Comunidades de crustáceos decápodos asociados a fondos superficiales de fanerógamas marinas y su relación con biotopos colindantes en el sur de España

AUTOR: Ángel Mateo Ramírez http://orcid.org/0000-0002-3825-3279 EDITA: Publicaciones y Divulgación Científica. Universidad de Málaga Esta obra está bajo una licencia de Creative Commons Reconocimiento-NoComercial-SinObraDerivada 4.0 Internacional: Cualquier parte de esta obra se puede reproducir sin autorización pero con el reconocimiento y atribución de los autores. No se puede hacer uso comercial de la obra y no se puede alterar, transformar o hacer obras derivadas. http://creativecommons.org/licenses/by-nc-nd/4.0/legalcode Esta Tesis Doctoral está depositada en el Repositorio Institucional de la Universidad de Málaga (RIUMA): riuma.uma.es D. JOSE ENRIQUE GARCÍA RASO, Profesor Catedrático del Departamento de Biología Animal de la Facultad de Ciencias de la Universidad de Málaga. CERTIFICA: Que la presente memoria titulada “Comunidades de crustáceos decápodos asociados a fondos superficiales de fanerógamas marinas y su relación con biotopos colindantes en el sur de España” presentada por el Licenciado en Ciencias Biológicas Ángel Mateo Ramírez, ha sido realizada bajo mi dirección y las publicaciones que la avalan no han sido utilizadas en tesis anteriores. Como director de tesis, considerando que la presente memoria reúne todos los requisitos para ser sometida a juicio de la Comisión correspondiente, por lo que autorizo su exposición y defensa para la obtención del grado de Doctor en Ciencias Biológicas con mención Europea. Y para que así conste, en cumplimiento de las disposiciones vigentes, firmo la presente acreditación en Málaga, a 09 de noviembre de 2015. Fdo.: Dr. José Enrique García Raso Visado en Málaga a 09 de noviembre de 2015 El director Fdo.: Dr. José Enrique García Raso Prof. Catedrático del departamento de Biología Animal Universidad de Málaga Memoria presentada para optar al grado de Doctor en Biología Fdo.: Ángel Mateo Ramírez A mis tres Marías, mi padre y hermana Familia y amigos “Muchas personas atacan al mar, yo le hago el amor”. Jacques Cousteau Agradeciminetos Enfrentarse a una tesis es duro, y más sin la ayuda de una beca de formación, ya que a los momentos duros a los que todos los doctorandos nos enfrentamos, se le suma el buscar medios para poder sustentarte. Una situación que es difícil de entender y que solo puede mantenerse por la devoción por lo que haces y por el apoyo de todas aquellas personas que han estado cerca ti. Han sido muchas personas las que he conocido en estos ocho años y de las que he aprendido mucho tanto a nivel profesional como personal. José Enrique García Raso, mi director de tesis y director del departamento, que me ha enseñado todo lo que se sobre decápodos y que ha estado ahí siempre que lo he necesitado, apoyándome y echándome una mano en todo. Carmen Salas y Sergio Gofas, que nos han ayudado en los muestreos, tanto buceando directamente como prestándonos su casa como “campamento base”. Así como buscando contratos y medios para ayudarme en la cruzada de la tesis. Todos ellos, grandes investigadores y personas, que me han enseñado todo lo que se sobre la fauna y flora marina, y de los que espero seguir aprendiendo mucho más durante mucho tiempo. Otros profesores y compañeros del departamento como Jesús Olivero (por resolverme las dudas estadísticas y por los ratos de charla), Ana Luz, Farfan, Antonio Román, Mario, Borja, Ana Carmen…por el día a día y por todo lo que en algún momento puede aprender, pues se puede hacer ciencia tanto en los pasillos como en un despacho. A los profesores de la Universidad de Alicante, donde realice junto con Javi y Pablo, los cursos de Doctorado. En especial a Jose Luis Sánchez Lizaso director de mi trabajo de Investigación Tutelada, Pablo Sánchez-Jerez y Alfonso Ramos Esplá. A todos los amigos que hicimos en Alicante, en especial Elena, María e Inés, por todos esos momentos inolvidables, que han hecho que hoy día sigamos siendo amigos. A Maite, a la cual conocí durante estos cursos, y tuve la suerte de trabajar y convivir con ella en Palma de Mallorca durante el año y medio que Introducción General 3 1.1 Praderas de fanerógamas marinas y fondos de algas infralitorales en el mar de Alborán De las alrededor de las 60 especies de fanerógamas marinas existentes en todo el mundo (Short y Coles 2001), cuatro son las especies presentes en el mar de Alborán: Posidonia oceanica (Linnaeus) Delile, Cymodocea nodosa (Ucria) Ascherson, Zostera noltei Hornemana, y Zostera marina Linnaeus. Las praderas del endemismo mediterráneo P. oceanica son las más extensas tanto en el mar Mediterráneo como en Alborán (Figura 1). Dentro de Alborán la mayoría de las praderas y las que mejor estado de conservación presentan, se encuentran en la provincia de Almería, representando un 82% del total de la superficie ocupada en toda Andalucía, unas 6139,56 ha (Junta Andalucía, 2013). En Granada las praderas se concentran en el parte oriental, entre Cala Chilches y Castel de Ferro, presentando un grado de conservación aceptable. En Málaga las praderas aparecen como parches discontinuos con formas y tamaños variables catalogándose como semipraderas. Estas semipraderas aparecen de forma irregular y dispersa sobre sustrato rocoso principalmente y se concentrán en tres sectores: sector oriental, en el tramo comprendido entre Molino de Papel (Paraje Natural de Maro y cerro Gordo)-Nerja; sector central, en el tramo comprendido entre Punta Calaburras-Calahonda; y sector occidental, Estepona y Punta Chullera, donde P. oceanica encuentre su límite occidental de distribución. La influencia atlántica y el aumento de la turbidez de las aguas hace que tanto los valores de densidad global (teniendo en cuenta la densidad de haces y la cobertura de la pradera), como el límite inferior de las praderas, disminuyan a medida que nos acercamos al estrecho de Gibraltar, siendo en Málaga, en donde las praderas presentan las densidades globales más bajas. Igualmente el límite inferior sube desde los 30m de profundidad del levante Almeriense, hasta unos 6 m en la zona de Punta Calaburras-Calahonda (Tintoré et al. 1988; Moreno y Guirado 2003; Luque y Templado 2004; García Raso et al. 2010; Junta de Andalucía 2013). La floración de P. ocenica es un hecho relativamente esporádico, en el cual, la temperatura del agua parece tener un papel importante, pues floraciones masivas se ha detectado tras veranos excesivamente calurosos (Díaz-Alamela et al. 2006; Urra et al. 2011b). De hecho, en 2009, se detecto una floración masiva a lo largo del litoral Capítulo 1 4 andaluz, registrándose unas densidades de flores de entre 40 flores/m2 en las praderas de Melicena (Granada), 50-200 flores/m2 en las del Paraje Natural de Maro y Cerro Gordo (Málaga) y 4-184 flores/m2 en las del LIC de Calahonda (Málaga) (Urra et al. 2011b; Junta de Andalucía 2013). En Almería, no obstante se han registrado floraciones durante tres años consecutivos (2001-2003), si bien, las densidades de flores fueron mucho menores (Moreno y Guirado 2006). La segunda especie en extensión es C. nodosa, especie de aguas cálidas (PérezLlorents et al. 2014), la cual presenta una extensión de 119,38 ha en toda Andalucía, de las cuales 100,43 ha corresponden al mar de Alboran (Almería principalmente) y 18,93 ha a la costa atlántica de Andalucía (Cádiz). Esta especie es la que presenta una distribución más continua a lo largo del litoral andaluz (Figura 2), formando tanto praderas mixtas con P. oceanica y Z. noltei como praderas monespecíficas. No obstante ha sufrido un elevado grado de regresión durante los últimos años desapareciendo praderas en las provincias de Málaga y Granada (Moreno y Guirado 2003; Junta de Andalucía 2013). Figura 1. Distribución de Posidonia oceanica en el litoral andaluz a partir de las observaciones realizados por el equipo de medio marino del Programa de Gestión Sostenible del Medio Marino Andaluz de la Consejería de Medio Ambiente y Ordenación del Territorio de la Junta de Andalucía entre para el periodo 2004-2013 (Modificada del Informe de Programa de Gestión Sostenible del Medio Marino 2013). Introducción General 5 Las dos últimas especies son las menos abundantes. Z. noltei es una especie templada que habita tanto aguas atlánticas (desde Noruega hasta Mauritania) como aguas mediterráneas, siendo especialmente abundante en los estuarios de la vertiente atlántica de Andalucía. Dentro de Alborán solo se encuentra en ciertas zonas someras del litoral almeriense y Algeciras (Phillips y Meñez 1988; Luque y Templado 2004) (Figura 3). Por otro lado, Z. marina, que es una especie circumboreal, que encuentra su límite de distribución meridional en una zona cercana a Gibraltar (Philips y Meñez 1988; Luque y Templado 2004; Perez-Llorents et al. 2014). La fuerte influencia atlántica que presenta el mar de Alborán hace que su abundancia sea mayor en él que en el resto del mar Mediterráneo (Rueda et al. 2008b). Sin embargo esta especie, es la que ha sufrido una regresión más severa, principalmente en Granada y Málaga, donde se considera desaparecida, pudiendose constatar su presencia solamente en el saco interno de la bahía de Cádiz (Junta de Andalucía 2013) (Figura 4). Figura 2. Observaciones de Cymodocea nodosa a lo largo del litoral Andaluz a partir de los datos obtenidos por parte del Programa de Gestión Sostenible del Medio Marino Andaluz. En verde se señalan las observaciones correspondientes a la presencia actual de biocenosis de esta especie. En blanco se señalan las biocenosis que por diversos motivos han regresionado y actualmente se consideran desaparecidas (Modificada del Informe de Programa de Gestión Sostenible del Medio Marino 2013). Capítulo 1 6 Figura 3. Observaciones de Zostera noltei a lo largo del litoral Andaluz a partir de los datos obtenidos por parte del Programa de Gestión Sostenible del Medio Marino Andaluz. En amarillo se señalan las observaciones correspondientes a la presencia actual de biocenosis de esta especie. En blanco se señalan las biocenosis que han regresionado y actualmente se consideran desaparecidas (Modificada del Informe de Programa de Gestión Sostenible del Medio Marino 2013). Figura 4. Observaciones de Zostera marina a lo largo del litoral Andaluz a partir de los datos obtenidos por parte del Programa de Gestión Sostenible del Medio Marino Andaluz. En verde se señalan las observaciones correspondientes a la presencia actual de esta especie. En blanco se señalan las biocenosis que por diversos motivos han sufrido una regresión y actualmente se consideran desaparecidas (provincias de Málaga y Granada) o que no se puede asegurar su presencia como ocurre en la provincia de Almería (Modificada del Informe de Programa de Gestión Sostenible del Medio Marino 2013). Introducción General 7 Por otra parte, las comunidades de algas fotófilas más característica del Mediterráneo occidental son aquellas dominadas por especies pertenecientes al género Cystoseira. Estas comunidades son muy diversas y están constituidas por diferentes estratos.1) formas incrustantes y pulviniformes (Valonia, Peyssonnelia, Mesophyllum,…); 2) formas cespitosas (Halimeda, Cladophora,..); 3) formas erectas bajas o “arbustivas” (Diginea, Dictyopteris,…) y formas erectas altas o “arbóreas” (Cystoseira spp) (Figura 5). Sin embargo, en Alborán, comunidades bien estructuradas con dominancia de especies de Cystoseira solo aparecen en Cabo de Gata. A medida que nos acercamos al estrecho de Gibraltar las comunidades de macrófitos van cambiando, aumentando la presencia y la abundancia de especies atlánticas. En las provincias de Málaga y Cádiz son muy abundantes las comunidades de la rodófita invasora Asparagopsis armata Harvey, la cual puede llegar a formar un ancho cinturón, desplazando a numerosas especies Figura 5. Esquema de la disposición en varios estratos de la vegetación algal de una comunidad de Cystoseira infralitoral. I, formas incrustantes y pulviniforme II, formas cespitosas. III, formas erectas bajas “arbustivas”. IV, formas erectas altas “arbóreas”. A, el estrato arbóreo está muy poco desarrollado o ausente en las comunidades de algas del ZEC”Calahonda”( Modificada de Margalef 1989). Capítulo 1 8 de algas autóctonas, tanto de zonas iluminadas como parcialmente umbrías. Además se observa una simplificación y miniaturización de la comunidad, como consecuencia de la sustitución del estrato arbóreo por el arbustivo, el cual se encuentra dominado por especies como la feofita Halopteris scoparia (Linnaeus) Sauvageau (Conde y Seoane 1982; Conde 1989; Margalef 1989; Flores-Moya 1989; Ballesteros 1993; Flores-Moya et al. 1995; Conde et al. 1996; Luque y Templado 2004; Cobos y Ortega 2010). 1.2 Pérdida de hábitat y sus efectos sobre las asociaciones de invertebrados El mar Mediterráneo, a pesar de representar solamente un 0,8% del volumen total de todos los océanos del mundo, está considerado como un punto caliente u “hot spot” de diversidad marina, presentando un 7% del total de las especies animales y vegetales conocidas actualmente, de las cuales, el 20,2% son endemismos mediterráneos. Las últimas estimaciones indican que en el mar Mediterráneo viven unas 17,000 especies de organismo, siendo los invertebrados los más numerosos con 11,000 especies. Sin embargo esta riqueza de especies no presenta una distribución homogénea, siendo la cuenca noroccidental más rica que la suroriental (Boudouresque 2004; Coll et al. 2010). De igual manera las aguas de la plataforma continental, por encima de los 200m, son las que acumulan la mayoría de la biodiversidad, especialmente la zona batimétrica comprendida por encima de los 50m (Fredj et al. 1992; Boudouresque 2004; Coll et al. 2010). Esta es la zona donde habitan las fanerógamas marinas y algas fotófilas, no obstante, la desaparición de estas marca el límite inferior del piso infralitoral (Pérès y Picard 1964). Desgraciadamente la mayoría de las presiones antropogénicas se concentran principalmente en este piso, impactando sobre las praderas de fanerógamas marinas y los fondos de algas fotófilas. Actividades como el desarrollo costero (construcción de puertos, espigones, etc.), la pesca ilegal con barcos de arrastre a profundidades menores de los 50m o el fondeo de embarcaciones deportivas tienen un efecto directo sobre las praderas de macrófitos y angiospermas marinas, al arrancarlas o eliminar el hábitat donde se desarrollan, ej. los sustratos duros sobre los que crecen las algas. Otras actividades Introducción General 9 tienen un efecto indirecto, en relación con la calidad del agua. La resuspensión de sedimentos generada por las obras costeras o la regeneración de playas, así como la eutrofización producida por la polución o por los aportes de materia orgánica provenientes de las industrias, emisarios o piscifactorias, traen consigo una reducción en la claridad del agua, disminuyendo la cantidad luz que llega al fondo. Además, las sustancias tóxicas o metales pesados que puedan venir en estos vertidos tienen también un efecto directo sobre la flora y fauna bentónica (Walker y Kendrick 1998; Delgado et al. 1999; Duarte 2002; Boudouresque 2006a; González-Correa 2008; Rueda et al. 2008a; Kiparissis et al. 2010). Otras amenazas relacionadas con actividades antropogénicas son las especies invasoras y el cambio climático, no obstante Boudouresque (2006a) cita que el 6,5% de la flora presente en el Mediterráneo tienen un origen haloctono y Marbà y Duarte (2010) encontraron una correlación positiva entre el calentamiento del Mediterráneo y el aumento de la mortalidad de haces en P. oceanica. Todas estas amenazas y presiones traen consigo una degradación de las praderas de fanerógamas marinas y fondos de macroalgas, pudiendo llegar a provocar la desaparición las mismas. Existen diferentes ejemplos de regresiones de praderas de P. oceanica en el Mediterráneo relacionadas principalmente con la polución, como en el Golfo de Trieste o en zonas cercanas del Lago de Venecia, con el desarrollo costero, como ocurrió en la Riviera Francesa o por una combinación de diferentes factores (pesca, regeneración de playas, acuicultura…) como se ha documentado en España, donde el 58% de las praderas de P. oceanica y el 52% de las fanerógamas de la costas catalanas y cercanas a Alicante respectivamente, han regresionado en las últimas décadas (Meinesz et al. 1991; Caressa et al. 1995; Rismondo et al. 1997; Airoldi y Beck 2007). En el caso de las praderas macroalgas (en referencia a fucales como Cystoseira spp. y Sargassum spp.) también hay ejemplos de regresiones locales en el Mediterráneo, tanto en España, Francia, Croacia, Italia como en Rumania (Sfriso 1987; Munda 1993; Rodríguez-Prieto y Polo 1996, Thibaut et al. 2005; Zaitsev 2006). Las causas son varias pero principalmente incluyen un aumento desproporcionado en las poblaciones de erizos y una disminución en la calidad del agua, como consecuencia de la polución y/o la sedimentación (Airoldi y Beck 2007). Con frecuencia, estas regresiones son mayores en las Capítulo 1 10 zonas más profundas, de manera que lo que se produce es una disminución en su rango batimétrico, convirtiéndolas en fondos de algas más someros. En otras ocasiones estas praderas son sustituidas por lo que se denomina en inglés “urchin barren/barren grounds”, la cual suele estar dominada por especies de coralinaceas, bancos de mejillones, macrófitos efímeros y erizos (Munda 1993; Boudouresque y Verlaque 2001; Guidetti et al. 2003; Thibaut et al. 2005). Esta degradación del hábitat suelen traer consigo una pérdida de diversidad, no solo a nivel de abundancia o riqueza específica, sino también en relación a la diversidad funcional (en relación con atributos funcionales como: el tamaño o la categoría trófica de las especies) y a la diversidad espacial o entre hábitats (ß diversidad) (en Arioldi y Beck 2007) (Figura 6). Las casusas de esta pérdida de diversidad son: 1) la desaparición de especies propias y el aumento de especies generalistas, lo que también se conoce como “homogenización biótica”, 2) la pérdida de fuentes de alimentos ,ya que por ejemplo, las praderas de fanerógamas marinas o los bosques de macroalgas, así como todos los animales asociados a estos hábitats crean un hábitat más complejo con muchos más recursos, parte de los cuales desaparecerían a consecuencia de la degradación del hábitat y por último, 3) la pérdida de funciones y propiedades ecosistemicas, en relación con la influencia de la complejidad estructural del hábitat sobre el medio, como por ejemplo la modificación del hidrodinamismo, de las condiciones lumínicas o de la sedimentación, así como de su función protectora (Mckinney y Lockwood 1999; Airoldi et al. 2008) (Figura 6). Por otro lado hay trabajos que demuestran que hábitats fragmentados formados por diferentes biotopos dan lugar a un hábitat más complejo y diverso (Boström et al. 2006; Mateo-Ramírez y García Raso 2012; Urra et al. 2013a). Algo similar encontraron Dean y Conell (1987 a, b, c) en comunidades de macroinvertebrados asociados a algas, cuyas máximas abundancias y riqueza de especies se obtuvieron durante las etapas intermedias de la sucesión, en relación con la mayor diversidad de algas y su mayor complejidad (biomasa y área). Esto puede estar relacionado con los efectos de la fragmentación de un hábitat sobre la fauna asociada, los cuales están a su vez condicionados por diferentes Introducción General 11 factores como 1) el cambio en la configuración del hábitat, 2) la distancia entre esos parches, 3) su tamaño y 4) la diversidad de hábitats presentes entre los diferentes parches (Farhrig, 2003). Boström et al. (2006), hicieron una revisión bibliográfica sobre las praderas de fanerógamas y su relación con la fauna asociada, y encontraron que la fragmentación de las fanerógamas no tiene por qué tener un efecto negativo sobre los animales asociados. Por ejemplo si esa fragmentación Figura 6. Diagrama mostrando la conexión entre la pérdida de hábitat y los patrones de diversidad, posibles feedbacks entre los diferentes procesos y el resultado global de la homogenización biótica. (Tomado de Airoldi et al. 2008). Capítulo 1 12 es consecuencia de la presencia de una especie de alga invasora o por adición de estructuras antropogenicas, estos nuevos hábitats pueden proporcionar beneficos para ciertas especies, pues ambos son hábitats estructurales (Boström et al. 2006). Por otro lado esa fragmentación del hábitat puede beneficiar a ciertas especies al disminuir la presión depredadora sobre ellas (ya que la búsqueda de alimento es más difícil en una pradera fragmentada, pues el depredador necesita moverse más y seleccionar el parche adecuado, lo que a su vez lo expone a otros depredadores) (Micheli y Peterson 1999; Hovel y Lipcius 2001) o al permitir la llegada de especies foráneas (Irlandi 1994). No obstante, la fragmentación del hábitat es algo a evitar pues es una amenaza para la conservación de la biodiversidad, tanto en los ecosistemas marinos como en los terrestres, ya que los hace más vulnerables (Bell et al. 2001; Tscharntke et al. 2002). 1.3 Decápodos y conectividad entre hábitats Los decápodos son animales que poseen una gran capacidad de desplazamiento, no solo por su capacidad locomotora sino también por su fase plantónica. Estás características les permite poder desplazarse entre diferentes hábitats y/o zonas, que pueden llegar a estar separadas por grandes distancias, estableciendo una conexión poblacional entre ellas (Chiswell et al. 2003). El ejemplo más extremo lo representan las especies pertenecientes al infraorden Achelata. Todas estas especies se caracterizan por la presencia de una etapa larvaria denominada filosoma, la cual presenta unas características (cuerpo transparente, dorsoventralmente aplanado y apéndices con muchas setas) que les permiten permanecer durante un largo periodo de tiempo en el plancton. Se ha constatado que filosomas de ciertas especies de langosta como Jasus edwardsii (Hutton, 1875), pueden persistir en el plancton hasta 24 meses (Booth 1994; von der Heyden et al. 2007). Esto combinado con las corrientes predominantes puede llegar a transportar a las filisomas de esta especie miles de kilómetros, estableciendo una conexión poblacional (Chiswell et al. 2003). Además ciertas especies realizan migraciones entre las zonas de cría y los hábitats donde residen los adultos, como es el caso de la langosta del Caribe, Panulirus argus (Latreille, 1804), Panulirus ornatus (Fabricius, 1798) o Panulirus cygnus George, Introducción General 19 y P. Calaburras (Cano y García La Fuente 1991) (Figura 9). La presencia o no de vientos del este (levante), determina la posición del “Atlantic Jet” (Macías et al. 2007). En ausencia de levante dicho chorro tiende a desplazarse hacia el norte, acercándose más a las costas de Andalucía, esto tiene un efecto directo sobre dos ramales que se desprenden del “Atlantic Jet” los denominados “núcleo septentrional”. Estos ramales se encuentran a unos 18Km de Marbella y discurren hacia P. Calaburras, alcanzando velocidades bastantes elevadas (40 m/s). El desplazamiento hacia el norte del “Atlantic Jet” provoca a su vez un acercamiento del “nucleo septentrional” a la costa andaluza, aumentando su velocidad e influencia sobre el tramo costero antes mencionado. Este fenómeno además se ve amplificado bajo condiciones meteorológicas en las que dominen los vientos de componente oeste (poniente) (Conde y Seoane 1982; Cano y García La Fuente 1991; Macías et al. 2007). En el mar de Alboran son abundantes los afloramientos de aguas profundas ricas en nutrientes. Estos afloramientos se extienden por la zona de Estepona, Figura 9. Esquema conceptual de la entrada del chorro atlántico (flecha negra) y los ramales que se desprenden de él “núcleo septentrional” (flechas negras discontinuas) (Modificado de Macías et al. 2008) Capítulo 1 20 Marbella (cercanas a la zona de muestreo), Málaga, Motril y Almería, así como, en el frente “Almería-Orán” (Tintoré et al. 1988; Cano y García La Fuente 1991; Macías et al. 2007). Los orígenes de estos afloramientos son diversos y pueden estar producidos por diferentes fenómenos, como 1) la fricción generada entre las aguas del “chorro atlántico” y las aguas mediterráneas (lo que se ve favorecido por la ausencia de vientos de levante) (Tintoré et al. 1991; Macías et al. 2007), 2) oscilaciónes norte-sur del “chorro atlántico”, 3) el transporte de Ekman en relación con los vientos de poniente (Sarhan et al. 2000) y 4) la topografía del fondo, debido a la presencia de isobatas curvadas y divergentes sobre la plataforma o cañones submarinos (Rodríguez et al. 2006). En el litoral comprendido entre P. Calaburras y Calahonda las variables ambientales relacionadas con la columna de agua varían estacionalmente, alcanzando los valores más altos principalmente en verano (temperatura superficial del agua de mar (T): 27ºC; irradiancia solar (IS): ca. 4 log lm m-2; concentración de clorofila a (Chl a): ca. 15 μg l-1) y los más bajos en invierno (T: 13,5ºC; IS: ca. 3,5 log lm m-2; Chl a: ca. 2 μg l-1) (datos personales). Por su parte, la salinidad permanece más o menos constante a lo largo del año (en torno a 36,7‰), con valores algo menores que en el resto del Mediterráneo debido a la influencia del agua atlántica (http://www.ma.ieo.es/gcc/sistemas_observaciones.htm). 1.4.4 Biogeografía En el litoral andaluz coinciden tres ecorregiones: el mar de Alborán, dentro de la provincia Mediterránea, la plataforma atlántica europea del Sur y los Afloramientos Saharauis, ambas dentro de la provincia Lusitánica (Spalding et al. 2007) (Figura 10). Esta variabilidad biogeográfica hace que en el mar de Alborán podamos encontrar especies con diferentes afinidades. 1) Especies atlánticas, como el hidrozoo Diphasia margareta (Hassall, 1841), la gorgonia Eunicella verrucosa (Pallas, 1766), los moluscos gasterópodos Rissoa guerinii Récluz, 1843 y Nassarius reticulatus (Linnaeus, 1758), los bivalvos Mytilus galloprovincialis y Cardita calyculata (Linnaeus, 1758) o el crustáceo decápodo Anapagurus hyndmanni Introducción General 21 (Bell, 1846), 2) especies mediterráneas, como el crinoideo Antedon mediterranea (Lamarck, 1816), la gorgonia Eunicella singularis (Esper, 1791), los gasterópodos Alvania lineata Risso, 1826 y Euspira macilenta (Philippi, 1844), el bivalvo Pinna nobilis Linnaeus, 1758 o el decápodo Pagurus chevreuxi (Bouvier, 1896), 3) especies subtropicales del norte de África, como las gorgonias Ellisella paraplexauroides (Stiasny, 1936), Eunicella labiata Thomson, 1927 y Eunicella filiformis (Studer, 1879) , los gasterópodos Mathilda quadricarinata (Brocchi, 1814) y Cymbula safiana (Lamarck, 1819), los bivalvos Modiolus lulat (Dautzenberg, 1891) y Ungulina cuneata (Spengler, 1798), los decápodos Pagurus mbizi (Forest, 1955) o Cryptosoma cristatum Brullé, 1837 o peces como la hurta (Pagrus auriga Valenciennes, 1843) o el poyo (Scorpaena maderensis Valenciennes, 1833) con poblaciones estables en Alborán. Por último mencionar aquellas especies cuyas poblaciones se concentran en el estrecho de Gibraltar y en sus proximidades, como los gasterópodos Aclis verduini van Aartsen, Menkhorst & Gittenberger, 1984, Alvania vermaasi van Aartsen, 1975 o Jujubinus dispar Curini-Galletti, 1982 (Pérès y Picard 1964; García Raso 1993; Gofas 1999; Rueda et al. 2010; Urra et al. 2013a,b; García Raso et al. 2013). Figura 10. Mapa de las provincias biogeográficas según el tratado clásico de Ekman (1953). En los últimos trabajos la provincia Mauritánica se encuadra dentro de la provincia Lusitánica (línea punteada) (Spalding et al. 2007). Capítulo 1 22 La influencia atlántica se hace notar también en las especies de macroalgas. Así en las cercanías de P. Calaburras especies atlánticas como Cystoseira mauritanica Sauvageau, Fucus spiralis Linnaeus, Saccorhiza polyschides (Lightfoot) Batters o Cystoseira usneoides (Linnaeus) M. Roberts, encuentran aquí su límite de distribución dentro del Mediterráneo (Conde y Seoane 1982; Conde 1989; Flores-Moya et al. 1995; Barceló-Martí et al. 2000; Flores-Moya 2012). A medida que nos acercamos a la parte oriental del litoral malagueño (Torre del Mar y Nerja) la comunidad macrofitobentónica va cambiando, disminuyendo el porcentaje de especies atlánticas y aumentando la presencia de especies Mediterráneas endémicas como Rissoella verruculosa (Bertoloni) J.Agardh, o especies pantropicales como Acetabularia acetabulum (Linnaeus) P.C.Silva, , Halimeda tuna (J.Ellis & Solander) J.V.Lamouroux, o Digenea simplex (Wulfen) C.Agardh, que no se encuentran hasta el cabo de Gata (Conde y Seoane 1982; Conde 1989). Esta diversidad de especies con diferentes afinidades hace que el mar de Alborán albergue la mayor diversidad biológica de todos los mares de Europa (Luque y Templado 2004; García Raso et al. 2010). Como ejemplo comentar, que solo dentro del ZEC de “Calahonda” y en un estudio preliminar ya se contabilizaron 355 especies de moluscos y 88 especies de decápodos, las cuales representan casi un 17 % y un 28 % del todas las especies de moluscos y decápodos presentes en todo el Mediterráneo y en el extremo occidental del mismo respectivamente (García Múñoz et al. 2008; Coll et al. 2010; Urra et al. 2011a,b; 2012a,b; 2013a,b; estudio presente). Introducción General 23 1.5.1 Marco de la Tesis Doctoral Hoy en día tanto las praderas de fanerógamas como los fondos de macroalgas están sufriendo una regresión a nivel mundial, que aunque puede deberse a causas naturales, está principalmente relacionada con actividades antrópicas (Walker y Kendrick 1998; Boudouresque et al. 2012). Actualmente, directivas como la Directiva Habitat, abogan por la protección y conservación de los hábitats y las poblaciones de las especies silvestres de la Unión Europea, mediante el establecimiento de una red ecológica, denominada Red Natura 2000, y de un régimen jurídico de protección de las especies. La Directiva Habitat, declara a las praderas de Posidonia oceanica como un hábitat prioritario para su conservación. La presencia de estas praderas en el ZEC ”Calahonda” fue primordial para la protección de esta zona. No obstante, la presencia de praderas de Cymodocea nodosa y de abundantes afloramientos rocosos (escasos en el litoral malagueño) cubiertos de macroalgas, así como de las especies asociadas a estos hábitats, ayudaron a la Figura 11. Visión general de los fondos mixtos formados por praderas fragmentadas de Posidonia oceanica, Cymodocea nodosa y fondos de macroalgas fotófilas dominados por Halopteris scoparia, presentes en el ZEC “Calahonda”. Capítulo 1 24 designación de esta zona como Zona Especial de Conservación (BOJA 153 de 07/08/2015). La presente tesis surgió como parte de los estudios que se realizaron por parte del grupo de investigación RNM 0141 del departamento de Biología Animal de la Universidad de Málaga, a partir de los cuales se constataron la presencia de dichos hábitats y especies asociadas. El objetivo principal de esta tesis es el estudio de las asociaciones de decápodos ligadas a los hábitats vegetales presentes en el ZEC ”Calahonda”. De igual manera y aprovechando los datos fenológicos y de densidad de haces obtenidos tanto para P. oceanica como para C. nodosa durante el año de muestreo faunístico (julio 2007 - mayo 2008), se decidió continuar tomando datos hasta marzo de 2010, para así poder tener un período de tiempo suficiente parar analizar la dinámica estacional de estas fanerógamas, en esta zona de transición entre aguas atlánticas y mediterráneas. 1.5.2 Objetivos En esta Tesis Docotoral se ha profundizado en el conocimiento de la fauna de decápodos presentes en la Zona Especial de Consevación (ZEC) “Calahonda” (Málaga). Los objetivos se han centrado en el estudio y análisis de la composición, estructura y dinámica temporal de las asociaciones de decápodos ligadas tanto a praderas de Posidonia oceanica y Cymodocea nodosa como a fondos de macroalgas fotófilas. Además de estudiar las relaciones existentes entre todas ellas. Los objetivos específicos que se abordaron fueron: 1. Estudiar la influencia de las características oceanográficas del ZEC ”Calahonda” sobre la fenología y el estado de conservación de las praderas de Posidonia oceanica y Cymodocea nodosa, analizando tanto la dinámica estacional e interanual de sus parámetros fenológicos y número de haces, como su relación con las variables ambientales (capítulo 2). 2. Estudiar la composición y estructura de las asociaciones de crustáceos decápodos ligadas a praderas de Cymodocea nodosa, evaluando sus cambios estacionales y su relación con las variables ambientales, sedimentológicas, fenología de la planta y el tamaño de los “parches” estudiados (capítulo 3). Introducción General 25 3. Caracterizar las asociaciones de crustáceos decápodos ligadas a praderas de Posidonia oceanica, y su dinámica temporal en relación tanto con las variables ambientales, hidrológicas y fenológicas (capítulo 4). 4. Analizar la estructura y composición de las asociaciones de crustáceos decápodos ligados a los fondos de macroalgas fotófilas dominadas principalmente por Halopteris scoparia, caracterizando tanto las diferencias entre los estratos (fronde-sedimento), como las diferencias estacionales y su relación con las variables ambientales, hidrológicas y los parámetros de las algas (altura, volumen, peso seco) (capítulo 5). 5. Estudiar la conectividad y flujo de especies entre los diferentes hábitats vegetales que constituyen los fondos del ZEC “Calahonda” y como esto influye en la diversidad global, estableciendo los hábitats preferentes para cada una de las especies dominantes y el uso de cada hábitat por parte de estas a lo largo del año de estudio. Además de analizar la diversidad taxonómica y trófica de cada una de las asociaciones de decápodos que están ligados a cada uno de estos hábitats (capítulo 6). CAPÍTULO 2 Estructura y dinámica de las praderas de Posidonia oceanica y Cymodocea nodosa en el noroeste del mar de Alborán: una región de transición entre el Mediterráneo y el Atlántico Este capítulo se basa/ This chapter is base on: Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows in the northwestern Alboran Sea: a Mediterranean–Atlantic transition region Mateo-Ramírez A., Urra J., Marina P., Rueda J.L., García Raso J.E., In preparation. Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows 35 also collected per day and site and transported in darkness at low temperature to the laboratory for chlorophyll a determination. Pigment analyses were carried out by filtering through Whatman GF/C glass filters, extraction using 100% acetone for 12h in cool and dark conditions and measurements using a spectrophotometer at wavelengths of 630, 647, 664 and 750 nm. The chlorophyll a concentrations were obtained using the equation proposed by Jeffrey and Humphrey (1975). Solar irradiance measurements during an annual cycle were obtained using HOBO Data Loggers that were deployed seasonally from winter 2009 to winter 2010. Each HOBO was placed daily close to the sampled seagrass meadows (at ca. 2 m depth) and set to record data with 10 min intervals between 10:00 and 14:00. Measurements were taken several days throughout each season (weeks before, during and after data collection) in both sites. Samples of sediment were also taken seasonally within the sampled meadows (n = 5 replicates per season and site) in order to estimate the organic content in the sediment. This percentage was calculated by the weight loss of dry sediment (3 subsamples of 20 gr. per replicate) after ignition at 500ºC for 1h. Statistical analyses Difference in the shoot density, number of leaves, shoot height and leaf width of P. oceanica and C. nodosa between sites (P. Calaburras; Calahonda) or patch areas (1-25m2; 25-50 m2;>50m2) were tested using a one-way PERMANOVAs, whereas two-way PERMANOVA design with months (M1;warm: JulySeptember ; M2:temperate: November-December; M3;coldest: January-March and M4;cold: May-June) and years (A1: July 2007-May 2008; A2: September 2008-June 2009; A3: September2009-March 2010) as fixed factors was used to analysed temporal and annual differences. The same PERMANOVA design was utilized to test statistical differences of environmental variables between months along the three annual cycles. Analyses were based on Euclidean distances (Anderson 2001) and the significance of P values was determinate through 9999 permutations of row data and residuals under a reduced model for one or two-way analysis respectively. When the numbers of permutation were low, Chapter 2 36 Monte-Carlo method was used to contrast the significance of P values. Twoway PERMANOVA pair-wise post hoc test (P < 0.05) was used for posterior multiple comparisons across months and years. Seasonal trends and values of water temperature and chlorophyll a obtained in the present study matched practically with all those obtained by the NOAA (http://www.noaa.gov), and the former data set was then used for multiple regressions. When seagrass parameters and sediment and environments variables displayed a similar pattern, forward stepwise multiple linear regression analysis was used to investigate the corresponding relationships. However irradiance was analyzed apart because we only have data from March 2009 to March 2010. Prior to carrying out the regression analyses, potential significant relationships between the independent variables (phenology, sediment and environmental variables) were investigated using Pearson’s correlation coefficient and variables that were significantly and highly correlated were eliminated from the analyses. Data used in the regression analyses were checked for normality using Kolmogorov and Smirnov test and transformed to ln when was necessary. These statistical procedures were performed using the software SPSS 20 and PRIMER 6.0 & PERMANOVA statistical package. In order to know the magnitude of annual fluctuations in the different plants parameters of both segrasses species the coefficient of variation (CV %) was calculated with all collected data. Finally a BIPo index was calculated (López y Royo et al. 2010), in order to know the conservation status of P. oceanica meadows studied. Results Environmental variables Mean seawater temperatures (T) showed a clear temporal trend throughout the 3 years in both sites, ranging from ca. 22 ºC in warm months (with exceptional values of 27 ºC in September 2009) to ca. 15 ºC in coldest moths (Fig. 2A, Table 1). Although Chl a showed higher values in Calahonda, both sites followed Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows 37 Figure 2. Environmental variables in P. Calaburras (empty bars) and Calahonda (solid bars): (A) Temperature, (B) concentration of chlorophyll a and (C) solar irradiance in different seasons. The discontinuous line shows the mean surface seawater temperature values from NOAA datasets collected in the study area. July, Jl; November, Nv; January,Ja; May, My; September, Sp; December, Dc; March, Mr; June, Ju. Mean ± standard error. Along the first (A1), second (A2) and third (A3) annual cycle. Letters above error bars display the results of PAIR-WISE test; different letters indicate signifcant differences at P< 0.05. Chapter 2 38 the same temporal trend with low values in coldest months and high in cold and warm months depend on the year (Fig. 2B; Table 1). Solar irradiance also displayed a temporal trend, with high values in warm months and low in coldest one, however these differences were not statistical different (Fig. 2C, Table 1). Percentage of organic matter (%OM) in Posidonia oceanica meadows showed the same seasonal trend in both sites, with high values in warm-cold months (ca. 1.6%) and low in temperate-coldest months (ca. 2.4%) (Table 1). In Cymodocea nodosa meadows, only in P. Calaburras % OM showed temporal differences. High values were recorded in warm-temperate months (maximum in July 07 with 4.09%) and minimum in March 10 (1.87%). In Calahonda the values were similar along the months and years (ca. 2%) (Table 1). Structure and dynamics of Posidonia oceanica meadows The coverage estimated for P. oceanica ranged between 14% in P. Calaburras and 20% in Calahonda. These meadows are composed by patches with different areas Sites Environmental variables Seawater temperature Solar irradiance Chlorophyll a Organic matter P. oceanica C. nodosa Seasonal P-F=75.677; P<0.001*** P-F=2.075; P=0.164 P-F=5.324; P<0.01** P-F=6.474; P<0.001*** P-F=10.81; P<0.001*** Punta Calaburras Inter-annual P-F=4.033; P<0.05* -- P-F=2.896; P=0.068 P-F=0.898; P=0.414 P-F=21.07; P<0.001** Season*year P-F=19.292; P<0.001*** -- P-F=4.971; P<0.01** P-F=1.315; P=0.263 P-F=2.964; P<0.05* Seasonal P-F=8.872; P<0.01** P-F=1.410; P=0.307 P-F=4.410; P<0.05* P-F=10.323; P<0.001*** P-F=0.865; P=0.470 Calahonda Inter-annual P-F=1.172; P=0.344 -- P-F=6.575; P<0.01** P-F=1.822; P=0.171 P-F=2.571; P=0.079 Season*year P-F=0.646; P=0.639 -- P-F=3.947; P<0.05* P-F=2.324; P<0.05* P-F=4.03; P<0.01** Among sites P-F=0.195; P=0.657 P-F=3.410; P=0.078 P-F=7.136; P<0.01** P-F=4.498; P<0.05* P-F=108.7; P<0.001** Table 1. Results of one and two-way PERMANOVA testing differences in environmental variables analyzed in shallow seagrass meadows of the northwestern Alboran Sea, between sites, moths and years. . * Significant differences P<0.05; **P<0.01; *** P< 0.001; P-F, Pseudo-F. Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows 39 (ranging from 1 to over 130 m2) which sometimes are interspaced with dead P. oceanica rhizome mattes. Shoot densities of P. oceanica were not significantly different between sites (one-way PERMANOVA; Pseudo-F= 0.031, P=0.862) when considering all measurements (849.93 ± 15.11 shoots m-2 in P. Calaburras; 845.48 ± 19.99 shoots m-2 in Calahonda; Mean ± SE) (Fig. 3A). Shoot densities in Calahonda displayed a temporal trend with high values in temperate months (maximum values observed in November 2007, ca.1055 shoots m-2). On the other hand in P. Calaburras values were very similar throughout the year and did not present monthly differences (Fig. 3A; Table 2, Appendix 1). Shoot densities were similar between annual cycles in both sites (Table 2) and between patches with different areas (all measurement; one-way PERMANOVA; Pseudo-F=0.433, P=0. 0.651). Negative correlations between shoot densities with shoot height (RPearson=-0.268, P<0.01) and leaf number (RPearson= -0.387, P<0.01) were found. Mature inflorescences were observed in March 2009 on orthotropic shoots of several patches with areas ranging 4-10 m2 at ca. 2 m depth in P. Calaburras, and displaying a maximum of 28 inflorescences m-2. A second and major flowering event was observed in nearly all patches in both P. Calaburras and Calahonda in November 2009. On this occasion, flowering densities were up to 144 inflorescences m-2 in P. Calaburras and up to 184 inflorescences m-2 in Calahonda (Urra et al. 2011b). Shoot heights displayed similar values in both sites ranging between 26.60 ± 1.40 cm in P. Calaburras and 24.90 ± 1.18 cm, in Calahonda (one-way PERMANOVA ; Pseudo-F= 0.863, P= 0.357) (Fig. 3B). A very clear temporal trend was observed in both sites presented high values in cold months (maxima in May 2008 in both sites, ca. 39 cm) (Fig. 3B; Table 2; Appendix 1). Inter-annual differences were detected in both sites with mean annual values ranging between 22.46 ± 1.74 and 27.56 ± 2.51 cm (Fig. 3B; Table 2; Appendix 1). Shoot heights did not display significant differences between patches with different areas (one-way PERMANOVA; Pseudo-F= 2.196, P=0.113). The mean leaf width displayed higher values in P. Calaburras (9.97 ± 0.03 mm) than in Calahonda (10.22 ± 0.03 mm) (one-way PERMANOVA; Pseudo-F= 14.486, P<0.001) (Fig. 3C, Table 2). Leaf width did not show a clear temporal trend, however Chapter 2 40 Figure 3. Phenological parameters of (A-D) Posidonia oceanica and (E-H) Cymodocea nodosa in P. Calaburras (dashed line) and Calahonda (solid line) Mean ± standard error. July, Jl; November, Nv; January, Ja; May, My; September, Sp; December, Dc; March, Mr; June, Ju. Mean±standard error. Along the first (A1), second (A2) and third (A3) annual cycles. Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows 43 between patches (one-way PERMANOVA; Pseudo-F= 0.016, P=0.984). Leaf width presented similar values (ca. 2 mm) in both sites (one-way PERMANOVA; Pseudo-F= 0.274, P=0.596) and in patches with different areas (oneway PERMANOVA; Pseudo-F= 0.9148, P=0.421) (Fig. 3G). Both sites displayed a similar temporal trend with high values in warm and cold months. Maxima were found in July 2007 and in September 2008 for P. Calaburras and Calahonda Table 3. Results of two-way PERMANOVA testing for differences in the shoot density shoot height, leaf width and leaf number of C. nodosa between months, year and mixed factor year x month, which compares each months with the rest of the months sampled in each year. * Significant differences P<0.05; **P<0.01; *** P<0.001. Punta Calaburras Calahonda Shoot density df MS Pseudo-F P df MS Pseudo-F P Year 2 41,999000 48.985 <0.001*** 2 650910 2.949 0.060 Month 3 8,369000 9.761 <0.001*** 3 987940 4.477 <0.01** Year x Month 5 3,578600 4.174 <0.01** 5 710290 3.218 <0.05* Residual 44 857390 44 220700 Total 54 54 Shoot height Year 2 608.95 139.05 <0.001*** 2 55 11.785 <0.001*** Month 3 213.38 48.725 <0.001*** 3 10.16 2.1769 0.090 Year x Month 5 217.2 49.597 <0.001*** 5 31.31 6.7087 <0.001*** Resisdual 44 4.379 44 4.667 Total 54 54 Leaf width Year 2 14.453 721.02 <0.001*** 2 0.088 1.6877 0.196 Month 3 1.5317 76.411 <0.001*** 3 0.471 9.0816 <0.001*** Year x Month 5 1.6708 83.348 <0.001*** 5 0.066 1.2694 0.296 Residual 44 0.0200 44 0.052 Total 54 54 Number of leaves Year 2 11.361 419.35 <0.001*** 2 0.304 14.074 <0.001*** Month 3 2.4185 89.273 <0.001*** 3 3.021 139.63 <0.001*** Year x Month 5 3.472 128.160 <0.001*** 5 0.138 6.3748 <0.001*** Residual 44 0.0271 44 0.022 Total 54 54 Chapter 2 42 The mean number of leaves per shoot was similar in both sites (ca.5.7 leaves shoot-1) (one-way PERMANOVA; F= 0.422, P >0.519), displayed a temporal trend with high values in coldest month. Maxima values were observed in January 2008 for P. Calaburras, and in March 2009 for Calahonda (Fig. 3D; Table 2; Appendix 1). No differences of number of leaves in relation to annual cycles and patches area were found (Table 2) (one-way PERMANOVA; Pseudo-F= 1.185, P=0.315). Structure and dynamics of Cymodocea nodosa The coverage estimated for C. nodosa was <5% in both sites, with patch dimensions ranging from 1 to over 60 m2 that sometimes were mixed with patches of P. oceanica. The mean shoot density of C. nodosa meadows was significantly higher (all measurements; one-way PERMANOVA; Pseudo-F= 14.920 P <0.001) in P. Calaburras (2640.71 ± 222.66 shoots m-2; Mean ± SE) than in Calahonda (2122.76 ± 75.16 shoots m-2) (Fig. 3E). In P. Calaburras shoot density displayed a temporal trend with high values in warm and cold months (maximum in July 2007, ca. 4592 shoots m-2) (Fig. 3E; Table 3; Appendix 2). In summer 2009, no C. nodosa shoots were observed in this site, however some of them were spotted again in November 2009 but none were found in March 2010. A clear decline was observed between the first and third annual cycles (3180.80 ± 309.29, 2724.80 ± 341.23 and 144 ± 26.84 shoot m-2) in P. Calaburras. On the other hand, in Calahonda shoot density displayed monthly and inter-annual changes, however were less acute than in P. Calaburras (maxima in July and November 2007) (Fig. 3E; Table 3; Appendix 2). Shoot densities were similar in patches with different area (one-way PERMANOVA; Pseudo-F= 1.217, P=0.309). Shoot height was not significantly different between sites with values ca. 9 cm (one-way PERMANOVA; Pseudo-F= 2.581, P= 0.11) (Fig. 3F). Although shoot height did not follow a temporal trend, inter-annual changes were found in both sites ranging from 6.89 ± 0.44 to 13.45 ± 2.08 cm (Table 3). However, monthly differences were only observed in P. Calaburras showed maximum values in July 2007 (Fig. 3F; Table 3; Appendix 2). Shoot heights did not display differences Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows 43 between patches (one-way PERMANOVA; Pseudo-F= 0.016, P=0.984). Leaf width presented similar values (ca. 2 mm) in both sites (one-way PERMANOVA; Pseudo-F= 0.274, P=0.596) and in patches with different areas (oneway PERMANOVA; Pseudo-F= 0.9148, P=0.421) (Fig. 3G). Both sites displayed a similar temporal trend with high values in warm and cold months. Maxima were found in July 2007 and in September 2008 for P. Calaburras and Calahonda Table 3. Results of two-way PERMANOVA testing for differences in the shoot density shoot height, leaf width and leaf number of C. nodosa between months, year and mixed factor year x month, which compares each months with the rest of the months sampled in each year. * Significant differences P<0.05; **P<0.01; *** P<0.001. Punta Calaburras Calahonda Shoot density df MS Pseudo-F P df MS Pseudo-F P Year 2 41,999000 48.985 <0.001*** 2 650910 2.949 0.060 Month 3 8,369000 9.761 <0.001*** 3 987940 4.477 <0.01** Year x Month 5 3,578600 4.174 <0.01** 5 710290 3.218 <0.05* Residual 44 857390 44 220700 Total 54 54 Shoot height Year 2 608.95 139.05 <0.001*** 2 55 11.785 <0.001*** Month 3 213.38 48.725 <0.001*** 3 10.16 2.1769 0.090 Year x Month 5 217.2 49.597 <0.001*** 5 31.31 6.7087 <0.001*** Resisdual 44 4.379 44 4.667 Total 54 54 Leaf width Year 2 14.453 721.02 <0.001*** 2 0.088 1.6877 0.196 Month 3 1.5317 76.411 <0.001*** 3 0.471 9.0816 <0.001*** Year x Month 5 1.6708 83.348 <0.001*** 5 0.066 1.2694 0.296 Residual 44 0.0200 44 0.052 Total 54 54 Number of leaves Year 2 11.361 419.35 <0.001*** 2 0.304 14.074 <0.001*** Month 3 2.4185 89.273 <0.001*** 3 3.021 139.63 <0.001*** Year x Month 5 3.472 128.160 <0.001*** 5 0.138 6.3748 <0.001*** Residual 44 0.0271 44 0.022 Total 54 54 Chapter 2 44 respectively. Leaf width showed differences between years in P. Calaburras, which presented a diminution from the first (2.46 ± 0.10 mm) to third annual cycle (1.63 ± 0.04 mm) (Fig. 3G; Table 3; Appendix 2). Number of leaves per shoot was not significantly different between sites (ca. 2.4 leaves shoot-1) (one-way PERMANOVA; Pseudo-F= 0.261, P=0.617) neither between patches with different area (one-way PERMANOVA; Pseudo-F= 1.975, P=0.157) (Fig. 3H). This variable present a similar temporal trend than leaf width and shoot density with maxima values in July 2007 and September 2008 for P. Calaburras and Calahonda respectively. Inter-annual changes were found in both sites with values ranging from 2.35 ± 0.08 to 2.52 ± 0.07 leaves shoot-1 (Fig. 3H; Table 3; Appendix 2). Relationship between phenological, sediment and environmental variables In P. oceanica the irradiance showed the highest values of R Square, explained 47.9% and 59.4% of the variability in relation to leaf number and shoot density respectively. In C. nodosa only sea water temperature showed a significant correlation with C. nodosa phenological variables, explained 24.4 % of shoot height variability (Table 4). Table 4. Results of multiple forward stepwise regression analyses of P. oceanica and C. nodosa variables in relation to environmental and sediment variables. T, temperature, I, irradiance, Chl a, concentration of chlorophyll a, % OM, percentage of organic matter in the sediment. Only environmental variables accepted in each regression model (P <0.05) are listed. Coefficient SE F ratio R2P Posidonia oceanica Leaf number 0.186 <0.01 constant 6.308 0.311 temperature -0.037 0.17 4.573 Leaf number 0.479 <0.05 constant 10.635 1.929 irradiance -1.319 0.52 6.424 Shoot density 0.594 <0.05 constant -571.8 431.4 irradiance 372.4 116.3 10.255 Cymodocea nodosa Shoot height 0.242 <0.05 constant 1.238 0.376 temperature 0.051 0.021 5.753 Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows 51 Appendixes Appendix 1. Two-way PERMANOVA pair-wise analysis of P. oceanica parameters among months and years using t-statistic. Values of significance were calculated used Monte Carlos (MC) method. M1, July, September; M2, November, December; M3 January, March; M4, May, June; A1, July 2007-May 2008; A2, September 2008-June 2009; A3, September 2009-March 2010.* Significant differences P<0.05; **P<0.01; *** P< 0.001. Only appears the values of the phenological parameters that presented differences in the mixed factor year x month. Punta Calaburras groups shoot density shoot height leaf weight number of leaf t P(MC) t P(MC) t P(MC) t P(MC) Month M1 A1 vs A2 2.536 <0.05* 6.036 <0.01** A1 vs A3 0.182 0.854 3.754 <0.01** A2 vs A3 5.121 <0.01** 2.859 <0.01** Month M2 A1 vs A2 5.223 <0.01** 0.953 0.371 A1 vs A3 5.034 <0.01** 4.545 <0.01** A2 vs A3 0.849 0.423 2.936 <0.01** Month M3 A1 vs A2 2.838 <0.05* 0.949 0.377 A1 vs A3 4.416 <0.01** 1.147 0.278 A2 vs A3 2.501 <0.05* 1.727 0.121 Month M4 A1 vs A2 0.181 0.867 0.745 0.477 Year A1 M1 vs M2 4.558 <0.01** 3.59 <0.01** M1 vs M3 1.231 0.257 3.65 <0.01** M1 vs M4 1.482 0.186 3.202 <0.05* M2 vs M3 6.139 <0.001*** 2.230 0.059 M2 vs M4 9.196 <0.001*** 1.914 0.095 M3 vs M4 3.876 <0.01** 0.107 0.916 Year A2 M1 vs M2 5.422 <0.001*** 3.066 <0.05* M1 vs M3 8.143 <0.001*** 2.572 <0.05* M1 vs M4 14.158 <0.001*** 3.7 <0.05* M2 vs M3 13.824 <0.001*** 0.109 0.915 M2 vs M4 23.049 <0.001*** 0.656 0.528 M3 vs M4 4.238 <0.01** 0.596 0.564 Year A3 M1 vs M2 7.804 <0.001*** 2.711 <0.05* M1 vs M3 2.805 <0.05* 1.566 0.150 M2 vs M3 11.511 <0.001*** 0.112 0.913 Chapter 2 52 Calahonda groups shoot density shoot height leaf weight number of leaf t P(MC) t P(MC) t P(MC) t P(MC) Month M1 A1 vs A2 0.713 0.507 2.619 <0.05* 1.530 0.172 A1 vs A3 0.461 0.660 0.406 0.687 3.981 <0.01** A2 vs A3 0.232 0.819 3.489 <0.01** 1.296 0.233 Month M2 A1 vs A2 0.745 0.474 2.051 0.076 0.843 0.432 A1 vs A3 2.829 <0.05* 2.840 <0.05* 5.622 <0.001*** A2 vs A3 2.635 <0.05* 6.738 <0.001*** 5.242 <0.01** Month M3 A1 vs A2 7.193<0.001*** 1.205 0.265 2.496 <0.05* A1 vs A3 2.918 <0.05* 3.038 <0.05* 2.389 <0.05* A2 vs A3 2.202 0.060 3.638 <0.01** 0.448 0.659 Month M4 A1 vs A2 6.328<0.001*** 2.922 <0.05* 1.329 0.220 Year A1 M1 vs M2 2.493 <0.05* 6.607 <0.001*** 0.442 0.668 M1 vs M3 0.963 0.370 0.118 0.908 3.702 <0.01** M1 vs M4 1.814 0.109 5.221 <0.01** 2.499 <0.05* M2 vs M3 2.784 <0.05* 7.503 <0.001*** 3.9 <0.01** M2 vs M4 6.863<0.001***13.818 <0.001*** 2.851 <0.05* M3 vs M4 8.288<0.001*** 6.124 <0.001*** 0.501 0.632 Year A2 M1 vs M2 1.956 0.088 8.261 <0.001*** 1.226 0.259 M1 vs M3 4.09 <0.01** 0.860 0.420 0.243 0.807 M1 vs M4 1.195 0.268 5.169 <0.001*** 0.434 0.668 M2 vs M3 8.076<0.001*** 5.326 <0.001*** 1.342 0.218 M2 vs M4 4.526 <0.01** 10.228 <0.001*** 0.571 0.582 M3 vs M4 5.423<0.001*** 3.254 <0.05* 0.304 0.771 Year A3 M1 vs M2 0.235 0.816 10.679 <0.001*** 2.312 0.052 M1 vs M3 1.424 0.191 3.511 <0.01** 2.081 0.067 M2 vs M3 2.480 <0.05* 7.412 <0.001*** 3.687 <0.01** Structure and dynamic of Posidonia oceanica and Cymodocea nodosa meadows 53 Appendix 2. Two-way PERMANOVA pair-wise analysis of C. nodosa parameters among months and years using t-statistic. Values of significance were calculated used Monte Carlos (MC) method. M1,July, September; M2, November, December; M3 January, March; M4, May, June; A1, July 2007-May 2008; A2, September 2008-June 2009; A3, September 2009-March 2010.* Significant differences P<0.05; **P<0.01; *** P< 0.001. Only appears the values of the phenological parameters that presented differences in the mixed factor year x month. Punta Calaburras groups shoot density shoot height leaf weight number of leaf t P(MC) t P(MC) t P(MC) t P(MC) Month M1 A1 vs. A2 0.179 0.863 9.096 <0.001*** 0.179 0.863 4.543 <0.01** A1 vs. A3 9.879 <0.001*** 13.586 <0.001*** 9.879 <0.001***22.113 <0.001*** A2 vs. A3 4.819 <0.001** 16.665 <0.001*** 4.819 <0.01** 43.818 <0.001*** Month M2 A1 vs. A2 0.826 0.428 2.535 <0.05* 0.826 0.428 0.283 0.781 A1 vs. A3 6.539 <0.001*** 2.009 0.084 6.539 <0.001*** 2.910 <0.05* A2 vs. A3 7.359 <0.001*** 0.065 0.952 7.359 <0.001*** 2.741 <0.05* Month M3 A1 vs. A2 1.366 0.214 1.52 0.172 1.366 0.214 3.138 <0.05* A1 vs. A3 8.554 <0.001*** 19.136 <0.001*** 8.554 <0.001*** 48.5 <0.001*** A2 vs. A3 6.745 <0.001*** 4.473 <0.01** 6.745 <0.001***66.408 <0.001*** Month M4 A1 vs. A2 2.062 0.070 1.622 0.137 2.062 0.070 0.643 0.539 Year A1 M1 vs. M2 4.6 <0.01** 7.475 <0.001*** 4.6 <0.01** 6.24 <0.001*** M1 vs. M3 4.42 <0.01** 9.83 <0.001*** 4.42 <0.01** 7.867 <0.001*** M1 vs. M4 1.052 0.323 9.305 <0.001*** 1.052 0.323 2.54 <0.05* M2 vs. M3 0.418 0.680 2.245 0.055 0.418 0.680 4.382 <0.01** M2 vs. M4 2.767 <0.05* 1.928 0.087 2.767 <0.05* 4.199 <0.01** M3 vs. M4 2.56 <0.05* 0.053 0.959 2.56 <0.05* 6.338 <0.001*** Year A2 M1 vs. M2 2.063 0.071 2.675 <0.05* 2.063 0.071 2.954 <0.05* M1 vs. M3 2.814 <0.05* 2.431 <0.05* 2.814 <0.05* 4.743 <0.01** M1 vs. M4 2.031 0.082 3.882 <0.01** 2.031 0.082 2.639 <0.05* M2 vs. M3 1.718 0.124 1.332 0.222 1.718 0.124 0.885 0.398 M2 vs. M4 0.15 0.882 2.164 0.062 0.15 0.882 4.423 <0.01** M3 vs. M4 1.164 0.274 0.369 0.723 1.164 0.274 5.297 <0.001*** Year A3 M1 vs. M2 2.683 <0.05* 7.741 <0.001*** 2.683 <0.05* 32.932 <0.001*** M1 vs. M3 M2 vs. M3 2.683 <0.05* 7.741 <0.001*** 2.683 <0.05* 32.932 <0.001*** Chapter 2 54 Calahonda groups shoot density shoot height leaf weight number of leaf t P(MC) t P(MC) t P(MC) t P(MC) Month M1 A1 vs. A2 2.101 0.071 2.756 <0.05* 8.552 <0.001*** A1 vs. A3 1.215 0.262 4.519 <0.01** 3.157 <0.05* A2 vs. A3 0.581 0.576 3.473 <0.01** 0.191 0.855 Month M2 A1 vs. A2 0.237 0.219 0.038 0.972 3.200 <0.05* A1 vs. A3 0.066 0.073 1.111 0.299 3.464 <0.05* A2 vs. A3 0.33 0.318 1.124 0.293 1.723 0.124 Month M3 A1 vs. A2 2.894 <0.05* 1.775 0.114 2.108 0.068 A1 vs. A3 0.215 0.835 2.084 0.069 2.557 <0.05* A2 vs. A3 2.633 <0.05* 0.71 0.501 3.772 <0.01** Month M4 A1 vs. A2 1.504 0.179 3.805 <0.01** 3.043 <0.05* Year A1 M1 vs. M2 0.297 0.775 1.49 0.184 19 <0.001*** M1 vs. M3 3.971 <0.01** 3.268 <0.01** 10.954 <0.001*** M1 vs. M4 3.824 <0.01** 0.255 0.805 0.535 0.602 M2 vs. M3 2.996 <0.05* 2.074 0.074 3.138 <0.05* M2 vs. M4 3.185 <0.05* 2.059 0.073 9.985 <0.001*** M3 vs. M4 1.046 0.323 4.061 <0.01** 6.893 <0.001*** Year A2 M1 vs. M2 0.746 0.471 1.71 0.122 10.25 <0.001*** M1 vs. M3 1.561 0.159 0.861 0.419 13.02 <0.001*** M1 vs. M4 0.274 0.789 0.059 0.957 M2 vs. M3 0.901 0.390 1.383 0.197 2.631 <0.05* M2 vs. M4 1.077 0.314 1.797 0.115 6.608 <0.001*** M3 vs. M4 1.89 0.095 0.877 0.410 8.746 <0.001*** Year A3 M1 vs. M2 0.877 0.397 1.871 0.098 9.26 <0.001*** M1 vs. M3 1.534 0.165 3.918 <0.01** 5.692 <0.01** M2 vs. M3 0.573 0.585 1.002 0.336 3.357 <0.05* CAPÍTULO 3 Cambios estacionales en la estructura de las asociaciones de crustáceos decápodos asociados a praderas de Cymodocea nodosa en el Mar de Alborán (Mediterráneo occidental) Este capítulo se basa/ This chapter is based on: Temporal changes in the structure of the crustacean decapod assemblages associated with Cymodocea nodosa meadows from the Alboran Sea (Western Mediterranean Sea) Mateo-Ramírez A., García Raso J.E. Marine Ecology: An Evolutionary Perspective (2012): 33, 302-316 Decapod assemblages associated with Cymodocea nodosa 57 Abstract The decapod assemblages associated with two shallow meadows of Cymodocea nodosa, located in the same geographical area (southern Spain) but on different substrates and with different patch size, have been analyzed. They display similar structure (diversity indices not significantly different), without a clear relation of richness and abundances to patch size, and with the same dominant species (the family Hippolytidae and, in particular, Hippolyte leptocerus are characteristic of this habitat). Composition of both crustacean assemblages is influenced by species that are common in neighbouring habitats. Therefore the connectivity among them is an important factor in the qualitative and quantitative structure of these decapod communities. Species richness appears to be higher than in Cymodocea meadows elsewhere in the Mediterranean and Atlantic at a similar depth, perhaps as consequence of the biogeographical location and the high diversity and connectivity with surrounding biotopes. High evenness values are the result of the structure and location of these meadows, which are fragmented and interspersed with other biotopes (sandy and rocky bottoms), resulting in an “ecotone effect”. On the other hand, the structures of the decapod assemblages differ significantly according to season. The abundance and species richness are both related to plant phenology and the dominant species present a positive correlation with the number of leaves per shoot. The maximum abundance of many species is coincident with the greatest seagrass development (spring – summer), which provide more resources (surface, biomass, protection, food). Therefore, seasonality is linked to plant life cycle, but also to the interrelationships and biology of the species, which are adapted and specialized to the environmental features of these shallow habitats. Keywords: Cymodocea nodosa, decapods, space - time variations, diversity, patch sizes, Western Mediterranean Sea. 5IJT BSUJDMF DBO CF EPXOMPBEFE GSPN IUUQXXXEFHSVZUFSDPNWJFXKCPUNJTTVFCPU CPUYNM Chapter 3 58 Introduction Marine seagrass meadows are essential elements in the biological structure and physical-chemical process of the coastal areas (Duarte and Sand-Jensen, 1990). They fulfil many important and varied functions such as the construction of new microhabitats with different ecological conditions, therefore enhancing biodiversity and biological interactions (Bellan-Santini et al. 1994; Pedersen et al. 1997; Orth et al. 1984; Phillips and Meriez 1988). Several studies deal with specific zoological groups associated with Cymodocea nodosa or with the total macrofauna in general (Ledoyer 1966, 1968; Scipione et al. 1996; Sánchez-Jerez et al. 1999; Guidetti and Bussotti 2000; Koutsoubas et al. 2000; Corbera et al. 2002; Barbera-Cebrián et al. 2002; Riera et al. 2003; Vizzini and Mazzolla 2004; Brito et al. 2005; Tuya et al. 2006, González et al. 2007; Como et al. 2008) but a few of them are focussed on the crustacean decapod community in the Atlantic Ocean (Schaffmeister et al. 2006) or the Mediterranean Sea (Ledoyer 1966, 1968; Števčić 1991; Reed and Manning 2000; García Raso et al. 2006a). On the other hand, the results found in these studies have shown quantitative and qualitative differences which could be attributed to different factors, such as the geographic area, sampling period, seagrass structural complexity, depth, substrata, sampling methodology, predation, among others. Some of these relationships have been analyzed in different geographical areas, such as Europe (Borg and Schembri 2000; Attrill et al. 2000; Hirst and Attrill 2008), North America (Lewis and Stoner 1983), Australia (Bell and Westoby 1986, Worthington et al. 1992, Kwak and Klumpp 2004), Indonesia (Unsworth et al. 2007) and seagrass species (Zostera marina, Z. capricorni, Posidonia oceanica, Posidonia australis, Cymodocea serratula, C. rotundata, Halodule spp., Halophilla ovalis, Thalassia testudinum, etc), sometimes with apparently contradictory results (e.g., the densities of the shrimp Hippolyte in two different studies (Lewis and Stoner 1983)). However, difficulties in macrofaunal estimates and in the interpretations should be taken into account, mainly when the species or groups under study prefer a specific microhabitat and/ or a particular food resource (for example, within decapods, the caridean shrimps and the penaeids; Mellors and Marsh 1993). Other factors to consider in the interpretations are the sampling methodology (Borg and Schembri 2000), the life Decapod assemblages associated with Cymodocea nodosa 59 cycles and larval settlement (Bell and Westoby 1986), the species adaptation and the predatory pressures (Unsworth et al. 2007). The aim of this study is to help to clarify, as far as possible, some of these situations, but focussing specifically on the seagrass Cymodocea nodosa and its decapod fauna. For this purpose, we compare two decapod assemblages from two very shallow meadows in the same geographic area. They are located only 7 kilometers apart, but on bottoms with different exposure level, substrata and surrounding habitats. Nevertheless there are no appreciable differences in depth, sea water temperature or plant phenology, and considering that the biogeographical setting is the same. Assemblage structures (values of diversity indices, richness, abundances, species composition) were analyzed spatially and temporally and, also in relation to the plant phenology. In addition, the existence of small patches of different sizes in one of the studied zones allowed us to evaluate the possible effect of fragmentation of habitat and species/area relationship. There are some studies dealing with those topics (Attrill et al. 2000, Bell et al. 2001; Barbera-Cebrián et al. 2002; Tanner 2005; Hirst and Attrill 2008; Macreadie et al. 2009) but they have never been carried out on the small seagrass Cymodocea nodosa and specifically on decapods. The study area The sampling area is located off the coast of Mijas (Málaga, southern Spain, western Mediterranean Sea) (Figure 1), within the marine Site of Community Importance (SCI) known as Calahonda (code ES6170030) in the Natura 2000 Network. It is one of the very few natural rocky outcrops existing on the shores of Malaga province and a “hotspot” for European biodiversity (García Raso et al. 2010) due to its high species richness with the presence of endangered and unique species and to the coexistence of species from different biogeographic origins (Atlantic Mediterranean, African-European), mainly studied for molluscs (Urra et al. 2011; Gofas, personal communication). This is a consequence of its geographical location (near the Straits of Gibraltar) and its oceanographic, physical-chemical and upwelling features (see Ekman 1953; Briggs 1974; Conde and Seoane 1982; Parrilla and Kinder 1987; Tintoré et al. 1991; Gofas 1999). Chapter 3 60 In the study area, small patches (1-10 m long) of Cymodocea nodosa can be found between 1 and 5 m depth, mixed with rocks, sand, seaweed and Posidonia oceanica (which is close to its westernmost biogeographical distribution limit). Two shallow sites were selected, “Punta de Calaburras” (PC) 36º30’23’’ N - 04º38’41’’W and Calahonda (CH) 36º29’21’’N - 04º41’55’’W. In Punta de Calaburras, the patch is of about 27 m2 at 0.6 - 1m, subjected to wave action (which determines the abundance of pebbles). Calahonda is a more sheltered site, and the meadow studied is composed of patches between 0.62 and 10.34 m2 (mean: 5 m2) at 1.5 - 2 m depth and protected by large rocks. Material and Methods Sampling methodology The samples were taken seasonally from July 2007 to May 2008 (July - summer “s”; November - autumn “a”; January or February - winter “w” and May - spring “p”); by two SCUBA divers during day. The specimens were collected with a manual airlift pump connected to an air tank, with a 0.5 mm mesh bag to prevent the loss of small specimens. The sampling methodology is friendly with the environment, and this represents a basic requirement in a protected area where the Cymodocea meadows are adjacent to fragile Posidonia meadows (the westernmost known site). In addition, this sampling gear is more effective than Fig. 1. Study area. Punta de Calaburras (A), Calahonda (B) (Southern Spain). Decapod assemblages associated with Cymodocea nodosa 67 frequent in deeper bottoms). The number of individuals (N) and species richness (S) throughout the year followed a similar trend: with peaks in spring and summer and minimal values in winter (Figure 2 A and B). Values were slightly higher in Punta de Calaburras than in Calahonda. Species richness ranged between 3 (PC w4) and 17 (PC s2) in Punta de Calaburras, and between 1 (CH w3) and 12 (CH p4) in Calahonda. Abundances (number of individuals) fluctuated between 4 (PC w4) and 121 ind. (PC s2) in Punta de Calaburras, and between 1 (CH w3) and 59 ind. (CH p4) in Table 4. SIMPER analyses. Species ranked according to their average within-group similarity at Punta de Calaburras (PC) and Calahonda (CH) (groups 1 and 2, 3, 5 in figure 4), through seasonality. Av.Abund: average abundance, Av.Sim: average similarity, Sim/SD: similarity/standard deviation, Contrib.%: average contribution to similarity, Cum.%: cumulative percentage of similarity. Species Av.Abund Av.Sim Sim/SD Contrib.% Cum.% Punta Calaburras (PC) Calcinus tubularis 1.79 8.58 1.54 19.84 19.84 Hippolyte leptocerus 1.68 6.29 1.1 14.54 34.38 Clibanarius erythropus 1.24 6.18 0.8 14.29 48.67 Anapagurus hyndmanni 1.30 5.09 0.74 11.76 60.42 Cestopagurus timidus 0.97 4.11 1.00 9.50 69.93 Hippolyte inermis 1.08 2.97 0.60 6.86 76.79 Athanas nitescens 0.90 2.28 0.61 5.28 82.07 Philocheras fasciatus 0.56 1.52 0.34 3.51 85.57 Polybius navigator 0.63 1.35 0.47 3.12 88.69 Xantho poressa 0.63 1.03 0.31 2.39 91.08 Calahonda (CH) Philocheras fasciatus 1.29 7.41 0.75 19.31 19.31 Hippolyte leptocerus 1.34 7.29 0.89 18.98 38.29 Anapagurus hyndmanni 0.89 6.02 0.75 15.67 53.96 Pirimela denticulata 0.78 5.64 0.98 14.69 68.65 Polybius navigator 0.37 2.10 0.36 5.47 74.12 Sycionia carinata 0.28 1.87 0.29 4.88 79.00 Hippolyte inermis 0.39 1.80 0.42 4.69 83.69 Processa edulis edulis 0.63 1.30 0.47 3.39 87.08 Sirpus zariquieyi 0.47 1.14 0.33 2.96 90.04 Chapter 3 68 Calahonda. The mean values found in each site (considering all replicates) were: S = 8.90 ± 3.78 species and N = 30.80 ± 26.12 ind. for Punta de Calaburras and S = 6.37 ± 3.35 species and N = 16.53 ± 15.96 ind. for Calahonda. Abundance and species richness of decapod assemblages were related to seagrass phenological variables. The number of leaves per shoot and number of shoot per square meter are similar at both sites (Mann-Whitney U test Z = - 0.627 P = 0.531 and Z= -1.921, P = 0.055). In the whole studied area, the number of leaves per shoot is positively correlated to abundance and species richness (RPearson Fig. 2. Seasonal evolution (mean ± SD) of species richness (S) and number of individuals (N) at Calahonda (A) and Punta de Calaburras (B) sites (Ju = July, Nov = November, Ja = January, May = May). Decapod assemblages associated with Cymodocea nodosa 69 = 0.621, P = 0.000025 and RPearson = 0.486, P = 0.002 respectively). Considering each site separately, the number of leaves per shoot is also positively correlated to the abundance and richness in Calahonda (RPearson = 0.623, P = 0.004 and RPearson = 0.620, P = 0.005 respectively). Nevertheless, a correlation with the number of individuals (RPearson = 0.635, P = 0.003) was only observed in Punta de Calaburras. Similarly, the abundance of the dominant species Hippolyte letocerus and Hippolyte inermis displayed a positive correlation with the number of leaves per shoot (RSpearman= 0.346, P = 0.029 and RSpearman=0.498, P = 0.001) and the abundance of Anapagurus hyndmanni is also positively correlated with the number of leaves per shoot but also with the number of shoots per square meter (RSpearman= 0.477, P = 0.001; RSpearman= 0.418, P = 0.007). Abundance and species richness did not display any correlation with the patch size (in Calahonda) (RSperman = -0.052, P = 0.827 and RSperman = -0.212, P = 0.370). This could be a consequence of the non-significant correlations between the patch area and the abundance of the three species before mentioned, Hippolyte letocerus, Hippolyte inermis and Anapagurus hyndmanni (RSpearman= 0.003, P = 0.989; RSpearman= -0.241, P = 0.306 and RSpearman= -0.334, P = 0.15 respectively). The results for the Calahonda assemblage (Figure 3A) displayed low-middle values in diversity indices and high evenness (H’= 1.00 ± 0.68 winter, 1.77 ± 0.23 spring; 1-l = 0.82 ± 0.03 spring, 0.893 ± 0.082 winter; J’= 0.83 ± 0.054 spring, 0.95 ± 0.04 winter). The mean values were H’= 1.49 ± 0.59, J’ = 0.90 ± 0.06 and 1-l = 0.85 ± 0.09. On the other hand, the assemblage in Punta de Calaburras shows higher values in diversity indexes and lower evenness (Figure 3B) (H’= 1.42 ± 0.66 winter, 2.04 ± 0.26 autumn; 1-l = 0.72 ± 0.29 summer, 0.87 ± 0.05 autumn; J’= 0.76 ± 0.22 summer, 0.87 ± 0.05 autumn). The mean values were H’= 1. 74 ± 0.54; 1-l = 0.79 ± 0.20 and J’= 0.82 ± 0.16. Only slight differences were observed between the values from both areas, but they are not significantly different (H’: FANOVA = 2.204 with P = 0.146; J’ and 1-l: MannWhitney U test Z = 131, P = 0.097 and Z = 184, P = 0.866, respectively). Significant difference was observed between seasonal samples of the global Chapter 3 70 decapod community (Punta de Calaburras + Calahonda), but with a low R value (RANOSIM = 0.225, P = 0.001) (the same occurs when the temperature is considered, RANOSIM = 0.466, P = 0.001). Difference was more significant when assemblages of Punta de Calaburras and Calahonda are considered separately across the seasonal groups (RANOSIM = 0.612, P = 0.001). Within sites (Table 5), seasonality was noted in Calahonda (RANOSIM = 0.425, P = 0.001): most seasonal groups show significant differences except autumn and winter. The species with highest contribution to the dissimilarity or characterization of summer assemblage were Anapagurus hyndmanni, Pirimela denticulada, Philocheras fasciatus, Hippolyte inermis and Sirpus zariquieyi (94.51% of cumulate contribution), and the autumn assemblage was mainly characterized by Syciona carinata, Polybius navigator, Anapagurus hyndmanni and Hippolyte leptocerus (with a 95.03%). Significant seasonality was also observed in Punta de Calaburras (RANOSIM = 0.390, P = 0.001). Summer and winter showed Fig. 3. Mean values and standard deviation of Pielou (J’), Shannon (H’) and Simpson (1l) diversity indexes at Calahonda (A) and Punta de Calaburras (B) sites (Ju = July, Nov = November, Ja = January, May = May). Decapod assemblages associated with Cymodocea nodosa 71 the highest differences, while the most similar were summer and autumn. In summer, Calcinus tubularis, Hippolyte inermis, Athanas nitescens, Hippolyte leptocerus, Cestopagurus timidus, Anapagurus hyndmanni and Sirpus zariquieyi represent a cumulated contribution of 91.39% and, in winter, Clibanarius erythropus, Philocheras fasciatus, Calcinus tubularis, Hippolyte leptocerus and Polybius navigator represent 92.44%. Table 5. ANOSIM pairwise test comparing seasonal groups at Calahonda (CH) and Punta de Calaburras (PC) sites. Calahonda (CH) Punta de Calaburras (PC) Groups R P R P Ju,Nov 0.706 0.016* 0.208 0.056 ns Ju,Ja 0.326 0.024* 0.684 0.008** Ju, May 0.468 0.008** 0.268 0.056 ns Nov, Ja 0.125 0.214 ns 0.256 0.063 ns Nov, May 0.656 0.008** 0.628 0.008** Ja, May 0.366 0.024* 0.600 0.008** P = significance level; *P < 0.05, **P < 0.01; n.s. = non significant; R = rank of similarities between samples; Ju = July; Nov = November; Ja = January; May = May. These two assemblages, 7 kilometers apart from each other and within the same habitat, are significantly different (RANOSIM = 0.362, P = 0.001). These differences and similarities are shown in the quantitative analyses of aggregation CLUSTER (Figure 4A) and ordination MDS (Figure 4B). Punta de Calaburras (PC) samples and replicates form a single group (only the replicates PCw3 and PCw4 are not included) organized into two seasonal subgroups: summer-autumn and winter-spring; in contrast, broader differences are observed in Calahonda (CH), because in spite of having the same similarity (28.62%), Calahonda replicates are organized into three groups with different seasonality (springsummer: group 2, autumn: group 3 and winter: group 4). Chapter 3 72 Fig. 4. Quantitative aggregation (Cluster) (A) and ordination (MDS) (B) analyses. Groups 1–5 discriminated at a similarity level of 28%. CH = Calahonda; PC = Punta de Calaburras; Ju = July; No = November; Ja = January; Ma = May. Numbers after month codes are replicate ranks. Decapod assemblages associated with Cymodocea nodosa 73 Discussion The Cymodocea nodosa meadows from Calahonda and Punta de Calaburras, in Southern Spain, show seasonal structural dynamics, with higher number of shoots m-2 than observed in other areas such as Ischia (Italy) (Cancemi et al., 2002), Mar Menor (Terrados and Ros 1992), Canary Islands (Reyes et al. 1995) and the Mauritanian Arguin Bank (van Lent et al. 1991), but with a lower number of leaves per shoot than that found in Tenerife or in the Adriatic Sea (Reyes et al. 1995, Guidetti et al. 2002). According to Terrados and Ros (1992) Cymodocea nodosa has a stable, genetically controlled leaf development model and differences may be attributed to variable nutrient availability affecting plant size. In fact, differences were observed in our data among patches located in two shallow sites characterized by slightly different environmental conditions. A comprehensive study on the phenology of this meadow from 2007 to 2010 is currently in preparation. As a general feature, the studied decapod community associated with shallow Cymodocea nodosa patches shows high species richness, with 34 species recorded. This richness is higher than that found in meadows at similar depths in the Adriatic Sea (1.5-3 m, 30 species, Števčič 1991) or Ischia (3 m, 18 species, Scipione et al. 1996) and lower than in Cabo de Gata (Almeria, SE Spain) when combining day and night samplings (48 species), but similar when only considering daytime samplings (34 species, García Raso et al. 2006). Nevertheless, it must be pointed out that the Cabo de Gata meadow is located at 10-14 m depth, therefore deeper and less influenced by the heavy stress caused by waves at shallower stations, which undoubtedly favours meadow habitability and diversity. In a study of decapod assemblages of Posidonia oceanica from Malta the diversity values were increased with increasing depth (maximum at 16 m) (Borg and Schembri 2000). The incidence of depth on species richness was noted nearby on soft-bottoms transects (García Muñoz et al. 2008), where 18 species of decapods were identified at 5 m (shallow sandy bottoms) and 29, 30 and 38 species at 15 m (in fine-sand, coarse-sand and coralligenous bottoms respectively). A similar influence of depth on faunal assemblages was observed in other groups such as molluscs (Rueda et al. 2000, Koulouri et al. 2006, Urra et al. 2011a). Chapter 3 74 The dominant decapod species at both sites was Hippolyte leptocerus, linked to the leaf stratum, as in daytime samples of Cymodocea nodosa from Almeria and other seagrass assemblages from French, Italian and Tunisian coasts (Ledoyer 1968, 1984; Scipione et al. 1996; Reed and Manning 2000; García Raso et al. 2006a). In addition, this confirms the dominance of the family Hippolytidae in these biotopes (Ledoyer 1966, 1968, 1969, 1984; Kikuchi and Pérès 1997; Templado 1984, Scipione et al. 1996), as it also occurs in the decapods assemblage of the seaweed Caulerpa prolifera (Lopez de la Rosa et al. 2006). This family in general and the genus Hippolyte in particular, are perfectly adapted to these environments and strata (leaves) by their shape, colour and behaviour, thus reducing their vulnerability to predation by visual hunters (Main 1987). Another dominant species at both sites was the hermit crab Anapagurus hyndmanni; but curiously, this species was absent on adjacent shallow sandy bottoms (5 m) and markedly abundant at deeper (15-25 m) detritic bottoms in the studied area (García Muñoz et al. 2008). Few others species show a total abundance of more than 5%, partly as a result of the environmental conditions and ecotone effect. Moreover, they also vary according to sites. Philocheras fasciatus, Processa edulis, Pilumnus hirtellus and Pirimela denticulata (the latter with a number of specimens similar to that of A. hyndmanni) are abundant in Calahonda, while Calcinus tubularis, Clibanarius erythropus, A. hyndmanni, Hippolyte inermis and Atanas nitescens are dominant in Punta de Calaburras. These differences are related to substrate composition, different sediment granulometry (as consequence of hydrodynamic conditions), and/or interaction with adjacent biotopes (see below). Spatial variations in seagrass faunal assemblages are common and have been attributed to the impact of nearby habitats (Skilleter et al. 2005) and physical factors. For example, Bowden et al. (2001) highlight a relation between the number of species and sediment variations due to tidal currents and waves. In our study, adjacent habitats determine the presence and relative abundance of specific species in these small meadows, since they are interspersed with Posidonia, sands, rocks/pebbles and seaweeds. Nevertheless H. leptocerus is always the dominant species and the diversity indices of the assemblages show similar values. Decapod assemblages associated with Cymodocea nodosa 75 The Calahonda meadow is located in a sheltered area protected by large rocks, which allows the settlement of well-sorted fine sands with little mud. The influence of sandy substratum is proven by species such as P. fasciatus, which lives buried in sediment and is a characteristic species of “SFBC” well-sorted fine sandy bottoms (Pérès and Picard 1964, Massé 1972, etc). Processa edulis is a typical species of seagrass leaf stratum (Ledoyer 1966, 1968; Harmelin 1964) and Caulerpa (López de la Rosa et al. 2006), but was also found in adjacent sandy bottoms, hiding during daytime in the sediment and eating during nighttime in the meadow, since it is a predator of other invertebrates (Ledoyer 1966, Chessa et al. 1989). Polybius navigator and Sycionia carinata, although less abundant, also contribute to the characterization of Calahonda Cymodocea patches but are found in adjacent algae bottoms, hidden in the sandy substrate (García Raso et al. 2006, Lopez de la Rosa et al. 2006, Števčić 1991). The meadow of Punta de Calaburras is not protected by large rocks and contains many pebbles. The dominance of the hermit crab Clibanarius erythropus in this shallow meadow is related to the presence and abundance of pebbles (Zariquiey 1968, Gherardi 1990, 1991). Other species, which also contribute to the characterization of this assemblage but with lower abundances, such as Calcinus tubularis, Athanas nitescens and Hippolyte inermis, characterize Posidonia oceanica seagrass; the latter in the leaf stratum (Ledoyer 1966) and the two former in the rhizome stratum (together with another hermit crab Cestopagurus timidus) (García Raso 1990). Calcinus tubularis and Athanas nitescens also live on shallow rocky bottoms; in fact, the rhizome stratum of Posidonia oceanica is considered to be comparable to a hard bottom (BellanSantini et al. 1994). These differences and similarities between stations are reflected in the aggregation and ordination analyses. Furthermore, decapods are mobile animals. Therefore, with some exceptions, their presence cannot be “exclusively” associated with a single biotope, unlike other groups such as molluscs, with reduced mobility or higher dependence on a specific sediment or biotope (Hemminga and Duarte 2000; Luque and Templado 2004). Actually, the dissimilarity and characterization of analogous decapod assemblages Chapter 3 76 (among “hard bottoms”, or “soft bottoms”) at similar depths relies heavily on quantitative values (García Raso et al. 1996, García Muñoz et al. 2008) and this holds true also for some other groups (Nakaoka et al. 2001). Analyzing the effect of habitat fragmentation on the macroinvertebrate communities associated with Zostera marina, Frost et al. (1999) mention: “differences were represented by subtle shifts in the relative abundance of a few common species rather than by larger shifts or wholesale species replacement”. The decapod abundance (mean = 18.4 ind m-2) is low when compared to that of Posidonia oceanica meadows (more complex habitat, with more strata and ecological niches, García Raso 1990) but higher than that found in Mediterranean Cymodocea meadows (García Raso et al. 2006, Števčić 1991). Middle to low values in the diversity indexes were observed, but higher than those found in Cymodocea assemblages from Almeria (only daytime samples, García Raso et al. 2006) and some seaweed assemblages of Caulerpa prolifera in Cadiz (López de la Rosa et al. 2002, 2006). This is the result of a more equitable distribution (highest evenness values), probably related to an ecotone effect, since the studied Cymodocea meadow is structured into patches interspersed with other habitats representing different ecological niches and including different associated species and dominances. Moreover, as already mentioned, mobility enables decapods easier access to new resources, thus resulting in a mixed and more equitable community. Hence, crustaceans (such as palaemonid shrimps and some peracarids) inhabiting fragmented Zostera seagrass meadows showed increased abundance at the boundary between sand and seagrass (Tanner 2005). In any case, the values found are well below those of Posidonia oceanica meadows (García Raso 1990). The positive correlation observed between specimen numbers, species richness, and seagrass phenology (with seasonal differences in the number of leaves per shoot) points out a direct (through their cycles) or indirect (through associated resources) relation between plant and animal communities. However, other factors (e.g, temperature, photoperiod) could be influencing both variables. Hence, high abundances of most species coincide with maximum seagrass development Decapod assemblages associated with Posidonia oceanica 83 Abstract The decapod assemblage associated with a Posidonia oceanica meadow located near its western limit of biogeographic distribution was studied over an annual cycle. Fauna samples were taken seasonally over a year (five replicates per season) in two sites located 7 km apart, using a non-destructive sampling method (airlift sampler) for the seagrass. The dominant species of the assemblage, Pisidia longimana, Pilumnus hirtellus and Athanas nitescens, were associated with the protective rhizome stratum, which is mainly used as a nursery. The correlations between decapod assemblage structure and some phenological parameters of the seagrass shoots and wave height were negative or null, which reflects that species associated with the rhizome had a higher importance than those associated with the leaf stratum. The abundance and composition of the decapods assemblage as well as the ecological indexes displayed a seasonality trend with maximum values in summer-autumn and minimum in winter-spring, which were related to the seawater temperature and the recruitment periods of the dominant species. The spatial differences found in the structure and dynamics of the assemblages may be due to variations in the recruitment of the dominant species, probably as a result of the influence of local factors (e.g. temperature, currents) and the high dispersal ability of decapods, together with the patchy configuration and the surrounding habitats. The studied meadows are fragmented and are integrated within a mosaic of habitats (Cymodocea nodosa patches, algal meadows, rocky and sandy bottoms), which promotes the movement of individuals and species among them, maintaining a high species richness and evenness. Keywords: Alboran Sea, decapod assemblages, fragmented, Posidonia oceanica, temporal dynamics. 5IJT BSUJDMF DBO CF EPXOMPBEFE GSPN IUUQPOMJOFMJCSBSZXJMFZDPNEPJNBFDBCTUSBDU Chapter 4 84 Introduction Posidonia oceanica (Linnaeus) Delile is the most abundant of the four indigenous Mediterranean seagrasses and forms the most extensive beds, with a total area in the Mediterranean Sea ranging between 25,000 and 45,000 km2 (Luque and Templado 2004). The architecture of P. oceanica beds, with different strata and microhabitats (leaves, rhizomes and interspersed soft sediments), supports different biotic assemblages (Pérès and Picard 1964; Kikuchi and Pérès 1977; Kikuchi 1980) and, in addition, offers shelter and acts as a nursery area for many animals, including some species of commercial importance (Gillanders 2006). Mixed photophilous and sciaphilic assemblages, including crustacean decapods, coexist (Luque and Templado 2004; García Raso et al. 2006a). Grazer species such as those belonging to the genus Hyppolite are associated with the upper, well-lit strata leaves), while the lower strata (rhizomes) acts as ‘hard and dark strata’, providing refuge and protection for many sciaphilic species. The high faunistic diversity results in a large number of interand intra-specific strategies and movements, which are further governed by intrinsic and extrinsic factors such as seasonal periods, food resources (e.g. epiphytes), depth, predation and the exchange of species from adjoining biotopes (García Raso et al. 1996; Borg et al. 2010). The latter can be favored by the seagrass meadows fragmentation because it produces a patchy configuration that may be interspersed with other habitats (other seagrass species, macroalgae, hard or soft bottoms) (Barberá-Cebrián et al. 2002; Jackson et al. 2006; Mateo-Ramírez and García Raso 2012; Urra et al. 2013a). In the Alboran Sea, Posidonia oceanica forms extensive meadows in its easternmost sector, close to the Almeria–Oran Front, whereas in the central and western sectors these are reduced from small meadows to patches, mainly existing over hard substrata (Luque and Templado 2004;Junta de Andalucía 2011). The westernmost distributional limit for P. oceanica is located between Cádiz and Málaga provinces (Southern Spain) in the Northern Alboran Sea, and in Sebkha-bou-Areg and the Chafarines Islands in the Southern Alboran Sea (PérezLloréns et al. 2014). In general, seagrass beds exhibit a natural fragmentation due to seasonal growth, die-off, storm events or hydrodynamic action (Duarte and Sand-Jensen 1990; Fonseca 1992; Abbate et al. 2000). The influence of Atlantic Decapod assemblages associated with Posidonia oceanica 85 waters with lower salinities likely plays a role in the fragmentation of meadows in the Alboran Sea (Luque and Templado 2004). However, anthropogenic impacts, such as coastal constructions (Ruíz and Romero 2003), illegal bottom trawling (González-Correa et al. 2005; Rueda et al. 2009), beach replenishment (González Correa et al. 2008) and organic pollution due to fish cultures (Ruíz et al. 2001), may also result in extensive fragmentation, which in turn may promote changes in the structure of the associated animal community. This alteration may have both positive and negative effects on the associated faunal assemblages (Eggleston et al. 1998; Urra et al. 2013a). Many studies have been carried out on the fauna associated with P. oceanica beds (e.g. Templado 1984; Gambi et al. 1992; Sánchez-Jerez et al. 1999, 2000; Barberá-Cebrián et al. 2002; Dimech et al. 2002; Ateş et al. 2005,2006; Como et al. 2008; Borg et al. 2010; Urra et al. 2013)a, but few of them have focused on the structure of the decapod assemblages (García Raso 1990a; Borg and Schembri 2000). Besides, there is no information about this topic for the westernmost distribution area of P. oceanica (close to the Strait of Gibraltar), where there is a high influence of Atlantic waters, and P. oceanica only forms small patches on rocky bottoms. The present study analysed the crustacean decapods assemblage associated with P. oceanica fragmented meadows inhabiting the Atlantic–Mediterranean transition area (close to the Strait of Gibraltar). The main goals were to determine if the geographic location, configuration of the meadows (i.e. patchiness), and phenological and environmental parameters play a role in the composition and structure of the associated decapod populations. Material and Methods Study area The study was carried out within the site of community importance (SCI) known as ‘Calahonda’ (code ES6170030), which is included in the Natura 2000 network and located off the Mijas coast (Málaga, Southern Spain; Fig. 1). Two sites were sampled: Punta de Calaburras (PC; 36º30´23´´ N, 04º38410´´ W) and Chapter 4 86 Calahonda (CH; 36º29´21´´ N, 04º41´55´´ W). PC is subjected to more wave action, resulting in a higher occurrence of pebbles than at CH, which is a more sheltered site with large rocks surrounded by soft sediments. The meadows studied are located close to the westernmost distributional limit of Posidonia oceanica (Luque and Templado 2004) and display a fragmented configuration formed by patches with diameters ranging between 1 and 130 m2, with a coverage of c. 14% (Urra et al. 2011b). In addition, it grows preferentially on hard substrata (rocks) and in areas that are sheltered by shallow rocky outcrops (2–5 m in depth). The presence of dead rhizomes covered by algae (i.e. Halopteris scoparia and Asparagopsis armata) interspersed with living seagrass meadows indicates that the P. oceanica meadows occupied a larger area in the past. Extensive macroalgae forests down to 5–7 m and patches of the small seagrass Cymodocea nodosa (coverage <5%), which frequently forms mixed meadows with P. oceanica, are also found in the area. Salinity remains almost constant throughout the year due to the low freshwater input in the area, ranging between 36.5 in fall and 36.8 in spring (salinity data given in practical salinity scale; and taken from Instituto Español de Oceanografía: http://www.ma.ieo.es/gcc/sistemas_ observacion.htm). Solar Fig. 1. Study area showing the location of sampling sites in the NW Alboran Sea. PC, Punta Calaburras; CH, Calahonda. Decapod assemblages associated with Posidonia oceanica 87 irradiance at 2 m depth at the sampling sites displays a seasonal pattern with high values in summer (3.95 ± 0.1 log lm·m-2), intermediate in spring and fall and low values in winter (3.47 ± 0.15log lm·m-2) (Á. Mateo-Ramírez, unpublished data). Sample collection and sampling methodology Samples were collected by SCUBA divers at a depth of 2 m in seagrass patches of different sizes: small (1–10 m2 area), medium (10–30 m2) and large (>30 m2). Sampling took place during daytime (in the morning) and seasonally in July (summer), November 2007 (fall), January (winter) and April 2008 (spring). In total, 40 faunistic samples (five replicates per site and season) were taken using a quadrat measuring 50 × 50 cm and an air-lift sampler (Borg and Schembri 2000; Mateo-Ramírez and García Raso 2012). The suction time was the same for each sample collected (3 min). This sampling methodology is environmentally friendly for seagrass and is a basic requirement in studies carried out in marine protected areas, such as this SCI, and is similar to the methodology practiced by MateoRamírez and García Raso (2012) and Urra et al. (2013a) in this SCI. In addition, it is more effective than the use of hand nets (Borg and Schembri 2000) because it collects the species living in all strata (leaves and rhizome) and therefore the assemblage can be better characterized. In the laboratory, every faunistic sample was sieved over mesh sizes down to 0.5 mm, storing each size fraction in 70% ethanol. Each species was identified and the numbers of individuals were counted. The abundance of decapods is expressed as the number of individuals per sample. The sizes of the dominant species were analysed in order to distinguish two groups: juveniles (small size) and adults (large size). The minimum size of adults was determined from the literature (Zariquiey Álvarez 1968; Manjón-Cabeza and García Raso 1994) and our personal data (collected samples). To determine the shoot density (shoots·m-2) in each season, five replicates of a quadrat measuring 50 × 50 cm were used in both sites (PC and CH) in the same patches where fauna was collected. In addition, 10 randomly chosen shoots in each quadrat were analysed in situ (n = 50 shoots per site and season) in order to determine the shoot parameters. Number of leaves per shoot (leaves·shoot-1), Chapter 4 88 the leaf height (from the basal part of the sheath to the blade tip) and the leaf width (at the mid-point between the sheath and the blade tip) of the largest leaf were measured. Seawater temperatures (T) and water samples for analysis of the concentration of chlorophyll a (Chl a) were measured on different days throughout each season (weeks before, during and after data collection). Two 1-l replicates of seawater (at the surface) were collected in each site and transported in darkness at low temperature to the laboratory for Chl a determination. Pigment analyses were carried out by filtering through Whatman GF/C glass filters. The pigments of the retained cells were then extracted using 100% acetone for 12 h in cool and dark conditions. The solution was measured using a spectrophotometer at wavelengths of 630, 647, 664 and 750 nm. The Chl a concentrations were obtained using the equation proposed by Jeffrey and Humphrey (1975). Samples of sediment were also taken on each sampling occasion (five replicates per season and site) in order to estimate the percentage of organic matter (%OM) within the sampled P. oceanica patches. This percentage was calculated by the weight loss of dry sediment (three subsamples of 20 g per replicate) after ignition at 500°C for 1 h. Wave height (m) information obtained with the High Resolution Limited Area Model and Wave Model numeric models was taken from the Instituto Nacional de Meteorologia (Spain). Statistical analyses The composition, structure and seasonal changes of the decapod assemblages were analysed from the values of abundance [N, total number; NJ, number of juveniles (small size); NA, number of adults (large size)], frequency index (Fi; percentage of samples in which a particular species is present), dominance index (Di; percentage of individuals of a particular species within the sample), species richness (S) and the ecological indexes of Shannon–Wiener diversity (H’) and evenness (J’). The ecological indices were obtained using the PRIMER v. 6 software (Clarke and Gorley 2006). Decapod assemblages associated with Posidonia oceanica 89 Two-way permutational multivariate analysis of variance (PERMANOVA) were used to test for differences in the environmental variables, the seagrass parameters and the ecological indexes between sites (PC, CH) and months/seasons July 2007 (summer), November 2007 (fall), January 2008 (winter) and April 2008 (spring). One-way PERMANOVA were used to test for differences between seagrass patch areas and to analyse the seasonality in the abundance of the main species. Analyses were based on Euclidean distances. The significance of P values was determined through 9999 permutations of residuals under a reduced model or of the raw data, for twoor one-way analysis, respectively (PERMANOVA + for PRIMER, Anderson et al. 2008). The analysis of similarity (ANOSIM) procedure was carried out for multivariate statistical comparisons of groups of samples according to the different factors considered (sites, months/seasons and patch areas) and the similarity percentages analysis (SIMPER) was used to identify those species that contributed to the similarity and dissimilarity among these groups. The similarity matrix used for ANOSIM analysis was calculated using Bray-Curtis index. The data were double square transformed. Both analyses were executed using the PRIMER v.6 software (Clarke and Gorley 2006). Correlations between ecological values (S, N, NJ, NA, H’ and J’) with seagrass parameters (shoot density, number of leaves per shoot, leaf height and width, patch area) and environmental variables (T, Chl a, %OM and wave height) were calculated by Pearson co-efficients. Finally, canonical correspondence analysis (using the software CANOCO) was performed to study the relationships among the seagrass parameters, environmental variables and the abundances of the top dominant decapod species. Results Environmental variables None of the analysed environmental variables displayed significant differences between sites (Table 1). Seawater temperature showed a significant temporal trend with maximum values in summer (20–24.5 °C) and minimum ones in winter (15–16.5 °C) at both sites (PC, Pseudo-F = 22.525, P < 0.01; CH, Pseudo-F = 4.478, P < 0.05). The Chl a concentration displayed a significant temporal pattern Chapter 4 90 at CH (Pseudo-F = 2.836, P < 0.05), with values ranging from 24.34 µg·l-1 in spring to 3.22 µg·l-1 in summer, whereas at PC the temporal changes were nonsignificant (Pseudo-F = 0.079, P = 0.969) and the highest values were detected in fall (12.31 µg·l-1). The %OM displayed a significant temporal trend at CH (Pseudo-F = 5.310, P < 0.01), with maximum values in summer (1.61–2.84%) and minimum ones in winter (0.99–2.11%), whereas at PC temporal changes were non-significant (Pseudo-F = 2.245, P = 0.097) and the highest value was observed in winter (1.44–2.65%). Wave height showed non-significant temporal differences at both sites (PC, Pseudo-F = 1.183, P = 0.334; CH, Pseudo-F = 2.852, P = 0.055), with maximum values (c. 2 m wave height) in fall for PC and in spring for CH, and minimum values in spring for PC (0.7 m) and in fall for CH (0.6 m). Phenology of Posidonia oceanica All phenological parameters of Posidonia oceanica displayed significant differences among seasons, but not between sites (Table 1). Shoot density showed a significant temporal trend at CH with maximum values in fall (856–1180 shoots·m-2; Pseudo-F = 10.871, P < 0.001), while at PC the maximum values occurred in winter (840–980 shoots·m-2), but without temporal significant differences (Pseudo-F = 0.926, P = 0.44) (Fig. 2A). Leaf number displayed significant temporal changes at CH (Pseudo-F = 7.681, P < 0.01) and PC (Pseudo-F = 33.518, P < 0.001), with maximum values in winter (6 leaves·shoot-1; Fig. 2B). Leaf height showed a significant temporal trend with maximum values in spring at both CH (29–46 cm, Pseudo-F = 56.863, P < 0.001) and PC (35–45 cm, Pseudo-F = 18.513, P < 0.001) (Fig. 2C). Leaf width displayed a significant temporal trend at both sites (CH, Pseudo-F = 7.410, P < 0.01; PC, Pseudo-F = 6.340, P < 0.01), with maximum values in winter (10–10.6 mm; Fig. 2D). The last two seagrass parameters presented non-significant differences between different patch areas (Table 1). Decapod assemblages associated with Posidonia oceanica 91 Table 1. Results of the two-way Permutational Multivariate Analysis of Variance to test the significance of the seagrass parameters (shoot density,leaf height and width in cm, number of leaves) and environmental variables (wave height, temperature, Chl a concentration and % organic matter) between sites and among seasons and areas of seagrass patches. Factors n Site Season Season*Site Environmental factors Wave heigth 54 P-F=0.228; P=0.644 P-F=2.189; P=0.103 P-F=1.081; P=0 369 Temperature 27 P-F=0.153; P=0.694 P-F=9.401; P<0.01** P-F=1.938; P=0.163 Chlorophyll a60 P-F=0.200; P=0.654 P-F=2 216; P=0.089 P-F=1.885; P=0.143 % Organic matter 106 P-F=6.641E-4; P=0.98 P-F=3.145; P<0.05* P-F=4.842; P<0.01** Phenological factors Shoot density 38 P-F=0.210; P=0.650 P-F=8.150; P<0.001*** P-F=6.651; P<0.01** Leaf height 38 P-F=9.420E-2; P=0.762 P-F=50441; P<0.001*** P-F=1.021; P=0.400 Leaf width 38 P-F=1.412; P=0 245 P-F=12.721; P<0.001*** P-F=1.585; P=0 208 Number of leaves 38 P-F=3.799; P=0.064 P-F=25.119; P<0.001*** P-F=5.425; P<0.01** Phenological factors nPatch area Patch area*Site Shoot density 19 P-F=0.278; P=0.753 P-F=0.555; P=0.586 Leaf height 19 P-F=2.240; P=0.138 P-F=0.130; P=0.883 Leaf width 19 P-F=2.629; P=0.105 P-F=0.218; P=0.797 Number of leaves 19 P-F=1.066; P=0.369 P-F=1.196; P=0 336 P-F, Pseudo-F values. *Significant differences at P < 0.05; ** at P < 0.01 and *** at P < 0.001. Composition and structure of the decapod assemblage A total of 1185 individuals were collected, belonging to 34 species and 18 families (Table 2). The family Inachidae (four spp.) was the best represented among decapods, followed by Hippolytidae, Alpheidae, Paguridae and Epialtidae (with three spp. each). The family Porcellanidae was the most abundant one [227 individuals (ind.)], followed by Alpheidae (184 ind.), Pilumnidae (183 ind.) and Paguridae (160 ind.). The assemblage displayed 16 species with dominance values higher than 1%, with Pisidia longimana and Pilumnus hirtellus as the top dominant species (18.7% and 15.4% Di, respectively), followed by Athanas nitescens, Cestopagurus timidus, Calcinus tubularis and Achaeus gracilis (>5% Di in all cases). Most of these species were also recorded as the most frequent species (Table 2). Chapter 4 92 Fig. 2. Seasonal trends of parameters of Posidonia oceanica. (A) Shoot density (shoots·m-2), (B) number of leaves (leaves·shoot-1), (C) leaf height (cm) and (D) leaf width (cm) at Punta Calaburras (PC, solid bars) and Calahonda (CH, empty bars). Mean + SE. Letters above error bars display the results of PAR-WISE tests; different letters indicate significantly different means at P < 0.05. Capital letters refer to PC comparisons and lower-case letters to CH comparisons. Decapod assemblages associated with Posidonia oceanica 99 Table 4. Results of similarity percentages analyses. Species ranked according to their average withingroup similarity within seasons (Punta de Calaburras and Calahonda). Species Av.Abund Av.Sim Sim/SD Contrib% Cum.% Summer P. hirtellus 1.54 14.72 5.94 30.17 30.17 P. longimana 1.55 13.24 1.74 27.13 57.30 C. timidus 0.95 4.83 0.91 9.90 67.20 C.tubularis 0.81 4.32 0.69 8.85 76.04 A. gracilis 0.73 3.68 0.68 7.54 83.58 P. anacharetus 0.52 2.31 0.52 4.74 88.32 Autumn A. nitescens 1.69 8.32 4.85 15.62 15.62 P. hirtellus 1.51 6.53 1.84 12.26 27.87 P. longimana 1.45 6.27 1.80 11.77 39.65 P. robusta 1.17 5.87 1.66 11.02 50.67 A. gracilis 1.12 4.32 1.24 8.10 58.77 H. leptocerus 0.79 3.40 0.92 6.38 65.15 X. hydrophilus 0.91 3.04 0.89 5.70 70.85 E. cranchii 0.91 2.96 0.90 5.55 76.40 C. tubularis 0.82 1.99 0.69 3.74 80.15 P. carinimana 0.65 1.99 0.68 3.73 83.87 C. timidus 0.74 1.89 0.70 3.55 87.43 E. edwardsii 0.57 1.38 0.53 2.59 90.02 Winter P. longimana 0.96 8.96 1.10 26.87 26.87 P. robusta 0.81 6.77 0.78 20.31 47.18 X. hydrophilus 0.67 5.96 0.78 17.87 65.05 P. hirtellus 0.47 2.22 0.42 6.67 71.72 A. nitescens 0.53 1.96 0.43 5.88 77.60 A. gracilis 0.35 1.29 0.30 3.86 81.46 E. cranchii 0.39 1.27 0.29 3.81 85.27 C. tubularis 0.50 1.05 0.29 3.16 88.44 E. edwardsii 0.35 1.03 0.29 3.10 91.53 Spring C. timidus 1.13 9.46 1.70 21.55 21.55 A. hyndmanni 1.10 8.55 1.20 19.47 41.02 P. hirtellus 0.82 5.35 0.89 12.18 53.20 P. longimana 0.89 4.45 0.68 10.13 63.34 C. tubularis 0.73 4.14 0.66 9.43 72.76 P. anacharetus 0.65 3.21 0.68 7.31 80.07 A. gracilis 0.60 2.79 0.52 6.36 86.43 X. hydrophilus 0.62 2.33 0.52 5.32 91.74 Av. abund., average abundance; ave sim., average similarity; sim./SD, similarity/standard deviation; contrib.%, average percentage contribution to similarity; cum.%, cumulative percentage similarity. Chapter 4 100 leaf height. No relationships were found between the ecological indexes and patch area or wave height (Table 5). Canonical correspondence analysis indicated that the first two axes accounted for 75.2% of the total variance in the species-environment relationships and 24.3% of the species variance. Leaf height (0.502), percentage of organic matter (0.423), leaf number (-0.428) and shoot density (-0.348) showed the highest correlations with axis I, whereas seawater temperature (0.686), leaf width (-0.614) and wave height (-0.591) presented the highest correlations with axis II. Forward selection indicated that leaf height (F = 4.03; P < 0.01) and seawater temperature (F = 3.56; P < 0.01) could explain most of the variance in the species data. The scatter diagram showed an ordination of the samples in relation to seasons of the year. The samples scattered along axis II following an increase in seawater Table 5. Relationships (Pearson correlation) between the species richness (S), total abundance (N), adult abundance (NA), juvenile (small size) abundance (NJ) evenness (J’) and Shannon–Wiener diversity (H’) with the seagrass parameters (shoot density, number of leaves, leaf height and width), patch area and the environmental variables [wave height, temperature, Chl a concentration and % organic matter (%OM)]. Index TChl a %O.M. Wave height Shoot density Number of leaves Leaf height Leaf width Patch area SR=0.377; P<0.05* R=-0.360; P<0.05* R=0.102; P=0.535 R=0.070; P=0.673 R=0.345; P<0.05* R=-0.158; P=0.336 R=-0.581; P<0.01** R=-0.353; P<0.05* R=-0.198; P=0.416 NR=0.393; P<0.05* R=-0.247; P=0.130 R=0.003; P=0.987 R=0.154; P=0.350 R=0.072; P=0.663 R=-0.145; P=0.378 R=-0.506; P<0.01** R=-0.205; P=0.211 R=-0.218; P=0.370 NA R=0.187; P=0.254 R=-0.163; P=0.322 R=0.039; P=0.812 R=0.266; P=0.102 R=0.112; P=0.499 R=-0.134; P=0.417 R=-0.337; P<0.05* R= -0.101; P=0.542 R=-0.151; P=0.424 NJ R=0.512; P<0.01** R=-0.283; P=0.080 R=-0.290; P=0.862 R=0.038; P=0.820 R=0.029; P=0.863 R=-0.134; P=0.414 R=-0.578; P<0.01** R=-0.264; P=0.104 R=-0.296; P=0.112 JR=-0.548; P<0.01** R=0.363; P<0.05* R=-0.270; P=0.871 R=0.248; P=0.127 R=0.085; P=0.609 R=0.372; P<0.05* R=0.147; P=0.370 R=0.411; P<0.01** R=0.065; P=0.793 HR=0.150; P=0.362 R=-0.245; P=0.132 R=0.166; P=0.313 R=0.138; P=0.401 R=0.393; P<0.05* R=-0.760; P=0.645 R=-0.438; P<0.01** R= -0.231; P=0.157 R=-0.089; P=0.718 *Significant differences at P < 0.05 and ** at P < 0.01. Decapod assemblages associated with Posidonia oceanica 101 temperature from the positive to the negative part of the axis (Fig. 5). The majority of the species is near the origin of axes (showing a poorly differentiated profile distribution). Nevertheless, some species appeared more separated in relation to axis I, indicating a possible relationship with the phenological parameters (such as Processa robusta; Fig. 5). Fig. 5. Plot from the canonical correspondence analysis of environmental variables, seagrass parameters and top dominant species in relation to axes I and II (eigenvalues: 0.118 and 0.085, respectively). Temp, seawater temperature; O.M., percentage of organic matter; Chl a, chlorophyll a concentration; Wave, wave height; Density, shoot density; Leaf N, leaf number; Height, leaf height and Width, leaf width. ▲ summer; △ fall; ■ winter; ■ spring. Chapter 4 102 Discussion Phenology of Posidonia oceanica The temporal trend observed for Posidonia oceanica meadows of PC and CH (Northwestern Alboran Sea) is similar to that reported for other Mediterranean meadows of this species (Eastern Spain: Sánchez Lizaso 1993; Italy: Guidetti et al. 2002). The meadows studied here were characterized by a patchy structure, showing higher shoot density and lower leaf height values in comparison to those from meadows located at similar depths in other Mediterranean areas (Northeast Spain: Marbà et al. 1996; Italy: Buia et al. 1992; Greece: Amoutzpoulou-Schina and Haritonidis 2005). However, Marbà et al. (1996) studied P. oceanica meadows in the Northeastern Alboran Sea (close to the Almería–Oran Front) and observed similar phenological characteristics (i.e. high shoot densities and low leaf heights) to those of this study (Northwestern Alboran), suggesting that local factors may play a key role in controlling the growth of P. oceanica. These factors and probably the shallow location of the studied meadows (high wave exposure) at PC and CH may have important effects upon the phenological parameters of P. oceanica. Composition and structure of the decapod assemblages The decapod assemblages inhabiting the Posidonia oceanica meadows of PC and CH were dominated by species mainly associated with the rhizome stratum, such as Pisidia longimana, Pilumnus hirtellus, Athanas nitescens, Calcinus tubularis and Cestopagurus timidus. This faunistic composition showed similarities with those reported for other P. oceanica meadows (García Raso 1990a; Borg and Schembri 2000; Sánchez-Jerez et al. 2000; Box Centeno 2008). García Raso (1990a) observed that abundances of the hermit crabs C. timidus and C. tubularis (dominant species in P. oceanica meadows in South-eastern Spain and abundant in the present study area) depended upon the bathymetry (e.g. C. tubularis prefers deeper bottoms) as well as local factors such as environmental quality, surface currents and shell availability (Zupo et al. 1989; García Raso 1990a; Belci et al. 2010). By contrast, the geographic location of the Alboran Sea, with the acute Atlantic influence, marks some differences such as the absence of some Decapod assemblages associated with Posidonia oceanica 103 Mediterranean species commonly found in P. oceanica meadows located eastward of the Almería–Oran Front, and the presence of typical Atlantic species that do not extend towards the Mediterranean Sea. This is the case for the hermit crab Pagurus chevreuxi, found in areas of the Central and Western Mediterranean Sea (Zariquiey Álvarez 1968; Templado 1984; Borg and Schembri 2000) but absent from the present study area, and Anapagurus hyndmanni, an Atlantic species that occurs in the Alboran Sea (García Raso 1982; García-Gómez 1994; García Muñoz et al. 2008) but is absent eastward of the Almería–Oran Front. A similar patron had been observed for mollusks (Urra et al. 2013a). The methodology used in this study (air-lift sampler) allowed the collection of species linked to the foliar stratum, such as the shrimp Hyppolite inermis and the hermit crab C. timidus, both cited as dominant species in other P. oceanica meadows (Ledoyer 1966; Borg and Schembri 2000; Belci et al. 2011). The hippolytid H. inermis is a characteristic species of this stratum, whereas C. timidus makes vertical movements towards the apical part of the shoots or along them in daytime to feed (Ledoyer 1966; Borg and Schembri 2000; Mateo-Ramírez and García Raso 2012). However, the species associated with the rhizome stratum generally make a greater contribution to the faunistic richness of P. oceanica meadows, with a higher diversity of decapods than the leaf stratum (Templado 1984; Borg and Schembri 2000). This was also observed in PC and CH, where the majority of the dominant decapods are sciaphilic species associated with hard substrata (i.e. the rhizomes), which provide ideal conditions for adults and shelter for juveniles of different species. This is the case for P. hirtellus, a common crab species in infralittoral rocky bottoms that also uses the P. oceanica rhizomes and associated calcareous algae concretions as nursery areas, as observed in this and other studies (Vadon 1981; García Raso and Fernández Muñoz 1987; GarcíaRaso 1988). The high dispersal ability of decapods, together with the patchy configuration of the studied meadows and the presence of different habitats, rocky outcrops and algal stands interspersed among them, may explain why the most common species in this assemblage also occur in adjacent habitats. In this context, the species Achaeus gracilis, Achantonyx lunulatus, Sirpus zariquieyi and Eualus cranchii were collected in the surrounding photophilous algae meadows (Mateo-Ramírez A., Chapter 4 104 Urra J., Marina P., Rueda J.L., García Raso J.E. unpublished data). The latter is an opportunistic species that can occur in biotopes with cavities and on photophilic seaweed (García Raso 1990b), as well as in the leaf stratum of P. oceanica (Borg and Schembri 2000). The species Xantho hydrophilus and Porcellana platycheles were found under pebbles and small rocks (Vadon 1981; Mateo-Ramírez A., Urra J., Marina P., Rueda J.L., García Raso J.E. unpublished data), and Processa robusta and A. hyndmanni were found inhabiting soft bottoms, preferentially detritic bottoms for A. hyndmanni (García Muñoz et al. 2008). The presence of juveniles of A. hyndmanni in spring in both P. oceanica and Cymodocea nodosa meadows (MateoRamírez and García Raso 2012) may indicate that this species uses the seagrass meadows as nursery areas. This connectivity between surrounding habitats, with high levels of movement of species, has been reported in P. oceanica meadows throughout the Mediterranean Sea (García Raso 1990a; Scipione et al. 1996; Borg and Schembri 2000), in other seagrasses such as Cymodocea nodosa (García Raso et al. 2006; Mateo-Ramíez and García Raso 2012; Daoulatli et al. 2014) and in other faunistic groups with lower mobility than decapods such as mollusks (Urra et al. 2013). The ability to inhabit different habitats and the R-reproductive strategies of certain species, such as Alpheus dentipes, P. hirtellus, C. timidus and P. longimana (with long reproductive periods, from January–February to October–November), allow them to display a high colonization capacity, with the presence of juveniles throughout the year and one or two annual recruitment peaks (García Raso and Fernández Muñoz 1987; García Raso 1988, 1990a; López de la Rosa and García Raso 1992; Manjón-Cabeza and García Raso 1994; Belci 2013). These peaks can be delayed or advanced temporally and can show different intensities at different sites or depths in response to environmental changes induced by local factors (e.g. temperature, food availability), entailing changes in the size of the species’ populations or decapod assemblages (e.g. C. timidus) (Gambi et al. 1992; Borg and Schembri 2000). Besides, the high dispersal capacity of many decapod species (adults and larvae; Grantham et al. 2003), the local currents and tide influence in the study area, close to the Atlantic Ocean, could intensify the incorporation and/ or recruitment of species from adjacent habitats occurring at different areas and depths (Belci 2013). Decapod assemblages associated with Posidonia oceanica 105 Temporal dynamics of decapod assemblages and relationships with the environmental variables and seagrass parameters The studied decapod assemblages showed a significant temporal trend mainly related to the recruitment events of the dominant species, as reflected by the low number of correlations between the ecological indexes and the seagrass parameters. Most of the species occurring in these seagrass beds are sciaphilic ones and therefore not strictly dependent upon the seagrass seasonal cycles associated with the leaf stratum. These species usually search for shelter in sciaphilous micro-habitats, such as seagrass rhizomes and calcareous concretions, especially for juveniles. This could be the reason why the dominant species clustered near the origin in the CCA ordination analysis. Nevertheless, other species such as Cestopagurus timidus and Processa spp. are more linked to the foliar stratum (Borg and Schembri 2000; Belci 2013; Daoulatli et al. 2014), as they usually feed on epiphytes and prey that occur on the seagrass leaves. The relationships with seagrass (leaves) can be indirect, and changes in meadow parameters may act as drivers for recruitment (sometimes with monthly delays, Daoulatli et al. 2014). Cestopagurus timidus can play a key role in the assemblage dynamics and in the food web at the leaf stratum level, as observed by Gambi et al. (1992), even more so than Hyppolite inermis because of its higher abundance. The total abundance (N), species richness (S) and diversity index (H‘) reached their maximum values in fall when Posidonia oceanica displayed its minimum leaf height (Figs 2C and 3B). Similar seasonal trends have been observed in decapod assemblages inhabiting P. oceanica meadows throughout the Mediterranean Sea (García Raso 1990a; Scipione et al. 1996; Borg and Schembri 2000), and also in calcareous algae concretions linked to rhizomes of P. oceanica (García Raso 1988). These trends are coincident with the recruitment peaks of some of the dominant species, such as Athanas nitescens and Pilumnus hirtellus in fall (November) and Pisidia longimana in summer (July) (Fig. 4B and C), which are in agreement with the observations made by López de la Rosa and García Raso (1992). Moreover, positive correlations were observed between the total abundance and the abundance of juveniles (small sizes; NJ) with the seawater temperature in both PC and CH (Table 5). All this may suggest that the temporal–seasonal trends of the studied Chapter 4 106 decapod assemblages are mainly related to recruitment events of the dominant species and to seawater temperature rather than to seasonal phenological changes of P. oceanica meadows. Influence of the patchy structure of the seagrass meadows on the decapod assemblages The studied decapod assemblages occur in fragmented Posidonia oceanica meadows located close to the westernmost geographic limit for this seagrass and, therefore, display a different configuration than those located at similar depths in the eastern part of the Alboran Sea or in the Mediterranean Sea (Luque and Templado 2004). The structure of seagrass meadows (continuous or fragmented) and the effect of fragmentation on the associated invertebrate communities are controversial themes. The process of habitat fragmentation, in addition to habitat loss, results in three other effects: an increase in the number of patches, a decrease in patch areas and an increase in the isolation of patches (Bell et al. 2001; Fahrig 2003). Thus, different combinations of these effects may promote different responses on the fauna associated with seagrass meadows. Healy and Hovel (2004) found that total epifaunal density was significantly lower in patchy beds than in continuous or very patchy beds, whereas species richness was higher in very patchy beds. In Zostera marina, Bowden et al. (2001) found a positive species-area relationship and Jackson et al. (2006) found that total decapod density increased when seagrass patches form part of a landscape. Other studies have indicated that fragmented seagrass meadows support more decapods than continuous meadows or that decapod density is not related to patch area (Eggleston et al. 1998; Hovel and Lipcius 2002). The integration of the PC and CH meadows with other habitats (e.g. patches of Cymodocea nodosa, algae bottoms, rocky and sandy bottoms) generates a macrohabitat, which may reduce the habitat loss effect produced by the fragmentation. This configuration promotes the movement of species and individuals between habitats, which generates relatively high values of species richness, diversity and equitability that are comparable to those reported in Decapod assemblages associated with Posidonia oceanica 107 decapod assemblages inhabiting continuous P. oceanica meadows (17–50 spp.; H’ = c. 1.2–2.9 bits; J’ = c. 0.70, García Raso 1990a, 1988; Borg and Schembri 2000; Box Centeno 2008). Therefore, seagrass fragmentation is probably affecting some species, but it is not necessarily detrimental for the whole associated faunistic assemblage; also, fragmented seagrass patches may still provide important ecosystem functions for a high number of species and therefore should receive the same attention as nonfragmented ones with regard to habitat conservation and protection, as suggested by Borg et al. (2010), especially when considering that fragmentation has already been identified as an important external agent of seagrass decline (Gera et al. 2013). In any case, the co-existence of two inter-related assemblages (leaves and rhizomes – photophilous and sciaphilic) in P. oceanica meadows, and the further development of the rhizomes may minimize the effects of fragmentation when compared with other small seagrasses occurring on soft bottoms (e.g. Zostera, Cymodocea). Acknowledgements We would like to express our sincere gratitude to Carmen Salas Casanova and Serge Gofas from the University of Málaga (Spain) for their help at different stages of this research. We thank Maria Cristina Gambi, Loïc Michel and two anonymous reviewers for their interesting suggestions, and Terence W. Edwards for the English revision of this manuscript. This work was partly supported by the Junta de Andalucía ‘Consejería de Medio Ambiente’ (reference 807/46.2283) and the RNM-0141 research group of the University of Málaga. Decapod assemblages associated with Macroalgal beds 115 Sample collection and laboratory procedures Faunistic samples were collected in different times along one year, including July (summer), November 2007 (autumn), January (winter) and April 2008 (spring), with the use of SCUBA. The line intercept method (eight 50 m transects in each site) was used to estimate the algal coverage in the study area. The distinct bionomic strata were sampled separately, first the algal fronds (hereafter ‘algae stratum’) and then the underlying substratum (hereafter ‘sediment stratum’), in each site and season. Algal fronds within a quadrat (0.25 m2) were collected carefully and enclosed within <0.5 mm mesh bags. Once algae were removed, the sediment below the algae within the same quadrat was collected using an air-lift sampler with a <0.5 mm mesh bag. Suction time was similar (3 minutes) in every sediment sample collected. Five replicate samples were collected per site, stratum and season, resulting in a total of 80 samples. This methodology has previously been used successfully by other authors for collecting samples in similar photophilous algal meadows (Poulicek, 1984; Chemello and Russo, 1997; Bégin et al., 2004), and a similar number of species and individuals were found with this and other sampling techniques (e.g. scraping off individuals from the holdfast (Poulicek, 1984). The air-lift sampling was not used for the fronds in order to avoid collecting individuals from the underlying sediment. In the laboratory, every faunistic sample was sieved over mesh sizes down to 0.5 mm, storing each size fraction in 70% ethanol. Decapods were separated and each species were identified and their individuals counted under a binocular microscope. We distinguished two groups for the dominant species, juvenilessmall sizes and adults-large sizes, for which the minimum size of adults was determined from literature (Zariquiey, 1968) and collected data. Wave height (m) information obtained with the High Resolution Limited Area Model and Wave Model numeric models was taken from the Instituto Nacional de Meteorología (Spain) (http:// www.puertos.es). Water and sediment variables Water samples for estimating the concentration of chlorophyll a (Chl a) and seawater temperatures were taken several days before, during and after Chapter 5 116 sampling at each site, in order to study the relation of these variables with the decapods and the algal assemblages. In each sampling event and site, two replicates of 1 litre of seawater were collected at the surface and transported in darkness at low temperature to the laboratory for Chl a determination. Water samples were filtered through Whatman GF/C glass filters with a 1.2 μm of pore size. The pigments of the retained cells were then extracted using 100% acetone for 12h in cool and dark conditions. The solution was measured using a spectrophotometer at wavelengths of 630, 647, 664 and 750 nm. The Chl a concentrations were obtained using the equation proposed by Jeffrey and Humphrey (1975). The salinity remains almost constant throughout the year due to the low fresh water input in the area, ranging between 36.5 p.s.u. in autumn and 36.8 p.s.u. in spring (Salinity data given in Practical Salinity Scale; GCC, 2014). Samples of sediment were also taken within the sampled meadows (5 replicates per season and site) in order to estimate the percentage of organic matter (% OM). This percentage was calculated by the weight loss of dry sediment (3 subsamples of 20 gr. per replicate) after ignition at 500ºC for 1h. Macroalgal characteristics Macroalgae collected in the samples were also identified and quantified in order to study the structure and composition of the algal assemblage throughout the seasons, as well as their influence on the decapod assemblages. Measurements made on the main collected macroalgal species included: (1) Dry weight (DW), algal weight after drying for 48h at 84ºC and after removal of the epiphytes and animals (Edgar 1983b) (measured in g dw m−2); (2) Volume (V), calculated by the displacement of a known volume of water (Bussell et al., 2007) (measured in cm3 m−2); (3) Average height of the top-dominant species (H. scoparia; HS), length from the holdfast to the distal tip of the plant (measured in cm) of five randomly-selected fronds in each replicate. Finally the sum of the weights and volumes of the different species collected were considered to the total algal dry weight and volume. Decapod assemblages associated with Macroalgal beds 117 Data analysis Abundance (N) (individuals per sample), Frequency index (%Fi) (percentage of samples in which a particular species is present), Dominance index (%Di) (percentage of individuals of a particular species within the sample) (Glémarec,1964) and the product of Di x Fi/100 (Lopez de la Rosa et al. 2002; 2006) were calculated for the species characterization in each time of the year (July, November, January and April), site (P. Calaburras, Calahonda) and stratum (algae stratum and sediment). The characterization of decapods assemblages was done according to several ecological indices, such as the abundance of decapods (N), species richness (S), the Shannon-Wiener diversity index (H´) and the evenness index (J´). These ecological indices were calculated using the software PRIMER 6.0. The trophic groups dominating were analyzed by assigning each decapod species to a trophic category, including deposit feeders (Dep; species that feed on fragmented particulate organic matter from the substratum), filter feeders (Fil.; species that feed on particulate organic matter suspended in the water column), scavengers (Sca; species that feed on dead organic material), grazers (Grz; species that feed on periphyton or other epiphytes) and predators (Pred; species that capture and feed directly on all or part of a living animal species) (Borja et al. 2000; Zupo 1993; Grall et al. 2006; MarLIN 2006; Zubikarai et al. 2014). Spatial and temporal statistical differences of environmental variables and macroalgal features were tested with a two-factor ANOVA (Analysis of Variance) analysis considering the factors “site” (P. Calaburras vs. Calahonda) and “times” (July vs. November vs. January vs. April). Another two-factor ANOVA design with the factors size and times, was used for analysing differences between adults and juveniles abundances in each stratum along the sampling year. Moreover, threefactor ANOVA analyses were carried out for testing statistical differences in values of ecological indices according to site, strata and times. These three analyses were carried out after verifying normality (Kolmogorov-Smirnov) and homogeneity of variances (Levene). Abundance values were transformed in Ln (N). A post hoc Tukey test (P < 0.05) was used for posteriori multiple comparisons. KruskalWallis and Mann-Whitney analyses were carried out when the homogeneity of variances did not adjust to ANOVA conditions. These statistical procedures were performed using the software SPSS. Chapter 5 118 Differences in the decapod assemblage were analyzed though a three-way PERmutational MANOVA (PERMANOVA) design with the fixed factors sites, strata and times. Data were square-root transformed prior to analysis to downweight the relevance of the most abundant species and analyses were based on Bray–Curtis similarities (Clarke and Gorley, 2006). The same PERMANOVA design, but based on Euclidian distances, was use to test these differences between trophic groups. The SIMilarity PERcentage (SIMPER) procedure was used to identify those species that contributed to the similarity and dissimilarity between these same groups of samples. ANalysis Of SIMilarity (ANOSIM) was carried out for statistical comparisons of groups of samples (macroalgae species composition) according to factor site. These analyses were executed using the software PRIMER 6.0. Finally, the relationships between N and ecological indices (S, H’) of the whole decapod assemblage (algae + sediment stratum), as well as the abundance of single species, with environmental variables (temperature, Chl a, % OM) and algae characteristics (DW, V, HS) were determined by the Pearson correlation. In addition a Canonical Correspondence Analysis was performed to analyze the relationships between environmental variables and the decapods assemblages (taking into account only the top dominant species; Di>2.5%). The statistical significance of the effect of each variable was tested by a Monte Carlo permutation test. Prior to this, the environmental variables were screened and those which presented a correlation of more than 0.9 were not considered further. Environmental data expressed as % were transformed by log (x + 1). These correlations and multivariate analysis were executed using the software SPSS and CANOCO. Results Environmental variables Seawater temperatures were significantly higher in July (maxima, 24.5 °C) and November than in January (minima, 15 °C) and April in both P. Calaburras and Calahonda (two-way ANOVA; factor time: F = 44.4, P < 0.001; factor site: F Decapod assemblages associated with Macroalgal beds 119 = 1.2, P > 0.05). The Chl a concentrations displayed significant temporal changes (two-way ANOVA; factor time: F = 3.3, P < 0.05; factor site: F = 0.01, P > 0.05), with the highest value recorded in April (11.58 μg l−1) and the lowest in November (5.20 μg l−1), in both cases in Calahonda. Although %OM displayed significant differences between times and sites (two-way ANOVA; factor time: F = 4.7, P < 0.05; factor site: F = 36.1, P < 0.001), these values seemed very stable throughout time ranging between 1.11-1.79%, with the maxima in July (3%) at P. Calaburras. Wave height also presented significant differences between sites (Mann-Whitney; U=25, P < 0.001) and times (Kruskal-Wallis; X2 = 19, P< 0.001 in both sites), with maxima in November at Calahonda (~1 m). Macroalgal composition and seasonal variability Macroalgal cover ranged between 74% at P. Calaburras and 61% at Calahonda A total of 13 species were identified, being Halopteris scoparia (3868.53 g dw m-2), Jania rubens (633.36 g dw m-2) and Ellisolandia elongata (183.58 g dw m-2) the three most abundant. Other species were less abundant with values between less than 1 and 90 g dw m-2 (e.g. Cladostephus spongiosus, Asparagopsis armata or Padina pavonica). The structure of the macroalgal assemblages displayed slight, but significant differences between sites (RANOSIM = 0.159; P < 0.05). There were certain species that only appeared in P. Calaburras (Amphiroa rigida; P. pavonica) or in Calahonda (Dyctiota dycotoma; Acrosorium ciliolatum; Sphaerococcus coronopifolius) with values ranging between 1 and 8 g dw m-2. In addition, the dominant species presented some differences in their biomass, for example H. scoparia was more abundant in P. Calaburras (2171.95 g dw m-2) than in Calahonda (1696.58 g dw m-2) and J. rubens (373.68 g dw m-2; 259.68 g dw m-2) and E. elongata (146.89 g m-2; 36.69 g dw m-2 respectively) were in Calahonda. The structure of these assemblages showed significant temporal variations (Table 1). Total macroalgal mean DW and V displayed significant high values in July-April and low values in November-January, but these differences were not significant between sites (Table 1; Fig. 2). Both parameters presented Chapter 5 120 maximum values in April at Calahonda (754.75 ± 87.01 g dw m-2 and 2424.80 ± 181.76 cm3 m-2, respectively; mean ± SE). The dominant species (H. scoparia) also presented differences in DW, V and HS between times (Table 1; Fig. 2). Maximum values were observed in P. Calaburras for DW (557.57 ± 44.46 g dw m-2) and in Calahonda for V (1838.4 ± 78.76 cm3 m-2) and HS (11.42 ± 0.33 cm) during April. The DW and HS showed a significant and high positive correlation with the Chl a (RPearson = 0.50 and RPearson = 0.83, P < 0.001) while V of Table 1. Two-way ANOVA analyses for testing differences in the values of algal parameters in relation to sites (Punta Calaburras and Calahonda) and times (July, November, January, April). df, Degree of Free; SS, Squares Sum; MS, Mean Square. Source of variation n df SS MS F P Macroalgal characteristic Biomass 40 Site 1 1868.962 1868.962 0.057 0.813 Time 3 735,514.02 245,171.34 7.436 <0.001 Error 32 1,055,060.56 32,970.64 Volume 40 Site 1 34.339.6 37,339.60 0.146 0.705 Time 3 4,615,930.40 1,538,643.50 6.557 <0.05 Error 32 7,509,401.60 234,668.80 Halopterys scoparia Biomass 40 Site 1 14,028.02 14,028.02 1.015 0.321 Time 3 303,168.61 1,001,056.20 7.311 <0.001 Error 32 442,292.00 13,821.63 Volume 40 Site 1 1768.9 1768.9 0.013 0.909 Time 3 2,098,009.10 699,336.37 5.222 <0.05 Error 32 4,285,177.60 133,911.80 Height 40 Site 1 0.876 0.876 0.731 0.399 Time 3 15.586 5.195 4.322 <0.05 Error 32 38.381 1.199 Decapod assemblages associated with Macroalgal beds 121 H. scoparia displayed the same with %OM (RPearson = 0.62; P < 0.001). Some other dominant species such as J. rubens and E. elongata displayed maximum values in April for DW and V at both sites (Fig. 2). The maximum values were obtained in P. Calaburras with 62.34 ± 29.93 g dw m-2 and 104 cm3 m-2 respectively. However, due to large differences between replicates, the temporal variations of DW (Kruskal-Wallis; X2 = 10.25, P < 0.05) and V (Kruskal-Wallis; c2 = 10.86, P < 0.05) were only statistically significant for J. rubens at Calahonda. DW and V of J. rubens showed a significant positive correlation with Chl a (RPearson = 0.44; RPearson = 0.47, P < 0.05). Fig. 2. Seasonal trends of dry weight, (A, B) and volume, (C, D) of dominant species Halopteris scoparia (grey bars), Jania rubens (empty bars) and Ellisolandia elongata (black bars) and less abundant species (dark grey bars) in Punta Calaburras and Calahonda. Mean ± standard error. Letters above error bars display the results of post hoc Tukey, different letters indicate significant differences at P<0.05. Chapter 5 122 Composition, structure and seasonal variability of the decapods assemblages A total of 35 species were identified from 2654 collected specimens. Hippolitidae and Pilumnidae represented the most abundant families (529 and 479 ind. respectively), and Epialtidae represented the most diverse family (4 species). The topdominant species were Hippolyte leptocerus (481 ind.), Pilumnus hirtellus, Athanas nitescens, Sirpus zariquieyi, Achaeus gracilis and Acanthonyx lunulatus. The decapod assemblage showed significant different between sites, strata and times (Table 2, 3, 4). For example, species such as Pisa carinima, Calcinus tubularis, Cestopagurus timidus or Anapagurus hyndmanni were more abundant in P. Calaburras, whereas A. nitescens, Philocheras fasciatus or Processa robusta was in Calahonda (Table 2). Regarding to strata; there were species that only were found in one of the strata, e.g. Pisa nodipes (Di = 0.14%; Fi = 0.35%) in the algae stratum or Atelecyclus rotundatus (0.10%; 0.25%) and Sicyonia carinata (0.35%; 0.89%) in the sediment stratum. According to the Di x Fi values, decapods assemblages associated with the algae stratum were dominated by P. hirtellus in July, H. leptocerus from November to January and A. lunulatus in April (Table 4), whereas in the sediment stratum it was dominated by P. hirtellus in July and November, A. gracilis and H. leptocerus in January and H. leptocerus in April (Table 4). The species richness showed significant differences between sites, strata and times (Table 5). The temporal trend observed was similar in both strata, with significant maximum values in November and January, being more acute and with higher values in the sediment stratum (Fig. 3A, B). The highest values were observed in November in the algae stratum (10.20 ± 1.39 species) and in January in the sediment stratum (14.60 ± 1.03 species), in both cases in P Calaburras (Fig. 3A). The abundance showed significant differences between strata and times, but not between sites (Table 5). The temporal trend was similar to that observed for the species richness, with significant higher values in November in both strata (Fig. 3C, D). The highest abundance values were observed in November in both the algae (50 ± 14.00 ind.) and the sediment stratum (110.80 ± 37.85 ind.) in P. Calaburras. Overall, abundance of adults was higher than that of juveniles in the algae stratum (two-way ANOVA, factor size: F = 15.580, P< 0.001; factor season: Decapod assemblages associated with Macroalgal beds 123 Table 2. SIMiliraty PERcentage analyses. Species ranked according to their average within-group similarity at sites (Punta de Calaburras and Calahonda) and strata (algae and sediment). Av. Abund, Average abundance; Av. Sim, average similarity; Sim ⁄ SD, similarity/ standard deviation; Contrib. %, average contribution to similarity; Cum.%, cumulative percentage of similarity. Sites Punta Calaburras Av.Abund Av.Sim Sim/SD Contrib% Cum.% H. leptocerus 2.14 8.25 1.4 21.73 21.73 P. hirtellus 2.09 7.49 1.41 19.73 41.46 S. zariquieyi 1.57 5.64 1.15 14.86 56.32 A. lunulatus 1.07 4.07 0.79 10.72 67.04 P. carinimana 0.97 2.99 0.71 7.88 74.91 A. gracilis 0.9 1.48 0.5 3.89 78.8 C. tubularis 0.66 1.41 0.47 3.71 82.51 P. denticulata 0.62 0.94 0.41 2.46 84.97 C. timidus 0.67 0.88 0.36 2.31 87.28 A. hyndmanni 0.62 0.77 0.26 2.03 89.31 P. longimana 0.51 0.67 0.32 1.76 91.06 Calahonda Av.Abund Av.Sim Sim/SD Contrib% Cum.% H. leptocerus 1.92 9.48 1.17 26.77 26.77 P. hirtellus 1.6 5.23 0.8 14.76 41.52 A. lunulatus 1.04 4.7 0.63 13.26 54.78 A. gracilis 1.16 4.12 0.75 11.64 66.42 P. denticulata 0.83 3.25 0.53 9.17 75.59 S. zariquieyi 1.01 2.57 0.59 7.26 82.85 A. nistencens 1.02 1.52 0.45 4.3 87.14 P. longimana 0.62 1.04 0.36 2.93 90.07 Strata Alga Stratum Av.Abund Av.Sim Sim/SD Contrib% Cum.% A. lunulatus 1.18 8.03 0.88 23.86 23.86 H. leptocerus 1.43 6.83 0.88 20.29 44.15 P. hirtellus 1.35 6.44 0.87 19.12 63.27 S. zariquieyi 0.84 3.41 0.64 10.11 73.38 P. carinimana 0.56 2.05 0.44 6.08 79.46 A. gracilis 0.59 1.82 0.39 5.4 84.86 P. tetraodon 0.3 0.93 0.24 2.77 87.63 C. tubularis 0.34 0.89 0.29 2.65 90.28 Sediment Stratum Av.Abund Av.Sim Sim/SD Contrib% Cum.% H. leptocerus 2.62 11.45 2.38 26.18 26.18 P. hirtellus 2.34 6.66 1.31 15.23 41.41 S. zariquieyi 1.73 4.95 1.08 11.32 52.73 P. denticulata 1.19 4.53 0.97 10.36 63.1 A. gracilis 1.48 3.78 0.93 8.64 71.74 A. lunulatus 0.93 2.04 0.67 4.66 76.4 P. fasciatus 0.87 1.95 0.52 4.45 80.86 P. carinimana 0.8 1.46 0.56 3.33 84.18 A. nistencens 1.29 1.3 0.47 2.96 87.14 P. longimana 0.74 1.05 0.39 2.39 89.54 C. tubularis 0.57 0.79 0.39 1.82 91.35 Chapter 5 124 F = 10.646, P< 0.001) and the sediment stratum (two-way ANOVA, factor size: F =8.231, P< 0.01; factor season: F = 10.396, P< 0.001) along the year. However some species presented a higher number of juveniles than adults in all times and in both strata, as in the case of P. hirtellus with percentages ranging between 61% in April (sediment stratum) and 93% in November (algae stratum). Other species showed higher number of juveniles in certain times and especially in the sediment stratum such as in the case of A. gracilis (59% of juveniles in January), A. nitescens and P. longimana (94% and 79% in July respectively), P. carinimana (87%-62% in November-January) or A. hyndmanni (77% in April). Table 3. PERmutational MANOVA analyses based on Bray-Curtis index for testing differences in the decapod assemblages in relation to sites (Punta Calaburras and Calahonda), strata (algae and sediment) and times (July, November, January, April). df, Degree of Free; SS, Squares Sum; MS, Mean Square. Source of varation df SS MS Pseudo-F P Sites 1 5956 5956 3.701 <0.001 Stratum 116256 16256 10.103 <0.001 Time 324562 8187.3 5.088 <0.001 Site * Stratum 1917.47 917.47 0.570 0.826 Site * Time 3 7726.2 2575.4 1.600 <0.05 Stratum * Time 310160 3386.5 2.104 <0.01 Site * Stratum * Time 35377.3 1792.4 1.114 0.329 Residual 64 102980 1609 Total 79 173930 Diversity values of Shannon-Wiener index displayed significant differences between sites, strata and times (Table 4). Values showed a temporal trend with significant seasonal differences in the sediment stratum in P. Calaburras with the maxima in January (2.39 ± 0.18) (Fig. 3E), and in the algae stratum in Calahonda with the maxima in November (1.90 ± 0.08) (Fig. 3F). Finally, evenness only showed significant differences between times (Table 4), and these differences were observed in P. Calaburras in both strata, with the maxima in the algae stratum (0.98 ± 0.01 in July) (Fig. 3G). In Calahonda, evenness values were very similar throughout the year in both strata and ranged between 0.8 and 0.9 (Fig. 3H). Decapod assemblages associated with Macroalgal beds 131 variance in the species data, followed by DW of J. rubens (F = 3.44, P < 0.01), wave height (F = 2.06, P < 0.05), DW of H. scoparia (F = 2.14, P < 0.05) and DW of E. elongata (F = 2.06, P < 0.05). Samples aggregated in two different groupings: 1) samples from July and November, which were located mainly in the negative part of axis I in correlation with high values of temperature and wave height; and 2) samples from January and April, which were located in the positive part of axis I in correlation with high values of HS and Chl a and low temperatures (Fig. 5). The analyses suggests that the seasonality of decapod assemblages associated with the sediment stratum is mainly related to the temperature and algal characteristics Fig. 5. Canonical correspondence analysis of environmental variables, algae parameters, decapod assemblages and single species associated to sediment stratum in relation to axes I and II (eigenvalues: 0.102 and 0.055 respectively). T, seawater temperature; %OM, percentage of organic matter; Chl a, chlorophyll a concentration; Wave, wave height; DWH, dry weight of H. scoparia; HH, height of H. scoparia; DWJ, dry weight of J. rubens; DWE and dry weight of E. elongata; An, A. hyndmanni; Ph, P. hirtellus; Pl, P. longimana; Pf, Philocheras fasciatus; Hl, H. leptocerus; Pd, P. denticulata; Al, A. lunulatus; Ct, C. timidus; Pc, P. carinimana; Ag, A. gracilis; Si, S. zariquieyi. Chapter 5 132 (DW and HS). The majority of the top dominant species are near the origin of the axes, which shows a poorly differentiated profile distribution in relation with the analyses of parameters. Discussion Seasonality and composition of the algal community The macroalgal beds considered for this study seem to correspond to that of Mediterranean sublittoral communities dominated by leafy macroalgae species without the erect stratum of Fucales (Ros et al. 1985; Ballesteros 1993). The species composition of macroalgal beds of SCA “Calahonda” are similar to the community from Tossa de Mar (NW Mediterranean Sea), although with some differences such as absence of Cladophora prolifera or Lithophyllum incrustans and presence of A. armata (Ballesteros and Pinedo 2004). The H. scoparia beds studied showed a more definite seasonal trend than those reported by Ballesteros (1993) in a NW Mediterranean algal bed which did not present seasonal biomass changes. This seasonal trend was similar to that observed in other Mediterranean and Atlantic beds with maximum development mainly in April (Ballesteros 1984; Borja 1986b). The other two dominant species J. rubens and E. elongata also presented high development in July, in relation to the high temperature values (Ballesteros 1993) and in April, related to the high concentration of nutrients as observed in the positive correlation between Chl a concentration and several algae variables. Decapod assemblages, temporality and microhabitat use The decapod assemblages associated with photophilous macroalgal beds of SCA “Calahonda” (northwestern Alboran Sea) showed a high species richness (35 species), and although they were mainly composed by species common in the Mediterranean, some African (P. carinima) and Atlantic (A. hyndmanni) species were also abundant. The number of species observed here was higher than that found in similar habitats in other Mediterranean areas (25 species) or in the Caribbean Sea (32-27 spp.), and similar to that found in the more Decapod assemblages associated with Macroalgal beds 133 complex seascape composed by seagrasses and macroalgae in Florida (38 species) (Gore et al. 1981; Castelló et al. 1987; Quirós and Campos 2010; Quirós et al. 2012). The decapod assemblages in the SCA “Calahonda” were dominated by both species commonly found in macroalgal beds (P. hirtellus, S. zariquieyi, A. lunulatus or A. gracilis), and species frequently associated with seagrass meadows (H. leptocerus or A. nitescens) (Pérès and Picard 1964; Zariquiey 1968; Vadon 1981; Castelló et al. 1987; López de la Rosa and García Raso 1992; Ballesteros and Pinedo 2004; Mateo-Ramírez and García Raso 2012; Mateo-Ramírez et al. 2015). Furthermore, characteristic species of soft and hard bottoms were also collected in high numbers due to the proximity of these habitats and the movement of species among them (García-Muñoz et al. 2008; Mateo-Ramírez and García Raso 2012; Mateo-Ramíre et al. 2015). These species were presented in both sampling sites; however C. tubularis, C. timidus, P. carinimana and A. hyndmanni were more abundant in P. Calaburras, whereas P. fasciatus and P. robusta were in Calahonda. These differences were related with the higher quantity of rocks covered by algae and soft bottoms present in P. Calaburras and Calahonda respectively. Regarding the trophic composition, decapods assemblages associated to macroalgae beds of SCA ”Calahonda” were dominated by scavengers and predators. However, most of the trophic groups displayed a certain equitable distribution (in reference to abundance values) during most of the year (November-January) in both strata. An equitable distribution of trophic groups would indicate a healthy ecosystem functioning, which together with a high number of predator species could reflect a higher functional diversity of the entire assemblage (Ngai and Srivastaba 2006; Bremner et al. 2006; Zubikarai et al. 2014). These data could be used as a reference point for decapods assemblages associated with low impactated habitats such as these macroalgal beds of SCA “Calahonda”, when considering the “Good Enviromental Status” (Marine Strategy Framework Directive) of European marine habitats. Our results indicate a decoupling between the decapod assemblages and the algal characteristics (biomass; volume), with higher values of species richness, Chapter 5 134 abundance and Shannon-Winner diversity during November and January when macroalgal beds showed its lowest development. Macroalgal beds are subjected to seasonal and inter-annual changes in its species composition and coverage (Ballesteros 1993), which can influence the abundance, structure and diversity of invertebrate benthic assemblages (Bustamante et al. 2014). Several studies have showed a positive correlation between algal biomass and mollusc and/or crustacean abundance, in relation to the availability of substrate and food (Borja 1986a; López de la Rosa et al. 2006; Guerra-García et al., 2011b; Antit et al. 2013). Moreover, Urra et al. (2013b) observed that the vegetative cycle of algae in the northwestern Alboran Sea played an important role in the abundance of some dominant epifaunal grazers, with high abundance and species richness values coinciding with high biomass of algae. However, other studies have reported similar decoupling than ours (Edgar 1983b, Taylor 1998; Langtry and Jacoby, 2006). This decoupling could be linked with the feeding strategies of most decapod species inhabiting this habitat (scavengers, A. lunulatus, S. zariquieyi or A. gracilis; predators, P. hirtellus, P. fasciatus or P. carinimana; or deposit feeders, A. nitescens) which are not directly dependent to the algal dynamic, since their food resources (MO, detritus, prey) are mainly present in the sediment. Indeed, deposit feeders are the dominant trophic group on the sediment and the basal layer of vegetation in the macroalgae beds of Alboran Sea and Basque Coast (Urra et al.2013b; Bustamante et al. 2014). In addition, movements of species between surrounding habitats (P. oceanica, C. nodosa, soft and hard bottoms) as well as vertical movements of certain species between algal fronds and the underlying sediment looking for refuge or food, could be also related to this. This type of movements and use of different microhabitats/strata have been described in others habitats (C. nodosa or P. oceanica meadows) and species such as the gastropod Rissoa parva or the hermit crab C. timidus, which make vertical movements between strata looking for food (Fernández et al. 1998; Mateo-Ramírez and García Raso 2012; Urra et al. 2013b; Mateo-Ramírez et al. 2015). Decapod assemblages associated with Macroalgal beds 135 Life cycle of species is another factor that could be related to this decoupling observed between decapods assemblages and algal dynamics. In fact, A. gracilis, P. carinimana, P. denticulata or P. hirtellus showed higher abundances in the sediment with maximun values during November or January. These peaks were related to recruitment events, which coincided with the lowest values of algal biomass and temperature (Vadon 1981; López de la Rosa and García Raso 1992). In the case of other species, such as Hippolyte spp., abundances usually display a correlation with maximum algal, seagrass or epiphytic development (Mazzella et al., 1989; Lopez de la Rosa et al., 2006). In the SCA “Calahonda”, high abundances of Hippolyte spp. (mainly juveniles) were observed in July and May in C. nodosa meadows (Mateo-Ramírez and García Raso 2012), and in November and January (mainly adults and especially ovigerous females) in macroalgal beds. These data suggest that Hippolyte spp. (e.g. H. leptocerus) mainly recruit on seagrasses in spring-summer, and then individuals migrate to algal beds for spawning under the protection of algal fronds in autumn-winter. The low decapods abundance and richness values observed in our study during April and July could be also linked with a high predation pressure from fishes. In this line, several studies have showed that fish assemblage associated with macroalgal beds of European waters present spawning peaks between these months increasing their presence and pressure in these months (AlonsoFernandez et al., 2011; Raposeiro and Azevedo, 2009). Similar temporal patterns have been observed in seagrass meadows of eastern (Alicante) and southern (Almería) Spain, with a negative relationship between predator and prey abundances (Sánchez-Jerez et al., 1999; García Raso et al., 2006a). Moreover, the larger size of decapods (6-21 mm) in comparison with molluscs (mainly 2-3 mm), amphipods or tanaids (3-5 mm), could force them to look for shelter and protection under the algal cover, as many fishes inhabiting vegetated environments (e.g. Thalassoma pavo, Coris julis, Serranus scriba and Diplodus annularis) feed directly on phytal invertebrates and large individuals are more efficiently detected by visual predators (Tortonese, 1975; Mazzoldi and De Girolamo, 1997; Edgar, 1983b). Chapter 5 136 Finally, the structural complexity of photophilous algal beds is another factor that can influence on the structure of the associated decapods assemblages. We are found a negative correlation between decapods abundance and DW of H. scoparia. According to Heck and Orth (1980), an increment of the surface per unit of substrate is related to an increment in epifauna density, but this relationship has a limitation when the vegetative development could prevent the access of new individuals. Probably, the low development (e.g.biomass) of algae in November-January could facilitate the input of species and individuals, increasing the abundance and diversity values. This trend was also found in epifaunal communities associated with E. elongata or Caulerpa prolifera beds, where the densest faunal populations were found during low or intermediate development stages of the algae (Sánchez-Moyano et al. 2001; Kelaher 2003; Guerra-García et al. 2011b). Conclusions Decapods assemblages inhabiting macroalgal beds dominated by H. scoparia in the northwestern Alboran Sea presented a higher number of species than other decapod assemblages inhabiting similar habitats in Mediterranean and Atlantic coastal areas, being composed of species associated with algal beds and others commonly found in surrounding habitats. Furthermore, these assemblages showed a healthy ecosystem functioning with a dominance of scavenger and predator species, as well as a relative equitable distribution of the majority of the trophic groups during most of the year. They displayed higher individual abundances and species number when algal assemblages showed lower biomass, which indicate that decapods assemblages displayed a temporal pattern not necessarily linked with algae dynamics. Important aspects to be considered behind these observations include different feeding and reproductive strategies, movement of species between surrounding habitats and strata as well as punctual predation pressure. Our results could be indicated that the decapod assemblages need not be present a closed relationship with the algal dynamics. A conceptual model is proposed here to explain this (Fig. 6). Decapod assemblages associated with Macroalgal beds 137 Fig. 6. Conceptual model proposed for explain the decoupling between decapod assemblages and algal dynamic. N sediment, abundances in the sediment; N algae, abundance in the algae stratum; S sediment, species richness in the sediment, S algae, species richness in the algae stratum. Chapter 5 138 Acknowledgements We would like to express our sincere gratitude to Carmen Salas Casanova and Serge Gofas from the University of Malaga (Spain) for their help at different stages of this research. We thank Terence W. Edwards for the English revision of this manuscript. This work was partly supported by the Junta de Andalucía “Consejería de Medio Ambiente” (reference 807/46.2283) and RNM-0141 Research Group from the University of Malaga. CAPÍTULO 6 Estructura, composición y conectividad de las asociaciones de decápodos ligados a un fondo infralitoral fragmentado Structure, composition and connectivity of decapod assemblages associated with an infralittoral fragmented seascape