Physical–biological coupling in the coastal upwelling system of the Ría de Vigo (NW Spain). I: In situ approach

MARINE ECOLOGY PROGRESS SERIES Ma Ecol P og Se Vol. 353: 27–40, 2008 doi: 10.3354/meps07197 Published Janua y 17 INTRODUCTION New p oduc ion (NP; see Table 1 o lis o abb e ia- ions) is he ac ion o p ima y p oduc ion (Pg) ha is suppo ed by alloch honous nu ien sal s (Dugdale & Goe ing 1967) and is a key a iable o unde s anding ma ine ecosys ems unc ioning (e.g. exploi a ion o ma ine esou ces global clima e change). The e o e, NP no only ep esen s he h eshold o biomass exploi a ion (Quiñones & Pla 1991), bu also a ec s he e iciency o he biological pump as a ap o an h opogenic CO2(Longhu s 1991). The mos s aigh o wa d way o es ima e NP is o conside he nu ien balance (inpu minus ou pu ) in a gi en ecosys em (Smi h & Hollibaugh 1993, Wollas 1993). Howe e , his concep ually simple me hod © In e -Resea ch 2008 · www.in - es.com*Email: [email p o ec ed] o [email p o ec ed] Physical–biological coupling in he coas al upwelling sys em o he Ría de Vigo (NW Spain). I: In si u app oach S. Pied acoba1, 2,*, M. Nie o-Cid1, C. Sou o2, M. Gilco o2, 3, G. Rosón2, X. A. Ál a ez-Salgado1, R. Va ela2, F. G. Figuei as1 1CSIC, Ins i u o de In es igacións Ma iñas, Edua do Cabello 6, 36208 Vigo, Spain 2Facul ade de Ciencias do Ma , Uni e sidade de Vigo, Ma cosende, 36200 Vigo, Spain 3School o Chemis y, Physics & Ea h Sciences, Flinde s Uni e si y, GPO Box 2100, 5001 Adelaide, Aus alia ABSTRACT: The a e o he ino ganic and o ganic ni ogen apped in he coas al upwelling sys em o Ría de Vigo (NW Spain) was s udied a he 2 o 4 d ime-scale du ing July and Sep embe 2002. A ansien geochemical model was applied o he measu ed esidual cu en s and concen a ions o ino ganic ni ogen (NT), and dissol ed (DON) and pa icula e (PON) o ganic N o ob ain (1) he ne balance o inpu s minus ou pu s (i – o); (2) he ne accumula ion; and (3) he ne ecosys em p oduc ion (NEP) o NT, DON and PON. P e iously unaccoun ed la e al a ia ions in esidual cu en s and N species concen a ions in he ía we e conside ed. The ía was au o ophic du ing July (a e age NEP, 107 mg N m–2 d–1). Abou 25% o his ma e ial was expo ed o he shel and he emaining 75% was ans e ed o he sedimen s o p omo ed o highe ophic le els. Du ing summe upwelling episodes, 30 o 70% o he o ganic N expo ed o he shel came om ma e ials p e iously accumu- la ed in he ía. By con as , du ing summe elaxa ion, 60 o 70% o he accumula ed o ganic N came om in si u con e sion o NTand he emaining 30 o 40% was alloch honous. As shown by o he au ho s, NEP was a he low du ing in ense upwelling and inc eased o high alues du ing i s sub- sequen elaxa ion. In Sep embe , he me abolism changed om he e o ophic o sligh ly au o ophic (a e age NEP, 26 mg N m–2 d–1). The DON and PON impo ed om he shel du ing au umn down- welling expe ienced a di e en a e: DON was consumed and PON accumula ed in he ía. By con as , DON and PON we e p oduced in si u a he expense o N nu ien s p e iously accumula ed in o he sys em du ing au umn elaxa ion. The oles o e ical con ec ion and u bulen mixing in he e iliza ion o he pho ic laye we e also assessed. Mixing was he mos impo an e iliza ion mechanism, ensu ing anspo o nu ien s unde upwelling and downwelling condi ions. KEY WORDS: Ni ogen luxes · Geochemical model · Upwelling · Downwelling · Rías Baixas (NW Spain) Resale o epublica ion no pe mi ed wi hou w i en consen o he publishe Ma Ecol P og Se 353: 27–40, 2008 in ol es ope a ional di icul ies: eloci y and u bulen di usion p o iles, oge he wi h nu ien concen a- ions, ha e o be de e mined wi h a equency highe han he lushing ime o he s udy sys em. Fo his ea- son, di e en indi ec me hods ha e been de eloped o e he las ew decades o e alua e NP. G a i a ional sinking and subsequen mine aliza ion is he main a e o NP in as a eas o he open ocean, whe e he ho izon al ci cula ion is qui e slow and mos o Pg is p ocessed wi hin he mic obial loop (Legend e & Rassoulzadegan 1995, Co ne & Biddanda 2002). The e o e, assuming s eady-s a e condi ions, NP should be equal o he sedimen a ion a e measu ed a he base o he pho ic laye (Eppley & Pe e son 1979). In i o app oaches ha e also been used o es ima e N; hese include he 15N-labeled ni a e/ammonium+ u ea me hod (Dugdale & Goe ing 1967) o he ligh – da k oxygen incuba ion me hod (S ickland & Pa sons 1972), in which i is assumed ha he ne communi y p oduc ion (NCP) es ima ed in he ligh bo le is equal o he NP (Quiñones & Pla 1991). Howe e , he appli- ca ion o he la e me hod in sub opical gy es has led o con o e sial esul s (au o ophic NCP > 0 o he e o- ophic NCP < 0; Williams 1998, Se e e al. 2001, del Gio gio & Dua e 2002). In any case, he alidi y o shallow sedimen aps (Buessele 1991, Ducklow e al. 1995), 15N-labeled incuba ions (Legend e & Gosselin 1989, Wa a e al. 1995), and ligh –da k oxygen incu- ba ions (Hansell e al. 2004) o es ima e NP has been openly challenged. The e alua ion o NP is mo e di icul in coas al sys- ems, due o he enhanced ho izon al ad ec ion, and he ans e ence o a signi ican ac ion o Pg o me a- zoans (Legend e & Rassoulzadegan 1995, Co ne & Biddanda 2002). In he pa icula case o coas al upwelling sys ems a empe a e la i udes, he sho lushing imes and he non-s eady-s a e condi ions inc ease he complexi y o NP es ima ion (Ál a ez- Salgado e al. 1996a, Smi h & Hollibaugh 1997). The mass balance app oach using sedimen aps and/o in i o incuba ions may no e lec an accu a e NP es ima e o hese ecosys ems (Smi h & Hollibaugh 1997, Ál a ez-Salgado e al. 2001). In his s udy and i s companion pape (Pied acoba e al. 2008 his olume) we es ima e he NP in a coas al upwelling sys em (Ría de Vigo, NW Spain) unde con- as ing upwelling (July) and downwelling (Sep em- be ) condi ions using mass balance, sedimen ap, and in i o incuba ion app oaches. Pa I is ocused on he es ima ion o he inpu s minus ou pu s balance o ni o- gen species ( o al nu ien s, NT; dissol ed o ganic ni ogen, DON; pa icula e o ganic ni ogen, PON) on a sho ime-scale (<4 d) by combining dynamic and chemical de e mina ions along a ans e se sec ion in he cen al segmen o he Ría de Vigo. A mass balance box model was de eloped o e alua e he ela i e impo ance o ne accumula ion and ne ecosys em p oduc ion (NEP = NP) as a mechanism o o ganic and ino ganic ni ogen apping in his ecosys em. P e i- ous geochemical balances applied in he Ría de Vigo (Ál a ez-Salgado e al. 2000, Gilco o e al. 2001) es i- ma ed he eloci y ield om salini y and empe a u e dis ibu ions. In he p esen s udy, NP was es ima ed om di ec cu en measu emen s. In addi ion, he la - e al a iabili y o hese cu en s was conside ed o he i s ime. In Pa II, he me abolism o he Ría de Vigo is assessed while conside ing he oxygen p oduc ion/ 28 A F ee su ace a ea o he ía A S , A B Su ace and bo om a eas o he ans e se sec ion ADCP Acous ic Dopple cu en p o ile chl aChlo ophyll a C S , C B A e age su ace and bo om concen a- ions o C DON Dissol ed o ganic ni ogen DOM Dissol ed o ganic ma e ENACW Eas e n No h A lan ic Cen al Wa e F x Flux o any species ca ied by a ho izon al con ec i e low ( Q S, Q B ) F Z Flux o any species ca ied by e ical con ec i e low ( Q Z ) KZ Tu bulen mixing coe icien LNM Le el o no mo emen <N> B un -Väisälä equency N T Dissol ed ino ganic ni ogen NEP Ne ecosys em p oduc ion NEP T Ne ecosys em p oduc ion o N T NEP D Ne ecosys em p oduc ion o DON NEP P Ne ecosys em p oduc ion o PON NEP SED Ne balance o PON sedimen a ion, PON and esuspension and N T di usion om he sedimen M Z Ve ical di usi e luxes o – i Ne budge o ou pu s minus inpu s (o – i) T Ne budge o ou pu s minus inpu s o N T (o – i) P Ne budge o ou pu s minus inpu s o PON (o – i) D Ne budge o ou pu s minus inpu s o DON PON Pa icula e o ganic ni ogen PAR Pho osyn he ic a ailable adia ion Q B Bo om ho izon al low Q S Su ace ho izon al low Q Z Ve ical low R i Richa dson numbe ρDensi y U 0 Tidal eloci y ampli ude V Volume V X Ho izon al con ec i e eloci y V Z Ve ical con ec i e eloci y VMADCP Vessel-moun ed acous ic Dopple cu en p o ile Δ(V·N T )兾Δ Accumula ion o N T Δ(V· DON)兾Δ Accumula ion o DON Δ(V· PON)兾Δ Accumula ion o PON Table 1. Glossa y o ele an e ms Pied acoba e al.: Physical–biological coupling in Ría de Vigo. I espi a ion balance. Sedimen a ion and mic ozoo- plank on g azing a es, es ima ed o he i s ime in he Ría de Vigo, a e also conside ed. The NP es ima- ions ob ained by he in si u (Pa I) and in i o (Pa II) app oaches a e compa ed and he possible causes o he obse ed disc epancies a e discussed. MATERIALS AND METHODS Su ey a ea. The Ría de Vigo is a la ge (2.76 km3) V-shaped inden a ion in he wes e n coas o he Ibe - ian Peninsula (Fig. 1). I expe iences wind-d i en upwelling o cold, sal y and nu ien - ich Eas e n No h A lan ic Cen al Wa e (ENACW) om Ap il- May o Sep embe -Oc obe (Woos e e al. 1976, F aga 1981). Upwelling e en s occu wi h a 1 o 2 wk e- quency (Blan on e al. 1987, Ál a ez-Salgado e al. 1993). Downwelling is he dominan p ocess he es o he yea ; i is cha ac e ized by he ad ec ion o wa m and nu ien -poo shel su ace wa e o he ía (Cas o e al. 1997). The main ibu a y is he i e Oi abén- Ve dugo, which d ains an a e age annual low o 15 m3s–1 (Noguei a e al. 1997) in o San Simón Bay, a semi-enclosed basin o ~20 km2(Fig. 1). Sampling s a egy. Fi e s a ions in he middle seg- men o he Ría de Vigo (41 ± 2 m max. dep h) we e sampled abou 1 h be o e sun ise on 15, 18, 22 and 26 July and 17, 19, 23 and 26 Sep embe 2002. Wa e samples om S n 00 (Fig. 1) we e aken wi h a ose e sample equipped wi h wel e 10 l PVC Niskin bo les wi h s ainless-s eel in e nal sp ings, while wa e sam- ples om S ns R01 and R03 loca ed a he no he n and sou he n ends o he segmen , we e aken wi h 5 l PVC Niskin bo les. A S n 00, salini y and empe a u e we e eco ded wi h an SBE 9/11 CTD p obe a ached o he ose e sample . Conduc i i y measu emen s we e con e ed in o p ac ical salini y scale alues wi h he equa ion o UNESCO (1985). Wa e samples we e collec ed om 5 dep hs: a he su ace o a he dep h o 50% o he pho osyn he ically a ailable adia ion (PAR, µE m–2 s–1) pene a ion (2.5 ± 0.3 m), a 25% o PAR pene a ion (7 ± 1 m), a 1% o PAR pene a ion (14.2 ± 2.2 m), a 26.2 ± 1.2 m, and a he bo om. 29 Vigo Pon e ed a A ousa Ri e Oi abén Ve dugo Sillei o buoy Rande S ai Ibe ian Peninsula c 2 4 5 9 1 00 a 1000 1500 2000 –40 –30 –20 –10 0 7.3-10.8 NS CS 5.9-15.2 SS 6.3-10.6 NB SB 6.2-15.3 N(R03) S(R01) 9.5-17.0 b 9.0º 8.9º 8.8º Longi ude (Wes ) 42.1º 42.2º La i ude (No h) Rande S ai San Simon Bay C. Ma Cíes Islands P. Bo nei a DCM12 & CTD CTD s a ion Ría de Vigo R03 R01 20 m 20 m 60 m 80 m Fig. 1. Ría de Vigo (NW Spain). (a) The longi udinal ansec epea ed du ing he 60 isi s om 1990 o 1997. The c oss-ha ched zone is he ee su ace a ea o he olume bounded by he s udy ans e se sec ion. (b) The 5 sec o s— 3 a he su ace (NS, CS and SS) and 2 a he bo om (NB and SB)— de ined o s udy he la e al a iabili y o he middle ía. (c) The Ría de Vigo in ela ion o he Rías Baixas and he loca ion o he Sea wa ch buoy o Pue os del Es ado o Cape Sillei o Ma Ecol P og Se 353: 27–40, 2008 A S ns R01 and R03, salini y and empe a u e we e eco ded wi h a CTD SBE-25 and wa e samples we e collec ed om 2 dep hs: su ace and bo om. Wind and cu en measu emen s. Coas al wind da a we e ob ained om he Seawa ch buoy o Cape Sillei o (www.pue os.es; ou Fig. 1), a ep esen a i e si e o winds blowing o he Ría de Vigo (He e a e al. 2005). Low-pass il e ed winds a e used o de ine he p e ailing oceanog aphic scena io in he ía du ing each pe iod. The hea balance and eshwa e lux o each sampling da e we e aken om Pied acoba e al. (2005). Cu en eloci y p o iles we e measu ed aboa d RV ‘My ilus’ wi h an RDI essel-moun ed b oad- band acous ic Dopple cu en p o ile (VMADCP) o 300 kHz along he ans e se sec ion joining S ns R01 and R03 on 18 and 22 July and 19 Sep embe 2002. In addi ion, cu en p o iles we e measu ed wi h an Aande a DCM12 acous ic Dopple cu en p o ile (ADCP) moo ed a S n 00. The moo ing was deployed om 10 o 29 July and om 17 o 26 Sep embe 2002. The DCM12 lay on he seabed, eco ding, a 30 min in e als, he cu en in eg a ed in 5 e ical cells 11 m hick. DCM12 and VMADCP da a we e o a ed coun e clockwise 30° o p oduce a longi udinal and a ans e sal componen ela i e o he main axis o he ía. A C++ e sion o he HAMSOM model (3D, z-coo di- na e A akawa-C g id), ba oclinic, semi-implici ini e– di e ence nume ical p edic ion model), ecen ly ali- da ed in he ou e (Sou o e al. 2003) and inne (Diz e al. 2004) Ría de Vigo, was used o ep oduce he la e al a iabili y o he sub idal cu en ield on 15 and 26 July and 17, 23 and 26 Sep embe 2002. The accu acy o he model o ep oduce he la e al cu en ield da a has been success ully es ed by Pied acoba e al. (unpubl.). Ho izon al ad ec ion. Fi e sec o s— 3 a he su ace (NS, CS and SS) and 2 a he bo om (NB and SB)— we e de ined o accoun o he la e al a iabili y in he middle ía (Fig. 1b). The c i e ion o di ide he su ace and bo om sec o s was based on he le el o no ho i- zon al mo ion, aking in o accoun he da a om he DCM12 and he VMADCP, and he ou pu s o he HAMSOM model. Ho izon al ad ec ion eloci ies we e a e aged in he 5 sec o s. Ve ical ad ec ion. The Ría de Vigo was isi ed o he pu pose o measu ing mass balances 60 imes om 1990 o 1997. Fig. 1a shows he longi udinal ansec epea ed du ing he 60 isi s. The good ela ionship be ween he bo om ho izon al low (QB) and he e i- cal low (QZ) ob ained om he mass balances (Fig. 1b) pe mi s calcula ion o he unknown QZ(in m3s–1) om he measu ed QB(in m3s–1) in his s udy (n a e he pai s o consecu i e sampling): QZ= –212(±43) + 0.73(±0.03)· QB = +0.92, p < 0.001, n = 47 (1) The e ical con ec i e eloci y (VZ) was calcula ed conside ing he su ace a ea o he shaded zone in he inse o Fig. 1a. The e ical displacemen o he le el o no mo emen (LNM) be ween 2 consecu i e su eys was also calcula ed. Ve ical mixing. The u bulen mixing coe icien , KZ(m2d–1), was pa ame e ized acco ding o Munk & Ande son (1948), using he ADCP and CTD da a: KZ= K0·(1 + 10 · Ri)1 /2(2) K0= 3 ×10–3·U0·Z(3) (4) whe e K0is he eddy coe icien o neu al s abili y, U0 is he ampli ude o he idal eloci y om 0 o Zm dep h, Riis he Richa dson numbe , ρis he a e age densi y, and gis he g a i y accele a ion. Veloci y dimensions we e ob ained by di iding KZby Z/2, he dis ance be ween he g a i y cen e s o he uppe and lowe laye . S abili y. The wa e column a e age B un -Väisälä equency, <N> (in min–1), a S n 00 (Fig. 1) was calcu- la ed as ollows: <N> = {4g(ρl– ρu)兾(ρl+ ρu)Z}1 /2, whe e ρland ρu a e he a e age densi y o he uppe and lowe laye s. Chemical a iables. Salini y (S, accu acy ± 0.003), N-nu ien s (NH4+, ±0.05 µM; NO2–, ±0.02 µM; and NO3–, ±0.1 µM), DON (±0.2 µmol N l–1), suspended PON (±0.1 µmol N l–1), and chl a(±0.05 mg m–3) we e measu ed as desc ibed elsewhe e. Ho izon al/ e ical luxes. A e age concen a ions (µmol kg–1) o all species we e calcula ed om he 3 s a ions loca ed along he middle segmen o he ía by means o an in e se dis ance in e pola ion me hod. The ho izon al lux o any ni ogen species (FX, in µmol m–2 d–1) was calcula ed as ollows: FX= VXCXρX(5) whe e VXis he ho izon al eloci y, and CXand ρXa e he a e age concen a ion and densi y, espec i ely, a each o 5 sec o s de ined abo e. Ve ical con ec i e luxes (FZ, in µmol m–2 d–1) we e ob ained as ollows: FZ= VZCZρZ(6) whe e VZis he e ical con ec i e eloci y, and CZ and ρZa e he dep h-a e age concen a ion and den- si y, espec i ely, in he s udy sec ion. Ri=− ⋅∂ ∂ ∂ ∂ () g z u z ρ ρ 2 30 Pied acoba e al.: Physical–biological coupling in Ría de Vigo. I Finally, e ical di usi e luxes (MZ, in µmol m–2 d–1), we e ob ained as: (7) whe e CSand CBa e he a e age concen a ions o any species in he su ace and bo om laye , espec i ely. Geochemical budge . The ne budge o inpu s minus ou pu s, i – o, o any species (C)in o a gi en ol- ume (in µmol C m–2 d–1) esul s om he accumula ion, Δ(C⋅V)兾Δ , and ne ecosys em p oduc ion, NEPC, e ms: (8) The accumula ion e m (in µmol C m–2 d–1) can be sep- a a ed in 2 e ms: (9) The i s e m on he igh o Eq. (9) indica es he ne accumula ion due o changes in concen a ion and he second e m due o changes in olume. The a e - age NEPCbe ween 2 consecu i e su eys can be es i- ma ed as: (10) whe e he balance o o – i (in mol C m–2 d–1) is calcu- la ed as: o – i = (FS⋅AS– FB⋅AS)兾A(11) FSand FBa e he a e age su ace and bo om ho izon- al luxes (in µmol m–2 d–1); ASand ABa e he a eas o he su ace and bo om laye o he ans e se sec ion (in m2); and Ais he ee su ace a ea o he ía bounded by he ans e se sec ion (35.6 ×106m2). RESULTS Wa e lows No he ly winds p e ailed o e he shel in July (Fig. 2a), while sou he ly winds we e dominan in Sep- embe (Fig. 3a). Since con inen al uno was e y low in July (<9 m3s–1; Fig. 2b) and Sep embe ( om 5.7 o 14.5 m3s–1; Fig. 3b), 6 pe iods we e de ined acco ding o he dominan shel wind. The i s , om 15 o 18 July, was cha ac e ized by an in ense upwelling episode ha eached a maximum on 15 July. A posi i e esidual ci cula ion pa e n wi h a s ong su ace ou - low o –12 cm s–1 was eco ded (Fig. 2d). Du ing he second pe iod, om 18 o 22 July, no he ly winds aba ed, p oducing a e e sal o he esidual ci cula ion pa e n (su ace in low and bo om ou low; Fig. 2d). The hi d pe iod (22 o 26 July) was domina ed by no he ly winds ha p omo ed an in ense upwelling on 26 July. A posi i e esidual ci cula ion pa e n esumed (Fig. 2d), wi h su ace and bo om cu en s simila o he i s pe iod. The ou h pe iod, om 17 o 19 Sep embe , was domina ed by sou heas e ly winds ha dec eased g adually and caused a nega i e esid- ual cu en pa e n wi h a maximum su ace in low o 10 cm s–1 and bo om ou low o –10 cm s–1 (Fig. 3d). The i h pe iod, om 19 o 23 Sep embe , was cha ac- e ized by he ansi ion om sou he ly o no he ly winds (Fig. 3a), when he e e sal o he esidual pa - e n pe sis ed bu cu en eloci ies diminished. On 23 Sep embe , a 3-laye ci cula ion de eloped, wi h ou - low h ough he su ace and bo om laye s and in low h ough he in e media e laye . Finally, he six h pe iod, om 23 o 26 Sep embe , was cha ac e ized by mode a e no heas e ly winds blowing o e he shel wi h lowe in ensi y han in July (Figs. 2a, 3a). As a esul , he 2-laye posi i e ci cula ion pa e n was e- es ablished. On 15 July, he salini y and empe a u e dis ibu- ions showed he ypical esponse o an upwelling e en . Despi e he subsequen elaxa ion o shel winds, he e ec o coas al upwelling s ill emained in he bo om laye 3 d la e (Fig. 2e, ). The e e sal o he ci cula ion in esponse o he wind calm (Fig. 2a) a ec ed he wa e column s a i ica ion: salini y was mo e homogeneous and he halocline was di icul o dis inguish, while su ace empe a u e inc eased and he he mocline was ound a 20 m dep h. The he mo- haline s uc u e ound on 17 Sep embe co esponded o a ypical au umn downwelling, wi h wa m (>17°C) and ela i ely sal y (>34.8) shel su ace wa e s in he su ace laye (Fig. 3e, ). The pe sis ence o shel sou he ly winds p oduced a deepening o he su ace salini y and empe a u e as a consequence o down- welling. Sou he ly winds changed o no heas e ly on 22 Sep embe (Fig. 3a), p oducing a sligh upli o he isohalines and iso he ms, sugges ing he en y o oceanic ENACW h ough he bo om laye s on 26 Sep embe . The analysis o he longi udinal componen o he esidual cu en shows ypical 2-laye ed posi i e and nega i e ci cula ion pa e ns unde upwelling condi- ions on 15 July (Fig. 4a) and downwelling condi ions on 19 Sep embe (Fig. 4 ), espec i ely. A a ie y o in e media e cases we e obse ed on he o he sam- pling da es, showing he complexi y o he la e al esidual ci cula ion o he ía. Ve ical con ec i e eloci ies we e upwa ds du ing he upwelling e en s o 15 and 26 July (Fig. 4a,d) and 23 and 26 Sep embe (Fig. 4g,h). They we e nea ly ze o du ing he elaxa ion o upwelling (Fig. 4b) and e e sed o downwa ds on 22 July (Fig. 4c), and on 17 and 19 Sep embe (Fig. 4e, ). In July, mixing eloci ies NEP o i (V ) C=−+ ⋅Δ Δ C Δ Δ Δ Δ Δ Δ (V)V VC C C ⋅=+⋅ io (V) NEPC −= ⋅− Δ Δ C MKCC ZZBS z =− () Δ 31 Ma Ecol P og Se 353: 27–40, 2008 eached a maximum du ing upwelling e en s, being wice he e ical con ec i e eloci y on 15 July (Fig. 4a). Th oughou Sep embe , mixing eloci y was highe han he con ec i e e ical eloci y, excep on 17 Sep embe , and i eached a maximum on 19 Sep- embe (Fig. 4 ). The s a i ica ion ha ook place du - ing he elaxa ion a e he i s upwelling episode caused mixing eloci y ha was hal he e ical con- ec i e eloci y on 26 July (Fig. 4d). The e ical dis- placemen s o he LNM we e negligible compa ed o he e ical con ec i e and mixing eloci ies in July and Sep embe . Acco ding o ou p e ious hyd og aphic and dy- namic desc ip ion, he 6 pe iods we e named as ol- lows: (i) spin-down o s ong summe upwelling (15 o 18 July), (ii) summe elaxa ion (18 o 22 July), and (iii) spin-up o weak summe upwelling (22 o 26 July), (i ) au umn downwelling (17 o 19 Sep embe ), ( ) au umn ansi ion (19 o 23 Sep embe ), and ( i) au umn upwelling (23 o 26 Sep embe ). Ni ogen luxes NO3–(Fig. 2k), he mos abundan N-nu ien in July (69% o NT= NO3–+ NO2–+ NH4+), was a a maximum a he bo om laye du ing he spin-down o he s ong summe upwelling (15 o 18 July) because o he en y o ENACW. In he su ace laye , maximum concen a- ions o chl a(Fig. 2i) and PON (Fig. 2j) we e eco ded. Du ing he summe elaxa ion (18 o 22 July), lowe chl aand PON concen a ions we e eco ded a he su ace compa ed wi h he p e ious pe iod (15 o 18 July) and a subsu ace chl amaximum de eloped a 20 m dep h ( he nu icline). DON (Fig. 2g) accumula- ion in he su ace laye ook place a he end o he 32 –10 –20 –30 –2 –2 2 2 0 2 4 4 0 0 0 0 d –40 –30 –20 –10 35.6 35.6 35.5 e –40 –30 –20 –10 15/7 18/7 22/7 26/7 12 12 14 16 –40 –30 –20 –10 6 4 g –40 –30 –20 –10 15/7 18/7 22/7 26/7 2 4 6 10 12 8 i –40 –30 –20 –10 2 4 4 h –40 –30 –20 –10 2 4 6 j –40 –30 –20 –10 2 4 8 10 6 k –40 –30 –20 –10 15/7 18/7 22/7 26/7 0.1 0.2 0.3 l 16/7 18/7 20/7 22/7 24/7 26/7 Q -P-E Q P E –2 0 2 4 6 8 10 12 14 b H 16/7 18/7 20/7 22/7 24/7 26/7 0 –200 –100 100 200 300 400 c Dep h (m) 16/7 18/7 20/7 22/7 24/7 –12 –6 0 6 12 m/s a Fig. 2. July 2002. (a) Time cou se o shel winds, in m s–1; (b) con inen al uno (Q ), p ecipi a ion (P) and e apo a ion (E), in m3s–1; (c) hea balance (H), in cal cm–2 d–1; (d) esidual cu en s, in cm s–1; dashed line ep esen s he pycnocline; (e) salini y; ( ) empe a- u e, in °C; (g) dissol ed o ganic ni ogen, in µmol l–1; (h) ammonium, in µmol kg–1; (i) chl a, in µg l–1; (j) pa icula e o ganic ni o- gen, in µmol l–1; (k) ni a e, in µmol kg–1; and (l) ni i e, in µmol kg–1. Hea and eshwa e luxes aken om Pied acoba e al. (2005) Pied acoba e al.: Physical–biological coupling in Ría de Vigo. I summe elaxa ion, eaching concen a ions >7 µmol kg–1. Du ing he spin-up o weak summe upwelling (22 o 26 July), nu ien - ich wa e en e ed h ough he bo om laye s, bu his inpu was weake compa ed wi h he i s upwelling pe iod (bo om NO3–< 11 µmol kg–1). In con as wi h July, N-nu ien s le els we e educed by hal du ing Sep embe and NH4+(Fig. 3h) was he mos abundan o m, ep esen ing 72% o NT. Du ing he au umn downwelling (17 o 19 Sep embe ), nu ien -poo wa e s en e ed om he shel . Concomi- an ly, lowe PON (~4 µmol kg–1) and chl a(~4 µg kg–1) concen a ions we e measu ed a he su ace laye compa ed o July (Fig. 3i,j). DON was highe han PON du ing his pe iod and p esen ed a homogeneous dis- ibu ion h oughou he wa e column (Fig. 3j). A he beginning o he au umn ansi ion (19 o 23 Sep em- be ), nu ien s le els we e s ill e y low in he su ace laye (NH4+< 1.6 µmol l–1 and NO3–< 0.5 µmol l–1), bu he wind change caused a p og essi e en y o ENACW in he bo om laye . The le els o NO3–and NH4+(Fig. 3h,k) inc eased especially a he beginning o he au umn upwelling (23 o 26 Sep embe ), al hough he concen a ion o NO3–a he bo om was 5 imes lowe han he le els measu ed du ing he weakes summe upwelling (22 o 26 July). As a conse- quence o inc easing nu ien le els, high su ace chl a (>8 µg l–1) and PON (>7 µmol kg–1) concen a ions we e measu ed (Fig. 3i,j). As o July, he DON maxi- mum (Fig. 3g) was delayed 3 o 4 d compa ed wi h he chl aand PON maxima. In July, no signi ican di e ences in ei he nu ien s ocks o luxes we e obse ed be ween he 2 bo om sec o s. By con as , signi ican di e ences (p < 0.05) in nu ien luxes be ween he 3 su ace sec o s we e obse ed on mos sampling da es (Table 2). DON and PON s ocks we e e y homogeneous h ough he su - 33 1 1 2 –40 –30 –20 –10 k 0.4 0.2 0.3 0.3 6 7 7 9 8 –40 –30 –20 –10 g 2 3 4 6 7 –40 –30 –20 –10 j 17 17 14 16 16 15 –40 –30 –20 –10 17/9 19/9 23/9 26/9 35 35.2 34.8 35.4 –40 –30 –20 –10 e 2 2 4 6 8 –40 –30 –20 –10 17/9 19/9 23/9 26/9 i 2 6 4 –40 –30 –20 –10 h 2 –6 –2 2 4 4 –4 0 0 –10 –20 –30 d 16/9 18/9 20/9 22/9 24/9 26/9 H 0 –200 –100 100 200 300 400 –40 –30 –20 –10 17/9 19/9 23/9 26/9 l 16/9 18/9 20/9 22/9 24/9 26/9 Q -P-E Q P E –2 0 2 4 6 8 10 12 14 b c 17/9 19/9 21/9 23/9 25/9 –12 –6 0 6 12 m/s a Dep h (m) Fig. 3. Sep embe 2002. (a) Time e olu ion o shel winds, in m s–1; (b) con inen al uno (Q ), p ecipi a ion (P) and e apo a ion (E), in m3s–1; (c) hea balance (H), in cal cm–2 d–1; (d) esidual cu en s, in cm s–1 (dashed line: pycnocline); (e) salini y; ( ) empe a u e, in °C; (g) dissol ed o ganic ni ogen, in µmol l–1; (h) ammonium, in µmol kg–1; (i) chl a, in µg l–1; (j) pa icula e o ganic ni ogen, in µmol l–1; (k) ni a e, in µmol kg–1; and (l) ni i e, in µmol kg–1. Hea and eshwa e luxes aken om Pied acoba e al. (2005) Ma Ecol P og Se 353: 27–40, 2008 ace and bo om laye s, whe eas DON and PON luxes equen ly exhibi ed signi ican di e ences (p < 0.05) be ween he 3 su ace sec o s, especially on 26 July. In Sep embe , mos o he concen a ion di e ences along he ans e se sec ion occu ed in o ganic ni o- gen. As in July, he luxes o ino ganic and o ganic species we e signi ican ly di e en be ween sec o s (p < 0.05), especially on 17 Sep embe (Table 3). Ni ogen budge A budge o o – i (in mg N m–2 d–1) o NT, PON and DON was es ima ed conside ing he luxes pe m2o ee su ace a ea om he middle sec ion o he inne eaches o he ía (Table 4). The la e al a iabili y epo ed in bo h he N species concen a ions and he longi udinal componen o he esidual cu en was conside ed in he calcula ion o N-nu ien s, PON and DON luxes and, he e o e, in he co esponding i – o balances. In addi ion, he accumula ion e ms o all ni ogen species we e also calcula ed (V · ΔNT兾Δ ), assuming ha he concen a ions o NT, PON and DON in he olume o he ía bounded by he middle sec ion do no di e om he concen a ions in he middle sec- ion. NEP was ob ained by summing up o – i and V·ΔNT兾Δ (see Eq. 10) o NT(NEPT), DON (NEPD) and PON (NEPP) be ween 2 consecu i e su eys. Du ing he s ong summe upwelling, he olume o he ía bounded by he middle sec ion apped –289 ± 17 mg N m–2 d–1, i.e. 62% o he NTen y, bu only 44% o –289 ± 17 mg N m–2 d–1 we e ans o med in o o ganic ma e (NEPT= –127 ± 37 mg N m–2 d–1), while 162 mg N m–2 d–1 accumula ed in o he olume. By con as , he ía was a nu ien sou ce du ing he sub- sequen upwelling elaxa ion (o – i = 101 ± 5 mg N m–2 d–1) a he expense o he nu ien s accumula ed du ing he p e ious pe iod. Du ing he ollowing weak up- welling episode, nu ien sal s we e apped again, wi h 57% u ilized o o ganic ma e p oduc ion and 43% accumula ed in o he s udy olume. Du ing he au umn downwelling he ía ac ed as a nu ien sou ce (o – i = 119 ± 9 mg N m–2 d–1) a he expense o he nu ien s p e iously accumula ed in he sys em, whe eas NEPTwas balanced (–18 ± 12 mg N m–2 d–1). Howe e , du ing he subsequen elaxa ion pe iod, he ía showed a he e o ophic me abolism (NEPT> 0), expo ing 28% o he egene a ed NT o he shel (o – i = 19 ± 1) and accumula ing he emaining 72% in o he ía. Finally, du ing he weak au umn upwelling, he ía ac ed as a nu ien ap, –96 ± 8 mg N m–2 d–1, one hi d lowe han du ing he weak summe upwelling. Only 27% o NTwas consumed o p oduce o ganic ni ogen and he emaining 73% accumula ed in o he sys em. As a esul , he sys em changed om he e o ophic o balanced condi ions (NEPT~ 0). Du ing July, he p opo ion o NEPTdue o NO3–and NH4+was 67% and 33% espec i ely, whe eas du ing Sep embe , he alues we e 45% and 55% espec i ely. The ac ion o NEPT ans e ed o he sedimen s (NEPSED) can be in e ed indi ec ly om he ollowing mass balance equa ion: NEPT+ NEPD+ NEPP+ NEPSED = 0 ± δ(12) NEPSED is he balance be ween PON sedimen a ion and PON esuspension + NTdi usion om he sedi- men (Ál a ez-Salgado e al. 1996a). I should be no ed ha NEPSED is calcula ed wi h a la ge unce ain y (±δ), esul ing om he indi idual unce ain ies associa ed wi h he calcula ion o NEPT, NEPDand NEPP. Du ing he s ong and weak upwelling pe iods, a la ge ac- 34 1.7±0.1 –0 –0. 6± 6±0.5 –0 –0. 1± 1±0.3 0. 1± 1±0.3 0. 0± 0±0.3 0 .4±0.1 .4±0.1 –0.6±0.5 –3.0±0.1 17/09/02 –0.6±0.2 0.2±0.0 0.6±0.2 0.5±0.1 0.1±0.0 3.8±0.1 0. 8± 8±0.3 0.8±0.3 2.2±0.4 23/09/02 –0.8±0.2 –0.7±0.2 –0.7±0.2 –0.7±0.2 –0.5±0.1 2.1±0.1 0 .0±0.3 .0±0.3 0.0±0.3 1.2±0.2 26/09/02 –1.1±0.3 –1.0±0.3 0.3±0.1 0.3±0.1 0.5±0.1 19/09/02 1.1±1.0 1.8±0.8 4.1±1.2 –4.8±0.9 –4.4±0.7 5.3±0.2 0.5±0.6 –3.7±0.1 SNSN 1.7±1.3 –2.9±1.5 –7.0±1.6 2.9±0.4 3.4±0.3 18/07/02 0.7±0.1 –0.1±0.3 0.0±0.5 22/07/02 9.7±1.6 9.3±0.5 7.3±1.8 8.4±0.9 6.6±0.6 1.5±0.1 0.1±0.3 –1.0±0.1 8.2±0.2 0.0±0.3 3.7±0.5 15/07/02 –4.5±1.4 –4.0±1.2 –3.7±1.1 2.3±0.7 1.7±0.5 2.4±0.1 0.4±0.1 4.6±1.4 26/07/02 –1.3±0.4 –2.1±0.6 –3.0±0.9 –0.5±0.1 2.1±0.6 SNSN ab cd e gh Fig. 4. Time cou se o ho izon al (km d–1) and e ical (m d–1) eloci ies o each sampling da e (dd/mm/yy). Solid ci cle, ou low (V < 0 o wes wa ds) and c ossed ci cle, in low (V > 0 o eas wa ds) in o he ía; solid a ow, e ical ad ec i e e- loci y (upwa d, >0; downwa d, <0); zigzag a ow, e ical mixing eloci y (KZ兾Δz); whi e a ow, e ical displacemen o he LNM. Sec o s a e desc ibed in Fig. 1b Pied acoba e al.: Physical–biological coupling in Ría de Vigo. I ion o NEPTsank o he bo om (75 ± 35% on 15 o 18 July and 97 ± 27% on 22 o 26 July). Du ing he au umn downwelling, he ne sedimen a ion (NEPSED) es ima ed wi h Eq. (12) is balanced by he loss o he DON p e iously accumula ed in he sys em, since NEPTwas nea ly ze o. T ans o ma ion o DON in o sinking pa icles implies he use o DON as a nu ien sou ce o mic ohe e o ophs (mainly bac e ioplank- on) and i s ans e ence o highe ophic le els (zoo- plank on, hanging mussels) o become sinking pa i- cles. The au umn ansi ion was he only pe iod when NEPSED was <0, which means ha esuspension o PON and/o di usion o NT om he sedimen s was la ge han POM deposi ion o e he bo om. The ela i e impo ance o accumula ion e sus ex- po o DON and PON was also assessed in each pe iod. 35 Da e Sec o s Concen a ions Fluxes NTDON PON NO3–NO2–NH4+NTDON PON NO3–NO2–NH4+ 15 July SS-CS **ns ns *ns *ns ns *ns * CS-NS ***ns ns ***ns *ns * SS-NS **ns ns ** **ns ns ns * NB-SB *ns ns ns ns ns ns ns ns ns ns ns 18 July SS-CS *ns **ns * ****** CS-NS *ns ns *** ****** SS-NS *ns **** ****** NB-SB nsnsnsnsnsns nsnsnsnsnsns 22 July SS-CS *ns ns *ns * ****** CS-NS *ns ns *ns **ns ns *** SS-NS ns ns ns ns ns ns ns **ns ns ns NB-SB *ns ns ns ns ns ****ns * 26 July SS-CS ns ns ns *** ***ns ns * CS-NS *ns ns ns ** ****** SS-NS *ns ns ns ** ****** NB-SB nsnsnsnsnsns ****** Table 2. - es o he concen a ions and ho izon al luxes o NT, DON, PON, NO3 –, NO2–and NH4+be ween he 3 su ace (NS, CS and SS) and he 2 bo om (NB and SB) sec o s, o each sampling da e in July 2002. The p-le el was calcula ed as ollows: i > 2.015, he di e ences be ween 2 sec o s a e signi ican a p < 0.05 (indica ed by *), whe eas i < 2.015, he di e ence is nonsigni ican (indica ed by ns). x _,σ2 Xand y _,σ2 Y he mean and a iance o he concen a ions o luxes in 2 sec o s espec i ely xy X 2Y 2 − () +σσxy X 2Y 2 − () +σσ Da e Sec o s Concen a ions Fluxes NTDON PON NO3–NO2–NH4+NTDON PON NO3–NO2–NH4+ 17 Sep SS-CS ***ns *ns ****** CS-NS ***ns ns ns ****** SS-NS *ns ns ns ns ns ****** NB-SB ns *ns ns ns ns ****** 19 Sep SS-CS ns **ns ns ns ***ns ** CS-NS ns *ns ns ns ns ns ns ns *ns ns SS-NS ns **ns *ns ****** NB-SB ns **ns ns ns ns ns **ns ns 23 Sep SS-CS ns ns ns ns ns ns ns ns ns ns ns ns CS-NS nsnsnsns *ns ns ns ns *ns * SS-NS ns ns ns ns ns ns *ns ns *ns * NB-SB *ns *ns ns ns *ns ns *** 26 Sep SS-CS ns **ns ns ns ****** CS-NS ns **ns ns ns ns ns ns ns ns ns SS-NS ns *ns ns ns ns ****** NB-SB nsnsnsnsnsns nsnsns **ns Table 3. - es o he concen a ions and ho izon al luxes as in Table 2, bu o Sep embe 2002