Efectos del estrés abiótico y factores

bióticos en las interacciones planta-

planta: implicaciones para el

funcionamiento y la restauración de los

ecosistemas semiáridos

Autor: Santiago Soliveres Codina1,2

Directores: Fernando T. Maestre Gil1, Adrián Escudero Alcántara1 y Fernando

Valladares Ros2

1Área de Biodiversidad y Conservación. Departamento de Biología y Geología, Escuela

Superior de Ciencias Experimentales y Tecnología, Universidad Rey Juan Carlos.

2 Instituto de Recursos Naturales, Centro de Ciencias Medioambientales, Consejo Superior de

Investigaciones Científicas.

Madrid, 2010

Dr. Fernando T. Maestre Gil y Dr. Adrián Escudero Alcántara, Profesor Titular y Catedrático

de Universidad del Departamento de Biología y Geología de la Universidad Rey Juan Carlos,

respectivamente, y Fernando Valladares Ros, Profesor de Investigación del Instituto de

Recursos Naturales (Centro de Ciencias Medioambientales) del Consejo Superior de

Investigaciones Científicas,

CERTIFICAN:

Que los trabajos de investigación desarrollados en la memoria de tesis doctoral: “Efectos del

estrés abiótico y factores bióticos en las interacciones planta-planta: implicaciones para

el funcionamiento y la restauración de los ecosistemas semiáridos”, son aptos para ser

presentados por el Ldo. Santiago Soliveres Codina ante el Tribunal que en su día se consigne,

para aspirar al Grado de Doctor en Ciencias Ambientales por la Universidad Rey Juan Carlos

de Madrid.

VºBº Director Tesis VºBº Director de Tesis

Dr. Fernando T. Maestre Gil Dr. Adrián Escudero Alcántara

VºBº Director Tesis

Dr. Fernando Valladares Ros

A mi madre, por ser mi inspiración

A Soraya, por ser mi Todo.

La transición de los datos a la teoría requiere imaginación

creativa. Las hipótesis y teorías científicas no se “derivan” de

los hechos observados, sino que se “inventan” para dar cuenta

de ellos.

Carl Hempel

AGRADECIMIENTOS

“Joer, que cuatro años más largos” es lo primero que me viene a la mente al empezar a

escribir estos agradecimientos. Pero después echo la vista atrás y me doy cuenta que he vivido

un montón de experiencias, muchas más buenas que malas, y que he aprendido mucho en este

proceso que es la tesis, he disfrutado muchísimo y me he formado como científico, y aún más

importante, como persona (nunca es tarde para esto último, jeje). Todo esto se lo tengo que

agradecer a la gente que me ha rodeado, apoyado y ayudado durante toda esta importante

etapa de mi vida, a ellos va dedicada esta sección.

En primer lugar les quiero agradecer a mis directores de tesis el haberme brindado esta

oportunidad, el haber trabajado mucho y muy duro para que pudiera colaborar con un montón

de gente distinta, aprender cuanto necesitara en el camino y no tener que preocuparme por

problemas de financiación, algo básico para poder hacer ciencia. Muy especialmente te quería

agradecer a ti, Fernando (Maestre), por tu apoyo continuo y por ser un director inmejorable.

Sin tu ayuda y apoyo, tus innumerables ideas y la inmensa cantidad de horas que has

invertido, mirando y revisando cada uno de los capítulos de esta tesis en cada una de sus

fases, esto no habría sido posible. Pero te quiero agradecer mucho más, el haberme enseñado

una forma de trabajar que no tiene precio, y el haberme dado un ejemplo a seguir como

profesional y como persona. A ti, Adrián, te tengo que agradecer tu infinita paciencia y todos

los comentarios y aportaciones que has hecho a muchos de los trabajos que aquí se presentan.

Me has ayudado muchísimo a entender la ecología de comunidades y muchas otras ecologías.

Las discusiones científicas que hemos tenido y todos tus comentarios al respecto, han sido

todo un placer y un aprendizaje estupendo, sin mencionar la estadística! Y sobre todo me has

enseñado a ser crítico con todo lo que leo y escribo, y a aprender a formarme un criterio

propio, lo cual sin duda me ayudará a ser mejor científico. Fernando (Valladares) es sin duda

el responsable de que yo pudiera tener una beca para poder hacer la tesis. Te agradezco

también que siempre hayas conseguido sacar un rato para ayudarme en muchas etapas de esta

tesis, tus continuas aportaciones a muchos de estos trabajos sin duda han mejorado su calidad.

Dudo que pocas personas sepan más del efecto de la sombra y de ecofisiología en general que

tú, sólo espero haber sido un buen alumno.

A Pablo decirte que has sido un compañero de tesis excepcional, no tengo palabras para

agradecerte tu continuo apoyo a lo largo de toda esta etapa de mi vida. He aprendido

muchísimo de ti y, para mí, eres todo un ejemplo a seguir por muchas cosas. Por tu inmensa

capacidad de trabajo, por no dejarte caer nunca y tirar siempre adelante, fueran cuales fueran

las circunstancias y porque tienes unos coj… como no los tiene nadie. Muchísimas gracias

también por todos los buenos momentos que he pasado contigo, en la uni, en los taludes, en

las fieldtrips, en las roadtrips, y en las trips callejeras y camperas nocturnas y diurnas, y por

acogerme en tu casa siempre que lo he necesitado. También por ser sin duda el maestro de

ceremonias del grupo, por las Aranjuez summer festival, las fiestas en tu casa, por enseñarme

las jam-sessions madrileñas, por las cenas y por todo lo demás, mi experiencia madrileña

hubiera sido mucho peor si no hubieras estado. Espero que nos encontremos muchas veces

más en el camino. Andrea, a ti también te quería agradecer el ser una compañera estupenda, y

tu ayuda en muchas fases de esta tesis. No te creas que me olvidé de quien se quedó conmigo

aquellos primeros días hasta las 10 de la noche cortando los p… protectores para los conejos.

Espero que te vaya muy bien en lo que emprendas en el futuro. Los tres hemos vivido muchos

momentos juntos que recordaré siempre con cariño (eso de subir garrafas monte arriba es que

une mucho, jeje), y que espero que se repitan en el futuro. Me gustaría agradecer a Jorgito

(alias Piruan) por ser un compañero estupendo durante mis primeros pasos en el mundillo de

la ecología, por tu incontenible curiosidad y tus ganas de aprender siempre cosas nuevas.

Tampoco me olvido de que me has apoyado siempre, y has confiado en mí como pocos.

Espero que te esté yendo genial allá por la Puna.

Tengo que decir que estoy tremendamente orgulloso de pertenecer y haber trabajado en el

departamento de Biología y Geología de la URJC. El nivel científico de los Biodiversos es

incuestionable, pero aún es mucho mayor el nivel humano, el buen ambiente que se respira y

lo fácil que es trabajar con todos vosotros. Desde luego hacéis que ir a trabajar cada día sea

más un placer que una obligación. Muchísimas gracias por haberme aceptado y acogido, y

sobre todo, por haberme soportado. Recuerdo cómo Rubén y yo hablábamos en susurros el

primer día, cuando me estaba enseñando los despachos y presentándome a la gente…aquella

fue la única vez que hablé flojo en ese departamento, y sin embargo, todos los biodiversos han

sabido perdonarme (o eso espero)… También espero que sepáis perdonarme mi sentido del

humor algo “brusco” y no siempre acertado, espero que sepáis que siempre fue sin mala

intención. En especial quería agradecerles a mis compis de despacho todos los buenos

momentos que he vivido con vosotros durante estos años. A Cris Escolar, la pequeña del

grupo fernandiano, por tu eterna alegría contagiosa y por ser una compañera genial. A Edu,

aquel señor bajito, con el pelo de dudosa procedencia, y oriundo de “un país chiquito, al ladito

del cielo”. Simplemente, personas como tú hacen que el mundo sea un lugar mejor. Muchas

gracias por preocuparte por todos nosotros siempre, por ser el alma de Biodispersos (Úbeda

queda como testigo) y por no parar de organizar cosas para que, aquellos que estuviéramos

fuera y añoráramos a nuestra gente, nos sintiéramos siempre como en casa. A todos los

demás, Ares, Sonia, Mari Carmen, Samu, Pablo, Alberto, Rubén Milla y Alf (si, si…os

escapasteis, pero para mí seguís siendo de mi despacho), muchas gracias por vuestra paciencia

y ayuda, por todo lo que nos hemos reído y por los buenos recuerdos que me llevo de todos

vosotros. Gracias a todos por hacer de ese despacho un lugar cálido donde diera gusto

trabajar. También quería agradecer a toda la gente que me ayudó en algún momento con el

trabajo de esta tesis, a Kike, Cristina Alcalá, Vicky, Patri, Dolo, Becky, Yoli, Chele, los

Rubenes (Milla y Torices), Mariajo, José Margalet, Luis Giménez, Ozeluí y muchos otros que

seguro que se me olvidan, pido perdón por ello. A Luisiño, por su optimismo y simpatía, y

por ser un gran amigo y compañero. También porque sin él los eventos lúdico-deportivos de

este departamento serían inexistentes. Me gustaría también agradecerles a todos mis otros

compañeros biodispersos el que me hayan ayudado cada vez que los he necesitado, y por

todos los buenos momentos que hemos vivido juntos. A Rocío, María y Mónica, que junto

con Andrea (Javi, Samu y Rubén Torices eran, supuestamente, la parte sensata) han sido el

terror de vigilantes de metro y porteros de discoteca, me lo he pasado genial con todos

vosotros cada vez que nos hemos juntado y espero que nos sigamos juntando muchas veces

más…sea en el país que sea. También a Isa, Ana Millanes, Cris Fernández, Raúl, Javi, Sonja,

Luis Giménez y Julián porque, sin duda, habéis ayudado a pasar estos años en la distancia de

una forma mucho más agradable. No quiero olvidar a los nuevos, Peska, Carlos, Gema, Sonia

(la otra) y los demás, me voy bien tranquilo sabiendo que las nuevas generaciones seguirán

formando un departamento cojonudo donde trabajar y que pasarán muchos años antes de que

haya un cochinillo aburrido en Navidad, espero eso sí, que mejoréis el nivel futbolístico

actual.

Resulta que a las pocas horas de llegar a Móstoles me vi metido en un coche con un tipo

asturiano con barba y pelo largo que me llevaba a ver un piso en la c/ Camino de Humanes,

12. Allí esperaba una andaluza, fumando como un carretero, y con un tembleque extraño en la

pierna. Resulta que Rubén y Mariajo eran esa gente, y que fueron mi familia durante más de 2

años. No puedo dejar de agradeceros que hicierais de esa casa un hogar en el que poder

desahogarme y estar completamente a gusto, vuestro continuo apoyo y el que hayáis sido mis

“hermanos mayores”, lo bien que lo hemos pasado y todo lo que he aprendido y disfrutado

con vosotros. Me alegro mucho de poder teneros como amigos. A la familia tengo que sumar

a Lucia y Horta, ha sido un gustazo conoceros y he disfrutado de cada momento que hemos

pasado juntos.

A toda la gente que ha colaborado en algún capítulo de esta tesis, a Lucia por ayudarme con

toda la parte de dendro y carbohidratos, y por enseñarme lo poco que se, sin tu ayuda el

capítulo 2 no habría sido posible. A Chemi Olano, que con sus ideas y participación hizo que

éste capítulo fuera mucho mejor. David Eldridge (alias Mr. Fantastic), the aussie guy with

italian accent, thanks for all your help and your extraordinary sense of humour; and especially

for completing the aussie gradient!!!. It has been a pleasure to meet and work with you. Matt

Bowker, thanks for a lot of things, it has been wonderful everytime I worked with you and

also the time we spent outside the job. Thanks also for an excellent feedback in our scientific

discussions, for your patience and for introducing me in the causal correlation world. I´m very

grateful because it was you who showed me that stop and think is the first thing you should do

to start doing good science. To the aussie team, Nick Reid and Matt Tighe, for hosting me in

Armidale and helping me with all the logistics during my stay. I especially thank Nick Schultz

for helping me during all the fieldwork in Australia. A Rubén, porque sin él el capítulo 5 no

existiría, por ser tan rápido y eficiente currando y porque ha sido un gustazo descubrir que,

además de un buen amigo, eres muy buen científico. A Estrella, por iniciarme con el mundillo

R y por ayudarme a buscar parcelas en Alicante, y a Bea Amat por echarme un cable con los

árboles de regresión.

A mi gente de Alicante. Pero sobretodo al “Kanutet Vilero”, sin Jorge, Curro, Juande, Joanmi

(gràcies per lo de la portada, rei), Raúl y Juanito mi vida en general, y los cuatro años de la

tesis en particular, no serían lo mismo (no sé si mejor o peor, pero desde luego, mucho más

aburrida). Sólo decir que tengo mucha suerte de teneros como amigos. Muchas gracias por

estar siempre ahí, por perdonarme mis repetidas ausencias, y por estar siempre dispuestos a

pasar conmigo el poco tiempo libre que he podido sacar, por las risas que nos hemos echado,

y por ayudarme en todo cuanto os he pedido. Lo de ayudarme a plantar 2300 árboles el plena

navidad no se me olvidará nunca…aunque con todo el alcohol que me habéis hecho filtrar,

cabrones, ni siquiera puedo entender cómo es que aún recuerdo mi propio nombre! Tampoco

me quiero olvidar de los alicantinos expatriados, Jorgito, JM (ex-expatriados), el Cuñao,

Kiket, Vicen, Carla, Paula y Aitana, gracias por hacer que me sintiera en Madrid como en

casa, o mejor dicho, que me sintiera acompañado fuera de ella.

Sin mi madre nada de esto (ni muchísimas otras cosas) nunca hubiera sido posible, tinc que

agraírte moltes coses, pero sobretot que sempre hages estat ahí, que sigues una lluitadora i tot

un exemple a seguir, i que ho hages donat tot per uns fills que no et mereixen, gràcies mare

PER TOT. También me gustaría agradecerle al resto de mi familia, A Benilde, Juan Antonio,

Jorge y Katia, porque sin su apoyo uno muchas veces no podría seguir adelante. Y como no, a

mi recua de mascotas, las que siguen y las que se fueron, a Stick, Peluso, Lluna, Mac, y a mis

últimos “compañeros de despacho”, a Acho y Legaña. Por estar siempre dispuestos a hacerme

compañía y darme cariño sin pedir nada a cambio, por lo que me habéis hecho reír y disfrutar,

y por enseñarme cosas tan útiles como que, si te echas al suelo en verano, se está más

fresquito que en el sofá. No sabéis leer (creo), pero valéis un imperio.

Y uno siempre se deja lo mejor para el final. A Soraya, por ser mi continuo apoyo y por estar

siempre junto a mí. Por soportar 3 años y medio de distancia y por acompañarme al final del

mundo cuando te lo pedí (11 vuelos en 13 días, y eso que te da miedo volar!). Por dejarte

robar horas por ese infame ordenador, unas horas que me gustaría, y debería, haber pasado

contigo, y nunca reprochármelo (yo sí que lo hago, créeme). Por ayudarme en todo, siempre.

Porque sin tí nada tiene mucho sentido y contigo y tu eterna sonrisa los problemas parecen

unos puntitos lejanos en el horizonte. Y porque no recuerdo haber vivido mejores momentos

en mi vida que los que paso junto a ti, es imposible pensar que haya alguien mejor con quien

uno podría estar.

Seguro que me he dejado a mucha gente, y espero que sepan perdonarme…

A TODOS, GRACIAS POR TODO

ÍNDICE

� Resumen

Antecedentes 3

Objetivos 26

Metodología general y área de estudio 29

Estructura general de la tesis 35

� Capítulo 1

Predicted climate change effects in rainfall regime modulate the outcome of grass-

shrub interactions in two semiarid communities. 39

� Capítulo 2

Spatio-temporal heterogeneity in abiotic factors modulate multiple ontogenetic shifts

between competition and facilitation. 65

� Capítulo 3

Temporal dynamics of herbivory and water availability interactively modulate the

outcome of a grass-shrub interaction in a semiarid ecosystem. 89

� Capítulo 4

On the relative importance of climate and biotic non-trophic interactions as drivers of

local plant species richness 111

� Capítulo 5

On the relative importance of environmental conditions, biotic interactions and

evolutionary relationships as drivers of the structure of semiarid communities. 153

� Discusión y conclusiones generales 185

� Bibliografía y afiliación de los coautores 218

RESUMEN

RESUMEN

2

3

ANTECEDENTES

Las llamadas “tierras secas” (drylands) engloban todos aquellos ecosistemas de ambientes

desde hiperáridos a secos-subhúmedas; representando en total un 41% de la superficie

emergida del planeta (Millenium Ecosystem Assessment [en adelante MEA] 2005, Reynolds

et al. 2007a,b; Fig. A1). Las tierras secas se caracterizan generalmente por tener

precipitaciones escasas y variables, temperaturas aéreas extremas y una evapotranspiración

potencial elevada (Noy-Meir et al. 1973, Whitford 2002, Reynolds et al. 2007b). Estas

características ambientales hacen que estos sistemas tengan una baja productividad, que es

altamente variable dependiendo de las condiciones de cada año y de una alta heterogeneidad

espacial en la disponibilidad de nutrientes y la productividad vegetal (Whitford 2002). Pese a

ello, las tierras secas representan una parte importante de la biodiversidad global (Convención

para la Lucha contra la Desertificación 2005; en adelante CLD), y son el hogar y la fuente de

sustento del 38% de la población mundial (Reynolds et al. 2007a,b). Los impactos

antropogénicos (i.e. cambios en el uso del suelo, sobre-explotación de recursos, aumento de

las infraestructuras) y el aumento de la aridez provocado por el cambio climático son algunas

de las causas más importantes de la degradación de las tierras secas, comúnmente llamada

desertificación (MEA 2005, Reynolds et al. 2005, 2007b). Un aumento del nivel de

degradación implica una pérdida del funcionamiento y de los servicios ecosistémicos,

afectando directamente al bienestar de una parte importante de la población humana (MEA

2005, Reynolds et al. 2007a). Una vez que una zona ha sido degradada, revertir estos cambios

es difícil, ya que se requieren profundas transformaciones socio-económicas que afecten al

desarrollo y manejo de estas áreas, inversiones sustanciales de recursos externos, y un

profundo conocimiento de los factores que condujeron a la merma de la productividad y de

funcionamiento ecosistémico (MEA 2005, Reynolds et al. 2007a,b). Un alto porcentaje de las

tierras secas, hasta el 70% según la CLD, están amenazadas de degradación, mientras que un

10-20% de ellas ya están degradadas en mayor o menor grado (MEA 2005).

Más de las dos terceras partes del territorio español pertenecen a lo que se define como

tierras secas, y hasta el 36% de su territorio está amenazado por la desertificación,

concentrándose la mayoría de este área en la mitad sur Peninsular (Ministerio de Medio

Ambiente, Rural y Marino; Fig A1). Aunque no es la única condición, un mejor conocimiento

sobre cómo funcionan los ecosistemas en las tierras secas, sobre los factores que afectan a su

RESUMEN

4

biodiversidad, y sobre cómo están respondiendo estos ecosistemas al cambio climático o a

incrementos en las perturbaciones que les afectan, y que están asociados con motores de

cambio global (i.e. herbivoría, incendios), nos permitirá desarrollar herramientas para

prevenir y combatir la desertificación (MEA 2005, Reynolds et al. 2005, 2007a). Los

ecosistemas naturales presentan un cierto nivel de resiliencia, de forma que pueden resistir

cierto nivel de perturbación o incremento de estrés sin verse severamente afectados, y

pudiendo recuperarse bajo condiciones ambientales promedio (Noy-Meir 1975, Westoby et al.

1989, Briske et al. 2003); el problema es que, una vez alcanzados los umbrales de resistencia

de estos ecosistemas, éstos pueden sufrir cambios repentinos a estados severamente

degradados, desde donde el retorno puede ser imposible pese a que las condiciones

ambientales que provocaron el tránsito del umbral se modifiquen (Briske et al. 2003, Cortina

et al. 2005, Kefi et al. 2007, Scheffer et al. 2009). Por tanto, es fundamental establecer cuales

son los diferentes umbrales (i.e. niveles de perturbación, aridez) que hacen colapsar los

mecanismos de resiliencia de estos ecosistemas a estados más degradados; como por ejemplo

cambios en la disponibilidad de hábitats (Bascompte y Rodríguez 2001), el colapso de la

expansión de nicho promovida por las plantas nodriza (Michalet et al. 2006), o cambios en el

patrón espacial y tamaño de los parches de vegetación que afecten a la captura y

redistribución de los recursos (Schlesinger et al. 1990, Tongway y Hindley 1995, Kefi et al.

2007). Esta tesis doctoral se centra en los sistemas semiáridos, y por tanto, es sobre ellos en

concreto sobre los que hablaremos a partir de ahora.

LA DINÁMICA Y EL FUNCIONAMIENTO DE LOS SISTEMAS SEMIÁRIDOS

La principal característica de los ambientes áridos y semiáridos es que llueve poco (índices de

aridez [precipitación anual/evapotranspiración potencial] de entre 0,5-0,05), que esta lluvia es

generalmente impredecible, y que normalmente se produce en pulsos más o menos extensos

seguidos de temporadas secas prolongadas (Noy-Meir 1973, Whitford 2002). Por

consiguiente, el reclutamiento de nuevas plántulas sólo se produce durante pulsos de elevada

disponibilidad de agua, que son poco frecuentes e irregulares a lo largo del tiempo; estos

pulsos suelen corresponder con años particularmente benignos desde el punto de vista

climático (p. ej. Eldridge et al. 1991, Escudero et al. 1999, Kitzberger et al. 2000, Whitford

2002, Holmgren et al. 2006). Aunque los pulsos de agua más abundantes y continuos existen,

son los pulsos de agua cortos y de escasa intensidad (<5mm) los que predominan (Whitford

2002, Huxman et al. 2004). La diferente frecuencia entre los distintos tipos de pulsos, y cómo

5

los distintos componentes del ecosistema (i.e. microorganismos del suelo, costra biológica y

distintos tipos funcionales de plantas) aprovechan estos pulsos de agua, son la razón de que

los ecosistemas áridos y semiáridos sustenten niveles de diversidad relativamente altos pese a

su escasa productividad (Sala y Lauenroth 1982, Fowler 1986, Ogle y Reynolds 2004,

Schwinning et al. 2004).

Figura A1 Distribución global de las tierras secas (arriba; Fuente: MEA 2005), y áreas amenazadas de desertificación en España (de verde a amarillo representa de menor a mayor grado de amenaza, imagen de abajo. Fuente: Ministerio de Medio Ambiente, Rural y Marino: http://www.mma.es/portal/secciones/biodiversidad/desertificacion/desertificacion_espnia/index.htm).

La cobertura vegetal discontinua, organizada en manchas discretas de vegetación

embebidas en una matriz de suelo desnudo, con escasa cobertura vegetal (Fig. A2),

RESUMEN

6

característica de muchos ambientes áridos y semiáridos es el origen de la dinámica fuente-

sumidero que define el funcionamiento de estos ecosistemas (Ludwig y Tongway 1995,

Aguiar y Sala 1999, Puigdefábregas et al. 1999). La dinámica fuente-sumidero consiste en el

arrastre de agua, nutrientes y semillas durante los eventos lluviosos de cierta magnitud desde

la matriz de suelo desnudo o de muy escasa cobertura (fuente) hasta las manchas de

vegetación existente (sumideros). Aunque la costra biológica del suelo y los microorganismos

presentes en la matriz de suelo desnudo tienen un papel importante en la captura de carbono y

el reciclado de nutrientes de estos ecosistemas (Belnap y Lange 2003, Huxman et al. 2004,

Castillo-Monroy et al. 2010); la mayor parte de la productividad, reciclado de nutrientes y

captura de carbono ocurre en las manchas de vegetación, dominadas en su mayoría por

arbustos y herbáceas graminoides perennes, que actúan como islas de recursos en un ambiente

mucho más pobre (Franco y Nobel 1987, Schlesinger y Pilmanis 1998, Aguiar y Sala 1999).

La concentración de los recursos en dichas manchas conlleva generalmente un aumento de la

productividad, heterogeneidad ambiental y diversidad a nivel local (Noy-Meir 1973, Huxman

et al. 2004, Schwinning et al. 2004), algo que sería más difícil de alcanzar si los recursos se

repartieran de forma homogénea en el ecosistema (Noy-Meir 1973, Aguiar y Sala 1999). A su

vez, es precisamente esta concentración de los recursos la que mantiene la estructura

discontinua, con la presencia de manchas discretas de vegetación, a largo plazo, ya sea en

áreas donde esta estructura existía de por sí (Ludwig y Tongway 1995, Puigdefábregas et al.

1999, Rietkerk y Van der Koppel 2008), o bien en lugares donde cambios en la composición

vegetal han generado una distribución heterogénea de los recursos (Schlesinger et al 1990,

Archer 1994). La dinámica fuente-sumidero mantiene esta estructura heterogénea mediante

procesos de retro-alimentación positivos a escala de mancha (más recursos, más

productividad en las manchas de vegetación, más captura de recursos), y negativos a mayores

escalas (concentración de recursos y semillas en las manchas de vegetación dificulta el

reclutamiento en sitios libres de vegetación, y por tanto, el desarrollo de patrones espaciales

uniformes a nivel del ecosistema; Schlesinger et al. 1990, Archer 1994, Rietkerk y Van de

Koppel 2008).

Las diferencias en las condiciones ambientales (sombra, fertilidad del suelo,

infiltración), junto con una mayor cantidad de semillas concentradas al quedar atrapadas en

las manchas de vegetación por el efecto de la escorrentía o por la deposición de aves y

mamíferos, hace que sea en estos lugares donde se concentra una gran parte del reclutamiento

de nuevas plántulas (Aguiar y Sala 1999), aunque esto dependerá de los requerimientos

ecológicos de cada especie (p. ej. Miriti et al. 1998, 2007, Caballero et al. 2008). El efecto

7

positivo que tienen unas plantas sobre otras mediante la mejora de las condiciones

ambientales (p. ej. reducción de temperatura y viento, mejora de las condiciones hídricas

mediante sombreo, mayor fertilidad del suelo) y/o la defensa frente a la herbivoría recibe el

nombre de “facilitación” (Callaway 1995; Figura A2), siendo uno de los factores

fundamentales que interviene en el reclutamiento de nuevas plántulas en estas manchas de

vegetación (Niering et al. 1963, McAuliffe 1988, Eccles et al. 1999). Sin embargo, la

intercepción del agua de lluvia por los doseles de las plantas adultas y la mayor competencia

por agua, luz y nutrientes debajo de estos manchas de vegetación, junto con los efectos

negativos de la caída hojarasca, tanto físicos como químicos, relacionados con la liberación

de compuestos alelopáticos, producen efectos negativos que pueden llegar a superar a los

efectos beneficiosos según las especies implicadas y las condiciones ambientales reinantes

(revisado en Fowler 1986, Callaway 2007). Además de esto, las mismas condiciones

ambientales que son beneficiosas para las plántulas (i.e. poca radiación incidente) pueden

resultar negativas para plantas más adultas (Schupp 1995), lo que hace que las interacciones

positivas o facilitativas puedan volverse competitivas a medida que las plantas facilitadas

avanzan en su desarrollo (Fowler 1986, Callaway y Walker 1007, Miriti 2006). Además, lo

que puede ser beneficioso para una plántula en un determinado momento y condiciones puede

ser pernicioso en otros, sin necesidad de cambios ontogenéticos (De la Cruz et al. 2008). El

equilibrio entre facilitación/neutralidad/competencia es un determinante fundamental de la

dinámica de las comunidades vegetales en los ecosistemas semiáridos (Fowler 1986, Aguiar y

Sala 1999). Por tanto, un mayor conocimiento sobre la importancia relativa de las

interacciones positivas frente a las negativas en distintos procesos y atributos ecosistémicos

(i.e. composición, estructura), así como un entendimiento de los condicionantes para que se

de uno u otro signo en la interacción, son fundamentales para poder entender el ensamblaje de

las especies, la dinámica y funcionamiento de estos ecosistemas y mejorar su restauración

(Fowler 1986, Cortina et al. 2005, Callaway 2007, Brooker et al. 2008, Gómez-Aparicio

2009).

RESUMEN

8

Figura A2 Estructura espacial heterogénea con parches de vegetación herbácea y arbustiva en un espartal de Stipa tenacissima en Zorita, España (arriba). Individuos de Austrostipa scabra reclutando bajo un arbusto adulto en un bosque abierto de Eucalyptus populnea en Nyngan, Australia (abajo).

9

LA HIPÓTESIS DEL GRADIENTE DE ESTRÉS Y LA INTRODUCCIÓN DE LA FACILITACIÓN EN LA TEORÍA

ECOLÓGICA

Aunque las interacciones positivas entre plantas son conocidas desde hace tiempo (Clements

1916, Shreve 1931, Niering et al. 1963), han sido ampliamente ignoradas en ecología hasta

hace 25 años (Callaway 1995). Durante los años previos enfatizó el papel de procesos como la

competencia entre plantas, el efecto de las perturbaciones o el estrés abiótico en el ensamblaje

de comunidades (p. ej. Grime 1973, Huston 1979, 1999, Tilman 1988), ignorando la

importancia de la facilitación en este proceso (ver revisión en Callaway 2007). Prueba de ello

es que sólo 27 artículos científicos versaron sobre facilitación en el período 1900-1989,

mientras que el número de artículos sobre este tema en los últimos 20 años asciende hasta

1252 (Pakeman et al. 2009). Además, la facilitación ha sido introducida como un proceso

importante en el ensamblaje de especies en comunidades, especialmente en medios

“estresantes” (i.e. medios áridos y semiáridos, alta montaña, ecosistemas salobres,

comunidades intermareales), ganándose un sitio en el marco de la teoría general ecológica

(Callaway 1997, Hacker y Gaines 1997, Stachowicz 2001, Bruno et al 2003, Lortie et al.

2004, Michalet et al. 2006, Callaway 2007).

Gran parte de este “viraje” en la atención prestada a la facilitación en ecología se la

debemos a Mark Bertness y Ragan Callaway, que en 1994 publicaron su Hipótesis del

Gradiente de Estrés (en adelante SGH, del inglés “Stress-Gradient hipótesis”). Estos autores

propusieron un modelo teórico sencillo y biológicamente muy plausible, en el cual predecían

un aumento de la importancia y la frecuencia de las interacciones facilitativas frente a las

competitivas a medida que el estrés abiótico o las perturbaciones aumentan. A partir de ese

momento, las interacciones positivas pasaron a ser un mecanismo a tener en cuenta a la hora

de estudiar la dinámica de las poblaciones, y luego las comunidades vegetales en todo el

mundo (Callaway 2007). Bajo el paraguas de la SGH, numerosos ecólogos han fijado su

atención en estas interacciones positivas y su efecto en la dinámica de ciertos pares de

especies en particular (p. ej. Valiente-Banuet et al. 1991, Maestre et al. 2001, Gómez-Aparicio

et al. 2004, Sthultz et al. 2007), en el efecto de ciertas especies clave sobre las demás especies

(p. ej. Holzapfel y Mahall 1999, Pugnaire y Luque 2001, Maestre y Cortina 2005, Badano y

Cavieres 2006), o, más raramente, en la importancia de la facilitación a nivel de comunidades

enteras (p. ej. Hacker y Bertness 1999, Kikvidze et al. 2005, Maestre et al. 2010). Además,

esta hipótesis no solo ha sido evaluada en plantas, sino que existen estudios sobre su

aplicabilidad a comunidades intermareales (p. ej. Stachowicz 2001, Kawai y Tokeshi 2007,

RESUMEN

10

Daleo y Iribarne 2009), costra biológica del suelo (Mulder et al. 2001, Maestre et al. 2008,

Bowker et al. 2010) e incluso entre herbívoros de diferente tamaño (Arsenault y Owen-Smith

2002). Prueba de su importancia son las más de 830 citas que ha recibido el artículo de

Bertness y Callaway (1994) hasta la fecha.

A pesar de constituir un punto de partida inmejorable para las investigaciones sobre

facilitación, la SGH ha encontrado numerosas excepciones que hacen poner en duda la

generalidad de sus predicciones. Así, resultados obtenidos cuando se evalúa la actividad

vegetal con diferentes parámetros (Goldberg y Novoplansky 1997, Maestre et al. 2005,

Gómez-Aparicio 2009), se estudian niveles extremadamente altos de estrés (Kitzberger et al.

2000, Tielbörger y Kadmon 2000a, Maestre y Cortina 2004), o se consideran de forma

simultánea distintos factores de estrés (Ibañez y Schupp 2001, Kaway y Tokeshi 2007,

LeRoux and McGeoch 2010) contradicen las predicciones de la SGH. Además, las

interacciones planta-planta son altamente específicas, independientemente del nivel de estrés

en el que ocurran (Callaway 2007). Esto hace pensar que los diferentes rasgos ecológicos de

las especies que interactúan (Choler et al. 2001, Liancourt et al. 2005, Gross et al. 2009,

Maestre et al. 2009a, Gómez-Aparicio 2009) o sus relaciones evolutivas (Valiente-Banuet et

al. 2006, Valiente-Banuet y Verdú 2007, Castillo et al. 2010) son factores fundamentales que

afectan al resultado de estas interacciones. Estos resultados contradictorios han sido el punto

de partida de numerosos debates y redefiniciones de la SGH, que han tratado de incorporar el

efecto de la identidad y características ecológicas de las especies implicadas, los distintos

niveles y tipos de estrés o las distintas medidas de rendimiento vegetal utilizadas (Holmgren

et al 1997, Maestre et al. 2005, 2006, 2009a, Lortie y Callaway 2006, Callaway 2007,

Malkinson y Tielbörger 2010, Holmgren y Scheffer 2010). De todas estas aportaciones se

puede concluir que: a) la facilitación será especialmente intensa y frecuente a niveles

intermedios de estrés; esto puede deberse a que, bajo niveles elevados de estrés, el efecto

negativo de la competencia supera al efecto positivo; o bien a que las plantas nodriza no son

capaces de mejorar suficientemente las condiciones ambientales bajo condiciones tan

extremas (Maestre y Cortina 2004, Michalet et al. 2006, Holmgren y Scheffer 2010), b) el

resultado de las interacciones dependerá de las características ecológicas de las especies

implicadas (competidoras o tolerantes al estrés, sensu Grime 2001) y de si el tipo de estrés

está directamente relacionado con un recurso (p. ej. agua o luz), o no (p. ej. temperatura o

salinidad; Maestre et al. 2009a), c) el resultado de las interacciones dependerá del efecto de la

nodriza sobre la disponibilidad de recursos (p. ej. luz, agua o temperatura), y las tolerancias

relativas de las especies facilitadas (p. ej. tolerancia a la sombra, a la sequía o al frío;

11

Holmgren et al. 1997, Choler et al. 2001, Liancourt et al. 2005, Malkinson y Tielbörger 2010,

Holmgren y Scheffer 2010), d) el resultado de las interacciones bajo distintos tipos de estrés

puede ser de muy diferente naturaleza, pudiendo darse jerarquías entre los distintos tipos de

estrés presentes (Baumeister y Callaway 2006), sinergias o anulaciones de sus efectos (Kawai

y Tokeshi 2007) o bien, simplemente que éstos sean aditivos (Riginos et al. 2005), y e) las

interacciones entre plantas tenderán a ser más positivas cuanto menos relacionadas (i.e. más

distantes en la evolución) estén las especies que interactúan (Valiente-Banuet et al. 2006,

Valiente-Banuet y Verdú 2007, Castillo et al. 2010).

Además de estas complicaciones añadidas a la simplicidad excesiva de la SGH,

hemos de tener en cuenta que, en medios áridos y semiáridos la disponibilidad hídrica y de

nutrientes se produce en pulsos, seguidos de largos períodos de tiempo con una disponibilidad

de agua o nutrientes muy limitada (inter-pulsos; Goldberg y Novoplansky 1997, Whitford

2002), lo que complica aún más la predicción sobre el resultado de las interacciones planta-

planta. Por ejemplo, Goldberg y Novoplansky (1997) desarrollaron un modelo en el que

predecían un efecto negativo de las nodrizas sobre el crecimiento durante los pulsos o épocas

benignas, mientras que este efecto cambiaría a positivo, aumentando la supervivencia, durante

los inter-pulsos o épocas más secas. El resultado final de la interacción dependería de: 1) el

efecto relativo de la competencia con la nodriza frente a los factores abióticos en el

agotamiento de los nutrientes durante los inter-pulsos, y 2) del efecto de la reducción en el

crecimiento durante los pulsos sobre la supervivencia posterior en los inter-pulsos (Goldberg

y Novoplansky 1997). A pesar de su importancia, esta variabilidad intra-anual en los recursos

y su importancia relativa sobre el resultado final de las interacciones planta-planta ha sido

poco estudiada (Barchuk et al. 2005, Kikvidze et al. 2006, Sthultz et al. 2007, de la Cruz et al.

2008).

Los diversos modelos existentes sobre los efectos del cambio climático para la región

Mediterránea semiárida predicen, además de una reducción de la cantidad anual en las

precipitaciones, un cambio sustancial en su patrón temporal (IPCC 2007). En el futuro es

esperable que las épocas secas o inter-pulsos sean más largos, las lluvias durante los pulsos

menos abundantes, y los eventos de lluvias torrenciales más frecuentes (IPCC 2007, Knapp et

al. 2008, Miranda et al. 2009). Es muy probable que estos cambios tengan efectos profundos

sobre la vegetación de los ecosistemas semiáridos (Ogle y Reynolds 2004, Holmgren et al.

2006, López et al. 2008, Heisler-White et al. 2009, Miranda et al. 2009, Pías et al. 2010). Por

consiguiente, es fundamental estudiar cómo las interacciones planta-planta van a verse

afectadas por estos cambios, o cómo pueden mitigar estos cambios aumentando la resiliencia

RESUMEN

12

del ecosistema ante el incremento de aridez y el cambio en el patrón temporal de las

precipitaciones (Brooker et al. 2008). Sin embargo, son muy pocos los estudios que han

evaluado el efecto de estos cambios en la frecuencia y abundancia de los eventos de lluvia

sobre las interacciones planta-planta (Zavaleta 2006, Knapp et al. 2008, Matías 2010).

Ya se ha comentado anteriormente que los cambios en los requerimientos ecológicos

de las distintas especies durante la ontogenia pueden modificar los resultados de sus

interacciones con otras plantas (Callaway y Walker 1997, Miriti 2006). Estos cambios han

sido relacionados con la existencia de interacciones positivas durante la germinación y el

desarrollo de plántula, debido a una mayor vulnerabilidad a la sequía o a la herbivoría

(Schupp 1995, Cavender-Bares y Bazzaz 2000, Ibañez y Schupp 2001) y con el incremento de

la competencia por la luz y el agua a medida que las plántulas crecen y se convierten en

individuos reproductivos (Miriti 2006, 2007, Schiffers y Tielbörger 2006, Armas y Pugnaire

2009). Sin embargo, esta relación no siempre es tan evidente, y parece depender del nivel de

estrés ambiental reinante en cada estado ontogenético y del estado fisiológico (Ibañez y

Schupp 2001, Sthultz et al. 2007, Butterfield et al. 2010), de las diferencias en la forma de

crecimiento de las especies que interactúan (Gómez-Aparicio 2009), o de las relaciones

evolutivas que definen las diferencias entre los rasgos ecológicos entre estas especies

(Valiente-Banuet y Verdú 2008). La inmensa mayoría de los estudios que evalúan cambios

ontogenéticos en las interacciones planta-planta se centran en ventanas temporales concretas a

lo largo del desarrollo de las especies facilitadas (Armas y Pugnaire 2005, 2009, Miriti 2006,

2007, Valiente-Banuet y Verdú 2008). Sin embargo, estas aproximaciones no dan una visión

global del tema, ya que no tienen en cuenta el efecto de arrastre que las condiciones

ambientales de años anteriores pueden tener en el presente. El efecto de las condiciones del

año anterior sobre el crecimiento presente es fundamental en medios semiáridos (Whitford

2002). Por consiguiente, el no tener en cuenta dicho efecto puede confundir la relación entre

el resultado de una interacción planta-planta dada y las condiciones ambientales reinantes en

un año concreto. Estudios que sigan el desarrollo de estas interacciones durante el tiempo

necesario para que estas especies de vida larga pasen de plántula a adulto son logísticamente

prohibitivos, por lo que el uso de medidas indirectas (i.e. estudios dendrocronológicos)

podrían ser una solución adecuada para evaluar la interacción clima-ontogenia en estas

especies (véase Armas y Pugnaire 2005, Sthultz et al. 2007, Aragón et al. 2008, Pías et al.

2010 para otras aproximaciones).

Además, es fundamental que estos estudios a lo largo de distintos estados

ontogenéticos se realicen con distintas combinaciones de formas de crecimiento que

13

interactúen entre ellas (Gómez-Aparicio 2009). Aunque muchas plántulas tienen escasa

tolerancia a la sequía y pueden estar menos estresadas bajo el dosel de una planta adulta

(Schupp 1995 pero ver Caballero et al. 2008, Matías 2010), las diferencias en los

requerimientos ecológicos de individuos adultos entre tipos funcionales o formas de

crecimiento diferentes pueden ser muy importantes. Por ejemplo, distintas tolerancias a la

sombra o la sequía (Holmgren et al. 1997, Hastwell y Facelli 2003) o diferentes

profundidades de enraizamiento (Sala y Lauenroth 1982, Schwinning et al. 2001, Ogle y

Reynolds 2004) pueden ser clave a la hora de definir el resultado de las interacciones entre

adultos de distintas especies. Los estudios sobre cambios de facilitación a competencia a lo

largo de la ontogenia se centran en interacciones arbusto-arbusto, con ambas especies

teniendo características morfológicas y funcionales parecidas, por lo que es esperable que

compartan sus nichos ecológicos de explotación de agua, luz y nutrientes (p. ej. Miriti 2006,

Armas y Pugnaire 2009). Sin embargo, estudios centrados en otro tipo de interacciones (i.e.

herbácea-arbusto o entre arbustos con estrategias y formas contrastadas), muy abundantes por

otro lado en medios semiáridos (Aguiar y Sala 1999), pueden arrojar resultados muy

diferentes. Ello es así debido a que ambas formas de crecimiento difieren en la toma de agua y

nutrientes por sus distintas profundidades de enraizamiento (Sala y Lauenroth 1982, Fowler

1986) y a que tienen alturas contrastadas, lo que puede reducir la competencia entre ambos

grupos por la luz. Por ejemplo, Armas y Pugnaire (2005) encontraron que los individuos

juveniles de la gramínea perenne Stipa tenacissima tenían efectos neutros o negativos sobre el

arbusto Cistus clusii, dependiendo del clima; pero que este efecto negativo se volvía neutro (o

incluso positivo, en condiciones más secas) cuando individuos adultos de ambas especies

interactuaban. En otro estudio, Gasque y García-Fayos (2004) encontraron que tanto las

plántulas como los adultos de Pinus halepensis se desarrollaban mejor cuando crecían cerca

de una macolla de S. tenacissima, sin encontrar cambios en estas interacciones positivas en

distintos momentos del desarrollo de P. halepensis, aunque los efectos positivos de S.

tenacissima sobre las plántulas de P. halepensis desaparecieron durante la sequía estival, ya

que ninguna plántula sobrevivió. Resultados que contrastan con una relación monotónica

entre el paso de facilitación a competencia con la edad (p. ej. Miriti 2006, Valiente-Banuet y

Verdú 2008) no se han encontrado sólo en interacciones leñosa-herbácea, si no que también

son comunes en otro tipo de interacciones. Tirado y Pugnaire (2003) encontraron efectos

facilitativos, independientemente del momento ontogenético, al estudiar la interacción entre

dos arbustos (Asparagus albus y Ziziphus lotus). Sthultz et al. (2007) también encontraron

que el efecto del arbusto Fallugia paradoxa sobre el árbol Pinus edulis era positivo en sitios

RESUMEN

14

de elevado estrés, independientemente de la edad de P. edulis, aunque el efecto positivo de F.

paradoxa se tornaba negativo a medida que P. edulis crecía en sitios menos estresantes. Las

interacciones entre especies herbáceas parecen estar normalmente dominadas por la

competencia (Gómez-Aparicio 2009), aunque esto depende en un grado importante de las

condiciones ambientales reinantes (Graff et al. 2007, Veblen 2008), y con toda probabilidad

de los rasgos funcionales de las herbáceas que interactúen (Cahill et al. 2008, Gómez-

Aparicio 2009). En consecuencia, la relación entre el signo de las interacciones planta-planta,

la ontogenia y las condiciones climáticas parece ser altamente específica de cada especie, o

por lo menos para cada forma de crecimiento (Gómez-Aparicio 2009), por lo que no sirve un

modelo simple para predecir la evolución de las interacciones. Sin duda, estudios destinados a

entender mejor estas relaciones son fundamentales para entender la dinámica de las

comunidades vegetales en los escenarios climáticos presentes y futuros (Fowler 1986,

Butterfield 2009).

EL PAPEL DE LA HERBIVORÍA Y SU INTERACCIÓN CON EL CLIMA EN LAS INTERACCIONES PLANTA-

PLANTA

La herbivoría es un factor fundamental que afecta al desarrollo de las plantas en condiciones

semiáridas (McNaughton 1978, Milchunas et al. 1989, Fowler 2002, Kefi et al. 2007).

Sabemos que la presencia de herbívoros y su preferencia por ciertas especies pueden influir en

la dinámica competitiva entre las plantas (Gurevitch et al. 2000, Fowler 2002), con profundos

efectos en la diversidad y abundancia relativa de las especies vegetales (Westoby et al. 1989,

Fuhlendorf et al. 2001, Briske et al. 2003). También se conoce el efecto que algunas especies

no palatables pueden ejercer en el mantenimiento de la diversidad en las comunidades

vegetales mediante su papel protector/facilitador sobre otras (Hay et al. 1986, Callaway et al.

2000, Rebollo et al. 2002, Baraza et al. 2006, Veblen 2008). De hecho, este efecto protector

puede compensar el efecto negativo derivado de la competencia por agua o nutrientes, dando

como resultado neto una asociación positiva entre pares de especies (Graff et al. 2007). La

herbivoría es, por tanto, un factor determinante de las interacciones planta-planta en

ambientes semiáridos (Rebollo et al. 2002, Baraza et al. 2006, Graff et al. 2007, Veblen

2008).

Sin embargo, al igual que pasa con la mejora microclimática, el efecto protector de

estas plantas nodriza tiene límites, y bajo niveles extremadamente altos de herbivoría u otras

perturbaciones, este efecto positivo puede llegar a desaparecer (Brooker et al. 2006, Smit et

15

al. 2007, 2009, Graff et al. 2007, Forey et al. 2009). En medios semiáridos, la presión de los

herbívoros puede aumentar a medida que la disponibilidad de forraje se reduce con menores

disponibilidades hídricas (Illius y O´Connor 1999, Chase et al. 2000) o por aumento de cargas

de estos herbívoros por dinámicas endógenas; a su vez, las condiciones abióticas pueden

acelerar o retrasar la recuperación de las comunidades vegetales después de la perturbación

(Fuhlendorf et al. 2001). Es lógico pensar que ambos factores, herbivoría y aridez, interactúan

a la hora de definir el resultado de las interacciones planta-planta y la dinámica de las

comunidades vegetales (Van Auken 2000, Roques et al. 2001, Smit et al. 2009). Sin embargo,

a pesar de los numerosos estudios realizados para evaluar las interacciones planta-planta bajo

niveles diferentes de estrés abiótico o herbivoría de forma separada (revisados en Callaway

2007), muy pocos trabajos han evaluado el efecto conjunto que la interacción entre ambos

tipos de estrés produce sobre estas interacciones (Ibañez y Schupp 2001, Veblen 2008,

Anthelme et al. 2009). La co-ocurrencia de ambos tipos de estrés puede afectar de forma

importante a las interacciones planta-planta, provocando profundos efectos en la dinámica de

las comunidades, que son difíciles de predecir si sólo consideramos ambos factores de estrés

por separado (Smit et al. 2009).

Además del efecto protector descrito anteriormente, las plantas nodriza pueden

incrementar la tolerancia de las plantas facilitadas a la herbivoría (Rand 2004, Acuña-

Rodríguez et al. 2006). Las mejores condiciones hídricas y la mayor concentración de

nutrientes presentes frecuentemente bajo el dosel de las plantas nodriza pueden ayudar a la

recuperación de ciertas especies después de la pérdida de biomasa producida por la herbivoría

(Crawley et al. 1998). Sin embargo esto no es tan sencillo, ya que el efecto que unas mejores

condiciones hídricas o nutricionales, pero una menor radiación solar, tengan sobre la

tolerancia a la herbivoría dependerá de cual es el recurso limitante en cada caso, y de cómo la

pérdida de biomasa producida por la herbivoría afecte a la toma de este recurso. Wise y

Abrahamson (2005, 2007) propusieron el Modelo de Recursos Limitados (“Limited Resource

Model”), que predice correctamente la inmensa mayoría de relaciones entre tolerancia a la

herbivoría y disponibilidad de recursos. Este modelo sugiere que la tolerancia a la herbivoría

será mayor en condiciones más húmedas cuando el agua es el factor más limitante (algo

común en sistemas semiáridos), ya que estas condiciones más húmedas favorecen mayores

tasas de fotosíntesis, permitiendo compensar las pérdidas de biomasa ocasionadas por los

herbívoros (Crawley et al. 1998). Sin embargo, en los casos en los que la luz es también un

factor limitante (p. ej. Marañón y Bartolomé 1993, Seifan et al. 2010a), la tolerancia a la

herbivoría será más baja conforme aumente la humedad (Baraza et al. 2004, Wise y

RESUMEN

16

Abrahamson 2005, 2007; Fig. A3). Esto se debe a que la fotosíntesis de estas plantas es

potencialmente más alta bajo condiciones más húmedas, lo que hace que la pérdida relativa de

rendimiento sea mayor cuando la fotosíntesis se ve limitada por la baja disponibilidad de luz

producida por la pérdida de biomasa causada por la herbivoría. En cambio, en condiciones

más secas, el crecimiento vegetal ya está limitado por la baja disponibilidad hídrica, por lo

que el efecto relativo que tiene una menor disponibilidad lumínica en la fotosíntesis es mucho

menor y la tolerancia es más alta (Fig. A3). Por consiguiente, tanto el efecto que las plantas

nodriza tendrán sobre la tolerancia de las plantas facilitadas a la herbivoría, como las

variaciones de este efecto bajo distintos niveles de disponibilidad hídrica son difíciles de

predecir.

Figura A3 Predicciones del Modelo de Recursos Limitados en el caso de que la herbivoría afecte a la toma del recurso principal (Hrp; en nuestro caso el recurso principal sería agua) o a la toma de un recurso alternativo (Hra; en nuestro caso luz). Estas predicciones varían si el nivel inicial del recurso principal es bajo (panel de la izquierda) o alto (panel de la derecha). Modificado de Wise y Abrahamson 2007.

Debido a la gran importancia de ambos factores (herbivoría y agua) en ambientes

semiáridos, parece prioritario llevar a cabo estudios destinados a resolver como las

interacciones planta-planta afectan a la incidencia de los herbívoros sobre las plantas

facilitadas, y a la tolerancia a la herbivoría de estas plantas bajo distintas disponibilidades

hídricas. Esto trabajos son fundamentales para entender la dinámica de las comunidades

vegetales semiáridas (Smit et al. 2009). Este conocimiento puede ayudar a desarrollar

estrategias de manejo adecuadas, que mantengan unos niveles razonables de pastoreo

dependiendo de las condiciones de cada año, permitiendo un desarrollo sostenible en

ambientes semiáridos (el 55% del área ocupada por estos ambientes se destina a la ganadería;

17

MEA 2005). Particularmente, aquellos estudios enfocados en las interacciones leñosa-

herbácea pueden ayudar a predecir el efecto de la herbivoría y el incremento de la aridez en

ecosistemas afectados por procesos de matorralización.

EL PAPEL DE LA FACILITACIÓN EN EL ENSAMBLAJE DE COMUNIDADES

Uno de los debates que más ha fascinado a los ecólogos a lo largo de su historia (ver Gotelli y

Graves 1996, Callaway 1997, 2007, Hubbell 2001, Alonso et al. 2006 para revisiones sobre

este debate) es el iniciado a partir de los trabajos de Clements (1916) y Gleason (1926), los

cuales proponen modelos opuestos sobre la concepción de las comunidades vegetales: un

superorganismo en el cual las interacciones bióticas juegan un papel fundamental (Clements

1916), o bien una coincidencia de especies que coexisten debido a sus adaptaciones

independientes a las condiciones ambientales de cada lugar (Gleason 1926). Resolver este

debate no sólo tiene un atractivo teórico, también tiene implicaciones fundamentales para la

conservación de los ecosistemas naturales. Comprender la naturaleza de una comunidad y ver

las distintas especies como entidades reemplazables y sustituibles, o como organismos con

fuertes relaciones de interdependencia puede conllevar cambios drásticos en como

entendemos y gestionamos la biodiversidad (discutido en Callaway 2007). Existen numerosos

estudios que apoyan ambas teorías, siendo clave la escala temporal, y sobretodo espacial, a la

que estos estudios se realizan (Hubbell 2001, Stokes y Archer 2010).

Estudios conducidos en áreas extensas, a niveles biogeográficos o regionales, parecen

indicar que las interacciones bióticas no son importantes, y que las comunidades por tanto, no

son estables, sino que cambian con el tiempo debido a procesos estocásticos de especiación,

dispersión y extinción (Hubbell 2001, pero ver Gotelli et al. 2010). Sin embargo, estudios a

escalas locales advierten sobre la suma importancia que las interacciones bióticas tienen en el

mantenimiento de estas especies en un lugar dado, así, procesos como la exclusión

competitiva, la segregación de nicho o los mutualismos entre numerosos organismos (p. ej.

insectos polinizadores y plantas) han sido demostrados como clave para mantener el conjunto

concreto de especies que coexisten en una comunidad dada (Levin 1970, Diamond 1975,

Huston 1979, 1999, Rezende et al. 2007), especialmente en aquellas más consolidadas o

menos pioneras (Stokes y Archer 2010). Por otro lado, sabemos que las interacciones

facilitativas son claves para mantener la diversidad en las comunidades naturales (Bruno et al.

2003, Callaway 2007). Así, muchas de las especies nodriza han sido catalogadas como

ingenieros del ecosistema (sensu Jones et al. 1996) en numerosos ecosistemas de todo el

RESUMEN

18

mundo (Hacker y Bertness 1999, Stachowicz 2001, Maestre y Cortina 2005, Badano y

Cavieres 2006). Como ya se ha comentado anteriormente, estas especies modifican las

condiciones microambientales en sus cercanías, lo que permite la existencia de especies

menos adaptadas a las condiciones ambientales particulares de ese sitio, ya que sin la

presencia de estas especies nodriza no podrían sobrevivir (expansión de nicho; Bruno et al.

2003).

Así pues, parece claro que ambas visiones de lo que es una comunidad son

complementarias, y que la escala espacial que consideremos (no tanto la temporal, como se

discutirá en la siguiente sección) es fundamental para entender la importancia de los procesos

estocásticos frente a las interacciones bióticas para el ensamblaje de las comunidades. De

hecho, Hubbell (2001) reconoce que “aunque la asunción de neutralidad completa es, sin

duda, falsa, pocos ecólogos negaran que las poblaciones y comunidades reales no están

sujetas sólo a los factores físicos y las interacciones bióticas, sino también a la estocasticidad

demográfica…Las comunidades ecológicas están indudablemente gobernadas por reglas de

ensamblaje de nicho y dispersión, junto con la estocasticidad demográfica, pero la pregunta

importante es: ¿Cuál es la importancia relativa y cuantitativa de estos procesos?”. Bien, han

pasado casi diez años desde que se hizo esta pregunta, y algo hemos aprendido en el camino.

Tanto las aproximaciones teóricas como empíricas parecen apuntar al mismo sitio: mientras

que son los procesos estocásticos a gran escala (especiación y dispersión) los que determinan

que especies de plantas “aparecen” en una comunidad dada, son los factores físicos (también

estocásticos en algún grado) y las interacciones bióticas los que determinan en mayor parte

cuales de estas especies permanecen (Huston 1999, Lortie et al. 2004a, Rajaniemi et al. 2006,

Rezende et al. 2007, pero ver Gotelli et al. 2010).

Si bien se presupone que tanto los factores ambientales (i.e. clima, perturbaciones)

como las interacciones bióticas condicionan el ensamblaje de comunidades a escalas locales,

poco se sabe sobre la importancia relativa de ambos factores como determinantes de dicho

ensamblaje (Butterfield et al. 2010). Callaway (2007) advierte sobre el efecto tampón que

algunas plantas nodriza pueden tener sobre la variabilidad inter-anual en las precipitaciones

(que podemos considerar estocástica) de algunos desiertos y zonas semiáridas, promoviendo

ambientes más estables en estos ecosistemas. Esta revisión sugiere, por tanto, que las posibles

interacciones entre las condiciones abióticas y las interacciones planta-planta son clave a la

hora de definir la dinámica de las comunidades naturales, ya que pueden modificar los efectos

de la estocasticidad climática en estas poblaciones.

19

Algunos estudios evaluaron la hipótesis de que las interacciones bióticas

(fundamentalmente la competencia) eran clave en los sistemas más productivos, mientras que

las condiciones ambientales, y la falta de adaptaciones fisiológicas de la mayoría de especies a

estas condiciones, determinaban la caída en la riqueza de especies observada en los ambientes

menos productivos (Grime 1973, Huston 1979). De estos estudios se deriva una relación

unimodal entre la riqueza y la productividad de las comunidades naturales, donde en ambos

extremos se vería una caída de la diversidad de especies respecto a niveles intermedios de

productividad, donde los niveles de diversidad serían máximos (Grime 1973). Tanto Hacker y

Gaines (1997) como Michalet et al. (2006) discutieron el papel que podrían jugar las

interacciones facilitativas en esta relación unimodal entre productividad y diversidad. Ambos

trabajos afirman que las plantas nodriza, mediante modificación del microambiente bajo su

dosel, pueden reducir el filtro abiótico que se da en las condiciones menos productivas,

aumentando así la diversidad en niveles medios-altos de “estrés”. Estos trabajos difieren en

que Hacker y Gaines (1997) asumen que la relación entre las interacciones positivas y el

incremento de estrés es positiva y monotónica, mientras que Michalet et al. (2006) apuntan a

un colapso de estos efectos facilitativos bajo niveles extremadamente altos de estrés, donde

incluso el crecimiento de las plantas nodriza, y por tanto su capacidad de modificación del

microambiente, estaría limitado.

Aunque estos modelos sencillos fueron un punto de partida excelente para empezar a

evaluar el papel relativo de los factores físicos y las interacciones bióticas en el ensamblaje de

las especies a nivel local, son insuficientes. Por un lado, se basan en una relación diversidad-

productividad unimodal, la cual no es, ni mucho menos, tan general como se esperaba (Grace

1999, Waide et al. 1999, Gillman y Wright 2006, sólo en 2010 se han publicado en la revista

Ecology cinco estudios discutiendo la generalidad de esta relación). Por otro lado, presuponen

la existencia de un gradiente de “estrés” que afecta de forma general todas las especies de una

comunidad, y que este nivel de estrés aumenta a medida que se reduce la productividad, lo

cual tampoco tiene por qué ser cierto (ver discusión en Körner 2003, 2004, Lortie et al.

2004b). Cada especie presenta unas características propias que le permitirán tener un óptimo

ambiental en unas condiciones particulares (sean estas más o menos productivas); por tanto, a

medida que nos alejamos de estas condiciones ambientales óptimas, esta especie en particular

verá aumentado su nivel de estrés (Chapin et al. 1987, Körner 2003). Sin embargo, las

distintas especies que coexisten en una comunidad difieren en mayor o menor grado en sus

óptimos ambientales y, por tanto, es incorrecto considerar que todas ellas se verán afectadas

de la misma manera a medida que cambien las condiciones ambientales (Chapin et al. 1987,

RESUMEN

20

Lortie et al. 2004b). Por lo tanto, el uso de aproximaciones basadas en el papel de las especies

“tolerantes al estrés” facilitando la existencia de las especies más “competidoras” a medida

que el estrés aumenta (p. ej. Travis et al. 2005) es inadecuado porque estas estrategias

ecológicas cambian con las condiciones ambientales (una especie tolerante a un determinado

factor de estrés no tiene porque ser tolerante a otros tipos de estrés; y una especie competidora

verá modificadas sus habilidades competitivas dependiendo de las condiciones ambientales en

las que se desarrolle). La revisión y cuestionamiento de estos tres supuestos (la existencia de

una relación unimodal entre riqueza y productividad, de un nivel de estrés único que afecte a

comunidades naturales enteras y de estrategias ecológicas que permanecen estables a lo largo

de gradientes ambientales amplios), permitirá explorar alternativas más realistas sobre el

papel que juega la expansión de nicho y las modificaciones microclimáticas promovidas por

las plantas nodriza a lo largo de gradientes ambientales. Sería de esperar, entonces, que el

efecto positivo de las plantas nodriza sobre la diversidad local se extienda con igual

importancia a lo largo de gradientes ambientales amplios, ya que este efecto positivo afectará

al mismo número de especies, aunque su identidad (tanto de las nodrizas como de las

facilitadas) vaya cambiando a medida que cambien las condiciones ambientales y unas

especies u otras se alejen de su óptimo ambiental (Greiner la Peyre 2001, Choler et al. 2001,

Liancourt et al. 2005, Chu et al. 2008, Holmgren y Scheffer 2010). La superación de los

supuestos aludidos debería tener profundas implicaciones en nuestra forma de ver la

importancia relativa de las interacciones positivas a lo largo de gradientes ambientales,

haciendo innecesario hablar de un nivel de estrés único a nivel de comunidad, ya que este

nivel cambiará con cada especie y condición ambiental. Esta es quizás, la razón fundamental

de los resultados contradictorios sobre los cambios en el signo y la intensidad de las

interacciones planta-planta a lo largo de gradientes ambientales (Maestre et al. 2005, 2006,

Lortie y Callaway 2006). Asimismo, la frecuencia e importancia de las interacciones positivas

a nivel de comunidad deberían mantenerse estables a lo largo de gradientes ambientales

amplios, ya que la identidad, pero no la cantidad, de las especies facilitadas es lo que

cambiará a lo largo de estos gradientes. Sin duda, estudios diseñados para evaluar estas

predicciones serán de ayuda para finalmente entender tanto el papel relativo de los factores

físicos y las interacciones bióticas en el mantenimiento de la diversidad local, como la

relación de esta diversidad con la productividad (Mulder et al. 2001, Callaway 2007).

La inmensa mayoría de las aproximaciones que versan sobre el efecto de la

facilitación en las comunidades naturales, tanto teóricas como empíricas, se centran en los

niveles medios-altos de “estrés” (Hacker y Gaines 1997, Lortie et al. 2004a, Travis et al.

21

2005, Michalet et al. 2006, Callaway 2007), otros estudios revelan que los efectos de ciertas

plantas sobre sus vecinas pueden extenderse en condiciones muy productivas (Levine 1999,

Laird y Schwamp 2006, 2009, Brooker et al. 2008), donde cabría esperar que la exclusión

competitiva jugara un papel fundamental (Grime 1973, 2001). Uno de los mecanismos

propuestos (la complementariedad de nicho, descrito a continuación) no puede considerase

como facilitación, ya que no implica que una determinada especie se beneficie por la

presencia de otra. En ambientes heterogéneos, una mayor diversidad de especies, o grupos

funcionales distintos, puede llevar a una mayor y más eficiente explotación de los recursos,

aumentando la productividad de la comunidad a mayores niveles de diversidad debido a la

complementariedad de nicho (p. ej. Hector et al. 1999).

Figura A4 Quizás el ejemplo más conocido de redes de competencia intransitiva es el juego de “Piedra, papel o tijeras”. Podemos observar como la complejidad de las redes intransitivas aumenta a medida que añadimos más actores (especies) al juego. Este tipo de redes de competencia intransitiva es más probable que ocurran en las manchas de vegetación cuanto mayor sean la riqueza de especies y la heterogeneidad en los recursos por los que compiten. Esto puede encontrarse a medida que nos movemos desde áreas de suelo desnudo hacia las manchas de vegetación de mayor complejidad (imágenes de la parte de debajo de la figura).

Sin embargo, otros mecanismos sí que están directamente relacionados con las

interacciones positivas entre plantas. Entre ellos destacan la facilitación indirecta, es decir el

efecto positivo de una especie determinada sobre otra, mediado por el efecto negativo de la

primera sobre una tercera especie (Levine 1999, Callaway 2007, Brooker et al. 2008, Cuesta

RESUMEN

22

et al. 2010). Si imaginamos una sencilla comunidad de tres especies (A, B y C), donde hay

una jerarquía competitiva marcada (A>B>C), entonces, es fácil de imaginar que A puede

facilitar a C mediante su efecto negativo sobre B. Otro mecanismo que atañe a las

interacciones positivas de una manera indirecta es la competencia intransitiva, esto es, la

inexistencia de una jerarquía marcada en las habilidades competitivas de las especies que

coexisten (Gilpin 1975). Si volvemos a nuestra comunidad de tres especies, será fácil de

entender que si A>B>C>A entonces el balance competitivo está más equilibrado y se pueden

mantener mayores niveles de diversidad (Laird y Schwamp 2006, 2009, Bowker et al. 2010;

Fig. A4). Esta competencia intransitiva sólo puede existir cuando la heterogeneidad en los

recursos y en los grupos funcionales que coexisten permite un equilibrio en las habilidades

competitivas de las especies en una comunidad (Grace 1993, Huston 1999). Ya hemos dicho

anteriormente que las plantas nodriza, y los parches que estas forman, son una de las mayores

fuentes de heterogeneidad en ecosistemas semiáridos (p. ej. Pugnaire et al. 1996a, Tracol et al.

2010). Bajo el dosel de estas plantas se dan condiciones heterogéneas de luz, agua, nutrientes

o redes micorrícicas (Pugnaire et al. 1996a, Holmgren et al. 1997, Wolfe et al. 2009 y

referencias en ese texto) que pueden generar las condiciones necesarias de heterogeneidad

para que se de competencia intransitiva o, alternativamente, segregación de nicho. Ambos

mecanismos pueden promover un aumento de la diversidad de especies que coexisten bajo su

dosel (Grace 1993, Pugnaire et al. 1996a, Hastwell y Facelli 2003, Silvertown 2004, Maestre

y Cortina 2005, Badano y Cavieres 2006, Laird y Schwamp 2006). Así pues, es probable que

se den procesos de retroalimentación positiva entre ambos procesos (más heterogeneidad y

más diversidad generan competencia intransitiva o segregación de nicho, que a su vez

aumentan la diversidad) que aumenten de forma desproporcionada la diversidad local de las

comunidades vegetales. No obstante, hasta la fecha sólo hay un estudio que evalúe el efecto

de las plantas nodriza sobre la dinámica competitiva de sus especies facilitadas (Tielbörger y

Kadmon 2000b), y no se ha evaluado este efecto conjuntamente con otros mecanismos como

la mejora microambiental y expansión de nicho. Estudios que evalúen los efectos de las

plantas nodriza sobre la riqueza de especies local, teniendo en cuenta en un mismo marco

general todos los posibles mecanismos por los que estas plantas pueden aumentar la

diversidad (expansión de nicho, competencia intransitiva o segregación de nicho), son

necesarios para entender finalmente el papel de las interacciones planta-planta en la

diversidad local y, por tanto, en la productividad y el funcionamiento ecosistémico (Mulder et

al. 2001, Hooper et al. 2005) a lo largo de gradientes ambientales (Callaway 2007).

23

EL PAPEL DE LA FACILITACIÓN EN LA EVOLUCIÓN DE LAS COMUNIDADES VEGETALES

Aunque se ha sugerido que las interacciones bióticas no son importantes en comparación con

los procesos estocásticos de especiación y dispersión a lo largo de escalas de tiempo

evolutivas (Hubbell 2001), numerosos estudios indican lo contrario (p. ej. Bascompte 2009).

Ejemplos clásicos de ello son los procesos de coevolución que pueden existir entre diferentes

especies de plantas y sus animales asociados, ya sean polinizadores o herbívoros (Darwin

1859). Rezende et al. (2007) encontraron una señal filogenética clara en redes de

interacciones animal-planta. Sus resultados indican que las interdependencias entre ambos

tipos de organismos pueden llevar a procesos de coextinción cuando una de las especies

desaparece, mostrando un alto grado de dependencia interespecífica que parece extenderse

durante toda la historia evolutiva de las especies que forman las comunidades naturales. Estos

procesos de coevolución han sido demostrados también en las interacciones entre plantas. Por

ejemplo, Callaway y Aschehoug (2000) evaluaron las interacciones entre Centaurea diffusa,

una planta nativa de Asia, e invasora en Estados Unidos, y sus vecinas en ambas regiones. En

este estudio vieron como, al añadir carbón activo para secuestrar los compuestos alelopáticos,

no se encontró ninguna diferencia entre las vecinas asiáticas (que por otro lado eran menos

sensibles a la competencia con C. diffusa). Sin embargo, las vecinas americanas

experimentaron un menor efecto competitivo de C. diffusa al añadir carbón activo, lo que

demuestra que estas especies estaban menos adaptadas a los compuestos alelopáticos. Estos

resultados sugieren que las vecinas asiáticas habían experimentado algún grado de adaptación

a estos compuestos debido a la coexistencia con C. diffusa. Pero sin duda, el mejor ejemplo de

interdependencia entre plantas a lo largo de escalas de tiempo evolutivas lo encontramos en

Valiente-Banuet et al. (2006). Estos autores encontraron que las especies de origen Terciario

(condiciones más húmedas que las actuales) dependen de la presencia de especies originadas

durante el Cuaternario (condiciones más áridas) para mantener su nicho de regeneración en

diversos ecosistemas Mediterráneos. La conclusión de estos autores fue que la mejora

microclimática promovida por las especies del Cuaternario ha sido clave para mantener a las

especies del Terciario en ambientes a los que no estaban adaptadas, lo que indica que las

interacciones planta-planta son fundamentales para mantener la diversidad de las

comunidades naturales a lo largo de escalas de tiempo evolutivas.

El reciente desarrollo de las filogenias moleculares ha permitido a los ecólogos evaluar

el efecto de distintos mecanismos (i.e. interacciones bióticas, factores físicos) en el

ensamblaje de las comunidades a lo largo de estas escalas de tiempo evolutivas (Webb et al.

RESUMEN

24

2002, Cavender-Bares et al. 2009). Si asumimos que muchos rasgos ecológicos están bien

conservados a lo largo de la evolución (Herrera 1984, 1992, Ackerly 2003, Valiente-Banuet et

al. 2006), una dominancia de las interacciones competitivas producirá la coexistencia de

especies con rasgos marcadamente diferentes, que permitan cierta segregación de nichos

ecológicos para su coexistencia; por tanto, el patrón filogenético de esa comunidad será más

disperso o equitativo de lo que sería esperable por azar (Webb et al. 2002). En cambio, si los

filtros abióticos son de mayor importancia para esa comunidad, el conjunto de especies que la

formen se caracterizará por una cierta homogeneidad en sus rasgos ecológicos, que

corresponde con sus adaptaciones morfológicas y fisiológicas a esas condiciones ambientales;

por tanto, el patrón filogenético de esa comunidad será más agregado de lo que sería esperable

por azar (Webb et al. 2002, Pausas y Verdú 2007). Esta asunción sencilla fue el punto de

partida de la explosión de estudios que durante los últimos diez años han tratado de inferir los

mecanismos dominantes en el ensamblaje de una comunidad dada a partir su patrón

filogenético (revisado en Cavender-Bares et al. 2009, Vamosi et al. 2009).

Estudios recientes advierten sobre otros posibles mecanismos que pueden afectar al

patrón filogenético de una comunidad y que antes no habían sido considerados. Por citar

algunos ejemplos, la preferencia de herbívoros o polinizadores por taxones filogenéticamente

relacionados (Webb et al. 2006, Sargent y Ackerly 2008), la escala a la que se realice el

estudio (Cavender-Bares et al. 2006, Kraft et al. 2007, Kraft y Ackerly 2010), diferencias en

el nicho de regeneración o las habilidades competitivas entre las especies que coexisten

(Myfield y Levine 2010), o las interacciones positivas entre plantas (Valiente-Banuet y Verdú

2007, Verdú et al. 2009), son algunos de los mecanismos que pueden afectar a la estructura

filogenética de las comunidades. Es por ello que, para inferir los mecanismos de ensamblaje a

partir de patrones filogenéticos se recomiendan medidas complementarias de otros procesos,

como patrones de co-ocurrencia (indicador de interacciones bióticas positivas y negativas;

Tirado y Pugnaire 2005), variables físicas (filtros abióticos) o la conservación de rasgos

ecológicos importantes a lo largo de la evolución (Cavender-Bares et al. 2009, Pausas y

Verdú 2010). Sin embargo, estudios que incluyan las medidas de estos otros mecanismos y las

posibles interacciones entre ellos, tanto a nivel de comunidad como a nivel de especie, son

aún muy escasos pese a que las interacciones entre algunos de esos procesos son clave para el

ensamblaje de las comunidades semiáridas (Holmgren y Scheffer 2010, Butterfield et al.

2010).

La idea de Darwin (1859) en relación a que las especies más parecidas necesariamente

tenían que competir de una forma más intensa ha permeado en la teoría ecológica durante 150

25

años (Webb et al. 2002, Cahill et al. 2008). Sin embargo, a pesar de sus profundas

implicaciones para la diversidad local y el ensamblaje de las comunidades naturales, esta idea

ha sido pobremente estudiada experimentalmente (Valiente-Banuet et al. 2006, Valiente-

Banuet y Verdú 2007, 2008, Cahill et al. 2008, Castillo et al. 2010). Cahill et al. (2008) no

encontraron ninguna relación entre la distancia filogenética y el efecto de la competencia al

evaluar una base de datos amplia que incluía la relación entre 50 especies objetivo y 92

especies competidoras distintas. Ellos atribuyeron esta falta de relación a la diferencia entre

las interacciones entre mono- y dicotelodóneas, ya que la intensidad de la competencia

aumentaba con la distancia filogenética para las monocotiledóneas, ocurriendo lo contrario

con las dicotiledóneas. Los trabajos de Valiente-Banuet y colaboradores (2006, 2007, 2008) y

Castillo et al. (2010), conducidos en su mayoría con especies dicotiledóneas, concluyen que la

competencia disminuye también con la distancia filogenética entre dos especies, siendo más

probable que se den interacciones positivas entre especies distanciadas en la evolución. Por

tanto, la idea de Darwin parece confirmarse en la mayoría de casos estudiados, al menos para

plantas dicotiledóneas. De estos estudios se concluye, por tanto, que la relación evolutiva es

clave para decidir el resultado de la interacción entre dos especies. Sin embargo, se ha

discutido con anterioridad en este texto, y durante 20 años en la literatura ecológica en

general, que las condiciones ambientales son fundamentales para definir el resultado de estas

interacciones. Entonces, ¿cuál es la importancia relativa de las condiciones ambientales frente

a las relaciones evolutivas a la hora de definir el resultado de las interacciones planta-planta?,

¿interactúan ambos factores a la hora de definir estos resultados? Hasta la fecha ningún

estudio se ha planteado responder a estas preguntas, las cuales son clave para establecer la

importancia de las relaciones planta-planta en el ensamblaje de las comunidades bajo distintas

condiciones ambientales y a lo largo de escalas de tiempo amplias. Asimismo, estudios

enfocados en la interacción entre la distancia filogenética entre las especies implicadas y el

clima en el que se dan estas interacciones nos pueden ayudar a mejorar nuestras inferencias

sobre los procesos reinantes en el ensamblaje de una comunidad a partir del estudio de los

patrones filogenéticos.

RESUMEN

26

27

OBJETIVOS

El objetivo general de esta tesis es evaluar el efecto de distintos niveles de estrés, tanto biótico

como abiótico, en el resultado de las interacciones entre pares de especies vegetales,

estudiando también cómo estas interacciones y los factores climáticos afectan a la diversidad

local de especies y a la estructura filogenética de las comunidades vegetales en medios

semiáridos. Para poder conseguir este objetivo general se han desarrollado siete objetivos

específicos, que se describen a continuación y se abordarán en los cinco capítulos que

conforman el cuerpo de esta tesis doctoral.

OBJETIVOS A NIVEL DE PAR DE ESPECIES

• Evaluar el efecto del cambio en el patrón temporal de las precipitaciones predicho por

diversos modelos de cambio climático en el resultado de la interacción entre Retama

sphaerocarpa (planta facilitada) y distintas especies herbáceas (plantas nodriza) en un

ecosistema natural (espartal dominado por Stipa tenacissima) y uno emergente

(herbazal de talud de carretera; Capítulo 1).

• Evaluar el efecto de la variabilidad espacio-temporal en la disponibilidad de agua

sobre la interacción entre el arbusto gipsófilo Lepidium subulatum (planta facilitada) y

la herbácea perenne S. tenacissima (planta nodriza) a lo largo de diferentes estados

ontogenéticos de L. subulatum (Capítulo 2).

• Determinar el efecto simultáneo de dos factores distintos de estrés (herbivoría y

aridez), así como de su dinámica temporal, en el resultado de la interacción entre R.

sphaerocarpa (planta facilitada) y la herbácea S. tenacissima (planta nodriza; Capítulo

3).

• Evaluar la generalidad de los modelos teóricos existentes para predecir el signo de las

interacciones planta-planta a lo largo de gradientes ambientales en dos regiones

semiáridas contrastadas (Capítulo 4).

RESUMEN

28

• Definir la importancia relativa de las relaciones filogenéticas y las condiciones

ambientales a la hora de definir el signo de las interacciones entre pares de especies

vegetales presentes en espartales de S. tenacissima a lo largo de un gradiente

ambiental amplio (Capítulo 5).

OBJETIVOS A NIVEL DE COMUNIDAD

• Estudiar la importancia relativa de distintos mecanismos de facilitación/competencia

(expansión de nicho, mejora microambiental, competencia intransitiva y segregación

de nicho) y de los factores climáticos, así como la interacción entre ambos, a la hora

de determinar la riqueza local de especies en dos comunidades semiáridas de

características contrastadas a lo largo de gradientes ambientales amplios (Capítulo 4).

• Evaluar la extensión del efecto de las interacciones bióticas, los factores climáticos, y

su interacción, sobre el patrón filogenético en espartales de Stipa tenacissima a lo

largo de un gradiente ambiental amplio (Capítulo 5).

29

METODOLOGÍA GENERAL

Y ÁREA DE ESTUDIO

REA DE ESTUDIO

Salvo dos excepciones (parte de los capítulos 1 y 4), esta tesis doctoral se centra en su

totalidad en los espartales de Stipa tenacissima situados en el centro y sudeste Peninsular.

Este ecosistema es uno de los más representativos de las zonas semiáridas de España y el

Norte de África (LeHoureu 2001). Los espartales se extienden sobre suelos pedregosos,

limosos, arcillosos, calizos o yesosos, en zonas desde el nivel del mar hasta 2000 m de altitud,

y con precipitaciones que pueden llegar hasta los 700 mm, aunque preferentemente se dan en

la franja entre 200-400 mm (revisado en Maestre et al. 2007). Los espartales son formaciones

vegetales abiertas, con coberturas que oscilan entre el 18% y el 60% (Maestre 2002, Ramírez

2006), de estructura y composición heterogéneas (Puigdefábregas y Sánchez 1996,

Puigdefábregas et al. 1999, Maestre et al. 2007). Al igual que otros sistemas semiáridos, la

estructura espacial de la vegetación en los espartales, caracterizada por la presencia de

manchas de vegetación discreta embebidos en una matriz de suelo desprovisto de plantas

vasculares, generan una dinámica fuente-sumidero que resulta clave en la dinámica hídrica y

ecológica de estas comunidades (Puigdefábregas et al. 1999, Maestre y Cortina 2004c,

Ramírez y Bellot 2009).

Los espartales han sido intensamente manejados por el hombre desde hace no menos

de 4000 años, principalmente para la explotación de fibras vegetales (Barber et al. 1997). Sin

embargo, la llegada de las fibras sintéticas y el abandono general del campo que ocurrió en

España a partir de los 1960s, promovió el cese del manejo humano de estos ecosistemas

(Maestre et al. 2007). Este cese ha provocado la recolonización, aunque muy lenta y poco

abundante, de los arbustos rebrotadores típicos de estos climas, que anteriormente eran

eliminados por su posible efecto negativo sobre el crecimiento del esparto (Cortina y Maestre

2005, Maestre et al. 2007, Maestre et al. 2009b). Así, arbustos como Pistacia lentiscus,

Quercus coccifera, Rhamnus lycioides o Ephedra fragilis, entre otros, han aumentado

levemente su cobertura en estos ecosistemas desde el abandono de su explotación (Maestre et

RESUMEN

30

al. 2007). Estos arbustos, a pesar de representar una parte pequeña en cuanto a cobertura en

estos ecosistemas, juegan un papel fundamental, ya que incrementan la heterogeneidad y

diversidad local (Cortina y Maestre 2005, Maestre y Cortina 2005), y afectan positivamente a

la fertilidad y el funcionamiento ecosistémico (Pugnaire et al. 1996a, Caravaca et al. 2003,

Maestre et al. 2009b). A diferencia de estos arbustos, que extienden su sistema radicular no

sólo bajo su dosel, si no también en las áreas de suelo desnudo circundantes, el esparto centra

sus raíces exclusivamente bajo su dosel, y a profundidades inferiores a 40 cm de profundidad,

dependiendo su rendimiento en gran parte de su capacidad para recoger el agua de escorrentía

generada durante los eventos de lluvia (Puigdefábregas et al. 1999; pero véase Ramírez et al.

2007 para una visión alternativa). Esta capacidad de capturar el agua de escorrentía, junto con

los efectos que su dosel produce sobre la reducción de la radiación incidente y la demanda

evaporativa, son claves para entender el efecto positivo que el esparto produce sobre otras

especies y su papel como “islas de recursos” (Maestre et al. 2001, 2003, Gasque y García-

Fayos 2004, Armas y Pugnaire 2005, Barberá et al. 2006, Navarro et al. 2008).

Los capítulos 1, 2 y 3 de esta tesis se centran en espartales del centro de la Península

Ibérica que crecen en suelos gipsícolas. Estos suelos presentan características químicas

(exceso de iones de sulfato o Calcio, baja retención de agua; Meyer 1986, Escudero et al.

1999, 2000) y físicas (costra superficial dura; Romao y Escudero 2005) que hacen que la

colonización vegetal difícil para muchas especies, siendo su composición florística

particularmente abundante en especialistas de estos sustratos (revisado en Caballero 2006,

Matesanz 2008). Esto, junto con la combinación de características climáticas adversas (estos

suelos se desarrollan sobretodo en medios semiáridos) hace que estos suelos yesosos

presenten una gran cantidad de endemismos, adaptados a las condiciones particulares de estos

suelos y climas. Estas características hacen de los ecosistemas Mediterráneos yesíferos

hábitats de interés para su conservación por la particularidad de las especies que los

conforman (Caballero 2006, Matesanz 2008 y referencias en esos textos). Por tanto, en los

espartales escogidos para el desarrollo de los capítulos 1, 2 y 3 pueden encontrarse especies

propias de los matorrales gipsófilos (especialistas de suelos yesíferos; Helianthemum

squamatum, Lepidium subulatum, Centaurea hyssopifolia) y “gipsovags” (generalistas que

pueden vivir en suelos yesíferos; Retama sphaerocarpa, Rosmarinus officinalis o Thymus

vulgaris). En cambio, los capítulos 4 y 5 se centran en los espartales presentes en los suelos

calcáreos desde Guadalajara a Murcia (ver Fig. A5). La amplia zona de distribución de estos

espartales sobre un suelo relativamente homogéneo hace posible la realización de

experimentos observacionales a lo largo de un gradiente climático amplio, que oscila entre

31

13-17 ºC de temperatura media, y 273-488 mm de precipitación media anual, con mucha

mayor cantidad y frecuencia de heladas en el extremo occidental (Guadalajara-Madrid) que en

el oriental (Alicante-Murcia) del gradiente.

Alternativamente, parte de los experimentos de esta tesis doctoral (Capítulo 1) se han

desarrollado sobre herbazales de talud de carretera, un tipo de ecosistema emergente que está

aumentando su importancia a nivel global año a año (García-Palacios et al. 2010). Estos

sistemas se caracterizan por una cobertura herbácea dominada por especies anuales,

generalmente ruderales que presentarán mayor o menor cobertura dependiendo del tipo de

talud (desmonte o terraplén; Matesanz et al. 2006) o de la disponibilidad de agua durante la

germinación y desarrollo de las plántulas (Bochet y García-Fayos 2004; ver García-Palacios

2010 para una revisión extensa sobre la dinámica de estos sistemas emergentes). En estos

herbazales, es común que la sucesión secundaria se vea ralentizada, bien por la escasa

disponibilidad de agua o nutrientes (Bolling y Walker 2000), o bien porque coberturas

herbáceas muy desarrolladas no dejan huecos libres para la colonización de nuevas especies

(Burke y Grime 1996). La introducción de especies leñosas en estos sistemas ha sido

recomendada para acelerar la sucesión secundaría (Booth et al. 1999); sin embargo, sabemos

muy poco sobre como las herbáceas dominantes en este tipo de sistemas afectan al éxito de

estas plantaciones, especialmente bajo diferentes niveles de disponibilidad hídrica. Un mejor

entendimiento de la interacción entre ambos grupos vegetales (herbáceas y leñosas) nos

permitirá conciliar los dos mayores retos en la restauración de estos taludes de carretera: el

establecimiento de una cobertura herbácea suficientemente densa como para prevenir

procesos de erosión (Andrés y Jorba 2000), y el establecimiento de especies leñosas para

acelerar su sucesión secundaria (Jorba y Vallejo 2008).

Por otro lado, en el Capítulo 4 se evalúa el efecto de las interacciones bióticas y el

clima en la diversidad local, no sólo de espartales, si no de ecosistemas semiáridos

australianos. Concretamente se han muestreado comunidades pertenecientes a las alianza de

Eucalyptus populnea y Callitris glaucophylla y de Casuarina pauper y Alectyron oleifolius

(Beadle 1948), comúnmente encontradas en las llamadas “tierras rojas” de este continente.

Estas tierras rojas se caracterizan por tener textura arenosa, suelos profundos y con contenidos

bajos en nutrientes (Isbell 1996). Estos ecosistemas se caracterizaban originalmente por

presentar coberturas herbáceas continuas, con individuos dispersos de Eucalyptus spp. y

algunos arbustos. Sin embargo, la elevada presión ganadera a la que han sido sometidas estas

áreas (Keith 1998) ha incrementado notablemente el reclutamiento de estos arbustos (p. ej.

Geijera parviflora, Eremophylla spp., Callitris glaucophylla, etc.) y ha reducido la cobertura

RESUMEN

32

herbácea. Lo que finalmente ha generado una estructura discontinua, de manchas discretas de

vegetación embebidas en una matriz de suelo desnudo, equivalentes a las de los espartales

anteriormente descritos (Tongway y Hindley 1995). Estas comunidades ocupan una amplia

superficie en el este de Australia, que permitió seleccionar 10 parcelas con vegetación

perteneciente a estas comunidades a lo largo de un gradiente climático amplio (16º–19º C y

280–630 mm) con el objetivo de complementar el muestreo realizado en España para el

Capítulo 4 (descrito con más detalle en la siguiente sección).

Figura A5. Distribución aproximada de las parcelas utilizadas a lo largo del gradiente ambiental referido en los capítulos 4 y 5. El diagrama ombroclimático (Fuente: www.globalbioclimatics.org) y una imagen general de los dos extremos del gradiente junto a una posición intermedia aparecen mostrados en la figura.

El clima de todas las áreas de estudio escogidas para esta tesis es Mediterráneo

semiárido, difiriendo en su grado de continentalidad y en sus precipitaciones medias. En las

parcelas situadas en el centro Peninsular, las temperaturas en invierno son mucho más frías y

las precipitaciones anuales son algo mayores. La lluvia sigue una distribución bimodal, con

máximos marcados en primavera (Abril-Mayo) y otoño (Septiembre-Octubre). La

variabilidad interanual de la precipitación es muy grande, siendo impredecibles tanto la

33

cantidad como el patrón temporal de las lluvias de cada año (revisado en Puigdefábregas et al.

1999; ver figuras con precipitaciones registradas durante el período de estudio en los capítulos

1 y 3). En las parcelas situadas en el sudeste peninsular, las temperaturas son más suaves en

invierno y las precipitaciones medias algo menores. La distribución temporal de los eventos

de lluvia en estas parcelas es más unimodal, centrándose las precipitaciones hacia finales del

verano y comienzos del otoño (Fig. A5). Las precipitaciones en el gradiente australiano

siguen una distribución relativamente homogénea a lo largo del año, con una ligera

predominancia de eventos de lluvia durante el verano (el 60% de estos eventos se concentra

en verano).

METODOLOGÍA GENERAL

En esta tesis doctoral se han realizado experimentos tanto manipulativos como

observacionales, en los que se incluyen distintos niveles de estrés abiótico (generalmente

aridez) y biótico (herbivoría). Todos ellos han sido llevados a cabo bajo condiciones

naturales. Los diferentes niveles de aridez se han conseguido mediante riegos en los

experimentos manipulativos, o mediante el uso de gradientes ambientales amplios en las

aproximaciones observacionales. En éstas últimas se han homogeneizado, tanto como ha sido

posible, la pendiente, orientación, tipo de suelo e historia de manejo previo de las parcelas

seleccionadas, con la intención de evitar la influencia de otros factores que no estuvieran

considerados en el experimento y que pudieran confundir la interpretación de sus resultados.

Los únicos herbívoros considerados han sido los conejos, que eran especialmente abundantes

en las parcelas seleccionadas (véanse los tres primeros capítulos). Su nivel de herbivoría ha

sido controlado mediante mallas de exclusión, combinado con el seguimiento de plantones no

protegidos a lo largo de un año. Los experimentos manipulativos se centran en interacciones a

nivel de par de especies. Las aproximaciones observacionales, en cambio, han sido utilizadas

para evaluar el efecto de estas interacciones tanto a nivel de especie como a nivel de

comunidad. La complejidad de las interacciones bióticas se ha evaluado mediante análisis de

co-ocurrencia (Gotelli 2000) o utilizando distintos indicadores del desarrollo vegetal

(crecimiento, supervivencia y eficiencia fotosintética en los manipulativos, crecimiento o

cobertura en los observacionales) en individuos que crecían bajo el dosel de la planta nodriza

escogida y en áreas libres de vegetación para cada estudio. La metodología utilizada en cada

caso se detalla en la sección dedicada a ese fin en los distintos capítulos de esta tesis doctoral

(ver siguiente sección).

RESUMEN

34

35

ESTRUCTURA GENERAL DE LA TESIS

Los cinco capítulos que conforman el cuerpo de esta tesis han sido escritos en inglés, para su

publicación en revistas científicas de ámbito internacional. A continuación se da una breve

descripción de estos capítulos y de la metodología utilizada en cada uno de ellos.

Capítulo 1. Las modificaciones en el régimen de lluvias predichas con el cambio

climático modulan las interacciones herbácea-arbusto en dos comunidades

semiáridas.

Pese al gran interés que ha suscitado en los últimos 20 años la relación entre las interacciones

planta-planta y el nivel de estrés, muy pocos estudios han evaluado cómo los cambios en el

patrón temporal de las precipitaciones dentro del mismo año afectan al resultado de las

interacciones planta-planta. Este patrón temporal es incluso más importante que la cantidad de

lluvia que cae durante un año determinado para el funcionamiento de los medios semiáridos.

El objetivo de este capítulo es evaluar el efecto de las modificaciones en la abundancia y

frecuencia de los eventos de lluvia predichos con el cambio climático en la interacción entre

plantones de R. sphaerocarpa y diversas especies herbáceas. Para ello aumentamos

experimentalmente la cantidad de agua disponible entre abril y julio, imitando a la inversa la

reducción de las precipitaciones durante este período (el tratamiento control, sin riego, sería el

futuro escenario de cambio climático y los tratamientos de riego serían los escenarios

actuales). El riego fue distribuido en dos o cuatro pulsos de lluvia, imitando el incremento en

la frecuencia de eventos torrenciales. El estudio se llevó a cabo durante tres años en dos

sistemas marcadamente diferentes situados sobre sustratos ricos en yeso: un espartal y un

herbazal de terraplén de carretera, dominados por S. tenacissima y diversas anuales nitrófilas,

respectivamente. Para evaluar el efecto de las herbáceas sobre R. sphaerocarpa, se midió la

supervivencia, crecimiento y eficiencia fotosintética durante estos tres años. Los objetivos de

este capítulo eran evaluar las diferencias entre 1) diferentes vecinas herbáceas, 2) el efecto de

las vecinas a lo largo de pulsos e interpulsos y 3) el cambio en el efecto de las vecinas bajo

diferentes abundancias o frecuencias de eventos lluviosos.

RESUMEN

36

Capítulo 2. La heterogeneidad espacio-temporal en los factores abióticos modula los

cambios entre competencia y facilitación que ocurren a lo largo de la ontogenia.

Las interacciones planta-planta están determinadas, en parte, por las condiciones ambientales

y la ontogenia de las especies implicadas. Pese a que el efecto de ambos factores en el

resultado de estas interacciones ha sido evaluado de forma separada, muy pocos estudios han

investigado su efecto conjunto. El objetivo de este capítulo era testar este efecto, así como su

variabilidad espacio-temporal. Para ello se evaluó la interacción entre el arbusto gipsófilo

Lepidium subulatum y la herbácea perenne S. tenacissima en tres zonas del centro Peninsular.

Dentro de estas tres zonas evaluamos el resultado neto de la interacción entre ambas especies

usando análisis de co-ocurrencia en laderas con orientación norte (menos estrés) y sur (más

estrés). En una de las tres zonas (Aranjuez) se evaluaron cambios en el signo de la interacción

a lo largo de distintas etapas ontogenéticas de L. subulatum en las dos orientaciones mediante

una combinación de experimentos de siembra y medidas dendrocronológicas, de floración y

de acumulación de carbohidratos. Este conjunto de técnicas nos permitió estudiar en detalle el

efecto de la ontogenia y de la variabilidad espacial en la disponibilidad hídrica, así como su

interacción en la relación entre L. subulatum y S. tenacissima. Las hipótesis principales

fueron: 1) la interacción entre ambas especies pasará de fuertemente positiva a fuertemente

negativa a lo largo del desarrollo de L. subulatum; 2) dado que ambas especies son tolerantes

al estrés, es de esperar que las interacciones positivas dominen a niveles intermedios de estrés

hídrico; y 3) un mayor nivel de estrés hídrico debería reducir el efecto negativo de S.

tenacissima en individuos adultos de L. subulatum.

Capítulo 3. Las dinámicas temporales de la herbivoría y la disponibilidad hídrica

interactúan modulando el resultado de una interacción herbácea-arbusto en un ecosistema

semiárido.

La herbivoría y la aridez son dos factores de estrés que comúnmente coinciden en los medios

semiáridos, jugando ambos un papel fundamental en el resultado de las interacciones planta-

planta. Sin embargo, su efecto conjunto ha sido pobremente estudiado. En este capítulo se

estudia el efecto conjunto de ambos tipos de estrés, y su variabilidad temporal a lo largo del

año, en el resultado de la interacción entre S. tenacissima y plantones del arbusto R.

sphaerocarpa. Para ello se utilizó una combinación de aproximaciones observacionales y

37

experimentales. Estas aproximaciones incluyen el seguimiento de la supervivencia y el nivel

de daño provocado por conejo en plantones sin protección a lo largo de un año, y el efecto de

la manipulación del nivel de aridez y el daño por herbivoría en el signo de la interacción entre

ambas especies. Las hipótesis iniciales de este estudio fueron: 1) S. tenacissima reducirá el

daño por conejo sobre R. sphaerocarpa mediante su papel de ocultadora frente a los

herbívoros; 2) las mejores condiciones hídricas, pero menores niveles de luz, reducirán la

tolerancia de R. sphaerocarpa bajo el dosel de S. tenacissima; y 3) el efecto conjunto de

niveles altos de ambos tipos de estrés (herbivoría y aridez) anularán el efecto positivo

derivado de la protección y la mejora microclimática de S. tenacissima.

Capitulo 4. Sobre la importancia relativa del clima y las interacciones bióticas no tróficas

como determinantes de la riqueza local de especies vegetales.

En este capítulo se pretende evaluar el efecto relativo de las condiciones climáticas y distintos

componentes de las interacciones bióticas (expansión de nicho y efecto sobre la dinámica

competitiva de las especies vecinas –intransitividad en la competencia o segregación de

nicho), así como la interacción entre ambos factores, a la hora de definir la riqueza local de

especies vegetales en comunidades semiáridas a lo largo de gradientes ambientales amplios.

Para ello se llevo a cabo un estudio observacional a diversas escalas en comunidades

semiáridas de España y Australia. Se evaluó el efecto de ocho variables climáticas, resumidas

mediante análisis de componentes principales, sobre la riqueza específica de cada localidad,

así como sobre diversos indicadores del signo, intensidad e importancia de las interacciones

bióticas a nivel de especie y de comunidad. Además, se estudió el efecto de estas

interacciones sobre la riqueza específica, y la variación de dicho efecto a lo largo del

gradiente climático escogido en cada región. Se midió la intensidad y la importancia de estas

interacciones, tanto a nivel de par de especies como de comunidad, utilizando índices

disponibles en la literatura. La frecuencia de las interacciones facilitativas a nivel de

comunidad fue cuantificada calculando el porcentaje de especies, con respecto al número total

de especies en la localidad, que se desarrollaron mejor (beneficiarias), o dependían

directamente (obligadas) de la presencia de dos especies nodrizas distintas en cada región. La

expansión de nicho fue evaluada mediante el número de especies obligadas, y también

utilizando un índice de similaridad entre las poblaciones de los microambientes nodriza (bajo

una de las dos especies nodriza seleccionadas en cada región) y claro (en suelo desnudo). Los

cambios en la dinámica competitiva de las vecinas fueron analizados mediante una

RESUMEN

38

aproximación observacional a escala de mancha, y utilizando modelos nulos de estructura

agrupada (“guild-structure null models”) para medir los patrones de co-ocurrencia. Nuestras

hipótesis iniciales predecían una igual importancia de las interacciones bióticas para la

diversidad a lo largo de los gradientes ambientales, provocada porque la expansión de nicho

afecta al mismo número de especies, aunque su identidad cambie, a lo largo de estos

gradientes. Alternativamente, proponemos que los cambios en la dinámica competitiva

(segregación de nicho o competencia intransitiva), junto con el aumento de la riqueza de

especies debido a la expansión de nicho, conducen a efectos positivos desproporcionados

sobre la riqueza local de especies debido a procesos de retroalimentación positiva.

Capitulo 5. Sobre la importancia relativa de las condiciones ambientales, las interacciones

bióticas y las relaciones evolutivas como determinantes de la estructura de las comunidades

semiáridas.

En este capítulo se evalúa el efecto relativo de las condiciones ambientales y las interacciones

bióticas (competencia/facilitación a nivel de comunidad) sobre el patrón filogenético de

espartales semiáridos dominados por S. tenacissima. En el mismo se estudia también cómo

interactúan estos dos factores a la hora de determinar dicho patrón, así como la variación en

sus efectos relativos a lo largo de un gradiente ambiental amplio. Para ello se utilizan parte de

las zonas de estudio e indicadores de interacciones bióticas descritos para el capítulo 4, junto

con medidas de co-ocurrencia a nivel de parcela. También se determinó el patrón filogenético

de cada una de las localidades muestreadas y se evaluó el efecto sobre este patrón de los

distintos mecanismos derivados de las condiciones climáticas, las interacciones bióticas, y la

interacción entre ambos factores, mediante regresiones lineares y correlaciones parciales.

Asimismo, se analizó el efecto de la distancia filogenética entre la planta facilitada y su

nodriza, de las condiciones climáticas de cada lugar, y de la interacción entre ambos factores,

como moduladores de la relación entre un total de 200 pares de especies a lo largo de este

gradiente ambiental mediante árboles de regresión. Las hipótesis principales de este capítulo

fueron: 1) la importancia relativa de las condiciones climáticas y las interacciones bióticas

para el ensamblaje de las comunidades, y por tanto para su patrón filogenético, varía a lo

largo del gradiente ambiental, y 2) la distancia filogenética y las condiciones climáticas

interactúan a la hora de definir el signo de las interacciones planta-planta.

Santiago Soliveres, Fernando T. Maestre, Pablo García-Palacios,

Adrián Escudero, and Fernando Valladares.

Manuscrito inédito

1

Predicted climate change effects in rainfall regime modulate the

outcome of grass-shrub interactions in two semi-ari d communities

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

40

CHAPTER 1

41

ABSTRACT

Much research has been devoted to understand how plant-plant interactions behave along water stress gradients in drylands. However, few studies have evaluated how changes in the magnitude and frequency of rainfall events, which are an important component of ongoing climate change, modulate the outcome of such interactions. We evaluated the response of the interaction between seedlings of the shrub Retama sphaerocarpa (L.) Boiss., our target plant, and different herbaceous neighbours to those changes in rainfall availability during three years. The experiment was conducted in natural and anthropogenic grasslands dominated by a perennial stress-tolerator and ruderal annual species, respectively. Competition between herbaceous plants and Retama seedlings prevailed, and increased with water stress. These negative effects were reduced through time, suggesting niche segregation between the interacting plants. Less frequent, but more intense rainfall events, accelerated this niche segregation in the natural grassland, where the stress-tolerator grass took more advantage of light rainfall events than Retama and competition was stronger. Thus, increases in the frequency of heavy rains could counteract the negative effects of the increased competition between grasses and shrubs expected under higher water stress conditions. However, in the anthropogenic grassland the phenology of the annuals made more frequent and lighter rainfall events more useful to avoid competition by water, being heavy rains uneffective in this case. Our findings suggest the existence of a trade-off between the shade tolerance of protégée plants and the effects of nurses on light and water availability that defines the outcome of a given plant-plant interaction. Our results challenge current predictions on the outcome of these interactions under climate change, and could be used to further refine our forecasts on how plant-plant interactions, and therefore plant communities, will respond to such change in ecosystems where grass-shrubs interactions are prevalent.

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

42

INTRODUCTION

he study of plant-plant interaction

dynamics along water stress

gradients has been a major topic

in dryland ecology during the last decade

(e.g. Pugnaire and Luque 2001, Maestre

and Cortina 2004a). The seminal “stress

gradient hypothesis” (SGH), which predicts

an increase in the frequency of facilitative

interactions as abiotic stress increases

(Bertness and Callaway 1994), constitutes a

paradigmatic framework for these studies

(see Callaway 2007 and Brooker et al. 2008

for reviews). However, the generality of its

predictions has been recently debated

(Maestre et al. 2005, 2006, Lortie and

Callaway 2006, Callaway 2007, Maestre et

al. 2009a, Smit et al. 2009, Malkinson and

Tielbörger 2010). Furthermore, the

relationship between plant-plant

interactions and environmental gradients is

especially complex in arid and semiarid

environments, where water availability is

highly pulsed, with erratic and typically

short periods of enough water availability

triggering ecosystem processes

(Schwinning and Sala 2004). Indeed, the

size and frequency of individual rain events

registered in a given period may have more

importance for the functioning of semiarid

ecosystems than the total rainfall

accumulated (Whitford 2002, Ogle and

Reynolds 2004). These temporal dynamics

should be considered when studying plant-

plant interactions and community dynamics

in water-limited ecosystems (Goldberg and

Novoplansky 1997, De la Cruz et al. 2008).

Understanding community responses to

water pulses in water limited ecosystems is

particularly timing because current rainfall

events are likely to be extremer as a

consequence of the ongoing climate change

(Knapp et al. 2008). For example, forecasts

for the Mediterranean Basin predict a

decrease in the amount of annual rainfall,

the lengthening of drought periods and the

increase of the frequency of heavy storms

(IPCC 2007). Grass-shrub interactions are

particularly instructive for studying the

effects of these climatic changes on the

outcome of plant-plant interactions because

of their contrasted water acquisition

strategies and their abundance in natural

and anthropogenic ecosystems worldwide

(e.g. Sala et al. 1989, Scholes and Archer

1997). While grasses tend to use more

efficiently the water derived from light and

sparse rainfall events (Sala et al. 1989,

Reynolds et al. 2004), shrubs generally

perform better after continuous rains, which

recharge deeper soil profiles (Sala et al.

1989, Schwinning and Ehleringer 2001,

Schwinning and Sala 2004). These grass-

shrub interactions are likely to be

particularly sensitive to the changes in

T

CHAPTER 1

43

overall water availability and frequency of

rainfall events predicted under future

climate change scenarios. For example,

shrub encroachment may be promoted by

the increase in the frequency of heavy

storms (Ogle and Reynolds 2004), or

reduced by the enlargement of summer

drought (López et al. 2008), two features of

the ongoing climate change. Alternatively,

grasses may foster shrub survival and

recruitment under moderate drought

conditions, but these positive effects may

be overcome by the increase in competition

registered under periods of very low water

availability (Kitzberger et al. 2000, Maestre

and Cortina 2004a, 2004b). The possible

responses of these interactions to changes

in rainfall amount and frequency seem very

complex and at times counter-intuitive, and

are further complicated by the fact that the

multiple responses described above are

likely to occur simultaneously.

A particular case of grass-shrub

interaction, and also farily common in

semiarid environments worldwide, occurs

when annuals act as nurse plants for shrubs

(e.g. Holzapfel and Mahall 1999 and

references therein). These annual-shrub

interactions are likely to behave

differentially along environmental gradients

because of the different life-strategy of the

former. However, they have been largely

ignored in the facilitation literature, as the

bulk of studies on grass-shrub interactions

have evaluated the effects of shrubs on

annuals (see Callaway 2007 for a review).

Thus, studies directed to clarify the effects

of grasses (both perennials and annuals) on

shrubs across realistic water stress gradients

are of crucial importance to understand

plant community responses to ongoing

climate change in those ecosystems where

grasses and shrubs coexist. In this study we

aimed to test the response of grass-shrub

interactions to changes in the degree of

abiotic stress. The study was conducted in

two different semiarid Mediterranean

communities: a natural Stipa tenacissima L.

steppe (hereafter called “natural grassland”)

and an annual-dominated grassland located

in a motorway embankment (hereafter

called “anthropogenic grassland”).

Degraded landscapes such as the latter

represent good examples of novel

ecosystems, which are increasing in

importance worldwide because of the rise

in anthropogenic disturbances (Hobbs et al.

2006). Given that the structure and

functioning of these novel ecosystems often

differs from that found in natural ones

(Hobbs et al. 2006), studies focusing on

both natural and novel ecosystems may

help to further refine our predictions of the

response of plant communities to climate

change (Brooker 2006). Our focal species

were the leguminous shrub Retama

sphaerocarpa (L.) Boiss. (hereafter

Retama) and different grass species (the

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

44

stress-tolerator tussock grass Stipa

tenacissima –hereafter Stipa– and several

ruderal annual species), the latter acting as

potential nurse plants for Retama.

The outcome of grass-shrub

interactions in semiarid ecosystems, either

natural or anthropogenic, depends strongly

on soil water availability (Eliason and Allen

1997, Maestre and Cortina 2004a), and

therefore, both grassland-types are excellent

study systems to test changes in the

outcome of grass-shrub interactions across

abiotic stress gradients. We obtained a

realistic water stress gradient by modifying

the amount and timing of water availability

separately, according to the most likely

climate change scenarios for the study area.

We tested the following hypotheses: i) both

Stipa and annual grasses will reduce

Retama growth because of competition by

water during wet seasons, when most of

vegetation growth concentrates in semiarid

environments (Goldberg and Novoplansky

1997, Escudero et al. 1999); ii) Stipa and

annuals will differ in their effect on Retama

survival. Although both Stipa and annuals

will increase Retama survival during

summer drought, mainly via microclimate

amelioration and improvement of soil

properties (Goldberg and Novoplansky

1997, Maestre et al. 2003), these effects

will change with the lengthening of summer

drought. Under these conditions, the

positive effects of annuals will be more

intense because they will be mainly derived

from the shade produced by their dry

tissues (annuals die during summer), but

Stipa will reduce Retama survival because

of the increased competition by water will

outweight the positive environmental

buffering promoted by shade (Maestre and

Cortina 2004a); and iii) the increase in the

frequency of heavy storms will reduce

competition between Retama and both Stipa

and annual grasses by recharging deeper

soil layers and promoting niche segregation

(Sala et al. 1989, Ogle and Reynolds 2004).

METHODS

STUDY AREA

Both the natural and anthropogenic

grasslands selected for this study are

located in the center of the Iberian

Peninsula (natural grassland: 40º03´60´´N,

3º54´91´´W, 545 m.a.s.l.; anthropogenic

grassland: 52º16´00´´N, 3º43´13´´W, 604

m.a.s.l.) and share the same climate and soil

type. The climate is semiarid

Mediterranean, with average annual

precipitation and temperature of 388 mm

and 15 ºC, respectively, and with a strong

summer drought (Aranjuez weather station;

1994-2005 period; Marqués et al. 2008).

Both communities are located on gypsum-

rich soils, classified as Xeric Haplogypsid

(Marqués et al. 2008), although in the

anthropogenic grassland the original

CHAPTER 1

45

substrate has been altered by the mixture

with gravels and components from external

sources during the construction of the

motorway. Vegetation in the natural

grassland is an open steppe dominated by

Stipa tenacissima, with a perennial plant

cover of 24%. Vegetation in the

anthropogenic grassland is dominated by

annuals, with a mean cover of 75% and

with Bromus rubens L., B. diandrus Roth.,

and Medicago sativa L. as the most

abundant species (16, 14 and 14% of the

total cover, respectively; García-Palacios et

al. 2010). Hereafter we refer to those

herbaceous annuals as annual grasses for

simplicity. Both study sites hold a high

density of rabbits (Oryctolagus cunniculus

L.), as suggested by visual contacts, and by

the high number of warrens and latrines

found (S. Soliveres, pers. obs.).

EXPERIMENTAL DESIGN

In December 2006, 176 two-year old

Retama seedlings, with a mean height of of

27 ± 2 cm, were planted in each site by

using manually-dug holes of 20×20×20 cm.

These seedlings came from a nursery in

central Spain (viveros Bárbol, Madrid). We

randomly assigned these seedlings to two

different microsites: “Nurse” and “Open”.

Because of the heterogeneous patch-

interpatch structure of the natural steppe

system, and the homogenous herbaceous

cover of the anthropogenic site, these

microsite types were defined differently in

the two ecosystems. Nurse microsites were:

1) located upslope and adjacent to Stipa

tussocks (stress tolerator nurse) of ca. 1 m

width (< 15 cm from the edge of the north

face of the tussock, where facilitative

effects of this species on target shrubs have

been found, e.g. Maestre et al. 2003) in the

natural grassland, and 2) located in the

center of a multi-specific 50-cm diameter

grass patch (ruderal nurse) of ca. 40 cm

height and 75-100% cover (in spring) in the

anthropogenic grassland. The rest of the

seedlings were assigned to randomly

selected Open microsites. These were either

located in bare ground areas at least 80 cm

away from any perennial plant (natural

grassland site), or placed in sites where all

aboveground vegetation within 80-cm

diameter circles surrounding the target

seedling was monthly clipped

(anthropogenic grassland site).

The seedlings were randomly allocated

to establish two full factorial experiments

with two factors each, which were run in

parallel in the two grasslands studied. In the

first experiment (hereafter Experiment 1)

the two factors were Microsite (Nurse vs.

Open) and Irrigation. This irrigation

treatment consisted in three different

amounts of water applied during eight

watering events (once every month between

April and July in both 2007 and 2008). In

each watering event, the 0%, 25% and 50%

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

46

of the median from the last 30 years for that

particular month was applied to the control

(hereafter +0%), amount 1 (hereafter

+25%), and amount 2 (hereafter +50%)

treatments, respectively. In the second

experiment (hereafter Experiment 2) the

factors were i) Microsite (Nurse vs. Open),

and ii) Irrigation frequency. In this case we

applied the same amount of water

employed in the +50% treatment, but with

two different frequencies: two and four

pulses (hereafter named as 2x and 4x

treatments, respectively). The first

frequency treatment (4x) was applied in

four monthly pulses, from April to July, as

explained above; the second frequency

treatment (2x) was applied every two

months, in May and July. In each of the

watering events applied in the 2x treatment,

the amounts of water added equaled the

sum of April and May, and June and July

irrigations applied in the 4x treatment,

respectively (Fig. 1.1, Table 1.1). Both

irrigation treatments were applied

irrespectively of the rainfall registered in

each month (Fig. 1.1). With these two

experiments we aimed to evaluate the

effects on the outcome of the interaction

studied of: i) an increase in summer drought

(summer drought was longer in less-

watered plants [+0% > +25% > +50%]

because of the low rainfall levels typically

registered during June and July in the study

area), ii) a reduction of rainfall during the

wet season (spring rainfall was less

abundant in non-watered plants [+0% <

+25% < +50%]), and iii) changes in the

frequency of heavy showers (magnitude of

individual events was higher in the 2x than

in the 4x treatment, despite both treatments

received the same amount of water, Fig.

1.1). All these effects, i.e. the increase in

summer drought, the reduction in the total

rainfall amount and the increase in the

frequency of heavy showers are predicted

by future climate change scenarios for the

Mediterranean Basin (IPCC, 2007; see

Table 1.1). Because of the high density of

rabbits observed, and to avoid seedling

predation, the seedlings were protected

from browsing by using a thin-wire mesh.

This mesh did not shade the seedlings, and

thus did not confound the effects of any of

the factors studied in the experiment.

CHAPTER 1

47

Figure 1.1 Climatic data (mean monthly temperature, black circles; and monthly rainfall, black bars) obtained from a meteorological station (Onset, Pocasset, MA, USA) located in the natural grassland. The increment in monthly rainfall promoted by the irrigation treatments applied during 2007 and 2008 is represented by different colors: +25% (dark grey) = irrigation of 25% of the median of April-July rainfall in four pulses, +50 -4x (light grey) = irrigation of 50% of the median of April-July rainfall in four pulses, and +50 -2x (white) = irrigation of 50% of the median of April-July rainfall in two pulses).

Table 1.1 Details of the water amount added (l · m-2) in each irrigation treatment (Experiments 1 and 2) and the periods when these pulses took place. Last row explains the specific climate change effect that each irrigation treatment emulated.

Irrigation

Pulses

EXPERIMENT 1:

Amount treatment

EXPERIMENT 2:

Frequency treatment

+0% +25% +50% 4x 2x

April 0 6 12 12

May 0 9.25 18.5 18.5 30.7

June 0 10.85 21.7 21.7

July 0 5.65 11.3 11.3 33 Total water

applied 0 31.75 63.5 63.5 63.5

Simulated

climate change

effect

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

48

MONITORING OF SOIL MOSITURE

AND PLANT PERFORMANCE

Soil moisture was measured by time-

domain reflectometry (TDR; Topp and

Davis 1985) using a Campbell TDR100

system (Campbell Scientific Ltd,

Loughborough, UK). In 10 randomly

selected planting holes per each of the eight

possible treatments combinations, 10 cm

long TDR probes were vertically installed

(n = 80). This soil depth was chosen

because we expected the concentration of

most of the roots of the interacting species

to be in the upper part of the soil during the

study period (herbs and young woody

seedlings should concentrate their roots in

the upper soil layers; Scholes and Archer,

1997). A strong relationship between TDR

values and soil gravimetric moisture has

been found in the natural grassland (R2 =

0.84; P < 0.0001; Soliveres et al. unpubl.

data); thus, this measurement can be

considered as a good proxy for soil

moisture availability. Soil moisture was

measured every two months during the

study period, starting and ending in April

2007 and September 2009, respectively. As

summer drought is considered the

bottleneck for plant recruitment in semiarid

Mediterranean environments (Escudero et

al. 1999), this sampling was conducted

monthly during the summer (June-

September).

As the magnitude of the effect of Stipa

on woody seedlings follows seasonal

dynamics of water availability (Maestre et

al. 2003), the effects of herbaceous

neighbours on soil moisture availability

were tested considering different dry/wet

periods. To evaluate how the irrigation

treatments and the presence of grasses

affected soil moisture, the relative

interaction index (RII; Armas et al. 2004)

was calculated for each sampling date as:

(TDRNu – TDROp)( (TDRNu + TDROp),

where TDRNu and TDROp are soil moisture

data obtained in Nurse and Open

microsites, respectively. To aid interpreting

our results, RII data were grouped in

intervals with percentages in soil moisture

above and below 10% (wet and dry

seasons, respectively). This limit

corresponds to the natural seasonal

dynamics in water availability; periods

below and above 10% moisture occur

mainly during summer drought and when

plant activity concentrates, respectively

(Schwinning and Sala 2004, Reynolds et al.

2004). We randomly paired the samples by

microsite (Nurse vs. Open) and between the

four irrigation treatments obtained with the

two experiments (+0%, +25%, +50%-4x,

and +50%-2x; 10 pairs for each microsite ×

irrigation combination). Then, RII data of

these paired samples were grouped for wet

and dry seasons separately. With these data,

the average RII for all wet and dry seasons

was obtained for each of the four irrigation

CHAPTER 1

49

treatments.

The outcome of plant-plant interactions

varies widely depending on the

performance measure used (Goldberg and

Novoplansky 1997, Maestre et al. 2005).

Thus, as recommended when studying these

interactions along stress gradients (Brooker

et al. 2008), we used several performance

measurements to test the effect of

neighbour grasses on Retama seedlings.

Seedling height, root collar diameter and

survival were measured after each summer,

in September 2007, 2008 and 2009. A

strong relationship between standing

biomass and root collar diameter has been

previously observed for Retama seedlings

(R2 = 0.823, P < 0.0001; Soliveres et al.

unpubl. data), so this measure was used as

our surrogate of seedling biomass in the

field. The relative growth rate (hereafter

RGR) of Retama seedlings for each year

was obtained as: ln RCD1 – ln RCD0)/(T1 –

To); being RCD1 and RCD0 the root collar

diameter at the precedent (T0) and current

(T1) sampling date, respectively. The

slenderness coefficient, which is considered

a good indicator of light competition

(Kurashige and Agrawal 2005), was also

calculated as the ratio between height and

root collar diameter lnv H/ln RCD. Light

competition can be an important factor

affecting plant-plant interactions, even in

water-limited environments (Seifan et al.

2010a, Soliveres et al. 2010). Therefore,

measuring it is important to understand its

relative role in the interactions studied.

Both plant growth and survival are

integrative measurements that include the

effects of microsite or irrigation on plant

performance during the whole year.

Therefore, they are not appropriate to detect

seasonal differences in the effect of the

assayed treatments on seedling

performance. To assess for such

differences, we estimated the potential

photochemical efficiency of Retama

seedlings by measuring the maximum

quantum yield of PSII (Fv/Fm) of dark-

adapted leaves (30 min, at midday) with a

pulse-modulated fluorometer (FMS2,

Hansatech Instruments, Norfolk, UK). This

parameter has been widely used as an

indicator of plant stress in semiarid regions

(e.g. Pugnaire et al. 1996b, Maestre et al.

2003, Aragón et al. 2008); small changes in

photochemical efficiency have been

associated with water limitations during

important stages of the lifecycle of woody

plants in these environments (Aragón et al.

2008), and have been found to match

results with other performance measures

(e.g. survival) when evaluating the outcome

of plant-plant interactions (e.g. Maestre et

al. 2003, 2004). Six seedlings per

combination of treatments and grassland

type were randomly selected for these

measurements (n = 48 per grassland).

Different randomly selected plants were

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

50

measured in each sampling period, which

took place on the same dates than TDR

surveys (see above). To test for differences

in the intensity of biotic interactions in

response to the treatments evaluated within

dry or wet seasons, the RII index was

calculated using Fv/Fm data as described

above for the TDR data.

STATISTICAL ANALYSES

RII data (calculated from both TDR and

Fv/Fm measurements) were organized by

dry/wet seasons to test treatment effects

during seasons with contrasted rainfall

availability during the year. We calculated

95% confidence intervals of this index for

each level of irrigation amount or

frequency, and per each wet/dry period to

test for significant differences from zero

(which means neutral effect of the nurse on

the response variable). If 95% confidence

intervals did not overlap with zero or with

each other, we took as significant the

differences of the RII from zero or between

treatments, respectively. These differences

were interpreted as increases (if

significantly higher) or decreases (if

significantly lower) of soil moisture or

Fv/Fm under the canopy of the herbaceous

neighbours in comparison with open areas.

Relative growth rate data obtained for

root collar diameter and slenderness

coefficient were correlated from one year to

another (r > 0.375, P < 0.0001 in all cases).

Thus, we used multivariate analysis of

variance (MANOVA) to test the effects of

microsite and irrigation on this variable at

each study site. These analyses were

conducted separately for Experiments 1 and

2. Slenderness coefficient data were

squared-root transformed to meet

MANOVA assumptions.

Damage derived from rabbit activity

(warrens and territory coverage) was an

important source of seedling death despite

of the grazing protection provided (see

results below). Thus, we separated the

survival status of Retama seedlings into

three levels: alive, death by drought or

death by rabbit. We analyzed survival

percentages of these seedlings separately

for each year and study site by using a

hierarchical log-linear analysis, with

microsite and irrigation as fixed factors. To

assess the effects of the factors assayed

during each year, only those seedlings that

survived the previous summer were taken

into account (for example, to analyze

survival of 2008, we only considered those

seedlings alive after the summer of 2007).

With this approach, we were able to assess

the consistency of the effect of the

treatments evaluated over the years. This

approach also avoids the potential

“dragging” that an extremely strong effect

of a given treatment during a given year

may have on the overall net results (e.g. if

the +50% treatment would had strong

CHAPTER 1

51

effects in 2007, but not in the rest of years

we could detect it with our approach; in

contrast, taking into account survival

percentage from the beginning may produce

an overall net significant effect of this

treatment over the study period). Survival

data from one year to another were not

correlated (Pearson r < 0.4; P > 0.2 in all

the cases); thus, independency is expected.

As with growth data, separate analyses

were performed for each experiment.

Statistical analyses were conducted using

SPSS 13.0 for Windows (Chicago, Illinois,

USA).

RESULTS

SOIL MOISTURE

The different irrigation treatments produced

contrasted results depending on the

grassland considered. In the natural

grassland, the frequency, but not the

amount, of water added modified the effect

of Stipa on soil moisture; the opposite

response was found in the anthopogenic

grassland. In the natural grassland, Stipa

had a mostly neutral effect on soil moisture

in both dry and wet periods, regardless of

the amount of water added (Fig. 1.2A). This

neutral effect shifted to slightly positive

when the +50% treatment was applied in

two pulses (2x treatment), a response not

observed in the 4x treatment (Fig. 1.2B). In

the anthopogenic grassland, the increase in

water amount reduced monotonically the

negative effect of annuals on soil moisture

(Fig. 1.2C). Annuals reduced soil moisture

in the +0% treatment, but did not affect, or

even increased, water availability

comparing to open microsites in the +25%

and the +50% treatments, respectively. The

latter had the same effect regardless of the

frequency of watering pulses (Fig. 1.2D).

These effects were consistent in both dry

and wet periods.

PLANT PERFORMANCE

In Experiment 1, the irrigation treatments

modified the effect of grasses on the Fv/Fm

of Retama seedlings in a way that

mimicked the effects of the former on soil

moisture in the natural grassland (Fig. 1.3).

However, the amount of water applied did

not affect the studied interaction in the

natural grassland, with the exception that

Stipa effect shifted from neutral to slightly

positive as water availability increased

during wet seasons (neutral in the +0%

treatment and slightly positive in the rest;

Fig. 1.3A). Although the amount of water

was not important affecting the outcome of

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

52

Figure 1.2 Relative effects of herbaceous nurse plants, as measured with the Relative Interaction Index (RII), on soil water availability during dry/wet seasons (periods with soil moisture above and below 10%, respectively). RII was calculated for the mean values of ten Nurse/Open pairs for each of the irrigation treatments assayed. Legend of the Experiment 1 (panels A and C) and Experiment 2 (panels B and D) treatments as follows: +0% = no irrigation, +25% = irrigation of 25% of the median of April-July, +50% = irrigation of 50% of the median of April-July in four pulses, 4x = irrigation of 50% of the median of April-July rainfall in four pulses, and 2x = irrigation of 50% of the median of April-July rainfall in two pulses. Data represent means ± 95% confidence interval (n = 10) in the natural (A and B) and the anthropogenic (C and D) grassland, respectively.

the studied interaction during dry seasons,

less frequent but heavier water inputs (2x

treatment) neutralized the negative effect of

Stipa on Retama found in dry periods (Fig.

1.3B). Alternatively, nurse annuals reduced

monotonically their negative effect on

Retama seedlings during dry seasons as the

water amount added increased (Fig 1.3C),

but this treatment did not affect the studied

interaction during wet periods. Contrary to

the results found in the natural grassland,

the 2x treatment did not affect the

interaction outcome neither in dry nor in

wet seasons in the anthropogenic grassland

(Fig. 1.3D).

CHAPTER 1

53

Figure 1.3. Relative effects of herbaceous nurse plants, as measured with the Relative Interaction Index (RII), on seedling photochemical efficiency during dry/wet seasons (periods with soil moisture above and below 10%, respectively). RII was calculated for the mean values of six Nurse/Open pairs for each of the irrigation treatments assayed. Data represent means ± 95% confidence interval (n = 6) in the natural (A and B) and the anthropogenic (C and D) grassland, respectively. Rest of legend as in Fig. 1.2.

Neither microsite nor irrigation

treatments affected the growth rate of

Retama in the natural grassland, which was

almost nill in all the cases (Table 1.2; see

detailed statistics in Appendix A in

Supplementary Material). The 2x treatment,

but not the rest, slightly decreased the

slenderness coefficient of Retama seedlings

(Table 1.3). This trend was constant for the

three years of study, but was significant

only in 2007 (F3,53 = 3.63; P = 0.019). The

growth rate of Retama was much higher in

the anthropogenic grassland than in the

natural grassland, and was negatively

affected by the presence of annuals

(MANOVA Pillai´s Trace: F2,53 = 4.6; P =

0.014). According to the results found with

the Fv/Fm measurements, increases in

water availability (+50% treatment)

neutralized this negative effect of annuals,

regardless of the frequency of its

application. The slenderness coefficient was

no affected neither by microsite nor by any

of the irrigation treatments in the

anthropogenic grassland.

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

54

Table 1.2. Relative growth rate (cm · cm-1 · year-1) of root collar diameter (RGR) of Retama sphaerocarpa seedlings in the different experiments and study sites. Values are means ± SE (n = 4-16, depending on the treatment and the sampling period). Open = areas without perennial vegetation, Nurse = areas under the canopy of annual plants (anthropogenic grassland) or Stipa tenacissima (natural grassland), +0% = unwatered seedlings, +25% = irrigation of 25% of the median of April-July rainfall in four pulses, 4X = irrigation of 50% of the median of April-July rainfall in four pulses, and 2X = irrigation of 50% of the median of April-July rainfall in two pulses. Different letters or asterisks indicate significant differences (MANOVA test; P < 0.05) between microsites or among irrigation treatments, respectively. See Appendix A in Supplementary Information for detailed statistical results.

Experiment 1: Amount treatment

Microsite Irrigation Natural grassland Anthropogenic grassland

2007-2008 2008-2009 2007-2008 2008-2009

Open

+0% 0.00 ± 0.02 0.02 ± 0.01 1.37 ± 0.13 1.66 ± 0.13a

+25% 0.06 ± 0.01 0.00 ± 0.01 1.54 ± 0.08 1.83 ± 0.06a

+50% 0.05 ± 0.01 0.00 ± 0.01 1.40 ± 0.08 1.63 ± 0.08a

Nurse

+0% 0.00 ± 0.02 0.01 ± 0.01 1.40 ± 0.09 1.70 ± 0.11b

+25% 0.02 ± 0.01 0.00 ± 0.01 1.59 ± 0.13 1.56 ± 0.12b

+50% 0.03 ± 0.01 0.00 ± 0.01 1.56 ± 0.22 1.75 ± 0.13b

Experiment 2: Frequency treatment

Open 4X 0.05 ± 0.01a 0.00 ± 0.01 1.40 ± 0.08 1.63 ± 0.08

2X 0.04 ± 0.01a 0.01 ± 0.02 1.41 ± 0.11 1.66 ± 0.11

Nurse 4X 0.03 ± 0.01b 0.00 ± 0.01 1.56 ± 0.22 1.75 ± 0.13

2X 0.00 ± 0.01b 0.00 ± 0.02 1.46 ± 0.09 1.60 ± 0.08

The presence of grasses reduced the

survival of Retama in the two grasslands

studied during 2007 and 2008, a negative

effect that disappeared in 2009 (Table 1.4,

see also Appendix B in Supplementary

Material for detailed statistical results). In

the anthropogenic grassland, we found a

negative effect of grasses on the ability of

Retama seedlings to resist summer drought

in 2007, but a positive effect was detected

by the protection against rabbit damage

during this year, resulting in a net neutral

effect (see Retama seedlings death by

drought or by rabbits in Table 1.4).

CHAPTER 1

55

Table 1.3. Slenderness coefficient (unitless) of Retama sphaerocarpa seedlings in the different experiments and study sites. Values are means ± SE (n = 4-16, depending on the treatment and the sampling period). Rest of legend as in Table 1.2

Experiment 1: Amount treatment

Microsite Irrigation Natural grassland Anthropogenic grassland

2007 2008 2009 2007 2008 2009

Open

+0% 2.4 ± 0.3 2.2 ± 0.2 1.8 ± 0.1 2.7 ± 0.2 2.2 ± 0.1 1.7 ± 0.1

+25% 3.2 ± 0.2 2.2 ± 0.2 2 ± 0.2 3.5 ± 0.7 2.1 ± 0.2 1.8 ± 0.1

+50% 3.4 ± 0.4 2 ± 0.2 2 ± 0.1 4 ± 0.5 2.5 ± 0.2 1.6 ± 0.1

Nurse

+0% 2.9 ± 0.6 3 ± 0.5 2.5 ± 0.5 2.7 ± 0.2 2.2 ± 0.1 1.7 ± 0.1

+25% 2.9 ± 0.2 2.4 ± 0.2 2.3 ± 0.2 3.1 ± 0.3 2.2 ± 0.2 2 ± 0.1

+50% 3.7 ± 0.5 2.7 ± 0.4 2.4 ± 0.2 3.4 ± 0.4 2.4 ± 0.2 1.9 ± 0.1

Experiment 2: Frequency treatment

Open 4X 3.4 ± 0.4 2 ± 0.2 2 ± 0.1 4 ± 0.5 2.5 ± 0.2 1.6 ± 0.1

2X 2.4 ± 0.2* 1.9 ± 0.1 1.9 ± 0.2 4.1 ± 0.5 2.6 ± 0.2 1.7 ± 0.1

Nurse 4X 3.7 ± 0.5 2.7 ± 0.4 2.4 ± 0.2 3.4 ± 0.4 2.4 ± 0.2 1.9 ± 0.1

2X 2.3 ± 0* 2.2 ± 0.4 2.1 ± 0.2 3.7 ± 0.3 2.7 ± 0.1 2.4 ± 0.2

These negative effects were independent of

the irrigation treatments, with the exception

of Experiment 1 in the anthropogenic

grassland in 2008. During this year, the

survival of Retama monotonically increased

with water availability (+0% < +25% <

+50%), and the negative effects of annuals

on the survival of this species disappeared

in the +50% treatment (Table 1.4,

Appendix B). Different irrigation

frequencies did not affect the survival of

Retama seedlings neither in the natural nor

in the anthropogenic grassland.

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

56

Table 1.4. Percentage of Retama sphaerocarpa seedlings death by drought or by rabbit (separated by /) during the three years of study in the different experiments and study sites. Percentage were calculated taking as 100% those seedlings which had survived after the summer of the previous year. Initial n was 22 for each combination of treatments. Different letters or asterisks indicate significant differences (Log-linear test; P < 0.05) between microsites or among irrigation levels within each experiment, respectively. See Appendix B in Supplementary Information for detailed statistical results. Rest of legend as in Table 1.2.

Amount treatment

Microsite Irrigation Natural grassland Anthropogenic grassland

2007 2008 2009 2007 2008 2009

Open

+0% 4 / 0 10 / 3a 33 / 3 0 / 12a 12 / 12a 33 / 15

+25% 4 / 8 18 / 0a 33 / 0 0 / 9a 10 / 19a 38 / 8

+50% 0 / 4 21 / 0a 44 / 0 0 / 13a 4 / 13a* 13 / 13

Nurse

+0% 13 / 0 60 / 0b 75 / 0 8 / 0b 30 / 13b 52 / 11

+25% 8 / 0 50 / 0b 75 / 0 9 / 4b 30 / 22b 40 / 20

+50% 0 / 0 53 / 11b 63 / 11 10 / 0b 19 / 0b* 38 / 5

Frequency treatment

Open 4X 0 / 4 21 / 0a 44 / 0 0 / 13a 4 / 13 13 / 13

2X 0 / 0 19 / 5a 40 / 15 0 / 16a 0 / 21 20 / 24

Nurse 4X 0 / 0 53 / 11b 63 / 11 10/ 0b 19 / 0 38 / 5

2X 17 / 4 53 / 5b 74 / 5 12 / 4b 21 / 4 55 / 3

DISCUSSION

WATER STRESS INCREASES THE

INTENSITY OF GRASS-SHRUB

COMPETITION

According to our first hypothesis, we found

a negative effect of both Stipa and annuals

on the growth of Retama seedlings

(Goldberg and Novoplansky 1997).

However, this effect was almost nill in the

natural grassland, maybe due to the low

growth rates found in this ecosystem

regardless of the treatment applied. The

negative effect of annuals found in the

anthropogenic grassland seems to be mainly

caused by competition by water, an effect

that increased with water stress. According

to previous studies (Knoop and Walker

1985, Davis et al. 1999, Maestre and

CHAPTER 1

57

Cortina 2004a), the highest competition

intensities were found under the higher

water stress levels (both summer drought

and non-watered treatments in the

Experiment 1), a response that contradicts

predictions of the SGH (Bertness and

Callaway 1994). Both Stipa and the annuals

negatively affected the growth, stress level

and survival of Retama seedlings. These

results partially reject our second

hypothesis, based on the model proposed by

Goldberg and Novoplansky (1997), because

herbaceous neighbours reduced shrub

survival and did not show any positive

effect during dry periods (Figs. 1.2 and

1.3). Also, and in contrast with our second

hypothesis, this effect remained equal

regardless of the water amount added in the

natural grassland, but was more negative in

the anthropogenic grassland as water stress

increased (see results of Experiment 1). The

most plausible explanation for these results

is that the negative effects of both Stipa and

annuals on the stress experienced by

Retama seedlings during wet periods

reduced growth and resource capture,

compromising survival during summer

regardless of the environmental stress

experienced during this season.

Ruderal plants have been suggested as

possible nurse plants in degraded areas,

where other nurse plants are absent, and its

potential use as a restoration tool has been

suggested (Brooker et al. 2008). However,

our results do not fully support this

recommendation because of the inexistence

of positive effects of these herbaceous

plants during the entire study period (see

also McDonald 1986, Eliason and Allen

1997). Interestingly, when considering the

impact of rabbit activity, which was

considerable in the anthropogenic

grassland, the negative effect of annuals on

the ability of seedlings to overcome

summer drought was compensated by the

protective role of these annuals (see Retama

seedlings death by drought/rabbit in Table

1.4). This result illustrates how the effects

that multiple stressors have on a particular

nurse-protégée interaction is a crucial factor

affecting the potential of this particular

nurse to have positive, neutral or negative

effects on the protégée depending on the

particular environment where this

interaction occur (Smit et al. 2009,

Soliveres et al. in press).

DIFFERENT NURSE GRASSES HAVE

UNEVEN EFFECTS ON RETAMA

SEEDLINGS

Competition between Retama and annuals

was clearly driven by water competition,

and therefore increases in soil moisture

reduced the negative effect of these species.

However, the effects of Stipa on soil

moisture (neutral or positive) did not fully

match those on seedling performance

(survival reduction and increased stress),

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

58

suggesting that competition by factors other

than water were also important drivers of

the final outcome of the interaction studied.

The shade intolerance of Retama

(Valladares and Pugnaire 1999, Valladares

et al. 2003), together with the trend to

increase slenderness coefficient underneath

Stipa canopy, seems to point to light

competition as a key factor defining the

outcome of the studied interaction in the

natural, but not in the anthropogenic,

grassland, where Stipa canopies could

shade Retama seedlings enough to limit

carbon gain (Davis et al. 1999, Soliveres et

al. 2010, Seifan et al. 2010a). Light

competition could explain the contrasting

results found between both grassland types,

where a lack of influence of water amount

in the natural, but not in the anthropogenic

grassland was found. Our findings, and

those from previous studies (Prider and

Facelli 2004, Seifan et al. 2010a), point to

the existence of a trade-off between the

shade and drought tolerances of the

protégée plant and how nurse plants affect

water and light supply for their protégées.

For example, although nurses may exert a

positive effect on the availability of one

resource (i.e. water), the overall effect of a

nurse on a particular target will be negative

if shade provided by this nurse is deeper

than the physiological limits of this

particular protégée plant can withstand,

regardless of its positive effect on water

availability (Malkinson and Tielbörger

2010). This trade-off should be taken into

account when studying specific plant-plant

interactions along stress gradients because

it will define the final outcome of a given

interaction and how it changes across

environmental gradients (Holmgren et al.

1997, Prider and Facelli 2004, Malkinson

and Tielbörger 2010).

CHANGES IN RAINFALL

FREQUENCY, HERBACEOUS

PHENOLOGY AND POTENTIAL FOR

NICHE SEGREGATION

Niche segregation is among the most

important factors fostering coexistence in

semiarid environments (Fowler 1986, Sala

et al. 1989, Scholes and Archer 1997).

Despite that our study only lasted three

years, it seems enough time for Retama

seedlings to avoid competition by grasses,

as demonstrated by the lack of effect of

nurse herbs on Retama survival during

2009 at both study sites. Retama seedlings

are able to reach deep soil layers soon after

its establishment (Padilla and Pugnaire

2007), and thus it is likely that they were

able to reach deeper soil profiles than those

achieved by its herbaceous neighbours in

three years. These results agree with

previous studies with other species (Brown

and Archer 1990), which found a high

ability of woody seedlings to reach soil

CHAPTER 1

59

resource partitioning with grasses very

early in their life cycle.

Contrary to our expectations (Ogle and

Reynolds 2004, Knapp et al. 2008) and

despite of the existence of niche

segregation described above, the assayed

differences in the frequency of heavy

storms (2x vs. 4x treatment in Experiment

2) did not increase seedling survival when

growing with grasses in any of the studied

sites. However, taking into account other

measurements (soil moisture and Fv/Fm),

the two assayed frequencies had differential

effects depending on the site considered.

While more frequent water inputs (4x

treatment) promoted higher soil moisture

and shrub performance in the anthropogenic

grassland, less frequent but heavier water

inputs (2x treatment) had the same effects

in the natural one. Following the inverse

soil texture hypothesis proposed by Noy-

Meir (1973), this could be caused by

differences in soil depth between both

grasslands, with deeper soils (i.e. the

natural grassland) allowing niche

segregation and shallow soils (i.e.

anthropogenic grassland) preventing it.

However, both soil types had similar

geological basis and thus it is likely that

both had similar textures and depths. Thus,

the most plausible explanation for this

differential response between grassland

types is that differences on the life-strategy

(annuals vs. perennials) of the herbaceous

nurses caused it (Gómez-Aparicio 2009). In

the anthropogenic grassland more water

pulses (those of spring and early summer)

in the 4x than in the 2x treatment increased

the amount of water available for seedlings

when the annual nurses were active. This

increase in water availability might

compensate competition by water between

annuals and Retama seedlings, as

demonstrated by the significant reduction of

negative effects on survival, growth or

stress found. On the other hand, in the 2x

treatment one of the two irrigation pulses

was applied during summer, when most

ruderal nurses were death. Therefore,

negative effects of annuals on Retama

seedlings were less important when the

plant received this irrigation, resulting in a

less effective compensation of water

competition when compared to the same

amount of water applied more frequently.

In contrast, Stipa is not only perennial and

therefore competitive during the entire year,

but also highly efficient taking water from

short and light rainfall inputs, even during

summer (Pugnaire et al. 1996b, Balaguer et

al. 2002). Thus, the 4x treatment more

likely benefited Stipa than reduced

competition between this species and

Retama seedlings in the natural grassland.

However, the less frequent but heavier

water pulses applied in the 2x treatment

could reach deeper soil profiles, where

Stipa is not able to take water, and therefore

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

60

may enhance niche segregation, increasing

shrub performance (Fv/Fm results; see also

Sala et al. 1989, Schwinning and Ehleringer

2001, Ogle and Reynolds 2004). In spite of

this competition was importantly affected

by light (see discussion above), an

improvement in water status could increase

the ability of Retama to compete for light

(Fahey et al. 1998), as demonstrated by the

significant increase in the slenderness

coefficient and Fv/Fm observed in the 2x

treatment. Unfortunately, we did not

measure root structural traits, but the

extremely low shoot growth rates of

Retama seedlings found in this ecosystem,

compared with those observed in the

anthropogenic grassland (Table 1.2),

suggests a major investment in root growth

to promote this niche segregation. The

differential responses to different water

availabilities and frequencies found

depending on the life-strategy of the

herbaceous plants could help to reconcile

contrasting results observed in the

literature, which support or reject niche

segregation between woody plants and

grasses.

CONCLUDING REMARKS

Despite the plethora of studies devoted to

test facilitation/competition shifts along

stress gradients (see Callaway 2007 for a

review), to our knowledge none of them

have experimentally tested these shifts

within the same site, avoiding site-to-site

confounding factors, and with more than

two points along stress gradients driven by

water availability. We did so in two

contrasting ecosystems with different

herbaceous potential nurses, and following

a four-point realistic water stress gradient,

derived from predictions for future climate

change scenarios for the study areas, within

each of the three study years. Our results

suggest that the expected increase in water

stress under climate change will reduce

shrub recruitment and performance in

semiarid grasslands, but these effects will

depend on the specific species involved and

the suggested trade-off between

shade/drought tolerances of protégée and

nurse effects on light and water availability.

However, the increase in the frequency of

heavy rains may counteract this effect by

enhancing niche segregation among

coexisting plants (Knapp et al. 2008), an

effect that will be mediated by the life-

strategy of the nurses involved. The results

presented here challenge current predictions

of plant-plant interactions in response to the

ongoing climate change in Mediterranean

ecosystems (Brooker 2006), and therefore

raise caution on current generalizations on

how grass-shrub interactions will respond

to climate change.

ACKNOWLEDGEMENTS We thank Matthew Bowker for revising the English of this manuscript. A.P. Castillo-Monroy, M.

CHAPTER 1

61

Carpio, E. Pigem, C. Alcalá, P. Alonso, R. Milla, L. Gimenez and J. Margalet for their help during fieldwork. We thank the Instituto Madrileño de Investigación y Desarrollo Rural, Agrario y Alimentario (IMIDRA) and CINTRA S.A. for allowing us to work in their sites. SS and PGP hold PhD fellowships from the EXPERTAL project, funded by Fundación Biodiversidad and CINTRA S.A. This work was supported by the EXPERTAL,

REMEDINAL2, INTERCAMBIO and EFITAL [B007/2007/3-10.2] projects, the latter funded by the Ministerio de Ciencia e Innovación (MICINN). FTM acknowledges support from the European Research Council under the European Community's Seventh Framework Programme [(FP7/2007-2013)/ERC Grant agreement n° 242658].

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

62

Supplementary Material for Chapter 1.

Appendix A. Detailed statistical results of the MANOVA perfomed with growth rates and

slenderness coefficients.

Statistical results of the MANOVA perfomed with growth rate and slenderness coefficient

data in the natural (A, B; separate analyses for each variable) and the anthropogenic (C, D;

separate analyses for each variable) grasslands. Significant (P < 0.05) or marginally

significant (0.05 < P < 0.1) are highlighted in bold and italics, respectively. RGR = relative

annual growth of root collar diameter; SC = slenderness coefficient.

Table S1. MANOVA results showing the overall effect of the factors assayed (microsite and irrigation) on the 5 variables tested (RGR 2007-08 and 2008-09, and SC for 2007, 2008 and 2009) in the natural grassland. This analysis was performed separately for Experiments 1 (Amount treatment) and 2 (Frequency treatment).

Experiment 1: Amount treatment

Factor Pillai´s trace F df P-value Microsite (M) 0.574 4.859 5,18 0.005

Irrigation (I) 0.394 0.933 10,38 0.515 M x I 0.192 0.404 10,38 0.937

Experiment 2: Frecuency treatment

Microsite (M) 0.386 2.518 5,20 0.063

Irrigation (I) 0.402 2.686 5,20 0.052

M x I 0.115 0.520 5,20 0.758

Table S2. Results from separate ANOVA analyses for each variable in the natural grassland. Analyses were conducted separately for Experiments 1 and 2 (Amount and Frequency treatments, respectively). Experiment 1: Amount treatment

Dependent variable

Microsite (M) Irrigation (I) M x I

F df P F df P F df P RGR 2007-08 3.9 1,42 0.06 0.4 2,42 0.66 0.1 2,42 0.91 RGR 2008-09 1.6 1,42 0.22 1.1 2,42 0.35 0.3 2,42 0.73

SC 2007 1.9 1,42 0.19 1.8 2,42 0.18 0.2 2,42 0.79 SC 2008 5.0 1,42 0.04 0.6 2,42 0.56 0.4 2,42 0.69 SC 2009 3.3 1,42 0.08 0.6 2,42 0.55 0.2 2,42 0.80

Experiment 2: Frecuency treatment RGR 2007-08 7.6 1,24 0.01 2.6 1,24 0.12 0.2 1,24 0.65 RGR 2008-09 0.5 1,24 0.48 0.9 1,24 0.77 0.0 1,24 0.93

SC 2007 0.7 1,24 0.80 9.1 1,24 0.01 0.5 1,24 0.48 SC 2008 3.4 1,24 0.08 1.8 1,24 0.20 0.5 1,24 0.49 SC 2009 2.9 1,24 0.1 0.9 1,24 0.36 0.5 1,24 0.48

CHAPTER 1

63

Table S3. MANOVA results showing the overall effect of the factors assayed (microsite and irrigation) on the 5 variables tested (RGR 2007-08 and 2008-09, and SC for 2007, 2008 and 2009) in the anthropogenic grassland. This analysis was performed separately for Experiments 1 (Amount treatment) and 2 (Frequency treatment).

Experiment 1: Amount treatment

Factor Pillai´s trace F Hypothesis/error df P-value

Microsite (M) 0.286 5.056 5,63 0.001

Irrigation (I) 0.238 1.731 10,128 0.081

M x I 0.054 0.358 10,128 0.962 Experiment 2: Frecuency treatment

Microsite (M) 0.341 4.653 5,45 0.002

Irrigation (I) 0.085 0.839 5,45 0.529 M x I 0.045 0.424 5,45 0.829

Table S4. Results from separate ANOVA analyses for each variable in the anthropogenic grassland. Analyses were conducted separately for Experiments 1 and 2 (Amount and Frequency treatments, respectively).

Experiment 1: Amount treatment Dependent

variable Microsite (M) Irrigation (I) M x I

F df P F df P F df P RGR 2007-08 3.3 1,67 0.07 0.3 2,67 0.73 0.0 2,67 0.97 RGR 2008-09 5.3 1,67 0.03 0.3 2,67 0.76 0.8 2,67 0.48

SC 2007 0.6 1,67 0.44 3.0 2,67 0.06 0.3 2,67 0.76 SC 2008 0.0 1,67 0.95 1.2 2,67 0.30 0.1 2,67 0.94 SC 2009 0.17 1,67 0.70 0.2 2,67 0.78 0.6 2,67 0.54

Experiment 2: Frecuency treatment RGR 2007-08 3.4 1,49 0.07 2.1 1,49 0.16 0.3 1,49 0.56 RGR 2008-09 3.4 1,49 0.07 3.7 1,49 0.06 1.8 1,49 0.19

SC 2007 1.0 1,49 0.31 0.4 1,49 0.55 0.1 1,49 0.80 SC 2008 0.7 1,49 0.79 1.0 1,49 0.33 0.2 1,49 0.70 SC 2009 0.3 1,49 0.62 2.7 1,49 0.10 1.6 1,49 0.22

CLIMATE CHANGE AFFECTS PLANT INTERACTIONS

Appendix B: Detailed statistical results of the log-linear analyses performed with survival frequencies.

Hierarchical log-linear results for the different models for both study sites and for each year of study (2007-2009). Table S5 shows the results of

the analyses conducted with the three amount treatments evaluated (+0%, +25% and +50%; Experiment 1), without considering the frequency

levels. Table S6 shows the results of the analyses conducted with the two frequency treatments evaluated (2x and 4x; Experiment 2), without

introducing the amount levels. P-values < 0.05 are shown in bold

Table S5

Factor df

Anthropogenic grassland Natural grassland

2007 2008 2009 2007 2008 2009

G P G P G P G P G P G P

Microsite 2 14.5 0.0007 6.9 0.031 0.54 0.763 5.6 0.062 23.6 >0.001 2 0.365

Irrigation 6 0.06 0.999 11.9 0.018 2.4 0.655 8.8 0.065 1.6 0.813 0.1 0.999

Table S6

Factor df

Anthropogenic grassland Natural grassland

2007 2008 2009 2007 2008 2009

G P G P G P G P G P G P

Microsite 2 12 0.0024 3.5 0.174 1.8 0.410 5.9 0.052 12.9 0.002 2.3 0.314

Irrigation 6 0.5 0.778 1.7 0.434 2.7 0.253 5.7 0.057 0.4 0.816 3.4 0.186

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

Santiago Soliveres, Lucía DeSoto, Fernando.T. Maestre and José Miguel Olano

Manuscrito publicado en:

Perspectives in Plant Ecology, Evolution and Systematics 12: 227-234.

2

Spatio-temporal heterogeneity in abiotic factors mo dulate multiple

ontogenetic shifts between competition and facilita tion

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

66

CHAPTER 2

67

ABSTRACT

Plant-plant interactions are largely influenced by both environmental stress and ontogeny. Despite the effects of each of these factors on the overall outcome of these interactions has received considerable attention during the last years, the joint effects of both factors as drivers of such outcome are poorly understood. We used the combination of spatial pattern analysis, fruit production surveys, carbohydrate assays, sowing experiments and dendrochronological techniques to explore the interaction between Stipa tenacissima (nurse) and Lepidium

subulatum (protégée) in two different slope aspects. This battery of techniques allows us to study the effects of the nurse plant during the whole life cycle of the protégée, and to assess the role of spatio-temporal variability in abiotic stress as a modulator of ontogenetic shifts in plant-plant interactions. Spatial pattern analyses suggested a net facilitative effect of S. tenacissima on L. subulatum. This effect was particularly important during the germination, shifting to competition (growth reduction) early after establishment. Competition was gradually reduced as the shrub aged, suggesting niche differentiation. The magnitude of competition was reduced under low rainfall levels in south-facing slopes, whereas this response was observed due to other abiotic factors in north-facing slopes. Our results highlight the crucial effect that positive interactions at early life-stages have to determine the long-term outcome of a given plant-plant interaction, and the existence of multiple shifts between facilitation and competition along different life-stages of the protégée. They also show how these ontogenetic shifts are modulated by abiotic factors, which differ among slope aspects. These findings may help to refine conceptual and theoretical models about shifts between facilitation and ontogeny under current climate change scenarios.

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

68

INTRODUCTION

he analysis of the spatio-temporal

variation of facilitative and

competitive interactions along

abiotic stress gradients has become a major

research topic in community ecology

during the last two decades (Kikvidze

1996, Maestre and Cortina 2004a,

Kikvidze et al. 2006, Brooker et al. 2008).

In arid and semiarid areas, such variations

are particularly important, as water, which

is the most limiting factor, shows a strong

spatio-temporal variability (Whitford 2002,

Holmgren et al. 2006). These environments

are characterized by a high inter-annual

variability in rainfall distribution, and as a

result, plant recruitment is limited to

particularly rainy years (Holmgren et al.

2006). Furthermore, water availability also

experiences a strong spatial variation

between slope aspects: radiation and

temperatures, and thus water stress, are

higher in south-facing slopes, while in

north-facing slopes water stress and

evapotranspiration are lower (Friedman et

al. 1977, Bellot et al. 2004, Aragón et al.

2007, Pueyo and Alados 2007). These

spatio-temporal changes in water

availability have been pointed as a major

factor defining the final outcome of plant-

plant interactions in drylands (Tielbörger

and Kadmon 2000a, Pugnaire and Luque

2001, Gómez-Aparicio et al. 2004, Miriti

2007).

The relationship between abiotic

stress and the final outcome of plant-plant

interactions is further complicated by the

ontogenetic changes that plants experience

throughout their life cycle, which can

strongly modulate facilitation/competition

shifts (Miriti 2006, Schiffers and

Tielbörger 2006, Armas and Pugnaire

2009). Short-term studies, which form the

core of facilitation/competition research

(see Callaway 2007 for a review), are

insufficient to fully understand the

magnitude of ontogenetic shifts in plant-

plant interactions, but long-term studies are

often logistically prohibitive because of

economic and temporal constraints. Some

studies have overcome these limitations by

using annual plants (Schiffers and

Tielbörger 2006), or by sampling specific

temporal windows of the plant life cycle

(Armas and Pugnaire 2005, 2009, Miriti

2006, Valiente-Banuet and Verdú 2008).

These approaches in isolation are

insufficient to test ontogenetically-driven

facilitation/competition shifts along the

whole plant life, particularly in long-lived

perennial plants, and to assess the effects

of spatio-temporal changes in abiotic stress

on such ontogenetic shifts. These problems

can be circumvented using

dendrochronological techniques, assigning

annual rings to calendar years

(Schweingruber 1988). Since xylem acts as

T

CHAPTER 2

69

conductive area for water and nutrients

within a plant (e.g. Dyer and Bailey 1987,

Bascietto and Scarascia-Mugnozza 2004),

this technique can reconstruct investment

in secondary growth along plant life, and

therefore, act as a measurement of plant

performance in each year during its whole

life.

Although it is known that the

outcome of plant-plant interactions may be

affected by the interaction between abiotic

stress and the ontogeny of the target

species (Goldberg et al. 2001), there is a

lack of studies evaluating the simultaneous

effects of abiotic stress, both in space and

in time, and ontogeny as drivers of on the

outcome of plant-plant interactions (but see

Schiffers and Tielbörger 2006; Sthultz et

al. 2007). Improving our understanding on

the interacting effects of these factors will

allow us to further refine current

conceptual and mathematical models

aiming to predict how plant-plant

interactions change along stress gradients

(Michalet 2007), and to increase the

precision of our estimates about how plant

individuals and communities will respond

to ongoing climate change (Brooker 2006).

In this study, we combine spatial pattern

analyses, sowing experiments,

dendrochronological and reproductive

surveys, and carbohydrate assays to

explore the relationship between the

tussock grass Stipa tenacissima L.

(Poaceae; the nurse plant) and the shrub

Lepidium subulatum L. (Brassicaceae; the

protégée plant) in two slope aspects (north

and south) with contrasting abiotic stress

environments. This combination of

approaches provides us with a continuous

set of data, suitable for testing the presence

of ontogenetic facilitation/competition

shifts throughout the entire life cycle of the

protégée. Furthermore, we aimed to assess

the role of spatio-temporal changes in

abiotic stress (differences between abiotic

factors controlling plant growth and

survival among slope aspects, and

differences between water availability

among years) as a modulator of these

shifts. We tested the following hypotheses:

(i) the germination and survival of L.

subulatum seeds and seedlings will be

higher under S. tenacissima than in

adjacent bare ground zones because of the

improvement of environmental conditions

under the canopy of this nurse plant

(Maestre et al. 2001, 2003, Barberá et al.

2006); (ii) the outcome of the interaction

will shift from facilitation to competition

with shrub age, resulting in less growth

and fruit production of L. subulatum when

growing under the canopy of S.

tenacissima (Miriti 2006); (iii) given that

both S. tenacissima and L. subulatum are

primarily stress-tolerant species (Pugnaire

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

70

et al., 1996b, Palacio et al. 2007), and that

water deficit (a resource-related stress) is

the main limiting factor, we expect to find

facilitative interactions mainly under

moderate abiotic stress levels, (Maestre et

al. 2009a). This will happen because the

positive effect of microclimatic

amelioration is limited after a threshold

under high levels of abiotic stress, where

competition, mainly for water, overcomes

the positive effect of this amelioration and

competition arises again (Maestre and

Cortina 2004a); and (iv) the magnitude of

facilitation and competition is modulated

by spatio-temporal changes in climatic

conditions.

METHODS

TARGET SPECIES

Lepidium subulatum is a dwarf summer

deciduous shrub linked to gypsum

outcrops. It is distributed along the

Western Mediterranean where it coexists

with S. tenacissima. Seeds are small,

exhibiting atelechory (Escudero et al.

2000) and forming a small permanent seed

bank (Caballero et al. 2008). Emergence

concentrates in winter, but spans to June,

with densities ranging from 19 to 700

seedlings m2 (Escudero et al. 2000, J.M.

Olano, unpubl. data). Plant recruitment

bottlenecks occur during the first summer

after germination, when survival rates

range from 0.1% to 10%, with high

survivorship linked to especially

favourable years (Escudero et al. 2000,

J.M. Olano, unpubl. data). Annual survival

rates increase sharply afterwards, reaching

71–95% for adults, depending mainly on

autumn and spring conditions.

Flowering starts at 2-4 years (M.

Eugenio, pers. com.). Primary growth

occurs in two pulses, from mid February to

June, and from September to November,

respectively (Palacio and Montserrat-Martí

2005, Palacio et al. 2007). Secondary

growth also occurs in spring. Flowering

and fruiting stages last from April to June.

A detailed description on the natural

history of S. tenacissima and the grasslands

it forms is given in Maestre et al. (2009b).

STUDY AREA

Three sites were selected in central Spain

for this study: Aranjuez (40º10´30´´N, 31º

54´09´´W; 545 m.a.s.l.); Tielmes (40º

12´40´´N, 31º25´02´´W; 595 m.a.s.l.); and

Noblejas (40º10´03´´N, 31º37´01´´W; 526

m.a.s.l.). Their climate is Mediterranean

semiarid, with average annual precipitation

of 388 mm, characterized by a high inter-

annual variability and a characteristic

strong summer drought. Mean annual

temperature is 14.6 ºC, ranking from 25 ºC

in July to 5.6 ºC in January (Data from

National Meteorological Service, 1994–

2005. Marqués et al. 2008). The three sites

CHAPTER 2

71

were located on gypsum-rich soils,

classified as Typic Gypsiorthid (Tielmes

and Noblejas) and Xeric Haplogypsid

(Aranjuez; Soil Survey Staff 1994).

Vegetation was in all cases an open steppe

dominated by S. tenacissima, and

contained shrub species like L. subulatum,

Retama sphaerocarpa (L.) Boiss. and

Helianthemum squamatum (L.) Dum.

Cours. Perennial plant cover is below 45%

in all cases.

EXPERIMENTAL DESIGN

At each site, experimental plots in north-

and south-aspect slopes were established,

with slopes varying between 13º and 22º.

Perennial cover was different depending on

the slope aspect in all the studied areas:

31% vs. 33%; 24% vs. 42%; and 32% vs.

35% for south- vs. north-aspect slopes in

Tielmes, Aranjuez and Noblejas,

respectively. Since cover can be a good

surrogate of productivity in ecosystems

such as those studied (Flombaum and Sala

2009), and productivity is a good proxy for

abiotic stress at the level of entire plant

communities (Lortie and Callaway 2006),

we assume that these differences are

related to higher levels of abiotic stress in

the south- than in the north-facing slopes.

This agrees with many studies conducted

in arid and semiarid areas showing higher

abiotic stress in south- vs. north-aspect

slopes (e.g. Friedman et al. 1977,

Sternberg and Shoshany 2001, Bellot et al.

2004; Aragón et al. 2008, Pueyo and

Alados 2007). Differences in perennial

cover between slope aspects were

particularly evident in the Aranjuez site

(Appendix C in Supplementary Material),

so this site was selected to carry out the

bulk of the fieldwork in this study.

OBSERVATIONAL MEASUREMENTS

In April 2008, five 25 m × 4 m transects

were randomly established in each of the

experimental plots (30 transects in total).

Every L. subulatum individual found along

the transect band was registered. Those

individuals located at distances shorter

than 20 cm and larger than 50 cm from the

edge of a S. tenacissima tussock were

considered as growing in association with

S. tenacissima and in isolation,

respectively. These situations are hereafter

called Tussock and Open microsites,

respectively. This distance has been used

as separation between microsites in other

studies with S. tenacissima, detecting

significant differences in both biotic and

abiotic features between Tussock and Open

microsites (e.g. Maestre et al. 2001, 2003,

2009b). L. subulatum individuals growing

at distances among 20–50 cm from the

edge of a S. tenacissima tussock were not

considered for further analyses.

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

72

SOWING EXPERIMENT

In October 2007, a seed germination

experiment was conducted in the Aranjuez

site. It was designed as a fully factorial

experiment with two treatments: slope

aspect (north vs. south) and microsite

(open areas devoid of vascular vegetation,

north and south-face of S. tenacissima

tussocks). Ten replicates were established

per treatment combination (each consisting

in a 25 cm × 25 cm plot), and 75

commercial seeds of L. subulatum were

seeded in each replicate (25 holes, 3 seeds

per hole). We chose this approach over the

alternative neighbor removal approach,

because neighbor removal does not erase

the facilitative legacy effects of a nurse

plant upon soil infiltration and fertility.

Seeds were buried at 0.5 cm to avoid ant

depredation, irrigated with 40 ml of water

and protected from rabbits (Oryctolagus

cuniculus L.) by using a metallic mesh that

did not shade the seeding site. A

germination test conducted under

controlled conditions revealed that the total

germination rate of the pool of seeds

employed was 89% after one month in a

growth chamber (16 light hours at 20ºC

and 8 dark hours at 10ºC). Seed emergence

and seedling survival were monitored

monthly until July 2008, when all

germinated seedlings died during the

summer drought. Because of this extreme

mortality event and the lack of germination

in Open sites, sapling survival data

analyses cannot be provided.

DENDROCHRONOLOGICAL

SURVEYS

In June 2007, adult individuals of L.

subulatum were randomly selected in the

Aranjuez site for dendrochronological

measurements. These individuals were

chosen among those naturally growing

under four different conditions, resulting

from the combination of two microsites

(Tussock vs. Open) and slope aspects

(north vs. south); 16 individuals were

selected for each combination (64 in total).

After harvesting, a section of the stem

including the root collar was selected to

measure the annual growth ring widths as

an indicator of plant growth over the

course of its life. Annual rings were dated

and measured following standard

denchronological techniques as detailed in

Appendix D in Supplementary Material. A

section of the main root of the same plants

was also collected to measure the content

of non-structural carbohydrates using the

anthrone method (see Olano et al. 2006 for

a full account of the methodology). Two

different fractions of non-structural

carbohydrates were measured in this study:

nonsoluble and soluble carbohydrates. In

L. subulatum, non-soluble carbohydrates

are used to overcome respiration rates in

the leafless plant during summer drought

CHAPTER 2

73

(Palacio et al. 2007). Therefore, low

contents in non-soluble carbohydrates may

compromise the ability of this species to

survive this critical moment. We interpret

differences in this variable between

microsites or slope aspects as differences

in the ability of L. subulatum to respond to

summer drought stress. On the other hand,

soluble carbohydrates provide a surplus of

sugars that can be stored for use under

favourable conditions (Chapin et al. 1990),

and that are susceptible to be immediately

allocated to functions such as growth.

Thus, higher contents of soluble

carbohydrates in a given combination of

microsite/slope aspect may indicate that

the plant experiences more benign

conditions there. Both soluble and non-

soluble carbohydrates inform us about the

status of the reserves of the plant, and thus

are an integrative measurement of plant

performance during the whole year. These

variables were measured in June, just

before summer drought, and in the main

root because this organ and date match

with the maximum starch content organ

and period of the year for L. subulatum,

respectively (Palacio et al. 2007).

FRUIT/INFRUCTESCENCE RATIO

SURVEYS

In June 2008, ten reproductive L.

subulatum individuals in each combination

of two slope aspects (north vs. south) and

microsites (Tussock vs. Open) were

randomly selected in the Aranjuez site. Ten

infructescences per plant were randomly

chosen, and the number of fruits in each

infructescence was registered. The canopy

area of each sampled individual was

calculated using the ellipse formula with

the diameters parallel and perpendicular to

slope. This measurement was introduced in

the analysis as covariate to control for

plant size.

STATISTICAL ANALYSES

The frequency of naturally occurring L.

subulatum individuals in the surveyed plots

was analyzed by using a Chi-square

goodness of fit test. Our null hypothesis

was that L. subulatum individuals have a

random spatial pattern (depending directly

on the cover of each microsite). Data were

tested for independence in 6 separate one-

way tables (resulting from each

combination of site and slope aspect)

including only the microsite factor

(Tussock vs. Open). A joint analysis that

would permit testing the interaction

between the factors included in the model

was not possible because the relative

Tussock/Open microsites cover (and

therefore the expected frequencies) in the

different site × slope aspect combinations

were not equiprobable, a general

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

74

assumption of the null hypothesis used

when analyzing contingency tables with

multiple factors. To calculate the expected

frequencies, the number of total plants

found in each transect were multiplied per

the percentage of cover of each microsite.

The sum of predicted/observed frequencies

of the five transects per each site × slope

aspect combination was used to run the

Chi-square test. To adjust for the increase

in Type I error because of multiple testing,

the Bonferroni correction was used

(corrected a: 0.05/6 = 0.0083).

The effects of slope aspect (north

vs. south) and microsite (open areas, north

face of S. tenacissima tussocks and south

face of S. tenacissima tussocks) on the

cumulative number of germinated seeds

were tested by using generalized linear

models (GLMs). GLMs were run with a

Poisson error distribution combined with a

log link function; Type I log-likelihood

ratios were used to analyze main effects.

As recommended to counteract data over-

dispersion and to adjust the statistics

properly, the scale parameter was

estimated by dividing the square root of the

Pearson’s Chi-square statistic by the

degrees of freedom (McCullagh and

Nelder, 1989). Differences between

microsites were tested using a post-hoc test

based on least-square means. Growth data

obtained from dendrochronological

measurements were analyzed using two

complementary approaches. First, the ring

width data of all individuals measured in a

given combination of microsite and slope

aspect were grouped according to L.

subulatum age, independently of

recruitment year, and averaged. The

average ring width of each of these four

groups was used to estimate the effect of S.

tenacissima on the growth of L. subulatum

throughout the ontogeny of the later by

using the relative interaction index (RII;

Armas et al. 2004). This index was

calculated as (Gst – Go)/(Gst + Go), where

Gst and Go are the average ring widths of

L. subulatum individuals growing in

Tussock and Open microsites, respectively.

RII values were obtained from 1- to 12-

year old individuals to maintain enough

sample size along the whole age range (n =

12–16 individuals in all the groups for all

the ages analyzed; sample size

dramatically decreased for older

individuals). This 12-year period recovers

different ontogenetic stages of the protégée

plant: seedling (first year), juvenile (from 2

to 4 years) and adult stage (from 5 years

onward). Therefore, it is sufficient to test

possible ontogenetic shifts in the sign of

the interaction studied. To remove

potential autocorrelation derived from

repeated measures of ring width, a Prais–

Winsten autoregression was performed to

test the relationship between the values of

the RII index and L. subulatum age. This

CHAPTER 2

75

analysis takes into account the non-

independence of the samples by estimating

a regression equation whose errors follow

a first-order autoregressive process (SPSS,

2004). Second, ring width data were

averaged for each calendar year, without

taking into account the age of each L.

subulatum individual, to evaluate the effect

of climate inter-annual variability on

annual growth rates. The RII index was

calculated with these data as described

above, evaluating the relationship between

this index and the rainfall registered in

March by using linear regression. March

rainfall was selected as the best possible

rainfall predictor of abiotic stress because

it was highly correlated with the

standardized ring width series of L.

subulatum (Pearson’s correlation index:

0.78 and 0.62, P < 0.05 for ring widths

growth of L. subulatum in Open microsites

in south- and north-aspect slopes,

respectively, see Appendix E in

Supplementary Material). To ensure that

no other important rainfall variable was

ruled out without the proper test, we

evaluated the bivariate correlations

between the width growth of L. subulatum

rings and the rest of possible rainfall

indicators of abiotic stress: rainfall of the

rest of the months, total annual rainfall,

monthly rainfall of September–December

of the previous year, cumulative rainfall of

spring (March-May), and the cumulative

rainfall of the two annual pulses of L.

subulatum growth (February-June and

September-November). Of all these

correlations, only March rainfall, which is

prior to the main pulse of L. subulatum

primary growth (Palacio et al. 2008), was

statistically correlated with the width

growth of L. subulatum rings (Appendix E

in Supplementary Material). Therefore, we

selected rainfall in this month as our

surrogate of abiotic stress. Separate

regression and autoregression analyses

were conducted for north and south slopes.

This approach allowed us to test

differences in the effects of rainfall on the

final outcome of the interaction depending

on the variability in the abiotic conditions

among the two slope aspects considered.

The relative importance of both ontogeny

and abiotic stress as drivers of the outcome

of the interaction studied was also

explored. For doing this, three sequential

analyses were run. First, the effects of

microsite (Tussock vs. Open) and slope

aspect (north vs. south) on the growth of L.

subulatum were evaluated using repeated

measures ANOVA. For doing this, ring

width data were grouped according to L.

subulatum age, independently of the

calendar year when they were formed. As

abiotic conditions are different in north and

south slopes due to different irradiation,

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

76

temperature and water stress levels

(Friedman et al. 1977, Bellot et al. 2004,

Aragón et al. 2008), separated analyses

were run for each slope aspect. This

procedure allowed us to test potential

differences in the nature of abiotic factors

controlling the final outcome of the

interaction at each slope aspect. The

second step was to introduce the

recruitment year as covariate in this

analysis to assess the specific weight of

climatic variability. L. subulatum

individuals were of contrasting ages

(recruitment dates of individuals analyzed

vary from 1981 to 2000), so the ring width

corresponding to each individual age may

be influenced by the particular abiotic

conditions in the period they grew, and

therefore a high intra-group variability was

expected to be found. As abiotic factors

can produce non-random differences

between individuals, and these differences

could mask the effect of an experimental

treatment, its inclusion as a covariate has

been recommended to gain power when

testing the effects of the factors of interest,

particularly when a high intra-group

variability is found (see Engqvist 2005 and

references therein). If the effect of this

covariate is significant, and changes that of

S. tenacissima, this would suggest that

abiotic stress is modulating the net effects

of S. tenacissima on the growth of L.

subulatum during ontogeny. Lastly, the

same analysis was performed but changing

the covariate to the median of March

rainfall during the lifetime of each

individual included in the analysis. March

rainfall is a key driver of L. subulatum

growth (Palacio and Montserrat-Martí

2005, Palacio et al. 2007; Appendix E in

Supplementary Material), and its median

would be a good estimator of the water

stress level suffered by each individual

during its entire life. With this analysis we

aimed to differentiate the effects of rainfall

from those of unmeasured abiotic factors

characterizing each year as modulators of

ontogenetic shifts in plant–plant

interactions. The effects of microsite

(Tussock vs. Open) and slope aspect (north

vs. south) on the mean

fruits/infructescence ratio and on soluble

and non-soluble carbohydrates content

were evaluated with a two-way ANCOVA,

where the size of L. subulatum individuals

was used as a covariate. GLM analyses

were carried out using the GENMOD

procedure of SAS 9.0 (SAS Institute, Cary,

NC, USA). The remaining statistical

analyses were conducted using SPSS 13.0

for Windows (Chicago, IL, USA). The

non-soluble carbohydrates content did not

meet the ANOVA assumptions (normality

and homocedasticity), and was transformed

by using the arcsin transformation. The rest

of the data met these assumptions, and

were not transformed.

CHAPTER 2

77

RESULTS

Lepidium subulatum was more frequent

than expected under S. tenacissima

canopies than in open microsites in five of

the six site × slope aspect combinations

studied (Table 2.1). The overall cumulative

germination rate was very low due to the

drought conditions of the study year (Fig.

2.1). However, it was higher under the

canopy of S. tenacissima than in Open

microsites. A significant microsite × slope

aspect interaction was also found,

suggesting that the magnitude of the

differences between microsites varied with

the slope aspect considered. More

germination events occurred in the north-

than in the south-face of S. tenacissima

canopies in the south slope (χ2 = 9.2; P =

0.002), but no differences between the

north- and south-face of the canopy were

found in the north slope (χ2 = 1.2; P =

0.277).

Table 2.1. Expected and observed frequency of Lepidium subulatum individuals found under the canopy of Stipa tenacissima (Stipa) and in bare ground areas (Open) in north- and south-facing slope aspects at the three studied sites. Data represent the sum of the number of individuals of the five transects measured in each site × slope aspect combination. Chi-square test results evaluate the effects of microsite (χ2 and P-value) in the various site and slope aspect combinations. Significant results (Bonferroni-corrected α) are in bold.

Site Slope

Aspect

Expected

frequencies

Observed

frequencies χ2

P-value

Open Stipa Open Stipa

Tielmes South 75.5 29.5 40 26 17.1 <0.0001

North 63.62 23.38 52 9 10.97 0.009

Aranjuez South 176.95 28.05 137 29 9.05 0.002

North 63.56 33.44 24 34 24.63 <0.0001

Noblejas South 99.44 33.56 89 17 9.27 0.002

North 146.42 63.58 119 49 8.48 0.004

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

78

Figure 2.1. Cumulative germinations registered in bare ground areas (Open), north face of Stipa tenacissima tussocks (Stipa north) and south face of S. tenacissima tussocks (Stipa south). Data represent mean + SE; n = 10. Different letters indicate significant differences between microsites in each slope aspect (post-hoc test based on the differences of least-square means).

A significant positive relationship between

the RII values and L. subulatum age was

found, suggesting that the negative effect

of S. tenacissima on the growth of L.

subulatum decreased as individuals aged

(Fig. 2.2A). This relationship was found in

both slope aspects (R2 = 0.83 and 0.51 for

south and north slopes, respectively). A

negative linear relationship between the

RII values and March rainfall was found in

the south slope, suggesting that negative

interactions dominated in years of high

March rainfall (Fig. 2.2B). No significant

relationships were found in the north slope.

Repeated-measures ANOVA showed a

negative effect of S. tenacissima on the

growth of L. subulatum individuals

(Appendix D in Supplementary Material),

which was particularly evident in the north

slope (north: F1,32 = 5.54, P = 0.025; south:

F1,27 = 3.88, P = 0.059). When recruitment

year was introduced as a covariate in this

analysis, the effects of microsite became

non-significant in both the south

(recruitment year: F1,26 = 6.71, P = 0.015;

microsite: F1,26 = 0.46, P = 0.503) and

north (recruitment year: F1,31 = 4.24, P =

0.048; microsite: F1,31 = 3.22, P = 0.082)

slopes. When the median of March

precipitation was used as a covariate, it did

not change substantially the effects of S.

tenacissima on the growth of L. subulatum

in the north slope (median March

precipitation: F1,31 = 0.89, P = 0.35;

microsite F1,31 = 4.66, P = 0.039), but it did

so in the south slope (median March

CHAPTER 2

79

precipitation: F1,26 = 3.63, P = 0.068;

microsite: F1,26 = 0.58, P = 0.454). Non-

structural carbohydrate content was higher

in Open than in Tussock microsites (F1,58 =

4.44; P = 0.039 and F1,58 = 4.71; P = 0.034,

for soluble and non-soluble carbohydrates,

respectively; Fig. 2.3).

Figure 2.2. (A) Relationships between the effect size of Stipa tenacissima on the growth of Lepidium subulatum, as measured by the RII index, and the age of L. subulatum. (B) Relationships between values of this index and the median of rainfall registered in March during the period 1995–2007. Results of significant autoregressions (A) and linear regressions (B) are shown. Each RII value is obtained by averaging growth data from 12 to 16 L. subulatum individuals.

B

A

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

80

Figure 2.3. Soluble and non-soluble root carbohydrates content (black and grey bars, respectively) of Lepidium subulatum individuals harvested into two different slope aspects (north vs. south) and growing underneath Stipa tenacissima canopy (Stipa) or in areas without perennial vegetation (Open). Data represent means ± SE; n = 16. Asterisks mark significant differences in root carbohydrates content between microsites within each slope aspect. Different letters mark significant differences among slope aspects for soluble (normal letters) and non-soluble carbohydrates (capital letters).

Figure 2.4. Number of fruits per infructescence (mean ± SE; n = 10) of Lepidium subulatum individuals growing into two different slope aspects (north vs. south) and microsites (underneath Stipa tenacissima canopy, Stipa, or in areas without perennial vegetation, Open).

CHAPTER 2

81

A higher content of soluble carbohydrates

was found in north- than in south-aspect

slopes (F1,58 = 28.82; P < 0.001), but non-

significant effects of slope aspect were

found when analyzing non-soluble

carbohydrates. The average number of

fruits/infrutescence ratio was affected by

the size of L. subulatum individuals (F1,35 =

10.70, P = 0.002), as larger plants had

more fruits. However, no significant

effects of slope aspect and microsite were

found (slope aspect: F1,35 = 0.91, P=0.346;

microsite: F1,35 = 0.04, P = 0.845. Fig. 2.4).

DISCUSSION

In contrast to studies showing a single

facilitation/competition shift throughout

the ontogeny of perennial plants (Miriti

2006, Valiente-Banuet and Verdú 2008,

Armas and Pugnaire 2009), our results

provide evidence of multiple ontogenetic

shifts between facilitation/competition

during the life cycle of the protégée. We

also found that spatio-temporal changes in

abiotic stress modulated these ontogenetic

shifts. To our knowledge, these responses

have not been described before. While the

importance of studying different abiotic

stressors and their interaction has been

highlighted (Holmgren et al. 1997, Riginos

et al. 2005, Baumeister and Callaway

2006), our study illustrates how the spatial

variability of these stressors, even at small

spatial scales, and their interaction with

ontogeny determine the final outcome of

plant–plant interactions.

As expected, we found higher

germination rates under the canopy of S.

tenacissima than in bare ground areas in

both slope aspects (Barberá et al. 2006,

Schiffers and Tielbörger 2006). This effect

seems crucial to define the final outcome

of the interaction studied, as indicated by

the strong net positive effect of S.

tenacissima on the abundance of L.

subulatum individuals in most of the

situations and sites studied. It is interesting

to note, however, that if only two

microsites (e.g. open vs. north face of S.

tenacissima tussocks) would had been

considered, as it has been done by most

facilitation studies (see Callaway 2007),

the higher positive effect of S. tenacissima

on the germination of L. subulatum found

in the south-aspect slope would point to a

higher facilitative effect with increases in

abiotic stress (Bertness and Callaway

1994). However, the contrasting results

found between slope aspects when

evaluating the germination of L. subulatum

in the north- and south-faces of S.

tenacissima tussocks suggest a complex

interplay between abiotic stress and

facilitation, which is strongly influenced

by spatial variability in these abiotic

stressors (e.g. light availability; Parker and

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

82

Muller 1982, Marañón and Bartolome

1993; S. Soliveres, unpubl. data). One

possible explanation for these results is

that in south-aspect slopes, where water

stress is higher than in north-aspect slopes

(Bellot et al. 2004, Aragón et al. 2008),

seeds predominantly germinate under the

north face of tussocks, where the

improvement in microclimate by S.

tenacissima is maximal due to water inputs

coming from run-off and to the shadow

provided by the canopy of this species

(Maestre et al. 2001, 2003). This microsite

preference is not so evident in the north-

aspect slope, as seed germination in both

faces of S. tenacissima tussocks was

similar. These results seems to suggest a

trade-off between microclimatic

amelioration in the north face of S.

tenacissima tussocks (Maestre et al. 2003)

and the increase in interference

competition with other neighboring plants

(Goldberg et al. 2001, Miriti 2006), as they

are less abundant under the south-face of

the tussocks (S. Soliveres, pers. obs.).

Stipa tenacissima had a negative

effect on the growth and carbohydrate

content of L. subulatum, suggesting that

not only sink activity (growth), but also

resource levels (carbohydrate content) are

lower in this microsite. Although we

cannot statistically differentiate the effect

found on L. subulatum growth from a

neutral one (we do not have error bars as

we grouped our data by the four possible

treatments combinations and averaged

them to calculate RII), there is a clear

reduction trend on the strength of the

negative effect on growth as L. subulatum

individuals aged. This effect, together with

the positive effect found on germination,

points to multiple facilitation/competition

shifts along the life cycle of L. subulatum.

Differences between our results and those

from previous studies (Miriti 2006,

Valiente-Banuet and Verdú 2008, Armas

and Pugnaire 2009) can be explained

because our study focuses on a grass–

woody plant interaction, while previous

studies have focused on woody–woody

plant interactions. In a grass–woody

interaction, the growth of the woody

individuals as they age helps to avoid

water and light competition from the

grasses (Fowler 1986, Van Auken 2000),

whereas the benefits of shade and

increased soil resources under the canopy

of the later still exist (Maestre et al. 2003).

This may render competition less

important, as effective nicheseparation is

likely to occur with increasing age (Fowler

1986, Van Auken 2000, Armas and

Pugnaire 2005). This is less likely to occur

when the nurse and protégée share the

same ecological traits (e.g. annuals:

Goldberg et al. 2001; Schiffers and

Tielbörger 2006; or shrubs: Miriti 2006),

as they are likely exhibiting greater niche

CHAPTER 2

83

overlap (Maestre et al. 2009a) and the

increase of standing biomass can lead to an

increase in resource competition (Schiffers

and Tielbörger 2006). This difference

suggests that differentiation in the traits of

the plants involved are likely to play an

important role in determining ontogenetic

facilitation/competition shifts (Armas and

Pugnaire 2009), and thus should be

explicitly taken into account when trying

to generalize the results of particular

studies.

Apart from the life story traits of

the interacting species, differences in the

tolerance to abiotic stress and the

competitive ability of the interacting

species are crucial when studying plant–

plant interactions (Liancourt et al. 2005,

Maestre et al. 2009a, Gross et al. 2010). In

our study, two stress-tolerant species

coexist along a temporal stress gradient

driven by water availability of each year.

As both species are stress tolerant, and

water is a resource-related factor, positive

interactions are expected to be dominant at

intermediate levels of abiotic stress

because of the existence of thresholds in

both sides of the gradient (Maestre et al.

2009a). This prediction was not met, as we

found less competition under more

stressful (i.e. lower rainfall) conditions in

the south slope, and no relationships

between abiotic stress and the outcome of

the interaction were found in the north

slope. The relationship with rainfall

observed in the south slope can be

explained by the compensation of the

negative effect of shade on growth with the

reduction in the water stress experienced

by L. subulatum, particularly in dry years

(Fig. 2.2B; Holmgren et al. 1997, Hastwell

and Facelli 2003). In our case, it seems that

rainfall modulated the trend toward

escaping light competition with age, being

shade less negative under dry years. This is

suggested by the reduction observed in the

negative effects of S. tenacissima on the

growth of L. subulatum when rainfall was

introduced as a covariate. In the north

slope, the ontogenetic trend was the same,

but abiotic factors other than rainfall

modulated this trend, as indicated by the

change in microsite effect along ontogeny

when recruitment year, but not rainfall,

was introduced in the analysis. Our study

shows how the same nurse effect, and its

interaction with ontogeny, depends on the

spatio-temporal changes in the overall

amount of abiotic stress experienced by the

interacting individuals, and on the

resources driving such stress (Holmgren et

al. 1997, Hastwell and Facelli 2003).

Moreover, we show how a longitudinal

track of competitive or facilitative

interactions and its relationship with

climatic conditions can be easily obtained

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

84

using secondary growth data present in

annual rings.

CONCLUDING REMARKS

In contrast with previous studies, which

found an increase in competition as

prote´ge´e plants grew (Miriti 2006,

Schiffers and Tielbörger 2006, Valiente-

Banuet and Verdú 2008), we found a

reduction of competition as plant aged.

This result may be influenced by

differences in ecological traits of species

involved in our study comparing to the

previous studies, which can result in an

effective niche separation between grasses

and shrubs, and therefore a reduction of

competition (e.g. Armas and Pugnaire

2005; our case) or in a niche overlap,

which together with the increase in

biomass can lead to higher competition

(e.g. Miriti 2006 for shrubs; Schiffers and

Tielbörger 2006 for annuals). Because

perennial grasses conform an important

component of many vegetation formations

worldwide (Zimmermann et al. 2010), our

results provide useful information to refine

current theoretical models about

facilitation/competition shifts along

ontogeny of perennial species. Our results

also provide important insights on how

spatio-temporal changes in abiotic stress

can modulate multiple

facilitation/competition shifts. As rainfall

increased competition escape in south-

aspect slopes, other factors were more

important in north-aspect slopes. Despite

the mostly negative effect of the nurse

plant on growth and reserve accumulation,

the positive effect found on early stages of

the life cycle of the protégée (germination)

may be driving the net positive sign of the

interaction, as demonstrated by the spatial

aggregation found between studied species.

Given the implications of understanding

how plant–plant interactions change along

stress gradients for accurately predicting

global change impacts on communities and

ecosystems (Brooker 2006), future studies

should pay special attention to the

interplay between abiotic stress and

ontogeny as joint drivers of

facilitation/competition shifts, and more

specifically, on the effect of nurse plants in

key stages of the life cycle under different

environmental conditions (Goldberg et al.

2001). This is particularly true when

working with long-lived species in

stressful environments.

ACKNOWLEDGMENTS We thank M. Bowker for an early revision on a previous version of this manuscript. E. Marcos, P. García-Palacios, A.P. Castillo, E. Pigem, C. Alcalá, J.C. Rubio (CESEFOR) and M. Méndez for their help during field and laboratory work. SS was supported by a fellowship from Fundación Biodiversidad-CINTRA (EXPERTAL project). LDS was supported by a fellowship from Junta de Castilla y León. FTM was supported by a ‘‘Ramón y Cajal’’ contract from the Spanish Ministerio de Ciencia e Innovación (MICINN), cofunded by the European Social Fund, by the British Ecological Society (Studentship 231/1975), and by the MICCIN project CGL2008-00986-E/BOS. JMO was supported by Junta de Castilla y León

CHAPTER 2

85

VA069A07 project. This research was supported by the EXPERTAL and INTERCAMBIO (BIOCON06/105) projects, funded by Fundación

Biodiversidad-CINTRA and Fundación BBVA, respectively.

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

86

Supplementary material for Chapter 2

Appendix C. Pictures showing differences among perennial cover in north (picture above) vs.

south (picture below) aspect slopes in the Aranjuez study site.

CHAPTER 2

87

Appendix D. Detailed description of growth measurements and ring-width growth data

used

After manually polishing Lepidium disc surfaces until the xylem cellular structure was clearly

visible, ring widths were visually dated by assigning calendar years (Stokes and Smiley 1968)

and measured to the nearest 0.001 mm using a sliding-stage micrometer (Velmex, Bloomfield

INC., NY, US) interfaced with a computer. The COFECHA software (Grissino-Mayer 2001)

was used to quantitatively check for cross-dating errors. The synchronised and highly inter-

correlated ring-width chronologies for each combination of microsite and slope were selected

from the pool of raw chronologies and used to build a master chronology. The master

chronologies represent the common ring-width growth for each combination of factors

(Figure S1). Master chronologies were standardized with the ARSTAN computer program

(Cook and Holmes 1996) by fitting to a spline function with a 50% frequency response of 32

yr, which was flexible enough to reduce the non-climatic variance by preserving high-

frequency climatic information.

Figure S1. Ring-width growth data in north and south slopes as Lepidium subulatum individuals aged. Stipa and Open indicate individuals located less than 20 cm and further away than 50 cm from the canopy of a Stipa tenacissima tussock. Most individuals from the Stipa microsite were located under its canopy. These data were used to calculate Relative Interaction Indexes (RII) shown in Figure 2.2 of the main text. Data represent means ± SE (n = 12-16).

MULTIPLE ONTOGENETIC SHIFTS IN PLANT INTERACTIONS

88

Appendix E. Climate Response of Lepidium subulatum growth

The obtained residual chronologies were correlated against monthly rainfall during the

growing season (September of the previous year to June) to assess which months determined

plant growth and can be used as indicators of stress level. Precipitation data were obtained

from the nearest meteorological station (period 1995–2007; Dosbarrios Meteorological

Station, located 11 km SE from Noblejas site).

Table S7. Pearson two-tailed correlations (r) between the standardized ring width growth data of Lepidium subulatum for the four slope aspect × microsite combinations and rainfall registered at the nearest weather station (39º53’04’’N, 3º28’33’’W. 714 m.a.s.l.). We consider each month separately and also the sum of the rainfall registered during February-June, September-November and March-May, which match with the growth pulses of L. subulatum. In all cases, n = 12. Stipa and Open indicate individuals located less than 20 cm and further away than 50 cm from the canopy of a Stipa tenacissima tussock, respectively. Most individuals from the Stipa microsite were located under its canopy. Correlations with P < 0.05 are in bold.

North -aspect slope

South -aspect slope

Open Stipa Open Stipa

r P r P r P r P Sep prev. 0.12 0.71 0.23 0.47 0.12 0.71 0.01 0.97 Oct. prev. 0.59 0.04 0.09 0.78 0.41 0.19 0.01 0.97 Nov. prev. -0.46 0.13 -0.09 0.78 -0.08 0.81 0.33 0.30 Dec. prev. -0.48 0.15 -0.19 0.56 -0.12 0.7 0.42 0.17 January -0.42 0.17 -0.21 0.51 0.06 0.85 0.33 0.29 February 0.22 0.49 -0.3 0.34 -0.1 0.82 -0.12 0.70 March 0.77 <0.01 -0.24 0.45 0.62 0.03 0.43 0.17 April -0.02 0.94 -0.03 0.92 -0.01 0.99 -0.04 0.91 May -0.31 0.34 -0.04 0.91 0.09 0.78 0.5 0.10 June -0.13 0.69 -0.27 0.39 0.18 0.58 -0.21 0.51 September -0.22 0.48 0.31 0.33 -0.06 0.86 0.46 0.13 October 0.33 0.29 0.18 0.58 -0.1 0.76 -0.4 0.20 November -0.25 0.44 -0.19 0.56 -0.07 0.83 0.06 0.86 December -0.23 0.48 -0.09 0.78 0.02 0.96 0.41 0.19 Annual -0.15 0.64 -0.34 0.29 0.14 0.67 0.46 0.13 Feb-June 0.17 0.61 -0.29 0.37 0.32 0.32 0.37 0.24 Sep-Nov. 0.26 0.42 0.14 0.67 0.39 0.21 0.28 0.38 March-May 0.15 0.65 -0.16 0.63 0.36 0.25 0.54 0.07

Santiago Soliveres, Pablo García-Palacios, Andrea P. Castillo-Monroy, Fernando T. Maestre,

Adrián Escudero and Fernando Valladares

Manuscrito publicado en: Oikos (en prensa). D.O.I. 10.1111/j.1600-0706.2010.18993.x

3

Temporal dynamics of herbivory and water availabili ty

interactively modulate the outcome of a grass-shrub interaction

in a semiarid ecosystem

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

90

CHAPTER 3

91

ABSTRACT

The study of plant-plant interactions along grazing and abiotic stress gradients is a major research topic in plant ecology, but the joint effects of both stressors on the outcome of plant-plant interactions remains poorly understood. We used two different factorial experiments conducted in a semiarid Mediterranean steppe to assess: 1) the role of the perennial grass Stipa tenacissima, a low-palatability species, providing protection from rabbit herbivory to the shrub Retama sphaerocarpa (Experiment 1), and 2) the effects of environmental amelioration provided by Stipa on the recovery of Retama after rabbit damage under two contrasted levels of water availability (Experiment 2). In the Experiment 1, water stress worked as an indirect modulator of herbivore protection by Stipa. This species protected Retama seedlings from rabbit herbivory during the wetter conditions of spring and winter, but this effect dissapeared when rabbit pressure on Retama increased during summer drought due to the decrease in alternative food resources. In the Experiment 2, Stipa exerted a negative effect on the survival of Retama seedlings during the three years of the experiment, regardless of inter-annual differences in rainfall or the watering level applied. This negative effect was mainly due to excessive shading. However, Stipa increased Retama recovery after initial rabbit impact, overriding in part this negative shade effect. Conversely, Stipa impact on the Fv/Fm of Retama seedlings depended on the intra-annual water dynamics and its experimental manipulation, overall contradicting predictions from the Stress-Gradient Hypothesis. The complex interactions found between herbivory, microclimatic amelioration from Stipa, and water availability as drivers of Retama performance illustrate the importance of considering the temporal dynamics of both biotic and abiotic stressors to fully understand the outcome of plant-plant interactions.

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

92

INTRODUCTION

nteractions among plants can be

negative, neutral, or positive,

depending their direction and

magnitude on the level and type of the

stressors governing a given community (see

Callaway 2007 for a review). Important

efforts have been devoted during the past

two decades to understand the dynamics of

plant-plant interactions along abiotic stress

or herbivory gradients, highlighting how

the presence of multiple abiotic stressors or

different levels of consumer pressure

importantly affect the outcome of plant-

plant interactions (e.g. Baumeister and

Callaway 2006, Graff et al. 2007).

However, and despite their common co-

occurrence in nature, the joint effects of

both herbivory and water stress on the

outcome of these interactions remain

largely ignored (Ibañez and Schupp 2001,

Veblen 2008, Smit et al. 2009). Considering

both abiotic stress and herbivory together is

crucial to understand the role of plant-plant

interactions in dryland ecosystems, where

these stressors are major factors influencing

plant community dynamics (Fischer and

Turner 1976, Whitford 2002).

The Stress-Gradient Hypothesis

(SGH; Bertness and Callaway 1994), a

framework in which most studies focused

on plant-plant interactions rely on, predicts

a continuous increase in the frequency of

positive interactions with increases in either

consumer pressure or abiotic stress.

However, several studies suggest that

positive plant-plant interactions may

collapse under extremely high levels of

both consumer pressure and abiotic

stressors directly related to resources, such

as water or light (e.g. Graff et al. 2007,

Maestre et al. 2009a). These studies caused

the generality of SGH predictions to be

challenged (Maestre et al. 2005, 2006,

Lortie and Callaway 2006). As a result of

this debate, predictions from the SGH have

been refined to consider the effect of

different abiotic stressors and the ecological

strategy of the species involved, and to

introduce consumer pressure as a major

factor affecting plant-plant interactions

along abiotic stress gradients (Maestre et al.

2009a, Smit et al. 2009). Furthermore, in

arid and semiarid areas water availability is

characterized by a strong inter- and intra-

annual variability, with marked temporal

dynamics that profoundly affect ecosystem

functioning (Whitford 2002). These

temporal dynamics add complexity to the

response of plant-plant interactions to

abiotic stress (Goldberg and Novoplansky

1997, Pugnaire and Luque 2001, de la Cruz

et al. 2008), and may also modulate the

I

CHAPTER 3

93

effect of herbivores on vegetation (Veblen

2008). Thus, they must be specifically

considered when studying the relationship

between plant-plant interactions and

different environmental biotic and abiotic

stressors in arid ecosystems.

Herbivory is also a major driver of

plant-plant interactions in these regions

(Baraza et al. 2006, Graff et al. 2007,

Veblen 2008). Unpalatable nurse plants

may protect their understorey vegetation

from herbivory damage by hiding more

palatable plants under their canopies, or by

sharing their defenses against herbivores

(Baraza et al. 2006, Veblen 2008), a process

commonly refered to as “associational

resistance” (sensu Hay 1986). This

protection against herbivory can strongly

affect vegetation dynamics and biodiversity

in grazed systems (Veblen 2008).

Alternatively, the improvement on water

status that nurse plants usually provide to

their target plants may also positively affect

target plant recovery from herbivory

damage (Rand 2004, Acuña-Rodriguez et

al. 2006). However, even in dry

environments, shade casted by nurse plants

can lead to light limitation for the protégée

plants (Seifan et al. 2010a, Soliveres et al.

2010). The joint effect that both shade and

an improved water status provided by nurse

plants have on the recovery from herbivory

of the protégée plants will depend on the

relative importance of water and light as

limiting factors for plant performance and

how herbivory affects their uptake (Wise

and Abrahamson 2005, 2007). Hence, the

final outcome on the protégée plants

response to herbivory is difficult to

generalize and predict, so more studies

along these lines are needed to refine

predictions on how plant communities

respond to different levels of abiotic and

biotic stressors (Graff et al. 2007, Smit et

al. 2009).

We conducted two simultaneous field

experiments to test the effects of rabbit

herbivory and water availability on the

interaction between the tussock grass Stipa

tenacissima L. (the nurse) and seedlings of

the leguminous resprouter shrub Retama

sphaerocarpa (L.) Boiss. (the protégée) in a

semiarid Mediterranean steppe. Stipa has an

overall low palatability (Ben Salem et al.

1994), and therefore could provide

herbivory protection for shrub seedlings by

associational resistance. Furthermore, the

positive effect that microclimatic

amelioration provided by Stipa has on the

survival of Mediterranean shrub seedlings

is well known (e.g. Maestre et al. 2001,

2003). Since water stress and the impact of

herbivores can prevent Retama

establishment in these environments

(Espigares et al. 2004, Rueda et al. 2008), it

is likely that the protection from herbivores

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

94

and the microclimatic amelioration

provided by Stipa canopies can play a key

role improving Retama colonization in

semiarid grasslands. We tested the

following hypotheses: i) Stipa provides

protection from rabbit herbivory to Retama,

enhancing its survival when rabbits are

present; ii) Shade provided by the Stipa

canopy enhances Retama water status

increasing its survival, but this positive

effect wane during extremely dry years due

to the overwhelming effect of competition

for water, iii) Regardless of plant water

status, light reduction produced by Stipa

will decrease Retama seedling recovery

from rabbit herbivory; and iv) Facilitation

of Retama seedlings by Stipa will collapse

under extreme levels of stress produced by

the joint action of herbivore damage and

drought stress.

METHODS

STUDY AREA

We conducted the study in the Aranjuez

Experimental Station, located in the center

of the Iberian Peninsula (40º03´60´´N,

3º54´91´´W). The climate is semiarid

Mediterranean, characterized by cold

winters and a strong summer drought, with

average annual precipitation and

temperature of 388 mm and 15 ºC,

respectively (1994-2005; Marqués et al.

2008, see also Appendix F in

Supplementary Material). The soil is

classified as Xeric Haplogypsid (Marqués

et al. 2008). Vegetation is an open steppe

dominated by Stipa (this species accounts

up to 90% of the total perennial cover),

with a perennial plant cover of 24%. Sparse

adult individuals of Retama and small

shrubs such as Lepidium subulatum L. and

Helianthemum squamatum (L.) Dum. Cours

are also present. The study site has a high

diversity of annual plants, which reach their

production peak in spring and constitute an

important part of plant productivity during

this period (Peco et al. 2009).

The study area harbours a high

density of rabbits (Oryctolagus Cunniculus

L.), as suggested by the high number of

visual contacts and the number of warrens

and latrines found (S. Soliveres, pers. obs.).

Domestic livestock or other large

herbivores are absent, and thus rabbits are

the only herbivores affecting vegetation

there. Rabbit activity tracks seasonal

variation in vegetation productivity. These

animals feed near their burrows to avoid

predation during spring and winter, when

their prefered food –mainly annuals– is

abundant; however, during summer drought

−when annuals dry out and food is scarcer−,

rabbits increase their exploration to obtain

enough food to survive (Rueda et al. 2008).

The selection of woody seedlings as a food

resource by rabbits increases during

summer (Maestre et al. 2001). Rabbit

CHAPTER 3

95

browsing, in the case of Retama seedlings,

may suppose the virtual removal of all the

aboveground biomass.

EXPERIMENTAL DESIGN

In December 2006, we planted 371 two-

year old Retama seedlings, with a mean

height of 27 ± 2 cm (mean ± SE, n = 20), by

using manually-dug holes of 20×20×20 cm.

The planted seedlings came from central

Spain (viveros Bárbol, Madrid, Spain), and

were maintained in the original nursery

until the week prior to planting. In the

nursery, seedlings were grown under

optimal conditions (full sunlight, fertilized

1:1 peat:coco-peat substrate, watering to

field capacity each week). We selected

these two-year old seedlings to ensure they

had enough size and reserves to resist rabbit

damage at least once (see Experiment 2

below). The selection of these two-year old

seedlings does not underestimate potential

facilitative effects of Stipa because Retama

seedlings were grown under optimal

conditions prior to their plantation, and thus

they were still sensitive to water stress and

to the environmental amelioration provided

by Stipa. We randomly planted these

seedlings on two different microsites:

“Tussock” and “Open”. The former

microsite was located < 15 cm from the

upslope face of an Stipa tussock (ca. 1 m

width and 80 cm height). Open microsites

were located in bare ground areas, > 50 cm

away from any perennial plant. One month

after planting, rabbits browsed some of the

seedlings, and we set up two parallel

experiments then. In the first experiment

(hereafter Experiment 1) we did not give

the seedlings any protection from further

grazing. This allowed us to test the

protection against herbivores provided by

Stipa canopy. In the second experiment

(hereafter Experiment 2) we evaluated the

joint effects of contrasted levels of water

availability and Stipa environmental

amelioration on the recovery of Retama

seedlings after rabbit browsing.

-Experiment 1: Nurse plant protection

against herbivores

From the 371 Retama seedlings planted, we

left a total of 195 seedlings without any

protection from herbivores (the remaining

plants after Experiment 2 was set up, see

below). From these seedlings, 103 and 92

were located in Tussock (refuge) and Open

(control) microsites, respectively. We

scored which of these plants had been

browsed by rabbits and which of them were

able to resprout and survive after the virtual

removal of their aboveground biomass (i.e.

the effect of rabbit damage) during January,

March and September 2007. Because of the

high number of rabbits present in the study

area and their repeated browsing, no plants

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

96

remained alive after September 2007, so we

stopped monitoring then.

-Experiment 2: Joint effects of herbivory

and abiotic stress on the outcome of the

interaction

This experiment had three factors: i)

Microsite (Tussock vs. Open), ii) Herbivory

(Retama seedlings whose aboveground

biomass was completely eaten by rabbits

during the first month in the field vs.

unbrowsed seedlings), and iii) Irrigation

(watered vs. non watered seedlings). Since

inter-annual variability is of crucial

importance in dry environments, and it may

strongly affect the results found and our

conclusions (Ibañez and Schupp 2001), we

conducted this experiment during a three

year-period (and therefore under three

contrasted environmental conditions) to add

confidence to the results found. We set up

22 replicates per each combination of

treatments (176 seedlings in total). All the

seedlings from this experiment were

protected from further herbivory after the

first month in the field using a metallic

mesh. The diameter of the openings in the

mesh was 5 cm, which casted no detectable

shade to the seedlings, and did not

confound the effects of any of the factors in

the experiment. In this experiment we

monitored the variables described below

(see Field monitoring section below).

We conducted a spatial analysis of

browsed seedlings using the Spatial

Analysis by Distance Indices (SADIE)

methodology (see Perry 1998 for details).

The spatial pattern of herbivory damage by

rabbits was random (SADIE’s Aggregation

Index [Ia] = 0.95; P = 0.56; n = 176). Thus,

we do not expect unmeasured variables

with spatial structure (e.g. soil depth,

distance to a rabbit burrow or slope

position) to influence seedling response to

the assayed treatments. We did not measure

seedling attributes that could influence

rabbit behaviour (e.g. plant C/N ratio, initial

plant height). However, the large number of

seedlings randomly assigned to each

treatment, and the fact that rabbit damage

was equally intense (i.e. complete removal

of aboveground biomass) regardless of the

microsite considered, should control for the

experimental noise that any unmeasured

factors potentially affecting rabbit behavior

could have on the results of this

experiment.

The irrigation treatment consisted in

eight supplementary pulses of water, once

every month, between April and July in

both 2007 and 2008. The wettest and driest

periods of the study area are spring (from

March to May) and summer (from June to

September), respectively. Thus, the

irrigation treatment affected both wet and

dry periods. In each monthly watering, we

applied an amount of water equivalent to

CHAPTER 3

97

the 50% of the median rainfall registered

during the past 30 years for this month,

irrespectively of the rainfall registered (i.e.

6, 9.3, 10.9 and 5.7 mm for April, May,

June and July, respectively; see Appendix 1

in Supplementary material). With this

irrigation scheme we aimed to evaluate the

joint effects of reducing the summer

drought (summer drought was longer in

unwatered plants because of the lack of

rainfall in June and July) and increasing

precipitation during the wet season (spring

rainfall was less abundant in unwatered

plants) on the outcome of the interaction

studied.

Field monitoring of Experiment 2

-Plant performance measurements

We monitored seedling height and survival

after each summer, which is the most

critical season for seedling survival in

Mediterranean semiarid regions (e.g.,

Maestre et al. 2001, 2003). We measured

these variables in September 2007, 2008

and 2009. Height was well correlated with

aboveground biomass, as demonstrated by

an allometric relationship performed with

seedlings of contrasted sizes and ages

(Spearman correlation: ρ = 0.65, P <

0.0001; n = 45), and thus was used as our

surrogate for seedling biomass in the field.

We measured photochemical

efficiency by using the in situ chlorophyll

fluorescence parameter Fv/Fm; it is

calculated from photosystem II (PSII)

fluorescence signals as the ratio between

the variable (Fv) and the maximum (Fm)

fluorescence signal, which are obtained

from a short light pulse after 20 minutes of

dark adaptation. Fv is the difference

between Fm and the minimal fluorescence

signal right before the saturating light pulse,

being Fm the light that plant cannot absorb

when its absortion capability has been

collapsed by a previous pulse. Fv/Fm was

determined with a pulse-modulated

fluorometer (FMS2, Hansatech Instruments,

Norfolk, UK). This variable is an estimator

of the overall plant stress (Maxwell and

Johnson 2000), and has been widely used as

an indicator of plant stress in numerous

studies in semiarid areas (e.g. Pugnaire et

al. 1996b, Maestre et al. 2001, 2003,

Aragón et al. 2008). Furthermore, small

changes in the concentration of chlorophyll

in leaf tissues associated with water

limitation, which can be tracked by

measuring Fv/Fm, can be crucial during

important stages of plant life (Aragón et al.

2008). Although the use of Fv/Fm can be

problematic for the detection of water stress

in some species (Resco et al. 2008), it is a

good proxy for plant stress in our case

because the canopy structure, lack of leaves

and the high tolerance of Retama to solar

radiation minimizes photoinactivation and

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

98

downregulation in this species (Valladares

and Pugnaire 1999), a confounding factor

that could mask the relationship between

Fv/Fm and water stress. We randomly

selected eight seedlings per treatment (n =

64) for Fv/Fm measurements. We measured

Fv/Fm in four key moments for seedling

performance in Mediterranean semiarid

environments: during the wettest period of

spring (April-May), in the middle and at the

end of the summer drought (July and

September, respectively) and during the

coldest month (December-January). Fv/Fm

was measured in the same plants from May

2007 to September 2009.

-Soil moisture measurements

We measured soil moisture by time-domain

reflectometry (TDR; Topp and Davis 1985)

using a Campbell TDR100 system

(Campbell Scientific Ltd, Loughborough,

UK). In eight randomly selected planting

holes per treatment, we vertically installed

10 cm long probes (n = 64). We chose this

soil depth because the vast majority of root

biomass of the planted seedlings

concentrates near the soil surface (the pots

where Retama seedlings were grown had

ca. 15 cm. depth), and because Stipa

tussocks also concentrate the majority of

their roots in the upper layers of the soil

(Puigdefábregas et al. 1999). We conducted

a site-specific calibration between Time

Domain Reflectometry measurements and

gravimetric soil moisture (R2 = 0.84; P <

0.0001; n = 68) to assess for the validity of

these measurements in gypsum soils and to

convert them in soil gravimetric moisture

data. We measured soil moisture in the

same plants in all the samplings, and in the

same sampling periods as Fv/Fm surveys

(see above). When these measurements

coincided with irrigation pulses, soil

moisture was measured at least one week

after the irrigation. With this approach we

avoided giving too much importance to the

occasional influence of irrigation in our

measurements, and assessed more

realistically the soil moisture available for

Retama seedlings during a given period. In

addition to these measurements, and to

further evaluate the effects of irrigation on

soil water availability, we measured soil

moisture one day after the irrigation pulse

was applied on June 2007 in the planting

holes of 10 undamaged plants (the eight

replicates selected for the previously

described soil moisture measurements plus

two extra replicates) of each of the four

combinations between microsite and

irrigation (n = 40). We measured soil

moisture only in unbrowsed plants because

we were only interested in assessing for

differences in soil moisture between the

irrigation treatments applied and the

microsites tested.

CHAPTER 3

99

STATISTICAL ANALYSES

As all Retama seedlings used in Experiment

1 died at the end of the monitoring, we did

not perform any statistical comparison with

the survival data at this period. However,

we used the log-rank statistic of the Kaplan-

Meyer procedure to compare the shape of

the survivorship and number of browsed

plants curves between Tussock and Open

microsites. We analyzed the survival of

Retama seedlings from Experiment 2

separately for each year by using a

hierarchical log-linear analysis, with

microsite, irrigation and herbivory as fixed

factors. To assess for the effects of the

factors assayed during the different

environmental conditions characterizing

each year, only those seedlings that

survived the previous summer were taken

into account for this analysis (for example,

to analyze survival of 2008, we only

considered those seedlings alive after the

summer of 2007). With this approach, we

were able to assess the consistency of

treatment effects over the years. It also

avoids the potential “dragging” that an

extremely strong effect of a particular

treatment during a given year may have on

the overall net results (i.e. we could detect

if herbivory had strong effects in 2007, but

not in the rest of years). Survival data from

one year to another were not correlated

(Spearman correlation: ρ < 0.4; P > 0.2 in

all the cases) and thus independency is

expected. Since browsing by rabbits

removed most aerial biomass of planted

seedlings, and initial seedling height after

rabbit impact was not measured, we did not

consider such height as a covariate in

further statistical analysis.

Months after planting0 2 4 6 8 10

unbr

owse

d se

edlin

gs (

%)

0

20

40

60

80

100

OPENTUSSOCK

Months after planting0 2 4 6 8 10

seed

ling

surv

ival

(%

)

0

20

40

60

80

100

OPENTUSSOCK

Figure 3.1. Dynamics of herbivore damage (A) and overall survivorship (B) of unprotected Retama sphaerocarpa seedlings growing under the canopy of Stipa tenacissima (Tussock) and in bare ground areas (Open) from December 2006 to September 2007.

We analyzed gravimetric soil

moisture (obtained from TDR data), Fv/Fm

A

B

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

100

and seedling height data by using repeated

measures ANOVA, with microsite,

irrigation and herbivory as fixed factors.

Data were square-root transformed to reach

normality and homocedasticity assumptions

when necessary. We found significant

interactions between time and the

treatments evaluated when analyzing soil

moisture and Fv/Fm data (data not shown).

These interactions can lead to the

misinterpretation of the effects of the fixed

factors (Quinn and Keough 2002), and are

of biological importance, as intra-annual

dynamics in water availability can strongly

affect the effects of nurse plants (Goldberg

and Novoplansky 1997, De la Cruz et al.

2008). To properly assess the effect of the

assayed treatments, and to explore how the

intra-annual dynamics in water availability

modulates them, we grouped both soil

moisture and Fv/Fm data for the three study

years in wet/dry periods (periods with soil

moisture values above and below 10%,

respectively) and analyzed them separately

using repeated measures ANOVA. We

established this 10% value to separate

wet/dry periods because it corresponds to a

biological threshold that separates periods

when most plant activity concentrates

(those with soil moisture > 10%) in

semiarid environments (Noy-Meir 1973,

Valladares et al. 2005). These analyses

could lack independency because results

obtained in dry periods were not completely

independent from those coming from wet

periods. However, by pooling the data from

the three study years together (we

conducted two separate RM ANOVA, one

with dry and another for wet periods data

from the three study-years) only consistent

results for the three years may result

significant and this lack of independency

disappears. Furthermore, with this approach

we removed the interactions with the

assayed treatments and time, avoiding the

confounding effect that the strong temporal

variability in water availability could have

on the interpretation of the main treatment

effects (Quinn and Keough 2002).

We evaluated differences in soil

moisture between watered vs. unwatered

plants (soil moisture measures after the

irrigation pulse of June 2007) using a two

factor (microsite and irrigation) ANOVA.

These data followed the assumptions of this

analysis, and thus were not transformed.

We conducted all statistical analyses using

SPSS 13.0 for Windows (Chicago, Illinois,

USA).

RESULTS

EXPERIMENT 1: NURSE PLANT

PROTECTION AGAINST HERBIVORES

-Survival and herbivore damage of

unprotected plants

Fewer plants were browsed by rabbits when

growing under Stipa canopies than in Open

CHAPTER 3

101

microsites during spring (Fig. 3.1A; Log-

rank test, P < 0.0001). Rabbit predation

intensity on Retama seedlings increased

during summer drought. This, together with

the removal of most aboveground biomass

produced after each browsing impact,

caused the depletion of the resprouting

ability of Retama seedlings, and all the

plants died during summer regardless of the

microsite where they were planted (Fig.

3.1B). Thus, we did not find differences in

survival among microsites during the study

period (Log-rank test, P = 0.466).

EXPERIMENT 2: JOINT EFFECTS OF

HERBIVORY AND ABIOTIC STRESS

-Plant performance

Stipa tenacissima reduced significantly the

survival of Retama seedlings during the

three years of study (2007: G2 = 12.7, df =

1, P = 0.005; 2008: G2 = 9.3, df = 1, P =

0.002 and 2009: G2 = 3.6, df = 1, P < 0.06,

Fig. 3.3. 2). Survival was 17%, 27% and

15% lower in Tussock than in Open

microsites for 2007, 2008 and 2009,

respectively. Irrigation increased the

survival of Retama seedlings during the

first summer by 27% (G2 = 10.29, df = 1, P

= 0.001), but did not affect the negative

effect of Stipa. Browsing damage did not

affect mortality rates per se in this

experiment, but reduced the negative effect

of Stipa on Retama (Microsite × Herbivory:

0

20

40

60

80

100

2009

Sur

viva

l (%

)

0

20

40

60

80

100

OPENTUSSOCK

2008

0

20

40

60

80

100

2007

H- H+

I-

H- H+

I+

I ; P < 0.0005M ; P < 0.0005M x H ; P = 0.005

M ; P < 0.005

M ; P = 0.06I x H ; P < 0.005

Figure 3.2. Survival of Retama sphaerocarpa seedlings during the three years of study in the eight combinations of treatments evaluated. Open = bare ground areas, Tussock = Stipa tenacissima canopies, I- = no irrigation, I+ = irrigation of 50% of the median of April-July period rainfall in four pulses, H- = no herbivore damage, and H+ = seedlings partially eaten by rabbits. Initial n = 22. G

2 = 7.7, df = 1, P = 0.005, Fig. 3.2). We

found a significant interaction between

prior herbivory damage and water

availability in 2009 (G2 = 8.85, df = 1, P =

0.003), being the survival of unbrowsed

seedlings higher than that of browsed

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

102

seedlings when these were unwatered.

When watered, seedlings that were browsed

showed a higher survival rate than those

that were not browsed (Fig. 3.2).

Herbivory decreased seedling height

during the three-years of study (RM

ANOVA: F1,171 = 4.6; P = 0.036), an effect

that was especially evident in 2008 (Fig.

3.3). We did not detect any microsite or

watering effect, neither any interaction

between the treatments evaluated or with

time, when analyzing seedling height. Stipa

tenacissima reduced the Fv/Fm ratio of

Retama seedlings during dry seasons (RM

ANOVA: F1,56 = 9.6; P = 0.003), but this

effect decreased when plants were watered

(Microsite x Irrigation: F1,56 = 3.9; P =

0.05). This Microsite x Irrigation

interaction was also found in wet periods,

when irrigation reduced the positive effects

of Stipa on Retama Fv/Fm (Fig. 3.4;

Appendix G.A).

-Soil moisture

2007 was the wettest year of the studied

period, with soil moisture levels well above

20% during spring (Appendix G.2B).

Conversely, 2009 was the driest year, with

soil moisture levels below 10% in three of

the four periods sampled (Appendix G.2B).

Stipa tenacissima slightly (< 2%) reduced

soil water availability during dry periods

(RM ANOVA: F 1,56 = 5.02; P = 0.029).

During wet periods, a significant microsite

× herbivory interaction was found (F 1,56 =

5.35; P = 0.024), with more water available

under Stipa canopies and Open microsites

for browsed and unbrowsed plants,

respectively (Fig. 3.5). Although irrigation

increased soil moisture values by an

average of 35% after watering (Two-way

ANOVA, F1,36 = 8.05; P < 0.001), it did not

affect soil moisture at the long-term, as this

treatment had no significant effects on this

variable when analyzing the data gathered

during the whole year (Repeated-Measures

ANOVA; P = 0.929).

DISCUSSION

The results of our study highlight the

importance of herbivory as a major factor

affecting the relationship between plant-

plant interactions and abiotic stress. The

increase in rabbit pressure during summer

drought, indirectly caused by the lack of

alternative food resources during this

season, overrided the herbivory protection

provided by Stipa during wetter periods,

when rabbit pressure upon Retama

seedlings was lower. Conversely, the initial

loss of biomass produced by rabbit

browsing shifted the interaction between

Stipa and Retama from negative to neutral.

CHAPTER 3

103

2009

10

20

30

40

2008

10

20

30

40

2007

Mea

n se

edlin

g he

ight

(cm

)

10

20

30

40

OPENTUSSOCK

H- H+

I-

H- H+

I+

RM ANOVA: H ; P < 0.005

Figure 3.3. Height of Retama sphaerocarpa seedlings during the three years of study in the eight combinations of treatments evaluated (mean ± SE; n depended on survival). Rest of Legend as in Figure 3.2.

Our results suggest that the negative effect

of Stipa on the performance (Fv/Fm) of

Retama was driven by water availability,

but that competition for other resources

rather than water modulated the effect of

Stipa on Retama survival. The complex

interactions between herbivory, abiotic

stress and their temporal dynamics as

drivers of the outcome of plant-plant

interactions highlight the importance of

considering these stressors together to fully

understand the outcome of plant-plant

interactions along environmental gradients

(Goldberg and Novoplansky 1997, de la

Cruz et al. 2008, Anthelme and Michalet

2009).

Pho

toch

emic

al e

ffici

ency

(F

v/F

m)

0.0

0.6

0.8

OPENTUSSOCK

DRY SEASON

0.0

0.6

0.8

WET SEASON

H- H+

I-

H- H+

I+

M ; P < 0.005M x I ; P = 0.005

M x I ; P = 0.05

Figure 3.4. Photochemical efficiency (Fv/Fm) of Retama sphaerocarpa seedlings during wet/dry periods (periods above/below 10% gravimetric soil moisture, respectively). Data are means ± SE of the three years-study period data pooled by wet/dry seasons; n = 8). Rest of Legend as in Figure 3.2.

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

104

Gra

vim

etric

soi

l moi

stur

e (%

)

0

2

4

6

8

10

OPENTUSSOCK

DRY SEASON

WET SEASON

H- H+

I-

H- H+

I+

0

5

10

15

20

25

30

M ; P < 0.05

M x H ; P < 0.05

Figure 3.5. Gravimetric soil moisture (inferred from time-domain reflectometry measurements) during wet/dry periods. Rest of Legend as in Figure 3.4.

PROTECTION AGAINST HERBIVORES

BY STIPA TENACISSIMA

Although Retama seedlings were protected

from herbivore damage when Stipa was

present, this effect dissapeared during

summer drought, when the higher rabbit

impact upon Retama seedlings

overshadowed the refuge effect of Stipa

(Fig. 3.1). Annual plants, which provide an

important fraction of plant productivity in

arid and semiarid systems (Fischer and

Turner 1978), have completed their life

cycle before the onset of summer drought in

our study area (Peco et al. 2009). Thus, the

corresponding increase in rabbit predation

upon perennials due to changes in diet

produced by the lack of annuals during

summer (Rueda et al. 2008) can explain the

suppression of this facilitative effect during

this season. Similar reductions of

facilitative effects under high herbivory

pressure have been previously reported

(Graff et al. 2007, Smit et al. 2007), and

should be common when food resources are

less abundant and the same number of

herbivores may exert higher pressure on the

remaining plants (but see Veblen 2008,

Anthelme and Michalet 2009).

DOES SHADE INTOLERANCE OF THE

PROTÉGÉE EXPLAIN THE NEGATIVE

EFFECT OF THE NURSE PLANT?

In contrast with previous studies using the

same nurse plant (e.g. Maestre et al. 2001,

2003), we found a net negative effect of

Stipa on survival of Retama seedlings.

Plant-plant interactions depend up to a great

degree on the identity of the species

involved (Callaway 2007), and thus these

contrasting results are not fully surprising.

Plant competition in drylands is generally

attributed to water or nutrients (Whitford

2002). Interestingly, most of the negative

effects of Stipa on Retama were not

influenced by increases in water

availability, neither were explained by the

effect of Stipa on this variable. The reduced

light availability under the canopy of Stipa

CHAPTER 3

105

tussocks (which suppose >80% of incident

PAR reduction; see Maestre et al. 2003),

has been identified as the major driver of

the positive effect of Stipa on shrub

seedlings (Maestre et al. 2003). However,

this same shade could be also a key factor

affecting the negative outcome of the

studied interaction (Seifan et al. 2010a,

Soliveres et al. 2010). The lack of leaves of

Retama and the concentration of the total

photosynthetic area on its cladodes, is

associated with the high light requirements

of this species (Valladares and Pugnaire

1999, Valladares et al. 2003, Espigares et

al. 2004), and suggests that a reduction in

available light might underlie the outcome

observed.

Most plant species adapted to

drought are not able to cope with deep

shade (Niinemets and Valladares 2006), and

therefore, it is likely that species that are

more adapted to full sunlight and drought,

which are abundant in dry environments, do

not benefit from the presence of a nurse

plant unless the positive effects of the nurse

on the water status of the protégée

overcome negative effects promoted by

light reduction (Holmgren et al. 1997). The

more drought-tolerant the protégée plant is,

the less positive the shade effect is expected

to be, according to the general ideas

discussed in recent revisions of the SGH

(Maestre et al. 2009a, Malkinson and

Tielbörger 2010). The same may happen

with different life stages of a given species,

as plants are often more shade tolerant early

during their ontogeny than in later stages of

development, and therefore are more likely

to benefit from nurse’s shade (Callaway and

Walker 1997, Miriti 2006). Our results

highlight the species-specific nature of such

effects (Callaway 2007), since the same

shade that is beneficial for some species

(Maestre et al. 2001, 2003) could

conceivably be negative for Retama or

other shade-intolerant plants (Marañón and

Bartolomé 1993, Seifan et al. 2010a). Thus,

more studies involving species with

different ecological strategies and drought

and shade relative tolerances are needed to

improve our understanding on the responses

of plant-plant interactions to abiotic stress

and herbivory at the entire community

level.

THE ROLE OF FACILITATION AND

RESOURCE AVAILABILITY ON THE

RECOVERY OF BROWSED PLANTS

Irrigation increased survival during part of

the studied period (Fig. 3.2) and also the

degree of stress experienced by Retama

seedlings was lower during wet periods.

Both results indicate that water was limiting

the performance of this species, regardless

of the microsite tested. However, light was

also an alternative limiting factor for plant

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

106

performance in Tussock but not in Open

microsites (discussed above). The Limited

Resource Model (LRM), which correctly

explains most of the relationships between

different resource levels and tolerance to

herbivory, predicts a differential recovery

from tolerance depending on the nature of

the limiting resources and the way

herbivory damage affect their uptake (Wise

and Abrahamson 2005, 2007). Particularly,

this model foresees a higher tolerance to

herbivory under wetter conditions when

water is the only limiting factor (such as in

Open microsites), but a lower tolerance

under these same wet conditions when

herbivory exacerbates the limitation of an

alternative resource, light in our case (such

as in Tussock microsites). This latter

outcome is explained because plants

growing under drought are already at their

optimum light conditions, and are mainly

limited by water availability, while plants

growing under Stipa canopies are closer to

their optimum moisture conditions, but also

limited by light. Thus, the latter plants will

be much more sensitive to the reduced

uptake of light (their alternative limiting

resource), and therefore to the loss of

biomass produced by herbivory, than the

former (Baraza et al. 2004, Wise and

Abrahamson 2005, 2007).

Following these predictions, we may

expect lower tolerance to herbivory in

Tussock than in Open microsites, especially

in watered plants, since plants growing

beneath the canopy of Stipa should be more

limited by light, while plants growing in

Open microsites should be limited only by

water. However, our results did not fully

match predictions from the LRM, maybe

because our target plant was a resprouter,

and therefore its tolerance to herbivory

might be affected not only by the amount of

resources available, but also by the reserves

of each seedling might have (Vesk et al.

2004). In contrast, a higher survival in

browsed plants was found when they were

watered, regardless of the microsite where

they were planted. This finding could

indicate an overcompensation of browsing

damage by plants when environmental

conditions were more benign (Crawley et

al. 1998). Why the LRM predictions did not

correctly explain the effects of shade

provided by Stipa on Retama recovery after

herbivory damage? Shade provided by

Stipa might increase the water status of

Retama seedlings by a reduction in

transpiration (Holmgren et al. 1997). Thus,

it is likely that this improvement in the

water status of Retama seedlings increased

their tolerance to herbivory in a similar way

that watering did it in Open microsites. This

positive effect in the recovery from

herbivory shifted the negative effect that

this same shade exerted on Retama survival

on unbrowsed plants (we found a

significant Microsite × Herbivory

CHAPTER 3

107

interaction when analyzing seedling

survival). The compensation of resource

competition due to herbivory protection has

been observed in previous studies in

semiarid environments (Graff et al. 2007).

Our results suggest that this compensation

also occurs when the nurse plant improves

recovery from herbivory, and highlight the

complex interaction between water stress

and herbivory as drivers of the outcome of

an interaction strongly influenced by shade.

TEMPORAL DYNAMICS IN WATER

AVAILABILITY MODULATE THE

EFFECT OF STIPA ON RETAMA

In contrast with our initial hypothesis, Stipa

negatively affected Retama survival during

the three years of study, regardless of the

inter-annual variation in rainfall availability

among years. This may be explained

because the studied interaction was mainly

driven by light competition, and therefore

water availability only played a secondary

role in its outcome. However, intra-annual

dynamics in water availability and our

irrigation treatment modulated the effect of

Stipa on the stress level experienced by

Retama seedlings. Goldberg and

Novoplansky (1997) proposed a conceptual

model to introduce the effect of intra-annual

resource dynamics on plant-plant

interactions. In their model, nurse plants

affected negatively protégée growth during

pulses (our wet seasons) due to competition

by resources, while increased survival

during periods with low nutrient availability

(our dry seasons). The final outcome would

depend on how much the negative effect on

growth during wet seasons affects survival

during dry periods, and on the relative

importance of plant uptake or abiotic

factors affecting resource availability

during these dry periods (Goldberg and

Novoplansky 1997, Hastwell and Facelli

2003). However, the effects of Stipa on the

Fv/Fm (our surrogate of plant stress) of

Retama seedlings during wet/dry seasons

found differed from the expected responses

arisen from the predictions of Goldberg and

Novoplansky (1997). Specifically, we

detected a trend towards facilitation and

competition during wet and dry seasons,

respectively (although it must be considered

that we only measured the degree of stress

experienced by Retama and not its growth

or survival seasonally, which could be a

better test for this model). Differences

produced in the outcome of the interaction

studied within these seasonal dynamics

varied with irrigation, which overall

suggest a reduction of competition intensity

at intermediate levels of abiotic stress

(watered plants during summer or

unwatered plant during spring), but a

prevalence of competition in the rest of

assayed situations, as suggested by the

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

108

refined SGH when the two species involved

are stress tolerators (Maestre et al. 2009a).

CONCLUDING REMARKS

Collectively, our experiments indicate the

existence of a complex hierarchy of

mechanisms affecting the interaction

studied (Baumeister and Callaway 2006). In

our case, water shortage per se was

irrelevant under extremely high herbivory

impacts (all seedlings died in Experiment 1

but none of them due to drought), but

indirectly modulated herbivory impacts by

affecting alternative rabbit food resources.

Conversely, the initial loss of biomass

produced by herbivory was a major factor

affecting the outcome of the interaction

studied. Stipa exerted a negative effect on

Retama seedlings mainly by light

competition, but this shade improved

seedlings recovery after herbivory,

overriding the negative effects found on

unbrowsed plants. Our findings illustrate

how the complex interactions between

herbivory and water stress jointly influence

the outcome of plant-plant interactions.

They provide insights to fully understand

the interplay between facilitation and

competition, and they can be used to further

refine conceptual models aiming to predict

the outcome of plant-plant interactions

along composite stress gradients.

ACKNOWLEDGEMENTS We wish to thank E. Chaneton, D. Eldridge, M. Seifan and three anonymous referees for their useful comments and corrections on a previous version of this manuscript. E. Pigem, C. Alcalá, S. Constán-Nava, J. Papadopoulos and E. Barahona helped during the fieldwork. We thank the Instituto Madrileño de Investigación y Desarrollo Rural, Agrario y Alimentario (IMIDRA) for allowing us to work in the Finca de Sotomayor (Aranjuez). SS and PGP hold PhD fellowships from the EXPERTAL grant, funded by Fundación Biodiversidad and CINTRA S.A. APC was supported by a PhD fellowship from the INTERCAMBIO (BIOCON06/105) grant, funded by Fundación BBVA. FTM acknowledges support from the European Research Council under the European Community's Seventh Framework Programme (FP7/2007-2013)/ERC Grant agreement n° 242658. This research was funded by the EXPERTAL grant, and with additional funds from INTERCAMBIO and REMEDINAL2 grants.

CHAPTER 3 .

109

Supplementary Material for Chapter 3

Appendix F. Climatic data (mean monthly temperature, black circles; and monthly rainfall,

grey bars) obtained from a meteorological station (Onset, Pocasset, MA, USA) located at the

study site. White bars represent the increment in monthly rainfall by the irrigation treatments

applied during 2007 and 2008.

Mc A

My Jn Jl Au S O N D Ja F

Mc A

My

Jn Jl

Au S O N D Ja

F M

c A

My

Jn

Jl

Au

S

Rai

nfal

l (m

m)

0

20

40

60

80

100

120

140monthly rainfallwatering

2007 2008 2009

Mea

n te

mpe

ratu

re (

ºC)

0

5

10

15

20

25

30

BIOTIC AND ABIOTIC STRESSORS AFFECT PLANT-PLANT INTERACTIONS

110

Appendix G. Photochemical efficiency of Retama sphaerocarpa seedlings (A) and

gravimetric soil moisture (A) in the different sampling dates for the eight possible

combinations of treatments. Open = bare ground areas, Stipa = Stipa tenacissima canopies, I-

= no irrigation, I+ = irrigation of 50% of the median of April-July period rainfall in four

pulses, H- =no herbivore damage, and H+ = seedlings partially eaten by rabbits. Data

represent means ± SE (n = 8).

Pho

toch

emic

al e

ffici

ency

(F

v/F

m)

2007 2008 2009

MayJuly

Sept.

January

May

July Sept.

January

April

July

Sept.

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Open I- H- Open I- H+ Open I+ H- Open I+ H+Stipa I- H-Stipa I- H+Stipa I+ H- Stipa I+ H+

MayJu

lySept.

January M

ay Ju

ly Sept.

JanuaryApril

July

Sept.

0

10

20

40Open I- H- Open I- H+ Open I+ H-Open I+ H+Stipa I- H-Stipa I- H+Stipa I+ H-Stipa I+ H+

Gra

vim

etric

soi

l moi

stur

e (%

)

2007 2008 2009

A

B

Santiago Soliveres, David J. Eldridge, Fernando T. Maestre, Matthew A. Bowker,

Matthew Tighe and Adrián Escudero.

Manuscrito en revisión en Ecological Monographs

4

On the relative importance of climate and biotic no n-trophic

interactions as drivers of local plant species rich ness in semiarid

communities

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

112

CHAPTER 4 .

113

ABSTRACT

Most studies including the role of positive plant-plant interactions as drivers of plant richness along environmental gradients assume an unimodal richness-productivity relationship, which is not as general as previously thought, and the existence of an overarching stress gradient which affects equally the different species forming a community. We aimed to evaluate the relative roles of facilitation/competition and environmental conditions as drivers of local species richness without these assumptions, and to clarify their contribution to the richness-productivity relationship. We conducted an observational experiment across wide environmental gradients in semiarid regions from Spain and Australia, assessing how the intensity, importance and frequency of positive interactions, and the degree of niche expansion provided by the nurse plants changed along these gradients. We also tested the particular mechanism (niche segregation, competitive exclusion or intransitivity) underlying the effects of nurses on their understorey vegetation. Nurse plants increased local richness by expanding the niche of the less adapted species both in Spain and Australia. The high variability of niches often found underneath their canopies may be the main reason why higher niche segregation and species coexistence was found under nurse than in open microsites. The outcome of the competition-facilitation continuum changed depending on the type of stress gradient considered. When it was driven by both rainfall and temperature (Spanish sites), the community-wide importance of nurse plants remained constant along the gradients. When the stress gradient was driven only by rainfall (Australian sites), the importance of nurses showed a unimodal relationship with the gradient, indicating a collapse of facilitation under both extremes of rainfall availability. Particular pairwise interactions outcomes were poorly predicted using abiotic measurements as an overarching stress level, and we propose to use each species distance to its environmental optimum as a better approach for this purpose. Our study provides a complete mechanistic understanding of the relative roles of plant-plant interactions and environmental conditions shaping local species richness in semiarid environments. These results can also be used to refine our predictions of the response of plant communities to environmental gradients, and clarify the relative importance of biotic interactions as a driver of such responses.

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

114

INTRODUCTION

he study of the mechanisms

controlling the composition of

communities has been a major

topic in ecology since the early days of this

science (see Gotelli and Graves 1996 for a

review). Recent research has highlighted

the fact that local-scale non-random

processes such as abiotic constraints and

biotic interactions determine the species

that are able to successfully colonize a

given environment (Huston 1999, Lortie et

al. 2004a, Rajaniemi et al. 2006).

Pioneering studies have suggested that,

among local-scale processes, competition

regulates richness at high levels of

productivity, while limited physiological

tolerances to abiotic stress or disturbance

reduce species recruitment, and thus

richness, at low levels of productivity

(Grime 1973, Huston 1979). An inherent

assumption of this observation is that

competition is less important at the lower

productivity end of the gradient. Overall,

the joint effects of abiotic constraints and

competition should generate a hump-shaped

relationship between species richness and

local productivity, which is reputed to be

ubiquituous in nature (Grime 1973, 2001,

Huston 1979). However, plant-plant

interactions are key drivers of community

structure in both high and low productivity

environments (Tilman 1988, Callaway

2007). Over the past decade, ecologists

have revisited the hump-shaped richness-

productivity relationship to explore the

potential effects of positive, non-trophic

interactions (hereafter ‘facilitation’; Hacker

and Gaines 1997, Michalet et al. 2006).

However, the relative effects of plant-plant

interactions and abiotic conditions on

changes in species richness along

environmental gradients, and therefore their

influence on the richness-productivity

relationship, remain uncertain (Rajaniemi et

al. 2006). Studies aimed at clarifying the

roles of these factors can help to explain

why several studies cast doubt on the

generality of the hump-shaped richness-

productivity relationship, particularly at

both local and regional scales (Waide et al

1999, Gillman and Wright 2006).

Many empirical studies (e.g. Hacker

and Bertness 1999, Kikvidze et al. 2005,

Valiente-Banuet et al. 2006, Cavieres and

Badano 2009) and theoretical models

(Bertness and Callaway 1994, Bruno et al.

2003, Lortie et al. 2004a) developed over

the past two decades have emphasized the

importance of facilitation for maintaining

community richness at low to moderate

levels of productivity (Hacker and Gaines

1997, Michalet et al. 2006). Environmental

buffering (both microclimatic amelioration

and protection from herbivory) by nurse

T

CHAPTER 4 .

115

species can increase the realized niche of

less adapted species, and therefore increase

local richness (niche expansion; Bruno et

al. 2003). Although the relationship

between the importance and frequency of

positive plant-plant interactions and

abiotic/biotic stress was originally thought

to be monotonic and positive (Bertness and

Callaway 1994), some studies have

suggested a waning of positive interactions

under either extreme abiotic stress levels

(e.g. Kitzberger et al. 2000, Ibañez and

Schupp 2001, Maestre and Cortina 2004a)

or intense disturbance (Ibañez and Schupp

2001, Smit et al. 2007, Forey et al. 2009),

thus obviating niche expansion (Michalet et

al. 2006). Explanations invoked to explain

this lack of niche expansion under such

conditions are: 1) the competitive effects of

nurse plants may outweight the positive

effects of environmental amelioration,

particularly when abiotic stress is promoted

by a resource such as water (Maestre and

Cortina 2004a, Maestre et al. 2009a), and 2)

nurse plants may not be able to attain a size

large enough to ameliorate harsh abiotic

conditions in extremely stressful

environments (Michalet et al. 2006). We

henceforth refer to both of these potential

mechanisms as facilitation waning models

(see Fig. 4.1).

Most studies describing facilitatory

mechanisms in relation to the diversity-

productivity curve have tended to focus on

unproductive to moderately-unproductive

environments (Hacker and Gaines 1997,

Michalet et al. 2006). Recent studies,

however, have revealed that ecological

processes such as niche segregation (Hector

et al. 1999, Silvertown 2004), competition

intransitivity (lack of hierarchy in

competition networks: Gilpin 1975, Laird

and Schwamp 2006, Bowker et al. 2010),

and indirect facilitation (Levine 1999,

Brooker et al. 2008) may be key

mechanisms enhancing species richness in

more productive conditions. More benign

conditions often found under nurse plants

(Franco and Nobel 1989) may increase the

local species pool and the heterogeneity of

available niches, also allowing finer

partitioning of variable resources. Although

largely ignored in facilitation research

(Brooker et al. 2008, but see Tielbörger and

Kadmon 2000b), these localized effects

may potentially increase niche segregation

(Pugnaire et al 1996,a Maestre and Cortina

2005) and competition intransitivity

beneath nurse plant canopies, enhancing

overall local richness and productivity.

Despite the interest generated in the joint

effects of nurse plants on both niche

expansion and changes in competition

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

116

outcomes, these mechanisms have largely

been explored separately in the literature.

There is a clear need, therefore, to develop

a unifying framework that integrates our

understanding of the roles of niche

expansion and competition in explaining

the role of plant-plant interactions on local

richness (Brooker et al. 2008).

Most conceptual models of

facilitation developed to date are based on

the notion that stress-tolerant plants

increase the performance of competitive

plants under “high levels of stress”. Thus

the underlying assumption of an

overarching stress gradient affecting plant

communities has permeated the facilitation

literature (Lortie et al. 2004a, Travis et al.

2005, Michalet et al. 2006). However,

stress is a complex concept (see Körner

2003, Lortie et al. 2004b, Körner 2004, for

a recent discussion) and is more

appropriately applied at the level of

individual species than at the community

level. Individual species are adapted to

tolerate particular environmental

conditions, and under these particular

conditions, a given species will experience

little or no stress (Körner 2003). However,

the morphological and physiological

adaptations to particular environmental

conditions constitute trade-offs in a species´

ability to cope with different stressors

(Tilman 1988, Niinemets and Valladares

2006). When a given species has to cope

with environmental conditions which it is

maladapted therefore, intuitively it could be

considered to be far from its environmental

optimum, and therefore limited or stressed

(Lortie et al. 2004b). Since species forming

a community do not need to be adapted to

exactly the same environmental conditions,

we can find, within a given community,

species differing in their optima and

therefore differences in the degree of stress

that they experience across environmental

gradients (Chapin et al. 1987, Greiner La

Peyre et al. 2001). We refer to this concept

henceforth as the individual-based stress

concept (Fig. 4.1). This may have profound

implications for our understanding on how

plant-plant interactions affect local richness

and its relationship with productivity, since

niche expansion does not necessarily need

to increase in the “moderate to high stress

direction” (Lortie et al. 2004a, Travis et al.

2005, Michalet et al. 2006). Rather, a

general mechanism operating across the

entire environmental gradient should be that

better adapted plants increase the realized

niche of less adapted species to a given set

of environmental conditions (Bruno et al.

2003). Thus when dealing with natural

plant communities, there is a clear need to

question the existence of an overarching

stress gradient for the whole community, or

of ecological strategies that remain constant

along these gradients (Greiner La Peyre et

al. 2001, Prider and Facelli 2004, Holmgren

CHAPTER 4 .

117

and Scheffer 2010). A reassessment of the

assumptions underlying these notions will

pave the way for an improved

understanding of the relative roles of biotic

interactions and abiotic conditions as

drivers of species diversity along

productivity gradients.

In this study we aimed to develop a

unifying framework that explains the

relative effects of plant-plant interactions

and environmental conditions on local

species richness across environmental

gradients. Our framework includes the two

main processes by which plant-plant

interactions may affect local diversity: 1)

niche expansion due to facilitation, and 2)

changes in competitive outcomes beneath

nurse canopies. We have included in this

framework an assessment of changes in the

intensity, importance and frequency of

positive interactions on community-level

plant richness and productivity along an

environmental gradient. We have also

evaluated changes in the intensity and

importance of a large number of pairwise

interactions along such gradients. To

develop our framework we established plots

across wide environmental gradients in

Spain and Australia, and assessed the

relative roles of plant-plant interactions and

abiotic constraints as drivers of the local

species richness at each plot. We chose

these two regions because they have

contrasting vegetation communities and

management histories, and both exhibit a

relatively wide diversity in species richness

across their respective environmental

gradients. Our main hypotheses were that:

1) nurse plants will enhance local species

richness via niche expansion and changes in

the competitive networks (niche segregation

and competition intransitivity) of their

understorey vegetation; 2) positive pairwise

interactions will wane at environmental

conditions corresponding to levels of

extreme stress for involved plants (Michalet

et al. 2006, Maestre et al. 2009a) and 3)

since different species co-ocurring in a

community differ in their relative tolerances

to given environmental conditions, the

intensity, importance and frequency of

facilitation at the community level, and

therefore niche expansion, will remain

constant across stress gradients as the

identity, but not the amount, of facilitated

species changes along such gradients

(Greiner La Peyre et al. 2001).

METHODS

STUDY AREA

Two semiarid regions were selected for this

study, one located in the Stipa tenacissima

steppes of central and south-eastern Spain,

and the other in the semiarid eucalypt

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

118

woodlands of eastern Australia. We

surveyed 10 (Australia) and 11 (Spain) sites

along climatic gradients ranging from 13º–

17º C mean annual temperature and 273–

488 mm average annual rainfall in Spain,

and from 16º–19º C and 280–630 mm in

Australia. Plot selection included the full

range of environmental conditions existing

within the distribution of both vegetation

communities, as recommended to test the

relationship between plant-plant

interactions, community processes and

environmental gradients (Brooker et al.

2008, Lortie 2010). Within each country the

selected plots shared a common soil type,

management style and vegetation

community type, and were selected with

similar orientation and slopes to minimize

any experimental noise that could

potentially influence the effect of climatic

conditions on the stress experienced by the

plant community. Stipa sites were located

on limestone soils. Vegetation was in all

cases an open steppe (mean cover 35 to

68%) dominated by the perennial tussock

grass Stipa, with sparse resprouting shrubs

such as Quercus coccifera, Pistacia

lentiscus and Rhamnus lycioides. Sites in

south-eastern Australia were open

woodlands located on clay loam soils.

Canopy cover, which ranged from 18-70%,

was dominated by Eucaliptus populnea, E.

intertexta, E. microtheca, Geijera

parviflora and several shrub species

(Eremophila mitchelii, E. sturtii, Dodonaea

viscosa, Acacia spp. and Senna spp.).

Details of the study sites and images of the

communities are given in Appendix H.

VEGETATION SURVEY

At each site we established a 30 m × 30 m

plot, containing the representative

vegetation of the surrounding area. This

plot size permitted the inclusion of several

shrub and tree patches within this area,

enough to conduct the facilitation surveys

described below. Within each plot we

centrally aligned three 30 m long transects,

8 m apart, down the slope for the vegetation

survey. Along each transect we placed 20

contiguous 1.5 m × 1.5 m quadrats, and

recorded the cover and abundance of all

perennial plant species within the quadrat.

These data provided us with a

presence/absence matrix of 80 columns

(four transects by 20 quadrats) for each

plot. The total cover of each plot, which has

been shown to be a good surrogate of

productivity in semiarid environments

(Flombaum and Sala 2009), was derived

from the average cover of perennial plants

across the 80 quadrats. This survey was

used to examine differences in community

composition derived from contrasting

environmental optima of the different

species forming each community across the

CHAPTER 4

119

Figure 4.1. Conceptual diagram synthesizing current facilitation models (upper panel) and our newly proposed model (lower panel). Current facilitation models are the Stress-Gradient Hypothesis (SGH), as originally presented by Bertness and Callaway (1994) and the Facilitation waning models, as a synthesis of new empirical data and several proposed modifications of the SGH (see main text). The SGH predicts an increase in positive interactions either under high consumer pressure or abiotic stress level (parts A, and E-F of the upper panel), being competition more important under moderate conditions (part C of the upper panel). Facilitation waning models propose that positive interactions dominate under high, but not extreme, levels of either abiotic stress or consumer pressure (parts B, D and E of the panel). In contrast with the SGH, these positive interactions collapse when consumer pressure reaches extreme high levels (e.g. Smit et al. 2007; part A in the upper panel) or when abiotic stress reaches this extreme levels (part F in the upper panel). Among facilitation waning models, some of them propose that negative interactions dominate again under such stressful conditions (e.g. Maestre et al. 2009a; discontinuous line), while others suggest that the sign of the interaction becomes neutral (e.g. Michalet et al. 2006). This differentiation is caused by the different explanations invoked to explain this facilitation waning (see main text). Our newly proposed model (Individual-based stress concept; lower panel) does not take into account if the “abiotic stress” or “consumer pressure” are extremely high, moderate or low, because their effects may differ depending on the physiological tolerances of the species involved, and thus are difficult to predict. Instead, we use the distance to the environmental optimum of each target species to predict the effects of nurse plants. As a given

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

120

particular species moves further away from its optimum, the effects of a nurse plant will become more positive (Greyner La Peyre et al. 2001, Holmgren and Scheffer 2010). This interaction will become more negative as the target species gets closer to its environmental optimum. This model also predicts a facilitation collapse when the given environmental conditions are so severe for a given target plants that the recruitment is impossible, even under the safe sites provided by the nurse plants (Kitzberger et al. 2000, Ibañez and Schupp 2001., Soliveres et al. in press)

gradient in each region (see Appendix I for

a detailed description of this analysis).

FACILITATION SURVEY

Because of differences in vegetation

physiognomy, we conducted facilitation

surveys using slightly different protocols in

Spain and Australia. In each plot located in

Spain, ten Stipa tussocks (hereafter Stipa

microsite) were randomly selected, and the

total area under their canopy sampled with

0.5 m × 0.5 m quadrats. The abundance and

cover of all perennial plant species were

recorded within each quadrat. Since most of

the species in the studied areas have

canopies smaller than the 0.5 m × 0.5 m

area, we believe that this is an appropriate

size to evaluate species co-ocurrence on a

distance closer enough to ensure that the

species found were interacting someway.

Ten paired open areas, located at least 1 m

from any Stipa tussock or resprouting shrub

(hereafter Open microsite), were randomly

selected adjacent to these tussocks. We

balanced our sampling effort among

microsites by sampling the same area (i.e

the same number of 0.25 m2 quadrats) of

Open microsites as that sampled under

Stipa. Finally, we sampled the same area

under the canopies of five Quercus

coccifera (or another resprouting shrub

species when Quercus was absent from the

plot; hereafter Shrub microsite).

For the Australian plots, we sampled

three different microsites; Open, Shrub and

Tree. Shrub microsites were represented as

inverse cone-shaped (sensu Whitford 2002)

shrubs such as Eremophila mitchellii,

Dodonaea viscosa, Senna artemisioides or

juvenile Callitris glaucophylla, depending

on the species present in each plot. Our

sampling protocol changed slightly

depending on the canopy size of shrubs.

Where the canopy was sufficiently large,

we sampled six 0.5 m × 0.5 m quadrats

under each of five shrubs. Where shrubs

were smaller, a larger number of shrubs

were sampled in order to sample a total of

30 quadrats. Paired Open microsites (> 1 m

from any shrub or tree) were also sampled

adjacent to these shrubs to yield the same

sampling area. Finally, we sampled the

same number of quadrats under the Tree

microsites, which were represented by

different species of eucalypts (E. populnea,

E. intertexta, E. microtheca), Casuarina

pauper or Geijera parviflora. Because of

the large area occupied by these tree species

CHAPTER 4 .

121

(up to 200 m2 for some Eucalyptus trees),

we sampled only three trees in each plot.

The canopy area of all shrubs and trees was

calculated based on the area of a circle. For

both Spain and Australia, plot-level

richness was calculated by summing up the

total number of different species found in

the 30 0.5 m × 0.5 m, and eighty 1.5 m ×1.5

m quadrats sampled.

STATISTICAL ANALYSES

-Reduction of climatic data

Eight climatic variables (annual radiation,

minimum, maximum and mean

temperature, and rainfall, temperature range

[maximum-minimum temperature], and

minimum and maximum temperatures for

the coldest and warmest month,

respectively) were collected for each site

using available climatic models (Ninyerola

et al. 2005) and data from the Bureau of

Meteorology (www.bom.gov.au) in Spain

and Australia, respectively. We reduced

these climatic variables to a single synthetic

variable for each country using PCA.

Summarizing environmental variables in a

PCA allowed us to obtain a more general

assessment of the influence of all of our

environmental variables at both community

and species-specific levels. This approach is

strongly recommended for testing

relationships between plant-plant

interactions and abiotic stress along

environmental gradients (Lortie 2010). We

used the first PCA axis as our surrogate for

the climatic gradient present at our sites in

both countries (hereafter referred to as

Climate). This axis explained 88.6%

(Eigenvalue = 8.08·103) and 86.2%

(Eigenvalue = 1.07·104) of the variance in

the climatic data for Spain and Australia,

respectively. This axis was highly

correlated with rainfall and radiation in

Spain (Eigenvectors = -0.864 and 0.502 for

rainfall and radiation, respectively; the

remainder of the eigenvectors were < 0.03

in all cases) but only with rainfall in

Australia (Eigenvector = 0.996; the rest of

eigenvectors were < 0.1 in all cases).

Principal Component Analyses were carried

out in Primer v. 6 statistical package for

Windows (PRIMER-E Ltd., Plymouth

Marine Laboratory, UK).

We evaluated the relationships

between Climate and both cover and

species richness at the plot level, and the

relationship between cover and richness,

using both linear and quadratic regressions

because either linear or unimodal

relationships between these variables are

expected from previous studies (e.g. Grime

1973, Whitford 2002). Regression analyses

were carried out using SPSS 13.0 for

Windows (Chicago, Illinois, USA).

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

122

-Assessing biotic interactions

Because plant-plant interactions cannot be

measured in a simple way (Maestre et al.

2005, Brooker et al. 2005), we applied three

complementary indices to quantify their 1)

intensity, 2) importance and 3) frequency.

The intensity and importance of plant-plant

interactions were assessed using both plant

richness and cover. To measure the

intensity of the interactions, i.e., the effect

that neighbours have on species richness

and cover regardless of other environmental

factors (Brooker et al. 2005), we used the

Relative Interaction Index (RII; Armas et

al. 2004). This index is calculated for each

microsite pair as (PNurse – POpen)/(PNurse +

POpen), where PNurse is either mean cover or

mean species richness under the canopy of

a nurse plant (Stipa, Shrub or Tree

microsites) and POpen is either mean cover

or mean species richness in the Open

microsite. This index has good statistical

properties, which make it suitable for

comparing the intensity of plant-plant

interactions across environmental gradients;

it has defined limits (-1,+1), is symmetrical

around zero, and has identical absolute

values for competition and facilitation. It is

also linear, unbiased at low intensity

interactions, and has no discontinuities in

its range (Armas et al. 2004). For each plot

we calculated the mean index obtained from

all the Nurse-Open microsite pairs sampled.

To assess the importance of plant-

plant interactions, i.e., the relative effect

that Stipa, shrubs and trees had on richness

and cover compared to that of other

environmental factors (Brooker et al. 2005),

we used the Interaction Importance Index

(Iimp; Seifan et al. 2010b), which has similar

statistical properties to RII and is therefore

comparable among sites located across the

environmental gradient sampled. This index

is calculated as Iimp= Nimp/│Nimp│+│Eimp│,

where Nimp and Eimp are the nurse plant and

environmental contributions to species

richness or total cover, respectively. Nimp is

calculated as PNurse – POpen, and Eimp as POpen

– MPOpen/Nurse, where MPOpen/Nurse is the

maximum value of species richness or

mean cover found in the entire gradient,

irrespective of the microsite sampled.

Finally, the frequency of positive

interactions was measured as the percentage

of either facilitation obligates and

facilitation beneficiaries (sensu Butterfield

2009), as a percentage of the total species

pool of each plot. We considered as

facilitation obligates those species found

only under the canopy of a given nurse

plant but not in the Open microsites

(regardless of the identity of the nurse

plant), while facilitation beneficiaries were

species with more individuals growing

under the canopy of a nurse than in the

Open microsites. We used the number of

recruited individuals because seedling

CHAPTER 4 .

123

germination and establishment, particularly

during the first year, are known to be the

principal bottlenecks in plant recruitment in

semiarid environments, such as those

sampled (Eldridge et al. 1991, Escudero et

al. 1999, Maestre et al. 2001). Thus, we

believe that a higher number of individuals

recruiting in a given microsite is indicative

of superior environmental conditions

experienced in this microsite for a given

species, an approach followed by previous

studies on the topic (Valiente-Banuet et al.

2006).

The three attributes used to quantify

plant-plant indicators (intensity, importance

and frequency) were plotted against

Climate and mean nurse size for each plot

(i.e. the area occupied by the nurse plants).

These relationships were tested with both

linear and quadratic regressions, following

predictions from previous facilitation

models (Bertness and Callaway 1994,

Michalet et al. 2006, Maestre et al. 2009a).

In the case of frequency, we calculated the

percentage of obligate and beneficiary

species for each microsite and country

separately, and then for the whole

community. For the whole community we

added together data for both nurse

microsites (Stipa and Shrub, Spain; or

Shrub and Tree, Australia) and also tested

them separately for each country. This

battery of approaches provided us with a

complete assessment of the relationships

between plant-plant interactions and abiotic

stress. Furthermore, it allowed us to

evaluate the importance of nurse area as a

driver of this relationship. This area

influences the effects of nurses on

microclimatic amelioration and niche

availability, and thus affects the richness

and cover of understorey plants (Pugnaire et

al. 1996a, Maestre and Cortina 2005,

Michalet et al. 2006).

Furthermore, to assess the relative

role of abiotic stress and nurse size at the

pairwise level, we selected for each country

separately, those species present in at least

three different sites across the sampled

environmental gradient, which ensured that

we had at least three points with which to

test the relationship between plant-plant

interactions and abiotic stress using climatic

features, as recommended (Lortie 2010).

With the selected species (16 and 13 for

Spain and Australia, respectively), we

calculated Iimp and RII for all the species

using cover of the target species as our

proxy of plant performance. If the intensity

or importance of facilitation increase with

abiotic stress, as predicted by the “stress-

gradient hypothesis” (SGH; Bertness and

Callaway 1994, Callaway 2007), we would

expect to detect an increasing trend of

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

124

positive interaction outcomes with

increasing stress. Alternatively, if

facilitative interactions collapse at

extremely high stress levels (i.e. low

rainfall), we would expect to find a

unimodal relationship between facilitation

indicators and climate, which may or may

not be related to nurse size (facilitation

waning; Maestre and Cortina 2004a,

Michalet et al. 2006, Maestre et al. 2009a).

To test the generality and validity of these

three models (SGH, facilitation waning

because of smaller nurses, and facilitation

waning because of increasing competition),

we evaluated the relationships between RII,

Iimp obtained with all the pairwise

interactions tested in each country, and the

climatic PCA axis or the nurse size using

Spearman correlations. To assist us in the

interpretation of these analyses, the

percentage of pairwise interactions that

followed each model prediction was also

calculated for each country and nurse plant.

If the Spearman correlation with Climate is

significant and negative, this will give

support to the SGH, if the relationship

between the interaction indices and Climate

is unimodal, this will give support to

facilitation waning models. If the latter is

also related with nurse size, this will

support the facilitation collapse derived

from nurse plant growth limitation. The

percentage of cases explained for a given

model is another measurement of the fitness

of each model to our data.

-Measuring changes in plant-plant

interaction outcomes depending on

microsite

Nurse plants may affect the competitive

outcomes of their understorey plants by

increasing competitive intransitivity or by

niche segregation (Brooker et al. 2008).

Both mechanisms are related to the

maintenance of higher species richness than

if competitive exclusion alone dominates

interactions among understorey plants

(Silvertown 2004, Laird and Schwamp

2006). Recent models have highlighted the

importance of competition intransitivity as

a key modulator of species richness (Laird

and Schamp 2006, 2008, Bowker et al.

2010). The degree of intransitivity can be

defined as the absence of a competitive

hierarchy among the species coexisting in a

community (Gilpin 1975). However, this

concept assumes that the competitive ability

of those species making up a given

community are constant along the whole set

of possible environmental conditions. To

adequately test intransitivity it is necessary

to measure the competitive ability of every

species against every other one (Grace et al.

1993), making it exceptionally difficult to

test empirically. Alternatively, niche

segregation has been identified as a crucial

mechanism increasing species richness (e.g.

CHAPTER 4 .

125

Silvertown 2004), and is more likely to

occur under the most heterogeneous

conditions found beneath nurse plants,

rather than in Open microsites (Pugnaire et

al. 1996a, Maestre and Cortina 2005).

However, niche segregation may change

along environmental gradients (Huston

1999), and is also difficult to measure

empirically when considering the whole

plant community.

We attempted to measure changes in

competition intransitivity and niche

segregation, and their relationship with the

plot-scale diversity of species found in each

microsite, by using null models of guild

structure based on patterns of species co-

occurrence (Gotelli and Graves 1996,

Gotelli et al. 2010). These null models are

organized a priori by groups of ecological

significance, such as different functional

groups or trophic levels (i.e. species guilds),

and allow testing of the role of competition

in structuring the community within each

guild separately (Gotelli and Graves 1996).

This analysis is not limited to grouping by

species guild. In reality, any type of a priori

group could be examined.

For co-occurrence analyses, we

organized our presence/absence data

(obtained from the 0.5 m × 0.5 m quadrats)

by microsite guilds, that is, we calculated

species co-occurrence (C-score index,

explained below) independently for each of

the Stipa, Shrub and Tree microsites by

pooling all the sampled quadrats of these

microsites within each plot (n = 30),

obtaining a unique value per microsite and

per plot. Most of the species sampled in

Spain and Australia are small shrubs or

grasses, and therefore the quadrat size used

is particularly suitable to include

interactions among them without including

random co-ocurrence or exclusions not

related to competition among them. We

estimated species co-occurrence with the C-

score index, a metric commonly used in this

kind of analyses (e.g. Dullinger et al. 2007,

Maestre et al. 2008, Rooney et al. 2008).

This index is calculated for each pair of

species as (Ri - S)(Rj - S), where Ri and Rj

are the number of total occurrences for

species i and j, and S is the number of

quadrats in which both species occur. This

score is then averaged over all possible

pairs of species in the matrix (Gotelli

2000). The C-score is related to the

competitive exclusion concept of

“checkerboardness” i.e., how many of the

possible species pairs in a given community

never appear in the same quadrat together.

Thus, positive and large values of this index

indicate that competition may be the

prevalent mechanism determining the co-

occurrence patterns observed (Gotelli

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

126

2000). To determine the strength of co-

occurrence in a sample, the observed C-

score value is compared against a set of null

models which serve as a baseline for what a

community unstructured by species

interactions would look like (Connor and

Simberloff 1979). As the values of the C-

score are dependent on the number of

species and co-occurrences observed within

each plot, we obtained a standardized effect

size (SES) as (Iobs - Isim) ⁄ Ssim, where Iobs is

the observed value of the C-score, and Isim

and Ssim are the mean and standard

deviation, respectively, of this index

obtained from the n simulations performed

(Gotelli and Entsminger 2006).

We used ‘fixed rows–equiprobable

columns’ null models and 5000

simulations. With this approach, each

species conserved its own abundance (rare

species remained rare and common species

remained common) and each quadrat was

assumed to have the same probability of

being colonized as the remainder,

regardless of the number of species found

in each quadrat, during the simulations.

This null model has been recommended for

standardized samples collected in

homogenous habitats (Gotelli 2000), such

as the ones gathered in this study. We also

used the “fixed rows-fixed columns”

algorithm (both species and quadrats

conserved their relative abundance and

richness, respectively) to add confidence to

our conclusions. The results obtained with

this analysis were similar to those obtained

with the ‘fixed rows–equiprobable

columns’ null model, and thus are not

shown.

Standardized Effect Size (SES)

values of the C-score less than or greater

than zero indicate prevailing spatial

segregation and aggregation among the

species within a community, respectively.

To assess the extent to which changes in

competitive outcomes affect local diversity,

we compared the SES obtained with the

plot-level richness found in each microsite

(hereafter ‘plot richness’). The logic

underlying the use of SES and plot richness

values to measure competition intransitivity

or niche segregation is that we assume that

SES will be higher when competitive

exclusion is more important at the quadrat-

scale. High SES values can lead to two

different outcomes: 1) a reduction in plot

richness because a few dominant species

occupy all of the available space (i.e. when

the differences in competitive ability among

co-existing species are high, competitive

transitivity leads to competitive exclusion

and low diversity), or 2) an increase in plot

richness, if there is a lack of a competitive

hierarchy and competitive dominants in

each quadrat, depending on the

microenvironmental conditions existing in

each different quadrat. In this case, a high

quadrat-scale competition will generate

CHAPTER 4 .

127

high turnover/heterogeneity in the

dominance of a given species, ultimately

leading to a high plot richness

(intransitivity increases plot richness; Laird

and Schwamp 2006, 2008). Alternatively,

SES values will be lower if 1) the

competitive ability of co-existing species is

more equilibrated at the quadrat-scale, and

thus niche segregation prevails, or 2) spatial

aggregation, and therefore positive effects

of the nurse on their understorey species,

prevail (Tirado and Pugnaire 2005). If we

analyze the effects of nurse plants on SES

and plot richness separately, there are four

possible responses: 1) Nurse plants have a

joint effect reducing SES and increasing

plot richness compared to Open microsites:

these plants promote the development of

understorey/neighbour species via niche

segregation and this has positive effects on

the overall plot-scale richness, 2) Nurse

plants increase both SES and plot richness:

these plants increase quadrat-scale

competition, but species with competitive

advantage vary among quadrats, generating

a high species turnover, and therefore

increasing plot-scale richness

(intransitivity), 3) Nurse plants increase

SES and reduce plot richness: competitive

exclusion is the dominant interaction

between understorey species and a smaller

set of competition winners dominate all

quadrats, 4) Nurse plants do not affect SES,

regardless of their effects on plot richness:

changes in the competitive outcomes are

not an important factor modulating the

effect of nurses on plot-scale diversity.

Differences in SES and plot

richness values obtained among microsites

were compared using separate one-way

ANCOVAs for each variable. In these

models, microsite (Open, Stipa and Shrub –

Spain; Open, Shrub and Tree –Australia)

was introduced as a fixed factor, and mean

plot cover (our surrogate of overall site

productivity) was used as a covariate.

Standardized Effect Size data were √(x+1)

transformed to meet assumption of

ANCOVA analyse (normal distribution of

residuals and homoscedasticity). Tukey’s

HSD post-hoc tests were used to assess for

differences among the three microsites of

each country. We tested for relationships

among the residuals of the ANCOVA and

nurse plant canopy area using Spearman

correlations. This was necessary in order to

assess the importance of nurse size, as we

could not use nurse size as a covariate in

our model because Open microsites do not

have a size and the relationships between

nurse area and their effect on SE and plot

richness might not necessarily be linear.

To add confidence to our results,

we also developed an alternative approach

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

128

to detect the relative importance of

competitive intransitivity and niche

segregation based on a modification of

Whittaker´s beta-diversity index (Whittaker

1972). Results from this approach were

very similar, and only suggested that it was

more sensible to detect niche segregation

(see Appendix J for methodological details

and results). Null model analyses were

conducted with Ecosim 7.22 (Gotelli and

Entsminger 2006). ANCOVA analyses

were carried out using SPSS 13.0 for

Windows (Chicago, Illinois, USA).

SPAIN AUSTRALIA

mean plot cover (%)

30 40 50 60

spec

ies

richn

ess

0

10

20

30

40

50

A

R2 = 0.66; P = 0.002

mean plot cover (%)

10 20 30 40 50 60 70 80

spec

ies

richn

ess

10

20

30

40

50

60

B

-200 -150 -100 -50 0 50 100

spec

ies

richn

ess

0

10

20

30

40

50

sp richness: R2 = 0.64; P = 0.017

cover: R2 = 0.77; P = 0.003

- RAINFALL +

+ MINIMUM TEMPERATURE/RADIATION -

Cov

er (

%)

30

35

40

45

50

55

60

65

C

- RAINFALL +

-200 -100 0 100 200 300

spec

ies

richn

ess

10

20

30

40

50

60

Cover: R2 = 0.4; P = 0.047

D

Cov

er (

%)

10

20

30

40

50

60

70

80

Figure 4.2. Relationships between cover, our surrogate of standing biomass, and species richness at the community level in Spain (A) and Australia (B), respectively. The relationship between both cover (open dots, continuous line) and richness (black dots, dashed line) and the first axis of a PCA derived from climatic variables is shown for both Spain (C) and Australia (D), respectively. Significant relationships (P < 0.05) are shown as bold lines.

CHAPTER 4

129

-Assessment of niche expansion

We calculated the Chao-Jaccard abundance-

based similarity index (hereafter Chao

index; Chao et al. 2005) to assess the

relative role of niche expansion by nurse

plants on community species richness. This

index is based on the probability that two

randomly chosen individuals, one from

each of two samples (referred to as a

“pair”), belong to species shared by both

samples. It takes into account not only the

number of shared species among different

microsites, but also differences in their

relative abundances (Chao et al. 2005). We

assume that, as the influence of niche

expansion increases, more species should

be present, or more abundant, under a given

nurse plant than in Open microsites. We

calculated the Chao index at the community

level by summing over the number of

individuals and species recorded in all

quadrats for a given microsite within each

plot. Thus, the higher the dissimilarity

among nurse microsites and open areas in a

given plot, the higher the effect of niche

expansion provided by nurse canopies on

the overall community richness in this plot.

Differences in the Chao index between

microsite pairs (Stipa/Shrub vs. Open,

Shrub/Tree vs. Open, Stipa/Shrub vs.

Shrub/Tree for Spain and Australia,

respectively) were compared with one-way

ANOVA, with microsite pair as fixed

factor. Tukey’s post-hoc HSD tests were

used to assess significant differences among

pairs. After conducting the analysis, and to

assess the influence of climate in niche

segregation, we evaluated the relationship

between Climate and Chao index using

both linear and quadratic regressions. With

both approaches we can correctly evaluate

the differences in the understorey

populations between each microsite (each

nurse-type may have different effects on a

given target plant, and this may translate in

a high dissimilarity not only between

Nurse/Open microsites, but also between

different nurses), and to account for the

possible non-linear relationships between

Climate and the effects of the different

nurse plants tested on their understorey

vegetation. The Chao index was calculated

using EstimateS 8.2.0 for Windows

(Colwell 2000;

http://viceroy.eeb.uconn.edu/estimates).

ANOVA and correlation analyses were

carried out using SPSS 13.0 for Windows

(Chicago, Illinois, USA).

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

130

Table 4.1. Relationship between the indicators of interaction intensity (Relative Interaction Index, RII) and importance (Interaction Importance Index, Iimp) calculated for particular pairwise interactions and Climate. Particular pairwise interactions for both studied regions and the two different nurse microsites tested in each region are included (Stipa/Shrub and Shrub/Tree in Spain and Australia, respectively). The relationship between RII/Iimp and Climate can be nil (0), monotonically positive or negative (+ and -, respectively) or hump-shaped (hs). The last column indicates the number of sites along the gradient in which each species was present.

Spain RII

Stipa

RII

Shrub

Iimp

Stipa

Iimp

Shrub Sites

Asparragus horridus + + 0 - 3

Asphodelus ramosus hs Hs hs hs 4

Brachypodium retusum 0 0 0 + 6

Cistus clusii - - 0 - 6

Fumana ericoides 0 - + 0 7

Fumana thymifolia 0 0 0 0 6

Helianthemum cinereum 0 - - 0 6

Helianthemum violaceum 0 - 0 - 5

Rosmarinus officinalis 0 Hs 0 0 8

Sedum sediforme 0 - 0 - 6

Stahelina dubia - + hs hs 4

Stipa offneri 0 + 0 hs 5

Teucrium capitatum 0 0 0 0 6 Teucrium

pseudochamaepytis hs Hs - 0 9

Thymus vulgaris 0 0 0 hs 11

Thymus zygis hs + 0 0 3

Australia RII

Shrub

RII

Tree

Iimp

Shrub

Iimp

Tree Sites

Austrodanthonia caespitosa 0 - - - 4

Austrostipa scabra 0 - 0 0 7

Boerhavia dominii 0 0 hs hs 4

Cenchrus ciliaris - - - 0 4

Chenopodium desertorum - - hs 0 4

Einadia nutans 0 0 0 + 6

Enteropogon acicularis 0 - 0 0 6

Maireana enchylaenoides hs 0 0 0 3

Maireana sclerolaenoides hs - hs + 4

Sclerolaena muricata - - + + 5

Sida corrugata - 0 0 + 5

Sida cunninghamii - - 0 0 4

Vittadinia cuneata 0 + 0 + 4

CHAPTER 4

131

Table 4.2. Percentage of pairwise interaction outcomes, as measured with the Relative Interaction Index (RII) and the Interaction importance index (Iimp), explained by the two major current models tested in this work, the Stress-Gradient Hypothesis and the waning of facilitation under extremely high stress (Waning). As a special case in the facilitation waning model, the percentage of cases that are positively related with nurse size (compared with the total number of cases tested) are indicated below. st = Stipa tenacissima, sh = resprouting shrubs in Spain (SP) or inverse cone-shaped shrubs in Australia (AU), and tr = Eucalyptus or Geijera parviflora trees.

Model

RII Iimp

SP

(st)

SP

(sh)

AU

(sh)

AU

(tr)

SP

(st)

SP

(sh)

AU

(sh)

AU

(tr)

Stress-Gradient Hypothesis 23.5 29.4 38.5 61.5 17.6 29.4 15.4 7.7

Facilitation

Waning

23.5 17.7 23 0 11.8 23.5 15.4 7.7

Nurse size 19 0 23 0 11.8 0 15.4 7.7

RESULTS

DIVERSITY-BIOMASS RELATIONSHIP

AND THE EFFECT OF CLIMATE

A total of 96 and 131 perennial species

were found in Spain and Australia,

respectively, with a plot-level richness

ranging from 9 to 47 in Spain, and from 16

to 51 in Australia, respectively. The number

of species found in each plot was linearly

and negatively related to mean plot cover in

Spain (Fig. 4.2A). This was particularly

evident near the centre of the climatic

gradient, under conditions of both moderate

drought and moderately low temperatures

(Fig. 4.2C). In contrast, richness was

largely independent of either cover or

climate in Australia (Figs. 4.2B and 4.2D).

Mean cover at the plot level showed a

monotonic and positive increase with water

availability in both Spain and Australia,

reaching its maximum value in the more

mesic plots.

PLANT-PLANT INTERACTIONS AND

ABIOTIC STRESS

Contrasting results were found between the

studied regions. While neither intensity nor

importance of plant-plant interactions were

related to abiotic stress in Spain, we

detected a hump-shaped relationship

between most RII and Iimp values and

rainfall for the Australian sites (Fig. 4.3).

Both the percentage of facilitation

beneficiaries and obligates tended to

decrease with rainfall. This negative trend

was found to be significant for beneficiaries

in Spain and for obligates in Australia (Fig.

4.4). The percentage of plants with more

individuals under the canopy of any nurse

plant than in Open microsites (facilitation

beneficiaries) decreased from about 50% in

the drier and warmer sites to about 30% in

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

132

SPAIN AUSTRALIA

-200 -100 0 100 200 300-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Shrub (richness) Tree (richness) Shrub (cover)Tree (cover)

-200 -100 0 100 200 300-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

B

D

Tree (cover): R2= 0.7; P = 0.005

Tree (richness): R2= 0.49; P < 0.05

Shrub (cover): R2= 0.56; P = 0.055

Tree (cover): R2= 0.57; P = 0.053

Tree (richness): R2

= 0.7; P < 0.05

-200 -150 -100 -50 0 50 100

Inte

ract

ion

inte

nsity

(R

II)

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

Stipa (richness)Shrub (richness)Stipa (cover)Shrub (cover)

-200 -150 -100 -50 0 50 100

Inte

ract

ion

impo

rtan

ce (

Iimp)

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

A

C

Climatic PCA axis values

Figure 4.3. Relationships between the indicators of interaction intensity (Relative Interaction Index, RII) and importance (Interaction importance index, Iimp), calculated for both community richness and cover, and the first axis of a PCA derived from climatic variables. Values are means ± SE per plot and microsite. Significant (P < 0.05) relationships are shown with a bold line; marginally significant ones (0.05 < P < 0.10) are also showed in the figure.

the wetter sites for Spain. This relationship

was different, however, for Australia, where

facilitation obligate plants followed a

unimodal relationship with climate. Only

10% of the species at the community level

required a nurse plant to occur at the wettest

sites, but this percentage increased up to

40% in the driest sites, and showed a

maximum (ca. 60%) at the middle of the

environmental gradient (Fig. 4.4). When we

tested the relationship between frequency of

positive interactions of each nurse plant and

climate separately, we found different

results depending on the nurse plant

analyzed at both studied regions. We

detected a similar relationship with climate

at the community level for Spanish Shrub

microsites and Australian Tree microsites,

but there were no significant relationships

for Stipa (Spain) or Shrub (Australia)

microsites. We found a marginally

significant linear relationship between nurse

CHAPTER 4

133

SPAIN AUSTRALIA

Shrub: R2 = 0.58; P < 0.005

Total: R2 = 0.61; P < 0.05

B

D

-200 -100 0 100 200 300

% fa

cilit

atio

n ob

ligat

es

0

10

20

30

40

50

60

70

ShrubTreeTotal

Climatic PCA axis values

-200 -100 0 100 200 300

% fa

cilit

atio

n be

nefic

iarie

s

0

20

40

60

80

100-200 -150 -100 -50 0 50 100

0

10

20

30

40

StipaShrubTotal

-200 -150 -100 -50 0 50 1000

10

20

30

40

50

60

C

ATree: R2 = 0.51; P = 0.02

Total: R2 = 0.61; P = 0.036

+ RAINFALL - Figure 4.4. Relationships between the percentage of facilitation beneficiaries (species with more individuals recruiting under nurse plants than in Open microsites) and facilitation obligates (species that only recruit under the canopy of nurse plants), regarding total species richness in each plot, and the first axis of a PCA derived from climatic variables. Significant (P < 0.05) relationships are shown with a bold line.

area and the percentage of the species

growing under Stipa (R2 = 0.34; P =

0.061); the same relationship for Shrub

microsites was not significant.

Relationships among nurse size and

interaction indices were not significant for

any microsites in Australia.

When testing the validity of the Stress-

Gradient hypothesis or alternative models

to predict pairwise interaction outcomes,

the Spearman correlations between

interaction indicators (RII and Iimp) and

Climate showed that facilitation tended to

decrease monotonically with rainfall in

three of the four cases in Australia (ρ

ranged from -0.27 to -0.46, P < 0.05 in all

the cases except Tree Iimp; see Fig. 4.5).

However, these interaction indicators were

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

134

not significantly correlated with Climate in

Spain (P > 0.6 in all cases). When we

visually assessed each pairwise interaction,

RII values appeared to decline with rainfall

in 38% of the total cases studied, according

to the SGH (Table 4.2). The relationship

between RII and Climate was more

common in Australia (Shrub: 38.5%, Tree:

61.5%) than in Spain (Stipa: 23.5%;

Shrub: 29.4%), which adds confidence to

the results obtained with the overall

Spearman correlations. In contrast, the

relationship between Iimp and Climate was

lower, in general. Iimp showed a

decreasing trend with rainfall in 18%

(Stipa) and 29% (Shrub) of cases for

Spain, and in 15% (Shrub) and 8% (Tree)

of cases for Australia, respectively. The

intensity and importance of the interactions

showed a hump-shaped relationship when

plotted against Climate in 20% and 14% of

cases for Spain and Australia, respectively

(Table 4.2). These cases provided support

for the facilitation waning model.

However, when we tested the relationships

between RII, Iimp and nurse size with

Spearman correlations, none of these

indices was correlated with nurse size in

any of the nurse microsites tested in both

countries (ρ < 0.2 in all cases).

Competitive exclusion, as indicated by

SES values, was lower under the canopy of

both Stipa and Shrub than in Open

microsites along the entire Spanish

gradient (Open: 1.03 ± 0.35; Stipa: 0.12 ±

0.14, Shrub: 0.03 ± 0.16; means ± SE; F2,29

= 3.01; P = 0.008). Tukey’s HSD post-hoc

tests revealed differences between Stipa

and Open (P = 0.028) and Shrub and Open

(P = 0.015) microsites, but not between

Stipa and Shrub microsites (P = 0.961).

Mean plot cover did not affect SES results,

but when analyzing the effects of microsite

on plot-level richness, we found a

significant effect of cover (F1,29 = 53.9; P <

0.0001) and a marginal positive effect of

microsite (F2,29 = 3.01; P = 0.065). Overall,

these results suggest an effective niche

expansion by nurse plants, indicating an

increase in local richness under both

nurses. However, this positive effect

decreased substantially with mean plot

cover (Spearman correlation between

residuals of the ANOVA and mean plot

cover = -0.78; P < 0.0001). Nurse size

showed no relationship with the residuals

of the ANCOVA models fitted with both

SES and plot richness in Spain or in

Australia. Differences in competitive

exclusion among microsites, as measured

with SES, were not found for the

Australian sites (F2,27 = 2.4; P = 0.101).

CHAPTER 4

135

SPAIN AUSTRALIA

-200 -100 0 100 200 300

-1.0

-0.5

0.0

0.5

1.0

Climatic PCA axis values

-200 -150 -100 -50 0 50 100

Inte

ract

ion

inte

nsity

(R

II)

-1.0

-0.5

0.0

0.5

1.0

StipaShrub

-200 -150 -100 -50 0 50 100

Inte

ract

ion

impo

rtan

ce (

Iimp)

-1.0

-0.5

0.0

0.5

1.0-200 -100 0 100 200 300

-1.0

-0.5

0.0

0.5

1.0

ShrubTree

- RAINFALL +

A

C

B

D

Shrub: ρ = -0.46; P < 0.0001Tree: ρ = -0.34; P = 0.003

Shrub: ρ = -0.27; P = 0.02

Figure 4.5. Scatter plot showing the relationship between pairwise facilitation indicators (intensity [RII, panels A and B for Spain and Australia, respectively] and importance [Iimp, panels C and D for Spain and Australia, respectively]) and Climate. Spearman correlation coefficients and P values are shown in each case. However, microsite significantly affected

plot richness in this country (F2,26 = 4.05; P

= 0.03). While the highest richness was

found in Shrub microsites (plot richness =

17±1.7 and 11 ± 1.5 for Shrub and Open

microsites, respectively; mean ± SE; Tukey

HSD: P = 0.029), this effect was less

marked in Tree microsites (plot richness =

15.8 ± 1.5; Tukey HSD: P = 0.093 for Tree

vs. Open microsites). In contrast to the

results found in Spain, mean plot cover did

not modify the effect of microsite on SES

or plot richness.

NICHE EXPANSION

Although the similarity index was slightly

lower for Shrub vs. Open microsites (0.64 ±

0.08, mean ± SE, n = 11) than for Stipa vs.

Open microsites (0.74 ± 0.08), we did not

find significant differences in the similarity

index among nurse microsites in Spain

(Stipa vs. Shrub, 0.68 ± 0.06, n = 11;

ANOVA: F2,29 = 0.622; P = 0.54). Shrub

and Stipa microsites shared about 70% of

their understorey populations with Open

sites. Significant differences in similarity

among microsite pairs were found,

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

136

however, in Australia (F2,26 = 4.57; P =

0.02), suggesting some degree of niche

expansion. Populations shared between

Shrub-Open or Shrub-Tree pairs were close

to 60%, though the similarity index for

Tree-Open microsites showed that they

shared about 40% of the individuals (Tukey

HSD post-hoc test; P = 0.034, 0.710 and

0.171 for Open-Shrub vs. Open-Tree,

Open-Shrub vs Shrub-Tree, Open-Tree vs

Shrub-Tree pairs, respectively).

-200 -150 -100 -50 0 50 100

Cha

o-Ja

ccar

d si

mila

rity

inde

x

0.0

0.2

0.4

0.6

0.8

1.0

Open vs Stipa Open vs Shrub Stipa vs Shrub

SPAIN

-200 -100 0 100 200 3000.0

0.2

0.4

0.6

0.8

1.0

Open vs ShrubOpen vs TreeShrub vs Tree

AUSTRALIA

Climatic PCA axis values

- RAINFALL +

Open vs Tree: R2 = 0.64; P = 0.028Shrub vs Tree: R2 = 0.63; P = 0.032

Figure 4.6. Relationships between Chao-Jaccard similarity index, our surrogate of niche expansion, and Climate in Spain and Australia. Data from the three different microsite pairs for each country (Open vs Stipa, Open vs Shrub, and Stipa vs Shrub for Spain; Open vs Shrub, Open vs Tree, and Tree vs Shrub for Australia) are shown. Significant relationships (P < 0.05) are shown as bold lines.

Overall, these results indicate that, while

Shrub microsites shared 60% of the

populations beneath their canopies with

both Tree and Open microsites, the

similarities between Tree and Open

populations were lower (ca. 40% shared).

Similarity among the microsites tested did

not show any relationship with Climate in

Spain, but two of the three indices

calculated showed a quadratic relationship

with this variable in Australia (Fig. 4.6),

suggesting a trend of increasing similarity

at both ends of the climatic gradient.

DISCUSSION

SPECIES RICHNESS-PRODUCTIVITY

RELATIONSHIP AND THE CONCEPT

OF STRESS

The results of our studies from both Spain

and Australia did not conform with the

hump-shaped relationship between richness

and productivity predicted by pioneering

studies (Grime 1973, Huston 1979). Several

reviews and meta-analyses have questioned

the universality of this hump-shaped

relationship (Grace 1999, Waide et al.

1999, Gillman and Wright 2006). Although

the unimodal richness-productivity

relationship is rarely proven empirically, it

continues to be used to invoke the role of

facilitative interactions on increasing plant

community diversity (Hacker and Gaines

1997, Michalet et al. 2006). Previous

CHAPTER 4 .

137

studies have described the roles of stress-

tolerant plants on niche expansion of more

competitive species (Lortie et al. 2004a,

Travis et al. 2005, Michalet et al. 2006).

However, we argue that, as different species

are adapted to different environmental

conditions, their environmental optima will

occur at different points along any given

productivity gradient (Chapin et al. 1987,

Ibañez et al. 2007, Holmgren and Scheffer

2010; see also Appendix I). A more general

framework should consider the species-

specific nature of ‘stress’ (Körner 2003)

and the ‘distance’ of each species from its

ecological optimum, rather than the

ecological strategy that it employs. Given

the absence of a clear overarching stress

level affecting whole plant communities

along environmental gradients (Chapin et

al. 1987), such a framework is clearly

needed. Our results suggest that plant-plant

interactions not only enhance diversity at

mid to low productivity levels (sensu

Hacker and Gaines 1997, Michalet et al.

2006), but that this effect extends to the

entire productivity gradient via

environmental amelioration by nurse plants

of the less adapted species to a given set of

environmental conditions (Holmgren and

Scheffer 2010). This positive effect is

reinforced because nurse plants not only

increase the available species pool in a

given plot, but also promote the existence

of a richer community because of greater

environmental heterogeneity (Pugnaire et

al. 1996a, Maestre and Cortina 2005).

PLANT-PLANT INTERACTIONS

ALONG ENVIRONMENTAL

GRADIENTS

Community productivity in both Spain and

Australia was limited mainly by water

availability, consistent with the

expectations for arid and semiarid

environments worldwide (Noy-Meir 1973,

Whitford 2002; Fig. 4.2). Thus, the

negative trend found in the frequency of

positive interactions with increasing rainfall

in both areas provides strong support for the

original predictions of the SGH (Bertness

and Callaway 1994). However, at the

Australian sites, where environmental stress

seemed to reach extremely high levels (plot

cover declined to 17% in some cases; see

Appendix H), positive effects of nurses on

community richness decreased at the

highest stress levels, consistent with results

from other semiarid environments

(Kitzberger et al. 2000, Maestre and Cortina

2004a, Anthelme et al. 2007). We found

some evidence that increased nurse size

could explain increased facilitative effects

of Stipa in the Spanish sites. Percentage of

facilitation obligate species and the

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

138

intensity of the positive effects on

community diversity increased with Stipa

size, suggesting an increased capacity to

ameliorate harsh environmental conditions

by larger tussocks. However, in contrast to

theoretical predictions (Michalet et al.

2006), nurse area was unrelated to either

importance, intensity or frequency of

positive interactions at the community level

in the other microsites tested. The model

developed by Michalet et al. (2006) was

based on empirical data using mostly

tussock-like nurses, plants similar to Stipa,

or cushion-like plants (e.g. Choler et al.

2001, Callaway et al. 2002, Liancourt et al

2005, Anthelme et al. 2007). Small

increases in the size of these nurses are

likely to have relatively large facilitatory

effects on understorey species (Michalet et

al. 2006, Anthelme et al. 2007). However,

relatively small changes in large nurse

plants, such as the Australian eucalypt trees

(with canopy areas up to 200 m2), would be

less likely to be influential. We suggest

caution, therefore, in generalizing the

effects of nurse plants in semiarid

environments without a consideration of

their size. Even when the smallest nurse

plants almost always exceed an area of 1.5

m2, such as the microsites tested in this

studies (excepting Stipa), we would not

expect a strong relationship between nurse

size and their tendency to alleviate

environmental stress. An alternative

explanation for the positive effects of nurse

size found in pairwise interactions (Table

4.2) could result from increased

heterogeneity in microclimate (e.g. shade,

temperature, light) at different parts of the

canopy, which would be expected to

increase niche segregation (Pugnaire et al.

1996a, Maestre and Cortina 2005). Thus,

the higher availability of different niches

and the associated decreases in inter-

specific competition via niche segregation

(Huston 1979, see discussion below) could

enhance the performance of particular

species more than any increase in the ability

of the nurse to buffer environmental

stressors per se.

Our results demonstrated

inconsistent relationships between

environmental stress and both interaction

intensity and importance, contrary to

prevailing facilitation theory (Brooker et al.

2005, Callaway 2007). In Spain, neither

interaction intensity nor importance were

related to the environmental gradient we

evaluated, but a hump-shaped relationship

for both facilitation/competition indicators

was found in Australia. How can we

account for this difference? Environmental

stress in the Spanish gradient was driven by

two negatively correlated and distinct

stressors; water and radiation/temperature.

Thus, it is likely that in the coldest or

warmest extremes of this gradient, less

cold- or drought-adapted plants would

CHAPTER 4 .

139

benefit from nurse canopies, respectively

(Choler et al. 2001, Liancourt et al. 2005).

This may explain why the net positive

effects of nurse plants were equally intense

or important at the community level along

the entire gradient (Tilman 1988). Indeed,

this is suggested by the fact that the

proportion of facilitation obligate species

was not significantly related to climate in

this region. The higher number of

facilitation beneficiary species under drier

conditions could be related to the fact that,

regardless of their physiological adaptions

or environmental optima, germination and

recruitment of most semiarid species are

limited by periods of adequate soil moisture

(Westoby 1978/79). These periods are more

easily achieved under the more shaded

onditions beneath nurse plants than in Open

microsites (Franco and Nobel 1989).

Therefore, under drier conditions we would

likely record more individuals of most of

species under the canopy of nurse plants

than in Open microsites (Kitzberger et al.

2000). The environmental stress in the

eastern Australian gradient was

predominantly driven by a single stressor,

rainfall, through its influence on soil

moisture availability. Consequently, nurse

plants could conceivably have an important

role in allowing recruitment and persistence

of taxa less adapted to low soil moisture

(Kitzberger et al. 2000, Soliveres et al.

2010). This facilitative role collapsed at

extremely high stress levels. The most

parsimonious explanation for this is that

facilitation collapsed because species with

low tolerances to drought and/or herbivory

were unable to overcome the environmental

filters controlling their recruitment,

regardless of the presence of nurse plants

(Kitzberger et al. 2000, Ibañez and Schupp

2001, Michalet et al. 2006, Soliveres et al.

in press), and therefore could not be

included in our pairwise analyses. Thus, the

positive effects on richness and productivity

at the community level collapsed under

these extremely stressful conditions (Forey

et al. 2009). This is consistent with

observations of higher percentage of

facilitation obligate species and lower

similarity at moderate levels of drought

stress, and with the breakdown of these

effects under extremely high levels of

drought (Fig. 4.4).

THE EFFECT OF PLANT-PLANT

INTERACTIONS ON DIVERSITY:

NICHE EXPANSION AND

SEGREGATION

The results from the Chao index of

similarity, and the lack of relationship

between this index and Climate (Fig. 4.6),

suggest that facilitation from nurse plants

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

140

promotes niche expansion for the less

adapted species to the given environmental

conditions present in a particular site across

the gradient, rather than increased niche

expansion of competitive species under

harsh conditions. Our results partially

contrast with previous studies, which have

suggested that stress-tolerant species

increase the realized niche of competitive

species on harsh (i.e. drier) environments

(Travis et al. 2005, Michalet et al. 2006).

We argue that this is not caused by the

stress-tolerator or competitor strategy of the

species involved, but rather the distance of

each target species to its ecophysiological

optimum. It should be realized that about

20-30% of the sampled species recruited

only under the canopies of nurse plants

(Fig. 4.4), regardless of the environmental

conditions present in each plot. The total

percentage of facilitation obligate species

(when considering both microsites together)

was greater than those for separate

microsites in both the Spanish and

Australian gradients surveyed. These results

suggest that the identity of facilitation

obligate species changed according to the

particular nurse plant examined. These

changes in the identity of facilitated species

depending on the nurse plants were,

perhaps, due to their different phylogenetic

relationships (Valiente-Banuet and Verdú

2007, 2008) or to differences in their

ecological strategy (Prider and Facelli 2004,

Maestre et al. 2009a).

In contrast with previous studies

(Tielbörger and Kadmon 2000b), our

results showed that nurse plants affected the

competition outcomes of their understorey

vegetation in comparison with Open areas

through increases in niche segregation, but

not by increasing competition intransitivity.

While this was apparent in Spain, it was not

in Australia. On the Spanish gradient,

competitive exclusion was significantly

lower under nurse canopies than in Open

microsites. This could be explained because

the more productive conditions found under

the nurse plants allowed more species to

recruit. Since nurse plants provide some

degree of microclimatic heterogeneity

(Pugnaire et al. 1996a), the joint effect of

both processes (increase in species pool and

variability in the resources that these

species compete for) might increase niche

segregation, and therefore local diversity

(Huston 1979, 1999, Silvertown 2004). We

found, however, that despite the relatively

constant effect of nurse plants upon the

competitive outcome of their understorey

vegetation, the effect of nurses on local

diversity decreased with productivity. It is

conceivable that the relative differences in

the microenvironmental conditions between

nurse and Open microsites that allowed

more species to recruit under nurses than in

the unvegetated interspaces declined under

CHAPTER 4 .

141

wetter conditions. In Australia, the lack of

nurse effects on competitive outcomes, but

their positive effect on local diversity, may

have resulted from environmental buffering

(both microclimatic amelioration and

grazing protection), allowing fewer

drought- and herbivory-tolerant species to

recruit, and therefore increased diversity by

direct facilitation. However, their relative

homogeneous microenvironmental

conditions did not allow niche segregation

to occur (Huston 1979). In contrast with

previous studies (Bowker et al. 2010), we

did not find an important contribution of

competition intransitivity to local species

richness in any of the studied regions, even

in the more productive and heterogeneous

environments beneath nurse plants. This

could be a matter of scale, since the

availability of slightly different niches (i.e.

differences in shade or nutrients), even at

the 0.5 m × 0.5 m scale, might prevent the

changes in competitive dominance of

different species depending on the different

conditions in each quadrat, causing the

niche segregation, rather than competitive

exclusion at this scale.

PAIRWISE INTERACTIONS AND THE

ABSENCE OF AN APPROPRIATE

EXPLANATORY MODEL

None of the models we tested i.e. SGH

(Bertness and Callaway 1994) or

Facilitation waning model (Michalet et al.

2006, Maestre et al. 2009a; see Fig. 4.1)

predicted more than 60% of the pairwise

interactions studied (see Tables 4.1 and

4.2). The percentage of cases explained,

however, varied strongly among microsites

and regions, as well as with the identity of

the particular nurse plant (Callaway 2007,

Table 4.2). These results highlight the

difficulties in establishing generalities when

predicting how the outcome of plant-plant

interactions change along environmental

gradients. Our results are not completely

unexpected, as plant-plant interactions are

driven by a complex set of factors including

the number and type of stressors considered

(Baumeister and Callaway 2006, Maestre et

al. 2009a, le Roux and McGeoch 2010), the

particular adaptations of species to cope

with current environmental conditions

(Choler et al. 2001, Liancourt et al. 2005),

the relative effect of nurse shade on target

plants (Holmgren et al. 1997, Prider and

Facelli 2004, Soliveres et al. 2010),

phylogenetic relationships (Valiente-Banuet

and Verdú 2007), and the ontogenetic stage

of interacting plants (Callaway and Walker

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

142

1997, Miriti 2006, Soliveres et al. 2010).

This complex array of mechanisms

involved in the outcome of particular

pairwise interactions makes it extremely

difficult to develop general predictions

about the response of a given pairwise

interaction to particular environmental

conditions without a detailed knowledge of

many of the attributes described above.

Extrapolation from a few pairwise

comparisons to the broader community of

species should only be made with extreme

caution. Future general models aimed at

predicting the outcomes of particular

pairwise interactions should therefore

consider the complexities and multiplicity

of mechanisms shaping such outcomes. We

suggest here that this complexity could be

better integrated by using a surrogate of the

distance of each target species from its

optimum, either from detailed knowledge

of its physiological tolerances (i.e Choler et

al. 2001, Liancourt et al. 2005) or by

phylogenetic derived assumptions

(Valiente-Banuet and Verdu 2007). The

outcomes of pairwise interactions could

therefore be better predicted by using a

distance to optimum approach rather than

an approach that considers all species in a

given community equally, by invoking the

stress gradient model.

CONCLUDING REMARKS

Our study has highlighted the fact that

nurse plants increase local richness of plant

communities across broad environmental

gradients. Nurses increase niche

segregation and species coexistence by

providing a range of available niches

beneath their canopies, thereby allowing

species that are less adapted to a particular

position within the gradient to survive and

recruit (niche expansion). Our results show

that the importance of nurse plants is

relatively constant along environmental

gradients when different independent

stressors (e.g. low temperature and rainfall,

herbivory) have differential effects on the

stress experienced by different species

within the community. However, the

importance of niche expansion increases

along environmental gradients when a

given stressor, or a combination of

positively correlated stressors (e.g. high

temperatures and salinity or drought),

affects the stress level of different species

within the community (e.g. Callaway et al.

2002). In the latter case, this positive effect

is likely to collapse under extremely high

levels of stress through several mechanisms

previously discussed in the literature

(Michalet et al. 2006, Smit et al. 2007,

Maestre et al. 2009a).

Given the clumped nature of

vegetation in arid and semiarid areas, and

thus the tendency for plants to interact, it is

CHAPTER 4 .

143

not surprising to find that plant-plant

interactions play a major role in shaping the

relationships between richness and

productivity in these environments.

However, our results indicate that the effect

of these interactions on increasing or

decreasing richness along environmental

gradients is more complex than previously

thought, and depends on the number of

stressors involved and their

interrelationships. Our findings also help to

explain why the unimodal relationship

between diversity and productivity is rarely

found in arid and semiarid environments

(Waide et al. 1999). We maintain that the

outcome of particular pairwise interactions

is best predicted by the distance of a given

plant to its environmental optimum rather

than by an overarching stress gradient. This

has profound implications for interpreting

previous studies evaluating the interplay

between facilitation and competition along

stress gradients, and should be considered

by future research on this important theme

of community ecology. We highlight the

fact that the complex array of mechanisms

shaping the outcomes of pairwise plant-

plant interactions makes it difficult to

develop a universal model that is able to

successfully predict their outcome along

environmental/productivity gradients. We

propose that approaches considering

multiple models, such as that followed in

this study, may provide important insights

into the mechanisms driving such

outcomes, and on the community and

ecosystem-level consequences of plant-

plant interactions. By using a multiplicity of

conceptual and analytical approaches, and

an appropriate dataset collected in two

contrasted semiarid regions, our study

provides a more complete mechanistic

understanding of the relative role of biotic,

non-trophic interactions and environmental

conditions shaping local richness. It also

helps to refine our predictions of the

response of plant communities to

environmental gradients, and clarifies the

relative importance of biotic interactions as

a driver of such responses.

ACKNOWLEDGMENTS We thank David Tongway and Nick Reid for their help during plot selection and fieldwork in Australia. Nick Reid also hosted SS during a research stay in his lab. Nick Schultz, Megan Good, María D. Puche, Pablo García-Palacios, Erin Roger, Ian Telford, James Val and Madeleine Rankin assisted with fieldwork and/or plant identification. Peter Weston, Anthony Gibson, Kevin Mitchell, Andrew Mosely and Patty Byrne allowed us access to their properties and gave us valuable information on land management issues in semiarid Australian woodlands. SS was supported by a PhD fellowship from the EXPERTAL project, funded by Fundación Biodiversidad and CINTRA S.A. This research was funded by the CEFEMED, INTERCAMBIO (BIOCON 06/105) and REMEDINAL projects, funded by the Universidad Rey Juan Carlos-Comunidad de Madrid, Fundación BBVA and Comunidad de Madrid, respectively. FTM acknowledges support from the European Research Council under the European Community's Seventh Framework Programme (FP7/2007-2013)/ERC

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

144

Grant agreement n° 242658. DJE is supported by grant LP0882630 from the Australian Research Co

CHAPTER 4 .

145

Supplementary material for Chapter 4.

Appendix H. Characteristics of vegetation types and sites selected

Stipa tenacissima grasslands, the vegetation type selected in Spain, constitute one of the more

important European/North African ecosystem types, as they can be found from the semi-

desert steppes of Ukraine to the western Mediterranean (Izco 1984), with the greatest

extensions being found in the Maghreb and the Iberian Peninsula (Le Houeróu 1986, 2001).

In the semiarid parts of the Mediterranean Basin, these grasslands are distributed over 32,000

km2 in a thin latitudinal fringe in North Africa, from Libya to Morocco, and in the

southeastern Iberian Peninsula (Le Houérou 2001). See Maestre et al. (2009b) for a detailed

account of the natural history of this ecosystem. Soils in the Spanish plots are Lithic

calciorthid (Soil Survey Staff 1994) and characterized by a high CaCO3 content, high pH

values, low depth, and a stony surface.

Sites in the semiarid woodlands in eastern Australia were located in the Bimble box

(Eucalyptus populnea)–White cypress pine (Callitris glaucophylla) alliance on the Cobar

Pediplain, the Belah (Casuarina pauper)–Rosewood (Alectryon oleifolius) woodlands in far

western New South Wales (NSW), and woodlands dominated by White box (Eucalyptus

albens) in central western NSW. While these communities have slightly different community

dominants, physiognomically they are similar and characterized by an open woodland on clay

loam soils with canopy cover ranging from 18-70% (Keith 1998). The midstorey shrub cover

at all sites varied depending on grazing intensity and rainfall (Beadle 1948). Soils in the

Australian plots were non-sodic Kandosols to Demosols under the Australian classification

(Isbell 1996), or a mixture of Luvisols, Yermosols and Ferrasols (FAO 1998) and are

commonly grouped as Red earths (Stace et al 1968). They are characterized by deep clay

loam to loamy surface textures with a gradual increase in clay content with depth. They have

relatively low available nutrient and water holding capability (Isbell 1996).

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

146

Table S7. Main charactertistics of the study sites. TSR = travelling stock reserve.

Country Site

Latitude

Longitude

Land use

Annual

rainfall

(mm)

Mean

projected

Cover

(%)

Australia Cowra 33º50’41”S 148º36’12”E TSR 630 70

Australia Quandialla 33º53’19”S 147º51’27”E TSR 570 51

Australia Nevertire 31º51'27"S 147º42´28”E TSR 490 70

Australia Condobolin 33º07’15”S 142º24’44”E State Forest 455 18

Australia Nyngan1 31º10’00”S 142º24’44”E Grazing 406 31

Australia Nyngan2 31º10’00”S 142º24’44”E Grazing 406 43

Australia Florida 31º33’22”S 146º18’35”E Grazing 398 44

Australia Truganini 32º07’00”S 146º39’50”E property 375 44

Australia Etiwanda 32º09’40”S 145º53’40”E Grazing 360 21

Australia Buronga 34º07’35’’S 141º05’09”E Grazing 280 35

Spain Barrax 39º02’91’’N 2º13’82’’W Hunting area 433 46

Spain

Camporreal 40º19’72’’N 3º25’36’’W Hunting area 457 54

Spain

Carrascoy 37º48’02’’N 1º18’32’’W Hunting area 282 37

Spain

Crevillente 38º14’15’’N 0º55’49’’W Hunting area 273 35

Spain

El Ventós 38º28’14’’N 0º37’03’’W Hunting area 319 36

Spain

Morata 40º27’62’’N 3º05’31’’W Hunting area 455 58

Spain

Sierra Espuña 37º49’27’’N 1º40’41’’W Hunting area 364 49

Spain

Titulcia 40º11’28’’N 3º30’13’’W Hunting area 440 57

Spain

Villarrobledo 39º12’64’’N 2º30’77’’W Hunting area 446 63

Spain

Yecla 38º35’40’’N 1º12’15’’W Hunting area 350 36

Spain

Zorita 40º21’30’’N 2º52’62’’W Hunting area 434 45

CHAPTER 4 .

147

SPAIN AUSTRALIA

Figure S2. General appearance of the vegetation types studied in Spain and Australia.

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

148

Appendix I. Measuring community composition changes across environmental

gradients

To test for differences in community composition across the environmental gradient, we

constructed for each country a data matrix comprising the relative frequency of each species

in the 80 1.5 m by 1.5 m quadrats within each plot. Thus, a species occurring in all 80

quadrats at a given site received a relative frequency of 100%. We examined differences in

community structure between the sites with Multi-Dimensional Scaling (MDS) after applying

a fourth-root transformation to the data to downweight the influence of the most dominant

species (Clarke and Warwick 2001). Extremely rare species (those appearing in less than the

1% of the sampled quadrats), were removed prior to analysis as recommended to improve the

clarity in the analyses (Anderson et al. 2008). These rare species accounted for about half of

the sampled species (46 of 96 for Spain, 57 of 131 for Australia). We used the Bray-Curtis

distance measure to construct the similarity matrix. This distance measure controls for the

relative abundance of each species in a given matrix, and helps to avoid the undue influence

of extremely abundant or extremely rare species (Clarke and Warwick 2001). We carried out

these analyses using PRIMER v6 statistical package for Windows (PRIMER-E Ltd.,

Plymouth Marine Laboratory, UK). We analysed the data in both two and three dimensions,

and then chose the 2-D ordination, because it provided satisfactory results (i.e. low stress

values) for both studied regions (Stress = 0.10 and 0.12 for Spain and Australia, respectively).

The stress value reflects how well the data can be represented in any given number of

dimensions. Stress values < 0.05 indicate that the n axes provide an excellent representation

of the relationships among samples, while values of > 0.20 are regarded as a poor

representation. We performed the analyses with 25 random starts to reduce the risk of finding

a local instead of the global minimum of this stress function.

To identify the species or abiotic factors responsible for the ordination patterns found,

the first two axes of the MDS biplot were correlated with the PCA axis obtained with climatic

data (Climate) and with the abundance of each species in each plot using Spearman

correlation coefficients. For the latter analyses, only the species present in at least three of the

80 plots were considered. Variables with a correlation coefficient > 0.5 are represented in the

MDS plot (Fig. B1).

The ordination (see Figure below) showed a significant effect of rainfall on

community composition in Spain, where different species were related both positively or

CHAPTER 4 .

149

negatively with this factor, depending on their ecophysiological tolerances. However, this

effect was less clear in Australia, suggesting the existence of unmeasured factors affecting

community composition. The relatively high density of domestic and native herbivores in the

Australian semiarid zone (Noble and Tongway 1986) suggests that different levels of grazing

pressure among plots could be one such factor.

MDS 1

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

MD

S 2

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

Barrax

Camporreal

El Ventós

Crevillente

Morata

Zorita

Villarrobledo

Sierra Espuña

Titulcia

Carrascoy

Yecla

Rainfall: 0.79 Hh: 0.70Gs: 0.67

Ac: 0.69 Ar: 0.64 Br: 0.61 Ss: 0.67Ft: 0.77 Hc: 0.87 Hv: 0.62

Ch:

0.8

5 T

z: 0

.70

A

MDS 1

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0

MD

S 2

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

CowraQuandialla

Gibson1

Gibson2

Condobolin

Nevertire

Buronga

Florida

TruganiniEtiwanda

Ed: 0.64 Des: 0.75 Cs: 0.65Pd: 0.75 Scl: 0.66

Rai

nfal

l: 0.

59

M

e: 0

.81

Cal

: 0.8

0

Al:

0.75

B

Fig. S3 . MDS plot showing the dissimilarity distance among communities of the plots sampled along environmental gradients in Spain (A) and Australia (B). Spearman correlation coefficients >0.5 are showed in a box for each axis. Legends are: Rainfall = Climatic PCA, highly correlated with rainfall in both countries. Spain: Tz = Thymus zygis, Gs = Genista scorpius; Ch = Carex humilis; Ac = Anthyllis citisoides; Ar = Asphodelus ramosus; Br = Brachypodium retusum; Ss = Sedum sediforme; Ft = Fumana thymifolia; Hh, Hc and Hv = Helianthemu hirtum, H. cinereum and H. violaceum, respectively. Australia: Al = Atriplex leptocarpa; Me = Maireana enchylaenoides; Cal = Calotis sp.; Ed = Eragrostys dielsii; Des = Desmodium sp.; Cs = Cheilantes sieberi; Scl = Sclerolaena sp.; Pd = Panicum decompositum.

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

150

Appendix J. An alternative approach to measure changes in competitive outcomes

depending on microsite

As stated in the main text, three principal mechanisms may determine competitive outcomes

in the vegetation in Open microsites and under nurse plants: competitive exclusion,

intransitivity and niche segregation. While competitive exclusion is likely to cause low

diversity at the plot-scale, both intransitivity and niche segregation are likely to increase it.

We aimed to to test for differences in the competitive outcomes among the three microsites

tested in each country. As an alternative approach to that presented in the main text, we

developed an index based on Whittaker´s beta-diversity index (Whittaker 1972): A/(B-1),

where A is the global species richness, and B the local species richness. In our case, A will be

the species richness found under each microsite at the whole plot level, while B will be the

species richness found in each one of the quadrats sampled for each microsite (n = 30). To

calculate B, those quadrats with no species occurring on them where not considered.

The rationale behind this approach is similar to that described in the main text: 1) if

competitive exclusion dominates, we will find a reduction of B causing a decrease in A (few

dominant species will control all the space available; low B, low A = competitive exclusion);

2) if intransitivity dominates, we will find a low B but a high A because of few dominant

species will dominate at the small-scale (quadrat): however, the identity of these dominant

species will change depending on the particular environmental conditions of each patch and

the original species mixture that colonized it. This will generate a high turnover of species

that will increase the plot-scale richness (low B, high A = competition intransitivity); 3) if

niche segregation dominates, both A and B will be high. At small spatial scales, the

differential exploitation of resources (niche segregation) will enhance species coexistence,

and therefore increase the chances of several species to recruit and coexist under a given

microsite (high B, high A = niche segregation). We analyzed separately both A and B with

univariate ANCOVA models, with microsite (three levels) as fixed factor and mean plot cover

(a surrogate of productivity) as a covariate (Table S7).

CHAPTER 4 .

151

Table S8. ANCOVA results for both small-scale (B) and plot-scale (B) richness for studied region.

Variable Factor df F P-value Effect Underlying

mechanism

Sp

ain

A Cover 1,29 53.9 <0.0001 Reduction of

microsite effect Higher niche segregation under nurse plants, this difference

reduces with higher

productivity

Microsite 2,29 3.0 0.65 Nurses increase A

B Cover 1,29 8.1 0.008 Reduction of

microsite effect

Microsite 2,29 4.7 0.017 Nurses increase B

Au

stra

lia

A Cover 1,26 0.2 0.628 None Higher niche

segregation under nurse

plants, regardless of ecosystem productivity

Microsite 2,26 4.0 0.030 Nurses increase A

B Cover 1,26 2.7 0.113 None

Microsite 2,26 4.0 0.030 Nurses increase B

BIOTIC INTERACTIONS DRIVE LOCAL-SCALE RICHNESS

152

Santiago Soliveres, Rubén Torices, Fernando T. Maestre.

Manuscrito en revisión en Journal of Ecology

5

On the relative importance of climate and biotic no n-

trophic interactions as drivers of local plant spec ies

richness in semiarid communities

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

154

CHAPTER 5

155

ABSTRACT

Molecular phylogenies are being increasingly used to get new insights on the mechanisms structuring plant communities. However, the large number of factors potentially affecting the phylogenetic structure of plant communities cautions against the sole use of this information to properly infer the mechanisms shaping them. We jointly evaluated the effects of environmental conditions and biotic interactions on the phylogenetic structure of 11 semiarid Stipa tenacissima L. communities along an ample environmental gradient. We also assessed the relative importance of phylogenetic relatedness (PD) and abiotic conditions as drivers of pairwise interactions across such gradient. Habitat filtering and biotic interactions promoted a random phylogenetic structure in most of the communities studied. While positive biotic interactions increased phylogenetic evenness by niche expansion and habitat differentiation, more benign environmental conditions reduced this evenness indirectly by reducing the effects promoted by nurse plants. Phylogenetic relatedness was the primary factor affecting pairwise interactions. Values of this variable between 207-272.8 Myr led to competition, those outside this range led to neutral or positive interactions, depending on climate. Our study illustrates, for the first time, the relative importance of climate and biotic interactions on the phylogenetic structure of plant communities, and shows how the evolutionary relationships and environmental conditions interact to determine particular pairwise interactions. We also provide a comprehensive set of easy-to-measure and interpret tools for avoiding misleading interpretations when inferring mechanisms from phylogenetic structure data in observational studies.

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

156

INTRODUCTION

he recent development of

molecular phylogenies has

provided ecologists with a

powerful tool to get new insights on the

mechanisms structuring plant communities

(Webb et al. 2002). The phylogenetic

structure of a given community has been

extensively used to assess the relative

importance of environmental conditions

(environmental filtering) or competition as

drivers of community structure (reviewed

in Webb et al. 2002, Cavender-Bares et al.

2009, Vamosi et al. 2009). Biotic

interactions and environmental conditions

are known to interactively affect plant

community structure and dynamics

(Butterfied et al. 2010). However, the

relative importance of both factors as

drivers of the phylogenetic structure of

plant communities is still poorly

understood (Cavender-Bares et al. 2004,

Verdú et al. 2009). Furthermore, the

phylogenetic structure of a given

community may be strongly affected by

other factors, including herbivore or

pollinator preference for closely related

taxa (Webb et al. 2006), the scale

considered (Kraft and Ackerly 2010), or

differences in niche and competitive ability

among co-occurring species (Myfield and

Levine 2010). This complex array of

factors makes the use of phylogenetic

structure alone insufficient to properly

infer the mechanisms shaping plant

communities (Cavender-Bares et al. 2009,

Myfield and Levine 2010). Thus, more

comprehensive approaches, including the

study of environmental factors and co-

occurrence patterns, have been

recommended to further refine the

conclusions drawn from phylogenetic

methods (Pausas and Verdú 2010).

Positive interactions among plants

have been shown to largely influence the

structure and diversity of plant

communities in virtually all terrestrial

ecosystems (Callaway 2007, Brooker et al.

2008), and can even promote the expansion

of realized species niches over

evolutionary time frames (Valiente-Banuet

et al. 2006). These interactions have been

shown to depend up to a great degree on

the abiotic environment (Callaway 2007;

Maestre et al. 2009), and on the

phylogenetic distance (hereafter PD)

between the interacting species (Valiente-

Banuet et al. 2006, Castillo et al. 2010).

Overall, these studies suggest that positive

interactions are more likely to occur

among phylogenetically distant species

pairs, or under harsh environmental

conditions (e.g. Callaway 2007, Valiente-

Banuet and Verdú 2007). In the same way

that the different ontogenetic stages of

involved species interact with their PD in

defining the outcome of pairwise plant-

T

CHAPTER 5

157

plant interactions (Valiente-Banuet and

Verdú 2008, Castillo et al. 2010), both

environmental conditions and PD are likely

to jointly determine the outcome of these

interactions, and hence the structure of

plant communities. Although it is expected

that harsh environmental conditions cause

phylogenetic clustering (Webb et al. 2002),

the prevalence of pairwise positive

interactions at the community level might

lead to an even phylogenetic structure

(Valiente-Banuet and Verdú 2007).

Therefore, knowing how PD and

environmental conditions jointly affect the

outcome of pairwise species interactions

will provide additional insights to

understand the relative importances of

environmental filtering and biotic

interactions as drivers of the phylogenetic

structure of entire plant communities

(Cavender-Bares et al. 2009).

To our knowledge, no previous

study has evaluated the relative

importances and joint effects of biotic

interactions and environmental conditions

as determinants of the phylogenetic

structure of whole plant communities along

wide environmental gradients, nor the joint

effects and the relative roles of PD and

environmental conditions as drivers of

pairwise plant-plant interactions. We

aimed to do so by simultaneously

measuring the phylogenetic structure,

different components of biotic interactions

(co-occurrence patterns, niche expansion

promoted by nurse plants, and differences

in the environmental filtering among

facilitated/non-facilitated species guilds),

and a set of environmental conditions in 11

semiarid Stipa tenacissima communities

located along an environmental gradient in

Spain. Additionally, we assessed how the

abiotic conditions, the PD between the

involved species, and the interaction

between both factors modulated a large set

of pairwise interactions outcomes along

such gradient. We addressed the following

questions: How do environmental filtering

and biotic interactions jointly affect the

community phylogenetic structure across

environmental gradients?, and What is the

relative importance of PD and abiotic

conditions in defining the outcome of

particular pairwise interactions?

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

158

Figure 5.1 Conceptual model synthesizing the unifying approach employed in this study at both community and pairwise levels. Arrows are the different processes considered affecting community assemblage and its phylogenetic structure. Circular boxes represent the surrogates of each process measured in this study. The phylogenetic structure of a given community will be affected by several factors such as abiotic conditions (i.e. environmental filtering) or biotic non-trophic interactions (i.e. facilitation/competition shifts, niche expansion or double environmental filtering [phylogenetic clustering among facilitated/non-facilitated species guilds, or differences in similarity among microsites]). However, these processes do not affect independently the community assemblage and they are likely to interact in many ways. We tested for shifts in the biotic interactions across a wide environmental gradient and how they affect the phylogenetic structure of the studied community. At the pairwise level, interaction outcomes should be more positive when these conditions represent higher stress (i.e. low rainfall, cold temperatures; Callaway 2007). A different line of inquiry highlights that the phylogenetic distance (PD) between involved species as a crucial factor affecting these interactions; the outcome will be more positive when higher the PD among the species involved

CHAPTER 5

159

(Valiente-Banuet et al. 2006, Valiente-Banuet and Verdú 2007, Castillo et al. 2010). We hypothesize that the PD distance between a nurse plant and a given beneficiary would need to increase for their interaction to be positive as abiotic conditions become less stressful.

MATERIALS AND METHODS

STUDY AREA

We studied 11 Stipa tenacissima

communities along a climatic gradient

spanning from the center to the south-east

of Spain (see Table S9). Our sites have

annual precipitation and temperature

values ranging from 273 mm to 488 mm,

and from 13ºC to 17ºC, respectively. To

minimize the experimental noise produced

by environmental factors other than

climate, which could affect our

conclusions, all the sites shared the same

general soil type (Lithic Calciorthid, Soil

Survey Staff 1994), and had similar

orientation and slope values. Vegetation

was in all cases an open grassland

dominated by S. tenacissima, with total

cover values ranging from 35% to 68%.

Sparse resprouting shrubs like Quercus

coccifera, Pistacia lentiscus or Rhamnus

lycioides were also present in all sites.

VEGETATION SURVEY

At each site we established a 30 m × 30 m

plot containing the representative

vegetation of the surrounding area. This

plot size allowed the inclusion of a number

of shrub patches large enough to conduct

the survey described below. In each site,

we located four 30 m long transect

downslope for the vegetation survey, each

8 m apart across the slope. Along each

transect, we placed 20 contiguous 1.5 m ×

1.5 m quadrats, and recorded the

presence/absence of each perennial plant

species within each quadrat.

To evaluate particular pairwise

interactions for both dominant nurse types

in the study region (sprouting shrubs and S.

tenacissima tussocks, hereafter Shrub and

Stipa microsites, respectively), and to

assess the total number of facilitated

species at each site, we established a

complementary sampling design. We

randomly selected ten Stipa tussocks in

each site, and sampled the total area under

their canopy using 0.5 m × 0.5 m quadrats

(~30 quadrats per site). Ten paired open

areas (areas located at least 1 m away from

any Stipa tussock or resprouting shrub,

hereafter Open microsite), were randomly

selected adjacent to these tussocks. The

same number of 0.5 m × 0.5 m quadrats

sampled in each Stipa microsite was

sampled in each Open microsite selected,

to balance the sampling effort. Finally, the

same area was also sampled under the

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

160

canopies of five sprouting shrubs (mostly

Q. coccifera, Appendix K). The abundance

(number of individuals) of all perennial

plant species was recorded within each

sampled quadrat.

ASSESSMENT OF PHYLOGENETIC

STRUCTURE

We assembled a phylogenetic tree for the

86 species included in this study using

Phylomatic2 (Webb et al. 2008). All the

families in our dataset matched the family

names of the angiosperm megatree used in

Phylomatic (R20091110.new), which was

based on the APG III phylogenetic

classification of flowering plant orders and

families (Angiosperm Phylogeny Group

2009). Within-family phylogenetic

relationships were further resolved based

on data from various published molecular

phylogenies (Appendix K, Table S10).

After assembling the phylogenetic tree, we

adjusted its branch lengths with the help of

the Phylocom BLADJ algorithm, which

fixes the age of internal nodes based on

clade age estimates, whereas undated

internal nodes in the phylogeny are spaced

evenly (Webb et al. 2008). According to

Vamosi and Vamosi (2010), we used

TimeTree (Hedges et al. 2006) to fix as

many nodes in the tree as possible (see

Appendix K for methodological details).

This procedure resulted in the fixation of

48 nodes (representing more than 70% of

internal nodes of our tree).

Once we assembled the

phylogenetic tree for all the species

surveyed (Fig. S8), we measured two

different indicators of phylogenetic

relationships among co-occurring plants at

each of our sites: the mean phylogenetic

distance (hereafter MPD; Webb et al.

2002), and the pairwise PD among every

possible pair of co-occurring species in

each site. Since MPD is related to the

species pool, we avoided this confounding

factor by calculating its standardized effect

size (SES) with the Picante package for R

(Kembel et al. 2010), version 2.10.1 (R

Development Core Team 2009). It was

calculated as (MPDobs - MPDsim) ⁄

sdMPDsim, where MPDobs was the

observed value of MPD, and MPDsim and

sdMPDsim were the mean and standard

deviation, respectively, of this index

obtained from the 1000 simulations

performed under the null model. Positive

SES, and with a P-value > 0.95 indicate

significant phylogenetic evenness in the

sampled community, while negative SES

with P-values < 0.05 indicate phylogenetic

clustering.

CHAPTER 5

161

Figure 5.2 Graphical synthesis of the results obtained. Different surrogates of biotic interactions (C-E), environmental filtering (A) and their interactions (B) are separated. Blue arrows indicate an increase in community phylogenetic evenness due to a given mechanism, while red arrows mean an increase in phylogenetic clustering (dashed lines indicate an indirect effect). Crossed-red circles indicate no effect (in the case of arrows) or no interaction with climate (in the case of variables). MPD = mean phylogenetic distance among all possible species pairs of a given site; MPDfac = mean phylogenetic distance among pairs of the cluster of facilitated species; C-score = standardised effect size of the C-score; Similarity = Chao-Jaccard abundance-based similarity index calculated for Stipa vs. Open or Shrub vs. Open microsites; % Obligates = percentage of species found only under a nurse plant comparing to the total species richness found in each site; Climate = values of the first axis of a PCA performed with eight environmental variables. Detailed statistical results from each relationship shown in this figure are in Fig. S10.

EVALUATING PLANT-PLANT

INTERACTIONS AT THE

COMMUNITY LEVEL

The 80 1.5 m �1.5 m quadrats surveyed at

each site were used to examine the co-

occurrence pattern at the whole community

level by using the SES of the C-score

index (Gotelli et al. 2000). This metric is

commonly used to assess the outcome of

biotic interactions at the community level

(e.g. Rooney et al. 2008, Bowker et al.

2010), and was calculated as described

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

162

above for the MPD (see Appendix K for

methodological details).

To measure niche segregation or

the “double habitat filtering” (see below)

promoted by nurse plants we used two

different measurements: 1) the Chao-

Jaccard abundance-based similarity index

(hereafter CI; Chao et al. 2005), and 2) the

difference between MPD for the whole set

of species in a given community (described

above), and the MPD for only the

facilitated/non-facilitated species guilds

(MPDfac and MPDnf, respectively). We

refer here to “double environmental

filtering” to describe the effect of the

different environmental conditions found

in Open or nurse microsites, which

generate different understorey species

guilds for each microsite (i.e., higher

dissimilarity among microsites; Badano

and Cavieres 2006). These guilds could be

phylogenetically clumped within each

microsite because of their shared

adaptations to the same environmental

conditions, but can be phylogenetically

even at the whole community level

because of the different environments

existing at each site, i.e., MPD > MPDfac

or MPDnf. The CI is based on the

probability that two randomly chosen

individuals, one from each of two selected

microsites (Open, Stipa or Shrub), belong

to species shared by both microsites. A

higher dissimilarity (lower CI) among

Stipa/Shrub and Open microsites will

indicate higher influence of niche

differentiation provided by nurse canopies.

We calculated the CI by summing over the

number of individuals and species recorded

in all the 0.5 m × 0.5 m quadrats per

microsite and site (n ~ 30) using EstimateS

8.2.0 for Windows (Colwell 2000). We

also compared the MPDfac and MPDnf

with the MPD index (see rationale in

Assessing the effects of environmental

filtering and biotic interactions on

community phylogenetic structure below).

Finally, to evaluate the degree of

the realized niche expansion (sensu Bruno

et al. 2003) provided by nurse plants, we

calculated the percentage of facilitation

obligates (sensu Butterfield 2009), in

comparison to the total number of species

found in each site. Facilitation obligates

were those species with individuals

recruiting only under a nurse plant, and

therefore, only able to colonize a given site

under the microclimatic protection

provided by nurses.

EVALUATING PLANT-PLANT

INTERACTIONS AT THE PAIRWISE

LEVEL

We measured facilitation intensity and

importance, i.e., the effect that neighbours

have on their target species regardless of

other environmental factors and the

relative effect of nurses on their target

CHAPTER 5

163

species compared to that of other

environmental factors, respectively

(Brooker et al. 2005), of each possible

Stipa- and Shrub-target species pairs at

each site. For doing this we used the

Relative Interaction Intensity index (RII)

and the Interaction Importance index

(Iimp) provided by Armas et al. (2004) and

Seifan et al. (2010), respectively (see

Appendix K for details). Both indices

oscillate between -1 and 1 (relative units);

positive and negative values indicate

facilitation and competition, respectively.

The higher the index value, the higher the

intensity (RII) or importance (Iimp) of

such effect. For these analyses we used the

total number of individuals found in the

~30 0.5 m × 0.5 m quadrats sampled at

each site, and calculated a unique RII and

Iimp index for each species and site. We

used the number of recruited individuals as

an indicator or each species performance to

calculate these indices, because the number

of recruited individuals indicates superior

environmental conditions for a given

species in a given microsite, an approach

followed by previous studies (Valiente-

Banuet et al. 2006, Valiente-Banuet and

Verdú 2007).

Alternatively, the pairwise

phylogenetic distances among both nurses

(Stipa and Shrub) and their co-occurring

species in each site were calculated using

the “cophenetic” command of R. The mean

PD of facilitation beneficiaries (species

with more individuals recruiting under a

nurse plant) and obligates (described

above) at both Stipa and Shrub microsites

was calculated in each site using these data

(see Fig. 5.1).

ASSESSING THE EFFECTS OF

ENVIRONMENTAL FILTERING AND

BIOTIC INTERACTIONS ON

COMMUNITY PHYLOGENETIC

STRUCTURE

Prevalent climatic conditions (rainfall,

radiation and a temperatures) for each site

were collected using available climatic

models (Ninyerola et al. 2005) and reduced

to a single synthetic variable using PCA

conducted with the Primer v. 6 statistical

package for Windows (PRIMER-E Ltd.,

Plymouth Marine Laboratory, UK).We

used the first axis of this PCA (hereafter

Climate) for the analyses explained below.

This axis explained 88.6% of the variance

and was inversely and positively correlated

with radiation and rainfall, respectively

(Appendix K).

The relationships between the

surrogates of biotic interactions used (C-

score, percentage of facilitation obligates,

and CI) and both MPD and Climate were

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

164

evaluated by using linear regressions. This

set of analyses gives us an idea of the

relative importance of environmental

conditions as drivers of biotic interactions,

and of the effects of these interactions on

the phylogenetic structure of the

community. Additionally, the relationship

between Climate and MPD was assessed

by using both linear regression (direct

effect) and partial correlations (indirect

effect, mediated by the effect of Climate

on C-score, percentage of facilitation

obligates, CI and MPD-MPDfac/nf,

respectively). If C-score, percentage of

facilitation obligates or CI increase MPD,

this will mean that either competition,

niche expansion or niche differentiation

provided by nurse plants, respectively,

increase the phylogenetic evenness of the

community. The relation between these

biotic interactions indicators and Climate

will give us an idea on how climatic

conditions influence such interactions.

Finally, if both linear regression and partial

correlations between Climate and MPD are

significant, this will mean that climatic

conditions have a direct effect on the

phylogenetic structure of the studied

communities. However, if only linear

regressions, but not partial correlations, are

significant, this will mean that climatic

conditions are affecting MPD only via

their indirect effects on a particular biotic

interaction mechanism (depending on the

partial correlation that is non-significant).

Finally, as a second evaluation of

the degree of habitat differentiation, and

how abiotic conditions affected it, we

evaluated the effect of Climate on the

difference between MPD and both MPDfac

and MPDnf, respectively. If this difference

is significantly higher than 0, niche

differentiation is producing a phylogenetic

clustering in facilitated or non-facilitated

species guilds, and thus the difference in

microclimatic conditions between nurse

and Open microsites affect the

phylogenetic structure of this particular

community. We first compared the MPD –

MPDfac/nf difference from zero by using

t-tests, with sites acting as replicates. Then,

the relationship between Climate and this

difference was evaluated using linear

regressions to asses the effect of climatic

conditions on this “double habitat filtering”

(Jones et al. 1997, Badano and Cavieres

2006; Fig. 5.1)

ASSESSING THE EFFECTS OF

CLIMATIC CONDITIONS AND PD ON

THE OUTCOME OF PAIRWISE

PLANT-PLANT INTERACTIONS

To evaluate the existence of an interaction

between PD and abiotic conditions, we

assessed the relationships between Climate

and the MPD obtained from all the nurse-

facilitated species pairs in each site by

CHAPTER 5

165

using linear regressions. These species

pairs were divided according to the type of

facilitative relationship (beneficiaries or

obligates) and nurse plant (Stipa or

shrubs). We also fitted the relationships

between Climate, PD, RII and Iimp by

using Regression trees (De´ath and

Fabricius 2000), as implemented in the

Tree package of R. We used 10-fold cross-

validation to fit the most parsimonious

model to each dataset (De´ath and

Fabricius 2000). Climate and the PD

between each target species and its nurse

were used as predictor variables in the four

regression trees performed (RII and Iimp

indices for Stipa and shrub nurses). Despite

this low number of predictors, we used this

procedure because of the heavily skewed

nature of the PD values (Fig. S5), a

characteristic commonly found with these

data (Castillo et al. 2010). Regression trees

also allow detecting non-linear

relationships, and are insensible to the

distribution of either the predictor or

response variables (De´ath and Fabricius

2000).

RESULTS

EFFECTS OF ENVIRONMENTAL

FILTERING AND BIOTIC

INTERACTIONS ON COMMUNITY

PHYLOGENETIC STRUCTURE

Most of the studied sites showed a random

phylogenetic structure (Table S9). Climate

was negatively related to MPD at the

community level, but this relationship was

only marginally significant (r = -0.53; P =

0.09). When we removed the indirect

effect mediated by the relationship

between Climate and the MPD – MPDfac

difference by using partial correlations,

this relationship disappeared (ρ = 0.32; P =

0.4), suggesting the lack of a direct effect

of Climate on MPD. In contrast, the

different measures of plant-plant

interactions were all positively, but

marginally, related to MPD at the

community level (r = 0.60, 0.56 and 0.57;

P = 0.07, 0.09 and 0.07 for C-score SES,

CI between Open and Shrub microsites and

percentage of facilitation obligates,

respectively; Fig. 5.2, Fig. S6). The

relationship between CI (Open vs. Stipa

microsites) and MPD was not significant (r

= - 0.11; P = 0.44). Most of the surrogates

of biotic interactions employed were not

related to Climate (P > 0.35 in all cases;

Fig. 5.2). However, the negative

relationship between the MPD-MPDfac

difference and Climate found (r = -0.82; P

= 0.006) suggests a decrease in

phylogenetic clustering among facilitated

species with increased rainfall availability.

This difference was also significantly

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

166

-200 -150 -100 -50 0 50 100

200

300

400

500

600

700

PCA climate

-200 -150 -100 -50 0 50 100

Mea

n P

D fr

om fa

cilit

atio

n be

nefic

iarie

s

200

250

300

350

400

StipaShrub

Total mean PD for Stipa

Total mean PD for Shrub

Mea

n P

D fr

om fa

cilit

atio

n ob

ligat

es

Total mean PD for Stipa

Total mean PD for Shrub

A

B

- RAINFALL +

Figure 5.3 Relationship between the PCA axis obtained from climatic values (PCA climate) and mean phylogenetic distance (PD) between facilitation beneficiaries (A) or obligates (B) species and their nurses. No relationships were found for any of the assayed variables and climate (R2 < 0.1 and P > 0.3 in all the cases).

CHAPTER 5

167

Figure 5.4 Regression trees conducted with both interaction intensity (RII; panel A) and importance (Iimp; panel B) indices. Split values for each predictor used (PD, in million years [Myr], or climatic PCA axis values) are shown in each branch. Terminal nodes show the mean value for each group of the response variable introduced and the number of cases in each node (between parenthesis; n = 200 cases for each tree). Positive and negative values indicate facilitative and competitive interactions, respectively. The higher the values, the more intense or important the effect of the nurse upon the target species. In each panel, the general fit of the model (D2, percentage of variance explained by the model), extracted from the null deviance (Deviance root), and the deviance of the final chosen tree after 10-fold cross-validation (Deviance tree) are shown.

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

168

different from zero, but the MPD-MPDnf

difference was not (mean difference = 1.09

± 0.15 and 0.69 ± 0.20; t = 2.36 and 1.15;

P = 0.046 and 0.284 for facilitated and

non-facilitated species, respectively).

These results suggest a phylogenetic

clustering, comparing with the general

species pool of each site, for the facilitated

but not for the non-facilitated species.

EFFECTS OF CLIMATIC CONDITIONS

AND PHYLOGENETIC DISTANCE ON

THE OUTCOME OF PAIRWISE

INTERACTIONS

The PD between nurse-facilitated species

remained constant across the entire

environmental gradient sampled (Fig. 5.3).

Climate and PD were poor predictors for

the RII and Iimp data calculated with the

target species tested and Stipa as nurse

plant, respectively (Regression trees D2 =

0.02 and 0 for RII and Iimp, respectively).

Conversely, regression trees predicted 25%

and 14% of the variance of the RII and

Iimp indices for Shrub microsites and their

target species (Fig. 5.4). Values of PD

between the shrubs and their target species

between 207 and 272.8 million years (Myr)

rendered negative or neutral interactions,

while values of PD outside this values

(<207 Myr or >272.8 Myr) indicated

positive results for both RII and Iimp

indices. Climate was a modulator of

secondary importance for these pairwise

interactions. When 207 Myr < PD < 272.8

Myr, values of Climate higher than 47.7

(the wettest sites) render negative RII

values, which were neutral otherwise.

When 207 Myr < PD > 272.8 Myr, RII

values were positive in the dryer sites

(Climate values < 127.2), but neutral in the

rest of climatic conditions (Fig. 5.4A).

Values of the Iimp index rendered slightly

different results; when PD > 272.8 Myr

and Climate values < -67 (dryer sites),

shrubs were not important for the

performance of their target species; when

Climate values where higher than -67 (mid

to wet sites), shrubs exerted a negative

effect upon their target species (Fig. 5.4B).

DISCUSSION

By jointly considering information on

abiotic conditions, state-of-the-art

phylogenetic tools, and different aspects of

biotic non-trophic interactions

(competition/facilitation shifts, niche

expansion and niche differentiation), we

were able to explore the relative

importances of plant-plant interactions and

the environment as drivers of the

phylogenetic structure of the studied

communities. Our results show that these

interactions are important to determine

phylogenetic structure along a wide

environmental gradient, but that the effect

of climatic conditions on this structure is

CHAPTER 5

169

indirectly mediated by its effects upon the

“double environmental filtering” provided

by the difference in microclimatic

conditions between nurse plants and Open

microsites. The phylogenetic evenness

promoted by biotic interactions and the

phylogenetic clustering promoted by

climatic conditions, mediated by its effects

on plant-plant interactions, caused a

random phylogenetic structure in most of

the studied communities. At the pairwise

level, we found that PD was a primary

modulator of plant-plant interactions, with

Climate playing a secondary role. PD

values between 207-278.2 Myr rendered

negative outcomes, while PD values

outside this range yielded positive or

neutral outcomes, depending on

environmental conditions.

DIRECT AND INDIRECT EFFECTS OF

ENVIRONMENTAL FILTERING AND

BIOTIC INTERACTIONS ON THE

COMMUNITY PHYLOGENETIC

STRUCTURE

In sharp contrast with previous studies

(Michalet et al. 2006, Callaway 2007), we

did not find any relationship between our

surrogates of biotic interactions or niche

expansion at the community level (C-score

and percentage of facilitation obligates

species, respectively) and the climatic

variables measured. This does not mean

that plant-plant interactions were not

important at this level of organization, but

rather that these interactions, mainly

positive ones, importantly influenced

community assemblage through niche

expansion and habitat differentiation

across the entire environmental gradient

studied (Fig. 5.2; Fig. S6). How does the

equal importance of biotic interactions

across wide environmental gradients affect

the relative roles of such interactions and

environmental filtering in determining the

phylogenetic structure of plant

communities?

When evaluating the effects of

climatic conditions or biotic interactions on

the phylogenetic structure of the studied

communities separately, we found support

for the patterns found by previous studies:

while competition increased the

phylogenetic evenness of a given

community, environmental filtering did the

contrary (Webb et al. 2002, Cavender-

Bares et al. 2004). However, when

considering both processes and their

interaction together, our results showed

that phylogenetic structure is determined

by both direct and indirect effects mediated

by climatic conditions and biotic non-

trophic interactions (Fig. 5.2), which

finally caused a random phylogenetic

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

170

pattern. To our knowledge, such

interaction has not been described before.

These results are in the line of the most

recent literature, which suggests that the

phylogenetic structure alone, even with a

perfect knowledge of the trait conservatism

of co-occurring species, may not be

enough to infer the implied mechanisms in

the assemblage of a given community

(Pausas and Verdú 2010). Our study

demonstrates that this is particularly true

when dealing with more than one

important mechanism shaping natural

communities, something fairly common in

nature (Butterfield et al. 2010). In contrast

with previous studies (e.g. Valiente-Banuet

and Verdú 2007, Verdú et al. 2009), we did

not find any increase in phylogenetic

evenness with the increase of facilitation

importance, as shown by the C-score

measurements at the community level.

This could be explained because of

contrasting ontogenetic stages comparing

to those studies (more adult-adult

interactions in our sampled sites than in the

others; Valiente-Banuet and Verdú 2008),

or because of some of the patchiness

registered with the C-score may be more

related to a shared dispersion syndrome

than to facilitation processes, and therefore

this could lead to a phylogenetic clustering

(see additional discussion below).

However, the increase of the relative

importance of niche expansion (sensu

Bruno et al. 2003), another important

surrogate of facilitation at the community

level, promoted phylogenetic evenness, in

agreement with previous findings

(Valiente-Banuet and Verdú 2007, Verdú

et al. 2009). This latter result highlights the

necessity of measuring different indicators

of plant-plant interactions, and not only co-

occurrence patterns, to correctly assess the

role of such interactions on the

phylogenetic structure of plant

communities.

We were able to establish how an

increase in rainfall counter-intuitively

acted as an environmental filtering,

increasing community phylogenetic

clustering, and how its effect was mediated

by the reduction in the habitat

differentiation provided by nurse plants

(more rainfall reduced the difference

between MPD and MPDfac; Fig. 5.2). Our

results also highlight the importance of

considering not only multiple processes,

but also employing different proxies to

identify their direct and indirect effects on

community assemblage and its

phylogenetic structure. Using only co-

occurrence patterns, as previously

suggested (Pausas and Verdú 2010), would

not suffice to detect these indirect effects.

To take into account the whole set of

processes affecting phylogenetic structure

may also help to disentangle the

importance of phylogenetic diversity per se

CHAPTER 5

171

on ecosystem functioning or other

important ecosystem services (e.g. Forest

et al. 2007, Maherali and Klironomos

2007, Cavender-Bares et al. 2009),

avoiding potential confounding effects

produced by different processes causing a

particular phylogenetic pattern (Mayfield

and Levine 2010).

IS THE EVOLUTIONARY

RELATIONSHIP MORE IMPORTANT

THAN ABIOTIC CONDITIONS

DEFINING PAIRWISE INTERACTIONS

OUTCOMES?

A hierarchy between both PD and climatic

conditions modulated the outcome of the

large set of pairwise interactions tested. In

our case, PD was the primary factor

affecting such interactions, while climatic

conditions played a secondary role (Fig.

5.4). We found a double threshold in PD

values that defined the sign of the

interactions, with PD values between 207

and 272.8 Myr always rendering

competition, and values outside these

thresholds always leading to facilitation.

While results from the upper threshold (PD

> 272.8 Myr leading to facilitation) agree

with current literature (Valiente-Banuet

and Verdú 2007, 2008, Castillo et al.

2010), those from the lower threshold

(facilitation with PD < 207 Myr) do not.

The idea that a threshold in the PD

between the involved species may define

the outcome of their interaction is

appealing, and fits surprisingly well –

although the authors did not discuss the

data in that way– results from a recent

study in the Mexican scrubland (Castillo et

al. 2010). These authors found that lower

PDs always indicate negative interactions,

while higher PDs could mean either

positive or negative outcomes (see Figs 2

and 3 in Castillo et al. 2010); in our case,

this threshold would be PD < 272.8 Myr.

Conversely, the lower PD threshold could

be related not with a facilitatory effect of

the nurse shrub, but with the role of the

shrubs studied as a refuge for animals and

their perch effect (Pausas et al. 2006). This

could promote an increase in the

deposition of seeds by animals of other

Tertiary and animal-dispersed shrubs, and

therefore foster the co-occurrence of

species phylogenetically related with the

studied shrubs by nucleation processes

(Herrera 1992, Verdú and García-Fayos

1996). The fact that these results may be

related mostly to a dispersion syndrome,

rather than to the facilitatory effect of the

shrubs themselves, could be a potential

explanation of the contrasting results found

in other studies regarding this low

threshold in PD values (Castillo et al.

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

172

2010), and regarding the effect of

facilitation in the phylogenetic structure

(we found an increase in phylogenetic

clustering with lower SES C-scores, but

Valiente-Banuet and Verdú [2007] or

Verdú et al. [2009] found the contrary

using other measurements of co-

occurrence patterns). What factors, then,

determine the outcome of pairwise

interactions when PD is high enough to

allow positive interactions to occur? Our

results show that environmental conditions

play a major role once this threshold has

been reached in some cases, with dryer

sites usually yielding more positive

outcomes (Callaway 2007). However, we

wish to highlight that only 25% of the

variance of the tested interaction outcomes

were predicted by our regression trees.

Therefore, other unmeasured factors, such

as different ontogenetic stages of the

involved plants (Valiente-Banuet and

Verdú 2008), differences in herbivory

pressure among sites (Smit et al. 2009), or

ecophysiological traits labile through

evolutionary time frames and therefore not

detected with our phylogenetic approach

(Cavender-Bares et al. 2004, 2009), could

be important factors affecting such

interactions (e.g. Liancourt et al. 2005).

Such unmeasured factors could play a

major role in defining the outcomes of the

interaction between Stipa and its target

species (e.g. Soliveres et al. 2010;

Soliveres et al. in press), since nor PD

neither climatic conditions were good

predictors of such outcomes.

CONCLUDING REMARKS

Inferring the mechanisms shaping plant

communities from their phylogenetic

structure alone may drive to misleading

conclusions (Myfield and Levine 2010).

The use of manipulative experiments

including different communities under

contrasting environmental conditions, and

the measurement of the direct and indirect

effects of biotic interactions have been

recommended to overcome these

limitations (Cavender-Bares et al. 2009,

Vamosi et al. 2009). However, such

experiments are often logistically

prohibitive. The observational and

analytical approach employed here may

serve as an alternative to experimentation,

and can help to avoid misleading

conclusions when inferring the several

possible mechanisms underlying the

assemblage of natural communities. Our

study illustrates the complexity of direct

and indirect effects of environmental

filtering and biotic interactions as drivers

of the phylogenetic structure of natural

communities across environmental

gradients. It also highlights the necessity of

taking into account environmental

conditions and different components of the

biotic non-trophic interactions when

CHAPTER 5

173

studying the processes shaping plant

communities and their phylogenetic

structure.

This study also represents, to our

knowledge, the first attempt to test the

relative importance and possible

interactions between the evolutionary

relationships of two interacting species and

climate as drivers of the outcome of their

interactions. Interactions between different

factors affecting plant-plant interactions

may lead to counterintuitive or antagonistic

responses and should be studied jointly to

properly infer their relative importances as

drivers of such interactions (Baumeister

and Callaway 2006, Soliveres et al. in

press). Regression trees revealed as a

useful tool for detecting the hierarchy and

the non-linear responses in the effect of

climate and PD on such interactions. We

found that PD was of primary importance

acting as a threshold, which could explain

the species-specific nature often found in

plant-plant interactions (Callaway 2007).

Conversely, climatic conditions affected

these interactions only once PD reached

this particular threshold. In contrast, other

unmeasured factors in the case of Stipa,

such as herbivory, ontogenetic stage or

specific ecophysiological traits not

included in the PD relationships seemed

more important drivers of such outcomes,

illustrating the difficulties of predicting

pairwise interactions outcomes with simple

models involving just one or few

predictors.

ACKNOWLEDGMENTS Estrella Pastor, Beatriz Amat, Luis Cayuela, María D. Puche, Matt Bowker, and Pablo García-Palacios assisted with fieldwork and/or statistical analyses. Adrián Escudero provided help with species identification. SS was supported by a PhD fellowship from the EXPERTAL project, funded by Fundación Biodiversidad and CINTRA S.A. FTM acknowledges support from the European Research Council under the European Community's Seventh Framework Programme (FP7/2007-2013)/ERC Grant agreement n° 242658. This research was funded by the CEFEMED, and INTERCAMBIO (BIOCON 06/105) projects, funded by the Universidad Rey Juan Carlos-Comunidad de Madrid, and Fundación BBVA, respectively.

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

174

Supplementary material for Chapter 5.

Appendix K. Detailed materials and methods.

SUITABILITY OF STIPA TENACISSIMA STEPPES

Stipa tenacissima steppes are well suited ecosystems for testing the relative roles of biotic

interactions and habitat filtering in the phylogenetic structure of plant communities, and to

assess the relative importances and possible interactions between PD and climate as drivers of

plant-plant interactions outcomes for different reasons. Important ecophysiological traits are

extremely well-conserved through evolutionary time in the Mediterranean flora (Herrera

1992; Valiente-Banuet et al. 2006), which avoid potential confounding factors in the

interpretation of a given phylogenetic structure (Webb et al. 2002). Conversely, the PD

between two interacting species is a good indicator of dissimilarity in their ecological niches

and traits (Valiente-Banuet et al. 2006, Valiente-Banuet and Verdú 2007). Furthermore, S.

tenacissima steppes are one of the most extended Mediterranean semiarid community types

(Le Houreu 2001), and are strongly shaped by facilitation and its interaction with abiotic

stress (e.g., Maestre and Cortina 2004a, Armas and Pugnaire 2005). See Maestre et al.

(2009b) for additional details on the natural history of these communities.

SYNTHESIZING CLIMATIC CONDITIONS WITH PCA

Eight climatic variables (annual radiation, minimum, maximum and mean temperature,

annual rainfall, temperature range [maximum-minimum], and minimum and maximum

temperatures for the coldest and warmest month, respectively) were collected for each site

using available climatic models (Ninyerola et al. 2005). We reduced them to a single synthetic

variable using PCA to obtain a more general assessment of the influence of all of our

environmental variables at both the community and pair of species levels. We used the first

axis of this PCA (referred in the main text to as Climate) as our surrogate of the climatic

gradient present at our sites. This axis explained 88.6% (Eigenvalue = 8.08·103) of the

variance in the climatic data, and was highly correlated with both rainfall and radiation

(Eigenvectors = -0.864 and 0.502 for rainfall and radiation, respectively; the remainder of the

eigenvectors were < 0.03 in all cases). PCA was carried out using the Primer v. 6 statistical

package for Windows (PRIMER-E Ltd., Plymouth Marine Laboratory, UK).

CHAPTER 5

175

METHODOLOGICAL DETAILS OF THE PHYLOGENETIC HYPOTHESIS

The phylogenetic distances between species pairs were estimated by assembling a

phylogenetic hypothesis for all the species included in this study using Phylomatic2 (avalaible

at: http://www.phylodiversity.net/phylomatic/phylomatic.html). All the families in our data

set matched the family names of the angiosperm megatree used in Phylomatic

(R20091110.new), which was based on the Angiosperm Phylogeny Group III phylogenetic

classification of flowering plant orders and families (APG 2009). Within-family phylogenetic

relationships were further resolved based on data from various published molecular

phylogenies (Asteraceae: Funk et al. 2005; Susanna et al. 2006; Cistaceae: Guzmán and

Vargas 2009; Guzmán et al. 2009; Fabaceae: Allan and Porter 2000, Allan et al. 2004,

Wojciechowski et al. 2004; Poaceae: Bouchenak-Khelladi et al. 2008, 2010; Rubiaceae:

Bremer and Eriksson 2009).

Once we had assembled the phylogenetic hypothesis, we adjusted its branch lengths

with the help of the Phylocom BLADJ algorithm (Webb et al. 2008), which fixes the age of

internal nodes based on clade age estimates, whereas undated internal nodes in the phylogeny

are spaced evenly between dated nodes to minimize tree-wide variance in branch length

(Webb et al. 2008). Thus, BLADJ is a simple tool that fixes the root node of a phylogeny at a

specified age and fixes the other nodes for which age estimates are available. It sets all other

branch lengths by placing the nodes evenly between dated nodes, as well as between dated

nodes and terminals (of Age 0). The Phylocom manual (Webb et al. 2008) suggests using the

age estimates from Wikström et al. (2001); however, new analyses estimating divergence

times for angiosperms have been published since the publication of this seminal work (e.g.

Bremer et al. 2004, Anderson et al. 2005, Magallón and Castillo 2009, Smith et al. 2010,

Wang et al. 2010). In addition, TimeTree (Hedges et al. 2006), a public knowledge-base of

divergence times among organism is publicly available online (http://www.timetree.net). This

utility allows exploration of the thousands of divergence times among organisms in the

published literature (Hedges et al. 2006). A tree-based (hierarchical) system is used to identify

all published molecular time estimates bearing on the divergence of two chosen taxa, such as

species, compute summary statistics, and present the results. We mainly used this database to

fix the ages of internal nodes on our phylogenetic hypothesis, completing TimeTree results

with other published sources when this database did not provide any date (Cistaceae: Guzmán

and Vargas 2009, Guzmán et al. 2009; Asteraceae: Kim et al. 2005, Torices 2010; Poaceae:

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

176

Bouchenak-Khelladi et al. 2010; Fabaceae: Lavin et al. 2005, Bello et al. 2009; Brassicaceae:

Franzke et al. 2009; Caryophyllaceae: Valente et al. 2010). Eventually, we fixed the age of 48

internal nodes (see Table S2).

Apart from the data showed in the text, we used this tree to calculate the mean nearest

phylogenetic taxon index (MNTD; Webb et al. 2002), which measures the phylogenetic

distance to the most closely related neighbor. This index was calculated using the Picante

package for R (Kembel et al. 2010). Since MNTD was highly correlated with MPD

(Pearson´s correlation: r = 0.87; P < 0.0001), and its results were very similar to those

obtained with the MPD index, we only used results from the latter (presented in Table S1).

MEASURING CO-OCCURRENCE PATTERNS AND PLANT-PLANT INTERACTIONS

OUTCOMES

We measured co-occurrence patterns in each community by using null models based on

patterns of species co-occurrence found with the 80 1.5 m × 1.5 m quadrats (Gotelli and

Graves 1996). We estimated species co-occurrence with the C-score index, a metric

commonly used by studies aiming to infer species interactions at the community level from

co-occurrence data (e.g. Rooney et al. 2008, Maestre et al. 2008, Bowker et al. 2010). This

index is calculated for each pair of species as (Ri - S)(Rj - S), where Ri and Rj are the number

of total occurrences for species i and j, and S is the number of quadrats in which both species

occur. This score is then averaged over all possible pairs of species in the matrix (Gotelli

2000). The C-score is related to the competitive exclusion concept of “checkerboardness”

i.e., how many of the possible species pairs in a given community never appear in the same

quadrat together. Thus, positive and large values of this index indicate that competition may

be a prevalent mechanism determining the co-occurrence patterns observed (Gotelli 2000).

As the values of the C-score are dependent on the number of species and co-

occurrences observed within each plot, we obtained a standardized effect size (SES) as (Iobs -

Isim) ⁄ Ssim, where Iobs is the observed value of the C-score, and Isim and Ssim are the mean and

standard deviation, respectively, of this index obtained from the n simulations performed

(Gotelli and Entsminger 2006). Standardized Effect Size (SES) values of the C-score less than

or greater than zero indicate prevailing spatial segregation (competition prevalence) and

aggregation (facilitation dominance; Tirado and Pugnaire 2005) among the species within a

community, respectively. We used ‘fixed rows–equiprobable columns’ null models and 5000

simulations. With this approach, each species conserved its own abundance (rare species

remained rare and common species remained common) and each quadrat was assumed to

CHAPTER 5

177

have the same probability of being colonized as the remainder, regardless of the number of

species found in each quadrat during the simulations. This null model has been recommended

for standardized samples collected in homogenous habitats (Gotelli 2000), such as the ones

gathered in this study. We also tried the “fixed rows-fixed columns” algorithm (both species

and quadrats conserved its relative abundance and richness, respectively) to add confidence to

our conclusions. The results obtained with these analyses were similar than those obtained

with the ‘fixed rows–equiprobable columns’ null model, and thus are not shown.

We also calculated interaction intensity and importance indices (RII; Armas et al.

2004 and Iimp; Seifan et al. 2010, respectively). RII indices were calculated as (PNurse –

POpen)/(PNurse + POpen), where PNurse was the number of individuals under the canopy of a nurse

plant (Stipa or Shrub) and POpen was the number or individuals recruited in the Open

microsite. Alternatively, Iimp indices were calculated as Iimp= Nimp/│Nimp│+│Eimp│, where

Nimp and Eimp were the nurse plant and environmental contributions to the total number of

individuals recruited for each species, respectively. Nimp was calculated as PNurse – POpen, and

Eimp as POpen – MPOpen/Nurse, where MPOpen/Nurse is the maximum number of recruited

individuals for a given species found in the entire gradient, irrespective of the microsite

sampled.

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

178

Table S9. Main characteristics of the study sites. MPD = Mean phylogenetic distance and SES = Standardized effect size. The SES of the MPD was calculated as MPDobs – meanMPDnull/ sdMPDnull; where MPDobs is the observed phylogenetic distance in our sample, and meanMPDnull and sdMPDnull are the mean and standard deviation of the phylogenetic distances obtained from the n simulated communities under the null model. Positive SES, and with a P-value > 0.95 indicate significant phylogenetic evenness in the sampled community, while negative SES with P-values < 0.05 indicate phylogenetic clustering.* Indicates a significant phylogenetic clustering in this community. The dominant shrub species is Quercus coccifera in all sites excepting at Carrascoy, where it was Rhamnus lycioides.

Site

Latitude

Longitude

Annual rainfall

(mm)

Species richness

MPD (SES) P (MPD)

Barrax 39º02’91’’N 2º13’82’’W 433 24 -0.88 0.22

Camporreal 40º19’72’’N 3º25’36’’W 457 9 -0.77 0.16

Carrascoy 37º48’02’’N 1º18’32’’W 282 38 -0.09 0.42

Crevillente 38º14’15’’N 0º55’49’’W 273 36 1.60 0.91

El Ventós 38º28’14’’N 0º37’03’’W 319 35 0.37 0.65

Morata 40º27’62’’N 3º05’31’’W 455 22 -1.01 0.15

Sierra Espuña 37º49’27’’N 1º40’41’’W 364 32 -1.49 0.04*

Titulcia 40º11’28’’N 3º30’13’’W 440 21 -0.84 0.24

Villarrobledo 39º12’64’’N 2º30’77’’W 446 18 -0.83 0.18

Yecla 38º35’40’’N 1º12’15’’W 350 47 0.52 0.74

Zorita 40º21’30’’N 2º52’62’’W 434 38 1.52 0.91

CHAPTER 5

179

Table S10. Estimated age of branching events.

Node

label

Estimated

time (Myr) Reference

1 334.5 Hedges et al. (2006) 2 298.0 Hedges et al. (2006) 3 150.1 Hedges et al. (2006) 4 147.8 Hedges et al. (2006) 5 122.0 Hedges et al. (2006) 6 101.0 Hedges et al. (2006) 7 67.0 Hedges et al. (2006) 8 112.0 Hedges et al. (2006) 9 57.0 Bouchenak-Khelladi et al. (2010) 10 44.7 Bouchenak-Khelladi et al. (2010) 11 28.0 Bouchenak-Khelladi et al. (2010) 12 125.0 Hedges et al. (2006) 13 121.0 Hedges et al. (2006) 14 109.0 Hedges et al. (2006) 15 98.0 Hedges et al. (2006) 16 69.0 Hedges et al. (2006) 17 89.0 Hedges et al. (2006) 18 94.0 Hedges et al. (2006) 19 84.0 Bello et al. (2009) 20 55.0 Lavin et al. (2005) 21 24.6 Lavin et al. (2005) 22 97.5 Hedges et al. (2006) 23 84.8 Hedges et al. (2006) 24 39.0 Wikström et al. (2001) 25 19.0 Franzke et al. (2009) 26 51.0 Hedges et al. (2006) 27 14.5 Guzmán and Vargas (2009); Guzmán et al. (2009) 28 2.0 Guzmán and Vargas (2009); Guzmán et al. (2009) 29 1.6 Guzmán and Vargas (2009); Guzmán et al. (2009) 30 6.1 Guzmán and Vargas (2009); Guzmán et al. (2009) 31 122.0 Hedges et al. (2006) 32 122.0 Hedges et al. (2006) 33 55.8 Valente et al. (2010) 34 16.6 Valente et al. (2010) 35 73.5 Wikström et al. (2001) 36 113.0 Magallón and Castillo (2009); Anderson et al. (2005); Janssens et al. 2009) 37 111.9 Magallón and Castillo (2009); Anderson et al. (2005); Janssens et al. 2009) 38 96.6 Magallón and Castillo (2009); Janssens et al. (2009) 39 47.0 Kim et al. (2005) 40 35.5 Kim et al. (2005) 41 24.7 Torices (2010) 42 108.0 Hedges et al. (2006) 43 17.0 Bremer and Eriksson (2009) 44 106.0 Hedges et al. (2006) 45 90.0 Hedges et al. (2006) 46 76.0 Hedges et al. (2006) 47 23.0 Wikström et al. (2001) 48 8.5 Stevens (2010)

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

180

Figure S4. Phylogenetic tree of the regional species pool. Phylogenetic hypothesis are developed in Appendix S1. Main nurse plants are highlighted in blue. Node labels are given in Table S9.

CHAPTER 5

181

0 100 200 300 400 500 600 700 800

num

ber

of c

ases

0

20

40

60

80

100

120

140

160

180

Phylogenetic distance nurse-target species

100 200 300 400 500 600 700 800

num

ber

of c

ases

0

20

40

60

80

100

120

A

B

Figure S5. Frequency histogram of the phylogenetic distances between the different target plants and their nurses, either Stipa tenacissima (A) or shrubs (B). Notice the discontinuous distribution and the skewness of the data.

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

182

ENVIRONMENTAL CONDITIONS

Climatic PCA axis values

-200 -150 -100 -50 0 50 100

MP

D (

SE

S)

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

R2= 0.30; P = 0.09

- RAINFALL +

Phylogenetic evenness

Phylogenetic clustering

CHAPTER 5

183

HABITAT × BIOTIC INTERACTIONS

-200 -150 -100 -50 0 50 100

C-s

core

-3

-2

-1

0

1

2

3

R2 = 0.13; P = 0.57

Climatic PCA axis values

-200 -150 -100 -50 0 50 100

% fa

cilit

atio

n ob

ligat

es

5

10

15

20

25

30

35

40

R2 = 0.12; P = 0.30

-200 -150 -100 -50 0 50 100

Sim

ilarit

y in

dex

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Open vs. Stipa: R2 = 0.05; P = 0.54 Open vs. Shrub: R2 = 0.01; P = 0.80

Climatic PCA axis values

-200 -150 -100 -50 0 50 100

MP

Dge

n -

MP

Dcl

ust

-2

-1

0

1

2

3

4

MPDfacilitated: R2 = 0.68; P = 0.006MPDnon-facilitated: R2 = 0.05; P = 0.56

Phylogenetic clusteringamong facilitated/non-facilitated species

Climatic PCA axis values

- RAINFALL +

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

184

BIOTIC INTERACTIONS

C-score-3 -2 -1 0 1 2 3

MP

D

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

R2 = 0.36; P = 0.07

Similarity index0.4 0.5 0.6 0.7 0.8 0.9 1.0

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

Open vs. Stipa: R2 = 0.01; P = 0.75Open vs. Shrub: R2 = 0.30; P = 0.09

% facilitation obligates

10 20 30 40

MP

D

-2.0

-1.5

-1.0

-0.5

0.0

0.5

1.0

1.5

2.0

R2 = 0.01; P = 0.77

MPDgen - MPDclust

-2 -1 0 1 2 3 4

Sim

ilarit

y in

dex

0.4

0.5

0.6

0.7

0.8

0.9

1.0

MPDfac vs Open-Stipa similarity MPDfac vs Open-Shrub similarity MPDnf vs Open-Stipa similarity MPDnf vs Open-Shrub similarity

Figure S6. Detailed results of the conceptual diagram shown in Figure 5.2. Results are organized following sections provided in Figure 5.2 (effects of environmental conditions, biotic interactions and habitat × biotic interactions). In all cases, linear regressions results are shown in each plot.

CHAPTER 5

185

BIOTIC AND ABIOTIC FILTERS FOR PHYLODIVERSITY

186

DISCUSIÓN Y CONCLUSIONES GENERALES

DISCUSIÓN GENERAL

188

189

DISCUSIÓN GENERAL

Los trabajos presentados en esta tesis doctoral constituyen una de las evaluaciones más

exhaustivas sobre la evolución de las interacciones planta-planta a lo largo de gradientes

ambientales realizadas hasta la fecha. Por un lado, se ha testado la evolución de la interacción

entre Stipa tenacissima y Retama sphaerocarpa a lo largo de cuatro puntos contrastados de

disponibilidad hídrica, derivados de predicciones realistas sobre futuros escenarios de cambio

climático (IPCC 2007, capítulo 1). Por otro lado, se ha testado el efecto de la co-ocurrencia de

la aridez con distintos factores de importancia en la determinación del signo y la intensidad de

las interacciones planta-planta, como son la herbivoría (p. ej. Baraza et al. 2006), la ontogenia

(p. ej. Miriti 2006) o la relación evolutiva entre las especies implicadas (p. ej. Castillo et al.

2010). Estas aproximaciones han revelado complejas interacciones y jerarquías existentes

entre todos estos factores (capítulos 2, 3 y 5). Los experimentos realizados a nivel de

comunidad han permitido evaluar la generalidad de los distintos modelos propuestos sobre la

evolución de las interacciones planta-planta a lo largo de gradientes ambientales (p. ej.

Bertness y Callaway 1994, Michalet et al. 2006, Maestre et al. 2009a), tanto a nivel de par de

especies como al de comunidades vegetales enteras (capítulo 4).

Los resultados obtenidos apuntan a que –por orden de importancia– la herbivoría, las

características ecológicas de las especies implicadas (deducidas tanto a partir de la tolerancia

a distintos factores de estrés o fases ontogenéticas como de su historia evolutiva) y, por

último, las condiciones climáticas y sus interacciones con los factores anteriormente

mencionados, son los factores clave que definen el signo y la intensidad de las interacciones

entre pares de especies vegetales. A pesar de que la mayoría de estudios sobre la dinámica de

las interacciones planta-planta en medios semiáridos se centran en la disponibilidad hídrica

como factor abiótico clave (p. ej. Holzapfel y Mahall 1999, Tielbörger y Kadmon 2000,

Pugnaire y Luque 2001, Maestre y Cortina 2004a), los trabajos presentados en esta tesis

apuntan a la disponibilidad de luz y a la intolerancia a la sombra de las especies beneficiarias,

o de sus distintas fases ontogenéticas, como un factor de gran importancia a la hora de definir

dichas interacciones y su relación con la disponibilidad hídrica (Valladares y Pearcy 2002,

Prider y Facelli 2004, Valladares et al. 2008, Seifan et al. 2010). Por otro lado, se ha

demostrado que el patrón temporal, y no sólo la cantidad, de lluvia es algo importante a

considerar si queremos entender la evolución de estas interacciones a lo largo de gradientes

DISCUSIÓN GENERAL

190

ambientales, o su respuesta frente a futuros escenarios climáticos (Zavaleta 2006, Knapp et al.

2008). Los resultados obtenidos apuntan a que, mientras que el aumento de aridez puede

conllevar un aumento de la competencia entre las especies estudiadas, el incremento de los

eventos de lluvia torrencial puede acelerar la segregación de nicho entre las especies

implicadas, reduciendo este efecto competitivo. Los efectos de ambos patrones (aumento de

aridez y eventos torrenciales) dependerán de las tolerancias relativas a la sombra y a la sequía

de la especie beneficiaria, y de la posibilidad de que se produzca una segregación de nicho

efectiva entre la planta nodriza y su beneficiaria (Holmgren et al. 1997, Knapp et al. 2008,

capítulos 1–3).

Los múltiples factores que afectan de forma conjunta a las interacciones planta-planta

(herbivoría, aridez y diferentes características ecológicas), que comúnmente coexisten en los

ecosistemas naturales, hacen que sea difícil desarrollar modelos generales que predigan la

evolución de estas interacciones a lo largo de gradientes ambientales. Especialmente

discutibles son aquellas aproximaciones que asumen un “gradiente de estrés” que afecta

igualmente a todas las especies de una comunidad dada, ya que éstas difieren en sus

adaptaciones ecofisiológicas y, por tanto, en sus tolerancias a los distintos factores de estrés

que representan unas condiciones ambientales dadas (Chapin et al. 1987, Greiner la Peyre et

al. 2001, Körner 2003).

Los resultados presentados en el capítulo 4 apuntan a que los efectos positivos de las

plantas nodriza sobre la riqueza de especies a nivel de comunidad están promovidos tanto por

la expansión de nicho como por la reducción en la exclusión competitiva entre plantas

vecinas. Estos efectos positivos se mantienen constantes a lo largo de gradientes ambientales

donde coexisten distintos factores de estrés inversamente relacionados, o no relacionados

entre sí (p. ej. aridez y bajas temperaturas), al contrario de lo que predicen los modelos

teóricos actuales (p. ej. Lortie et al. 2004a, Michalet et al. 2006). Esta falta de variación de la

importancia de las interacciones positivas a lo largo de estos gradientes ambientales deriva de

la inexistencia de una gradiente de estrés único que afecta a todas las especies presentes en

una comunidad. Las distintas condiciones microclimáticas que se encuentran bajo el dosel de

las plantas nodriza beneficiarán a las especies menos adaptadas a las condiciones locales, sean

éstas las que sean (Greyner la Peyre et al. 2001, Liancourt et al. 2005). Sin embargo, en

gradientes ambientales gobernados por factores de estrés únicos o correlacionados entre sí (p.

ej. aridez, bajas temperaturas), las interacciones entre plantas seguirán una relación unimodal,

con predominio de las interacciones positivas a niveles medios-altos de estrés y de las de

191

competencia en ambos extremos del gradiente, tal como predicen los modelos teóricos en la

actualidad (Michalet et al. 2006, Holmgren y Scheffer 2010).

Por último, en esta tesis se muestran, por primera vez, los efectos conjuntos, directos e

indirectos, de las condiciones climáticas (filtro de hábitat) e interacciones bióticas (distintos

indicadores de competencia/facilitación) en el patrón filogenético de comunidades vegetales

semiáridas (capítulo 5). El patrón filogenético encontrado fue aleatorio en la mayoría de

casos, debido al efecto conjunto de un incremento de la dispersión filogenética promovida por

las interacciones bióticas y del aumento de la agregación en este patrón causado por los

efectos indirectos del clima sobre dichas interacciones. Los procesos de nucleación, derivados

de la dispersión zoocora concentrada en arbustos y árboles remanentes que comúnmente se

encuentra en ambientes Mediterráneos (Verdú y García-Fayos 1996, Méndez et al. 2008),

también parecen importantes en esta agregación del patrón filogenético. Los arbustos

remanentes tienden a concentrar semillas de otros arbustos dispersados por animales, un

síndrome de dispersión altamente conservado a lo largo de la evolución en la flora

Mediterránea (Herrera 1992); por tanto, estos procesos de nucleación se dan entre especies

evolutivamente próximas entre sí, causando agregación en el patrón filogenético. Se ofrecen

en este último capítulo una serie de herramientas, fáciles de medir e interpretar, que pueden

ayudar a evitar conclusiones erróneas derivadas de la inferencia de los mecanismos

implicados en el ensamblaje de especies vegetales a partir del patrón filogenético de las

comunidades en estudios observacionales (Cavender-Bares et al. 2009, Mayfield y Levine

2010, Pausas y Verdú 2010).

HACIA UN NUEVO MODELO SOBRE LA EVOLUCIÓN DE LAS INTERACCIONES ENTRE PARES DE

ESPECIES A LO LARGO DE GRADIENTES BIÓTICOS Y ABIÓTICOS

Las numerosas excepciones a la Hipótesis del Gradiente de Estrés (Bertness y Callaway 1994)

comúnmente encontradas en los sistemas naturales, han originado un debate sobre la

generalidad de sus predicciones (Maestre et al. 2005, 2006, Lortie y Callaway 2006, Callaway

2007, Lortie 2010) que ha dado lugar a nuevas interpretaciones y revisiones del modelo

inicial propuesto (Maestre et al. 2009a, Smit et al. 2009, Malkinson y Tielbörger 2010,

Holmgren y Scheffer 2010). De estas nuevas aproximaciones se concluye que la naturaleza

del estrés (si está relacionado o no con recursos directamente tomados por las plantas, p. ej.

luz o agua frente a temperatura o salinidad), la co-ocurrencia de distintos factores de estrés

abiótico o de éstos con la herbivoría, o las respuestas no lineales derivadas de las distintas

DISCUSIÓN GENERAL

192

tolerancias fisiológicas de las especies beneficiarias a los cambios microclimáticos

promovidos por las plantas nodriza, son factores clave a considerar para entender la evolución

de las interacciones planta-planta a lo largo de gradientes ambientales. Sin embargo, incluir

todos estos factores en un único modelo, que permita además inferir lo que ocurre en

comunidades vegetales enteras a partir del estudio de las interacciones entre uno o pocos

pares de especies es difícil. Incluso los modelos más recientes no son lo suficientemente

generales a la hora de arrojar predicciones sobre la evolución de la frecuencia, intensidad e

importancia de las interacciones planta-planta en la estructura y composición de las

comunidades vegetales a lo largo de distintos gradientes ambientales (pero véase Holmgren y

Scheffer 2010 para una aproximación muy meritoria). Esto ocurre porque es extremadamente

difícil que los modelos relativamente sencillos, que podrían ser extrapolables a numerosos

ecosistemas y fácilmente evaluables, incluyan la gran variedad de factores que influyen en las

interacciones planta-planta. Y modelos que incluyan todos los factores serían extremadamente

complejos y, por tanto, poco útiles.

La dicotomía entre complejidad y utilidad se ilustra con un modelo conceptual que

predice la evolución en las interacciones planta-planta a lo largo de gradientes ambientales en

zonas semiáridas a nivel de par de especies (Fig. B1). Para su elaboración se han tenido en

cuenta algunas de las aportaciones más significativas de los últimos estudios publicados, así

como los resultados presentados en esta tesis doctoral. En este modelo se evitan aspectos

problemáticos a la hora de elaborar predicciones generales, como son la existencia de un

gradiente general de estrés, que afecte a todas las especies de una comunidad por igual, o la

existencia de estrategias ecológicas que permanecen constantes a lo largo de gradientes

ambientales amplios (Travis et al. 2005, Michalet et al. 2006, Maestre et al. 2009a, Smit et al.

2009). Para ello, el modelo se basa en las tolerancias relativas de las especies implicadas a

factores ambientales concretos (derivadas bien de sus rasgos ecofisiológicos o de su historia

evolutiva), medibles y extrapolables a cualquier ambiente.

El modelo presentado en la Figura B1 toma como base el modelo teórico presentado

por Holmgren et al. (1997), que considera la eficacia biológica de las especies beneficiarias a

lo largo de un gradiente, desde el centro del dosel de la planta nodriza hasta un claro libre de

vegetación, bajo distintas condiciones de aridez. A este modelo se le ha añadido la respuesta

diferencial, y no lineal, promovida por las distintas tolerancias ecofisiológicas de las especies

implicadas o de sus distintas fases ontogenéticas, así como el efecto conjunto de distintos

niveles de aridez con el impacto de la herbivoría. Por simplicidad se asume que la carga de

herbívoros es constante a lo largo de las distintas condiciones ambientales (Smit et al. 2009).

193

Esta asunción podría ser problemática, ya que se ha demostrado que ambos factores de estrés

interactúan de diversas maneras y que, por tanto, no se puede considerar que varíen

independientemente (Illius y O´Connor 1999, Ibañez y Schupp 2001, Silliman et al. 2005,

Veblen 2008, capítulo 3). Además, el efecto de la herbivoría depende de la tolerancia a la

misma de las especies beneficiarias así como de la palatabilidad de las especies nodriza

(Baraza et al. 2006, Zamora et al. 2008). Ambos factores (tolerancia y palatabilidad) son

directamente dependientes del nivel de recursos existente (Crawley 1998, Baraza et al. 2004,

Wise y Abrahamson 2005, 2007, Smit et al. 2009). Las tolerancias ecofisiológicas han sido

organizadas en tres grupos: 1) especies con una tolerancia media a la sequía y a la sombra,

que constituyen la mayoría de especies existentes (Niinemets y Valladares 2006), 2) especies

con una tolerancia a la sombra muy alta, pero intolerantes a la sequía (p. ej., Hedera helix,

Rhagodia spinescens), y 3) especies altamente tolerantes a la sequía, pero intolerantes a la

sombra (p. ej., Retama sphaerocarpa, Helianthemum squamatum, Ambrosia dumosa). Debido

a la existencia de compromisos en las adaptaciones fisiológicas a ambos factores, no se

considera la existencia de especies con una alta tolerancia conjunta a la sombra y a la sequía,

ya que es poco probable que estas especies existan en la naturaleza (Niinemets y Valladares

2006). Por simplicidad, tampoco se pueden considerar en el modelo otras tolerancias

fisiológicas que podrían ser relevantes en zonas áridas y semiáridas, como la tolerancia a

suelos salinos o pobres en nutrientes (Pugnaire et al. 2004, Brady et al. 2005, Riginos et al.

2005, Armas y Pugnaire 2009), a bajas temperaturas (capítulo 4), o a vientos (Baumeister y

Callaway 2006). Asimismo, las tolerancias a la aridez y a la sombra no se dividen en

categorías discretas, ya que cada especie se localizará en un punto concreto dentro de un

continuo de rasgos ecológicos que les permitirán una mayor o menor tolerancia a las

condiciones ambientales (Kobe et al. 1995, Ackerly 2003, Niineemets y Valladares 2006).

En el modelo que se introduce en la Figura B1 se consideran también tres fases

ontogenéticas distintas: germinación y establecimiento temprano de las plántulas, juvenil, e

individuos adultos reproductivos. Estas fases son claves a la hora de definir 1) el

reclutamiento de nuevas especies en la comunidad (Escudero et al. 1999, Maestre et al. 2001,

Holmgren et al. 2006), y 2) la demografía de estas especies (Escudero et al. 2000, Miriti et al.

2007), contribuyendo todos ellos de forma sustancial a la dinámica de las comunidades

vegetales en las tierras secas (McAuliffe 1988, Eldridge et al. 1991, Eccles et al. 1999,

Butterfield et al. 2010). Además, se considera la posibilidad de una segregación de nicho

efectiva entre individuos adultos de las dos especies que interactúan, de forma que se predicen

diferentes cambios en las interacciones entre adultos si existe dicha segregación (Fowler

DISCUSIÓN GENERAL

194

1986, Sala et al. 1989, Stokes y Archer 2010), o si por el contrario sus nichos ecológicos se

solapan (Ludwig et al. 2004, Miriti 2006, Armas y Pugnaire 2009). Esta última separación

puede tener implicaciones importantes en la respuesta de las comunidades vegetales a los

cambios en el patrón de las precipitaciones predichos con el cambio climático (Schwinning y

Sala 2004, Knapp et al. 2008, capítulo 1). Para hacer más entendibles los resultados de las

interacciones derivados de este modelo, se han calculado también los resultados que serían

esperables en la intensidad e importancia de las interacciones (sensu Brooker et al. 2005, ver

capítulo 4) para cada estrategia ecológica, cada fase ontogenética y cada uno de los tres

niveles de aridez propuestos (Fig. B2).

195

Figura B1. Modelo teórico donde se predice la evolución de la eficacia biológica (fitness) en un continuo dosel de planta nodriza-claro libre de vegetación a lo largo de un gradiente de aridez (modificado de Holmgren et al. 1997). Las estrategias ecológicas incluidas son tolerancia media a la sequía y a la sombra (en verde), intolerancia a la sequía (en azul), e intolerancia a la sombra (en rojo). Se indican variaciones en dicha evolución a lo largo de distintas fases ontogenéticas para cada estrategia: germinación y establecimiento inicial de plántulas (panel superior), crecimiento y supervivencia de individuos juveniles (2º panel) y coexistencia entre individuos adultos sin (3er panel) o con segregación de nicho (4º panel). Las flechas en el interior de cada panel indican el efecto de una carga constante de herbívoros a lo largo del gradiente de aridez, que es indiferente a las tolerancias a la sombra o a la sequía. Este efecto consiste en un aumento del fitness más cerca del dosel de la planta nodriza, que incrementará su importancia a niveles medios de aridez. A niveles más elevados de aridez, la escasez de fuentes alternativas de forraje hará que los herbívoros incrementen su esfuerzo de búsqueda y acaben perjudicando incluso a las plantas bajo el dosel (flecha diagonal en los paneles de la derecha). Este efecto será mucho más marcado en juveniles que en adultos, como muestran las flechas en los paneles.

DISCUSIÓN GENERAL

196

Figura B2. Intensidad e importancia de las interacciones planta-planta predichas a lo largo de un gradiente de aridez y en distintas fases ontogenéticas (germinación, juvenil y adulto; de arriba a abajo) predichos a partir del modelo teórico presentado en la Figura B1. Las distintas estrategias ecológicas incluidas siguen el mismo código de colores de dicha figura. La línea discontinua en medio de cada panel corresponde con el 0 (interacción neutra) de los indicadores tanto de intensidad como de importancia de las interacciones. Valores por encima o debajo de 0 implican facilitación o competencia, respectivamente. Las líneas continuas y discontinuas del panel inferior indican interacciones entre adultos con y sin segregación de nicho, respectivamente.

TRES ESTRATEGIAS, TRES RESPUESTAS

Las especies tolerantes a la sombra se comportarán prácticamente del mismo modo a lo largo

de su ontogenia (Fig. B1). Su elevada tolerancia a la sombra hace que sean capaces de

aprovechar la mayor fertilidad en el suelo y disponibilidad hídrica comúnmente encontradas

bajo el dosel de la planta nodriza (p. ej. Franco y Nobel 1989, Callaway 2007). Por tanto, su

eficacia biológica siempre será mayor bajo una planta nodriza que en lugares libres de

vegetación, independientemente de las condiciones de aridez. Un ejemplo clásico de este tipo

de interacciones se da entre arbustos tolerantes a las altas temperaturas y los cactus

columnares asociados en el sur de Estados Unidos y norte de Méjico. Así, los cactus

columnares, como por ejemplo Carnegia gigantea (saguaro) o Neobuxbaumia tetetzo, ambos

197

intolerantes a las altas temperaturas, crecen y sobreviven únicamente asociados a la sombra de

plantas nodriza, que reducen la temperatura bajo su dosel (Shreve 1931, Niering et al. 1963,

Valiente-Banuet y Ezcurra 1991). Este comportamiento también se ha observado en la

herbácea Brachypodium retusum a lo largo del gradiente ambiental estudiado en el capítulo 4

(Fig. B3; ver también Maestre y Cortina 2004b), o en las especies Enchylaena tomentosa y

Rhagodia spinescens en el semiárido australiano (Hastwell y Facelli 2003, Prider y Facelli

2004). Especies intolerantes a altas radiaciones también se benefician de la presencia de

especies nodriza en condiciones más mésicas. Por ejemplo, Gómez-Aparicio et al. (2006)

encontraron efectos muy positivos de la reducción de la luz incidente para las especies Acer

opalus y Quercus pyrenaica en una zona seco-subhúmeda (871 mm de lluvia anual), por lo

que infirieron que el nicho de regeneración de estas especies se encontraría mayoritariamente

bajo el dosel de especies nodriza, fundamentalmente arbustos (Castro et al. 2002, Gómez-

Aparicio et al. 2004, 2005). En todos los casos mencionados, estas especies mostraron una

distribución restringida a la sombra de alguna especie nodriza bajo un amplio rango de

condiciones de aridez. Esto da lugar a interacciones positivas muy intensas (al haber una

eficacia biológica muy baja en los claros, cualquier índice que mida la intensidad de la

interacción siempre será muy positivo). Sin embargo, a medida que aumente la aridez,

algunas de estas especies se irán alejando de su óptimo ambiental, siendo más difícil su

reclutamiento en medios muy áridos (nótese la bajada en la eficacia biológica de estas

especies a medida que aumenta la aridez en la Fig. B1). Esta reducción de la eficacia

biológica puede llegar incluso a la desaparición de esta especie bajo esas condiciones en

particular, pese a la mejora microclimática promovida por la planta nodriza; esto provocaría

un colapso de las interacciones positivas para estas especies (Michalet et al. 2006, Forey et al.

2009). Si se diera esta desaparición, obviamente la intensidad de la facilitación pasaría en un

cambio brusco de muy positiva a cero (Fig. B2), ya que la especie ni siquiera estaría presente.

Al contrario que la intensidad, la importancia de la facilitación para las especies intolerantes a

la sequía se reducirá con el nivel de aridez. Esto ocurre porque la importancia es el efecto

relativo de la planta nodriza con respecto a otras condiciones ambientales (Brooker et al.

2005) y, como se aprecia en la Fig. B1, la eficacia biológica de estas especies será mucho

mayor en condiciones más húmedas, independientemente del efecto de la planta nodriza.

Además, el efecto de esta planta nodriza no será tan positivo en condiciones más áridas, ya

que el consumo de agua de esta nodriza puede llegar a superar a su mejora microclimática,

dando lugar a condiciones de “sombra seca” y a efectos netos negativos sobre las especies

vecinas (Valladares 2001, Valladares y Pearcy 2002).

DISCUSIÓN GENERAL

198

PCA climático-150 -100 -50 0 50 100

Índi

ces

de in

tera

cció

n

-1.0

-0.5

0.0

0.5

1.0

RII ST RII SH Iimp ST Iimp SH

FA

CIL

ITA

CIÓ

NC

OM

PE

TE

NC

IA

- LLUVIA + Figura B3. Índices de intensidad (RII) e importancia (Iimp) de la interacción de Brachypodium retusum con sus especies nodriza Stipa tenacissima (ST) y arbustos rebrotadores (SH) a lo largo del gradiente climático estudiado en los capítulos 4 y 5. Nótese la nula relación con el clima, y los relativamente altos valores de los índices calculados.

Las especies con una tolerancia media a la sombra y a la sequía presentarán un

comportamiento muy distinto a lo largo de su ontogenia. Por un lado, la germinación y

establecimiento temprano de estas especies será mayor bajo las plantas nodriza, debido

fundamentalmente a un suelo más fértil y a una menor radiación solar (p. ej. McAuliffe 1988,

Barberá et al. 2006). Este efecto positivo puede colapsar a medida que aumente la aridez, ya

que las plántulas de estas especies son más sensibles al estrés hídrico y pueden no germinar o

morir bajo niveles muy elevados de aridez, incluso considerando los cambios microclimáticos

de la planta nodriza (Kitzberger et al. 2000, Ibañez y Schupp 2001, Gasque y García-Fayos

2004). Esto dará lugar a una relación unimodal de la importancia e intensidad de las

interacciones a medida que aumente la aridez (Fig. B2). Los juveniles de estas especies, por

otro lado, competirán con la nodriza en condiciones más húmedas, tal como predicen todos

los modelos teóricos (p. ej. Bertness y Callaway 1994, Brooker y Callaghan 1998, Michalet et

al. 2006, Maestre et al. 2009a). Sin embargo, estas especies aumentarán su eficacia biológica

bajo el dosel de una planta nodriza en condiciones intermedias de aridez, fundamentalmente

por el efecto positivo de la sombra sobre el estado hídrico de los juveniles (Valiente-Banuet y

Ezcurra 1991, Maestre et al. 2003, Callaway 2007). A medida que aumente la aridez, la

199

intercepción de la escasa lluvia por el dosel de la planta nodriza (Tielbörger y Kadmon

2000a), o el incremento de la competencia entre ambas especies por el agua (Maestre y

Cortina 2004a), generarán condiciones de “sombra seca” que reducirán la eficacia biológica

de las especies beneficiarias bajo el dosel de la planta nodriza. Esto originará una relación

unimodal a lo largo del gradiente de aridez, tanto de la intensidad como de la importancia de

la facilitación sobre especies con una tolerancia media a la sombra y a la sequía (Tielbörger y

Kadmon 2000a, Maestre y Cortina 2004, Barchuck et al. 2005). En individuos adultos, la

competencia se extenderá hasta niveles medios de aridez cuando no haya separación de nicho,

ya que las especies competirán por el agua y perderán parte de los beneficios de la sombra

promovida por la planta nodriza al sobrepasar la altura de su dosel (Miriti 2006, Callaway

2007, Valiente-Banuet y Verdú 2008). Sin embargo, en condiciones más áridas, procesos

como el levantamiento hidráulico (hydraulic lift) o, simplemente, el efecto de la sombra sobre

la humedad del suelo puede incrementar el efecto positivo de la nodriza (Callaway 2007).

Esto generará una relación monotónica o unimodal entre la intensidad de la interacción y la

aridez, dependiendo de los efectos relativos de la sombra o el levantamiento hidráulico y la

competencia en el agua disponible para las plantas adultas facilitadas (podemos encontrar

efectos contrastados en Dawson 1993, Maestre et al. 2003 y Ludwig et al. 2004, Maestre y

Cortina 2004a). Sin embargo, si existe segregación de nicho, la intensidad (y la importancia)

de la facilitación aumentará de forma lineal con la aridez, ya que la competencia entre ambas

especies se verá reducida por la segregación de nicho a la vez que el efecto de la sombra y la

mayor fertilidad del suelo continuarán, incluso en condiciones más áridas (Armas y Pugnaire

2005, Sthultz et al. 2007).

Las especies intolerantes a la sombra, por otro lado, experimentarán mayores tasas de

germinación y supervivencia temprana lejos de plantas adultas a niveles bajos e intermedios

de aridez (p. ej. Veblen 1989, Baskin y Baskin 1998), pero pueden verse beneficiadas por

cierto nivel de sombreo en lugares más áridos (Escudero et al. 2005, Pueyo et al. 2009,

capítulo 2, pero ver Olano et al. 2005). Aunque esto puede depender de otros factores, como

el momento en que emergen estas plántulas (de la Cruz et al. 2008). La relación entre

juveniles de especies intolerantes a la sombra y una planta nodriza será casi siempre negativa,

como se ha visto en numerosos sistemas semiáridos de todo el mundo (Parker y Müller 1982,

Marañón y Bartolomé 1993, Escudero et al. 1999, Forseth et al. 2001, Prider y Facelli 2004,

Seifan et al. 2010, capítulos 1–3). Este efecto negativo se incrementará con el nivel de aridez

(Eliason y Allen 1997, Davis et al. 1999, Espigares et al. 2004, Valladares et al. 2008,

DISCUSIÓN GENERAL

200

capítulos 1 y 3, Soliveres et al. en preparación; Fig. B2), al contrario de lo que predice la

Hipótesis del Gradiente de Estrés (Bertness y Callaway 1994).

Por último, la interacción entre individuos adultos de una especie intolerante a la

sombra y una planta nodriza dependerá en buena medida de si esta especie puede “escapar” de

la competencia por la luz u otros recursos a medida que crece o no. En el primer caso se

puede reducir el efecto negativo de la nodriza de forma gradual a medida que la especie

beneficiaria crece (capítulo 2); mientras que en el segundo, incluso los individuos adultos van

a seguir compitiendo por este recurso, manteniendo así la interacción negativa a lo largo de

toda la vida del individuo (Miriti 2006, Miriti et al. 2007).

En general, la herbivoría aumentará el efecto positivo de las nodrizas en todos los

casos, sin importar las tolerancias relativas de las especies implicadas a otros factores

ambientales. Se ha visto una reducción de las interacciones competitivas cuando los

herbívoros están presentes en numerosos ecosistemas semiáridos de todo el mundo

(McNaughton 1978, Gurevitch et al. 2000, Fowler 2002, Rebollo et al. 2002, Veblen 2008).

De hecho, este cambio en las interacciones debido a la herbivoría puede llegar a compensar

los efectos negativos de la competencia por recursos, generando efectos netos positivos (Graff

et al. 2007). Sin embargo, bajo presiones muy elevadas de herbivoría, como las que se pueden

dar si aumenta la carga de herbívoros con las mismas condiciones climáticas o se reduce la

productividad vegetal con la misma carga de herbívoros (Illius y O´Connor 1999), el efecto

protector de las plantas nodriza frente a la herbivoría puede colapsar, dando como producto

interacciones ligeramente positivas o neutras (Graff et al. 2007, Smit et al. 2007, capítulo 3,

Figs. B1 y B2). No obstante, esto dependerá de la palatabilidad de la planta beneficiaria y sus

vecinas, el grado de tolerancia a la herbivoría de las especies implicadas y sus diferentes fases

ontogenéticas, y el tipo de herbívoro presente en la zona de estudio (revisado en Zamora et al.

2008).

¿EXISTE UN MODELO SENCILLO Y GENERAL QUE PREDIGA LOS RESULTADOS

DE LAS INTERACCIONES ENTRE PARES DE ESPECIES?

Incluso evitando, por simplicidad, incluir en un mismo modelo algunos de los factores clave

que modulan las interacciones planta-planta –como la presencia de factores de estrés

relacionados o no con recursos (Maestre et al. 2009a), la continuidad de rasgos ecológicos que

definen la tolerancia de las especies a dichos factores (Kobe et al. 1995, Ackerly 2003,

Liancourt et al. 2005), o la co-ocurrencia de diversos tipos de estrés a la vez (Riginos et al.

2005, Baumeister y Callaway 2006, Kawai y Tokeshi 2007), por nombrar unos cuantos– el

201

modelo presentado en la Figura B1 ha generado un total de 36 escenarios distintos,

modulados a su vez por el nivel de herbivoría presente en cada sitio. Algo que la mayoría de

ecólogos definirían quizás como un modelo demasiado complejo y poco útil. Este modelo, sin

embargo, es el único que ilustra la gran complejidad existente y los numerosos factores que

afectan al comportamiento de las interacciones planta-planta en los sistemas semiáridos a lo

largo de gradientes climáticos (en este caso de aridez). La gran cantidad de factores que

intervienen en las interacciones planta-planta haría necesario crear modelos de esta

complejidad para cada bioma o sistema de estudio (p. ej. sistemas de alta montaña o

ecosistemas salobres requerirían de modelos diferentes al presentado, pero igualmente

complejos). Además, a la hora de aplicar estos modelos a nivel de comunidad, deberíamos

incluir los efectos derivados de que las especies no están organizadas en pares de individuos

aislados, sino que están organizadas en manchas discretas donde coexisten numerosos

individuos de diversas especies (p. ej. Aguiar y Sala 1999). Esto podría generar fenómenos de

facilitación indirecta (Levine et al. 1999, Cuesta et al. 2010), segregación o

complementariedad de nicho (Hector et al. 1999, Silvertown 2004, Stokes y Archer 2010), o

competencia intransitiva (Laird y Schwamp 2006, 2009), que podrían variar todos los

resultados predichos en los 36 escenarios mostrados en el modelo.

Se sugiere, por tanto, que el único modelo sencillo, que podría resultar útil para

predecir los resultados de las interacciones entre pares de especies e inferir estos resultados a

redes de interacciones más complejas, como las encontradas en las comunidades naturales, y

en todos los sistemas de estudio de forma general, es el modelo basado en el estrés individual,

descrito en el capítulo 4. Este modelo resume todos los posibles resultados de todos los

posibles escenarios prediciendo que las interacciones planta-planta se tornarán más positivas a

medida que la especie beneficiaria se aleja de su óptimo ambiental (ya sean estas condiciones

más o menos áridas, con o sin herbivoría, frío, salinidad, etc.), llegando a un colapso cuando

las condiciones ambientales son tan duras para esta especie en concreto que ni siquiera puede

reclutar bajo las plantas nodriza. Definir el óptimo ambiental de cada especie es una tarea

ardua, pero sin embargo, es la única forma de predecir de forma general el resultado de la

interacción de una especie en concreto con sus vecinas. Para inferir la distancia de una especie

en concreto a su óptimo podría considerarse utilizar el número de individuos reclutados en

áreas libres de vegetación como un indicador de la distancia al óptimo de esa especie (más

individuos reclutados significarán mejores condiciones ambientales para esa especie

[Valiente-Banuet et al. 2006, Valiente-Banuet y Verdú 2007]). También se pueden inferir las

tolerancias relativas a cualesquiera que sean los factores de estrés dominantes en cada área de

DISCUSIÓN GENERAL

202

estudio a partir de las bases de datos y publicaciones disponibles sobre rasgos ecológicos de

las especies implicadas. Por ejemplo, podemos encontrar algunas bases de datos de libre

acceso para especies mediterráneas (i.e. BROT [Paula et al. 2009] o BASECO [Gachet et al.

2005]) o artículos científicos que incluyen rasgos ecológicos de multitud de especies de todo

el mundo (p. ej. Niinemets y Valladares 2006, Poorter et al. 2009, Wright et al. 2004). Un

ejemplo de este tipo de inferencia lo podemos encontrar en Pavoine et al. (2010).

Alternativamente, se podrían utilizar modelos de distribución potencial (p. ej. Guisan y

Zimmermann 2000, Loiselle et al. 2003) de las especies beneficiarias. Según el modelo

basado en el estrés individual, estas especies experimentarán un efecto más positivo de la

presencia de una nodriza a medida que nos alejemos del centro de su área de distribución

potencial.

FACILITACIÓN A NIVEL DE COMUNIDAD: IMPLICACIONES PARA LA ESTRUCTURA Y EVOLUCIÓN DE

LOS ECOSISTEMAS SEMIÁRIDOS

El modelo basado en el estrés individual soluciona algunos de los supuestos más

problemáticos anteriormente mencionados (existencia de un gradiente de estrés general para

todas las especies de una comunidad dada, y de estrategias ecológicas que se mantienen

estables a lo largo de gradientes ambientales amplios) mediante el establecimiento de un

gradiente de estrés único para cada especie. Este gradiente está basado en medidas de la

tolerancia ecológica de cada especie a distintos factores ambientales, cuantitativas y

extrapolables a cualquier sistema de estudio. Así pues, la ausencia de estos supuestos, junto

con la consideración de la diferente naturaleza de los gradientes ambientales que podemos

encontrar en la naturaleza (compuestos por factores de estrés independientes o

correlacionados entre sí) y las diferencias en las tolerancias relativas entre las especies que

coexisten en una comunidad dada, nos permitirá considerar finalmente la evolución de la

importancia de las interacciones planta-planta a nivel de comunidad bajo distintas condiciones

ambientales.

En la Figura B4 se muestra un modelo conceptual basado en una sencilla comunidad

de tres especies (A, B y C). Cada especie difiere en su óptimo ambiental, y por tanto, tendrá

una distribución diferente (basada en su eficacia biológica bajo distintas condiciones) a lo

largo de un gradiente ambiental. Si no consideramos las interacciones positivas entre plantas,

podemos observar un solapamiento mínimo en las distribuciones de estas tres especies, que

vendría dado por: 1) las diferencias en sus tolerancias específicas a los distintos factores de

203

estrés encontrados a lo largo de ese gradiente ambiental (Chapin et al. 1987), y 2) la exclusión

competitiva que ejerce la especie más adaptada a las condiciones locales sobre las demás

(Grime 1973; panel izquierdo de la Fig. B4). En cambio, si consideramos las interacciones

facilitativas, se puede observar como la distribución espacial de cada especie aumenta

(expansión de nicho; Bruno et al. 2003), ya que las especies más adaptadas a las condiciones

locales aumentan las eficacia biológica de las menos adaptadas (panel derecho de la Fig. B4).

Sin embargo, incluso teniendo en cuenta estas interacciones positivas, llega un momento en

que las condiciones ambientales son demasiado duras para una especie dada, y el

reclutamiento de esta especie en concreto es imposible, dando como resultado una eficacia

biológica igual a cero, pese a la mejora microambiental que pudiera ejercer la especie nodriza

(Kitzberger et al. 2000, Ibañez y Schupp 2001; flechas en el panel derecho de la Fig. B4). Sin

embargo, el que el efecto positivo de la facilitación desaparezca para esa especie en concreto

no significa que haga lo propio a nivel de la comunidad entera, ya que habrá otras especies

capaces de reclutar bajo esas condiciones locales, pero lo suficientemente poco adaptadas

como para verse beneficiadas por la presencia de una especie nodriza. Esto explicaría por qué

en diversas comunidades se observa una reducción de la facilitación a escala de mancha, pero

no un efecto significativo de esta reducción en la facilitación sobre la diversidad local

(revisado en Michalet et al. 2006).

DISCUSIÓN GENERAL

204

Figura B4. Modelo teórico donde se predice la eficacia biológica de tres especies con rasgos ecológicos contrastados a lo largo de un gradiente ambiental. En el panel izquierdo se consideran sólo las tolerancias ecofisiológicas de cada especie y la exclusión competitiva. En el panel de la derecha se considera la expansión de nicho promovida por las especies nodriza (especies más adaptadas a las condiciones locales). Bajo condiciones demasiado severas para cada especie, su reclutamiento es imposible, incluso incluyendo la expansión de nicho, y la facilitación colapsa a nivel de esa especie (flechas en el panel). Cuando dejan de existir especies adaptadas a las condiciones locales a lo largo de este gradiente ambiental, la facilitación colapsa a nivel de comunidad (línea superior en el panel derecho). Hasta llegar a ese punto, la importancia de las interacciones positivas a nivel de comunidad permanece estable, ya que la identidad, pero no la cantidad, de especies facilitadas es lo que cambia a lo largo del gradiente.

Cuando el gradiente ambiental esté formado por diversos factores de estrés,

independientes entre sí, la importancia de la facilitación a nivel de la comunidad entera se

mantendrá constante hasta que ya no quede ninguna especie adaptada a las condiciones

locales que pueda facilitar a las demás, lo que llevaría al colapso de la facilitación (Silliman et

al. 2005, Michalet et al. 2006; línea superior del panel izquierdo de la Fig. B4). Casos en que

el gradiente ambiental esté dominado por un solo factor de estrés, o varios correlacionados

entre sí (p. ej. Bertness y Shumway 1993, Callaway et al. 2002), serían simplemente una

fracción del escenario anteriormente descrito. Ambos casos están separados por la línea

discontinua en la Figura B4. Este sencillo modelo conceptual resume el por qué las

interacciones positivas son importantes en numerosos ecosistemas de todo el mundo, y no

sólo en los considerados “estresantes” y el por qué las interacciones positivas serán más

importantes bajo niveles “intermedios de estrés”, se trate del bioma que se trate (revisado en

Holmgren y Scheffer 2010). También explica el colapso de las interacciones positivas bajo

205

elevados niveles de estos factores de estrés a lo largo de gradientes ambientales, ya sea para

especies concretas (Kitzberger et al. 2000, Ibañez y Schupp 2001), o en comunidades enteras

(Silliman et al. 2005, Michalet et al. 2006). La Figura B4 ayuda a entender el papel de la

facilitación en la evolución y resiliencia de las comunidades naturales. Estas interacciones

positivas asegurarán el nicho de regeneración de las especies menos adaptadas a las

condiciones locales (Valiente-Banuet et al. 2006), sean estas las que sean, o el mantenimiento

de la diversidad y el funcionamiento ecosistémico en condiciones ambientales menos

productivas (Mulder et al. 2001, Kikvidze et al. 2005, Badano y Cavieres 2006), como las

predichas con el cambio climático para la cuenca Mediterránea (Brooker 2006). Esto ocurrirá

siempre y cuando queden especies que presenten tolerancias ecofisiológicas a las condiciones

ambientales existentes y puedan suavizar estas condiciones para las especies menos

adaptadas.

Ahora bien, al igual que la mayoría de artículos relacionados con las interacciones

entre plantas, el modelo propuesto en la Figura B4 no considera el hecho de que las especies

de una comunidad no se relacionan par a par, si no que forman ensamblajes más o menos

complejos que atañen a todas las especies que coexisten en esa comunidad (Maestre et al.

2010). Un ejemplo de ello serían las manchas de vegetación de las tierras secas, donde

coexisten un elevado número de especies, y donde es probable que se den interacciones

positivas y negativas entre multitud de ellas, dando lugar a complejas redes de interacción

(Verdú et al. 2010). Como se ha mencionado antes, esto puede dar lugar a procesos de

facilitación indirecta, competencia intransitiva, segregación o complementariedad de nicho

(Grace et al. 1993, Hector et al. 1999, Levine et al. 1999, Silvertown 2004). El incremento de

la diversidad mediante la expansión de nicho, junto con el aumento en la heterogeneidad en

los recursos por el que estas especies compiten, promovidos ambos por la presencia de

especies nodriza, puede dar lugar a aumentos desproporcionados en la diversidad local

(Bowker et al. 2010). Aunque estos efectos ya se conocían y han sido estudiados por separado

(revisado en Brooker et al. 2008), hasta ahora ningún modelo teórico ha incluido ambos

mecanismos a la hora de predecir el papel de la facilitación en la diversidad de las

comunidades naturales (Callaway 2007). Se pretende dar ese paso con el modelo mostrado en

la Figura B5. Este modelo se basa en la distribución que podemos encontrar para cualquier

especie a lo largo de un gradiente espacial o temporal en cualquier libro de ecología, esto es,

una campana de Gauss más o menos apuntada. Como hemos explicado antes, los rasgos

ecológicos de cada especie, y sus tolerancias relativas a los diferentes factores de estrés

presentes a lo largo de ese gradiente, definirán su óptimo ambiental, que variará según la

DISCUSIÓN GENERAL

206

especie. En este modelo, se utiliza un conjunto inicial de siete especies (de la A a la G). El

nicho potencial de estas especies (panel superior izquierdo de todos los escenarios de la Fig.

B5) vendrá indicado por los diversos filtros de establecimiento, esto es, llegada de propágulos

(dispersión) y tolerancia a las condiciones locales. Una vez llegados a este punto es donde las

interacciones bióticas adquieren mayor importancia (p. ej. Huston 1999, Rajaniemi et al.

2006, pero ver Mitchell et al. 2009, Gotelli et al. 2010).

207

Figura B5. Modelo teórico donde se predicen los nichos potenciales y ocupados de distintas especies a lo largo de un gradiente ambiental, y como esto se refleja en la riqueza local de especies de una comunidad dada a partir de un conjunto inicial de siete especies (A–G). Las predicciones se realizan en cuatro escenarios distintos teniendo en cuenta: A) homogeneidad en los recursos por los que compiten las plantas, la cual hace imposible la intransitividad en la competencia o la segregación de nicho, B) heterogeneidad en los recursos, haciendo posible cierto grado de segregación de nicho o competencia intransitiva entre las especies existentes, C) expansión de nicho de las especies menos adaptadas a las condiciones locales promovida por especies nodriza, lo cual genera la llegada de nuevas especies que inicialmente no estaban en la comunidad, pero condiciones homogéneas en los recursos por los que estas especies compiten, y D) el doble efecto de las plantas nodriza (incrementar la heterogeneidad de nutrientes y aumentar el conjunto de especies mediante expansión de nicho) en un sistema ya de por sí heterogéneo. En los escenarios A y C se han sombreado los nichos ocupados por cada especie, para facilitar su comprensión. En los escenarios B y D esto no ha sido posible por el alto grado de solapamiento entre nichos que se predice. Nótese como la relación entre la riqueza de especies y la productividad (panel de abajo a la derecha en cada escenario) va cambiando según consideremos unos procesos u otros.

DISCUSIÓN GENERAL

208

En un primer escenario suponemos una disposición homogénea de los recursos por los

que compiten estas siete especies (escenario A en la Fig. B5). En estas condiciones, es

imposible que exista segregación de nicho o intransitividad en la competencia (Huston 1979,

1999). También obviamos aquí el efecto facilitador que pueden ejercer unas especies sobre

otras. Por tanto, la exclusión competitiva es la única interacción planta-planta que nos queda.

En unas condiciones más o menos homogéneas, las especies más adaptadas a las condiciones

locales (competitivas, sensu Grime 1979) pueden desplazar a las menos adaptadas, haciendo

que su nicho real sea menor que el potencial. Poniendo como ejemplo las especies A, B y C,

cumpliéndose que A>B>C tenemos que A excluirá a B y C en los lugares donde esté presente,

reduciendo el nicho real de B y C. Y B hará lo mismo con C (áreas sombreadas del escenario

A).

En el escenario B se considera el efecto de la heterogeneidad ambiental, esto es, la

variabilidad espacial o temporal en los recursos por los que compiten las especies presentes en

una comunidad. Esto permitirá 1) que no haya una jerarquía marcada en la competencia

(competencia intransitiva), si no que, dependiendo de las condiciones microambientales, la

especie A pueda desplazar a las demás o pueda ser desplazada por otras (Gilpin 1975, Grace

et al. 1993), y 2) que se pueda dar segregación de nicho, esto es, los nichos reales de

diferentes especies pueden mantenerse aunque éstas compitan por el mismo recurso, ya que

difieren en la toma de este recurso en el espacio o en el tiempo (Silvertown 2004). Ambos

procesos son importantes, ya que pueden aumentar la diversidad a nivel local (Grace 1993,

Huston 1999, Laird y Schwamp 2006; ver Tilman 1994 para una aproximación alternativa).

En el escenario B de la Figura B5 podemos ver, por ejemplo, como A no excluye

competitivamente a B de su nicho potencial (aunque reduce su nicho real considerablemente)

pese a su superioridad competitiva. Esto ocurre debido al efecto indirecto de D, una especie

que supera competitivamente a A. Por tanto, A, B y D coexisten (competencia intransitiva:

A>B>D>A, que genera facilitación indirecta de D sobre B). Alternativamente, G y F pueden

coexistir a pesar de que sus nichos potenciales se solapan. Esto se debe a que ambas especies

no compiten exactamente entre ellas pese a necesitar los mismos recursos, ya que los toman

en lugares o momentos diferentes (segregación de nicho, p. ej. Sala et al. 1989, capítulo 1). En

sistemas heterogéneos y relativamente ricos en especies, estas especies pueden

complementarse, de forma que toman los recursos disponibles de forma más eficiente,

incrementando la productividad a nivel del ecosistema (Hector et al. 1999). Aunque esta

complementariedad de nicho es clave para entender la relación entre diversidad y

productividad, por lo que se ha incluido en el modelo, no afecta a la riqueza de especies, si no

209

a la relación de la riqueza con la productividad total del sistema, por lo que no se discutirá

ahora el papel de este proceso.

El escenario C, por otro lado, es exactamente idéntico al A, salvo que se ha incluido la

expansión de nicho (sensu Bruno et al. 2003). Hasta el momento, este es el único efecto de las

interacciones positivas que se había incluido en los modelos teóricos sobre el efecto de las

interacciones planta-planta en la riqueza local de especies (Hacker y Gaines 1997, Bruno et al.

2003, Lortie et al. 2004a, Michalet et al. 2006). Como predicen todos estos modelos, y se ha

matizado en esta discusión, las especies más adaptadas a las condiciones locales

incrementarán la eficacia biológica de las especies menos adaptadas, e incluso pueden

permitir la colonización de otras especies, que no habían podido cruzar los filtros ambientales

por ellas mismas (H e I, en el escenario C). Al haber ahora más especies en la comunidad, hay

mayores posibilidades de que se de segregación o complementariedad de nicho. Sin embargo,

este escenario sigue asumiendo unas condiciones homogéneas, que permiten la existencia de

especies “competitivas” que excluyan a otras especies de sus nichos potenciales. Esto daría

lugar al incremento de la riqueza, gracias a la expansión de nicho, predicho en condiciones

más improductivas debido a que las especies “tolerantes al estrés” facilitan a las especies

“competitivas” en condiciones improductivas (Hacker y Gaines 1997, Travis et al. 2005,

Michalet et al. 2006; especie I en el escenario C). Aunque también puede incrementar la

diversidad en situaciones mucho más productivas, donde la especie más adaptada a las

condiciones locales (A en este caso), facilita la entrada de una especie menos adaptada (H).

Este fenómeno ha sido observado en sistemas considerados muy productivos, como el bosque

lluvioso tropical o bosques de ribera, entre otros (revisado en Holmgren y Scheffer 2010).

Finalmente, en el escenario D se considera un medio de por sí heterogéneo, como la

mayoría de medios naturales, donde además se incluye el doble efecto positivo que ejercen las

especies nodrizas sobre la diversidad local en una comunidad dada. Estos efectos son: 1)

incrementar el conjunto de especies disponibles por medio de la expansión de nicho (Bruno et

al. 2003, Lortie et al. 2004a, Badano y Cavieres 2006), y 2) aumentar la heterogeneidad

ambiental bajo su dosel, lo que incrementa aún más la heterogeneidad en los recursos por los

que estas especies compiten (p. ej. Pugnaire et al. 1996a, Cuesta et al. 2010). Este escenario

no había sido considerado hasta el momento para evaluar el efecto de la facilitación sobre la

diversidad local (Callaway 2007). Sin embargo, es en este escenario donde las posibilidades

de que se de competencia intransitiva y la segregación o complementariedad de nicho son

máximas, lo que llevaría a un aumento de la diversidad desproporcionado provocado por

procesos de retroalimentación entre ambos procesos (más especies disponibles y más

DISCUSIÓN GENERAL

210

heterogeneidad en la competencia generan competencia intransitiva y/o segregación de nicho,

que a su vez permiten que más especies puedan coexistir). Este proceso de retroalimentación

positiva aumenta la diversidad local gracias a dos efectos fundamentales. Primero, porque

aumenta el nicho potencial de la mayoría de las especies del conjunto, y segundo, porque se

reducen las posibilidades de exclusión competitiva. Pocas especies son excluidas de sus

nichos potenciales ya que, o bien difieren en el nicho de las especies con las que coexisten (H,

D y A, en el escenario D), o bien, terceras especies evitan que estas sean excluidas (D permite

que B no sea excluida por A). Este proceso de retroalimentación positiva, que lleva a la

reducción de la exclusión competitiva para muchas especies, podría ser la causa del

incremento de diversidad a lo largo de la sucesión en algunos ecosistemas (Johnston y Odum

1956, McKindsey y Bourget 2001), o de los resultados contrastados en la relación entre

diversidad e invasibilidad de los ecosistemas (revisado en Bruno et al. 2003). Sistemas más

diversos podrán atraer nuevas especies (sean estas invasoras o no) si las características del

sistema de estudio (i.e. heterogeneidad en los recursos por los que compiten las especies)

permiten que se den estos procesos de retroalimentación, pero evitarán la entrada de nuevas

especies si estos procesos de retroalimentación no se dan y domina la exclusión competitiva.

La coexistencia o no de todos estos factores (expansión, segregación y

complementariedad de nicho, y competencia intransitiva) puede modificar de forma sustancial

la relación entre la riqueza de especies y la productividad del ecosistema, incluso con las

mismas especies, estudiadas a la misma escala y con la misma amplitud del gradiente

estudiado (panel inferior derecho de todos los escenarios; ver Whitakker 2010, Mittlebach

2010 y referencias en estos textos para una discusión detallada sobre los efectos de la escala

de estudio en la relación riqueza-productividad). Esto podría ayudar a explicar la gran

cantidad de excepciones que se han encontrado a esta relación unimodal entre riqueza y

diversidad en numerosos ecosistemas de todo el mundo a nivel local (p. ej. Grace 1999,

Waide et al. 1999, Gillman y Wright 2006). El modelo propuesto contradice no sólo la idea de

que la facilitación es sólo importante en los sistemas más improductivos y estresantes (p. ej.

Bertness y Callaway 1994), si no también que la facilitación sólo afecte a esta relación entre

diversidad y productividad a niveles medios-bajos de ésta última (Michalet et al. 2006). Se

propone en esta discusión que evaluar el papel relativo de la expansión, complementariedad y

segregación de nicho, junto con el grado de intransitividad de la competencia nos permitirá,

por un lado, evaluar de una forma completa el papel de las interacciones planta-planta en la

diversidad local (Brooker et al. 2008), y por otro, definir la relación entre esta diversidad local

y la productividad y el funcionamiento del ecosistema (Mulder et al. 2001, Callaway 2007,

211

Maestre et al. 2010). Ambas cuestiones no son sólo atractivas desde el punto de vista teórico,

ya que nos permiten arrojar algo de luz sobre los mecanismos implicados en la relación entre

la diversidad y la productividad de los ecosistemas (Whitakker 2010), sino que también son

cruciales para finalmente entender el papel de la biodiversidad en el funcionamiento de los

ecosistemas y los servicios que éstos proveen (Hooper et al. 2005, Michalet et al. 2006) y, por

tanto, en como entendemos y manejamos la naturaleza. Sin embargo, evaluar el papel de cada

uno de los procesos mencionados a nivel de comunidad requeriría experimentos muy costosos

y logísticamente inabordables, por lo que se ha sugerido el uso de estudios observacionales

para evaluar la importancia relativa de las interacciones bióticas a nivel de comunidad (p. ej.

Gotelli y Graves 1996, Brooker et al. 2008, Pausas y Verdú 2010). A pesar de que las

herramientas existentes han permitido un mejor entendimiento de los mecanismos de

ensamblaje de especies y el papel de las interacciones bióticas en este proceso (p. ej. Gotelli y

Graves 1996, Dullinger et al. 2007, Maestre et al. 2008, 2010, Rooney 2008, Bowker et al.

2010, Gotelli et al 2010), éstas son insuficientes por el momento para discernir entre los

procesos anteriormente mencionados. Aunque queda mucho por hacer en este aspecto, se

ofrecen en los trabajos desarrollados en esta tesis doctoral (capítulos 4 y 5) una serie de

herramientas para diferenciar entre los procesos de competencia intransitiva, expansión y

segregación de nicho, y su efecto sobre la riqueza de especies y el patrón filogenético local, a

partir de estudios observacionales. El refinamiento de estas técnicas ayudará a entender

finalmente el papel de las interacciones bióticas no tróficas en el ensamblaje de las especies

que forman las comunidades naturales, y la relación entre la diversidad de estas especies y la

productividad de los ecosistemas.

IMPLICACIONES DE LA FACILITACIÓN EN LA RESTAURACIÓN DE LOS SISTEMAS SEMIÁRIDOS

El papel que la facilitación puede jugar en la restauración de los ecosistemas Mediterráneos

en general, y los semiáridos en particular, ya ha sido discutido ampliamente con anterioridad

(Pugnaire et al. 1996, Maestre et al. 2001, Castro et al. 2002, Gómez-Aparicio et al. 2004,

Cortina y Maestre 2005, Padilla y Pugnaire 2006, Valladares y Gianoli 2007, Pueyo et al.

2009, Cortina et al. 2010). Sin embargo, en la inmensa mayoría de estos trabajos se trata el

papel de las interacciones facilitativas en un contexto puramente climático, asumiendo que la

importancia de estas interacciones para el éxito de la restauración aumentará con el nivel de

aridez. Muy pocos trabajos han tenido en cuenta el colapso de estas interacciones facilitativas

bajo niveles extremadamente altos de aridez o herbivoría (Valladares y Gianoli 2007, Pueyo

DISCUSIÓN GENERAL

212

et al 2009, Cortina et al. 2010), o la importancia relativa de las interacciones facilitativas

frente a otros múltiples procesos y herramientas importantes para la restauración (Méndez et

al. 2008, Pueyo et al. 2009, Cortina et al. 2010). Se discute en esta sección la idoneidad de

utilizar las interacciones facilitativas en restauración en sistemas semiáridos, con respecto a

otras técnicas disponibles. También se discute como puede adaptarse el manejo de los

ecosistemas semiáridos a las futuras condiciones ambientales que se esperan con el cambio

climático, y el papel que la facilitación puede jugar en dicho proceso.

El funcionamiento ecosistémico en la mayoría de ambientes semiáridos viene dado por

la dinámica fuente-sumidero, que permite a las manchas de vegetación retener el agua de

escorrentía, recursos y semillas procedentes de las áreas de suelo desnudo. Esto favorece la

captura y reciclaje de nutrientes y aumenta la colonización de nuevas especies en las manchas

de vegetación (Ludwig y Tongway 1995, Aguiar y Sala 1999, Puigdefábregas et al. 1999).

Como ya se ha comentado en la introducción general, la degradación de estos sistemas viene

dada por la pérdida de esta estructura de la vegetación (reducción del tamaño y

distanciamiento entre los parches [Maestre et al. 2006, Kefi et al. 2007], o reducción de la

cobertura [Maestre y Escudero 2009]). La degradación de los sistemas semiáridos no es un

proceso lineal, sino que experimenta distintos umbrales de degradación correspondientes con

cambios drásticos en la estructura, composición y funcionamiento ecosistémicos (Van de

Koppel et al. 1997; ver Fig. B6, modificado de Cortina et al. 2010). Aunque se ha visto que el

funcionamiento ecosistémico no está necesariamente unido a su “restaurabilidad” (Cortina et

al. 2005, Maestre et al. 2006), conocer en qué estado funcional se haya el ecosistema es

básico para: 1) evaluar las medidas más adecuadas para su restauración, y 2) definir áreas

prioritarias para la conservación o restauración (Cortina et al. 2005, 2010, Méndez et al.

2008). Así por ejemplo, en el caso de los espartales de Stipa tenacissima, aunque dependerá

también de objetivos sociales, políticos y económicos, dos tipos de áreas prioritarias para la

restauración de las estepas de Stipa tenacissima serían aquellas con niveles de funcionalidad

intermedios (“estepas” y “estepas empobrecidas” sensu Cortina et al. 2010). En el caso de las

estepas empobrecidas, para asegurarnos que la pérdida en la cobertura vegetal y la diversidad

no alcanza estados de degradación severos (Maestre y Escudero 2009). En el caso de las

estepas, en cambio, la prioridad se debería a que podemos ganar bastante en funcionalidad y

servicios ecosistémicos con relativamente poca inversión.

213

Figura B6. Esquema donde se resumen los cuatro posibles estados de degradación de una estepa de Stipa tenacissima: estepa con matorral (A), estepa sin matorral (B), estepa degradada o empobrecida (C) y área desertificada o “badland” (D). Se indican de forma esquemática las actuaciones prioritarias en cada etapa. La introducción de especies leñosas, prestando atención a la procedencia y calidad del material utilizado, es recomendable en todas las etapas, aunque el grosor de la flecha indica en cada caso el grado de prioridad de esta práctica. Se señalan las áreas prioritarias para la restauración con un círculo. EI = estructuras inertes para la captura y retención de escorrentía; ME = mallas de exclusión de herbívoros.

Hemos de tener en cuenta que el principal cuello de botella para el establecimiento de

las plantas en medios semiáridos es el agua disponible (Eldridge et al. 1991, Whitford 2002,

Holmgren et al. 2006). Por tanto, sea cual sea la herramienta que se utilice y el estado de

degradación de la zona donde se actúe, un año seco dará un rendimiento menor del esfuerzo

DISCUSIÓN GENERAL

214

invertido (Navarro et al. 2006 y referencias en dicho texto). Por tanto, predecir de alguna

manera las condiciones climáticas venideras nos daría una información crucial sobre cuando

actuar para maximizar el éxito de los esfuerzos de restauración (Cortina et al. 2010). La gran

importancia de la alternancia de los eventos climáticos globales (El Niño/La Niña) para el

patrón de precipitaciones de numerosas regiones de todo el mundo, incluida España (ver

Holmgren et al. 2006), nos ofrece esta herramienta de predicción (Valladares y Gianoli 2007).

Así, la alternancia entre las fases del Niño y la Niña nos ofrece ventanas temporales de vital

importancia para predecir estos años lluviosos y centrar en ellos los proyectos de restauración.

Si bien la alternancia de estos fenómenos no es predecible al 100%, presenta un tiempo de

recurrencia aproximado de entre tres y seis años (Holmgren et al. 2006 y referencias en ese

texto). Por tanto, mientras no se desarrollen mejores herramientas de predicción, cuando en un

año en concreto se de la fase del Niño (condiciones más lluviosas para España), podemos

inferir que los dos siguientes años serán igualmente benignos. Los esfuerzos destinados a la

restauración de los sistemas semiáridos deberían centrarse exclusivamente en estos años,

probablemente más lluviosos que la media, donde el éxito de los proyectos de restauración

seguramente será mayor, sean cuales sean las herramientas que apliquemos.

Una vez establecido el cuándo y el dónde es mejor actuar, podemos centrarnos en

cómo es mejor hacerlo. Se utiliza aquí como base el diagrama modificado de Cortina et al.

(2010) para ilustrar el papel relativo de las interacciones planta-planta frente a otros procesos

y herramientas útiles para asegurar el éxito de la restauración en los espartales de Stipa

tenacissima (Fig. B6). En este diagrama se establecen cuatro estados alternativos en la

degradación del funcionamiento ecosistémico en el caso de estos espartales que varían desde

el suelo desnudo a la estepa con vegetación arbustiva. En las etapas menos degradadas (estepa

con vegetación arbustiva, estepa y estepa empobrecida) interesaría incrementar la cobertura

de vegetación arbustiva, lo cual se ha demostrado muy positivo para el funcionamiento y la

diversidad de los espartales ibéricos (Maestre y Cortina 2005, Cortina y Maestre 2005,

Maestre et al. 2009b). Para ello, en el caso de la estepa con vegetación arbustiva (no en el

resto de etapas de degradación), podemos confiar en los procesos de nucleación que

comúnmente suceden en sistemas Mediterráneos, donde los animales dispersan de forma

natural propágulos de numerosas especies bajo los arbustos o árboles remanentes (efecto

percha), que a su vez se establecerán con más éxito bajo estos arbustos debido a las

condiciones de sombreo y mayor fertilidad del suelo (Verdú y García-Fayos 1996, Pausas et

al. 2006, Méndez et al. 2008). Si bien este proceso natural puede incrementar la diversidad y

la cobertura de arbustos rebrotadores de forma natural en estos espartales y en numerosos

215

ecosistemas Mediterráneos (restauración pasiva), es poco probable que se aumente de forma

significativa la colonización de estas especies en lugares donde no haya arbustos o árboles

remanentes que atraigan a los animales dispersores. En otras palabras, la colonización natural

sólo se dará en lugares donde previamente ya haya algún arbusto presente (los animales

depositarán las semillas en esos lugares), no creándose nuevos parches de vegetación

arbustiva en otros lugares (Cortina y Maestre 2005). Por tanto, incrementar la tasa de

colonización y el número de parches formados por estos arbustos es deseable, especialmente

en las estepas donde éstos no estén presentes (Cortina y Maestre 2005, Maestre y Cortina

2005). Para ello, utilizar las interacciones positivas del esparto con estas especies ha sido

recomendado (p. ej. Maestre et al. 2001, Gasque y García-Fayos 2004, Barberá et al. 2006,

Navarro et al. 2008). Sin embargo, hay que considerar el hecho de que estas interacciones

pueden colapsar bajo niveles extremos de aridez o herbivoría (Maestre y Cortina 2004a,

capítulo 3). Bajo estas condiciones, el establecimiento de pilas de ramas u otras estructuras

inertes (p. ej. microcuencas) que provean mejores condiciones hídricas y de fertilidad del

suelo, o bien que reduzcan los efectos de la herbivoría (p. ej. mallas protectoras), para los

plantones introducidos o para la colonización natural son recomendables (Tongway y Ludwig

1996, Ludwig y Tongway 1996, Holmgren et al. 2006, Soliveres et al. 2008, Pueyo et al.

2009).

El uso de estructuras inertes destinadas a aumentar la retención del agua de escorrentía

y otros recursos, también es recomendable en lugares muy degradados, independientemente

de su nivel de aridez, debido a la ausencia de plantas que puedan ejercer como nodriza (estado

más degradado en la Fig. B6). También en estas zonas más degradadas, donde con seguridad

se ha perdido parte, o todo el horizonte orgánico del suelo, puede ser conveniente incrementar

la fertilidad del suelo mediante el uso de enmiendas orgánicas. Si bien esto dependerá de las

características físico-químicas del sitio a restaurar y de las condiciones climáticas (Soliveres

et al. en prensa y referencias en ese texto). Otra alternativa para incrementar la cobertura

vegetal, especialmente en los sitios más áridos o degradados, es la introducción de especies

más heliófilas (p. ej. Retama sphaerocarpa; Moro et al. 1997, Caravaca et al 2003). Los

capítulos 1 y 3 de esta tesis doctoral han demostrado que el uso de las interacciones

facilitativas tampoco sería recomendable para la introducción de estas especies, siendo

preferible introducir los plantones en lugares libres de vegetación. Las interacciones

facilitativas, por tanto, aunque son una herramienta de restauración muy útil en determinados

escenarios de restauración, no son recomendables en condiciones de aridez extrema o

herbivoría muy intensa. En estas situaciones, otras medidas como las anteriormente

DISCUSIÓN GENERAL

216

explicadas (y las discutidas más abajo) pueden dar mejores resultados en los proyectos de

restauración.

En algunas ocasiones, especialmente en los futuros escenarios predichos con el cambio

climático, las condiciones ambientales pueden ser extremadamente adversas para la gran

mayoría de especies de interés, de forma que su reclutamiento sea imposible, incluso bajo el

dosel de una especie nodriza o con la presencia de las estructuras inertes mencionadas con

anterioridad. En este caso, una cuidadosa selección de las poblaciones de donde se obtiene el

material para la restauración, y el establecimiento de una calidad de planta adecuada para las

condiciones ambientales puede mejorar el rendimiento de algunas especies (Cortina et al.

2006, Villar et al. en prensa). Aunque esta selección y preparación del material es importante

en cualquier proyecto de restauración, cobra especial importancia en sistemas particularmente

áridos, donde la procedencia o la calidad de la planta pueden decidir el destino de las

plántulas introducidas (Navarro et al. 2006, Trubat et al. 2008). La aplicación conjunta de la

selección del material vegetal adecuado (especies y procedencias), en conjunto con una buena

calidad de planta y el uso de estructuras inertes, puede llevarnos al éxito en la restauración

bajo estas condiciones más áridas o degradadas (p. ej. proyecto FUNDIVFOR en el sureste

ibérico, http://80.24.165.149/fundivfor/).

A pesar de que se ha sugerido que los arbustos pueden ser la mejor opción posible

como planta nodriza en ecosistemas Mediterráneos en las actuales condiciones climáticas

(Gómez-Aparicio et al. 2004, Gómez-Aparicio 2009), se debería conceder un mayor crédito

como especies nodriza a las herbáceas perennes como Stipa tenacissima en el futuro. Estas

herbáceas es probable que ejerzan efectos más positivos sobre las especies leñosas de interés

en las futuras condiciones predichas con el cambio climático, ya que el incremento en la

frecuencia de eventos de lluvia torrenciales puede incrementar la segregación de nicho entre

S. tenacissima y sus arbustos vecinos, pero no así en el caso de que la nodriza sea otro

arbusto.

217

CONCLUSIONES GENERALES

De los trabajos desarrollados en la presente tesis doctoral se extraen las siguientes conclusiones generales: 1) La tolerancia a la sombra es un factor clave a la hora de entender las interacciones planta-planta en medios semiáridos. La sombra generada por Stipa tenacissima ejerció un efecto negativo sobre Retama sphaerocarpa y Lepidium subulatum, ambas especies intolerantes a la sombra. 2) Los efectos de los cambios en el patrón de las precipitaciones predichos con el cambio climático varían dependiendo de las especies implicadas. En general, la reducción en las lluvias de primavera aumentó los efectos negativos sobre las especies intolerantes a la sombra. 3) El efecto competitivo de las especies herbáceas sobre las leñosas se reduce con la edad de éstas últimas, lo que sugiere la existencia de procesos de segregación de nicho. El aumento de los eventos de lluvia torrencial, predicho con el cambio climático, puede acelerar la segregación de nicho entre especies leñosas y herbáceas. 4) Los factores climáticos que influyen en las interacciones planta-planta a lo largo de su ontogenia presentan una heterogeneidad espacial marcada. Mientras que la lluvia fue el modulador fundamental de la interacción entre Lepidium y Stipa en la ladera de solana, otros factores climáticos fueron importantes en la ladera de umbría. 5) La herbivoría es un factor clave en las interacciones planta-planta, que puede llegar a compensar los efectos negativos derivados de la competencia por nutrientes (i.e. luz o agua). La reducción del impacto de la herbivoría promovida por el efecto protector de Stipa sobre Retama redujo la competencia entre ambas especies, dando como resultado un efecto neto positivo de Stipa sobre Retama. 6) La herbivoría y el estrés hídrico presentan interacciones complejas, existiendo una jerarquía entre ambos factores. La aridez por sí misma no tuvo un efecto importante en la interacción entre Stipa y Retama bajo niveles muy altos de herbivoría. Sin embargo, la aridez moduló indirectamente la presión de herbivoría afectando a la productividad vegetal, y por tanto, a la cantidad de forraje disponible para los herbívoros y a su impacto sobre la especie estudiada. 7) Existe una jerarquía entre las condiciones climáticas y la distancia filogenética entre los arbustos rebrotadores y sus vecinas a la hora de definir la interacción entre ambas. Distancias filogenéticas entre 207 y 273 millones de años produjeron siempre interacciones competitivas. Mientras que valores fuera de este rango dieron lugar a interacciones neutras o facilitativas, dependiendo de las condiciones climáticas. Ni la distancia filogenética ni el clima fueron factores clave entre la interacción entre Stipa y sus especies vecinas.

CONCLUSIONES

218

8) El efecto positivo de Stipa sobre Lepidium durante la germinación de éste último fue clave a la hora de definir el resultado neto de la interacción entre ambas especies. Esto quedó demostrado por el elevado grado de co-ocurrencia encontrado a pesar de la dominancia de interacciones negativas entre ambas especies a lo largo de la vida de Lepidium. 9) Las condiciones climáticas, interpretadas como un gradiente general de estrés, son predictores muy pobres del resultado de las interacciones entre pares de especies, ya que estas interacciones dependen de multitud de factores además del clima. Por tanto, ninguno de los modelos conceptuales dominantes en la actualidad predice una suficiente cantidad de estas interacciones como para ser aceptado de forma universal. 10) Las plantas nodriza aumentan la riqueza de especies y la diversidad filogenética a nivel local no sólo por la expansión de nicho, promovida por la mejora de las condiciones microclimáticas bajo su copa, sino también por un aumento de la segregación de nicho entre sus especies beneficiarias, derivado de una mayor heterogeneidad ambiental bajo su dosel. 11) La evolución del efecto positivo que las plantas nodriza ejercen sobre la riqueza local de especies a lo largo de gradientes ambientales depende de la naturaleza de dichos gradientes. Mientras que en España se mantuvo constante a lo largo de todo el gradiente estudiado (definido de forma opuesta por la aridez y las bajas temperaturas), este efecto aumentó con la aridez en Australia, donde la lluvia era el factor de estrés dominante en las comunidades estudiadas. En este último caso, se detectó un colapso del efecto positivo de las nodrizas bajo niveles extremos de aridez, como apuntan algunos modelos teóricos existentes. 12) Las condiciones climáticas y las interacciones bióticas interactúan a la hora de definir el patrón filogenético de las comunidades vegetales. Mientras que la facilitación aumentó la dispersión del patrón filogenético, condiciones climáticas “más benignas” (p. ej. más lluvia) incrementaron la agregación de este patrón mediante su efecto indirecto en la reducción de la diferenciación de nicho entre especies facilitadas y no facilitadas. El efecto conjunto de ambos factores generó un patrón filogenético aleatorio en la mayoría de comunidades estudiadas. Estos resultados advierten sobre la problemática derivada de inferir los mecanismos dominantes en el ensamblaje de una comunidad dada a partir únicamente de su patrón filogenético, aunque tengamos un conocimiento detallado del grado de conservación de importantes rasgos ecofisiológicos en las especies de esa comunidad a lo largo de la evolución.

BIBLIOGRAFÍA Y AFILIACIÓN DE LOS CO AUTORES

BIBLIOGRAFÍA

220

BIBLIOGRAFÍA

Ackerly, DD. 2003. Community assembly,

niche conservatism and adaptive evolution in changing environments. International Journal of Plant Sciences 164: S165–S184.

Acuña-Rodríguez I, LA Cavieres, and E Gianoli. 2006. Nurse effect in seedling establishment: facilitation and tolerance to damage in the Andes of central Chile. Revista Chilena de Historia Natural 79: 329–336.

Aguiar MR, and OE Sala. 1999 Patch structure, dynamics and implications for the functioning of arid ecosystems. Trends in Ecology and Evolution 14: 273–277.

Allan GJ, J Francisco-Ortega, A Santos-Guerra, E Boerner, and A Zimmer. 2004. Molecular phylogenetic evidence for the geographic origin and classification of Canary Islands Lotus (Fabaceae: Loteae). Molecular Phylogenetics and Evolution 32: 123–138.

Allan GJ, and JM Porter. 2000. Tribal delimitation and phylogenetic relationships of Loteae and Coronilleae (Faboideae: Fabaceae) with special reference to Lotus: evidence from nuclear ribosomal ITS sequences. American Journal of Botany 87: 1871-1881.

Anderson CL, K Bremer, and EM Friis. 2005. Dating phylogenetically basal eudicots using rbcL sequences and multiple fossil reference points. American Journal of Botany 92: 1737-1748.

Anderson MJ, RN Gorley, and KR Clarke. 2008. PERMANOVA+ for PRIMER: Guida to software and statistical methods. PRIMER-E, Plymouth, UK.

Andrés, P, and M Jorba. 2000. Mitigation strategies in some motorway embankments (Catalonia, Spain). Restoration Ecology 8: 268-275.

Angiosperm Phylogeny Group. 2009. An update of the Angiosperm Phylogeny Group classification for the orders and families of flowering plants: APG III. Botany Journal of the Linnean Society 161: 105–121.

Anthelme F, and R Michalet. 2009. Grass-to-tree facilitation in an arid grazed environment (Aïr Mountains, Sahara). Basic and Applied Ecology 10: 437-446.

Anthelme F, R Michalet, and M Saadou. 2007. Positive associations involving the tussock grass Panicum turgidum Forssk. In the Aïr-Ténéré Reserve, Niger. Journal of Arid Environments 68: 348–362.

Aragón CF, A Escudero, F. Valladares. 2008. Stress-induced dynamic adjustments of reproduction differentially affect fitness components of a semiarid plant. Journal of Ecology 96: 222–229.

Archer SR. 1995. The role of herbivores in mediating grass-woody plant interactions. Tropical Grasslands 29:218–235.

Archer SR, TW Boutton, and KA Hibbard. 2001. Trees in grasslands. In: Biogeochemical consequences of woody plant expansion in global biogeochemical cycles in the climate system. ED Schulze, SP Harrison, M Heimann, EA Holland, J Lloyd, IC Prentice, and D Schimel (Eds.). Academic press, San diego.

Armas C, R Ordiales, and FI Pugnaire. 2004. Measuring plant interactions: A new comparative index. Ecology 85: 2682–2686.

Armas C, and FI Pugnaire. 2005. Plant interactions govern population dynamics in a semi-arid plant community. Journal of Ecology 93: 978–989.

Armas C, and FI Pugnaire. 2009. Ontogenetic shifts in interactions of two dominant shrub species in a semi-arid coastal sand dune system. Journal of Vegetation Science 20: 535–546.

Aragón CF, A Escudero, and F Valladares . 2008. Stress-induced dynamic adjustments of reproduction differentially affect fitness components of a semiarid plant. Journal of Ecology 96: 222–229.

Arsenault R, and N Owen-Smith. 2002. Facilitation versus competition in grazing herbivore assemblages. Oikos 97: 313–318.

Badano EI, and LA Cavieres. 2006. Ecosystem engineering across ecosystems: do engineer species sharing common features have generalized or idiosyncratic effects on species diversity? Journal of Biogeography 33: 304–313.

Balaguer L, FI Pugnaire, E Martínez-Ferri, C Armas, F Valladares, and E Manrique . 2002. Ecophysiological significance of chlorophyll loss and reduced

BIBLIOGRAFÍA

222

photochemical efficiency under extreme aridity in Stipa tenacissima L. Plant and Soil 240: 343–352.

Baraza E, JM Gómez, JA Hódar, and R Zamora. 2004. Herbivory has a greater impact in shade than in sun: response of Quercus pyrenaica seedlings to multifactorial environmental variation. Canadian Jorunal of Botany 82: 357–364.

Baraza E, R Zamora, and JA Hódar. 2006. Conditional outcomes in plant-herbivore interactions: neighbours matter. Oikos 113: 148–156.

Barber A, MR Cabrera, and I Guardiola. 1997. Sobre la cultura de l'espart al territori valencià. Fundació Bancaixa, Spain.

Barberá GG, JA Navarro-Cano, VM Castillo. 2006. Seedling recruitment in a semi-arid steppe: The role of microsite and post-dispersal seed predation. Journal of Arid environments 67: 701–714.

Barchuk AH, A Valiente-Banuet, MP Díaz. 2005. Effect of shrubs and seasonal variability of rainfall on the establishment of Aspidosperma quebracho-blanco in two edaphically contrasting environments. Austral Ecology 30: 695–705.

Bascietto M., and G Scarascia-Mugnozza. 2004. A collection of functions to determine annual tree carbon increment via stem-analysis. Annals of Forest Science 61: 597–602.

Bascompte J. 2009. Disentangling the Web of Life. Science 325: 416–419.

Bascompte J, and MA Rodríguez. 2001. Habitat patchiness and plant species richness. Ecology Letters 4: 417–420

Baskin CC, and JM Baskin. 1998. Seeds: ecology, biogeography, and evolution of dormancy and germination. Academic Press, San Diego, USA.

Baumeister D, and RM Callaway. 2006. Facilitative effects of Pinus flexilis during succession: a hierarchy of mechanisms benefits other plant species. Ecology 87: 1816–1930.

Beadle NCW. 1948. The Vegetation and Pastures of Western New South Wales with Special Reference to Soil Erosion. Department of Conservation of NSW, Sydney.

Bello MA, A Bruneau, F Forest, and JA Hawkins. 2009. Elusive relationships within Order Fabales : Phylogenetic Analyses using matK and rbcL sequence data. Systematics and Botany 34: 102-114.

Bellot J, FT Maestre, and N Hernández. 2004. Spatio-temporal dynamics of chlorophyll fluorescence in a semi-arid shrubland. Journal of Arid Environments 58: 295–308.

Belnap J, and OL Lange (Eds). 2003. Biological Soil Crusts: Structure, Function, and Management, Springer-Verlag: Berlin.

Ben Salem H, A Nefzaoui, and H Abdouli. 1994. Palatability of shrubs and fodder trees measured on sheep and camels. CIHEAM Options Mediterraneennes 4: 35–48.

Bertness MD, and RM Callaway. 1994. Positive interactions in communities. Trends in Ecology and Evolution 9: 191–193.

Bertness MD, and SW Shumway. 1993. Competition and facilitation in marsh plants. American Naturalist 142: 718–724.

Bochet, E, and P García-Fayos. 2004. Factors controlling vegetation establishment and water erosion on motorway slopes in Valencia, Spain. Restoration Ecology 12:166-174.

Bolling, JD, and LR Walker. 2000. Plant and soil recovery along a series of abandoned desert roads. Journal of Arid environments 46:1-24.

Booth, DT, JK Gores, GE Schuman, and RA Olson. 1999. Shrub Densities on Pre-1985 Reclaimed Mine Lands in Wyoming. Restoration Ecology 7 :24-32.

Bouchenak-Khelladi Y, N Salamin, V Savolainen, F Forest, M van der Bank, MW Chase, and TR Hodkinson. 2008. Large multi-gene phylogenetic trees of the grasses (Poaceae): Progress towards complete tribal and generic level sampling. Molecular and Phylogenetic Evolution 47: 488-505.

Bouchenak-Khelladi Y, GA Verboom, V Savolainen, and TR Hodkinson. 2010. Biogeography of the grasses (Poaceae): A phylogenetic approach to reveal evolutionary history in geographical space and geological time. Biological Journal of the Linnean Society 162: 543-557.

Bowker MA, S Soliveres, and FT Maestre. 2010. Competition increases with abiotic stress and regulates the diversity of biological soil crusts. Journal of Ecology 98: 551–560.

Brady KU, AR Kruckeberg, and HD Bradshaw Jr. 2005. Evolutionary ecology of plant adaptation to serpentine soils. Annual Review of Ecology, Evolution, and Systematics 36: 243–266.

Bremer K, EM Friis, and B Bremer. 2004. Molecular phylogenetic dating of asterid flowering plants shows Early Cretaceous diversification. Systematics and Biology 53: 496–505.

Bremer, B., and Eriksson, O. (2009). Time tree of Rubiaceae: Phylogeny and dating the

223

family, subfamilies and tribes. Int. J. Plant Sci., 170, 766-793.

Briske DD, SD Fuhlendorf, and FE Smeins. 2003. Vegetation dynamics on rangelands: a critique of the current paradigms. Journal of Applied Ecology 40: 601–614.

Brooker RW. 2006. Plant–plant interactions and environmental change. New Phytologist 171: 271–289.

Brooker RW, and TV Callaghan. 1998. The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. Oikos 81: 196–207.

Brooker RW, Z Kikvidze, FI Pugnaire, RM Callaway, P Choler, CL Lortie, and R Michalet. 2005. The importance of importance. Oikos 109: 63–70.

Brooker RW, FT Maestre, RM Callaway, CL Lortie, LA Cavieres, G Kunstler, P Liancourt, K Tielbörger, JMJ Travis, F Anthelme, C Armas, L Coll, E Corcket, S Delzon, E Forey, Z Kikvidze, J Olofsson, FI Pugnaire, P Saccone, K Schiffer, M Seifan, B Touzard, and R Michalet. 2008. Facilitation in plant communities: the past, the present, and the future. Journal of Ecology 96: 18–34.

Brown JR, and SR Archer. 1990. Water relations of a perennial grass and seedling versus adult woody plants in a subtropical savanna, Texas. Oikos 57: 366–374.

Bruno JF, JJ Stachowicz, and MD Bertness. 2003. Inclusion of facilitation into ecological theory. Trends in Ecology and Evolution 18: 119–125.

Burke, MJW, and JP Grime. 1996. An Experimental Study of Plant Community Invasibility. Ecology 77: 776-790.

Butterfield BJ. 2009 Effects of facilitation on community stability and dynamics: synthesis and future directions. Journal of Ecology 97: 1192–1201.

Butterfield BJ, JL Betancourt, RM Turner, and JM Briggs. 2010. Facilitation drives 65 years of vegetation change in the Sonoran Desert. Ecology 91: 1132–1139.

Caballero I. 2006. Estructura espacio-temporal de un banco de semillas: Las comunidades gipsícolas del centro de la Península Ibérica. PhD Thesis.

Caballero I, JM Olano, J Loidi, and A Escudero. 2008. A model for small-scale seed bank and standing vegetation connection along time. Oikos 117: 1788–1795.

Caballero I, JM Olano, A Escudero, and J Loidi. 2008. Seed bank spatial structure in semi-arid environments: beyond the patch-

bare area dichotomy. Plant Ecology 195: 215–223.

Cahill JF, SW Kembel, EG Lamb, and PA Keddy. 2008. Does phylogenetic relatedness influence the strength of competition among vascular plants? Perspectives in Plant Ecology, Evolution, and Systematics 10: 41–50.

Callaway RM. 1995. Positive interactions among plants. The Botanical Review 61: 306–349.

Callaway RM. 1997. Positive interactions in plant communities and the individualistic-continuum concept. Oecologia 112: 143–149.

Callaway RM. 2007. Positive Interactions and Interdependence in Plant Communities. Springer; New York.

Callaway RM, and ET Aschehoug. 2000. Invasive plants versus their new and old neighbors: a mechanism for exotic invasion. Science 290: 521–523.

Callaway RM, RW Brooker, P Choler, Z Kikvidze, CJ Lortie, R Michalet, L Paolini, FI Pugnaire, B Newingham, ET Aschehoug, C Armas, D Kikodze, and BJ Cook. 2002. Positive interactions among alpine plants increase with stress. Nature 417: 844–888.

Callaway RM, D Kikodze, and Z Kikvidze. 2000. Facilitation by unpalatable weeds may conserve plant diversity in overgrazed meadows in the Caucasus Mountains. Oikos 89:275–282.

Callaway RM, and LR Walker. 1997. Competition and facilitation: A synthetic approach to interactions in plant communities. Ecology 78: 1958–1965.

Caravaca MF, MM Alguacil, D Barea, JM Navarro, JM Roldán, and A Garrigos. 2003. Re-establishment of Retama sphaerocarpa as a target species for reclamation of soil physical and biological properties in a semi-arid Mediterranean area. Forest Ecology and Management 182: 49–58.

Castillo JP, M Verdú, and A Valiente-Banuet. 2010. Neighborhood phylodiversity affects plant performance. Ecology (in press)

Castillo-Monroy, AP, FT Maestre, M Delgado-Baquerizo, and A Gallardo. 2010. Biological soil crusts modulate nitrogen availability in semi-arid ecosystems: Insights from a Mediterranean grassland. Plant and Soil 333: 21-34.

Castro J, R Zamora, JA Hódar, and JM Gómez. 2002. The use of shrubs as nurse plants: a new technique for reforestation in Mediterranean mountains. Restoration Ecology 10: 297-305.

BIBLIOGRAFÍA

224

Cavender-Bares J, DA Ackerly, D Baum and FA Bazzaz. 2004. Phylogenetic overdispersion in Floridian oak communities, American Naturalist, 163: 823-843.

Cavender-Bares J, and FA Bazzaz. 2000. Changes in drought response strategies with ontogeny in Quercus rubra: Implications for scaling from seedlings to mature trees. Oecologia 124: 8–18.

Cavender-Bares J, A Keen, and B Miles. 2006. Phylogenetic structure of Floridian plant communities depends on taxonomic and spatial scale. Ecology 87: S109-S122.

Cavender-Bares J, K. Kozak, P Fine, and S Kembel. 2009. The merging of community ecology and phylogenetic biology. Ecology Letters 12: 693–715.

Cavieres LA, and EI Badano. 2009. Do facilitative interactions increase species richness at the entire community level? Journal of Ecology 97: 1181–1191.

Chao A, RL Chazdon, RK Colwell, and T Shen. 2005. A new statistical approach for assessing similarity of species composition with incidence and abundance data. Ecology letters 8: 148–159.

Chapin III, FS, AJ Bloom, CB Field, and RH Waring. 1987. Plant responses to multiple environmental factors. BioScience 37: 49–57.

Chapin III, FS, ED Schulze, and HA Mooney. 1990. The ecology and economics of storage in plants. Annual Review in Ecology and Systematics 21: 423–447.

Chase JM, MA Leibold, MA, AL Downing, and JB Shurin. 2000 The effects of productivity, herbivory, and plant species turnover in grassland food webs. Ecology 81: 2485–2497.

Choler Ph, R Michalet, and RM Callaway. 2001. Facilitation and competition on gradients in alpine plant communities. Ecology 82: 3295–3308.

Chu CJ, YS Wang, FT Maestre, and G Wang. 2008. On the relevance of facilitation in alpine meadow communities: An experimental assessment with multiple species differing in their ecological optimum. Acta Oecologica 33: 108–113.

Clarke KR, and RM Warwick. 2001. Change in Marine Communities: An Approach to Statistical Analysis and Interpretation. 2nd edition: PRIMER-E, Plymouth, UK.

Convención para la Lucha contra la Desertificación (CLD) 2005. Economic opportunities in the drylands under the united nations convention to combat desertification:

http://www.unccd.int/cop/cop7/docs/ss_background-eng.pdf.

Clements FE. 1916. Plant succession: An analysis of the development of vegetation. Washington, DC: Carnegie Institute. Washington.

Crawley MJ. 1998. Herbivory. In: Physiological plant ecology, MC Press et al. (Eds.), Blackwell Scientific Publications, Germany.

Colwell RK. 2000 EstimateS: statistical estimation of species richness and shared species from samples. V 8.2.0. Dept. of Ecol. and Evol. Biol., Univ. of Connecticut, USA. http://viceroy.eeb.uconn.edu/estimates.

Connor EF, and D Simberloff. 1979. The assembly of species communities: chance or competition? Ecology 60: 1132–1140.

Cook ER, and RL Holmes. 1996. Guide for computer program ARSTAN. In: Grissino-Mayer HD, Holmes RL, Fritts HC, eds. The international treering data bank program library version 2.0 user’s manual. Tucson, AZ, USA: Laboratory of Tree-Ring Research, University of Arizona. 75–87.

Cortina J, and FT Maestre. 2005. Plant effects on soils in drylands: implications on community dynamics and ecosystem restoration. In: Tree Species effects in Soils: Implications for Global Climate Change. Binkley D, and O. Menyailo (Eds.). NATO Science Series, Kluwer Academic Publishers, Dordrecht.

Cortina J, FT Maestre, and D Ramírez. 2010. Innovations in semiarid restoration. The case of Stipa tenacissima L. grass steppes. In: Bautista S, J Aronson, and R Vallejo (Eds.). Land Restoration to Combat Desertification: Innovative Approaches, Quality Control and Project Evaluation. Fundación CEAM, Spain.

Cortina J, FT Maestre, R Vallejo, MJ Baeza, A Valdecantos, and M Pérez-Devesa. 2005. Ecosystem structure, function, and restoration success: Are they related? Journal of Nature Conservation 14: 152–160.

Cortina J, RM Navarro, and A del Campo. 2006. Evaluación del éxito de la reintroducción de especies leñosas en ambientes mediterráneos. In: Cortina J, JL Peñuelas, J Puértolas, R Savé, and A Vilagrosa (Eds.). Calidad de planta forestal para la restauración de ambientes mediterráneos. Ministerio de Medio Ambiente, Spain.

Daleo P, and O Iribarne. 2009. Beyond competition: the stress-gradient hypothesis tested in plant– herbivore interactions. Ecology 90: 2368–2374

225

Darwin C. 1859. On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. John Murray, London.

Davis MA, KJ Wrage, PB Reich, MG Tjoelker, T Schaeffer, C Muermann. 1999. Survival, growth, and photosynthesis of tree seedlings competing with herbaceous vegetation along a water-light-nitrogen gradient. Plant Ecology 145: 341–350.

Dawson, TE. 1993. Hydraulic lift and water use by plants: implications for water balance, performance and plant-plant interactions. Oecologia 95: 565-574.

De´ath G, and KE Fabricius. 2000. Classification and regression trees: a powerful yet simple technique for ecological data analysis. Ecology 81: 3178-3192.

de la Cruz, M, RL Romao, A Escudero, FT Maestre. 2008. Where do seedlings go? A spatio-temporal analysis of seedling mortality in a semi-arid gypsophyte. Ecography 31: 720–730.

Diamond JM. 1975. Assembly of species communities. In: Ecology and evolution of communities, Cody ML, and JM Diamond (Eds), Harvard Press, USA.

Dullinger S, I Kleinbauer, H Pauli, M Gottfried, R Brooker, L Nagy, JP Theurillat, JI Holten, O Abdaladze, JL Benito, JL Borel, G Coldea, D Ghosn, R Kanka, A Merzouki¸ C Klettner, P Moiseev, U Molau, K Reiter, G Rossi, A Stancisi, M Tomaselli¸ P Unterlugauer, P Vittoz, and G Grabherr. 2007. Weak and variable relationships between environmental severity and small-scale co-occurrence in alpine plant communities. Journal of Ecology 95: 1284–1295.

Dyer ME, and RL Bailey. 1987. A Test of Six Methods for Estimating True Heights from Stem Analysis Data. Forest Science 33: 3–13.

Eccles N, KJ Esler, and RM Cowling. 1999. Spatial pattern analysis in Namaqualand desert plant communities: evidence for general positive interactions. Plant Ecology 142: 71–85.

Eldridge DJ, M Westoby, and KG Holbrook. 1991. Soil-surface characteristics, microtopography and proximity to mature shrubs: effects on survival of several cohorts of Atriplex vesicaria seedlings. Journal of Ecology 79: 357–364.

Eldridge DJ, and D Freudenberger. 2005. Ecosystem wicks: woodland trees enhance water flow in a fragmented agricultural landscape in eastern Australia. Austral Ecology 30: 336–347

Eliason SA, and EB Allen. 1997. Exotic grass competition in supressing native shrubland re-establishment. Restoration Ecology 5: 244–255.

Engqvist L. 2005. The mistreatment of covariate interaction terms in linear model analyses of behavioral and evolutionary ecology studies. Animal Behavior 70: 967–971.

Escudero A, JM Iriondo, JM Olano, A Rubio, and R Somolinos. 2000. Factors affecting the establishment of a gypsophyte: The case of Lepidium subulatum. American Journal of Botany 87: 861–871.

Escudero A, RC Somolinos, JM Olano, and A Rubio. 1999. Factors controlling the establishment of Helianthemum squamatum, an endemic gypsophile of semiarid Spain. Journal of Ecology 87: 290–302.

Espigares T, A López-Pintor, and JM Rey-Benayas. 2004. Is the interaction between Retama sphaerocarpa and its understorey herbaceous vegetation always reciprocally positive? Competition-facilitation shift during Retama sphaerocarpa establishment. Acta Oecologica 26: 121–128.

Fahey TJ, JJ Battles, GF Wilson. 1998. Responses of early successional northern hardwood forests to changes in nutrient availability. Ecological Monographs 68: 183–212.

Fischer RA, and NC Turner. 1978. Plant productivity in the arid and semi-arid zones. Annual Review in Plant Physiology 29: 277–317.

Flombaum, P., and O. E. Sala. 2009. Cover is a good predictor of biomass in arid systems. Journal of Arid Environments 73: 597–598.

Forest F, R Grenyer, M Rouget, TJ Davies, RM Cowling, DP Faith, A Balmford, JC Manning, S Proches, and M van der Bank. 2007. Preserving the evolutionary potential of floras in biodiversity hotspots. Nature 445: 757–760.

Forey E, B Touzard, and R Michalet. 2009. Does disturbance drive the collapse of biotic interactions at the sever end of a diversity-biomass gradient? Plant ecology 206: 287–295.

Forseth IN, DA Wait, and BB Casper. 2001. Shading by shrubs in a desert system reduces the physiological and demographic performance of an associated herbaceous perennial. Journal of Ecology 89: 670–680.

Fowler N. 1986. The role of competition in plant communities in arid and semiarid

BIBLIOGRAFÍA

226

regions. Annual Review of Ecology and Systematics 17: 89–110.

Fowler N. 2002. The joint effects of herbivory, competition, and topographic position upon six grasses. Ecology 83: 2477–2488.

Franco AC, and PS Nobel. 1989. Effect of nurse plants on the microhabitat and growth of cacti. Journal of Ecology 77: 870–886.

Franzke A, D German, IA Al-Shehbaz, and K Mummenhoff. 2009. Arabidopsis family ties: Molecular phylogeny and age estimates in Brassicaceae. Taxon 58: 425-437.

Friedman JJ, G Orshan, and Y Ziger-Cfir. 1977. Suppression of Annuals by Artemisia herba-alba in Negev Desert of Israel. Journal of Ecology 65: 413–426

Fuhlendorf SD, DD Briske, and FE Smeins. 2001. Herbaceous vegetation change in variable rangeland environments: the relative contribution of grazing and climatic variability. Applied Vegetation Science 4: 177–188.

Funk VA, R Bayer, S Keeley, R Chan, L Watson, B Gemeinholzer, E Schilling, J Panero, B Baldwin, NT García-Jacas, A Susanna, and RK Jansen. 2005. Everywhere but Antarctica: Using a supertree to understand the diversity and distribution of the Compositae. Biol. Skrifter 55: 343-374.

Gachet S, E Véla, and T Tatoni. 2005. BASECO: a floristic and ecological database of Mediterranean French flora. Biodiversity and Conservation 14: 1023–1034.

García-Palacios P. 2010. Interacciones planta-suelo en taludes de carretera: implicaciones para la restauración. PhD Thesis.

García-Palacios P, S Soliveres, FT Maestre, A Escudero, A Castillo-Monroy, and F Valladares. 2010. Dominant plant species modulate responses to hydroseeding, irrigation and fertilization during the restoration of semiarid motorway slopes. Ecological Engineering, (In press).

Gasque M, and P García-Fayos. 2004. Interaction between Stipa tenacissima and Pinus halepensis: consequences for reforestation and the dynamics of grass steppes in semiarid slopes of western Mediterranean areas. Forest Ecology and Management 189: 251–261

Gillman LN, and SD Wright. 2006. The influence of productivity on the species richness of plants: a critical assessment. Ecology 87: 1234–1243.

Gilpin ME. 1975. Limit cycles in competition communities. American Naturalist 109: 51–60.

Gleason HA. 1926. The Individualistic Concept of the Plant Association. Bulletin of the Torrey Botanical Club 53: 7–26.

Goldberg DE, and A Novoplansky. 1997. On the relative importance of competition in unproductive environments. Journal of Ecology 85: 409–418.

Goldberg DE, R Turkington, L Olsvig-Whittaker, and AR Dyer. 2001. Density dependence in an annual plant community: variation among life history stages. Ecological Monographs 71: 423–446.

Gómez-Aparicio L. 2009. The role of plant interactions in the restoration of degraded ecosystems: a meta-analysis across life-forms and ecosystems. Journal of Ecology 97: 1202–1214.

Gómez-Aparicio L, JM Gómez, and R Zamora. 2005. Microhabitats shift rank in suitability for seedling establishment depending on habitat type and climate. Journal of Ecology 93: 1194–1202.

Gómez-Aparicio L, F Valladares, and R Zamora. 2006. Differential light responses of Mediterranean tree saplings: linking ecophysiology with regeneration niche in four co-occurring species. Tree Physiology 26: 947–958.

Gómez-Aparicio L, R Zamora, JM Gómez, JA Hódar, J Castro, and E Baraza. 2004. Applying plant facilitation to forest restoration in Mediterranean ecosystems: a meta-analysis of the use of shrubs as nurse plants. Ecological Applications 14: 1128–1138.

Gotelli NJ. 2000. Null model analysis of species co-occurrence patterns. Ecology 81: 2606–2621.

Gotelli NJ, and GL Entsminger. 2006. EcoSim: Null Models Software for Ecology, Version 7. Acquired Intelligence Inc. and Kesey-Bear. Available at: http://garyentsminger.com/ecosim.htm.

Gotelli NJ, and GR Graves. 1996. Null Models in Ecology, Smithsonian Institution Press, Washington, DC.

Gotelli NJ, GR Graves, and C Rahbek. (2010). Macroecological signals of species interactions in the Danish avifauna. Proceedings of the National Academy of Sciences USA (In press).

Graff P, MR Aguiar, and EJ Chaneton. 2007. Shifts in positive and negative plant interactions along a grazing intensity gradient. Ecology 88: 188–199.

227

Grace JB. 1999. The factors controlling species density in herbaceous plant communities: an assessment. Perspectives in Plant Ecology, Evolution and Systematics 2: 1–28.

Grace JB, GR Guntenspergen, and JR Keough. 1993. The examination of a competition matrix for transitivity and intransitive loops. Oikos 68: 91–98.

Greiner La Peyre MK, JB Grace, E Hahn, and IA Mendelssohn. 2001. The importance of competition in regulating species abundance along a salinity gradient. Ecology 82: 62–69.

Grime JP. 1973. Competitive exclusion in herbaceous plant communities. Nature 242: 344–347.

Grime JP. 2001. Plant Strategies, Vegetation Processes, and Ecosystem Properties, 2nd edn.Wiley, Chichester, UK.

Grissino-Mayer HD. 2001. Evaluating crossdating accuracy: a manual and tutorial for the computer program COFECHA. Tree-Ring Research 57: 205–221.

Gross N, P Liancourt, Ph Choler, KN Suding, and S Lavorel. 2010. Strain and vegetation effects on local limiting resources explain the outcomes of biotic interactions. Perspectives in Plant Ecology, Evolution and Systematics 12: 9–19.

Guisan A, and NE Zimmermann. 2000. Predictive habitat distribution models in ecology. Ecological Modelling 135: 147-186.

Gurevitch J, JA Morrison, and LV Hedges. 2000. The interaction between competition and predation: a meta-analysis of field experiments. American Naturalist 155: 435–453.

Guzmán B, and P Vargas. 2009. Historical biogeography and character evolution of Cistaceae (Malvales) based on analysis of plastid rbcL and trnL-trnF sequences. Organisms Divers. Evolution 9: 83-99.

Guzmán B, MD Lledó, P Vargas. 2009. Adaptive radiation in Mediterranean Cistus (Cistaceae). Plos ONE, 4, e6362

Hacker SD, and SD Gaines. 1997. Some implications of direct positive interactions for community species diversity. Ecology 78: 1990–2003.

Hacker SD, and MD Bertness. 1999. The role of positive interactions in the plant species diversity of salt marsh plant communities. Ecology 80: 2064–2073.

Hastwell GT, and JM Facelli. 2003. Differing effects of shade-induced facilitation on growth and survival during the

establishment of a chenopod shrub. Journal of Ecology 91: 941–950.

Hay ME. 1986. Associational plant defenses and the maintenance of species diversity: turning competitors into accomplices. The American Naturalist 128: 617–641.

Hector A, B Schmid, C Beierkuhnlein, CMC Caldeira, M Diemer, PG Dimitrakopoulos, JA Finn, H Freitas, PS Giller, J Good, R Harris, P Högberg, K Huss-Danell, J Joshi, A Jumpponen, C Körner, PW Leadley, M Loreau, A Minns, CPH Mulder, G O'Donovan, SJ Otway, JS Pereira, A Prinz, DJ Read, M Scherer-Lorenzen, ED Schulze, ASD Siamantziouras, EM Spehn, AC Terry, AY Troumbis, FI Woodward, S Yachi, and JH Lawton. 1999. Plant diversity and productivity experiments in European grasslands. Science 286: 1123–1127.

Hedges SB, J Dudley, and S Kumar. 2006. TimeTree: a public knowledge-base of divergence times among organisms. Bioinformatics 22: 2971-2972.

Heisler-White JL, JM Blair, EF Kelly, K Harmoney, and AK Knapp. 2009 Contingent productivity responses to more extreme rainfall regimes across a grassland biome. Global Change Biology 15: 2894–2904.

Herrera CM. 1984. Tipos morfológicos y funcionales en plantas leñosas del matorral mediterráneo del sur de España. Studia Oecologica 5: 7–34.

Herrera CM. 1992. Historical effects and sorting processes as explanations of contemporary ecological patterns: character syndromes in Mediterranean woody plants. The American Naturalist 140: 421-446.

Hobbs RJ, S Arico, J Aronson, JS Baron, P Bridgewater, VA Cramer, PR Epstein, JJ Ewel, CA Klink, AE Lugo, D Norton, D Ojima, DM Richardson, EW Sanderson, F Valladares, M Vilá, R Zamora, and M Zobel. 2006. Novel ecosystems: theoretical and management aspects of the new ecological world order. Global Ecology and Biogeography 15: 1–7.

Holmgren M, and M Scheffer. 2010. Strong facilitation in mild environments: the stress gradient hypothesis revisited. Journal of Ecology (In press). doi: 10.1111/j.1365-2745.2010.01709.x

Holmgren M, M Scheffer, and MA Huston. 1997. The interplay of facilitation and competition in plant communities. Ecology 78: 1966–1975.

Holmgren M, P Stapp, CR Dickman, C Gracia, S Graham, JR Gutiérrez, C Hice, F Jaksic,

BIBLIOGRAFÍA

228

DA Kelt, M Letnic, M Lima, BC López, PL Meserve, WB Milstead, GA Polis, MA Previtali, M Richter, S Sabaté, and FA Squeo. 2006. Extreme climatic events shape arid and semiarid ecosystems. Frontiers in Ecology and the Environment 4: 87–95.

Holzapfel C, and BE Mahall. 1999. Bidirectional facilitation and interference between shrubs and annuals in the mojave desert. Ecology 80: 1747–1761.

Hooper DU, FS Chapin III, JJ Ewel, A Hector, P Inchausti, S Lavorel, JH Lawton, D Lodge, M Loreau, S Naeem, B Schmid, H Setälä, AJ Symstad, J Vandermeer, and DA Wardle. 2005. Effects of biodiversity on ecosystem functioning: a consensus of current knowledge. Ecological Monographs 75: 3–35.

Hubbell SP. 2001. The Unified Neutral Theory of Biodiversity and Biogeography. Princeton University Press, USA.

Huston MA. 1979. A general hypothesis of species diversity. American Naturalist 113: 81–101.

Huston MA. 1999. Local processes and regional patterns: appropriate scales for understanding variation in the diversity of plants and animals. Oikos 86: 393–401.

Huxman TE, JM Cable, DD Ignace, AJ Eilts, N English, J Weltzin, and DG Williams. 2004. Response of net ecosystem gas exchange to a simulated precipitation pulse in a semiarid grassland: the role of native versus non-native grasses and soil texture. Oecologia 141:295–305.

Ibáñez I, JS Clark, S LaDeau, and J Hille Ris Lambers. 2007. Exploiting temporal variability to understand tree recruitment response to climate change. Ecological Monographs 77: 163–177.

Ibañez I, and EW Schupp. 2001. Positive and negative interactions between environmental conditions affecting Cercocarpus ledifolius seedling survival. Oecologia 129: 543–550.

Illius AW, and TG O’Connor. 1999. On the relevance of non-equilibrium concepts to arid and semi-arid grazing systems. Ecological Applications 9: 798–813.

IPCC. 2007. Climate Change 2001: Impacts, Adaptation and Vulnerability. International Panel of Climate Change, Geneva, Switzerland

Isbell RF. 1996. The Australian soil classification. CSIRO Publishing. Collingwood, Australia.

Izco J. 1984. Madrid Verde. Ministerio de Agricultura, Pesca y Alimentación, Spain.

Janssens SB, EB Knox, S Huysmans, EF Smets, and V Merckx. 2009. Rapid radiation of Impatiens (Balsaminaceae) during Pliocene and Pleistocene: Result of a global climate change. Molecular Phylogenetics and Evolution 52: 806–824.

Jones CG, JH Lawton, and M Shachak. 1997. Positive and negative effects of organisms as physical ecosystem engineers. Ecology 78: 1946–1957.

Jorba M, and RVallejo. 2008. The ecological restoration of quarries: a case with application of organic amendment and irrigation. Ecosistemas 17: 119-132.

Kawai T, and M. Tokeshi. 2007. Testing the facilitation-competition paradigm under the stress-gradient hypothesis: decoupling multiple stress factors. Proceedings of the Royal Society London B 274: 2503–2508.

Kéfi S, M Rietkerk, CL Alados, Y Pueyo, A ElAich, V Papanastasis, and PC de Ruiter. 2007. Spatial vegetation patterns and imminent desertification in Mediterranean arid ecosystems. Nature. 449: 213–217.

Keith D. 1998. Ocean shores to desert dunes: The native vegetation of NSW and the ACT. Department of Environment and Climate Change, Australia.

Kembel SW, PD Cowan, MR Helmus, WK Cornwell, H Morlon, DD Ackerly, SP Blomberg, and CO Webb. 2010. Picante: R tools for integrating phylogenies and ecology. Bioinformatics 26: 1463-1464.

Kikvidze Z. 1996. Neighbour interaction and stability in subalpine meadow communities. Journal of Vegetation Science 7: 41–44.

Kikvidze Z, L Khetsuriana, D Kikodze, and RM Callaway. 2006. Seasonal shifts in competition and facilitation in subalpine plant communities of the central Caucasus. Journal of Vegetation Science 17: 77–82.

Kikvidze Z, FI Pugnaire, RW Brooker, Ph Choler, CJ Lortie, R Michalet, and RM Callaway. 2005. Linking patterns and processes in alpine plant communities: a global study. Ecology 86: 1395–1400.

Kim KJ, KS Choi, and RK Jansen. 2005. Two chloroplast DNA inversions originated simultaneously during the early evolution of the sunflower family (Asteraceae). Molecular Biology and Evolution 22: 1783–1792.

Kitzberger T, DF Steinaker, and TT Veblen. 2000. Effects of climatic variability on facilitation of tree establishment in northern Patagonia. Ecology 81: 1914–1924.

Knapp AK, C Beier, DD Brisker DD, AT Classen, Y Luo, M Reichstein, MD Smith,

229

SD Smith, JE Bell, PA Fay, JL Heisler, SW Leavitt, R Sherry, B Smith, and E Weng. 2008. Consequences of more extreme precipitation regimes for terrestrial ecosystems. BioScience 58: 811–821.

Knoop WT, and BH Walker. 1985. Interactions of woody and herbaceous vegetation in a Southern African savanna. Journal of Ecology 73: 235–253.

Kobe RK, SW Pacala, JA Silander, and CD Canham. 1995. Ecological Applications 5: 517–532.

Körner C. 2003: Limitation and stress - always or never? Journal of Vegetation Science 14: 141–143.

Körner C. 2004. Individuals have limitations, not communities: A response to Marrs, Weiher and Lortie et al. Journal of Vegetation Science 15: 581–582.

Kraft NJB, and D Ackerly. 2010. Functional trait and phylogenetic tests of community assembly across spatial scales in an Amazonian forest. Ecological Monographs 80: 401–422

Kraft NJB, WK Cornwell, CO Webb, and DD Ackerly. 2007. Trait evolution, community assembly and the phylogenetic structure of ecological communities. American Naturalist 170: 271–283.

Kurashige NS, and AA Agrawal. 2005. Phenotypic plasticity to light competition and herbivory in Chenopodium album (Chenopodiaceae). American Journal of Botany 92: 21–26.

Laird RA, and BS Schamp. 2006. Competitive intransitivity promotes species coexistence. American Naturalist 168: 182–193.

Laird RA, and BS Schamp. 2008. Does local competition increase the coexistence of species in intransitive networks? Ecology 89: 237–247.

Lavin M, PS Herendeen, and MF Wojciechowski. 2005. Evolutionary rates analysis of Leguminosae implicates a rapid diversification of lineages during the Tertiary emergence. Systematics and. Biology 54: 575-594.

Le Houérou HN. 1986. The desert and arid zones of Northern Africa. In: Ecosystems of the world 12B. Hot deserts and arid shrublands. Evenary, M, I Noy-Meir, and DW Goodall (Eds). B, Elsevier, Amsterdam.

Le Houérou HN. 2001. Biogeography of the arid steppeland north of the. Sahara. Journal of Arid Environments 48: 103–128.

Le Roux PC, and MA McGeoch. 2010. Interaction intensity and importance along two stress gradients: adding shape to the

stress-gradient hypothesis. Oecologia 162: 733–745.

Levin SA. 1970. Community equilibria and stability, and an extension of the competitive exclusion principle. American Naturalist 104: 413–423

Levine JM. 1999. Indirect facilitation: evidence and predictions from a riparian community. Ecology 80: 1762–1769.

Liancourt P, RM Callaway, and R Michalet. 2005. Stress tolerance and competitive-response ability determine the outcome of biotic interactions. Ecology 86: 1611–1618.

Loiselle BA, CA Howell, CH Graham, JM Goerck, T Brooks, KG Smith, and PH Williams. 2003. Avoiding pitfalls of using species distribution models in conservation planning. Conservation Biology 17: 1591–1600.

López BC, M Holmgren, S Sabaté, CA Gracia. 2008. Estimating annual rainfall threshold for establishment of tree species in water-limited ecosystems using tree-ring data. Journal of Arid Environments 72: 602–611.

Lortie CJ. 2010. Synthetic analysis of the stress-gradient hypothesis. In: Positive Plant Interactions and Community Dynamics. Pugnaire FI (Ed.). C.R.C. Press, USA.

Lortie CJ, RW Brooker, Ph Choler, Z Kikvidze, R Michalet, FI Pugnaire, and RM Callaway. 2004a. Rethinking plant community theory. Oikos 107: 433–438.

Lortie CJ, RW Brooker, Z Kikvidze, and RM Callaway. 2004b. The value of stress and limitation in an imperfect world: a reply to Körner. Journal of Vegetation Science 15: 577–580.

Lortie CJ, and RM Callaway. 2006. Re-analysis of meta-analysis: support for the stress-gradient hypothesis. Journal of Ecology 94: 7–16.

Ludwig F, TE Dawson, HHT Prins, F Berendse, and H de Kroon. 2004. Below-ground competition between trees and grasses may overwhelm the facilitative effects of hydraulic lift. Ecology letters 7: 623–631.

Ludwig JA, and DJ Tongway. 1995. Spatial organization of landscapes and its function in semi-arid woodlands, Australia. Landscape Ecology 10: 51–63

Ludwig JA, and DJ Tongway. 1996. Rehabilitation of semi-arid landscapes in Australia II: restoring productive soil patches. Restoration Ecology 4: 398-406.

Maestre FT. 2002. La restauración de la cubierta vegetal en zonas semiáridas en función del patrón espacial de factores bióticos y abióticos. PhD Thesis.

BIBLIOGRAFÍA

230

Maestre FT, S Bautista, J Cortina, and J Bellot. 2001. Potential of using facilitation by grasses to establish shrubs on a semiarid degraded steppe. Ecological Applications 11: 1641–1655.

Maestre FT, S Bautista, and J Cortina. 2003. Positive, negative and net effects in grass-shrub interactions in semiarid Mediterranean steppes. Ecology 84: 3186–3197.

Maestre FT, MA Bowker, C Escolar, MD Puche, S Soliveres, S Mouro, P García-Palacios, AP Castillo-Monroy, I Martínez, and A Escudero. 2010. Do biotic interactions modulate ecosystem functioning along abiotic stress gradients? Insights from semi-arid Mediterranean plant and biological soil crust communities. Philosophical Transactions of the Royal Society B 365: 2057–2070.

Maestre FT, RM Callaway, F Valladares, and CJ Lortie. 2009a. Refining the stress-gradient hypothesis for competition and facilitation in plant communities. Journal of Ecology 97: 199–205.

Maestre FT, and J Cortina. 2004. Do positive interactions increase with abiotic stress? A test from a semi-arid steppe. Proceedings of the Royal Society of London, Series B (Supplement) 271: S331–S333.

Maestre FT, and J Cortina. 2004b. Are Pinus halepensis afforestations useful as a restoration tool in semiarid Mediterranean areas? Forest ecology and management 198: 303–317.

Maestre FT, and J Cortina. 2004c. Insights on ecosystem composition and function in a sequence of degraded semiarid steppes. Restoration Ecology 12: 494–502

Maestre FT, J Cortina, and S Bautista. 2004. Mechanisms underlying the interaction between Pinus halepensis and the native late-successional shrub Pistacia lentiscus in a semiarid plantation. Ecography 27: 776–786.

Maestre FT, and J Cortina. 2005. Remnant shrubs in Mediterranean semi-arid steppes: effects of shrub size, abiotic factors and species identity on understorey richness and occurrence. Acta Oecologica 27: 161–169.

Maestre FT, C Escolar, I Martínez, and A Escudero. 2008. Are soil lichen communities structured by biotic interactions? A null model analysis. Journal of Vegetation Science 19: 261–266.

Maestre FT, and A Escudero. 2009. Is the patch-size distribution of vegetation a

suitable indicator of desertification processes? Ecology 90: 1729-1735.

Maestre FT, MD Puche, MA Bowker, MB Hinojosa, I Martínez, P García-Palacios, A P Castillo, S Soliveres, AL Luzuriaga, AM Sánchez, JA Carreira, AGallardo, and A Escudero. 2009. Shrub encroachment can reverse desertification in Mediterranean semiarid grasslands. Ecology Letters 12: 930–941.

Maestre FT, DA Ramírez, and J Cortina. 2007. Ecología del esparto (Stipa tenacissima L.) y los espartales de la Península Ibérica. Ecosistemas 16: 117–136.

Maestre FT, F Valladares, and JF Reynolds. 2005. Is the change of plant-plant interactions with abiotic stress predictable? A meta-analysis of field results in arid environments. Journal of Ecology 93: 748–757.

Maestre FT, F Valladares, and JF Reynolds. 2006. The stress-gradient hypothesis does not fit all relationships between plant–plant interactions and abiotic stress: further insights from arid environments. Journal of Ecology 94: 17–22.

Magallón S, and A Castillo. 2009. Angiosperm diversification through time. American Journal of Botany 96: 349-365.

Maherali H, and JN Klironomos. 2007. Influence of phylogeny on fungal community assembly and ecosystem functioning. Science 316: 1746–1748.

Malkinson D, and K Tielbörger. 2010. What does the stress-gradient hypothesis predict? Resolving the discrepancies. Oikos 119: 1546–1552.

Marañón T, and JW Bartolome. 1993. Reciprocal transplants of herbaceous communities between Quercus agrifolia woodland and adjacent grassland. Journal of Ecology 81: 673–682.

Marqués MJ, R Bienes, R Pérez-Rodríguez, and L Jiménez. 2008. Soil degradation in Central Spain due to sheet water erosion by low-intensity rainfall events. Earth Surface Processes and Landforms 33: 414–423.

Matesanz S. 2008. Ecología funcional de plantas de yesos: respuestas y vulnerabilidad al cambio global. PhD Thesis.

Matesanz S, F Valladares, D Tena, M Costa-Tenorio, and D Bote. 2006. Early Dynamics of Plant Communities on Revegetated Motorway Slopes from Southern Spain: Is Hydroseeding Always

231

Needed?. Restoration Ecology 14: 297–307.

Matías, L. 2010. Efectos del Cambio Climático sobre la regeneración del bosque mediterráneo: una aproximación experimental. PhD Thesis.

Maxwell K, and GN Johnson. 2000. Chlorophyll fluorescence: a practical guide. Journal of Experimental Botany 51: 659–668.

Mayfield MM, and JM Levine. 2010. Opposing effects of competitive exclusion on the phylogenetic structure of communities. Ecology Letters (In press).

McAuliffe JR. 1988. Markovian dynamics of simple and complex desert plant communities. American Naturalist 131: 459–490.

McCullagh P, and J Nelder. 1989. Generalized Linear Models, Second Edition. Chapman and Hall, London.

McDonald PM. 1986. Grasses in young conifer plantations-Hindrance and help. Nortwest Science 60: 271–278.

McNaughton SJ. 1978 Serengeti Ungulates: Feeding Selectivity Influences the Effectiveness of Plant Defense Guilds. Science 17: 806–807.

Méndez M, D García, FT Maestre, and A Escudero. 2008. More ecology is needed to restore Mediterranean ecosystems. A reply to Valladares and Gianoli. Restoration Ecology 16: 210-216.

Meyer SE. 1986. The ecology of gypsophile endemism in the Eastern Mojave desert. Ecology 67: 1303–1313.

Michalet R. 2007. Highlighting the multiple drivers of change in the interactions along stress gradients. New Phytologist 173: 1–2.

Michalet R, RW Brooker, LA Cavieres, Z Kikvidze, CJ Lortie, FI Pugnaire, A Valiente-Banuet, and RM Callaway. 2006. Do biotic interactions shape both sides of the humped-back model of species richness in plant communities? Ecology Letters 9: 767–773.

Milchunas DG, WK Lauenroth, PL Chapman, and MK Kazempour. 1989. Effects of grazing, topography, and precipitation on the structure of a semiarid grassland. Vegetatio 80: 11–23.

Millennium Ecosystem Assessment. 2005. Ecosystems and Human Well-being: Desertification Synthesis. World Resources Institute, Washington, DC.

Miranda JD, FM Padilla, and FI Pugnaire. 2009. Response of a Mediterranean semiarid community to changing patterns of water

supply. Perspectives in Plant Ecology, Evolution, and Systematics, 11: 255–266.

Miriti MN. 2006. Ontogenetic shift from facilitation to competition in a desert shrub. Journal of Ecology 94: 973–979.

Miriti MN, HF Howe, and SJ Wright. 1998. Spatial patterns of mortality in a Colorado desert plant community. Plant Ecology 136: 41–51.

Miriti MN, SJ Wright, and HF Howe. 2007. Twenty years of changes in spatial association and community structure in a community of desert perennials. Ecology 88: 1177–1190.

Mitchell, MGE, JF Cahill, and DS Hik. 2009. Plant interactions are unimportant in a subarctic-alpine plant community. Ecology 90: 2360-2367.

Mittlebach GG. 2010. Understanding species richness–productivity relationships: the importance of meta-analyses. Ecology 91: 2540-2544.

Moro MJ, FI Pugnaire, P Haase, and J Puigdefábregas. 1997. Effect of the canopy of Retama sphaerocarpa on its understorey in a semiarid environment. Functional Ecology 11: 175-184.

Mulder CPH, DD Uliassi, and DF Doak. 2001. Physical stress and diversity-productivity relationships: the role of positive interactions. Proceedings of the National Academy of Sciences 98: 6704–6708.

Navarro RM, A del Campo, and J Cortina. 2006. Factores que afectan al éxito de una repoblación y su relación con la calidad de planta. In: Cortina J, JL Peñuelas, J Puértolas, R Savé, and A Vilagrosa (Eds.). Calidad de planta forestal para la restauración de ambientes mediterráneos. Ministerio de Medio Ambiente, Spain.

Navarro FB, J Lorite, E Fernández-Ondoño, MA Ripoll, and MN Jiménez. 2008. Stipa tenacissima as a nurse planto f the endemic species Haplophyllum bastetanum near Granada, SE Spain. Applied Vegetation Science 11: 63–72.

Niering WA, RH Whittaker, and CH Lowe. 1963. The saguaro: a population in relation to environment. Science 142: 15–23.

Niinemets U, and F Valladares. 2006. Tolerance to shade, drought, and waterlogging of temperate Northern Hemisphere trees and shrubs. Ecological Monographs 76: 521–547.

Ninyerola M, X Pons, and JM Roure. 2005. Atlas Climático Digital de la Península Ibérica. Metodología y aplicaciones en bioclimatología y geobotánica.

BIBLIOGRAFÍA

232

ISBN932860-8-7. Universidad Autónoma de Barcelona, Bellaterra, Spain.

Noble JC, and DJ Tongway. 1986. Herbivores in arid and semi-arid rangelands. In: Australian Soils: The Human Impact. Russell JS, and RF Isbell (Eds.). University of Queensland Press, Australia.

Noy-Meir I. 1973. Desert ecosystems: environment and producers. Annual Review in Ecology and Systematics 4: 25–51.

Noy-Meir I. 1975. Stability of grazing systems: an application of predator–prey graphs. Journal of Ecology 63: 459–481.

Olano JM, I Caballero, A Escudero, and J Loidi. 2005. Prediction of plant cover from seed bank analysis in a semi-arid plant community on gypsum. Journal of Vegetation Science 16: 215-222.

Olano JM, ES Menges, and E Martinez. 2006. Carbohydrate storage in five resprouting Florida scrub plant across a fire chronosequence. New Phytologist 170: 99−106.

Ogle K, and JF Reynolds. 2004. Plant responses to precipitation in desert ecosystems: integrating functional types, pulses, thresholds, and delays. Oecologia 141: 282–294.

Padilla FM, and FI Pugnaire. 2007. Rooting depth and soil moisture control Mediterranean woody seedling survival during drought. Functional Ecology 21: 489–495.

Pakeman RJ, FI Pugnaire, R Michalet, CJ Lortie, K Schiffers, FT Maestre, and J Travis. 2009. Is the cask of facilitation ready for bottling? A symposium on the connectivity and future directions of positive plant interactions. Biology Letters 5: 577–579.

Palacio S, A Escudero, G Montserrat-Martí, M Maestro-Martínez, R Milla, and MJ Albert. 2007. Plants living in gypsum: beyond the specialist model. Annals of Botany 99: 333–343.

Palacio S, M Maestro, and G Montserrat-Martí. 2007. Seasonal dynamics of nonstructural carbohydrates in two species of Mediterranean sub-shrubs with different leaf phenology. Environmental and Experimental Botany 59: 34–42.

Palacio S, R Milla, J Albuixech, C Pérez-Rontomé, JJ Camarero, M Maestro, and G Montserrat-Martí. 2008. Seasonal variability of dry matter content and its relationship with shoot growth and nonstructural carbohydrates. New Phytologist 180: 133–142.

Palacio S, and G Montserrat-Martí. 2005. Bud morphology and shoot growth dynamics in two species of Mediterranean sub-shrubs co-existing in gypsum outcrops. Annals of Botany 95: 949–958.

Parker VT, and CH Muller. 1982. Vegetational and environmental-changes beneath isolated live oak trees (Quercus agrifolia) in a California annual grassland. American Midland Naturalist 107: 69–81.

Paula S, M Arianoutsou, D Kazanis, Ç Tavsanoglu, F Lloret, C Buhk, F Ojeda, B Luna, JM Moreno, A Rodrigo, JM Espelta, S Palacio, B Fernández-Santos, PM Fernandes, and JG Pausas. 2009. Fire-related traits for plant species of the Mediterranean Basin. Ecology 90: 1420.

Pausas JG, A Bonet, FT Maestre, and A Climent. 2006. The role of the perch effect on the nucleation process in Mediterranean semi-arid oldfields. Acta Oecologica 29: 346–352.

Pausas JG, and M Verdú. 2010. The jungle of methods for evaluating phenotypic and phylogenetic structure of communities. BioScience 60: 614–625.

Pavoine S, E Vela, S Gachet, G de Bélair, and MB Bonsall. 2010. Linking patterns in phylogeny, traits, abiotic variables and space : a novel approach to linking environmental filtering and plant community assembly. Journal of Ecology. In press.

Peco B, L Rico, and FM Azcárate. 2009. Seed size and response to rainfall patterns in annual grasslands: 16 years of permanent plots. Journal of Vegetation Science 20: 8–16.

Perry JN. 1998. Measures of spatial pattern for counts. Ecology 79: 1008–1017.

Pias B, S Matesanz, A Escudero, TE Gimeno, and F Valladares. 2010. Effects of Global Change on early fitness components and offspring performance of an endemic Mediterranean gypsophyle. Oikos (In press) doi: 10.1111/j.1600-0706.2010.18232.x

Prider JN, and JM Facelli. 2004. Interactive effects of drought and shade on three arid zone chenopod shrubs with contrasting distributions in relation to tree canopies. Functional Ecology 18: 67–76.

Poorter H, Ü Niinemets, I Poorter, IJ Wright, and R Villar. 2009. Causes and consequences of variation in leaf mass per area (lma): a meta-analysis. New Phytologist 182: 565–588.

Pueyo Y, and CL Alados. 2007. Effects of fragmentation, abiotic factors and land use on vegetation recovery in a semi-arid

233

Mediterranean area. Basic and Applied Ecology 8: 158–170.

Pueyo Y, CL Alados, B García-Ávila, S Kéfi, M Maestro, and M Rietkerk. 2009. Comparing direct abiotic amelioration and facilitation as tools for restoration of semiarid grasslands. Restoration Ecology 17: 908–916.

Puigdefábregas J, A Sole, L Gutierrez, G del Barrio, and M Boer. 1999. Scales and processes of water and sediment redistribution in drylands: results from Rambla Honda study site in Southeast Spain. Earth-Science Reviews 48: 39–70.

Puigdefábregas J, and G Sánchez. 1996. Geomorphological implications of vegetation patchiness on semiarid slopes. In: Advances in hillslope processes, Volume 2. Anderson MG, and SM Brooks (Eds.). John Willey y Sons, London.

Pugnaire FI, C Armas, and F Valladares. 2004. Soil as a mediator in plant-plant interactions in a semi-arid community. Journal of Vegetation Science 15: 85-92.

Pugnaire FI, P Haase, J Puigdefábregas, M Cueto, LD Incoll, and SC Clark. 1996a. Facilitation and succession under the canopy of the leguminous shrub, Retama sphaerocarpa, in a semi-arid environment in south-east Spain. Oikos 76: 455–464.

Pugnaire FI, P Haase, LD Incoll, and SC Clark. 1996b. Response of the tussock grass Stipa tenacissima to watering in a semi-arid environment. Functional Ecology 10: 265–274.

Pugnaire FI, and MT Luque. 2001. Changes in plant interactions along a gradient of environmental stress. Oikos 93: 42–49.

Quinn GP, and MJ Keough. 2002. Experimental design and data analysis for biologists. Cambridge University Press.

Rajaniemi TK, DE Goldberg, R Turkington, and AR Dyer. 2006. Quantitative partitioning of regional and local processes shaping regional diversity patterns. Ecology Letters 9: 121–128.

Ramírez DA. 2006. Estudio de la transpiración del esparto (Stipa tenacissima L.) en una cuenca del semiárido alicantino: un análisis pluriescalar. PhD Thesis.

Ramírez DA, and J Bellot. 2009. Linking population density and habitat structure to ecophysiological responses in semiarid Spanish steppes. Plant Ecology 200: 191–204.

Ramírez DA, J Bellot, F Domingo, and A Blasco. 2007. Can water responses in Stipa tenacissima L. during summer season be promoted by non-rainfall water gains in soil? Plant Soil 291: 67–79.

Rand TA. 2004. Competition, facilitation, and compensation for insect herbivory in an annual salt marsh forb. Ecology 85: 2046–2052.

Rebollo S, DG Milchunas, I Noy-Meir, and PL Chapman. 2002. The role of a spiny plant refuge in structuring grazed shortgrass steppe communities. Oikos 98: 53–64.

Resco V, DD Ignace, W Sun, TE Huxman, JF Weltzin, and DG Williams. 2008. Chlorophyll fluorescence, predawn water potential and photosynthesis in precipitation pulse-driven ecosystems: implications for ecological studies. Functional Ecology 22: 479–283.

Reynolds JF, PR Kemp, K Ogle, and RJ Fernandez. 2004. Modifying the 'pulse-reserve' paradigm for deserts of North America: precipitation pulses, soil water, and plant responses. Oecologia 141:194–210.

Reynolds JF, FT Maestre, E Huber-Sannwald, JE Herrick, and PR Kemp. 2005. Aspectos socioeconómicos y biofísicos de la desertificación. Ecosistemas 14: 3–21.

Reynolds, JF, DM Stafford Smith, EF Lambin, BL Turner II, M Mortimore, SPJ Batterbury, TE Downing, H Dowlatabadi, RJ Fernández, JE Herrick, E Huber-Sannvald, R Leemans, T Lynam, FT Maestre, M Ayarza, and B Walker. 2007a. Global desertification: Building a science for dryland development. Science 316: 847–851.

Reynolds JF, FT Maestre, PR Kemp, DM Stafford-Smith, and E Lambin. 2007b. Natural and human dimensions of land degradation in drylands: causes and consequences. In: Terrestrial Ecosystems in a Changing World. Canadell J, D Pataki, and LF Pitelka (Eds.). Springer-Verlag, Berlin.

Rezende E, J Lavabre, P Guimaraes, P Jordano, and J Bascompte. 2007. Non-random coextinctions in phylogenetically structured mutualistic networks. Nature 448: 925–928.

R Development Core Team. 2009. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna.

Rietkerk MG, and J van de Koppel. 2008. Regular pattern formation in real ecosystems. Trends in ecology and evolution 23: 169–175.

Riginos C, SJ Milton, and T Wiegand. 2005. Context-dependent interactions between adult shrubs and seedlings in a semi-arid shrubland. Journal of Vegetation Science 16: 331–340.

BIBLIOGRAFÍA

234

Riginos C, and TP Young. 2007. Positive and negative effects of grass, cattle, and wild herbivores on Acacia saplings in an East African savanna. Oecologia 153: 985–995.

Romao R, and A Escudero. 2005. Gypsum physical soil surface crusts and the existence of gypsophytes in semi-arid central Spain. Plant Ecology 181: 1–11.

Roques KG, TG O'Connor, and AR Watkinson. 2001. Dynamics of shrub encroachment in an African savanna: relative influences of fire and herbivory, rainfall and density dependence. Journal of Applied Ecology 38: 268–280.

Rooney TP. 2008. Comparison of co-occurrence structure of temperate forest herb-layer communities in 1949 and 2000. Acta Oecologica 34: 354–360.

Rueda M, S Rebollo, L Gálvez-Bravo, and A Escudero. 2008. Habitat use by large and small herbivores in a fluctuating Mediterranean ecosystem: Implications of seasonal changes. Journal of Arid Environments 72: 1698–1708.

Sala OE, RA Golluscio, WK Lauenroth, and CA Soriano. 1989. Resource Partitioning Between Shrubs and Grasses in the Patagonian Steppe. Oecologia 81: 501–505.

Sala OE, and WK Lauenroth. 1982. Small rainfall events: an ecological role in semiarid regions. Oecologia 53: 301–304.

Sargent R, and DD Ackerly. 2008. Plant-pollinator interactions and community assembly. Trends in Ecology and Evolution 23: 123–130.

Scheffer M, J Bascompte, WA Brock, V Brovkin, SR Carpenter, V Dakos, H Held, EH van Nes, M Rietkerk, and G Sugihara, 2009. Early warning signals for critical transitions, Nature 461: 53–59.

Schiffers K, and K Tielbörger. 2006. Ontogenetic shifts in interactions among annual plants. Journal of Ecology 94: 336–341.

Schlesinger WH, and AM Pilmanis. 1998. Plant-soil interactions in desert. Biogeochemistry 42: 169–187.

Schlesinger WH, JF Reynolds, GL Cunningham, LF Huenneke, WM Jarrell, RA Virginia, and WG Whitford. 1990. Biological feedbacks in global desertification. Science 247: 1043–1048.

Scholes RJ, and SR Archer. 1997. Tree-grass interactions in savannas. Annual Review in Ecology and Systematics 28: 517–544.

Schupp EW. 1995. Seed-seedling conflicts, habitat choice and patterns of plant recruitment. American Journal of Botany 82: 399–409.

Schweingruber FH. 1988. Tree Rings: Basics and Applications of Dendrochronology. D. Reidel, Dordrecht; The Netherlands.

Schwinning S, and JR Ehleringer. 2001. Water use trade-offs and optimal adaptations to pulse-driven arid ecosystems. Journal of Ecology 89: 464–480.

Schwinning S, and OE Sala. 2004. Hierarchy of responses to resource pulses in and semiarid ecosystems. Oecologia 141: 211–220.

Seifan M, K Tielbörger, and R Kadmon. 2010a. Direct and indirect interactions among plants explain counterintuitive positive drought effects on an eastern Mediterranean shrub species. Oikos 119: 1601-1609.

Seifan M, T Seifan, C Ariza, and K Tielbörger. 2010b. Facilitating an importance index. Journal of Ecology 98: 356–361.

Shreve F. 1931. Physical conditions in sun and shade. Ecology 12: 96–104.

Silliman BR, J Van de Koppel, MD Bertness, L Stanton, and I Mendelssohn. 2005. Drought and herbivory lead to catastrophic salt marsh collapse. Science 310:1803–1805

Silvertown J. 2004. Plant coexistence and the niche. Trends in Ecology and Evolution 19: 605–611.

Smit C, M Rietkerk, and MJ Wassen. 2009. Inclusion of biotic stress (consumer pressure) alters predictions from the stress gradient hypothesis. Journal of Ecology 97: 1215– 1219.

Smit C, C Vanderberghe, J den Ouden, and H Müller-Schärer. 2007. Nurse plants, tree saplings and grazing pressure: changes in facilitation along a biotic environmental gradient. Oecologia 152: 265–273.

Smith SA, JM Beaulieu, and MJ Donoghue. 2010. An uncorrelated relaxed-clock analysis suggests an earlier origin for flowering plants. Proceedings of the National Academy of Sciences USA 107: 5897-5902.

Soliveres S, V Contreras, JP Fernandes, J Cortina, P García-Palacios, M Martínez, and J Fort. Ecotecnología aplicada a la restauración de infraestructuras lineales. In. F Valladares et al. (Eds.). Ecología e infraestructuras lineales: bases científicas para soluciones técnicas. Ministerio de Medio Ambiente, Spain. In press.

Soliveres S, L DeSoto, FT Maestre, and JM Olano. 2010. Spatio-temporal heterogeneity in abiotic factors modulate multiple ontogenetic shifts between competition and facilitation. Perspectives

235

in Plant Ecology, Evolution and Systematics 12: 227–234.

Soliveres S, P García-Palacios, AP Castillo-Montroy, FT Maestre, A Escudero, and F Valladares. 2010. Temporal dynamics of herbivory and water availability interactively modulate the outcome of a grass-shrub interaction in a semi-arid ecosystem. Oikos (In press) Doi: 10.1111/j.1600-0706.2010.18993.x

Soliveres S, J Monerris, and J Cortina. 2008. El uso de parches artificiales mejora el rendimiento de una repoblación de Rhamnus lycioides en medio semiárido. Cuadernos de la Sociedad Española de Ciencias Forestales 28: 125-130.

Soil Survey Staff. 1994. Keys to Soil Taxonomy, Sixth edition. USDA Soil Conservation Service, Pocahontas Press; Blacksburg.

SPSS. 2004. SPSS® Trends 13.0 . SPSS Inc.; Chicago.

Stace HCT, GD Hubble, R Brewer, KH Northcote, JR Sleeman, MJ Mulcahy, and EG Hallsworth. 1968. A Handbook of Australian Soils. Rellim Technical Publications, CSIRO and International Society of Soil Science, Glenside, South Australia.

Stachowicz JJ. 2001. Mutualisms, positive interactions, and the structure of ecological communities. BioScience 51: 235–246.

Sternberg M, and M Shoshany. 2001. Influence of slope aspect on Mediterranean woody formations: comparison of a semiarid and an arid site in Israel. Ecological Research 16: 335–345.

Stevens PF. 2010. Angiosperm Phylogeny Website: http://www.mobot.org/MOBOT/research/APweb/.

Sthultz CM, CA Gehring, and TG Whitham. 2007. Shifts from competition to facilitation between a foundation tree and a pioneer shrub across spatial and temporal scales in a semiarid woodland. New Phytologist 173: 135–145.

Stokes MA, and TL Smiley. 1968. An Introduction to Tree-ring Dating. University of Chicago Press; Chicago.

Susanna A, N Garcia-Jacas, O Hidalgo, R Vilatersana, and T Garnatje. 2006. The Cardueae (Compositae) revisited: insights from ITS, trnL-trnF, and matK nuclear and chlorpoplast DNA analysis. Annals of the Missouri Botanical Garden 93: 150-171.

Tielbörger K, and R Kadmon. 2000a. Temporal environmental variation tips the balance

between facilitation and interference in desert plants. Ecology 81: 1544–1553.

Tielbörger K, and R Kadmon. 2000b. Indirect effects in a desert plant community: is competition among annuals more intense under shrub canopies? Plant Ecology 150: 53–63.

Tilman D. 1988. Plant strategies and the dynamics and structure of plant communities. Princeton University Press, New Jersey.

Tilman D. 1994. Competition and biodiversity in spatially structured habitats. Ecology 75: 2-16.

Tirado R, and FI Pugnaire. 2003. Shrub spatial association and consequences for reproductive success. Oecologia 136: 296–301.

Tirado R, and FI Pugnaire. 2005. Community structure and positive interactions in constraining environments. Oikos 111: 437–444.

Tongway DJ, and N Hindley. 1995. Assessment of soil condition of tropical grasslands. CSIRO Ecology and Wildlife, Canberra, Australia.

Tongway DJ, and JA Ludwig. 1996. Rehabilitation of semi-arid landscapes in Australia II: restoring productive soil patches. Restoration Ecology 4: 388-397.

Tongway DJ, AD Sparrow, and MH Friedel. 2003 Degradation and recovery processes in arid grazing lands of central Australia. Part 1: soil and land resources Journal of arid environments 55: 301-326.

Topp GC, and JL Davis. 1985. Measurement of soil water control using time domain reflectometry (TDR): a field application. Journal of Soil Science 49: 9–22.

Torices R. 2010. Adding time-calibrated branch lengths to the Asteraceae supertree. Journal of Systematics and Evolution 48: 271-278.

Tracol Y, JR Gutiérrez, and FA Squeo. 2010 Plant Area Index and microclimate underneath shrub species from a Chilean semiarid community. Journal of arid environments (In press).

Travis JMJ, RW Brooker, and C Dytham. 2005. The interplay of positive and negative species interactions along an environmental gradient: insights from an individual based model. Biology Letters 1: 5–8.

Trubat R, J Cortina, and A Vilagrosa. 2008. Short-term nitrogen deprivation increases field performance in nursery seedlings of Mediterranean woody species. Journal of Arid Environments 72: 879–890.

BIBLIOGRAFÍA

236

Valente LM, V Savolainen, and P Vargas. 2010. Unparalleled rates of species diversification in Europe. Proceedings of the Royal Society of London, series B 277:1489-1496.

Valiente-Banuet A, and E ezcurra. 1991. Shade as a cause of the association between the cactus Neobuxbaumia tetetzo and the nurse plant Mimosa luisana in the Tehuacán valley, mexico. Journal of Ecology 79: 961–970.

Valiente-Banuet A, and M. Verdú. 2007. Facilitation can increase the phylogenetic diversity of plant communities. Ecology Letters 10: 1029–1036.

Valiente-Banuet A, and M. Verdú. 2008. Temporal shifts from facilitation to competition occur between closely related taxa. Journal of Ecology 96: 489–494.

Valiente-Banuet A, A Vital, M Verdú, and RM Callaway. 2006. Modern Quaternary plant lineages promote diversity through facilitation of ancient Tertiary lineages. Proceedings of the National Academy of Sciences USA 103: 16812–16817.

Valiente-Banuet A, F Vite, and A Zavala-Hurtado. 1991. Interaction between the cactus Neobuxbaumia tetetzo and the nurse shrub Mimosa luisana. Journal of Vegetation Science 2: 11–14.

Valladares F. 2001. Características mediterráneas de la conversión fotosintética de la luz en biomasa: de órgano a organismo. pp: 67–94 In: Zamora R, and FI Pugnaire (Eds). Aspectos funcionales de los ecosistemas mediterráneos. CSIC-AEET, Granada.

Valladares F, I Aranda, and D Sánchez-Gómez. 2008. La luz como factor ecológico y evolutivo para las plantas y su interacción con el agua. pp 337–372 In: Valladares F. (Ed.). Ecología del bosque mediterráneo en un mundo cambiante (segunda edición). Ministerio de Medio Ambiente. Egraf, S.A., Madrid.

Valladares F, LG Hernández, I Dobarro, C García-Pérez, R Sanz, and FI Pugnaire. 2003. The ratio of leaf to total photosynthetic area influences shade survival and plastic response to light of green-stemmed leguminous shrub seedlings. Annals of Botany 91: 577–584.

Valladares F, and FI Pugnaire. 1999. Tradeoffs between irradiance capture and avoidance in semiarid environments assessed with a crown architecture model. Annals of Botany 83: 459–469.

Valladares F, I Dobarro, D Sánchez-Gómez, and RW Pearcy. 2005. Photoinhibition and drought in Mediterranean woody saplings:

scaling effects and interactions in sun and shade phenotypes. Journal of Experimental Botany 56: 483–494.

Valladares F, and E Gianoli. 2007. How much ecology do we need to know to restore Mediterranean ecosystems? Restoration Ecology 15: 363-368.

Valladares F, and RW Pearcy. 2002. Drought can be more critical in the shade than in the sun: a field study of carbon gain and photoinhibition in a Californian shrub during a dry El Niño year. Plant Cell and Environment 25: 749.

Valladares F, J Zaragoza-Castells, D Sánchez-Gómez, S Matesanz, B Alonso, A Portsmuth, A Delgado, and OK Atkin. 2008. Is shade beneficial for mediterranean shrubs experiencing periods of extreme drought and late-winter frosts? Annals of Botany 102: 923–933.

Vamosi SM, SB Heard, JC Vamosi, and CO Webb. 2009. Emerging patterns in the comparative analysis of phylogenetic community structure. Molecular Ecology 18: 572–592.

Vamosi JC, and SM Vamosi. 2010. Key innovations within a geographical context in flowering plants: towards resolving Darwin's abominable mystery. Ecology Letters 13: 1270–1279.

Van Auken OW. 2000. Shrub invasions of North American semiarid grasslands Annual Review of Ecology and Systematics 31: 197–215.

Van de Koppel J, M Rietkerk, and FJ Weissing. 1997. Catastrophic vegetation shifts and soil degradation in terrestrial grazing systems. Trends in Ecology and Evolution 12: 352-356.

Veblen KE. 2008. Season- and herbivore-dependent competition and facilitation in a semi-arid savanna. Ecology 89: 1532–1540.

Veblen T. 1989. Tree regeneration responses to gaps along a transandean gradient. Ecology 77: 541–543.

Verdú M, and P García-Fayos. 1996. Nucleation processes in a Mediterranean bird-dispersed plant. Functional Ecology 10: 275–280.

Verdú M, P Jordano, and A Valiente-Banuet. 2010. The phylogenetic structure of plant facilitation networks changes with competition. Journal of Ecology, In press.

Verdú M, and JG Pausas. 2007. Fire drives phylogenetic clustering in Mediterranean Basin woody plant communities. Journal of Ecology 95: 1316–1323.

Verdú M, PJ Rey, JM Alcántara, G Siles, and A Valiente-Banuet. 2009. Phylogenetic

237

signatures of facilitation and competition in successional communities. Journal of Ecology 97; 1171-1180.

Vesk PA, DI Warton, and M Westoby. 2004. Sprouting by semi-arid plants: testing a dichotomy and predictive traits. Oikos 107: 72–89.

Villar P, S Soliveres, and JL Quero. Introducción de especies leñosas. In. F Valladares et al. (Eds.). Ecología e infraestructuras lineales: bases científicas para soluciones técnicas. Ministerio de Medio Ambiente, Spain. In press.

Waide RB, MR Willig, CF Steiner, G Mittelbach, L Gough, SI Dodson, GP Juday, and R Parmenter. 1999. The relationship between productivity and species richness. Annual Review in Ecology and Systematics 30: 257–300.

Wang H, MJ Moore, PS Soltis, CD Bell, SF Brockington, R Alexandre, CC Davis, M Latvis, SR Manchester, and DE Soltis. 2010. Rosid radiation and the rapid rise of angiosperm-dominated forests. Proceedings of the National Academy of Sciences USA 106: 3853-3858.

Webb CO, DD Ackerly, M McPeek, and MJ Donoghue. 2002. Phylogenies and community ecology. Annual Review of Ecology and Systematics 33: 475–505.

Webb CO, DD Ackerly, and SW Kembel. 2008. Phylocom: software for the analysis of phylogenetic community structure and trait evolution. Bioinformatics 24: 2098–2100.

Webb CO, GS Gilbert, and MJ Donoghue. 2006. Phylodiversity dependent seedling

mortality, size structure, and disease in a Bornean rain forest. Ecology 87: S123–S131.

Westoby M. 1979/80. Elements in a theory of vegetation dynamics. Israel Journal of Botany 28: 169–194.

Westoby M, BH Walker, and I Noy-Meir. 1989. Opportunistic management for rangelands not at equilibrium. Journal of Rangeland Management 42: 266–274.

Whitford WG. 2002. Ecology of desert systems. Academic Press; California.

Whittaker RJ. 2010. Meta-analyses and mega-mistakes: calling time on meta-analysis of the species richness-productivity relationship. Ecology 91: 2522-2533.

Wikström N, V Savolainen, and MW Chase. 2001. Evolution of the angiosperms: calibrating the family tree. Proceeding of

the Royal Society of London, series B 268: 2211-2219.

Wise MJ, and WG Abrahamson. 2005. Beyond the compensatory continuum: environmental resource levels and plant tolerance of herbivory. Oikos 109: 417–428.

Wise MJ, and WG Abrahamson. 2007. Effects of resource availability on tolerance of herbivory: a review and assessment of three opposing models. The American Naturalist 169: 443–454.

Wojciechowski MF, M Lavin, and MJ Sanderson. 2004. A phylogeny of legumes (Leguminosae) based on analysis of the plastid matK gene resolves many well-supported subclades within the family. American Journal of Botany 91: 1846-1862.

Wolfe BE, JL Parrent, AM Koch, BA Sikes, M Gardes, and JN Klironomos. 2009 Spatial Heterogeneity in Mycorrhizal Populations and Communities: Scales and Mechanisms. In: Mycorrhizas - Functional Processes and Ecological Impact. Azcón-Aguilar C, et al. (Eds.). Springer-Verlag, Berlin.

Wright IJ, PB Reich, M Westoby, DD Ackerly, Z Baruch, F Bongers, J Cavender-Bares, FS Chapin, JHC Cornelissen, M Diemer, J Flexas, E Garnier, PK Groom, J Gulias, K Hikosaka, BB Lamont, T Lee, W Lee, C Lusk, JJ Midgley, ML Navas, Ü Niinemets, J Oleksyn, N Osada, H Poorter, P Poot, I Prior, VI Pyankov, C Roumet, SC Thomas, MG Tjoelker, E Veneklaas, and R Villar. 2004. The world-wide leaf economics spectrum. Nature 428: 821-827.

Zamora R, P García-Fayos, and L Gómez-Aparicio. 2008. Las interacciones planta-planta y planta animal en el contexto de la sucesión ecológica. pp 337–372 In: Valladares F. (Ed.). Ecología del bosque mediterráneo en un mundo cambiante (segunda edición). Ministerio de Medio Ambiente. Egraf, S.A., Madrid.

Zavaleta ES. 2006. Shrub establishment under experimental global changes in a California grassland. Plant Ecology 184: 53–63.

Zimmermann J, S Higgins, V Grimm, J Hoffmann, and A Linstädter. 2010. Grass mortality in semi-arid savanna: The role of fire, competition and self-shading. Perspectives in Plant Ecology, Evolution and Systematics 12: 1–8.

COAUTORES

238

AFILIACIÓN DE LOS COAUTORES

Fernando T. Maestre Gil

Área de Biodiversidad y Conservación, Departamento de Biología y Geología, Escuela

Superior de Ciencias Experimentales y Tecnología, Universidad Rey Juan Carlos, 28933

Móstoles, Spain. E-mail: fernando.maestre@urjc.es

Adrián Escudero Alcántara

Área de Biodiversidad y Conservación, Departamento de Biología y Geología, Escuela

Superior de Ciencias Experimentales y Tecnología, Universidad Rey Juan Carlos, 28933

Móstoles, Spain. E-mail: adrian.escudero@urjc.es

Fernando Valladares Ros

Instituto de Recursos Naturales, Centro de Ciencias Medioambientales, C.S.I.C., Serrano 115,

E-28006 Madrid, Spain. E-mail: valladares@ccma.csic.es

Pablo García-Palacios

1) Área de Biodiversidad y Conservación, Departamento de Biología y Geología, Escuela

Superior de Ciencias Experimentales y Tecnología, Universidad Rey Juan Carlos, 28933

Móstoles, Spain.

2) Instituto de Recursos Naturales, Centro de Ciencias Medioambientales, C.S.I.C., Serrano

115, E-28006 Madrid, Spain. E-mail: pablo.palacios@urjc.es

Andrea P. Castillo Monroy

Área de Biodiversidad y Conservación, Departamento de Biología y Geología, Escuela

Superior de Ciencias Experimentales y Tecnología, Universidad Rey Juan Carlos, 28933

Móstoles, Spain. E-mail: teamatea@yahoo.es

239

Lucia DeSoto Suárez

Centro de Ecogía Funcional, Departamento de Ciencias da Vida. Universidade de Coimbra

3001 – 455 Coimbra. Portugal. E-mail: luciadesoto@gmail.com

Jose Miguel Olano

Área de Botánica, Departamento de Ciencias Agroforestales, E.U.I. Agrarias de Soria,

Universidad de Valladolid, Campus de los Pajaritos, 42004 Soria, Spain. E-mail:

jmolano@agro.uva.es

David Eldridge

Department of Environment, Climate Change and Water, Evolution and Ecology Research

Centre, School of Biological, Earth and Environmental Sciences, University of New South

Wales, Sydney, NSW 2052, Australia. E-mail: d.eldridge@unsw.edu.au

Matthew A. Bowker

Colorado Plateau research Station, US Geological Survey, Flagstaff (USA). E-mail:

matthew.bowker@nau.edu

Matthew K. Tighe

Ecosystem Management, School of Environmental and Rural Science, University of New

England, Armidale, NSW 2351, Australia. E-mail: mtighe2@une.edu.au

Rubén Torices Blanco

Área de Botánica, Departamento de Ciencias Agroforestales, E.U.I. Agrarias de Soria,

Universidad de Valladolid, Campus de los Pajaritos, 42004 Soria, Spain. E-mail:

rubentorices@gmail.com

COAUTORES

240

Diseño portada e imágenes:

Joan Miquel Fuster Mollà y Santiago Soliveres Codina

Maquetación:

Soraya Constán Nava y Santiago Soliveres Codina