Bottle effects, toxicity and cell viability during incubation experiments to asses community respiration

Bo le e ec s, oxici y and cell iabili y du ing incuba ion expe imen s o asses communi y espi a ion. Isabel Baños Ce ón Di ec o es académicos: Ja ie A ís egui Ruíz y Ma ía Fe nanda Mon e o del Pino Ins i u o de Oceanog a ía y Cambio Global, Uni e sidad de Las Palmas de G an Cana ia, Campus de Ta i a. 35017 G an Cana ia, Spain. Tesina de Más e en Oceanog a ía Uni e sidad de Las Palmas de G an Cana ia Diciemb e 2012 1 ABSTRACT Mic obial espi a ion is a key me abolic p ocess in he ocean ca bon cycle. Howe e , so a he e is no a “gold s anda d” me hod o es ima e espi a ion in he sea. Di ec oxygen consump ion “in i o” me hods ely on long incuba ions (24h) inside glass bo les, which may enhance changes in he enclosed na u al popula ions. Recen ly, an enzyma ic in i o ETS assay has been p oposed as an al e na i e o es ima e “ac ual” espi a ion in sho e incuba ion pe iods (<6 h). He e we ha e in es iga ed wo main ac o s ha we conside may a ec incuba ions in bo h he “in i o” oxygen consump ion expe imen s and he “in i o” ETS assays: (i) The e ec o enclosing na u al popula ion in bo les o di e en olume (125, 250 and 1000 mL) du ing 24 h, and (ii) The po en ial oxic e ec o he INT (which is educed o INT-F du ing he ETS assay) o e na u al popula ions, du ing he incuba ion pe iod. We es ed he “bo le e ec ” in he whole plank onic communi y (C) as well as in he il e ed ( h ough 1.2 μm il e ) mic obial communi y (F), along h ee expe imen s. Ou esul s show a iable esponses o na u al communi ies du ing he incuba ion inside bo les, bu wi hou any signi ican di e ence be ween he bo les’ olume used. The gene al pa e n o hese changes (bo h in he o al and ac iona ed communi ies) is a mode a e dec ease in phy oplank on popula ions and a sha p inc ease in bac e ial assemblages ( a ou ing he g ow h o HNA cells), which would p esumably a ec he inal me abolic a es. On he o he hand, we es ed he oxici y o he INT du ing a ypical in i o ETS assay (6 h incuba ion, 0.2 mM INT). The esul s o his single expe imen show ha , a e he INT addi ion, he e is (i) a comple e loss o P ochlo ococcus cells, and a mode a e loss o Synechococcus and picoeuka yo es, (ii) an inc ease in HNA bac e ia (p esumably by he up ake o newly o ganic subs a es om phy oplank on dea h) bu a dec ease in LNA 2 bac e ia, and (iii) a decline in bac e ial iabili y, which is p oduced in he i s 0.5 h, supe imposed o he g ow h o HNA bac e ia. In summa y, ou esul s highligh he impo ance o moni o ing he e olu ion o mic obial communi ies along incuba ions (bo h in he “in i o” oxygen consump ion expe imen s and “in i o” ETS assays) and wa n o some impo an me hodological cons ain s ha may d as ically in luence he ac ual espi a o y a es. 3 INTRODUCTION Ma ine bac e ia a e conside ed he g ea es con ibu o s o communi y espi a ion in he oceans, by up aking dissol ed o ganic ma e and emine alizing in o ca bon dioxide (Ri kin and Leg ende 2001). This ole makes hem key playe s in ma ine ca bon luxes (Gasol e al. 2008), being ele an in he assessmen o he global ocean ca bon cycle (Azam and Mal a i 2007, A ís egui e al. 2009). Di ec measu emen s o bac e ial espi a ion a e howe e cumbe some and edious, and equi e p e ious sepa a ion o bac e ia om he es o he plank onic communi y (by means o il a ion p ocesses), as well as long in i o incuba ions (gene ally 24 h o mo e). Al hough incuba ions may las some imes se e al days (e.g. del Gio gio and Cole 1998; She e al. 1999), he esponse o he p oka yo ic assemblage incuba ed inside he bo le has been gene ally assumed o be ep esen a i e o he in si u bac e ial beha iou (Ki chman e al. 1990). Ne e heless, he e is s ill an un esol ed deba e on whe he me abolic a es ob ained by in i o incuba ions -las ing om hou s o days- ep oduce ealis ic in si u a es, no only om bac e ial assemblages bu also om he whole mic obial communi y, due o “bo le e ec s” (e.g. Se e e al. 2009, Cal o-Díaz e al. 2011). As Robinson and Williams (2005) de ined, “bo le e ec s” include hose una oidable changes ha popula ions enclosed in bo les expe ience du ing he incuba ion p ocess, a ec ing he abundance (Lee and Fuh man 1991), bac e i o y p essu e (Hopkinson e al. 1989, Ma asé e al. 1992, Suzuki 1999), me abolic ac i i ies (She e al. 1999), ino ganic and o ganic nu ien s up ake (Gasol and Mo án 1999), phylogene ic composi ion (Massana e al. 2001) and communi y size s uc u e (Fe guson 1984). Due o his, expe imen s mus be handled wi h ex emely ca e, and mic obial abundances and a es moni o ed h ough he leng h o he incuba ion, o ealis ically in e p e he p ocesses ha ake 4 place inside he bo les in o de o ex apola e o in si u condi ions (Williams e al. 2013 in p ess). Pa icula ly, in oligo ophic egions whe e nu ien s a e sca ce and small picoplank on domina e, he p esumably non-na u al al e a ion o communi y s uc u e and me abolism du ing incuba ions has ecei ed a special a en ion (e.g. Cal o-Díaz e al. 2011). Recen s udies ha e b ough back in o discussion an old scien i ic dispu e ha a ose in he 80s abou quan i ying he magni ude o p ima y p oduc ion in he oceans based on geochemical e sus ecological in i o app oaches (Shulenbe ge and Reid 1981, Pla 1984), o unde s and i s ole in he global ca bon cycle. This con o e sy has been ansla ed a p esen o he s udy o he me abolic balance ( he P oduc ion / Respi a ion a io) o plank onic communi ies in open-ocean oligo ophic egions (Ducklow and Doney 2013 in p ess). Two con as ing iews con on he hypo hesis o an au o ophic ocean (P oduc ion > Respi a ion)-mos ly based on he in e p e a ion o la ge-scale and long- e m biogeochemical es ima es (Williams e al. 2013 in p ess) e sus he hypo hesis o a he e o ophic ocean (Respi a ion > P oduc ion) -based on sho - e m oxygen-de i ed a es ob ained om in i o incuba ions (Dua e e al. 2013 in p ess). In he middle o his deba e, he e s ill emain la ge unce ain ies on whe he “in i o” app oaches ep oduce eliable es ima es o mic obial/bac e ial espi a ion due o he ex ended leng h (24h o mo e) needed o ob ain signi ican changes in oxygen consump ion. This has os e ed he sea ch o al e na i e non-incuba ion (o sho e - incuba ion imes) me hods o ob ain mic obial espi a ion ac i i ies in he open ocean. In he 1970s, Packa d and co-wo ke s de eloped an enzyma ic in i o assay (ETS i o) o es ima e po en ial espi a o y ac i i y (unde sa u a ing subs a es). Yea s la e , A ís egui and Mon e o (1995) ound a s ong global co ela ion be ween ac ual a es o 5 oxygen consump ion and ETS i o, including da a om many di e en su ace oceanic egions, spanning om eu ophic o oligo ophic ecosys ems; al hough he co ela ions we e weake a almos e e y gi en pa icula egion. The ETS i o as an es ima e o espi a o y ac i i y was c i icized by Bams ed (2000), who conside ed ha enzyma ic ac i i ies should be measu ed a na u al le els o subs a es. Recen ly, Ma ínez-Ga cía e al. (2009) de eloped an in i o assay o he elec on anspo sys em (ETS) me hod, based on he educ ion o he e azolium INT sal o INT o mazan (INT-F) by na u al mic obial communi ies du ing sho incuba ions (2-6 h), claiming i s use o expand he s ill meag e da a base on mic obial espi a ion. They es ed he me hod in a wide ange o cul u es and su ace-wa e ma ine en i onmen s, inding a s ong o e all co ela ion ( 2=0.97) be ween oxygen consump ion (using he Winkle me hod; R) and in i o INT- F o ma ion a es (ETS), yielding a su p ising R/ETS = 12.8. Ne e heless, i mus be no iced ha mos o he R a es used in hei s udy o de i e he co ela ion we e se e al imes highe (> 2µM h-1) han published R a es om open-ocean na u al communi ies (e.g. A ís egui and Mon e o 1995). Mo eo e , he ew R alues used in his s udy om oligo ophic open ocean communi ies (<2µM h-1) do no seem o co ela e he same wi h he in i o ETS, as obse ed also in mo e ecen s udies om he same au ho s (A angu en-Gassis e al. 2011). I he in i o ETS yields eliable and obus es ima es o R, he me hod would de ini i ely allow imp o ing he global da a base on R, a oiding “bo le e ec s” (Robinson and Ramaiah 2011), pa icula ly in he da k ocean, whe e he Winkle me hod is close o o unde i s limi o sensi i i y (A ís egui e al. 2005). Howe e , a ecen s udy by Maldonado e al. (2012) ques ions he alidi y o he me hod. These au ho s show ha common cellula subs ances no in ol ed in he espi a o y elec on anspo sys em (ETS) may educe he INT sal o INT-F as well, yielding misleading esul s ha –acco ding o hem- would be equi alen o he blank in 6 an in i o ETS assay (Packa d 1971) whe e NADH, NADPH and succina e a e added a sa u a ion le els. Fu he mo e, he addi ion o a e azolium sal o na u al communi ies may ha e oxic e ec s o e he cells, educing hei iabili y and hence dec easing he me abolic a es o he whole communi y. Indeed, Gasol and A ís egui (2007) obse ed ha he addi ion o he luo ogenic e azolium sal CTC a concen a ions close o 5 mM p oduced a dec ease in cell abundance o 22% in coas al picoplank on, a e 30 min o incuba ion. Ma ínez-Ga cía e al. (2009) used in hei s udy concen a ions anging om 0.2 o 1 mM o INT, 5 o 25 lowe han hose used by Gasol and A ís egui, bu a he same ime, incuba ion imes 4-12 imes longe . Thus, i should be es ed whe he hese lowe concen a ions o e azolium added o cells du ing longe incuba ion imes ha e o no a signi ican oxic e ec on mic obial popula ions. I so, independen ly o he po en ial bias due o he INT educ ion caused by subs ances no in ol ed in he ETS, he oxici y o he INT would lead o an unde es ima ion o he espi a o y ac i i y. In he p esen s udy we wan ed o es wo main ac o s ha we conside ed c i ical in espi a ion s udies: he e ec o enclosing samples inside bo les and he po en ial oxic e ec o he use o o mazan sal s o he in i o assay o espi a o y ac i i y. These ep esen he main wo objec i es o his wo k: (1) To analyze he changes wi h ime in picoplank on communi y s uc u e and abundances caused by he enclosu e o na u al popula ions in bo les o di e en olume, in o de o elucida e he selec i e e ec o “bo le size” on he g ow h and mo ali y o bac e ia, cyanobac e ia and picoeuka yo es. 7 (2) To e alua e he po en ial oxic e ec s o he e azolium sal INT du ing he leng h o ime o a ypical incuba ion o in i o ETS assay a concen a ions epo ed in he li e a u e. We ha e s a ed wo king on a hi d objec i e, bu due o iming cons ain s in he da es scheduled o de end his memo y we ha e been unable o p esen he e he esul s, which a p esen a e a hei ini ial s age. The objec i e (which will con inue add essing in he nea u u e) is he ollowing: (3) To look a he a iabili y in he R:ETS i o and R:ETS i o a ios a he s a and end o a 24 h incuba ion, used o es ima e oxygen consump ion by he Winkle me hod. Ou aim is o see how does he po en ial g ow h (o loss) o mic obial popula ions inside he incuba ion bo les a ec he inal R/ETS a ios. MATERIALS & METHODS Plank on samples o he expe imen s we e collec ed in he eas e n coas o G an Cana ia (27º 51’ 26” N, 15º 23’15” W) Spain (Fig.1). These wa e s a e cha ac e ized by a picoplank on communi y p esen ing popula ions o he e o ophic bac e ia, P ochlo ococcus, Synechococcus and picoeuka yo es, ypical o oligo ophic en i onmen s (A ís egui and Mon e o 2005, Bal a e al. 2009). All he samples we e collec ed in 10-li e polyca bona e ca boys, a he same ime o he day, om 0 o 1m dep h. In all he expe imen s, wa e was il e ed wi hin 1-2h a e collec ion in he labo a o y. To sepa a e he bac e ial communi y om he o al communi y (conside ed i as he wa e p e- il e ed by 100µm mesh), we employed a Polyga d-CN ca idge il e (1.2µm inal po e size). These il e s yield a high pe o mance wi h a minimum 8 di e en ial p essu e, minimizing cells’ b eakage. Ne e heless, we ealized ha many o he picophy oplank on o ms also passed h ough hese il e s. Be o e each expe imen , all he consumable ma e ial and he bo les (glasswa e and Te lon-coa ed caps) we e cleaned in acid (3% inal concen a ion) o emo e any o ganic ma e which could in luence he mic obial communi y g ow h. The ca idge il e s we e disin ec ed du ing 30minu es in wa e a 80º C and subsequen ly au ocla ed a 126º C du ing 30 minu es. Only he oxici y assay wi h INT was pe o med in a ba h wa e a 18º C. Fo a leas 0.5-1 h, he bo les illed wi h seawa e we e empe a e in he ba h un il he desi ed empe a u e was eached. 1. Expe imen al design. 1.1. Volume ic bo le e ec . To assess he olume ic bo le e ec we conduc ed h ee se o expe imen s: Exp 1(July 2012), Exp 2 and Exp 3(Sep embe 2012). Each o he samples ( o bo h he whole plank onic communi y and he 1.2μm il e ed communi y) was incuba ed in h ee di e en olumes (125mL, 250mL and 1000mL) a in si u empe a u e (22º C) and in da k condi ions. A o al o wel e bo les we e employed in Exp 1 and eigh een in Exp 2 and Exp 3. Subsamples o To al O ganic Ca bon (TOC) and picoplank on abundances we e ob ained a =0 h (conside ed as ini ial) and a e 8, 12, 24 hou s o incuba ion, excep in Exp 1 whe e he 12 h measu emen was no conside ed. Bac e ial iabili y was also measu ed bu he esul s a =24 h a e no p esen ed he e. The high bac e ial abundance a his ime caused ha mo e han 1000 e en s s-1 we e eached when he sample was coun ed by cy ome y. The dilu ion equi ed by he sample caused an osmo ic shock, and hence an o e es ima ion o he damaged cells. 15 P ochlo ococcus, he e we e di e ences in he ini ial concen a ions o Synechococcus, wi h >6000 cells mL-1 in Exp 2 and <2000 cells mL-1 in Exp 3 (Figs. 3B-C). Picoeuka yo es p esen ed an almos s able concen a ion in he C samples du ing he i s 12h, gi ing ise o a sha p decline a 24h (Figs 4A-C). This a iabili y was no e iden in he F samples (excep a T0-8 in Exp 1), whe e cell concen a ions emain low bu p e y s able. Nanophy oplank on cells (>2μm) (Figs. 5A-C) did no pass he 1.2 μm il e , bu popula ions in he C sample d opped mo e sha ply han o he phy oplank on g oups in he i s 8 hou s o incuba ion. Al hough nanophy oplank on p esen ed concen a ions (300-500 cells mL-1) much lowe han he picoplank on g oups, hei la ge biomasses make hem key playe s in he communi y me abolism. Thus, he decline wi h ime o hei abundances down o hal o less he ini ial concen a ions a e 24 h would ep esen a s ong handicap when es ima ing me abolic a es. The e a e e y ew s udies looking a he ac o s ha p oduce he decline in plank onic communi ies enclosed in incuba ion bo les. Some o hese s udies indica e ha Synechococcus esponds wi h a g ea e esis ance o he manipula ion and o a d as ic change in he ligh egime o da kness han P ochlo ococcus (Binde and Chisholm 1995, Jacque e al. 1998), while P ochlo ococcus appea ed o be mo e in luenced by a highe g azing p essu e (Guillou e al. 2001). Ins ead, picoeuka yo es and nanophy oplank on, seem o be mo e sensible o con inemen han cyanobac e ia due o a highe nu ien eques (Ra en 1998, Veldhuis e al. 2005).. In any case, i is likely ha he exhaus ion o ino ganic nu ien s and he cascading ophic e ec s (by enhanced g azing and dec easing g ow h a es) p oduce changes in all popula ions o phy oplank on, la gely a ec ing o he mos sensible g oups (phy oplank on >2 μm). 16 1.2. Time e olu ion o he bac e ial assemblages in he “ o al mic obial communi y” (C) and he “1.2 μm il e ed” (F) samples. In con as o he ecu en decline wi h ime in phy oplank on popula ions, bac e ia show a clea opposi e end wi h a sha p enhancemen , which is mo e acu e a e he i s 12 h. In all he h ee expe imen s, he C samples espond mo e apid han he F samples o bac e ial inc ease (Fig. 6). Indeed, a e he i s 8 h an almos exponen ial g ow h o bac e ial cells can be obse ed in he h ee C expe imen s, eaching alues om 4 o 10 imes mo e a e 24 h (Fig. 6). The F samples, howe e , display a lag- phase ( om 8 o 12 h) be o e g owing exponen ially o simila o e en highe concen a ions. This lag-phase is cha ac e is ic o new cul u es ha need some ime o adap be o e s a g owing. The imp ession is ha while he bac e ial abundances in he C samples seem o end o s abilize a 24 h, a leas in Exp s 2 and 3 (Figs. 6B-C), in all h ee expe imen s he F samples look like s a ing he exponen ial g ow h phase a 24 h (p obably eaching much highe concen a ions in he ollowing hou s) By means o low cy ome y, we can iden i y wo b oad assemblages o bac e ia ha Li e al. (1995) desc ibed as low-nucleic acid con en (LNA) and high-nucleic acid con en (HNA). In ac hese wo ca ego ies comp ise a complex assemblage o g oups: one wi h less luo escence and small size (LNA) and he o he wi h highe luo escence and la ge sizes (HNA). LNA uses o be domina ed by he small SAR11clade g oup, whe eas he HNA assemblage is mo e complex, and may con ain la ge cells om he Rhodobac e ales, Bac e oide es and Alphap o eobac e ia g oup (Bal a e al. 2012, Vila-Cos a e al. 2012). Al hough bo h assemblages (HNA and LNA) a e ac i e, as hey may exhibi simila a es o leucine inco po a ion (Zubko e al. 2001), HNA a e conside ed he “ eady o espond” ac ion (Gasol e al. 1999) ou c oping LNA bac e ia 17 unde a ou able condi ions (e.g. excess o ganic ma e ). This si ua ion is e iden along he h ee expe imen s and he wo ea men s (C and F), showing a clea inc easing end in he HNA/LNA a io wi h ime (Fig. 7). Howe e , he HNA/LNA a io inc eases mo e in he F ea men compa ed o he C sample (Fig. 7), p obably because he a ailabili y o o ganic ca bon is e en highe . Indeed, he o al o ganic ca bon (TOC) da a measu ed a he ini ial incuba ion ime inside he bo les (Tables 2-4) show ha he TOC alue in he F sample is consis en ly highe han in he un il e ed sample (C). I is impossible o asc ibe he amoun o semi-labile o semi- e ac o y o ganic ca bon om TOC a ailable o p oka yo es, bu he la ge di e ences in TOC be ween F and C samples (a almos all he sampling imes) sugges ha il a ion p oduces o ganic deb is and dissol ed ca bon ha migh be used in pa by bac e ial cells o g ow up. Mo eo e , he a iabili y in TOC a each sampling pe iod and ea men may esul om he balance be ween he usage o bio-a ailable TOC by p oka yo es and he p oduc ion o TOC by he dea h and exc e ion o mic oo ganisms (Ca lson e al. 1998) In any case, o ha e a mo e p ecise es ima e o he TOC a ailable o p oka yo es we i s should sub ac he biomass o picoplank on om TOC a each sampling ime, an exe cise ha we will do in he nea u u e. A c i ical issue (no mally igno ed in me abolic s udies) is o moni o he abundance o ac i e ( s inac i e) cells, which would mos ly con ibu e o he o e all mic obial me abolism. Some su p ising esul s indica e ha al hough he e a e changes in bac e ial assemblages inside he bo les du ing long- e m incuba ions (24h o mo e), cons an communi y me abolic a es can be main ained as a esul o shi s in communi y composi ion (e.g. Bal a e al. 2012). Whe he hese shi s e lec a iabili y in he dea h/g ow h o ce ain speci ic bac e ial g oups is unclea . He e we ha e moni o ed he 18 iabili y o bac e ial cells along he expe imen s, looking a hei memb ane in eg i y (NADS me hod), and ound consis en esponses (Tables 2-4). The ini ial pe cen age o li e cells (NADS+, wi h an in eg al memb ane) in Exp s 2 and 3 was e y simila (91% in he F ea men and 80% in C ea men ), keeping also simila ends in bo h expe imen s a T12: inc easing in C and dec easing in F. Ne e heless, he cell iabili y in he F ea men s showed a mo e p onounced dec ease in Exp 3 han in Exp 2. This a iabili y migh e lec changes also in he shi o bac e ial assemblages, bu a e c i ical o moni o du ing me abolic expe imen s, since p obably will gi e he clue o a possible ela ionship be ween ac i e bac e ia abundance and me abolic a es. 1.3. Time e olu ion o he Au o ophic: He e o ophic biomasses a io (AB:HB) in he “ o al mic obial communi y” (C) and he “1.2 μm il e ed” (F) samples. We con e ed he plank onic cell abundances o biomasses (see me hods) in o de o es ima e he a iabili y in he AB:HB. This a io p o ides a balance be ween he con ibu ion o au o ophic phy oplank on and he e o ophic mic obes o he me abolic a es o a gi en ecosys em (Odum 1971, A ís egui and Mon e o 2005). Al hough would be somewha naï e o equa e a biomass a io o a me abolic p oxy (see o ins ance he discussion abo e abou he ac i e s inac i e bac e ia), he AB:HB gi es a b oad pic u e o he shi in he me abolic s a e o he ecosys em along he expe imen . Figu e 8 shows a clea dec easing end o he AB:HB a io along he expe imen , ending o nea 0 a T24. A T0 and T8 he a io is close o 1 (some imes e en highe ), sugges ing a me abolic balance be ween au o ophy and he e o ophy. As soon as phy oplank on popula ions decline and HNA bac e ia g ow exponen ially, he a io u ns ou o a s ic he e o ophy. Simila esul s we e ecen ly published by Cal o-Diaz e al. (2011), du ing a se o expe imen s ca ied ou in he No h A lan ic sub opical Gy e. I 19 emains o know wha is he ue con ibu ion o he inc easing bac e ia popula ion (and concomi an dec easing AB:HB a io) o he bias in ac ual espi a ion a es. This opic is pa o ou u u e wo k and needs o be add essed h ough sho - e m espi a ion expe imen s wi h na u al communi ies. O e all, ou esul s e idence di e en esponses o na u al communi ies when hey a e enclosed and incuba ed inside bo les o di e en olume. Ne e heless, he gene al pa e n o hese changes (bo h in he o al and ac iona ed communi ies), and wi hou any signi ican di e ence be ween olume bo les < 1000 mL, is a mode a e dec ease in phy oplank on popula ions and a sha p inc ease in bac e ial assemblages ( a ou ing he g ow h o HNA cells). I is, howe e , necessa y o moni o he iable bac e ial – ma ching oxygen consump ion a es- o see whe he o no hese cells a e esponsible o he bulk o espi a o y a es. 2. Responses o picophy oplank on and he e o ophic bac e ia o INT oxici y du ing an “in i o ETS” assay. The esponse o picoplank onic (au o ophic and he e o ophic) popula ions o he addi ion o INT in concen a ions equi alen o hose used in “in i o ETS assays” (Ma ínez-Ga cía e al. 2009) a ied depending on he g oups. The mos s iking esponse was ha o P ochlo ococcus. A e adding INT o he sample, i s popula ions (wi h a concen a ion close o 6 x 104 cells mL-1) comple ely anished a e he i s 0.5 h, while in he con ol sample he popula ions emained almos s able along he 6 hou s o incuba ion (Fig. 9A). Synechococcus abundance expe ienced an a e age dec ease o 9-12% in he INT sample, wi h espec o hei ini ial abundance. The dec ease was mo e e iden a e 1 h o incuba ion (Fig. 9B). Picoeuka yo es esponded simila o Synechococcus , wi h a d op in hei abundances o 11-20% along he incuba ion in he 20 INT samples (Fig. 9C). I is no clea why P ochlo ococcus is so sensi i e o he oxic e ec o INT (compa ed wi h Synechococcus and picoeuka yo es). Pe haps i s pe ec adap a ion o oligo ophic en i onmen s, wi h a d as ic educ ion o cell and genome size (Pa ensky and Ga cza ek 2010) - also known as “The Black Queen Hypo hesis” (Mo is e al. 2012)- as well as i s dependence on co-occu ing mic oo ganisms in c oss-p o ec ing om oxida i e damage (Mo is e al. 2011), makes P ochlo ococcus a a he suscep ible o ganism unde ex e nal damage, albei i s pe ec adap a ion o na u al condi ions. P ochlo ococcus is he mos abundan pho osyn he ic o ganism in oligo ophic open- ocean en i onmen s (wi h abundances close o he e o ophic bac e ia), and hence mus play a key ole in he o e all mic obial me abolism (e.g. espi a ion). Ou esul s hus ale ha he addi ion o INT o assessing “in i o” ETS ac i i y in na u al samples (wi h he consequen loss o all P ochlo ococcus cells, and o a lesse ex en o Synechococcus and picoeuka yo es) migh clea ly unde es ima e he ac ual espi a o y ac i i y o he mic obial communi y. The esponse o he e o ophic bac e ia o he addi ion o INT was di e en o phy oplank on, bu a iable among assemblages (Fig.9 D-F). HNA bac e ia esponded inc easing hei abundances in he INT samples (wi h espec o he con ol sample), whe eas he LNA bac e ia sligh ly dec eased in abundance. HNA bac e ia inc eased 23% in abundance in he i s 0.5 h a e he INT addi ion, being 11-14% mo e abundan along he expe imen , wi h espec o he con ol. The ini ial inc ease o HNA could ha e bene i ed o he dea h (and dissol ed o ganic leaking) o picophy oplank on. LNA bac e ia (p esumably less oppo unis ic han he HNA bac e ia) didn’ seem howe e o ake bene i o he po en ial inc ease in o ganic ca bon subs a es, 21 dec easing in hei abundances 6-10% a e INT addi ion. O e all, he balance is a sligh inc ease in o al bac e ial along he 6 h o he incuba ion expe imen (al hough non signi ican , excep in he i s 0.5 h). Ne e heless, i we look a he iabili y o bac e ial cells a e he INT addi ion (Fig. 10) we obse e ha jus a e he i s 0.5 h he e is a signi ican 20% decline in he % o iable bac e ia cells wi h espec o he con ol, which emains cons an along he 6 h. Al hough we ha en’ had ime o sea ch mo e on he cy ome ic da a, we will be able o disce n whe he he d op in he iabili y is mo e due o HNA o LNA bac e ia. Gasol and A ís egui (2007) ound a s ong co ela ion be ween he inc ease in he o ma ion o CTC-F (ano he o mazan sal de i a i e o he educ ion o he luo ogenic e azolium dye CTC) and he disappea ance o HNA bac e ia. In hei expe imen s -pe o med wi h coas al communi ies simila o hose used in ou expe imen s- hey obse ed ha he CTC-F g anules o med inside he cells became so la ge ha las ed b eaking and killing he cells. In ou case, we did no ind any co ela ion be ween he INT-F o ma ion and he disappea ance o bac e ia cells (nei he HNA o LNA; Fig. 9F), no wi h he % o iable cells. Pe haps, he INT-F c ys als o med inside he cells we e no la ge enough as o b eak he cells, bu wo o he easons may also explain his lack o co ela ion: (i) The INT concen a ion was low enough as o a ec only o a small p opo ion o he o al cells, and hence he co ela ion is no e iden (which in ha case would ques ion he use o 0.2 mM o INT as a sa u a ing le el du ing in i o ETS assays). (ii) The INT is mos ly educed o INT- F by subs a es no in ol ed in he ETS eac ions, as Maldonado e al. (2012) a gued. In summa y, ou esul s sugges ha he addi ion o INT o he na u al samples p oduce he ollowing e ec s (mos ly in he i s 0.5 h): i) A comple e loss o P ochlo ococcus 22 cells, and a mode a e loss o Synechococcus and picoeuka yo es, wi h a leak o dissol ed o ganic subs a es ha could be used by bac e ia. ii) An inc ease in HNA bac e ia (p esumably by he up ake o newly o ganic subs a es om phy oplank on dea h), bu a dec ease in LNA bac e ia. iii) A decline in bac e ial iabili y, which is p oduced in he i s 0.5 h, supe imposed o he g ow h o HNA bac e ia. O e all, he e ec o adding 0.2mM o INT o a na u al sample du ing an in i o ETS assay would lead o an unde es ima ion o he espi a o y ac i i y, al hough i emains o quan i y i s magni ude. This will de ini i ely in luence also he a iabili y in he R/ETS a ios used o con e ETS ac i i y o ac ual espi a ion a es. CONCLUSIONS AND FUTURE WORK Ou esul s may be summa ized in he ollowing main conclusions: 1. The enclosu e o na u al popula ions inside BOD bo les du ing “oxygen consump ion expe imen s” (along 24 h), p oduce a mode a e (a e age: 20%) dec ease in phy oplank on popula ions, which is mo e e iden in picoeuka yo es and nanophy oplank on. 2. He e o ophic bac e ia, and pa icula ly he HNA assemblages, inc ease up o 10 imes mo e, p obably a ou ed by he enhancemen in o ganic subs a es a ailable due o phy oplank on mo ali y. 3. P e- il a ion o na u al samples, o achie e bac e ial espi a ion, leads o an enhancemen o bac e ial g ow h, which is mo e e iden a e a lag phase o abou 12 h. 23 4. The au o ophic/he e o ophic biomass a io shi s along he expe imen s, om alues >1 o close o 0, due o he mode a e dec ease in phy oplank on biomass and he sha p inc ease in bac e ial biomass. 5. We did no ind any signi ican a iabili y in he e olu ion o mic obial communi ies (nei he in he whole communi y no in he ac iona ed one) ac oss he di e en bo les used (wi h olumes anging om 125 mL o 1 li e). Ne e heless, we canno disclaim ha la ge olumes may educe “bo le e ec s”. 6. The addi ion o INT (a 0.2mM inal concen a ion) du ing an “in i o” ETS assay p oduced he mo ali y o all P ochlo ococcus cells du ing he i s 0.5 h, as well as a mode a e loss o Synechococcus and picoeuka yo es. 7. He e o ophic bac e ia abundances, howe e , did no show a clea esponse o INT addi ions, al hough bac e ial iabili y d opped a 20% in he i s 0.5 h. All he abo e esul s ale o he impo ance o moni o ing he e olu ion o mic obial popula ions du ing incuba ion expe imen s o add ess espi a o y a es, and open new esea ch ques ions ha we will add ess in he nea u u e: 1. Does la ge olumes o wa e minimize he “bo le e ec s”?. 2. Does p e- il a ion inc ease me abolic a es as seems o do wi h bac e ial abundances in >24 h expe imen s?. 3. Why P ochlo ococcus is mo e sensi i e han o he picophy oplank on o he oxici y o INT, and how his a ec o he inal espi a o y a es in he “in i o” ETS assays?. 4. Why HNA bac e ia inc ease du ing he in i o ETS assay? 24 5. A e he cell memb anes o bac e ia less pe meable han hose o phy oplank on cells o INT?. 6. Wha ac o s a e he main esponsible in he a iabili y in he R/ETS i o and R/ETS i o a ios? 7. Wha a e he main cons ains o add ess espi a o y a es du ing incuba ion expe imen s?. REFERENCES A angu en-Gassis M, Se e P, Fe nández E, He e a JL, Domínguez JF, Valesca P, Escanez J (2011) P oduc ion and espi a ion con ol he ma ine mic obial me abolic balance in he eas e n No h A lan ic sub opical gy e. Deep-Sea Res I 58:768–75 A is egui J, Mon e o MF (1995) The ela ionship be ween communi y espi a ion a e and ETS ac i i y in he ocean. J Plank on Res 17:1563-1571 A ís egui J, Agus í S, Middelbu g JJ, Dua e CM (2005) Respi a ion in he mesopelagic and ba hypelagic zones o he oceans. In: del Gio gio PA, Williams PJLeB (eds) Respi a ion in Aqua ic Ecosys ems. Ox o d Uni e si y P ess, New Yo k, p 147- 180 A ís egui J, Mon e o MF (2005) Tempo al and spa ial changes in plank on espi a ion and biomass in he Cana y Islands egion: he e ec o mesoscale a iabili y. J Ma Sys 54:65-82 A ís egui J, Gasol JM, Dua e CM, He ndl GJ (2009) Mic obial oceanog aphy o he da k ocean’s pelagic ealm. Limnol Oceanog 54:1501-1529 31 Figu e 2. Changes in he abundance o P ochlo ococcus du ing olume ic bo le e ec expe imen s in Exp 1 (A), Exp 2 (B) and Exp 3 (C). Con inuous line shows he whole plank onic communi y and do ed line he il e ed communi y. No e he di e en scales on he y axes. 32 Figu e 3. Changes in he abundance o Synechococcus du ing olume ic bo le e ec expe imen s in Exp 1 (A), Exp 2 (B) and Exp 3 (C). Con inuous line shows he whole plank onic communi y and do ed line he il e ed bac e ial communi y. No e he di e en scales on he y axes. 33 Figu e 4. Changes in he abundance o picoeuka yo es du ing olume ic bo le e ec expe imen s in Exp 1 (A), Exp 2 (B) and Exp 3 (C). Con inuous line shows he whole plank onic communi y and do ed line he il e ed bac e ial communi y. No e he di e en scales on he y axes. 34 Figu e 5. Changes in he abundance o nanoplank on du ing olume ic bo le e ec expe imen s in Exp 1 (A), Exp 2 (B) and Exp 3 (C). Con inuous line shows he whole plank onic communi y and do ed line he il e ed bac e ial communi y. No e he di e en scales on he y axes. 35 Figu e 6. Changes in he abundance o bac e ia du ing olume ic bo le e ec expe imen s in Exp 1 (A), Exp 2 (B) and Exp 3 (C). Con inuous line shows he whole plank onic communi y and do ed line he il e ed bac e ial communi y. No e he di e en scales on he y axes. 36 Figu e 7.Tempo al e olu ion o he High DNA and Low DNA bac e ia a io (HNA:LNA a io) du ing olume ic bo le e ec in Exp 1 (A), Exp 2 (B) and Exp 3 (C). Solid symbols e e o he whole plank onic communi y and emp y symbols o he il e ed bac e ial communi y. No e he di e en scales on he y axes. 37 Figu e 8.Tempo al e olu ion o he au o ophic: he e o ophic biomass a io du ing olume ic bo le e ec in Exp1 (A), Exp 2 (B) and Exp 3(C). Solid symbols e e o he whole plank onic communi y and emp y symbols o he il e ed bac e ial communi y. No e he di e en scales on he y axes. 38 Figu e 9. Cellula abundance o picophy oplank on (P ochlo ococcus, Synechococcus and picoeuka yo es) and he e o ophic bac e ia (HNA and LNA) du ing he oxici y assay wi h INT 39 Fig10. Bac e ial iabili y (%) du ing he oxici y assay wi h INT 40 Table 1. Con e sion ac o s used o ob ain he picoplank on biomass in ou s udy a ea, desc ibed by Mon e o (2012) Øbio olume biomass (μm)(μm 3 ) gCcel -1 HNA bac e ia 0.46 0.05 10 LNA bac e ia 0.63 0.13 26 P ochlo ococcus 0.68 0.16 33 Synechococcus 0.97 0.48 96 Picoeuka yo es 0.85 0.32 64 Cells