Centro Andaluz para la Evaluación y Seguimiento del · PDF fileMaría Jacoba, Yolanda Cantón y Paco Domingo son de esas personas que basta con que pasen cerca de tu vida para que

Centro Andaluz para la Evaluación y Seguimiento del Cambio

Global

Universidad de Almería

Ecología del Tejón europeo (Meles meles) en paisajes áridos

Mediterráneos

Avanzando en el conocimiento de su distribución espacial, hábitos

alimenticios y predicción de su distribución futura frente al cambio

climático

Ecology of the European badger (Meles meles) in Mediterranean arid

landscapes

Advancing knowledge of its spatial distribution, feeding habits, and

forecasting future climate driven distributions

Memoria presentada por D. Juan Miguel Requena Mullor para optar al

Grado de Doctor en Ciencias Biológicas por la Universidad de Almería

Esta tesis ha sido dirigida por el Dr. Enrique López Carrique, Profesor

Contratado Doctor del Departamento de Educación de la Universidad de

Almería y codirigida por el Dr. Hermelindo Castro Nogueira, Profesor

Titular del Departamento de Biología y Geología de la Universidad de

Almería, y por el Dr. Antonio J. Castro Martínez, investigador en el

Oklahoma Biological Survey y Profesor Asociado en la Universidad de

Oklahoma (EEUU).

Vº Bº Director Tesis Vº Bº Co-Director Tesis Vº Bº Co-Director Tesis

Enrique López Carrique Hermelindo Castro Nogueira Antonio J. Castro Martínez

Febrero 2015

A mis padres Juan y Juani,

mi esposa Alicia e hijas Lola y Macarena.

Quién lo diría, los débiles de veras nunca se rinden. (Rincón de Haikus,

1999).

Mario Benedetti, (1920-2009).

1

AGRADECIMIENTOS

Probablemente ésta será la sección más leída con diferencia de toda la

tesis doctoral, por lo que intentaré que algo de ella se os quede en la

memoria. Empecemos...

La cualidad más valiosa que debe poseer una persona es la imaginación.

A partir de este dogma nace el concepto del número imaginario i

(notación empleada para referirse a la raíz cuadrada de -1). En el pasado,

científicos y sabios interpretaban la aparición del número i en la

resolución de una ecuación como un síntoma de que dicha ecuación no

tenía solución. Sin embargo, lejos de desalentarse en su resolución

continuaron tratando al número i como al resto de números. De esta

forma, con un poquito de paciencia y saber hacer al final el número i

desaparecía y la ecuación se resolvía.

Tras los números imaginarios hay un mensaje muy importante. Ninguna

dificultad es insalvable. Si no puedes derribar un muro, sáltalo, rodéalo

pero no te pares y des media vuelta. Porque detrás de él existen

soluciones. De hecho, la raíz cuadrada de -1 no tiene solución real pero

aún así la puedes llevar como compañera de viaje que al final del camino

todo se arregla.

A lo largo del desarrollo de esta tesis doctoral han aparecido problemas

de todo tipo, algunos de ellos siguen sin solución aún hoy, pero aquí la

tienes entre tus manos (menos probablemente) o en la pantalla de tu

ordenador (más probablemente). Durante este proceso de formación

han pasado por mi vida numerosas personas con los bolsillos repletos de

números i, pero seguían incansablemente hacia adelante a pesar de

todo. Y yo... impregnado de ese ímpetu aprendí a convivir con la raíz

cuadra de -1.

Durante este tiempo he aprendido tantas cosas que puedo decir sin

temor a errar que antes de comenzar la tesis no sabía absolutamente

nada. La oportunidad de ser doctor me cogió ya con unos años y una

familia a mi lado, y por este motivo, el objetivo de conseguirlo se

convirtió desde el principio en una necesidad obsesiva, pero a la vez, en

una oportunidad irrepetible.

2

Todo comenzó con Emilio González Miras, amigo y compañero en

aquellos locos años universitarios de Granada en los que mi prioridad en

la vida era bien distinta a la que es hoy. Él fue quien me abrió la puerta a

esta aventura. El siguiente escalón lo subí de la mano de Enrique López

(director de esta tesis), responsable de dar el sí quiero a mi entrada en

esta profesión. Ha sabido siempre comprender y ceder el espacio que yo

necesitaba para mi desarrollo como doctor.

Siguiendo el orden cronológico de las personas que han formado parte

de esta tesis llegamos a Antonio Castro (Codirector). El azar juguetón del

destino corrió de mi lado e hizo que entrara en escena una de las

personas más honestas y buenas que he conocido jamás. Desde el primer

minuto que compartimos despacho en el fondo de la segunda planta del

edificio CITE-IIB se ofreció a estar a mi lado. Ha sido un apoyo

imprescindible para mí, guía en todos los acontecimientos y fases por las

que he ido pasando y amigo incondicional y sincero. Nunca tendré

tiempo suficiente para agradecerte lo que me has ayudado.

Eje clave en todos estos años ha sido y es Hermelindo Castro

(Codirector). Su papel es claro y determinante: “es el chamán de la

tribu”. Sin su “magia” nada de esto sería posible. Además, posee un

sentido común y calidad humana de las que dejan huella.

Emilio Virgós, la quinta pata de la tesis. Su humanidad y criterio científico

son apabullantes. Todo un lujo trabajar con él, Emilio es al Tejón lo que

Manolo Caracol al flamenco: un creador.

Javier Cabello, un inmejorable compañero para discutir de ciencia, un

enamorado de su trabajo, gran clarificador de ideas y estimulador de

pensamiento, gracias por estar ahí.

Cecilio Oyonarte, pa' comérselo... sin duda una de las personas con las

que más conecto profesional e intelectualmente, gracias por confiar en

mí.

Domingo Alcaraz, el apoyo en la sombra, un referente, gracias por creer

en mí.

3

Mis queridas/os compañeras/os de batalla; María López (jamón jamón,

acojonante persona, la mejor), Patricia (siempre dispuesta a ayudar y a

escuchar), Ricardo (Don Antonio, chapó, el compañero que todos

soñamos tener), Emilio Rodríguez (otra bestia de persona, el pankiteras

mejor persona de la escena alternativa), Andrew Reyes (rockero de

corazón noble y más cerca de los satélites que del planeta tierra), Cristina

Quintas (la eterna sonrisa), Mar Molina (esa chispa que cualquier lugar

de trabajo que se precie necesita), las chicas y chicos del “tupper

revolution” (muy buenos postres, confesiones y risas que nunca

olvidaré): Josema (un crack), Eva (…), José Luis (donador desinteresado),

Ismael (muy buen tío), Sonia (mmm), Olga (pedazo de hembra), Ángela

(mi alma gemela).

María Jacoba, Yolanda Cantón y Paco Domingo son de esas personas que

basta con que pasen cerca de tu vida para que quieras que se metan de

lleno.

Mis compañerAs de Nexa (Enemies and eXpectations for Agroecology),

Jordi, Estefanía, Marta, Mónica y Eva. Formar parte de este equipo es

como tocar en el sexteto de Paco de Lucía, un privilegio que la vida me

ha dado.

Reyes Tirado está también entre estas páginas. Mi gran amiga y la

primera persona que me dio la oportunidad de inmiscuirme en el mundo

de la ecología. ¡Qué vueltas da la vida Amparito!.

¡Y llegamos al final!.

Si a alguien debo agradecer que yo haya sido capaz es a Alicia, mi mujer.

Ella lo es todo, sin ella no hay rumbo, no hay un por qué, es mi gran

maestra en ser feliz. (Me pediste un beso y yo… te di toda una vida).

A mis hijas Lola y Macarena, cuando leáis esto dentro de unos años solo

espero que el tiempo que os he robado haya merecido la pena y que

hayamos desplazado al destino un poquito a nuestro favor.

Mis hermanas Carmen y Taté, siempre estaréis ahí.

4

A mis padres Juan y Juani porque sé que esto significa mucho para

vosotros; esta tesis comenzó con vuestro sacrificio, entrega y afán de

superación. ¡Gracias eternas!.

…….. Es un libro que habla de lo que hablan casi todos los libros —continuó el

viejo—. De la incapacidad que las personas tienen para escoger su propio

destino. Y termina haciendo que todo el mundo crea la mayor mentira del

mundo.

—¿Cuál es la mayor mentira del mundo? —indagó, sorprendido, el muchacho.

—Es ésta: en un determinado momento de nuestra existencia, perdemos el

control de nuestras vidas, y éstas pasan a ser gobernadas por el destino. Ésta es

la mayor mentira del mundo.

Fragmento extraído de “El Alquimista”. Paulo Coelho.

5

RESUMEN

El Tejón europeo es un carnívoro mustélido de mediano tamaño con una

amplia distribución en la Península Ibérica. Sin embargo, los ambientes

áridos Mediterráneos se encuentran lejos de su óptimo centro europeo.

Por ello, la supervivencia de la especie puede verse amenazada en el

futuro como consecuencia de los efectos derivados del Cambio Global.

Esta tesis doctoral supone un avance sobre tres rasgos clave de la

ecología del Tejón europeo en paisajes áridos Mediterráneos: (1) ¿qué

factores ambientales impulsan su distribución espacial?, (2) ¿cuál es la

variabilidad de los hábitos alimenticios de la especie en un contexto árido

Mediterráneo?, y (3) ¿qué cambios potenciales en su distribución

espacial pueden derivarse ante escenarios climáticos futuros?.

Los resultados obtenidos identifican la dinámica espacio-temporal de la

producción primaria como uno de los factores que impulsa la distribución

espacial de la especie. Los modelos de distribución obtenidos muestran

que los paisajes con mayor idoneidad de hábitat para el Tejón son

aquellos que poseen una producción vegetal elevada, espacialmente

heterogénea y poco variable a lo largo del año. En este sentido, las

variables derivadas de la teledetección y relacionadas con el

funcionamiento ecosistémico describen muy bien estos paisajes,

mejorando considerablemente la fiabilidad de la distribución predicha

por los modelos.

Aunque el comportamiento alimenticio del Tejón en paisajes

Mediterráneos ha sido descrito principalmente como frugívoro, los

recursos tróficos clave explotados por la especie varían

significativamente entre paisajes con diferente cobertura vegetal y uso

6

del suelo. Así, insectos, algarrobas y micromamíferos fueron relevantes

en el paisaje de maquia; higos y naranjas en el matorral xérico y

lombrices e insectos en la media montaña arbolaba.

Debido a que los principales ítems consumidos por el Tejón dependen del

clima y del uso que el hombre hace del suelo, sus hábitos alimenticios

podrían modificarse como consecuencia de la aridificación,

intensificación de los cultivos y/o el abandono rural. De forma particular,

la calidad del hábitat para el Tejón podría verse significativamente

reducida en algunos paisajes agrícolas a finales del siglo XXI como

consecuencia de una homogeneización espacial en la producción

primaria.

Por consiguiente, es necesario el seguimiento y conservación de su

calidad de hábitat mediante iniciativas encaminadas al mantenimiento

del paisaje rural Mediterráneo. Para tal fin, las políticas diseñadas con el

objetivo de prevenir el abandono rural y preservar la heterogeneidad de

los paisajes agrícolas tradicionales resultan de gran importancia. Así

mismo, el seguimiento de la dinámica espacio-temporal de la producción

primaria a través de herramientas derivadas de la teledetección puede

ayudar a identificar zonas susceptibles de disminuir su idoneidad de

hábitat y a optimizar así programas de seguimiento de la especie.

Una síntesis de los principales resultados de la Tesis doctoral y otros

recursos gráficos pueden ser consultados a través del siguiente enlace

web:

http://european-badger-mediterranean-landscapes.site40.net/index.html

7

SUMMARY

The European badger (Meles meles, family Mustelidae) is a medium-sized

carnivore that is widely distributed in Europe. Arid Mediterranean

environments at the western extent of its range, however, differ greatly

from the optimal habitat conditions of Central Europe. For this reason,

Global Change may adversely affect the survival of badgers in the Iberian

Peninsula.

This PhD thesis addresses this concern using spatial distribution models.

The analyses considered three key aspects of the ecology of the badger

in Mediterranean arid landscapes: (1) environmental factors meditating

its spatial distribution, (2) the variability of badger feeding habits, and (3)

potential changes in distribution under future climate scenarios.

The analysis of environmental factors indicates that spatio-temporal

dynamics of primary production is a leading factor affecting the spatial

distribution of badgers. High quality habitats tended to occur in

landscapes with high primary production and spatial heterogeneity

throughout the year. The remote sensing-derived variables and related

with the ecosystem functioning depict very well these landscapes and

their incorporation in the models improves the performance of the

predicted distribution.

Although the badger is considered a frugivore in much of its

Mediterranean range, the results of this thesis demonstrate that its

feeding behavior is more complex in the arid Iberian Peninsula.

Specifically, feeding behavior differs with land cover and land use. In this

region, insects, carobs and small mammals were important in the

8

maquia; figs and oranges in the xeric shrubland and earthworms and

insects in mid-elevation forests.

This PhD thesis discusses that the availability and abundance of food

items for the badger, which are dependent on climate and land use, may

be altered by aridification, intensification of crop production, and/or

rural abandonment. In particular, the habitat quality for the badger may

decrease significantly in some agricultural landscapes due to a spatial

homogenization of primary production, as predicted for the end of this

century.

Therefore, monitoring and conservation of habitat quality are necessary

in the Mediterranean rural landscapes to insure the persistence of this

species. To achieve this objective, policies designed to prevent rural

emigration and promote the preservation of the heterogeneity of

traditional agricultural landscapes are crucial. Likewise, the monitoring of

spatio-temporal dynamics of primary production by remote sensing

could identify areas of decreasing habitat suitability.

A synthesis of the main results of this PhD Thesis and some graphic

resources can be accessed by the following web link:

http://european-badger-mediterranean-landscapes.site40.net/index.html

9

ÍNDICE

AGRADECIMIENTOS .......................................................................................... 1

RESUMEN ......................................................................................................... 5

SUMMARY ........................................................................................................ 7

ÍNDICE ............................................................................................................... 9

ÍNDICE DE TABLAS ........................................................................................... 12

ÍNDICE DE FIGURAS ......................................................................................... 14

1. INTRODUCCIÓN........................................................................................... 18

1.1 EL TEJÓN EUROPEO: RASGOS DE SU ECOLOGÍA EN UN CONTEXTO ÁRIDO MEDITERRÁNEO

................................................................................................................... 19

1.2 JUSTIFICACIÓN ........................................................................................... 24

1.3 OBJETIVO GENERAL E HIPÓTESIS DE TRABAJO ..................................................... 25

2. ÁREA DE ESTUDIO ....................................................................................... 28

3. MATERIAL Y MÉTODOS ............................................................................... 35

3.1 BLOQUE I: TRABAJO DE CAMPO Y LABORATORIO ............................................... 36

3.2 BLOQUE II: INFORMACIÓN SATELITAL Y FUNCIONAMIENTO ECOSISTÉMICO ............... 37

3.3 BLOQUE III: ANÁLISIS ESTADÍSTICO ............................................................... 38

3.3.1 Técnicas paramétricas .................................................................... 39

3.3.2 Técnicas no paramétricas ............................................................... 40

4. RESULTADOS ............................................................................................... 42

RESULTADO 4.1: MODELING SPATIAL DISTRIBUTION OF EUROPEAN BADGER IN

ARID LANDSCAPES: AN ECOSYSTEM FUNCTIONING APPROACH .................... 42

ABSTRACT ...................................................................................................... 44

4.1.1 INTRODUCTION ....................................................................................... 45

4.1.2 MATERIAL AND METHODS .......................................................................... 48

4.1.2.1 Study area ................................................................................... 48

4.1.2.2 Field survey data .......................................................................... 48

4.1.2.3 Environmental data ..................................................................... 49

4.1.2.4 Model building............................................................................. 52

4.1.2.5 Model evaluation ......................................................................... 53

4.1.3 RESULTS ................................................................................................ 56

4.1.3.1 Occurrence of European badger ................................................... 56

4.1.3.2 Threshold-independent test ......................................................... 57

4.1.3.3 Information criteria...................................................................... 58

10

4.1.3.4 Relevant variables and their effects .............................................. 59

4.1.4 DISCUSSION ........................................................................................... 61

4.1.4.1 Did the EVI-derived variables improve ecological niche modeling of

the European badger in arid landscapes?................................................. 61

4.1.4.2 Was the predicted spatial distribution across arid lands consistent

with the ecological preferences of the European badger? ........................ 63

4.1.4.3 Ecosystem functional dimension in species ecological modeling and

conservation ........................................................................................... 65

RESULTADO 4.2: FEEDING HABITS OF EUROPEAN BADGER (MELES MELES) IN

MEDITERRANEAN ARID LANDSCAPES ........................................................... 68

ABSTRACT ...................................................................................................... 70

4.2.1 INTRODUCTION ....................................................................................... 71

4.2.2 MATERIALS AND METHODS ........................................................................ 74

4.2.2.1 Localization and description of landscapes ................................... 74

4.2.2.2 Diet analysis ................................................................................ 78

4.2.2.3 Data analyses .............................................................................. 79

4.2.3 RESULTS ................................................................................................ 80

4.2.3.1 Diet composition .......................................................................... 80

4.2.3.2 Effects of landscape and season on diet ....................................... 83

4.2.3.3 Earthworm consumption .............................................................. 86

4.2.3.4 Diet diversity and food consumption ............................................ 87

4.2.3.5 Nonparametric multidimensional scaling ..................................... 87

4.2.4 DISCUSSION ........................................................................................... 89

4.2.4.1 Spatial and temporal variation of badger diet .............................. 89

4.2.4.2 Potential implications of climatic change and land use change on

feeding habits ......................................................................................... 92

RESULTADO 4.3: MODELING AND MONITORING HABITAT QUALITY FROM

SPACE: THE EUROPEAN BADGER .................................................................. 95

ABSTRACT ...................................................................................................... 97

4.3.1 INTRODUCTION ....................................................................................... 99

4.3.2 MATERIALS AND METHODS ...................................................................... 102

4.3.2.1 Species, study area and presence records ................................... 102

4.3.2.2 Environmental variables............................................................. 103

4.3.2.3 Modeling approach .................................................................... 104

4.3.2.4 Spatial distribution modeling ..................................................... 105

4.3.2.5 Model evaluation and variable relative importance .................... 106

4.3.2.6 Future forecasting ..................................................................... 107

4.3.2.7 Comparing current and future spatial distributions: identification of

sensitive areas to loss habitat suitability ................................................ 108

4.3.2.8 Limiting factors of habitat suitability .......................................... 109

11

4.3.3 RESULTS .............................................................................................. 110

4.3.3.1 Model performance and current projected distribution ............... 110

4.3.3.2 Forecasted future distributions................................................... 112

4.3.3.3 Potential areas to lose habitat suitability and involved

environmental drivers ........................................................................... 114

4.3.4 DISCUSSION ......................................................................................... 117

4.3.4.1 EVI descriptors of ecosystem functioning to forecast species

distributions .......................................................................................... 117

4.3.4.2 Global and local implications for wildlife monitoring and

management ........................................................................................ 120

5. DISCUSIÓN ................................................................................................ 122

6. CONCLUSIONES ......................................................................................... 131

REFERENCIAS ................................................................................................ 132

ANEXOS ........................................................................................................ 158

RESULTADO 4.1: MODELING SPATIAL DISTRIBUTION OF EUROPEAN BADGER IN ARID

LANDSCAPES: AN ECOSYSTEM FUNCTIONING APPROACH ......................................... 158

Appendix A ............................................................................................ 158

Appendix B ............................................................................................ 159

Appendix C ............................................................................................ 160

Appendix D............................................................................................ 161

Appendix E ............................................................................................ 162

RESULTADO 4.2: FEEDING HABITS OF EUROPEAN BADGER (MELES MELES) IN

MEDITERRANEAN ARID LANDSCAPES .................................................................. 163

Appendix F ............................................................................................ 163

Appendix G ........................................................................................... 168

Appendix H ........................................................................................... 171

RESULTADO 4.3: MODELING AND MONITORING HABITAT QUALITY FROM SPACE: THE

EUROPEAN BADGER........................................................................................ 172

Appendix I ............................................................................................. 172

Appendix J ............................................................................................. 174

12

ÍNDICE DE TABLAS

4. RESULTADOS

RESULTADO 4.1: MODELING SPATIAL DISTRIBUTION OF EUROPEAN

BADGER IN ARID LANDSCAPES: AN ECOSYSTEM FUNCTIONING

APPROACH

Table 4.1.1. Groups of variables used for constructing models. Group 1:

topography and climate, group 2: Land cover and Land uses, group 3: EVI

variables, group 4: EVI of land cover. Each model contained group 1, the

ALL model all four groups, LC & LU model groups 1 and 2, the EVI model

groups 1 and 3, and the EVI LC model groups 1 and 4. Thus, three models

included the ecosystem functional variables: EVI, EVI LC and ALL, and

only LC & LU model did not include these variables. .................................... 55

Table 4.1.2. Comparison of threshold-independent receiver operating

characteristic (ROC) results for European badger using LC & LU, EVI, EVI

LC and ALL models. For each random partition of occurrence records, the

maximum AUCPO is marked in bold, the minimum underlined, and if the

observed difference between the maximum AUCPO and the rest is

statistically significant (under a null hypothesis that true AUCPOs are

equal), it is marked with an asterisk ............................................................ 57

Table 4.1.3. Number of estimated parameters (K), AICc differences

(∆AICc) and Akaike weights (Wi). The maximum Wi for each random

partition of occurrence records is marked in bold. ...................................... 58

RESULTADO 4.2: FEEDING HABITS OF EUROPEAN BADGER (MELES MELES)

IN MEDITERRANEAN ARID LANDSCAPES

Table 4.2.1. Mean relative volume (%) and frequency of occurrence (in

parentheses, %) values for the different wide categories considered in

13

each landscape and season. We reported values of frequency of

occurrence for comparative purposes with other studies using these

categories ................................................................................................... 81

Table 4.2.2. Results of the two-way ANOVA with season and landscape

type as fixed factors and the relative volume of the wide categories as

the response variable. No effect varied its signification by applying the

bootstrap resampling (see Table F.1 Appendix F) ........................................ 83


SPACE: THE EUROPEAN BADGER

Table 4.3.1. MaxEnt models performance for the European badger in SE

Spain under current climate conditions. For each model, the training data

(80% of the total) used were different. The table shows maximised log-

likelihood function (log(L)), number of estimated parameters (K),

corrected Akaike Information Criterion values (AICc), AICc differences

from M9 (∆AICc) and Akaike weights (Wi); *most parsimonious model ..... 110

Table 4.3.2. Most important environmental variables (*) in the MaxEnt

model for habitat suitability of the European badger in SE arid Spain

under current climate conditions (1971-2000). Relative importance of

variables was evaluated by a jackknife test on the training and test gains.

The gains obtained using all variables were 0.227 for training data and

0.397 for test data, so these were the reference values. (see

“Environmental variables” subsection for variables abbreviations) ........... 111

Table 4.3.3. Pearson coefficient correlation (rho) between habitat

suitability predicted in the presence records under current climate

conditions (1971-2000), and each environmental variable. (*P < 0.05; **P

< 0.001). (see “Environmental variables” subsection for variables

abbreviations) ........................................................................................... 112

14

ÍNDICE DE FIGURAS

1. INTRODUCCIÓN

Figura 1.1. Esquena general de la tesis. La tesis persigue dar respuesta a

tres objetivos específicos relacionados con aspectos relevantes de la

ecología del Tejón: distribución espacial, hábitos alimenticios y previsión

de su distribución bajo escenarios de cambio climático. Para abordar

dichos objetivos, se han llevado a cabo diferentes metodologías,

aunando el trabajo de campo y laboratorio, la obtención y

procesamiento de información satelital, y por último, el análisis

estadístico. ................................................................................................. 27

2. ÁREA DE ESTUDIO

Figura 2.1. Localización del área de estudio y principales usos del suelo y

tipos de vegetación. Los límites han sido definidos en base al Índice de

aridez de Martonne (Martonne, 1926). ....................................................... 29

Figura 2.2. Selección de algunos paisajes habitados por el Tejón dentro

del área de estudio. La diversidad de tipos de vegetación, así como de

usos del suelo, definen un territorio donde la riqueza y singularidad de

sus paisajes hacen de él un laboratorio natural para el estudio de la

ecología de esta especie. (a) Paisaje rural de huertas tradicionales

embebidas dentro de un entorno xérico. Algunas terrazas se encuentran

en evidente estado de abandono. (b) Ambiente de marisma con grandes

charcones salinos rodeados de dunas y matorrales de porte elevado. (c)

Estepa árida de relieve llano cercano a la costa, interrumpido solo por

serpenteantes ramblas sujetas a la estacionalidad de las precipitaciones.

(d) Garriga Mediterránea de perfil irregular y de vegetación frondosa,

donde se entremezclan frutales silvestres y asilvestrados. (e) Masas de

encinas relictas y pinares de repoblación dominan desde las alturas uno

de los pocos subdesiertos naturales de Europa. (f) Grandes extensiones

15

de cultivo del Almendro en las altiplanicies del norte del área de estudio,

se entremezclan con pequeños parches de Encinas..................................... 33

3. MATERIAL Y MÉTODOS

Figura 3.1. Esquema gráfico de la metodología empleada en la tesis

doctoral. EVI: Índice de Vegetación Mejorado; MaxEnt: Máxima Entropía;

GLMs: Modelos Lineales Generalizados; GAMs: Modelos Aditivos

Generalizados; NMDS: Escalamiento Multidimensional No Paramétrico;

R: software R .............................................................................................. 35

Figura 3.2. Atributos funcionales derivados del Índice de Vegetación

Mejorado (EVI) y relacionados con el funcionamiento del ecosistema.

Figura modificada de G. Baldi (http://lechusa.unsl.edu.ar) y Alcaraz-

Segura (2005).............................................................................................. 38

4. RESULTADOS

RESULTADO 4.1: MODELING SPATIAL DISTRIBUTION OF EUROPEAN

BADGER IN ARID LANDSCAPES: AN ECOSYSTEM FUNCTIONING

APPROACH

Figure 4.1.1. Study area location. ................................................................ 47

Figure 4.1.2. Jackknife test of variable importance for European badger in

the ALL model with maximum AUCPO. (a) Bars show the AUCPO with each

variable modeled separately. Ratios above the bars show the AUCPO

percentage of the reference value (0.831); (b) Bars show the AUCPO, when

each variable is extracted from the model. The ratios above the bars

show the ratio decreased by the AUCPO with respect to the reference

value (0.831). .............................................................................................. 60

16

RESULTADO 4.2: FEEDING HABITS OF EUROPEAN BADGER (MELES MELES)

IN MEDITERRANEAN ARID LANDSCAPES

Figure 4.2.1. Location of landscapes within the study area in Almería

province, Spain. In each landscape, we identified a zone with latrines

frequently used by European badger (Meles meles). Then, we drew a 3

km-radius buffer zone using the latrines as centroid. .................................. 76

Figure 4.2.2. Estimated relative volume (%) for each main category

considered in the different seasons and landscapes. Whiskers represent

the standard error of the mean values of categories. .................................. 85

Figure 4.2.3. Interaction between landscape and season for earthworm

relative volume (%) in the diet. Whiskers represent the standard error of

the mean values of categories. .................................................................... 86

Figure 4.2.4. Nonparametric multidimensional scaling (NMDS). The axis

NMDS1 and NMDS2, show the range of the distances reached between

seasons in the three landscapes. Seasons are arranged so that the

distances between them are as close to the real differences between the

mean relative volume (%) of fruits, vertebrates and invertebrates

consumed in each landscape. A lower distance between seasons means

greater similarity between them and vice versa. Isoplets are based on the

Shannon´s diversity index. .......................................................................... 88


SPACE: THE EUROPEAN BADGER

Figure 4.3.1. (a) MaxEnt-modeled decrease in habitat suitability of the

European badger in SE arid Spain from current climate conditions (1971-

2000) to two future climate scenarios (IPCC A2 and B1 for 2071-2099).

Habitat suitability maps with mean suitability computed by rows and

columns (cell size 100 x 100 m) in the margins. X and Y axes show UTM

coordinates (Zone 30, Datum ED1950). Cells in grey contain greenhouses,

17

so they were removed before computing the predictions (see

“Environmental variables” subsection). Histograms (Y axis: number of

cells/number total of cells) of the habitat suitability values. (b) Study area

(7051km2) and location of the 179 badger presence records used in this

study. The area only includes arid climate; based on Martonne aridity

index. (c) MaxEnt-modeled maps of significant differences in habitat

suitability between current and predicted climate conditions under the

A2 and B1scenarios for the European badger in SE arid Spain. In pale red,

areas where the variables are expected to significantly decrease (SD >

0.975); in blue, areas where the variables are expected to significantly

increase (SD < 0.025); and in grey, areas where there was no significant

difference. ................................................................................................ 113

Figure 4.3.2. MaxEnt-modeled maps of the limiting factors of habitat

suitability for the European badger in SE arid Spain under current climate

conditions (1971-2000) and two future climate scenarios (IPCC A2 and B1

for 2071-2099). The limiting factor is the environmental variable whose

value at one cell most influences the model suitability prediction. (see


Figure 4.3.3. Maps of the significant differences in the EVI descriptors of

ecosystem functioning and in the climate variables between current

climate conditions (1971-2000) and two future climate scenarios (IPCC

A2 and B1 for 2071-2099) in SE Spain. In pale red, areas where the

variables are expected to significantly decrease (SD > 0.975); in blue,

areas where the variables are expected to significantly increase (SD <

0.025); and in grey, areas where there was no significant difference. (see


INTRODUCCIÓN El Tejón europeo: rasgos de su ecología en un contexto árido Mediterráneo. Justificación. Objetivo general e hipótesis de trabajo

18

1. INTRODUCCIÓN

(Fotografía: letrina de Tejón en el inferior de la imagen dominando los

rebosantes desiertos del sureste europeo. Paraje Natural del Desierto de

Tabernas, Almería).


19

1.1 El Tejón europeo: rasgos de su ecología en un

contexto árido Mediterráneo

El Tejón europeo (Meles meles L., 1758) es un carnívoro de mediano

tamaño perteneciente a la familia Mustelidae. Su cuerpo es robusto y

alargado, con cabeza pequeña y cuello muy corto. Posee fuertes patas

acabadas en largas y poderosas uñas que le sirven para excavar. Su

diseño facial es muy característico y consiste en un fondo blanco surcado

por dos bandas negras que cubren la zona de los ojos (Virgós, 2005). La

especie está presente en casi toda Eurasia, aunque su abundancia y/o

presencia no es homogénea a lo largo de su rango de distribución (Virgós

& Casanovas, 1999). Así, en paisajes humanizados del Reino Unido, se

han registrado densidades de más de 40 tejones/km2 (Macdonald &

Newman, 2002) mientras que en algunas zonas del sur de la Península

Ibérica, las densidades no superan 1 tejón/km2 (Revilla et al., 2001a). Esta

variabilidad en su abundancia pone de manifiesto su gran versatilidad

ecológica, pudiendo sobrevivir en una amplia variedad de paisajes,

aunque no con el mismo éxito reproductivo. En Europa centro-

occidental, países escandinavos y Reino Unido, los tejones viven

generalmente en bosques de hoja caduca con alternancia de pastizales

(Kruuk, 1989; Feroe & Montgomery, 1999). En zonas del suroeste de la

Península Ibérica, el matorral mediterráneo representa su hábitat

preferido (Revilla et al., 2000), mientras que en el sureste, donde

aumentan las condiciones de aridez, los tejones seleccionan paisajes

mosaico constituidos por cultivos extensivos mezclados con parches de

vegetación natural (Lara-Romero et al., 2012). Esta capacidad para

adaptarse y sobrevivir en distintos paisajes viene determinada por la

amplitud de estrategias tróficas que es capaz de adoptar. El Tejón es


20

considerado un especialista en el consumo de lombrices (Lumbricus spp.)

en Gran Bretaña y otras zonas del noroeste de Europa (Kruuk & Parish,

1981). En la región Mediterránea, la disponibilidad de lombrices es

menor debido principalmente a una menor precipitación y a un manejo

del suelo distinto al de otras zonas del norte de Europa (Virgós et al.,

2005a). En estos ambientes, la especie se comporta como un generalista

trófico (Roper, 1994), y consume frutos, insectos y vertebrados en las

zonas más áridas (Piggozi, 1991; Rodríguez & Delibes, 1992; Barea-Azcón

et al., 2010). No obstante, puede mostrar especialización en el consumo

de lombrices en zonas montañosas más húmedas, comportándose por

tanto, como un especialista facultativo bajo circunstancias específicas

(Virgós et al., 2004).

El Tejón europeo ha sido ampliamente estudiado, y sus tendencias

poblacionales y distribución seguidas con interés en el área templada del

continente europeo, especialmente en las Islas Británicas. Gran parte de

este interés se debe a que en dichas zonas la especie representa un

reservorio de Mycobacterium bovis, una micobacteria causante de la

Tuberculosis bovina en el ganado vacuno (Muirhead et al., 1974). Las

investigaciones han mostrado una asociación entre las infecciones de los

rebaños y la presencia de tejones afectados en la misma zona (Muirhead

et al., 1974; Wilesmith, 1983). Por el contrario, en la región

Mediterránea, los estudios sobre la ecología y conservación del Tejón

fueron muy escasos hasta comienzos del siglo XXI, llegando incluso a

catalogarse como “especie insuficientemente conocida” en el Libro Rojo

de los Vertebrados de España (Blanco & González, 1992). En el caso

particular de la Península Ibérica, es a partir del año 2000 cuando

aumenta considerablemente el conocimiento de la especie gracias a los


21

estudios científicos realizados por diferentes equipos de investigación y a

la labor del Grupo de Carnívoros Terrestres de la SECEM (Sociedad

Española para la Conservación y Estudio de los Mamíferos) (Virgós et al.,

2005b). La información acumulada durante los últimos 20 años ha

permitido su catalogación actual como especie en Riesgo menor (LC)

(Palomo et al., 2007).

En el sur de la Península Ibérica, los trabajos sobre la ecología del Tejón

se han centrado en el Parque Nacional de Doñana y algunas zonas

puntuales en el sureste de Andalucía (Rodríguez & Delibes, 1992; Barea-

Azcón et al., 2010; Lara-Romero et al., 2012). Los paisajes áridos

Mediterráneos, particularmente los situados en el sureste ibérico,

suponen un reto para la supervivencia del Tejón. Estos ambientes

representan el límite de su rango de distribución (Del Cerro et al., 2010),

situándose muy por debajo de la idoneidad de hábitat de los paisajes

centroeuropeos (Lara-Romero et al., 2012). Dicha idoneidad puede

incluso verse disminuida en el futuro, dado que la región Mediterránea

es una de las zonas más susceptibles a sufrir los efectos derivados de

algunos impulsores directos de Cambio Global (ej., cambio climático y

cambios en la cobertura vegetal y de uso del suelo) (Sala et al., 2000;

Giorgi & Lionello, 2008). En el sureste de la Península Ibérica se espera un

incremento considerable de las condiciones de aridez debido a un

aumento de la temperatura y disminución de la precipitación (Giorgi &

Lionello, 2008), pudiendo crear condiciones particularmente difíciles

durante el periodo estival (De Luís et al., 2001). Sin embargo, a pesar de

las consecuencias que los cambios en el clima y en los usos del suelo

pueden suponer para la supervivencia del Tejón (Virgós et al., 2005c), no

existe mucho conocimiento sobre la repercusión que estos impulsores


22

pueden tener para la especie en ambientes áridos Mediterráneos. Esto

enfatiza la necesidad de avanzar en la comprensión de aspectos clave de

su ecología como son: (1) qué factores ambientales impulsan su

distribución espacial, (2) variabilidad de los hábitos alimenticios de la

especie en un contexto árido Mediterráneo, y (3) cambios potenciales en

su distribución espacial derivados de la proyección a condiciones

climáticas futuras.

Los modelos de distribución espacial tratan de estimar la idoneidad

relativa de hábitat requerida por una especie en un área geográfica

determinada (Warren & Seifert, 2011; Menke et al., 2009). Estas técnicas

representan una herramienta muy útil en la biología de la conservación,

por lo que su uso se está viendo incrementado considerablemente en los

últimos años (Austin, 2007). Actualmente, los modelos de distribución

desarrollados para el Tejón, integran observaciones de presencia de la

especie junto con variables ambientales de tipo topográfico, climático y

de cobertura y uso del suelo, implementados en Sistemas de Información

Geográfica (SIGs) (Virgós & Casanovas, 1999; Jepsen et al., 2005;

Newton-Cross et al., 2007). Estos modelos han mejorado nuestra

comprensión de su distribución y abundancia (Newton-Cross et al.,

2007), reduciendo muchas de las limitaciones asociadas al muestreo de

campo (ej., alto coste económico, limitación en la extensión del área

geográfica estudiada). Sin embargo, la información derivada de

cartografía SIG también posee limitaciones de representatividad

ecológica, tales como, no representar características del paisaje

relevantes para la especie objeto de estudio, o mostrar una resolución

espacial inadecuada a la escala de trabajo seleccionada (Pearce et al.,

2001). Con el fin de avanzar sobre estos problemas, algunos autores


23

proponen utilizar información satelital para estimar variables que

describan atributos relacionados con el funcionamiento del ecosistema a

través de la producción vegetal (Pettorelli et al., 2005; Cabello et al.,

2012a). Estos atributos describen la variabilidad espacio - temporal de la

producción primaria (Alcaraz-Segura et al., 2013; Requena-Mullor et al.,

2014), por lo que podrían resultar útiles en la modelización de la

distribución espacial del Tejón. Además, gracias a su respuesta rápida

ante los cambios ambientales (Pettorelli et al., 2011), pueden ayudar a

detectar zonas susceptibles de disminuir su calidad de hábitat bajo

condiciones climáticas futuras.

Otro aspecto clave de la ecología del Tejón en paisajes áridos, es su

alimentación. Se ha demostrado que la diversidad de paisajes y de

condiciones ambientales propician distintas estrategias alimenticias en

los tejones, lo que puede condicionar a su vez diferentes organizaciones

sociales, densidades, y afectar a otros aspectos socio-ecológicos (Virgós

et al., 2005a). Por tanto, conocer los hábitos alimenticios de la especie en

paisajes áridos Mediterráneos, es fundamental para comprender de qué

manera pueden variar en respuesta a los principales impulsores de

Cambio Global, y afectar a rasgos importantes de su ecología. Los

estudios sobre la dieta del Tejón en ambientes áridos de la Península

Ibérica son muy escasos. Rodríguez & Delibes (1992) describieron la dieta

únicamente durante la estación de verano en un paisaje con vegetación

xerofítica y cultivos. Barea-Azcón et al. (2010), estudiaron la dieta anual

del Tejón en una zona con clima continental pero en un año

especialmente seco, y un paisaje dominado por Olivos (Olea europea),

pinares (Pinus halepensis) y encinas (Quercus rotundifolia). En este

sentido, no existen estudios que hayan realizado un seguimiento anual


24

de la dieta en ambientes con aridez constante en el tiempo y analizando

además la influencia de diferentes coberturas vegetales y usos del suelo.

Por último, profundizar en cómo la distribución del Tejón puede variar

ante futuros escenarios climáticos resulta fundamental para entender los

procesos relacionados con la pérdida de su calidad de hábitat. De forma

general, Levinsky et al. (2007) prevén una reducción drástica de la

riqueza potencial de mamíferos en la región Mediterránea, y

particularmente, Maiorano et al. (2014) predicen una disminución de

hasta un 50% en la distribución de las especies de la familia Mustelidae

en la cuenca Mediterránea. Dado que el Tejón es escaso o está ausente

en ambientes áridos Mediterráneos (Virgós et al., 2005c), sería de

esperar que su distribución se viera modificada en respuesta a cambios

ambientales. Estas previsiones ofrecen una excelente oportunidad para

mejorar nuestro conocimiento acerca de los impactos potenciales que el

cambio climático pudiera tener sobre la especie y analizar posibles

respuestas frente a ellos. A su vez, arrojaría información útil para el

desarrollo de planes de conservación, tanto para el Tejón, como para

otros meso carnívoros Mediterráneos.

1.2 Justificación

El Tejón europeo posee una amplia distribución en la Península Ibérica

(Revilla et al., 2002). Sin embargo, a una menor escala, la especie

presenta claras tendencias en sus preferencias de hábitat (Virgós &

Casanovas, 1999; Revilla et al., 2000), y puede llegar a ser localmente

raro o incluso estar ausente, como ocurre en algunas zonas del sureste

árido (Virgós, 1994; Virgós et al., 2005a). El sureste de la Península

Ibérica representa uno de los límites de su distribución, y por tanto, las


25

condiciones ambientales se encuentran lejos de su óptimo

Centroeuropeo (Virgós & Casanovas, 1999). Por ello, sería razonable

pensar que la supervivencia de la especie en esta región pueda verse

amenazada en un contexto de Cambio Global, por lo que mejorar nuestra

comprensión sobre la ecología de la especie en estos ambientes resulta

determinante. Aunque existe un amplio conocimiento en otras zonas de

su rango de distribución (ej., Europa central e Islas Británicas), en la

región Mediterránea es aún insuficiente, particularmente en ambientes

áridos. A pesar de que en los últimos años se ha avanzado en esta

dirección (Barea-Azcón et al., 2010; Lara-Romero et al., 2012), aún

quedan aspectos importantes por comprender como cuáles son los

factores ambientales que impulsan su distribución espacial, la

variabilidad de sus hábitos alimenticios dentro de un contexto árido o los

patrones futuros de dicha distribución frente a escenarios de cambio

climático.

1.3 Objetivo general e hipótesis de trabajo

El objetivo general de la tesis es avanzar en el conocimiento sobre la

ecología del Tejón europeo en paisajes áridos Mediterráneos (Fig. 1.1).

Para lograr dicho objetivo se proponen tres objetivos específicos:

1. Desarrollar modelos de distribución espacial incorporando

información satelital relacionada con el funcionamiento del

ecosistema, para testar si mejora la fiabilidad de los mismos.

2. Describir y comparar la dieta del Tejón a través de paisajes

áridos con diferente cobertura vegetal y uso del suelo.


26

3. Proyectar la distribución espacial del Tejón utilizando

escenarios de cambio climático futuros propuestos por el

Intergovernmental Panel on Climate Change (IPCC).

Para ello, se plantean las siguientes hipótesis de trabajo:

1) Dado que los recursos tróficos del tejón están relacionados

directa y/o indirectamente con la producción primaria del

ecosistema, cabría esperar que información derivada de sensores

remotos, relacionada a través de la producción vegetal con el

funcionamiento del ecosistema, resultara útil en la modelización

de la distribución espacial del Tejón.

2) Debido al carácter generalista descrito para el Tejón en

ambientes Mediterráneos (Roper, 1994), el comportamiento

trófico de la especie podría variar entre paisajes áridos con

diferente cobertura vegetal y uso del suelo, y explotar distintos

recursos alimenticios.

3) Los cambios en los patrones de precipitación y temperatura

previstos por los modelos de circulación general de la atmósfera

(IPCC, 2013) podrían reducir hasta en un 50% el rango de

distribución de las especies de la familia Mustelidae en la cuenca

Mediterránea (Maiorano et al., 2014). Por tanto, cabe esperar

una reducción general de la calidad del hábitat para el Tejón en

el sureste árido de la Península Ibérica.


27

Figura 1.1. Esquena general de la tesis. La tesis persigue dar respuesta a tres objetivos específicos relacionados con aspectos relevantes de la ecología

del Tejón: distribución espacial, hábitos alimenticios y previsión de su distribución bajo escenarios de cambio climático. Para abordar dichos objetivos,

se han llevado a cabo diferentes metodologías, aunando el trabajo de campo y laboratorio, la obtención y procesamiento de información satelital, y

por último, el análisis estadístico.

ÁREA DE ESTUDIO

28

2. ÁREA DE ESTUDIO

La tesis ha sido llevada a cabo en el sureste de la Península Ibérica

(3606’N, 217’E) (Fig. 2.1). Esta región ocupa una posición biogeográfica

singular en el contexto del Mediterráneo occidental. Situada en una zona

de "sombra de lluvias" al abrigo del macizo montañoso de Sierra Nevada,

se encuentra protegida de las borrascas atlánticas que entran por el

oeste, y expuesta a las particularidades climáticas del Mar de Alborán,

con perturbaciones estacionales que dejan intensas lluvias de carácter

torrencial. Así, el rango de precipitación media oscila entre 165-419

mm/año. Dicha personalidad pluviométrica, unida a la rigurosidad

térmica de los veranos mediterráneos y a unas temperaturas suaves el

resto del año (temperatura media de las mínimas: -1.6–15 °C,

temperatura media de las máximas 17-24.5 °C), dibujan uno de los

entornos de aridez más intensos de Europa, con una evapotranspiración

potencial media de 343-1038 mm/año. Junto a todo esto, la región de

estudio ofrece gradientes altitudinales que van desde el nivel del mar

hasta los 1500 m, así como una variedad de litologías que derivan en una

variedad paisajística notable y por tanto, en una oferta de nichos

ecológicos muy diversa.

La región acoge tres pisos termoclimáticos: termo, meso y

supramediterráneo dentro de los sectores biogeográficos almeriense,

alpujarreño-gadorense, nevadense y guadiciano-bacense. Su litología

está caracterizada por la presencia de materiales metamórficos

paleozoicos y paleozoico-triásicos, representados mayoritariamente por

micaesquistos, además de materiales sedimentarios con calizas, margas,

yesos, y arenas terciarias, junto a conglomerados y arcillas cuaternarias.

ÁREA DE ESTUDIO

29 Figura 2.1. Localización del área de estudio y principales usos del suelo y tipos de vegetación. Los límites han sido definidos en base al

Índice de aridez de Martonne (Martonne, 1926).

ÁREA DE ESTUDIO

30

La vegetación más extendida se corresponde con las series termo-

mediterránea semiárida del Arto (Maytenus senegalensis subsp.

europaeus) y termo- y meso-termo-mediterránea del Lentisco,

representadas por Lentiscares con Pistacia lentiscus, Chamaerops humilis

y Rhamnus spp., junto con extensas zonas de matorral de Albaida

(Anthyllis spp.), Esparto (Macrochloa tenacissima) y Tomillos (Thymus

spp.). Puntualmente importantes son los complejos politeselares de

vegetación edafoxerófila sobre yesos con matorrales de pequeño porte.

La vegetación arbórea se asienta en las zonas más elevadas, con

encinares basófilos y silíceos junto a extensas plantaciones de pino (P.

halepensis, P. nigra y P. silvestris). Cabe destacar, por la relevancia

ecológica para el Tejón (Corbacho et al., 2003), las ramblas y cauces

ocupados por geoseries edafohigrófilas donde destacan: Aneas (Typha

spp.), Carrizo (Phragmites autralis), Tarays (Tamarix spp.), Adelfa (Nerium

oleander) y frutales como la Higuera (Ficus carica) y el Algarrobo

(Ceratonia siliqua).

Pero sin duda, uno de los rasgos del área de estudio más destacados y de

mayor trascendencia sobre la ecología del Tejón europeo, son los usos

agrícolas del suelo derivados de la actividad humana (Lara-Romero et al.,

2012). El paisaje rural Mediterráneo ha sido definido como un mosaico

cambiante constituido por cultivos extensivos mezclados con parches de

vegetación natural, favorecedor de la diversidad y abundancia de

carnívoros (Pita et al., 2009). Dichos paisajes mantienen una alta

heterogeneidad paisajística con gran variedad de cultivos, pero

manteniendo a su vez ribazos y linderos. Aunque en retroceso por

abandono, aún se conservan "vegas" asociadas a los cursos de agua,

especialmente en los tramos medios, donde el cultivo de olivos y cítricos

ÁREA DE ESTUDIO

31

son los más destacados. En contraposición, existen extensas áreas de

cultivo intensivo en regadío de olivos, así como de especies herbáceas

(ej., lechuga) y en secano (almendro y cereal), éstos últimos

especialmente importantes hacia el norte. De igual forma, por debajo de

la mitad sur y cercanos al litoral, aparecen tres importantes núcleos de

cultivo hortícola intensivo bajo plástico. El grado de antropización

urbanística, y por tanto, de ocupación humana, es máximo hacia la costa,

disminuyendo en el interior. El éxodo de la población rural desde las

zonas del interior hacia el litoral, especialmente en el sur, representa el

motivo principal de la disminución y deterioro del paisaje rural

tradicional sufrido en las últimas décadas en la región de estudio (Castro

et al., 2011).

Tanto el desarrollo urbanístico como la actividad agrícola intensiva,

ejercen importantes presiones, y por tanto amenazas, sobre la

conservación de los hábitats ocupados por el Tejón, plasmadas por

ejemplo, en la pérdida y fragmentación del hábitat (Virgós et al., 2005c).

En relación a este aspecto, el área de estudio y zonas aledañas, cuentan

con diferentes figuras de protección fruto de las políticas de

conservación del territorio en las últimas décadas. Así, las zonas

montañosas gozan en general de un buen estatus de protección, ej.,

Parque Nacional y Natural de Sierra Nevada, Parque Natural de Sierra

María-Los Vélez y Paraje Natural de Sierra Alhamilla. Los ambientes de

humedal poseen un moderado nivel de protección hacia el sur (ej., Paraje

Natural de Punta Entinas-Sabinar) y casi inexistente en zonas del noreste.

El paisaje estepario, muy extendido a lo largo del área, presenta

diferentes grados de protección. Así, el Parque Natural de Cabo de Gata-

Níjar, Paraje Natural del Desierto de Tabernas y Karst en Yesos de Sorbas,

ÁREA DE ESTUDIO

32

y varios LICs (Lugares de Interés Comunitario) repartidos por el centro y

sur de la región, conforman un gradiente decreciente de conservación de

dicho paisaje. Los ecosistemas esteparios han sido poco valorados

tradicionalmente por el ser humano, sin embargo, son también

explotados por el Tejón aunque en menor proporción. Por último, cabe

destacar la ausencia en la mayoría de los casos, de figuras de protección

que recaigan directamente sobre agroecosistemas tales como las vegas

fluviales tradicionales comentadas antes, o los paisajes cerealistas del

altiplano en la zona norte aledaña al área de estudio. Ambos representan

paisajes humanizados pero de gran valor ecológico, muy importantes

para el Tejón a escala regional (Virgós et al., 2002; Lara-Romero et al.,

2012).

A modo de ejemplo representativo de la diversidad de paisajes presentes

en el área de estudio, la Fig. 2.2 muestra una selección de localidades en

las cuales se ha detectado la presencia de la especie a lo largo de la

realización de la tesis doctoral.

(a) Paisaje rural de huertas tradicionales. (b) Ambiente de marisma.

ÁREA DE ESTUDIO

33

(c) Estepa árida de relieve llano cercano a la costa.

(d) Garriga Mediterránea de perfil irregular.

(e) Masas de encinas relictas y pinares de repoblación.

(f) Grandes extensiones de cultivo del Almendro entremezclados con pequeños

parches de Encinas.

Figura 2.2. Selección de algunos paisajes habitados por el Tejón dentro del área

de estudio. La diversidad de tipos de vegetación, así como de usos del suelo,

definen un territorio donde la riqueza y singularidad de sus paisajes hacen de él

un laboratorio natural para el estudio de la ecología de esta especie. (a) Paisaje

rural de huertas tradicionales embebidas dentro de un entorno xérico. Algunas

terrazas se encuentran en evidente estado de abandono. (b) Ambiente de

marisma con grandes charcones salinos rodeados de dunas y matorrales de

porte elevado. (c) Estepa árida de relieve llano cercano a la costa, interrumpido

solo por serpenteantes ramblas sujetas a la estacionalidad de las

precipitaciones. (d) Garriga Mediterránea de perfil irregular y de vegetación

frondosa, donde se entremezclan frutales silvestres y asilvestrados. (e) Masas de

encinas relictas y pinares de repoblación dominan desde las alturas uno de los

pocos subdesiertos naturales de Europa. (f) Grandes extensiones de cultivo del

Almendro en las altiplanicies del norte del área de estudio, se entremezclan con

pequeños parches de Encinas.

ÁREA DE ESTUDIO

34

De forma particular, y con el fin de testar cada una de las hipótesis

planteadas en la tesis doctoral, se han definido con posterioridad

distintas zonas dentro del área de estudio. Su definición se ha realizado

atendiendo a los requerimientos derivados del planteamiento conceptual

y metodológico de cada hipótesis, y son expuestos detalladamente en el

apartado de resultados.

MATERIAL Y MÉTODOS Trabajo de campo y laboratorio, información satelital - funcionamiento ecosistémico y análisis estadístico

35

3. MATERIAL Y MÉTODOS

La metodología desarrollada en esta memoria de tesis doctoral integra

tres bloques de trabajo principales (Fig. 3.1). BLOQUE I: técnicas de

muestreo en campo y trabajo de laboratorio; BLOQUE II: descarga y

procesamiento de información satelital para la estima de atributos

funcionales del ecosistema empleados como subrogados de la dinámica

espacio - temporal de la producción primaria; y por último, BLOQUE III:

análisis estadístico paramétrico y no paramétrico empleando el software

libre R (R Core Team, 2014) para, a partir de la información obtenida en

los dos bloques anteriores, testar las hipótesis planteadas.

Figura 3.1. Esquema gráfico de la metodología empleada en la tesis doctoral.

EVI: Índice de Vegetación Mejorado; MaxEnt: Máxima Entropía; GLMs: Modelos

Lineales Generalizados; GAMs: Modelos Aditivos Generalizados; NMDS:

Escalamiento Multidimensional No Paramétrico; R: software R.


36

3.1 BLOQUE I: Trabajo de campo y laboratorio

El trabajo de campo ha consistido en la búsqueda activa de indicios de

presencia de Tejón y la recolección de excrementos. Existen numerosas

técnicas de campo aplicables al muestreo de la presencia de Tejón (ver

Virgós & Revilla, 2005 para una revisión y resumen). Sin embargo, no

todas son válidas en zonas con baja densidad de la especie como ocurre

en el sureste árido de la Península Ibérica (Lara-Romero et al., 2012). En

este contexto, un método fiable, rápido y barato para cubrir grandes

extensiones, es la búsqueda activa de indicios de presencia (huellas

principalmente) en cuadrículas de igual área y durante un tiempo

determinado (Revilla et al., 2001b). No obstante, dado que la impresión

de huellas es muy dependiente del tipo y estado del sustrato, es

recomendable no limitar la búsqueda únicamente a huellas y ampliarla

también a letrinas y tejoneras (Virgós & Revilla, 2005). La información

levantada en campo, ofrece no solo una serie de localizaciones con

presencia de la especie (coordenadas UTM) imprescindibles para la

modelización de su distribución espacial, sino que posibilita además el

conocimiento de zonas de marcaje con letrinas para la recolecta de

excrementos y su posterior análisis en laboratorio.

El trabajo de laboratorio se ha basado en el análisis visual de

excrementos. Existen diversos protocolos estandarizados para el

reconocimiento y conteo de restos de alimento en heces animales (Kruuk

& Parish, 1981; Pigozzi, 1991). La disgregación de las muestras se realiza

en medio acuoso, para posteriormente, tamizarlas y separar los

componentes. Para el reconocimiento y determinación de los restos de

presas consumidas, se utiliza la lupa binocular, en caso de ítems

macroscópicos (ej., semillas, huesos, restos quitinosos de invertebrados,


37

etc.), o el microscopio óptico para la detección de restos más diminutos

como las quetas de las lombrices de tierra.

3.2 BLOQUE II: Información satelital y funcionamiento

ecosistémico

La información espectral emitida por la vegetación y obtenida a partir de

sensores remotos representa un nuevo paradigma para el estudio y

seguimiento de la fauna y su conservación (Pettorelli et al., 2011; Cabello

et al., 2012a). La respuesta espectral de la cobertura vegetal en

longitudes de onda en el rango del rojo e infrarrojo permite estimar

diferentes atributos funcionales sobre grandes extensiones del territorio

(Running et al., 2000; Paruelo et al., 2005), posibilitando a su vez el

estudio del funcionamiento de la vegetación a escala de ecosistema

(Lloyd, 1990). Un ejemplo de ello lo constituye el Índice de Vegetación

Mejorado (EVI), el cual ha sido ampliamente usado como un subrogado

de la Producción Primaria (PP) y su dinámica estacional (Pettorelli et al.,

2005; Alcaraz-Segura et al., 2013) (Fig. 3.2). La ecuación utilizada para su

obtención se indica a continuación:

(1)

donde G es un factor de ganancia; RIRC, RR y RA son respectivamente los

valores de reflectancia bidireccional de la superficie de la tierra para las

bandas del infrarrojo cercano, del rojo y del azul con una corrección de

los efectos de la atmósfera (Absorción de ozono y Rayleigh); C1 y C2 son

los coeficientes de resistencia de aerosoles, que usan la banda azul para

corregir la influencia del aerosol en la banda roja y L es un ajuste del

fondo del dosel que toma en cuenta la transferencia radiante diferencial

del infrarrojo cercano y el rojo a través del dosel. Los coeficientes


38

adoptados en el algoritmo del cálculo del EVI son L = 1, C1 = 6, C2 = 7.5 y G

= 2.5. (Fuente: http://modis.gsfc.nasa.gov/data/atbd/atbd_mod13.pdf).

Este índice es un estimador lineal de la fracción de la radiación

fotosintéticamente activa absorbida por la vegetación (fAPAR) (Ruimy et

al., 1994; Huete et al., 2002), el principal control de las ganancias de

carbono (Monteith, 1981), y está siendo cada vez más utilizado en

ecología animal para describir atributos del funcionamiento de los

ecosistemas (Wang et al., 2010; Meynard et al., 2012; Bardsen & Tveraa,

2012; Requena-Mullor et al., 2014).

Figura 3.2. Atributos funcionales derivados del Índice de Vegetación Mejorado

(EVI) y relacionados con el funcionamiento del ecosistema. Figura modificada de

G. Baldi (http://lechusa.unsl.edu.ar) y Alcaraz-Segura (2005).

3.3 BLOQUE III: Análisis estadístico

Con el fin de testar las hipótesis planteadas, la información obtenida en

los bloques I y II ha sido analizada con diversas técnicas estadísticas

utilizando el software libre R.


39

3.3.1 Técnicas paramétricas

La modelización de la distribución espacial del Tejón, se ha basado en el

principio de máxima entropía implementado en el software libre MaxEnt

por Phillips et al. (2006). MaxEnt utiliza datos de presencia de la especie

junto a un grupo de pseudo-ausencias escogidas al azar del área de

estudio donde la ausencia y presencia es posible. El objetivo es encontrar

la función de probabilidad que posea la máxima entropía, esto es, la más

cercana a la uniformidad. MaxEnt tiene en cuenta una serie de

restricciones expresadas como funciones simples de las variables

ambientales utilizadas, de tal forma que, el algoritmo aplicado por

MaxEnt obliga a que el promedio de valores obtenidos a partir de las

funciones para cada variable esté próximo a la media empírica conocida

en las localidades con presencia de la especie (Phillips et al., 2004).

Finalmente, el algoritmo otorga un valor de probabilidad de presencia

(asumiendo por defecto que la prevalencia es 0.5) para cada una de las

unidades espaciales en las que haya sido dividida el área de estudio.

Para testar potenciales diferencias entre las estrategias tróficas

desarrolladas por el Tejón en los paisajes estudiados, se realizó un

análisis de varianzas mediante Modelos Lineales Generalizados (GLMs).

Los GLMs son una extensión de los modelos lineales que permiten utilizar

distribuciones de probabilidad no normales para la variable respuesta.

Un GLM consiste en tres componentes:

1.- La distribución de probabilidad de la variable respuesta. Si la variable

respuesta es continua, puede asumirse que su distribución será normal,

con media μ y varianza σ2.


40

2.- Componente sistemático. Es una combinación lineal de variables

predictoras continuas y/o categóricas. Cuando estas variables son

únicamente categóricas, el GLM es equivalente al análisis de varianzas

empleado en estadística aplicada (Sokal & Rohlf, 1995; Underwood,

1997).

3.- Función de enlace. Define la relación entre la media de la variable

respuesta y el componente sistemático. Cuando se asume que la relación

entre ambos es lineal, se emplea la función identidad, donde g(μ) = μ.

En los GLMs la estimación de los parámetros se realiza mediante el

método de máxima verosimilitud. El principio en el cual se basa la

estimación por máxima verosimilitud es simple: dada una muestra de

observaciones, el valor estimado para un parámetro es aquél que

maximiza la probabilidad de dichas observaciones.

Los GLMs son considerados modelos paramétricos porque debe

especificarse una distribución de probabilidad para la variable respuesta,

y por tanto, para el término de error del modelo. Así mismo, se

consideran lineales porque la variable respuesta es descrita como una

combinación lineal de variables predictoras. Más información sobre estos

modelos puede encontrase en Agresti (1996) y Myers & Montgomery,

(1997).

3.3.2 Técnicas no paramétricas

Los Modelos Aditivos Generalizados (GAMs) son modificaciones no

paramétricas de los GLMs donde cada variable predictora es incluida en

el modelo como una función no paramétrica de "suavizado" (del inglés

smoothing) (Hastie & Tibshirani, 1990). Estas funciones suelen ser LOESS

smoothing (local regression smoother) o cubic splines (para una


41

explicación en detalle ver respectivamente Keele, 2008; Wood, 2006;

entre otros). De forma breve, LOESS smoothing, consiste en un ajuste de

modelos de regresión lineal de forma local a través de pequeñas

"ventanas" de tamaño ajustable a lo largo del eje de abscisas, donde los

puntos dentro de cada ventana son utilizados en un modelo de regresión

para predecir el valor de la variable respuesta correspondiente al valor

medio o a la mediana de la variable predictora dentro de la

correspondiente ventana. En cubic splines, el eje de abscisas es dividido

en varios segmentos, y en cada uno de ellos, una función polinómica de

grado 3 es ajustada.

Los GAMs permiten relaciones no lineales entre la variable respuesta y

las variables predictoras, por lo que resultan muy útiles en situaciones

donde las relaciones lineales no son esperables. Este puede ser el caso de

la relación entre el EVI medio anual y la precipitación media anual (ver

Figura J.1 en Appendix J).

Por último, la búsqueda de similaridad-disimilaridad entre paisajes y

estaciones del año en relación a los alimentos consumidos por el Tejón

requiere de técnicas multivariantes que no asuman normalidad entre las

variables predictoras. El Escalamiento Multidimensional No Paramétrico

(NMDS) es una técnica multivariante de interdependencia que trata de

representar en un espacio geométrico de pocas dimensiones

(normalmente dos) las similaridades existentes entre un conjunto de

objetos (ej., las estaciones del año). De esta manera, una menor distancia

entre los objetos indica mayor semejanza entre las variables predictoras

utilizadas (ej., alimentos consumidos).

RESULTADO 4.1 Modeling spatial distribution of European badger in arid landscapes: an ecosystem functioning approach

42

4. RESULTADOS

RESULTADO 4.1: MODELING SPATIAL DISTRIBUTION

OF EUROPEAN BADGER IN ARID LANDSCAPES: AN

ECOSYSTEM FUNCTIONING APPROACH

Tejón inspeccionando madrigueras de conejo en busca de presas. Escena

capturada con fototrampeo en la Rambla de las Amoladeras. Mayo 2011. Parque

Natural de Cabo de Gata-Níjar.

Basado en:

Juan M. Requena-Mullor, Enrique López, Antonio J. Castro, Javier Cabello,

Emilio Virgós, Emilio González-Miras, Hermelindo Castro. (2014).

Landscape Ecology, 29:843-855.


43

Objetivo 1: “Desarrollar modelos de distribución espacial incorporando

información satelital relacionada con el funcionamiento del ecosistema,

para testar si mejora la fiabilidad de los mismos.”

Hipótesis 1: “Dado que los recursos tróficos del tejón están relacionados

directa y/o indirectamente con la producción primaria, cabría esperar

que la información satelital relacionada con el funcionamiento del

ecosistema a través de la producción vegetal, resultara útil en la

modelización de la distribución espacial del Tejón.”

Tengo mis resultados hace tiempo, pero no sé cómo llegar a ellos.

C. F. Gauss


44

Abstract

Understanding the factors determining the spatial distribution of species

is a major challenge in ecology and conservation. This study tests the use

of ecosystem functioning variables, derived from satellite imagery data,

to explore their potential use in modeling the distribution of the

European badger in Mediterranean arid environments. We found that

the performance of distribution models was enhanced by the inclusion of

variables derived from the Enhanced Vegetation Index (EVI), such as

mean EVI (a proxy for primary production), the coefficient of variation of

mean EVI (an indicator of seasonality), and the standard deviation of

mean EVI (representing spatial heterogeneity of primary production). We

also found that distributions predicted by remote sensing data were

consistent with the ecological preferences of badger in those

environments, which may be explained by the link between EVI-derived

variables and the spatial and temporal variability of food resource

availability. In conclusion, we suggest the incorporation of variables

associated with ecosystem function into species modeling exercises as a

useful tool for improving decision-making related to wildlife conservation

and management.

Keywords: Ecological niche modeling, MaxEnt, remote sensing, EVI, land

use-land cover, Mediterranean ecosystems, Spain, Meles meles


45

4.1.1 Introduction

Understanding the factors determining the spatial distribution of species

is a major challenge in ecology and conservation biology (Brown et al.,

1995). The European badger (Meles meles) is medium-sized carnivore

widely distributed across Europe. In Mediterranean arid landscapes the

species is not abundant or is absent due to extreme aridity (Virgós et al.,

2005). Current spatial distribution models for the European badger use

occurrence data in conjunction with environmental variables derived

from GIS data sources, such as topographic, climatic, and land cover/use

(Virgós & Casanovas, 1999a; Jepsen et al., 2005; Newton-Cross et al.,

2007). These models have improved our understanding of badger

distribution and abundance (Newton-Cross et al., 2007) by reducing

limitations associated with field sampling (e.g., high economic cost and

limited geographic range). However, data derived by GIS cartography

could include limitations of ecological representativeness such as not

representing relevant landscape features for the target species or

inadequate spatial resolution (Pearce et al., 2001).

The use of ecosystem functioning variables could improve spatial

distribution modeling due to their capacity to reflect spatial variability of

landscape features and faster response to environment changes

(Pettorelli et al., 2011). Ecosystem functioning variables can be extracted

from remote sensing imagery, available continuously, both spatially and

temporally. This allows the employment of standardized spectral indexes

for monitoring species on different spatiotemporal scales (Nilsen et al.,

2005) reducing extrapolations. An example of potentially useful

ecosystem functioning variables are the functional attributes derived by

the Enhanced Vegetation Index (EVI). The EVI has been used in mammal


46

ecology by Wang et al. (2010), Meynard et al. (2012), and Bardsen &

Tveraa (2012). The EVI is linearly related to ecosystem carbon gains, and

therefore, to net primary productivity (NPP) (Monteith, 1981), which is

used as a surrogate of ecosystem functioning (Alcaraz et al., 2006;

Cabello et al., 2012a). Thus, measures derived from EVI can describe

ecosystem functional attributes (Pettorelli et al., 2005). These attributes

include the mean annual EVI (i.e., surrogate of primary production)

(Huete et al., 1997; Sims et al., 2006) and the coefficient of variation of

mean annual EVI (i.e., indicator of seasonality) (Alcaraz-Segura et al.,

2013).

The Resource dispersion hypothesis posits that the size of badger

territories is mainly linked to the dispersion of food resources

(Macdonald, 1983; Kruuk, 1989; Macdonald & Carr, 1999). This

hypothesis emphasizes the key role of patchiness of food quality in

determining how large badger territories are. For example, habitat

production tends to drive body condition, ultimately influencing fitness

(Woodroffe, 1995). As a consequence, reproductive success of females is

largely dependent on food conditions, which in badgers are mainly linked

to climate factors mediating food abundance (e.g., production of

habitats) (Woodroffe & Macdonald, 1995). Therefore, badger

demography, abundance and social life is mainly shaped by food

availability and predictability (seasonality), which can be assessed by

ecosystem functional attributes derived of spectral vegetation indices

(e.g., Nilsen et al., 2005; Pettorelli et al., 2005; Pettorelli et al., 2006).

The purpose of this study is to test the use of ecosystem functional

variables derived from EVI (e.g., mean annual EVI, coefficient of variation

of mean annual EVI, and spatial deviation of mean annual EVI) to


47

improve spatial distribution modeling of the European badger. With this

aim, we first sampled badger occurrence in a representative arid

landscape located in the southeastern Iberian Peninsula (Fig. 4.1.1).

Secondly, we designed a variety of spatial distribution models based on

environmental variables, with and without including EVI-derived

variables. We also explored their performance based on a subset of

previously sampled presence data and the habitat preferences of badger

as described by other authors. Finally, we discuss the role of ecosystem

functional dimension in species ecological modeling and conservation.

Figure 4.1.1. Study area location.


48

4.1.2 Material and methods

4.1.2.1 Study area

We selected a representative area of arid landscapes in the southeastern

Iberian Peninsula based on the Martonne aridity index (Martonne, 1926)

(Fig. 4.1.1), as it is easily calculated and mapped with GIS layers. Inside

this area defined as Arid steppic by the Martonne index (range: 5-15), we

drew a 3.5 km-radius buffer zone on both sides of the two major rivers

basins in the region, and then joined the two buffers (Fig. 4.1.1). In this

form, we ensure inclusion of the potential home range estimated for

European badgers in these environments (i.e., 9 km2) (Lara-Romero et al.,

2012). The study area comprised 835 km2, with a temperature gradient

(range of minimum mean temperatures: 7-12 °C, range of maximum

mean temperatures: 23-28 °C) and an annual precipitation gradient (200-

600 mm/year) associated with a wide altitudinal gradient (0-1400 m).

Evapotranspiration ranges from 93 to 945 mm/year. Another important

feature of the area is the diversity of land cover/use: xerophytic scrubs

represent 48% of the area, where Stipa tenacissima is the most

abundant. Forested habitat is very scarce, corresponding mostly to

scattered pine forests (Pinus halepensis). Crops occupy 27% of the study

area and include fruit orchards (especially abundant near the rivers),

arable crops and greenhouses, in similar proportions.

4.1.2.2 Field survey data

A field survey was conducted from September 2010 to February 2011.

The study area was divided into 5 x 5 km UTM (Universal Transverse

Mercator) plots (out of total 66 plots) to organize the field surveys and

not as the sampling unit. A survey to identify signs of badgers (i.e.,


49

footprints, latrines and setts) was carried out for 6 hours in each plot. To

maximize the detection of the species with the least effort, we selected

places for survey such as paths and catchments for footprints, hills for

latrines and easy to dig sloping areas for setts. These places are known to

be usually used by badgers. The GPS (UTM) coordinates of each sign

were noted with a measurement error of up to 10 m using a GPSmap®

60CSx-Garmin. To avoid spatial autocorrelation of environmental

variables (see below), no signs within 100 m from each other were

considered (see Appendix A).

4.1.2.3 Environmental data

The study area was characterized based on twenty predictor variables

(Table 4.1.1 and Appendix B). Nine of these variables are commonly used

in European badger ecology studies (e.g., Virgós and Casanovas 1999a;

Revilla et al 2000; Jepsen et al., 2005; Macdonald & Newman, 2002;

Rosalino et al., 2008; Lara-Romero et al., 2012), and were comprised by

two climate variables, one topographic variable and six variables related

to habitat structure represented by different land cover/use. The eleven

remaining variables were derived from remote sensing data.

Final resolution of environmental data sets was adjusted to 100 x 100 m

pixel size (i.e., sample unit), to agree with the predominant smallest

spatial resolution of data (Ferrier & Watson, 1997; Elith & Leathwick,

2009). Some variables (i.e., land cover/use variables) were scaled to the

relevant scale for badgers (i.e., their home range).

Topographic and climate variables

Topographic and climate variables were derived from spatial data layers

of the Environmental Information Network of Andalusia


50

(http://www.juntadeandalucia.es/medioambiente/site/web/rediam).

ESRI ArcMapTM 9.3 was used for their handling and processing. The

topographic variable was mean slope, which has been described as a

factor relevant for sett digging (Jepsen et al., 2005). It was estimated

from the digital elevation model of Andalusia with a spatial and elevation

pixel resolution of 20 x 20 m. This layer was resampled to 100 x 100 m.

Climate variables (Virgós & Casanovas, 1999a; Johnson et al., 2002;

Macdonald & Newman, 2002) were mean annual rainfall and the mean

maximum temperature, acquired with a resolution of 100 x 100 m, so no

transformation was made.

Land cover and land use variables

The land cover and land use variables (Virgós & Casanovas, 1999a; Revilla

et al., 2000; Rosalino et al., 2008; Lara-Romero et al., 2012) were derived

from Andalusian Land use/Land cover map (scale 1:25000 from 2007), in

vector format. This layer included the following classes: scattered scrub,

dense scrub, woody crops, arable crops, mixed crops (woody and arable)

and mosaic crops (crops and natural vegetation). The study area was first

divided into 3 x 3 km plots. We estimated the area (km2) of each class

and then we rasterized to 100 x 100 m pixel size. These variables were

scaled because the percent cover is relevant for badgers (Lara-Romero et

al., 2012), instead of using the class of land cover/use as categorical

variable. We considered a 9 km2 area as the probable home range of the

European badger in areas of low habitat suitability (Lara-Romero et al.,

2012).

http://www.juntadeandalucia.es/medioambiente/site/web/rediam


51

EVI variables

The eleven variables derived from remote sensing data were estimated

based on the MOD13Q1 EVI product, generated from images captured by

the MODIS sensor aboard the NASA’s TERRA satellite

(www.modis.gsfc.nasa.gov) for a period of seven years. These images

have the advantages of its high temporal resolution of 16 days (23

images/year) and spatial resolution appropriate to the scale of the study

(231 x 231 m). The images were subjected to pixel quality filtering, in

which those affected by heavy content of aerosols, clouds, shadows,

snow or water were eliminated. The EVI is the index least affected by

atmospheric conditions and presents fewer saturation problems for high

levels of biomass (Huete et al., 2002).

The mean annual EVI is linearly related to total carbon gain (Running et

al., 2000), and has been used as a surrogate of vegetation production

(Alcaraz-Segura et al., 2013). The standard deviation of mean annual EVI

is an indirect measure of spatial heterogeneity, so that a high standard

deviation may indicate mixed patches, while a low standard deviation is

common in homogeneous landscapes. This variable was estimated by

calculating the standard deviation of mean annual EVI in the 3 x 3 km

plots used to estimate the land cover/use variables. The coefficient of

variation of mean annual EVI is a seasonal carbon gain descriptor (Alcaraz

et al., 2006) that has been used as an indicator of ecosystem seasonality

(Alcaraz-Segura et al., 2013). Furthermore, seasonality, although

described by other variables, has proven decisive in modeling the habitat

of several other species (Boyce, 1978; Ferguson & McLoughlin, 2000;

Wiegand et al., 2008). In addition to these, the EVI autumn mean

(September-November) and EVI spring mean (March-May) were also


52

included as variables, because they represent the two growing seasons in

Mediterranean arid landscapes (Cabello et al., 2012b).

EVI variables were resampled to 100 x 100 m by a bilinear resampling

technique. It determines the new value of a cell based on a weighted

distance average of the four nearest input cell centers. This is likely more

realistic than using nearest-neighbor interpolation method (Phillips et al.,

2006).

EVI of land cover and land uses variables

Five variables were created by calculating mean EVI for each class of land

cover/use referred to above. These variables were also resampled to 100

x 100 m by a bilinear resampling technique.

4.1.2.4 Model building

MaxEnt

We used MaxEnt v. 3.3.3k (Phillips et al., 2006) to model the spatial

distribution of the European badger. The MaxEnt algorithm uses

presence-only data. This is an advantage when working with a very low

density of target species at large scales, as we expected in the study area

based on Lara-Romero et al. (2012), due to the uncertainty in absences.

Although MaxEnt has been criticized on several occasions (see recently

Veloz, 2009; Yackulic et al., 2012), it is widely used for modeling the

spatial distribution of species for various purposes, e.g., testing model

performance against other methods (Elith et al., 2006) and using several

types of variables (Buermann et al., 2008), predicting species richness or

diversity (Graham & Hijmans, 2006), or forecasting distributions to

estimate variations with climate change/land transformation (Yates et


53

al., 2010). Finally, given that 1) the main goal of this study is to test the

performance of models using ecosystem functional variables, and 2)

prediction maps generated by MaxEnt are of interest as assessment

tools, but are not the goal itself, we considered MaxEnt a valid tool for

achieving our objectives.

Models

To test the utility of environmental functional variables in modeling the

spatial distribution of the badger, we combined the twenty variables into

four groups, with and without including ecosystem functional variables

(Table 4.1.1). We defined these four groups because they were the most

ecologically reasonable and of interest for comparison in keeping with

the objectives of this study. These groups of variables, along with the

badger presence data, were input to compute models. We used 10-fold

cross-validation of the occurrence locations. Each partition was made by

randomly selecting 75% of the occurrence locations as training data, and

the remaining 25% as test data. Then, each one of the partitions, along

with each of the four combinations of variables, was run in MaxEnt to

compute the models. We made 10 random partitions rather than a single

one in order to assess the average model behavior, and to allow for

statistical testing of observed differences in performance (Phillips et al.,

2006).

4.1.2.5 Model evaluation

Threshold-independent evaluation

We evaluated the performance of models created from different

combinations of variables using all discriminating thresholds within the

predicted area as suitable or unsuitable for badgers. We used (threshold-


54

independent) receiver operating characteristic (ROC) analysis for this, as

it uses a single measure, the area under the curve (AUC), to show model

performance. With presence-only data, the AUCPO (i.e., AUC estimated

with presence-only data) maximum was less than 1 (Wiley et al., 2003),

so we do not know how close to optimal a given AUCPO was.

Nevertheless, we were able to determine the statistical significance of

the AUCPO and compare the performance of different models (Phillips et

al., 2006). We employed a DeLong test (DeLong et al., 1998) to compare

AUCPO values for each combination of variables. The DeLong test is

designed to nonparametrically compare the difference between two

AUCs from two correlated ROC curves. The Z score is defined as the

difference of AUC divided by its standard error. Under the null hypothesis

(the difference in AUC is zero) Z has a standard normal distribution (Chen

et al., 2013). This test was computed in R (R Development Core Team,

2011).


55

Table 4.1.1. Groups of variables used for constructing models. Group 1: topography and climate, group 2: Land cover and Land uses, group 3: EVI

variables, group 4: EVI of land cover. Each model contained group 1, the ALL model all four groups, LC & LU model groups 1 and 2, the EVI model

groups 1 and 3, and the EVI LC model groups 1 and 4. Thus, three models included the ecosystem functional variables: EVI, EVI LC and ALL, and only

LC & LU model did not include these variables.

Variable Short name Groups of variables

Group 1 Group 2 Group 3 Group 4

Mean slope SLO X Annual Mean Rainfall MRAIN X Mean value of the maximum temperatures MMT X

Area of scattered scrub SSCRUB X

Area of dense scrub SDCRUB X

Area of woody crop SWCROP X

Area of arable crop SACROP X

Area of mixed crop SMICROP X

Area of mosaic crop SMOCROP X

EVI annual mean EVIMEAN X Standard deviation of EVI annual mean EVISTD X Coefficient of variation of EVI annual mean EVICV X EVI autumn mean AEVI X

EVI spring mean SEVI X EVI annual mean of scattered scrub SSCEVI X

EVI annual mean of dense scrub DSCEVI X

EVI annual mean of woody crop WCEVI X

EVI annual mean of arable crop ACEVI X

EVI annual mean of mixed crop MICEVI X

EVI annual mean of mosaic crop MOCEVI X

Group 1: topography and climate, group 2: Land cover and Land use, group 3: EVI variables, group 4: EVI of land cover. Each model

contained group 1, the ALL model all four groups, LC & LU model groups 1 and 2, the EVI model groups 1 and 3, and the EVI LC

model groups 1 and 4. Thus, three models included the ecosystem functional variables: EVI, EVI LC and ALL, and only LC & LU model

did not include these variables.


56

Information criteria

Following Warren & Seifert (2011), we implemented an Akaike

information criterion corrected for small sample size (AICc) (Burnham &

Anderson, 2002) in the MaxEnt models. We standardized raw scores for

each model, so that all scores within the study area added up to 1. Then

we calculated the likelihood of the data in each model by taking the

product of the suitability scores for each pixel showing presence. Both

training and test data were used in calculating likelihood. The number of

parameters was measured by counting all parameters with a nonzero

weight in the .lambda file produced by MaxEnt. All AICcs were computed

using ENMTools software (Warren et al., 2010).

Variable relative importance and response curves

We evaluated the relative importance of the variables using a jackknife

test on the AUCPO found from test data. Thus AUCPO was estimated by 1)

removing the corresponding variable, and then creating a model with the

remaining variables, 2) creating a model using each variable alone, and 3)

using all variables. Furthermore, we plotted the response curves for the

variables which caused the widest variations in the AUCPO. Curves were

estimated by generating a model using only the corresponding variable

and disregarding those remaining (Phillips et al., 2006).

4.1.3 Results

4.1.3.1 Occurrence of European badger

The field survey yielded 94 presence locations, mainly associated with

the two main rivers in the study area (see Appendix C). Landscapes near

the rivers had a larger supply of food resources for the European badger,


57

because crops are abundant there (Fig. 4.1.1). These presence records

are enough for this study since MaxEnt algorithm has been proved to

works well at different sample size (Hernández et al., 2006). 51 of the

records were footprints, 26 latrines, 15 setts and 2 road casualties.

4.1.3.2 Threshold-independent test

In 6 of the 10 partitions, combinations with all variables (ALL) yielded the

models with the highest AUCPO (Table 4.1.2). In 8 of the 10 partitions, the

AUCPO was higher for EVI and EVI LC than for the Land cover & Land uses

models, which were the lowest in most of the partitions.

Table 4.1.2. Comparison of threshold-independent receiver operating

characteristic (ROC) results for European badger using LC & LU, EVI, EVI LC and

ALL models. For each random partition of occurrence records, the maximum

AUCPO is marked in bold, the minimum underlined, and if the observed

difference between the maximum AUCPO and the rest is statistically significant

(under a null hypothesis that true AUCPOs are equal), it is marked with an

asterisk.

Data partition

LC & LU

EVI EVI LC

ALL

AUCPO AUCPO AUCPO AUCPO

1 0.722 0.669 0.725 0.722 2 0.63 0.701 0.634 0.674 3 0.644* 0.705 0.745 0.658 4 0.753* 0.756* 0.761 0.816 5 0.625* 0.785 0.69 0.69* 6 0.742* 0.793* 0.746* 0.808 7 0.734 0.726 0.707 0.741 8 0.673* 0.711 0.682 0.735 9 0.772 0.788 0.753 0.831 10 0.718* 0.758* 0.77 0.82

Average 0.701 0.739 0.721 0.749 Standard deviation

0.053

0.042 0.042

0.065

Maximum 0.772 0.793 0.770 0.831 Minimum 0.625 0.669 0.634 0.658


58

LC & LU: Land cover and Land uses. AUCPO: AUC estimated with presence-only data.

4.1.3.3 Information criteria

Table 4.1.3 shows Akaike weights found by models. It is accepted that

models with AICc differences (∆AICc) < 2 are plausible while models with

∆AICc values > 10 are rejectable (Burnham & Anderson, 2002). Thus, 6 of

the 10 data partitions accepted EVI and LC & LU as the most

parsimonious models, while one of the partitions accepted the EVI LC

model. ALL models were not plausible in any of the partitions.

Table 4.1.3. Number of estimated parameters (K), AICc differences (∆AICc) and

Akaike weights (Wi). The maximum Wi for each random partition of occurrence

records is marked in bold.

Partition data

LC & LU EVI EVI LC ALL

K ΔAICc Wi K ∆AICc Wi K ∆AICc Wi K ∆AICc Wi

1 24 0.00 0.73 30 1.98 0.27 33 29.28 0.0 46 40.63 0.0

2 26 17.72 0.00 27 0.00 1.00 28 35.18 0.0 49 91.35 0.0

3 27 29.05 0.00 28 0.00 1.00 37 58.81 0.0 51 107.23 0.0

4 28 0.00 0.55 33 0.41 0.45 37 44.61 0.0 47 34.45 0.0

5 25 0.00 1.00 34 15.58 0.00 29 11.82 0.0 47 53.23 0.0

6 29 13.74 0.00 30 0.00 1.00 34 38.57 0.0 54 108.75 0.0

7 27 15.15 0.00 26 0.00 1.00 31 39.43 0.0 43 27.42 0.0

8 30 0.00 0.94 33 5.39 0.06 35 25.30 0.0 47 34.13 0.0

9 25 0.00 0.69 34 14.69 0.00 27 1.61 0.3 53 94.24 0.0

10 25 0.00 1.00 34 19.93 0.00 28 12.72 0.0 50 66.34 0.0

LC & LU: Land cover and Land uses.


59

4.1.3.4 Relevant variables and their effects

We only analyzed the relative importance of variables from the ALL

model, with the maximum AUCPO value. Area of mosaic crop (SMOCROP)

caused a 2% reduction in AUCPO (Fig. 4.1.2b). Therefore, this variable,

along with others that caused a reduction of over 1% (Area of scattered

scrub (SSCRUB), EVI of mosaic crop (MOCEVI), mean maximum

temperature (MMT) and coefficient of variation of mean EVI (EVICV)),

provided the most useful information not present in the other variables.

We considered reductions about 2% and 1% as relevant, because these

percentages were above the third quartile (0.84%) of reduction values

percentage. EVIMEAN alone had the highest AUCPO (87.4% AUCPO with all

variables) (Fig. 4.1.2a) and therefore, this variable provided the most

useful information by itself. Apart from this, others like EVI spring, EVI

autumn, EVI scattered scrub and standard deviation of mean EVI, were

over 79%.


60

Figure 4.1.2. Jackknife test of variable importance for European badger in the

ALL model with maximum AUCPO. (a) Bars show the AUCPO with each variable

modeled separately. Ratios above the bars show the AUCPO percentage of the

reference value (0.831); (b) Bars show the AUCPO, when each variable is

extracted from the model. The ratios above the bars show the ratio decreased

by the AUCPO with respect to the reference value (0.831).

Variables such as scattered scrub area, mean maximum temperature,

standard deviation of mean EVI and EVI of scattered scrub exerted a

nonlinear effect on European badger habitat suitability, as predicted by

MaxEnt (Appendix D). On the contrary, mosaic crop area and mean

annual EVI, exerted a positive linear effect, while EVI crop mosaic and

coefficient of variation of mean EVI had a negative linear effect.


61

4.1.4 Discussion

4.1.4.1 Did the EVI-derived variables improve ecological

niche modeling of the European badger in arid landscapes?

EVI variables provided useful information that improved the ecological

niche modeling of European badger in arid Mediterranean landscapes.

Based on the AUCPO and AICc criteria, models built with EVI variables,

performed well in predicting the spatial distribution, while models

without them were inferior (based on AUCPO). We suggest that the

variables included in the EVI models underlie the spatiotemporal

dynamic of badger food resources by describing vegetation production

(EVIMEAN), seasonality (EVICV) and spatial heterogeneity (EVISTD). Thus,

areas with high EVI mean and EVI spatial heterogeneity represented

more suitable habitats for the European badger, while they rejected

areas with high EVI seasonality.

Our study showed that despite the fact that rainfall (expressed here as

mean annual rainfall, MRAIN) is considered the main driver of vegetation

growth in Mediterranean environments (Nemani et al., 2003), it did not

prove to be as good a predictor as the mean EVI (as proxy of primary

production) for European badger distribution. The higher performance of

EVI mean can be explained by the findings of Cabello et al. (2012b), in

which production derived from EVI in drylands reflects the variation of

the water use efficiency and its availability due to the features of

vegetation and lithology. In addition, EVI mean also reflected the NPP for

irrigated crops, which do not depend directly on rainfall (33% of crops in

the study area are irrigated).


62

Additionally, more seasonal environments in the study area (i.e., with

high EVICV values) represented zones with low habitat quality for

European badger. Johnson et al. (2002) suggested that badger densities

across Europe are associated with seasonal constraints, or some other

constraint(s) that covary with seasonality. EVI models predicted as

suitable, landscapes with little annual variation in EVI values,

corresponding with sites where the availability of food may be assured

even in summer, the season experiencing the most extreme shortages in

food. Similarly, Virgós & Casanovas (1999a) showed that a decrease in

summer rainfall reduces badger occurrence in Mediterranean mountains.

We also found that badgers selected areas with high EVI spatial

heterogeneity. Pita et al. (2009) described the Mediterranean rural

landscape as a shifting mosaic that benefits diversity and presence of

species as the European badger. The different types of traditional crops,

along with patches of semi-natural vegetation, especially scrub and/or

forest, yield a wide variety of food resources. We argue that the EVISTD

variable might detect these mosaic landscapes. However, although this

variable contributed positively to European badger habitat suitability, its

effect was nonlinear, suggesting that badgers would not need such

heterogeneity to survive in certain landscapes.

Both EVICV and EVISTD might depict variability of resources availability.

EVICV represents temporal variability in the availability of resources

because it is the dispersion of mean EVI throughout the year. In this

sense, if EVI in summer and winter are significantly different, the annual

temporal variability of EVI will be large. On the other hand, EVISTD

represents spatial variability because it is the standard deviation of mean


63

EVI into the potential territories of badgers. In consequence, high values

indicate that a landscape will be more heterogeneous.

4.1.4.2 Was the predicted spatial distribution across arid

lands consistent with the ecological preferences of the

European badger?

The distribution predicted by the EVI models was coherent with the

habitat preferences described for the European badger (see Appendix E

for further details of predicted distributions by models). Our results

reveal that badger´s presence in the study area was mainly associated

with sites near rivers where there were several different types of crops

and patches of natural vegetation. According to Lara-Romero et al.

(2012), in Mediterranean drylands the European badger prefers mosaic

landscapes consisting of fruit orchards and natural vegetation, which

provide shelter and food resources. In these environments, the diet is

diversified, with consumption of fruit increasing in some seasons (Barea-

Azcón et al., 2010). Fruits, insects and vertebrates have also been

described as relevant food resources for European badger in

Mediterranean environments (Rodríguez & Delibes, 1992; Revilla &

Palomares, 2002). Likewise, other authors have related the occurrence or

abundance of these items with satellite-derived vegetation indices, such

as EVI or Normalized Difference Vegetation Index (NDVI) (see Willems et

al., 2009; Lafage et al., 2013; Tapia et al., 2013).

EVI and EVI for Land cover models discriminated better between suitable

(i.e., mosaic landscapes with crops) and unsuitable areas (homogeneous

patches of dense xerophytic scrubs) than the LC & LU models (see Table

4.1.2 and Appendix E). EVI variables provided information for

discriminating between two patches with the same type of land use and


64

cover, but with different primary production, seasonality and spatial

heterogeneity. The EVI for Land cover models exhibited an intermediate

performance (Table 4.1.2). These models also used variables related to

primary production. However, such variables were averaged based on

the spatial classification derived by GIS cartography. These maps may not

represent relevant landscape features for the target species or

inadequate spatial resolution (Pearce et al., 2001).

Sites with high EVIMEAN and SMOCROP (area of mosaic crops) values

represented the most suitable habitats for the European badger.

However, the variable EVI mean of mosaic crops (MOCEVI) showed a

negative effect on badger presence, which could be explained by the fact

that 1- mosaic crop variable, in turn, encompasses different types of

crops, and 2- badger presence records with high EVI values, are

associated with non-irrigated almond crop, which would not favor

badger presence in those areas. This suggests that in particular

landscapes, the type of land use would be more decisive for badger than

its associated production.

Removal of variables such as SWCROP (area of woody crop) and AEVI (EVI

autumn), did increase performance, meaning that such variables reduced

the generality of the model. This is, models made with these variables

appear to be less transferable to other geographic areas or to projected

future distributions by applying future conditions (Phillips, 2006).

Regarding the potential bias of the selected study area on results, we

consider that the study area contained enough variability to ensure that

its effect was minimized. Probably, a larger buffer would provide similar

results because the area between both rivers has not crops. In


65

Mediterranean arid landscapes, the major landscape variability is

generally associated with areas near rivers (Corbacho et al., 2003) and

along altitudinal gradients, just what we defined with our study area.

4.1.4.3 Ecosystem functional dimension in species

ecological modeling and conservation

The incorporation of remotely sensed characterization of the ecosystem

functional dimension in management and monitoring of species and

populations is gaining attention in conservation biology (Cabello et al.,

2012a). Ecosystem functional dimension provides proxies showing

biodiversity patterns and new tools and criteria that can assist in

designing conservation planning and actions. Some examples are shown

by Bardsen & Tveraa (2012), who used vegetation production estimated

by EVI to advance knowledge of the reproductive biology of reindeer

(Rangifer tarandus) in Norway; Oindo (2002), who predicted mammal

species richness and abundance using multi-temporal NDVI data; or

Wiegand et al. (2008), who studied the relationship between brown bear

(Ursus arctos) habitat quality and the seasonal course of NDVI as a proxy

for ecosystem functioning in the northern Iberian Peninsula.

Ecological modeling of the European badger in the Iberian Peninsula has

to date been addressed mainly using landscape structural variables

estimated from visual field observation (transect scale) (Virgós &

Casanovas, 1999b) and by GIS information (regional scale) (Rosalino et

al., 2004). Even though these variables that reflect landscape structure

are essential to modeling the species distribution (Rosalino et al., 2008),

they do not reflect the role of ecosystem functioning indicators or their

bidirectional relationship with the conservation of biodiversity and

ecosystem processes (Cabello et al., 2012a). However, Pettorelli et al.


66

(2005) and García-Rangel & Pettorelli (2013) point out some constrains of

remote sensing data to wildlife studies such as select the most suitable

processing to eliminate noise in the data, insufficient temporal resolution

to precisely date phenological phenomena, and economic disadvantages

due to many satellites still produce data that are not free.

Our study is the first to show that incorporation of ecosystem functional

variables (EVI-derived) improves the prediction of spatial distribution

modeling of the European badger in arid landscapes, considered

especially sensitive to Global Change (Lavorel et al., 1998). In this sense,

Pettorelli et al. (2005) suggested that satellite-derived indexes, such NDVI

or EVI, could be used to predict the ecological effects of environmental

change on ecosystems functioning and animal population dynamics and

distributions, due to their correlation with vegetation biomass and

relationship with climate variables.

Finally, we found that EVI variables represented relevant ecological

parameters for the description of the distribution of the European

badger as they can indicate 1) a high NPP associated with orchards or

fruit crops, very important for its survival in Mediterranean arid

landscapes (Rodríguez & Delibes, 1992; Lara-Romero et al., 2012), 2)

seasonality in the primary production, which can be seen as a surrogate

of habitat quality (Johnson et al., 2002), and 3) spatially heterogeneous

landscapes which provide different food resources (Pita et al., 2009).

However, these variables should be tested in other areas of its

distribution range. Models including EVI variables perform better (based

on AUCPO) than models not including these variables. Additionally,

continuous availability, both spatially and temporally, of remote sensing


67

data can improve the accuracy of monitoring and modeling wildlife for

conservation purposes in arid ecosystems throughout the world.

RESULTADO 4.2 Feeding habits of European badger (Meles meles) in Mediterranean arid landscapes

68

RESULTADO 4.2: FEEDING HABITS OF EUROPEAN

BADGER (Meles meles) IN MEDITERRANEAN ARID

LANDSCAPES

Tejón en busca de alimento en la quietud de la noche entre la densidad de la

vegetación de ribera. Escena capturada con fototrampeo en la Rambla de los

Molinos. Abril 2013. Paraje Natural del Desierto de Tabernas.

Basado en:

J. M. Requena-Mullor, E. López, A. J. Castro, E. Virgós, H. Castro. Hábitos

alimenticios del Tejón europeo en un paisaje árido Mediterráneo.

Galemys, aceptado con revisión menor.

J. M. Requena-Mullor, E. López, A. J. Castro, E. Virgós, H. Castro.

Landscape influence in feeding habits of European badger (Meles meles)

in arid Spain. Mammal Research, en revisión.


69

Objetivo 2: “Describir y comparar la dieta del Tejón a través de paisajes

áridos con diferente cobertura vegetal y uso del suelo.”

Hipótesis 2: “Debido al carácter generalista descrito para el Tejón en

ambientes Mediterráneos (Roper, 1994), el comportamiento trófico de la

especie podría variar entre paisajes áridos con diferente cobertura

vegetal y uso del suelo, y explotar en cada caso distintos recursos

alimenticios.”

Somos lo que comemos.

Ludwig Feuerbach.


70

Abstract

This study evaluates the influence of landscape on feeding behavior of

European badger (Meles meles) in southern Iberian Peninsula. We

particularly explore whether different vegetation types and land uses

affect its feeding habits across three arid landscapes; maquia, xeric

shrubland, and forestry. Although diet of badger has been described as

frugivory in Mediterranean environments, with cultivated and wild fruits

as key items (e.g., olives or figs), its feeding strategy may vary in response

to the composition of landscape consuming different key items in an arid

Mediterranean context. Based on 252 scats monthly collected from June

2011 to May 2012, we found that diet significantly varied among

landscapes studied: insects, carob, and small mammals were the key

items in the maquia, figs and oranges in the xeric shrubland, and

earthworms and insects in the forestry. This shows that within an arid

context, badger adapt its feeding behavior to particular landscape

conditions, and specifically, the diet of badgers shift from an animal-

based to another where cultivated fruits reaches the high importance.

Our results support the key role of human activities in shaping badger

behavior and diet and illuminating the contrasting dietary differences of

badger living on more pristine habitats versus those inhabiting human-

made habitats. Based on the proved effect of precipitation and land

management practices on items here identified, we also discuss

implications of Global Change drivers in badgers feeding habits for the

arid Mediterranean region.

Keywords: Meles meles, European badger, mustelids, diet, frugivory,

Iberian Peninsula, drylands, land use-land cover


71

4.2.1 Introduction

Feeding habits of the European badger Meles meles Linnaeus, 1758 have

been extensively studied and are one of its best-known ecological

features (Goszcynski et al., 2000; Virgós et al., 2005a). Their diet is very

diverse throughout their distribution range, with a wide assortment of

trophic strategies (Melis et al., 2002). The European badger is considered

a specialist forager for earthworms in Britain and others areas of

northwest Europe (Kruuk & Parish, 1981; Kruuk, 1989; but see Roper,

1994). In the Mediterranean region, earthworm availability is lower than

in north Europe due to a low precipitation and different landscape

composition (Virgós et al., 2004). In these environments, the species is a

trophic generalist (Roper et al., 1994) consuming fruits, insects, and

vertebrates (Piggozi, 1991; Rodríguez & Delibes, 1992; Barea-Azcón et al.,

2010). However, badgers can also specialize in earthworm consumption

in rainy mountainous Mediterranean areas, so it can be considered as a

facultative specialist at local scale, taking the most profitable resource

depending on supply and availability (Martín et al., 1995; Virgós et al.,

2004).

Despite the discussion about the feeding specialization of badgers, it has

been suggested an important effect of consumption of earthworms as

compared to other trophic resources on life-history traits of the species

such as population density and reproductive success (da Silva et al.,

1993; Woodroffe & Macdonald, 1993; Virgós et al., 2005a). However,

badgers can survive at low densities in extreme arid landscapes (Lara-

Romero et al., 2012; Requena-Mullor et al., 2014), where earthworms

are absent or very scarce. Despite the interest of these regions to a wide

interpretation of feeding strategies of badgers, only two studies dealing


72

with badger diet in arid environments (Rodríguez & Delibes, 1992; Barea-

Azcón et al., 2010). Rodríguez & Delibes (1992) described the diet only

during summer in a landscape with xerophitic vegetation and crops.

Barea-Azcón et al. (2010) analyzed the annual diet in a region with

continental climate (14 °C/year and 620 mm/year) but during an

especially dry year (250 mm), and in a landscape dominated by olive tree

plantations (Olea europea), dense pine forest reforestations (Pinus

halepensis) and some holm oak patches (Quercus rotundifolia) and both

studies emphasized the importance of cultivated fruits and rabbits

(Oryctolagus cuniculus) in the diet of badgers living at these

environments.

Fruits are largely consumed for badgers in some regions, Rosalino &

Santos-Reis (2009) detected an increase in fruit consumption along a

west to east Mediterranean gradient in a diverse guild of frugivory

mammals including the European badger. These authors argued that fruit

consumption depends on several factors such as fruit characteristics

(e.g., pulp content), availability of wild and cultivated fruits, and

abundance of other food resources (see also Herrera, 1989), which vary

throughout the Mediterranean basin. In fact, some authors have

highlighted cultivated fruits as a key food resource for badgers in

Mediterranean arid environments (Pigozzi, 1991; Rodríguez & Delibes,

1992; Barea-Azcón et al., 2010) which encourage the importance of the

orchards as a key habitat for the species in these environments (Lara-

Romero et al., 2012; Requena-Mullor et al., 2014). Thus, in arid habitats,

cultivated fruits could replace earthworms as the key food in badger life-

history traits, allowing increase its abundance and performance.


73

Feeding strategies of badgers are very flexible and current tactics can be

modified if environmental conditions change. The Mediterranean region

is especially vulnerable to Global Change drivers (Sala et al., 2000; Giorgi

& Lionello, 2008). Aridity is expected to increase in the Iberian Peninsula,

especially in arid zones (Giorgi & Lionello, 2008) due to increasing

temperatures and decreasing rainfall, particularly during summer months

(De Luis et al., 2001). Additionally, agricultural lands represent more than

half of the Mediterranean region by area (Olesen & Bindi, 2002). The

reform of the Common Agricultural Policy (PAC) for 2014-2020 outlines

steps to promote crop diversification, establish and maintain permanent

pastures, and leave some land fallow to restore natural ecological

processes (Martínez & Palacios, 2012). These measures would benefit

the conservation of badger populations in Mediterranean environments

(Virgós et al., 2005b). Nevertheless, aging of the rural population and

young peoples’ exodus to the cities, is leading to widespread land

abandonment in last decades (Castro et al., 2011). This demographic

change is causing a deterioration of the rural Mediterranean landscape

which is very important for the conservation of some species that

depend on agricultural areas (Zamora et al., 2007; Pita et al., 2009). This

is true for badgers, which are very linked to traditional human activities

and agricultural practices, especially in arid environments (Kruuk, 1989;

Virgos et al., 2005b; Lara-Romero et al., 2012). Consequently, in the next

future two main changes can modify Mediterranean regions, first most

areas of the Mediterranean can be transformed to arid habitat; second,

most agricultural areas will be abandoned. Under these new scenarios,

food resources for badgers can change dramatically, and this can affect

other life-history traits such as population density, social organization or

population growth (Macdonald & Newman, 2002; Macdonald et al.,


74

2010). Therefore, the diet of badgers in arid regions of the

Mediterranean can offer an opportunity to forecast what resources can

be essentials for badgers in the new scenario of increasing aridity for

most of the Mediterranean region, and elucidate how changes in

agricultural practices can impact on badger diet and then other

important traits of the species.

This study uses three landscapes (maquia, xeric shrubland, forestry) in

environments of southern Iberia Peninsula which represents the extreme

ecological and geographical range of the species to explore whether

different vegetation types and land uses affect the feeding behavior of

badger. Based on previous findings regarding the effect of precipitation

and land management practices on food resources for badgers, potential

implications of Global Change drivers in badgers feeding habits are

discussed for arid Mediterranean areas.

4.2.2 Materials and methods

4.2.2.1 Localization and description of landscapes

This study was conducted in the southeastern Iberian Peninsula (36°06’N,

2°17’E) (Fig. 4.2.1). This region is the most arid in Europe (Armas et al.,

2011), represents some of the most extreme arid conditions inhabited by

badgers, and contains a wide variety of mixed arid environments with

Mediterranean rural landscapes. The aridity of target landscapes was

characterized based on the Martonne aridity index (Martonne, 1926) (Ia),

considering arid climate a Ia between 5 and 15 (Requena-Mullor et al.,

2014).

On a local scale, diet composition of the European badger depends

primarily on the management and soil use by humans (Kruuk, 1989;


75

Fischer et al., 2005). Therefore, to compare the diet among landscapes,

we selected three landscapes with different land cover and use (maquia,

xeric shrubland and forestry). In each landscape, first we identified a zone

with latrines frequently used by the species. Then, we drew a 3 km-

radius buffer zone using the latrines as the centroid (Fig. 4.2.1). This

method ensures the inclusion of the potential home range size estimated

for European badgers living at these poor environments (i.e., 9 km2)

(Lara-Romero et al., 2012). Finally, we characterized the type of land

cover and use within the buffers based on GIS cartography from

Andalusia Land use/Land cover map (scale 1:25.000, 2007).


76

Figure 4.2.1. Location of landscapes within the study area in Almería province,

Spain. In each landscape, we identified a zone with latrines frequently used by

European badger (Meles meles). Then, we drew a 3 km-radius buffer zone using

the latrines as centroid.

Landscape 1: Maquia (37°08’N, 1°55’E)

Maquia has the lowest altitude (102 m) and an Ia of 11.72 (0.004). The

mean annual rainfall is 340 mm/year and mean annual temperature is


77

19°C. The area is mainly dominated by non-forested natural vegetation

(65% of the total area) including dense shrubland (e.g., Macrochola

tenacissima, Pistacia lentiscus) and sparse shrubland (Rhamnus lycioides,

Anthyllis cytisoides). Agricultural uses comprise the 22% of the area,

where homogeneous herbaceous crops (i.e., without natural vegetation)

are predominant, with irrigated and rainfed crops in similar proportions.

It is important to highlight the abundance of wild vegetation with fleshy

fruits (e.g., Ceratonia siliqua, Ficus carica, Chamaerops humilis, Vitis sp.,

O. europea var. silvestris) in the watercourses.

Landscape 2: Xeric shrubland (36°58’N, 2°29’E)

Xeric shrubland is located at 228 m altitude, and has an Ia of 6.98

(0.012). The mean annual rainfall is 200 mm/year and the mean annual

temperature is 18 °C. 70% of the area is covered by sparse xeric

shrubland (M. tenacissima, Salsola genistoides, Anthyllis terniflora).

Crops occupy only the 12% of the landscape, where the 44% of them are

woody irrigated crops (e.g., Citrus sp.) and the rest are greenhouses. The

remaining 18% of the area is occupied by minority uses.

Landscape 3: Forestry (37°06’N, 2°46’E)

Forestry is located at 1320 m altitude, and has an Ia of 12.54 (0.028).

Mean annual rainfall is 310 mm/year and mean annual temperature is

12°C. Natural vegetation covers the 84% of the area, with forestry (e.g.,

Pinus spp.) and shrublands (e.g., Genista spp., Adenocarpus decorticans)

in similar proportions. Agricultural uses represent the 16% of the total

area, in which mosaics of natural vegetation with rainfed crops (O.

europea and Prunus dulcis) are predominant.


78

4.2.2.2 Diet analysis

The analysis of the diet consisted in collecting scats in badger latrines and

examining composition in the laboratory. Faeces were collected once a

month from the latrines across the three target landscapes. Sampling

was conducted from June 2011 to May 2012. In each visit, we recorded

the number of latrines and scats contained. Prior to the collection period,

all scats were removed from latrines at the end of May 2011. Faeces

were classified based on their water content, shape and colour. If this

was impossible, the entire latrine content was taken as a single faeces

(Pigozzi, 1991). Washing and sieving was carried out according to the

Kruuk & Parish (1981) protocol. For each scat, the total number of items

were counted or extrapolated from the remains, following the Kruuk &

Parish (1981) and Pigozzi (1991) methods. Items were gathered in three

broad categories: fruits, vertebrates, invertebrates. Nevertheless, in

order to identify what food resources were dominant within each

landscape, we carried out a finer grouping, classifying each remainder to

the lowest taxonomic level in each case (i.e., species). The food remains

were compared with reference collections to ensure a correct taxonomic

determination. Lastly, data collected was grouped by annual seasons, i.e.,

summer (June, July, August), autumn (September, October, November),

winter (December, January, February), spring (March, April, May). For all

categories (both in the wide and fine classification), we estimated the

frequency of occurrence FO (%) and the relative volume RV (%), the latter

was assessed visually based on the Kruuk & Parish (1981) method. In

addition, we calculated the Shannon´s diversity index for each landscape

and season (Shannon, 1948):

- ∑ lo i (1)


79

where Pi= proportion of each food category. The diversity is minimum

when H = 0 and it is maximum when H = log n (n = number of categories).

We considered three wide categories, so H maximum = 0.47.

4.2.2.3 Data analyses

Landscape type and seasonal differences in the relative volume of wide

categories (i.e., fruits, vertebrates and invertebrates) were analyzed by a

two-way ANOVA with the relative volume of each category as a response

variable and the landscape type and season as fixed factors (Virgós et al.,

2004). When the effects of landscape or season were significant, we

employed a Duncan´s test to reveal which landscapes or seasons showed

large differences in relative volume for each item considered. All

residuals were checked for normality (Shapiro-Wilk Normality test)

(Shapiro & Wilk 1965), and homogeneity of variances (Bartlett test)

(Snedecor & Cochran, 1989). In general, the residuals showed no-

normality and heterocedasticity, so we resampled the data by

bootstrapping (10000 replicates) for estimate the F´s distribution and

recalculated the critical value for signification (i.e., 0.05) (Appendix F).

Due to the importance of earthworms for badgers in others areas of its

distribution (Kruuk, 1989; Virgós et al., 2004), we also analyzed

separately the consumption of this item so that our data can be

compared to previous studies.

Correlations between diet diversity (measured by the Shannon index)

and the relative volume of wide categories, between diet diversity and

main consumed fine categories, and among all wide categories, were

analyzed using the Spearman´s rho statistic (Best & Roberts, 1975).


80

Finally, we used a nonparametric multidimensional scaling (NMDS) to

explore similarities in the diet based on both spatial (landscapes) and

temporal (seasonal) variations. With this aim, we arranged the seasons

for each landscape in a Cartesian axis based on the grouping

consumption of fruits, vertebrates and invertebrates (in RV). Regarding

this, a shorter distance between seasons means greater similarity and

vice versa. We used the Bray-Curtis distance (djk) to compute the NMDS.

∑ – ∑ ii (2)

where x is the RV (%) of food category i (i.e., fruits, vertebrates,

invertebrates) in the j and k seasons. To check the goodness of NMDS we

measured the concordance in the rank order of the observed inter-

season distances and those predicted from the similarities. One measure

of fit is the Kruskal´s stress (Kruskal, 1964). Clarke (1993) provided some

guidelines for stress values, so stress values greater than 0.3 indicate the

achieved configuration is no better than random.

All statistical analyses were carried out using R software version 2.14.2 (R

Development Core Team, 2012)

4.2.3 Results

4.2.3.1 Diet composition

A total of 252 faeces were collected, 54 in maquia, 140 in xeric

shrubland, and 58 in forestry (Table 4.2.1). In maquia, fruits were

consumed throughout the year (annual FO = 81%), although the mean

relative volume was greatly reduced in summer (VR = 29.3% ±4.73 SE).

Carobs were the most frequent fruit in excrements (annual FO = 46.29%)

(see Table G.1 in Appendix G). Grapes (47.5% ±12.5 SE) (summer), carobs

(82% ±8.98 SE and 96.4% ±2.82 SE) (autumn and winter) and oranges


81

(63% ±27.5 SE) (spring) were the fruits with the highest mean relative

volume. Vertebrates were more frequent in spring (FO = 72.7%) (Table

4.2.1), with small-mammals (e.g., rodentia) as key prey and obtaining a

relevant mean volume (RV = 70% ±17.75 SE spring and RV = 82.5% ±8.63

SE winter) (Table G.1). Invertebrates were very frequent in summer (FO =

100%) (Table 4.2.1) (mainly Coleoptera) but in a low mean volume

(25.3% ±5.63 SE) (Table G.1).

Table 4.2.1. Mean relative volume (%) and frequency of occurrence (in

parentheses, %) values for the different wide categories considered in each

landscape and season. We reported values of frequency of occurrence for

comparative purposes with other studies using these categories.

Landscape and prey item Summer Autumn Winter Spring

Maquia

Fruits 29.3 (100) 51.2 (100) 68.0 (72.0) 42.5 (72.7)

Vertebrates 20.3 (16.7) 45.8 (16.7) 76.5 (40.0) 64.1 (72.7)

Invertebrates 15.0 (100) 8.5 (58.3) 34.8 (24.0) 21.1 (36.4)

N 6 12 25 11

Xeric shrubland

Fruits 61.8 (96.2) 88.6 (100) 75.2 (95.5) 74.7 (92.7)

Vertebrates 73.8 (15.4) 20.0 (13.3) 39.5 (22.7) 23.3 (21.8)

Invertebrates 6.8 (42.3) 6.0 (40.0) 18.7 (54.5) 30.3 (70.9)

N 26 15 44 55

Forestry

Fruits 54.6 (41.7) 66.1 (70.0) 68.5 (40.0) 50.1 (9.5)

Vertebrates 30.0 (8.3) 27.6 (30.0) 61.3 (26.7) 53.4 (57.1)

Invertebrates 34.2 (100) 40.0 (80.0) 39.5 (86.7) 35.8 (81.0)

N 12 10 15 21


82

In xeric shrubland, fruits appear to be the key food resource for badger

based on both the frequency of occurrence (annual FO = 95%) and

relative volume (mean annual RV = 73.41% ±2.49 SE). Two types of fruits

were mainly consumed, figs (FO = 73% and 93.3%; mean RV = 78.4%

±6.27 SE and 94% ±2.91 SE) (from summer to autumn) and oranges (FO =

84% and 67.2%; mean RV = 80% ±4.52 SE and 83.6% ±4.42 SE) (from

winter to spring) (Table G.2). Vertebrates had low relevance throughout

the year, appearing only with a considerable volume in summer (mean

RV = 73.8% ±18.18 SE) (Table 4.2.1), with anura (mean VR= 100%) and

rabbits (VR= 87.5% ±2.5 SE) being the predominant taxonomic groups

(Table G.2). Invertebrates were more frequent in spring (FO = 70.9%) and

winter (FO = 54.5%) (Table 4.2.1), although their volume was not

important in any season (mean annual RV = 17.62% ±2 SE). Orthoptera

was the most frequent along the year (annual FO = 34%) (Table G.2).

In forestry there was a high consumption of invertebrates throughout the

year (annual FO = 86%), although with a low relative volume (mean

annual RV = 36% ±3 SE). Scorpions (FO = 50%; mean RV = 35.8% ±5.78 SE)

(summer), Hymenoptera (FO = 60%; mean RV = 54% ±11.57 SE)

(autumn), earthworms (FO = 53.3%; mean RV = 81.8% ±4.99 SE) (winter),

and caterpillars (FO = 66.6%; mean RV = 44.1% ±8.35 SE) (spring) were

the most relevant invertebrates (Table G.3). Fruits were consumed

mainly in autumn (FO = 70%) (Table 4.2.1), with figs (mean RV = 60%

±5.77 SE), wild blackberries (mean RV = 90% ±10 SE) and almonds (mean

RV = 50% ±) (pericarp fleshly) most abundant (Table G.3). The relative

volume of fruits was considerable throughout the year (mean annual RV

= 62% ±6.64 SE). Vertebrates were consumed more frequently in spring

(FO = 57.1%) (Table 4.2.1). Spring and winter were the seasons with the


83

highest volume of vertebrates in the faeces (mean RV = 53.4% ±11.08 SE

and mean RV = 61.3% ±12.31 SE respectively), mainly due to reptiles and

rabbits (Table G.3).

4.2.3.2 Effects of landscape and season on diet

Consumption of fruits showed statistically significant difference for

landscape and season when the two factors were separately considered,

however, the interaction between both factors was not significant (Table

4.2.2). Relative volume was higher in xeric shrubland than in maquia

(Duncan´s test, p < 0.01) (see Fig. 4.2.2). Moreover, fruit consumption

was higher in winter than in summer (Duncan´s test, p < 0.01) (Fig. 4.2.2).

Table 4.2.2. Results of the two-way ANOVA with season and landscape type as

fixed factors and the relative volume of the wide categories as the response

variable. No effect varied its signification by applying the bootstrap resampling

(see Table F.1 Appendix F).

Relative volume Effect df F p

Fruits Season 3 4.04 <0.01

Landscape 2 13.3 <0.001

Season x Landscape 6 1.22 0.29

Vertebrates Season 3 2.07 0.11

Landscape 2 5.56 <0.01


Invertebrates Season 3 2.85 <0.05

Landscape 2 11.61 <0.001


We also observe significant differences in the consumption of

vertebrates between landscapes, but not between seasons (Table 4.2.2).


84

Vertebrates showed significantly higher relative volume in maquia than

in xeric shrubland (Duncan´s test, p < 0.01).

For invertebrate consumption, we observed differences among

landscape and season even though the interaction was marginally not

significant (Table 4.2.2). More invertebrates were consumed by badgers

in forestry than those in maquia and xeric shrubland (Duncan´s test, p <

0.01). Invertebrates were higher consumed in spring than in autumn

(Duncan´s test, p < 0.05).


85

Figure 4.2.2. Estimated relative volume (%) for each main category considered in

the different seasons and landscapes. Whiskers represent the standard error of

the mean values of categories.


86

4.2.3.3 Earthworm consumption

Earthworms were consumed in the three landscapes but always with a

low frequency of occurrence (mean annual FO = 3.7%, 2.8% and 17.2%

for maquia, xeric shrubland and forestry respectively) (Tables G.1, G.2

and G.3). However, a high relative volume was found (mean annual RV =

90% ±10 SE, 50% ±12.24 SE and 77.7% ±4.82 SE for maquia, xeric

shrubland and forestry).

For the relative volume, there was only difference between seasons

(ANOVA, F = 59.2, p < 0.0001) and it was higher in winter than in spring

(Duncan´s test, p < 0.0001) (Fig. 4.2.3).

Figure 4.2.3. Interaction between landscape and season for earthworm relative

volume (%) in the diet. Whiskers represent the standard error of the mean

values of categories.


87

4.2.3.4 Diet diversity and food consumption

Diet diversity was similar in the three landscapes, but the more diverse

annual diet was obtained in maquia (H = 0.39), followed by forestry (H =

0.37) and xeric shrubland (H = 0.30). Diet diversity was positively

correlated with fruit and vertebrate consumption (rho = 0.44, p = <0.001,

n = 36; rho = 0.58, p < 0.001, n = 36). Conversely, the diversity was

negatively, but not significantly, correlated with the invertebrate

consumption (rho = -0.17, p = 0.30, n = 36). Food resources with a high

consumption (e.g., carobs, figs, oranges, coleopteran and earthworms,

see Tables G.1, G.2 and G.3) showed a no significant correlation with the

diet diversity, except rodentia which showed positive correlation (rho =

0.41, p < 0.05, n = 36). Between wide categories, relative volume of fruits

and invertebrates showed negative correlation (rho = -0.41, p < 0.05, n =

36), indicating higher consumption of fruits in landscapes or seasons

where invertebrates are not very important in the diet of badgers. Fruits

with vertebrates and vertebrates with invertebrates, showed a low non

significant correlation (rho = -0.03, p = 0.84, n = 36; rho = -0.10, p = 0.53,

n = 36).

4.2.3.5 Nonparametric multidimensional scaling

The spatial configuration reached by the NMDS (Fig. 4.2.4) was better

than random (Kruskal´s stress < 0.2). By seasons, the summer in the

maquia (mean djk = 0.37) and the autumn in the xeric shrubland (mean djk

= 0.31) had the greatest distance (Fig. 4.2.4 and Appendix H).

The diet in the maquia had the greatest distance between seasons (mean

djk = 0.34), while in the xeric shrubland the distances were intermediate

(mean djk = 0.26), and in the forestry, the seasons were the closest (mean


88

djk = 0.15) (Appendix H). The summer in the maquia and the autumn in

the xeric shrubland showed the lowest Shannon´s diversity index (0.26

and 0.27 respectively) (Fig. 4.2.4).

Figure 4.2.4. Nonparametric multidimensional scaling (NMDS). The axis NMDS1

and NMDS2, show the range of the distances reached between seasons in the

three landscapes. Seasons are arranged so that the distances between them are

as close to the real differences between the mean relative volume (%) of fruits,

vertebrates and invertebrates consumed in each landscape. A lower distance


89

between seasons means greater similarity between them and vice versa.

Isoplets are based on the Shannon´s diversity index.

4.2.4 Discussion

4.2.4.1 Spatial and temporal variation of badger diet

Despite of the common arid environmental features across all locations,

badger diet varied significantly across the three landscapes. One or two

dominants food resources in each of the cases studies were identified.

This supports Virgós et al. (2004), in other Mediterranean areas where

feeding behaviour of badgers can vary among close locations but with

different habitat type, rainfall regimen or human land use.

We found that in maquia, badgers were mainly frugivores although

vertebrates were also important in winter and spring. This feeding

strategy has been also described in other Mediterranean areas (Piggozi,

1991; Rosalino et al., 2004). Carobs were the most consumed fruit

throughout the year. This is the first time that this food resource is

described in this type of landscape. Carobs were available and abundant

along the year (personal observation), highlighting the generalist-

opportunist character of the European badger (Pigozzi, 1991). The fruit

consumption was however lower than in xeric shrubland, besides this,

the fruits consumed were mainly wild, so the direct dependence of

badgers for orchards would be less decisive. Small-mammals were the

second most important item, probably providing the necessary proteins

in a vegetarian diet (Ciampalini & Lovari, 1985).

In xeric shrubland, fruits, especially cultivated, were the most important

food resource. Oranges from winter to spring and figs from summer to

autumn as the most consumed items. While figs have been described in


90

dry Mediterranean landscapes as an usual food resource for badgers

(Barea-Azcón et al., 2010), this is the first time that oranges are described

across these environments. This landscape is the most arid of the cases

studies; therefore, the importance of orchards for the survival of badgers

in these environments is reinforced (Lara-Romero et al., 2012).

In forestry, badger diet mainly consisted in insects when frequency of

occurrence was considered, but fruits prevailed again when relative

volume was taken into account. A large consumption of insects supports

findings by Virgós et al. (2004) in some of the habitats sampled in the

mountain areas of central Spain. However, the relevance of fruits also in

this habitat highlights the key importance of this food resource for

badger in all the arid landscapes.

When badgers preferentially consume the same type of food resource

along the year, such as fruits in xeric shrubland or invertebrates in

forestry, the diet similarity between seasons is higher. On the other

hand, the highest distances between seasons (e.g., summer vs. winter in

maquia, summer in xeric shrubland vs. summer in maquia) were mainly

due to the differences in the vertebrates and invertebrates consumption.

In Mediterranean arid landscapes, the availability of earthworms is

scarce (Virgós et al., 2005a). Our results show that badgers consumed

this item across maquia and xeric shrubland, supporting results obtained

by Barea-Azcón et al. (2010), for other Mediterranean area. In forestry,

the consumption of earthworms was even larger than in the other

landscapes. This increase may be due to a greater availability of

earthworms in mid-mountain areas because of the altitude effect, which

causes that rainfall is larger (Virgós et al., 2005a). However, the


91

consumption of earthworms was not uniform across the year, appearing

only in winter and spring, because of this prey item is mainly available in

the rainy and mild conditions of spring and autumn-winter (Edwards &

Lofty, 1977; Kruuk & Parish, 1981). All these results combined reinforce

the usefulness of this trophic resource for badgers when available, and

its large importance even in the extreme arid conditions of the south

edge of its distribution range, contradicting the original ideas about the

unimportance of earthworms in most of the southern badger range

(Ciampalini & Lovari, 1985; Pigozzi, 1991; Roper, 1994; Martín et al.,

1995; but see Virgós et al., 2004).

Olives were not relevant in any of the three landscapes. This disagrees

with findings in other Mediterranean areas (Kruuk & de Kock, 1981;

Rosalino et al., 2004; Barea-Azcón et al., 2010). This may be due to their

low availability in the sampled localities (personal observation), as

expected of a generalist (or facultative specialist) species such as

European badger (Pigozzi, 1991). Badgers shifted their preferences for

different fruits when availability changes.

Our results showed that the diet diversity did not decrease with the

consumption of specific items. This supports the generalist character of

badger in the arid Mediterranean environments described by Rodríguez

& Delibes (1992) and Barea-Azcón et al. (2010). Despite the large

importance of fruits, this cannot be viewed as of similar importance of

earthworms or rabbits in other areas, where diversity is strongly

associated to the consumption of these resources (Kruuk, 1989; Martín

et al., 1995). On the other hand, fruits and invertebrates showed a

negative correlation. An explanation for this could be badgers replace the

shortage of fruits per invertebrates (mainly insects which are predictable


92

and abundant). This supports similar strategy of badgers from the central

Spain mountains when earthworms are scarce (Virgós et al., 2004).

Our results reinforce the feeding behavior described for the European

badger in Mediterranean landscapes by Rodríguez & Delibes (1992) and

Barea-Azcón et al. (2010). We also found that two food resources (i.e.,

oranges and carob trees) have a high presence in the diet, being

described for first time as key food resource in badger survival across

Mediterranean arid landscapes. These items have a clumped distribution,

showing a high seasonal abundance and a high energetic value (Agencia

Andaluza del Agua, 2010). The same features have been cited for other

key resources of the species (Kruuk, 1989; Martín et al., 2005; Rosalino et

al., 2005).

4.2.4.2 Potential implications of climatic change and land

use change on feeding habits

Climate change and land use change have been described as two of the

main direct drivers of Global Change in Mediterranean environments

(Vitousek, 1994; Giorgi & Lionello, 2008). On one hand, it has been

proved that the abundance of European badger is related with climatic

characteristics (Virgós & Casanovas, 1999) and their seasonality (Johnson

et al., 2002). Many of the items consumed by badgers across the three

landscapes depend directly (e.g., carobs, figs, blackberries, fan palm

fruits, earthworms) or indirectly (e.g., insects, small-mammals, rabbits)

on precipitation. Theoretical models about species coexistence predict

that a larger temporal variation of resources reduces probability of

specialization in feeding habits (Wilson and Yoshimura 1994). Regarding

this, badgers would have a facultative specialist behaviour in more

variable environments (Virgós et al., 2004), modifying their feeding


93

strategy and/or the choice of food, in order to maximize the intake in

accordance with their availability (Pigozzi, 1991). Accordingly, the species

would prefer landscapes with a high spatial-temporal predictability in

food resources (Mellgren & Roper, 1986; Kruuk, 1989). These landscapes

could be associated with orchards presence, representing a predictable,

clumped and profitable habitat to provide food for badgers. This has

been recently described by Requena-Mullor et al. (2014), showing that

the presence of agricultural humanized landscapes increases the

opportunity of use different food resources over time and space.

Traditional agricultural practices provide badgers with other food

resources such as figs, loquats, apples, and apricots that also indirectly

depend on irrigation. In addition, because of lower use of pesticides,

traditional agricultural practices also offer a great diversity and

abundance of arthropods (Bengtsson et al., 2005). In the Mediterranean

region, the species selects mosaic landscapes consisting of fruit crops and

orchards, mixed with patches of natural vegetation that provide food and

shelter (Lara-Romero et al., 2012). We found that in xeric shrubland,

badger showed a large preference for oranges, which directly depends

on the irrigation (Agencia Andaluza del Agua, 2010) and for figs, which

depends more directly on the precipitation. Therefore, a reduction of

these food resources by crop abandonment and less precipitation can

affect regional occurrence and abundance, including local extinction

where habitats can be extremely arid.

On the other hand, land cover and use changes influence the feeding

habits of European badger by modifying its feeding strategy and/or the

choice of preys (Kruuk, 1989). On a local scale, the composition of badger

diet depends on the land management and use (Fischer et al., 2005),


94

particularly in agricultural areas (Rosalino et al., 2004; Barea-Azcón et al.,

2010). Kruuk & Parish (1985) described how the feeding of badger varied

with changes in agricultural production in north Scotland. These authors

found that earthworm’s availability decreased during eight years and

badgers offset this reduction eating cereals in the later months. Despite

this, as a consequence they found a body weight reduction in males (i.e.,

14%) and females (17%). This finding suggests that due to extreme

conditions and low badger abundance in Mediterranean arid

environments, a reduction of key food resources may strongly affect

badger fitness (Virgós et al., 2005b).

In conclusion, food resources consumed by badgers depend on both

precipitation and land use-land cover. Therefore, variations predicted on

these factors under future climate change and land use scenarios (IPCC,

2013) may affect badger feeding habits across Mediterranean arid

environments, and be probably translated to changes in regional

distribution, abundance and other life-history traits.

RESULTADO 4.3 Modeling and monitoring habitat quality from space: the European badger

95

RESULTADO 4.3: MODELING AND MONITORING

HABITAT QUALITY FROM SPACE: THE EUROPEAN

BADGER

Tejón aseándose y desparasitándose en las cercanías de su tejonera. Escena

captada con fototrampeo en el paraje Venta de los Yesos. Marzo 2013. Tabernas.

Basado en:

Juan M. Requena-Mullor, Enrique López, Antonio J. Castro, Domingo

Alcaraz-Segura, Hermelindo Castro, Andrés Reyes, Javier Cabello.

Modeling and monitoring habitat quality from space: the European

badger. Journal of Applied Ecology, en revisión.


96

Objetivo 3: “Proyectar la distribución espacial del Tejón utilizando

escenarios de cambio climático futuros propuestos por el IPCC.”

Hipótesis 3: “Los cambios en los patrones de precipitación y temperatura

previstos por los modelos de circulación general de la atmósfera (IPCC,

2013), reducirán hasta en un 50% el rango de distribución de las especies

de la familia Mustelidae en la cuenca Mediterránea (Maiorano et al.,

2014). Por tanto, cabría esperar una reducción general de la calidad del

hábitat para el Tejón en el sureste árido de la Península Ibérica.”

Una característica básica de la ciencia experimental es la necesidad de obtener

conclusiones a partir de información incompleta.

M. Anthony Schork y Richard D. Remington.


97

Abstract

1. As climate change is expected to have a significant impact on species

distributions, there is an urgent need to enhance the efficiency of

biodiversity monitoring programs and provide managers with valuable

information to guide and adapt their actions.

2. For the European badger, a species not abundant and at risk of local

extinction due to climate change in the arid environments of

southeastern Spain, we identify which areas are prone to loss or gain of

habitat suitability as a result of climate change. Using MaxEnt, we

designed spatial distribution models for the badger using presence-only

data and climate and EVI-derived variables, and forecast the badger

potential spatial distribution for the 2071- 2099 period based on the IPCC

scenarios.

3. Including remotely sensed descriptors of the temporal dynamics and

spatial patterns of ecosystem functioning into spatial distribution

models, results suggest that high suitability areas for European badgers

may decrease in response to future climate scenarios. Primary

production and ecosystem functional heterogeneity seem to be these

main drivers of change.

4. Synthesis and applications. The incorporation of ecosystem functional

attributes derived from remote sensing in the modeling of future

forecast may contribute to the improvement of the detection of

ecological responses. Because conservation resources are limited, the

monitoring efforts for biodiversity should focus on areas which are more

likely to experience local extinctions or occurrence shifts and on the

involved environmental drivers of change. We suggest that our approach


98

can be applied in a variety of ecosystems worldwide and under diverse

climatic change scenarios, thereby supporting the design of

optimized/cost-efficient monitoring schemes and improving the ever-

increasing need for monitoring of biodiversity across space and time in a

rapidly changing planet.

Keywords: remote sensing, IPCC, biodiversity, drylands, mammals,

Enhanced Vegetation Index, ecosystem functioning, MaxEnt, future

forecasted niche


99

4.3.1 Introduction

As climate change is expected to have a significant impact on species

distributions, there is an urgent need to enhance the efficiency of

biodiversity monitoring programs and provide managers with valuable

information to guide and adapt their actions (Mawdsley et al., 2009). In

particular, mammalian species richness will be dramatically reduced

throughout the Mediterranean basin. However, the trend will not be

uniform for all taxa. For instance, Mustelidae (e.g., badger, weasel) will

decrease while Canidae (e.g., wolf), Hyaenidae (e.g., hyena) and some

families of Chiroptera (bats) will increase (Maiorano et al., 2011). These

findings highlight the complexity of species response to climate change

and the necessity of focusing monitoring efforts on areas where species

are likely to gain or lose suitable habitat to detect potential range shifts

driven by climate change (Amorim et al., 2014). An optimal tool for this

purpose is correlative models exploring the relationship between species

occurrences and environmental predictors (Araújo & Peterson, 2012).

Species distribution models (SDMs) aim to capture relationships between

a species (occurrence) and its environment. SDMs are often used to

forecast future distributions under environmental change scenarios. For

instance, bioclimate envelope models in their purest form only consider

climatic variables as predictors of species distribution. However, many

other environmental factors play an important role in determining

species distribution patterns and their dynamics over time. Such factors

may include from land cover, land use and management, soil types

(Pearson & Dawson, 2003) or ecosystem functioning descriptors such as

ecosystem production and seasonality (Requena-Mullor et al., 2014). As

a result, remotely sensed indicators of ecosystem functioning are


100

increasingly being used in animal research. In particular, spectral

Vegetation Indices (VIs) have been used to great success in mammal

ecology (Cabello et al., 2012a). VIs are conceptually and empirically

linked with primary production (Paruelo et al., 1999), which determines

the amount of green biomass available for herbivores and is referred as

the main descriptor of ecosystem functioning (Alcaraz-Segura et al.,

2006).

Functional attributes derived from VIs are usually expressed as average

temporal summaries, such as the annual mean (i.e., surrogate of mean

annual primary production) or the seasonal coefficient of variation (i.e.,

indicator of seasonality or temporal variation within the year) (Alcaraz-

Segura et al., 2013). Of particular note are spectral Vegetation Indices

(VIs), such as the Normalized Difference Vegetation Index (NDVI) and the

Enhanced Vegetation Index (EVI). Both of these VIs are directly related

with the fraction of photosynthetically active radiation (fAPAR)

intercepted by green vegetation (Ruimy et al., 1994). This relationship

allows the derivation of regional maps of primary production from

radiation use efficiency values (Castro et al., 2013). Landscape functional

heterogeneity has also been suggested as a significant driver of species

(Davidowitz & Rosenzweig, 1998) and ecosystem diversity (Alcaraz-

Segura et al., 2013), particularly in the Mediterranean Region. Many

animal species have proved to be especially sensitive to spatial

heterogeneity (Fryxell et al., 2005). Recent findings suggest that this

sensitivity is related more to functional heterogeneity than to structural

heterogeneity (Zaccarelli et al., 2013). For instance, Requena-Mullor et

al. (2014) found modeled spatial distribution of the European badger

(Meles meles) in SE Spain was significantly improved when augmenting


101

climate variables with EVI-derived functional attributes, (esp. the spatial

variability of EVI) rather than land-cover and land-use variables.

The purpose of this study is to explore the benefits of including remotely

sensed descriptors of the temporal dynamics and spatial patterns of

ecosystem functioning into spatial distribution models for suitability

monitoring of terrestrial mammal habitat. We used the European badger

in an arid region of southeastern Spain as a case study. In this region the

species is not abundant and at risk of local extinction due to climate

change (Virgós et al., 2005c). Our objective was to guide the species

monitoring by identifying which areas are prone to loss or gain of habitat

suitability as a result of climate change to inform managers about the

primary direct environmental drivers of change. To begin the modeling

process, we designed spatial distribution models for the badger using

presence-only data and climate and EVI-derived variables. Next, we

forecast the badger potential spatial distribution for the 2071- 2099

period based on the IPCC scenarios (IPCC, 2007), and identified sensible

areas that may lose habitat suitability based on the forecast. Finally, we

discuss how the findings of this study inform both the local monitoring of

this single species and are broadly applicable to global conservation

efforts for species monitoring and management.


102

4.3.2 Materials and methods

4.3.2.1 Species, study area and presence records

The European badger is a medium-sized carnivore widely distributed

across Europe. In the arid southeastern most limits of its range (i.e., the

Mediterranean drylands of Iberian Peninsula) the European badger

prefers mosaic landscapes consisting of fruit orchards and natural

vegetation, which provide shelter and food resources (Lara-Romero et

al., 2012). The potential effects that climate change on life-history traits

such as population density, social organization or population growth has

also been highlighted for this species (Macdonald et al., 2010). For these

reasons, European badger is an ideal study organism for the purpose of

this paper.

The study was conducted in the southeastern Iberian Peninsula (36°06’N,

2°17’E) (Fig. 4.3.1b). This region is the most arid in all of Europe and

presents the most extreme arid conditions in the specie range. We

defined “arid” using the Martonne aridity index (Ia), including values

between 5 and 15 (Martonne, 1926).

The presence records for the badgers were obtained from published data

(Requena-Mullor et al., 2014) and personal databases from the authors.

We reduced locally dense sampling by thinning the records to one per

100x100 m grid cell. A total of 179 presence records were used in the

modeling (Fig. 4.3.1b). The samples were distributed across a wide

gradient of altitude (0-1500 m), temperature (minimum mean

temperatures: -1.6–15 °C, maximum mean temperatures 17-24.5 °C),

precipitation (165-419 mm/year), and evapotranspiration (343-1038

mm/year).


103

4.3.2.2 Environmental variables

We selected six variables connected to ecological requirements of the

European badger and which have high predictive power in terms of

habitat suitability (Virgós & Casanovas, 1999; Requena-Mullor et al.,

2014). Variables were related to climate (mean annual precipitation

(PREC) and mean value of monthly maximum temperatures

(TMEDMAX)), relief (mean slope (SLOPE)), and spatio-temporal patterns

of primary production (EVI annual mean (EVIMEAN), intra-annual

coefficient of variation of EVI (EVIC), and spatial standard deviation of EVI

annual mean (EVISTD)). To avoid collinearity between predictors, we

checked that pairwise Pearson correlations were less than 0.85 (Booth,

Niccolucci & Schuster, 1994). The maximum Pearson correlation value

was 0.36, corresponding to EVICV and TMEDMAX.

PREC, TMEDMAX (for the 1971 to 2000 period) and SLOPE were derived

from spatial data layers of the Environmental Information Network of

Andalusia

(http://www.juntadeandalucia.es/medioambiente/site/web/rediam).

PREC and TMEDMAX had an intermediate cell size (100 x 100 m), so all

remaining variables were adjusted to this spatial resolution in QGIS 2.0 to

fit that grid following Elith & Leathwick (2009). SLOPE was calculated

from a 20 x 20 m pixel digital elevation model of Andalusia. We

resampled to the 100 x 100 m grid using bilinear resampling, which is

more realistic than nearest-neighbor interpolation (Phillips et al., 2006).

Our three functional descriptors of the spatiotemporal patterns of

primary production were derived from satellite images captured by the

MODIS sensor onboard the NASA TERRA satellite


104

(www.modis.gsfc.nasa.gov/). We used the MOD13Q1 EVI product, which

consists of 16-day maximum value composite images (23 per year) of the

EVI at a 231x231 m pixel size. This product has atmospheric, radiometric

and geometric corrections. We used EVI instead of NDVI because it is less

influenced by soil background and saturation problems at high biomass

levels (Huete et al., 2002). We first used the Quality Assessment (QA

band) information of this product to filter out those values affected by

high content of aerosols, clouds, shadows, snow or water. Next, we

calculated the mean seasonal EVI profile (average year) for the 2001-

2013 period and derived the EVI annual mean (EVIMEAN) as the mean of

the 23 images of the average year and the intra-annual coefficient of

variation of EVI (EVICV) as the intra-annual standard deviation divided by

EVIMEAN. The spatial standard deviation of EVIMEAN (EVISTD) was

calculated in windows of 3x3 km (13x13 MOD13Q1 pixels) throughout

the study area. The size of this window was determined based on the

suggested 9 km2 home range of the European badger for low suitability

habitats (Lara-Romero et al., 2012). The three EVI variables were

resampled to the 100 x 100 m grid by the bilinear resampling technique.

Horticultural greenhouses are intensively used in this area (Quintas-

Soriano et al., 2014), and because the EVI values of greenhouses cannot

be interpreted as vegetation greenness (Huete et al., 2002), we removed

all grid cells containing greenhouses (5% of the study area) to avoid their

influence of species distribution modeling.

4.3.2.3 Modeling approach

We designed SDMs for the badger based on the current environmental

conditions and evaluated these models and the relative importance of

the variables. Using the most parsimonious model to forecast the spatial


105

distribution under future scenarios we compared the current and

forecasted distributions and identified potential areas that may lose

habitat suitability and explored potential limiting factors of habitat

suitability.

4.3.2.4 Spatial distribution modeling

We used MaxEnt, v. 3.3.3k to build our models (Phillips et al., 2006).

Using the principle of maximum entropy on presence-only data, MaxEnt

estimates a set of functions that relate environmental variables and

habitat suitability and approximates the species niche and potential

spatial distribution. The MaxEnt algorithm has been frequently used in

recent years to forecast species spatial distributions and estimate shifts

in distribution due to climate change (Bateman et al., 2012).

While MaxEnt performs well both for spatial and temporal projections,

recent studies have demonstrated the sensitivity of SDM performance to

model specification (Muscarella et al., 2014). Thus, it is important to

implement species-specific tuning of settings and to use different training

and testing datasets for model evaluation.

In this sense, to obtain response curves more ecologically realistic and

more general predictions, we parameterized MaxEnt removing threshold

and hinge features (see Elith et al. (2011) for explanation of MaxEnt

features), and increased the regularization parameter β from 1 to10 (Hill

et al., 2014). β parameter regulates the smoothness and regularity of the

models. Therefore, by increasing the β parameter, such models are less

likely to overfit, because they have fewer parameters (Phillips et al.,

2006). Additionally, Warren & Seifert (2011) have highlighted that

inappropriately complex or inappropriately simple models show reduced


106

transferability to other time periods. Consequently, we used corrected

Akaike information criterion (AICc) (Burnham & Anderson, 2002) to

examine effects of changes in β as implemented in the software

ENMTools (Warren et al., 2010). This measure evaluates the information

loss, when a given model should describe reality and it can be

interpreted as a trade-off between model performance and complexity.

The model built with the value of β equal to 5 gained the lowest score for

the AICc, therefore it has been set to build the final model used to

forecast the spatial distribution (see “Model evaluation” subsection).

Likewise, in cases where a model is transferred to a new time period

which holds conditions more extreme than those available in model

training, the species response curve is said to be truncated. Accordingly,

current distributions were forecasted into the future scenarios by

“campling” the species response at that of the most-similar conditions in

the study area. In this way, if a variable projected to future conditions

reaches values greater than the maximum of the corresponding variable

used during training, those values are reduced to the maximum, and

similarly for values below the corresponding minimum (Phillips et al.,

2006).

To deal with the sample bias (i.e., some sites are more likely to be

surveyed than others), we upweighted records with few neighbors in

geographic space using a bias grid in MaxEnt (Elith et al., 2010) (see

Appendix I).

4.3.2.5 Model evaluation and variable relative importance

To select the most parsimonious model to forecast into future scenarios

we choose the lowest AICc and checked the relative importance of


107

variables. Models were run across 10 cross-validation replicates and their

AICc values were estimated. In each replicate, we partitioned presence

data by randomly selecting 80% of occurrence localities as training data

and the remaining 20% as test data. Models selected by AICc more

accurately estimated the habitat suitability when such models are

transferred to a different time period (Warren & Seifert, 2011).

We used a jackknife test to evaluate the relative importance of each

variable on both the regularized training gain (using training data) and

regularized test gain (using test data) of the models. It is important to

highlight that this gain is referred to likelihood of the models and not to

the increase of habitat suitability. We then estimated the regularized

model gain by (1) creating a model using each variable alone, (2)

removing the corresponding variable, and then creating a model with the

remaining variables, and (3) using all variables (Phillips et al., 2006).

4.3.2.6 Future forecasting

We forecast the badger spatial distribution for future climate scenarios

by using the best model based on AICc developed for the current climate

conditions. Two scenarios proposed by the Inter-governmental Panel of

Climate Change were considered: A2 and B1 (IPCC, 2007). The A2

scenario assumes a continuously increasing global population, economic

development that is primarily regionally oriented and focused on

economic growth and technological changes that are more fragmented

and slower than in other storylines. The B1 scenario assumes that

economic structures rapidly change towards a service and information

economy, and that resource-efficient technologies are introduced. The

projected PREC and TMEDMAX variables were obtained from the spatial


108

downscaling of these IPCC scenarios available at the Environmental

Information Network of Andalusia (see “Environmental variables”

subsection; Government of Andalusia, 2014). The SLOPE variable was

assumed to remain constant. The three projected functional descriptors

derived from EVI (i.e., EVIMEAN, EVICV and EVISTD) were predicted by

generalized additive models (GAMs). Precipitation and temperature are

the two major climate factors that govern the primary production of the

biosphere, although the response of primary production to these climate

drivers can vary both spatially and temporality (Cabello et al., 2012a). At

the regional scale, precipitation is the main climatic constrain of vegetal

growth in the Mediterranean climate (Nemani et al., 2003). However, the

seasonality of primary production remains modulated by both

precipitation and temperature (Schloss et al., 1999). Therefore, projected

EVIMEAN was predicted by a GAM that used the projected PREC variable

of the corresponding scenario as an explanatory variable. The EVICV

variable was predicted in the same manner but used projected PREC and

TMEDMAX as explanatory variables. Finally, the projected EVISTD

variable was obtained by using a 3x3 km moving window on the

projected EVIMEAN variable (as explained in “Environmental variables”

subsection) (see Appendix J).

4.3.2.7 Comparing current and future spatial distributions:

identification of sensitive areas to loss habitat suitability

To identify sensitive areas prone to losing habitat suitability for badgers,

we identified areas that showed a significant decrease of habitat

suitability between current and projected future conditions by applying

the methodology of Januchowski et al. (2010), using the SigDiff function

in the R package SDMTools. SigDiff is a “local” metric that identifies


109

where spatial distributions significantly differed. This metric represents

the significance of the pairwise differences relative to the mean and

variance of all differences between the two predicted spatial

distributions. The probability values predicted by this function represent

the area under the curve of a Gaussian distribution defined by the mean

and variance across all cells. These probability values were reclassified to

indicate areas where habitat suitability significantly decreased (SD ≥

0.975), where habitat suitability significantly increased (SD ≤ 0.025) and

where there was no significant difference between distributions

(Bateman et al., 2012). In addition, we also computed a “global” metric,

the I similarity statistic (Warren et al., 2008). This metric sums the pair-

wise differences between two predicted distributions to create a single

value representing the similarity of the two distributions. The I similarity

statistic ranges from a value of 0, where two distributions have no

overlap, to 1, where they are identical. It was computed with the Istat

function in the R package SDMTools.

4.3.2.8 Limiting factors of habitat suitability

We calculated the most limiting factor in each cell of the study area, the

value of each variable was changed to the mean value of that variable

over the species occurrences, and then, the resulting suitability value was

recorded. Thus, the limiting factor at that cell was the variable for which

the change results in the largest suitability value (Elith et al., 2010).


110

4.3.3 Results

4.3.3.1 Model performance and current projected

distribution

Comparing the ten MaxEnt models, model 9 (hereafter M9) was the most

plausible (i.e., ∆AICc < 2) (Burnham & Anderson, 2002) (Table 4.3.1) and

was selected to forecast the badger spatial distribution into future

scenarios.

Table 4.3.1. MaxEnt models performance for the European badger in SE Spain

under current climate conditions. For each model, the training data (80% of the

total) used were different. The table shows maximised log-likelihood function

(log(L)), number of estimated parameters (K), corrected Akaike Information

Criterion values (AICc), AICc differences from M9 (∆AICc) and Akaike weights

(Wi); *most parsimonious model.

Model Log (L) K AICc ∆AICc Wi

1 -2312.33 15 4657.63 8.62 0.01

2 -2312.80 13 4653.81 4.81 0.04

3 -2310.58 14 4651.73 2.72 0.13

4 -2311.42 13 4651.06 2.05 0.18

5 -2312.92 13 4654.07 5.06 0.04

6 -2313.00 14 4656.57 7.56 0.01

7 -2314.75 13 4657.73 8.72 0.01

8 -2312.15 13 4652.52 3.51 0.08

9* -2311.56 12 4649.01 0.00 0.49

10 -2314.81 12 4655.51 6.50 0.02

Current spatial distribution of European badger predicted for the M9

model is shown in Fig. 4.3.1a. Habitat suitability in the region ranged

from 0.008 to 0.96, with an average of 0.401 ± 0.135 SD. EVISTD was the


111

most important variable based on the jackknife test using the regularized

training gain; EVISTD caused the highest gain when used alone in the

model, and the greatest decrease in gain when removed from the model

(Table 4.3.2). However, using the regularized test gain, EVISTD also

achieved the highest gain when used alone, but PREC caused the greatest

decrease in gain when removed from the model.

Table 4.3.2. Most important environmental variables (*) in the MaxEnt model

for habitat suitability of the European badger in SE arid Spain under current

climate conditions (1971-2000). Relative importance of variables was evaluated

by a jackknife test on the training and test gains. The gains obtained using all

variables were 0.227 for training data and 0.397 for test data, so these were the

reference values. (see “Environmental variables” subsection for variables

abbreviations).

Training gain Test gain

Variable With only Without With only Without

PREC 0.042 0.193 0.007 0.276*

TMEDMAX 0.075 0.211 -0.051 0.435

SLOPE 0.028 0.192 0.001 0.334

EVIMEAN 0.038 0.206 0.161 0.285

EVICV 0.006 0.205 0.057 0.348

EVISTD 0.111* 0.170* 0.204* 0.30

The correlations between habitat suitability in the presence records and

each environmental variable under current conditions suggest which

environmental factors influence the presence of European badger. Thus,

EVISTD and TMEDMAX had the two highest positive values (Table 4.3.3),

while PREC and EVICV reached the two highest negative values.


112

Table 4.3.3. Pearson coefficient correlation (rho) between habitat suitability

predicted in the presence records under current climate conditions (1971-2000),

and each environmental variable. (*P < 0.05; **P < 0.001). (see “Environmental

variables” subsection for variables abbreviations).

Variable rho

PREC -0.49**

TMEDMAX 0.56**

SLOPE -0.18*

EVIMEAN 0.15*

EVICV -0.25**

EVISTD 0.73**

4.3.3.2 Forecasted future distributions

MaxEnt model M9 forecasted a decrease in high habitat suitability sites

for the European badger in the arid region of SE Spain under both A2 and

B1 future climate scenarios (Fig. 4.3.1a). The sites with high habitat

suitability (> 0.67) completely disappeared under both scenarios

(suitability ranged from 0.009 to 0.67 (A2) and 0.62 (B1)), changing the

skewness of the habitat suitability values from 0.40 (current) to -0.64

(A2) and -0.54 (B1) (skewness was estimated using the type 3 method in

Joanes & Gill, 1998). However, the average of habitat suitability for the

study area were similar (current conditions: 0.40 ± 0.13 SD; A2: 0.43 ±

0.11 SD and B1: 0.39 ± 0.11 SD). Therefore, the I similarity statistic

showed a low difference between current and forecasted future

distributions (0.93 and 0.92 for A2 and B1 respectively), which means a

high similitude between spatial distributions at regional scale.


113

Figure 4.3.1. (a) MaxEnt-modeled decrease in habitat suitability of the European

badger in SE arid Spain from current climate conditions (1971-2000) to two

future climate scenarios (IPCC A2 and B1 for 2071-2099). Habitat suitability

maps with mean suitability computed by rows and columns (cell size 100 x 100

m) in the margins. X and Y axes show UTM coordinates (Zone 30, Datum

ED1950). Cells in grey contain greenhouses, so they were removed before

computing the predictions (see “Environmental variables” subsection).

Histograms (Y axis: number of cells/number total of cells) of the habitat

suitability values. (b) Study area (7051km2) and location of the 179 badger


114

presence records used in this study. The area only includes arid climate; based

on Martonne aridity index. (c) MaxEnt-modeled maps of significant differences

in habitat suitability between current and predicted climate conditions under

the A2 and B1scenarios for the European badger in SE arid Spain. In pale red,

areas where the variables are expected to significantly decrease (SD > 0.975); in

blue, areas where the variables are expected to significantly increase (SD <

0.025); and in grey, areas where there was no significant difference.

4.3.3.3 Potential areas to lose habitat suitability and

involved environmental drivers

The MaxEnt model M9 revealed more loss than gain of suitable habitats

for the European badger in SE Spain under both A2 and B1 climate

change scenarios. Habitat suitability is expected to significantly decrease

in both scenarios: A2 by 3.85% and B1 by 4.04 % of the study area. In

both scenarios, significant decreases occurred along the two main river

valleys and along the eastern coastline (Fig. 4.3.1c). In contrast, some

areas are also expected to significantly gain habitat suitability in both

scenarios (though with lower extension than losses): A2 by 1.32% and B1

by 1.15%. In both scenarios, significant gains occurred in the central and

northwestern parts of the study area.

According to MaxEnt model M9 under current climate conditions, the

main limiting factor of habitat suitability for the European badger in SE

Spain strongly varied across the region (Fig. 4.3.2) Slope was important in

mountains, as well as precipitation (PREC) and temperature (TMEDMAX),

particularly in inner mountains. The EVI descriptors of temporal and

spatial patterns of ecosystem functioning (i.e., EVIMEAN, EVICV and

EVISTD) were more limiting at low-altitude foothills, plains and valleys.

Conversely, under both future climate scenarios, the spatial standard

deviation of EVIMEAN (EVISTD) was the main limiting factor throughout


115

the region (Fig. 4.3.2). In addition, slope was important in the highest

altitudes, EVIMEAN and EVICV in the northwestern foothills and plateaus

(particularly in A2), and precipitation (PREC) in the northern and central

western foothills (only in B1).

Figure 4.3.2. MaxEnt-modeled maps of the limiting factors of habitat suitability

for the European badger in SE arid Spain under current climate conditions (1971-

2000) and two future climate scenarios (IPCC A2 and B1 for 2071-2099). The

limiting factor is the environmental variable whose value at one cell most

influences the model suitability prediction. (see “Environmental variables”

subsection for variables abbreviations).

The areas that are expected to experience significant changes in habitat

suitability (Fig. 4.3.1c) tended to occur more in areas that experience

significant changes in the EVI descriptors of temporal and spatial patterns

of ecosystem functioning than in areas that only experience changes in

the climate variables (Fig. 4.3.3). Areas where EVIMEAN and EVISTD

significantly decreased (main river valleys, Fig. 4.3.3) were also those

where habitat suitability significantly decreased (Fig. 4.3.1c), which

agrees with their positive correlation (Table 4.3.3) and areas where

EVICV significantly decreased (inner plateaus, Fig. 4.3.3) were mainly

those where habitat suitability significantly increased (Fig. 4.3.1c), which

agrees with their negative correlation (Table 4.3.3).


116


117

Figure 4.3.3. Maps of the significant differences in the EVI descriptors of

ecosystem functioning and in the climate variables between current climate

conditions (1971-2000) and two future climate scenarios (IPCC A2 and B1 for

2071-2099) in SE Spain. In pale red, areas where the variables are expected to

significantly decrease (SD > 0.975); in blue, areas where the variables are

expected to significantly increase (SD < 0.025); and in grey, areas where there

was no significant difference. (see “Environmental variables” subsection for

variables abbreviations).

4.3.4 Discussion

4.3.4.1 EVI descriptors of ecosystem functioning to forecast

species distributions

Our results suggest that high suitability areas for European badgers may

decrease in response to future climate scenarios. Because conservation

resources are limited, the monitoring efforts for badgers should focus on

areas which are more likely to experience local extinctions or occurrence

shifts and on the involved environmental drivers of change. Primary

production (EVIMEAN) and ecosystem functional heterogeneity (EVISTD)

seem to be these main drivers of change. Macdonald et al. (2010) found

that badger life history parameters (such as survival, fecundity or body-

weight) are correlated with annual variability of both temperature and

rainfall mediated by food supply. Therefore, climate trends might

influence badgers population growth directly and through interactions

with food availability (Nouvellet et al., 2013).

The results also show that the local loss of the best suitable habitats for

badgers occurs in the same sites where primary production (represented

by EVIMEAN) is expected to significantly decrease (mainly along river

valleys). This green biomass is at the base of food webs, and therefore a

decrease would translate into lower food availability for the rest of

levels. The forecasted local loss of the best suitable habitats may be


118

driven by lower food availability under a new scenario of increasing

aridity. This will likely translate to changes in regional distribution,

abundance and life-history traits (such as fitness or body-weight).

In addition, the local loss of the best suitable habitats for badgers occurs

in the same sites where the spatial heterogeneity of primary production

(represented by EVISTD) is expected to significantly decrease under

future climate scenarios. At the regional scale, they also display that

EVISTD is expected to become the main limiting factor of habitat

suitability throughout the study area. In the arid Mediterranean climate,

badgers prefer to live in rural landscapes consisting in a mosaic of fruit

crops and orchards mixed with patches of semi-natural vegetation,

where the heterogeneity provides diverse food resources and shelters

(Lara-Romero et al. 2012). Several authors have suggested that this

preference is a response to food shortage (Lara-Romero et al., 2012;

Requena-Mullor et al., 2014). Mosaic landscapes represent a predictable,

clumped and profitable habitat to maximize the regional stability of food

supply such as fruits, insects or vertebrates in an arid region. In this way,

badgers would offset the lower availability of key prey items (i.e.,

earthworms) in arid regions compared to more rainy regions within its

distribution range (Kruuk, 1978). The forecasted landscape

homogenization in terms of ecosystem functioning would imply a

decrease in habitat suitability.

Conversely, our study also revealed some new spots that would reach

moderately high habitat suitability (from 0.5 to 0.7) under future climate

change scenarios. In such areas, the intra-annual variability of primary

production (seasonal coefficient of variation of EVI; EVICV) is expected to

significantly decrease. EVICV has been used as an indicator of carbon


119

gains seasonality that describes the variability of green biomass

production across seasons (Alcaraz-Segura et al., 2009). At the European

scale, badger densities are known to be limited by seasonal temperature,

or some other constraint(s) that covary with seasonality (Johnson et al.,

2002). At the local scale, it is also known that badger survival is especially

sensitive to seasonal weather extremes, such as extended summer

droughts and winter frosts (Macdonald & Newman 2002). For instance,

Woodroffe & Macdonald (2000) found that summer rainfall is a

significant predictor of cub survivorship. In SE Spain, seasonality is

expected to increase due to increasing temperatures and decreasing

rainfall, particularly during summer months (De Luis et al., 2001). Greater

seasonality would lead to lower habitat quality for badgers, monitoring

and management actions should pay special attention to such critical

periods that increase seasonality of primary production. However, for

monitoring purposes, it is important to bear in mind that changes in

primary production may be time-lagged with regard to changes in

precipitation or temperature (Cabello et al., 2012b). For example, in the

study area, Cabello et al. (2012b) observed from 2000 to 2010 that slight

precipitation increases during late summer led to relatively much higher

EVI increases in late autumn, which persisted throughout the winter

(critical periods in badger population dynamics, Macdonald et al., 2010).

Our results highlight seasonality of primary production, but not

seasonality of precipitation or temperatures, to better determine habitat

suitability for badgers, so focusing solely on weather extremes is not

enough for an effective monitoring of badgers populations in the study

area.


120

4.3.4.2 Global and local implications for wildlife monitoring

and management

One of the strategies suggested to adapt species monitoring and

management to climate change is evaluating and enhancing monitoring

programs for wildlife on a global scale (Mawdsley et al., 2009). However,

the costs to adapt existing monitoring systems are likely to be high.

Therefore new tools and approaches are required. Here we show that

cost-efficient monitoring schemes using globally available data sets (i.e.,

variables related to ecosystem functional attributes derived from remote

sensing) are sensitive to both long-term and rapid environmental

changes affecting the distribution range of a species (Crabtree et al.,

2009).

At the local scale of our study, the arid Mediterranean rural landscape is

a shifting mosaic that benefits carnivore diversity and abundance (Pita et

al., 2009). In this challenging environment, monitoring of future

landscape trajectories is vital for biodiversity conservation and

maintaining connectivity between suitable habitat areas (Piquer-

Rodríguez et al., 2012). Our results suggest to local managers that EVI-

derived functional attributes are useful to detect priority areas for

monitoring based on their loss or gain habitat suitability for badgers. For

instance, the areas where EVIMEAN significantly decreased were also

those where habitat suitability significantly decreased, which suggest

monitoring the spatial and temporal variability of EVI functional

attributes detects changes in habitat suitability. Likewise, seasonal

dynamic of primary production (represented by EVICV) could allow to

identify critical periods where focus the monitoring actions. These results

enhance previous findings by Requena-Mullor et al. (2014), which


121

showed that the distribution of European badgers in this region was

predicted by functional attributes such as high primary production

(EVIMEAN), seasonality in the primary production (EVICV) and spatially

heterogeneous landscapes (EVISTD). However, advances in the

knowledge of the relationship between these functional attributes and

food resources availability are still necessary for effective monitoring of

the European badger.

In conclusion, the incorporation of ecosystem functional attributes

derived from remote sensing in the modeling of future forecast may

contribute to the improvement of the detection of ecological responses.

We believe that our approach can be applied in a variety of ecosystems

worldwide and under diverse climatic change scenarios, thereby

supporting the design of optimized/cost-efficient monitoring schemes.

The use of predictive models, such as ours, can address and improve the

ever-increasing need for monitoring of biodiversity across space and time

in a rapidly changing planet.

DISCUSIÓN

122

5. DISCUSIÓN

El Tejón europeo se enfrenta a una serie de amenazas que pueden poner

en jaque la viabilidad de sus poblaciones, o en última instancia, su

presencia en los ecosistemas Mediterráneos a finales del siglo XXI. Esta

situación se agrava en algunos paisajes áridos Mediterráneos, donde la

especie es poco abundante y con distribución parcheada, lo cual hace su

supervivencia aún más complicada. Entre las amenazas más destacadas

se encuentran el cambio climático, consecuencia del calentamiento

atmosférico global, y la pérdida-fragmentación de su hábitat, ocasionada

por los cambios en el uso y la cobertura del suelo (Virgós et al., 2005c).

Varios trabajos han destacado los efectos potenciales que las variaciones

en los ciclos de precipitación y temperatura tienen sobre la dinámica

poblacional del Tejón (Macdonald et al., 2010; Nouvellet et al., 2013). Sin

embargo, la respuesta de la especie puede variar de unas regiones a

otras, por lo que se requieren estudios en diferentes áreas de su rango

de distribución (Virgós et al., 2005c). Por otro lado, el ser humano está

modificando la composición del paisaje y sustituyendo las prácticas

agrícolas tradicionales por otras más intensivas y menos respetuosas con

el entorno (Pita et al., 2009). Por este motivo, y dado que los paisajes

rurales Mediterráneos son el hábitat preferido del Tejón en ambientes

áridos, la calidad de su hábitat, y por tanto, la abundancia de la especie,

podría verse significativamente reducida (Lara-Romero et al., 2012).

En relación a esto, el último informe del Panel Intergubernamental

contra el Cambio Climático es claro: sin una gestión eficaz, derivada de

políticas de conservación que tengan en cuenta los efectos potenciales

del Cambio Global, algunas especies en riesgo menor en la actualidad,

pueden llegar a ser mucho más raras, e incluso desaparecer localmente

DISCUSIÓN

123

durante el siglo XXI (IPCC, 2013; Inger et al., 2015). La pérdida de

especies puede llevar a su vez a una reducción local de la biodiversidad y

a una modificación de la estructura y funcionamiento del ecosistema

(National Research Council, 1999). De forma particular, los meso-

carnívoros como el Tejón, juegan un papel relevante en los sistemas

naturales como predadores, competidores y especies “paragua”. De este

modo, variaciones en su abundancia o en la diversidad de sus

comunidades pueden inducir cambios a nivel ecosistema (Roemer et al.,

2009). Todo ello pone de manifiesto la necesidad de mejorar el

conocimiento sobre el rango actual de distribución del Tejón y los efectos

potenciales que sobre el mismo pudiera tener el Cambio Global. De esta

forma, los gestores pueden optimizar las medidas de conservación y

diseñar programas de seguimiento más eficientes, no solo de cara a la

conservación de la especie, sino también de los paisajes que habita

(Proyecto GLOCHARID, 2014).

Esta tesis doctoral supone un avance en esta línea de trabajo, y propone

algunas recomendaciones para el seguimiento de su distribución y

conservación del hábitat en un contexto Mediterráneo árido.

Uno de los principales impulsores de la distribución espacial del Tejón en

paisajes áridos Mediterráneos es la dinámica espacio-temporal de la

producción primaria (Requena-Mullor et al., 2014) (ver RESULTADO 4.1).

La producción primaria está situada en la base de la cadena trófica y

sintetiza aspectos funcionales clave de los ecosistemas, por lo que es

considerada un atributo integrador del funcionamiento ecosistémico

(McNaughton et al., 1989; Virginia & Wall, 2001). En un planeta que

cambia rápidamente, disponer de herramientas que informen casi a

tiempo real de la dinámica de la producción vegetal resulta de vital

DISCUSIÓN

124

importancia en tareas de seguimiento ecológico. En la era de los

satélites, la traslación de la información espectral a variables

relacionadas con el funcionamiento ecosistémico está ampliando el uso

tradicional de las imágenes de satélite en la biología de la conservación,

ofreciendo descriptores de procesos clave del funcionamiento de los

ecosistemas (ej., dinámica de la producción primaria, evapotranspiración,

eficiencia en el uso del agua por la vegetación, etc.) (Cabello et al., 2012).

Estos descriptores están resultando muy valiosos a la hora de modelizar

rasgos de la ecología de las especies y su respuesta frente a los cambios

ambientales (Cabello et al., 2012a). Un ejemplo de ello son los atributos

funcionales derivados de índices de vegetación, tales como el EVI. Estos

atributos son sensibles a factores que afectan al rango de distribución de

las especies (Crabtree et al., 2009), y han resultado de gran utilidad en la

modelización de la distribución espacial del Tejón europeo (Requena-

Mullor et al., 2014).

De cara al reto de monitorear los efectos que el Cambio Global pueda

tener sobre la distribución de la especie es indudable que resulta

determinante el uso de información satelital relacionada con la

producción primaria. Bajo la asunción de que la producción vegetal de un

área influencia la red trófica completa (McNaughton et al., 1989), los

atributos funcionales pueden ayudar a comprender (de abajo a arriba) las

alteraciones en la dinámica de las comunidades como respuesta a los

cambios ambientales. Frente al cambio climático, esta visión puede

resultar particularmente útil en ambientes Mediterráneos, debido a que

la precipitación es el principal factor limitante del crecimiento vegetal

(Nemani et al., 2003) (ver RESULTADO 4.3). La respuesta rápida de estos

atributos frente a cambios ambientales, su disponibilidad en continuo,

DISCUSIÓN

125

espacial y temporal a resoluciones óptimas regionales, hacen de ellos

una herramienta muy eficaz en el estudio y seguimiento de la

distribución de las especies.

Es importante resaltar que el uso de imágenes satelitales para derivar

atributos funcionales requiere de un filtrado previo de calidad de las

mismas (Reyes et al., 2015) y de una normalización-estandarización de

los algoritmos empleados en la estima de los mismos. Todo ello pone de

manifiesto la necesidad de protocolos internacionales consensuados que

permitan extender su uso a escalas más amplias y poder comparar los

resultados entre distintas áreas del planeta. De esta forma, avances en

distintos ámbitos de conocimiento y ecosistemas del planeta pueden

integrarse y servir de apoyo en la búsqueda de criterios generales sobre

los que apoyar Normas y Directrices internacionales en la lucha contra el

Cambio Global y la pérdida de biodiversidad.

Si analizamos el caso concreto del Tejón en el contexto árido

Mediterráneo, la dinámica espacio-temporal de la producción primaria

condiciona en gran medida su distribución espacial. Este resultado

(RESULTADO 4.1) hace pensar que existe algún tipo de correlación entre

los atributos funcionales que describen dicha distribución y la dinámica

espacio-temporal de los recursos alimenticios explotados por la especie.

El Tejón selecciona áreas con altos valores de producción vegetal y

heterogeneidad espacial, mientras que rechaza zonas con alta

variabilidad temporal en dicha producción. Estas condiciones suelen ser

representativas de paisajes rurales tradicionales bien conservados que

ofrecen al Tejón una oferta de recursos alimenticios variada

espacialmente y estable a lo largo del año (Requena-Mullor et al., 2014).

DISCUSIÓN

126

Sin embargo, poco se sabe de la relación entre la dinámica espacio-

temporal de los atributos funcionales del EVI y los recursos alimenticios

explotados por el Tejón. De entre los ítems consumidos por la especie en

ambientes Mediterráneos, son especialmente importantes los frutos,

insectos y vertebrados (Rodríguez & Delibes, 1992; Revilla & Palomares,

2002; Barea-Azcón et al., 2010, Requena-Mullor et al., 2015). Algunas de

las cuestiones a responder en relación a esto son: ¿hay una relación

directa positiva entre la producción primaria y la disponibilidad de

frutos?, ¿existe desfase temporal entre la fenología de la producción

vegetal y la disponibilidad de alimento?, y en caso afirmativo, ¿cuál es su

magnitud?, y ¿un paisaje heterogéneo, ofrece realmente una oferta más

variada de alimento?. En esta dirección, algunos autores han relacionado

la ocurrencia o abundancia de estos ítems con índices de vegetación tales

como el EVI o el Índice de Vegetación de Diferencia Normalizada (del

inglés, NDVI) (Willems et al., 2009; Lafage et al., 2013; Tapia et al., 2013),

pero poco más se sabe.

Para seguir avanzando en esta línea es imprescindible identificar primero

qué recursos alimenticios explota el Tejón, y si existe variabilidad en las

estrategias tróficas adoptadas entre paisajes dentro de un contexto árido

Mediterráneo. En este sentido, los resultados encontrados muestran

cómo los hábitos alimenticios del Tejón varían de unos paisajes a otros

dentro del contexto árido (ver RESULTADO 4.2). Aunque la frugivoría

sigue apareciendo como la principal estrategia adoptada por la especie

(Pigozzi, 1991; Rodríguez & Delibes, 1992; Barea-Azcón et al., 2010), en

zonas montañosas, donde la repoblación con pinos es el rasgo

dominante del paisaje, los invertebrados (ej., insectos y lombrices)

adquieren una importancia significativa. Este aumento en el consumo de

DISCUSIÓN

127

invertebrados compensaría la escasez de frutos cultivados y/o silvestres

de mayor contenido energético (ej., naranjas, aceitunas, higos,

algarrobas, etc.), los cuales son más abundantes en las vegas fluviales y

matorrales de la maquia mediterránea situados a menor altitud.

La dependencia directa y/o indirecta del Tejón hacia el manejo que el ser

humano hace del territorio se pone de manifiesto cuando se analizan sus

hábitos alimenticios. De esta forma, las prácticas agrícolas tradicionales

aportan a la especie no solo frutos cultivados (ej., naranjas), sino otros

como higos, nísperos, manzanas, albaricoques, etc., cultivados

secundariamente o en muchos casos naturalizados, pero dependientes

en cierto grado del regadío (Requena-Mullor et al., 2015). Además,

debido al escaso uso de pesticidas y maquinaria agrícola, esto ambientes

pueden ofrecer una gran diversidad y abundancia de artrópodos

(Bengtsson et al., 2005). Por todo ello, es importante tener en cuenta el

efecto potencial de las políticas agrarias sobre una especie como el

Tejón, en cuanto que pueden ser impulsoras de cambios en el uso del

suelo (Virgós et al., 2005c). En este sentido, la reforma de la Política

Agraria Común (PAC) para el período 2014-2020, contempla un paquete

de medidas que promueven la diversificación de cultivos, mantenimiento

de pastos permanentes y el destino de parte de la propiedad como

superficie de interés ecológico (Martínez & Palacios, 2012). Estas

prácticas, favorecerían a priori la conservación de las poblaciones de

Tejón en ambientes Mediterráneos (Virgós et al., 2005c), sin embargo, el

envejecimiento de la población rural y el éxodo de los más jóvenes a las

ciudades está propiciando un abandono generalizado de tierras en las

últimas décadas (Castro et al., 2011). Esto trae consigo un deterioro del

DISCUSIÓN

128

paisaje rural Mediterráneo, y por tanto, una amenaza para el futuro de la

especie.

Por otro lado, algunos ítems consumidos por el Tejón dependen

directamente de la precipitación, (ej., algarrobas, palmitos, moras,

lombrices), o indirectamente de la producción primaria (ej., insectos,

roedores, conejos). Ante un escenario de cambio climático futuro, la

disponibilidad de estos recursos, o la modificación de su fenología,

podría tener consecuencia en sus hábitos alimenticios, adaptándolos a la

nueva situación (Virgós et al., 2004). En casos de periodos extremos de

aridez, podría verse obligado a abandonar el territorio y buscar nuevas

zonas, lo que en condiciones de distribución parcheada, puede

incrementar el riesgo de desaparición local (Virgós et al., 2005c). En

relación a esto último, la calidad del hábitat para el Tejón en paisajes

áridos Mediterráneos podría verse significativamente reducida de aquí a

finales de siglo (ver RESULTADO 4.3). No obstante, esta reducción no

sería uniforme en el espacio, pudiendo incluso aumentar en algunas

zonas. Aquellas que ofrecen en la actualidad las mejores condiciones

para la supervivencia del Tejón (esto es, los paisajes agrícolas

tradicionales (Lara-Romero et al., 2012)), verían reducida su idoneidad de

hábitat debido principalmente a una homogeneización del paisaje desde

el punto de vista de la producción primaria. Por el contrario, algunas

áreas del centro y norte del área de estudio, aunque de menor extensión,

aumentarían su idoneidad gracias a una disminución en la estacionalidad

de la producción primaria. A pesar de que estos resultados hay que

tomarlos con precaución, debido a que no se tienen en cuenta de

manera directa los cambios en el uso del suelo (Pearson & Dawson,

2003) y la incertidumbre asociada a los modelos climáticos (Deser et al.,

DISCUSIÓN

129

2012), otros autores predicen igualmente una homogeneización del

paisaje derivada de la intensificación de los cultivos, y a un abandono del

uso agrícola tradicional ocasionado por el éxodo humano a áreas urbanas

(Silvestre, 2002; De Stefano, 2004; Piquer-Rodríguez et al., 2012). Por

tanto, estas previsiones dibujarían igualmente un escenario poco

alentador para la especie. De hecho, Maiorano et al. (2014) predicen, en

base a los mismos escenarios de cambio y período de tiempo que los

empleados en esta tesis doctoral, una disminución de hasta un 50% en la

distribución de las especies de la familia Mustelidae en la cuenca

Mediterránea. En un contexto global cambiante, el último informe del

IPCC para la conservación de los mamíferos predice que el 25% de las

especies de mamíferos del mundo, aproximadamente unas 1,125

especies, están en riesgo de extinción global (IPCC, 2013). Paralelamente,

Inger et al. (2015) alertan de que las especies que en la actualidad no

están amenazas corren serio peligro de estarlo en el futuro debido a que

toda la atención se centra en aquellas que sí lo están.

Desde esta perspectiva, resulta de vital importancia para la conservación

del Tejón en ambientes áridos Mediterráneos el seguimiento y

preservación de la calidad de su hábitat de cara a prever un potencial

deterioro y por tanto, un retroceso en su rango de distribución. Por ello,

son claves las iniciativas encaminadas al mantenimiento del paisaje rural

Mediterráneo. Las nuevas políticas en materia de conservación a nivel

nacional e internacional en los países Mediterráneos donde el Tejón está

presente deberían centrar esfuerzos en la protección y conservación de

dichos paisajes, e incluir los atributos funcionales derivados de sensores

remotos como herramienta para la caracterización y seguimiento del

estado de conservación (Proyecto GLOCHARID, 2014). Dichas políticas

DISCUSIÓN

130

deben contemplar acciones a llevar a cabo tanto dentro como fuera de

las áreas naturales protegidas debido a que una gran proporción de estos

paisajes se encuentran fuera de dichos límites (Cox & Underwood, 2011).

Una de las estrategias propuestas es el desarrollo de medidas

específicamente diseñadas para prevenir el abandono rural y preservar

las tierras agrícolas (Falcucci et al., 2007). Por un lado, el éxodo humano

de los entornos rurales a la ciudad debe de frenarse con programas que

estimulen las oportunidades de futuro, y por otro lado, medidas

orientadas a mantener la heterogeneidad paisajística y las prácticas

agrícolas tradicionales, ayudarían a preservar la biodiversidad asociada a

los agro-ecosistemas Mediterráneos. Por último, es imprescindible

aplicar de manera eficaz y rigurosa la normativa que rige la ordenación

del territorio y regula los cambios en el uso del suelo para evitar la

intensificación de los cultivos y la homogeneización del paisaje.

CONCLUSIONES

131

6. CONCLUSIONES

1. Los descriptores del funcionamiento ecosistémico, obtenidos a partir

de información satelital, representan una herramienta útil e innovadora

en la modelización de la distribución espacial de meso-carnívoros como

el Tejón europeo. Esto reafirma la necesidad de poner a disposición de

gestores y tomadores de decisiones herramientas que incorporen esta

información para elaborar e implementar programas de seguimiento

como parte de estrategias orientadas a la conservación de la

biodiversidad.

2. Los resultados obtenidos identifican las huertas tradicionales como

uno de los paisajes clave para la supervivencia del Tejón europeo en

paisajes áridos Mediterráneos. Dentro de un termoclima árido, esta tesis

sugiere que los cambios en la dinámica de la precipitación y/o la

modificación del paisaje pueden ejercer efectos sobre los hábitos

alimenticios del Tejón europeo, pudiendo en última estancia poner en

jaque la viabilidad de sus poblaciones a finales del siglo XXI.

3. El debate abierto asociado a que especies no amenazadas en la

actualidad podrían estarlo en el futuro pone de manifiesto la necesidad

de políticas de conservación que incorporen una visión de la

biodiversidad más general, sin particularizar tanto en especies

“emblemáticas”.

4. Finalmente, esta tesis representa un ejemplo de la aplicabilidad de las

técnicas de muestreo tradicional de mamíferos combinado con el uso de

información satelital en el estudio de procesos claves en la biología de la

especies.

REFERENCIAS

132

REFERENCIAS

Agencia Andaluza del Agua. (2010). Río Andarax. Agencia Andaluza del

Agua. Consejería de Medio Ambiente (Editores). Junta de Andalucía.

Sevilla.

Agresti, A. (1996). An Introduction to Categorical Data Analysis. Wiley,

New York.

Alcaraz-Segura, D. (2005). Caracterización mediante teledetección del

funcionamiento de los ecosistemas ibéricos. Bases para la conservación

de la biodiversidad en un escenario de cambio global. (Tesis doctoral

inédita). Universidad de Almería. Almería.

Alcaraz-Segura, D., Paruelo, J.M. & Cabello J. (2006). Identification of

current distribution of Ecosystem Functional Types in the Iberian

Peninsula. Global Ecology and Biogeography, 15(2):200–210.

Alcaraz-Segura, D., Cabello, J. & Paruelo, J. (2009). Baseline

characterization of major Iberian vegetation types based on the NDVI

dynamics. Plant Ecology, 202:13-29.

Alcaraz-Segura, D., Paruelo, J.M., Epstein, H.E. & Cabello, J. (2013).

Environmental and Human Controls of Ecosystem functional Diversity in

Temperate South America. Remote sensing, 5(1):127-154.

Amorin, F., Carvalho, S.B., Honrado, J. & Rebelo H. (2014). Designing

Optimized Multi-Species Monitoring Networks to Detect Range Shifts

Driven by Climate Change: A Case Study with Bats in the North of

Portugal. PLOS ONE, 9, e87291. doi:10.1371/journal.pone.0087291.

REFERENCIAS

133

Andalusia Land use/Land cover map (scale 1:25.000). (2007). Disponible

en http://www.juntadeandalucia.es/medioambiente/site/rediam.

Araújo, M.B. & Peterson, A.T. (2012). Uses and misuses of bioclimatic

envelope modeling. Concepts & Synthesis, 93:1527-1539.

Armas, C., Miranda, J. D., Padilla, F. M. & Pugnaire, F. I. (Editores). (2011).

Special issue: The Iberian Southeast. Journal of Arid Environments,

75:1241-1243.

Austin, M. (2007). Species distribution models and ecological theory: A

critical assessment and some possible new approaches. Ecological

Modelling, 200:1-19.

Bardsen, B.J. & Tveraa, T. (2012). Density-dependence vs. density-

independence - linking reproductive allocation to population abundance

and vegetation greenness. Journal of Animal Ecology, 81:364-376.

Barea-Azcón, J.M., Ballesteros-Duperón, E., Gil-Sánchez, J.M. & Virgós, E.

(2010). Badger Meles meles feeding ecology in dry Mediterranean

environments of the southwest edge of its distribution range.

ActaTheriologica. 55(1):45-52.

Bateman, B.L., VanDerWal, J., Williams, S.E. & Johnson, C.N. (2012).

Biotic interactions influence the projected distribution of a specialist

mammal under climate change. Diversity and Distributions, 18:861-872.

Bengtsson, J., Ahnström, J. & Weibull, A. (2005). The effects of organic

agriculture on biodiversity and abundance: a meta-analysis. Journal of

Applied Ecology, 42(2):261-269.

http://www.juntadeandalucia.es/medioambiente/site/rediam

REFERENCIAS

134

Best, D. J. & Roberts, D. E. (1975). The Upper Tail Probabilities of

Spearman's rho. Journal of the Royal Statistical Society. Series C (Applied

Statistic), 24:377–379.

Blanco, J. C. & González, J. L. (1992). Libro Rojo de los Vertebrados de

España. ICONA, Ministerio de Agricultura, Pesca y Alimentación, Madrid,

España, 714 p.

Booth, G.D., Niccolucci, M.J. & Schuster, E.G. (1994). Identifying proxy

sets in multiple linear regression: an aid to better coefficient

interpretation. Intermountain Research Station, USDA Forest Service,

Ogden, Utah, USA.

Boyce, M.S. (1978). Climatic variability and body size variation in the

muskrast (Ondatra zibethicus) of North America. Oecologia, 36:1-19.

Brown, J.H., Mehlman, D.W. & Steven, G.C. (1995). Spatial variance in

abundance. Ecology, 76:2028-2043.

Buermann, W., Saatchi, S., Smith, T.B., Zutta, B.R., Chaves, J.A., Milá, B. &

Graham, C.H. (2008). Prediction species distributions across the

Amazonian and Andean regions using remote sensing data. Journal of

Biogeography 35(7):1160-1176.

Burnham, K.P. & Anderson, D.R. (2002). Model selection and multimodel

inference. A practical information-theoretic approach. Springer-Verlag.

New York.

Cabello, J., Fernández, N., Alcaraz-Segura, D., Oyonarte, C., Piñeiro, G.,

Altesor, A., Delibes, M. & Paruelo, J.M. (2012a). The ecosystem

functioning dimension in conservation: insights from remote sensing.

Biodiversity and Conservation, 21:3287-3305.

REFERENCIAS

135

Cabello, J., Alcaraz-Segura, D., Ferrero, R., Castro, A.J. & Liras, E. (2012b).

The role of vegetation and lithology in the spatial and inter-annual

response of EVI to climate in drylands of Southeastern Spain. Journal of

Arid Environment, 79:76-83.

Castro, A.J., Martín-López, B., García-Llorente, M., Aguilera, P.A., López,

E. & Cabello, J. (2011). Social preferences regarding the delivery of

ecosystem services in a semiarid Mediterranean region. Journal of Arid

Environment, 75(11):1201-1208.

Castro, A.J., Paruelo, J.M., Alcaraz-Segura, D., Cabello, J., Oyarzabal, M. &

López-Carrique, E. (2013). Missing gaps in the estimation of the carbon

gains service from Light Use Efficiency models. En: Alcaraz-Segura, D. &

Di Bella, C. (Editores). Earth Observation of Ecosystem Services, 105-124

p.

Chen, W., Samuelson, F.W., Gallas, B.D., Kang, L., Sahiner, B. & Petrick, N.

(2013). On the assessment of the added value of new predictive

biomarkers. BMC Medical Research Methodology, 13:98.

Ciampalini, B. & Lovari, S. (1985). Food habits and trophic niche overlap

of the badger (Meles meles L.) and the red fox (Vulpes vulpes L.) in a

Mediterranean coastal area. Z. Säugetier-Kunde, 50:226-234.

Clarke, K. R. (1993). Non-parametric multivariate analyses of changes in

community structure. Austustralian Journal of Ecology, 18:117-143.

Corbacho, C., Sánchez, J.M. & Costillo, E. (2003). Patterns of structural

complexity and human disturbance of riparian vegetation in agricultural

landscapes of a Mediterranean area. Agriculture, Ecosystems and

Environment, 95:495-507.

REFERENCIAS

136

Cox, R.L. & Underwood, E.C. (2011). The importance of Conservation

Biodiversity Outside of Protected Areas in Mediterranean Ecosystems.

PLOS ONE, 6(1):e14508.

Crabtree, R., Potter, C., Mullen, R., Sheldon, J., Huang, S., Harmsen, J.,

Rodman, A. & Jean, C. (2009). A modeling and spatio-temporal analysis

framework for monitoring environmental change using NPP as an

ecosystem indicator. Remote Sensing of Environment: 113:1486-1496.

Da Silva, J., Woodroffe, R. & Macdonald, D. W. (1993). Habitat, food

availability and group territoriality in the European badger, Meles meles.

Oecologia, 95:558-564.

Davidowitz, G. & Rosenzweig, M.L. (1998). The latitudinal gradient of

species diversity among North American grasshoppers within a single

habitat: a test of the spatial heterogeneity hypothesis. Journal of

Biogeography, 25:553-560.

De Stefano, D. (2004). Freshwater and tourism in the Mediterranean.

Rome, Italy: WWF Mediterranean Programme.

De Luis, M., García-Cano, M.F., Cortina, J., Raventós, J., González-Hidalgo,

J. C. & Sánchez, J.R. (2001). Climatic trends, disturbances and short-term

vegetation dynamics in a Mediterranean shrubland. Forest Ecology and

Management, 147(1):25-37.

Del Cerro, I., Marmi, A., Ferrando, P. & Chashchin P. (2010). Nuclear and

mitochondrial phylogenies provide evidence for four species of Eurasian

badgers (Carnivora). Zoologica Scripta, 39:415-425.

REFERENCIAS

137

DeLong, E.R., DeLong, D.M. & Clarke-Pearson, D.L. (1998). Comparing the

areas under two or more correlated receiver operating characteristic

curves: a non-parametric approach. Biometrics, 44:837-845.

Deser, C., Phillips, A., Bourdette, V. & Teng, H. (2012) Uncertainty in

climate change projections: the role of internal variability. Climate

Dynamics, 38(3-4):527-546.

Edwards, C. A. & Lofty, J. R. (1977). Biology of earthworms. Chapman and

Hall, London, U.K.

Elith, J., Graham, C.H., Anderson, R.P. Dudík, M., Ferrier, S., Guisan, A.,

Hijmans, R.J., Huettmann, F., Leathwick, J.R., Lehmann, A., Li, J.,

Lohmann, L.G., Loiselle, B.A., Manion, G., Moritz, C., Nakamura, M.,

Nakazawa, Y. & Overton, J.M., Townsend, A., Phillips, S.J., Richardson, K.,

Scachetti-Pereira, R., Schapire, R.E., Soberón, J., Williams, S., Wisz, M.S. &

Zimmermann N.E. (2006). Novel methods improve prediction of species´

distributions from occurrence data. Ecography, 29:129-151.

Elith, J. & Leathwick, J.R. (2009). Species Distribution Models: Ecological

Explanation and Prediction Across Space and Time. Annual Review of

Ecology, Evolution, and Systematics, 40:677-697.

Elith, J., Kearney, M. & Phillips, S. (2010). The art of modelling range-

shifting species. Methods in Ecology and Evolution, 1:330-342.

Elith, J., Phillips, S.J., Hastie, T., Dudík, M., Chee, Y.E. & Yates, C.J. (2011).

A statistical explanation of MaxEnt for ecologists. Diversity and

Distributions, 17:43-57.

REFERENCIAS

138

Falcucci, F., Maiorano, L. & Boitani, L. (2007). Changes in land-use/land-

cover patterns in Italy and their implications for biodiversity

conservation. Landscape Ecology, 22:617-631.

Ferguson, S.H. & McLoughlin, P.D. (2000). Effect of energy availability,

seasonality and geographic range on brown bear life history. Ecography,

23:193-200.

Feroe, S. & Montgomery, W.I. (1999). Habitat effects on the spatial

ecology of the European badger (Meles meles). Journal of Zoology,

247:537-549.

Ferrier, S. & Watson, G. (1997). An Evaluation of the effectiveness of

environmental surrogates and modeling techniques in predicting the

distribution of biological diversity. Environment Australia, Canberra

Australia. Disponible en:

http://www.environment.gov.au/archive/biodiversity/publications/techn

ical/surrogates/ .

Fischer, C., Ferrari, N. & Weber, J. M. (2005). Exploitation of food

resources by badgers (Meles meles) in the Swiss Jura Mountains. Journal

of Zoology of London, 266(2):121-131.

Fryxell, J.M., Wilmshurst, J.F., Sinclair, A.R.E., Haydon, D.T., Holt, R.D. &

Abrams, P.A. (2005). Landscape scale, heterogeneity, and the viability of

Serengeti grazers. Ecology Letters, 8:328-335.

García-Rangel, S. & Pettorelli, N. (2013). Thinking spatially: The

importance of geospatial techniques of carnivore conservation.

Ecological Informatics, 14:84-89.

http://www.environment.gov.au/archive/biodiversity/publications/technical/surrogates/

http://www.environment.gov.au/archive/biodiversity/publications/technical/surrogates/

REFERENCIAS

139

Giorgi, F. & Lionello, P. (2008). Climate change projections for the

Mediterranean region. Global Planet Change. 63(2-3):90-104.

Goszczynski, B., Jedrzejewska, B. & Jedrzejewski, W. (2000). Diet

composition of badgers (Meles meles) in a pristine forest and rural

habitats of Poland compared to other European populations. Journal of

Zoology of London, 250(4):495-505.

Graham, C.H. & Hijmans, R.J. (2006). A comparison of methods for

mapping species ranges and species richness. Global Ecology and

Biogeography, 15:578.

Hastie, T. & Tibshirani, R. (1990). Generalized Additive Models. Chapman

& Hall, London.

Hernández, P.A., Graham, C.H., Master, L.L. & Albert D.L. (2006). The

effect of sample size and species characteristics on performance of

different species distribution modeling methods. Ecography, 29:773-785.

Herrera, C.M. (1989). Frugivory and seed dispersal by carnivorous

mammals and associated fruit characteristics, in undisturbed

Mediterranean habitats. Oikos, 55:250-262.

Hill, M.P, Axford, J.K. & Hoffmann, A.A. (2014). Predicting the spread of

Aedes albopictus in Australia under current and future climates: Multiple

approaches and datasets to incorporate potential evolutionary

divergence. Austral Ecology, 39:469-478.

Huete, A.R., Liu, H.Q., Batchil. K. & van Leeuwen, W. (1997). A

comparison of vegetation indices global set of TM images for EOS-

MODIS. Remote Sensing of Environment, 59:440-451.

http://modis.gsfc.nasa.gov/sci_team/pubs/abstract.php?id=03629



REFERENCIAS

140

Huete, A.R., Didan, K., Miura, T., Rodriguez, E.P., Gao, X. & Ferreira, L.G.

(2002). Overview of the radiometric and biophysical performance of the

MODIS vegetation indices. Remote Sensing of Environment, 83:195-213.

Inger, R., Gregory, R., Duffy, J.P., Stott, Vorïsek, P. & Gaston, K.J. (2015).

Common European birds are declining rapidly while less abundant

species’ numbers are rising. Ecology Letters, 18:28-36.

Intergovernmental Panel on Climate Change (2007) Climate Change

2007: Synthesis Report. Contribution of Working Groups I, II and III to the

Fourth Assessment Report of the Intergovernmental Panel on Climate

Change. Core Writing Team, Pachauri, R.K & Reisinger, A. (Editores). IPCC,

Geneva, Switzerland, 104 p.

IPCC, (2007): Cambio climático 2007: Informe de síntesis. Contribución de

los Grupos de trabajo I, II y III al Cuarto Informe de evaluación del Grupo

Intergubernamental de Expertos sobre el Cambio Climático [Equipo de

redacción principal: Pachauri, R.K. & Reisinger, A. (directores de la

publicación)]. IPCC, Ginebra, Suiza, 104 p.

IPCC, (2013): Climate Change 2013: The Physical Science Basis.

Contribution of Working Group I to the Fifth Assessment Report of the

Intergovernmental Panel on Climate Change. Stocker T.F., Qin G.K.

Plattner M., Tignor S.K., Allen J., Boschung A., Nauels Y., Xia V., Bex P.M.

(Editores). Cambridge University Press, Cambridge, United Kingdom and

New York, NY, USA, 1535 p.

IPCC, (2013): Summary for Policymakers. En: Climate Change 2013: The

Physical Science Basis. Contribution of Working Group I to the Fifth

Assessment Report of the Intergovernmental Panel on Climate Change.

REFERENCIAS

141

Stocker, T.F., Qin, D., Plattner, G.K., Tignor, M., Allen, S.K., Boschung, J.,

Nauels, A., Xia, Y., Bex, V. & Midgley, P.M. (Editotes). Cambridge

University Press, Cambridge, United Kingdom and New York, NY, USA, 33

p.

Januchowski, S., Pressey, B., Vanderwal, J. & Edwards, A. (2010).

Characterizing errors in topographic models and estimating the financial

costs of accuracy. International Journal of Geographical Information

Science, 24:1327-1347.

Jepsen J.U., Madsen A.B., Karlsson M. & Groth D. (2005). Predicting

distribution and density of badger (Meles meles) setts in Denmark.

Biodiversity and Conservation, 14:3235-3253.

Joanes, D.N. & Gill, C.A. (1998). Comparing measures of sample skewness

and kurtosis. The Statistician, 47:183-189.

Johnson, D.D., Jetz, W. & Macdonald, D.W. 2002. Environmental

correlates of badger social spacing across Europe. Journal of

Biogeography, 29(3):411-425.

Junta de Andalucía. (2014). El clima de Andalucía en el siglo XXI.

Consejería de Medio Ambiente y Ordenación del Territorio de la Junta de

Andalucía. (Editores). Sevilla.

Kruskal, J.B. (1964). Multidimensional scaling by optimizing goodness of

fit to a nonmetric hypotheses. Psychometrika, 29(1):1-27.

Kruuk, H. (1978). Spatial organisation and territorial behavior of the

European badger, Meles meles. Journal of Zoology, 184:1-19.

REFERENCIAS

142

Kruuk H. (1989). The social badger: ecology and behaviour of a group-

living carnivore (Meles meles). Oxford University Press.

Kruuk, H. & de Kock, L. (1981). Food and habits of badgers (Meles meles)

on Monte Baldo, northern Italy. Z. Säugetierkunde, 46:295-301.

Kruuk H. & Parish T. (1981). Feeding specialization of the European

badger Meles meles in Scotland. Journal of Animal Ecology, 50:773-788.

Kruuk, H. & Parish, T. (1985). Food, food availability and weight of

badgers (Meles meles) in relation to agricultural changes. Journal of

Applied Ecology, 22:705-715.

Lafage, D., Secondi, J., Georges, A., Bouzillé, J.B. & Pétillon, J. (2013).

Satellite-derived vegetation indices as surrogate of species richness and

abundance of ground beetles in temperate floodplains. Insect

Conservation and Diversity, doi: 10.1111/icad.12056.

Lara-Romero, C., Virgós, E., Escribano-Ávila, G., Mangas, J.G., Barja, I. &

Pardavila, X. (2012). Habitat selection by European badgers in

Mediterranean semi-arid ecosystems. Journal of Arid Environment,

76:43-48.

Lavorel, S., Canadell, J., Rambal, S. & Terradas, J. (1998). Mediterranean

terrestrial ecosystems: research priorities on global change effects.

Global Ecology and Biogeography Letters, 7:157-166.

Levinsky, I., Skov, F., Svenning J.C. & Rahbek C. (2007). Potential impacts

of climate change on the distributions and diversity patterns of European

mammals. Biodiversity and Conservation, 16:3803-3816.

REFERENCIAS

143

Lloyd, D. (1990). A phenological classification of terrestrial vegetation

cover using shortwave vegetation index imagery. International Journal of

Remote Sensing, 11:2269-2279.

Macdonald, D.W. (1983). The ecology of carnivore social behaviour.

Nature, 301:379-384.

Macdonald, D.W. & Carr, G.M. (1999). Food security and the rewards of

tolerance. En: Comparative socioecology: the behavioural ecology of

humans and animals, vol 8. Standen, V. & Foley, R. (Editores). Blackwell

Scientific, 75-79 p.

Macdonald, D.W. & Newman, C. (2002) Population dynamics of badgers

(Meles meles) in Oxford shires, U.K.: numbers, density and cohort life

histories, and a possible role of climate change in population growth.

Journal of Zoology of London, 256(1):121-138.

Macdonald, D.W., Newman, C., Buesching, C.D. & Nouvellet, A. (2010).

Are badgers ‘Under The Weather’? Direct and indirect impacts of climate

variation on European badger (Meles meles) population dynamics. Global

Change Biology, 16:2913-2922.

Maiorano, L., Falcucci, A., Zimmermann, N.E., Psomas, A., Pottier, J.,

Baisero, D., Rondinini, C., Guisan, A. & Boitani, L. (2011). The future of

terrestrial mammals in the Mediterranean basin under climate change.

Philosophical Transactions of The Royal Society, 366:2681-2692.

McNaughton, S.J., Oesterheld, M., Frank, D.A. & Williams, K.J. (1989).

Ecosystem-level patterns of primary productivity and herbivory in

terrestrial habitats. Nature, 341:142-144.

REFERENCIAS

144

Martín, R., Rodríguez, A. & Delibes, M. (1995). Local feeding

specialization by badgers (Meles meles) in a Mediterranean environment.

Oecologia, 101:45-50.

Martínez, F., & Palacios, A.M. (2012). Guía práctica de la nueva PAC.

Editorial Agrícola Española, S.A., Madrid.

Martonne, E. (1926). Areisme et indice d’aridité. Geographical Review,

17:397-414.

Mawdsley, J.R., O’Malley, R. & Ojima, D.S. (2009). A Review of Climate-

Change Adaptation Strategies for Wildlife Management and Biodiversity

Conservation. Conservation Biology, 23:1080-1089.

Melis, C., Cagnacci, F. & Bargagli, L. (2002). Il tasso. Habitat, 122:44-52

Mellgren, R.L., & Roper, T.J. (1986). Spatial learning and discrimination of

food patches in the European badger (Meles meles L.). Animal Behaviour,

34(4):1129-1134.

Menke S.B., Holway R.N., Fisher R.N. & Jetz W. (2009). Characterizing and

predicting species distributions across environments and scales:

Argentine ant occurrences in the eye of the beholder. Global Ecology and


Meynard, C.N., Pillay, N., Perrigault, M., Caminade, P. & Ganem, G.

(2012). Evidence of environmental niche differentiation in the striped

mouse (Rhabdomys spp.): inference from its current distribution in

southern Africa. Ecology and Evolution, 2(5):1008-1023.

Monteith J.L. (1981). Evaporation and surface temperature. Royal

Meteorological Society, 107:1-27.

REFERENCIAS

145

Morueta-Holme, N., Flojgaard, C. & Svenning, J.C. (2010). Climate Change

Risks and Conservation Implications for a Threatened Small-Range

Mammal Species. PLOS ONE, 5(4):e10360.

Muirhead, R.H., Gallagher, J. & Burn, K.J. (1974). Tuberculosis in wild

badgers in Gloucestershire: Epidemiology. Veterinary Record, 95:552-

555.

Muscarella, R., Galante, P.J., Soley-Guardia, M., Boria, R.A. Kass, J.M.,

Uriarte, M. & Anderson, R.P. (2014). ENMeval: An R package for

conducting spatially independent evaluations and estimating optimal

model complexity for MAXENT ecological niche models. Methods in

Ecology and Evolution, doi: 10.1111/2041-210X.12261.

Myers, R.H. & Montgomery, D.C. (1997). A tutorial on generalized linear

models. Journal of Quality Technology, 29:274-291.

National Research Council. (1991). Policy Implications of Greenhouse

Warming. Report of the Committee on Science, Engineering and Public

Policy, National Research Council/National Academy of Sciences

(NRC/NAS), National Academy Press, Washington, DC, USA, 127 p.

Nemani, R.R., Keeling, C.D., Hashimoto, H., Jolly, W.M., Piper, S.C.,

Tucker, C.J., Myneni, R.B. & Running, S.W. (2003). Climate-Driven

Increases in Global Terrestrial Net Primary Production from 1982 to

1999. Science, 300:1560.

Newton-Cross, G., White, P.C. & Harris, S. (2007). Modelling the

distribution of badgers Meles meles: comparing predictions from field-

based and remotely derived habitat data. Mammal Review, 37(1):54-70.

REFERENCIAS

146

Nilsen, E.B., Herfindal, I. & Linnell, J.D. (2005). Can intra-specific variation

in carnivore home-range size be explained using remote-sensing

estimates of environmental productivity?. Ecoscience, 12(1):68-75.

Nouvellet, P., Newman, C., Buesching, C.D. & Macdonald, D.W. (2013). A

Multi-Metric Approach to Investigate the Effects of Weather Conditions

on the Demographic of a Terrestrial Mammal, the European Badger

(Meles meles). PLOS ONE, 8:e68116.

Oindo, B.O. (2002). Predicting Mammal Species Richness and Abundance

Using Multi-Temporal NDVI. Photogrammetric Engineering and Remote

Sensing, 68(6):623-629.

Olesen, J.E. & Bindi, M. (2002). Consequences of climatic change for

European agricultural productivity, land use and policy. European Journal

of Agronomy, 16(4):239-262.

Palomo, L.J., Gisbert, J. & Blanco, J.C. (2007). Atlas y Libro Rojo de los

Mamíferos Terrestres de España. Dirección General para la

Biodiversidad-SECEM-SECEMU, Madrid, 588 p.

Paruelo, J.M., Lauenroth, W.K., Burke, I.C. & Sala, O.E. (1999). Grassland

precipitation-use efficiency varies across a resource gradient.

Ecosystems, 2:64-68.

Paruelo, J.M., Piñeiro, G., Oyonarte, C., Alcaraz, D., Cabello, J. &

Escribano, P. (2005). Temporal and spatial patterns of ecosystem

functioning in protected arid areas of Southeastern Spain. Applied

Vegetation Science, 8:93-102.

Pearce, J.L., Cherry, K., Drielsma, M., Ferrier, S. & Whish, G. (2001).

Incorporating expert opinion and fine-scale vegetation mapping into

REFERENCIAS

147

statistical models of faunal distribution. Journal of Applied Ecology,

38:412-424.

Pearson, R. & Dawson, T. (2003). Predicting the impacts of climate

change on the distribution of species: are bioclimate envelope models

useful?. Global Ecology & Biogeography, 12:361-371.

Pettorelli, N., Vik, J.O., Mysterud, A., Gaillard, J.M., Tucker, C.J. &

Stensethet, N.C. (2005). Using the satellite-derived NDVI to assess

ecological responses to environmental change. Trends in Ecology and

Evolution, 20(9):503-510.

Pettorelli, N., Gaillard, J.M., Mysterud, A., Duncan, P., Stenseth, N.C.,

Delorme, D., Van Laere, G., Toïgo, C. & Klein, F. (2006). Using a proxy of

plant productivity (NDVI) to track animal performance: the case of roe

deer. Oikos, 112:565-572.

Pettorelli N., Ryan S., Mueller T., Bunnefeld, N., Jedrzejewsk, B. & Lima,

M. (2011). The Normalized Difference Vegetation Index (NDVI):

unforeseen successes in animal ecology. Climate Research, 46:15-27.

Phillips, S.J., Dudík, M. & Schapire R.E. (2004). A Maximun Entropy

Approach to Species Distribution Modeling. Twenty-First International

Conference on Machine Learning, 655-662 p.

Phillips, S.J. (2006) A brief tutorial on MaxEnt. AT & T Research.

Disponible en:

http://www.cs.princeton.edu/~schapire/maxent/tutorial/tutorial.doc.

Phillips S.J., Anderson, R.P. & Schapire, R.E. (2006). Maximum entropy

modeling of species geographic distribution. Ecological Modeling,

190:231-259.

REFERENCIAS

148

Phillips, S.J., Dudík, M., Elith, J., Graham, C.H., Lehmann, J.L. & Ferrier, S.

(2009). Sample selection bias and presence-only distribution models:

implications for background and pseudo-absence data. Ecological

Applications, 19(1):181-197.

Pigozzi, G. (1991). The diet of the European badger in a Mediterranean

coastal area. Acta Theriologica, 36(3-4):293-306.

Piquer-Rodríguez, M., Kuemmerle, T., Alcaraz-Segura, D., Zurita-Milla, R.

& Cabello, J. (2012). Future land use effects on the connectivity of

protected area networks in southeastern Spain. Journal of Nature

Conservation, 20:326-336.

Pita, R., Mira, A., Moreira, F., Morgado, R. & Beja, P. (2009). Influence of

landscape characteristics on carnivore diversity and abundance in

Mediterranean farmland. Agriculture, Ecosystems and Environment,

132(1-2):57-65.

Proyecto GLOCHARID (2014). Centro Andaluz para el Estudio y

Seguimiento del Cambio Global. Informes I, II, III & IV. Disponible en:

http://caescg.org.

QGIS Development Team, (2014). QGIS Geographic Information System.

Open Source Geospatial Foundation Project. Disponible en:

http://qgis.osgeo.org.

Quintas-Soriano, C., Castro, A.J., García-Llorente, M., Cabello, J. & Castro,

H. (2014). From supply to social demand: a landscape-scale analysis of

the water regulation service, Landscape Ecology, 29:1069-1082.

http://qgis.osgeo.org/

REFERENCIAS

149

R Core Team, (2014). R: A language and environment for statistical

computing. R Foundation for Statistical Computing, Vienna, Austria.

Disponible en: http://www.R-project.org/.

Requena-Mullor, J.M., López, E., Castro, A.J., Cabello, J., Virgós, E.,

González-Miras, E., Castro, H. (2014). Predicting spatial distribution of

European badger in arid landscapes: an ecosystem functioning approach.

Landscape Ecology, 29(5):843-855.

Requena-Mullor, J.M., López, E., Castro, A.J., Virgós, E., Castro, H. (2015).

Hábitos alimenticios del Tejón europeo en un paisaje árido

Mediterráneo. Galemys, aceptado con revisión menor.

Revilla, E., Palomares, F. & Delibes, M. (2000). Defining key habitats for

low density populations of Eurasian badgers in Mediterranean

environments. Biological Conservation, 95:269-277.

Revilla, E., Palomares, F. & Delibes, M. (2001b). Edge-core effects and the

effectiveness of traditional reserves in conservation: Eurasian badgers in

Doñana National Park. Conservation Biology, 15:148-158.

Revilla, E., Palomares, F. & Fernández, N. (2001a). Characteristics,

location and selection of diurnal resting dens by Eurasian badger (Meles

meles) in a low density area. Journal of Zoology of London, 255:291-299.

Revilla, E. & Palomares, F. (2002). Spatial organization, group living and

ecological correlates in low-density populations of Eurasian badgers,

Meles meles. Journal of Animal Ecology, 71:497-512.

Revilla, E., Casanovas, J.G. & Virgós E. (2002). Meles meles Linnaeus,

1758. Tejón. En: L. J. Palomo y J. Gisbert. (Editores.) Atlas de los

http://www.r-project.org/

REFERENCIAS

150

mamíferos terrestres de España. Dirección General de Conservación de la

Naturaleza, SECEM-SECEMU, Madrid, 274-277 p.

Reyes, A., Cabello, J. & Alcaraz-Segura, D. (2015). Implicaciones del

filtrado de calidad del índice de vegetación EVI para el seguimiento

funcional de ecosistemas. Revista de teledetección, en prensa.

Rodríguez, A. & Delibes, M. (1992) Food habits of badgers (Meles meles)

in an arid habitat. Journal of Zoology, 227:347-350.

Rodríguez, J.P., Brotons, L., Bustamante, J. & Seoane, J. (2007). The

application of predictive modelling of species distribution to biodiversity.

Diversity and Distributions, 13:243-251.

Roemer, G.W., Gompper, M.E. & Valkenburgh, B.V. (2009). The ecological

role of the mammalian mesocarnivore. BioScience, 59:165-173.

Roper, T.J. (1994). The European badger Meles meles: food specialist or

generalist?. Journal of Zoology of London, 234(3):437-452.

Rosalino, L.M., Macdonald, D.W. & Santos-Reis, M. (2004). Spatial

structure and land cover use in a low density Mediterranean population

of Eurasian badgers. Canadian Journal of Zoology, 82(9):1493-1502.

Rosalino, L.M., Loureiro, F., Macdonald, D.W. & Santos-Reis, M. (2005).

Dietary shifts of the badger (Meles meles) in Mediterranean woodlands:

an opportunistic forager with seasonal specialisms. Mammal Biology,

70(1):12-23.

Rosalino, L.M., Santos, M.J., Beiber, P. & Santos-Reis, M. (2008) Eurasian

badger habitat selection in Mediterranean environments: Does scale

really matter?. Mammal Biology, 73:189-198.

REFERENCIAS

151

Rosalino, L.M., & Santos-Reis, M. (2009). Fruit consumption by carnivores

in Mediterranean Europe. Mammal Review, 39(1): 67-78.

Ruimy, A., Saugier, B. & Dedieu, G. (1994). Methodology for the

estimation of terrestrial net primary production from remotely sensed

data. Journal of Geophysical Research, 99:5263-5283.

Running, S.W., Thornton, P.E., Nemani, R. & Glassy, J.M. (2000) Global

terrestrial gross and net primary productivity from the earth observing

system. En: Sala, O., Jackson, R., Mooney, H. (Editores) Methods in

Ecosystem Science, New York. Springer-Verlag, 44-57 p.

Sala, O.E., Stuart, F., Armesto, J.J., Berlow, E., Bloomfield, J., Dirzo, R.,

Huber-Sanwald, E., Huenneke, L.F., Jackson, R.B., Kinzig, A., Leemans, R.,

Lodge, D.M., Mooney, H.A., Oesterheld, M., Poff, N.L., Sykes, M.T.,

Walker, B.H., Walker, M. & Wall, D.H. (2000). Global Biodiversity

Scenarios for the Year 2100. Science, 287(5459):1770-1774.

Schloss, A.L., Kickligther, D.W., Kaduk, J. & the participants of the

Postdam NPP Model Intercomparison. (1999). Comparing global models

of terrestrial net primary productivity (NPP): comparison of NPP to

climate and the Normalized Difference Vegetation Index. Global Change

Biology, 5(Suppl. 1):25-34.

Shannon, C.E. (1948). A mathematical theory of communications. The

Bell System Technical Journal, 27:379-423.

Shapiro, S.S. & Wilk, M.B. (1965). An analysis of variance test for

normality (complete samples). Biometrika, 52(3-4):591-611.

Silvestre, J. (2002). Las emigraciones interiores en España durante los

siglos XIX y XX: una revisión bilbliográfica. Ager, 2:227-248.

REFERENCIAS

152

Sims, D.A., Rahman, A.F., Cordova, V.D., Bassil, B.Z., Baldocchi, D.D.,

Flanagan, L.B., Goldstein, A.H., Hollinger, D.Y., Misson, L., Monson, R.K.,

Oechel, W.C., Schmid, H.P., Wofsy, S.C., & Xu, L. (2006). On the use of

MODIS EVI to asses gross primary productivity of North American

ecosystem. Journal of Geophysical Research: Biogeosciences,

111:G04015.

Singh, N.J., Yoccoz, N.G., Bhatnagar, Y.V. & Fox, J.L. (2009). Using habitat

suitability models to sample rare species in high-altitude ecosystems: a

case study with Tibetan argali. Biodiversity and Conservation,

18(11):2893-2908.

Snedecor, G.W. & Cochran, W.G. (1989). Statistical Methods, Eighth

Edition, Iowa State University Press.

Sokal, R.R. & Rohlf, F.J. (1995). Biometry. 3rd edition. W.H. Freeman,

New York.

Tapia, L., Domínguez, J., Regos, A. & Vidal, M. (2013). Using remote

sensing data to model European wild rabbit (Oryctolagus cuniculus)

occurrence in a highly fragmented landscape in northwestern Spain. Acta

Theriologica, doi: 10.1007/s13364-013-0169-2.

Underwood, A.J. (1997). Experiments in Ecology. Their Logical Design and

Interpretation Using Analysis of Variance. Cambridge University Press,

Cambridge.

Veloz, S.D. (2009). Spatially autocorrelated sampling falsely inflates

measures of accuracy for presence-only niche models. Journal of


REFERENCIAS

153

Virginia, R.A. & Wall, D.H. (2001). Ecosystem function, principles of.

Encyclopedia of Biodiversity (ed. by S.A.Levin), Academic Press, San

Diego, 345-352 p.

Virgós, E. (1994). Consideraciones sobre la situación del tejón (Meles

meles) en algunas áreas del centro de la Península Ibérica. Aegypius,

12:37-40.

Virgós, E. & Casanovas, J.G. (1999a). Environmental constraints at the

edge of a species distribution, the Eurasian badger (Meles meles L.): a

biogeographic approach. Journal of Biogeography, 6:559-564.

Virgós, E., Casanovas, J.G. (1999b). Badger Meles meles sett site selection

in low density Mediterranean areas of Central Spain. Acta Theriologica,

44(2):173-182.

Virgós, E., Tellería, J.L. & Santos, T. (2002). Effects of geographic location,

vegetation type, isolation and fragment features on the richness and

occurrence of medium-sized mammals in forest fragments of central

Spain. Biosiversity and Conservation, 11:1063-1079.

Virgós, E., Mangas, J.G., Blanco-Aguilar, J.A., Garrote, G., Almagro, N. &

Pérez-Viso, R. (2004). Food habits of the European badger (Meles meles)

along an altitudinal gradient of Mediterranean environments: a field test

of the earthworm specialization hypothesis. Canadian Journal of Zoology,

82:41-51.

Virgós, E. (2005). Tejón – Meles meles. En: Enciclopedia Virtual de los

Vertebrados Españoles. Carrascal, L. M., Salvador, A. (Editores). Museo

Nacional de Ciencias Naturales, Madrid. Disponible en:

http://www.vertebradosibericos.org/.

REFERENCIAS

154

Virgós, E., Revilla, E., Domingo-Roura, X. & Mangas, J. G. (2005c).

Conservación del tejón en España: síntesis de resultados y principales

conclusiones. En: Virgós, E., Revilla, E., Mangas, J.G. & Domingo-Roura X.

(Editores). Ecología y conservación del tejón en ecosistemas

mediterráneos. Sociedad Española para la Conservación y Estudio de los

Mamíferos (SECEM), Málaga., 283-294 p.

Virgós, E., Revilla, E., Mangas, J.G., Barea-Azcón, J.M., Rosalino, L.M. & De

Marinis, A.M. (2005a). Revisión de la dieta del Tejón (Meles meles) en la

Península Ibérica: comparación con otras localidades de su área de

distribución natural. En: Virgós, E., Revilla, E., Mangas, J.G. & Domingo-

Roura X. (Editores). Ecología y conservación del tejón en ecosistemas


Mamíferos (SECEM), Málaga, 67-80 p.

Virgós, E., Revilla, E., Mangas, J.G., Domingo-Roura, X. (Editores).

(2005b). Introducción. Ecología y conservación del tejón en ecosistemas



Virgós, E. & Revilla, E. (2005). Métodos de estudio de la abundancia del

Tejón (Meles meles). En: Virgós, E., Revilla, E., Mangas, J.G. & Domingo-

Roura X. (Editores). Ecología y conservación del tejón en ecosistemas



Vitousek, P.M. (1994). Beyond Global warming: Ecology and Global

Change. Ecology, 75:1861-1876.

REFERENCIAS

155

Wang, T., Ye, X., Skidmore, A.K. & Toxopeus, A.G. (2010). Characterizing

the spatial distribution of giant pandas (Ailuropoda melanoleuca) in

fragmented forest landscape. Journal of Biogeography 37:865-878.

Warren, D.L., Glor, R.E. & Turelli, M. (2008). Environmental niche

equivalency versus conservatism: quantitative approaches to niche

evolution. Evolution, 62:2868-2883.

Warren, D.L., Glor, R.E. & Turelli, M. (2010). ENMTools: a toolbox for

comparative studies of environmental niche models. Ecography, 33:607-

611.

Warren, D.L. & Seifert, S.N. (2011). Ecological niche modeling with

Maxent: the importance of model complexity and the performance of

model selection criteria. Ecological Applications, 21(2):335-342.

Wiegand, T., Naves, J., Garbulsky, M.F. & Fernández, N. (2008). Animal

habitat quality and ecosystem functioning: exploring seasonal patterns

using NDVI. Ecological Monographs, 78(1):87-103.

Wilesmith, J.W. (1983). Epidemiological features of bovine tuberculosis in

cattle herds in Great Britain. Journal of Hygiene, 90:159-176.

Wiley, E.O., McNyset, K.M., Peterson, A.T., Robins, C.R. & Stewart, A.M.

(2003). Niche modeling and geographic range predictions in the marine

environment using a machine-learning algorithm. Oceanography,

16(3):120-127.

Willems, E.P., Barton, R.A. & Hill, R.A. (2009). Remotely sensed

productivity, regional home range selection, and local range use by an

omnivorous primate. Behavioral Ecology, 20:985-992.

REFERENCIAS

156

Wilson, D.S. & Yoshimura, J. (1994). On the coexistence of specialist and

generalist. The American Naturalist, 144(4):692-707.

Wood, S.N. (2011). Fast stable restricted maximum likelihood and

marginal likelihood estimation of semiparametric generalized linear

models. Journal of the Royal Statistical Society: Series B (Statistical

Methodology), 73(1):3-36.

Woodroffe, R. & Macdonald, D.W. (1993). Badger sociality: models of

spatial grouping. Symposium of the Zoological Society of London, 65:145-

169.

Woodroffe, R. (1995). Body condition affects implantation date in the

European badger, Meles meles. Journal of Zoology, 236:183-188.

Woodroffe, R. & Macdonald, D.W. (1995). Female/female competition in

European badgers Meles meles: effects on breeding success. Journal of

Animal Ecology, 64:12-20.

Woodroffe, R. & Macdonald, D.W. (2000). Helpers provide no detectable

benefits in the European badger (Meles meles). Journal of Zoology,

250:113-119.

Yackulic, C.B., Chandler, R., Zipkin, E.F., Royle, J.A., Nichols, J.D., Grant,

E.C. & Veranet, S. (2012). Presence-only modelling using MAXENT: when

can we trust the inferences?. Methods in Ecology and Evolution,

4(3):236-243.

Yates, C.J., McNeill, A., Elith, J. & Midgley, G. (2010). Assessing the

impacts of climate change and land transformation on Banksia in the

South West Australian Floristic Region. Diversity and Distribution, 16:187-

201.

REFERENCIAS

157

Zaccarelli, N., Li, B.L., Petrosillo, I. & Zurlini, G. (2013). Order and disorder

in ecological time-series: Introducing normalized spectral entropy.

Ecological Indicators, 28:22-30.

Zamora, J., Verdú, J.R. & Galante, E. (2007). Species richness in

Mediterranean agroecosystems: spatial and temporal analysis for

biodiversity conservation. Biological Conservation, 134(1):113-121.

ANEXOS Resultado 4.1: Modeling Spatial Distribution of European Badger in Arid Landscapes: an Ecosystem Functioning Approach

158

ANEXOS

Resultado 4.1: Modeling Spatial Distribution of

European Badger in Arid Landscapes: an Ecosystem

Functioning Approach

Appendix A

Table A.1. Moran´s Index for the environmental variables.

Variable Short name Moran´s I Z score p-value

Mean slope SLO 0.015 2.268 0.02

Annual Mean Rainfall MRAIN 0.103 10.059 0

Mean value of the maximum

temperatures MMT 0.227 21.103 0

Area of scattered scrub SSCRUB 0.065 6.807 0

Area of dense scrub SDCRUB 0.004 1.711 0.086

Area of woody crop SWCROP -0.02 -0.958 0.337

Area of arable crop SACROP -0.004 1.404 0.16

Area of mixed crop SMICROP 0.013 2.756 0.005

Area of mosaic crop SMOCROP -0.006 0.371 0.709

EVI annual mean EVIMEAN 0.024 3.165 0.001

Standard deviation of EVI annual mean EVISTD 0.279 25.706 0

Coefficient of variation of EVI annual

mean EVICV 0.161 15.386 0

EVI autumn mean AEVI 0.059 6.285 0

EVI spring mean SEVI 0.031 3.731 0

EVI annual mean of scattered scrub SSCEVI -0.022 -1.283 0.199

EVI annual mean of dense scrub DSCEVI -0.014 -0.635 0.525

EVI annual mean of woody crop WCEVI -0.018 -0.715 0.474

EVI annual mean of arable crop ACEVI -0.005 0.499 0.617

EVI annual mean of mixed crop MICEVI 0.027 3.53 0

EVI annual mean of mosaic crop MOCEVI -0.015 -0.509 0.61


159

Appendix B

Table B.1. Environmental variables used for spatial distribution modeling.

Variable Layer type

Spatial

resolution

(m)

Temporal

resolution

Mean slopea Raster 20 2005

Annual Mean Rainfalla

Raster 100 1961-1990

Mean value of the maximum temperaturesa Raster 100

EVI annual meanb

Raster

250

16-days

EVI autumn meanb Raster

EVI spring meanb Raster

Coefficient of variation of EVI annual meanb Raster

Standard deviation of EVI annual meanb Raster

EVI annual mean of scattered scruba,b Raster

250

EVI annual mean of dense scruba,b Raster

EVI annual mean of woody cropa,b Raster

EVI annual mean of arable cropa,b Raster

EVI annual mean of mixed cropa,b Raster

EVI annual mean of mosaic cropa,b Raster

Area of scattered scruba Vector

250 2007

Area of dense scruba Vector

Area of woody cropa Vector

Area of arable cropa Vector

Area of mixed cropa Vector

Area of mosaic cropa Vector

Source data: aEnvironmental Infomation Network of Andalusia;

bMODIS sensor


160

Appendix C

Figure C.1. Occurrence of the European badger in the study area. We

emphasize that badger presence is mainly associated with cultivated

landscapes.


161

Appendix D

Figure D.1. The response curves show the contribution of each

environmental variable separately in the ALL model to the MaxEnt raw

prediction. The MaxEnt prediction may be seen as the suitable habitat

predicted by each pixel cell (Elith et al., 2011), so maximum values

correspond to the highest predicted suitability for European badger.


162

Appendix E

Figure E.1. Spatial distribution for European badger predicted by the best models from each combination of variables. It is

important to note here that suitability represented in the figures below was estimated under the assumption that the

probability of presence under average conditions was 0.5.

ANEXOS Resultado 4.2: Feeding Habits of European Badger (Meles meles) in Mediterranean Arid Landscapes

163

Resultado 4.2: Feeding Habits of European Badger

(Meles meles) in Mediterranean Arid Landscapes

Appendix F

R code to estimate the F´s distribution from ANOVA analysis by bootstrap

resampling technique.

#NOTE: This script is modified from the William B. King´s script in R #tutorials (http://ww2.coastal.edu/kingw/statistics/R-tutorials/resample.html) to carry out #a two-way ANOVA model with interactions.

#This example estimates the F´s distribution for the effects of #landscape and season factors on the fruit relative volume.

######################################################################

#we estimate the mean values in each level

meanblocks = with(scatsVRcomplet, tapply(fruits_v,fruits_season:fruits_landsca,mean,na.rm=T))

#we center all the groups on the same mean (zero), but we leave the variance and shape of the individual group distributions undisturbed

grpA = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="autumn"&scatsVRcomplet$fruits_landsca=="Forestry"] - meanblocks[1]))

grpB = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="autumn"&scatsVRcomplet$fruits_landsca=="Maquia"] - meanblocks[2]))

grpC = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="autumn"&scatsVRcomplet$fruits_landsca=="Xeric_shrubland"] - meanblocks[3]))

http://ww2.coastal.edu/kingw/statistics/R-tutorials/resample.html

http://ww2.coastal.edu/kingw/statistics/R-tutorials/resample.html


164

grpD = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="spring"&scatsVRcomplet$fruits_landsca=="Forestry"] - meanblocks[4]))

grpE = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="spring"&scatsVRcomplet$fruits_landsca=="Maquia"] - meanblocks[5]))

grpF = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="spring"&scatsVRcomplet$fruits_landsca=="Xeric_shrubland"] - meanblocks[6]))

grpG = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="summer"&scatsVRcomplet$fruits_landsca=="Forestry"] - meanblocks[7]))

grpH = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="summer"&scatsVRcomplet$fruits_landsca=="Maquia"] - meanblocks[8]))

grpI = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="summer"&scatsVRcomplet$fruits_landsca=="Xeric_shrubland"] - meanblocks[9]))

grpJ = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="winter"&scatsVRcomplet$fruits_landsca=="Forestry"] - meanblocks[10]))

grpK = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="winter"&scatsVRcomplet$fruits_landsca=="Maquia"] - meanblocks[11]))

grpL = as.vector(na.omit(scatsVRcomplet$fruits_v[scatsVRcomplet$fruits_season=="winter"&scatsVRcomplet$fruits_landsca=="Xeric_shrubland"] - meanblocks[12]))

season = scatsVRcomplet[!is.na(scatsVRcomplet$fruits_v),2]


165

landscape= scatsVRcomplet[!is.na(scatsVRcomplet$fruits_v),3]

#number of replicates for bootstrap

R = 10000

Fstar = numeric(R)

Fstar1 = numeric(R)

Fstar2 = numeric(R)

for (i in 1:R) {#loop for the replicates

groupA = sample(grpA, size=length(grpA), replace=T)

groupB = sample(grpB, size=length(grpB), replace=T)

groupC = sample(grpC, size=length(grpC), replace=T)

groupD = sample(grpD, size=length(grpD), replace=T)

groupE = sample(grpE, size=length(grpE), replace=T)

groupF = sample(grpF, size=length(grpF), replace=T)

groupG = sample(grpF, size=length(grpG), replace=T)

groupH = sample(grpF, size=length(grpH), replace=T)

groupI = sample(grpF, size=length(grpI), replace=T)

groupJ = sample(grpF, size=length(grpJ), replace=T)

groupK = sample(grpF, size=length(grpK), replace=T)

groupL = sample(grpF, size=length(grpL), replace=T)


166

simfacto = c(groupA,groupB,groupC,groupD,groupE,groupF,groupG,groupH,groupI,groupJ,groupK,groupL)

simdata = data.frame(simfacto,season,landscape)

Fstar[i] <- summary(aov(simfacto~season*landscape, data=simdata))[[1]]$F[1]

Fstar1[i]<- summary(aov(simfacto~season*landscape, data=simdata))[[1]]$F[2]

Fstar2[i]<- summary(aov(simfacto~season*landscape, data=simdata))[[1]]$F[3]

}

#Fstar: F´s distribution for the season effect

#Fstar1: F´s distribution for the landscape effect

#Fstar2: F´s distribution for the interaction effect

resampling<-matrix(c(Fstar,Fstar1,Fstar2),nrow=10000,ncol=3)

#critical value for the probability 0.95, with corresponding freedom #degrees but assuming normality.

qf(.95,df1,df2)#df1: freedom degrees of the factor; df2: freedom #degrees of the residuals

#critical value for the probability 0.95, with corresponding freedom #degrees but no assuming normality.

quantile(Fstar,.95)

#this script must be run for the relative volume of vertebrates and invertebrates.


167

The results are shown in the Table F.1. For the two-way ANOVA analysis

with season and landscape as fixed factors and the relative volume of the

wide categories as the response variable, no effect varied its signification

by applying the bootstrap resampling.

Table F.1. Critical values for 0.95 probabilities. With * no assuming normality,

without * assuming normality.

Relative volume Effect F Critical value *Critical value

Fruits Season 4.04 2.646 2.62

Landscape 13.3 3.037 2.901

Season x Landscape 1.22 2.14 2.059

Vertebrates Season 2.07 2.76 2.26

Landscape 5.56 3.153 3.05


Invertebrates Season 2.85 2.53 2.645

Landscape 11.61 3.036 3.058



168

Appendix G

Food consumption by fine categories.

Table G.1. Seasonal composition of the diet considering frequency of occurrence (FO) and

mean relative volume (RV) for each fine category in maquia.

Maquia

Summer n=6 Autumn n=12 Winter n=25 Spring n=11

Fruits Fruits Fruits Fruits

FO RV FO RV FO RV

FO RV

Olea

europaea 50.0 18.3

Chamaerops

humilis 33.3 35.0

Chamaerops

humilis 28.0 52.8

Olea

europaea 18.1 58.0

Chamaerops

humilis 33.3 5.0

Ceratonia

siliqua 58.3 82.0

Olea

europaea 16.0 47.5

Ceratonia

siliqua 63.6 32.0

Ceratonia

siliqua 66.6 30.0 Ficus carica 33.3 52.0

Ceratonia

siliqua 28.0 96.4 Citrus sp. 18.1 63.0

Vitis sp. 33.3 47.5 Olea

europaea 25.0 10.0 Citrus sp. 8.0 60.0

Ficus carica 33.3 35.0 Vitis sp. 16.6 38.0

Vertebrates Vertebrates Vertebrates Vertebrates

FO RV FO RV FO RV

FO RV

Birds 16.6 20.0 Rodentia 16.6 45.0 Rodentia 40.0 82.5 Rodentia 54.5 70.0

Eggs 33.0 15.0

Rabbits 8.0 42.5 Sauria 18.1 45.0

Invertebrates Invertebrates Invertebrates Invertebrates

FO RV FO RV FO RV

FO RV

Orthoptera 33.3 4.0 Orthoptera 16.6 19.0 Coleoptera 12.0 11.6 Orthoptera 18.1 13.0

Coleoptera 100.0 25.3 Gastropoda 16.6 4.0 Buthus

occitanus 4.0 10.0 Gastropoda 27.2 6.0

Buthus

occitanus 16.6 10.0 Coleoptera 25.0 10.0

Lumbricus

sp. 8.0 90.0 Others 9.0 84.0

Isopoda 50.0 4.3 Isopoda 25.0 7.0 Orthoptera 4.0 10.0

Buthus

occitanus 16.6 5.5


169


mean relative volume (RV) for each fine category in xeric shrubland.

Xeric shrubland



FO RV FO RV

FO RV

FO RV

Opuntia sp. 15.4 45.0 Olea

europaea 6.6 2.0

Ceratonia

siliqua 4.5 98.0 Citrus sp. 67.2 83.6

Malus

domestica 26.9 51.4

Ceratonia

siliqua 6.6 96.0 Citrus sp. 84.0 80.0 Others 3.6 95.0

Ficus carica 73.0 78.4 Ficus carica 93.3 94.0 Others 11.3 57.0 Prunus

armeniaca 21.8 44.3

Vitis sp. 11.5 46.6

Phoenix

dactylifera 2.2 15.0

Eriobotrya

japonica 1.8 5.0

Ceratonia

siliqua 7.6 5.0

Annona

cherimola 2.2 5.0

Other

vegetals 15.3 27.2


FO RV FO RV

FO RV

FO RV

Rodentia 3.8 20.0 Birds 6.6 10.0 Rodentia 15.9 49.0 Rodentia 9.0 22.0

Anura 3.8 100.0 Rodentia 6.6 30.0 Sauria 2.2 10.0 Rabbits 1.8 35.0

Rabbits 7.6 87.5

Birds 4.5 20.0 Birds 5.4 31.6

Sauria 5.4 13.3


FO RV FO RV

FO RV

FO RV

Orthoptera 15.4 6.2 Gastropoda 6.6 2.0 Orthoptera 25.0 18.0 Orthoptera 54.5 38.0

Buthus

occitanus 7.6 7.5 Coleoptera 20.0 5.0 Hymenoptera 4.5 30.0 Coleoptera 32.7 19.4

Isopoda 3.8 5.0 Orthoptera 20.0 8.3 Lumbricus sp. 6.8 30.0 Lumbricus sp. 1.8 60.0

Coleoptera 15.4 8.0

Coleoptera 18.1 8.1 Hymenoptera 3.6 7.5

Hymenoptera 7.6 4.0

Others 4.5 10.0 Others 1.8 10.0

Arachnida 3.8 10.0

Odonata 1.8 15.0

Gastropoda 1.8 30.0


170


mean relative volume (RV) for each fine category in forestry.

(c) Forestry



FO VR FO VR FO VR

FO VR

Juglans regia 8.3 70.0 Morus sp. 20.0 90.0 Others 33.3 80.0 Prunus dulcis 4.7 50.0

Morus sp. 33.3 50.0 Ficus carica 30.0 60.0 Castanea

sativa 6.6 5.0

Prunus dulcis 20.0 50.0

Other

vegetals 30.0 20.0


FO VR FO VR FO VR

FO VR

Rodentia 8.3 30.0 Rodentia 20.0 30.0 Rodentia 20.0 53.3 Rabbits 9.5 100.0

Eggs 8.0 30.0 Sauria 10.0 20.0 Sauria 6.6 85.0 Rodentia 14.2 46.6

Birds 4.7 10.0

Sauria 28.5 46.6


FO VR FO VR FO VR

FO VR

Orthoptera 25.0 28.3 Orthoptera 10.0 10.0 Orthoptera 53.3 11.8 Hymenoptera 57.1 33.3

Isopoda 16.6 5.0 Hymenoptera 60.0 54.0 Hymenoptera 20.0 16.6 Buthus

occitanus 4.7 5.0

Hymenoptera 41.6 27.8 Caterpillars 10.0 10.0 Lumbricus sp 53.3 81.8 Coleoptera 28.5 25.8

Buthus

occitanus 50.0 35.8

Buthus

occitanus 10.0 50.0 Coleoptera 13.3 17.5 Caterpillars 66.6 44.1

Diptera 25.0 11.6 Coleoptera 10.0 5.0 Others 6.6 15.0 Lumbricus sp 9.5 61.0

Coleoptera 58.3 52.8


171

Appendix H

Table H.1. Dissimilarity matrix based on the Bray-Curtis distance (dij).

MQ XS F

Sum Aut Win Spr Sum Aut Win Spr Sum Aut Win

MQ

Aut 0.321

Win 0.471 0.262

Spr 0.330 0.177 0.167

XS Sum 0.459 0.165 0.111 0.162

Aut 0.384 0.296 0.358 0.437 0.317

Win 0.351 0.170 0.190 0.231 0.216 0.184

Spr 0.334 0.291 0.206 0.324 0.321 0.171 0.108

F

Sum 0.297 0.198 0.203 0.241 0.303 0.312 0.183 0.126

Aut 0.350 0.274 0.183 0.308 0.306 0.257 0.162 0.085 0.084

Win 0.448 0.235 0.058 0.159 0.165 0.335 0.163 0.179 0.175 0.123

Spr 0.366 0.149 0.133 0.125 0.217 0.398 0.203 0.222 0.109 0.173 0.098

MQ: maquia; XS: xeric shrubland; F: forestry; Sum: summer; Aut: autumn, Win: winter; Spr: spring.

ANEXOS Resultado 4.3: Modeling and monitoring habitat quality from space: the European badger

172

Resultado 4.3: Modeling and Monitoring Habitat

Quality from Space: the European Badger

Appendix I

Weighting the species records to deal with the sample selection bias.

Due to the species records used in the study derived from different

datasets, some sites may be more poorly sampled than other. Phillips et

al. (2009) named this bias as sample selection bias and highlighted that it

can severely impact model quality. To consider this problem, we followed

the strategy used in Elith et al. (2010). These authors noted that without

records on survey effort we cannot distinguish between areas that are

environmentally unsuitable and those that are under-sampled. Thus, we

upweighted records with few neighbours in geographic space, to give

them prominence. In this form, we projected the species records into

World Geodetic System (revision WGS 84) and calculated the number of

records in each 3 x 3 km grid divided by the area of grid cell (i.e., 9 km2)

(to avoid edge effects at study area borders and coast). Then, we added

one to all values to avoid zero weights and therefore all pixels can be

chosen as background in MaxEnt algorithm. The result was a grid file with

high weights in densely sampled plots and low in sparsely sampled plots

(Fig. I.1), and it was used as bias grid in MaxEnt (see online help in

MaxEnt software).


173

Figure I.1. Bias map used to run MaxEnt.


174

Appendix J

Methods to predict the EVI variables into future conditions using the

climate variables.

To predict the EVIMEAN (surrogate of primary production) and EVICV

(indicator of seasonality) variables onto future scenarios (i.e., A2 and B1),

we estimated a first Generalized Additive Model (GAM) (Hastie &

Tibshirani, 1990) using EVIMEAN as response variable and only PREC as

explanatory variable, and then, a second GAM where we used EVICV as

response and PREC and TMEDMAX as explanatory variables. All these

variables were incorporated for current conditions. Finally, we used the

projected values of PREC and TMEDMAX variables onto future scenarios

to run the models and predict future values of both EVI variables. GAM

models were run in R, v. 3.0.3 (R Development Core Team, 2014) using

the “mgcv” library, v. 1.7-27 (Wood, 2011).

In GAMs each predictor is included in the model as a non-parametric

smoothing function allowing to fit nonlinear relationships between

response and predictors. We specified the error distribution of the

response as Gaussian and the link function as identity. We trained the

models on all cells from the study region. The optimal amount of

smoothing was automatically determined by a cross-validation process

(Wood, 2011). Parameters and predictor effects of the GAMs for

EVIMEAN and EVICV variables are shown in the Table J.1 and Fig. J.1

respectively.


175

Table J.1. Summary of the GAMs for EVIMEAN and EVICV variables.

EVIMEAN generalized additive model

Predictor edf F p-value Deviance explained

s(PREC) 8.99 9579 < 0.0001 12.1%

EVICV generalized additive model

Predictor edf F p-value Deviance explained

s(PREC) 8.99 753.2 < 0.0001 17.6%

s(TMEDMAX) 8.99 12213.2 < 0.0001

s(): smoothing function; edf: effective degrees of freedom (see the manual of

“mgcv” R package for explanation).

Figure J.1. Estimated smoothing curves of the GAMs for EVIMEAN and EVICV

variables in the study region. The x-axes show the values of the predictors (i.e.,

(a) precipitation; (b) precipitation and temperature) and the y-axes the

contribution of the smoother to the fitted values. The solid lines are the

smoother and the dotted lines 95% confidence bands.

Documents

Centro Andaluz para la Evaluación y Seguimiento del · PDF fileMaría Jacoba, Yolanda Cantón y Paco Domingo son de esas personas que basta con que pasen cerca de tu vida para que