Importance of mixotrophic nanoplankton in Aysén Fjord (Southern Chile) during austral winter

Continental Shelf Research 31 (2011) 216ndash224

Contents lists available at ScienceDirect

Continental Shelf Research

0278-43

doi101

n Corr

Ambien

Chile T

E-m

journal homepage wwwelseviercomlocatecsr

Importance of mixotrophic nanoplankton in Aysen Fjord (Southern Chile)during austral winter

Till Czypionka ab Cristian A Vargas bcn Nelson Silva d Giovanni Daneri ceHumberto E Gonzalez cef Jose Luis Iriarte cg

a University of Southern Denmark Institute of Biology Campusvej 55 Odense Denmarkb Unidad de Sistemas Acuaticos Centro de Ciencias Ambientales EULA Chile Universidad de Concepcion Casilla 160-C Concepcion Chilec Centro de Investigacion en las Ecosistemas de la Patagonia (CIEP) Bilbao 449 Coyhaique Chiled Escuela de Ciencias del Mar P Universidad Catolica de Valparaıso Casilla 1020 Valparaıso Chilee Centro de Investigacion Oceanografica del Pacıfico Sudoriental (COPAS) Universidad de Concepcion Casilla 160-C Concepcion Chilef Instituto de Biologıa Marina Universidad Austral de Chile Casilla 567 Valdivia Chileg Instituto de Acuicultura Facultad de Pesquerıas y Oceanografıa Universidad Austral de Chile Casilla 1327 Puerto Montt Chile

a r t i c l e i n f o

Article history

Received 8 October 2009

Received in revised form

14 June 2010

Accepted 21 June 2010Available online 3 July 2010

Keywords

Mixotrophy

Nanoplankton

Flagellates

Protists

Light limitation

Fjord ecosystem

43$ - see front matter amp 2010 Elsevier Ltd A

016jcsr201006014

esponding author at Unidad de Sistemas A

tales EULA Chile Universidad de Concepcion

el +56 41 2204032 fax +56 41 2207076

ail address crvargasudeccl (CA Vargas)

a b s t r a c t

Mixotrophy the combination of autotrophic and heterotrophic nutrition in the same organism is

widespread in planktonic algae Several reports from temperate and high-latitude fjords in Scandinavia

suggest the occurrence of a niche in late summer and autumn during post-bloom conditions in which

mixotrophic algae can become important grazers in pelagic ecosystems accessing the nutrients bound

in their prey to overcome nutrient limitation Here we experimentally determined the trophic modes

and bacterivory rates for the nanoplankton community (2ndash20 mm) in Aysen Fjord located in the Chilean

Northern Patagonia during two contrasting seasons winter and spring While mixotrophic

nanoplankton was virtually absent from the system in spring in winter at occasions it even constituted

the dominant trophic group of the nanoplankton with abundances of 4900 cells mL1 This indicates a

second niche for mixotrophs in winter when mixotrophy allows overcoming light limitation

amp 2010 Elsevier Ltd All rights reserved

1 Introduction

Mixotrophy is defined as the ability of an organism to combineautotrophic and heterotrophic nutrition The pervasive occurrenceof mixotrophs in the plankton of both marine and freshwaterecosystems is now realized (Jones 2000) Mixotrophy can be seen asa special form of omnivory allowing organisms to overcomenutrient carbon light or energy limitation (Elser and Hessen2005) Surveys in Scandinavian fjords and coastal embayments haveshown that mixotrophic nanoflagellates (nanoplankton size range2ndash20 mm) can play an important role mixotrophic flagellate algaedominate the protist community and constitute up to 86 of thetotal flagellate grazing (Havskum and Riemann 1996) Thisphenomenon has been confirmed by less extreme but significantreports of mixotrophic nanoflagellates in such environments(Nygaard and Tobiesen 1993 Havskum and Hansen 1997) In theChilean fjord region (43ndash561S) the phytoplankton is typicallydominated by large diatoms in spring and nanoflagellates in winter

ll rights reserved

cuaticos Centro de Ciencias

Casilla 160-C Concepcion

(Pizarro et al 2005) Small algae are more effective than their largercompetitors at dissolved nutrient uptake and in their use of light forphotosynthesis (Raven 1998) which should favor them in periodsof light or nutrient limitation as well as in situations in whichmixotrophy might also offer them an advantage However up to thispoint the use of mixotrophy as a nutritional strategy has not beenanalyzed in the Chilean fjord region Here we experimentallyinvestigate the winter and spring communities of nanoflagellatealgae (2ndash20 mm) asking one simple question Is mixotrophy a majorstrategy in small algae (to overcome light limitation) during winterin high-latitude oligotrophic fjord environments

Although simple this question has at least three major implica-tions First mixotrophs complicate the flow of energy and nutrientsin fjord food webs by functioning as both producers and consumersrendering classical models of ecosystem functioning incompleteSecond as a consequence mixotrophy requires considering thenutrients bound in the prey of the mixotrophs which might changethe nutrients finally limiting photosynthesis (Jansson 1998) Thirdtheoretical food-web models suggest that mixotrophy has consider-able effects eg changes in the overall primary production andcarbon and nutrient channeling systems as well as new competitivepressures on phyto- and zooplankton (Jones 2000 Tittel et al2003) Field studies contrasting the role and impact of mixotrophic

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224 217

nanoplankton between seasons are scarce for marine ecosystemsand most were carried out at low-latitude environments(eg Christaki et al 1999 Tsai et al 2007) Although mixotrophydescribes many forms of nutritional strategies ranging from absoluteautotrophy to absolute heterotrophy (Jones 2000) herein ouremphasis will be on algae that combine phototrophic autotrophywith phagotrophic heterotrophy These algae were evaluatedfunctionally by the possession of photo-pigments and theirexperimentally demonstrated ability to take up prey We performedfeeding experiments with fluorescently labeled bacteria (FLB) takenfrom a fjord in contrasting seasons and at different depths relatingour findings to various environmental variables This work there-fore bridges a remaining gap in our knowledge suggesting anecessity to incorporate mixotrophy into the analyses of ecosystemfunctioning in the Chilean fjord region and in high-latitude fjordenvironments in general

2 Material and methods

21 Study area

The study was carried out in the inner sound of Aysen Fjord inSouthern Chile (451220S 731040W see Fig 1 for map) during austral

Fig 1 Location of the study area in the context of the Patagonian fjord system as

indicated by the arrow (right) location of station 79 in Aysen Fjord (cross)

Table 1Overview over the experiments sample date sample depth and in situ temperature s

Experiment Date (mmddyyyy) Depth (m) Temperatu

1 08022007 1 72

2 08022007 10 95

3 08022007 25 99

4 08042007 1 87

5 08042007 5 94

6 08042007 25 99

7 08062007 1 94

8 08062007 5 94

9 08062007 25 103

10 11082007 5 92

11 11082007 10 95

12 11082007 25 95

13 11092007 5 97

14 11092007 10 95

15 11092007 25 95

16 11112007 5 98

17 11112007 10 96

18 11112007 25 95

winter (2ndash6 August 2007) and austral spring (8ndash11 November 2007)on board the RV AGOR Vidal Gormaz Aysen Fjord is a typical silledhigh-latitude fjord with a freshwater inflow (Caceres et al 2002)Salinity temperature and density were measured with a Seabird 19CTD Salinity and density measurements were visualized using thesoftware Ocean Data View (Schlitzer 2008) Seawater samples weregathered with a Teflon-coated Niskin bottle rosette system Nutrientanalyses comprising orthophosphate (PO4

3) nitrate (NO3) and silicic

acid (Si(OH)4) were performed for standard depths (0 2 5 10 25 5075 100 m) every 3 h from 2 to 3 August and from 7 to 8 November2007 Samples were stored at ndash25 1C in aseptic polyethylene-bottles(50 mL) until their analysis using an automatic nutrient analyzeraccording to Atlas et al (1971) Three qualitative depths were definedand used to determine the size-fractionated chlorophyll-a (Chl-a)concentrations and primary production (PP) bacterial abundance(BA) and bacterial production (BP) these were designated surface(0 m in winter 5 m in spring) chlorophyll maximum (chlor max)and dark (25 m) The chlor max depth was designated dailyfollowing the CTD profile analysis in the halocline between thefreshwater influenced upper layer and the underlying saltier watermasses (see Table 1 for overview) Measurements of PP by size classes(o2 2ndash20 420 mm) were performed for the three depths inanalysis using the 14C-isotope incorporation method described bySteemann-Nielsen (1952) Chl-a concentrations were measured forthe same depths and size classes using a digital Turner PS 700fluorometer after Chl-a extraction in 90 vv acetone For a detaileddescription of both methods refer to Iriarte et al (2007 and citationstherein) Samples for BA were preserved in glutaraldehyde (60 WVin 02 mm prefiltered seawater) stained with DAPI (406-diamidino-2-phenylindole final conc of 1 mg mL1) and subsequently collectedon 02-mm black Nucleopores filters (Porter and Feig 1980)

Water for BP estimations was incubated in parallels using watersamples from the same oceanographic bottle as for PP experimentsBP was estimated by the incorporation of L-[14C(U)]-leucine intoproteins at a final concentration of 10 nM (Simon and Azam 1989)Triplicate samples of 10 mL and a formalin-killed control wereincubated at the in situ temperature and darkness for 1 h Afterincubation samples were filtered over 022 mm membrane filtersand extracted with cold 5 trichloroacetic acid Samples werefrozen on board and counted in a liquid scintillation counter in thelaboratory Leucine incorporation was converted into biomassproduction using the carbon fraction of proteins of 086 (Simon andAzam 1989) The bacterial volume was calculated from length andwidth measurements of at least 20 cells per sample by using

alinity and bacterial abundance

re (1C) Salinity (PSU) Bacterial abundance (106 cells mLndash1)

187 49

290 36

302 2

87 59

291 3

303 22

292 35

293 13

304 22

85 97

277 39

296 17

276 5

286 28

294 26

271 88

283 35

295 23

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224218

standard geometric forms for a sphere and an ellipsoid (ie rodsand bacilli) respectively The volume-to-carbon conversion wascarried out using a factor of 035 pg C mm3 (Bjoslashrnsen 1986) Fornanoflagellates the carbon content was estimated in similarmanner using a conversion factor of 011 pg C mm3 (Havskumand Hansen 1997) Trapezoidal integration was used for depthintegration of PP BP and nanoplankton biomass The conversion ofPP per hour to PP per day was performed using daylight hourinformation retrieved from the website of the AstronomicalApplications Department of the US Naval Observatory (httpaausnonavymil)

22 Functional analysis of the nanoplankton community

Seawater samples from three depths (surface chlor maxdark see Table 1) were filtered through a Nitex net (20-mm mesh)to obtain the nanoplankton fraction and left undisturbed for 1 h atthe in situ temperature until setting up the experiment Thefunctional analysis of the nanoplankton community was madeusing FLB which were prepared by staining natural marinebacterial assemblages from the coastal waters of centralChile (361320S 721560W) with 5-(46-dichlorotriazin-2-yl)aminofluorescein (DTAF) according to Sherr et al (1987) FLBuptake experiments were carried out in triplicate in 250-mL acid-cleaned polycarbonate Nalgenes bottles and incubated in atemperature-regulated on-deck rack equipped with a shadingdevice mimicking the in situ light regime when necessary FLBwere added to a final concentration of 105 FLB mL1 in the assay(Sherr et al 1987) Subsamples of 30 mL were collected at 0 1030 and 50 min fixed with glutaraldehyde (final concentration of2 Christaki et al 1999) and stained with a combinationof 1200 mL of DAPI (final concentration 4 mg mL1) and 600 mL ofproflavine (final concentration 0033 Haas 1982) Stainedsamples were filtered on 08-mm Nucleopores filters and storedat 20 1C until analysis Analysis was performed at 1000

Fig 2 (a) Salinity distribution over 30 m depth for the study period overlaid with

size-fractionated PP integrated over the upper 25 m water column for the respective sam

over 25 m for the respective sampling days subdivided by trophic mode and size

magnification with an Olympus IX-51 fluorescence microscopeusing one filter (excitation UV-light 340ndash380 nm emission435ndash485 nm) to observe DTAF and DAPI fluorescence and asecond filter (excitation 450ndash490 nm emission 515 nm) toobserve proflavine and chlorophyll autofluorescence At least250 flagellates per filter were classified as autotrophs(ie chloroplast autofluorescence) heterotrophs (ie proflavinestain without autofluorescence) or mixotrophs (ie chloroplastautofluorescence and FLB ingestion) as well as being classifiedaccording to their size (2ndash5 and 5ndash20 mm) Food vacuole contentswere estimated for both mixotrophic and heterotrophic cellsbased on the FLB ingestion We calculated the number of ingestedFLB ind1 h1 from the slope of food vacuole content versusincubation time for each bottle and the derived clearance rate viadivision by the FLB concentration Average values of ingestedFLB ind1 h1 and clearance rates were calculated for triplicatebottles over the linear portion of the uptake curve The ingestionrate was then calculated as the product of the clearance rate andthe ambient bacterial abundance Carbon balances were madeusing ingestion rate data assuming a gross growth efficiency of32 (Straile 1997) Mixotrophs were assumed to be only 52 asefficient in fixing inorganic carbon as strict autotrophs (Hansenet al 2000 Adolf et al 2006) which was used for calculatingtheir share of total nanoplankton PP Carbon balances werecalculated for an entire day using daylight information obtainedas described at the end of Section 21

3 Results

Contrasting environmental conditions were found betweenaustral winter and spring As is typical in most high-latitude fjordecosystems the physicalchemical conditions in the upper layer ofthe fjord were heavily influenced by freshwater runoff (Fig 2a)The influence of freshwater runoff in the upper water column was

(sigma-T-) isopycnals vertical dotted lines indicate sampling days (b) BP and

pling days Note different axes for PP and BP (c) Nanoplankton biomass integrated

Fig 3 Concentrations of dissolved inorganic nutrients at the time other

parameters were sampled as well nitrate (black dot) silicic acid (white dot)

and phosphate (black cross) over 25 m depth in winter (02 August 2007) and

spring (08 November 2007) Symbols indicate nutrient sampling depths

Horizontal dotted lines indicate depths where nanoplankton was sampled

Fig 4 Distribution of nanoflagellate abundance from the winter sampling (2 4 6

August 2007) subdivided by trophic mode and size range at lsquolsquosurfacersquorsquo (0 m in

winter 5 m in spring) chlorophyll maximum (lsquolsquochlor maxrsquorsquo) and lsquolsquodarkrsquorsquo (25 m)

For 2 and 4 August carbon balances for MNFs in the 5ndash20 mm size range at the

lsquolsquosurfacersquorsquo and lsquolsquochlor maxrsquorsquo depths are given as insets gray barsfrac14autotrophic

carbon gain white barsfrac14heterotrophic carbon gain

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224 219

more intense in spring perhaps due to the increased glacial icemelting occurring in this season in Patagonia (Caceres et al2002) Integrated PP was considerably lower in winter (maxvalues 1465 mg C m2 d1) than in spring (max values2904 mg C m2 d1) BP was very low compared to PP withvalues reaching 22 mg C m2 d1 in winter and524 mg C m2 d1 in spring contributing a maximum share ofonly 35 of the total integrated production (Fig 2b) Theinfluence of freshwater runoff was also visible in the low levelsof NO3

and PO43 and high Si(OH)4 concentrations in the brackish

waters of the upper 5 m (Fig 3) suggesting that phytoplankton inbrackish waters may be able to grow under PO4

3-deficientconditions (see Discussion) The PO4

3 concentration wasdetectable at the time we did the experiments but droppedbelow the detection limit during the course of the day andreached concentrations up to 094 mM We believe thesefluctuations show the influence of tides the ship from whichthe measurements were taken remained at a fixed position whilethe tide drove the water masses with their different freshwaterinfluxes back and forth below the vessel In spring the PO4

3 andNO3 depletion was even more persistent in the surface layer

(above our lsquolsquosurfacersquorsquo sample depth of 5 m)Fig 4 illustrates the itemization (by trophic mode and size) of

the nanoflagellate community over depth for the winter samplingperiod On 2 and 4 August 2007 (austral winter) thenanoplankton community in the surface layer was dominatedby large nanoflagellates (5ndash20 mm) (Fig 2c) which were in turndominated (490) by cryptophyte algae In addition to thedominant large cryptophyte algae there was always abackground of small heterotrophic flagellates for which FLBingestion was rarely observed In contrast when mixotrophiccryptophytes fed they did so heavily on FLB with individual foodvacuole contents of up to 12 FLB cell1 At the chlor max depthmixotrophy became less important whereas small autotrophs(ie non-cryptophytes) loomed large although their share neversurpassed that of the typical ubiquitous small lsquolsquobackgroundrsquorsquoheterotrophs At 25 m the depth defined as dark the generalflagellate abundance was low The data from 6 August 2007differed substantially from the other two sampling days in thewinter period This was due to a heavy storm the night before

sampling started that presumably broke down stratification Inspring the data were pretty consistent over the entire studyperiod (see Fig 5) In general abundances were higher comparedto winter In the surface layer small phototrophs (o5 mm) werequite important (max value 4200 cells mL1 min value800 cells mL1) and co-dominated the usual small lsquolsquobackgroundrsquorsquoheterotrophs (max value 5500 cells mL1 min value3800 cells mL1) The abundance of small phototrophs droppedat the chlorophyll maximum depth and downright disappeared at

Fig 5 Distribution of nanoflagellate abundance from the spring sampling (8 9 11

November 2007) subdivided by trophic mode and size range at the lsquolsquosurfacersquorsquo

(0 m in winter 5 m in spring) chlorophyll maximum (lsquolsquochlor maxrsquorsquo) depth and

lsquolsquodarkrsquorsquo (25 m)

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224220

25 m depth Values deviated for different sampling days but theobserved pattern was consistent over the study periodMixotrophy was virtually absent in spring at all analyzeddepths Small heterotrophs were mostly about 3 mm FLBingestion by these small heterotrophs was observed but wasscarce compared to grazing on FLB by cryptophyte algae in winter

Table 2 presents the values for size-fractionated PP (Table 2a)and the Chl-a concentrations (Table 2b) measured for therespective size fractions from 2 and 4 August 2007 plus 11November 2007 as an example of the springtime situation Total

PP production was relatively high during the two first sampledays in austral winter (229 and 261 mg C m3 h1) in thesurface layer whereas the corresponding value for 11 Novemberonly reached 107 mg C m3 h1 This spring-winter inversion ofthe expected PP values did not hold over the whole sample periodand values peaked at 236 mg C m3 h1 in spring and dropped aslow as 63 mg C m3 h1 in winter However Chl-a values wereon average higher in spring than in winter as would be expectedThe assimilation index defined as the mass carbon fixed by massof Chl-a as a measure of the autotrophic growth rate (Gocke et al2003) was calculated for the same dates and depths (Table 2c)For the two days of August the assimilation index of nano-plankton for the surface layer was 258 and 312 considerablyhigher than that of the pico- and microplankton such extremedifferences were not observed in spring when values were muchlower

On the occasions when mixotrophy played a major role (2 and4 August 2007 surface and chlor max size range 5ndash20 mm)carbon balances were made (see insets in Fig 4 for the respectivedates) The total autotrophic share of carbon acquisition washigher at the surface (76 and 62) than at the chlor max depthwhere the heterotrophic carbon acquisition was more important(52 and 62)

To evaluate the degree of prey saturation we plotted theingestion rates (IR) of the phagotrophic nanoflagellates (subdi-vided by trophic mode size and season sampled) against BA(Fig 6) Regression analyses were performed on the IR relation asa function of BA using a linear function and an exponential growthfunction Relations were better described by an exponentialfunction for data from winter (R2 values of 095ndash098) and by alinear function for spring (R2 values 061ndash085) (see Table 3) Thisdoes not apply to results for small mixotrophs in winter and largeheterotrophs in spring for which no convincing fits were foundlikely due to the low sample size of these compartments The bestfits according to the coefficient of determination (R2) are plottedin Fig 6

4 Discussion

41 Abundance of mixotrophic nanoflagellates

This study indicates that mixotrophic algae formed a majorpart of the nanophytoplankton community in an oligotrophichigh-latitude fjord environment during winter conditions whenautotrophic organisms were most likely limited in theirphotosynthetic carbon synthesis by light deficiency Due to largedeviations between the replicates no significant differences werefound between the concentrations of autotrophic heterotrophicand mixotrophic nanoplankton within the same size range (t-testpfrac14044) Nonetheless mixotrophic cryptophytes could be thedominant nanoplankton since our methodology tends tounderestimate their concentration and overestimate that ofhetero- and autotrophs The mixotrophs were identified basedon their FLB uptake Many studies in the literature reportdiscrimination against FLB compared to living bacteria (egLandry et al 1991) and in particular against motile bacteria(Gonzales et al 1993) Assuming a similar share of feeding-inactive cryptophytes in the mixotrophic compartment we couldhave misclassified a large fraction of the actual mixtotrophs asautotrophs In fact inactive grazers are a common phenomenon infeeding experiments (eg Schmidtke et al 2006) leading in thiscase to an underestimation of mixotroph abundance and anoverestimation the abundance of autotrophs

Table 2(a) Primary production and (b) Chl-a concentrations for winter (02 August 2007 04 August 2007) and spring (11 November 2007) for comparison In (c) the assimilation

index for 02 August 2007 and 04 August 2007 is also given

(a) Distribution of PP (mg C m3 h1) over depth Absolute values and relative contribution to total PP of the different size compartments

Date (ddmmyyyy) Depth (m) PP (mg C m3 h1) Contribution ()

Total 20ndash2 lm o2 lm 420 lm 20ndash2 lm o2 lm 420 lm

02082007 0 229 115 529 607 504 231 265

10 167 067 038 063 40 226 374

25 082 033 007 042 402 86 512

04082007 0 261 930 566 111 356 217 427

5 174 085 063 025 49 364 146

25 020 012 008 001 571 381 48

11112007 5 107 632 298 143 589 278 133

10 246 087 061 098 353 248 399

25 047 031 005 012 65 106 244

(b) Distribution of Chl-a (mg m3) over depth Absolute values and relative contribution to total Chl-a of the different size compartments

Date (ddmmyyyy) Depth (m) Chl-a (mg m3) Contribution ()

Total 20ndash2 lm o2 lm 420 lm 20ndash2 lm o2 lm 420 lm

02082007 0 148 045 033 070 303 226 471

10 052 025 018 009 486 342 182

25 019 005 002 013 232 81 687

04082007 0 202 030 044 128 148 22 632

5 084 048 017 020 566 20 234

25 015 007 004 003 5000 286 214

11112007 5 348 109 087 153 312 25 438

10 085 042 024 019 494 282 224

25 025 019 004 002 752 153 95

(c) Assilimation index (PP per Chl-a) for 2 August 2007 and 4 August 2007 at different sample depths Values for total phytoplankton and different size fractions

Date (ddmmyyyy) Depth (m) Assimilation index (mg C mg Chl-a1 h1)

Total 20ndash2 lm o2 lm 420 lm

02082007 0 155 258 158 872

10 322 265 213 663

25 423 733 448 315

04082007 0 129 312 128 873

5 206 178 375 128

25 139 162 196 036

11112007 5 308 582 343 094

10 290 207 256 513

25 187 161 129 481

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224 221

42 The benefits of mixotrophy in austral winter

Mixotrophy might help overcome dissolved nutrient limitationby granting access to the pool of particulate nutrients or may aidin overcoming carbon andor energy limitation (Jones 2000) Forour study the question is Did an organic carbonenergylimitation or a phosphate limitation trigger the rise of mixotrophyduring austral winter conditions in an oligotrophic fjordenvironment

43 Indications for phosphate limitation inducing mixotrophy

For our lsquolsquosurfacersquorsquo-measuring depth at 1 m NO3PO4

3 ratios ingeneral indicate a phosphate limitation reaching a maximum ofNO3PO4

3 50 during the 24-h sampling period in winter (rangeNO3

ndashPO43 at 0 m 480 at 2 m 30 data not shown) The

mixotroph abundance peaked in these surface brackish watersthat are relatively poor in nutrients year round as documented bymany authors for Patagonian fjords and channels (eg Palma and

Silva 2004) So the idea of phosphate limitation is rather plausiblein the absence of any other limiting factor (eg light)

44 Indications for carbonenergy limitation inducing mixotrophy

Carbon balances for mixotrophs in the 5ndash20 mm size range on2 and 4 August 2007 when the mixotroph abundance peaked (seeinsets in Fig 4) showed that both autotrophy and heterotrophycontributed considerably to the carbon acquisition (72 and 62autotrophic carbon gain at surface) Thus the heterotrophiccarbon acquisition cannot be identified as irrelevant and there-fore carbon acquisition cannot be dismissed as a trigger It makessense that autotrophy would become less important deeper in thewater column (38 and 48 autotrophic carbon gain at the chlormax depth) because of the reduced light availability at depth dueto sediments from glacial and river runoff in this region (Pizarroet al 2005) Carbon balances from 2 and 4 August 2007 might notbe representative of the winter season Both days were unusuallysunny for winter conditions a fact which is reflected in the

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224222

relatively high PP values (exceeding 22 and 26 mg C m3 h1 atthe surface) (Table 2a) whereas winter values are usuallyreported to be below 4 mg C m3 h1 in this region (Iriarteet al 2007)

The assimilation index (ie photosynthetic carbon fixation perChl-a) reached values more common for high-nutrient high-temperature regions such as tropical coastal lagoons (Gocke et al2003) which we suggest indicated that light conditions duringthese two days were well above those to which the phytoplanktonis usually exposed Thus the average share of heterotrophiccarbon acquisition is likely to be greater for the entire australwinter Iriarte et al (2007) mentioned high day-to-day variabilityin radiation due to changing cloud cover in this region butpointed out poor overall light conditions in winter An alternativeexplanation for the exceptionally high assimilation index valuesof the nanoplankton on 2 and 4 August 2007 might be thatphotosynthetic products leaking from the algae already contain-ing 14C as a marker were assimilated by the bacteria in the assayThis would allow for the possibility of redirecting organic 14C tobacterial grazers via the microbial loop The exceptionally highassimilation indices were found on the occasions with the highestbacterivory rates while the differences in the assimilation indices

Fig 6 Plots of in situ bacterial abundances (105 cells mL1) (x-axis) against

ingestion rates of flagellates (bacteria cell1 h1) (y-axis) Mind different ranges of

axes Top (A and B) flagellates in the 5ndash20 mm size range Bottom (C and D)

flagellates o5 mm Left column winter Right column spring Ingestion rates of

mixotrophs (white dots) and heterotrophs (black triangles) Curves indicate best

significant (po005) regression (linear or exponential function)

Table 3Results of correlation analyses between and predation rates (y) as a function of bacteria

bold Values with significance level p45 are in italics

Season Size Trophic mode yfrac14a+b x

R2 a b

Winter 5ndash20 mm Heterotrophy 084 626292 0000

Mixotrophy 085 840619 0000

o5 mm Heterotrophy 078 1967 883E

Mixotrophy 008 110859 94

Spring 5ndash20 mm Heterotrophy 002 47495 774E

Mixotrophy 064 142781 513E

o5 mm Heterotrophy 085 235595 0000Mixotrophy 061 8425 0000

between the size classes were less extreme when bacterivoryrates were low (Table 2c) This in turn would support the validityof our bacterivory rate measurements

45 Regression analysis between IR and BA

The regression analysis of the IR of the bacterial prey of thephagotrophs as a function of the in situ bacterial abundance offersstrong support for the hypothesis of a wintertime carbonenergylimitation According to the lsquolsquooptimal foraging theoryrsquorsquo the IRlevels off at saturating food concentrations as a functionalresponse equaling a decreased clearance rate and thus areduction in the costs of predationfiltering (Jumars 1993) Inwinter an exponential growth function best described the IR as afunction of BA for large mixotrophs which according to Jumars(1993) suggests that this group is far from food saturation (Fig 6Table 3) This also holds true for heterotrophs in winter they arenot suspected of feeding to overcome phosphate limitation sincetheir prey (ie bacteria) are known to have high phosphatecontents (Vadstein 2000) Therefore we concluded that largemixotrophs were carbonenergy limited in austral winter Eventhough the correlation for mixotrophs o5 mm in winter was notsignificant it is safe to assume that they were not limited bydissolved phosphate either because their higher surface-to-volume ratio makes them more unlikely to be limited by dissolvednutrients than their larger counterparts In spring IR as functionof BA was best described by a linear regression (except forheterotrophs in the 5ndash20 mm size range for which no satisfying fitcould be made) This indicates that food saturation was notreached but that its limitation was not as severe as in winter(Jumars 1993) Therefore even though dissolved phosphateconcentrations were low in the surface layer we assumed thatthe main benefit of mixotrophy (observed for the winter period)was carbon and energy acquisition in an environment of reducedphotosynthetic active radiation (PAR) This agrees with otherreports of mixotrophic cryptophytes for other ecosystems withreduced PAR (eg Marshall and Laybourn-Parry 2002 Hammerand Pitchford 2006) We are well aware of the risks implied indrawing conclusions from such correlations under non-laboratoryconditions with many varying factors not taken into accountFurthermore the conclusions drawn from the regression analysisof IR and BA are potentially flawed The lsquolsquooptimal foraging theoryrsquorsquois well established for predatorsfilter feeders with complexbehaviors It might be asked if these relationships really apply forthe organisms in this study Therefore although we weredoubtlessly able to show mixotrophy to be a major strategy forthe nanoplankton in winter we were not able to definitely settle

l abundance (x) Values for the function with the higher R2 as given in Fig 3 are in

yfrac14a exp(b x)

p of b R2 a b p of b

3 00107 09618 06214 867E06 000464 00085 0951 423 636E06 00032

06 00197 09843 00239 872E06 000065E06 05867 0 76555 678E16 1

06 07118 00151 61032 641E07 07723

05 00006 05603 19564 301E06 00329

1 00004 07699 84229 259E06 00078

1 00002 05055 152284 182E06 00038

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224 223

on the environmental factor causing the mixotrophy Sincemixotrophic nanoflagellates were recently shown to also play aprominent role in the Southern Ocean (Moorthi et al 2009) weencourage subsequent laboratory studies on mixotrophic nano-flagellates and especially cryptophytes to clarify this matterMoreover once an organism has invested in both photosyntheticand phagotrophic apparatuses the benefits do not have toappertain strictly to these categories since the sum of thebenefits can also make mixotrophy a beneficial strategy (Tittelet al 2003)

46 Integration of the findings in the regional and global context

Although limited in terms of spatial coverage we assume ourfindings are not lsquolsquoaccidentsrsquorsquo but are representative of a largerrecurring relevant pattern Iriarte et al (2007) confirmed thedominance of smaller phytoplankton size classes (cryptophytessmall naked dinoflagellates prymnesiophytes) in winter andpost-bloom periods in the region Furthermore the generareported by Iriarte et al (2007) are all known to containmixotrophic members (Safi and Hall 1999)

Whereas our study identified light and thus the energycarbon limitation as the main trigger for mixotrophy inaustral winter conditions it is important to realize that thissubstantially differs from the reports of mixotrophic abundancein coastal embayments mentioned in the Introduction (Nygaardand Tobiesen 1993 Havskum and Riemann 1996 Havskumand Hansen 1997) These studies described situations duringpost-bloom conditions when dissolved inorganic nutrientsbecame limiting so that mixotrophy was a beneficial strategyin the phytoplankton in order to overcome an inorganicnutrient limitation

Whether these expectations of a second niche for mixotrophyduring post-bloom conditions also apply to the Southern Chileanfjord region remains to be validated The region is renowned forrecurrent harmful algae blooms consisting mainly of the dino-flagellates Alexandrium catenella Dinophysis acuminata andGymnodinium spp among others (Iriarte and Gonzalez 2008)These species are also reported to be mixotrophic (Jacobson andAndersen 1994 Legrand and Carlsson 1998 Li et al 2000) butmostly belong to the microplankton (size range 20ndash200 mm) andthus may feed on larger eukaryotic prey They are therefore notdirectly comparable to the nanoplankton community analyzed byus but these harmful algae bloom-forming species might be usedas proxies for mixotrophy thereby motivating further research onthis issue in the future Even though the patterns are diverseharmful algae blooms are reported to accumulate in summerautumn during post-bloom conditions (Uribe and Ruiz 2001Guzman et al 2002 Molinet et al 2003 Seguel et al 2005) Thisholds true not only for the Southern Chilean fjord region but alsofor its counterparts in the Northern Hemisphere where analogouspatterns in plankton succession (Druzkhova and Makarevich2008) as well as the formation of harmful algae blooms aretypically observed (Cembella et al 2005)

To our knowledge we have comparatively measured thebacterivory rates of phototrophic protists and heterotrophicprotists inhabiting a Chilean Patagonian fjord ecosystem for thefirst time showing the significant contribution of mixotrophicalgae during environmentally limiting conditions

Acknowledgements

We are grateful for the assistance of the captains and crew ofthe RV AGOR Vidal Gormaz (Chilean Navy) and the manyundergraduate and graduate students who participated in our

cruises (especially Ms Paulina Contreras) This work was fullysupported by Fondecyt Project No 1060709 to CV and CIMARFiordos Program Project Nos CONA-C12F 06-08 to NS and CONA-C12F 06-10 to all other co-authors TC was supported by anInternationaliseringspuljen grant from his home university duringhis stay at the Aquatic Ecosystem Functioning Lab (CA VargasEULA Centre UdeC) in Chile

References

Adolf JE Stoecker DK Harding LW 2006 The balance of autotrophy andheterotrophy during mixotrophic growth of Karlodinium micrum (Dinophy-ceae) J Plankton Res 28 737ndash751

Atlas EL Hager SW Gordon LI Park PK 1971 A practical manual for use ofthe Technicon AutoAnalyzer in sea water nutrient analyses Oregon StateUniversity Technical Report

Bjoslashrnsen PK 1986 Automatic determination of bacterioplankton biomass byimage analysis Appl Environ Microbiol 51 1199ndash1204

Caceres M Valle-Levinson A Sepulveda HH Holderied K 2002 Transversevariability of flow and density in a Chilean fjord Cont Shelf Res 221683ndash1698

Cembella AD Ibarra DA Diogene J Dahl E 2005 Harmful alga and theirassessment in fjords and coastal embayments Oceanography 18 158ndash171

Christaki U van Wambeke F Dolan JR 1999 Nanoflagellates (mixotrophsheterotrophs and autotrophs) in the oligotrophic Eastern Mediterraneanstanding stocks bacterivory and relationships with bacterial production MarEcolmdashProg Ser 181 297ndash307

Druzhkova EI Makarevich PR 2008 Annual cycle of nanophytoplankton incoastal waters of the Barents Sea Biol Bull 35 428ndash435

Elser JJ Hessen DO 2005 Biosimplicity via stoichiometry the evolution offood-web structure and processes In Belgrano S Scharler UM Dunne JUlanowicz RE (Eds) Aquatic Food Webs An Ecosystem Approach OxfordUniversity Press Oxford UK pp 7ndash18

Gocke K Mancera JE Vidal LA Fonseca D 2003 Planktonic primaryproduction and community respiration in several coastal lagoons of the outerdelta of the Rio Magdalena Colombia Bol Invest Mar Cost 32 125ndash144

Gonzales JM Sherr EB Sherr BF 1993 Differential feeding by marineflagellates on growing versus starving and on motile versus nonmotilebacterial prey Mar EcolmdashProg Ser 102 257ndash267

Guzman L Pacheco H Pizarro G Alarcon C 2002 Alexandrium catenella yveneno paralizante de los mariscos en Chile In Sar EA Ferrario MEReguera B (Eds) Floraciones algales nocivas en el Cono Sur AmericanoInstituto Espanol de Oceanografıa Madrid Spain pp 235ndash256

Haas LW 1982 Improved epifluorescent microscopic technique Ann InstOceanogr Paris 58 (Suppl) 261ndash266

Hammer AC Pitchford PW 2006 Mixotrophy allelopathy and the populationdynamics of phagotrophic algae (cryptophytes) in the Darss Zingst Boddenestuary Southern Baltic Mar EcolmdashProg Ser 328 105ndash115

Hansen PJ Skovgaard A Glud RN Stoecker DK 2000 Physiology of themixotrophic dinoflagellate Fragilidium subglobosum II Effects of time scale andprey concentration on photosynthetic performance Mar EcolmdashProg Ser 201137ndash146

Havskum H Hansen A 1997 Importance of pigmented and colourless nano-sized protists as grazers on nanoplankton in a phosphate-depleted Norwegianfjord and in enclosures Aquat Microb Ecol 12 139ndash151

Havskum H Riemann B 1996 Ecological importance of bacterivorouspigmented flagellates (mixotrophs) in the bay of Aarhus Denmark MarEcolmdashProg Ser 137 251ndash263

Iriarte JL Gonzalez HE 2008 Phytoplankton bloom ecology of the inner sea ofChiloe Southern Chile Nova Hedwigia 133 67ndash79

Iriarte JL Gonzalez HE Liu KK Rivas C Valenzuela C 2007 Spatial andtemporal variability of chlorophyll and primary productivity in surface watersof Southern Chile (415ndash431S) Estuarine Coastal Shelf Sci 74 471ndash480

Jacobson DM Andersen RA 1994 The discovery of mixotrophy in photosyn-thetic species of Dinophysis (Dinophyceae) light and electron microscopicalobservations of food vacuoles in Dinophysis acuminata D norvegica and twoheterotrophic dinophysoid dinoflagellates Phycologia 33 97ndash110

Jansson M 1998 Nutritional limitation and bacteriandashphytoplankton interactionsin humic lakes In Hessen DO Tranvik LJ (Eds) Aquatic Humic SubstancesEcology and Biochemistry Springer Heidelberg Germany pp 177ndash195

Jones RI 2000 Mixotrophy in planktonic protists an overview Freshwater Biol45 219ndash226

Jumars P 1993 Concepts in Biological Oceanography An Interdisciplinary PrimerOxford University Press New York USA

Landry MR Lehner-Fournier JM Sundstrom JA Fagerness VL Selp SE 1991Discrimination between living and heat-killed prey by a marine zooflagellateParaphysomomnas vestita (Stokes) J Exp Mar Biol Ecol 146 139ndash151

Legrand C Carlsson P 1998 Uptake of high molecular weight dextran by thedinoflagellate Alexandrium catenella Aquat Microb Ecol 16 81ndash86

Li A Stoecker DK Coats DW 2000 Mixotrophy in Gyrodinium galatheum(Dinophyceae) grazing responses to light intensity and inorganic nutrientsJ Phycol 36 33ndash45

T Czypionka et al Continental Shelf Research 31 (2011) 216ndash224224

Marshall W Laybourn-Parry J 2002 The balance between photosynthesis andgrazing in Antarctic mixotrophic cryptophytes during summer FreshwaterBiol 47 2060ndash2070

Molinet C Lafon A Lembeye G Moreno CA 2003 Spatial and temporaldistribution patterns of blooms of Alexandrium catenella (Whedon amp Kofoid)Balech 1985 on inland seas of Northwest Patagonia Chile Rev Chil Hist Nat76 681ndash698

Moorthi SD Caron DA Gast R Sanders RW 2009 Mixotrophy awidespread and important ecological strategy for planktonic and sea-icenanoflagellates in the Ross Sea Antarctica Aquat Microb Ecol 54269ndash277

Nygaard K Tobiesen A 1993 Bacterivory in algae a survival strategy duringnutrient limitation Limnol Oceanogr 38 273ndash279

Palma S Silva N 2004 Distribution of Siphonophores Chaetognaths Euphau-siids and oceanographic conditions in the fjords and channels of SouthernChile Deep-Sea Res II 51 513ndash515

Pizarro G Montecino V Guzman L Munoz V Chacon V Pacheco HFrangopulos M Retamal L Alarcon C 2005 Patrones locales recurrentesdel fitoplancton en fiordos y canales australes (43ndash56 1S) en primavera yverano Cienc Tecnol Mar 28 63ndash83

Porter KG Feig YS 1980 The use of DAPI for identifying and counting aquaticmicroflora Limnol Oceanogr 25 943ndash948

Raven JA 1998 The twelfth Tansley lecture small is beautiful the picophyto-plankton Funct Ecol 12 503ndash513

Safi KA Hall JA 1999 Mixotrophic and heterotrophic nanoflagellate grazing inthe convergence zone east of New Zealand Aquat Microb Ecol 20 83ndash93

Schlitzer R 2008 Ocean Data View httpodvawideS

Schmidtke A Bell EM Weithoff G 2006 Potential grazing impact of themixotrophic flagellate Ochromonas sp (Chrysophyceae) on bacteria in anextremely acidic lake J Plankton Res 28 991ndash1001

Seguel M Tocornal MA Sfeir A 2005 Floraciones algales nocivas en los canalesy fiordos del sur de Chile Cienc Tecnol Mar 28 5ndash13

Sherr BF Sherr EB Fallon RD 1987 Use of monodispersed fluorescentlylabeled bacteria to estimate in situ protozoan bacterivory Appl EnvironMicrobiol 53 958ndash965

Simon M Azam F 1989 Protein content and protein synthesis rates ofplanktonic marine bacteria Mar EcolmdashProg Ser 51 201ndash213

Steemann-Nielsen E 1952 The use of radiocarbon (14C) for measuring organicproduction in the sea J Cons Int Explor Mer 18 117ndash140

Straile D 1997 Gross Growth Efficiencies of protozoan and metazoan zooplank-ton and their dependence on food concentration predator prey weight ratioand taxonomic group Limnol Oceanogr 42 1375ndash1385

Tittel J Bissinger V Zippel B Gaedke U Bell E Lorke A Kamjunke N 2003Mixotrophs combine resource use to outcompete specialists implications foraquatic food webs Proc Natl Acad Sci USA 100 12776ndash12781

Tsai A Chiang K Chan Y Lin Y Chang J 2007 Pigmented nanoflagellates inthe coastal western subtropical Pacific are important grazers on Synechococcuspopulations J Plankton Res 29 71ndash77

Uribe JC Ruiz M 2001 Gymnodinium brown tide in the Magellanic fjordsSouthern Chile Rev Biol Mar Oceanogr 36 155ndash164

Vadstein O 2000 Heterotrophic planktonic bacteria and cycling of phosphorusPhosphorus requirements competitive ability and food web interactions InSchink B (Ed) Advances in Microbial Ecology Plenum Press New York USApp 115ndash167