zooplankton grazing on natural algae and bacteria under … · 2015-12-13 · romo et al. (2008)....

Limnetica, 29 (2): x-xx (2011)Limnetica, 34 (2): 541-560 (2015)c© Asociación Ibérica de Limnología, Madrid. Spain. ISSN: 0213-8409

Zooplankton grazing on natural algae and bacteria under hypertrophicconditions

Gabriela Onandia1,∗, Juliana D. Dias2 and Maria Rosa Miracle1

1 Departamento de Microbiología y Ecología, ICBiBE, Universidad de Valencia, Calle Dr. Moliner 50, E-46100Burjassot (Valencia), España.2 Nupélia, Programa de Pós Graduação em Ecologia de Ambientes Aquáticos Continentais, UniversidadeEstadual de Maringá, Maringá, PR, Brazil.

∗ Corresponding author: [email protected]

Received: 24/11/14 Accepted: 04/08/15

ABSTRACT

Zooplankton grazing on natural algae and bacteria under hypertrophic conditions

The grazing characteristics of the natural zooplankton community in the shallow hypertrophic lagoon Albufera de Valencia(Spain) were compared in two periods of the annual cycle: (1) at the onset of spring, after a period of increased waterflow through the lagoon that reduces phytoplankton density and cyanobacterial prevalence, and (2) in mid-spring, after thespring phytoplankton bloom and under conditions of cyanobacterial dominance. Clearance, ingestion and assimilation rateswere measured by labelling natural seston with 14C-bicarbonate and 3H-thymidine/leucine. The clearance rates (CRs) ofDaphnia magna, the dominant species at the onset of spring, were low, very likely as a consequence of the high abundanceof filamentous cyanobacteria in the lagoon. The CRs were nearly equal for phytoplankton and bacteria, corroborating theunselective feeding of Daphnia. Bacteria supplied ≈ 40 % of the planktonic carbon ingested byD. magna, and this carbon wasassimilated with the same efficiency as that provided by phytoplankton. This suggests a strong coupling between the microbialloop and the classic grazer trophic web at the onset of spring in the lagoon and, consequently, an important contribution ofthe microbial loop to energy transfer to higher trophic levels (fish). The CRs of Bosmina longirostris, the dominant speciesin mid-spring, were also low probably because the food concentrations were well above the incipient limiting level. TheCRs were higher for 1-15 µm algae than for < 1 µm bacteria, indicating positive selective feeding towards larger cells. Algaesized 3-15 µm provided most of the planktonic carbon ingested by B. longirostris. Similarly, regarding the total mid-springzooplankton community, phytoplankton contributed 90 % of the planktonic carbon ingested. This suggests that during thespring phytoplankton bloom when cyanobacteria dominate, the energy transferred to zooplankton via bacteria constitutes asmaller fraction than during periods of lower phytoplankton density. Finally, our results indicate that zooplankton consumed avery small proportion of phytoplankton and bacterioplankton production during the two studied periods, which highlights thereduced ability of zooplankton to control phytoplankton growth in hypertrophic systems.

Key words: Daphnia magna, Bosmina longirostris, clearance rates, shallow lake, carbon flux.

RESUMEN

Alimentación del zooplancton sobre comunidades naturales de algas y bacterias en condiciones hipertróficas

Las características filtradoras de la comunidad natural de zooplancton en el lago somero hipertrófico Albufera de Valencia(España) fue comparada en dos periodos del ciclo anual: (1) al inicio de la primavera, tras un periodo de intenso flujo en ellago que reduce la densidad fitoplantónica y la prevalencia de cianobacterias y (2) a mediados de primavera, tras el pico decrecimiento de fitoplancton y en condiciones de dominancia de cianobacterias. Las tasas de filtración, ingestión y asimilaciónfueron medidas marcando el seston natural con 14C-bicarbonato y 3H-Timidina/Leucina. Las tasas de filtración (CR’s) deDaphniamagna, la especie dominante al inicio de la primavera, fueron bajas, probablemente como consecuencia de la elevadaabundancia de cianobacterias filamentosas en el lago. Las CR’s fueron prácticamente iguales para fitoplancton y bacterias,corroborando la alimentación no selectiva de Daphnia. Las bacterias proporcionaron el ≈ 40 % del carbono planctónicoingerido por D. magna, y este carbono fue asimilado con la misma eficiencia que el suministrado por el fitoplancton. Estosugiere un estrecho acoplamiento entre el bucle microbiano y la red trófica clásica a principios de primavera en el lago y enconsecuencia una contribución importante del bucle microbiano a la transferencia de energía a niveles tróficos superiores

16159_Limnetica 34(2), pàgina 283, 24/11/2015

542 Onandia et al.

(peces). Las CR’s de Bosmina longirostris, la especie dominante a mediados de primavera, fueron bajas, posiblemente debidoa que las concentraciones de alimento superaban el “nivel saturante de ingestión”. Las CR’s fueron más altas para algasde 1-15 µ m que para bacterias < 1 µ m, indicando una selección positiva de las células de mayor tamaño. Las algas de3-15 µ m proporcionaron la mayor parte del carbono planctónico ingerido por B. longirostris. Asimismo, si se considera todala comunidad zooplanctónica de mediados de primavera, el fitoplancton aportó el 90 % del carbono planctónico ingerido.Esto sugiere que durante picos de crecimiento fitoplantctónico en los que dominan las cianobacterias, la energía transferida alzooplancton a través de las bacterias constituye una fracción más pequeña en comparación con periodos de menor densidadfitoplanctónica. Finalmente, nuestros resultados indican que durante los periodos estudiados, el zooplancton consumió unamínima proporción de la producción fitoplanctónica y bacteriana, lo que evidencia la reducida capacidad del zooplanctonpara controlar el crecimiento de fitoplancton en sistemas hipertróficos.

Palabras clave: Daphnia magna, Bosmina longirostris, tasas de filtración, lago somero, flujo de carbono.

INTRODUCTION

Hypertrophic lakes are characterized by sus-taining an exceptionally high phytoplanktonbiomass. However, the majority of this biomassgenerally consists of filamentous and colonialcyanobacteria, inedible to most zooplankton andresponsible for additional distress to the feedingmechanism of large cladocerans (DeMott et al.,2001). Moreover, eutrophication typically in-volves an increase in the proportion of plank-tivorous and omnivorous fish at the expense ofpiscivorous species, which results in increasedpredation control over zooplankton (Jeppesenet al., 2007). Both feeding inhibition by fila-mentous cyanobacteria and fish predation havebeen found to favour small-bodied species ineutrophic lakes (Gliwicz, 1990).Consequently, hy-pertrophic lakes are often dominated by cyclo-poid copepods and rotifers and present a low zoo-plankton:phytoplankton biomass ratio (Havenset al., 2007; Fermani et al., 2013).In addition to phytoplankton, heterotrophic

bacteria might provide an alternative food sourcefor zooplankton. The bacterivory of zooplanktonhas been widely described in the literature (Tóth& Kato, 1997; Work & Havens 2003; Miracleet al., 2014), which leads to the question of theimportance of bacteria in channelling energy tozooplankton through the microbial food web inhypertrophic lakes.Numerous studies have investigated the food

selectivity and grazing characteristics of zoo-

plankton. However, the majority of them in-volved the use of cultured cell suspensions thatwere often monospecific (Lair, 1991; Tóth & Ka-to, 1997; He & Wang, 2006). There are few dataon grazing estimations based on natural planktonsuspensions, especially containing both algaeand bacteria from the natural assemblage. Fur-ther, the available information in this regard isespecially scarce for hypertrophic lakes.The shallow hypertrophic lagoon Albufera de

Valencia provides a representative case study forthis type of systems. This lagoon shifted froma clear to a turbid state as a consequence of aneutrophication process that began in the 1960s(Vicente & Miracle, 1992; Romo et al., 2005).Today, the lagoon completely lacks macrophytesand is in a turbid state dominated by phyto-plankton, namely cyanobacteria (Onandia et al.,2014a). Nonetheless, the lagoon experiencesshort “clear-water” events in late winter (Janua-ry-March). These events are the result of themanagement of the lagoon, which is “flushed”in winter when the water contained in the neigh-bouring rice paddies flows through the lagooninto the adjacent Mediterranean Sea. Duringthis period, there is a decrease in algal biomassmainly caused by the decline of cyanobacteria.After the flushing period, the water flow throughthe lagoon is strongly reduced, and cyano-bacterial dominance is re-established.The main objective of this work was to stu-

dy the grazing characteristics of the natural zoo-plankton community in the Albufera de Valencia


Zooplankton grazing under hypertrophic conditions 543

lagoon during periods of contrasting cyanobacte-rial dominance, i.e., the onset of spring vs mid-spring, characterized by non-cyanobacterial vscyanobacterial dominance. More specifically, weaimed to assess the importance of different algaland bacterioplankton size fractions to zooplank-ton feeding.

MATERIALS ANDMETHODS

Study site and general methods

Albufera de Valencia is a shallow, oligohaline(salinity≈1�) lagoon on theMediterraneancoast,15 km south of the town of Valencia (Spain).With an extension of roughly 24 km2 and a meandepth of approximately 1 m, the Albufera de Va-lencia is currently a hypertrophic, cyanobacteria-dominated lagoon with chlorophyll-a (Chl-a)concentrations over 100 µg/L during most of theyear and over 300 µg/L in spring (Onandia et al.,2015). More detailed information on the historyand limnological characteristics of the lagooncan be found in Vicente & Miracle (1992) andRomo et al. (2008).The study consisted of two sets of exper-

iments aimed to assess the grazing aspects atthe onset of spring and mid-spring, namely onMarch 26, 2011 and May 16, 2012, respectively.Additionally, a number of selected variables weremeasured at the experimental site on these datesas described in Onandia et al. (2014b). Waterconductivity (Cond), pH, salinity and tempera-ture (T) were measured in situ with a multipa-rameter probe (WTW Multi 350i). The Secchidisk depth was recorded, and vertical profilesof photosynthetically active radiation were ob-tained with a flat-faced quantum sensor (LiCorLi-1000) to calculate the light attenuation co-efficient Kd. Water samples for chemical andbiological analyses were taken with a Ruttnerbottle (50 cm length) by placing the top of thebottle at approximately 30-40 cm depth. Total ni-trogen (TN), ammonia, nitrate, nitrite, total phos-phorus (TP) and alkalinity (Alk) were estimatedfollowing the methodology described in APHA(1992). Dissolved inorganic nitrogen (DIN),

calculated as the sum of ammonia, nitrate andnitrite, was subsequently substracted from TN(TN-DIN). Chl-a concentrations were deter-mined spectrophotometrically as described byShoaf & Lium (1976). Samples from lagoonwater and feeding suspensions were preservedin Lugol’s solution until the enumeration ofalgal cells and protists (ciliates and heterotrophicflagellates) in 3 mL sedimentation chamberswith an inverted microscope at 1000×. Giventhe low abundance of heterotrophic flagellates,we opted to enumerate them with an invertedmicroscope in the algal samples because thewater volume explored was notably larger thanthat of the bacterial DAPI-stained samples. Cellbiovolume was calculated as described by Hille-brand et al. (1999), and phytoplankton biovolu-me was transformed into biomass using a con-version factor of 0.225 pg C/µm3 for mixedphytoplankton populations (Reynolds, 2006).Samples of lagoon water and feeding sus-

pensions for the enumeration of bacteria werepreserved in a mixture of paraformaldehyde andglutaraldehyde (100 g/L wv and 0.5 % final con-centrations, respectively). Subsamples of 200/250 µL were filtered on black polycarbonatefilters (Millipore, 0.2 µm) and stained with DAPI(Porter & Feig, 1980). Bacteria were then coun-ted on epifluorescence microscope micropho-tographs (Zeiss III RS). Cell volume was estima-ted based on the dimensions measured in thesemicrophotographs using a general formula (Som-maruga, 1995) and transformed to carbon byapplying a conversion factor of 0.2 pg C/µm3

(Simon & Azam, 1989). Autotrophic picocyano-bacteria were also evaluated in the same micros-cope slides under green light excitation.Zooplankton samples were taken from lagoon

water by filtering two Ruttner bottles (2.7 L ca-pacity each) through a 30 µm nytal mesh andpreserved with formalin 4 %. Zooplankton spec-imens from these samples and from the exper-imental vials were counted and identified usingan inverted microscope. The rotifer biomass wascalculated using species-specific carbon conver-sion factors (Latja & Salonen, 1978; Telesh et al.,1998). Microcrustacean length, once measured,was converted into biomass according to Dumont


544 Onandia et al.

et al. (1975) and Bottrell et al. (1976) and into Cby assuming that the C content was 0.48 of thedry weight (Andersen & Hessen, 1991).

Grazing experiments at the onset of spring

Our experimental design targeted the estimationof clearance rates (CRs), ingestion rates (IRs),mass-specific ingestion rates (MSIRs), assim-ilation rates (ARs) and assimilation efficiency(AE) of natural algae and bacteria by D. magna,the dominant zooplankton species at the end ofMarch 2011, under natural-resembling conditions.Twenty-four hours prior to the grazing experi-

ments, a suspension of radiolabelled phytoplank-ton and bacteria was prepared. Freshly collectedwater from the Albufera de Valencia lagoon wasfiltered through a 25 µm nytal mesh and labelledwith both [methyl-3H]-thymidine (Perkin Elmer;specific activity = 2.59 to 3.33 TBq/mmol; fi-nal concentration = 5 nM) and NaH14CO3 (DHI;specific activity = 39035 kBq/mL; final concen-tration = 4 µCi/mL). The suspension was incu-bated for 24 hours at 15 ◦C and 110 µEinsteinsm−2 s−1 (16 h light:8 h dark cycle). The goal ofthe incubation period was to label bacteria with[methyl-3H]-thymidine and phytoplankton with14C. Nonetheless, a small proportion of both ra-diotracers might have been transferred to othercomponents of the food web (namely to het-erotrophic flagellates and ciliates, as well as toheterotrophic bacteria in the case of 14C). Afterthe incubation period, the suspension was cen-trifuged (12 m at 3500 r.p.m) and aspirated, andthe remaining pellets were rinsed with lagoonwater that had been filtered through glass-fibrefilters (GF/F, Whatman, hereafter referred to aspure lagoon water), centrifuged, aspirated again,and resuspended in pure lagoon water.One day before the grazing experiments,

500 mL of freshly collected water from the la-goon was filtered through a 25 µm sieve and dis-tributed into 10 plastic vessels: 8 replicates and2 controls (containing 50 mL each). Five adultDaphnia magna individuals (collected from thelagoon at the same time) were added to each ves-sel and allowed to acclimate for 24 h underthe same conditions as in the radiolabelled sus-

pension. To estimate the CRs, 5 mL of waterwere extracted from 5 zooplankton vessels andreplaced by 5 mL of radiolabelled solution.Daphnia were allowed to feed for 8 minutes,then anaesthetized with carbonized water andsubsequently fixed with sucrose-formalin (40 %).To determine the ARs, 5 mL of water was ex-

tracted from the 5 remaining zooplankton ves-sels and replaced by 5 mL of radiolabelled foodsolution. Daphnia were allowed to feed for 30minutes, then filtered through a 50 µm sieve andimmediately transferred to a vessel containing50 mL of unlabelled lagoon water that had beenfiltered through a 25 µm sieve. The animals wereallowed to evacuate their guts for a period of 45min, then anaesthetizedwith carbonized water andsubsequently fixed with sucrose-formalin (40 %).In all experiments, after the animals were

fixed, they were collected on a 90 µm sieve andfully rinsed with a 4 % sucrose-water solutionto eliminate any radiolabelled cells potentiallyattached to them. The same procedure was follo-wed for the blanks, although the animals werekilled with sucrose-formalin (40 %) before thestart of the CRs or ARs experiments. The animalscontained in each microcosm were then placed inglass vessels containing a 4 % sucrose-water so-lution and stored at 4 ◦C. Within 1 day, these ani-mals were transferred with a minimum amount ofliquid to corresponding scintillation vials.

Mid-spring grazing experiments

Our experimental design targeted the estimationof the CRs, IRs, MSIRs, ARs and AE of Bosminalongirostris, the dominant zooplankton species inmid-May 2012. Additionally, we aimed to esti-mate the CRs of the 50-100 µm zooplankton sizefraction, together with the CRs, IRs and MSIRsof the entire zooplankton community within thisperiod.Twenty-four hours prior to the grazing experi-

ments, two food suspensions were prepared. Thepreparation of the first suspension aimed to radi-olabel the < 15 µm phytoplankton cells presentin the natural assemblage. To this end, freshlycollected water from the Albufera de Valencialagoon was filtered through a 15 µm nytal mesh



and labelled with NaH14CO3 (same activity asin the early spring experiment). The preparationof the second suspension aimed to radiolabelthe < 3 µm heterotrophic bacteria naturally pre-sent in the lagoon water. To do so, freshlycollected water was filtered through a 3 µm filterand labelled with both [methyl-3H]-thymidine(Perkin Elmer; same activity as in the earlyspring experiment) and L-[3,4,5-3H (N)]-leucine (Perkin Elmer; specific activity = 3.7 to5.56 TBq/mmol; final concentration = 5 nM).Both suspensions were incubated for 20 hoursat 20 ◦C and 110 µEinsteins m−2 s−1 (16 h light:8 h dark cycle). Thereafter, the suspension con-taining the radiolabelled < 15 µm phytoplank-ton cells was filtered through a 3 µm filter(Whatman Nuclepore Track-Etched MembranePC MB, 47 mm) to obtain the 3-15 µm and< 3 µm algal size fractions. Likewise, the sus-pension containing the radiolabelled < 3 µmheterotrophic bacteria was serially fractioned toobtain the 1-3 µm (Whatman Nuclepore Track-Etched Membrane Nuclpo, 47 mm, 1 µm) and0.2-1 µm (Whatman Cyclopore Track-Etchedmembrane Cyclpo PE, 47 mm, 0.2 µm) bacte-rial size fractions. To maximize the similaritybetween the food suspensions and the natu-ral plankton community (< 15 µm), the sizefractions of the labelled algae/bacteria were re-suspended in unlabelled lagoon water containingthe complementary plankton size fractions, i.e.,the 3-15 µm algal fraction was resuspended inunlabelled < 3 µm filtrate.The next step involved the preparation of 7

sets of 4 microcosms (3 replicates plus a control)consisting of 60 mL of freshly collected unfil-tered lagoon water containing the natural zoo-plankton community. To estimate the CRs of thezooplankton community on food suspensions con-taining combinations of different natural plank-ton size fractions, every microcosm set was sub-jected to one of the following treatments, whichinvolved the additionof the following suspensions:

a. 5 mL 3-15 µm 14C + 5 mL< 3 14C

b. 5 mL 3-15 µm 14C + 5 mL 1-3 3H

c. 5 mL 3-15 µm 14C + 5 mL< 1 3H

d. 5 mL< 3 14C + 5 mL 1-3 3H

e. 5 mL< 3 14C + 5 mL< 1 3H

f. 5 mL 1-3 3H

g. 5 mL< 1 3H

A volume equivalent to that to be added to thesuspensions was removed from the microcosmsprior to the suspension addition. The volume wasremoved by aspiration through a 20 µm filter. Alltreatments had the same total concentrations ofphytoplankton and bacteria found in the < 15 µmnatural water filtrate. After the radiolabelled sus-pensions were added, the animals were allowedto feed for 15 minutes, anaesthetized with car-bonized water, filtered through a 50 µm sieveand fully rinsed with water (pH = 7.8) to elimi-nate any radiolabelled cells potentially attachedto them.For the assimilation experiments, two sets of

4 microcosms (3 replicates + 1 control) wereprepared with 60 mL of unfiltered lagoon water.Similar to the CR experiments, a volume equiv-alent to that to be added to the suspensions wasremoved from the microcosms prior to the sus-pension addition by aspiration through a 20 µmfilter. Each set of microcosms was either sub-jected to treatment b or e. After the radiolabelledsuspensions were added, the zooplankton wereallowed to feed for 50 minutes, filtered througha 50 µm sieve, rinsed carefully with water andimmediately transferred to a vessel containing50 mL of unlabelled lagoon water that had beenfiltered through a 15 µm sieve. The animals wereallowed to evacuate their guts for a period of45 min, anaesthetized with carbonized water, fil-tered through a 50 µm sieve and fully rinsed asexplained above.In all experiments, once the animals were fil-

tered and rinsed, they were immediately trans-ferred to a Petri dish containing a 4 % sucrose-water solution, fixed with formalin (4 %) andstored at 4 ◦C. The same procedure was followedfor the blanks, although the animals were killedwith sucrose-formalin (4 %) before the start ofthe CR or AR experiments. Within 1 day, the ani-mals from each microcosmwere separated as fol-


546 Onandia et al.

lows: 20 individuals of B. longirostris were firstremoved from the sucrose-water solution and therest of the animals were counted and then seriallyfiltered, first through a 100 µm and then througha 50 µm nytal sieve. The 20 individuals of B. lon-girostris and the nytal sieves (15 mm diameter)with the collected animals were transferred witha minimum amount of liquid to correspondingseparate scintillation vials.

Radioactivity measurements and calculations

In all experiments (involvingD. magna or B. lon-girostris and others), after the animals were trans-ferred to scintillation vials and the remainingliquid was evaporated, 250 mL of Soluene-350(Perkin Elmer) was added to them. After 24hours, 10 mL of scintillation cocktail (Sigma-Fluor TM High Performance LSC Cocktail, foraqueous samples, S4023) was added. The radio-activity accumulated by the animals and the ra-dioactivity in the grazing suspensions of each mi-crocosmwere measured with a PerkinElmer LSATri-Carb 2810TR liquid scintillation counter.The CRs were calculated as follows:

CR=(dpmindividual−dpmblank)/dpmgrazingsuspension×t,where CR = µL individual−1 h−1 and t = time inhours. The IRs (µm3 ind−1 h−1) were calculatedby multiplying the CR by the corresponding bio-volume of algae or bacteria in the feeding sus-pension. To estimate the IRs on algae, only thebiovolume corresponding to the edible organisms

Table 1. Main limnological characteristics on the experimentaldates (March 2011: 03/26/2011 and May 2012: 05/16/2012). Seetext for abbreviations.Principales características limnológicas enlas fechas experimentales (Marzo 2011: 26/03/2011 yMayo 2012:16/05/2012).Consultar enel texto las abreviaturas.

March 2011 May 2012

T (◦C) 16.2 21.2pH 8.9 9.6Cond (µS) 1426 1656Chl-a (µg/L) 177 301Secchi (m) 0.4 0.35Kd 4.7 7.1TP (µg/L) 174 262TN (mg/L) 4.7 5TN-DIN (mg/L) 2.8 4.3Alk (mg/L) 24.4 15.8

present in the radiolabelled solution was takeninto account; chlorophytes, diatoms and a mi-nor part of cyanophyceae (excluding filaments ofthe genus Planktolyngbya and Pseudanabaena,as well as colonies of the genus Merismopedia,Aphanothece and Aphanocapsa) were considerededible. In the case of the IR of heterotrophicbacteria, all present morphotypes were consid-ered edible. To estimate the MSIRs (pg C pg C−1h−1), the carbon content of ingested biovolumeper hour was calculated and divided by the zoo-plankton carbon content in the corresponding mi-crocosms. Assimilation rates (AR) were calcu-lated as described for the CR, and the AE wassubsequently calculated as, AE = CR/AR× 100.Although the algal and bacterial abundance in

the food solutions was lower than in the lake, theIRs estimated in the microcosms could be ex-trapolated to those expected in the lake becausethe food concentrations in the feeding solutionswere above the incipient level. Thus, we couldestimate the relative grazing rate of lake’s phy-toplankton and bacterioplankton biomass as wellas the relative grazing rate of primary productionand bacterial production. These relative grazingrates were calculated as the percentage of thedaily IRs with respect to the i) total phytoplank-ton and bacterioplankton biomass and ii) dailyprimary (PPday) and bacterial production (PBday),respectively. The values for PPday (g C m−3 d−1)and BPday (mg C m−3 d−1) were obtained in pre-vious studies carried out in the same locality andseasonal period (Onandia et al., 2014a; b).

RESULTS

The physicochemical conditions at the time ofthe zooplankton grazing experiments in the la-goon are shown in Table 1. The hypertrophic stateof the lagoon is illustrated by the high TN, TPand Chl-a concentrations, together with the wa-ter turbidity (Wetzel, 2001). The water temper-ature was 5 ◦C higher in the mid-spring period.The elevated phytoplankton spring growth is in-dicated by the markedly high values of Chl-a inMay, which were nearly double those found inMarch. Correspondingly, the water turbidity (il-



lustrated by Kd and the Secchi depth), TN and TPwere also higher. Similarly, variables associatedwith photosynthetic activity reflected this springupsurge (high pH, low alkalinity).


On 26 March 2011, cladocera represented morethan 50 % of the total zooplankton biomass(Fig. 1). Daphnia magna was the dominant spe-cies (31 %), although Alona rectangula (14 %)and Chydorus sphaericus (6 %) were also wellrepresented. The cyclopoid copepod Acanthocy-clops americanus and rotifera each contributedroughly 25 % of the zooplankton biomass. Themain rotifer species were Brachionus variabilis

(D. magna epibiont) and Keratella cochlearis,with a moderate presence of Polyarthra sp. anda minor presence of Synchaeta sp. and Keratellatropica. However, these proportions differed intermsof individuals (Table 2).The phytoplankton community (biomass 38

mm3/L) was noticeably dominated by diatoms(mostly Nitzschia sp. and Fragilaria sp.), whichcontributed 42 % of the total phytoplankton bio-mass (Fig. 1). The conjugatophycea Cosmariumabbreviatum also represented a significant bio-mass fraction (20 %), slightly higher than thefraction composed by a wide variety of pooledchlorophytes species (e.g., Scenedesmus sp., Des-modesmus sp., Monoraphidium sp., Tetraedronsp.). Cyanophyceae amounted to 11 % of the

Oscillatoriales

Synechococcales

Nostocales

Chrooccoclaes

Bacillariophyceae

Chlorophyta

Cryptophyta

Conjugatophyceae

Chrysophyceae

Other

Cocci

Bacilli

Curved bateria

Filamntous

bacteria

Bacteria < 0.6

μm

March 2011 May 2012

Zooplankton

Phytoplankton

Bacteria

B: 0.32 μgC/mL B: 0.14 μ gC/mL

B: 38 mm3/L : 143 mmB 3/L

B: 10 mm3/L B: 8 mm3/L

Copepoda

Ro�fera

Daphnia magna

Other cladocera

Bosmina longirostris

Filamentous

Colonial

Chroococcaceae

Bacillariophyceae

Chlorophyta

Cryptophyta

Conjugatophyceae

Chrysophyceae

Other

Cocci

Bacilli

Curvedbacteria

Bacteria < 0.6 μm

Filamentous bacteria

cyanobacteria

cyanobacteria

Figure 1. Relative contribution of the different groups of zooplankton, phytoplankton and bacterioplankton to the natural planktonbiomass in the lake on the experimental dates (March 2011: 03/26/2011 and May 2012: 05/16/2012). B: biomass (µg C/mL ormm3/L). Contribución relativa de los diferentes grupos de zooplancton, fitoplancton y bacteria a la biomasa planctónica naturaldel lago en durante la realización de los experimentos en Marzo de 2011 (26/03/2011) y Mayo de 2012 (16/05/2012). B: biomasa(µgC/mL o mm3/L).


548 Onandia et al.

total pool, with the main contributors beingoscillatoriales such as Pseudanabaena galeata.Minor but still well represented groups werecryptophytes, namely Cryptomonas sp., andchrysophytes. In terms of relative abundance,the proportions changed: cyanobacterial fila-ments (mainly Pseudanabaena galeata) andcolonies represented 10 % and 1 %, respectively.Regarding the bacterial community, the mostsignificant biomass contribution was from fil-amentous bacteria (40 %, Fig. 1), although itsrelative abundance with respect to total bacterialindividuals was small (Table 2).In the suspension containing labelled organ-

isms< 25 µm, the proportion of the different phy-toplankton groups was very similar to that ob-served in the natural assemblage. However, to-tal phytoplankton biomass was 17.2 mm3/L, ap-proximately half of that found in the unfilteredlagoon sample. Fragilaria sp. and Nitzschia spp.were the dominant species (42 %) in terms ofbiomass, but Cosmarium abbreviatum also ac-counted for an important fraction (25 %). Chloro-phytes and cyanobacteria contributed 17 % and

12 % to the total biomass, whereas cryptophytesand chrysophytes represented 2.6 % and 0.7 %,respectively. Similarly, bacteria in the labelledsuspension were distributed as described in thewater sample (Fig. 1, Table 2). In terms of in-dividuals, flagellates and ciliates presented verylow relative abundances (lower than that in thewhole water sample, Table 2). Flagellates and cil-iates amounted to approximately 1 % and below0.1 %, respectively, of the abundance of phyto-plankton individuals and were an order of mag-nitude lower than the bacterial numbers. Whencompared to phytoplankton biomass, flagellatesand ciliates contributed at somewhat higher per-centages, but still below 2 % and 1 %, respec-tively; conversely, these percentages were 4 timeshigher than the bacterial biomass.

Daphnia magna CRs were nearly equal forboth groups of radiolabelled organisms. Duringthe period in which the food suspension was in-cubated with 14C, this carbon was incorporatedinto the phytoplankton biomass, but part of thiscarbon was presumably released as dissolved or-ganic matter and subsequently incorporated by

Table 2. Abundance (ind/mL) of the main zooplankton, phytoplankton and bacterial groups.Abundancia (ind/mL) de los principalesgrupos de zooplancton, fitoplancton y bacterias.

Abundance (ind/mL)

March 2011 May 2012

Zooplankton

Daphnia magna + other cladocera 0.04 + 0.02 —Bosmina longirostris + other cladocera — 1.4 + 0.04Copepoda 0.5 0.5

Rotifera 0.9 0.6

Ciliates 300 643Heterotrophic flagellates 3600 19166

Phytoplankton (×103)Filamentous and colonial cyanobacteria (inedible) 69 764

Edible cyanobacteria 7 200

Diatoms 43 107Chlorophytes 31 44

Other 51 6

Heterotrophic bacteria (× 106)Bacteria< 0.6 µm 18.2 17

Bacteria 0.7-3 µm 12.8 8.4

Filamentous bacteria 1.3 1.9



heterotrophic bacteria. However, previous resultsin the lagoon showed that, in spring, less than15 % of the extracellular dissolved organic car-bon released by phytoplankton is consumed byheterotrophic bacterial gross production (see bac-terial carbon demand in Table 2 in Onandia et al.,2014a); therefore, we assumed that the labelledcarbon was mainly incorporated into phytoplank-ton biomass. Likewise, during the incubation pe-riod, a fraction of the radiolabelled compoundswas very likely transferred to different protists.Given the low relative abundance of ciliates andheterotrophic flagellates with respect to phyto-plankton and bacteria, we considered that the la-belled thymidine was largely incorporated intothe heterotrophic bacterial biomass, although la-belled flagellates were very likely present.Taking into account these considerations,

the measure of CR as the suspension volumefiltered by Daphnia can be considered a goodestimate and therefore used to calculate the IRfor the algae and bacteria in the suspension byknowing their corresponding concentrations,even though we had an unevenly labelled poolof organisms. Moreover, because no significantdifferences were found between CRs using thetwo radiotracers, our results indicate that therewere no differences between the CRs of phyto-plankton and heterotrophic bacteria. However,the IRs were significantly higher for algae thanfor bacteria (Fig. 4). Regarding the bacterialcarbon ingested by D. magna, bacteria in thesize range of 1-3 µm amounted to 83 % of thetotal bacterial biomass, and therefore, assuming

even CRs for all bacteria, this size range wouldprobably provide most of the bacterial carboningested by D. magna.The mean MSIRs of D. magna for algae and

bacteria are shown in Table 3. These results in-dicate that bacteria contributed ≈ 40 % of the to-tal (algae + bacteria) ingested carbon. The AEsfor phytoplankton, as well as for bacteria, wereapproximately 50 % and did not differ signifi-cantly (p > 0.05, n = 8, ANOVA test). Accordingto these results, at the end of March, D. magnagrazing removed 5 %/d and 15 %/d of the stand-ing stock of total phytoplankton and bacteria, re-spectively, in the Albufera de Valencia lagoon.The relative grazing rates of PPday and BPday byD. magna were 9 %/d and 75 %/d, respectively.


In mid-spring 2012, the zooplankton communitywas clearly dominated by the cladocera Bosminalongirostris, which amounted to 72 % of thezooplankton biomass (Fig. 1). Other cladoceraspecies such as Chydorus sphaericus, Alona rec-tangula and Ceriodaphnia laticaudata werefound, but they represented a negligible biomassfraction. The rest of the zooplankton biomassconsisted of Acanthocyclops americanus androtifers in similar proportions. The main rotiferspecies were Brachionus calyciflorus and Ker-atella cochlearis, although Keratella tropica,Lecane bulla, Polyarthra sp. and Hexarthra fen-nica were also present in low numbers (Fig. 1).The proportions varied in terms of their abun-

Table 3. Mass-specific ingestion rates (MSIRs, pg C pg C –1 h–1) and assimilation efficiency (AE,%) ofD. magna and B. longirostrisfor different labelled organisms. Tasas de ingestión por unidad de biomasa (MSIRs, pgC pgC–1 h–1) y eficiencia de asimilación (AE,%) de D. magna y B. longirostris para los distintos organismos marcados.

Labelled size in food suspensions (µm)

Algae Bacteria

Daphnia magna < 25 < 3

Mass specific ingestion rate (pgC pgC−1 h−1) 0.054± 0.005 0.036± 0.004Assimilation efficiency (%) 56± 2 51±Bosmina longirostris 1-15 3-15 1-3 1-3 < 1

Mass specific ingestion rate (pgC pgC−1 h−1) 32.5 21.4± 5.9 3± 0.3 4.6± 1.4 0.7± 0.2Assimilation efficiency (%) 44± 20 52± 8 53± 20 73± 11


550 Onandia et al.

dances (Table 2). Cyanobacteria were the mainalgal group, contributing 75 % of total phyto-plankton biomass. We assumed that only a smallproportion of the cyanobacterial biomass (24 %)

was ingested by zooplankton because the specieswith larger biomass contributions were consid-ered inedible, such as the filamentous Pseudana-baena galeata, (37 % abundance, 37 % biomass)

0

10

20

30

40

101

102

103

104

105

106

107

o s

Chrooc

coca

ceae b

Chlorop

hyta

Conjug

atoph

ycea

e

Cryptop

hyta

0

10

20

30

40

101

102

103

104

105

106

107

*Meris

mopedia sp

p. +

Aphanotece s

p. + Aphanocapsa sp

.

228 x 10399 x 103

108 x 103275 x 103 3 - 15 m

< 3 m

Alga

l bio

mas

s (m

m3 /L

)

Alga

l abu

ndan

ce (

cell/

mL)

Fragilaria

sp. +

Nitz

schia sp

. +

Cyclotella sp

.

*P. g

aleata +

G. amphibium +

other

Figure 2. Phytoplankton biomass (bars) and abundance (squares) in the radiolabelled food suspensions containing (a) 3-15 µm and(b) < 3 µm algae used in the mid-spring experiments (May 2012). Bar fractions are ordered corresponding to taxon names on thex-axis. *Filamentous and colonial cyanobacteria were also counted as individuals: numbers under the corresponding squares indicateabundance (ind/mL). Biomasa (barras) y abundancia (cuadrados) de fitoplancton en las suspensiones de alimentación radioactivascon algas de (a) 3-15 µm y (b) < 3 µm en los experimentos de final de primavera (mayo de 2012). Las fracciones de las barras estánordenadas siguiendo los nombres de los taxones del eje horizontal. *Las cianobacterias filamentosas y coloniales fueron tambiéncontados como individuos: la abundancia (ind/mL) se indica bajo los cuadrados correspondientes.



and the colonial Merismopedia sp. (38 % abun-dance, 25 % biomass). Next in importance werebacillariophytes (18 %), mainly represented byspecies of the genus Fragilaria sp. and Nitzschiasp., in addition to Cyclotella sp. and chloro-phytes (6 %), with species such as Oocystis sp.,Scenedesmus sp., Tetraedron sp., or Lagerheimiasp. (Fig. 1). Filamentous bacteria also dominatedthe heterotrophic bacterial community in termsof biomass (50 %), but their relative abundancewas low (7 %). Bacteria of the smallest size(< 0.6 µm) were always more numerous (62 %of the relative abundance), but the other bacteria(mostly 1-3 µm) were also abundant (Fig. 1).The phytoplankton community composi-

tion of the two radiolabelled food suspensions(15-3 µm and < 3 µm seston fractions) resembledthat of the natural phytoplankton assemblage(Figs. 1-2). P. galeata and Merismopedia sp.dominated in terms of biomass, but Fragi-laria sp. also made an important contribution.However, the relative abundance of filamentsor colonies was low with respect to the totalphytoplankton entities. In the 3-15 µm fraction,the phytoplankton abundance and biomass weremarkedly higher than in the < 3 µm fraction.The former phytoplankton biomass amounted to

56.6 mm3/L, and the latter was 9.2 mm3/L. Like-wise, the phytoplankton abundance decreasedfrom 7.3 × 106 cell/mL in the 3-15 µm fraction to1.2 × 106 cell/mL in the < 3 µm fraction (Fig. 2).The bacterial abundance was negligible in the3-15 µm fraction (although a small proportionof filamentous bacteria were present), and therelative proportion of the 1-3 µm and < 1 µmbacterial sizes was very similar to that of the lakewater. The relative contribution of flagellatesand ciliates in these fractions was very low;flagellates amounted to approximately 0.03 % ofthe number of phytoplankton individuals in bothfractions (and less than 2 % of phytoplanktonbiomass), while ciliates amounted to 0.003 % ofthe number of phytoplankton individuals in the3-15 µm (0.3 % of phytoplankton biomass) andto a negligibleproportionboth in numbers and bio-mass in the < 3 µm fraction.

B. longirostris was found to clear all radio-labelled suspensions (Fig. 3). However, the CRsdiffered significantly across the suspensions con-taining organisms with different size ranges. TheCRs on algae in the size range of 1-15 µm weresignificantly higher than the CRs on < 1 µm bac-teria (p < 0.05, n = 31, ANOVA test; Tukey-btest). No significant differences were found be-

Figure 3. Mean clearance rates (CR, µL ind–1 h–1) of D. magna and B. longirostris in March 2011 and May 2012 for the differentlabelled organisms. Error bars indicate the standard deviation. Tasas de filtración medias (CR, µL ind–1 h–1) de D. magna y B.longirostris en marzo de 2011 y mayo de 2012 para los diferentes organismos marcados.


552 Onandia et al.

Figure 4. Mean ingestion rates (IR, µm3 ind–1 h–1) of D. magna and B. longirostris in March 2011 and May 2012 for the differentlabelled organisms. Tasas de ingestión media (IR, µm3 ind–1 h–1) de D. magna y B. longirostris en marzo de 2011 y mayo de 2012para los diferentes organismos marcados.

tween the CRs on the two algal size fractions.Similarly, no significant differences were foundbetween the CRs on the two bacterial size frac-tions, but the minimum CR was found for bac-teria of the 0.2-1 µm size class (Fig. 3). The IRsof B. longirostris are shown in Figure 4. Giventhe important contribution of Fragilaria sp. tothe edible phytoplankton biomass in the radiola-belled suspensions, this species probably propor-tionated the largest amount of carbon ingested byB. longirostris.The mean MSIRs for B. longirostris de-

creased with the size of food particles in thetested food suspensions (Table 3). Bacteria con-tributed ≈ 10 % of the total (algae + bacteria)ingested carbon. B. longirostris displayed an AEof approximately 50 % for the different planktonsize fractions contained in the food suspensions,with the exception of higher values for <1 µmbacteria (Table 3).The CRs of the 50-100 µm zooplankton

size fraction, consisting of Keratella cochlearisand a small proportion of A. americanus nauplii(Fig. 5), significantly differed across the differentfood suspensions. The CRs on algae in the sizerange of 3-15 µm were significantly higher thanthose on algae in the size range of 1-3 µm andon bacteria (p < 0.05, n = 28, Kruskal-Wallistest; pairwise comparison). No signal was found

Figure 5. Clearance rates of 50-100 µm zooplankton (CR,µL ind–1 h–1) for the different labelled organisms in May2012. Zooplankton consisted mainly of Keratella cochlearisand some Acanthocyclops americanus nauplii. The line wi-thin the box represents the median, the black rectangles indicatethe means, and the whiskers indicate the 95 % confidence inter-vals. Tasas de filtración del zooplancton de 50-100 µm (CR, µLind–1 h–1) para los diferentes organismos marcados en mayo de2012. El zooplancton estaba constituido por K. cochlearis y al-gunos naplius de A. americanus. La línea dentro de las cajasindica la mediana, los rectángulos negros la media y las barrasde error los intervalos de confianza del 95 % .

in the separately analysed adult copepods or thelast copepodite stages.The CRs computed globally, pooling together

cladocerans, cyclopoid nauplii and rotifers (main-ly Bosmina, cyclopoid nauplii and Keratella),



Figure 6. Total zooplankton community mass-specific inges-tion rates (MSIR, pg C pg C–1 h–1) in May 2012. The line withinthe box represents the median, the black rectangles indicate themeans, and the whiskers indicate the 95 % confidence intervals.Tasa de ingestión por unidad de biomasa de la comunidad totalde zooplancton en (MSIR, pgC pg C–1 h–1) mayo de 2012. Lalínea dentro de las cajas indica la mediana, los rectángulos ne-gros la media y las barras de error los intervalos de confianzadel 95 %.

were relatively higher for algae than for bacteria.There were significant differences across thetested food suspensions. Namely, CRs weresignificantly higher in the treatment with algaein the size range 1-15 µm than in the other treat-ments. Likewise, CRs on algae in the size rangeof 3-15 µm were significantly higher than theCRs for 1-3 µm algae and for the two bacterialsize ranges (p < 0.05, n = 31, Kruskal-Wallistest; pairwise comparisons). The Cladocera +rotifer + nauplii IR was estimated to be 98 ×103 µm3/h and 37 × 103 µm3/h for edible algaeand bacteria, respectively. The mean estimatesof MSIRs are shown in Figure 6 and amountedto a total of 0.08 pg C pg C−1 h−1 for algaeand approximately 0.006 pg C pg C−1 h−1 forbacteria. These results indicate that bacteriacontributed ≈ 10 % of the total carbon ingested(algae + bacteria). Subsequently, we couldestimate that the total zooplankton communityconsumed 1 %/d and 4 %/d of the total algal andbacterial standing stock in Albufera de Valenciain mid-spring. The relative grazing rates of PPdayand BPday by the total zooplankton communitywere 2 %/d and 0.6 %/d, respectively.

DISCUSSION


D. magna is markedly seasonal in Albufera deValencia, occurring mainly during winter months(Miracle et al., 2002) and peaking at the end ofthis season, namely at the time of winter flush-ing, thus aiding in water clearing. Its occurrenceis usually extended to early spring. Daphnia isknown to be an efficient filter feeder. However,the CRs of D. magna that we measured werequite low (0.3 mL ind−1 h−1). Our rates were atthe bottom of the 0.05 – 2.1 mL ind−1 h−1 rangefound at 15 ◦C in Dutch lakes and were well be-low the CR of ≈ 2 mL ind−1 h−1 on Scenedesmussp. reported for 1.8 mm length individuals of 4Daphnia species (Gulati, 1978; DeMott et al.,2001).These low rates were very likely caused by

the high abundance of filamentous cyanobacteria(30.6 × 103 filaments/mL). Previous experimentsin Albufera de Valencia with seston dilutionsshowed a marked decrease inD. magna CRs withincreasing filament concentrations (Sahuquilloet al., 2007). By using a different methodology(beads), these authors reported a CR ≈ 0.3 mLind−1 h−1 for a filament concentration very closeto that tested in our experiments, thus completelyagreeing with our results. Similarly, a study com-prising different food suspensions with differentproportions of Cylindrospermopsis raciborskiithat resembled eutrophic conditions indicatedthat the presence of C. raciborskii relative to thesuitable food resulted in reduced D. magna CRs(Panosso & Lürling, 2010). The extremely lowCR of ≈ 0.02 mL ind−1 h−1 on filtered sestonfrom hypertrophic Lake Breukeleveen (with ≈2 × 105 filaments/mL) obtained for individuals1.8 mm in length further supports our hypothesis(DeMott et al., 2001). With regard to daphnidsgrazing on bacteria, the CRs of D. pulex (1.8 mmin length) estimated in the laboratory with natu-ral bacterioplankton from Toolik Lake in Alaskawere approximately 0.75 mL ind−1 h−1 (Petersonet al., 1978), a value that is approximately doubleour estimations. However, the CRs on bacteriawere practically equal to those reported for D.


554 Onandia et al.

magna feeding on the yeast Rhodotorula glutinisat 15 ◦C (Burns, 1968). Given that the CRs onalgae and bacteria were simultaneously esti-mated from a common food suspension, the me-chanical interference caused by the high abun-dance of filamentous cyanobacteria very likelyindirectly reduced the ability of D. magna toingest bacteria.The D. magna CRs were nearly identical for

the organisms labelled with the two radiotracers.Despite the fact that the radiolabelling foodsuspension procedure targeted the incorporationof 14C and 3H-thymidine by phytoplankton andbacteria, respectively, a fraction of the radio-tracers might have been channelled to differentplanktonic groups. However, as we indicated inthe results section, because the amount of dis-solved organic carbon excreted by algae was lowwith respect to the total dissolved organic carbonand the relative abundance of flagellates andciliates was low with respect to bacteria in thegrazing suspensions, phytoplankton and bacteriashould be the main labelled groups of organisms.Therefore, while acknowledging a certain rough-ness in our methodology, our results corroboratethe unselective feeding of Daphnia (Lampert,1987). Few data are available on D. magna fee-ding under natural-resembling conditions, but,in accordance with our results, individual D. ga-leata CRs estimated from the natural planktoniccommunity of the eutrophic Bautzen reservoirwere practically equal for phytoplankton and bacte-ria (Kamjunke et al., 1999). Likewise, on a studyinvolving cultures of Chlamydomonas reinhard-tii (5-7 µm) and Aerobacter aerogenes (0.85 µm)as food suspensions, theD.magnaCRswere foundto be similar for both food types (DeMott, 1982).The D. magna IRs were higher on phyto-

plankton than on bacteria as a consequence ofthe higher contribution of phytoplankton to thetotal biomass of the plankton community on theexperimental date. In our study, the largest frac-tion of ingested carbon (measured as MSIRs) wasproportionated by phytoplankton, but the shareof bacteria in the total ingested pool was im-portant (≈ 40 %). A high bacterial contribution(42 %) to the total ingested carbon by D. galeatawas reported during the clear water phase in the

Bautzen reservoir, associated with a low algalbiomass (Kamjunke et al., 1999). In Albufera deValencia, the algal biomass typically decreases atthe end of winter (Onandia et al., 2014a; Romoet al., 2005), which might explain the impor-tant contribution of bacteria to the D. magna car-bon ingestion during this period. Additionally,we would like to highlight that detritus as well asprotozoa probably provided a part of the ingestedcarbon but was not considered in the experiment.According to our experiments, D. magna

assimilated approximately 50 % of the ingestedcarbon when feeding on an algal concentrationof 3.3 mg C/L. Our values are in good agree-ment with those found for D. magna in Dutchlakes with a seston concentration of 3.5 mg C/Land were close to the range of 34-49 % estimatedfor D. magna feeding on algal suspensions (con-taining Chlamydomonas reinhardtii andScenedes-mus obliquus) with a concentration of 3 mg C/L(Gulati, 1978; He & Wang, 2006). There are fewstudies combining assimilation data on algae andbacteria, but it has been reported that zooplank-ton assimilates algae (15-90 %) more efficientlythan bacteria (3-50 %) (Wetzel, 2001). Althoughour values lay within the mentioned range, thefact that the MSIRs as well as the AEs werecomparable for algae and bacteria highlightsthe role of bacteria as an important food sourcefor zooplankton at the onset of spring in thelagoon. Moreover, some algae such as C. abbre-viatum, which contributed a significant fractionof the algal biomass and therefore very likely tothe phytoplankton carbon ingested, might nothave been assimilated because of the mucilagecover (Coesel, 1997). These results might havebroader implications for the annual successionof zooplankton because it has been suggestedthat bacteria could allow the continuance of ahigh Daphnia biomass during periods of foodlimitation (i.e., low algal biomass (Kamjunke etal., 1999).


We found that B. longirostris, the dominant zoo-plankton species in this period and essentiallythe only cladoceran species, fed preferably on



algae (size ranges of 1-3, 3-15 and 1-15 µm) over< 1 µm bacteria, indicating positive selectivefeeding towards larger cells. These rates werelow because the food concentrations were veryhigh, i.e., well above the incipient limiting level.Nonetheless, they lay within the compiled CRrange for B. longirostris from numerous grazingstudies on algae and bacteria reviewed in Tóth& Kato (1997). There are few studies in whichthe CRs were estimated on natural plankton, butthe referred ones are in good agreement withour results. For instance, the B. longirostris CRsestimated in situ in three oligotrophic Ontariolakes indicated that it filteredCosmarium impres-sulum (≈ 28 × 17 × 12 µm) more efficiently thanChlorella vulgaris (3-7 µm). Similarly, CRs onnearly natural bacterioplankton were higher for0.9-8.4 µm bacteria than for ≤ 0.45 µm bacteria(Tóth & Kato, 1997). Moreover, higher Eu-bosmina CRs on natural bacterioplankton wereobtained for attached bacteria (< 3 µm) thanfor free-living bacteria (0.2-1 µm) (Schoenberg,1989). Regarding the results obtained underlaboratory conditions, higher B. longirostrisCRs were found for Chlamydomonas rein-hardtii (5-7 µm) than for Aerobacter aerogenes(0.85 µm) on separate or combined cultures ofboth species (DeMott, 1982).The fact that we did not find significant dif-

ferences between the CRs on 1-3 µm and < 1 µmsized bacteria but the CRs on 1-3 µm algae werehigher than those on <1 µm bacteria is possi-bly due to the higher mean size of algae com-pared to bacteria (1.5 vs 1.2 µm); it also sug-gests that B. longirostris does not discriminatebetween particles below a certain threshold size(i.e., < 1.2 µm).Unfortunately, we cannot infer mechanistic

information on the selective feeding pattern fromour data. Tóth & Kato (1997) attributed the selec-tive feeding to passive selection through the filtermesh size on the fine gnathobasic filter platesrather than to active particle discrimination.However, based on mechanistic observations,DeMott & Kerfoot (1982) ascribed the differ-ential feeding to a bimodal food collection; thefirst two pairs of thoracic limbs can be spreadout to capture particles, and the 3rd to 5th limbs

can be used to filter fine particles. In view of ourresults, we cannot conclude that B. longirostris isa highly effective bacterial grazer. The outcomesof previous studies are controversial. Tóth &Kato (1997) concluded from a literature reviewthat the feeding efficiency of B. longirostrison both natural and cultured bacteria is veryoften low. However, these authors found thatB. longirostris fed efficiently on natural bacte-rioplankton and suggested that this cladocerancould importantly shape the structure and size ofthe bacterial community.The IRs of B. longirostris were markedly

higher on phytoplankton than on bacteria as aconsequence of the higher contribution of phy-toplankton to the total biomass of the planktoncommunity in May 2012. The mean MSIRsfor B. longirostris were proportional to the cellsize in the tested solutions, with 3-15 µm algaerepresenting the main ingested carbon source forthis species.The AEs for 3-15 µm algae were approxi-

mately 50 %, substantially lower than the va-lues of 74-92 % estimated for B. longirostrisfeeding on an algal mixture (containing mostlyScenedesmus and Chlorella) with concentrationsof 0.05-2.50 mg C/L (Urabe, 1991). The highfood concentration of our 3-15 µm feedingsuspension (5.4 mg C/L) might explain thisdiscrepancy because, in agreement with the “su-perfluous feeding” theory, a negative relationshiphas been found between the AE of cladoceransand food concentrations (He & Wang, 2006).The AEs were markedly higher for bacteria sized1-3 µm, but our values lay within the rangesusually reported for zooplankton (see above).The higher AEs for < 1 µm bacteria suggesta higher digestibility of this bacterial group;however, given their small mean mass and thelow CRs on this group, these results should beconsidered with caution; for the same reasons,the contribution of the smallest bacteria to thetotal ingested carbon by B. longirostris can beregarded as of low relevance.The 50-100 µm zooplankton size fraction

consisting mainly of Keratella cochlearis and asmall proportion of nauplii, showed the highestCRs on 3-15 µm algae. Keratella cochlearis is


556 Onandia et al.

able to feed on particles of different sizes (Aer-obacter, Rhodotorula and Chlamydomonas) in-cluding detritus and killed algal cells (Bogdan etal., 1980; Weisse & Frahm, 2002; Starkweather& Bogdan, 1980). However, the IRs of bacteria-sized particles is low (Ooms-Wilms et al. 1995).Moreover, it rarely ingests “flavoured beads” andappears to have significant selection capabilities(DeMott, 1986). Furthermore, when culturedin filtered lake water, it did not reproduce andexhibited poor survival in the < 3 or < 1 µmfiltrates, but it survived and reproduced well inthe < 15 µm filtrates (Ooms-Wilms, 1997). Ourresults confirm K. cochlearis bacterivory butalso show their preference for phytoplankton.Regarding the zooplankton community as

a whole, bacteria contributed 25 % of the totalingested carbon by zooplankton as a conse-quence of its lower contribution to the phyto-bacterioplankton biomass. Furthermore, thebacteria with very high relative abundances arevery small (< 0.6 µm, Table 2) and representonly 20 % of total bacterial biomass (Fig. 1)thus, their contribution to the ingested C wasonly 5 %. These results probably reflect thefeeding preferences of the dominant Bosminalongirostris. Nevertheless, the consumed bacte-ria represented a much greater percentage thanphytoplankton with respect to their respectivebiomass concentration in the water because ofthe dense algal concentrations.

Onset of spring versus mid-spring grazingexperiments

The physicochemical and biological parametersdiffered substantially between the study periods.The composition of the phytoplankton com-munity varied markedly. The variations in thephytoplankton assemblage are closely relatedto the hydrological management of the lagoon.The “flushing” experienced by the lagoon duringwinter favours the decrease in cyanobacterialbiomass and the appearance of diatoms andchlorophytes, which are edible to daphnids.Thereafter, in mid-spring, the strong reductionof the flow through the lagoon promotes theincrease in filamentous and colonial cyanobac-

terial biomass and the reestablishment of thesegroups’ dominance (Onandia et al., 2014a). Thiswas accompanied by a shift from a zooplanktoncommunity dominated by the large cladocera D.magna in early spring to a community dominatedby the small B. longirostris later in the season.Early works on the annual zooplankton suc-

cession in Albufera de Valencia described theoccurrence of D. magna in winter-spring, fol-lowed by the sporadic appearance ofD. pulex anda brief peak of B. longirostris (Blanco, 1974).More recent studies in this lagoon have linkedthe appearance of D. magna to the mentioneddecrease in cyanobacteria in early spring duringthe “flushing” periods (Oltra & Miracle, 1992;Miracle & Sahuquillo, 2002). However, when fil-amentous cyanobacteria bloomed again in mid-spring, D. magna failed to grow, in accordancewith what has been observed in the hypertrophicLoosdrecht Lakes (DeMott et al., 2001). Fila-mentous cyanobacteria have been found to in-hibit the feeding of Daphnia (Sahuquillo et al.,2007; Panosso & Lürling, 2010) and often rep-resent an inadequate food source for several rea-sons. Cyanobacteria might be toxic, nutritionallydeficient or difficult to ingest because of mechan-ical interference (Lampert, 1987; Brett &Müller-Navarra, 1997; DeMott et al., 2001). In Albuferade Valencia, a large part of the D. magna popula-tion at the onset of spring is composed of ephip-pial females that disappear from the plankton,leaving ephippia that sink to the bottom of thelake (Miracle& Sahuquillo, 2002). The inductionof resting eggs could be both due to poor foodconditions and the entrance of juvenile fish earlyin the year in the lagoon (Slusarczyk et al., 2013).In contrast, the spring bloom of filamentous

and colonial cyanobacteria favoured the ap-pearance of B. longirostris. Small cladoceransoften outcompete larger ones under conditionsof cyanobacterial dominance (Lampert, 1982;Abrantes et al., 2009), with Bosmina sp. beingless sensitive than daphnids to cyanobacterialfilaments (Gliwicz & Lampert, 1990). B. lon-girostris is typically considered an indicator ofeutrophy (Alonso, 1996; Haberman & Haldna,2014). Therefore, the observed changes in thezooplankton community may reflect changes in



the amount and quality of phytoplankton. Ourobservations support the idea that cladoceransprovide a useful indicator of the trophic status inshallow lakes (Haberman et al., 2007).Our results revealed different feeding strate-

gies of the dominant cladocerans in the twostudied periods.D. magna showed no differentialCRs on phytoplankton and bacteria. On thecontrary, B. longirostris grazed preferentially onalgae (size range of 1-15 µm) than on bacteria(mainly < 1 µm). These findings agree with theresults from in vitro grazing experiments onnatural radiolabelled plankton in the shallowhypertrophic Lake Søbygaard (Denmark), indi-cating that, whereas B. longirostris fed mainly onphytoplankton; daphnids equally cleared bacteriaand algae (Jeppesen et al., 1996).Our estimations of the relative grazing rates

of bacteria and phytoplankton indicate that onlya minor proportion of the standing stock ofbacteria and especially of phytoplankton wasremoved by zooplankton grazing, which is ingood agreement with the results from othershallow eutrophic lakes (Agasild & Nõges,2005). Furthermore, the greater part of the phy-toplankton production and secondary bacterialproduction was not consumed by grazing. Thelow phytoplankton and bacterial biomass lossesdriven by grazing illustrate the limited abilityof the zooplankton to control the lower trophiclevels in the lagoon and support the idea ofreduced top-down control in warm-temperatelakes (Jeppesen et al., 2007).We found that bacteria supplied a substan-

tial part of the carbon ingested by D. magna(≈ 40 %) at the onset of spring. However, inmid-spring, bacteria only provided a minor frac-tion (≈ 10 %) of the carbon ingested by thezooplankton community. Our results suggest thatduring periods in which algal biomass is reducedand Daphnia dominates, the trophic couplingbetween the microbial loop and the classicalgrazer food web might play a more importantrole in energy transfer in hypertrophic lakes, notonly through the microbial pathway but also viadirect bacterial carbon flow to macrozooplank-ton. However, during peaks in phytoplanktongrowth and cyanobacterial dominance in which

B. longirostris dominates the zooplankton com-munity, the amount of energy transferred viabacterial secondary production is smaller than inthe period of lower phytoplankton density.

ACKNOWLEDGEMENTS

We are grateful to E. Vicente and C. Blanco fortheir help in the experimental activities and to J.M. Soria, O. Kramer andM. L. Sendra for chemi-cal and phytoplankton data. This work was sup-ported by the Spanish Ministry of Economy andCompetitiveness through the project CGL2009–12229 and grant BES–2010–481 032212. J. D.Dias acknowledges the Capes for sandwich doc-torate scholarship (Process no. 9144-11-0).

REFERENCES

ABRANTES, N., A. NOGUEIRA & F. GONCAL-VES. 2009. Short-term dynamics of cladocerans ina eutrophic shallow lake during a shift in the phy-toplankton dominance. Annales De Limnologie-International Journal of Limnology, 45: 237–245.

AGASILD, H. & T. NÕGES. 2005. Cladoceran androtifer grazing on bacteria and phytoplankton intwo shallow eutrophic lakes: in situ measurementwith fluorescent microspheres. Journal of Plank-ton Research, 27(11): 1155–1174.

ALONSO,M. 1996. Fauna Ibérica. Vol. 7. Crustacea.Branquiopoda. Museo Nacional de Ciencias Natu-rales, Consejo Superior de Investigaciones Cientí-ficas. Madrid. Spain.

ANDERSEN, T. & D. O. HESSEN. 1991. Carbon, ni-trogen, and phosphorus content of freshwater zoo-plankton. Limnology and Oceanography, 36(4):807–814.

BLANCO, C. 1974. Estudio de la contaminación dela Albufera de Valencia y de los efectos de dichacontaminación sobre la fauna y flora del lago.Ph.D. Thesis. University of Valencia, Spain.

BOGDAN,K. G., J. J. GILBERT & P. L. STARK-WEATHER. 1980. In situ clearance rates of plank-tonic rotifers. Rotatoria. Developments in Hydro-biology, 1: 73–77.

BOTTRELL, H. C. 1976. A review of some prob-lems in zooplankton production studies. Norwe-gian Journal of Zoology, 24: 419–456.


558 Onandia et al.

BRETT, M. & D. MÜLLER-NAVARRA. 1997. Therole of highly unsaturated fatty acids in aquaticfood web processes. Freshwater Biology, 38(3):483–499.

BURNS, C. W. 1968. The relationship between bodysize of filter-feeding Cladocera and the maximumsize of particle ingested. Limnology and Oceanog-raphy 13(4): 675–678.

COESEL, P. M. 1997. The edibility of Staurastrumchaetoceras and Cosmarium abbreviatum (desmi-diaceae) for Daphnia galeata/hyalina and the roleof desmids in the aquatic food web. Aquatic Ecol-ogy, 31(1): 73–78.

DEMOTT,W. R. 1982. Feeding selectivities and rela-tive ingestion rates of Daphnia and Bosmina. Lim-nology and Oceanography, 27(3): 518–527.

DEMOTT,W. R. 1986. The role of taste in food selec-tion by freshwater zooplankton. Oecologia, 69(3):334–340.

DEMOTT,W. R. & W. C. KERFOOT. 1982. Compe-tition among cladocerans: nature of the interactionbetween Bosmina and Daphnia. Ecology, 63(6):1949–1966.

DEMOTT,W. R., R. D. GULATI & E. VAN DONK.2001. Daphnia food limitation in three hypereu-trophic Dutch lakes: evidence for exclusion of lar-ge-bodied species by interfering filaments of cya-nobacteria. Limnology and Oceanography, 46(8):2054–2060.

DUMONT, H. J., I. VAN DE VELDE & S. DU-MONT. 1975. The dry weight estimate of biomassin a selection of Cladocera, Copepoda and Ro-tifera from the plankton, periphyton and benthosof continental waters. Oecologia, 19(1): 75–97.

FERMANI, P., N. DIOVISALVI, A. TORREMO-RELL, L. LAGOMARSINO, H. E. ZAGARESE& F. UNREIN. 2013. The microbial food webstructure of a hypertrophic warm-temperate shal-low lake, as affected by contrasting zooplanktonassemblages. Hydrobiologia, 714(1): 115–130.

GLIWICZ, Z. M. 1990. Why do cladocerans fail tocontrol algal blooms? Hydrobiologia, 200/201:83–97.

GLIWICZ, Z. M. &W. LAMPERT. 1990. Food thres-holds in Daphnia species in the absence and pres-ence of blue-green filaments. Ecology, 71(2): 691–702.

GULATI, R. 1978. Vertical changes in the filtering,feeding and assimilation rates of dominant zoo-plankters in a stratified lake. Verhandlungen desInternationalen Verein Limnologie, 20: 950–956.

HABERMAN, J., R. LAUGASTE & T. NÕGES.2007. The role of cladocerans reflecting the tro-phic status of two large and shallow Estonian la-kes. Hydrobiologia, 584: 157–166.

HABERMAN, J. & M. HALDNA. 2014. Indices ofzooplankton community as valuable tools in as-sessing the trophic state and water quality of eu-trophic lakes: long-term study of Lake Võrtsjärv.Journal of Limnology, 73(2): 263–273.

HAVENS, K. E., J. R. BEAVER & T. L. EAST. 2007.Plankton biomass partitioning in a eutrophic sub-tropical lake: comparison with results from tem-perate lake ecosystems. Journal of Plankton Re-search, 29(12): 1087–1097.

HE, X. & W. X. WANG. 2006. Releases of ingestedphytoplankton carbon by Daphnia magna. Fresh-water Biology, 51(4): 649–665.

HILLEBRAND, H., C. D. DURSELEN, D. KIRSCH-TEL, U. POLLINGHER & T. ZOHARY. 1999.Biovolume calculation for pelagic and benthic mi-croalgae. Journal of Phycology, 35(2): 403–424.

JEPPESEN, E., M. SØNDERGAARD, J. P. JENSEN,E. MORTENSEN & O. SORTKJÆR. 1996. Fish-induced changes in zooplankton grazing on phyto-plankton and bacterioplankton: a long-term studyin shallow hypertrophic Lake Søbygaard. Journalof Plankton Research, 18(9): 1605–1625.

JEPPESEN, E., M. MEERHOFF, B. JACOBSEN, R.HANSEN, M. SØNDERGAARD, J. JENSEN, T.LAURIDSEN, N. MAZZEO & C.BRANC. 2007.Restoration of shallow lakes by nutrient controland biomanipulation – the successful strategy va-ries with lake size and climate. Hydrobiologia,581: 269–285.

KAMJUNKE, N., A. BENNDORF, C. WILBERT, M.OPITZ, J. KRANICH, M. BOLLENBACH & J.BENNDORF. 1999. Bacteria ingestion by Daph-nia galeata in a biomanipulated reservoir: amechanism stabilizing biomanipulation? Hydrobi-ologia, 403: 109–121.

LAIR, N. 1991. Grazing and assimilation rates of nat-ural populations of planktonic cladocerans in a eu-trophic lake. Hydrobiologia, 215: 51–61.

LAMPERT,W. 1982. Further studies on the inhibitoryeffect of the toxic blue-green Microcystis aerug-inosa on the filtering rate of zooplankton. Archivfür Hydrobiologie, 95: 207–220.

LAMPERT,W. 1987. Laboratory studies on zoo-plankton-cyanobacteria interactions. New ZealandJournal of Marine and Freshwater Research,21(3): 483–490.



LATJA, R. & K. SALONEN. 1978. Carbon analy-sis for the determination of individual biomassesof planktonic animals. Verhandlungen des Interna-tionalen Verein Limnologie, 20: 2556–2560.

MIRACLE, M. R. &M. SAHUQUILLO. 2002. Chan-ges of life-history traits and size inDaphnia magnaduring a clear-water phase in a hypertrophiclagoon (Albufera of Valencia, Spain). Verhandlun-gen des Internationalen Verein Limnologie, 28:1203–1208.

MIRACLE, M. R., E. VICENTE, S. SARMA & S.NANDINI. 2014. Planktonic rotifer feeding in hy-pertrophic conditions. International Review of Hy-drobiology, 99: 141–150.

OLTRA, R. & M. MIRACLE. 1992. Seasonal suc-cession of zooplankton populations in the hyper-trophic lagoon Albufera of Valencia (Spain).Archiv Fur Hydrobiologie, 124(2): 187–204.

ONANDIA, G., M. R. MIRACLE & E. VICENTE.2014a. Primary production under hypertrophicconditions and its relationship with bacterialproduction. Aquatic Ecology, 48(4): 447–463.

ONANDIA, G., M. ROSA MIRACLE, C. BLASCO& E. VICENTE. 2014b. Diel and seasonal vari-ations in bacterioplankton production in a hyper-trophic shallow lagoon. Aquatic Microbial Ecol-ogy, 72(3): 255–267.

ONANDIA, G., A. GUDIMOV,MR. MIRACLE&G.ARHONDITSIS. 2015. Towards the developmentof a biogeochemical model for addressing the eu-trophication problems in the shallow hypertrophiclagoon of Albufera de Valencia, Spain. EcologicalInformatics, 26: 70–89.

OOMS-WILMS, A. L. 1997. Are bacteria an impor-tant food source for rotifers in eutrophic lakes?Journal of Plankton Research, 19(8): 1125–1141.

OOMS-WILMS, A., G. POSTEMA & R. GULATI.1995. Evaluation of bacterivory of Rotifera basedon measurements of in situ ingestion of fluorescentparticles, including some comparisons with Clado-cera. Journal of Plankton Research, 17(5): 1057–1077.

PANOSSO, R. &M. LÜRLING. 2010.Daphnia mag-na feeding on Cylindrospermopsis raciborskii: therole of food composition, filament length and bodysize. Journal of Plankton Research, 32(10): 1393–1404.

PETERSON, B. J., J. E. HOBBIE & J. F. HANEY.1978. Daphnia grazing on natural bacteria. Lim-nology and Oceanography, 23(5): 1039–1044.

PORTER, K. G. & Y. S. FEIG. 1980. The use of Dapifor identifying and counting aquatic microflora.Limnology and Oceanography, 25(5): 943–948.

REYNOLDS, C. S. 2006. Ecology of phytoplankton.Cambridge University Press Cambridge. NY.USA.

ROMO, S., M. J. VILLENA, M. SAHUQUILLO, J.M SORIA, M. GIMÉNEZ, T. ALFONSO, E. VI-CENTE & M. R. MIRACLE. 2005. Response of ashallow Mediterranean lake to nutrient diversion:does it follow similar patterns as in northern shal-low lakes? Freshwater Biology, 50: 1706–1717.

ROMO, S., A. GARCÍA-MURCIA, M. J. VILLENA,V. SÁNCHEZ & A. BALLESTER. 2008. Tenden-cias del fitoplancton en el lago de la Albufera deValencia e implicaciones para su ecología, gestióny recuperación. Limnetica, 27: 11–28.

SAHUQUILLO,M., M. MELÃO & M. MIRACLE.2007. Low filtering rates of Daphnia magna in ahypertrophic lake: laboratory and in situ experi-ments using synthetic microspheres. Hydrobiolo-gia, 594: 141–152.

SCHOENBERG, S. A. 1989. Effects of algal concen-tration, bacterial size and water chemistry on theingestion of natural bacteria by cladocerans. Jour-nal of Plankton Research, 11(6): 1273–1295.

SHOAF, T. W., & B. W. LIUM. 1976. Improved ex-traction of chlorophyll a and b from algae usingdimethylsulphoxide. Limnology and Oceanogra-phy, 21: 926–928.

SIMON,M. & F. AZAM. 1989. Protein content andprotein synthesis rates of planktonic marine bacte-ria.Marine Ecology Progress Series, 51: 201–213.

SLUSARCZYK,M., A. OCHOCKA & P. BIECEK.2013. Prevalence of kairomonone-induced dia-pause in Daphnia magna from habitats with andwithout fish. Hydrobiologia, 715: 225–232.

SOMMARUGA,R. 1995. Microbial and classicalfood webs: A visit to a hypertrophic lake. FemsMicrobiology Ecology, 17: 257–270.

STARKWEATHER, P. L. & K. G. BOGDAN. 1980.Detrital feeding in natural zooplankton communi-ties: discrimination between live and dead algalfoods. Hydrobiologia, 73(1-3): 83–85.

TELESH, I. V., M. RAHKOLA & M. VILJANEN.1998. Carbon content of some freshwater rotifers.Hydrobiologia, 387: 355–360.

TÓTH, L. G. & K. KATO. 1997. Size-selective gra-zing of bacteria by Bosmina longirostris – an ima-ge-analysis study. Journal of Plankton Research,19(10): 1477–1493.


560 Onandia et al.

VICENTE, E. & M. MIRACLE. 1992. The coastallagoon Albufera de Valencia: An ecosystem understress. Limnetica, 8: 87–100.

WEISSE, T. & A. FRAHM. 2002. Direct and indirectimpact of two common rotifer species (Keratellaspp.) on two abundant ciliate species (Urotrichafurcata, Balanion planctonicum). Freshwater Bio-

logy, 47(1): 53–64.WETZEL, R. 2001. Limnology: Lake and River

Ecosystems. Springer. C. USA.WORK, K. A. & K. E. HAVENS. 2003. Zooplanktongrazing on bacteria and cyanobacteria in a eutro-phic lake. Journal of Plankton Research, 25(10):1301–1306.


zooplankton grazing on natural algae and bacteria under … · 2015-12-13 · romo et al. (2008)....

Documents