundance of live prey such as
arthropods and reptiles can also increase along road edges (Dangerfield et al., 2003;
Andrews et al., 2006). Contrary to the above mentioned avian species, bee-eaters feed
almost exclusively on live flying insects captured on the air (Krebs and Avery, 1984;
1985; Kossenko and Fry, 1998). Although a traffic-mediated increase in the abundance
or detectability of flying insects was not visually evident to us in Doñana, closely
related species such as the northern carmine bee-eater Merops nubicus have been
reported to feed on insects flushed by wild and domesticated mammals in Africa
(Grimwood, 1964; Boswall, 1970; Cunningham Van Someren, 1970), and to forage in
association with large birds, man and motor vehicles (Jackson, 1945). Similarly,
southern carmine bee-eaters Merops nubicoides have been observed feeding on insects
flushed by mammals and vehicles in the Kruger National Park (Dean and MacDonald,
1981).
A second hypothesis potentially explaining the reported pattern is that bee-eaters
would be attracted to traffic because this stimulus is an effective hunting clue in
locations other than our study site (unmatched cue hypothesis). The lack of asphalt
pavements in the roads of Doñana generates conspicuous clouds of dust behind passing
vehicles, and birds could have learned to use this cue for detecting foraging patches in
their African winter areas (i.e. herds of African ungulates or running cars where flushed
insects may abound; see references above). The learned behavior would lead to a trafficrelated increase of hunting effort in Doñana regardless of any real effects on local prey
availability.
158
Finally, a third potential hypothesis is that passing vehicles interrupt the normal
progression of adults’ daily activities (e.g. resting, preening, brooding, hunting for selfmaintenance) and disturbed birds redirect their behavior towards nestling feeding
(interference-displacement hypothesis). Here motor traffic would increase feeding rates
only secondarily, through displacement behaviors triggered by the perturbation. For
example, adults hunting for self-maintenance could deliver prey to nestlings just
because the alarm calls of conspecific (elicited by vehicle passage) triggered their return
to the colony. In the same line of arguments, adults resting, preening or brooding at the
colony would take-off in response to vehicle passage. Once on the air, a hunting mode
could be naturally activated in this primarily aerial predator. In the latter two examples
traffic can indirectly result in increased feeding rates despite the primary responses were
predator-avoidance behaviors.
Our results provide evidence that motor traffic in Doñana Biosphere Reserve
exerts relevant effects on the behavior of roadside breeding bee-eaters, but its net
consequences on individual health and performance remain unclear. On the one hand,
repeated exposure to traffic could bring fitness costs. Extrapolating the reported
association between vehicle passage and predator-avoidance responses, bee-eaters
breeding beside the studied high-traffic areas experienced 379 flock flushing events and
235 alarm call episodes during one single week, compared to 14 flock flushing events
and nine alarm calls in low-traffic colonies. In addition to the direct energy demands
associated to the alarm responses themselves (e.g. Bergstrom and Lachmann, 2001),
disturbance stimuli can indirectly affect individual health and fitness through trade-offs
159
between perceived risk and energy intake or lost opportunity costs. For example,
passerine birds may decrease singing activity in response to human disturbance, tradingoff predation-avoidance behaviors with the ability to defend territories and attract mates
(Gutzwiller et al., 1994; Slabbekoorn and Peet, 2003). Similarly, lekking species such as
the great snipe Gallinago media respond to human disturbance fleeing away from the
display grounds, trading-off access to mates (Kålås et al., 1995). Vehicle exposure and
road proximity has been shown to elevate physiological stress in birds (i.e.
corticosterone levels, Sapolsky et al., 2000) including spotted owls Strix occidentalis
(Hayward and Wasser, 2006), adult and nestling white-crowned sparrows Zonitrichia
leucophrys oriantha (Crino et al., 2006) and female American kestrels Falco sparverius
(Strasser and Heathn, 2013). We may thus expect increased mating efforts and higher
levels of corticosterone in bee-eaters breeding beside high-traffic roads. In the shortterm, corticosterone elevations may allow traffic-exposed bee-eaters meeting the
increased energy demands of stress-copying mechanisms (such as the recorded
predation-avoidance behaviors) through its effects on gluconeogenesis, mobilization of
body energy stores and locomotion (Sapolsky et al., 2000). However, in the long-term
sustained corticosterone elevations are known to exert deleterious effects on health,
reproduction and survival (Wingfield and Sapolski, 2003) and we may predict reduced
fat depots and body condition with potential cascading effects on reproductive function
(Miller et al., 2009; Strasser and Heath, 2013).
On the other hand, the recorded positive effects of traffic on parental feeding
rates could bring fitness benefits to bee-eaters. Parental feeding rates were 2.75 times
higher in traffic exposed nests compared to control areas, and this might result in overall
160
higher offspring quality (e.g. improved body condition and growth rate; Ardia, 2007)
and quantity (Clutton-Brock, 1991). However, the consequences for adults’ health and
condition would be very different depending on the preponderant mechanism mediating
the traffic-related increase in feeding rates. If the facilitation hypothesis underlies our
results, adult bee-eaters would optimize hunting efficiency without necessarily
increasing hunting effort, and higher reproductive outcome could be achieved without
compromising adult health and condition. This mechanism may explain why breeding
success in magpies Pica pica is significantly higher near highways (Yamac and Kirazli,
2012). The net fitness outcome for adult birds should nonetheless depend on the relative
contribution of the costs associated to predator-avoidance behaviors (e.g. increased
stress levels, decreased body condition, higher mortality risk) and the benefits
associated to increased offspring quality or quantity. Alternatively, if the unmatched cue
or the interference-displacement hypotheses underlie our results, traffic-related
elevations in feeding rates would be a consequence of increased feeding effort at a time
when prey is neither more abundant nor more detectable. Energy expenditure has been
shown to increase exponentially with food provisioning rates in closely related bee-eater
species (i.e. blue-throated bee-eaters Merops viridis; Bryant, 1988). Under constant
conditions of food availability, increased feeding rates could lead to a loss in parental
mass, increasing the chances of starvation or the susceptibility to parasites and disease.
The expected reproductive benefits would then occur at the expense of compromising
adult health and condition in birds already suffering the costs of traffic. Disentangling
which of the three hypotheses above explain the reported traffic-related increase in
feeding rates will have important consequences for management decisions, as it will
161
allow determining if the study roadside environments are ecological traps or not
(Kuitunen et al., 2003; Orlowski, 2008). Whether the final outcome of road traffic is a
positive, negative or neutral effect on bee-eaters’ net fitness is therefore difficult to
predict, and should be the focus of future ad-hoc research.
Our results should be interpreted in context with the actual traffic pressure
exerted by management and research activities within a biosphere reserve. This pressure
was moderate even in the high-traffic colonies, with a maximum recorded count of 91
vehicles/day during our survey, while in contrast the adjacent public road A483
intersecting with our high-traffic study road (see Figure 1) holds an average traffic
above 6500 vehicles/day (Revilla et al., 2010). Bee-eaters clearly avoid this public road
and nearby areas (with the closest nests in the intersecting study road being located 380
m away) likely due to the elevated transit, high vehicle speed and asphalt pavement.
Although we are not aware of the species’ thresholds of traffic tolerance, bee-eaters
modified their behavior in response to management- and research-related traffic,
showing lack of habituation and experiencing mortality through vehicle collisions.
Road kills were never detected in our low-traffic colonies, but 3-7 deaths are
normally reported on a yearly basis along the studied high-traffic road leading to the
Palacio de Doñana (authors pers. obs.). The real impact of management- and researchrelated road traffic on bee-eater mortality should be higher than currently appreciated
due to the rich community of scavengers and predators (e.g. wild boars, carnivores,
corvids, raptors; Sebastián-González et al., 2013) likely removing injured and dead
roadside birds (Santos et al., 2011; Teixeira et al., 2013). Bee-eaters probably collide
162
with vehicles as a consequence of their risk-avoidance flushing responses. This risk
seems more evident during the mating and nest-digging period when flock size is
typically larger, birds are distracted by the intense socio-sexual activities of the colonies
and flushing distances decrease (Lessells et al., 1994; Petrescu and Adam, 2001) forcing
drivers to reduce vehicle speed below the current limit of park regulations (i.e. 40
Km/h) in order to avoid collisions (authors pers. obs.).
Until the net consequences of traffic on individual health, stress levels and
fitness are determined, managers are encouraged to maintain the present traffic levels
(i.e. below 100 vehicles/day on roads presenting bee-eater colonies). Furthermore,
lowering speed limits below 40 Km/h beside colony locations could minimize trafficrelated mortality even if this measure is applied only a few weeks per year.
Our results illustrate how an avian population can show contrasting behavioral
responses to the same type of disturbance occurring at different times of its life cycle.
Because the expected outcome of traffic on individual performance can be opposed if
the responses are only monitored during either mating or brooding, we recommend
caution before inferring net fitness consequences of traffic from isolated behavioral
responses or specific life history sub-stages.
Acknowledgements
We thank the staff of the Doñana Biological Reserve and especially Fernando
Ibañez for their invaluable logistic support. JB and MC were supported by the Ramón y
Cajal Program of the CSIC and the Spanish Ministry of Economy and Competitiveness,
163
and MD by a JAE-PRE fellowship. Financial support was provided by Research project
CGL2012-32544 (Spanish Ministry of Economy and Competitiveness and Feder funds),
and grant 511/2012 (Organismo Autónomo de Parques Nacionales; Spanish Ministry of
Agriculture, Food and the Environment).
References
Alvarez, F. and Hiraldo, F. (1974). Estructura de las galerías de nidificación del abejaruco (Merops
apiaster) en Doñana [Bee-eater (Merops apiaster) nesting tunnel structure in Doñana]. Doñana Acta
Vertebrata. 1: 61-67.
Ardia, D.R. (2007). Site- and sex-level differences in adult feeding behaviour and its consequences to
offspring quality in tree swallows (Tachycineta bicolor) following brood-size manipulation. Canadian
Journal of Zoology. 85: 847-854.
Baudains, T.P. and Lloyd, P. (2007). Habituation and habitat changes can moderate the impacts of human
disturbance on shorebird breeding performance. Animal Conservation. 10: 400-407.
Bautista, L.M., Garcia, J.T., Calmaestra, R.G., Palacín, C., Martín, C.A., Morales, M.B., Bolnal, R. and
Viñuela, J. (2004). Effect of weekend road traffic on the use of space by raptors. Conservation
Biology. 18: 726-732.
Bee, M.A. and Swanson, E.M. (2007). Auditory masking of anuran advertisement calls by road traffic
noise. Animal Behaviour. 74: 1765-1776.
Bennett, V.J., Sparks, D.W. and Zollner, P.A. (2013). Modeling the indirect effects of road networks on
the foraging activities of bats. Landscape Ecology. 28: 979-991.
Bergstrom, C.T. and Lachmann, M. (2001). Alarm calls as costly signals of antipredator vigilance: the
watchful babbler game. Animal Behavior. 61: 535-543.
Blasco, M., Domeño, C. and Nerín, C. (2008). Lichens biomonitoring as feasible methodology to assess
air pollution in natural ecosystems: combined study of quantitative PAHs analyses and lichen
biodiversity in the Pyrenees Mountains. Analytical and Bioanalytical Chemistry. 391: 759-771.
Boswall, S. (1970). The association of the northern carmine bee-eater Merops n. nubicus with mammals,
birds and motor vehicles in Ethiopia. Bulletin of the British Ornithology Club. 90: 92-96.
Bouskila, A. and Blumstein, D.T. (1992). Rules of thumb for predation hazard assessment: predictions
from a dynamic model. American Naturalist. 139: 161-176.
Boyle, S.A. and Samson, F.B. (1985). Effects of nonconsumptive recreation on wildlife: a review.
Wildlife Society Bulletin. 13: 110-116.
Bryant, D.M. (1988). Energy expenditure and body mass changes as measures of reproductive costs in
birds. Functional Ecology. 2: 23-34.
Byrnes, P., Goosem, M. and Turton, S.M. (2012). Are less vocal rainforest mammals susceptible to
impacts from traffic noise? Wildlife Research. 39: 355-365.
Casas-Criville, A. and Valera, F. (2005). The European bee-eater (Merops apiaster) as an ecosystem
engineer in arid environments. Journal of Arid Environments. 60: 227-238.
vertebrate fauna road-kill aggregations. Biological Conservation. 109: 15-26.
Clutton-Brock, T.H. (1991). The evolution of parental care. Princeton University Press, 353 pp.,
Princeton (USA).
164
Coelho, H.P. and Kindel, A. (2008). Roadkills of vertebrate species on two highways through the Atlantic
Forest Biosphere Reserve, southern Brazil. European Journal of Wildlife Research. 54: 689-699.
Cramp, S. (1985). Handbook of the birds of Europe, the Middle East and North Africa, Vol. IV. Oxford
University Press, 960 pp., Oxford (UK).
Crino, O.L., Van Oorschot B.K., Johnson, E.E., Malisch, J.L. and Breuner, C.W. (2011). Proximity to a
high traffic road: glucocorticoid and life history consequences for nestling white-crowned sparrows.
General and Comparative Endocrinology. 173: 323-332.
Cunningham Van Someren, G.R. (1970). Animated perches and feeding associations of birds in the
Sudan. Bulletin of the British Ornithology Club. 90: 120-122.
Dalal, H., Gabriel, B-D., Wilson, D.R. and Mennill, D.J. (2011). Anthropogenic noise affects song
structure in red-winged blackbirds (Agelaius phoeniceus). Journal of Experimental Biology. 214:
3549-3556.
D'Amico, M., Rouco, C., Russell, J.C., Román, J. and Revilla, E. (2013). Invaders on the road:
synanthropic bird foraging along highways. Oecologia Australis. 17: 86-95.
Dangerfield, J.M., Pik, A.J., Britton, D., Holmes, A., Gillings, M., Oliver, I., Briscoe, D. and Beattie, A.J.
(2003). Patterns of invertebrate biodiversity across a natural edge. Austral Ecology. 28: 277-236.
Dean, W.R.J and MacDonald, I.A.W. (1981). A review of African birds feeding in association with
mammals. Ostrich. 52: 135-155.
Fahrig, L. and Rytwinski, T. (2009). Effects of roads on animal abundance: an empirical review and
synthesis. Ecology and Society. 14: 21 [online].
Fernández-Juricic, E., Sallent, A., Sanz, R. and RodrÍguez-Prieto, I. (2003). Testing the risk-disturbance
hypothesis in a fragmented landscape: nonlinear responses of house sparrows to humans. The Condor.
105: 316-326.
Ferreras, P., Aldama, J.J., Beltrán, J.F. and Delibes, M. (1992). Rates and causes of mortality in a
fragmented population of Iberian lynx Felis pardina Temminck, 1824. Biological Conservation. 61:
197-202.
Findlay, C.S.T. and Bourdages, J. (2000). Response time of wetland biodiversity to road construction on
adjacent lands. Conservation Biology. 14: 86-94.
Forman, R.T.T., Sperling, D., Bissonette, J.A. , Clevenger, A.P., Cutshall, C.D., Dale, V.H., Fahrig, L.,
(2003). Road Ecology: science and solutions. Island Press, 481 pp., Washington (USA).
Frid, A. and Dill, L. (2002). Human-caused disturbance stimuli as a form of predation risk. Conservation
Ecology. 6: 11 [online].
García-Novo, F. and Marín, C. (2006). Doñana: water and biosphere. UNESCO-sponsored programmes
and publications. Ministerio de Medio Ambiente, Confederación Hidrográfica del Guadalquivir, 365
pp., Seville (Spain).
Gibbs, J.P. and Shriver, W.G. (2002). Estimating the effects of road mortality on turtle populations.
Gill, J.A., Sutherland, W.J. and Watkinson, A.R. (1996). A method to quantify the effects of human
disturbance on animal populations. Journal of Applied Ecology. 33: 786-792.
Grimwood, I. (1964). The northern carmine bee-eater Merops nubicus Gmelin using animal perches.
Bulletin of the British Ornithology Club. 84: 105-106.
Gutzwiller, K.J., Wiedemann, R.T., Clements, K.L. and Anderson, S.H. (1994). Effects of human
intrusion on song occurrence and singing consistency in subalpine birds. The Auk. 111: 28-37.
Hayward, L.S. and Wasser, S.K. (2006). Physiological measures of disturbance related to vehicle
exposure in the northern spotted owl. Integrative and Comparative Biology. 46: E57.
Inglisa, M. and Galeotti, P. (1993). Daily activity at nests of the European bee-eaters (Merops apiaster).
Ethology Ecology & Evolution. 5: 107-114.
Jackson, T.H.E. (1945). Some Merops-Ardeotis perching associations in northern Kenya. Ibis. 87: 284286.
Jilka, A. and Ursprung, J. (1980). Zur stimme des bienenfressers (Merops apiaster) und ihrer rolle im
sozialen verkehr der artgenossen. Egretta, 23, 8-19.
Jones, C.S., Lessells, C.M. and Krebs, J.R. (1991). Helpers-at-the-nest in European bee-eaters (Merops
apiaster): a genetic analysis. In: DNA fingerprinting: approaches and applications. Burke, T., Dolf, G.,
Jeffreys, A.J. and Wolff, R. (Eds.). Birkhäuser Basel (Switzerland), pp. 169-192.
165
Kålås, J.A., Fiske, P. and Sæther, S.A. (1995). The effect of mating probability on risk taking: an
experimental study in lekking great snipe. American Naturalist. 146: 59-71.
Kossenko, S.M. and Fry, C.H. (1998). Competition and coexistence of the European bee-eater Merops
apiaster and the blue-cheeked bee-eater Merops persicus in Asia. Ibis. 140: 2-13.
Krebs, J.R. and Avery, M.I. (1984). Chick growth and prey quality in the European bee-eater (Merops
apiaster). Oecologia. 64: 363-368.
Krebs, J.R. and Avery, M.I. (1985). Central place foraging in the European bee-eater, Merops apiaster.
Journal of Animal Ecology. 54: 459-472.
Kuitunen, M.T., Viljanen, J., Rossi, E. and Stenroos, A. (2003). Impact of busy roads on breeding success
in pied flycatchers Ficedula hypoleuca. Environmental Management. 31: 79-85.
Lessells, K. and Avery, M.I. (1989). Hatching asynchrony in European bee-eaters Merops apiaster.
Journal of Animal Ecology. 58: 815-835.
Lessells, C.M., Avery, M.I. and Krebs, J.R. (1994). Nonrandom dispersal of kin: why do European beeeater (Merops apiaster) brothers nest close together? Behavioral Ecology. 5: 105-113.
Lima, S.L., Blackwell, B.F., DeVault, T.L. and Fernández-Juricic, E. (2014). Animal reactions to
oncoming vehicles: a conceptual review. Biological Reviews.
Littell, R.C., Milliken, G.A., Stroup, W.W. and Wolfinger, R.D. (1996). SAS System for Mixed Models,
SAS Institute, 840 pp., Cary (USA).
McCullagh, P. and Nelder, J.A. (1989). Generalized Linear Models. Chapman and Hall, 511 pp., New
York (USA).
McGowan, C.P. and Simons, T.R. (2006). Effects of human recreation on the incubation behavior of
American oystercatchers. Wilson Journal of Ornitholology. 118: 485-493.
McLellan, B.N. and Shackleton, D.M. (1988). Grizzly bears and resource-extraction industries: effects of
roads on behaviour, habitat use and demography. Journal of Applied Ecology. 25: 451-460.
Miller, D.A., Vleck, C.M. and Otis, D.L. (2009). Individual variation in baseline and stress-induced
corticosterone and prolactin levels predicts parental effort by nesting mourning doves. Hormones and
Behavior. 56: 457-464.
Morgan, G.M., Boughton, R.K., Rensel, M.A. and Schoech, S.J. (2010). Road effects on food availability
and energetic intake in Florida scrub-jays (Aphelocoma coerulescens). The Auk. 127: 581-589.
Mortensen, D.A., Rauschert, E.S.J., Nord, A.N. and Jones, B.P. (2009). Forest roads facilitate the spread
of invasive plants. Invasive Plant Science and Management. 2: 191-199.
Orlowski, G. (2008). Roadside hedgerows and trees as factors increasing road mortality of birds:
implications for management of roadside vegetation in rural landscapes. Landscape and Urban
Planning. 86: 153-161.
Peris, S.J. and Pescador, M. (2004). Effects of traffic noise on passerine populations in Mediterranean
wooded pastures. Applied Acoustics. 65: 357-366.
Petrescu, A. and Adam, C. (2001). Interspecific relations in the populations of Merops apiaster L. (Aves:
Coraciiformes) of southern Romania. Travaux du Museum National d’Histoire Naturelle Grigore
Antipa. 43: 305-322.
Ramp, D., Wilson, V.K. and Croft, D.B. (2006). Assessing the impacts of roads in peri-urban reserves:
road-based fatalities and road usage by wildlife in the Royal National Park, New South Wales,
Australia. Biological Conservation. 129: 348-359.
Rees, E.C., Bruce, J.H., and White, G.T. (2005). Factors affecting the behavioural responses of whooper
swans (Cygnus c. cygnus) to various human activities. Biological Conservation. 121: 369-382.
Reid, J.C. (1974). Bienenfresser-Beobachtungen im ostlichen Osterreich. Egretta. 17: 15-22.
Revilla, E., Barón, A. and Román, J. (2010). Número de vehículos que circulan por Doñana: intensidad
de tráfico [The number of vehicles circulating in Doñana: traffic intensity]. In: Evaluación de los
efectos del tránsito a motor sobre especies y comunidades de interés en el Espacio Natural de
Doñana [Evaluation of the effects of motorized traffic on species and communities of interest within
Doñana Natural Area]. Román, J., Barón, A. and Revilla, E. (Eds.). Consejería de Medio Ambiente,
Junta de Andalucía. Estación Biológica de Doñana CSIC, Sevilla (España), pp. 41-67.
[online].
166
Sapolsky, R.M., Romero, L.M. and Munck, A.U. (2000). How do glucocorticoids influence stress
responses? Integrating permissive, suppressive, stimulatory, and preparative actions. Endocrine
Reviews. 21: 55-89.
Schroeder, J., Nakagawa, S., Cleasby, I.R. and Burke, T. (2012). Passerine birds breeding under chronic
noise experience reduced fitness. PLoS ONE. 7: e39200 [online].
Slabbekoorn, H. and Peet, M. (2003). Birds sing at a higher pitch in urban noise. Nature. 424: 267.
Sebastián-González, E., Sánchez-Zapata, J.A., Donázar, J.A., Selva, N., Cortés-Avizanda, A., Hiraldo, F.,
Blázquez, M., Botella, F. and Moleón, M. (2013). Interactive effects of obligate scavengers and
scavenger community richness on lagomorph carcass consumption patterns. Ibis. 155: 881-885.
Smalley, I., O'Hara-Dhand, K., McLaren, S., Svircev, Z. and Nugent, H. (2013). Loess and bee-eaters I:
ground properties affecting the nesting of European bee-eaters (Merops apiaster L.1758) in loess
deposits. Quaternary International. 296: 220-226.
Sorice, M.G., Shafer, C.S. and Ditton, R.B. (2006). Managing endangered species within the usepreservation paradox: the Florida manatee (Trichechus manatus latirostris) as a tourism attraction.
Stankowich, T. and Blumstein, D.T. (2005). Fear in animals: a meta-analysis and review of risk
assessment. Proceedings of the Royal Society B: Biological Sciences. 272: 2627-2634.
Strasser, E.H. and Heath, J.A. (2013). Reproductive failure of a human-tolerant species, the American
kestrel, is associated with stress and human disturbance. Journal of Applied Ecology. 50: 912-919.
Teixeira, F.Z., Coelho, A.V.P., Esperandio, I.B. and Kindel, A. (2013). Vertebrate road mortality
estimates: effects of sampling methods and carcass removal. Biological Conservation. 157: 317-323.
Valera F. (2009). Abejaruco europeo [European bee-eater] - Merops apiaster Linnaeus, 1758:
http://www.vertebradosibericos.org/aves/merapi.html [online]. In: Enciclopedia Virtual de los
Vertebrados Españoles [Virtual Encyclopedia of the Vertebrates of Spain]. Salvador, A. and Bautista,
L.M. (Eds.). Sociedad de Amigos del MNCN (SAM) y Museo Nacional de Ciencias Naturales
(MNCN - CSIC). Madrid (Spain).
Valverde, J.A. (1958). An ecological sketch of the Coto Doñana. British Birds. 4: 1-23.
Velando, A . and Munilla, I. (2011). Disturbance to a foraging seabird by sea-based tourism: implications
for reserve management in marine protected areas. Biological Conservation. 144: 1167-1174.
Vos, C.C. and Chardon, J.P. (1998). Effects of habitat fragmentation and road density on the distribution
pattern of the moor frog Rana arvalis. Journal of Applied Ecology. 35: 44-56.
Williams, R., Lusseau, D. and Hammond, P.S. (2006). Estimating relative energetic costs of human
disturbance to killer whales (Orcinus orca). Biological Conservation. 133: 301-311.
Wingfield, J.C. and Sapolsky, R.M. (2003). Reproduction and resistance to stress: when and how?
Journal of Neuroendocrinology. 15: 711-724.
Yamac, E. and Kirazli, C. (2012) Road effect on the breeding success and nest characteristics of the
Eurasian magpie (Pica pica). Ekoloji. 21: 1-10.
Zuur, A.F., Leno, E.N., Walker, N.J., Saveliev, A.A. and Smith, G.M. (2007). Mixed effects models and
extensions in ecology with R. Springer, 523 pp., New York (USA).
167
Capítulo 5:
Invasiones Biológicas
Invasores en ruta:
aves sinantrópicas alimentándose en las carreteras.
______________________________________________________________________
D'Amico, M. a, Rouco, C. b, Russell, J.C. c, Román, J.
a
and Revilla, E.
a
(2013).
Invaders on the road: synanthropic bird foraging along highways. Oecologia Australis.
17(1 - Road Ecology Special Issue), 86-95.
a
Doñana Biological Station (EBD-CSIC), Seville (Spain).
b
Manaaki Whenua - Landcare Research, Alexandra (New Zealand).
c
University of Auckland, Auckland (New Zealand).
Invaders on the road: synanthropic bird foraging along highways
Resumen: Las carreteras intermunicipales ofrecen potencialmente alta
disponibilidad de alimentos de fácil acceso y pueden representar un área preferente para
la alimentación de aves sinantrópicas. Como consecuencia, una red de carreteras podría
actuar como corredor de invasión de aves urbanas no nativas. Este estudio es una
primera aproximación para verificar tal patrón de dispersión, determinando si la
búsqueda de alimento es realmente la principal actividad desarrollada a lo largo de
carreteras intermunicipales. Con este objetivo muestreamos 200 puntos de la red de vías
intermunicipales de Nueva Zelanda, realizando observaciones sobre cuatro especies de
aves sinantrópicas invasoras (gorrión Passer domesticus, mirlo Turdus merula y
estornino Sturnus vulgaris, todos euroasiáticos; y miná Acridotheres tristis, de origen
indio).
Para
cada
especie
también
investigamos:
preferencias
alimenticias,
características de carretera que puedan afectar la búsqueda de alimento, y selección de
micro-hábitat de alimentación. Los comportamientos observados predominantemente
estuvieron relacionados con actividad de alimentación, con diferencias en estrategia de
cada especie. Los gorriones fueron observados alimentándose principalmente de
semillas e invertebrados, mientras estorninos y minás mostraron comportamientos
alimenticios más generalistas. Estas tres especies fueron observadas principalmente a lo
largo del borde de carreteras, mientras que los mirlos fueron detectados principalmente
depredando invertebrados en prados de las cunetas. De forma semejante a las áreas
urbanas, las carreteras intermunicipales proveen alimentos de fácil acceso, disponibles
para las especies de estudio. Proponemos la existencia de un patrón de invasión por aves
sinantrópicas mediado por carreteras, e incentivamos futuros estudios para optimizar
estrategias de manejo, focalizado hacia carreteras intermunicipales como áreas de
control más adecuadas para evitar la propagación de esas especies.
169
Capítulo 5: Invasiones Biológicas
Abstract: Intercity roads potentially offer a high availability of easily accessible
food and could therefore be expected to represent a preferential area for synanthropic
bird foraging. As a consequence, a road-network could act as an invasion corridor for
non-native urban birds. This study is a first approximation to verify such pattern of
spread, by determining if foraging is actually the main activity carried out along
intercity roads. With this aim, during the spring of 2011 we surveyed 200 points along
the New Zealand intercity road-network, carrying out observations on four invasive
synanthropic bird species (Eurasian house sparrow Passer domesticus, Eurasian
blackbird Turdus merula, Eurasian starling Sturnus vulgaris and Indian myna
Acridotheres tristis). For every species we also investigated: food preferences, the
possibility that certain road features could affect the occurrence of feeding activity
along roads, and foraging micro-habitat selection along roads. Predominant observed
behaviors related to feeding activity, with differences concerning species-specific
foraging strategies. Sparrows were mainly observed feeding on small-sized items (such
as seeds and invertebrates), while both starlings and mynas showed more generalist
feeding habits. These three species were mostly observed along the side of the lane, in
contrast with blackbirds, mainly detected preying on invertebrates along the mown
grassy verges of the road. Similarly to urban areas, intercity roads provide easily
accessible food, available for all the study species. We propose a road-mediated
invasion pattern for synanthropic birds, and encourage future studies to optimize
management strategies, and targeting intercity roads as the most suitable control areas to
prevent spread.
170
Introduction
Human dominated landscapes are increasing worldwide, with associated
potential threats to biodiversity and ecosystems (Vitousek et al., 1997; Hooper et al.,
2012). Paradoxically, there are species taking advantage of global change, shifting their
trophic and spatial preferences to the novel niches of recently degraded habitats
(Rebele, 1994; Marzluff, 2001). In the case of urban and rural areas, these species are
usually defined as synanthropic (di Castri et al., 1990), including human commensal
birds (Luniak et al., 2004; Bonier et al., 2007). Urbanized environments are also
selected by synanthropic birds because they have lower predator diversity and milder
winter temperatures compared with natural habitats (Jokimäki et al., 1996; Gering and
Blair, 1999; McKinney, 2006). Synanthropic bird species have been introduced
worldwide for food, pest control and nostalgic purposes, and due to their adaptation to
coexist with humans they have proliferated and also spread (Marzluff, 2001; McKinney,
2006).
Intercity roads usually cross rural and natural habitats, and probably show to a
lesser extent the above mentioned characteristics of urbanized areas, including the
availability of easily accessible feeding resources. Food availability is one of the main
drivers of dispersal (Bowler and Benton, 2005), and for urban birds it should be easier
to forage along intercity roads than in certain natural habitats. As a consequence, urban
bird dispersal routes might not be randomly selected, and these species could
preferentially move near this kind of linear infrastructure. In this case, road-networks
might therefore act as invasion corridors for synanthropic birds. A first approximation
171
to verify this spreading pattern would be an observational study describing the
synanthropic bird behaviors along intercity roads, trying to confirm the relative
significance of feeding activities. A positive association between foraging and roads
would encourage a simplification of current management strategies for these introduced
species, by determining the suitable target control areas for carrying out actions focused
on their containment.
New Zealand is an insular country with a distinctive biota mainly characterized
by its endemic bird communities (Cooper and Millener 1993; Daugherty et al., 1993).
The principal threats to New Zealand biodiversity are biological invasions (Clout, 2001;
Norton, 2009), including the introduction of synanthropic birds mostly carried out since
1860 by Acclimatization Societies (Druett, 1983; McDowall, 1994). Urban birds
compete with native species and prey on endemic fauna (Meads et al., 1984; Greene and
Jones, 2003). Moreover, they contribute to the spread of infectious diseases and
invasive plants (Williams and Karl, 1996; Tompkins and Gleeson, 2006), and they also
impact several human activities such as orchard cultivation (Dawson and Bull, 1970,
Kross et al., 2012). Introduced synanthropic birds represent the most abundant and
widespread avian species in New Zealand, particularly the Eurasian house sparrow
Passer domesticus, the Eurasian blackbird Turdus merula, the Eurasian starling Sturnus
vulgaris and the Indian myna Acridotheres tristis (Heather and Robertson, 2005). They
are regionally subjected to lethal control methods including shooting, trapping and
baiting by avicides, principally alphachloralose and DRC-1339 (Nelson, 1990; 1994).
All these methods and especially the use of poison also entail mortality risk for nontarget bird species (Nelson 1990, 1994), but the society generally supports control
172
programs (Green and Rohan, 2012). However, their planning nonetheless requires
prudence.
The main purpose of the present study is to determine the principal activity
carried out by these four synanthropic bird species along the intercity roads of New
Zealand, where they are abundant. Our hypothesis advances that urban birds use such
areas mostly to forage, with implications for their establishment and spread patterns and
consequent management strategies. Moreover, with the aim of further improving the
control and containment of these introduced birds, for every species we also
investigated: food preferences, the possibility that certain road features could affect the
occurrence of feeding activity along roads, and foraging micro-habitat selection along
roads. All that considered, this species-specific observational study about the feeding
habits of non-native synanthropic birds is of high interest to understand their activity
along intercity roads, with several possible implications for the improved planning of
their management.
Study area and Data collection
Our study was carried out along several New Zealand intercity roads, across
both the North and the South Islands (Figure 1), during three weeks of the Austral
spring 2011. We selected 200 sample points (50 for every surveyed species: sparrows,
blackbirds, starlings and mynas) depending on features characterizing road and
surrounding environment: traffic intensity (high or low), habitat (open or with
173
shrubs/trees) and presence/absence of mown grassy verges. Each variable level was
equally sampled for all the four species. Every sample point was at least five kilometers
from another to avoid resurveying the same individual. At each sample point the
observer waited inside a car for the arrival of an individual of our study species. The
individual detection was the beginning of a binocular focal observation of five minutes,
recording the main activity carried out. In the case of foraging behaviors, we also
registered the most frequent feeding choice and the section of the road in which the
action was mostly carried out.
Data analysis
To determine the principal activity carried out by introduced synanthropic birds
along New Zealand intercity roads we carried out a first Pearson χ2 test (goodness-offit; SAS Institute Inc., 2008) for each study species. In such test we analyzed if the
observed frequency of every detected behavior significantly diverged from the
distribution of behaviors if they were random.
Furthermore, selecting only the observations of foraging activities, with a second
Pearson χ2 test (goodness-of-fit) we also determined the food resources most often
consumed by every study species along roads.
With the aim to analyze if certain roadway and surrounding environment
features were affecting the appearance of feeding activity along intercity roads, we
carried out a logistic generalized linear model (GLM: McCullagh and Nelder, 1989) for
each study species. The response variable was the presence/absence of foraging
behaviors as the main detected activity of the observation, while the explanatory
174
categorical variables were: traffic volume (high or low), habitat (open or with shrubs/
trees) and presence/absence of mown grassy verges. In all cases, we applied a binomial
error distribution and logit link function.
Finally, the fourth analyses also were Pearson χ2 tests (goodness-of-fit), in this
case with the aim to determine in which sections of the road (middle, shoulders and
verges) the species carried out the main part of their foraging activities. Also for these
analyses, we performed a test for each study species.
175
Figure 1: Sample point distribution for each study species. 1) sparrows, 2) blackbirds, 3) starlings and
4) mynas.
176
Results
We mostly detected four types of activities along the New Zealand intercity
roads: foraging, searching, perching and interacting. The latter included attacks and
fighting, courtship, copulations, and also parental care of fledglings. Other activities
were also detected (passing, alert, escaping, preening, defecating), but none of them
reached the status of predominant behavior during an observation.
Sparrows were observed carrying out all the four main activities, especially
foraging (76% of the observations: Figure 2). The first Pearson χ2 test indicated that
such behavior was more frequent (p < 0.0001) than searching, perching and interacting
(respectively 8%, 10% and 6% of the observations). We did not detect interactions
among blackbirds, and also in their case foraging was more frequent (74% of the
observations; p < 0.0001) than both searching and perching (16% and 10%). Starlings
carried out all the four main activities, with feeding being more frequent (80% of the
observations; p < 0.0001) than the other behaviors (searching 8%, perching 4% and
interacting 8%). The first Pearson χ2 test for mynas also revealed the same pattern
(foraging 74%, searching 4%, perching 12% and interacting 10%; p < 0.0001; Figure 2).
177
Figure 2: Non-native urban bird main activity for each of 50 observations. For all the four species,
foraging activity significantly represents the main observed behavior.
Concerning the main food resources consumed along intercity roads, we
detected six categories: garbage, identifiable carrions (mostly introduced mammals;
hedgehogs Erinaceus europaeus, possums Trichosurus vulpecula and rabbits
Oryctolagus cuniculus), flattened organic matter, invertebrates, seeds/fruits and
unrecognizable items.
We could not identify the most part of the items selected by sparrows (55% of
foraging observations; Table 1). The second Pearson χ2 test showed that foraging
activity on unrecognizable resources was more frequent (p = 0.0003) than on garbage,
flattened organic matter and seeds (respectively the 11%, 18% and 16%). For sparrows,
we did not detect scavenging behavior and identifiable predation on invertebrates.
Conversely, blackbirds mainly fed on invertebrates, and also on unrecognizable items
(respectively 35% and 46% of foraging observations). Flattened organic matter and
178
seeds/fruits were consumed less (8% and 11%; p = 0.0016), and foraging on garbage
and carrions was not detected. We could not categorize the most part of items selected
by starlings (50% of the feeding observations). Foraging activity on flattened organic
matter and invertebrates was less frequent (both of them respectively 20%; p < 0.0001),
followed by consumption of garbage and seeds/fruits (both of them respectively 5%).
Scavenging behavior was not detected for starlings. Finally, there were no significant
differences in the food resources consumed by mynas along intercity roads. The most
exploited item was flattened organic matter (30% of foraging observations), followed
by invertebrates (17%), garbage, carrions, seeds/fruits and unrecognizable items (each
of them respectively 14%; Table 1).
Table 1: Number of foraging observations for every main consumed food item. The meaning of the
acronym FOM is flattened organic matter, and for NII is not identified items. Total represents the overall
number of foraging observation per species. The only species without significant differences among
consumed items is the myna.
Garbage
Carrions
FOM
Invertebrates
Seeds/Fruits
NII
Total
Sparrows
4
0
7
0
6
21
38
Blackbirds
0
0
3
13
4
17
37
Starlings
2
0
8
8
2
20
40
Mynas
5
5
11
6
5
5
32
179
Regarding the GLMs carried out, the only candidate variable affecting the
appearance of synanthropic birds foraging along intercity roads was the presence of
mown grassy verges, increasing the occurrence of blackbirds (p = 0.007; Figure 3).
Figure 3: Probability to observe a blackbird foraging along the highway during an observation, in
relation to the presence/ absence of mown grassy verges.
Finally, about the road sections preferably used for feeding activities by
introduced synanthropic birds, the third Pearson χ2 test for sparrows revealed they
significantly selected the shoulders of the lanes (76% of foraging observations; p <
0.0001; Figure 4), compared with the middle of the roadway and its verges (respectively
11% and 13%). Conversely, blackbirds are the only species actively selecting the verges
(76%; p < 0.0001), followed by the sides (21%) and the middle of the road (3%). On the
other hand, similarly to sparrows, the starlings also preferred the shoulders (60%; p =
0.0005), and used the middle of the road and the verges less (10% and 30%). The mynas
180
mostly selected the external sides of the lanes (81%; p < 0.0001), compared with the
middle of the roadway or its verges (respectively 16% and 3% of foraging
observations).
Figure 4: Number of foraging observations for each species and area of the road (verges, shoulders
and the middle of the roadway).
181
Discussion
The present study represents a simple approximation to the use of behavioral
observations with the aim to understand habitat use and diet, and optimize control
strategies of invasive species. With a moderate effort we recorded that introduced
synanthropic birds are generally using intercity roads mainly for foraging purposes,
affirming the plausibility of a road-mediated invasion pattern for synanthropic birds.
This could have consequences on the current management strategies for these species.
Our results show that the four study species were detected carrying out feeding
activities during approximately 75% of the observations. This proportion might also
increase considering that two of the three other detected behaviors (both searching and
perching) are probably related to foraging in several cases. We did not categorize such
behaviors as feeding activities because searching can also concern grit and nest
materials, and perching may just correspond to a resting activity. Such prevalence of
foraging activities along intercity roads can be related to the easily accessible food
availability that we also detected during the fieldwork of the present study. This
availability represents a pronounced feature of urban areas as well, and probably one of
the main causes of synanthropization of several bird species (Marzluff, 2001; Davies et
al.; 2009, Evans et al.; 2009). Therefore, considering the importance of feeding
resources for dispersal (Bowler and Benton, 2005), intercity roads might represent a
preferential first step of invasion spread for urban birds, prior to the colonization of
natural habitats. The confirmation of the road-mediated invasion pattern that we suggest
could enable the optimization of introduced synanthropic bird containment strategies,
182
which might be mainly focused on urban areas and the surrounding intercity roads. This
management choice would have less probability to accidentally target native New
Zealand birds, which tend to avoid human dominated environments (van Heezik et al.,
2008). All that considered, we strongly recommend further research regarding our
suggestion of road-mediated invasion by non-native synanthropic bird species.
Concerning the main food items consumed along intercity roads, we can
differentiate them in two categories, directly and indirectly related with human activity.
The first category includes garbage and carrions resulting from road-kills and the
flattened organic matter consequent to the repeated transit of vehicles. The second
category consists of invertebrates, seeds/fruits and unrecognizable items. We could not
identify the latter items due to their small size, and therefore we suspect they should
also correspond with invertebrates and seeds. We consider that invertebrates and
seeds/fruits availability along verges were indirectly related with human activity
because these might respectively depend on roadway/roadside building materials and
verge vegetation management. Indeed, temperatures are usually higher along paved
roads than in natural surrounding environments, attracting arthropods and their
predators (Hannay, 2001; Jackson, 2003). Moreover, roadside vegetation management
can affect invertebrate availability for insectivorous birds, facilitating their foraging
activities in the presence of mown grassy verges (Romanowski and Zmihorski, 2008,
Devereux et al., 2006). Roadside hedgerows and plantations also determine seed and
fruit availability along the verges. All of our four study species fed along intercity roads
according to their previously known diet preferences (Moeed, 1976; 1980; MacMillan,
1981; Williams and Karl, 1996), but we also detected the consumption of food items
183
directly related with human activity. Mynas were observed feeding on road-killed
carrions, and all the species consumed garbage (excluding blackbirds) and flattened
organic matter. This exploitation of human-related food items suggests a plausible lack
of neophobia among urban birds, which could be capitalized by wildlife managers
selecting live-trapping as the main strategy to control such invasive species, also
eliminating mortality risk for autochthonous species. The previously mentioned baiting
with avicides (Nelson, 1990; 1994) could then be limited to particular situations, such as
intercity roads crossing degraded environments with consequent low densities of native
avifauna (van Heezik et al., 2008).
Regarding the influence of certain features of roadway and surrounding
environment on the presence of urban bird foraging along intercity roads, the only
significant variable was the presence of mown grassy verges increasing the occurrence
of blackbirds. As previously suggested, this kind of roadside management could
facilitate the feeding activities of insectivorous birds (Romanowski and Zmihorski,
2008, Devereux et al., 2006), and we largely detected blackbirds preying on
invertebrates (especially earthworms) during our observations. The mowing of grassy
verges could therefore be an effective strategy to attract blackbird individuals in order to
expose their populations to management, should it be required.
We also observed blackbird preference for verges through the analysis of
foraging micro-habitat selection, as a further confirmation of their relative dependence
on roadside food resources. Conversely, the other three study species significantly used
road shoulders to carry out their feeding activities, probably because there they can
184
access both the roadway and roadside food resources. Moreover, comparing mynas with
sparrows and starlings, we can appreciate how they additionally used the middle of the
roads too, confirming their preference for human-related food items. All that
considered, the use of the roadway shows us that road-kill mortality seems to be a
tolerable risk for these three species, denoting a boldness that can indicate again the
pertinence of trapping as a main control method.
The results of the present study imply relevant suggestions for the management
of these introduced species. We also reaffirmed the plausibility of a road-mediated
invasion pattern for synanthropic birds and we strongly suggest its confirmation through
future studies, due to its potential power of optimization for the containment of these
species, especially determining if urban areas and intercity roads are the most suitable
control target areas. We also encourage further and more exhaustive studies regarding
environmental variables affecting the presence of introduced synanthropic birds, with
the aim to improve management protocols and consequently to limit associated risks for
New Zealand native avifauna.
Acknowledgements
This research was supported by the JAE-PRE PhD stay fellowship granted to
MD by the CSIC. MD was also supported by a JAE-PRE fellowship from CSIC. Sofia
Conradi-Fernandez provided logistic support, Pablo Lucas helped with GIS, Zulima
Tablado, Simone Santoro, Manuela González-Suárez and two anonymous reviewers
provided useful comments.
185
References
Bonier, F., Martin, P.R. and Wingfield, J.C. (2007). Urban birds have broader environmental tolerance.
Biology Letters. 3: 670-673.
Bowler, D.E. and Benton, T.G. (2005). Causes and consequences of animal dispersal strategies: relating
individual behaviour to spatial dynamics. Biological Reviews. 80: 205-225.
Clout M. (2001). Where protection is not enough: active conservation in New Zealand. Trends in Ecology
& Evolution. 16: 415-416.
Cooper, R.A. and Millener, P.R. (1993). The New Zealand biota: historical background and new research.
Daugherty, C.H., Gibbs, G.W. and Hitchmough, R.A. (1993). Mega-island or micro-continent? New
Zealand and its fauna. Trends in Ecology & Evolution. 8: 437-442.
Davies, Z.G., Fuller, R.A., Loram, A., Irvine, K.N., Sims, V. and Gaston, K.J. (2009). A national scale
inventory of resource provision for biodiversity within domestic gardens. Biological Conservation.
142: 761-771.
Dawson, D.G. and Bull, P.C. (1970). A questionnaire survey of bird damage to fruit. New Zealand
Journal of Agricultural Research. 13: 362-371.
Devereux, C.L., Whittingham, M.J., Krebs, J.R., Fernández-Juricic, E. and Vickery, J.A. (2006). What
attracts birds to newly mown pasture? Decoupling the action of mowing from the provision of short
swards. Ibis. 148: 302-306.
di Castri, F., Hansen, A.J. and Debussche, M. (1990). Biological invasions in Europe and the
Mediterranean Basin. Monographiae Biologicae. 65: 1-463.
Druett, J. (1983). Exotic intruders: the introduction of plants and animals into New Zealand. Heinemann,
291 pp., Auckland (New Zealand).
Evans, K.L., Newson, S.E. and Gaston, K.J. (2009). Habitat influences on urban avian assemblages. Ibis.
151: 19-39.
Gering, J.C. and Blair, R.B. (1999). Predation on artificial bird nests along an urban gradient: predatory
risk or relaxation in urban environments? Ecography. 22: 532-541.
Green, W. and Rohan, M. (2012). Opposition to aerial 1080 poisoning for control of invasive mammals in
New Zealand: risk perceptions and agency responses. Journal of the Royal Society of New Zealand.
42: 185-213.
Greene, T.C. and Jones, A. (2003). Observed responses of captive stoats (Mustela erminea) to nest boxes
and metal collars used to protect kaka (Nestor meridionalis) nest cavities. New Zealand Journal of
Ecology. 27: 139-145.
Hannay, L. (2001). Effect of roads on arthropods. The Road-RIPorter. 6: 10-11.
Heather, B.D. and Robertson, H.A. (2005). The field guide to the birds of New Zealand. Penguin Books,
440 pp., North Shore (New Zealand).
Hooper, D.U., Adair, E.C., Cardinale, B.J., Byrnes, J.E.K., Hungate, B.A., Matulich, K.L., Gonzalez, A.,
Duffy, J.E., Gamfeldt, L. and O'Connor, M.I. (2012). A global synthesis reveals biodiversity loss as a
major driver of ecosystem change. Nature. 486: 105-108.
Jackson, H.D. (2003). Another reason for nightjars being attracted to roads at night. Ostrich. 74: 228-230.
Jokimäki, J., Suhonen, J., Inki, K. and Jokinen, S. (1996). Biogeographical comparison of winter bird
assemblages in urban environments in Finland. Journal of Biogeography. 23: 379-386.
Kross, S.M., Tylianakis, J.M. and Nelson, X.J. (2012). Effects of introducing threatened falcons into
vineyards on abundance of passeriformes and bird damage to grapes. Conservation Biology. 26: 142149.
Luniak, M. (2004). Synurbization - adaptation of animal wildlife to urban development. In: Proceedings
4th International Urban Wildlife Symposium (Shaw, W.W., Harris, L.K. and VanDruff, L., Eds.).
University of Arizona, Tucson (USA), pp. 50-55.
MacMillan, B.W.H. (1981). Food of house sparrows and greenfinches in a mixed farming district,
Hawke's Bay, New Zealand. New Zealand Journal of Zoology. 8: 93-104.
Marzluff, J.M. (2001). Worldwide urbanization and its effects on birds. In: Avian ecology and
conservation in an urbanizing world (Marzluff, J.M., Bowman, R. and Donnelly, R., Eds.). Norwell
(USA), pp. 19-47.
186
McCullagh, P. and Nelder, J.A. (1989). Generalized linear models. Chapman and Hall, 511 pp., London
(UK).
McDowall, R.M. (1994). Gamekeepers for the nation: the story of New Zealand's acclimatisation
societies, 1861-1990. Canterbury University Press, 508 pp., Christchurch (New Zealand).
McKinney, M.L. (2006). Urbanization as a major cause of biotic homogenization. Biological
Meads, M.J., Walker, K.J. and Elliott, G.P. (1984). Status, conservation, and management of the land
snails of the genus Powelliphanta (Mollusca: Pulmonata). New Zealand Journal of Ecology. 11: 277306.
Moeed, A. (1976). Foods of the common myna (Acridotheres tristis) in central India and in Hawke's Bay,
New Zealand. Notornis. 23: 246-249.
Moeed, A. (1980). Diets of adult and nestling starlings (Sturnus vulgaris) in Hawke's Bay, New Zealand.
New Zealand Journal of Zoology. 7: 247-256.
Nelson, P.C. (1990). Bird problems in New Zealand - Methods of control. In: Proceedings of the
Fourteenth Vertebrate Pest Conference (Davis, L.R. and Marsh, R.E., Eds.). Sacramento (USA), pp.
361-364.
Nelson, P.C. (1994). Bird control in New Zealand using alpha-chloralose and DRC1339. In: Proceedings
of the Sixteenth Vertebrate Pest Conference (Halverson, W.S. and Crabb, A.C., Eds.). Santa Clara
(USA), pp. 259-264.
Norton, D.A. (2009). Species invasions and the limits to restoration: learning from the New Zealand
experience. Science. 325: 569-571.
Rebele, F. (1994). Urban ecology and special features of urban ecosystems. Global Ecology and
Biogeography Letters. 4: 173-187.
Romanowski, J. and Zmihorski, M. (2008). Selection of foraging habitat by grassland birds: effect of prey
abundance or availability? Polish Journal of Ecology. 56: 365-370.
SAS Institute Inc. (2008). SAS user's guide, version 9.2., [online].
Tompkins, D.M. and Gleeson, D.M. (2006). Relationship between avian malaria distribution and an
exotic invasive mosquito in New Zealand. Journal of the Royal Society of New Zealand. 36: 51-62.
van Heezik, Y., Smyth, A. and Mathieu, R. (2008). Diversity of native and exotic birds across an urban
gradient in a New Zealand city. Landscape and Urban Planning. 87: 223-232.
Vitousek, P.M., Mooney, H.A., Lubchenco, J. and Melillo, J.M. (1997). Human domination of Earth's
ecosystems. Science. 277: 494-499.
Williams, P.A. and Karl, B.J. (1996). Fleshy fruits of indigenous and adventive plants in the diet of birds
in forest remnants, Nelson, New Zealand. New Zealand Journal of Ecology. 20: 127-145.
187
Discusión general y Conclusiones
Síntesis y Perspectivas
Esta tesis doctoral consta de cinco estudios que pretenden describir patrones de
impactos viarios sobre poblaciones animales, tratando de investigar los mecanismos que
generan estos patrones. Cada estudio se refiere a uno de los temas principales de la que
se define como Ecología de carreteras: atropello, efecto barrera, medidas de mitigación,
impacto sobre comportamiento y reproducción, e invasiones biológicas.
En el primer capítulo investigamos los factores específicos intrínsecos,
temporales y espaciales que afectaban la probabilidad y los patrones de atropello en
vertebrados mediterráneos. Tanto las especies comunes como las abundantes tuvieron
más probabilidad de ser atropelladas que las raras, y dentro del conjunto de especies
comunes y abundantes el rasgo de historia natural que más afectó a la probabilidad de
colisión fue la estrategia metabólica, siendo los ectotermos más atropellados de los
endotermos. Las lluvias se asociaron a los picos de atropello de anfibios, mientras las
altas temperaturas coincidieron con un aumento de atropellos en serpientes, lagartijas y
aves nidificantes. Lagartijas y micromamíferos fueron atropellados sobre todo a lo largo
de grandes vías con mucho tráfico y altos límites de velocidad. Sugerimos priorizar las
medidas de mitigación capaces de reducir permanentemente la mortalidad de ectotermos
(anfibios y reptiles), como pasos de fauna específicos y cercados impermeables en
correspondencia de sus puntos calientes de atropello. Al mismo tiempo, recomendamos
aplicar las medidas temporales de mitigación durante los picos estacionales de
atropellos.
188
En el segundo capítulo medimos el efecto barrera en ungulados (ciervo Cervus
elaphus y jabalí Sus scrofa) a lo largo de una red viaria heterogénea (i.e. con variación
de tráfico y superficie viaria) y estimamos el consecuente deterioro de hábitat dentro de
una emblemática área protegida, el Espacio Natural de Doñana. La probabilidad de
presencia de ciervos y jabalíes fue afectada por la mera distancia a la carretera más
cercana (en la mayoría de casos caminos sin pavimentación ni tráfico), y los primeros
sufrieron también la cercanía a carreteras de más envergadura y tránsito. El efecto
barrera debido a la carretera más cercana sugiere una evitación de espacios artificiales
vacíos que podría estar basada en la asociación entre estructuras lineares y la posibilidad
de vehículos transitando. La consecuencia a nivel de paisaje es una reducción de la
probabilidad de presencia del 32% por ciervos y del 45% por jabalíes, con una
consecuente reducción de calidad de hábitat, muy grave dentro de un enclave protegido.
Por esta razón sugerimos el establecimiento de grandes áreas sin infra-estructuras
lineares, a través de una optimización de la red viaria que incluía también el cierre de
caminos no pavimentados.
En el tercer capítulo hipotetizamos los tradicionales muestreos genéricos de
eficacia de pasos de fauna para micromamíferos pudieran estar sesgados hacía las
especies generalistas de hábitat en detrimento de las especialistas, considerado que las
carreteras a menudo están rodeadas por ambientes degradados. Por esta razón,
comparamos un muestreo genérico con uno específico, seleccionando tres especies
modelo: una generalista (Peromyscus maniculatus), una especialista de bosque (Myodes
gapperi) y una de pradera (Microtus pennsylvanicus). Muestreamos en cuatro tipos de
pasos de fauna (ecoductos, puentes, a sección elíptica y a sección cuadrangular) en el
Parque Nacional Banff (Canadá). Posicionamos trampas de huellas a lo largo de los
pasos, y como controles también delante de sus entradas y en los bosques colindantes.
189
El muestreo genérico no detectó diferencias en la ocurrencia de micromamíferos entre
pasos de fauna y controles, mientras que el específico evidenció que la única especie
que usaba consistentemente los pasos era la generalista, seleccionando principalmente
los ecoductos (que además fueron los únicos pasos usados por las especialistas). La
eficiencia de los pasos de fauna para micromamíferos especialistas podría ser
sobrestimada, con posibles implicaciones para su conservación. Sugerimos entonces
mejorar el hábitat (o por lo menos la disponibilidad de cobertura) a lo largo de los pasos
y en el espacio entre estos y los ambientes colindantes.
En el cuarto capítulo investigamos como el estrés asociado al tráfico puede
amenazar el éxito individual de aves que acuden a carreteras para la búsqueda de
alimento y lugares de nidificación. Intentamos averiguar si los abejarucos Merops
apiaster que crían en las áreas laterales de los caminos perciben el tráfico como un
elemento de estrés que estimule alarmas, estampidas y un decremento de los cuidados
parentales, o si en cambio demuestran habituación. Comparamos individuos sometidos
a diferentes niveles de tráfico dentro de la Reserva de la Biosfera de Doñana. Alarma y
estampidas resultaron aumentar en presencia de tráfico a motor, siendo mayores en las
colonias de cría establecidas en zonas de tránsito intenso. Sin embargo, las tasas de ceba
parentales también cambiaron de acuerdo al nivel de tráfico, pero contrariamente a las
predicciones fueron mayores en los nidos situados en carreteras de tránsito elevado,
aumentando durante los días laborales y con picos en las horas punta. El tráfico parece
afectar positivamente las tasas de ceba, aumentando potencialmente el éxito
reproductivo de los abejarucos. Por otro lado, hasta que no se determinen las
consecuencias netas del tráfico sobre salud individual, niveles de estrés y eficacia
biológica, sugerimos mantener los actuales niveles de tráfico.
190
En el quinto capítulo hipotetizamos que las carreteras intermunicipales ofrecen
una alta disponibilidad de alimentos de fácil acceso y pueden representar un área
preferente para la alimentación de aves sinantrópicas. Como consecuencia, una red de
carreteras podría actuar como corredor de invasión de aves urbanas no nativas. Por esta
razón intentamos proporcionar esta primera aproximación para verificar la posible
existencia de este patrón de dispersión, determinando si la búsqueda de alimento es
realmente la principal actividad desarrollada a lo largo de las carreteras
intermunicipales. Muestreamos entonces 200 puntos de la red de vías intermunicipales
de Nueva Zelanda, realizando observaciones sobre cuatro especies de aves sinantrópicas
invasoras (el gorrión Passer domesticus, el mirlo Turdus merula y el estornino Sturnus
vulgaris, todos de origen euroasiático; y el miná Acridotheres tristis, de origen indio).
Para cada especie también observamos: preferencias alimenticias, características de las
carreteras que puedan afectar la búsqueda de alimento, y selección de micro-hábitat de
alimentación.
Los
comportamientos observados estuvieron predominantemente
relacionados con actividad de alimentación, con diferencias en las estrategias de cada
especie. Los gorriones fueron observados alimentándose principalmente de semillas e
invertebrados, los estorninos y los minás mostraron comportamientos alimenticios más
generalistas, mientras que los mirlos fueron detectados principalmente depredando
invertebrados en los prados de las cunetas. Las carreteras intermunicipales realmente
proveen alimentos de fácil acceso, disponibles para todas las especies de estudio.
Proponemos entonces la existencia de un probable patrón de invasión por aves
sinantrópicas mediado por carreteras, e recomendamos futuros estudios para optimizar
el manejo focalizado hacia las carreteras intermunicipales como las áreas de control más
adecuadas para evitar la propagación de esas especies.
191
Esta tesis doctoral proporciona cinco ejemplos de estudios que no se limitan a
describir patrones, como a menudo se ha venido haciendo en Ecología de carreteras,
sino que pretenden investigar los mecanismos que generan esos patrones. El primer
capítulo, por ejemplo, trataba el tema de los atropellos de fauna intentando proporcionar
una información más completa que un listado de especies afectadas y las posibles
causas, ya que pretendía entender que factores climáticos, ambientales y de tráfico
determinaban la distribución temporal y espacial de las colisiones de distintos grupos
funcionales de especies.
El enfoque al estudio de mecanismos nos ha permitido sugerir medidas de
mitigación que pudieran integrar las que han sido realizadas hasta ahora, y que podrían
resultar muy efectivas, ya que pretenden resolver los impactos a través de un
conocimiento más profundo de su forma de afectar especies y ecosistemas. En el
segundo capítulo, por ejemplo, vimos que no es el tráfico motorizado el único factor
que determina efecto barrera en ungulados, considerado que estos han mostrado una
gran tendencia a evitar incluso caminos que apenas tenían tránsito. Como consecuencia,
sugerimos que en estos casos habría que integrar los pasos de fauna a menudo ya
existentes a lo largo de las carreteras principales con acciones de optimización de las
redes viarias, eliminando los caminos redundantes por lo menos dentro de las áreas
protegidas.
Por otro lado, este enfoque al estudio de mecanismos puede ser positivo no solo
para la resolución de problemas concretos como el de las medidas de mitigación, sino
que más en general puede contribuir considerablemente al avance del conocimiento en
Ecología de carreteras, ya que este tipo de enfoque a menudo proporciona la posibilidad
de hacer inferencia a una escala más amplia. En el tercer capítulo, por ejemplo,
observamos que los micromamíferos generalistas de hábitat eran los únicos que
192
realmente usaban los pasos de fauna en los bosques de las Montañas Rocosas
Canadienses, con toda probabilidad gracias precisamente a su tolerancia a entornos
degradados, mientras los especialistas de hábitat se quedaban confinados en los
ambientes que les eran más propicios. Haber diseñado este estudio con la finalidad de
entender que uso hacían de los pasos de fauna tanto las micromamíferos generalistas
como los especialistas nos ha hecho entender el probable mecanismo que determina la
eficiencia de los pasos de fauna para estos grupos de especies, y nos permite inferir que
estos patrones se podrían repetir también en micromamíferos de otras áreas e incluso
con otros taxa, siempre que sean generalistas y especialistas de hábitat.
Finalmente, los estudios enfocados en investigar mecanismos proporcionarán un
tipo de información que podrá resultar extremadamente útil para el desarrollo de la
Ecología de carreteras, que probablemente se centrará en un futuro próximo en efectos
crónicos como por ejemplo la distribución a nivel de paisaje de contaminantes y
personalidades, para luego enfocarse finalmente en estudios de síntesis y meta-análisis
necesarios para obtener patrones más globales y fiables de cómo funcionan los impactos
asociados a las redes viarias.
193
Conclusiones de esta tesis doctoral
1) Existen distintos factores intrínsecos a las especies animales que hacen que
algunas de ellas sean más susceptibles a ser atropelladas que otras, tanto en
nuestro caso de estudio (los vertebrados mediterráneos) como con toda
seguridad también en el resto de vertebrados e invertebrados del mundo. El
más lógico entre estos factores intrínsecos es la abundancia de una especie,
pero pueden haber otros factores, como los relacionados con su capacidad de
movimiento y su estrategia metabólica (endotermia versus ectotermia).
2) Existen factores temporales como el fotoperiodo, las precipitaciones y la
temperatura que determinan cambios en actividad y abundancia de algunos
grupos funcionales de especies a lo largo del año y consecuentemente la
variación de su probabilidad de atropello.
3) Existen factores espaciales como las dimensiones de una carretera, su tráfico
y la velocidad máxima permitida en ella que hacen que los atropellos de
algunos grupos funcionales de especies se concentren en algunos puntos
concretos. Por otro lado, también existen otros factores espaciales
relacionados con la variabilidad de abundancia de algunos grupos
funcionales de especies (e.g. la distancia a un cuerpo de agua) que
contribuyen a la variabilidad de la distribución espacial de los atropellos.
194
4) Algunas especies de ungulados (y probablemente muchas otras especies de
larga vida y con suficiente capacidad cognitiva) tienen tendencia a evitar las
infra-estructuras
lineares,
aunque
estas
sean
simple
caminos
sin
pavimentación ni tráfico motorizado circulante; lo que comporta una
reducción sensible de la calidad de hábitat de grandes áreas naturales
(incluso protegidas, como en nuestro caso de estudio).
5) En algunas especies de ungulados (y probablemente en muchas otras
especies) a la evitación de meras infra-estructuras lineares sin transito se
puede sumar la evitación de tráfico motorizado, que puede contribuir a su
vez a la mencionada reducción de la calidad de hábitat.
6) Las especies generalistas de hábitat (en nuestro caso de estudio micromamíferos) pueden usar las áreas degradadas como las carreteras y sus
alrededores con más frecuencia que las especialistas de hábitat; lo que puede
comportar sesgos en los muestreos de eficiencia de pasos de fauna que no
diferencien a nivel de especie.
7) Como se ha descrito ya en otras especies, también los micromamíferos
parecen preferir los ecoductos a los pasos de fauna subterráneos. Aún así, las
especies generalistas de hábitat los usan con más frecuencia que las
especialistas, así que para aumentar la eficacia de los ecoductos son
necesarias acciones específicas de mejora de su hábitat.
195
8) Algunas especies que potencialmente se benefician de la presencia de
carreteras (como los abejarucos en nuestro caso de estudio, que llegan a
nidificar en taludes de los caminos de mucho tráfico) siguen reaccionando
frente a los vehículos a motor como si estuvieran delante de depredadores
naturales (e.g. alarmas y estampidas), lo que podría provocar a largo plazo
que su eficacia biológica pueda verse afectada.
9) Los estudios de comportamiento animal a lo largo de infra-estructuras viarias
pueden revelar patrones sorprendentes, como en nuestro caso de estudio en
que el tránsito de vehículos a motor parece facilitar (directamente o
indirectamente) el forrajeo de aves insectívoras y por lo tanto sus tasas de
ceba al nido.
10) Las aves sinantrópicas que son invasoras en Nueva Zelanda (y
probablemente la mayoría de vertebrados sinantrópicos e invasores a lo largo
del globo) usan las carreteras para la búsqueda de alimento, lo que podría
facilitar en estas especies un patrón de invasión biológica mediado por la
presencia de infra-estructuras viarias.
196

On the road - RIO Principal

Transcription

Similar documents

23-June-The-influenc..

libroALTROper TIPO

Pedro Fernández began his international career when he was

After Party Flyer

Diptico de la Comarca de Guadix

St. Therese - communityofsttherese.org

Enero 3, 2016 Sagrada Familia - St. Anthony of Padua Catholic

August 3, 2014 - Holy Family Catholic Church