1
IDENTIFICACIÓN DE PUNTOS CRÍTICOS DE PERMEABILIDAD PARA LA FAUNA EN LAS VÍAS DE TRANSPORTE ANDALUZAS
EN ESCENARIOS DE CAMBIO CLIMÁTICO
Memoria Final
Universidad de Málaga
2
3
Identificación de puntos críticos de permeabilidad para la fauna en las vías de
transporte andaluzas en escenarios de cambio climático
Memoria Final
4
© Agencia de Obra Pública de la Junta de Andalucía. Consejería Fomento y Vivienda. Junta de Andalucía. 2013
Universidad de Málaga
Málaga. 20-01-2014
5
ÍNDICE DE LA MEMORIA FINAL
1. Introducción general ............................................................................................................................1 2. Objetivos .............................................................................................................................................2 3. Metodología.........................................................................................................................................3 4. Análisis de resultados ..........................................................................................................................7 5. Conclusiones.....................................................................................................................................68 6. Referencias .......................................................................................................................................69
Anejo nº1. Modelo de regresión logística
Anejo nº2. Difusión científica
6
1
1. Introducción general
Andalucía participa de una de las regiones de más alta biodiversidad de la Tierra (Myers & Giller, 1988). La
Unión Internacional para la Conservación de la Naturaleza mantiene que los ecosistemas andaluces suelen
ser pobres en nutrientes y sometidos a estrés estacional, pero ricos en especies. La mayoría de los taxones
presentan altos niveles de diversidad, tanto a nivel de especie como de subespecie. Muchos de los
mecanismos históricos y evolutivos de esta diversificación han sido atribuidos a la peculiar situación y
configuración geográfica de este territorio (incluyendo la variabilidad climática y la heterogeneidad de hábitat),
así como a la larga historia de cambios en la ordenación del territorio por parte de los pueblos que han
habitado esta zona, lo que ha creado y mantenido un amplio espectro de nuevos hábitats.
Andalucía, como la mayoría de las regiones europeas, presenta un elevado desarrollo económico, de forma
que, en general, los ecosistemas naturales han sido alterados o destruidos por los usos del suelo
dominantes, como son los relacionados con la agricultura o el desarrollo urbano. Por esta causa, uno de los
grandes desafíos del siglo XXI se refiere al esfuerzo concertado para desarrollar y aplicar estrategias
diseñadas para evitar la pérdida de biodiversidad. Esta tarea se hace cada vez más crucial debido al
crecimiento de la población humana, los elevados niveles de uso de energía y las emisiones de gases de
efecto invernadero asociadas, que pueden afectar al clima general de la Tierra. A esto hay que añadir que el
aumento de la demanda de suelo por parte del ser humano y la construcción de estructuras viarias se
traduce en altas tasas de pérdida y fragmentación de hábitat para otras especies. La asociación entre cambio
climático y fragmentación de hábitat de origen humano amenaza, por tanto, a un buen número de
poblaciones (Gaston, 2000).
El Grupo Intergubernamental de Expertos sobre el Cambio Climático de las Naciones Unidas (IPCC), en su
informe del año 2007, ha puesto de manifiesto el consenso científico básico sobre la realidad de este
cambio, aunque no faltan disidentes, así como la necesidad de implementar medidas mitigadoras de sus
efectos. Como consecuencia, la comunidad científica y las administraciones con competencia en
conservación muestran un interés creciente en predecir la respuesta espacial de las especies a los cambios
previsibles en el clima, para anticipar la necesaria adaptación de las medidas de conservación a tomar
(Araújo et al., 2004, Real et al., 2010). Evaluar las posibilidades de cambio de las distribuciones de las
especies en función de los diferentes escenarios de cambio climático se hace por tanto necesario para, por
ejemplo, conocer cómo cambiarán las áreas favorables para las especies en el futuro y cómo se pueden ver
afectados tales cambios por el efecto barrera de las infraestructuras públicas.
2
Los vertebrados incluyen algunas de las especies más carismáticas de la fauna ibérica, particularidad que en
parte deriva de su carácter endémico y del gran riesgo de extinción a que están expuestas. Numerosas
especies ibéricas de vertebrados se encuentran claramente amenazadas. De especial importancia es el
rápido declive de la herpetofauna, ya que son las especies más vulnerables a los rápidos cambios en el
hábitat (Araújo et al., 2006), debido en parte a que la mayoría de ellas presentan áreas de distribución
reducidas y poca capacidad de dispersión. Es también conocido que los tetrápodos, especialmente los no
voladores, que habitan los humedales (anfibios, reptiles y mamíferos acuáticos) son sensibles a los cambios
en el régimen de lluvias y de evapotranspiración (Jackson et al., 2001). Por todo ello es previsible que las
áreas favorables para estas especies se vean modificadas si el clima cambia como prevé el IPCC, por lo que
las especies deberán distribuirse por áreas diferentes a las actuales para adaptarse a esas modificaciones.
Cada vez más, nuevos análisis espaciales y sistemas de observación de la Tierra juegan un papel central en
el esfuerzo para prever las respuestas adaptativas de las especies a los cambios ambientales. Uno de los
enfoques más usados ha sido la estimación, mediante modelos biogeográficos, de la respuesta potencial de
las especies a los cambios futuros en caso de variación, a gran escala, de los parámetros eco-climáticos y
biofísicos (Guisan & Zimmermann, 2000, Márquez, et al., 2011). Los modelos espaciales resultan
apropiados para prever la evolución de la distribución de las especies en distintos escenarios de cambio
global. En los últimos tiempos, los estudios de modelación biogeográfica se están convirtiendo, cada vez
más, en herramientas fundamentales para la gestión racional de la fauna y para la conservación del medio
ambiente. Los modelos espaciales permiten hacer inferencias sobre la distribución de las especies en
función de las características ambientales, bióticas, espaciales y humanas del territorio estudiado, aún
cuando no exista un conocimiento exhaustivo de la distribución de las especies. Este tipo de modelos serán
utilizados en nuestra propuesta para pronosticar la evolución de las condiciones favorables para los
vertebrados hasta el año 2040 en distintos escenarios de cambio climático.
En la actualidad se tiene un conocimiento suficiente de la distribución de los tetrápodos no voladores
amenazados en Andalucía (Franco y Rodríguez, 2001), lo que constituye el punto de partida del presente
proyecto.
2. Objetivos
Evaluar las posibilidades de cambio de las distribuciones de distintas especies faunísticas en función de los
diferentes escenarios de cambio climático se hace necesario para conocer cómo cambiarán las áreas
favorables y como pueden afectar factores externos, como es el caso del efecto barrera de las
infraestructuras públicas.
3
Los objetivos básicos que se establecen en el presente proyecto de I+D+i denominado “Identificación de
puntos críticos de permeabilidad para la fauna en las vías de transporte andaluzas en escenarios de cambio
climático” son:
1. Elaborar modelos de distribución en Andalucía de las especies de tretrápodos no voladores
amenazados según las predicciones de los diferentes modelos de circulación y escenarios de emisiones
en el periodo 1961-1990.
2. Proyectar al futuro (2011-2040) los modelos obtenidos para determinar los cambios previstos en
sus distribuciones.
3. Establecer cuales serían los “puntos negros” de potenciales conflictos entre las áreas favorables
futuras para las especies y las infraestructuras de transporte andaluzas.
Los resultados previsibles serán mapas de riesgos asociados a las infraestructuras viarias andaluzas en
escenarios de cambio climático, de forma que se pueda enfocar en ellos las actuaciones para facilitar las
migraciones de las especies amenazadas forzadas por el cambio en el clima.
3. Metodología
Para el estudio se han considerado 19 especies de tetrápodos no voladores con algún grado de amenaza en
Andalucía (Franco y Rodríguez, 2001, Ley8/2003 de 28 de octubre, de la flora y la fauna silvestre, BOJA
128: 23790-23810), se ha incluido también Triturus pygmaeus porque las poblaciones más orientales
presentan problemas de conservación (Pleguezuelos, et al. 2004) (Tabla 1).
Se elaboró una base de datos con la distribución en Andalucía de las 20 especies de tetrápodos no voladores
amenazados, otra con las variables de cambio climático para dos modelos de circulación: CGCM2 del
Canadian Climate Centre for Modeling y ECHAM4 del Max Planck Institut für Meteorologie, y dos escenarios
de emisiones del IPCC: A2 y B2 (Nakicenovic et al. 2000) para el presente (1961-1990) y para el futuro
próximo (2011-2040). Las variables de clima se obtuvieron de la Agencia Estatal de Meteorología (AEMET).
Tabla 1:. Especies con algún grado de amenaza en Andalucía. (CR): en peligro crítico, (EN): en peligro,
(VU): vulnerable, (*): con interés de conservación.
Anfibios Reptiles Mamíferos
Salamandra salamandra
longirostris (VU)
Emys orbicularis (VU) Atelerix algirus (EN)
Alytes dickhilleni (VU) Testudo graeca (EN) Neomys anomalus (EN)
Triturus pygmaeus (*) Lacerta schreiberi (CR) Talpa occidentalis (VU)
4
Algyroides marchi (VU) Canis lupus (CR)
Coronella austriaca (EN) Lutra lutra (VU)
Vipera latasti (VU) Lynx pardinus (EN)
Capreolus capreolus (VU)
Sciurus vulgaris (VU)
Arvicola sapidus (VU)
Chionomys nivalis (EN)
Microtus cabrerae (CR)
Además, se creó otra base de datos con variables ambientales que incluye la situación geográfica, topografía,
actividad humana y clima en Andalucía. La unidad geográfica operativa utilizada en estas bases de datos fue
siempre la cuadrícula UTM de 10x10 km2. Las variables ambientales consideradas en este estudio se
muestran en la Tabla 2.
Se elaboraron modelos de favorabilidad ambiental para las 20 especies con las variables del modelo de
circulación CGCM2 y los escenarios A2 y B2, usando la regresión logística por pasos hacia delante y la
función de favorabilidad (Real et al. 2006) y considerando las tendencias espaciales (Legendre & Legendre,
1998). Como punto de partida se usaron los modelos de favorabilidad ambiental para España continental
obtenidos comprobando la relación de cada una de las variables con la distribución de cada especie y
conservando en el modelo sólo aquellas relaciones significativas (p< 0.05). Para evitar el incremento de error
tipo I (Benjamini & Hochberg 1995; García 2003) se controló la tasa de descubrimiento falso (FDR) usando
el procedimiento propuesto por Benjamini & Hochberg (1995) que acepta solo las variables que cumplen la
condición q< 0.05. De esta forma se asegura que las variables del modelo final realizan una aportación
efectiva al poder explicativo de dicho modelo.
Se usó la variación del factor de inflación (VIF) para cuantificar la colinealidad entre las variables de un
modelo. El factor de inflación representa la correlación entre el predictor y el resto de variables del modelo.
Un valor alto del factor de inflación no supone ningún problema a la hora de modelar distribución actual de
una especie porque el proceso de selección de variables en la regresión logística por pasos ya tiene en
cuanta ese factor. Sin embargo, si es un problema a la hora de transferir el modelo a las condiciones futuras
ya que las correlaciones entre las variables en el futuro puede ser diferente a las actuales. Así, en los
modelos nacionales de partida solo se incluyeron las variables climáticas estacionales de precipitación y
temperatura (p13, p14, p15, p16, tx13, tx14, tx15, tx16).
Se evaluaron tres aproximaciones para elaborar los modelos de distribución en Andalucía de las 20 especies
consideradas en este estudio.
5
- Modelo nacional. Consistió en usar los modelos de las especies obtenidos para toda España
proyectados solo en Andalucía.
- Modelo regional. Se elaboró un nuevo modelo para cada especie exclusivamente con los datos de
Andalucía. Se uso el mismo procedimiento que en los modelos nacionales.
- Modelo actualizado. Se actualizó el modelo nacional usando las mismas variables y recalibrándolo
con los datos de las variables para Andalucía. En este caso en lugar de una regresión por pasos
hacia delante se usó en método enter para asegurar la recalibración en el área de estudio de las
variables seleccionadas en el modelo nacional.
Tabla 2:. Factores explicativos y variables utilizadas para modelar las distribuciones de las especies.
Fuentes: (1)I.G.N. (1999); (2)US Geological Survey (1996); (3)Farr and Kobrick (2000); (4)ORNL (2001); (5)Agencia Estatal de Meteorología of Spain (AEMET), Ministerio de Medio Ambiente (http://www.aemet.es/es/elclima/cambio_climat/escenarios).
Factores Código Variables
La Latitude (ºN)(1) Situación espacial
Lo Longitude (ºE)(1)
A Mean altitude (m)(2)
S Slope (º) (calculated from altitude)
orS Grados orientación sur(3)
Topografía
orW Grados orientació oeste(3)
Daut Distancia a la autovía más próxima (km)(1)
U100 Distancia al centro urbano más próximo de más de 100 000 (km)(1)
U500 Distancia al centro urbano más próximo de más de 500 000 (km)(1)
Actividad humana
HPd Densidad de población en 2000 (Nº habitantes/Km2)(4)
p0 Precipitación anual (mm)(5)
p13 Precipitación en primavera(mm)(5)
p14 Precipitación en verano (mm)(5)
p15 Precipitación en otoño (mm)(5)
p16 Precipitación en invierno (mm)(5)
tx0 Temperatura máxima anual(5)
tx1 Temperatura máxima de enero(5)
tx7 Temperatura máxima de julio (5)
tx13 Temperatura máxima en primavera (5)
tx14 Temperatura máxima en verano (5)
tx15 Temperatura máxima en otoño(5)
Clima
tx16 Temperatura máxima en invierno (5)
c2: CGCM2-A2, c4: CGCM2-B2, e2: ECHAM4-A2, e4: ECHAM4-B2
6
Los criterios para evaluar los modelos resultantes de las diferentes aproximaciones han sido: Cohen’s kappa
con un umbral de favorabilidad de 0.5, AUC (Area Under the Curve) y AIC (Akaike Information Criterion) (Real
et al. 2010, Márquez et al 2011) para determinar su capacidad de clasificación, de discriminación y la
parsimonia de los mismos, respectivamente. La calibración de los modelos se estimó usando RSME (Root
Mean Square Error) de las presencias predichas y observadas en diez intervalos de igual probabilidad (bins).
Se consideró como bien calibrados aquellos modelos con el menor RSME de las tres aproximaciones
(nacional, regional y actualizada) que cumplieran la condición de que en los 10 intervalos de igual
probabilidad (bins) considerados haya más de 15 cuadrículas y su probabilidad media * nº cuad ≥ 5.
Los modelos mejor evaluados y bien calibrados se proyectaron al futuro (2011-2040). Para esos casos se
calculó las diferencias en la favorabilidad futura y presente. Se trató de encontrar para qué especies sus
distribuciones actuales se prevé que van a experimentar un desplazamiento en el futuro en respuesta al
cambio de las condiciones climáticas. De esta forma cuando se representen las diferencias de favorabilidad
futura y presente se apreciarán zonas donde se producirán perdidas de favorabilidad y zonas donde ésta se
incrementará. Las infraestructuras viarias que dificulten el desplazamiento de las distribuciones de las
especies desde las zonas con perdida de favorabilidad a las zonas donde se pronostica un incremento de la
misma serán las que podrían ejercer un efecto barrera.
Para la evaluación del efecto barrera se estimó el posible flujo entre áreas separadas por una infraestructura
viaria. En cada área se promedio la diferencia de favorabilidad futura y presente. Se considera que el flujo va
desde zonas con mayor pérdida de favorabilidad (valores negativos, que indican que esa zona será más
desfavorable en el futuro de lo que es en el presente) hacia zonas con menor pérdida o con aumento de
favorabilidad (valores positivos, que indican que esa zona será más favorable en el futuro de lo que es en el
presente). La intensidad del flujo se establece como la diferencia entre los promedios de cambio de
favorabilidad de cada zona separada por una infraestructura viaria.
7
4. Análisis de resultados
El Anexo 1 recoge los modelos de regresión logística resultantes para cada una de las tres aproximaciones
consideradas en este estudio.
Las Figuras 1-3 muestran los mapas de distribución de las especies y los mapas de favorabilidad obtenidos a
partir de cada aproximación para los modelos de circulación CGCM2-A2, CGCM2-B2 y ECHAM4-A2/B2,
respectivamente.
Fig. 1: Mapas de distribución de cada especie y mapas de favorabilidad ambiental para las condiciones actuales de acuerdo con la aproximación del modelo nacional (NAC), del modelo regional (REG) y del modelo nacional actualizado (ACT) para el modelo de circulación CGCM2-A2.
8
Fig. 1: cont.
9
Fig. 1: cont.
10
Fig. 1: cont.
11
Fig. 1: cont.
12
Fig. 1: cont.
13
Fig. 1: cont.
14
Fig. 1: cont.
15
Fig. 1: cont.
16
Fig. 1: cont.
17
Fig. 2: Mapas de distribución de cada especie y mapas de favorabilidad ambiental para las condiciones actuales de
acuerdo con la aproximación del modelo nacional (NAC), del modelo regional (REG) y del modelo nacional actualizado
(ACT) para el modelo de circulación CGCM2-B2.
18
Fig. 2: cont.
19
Fig. 2: cont.
20
Fig. 2: cont.
21
Fig. 2: cont.
22
Fig. 2: cont.
23
Fig. 2: cont.
24
Fig. 2: cont.
25
Fig. 2: cont.
26
Fig. 2: cont.
27
Fig. 3: Mapas de distribución de cada especie y mapas de favorabilidad ambiental para las condiciones actuales de acuerdo con la aproximación del modelo nacional (NAC), del modelo regional (REG) y del modelo nacional actualizado (ACT) para el modelo de circulación ECHAM4-A2/B2.
28
Fig. 3: cont.
29
Fig. 3: cont.
30
Fig. 3: cont.
31
Fig. 3: cont.
32
Fig. 3: cont.
33
Fig. 3: cont.
34
Fig. 3: cont.
35
Fig. 3: cont.
36
Fig. 3: cont.
37
La Tabla 3 muestra los valores de los criterios de evaluación (Cohen’s kappa con un umbral de favorabilidad
de 0.5, AUC (Area Under the Curve) y AIC (Akaike Information Criterion)) para cada especie, aproximaciones
y modelo de circulación consideradas en este estudio. Además recoge la calibración de cada modelo (RSME).
Para los tres modelos de circulación el mayor número de modelos de favorabilidad ambiental con la mejor
evaluación se dio en aproximación del modelo actualizado (ACT). La aproximación regional (REG) también es
aceptable para algunas especies. En ningún caso la aproximación nacional (NAT) fue la mejor, aunque para
T. graeca y L. schreiberi superó en valoración a la regional.
El hecho de que la aproximación nacional no haya sido, en ningún caso, la mejor contradice claramente la
tesis que esgrimen muchos modeladores de que siempre se debería modelar la distribución completa de una
especie, que parte de ella. La heterogeneidad ambiental de Andalucía no se da en la misma medida en el
resto de la España peninsular.
38
Tabla 3:. Criterios de evaluación y calibración de los modelos para cada especie y aproximación. En azul la mejor valoración de las tres aproximaciones. Especies en las que la mejor la aproximación es la regional al menos para dos de los criterios de evaluación. Especies en las que la mejor la aproximación es la actualización al menos para dos de los criterios de evaluación.
NAT REG ACT CGCM2-A2 AIC AUC kappa RMSE AIC AUC kappa RMSE AIC AUC kappa RMSE
S. salamandra 903.15 0.833 0.460 0.0902 710.63 0.906 0.658 0.1266 605.70 0.932 0.702 0.0830
A. marchi 57.37 0.994 0.564 0.3316 55.69 0.995 0.532 0.2790 51.30 0.996 0.555 0.2190 C. austriaca 72.38 0.901 -0.0031 0.1470 31.37 0.982 0.099 0.2691 26.01 1 1 0.0002
L. schreiberi 72.50 0.480 -0.0037 0.0854 -- -- -- - 24.00 1 1 0.0000
T. graeca 103.51 0.983 0.220 0.2376 113.90 0.963 0.203 0.1393 97.06 0.985 0.310 0.2030 V. latasti 842.01 0.725 0.225 0.0741 785.89 0.751 0.277 0.1067 741.38 0.813 0.345 0.1158
N. anomalus 263.95 0.733 -0.002 0.0775 206.77 0.867 0.120 0.1236 206.96 0.886 0.157 0.2091
T. occidentalis 617.59 0.693 0.106 0.2912 438.90 0.884 0.357 0.1013 450.93 0.893 0.357 0.3297 C. capreolus 530.70 0.709 0.065 0.2559 331.87 0.914 0.382 0.0559 343.39 0.920 0.384 0.1034
S. vulgaris 484.96 0.902 0.472 0.1082 333.30 0.955 0.550 0.1186 324.20 0.963 0.568 0.0637
C. nivalis 64.33 0.996 0.404 0.2393 25.76 0.999 0.660 0.2822 22.00 1 1 0.0000 A. dickhilleni 388.94 0.929 0.357 0.0785 284.90 0.966 0.407 0.0455 363.61 0.940 0.272 0.0755
T. pygmaeus 786.02 0.817 0.232 0.0766 705.57 0.863 0.458 0.1127 752.59 0.837 0.408 0.0639
C. lupus 487.91 0.640 -0.0042 0.1499 125.88 0.990 0.621 0.1244 136.30 0.988 0.591 0.1176 L. lutra 1044.14 0.804 0.458 0.0765 876.32 0.853 0.604 0.0812 888.29 0.856 0.595 0.0636
L. pardinus 174.97 0.924 0.128 0.2380 153.30 0.947 0.227 0.1100 167.88 0.933 0.181 0.0979
A. sapidus 903.66 0.653 0.196 0.2404 840.51 0.721 0.273 0.1265 846.30 0.717 0.243 0.0947 M. cabrerae 99.92 0.929 0.014 0.1294 10.00 1 1 0.0000 23.35 0.999 0.269 0.4856
E. orbicularis 612.74 0.814 0.197 0.1977 519.65 0.881 0.361 0.1204 534.13 0.880 0.371 0.1266
A. algirus 80.48 0.972 0.250 0.2189 65.78 0.969 0.142 0.1864 70.29 0.985 0.206 0.2815
39
Tabla 3: cont.
NAT REG ACT CGCM2-B2 AIC AUC kappa RMSE AIC AUC kappa RMSE AIC AUC kappa RMSE
S. salamandra 896.59 0.835 0.461 0.0904 724.92 0.900 0.652 0.1105 587.40 0.936 0.714 0.0665
T. pygmaeus 789.39 0.815 0.231 0.0794* 762.67 0.839 0.422 0.0720* 752.64 0.839 0.418 0.0821 A. marchi 57.37 0.994 0.564 0.3316 55.57 0.995 0.532 0.2792 51.30 0.996 0.555 0.2190
C. austriaca 72.45 0.900 -0.0031 0.3318 35.37 0.982 0.099 0.3475 26.01 1.000 1 0.0002
E. orbicularis 626.68 0.810 0.184 0.1010 530.96 0.877 0.351 0.1156 536.71 0.880 0.363 0.1141 L. schreiberi 71.40 0.486 -0.0036 0.0775 -- -- -- - 24.00 1.000 1 0.0000
T. graeca 103.64 0.983 0.219 0.1567 113.56 0.963 0.205 0.1033 97.21 0.986 0.346 0.2261
V. latasti 842.38 0.725 0.226 0.0697 785.34 0.752 0.279 0.1066 741.06 0.814 0.340 0.0927 N. anomalus 260.33 0.757 -0.0059 0.0931 210.78 0.867 0.120 0.1227 205.25 0.895 0.160 0.1353
T. occidentalis 617.96 0.693 0.109 0.2928 437.73 0.885 0.352 0.1148 451.11 0.893 0.358 0.3307
C. lupus 487.66 0.639 -0.0042 0.1508 144.21 0.985 0.591 0.0753 136.79 0.988 0.591 0.1207 C. capreolus 533.07 0.706 0.056 0.2465 340.22 0.915 0.372 0.0117 341.36 0.922 0.411 0.1150
S. vulgaris 478.18 0.909 0.489 0.1028 343.51 0.957 0.540 0.1211 335.51 0.957 0.561 0.0000
C. nivalis 73.55 0.997 0.374 0.3994 29.76 0.999 0.660 0.2822 28.00 1 1 0.0000 L. pardinus 173.60 0.929 0.131 0.2527 150.32 0.951 0.238 0.1854 164.77 0.937 0.198 0.0740
A. sapidus 902.81 0.654 0.189 0.1873 827.18 0.735 0.259 0.1467 849.74 0.713 0.252 0.0940
M. cabrerae 99.22 0.930 0.0145 0.1292 16.00 1 1 0.0000 23.38 0.999 0.269 0.4867 A. dickhilleni 371.82 0.940 0.357 0.0692 284.04 0.966 0.386 0.0472 289.43 0.966 0.392 0.0448
A. algirus 75.94 0.972 0.245 0.2391 67.84 0.960 0.142 0.1015 65.00 0.987 0.203 0.2318
L. lutra 1022.74 0.807 0.442 0.0615 896.66 0.852 0.602 0.0892* 891.02 0.853 0.517 0.0232
40
Tabla 3: cont. NAT REG ACT
ECHAM4-A2/B2 AIC AUC kappa RSME AIC AUC kappa RSME AIC AUC kappa RSME
S. salamandra 847.75 0.849 0.498 0.0755 719.40 0.899 0.658 0.0881 648.30 0.924 0.662 0.0626
A. marchi 59.37 0.994 0.564 0.3316 57.21 0.994 0.485 0.2461 53.30 0.996 0.555 0.2190 C. austriaca 75.71 0.823 -0.002 0.0651 33.37 0.982 0.099 0.3475 38.64 0.998 0.192 0.3666
L. schreiberi 52.88 0.984 0.000 0.0278 -- -- -- -- 0.00011 1.000 1.000 0.0000
V. latasti 851.85 0.720 0.212 0.1092 790.18 0.747 0.260 0.0677 759.41 0.806 0.346 0.1306 N. anomalus 263.40 0.789 0.000 0.0824 209.76 0.861 0.115 0.1265 206.98 0.906 0.181 0.2293
T. occidentalis 580.79 0.740 0.199 0.2934 454.06 0.872 0.332 0.1849 459.62 0.882 0.349 0.1350
C. capreolus 447.63 0.826 0.263 0.1224 334.31 0.914 0.375 0.1057 342.37 0.915 0.390 0.1349 A. dickhilleni 312.77 0.962 0.436 0.0320 304.96 0.961 0.537 0.0410 301.47 0.965 0.571 0.0736
T. pygmaeus 807.32 0.813 0.168 0.0975 773.27 0.824 0.388 0.0967 761.10 0.843 0.425 0.1157
C. lupus 548.81 0.668 0.000 0.0736 156.42 0.983 0.587 0.1418 152.85 0.984 0.549 0.1144 C. nivalis 71.30 0.998 0.329 0.3211 27.76 0.999 0.660 0.2822 32.00 1.000 1.000 0.0000
A. sapidus 896.61 0.662 0.169 0.2924 865.90 0.684 0.221 0.0862 874.25 0.685 0.231 0.2783
A. algirus 81.52 0.973 0.194 0.2795 66.69 0.963 0.155 0.1753 72.15 0.982 0.223 0.1949 T. graeca 91.63 0.987 0.251 0.1544 75.30 0.994 0.476 0.3728 78.41 0.992 0.495 0.2652
S. vulgaris 474.44 0.909 0.499 0.1570 321.68 0.958 0.554 0.0732 371.30 0.942 0.545 0.0741
L. lutra 1017.76 0.818 0.497 0.0915 875.15 0.855 0.604 0.0658 910.38 0.851 0.562 0.0513 L. pardinus 178.85 0.921 0.101 0.3589 130.62 0.974 0.340 0.1341 177.04 0.920 0.179 0.3735
M. cabrerae 43.90 0.993 0.246 0.1315 8.00 1.000 1.000 0.7047 24.00 1.000 0.372 0.0000
E. orbicularis 675.82 0.744 0.112 0.1008 518.03 0.883 0.400 0.1174 629.91 0.807 0.237 0.1968
45
Fig. 4: Gráficas de calibración de los modelos para el escenario CGCM2-A2 para las tres aproximaciones. Se ha resaltado el modelo mejor calibrado y que cumple estrictamente los criterios de calibración (todos los bin de probabilidad tienen más de 15 cuadrículas y nº cuad*Probabliidad media = 5)
46
Fig. 4: cont.
47
Fig. 4: cont.
48
Fig. 4: cont.
49
Fig. 5: Gráficas de calibración de los modelos para el escenario CGCM2-B2 para las tres aproximaciones. Se ha resaltado el modelo mejor calibrado y que cumple estrictamente los criterios de calibración (todos los bin de probabilidad tienen más de 15 cuadrículas y nº cuad*Probabliidad media = 5)
50
Fig. 5: cont.
51
Fig. 5: cont.
52
Fig. 5: cont.
53
Fig. 6: Gráficas de calibración de los modelos para el escenario ECHAM4-A2/B2 para las tres aproximaciones. Se ha resaltado el modelo mejor calibrado y que cumple estrictamente los criterios de calibración (todos los bin de probabilidad tienen más de 15 cuadrículas y nº cuad*Probabliidad media = 5)
54
Fig. 6: cont.
55
Fig. 6: cont.
56
Fig. 6: cont.
57
Para no incrementar la incertidumbre solo se proyectaron al futuro los modelos bien calibrados que han
resultado ser:
Escenario CGCM2–A2:
S. salamandra
S. vulgaris
L. lutra
Escenario CGCM2–B2:
S. salamandra
L. lutra
Escenario EHAM–A2/B2:
S. salamandra
L. lutra
Las figuras 7 - 15 muestran los mapas de favorabilidad ambiental presente y futura (2040) para las especies
con modelos bien calibrados.
Fig. 7: Mapas de favorabilidad ambiental de S. salamandra para el escenario CGCM2-A2
58
Fig. 8: Mapas de favorabilidad ambiental de S. vulgaris para el escenario CGCM2-A2
Fig. 9: Mapas de favorabilidad ambiental de L. lutra para el escenario CGCM2-A2
59
Fig. 10: Mapas de favorabilidad ambiental de S. salamandra para el escenario CGCM2-B2
Fig. 11: Mapas de favorabilidad ambiental de L. lutra para el escenario CGCM2-B2.
60
Fig. 12: Mapas de favorabilidad ambiental de S. salamandra para el escenario ECHAM-A2
Fig. 13: Mapas de favorabilidad ambiental de L. lutra para el escenario ECHAM-A2
61
Fig.14: Mapas de favorabilidad ambiental de S. salamandra para el escenario ECHAM-B2
Fig. 15: Mapas de favorabilidad ambiental de L. lutra para el escenario ECHAM-B2
Los resultados muestran que las áreas favorables futuras de la mayoría de las especies y escenarios
experimentarán una contracción, no se observan grandes desplazamientos de las áreas favorables. (Figuras
7 - 15).
62
Las diferencias mínimas y máximas entre la favorabilidades ambientales futuras y presentes de cada especie
se muestran en la Tabla 4.
Tabla 4: Diferencias mínimas y máximas entre las favorabilidades ambientales futuras (2040) y presentes para cada una de las especies con modelo bien calibrado. Se destaca el caso con mayor diferencia entre pérdidas y ganancias futuras.
Diferencias favorabilidad futura-presente
Escenarios Especies Mínimo Máximo
S. salamandra -0.8311 0.0000
S. vulgaris -0.0830 0.4524 CGCM2-A2
L. lutra -0.8367 0.0988
S. salamandra -0.9296 0.0000 CGCM2-B2
S. vulgaris -0.5165 0.0067
S. salamandra -0.7181 0.4765 ECHAM4-A2
L. lutra -0.6806 -0.0008
S. salamandra -0.1993 0.6566 ECHAM4-B2
L. lutra -0.4324 0.0105
Tabla 5: Diferencias mínimas y máximas entre las favorabilidades ambientales futuras (2040) y presentes para cada una de las especies con modelo bien calibrado considerando el efecto del clima puro. Diferencias favorabilidad futuro-
presente Mínimo Máximo
S. salamandra -0.2042 0.0000
S. vulgaris -0.0088 0.0000 CGCM2-A2
L. lutra 0.0000 0.0951
S. salamandra -0.1049 0.0000 CGCM2-B2
S. vulgaris 0.0000 0.0397
S. salamandra -0.1029 0.0634 ECHAM4-A2
L. lutra -0.1897 -0.0003
S. salamandra -0.0218 0.0770 ECHAM4-B2
L. lutra -0.1054 0.0025
Además, para estas especies (S. salamandra, S. vulgaris y L. lutra) se consideró un factor de corrección del
efecto puro del clima, pero resultó que se atenuaban las diferencias entre las favorabilidades ambientales
presentes y futuras. Para ninguna de las especies se apreció un rango de diferencias en las favorabilidades
63
de los dos periodos que permitiera evaluar el posible efecto barrera de las infraestructuras frente al
desplazamiento de la distribución de las especies como respuesta frente al cambio del clima (Tabla 5).
El máximo rango de diferencias entre las áreas favorables en el 2040 y las favorables en la actualidad se dio
para salamandra en el escenario ECHAM4-A2 (Tabla 4, Figura12). Por tanto se ha elegido esta especie en
dicho escenario para evaluar el efecto barrera de las infraestructuras. El mapa de diferencias entre las
favorabilidades ambientales futura y presente de la especie y las principales infraestructuras viarias se
muestra en la Figura 16. La favorabilidad ambiental de S. salamandra en Andalucía, de acuerdo con las
predicciones climáticas del modelo de circulación ECHAM4-A2, se verá reducida en la mitad occidental, en
cambio, en el cuadrante nororiental se incrementará la favorabilidad ambiental para la especie.
S. salamandra presenta dos subespecies en Andalucía, S. salamandra morenica, que ocupa áreas
montañosas al norte del río Guadalquivir en Sierra Morena y sierras de Cazorla, Segura y Alcaraz, y S.
salamandra longirostris, que ocupa las sierras de Cádiz y Málaga. La reducción de las áreas favorables va a
afectar más a longirostris que a morenica porque esta última podrá encontrar zonas favorables en el
cuadrante nororiental (Figura 16). Sin embargo, morenica tendrá que superar las barreras de la línea férrea
Córdoba – Almorchón y la línea del Ave Sevilla-Córdoba-Madrid, en ambos casos en sus tramos entre
Córdoba capital y el límite de Andalucía, para alcanzar las zonas más favorables.
No se predice que pueda haber un solapamiento de las zonas favorables de las dos subespecies.
Como se puede comprobar en la Figura 17, para las dos subespecies algunas infraestructuras podrían
ejercer un efecto barrera en su posible desplazamiento a zonas climáticamente más favorables. La intensidad
de los posibles flujos para las dos subespecies se muestra en la Figura 18. Los flujos son más intensos en la
zona ocupada por la subespecie morenica
64
Fig. 16: Diferencias de favorabilidad ambiental entre el futuro (2040) y el presente para S. salamandra en el escenario ECHAM4-A2. Las zonas en tonos rojizos tendrán pérdidas de favorabilidad ambiental y las zonas en tonos verdes incrementarán la favorabilidad ambiental con respecto a la situación actual
65
Fig. 17: Distribución de presencias de las dos subespecies de salamandra en Andalucía (S. salamandra longirostris y S. salamandra morenica)
66
Fig. 18: Posible flujo de desplazamiento de los individuos en el 2040 desde zonas desfavorables a zonas más favorables para las dos subespecies de salamandra (S. salamandra longirostris y S. salamandra morenica). El grosor de las flechas indica el efecto barrea de la infraestructura sobre la que se sitúa.
67
Los resultados del proyecto se han presentado en el 6th International Conference of the International
Biogeography Society” que se celebró del 9 al 13 de enero de 2013, en Miami, Florida (USA), y en el 11th
International Mammalogical Congress que se celebró en Belfast (UK) del 11 al 16 de agosto de 2013 . Se
adjunta pdf de los pósters (Anexo 2).
Además de estos estudios directamente derivados del proyecto, el investigador principal del mismo y la
investigadora contratada para desarrollarlo han participado en dos artículos relacionados con la modelación
de especies y las infraestructuras viarias, en los que se ha volcado conocimiento generado en el transcurso
de este proyecto:
- Muñoz, A.R & Real, R. (2013) Distribution of Bonelli’s Eagle Aquila fasciata in southern Spain: scale may
matter. Acta Ornithologica 48(1):93-101. ISSN 0001-6454, Editorial: Museum & Inst Zoology, Wilcza 64, pl-
00-679 Varsovia, Polonia. (Factor de Impacto= 1,681, 5/22 Q1 en Ornitología). Se adjunta pdf del artículo
(Anexo 2).
- Fa, J. E., Farfán, M. A., Márquez, A. L., Duarte, J., Nackoney, J., Hall, A, Dupain, J. Seymour, S.,1
Johnson, P. J., Macdonald, D. W., Real, R., Vargas, J. M. (under revision) Bushmeat in markets tell a
depletion story. Conservation Biology.
68
5. Conclusiones
- De las tres aproximaciones utilizadas en este estudio para modelar la distribución de las especies en
Andalucía la actualización del modelo nacional para el territorio andaluz (ACT) resulto ser la mejor
aproximación en la mayoría de los casos.
- Para tres especies (S. salamandra, S. vulgari, y L. lutra) se han conseguido modelos bien calibrados.
Solo dichos modelos fueron los proyectados al futuro para evitar añadir más incertidumbre a las
predicciones futuras de los modelos.
- Las áreas favorables futuras predichas por la mayoría los modelos para estas especies
experimentarán una contracción, no se observan grandes desplazamientos de las áreas favorables.
Por tanto, en Andalucía las infraestructuras viarias no serán una barrera para el desplazamiento de
las distribuciones de la mayoría de las especies amenazadas en la respuesta al cambio climático
predicho para el futuro.
- Solo para salamandra en el escenario ECHAM4-A2 se detectó un desplazamiento patente de la
distribución de las áreas favorables en el futuro con respecto a las áreas favorables del presentes.
Según este modelo, la favorabilidad ambiental de la salamandra en Andalucía se verá reducida en la
mitad occidental y, por el contrario, se incrementará en el cuadrante nororiental.
- La reducción de las áreas favorables va a afectar más a S. salamandra longirostris que a S.
salamandra morenica porque esta última podrá encontrar zonas favorables en el cuadrante
nororiental. Sin embargo, S. s. morenica tendrá que superar la barreras de la línea férrea Córdoba –
Almorchón y la línea férrea de alta velocidad Sevilla-Córdoba-Madrid en los tramos que van desde
Córdoba capital hasta el límite septentrional de Andalucía.
69
6. Referencias
Agencia Estatal de Meteorología of Spain (AEMET), Ministerio de Medio Ambiente
(http://www.aemet.es/es/elclima/cambio_climat/escenarios).
Araújo, M.B., Cabezas, M., Thuiller, W. et al. (2004) Would climate change drive species out of reserves? An
assessment of existing reserve-selection methods. Glob Change Biol., 10: 1618-1626.
Araújo, M.B., Thuiller W. and Pearson, R.G. (2006) Climate warming and the decline of amphibians and
reptiles in Europe. J. Biogeogr., 33: 1712–1728.
Benjamini, Y. and Hochberg Y. (1995) Controlling the false discovery rate: a practical 8 and powerful
approach to multiple testing. Journal of the Royal Statistical Society Series B, 57: 289-300.
Farr and Kobrick (2000) Shuttle Radar Topography Mission produces a wealth of data.� EOS Trans. Am.
Geophys. Union, 81: 583-585.
Franco A. y Rodríguez M (2001) Libro Rojo de los Vertebrados Amenazados de Andalucía. Consejería de
Medio Ambiente. Junta de Andalucía.
García, L.V. (2003) Controlling the false discovery rate in ecological research. Trends Ecol. Evol., 18: 553–
554.
Gaston, K. J. (2000). Global patterns in biodiversity. Nature, 40: 220-227.
Guisan, A., & Zimmermann, N. E. (2000). Predicitive habitat distribution models in ecology. Ecological
Modelling, 135: 147 - 186.
I.G.N. (1999) Mapa de carreteras. Península Ibérica, Baleares y Canarias. Inst. Geográfico Nacional/Ministerio
de Fomento, Madrid.
Jackson, R.B., Carpenter, S.R., Dahm, C.N. et al. (2001) Water in a changing World. Ecol Appl., 11: 1027-
1045.
Legendre P, Legendre L (1998) Numerical ecology, 2nd edn. Elsevier, Amsterdam
Márquez A.L., Real, R., Olivero, J. and Estrada, A. (2011) Combining climate with other influential factors for
modelling climate change impact on species distribution. Climate Change, 108:135-157.
Palomo, L. J. et al. 2007. Atlas y Libro Rojo de los Mamíferos Terrestres de España. Dirección General para la
Biodiversidad- SECEM-SECEMU, Madrid.
Pleguezuelos, J. M. et al. (eds) 2004. Atlas y libro rojo de los anfibios y reptiles de España. Dirección General
de Conservación de la Naturaleza-Asociación Herpetológica Española, Madrid, Spain.
ORNL (2001) LandScan 2000 Global Population Database._Oak Ridge National Laboratory (ORNL), Oak
Ridge, TN.
Real, R., Barbosa, A.M., Vargas, J.M. (2006) Obtaining environmental favourability functions from logistic
regression. Environ. Ecol. Stat., 13: 237-245.
70
Real, R., Márquez A.L., Olivero, J. and Estrada, A. (2010) Species distribution models in climate change
scenarios are still not useful for informing policy planning: an uncertainty assessment using fuzzy logic.
Ecography 33: 304-314.
US Geological Survey (1996) GTOPO30. Land Processes Distributed Active Archive Center (LP DAAC), EROS
Data Center. _ <http://edcdaac.usgs.gov/gtopo30/gtopo30.asp>, accessed 22 September 1999. Myers,
A., & Giller, P. S. (Eds.). (1988). Analytical biogeography: An integrated approach to the study of animal
and plant distributions. London: Chapman & Hall.
ANEJO Nº1.
Anexo 1. Modelo de regresión logística obtenido para cada especie y para cada aproximación nacional, regional y actualizada. Se representa el coeficiente de cada variable en la función logit (B), el error típico asociado a B (ET), el parámetro obtenido para cada variable en el test de Wald y su significación. Abreviaturas de las variables como en la tabla 1. Modelo nacional Modelo regional Modelo actualizado
Variables en la ecuación Variables en la ecuación Variables en la ecuación
B E.T. Wald Sig. B E.T. Wald Sig. B E.T. Wald Sig.pend 0.205 0.05 19.97 0.00 pend 0.828 0.21 15.02 0.00 pend 0.283 0.05 28.50 0.00
alti2 -0.0000027 0.00 28.62 0.00 p14c2 -0.477 0.17 7.81 0.01 alti2 -0.0000024 0.00 21.02 0.00
alti 0.0077 0.00 41.69 0.00 pend2 -0.029 0.01 7.97 0.00 alti 0.0066 0.00 27.34 0.00
p15c2 -0.616 0.06 94.87 0.00 la2lop 0.014 0.00 31.10 0.00 p15c2 -0.152 0.12 1.53 0.22
tx14c2 -0.241 0.10 5.29 0.02 p16c2 0.223 0.04 33.48 0.00 tx14c2 -0.486 0.13 12.97 0.00
p13c2 0.168 0.05 10.48 0.00 u500 -0.181 0.03 33.11 0.00 p13c2 -0.190 0.09 4.47 0.03
p16c2 0.149 0.03 25.35 0.00 u100 0.0305 0.01 4.99 0.03 p16c2 0.220 0.06 13.78 0.00
lop 0.783 0.38 4.18 0.04 Cte -117.23 19.85 34.89 0.00 lop 3.18 0.65 24.17 0.00
daut 0.031 0.01 12.91 0.00 daut -0.0160 0.01 1.57 0.21
alyd
ic_c
2
Cte -60.02 3.73 2.60 0.11 Cte -14.51 4.55 10.17 0.00
alti 0.002 0.00 36.31 0.00 p15c2 0.638 0.07 93.64 0.00 alti 0.0031 0.00 4.58 0.03
alti2 -0.00000046 0.00 5.96 0.01 p13c2 0.331 0.04 58.28 0.00 alti2 -0.00000062 0.00 0.59 0.44
pend 0.758 0.05 213.54 0.00 p16c2 -0.221 0.03 52.84 0.00 pend 1.15 0.19 38.29 0.00
pend2 -0.034 0.00 110.46 0.00 lalop -0.298 0.09 11.04 0.00 pend2 -0.042 0.01 17.17 0.00
orS 0.010 0.00 10.36 0.00 lalop2 0.0705 0.02 14.22 0.00 orS 0.0120 0.01 2.06 0.15
lop2 -1.594 0.12 168.03 0.00 lop3 -0.193 0.05 16.62 0.00 lop2 1.20 2.82 0.18 0.67
lalop2 0.042 0.00 196.73 0.00 orS 0.0214 0.01 9.38 0.00 lalop2 -0.052 0.07 0.51 0.48
la2lop -0.005 0.00 76.19 0.00 Cte -4.99 5.23 0.91 0.34 la2lop 0.023 0.01 3.11 0.08
lalop 0.156 0.02 44.64 0.00 lalop -0.786 0.51 2.40 0.12
sals
al_c
2
daut 0.010 0.00 20.16 0.00 daut -0.060 0.01 31.01 0.00
u500 0.003 0.00 7.61 0.01 u500 0.014 0.01 4.87 0.03
pobl 0.00049 0.00 15.14 0.00 pobl 0.000605 0.00 3.38 0.07
p13c2 0.163 0.01 257.97 0.00 p13c2 0.468 0.15 10.24 0.00
p14c2 -0.151 0.01 181.24 0.00 p14c2 0.535 0.21 6.48 0.01
p16c2 -0.055 0.00 179.32 0.00 p16c2 -0.138 0.04 11.09 0.00
Cte -0.835 0.66 1.61 0.20 Cte -12.66 3.00 17.81 0.00alti 0.0035 0.00 23.99 0.00 p16c2 0.159 0.02 44.47 0.00 alti 0.0071 0.00 21.79 0.00
alti2 -0.0000012 0.00 7.23 0.01 p15c2 -0.296 0.08 13.38 0.00 alti2 -0.0000031 0.00 7.86 0.01
orS 0.017 0.00 13.95 0.00 lop2 -6.65 0.75 77.95 0.00 orS 0.013 0.01 3.42 0.06
lop -156.53 20.12 60.51 0.00 lalop2 0.180 0.02 76.95 0.00 lop 295.71 121.62 5.91 0.02
la2lop -0.091 0.01 50.52 0.00 p14c2 -0.855 0.15 33.93 0.00 la2lop 0.251 0.09 8.58 0.00
lalop 7.58 1.02 55.74 0.00 alti 0.0068 0.00 22.93 0.00 lalop -17.32 6.46 7.20 0.01
la3 -0.00045 0.00 61.44 0.00 alti2 -0.0000031 0.00 8.93 0.00 la3 -0.0019 0.00 25.92 0.00
u100 -0.0075 0.00 9.30 0.00 tx13c2 3.19 0.81 15.65 0.00 u100 -0.0120 0.01 4.18 0.04
p13c2 0.310 0.04 65.29 0.00 tx16c2 1.30 0.23 32.27 0.00 p13c2 -0.177 0.11 2.72 0.10
p14c2 -0.547 0.06 94.67 0.00 tx15c2 -5.26 0.89 35.23 0.00 p14c2 -0.170 0.14 1.56 0.21
p16c2 -0.050 0.01 19.89 0.00 Cte 41.98 7.16 34.38 0.00 p16c2 0.085 0.03 6.51 0.01
Trip
yg_c
2
Cte 17.92 3.29 29.71 0.00 Cte 97.30 19.74 24.31 0.00
alti 0.00803 0.00 30.56 0.00 alti 0.0062 0.00 17.99 0.00 alti 0.0064 0.00 15.88 0.00
alti2 -0.0000030 0.00 27.60 0.00 alti2 -0.0000025 0.00 17.26 0.00 alti2 -0.0000025 0.00 16.81 0.00
pend 0.218 0.05 16.72 0.00 pend 0.281 0.06 20.59 0.00 pend 0.255 0.07 14.59 0.00
lop 4.46 0.84 28.41 0.00 p14c4 -0.653 0.17 15.65 0.00 lop 14.11 2.63 28.68 0.00
la2 -10.09 0.31 12.29 0.00 la2lop 0.005 0.00 5.16 0.02 la2 -1.51 1.30 1.35 0.24
la 78.62 23.45 11.24 0.00 lop2 0.614 0.22 7.68 0.01 la 115.53 97.09 1.42 0.23
daut 0.049 0.01 17.50 0.00 p15c4 0.923 0.15 36.97 0.00 daut 0.0098 0.02 0.29 0.59
alyd
ic_c
4
u500 -0.012 0.00 7.85 0.01 u500 -0.108 0.02 35.61 0.00 u500 -0.096 0.03 11.87 0.00
p14c4 -0.315 0.10 10.79 0.00 Cte -79.28 14.21 31.12 0.00 p14c4 -0.485 0.16 9.23 0.00
p15c4 0.357 0.08 19.29 0.00 p15c4 0.897 0.16 32.39 0.00
tx16c4 -0.608 0.14 17.64 0.00 tx16c4 0.426 0.26 2.63 0.10
Cte -1453.57 444.31 10.70 0.00 Cte -2309.00 1817.23 1.61 0.20
alti 0.0024 0.00 38.29 0.00 p15c4 0.342 0.08 18.28 0.00 alti 0.0039 0.00 7.65 0.01
alti2 -0.00000048 0.00 6.64 0.01 p13c4 0.477 0.06 68.57 0.00 alti2 -0.00000086 0.00 1.51 0.22
pend 0.745 0.05 202.02 0.00 p16c4 -0.202 0.03 42.51 0.00 pend 1.09 0.19 34.47 0.00
pend2 -0.034 0.00 105.93 0.00 lop3 -0.012 0.00 14.75 0.00 pend2 -0.037 0.01 14.18 0.00
orS 0.0101 0.00 10.31 0.00 la3 0.00037 0.00 19.75 0.00 orS 0.0073 0.01 0.76 0.38
lop2 -1.49 0.14 114.77 0.00 orS 0.022 0.01 9.57 0.00 lop2 -6.26 3.52 3.15 0.08
lalop2 0.039 0.00 140.92 0.00 Cte -34.77 4.44 61.46 0.00 lalop2 0.149 0.09 2.60 0.11
la2lop -0.0042 0.00 38.69 0.00 la2lop -0.017 0.02 0.97 0.32
lalop 0.128 0.03 19.36 0.00 lalop 0.739 0.67 1.22 0.27
daut 0.010 0.00 20.29 0.00 daut -0.052 0.01 23.47 0.00
u500 0.003 0.00 7.88 0.00 u500 0.022 0.01 11.32 0.00
pobl 0.00049 0.00 15.65 0.00 pobl 0.000604 0.00 3.80 0.05
p13c4 0.168 0.01 261.35 0.00 p13c4 0.327 0.15 4.81 0.03
p14c4 -0.140 0.01 114.84 0.00 p14c4 0.119 0.22 0.29 0.59
p15c4 -0.031 0.02 3.37 0.07 p15c4 0.670 0.18 14.40 0.00
p16c4 -0.045 0.01 35.37 0.00 p16c4 -0.268 0.05 25.29 0.00
sals
al_c
4
Cte -0.570 0.70 0.65 0.42 Cte -23.87 4.40 29.47 0.00
alti 0.0034 0.00 23.49 0.00 p16c4 0.044 0.01 67.95 0.00 alti 0.0072 0.00 22.71 0.00
alti2 -0.0000011 0.00 6.82 0.01 p14c4 -0.302 0.07 17.54 0.00 alti2 -0.0000033 0.00 8.44 0.00
orS 0.017 0.00 14.18 0.00 alti 0.0053 0.00 20.59 0.00 orS 0.013 0.01 3.46 0.06
lop -155.28 20.12 59.57 0.00 alti2 -0.0000027 0.00 10.36 0.00 lop 350.69 126.78 7.65 0.01trip
yg_c
4
la2lop -0.091 0.01 49.86 0.00 u500 0.018 0.00 38.73 0.00 la2lop 0.291 0.09 10.52 0.00
lalop 7.53 1.02 54.94 0.00 tx13c4 3.37 0.64 28.00 0.00 lalop -20.26 6.74 9.05 0.00
la3 -0.00044 0.00 61.44 0.00 tx15c4 -3.83 0.70 29.86 0.00 la3 -0.002 0.00 28.06 0.00
u100 -0.0071 0.00 8.59 0.00 Cte 18.38 5.20 12.51 0.00 u100 -0.013 0.01 4.55 0.03
p13c4 0.299 0.04 61.85 0.00 p13c4 -0.234 0.11 4.39 0.04
p14c4 -0.540 0.06 91.89 0.00 p14c4 -0.118 0.14 0.75 0.39
p16c4 -0.046 0.01 17.29 0.00 p16c4 0.101 0.03 8.53 0.00
Cte 17.16 3.21 28.52 0.00 Cte 101.69 19.76 26.48 0.00
alti 0.010 0.00 40.79 0.00 alti 0.0053 0.00 14.91 0.00 alti 0.0103 0.00 30.77 0.00
alti2 -0.0000035 0.00 32.44 0.00 alti2 -0.0000022 0.00 14.91 0.00 alti2 -0.0000035 0.00 27.57 0.00
pend 0.246 0.06 14.99 0.00 pend 0.255 0.06 18.04 0.00 pend 0.229 0.07 12.04 0.00
la2 -3.73 0.56 43.78 0.00 tx15e2 0.909 0.35 6.81 0.01 la2 -4.50 1.76 6.51 0.01
la 276.67 41.96 43.48 0.00 lalop 0.377 0.09 19.24 0.00 la 331.85 134.04 6.13 0.01
daut 0.029 0.01 4.39 0.04 lop2 -0.332 0.21 2.40 0.12 daut 0.0096 0.02 0.29 0.59
p13e2 1.88 0.35 28.65 0.00 u500 -0.089 0.01 38.39 0.00 p13e2 1.75 0.36 23.04 0.00
p14e2 -0.622 0.27 5.31 0.02 p16e2 0.268 0.05 31.31 0.00 p14e2 0.307 0.46 0.44 0.51
p16e2 -0.946 0.18 26.24 0.00 Cte -105.92 18.92 31.35 0.00 p16e2 -0.9564 0.20 22.93 0.00
tx13e2 13.53 2.87 22.23 0.00 tx13e2 14.20 3.61 15.44 0.00
tx14e2 30.08 1.13 7.42 0.01 tx14e2 3.62 1.25 8.42 0.00
tx15e2 -31.93 5.33 35.94 0.00 tx15e2 -33.93 5.84 33.79 0.00
tx16e2 14.19 2.65 28.68 0.00 tx16e2 16.23 2.93 30.63 0.00
Alyd
ic_e
2
Cte -5012.77 767.58 42.65 0.00 Cte -6024.75 2538.45 5.63 0.02
alti 0.0030 0.00 58.99 0.00 p15e2 -0.931 0.12 58.57 0.00 alti 0.0042 0.00 9.21 0.00
alti2 -0.00000066 0.00 11.62 0.00 p16e2 0.314 0.05 33.64 0.00 alti2 -0.0000013 0.00 2.39 0.12
pend 0.587 0.05 119.87 0.00 p13e2 0.563 0.13 18.61 0.00 pend 0.709 0.17 16.81 0.00
pend2 -0.027 0.00 67.86 0.00 lop 10.07 4.00 6.35 0.01 pend2 -0.022 0.01 5.26 0.02sals
al_e
2
orS 0.0093 0.00 8.47 0.00 la2lop -0.0080 0.00 7.77 0.01 orS 0.018 0.01 4.99 0.03
lop2 -0.906 0.05 342.03 0.00 p14e2 0.580 0.16 12.59 0.00 lop2 -1.95 0.85 5.23 0.02
lalop2 0.027 0.00 425.88 0.00 la 5.15 1.07 23.08 0.00 lalop2 0.040 0.02 3.07 0.08
la2lop -0.002 0.00 258.93 0.00 orS 0.024 0.01 11.40 0.00 la2lop 0.0007 0.00 1.94 0.16
daut 0.0084 0.00 16.77 0.00 u100 -0.012 0.01 4.89 0.03 daut -0.035 0.01 15.47 0.00
pobl 0.00049 0.00 16.12 0.00 Cte -198.82 40.70 23.86 0.00 pobl 0.0005 0.00 2.77 0.10
p13e2 0.228 0.01 298.70 0.00 p13e2 0.720 0.12 39.14 0.00
p14e2 -0.145 0.02 86.83 0.00 p14e2 0.474 0.19 6.56 0.01
p15e2 -.165 0.02 101.25 0.00 p15e2 -0.487 0.10 23.50 0.00
Cte 1.11 0.59 3.56 0.06 Cte -9.39 2.44 14.82 0.00
orS 0.015 0.00 12.15 0.00 lop3 -0.023 0.00 101.32 0.00 orS 0.0088 0.01 1.57 0.21
orW -0.013 0.01 4.37 0.04 tx15e2 -2.06 0.64 10.46 0.00 orW -0.0194 0.01 3.65 0.06
lop -640.01 12.21 27.50 0.00 tx13e2 1.20 0.49 6.03 0.01 lop -60.39 84.94 0.51 0.48
lop2 6.30 1.34 22.04 0.00 daut -0.028 0.01 18.27 0.00 lop2 0.586 8.54 0.00 0.95
lop3 -0.127 0.02 28.73 0.00 u500 0.017 0.00 36.62 0.00 lop3 -0.393 0.07 29.93 0.00
lalop2 -0.120 0.03 15.14 0.00 pend 0.108 0.04 7.57 0.01 lalop2 0.126 0.23 0.29 0.59
lalop 1.483 0.30 24.60 0.00 Cte 23.77 5.68 17.52 0.00 lalop 0.977 2.27 0.18 0.67
la2 0.657 0.10 40.37 0.00 la2 1.65 0.62 6.99 0.01
la3 -0.012 0.00 48.51 0.00 la3 -0.031 0.01 7.53 0.01
p16e2 0.065 0.01 87.69 0.00 p16e2 0.034 0.02 3.21 0.07
tx14e2 -0.381 0.10 14.92 0.00 tx14e2 -0.687 0.20 11.48 0.00
tx16e2 0.386 0.14 7.33 0.01 tx16e2 -0.324 0.26 1.57 0.21
trip
yg_e
2
Cte -267.33 53.67 24.81 0.00 Cte -631.41 292.06 4.67 0.03
pend .732 0.15 22.73 0.00 p14c2 0.896 0.22 17.13 0.00 pend 0.555 0.19 8.19 0.00
alti 0.0065 0.00 23.23 0.00 pend 0.584 0.16 13.18 0.00 alti 0.005 0.00 9.79 0.00
daut 0.0424 0.02 5.91 0.02 tx14c2 0.886 0.50 3.19 0.07 daut 0.078 0.06 1.67 0.20
la3 -0.0738 0.02 13.70 0.00 Cte -49.50 18.31 7.31 0.01 la3 -0.164 0.12 1.88 0.17algm
ar_c
2
la 330.65 88.20 14.05 0.00 la 715.99 522.82 1.88 0.17
Cte -8532.96 2258.62 14.27 0.00 Cte -18217.55 13297.35 1.88 0.17
alti 0.0023 0.00 83.46 0.00 pend 0.734 0.20 14.17 0.00 alti -0.044 0.26 0.03 0.87
pend 0.406 0.06 41.62 0.00 Cte -12.63 2.82 20.03 0.00 pend 60.11 184.78 0.11 0.74
pend2 -0.018 0.00 22.98 0.00 pend2 -1.25 4.44 0.08 0.78
lop 0.214 0.08 7.29 0.01 lop -189.55 616.04 0.09 0.76
la3 -0.0011 0.00 14.96 0.00 la3 -8.38 27.01 0.10 0.76
la 5.94 1.38 18.57 0.00 la 33851.23 112494.02 0.09 0.76
u500 -0.0031 0.00 6.30 0.01 u500 1.75 6.87 0.06 0.80
p13c2 0.037 0.01 37.94 0.00 p13c2 1.63 117.68 0.00 0.99
p14c2 -0.048 0.01 18.09 0.00 p14c2 -92.23 514.22 0.03 0.86
tx13c2 -0.728 0.16 20.60 0.00 tx13c2 360.93 1615.70 0.05 0.82
tx14c2 -0.437 0.07 37.00 0.00 tx14c2 93.23 760.22 0.02 0.90
tx15c2 1.221 0.20 38.25 0.00 tx15c2 -934.73 2539.43 0.14 0.71
cora
us_c
2
Cte -176.87 37.47 22.27 0.00 Cte -815457.93 2761439.06 0.09 0.77
alti -0.0018 0.00 25.33 0.00 p15c2 -0.266 0.10 7.37 0.01 alti 0.0029 0.00 10.89 0.00
orS 0.014 0.00 12.62 0.00 lop -42.44 7.57 31.40 0.00 orS 0.011 0.01 1.47 0.22
lop -124.76 21.54 33.56 0.00 la2lop 0.029 0.01 28.24 0.00 lop 16.46 346.23 0.00 0.96
la2lop -0.073 0.01 30.67 0.00 p13c2 0.360 0.10 12.56 0.00 la2lop 0.064 0.24 0.07 0.79
lalop 60.05 1.07 32.20 0.00 p14c2 -0.563 0.15 14.02 0.00 lalop -2.8972 18.24 0.03 0.87
la2 0.200 0.07 7.79 0.01 la -5.42 1.91 8.03 0.00 la2 -0.064 0.04 3.13 0.08
daut 0.013 0.00 22.79 0.00 pend 0.275 0.07 16.39 0.00 daut -0.0027 0.01 0.05 0.83
p13c2 0.091 0.02 19.77 0.00 Cte 207.60 70.98 8.55 0.00 p13c2 0.471 0.17 7.92 0.00
p14c2 -0.145 0.02 51.96 0.00 p14c2 -0.621 0.24 6.59 0.01
p15c2 -0.055 0.02 8.61 0.00 p15c2 -0.312 0.15 4.22 0.04
tx13c2 -1.40 0.24 35.47 0.00 tx13c2 0.785 1.09 0.52 0.47
emyo
rb_c
2
tx14c2 -0.229 0.07 9.46 0.00 tx14c2 0.399 0.70 0.32 0.57
tx15c2 1.81 0.30 37.78 0.00 tx15c2 -1.12 1.48 0.58 0.45
Cte 327.57 112.46 8.48 0.00 Cte 89.91 48.82 3.39 0.07
alti 0.0014 0.00 16.23 0.00 alti -0.051 5.49 0.00 0.99
pend 0.674 0.09 61.07 0.00 pend 112.07 2231.96 0.00 0.96
pend2 -0.034 0.01 35.90 0.00 pend2 -5.11 127.49 0.00 0.97
lop2 -0.110 0.01 109.94 0.00 lop2 -8.20 622.55 0.00 0.99
daut -0.008 0.00 5.21 0.02 daut -0.629 62.96 0.00 0.99
u100 0.011 0.00 15.49 0.00 u100 1.66 47.64 0.00 0.97
u500 -0.011 0.00 57.40 0.00 u500 1.16 99.16 0.00 0.99
p14c2 -0.043 0.02 7.11 0.01 p14c2 -13.43 1835.44 0.00 0.99
p15c2 0.210 0.03 63.01 0.00 p15c2 3.93 3508.48 0.00 1.00
p16c2 -0.083 0.01 59.42 0.00 p16c2 -4.27 1621.39 0.00 1.00
tx16c2 -0.173 0.05 10.47 0.00 tx16c2 -20.10 3115.03 0.00 0.99
lacs
ch_c
2
Cte -0.836 0.82 1.04 0.31 Cte 154.45 65128.90 0.00 1.00
la3 -0.058 0.02 6.25 0.01 p13c2 -0.372 0.07 29.30 0.00 la3 -0.115 0.06 3.33 0.07
la 232.19 96.58 5.78 0.02 tx13c2 9.63 2.53 14.45 0.00 la 462.82 256.70 3.25 0.07
daut 0.119 0.03 12.32 0.00 tx15c2 -8.99 2.56 12.34 0.00 daut 0.115 0.04 7.82 0.01
u500 0.060 0.01 16.29 0.00 Cte 29.39 13.30 4.88 0.03 u500 0.056 0.02 5.02 0.02
p13c2 -0.163 0.10 2.66 0.10 p13c2 -0.263 0.17 2.37 0.12
tx13c2 190.09 3.74 26.07 0.00 tx13c2 19.85 7.37 7.26 0.01
tx14c2 -80.04 1.59 25.52 0.00 tx14c2 -7.98 2.64 9.15 0.00
tx16c2 -11.55 2.42 22.71 0.00 tx16c2 -12.69 5.63 5.08 0.02
tesg
ra_c
2
Cte -5644.34 2398.67 5.54 0.02 Cte -11290.06 6307.21 3.20 0.07
alti 0.0041 0.00 69.01 0.00 alti 0.0031 0.00 60.33 0.00 alti 0.0031 0.00 5.33 0.02
alti2 -0.00000080 0.00 12.56 0.00 p16c2 0.0307 0.00 61.22 0.00 alti2 0.00000018 0.00 0.08 0.78
pend 0.269 0.06 22.97 0.00 tx14c2 -0.228 0.06 13.84 0.00 pend -0.0802 0.16 0.26 0.61
vipl
at_c
2
pend2 -0.013 0.00 13.82 0.00 tx13c2 0.494 0.12 16.38 0.00 pend2 0.0120 0.01 2.26 0.13
orS 0.010 0.00 11.34 0.00 orS 0.024 0.01 12.76 0.00 orS 0.0199 0.01 7.82 0.01
lop -31.22 2.79 124.91 0.00 Cte -9.80 2.58 14.47 0.00 lop -175.97 94.13 3.49 0.06
lop2 20.036 0.21 90.92 0.00 lop2 12.97 8.78 2.18 0.14lalop2 -0.050 0.01 93.05 0.00 lalop2 -0.345 0.23 2.17 0.14
lalop 0.788 0.07 131.45 0.00 lalop 4.7030 2.51 3.51 0.06
la3 -0.00052 0.00 114.04 0.00 la3 -0.0032 0.00 4.26 0.04
u100 0.0092 0.00 26.12 0.00 u100 0.0144 0.01 4.73 0.03
u500 -0.002 0.00 3.04 0.08 u500 0.0071 0.01 1.99 0.16
p14c2 -0.131 0.01 207.45 0.00 p14c2 -0.355 0.13 7.57 0.01
p16c2 0.022 0.00 48.15 0.00 p16c2 0.052 0.01 14.17 0.00
tx13c2 -0.702 0.07 102.10 0.00 tx13c2 -0.501 0.29 3.09 0.08
tx16c2 0.794 0.07 123.07 0.00 tx16c2 0.824 0.29 8.20 0.00
Cte 28.81 3.54 66.34 0.00 Cte 157.27 80.39 3.83 0.05
pend 0.732 0.15 22.73 0.00 p14c4 0.911 0.22 16.47 0.00 pend 0.555 0.19 8.19 0.00
alti 0.0065 0.00 23.23 0.00 pend 0.589 0.16 13.31 0.00 alti 0.0047 0.00 9.79 0.00
daut 0.042 0.02 5.91 0.02 tx14c4 0.980 0.52 3.50 0.06 daut 0.078 0.06 1.67 0.20
la3 -0.074 0.02 13.70 0.00 Cte -52.72 19.39 7.39 0.01 la3 -0.164 0.12 1.88 0.17
la 330.65 88.20 14.05 0.00 la 715.99 522.82 1.88 0.17algm
ar_c
4
Cte -8532.96 2258.62 14.27 0.00 Cte -18217.55 13297.35 1.88 0.17
p13c4 0.037 0.01 37.74 0.00 pend 0.734 0.20 14.17 0.00 p13c4 1.80 793.57 0.00 1.00
tx14c4 -0.425 0.07 35.62 0.00 Cte -12.63 2.82 20.03 0.00 tx14c4 82.03 4773.21 0.00 0.99
p14c4 -0.048 0.01 17.66 0.00 p14c4 -83.89 3234.95 0.00 0.98
la3 -0.0011 0.00 14.78 0.00 la3 -7.27 126.81 0.00 0.95
pend 0.406 0.06 41.53 0.00 pend 54.03 484.75 0.01 0.91
la 5.94 1.38 18.43 0.00 la 29380.04 545387.36 0.00 0.96
tx13c4 -0.75 0.16 21.29 0.00 tx13c4 335.73 11184.87 0.00 0.98
cora
us_c
4
tx15c4 1.23 0.20 38.84 0.00 tx15c4 -837.16 9079.38 0.01 0.93
pend2 -0.017 0.00 22.85 0.00 pend2 -1.0956 10.25 0.01 0.91
alti 0.0023 0.00 82.81 0.00 alti -0.047 1.19 0.00 0.97
u500 -0.0031 0.00 6.33 0.01 u500 1.7737 11.13 0.03 0.87
lop 0.207 0.08 6.75 0.01 lop -150.28 916.05 0.03 0.87 Cte -177.14 37.61 22.19 0.00 Cte -707869.43 13617427.23 0.00 0.96
alti -0.0015 0.00 18.27 0.00 lop -39.56 6.62 35.72 0.00 alti 0.0024 0.00 7.95 0.00
orS 0.015 0.00 14.74 0.00 la2lop 0.027 0.00 33.91 0.00 orS 0.014 0.01 2.16 0.14
lop -59.37 8.13 53.35 0.00 p14c4 -0.220 0.09 6.60 0.01 lop -110.10 387.43 0.08 0.78
lalop2 0.00081 0.00 8.01 0.00 la3 0.0077 0.00 7.38 0.01 lalop2 -0.0030 0.01 0.25 0.62
la2lop -0.034 0.00 48.37 0.00 la -39.46 11.79 11.19 0.00 la2lop -0.0520 0.27 0.04 0.85
lalop 2.83 0.40 50.85 0.00 pend 0.346 0.07 27.80 0.00 lalop 4.86 20.37 0.06 0.81
daut 0.015 0.00 23.73 0.00 Cte 1080.59 295.20 13.40 0.00 daut -0.012 0.01 1.00 0.32
u500 -0.0037 0.00 10.10 0.00 u500 0.0105 0.01 1.50 0.22
p13c4 0.088 0.02 21.05 0.00 p13c4 0.471 0.18 7.14 0.01
p14c4 -.138 0.02 48.27 0.00 p14c4 -0.479 0.28 3.02 0.08
p15c4 -0.065 0.02 12.31 0.00 p15c4 -0.365 0.16 5.07 0.02
tx13c4 -1.31 0.22 34.75 0.00 tx13c4 1.00 1.19 0.71 0.40
tx14c4 -.233 0.07 10.24 0.00 tx14c4 0.180 0.65 0.08 0.78
tx15c4 1.76 0.28 38.37 0.00 tx15c4 -1.30 1.60 0.66 0.42
emyo
rb-c
4
Cte -8.70 2.19 15.80 0.00 Cte 4.74 10.33 0.21 0.65
alti 0.0014 0.00 15.82 0.00 alti -0.052 4.69 0.00 0.99
pend 0.677 0.09 61.49 0.00 pend 111.65 1672.01 0.00 0.95
pend2 -0.034 0.01 35.87 0.00 pend2 -5.10 99.97 0.00 0.96
lop2 -0.111 0.01 112.06 0.00 lop2 -8.55 296.83 0.00 0.98
daut -0.0083 0.00 5.36 0.02 daut -0.655 28.40 0.00 0.98
u100 0.011 0.00 16.39 0.00 u100 1.67 59.33 0.00 0.98
lacs
ch_c
4
u500 -0.012 0.00 58.60 0.00 u500 1.10 28.77 0.00 0.97
p14c4 -0.038 0.02 5.74 0.02 p14c4 -12.08 812.81 0.00 0.99
p15c4 0.198 0.03 61.80 0.00 p15c4 4.87 712.09 0.00 0.99
p16c4 -0.079 0.01 57.79 0.00 p16c4 -4.82 290.51 0.00 0.99
tx16c4 -0.166 0.05 9.60 0.00 tx16c4 -20.65 1971.95 0.00 0.99 Cte -0.789 0.82 0.93 0.34 Cte 179.41 28043.73 0.00 0.99
tx16c4 -11.74 2.47 22.57 0.00 p13c4 -0.406 0.08 27.37 0.00 tx16c4 -12.49 5.45 5.25 0.02
la 470.32 190.21 6.11 0.01 tx13c4 10.44 2.77 14.17 0.00 la 906.38 507.96 3.18 0.07
la2 -6.42 2.55 6.34 0.01 tx15c4 -10.03 2.87 12.22 0.00 la2 -12.35 6.88 3.22 0.07
tx13c4 19.41 3.82 25.84 0.00 Cte 38.57 15.75 6.00 0.01 tx13c4 19.71 7.22 7.45 0.01
u500 0.061 0.01 16.91 0.00 u500 0.055 0.02 4.86 0.03
p13c4 -0.168 0.10 2.84 0.09 p13c4 -0.261 0.17 2.41 0.12
daut 0.119 0.03 12.30 0.00 daut 0.117 0.04 8.06 0.00
tx14c4 -8.36 1.66 25.28 0.00 tx14c4 -8.15 2.69 9.17 0.00
tesg
ea_c
4
Cte -85870.04 3545.49 5.87 0.02 Cte -16602.72 9367.14 3.14 0.08
alti 0.0041 0.00 68.34 0.00 alti 0.0031 0.00 60.63 0.00 alti 0.0032 0.00 5.53 0.02
alti2 -0.00000079 0.00 12.33 0.00 p16c4 0.031 0.00 61.37 0.00 alti2 0.00000017 0.00 0.07 0.79
pend 0.269 0.06 23.04 0.00 tx14c4 -0.235 0.06 14.60 0.00 pend -0.077 0.16 0.24 0.62
pend2 -0.013 0.00 13.86 0.00 tx13c4 0.506 0.12 16.91 0.00 pend2 0.012 0.01 2.15 0.14
orS 0.0103 0.00 11.40 0.00 orS 0.024 0.01 12.83 0.00 orS 0.0201 0.01 7.96 0.00
lop -31.55 2.80 127.15 0.00 Cte -9.81 2.57 14.62 0.00 lop -180.70 93.71 3.72 0.05
lop2 20.068 0.21 93.22 0.00 lop2 13.46 8.72 2.38 0.12
lalop2 -0.051 0.01 95.37 0.00 lalop2 -0.358 0.23 2.37 0.12
lalop 0.796 0.07 133.64 0.00 lalop 4.8284 2.50 3.73 0.05
la3 -0.00052 0.00 116.47 0.00 la3 -0.0032 0.00 4.48 0.03
u100 0.0092 0.00 26.29 0.00 u100 0.015 0.01 4.87 0.03
u500 -0.0019 0.00 3.18 0.07 u500 0.0072 0.01 2.03 0.15
vipl
at_c
4
p14c4 -0.130 0.01 206.05 0.00 p14c4 -0.346 0.13 7.56 0.01
p16c4 0.022 0.00 47.26 0.00 p16c4 0.051 0.01 14.18 0.00
tx13c4 -0.696 0.07 100.61 0.00 tx13c4 -0.508 0.28 3.19 0.07
tx16c4 0.787 0.07 121.11 0.00 tx16c4 0.854 0.29 8.73 0.00
Cte 29.17 3.53 68.29 0.00 Cte 160.71 80.14 4.02 0.04
pend 0.732 0.15 22.73 0.00 p14e2 0.988 0.19 26.10 0.00 pend 0.555 0.19 8.19 0.00
alti 0.0065 0.00 23.23 0.00 pend 0.578 0.15 14.19 0.00 alti 0.0047 0.00 9.79 0.00
daut 0.042 0.02 5.91 0.02 Cte -19.60 3.76 27.17 0.00 daut 0.078 0.06 1.67 0.20
la3 -0.074 0.02 13.70 0.00 la3 -0.164 0.12 1.88 0.17
la 330.65 88.20 14.05 0.00 la 715.99 522.82 1.88 0.17algm
ar_e
2
Cte -8532.96 2258.62 14.27 0.00 Cte -18217.55 13297.35 1.88 0.17
p14e2 -0.026 0.01 23.79 0.00 pend 0.734 0.20 14.17 0.00 p14e2 23.32 24.67 0.89 0.34
p15e2 0.029 0.00 108.14 0.00 Cte -12.63 2.82 20.03 0.00 p15e2 -1.88 1.93 0.95 0.33
tx13e2 -1.61 0.22 55.96 0.00 tx13e2 72.68 72.19 1.01 0.31
la3 -0.0012 0.00 17.25 0.00 la3 -1.35 1.23 1.20 0.27
pend 0.413 0.06 43.10 0.00 pend 6.23 7.00 0.79 0.37
la 6.49 1.40 21.42 0.00 la 5382.71 4930.55 1.19 0.27
tx15e2 1.66 0.23 52.29 0.00 tx15e2 -86.40 87.15 0.98 0.32
pend2 -0.019 0.00 28.40 0.00 pend2 -0.136 0.19 0.52 0.47
alti 0.0022 0.00 81.37 0.00 alti -0.0075 0.01 0.78 0.38
u500 -0.0055 0.00 22.36 0.00 u500 0.202 0.30 0.45 0.50
cora
us_e
2
Cte -1940.06 38.67 25.19 0.00 Cte -130587.44 119741.26 1.19 0.28
alti -0.0027 0.00 71.52 0.00 lalop -2.65 0.38 47.77 0.00 alti -0.000404 0.00 0.36 0.55
orS 0.014 0.00 13.86 0.00 la2lop 0.0697 0.01 47.00 0.00 orS 0.0280 0.01 12.10 0.00
la3 -0.00065 0.00 10.72 0.00 p15e2 -0.266 0.13 4.37 0.04 la3 0.0035 0.00 1.27 0.26
la 3.51 0.91 14.90 0.00 p16e2 0.203 0.06 11.45 0.00 la -13.27 13.09 1.03 0.31
daut 0.021 0.00 58.71 0.00 la3 -0.0017 0.00 26.09 0.00 daut 0.031 0.01 15.16 0.00emyo
rb_e
2
u500 -0.0061 0.00 34.20 0.00 pend 0.274 0.07 17.12 0.00 u500 0.0016 0.00 0.24 0.62
p14e2 -0.047 0.01 19.03 0.00 p14e2 -0.522 0.24 4.86 0.03 p14e2 0.102 0.10 1.01 0.31
tx13e2 -20.06 0.35 34.61 0.00 Cte 93.97 17.99 27.29 0.00 tx13e2 -1.53 1.37 1.24 0.27
tx14e2 10.02 0.18 31.36 0.00 tx14e2 0.755 0.71 1.13 0.29
tx16e2 1.21 0.23 28.54 0.00 tx16e2 1.92 0.90 4.56 0.03 Cte -105.32 23.45 20.18 0.00 Cte 285.12 325.19 0.77 0.38
alti 0.0012 0.00 10.64 0.00 alti -0.084 10.04 0.00 0.99
pend 0.611 0.09 48.11 0.00 pend 106.69 5376.93 0.00 0.98
pend2 -0.030 0.01 26.47 0.00 pend2 -5.02 424.91 0.00 0.99
lop2 0.834 0.12 44.96 0.00 lop2 30.34 6556.83 0.00 1.00
lop3 -0.073 0.01 55.25 0.00 lop3 -0.341 488.85 0.00 1.00
la2lop -0.0019 0.00 41.49 0.00 la2lop -0.242 19.97 0.00 0.99
la2 -0.360 0.05 48.45 0.00 la2 168.83 13888.33 0.00 0.99
la 30.45 4.16 53.64 0.00 la -12641.44 1092765.68 0.00 0.99
daut -0.018 0.00 17.99 0.00 daut -1.62 200.98 0.00 0.99
p13e2 0.055 0.01 56.41 0.00 p13e2 -10.15 510.71 0.00 0.98
tx13e2 -30.08 0.54 31.89 0.00 tx13e2 88.29 31235.24 0.00 1.00
tx14e2 0.517 0.19 7.62 0.01 tx14e2 64.73 21230.99 0.00 1.00
tx15e2 2.62 0.44 34.63 0.00 tx15e2 -189.32 18370.64 0.00 0.99
lacs
ch_e
2
Cte -652.25 82.79 62.07 0.00 Cte 237930.27 21027232.02 0.00 0.99
la 2241.17 451.48 24.64 0.00 p13e2 -4.09 1.11 13.64 0.00 la 4916.23 1286.51 14.60 0.00
la2 -30.78 6.17 24.87 0.00 lop3 -0.0248 0.01 5.90 0.02 la2 -67.06 17.52 14.66 0.00
tx15e2 80.07 1.49 29.19 0.00 orW -0.0839 0.04 3.80 0.05 tx15e2 8.41 1.77 22.53 0.00
p14e2 12.93 2.47 27.51 0.00 p14e2 19.98 5.21 14.71 0.00 p14e2 16.64 3.89 18.32 0.00
p13e2 -2.83 0.50 32.64 0.00 tx16e2 8.33 1.88 19.71 0.00 p13e2 -3.31 0.71 21.70 0.00
Cte -40993.52 8288.10 24.46 0.00 la3 -0.684 0.18 13.87 0.00 Cte -90309.74 23657.97 14.57 0.00
la 2766.93 743.25 13.86 0.00
tesg
ra_e
2
Cte -67857.38 18223.41 13.87 0.00
alti 0.0029 0.00 43.42 0.00 alti 0.0017 0.00 20.91 0.00 alti 0.0012 0.00 0.98 0.32
alti2 -0.00000053 0.00 6.29 0.01 la2lop -0.00027 0.00 13.59 0.00 alti2 0.00000056 0.00 0.81 0.37
pend 0.434 0.05 64.95 0.00 orS 0.0206 0.01 9.95 0.00 pend 0.016 0.15 0.01 0.92
pend2 -0.0195 0.00 33.40 0.00 p15e2 0.0567 0.02 13.83 0.00 pend2 0.0077 0.01 1.00 0.32
orS 0.0077 0.00 6.23 0.01 pend 0.113 0.03 10.50 0.00 orS 0.016 0.01 4.89 0.03
lop -380.07 3.06 154.49 0.00 Cte -4.66 1.06 19.37 0.00 lop -315.98 87.54 13.03 0.00
lop2 2.105 0.22 88.01 0.00 lop2 25.81 8.21 9.87 0.00
lop3 0.010 0.00 6.72 0.01 lop3 -0.051 0.04 1.77 0.18
lalop2 -0.058 0.01 81.63 0.00 lalop2 -0.666 0.22 9.35 0.00
lalop 10.00 0.08 150.25 0.00 lalop 8.31 2.33 12.71 0.00
la3 -0.001 0.00 175.28 0.00 la3 -0.0052 0.00 13.42 0.00
u100 0.0054 0.00 11.90 0.00 u100 0.014 0.01 4.33 0.04
p13e2 -0.137 0.02 35.18 0.00 p13e2 0.0051 0.13 0.00 0.97
p14e2 -0.123 0.02 64.66 0.00 p14e2 -0.688 0.21 10.57 0.00
p15e2 0.158 0.02 49.21 0.00 p15e2 0.190 0.10 3.69 0.05
tx14e2 -0.290 0.03 78.66 0.00 tx14e2 -0.345 0.12 8.48 0.00
vipl
at_e
2
Cte 47.18 3.69 163.23 0.00 Cte 285.76 75.34 14.39 0.00
alti -0.004 0.00 40.18 0.00 p13c2 -0.239 0.07 12.86 0.00 alti 0.0013 0.00 0.30 0.58
lop2 -10.22 1.57 42.53 0.00 la -3.19 1.71 3.46 0.06 lop2 -2.24 102.38 0.00 0.98
lalop2 0.217 0.03 41.29 0.00 Cte 120.55 63.29 3.63 0.06 lalop2 0.060 2.72 0.00 0.98
la2lop -0.071 0.01 42.92 0.00 la2lop -0.031 0.59 0.00 0.96
lalop 3.59 0.54 44.45 0.00 lalop 1.10 22.46 0.00 0.96
daut -0.0403 0.01 9.56 0.00 daut -0.131 0.10 1.73 0.19
u100 -0.014 0.00 8.03 0.00 u100 0.127 0.09 2.17 0.14
atea
lg_c
2
p14c2 -0.169 0.05 10.91 0.00 p14c2 -0.582 2.13 0.07 0.78
p15c2 0.181 0.08 4.58 0.03 p15c2 -0.778 1.12 0.48 0.49
tx14c2 -0.419 0.22 3.54 0.06 tx14c2 -0.525 3.58 0.02 0.88
Cte -133.11 20.75 41.16 0.00 Cte 34.63 273.06 0.02 0.90
alti 0.0034 0.00 41.55 0.00 pend 0.214 0.05 18.81 0.00 alti 0.0033 0.00 2.62 0.11
alti2 -0.00000088 0.00 12.30 0.00 tx15c2 -0.909 0.20 21.57 0.00 alti2 -0.0000011 0.00 2.01 0.16
lop2 -0.089 0.01 58.09 0.00 Cte 14.65 4.13 12.58 0.00 lop2 -0.385 0.29 1.80 0.18
la2lop 0.0029 0.00 97.24 0.00 la2lop -0.002 0.01 0.11 0.74
lalop -0.080 0.01 32.56 0.00 lalop 0.242 0.24 0.97 0.32
u100 -0.006 0.00 8.77 0.00 u100 -0.045 0.01 9.49 0.00
p14c2 -.122 0.01 110.27 0.00 p14c2 -0.144 0.26 0.30 0.59
p15c2 0.044 0.01 42.08 0.00 p15c2 0.235 0.14 2.88 0.09
tx13c2 -0.837 0.09 95.24 0.00 tx13c2 -1.53 0.80 3.62 0.06
tx16c2 0.835 0.08 105.85 0.00 tx16c2 0.375 0.48 0.61 0.43
neoa
no_c
2
Cte -3.18 1.12 8.13 0.00 Cte -8.71 14.83 0.35 0.56
alti 0.0013 0.00 29.65 0.00 alti 0.0055 0.00 23.79 0.00 alti 0.0024 0.00 16.63 0.00
pend 0.083 0.02 11.45 0.00 alti2 -0.0000014 0.00 9.29 0.00 pend 0.0067 0.05 0.02 0.90
lop 1.976 0.26 55.85 0.00 tx15c2 9.73 2.23 19.10 0.00 lop -2.40 2.14 1.26 0.26
lalop2 -0.006 0.00 129.45 0.00 tx14c2 -4.47 1.08 17.18 0.00 lalop2 0.0045 0.00 1.12 0.29
la3 -0.002 0.00 97.90 0.00 la -40.91 18.95 4.66 0.03 la3 0.0109 0.01 1.59 0.21
la 110.05 1.12 97.34 0.00 la3 0.0090 0.00 4.05 0.04 la -50.17 38.00 1.74 0.19
daut 0.0067 0.00 7.50 0.01 tx16c2 -5.68 1.18 23.15 0.00 daut -0.0061 0.01 0.24 0.63
u100 -0.0087 0.00 26.03 0.00 orS -0.0204 0.01 6.80 0.01 u100 -0.015 0.01 2.37 0.12
p13c2 -0.037 0.01 11.09 0.00 Cte 1039.91 474.38 4.81 0.03 p13c2 0.267 0.13 4.29 0.04
p15c2 0.107 0.02 26.24 0.00 p15c2 0.277 0.32 0.76 0.38
p16c2 -0.029 0.01 13.37 0.00 p16c2 -0.202 0.08 7.04 0.01
tx14c2 -0.939 0.18 27.82 0.00 tx14c2 -4.08 1.31 9.72 0.00
talo
cc_c
2
tx15c2 1.17 0.38 9.68 0.00 tx15c2 10.02 2.79 12.90 0.00
tx16c2 -0.335 0.20 2.82 0.09 tx16c2 -6.13 1.56 15.34 0.00
Cte -292.65 29.67 97.29 0.00 Cte 1275.67 961.20 1.76 0.18
alti 0.0083 0.00 121.88 0.00 la3 -0.461 0.12 13.84 0.00 alti -0.014 0.01 4.29 0.04
alti2 -0.0000027 0.00 52.58 0.00 la2 26.31 7.06 13.89 0.00 alti2 0.000018 0.00 9.97 0.00
lop2 -0.058 0.01 109.41 0.00 tx14c2 45.08 9.30 23.51 0.00 lop2 -0.326 0.20 2.58 0.11
la3 0.0042 0.00 142.82 0.00 u500 0.099 0.04 4.88 0.03 la3 -0.417 0.09 21.23 0.00
la -18.82 1.65 129.59 0.00 tx16c2 18.60 6.04 9.48 0.00 la 1837.52 397.57 21.36 0.00
p14c2 -0.047 0.01 19.91 0.00 u100 -0.216 0.05 16.83 0.00 p14c2 -1.98 0.60 10.88 0.00
tx14c2 0.812 0.13 40.79 0.00 tx15c2 -69.12 15.61 19.61 0.00 tx14c2 18.38 4.17 19.40 0.00
tx15c2 -0.894 0.15 35.23 0.00 orS 0.067 0.02 7.41 0.01 tx15c2 -16.49 3.64 20.56 0.00
canl
up_c
2
Cte 475.17 43.11 121.49 0.00 la2lop -0.0075 0.00 9.52 0.00 Cte -47096.22 10152.61 21.52 0.00
alti -0.0015 0.00 54.79 0.00 p13c2 0.415 0.06 46.30 0.00 alti -0.0018 0.00 8.85 0.00
pend 0.566 0.05 148.08 0.00 p16c2 -0.126 0.02 29.53 0.00 pend 0.460 0.12 13.64 0.00
pend2 -0.026 0.00 84.91 0.00 lop -15.78 1.59 97.95 0.00 pend2 -0.0144 0.01 4.33 0.04
lop -4.12 0.50 66.68 0.00 la2lop 0.0102 0.00 93.32 0.00 lop -30.90 11.28 7.50 0.01
lalop 0.095 0.01 61.72 0.00 alti -0.0021 0.00 17.47 0.00 lalop 0.776 0.30 6.50 0.01
la3 -0.000081 0.00 20.22 0.00 tx14c2 -0.289 0.08 12.03 0.00 la3 -0.00031 0.00 0.57 0.45
daut 0.0038 0.00 3.53 0.06 pend 0.2142 0.05 19.77 0.00 daut -0.0108 0.01 1.88 0.17
u100 0.0087 0.00 32.98 0.00 u500 -0.012 0.00 12.99 0.00 u100 0.0055 0.01 1.05 0.30
u500 0.0057 0.00 48.49 0.00 Cte 10.28 2.98 11.86 0.00 u500 -0.0088 0.00 6.12 0.01
p13c2 -0.042 0.01 69.16 0.00 p13c2 0.463 0.08 34.03 0.00
p15c2 0.038 0.01 9.28 0.00 p15c2 0.160 0.09 2.86 0.09
p16c2 0.011 0.01 4.20 0.04 p16c2 -0.195 0.03 45.35 0.00
lutlu
t_c2
tx16c2 -0.245 0.04 36.51 0.00 tx16c2 -0.172 0.16 1.14 0.29
Cte 8.17 1.48 30.37 0.00 Cte 19.08 21.64 0.78 0.38
alti 0.0039 0.00 6.58 0.01 tx13c2 14.41 2.70 28.47 0.00 alti 0.0035 0.00 5.69 0.02
lynp
ar_c
2
lop3 -0.0066 0.00 9.89 0.00 tx15c2 -15.26 3.07 24.65 0.00 lop3 -0.0030 0.00 0.96 0.33
u500 0.015 0.01 4.39 0.04 p14c2 0.322 0.17 3.79 0.05 u500 0.021 0.01 9.70 0.00
tx13c2 7.63 1.59 23.15 0.00 tx14c2 1.46 0.43 11.41 0.00 tx13c2 8.75 1.70 26.54 0.00
tx15c2 -5.30 1.45 13.42 0.00 u500 0.018 0.01 6.00 0.01 tx15c2 -6.97 1.56 19.88 0.00
Cte -30.84 11.20 7.58 0.01 Cte 19.78 14.23 1.93 0.16 Cte -14.50 11.42 1.61 0.20
alti 0.0043 0.00 85.56 0.00 la -11.83 2.74 18.63 0.00 alti 0.0037 0.00 1.14 0.29
alti2 -0.0000020 0.00 88.46 0.00 pend 0.449 0.07 39.69 0.00 alti2 -0.0000025 0.00 0.89 0.35
pend 0.601 0.05 127.15 0.00 tx16c2 0.735 0.31 5.64 0.02 pend 0.490 0.26 3.56 0.06
pend2 -0.027 0.00 63.50 0.00 lop3 -0.074 0.02 19.59 0.00 pend2 -0.0049 0.01 0.13 0.72
orS 0.0055 0.00 3.42 0.06 la2lop 0.266 0.09 9.01 0.00 orS -0.0036 0.01 0.08 0.77
lalop2 -0.0024 0.00 95.80 0.00 lalop -17.14 6.94 6.10 0.01 lalop2 -0.035 0.01 9.75 0.00
la2lop 0.0064 0.00 242.87 0.00 lop 274.71 135.69 4.10 0.04 la2lop 0.101 0.03 11.92 0.00
lalop -0.221 0.02 195.21 0.00 Cte 411.51 99.62 17.06 0.00 lalop -3.37 1.01 11.04 0.00
la -1.11 0.13 68.05 0.00 la -18.55 5.77 10.35 0.00
daut -0.009 0.00 16.79 0.00 daut -0.0069 0.02 0.19 0.66
u100 0.019 0.00 124.26 0.00 u100 0.0026 0.01 0.05 0.83
u500 0.0065 0.00 39.93 0.00 u500 -0.0085 0.02 0.29 0.59
p14c2 -0.110 0.01 80.42 0.00 p14c2 -0.354 0.34 1.08 0.30
p15c2 0.144 0.02 47.08 0.00 p15c2 0.565 0.40 1.96 0.16
p16c2 -0.037 0.01 22.22 0.00 p16c2 -0.106 0.12 0.75 0.39
tx14c2 -1.15 0.14 66.11 0.00 tx14c2 4.40 2.11 4.34 0.04
tx15c2 1.94 0.32 37.59 0.00 tx15c2 -9.93 4.73 4.41 0.04
tx16c2 -0.968 0.17 31.74 0.00 tx16c2 5.85 2.44 5.77 0.02
capc
ap_c
2
Cte 35.15 5.81 36.61 0.00 Cte 660.44 209.86 9.90 0.00
alti 0.0027 0.00 48.98 0.00 p14c2 -0.234 0.13 3.11 0.08 alti 0.0044 0.00 8.00 0.00
alti2 -0.0000011 0.00 42.08 0.00 pend 0.184 0.04 17.24 0.00 alti2 -0.0000019 0.00 10.83 0.00
pend 0.250 0.02 151.52 0.00 la2lop 0.0067 0.00 31.50 0.00 pend 0.250 0.07 13.75 0.00
sciv
ul_c
2
orS 0.010 0.00 13.89 0.00 u500 -0.087 0.01 36.59 0.00 orS -0.0095 0.01 0.98 0.32
lop 1.44 0.32 20.29 0.00 p13c2 0.323 0.06 26.12 0.00 lop 26.44 54.72 0.23 0.63
la2lop 0.000 0.00 6.00 0.01 Cte -61.71 10.09 37.42 0.00 la2lop -0.0089 0.04 0.05 0.82
la 0.609 0.12 25.13 0.00 la 7.0959 23.56 0.09 0.76
u100 -0.0071 0.00 18.00 0.00 u100 0.0185 0.01 1.59 0.21
u500 0.0023 0.00 6.95 0.01 u500 -0.134 0.05 6.18 0.01
pobl 0.00048 0.00 14.93 0.00 pobl 0.0018 0.00 5.64 0.02
p14c2 -0.029 0.01 15.75 0.00 p14c2 -0.092 0.26 0.13 0.72
p15c2 0.034 0.00 63.67 0.00 p15c2 0.538 0.14 15.61 0.00
tx13c2 -0.645 0.12 27.76 0.00 tx13c2 -1.6516 1.30 1.62 0.20
tx14c2 -0.230 0.05 20.56 0.00 tx14c2 0.246 0.72 0.12 0.73
tx15c2 0.934 0.15 39.36 0.00 tx15c2 2.16 1.64 1.74 0.19
Cte -35.56 5.70 38.86 0.00 Cte -381.54 878.62 0.19 0.66
pend 0.074 0.01 28.79 0.00 pend 0.109 0.03 17.32 0.00 pend 0.128 0.03 18.68 0.00
la2lop 0.000104 0.00 71.05 0.00 u100 -0.026 0.01 23.52 0.00 la2lop -0.000286 0.00 10.43 0.00
la3 0.00053 0.00 20.96 0.00 pobl 0.0005 0.00 2.97 0.08 la3 -0.0073 0.00 6.40 0.01
la -2.14 0.54 15.42 0.00 p14c2 0.242 0.04 32.97 0.00 la 29.88 11.92 6.29 0.01
u100 -0.0087 0.00 47.84 0.00 u500 -0.0082 0.00 13.16 0.00 u100 -0.031 0.01 25.26 0.00
pobl 0.00031 0.00 9.14 0.00 Cte -2.43 0.26 84.78 0.00 pobl 0.0005 0.00 3.01 0.08
p14c2 -0.041 0.00 101.01 0.00 p14c2 0.355 0.07 24.35 0.00
tx13c2 -0.510 0.09 31.89 0.00 tx13c2 -0.114 0.51 0.05 0.82
tx15c2 0.432 0.11 14.99 0.00 tx15c2 0.280 0.60 0.22 0.64
arvs
ap_c
2
Cte 50.39 14.37 12.30 0.00 Cte -743.54 296.72 6.28 0.01
pend2 -0.029 0.01 27.22 0.00 alti 0.012 0.00 10.07 0.00 pend2 1.21 20.99 0.00 0.95
pend 0.943 0.12 63.49 0.00 u100 -0.129 0.05 6.74 0.01 pend -24.53 531.26 0.00 0.96
alti 0.0034 0.00 74.75 0.00 Cte -15.14 4.93 9.43 0.00 alti 0.079 1.50 0.00 0.96
tx13c2 -20.04 0.30 46.33 0.00 tx13c2 -258.65 5359.58 0.00 0.96chin
iv_c
2
tx15c2 2.49 0.35 49.85 0.00 tx15c2 245.67 7142.49 0.00 0.97
p14c2 -0.015 0.01 6.20 0.01 p14c2 27.41 685.63 0.00 0.97
la3 -0.0011 0.00 10.34 0.00 la3 -1.64 21.35 0.01 0.94
la 6.48 1.72 14.20 0.00 la 6643.89 87590.64 0.01 0.94
daut -0.029 0.01 17.15 0.00 daut 3.22 63.81 0.00 0.96
u100 -0.012 0.00 6.71 0.01 u100 -3.70 54.36 0.00 0.95
Cte -211.54 46.37 20.82 0.00 Cte -163984.07 2169393.57 0.01 0.94
alti 0.0034 0.00 11.82 0.00 p14c2 41.55 457.49 0.01 0.93 alti 0.069 0.11 0.43 0.51
daut 0.0060 0.00 4.25 0.04 lop3 39.95 812.42 0.00 0.96 daut 0.464 0.44 1.13 0.29
alti2 -0.0000012 0.00 7.39 0.01 lop2 -623.74 14234.66 0.00 0.97 alti2 -0.000026 0.00 0.35 0.55
tx14c2 0.825 0.09 88.80 0.00 lop 2915.17 78447.75 0.00 0.97 tx14c2 -10.40 11.68 0.79 0.37
tx13c2 -0.854 0.12 52.78 0.00 Cte -4646.33 138776.65 0.00 0.97 tx13c2 4.46 5.20 0.73 0.39
p13c2 0.022 0.01 15.59 0.00 p13c2 -0.246 0.45 0.29 0.59
mic
cab_
c2
Cte -15.42 1.51 103.78 0.00 Cte 157.04 253.78 0.38 0.54
alti -0.0040 0.00 60.67 0.00 p13c4 -0.283 0.08 13.93 0.00 alti 0.000054 0.00 0.00 0.98
lop -158.61 27.34 33.65 0.00 Cte 3.4567 1.65 4.40 0.04 lop -2412.28 1616.24 2.23 0.14
lop3 0.201 0.05 18.86 0.00 lop3 12.63 9.24 1.87 0.17
lalop2 -0.188 0.04 23.84 0.00 lalop2 -8.02 5.81 1.90 0.17
lalop 6.19 1.10 31.73 0.00 lalop 127.85 88.35 2.09 0.15
la3 -0.0088 0.00 34.56 0.00 la3 -0.126 0.08 2.30 0.13
daut -0.038 0.01 8.79 0.00 daut -0.117 0.10 1.47 0.23
u100 -0.015 0.00 9.54 0.00 u100 0.128 0.09 1.89 0.17
atea
lg_c
4
Cte 220.41 36.64 36.19 0.00 Cte 353.03 832.69 0.18 0.67
alti 0.0033 0.00 37.49 0.00 pend 0.214 0.05 18.81 0.00 alti 0.0024 0.00 1.32 0.25
alti2 -0.00000090 0.00 12.39 0.00 tx15c4 -0.913 0.20 21.55 0.00 alti2 -0.00000080 0.00 1.13 0.29
lop2 -0.245 0.03 93.60 0.00 Cte 14.75 4.15 12.62 0.00 lop2 0.561 0.96 0.34 0.56
lop3 0.0062 0.00 20.91 0.00 lop3 -0.0504 0.06 0.69 0.41neoa
no_c
4
la2lop 0.0018 0.00 205.96 0.00 la2lop 0.0013 0.00 0.11 0.73
u100 -0.0066 0.00 11.40 0.00 u100 -0.037 0.01 6.70 0.01
u500 0.0022 0.00 4.00 0.05 u500 -0.014 0.01 1.19 0.28
p14c4 -0.134 0.01 119.14 0.00 p14c4 -0.200 0.26 0.58 0.45
p15c4 0.061 0.01 85.49 0.00 p15c4 0.316 0.14 5.34 0.02
tx13c4 -0.879 0.08 110.99 0.00 tx13c4 -1.45 0.92 2.49 0.11
tx16c4 0.866 0.08 113.73 0.00 tx16c4 0.341 0.52 0.43 0.51
Cte -6.88 1.24 30.77 0.00 Cte -5.69 11.04 0.27 0.61
alti 0.0013 0.00 28.29 0.00 alti 0.0054 0.00 22.83 0.00 alti 0.0024 0.00 16.45 0.00
pend 0.086 0.02 12.30 0.00 alti2 0.0000 0.00 8.67 0.00 pend 0.0082 0.05 0.02 0.88
lop 1.967 0.26 55.33 0.00 tx15c4 9.93 2.22 20.02 0.00 lop -2.19 2.15 1.04 0.31
lalop2 -0.0057 0.00 130.32 0.00 tx14c4 -4.52 1.06 18.01 0.00 lalop2 0.0039 0.00 0.84 0.36
la3 -0.0023 0.00 96.42 0.00 la -42.13 18.81 5.02 0.03 la3 0.0098 0.01 1.22 0.27
la 110.01 1.13 95.27 0.00 la3 0.0093 0.00 4.37 0.04 la -44.96 38.72 1.35 0.25
daut 0.0066 0.00 7.18 0.01 tx16c4 -5.7906 1.18 24.15 0.00 daut -0.0056 0.01 0.20 0.66
u100 -0.0086 0.00 25.35 0.00 orS -0.0199 0.01 6.57 0.01 u100 -0.014 0.01 2.26 0.13
p13c4 -0.037 0.01 11.36 0.00 Cte 1069.07 470.68 5.16 0.02 p13c4 0.270 0.13 4.04 0.04
p15c4 0.110 0.02 28.78 0.00 p15c4 0.222 0.32 0.47 0.49
p16c4 -0.031 0.01 15.02 0.00 p16c4 -0.183 0.07 6.06 0.01
tx14c4 -0.964 0.18 28.70 0.00 tx14c4 -4.08 1.27 10.33 0.00
tx15c4 1.24 0.38 10.52 0.00 tx15c4 10.05 2.78 13.08 0.00
tx16c4 -0.386 0.20 3.55 0.06 tx16c4 -6.13 1.58 15.10 0.00
talo
cc_c
4
Cte -291.47 29.91 94.95 0.00 Cte 1142.14 979.12 1.36 0.24
alti 0.0082 0.00 121.37 0.00 la3 -0.528 0.12 17.95 0.00 alti -0.014 0.01 4.26 0.04
alti2 -0.0000027 0.00 51.94 0.00 la2 30.40 7.15 18.07 0.00 alti2 0.000018 0.00 9.99 0.00
lop2 -0.056 0.01 101.39 0.00 tx14c4 16.22 3.45 22.07 0.00 lop2 -0.320 0.20 2.60 0.11
la3 0.0042 0.00 143.37 0.00 tx15c4 -12.31 2.87 18.38 0.00 la3 -0.422 0.09 21.49 0.00canl
up_c
4
la -18.94 1.66 130.12 0.00 orS 0.0660 0.02 8.87 0.00 la 1857.02 399.33 21.63 0.00
p14c4 -0.0500 0.01 21.60 0.00 pend2 0.052 0.01 12.08 0.00 p14c4 -1.98 0.59 11.04 0.00
tx14c4 0.800 0.13 40.27 0.00 la2lop -0.0050 0.00 19.36 0.00 tx14c4 18.10 4.10 19.48 0.00
tx15c4 -0.878 0.15 34.42 0.00 Cte -15129.96 3510.18 18.58 0.00 tx15c4 -16.17 3.57 20.48 0.00
Cte 478.38 43.31 122.02 0.00 Cte -47592.79 10197.74 21.78 0.00
alti -0.0018 0.00 82.95 0.00 p15c4 -0.204 0.10 3.85 0.05 alti -0.0010 0.00 3.33 0.07
pend 0.611 0.05 171.79 0.00 p13c4 0.405 0.06 43.11 0.00 pend 0.4080 0.13 9.91 0.00
pend2 -0.026 0.00 84.97 0.00 p16c4 -0.065 0.04 3.22 0.07 pend2 -0.015 0.01 4.41 0.04
lop -117.29 12.26 91.56 0.00 lop -19.94 2.75 52.72 0.00 lop 716.53 160.25 19.99 0.00
lop2 -0.806 0.24 10.93 0.00 la2lop 0.0130 0.00 51.24 0.00 lop2 -1.63 10.23 0.03 0.87
lop3 0.0080 0.00 14.06 0.00 alti -0.0022 0.00 17.60 0.00 lop3 -0.129 0.06 5.01 0.03
lalop2 0.014 0.01 5.16 0.02 tx14c4 -0.449 0.12 13.10 0.00 lalop2 0.108 0.27 0.16 0.69
la2lop -0.075 0.01 88.13 0.00 pend 0.227 0.05 21.54 0.00 la2lop 0.519 0.10 29.18 0.00
lalop 5.969 0.62 91.72 0.00 u500 -0.015 0.00 15.67 0.00 lalop -38.97 7.39 27.80 0.00
la2 -0.412 0.05 82.42 0.00 Cte 18.67 5.67 10.84 0.00 la2 0.013 0.10 0.02 0.89
daut 0.0042 0.00 4.12 0.04 daut -0.0093 0.01 0.96 0.33
u100 0.0051 0.00 9.95 0.00 u100 0.0056 0.01 0.74 0.39
u500 0.0097 0.00 97.27 0.00 u500 -0.029 0.01 23.20 0.00
p13c4 -0.049 0.01 39.90 0.00 p13c4 0.220 0.10 4.57 0.03
p15c4 0.075 0.01 40.54 0.00 p15c4 -0.185 0.12 2.36 0.12
tx16c4 -0.293 0.04 45.12 0.00 tx16c4 -0.135 0.21 0.41 0.52
lutlu
t_c4
Cte 222.32 24.45 82.66 0.00 Cte 11.65 133.53 0.01 0.93
alti 0.0041 0.00 7.26 0.01 tx13c4 15.14 2.77 29.84 0.00 alti 0.0037 0.00 6.39 0.01
lop3 -0.0061 0.00 9.56 0.00 tx15c4 -15.99 3.13 26.16 0.00 lop3 -0.0017 0.00 0.28 0.59
u500 0.0150 0.01 4.30 0.04 p14c4 0.391 0.17 5.23 0.02 u500 0.021 0.01 9.58 0.00
tx13c4 7.96 1.62 24.18 0.00 tx14c4 1.36 0.42 10.60 0.00 tx13c4 9.61 1.82 27.87 0.00
tx15c4 -5.62 1.45 15.03 0.00 u500 0.019 0.01 6.00 0.01 tx15c4 -7.93 1.70 21.81 0.00lynp
ar_c
4
Cte -30.22 10.77 7.87 0.01 Cte 24.44 14.70 2.76 0.10 Cte -9.57 11.81 0.66 0.42
alti 0.0042 0.00 84.62 0.00 la -43.22 15.88 7.40 0.01 alti 0.0035 0.00 1.10 0.29
alti2 -0.0000020 0.00 87.37 0.00 la3 0.0086 0.00 4.53 0.03 alti2 -0.0000024 0.00 0.85 0.36
pend 0.597 0.05 126.01 0.00 pend 0.451 0.07 40.01 0.00 pend 0.508 0.26 3.87 0.05
pend2 -0.027 0.00 63.70 0.00 tx16c4 0.810 0.31 6.78 0.01 pend2 -0.0057 0.01 0.18 0.67
orS 0.0059 0.00 3.90 0.05 lalop2 0.223 0.05 21.98 0.00 orS -0.0037 0.01 0.09 0.76
lalop2 -0.0023 0.00 89.30 0.00 lop2 -9.68 1.99 23.67 0.00 lalop2 -0.035 0.01 10.07 0.00
la2lop 0.0064 0.00 234.76 0.00 la2lop 0.0094 0.00 17.05 0.00 la2lop 0.1001 0.03 12.63 0.00
lalop -0.221 0.02 191.78 0.00 Cte 1120.06 386.48 8.40 0.00 lalop -3.3402 0.98 11.65 0.00
la -10.09 0.14 62.96 0.00 la -19.60 5.61 12.19 0.00
daut -0.0088 0.00 17.78 0.00 daut -0.0075 0.02 0.22 0.64
u100 0.019 0.00 130.24 0.00 u100 0.00067 0.01 0.00 0.95
u500 0.0062 0.00 36.72 0.00 u500 -0.0058 0.02 0.14 0.71
p14c4 -0.100 0.01 68.03 0.00 p14c4 -0.385 0.33 1.38 0.24
p15c4 0.125 0.02 36.69 0.00 p15c4 0.750 0.40 3.53 0.06
p16c4 -0.031 0.01 15.77 0.00 p16c4 -0.157 0.12 1.75 0.19
tx14c4 -1.12 0.14 62.45 0.00 tx14c4 4.69 2.09 5.04 0.02
tx15c4 1.92 0.32 35.75 0.00 tx15c4 -10.47 4.72 4.92 0.03
tx16c4 -0.957 0.17 29.92 0.00 tx16c4 5.97 2.44 5.97 0.01
capc
ap_c
4
Cte 34.49 5.88 34.38 0.00 Cte 698.87 204.46 11.68 0.00
alti 0.0028 0.00 55.94 0.00 pend 0.212 0.04 22.42 0.00 alti 0.0060 0.00 15.94 0.00
alti2 -0.0000011 0.00 42.36 0.00 lalop 0.349 0.06 35.41 0.00 alti2 -0.0000023 0.00 16.73 0.00
pend 0.220 0.02 131.18 0.00 p15c4 0.496 0.06 61.73 0.00 pend 0.1885 0.06 8.75 0.00
orS 0.0092 0.00 11.24 0.00 tx14c4 0.609 0.19 10.40 0.00 orS -0.0091 0.01 0.93 0.34
lop 2.97 0.35 72.69 0.00 u500 -0.122 0.02 31.73 0.00 lop 28.65 20.75 1.91 0.17
sciv
ul_c
4
lop3 0.015 0.00 57.18 0.00 Cte -101.82 17.80 32.73 0.00 lop3 -0.034 0.19 0.03 0.86
lalop2 -0.0082 0.00 56.45 0.00 lalop2 -0.042 0.09 0.21 0.65
la 0.605 0.08 64.32 0.00 la -3.3661 4.66 0.52 0.47
u100 -0.005 0.00 10.15 0.00 u100 -0.013 0.01 1.02 0.31
pobl 0.00042 0.00 12.82 0.00 pobl 0.0015 0.00 5.65 0.02
p14c4 -0.030 0.01 16.09 0.00 p14c4 0.559 0.22 6.23 0.01
p15c4 0.054 0.01 75.74 0.00 p15c4 0.335 0.12 8.16 0.00
tx13c4 -0.366 0.14 6.90 0.01 tx13c4 1.21 1.48 0.67 0.41
tx14c4 -0.194 0.04 18.88 0.00 tx14c4 0.068 0.43 0.02 0.87
tx15c4 0.629 0.16 14.89 0.00 tx15c4 -0.153 1.84 0.01 0.93
Cte -40.23 4.47 80.99 0.00 Cte -20.93 205.46 0.01 0.92
tx15c4 0.428 0.11 14.54 0.00 alti2 0.00000173 0.00 16.67 0.00 tx15c4 0.202 0.64 0.10 0.75
tx13c4 -0.510 0.09 31.30 0.00 pend 0.089 0.03 6.72 0.01 tx13c4 -0.067 0.54 0.02 0.90
la3 0.00052 0.00 20.51 0.00 alti -0.0026 0.00 12.11 0.00 la3 -0.0067 0.00 5.52 0.02
la -2.11 0.54 15.05 0.00 u100 -0.024 0.01 20.19 0.00 la 27.38 11.75 5.43 0.02
u100 -0.0086 0.00 47.45 0.00 p14c4 0.256 0.05 30.13 0.00 u100 -0.029 0.01 23.40 0.00
p14c4 -0.042 0.00 100.01 0.00 u500 -0.0090 0.00 15.77 0.00 p14c4 0.331 0.07 21.17 0.00
pobl 0.00031 0.00 9.08 0.00 Cte -1.94 0.29 45.91 0.00 pobl 0.00051 0.00 3.12 0.08
la2lop 0.00010 0.00 70.78 0.00 la2lop -0.00028 0.00 9.33 0.00
pend 0.074 0.01 28.41 0.00 pend 0.130 0.03 19.28 0.00
arvs
ap_c
4
Cte 49.75 14.36 12.00 0.00 Cte -681.17 292.67 5.42 0.02
alti2 -0.000010 0.00 6.33 0.01 alti 0.012 0.00 10.07 0.00 alti2 -0.00030 0.00 0.05 0.82
pend2 -0.030 0.01 26.86 0.00 u100 -0.129 0.05 6.74 0.01 pend2 1.14 30.93 0.00 0.97
pend 0.919 0.12 55.28 0.00 Cte -15.14 4.93 9.43 0.00 pend -20.51 421.59 0.00 0.96
alti 0.0066 0.00 27.22 0.00 alti 1.1249 3.59 0.10 0.75
tx13c4 -2.59 0.34 57.95 0.00 tx13c4 -126.77 5880.22 0.00 0.98
tx15c4 3.10 0.39 63.69 0.00 tx15c4 -179.53 7116.08 0.00 0.98
chin
iv_c
4
p14c4 -0.149 0.02 39.72 0.00 p14c4 9.6122 982.25 0.00 0.99
la3 -0.00026 0.00 12.38 0.00 la3 -0.133 2.59 0.00 0.96
p15c4 0.335 0.05 51.52 0.00 p15c4 -45.22 249.58 0.03 0.86
p16c4 -0.137 0.02 52.79 0.00 p16c4 4.3125 40.57 0.01 0.92
la2lop 0.0074 0.00 32.49 0.00 la2lop 1.97 73.64 0.00 0.98
daut -0.040 0.01 34.84 0.00 daut -0.221 31.47 0.00 0.99
lalop -0.315 0.06 32.11 0.00 lalop -89.08 2682.57 0.00 0.97
Cte -10.14 7.00 2.10 0.15 Cte 16154.47 179366.90 0.01 0.93
alti 0.0034 0.00 11.56 0.00 p14c4 43.58 429.11 0.01 0.92 alti 0.0689 0.10 0.43 0.51
daut 0.0060 0.00 4.14 0.04 lop3 43.12 951.12 0.00 0.96 daut 0.458 0.43 1.14 0.29
alti2 -0.0000012 0.00 7.30 0.01 lop2 -701.21 16651.32 0.00 0.97 alti2 -0.000026 0.00 0.35 0.55
tx14c4 0.837 0.09 90.50 0.00 la2 -0.389 34.35 0.00 0.99 tx14c4 -10.26 11.53 0.79 0.37
tx13c4 -0.873 0.12 54.24 0.00 lop 3464.24 93293.95 0.00 0.97 tx13c4 4.37 5.12 0.73 0.39
p13c4 .023 0.01 16.46 0.00 Cte -5234.17 211563.49 0.00 0.98 p13c4 -0.248 0.46 0.29 0.59
mic
cab_
c4
Cte -15.49 1.52 103.30 0.00 Cte 155.50 251.50 0.38 0.54
alti -0.0036 0.00 36.69 0.00 p13e2 -0.455 0.13 13.03 0.00 alti 0.00044 0.00 0.04 0.85
lop2 -9.77 1.60 37.09 0.00 Cte 5.62 2.25 6.27 0.01 lop2 -121.71 94.50 1.66 0.20
lalop2 0.211 0.03 37.21 0.00 lalop2 3.12 2.39 1.70 0.19
la2lop -0.067 0.01 35.94 0.00 la2lop -0.758 0.59 1.66 0.20
lalop 3.318 0.55 35.99 0.00 lalop 31.07 24.65 1.59 0.21
daut -0.046 0.01 12.96 0.00 daut -0.057 0.10 0.31 0.58
u500 -0.0072 0.00 5.55 0.02 u500 -0.189 0.52 0.13 0.71
p13e2 -0.287 0.10 8.65 0.00 p13e2 4.91 4.91 1.00 0.32
p15e2 0.186 0.08 5.12 0.02 p15e2 -5.11 5.24 0.95 0.33
atea
lg_e
2
Cte -122.96 21.01 34.25 0.00 Cte -451.43 559.77 0.65 0.42
alti 0.0035 0.00 44.46 0.00 pend 0.206 0.05 17.05 0.00 alti 0.0017 0.00 0.74 0.39
alti2 -0.0000011 0.00 18.67 0.00 tx15e2 -0.885 0.20 20.36 0.00 alti2 -0.00000058 0.00 0.63 0.43
neoa
no_e
2
lop 15.40 4.53 11.54 0.00 Cte 16.33 4.62 12.50 0.00 lop 496.80 341.36 2.12 0.15
lop2 -1.70 0.38 20.10 0.00 lop2 -36.52 29.08 1.58 0.21
lalop2 0.039 0.01 18.63 0.00 lalop2 0.973 0.77 1.61 0.21
lalop -0.337 0.11 9.94 0.00 lalop -13.20 9.03 2.14 0.14
la 20.08 0.36 34.01 0.00 la 47.16 25.99 3.29 0.07
u100 -0.010 0.00 26.46 0.00 u100 -0.0440 0.02 7.93 0.00
u500 -0.0029 0.00 6.64 0.01 u500 -0.0046 0.02 0.03 0.85
p13e2 -0.054 0.01 15.49 0.00 p13e2 0.930 0.47 3.88 0.05
p14e2 -0.064 0.02 14.98 0.00 p14e2 -1.66 1.04 2.58 0.11
p16e2 0.052 0.01 28.63 0.00 p16e2 -0.143 0.15 0.89 0.34
tx13e2 -2.16 0.23 90.91 0.00 tx13e2 -1.09 2.11 0.27 0.60
tx15e2 2.28 0.25 83.94 0.00 tx15e2 -0.741 2.52 0.09 0.77
Cte -96.00 15.77 37.08 0.00 Cte -1747.32 977.85 3.19 0.07
alti 0.0015 0.00 44.48 0.00 alti 0.0063 0.00 48.93 0.00 alti 0.0019 0.00 12.82 0.00
pend 0.047 0.02 4.37 0.04 alti2 -0.00000168
0.00 18.16 0.00 pend 0.032 0.05 0.42 0.52
lop 1.36 0.18 54.55 0.00 la -1.92 0.27 49.52 0.00 lop 6.65 2.01 10.90 0.00
lalop2 -0.004 0.00 111.57 0.00 orS -0.023 0.01 8.68 0.00 lalop2 -0.016 0.00 11.81 0.00
la2 -0.297 0.03 95.78 0.00 p13e2 0.103 0.02 28.05 0.00 la2 3.11 0.77 16.48 0.00
la 24.26 2.43 99.87 0.00 Cte 65.19 9.86 43.70 0.00 la -236.10 57.71 16.73 0.00
u100 -0.0092 0.00 33.96 0.00 u100 -0.0025 0.01 0.10 0.75
tx13e2 -1.13 0.29 14.95 0.00 tx13e2 -6.07 1.79 11.47 0.00
tx14e2 -0.868 0.24 13.39 0.00 tx14e2 -1.80 1.06 2.89 0.09
tx15e2 2.81 0.43 42.93 0.00 tx15e2 14.94 3.47 18.53 0.00
tx16e2 -.905 0.35 6.76 0.01 tx16e2 -7.39 2.19 11.42 0.00
talo
cc_e
2
Cte -496.14 47.93 107.16 0.00 Cte 4417.14 1078.26 16.78 0.00
alti 0.0077 0.00 89.59 0.00 la3 -0.309 0.10 9.98 0.00 alti -0.0020 0.01 0.11 0.74
can_
lup_
e2
alti2 0.000 0.00 34.45 0.00 la2 17.76 5.59 10.10 0.00 alti2 0.0000044 0.00 0.65 0.42
lop2 -.174 0.01 188.69 0.00 tx14e2 2.95 0.57 27.11 0.00 lop2 -1.22 1.36 0.80 0.37
la2lop 0.00083 0.00 109.21 0.00 orS 0.054 0.02 7.33 0.01 la2lop 0.0091 0.01 0.59 0.44
u100 -0.0075 0.00 11.35 0.00 pend 0.548 0.16 11.90 0.00 u100 -0.146 0.04 11.77 0.00
u500 0.0077 0.00 37.66 0.00 Cte -8827.53 2712.81 10.59 0.00 u500 -0.046 0.05 0.72 0.40
pobl -0.0035 0.00 10.41 0.00 pobl -0.0027 0.00 0.43 0.51
tx13e2 30.01 0.30 100.28 0.00 tx13e2 20.42 7.33 7.77 0.01
tx14e2 -.611 0.07 82.54 0.00 tx14e2 20.61 4.45 21.46 0.00
tx15e2 -2.708 0.25 120.42 0.00 tx15e2 -40.13 9.00 19.86 0.00
Cte 7.00 1.05 44.14 0.00 Cte -234.82 94.53 6.17 0.01
alti -0.0015 0.00 46.59 0.00 p15e2 -0.595 0.10 38.23 0.00 alti -0.0014 0.00 7.08 0.01
pend 0.541 0.05 137.38 0.00 p13e2 0.554 0.09 41.91 0.00 pend 0.402 0.12 10.43 0.00
pend2 -0.023 0.00 73.13 0.00 p16e2 0.078 0.04 3.30 0.07 pend2 -0.0110 0.01 2.60 0.11
lop -15.42 4.70 10.78 0.00 lop -27.89 3.24 74.28 0.00 lop 242.75 189.59 1.64 0.20
la2lop -0.0075 0.00 6.46 0.01 lalop 0.704 0.09 66.98 0.00 la2lop 0.190 0.13 2.07 0.15
lalop 0.678 0.23 8.33 0.00 tx14e2 -0.446 0.08 30.43 0.00 lalop -13.62 10.00 1.85 0.17
daut 0.0045 0.00 4.29 0.04 alti -0.0018 0.00 13.29 0.00 daut -0.023 0.01 7.61 0.01
u100 0.0076 0.00 22.50 0.00 pend 0.201 0.05 18.24 0.00 u100 0.018 0.01 10.89 0.00
u500 0.0084 0.00 102.15 0.00 Cte 20.28 3.48 33.93 0.00 u500 -0.0098 0.00 9.01 0.00
p14e2 0.059 0.01 18.62 0.00 p14e2 0.360 0.18 4.02 0.04
p15e2 -0.162 0.02 89.89 0.00 p15e2 -0.270 0.12 5.14 0.02
p16e2 0.122 0.01 103.70 0.00 p16e2 0.146 0.06 6.30 0.01
tx14e2 0.890 0.13 45.49 0.00 tx14e2 -0.803 0.68 1.40 0.24
tx15e2 -2.19 0.35 39.46 0.00 tx15e2 1.37 2.10 0.42 0.52
tx16e2 1.30 0.24 29.65 0.00 tx16e2 -0.900 1.45 0.38 0.54
lulu
t_e2
Cte 1.66 1.04 2.57 0.11 Cte 11.08 7.47 2.20 0.14
orS 0.029 0.01 4.79 0.03 tx13e2 11.83 5.01 5.59 0.02 orS 0.0245 0.01 3.38 0.07
lynp
ar_e
2
lop3 -0.006 0.00 5.13 0.02 tx15e2 -80.80 20.57 15.43 0.00 lop3 -0.0069 0.00 3.25 0.07
p13e2 -0.326 0.10 11.24 0.00 tx16e2 48.51 15.08 10.34 0.00 p13e2 -0.314 0.09 12.80 0.00
tx13e2 1.48 0.27 29.33 0.00 alti 0.0047 0.00 4.77 0.03 tx13e2 1.10 0.34 10.37 0.00
Cte -32.53 7.77 17.53 0.00 lop3 -0.0277 0.01 6.97 0.01 Cte -22.63 9.77 5.37 0.02
tx14e2 22.69 7.95 8.15 0.00
p13e2 -1.06 0.26 16.67 0.00
Cte 174.60 38.26 20.83 0.00 alti 0.0036 0.00 70.06 0.00 p16e2 0.146 0.08 3.64 0.06 alti 0.0053 0.00 3.98 0.05
alti2 0.00000 0.00 71.99 0.00 p13e2 0.471 0.14 10.84 0.00 alti2 -0.0000029 0.00 2.06 0.15
pend 0.573 0.05 114.47 0.00 p15e2 -0.690 0.17 15.76 0.00 pend 0.427 0.24 3.14 0.08
pend2 -0.024 0.00 50.93 0.00 la2 -4.25 0.93 20.77 0.00 pend2 -0.011 0.01 0.80 0.37
orS 0.0073 0.00 5.84 0.02 la3 0.0760 0.02 21.67 0.00 orS 0.011 0.01 1.00 0.32
la3 0.00025 0.00 170.60 0.00 pend 0.755 0.20 14.92 0.00 la3 0.00083 0.00 2.96 0.09
u100 0.017 0.00 96.27 0.00 pend2 -0.022 0.01 4.70 0.03 u100 -0.0016 0.01 0.02 0.89
u500 0.003 0.00 10.45 0.00 tx16e2 1.32 0.52 6.35 0.01 u500 0.000798 0.01 0.01 0.92
p13e2 0.152 0.01 205.03 0.00 tx14e2 0.701 0.32 4.80 0.03 p13e2 0.4002 0.24 2.82 0.09
p14e2 0.072 0.01 52.94 0.00 Cte 1911.26 444.51 18.49 0.00 p14e2 0.699 0.45 2.41 0.12
p15e2 -0.402 0.02 387.44 0.00 p15e2 -1.1556 0.21 29.87 0.00
p16e2 0.190 0.01 248.96 0.00 p16e2 0.410 0.11 13.50 0.00
tx13e2 -0.540 0.23 5.63 0.02 tx13e2 1.4602 3.08 0.23 0.64
tx14e2 1.88 0.17 116.15 0.00 tx14e2 3.8934 1.91 4.18 0.04
tx15e2 -4.47 0.43 106.92 0.00 tx15e2 -13.38 3.42 15.30 0.00
tx16e2 3.49 0.30 133.93 0.00 tx16e2 9.9570 3.11 10.25 0.00
capc
ap_e
2
Cte -21.12 2.55 68.38 0.00 Cte -54.29 31.99 2.88 0.09
alti 0.0030 0.00 65.36 0.00 alti 0.0053 0.00 15.74 0.00 alti 0.0053 0.00 13.59 0.00
alti2 -0.000010 0.00 36.62 0.00 alti2 -0.00000203
0.00 14.61 0.00 alti2 -0.0000021 0.00 15.10 0.00
sciv
ul_e
2
pend 0.228 0.02 115.78 0.00 pend 0.194 0.06 9.73 0.00 pend 0.291 0.06 23.19 0.00
orS 0.014 0.00 26.17 0.00 tx16e2 2.32 0.56 17.33 0.00 orS -0.012 0.01 1.73 0.19
lop2 0.030 0.00 111.24 0.00 lalop2 -0.0403 0.01 13.80 0.00 lop2 0.0095 0.05 0.04 0.85
la2 -0.073 0.02 12.03 0.00 lop 24.35 5.39 20.43 0.00 la2 -1.64 1.27 1.67 0.20
la 60.07 1.66 13.31 0.00 p14e2 2.01 0.42 22.60 0.00 la 119.26 96.17 1.54 0.21
u100 -0.0040 0.00 6.70 0.01 u500 -0.073 0.01 27.44 0.00 u100 -0.0203 0.01 3.57 0.06
pobl 0.00045 0.00 13.49 0.00 p16e2 -0.173 0.09 4.13 0.04 pobl 0.0013 0.00 7.91 0.00
p13e2 0.046 0.01 26.31 0.00 Cte -135.80 24.91 29.72 0.00 p13e2 0.316 0.20 2.39 0.12
p14e2 -0.065 0.01 32.39 0.00 p14e2 0.558 0.45 1.54 0.21
p15e2 0.056 0.02 8.30 0.00 p15e2 0.655 0.32 4.31 0.04
p16e2 -0.051 0.01 15.06 0.00 p16e2 -0.498 0.12 16.75 0.00
tx14e2 -0.239 0.04 34.81 0.00 tx14e2 -1.1414 0.38 9.02 0.00
tx15e2 0.163 0.06 6.71 0.01 tx15e2 1.8185 0.65 7.82 0.01
Cte -128.51 32.55 15.59 0.00 Cte -2182.49 1816.52 1.44 0.23
tx13e2 -0.158 0.03 30.86 0.00 pend 0.1015 0.03 9.44 0.00 tx13e2 -0.138 0.10 1.99 0.16
la3 0.000051 0.00 34.80 0.00 alti 0.00084 0.00 8.80 0.00 la3 -0.0000068 0.00 0.02 0.90
u100 -0.010 0.00 80.06 0.00 u100 -0.0146 0.00 15.32 0.00 u100 -0.013 0.00 7.56 0.01
pobl 0.00031 0.00 9.07 0.00 pobl 0.0007 0.00 5.93 0.01 pobl 0.0007 0.00 5.11 0.02
la2lop -0.000049 0.00 30.78 0.00 Cte -1.67 0.22 57.89 0.00 la2lop 0.000022 0.00 0.12 0.73
pend 0.071 0.01 31.05 0.00 pend 0.134 0.03 20.77 0.00
arvs
ap_e
2
Cte 0.639 1.06 0.37 0.55 Cte 1.91 2.90 0.43 0.51
alti2 -0.0000012 0.00 7.80 0.01 alti 0.012 0.00 10.07 0.00 alti2 0.0000094 0.01 0.00 1.00
pend2 -0.027 0.01 19.76 0.00 u100 -0.129 0.05 6.74 0.01 pend2 0.618 59.39 0.00 0.99
pend 0.867 0.13 42.67 0.00 Cte -15.14 4.93 9.43 0.00 pend -13.85 1418.87 0.00 0.99
alti 0.0073 0.00 30.87 0.00 alti 0.097 32.47 0.00 1.00
tx13e2 -3.41 0.71 23.18 0.00 tx13e2 -242.59 24947.78 0.00 0.99
tx15e2 -2.61 1.39 3.52 0.06 tx15e2 -1372.22 30507.84 0.00 0.96
chin
iv_e
2
tx14e2 2.39 0.60 16.08 0.00 tx14e2 624.43 24268.35 0.00 0.98
tx16e2 4.39 0.97 20.56 0.00 tx16e2 1009.88 32679.58 0.00 0.98
p13e2 0.133 0.03 19.29 0.00 p13e2 53.10 1541.12 0.00 0.97
la2 -0.264 0.07 15.30 0.00 la2 -0.059 578.22 0.00 1.00
p15e2 -0.150 0.03 25.80 0.00 p15e2 -54.17 1782.34 0.00 0.98
la2lop 0.0063 0.00 20.58 0.00 la2lop -1.07 243.28 0.00 1.00
daut -0.031 0.01 19.17 0.00 daut 1.96 42.39 0.00 0.96
lalop -0.275 0.06 22.21 0.00 lalop 38.27 9231.27 0.00 1.00
Cte -412.50 103.14 15.99 0.00 Cte 2155.99 970426.61 0.00 1.00
alti 0.0010 0.00 10.23 0.00 p14e2 37.62 144.56 0.07 0.79 alti 0.053 2.96 0.00 0.99
tx14e2 0.431 0.13 11.59 0.00 lop2 31.30 121.31 0.07 0.80 tx14e2 50.60 2308.98 0.00 0.98
tx16e2 -0.541 0.15 12.47 0.00 Cte -2380.52 9198.12 0.07 0.80 tx16e2 -4.40 2199.43 0.00 1.00
p13e2 0.281 0.05 31.55 0.00 p13e2 -23.01 1993.52 0.00 0.99
lop3 0.013 0.00 6.95 0.01 lop3 -0.548 345.70 0.00 1.00
p15e2 -0.155 0.05 10.89 0.00 p15e2 27.21 1726.64 0.00 0.99
lalop2 -0.029 0.01 20.15 0.00 lalop2 21.88 454.43 0.00 0.96
lop2 10.01 0.24 17.94 0.00 lop2 -807.88 14448.62 0.00 0.96
la3 -0.021 0.00 110.78 0.00 la3 -5.21 124.66 0.00 0.97
la2 1.28 0.12 111.29 0.00 la2 280.05 6916.54 0.00 0.97
mic
cab_
e2
Cte -704.35 65.12 117.00 0.00 Cte -121466.38 3143944.03 0.00 0.97
ANEJO Nº2.
INTRODUCTION
This work presents the preliminary results of a project aiming to identify the critical tracks of the Andalusian
(Spain) road network which could hinder the needed permeability for the threatened non-volant
tetrapod
species in their geographic response to climate change. We propose to take advantage of national species distribution models, but proceeding to update them with variables yielded by climatic models for the regional target area before projecting them to the future.
Ana-Luz Márquez1
and Raimundo
Real1
1: Biogeography, Diversity, and Conservation Research Team, Dept. of Animal Biology, Faculty of Sciences, University of Malaga, E-29071, Malaga, Spain
Applying modelling
updating approaches to identify potential conflicts with the main roads in the expected geographic adaptation of species to climate change
IBS 6th Biennial
Conference, Miami, Florida
EUROPE
Andalusia
SPAIN
STUDY AREA
We assessed the classification power of the models by calculating Cohen’s kappa using the favourability
value of F = 0.5 as the classification threshold, and the discrimination power using the area under the Receiver Operating Characteristic Curve (AUC). We compared the parsimony of the models using the Akaike
Information Criterion (AIC).
METHODS
We used the favourability
function (Real et al 2006) to model the 20 non-volant
tetrapod
species, with some degree of threat in Andalusia, under different emission scenarios (A2 and B2) and the general circulation model GGCM2. We evaluated different approaches to build the models for Andalusia. The first approach was to use the same
models for mainland Spain projected only on Andalusia (NAT). The second approach was to update the national models by using the same variables and recalibrating the models using the data for Andalusia (UPD). A third approach was to perform a new model for Andalusia (REG).
RESULT
REFERENCES-
Real R, Barbosa
AM, Vargas JM (2006) Obtaining environmental favourability
functions from logistic regression. Environ Ecol Stat 13: 237 –
245.
Values of AIC, AUC and Cohen’s kappa for each approach for the
emission scenario CGCM2-B2
Amphibian
Salamandra
salamandra
Alytes
dickhilleniTriturus
pygmaeus
Mammal
Atelerix
algirus
Neomys
anomalusTalpa
occidentalisCanis
lupus Lutra
lutraLynx pardinus
ReptileEmys
orbicularisTestudo
graecaLacerta
schreiberiAlgyroides
marchiCoronella
austriacaVipera
latasti
Capreolus
capreolusSciurus
vulgarisArvicola
sapidusChionomys
nivalisMicrotus
cabrerae
SPECIES
CONCLUSION
CGCM2-A2 scenario favourability
mapsDistribution map
C. l
upus
Values of AIC, AUC and Cohen’s kappa for each approach for the
emission scenario CGCM2-A2
Species for which the best model is the resulting from performing a new model for Andalusia (REG) for at least two of the criteria considered.
Species for which the NAT approach is better than the REG approach
(although not better than the UPD approach) for all the criteria
considered.
Species for which the best model is that resulting from the recalibration of the national model for Andalusia (UPD) at least for two of the criteria considered.
Our results challenge the general opinion that it is better to perform a species distribution model at a larger distribution area of the species. Only for T. graeca
that approach is better than building the model for a part of the total distribution area.However, updating the national models by using the same variables and recalibrating the models using the data for Andalusia (UPD) resulted to be the best approach for building species distribution models for most of the species
considered in this study.
Distribution map CGCM2-B2 scenario favourability
maps
V. l
atas
ti0.00 -
0.200.21 -
0.400.41 –
0.600.61 –
0.800.81 –
1.00Favourability
AIC AUC kappa AIC AUC kappa AIC AUC kappa
S. salamandra 903.15 0.833 0.460 710.63 0.906 0.658 605.70 0.932 0.702A. marchi 57.37 0.994 -0.010 55.69 0.995 0.532 51.30 0.996 0.555C. austriaca 72.38 0.901 0.405 31.37 0.982 0.099 26.01 1 1 L. schreiberi 72.50 0.480 -0.004 -- -- -- 24.00 1 1 T. graeca 103.51 0.983 0.220 113.90 0.963 0.203 97.06 0.985 0.310 V. latasti 842.01 0.725 0.225 785.89 0.751 0.277 741.38 0.813 0.345 N. anomalus 263.95 0.733 -0.002 206.77 0.867 0.120 206.96 0.886 0.157 T. occidentalis 617.59 0.693 0.106 438.90 0.884 0.357 450.93 0.893 0.357C. capreolus 530.70 0.709 0.065 331.87 0.914 0.382 343.39 0.920 0.384 S. vulgaris 484.96 0.902 0.472 333.30 0.955 0.550 324.20 0.963 0.568 C. nivalis 64.33 0.996 0.404 25.76 0.999 0.660 22.00 1 1 A. dickhilleni 388.94 0.929 0.357 284.90 0.966 0.407 363.61 0.940 0.272 T. pygmaeus 786.02 0.817 0.232 705.57 0.863 0.458 752.59 0.837 0.408 C. lupus 487.91 0.640 -0.004 125.88 0.990 0.621 136.30 0.988 0.591 L. lutra 1044.14 0.804 0.458 876.32 0.853 0.604 888.29 0.856 0.595 L. pardinus 174.97 0.924 0.128 153.30 0.947 0.227 167.88 0.933 0.181 A. sapidus 903.66 0.653 0.196 840.51 0.721 0.273 846.30 0.717 0.243 M. cabrerae 99.92 0.929 0.014 10.00 1 1 23.35 0.999 0.269 E. orbicularis 612.74 0.814 0.197 519.65 0.881 0.361 534.13 0.880 0.371 A. algirus 80.48 0.972 0.250 65.78 0.969 0.142 70.29 0.985 0.206
REGNAT UPD
AIC AUC kappa AIC AUC kappa AIC AUC kappa
S. salamandra 903.15 0.835 0.461 724.92 0.900 0.652 587.40 0.936 0.714 T. pygmaeus 788.02 0.815 0.231 762.67 0.839 0.422 752.64 0.839 0.418A. marchi 57.37 0.994 -0.010 55.57 0.995 0.532 51.30 0.996 0.555C. austriaca 72.38 0.900 0.411 35.37 0.982 0.099 26.01 1.000 1 E. orbicularis 612.74 0.810 0.184 530.96 0.877 0.351 536.71 0.880 0.363 L. schreiberi 72.50 0.486 -0.004 -- -- -- 24.00 1.000 1 T. graeca 103.51 0.983 0.219 113.56 0.963 0.205 97.21 0.986 0.346V. latasti 842.01 0.725 0.226 785.34 0.752 0.279 741.06 0.814 0.340 N. anomalus 265.95 0.757 -0.006 210.78 0.867 0.120 205.25 0.895 0.160 T. occidentalis 617.59 0.693 0.109 437.73 0.885 0.352 451.11 0.893 0.358 C. lupus 487.91 0.639 -0.004 144.21 0.985 0.591 136.79 0.988 0.591C. capreolus 530.70 0.706 0.056 340.22 0.915 0.372 341.36 0.922 0.411 S. vulgaris 484.96 0.909 0.489 343.51 0.957 0.540 335.51 0.957 0.561 C. nivalis 70.33 0.997 0.374 29.76 0.999 0.660 28.00 1 1 L. pardinus 174.97 0.929 0.131 150.32 0.951 0.238 164.77 0.937 0.198A. sapidus 903.66 0.654 0.189 827.18 0.735 0.259 849.74 0.713 0.252 M. cabrerae 99.92 0.930 0.014 16.00 1 1 23.38 0.999 0.269 A. dickhilleni 392.94 0.940 0.357 284.04 0.966 0.386 289.43 0.966 0.392 A. algirus 76.48 0.972 0.245 67.84 0.960 0.142 65.00 0.987 0.203 L. lutra 1052.14 0.807 0.442 896.66 0.852 0.602 891.02 0.853 0.517
UPDREGNAT
Project G-GI3000/IDIG
INTRODUCTION
This work presents the preliminary results of a project aiming to identify the critical tracks of the Andalusian
(Spain) road network which could hinder the needed permeability for the threatened non-volant
tetrapod
species in their geographic response to climate change.
Raimundo
Real and Ana-Luz MárquezBiogeography, Diversity, and Conservation Research Team, Dept. of Animal Biology, Faculty of Sciences, University of Malaga, E-29071, Malaga, Spain
Do national distribution models capture the environmental response of species better than regional ones?
EUROPE
Andalusia
SPAIN
STUDY AREA
CONCLUSION
METHODS
We modelled the ecogeographical
favourability for 11 threatened mammal species in Andalusia (Spain), under the climate change scenario GGCM2-A2. We considered three approaches:
a)
Using national models (trained in the whole mainland Spain) and
projected on Andalusia
(NAT-
model)
b) Performing regional models (trained in Andalusia only) (REG-model)
c) Using the national models recalibrating them with the distribution data for Andalusia (UPD-
model)
For all the models we assessed their calibration power, by calculating the Root Mean Square Error (RMSE) of their predicted and observed presences in probability bins, their parsimony, discrimination and classification power and the probability function domain. For each species we selected for projection to the future the model approach with the best assessment scores.
11th
International Mammalogical
Congress
Updating the national models (UPD) or performing the models for
Andalusia only (REG) resulted to be the best approaches for building species distribution models for the mammal species considered in this study. The models with the
best classification and discrimination power and the higher parsimony
not always were the best calibrated models. The updating approach (UPD) produced most of the best calibrated models. Wider national distribution models do not capture the environmental response of species better than more focused regional ones, although they may include valid information to be updated with regional distribution data.
RESULT
Values of AIC, AUC, Cohen’s kappa and RMSE for each approach for the emission scenario CGCM2-A2
M. cabrerae
calibration graphs and favourability maps.
The best calibrate models
Species for which the best model is that resulting from the recalibration of the national model for Andalusia (UPD )for at least
two of the criteria considered (AIC, AUC, kappa)
Species for which the best model is the resulting from performing a new model for Andalusia (REG) for at least two of the criteria considered (AIC, AUC, kappa).
S. vulgaris
calibration graphs and favourability maps
REG-models
UDP-modelsC. capreolus
calibration graphs and favourability maps
L. lutra
calibration graphs and favourability maps
AIC AUC kappa RMSE AIC AUC kappa RMSE AIC AUC kappa RMSE
N. anomalus 263.95 0.733 -0.002 0.0775 206.77 0.867 0.120 0.1236 206.96 0.886 0.157 0.2091 T. occidentalis 617.59 0.693 0.106 0.2912 438.90 0.884 0.357 0.1013 450.93 0.893 0.357 0.3297C. capreolus 530.70 0.709 0.065 0.2559 331.87 0.914 0.382 0.0559 343.39 0.920 0.384 0.1034 S. vulgaris 484.96 0.902 0.472 0.1082 333.30 0.955 0.550 0.1186 324.20 0.963 0.568 0.0637 C. nivalis 64.33 0.996 0.404 0.2393 25.76 0.999 0.660 0.2822 22.00 1 1 0.0000 C. lupus 487.91 0.640 -0.0042 0.1499 125.88 0.990 0.621 0.1244 136.30 0.988 0.591 0.1176 L. lutra 1044.14 0.804 0.458 0.0765 876.32 0.853 0.604 0.0812 888.29 0.856 0.595 0.0636 L. pardinus 174.97 0.924 0.128 0.2380 153.30 0.947 0.227 0.1100 167.88 0.933 0.181 0.0979 A. sapidus 903.66 0.653 0.196 0.2404 840.51 0.721 0.273 0.1265 846.30 0.717 0.243 0.0947 M. cabrerae 99.92 0.929 0.014 0.1294 10.00 1 1 0.0000 23.35 0.999 0.269 0.4856 A. algirus 80.48 0.972 0.250 0.2189 65.78 0.969 0.142 0.1864 70.29 0.985 0.206 0.2815
UPDREGNAT
Squares outside the probability function domain.
2
1
1
N
i ii OFN
RSMEFi
= mean probability
of
presence
predicted
in bin
iOi
= porportion
of
sites
in bin
i with
observed
presencesN = number
of
probability
bins
Project G-GI3000/IDIG
Distribution of Bonelli’s Eagle Aquila fasciata in southern Spain: scalemay matter
Antonio-Román MUÑOZ1,2,3 & Raimundo REAL1
1Biogeography, Diversity, and Conservation Research Team, Dept. of Animal Biology, Faculty of Sciences, University ofMalaga, E-29071, Malaga, SPAIN, e-mail: [email protected]ón Migres, N-340, Km. 96 Huerta Grande, Pelayo, E-11390 Algeciras, SPAIN3Didactics of Mathematics, Didactics of Social Sciences and Experimental Sciences, Faculty of Education Sciences,University of Malaga, E-29071, Malaga, Spain.
Muñoz A.-R., Real R. 2013. Distribution of Bonelli’s Eagle Aquila fasciata in southern Spain: scale may matter. ActaOrnithol. 48: 93–101. DOI 10.3161/000164513X670043
Abstract. Understanding factors that determine the distribution of the endangered Bonelli’s Eagle requires differentapproaches and analytical tools. These factors may differ depending on the spatial scale at which they act. Bonelli’sEagle distribution in Spain has been studied previously using local and large (nation-wide) study area sizes, and humanactivities seemed not to affect negatively the occupancy of breeding territories. To study the factors affecting the speciesat an intermediate spatial scale we modelled Bonelli's Eagle distribution in Málaga province (S Spain), where the breed-ing density is the highest known in Europe. We applied a favourability function based on generalized linear modelsusing the presence/absence of breeding territories of the species, and the values of a set of variables related to climate,topography, interspecific competition with Golden Eagle Aquila chrysaetos and human activity. We obtained a parsimo-nious model that included cliff availability and distance to highways as predictors of Bonelli’s Eagle distribution. Ashighways may be seen as surrogates of intensive human activity, we conclude that, contrary to what was previouslyfound at local or at nation-wide scales, human actions negatively affect the distribution of breeding territories at anintermediate scale. The construction of new roads and highways in the Mediterranean area of mainland Spain, whichis the most climatically favourable region for the species, could have negative consequences for the Spanish metapop-ulation of Bonelli's Eagle, particularly in peripheral populations or distant areas that depend on the arrival of immi-grants to persist.
Key words: Aquila fasciata, human disturbance, Hieraaetus fasciatus, predictive models, spatial scale, variation partitioning
Received — Nov. 2010, accepted — Jan. 2013
ACTA ORNITHOLOGICAVol. 48 (2013) No. 1
INTRODUCTION
Spatial distribution patterns of animal popula-tions are mainly induced by environmental het-erogeneity that determines gradients of habitatquality and may have profound effects on the spa-tial structure and abundance of species (Brown1984, Orians & Wittenberger 1991, Rahbek et al.2007). Local studies analysing the surroundings ofbreeding sites are normally biased to good qualityhabitats, as they focus on relatively homogeneousareas. Local factors influencing the distribution ofa species may be different from those acting overlarger areas that include a broader range of habi-tat quality. Indeed some recent studies do includemore than one spatial scale in their analyses (e.g.Sergio et al. 2003). Determining the relative contri-butions of local versus regional factors affecting
the distribution of a species may be the key tounderstanding its overall distribution pattern.This requires linking the spatial scale being con-sidered to the effects of the hypothesized process-es operating at that scale. If we are to attribute rel-ative impacts of various factors influencing thedistribution of the species, we should considerhow results vary as a function of scale, and alsosearch for consistent patterns across scales.
During the second half of the 20th century,Bonelli’s Eagle Aquila fasciata suffered one of themost severe population declines recorded amongbirds of prey, and was consequently listed asendangered in Europe (BirdLife International2004). Although the population appears to havestabilized in recent years, there are still areaswhere threats persist for this species, mainly dueto habitat degradation and unnatural mortality
(Ponchon 2011, López-López et al. 2012). In Spain,which supports about 80% of the European breed-ing population, it has changed its status from“Vulnerable” (Blanco & González 1992) to“Endangered” (MadroZo et al. 2004), and localextinction rates ranged from 32.1% to 48.6% forthe years 1980–1997 (Real & MaZosa 1997, Carreteet al. 2002). Because of this, different authors haveanalysed influential factors acting at diverse scaleson this species. Hengeveld (1990) distinguishedbetween the scale of variation that refers to thesize of the territory analysed, and the scale whichrefers to the size of the unit used to lattice the ter-ritory. López-López et al. (2006) analysed theeffects of different scales of resolution on Bonelli’sEagle’s habitat preferences maintaining the scaleof variation. As regards the scale of variation, mostanalyses of Bonelli’s Eagle distribution are localstudies involving nest-centred areas smaller than100 km2 (e.g. Gil-Sánchez et al. 2004, but seeNiamir et al. 2011), with recent contributions ofnational scale studies (MuZoz et al. 2005, Carrascal& Seoane 2009), including different scales of vari-ation.
Bonelli’s Eagle is traditionally considered asremarkably tolerant of human activity. Most of thelocal-scale ecological studies (e.g. Gil-Sánchez etal. 1996, Rico et al. 1999, Carrete et al. 2002, Gil-Sánchez et al. 2004, López-López et al. 2006) dis-counted human-related activities as a main causeof nest-site selection or abandonment of breedingterritories. MuZoz et al. (2005) considered humanactivity as only a secondary factor at a national-scale, with cliff availability and climate as the pri-mary explanatory factors, and Carrascal & Seoane(2009) did not find effects of the variables describ-ing the degree of human pressure on the distribu-tion of the species in Spain. In order to bridge thegap between these two approaches, local versusnational analyses, we studied the factors affectingthe absence/presence of breeding territories of thespecies at an intermediate spatial scale.
In this paper we model the distribution ofBonelli’s Eagle to provide understanding of thespecies distribution using an intermediate studyarea size (larger than a territory but smaller than anational analysis) in Málaga province (southernSpain), which holds the highest breeding densityin Europe. We then assess how the scale affectsour perception of patterns of Bonelli’s Eagle distri-bution, and our inference of causal processes, par-ticularly human activity. We consider the implica-tions of the obtained results for Bonelli’s Eagleconservation.
94 A.-R. Muñoz & R. Real
MATERIALS AND METHODS
Study areaThe study area comprises Málaga province (S Spain, Fig. 1), a mountainous region of 7267km2 with a typically Mediterranean climate (meanannual rainfall ranging from 400–1200 mm andannual temperature ranging from 12.6 to 19.2 °C).At present Málaga supports 80–82 breeding terri-tories of Bonelli’s Eagle (Bautista et al. 2003,Jiménez & MuZoz 2008), and is considered to bepart of one of the last strongholds of the species inEurope (Balbontín et al. 2003).
Bird censusesThe breeding population of Bonelli’s Eagles wasmonitored from 2001 to 2005 by visiting at leastthree times all potential territories during thebreeding season. The first inspection was made inJanuary to check for territory occupancy. The sec-ond check of the nesting areas was made duringFebruary and early March to record egg laying,and the third visit occurred during April or May,to confirm the reproduction (see Del Moral 2006).The breeding territories were considered as occu-pied if we observed the nests obviously repairedwith green branches, typical pair behaviour,courtship, brood rearing activity or young eagles(Rico et al. 1999, Carrete et al. 2001). Golden EaglesAquila chrysaetos were surveyed from 1999 to 2001,and territorial pairs were censused each year dur-ing the pre-incubation period. The populationconsisted of 20–22 breeding pairs.
Breeding density of Bonelli’s Eagle reaches values up to 4 pairs/100 km2 in contiguous 10×10-km squares. We took the presence/absence
Fig. 1. Iberian Peninsula and location of the study area (Málagaprovince, in black).
of breeding territories of Bonelli’s Eagle on the 104UTM 10×10-km squares of Málaga province fromJiménez & MuZoz (2008), breeding territories werepresent in 67 squares. We selected this size ofsquares because it is considered to be a landscaperesolution scale, as home ranges are typicallysmaller (López-López et al. 2006).
Predictor variablesTo model Bonelli’s Eagle distribution we used 31independent variables related to spatial situation,topography, climate, lithology, human activity, andGolden Eagle, its main competitor, breeding terri-tories presence/absence (Table 1). These variableswere chosen on the basis of potential predictiveand explanatory power, include most of the fac-tors influencing its distribution, and wereassumed to be at least correlated with more prox-imal causal factors (MuZoz et al. 2005). AsRobertson et al. (2003) suggested, models that relyon indirect links between species distributionrecords and environmental variables can predict
Scale-related effect of human activity on Bonelli's Eagle distribution 95
distributions at least as well as mechanistic modelsthat use more proximal variables (Austin 2002).
We digitized the variables, except for altitude,which is released as a DEM by US GeologicalSurvey (1996), using the CartaLinx 1.2 softwareand processed them using the Idrisi32 GIS soft-ware. Isoline variables, from „mean relative airhumidity in January” to „longitude” (Table 1)were interpolated from a triangulated irregularnetwork performing parabolic bridge and tunneledge removal, obtaining values in approximately1-km2 pixels. Area was calculated using theIdrisi32 AREA module. Secondary variables,defined in Table 1 were calculated from the mapsof the primary variables by an algebraic operationin parentheses using the Idrisi Image Calculator.We also digitized the highways and major urbancenters and calculated their distance to each 1-km2
pixel using the Idrisi DISTANCE module. Soil per-meability (Perm) was obtained from a map of syn-theses of ground-water aquifers, a categorical mapwith three different permeability classes (IGME
Table 1. Environmental variables used to model the distribution of Bonelli’s Eagle in Málaga province. Data sources: 1 — US Geological Survey (1996), 2 — Font (1983), 3 — Font (2000), 4 — Montero de Burgos & González-Rebollar (1974), 5 — IGME (1979), 6 — IGN (1999), 7 — own data. Data on the number of inhabitants of urban centers taken from the InstitutoNacional de Estadística (http://www.ine.es).
Code Variable Source
Area Surface area (km2)
Alti Altitude (m) 1
EleR Elevation Range (m) 1
Slop Slope (degrees) (calculated from Alti)HJan Mean relative air humidity in January at 07:00 hours (%) 2
HJul Mean relative air humidity in July at 07:00 hours (%) 2
HRan Annual relative air humidity range (%) (=|HJan-HJul|)PET Mean annual potential evapotranspiration (mm) 2
AET Mean annual actual evapotranspiration (mm) (=min[PET, Prec])Inso Mean annual insolation (hours/year) 2
SRad Mean annual solar radiation (kwh/m2/day) 2
TJan Mean temperature in January (°C) 2
TJul Mean temperature in July (°C) 2
Temp Mean annual temperature (°C) 2
TRan Annual temperature range (°C) (=TJul-TJan)
DFro Mean annual number of frost days (minimum temperature ≤ 0°C) 2
DPre Mean annual number of days with precipitation ≥ 0,1 mm 2
Prec Mean annual precipitation (mm) 2
MP24 Maximum precipitation in 24 hours (mm) 2
RMP Relative maximum precipitation (=MP24/Prec)Cont Continentality index 3
Humi Humidity index 3
PIrr Pluviometric irregularity 4
ROff Mean annual run-off (mm) 5
Perm Soil permeability 5
Lati Latitude (degrees N) 6
Long Longitude (degrees E) 6
Aqch Presence/absence of Golden Eagle 7
DHi Distance to the nearest highway (km) 6
U100 Distance to the nearest town with more than 100,000 inhabitants (km) 6
U500 Distance to the nearest town with more than 500,000 inhabitants (km) 6
96 A.-R. Muñoz & R. Real
1979). For every variable, we determined the val-ues of each UTM 10×10-km square by calculatingthe average of the values assigned to the pixelswithin the square.
Latitude and longitude were included to takeinto account the spatial structure in the distribu-tion of the species. Legendre (1993) argued thatthis structure should be included in ecologicalmodels, as it is functional in ecosystems. Spatialstructuring in species distributions may resultfrom the influence of spatially autocorrelated conditioning factors (Legendre & Fortin 1989,Borcard et al. 1992), or from pure spatial effectsdue to contagious biotic processes inherent to their own population dynamics, such as migra-tion (Legen dre 1993, Real et al. 2003, Castro et al.2008).
Statistical analyses and distribution modellingTo select a subset of significant predictors for eachmodel we related each of the variables separatelywith the distribution of the species using general-ized linear models (GLMs), and only those whoserelationship was significant with α < 0.05 wereretained. Statistical theory predicts an increase ofspurious findings when a large number of vari-ables is analysed, due to the increase of type Ierror under repeated testing (i.e. the familywiseerror rate, FWER). García (2003) recommended con -trolling the FWER in ecological research by evalu-ating the false discovery rate (FDR, Benja mini &Hochberg 1995). We controlled the FWER usingthe procedure for all forms of dependency amongtest statistics (Benjamini & Yekutieli 2001) under aFDR value of q = 0.05. We only accepted thosevariables which were significantly related to thedistribution of the species with α < 0.05 under aFDR of q < 0.05.
We then grouped the resulting significant vari-ables according to the explanatory factors towhich they were related (see predictor variables),and used them to build different uni- and two-fac-tor explanatory models for the species distributionusing the method enter to include the variables inthe model. Single-factor models were built usingthe most significant variable for each factor. Sincetopography has been found to be a consistent fac-tor influencing the distribution of Bonelli’s Eagleat different scales (Gil-Sánchez et al. 1996, Onti -veros & Pleguezuelos 2003, MuZoz et al. 2005), wetested a set of two-factor models that includedtopography combined with other factors (climate,human activity and interspecific competition) pre-viously reported as important for the species. We
compared the different models using the AkaikeInformation Criterion (AIC) (Akaike 1973), whichweighs the deviance of a model by the number ofparameters (Burnham & Anderson 2002). AIC isdefined as:
AIC = −2 logH+2h
where H refers to the value of the maximized loglikelihood and h to the number of parameters inthe model. The smaller the AIC, the better themodel. For each model we computed the AIC dif-ferences Δi (Burnham & Anderson 2002, p. 71).
We then used the environmental favourabilityfunction explained by Real et al. (2006). We select-ed this favourability function because probabilityvalues derived from logistic regression are affect-ed by the prevalence, whereas the favourabilityfunction assesses the variation in the probabilityof occurrence of the species in certain local condi-tions with respect to the overall species preva-lence, and so favourability values reflect only theenvironmental conditions which are appropriatefor the species (Acevedo & Real 2012).This function may be expressed as:
where P is the logistic probability value, n1 isthe number of presences and n0 is the number ofabsences. We so obtained the favourability valuesfor Bonelli’s Eagle in each 10×10-km square inMálaga. We preferred the concept of favourabilityto that of suitability because favourability valuesare interpretable in absolute terms, as they indi-cate how local presence’s probability differs fromthat expected by chance in the whole study area.Suitability values rank local sites according totheir capacity to hold the species but they are notrelated to probability and thus are uninformativein absolute terms.
To assess the classification accuracy of themodel we started from the values of a confusionmatrix and calculated the sensitivity (ratio of correctly predicted presences to total number of presences), specificity (ratio of correctly predicted absences to total number of absences),correct classification rate (ratio of correctly pre-dicted presences and absences to total number oflocalities), and Cohen’s kappa (Fielding & Bell1997, MuZoz & Real 2006). To take into account theinteractions between the determinant variables,which often result in an overlaid effect in space
P(1-P)
n1
n0
P(1-P)
F=+
Scale-related effect of human activity on Bonelli's Eagle distribution 97
ranged from 101.40 to 126.99 (Table 3). The evalu-ation scores of all previous models (sensitivity,specificity, correct classification rate, and Cohen´skappa) are shown in Table 4.
The lowest AIC was obtained for the modelcombining topography and human activity, andso it was considered the best model. FollowingAltman (1991) the predictions of this model agree"good" with the observations (0.6 < k < 0.8). Theresults of the model variation partitioning indicatethat 54% of variation is accounted for by elevationrange, independent to the distance to highways;24% is explained by distance to highways, inde-pendent to elevation range; and 22% is explainedby the combined effects between both factors, thatis to say, by the existence of squares that simulta-neously have a high elevation range and are locat-ed far away from highways, or have little eleva-tion range and are close to highways. Fa -vourability values for Bonelli’s Eagle in Málagaprovince UTM grid cells, for the model combiningtopography and human activity, are representedin Fig. 2, showing also the distributions of thehighways.
due to collinearity between them (Borcard et al.1992, Legendre 1993), we performed a variationpartitioning procedure to specify how much ofthe variation of the final model accounted for thepure effect of each variable, which proportion wasattributable to their interaction, and how thesevariables interact affecting the target variable(Legendre 1993, Legendre & Legendre 1998,MuZoz et al. 2007). The part of the variation of thefinal model explained by each variable (Ri
2) wasobtained by using the squared value of thePearson correlation coefficient between the valuesobtained in the final model and those yielded bythe models based on each variable included in themodel. The amount of variation explained by eachpair, trio, etc. of explanatory variables (Ri+j+…+n
2)may be obtained by correlating the final modelvalues with those yielded by the model usingthese variables. Then, the pure effect of each vari-able (RPi
2) may be assessed by subtracting the vari-ation explained by the other variables togetherfrom the variation explained by all explanatoryvariables together (RPi
2=Ri+j+ +n2-Rj+…+n
2). Thevariation attributable to the interaction of pairs ofvariables (Rij
2) may be obtained by subtractingfrom Ri+j+…+n
2 the pure effect of the two variables(RPi
2+ RPj2) and the variation explained by the
other variables together (Rk+…+n2) (see Whittaker
1984, Legendre & Legendre 1998 pp. 532–534,MuZoz et al. 2005).
RESULTS
Twelve of the variables had a univariate signifi-cant relationship with the presence of breedingterritories of Bonelli’s Eagle, and were related totopography, climate, human activity, and inter-specific competition (Table 2). The AIC values forthe single and two-factor models derived from thecombination of topography with the other factors
Table 2. Variables significantly related with the distribution ofBonelli's Eagle in Málaga province, selected with the univariateanalysis, and their significance (p). Variables are grouped intoexplanatory factors. Codes as in Table 1.
Factor Variable p
Topography Area 0.0011
Alti 0.0012
EleR 0.0000
Slop 0.0000
Climate Inso 0.0012
Temp 0.0006
DPre 0.0037
Prec 0.0044
MP24 0.0068
Humi 0.0044
Interspecific competition Aqch 0.0010
Human activity DHi 0.0000
Table 3. Most significant models explaining Bonelli's Eagle distribution obtained for each combination of explanatory factors usingthe variables in Table 2. Models are ranked according to their goodness. AIC: Akaike’s Information Criterion. Δi: AIC differences.Variable codes as in Table 1.
Explanatory factor Logit function AIC Δi
Topography + Human activity 2.49+0.07DHi+0.003EleR 101.403 0
Topography + Climate -5.52+0.04EleR+0.07DPre 104.231 2.828
Topography -1.87+0.04EleR 108.496 7.093
Topography + Interspecific competition - 108.794 7.391
Human activity -0.51+0.10DHi 121.581 20.178
Interspecific competition 0.27+2.73Aqch 125.641 24.238
Climate 9.78-0.56Temp 126.988 25.585
DISCUSSION
The environmental model for Bonelli’s Eagle dis-tribution in Málaga province is remarkably parsi-monious, since it only includes two predictors toaccount for its distribution in 104 squares. The firstvariable entering the model is elevation range,which is closely related with mountainous areas.Mountains have been consistently found to beimportant in the distribution of Bonelli’s Eagle atlocal (Gil-Sánchez et al. 1996, Sánchez-Zapata etal. 1996, Ontiveros 1999), large (Ontiveros &Pleguezuelos 2003, MuZoz et al. 2005, Carrascal &Seoane 2009), and intermediate spatial scales(López-López et al. 2006, and this paper). Thepure effect of cliff availability on the distributionof the species is similar in Málaga province (54%)and mainland Spain (53%, MuZoz et al. 2005),which confirms the profound dependence of thespecies on mountain ranges at these two differentscales in the studied areas. Mountains are normal-ly associated with nest site availability for cliffnesters (Newton 1979), and Pérez-García et al.(2013) indicated for Bonelli’s Eagles that topogra-phy and natural land-marks play an importantrole in home range segregation, helping to min-imise antagonist neighbour conflicts.
Climatic variables are absent from our interme-diate-scale model, but they were important in thenation-wide model of MuZoz et al. (2005), whichwas performed using the same resolution scale(10×10-km squares). Climate is the second factordetermining the distribution of the species inSpain, as suitable areas for Bonelli’s Eagle aremountains with a Mediterranean climate (MuZozet al. 2005). However, climate is a factor that typi-cally changes over large areas, and the climate ofMalaga province is completely Mediterranean, sothe climatic factor loses its predictive and explana-tory capacities at this intermediate study area size.As already stated Mackey & Lindenmayer (2001)the processes dominating the delivery of the pri-mary environmental resources relevant to the dis-tribution of animals are scale-specific, being theclimate the more influential factor at a largeextent, whereas topography create the finer-scalevariations in climate that influence species distri-butions (see also Elith & Leathwick 2009).
The most important difference between themodel obtained in this study (at an intermediatescale) and those made at local and nation-widescales is the role of human-related variables.Distance to highways is the second most impor-tant variable in our analysis, explaining almost25% of the species distribution at intermediatestudy area size, while it is not reflected at a largerscale. Maurer (1996) considered that roads shouldbe seen as a surrogate for human density, eco-nomic activity and intervention in the landscape,which is certainly true in Malaga, where humanactivity is clearly associated to highways.Consequently, factors which were not available onUTM 10×10-km squares, such as prey availability,a factor influencing the selection of settlementareas by juvenile Bonelli’s Eagles (MaZosa et al.1998, Moleón et al. 2009), or density of powerlines, a man-induced cause of mortality for bigraptors (Rollan et al. 2010, López-López et al.2011), could be partially affected by this variable.In general terms both abundance and occurrence
98 A.-R. Muñoz & R. Real
Table 4. Measure of performance of the best uni- and two-factor explanatory models for the species distribution, indicating sensitivity, specificity, correct classification rate, and Cohen´s kappa.
Explanatory factor Sensitivity Specificity CCR Cohen´s kappa (k)
Topography + Human activity 0.91 0.70 0.83 0.63
Topography + Climate 0.86 0.67 0.79 0.55
Topography 0.88 0.62 0.79 0.52
Topography + Interspecific competition 0.88 0.62 0.79 0.52
Human activity 0.77 0.45 0.66 0.24
Interspecific competition - - 0.64 0
Climate 0.86 0.43 0.71 0.32
Fig. 2. Favourability values for Bonelli's Eagle in each UTM10×10 km square of Málaga province, shown on a scale rangingfrom 0 (unfavourable, white) to 1 (favourable, black).Highways are also shown.
0
0.5
1
of breeding birds are depressed near major roads,representing areas with lower-quality territories(Moreno-Mateos et al. 2011, Summers et al. 2011,Silva et al. 2012). Furthermore, some diseases suchas trichomoniasis, one of the most importantnestling mortality factor for Bonelli’s eagle (Real etal. 2000), could be related to the consumption ofdomestic pigeons, and thus to human activity.
The results we present here show that the fac-tors explaining the species distribution is likely tobe scale dependent, as the ranking of the involvedvariables is dependent on the considered spatialscale. Studies on small study area sizes failed todetect any adverse effect of human disturbanceon Bonelli’s Eagle (e.g. Gil-Sánchez et al. 2004),and the national scale analyses detected this effectonly marginally (MuZoz et al. 2005) or did notnotice it (Carrascal & Seoane 2009), so the scale ofthe analyses is key for detecting the human effecton Bonelli’s Eagle distribution. Local studiessearched for disturbances in the near vicinity (< 5km) of the nesting sites, whereas the distributionof the species in Spain varies over 500,000 km2.
In our analysis, the mean distance of occupiedsquares to highways was 16.5 km, versus 7.7 km ofmean distance to highways for unoccupiedsquares (F = 17.6, p < 0.001). As this process occursat intermediate scale, local scale (< 5 km radiusaround the nest) analyses are unable to detect itbecause they are mainly affected by cliff availabil-ity, whereas broad-scale analyses (national distri-bution) are too coarse to account for it satisfactori-ly and are more influenced by climate. Althoughour model indicates that the species prefers tobreed in cliffs separated from intense human dis-turbance, more than 50% of presences are ac -counted for by the pure effect of cliff availability,independently of the distance to highways, whichrenders the effect of highways quite difficult todetect intuitively during a field survey. Interest -ingly, López-López et al. (2006), who also analysedan area of similar size (aprox. 7000 km2), alsofound some negative effect of roads on Bonelli’sEagle, although they did not quantified the rela-tive influence of human disturbance on the species.
The influence of human activity in Málagaprovince, where the present status of the speciesis considered to be optimal is of particular con-cern. This province is part of a favourable areafrom which juveniles may disperse to other, lessfavourable territories (MuZoz et al. 2005), whichmeans that human disturbance in this area may have far-reaching effects. In fact, a satellitetracking study (Cadahía et al. 2009) demonstrates
Scale-related effect of human activity on Bonelli's Eagle distribution 99
high distances for natal dispersal in this species,with birds born in favourable areas attempting thefirst breeding in less favourable areas separatedmore than 400 km. Several EU funded LIFE proj-ects have aimed at promoting Bonelli’s Eagle con-servation although, unfortunately, the effectsassociated to highways are not easily amelioratedby this kind of conservation efforts. Since morehighways are planned for the near future inMálaga, and other favourable areas in Spain, weshould take into account that an abusive develop-ment has a cost in biodiversity, even for a “humantolerant” species.
In summary, our results show that topographyis the main driver of Bonelli’s Eagle breeding dis-tribution at all scales, climate is a major driver onlyat a broad scale, whereas human impact on thespecies distribution, and probably on the factorsaffecting human-induced mortality, is only detect-ed as a major driver at intermediate scale. In thisway, systematic analyses of the relative impor-tance of the different factors affecting the distribu-tion of a species as a function of spatial scale mayprovide new insights in distribution modellingstudies and a better understanding of the patternsof occurrence and most likely in the abundance ofa species.
ACKNOWLEDGEMENTS
We are grateful to M. Skierczynski for constructivecomments on the manuscript. The authors wouldlike to thank FEDER of European Union andAgencia de Obra Pública-Consejería de ObrasPúblicas y Vivienda, Junta de Andalucía, for finan-cial support (project G-GI3000/IDIG).
REFERENCES
Acevedo P., Real R. 2012. Favourability: concept, distinctivecharacteristics and potential usefulness. Naturwissen -schaften 99: 515–522.
Akaike H. 1973. Information theory and an extension of themaximum likelihood principle. In: Petrov B. N., Csaki F.(eds). Proceedings of the Second International Symposiumon Information Theory. Akademiai Kiado, Budapest, pp.267–281.
Altman D. G. 1991. Practical statistics for medical research.Chapman and Hall. London.
Austin M. P. 2002. Spatial prediction of species distribution: aninterface between ecological theory and statistical model-ling. Ecol. Model. 157: 101–118.
Balbontín J., Penteriani V., Ferrer M. 2003. Variations in the ageof mates as an early warning signal of changes in popula-tion trends? The case of Bonelli’s Eagle in Andalusia. Biol.Conserv. 109: 417–423.
Bautista J., MuZoz A. R., Jiménez J. J., Luque J. J., Fernández F.2003. [Malaga main Iberian sanctuary for Bonelli’s eagle].Quercus 204: 18–22.
Benjamini Y., Hochberg Y. 1995. Controlling the false discoveryrate: a practical and powerful approach to multiple testing.J. Roy. Stat. Soc. B 57: 289–300.
Benjamini Y., Yekutieli D. 2001. The control of the false discov-ery rate in multiple testing under dependency. Ann. Stat.29: 1165–1188.
BirdLife International 2004. Birds in Europe: population esti-mates, trends and conservation status. BirdLife Conserva -tion Series n 12. BirdLife International, Cambridge.
Blanco J. C., González J. L. 1992. [Red List of SpanishVertebrates]. Colección Técnica ICONA. Madrid.
Borcard D., Legendre P., Drapeau P. 1992. Partialling out thespatial component of ecological variation. Ecology 73:1045–1055.
Brown J. H. 1984. On the relationship between abundance anddistribution of species. Am. Nat. 124: 255–279.
Burnham K. P., Anderson D. R. 2002. Model selection and mul-timodel inference: a practical information-theoreticapproach. 2nd ed. Springer-Verlag. New York.
Cadahía L., López-López P., Urios V., Soutullo Á., Negro J. J.2009. Natal dispersal and recruitment of two Bonelli’sEagles Aquila fasciata: a four-year satellite tracking study.Acta Ornithol. 44: 193–198.
Carrascal L. M., Seoane J. 2009. Factors affecting large-scale dis-tribution of the Bonelli’s Eagle Aquila fasciata in Spain. Ecol.Res. 24: 565–573.
Carrete M., Sánchez-Zapata J. A., Martínez J. E., Palazón J. A.,Calvo J. F. 2001. [Spatial patterns of distribution ofBonelli`s Eagles (Hieraaetus fasciatus) and Golden Eagles(Aquila chrysaetos) in southeastern Spain]. Ardeola 48:175–182.
Carrete M., Sánchez-Zapata J. A., Martínez J. E., Sánchez M. A.,Calvo J. F. 2002. Factors influencing the decline of aBonelli’s Eagle Hieraaetus fasciatus population in southeast-ern Spain: demography, habitat or competition? Biodivers.Conserv. 11: 975–985.
Castro A., MuZoz A. R., Real R. 2008. Modelling the spatial dis-tribution of Tengmalm’s owl Aegolius funereus in itsSouthwestern Palaeartic limit (NE Spain). Ardeola 55:71–85.
Del Moral J. C. 2006. [The Bonelli’s Eagle in Spain: populationin 2005 and census method]. SEO/BirdLife. Madrid. avail-able at http://www.seo.org/media/fotos/ 9_perdicera05.pdf(accessed 2 November 2012).
Elith J., Leathwick J. R. 2009. Species Distribution Models: eco-logical explanation and prediction across space and time.Annu. Rev. Ecol. Evol. Syst. 40: 677–697.
Fielding A. H., Bell J. F. 1997. A review of methods for theassessment of prediction errors in conservationpresence/absence models. Environ. Conserv. 24: 38–49.
Font I. 1983. [Climatic atlas of Spain]. Instituto Nacional deMeteorología. Madrid.
Font I. 2000. [Climatology for Spain and Portugal]. EdicionesUniversidad de Salamanca. Salamanca.
García L. V. 2003. Controlling the false discovery rate in ecolog-ical research. Trends Ecol. Evol. 18: 553–554.
Gil-Sánchez J. M., Moleón M., Otero M., Bautista J. 2004. Anine-year study of successful breeding in a Bonelli’s Eaglepopulation in southeast Spain: a basis for conservation.Biol. Conserv. 118: 685–694.
Gil-Sánchez J. M., Molino F., Valenzuela G. 1996. [Nest-siteselection by Bonelli’s Eagle (Hieraaetus fasciatus) inGranada (southeastern Spain)]. Ardeola 43: 189–197.
Hengeveld R. 1990. Dynamic Biogeography. CambridgeUniversity Press. Cambridge.
IGME 1979. [Spanish hydrogeological map. Explanations ofrainfall maps, hydrogeological survey and groundwatersystems]. 2nd edition. Instituto Geológico y Minero deEspaZa. Madrid.
IGN 1999. [Road map. Iberian Peninsula, Balearics and CanaryIslands]. Instituto Geográfico Nacional/Ministerio deFomento. Madrid.
Jiménez J. J., MuZoz A. R. (eds). 2008. [Atlas of raptors in theprovince of Málaga (breeding, migration and wintering)].Centro de Ediciones de la Diputación de Málaga. Málaga.
Legendre P. 1993. Spatial autocorrelation: trouble or new para-digm? Ecology 74: 1659–1673.
Legendre P., Fortin M. J. 1989. Spatial pattern and ecologicalanalysis. Vegetatio 80: 107–138.
Legendre P., Legendre L. 1998. Numerical Ecology. 2nd ed.Elsevier Science, Amsterdam.
López-López P., Ferrer M., Madero A., Casado E., McGrady M.2011. Solving man-induced large-scale conservation prob-lems: the Spanish Imperial Eagle and power lines. PLOSONE 6 (3): e17196. doi:10.1371/journal.pone.0017196.
López-López P., García-Ripollés C., Aguilar J. M., García-LópezF., Verdejo J. 2006. Modelling breeding habitat preferencesof Bonelli’s Eagle (Hieraaetus fasciatus) in relation to topog-raphy, disturbance, climate and land use at different spa-tial scales. J. Ornithol. 147: 97–107.
López-López P., Sarŕ M., Di Vittorio M. 2012. Living on theedge: assessing the extinction risk of critically endangeredBonelli’s Eagle in Italy. PLOS ONE 7(5): e37862.doi:10.1371/journal.pone.0037862.
Mackey B. G., Lindenmayer D. B. 2001. Towards a hierarchicalframework for modelling the spatial distribution of ani-mals. J. Biogeogr. 28: 1147–1166.
MadroZo A., González C., Atienza J. C. (eds). 2004. [Redbook ofthe birds of Spain]. Ministerio de Medio Ambiente-Sociedad EspaZola de Ornitología. Madrid.
MaZosa S., Real J., Codina J. 1998. Selection of settlement areasby juvenile Bonelli’s Eagle in Catalonia. J. Raptor Res. 3:208–214.
Maurer B. A. 1996. Relating human population growth to theloss of biodiversity. Biodiv. Lett. 3: 1–5.
Moleón M., Bautista J., Sánchez-Zapata J. A., Gil-Sánchez J. M.2009. Diet of non-breeding Bonelli’s Eagles Hieraaetus fas-ciatus at settlement areas of southern Spain. Bird Study 56:142–146.
Montero de Burgos J. L., González-Rebollar J. L. 1974. [Climaticdiagrams]. ICONA. Madrid.
Moreno-Mateos D., Rey-Benayas J. M., Pérez-Camacho L., dela MontaZa E., Rebollo S., Cayuela L. 2011. Effects of landuse on nocturnal birds in a Mediterranean agriculturallandscape. Acta Ornithol. 46: 173–182.
MuZoz A. R., Altamirano M., Takasu F., Nakamura H. 2007.Nest light environment and the potential risk of CommonCuckoo (Cuculus canorus) parasitism. Auk 124: 619–627.
MuZoz A. R., Real R. 2006. Assessing the potential range expan-sion of the exotic monk parakeet in Spain. Divers. Distrib.12: 656–665.
MuZoz A. R., Real R., Barbosa A. M., Vargas J. M. 2005.Modelling the distribution of Bonelli’s Eagle in Spain:Implications for conservation planning. Divers. Distrib. 11:477–486.
Newton I. 1979. Population Ecology of Raptors. T and ADPoyser. Berkhamsted.
Niamir A., Skidmore A. K., Toxopeus A. G., MuZoz A. R., RealR. 2011. Finessing atlas data for species distribution mod-els. Divers. Distrib. 17: 1173–1185.
Ontiveros D. 1999. Selection of nest cliffs by Bonelli’s Eagle(Hieraaetus fasciatus) in southeastern Spain. J. Raptor Res.33: 110–116.
100 A.-R. Muñoz & R. Real
Ontiveros D., Pleguezuelos J. M. 2003. Influence of climate onBonelli’s Eagle’s (Hieraaetus fasciatus V. 1822) breeding suc-cess through the western Mediterranean. J. Biogeogr. 30:755–760.
Orians G. H., Wittenberger J. F. 1991. Spatial and temporalscales in habitat selection. Am. Nat. 137: 29–49.
Pérez-García J. M., Margalida A., Afonso I., Ferreiro E.,Gardiazábal A., Botella F., Sánchez-Zapata J. A. 2013.Interannual home range variation, territoriality and over-lap in breeding Bonelli’s Eagles (Aquila fasciata) tracked byGPS satellite telemetry. J. Ornithol. 154: 63–71.
Ponchon C. 2011. World distribution and dynamic of theBonelli’s eagle Mediterranean population Aquila fasciata.In: Scher O., Lecacheur M. (eds). La conservation de l’Aiglede Bonelli. Actes du colloque international, Montpellier,pp. 14–15.
Rahbek C., Gotelli N. J., Colwell R. K., Entsminger G. L., RangelT. G. L. V. B., Graves G. R. 2007. Predicting continental-scalepatterns of bird species richness with spatially explicitmodels. Proc. R. Soc. B 274: 165–174.
Real J., MaZosa S. 1997. Demography and conservation ofWestern European Bonelli’s Eagle (Hieraaetus fasciatus)populations. Biol. Conserv. 79: 59–66.
Real J., MaZosa S., MuZoz E. 2000. Trichomoniasis in a Bonelli’sEagle population in Spain. J. Wildl. Dis. 36: 64–70.
Real R., Barbosa A. M., Porras D., Kin M. S., Márquez A. L.,Guerrero J. C., Palomo L. J., Justo E. R., Vargas J. M. 2003.Relative importance of environment, human activity andspatial situation in determining the distribution of terres-trial mammal diversity in Argentina. J. Biogeogr. 30:939–947.
Real R., Barbosa A. M., Vargas J. M. 2006. Obtaining environ-mental favourability functions from logistic regression.Environ. Ecol. Stat. 13: 237–245.
Rico L., Sánchez-Zapata J. A., Izquierdo A., García J. R., MoránS., Rico D. 1999. [Recent trends in the status of GoldenAquila chrysaetos and Bonelli’s Eagles Hieraaetus fasciatus inValencia]. Ardeola 46: 235–238.
Robertson M. P., Peter C. I., Villet M. H., Ripley B. S. 2003.Comparing models for predicting species’ potential distri-butions: a case study using correlative and mechanisticpredictive modelling techniques. Ecol. Model. 164:153–167.
Rollan A., Real J., Bosch R., Tintó A., Hernández-Matías A.2010. Modelling the risk of collision with power lines inBonelli’s Eagle Hieraaetus fasciatus and its conservationimplications. Bird Conserv. Int. 20: 279–294.
Sánchez-Zapata J. A. 1996. [Recovery Plan for Bonelli’s Eagle inMurcia, Spain]. Ambiental S.L.-Consejería de Agricultura,Agua y Medio Ambiente. Murcia.
Sergio F., Pedrini P., Marchesi L. 2003. Spatio-temporal shifts ingradients of habitat quality for an opportunistic avianpredator. Ecography 26: 243–255.
Silva C. C., Lourenço R., Godinho S., Gomes E., Sabino-Marques H., Medinas D., Neves V., Silva C., Rabaça J. E.,Mira A. 2012. Major roads have a negative impact on theTawny Owl Strix aluco and the Little Owl Athene noctuapopulations. Acta Ornithol. 47: 47–54.
Summers P. D., Cunnington G. M., Fahrig L. 2011. Are the neg-ative effects of roads on breeding birds caused by trafficnoise? J. Appl. Ecol. 48: 1527–1534.
US Geological Survey 1996. GTOPO30. Land ProcessesDistributed Active Archive Center (LP DAAC), EROS Data Center. http://edcdaac.usgs.gov/gtopo30/gtopo30.asp.Cited 22 Sep 1999.
Whittaker J. 1984. Model interpretation from the additive ele-ments of the likelihood function. Appl. Statist. 33: 52–64.
STRESZCZENIE
[Czynniki wpływające na występowanie orzełkapołudniowego w południowej Hiszpanii].Identyfikacja czynników, które mogą wpływać narozmieszczenie zagrożonych gatunków zwierzątwymaga zastosowania różnych podejść oraznarzędzi analitycznych, gdyż ich znaczenie możezależeć od skali przestrzennej, w jakiej są one roz-patrywane. Do tej pory rozmieszczenie orzełkapołudniowego badano w skali lokalnej (w oparciuo miejsca lęgowe) oraz w skali całej Hiszpanii. W obu tych skalach nie stwierdzono, aby różniedefiniowane zmiany powodowane działaniamiczłowieka miały istotny wpływ na zajmowanieterytoriów lęgowych.
W pracy analizowano czynniki potencjalniewpływające na występowanie orzełka, rozpatru-jąc je w skale regionalnej. Analizy oparto orozmieszczenie lęgowych par w prowincji Malaga,gdzie zagęszczenie tych ptaków jest najwyższe w Europie. Określano czynniki potencjalniewpływające na występowanie orzełka, na pod-stawie występowania par lęgowych w 104kwadratach 10 × 10 km. Pod uwagę wzięto 31 zmiennych (Tab. 1), przyporządkowanych następ-nie do grup zmiennych opisujących topo grafię,klimat, konkurencję międzygatunkową (określa -ną jako obecność orła przedniego) oraz działal-ność człowieka (Tab. 2).
Występowanie orzełka najlepiej wyjaśnianebyło przez czynniki związane z topografią (róż -nice w wysokości) oraz odległością od autostrad(Tab. 3). Model uwzględniający te czynniki byłnajlepiej dopasowany do danych (Tab. 4). W prze-ciwieństwie do wyników wcześniejszych pracwykazano negatywny wpływ działalności czło -wieka — tereny położone bliżej autostrad byłymniej sprzyjające występowaniu orzełka (Fig. 2).Autorzy wskazują, ze dalsza planowana rozbu-dowa sieci autostrad może negatywnie wpływaćna ten zagrożony gatunek.
Scale-related effect of human activity on Bonelli's Eagle distribution 101