The influence of geomorphology on the composition of aquatic flora and fauna withim a temporary pond network

Limne ica, 29 (2): x-xx (2011) Limne ica, 33 (2): 327-340 (2014) c Asociación Ibé ica de Limnología, Mad id. Spain. ISSN: 0213-8409 The in luence o geomo phology on he composi ion o aqua ic lo a and auna wi hin a empo a y pond ne wo k Ma ga i a Flo encio1,4,∗, Lau a Se ano2, Pa icia Siljes om3, Rocio Fe nandez-Zamudio1,2, Pablo Ga cia-Mu illo2and Ca men Díaz-Paniagua1 1Es ación Biológica de Doñana (CSIC), Se illa, Spain. 2Dep . Plan Biology and Ecology, Uni e si y Se illa, Se illa, Spain. 3Ins i u o de Recu sos Na u ales y Ag obiología (CSIC), Se illa, Spain. 4P esen add ess: Azo ean Biodi e si y G oup (CITA-A) and Pla o m o Enhancing Ecological Resea ch and Sus ainabili y (PEERS), Depa amen o de Ciências Ag á ias, Uni e sidade dos Aço es, Azo es, Po ugal. ∗Co esponding au ho : [email p o ec ed] 2 Recei ed: 03/03/2014 Accep ed: 08/08/2014 ABSTRACT The in luence o geomo phology on he composi ion o aqua ic lo a and auna wi hin a empo a y pond ne wo k Geomo phological and hyd ological ea u es can p o ide a sound basis o global we land classi ica ion. Tempo a y ponds loca ed on he sandy a ea o Doñana can be classi ied in o i e di e en geomo phological a eas. We hypo hesised ha hese ponds would di e in soil cha ac e is ics, which may also explain di e ences in he composi ion o hei mac oin e eb a e, amphibian, and aqua ic plan species assemblages. The s udy ponds we e signi ican ly seg ega ed in o sou he n and no he n geomo phological a eas based on he dep h o he wa e able. The esul s o he mul i a ia e o dina ion o soil a iables we e consis en wi h a no h-sou h seg ega ion, which was explained by he ca bona e and o he main ion concen a ions in he pond basins. The ponds loca ed in eco ones (ma sh-sand and s able-mobile dunes) we e pa icula ly ich in amphibians and mac oin e eb a es. This esul may be explained because he ponds wi h he longes pe manence o wa e we e loca ed in hese a eas, which ep esen ed he only looded ponds du ing d ough s. In addi ion, hey also con ained a high numbe o empo a y ponds, hus a ou ing connec i i y and en i onmen al he e ogenei y in hese a eas. Key wo ds: Tempo a y ponds, geomo phology, Doñana, aqua ic assemblages, mac oin e eb a es, amphibians, aqua ic plan s. RESUMEN In luencia de la geomo ología sob e la composición de la auna y lo a acuá icas de un sis ema de lagunas empo ales Las ca ac e ís icas hid ológicas y geomo ológicas pueden se de g an u ilidad pa a la clasi icación de humedales. Las lagunas empo ales del man o eólico de Doñana se pueden clasi ica en unción de las cinco á eas geomo ológicas en que se encuen an. Nues a hipó esis se basa en que las lagunas de es as á eas di ie en en las ca ac e ís icas de sus suelos, lo cual puede explica ambién las di e en e composición de las comunidades de mac oin e eb ados, an ibios y plan as que las habi an. Las lagunas del su de Doñana se di e encia on de las del no e p incipalmen e en la dis ancia a la capa eá ica. Se ealizó una o denación de las lagunas a pa i de un análisis mul i a ian e de las ca ac e ís icas del suelo que esul ó consis en e con la seg egación no e-su . Se obse ó una mayo seg egación en las comunidades de plan as acuá icas en es e g adien e no e-su , que se explica en unción de las concen aciones de ca bona os y de o os iones p incipales en el suelo de las lagunas. Las lagunas localizadas en á eas de eco ono (ma isma-a enas y a enas es abilizadas-a enas mó iles) esul a on se las de mayo iqueza de an ibios y mac oin e eb ados. Es o pod ía debe se a que es as á eas con ienen las lagunas de mayo du ación, siendo los únicos medios inundados en épocas de sequía, y al al o núme o de lagunas empo ales que p esen an, a o eciendo la conec i idad y he e ogeneidad de ambien es. Palab as cla e: Lagunas empo ales, comunidades acuá icas, geomo ología, Doñana, mac oin e eb ados, an ibios, plan as acuá icas. 15624_Limne ica 33(2), pàgina 327, 09/11/2014 328 Flo encio e al. INTRODUCTION The leng h o he we phase has long been ega ded as a undamen al ac o ha in lu- ences all o he ea u es in empo a y aqua ic sys ems (Rzóska, 1961). The hyd ological egi- mes o empo a y ponds may di e due o many ac o s such as he pond basin soil, which plays an impo an ole in empo a y pond ecosys ems. O e long- e m pe iods, an a ea’s inunda ion equency and du a ion a ec he physical and chemical p ocesses in ol ed in he de elopmen o hyd ic soils (Rabenho s , 2004). Hyd ogeomo phic-based app oaches a e p e e en ially applied in we land unc ional as- sessmen s (Cole e al., 1997). These app oaches a e pa icula ly use ul in he classi ica ion o empo a y we lands in semi-a id egions, whe e he unp edic abili y o he clima e esul s in a wide ange o hyd ological condi ions in a single wa e body be ween yea s. Medi e anean empo a y ponds a e ound in many pa s o he wo ld ha a e cha ac e ized by Medi e anean clima es. Mos o hese ponds o m in dep essions in soils o e lying an impe - meable subs a e and esul in a pe ched wa e able (Keeley & Zedle , 1998), as desc ibed in Cali o nia, Chile, Sou h A ica, and wes e n Aus- alia. By con as , in he Medi e anean basin empo a y ponds also occu in loodplains ha d ain ela i ely la ge wa e shed sys ems (Keeley & Zedle , 1998) and a e no necessa ily associ- a ed wi h impe meable ine sedimen s. They a e common in a eas wi h a Medi e anean clima e, in which he desicca ion pe iod gene ally occu s in he summe , and he looding pe iod ollows hea y win e o au umn ain all (Zedle , 1987; G illas e al., 2004; Gómez-Rod iguez e al., 2009; Díaz-Paniagua e al., 2010). Despi e he wo ldwide dis ibu ion o empo a y we lands, he e a e ew s udies add essing he classi ica ion o bo h lo a and auna based on he geomo pho- logical cha ac e is ics o hei habi a s. Howe e , such an app oach has been common in we land plan classi ica ions (Haslam, 2003). The abili y o hyd ogeomo phic classes o disc imina e among o he biological communi ies has been occasionally es ed in se e al we land a eas, wi h soil o ganic ma e and mois u e as he s onges d i e s o he ypi ica ion o mac oin e eb a e assemblages (USEPA, 2002). In he Ibe ian Pe- ninsula, Alonso (1998) de eloped a ypology o c us acean assemblages closely ela ed o subs a e, clima e and geog aphic a ibu es. Recen ly, Sahuquillo & Mi acle (2013) p o ided a ypology o Medi e anean ponds using c us- acean assemblages based on wa e pe manence, clima ic and o og aphic ea u es. Howe e , pond soil cha ac e isa ion appea s o p o ide aluable in o ma ion abou he main cha ac e is ics o an a ea, which is ela ed o he composi ion o aqua ic species, and may p o- ide a no el pe spec i e o he classi ica ion o aqua ic communi ies. A la ge ne wo k o empo a y ponds is loca ed in Doñana Na ional Pa k (SW Spain), whe e mo e han 3 000 ponds may be inunda ed du ing yea s o hea y ain all (Díaz-Paniagua e al., 2010). These ponds cons i u e impo an habi a s o amphibians, mac oin e eb a es, mic oc us- aceans, and aqua ic ege a ion (Díaz-Paniagua e al., 2010). To manage and conse e hese habi a s, i is o c ucial impo ance o assess hei wa e egime and e alua e he main ac o s sus- aining hei we phase. The se e al a emp s by sho -scale s udies o classi y Doñana ponds in o di e en hyd ological ca ego ies ha e p oduced con adic o y esul s (Ga cía-No o e al., 1991; B a o & Mon es, 1993; Manzano, 2001). How- e e , he cha ac e is ics o he soil o hei pond basins could p o ide a geomo phological ame- wo k upon which o base a classi ica ion sys em. In his s udy, we analysed he cha ac e is ics o he soils o empo a y ponds loca ed in he di e en geomo phological a eas o Doñana Na- ional Pa k. We hypo hesised ha he geomo - phological classi ica ion o ponds was obus and ha ponds in di e en geomo phological a eas would ha e di e en composi ions o species as- semblages (mac oin e eb a es, amphibians, and aqua ic plan s). We we e pa icula ly in e es ed in de e mining i idiosync a ic species we e ex- clusi ely associa ed wi h pa icula geomo pho- logical a eas. 15624_Limne ica 33(2), pàgina 328, 09/11/2014 Pond geomo phology in luences lo a and auna 329 METHODS Geomo phology o he Doñana pond ne wo k Doñana Na ional Pa k is a coas al dune a ea ha is loca ed o e a la ge sedimen a y aqui e sys- em o app oxima ely 3000 km2. The Doñana e- gion was o med in he Qua e na y age when he es ua y o he Guadalqui i Ri e was enla ged and eshaped by he o ma ion o sandy spi s a e he las pos glacial ansg ession (Siljes om e al., 1994). One hal o he pa k includes an ex en- si e ma sh, and he o he hal is composed o ae- olian sands ha o m a pe meable laye h ough which wa e s d ain o an uncon ined aqui e . The wa e able is gene ally shallow in his a ea and in e sec s wi h smoo h dep essions in he soil su ace, o ming bo h pe manen and empo a y ponds (Manzano & Cus odio, 2006). This aeo- lian uni has been sepa a ed in o he ollowing i e geomo phological a eas ha di e in hei dynamics, pa en ma e ial, and age: no he n s a- bilised dunes, sou he n s abilised dunes, mobile dunes, and wo con ac a eas, o eco ones, which include he pe idune ponds and he ma sh-dune bo de (Siljes om e al, 1994) (Fig. 1). Figu e 1. Loca ion o he pond sample soils () and he lo a and auna pond samples (•) in he i e geomo phological a eas desc ibed o Doñana Na ional Pa k. Localización de las lagunas mues eadas pa a el análisis de los suelos (), y pa a la de e minación de su lo a y auna acuá icas (•) en las cinco á eas geomo ológicas desc i as en el Pa que Nacional de Doñana. 15624_Limne ica 33(2), pàgina 329, 09/11/2014 330 Flo encio e al. The no he n s able dune a ea is an ancien dune sys em (6000-12000 BP) whe e empo a y ponds o m in he lowe a eas a e he ise o he wa e able by ain all in il a ion. The sou he n s able dune a ea, which is o mo e ecen o igin (2000-4000 BP), has a smalle ex ension, and he soils di e in he physico-chemical and mine - alogical cha ac e is ics ( ich in CaCO3and o he sal s) om hose ound in he no he n s able dunes, which a e mos ly siliceous (Siljes om e al., 1994). In he mobile dunes, discha ges om a local dune aqui e o m empo a y ponds in de- p essions in in e dune alleys (o slacks). The wo in e aces be ween hese geomo phological a eas play a special ole in he o ma ion o ponds because hey a e cha ac e ised by ph ea ic dis- cha ges o g oundwa e . The pe idune eco one is he a ea in which mobile dunes and no he n s a- ble dunes come in con ac . I includes he la ges and mos pe manen ponds o he pa k ha a e mainly ed by he dune aqui e . The ma sh-s able dune in e ace is locally known as “La Ve a”. In his eco one, he discha ges o deep wa e le els o igina ing om he ma sh and s able dunes coin- cide, o en esul ing in wa e unning o in shal- low and empo a y s eams a e hea y ain all (Clemen e e al., 1998). Doñana ponds a e ed by eshwa e ( ain all, uno , and g oundwa e discha ge) and ha e no su ace o g oundwa e connec ion o he sea, al hough hey do ecei e sal s o ma ine o i- gin h ough ai bo ne deposi ion (Se ano e al., 2006). Tempo a y ponds a e looded a e hea y ains, which in his a ea may occu in au umn, win e , o less equen ly, in he sp ing. The d y phase coincides wi h summe ; i s a s in he mos epheme al ponds in he sp ing, whe eas hose wi h he longes hyd ope iods (i.e., du a- ion o he looded pe iod) may e ain wa e un il he end o summe (Díaz-Paniagua e al., 2010). Soil sampling To ela e he soil geomo phological cha ac e is- ics o ponds wi h hei aqua ic assemblages, soil samples o 43 pond basins loca ed ac oss he i e di e en geomo phological a eas we e ob ained by pe o ming bo ehole sampling du ing pond d y pe iods in 2007 (n=34) and 2012 (n=9). We collec ed samples om each ho izon in all p o iles o labo a o y analyses. Chemical analy- ses (pH, elec ical conduc i i y, ca ion exchange capaci y (CEC), phospho us ex ac ion, o ganic ma e cha ac e isa ion, assimilable ca ions, and ca bona es) we e conduc ed ollowing Spa ks & Ba els (1996). Soil ex u e was de e mined using a Boyoucos densime e (Day, 1986); he pe cen - ages o sil and clay we e summed o p o ide he o e all pe cen age o ine sedimen s. Hea y me als we e ob ained using humid diges ion and Table 1. Pe cen ages o ponds wi h di e en wa e egimes o which mac oin e eb a e, amphibian, and aqua ic plan species composi ion was cha ac e ised in 2007. Ponds a e ca ego ised in o one o he i e geomo phological a eas desc ibed o Doñana Na ional Pa k. Wa e egime was classi ied as ollows: empo a y ponds (na u al empo a y ponds wi h an annual d y phase), exca a ed ponds (isola ed ponds ha ha e been a i icially exca a ed o main ain wa e in summe ), empo a y+exca a ed ( empo a y ponds ha include an exca a ed pond in hei basin o e ain wa e in summe ), pe manen ponds (la ge na u al ponds ha do no d y up in summe ) and s eams ( unning wa e s a e hea y ains han hen pe sis as empo a y ponds). Fo amphibians, we sampled an addi ional pond; he modi ied pe cen age is included in b acke s. Po cen ajes de lagunas con di e en e égimen híd ico sob e el o al de lagunas en las que se mues a on mac oin e eb ados, an ibios y plan as acuá icas en 2007. Las lagunas se han clasi icado den o de cada una de las cinco á eas geomo ológicas desc i as en el Pa que Nacional de Doñana. En los mues eos se an ibios se mues eó una laguna más que en los o os g upos, po lo que el po cen aje modi icado pa a es e g upo se mues a en e pa én esis. No he n s able dunes Sou he n s able dunes Mobile dunes Pe idune eco one La Ve a eco one Tempo a y ponds 61.3 (62.5) 11.8 57.1 87.5 65.4 Exca a ed ponds 19.4 88.2 42.9 0 11.5 Tempo a y+exca a ed ponds 19.4 0 0 0 0 Pe manen ponds 0 0 0 12.5 0 S eams 0 0 0 0 23.1 N ponds sampled 31(32) 17 7 8 26 15624_Limne ica 33(2), pàgina 330, 09/11/2014 Pond geomo phology in luences lo a and auna 331 ICP/OES lec u es (Williamson, 1994). To assess he di e ences in he g oundwa e le els o he geomo phological a eas, we also measu ed he dep h o he g oundwa e able in he sampled ponds and 36 addi ional ponds (n=79 o al si es) h oughou he di e en a eas o he pa k. Ponds, auna, and lo a To ela e ou da a on pond soils o he p esence o pa icula species o aqua ic auna and lo a, we used p esence/absence da a on he ege a ion, mac oin e eb a es, and amphibians o empo- a y ponds ac oss he en i e pa k. Da a om 89 ponds we e collec ed om Ma ch-June 2007 (an addi ional pond was sampled o gain mo e in o ma ion on amphibians: n=90 ponds o amphibian da a). The ponds’ geomo phological loca ions and wa e egimes a e indica ed in Table 1. The wa e egimes we e i s classi ied as ollows: empo a y, pe manen , s eam, o a i icially exca a ed. S eams e e o in e mi - en unning wa e s ha only low owa ds he ma sh a e hea y ains and hen o m len ic ponds. A i icially exca a ed ponds e e o holes exca a ed in he deepe pa s o na u al ponds o each he ph ea ic le el du ing d y pe iods and may be conside ed as pe manen wa e bodies; hey we e mainly buil o p o ide wa e o ca le o wild auna du ing he summe ( empo a y+exca a ed ponds). These exca a ed ponds some imes occupy he en i e pond basin, a e isola ed om na u al empo a y ponds, and occu mainly in he sou he n s able dunes. Amphibians and mac oin e eb a es we e sampled using a dip-ne wi h a 1-mm mesh size. We pe o med h ee successi e sweep-ne s ac oss a s e ch o wa e o c.1.5 m in each sampling uni o sample he he e ogeneous en i onmen s mainly based on dep h and aqua ic plan co e wi hin each pond; we ollowed he same p ocedu e o bo h g oups o o ganisms Figu e 2. Non-me ic mul i-dimensional scaling o dina ion pe o med on he simila i y ma ix o pond soil samples acco ding o pond loca ion in he di e en geomo phological a eas o he Doñana aeolian sands: no he n s able dunes (2), sou he n s able dunes (N), pe idune (s able-mobile dunes) eco one (◦), mobile dunes (▽), and La Ve a (ma sh-sand) eco one (•). The ela i e magni ude o Spea man’s co ela ion alues be ween each soil a iable and he pond o dina ion is indica ed by he line size (Dep h: ph ea ic dep h; Fine: ine ac ion; O.M.: o ganic ma e ; EC: elec ical conduc i i y; ClayM: % o clay in he clay- ich ho izon; ClayS: % o clay in he su ace ho izon). O denación mul idimensional no mé ica ealizada sob e la ma iz de simila idad de las mues as de suelo de las lagunas localizadas en las di e en es á eas geomo ológicas desc i as en el man o eólico de Doñana: A enas es abilizadas del (2), a enas es abilizadas del su (N), eco ono pe iduna (a enas es abilizadas-dunas mó iles (◦), dunas mó iles (▽), y eco ono de La Ve a (ma isma-a enas) (•). Se indica la magni ud de los alo es de la co elación de Spea man en e los ipos de suelos y la o denación de las lagunas. (Dep h: ni el eá ico, Fine: acción ina, O.M.: ma e ia o gánica; EC: conduc i idad eléc ica, ClayM: % de a cilla en el ho izon e con mayo p opo ción de a cillas, ClayS: % de a cillas en el ho izon e supe icial). 15624_Limne ica 33(2), pàgina 331, 09/11/2014 332 Flo encio e al. (see, e.g., Flo encio e al. 2009a o a de ailed desc ip ion o he p ocess). Aqua ic plan s we e iden i ied a he same loca ions. In ou analyses, we examined adul and la al mac oin e eb a es sepa a ely (he ea e e e ed o as “ axa” o simplici y). Mac oin e eb a es we e iden i ied o species in he case o mos adul s o genus in he case o mos la ae, al hough mos membe s o he o de s Bassoma opho a, Dip e a, Haplo- axida, and Lumb icula, as well as o he amily Saldidae (He e op e a), we e only iden i ied o he amily le el. Amphibians and aqua ic plan s we e iden i ied o he species le el. Da a analyses A mul i a ia e o dina ion analysis (non-me ic MDS) was pe o med on he simila i y ma ix o he 14 physico-chemical soil a iables using P ime 6.0. These a iables we e measu ed in he 43 pond basins ha included he deep clayey laye o he opsoil laye when clay was no p esen . The a iables included in his analysis we e all log ans o med (X+1), wi h he excep ion o he non- ans o med pH; elec ical conduc i i y, o ganic ma e con en , ine sedi- men s (clay and sil ), ca bona es, To -P, To Ca, Mg, S, Mg, K, and Fe in he su ace ho izon; clay and o al ine sedimen con en s in he clay- ich ho izon; and dis ance o he g oundwa e able. We hen ob ained a iangula simila i y ma ix using Euclidean dis ances. A mul i a ia e ANOVA analogue (one-way ANOSIM) was used o es o mul i a ia e di e ences in he soil cha ac e is ics o samples om ponds loca ed in di e en a eas, i.e., using he geo- mo phological a eas as g ouping ac o s in he one-way ANOSIM analyses (P ime 6). The ANOSIM s a is ic R is close o one when g oups a e comple ely di e en (Cla ke & Wa wick, 2001). The ANOSIM analyses also allowed o pai -wise compa isons be ween he di e en geomo phological a eas. Spea man’s ank o de co ela ions we e also pe o med on he un ans- o med a iables. In addi ion, di e ences in he dep h o he g oundwa e able among he geo- mo phological a eas we e es ed using ANOVA and Tukey’s pos hoc es (STATISTICA 6.0). Table 2. Pai -wise compa isons ob ained a e a one-way ANOSIM analysis pe o med on pond soil a iables be ween he i e geomo phological a eas (no he n: no he n s able dunes; sou he n: sou he n s able dunes; mobile dunes; pe idune: pe idune eco one; and La Ve a: La Ve a eco one) indica e sig- ni ican di e ences ( p<0.05) be ween he pai -wise compa - isons excep o wo cases ( p>0.1); ANOSIM s a is ic R is close o one when g oups a e comple ely di e en . Compa a- ciones pa eadas ob enidas en el análisis ANOSIM de una ía sob e las a iables del suelo de las lagunas en e las á eas geo- mo ológicas (No he n: a enas es abilizadas del no e, Sou h- e n: a enas es abilizadas del su , Mobile Dunes: dunas mó iles, Pe idune: eco ono pe iduna y Ve a: eco ono de La Ve a) in- dicando di e encias signi ica i as en odas las compa aciones pa eadas ( p<0.05) excep o en dos ocasiones ( p>0.1); el es adís ico R de ANOSIM se ap oxima a 1 cuando los g upos son comple amen e di e en es. Rp No he n ×Sou he n 0.99 0.0001 No he n x Mobile Dunes 0.86 0.0006 No he n ×Pe idune 0.12 0.11 No he n ×Ve a 0.26 0.012 Sou he n ×Mobile Dunes 0.85 0.029 Sou he n ×Pe idune 0.99 0.002 Sou he n ×Ve a 0.70 0.001 Pe idune ×Mobile Dunes 0.87 0.006 Pe idune ×Ve a 0.16 0.04 Mobile Dunes ×Ve a 0.16 0.177 We analysed he simila i y in assemblage composi ion among he sampled ponds using species p esence/absence da a; sepa a e analyses we e pe o med o amphibians, aqua ic plan s, and mac oin e eb a es. We ob ained iangula simila i y ma ices o each g oup using he Sø ensen simila i y coe icien (i.e., he in e se o he Sø ensen dissimila i y index). To examine di e ences in he species composi ions o he ponds, we pe o med h ee addi ional one-way ANOSIM analyses using he co esponding ian- gula simila i y ma ices o amphibians, aqua ic plan s and mac oin e eb a es; geomo phological a ea was he g ouping ac o in all h ee analyses. RESULTS Pond soil classi ica ion A non-me ic MDS o dina ion o he 14 pond soil a iables showed ha di e ences among he geomo phological a eas we e mainly a - ibu able o a s ong no h-sou h spa ial g adien 15624_Limne ica 33(2), pàgina 332, 09/11/2014 Pond geomo phology in luences lo a and auna 333 Table 3. Exclusi e axa ound in pa icula geomo phological a eas in Doñana Na ional Pa k (p esence/absence is indica ed as 1/0). The o al numbe o axa eco ded in each a ea is indica ed o each gene al g oup (a e Flo encio e al. 2009a; Díaz-Paniagua e al., 2006; Fe nández-Zamudio, 2011). Taxa in bold we e ound exclusi ely in a pa icula a ea (Adul - A and La ae- L we e di e en o Coleop e a and He e op e a; Odona a e e s only o aqua ic la ae). Taxones que esul a on exclusi os o ca ac e ís icos de lagunas de de e minadas á eas geomo ológicas del Pa que Nacional de Doñana (la p esencia/no de ección se indica como 1/0). Se indica el núme o o al de axa egis ados en cada á ea (según da os de Flo encio e al. 2009; Díaz-Paniagua e al. 2006). Se mues an en neg i as los axa exclusi os de un á ea de e minada (se di e encia en e Adul os -A y La as -L pa a Coleop e a y He e op e a, Odona a se e ie e sólo a la as acuá icas). No he n s able dunes Sou he n s able dunes Mobile dunes Pe idune eco one La Ve a eco one Amphibians 8 6 6 10 10 Bu o bu o (Linnaeus, 1758) 0 0 0 1 1 Pelody es ibe icus Sánchez-He áiz, Ba badillo, Macho dom & Sanchiz, 2000 0 0 0 1 1 Coleop e a 42 39 28 32 40 Yola bica ina a (La eille, 1804)-A 0 1 1 0 0 Hyda icus leande (Rossi, 1790) -L 0 0 0 0 1 Hyg o us inaequalis (Fab icius, 1777) -A 0 1 0 0 0 Laccobius e elie ei Pe is, 1864 -A 1 0 0 0 0 Hyd ophilus pis aceus (Lapo e, 1840) -L 1 0 0 0 1 Hyd ocha a la ipes (S e en, 1808) -L 0 0 0 0 1 Limnebius u ca us Baudi, 1872 -A 0 1 0 0 0 Gy inus –L 1 0 0 1 0 He e op e a 18 16 14 14 17 Mic o elia pygmaea (Du ou , 1833) -A 1 0 0 1 0 Odona a 9 7 5 10 8 Les es mac os igma (E e sm., 1836) 0 0 0 1 0 La ge C us aceans 4 0 0 0 3 Magh ebes he ia ma occana Thié y, 1988 1 0 0 0 0 Cyzicus g ubei Simon, 1886 0 0 0 0 1 B anchipus co esi Alonso y Jaume, 1991 1 0 0 0 0 Tanymas ix s agnalis (Linnaeus, 1758) 1 0 0 0 0 B anchipus scha e i Fische de Waldheim, 1834 1 0 0 0 0 S ep ocephalus o ico nis (Waga, 1842) 0 0 0 0 1 T iops mau i anicus (Ghigi, 1921) 0 0 0 0 1 Aqua ic Plan s 65 15 28 50 60 Cha a ulga is L. 0 1 1 0 0 Cha a agilis Des . 0 1 1 0 0 Isolepis lui ans (L.) R. B . 1 0 0 1 0 Juncus he e ophyllus Du ou 1 0 0 1 0 Ni ella anslucens (Pe s.) C. Aga dh 1 0 0 0 1 Po amoge on na ans L. 1 0 0 0 0 Lemna isulca L. 0 1 1 0 0 Wol ia a hiza (L.) Ho kel ex Wimme 0 1 1 0 0 To al numbe o species 146 83 81 116 138 15624_Limne ica 33(2), pàgina 333, 09/11/2014 334 Flo encio e al. in ca bona es and an eas -wes g adien in clay (Fig. 2). An ANOSIM analysis showed signi i- can di e ences be ween he soil cha ac e is ics o he i e geomo phological a eas (Global R = 0.46, p=0.0001); all he pai -wise com- pa isons be ween hese a eas we e signi ican (p<0.05), excep he compa isons be ween he pe idune eco one and he no he n s able dunes and be ween La Ve a eco one and he mobile dunes (p>0.1, Table 2). The pond soils loca ed in he sou he n s a- ble dunes and he mobile dunes had signi ican ly shallowe ph ea ic dep hs han he es o he ponds (ANOVA, F4,73 =8.7128; p=0.0001, and Tukey’s es ) (Fig. 3). We ound la ge pe cen - ages o exca a ed ponds in hese a eas (Table 1), which accoun ed o he shallowe dep h o he wa e able a hese si es. Pe manen ponds only occu ed a si es loca ed in he pe idune a ea (Ta- ble 1), oge he wi h o he ex ensi e ponds wi h long hyd ope iods. Fauna and lo a associa ed wi h pa icula geomo phological a eas The Doñana pond ne wo k p o ides aqua ic habi a s o a la ge numbe o species o lo a and auna (Table 3). Fo mac oin e eb a es, he highes axon ichness was ound in ponds loca ed in he no he n s able dunes and La Ve a eco one, wi h he excep ion o Odona a, which we e iches in he pe idune eco one. We de ec ed idiosync a ic axa ha we e exclusi ely ound in pa icula a eas, such as Coleop e a, Odona a and pa icula ly he g oup o la ge c us aceans. The highes amphibian species ichness was ound in he pe idune and La Ve a eco ones; wo species we e exclusi e o hese habi a s. The g adien o conduc i i y la gely con ibu ed o a clea di e en ia ion be ween aqua ic plan composi ions om he no he n (no he n s able dunes, Ve a and pe idune a ea) and sou he n a eas (mobile dunes and sou he n s able dunes). The lowes ichness o aqua ic plan s was ound in he sou he n s able dunes and mobile dunes. We did no obse e any plan species ha exclusi ely occu ed in one pa icula a ea, bu se e al species we e p ima ily associa ed Figu e 3. Mean ±s anda d de ia ion o he dep h o he g oundwa e able (ph ea ic le el) o he ponds loca ed in he i e geomo phological a eas desc ibed o Doñana Na- ional Pa k (no he n: no he n s able dunes; La Ve a: La Ve a eco one-con ac a ea be ween he no he n s able dunes and ma shes; pe idune: pe idune eco one-con ac a ea be ween he no he n s able dunes and mobile dunes; mobile dunes; sou h- e n: sou he n s able dunes). Di e en le e s indica e signi ican di e ences a e Tukey’s pos hoc es s o he i e geomo - phological a eas ( p<0.05). Media y des iación ípica de la dis ancia a la capa eá ica de las lagunas localizadas en las cinco á eas geomo ológicas desc i as en el Pa que Nacional de Doñana (No he n: a enas es abilizadas del no e; La Ve a: Eco ono de la Ve a-á ea de con ac o en e las a enas es abi- lizadas del no e y las ma ismas; Pe idune: eco ono pe iduna - á ea de con ac o en e las a enas es abilizadas del no e y las dunas mó iles; Mobile dunes: Dunas mó iles; Sou he n: A e- nas es abilizadas del su ). Las di e en es le as en la igu a indican di e encias signi ica i as después de ealiza el análi- sis pos hoc de Tukey en e las cinco á eas geomo ológicas (p<0.05). wi h wo con iguous a eas. Thus, pa icula plan species only appea ed in a eas wi h pe manen wa e s, i.e., in he pe idune eco one ha con- ained pe manen ponds and in he sou he n s a- ble dunes ha mainly con ained exca a ed ponds (Table 1 & Table 3). Consequen ly, ANOSIM analyses e ealed ha geomo phological a eas di e ed signi i- can ly in he composi ion o hei mac oin e e- b a e (Global R = 0.12, p=0.0002), amphibian (Global R = 0.14, p=0.0001), and aqua ic plan assemblages (Global R = 0.33, p=0.0001). Pai -wise compa isons e ealed ha he assem- blage composi ion o he h ee o ganism g oups signi ican ly di e ed ( p<0.05) be ween he mobile dunes and no he n s able dunes, no he n and sou he n s able dunes, and sou he n s able dunes and La Ve a eco one (Table 4). Di e ences 15624_Limne ica 33(2), pàgina 334, 09/11/2014 Pond geomo phology in luences lo a and auna 335 Table 4. Pai -wise compa isons ob ained a e one-way ANOSIM analyses pe o med o assemblage composi ion among he i e geomo phological a eas (no he n: no he n s able dunes; sou he n: sou he n s able dunes; mobile dunes; pe idune: pe idune eco one and La Ve a: La Ve a eco one). A o al o h ee one-way ANOSIM analyses we e pe o med, one o each bio ic assemblage: mac oin e eb a es, amphib- ians, and aqua ic plan s. Only p- alues o signi ican pai -wise compa isons a p<0.05 be ween he biological assemblages o he geomo phological a eas a e shown o each g oup o o - ganisms (ANOSIM s a is ic R is close o one when g oups a e comple ely di e en ; all non-signi ican pai -wise compa isons a p>0.06). Compa aciones pa eadas ob enidas en los análi- sis ANOSIM de una ía compa ando la composición de las co- munidades en e las á eas geomo ológicas (No he n: a enas es abilizadas del no e, Sou he n: a enas es abilizadas del su ; Mobile Dunes: dunas mó iles, Pe idune: eco ono pe iduna y Ve a: eco ono de La Ve a). Se ealiza on es análisis ANOSIM de una ía, uno pa a cada g upo de especies: mac oin e - eb ados, an ibios y plan as acuá icas. Sólo se mues an los p- alo es ob enidos en compa aciones pa eadas signi ica i- as con p<0.05 en e las composiciones biológicas de las di e en es á eas mo ológicas (El es adís ico R de ANOSIM se ap oxima a 1 cuando los g upos son comple amen e di e- en es; odas las compa aciones pa eadas no ue on signi ica i- as con un alo de p>0.06). Rp Mac oin e eb a es Mobile Dunes ×No he n 0.21 0.05 Mobile Dunes ×Pe idune 0.22 0.013 No he n ×Sou he n 0.28 0.0001 Sou he n ×Ve a 0.17 0.002 Amphibians Mobile Dunes ×No he n 0.31 0.015 No he n ×Sou he n 0.37 0.0001 No he n ×Ve a 0.12 0.001 Sou he n ×Ve a 0.13 0.009 Mac ophy es Mobile Dunes ×No he n 0.43 0.001 Mobile Dunes ×Sou he n 0.19 0.038 Mobile Dunes ×Ve a 0.21 0.041 No he n ×Sou he n 0.68 0.0001 No he n ×Ve a 0.09 0.003 Pe idune ×Sou he n 0.58 0.0001 Sou he n ×Ve a 0.57 0.0001 among addi ional pai ed a eas we e also ound o mac oin e eb a es(1pai edcompa ison,Table 4), amphibians (1 pai ed compa ison, Table 4) and aqua ic plan s (4 pai ed compa isons, Table 4). DISCUSSION The ca ego isa ion o ponds acco ding o hei geomo phological a eas esul ed in a use ul and obus classi ica ion sys em ha e ealed impo - an di e ences in he soil ypes and assemblages o aqua ic lo a and auna. Because we lands in any egion de elop acco ding o geomo pholog- ical p ocesses ha a e d i en by he in e ac ion o clima e and geology (Gilman, 1994), geomo - phological and hyd ological ea u es can p o ide a sound basis o a global we land classi ica ion sys em in which biological a ia ion is indepen- den ly add essed (Semeniuk & Semeniuk, 1997). The Doñana we lands ha e been adi ionally g ouped in o dune ponds and ma shes acco ding o hei loca ion. Bo h ypes ha e been u he classi ied by size, hyd ological ea u es, and li o al ege a ion (as e iewed by Se ano e al., 2006). A quan i a i e hyd o-geomo phological app oach es ablished a h eshold o 1.6 mS cm−1 o seg ega e dune ponds om ma shes du ing hea y loods (Espina & Se ano, 2009). Ga cía- No o e al. (1991) sepa a ed dune ponds in o wo ca ego ies based on he Na+/Mg2+ a io o hei su ace wa e s, which dis inguished g oundwa e echa ge and discha ge a eas du ing loods. Fu he disc imina ion among ponds has no been achie ed, mos likely due o he limi ed amoun o in o ma ion ha can be ob ained by measu ing he concen a ions o dissol ed o suspended ma e ials unde wa e le el egimes ha show a high deg ee o luc ua ion (bu see Flo encio e al., 2009b). By con as , we used o dina ion o de e mine ha he pa e ns o pond soil a iables ag ee wi h he dis ibu ion o he di e en geomo phological a eas loca ed wi hin he Doñana aeolian sands and ha he sou he n s able dune pond soils we e seg ega ed om he pond soils in he o he a eas. The geomo phological a eas di e ed sig- ni ican ly in he assemblage composi ion o he main g oups o o ganisms ha inhabi ed hei ponds: mac oin e eb a es, amphibians, and aqua ic plan s. We ound ha he sou he n s able dunes and mobile dunes in pa icula had dis inc lo al and aunal assemblages, likely as a esul o he pa icula ea u es desc ibed 15624_Limne ica 33(2), pàgina 335, 09/11/2014