Effects of Environmental Stress on Ascidian Populations in Algeciras Bay (Southern Spain). Possible Marine Bioindicators?

MARINE ECOLOGY PROGRESS SERIES Ma Ecol P og Se Published Decembe 5 E ec s o en i onmen al s ess on ascidian popula ions in Algeci as Bay (sou he n Spain). Possible ma ine bioindica o s? S. A. Na anjo*, J. L. Ca ballo, J. C. Ga cia-Gomez Labo a o io de Biologia Ma ina, Dp o. de Fisiologia y Biologia Animal, Facul ad de Biologia, Apdo 1095, E-41080 Se illa, Spain ABSTRACT: The dis ibu ion and abundance o li o al ascidians we e analyzed wi h espec o hel possible ela ionships wi h en i onmen al s ess. As pa o a mul idisciplina y esea ch p ojec on he ben h~c communi ies in Algeci as Bay, sou he n Spain, a sui e o en i onmen al a iables was mea- su ed (hyd odynamism, sil ing, suspended solids and o ganic ma e ). A e displaying he similali ies o auna h ough clus e ing and o dina ion o sampling si es, he ela ionships be ween communi y di e ences and changes in he abio ic componen we e es ablished based on he BIO-ENV p ocedu e and Canonical Co espondence Analysis. Hyd odynamism and he pe cen age o o ganic ma e in he sil is he a iable combina ion ha bes explains (Spea man co ela ion o 0.82) he bio ic s uc u e. While all ascidians show a ce ain ole ance o di e se en i onmen al ac o s, some species such as Ciona in es inalis. Diplosoma spongi o me, Phallusia mammilla a, Mic ocosmus squamige , S yela pli- ca a and Synoicum a gus could be conside ed as indica o s o a eas which ha e been subjec o in ense s ess (subs a e ans o ma ion, wa e s agna ion and sedimen a ion excess) o e long pe iods o ime, whe eas o he s such as Aplidium conicu n, Aplidium punc um. Cla elina della allei, Halocyn hia papillosa and S olonica socialis, which li e only in na u al and non-pe u bed ock a eas, could be ca - ego ized as species e y sensi i e o s ess, as well as ind~ca o s o good condi ions. KEY WORDS: Ascidian communi ies . En i onmen al s ess . hdul i a ia e communi y measu emen s B~oindica o s INTRODUCTION The e is now widesp ead ecogni ion ha chemical moni o ing o pollu ion alone is no enough, and ha pollu ion is essen ially a biological phenomenon be- cause o i s impac on li ing o ganisms (W igh e al. 1994). The analysis o changes in ben hic communi y s uc u e has now become one o he mains ays in de ec ing and moni o ing he biological e ec s o ma ine pollu ion (Wa wick & Cla ke 1993). Ne e - heless, knowledge o he dis ibu ion pa e ns o ben- hic o ganisms along na u al en i onmen al g adien s (e.g. om u bulen o calm condi. ions) is necessa y be o e a possible an h opogenic dis u bance can be p o ed. Algeci as Bay ea u es impo an indus ial de elop- men s (wi h chemical indus ies, e ine ies, he mal powe plan s, i onwo ks and pape mills) as well as in- ense ha bou ac i i y. All o his esul s in a highly ans o med coas line (mainly in he cen al zone) whe e shipya ds, pie s and b eakwa e s, among o he po cons uc ions, a ec he no mal wa e low. Some a eas ha a e be e conse ed and ha e na u al ocks exis in he ou e zone. Recen s udies show he amoun o di ec was e in his a ea o be close o 2800 m"', wi h much mo e u ban sewage han in- dus ial was e-wa e s (Wai e al. 1990). consequen ly, Algeci as Bay can be conside ed o be a pa chwo k o small a eas unde di e en le els o s ess due o he in luence o en i onmen al and an h opogenic ac o s. Thus, a mul i ude o small en i onmen s can be iden i- ied and, co espondingly, a simila le el o he e o- genei y can be expec ed in he s uc u e o he ben hic communi ies wi hin hese en i onmen s. Algeci as Bay is conside ed o be an especially sui - able a ea o biomoni o ing s udies (Nai-anjo & Ga cia- O In e -Resea ch 1996 Resale o ull a icle no pe mi ed Ma Ecol P og Se 144: 119-131, 1996 Gomez 1993, Ca ballo e al. 1994, Con adi & Ce e a 1995, Ca ballo e al. 1996). In his sense, a mul idiscipli- na y esea ch p og am was conduc ed om 1991 o 1995 in o de o gain knowledge o he mac oben hic auna and i s ela ion o majo en i onmen al a iables in his a ea. Among all sessile o ganisms s udied, he ascidians a e a pa icula ly in e es ing g oup in his espec , since hey display bo h o he main ecological s a egies-soli a y and colonial-and a a ie y o mo phological ypes which a e o adap i e signi icance (Jackson 1977a, 1979) and which co ela e wi h dis inc abili ies o colonizing new su aces. Acco dingly, i is a good diagnos ic g oup whose d~s ibu ion may e lec some o he p e ailing s uc u ing ac o s a ec ing ben- hic assemblages (Tu on 1990) Ascidians a e ound on all subme ged ha d su aces, including conc e e, i on, ope and plas ics. Some species a e mos abundan in highly ans o med and pnll~~ e i ~n i nn nen q 5~1 -h as po s, ha ho s and in- dus nal a eas, whe e he excess o pa icula e o ganic ma e and he p oli e a ion o bac e ia a e a sou ce o ood o il e - eeding in e eb a es, pa icula ly ascidia.ns (Monnio e al. 1991). Ascidians can also ake up a ious me als om he wa e ( anadium is one o he mos common) and some oxic subs ances (hea y me als and hyd oca - bons). The e o e, ascidians con ibu e o he cleaning and pu i ica ion o was e-wa e s. Fo his eason, nume ous s udies ha e ou lined he impo ance o his g oup as pollu ion bioindica o s (Papadopoulou e al. 1972. Papadopoulou & Kanias 1977, Monnio 1978, Bell e al. 1982). In spi e o his, li le is known abou he in luence o local en i onmen al ac o s (such as wa e mo emen , suspended solids o sil ing) on he p esence o absence o he species in a gi en habi a . In his sense, we ha e s udied he composi ion and abundance o he ascidian communi ies in Algeci as Bay, as well as hei le el o adap a ion o abio ic com- ponen s. MATERIAL AND METHODS Ascidians we e sampled a 11 coas al s a ions dis- ibu ed a ound Algeci as Bay, a dep hs om 3 o 15 m. Thei loca ions a e shown in Fig 1. Due o h.e he e ogenei y o he ha d-bo om habi a s (na u al ock o ma ions, po cons uc ions, shipya ds and o he a i icial subs a es), he s a ions we e selec ed on he basis o en i onmen al a iabili y. In en o ies o species pe s a ion we e aken o he specimens collec ed by SCUBA-di e s along con inuous ansec s (app oxima ely 50 m long). These in en o ies we e comple ed by isual and pho og aphic sampling along he same ansec s, wi h a simila imme sion ime a Fig. 1 Loca ion o he sampling s a ions in Algeci as Bay, sou he n Spain all s a ions ( he e we e no eplica es on any occasion). All ascidians sigh ed a a dis ance o 1 m on ei he side o he ansec s we e collec ed and/o pho og aphed. Su eys we e aken om Sep embe 1992 o No embe 1993. Table 1 sho s he species abundance in e ms o semi-quan i a i e abundance codes. Coding can be use ul when di e en sampling me hods a e u ilized, and in his s udy i pe mi ed mo e objec i e compa - isons o he aunis ic composi ion a di e en s a ions (Ca ballo e al. 1996). Fu he mo e, his p ocedu e is p esumed o ha e an e ec sim~la o s anda diza ion by columns, which minimizes da a a iabili y a ising om sampling he e ogenei y (Maldonado S1 U iz 1995). The es abl~shmen o en i onmen al a iabili y among s a ions was based on he analysis o 5 abio ic a iables ha could be ela ed o en i onmen al s ess. These we e hyd odynamism, sil ing, suspended solids, sus- pended o ganic ma e and o ganic ma e in he sil (SOM) ga he ed in collec ing bo les. The me hods used o es ima e hese pa ame e s a e de ailed in Ca - ballo e al. (1996); measu emen was made as in Gambi e al. (1989) o hyd odynamism, Moo e (1972) o sill- ing and S ickland & Pd sons (1960) o he o he h ee. As a p epa a o y s ep o he s a is ical analysis, hese abio ic a iables we e log- ans o med. Na anjo e al. E ec s o en i onmen al s ess on ascidlans 121 - Compa isons among sampling s a ions (on he basis o species abundance) we e conduc ed acco ding o he s a egy ou lined by Field e al. (1982) o he analysis o da a on communi y s uc u e: (a) The bio ic ela ionship be ween any 2 samples (sampling s a ions in ou case) was dis illed in o a coe icien measu ing simila i y (o dissimila i y) in species composi ion. The B ay-Cu is lndex (B ay & Cu is 1957), which is no a unc ion o join absence, was chosen o his pu pose. I is ecognized as an e icien measu e o he e a- lua ion o a ini ies on semi-quan i a i e da a om la ge aunis ic assemblages (Gami o & Ra aelli 1992). (b) S a ions we e classi ied in o g oups (using he i- angula ma ix o simila i y ob ained be ween e e y pai o s a ions) ei he by hie a chical agglome a i e clus e ing, wl h g oup-a e age linking (e.g. Snea h & Sokal 1973), o by mapping he s a ion in e - ela ionships in o o dina ion using non-me ic mul i- dimensional scaling (MDS) (K uskal & Wish 1978). (c) Rela ionships be ween species we e analyzed by ansposing he da a ma ix and epea ing he classi i- ca ion and o dina ion me hods on he new simila i y ma ix be ween e e y pai o species. Species which we e indica i e o he dissimila i y be ween g oups o s a ions we e de e mined as in Cla ke (1993), using he SIMPER compu e p og am. In o de o link he mul i a ia e communi y s uc u e ( e lec ed in he aunis ic he e ogenei y o he di e en sampling s a ions) o en i onmen al a iables (which also change wi h each s a ion), we used 2 di e en , bu complemen a y, s a is ical echniques: (1) BIO-ENV p ocedu e (Cla ke & Ains~ o h 1993). In summa y, his consis s o a sepa a e compa ~son o he among-s a ion simila i y ma nx o he bio a wi h equi- alen iangula ma ices o all combina ions o abio ic a iables, and he consequen choice o he subse o en i onmen al a iables which p o ides a good ma ch be ween he 2 con igu a ions. A ma ch was measu ed by he Spea man ank co ela ion coe icien (p,), which was compu ed o he dissimila i y ma ices o he bio ic and abio ic da a. Al hough no speci ic plo is p o ided showing he esul , a diag am displaying he bio ic MDS (based on he B ay-Cu is dissimila i y index) in con- junc ion wi h he abio ic MDSs (compu ed om Euclidean dis ances) o he subse o en l onnlen al a iables selec ed by he BIO-ENV p ocedu e as op imal would be app opia e, especially when high alues o p, a e ob ained (e.g. p, % 0.8) (Cla ke & Ainswo h 1993). (2) Canonical Co espondence Analysis (CCA). This is a di ec g adien echnique, and ep esen s a special case o mul i a ia e eg ession (Palme 1993). In he e- sul ing o dina ion diag am, en i onmen al a iables can be ep esen ed by a ows along wi h he species and/o s a ion sco es. De ails o how o in e p e CCA diag ams a e gi en in e B aak (1986) and Palme (1993). The so wa e used was: PRIMER e sion 3.1 om Plymou h Ma ine Labo a o y, UK, ol- clus e , MDS, SIMPER and BIO-ENV analysis; STATGRAPHICS 6.0 o eg ession analysis; and CANOCO 3.12 o Co e- spondence Analysis, CCA and Mon eca lo pe mu a- ion es s. RESULTS S a ion g oupings based on species composi ion Table 1 indica es la ge a iabili y in he na u e, ype and a ailabili y o ha d subs a es cha ac e izing he sampling s a ions. A p elimina y s udy o he local dis- ibu ion o he mos common ascidians indica ed 3 basic g oups: (a) species exclusi ely dis ibu ed in he ou e zone o he bay which se le mainly on na u al ock subs a es; (b) species mo e abundan in he inne zone o he bay which se le mainly on a i icial sub- s a es; (c) species ound h oughou he bay. Analysis o he dend og am in Fig. 2 shows he dl - e ences among he s a ions as s a ed abo e. This clas- si ica ion echnique sepa a es he s a ions in o 3 majo g oups (A, B, C) which a e ela ed, in pa , o he na u e o he subs a e. In g oup A, he s a ions a e cha ac e ized by he p esence o ex ensi e na u al ock o ma ions (in he ex emes o he bay). G oups B and C include he es o he s a ions (In he inne pa s o he bay), whe e a i icial subs a es a e dominan (walls and columns in po s, b eakwa e s, shipya ds, e c.). The a e age dissimila i y be ween g oups A and B 1s 59.74%. As is shown in Table 2, 17 o he 38 specles explain he dissimila i y o 70% be ween g oups A and B. Pseudodis o na obscu u n, Polyci o ad ia icu n, Fig 2. Dend ogl-am o s a ions using g oup-a e age clus e - lng om B ay-Cu ~s simila i y on specles abundances The 3 g oups o s a ions [A o C) sepa a ed a a 50% simila i y h eshold (do ed line) a e indica ed 122 Ma Ecol P og Se 144 119-131, 1996 Table 1. Asc d~an species abundance In Algec~ as Bay and subs a e cha ac e ~s ~cs a sampl~ng s a lons Sem~quan ~ a ~ e abundance codes, 0 = species abhen , 1 = a e (1 o 2 speclmens eco ded), 2 = equen ( om 3 o 10 specimens pe ~ nme s~on) 3 = e y common (an a e age o mo e han 10 speclmens pe imme sion) Subsi a e onen a ion H = ho izon al su aces, V = e ical su aces, 0 = o e hangs U = unde boulde s; Cc = c e lces and can les. E = ep~b~on Na u al ock ypes a = la ge sub- me ged boulde s, b = s nall s ones on sandy bo oms; c = ock slabs, d = biode ~ ~c bo oms, e = Caule pa p ol e a weeds A l- ic~al subs a e ypes a = piling o conc e e blocks in b eakwa e s, b = e lcal su aces In po walls, c = ple s in sh~pya cis, d = o he s a l ic~al s uc u es sca e ed on bo om Dep h (m): 3 o 7 8 o 15 3 o 5 3 o 10 0 o 4 2 o 6 10 o 18 3 o 9 2 o 5 l o 6 4 o 10 Subs a e A ~ lc~al subs a e: - - - b,c a,c a,c,d b,c.d a,d d a,c,d b,d onen a lon Na u al ock a,b a,b a,b,c c c c,d c,d c c,e c,d a,c S n 1 S n 2 S n 3 S n 4 S n 5 S n 6 S n 7 S n 8 S n 9 S n 10 S n 11 Apli d~ um conicum Aplidiu n elegans Apl~diu n pallidum Apl~d~um punc um Ascldia men ~lla Asadiella aspe sa Ascidjella sca b a Bo nllus leach1 Ro yllus schlossen Qon<~ edwa ds~ Ciona ~n es lnalis Cla elina de!Ja allej Cla el~na lepadi o mis Cla el~na nana Dide nn u n conaceu n Didemnum maculosun~ D~plosoma lis enanun~ Diplosoma spong~io me Dls om~ls a iolosus Halocyn h~a pap~llosa Mic~ ocosm us squamige Molgula bleize! Molgula occlden al s Phallus~a umiga a Phallusla inge ia Phallusla mammilla a Polyu o ad ial~cum Polyci o c ~s aljn Llm Polysync a on lacazei Pseudodls oma obscu um Pyu a nlc ocos n us Rhopalaea neapol~ ana S oloiuca soclal~s S yela canopus S yela phca a Synoicum a]-gus Syno~cu n blochmann~ T ld~demnu n cei-eum A4ic ocosmus squamige and Synoicum a gus can be conside ed as good disc ~.mina o s because hey con- nbu e he highes alues o he a e age disslmilan y (6) and because hey ha e easonably high alues o he S,/SD(~,) a io (Cla ke 1993) The s a ions whlch make up g oup C (5 and 9) show he lowes le els o speci ic ichness in he bay and ha e a aunis ical composi ion ypical o ha bou a eas (h/lic ocosmus squamige , S yela plica a, Synoicum a gus, e c.); hus, hey a e mo e closely ela ed o he s a ions in g oup B (40 %, simila i y). MDS o dina ion was pe olmed using he smla i les be ween species In he esul ing 2-dimensional con lg- u a lon a de ~ni e species agg ega ion can be obse ed, bu a di ec in ep e a ion could be inaccu a e Molgula bleizel and Ascidiella aspe sa a e abundan bu exclu- si e o S n 9, so hei simila i y wl h he o he species is e y low The emo al o hese ou lie specles 1s ecom- mended (Gauch 1982, Cla ke & G een 1988) in o de o ob ain a mo e eliable pic u e o he emaining species Al hough s ess ends o dec ease as dlmenslonal~ ~ is inc eased (Field e a1 1982), when MDS was pe o med Na anjo e al.: E ec s o en i onmen al s ess on ascidlans 123 in 3 dimensions IFia. 3). he swecies' ela- Table 2. A e aqe abundance (7) o impo an ascidian species In q oups A [l , ., . ionships became mo e e idei , n F,~, 3 i o 3, I l) and B (4, 6 o 8, 10) a .kllgeci as Bay s a ions. species a eiis ed in 0;- de o hei con ibu ion (8,) o a e age dissimila i y (8) (= 59.74) be ween is o iden i y o he 2 g oups (only pa ly gi en) Species names gi en In ull in Table 1 species based on he maio ends o dis i- which li e on na u al ock a e on he igh S. a gus 2.8 0.25 4.15 1.38 3.01 side o he diaa am IA~lidium ~allidum, l? ad ia icum 0.6 3.0 4.13 2.41 1.71 13.86 695 1 bu ion in he bay: species mainly dis ib- u ed a he ou e pa o bay (S ns 1 o 3) - - Species Ys YA 6, SD(6,) 8, ,'SD(6,) E$,(%8) ga a, Rhopalaea neapoli ana, Synoicum I A. eleqans 2.0 1.25 2 14 1.73 123 39.41 1 d . . S olonica socialis, Cla eljna della allei, Tn- didenlnum ce eu n, Pseudodis oma c is al- linum, Aplidium punc um, Phallusia u nl- M. s4Uamige 2.6 0.25 4.02 2.18 20.58 P obscu u n 0.0 2.5 3 97 0.64 6.25 c is allnus 0,0 63 ,73 27.24 1 52 31.63 A, 1.2 2.5 251 1.90 132 35.84 conjcum. Ciona edwa dsi, A~lidium ele- I D. sponqi o me 1.8 1.0 186 1.10 170 59.48 1 blochmanhi, ~seudodis oma obscu um). Species which a e conside ed o ha e a medium le el o ole ance in ela ion o he subs a e (Polyci o ad ia icum, Aplidium S. b~~i ~mannj 0.4 1.5 2.13 1.20 1.78 42.98 A. punc "m 0.2 2.06 1.95 1.06 46.44 C.della allei 0.0 1.5 2.02 2.11 0.96 49.82 C, lepadi o m,s 2,2 1.99 1.46 1.37 53.15 P mammilla a 1.4 0.5 1 92 1.52 1.26 56.36 na u al ock, bu hey can also colonize , . gans, Didemnum n~aculosum, Cla elina nana, Bo iyllus leachi, e c.) a e g ouped in he middle; hey a e mo e abundan o e a i icial subs a es. These species a e dis- . - ce eum 0.2 1.25 173 1.61 1 08 62.38 D. a iolosus 1.6 1.0 1.71 1.18 1.45 D, maculosum 65.24 1.58 1.17 1.35 67.90 D, l;s e janum 1.4 0.75 1.55 1.11 1.40 70.49 ibu ed h oughou p ac ically all o he bay, excep o in luence o hyd ological ac o s along i s ma gins is un- he mos in e nal a eas (pie s and shipya ds). Finally, e en. The 5 en i onmen al a iables eco ded showed a species g ouped on he le (Ciona in es inalis, Mic o- no able a iabili y a bo h spa ial and empo al le els cosmus squamige , Synoicum a gus, S yela plica a, ( he la e is no aken in o accoun in his wo k). Fig. 4 Phallusia mammilla a and Cla elina lepadi o mis) a e (b o g) shows a mul iple box and whiske plo o each ypical species o ha bou a eas and colonize e e y ype a iable analyzed. In spi e o some sampling s a ions be- o a i icial su ace ing placed di e en ly, he ela ionships among s a ions we e simila o he ones in Ca ballo e al. (1996). Due o he p oximi y o he S ai s o Gib al a , he Rela ionship be ween species and en i onmen al hyd odynamic condi ions depend on majo coas al a iables cu en s and, in consequence, on idal lows and p e- ailing winds. Ex eme alues o hyd odynamism we e Abio ic cha ac e is ics o he s a ions egis e ed a ou e (and mo e exposed) s a ions (1 o 3 and 11). The high hyd odynamism eco ded a S n 8 Because o he many coas al ans o ma ions in Alge- can be explained by he in luence o 2 li o al cu en s ci as Bay (due mainly o indus ial de elopmen ), he which appea du ing low- ide pe iods (Camiiias 1987). Fig. 3. Th ee-dimensional MDS con igu a ion o species slmi- lan y ma nx (s ess = 0.09). Values o Asc~diella aspe sa and Molgula ble~zei we e e- mo ed p e ious o analyses. Species names a e abb e ia ed om Table 1 124 Ma Ecol P og Se 144: 119-131, 1996 l D,, = 4 S a ~on S a ion Fig 4 (a) Speci ~c nchness (D,,,) o ascidians a he sampl~ng s a ions; (b o g) mul ~ple box and wh~ske 'O plo o each abio ic ac o egis e ed a he 11 sam- pling s a ions. (d) as (c] bu wi hou S n 5 S a ~on S a ion In gene al, hyd odynamism is in e sely ela ed o sil ing, bu a S n 5 hese pa ame e s seem occasionally pa allel (e.g. du ing he ainy season). The unusual condi ions o his a ea a e due o he in luence o he Palmones Ri e (sedimen a ion is inc eased by he i e 's low) Thus, bo h he ex eme and he highes a e age annual alues o sil ing (195.6 and 74.04 g m-2 mo-', espec i ely) and o SOM (13.2 and 7 g m-2 mo-', espec i ely) we e egis e ed he e. On he o he hand, suspended solids and suspended o ganic ma e eached he highes alues a S n 9, whe e wa e enewal is limi ed by a pa ially enclosed ha bou zone. Hyd odynamism seems o in luence he o he ac- o s. Se e al en i onmen al ela ionships can be es ab- lished among he se o anables measu ed (and log- ans o med) by using co ela ion analysis. Whe eas hyd odynamism and suspended solids show a nega i e co ela ion ( = -0.72; p = 0.018), a posi i e linea ela- ionship exis s be ween sil ing and suspended solids ( = 0.54; p < 0.1), as well as be ween bo h a iables and hei o ganic ma e con en s (absolu e alues only, see Fig. 5). Uni a ia e and mul i a ia e communi y measu es A e he a ia ion le els o majo hyd ological ac o s we e eco ded, di e en s a is ical me hods we e ca - ied ou in o de o es ablish a ela ionship be ween abio ic he e ogenei y and auna1 dis ibu ion ( om as- cidian species abundance) h oughou he su ey a ea. As a i s app oxima ion, he bio ic in o ma ion ela- i e o each sampling s a ion was summanzed by sim- ple uni a ia e measu es such as he numbe o species ound (S) o he speci ic ichness. Fo his pu pose, speci ic ichness has been es ablished om he Ma - gale index: D*,, = S - l/ln(N), whe e N is he o al numbe o specimens collec ed (e.g. Bakus 1990). Al- hough S and DM, a e usually co ela ed ( o his case: = 0.67; p < 0.05), speci ic ichness gi es a quan i a i e measu e o di e si y a oiding some p oblems inhe - en o o he indexes such as Shannon's o Simpson's (Magu an 1989). Fig 5 Llnea co ela ions be ween he o ganic ma e (0 M ) con en ( o al and pe cen age) wi h sil ing and suspended solids + posi i e, and - nega l e co ela ions 'p < 0 1, '"p < 0 001 Na anjo e al.: E ec s o en i onmen al s ess on ascidians 125 - Fig. 4a shows he speci ic ichness h oughou he bay. The low alues o his pa ame e (DblEI < 3.1) ob ained a S ns 5, 6 and 9 a e ema kable, as well as he highes speci ic ichness (Db1, > 7) a S n 2. This bio ic in o ma ion p o ided no signi ican ela ionship wi h he se o en i onmen al a iables conside ed he e; howe e , a de ini e posi i e ela ionship be ween speci ic ichness and hyd odynamism and an in e se one wi h espec o suspended solids is sugges ed ( = 0.53 and = -0.51, espec i ely; p < 0.1). The biological da a a e bes ep esen ed using a mul i a ia e summa y, such as an MDS o dina ion (Cla ke & Wa wick 1994). Fig. 6 shows he spa ial o dina ion o s a ions based on species abundances o he se o abio ic a iables. In hese plo s, especially in he i s one, he o dina ion o s a ions wi h ega d o he ho izon al axis can be explained on he basis o a hypo he ical hyd odynamism g adien , so ha he si es mos exposed o cu en s a e clus e ed o he le and hose in calm condi ions o he igh . Howe e , S n 5 modi ies his gene al pa e n and, especially wi h sil - ing, i s ex eme alues can lead o a misin e p e a ion o he plo s (Gami o & Ra aelli 1992). The ep esen a- ions o he es o he abio ic a iables do no explain he s a ion g oupings in a consis en manne ei he . The limi a ions o he MDS echnique a e e iden , bu i can be used p e ious o mo e de ailed analyses. The BIO-ENV p ocedu e is conside ed o be a be e app oach o he p oblem. Table 3 shows he combina- ions o en i onmen al a iables which gi e ise o he la ges ank co ela ion (p,) be ween bio ic and abio lc sample dissimila i ies. The combina ion o a iables which bes g ouped he s a ions, in a manne consis- en w~ h he ascidian composi ion, in ol es hyd ody- namism oge he wi h SOM (ps = 0.82), and he com- bina ion o hese wi h suspended solids (p, = 0.78). Hyd odynamism is he single ac o which eaches he maximum ma ching coe icien (p, = 0.64), and when u he explana o y a iables a e added p, inc eases. Howe e , i sil ing o suspended o ganic ma e a e included he coe icien dec eases, p obably due o hei mino e ec s on communi y s uc u e (Cla ke & Ainswo h 1993). I he mos impo an en i onmen al ac o s o s uc u ing he communi y we e known, he locali ies ha ing a he simila alues o hese ac o s would be expec ed o p esen a he simila ascidian composi- ion, and an o dina ion based on his en i onmen al da a would g oup s a ions in he same way as o he bio ic plo (Cla ke & Wa wick 1994). Fig. ?a shows he MDS o species abundance a 10 s a ions (S n 5 emo ed). The emaining plo s in his igu e a e hose o he bes 2- and 3- a iable combina ions selec ed by BIO-ENV and ha o he comple e se o a iables measu ed. As Cla ke & Ainswo h (1993) s a ed, o consis ency o p esen a ion, hese plo s a e also MDS o dina ions, ye based on he Euclidean dis ance o no malised abio ic a iables. This app oach would he e o e be accep able, since he small numbe o a iables lead o only mino di e ences be ween MDS and he co esponding P incipal Componen s Analy- sis, which is expec ed when he highe -dimensional s uc u e is well- ep esen ed in 2 dimensions (no e he low alues o s ess ob ained). Fig. 6. Two-dimensional MDS o dina ion o s a- ions. Supe imposed ci - cles ep esen he annual a e ages o each o he abio ic a iables (s ess = 0.05) hyd odynamism b b.. 36 c, b ? s~ll ng am o 0 SOM suspended sol~ds ' 0 00 0 '0 0 0 suspended o ganlc malle -0 "63 O @ 0 126 Ma Ecol P og Se 144: 119-131, 1996 Table 3. Resul s o BIO-ENV analysis combina ions o en i onmen al a iables, aken k a a ime, y~elding he 'bes ma ches' o bio ic and abio ic simila i y ma nxes o each k, as measu ed by s anda d Spea man coe icien p,. Bold ype indica es he combi- na ion w~ h maxlmum p, o e all. En i onmen al ac o s hyd - hyd odynamism; susOM = suspended o ganic ma e (pe cen age), SOM = o ganlc ma e in he sil (pe cen age); sus = suspended solids; sil = s~l ~ng k 1 hyd (0.64) 2 hyd, SOM (0.82) 3 hyd, sus, SOM (0.78) 4 hyd, susOM, SOM, sus (0.67) 5 hyd, susOM, SOM, sus, sil (0.62) Bes a iable comb~na ~ons (p,) SOM SUS susOM sil (0.50) (0.21) (0.17) (-0.07) SOM, sus SOM, susOM hyd. sus . . . (0.66) (0.56) (0 491 hyd, SOM, susOM hyd, SOM, sll (0.74) (0.64) hyd, susOM, SOM, sil hyd, susOM, SOM, sil (0.65) (0.61) The high deg ee o conco dance be ween bio ic and abic ic plo s in Fig. 7 is e ?a kab!e, pa icula ly o dina ions (a) and (b), which g oup he si es in a e y simila ashion. The clus e o med by S ns 1 o 3 and 11, which a e loca ed in he ou e zone o he bay, a e bo h consis en in he i s 3 plo s as is he poo ela ionship o S n 9 o he o he s. Table 3 shows how p, declines sligh ly when u he a iables, apa om hose mos ele an , a e added; his is also e lec ed in he co esponding MDS. In ac , he BIO- ENV p ocedu e has a na u al s opping ule o he cases whe e p, dec eases wi h inclusion o unimpo - an a iables (Cla ke & Ainswo h 1993). This is shown in he MDS based on comple e abio ic da a (Fig. ?d), whe e he lack o conco dance wi h he bio ic MDS is ob ious. hyd odynamlsm + SOM 77 hyd odynamism + SOM + sus~ended solids all en l onmenlal a iables 10 On he o he hand, he highes p, ob ained (close o 1) shows a s ong in luence o he abio ic a iables on he ascidian popula ions, bu i is di icul o see i hey a e ue en i onmen al g adien s, since hei le el o in lu- ence on each species emains c yp ic. A CCA could be use ul o sol ing his p oblem, gi en ha in he esul - ing biplo s he species o dina ion is di ec ly ela ed o he a ia ion o abio ic ac o s ( e B aak 1986). In o de o a oid dis o ions which can be caused by a e da a (bo h bio ic and abio ic a iables), all da a om S n 9 and o AscidieUa aspe sa and Molgula bleizei we e emo ed. Fu he mo e, as suspended solids and he e- spec i e o ganic ma e p opo ions we e s ongly co - ela ed ( = -0.92; p c 0.051, he o me a iable was omi - ed so ha mul i-colinea i y p oblems would no occu . In his manne , CCA was conduc ed on a e ised se o da a om 4 en i onmen al a iables eco ded a 10 s a ions wi h a o al o 36 species ound. The biplo o species and hy- d ological ac o s p o ided by CCA is shown in Fig. 8. Axes I and I1 expla.in 73.4 O/n o o al a iance in he da a, wi h eigen al- ues o 0.32 and 0.11, espec i ely. The abi- o ic ac o s mos ela ed o axis I a e hyd o- dynamism and SOM, whe eas sil ing is close o axis I1 (see Table 4). The obus ness o he analysis was de e mined using he Mon e-Ca lo pe mu a ion es (Hope 1968, in e B aak 1988). The esul s we e as ollows: F- a io = 2.1; p c 0.01. The ou come o CCA was also checked by compa ing he species- ac o co ela ions wi h hose ob ained by Co espondence Analysis (CA) and De ended Co espon- dence analysis (DCA), whe e no e iden Fig. 7. MDS plo s o 10 sampling s a lons based on: (a) abundance da a o 38 ascidian specles; (h) and (c) he bes 2- and 3- a iable combina ions o di e ences we e ound. ans o med en i onmen al a iables: and Id1 all en i onmen al a ~ables In he biplo , specles a e exp essed as ~, (see ex ). S ess alues a e: 0.04; 0; 0.01; and 0.03, espec i ely poin s and en i onmen al ac o s as Na anjo e al.: E ec s o en i onmen al s ess on ascidians SOM 1 0 C ~n 0 I A.palo suspended o ganic Fig. 8. Species- ac o biplo based on Canonical Co - ma e espondence Analysis o a 36 species da a se wi h e- spec o he abio ic a iables ( ep esen ed by a ows). Each a ow poin s in he di ec ion o maximum change o he espec i e en i onmen al a iable sil ing ac oss he diag am, and ~ s leng h IS p opo ional o he a e o change in his di ec ~on. Specles names a e abb e ia ed om Table 1 a ows. The leng h o an a ow indica es he impo - ance o his ac o . Each a ow de e mines a di ec ion o axis in he diag am, ob ained by ex ending he a ow in bo h di ec ions. The p ojec ions o a species on his axis shows i s p e e ence o high o low alues o his en i onmen al g adien ( e B aak 1986). We hus in e ha Pseudodis oma obscu um, Polyci o c is alinus, Polyci oi- ad ia icu n, Synoicum bloch- manni, Halocyn hia papillosa and Rhopalaea neapoli- ana p e e a eas highly exposed o cu en s and high le els o SOM; species such as Aplidium punc um, T j- didemnum ce eum, Cla elina della allei and S olonica socialis ha e he highes weigh ed a e age wi h espec o suspended o ganic ma e ; while Ciona in es inalis, Ascidiella scab a, S yela plica a, Synoicum a gus, Mic ocosmus squamige , Ascidia men ula and Phallusia mam nilla a a e mos abundan in calm si es whe e highes con en s o SOM (absolu e alues) a e eached. The emaining specles p e e a e age alues o show no clea p e e ences. Table 4. Canonical coe ~cien s o and in ase co ela ions o DISCUSSION Rela ion o abio ic a iables In ben hic communi ies, physical dis u bance, p e- da ion and compe i ion a e conside ed o be he mos impo an ac o s a ec ing he abundance and dis i- bu ion o species al hough, in ela ion o ascidians, he signi icance o la al beha iou mus also be aken in o conside a ion (Buss 1979, S ane & Lundal 1982, Da is 1987, S ane & Young 1989). Along wi h hese ac o s, he subs a e ea u es (na u e, size and a ailabili y) de e mine he composi ion o ben hic sys ems (Jackson 197713). In Algeci as Bay, we ind ha he he e ogenei y o he en i onmen s in es iga ed was condi ioned by he ha bou wo ks and u ban de elopmen s exis ing along i s li o al. This physical a iabili y, along wi h he di - e en ypes o u ban and indus ial was es, clea ly a ec s he numbe and composi ion o species p esen , which ha e di e en le els o adap a ion and ole - ance. The e o e, he dis inc in luence ha he maio 1 . P 1 and loca ion o subs a es in he bay. Those species en ~ onn~en al a iables wi h he i s 2 axes o CCA. susOM: en i onmen al ac o s ha e along he coas line should suspended o ganic ma e (pe cen age); SOM: o ganic ma e in he sil (pe cen age) be conside ed be o e he p incipal ends in he spa ial dis ibu ion o ascidians can be es ablished. Va iables Canonical coe icien s In ase co ela ions Axis 1 Axis 2 Axis 1 Axis 2 and Cla elina lepadi o mis) we e mo e abundan in The simila i y analyses show a clea co espondence be ween he le els o species g ouping and he ype Sil ing -O 039 -0. 13 -0.041 -O 770 SUSOM -0 223 -0.298 -0.236 -0 322 Hyd odynamlsm -O 695 0,144 -0,737 0,156 SOM -0 779 0.260 -0.827 0.281 which p e e ligh , shallow en i onmen s and which colonize a i icial su aces (Mic ocosmus squamige , S yela plica a, Phallusia mam nilla a, Synoicum a gus